Cell Signalling Biology
| Full text PDF Full text PDF
Please cite as Berridge, M.J. (2014) Cell Signalling Biology; doi:10.1042/csb0001002
Home
PDF
Animations
InsP3 and DAG formation
NF-κB activation
Contents
Index
Rights and permissions
Terms and conditions of usage
Send us your comments
FAQs
Cell Signalling Pathways
Michael Berridge
Synopsis
Synopsis

Cells use a large number of clearly defined signalling pathways to regulate their activity. In this module, attention is focused on the On mechanisms responsible for transmitting information into the cell. These signalling pathways fall into two main groups depending on how they are activated. Most of them are activated by external stimuli and function to transfer information from the cell surface to internal effector systems. However, some of the signalling systems respond to information generated from within the cell, usually in the form of metabolic messengers. For all of these signalling pathways, information is conveyed either through protein–protein interactions or it is transmitted by diffusible elements usually referred to as second messengers. Cells often employ a number of these signalling pathways, and cross-talk between them is an important feature. In this section, attention is focused on the properties of the major intracellular signalling pathways operating in cells to regulate their cellular activity.

During the processes of development, specific cell types select out those signalling systems that are suitable to control their particular functions. One of the aims of this website is to understand how these unique cell-specific signalsomes function to regulate different mammalian cell types.


Intracellular signalling pathways

There are a large number of intracellular signalling pathways responsible for transmitting information within the cell. They fall into two main categories. The majority respond to external stimuli arriving at the cell surface, usually in the form of a chemical signal (neurotransmitter, hormone or growth factor), which is received by receptors at the cell periphery that function as molecular antennae embedded in the plasma membrane. These receptors then transfer information across the membrane using a variety of transducers and amplifiers that engage a diverse repertoire of intracellular signalling pathways (Pathways 1–16 in /1Module 2: Figure cell signalling pathways). The /1phosphoinositide signalling and /1Ca2+ signalling systems (Pathways 2–6) have been grouped together because they contain a number of related signalling pathways that often interact with each other. The other categories are the pathways that are activated by signals generated from within the cell (Pathways 17 and 18). There are a number of /1metabolic messengers that act from within the cell to initiate a variety of signalling pathways. /1GTP-binding proteins often play a central role in the transduction process responsible for initiating many of these signalling pathways.

All of these signalling pathways generate an internal messenger that is responsible for relaying information to the sensors that then engage the effectors that activate cellular responses. The main features of the signalling pathways summarized in Module 2: Figure cell signalling pathways are outlined below:

  1. Cyclic AMP signalling pathway. One of the first signalling systems to be characterized was the cyclic AMP signalling pathway, which led to the second messenger concept that applies to many other signalling systems. The idea is that the external stimulus arriving at the cell surface is the first messenger, which is then transformed at the cell surface by adenylyl cyclase (AC) into a second messenger, cyclic AMP, which is part of the signalling cascade that then activates downstream effectors.
  2. Cyclic ADP-ribose (cADPR) signalling and nicotinic acid–adenine dinucleotide phosphate (NAADP) signalling systems function in Ca2+ signalling. An ADP–ribosyl cyclase (ADP-RC) is responsible for generating these two second messengers.
  3. Voltage-operated channels (VOCs) contribute to Ca2+ signals by controlling the entry of external Ca2+ in excitable cells.
  4. Receptor-operated channels (ROCs) contribute to Ca2+ signals by controlling Ca2+ entry in both excitable and non-excitable cells.
  5. Stimuli that activate phospholipase C (PLC) to hydrolyse PtdIns4,5P2 (also known as PIP2) generate a number of signalling pathways:
  6. PtdIns 3-kinase signalling is activated by stimuli that stimulate PtdIns 3-kinase to phosphorylate PIP2 to form the lipid second messenger PtdIns3,4,5P3 (PIP3).
  7. Nitric oxide (NO)/cyclic GMP signalling pathway. Nitric oxide (NO) synthase (NOS) generates the gas NO that acts both through cyclic GMP and nitrosylation reactions. NO has a particularly important role in modulating the activity of other pathways such as Ca2+ signalling.
  8. Redox signalling. Many receptors act through NADPH oxidase (NOX) to form reactive oxygen species, such as the superoxide radical (O2−•) and hydrogen peroxide (H2O2), which act to regulate the activity of specific signalling proteins such as tyrosine phosphatases, transcription factors and ion channels. The O2−• participates in the nitrosylation reaction in Pathway 7.
  9. Mitogen-activated protein kinase (MAPK) signalling. This is a classical example of a protein phosphorylation cascade that often begins with Ras and consists of a number of parallel pathways that function to control many cellular processes and particularly those related to cell proliferation, cell stress and apoptosis.
  10. Nuclear factor κB (NF-κB) signalling pathway. This signalling system has a multitude of functions. It is particularly important in initiating inflammatory responses in macrophages and neutrophils as part of an innate immune response to invading pathogens.
  11. Phospholipase D (PLD) signalling pathway. This is a lipid-based signalling system that depends upon the hydrolysis of phosphatidylcholine by phospholipase D (PLD) to give phosphatidic acid (PA), which functions as a second messenger to control a variety of cellular processes.
  12. Sphingomyelin signalling pathway. Certain growth factors and cytokines hydrolyse sphingomyelin to generate two second messengers that appear to have opposing effects in the cell. Ceramide seems to promote apoptosis, whereas sphingosine 1-phosphate (S1P) stimulates cell proliferation. S1P may also release Ca2+ from internal stores. The action of S1P is complicated in that it is released from the cell, where it can act as a hormone to stimulate external receptors.
  13. Janus kinase (JAK)/signal transducer and activator of transcription (STAT) signalling pathway. This is a fast-track signal transduction pathway for transferring information from cell-surface receptors into the nucleus. The Janus kinases (JAKs) are tyrosine kinases that phosphorylate the signal transducers and activators of transcription (STATs), which carry the information into the nucleus.
  14. Smad signalling pathway. This pathway mediates the action of the transforming growth factor β (TGF-β) superfamily, which controls transcription through the Smad transcription factors.
  15. Wnt signalling pathways. These pathways play an important role in both development and cell proliferation.
  16. Hedgehog signalling pathway. This pathway resembles the Wnt signalling pathway in that it also functions to regulate development and cell proliferation. The ligand Hedgehog (Hh) acts through the transcription factor GLI.
  17. Hippo signalling pathway. This pathway has a core protein kinase cascade that has some similarities to the MAP kinase cascade in that a series of serine/threonine protein kinases function to regulate the transcription of a number of genes that function in cell growth, proliferation and apoptosis.
  18. Notch signalling pathway. This is a highly conserved signalling system that functions in developmental processes related to cell-fate determination particularly in stem cells. The notch receptor generates the transcription factor NICD (Notch intracellular domain).
  19. Endoplasmic reticulum (ER) stress signalling. The endoplasmic reticulum (ER) stress signalling pathway concerns the mechanisms used by the ER to transmit information to the nucleus about the state of protein processing within the lumen of the ER.
  20. Metabolic messengers. There are a number of metabolites that can function as messengers to regulate a number of different processes. The AMP signalling pathway is regulated by adenosine monophosphate (AMP), which functions as a metabolic messenger to activate an important pathway for the control of cell proliferation.

Not included in /1Module 2: Figure cell signalling pathways are some additional signalling pathways that are related to cellular metabolism. The /1sterol sensing and cholesterol biosynthesis controls the level of cholesterol in cell membranes. Another example is found in the /1NAD signalling pathways, where NAD+ functions to regulate a number of cellular processes, including energy metabolism, gene transcription, DNA repair and perhaps /1ageing as well. The gastrotransmitter /1hydrogen sulphide (H2S), which is released from cysteine, acts to regulate multiple cellular processes (/1Module 2: Figure H2S generation).

These cassettes then engage a variety of effectors that are responsible for activating cellular responses. All of these mechanisms (/1Module 1: Figure signal transmission mechanisms) depend upon /1information transfer mechanisms whereby information is transferred along an orderly sequence of events to activate the internal effectors responsible for inducing a great variety of cellular responses.

Cyclic AMP signalling pathway

Cyclic AMP is a ubiquitous second messenger that regulates a multitude of cellular responses. Cyclic AMP formation usually depends upon the activation of G-protein-coupled receptors (GPCRs) that use heterotrimeric G proteins to activate the amplifier adenylyl cyclase (AC), which is a large family of isoforms that differ considerably in both their cellular distribution and the way they are activated. There are a number of cyclic AMP signalling effectors such as protein kinase A (PKA), the exchange proteins activated by cyclic AMP (EPACs) that activate the small GTP-binding protein Rap1 and the cyclic nucleotide-gated channels (CNGCs). These various effectors are then responsible for carrying out the cyclic AMP signalling functions that include control of metabolism, gene transcription and ion channel activity. In many cases, these functions are modulatory in that cyclic AMP often acts to adjust the activity of other signalling pathways and thus has a central role to play in the cross-talk between signalling pathways. This modulatory function is particularly evident in the case of Ca2+ signalling in both neuronal and muscle cells. Many of the functions of cyclic AMP depend upon the precise location of PKA relative to both its upstream and downstream effectors. A family of A-kinase-anchoring proteins (AKAPs) determines this cellular localization of PKA as well as a number of other signalling components. The OFF reactions responsible for removing cyclic AMP are carried out by either cyclic AMP hydrolysis or by cyclic AMP efflux from the cell.

Cyclic AMP formation

The formation of cyclic AMP can be activated by a very large number of cell stimuli, mainly neurotransmitters and hormones (Module 1: Figure stimuli for cyclic AMP signalling). All these stimuli are detected by G-protein-coupled receptors (GPCRs) that use heterotrimeric G proteins, which are the transducers that are responsible for either activating or inhibiting the enzyme adenylyl cyclase (AC) (Module 2: Figure heterotrimeric G protein signalling). In the case of AC stimulation, the external stimulus binds to the GPCR that functions as a guanine nucleotide exchange factor (GEF) to replace GDP with GTP, which dissociates the heterotrimeric complex into their Gβγ and Gα subunits (Module 2: Figure cyclic AMP signalling). The GαS·GTP complex activates AC, whereas Gαi·GTP inhibits AC. The Gα subunits have GTPase activity that hydrolyses GTP to GDP, thus terminating their effects on AC. The endogenous GTPase of GαS·GTP is inhibited by cholera toxin and this causes the persistent activation of the intestinal fluid secretion that results in the symptoms of cholera.

Adenylyl cyclase (AC)

The adenylyl cyclase (AC) family is composed of ten isoforms: nine of them are membrane-bound (AC1–AC9), while one of them is soluble (AC10) (/1Module 2: Table adenylyl cyclases). The domain structure of AC1–AC9 is characterized by having two regions where there are six transmembrane regions (/1Module 2: Figure adenylyl cyclase structure). The large cytoplasmic domains C1 and C2, which contain the catalytic region, form a heterodimer and co-operate with each other to convert ATP into cyclic AMP. Some of the isoforms are /1Ca2+-sensitive adenylyl cyclases.

Ca2+-sensitive adenylyl cyclases

Four of the AC isoforms are sensitive to Ca2+. AC1 and AC8 are stimulated by Ca2+ acting through calmodulin (CaM), while AC5 and AC6 are inhibited by Ca2+ through a direct action. Entry of Ca2+ through store-operated channels (SOCs) seems to be particularly important for activating AC1 and AC8. This tight coupling occurs at lipid rafts where there is a direct coupling between AC8 and the Orai1 entry channel. This Ca2+-sensitive AC8 is thus positioned close to the microdomain of Ca2+ that forms as Ca2+ enters the cell through the Orai1 channel.

Cyclic AMP signalling effectors

Cyclic AMP is a highly versatile intracellular messenger capable of activating a number of different effectors (Module 2: Figure cyclic AMP signalling). An example of such effectors is the exchange proteins activated by cyclic AMP (EPACs), which act to stimulate Rap. Another group of effectors are the cyclic nucleotide-gated channels (CNGCs) (Module 3: Figure Ca2+ entry mechanisms) that play a particularly important role in the sensory systems responsible for smell and taste. Most of the actions of cyclic AMP are carried out by protein kinase A (PKA).


Module 2: Table adenylyl cyclases Regulatory properties and distribution of adenylyl cyclase.

Adenylyl cyclase(AC) isoform Gβγ i or Gαo Modulation by Ca2+, calmodulin (CaM), Ca2+/CaM kinase II (CaMKII), protein kinase C (PKC), protein kinase A (PKA) Tissue distribution
AC1 ↓ Gαo ↑ CaM and PKC Brain, adrenal medulla
↓ CaMKII
AC2 ↑ PKC Brain, skeletal and cardiac muscle, lung
AC3 ↑ CaM and PKC Brain, olfactory epithelium
↓ CaMKII
AC4 ↑ PKC Brain, heart, kidney, liver
AC5 ↓ Ca2+ and PKCα Brain, heart, kidney, liver, lung, adrenal
↓ PKA
AC6 ↑ PKC Ubiquitous
↓ Ca2+ and PKA
AC7 ↑ PKC Ubiquitous, high in brain
AC8 ↑ CaM Brain, lung, heart, adrenal
AC9 Brain, skeletal muscle
AC10 Activated by HCO3 Testis


The membrane-bound adenylyl cyclases (AC1–AC9) are widely distributed. They are particularly rich in brain, but are also expressed in many other cell types. The soluble AC10 is restricted to the testis. The primary regulation of the AC1–AC9 isoforms is exerted through components of the heterotrimeric G proteins, which are dissociated upon activation of neurotransmitter and hormonal receptors into their α and βγ subunits. They are all activated by Gαs, but only some of the isoforms are inhibited by Gαi. The Gβγ subunit is able to activate some isoforms, but inhibits others. The isoforms also differ in the way they are modulated by components of other signalling pathways such as Ca2+ and PKC. Some of the others are inhibited by PKA, which thus sets up a negative-feedback loop whereby cyclic AMP can inhibit its own production. Modified from Table I in Whorton and Sunahara 2003. Reproduced from Handbook of Cell Signaling, Volume 2 (edited by R.A. Bradshaw and E.A. Dennis), Whorton, M.R. and Sunahara, R.K., Adenylyl cyclases, pp. 419–426. Copyright (2003), with permission from Elsevier.



Protein kinase A (PKA)

Many of the actions of cyclic AMP are carried out by protein kinase A (PKA), which phosphorylates specific sites on downstream effector processes (Module 2: Figure cyclic AMP signalling). PKA is composed of two regulatory (R) subunits and two catalytic (C) subunits. The way in which cyclic AMP activates PKA is to bind to the R subunits, which then enables the C subunits to phosphorylate a wide range of different substrates [Module 2: Figure protein kinase A (PKA)]. Of the two types of PKA, protein kinase A (PKA) I is found mainly free in the cytoplasm and has a high affinity for cyclic AMP, whereas protein kinase A (PKA) II has a much more precise location by being coupled to the A-kinase-anchoring proteins (AKAPs). The AKAPs are examples of the scaffolding proteins that function in the spatial organization of signalling pathways by bringing PKA into contact with its many substrates. The scaffolding function of the AKAPs is carried out by various domains such as the conserved PKA-anchoring domain [yellow region in Module 2: Figure protein kinase A (PKA)], which is a hydrophobic surface that binds to an extended hydrophobic surface on the N-terminal dimerization region of the R subunits. At the other end of the molecule, there are unique targeting domains (blue) that determine the way AKAPs identify and bind specific cellular targets in discrete regions of the cell.

Protein kinase A (PKA) I

Type I protein kinase A (PKA) associates with the RI isoforms. As for all isoforms, the R subunits form dimers through their N-terminal dimerization/docking domains [yellow bar in Module 2: Figure protein kinase A (PKA)]. In addition to holding two R subunits together, this N-terminal region is also responsible for docking to the A-kinase-anchoring proteins (AKAPs), as occurs for PKA II. However, the RI isoforms have a very low affinity for the AKAPs and are thus mainly soluble. Cyclic AMP acts by binding to the tandem cyclic AMP-binding domains to release active C subunits that then phosphorylate specific substrates. Since the RI subunits have a higher cyclic AMP-binding affinity, PKA I will be able to respond to the lower cyclic AMP concentrations found globally within the bulk cytoplasm.

Protein kinase A (PKA) II

A characteristic feature of Type II protein kinase A (PKA) is that the regulatory dimer is made up of RII subunits. Since this RII subunit has a much higher affinity for the A-kinase-anchoring proteins (AKAPs), PKA II is usually docked to this scaffolding protein and thus has a much more precise localization to specific cellular targets.

The substrates phosphorylated by cyclic AMP (Module 2: Figure cyclic AMP signalling) fall into two main groups: the cyclic AMP substrates that regulate specific cellular processes and the cyclic AMP substrates that are components of other signalling systems.

Cyclic AMP substrates that regulate specific cellular processes:

Cyclic AMP substrates that are components of other signalling systems:

Salt-inducible kinase 2 (SIK2)

The salt-inducible kinase 2 (SIK2) functions to phosphorylate the transducer of regulated CREB (TORC2), thereby preventing it from entering the nucleus to facilitate the activity of the transcriptional factor cyclic AMP response element-binding protein (CREB) (Module 4: Figure CREB activation). It functions to regulate TORC2 in liver cells (Module 7: Figure liver cell signalling) and in insulin-secreting β-cells (Module 7: Figure β-cell signalling).

Exchange proteins activated by cyclic AMP (EPACs)

Other targets for the second messenger cyclic AMP are the exchange proteins activated by cAMP (EPACs). One of the functions of EPACs is to activate Rap1 and Rap2B, which have many functions, many of which are related to controlling actin dynamics. In addition, the EPAC/Rap pathway can activate phospholipase Cε (PLCε) and this mechanism has been implicated in the control of autophagy (Module 11: Figure autophagy signalling mechanisms).

Cyclic AMP signalling functions

The cyclic AMP signalling pathway functions in the control of a wide range of cellular processes:

Cyclic AMP hydrolysis

There are two OFF reactions of the cyclic AMP signalling pathway, cyclic AMP efflux from the cell and cyclic AMP hydrolysis. The latter is carried out by a family of phosphodiesterase enzymes that hydrolyse cyclic AMP to AMP (Module 2: Figure cyclic AMP signalling).

Cyclic AMP efflux

There are two OFF reactions for the cyclic AMP signalling pathway, cyclic AMP hydrolysis and cyclic AMP efflux from the cell (Module 2: Figure cyclic AMP signalling). The latter is carried out by ABCC4, which is one of the ATP-binding cassette (ABC) transporters (Module 3: Table ABC transporters).


Module 2: Table heterotrimeric G proteins The heterotrimeric proteins are assembled from subunits taken from the G protein α (Gα), G protein β (Gβ) and G protein γ (Gγ) families.

Heterotrimeric G protein Function
G protein α (Gα) subunits
 Gαs Stimulate adenylyl cyclase
 Gαolf Stimulate adenylyl cyclase
 Gαi1 Inhibit adenylyl cyclase
 Gαi2 Inhibit adenylyl cyclase
 Gαi3 Inhibit adenylyl cyclase
 Gαo1 Inhibit adenylyl cyclase
 Gαo2 Inhibit adenylyl cyclase
 Gαt1 Stimulate cyclic GMP phosphodiesterase in rod photoreceptors
 Gαt2 Stimulate cyclic GMP phosphodiesterase in rod photoreceptors
 Gz Close K+ channels. Inhibits exocytosis (see Module 10: Figure lactotroph regulation)
 Gαgust Stimulate phospholipase Cβ (PLCβ) (see Module 10: Figure taste receptor cells and Module 7: Figure L cells)
 Gαq Stimulate phospholipase Cβ (PLCβ)
 Gα11 Stimulate phospholipase Cβ (PLCβ)
 Gα14 Stimulate phospholipase Cβ (PLCβ)
 Gα15 Stimulate phospholipase Cβ (PLCβ)
 Gα16 Stimulate phospholipase Cβ (PLCβ)
 Gα12 Stimulate RhoGEFs to activate Rho (Module 2: Figure Rho signalling)
 Gα13 Stimulate RhoGEFs to activate Rho (Module 2: Figure Rho signalling)
G protein β (Gβ) subunits; β1–β5 These β subunits combine with γ subunits to form βγ dimers that have a number of control functions (for details see Module 2: Figure heterotrimeric G protein signalling)
G protein γ (Gγ) subunits; γ1–γ11 These γ subunits combine with β subunits to form βγ dimers that have a number of control functions (for details see Module 2: Figure heterotrimeric G protein signalling)



GTP-binding proteins

There are a large number of GTP-binding proteins, which usually are referred to as G proteins, which play a central role in cell signalling as molecular switches (Module 2: Figure G protein binary switching). These G proteins are also GTPases, and the switching is driven by both the binding of GTP and its hydrolysis to GDP. The G protein is inactive when bound to GDP, but when the GDP is exchanged for GTP, the G protein/GTP complex is active and transfers information down the signalling pathway until the endogenous GTPase activity hydrolyses GTP to GDP. The G proteins belong to two groups, the heterotrimeric G proteins and the monomeric G proteins, which have separate, but overlapping, functions.

Guanine nucleotide-exchange factors (GEFs)

The guanine nucleotide-exchange factors (GEFs) are essential signal transducing components in that they receive information from a variety of signalling pathways in order to activate the GTP-binding proteins (Module 2: Figure G protein binary switching). There are many different GEFs (Module 2: Table monomeric G protein toolkit).

Trio

Trio is a typical guanine nucleotide-exchange factor (GEF) (Module 2: Table monomeric G protein toolkit) that functions in cell proliferation where it couples G protein-coupled receptors (GPCRs) to components of the MAPK signalling pathway (Module 9: Figure growth factor signalling).

Heterotrimeric G proteins

The heterotrimeric G proteins function as transducers for the G-protein-coupled receptors (GPCRs) that activate a number of cell signalling pathways (Module 1: Table G-protein coupled receptors). These G proteins are made up from 16 Gα, five Gβ and 11 Gγ genes (Module 2: Table heterotrimeric G proteins). These different subunits are characterized by having various lipid modifications that serve to insert them into the plasma membrane, where they are positioned to detect information coming in from the GPCRs and to relay it to various amplifiers. The Gα subunits are either palmitoylated or myristoylated near the N-terminus, whereas the Gβγ subunits are prenylated. The heterotrimeric G proteins are extremely versatile signalling elements, and both the Gα subunit and the Gβγ subunit are able to relay information to downstream components (Module 2: Figure heterotrimeric G protein signalling).

These heterotrimeric G proteins exist in two states. When the Gα subunit is bound to GDP, it forms a complex with Gβγ subunits to form the inactive heterotrimeric complex. When the GPCRs, which are sensitive to a wide range of stimuli (Module 1: Figure stimuli for cyclic AMP signalling), are activated, they function as a guanine nucleotide exchange factor (GEF) to replace the GDP with GTP. The binding of GTP not only activates the Gα subunit, but also liberates the Gβγ subunit, both of which can activate a range of signalling systems. This signal transduction process is switched off when the endogenous GTPase activity of the Gα subunit hydrolyses GTP to GDP and the Gα/GDP complex recombines with the Gβγ subunit to reform the inactive heterotrimeric complex. The normal rate of GTPase activity is very low (four to eight conversions per s), which means that the two subunits have a long time to find their targets. However, there are two mechanisms for speeding up the GTPase activity. Firstly, some of the targets can act to accelerate the GTPase activity. Secondly, a family of regulators of G protein signalling (RGS) proteins function as GTPase-activating proteins (GAPs) that accelerate the Gα subunit GTPase activity more than 1000-fold.

There are G-protein receptor kinases (GRKs) such as β-adrenergic receptor kinase 1 (βARK1), which phosphorylate active receptors to provide binding sites for arrestin that prevent the heterotrimeric proteins from binding the receptor and this leads to receptor desensitization (Module 1: Figure homologous desensitization).

The active Gα/GTP and Gβγ subunits are able to relay information to a large number of signalling pathways, which are described in more detail for the different signalling pathways:

Mutations in CNAQ and GNA11, which encode Gαq and the closely related Gα11 respectively, have been identified in most cases of uveal melanoma,

Regulators of G protein signalling (RGS)

The regulators of G protein signalling (RGS) are a large family of approximately 30 proteins that function as GTPase-activating proteins (GAPs) for the heterotrimeric G proteins (Module 2: Figure heterotrimeric G protein signalling). The Gα subunit has a low intrinsic GTPase activity and this is greatly increased by the RGS proteins. RGS structure is defined by an RGS-box region that is responsible for binding to Gα/GTP. However, many of the RGS proteins contain a number of other protein–protein interaction domains, such as PDZ, phosphotyrosine-binding (PTB), pleckstrin homology (PH) and phox homology (PX) domains, indicating that, in addition to their GAP activity, they may have other functions. One of these might be the regulation of the G-protein-activated inwardly rectifying K+ (GIRK) channel.

The function of RGS proteins has been clearly defined in phototransduction, where RGS9 acts to accelerate GTP hydrolysis by Gαt (step 6 in Module 10: Figure phototransduction). The level of RGS4, which modulates the activity of the Ca2+-sensing receptor (CaR), is reduced in the brain in Alzheimer's disease (AD) and could contribute to the onset of inflammation in Alzheimer's disease (Module 12: Figure Inflammation and Alzheimer's disease).

Monomeric G proteins

The monomeric GTP-binding proteins (G proteins) belong to a large family of approximately 150 members. Ras was the founding member, and the family is often referred to as the Ras family of small G proteins. Within this family, it is possible to recognize five subfamilies: Ras, Rho, Rab, Ran and ADP-ribosylation factor (Arf) (Module 2: Table monomeric G protein toolkit). The Ran family plays a role in nuclear transport, whereas the large Rab family functions in membrane trafficking. The Ras and Rho family are primarily involved in cell signalling, where they function as binary switches to control a number of cell signalling systems. This binary switch is driven by the binding of GTP, which represents the ON reaction, and is followed by the hydrolysis of the GTP by the endogenous GTPase activity. Most attention has focused on a small number of these signal transducers, and the following will be described in detail to illustrate their role in cell signalling:

Ras signalling mechanisms

Ras is a small (21 kDa) GTPase that plays a central role as a signal transducer for a number of signalling systems. It was first identified as a transducer for the tyrosine kinase-linked receptors (Module 1: Figure stimuli for enzyme-linked receptors), where it functions to relay information to the mitogen-activated protein kinase (MAPK) signalling pathway. It is now known that activated Ras is able to relay information to a number of other signalling pathways. Like other G proteins (Module 2: Figure G protein binary switching), Ras functions as a binary switch that is activated by binding GTP. There are thus two critical aspects to Ras action: how is the switch controlled and how is information relayed out to different signalling pathways (Module 2: Figure Ras signalling)? With regard to the first question, the ON reaction of the binary switch is controlled by a number of Ras guanine nucleotide exchange factors (RasGEFs) such as Son-of-sevenless (SoS), Ras guanine nucleotide release-inducing factors (RasGRFs) and Ras guanine nucleotide releasing proteins (RasGRPs) (Module 2: Table monomeric G protein toolkit). This conversion of Ras·GDP into Ras·GTP is inhibited by the tumour suppressor protein merlin. Neurofibromatosis type 2 is caused by mutations in merlin. The inactivation of Ras·GTP is mediated by a family of Ras GTPase-activating proteins (RasGAPs).

The Ras signalling mechanism relays information to a number of signalling targets:

Son-of-sevenless (SoS)

Son-of-sevenless (SoS) is a classical Ras guanine nucleotide exchange factor (RasGEF). Its mode of action is evident in the way the platelet-derived growth factor receptor (PDGFR) is linked to the mitogen-activated protein kinase (MAPK) signalling pathway (Module 1: Figure PDGFR activation). SoS binds to the Src homology 3 (SH3)-containing adaptor growth factor receptor-bound protein 2 (Grb2), which is attached to the phosphotyrosine residues of the activated receptor. Once it is associated with the receptor, SoS comes into contact with Ras and can begin to facilitate the exchange of GDP for GTP, thus creating the active Ras/GTP complex that begins to stimulate the MAPK signalling pathway (for further details see Module 2: Figure ERK signalling).

Ras guanine nucleotide release-inducing factors (RasGRFs)

The Ras guanine nucleotide release-inducing factors (RasGRFs) and Ras guanine nucleotide releasing proteins (RasGRPs) are also important GEFs that can relay information to Ras. The RasGRFs, which are strongly expressed in brain, are particularly important for the activation of the MAPK signalling pathway in neurons, where they respond to activation of N-methyl-D-aspartate (NMDA) receptors. One of the functions of the latter is to gate Ca2+, which then acts through calmodulin (CaM) to stimulate RasGRF. This is a highly localized signalling event, because the RasGRF is associated with the NMDA receptor and thus responds to the microdomain of Ca2+ near the mouth of the channel. This signalling mechanism functions in neuronal gene transcription (Module 10: Figure neuronal gene transcription).

Ras guanine nucleotide releasing proteins (RasGRPs)

The Ras guanine nucleotide releasing proteins (RasGRPs) are mainly expressed in haematopoietic cells (Module 2: Table monomeric G protein toolkit). These GRPs are also known as CalDAG-GEFs because they are sensitive to diacylglycerol (DAG) and Ca2+. They contain a C1 domain that binds DAG and they also have a pair of Ca2+-binding EF-hands. The sensitivity to both DAG and Ca2+ suggests that these RasGRPs may couple phosphoinositide signalling to the activation of the various pathways that are linked to Ras (Module 2: Figure Ras signalling).

Ras GTPase-activating proteins (RasGAPs)

There are a number of Ras GTPase-activating proteins (RasGAPs) that accelerate the OFF reaction by speeding up the hydrolysis of GTP to GDP by Ras. Examples of such RasGAPs include p120 RasGAP, neurofibromin, synaptic Ras GTPase activating protein 1 (SYNGAP1), Ca2+-promoted Ras inactivator (CAPRI) and Ras GTPase-activating-like (RASAL). The function of some of these RasGAPs, such as CAPRI and RASAL, are attracting considerable attention because they are Ca2+-sensitive proteins that rapidly translocate to the membrane through their C2 domains whenever there is an increase in the level of Ca2+ (Module 2: Figure Ras signalling).

SYNGAP1, which is expressed mainly in brain, is also activated by Ca2+ following its phosphorylation by Ca2+/calmodulin-dependent protein kinase II (CaMKII). The role of Ca2+ in modulating the activity of both the RasGEFs and RasGAPs indicates an important feedback interaction between the Ca2+ and Ras signalling pathways. Mutation of the SYNGAP1 gene during brain development brings about neuronal defects characterized by an excitation-inhibition imbalance that results in autism spectrum disorders (ASD) and intellectual disability (ID).

Rap signalling mechanisms

The Ras-related protein (Rap) family are typical small monomeric G-proteins that belong to the Ras superfamily (Module 2: Table monomeric G-protein toolkit). The Rap family has four members Rap1A, Rap1B, Rap2A and Rap2B. One of the guanine nucleotide-exchange factors (GEFs) is the cyclic AMP-dependent GEF EPAC that regulates the activation of Rap1 and Rap2B. This pathway can activate phospholipase Cε (PLCε) (Module 2: Figure cyclic AMP signalling). Such a mechanism has been implicated in the control of autophagy (Module 11: Figure autophagy).

Rab signalling mechanism

The Ras-associated binding (Rab) proteins are typical small monomeric G-proteins that belong to the Ras superfamily. The primary function of the Rabs is to control multiple phases of membrane and protein trafficking such as vesicle formation, motility, tethering and fusion to the acceptor membrane and signalling to other organelles. There are more than 60 human Rabs (Module 2: Table monomeric G-protein toolkit). The enormous proliferation of this family reflects their central role in the regulation of so many vesicle trafficking processes.

The continuous cycling between the cytosol and different membrane-bound compartments is shepherded by a number of protein factors and also requires the prenylation of Rab. This post-translational lipid modification depends on the addition of isoprenes, which are intermediaries in cholesterol biosynthesis. Following its synthesis, Rab is bound to GDP and is soluble in the cytoplasm (Module 2: Figure Rab signalling). This Rab.GDP is then recognized by the chaperone protein Rab escort protein (REP), which transfers it to the Rab geranylgeranyl transferase (GGT) that adds either one or two geranylgeranyl groups to cysteine residue(s) located at the C-terminus of the Rab. Mutation of REP causes choroideraemia, a form of blindness resulting from degeneration of the retinal pigment epithelium.

The prenylated Rab enters the Rab membrane cycle by binding to the GDP dissociation inhibitor (GDI). The Rab.GDP/GDI complex is then targeted to specific membrane sites by interacting with the GDI displacement factor (GDF). Once Rab has performed its function it is extracted from the membrane by GDI and remains in the cytosol until it is required again to carry out the same or some other function. Some patients with X-linked non-specific mental retardation have mutations in one of the genes for GDI.

Rabs carry out their multiple tasks through the operation of the Rab GTP/GDP cycle regulated by the co-ordinated activity of guanine nucleotide-exchange factors (GEFs) and GTPase-activating proteins (GAPs) (Module 2: Figure Rab signalling). Examples of the RAB GEFs include Rabex-5, Rabin3 and TRAPP1. Examples of some of the Rab GAPs are Rab3GAP, Rin1, Tbc1d1 and Tbc1d4 (also known as AS160). Tbc1d1 and Tbc1d4/AS160 are the targets of insulin action resulting in insertion of GLUT4-containing vesicles in skeletal muscle (Module 7: Figure skeletal muscle E-C coupling). The Rab.GTP form has the dynamic conformation and mobility to interact with many different downstream effectors that drive the multiple Rab responses (Module 2: Figure Rab signalling).

An allele of Tbc1d4, which reduces the translocation of GLUT4 to the plasma membrane of skeletal muscle (Module 7: Figure skeletal muscle E–C coupling), is linked to a susceptibility to obesity.

Rab1

Rab1 functions in vesicle tethering and fusion especially during both ER-to-Golgi and Golgi-to-ER transport (see steps 1 and 2 in Module 4: Figure membrane and protein trafficking). During the COPII-mediated transport from the ER to the Golgi (Module 4: Figure COPII-coated vesicles), TRAPP1 is the guanine nucleotide-exchange factor (GEF) that converts inactive Rab1.GDP into active Rab1.GTP during vesicle tethering prior to fusion to the Golgi membrane. The transport from the Golgi back to the ER begins with the golgin protein p115 recruiting Rab1b, which then activates the Golgi-specific brefeldin A resistant factor 1 (GBF1), which is a guanine nucleotide-exchange factor (GEF) for the small monomeric G-protein ADP-ribosylation factor 1 (Arf1) that has an important role in controlling key events such as coat formation, actin polymerization and Golgi vesicle budding during COPI-mediated transport from the Golgi to the ER (Module 4: Figure COPI-coated vesicles).

Rab3

There are three Rab3 proteins (Rab3A, Rab3B, Rab3C and Rab3D) that function in regulated exocytosis. These Rab3 members tend to be restricted to cells types with high rates of exocytosis such as neurons and neuroendocrine cells (Rab3A and Rab3C). Rab3B is located on tight junctions and secretory granules in epithelial cells. The more widely expressed Rab3D is located on the secretory granules of non-neuronal cells such as fat cells, adipocytes, exocrine glands and certain haematopoietic cells. Rab3A and Rab3C tend to reduce exocytosis, whereas Rab3B and Rab3D have a more positive role.

The activation of these Rabs is regulated by a Rab3GAP which has two components: a catalytic RabGAP1 subunit and a non-catalytic RabGAP2 subunit. Mutations in RabGAP1 have been linked to Warburg Micro syndrome whereas a milder form called Martsolf syndrome is cause by mutations in RabGAP2.

Rab4

Rab 4, together with myosin V, functions in early endosome to plasma membrane trafficking (Module 4: Figure early endosome budding). Rab 4, together with myosin V, also functions in early endosome to trans-Golgi network (TGN) trafficking (Module 4: Figure early endosome budding).

Rab5

Rab5 plays a role in endosome vesicle fusion to early endosomes by targeting vesicles to the early endosome (Module 4: Figure endosome vesicle fusion). Such a role is manifest during phagosome maturation (Module 4: Figure phagosome maturation). Rab5 also has a role in activating hVPS34 to promote vesicle nucleation during autophagy (Module 11: Figure autophagy signalling mechanisms).

Rabex-5 is the guanine nucleotide-exchange factor (GEFs) that activates Rab5, which then interacts with the effector Rabaptin-5. Rab22 may play a role in recruiting Rabex-5 to the early endosome.

Rab6

During the process of COPII-mediated transport from ER to Golgi (Module 4: Figure COPII-coated vesicles), Rab6 binds to Bicaudal D and thus provides a mechanism to attach the dynein motor complex to the vesicle for its transfer between the ER and Golgi membranes (Module 4: Figure dynein)

Rab 7

The two Rab7 proteins (Rab7A and Rab7B), which are expressed ubiquitously, play a role in regulating vesicle trafficking between the endocytic and autophagic compartments. Rab7 is found on the acidic and degradative organelles, such as the late endosomes, lysosomes, multivesicular bodies, phagosomes and autophagolysosomes. These mechanisms regulate the degradation of endocytic cargos such as EGF receptors and neurotrophic factor receptors such as TrkA and internalized cholesterol. The role of RAB7 in lysosomal transport depends on the effector protein Rab7-interacting lysosomal protein (RILP). Rab7 may also play a role in the retrograde trafficking of the signalling endosomes that supply trophic support to peripheral neurons.

Charcot–Marie–Tooth disease type 2B is caused by mutations in Rab7A.

Rab8

Rab8A interacts with myosin Vb to carry out non-clathrin-dependent endocytosis. It also plays a role in the biogenesis of cilia by interacting with cenexin/ODF2.

A defect in the transcriptional regulation of Rab8A has been linked to Microvillus inclusion disease.

Rab11

There are three Rab11 proteins (Rab11A, Rab11B and Rab11C), which function in both endocytic recycling and cytokinesis. Rab11 plays an important role in the transport of synaptic vesicles where it is attached to myosin Vb through the Rab11-family interacting proteins (FIPs) (Module 4: Figure myosin motor). Such trafficking events leading to receptor insertion are key components in the relationship between Ca2+ and synaptic plasticity during the process of learning (Module 10: Figure Ca2+-induced synaptic plasticity).

Rab11, together with its effector FIP2, also plays a role in the trafficking and insertion of aquaporin-2 (AQP2) (Module 7: collecting duct function) and also plays a role in controlling the trafficking and insertion of new vesicles to form the cleavage furrow during cytokinesis (Module 9: Figure cytokinesis).

Rab27

There are two Rab27 proteins (Rab27A and Rab27B), which function in intracellular transport of secretory granules and melanosomes (Module 4: Figure myosin motor). Rab27 exerts its actions through various effectors that fall into three families that have distinct functions. There are synaptotagmin-like proteins (Slps), Slp lacking C2 domains (Slac2s) and the Rab3-binding domain (RBD) family. They all have an N-terminal Rab27-binding domain known as the Slp-homology domain (SHD), which binds to the GTP-bound form of Rab27. The Slp and RBD families also have two C-terminal C2 domains, which enables these effectors to bind to phospholipids at the plasma membrane and thus enables Rab27a to regulate the secretory machinery at the vesicle docking step. The Slac2-a/Melanophilin and Slac2-c/MyRIP complexes link Rab27a to molecular motors such as MyoVa (Module 4: Figure myosin motor). The active Rab27.GTP carries out a number of functions:

  • One of the functions of RAB27A is to move melanosomes along actin filaments in retinal pigment epithelium cells and in melanocytes. A defect in the transport of melanocytes results in choroideraemia. In the case of melanogensis, Rab27A interacting with melanophilin that then uses myosin Va to move melanosomes to the cell periphery of the melanocytes (Module 7: Figure melanogenesis).
  • Another function of Rab27A is to regulate the secretion of perforin- and granzyme-containing cytotoxic granules at the immunological synapse that forms when cytotoxic T lymphocytes (CTLs) and natural killer (NK) cells interacts with their target cells, such as virus-infected cells and tumour cells. Interaction of the T cell receptor (TCR) with an antigen-presenting target cell induces the rapid polarization of the microtubule-organizing center (MTOC), which orchestrates secretion of the cytotoxic granules at the immunological synapse. The Rab27 effector hMunc13-4 contributes to the maturation and exocytosis of the perforin-containing granules. Several inherited conditions, which lead to functional impairment of this cytotoxic pathway result from defects in either Rab27A or hMunc13-4. Mutations in Rab27A have been linked to Griscelli syndrome type 2 (GS2), whereas mutations in hMunc13-4 cause familial hemophagocytic lymphohistiocytosis type 3 (FHL3).
  • Insulin granules are tethered to the plasma membrane using the effector Slp4-a/granuphilin-a to interact with both Rab27A and the SNARE protein syntaxin-1a.
  • In glucagon-secreting α-cells in the pancreas, glucagon granules are targeted to the plasma membrane by Slp2a.

Rab11-family interacting proteins (FIPs)

The Rab11-family interacting proteins (FIPs) serve as effectors for a number of Rabs and ADP-ribosylation factors (Arfs). There are five members, which all have a Rab11-binding domain (RBD) at the C-terminus and are divided into two classes:

Class I (have a phospholipid-binding C2 domain)

  • Rip11, which is also known as pp75 or FIP5, functions in the transport of GLUT-4 vesicles to the cell surface when adipocytes are stimulated by insulin.
  • FIP2, which is also known as optineurin (OPTN), functions to couple Rab 11 to the AMPA-containing synaptic vesicles (Module 4: Figure myosin motor). Such trafficking events leading to receptor insertion are key components in the relationship between Ca2+ and synaptic plasticity during the process of learning (Module 10: Figure Ca2+-induced synaptic plasticity).
  • Rab-coupling protein (RCP) is also known as FIP1C

ClassII (have typical Ca2+-binding EF-hands)

Mutations in FIP2/optineurin have been linked to some forms of glaucoma and amyotrophic lateral sclerosis (ALS).

Rac signalling mechanisms

The Rac signalling pathway has an important role in activating a number of signalling pathways (Module 2: Figure Rac signalling). Like other G proteins, Rac functions as a binary switch. It is inactive when bound to GDP, but when this GDP is exchanged for GTP, the Rac/GTP complex becomes active. External stimuli can activate this switch using different guanine nucleotide exchange factors (GEFs) such as T cell lymphoma invasion and metastasis (Tiam), Kalirin, Vav, SoS and P-Rex. Arf6, which participates in the Arf signalling pathway, stimulates the Rac GEF called downstream of Crk-180 homologue (DOCK180). Kalirin functions in the Ephrin (Eph) receptor signalling pathway, where it links the EphB receptor to Rac and actin assembly (Module 1: Figure Eph receptor signalling). The lipid second messenger PtdIns3,4,5P3, which is produced by the PtdIns 3-kinase signalling pathway, is particularly effective in stimulating many of these GEFs.

The activated Rac/GTP complex has a number of important actions:

  1. Rac is responsible for mediating the action of the PtdIns 3-kinase signalling pathway in stimulating the NADPH oxidase to initiate the reactive oxygen species (ROS) signalling pathway (Module 2: Figure plasma membrane ROS formation).
  2. It stimulates the JNK (c-Jun N-terminal kinase) signalling pathway (Module 2: Figure JNK signalling).
  3. It promotes actin stability by stimulating p21-activated kinase (PAK) to phosphorylate LIM kinase, which phosphorylates cofilin to inhibit its ability to cut actin.
  4. It contributes to focal adhesion integrin signalling (Module 6: Figure integrin signalling).
  5. It promotes actin remodelling in several ways. It stimulates PtdIns4P 5-kinase to increase the formation of PtdIns4,5P2 to control actin remodelling (Module 4: Figure actin remodelling). It also acts through insulin receptor substrate p53 (IRSp53), which activates Wiskott–Aldrich syndrome protein (WASP) to co-ordinate the activity of the actin-related protein 2/3 complex (Arp2/3 complex) that initiates actin polymerization (Module 4: Figure actin remodelling). Such a role is evident in neurons during spine morphogenesis (Module 10: Figure postsynaptic density) and during the Ca2+ and synaptic plasticity changes that occur during long-term potentiation (LTP) (Module 10: Figure Ca2+-induced synaptic plasticity).

Rho signalling mechanisms

The Rho (Ras homologue) signalling pathway has an important role in regulating a number of systems, in particular the operation of the cytoskeleton. It regulates the function of actin in two main ways. Firstly, it controls some of the processes that function in actin assembly and it also operates one of the control mechanisms that regulate the myosin II–actin interaction responsible for contraction both in smooth muscle cells and non-muscle cells (Module 2: Figure Rho signalling). Like other G proteins, Rho functions as a binary switch. It is inactive when bound to GDP, but when this GDP is exchanged for GTP, the Rho/GTP complex becomes active. External stimuli can activate this switch using different Rho guanine nucleotide exchange factors (RhoGEFs). There are a large number of such RhoGEFs that are characterized by having a Dbl homology (DH) domain. The term Dbl comes from diffuse B cell lymphoma, which is the cell line where the first RhoGEF was identified. The family of Dbl-containing GEFs has now expanded to 69 members, some of which are shown in Module 2: Table monomeric G protein toolkit. These GEFs are not necessarily specific for Rho: some are rather promiscuous and will also function as Rac and Cdc42 GEFs. In the case of Rho, ephexin appears to be specific for coupling ephrin receptors to Rho (Module 2: Figure Rho signalling). Also leukaemia-associated RhoGEF (LARG), p115-RhoGEF and PDZ-RhoGEF are specific for coupling G-protein-coupled receptors (GPCRs) to Rho activation. Net-1, which appears to be sensitive to PtdIns3,4,5P3 (PIP3), is normally found in the nucleus, but translocates to the cytoplasm, where it functions to activate RhoA. Another important GEF is Ect2, which is also found in the nucleus during interphase, but then associates with microtubules during mitosis, where it stimulates the activation of Rho cytokinesis (Module 9: Figure cytokinesis).

The activated Rho/GTP complex is able to stimulate a number of signalling processes, many of which are directed towards remodelling of actin and its contraction. Many of the actions of Rho are carried out through a Rho kinase (ROCK). Diaphanous-related forming 1 (Dia1) is one of the downstream effectors that responds to Rho/GTP by interacting with profilin to increase actin polymerization.

Oligophrenin-1 (OPHN1) is a Rho Gap that functions to stabilize AMPA receptors at neuronal synapses and thus plays an important role in both synapse formation and synaptic plasticity. Mutations in the OPHN1 gene have been linked to mental retardation.

The following are some examples of the function of Rho signalling:

Cdc42 signalling mechanisms

The Cdc42 signalling mechanism has an important role in controlling the actin cytoskeleton. It regulates some of the processes that function in actin assembly (Module 2: Figure Cdc42 signalling). Like other G proteins, Cdc42 functions as a binary switch. It is inactive when bound to GDP, but when this GDP is exchanged for GTP, the Cdc42/GTP complex becomes active. External stimuli activate this switch using different guanine nucleotide exchange factors (GEFs), such as intersectin-long (ITSN-L) and PAK-interacting exchange factor α (α-Pix). Just how these Cdc42 RhoGEFs are activated is still not properly understood. In the case of the Ephrin (Eph) receptor signalling pathway, intersectin is activated by binding to the EphA receptor (Module 1: Figure Eph receptor signalling). The activated Cdc42/GTP complex then stimulates various elements that regulate actin polymerization. It promotes actin stability by stimulating p21-activated kinase (PAK) to phosphorylate LIM kinase, which phosphorylates cofilin to inhibit its ability to cut actin. One of its main actions is to stimulate the Wiskott–Aldrich syndrome protein (WASP) that acts on the actin-related protein 2/3 complex (Arp2/3 complex) to initiate actin polymerization (Module 4: Figure actin remodelling). This action is facilitated by Cdc42 acting on profilin through insulin receptor substrate p53 (IRSp53) and Mena.

The Cdc42 signalling mechanism plays an important role in actin polymerization in synaptic spines as part of the events associated with Ca2+ and synaptic plasticity (Module 10: Figure Ca2+-induced synaptic plasticity).


Module 2: Table monomeric G protein toolkit Summary of the monomeric G proteins with their guanine nucleotide exchange factors (GEFs) and GTPase-activating proteins (GAPs).

Monomeric G protein, GEF or GAP Comments
Monomeric G proteins (approximately 150 members)
Ras family (36 members) Contribute to the activation of multiple signalling pathways (Module 2: Figure Ras signalling)
  H-Ras
  K-Ras
  N-Ras
  R-Ras
  RalA
  RalB
  Rap1
  Rap2 Rap2B has been implicated in the control of autophagy (Module 11: Figure autophagy)
  RagA RagB RagC RagD The RagB/RagD heterotrimer interacts with TORC1 to control autophagy (Module 11: Figure autophagy)
Rho family (22 members)
  Rho Primary function is cytoskeletal remodelling (Module 2: Figure Rho signalling)
   RhoA
   RhoD
   RhoE
   RhoF
   RhoH
   RhoV
   Mito Rho
   Rho BTB
  Rac Cytoskeletal remodelling and ROS signalling (Module 2: Figure Rac signalling)
   Rac1
   Rac2
   Rac3
   Cdc42 Primary function is cytoskeletal remodelling (Module 2: Figure Cdc42 signalling)
Rab family A large family of GTPases that function in membrane trafficking
  Rab1A ER-to-Golgi trafficking. Rab1 interacts with the golgins (golgin-84, golgin-95, p115 and giantin)
 Rab1B
  Rab2A ER-to-Golgi trafficking. Rab2 interacts with the golgins (golgin-45)
 Rab2B
  Rab3A Control exocytosis by interacting with effectors such as Rabphilin, Rabin3, RIM1/2 and synapsin
 Rab3B
 Rab3C
 Rab3D
  Rab4A Endocytic recycling and functions in GLUT4 biogenesis, sorting and exocytosis
 Rab4B
 Rab4C
 Rab4D
  Rab5A Functions in early endosome fusion and phagosome maturation. Interacts with Rabaptin-5, Rabenosyn-5, EEA1 and Rabanykyrin-5 (see Module 4: Figure endosome vesicle fusion)
 Rab5B
 Rab5C
 Rab5D
  Rab6A Functions at the Golgi where it interacts with Giantin, Rabankyrin. Couples to dynein through Bicaudal D1 (see Module 4: Figure dynein)
 Rab6B
 Rab6C
 Rab6D
  Rab7A Operates in the late endosomal pathway during formation of lysosomes. Rab7A is mutated in Charcot–Marie–Tooth type 2B disease
 Rab7B
  Rab8A Functions in trafficking from Golgi to the plasma membrane. Interacts with optineurin and FIP-2. Rab8A binds to myosin Vb and functions in non-clathrin-dependent endocytosis
 Rab8B
 Rab8C
  Rab9C Functions in trafficking from late endosomes to the Golgi.
 Rab9D
  Rab10 Functions in trafficking from Golgi to the plasma membrane. Interacts with Rim1.
  Rab11A Located in recycling endosomes. Functions in trafficking and internalization of aquaporin-2 (AQP2). Interacts with multiple effectors (FIPs, myosin Vb and PtdIns 4-kinaseβ)
 Rab11B
  Rab12 Functions in trafficking from Golgi to the plasma membrane.
  Rab13 Functions in biogenesis of tight junctions
  Rab14 Functions in trafficking from Golgi to the early endosomes.
  Rab15 Early endosome trafficking
  Rab17 Operates in the apical recycling endosomes and functions in ciliogenesis
  Rab18 Functions in formation of lipid droplets from the ER
  Rab19A Located in the Golgi
  Rab19B
  Rab20 Trafficking of connexin-43
  Rab21 Located on early endosomes and functions in cytokinesis
  Rab22 Located on early endosomes
  Rab23 Functions in formation of phagosomes
  Rab24 Functions in autophagy
  Rab25 Functions in integrin recycling
  Rab26 Interacts with Rim1 in exocytosis
  Rab27A Intracellular transport of secretory granules and melanosomes (Module 4: Figure myosin motor)
 Rab27B
 Rab28
 Rab29
  Rab30 Located on Golgi
  Rab31 Located on early endosomes
  Rab32 Located on early melanosmes and mitochondria
  Rab33A ER-to-Golgi trafficking
 Rab33B
  Rab34 Functions in macropinocytosis
  Rab35 Located on recycling endosomes and functions in cytokinesis and formation of the immunological synapse
 Rab36
  Rab37 Located on secretory granules and functions in mast cell degranulation
  Rab38 Functions in formation of melanosomes
  Rab39 Located on the Golgi
 Rab40A
 Rab40B
 Rab40C
  Rab41
  Rab42
  Rab43 ER-to-Golgi trafficking
  Rab44
Arf family There are six ADP-ribosylation factors that function in actin remodelling and membrane trafficking (Module 2: Figure PLD signalling)
Class I Operate mainly at the Golgi
 Arf1
 Arf2
 Arf3
Class II Operate mainly at the Golgi
 Arf4
 Afr5
Class III
 Arf6 Operates mainly at the plasma membrane. It functions in endocytosis of AMPA receptors at synaptic endings (Module 10: Figure Ca2+-induced synaptic plasticity)
Arf-like (ARL) A small GTPase that interacts with golgins
 Ran A small GTPase that functions in nucleocytoplasmic transport
Guanine nucleotide-exchange factors (GEFs)
RasGEFs
  SoS1 Son-of-sevenless-1
  SoS2 Son-of-sevenless-2
  RasGRF1 Expressed in brain, particularly hippocampus
  RasGRF2
  RasGRF3 Expressed in brain
  RasGRP1 Expressed in T cells
  RasGRP2 Expressed in neutrophils and platelets
  RasGRP3 Expressed in B cells
  RasGRP4 Expressed in mast cells
Rho family GEFs
  The Dbl family The 69 family members have a Dbl homology domain (DH)
   Dbl Diffuse B cell lymphoma
   Dbs Dbl's big sister
   Ect2 Epithelial-cell transforming gene-2 that functions in cytokinesis (Module 9: Figure cytokinesis)
   Ephexin Functions in growth cone collapse mediated by EphA receptors (Module 1: Eph receptor signalling)
   DOCK180 Downstream of Crk-180 homologue is a Rac GEF (Module 2: Figure Arf signalling), which functions in focal adhesion integrin signalling (Module 6: Figure integrin signalling)
   Fgd1 Facial genital dysplasia
   ITSN-L Intersectin-long controls Cdc42 activity during EphB receptor signalling (Module 1: Eph receptor signalling)
   Kalirin Functions to activate Rac during the action of Eph receptors on dendritic spine morphogenesis (Module 1: Eph receptor signalling)
   LARG Leukaemia-associated RhoGEF
   NET1 Neuroepithelioma transforming gene 1
   p115-RhoGEF Links G protein-coupled receptors to Rho
   PDZ-RhoGEF Links G protein-coupled receptors to Rho
   P-Rex1 PtdIns3,4,5P3-dependent Rac exchanger 1
   α-Pix PAK-interacting exchange factor α
   β-Pix PAK-interacting exchange factor β
   Tiam1 T cell lymphoma invasion and metastasis-1
   Tiam2 T cell lymphoma invasion and metastasis-2
   Trio Links G protein-coupled receptors to Rho and Rac (Module 9: Figure growth factor signalling)
   Vav1
   Vav2
   Vav3
RAB GEFs See Module 2: Figure Rab signalling
  Rabex-5 A GEF for Rab 5 (Module 4: Figure endosome vesicle fusion)
  Rabin3
  TRAPP1 A GEF for Rab1 (Module 4: Figure COPII-coated vesicles)
Arf GEFs
  Cytohesin 1
  Cytohesin 4
  GBF1 Golgi-specific brefeldin A resistant factor 1 functions in COPI-mediated protein transport from the Golgi to the ER (Module 4: Figure COPI-coated vesicles)
  ARNO/Geα2 Arf nucleotide-binding site opener (ARNO)
  BRAG2 This Arf6 GEF binds to AMPA receptors to promote endocytosis
GTPase-activating proteins (GAPs)
Ras GAPs
  p120RasGAP
  Neurofibromin This tumour suppressor is lost in the inherited disorder neurofibromatosis type 1
  SynGAP
  GAP1m
  GAP1IP4BP Ins1,3,4,5P4 binds to this protein, causing it to dissociate from the plasma membrane
  CAPRI Ca2+-promoted Ras inactivator
  RASAL Ras GTPase-activating-like
  DAB2IP
  Miro Mitochondrial Rho-GTPase functions in mitochondrial motility (Module 5: Figure mitochondrial motility)
Rac GAPs
  3BP-1 3BP-1 inactivates Rac.GTP by enhancing GTP hydrolysis (Module 2: Figure Rac signalling)
Rho GAPS
  DLC1 (p122RhoGAP in mice) Deleted in liver cancer 1 is altered in many tumours. It regulates actin formation in adhesion complexes (Module 6: Figure integrin signalling)
  p190-RhoGAP Inhibits Rho activity during neutrophil chemotaxis (Module 11: Figure neutrophil chemotactic signalling)
  Oligophrenin-1 (OPHN1) Mutations in OPHN1 have been linked to mental retardation
Rab GAPs See Module 2: Figure Rab signalling
  Rab3GAP
  Rin1
  Tbc1d1 These Rab GAPs function in the insertion of the GLUT4 transporter (Module7: Figure skeletal muscle E-C coupling)
 Tbcd14/AS160
Arf GAPS Function to inactivate Arf.GTP (Module 2: Figure Arf signalling)
  Arf GAP1 group
   ArfGAP1 Function in removing the coat from COPI-coated vesicles (Module 4: Figure COPI-coated vesicles)
   ArfGAP2
   ArfGAP3
   SMAP1 Stromal membrane-associated protein
   SMAP2
   G-protein-coupled receptor kinase interactors (Module 10: Figure Ca2+-induced synaptic plasticity)
 Git1
 Git2
  AZAP group Arf GAP with ANK repeats and PH domain group of Arf GAPs
   ACAP1 Arf GAP with coiled-coil, ANK repeats and PH domain (ACAP)
  ASCP2
  ASCP3
   ASAP1 Arf GAP with Src homology 3, ANK repeats and PH domain (ASAP)
  ASAP2
  ASAP3
   AGAP1 Arf GAP with GLD domain, ANK repeats and PH domain (AGAP)
  AGAP2
  AGAP3
   ARAP1 Arf GAP with Rho GAP, ANK repeats and PH domain (ARAP)
  ARAP2
  ARAP3


Since some of the families are very large, only representative members have been included. Data for the large Rab family was based on that from Supplementary Table S1 from Stenmark (2009).



Arf signalling mechanisms

The ADP-ribosylation factors (Arfs) are best known for their role in membrane and protein trafficking. There are six mammalian Arfs that are divided into three classes (Module 2: Table monomeric G-protein toolkit). The Class I and Class II Arfs are found mainly at the Golgi and also on various endosomal membranes where they function in actin remodelling and formation of the coat protein complexes responsible for the vesicle budding that transfers cargo form one compartment to another. Such a role is evident in the case of COPI-mediated transport from Golgi to ER (Module 4: Figure COPI-coated vesicles). Class III has a single member Arf6, which functions at the plasma membrane where it can have marked effects on actin polymerization. One action of Arf6 is to induce the trafficking of Rac1 to the plasma membrane where it is activated by the Rac GEF called downstream of Crk-180 homologue (DOCK180) to increase membrane ruffling and focal adhesion actin attachment (Module 6: Figure integrin signalling). The DOCK180 acts as a complex in association with a protein called engulfment and cell motility (ELMO). Arf6 has also been implicated in the endocytosis of AMPA receptors during the process of Ca2+-induced synaptic plasticity at synaptic endings (See step 4 in Module 10: Figure Ca2+-induced synaptic plasticity). This function of Arf6 depends on the activity of BRAG2, which is an Arf GEF that binds to the Glu2 subunit of the AMPA receptor.

Like other monomeric GTP-binding proteins (G-proteins), Arf activity is regulated by a balance between Arf guanine nucleotide-exchange factors (Arf GEFs) and the Arf GTPase-activating proteins (Arf GAPs) (Module 2: Figure Arf signalling). Examples of Arf GEFs are Golgi-specific brefeldin A resistant factor 1 (GBF1) and Arf nucleotide-binding site opener (ARNO). These activation mechanisms are counteracted by the Arf GAPs that facilitate the hydrolysis of GTP to GDP thus converting the complex back into its inactive Arf.GDP state. There are 24 human genes coding for Arf GAPs, which are characterized by having an Arf GAP domain, that can be separated into two main groups: the Arf GAP1 type and the Arf GAP with ANK repeats and PH domain (AZAPs) (Module 2: Table monomeric G protein toolkit).

The Arf GAP1 group has three genes coding for Arf GAP1-3, two genes for stromal membrane-associated protein 1 and 2 (SMAP1 and SMAP2) and two genes coding for G-protein-coupled receptor kinase interactors 1 and 2 (Git1 and Git2). Regulation of Arf1 by activation of Git1 contributes to the actin depolymerization that occurs during synaptic plasticity (Module 10: Figure Ca2+-induced synaptic plasticity). ArfGAP1 and ArfGAP2/3 inactivates Arf1.GTP resulting in the shedding of the COPI coat as the vesicles approach the endoplasmic reticulum (See step 7 in Module 4: Figure COPI-coated vesicles).

Another group of Arf GAPs with ANK repeats and PH domain (AZAPs) have 12 genes that fall into four groups and the ‘Z’ refers to the protein domain that characterizes each subgroup:

  • Arf GAP with coiled-coil, ANK repeats and PH domain (ACAP)
  • Arf GAP with GLD domain, ANK repeats and PH domain (AGAP)
  • Arf GAP with Rho GAP, ANK repeats and PH domain (ARAP)
  • Arf GAP with Src homology 3, ANK repeats and PH domain (ASAP)

The ASAP family has three genes (ASAP1–3). The ASAP3–Arf6 complex is of interest because it is sensitive to Ca2+ that binds to a specific region at the complex interface. This sensitivity to Ca2+ suggests that there may be cross-talk between the Arf and Ca2+ signalling pathways.

To carry out its role in vesicle trafficking, the Arfs have two main actions. First, they catalyse the nucleation and assembly of coat protein complexes during the process of vesicle budding. This role is particularly evident at the Golgi during the formation of COPI vesicles (Module 4: Figure COPI-coated vesicles). The Arfs also play a role in recruiting various Arf-associated golgins, such as GMAP-210, to the Golgi membrane.

The Arfs also function by activating the PtdIns4,5P2 signalling cassette responsible for controlling multiple functions (Module 2: Figure PtdIns4,5P2 signalling). The Arfs act by recruiting and stimulating both PtdIns 4 kinase IIIβ and the PtdIns4P 5-kinase (Module 2: Figure Arf signalling). All of the Arfs are responsible for phospholipase D (PLD) activation (Module 2: Figure PLD signalling). The subsequent formation of phosphatidic acid (PA), which is a bioactive lipid, has two actions: it acts synergistically to stimulate PtdIns4P 5-kinase and it alters membrane curvature during vesicle budding (Module 2: Figure Arf signalling).

Dedicator of cytokinesis 8 (DOCK8)

There is a family of eleven dedicator of cytokinesis (DOCK) proteins that are guanine nucleotide-exchange factors (GEFs) that appear to act primarily on the Rac signalling mechanism and the Cdc42 signalling mechanism. A typical example is DOCK8, which has an N-terminal DOCK Homology Region-1 (DHR-1) responsible for the guanine nucleotide-exchange reaction and a C-terminal DHR-2 domain that is thought to bind phospholipids. This family of DOCK proteins may act by remodelling actin to carry out a variety of cellular functions such as the control of cell migration, morphology, adhesion and proliferation. During the process of B-cell differentiation in the lymph node, DOCK8 is activated following its interaction with the integrin intercellular adhesion molecule 1 (ICAM1) (Module 8: Figure B cell maturation signalling).

Loss-of-function mutations in DOCK8 have been linked to variants of hyper-IgE syndromes (HIES) and mental retardation.

p21-activated kinase (PAK)

The p21-activated kinases (PAKs) are some of the major downstream targets of Rac and Rho (Module 2: Figure Rho-regulated kinases). They have been implicated in a large number of processes such as actin remodelling, cell cycle control, transcription and apoptosis. There are six isoforms (PAK1–PAK6) with slightly different functions. Rac acts through LIM kinase 1 (LIMK1) to phosphorylate cofilin (Module 2: Figure Rac signalling). Rho acts through PAK4 to phosphorylate the same LIMK1 (Module 2: Figure Cdc42 signalling). Another function of PAK1 is to phosphorylate myosin light chain kinase (MLCK), resulting in a decrease in the activity of the actin–myosin contractile system. PAK1 can also be activated independently of Rac through the calcium and integrin-binding protein 1 (CIB1) that acts by stimulating PAK1 (Module 2: Figure Rac signalling).

The fragile X mental retardation protein 1 RMRP1, which is mutated in fragile X syndrome (FXS), may act to inhibit the function of PAK particularly during actin remodelling in neuronal spines (Module 10: Figure Ca2+-dependent synaptic plasticity).

Myotonic dystrophy kinase-related Cdc42-binding kinase (MRCK)

Myotonic dystrophy kinase-related Cdc42-binding kinase (MRCK) has a number of functional domains and belongs to the same family as Rho kinase (ROK) (Module 2: Figure Rho-regulated kinases). MRCK, which is activated by Cdc42·GTP, may play a role in actin formation during the growth of neurites. One of the downstream targets of MRCK is the MYPT1 subunit (Module 5: Table PP1 regulatory and inhibitory subunits and proteins) that binds protein phosphatase 1δ (PP1δ) and which dephosphorylates the myosin light chain (MLC).

Rho kinase (ROCK)

Rho kinase (ROCK) is a serine/threonine kinase that phosphorylates key regulators of actin and myosin function (Module 2: Figure Rho-regulated kinases). ROCK acts to stabilize actin by phosphorylating LIM, which then phosphorylates cofilin to prevent it from severing actin. Rho also acts on diaphanous-related formin protein (Dia), which belongs to a formin-related protein family that interacts with profilin to promote actin polymerization.

The other major function of ROCK is to control contraction by activating myosin II in smooth muscle cells and in non-muscle cells. The activity of myosin II is regulated by the phosphorylation status of the myosin light chain (MLC) that is associated with the myosin head. When MLC is dephosphorylated, myosin is inactive, but when it is phosphorylated by the Ca2+-sensitive myosin light chain kinase (MLCK), myosin can begin to interact with actin to induce contraction. ROCK can influence the phosphorylation of MLC by inhibiting the protein phosphatase 1 by phosphorylating the scaffolding protein myosin phosphatase targeting subunit 1 (MYPT1) (Module 5: Table PP1 regulatory, targeting and inhibitory subunits).

ROCK may also function to inhibit endocytosis by phosphorylating endophilin.

Ca2+ signalling

Ca2+ signalling is one of the major signalling systems in cells (/1Module 2: Figure cell signalling pathways). It functions to regulate many different cellular processes throughout their life history. It triggers new life at the time of fertilization. It controls many processes during development, and once cells have differentiated, it governs the activity of most cellular processes, effectively determining how we metabolize, secrete, move and think. There also is a darker side to its action, because larger than normal elevations can cause cell death, either in the controlled manner of programmed cell death (apoptosis) or in the more catastrophic necrotic changes that occur during processes such as stroke or cardiac ischaemia.

The /1basic mechanism of Ca2+ signalling is relatively simple in that it depends upon an increase in the intracellular concentration of this ion. The Ca2+ concentration is low when cells are at rest, but when an appropriate stimulus arrives, there is a sudden elevation, which is responsible for a change in cellular activity. However, there are multiple variations of this relatively simple theme. The versatility of Ca2+ signalling is achieved by having an extensive Ca2+ toolkit from which a large number of /1Ca2+ signalling signalsomes are assembled. This large toolkit contains many different components that can be mixed and matched to create many different /1Ca2+ signalling modules. There are /1Ca2+ entry channels, which control the entry of Ca2+ from the outside. There are /1Ca2+ release channels, which control the release of Ca2+ from internal stores. The /1Ca2+ buffers ensure that the concentration of Ca2+ remains within its operation range and does not rise to levels that can induce cell death. Once Ca2+ has carried out its signalling function, there are /1Ca2+ pumps and exchangers that remove it from the cytoplasm by either extruding it from the cell or returning it to the internal stores. /1Ca2+ signalling functions are carried out by various /1Ca2+ sensors and /1Ca2+ effectors that are responsible for translating Ca2+ signals into a change in cellular activity.

Basic mechanism of Ca2+ signalling

Cells at rest maintain a low intracellular concentration of Ca2+ (approximately 100 nM), but this increases rapidly into the micromolar range when cells are stimulated (/1Module 2: Figure basic Ca2+ signalling mechanism). This increase in intracellular Ca2+ can operate over a very wide time domain (e.g. microseconds to hours) to regulate many different cellular processes. This very wide temporal range of Ca2+ signalling is an intrinsic property of the /1Ca2+ signalling modules.

An important feature of Ca2+ signalling is its dynamic nature, as exemplified by the fact that Ca2+ signals invariably appear as a brief transient. The rising phase of each transient is produced by the ON reactions, whereas the falling phase depends on the OFF reactions (Module 2: Figure Ca2+ signalling dynamics). At any moment in time, the level of Ca2+ is determined by the balance between the Ca2+ ON reactions that increase Ca2+ and the Ca2+ OFF reactions that remove it from the cytosol. An important aspect of Ca2+ dynamics is that there must be a precise balance between these ON and OFF reactions in order to maintain intracellular Ca2+ homoeostasis. Such homoeostasis is of critical importance where Ca2+ is being used repetitively to control processes such as the cardiac ventricular cell Ca2+ release and neuronal dopaminergic (DA) neuronal pacemaker mechanisms (Module 10: Figure tonic oscillation in DA neurons).

Another important aspect of the ON and OFF reactions is their spatial location. An example of this spatial organization is the ER/mitochondrial Ca2+ shuttle (Module 5: Figure ER/mitochondrial shuttle), where events at the ER are closely linked to those in the mitochondria. Another important spatial aspect is that the ON reactions are often closely associated with the effector systems that respond to Ca2+. For example, voltage-operated channels (VOCs) in presynaptic endings are associated with the synaptic vesicles, thus producing a highly localized puff of Ca2+ to trigger exocytosis (Module 4: Figure Ca2+-induced membrane fusion). Similarly, the type 2 ryanodine receptors (RYR2s) of cardiac cells are lined up close to the contractile filaments to ensure that Ca2+ will rapidly stimulate contraction.

In cases where cells need to be stimulated over a long time, these transients are repeated at set intervals to set up /1Ca2+ oscillations. These oscillations are part of the spatiotemporal aspects of Ca2+ signalling.

Ca2+ signalling signalsome

Cells have access to a large Ca2+ signalling toolkit (/1Module 2: Table Ca2+ signalling toolkit). Many of the toolkit components have similar functions (/1Module 2: Figure Ca2+ signalling toolkit), which represents a generic Ca2+ signalling system. In reality, however, each cell type has a clearly defined subset of toolkit components that will be referred to as a Ca2+ signalling signalsome. These cell-specific signalsomes are put in place during development when a process of /1signalsome expression enables each differentiating cell to select out those signalling components it will require to control its particular functions (/1Module 8: Figure signalsome expression). There are an enormous number of cell-specific Ca2+ signalsomes (/1Module 2: Figure cell-specific Ca2+ signalsomes). The important point is that each signalsome generates a cell-specific Ca2+ signal with characteristic spatial and temporal properties.

A signalsome is defined here as the collection of signalling components that make up each cell-specific signalling system. The Ca2+ signalsomes of different cell types often display recurring themes in the form of /1Ca2+ signalling modules (/1Module 2: Figure Ca2+ modules).

Ca2+ signalling modules

The Ca2+ signalling system in specific cell types is often not a single entity, but is made up of distinct modules, which are mixed and matched to produce the cell-specific systems found in different cell types. Some of the main modules used by cells are summarized in /1Module 2: Figure Ca2+ modules:

  1. Agonists such as the neurotransmitters glutamate and ATP act directly on receptor-operated channels (ROCs) in the plasma membrane to allow external Ca2+ to enter the cell.
  2. Second messengers such as diacylglycerol (DAG), cyclic AMP, cyclic GMP and arachidonic acid acting from the cytoplasmic side open second messenger-operated channels (SMOCs) in the plasma membrane.
  3. Membrane depolarization (ΔV) activates voltage-operated channels (VOCs) in the plasma membrane to allow a rapid influx of external Ca2+.
  4. Membrane depolarization (ΔV) activates a specific VOC isoform, the CaV1.1 L-type channel, that activates the ryanodine receptor 1 (RYR1) in skeletal muscle through a direct conformational-coupling mechanism.
  5. Membrane depolarization (ΔV) activates voltage-operated channels (VOCs) in the plasma membrane to allow a rapid influx of external Ca2+ (see Module 3) to provide a Ca2+ trigger that then activates the ryanodine receptor 2 (RYR2) to release Ca2+ stored in the sarcoplasmic reticulum (SR) through a process of Ca2+-induced Ca2+ release (CICR). This mechanism is found in cardiac muscle and neurons.
  6. Agonists acting on cell-surface receptors generate inositol 1,4,5-trisphosphate (InsP3) (Module 2: Figure InsP3 and DAG formation), which then diffuses into the cell to activate the InsP3 receptor (InsP3R) to release Ca2+ from the ER.

Ca2+ ON reactions

In response to external stimuli, channels in the plasma membrane or ER are opened, and Ca2+ flows into the cytoplasm to bring about the elevation of cytosolic Ca2+ responsible for cell activation (/1Module 2: Figure Ca2+ signalling dynamics). During these ON reactions, the cell employs a variety of both /1Ca2+ entry channels and /1Ca2+ release channels to create Ca2+ signals with markedly different spatial and temporal properties.

The entry of Ca2+ across the plasma membrane is carried out by many different channels whose names indicate how they are opened:

  • /1Voltage-operated channels (VOCs)
  • /1Agonist-operated channels (AOCs)
  • /1Receptor-operated channels (ROCs)
  • /1Second messenger-operated channels (SMOCs)
  • /1Store-operated channels (SOCs)

Release of Ca2+ from internal stores is carried out by different types of channels and control mechanisms:

One of the major problems in Ca2+ signalling has been to determine how stimuli arriving at the cell surface gain access to these internal stores. Two main mechanisms have been identified:

  1. Conformational coupling through a protein–protein interaction. This is a very fast mechanism that depends upon a sensor in the plasma membrane interacting directly with an internal release channel. The receptor on the cell surface is the /1CaV1.1 L-type channel (a voltage sensor), which is coupled to the type 1 /1ryanodine receptor 1 (RYR1) (Ca2+ module 4 in /1Module 2: Figure Ca2+ modules). Information is transferred through a process of conformational coupling. This mechanism is restricted to skeletal muscle (/1Module 7: Figure skeletal muscle E-C coupling) and perhaps also to some neurons.
  2. Generation of diffusible second messengers. Activation of receptors or channels on the cell surface generate second messengers that then diffuse into the cell to activate release channels. One of the most significant Ca2+-mobilizing messengers is Ca2+ itself, which is a potent activator of the two main internal release channels, the /1ryanodine receptors (RYRs) and the /1inositol 1,4,5-trisphosphate receptors (InsP3Rs). This /1Ca2+-induced Ca2+ release (CICR) mechanism has the unique property of being autocatalytic and plays a central role in generating those Ca2+ signals that appear as regenerative Ca2+ waves (/1Module 2: Figure Ca2+-induced Ca2+ release).


Module 2: Table Ca2+ signalling toolkit The Ca2+ signalling toolkit.

Component Comments
RECEPTORS AND TRANSDUCERS
G protein-coupled receptors (GPCRs) See Module 1: Table G-protein-coupled receptors for a list of those GPCRs that act by stimulating phospholipase Cβ
Tyrosine-kinase-linked receptors See Module 1: Figure tyrosine kinase-linked receptors. (The receptors shown opposite can activate phospholipase Cγ)
 Platelet-derived growth factor receptor (PDGFR) See Module 1: Figure PDGFR activation
  PDGFRα
  PDGFRβ
 Epidermal growth factor receptor (EGFR)
  ERBB1–ERBB4
 Vascular endothelial growth factor receptor (VEGFR) See Module 9: Figure VEGF-induced proliferation
  VEGFR1–VEGFR3
G Proteins
 Gαq, Gα11, Gα14, Gα15, Gα16, Gβγ These G protein subunits activate phospholipase C (Module 2: Figure heterotrimeric G protein signalling)
Guanine nucleotide exchange factors
RasGRF1 A Ca2+-activated GEF expressed in brain (Module 2: Figure Ras signalling)
Regulators of G protein signalling (RGS)
 RGS1, RGS2, RGS4, RGS16
Phospholipase C (PLC)
 PLCβ1–4, PLCγ1–2, PLCδ1–4, PLCε, PLCζ PLC hydrolyses PtdIns4,5P2 to form InsP3 and DAG (Module 2: Figure PLC structure and function)
ADP-ribosyl cyclase Functions to generate cADPR and NAADP (Module 2: Figure cADPR/NAADP function)
 CD38
CHANNELS
Plasma membrane channels
Voltage-operated channels (VOCs) See Module 3: Table VOC classification for further details
  CaV1.1 (L-type)
  CaV1.2 (L-type)
  CaV1.3 (L-type)
  CaV1.4 (L-type)
  CaV2.1 (P/Q-type)
  CaV2.2 (N-type)
  CaV2.3 (R-type)
  CaV3.1 (T-type)
  CaV3.2 (T-type)
  CaV3.3 (T-type)
Ca2+-sensitive ion channels
  Ca2+-activated K+ channels See Module 3: Table properties of Ca2+-sensitive K+ channels for further information
   SK (small conductance Ca2+-sensitive channel)
   IK (intermediate conductance Ca2+-sensitive channel)
   BK (large conductance Ca2+-sensitive channel)
  Ca2+-activated chloride channel
   HCLCA1 (human chloride channel, Ca2+-activated)
Transmembrane protein (TMEM16) family
  TMEM16F A phospholipid scramblase (Module 11: Figure platelet activation)
  TMTM16a and TMEM16b Putative Ca2+-activated Cl channels
Receptor-operated channels (ROCs) See Module 3: Table receptor-operated channel toolkit for further details)
  Nicotinic acetylcholine receptors
  5-HT3 5-Hydroxytryptamine receptor
  AMPA receptors α-Amino-3-hydroxy-5-methylisoxazole-4-propionic acid receptor
  NMDA receptors N-methyl-D-aspartate receptor
  P2X receptors
Second messenger-operated channels (SMOCs) Module 3: Figure Ca2+ entry mechanisms
  Arachidonate-regulated Ca2+ channel (IARC)
Cyclic nucleotide-gated channels (CNGs) Module 3: Figure cyclic nucleotide-gated channels
  CNGA1–CNGA4, CNGB1, CNGB3
Transient receptor potential (TRP) ion channel family Module 3: Figure TRP channel phylogeny
  TRPC1–TRPC7 Canonical
  TRPV1–TRPV6 Vanilloid
  TRPM1–TRPM8 Melastatin
Polycystins Module 3: Figure polycystin domain structure
  PC1
  PC2
Endoplasmic reticulum (ER)/sarcoplasmic reticulum (SR) Ca2+ release channels
InsP3 receptors (InsP3Rs) Module 3: Figure InsP3R structure
  InsP3R1
  InsP3R2
  InsP3R3
Ryanodine receptors (RYRs) Module 3: Figure ryanodine receptor structure
  RYR1
  RYR2
  RYR3
Channel regulators
 Triadin
 Junctin
 Sorcin EF-hand protein that can bind to annexin 7
 FKBP12 FK506-binding protein 12 kDa
 FKBP12.6 FK506-binding protein 12.6 kDa
 Phospholamban Module 5: Figure phospholamban mode of action
 MICU1 Mitochondrial calcium uptake 1 regulates the mitochondrial uniporter
Ca2+ BUFFERS
Cytosolic buffers
 Calbindin D-28k
 Calretinin
 Parvalbumin
Endoplasmic reticulum (ER)/sarcoplasmic reticulum (SR) buffers and chaperones
 Calnexin
 Calreticulin
 Calsequestrin
 GRP78 Glucose-regulatory protein of 78 kDa. Also known as BiP
 GRP94 Glucose-regulatory protein of 94 kDa. Also known as endoplasmin
Ca2+ SENSORS
EF-hand Ca2+-binding proteins
 Calmodulin (CaM) A ubiquitous Ca2+ sensor
 Calcineurin B (CaNB) A sensor that resembles CaM and controls the function of calcineurin (Module 4: Figure calcineurin)
 Calcium and integrin-binding protein 1 (CIB1) Functions to control multiple cellular processes
 Calcium Homeostasis Endoplasmic Reticulum Protein (CHERP) Acts together with ALG-2 to regulate both InsP3Rs and RYRs
 Troponin C (TnC) The Ca2+ sensor in striated muscle
 Miro Mitochondrial Rho-GTPase functions in mitochondrial motility (Module 5: Figure mitochondrial motility)
 DAG kinase α A Ca2+-sensitive enzyme that phosphorylates diacylglycerol (DAG) (Module 2: Figure InsP3/DAG recycling)
 Stromal interaction molecule (STIM) Senses Ca2+ in the ER lumen (Module 3: Figure STIM-induced Ca2+ entry)
 Apoptosis-linked gene 2 (ALG-2) A penta-EF-hand protein that may control ER-to-Golgi trafficking (Module 4: Figure COPII-coated vesicles)
S100 proteins
S100 protein clustered on chromosome 1 S100A1-S100A14 are all clustered on 1q21 (Module 4: Figure S100 phylogenetic tree)
  S100A1
  S100A2
  S100A3
  S100A4
  S100A5
  S100A6
  S100A7
  S100A8
  S100A9
  S100A10
  S100A11
  S100A12
  S100A13
  S100A14
S100 protein clustered on other chromosomes
  S100B
  S100C
  S100P
Neuronal Ca2+ sensor proteins (NCS) A family of 14 EF-hand Ca2+-binding proteins
 NCS-1 Brain and retina
 Hippocalcin Restricted to hippocampal neurons
 Neurocalcin δ Brain and retina
 Recoverin Retina
Visinin-like proteins (VILIPs)
 VILIP-1 Brain and retina
 VILIP-2 Retina
 VILIP-3 Retina and cerebellar Purkinje cells
Gelsolin/villin superfamily
 Adseverin Severs the cortical actin filaments during adrenal secretion in chromaffin cells (Module 7: Figure chromaffin cell secretion)
 CapG
 Gelsolin Modulated actin remodelling (Module 4: Figure actin remodelling)
 Flightless I
 Advillin
 Villin
 Villin-like protein
 Supervillin
Guanylyl cyclase-activating proteins (GCAPs) Module 10: Figure phototransduction
 GCAP1 Retina
 GCAP2 Retina
 GCAP3 Retina
Kv-channel-interacting proteins (KChIPs)
 KChIP1 Brain
 KChIP2 Brain
 KChIP3/downstream regulatory element modulator (DREAM)/calsenilin This protein has three names which are currently in use
 KChIP4 Brain
Ca2+-binding proteins (CaBPs)
 Caldendrin Brain and retina
 L-CaBP1 A long splice variant of caldendrin
 S-CaBP1 A short splice variant of caldendrin
 CaBP2 Retina
 CaBP3 Retina
 CaBP4 Brain and retina
 CaBP5 Retina
 Calneuron-1 The calneurons act to regulate the trafficking function of PtdIns 4-KIIIα at the Golgi
 Calneuron-2
C2 domain Ca2+-binding proteins
Synaptotagmins The Ca2+ sensors that control exocytosis (Module 4: Figure Ca2+-induced membrane fusion)
  Synaptotagmin I
  Synaptotagmin II
  Synaptotagmin III
 Otoferin The Ca2+ sensors that functions in hair cell transmitter release
Annexins Ca2+-dependent phospholipid-binding proteins (Module 4: Figure annexin structure)
  Annexin A1
  Annexin A2
  Annexin A3
  Annexin A4
  Annexin A5
  Annexin A6
  Annexin A7
  Annexin A8
  Annexin A9
  Annexin A10
  Annexin A11
  Annexin A13
Ca2+-SENSITIVE ENZYMES AND PROCESSES
Ca2+-regulated enzymes
Ca2+-dependent protein kinases (CaMKs)
  CaMKI
  CaMKII
  CaMKIII
  CaMKIV
  Myosin light chain kinase (MLCK)
  Phosphorylase kinase
  Ins1,4,5P3 3-kinase
  Proline-rich tyrosine kinase 2 (Pyk2) A Ca2+-sensitive tyrosine kinase that functions in osteoclast podosomes (Module 7: Figure osteoclast podosome)
Lipid kinase
  hVps34 See Module 9: Figure target of rapamycin signalling
   PKC-α
   PKC-βI
   PKC-βII
   PKC-γ
Phosphodiesterases
  Cyclic GMP phosphodiesterase (PDE) Some of the phosphodiesterases are sensitive to Ca2+ see Module 5: Table PDE family properties
   PDE1A
   PDE1B
   PDE1C
Adenylyl cyclases (ACs) See Module 2: Table adenylyl cyclases for details of these Ca2+-sensitive enzymes
   AC-1
   AC-III
   AC-VIII
   AC-V
   AC-VI
Dual oxidases (DUOX1-2) See Module 2: Figure redox signalling components
  Nitric oxide synthase (NOS) Module 2: Figure NO and cyclic GMP signalling
   Endothelial NOS (eNOS)
   Neural NOS (nNOS)
Ca2+-activated proteases
   Calpain I
   Calpain II
Transcription factors
 Nuclear factor of activated T cells (NFAT) Module 4: Figure NFAT activation
  NFATc1
  NFATc2
  NFATc3
  NFATc4
 Cyclic AMP response element-binding protein (CREB) Module 4: Figure CREB activation
 Downstream regulatory element modulator (DREAM)
 CREB-binding protein (CBP)
Ca2+ PUMPS AND EXCHANGERS
Na+/Ca2+ exchanger (NCX) See Module 5: Table Ca2+ pumping toolkit for further details
 NCX1
 NCX2
 NCX3
Na+/Ca2+/-K+ exchanger (NCKX)
 NCKX1
 NCKX2
 NCKX3
 NCKX4
Mitochondrial channels and exchangers Module 5: Figure mitochondrial Ca2+ signalling
 Permeability transition pore (PMT)
 Na+/Ca2+ exchanger
 MCU Mitochondrial Ca2+ uniporter
Plasma membrane Ca2+-ATPases (PMCAs) See Module 5: Table Ca2+ pumping toolkit for further details
 PMCA1
 PMCA2
 PMCA3
 PMCA4
Sarco/endo-plasmic reticulum Ca2+-ATPases (SERCAs) See Module 5: Table Ca2+ pumping toolkit for further details
 SERCA1
 SERCA2
 SERCA3
Secretory-pathway Ca2+-ATPase (SPCA) pumps
 SPCA1 Ubiquitous; located in the Golgi
 SPCA2


The Ca2+ signalling system has a very large toolkit of signalling components. The ways in which the major components are organized are summarized in Module 2: Figure Ca2+ signalling toolkit. Components from this toolkit can be mixed and matched to create a diverse array of cell-specific Ca2+ signalsomes (Module 2: Figure cell-specific Ca2+ signalsomes) that are capable of delivering Ca2+ signals with very different spatial and temporal properties.



Another classical process for releasing internal Ca2+ is the /1inositol 1,4,5-trisphosphate (InsP3)/Ca2+ signalling cassette (Ca2+ module 6 in /1Module 2: Figure Ca2+ modules). Other Ca2+-mobilizing messengers have been described such as /1sphingosine 1-phosphate (S1P), /1cyclic ADP ribose (cADPR) and /1nicotinic acid–adenine dinucleotide phosphate (NAADP).

The channels responsible for these Ca2+ ON reactions usually have powerful inactivation mechanisms that rapidly curtail the entry or release processes to prevent the cell being swamped with Ca2+, which can result in cell stress and apoptosis. Once the ON reactions have been curtailed, the /1Ca 2+ OFF reactions rapidly take over to return the activated level of Ca2+ back to its resting level.

Ca2+-induced Ca2+ release (CICR)

A process of Ca2+-induced Ca2+ release (CICR) plays a central role in the way Ca2+ signals are generated. This positive-feedback mechanism whereby Ca2+ triggers its own release has two important functions in cells. It enables Ca2+ entering across the plasma membrane to function as a messenger to release Ca2+ from the internal store (/1Module 2: Figure Ca2+-induced Ca2+ release). This function of CICR was first described in cardiac cells, where the /1CaV1.2 L-type channel provides an influx of trigger Ca2+ that then diffuses into the cell to activate the /1ryanodine receptor 2 (RYR2) (Ca2+ module 5 in /1Module 2: Figure Ca2+ modules). A similar interaction is particularly evident for /1neuronal Ca2+ entry and release channels.

The other main function of CICR is to set up /1intracellular Ca2+ waves where an elevated level of Ca2+ in one region of the cell (the initiation site) propagates throughout the rest of the cytoplasm as a regenerative Ca2+ wave. Waves can progress by recruiting either RYRs or inositol 1,4,5-trisphosphate receptors (InsP3Rs) (/1Module 2: Figure Ca2+-induced Ca2+ release). These Ca2+ waves are made up of elementary Ca2+ events such as the /1sparks and /1puffs produced by the RYRs and InsP3Rs respectively. It is these unitary events that are used to generate the regenerative waves that make up /1global Ca2+ signals.

Ca2+ OFF reactions

Cells use a variety of mechanisms to remove Ca2+ from the cytoplasm (/1Module 2: Figure Ca2+ signalling dynamics). The introduction of Ca2+ into the cell during the /1Ca2+ ON reactions usually occurs for a relatively brief period during which there is a rapid increase in the intracellular concentration of Ca2+. In fact, the increase in free Ca2+ that can be measured in the cytoplasm using aequorin or fluorescent indicators is a very small proportion of the total amount of Ca2+ that enters during the ON reactions. Much of this Ca2+ is rapidly bound by the cytosolic buffers or is taken up by the mitochondria. As the Ca2+ concentration returns to its resting level, Ca2+ leaves the buffers and the mitochondria and is returned to the ER or is pumped out of the cell resulting in a brief Ca2+ transient.

The recovery process thus depends upon a complex interplay between cytosolic Ca2+ buffers, mitochondria and Ca2+ pumps and exchangers on the internal stores and on the plasma membrane (Module 2: Figure Ca2+ signalling dynamics). These Ca2+ OFF reactions operate at different stages during the recovery phase of a typical Ca2+ spike. The buffers and mitochondria operate early, and the Ca2+ pumps and exchangers are responsible for restoring the status quo by pumping Ca2+ out of the cell or back into the ER. These pumps and exchangers operate at different times during the recovery process. The Na+/Ca2+ exchangers have low affinities for Ca2+, but have very high capacities and this enables them to function at the beginning of the recovery process to rapidly remove large quantities of Ca2+. On the other hand, the plasma membrane Ca2+-ATPase (PMCA) and sarco/endo-plasmic reticulum Ca2+-ATPase (SERCA) pumps have lower capacities, but their higher affinities mean that they can complete the recovery process and can continue to pump at lower Ca2+ levels, thus enabling them to maintain the internal stores and the resting level (Module 5: Figure Ca2+ uptake and extrusion).

Some of these OFF reactions interact with each other during the recovery period, and this is particularly evident in the case of the /1ER/mitochondrial Ca2+ shuttle (/1Module 5: Figure ER/mitochondrial shuttle). When Ca2+ is released from the ER, Ca2+ is rapidly taken up by the mitochondria and this is then released slowly back to the ER once Ca2+ has returned to the resting level.

In addition to its role of returning the level of Ca2+ to its resting level following a stimulus, the Ca2+ OFF reactions are in constant operation to maintain the resting level of Ca2+ within a fairly narrow range. This calciostat is a dynamic system in that the OFF reactions are operating continuously to reverse the constant basal rate of Ca2+ entry through various ‘Ca2+ leak pathways’ that remain poorly defined. There is increasing evidence that /1presenilins may function as such a leak channel and this has been incorporated into the /1calcium hypothesis of Alzheimer's disease (see step 10 in /1Module 12: Figure amyloid processing and Ca2+ signalling). The latter can result in an elevation of cytosolic Ca2+ if the OFF reactions are inhibited. For example, agents such as thapsigargin, which inhibit the SERCA pumps on the ER, will result in an increase in Ca2+ concentration. Indeed, the rate at which the Ca2+ concentration rises following pump inhibition provides a measure of the activity of these leak pathways.

In pathological situations, the lack of oxygen reduces the supply of energy, thus compromising the function of the OFF reactions, resulting in the rise of Ca2+ that is so damaging during stroke or cardiac ischaemia. Under normal circumstances, however, the fully energized OFF reactions can rapidly reduce the pulse of Ca2+ introduced by the ON mechanisms, thus generating a brief Ca2+ transient (/1Module 2: Figure Ca2+ transient mechanisms). Such brief pulses of Ca2+ are a characteristic feature of the Ca2+ signalling pathway and contribute to the /1spatiotemporal aspects of Ca 2+ signalling.

Ca2+ buffers

Cells express a large number of Ca2+-binding proteins, which fall into two main groups: Ca2+ sensors and Ca2+ buffers. The Ca2+ sensors respond to changes in intracellular Ca2+ by activating some downstream effector process. In a sense, all proteins capable of binding Ca2+ will act as a buffer, and this applies to the sensors. However, the concentration of these sensors is usually rather low, so they have little buffering capacity. The role of Ca2+ buffering is carried out by the other major group of Ca2+-binding proteins. Cells have a number of buffers (Module 2: Table Ca2+ signalling toolkit) that are capable of binding to Ca2+ both in the cytoplasm and within the lumen of the endoplasmic reticulum (ER). The sole function of these Ca2+ buffers is to bind Ca2+, and this has an important function in shaping both the spatial and temporal properties of Ca2+ signals.

The major cytosolic buffers in cells are parvalbumin (PV), calbindin D-28k (CB) and calretinin (CR). The major buffers that operate within the lumen of the ER are calsequestrin (CSQ) in the sarcoplasmic reticulum of muscle cells and calreticulin (CRT) in the ER of non-muscle cells. The latter is unusual in that it functions both as a cytosolic and a luminal buffer. The role of calsequestrin is discussed elsewhere and here we concentrate on the function of the cytosolic buffers. The latter have subtly different Ca2+-binding properties and are expressed in cells in differing combinations and concentrations to create Ca2+ signals that are tailored to carry out different functions. For example, neurons such as Purkinje cells express large amounts of PV and CB. As a consequence, Purkinje cells have a large endogenous Ca2+ buffering capacity, e.g. their buffers bind approximately 2000 Ca2+ ions for each free ion. Lower capacities of 50–100:1 are found in other cells. Motor neurons have a very low buffering capacity and consequently have large Ca2+ signals in both the soma and dendrites during normal physiological responses, and this makes them particularly susceptible to neurodegeneration. Buffer concentration is one of the important parameters in determining buffer capacity. The other key parameters include affinity for Ca2+ and other metal ions, the kinetics of Ca2+ binding, release and mobility.

Alterations in Ca2+ buffers have been linked to schizophrenia.

Parvalbumin (PV)

Parvalbumin (PV) is a slow-onset buffer. It has relatively low on and off rates, which means that it cannot respond to the rapid onset of most Ca2+ signals. However, PV can soak up Ca2+ once the signal has appeared and thus influences the rate at which Ca2+ recovers. It is strongly expressed in skeletal muscle, where it plays an important role in facilitating the rate of relaxation. In PV−/− mice, there is a slowing in the recovery of the Ca2+ transient. Some of the effects of removing PV are compensated for by an increase in the volume of mitochondria that has a similar ability to PV of removing free Ca2+ from the cytoplasm during the recovery phase.

Reductions in the level of PV have been recorded in schizophrenia.

Calbindin D-28k (CB)

Calbindin D-28k (CB) is one of the major cytosolic buffers, particularly in neurons. It is a fast buffer that can have a major effect on both the spatial and temporal properties of Ca2+ transients. CB thus plays a major role in restricting the size of the elementary Ca2+ events that form around Ca2+ channels. CB is thus of central importance for the ability of neurons to create the highly localized Ca2+ events that occur in spines. The existence of these buffers has enabled neurons to increase the numbers of their synaptic connections, and this neuronal miniaturization greatly enhances the signal processing capacity of the brain. Some support for such a notion emerged from the finding of a remarkable compensatory mechanism whereby the volume of the spines increases markedly in neurons when these buffers are knocked out. In effect, the lack of these buffers reduces the capacity of neurons to miniaturize the Ca2+ component of their signalling systems.

The ability of CB to buffer internal Ca2+ may play an important role in facilitating Ca2+ reabsorption by the kidney tubule (Module 7: Figure kidney Ca2+ reabsorption). In the case of Ca2+ reabsorption by the intestine, the flux of Ca2+ is facilitated by calbindin D-9k (Module 7: Figure intestinal Ca2+ reabsorption).

An important aspect of the calcium hypothesis of Alzheimer's disease is that there is a decrease in the expression of CB, which increases the sensitivity of neurons to the enhanced Ca2+ signals that arise during the onset of Alzheimer's disease (Module 12: Figure amyloid cascade hypothesis).

Calretinin (CR)

Calretinin (CR) is an EF-hand Ca2+-binding protein that is closely related to calbindin D-28k (CB). It has six EF-hand domains all of which can bind Ca2+ except for domain VI. CR is usually located in the cytoplasm, but can also be found localized to specific sites in the cell where it may function as a buffer to regulate microdomains of Ca2+. This Ca2+ buffer is an excellent biochemical marker for certain GABAergic inhibitory interneurons such as the Cajal–Retzius neurons and double bouquet neurons in the dorsolateral prefrontal cortex (DLPFC) (Module 10: Figure dorsolateral prefrontal cortex).

Calreticulin (CRT)

Calreticulin (CRT) is a low-affinity Ca2+-binding protein that is located within the lumen of the endoplasmic reticulum (ER). It has also been detected in the nucleus and cytoplasm and it may also be secreted into the extracellular environment during periods of cell stress. Its primary location, however, is within the lumen of the ER, where it functions both as a chaperone protein and as the major ER Ca2+ buffer. CRT has three main domains; there is a globular N-domain of unknown function followed by a proline-rich P-domain, which has sequences unique to CRT and its homologous proteins calnexin and calmegin. Finally, there is a highly acidic C-terminal domain, which has a large number of low-affinity Ca2+-binding sites capable of binding 20–30 mol of Ca2+/mol of protein. This C-terminal region terminates in the KDEL sequence responsible for retaining CRT in the lumen of the ER.

The chaperone function of CRT is carried out in conjunction with a related chaperone calnexin. A calnexin/calreticulin cycle ensures the correct folding and subunit assembly of glycoproteins and is thus essential for protein trafficking and secretion. The proper folding and assembly of proteins is very dependent on a constant level of Ca2+ within the ER lumen and the chaperones. The ability of CRT to bind large amounts of Ca2+ enables it to function as an ER Ca2+ buffer to help maintain this constancy of ER Ca2+. In addition to functioning as a passive buffer, CRT may play a more direct role in maintaining homoeostatic control of their working environment by modulating the activity of the Ca2+ channels and pumps. When the Ca2+ level within the lumen gets too high, CRT and calnexin inhibit the sarco/endo-plasmic reticulum Ca2+-ATPase (SERCA) pump and, when Ca2+ is released, this inhibition is relieved and the SERCA pumps can restore Ca2+ to its normal level. In addition to its role as an ER buffer, CRT may play an active role in ensuring that the Ca2+ concentration within the ER lumen is maintained at the optimal level for protein folding to occur.

Ca2+ signalling function

The function of Ca2+ as an intracellular second messenger is carried out by a combination of /1Ca2+ sensors and /1Ca2+ effectors (/1Module 2: Figure Ca2+ signalling toolkit). The major sensors are the EF-hand proteins /1troponin C (TnC), /1calmodulin (CaM), /1neuronal Ca2+ sensor proteins (NCS) and the /1S100 proteins. However, there are a number of other sensors (/1Module 2: Table Ca2+ signalling toolkit).

These sensors are then responsible for relaying information through a range of effectors:

Spatiotemporal aspects of Ca2+ signalling

The use of Ca2+ as a universal signal for cell regulation is somewhat paradoxical because this ion can be very toxic to cells if its level remains high for a prolonged period. Such toxicity is avoided by presenting Ca2+ signals in a pulsatile manner (/1Module 2: Figure temporal aspects). In addition to this temporal aspect, the Ca2+ signal is also highly organized in space. The /1elementary and global aspects of Ca2+ signalling greatly increase the versatility of the Ca2+ signalling system in that it can act either locally or globally. For example, muscle contraction is activated by a global elevation in Ca2+, whereas the release of neurotransmitters results from a minute punctate pulse of Ca2+ delivered directly to the docked vesicle by a Ca2+ sensor tightly associated with exocytotic machinery (/1Module 4: Figure Ca2+-induced membrane fusion). In between these two extremes, there are many variations in the way the Ca2+ signal is presented to cells. Perhaps the most dramatic are the Ca2+ waves that initiate at fixed localities and then process through the cytoplasm in a regenerative manner through the process of CICR (/1Module 2: Figure Ca2+-induced Ca2+ release).

These spatiotemporal aspects greatly enhance the versatility of Ca2+ signalling, thus providing the flexibility to regulate so many cellular processes.

Cyclic ADP-ribose (cADPR) signalling

Cyclic ADP-ribose (cADPR) is one of the messengers associated with the NAD+ signalling pathways. cADPR has attracted considerable attention as a putative messenger to regulate the Ca2+ signalling pathway. cADPR generation and metabolism is unusual in that it is carried out by the same enzyme that possesses both synthase and hydrolase activity. The cADPR control of Ca2+ release is somewhat controversial, as its precise mode of action is still unclear. A cADPR working hypothesis has been put forward to provide a framework to understand some of the current information on how this messenger appears to operate. A relationship between cADPR and cell regulation has been established in a number of different cell types.

cADPR working hypothesis

There is sufficient evidence to take seriously the possibility that cADPR functions to regulate Ca2+ signalling. What seems to be in question is exactly how it functions. This working hypothesis has two main components:

  • Firstly, the generation of cADPR is closely coupled to cellular metabolism, as described in cADPR generation and metabolism (Module 2: Figure cADPR/NAADP function). The idea is that cADPR may function as a metabolic messenger responsible for relaying information about the state of metabolism to various systems in the cell, especially those that require a heavy expenditure of energy. An example would be Ca2+ signalling and the downstream elements regulated by Ca2+. When energy is abundant, an increase in cADPR will set the stage for Ca2+ signalling to occur.
  • Secondly, the cADPR control of Ca2+ release may occur indirectly through an activation of the sarco/endo-plasmic reticulum Ca2+-ATPase (SERCA) pump to increase the uptake of Ca2+ into the endoplasmic reticulum/sarcoplasmic reticulum. This increased loading of the internal store will serve to sensitize release channels such as the RYRs, thus leading to an increase in Ca2+ signalling. This messenger should thus be considered as a modulator rather than a mediator of Ca2+ signalling.

cADPR generation and metabolism

The generation and metabolism of cADPR are described together because both processes are carried out by the same enzyme: the ADP-ribosyl cyclase (Module 2: Figure cADPR metabolism). The hydrolysis of cADPR produces ADPR, which has been implicated as a messenger regulating melastatin-related transient receptor potential 2 (TRPM2), which is a Ca2+ channel in the plasma membrane (Module 2: Figure cADPR/NAADP function). This is a highly versatile enzyme. In addition to synthesizing and metabolizing cADPR, it is also responsible for synthesizing NAADP. In mammals, this bifunctional enzyme appears to be the lymphocyte antigen CD38, which is expressed widely and is located both in the plasma membrane and at internal sites. With regard to the former location, its enzymatic region is located on the outside. This is unusual because its substrate and its site of action are on the inside of the cell. It has been suggested that the enzyme might use NAD+ derived from dying cells at sites of infections to generate cADPR, which is then transported into the cell.

The intracellular enzyme is more likely to be the one that functions in most cells. Just how the cyclase is activated is still unclear. One suggestion is that the formation of cADPR and NAADP is sensitive to cellular metabolism (Module 2: Figure cADPR/NAADP function). In other words, cADPR and NAADP might be metabolic messengers that are capable of relaying information about the state of cellular metabolism to the Ca2+ signalling pathways. Such a notion is supported by the fact that cADPR metabolism by the hydrolase is inhibited by either ATP or NADH. Another suggestion is that it might be activated by agonists acting through cell-surface receptors, but the coupling mechanism remains to be established (Module 2: Figure cADPR/NAADP function). This absence of a coupling mechanism might be explained by the cADPR working hypothesis if the external agonist enhanced cADPR formation indirectly by first increasing cellular metabolism. Such a mechanism could explain the ability of β-adrenergic agents to increase cADPR levels in heart. Likewise, the glucose-dependent increase in cADPR in β-cells can be directly linked to the metabolism of glucose, with the resulting increase in ATP acting to reduce the hydrolase activity (Module 2: Figure cADPR/NAADP function).

In CD38−/− mice, there is a decrease in the amount of oxytocin (OT) released from hypothalamic neurons. Such mice show defects in maternal nurturing and in social behaviour.

cADPR control of Ca2+ release

One of the major uncertainties about cADPR is its mode of action in controlling the release of Ca2+. There have been suggestions that it is a Ca2+-mobilizing second messenger that acts by stimulating the ryanodine receptors (RYRs) to release Ca2+. However, direct evidence for this assertion is not particularly convincing. Early single channel recordings seemed to provide such evidence by showing that cADPR could open RYRs in lipid membranes, but these observations were challenged on the basis that the cADPR was acting through the ATP-binding site. When cADPR is injected into cells, it usually fails to release Ca2+, but after a period of time, it can begin to enhance the sensitivity of the RYRs. An example of such an effect is shown in heart cells, where there is a gradual increase in spark frequency following addition of cADPR (Module 2: Figure cADPR action in heart cells). This observation on cardiac cells forms the basis of the second part of the cADPR working hypothesis, which argues that cADPR acts indirectly as a modulator of Ca2+ release (Module 2: Figure cADPR/NAADP function) by stimulating the sarco/endo-plasmic reticulum Ca2+-ATPase (SERCA) pump to increase the load of Ca2+ within the lumen of the store. This increase in luminal Ca2+ then sensitizes the RYRs so that they either begin to open spontaneously to give Ca2+ sparks (Module 2: Figure cADPR action in heart cells) or begin to produce larger Ca2+ transients when activated by a pulse of trigger Ca2+ as occurs in neurons (Module 2: Figure cADPR action in neurons).

cADPR and cell regulation

A role for cADPR has been implicated in a number of different cell types:

Nicotinic acid–adenine dinucleotide phosphate (NAADP) signalling

NAADP is one of the messengers associated with the NAD signalling pathways. NAADP has attracted considerable attention as a Ca2+-mobilizing second messenger. NAADP generation and metabolism is unusual in that it is closely allied to that of cADPR in that the same enzyme synthesizes both messengers. The NAADP control of Ca2+ release depends upon this messenger acting on a store that is different from that controlled by the inositol 1,4,5-trisphosphate receptors (InsP3Rs) or the ryanodine receptors (RYRs). A relationship between NAADP and cell regulation seems to depend upon its interaction with the other Ca2+-mobilizing messenger systems.

NAADP generation and metabolism

NAADP (Module 2: Figure NAADP structure) is another member of the NAD signalling pathways. Its synthesis from NADP is closely linked to that of cADPR in that they both share the same enzyme: the ADP-ribosyl cyclase (Module 2: Figure cADPR/NAADP function). This enzyme uses NAD+ to make cADPR through a cyclization reaction, or it can use NADP as a substrate to produce NAADP through a base-exchange reaction during which the nicotinamide group is exchanged for nicotinic acid.

Unlike cADPR, which is hydrolysed by the same enzyme, NAADP is degraded to NAAD by phosphatases such as alkaline phosphatase.

NAADP control of Ca2+ release

NAADP functions as a Ca2+-mobilizing messenger to release Ca2+ from an internal store (Module 2: Figure cADPR/NAADP function). This NAADP-sensitive store, which is distinct from the endoplasmic reticulum/sarcoplasmic reticulum store that is regulated by the inositol 1,4,5-trisphosphate receptors (InsP3Rs) and the ryanodine receptors (RYRs), is still being defined. On the basis of studies carried out in sea urchin eggs, this novel store appears to reside in a lysosome-related organelle. The receptors for NAADP appear to be a family of two-pore channels (TPC), with TPC1 being located on the endosomal membranes, whereas TPC2 is on lysosomal membranes.

NAADP and cell regulation

The amount of Ca2+ released by NAADP is relatively small and is unlikely to have a direct role in Ca2+ signalling. However, there are suggestions that it may act indirectly to trigger the release of Ca2+ by the other release channels [ryanodine receptors (RYRs) or the inositol 1,4,5-trisphosphate receptors (InsP3Rs) (Module 2: Figure cADPR/NAADP function)].

NAADP has been implicated in the control of insulin release by β-cells (Module 7: Figure β-cell signalling) and in the control of secretion by pancreatic acinar cells (Module 7: Figure control of pancreatic secretion).

ADP-ribosyl cyclase

ADP-ribosyl cyclase is the enzyme responsible for synthesizing the Ca2+ mobilizing messengers cADPR and NAADP. During cADPR generation and metabolism it uses NAD+ to make cADPR through a cyclization reaction (Module 2: Figure cADPR/NAADP function). The same enzyme is also responsible for the hydrolysis of cADPR to ADPR. This enzyme can also use NADP as a substrate to produce NAADP through a base-exchange reaction during which the nicotinamide group is exchanged for nicotinic acid.

In mammals, there are two ribosyl cyclases: CD38 and bone marrow stromal cell antigen-1 (BST-1). Most is known about CD38, which takes its name from the fact that it was originally identified as a lymphocyte antigen CD38. This bifunctional enzyme is expressed widely and is located both in the plasma membrane and at internal sites.

Phosphoinositide signalling

Signalling through the phosphoinositide lipids is complex because there are a number of signalling cassettes associated with both the synthesis and hydrolysis of the phosphoinositides. Phosphoinositide metabolism can be divided into two main parts. Firstly, there is inositol lipid metabolism, which describes the way in which the parent molecule PtdIns is metabolized to form a number of lipid intermediates, some of which are key elements in different signalling cassettes. Secondly, there is inositol phosphate metabolism, which is a complex pathway responsible for metabolizing soluble inositol phosphates. Some of the inositol phosphates generated by this metabolism have been implicated as messengers operating within the multipurpose inositol polyphosphate signalling pathway.

PtdIns4,5P2 is of particular interest because it is the precursor used to generate the second messengers inositol 1,4,5-triphosphate (InsP3) and diacylglycerol (DAG) that function in the inositol 1,4,5-trisphosphate (InsP3)/Ca2+ signalling cassette and the diacylglycerol (DAG)/protein kinase C (PKC) signalling cassette respectively. PtdIns4,5P2 is also a precursor that is metabolized to PtdIns3,4P2 and PtdIns3,4,5P3, which function as second messengers to activate protein kinases in the PtdIns 3-kinase signalling pathway. In addition to being a precursor for these various signalling pathways, PtdIns4,5P2 can also function as a messenger for the PtdIns4,5P2 signalling cassette, where localized synthesis or hydrolysis of this lipid functions to regulate a number of cellular systems such as the actin cytoskeleton, membrane trafficking, exocytosis, ion channels and exchangers.

Alterations in the activity of this phosphoinositide lipid signalling pathway have been implicated in various diseases such as manic-depressive illness, Lowe's oculocerebrorenal (OCRL) syndrome and Cowden's disease.

Phosphoinositide metabolism

Phosphoinositide metabolism can be separated into two main components (/1Module 2: Figure phosphoinositide metabolism):

These two pathways are connected by the processes of lipid hydrolysis and lipid synthesis. The hydrolysis occurs when external signals activate /1phospholipase C (PLC) to hydrolyse PtdIns4,5P2 to form the second messengers inositol 1,4,5-trisphosphate (Ins1,4,5P3) and diacylglycerol (DAG). The Ins1,4,5P3 is one of the major inputs into the inositol phosphate metabolic pathway that produces the inositol required for lipid synthesis.

This complex phosphoinositide metabolic network participates in a number of highly versatile signalling cassettes (Module 2: Figure phosphoinositide signalling systems):

Inositol lipid metabolism

The different phosphoinositide signalling pathways are derived from the parent molecule phosphatidylinositol (PtdIns) (/1Module 2: Figure PtdIns structure). Unlike the other phospholipids found in cellular membranes, PtdIns is unique in that the free hydroxy groups at the 3-, 4- and 5-positions on the inositol ring can be phosphorylated further to create a family of phosphoinositides (/1Module 2: Figure phosphoinositide metabolism). One of the more important lipids involved in phosphoinositide signalling is phosphatidylinositol 4,5-bisphosphate (PtdIns4,5P2), which is a nodal point for a number of signalling systems (/1Module 2: Figure phosphoinositide signalling systems).

The metabolism of PtdIns is carried out by a collection of /1inositol lipid kinases and /1inositol lipid phosphatases. The numbers below refer to the different reactions shown in /1Module 2: Figure phosphoinositide metabolism:

  1. The Class III /1PtdIns 3-kinase (PtdIns 3-K) adds a phosphate to the 3-position of PtdIns to form PtdIns3P.
  2. /1PtdIns 4-kinase (PtdIns 4-K) adds a phosphate to the 4-position of PtdIns to form PtdIns4P. It can also be formed by removal of the 3-phosphate from PtdIns3,4P2 by /1phosphatase and tensin homologue deleted on chromosome 10 (PTEN).
  3. The /1PIKxyve kinase is thought to be responsible for adding a phosphate to the 5-position of PtdIns to form PtdIns5P.
  4. The PIKfyve also adds a phosphate to the 5-position of PtdIns3P to form PtdIns3,5P2. The formation of PtdIns3,5P2, which changes in cells following osmotic stress, is a key component of the /1PtdIns3,5P2 signalling cassette (/1Module 2: Figure PIKfyve activation).
  5. Types I and II /1PtdIns 3-kinase (PtdIns 3-K) add a phosphate to the 3-position of PtdIns4P to form PtdIns3,4P2. PtdIns3,4P2 has been suggested to function as a messenger operating within the plasma membrane, where it serves to recruit and activate protein kinases such as /1protein kinase B (PKB). PtdIns3,4P2 can also be formed by other enzymes (see Steps 6 and 7).
  6. /1PtdInsP kinase II adds a phosphate to the 4-position of PtdIns3P to form PtdIns3,4P2. The same enzyme can add a phosphate to the 4-position of PtdIns5P to form PtdIns4,5P2.
  7. /1Inositol polyphosphate 5-phosphatase such as Src homology 2 (SH2) domain-containing inositol phosphatase (SHIP) removes a phosphate from the 5-position of PtdIns 3,4,5P3 to form PtdIns3,4P2.
  8. A /1PtdIns4P 5-kinase (PtdIns4P 5-K) adds a phosphate to the 5-position of PtdIns4P to form PtdIns4,5P2. The latter can also be formed by Steps 6 and 9. PtdIns4P5-K can also phosphorylate PtdIns3,4P2 to form PtdIns3,4,5P3.
  9. /1Phosphatase and tensin homologue deleted on chromosome 10 (PTEN) and transmembrane phosphoinositide 3-phosphatase and tensin homologue 2 (TPTE2) removes a phosphate from the 3-position of PtdIns3,4,5P3.
  10. A Class I PtdIns 3-kinase adds a phosphate to the 3-position of PtdIns4,5P2 to form PtdIns3,4,5P3. PtdIns3,4,5P3, which functions as one of the second messengers in the /1PtdIns 3-kinase signalling pathway, can also be formed by the addition of a phosphate to the 5-position of PtdIns3,4P2 by the /1PtdIns4P 5-kinase (PtdIns4P 5-K) (i.e. the same enzyme used for Step 8).
  11. The /1myotubularins remove the 3-phosphate from PtdIns3P and also from PtdIns3,5P2.
  12. The 5-phosphate on PtdIns4,5P2 is removed by various lipid phosphatases such as /1oculocerebrorenal syndrome of Lowe (OCRL), /1inositol polyphosphate 5E-phosphatase (INPP5E), inositol polyphosphate-5 phosphatase-like 1 (INPPL1) and the /1synaptojanins.
  13. The 5-phosphate on PtdIns3,5P2 is removed by /1inositol polyphosphate 5E-phosphatase (INPP5E) and by the Sac domain phosphatase /1Sac3, which is an orthologue of yeast Fig4.
  14. The /1inositol polyphosphate 4-phosphatase type II (INPP4B) dephosphorylates PtdIns3,4P2 to form PtdIns3P.
  15. /1Synaptojanin 1 (SJ1) can remove the phosphates from PtdIns3P and PtdIns4P to produce PtdIns. The suppressor of actin mutation 1-like (SACM1L) can also dephosphorylate PtdIns4P to regulate Golgi membrane morphology and the mitotic spindle.


Module 2: Table inositol lipid kinases and phosphatases Inositol lipid kinases and phosphatases

Substrate Product Comments
Inositol lipid kinases
Class III PtdIns 3-kinase PtdIns PtdIns3P Also known as hVps34 (Step 1 in /1Module 2: Figure phosphoinositide metabolism). Functions in protein trafficking (/1Module 2: Figure localized inositol lipid signalling)
Phosphatidylinositol 4-kinase (PtdIns 4-K) Step 2 in /1Module 2: Figure phosphoinositide metabolism.
 PtdIns 4-KIIα PtdIns PtdIns4P
 PtdIns 4-KIIβ PtdIns PtdIns4P
 PtdIns 4-KIIIα PtdIns4P
 PtdIns 4-KIIIβ PtdIns4P
Phosphatidylinositol 5-kinase (PtdIns 5-K) PtdIns PtdIns5P Step 3 in /1Module 2: Figure phosphoinositide metabolism.
Class II PtdIns 3-kinase PtdInsPtdIns4P PtdIns3PPtdIns3,4P2 Step 5 in /1Module 2: Figure phosphoinositide metabolism.
Phosphatidylinositol4P 5-kinase (PtdIns4P 5-K) Type I Step 8 in /1Module 2: Figure phosphoinositide metabolism.
 PtdIns4P 5-KIα PtdIns4P PtdIns4,5P2
 PtdIns4P 5-KIβ PtdIns4P PtdIns4,5P2
 PtdIns4P 5-KIγ PtdIns4P PtdIns4,5P2
PhosphatidylinositolP kinase (PtdInsP 4-K) Type II PtdIns3PPtdIns5P PtdIns3,4P2PtdIns4,5P2 Step 6 in /1Module 2: Figure phosphoinositide metabolism.
PhosphatidylinositolP 5-kinase (PIKfyve) Type III PtdIns3P PtdIns3,5P2 Step 4 in /1Module 2: Figure phosphoinositide metabolism. Functions in protein trafficking (/1Module 2: Figure localized inositol lipid signalling). Mutations of this kinase are associated with Francois–Neetens mouchetee fleck corneal dystrophy (CFD)
Class I PtdIns 3-kinases Step 10 in /1Module 2: Figure phosphoinositide metabolism.
Class IA PtdIns 3-kinase The structure of these lipid kinases is shown in /1Module 2: Figure PI3-K family
 p110α (catalytic subunit) PtdIns4,5P2 PtdIns3,4,5P3
 p110β (catalytic subunit) PtdIns4,5P2 PtdIns3,4,5P3
 p110δ (catalytic subunit) PtdIns4,5P2 PtdIns3,4,5P3
 p85α (regulatory subunit)
 p85β(regulatory subunit)
 p55α(regulatory subunit)
 p50(regulatory subunit)
 p55γ(regulatory subunit)
Class IB PtdIns 3-kinase
 p110γ PtdIns4,5P2 PtdIns3,4,5P3
 p84 (regulatory subunit)
 p101 (regulatory subunit)
Class II PtdIns 3-kinase
 PtdIns 3KC2α PtdIns4,5P2 PtdIns3,4,5P3
 PtdIns 3KC2β PtdIns4,5P2 PtdIns3,4,5P3
 PtdIns 3KC2γ PtdIns4,5P2 PtdIns3,4,5P3
Inositol lipid phosphatases
Src homology 2 (SH2) domain-containing inositol phosphatase (SHIP) Step 7 in /1Module 2: Figure phosphoinositide metabolism.
 SHIP1 PtdIns3,4,5P3 PtdIns3,4,P2
 SHIP2 PtdIns3,4,5P3 PtdIns3,4,P2
Phosphatase and tensin homologue deleted on chromosome 10 (PTEN) PtdIns3,4P2 PtdIns4P Step 9 in /1Module 2: Figure phosphoinositide metabolism.
PtdIns3,4,5P3 PtdIns4,5P2 This lipid phosphatase is mutated in many human cancers
Myotubularins (MTMR1–MTMR13) Step 11 in /1Module 2: Figure phosphoinositide metabolism.
 MTM1 PtdIns3 PtdIns Mutations of MTM1 are responsible for X-linked recessive myotubular myopathy
 MTM2 PtdIns3 PtdIns MTMR2 is mutated in Charcot–Marie–Tooth disease 4B
Synaptojanin 1 (SJ1) PtdIns4,5P2 PtdIns4P Step 8 in /1Module 2: Figure phosphoinositide metabolism.
Synaptojanin 1 (SJ1) PtdIns4,5P2 PtdIns4P Removes protein coates following scission of endocytic vesicles (/1Module 4: Figure scission of endocytic vesicles)
Oculocerebrorenal syndrome of Lowe (OCRL) PtdIns4,5P2 PtdIns4P Mutations of OCRL cause Lowe's syndrome
PtdIns3,4,5P3 PtdIns3,4P2
Sac3 PtdIns3,5P2 PtdIns3P Mammalian homologue of yeast Fig4 is part of the PAS complex (/1Module 2: PIKfyve activation)



Inositol lipid kinases

There are a number of inositol lipid kinases that function to phosphorylate hydroxyls on the inositol headgroup of phosphatidylinositol.

PtdIns 3-kinase (PtdIns 3-K)

The PtdIns 3-kinases (PtdIns 3-Ks) are a family of enzymes that have been classified into three classes (/1Module 2: Figure PI 3-K family).

Class I PtdIns 3-kinases

The primary function of the Class I PtdIns 3-kinases is to phosphorylate PtdIns4,5P2 to form the lipid second messenger PtdIns3,4,5P3 (/1Module 2: Figure PtdIns 3-kinase signalling). These Class I enzymes are divided into two groups: five regulatory subunits and three catalytic subunits (/1Module 2: Figure PI 3-K family).

The Class IA enzymes are heterodimers that are formed from a regulatory subunit combining with a catalytic subunit. There are five regulatory subunits (p85α, p85β, p55α, p50α and p55γ), which are typical adaptor proteins. The p85 regulatory subunit contains an Src homology 3 (SH3) domain, a breakpoint-cluster-region homology (BH) domain, two proline-rich regions (P) on either side of the BH domain and two Src homology (SH2) domains that are separated by a p110-binding domain, enabling them to interact with a p85-binding domain located on the three catalytic subunits (p110α, p110β and p110δ). The different binding domains on the regulatory subunits enhance the versatility of the Class IA enzyme by enabling the catalytic subunit to be activated by interacting with a variety of signalling molecules:

The Class IB enzyme has a different activation mechanism in that it is stimulated by heterotrimeric G proteins (Module 2: Figure PtdIns 3-kinase signalling). It has a p110γ catalytic subunit and two regulatory subunits (p101 and p84) (Module 2: Figure PI 3-K family). These regulatory subunits may mediate the translocation of the p110γ catalytic subunit to the membrane by binding to the Gβγ subunit.

The Class I enzymes interact strongly with the oncogene Ras (Module 2: Figure PtdIns 3-kinase signalling) and this can have important consequences for both upstream and downstream events, depending on the cellular context. With regard to the former, the binding of Ras can enhance enzyme activation and the generation of lipid messengers. On the other hand, the association with Ras might mediate some of the downstream effects of the PtdIns 3-kinases.

Class I PtdIns 3-kinases are strongly inhibited by the fungal metabolite wortmannin (an irreversible inhibitor) and by LY294002 (a reversible inhibitor).

Class II PtdIns 3-kinase

The Class II PtdIns 3-kinase has a more restricted substrate range in that it phosphorylates just PtdIns and PtdIns4P. There are no regulatory subunits, and the three catalytic subunits (PI3KC2α, PI3KC2β and PI3KC2γ) are characterized by having a C-terminal phox homology (PX) domain and a C2 domain (Module 2: Figure PI 3-K family). The PI3KC2α plays an important role in the maturation of clathrin-coated pits during endocytosis (Module 2: Figure localized inositol lipid signalling).

Class III PtdIns 3-kinase

The Class III PtdIns 3-kinase has a restricted substrate range in that it phosphorylates just PtdIns to form PtdIns3P. The prototype of this class is vacuolar protein sorting 34 (Vps34) protein found in Saccharomyces cerevisiae of which there is a human homologue (hVps34). The activity of hVps34 is regulated by Ca2+ and this contributes to the regulation of the target of rapamycin (TOR), which functions in cell growth control (see steps 4 and 5 in Module 9: Figure target of rapamycin signalling). There is a Ca2+-dependent CaM-binding domain located between residues 318 and 334 in the accessory domain (Module 2: Figure PI3-K family).

The hVps34 plays an important role in intracellular trafficking such as the early endosome to plasma membrane trafficking (Module 4: Figure early endosome budding), early endosome protein sorting and intraluminal vesicle formation (Module 4: Figure intraluminal endosomal vesicle formation) and endosome vesicle fusion to early endosomes (Module 4: Figure endosome vesicle fusion). In the case of autophagy, hVps34 plays a central role in driving the early induction, nucleation and elongation steps responsible for forming the autophagic vacuole (Module 11: Figure autophagy signalling mechanisms).

PtdIns 4-kinase (PtdIns 4-K)

PtdIns 4-kinase catalyses phosphorylation of PtdIns on the 4-position of the inositol ring to produce PtdIns4P (Step 2 in Module 2: Figure phosphoinositide metabolism). Enzyme activity is found in most cellular membranes (plasma membrane, endoplasmic reticulum, Golgi, nuclear membrane and secretory vesicles). Cells have two classes of PtdIns 4-K: Type II and Type III.

The Type II PtdIns 4-K, which contains wortmannin-insensitive α- and β-isoforms, may be responsible for generating the lipid substrates in the membrane that are used for signalling. The enzymatic activity of PtdIns 4-kinase is greatly enhanced in various cancer cells.

The Type III PtdIns 4-K, which has α- and β-isoforms, is sensitive to the drug wortmannin. The PtdIns 4-KIIIα found in the endoplasmic reticulum and the PtdIns 4-KIIIβ found in the cytosol and Golgi are responsible for producing PtdIns4P, which is one of the localized phosphoinositide lipid signalling molecules used for vesicle trafficking (Module 2: Figure localized inositol lipid signalling). The PtdIns 4-KIIIβ at the Golgi is controlled by the Arf signalling pathway (Module 2: Figure Arf signalling). It is the Type III PtdIns 4-K isoform that is activated by the neuronal Ca2+ sensor-1 protein (NCS-1) to enhance exocytosis. Originally, it was considered that PtdIns4P functioned solely as a precursor for the formation of PtdIns4,5P2, but there is increasing evidence for a role as part of a PtdIns4P signalling cassette.

PtdIns phosphate kinases

There is a family of PtdIns phosphate kinases that are divided into three subfamilies (Types I–III) that are localized to different compartments and function to phosphorylate different phosphoinositides.

PtdIns4P 5-kinase (PtdIns4P 5-K)

This versatile family of Type I PtdIns phosphate kinases (also abbreviated as PIPKI) is composed of three isoforms (α, β and γ) that are located in different cellular regions such as the plasma membrane, focal adhesions, Golgi and nucleus. Their primary role is to phosphorylate PtdIns4P to PtdIns4,5P2 (Step 8 in Module 2: Figure phosphoinositide metabolism).

PtdIns4P 5-kinase Iα

This enzyme functions at multiple locations in the cell. It is the major enzyme responsible for forming PtdIns4,5P2 at the plasma membrane. It is also found within the nucleus at nuclear speckles that are sites of mRNA processing.

PtdIns4P 5-kinase Iβ

This isoform is found on membranes surrounding the nucleus.

PtdIns4P 5-kinase Iγ

The PtdIns4,5P2 function in focal adhesions depends upon a spliced form of this enzyme, which is targeted to focal adhesions by interacting with talin (Module 6: Figure integrin signalling). Talin contains a FERM domain, which is used to bring PtdIns4P 5-kinase Iγ into the adhesion complex. Similarly, this isoform plays a major role the control of endocytosis (Module 4: Figure endocytosis).

This Type I enzyme can be regulated by two separate mechanisms. Firstly, its activity is stimulated by phosphatidic acid (PA) that is formed either by the /1phospholipase D (PLD) signalling pathway where phosphatidylcholine (PC) is hydrolysed by phospholipase D (/1Module 2: Figure PLD signalling) or by the phosphorylation of diacylglycerol (DAG) by /1diacylglycerol kinase (DAG) kinase (/1Module 2: Figure InsP3/DAG recycling). Secondly, enzyme activity can be enhanced by small /1monomeric G proteins such as Rho (/1Module 2: Figure Rho signalling). During its activation, PtdIns4P 5-kinase Iα translocates to the plasma membrane through an activation mechanism that is dependent upon Rac and Rho. This G protein regulation results in the localized synthesis of PtdIns4,5P2, which then acts as a second messenger for the /1PtdIns4,5P2 signalling cassette. For example, it can influence a variety of proteins to regulate processes such as /1actin remodelling, /1exocytosis, the PLD signalling pathway (/1Module 2: Figure PLD signalling) and the permeability of ion channels.

The PtdIns4P 5-Kα has an additional role in that it can phosphorylate PtdIns3,4P2 to form the lipid second messenger PtdIns3,4,5P3 (Step 8 in /1Module 2: Figure phosphoinositide metabolism).

PtdInsP kinase II

This enzyme, which is found in the cytosol, endoplasmic reticulum and nucleus, but not in the plasma membrane, is a PtdInsP 4-kinase (PIP4K) capable of phosphorylating both PtdIns3P and PtdIns5P to produce PtdIns3,4P2 and PtdIns4,5P2 respectively (Steps 6 in /1Module 2: Figure phosphoinositide metabolism).

PtdInsP kinase III (PIKfyve)

This type III PtdInsP kinase, which is known as PhosphoInositide Kinase for five position containing a Fyve finger (PIKfyve), phosphorylates PtdIns3P to PtdIns3,5P2 (Step 4 in Module 2: Figure phosphoinositide metabolism). Despite being classified as a PtdInsP kinase, this kinase may also phosphorylate PtdIns on the 5-position to form PtdIns5P (see Step 3). This PIKfyve kinase has a FYVE domain to target it to endomembranes where it functions in the PtdIns3,5P2 signalling cassette (Module 2: Figure PIKfyve activation).

Mutations in PIKfyve have been linked to François–Neetens Mouchetée fleck corneal dystrophy (CFD).

Inositol lipid phosphatases

There are a large number of inositol lipid phosphatases that contribute to inositol lipid metabolism (Module 2: Figure phosphoinositide metabolism). The situation is complicated in that some of these phosphatases have a broad range of substrates and can act on both inositol lipids and inositol phosphates. The following are the phosphatases that act predominantly on inositol lipids:

5-Phosphatase II

5-Phosphatase II, which is also known as inositol polyphosphate 5B-phosphatase (INPP5B), is a 75 kDa protein that preferentially hydrolyses PtdIns4,5P2 and PtdIns3,4,5P3, but can also act on the inositol phosphates Ins1,4,5P3 and Ins1,3,4,5P4. This 5-phosphatase II closely resembles oculocerebrorenal syndrome of Lowe (OCRL) both in structure and function. Like OCRL, it has a central 5-phosphatase catalytic domain followed by an ASH domain and a RhoGAP domain. The ASH domain enables it to interact with Rab5. The C-terminal end has a CAAX region that can be prenylated to facilitate its membrane localization.

5-Phosphatase II is located on the Golgi where it may function in the trafficking of vesicles to both the ER and to the early endosome.

Inositol polyphosphate 5E-phosphatase (INPP5E)

Inositol polyphosphate 5E-phosphatase (INPP5E), which is also known as Type IV 5-phosphatase, is widely expressed where it functions to remove the 5-phosphate from PtdIns4,5P2, PtdIns3,5P2 and PtdIns3,4,5P3 (Module 2: Figure phosphoinositide metabolism). INNP5E plays a vital role in the function of the primary cilium in regulating both development and multiple cellular processes. Mutations in the INNP5E gene have been linked to MORM syndrome.

Phosphatase and tensin homologue deleted on chromosome 10 (PTEN)

PTEN (phosphatase and tensin homologue deleted on chromosome 10) is a lipid phosphatase that is mutated in many human cancers. Its gene is a tumour suppressor, which functions in cell migration, proliferation and survival. PTEN inactivates the 3-phosphorylated lipid second messengers that operate within the PtdIns 3-kinase signalling cassette. It functions to remove the phosphate from the 3-position of PtdIns3,4P2 and PtdIns3,4,5P3 (Module 2: Figure phosphoinositide metabolism).

PTEN is a redox-sensitive enzyme, which may contribute to a positive-feedback loop that enhances redox signalling (Module 2: Figure plasma membrane ROS formation). It is inhibited by H2O2, which induces Cys-124 in the active site to form a disulphide bond with Cys-71. This disulphide bond formation is specifically reversed by thioredoxin. A reversible inactivation of PTEN may thus contribute to the accumulation of PtdIns3,4,5P3 (Module 2: Figure plasma membrane ROS formation), which may thus help to promote proliferation by switching off the metabolism of the 3-phosphorylated lipid second messengers. This reversible inactivation of PTEN contributes to the accumulation of PtdIns3,4,5P3 by setting up a positive-feedback loop, since the formation of this lipid messenger is responsible for stimulating the production of H2O2 (Module 2: Figure plasma membrane ROS formation).

There are additional enzymes that hydrolyse the 3-phosphorylated position. One of these is transmembrane phosphoinositide 3-phosphatase and tensin homologue 2 (TPTE2), which is localized at the plasma membrane. The other is PTEN homologous inositol lipid phosphatase (TPIP) that occurs as α and β isoforms. The TPIPα isoform hydrolyses PtdIns3,4,5P3, PtdIns3,5P2, PtdIns3,4P2 and PtdIns3P. It has N-terminal transmembrane domains that appear to localize the protein to the ER.

PTEN is a potent tumour suppressor that is frequently inactivated in many different cancers, e.g. endometrial, prostate, mammary carcinomas, melanomas and thyroid tumours. When PTEN is inactivated in neurons, there is a progressive increase in cell size and increased phosphorylation of protein kinase B (PKB) that create cerebellar abnormalities resembling those seen in human Lhermitte–Duclos disease (LDD).

Germline mutations in the PTEN gene have been implicated in the development of Cowden's disease and Bannayan–Zonana syndrome, where there is an increased risk of breast and thyroid cancers. An increase in the expression of PTEN by the tumour suppressor p53 may contribute to p53-induced apoptosis.

Myotubularins

There is a large family of myotubularins that function as lipid phosphatases that dephosphorylate PtdIns3P and also PtdIns3,5P2 (Module 2: Figure phosphoinositide metabolism). They may also act on PtdIns3,5P2. Mammalian cells express 15 myotubularins. The founding member is MTM1 and the remainder are MTMR2–MTMR14. These enzymes have a phosphatase domain, and a GRAM domain, which associates with membranes and a C-terminal coiled-coil domain that links these enzymes to other proteins. One of the isoforms, MTMR6, appears to have a specific role in the PtdIns3P signalling cassette. Hydrolysis of PtdIns3P may control the activity of the Ca2+-activated intermediate-conductance (IK) channel (KCa3.1), and this may be of particular significance for the relationship between K+ channels and cell proliferation.

These enzymes are of interest because mutations in MTM1 are responsible for X-linked recessive myotubular myopathy and MTMR2 is mutated in Charcot–Marie–Tooth disease 4B. Missense mutations in MTMR14 have also been described in centronuclear myopathy (CNM).

Oculocerebrorenal syndrome of Lowe (OCRL)

Oculocerebrorenal syndrome of Lowe (OCRL) is a Rho-GAP-domain-containing enzyme that can hydrolyse both inositol phosphates (Ins1,4,5P3 and Ins1,3,4,5P4) and the phosphoinositides PtdIns4,5P2 and PtdIns3,4,5P3. OCRL also contains an ASH [ASPM (abnormal spindle-like microcephaly-associated protein/SPD2 (spindle pole body2)/hydin] domain, which is thought to provide a binding site for microtubules. One of the functions of OCRL is to control membrane and protein trafficking during endocytosis where it operates primarily at the interface between the trans-Golgi network (TGN) and the endosomes.

A deficiency of this enzyme is responsible for Lowe's oculocerebrorenal (OCRL) syndrome. Some patients with Dent's disease also carry mutations in OCRL.

Synaptojanin

There are two synaptojanins: synaptojanin 1 (SJ1) and synaptojanin 2 (SJ2).

Synaptojanin 1 (SJ1)

Synaptojanin 1 (SJ1) is a multi-functional inositol polyphosphate 5-phosphatase that has an important role in the process of endocytosis (Module 4: Figure endocytosis). There are two lipid phosphatase domains arranged in tandem: an N-terminal Sac1 domain, which dephosphorylates PtdIns3P and PtdIns4P, and an adjacent 5-phosphatase that acts on PtdIns4,5P2 and PtdIns3,4,5P3. The C-terminal region has a proline-rich domain (PRD) that targets synptojanins to the SH3 domains of dynamin, endophilin, intersectin and amphiphysin on clathrin-coated pits (Module 4: Figure scission of endocytic vesicles). Alternative splicing of this region gives rise to the SJ1-145 and SJ1-170 isoforms. The former is strongly expressed in presynaptic terminals. SJ1 has a primary role in coat removal in the final stages of endocytosis and is particularly important for synaptic vesicle recycling. Since SJ1 is distributed throughout the neuron, it also plays a role in postsynaptic endocytosis.

The activity of SJ1 is regulated by a phosphorylation/dephosphorylation cycle operated by the dual-specificity tyrosine-phosphorylation regulated kinase 1A (DYRK1A) and calcineurin (CaN) respectively.

Synaptojanin 2 (SJ2)

Synaptojanin 2 (SJ2) has a catalytic domain that closely resembles that of synaptojanin 1 (SJ1). Despite this close similarity, SJ2 does seem to have some distinct functions. Like SJ1, SJ2 has also been implicated in clathrin-mediated endocytosis.

Proline-rich inositol polyphosphate 5-phosphatase (PIPP)

The proline-rich inositol polyphosphate 5-phosphatase (PIPP) has N- and C-terminal proline-rich domains. There also is a SKICH domain responsible for attaching PIPP to the plasma membrane where it acts to hydrolyse PtdIns4,5P2 and PtdIns3,4,5P3. It can also hydrolyse Ins1,4,5P3 and Ins1,3,4,5P4. It is highly expressed in the brain where it appears to function in neurite extension.

Skeletal muscle and kidney enriched inositol phosphatase (SKIP)

As its name implies, skeletal muscle- and kidney-enriched inositol phosphatase (SKIP) is strongly expressed in skeletal muscle and kidney, but is also found in heart and brain. It's C-terminal SKICH domain attaches it to the plasma membrane and to membranes in the perinuclear region. It acts to hydrolyse both PtdIns4,5P2 and PtdIns3,4,5P3 and has been implicated in insulin signalling.

Sac3/Fig4

The suppressor of actin 3 (Sac3), which is the mammalian orthologue of Fig4 in yeast, is a SAC domain lipid phosphatase that removes the 5-phosphate from PtdIns3,5P2 to form PtdIns3P (see step 13 in Module 2: Figure phosphoinositide metabolism). Sac3/Fig4 is part of the PAS complex and has an important structural role in that it also helps to facilitate the kinase activity of PIKfyve that operates in the PtdIns3,5P2 signalling cassette (Module 2: Figure PIKfyve activation).

Mutations in Sac3/Fig3 have been linked to Charcot–Marie–Tooth disease 4J.

Inositol phosphate metabolism

The Ca2+-mobilizing messenger function of inositol 1,4,5-trisphosphate (InsP3) is terminated through its metabolism by a complex inositol phosphate metabolic pathway (Module 2: Figure inositol phosphate metabolism) that has two main functions. Firstly, it produces free inositol that is resynthesized to PtdIns, which can be reused for further signalling. Secondly, it generates an assortment of inositol phosphates, some of which contribute to the multipurpose inositol polyphosphate signalling pathway.

The Ins1,4,5P3 that enters the metabolic pathway is metabolized via two pathways. It is dephosphorylated by Type I inositol polyphosphate 5-phosphatase (step 1) to form Ins1,4P2 or it can be phosphorylated by InsP3 3-kinase to form Ins1,3,4,5P4 (step 4). Both Ins1,4P2 and Ins1,3,4,5P4 are putative messengers in the multipurpose inositol polyphosphate signalling pathway. The Ins1,4P2 is sequentially dephosphorylated to free inositol. The Ins1,3,4,5P4 is dephosphorylated to Ins1,3,4P3, which occupies an important position in the metabolic pathway in that either it is dephosphorylated down to inositol, as part of an inositol recycling pathway, or it can be phosphorylated further to form additional inositol polyphosphates. The inositol phosphates in the pink boxes in Module 2: Figure inositol phosphate metabolism have been implicated as intracellular messengers. InsP3 mobilizes internal Ca2+ and the putative functions of the others are described in the multipurpose inositol polyphosphate signalling pathway.

Metabolism of the inositol phosphates is carried out by a large number of inositol phosphate kinases and phosphatases (Module 2: Figure inositol phosphate metabolism):

Inositol polyphosphate 5-phosphatase (INPPs)

This large family of enzymes, which are products of multiple genes and splice variants, removes the 5-phosphate group from inositol lipids and, in some cases, can also act on inositol phosphates. All members of the family share a central conserved catalytic domain:

5-Phosphatase I

The 5-phosphatase I functions to hydrolyse the inositol phosphates Ins1,4,5P3 and Ins1,3,4,5P4 (Steps 1 in Figure 2 inositol phosphate metabolism). If this enzyme is deleted, there is an increase in the level of InsP3, and this was associated with a transformed phenotype. This enzyme is inhibited following phosphorylation by Ca2+/calmodulin-dependent protein kinase II (CaMKII). Type I, which hydrolyses the inositol phosphates, is directed to the membrane by isoprenylation of the C-terminal region. If these membrane attachment sites are removed, allowing the enzyme to disperse into the cytosol, InsP3 metabolism is severely curtailed. It seems that much of the InsP3 is metabolized close to its site of action at the plasma membrane.

Src homology 2 (SH2) domain-containing inositol phosphatases (SHIP)

The Src homology 2 (SH2) domain-containing inositol phosphatases (SHIP1 and SHIP2) hydrolyse Ins1,3,4,5P4 and PtdIns3,4,5P3. The SHIPs, which come in different forms, are found in haematopoietic cells, where they play a role in inhibiting signalling. In B cells, for example, the FcγRIIB receptor has an immunoreceptor tyrosine-based inhibitory motif (ITIM) to which SHIPs attach causing a reduction in Ca2+ signalling by hydrolysing PtdIns3,4,5P3 and thereby reducing the stimulatory effect of the Tec tyrosine kinase family on phospholipase Cγ (Module 2: Figure ROS effects on Ca2+ signalling). SHIPs perform a similar function in mast cells (Module 11: Figure mast cell inhibitory signalling). SHIP1 functions to localize PIP3 at the front of the cell during neutrophil chemotaxis (Module 11: Figure neutrophil chemotaxis).

Inositol polyphosphate 1-phosphatase

Steps 2 in Module 2: Figure inositol phosphate metabolism. This enzyme dephosphorylates both Ins1,3,4P3 and Ins1,4P2. This enzyme has been implicated in the signalling pathways of compensatory hypertrophy and is up-regulated in human colorectal cancer.

Inositol monophosphatase

Steps 3 in Module 2: Figure inositol phosphate metabolism. The free inositol that is formed by this enzyme is then used to resynthesize PtdIns, which is returned to the plasma membrane, where it can be reused to function as the precursor for the phosphoinositide signalling lipids (Module 2: Figure phosphoinositide metabolism). This inositol monophosphatase is inhibited by lithium (Li+), which thus acts to reduce the supply of inositol (Module 2: Figure InsP3/DAG recycling). This ability of Li+ to lower the level of inositol is the basis of an inositol depletion hypothesis to account for its action in controlling manic-depressive illness.

Ins1,4,5P3 3-kinase

Step 4 in Module 2: Figure inositol phosphate metabolism. This kinase is one of the major pathways for metabolizing the second messenger Ins1,4,5P3. There are three isoforms of this enzyme (InsP3 3-kinase A, B and C). Both type A and type B are activated by Ca2+/calmodulin, with the type B being the most sensitive. This enzyme is also sensitive to other signalling pathways in that both forms are inhibited by PKC. On the other hand, protein kinase A (PKA) activates type A but inhibits type B.

Inositol polyphosphate 4-phosphatase type II (INPP4B)

Steps 5 in Module 2: Figure inositol phosphate metabolism. A multifunctional phosphatase capable of dephosphorylating both Ins3,4P2 and Ins1,3,4P3, as well as the inositol lipid PtdIns3,4P2. This enzyme might be a regulator of cell proliferation, since it is absent in hyperproliferative megakaryocytes that lack the transcription factor GATA-1. The INPP4B also plays an important role in the maturation of clathrin-coated pits during endocytosis (Module 2: Figure localized inositol lipid signalling).

Inositol polyphosphate 3-phosphatase

Steps 6 in Module 2: Figure inositol phosphate metabolism. A multifunctional phosphatase capable of dephosphorylating both Ins1,3P2 and PtdIns3P.

Ins1,3,4P3 6-kinase

Step 7 in Module 2: Figure inositol phosphate metabolism.

Ins1,3,4,6P4 5-kinase

Step 8 in Module 2: Figure inositol phosphate metabolism.

Ins3,4,5,6P4 1-kinase

Step 9 in Module 2: Figure inositol phosphate metabolism. This enzyme is responsible for inactivating the Ins3,4,5,6P4 that regulates chloride channels. It is an unusual enzyme in that it has alternative positional specificity (5/6-kinase activity against Ins1,3,4P3). The enzyme is the same as the Ins1,3,4,5,6P5 1-phosphatase.

Ins1,3,4,5,6P5 1-phosphatase

Step 10 in Module 2: Figure inositol phosphate metabolism. The enzyme is equivalent to the Ins 3,4,5,6P4 1-kinase.

Ins1,4,5,6P4 3-kinase

Step 11 in Module 2: Figure inositol phosphate metabolism. This is a multi-kinase that can phosphorylate both Ins1,4,5,6P4 and Ins1,4,5P3.

Ins1,3,4,5,6P5/InsP6 3-phosphatase

Steps 12 in Module 2: Figure inositol phosphate metabolism. This enzyme is also known as the multiple inositol phosphate phosphatase (MIPP). Its function is somewhat unclear because it is located within the lumen of the ER, apparently insulated from its two substrates.

Ins1,3,4,5,6P5 2-kinase

Step 13 in Module 2: Figure inositol phosphate metabolism. The production of InsP6 occurs primarily through the phosphorylation of Ins1,3,4,5,6P5 by a 2-kinase.

InsP6 kinase

Steps 14 in Module 2: Figure inositol phosphate metabolism. This kinase adds an additional phosphate to that already present on the 5-position to produce InsP7 (PP-InsP5). This kinase can run in reverse and is thus potentially capable of reforming ATP. The same enzyme may act on Ins1,3,4,5,6P5 to produce PPInsP4. There are three isoforms (InsP6K1–InsP6K3). InsP6K2 has been implicated in the action of interferons on cell growth and apoptosis.

Diphosphoinositol phosphate phosphohydrolase (DIPP)

Steps 15 in Module 2: Figure inositol phosphate metabolism. This enzyme dephosphorylates both PP-InsP5 (InsP7) and [PP]2-InsP4 (InsP8), during which it removes the β-phosphate from the diphosphate groups.

PP-InsP5 kinase

Step 16 in Module 2: Figure inositol phosphate metabolism. This kinase adds a further phosphate to PP-InsP5 to form [PP]2-InsP4, but the positional specificity of the latter remains to be established. Like the InsP6 kinase, this enzyme is capable of forming ATP by running in a reverse mode.

Mammalian inositol phosphate multikinase (mIPMK)

Steps 17 in Module 2: Figure inositol phosphate metabolism. This mammalian inositol phosphate multikinase (mIPMK) is a multifunctional kinase capable of the following phosphorylations:

  • Ins4,5P2 → Ins1,4,5P3
  • Ins1,4,5P3 → Ins1,3,4,5P4
  • Ins1,3,4,5P4 → InsP5
  • InsP5 → PP-InsP4

Inositol 1,4,5-trisphosphate (InsP3)/Ca2+ signalling cassette

This signalling cassette uses the second messenger inositol 1,4,5-trisphosphate (InsP3) to mobilize Ca2+ from internal stores. When external stimuli engage receptors on the cell surface, they activate the enzyme phospholipase C (PLC), which hydrolyses PtdIns4,5P2 to yield InsP3 and diacylglycerol (DAG) (Module 2: Figure InsP3 and DAG formation). This is a bifurcating signalling pathway in that InsP3 generates a Ca2+ signal, whereas DAG functions to stimulate protein kinase C (PKC). The InsP3 released from the membrane diffuses into the cytosol where it engages the InsP3 receptors (InsP3Rs) to release Ca2+ from the endoplasmic reticulum.

The operation of this InsP3/Ca2+ signalling cassette can be divided into separate components:

PtdIns conversion into the precursor lipid PtdIns4,5P2

The precursor PtdIns4,5P2, which is hydrolysed to give InsP3, is part of the complex inositol lipid metabolic pathway (Module 2: Figure phosphoinositide metabolism). The synthesis of the lipid precursor PtdIns4,5P2 from PtdIns is controlled by two substrate cycles involving lipid kinases and phosphatases (Module 2: Figure InsP3 and DAG formation). In the first step, the hydroxy group at the 4-position of the inositol ring is phosphorylated to form PtdIns4P. In the next step, the PtdIns4P is phosphorylated further to yield the precursor PtdIns4,5P2. The next step is the agonist-dependent hydrolysis of PtdIns4,5P2 to generate InsP3 and diacylglycerol (DAG).

Hydrolysis of PtdIns4,5P2 to generate InsP3 and diacylglycerol (DAG)

External stimuli (e.g. hormones, neurotransmitters and growth factors) gain access to this signalling pathway by activating cell-surface receptors. The latter fall into two main classes, the G protein-coupled receptors (GPCRs) and the protein tyrosine kinase-linked receptors (PTKRs). During the transduction process, the precursor lipid PtdIns4,5P2 is hydrolysed by phospholipase C (PLC) to produce both InsP3 and DAG. The family of PLCs can be distinguished by the way they are coupled to cell-surface receptors. In general, the GPCRs use the PLCβ isoforms, whereas the receptor tyrosine kinases (RTKs) are coupled to the PLCγ isoforms (Module 2: Figure PLC structure and function).

DAG functions to activate the diacylglycerol (DAG)/protein kinase C (PKC) signalling cassette, whereas the InsP3 diffuses into the cytosol to activate the InsP3 receptors to release Ca2+ stored in the endoplasmic reticulum. The signalling function of this bifurcating signalling pathway is curtailed by the metabolism of InsP3 and the resynthesis of PtdIns.

Metabolism of InsP3 and DAG and the resynthesis of PtdIns

The metabolism of InsP3 and DAG and the resynthesis of PtdIns are the OFF reactions that terminate the actions of these two messengers. There are two pathways for diacylglycerol (DAG) metabolism. Similarly, there are two separate mechanisms of inositol 1,4,5-trisphosphate (InsP3) metabolism.

Diacylglycerol (DAG) metabolism

The second messenger DAG is metabolized via two separate pathways. It can be phosphorylated by diacylglycerol (DAG) kinase to form phosphatidic acid (PA) or it is hydrolysed by diacylglycerol (DAG) lipase. The PA is transferred to the endoplasmic reticulum, where it interacts with CTP to form the CDP/DAG complex, which is a precursor in the resynthesis of PtdIns (Module 2: Figure InsP3/DAG recycling).

Diacylglycerol (DAG) kinase

Diacylglycerol (DAG) kinase α (DAGKα), which is one of a family of nine mammalian isotypes, has a number of domains, including Ca2+-binding EF-hand motifs and an N-terminal recoverin homology domain that is related to the recoverin family of neuronal Ca2+ sensors. These two domains appear to function as a unit during Ca2+-induced activation of DAGKα. In response to an increase in Ca2+, DAGKα translocates to the membrane, where it phosphorylates DAG to phosphatidic acid (PA) (Module 2: Figure InsP3/DAG recycling). DAG kinase functions in the scission of COPI-coated vesicles that bud off from the Golgi (Module 4: Figure COPI-coated vesicles).

Diacylglycerol (DAG) lipase

A diacylglycerol (DAG) lipase is responsible for removing one of the fatty acid tails from the sn-position of DAG to form monoacylglycerol (Module 2: Figure InsP3/DAG recycling). This hydrolysis of DAG, which is a Ca2+-sensitive process, may represent the primary mechanism for removing the DAG that is produced following the hydrolysis of inositol lipids.

The 2-arachidonylglycerol (2-AG) formed by DAG lipase is one of the endocannabinoids that has multiple messenger functions.

Inositol 1,4,5-trisphosphate (InsP3) and Ca2+ signalling

The primary function of inositol 1,4,5-trisphosphate (InsP3) is to function as a second messenger to release Ca2+ from the internal stores. The 1,4,5-trisphosphate receptors (InsP3Rs) located on the endoplasmic reticulum respond to InsP3 by releasing puffs of Ca2+ (Module 3: Figure InsP3R activation). These InsP3Rs are sensitive to both InsP3 and Ca2+ and can thus function as co-incident detectors (Module 2: Figure Ca2+-induced Ca2+ release). This InsP3/Ca2+ signalling system controls many different cellular processes in a large number of different cell types (Module 2: Figure InsP3/Ca2+ signalling functions).

Inositol 1,4,5-trisphosphate (InsP3) metabolism

The inositol 1,4,5-trisphosphate (InsP3) that is formed during signalling enters the pathways of inositol phosphate metabolism from which it emerges as free inositol. It is dephosphorylated by Type I inositol polyphosphate 5-phosphatase (Steps 1 in Module 2: Figure inositol phosphate metabolism) to form Ins1,4P2 or it can be phosphorylated by InsP3 3-kinase to form Ins1,3,4,5P4 (Step 4). These two products then enter a complex metabolic pathway that plays an important role in recycling the inositol headgroup. Cells have access to three separate sources of inositol: recycling the second messenger InsP3, de novo synthesis from glucose 6-phosphate or uptake of dietary inositol circulating in the plasma (Module 2: Figure InsP3/DAG recycling). Drugs such as lithium and valproate may control manic-depressive illness by reducing the supply of inositol by inhibiting the inositol monophosphatase and inositol synthase respectively. The supply of free inositol is one of the essential precursors for the synthesis of PtdIns.

Synthesis of PtdIns

The two second messengers that are formed during phosphoinositide signalling can be recycled back to the precursor lipid PtdIns4,5P2 through a series of steps (Module 2: Figure InsP3/DAG recycling):

  1. Conversion of PtdIns into the precursor lipid PtdIns4,5P2.
  2. The agonist-dependent activation of phospholipase C (PLC), which hydrolyses PtdIns4,5P2 to generate the second messengers inositol 1,4,5-trisphosphate (InsP3) and diacylglycerol (DAG).
  3. The InsP3 is recycled back to free inositol by a complex metabolic pathway illustrated in Module 2: Figure inositol phosphate metabolism.
  4. DAG is phosphorylated by the Ca2+-sensitive enzyme DAG kinase to form phosphatidic acid (PA) or it is hydrolysed by DAG lipase to form monoacylglycerols (MAGs) such as 2-arachidonylglycerol (2-AG), which is one of the endocannabinoids. The 2-AG is hydrolysed by monoacylglycerol lipase (MAGL) to glycerol and arachidonic acid.
  5. PA is transferred from the plasma membrane to the ER where it interacts with CTP to form CDP/DAG by a CDP/DAG synthetase. The inositol, which can be produced through three mechanisms (de novo synthesis, recycling or uptake of dietary inositol), is attached to CDP/DAG by the PtdIns synthetase [cytidine diphosphate (CDP)/diacylglycerol (DAG):myo-inositol 3-phosphatidyltransferase] located on the endoplasmic reticulum (ER) to form PtdIns.
  6. The PtdIns is then transported from the ER back to the plasma membrane by a PtdIns transfer protein (PITP). Cells express two PITPs, an α and a β isoform, produced by separate genes. The latter appears to be the housekeeping isoform that is essential for cell survival, whereas the α isoform has a more specialized function. When PtdIns is added back to the plasma membrane, it can once again be converted through the two phosphorylation reactions to maintain the supply of the PtdIns4,5P2 that is the precursor for the phosphoinositide signalling pathway.
  7. Cells can also use plasma inositol, which is taken up by a sodium-dependent myo-inositol cotransporter-1 (SMIT1).

Phospholipase C (PLC)

Phospholipase C (PLC) hydrolyses the lipid precursor PtdIns4,5P2 to produce both InsP3 and DAG. It is made up of five subclasses, which have variable isoforms PLCβ (β1–β4), PLCδ (δ1–δ4), PLCγ (γ1 and γ2), PLCε and PLCζ (Module 2: Figure PLC structure and function). All forms of the enzyme have an absolute requirement for Ca2+ that plays a critical role in the catalytic site. The PtdIns4,5P2 is cleaved through two sequential reactions: first, the phosphodiester bond is cleaved to DAG and inositol 1,2-cyclic phosphate, the latter is then hydrolysed to give the acyclic InsP3 that is released into the cytoplasm. The domain structure of the different PLC isoforms reveals a number of common structural features related to the way the enzyme associates with the membrane and functions to hydrolyse PtdIns4,5P2. The catalytic domain is made up from the X and Y regions. They have at least two potential lipid-binding domains: the pleckstrin homology (PH) domain and the C2 domain (Module 6: Figure modular lipid-binding domains). In the case of PLCδ1, the enzyme may first associate with the membrane through its PH domain, which has a high affinity for PtdIns4,5P2. Further interactions may then occur through the C2 domain, which has an extensive interface with the catalytic domain and may thus enable the catalytic site to integrate itself into the membrane to hydrolyse the lipid. The main difference between these PLC isoforms concerns the way in which they are activated.

Phospholipase Cβ (PLCβ)

The PLCβ isoforms are mainly activated by G protein-coupled receptors (GPCRs). PLCβ1 and PLCβ3 are fairly ubiquitous, whereas PLCβ2 and PLCβ4 have more limited tissue distributions.

PLCβ isoform distribution and function

PLCβ1 is highly concentrated in brain (pyramidal cells of hippocampus, Purkinje cells and granule cells). The knockout phenotype is characterized by seizures leading to sudden death. These seizures are similar to those seen during epilepsy. Many of the defects are found in the central nervous system (CNS). Activation of PLC by muscarinic receptors is suppressed in the temporal lobe, cerebellum and hippocampus, and this could decrease the inhibitory tone, leading to the seizures.

PLCβ2 is mainly expressed in cells of the immune system. Knockout mice show some disruption of chemokine signalling. For example, neutrophils fail to respond to the chemoattractant fMet-Leu-Phe with the usual changes in PLC activation, Ca2+ release and superoxide radical (O2−•) production. However, the response to lipopolysaccharide (LPS) is normal. Despite this absence of PLC activation in leucocytes, chemotactic responses are enhanced, suggesting that this signalling pathway may normally antagonize the signalling pathways normally used to control chemotaxis.

PLCβ3 is found mainly in brain, parotid, smooth muscle and liver. Disruption of this gene is lethal, with the mice dying by day 2.5. The embryos are highly disorganized and have low cell numbers, suggesting a role for this isoenzyme in cell division. This isoform has two sites (Ser-26 and Ser-105) that are phosphorylated by cyclic GMP-dependent protein kinase (PKG) resulting in a decrease in enzyme activity (Module 7: Figure smooth muscle cell cyclic GMP signalling).

PLCβ4 is found in cerebellum and granule cells. Knockout mice appear to have defects in the cerebellum, resulting in poor motor co-ordination due to a decrease in PLC stimulation through metabotropic glutamatergic and muscarinic receptors. There also are defects in the processing of visual information.

The primary activation mechanism of the PLCβ isoforms is through the Gq family of heterotrimeric G proteins (Gq, G11, G14, G15 and G16) (Module 2: Figure PLC structure and function). Gq is a highly versatile G protein in that it can relay information to a number of signalling pathways. For example, the Gq dissociates into Gα that can stimulate PLCβ and it can also activate Trio that then activates both the JNK signalling pathway and the p38 pathway. These two pathways have very different kinetics: stimulation of PLCβ is very fast and terminates quickly because the PLCβ has endogenous GAP activity that hydrolyses the GTP on Gα to inactivate signalling. On the other hand, the Gα associated with Trio inactivates very slowly thus producing long-lasting signals ideally suited to stimulate proliferation (Module 9: Figure growth factor signalling). The Gβγ is also active and can stimulate the extracellular-signal-regulated kinase (ERK) pathway (Module 2: Figure ERK signalling).

While Gq and G11 are found in most tissues, the other three are restricted to cells of haematopoetic origin. The function of these G proteins is complicated because the PLCβ isoforms are sensitive to both the α and βγ components of the heterotrimeric complex. PLCβ1 and PLCβ4 are most sensitive to stimulation through α subunits, whereas βγ is more effective at activating PLCβ2 and PLCβ3. This sensitivity to βγ subunits may explain the pertussis-toxin-sensitive stimulation of PLC by the Gi family of G proteins. For example, the adenosine A1, muscarinic M2, somatostatin and μ-, δ- and κ-opioid receptors can couple to PLC through Gi and Go. Of the five Gβ and 11 Gγ subunits, there appears to be little specificity with regard to the activation of PLC. In general, Gβγ is less efficacious than Gα in stimulating PLC.

The two G protein subunits interact with PLCβ at different sites. The α subunit interacts with a region within the long C-terminal tail, whereas the βγ subunit interacts with a site within the Y region of the catalytic site (Module 2: Figure PLC structure and function).

PLCβ also contains GAP (GTPase-activating protein) activity that stimulates the intrinsic GTPase activity of the α subunit. When an agonist binds to the GPCRs, it induces a conformational change that is transmitted to the underlying G protein, causing it to dissociate. At this time the GDP bound to the α subunit is exchanged for GTP. The active GTP/α subunit complex can then activate PLCβ and this activation process is terminated when the intrinsic GTPase activity of the α subunit hydrolyses GTP back to GDP allowing the GDP/α subunit to once again bind to the βγ subunit to form the inactive complex. The intrinsic GTPase activity of the α subunit is rather low, but is greatly enhanced by two mechanisms. There is a GAP activity associated with the long C-terminal tail of PLCβ. Therefore PLCβ plays a direct role in terminating its own activity. In addition, there are regulators of G protein signalling (RGS) proteins. There are about 20 of these RGS proteins, of which RGS2–RGS4 seem particularly effective in interacting with Gαq. The GTPase activity of Gαq is enhanced 25-fold by RGS4. There is considerable discrimination with regard to how individual RGS proteins bind to specific receptor/Gαq complexes enabling them to exert an agonist-specific inhibitory mechanism. This mechanism could explain the different types of Ca2+ signals observed in hepatocytes following stimulation with different agonists.

Modulation of PLCβ by other signalling pathways

In general, the phosphorylation of PLCβ, particularly isoforms 2 and 3, by protein kinase A (PKA) results in inhibition of the enzyme.

Phospholipase Cγ (PLCγ)

PLCγ functions predominantly in early development and in the signalling pathways that control cell proliferation. PLCγ1 knockout mice die at embryonic day 9, by which time the embryos appear normal, albeit somewhat smaller.

PLCγ is characterized by having a large insert between the X and Y domains (Module 2: Figure PLC structure and function). This insert contains an additional pleckstrin homology (PH) domain, which is itself split by another insert containing two Src homology 2 (SH2) domains and a single Src homology 3 (SH3) domain. The SH2 domains play a critical role in the activation of PLCγ because they provide a docking module that enables the enzyme to translocate to the membrane to dock to phosphorylated tyrosine residues on activated receptors or associated scaffolding proteins. This tyrosine phosphorylation occurs during activation of protein tyrosine kinase-linked receptors (PTKRs) or the non-receptor protein tyrosine kinases (e.g. Src, Syk, Btk, Lck, Fyn). This sequence of events is well illustrated by the activation of PLCγ1 during activation of the T cell receptor (Module 9: Figure TCR signalling) or PLCγ2 following B-cell antigen receptor (BCR) activation (Module 9: Figure B cell activation). This translocation to the cell surface has two important consequences for the activation of the enzyme. Firstly, it brings the enzyme close to its substrate in the membrane. Secondly, by interacting with tyrosine kinases, PLCγ1 is itself phosphorylated on tyrosine residues located at positions 771, 783 and 1254 (black triangles in Module 2: Figure PLC structure and function). Phosphorylation at Tyr-783 is particularly important for switching on enzyme activity. The PLCγ2 isoform is activated by phosphorylation of tyrosine residues 753 and 759, mainly by the tyrosine kinase Btk (Module 2: Figure ROS effects on Ca2+ signalling).

The enzymatic activity of PLCγ is also stimulated by the PtdIns3,4,5P3 produced by the PtdIns 3-kinase signalling pathway and thus represents a major point of interaction between these two signalling pathways. This interaction is particularly important during lymphocyte activation, where the Tec tyrosine kinase family are activated by PtdIns3,4,5P3 to phosphorylate and activate both PLCγ1 and PLCγ2. Both isoforms bind strongly to this highly charged lipid through the N-terminal PH domain, and this interaction enhances enzymatic activity.

The Src homology 3 (SH3) domain, which binds to proline-rich sequences, may enable PLCγ to bind to other components, such as the cytoskeleton and the protein dynamin.

Phospholipase Cδ (PLCδ)

Less is known about the activation of PLCδ. Unlike PLCβ and PLCγ, PLCδ appears not to be regulated by receptors directly. Once PLCδ1 has associated with the membrane, it appears to be activated by an elevation in cytosolic Ca2+. This sensitivity to changes in intracellular Ca2+ is thought to depend on the C2 domain (Module 2: Figure PLC structure and function).

Crystallographic studies of PLCδ1 have begun to reveal how PLC enzymes function to cleave their lipid substrates that are embedded in the plasma membrane.

Phospholipase Cε (PLCε)

Phospholipase Cε (PLCε) differs from the other isoforms by having two Ras-association domains (RA) at its C-terminus (Module 2: Figure PLC structure and function). It binds specifically to the GTP-bound forms of Ha-Ras and Rap1A. The enzyme is targeted to the membrane through its ability to bind to these G proteins. It also has a Cdc25 homology domain that is a guanine nucleotide exchange factor (GEF) motif for Rap1, which enables it to translocate to the perinuclear region following epidermal growth factor (EGF) stimulation, where it can activate the mitogen-activated protein kinase (MAPK) signalling pathway. There is evidence that cyclic AMP acting through the exchange protein activated by cAMP (EPAC) can stimulate Rap1 to activate PLCε (Module 2: Figure cyclic AMP signalling) and this mechanism has been implicated in the control of autophagy (Module 11: Figure autophagy signalling mechanisms).

Phospholipase Cζ (PLCζ)

Phospholipase Cζ (PLCζ) has a restricted expression in that it is only found in mammalian sperm. PLCζ lacks the N-terminal PH domain (Module 2: Figure PLC structure and function), which means that there is no obvious mechanism for it to bind to phospholipids. PLCζ has a sensitivity to Ca2+ that is 100-fold greater than that of PLCδ1. Both the C2 and the tandem EF-hands are necessary for PLCζ to hydrolyse PtdIns4,5P2 to generate the InsP3 that plays a central role in sperm-induced oocyte activation during fertilization (Module 8: Figure mammalian fertilization).

Mutations in PLCζ have been linked to male infertility.

Diacylglycerol (DAG)/protein kinase C (PKC) signalling cassette

The diacylglycerol (DAG) that is formed when PtdIns4,5P2 is hydrolysed by phospholipase C (PLC) remains within the plane of the membrane where it acts as a lipid messenger to stimulate some members of the /1protein kinase C (PKC) family (/1Module 2: Figure InsP3 and DAG formation). This PKC family contains a number of isoforms with diverse structures and activation mechanisms. The /1PKC signalling function is equally diverse, and it has proved difficult to pin down its precise function in the regulation of specific cellular processes.

Protein kinase C (PKC)

The protein kinase C family has been divided into three subgroups (Module 2: Figure PKC structure and activation):

  • Conventional PKCs (cPKCs: α, β1, β2 and γ) contain C1 and C2 domains
  • Novel PKCs (nPKCs: δ, ε, η and θ) contain a C1 and a C2-like domain
  • Atypical PKCs (aPKCs: ζ, ι and λ) contain a truncated C1 domain

The activation of these different PKC families is complex in that it depends on a number of processes such as priming, translocation and association with scaffolding proteins. The priming process is driven by a sequence of multisite phosphorylation events that have to occur before the enzyme can perform its signalling function. The newly synthesized proteins are inactive and undergo a priming process that begins when phosphoinositide-dependent protein kinase 1 (PDK1), which is part of the PtdIns 3-kinase signalling pathway, phosphorylates a site in the C4 domain (Module 2: Figure PKC structure and activation). This is then followed by intramolecular autophosphorylation reactions of two further sites in the C-terminal tail. The primed enzyme is now ready to perform its signalling function by responding to both diacylglycerol (DAG) and Ca2+. The cPKCs and most of the nPKCs are lipid-sensitive enzymes in that they are activated by DAG that binds to the C1 domain. This C1 domain is also responsible for binding the tumour-promoting phorbol esters. In the case of the cPKCs, Ca2+ also plays an important role as a cofactor that binds to the C2 domain to increase its affinity for acidic phospholipids and is thus responsible for inducing the translocation of cPKCs to the plasma membrane.

The activation of the cPKC isoforms follows a set sequence of events that begins with the hydrolysis of PtdIns4,5P2 to form DAG and inositol 1,4,5-trisphosphate (InsP3). The InsP3 releases Ca2+, which initiates the activation process by binding to the C2 domain to induce a conformational change that greatly enhances the membrane affinity of cPKC and thus promotes its translocation to the membrane where it is activated when it makes contact with DAG. The pseudosubstrate (PS) domain swings away from the active site, which is now free to phosphorylate its substrates.

The nPKCs, which are also activated by DAG, have a much slower activation process because the Ca2+-dependent facilitation of membrane translocation is absent.

Protein kinase Cζ (PKCζ)

This is one of the atypical PKC isoforms (aPKC), which fails to respond to diacylglycerol (DAG) or Ca2+, but is sensitive to low levels of ceramide (Module 2: Figure sphingomyelin signalling). PKCζ is also activated following phosphorylation by kinases from other signalling pathways. For example, PKCζ is activated by phosphoinositide-dependent kinase 1/2 (PDK1/2) (Module 2: Figure PtdIns 3-kinase signalling). In some cells, it is responsible for activating nuclear factor κB (NF-κB). PKCζ can also phosphorylate the glucose transporter in skeletal muscle (Module 7: Figure skeletal muscle E-C coupling).

Protein kinase Cθ (PKCθ)

Protein kinase Cθ (PKCθ) is one of the novel PKCs (nPKCs) that plays an important role in lymphocyte activation where it is concentrated in the central region of the immunological synapse (Module 9: Figure immunological synapse structure). The PKCθ functions by activating the nuclear factor κB (NF-κB) signalling pathway by phosphorylating the scaffolding protein CARMA1 (Module 9: Figure TCR signalling).

PKC signalling functions

The different PKC isoforms function to regulate a wide range of cellular processes:

Myristoylated alanine-rich C-kinase substrate (MARCKS)

The myristoylated alanine-rich C-kinase substrate (MARCKS) protein can interact with a variety of molecules such as actin, calcium/calmodulin and membrane phospholipids and particularly phosphatidylinositol 4,5-bisphosphate (PtdIns4,5P2). This lipid binding enables MARCKS to associate with cell membranes.

The function of MARCKS is regulated through both Ca2+ signalling and the diacylglycerol (DAG)/protein kinase C (PKC) signalling cassette. Under resting conditions, MARCKS is associated with membranes and this membrane attachment and interaction with actin is reversed following its phosphorylation by PKC or its interaction with Ca2+ that associates with calmodulin that is bound to MARCKS.

In neurons, MARCKS protein is located both pre- and post-synaptically. It may function to attach actin filaments to the plasma membrane (see 3 in Module 10: Figure Ca2+-induced synaptic plasticity). There are indications that MARCKS may function in the stabilization of the actin cytoskeleton in membrane spines during learning. The main role for this protein seems to be the clustering of PtdIns4,5P2 at the membrane where it performs the multiple functions of the PtdIns4,5P2 signalling cassette.

In endothelial cells it can regulate movement and may act to link actin filaments to the plasma membrane. MARCKS may be susceptible to redox-sensitive reorganization of the cytoskeleton as an increase in ROS seems to contribute to the control of endothelial regulation of paracellular permeability by disrupting the cytoskeleton.

Phosphoinositide lipid signalling molecules

Some of the phosphoinositides derived from the metabolism of PtdIns (Module 2: Figure phosphoinositide metabolism) have important signalling functions that are often located in specific cellular membranes (Module 2: Figure localized inositol lipid signalling). One of the most versatile systems is the PtdIns4,5P2 signalling cassette. PtdIns4,5P2 is the precursor used to generate second messengers such as InsP3, DAG and PtdIns3,4,5P3. The latter is a particularly important messenger because the formation of PtdIns3,4,5P3 drives the PtdIns 3-kinase signalling system that controls a great variety of cellular processes (Module 2: Figure PtdIns 3-kinase signalling). In addition, PtdIns4,5P2 is a messenger in its own right in that localized formation of this lipid can control endocytosis, actin remodelling, ion channels and phagocytosis (Module 2: Figure localized inositol lipid signalling).

There is increasing evidence that some of the other lipids may also play signalling roles as described in the PtdIns3P signalling cassette, the PtdIns4P signalling cassette, the PtdIns5P signalling cassette and the PtdIns3,5P2 signalling cassette. In the case of the latter, the PtdIns3,5P2 in the late endosome activates the TRPML1 channels to produce the Ca2+ signals necessary to trigger the membrane fusion events that form the lysosomes.

PtdIns4,5P2 signalling cassette

The phosphoinositide PtdIns4,5P2 is widely distributed throughout the cell. Not only is it found in the plasma membrane, but also it occurs in the Golgi, endosomes, endoplasmic reticulum (ER) and within dense structures within the nucleus. This lipid has two main roles. Firstly, it is hydrolysed to form InsP3 and diacylglycerol (DAG) or it can be phosphorylated further to form the lipid messenger PtdIns3,4,5P3 (Module 2: Figure localized inositol lipid signalling). Secondly, PtdIns4,5P2 can also function as a lipid messenger (Module 2: Figure PtdIns4,5P2 signalling). Such a signalling role is supported by the observation that cell stimulation can result in highly localized increases in PtdIns4,5P2. The localized formation of this lipid messenger can be activated by the monomeric G proteins and in particular the family of Rho proteins. For example, Rac and Rho can function to target PtdIns4P 5-kinase I to the plasma membrane. In addition, Rho-GTP can activate this kinase directly to form PtdIns4,5P2 (Module 2: Figure Rho signalling).

There is increasing evidence that a variety of cellular processes are activated in response to a change in the level of this lipid:

PtdIns4,5P2 regulation of actin remodelling

There are a number of cellular processes, such as chemotaxis, locomotion, phagocytosis, shape change and cytokinesis, where there are rapid changes in the cytoskeleton in response to external stimuli such as growth factors and cytokines. Inositol lipids play a significant role in transmitting information from the plasma membrane to a variety of cell signalling systems (Module 2: Figure localized inositol lipid signalling). The formation of PtdIns4,5P2 functions to control actin remodelling (Module 4: Figure actin remodelling). Its main actions are to uncap the barbed end of actin filaments and to facilitate actin nucleation and polymerization by controlling a variety of actin regulatory proteins. A critical component of the signalling network is the Rho family of monomeric G proteins (Arf, Rho, Cdc42 and Rac) that relay information from various signalling pathways such as the PtdIns 3-kinase signalling pathway to the effector molecules that are directly responsible for remodelling the actin network. One of the signalling molecules is PtdIns4,5P2.

The signalling cascade begins when external signals acting through guanine nucleotide exchange factors (GEFs) stimulate members of the Rho family of G proteins by catalysing the exchange of GDP for GTP. The active G proteins can then have a number of functions. For example, one function of Rac-GTP is to activate actin polymerization (Module 2: Figure Rac signalling). Similarly, Cdc42 also stimulates actin remodelling (Module 2: Figure Cdc42 signalling). The consequence of this remodelling varies depending on which G protein is operating. For example, the Cdc42 and Rac polymerization of actin results in the formation of filopodia and ruffles respectively (Module 4: Figure actin remodelling).

The formation of PtdIns4,5P2 contributes to actin remodelling by acting on a number of the molecules that control actin polymerization such as the Wiskott–Aldrich syndrome protein (WASP) and the related Wiskott–Aldrich syndrome protein (WASP) verprolin homologous (WAVE) proteins. WASP and WAVE relay information from the upstream signalling elements to the downstream cytoskeletal regulators. Perhaps the most important example of the latter is the actin-related protein 2/3 (Arp2/3) complex (containing seven strongly associated subunits of which two are the actin-related proteins Arp2 and Arp3) that is responsible for catalysing the polymerization of actin. PtdIns4,5P2 appears to have two actions. It binds to and activates WASP and to Scar and thus contributes to the relay of information to Arp2/3. In addition, it can alter the activity of various proteins that modify the structure of actin. For example, it is an indirect activator of Cdc42 and it may loosen the gelsolin caps on the end of the barbed ends, opening up new sites for actin polymerization. The PtdIns4,5P2 may also induce a conformational change in vinculin that enables it to bind to talin during the formation of the focal adhesion complex (Module 6: Figure vinculin function).

PtdIns4,5P2 function in focal adhesions

The Iγ isoform of PtdIns4P 5-kinase (PtdIns4P 5-K) is highly localized at the focal adhesion complex, where it is tightly bound to talin to regulate the local formation of PtdIns4,5P2 (Module 6: Figure integrin signalling).

PtdIns4,5P2 regulation of exocytosis

The synthesis of PtdIns4,5P2 has been implicated in the ATP-dependent processes of priming vesicles as part of the exocytotic mechanism. The priming process seems to depend upon a number of steps that begin with the PtdIns transfer protein carrying PtdIns to the vesicle membrane where it is phosphorylated to PtdIns4P by PtdIns 4-kinase (PtdIns 4-K) and then to PtdIns4,5P2 by the PtdIns4P 5-kinase (PtdIns4P 5-K). It is still not clear exactly how this PtdIns4,5P2 functions to prime the vesicle for exocytosis.

The ability of the neuronal Ca2+ sensor-1 (NCS-1) to facilitate exocytosis may depend on its ability to activate the PtdIns 4-kinase (PtdIns 4-K).

PtdIns4,5P2 regulation of membrane trafficking and endocytosis

PtdIns4,5P2 plays a role in endocytosis by targeting adaptor protein 2 (AP2) and various clathrin-associated sortin proteins (CLASPs) (Module 4: Figure cargo sorting proteins) to endocytic vesicles to provide a scaffold to bind clathrin to form clathrin-coated pits (Module 4: Figure endocytosis). The PtdIns4P 5-kinase Iγ is recruited to the plasma membrane, where it phosphorylates PtdIns4P to PtdIns4,5P2, which recruits the adaptor protein AP-2 to form the clathrin-coated pits.

PtdIns4,5P2 regulation of phagocytosis

When macrophages engulf foreign particles, PtdIns4P 5-kinase (PtdIns4P 5-K) Iα is recruited to the plasma membrane at the phagosome cup, where it induces a local pulse of PtdIns4,5P2, which is then rapidly degraded by phospholipase C (PLC) to leave behind diacylglycerol (DAG). This local formation of PtdIns4,5P2 seems essential for the early process of phagocytosis.

The localized transient increase in PtdIns4,5P2 may function in the initial recruitment of actin to the phagocyte.

PtdIns4,5P2 activation of phospholipase D

Phospholipase D (PLD), which catalyses the hydrolysis of phosphatidylcholine (PC) to generate phosphatidic acid (PA), exists as two isoforms, PLD1 and PLD2. The PLD1 is primarily located on vesicles associated with the endosomal/lysosomal pathway, whereas PLD2 is mainly found on the plasma membrane. An interaction between PtdIns4P 5-kinase α (PtdIns4P 5-Kα) and the PLDs ensures that there is a local generation of PtdIns4,5P2 that plays an important role in regulating PLD activity (Module 2: Figure PLD signalling).

PtdIns4,5P2 regulation of ion channels and exchangers

A number of ion channels and exchangers appear to be regulated by PtdIns4,5P2. In some cases, the hydrolysis of PtdIns4,5P2 results in channel opening as occurs for the transient receptor potential (TRP) vanilloid 1 (TRPV1) channel, whereas the TRPC6 channel, TRP melastatin 4 (TRPM4), TRP melastatin 7 (TRPM7) and TRP melastatin 8 (TRPM8) channels are closed (Module 2: Figure PtdIns4,5P2 regulation of TRP channels). In the case of the inward rectifying K+ channels, lipid hydrolysis results in channel opening (Module 2: Figure PtdIns4,5P2 regulation of K+ channels). Modulation of the TRPV1 channel is particularly important in hyperalgesia (Module 10: Figure nociception).

Much of the information on this regulatory function of PtdIns4,5P2 has emerged from studies on K+ channels. Stimulation of phospholipase C (PLC) by G protein-coupled receptors (GPCRs) causes a reduction in the level of PtdIns4,5P2, which results in a reduction in the activity of some of the voltage-dependent K+ (Kv) channels:

PtdIns 3-kinase signalling

The PtdIns 3-kinase signalling cassette, which generates the lipid second messenger phosphatidylinositol 3,4,5-tetrakis phosphate (PtdIns3,4,5P3), has multiple functions in regulating a wide range of cellular processes such as the control of glycogen metabolism, lipid synthesis, protein synthesis, gene transcription and cell growth, inflammation, cytoskeletal rearrangement and apoptosis (Module 2: Figure PtdIns 3-kinase signalling). With regard to the latter, the PtdIns 3-kinase signalling cassette has a special function in cell survival through its contribution to the hormonal modulation of apoptosis and by enhancing the activity of the target of rapamycin (TOR). PtdIns 3-kinase signalling is one of the major pathways used by the insulin receptor to regulate energy uptake and storage.

This lipid signalling pathway has a special relationship to Ca2+ signalling in that it can regulate the way in which external signals can maintain information flowing through the /1inositol 1,4,5-trisphosphate (InsP3)/Ca2+ signalling cassette.

Operation of the PtdIns 3-kinase signalling cassette can be considered in two parts:

  • /1Generation and metabolism of the 3-phosphorylated lipid messengers
  • /1Mode of action of the 3-phosphorylated lipid messengers

One of the important functions of this pathway is to regulate cell proliferation. Since a number of human cancers are found to have mutations in certain components of this pathway, there is considerable interest in the relationship between PtdIns 3-kinase signalling and cancer. A decline in the activity of this signalling pathway is the cause of insulin resistance that leads to diabetes (Module 12: Figure insulin resistance).

Generation and metabolism of the 3-phosphorylated lipid messengers

Cells contain a number of phosphoinositides carrying a phosphate on the 3-position (Module 2: Figure phosphoinositide metabolism). One of these is PtdIns3,4,5P3, which is the main second messenger operating in the PtdIns 3-kinase signalling pathway. This highly phosphorylated lipid is generated in cells following activation of either G protein-coupled or tyrosine kinase-coupled receptors (Module 2: Figure PtdIns 3-kinase signalling).

These 3-phosphorylated lipid messengers are metabolized either by the type II inositol polyphosphate 5-phosphatases or by phosphatase and tensin homologue deleted on chromosome 10 (PTEN).

Mode of action of the 3-phosphorylated lipid messengers

The lipid messengers PtdIns3,4P2 and PtdIns3,4,5P3 act within the plane of the plasma membrane by binding to a great variety of target proteins (/1Module 2: Figure PtdIns 3-kinase signalling). Most of these downstream targets are soluble proteins that translocate to the membrane by binding to the lipid messengers through various lipid-binding domains [e.g. /1pleckstrin homology (PH), /1Phox homology (PX), /1C2 domains and basic amino acid regions] (/1Module 6: Figure modular lipid-binding domains). Once drawn on to the membrane, these target proteins are activated and function as effectors to control a large number of cellular processes:

Phosphoinositide-dependent kinase 1 (PDK1)

One of the main functions of phosphoinositide-dependent kinase 1 (PDK1) is to phosphorylate protein kinase B (PKB), which translocates to the membrane, where it binds to PtdIns3,4,5P3 (Module 2: Figure PtdIns 3-kinase signalling). This binding alters the conformation of PKB so that critical sites become available to PDK1. Like PKB, PDK1 has pleckstrin homology (PH) domains that also bind to the 3-phosphorylated lipid messengers that serve to activate their kinase activity. PDK1 is also responsible for phosphorylating and activating other signalling molecules such as ribosomal S6 protein kinase 1 (S6K1) and atypical protein kinase Cζ (PKCζ).

Protein kinase B (PKB)

Protein kinase B (PKB), which is also known as Akt, is a serine/threonine protein kinase that functions in the PtdIns 3-kinase signalling pathway (Module 2: Figure PtdIns 3-kinase signalling). PKB has three members (PKBα, PKBβ and PKBγ) that are activated through a two-stage process. Firstly, it translocates to the membrane by binding to either PtdIns3,4P2 or PtdIns3,4,5P3 through pleckstrin homology (PH) domains. The latter appears to be particularly important when cells are studied in vivo. The next stage depends upon its interaction with phosphoinositide-dependent kinase 1 (PDK1), which then completes the activation process by phosphorylating PKB on Thr-308. In addition, a DNA-dependent protein kinase (previously called PDK2) phosphorylates Ser-473. The activated PKB then functions to stimulate a variety of molecular targets, including glycogen synthase kinase-3 (GSK-3) (which mediates the effect of insulin on glycogen metabolism) and the pro-apoptotic factor Bad. PKB also plays a role in redox signalling in apoptosis.

PKB is a key player in cell growth control, where it functions by phosphorylating tuberous sclerosis 1 and 2 (TSC1/2), which integrate a number of inputs that control the activity of the target of rapamycin (TOR) (Module 9: Figure target of rapamycin signalling). PKB can also translocate into the nucleus, where it phosphorylates the Forkhead box O (FOXO) transcription factors (Module 4: Figure FOXO control mechanisms) that play an important role in regulating the cell cycle.

A reduction in PKB activity by various mechanisms seems to be responsible for the onset in insulin resistance (Module 12: Figure insulin resistance).

Ribosomal S6 protein kinase 1 (S6K1)

This kinase plays an important role in regulating protein synthesis by phosphorylating the S6 ribosomal protein (a component of the 40S ribosomal subunit), which then enhances the translation of those mRNA transcripts that contain a polypyrimidine tract at the 5' transcriptional start site. The control of S6K1 is complex in that it depends upon a priming step that is followed by series of phosphorylation events at multiple sites that are sensitive to a number of kinases. The priming step depends on Ca2+, which appears to act by opening up the enzyme so that it becomes sensitive to phosphorylation by different kinases. One of these is phosphoinositide-dependent kinase 1 (PDK1), which carries out the wortmannin-sensitive phosphorylation. In addition there are rapamycin-sensitive sites phosphorylated by the target of rapamycin (TOR) (Module 9: Figure target of rapamycin signalling).

Glycogen synthase kinase-3 (GSK-3)

One of the primary functions of insulin is to act through glycogen synthase kinase-3 (GSK-3) to stimulate the conversion of glucose into glycogen by increasing the activity of glycogen synthase (Module 7: Figure skeletal muscle E-C coupling). GSK-3 is a proline-directed protein kinase that usually requires a priming kinase to add a phosphate to its substrate before it can carry out further phosphorylations. Casein kinase I (CKI) often acts as the priming kinase as it does for β-catenin in the Wnt signalling pathway (Module 2: Figure Wnt canonical pathway).

GSK-3 also has a number of other signalling functions:

Tec tyrosine kinase family

All members of this tyrosine kinase family are cytosolic, but they translocate rapidly to the membrane through pleckstrin homology (PH) domains that are particularly sensitive to PtdIns3,4,5P3. A prominent member of this family is Bruton's tyrosine kinase (Btk). Once bound to the membrane, these enzymes are also phosphorylated by Src family tyrosine kinases such as Lyn and Src. Activation thus requires both translocation to the membrane and phosphorylation on tyrosine residues. One of the functions of these Tec tyrosine kinases is to enhance Ca2+ signalling by maintaining the activity of phospholipase Cγ (PLCγ). This enhancement of PLCγ activity by the Tec kinases represents a major point of interaction between the inositol 1,4,5-trisphosphate (InsP3)/Ca2+ signalling cassette and the PtdIns 3-kinase signalling cassette.

Bruton's tyrosine kinase (Btk)

Bruton's tyrosine kinase (Btk) is one of the non-receptor protein tyrosine kinases (Module 1: Figure non-receptor tyrosine kinases) that plays an important role in B cells where it functions to activate PLCγ2 (Module 9: Figure B cell activation). It has a similar mode of action in the mast cell FcεRI signalling pathway (Step 7 in Module 11: Figure FcεRI mast cell signalling).

Inactivation of Btk results in Bruton's type X-linked agammaglobulinaemia.

Inducible T cell kinase (Itk)

As for Bruton's tyrosine kinase (Btk) in B cells, inducible T cell kinase (Itk) has a similar function in controlling the maturation of T cells by activating phospholipase Cγ1 (PLCγ1) (Module 9: Figure TCR signalling).

Insulin receptor

Insulin has a major role to play in regulating a variety of cellular processes, with particular emphasis on the regulation of energy uptake and storage. The action of insulin is carried out by the insulin receptor, which is a disulphide-linked homodimer that belongs to the large family of receptors that have tyrosine kinase domains (Module 1: Figure tyrosine kinase-linked receptors). An important component of the signal transduction mechanism used by the insulin receptor is the insulin receptor substrate (IRS), which is a classical scaffolding protein (Module 6: Figure IRS domain structure).

The function of the insulin receptor depends upon the activation of the PtdIns 3-kinase signalling pathway that occurs through the sequence of events shown in Module 2: Figure insulin receptor:

  1. Insulin binds to the extracellular domain to induce a conformational change in the receptor resulting in the activation of the intracellular tyrosine kinase domains.
  2. Once activated, the tyrosine kinase domains undergo autophosphorylation whereby they phosphorylate up to six tyrosine residues on the opposite β chain.
  3. The insulin receptor substrate (IRS), which is a scaffolding domain (Module 6: Figure IRS domain structure), attaches itself to juxtamembrane phosphotyrosine residue 972. Once it is drawn into the vicinity of the receptor, the tyrosine kinase domain also phosphorylates IRS on multiple residues.
  4. The phosphotyrosine residues on IRS then function as docking sites for Class IA PtdIns 3-kinase (PI 3-K).
  5. The PI 3-K then phosphorylates PtdIns4,5P2 to form the lipid second messenger PtdIns3,4,5P3.
  6. The PtdIns3,4,5P3 then activates the PtdIns 3-kinase signalling pathway (Module 2: Figure PtdIns 3-kinase signalling) that controls a number of cellular processes:

The onset of diabetes (i.e. Type 2 diabetes) begins when cells become resistant to the action of insulin in energy uptake and storage.

PtdIns3,5P2 signalling cassette

The PtdIns3,5P2 signalling cassette functions in membrane and protein trafficking and particularly in the events that occur during the early endosome maturation to lysosome (see step 10 in Module 4: Figure membrane and protein trafficking). The early endosome is rich in PtdIns3P that is converted into PtdIns3,5P2 (Module 2: Figure localized inositol lipid signalling), which then orchestrates the trafficking events responsible for the transformation of the early endosome into the multivesicular endosome (MVE) that end up as lysosomes. The PtdIns3,5P2 seems to act by stimulating TRPML1 channels to create the local domains of Ca2+ necessary to activate the endosome fusion to form the lysosomes. The formation of PtdIns3,5P2 is carried out by PhosphoInositide Kinase for five position containing a Fyve finger (PIKfyve), which has a number of protein interaction domains that enables it to interact with a number of other proteins to form a functional PIKfyve–ArPIKfyve–Sac3 (PAS) complex (Module 2: Figure PIKfyve activation). The N-terminal FYVE domain binds to the substrate PtdIns3P. The Cpn160_TCP1 domain interacts with the associated regulator of PIKfyve (ArPIKfyve), which forms a dimer and functions as a scaffolding protein to assemble the PAS complex. The C-terminal kinase domain is then positioned to catalyse the conversion of PtdIns3P into PtdIns3,5P2. The suppressor of actin 3 (Sac3), which is the mammalian orthologue of Fig4 in yeast, is a SAC domain lipid phosphatase that removes the 5-phosphate from PtdIns3,5P2 to form PtdIns3P. Sac3 is part of the PAS complex and has an important structural role in that it also helps to facilitate the kinase activity of PIKfyve.

Mutations in Sac3/Fig3 have been linked to Charcot–Marie–Tooth disease 4J.

PtdIns3P signalling cassette

PtdIns3P, which is formed by the phosphorylation of PtdIns by the Class III PtdIns 3-kinase, has a specific role to play in controlling a number of cellular processes.

One of the main functions of PtdIns3P is to regulate intracellular vesicle trafficking especially at the level of the early endosome (Module 2: Figure localized inositol lipid signalling). Most of the PtdIns3P is found on internal membranes and particularly those that function in vesicle trafficking:

PtdIns3P levels can be reduced through the action of the myotubularin family of 3-phosphatases (MTM1 and MTMR1–MTMR14). These enzymes are of interest because MTM1 mutations are responsible for X-linked recessive myotubular myopathy, MTMR2 mutations occur in Charcot–Marie–Tooth disease 4B and MTMR14 mutations have been linked to centronuclear myopathy (CNM).

PtdIns4P signalling cassette

For a long time, PtdIns4P was considered to be only a precursor of PtdIns4,5P2, but now there are indications that it might have a signalling role to regulate membrane and protein trafficking. One role is to control coat removal following the formation of clathrin-coated vesicles (Module 4: Figure scission of endocytic vesicles). Another role is to control trafficking through the endoplasmic reticulum/Golgi transport system (Module 2: Figure localized inositol lipid signalling), where it is formed by at least two PtdIns 4-kinases (PtdIns 4-Ks). The small GTPase ADP-ribosylation factor Arf-1 recruits the PtdIns 4-KIIIβ to the Golgi. At resting levels of Ca2+, this PtdIns 4-KIIIβ is kept inactive when bound to the Ca2+-sensing proteins calneuron-1 or calneuron-2. In response to a local pulse of Ca2+, the inhibitory calneurons are replaced by neuronal Ca2+ sensor 1 (NCS-1) that stimulates PtdIns 4-KIIIα to begin to produce the PtdIns4P necessary for vesicle formation.

In addition, the Golgi also has a PtdIns 4-kinase α that produces the PtdIns4P that associates with activating protein 1 (AP-1) to regulate the trafficking of clathrin-coated vesicles through the trans-Golgi network.

PtdIns5P signalling cassette

PtdIns5P has been implicated in several signalling events. It may modulate the PtdIns 3-kinase signalling pathway by interfering with the metabolism of the lipid messenger PtdIns3,4,5P3. It may also have a signalling role within the nucleus to control the response to DNA damage. PtdIns5P located on chromatin may provide an anchor to bind inhibitor of growth family, member 2 (ING2), which then results in acetylation of the p53 tumour suppressor.

Multipurpose inositol polyphosphate signalling pathway

The process of /1inositol phosphate metabolism generates a large number of inositol phosphates (/1Module 2: Figure inositol phosphate metabolism). Many of these are metabolic intermediates, but some have been implicated in a variety of control functions:

Ins1,4P2

There is some evidence to suggest that Ins1,4P2 may function within the nucleus to activate DNA polymerase. This inositol phosphate has also been implicated in Ca2+ signalling and cardiac hypertrophy.

Ins1,3,4P3

This inositol phosphate has an important signalling function as a negative regulator of the /1Module 2: Figure inositol phosphate metabolism) that controls the level of Ins3,4,5,6P4, which is an inhibitor of Ca2+-sensitive Cl- channels.

Ins1,3,4,5P4

This inositol phosphate, which is formed by phosphorylating Ins1,4,5P3, has been implicated in the control of Ca2+ entry into cells. However, its mode of action is unknown. Some clues concerning its action may come from the identification of GAP1IP4BP, which is one of the GTPase-activating protein (GAP) family (Module 2: Table monomeric G protein toolkit). GAP1IP4BP normally associates with the plasma membrane. When Ins1,3,4,5P4 binds to the membrane-anchoring domain of GAP1IP4BP, it causes this GAP to come off the membrane. A closely related GAP1m is located on the ER.

Ins1,4,5,6P4

Ins1,4,5,6P4, which can be formed by the phosphorylation of Ins1,4,5P3 or by the dephosphorylation of Ins1,3,4,5,6P5 (/1Module 2: Figure inositol phosphate metabolism), may function as a regulator of the Class I histone deacetylases (HDACs).

Ins3,4,5,6P4

Ins3,4,5,6P4 functions as an inhibitor of the Ca2+-sensitive Cl channels (CLCAs) in epithelial cells, which regulate salt and fluid secretion, cell volume homoeostasis and electrical excitability in neurons and smooth muscle cells. It appears to act by preventing Ca2+/calmodulin-dependent protein kinase II (CaMKII) from activating the channel.

Ins3,4,5,6P4 is formed by an /1Module 2: Figure inositol phosphate metabolism).

InsP6

There are a number of suggestions concerning the possible messenger role of InsP6. Cells contain high levels of InsP6 (approximately 15–100 μM), and much of this is probably not in solution, but is probably attached to the phospholipids in membranes through electrostatic interactions with bivalent cations. The level of InsP6 does not change much during acute stimulation, but its level can vary over the long term, as occurs during the cell cycle and during cellular differentiation. The various proposed messenger functions outlined below may thus depend upon highly localized fluctuations in specific cellular compartments.

Trafficking of vesicles

InsP6 may inhibit clathrin cage assembly by binding to adaptor protein (AP)-2 and -3. Such a role in membrane trafficking is also consistent with the observation that InsP6 binds to synaptotagmin by competing with the inositol lipid-binding site. It therefore seems that InsP6 may function as a negative regulator of endocytic vesicle traffic.

Such a possibility may explain the observation that GRAB, which is a guanine nucleotide exchange factor that acts on Rab3A, interacts with InsP6 kinase which converts InsP6 into InsP7.

Endocytosis

In insulin-secreting β-cells, InsP6 appears to promote dynamin I-mediated endocytosis through a mechanism that depends upon protein kinase C (PKC), which may act to inhibit the phosphoinositide phosphatase synaptojanin, thereby raising the level of PtdIns4,5P2 that has been implicated in vesicle dynamics.

Regulation of Ca2+ channels

InsP6 is present in many brain regions and appears to be elevated following neural activity. It may act by stimulating adenylyl cyclase to produce cyclic AMP, which then increases the activity of L-type Ca2+ channels. Alternatively, it may activate L-type Ca2+ channels through an inhibition of protein phosphatases. In vascular smooth muscle cells, InsP6 appears to act through a protein kinase C (PKC)-dependent pathway.

Regulation of gene transcription

Studies on yeast have revealed that ARG82, which is an Ins1,4,5P3 6-kinase that phosphorylates Ins1,4,5P3 within the nucleus to form Ins1,4,5,6P4, functions as a transcriptional regulator.

Regulation of mRNA export from the nucleus

Studies on yeast indicate that InsP6 formed by an inositol polyphosphate kinase located on the nuclear pores may facilitate the export of mRNA from the nucleus.

PP-InsP4

This diphosphorylated inositol phosphate has been considered as an orphan signal, as its function is unknown. Its level declines in response to either cyclic AMP or cyclic GMP.

InsP7 (PP-InsP5)

This diphosphoinositide, which is formed by the phosphorylation of InsP6 by an InsP6 kinase (see Step 14 in Module 2: Figure inositol phosphate metabolism), has been implicated in a number of functions. It may serve as an energy store as has been suggested for InsP8 ([PP]2-InsP4). InsP7 has been shown to regulate insulin signalling by inhibiting the activity of protein kinase B (PKB), which has a central role in the insulin receptor pathway responsible for regulating a large number of insulin-sensitive processes (Module 2: Figure insulin receptor). Through this inhibitory effect on insulin signalling, InsP7 may contribute to the onset of insulin resistance (Module 12: Figure insulin resistance).

InsP8 ([PP]2-InsP4)

InsP8 is a diphosphoinositide that may function as a high-energy store in that it can donate a high-energy phosphate to ADP to form ATP.

Nitric oxide (NO)/cyclic GMP signalling pathway

Nitric oxide (NO), which is often referred to as a gastrotransmitter, is a highly diffusible messenger that passes rapidly through cell membranes. It can act as a second messenger within its cell of origin or it can diffuse across membranes to act on neighbouring cells as a paracrine signalling agent. NO synthesis is carried out by nitric oxide synthase (NOS), which comes in three different forms: neuronal nitric oxide synthase (nNOS), inducible nitric oxide synthase (iNOS) and endothelial nitric oxide synthase (eNOS). These different isoforms share a similar NO synthetic reaction mechanism, which uses L-arginine as a substrate and O2 and NADPH as co-substrates to form NO. NOS regulation is very different for the three isoforms and is partly dependent on the way they are located in different parts of the cell. The action of NO is complex in that it can transmit information in markedly different ways. One of its actions is mediated through the cyclic GMP signalling pathway, where it stimulates soluble guanylyl cyclase to produce the cyclic GMP that can modify the properties of ion channels, protein phosphatases or cyclic nucleotide phosphodiesterase. NO can also act through the reactive nitrogen species (RNS) signalling pathways, whereby the NO alters the activity of a variety of protein targets through a nitrosylation reaction. This diverse NO/cyclic GMP signalling pathway operates to control the following cellular processes:

NO synthesis

Nitric oxide (NO) synthesis is carried out by NO synthase (NOS), of which there are three isoforms named either after the tissues where they were first discovered, i.e. neuronal nitric oxide synthase (nNOS) and endothelial nitric oxide synthase (eNOS) or by the way in which they are controlled, i.e. inducible nitric oxide synthase (iNOS) (Module 2: Figure NO and cyclic GMP signalling). The expression of these enzymes is not as restricted as their names imply, but are widely expressed and can coexist in many cell types. Even though these isoforms are regulated differently and have different cellular locations, they all seem to use the same NO synthetic reaction mechanism.

The excessive production of NO can have pathological consequences and has been linked to various disease states such as Huntington's disease, Alzheimer's disease and hypertension.

NO synthetic reaction mechanism

The different nitric oxide synthase (NOS) enzymes function as homodimers, which are arranged in a head-to-head orientation with the N-terminal oxidase domain of one monomer lined up alongside the C-terminal reductase domain of its neighbour. The substrates for the enzymatic reaction mechanism are L-arginine, oxygen and NADPH, which combine to form citrulline with the liberation of NO (Module 2: Figure NO synthase mechanism). NOS regulation is complicated because each isoform appears to be regulated by different mechanisms.

Endothelial nitric oxide synthase (eNOS)

As its name implies, endothelial nitric oxide synthase (eNOS) was first described in endothelial cells, where it generates NO in response either to agonists such as acetylcholine and bradykinin that elevate Ca2+ or to blood flow-induced shear stress. It is now evident that eNOS is expressed in many different cell types (lung epithelial cells, blood platelets, cardiac myocytes and hippocampal neurons). It has a complex regulation, which is very dependent on its attachment to caveolin, one of the proteins in caveolae (Module 6: Figure caveolae molecular organization), where it contributes to their signalling function. One of the key regulators of eNOS is Ca2+, which acts through calmodulin (CaM) to stimulate the enzyme to release NO (Module 2: Figure eNOS activation).

The NO released from endothelial cells diffuses out to regulate smooth muscle cell contraction and hence controls blood pressure, smooth muscle cell proliferation, aggregation of blood platelets and leucocyte adhesion. Studies on NO/cyclic GMP and cardiac hypertrophy have revealed that expression of eNOS in endothelial cells can inhibit hypertrophy in neighbouring cardiac cells. Given its central role in regulating so many cellular processes, alterations in the endothelial production of NO have been implicated in many disease states, such as hypertension, diabetes and hypercholesterolaemia.

Inducible nitric oxide synthase (iNOS)

Inducible nitric oxide synthase (iNOS), which is also known as immunocyte NOS, was first described in macrophages, where its expression is up-regulated by inflammatory mediators. Although macrophages are the main cells that express iNOS, it is also found in other cell types (cardiac cells, vascular smooth muscle cells and glial cells). Unlike the other isoforms, the activation of iNOS does not require an elevation of Ca2+. However, iNOS does bind calmodulin (CaM), which is essential for its activation. Since the enzyme is constitutively active, its primary regulation depends upon its induction by inflammatory receptors such as those that respond to interferon-γ (IFN-γ) or lipopolysaccharide (LPS). As large amounts of enzyme are produced, this is a high-output pathway capable of delivering NO for prolonged periods as part of the cells defence against invading micro-organisms. As such, it does not strictly function as a messenger. However, the large production of NO at the sites of inflammation will spill over to affect neighbouring cells. The large up-regulation of iNOS may account for the fall in blood pressure during endotoxic shock.

Neuronal nitric oxide synthase (nNOS)

Neuronal nitric oxide synthase (nNOS) was first described in neurons, but has since been found in other cell types such as skeletal muscle, which has an alternatively spliced variant that has a 34-amino-acid insert between exons 16 and 17. In both neurons and muscle, nNOS is closely associated with the plasma membrane where it binds to various proteins through its PDZ domains. In neurons, nNOS binds to the postsynaptic density (PSD) proteins such as PSD-93 and PSD-95 (Module 10: Figure postsynaptic density), whereas in skeletal muscle it interacts with α1-syntrophin (a binding partner of dystrophin).

Cyclic GMP signalling pathway

The cyclic GMP signalling pathway (Module 2: Figure NO and cGMP signalling) is governed by the second messenger cyclic GMP, which is synthesized by guanylyl cyclase (GC). The latter comes in two different forms: there are the soluble GC (sGC) and membrane-bound particulate guanylyl cyclases (pGCs). Many of the signalling functions of cyclic GMP are carried out by cyclic GMP-dependent protein kinase (cGK). In addition, cyclic GMP can act directly to open cyclic nucleotide-gated channels. Cyclic GMP hydrolysis is carried out by a cyclic GMP-specific phosphodiesterase (PDE5). Through these different signalling pathways, cyclic GMP functions to regulate a diverse collection of cellular processes. There are those where cyclic GMP mediates the action of NO:

In those cases where cyclic GMP is formed by the pGC, cyclic GMP acts independently of NO to regulate cellular processes such as phototransduction (Module 10: Figure phototransduction overview). Certain strains of Escherichia coli, which secrete the STa toxin, increase intestinal secretion and cause diarrhoea by activating the cyclic GMP signalling pathway by stimulating the particulate guanylyl cyclase C (pGC-C) receptor that is normally activated by guanylin (Module 7: Figure intestinal secretion).

NO/cyclic GMP and synaptic plasticity

The enzymes responsible for NO formation and its action are richly expressed in the nervous system. Although the precise function of NO is still debated, there are indications that it might function in synaptic plasticity by contributing to cerebellar cell long-term potentiation (LTP) at the parallel fibre/Purkinje cell synapse.

Guanylyl cyclase (GC)

Guanylyl cyclase (GC) is the enzyme that synthesizes cyclic GMP from ATP (Module 2: Figure NO and cyclic GMP signalling). It comes in two main forms, soluble GC (sGC), which is activated by NO, and membrane-bound particulate guanylyl cyclases (pGCs). The latter belongs to a group of single membrane-spanning receptors that use an enzyme to transduce information. In this case, it is the guanylyl cyclase region of the cytoplasmic domain that functions both as a transducer and an amplifier to generate the second messenger cyclic GMP.

Cyclic GMP hydrolysis

The enzyme that reverses the second messenger action of cyclic GMP by hydrolysing it to GMP is the cyclic GMP-specific phosphodiesterase (PDE5), which can be activated by cGKI, thus setting up a negative-feedback loop. Excessive signalling through the cyclic GMP signalling pathway will thus be curtailed through this ability of cyclic GMP to enhance its own hydrolysis. PDE5 is of considerable interest as it is the target of Viagra.

Cyclic GMP-dependent protein kinase (cGK)

There are two cyclic GMP-dependent protein kinases (cGKs): cGKI and cGKII. The cGKI comes in two alternatively spliced forms, cGKIα and cGKIβ. These cGKs are serine/threonine protein kinases that exist as homodimers that are held together by leucine zippers in their N-terminal domains, which fold over to inhibit the catalytic domain. The binding of cyclic GMP to a regulatory site induces a conformational change that relieves this inhibition, thus enabling the catalytic domain to phosphorylate its substrates.

cGK targets

These cGKs are targeted to specific sites in the cell. The N-terminal domain is responsible for targeting cGKIα and cGKIβ to specific cellular regions as is particularly important for smooth muscle relaxation mediated by nitric oxide (NO) and cyclic GMP (Module 7: Figure smooth muscle cell cyclic GMP signalling).

The cGKII has an N-terminal Gly-2 myristic acid residue that serves to target it to the plasma membrane.

cGK protein substrates

The cGMP-sensitive target proteins regulated by cGKs include inositol 1,4,5-trisphosphate receptor-associated cGKI substrate (IRAG), large conductance (BK) channels, cGMP-specific phosphodiesterase (PDE5), cerebellar G substrate, vesicle-associated membrane protein (VASP) and telokin.

Inositol 1,4,5-trisphosphate receptor-associated cGKI substrate (IRAG)

Inositol 1,4,5-trisphosphate receptor (InsP3R)-associated cGKI substrate (IRAG) is located on the endoplasmic reticulum (ER), where it appears to associate with the InsP3R. When it is phosphorylated by cGKIβ, IRAG acts to inhibit channel opening (Module 7: Figure smooth muscle cell cyclic GMP signalling).

Reactive nitrogen species (RNS) signalling

One of the ways by which NO functions in cells is through an S-nitrosylation reaction (/1Module 2: Figure NO and cGMP signalling). This covalent modification results from the addition of NO to reactive cysteine residues on specific target proteins. NO does not react directly with these proteins, but is first of all converted into /1reactive nitrogen species (RNS) that are then responsible for carrying out the /1nitrosylation reaction. These nitrosylation reactions are reversed by /1denitrosylation reactions. Many of the functions of NO are carried out by this nitrosylation-sensitive signalling pathway and there is growing evidence for nitrosylation dysfunction in disease.

Reactive nitrogen species (RNS)

The reactive nitrogen species (RNS) that carry out the nitrosylation reaction are formed when NO interacts with various acceptors such as superoxide radical (O2−•), cysteine (Cys), glutathione (GSH) or transition metal ions (Mn+, e.g. Fe3+ or Cu2+) (Module 2: Figure NO and cGMP signalling):

NO + O2 → ONOO (peroxynitrite)
NO + Mn+ → Mnn+ − NO
NO + GSH → GS − NO
NO + Cys → Cys − NO

The interaction between NO and O2−• forms the strong oxidant peroxynitrite (ONOO), which is very much more reactive than the two parent molecules. It can react with electron-rich groups such as protein sulfhydryls as part of the nitrosylation reaction to form the S-nitrosothiols (SNOs). Although peroxynitrite has a brief half-life (approx. 15 ms), it exists for long enough to diffuse through a cell and perhaps also to neighbouring cells enabling it to function as an intra- and inter-cellular messenger.

Low molecular mass thiols, such as glutathione (GSH), can also be nitrosylated by NO to form GS-NO.

Nitrosylation reaction

The S-nitrosylation reaction depends upon the transfer of NO from one of the reactive nitrogen species (RNS) to a peptidyl cysteine thiol group of target proteins (R-SH) (Module 2: Figure NO and cGMP signalling):

R-SH + ONOO → R-SNO
R-SH + Mn+ − NO → R-SNO
R-SH + GS − NO → R-SNO
R-SH + Cys − NO → R-SNO

Specificity is determined by the fact that the target proteins (R-SH) have hyperreactive cysteine groups where the thiol moiety exists as a thiolate anion due to the presence of positively charged amino acids in the immediate vicinity of the protein chain.

Denitrosylation reactions

Denitrosylation is carried out by two main denitrosylases (Module 2: Figure NO and cGMP signalling). One mechanism depends on the oxidoreductase thioredoxin (Trx) operating in concert with Trx reductase (TrxR). In its reduced state, the two cysteine residues in the highly conserved -Cys-Gly-Pro-Cys- active site are reduced and the hydrogens are used to release the NO as HNO and to convert the SNO group on the target protein into an SH group. During this denitrosylation reaction, a disulphide bond is formed between the vicinal cysteine residues at the active site. This oxidized Trx is then reduced by NADPH through the activity of the seleno-flavoprotein TrxR.

The other mechanism depends on S-nitrosoglutathione reductase (GSNOR), which is also known as GSH-dependent formaldehyde dehydrogenase and Class III alcohol dehydrogenase (ADH3). The GSNOR metabolizes the GSNO, which is one of the reactive nitrogen species (RNS), which is produced through two mechanisms. It is formed when NO interacts with GSH and is also released following the denitrosylation of target proteins (Module 2: Figure NO and cGMP signalling). The GSNOR uses NADH as an electron donor to irreversibly convert GSNO into GSNHOH. In the absence of GSNOR, the level of GSNO increases as does the amount of S-nitrosylated proteins, which suggest that these SNO proteins might be in equilibrium with GSNO and this might account for the fact that certain cellular proteins seem to by constitutively nitrosylated.

Redox signalling

Cells have evolved a sophisticated mechanism of intracellular signalling based on localized changes in the oxidation state of specific proteins. The internal environment of cells is normally highly reduced. Certain forms of stress are associated with an increase in the oxidative state, and this can induce apoptosis. It is also well known that certain phagocytic cells, such as neutrophils, can rapidly generate superoxide radical (O2−•) and hydrogen peroxide (H2O2) that are used to kill other cells during inflammatory responses. In addition to these pathological effects, there is increasing evidence that the redox system has been adapted to perform a variety of signalling functions and can modulate the activity of other signalling pathways. As such, they can control many cellular processes, including cell proliferation, apoptosis and cellular senescence. Some of these effects are exerted through a two-way interaction with Ca2+ signalling. For example, redox signalling can help to promote the tyrosine phosphorylation events that generate many signalling cascades and it can modulate the activity of the ryanodine receptors (RYRs) and inositol 1,4,5-trisphosphate receptors (InsP3Rs) that release Ca2+. Conversely, Ca2+ can stimulate redox signalling, particularly within the mitochondrion, indicating that there are dynamic interactions operating between these signalling pathways.

There are two main types of redox signalling. The first type is reactive oxygen species (ROS) signalling, which depends on the formation of ROS. The second type is reactive nitrogen species (RNS) signalling, which is carried out by RNS and is linked to the nitric oxide (NO)/cyclic GMP signalling pathway.

Reactive oxygen species (ROS) signalling

ROS signalling has all the hallmarks of a classical signalling mechanism. The second messengers are the reactive oxygen species (ROS) formed in response to many agonists. Reactive oxygen species (ROS) formation depends on the stimulation of a NADPH oxidase that removes an electron from NADPH and adds it to oxygen to create superoxide radical (O2−•), which is one of the ROS found in cells. Superoxide dismutase (SOD) rapidly converts the O2−• into hydrogen peroxide (H2O2), which is one of the main messenger molecules used by the redox signalling pathway (Module 2: Figure summary of redox signalling).

Rapid reactive oxygen species (ROS) metabolism ensures that H2O2, like other intracellular messengers, has a short half-life. This metabolism of H2O2 is carried out by a range of enzymes, including catalase, glutathione peroxidase (GPx) and peroxiredoxin (Prx). Since the inside of the cell is a highly reducing environment, the last two enzymes can draw upon a large reservoir of reducing equivalents in order to metabolize the ROS. The cell has a large redox buffer capacity in the form of glutathione, which functions to maintain the redox balance in the cell. The fact that H2O2 is metabolized so rapidly means that its site of action is highly localized close to its site of production.

The reactive oxygen species (ROS) messenger action of H2O2 depends upon its ability to react with the cysteine residues of a certain group of target proteins. The latter are marked out by virtue of having hyperreactive thiol groups that are rapidly oxidized to form a disulphide bond. The recovery from this oxidized state back to a fully reduced thiol group is carried out by the glutaredoxin and/or the thioredoxin systems.

Finally, the oxidized target proteins activate a number of oxidation-sensitive processes that bring about a number of cellular responses such as gene activation, modulation of ion channels and the activity of other signalling pathways [mitogen-activated protein kinase (MAPK) cascade and Ca2+ signalling]. It is therefore not surprising to find a role for redox signalling in proliferation and cancer. There is also a strong relationship between redox signalling and schizophrenia (Module 12: Figure schizophrenia).


Module 2: Table redox signalling components The major components of the redox signalling pathway

Component Comment
NADPH oxidases (NOXs)
 NOX1 Inducible enzyme found in colon and smooth muscle
 NOX2 (gp91phox) Major NOX in phagocytes
 NOX3 Foetal kidney
 NOX4 Widespread
 NOX5 Brain, spleen and sperm
 DUOX1 (dual oxidase 1) A Ca2+-sensitive isoform
 DUOX2 (dual oxidase 2) A Ca2+-sensitive isoform
NOX/DUOX regulatory factors
 p47phox
 p67phox
 p40phox
 p22phox
 NOXO 1
 NOXA 1
 Rac1/Rac2
ROS metabolism
 Superoxide dismutase (SOD)
 Catalase Localized in peroxisomes
 Glutathione peroxidase (GPx) Localized in cytosol and mitochondria
 Peroxiredoxin (Prx)
  Prx I–IV (2-Cys) I and II in cytosol; III in mitochondria; IV in endoplasmic reticulum
  Prx V (atypical 2-Cys)
  Prx VI (1 Cys)
Thiol-containing proteins/peptides
 Glutathione (GSH)
 Glutaredoxin (Grx)
 Thioredoxin (Trx)
  Trx-1
  Trx-2 Mitochondria-specific
Reductases
 Glutathione reductase
 Glutaredoxin reductase
 Thioredoxin reductase (TrxR)
 Sulphiredoxin (Srx) Catalyses reduction of hyperperoxidized proteins



Reactive oxygen species (ROS)

Reactive oxygen species (ROS) is a collective term that refers to those oxygen species [superoxide (O2−•), hydrogen peroxide (H2O2) and the hydroxyl radical (OH)] that are more reactive than ground-state oxygen [Module 2: Figure reactive oxygen species (ROS)]. These are sometimes considered synonymous with free radicals, but not all ROS are free radicals. The latter are defined as atoms or molecules that contain one or more unpaired electrons. With regard to signalling, it is the O2−• and H2O2 that appear to be the most important messengers. Also, there are indications that these two messengers may perform different functions.

There are two important sources of reactive oxygen species (ROS): one is at the plasma membrane and the other is within the mitochondria (Module 2: Figure sites of ROS formation).

Superoxide (O2−•)

The superoxide radical (O2−•) that is formed by the one-electron reduction of O2 is short-lived (half-life of 10−6 s) in that it rapidly dismutates into hydrogen peroxide (H2O2). This conversion can occur spontaneously, but the reaction is greatly accelerated by the enzyme superoxide dismutase (SOD).

Hydrogen peroxide (H2O2)

Much attention is focused on hydrogen peroxide (H2O2) because it appears to be the primary messenger molecule functioning in the redox signalling pathway. Since it has no unpaired electrons, H2O2 is not a free radical and thus is not a particularly powerful oxidizing agent. This means that it can function as a messenger by diffusing away from its site of action to interact with more distant targets. However, its sphere of influence is restricted by its short half-life, which is determined by the rapid reactive oxygen species (ROS) metabolism of H2O2. The H2O2 may also act as a paracrine signal that diffuses away from stimulated cells to alter the activity of neighbouring cells. Release of H2O2 during wound healing recruits leucocytes as part of the inflammatory response.

Hydroxyl radical (OH)

While H2O2 is relatively benign, it can be converted into the highly toxic hydroxyl radical (OH) through a reduction process catalysed by transition metals (Fe3+ or Cu2+). OH has a half life of 10−9 s indicative of its very high reactivity in that it reacts immediately and indiscriminately with the first molecule it finds. Much of the oxidative damage cause by ROS is mediated by OH.

Superoxide dismutase (SOD)

A family of metalloproteinases that converts superoxide radical (O2−•) into hydrogen peroxide (H2O2) (Module 2: Figure plasma membrane ROS formation):

2O2−• + 2H+ → H2O2 + O2

There are four families of SOD enzymes:

  • Copper/zinc-containing superoxide dismutases (CuZnSODs)
  • Manganese-containing superoxide dismutases (MnSODs)
  • Copper-containing superoxide dismutases (CuSODs)
  • Iron-containing superoxide dismutases (FeSODs)

Expression of MnSOD is regulated by the FOXO3a transcription factor (Module 4: Figure FOXO control mechanisms).

Mutation of SOD is the cause of amyotrophic lateral sclerosis (ALS), which is a debilitating and progressive neurological disease.

Reactive oxygen species (ROS) formation

Reactive oxygen species (ROS) are formed at two main sites: there is plasma membrane reactive oxygen species (ROS) formation and mitochondrial reactive oxygen species (ROS) formation (Module 2: Figure sites of ROS formation). This production of ROS appears to be highly localized suggesting the existence of reactive oxygen species (ROS) microdomains.

Plasma membrane reactive oxygen species (ROS) formation

A large number of receptors responding to external signals stimulate the formation of reactive oxygen species (ROS), including cytokine receptors (tumour necrosis factor α, interleukin-1 and interferon-γ), growth factor receptors [platelet-derived growth factor (PDGF), epidermal growth factor (EGF) and basic fibroblast growth factor (bFGF) receptors] and G protein-coupled receptors (GPCRs) (5-hydroxytryptamine, bradykinin, angiotensin II, thrombin and endothelin receptors). Just how all these receptors are coupled to the formation of ROS is still somewhat uncertain, but there is increasing evidence that the mechanism might resemble that found in phagocytic cells.

In phagocytes, external signals acting through cell-surface receptors stimulate the PtdIns 3-kinase signalling pathway to form the lipid messenger PtdIns3,4,5P3 (Module 2: Figure PtdIns 3-kinase signalling), which then acts through Rac to stimulate NADPH oxidase. This multicomponent enzyme uses NADPH as an electron donor to carry out the first step of ROS formation i.e. the reduction of oxygen to superoxide radical (O2−•) (Module 2: Figure plasma membrane ROS formation). In addition to this activation through Rac, it seems that the enzyme can also be regulated through diacylglycerol (DAG) and Ca2+ acting through protein kinase C (PKC), which phosphorylates p47phox, one of the cytoplasmic components of NADPH oxidase, to bring about the assembly of the multicomponent complex. Non-phagocytic cells that generate ROS for signalling use a similar, but genetically distinct, NADPH oxidase called Nox1, which plays a significant role in redox signalling in proliferation and cancer.

In phagocytes, the O2−• is released to the outside of the cell, where it is able to attack invading micro-organisms, whereas, in non-phagocytic cells, it functions as a messenger to activate a signalling cascade on the inside of the cell (Module 2: Figure plasma membrane ROS formation). Superoxide dismutase (SOD) rapidly converts the O2−• into hydrogen peroxide (H2O2), which appears to be the primary messenger molecule of this redox signalling pathway.

NADPH oxidase

The NADPH oxidase (NOX/DUOX) family (Module 2: Table redox signalling components) consists of a number of enzymes with different cellular locations.

NOX2, also known as gp91phox, has been described best for phagocytes, where it is made up of a number of subunits. The catalytic component is cytochrome b558, which is a heterodimer formed from gp91phox and p22phox. In addition to this heterodimer, which is located in the membrane, there are cytoplasmic components (e.g. p47phox, p67phox, Rap1A and Rac) that play a role in regulating enzyme activity. The DUOX (dual oxidase) enzymes are sensitive to Ca2+ and play an important role in interacting with the Ca2+ signalling pathway (Module 2: Figure ROS effects on Ca2+ signalling).

There is a strong relationship between redox signalling and schizophrenia during which the induction of NOX2 may play an important role in triggering the formation of the peroxynitrite that acts to inhibit NMDA receptor function (Module 12: Figure schizophrenia).

Mitochondrial reactive oxygen species (ROS) formation

Most of the electrons that enter the electron transport chain are transferred to oxygen in an orderly manner, but there is always a 1–2% leakage during which an electron is transferred directly to oxygen to form superoxide (O2−•) and this is the source of mitochondrial reactive oxygen species (ROS) (Module 5: Figure mitochondrial Ca2+ signalling). This orderly electron transfer to oxygen occurs during energy metabolism, where oxygen is reduced to water by accepting four electrons (e) from cytochrome c oxidase:

O2 + 4e + 4H+ → 2H2O

Mitochondrial energy metabolism is inherently dangerous, because the electron transport chain is somewhat leaky in that some of the molecular oxygen is diverted into the formation of the superoxide radical (O2−•), which is then converted into hydrogen peroxide (H2O2) and the hydroxyl radical (OH). Such mitochondrial ROS may play an important role in apoptosis by acting synergistically with Ca2+ to stimulate the formation of the mitochondrial permeability transition pore (mPTP) (Module 5: Figure mitochondrial Ca2+ signalling). The formation of mitochondrial ROS appears to be highly localized in that small superoxide flashes have been recorded in single mitochondria. This is another example of reactive oxygen species (ROS) microdomains.

ROS formation can also have a regenerative component in that there is a process of ROS-induced ROS release (RIRR).

ROS-induced ROS release (RIRR)

The generation of ROS by mitochondria might be a regenerative process in that a local release of ROS in cardiac myocytes causes a rapid mitochondrial depolarization due to formation of a permeability transition pore (mPTP) and a concomitant increase in intrinsic ROS production, i.e. a process of ROS-induced ROS release (RIRR). This RIRR has a regenerative property in that it can spread through long chains of adjacent mitochondria in the form of a wave.

Redox balance

The redox balance of the cell is maintained by energy metabolism, primarily the pentose phosphate shunt that feeds reducing equivalents into the cell in the form of NADPH. The latter is then used to maintain redox buffers such as glutathione (GSH) (Module 2: Figure GSH/GSSG couple) and thioredoxin (Trx) in their reduced forms. The concentration of GSH in the cytoplasm lies within the 1–10 mM range, with over 99% existing as the reduced GSH form. The 2GSH/GSSG redox couple is thus a measure of the redox balance in the cell. A similar balance exists for the oxidized and reduced forms of TRX. The reduced GSH and Trx are used for a number of reductive processes such as the metabolism of hydrogen peroxide by glutathione peroxidase (GPx) (Module 2: Figure H2O2 metabolism) or as a source of reducing equivalents for the glutaredoxin system (Module 2: Figure recovery of protein oxidation). The GSSG (the oxidized form of glutathione) is converted back into GSH by glutathione reductase.

Similar redox control enzymes and buffers are located in both the mitochondria and within the lumen of the endoplasmic reticulum (ER). Like the cytoplasm, the mitochondrial matrix maintains a reducing environment and uses similar enzymatic mechanisms to control the ROS emanating from the electron transport chain. The ER, however, is somewhat different in that the GSH/GSSG ratio is close to 1 and this more oxidizing environment is necessary for the formation of the disulphide bonds that are an integral component of the extracellular proteins that are processed and packaged within the ER. Protein disulfide isomerase (PDI) catalyses the formation of protein disulfide bonds within the ER. This oxidative process of disulfide bond formation utilises oxidoreductases such as ER oxidoreductin 1 (ERO1), which acts to transfer oxidizing equivalents to PDI and is thus reduced during the thiol-disulfide exchange reaction.

Since the cytoplasm and the ER lumen have different redox potentials, there is a redox potential gradient across the ER membrane and this might be used to modulate Ca2+signalling by altering the activity of the ion channels that release Ca2+.

Reactive oxygen species (ROS) metabolism

Like all other intracellular signalling molecules, ROS are metabolized rapidly. Superoxide dismutase (SOD) rapidly converts superoxide radical (O2−•) into hydrogen peroxide (H2O2), which is then metabolized by a number of enzyme systems including catalase, glutathione peroxidase (GPx) and peroxiredoxin (Prx) (Module 2: Figure H2O2 metabolism).

With so many enzyme systems co-operating to metabolize H2O2, it is likely that this messenger will have a highly restricted sphere of influence localized to its site of production either at the plasma membrane or within the mitochondrion.

Catalase

Catalase is a haem-containing protein that decomposes hydrogen peroxide (H2O2) to water and oxygen (Module 2: Figure H2O2 metabolism).

Most of the catalase in cells is found in peroxisomes, thus restricting the role of the enzyme in dealing with the H2O2 generated during the redox signalling mechanism at the plasma membrane.

Glutathione (GSH)

Glutathione (GSH) is a redox buffer that regulates the cellular redox balance (Module 2: Figure GSH/GSSG couple). GSH is a tripeptide consisting of glutamic acid, cysteine and glycine. It is synthesized by two enzymes. First, there is glutamate cysteine ligase (GCL), which is made up of two subunits: a GCL catalytic subunit (GCLC) and a GCL modifier subunit (GCLM). The GCLC carries out an ATP-dependent condensation reaction between cysteine and glutamate to form gamma-glutamylcysteine. Secondly, a GSH synthetase (GSS), which is also known as GSH S-transferase (GST), adds a glycine to the dipeptide gamma-glutamylcysteine to form GSH.

A dysregulation of GSH metabolism has been implicated in schizophrenia (Module 12: Figure schizophrenia) and in the role of inflammation in Alzheimer's disease (Module 12: Figure inflammation in Alzheimer's disease).

Glutathione peroxidase (GPx)

The glutathione peroxidase (GPx) family uses the reducing power of glutathione to convert H2O2 into water (Module 2: Figure H2O2 metabolism). GPx is one of the detoxifying enzymes induce by the nuclear response factor-1 (NRF-1).

Peroxiredoxin (Prx)

The peroxiredoxins (PrxI to PrxIV) are a family of small antioxidant proteins that function to metabolize hydrogen peroxide (H2O2) to water, thus curtailing its messenger action. PrxI and PrxII are cytosolic, whereas Prx III is found on mitochondria and PrxIV is on the endoplasmic reticulum. The operation of the catalytic cycle goes through the following steps (Module 2: Figure peroxiredoxin catalytic cycles):

  1. H2O2 is generated near the plasma membrane when the PtdIns3,4,5P3 (PIP3) formed by receptor activation stimulates NADPH oxidase.
  2. H2O2 interacts with the reduced cysteine residues (Cys-SH) in the N-terminal regions of the thioredoxin (Trx) dimers to form two oxidized sulphenic residues (Cys-SOH).
  3. The Cys-SOH can then interact with the conserved Cys-SH on the C-terminal regions of the neighbouring dimer to form two intermolecular disulphides.
  4. The Prx disulphide is converted back into the reduced form by Trx, which is regenerated by thioredoxin reductase (Module 2: Figure recovery of protein oxidation).
  5. The sulphenic residues formed by Reaction 2 can undergo hyperperoxidation by interacting with further molecules of H2O2 to form the sulphinic acid residues.
  6. This hyperperoxidation reaction can be reversed by a reaction that requires ATP catalysed by the enzyme sulphiredoxin (Srx).
  7. The phosphorylated intermediate is reduced back to the reduced form of Prx by the Trx system.
  8. The Prx system is very efficient at limiting the size of reactive oxygen species (ROS) microdomains. However, cell-surface receptors are capable of enlarging this microdomain by inducing a local inactivation of Prx. The activated receptor, perhaps acting through the non-receptor protein kinase Src, brings about an inactivation of the Prx molecules in the local vicinity by phosphorylating Prx on Tyr-194 thus enabling the plume of H2O2 to spread away from the receptor.

The reducing equivalents derived from thioredoxin (Trx) are used to regenerate Prx-(SH)2:

Prx-S2 + Trx-(SH)2 → Prx-(SH)2 + Trx-S2

When Prx1 in mice is knocked out, animals develop haemolytic anaemia and malignant cancers.

Reactive oxygen species (ROS) messenger action

The primary action of hydrogen peroxide (H2O2) is to reversibly oxidize a variety of target proteins with a high degree of specificity (Steps 1–7 in Module 2: Figure reversible and irreversible ROS oxidations):

  1. The primary action of H2O2 is to oxidize the hyperreactive cysteine to form a sulphenic acid group (-SOH), which can be metabolized further along a number of pathways.
  2. The sulphenic acid residue can interact with nitrogen on a neighbouring serine residue to form an intramolecular cyclic sulphenyl amide as occurs during the oxidation of protein tyrosine phosphatases (Module 2: Figure ROS formation and action).
  3. The sulphenic acid residue can be converted into an intramolecular disulphide bond with the elimination of water.
  4. The sulphenic acid residue can interact with glutathione (GSH) to form an intermolecular disulphide bond.
  5. The sulphenic acid residue can undergo hyperperoxidation by interacting with another molecule of H2O2 to form a sulphinic acid intermediate (Cys-SO2H).
  6. The sulphinic acid intermediate undergoes further hyperperoxidation to form the sulphonic acid intermediate (Cys-SO3H).
  7. The sulphinic acid group (Cys-SO2H) can be reduced by a reaction that requires ATP and is catalysed by the enzyme sulphiredoxin (Srx).

The various oxidated intermediates can be converted back into the initial reduced state by either the thioredoxin (Trx) or the glutaredoxin (Grx) system (Module 2: Figure recovery of protein oxidation).

These oxidation processes can occur through a number of mechanisms. What is remarkable about this process is its specificity. Only a subset of proteins are modified, and within these there is a high degree of specificity in that only certain thiols are modified. How is it that an oxidizing agent such as H2O2 is able to seek out and selectively modify specific target proteins? The answer lies in the fact that proteins vary considerably in their sensitivity to mild oxidizing agents, such as H2O2. Most of the cysteine residues in proteins have a high acidic constant (i.e. pKa values of approximately 8.5), which means that they are resistant to attack by H2O2. However, some of the cysteine residues, particularly those located next to positively charged amino acids, have pKa values between 4 and 5 and thus exist as a thiolate anion (Cys-S), which is very vulnerable to oxidation, and these have been referred to as hyperreactive cysteine residues (the S group marked with an asterisk in Module 2: Figure reversible and irreversible ROS oxidations).

The sulphenic acid intermediate is somewhat unstable and can be converted into a number of intermediates by either eliminating water or causing it to interact with GSH (Reactions 2, 3 and 4 in Module 2: Figure reversible and irreversible ROS oxidations). In addition, sulphenic acid can be oxidized further by H2O2 to sulphinic acid and sulphonic acid. The formation of sulphinic acid is reversible through a reaction that requires ATP and sulphiredoxin (Srx). However, the final step to form sulphonic acid is irreversible and can result in serious damage and has been implicated in the ageing process. One idea is that this irreversible change may accumulate with time and the progressive damage may result in ageing.

There is a suggestion that the hyperperoxidation reactions that lead to the irreversible oxidation may be avoided by an internal reaction whereby the sulphenic acid is rapidly converted into a sulphenylamide species by interacting with the main-chain nitrogen atom of an adjacent serine residue (Step 2 in Module 2: Figure reversible and irreversible ROS oxidations). The formation of this sulphenyl-amide intermediate protects against further oxidation, and the enzyme can be reactivated by converting the intermediate back into a thiol group via a mixed disulphide reaction involving GSH.

Two of the main signalling molecules whose activities are reduced by oxidation are the protein tyrosine phosphatases and phosphatase and tensin homologue deleted on chromosome 10 (PTEN) (Module 2: Figure ROS formation and action). An analysis of tyrosine phosphatase structure and function reveals the presence of a hyperreactive cysteine residue in the catalytic domain that is sensitive to oxidants resulting in inactivation of the enzyme. Other examples of proteins that have such hypersensitive residues include the cell cycle regulatory enzyme Cdc25C, the Ca2+-release channels ryanodine receptors (RYRs) and inositol 1,4,5-trisphosphate receptors (InsP3Rs). The specificity with which this redox signalling system exerts its effects therefore depends on the fact that H2O2 will only modify hyperreactive cysteine residues that exist in these target proteins of the different oxidation-sensitive processes.

Recovery of oxidation-sensitive processes

Like other signalling pathways, there are mechanisms in place for the recovery of oxidation-sensitive processes based on the thioredoxin (Trx) and glutaredoxin (Grx) systems (Module 2: Figure recovery of protein oxidation). Although the two systems have much in common, there are some differences, not least of which are their substrate specificity and the kinds of disulphide bonds that they can reduce. For example, Trx is more effective at reducing protein tyrosine phosphatase 1B than is Grx.

Trx and Grx exist in a reduced or oxidized state, and it is the former that enables them to reduce their substrates. In doing so, they become oxidized and have to be converted back into a reduced state by the Trx and Grx systems respectively. In the case of Trx, this is carried out by thioredoxin reductase. The Grx system is somewhat more complicated in that it depends upon glutathione that is regenerated by a glutathione reductase.

The level of Trx is markedly elevated during rheumatoid arthritis and this may influence the rate of secretion of matrix metalloproteinases (MMPs).

Thioredoxin (Trx)

Thioredoxin (Trx) is a redox buffer that operates together with thioredoxin reductase (TrxR) and NADPH to regulate the redox state of many different proteins (Module 2: Figure recovery of protein oxidation). It is a major dithiol reductase that functions to re-activate signalling proteins such as tyrosine phosphatases and PTEN (Module 2: Figure ROS formation and action). One of its other functions is to re-juvenate the peroxiredoxins (Prxs), which are a family of small antioxidant proteins that function to metabolize hydrogen peroxide (H2O2) to water, thus curtailing its messenger action (Module 2: Figure peroxiredoxin catalytic cycles). Trx also plays a role in the denitrosylation reaction (Module 2: Figure NO and cGMP signalling).

One of the functions of Trx is to regulate apoptosis signal-regulating kinase 1 (ASK1). Reduced Trx-(SH)2 is known to bind to ASK1, but when it is oxidized to Trx-S2, the ASK1 is released, and proceeds to induce apoptosis.

Thioredoxin-2 (Trx-2) is a mitochondrial-specific member of the Trx family. It functions together with mitochondrial thioredoxin reductase 2 (TrxR2) to regulate the mitochondrial membrane potential and contributes to the inhibition of apoptosis by regulating the nitrosylation levels of the mitochondrial caspases. Activation of the FAS signalling system can increase the activity of Trx2, which switches the balance towards the more active denitrosylated form of the caspases.

Expression of the thioredoxin-interacting protein (Txnip), which binds to Trx and inhibits its participation in the denitrosylation reaction, is reduced in schizophrenia and thus contributes to the increase in redox signalling in schizophrenia (Module 12: Figure schizophrenia).

Thioredoxin reductase (TrxR)

Thioredoxin reductase (TrxR) together with thioredoxin (Trx) is an important oxidoreductase system that has a significant role in regulating the redox state (Module 2: Figure recovery of protein oxidation).

Trx reductase is unusual in that it contains selenocysteine (SeCys) located in the C-terminal active site, which has a highly conserved -Gly-Cys-SeCys-Gly- sequence. The N-terminal region contains flavin–adenine dinucleotide (FAD). The enzyme operates by transferring electrons from NADPH to FAD and then on to the active site in the C-terminus.

The level of Trx reductase is greatly increased in various tumour cells, where it may play an important role in inhibiting apoptosis by regulating the activity of Trx that inhibits apoptosis signal-regulating kinase 1 (ASK1). An enhanced level of TrxR may prevent apoptosis by ensuring that the Trx remains reduced to ensure that ASK1 remains inactive.

The ability of this thioredoxin system to reverse redox signalling is inhibited by Ca2+, which acts to convert a large proportion of the reduced Trx-(SH)2 into its oxidized form. Such an action would enhance the growth-promoting activity of the redox signalling system.

Thioredoxin-2 (Trx-2) is a mitochondria-specific member of the Trx family. It functions to regulate the mitochondrial membrane potential and contributes to the inhibition of apoptosis.

Glutathione reductase

Glutathione reductase is responsible for converting oxidized GSSG back into the reduced GSH (Module 2: Figure recovery of protein oxidation):

GSSG + NADPH + H+ → 2GSH + NADP+

Oxidation-sensitive processes

The redox signalling pathway acts to regulate a variety of oxidation-sensitive processes (Module 2: Figure plasma membrane ROS formation). The cellular proteins that are sensitive to oxidation are those that contain hyperreactive cysteine residues. One of the main functions of redox signalling is to modulate the activity of other signalling systems that contain protein components with such hyperreactive cysteine residues, e.g. Ca2+-releasing channels [inositol 1,4,5-trisphosphate receptors (InsP3Rs) and ryanodine receptors (RYRs)], transcription factors, protein tyrosine phosphatases and by activating the mitochondrial permeability transition pore (MTP) to induce apoptosis. As so many key signalling functions are being influenced, it is not surprising to find that there is a role for redox signalling in many cellular processes:

Redox signalling in proliferation and cancer

One of the main actions of redox signalling is to control growth. During the action of many growth factors there is an increase in the production of hydrogen peroxide (H2O2) (Module 2: Figure plasma membrane ROS formation), which facilitates growth factor signalling by inhibiting tyrosine phosphatases and the tumour suppressor phosphatase and tensin homologue deleted on chromosome 10 (PTEN). The latter inhibits the hydrolysis of PtdIns3,4,5P3 (PIP3), which functions in cell migration, proliferation and survival. H2O2 inactivates PTEN by inducing a disulphide bond to form between Cys-124 in the active site and Cys-71. This disulphide is specifically reversed by thioredoxin. A reversible inactivation of PTEN may thus contribute to the accumulation of PIP3, which will thus set up a positive-feedback loop, since the formation of PIP3 is responsible for stimulating the production of H2O2.

Many cancer cells are known to use ROS to control their proliferation. There are five human homologues of Nox1. When Nox1 is overexpressed in fibroblasts, there is an increase in proliferation and tumour formation. The addition of antioxidants can reduce the growth of cancer cells.

Many cancer cells, like a number of stem cells, have enhanced ROS defences in the form of elevated levels of GSH and thioredoxin, which make them particularly resistant to apoptosis. Thioredoxin may also play a role in increasing the expression of the hypoxia-inducible factor 1α (HIF-1α), resulting in an increase in vascular endothelial growth factor (VEGF) and tumour angiogenesis.

Redox signalling in apoptosis

In addition to playing a role in cell proliferation, there is a darker side to reactive oxygen species (ROS) in that they can also activate apoptosis. One of the actions of redox signalling is that it contributes to Ca2+-induced apoptosis at the level of the mitochondria. The uptake of Ca2+ and the resulting increase in ROS formation act synergistically to open the mitochondrial permeability transition pore (MTP) (Module 5: Figure mitochondrial Ca2+ signalling). ROS can also increase apoptosis by stimulating the acidic sphingomyelinases (SMases) that produce ceramide (Module 2: Figure sphingomyelin signalling).

Cells have different ways of suppressing this ROS-induced apoptosis. One mechanism is carried out by PtdIns 3-kinase signalling during which protein kinase B (PKB) plays a prominent role through the Forkhead box O 3a (FOXO3a) transcription factor, which acts by increasing the amount of manganese superoxide dismutase (MnSOD) to provide greater protection against ROS (Module 4: Figure FOXO control mechanisms).

An overactive redox signalling system may contribute to the increase in neuronal cell death that characterizes Alzheimer's disease (Module 12: Figure astrocyte-induced neuronal death) and Down's syndrome.

Redox signalling and DNA damage

One of the major pathological consequences of excess ROS formation is DNA damage. When this damage occurs during the G1 phase of the cell cycle, specific repair mechanisms function to repair the damage and they also induce the process of G1 checkpoint signalling to DNA double-strand breaks (DSBs) (Module 9: Figure G1 checkpoint signalling).

Redox signalling in vascular homoeostasis

Hydrogen peroxide (H2O2) may function as an endothelium-derived hyperpolarizing factor (EDHF) that diffuses across to relax neighbouring smooth muscle cells. This action may be particularly important in regulating the tone of cerebral arteries.

Redox signalling and gene transcription

The redox signalling pathway has been implicated in the control of gene transcription, particularly with regard to the activation of nuclear factor κB (NF-κB). Whether this activation is due to a direct modulation of the transcription factor by messengers such as H2O2 or indirectly through activation of other signalling pathways remains to be determined.

A large number of other transcription factors and proteins [activating protein 1 (AP-1), DJ-1, specificity protein 1 (Sp1), c-Myb, p53 and Egr-1] are redox-sensitive. Many of these have a highly conserved cysteine residue located within their DNA-binding domains that has to be reduced in order for the factor to bind DNA (Module 4: Figure SRF and AP-1). In theory, therefore, such factors would be inhibited by oxidation. There is a nuclear redox factor 1 (Ref-1) that functions to control transcription by reducing this cysteine.

HDAC oxidation is an important mechanism used to control chromatin remodelling and gene transcription.

DJ-1

DJ-1 has come to prominence as one of the autosomal recessive genes responsible for familial Parkinson's disease (PD). It is a small protein (189 amino acids) and it appears to function as a dimer. It also has a propensity to bind to other proteins such as Parkin and mortalin. The precise function of DJ-1 is not fully understood, but it seems to play a role either as an antioxidant protein or to enable cells to respond to oxidant stress. One of its functions is to increase the expression of the uncoupling protein 4 (UCP4) and uncoupling protein 5 (UCP5) channels to leak H+ that depolarizes the inner mitochondrial membrane potential to reduce respiration and the damaging effects of excess Ca2+ (Module 12: Figure signalling pathways in Parkinson's disease).

The expression of DJ-1 is regulated by specificity protein 1 (Sp1), through mechanisms that are tightly regulated by histone deacetylases (HDACs) that deacylate histones H3 and H4 and suppress transcription. Inhibition of these HDACs by phenylbutyrate has been shown to alleviate some of the symptoms of PD by increasing the expression of DJ-1.

One of the actions of DJ-1 is to stabilize nuclear factor erythroid 2 related factor 2 (NRF-2) by inhibiting its interaction with Keap1 and thus reducing the degradation of NRF-2 (Module 4: Figure p62 function).

Redox factor 1 (Ref-1)

Ref-1 plays a role in the nucleus to promote gene transcription (Module 4: Figure SRF and AP-1) and to protect cells against oxidative stress. There is a possibility that it might also have a function within the cytoplasm to reduce the Rac-1-regulated production of reactive oxygen species (ROS).

Redox signalling and modulation of Ca2+ signalling

There are reciprocal interactions operating between the redox and Ca2+ signalling pathways. There are Ca2+ signalling effects on redox signalling and there are redox signalling effects on Ca2+ signalling.

Ca2+ signalling effects on redox signalling

One of the actions of Ca2+ is to enhance redox signalling by interfering with the recovery of the oxidation-sensitive processes. Ca2+ acts by turning down the thioredoxin system by inhibiting the thioredoxin reductase that normally switches off redox signalling.

Redox signalling effects on Ca2+ signalling

There are numerous examples of redox signalling acting to enhance Ca2+ signalling. For example, the two Ca2+ release channels ryanodine receptors (RYRs) and the InsP3 receptors (InsP3Rs) can be activated by oxidation of key cysteine residues. In the case of the latter, the oxidizing agent thimerosal faithfully reproduces the Ca2+ transients normally induced by sperm fusion during fertilization (Module 2: Figure thimerosal-induced Ca2+ signalling). Redox signalling can also influence Ca2+ homoeostasis by oxidizing the plasma membrane Ca2+-ATPase (PMCA), which will result in an elevation of the resting level of Ca2+. Such a mechanism may be particularly relevant for the effect of inflammation in Alzheimer's disease (Module 12: Figure inflammation and Alzheimer's disease).

Another example is that hydrogen peroxide (H2O2) can markedly enhance Ca2+ signalling by inhibiting the Src homology 2 (SH2) domain-containing protein tyrosine phosphatase-1 (SHP-1), which normally acts to keep in check the protein tyrosine phosphorylation cascade that occurs during B cell receptor (BCR) activation (Module 2: Figure ROS effects on Ca2+ signalling). This is a reciprocal interaction because it is the increase in Ca2+ that activates the formation of H2O2 through a Ca2+-dependent activation of dual oxidase (DUOX). This is a good example of the cross-talk that can exist between signalling pathways.


Module 2: Table MAPK signalling toolkit Mitogen-activated protein kinase (MAPK) signalling toolkit

Component Comment
Transducers
 GCK Germinal centre kinase
 GLK GCK-like kinase
 HPK1 Haematopoietic progenitor kinase 1
 MST1 Mammalian Ste20-like protein kinase
MAPKKKs
 ASK1 Apoptosis signal-regulating kinase 1
 ASK2 Apoptosis signal-regulating kinase 2
 DLK Dual leucine-zipper-bearing kinase
 MEKK1 MAPK/ERK kinase kinase 1
 MEKK2 MAPK/ERK kinase kinase 2
 MEKK3 MAPK/ERK kinase kinase 3
 MEKK4 MAPK/ERK kinase kinase 4
 MLK1 Mixed lineage kinase 1
 MLK2 Mixed lineage kinase 2
 MLK3 Mixed lineage kinase 3
 Mos
 Raf-1
 B-Raf
 PAK p21-activated kinase
 TAK1 TGFβ-activated protein kinase 1
 Tpl2 Tumour progression locus 2
MAPKKs
 MEK1 MAPK/ERK kinase 1
 MEK2 MAPK/ERK kinase 2
 MKK3 MAPK kinase 3
 MKK4 MAPK kinase 4
 MEK5 MAPK/ERK kinase 5
 MKK6 MAPK kinase 6
 MKK7 MAPK kinase 7
MAPKs
 ERK1 Extracellular-signal-regulated kinase 1
 ERK2 Extracellular-signal-regulated kinase 2
 ERK3-related
 ERK3 Extracellular-signal-regulated kinase 3
 ERK5 Extracellular-signal-regulated kinase 5
 ERK7 Extracellular-signal-regulated kinase 7
 ERK8 Extracellular-signal-regulated kinase 8
 JNK1 c-Jun N-terminal kinase 1
 JNK2 c-Jun N-terminal kinase 2
 JNK3 c-Jun N-terminal kinase 3
 p38α
 p38β
 p38γ
 p38δ
MAPK scaffolding proteins
 β-Arrestin-2
 JIP1 JNK-interacting protein 1
 JIP2 JNK-interacting protein 2
 JIP3 JNK-interacting protein 3
 KSR1 Kinase suppressor of Ras 1
 KSR2 Kinase suppressor of Ras 2
 MP1 MEK partner 1
 MKPX Phosphatase that may also function as a scaffold
 SKRP1 Stress-activated protein kinase (SAPK) pathway-regulating phosphatase 1
MAPK target proteins
 ATF2 Activating transcription factor 2
 Cytoplasmic PLA2 Cytoplasmic phospholipase A2
 ETS
 Elk-1
 c-Jun
 Jun-B
 Jun-D
 MAPKAP-kinase MAPK-activated protein kinase
 MEF2 Myocyte enhancer factor 2
 MSK1 Mitogen- and stress-activated protein kinase 1
 MSK2 Mitogen- and stress-activated protein kinase 2
 SAP-1 Stomach cancer-associated protein tyrosine phosphatase-1
MAPK phosphatases
 MKP-1 MAPK phosphatase-1 (Module 2: Figure ERK signalling)
 MKP-2 MAPK phosphatase-2
 MKP-3 MAPK phosphatase-3, acts specifically to dephosphorylate ERK2
 MKP-4 MAPK phosphatase-4
 MKP-5 MAPK phosphatase-5
 VHR VH1-related
 hVH3/B23
 hVH5/M3/6
 PAC1
 Pyst2
 STEP Striatal-enriched protein tyrosine phosphatase; dephosphorylates ERK in neurons (Module 10: Figure medium spiny neuron signalling).



Hydrogen sulfide (H2S)

Hydrogen sulfide (H2S) is one of the gasotransmitters, which include nitric oxide (NO) and carbon monoxide (CO), that functions to regulate numerous cellular processes. The synthesis of H2S is catalysed by a number of pathways (Module 2: Figure H2S generation):

  1. Cystathionine β-synthase (CBS) acts on L-cysteine to release H2S. The CBS is mainly localized in the brain (astrocytes and neurons) but not in the microglia.
  2. Cystathionine γ-lyase (CSE) liberates H2S from L-cysteine using vitamin B6 as a cofactor. The CSE is expressed in the cardiovascular system.
  3. The two-step process begins with L-cysteine and α-ketoglutarate being converted into 3 mercaptopyruvate (3MP) using vitamin B6 as a cofactor. β-Mercaptopyruvate sulfurtransferase (3MST) then liberates H2S from 3MP. D-cysteine can also be used to generate H2S following its conversion into 3MP by D-amino acid oxidase (DAO). The MST is localized in various neurons (cerebellar Purkinje neurons, hippocampal and mitral neurons), aortic endothelium and smooth muscle cells.

The way in which the formation of H2S is regulated is unclear, but there is evidence that an elevation in cytosolic Ca2+ can inhibit some of the enzymes such as CSE, 3MST and CAT (Module 2: Figure H2S generation).

H2S acts through an S-sulfhydration reaction (Module 2: Figure H2S generation) to modify specific cysteine residues in target proteins such as actin, β-tubulin and glyceraldehyde dehydrogenase (GAPDH). Another target is protein tyrosine phosphatase 1B (PTP1B), which is located on the surface of the endoplasmic reticulum. One of its functions is to regulate the activity of protein kinase RNA–like endoplasmic reticulum kinase (PERK), that has a role in regulating protein synthesis through its ability to phosphorylate the eukaryotic initiation factor 2α (eIF-2α) (Module 9: Figure regulation of eIF-2α cycling).

H2S is an important signalling molecule that can regulate multiple processes: neuronal synaptic transmission, smooth muscle relaxation, insulin secretion and inflammatory responses. Vasorelaxation was one of the first signalling functions identified for H2S that seems to mediate its action through the ATP-sensitive K+ (KATP) channels, but is now known to regulate many other smooth muscle cell processes. For example, there also are indications that H2S may act as an inositol 1,4,5-trisphosphate receptor (InsP3R) antagonist to reduce the release of Ca2+ from the ER (Module 3: Figure InsP3R regulation) and this may play a particularly important role in airway smooth muscle cells.

A decline in the levels of H2S may contribute to various bronchial diseases such as asthma and chronic obstructive pulmonary disease (COPD).

Mitogen-activated protein kinase (MAPK) signalling

Overview

The multifunctional mitogen-activated protein kinase (MAPK) signalling system consists of separate pathways that function to control a number of different cellular processes such as gene transcription, metabolism, motility, cell proliferation, apoptosis, synaptic plasticity and long-term memory. These different downstream effectors are activated by the final MAPK components associated with the three main signalling pathways:

These different pathways are assembled by combining components from an extensive mitogen-activated protein kinase (MAPK) signalling toolkit.

The mitogen-activated protein kinase (MAPK) signalling properties such as their spatio-temporal control mechanisms help to explain how they operate to regulate so many cellular processes.

The activity of the MAPK signalling pathway is reversed by the mitogen-activated protein kinase (MAPK) phosphatases.

Mitogen-activated protein kinase (MAPK) signalling toolkit

There is an extensive mitogen-activated protein kinase (MAPK) signalling toolkit, which can be divided into different functional components such as the transducers, the MAPK kinase kinases (MAPKKKs), MAPK kinases (MAPKKs), MAPKs, MAPK scaffolding proteins and MAPK target proteins (Module 2: Table MAPK signalling toolkit). Specific components from this toolkit are then assembled into the different signalling pathways (Module 2: Figure MAPK signalling).

Extracellular-signal-regulated kinase (ERK) pathway

The extracellular-signal-regulated kinase (ERK) pathway is one of the major signalling cassettes of the mitogen-activated protein kinase (MAPK) signalling pathway (Module 2: Figure MAPK signalling). It performs a number of important signalling functions, including the control of cell proliferation and the synaptic plasticity responsible for learning and memory. The main MAPK/ERK kinase kinase (MEKK) components are the Raf family members Raf-1, A-Raf and B-Raf that phosphorylate two serine residues on the MAPK/ERK kinase (MEK) components MEK1/2. The latter are dual-specificity protein kinases that phosphorylate the tyrosine and threonine residues of the characteristic MAPK components ERK1/2 that are responsible for stimulating the downstream effectors, many of which are transcription factors (Module 2: Figure ERK signalling). There is thus a linear transfer of information through a phospho-relay system based on a sequential series of phosphorylation events.

An important feature of this ERK pathway, which can be activated by both protein tyrosine kinase-linked receptors (PTKRs) and by G protein-coupled receptors (GPCRs), is its spatial organization (Module 2: Figure ERK signalling).

In the case of the PTKRs, growth factors such as platelet-derived growth factor (PDGF) usually cause receptor dimerization, which allows the cytosolic tyrosine kinase domains to come together and to phosphorylate each other (Module 1: Figure PDGFR activation). These phosphorylated residues then function as docking motifs to pull in signalling components such as Shc, growth factor receptor-bound protein 2 (Grb2) and Son-of-sevenless (SoS) that then activate the small GTP-binding protein Ras. Activated Ras then interacts with the protein kinase Raf, which initiates the phosphorylation cascade of the ERK pathway. The three components of this pathway (Raf-1, MEK1/2 and ERK1/2) are held in place at the cell surface by the scaffolding protein kinase suppressor of Ras 1 (KSR1). Up to this point, all the signal transduction processes have occurred at the cell surface, and the next information transfer step depends upon the diffusion of the activated enzyme from the cell surface to the nucleus. Once it is phosphorylated, the activated phospho-ERK1/2 leaves the plasma membrane to diffuse into the cytoplasm and then into the nucleus, where it phosphorylates and activates a number of transcription factors (Module 2: Figure ERK signalling). Some of these genes code for MAPK signalling components, such as MAPK phosphatase 1 (MKP-1), which sets up a negative-feedback loop. In addition, phospho-ERK1/2 can also act together with Ca2+ to stimulate cytoplasmic phospholipase A2 (cPLA2). The Ca2+ induces the cPLA2 to associate with cytosolic membranes, whereas the ERK1/2 phosphorylates Ser-505 and Ser-727 to stimulate the enzymatic release of arachidonic acid from phospholipid precursors.

There are a number of putative mechanisms for linking the activation of G protein-coupled receptors (GPCRs) to the ERK pathway. The release of βγ subunits may activate Src, which can then feed into the processes that activate Ras. The arrestins can also function as scaffolds to assemble components of the ERK pathway such as Raf1 and ERK1/2. Alternatively, activation of phosphoinositide hydrolysis by Gq to produce inositol 1,4,5-trisphosphate (InsP3) and diacylglycerol (DAG) can access the ERK pathway via two mechanisms. The InsP3/Ca2+ can act through proline-rich tyrosine kinase 2 (Pyk2), whereas DAG and Ca2+ act through protein kinase C (PKC).

One of the major functions of the ERK pathway is to activate a range of different transcription factors such as cyclic AMP response element-binding protein (CREB) (Module 4: Figure CREB activation) and Elk-1 (Module 4: Figure ETS activation).

This ERK pathway contributes to the control of a large number of cellular processes:

c-Jun N-terminal kinase (JNK) pathway

The c-Jun N-terminal kinase (JNK) pathway is one of the major signalling cassettes of the mitogen-activated protein kinase (MAPK) signalling pathway. It functions in the control of a number of cellular processes, including proliferation, embryonic development and apoptosis. The pathway takes its name from the c-Jun N-terminal kinases 1–3 (JNK1–JNK3), which are the MAPKs that interact with the final effectors (Module 2: Figure MAPK signalling). They contain the dual phosphorylation motif Thr-Pro-Tyr, which is phosphorylated following activation of the upstream phosphorylation cascade.

The JNK pathway is activated by a bewildering number of mechanisms. This complexity is evident by the fact that there are 13 MAPK kinase kinases (MAPKKKs) responsible for feeding information into the JNK pathway. The apoptosis signal-regulating kinase 1 (ASK1) is an example of such a kinase that initiates the signalling cascade that leads to JNK activation. One way of trying to cope with this complexity is to examine specific examples such as the activation of JNK by the interleukin-1 receptor (Module 2: Figure JNK signalling).

The JNK pathway can also be activated through G protein-coupled receptors (GPCRs) using G proteins such as G12/13 and Gγq/11. Just how G proteins feed into the cascade is unclear, but it seems that they activate the GTP-binding proteins such as Rac, Rho and Cdc42. For example, the Gγq/11, which acts through Trio that is a GEF capable of activating Rac and Rho functions in growth factor signalling pathways (Module 9: Figure growth factor signalling). Alternatively, the arrestins that associate with GPCRs during the process of receptor desensitization may function as a scaffold to bring together components of the JNK pathway such as MKP7 and JNK3.

The MAPK signalling system operates a negative-feedback loop in that some of the genes activated by the JNK pathway code for signalling components such as the scaffold protein JNK-interacting protein 1 (JIP1), which will bind JNK and thus limits its action.

This JNK pathway contributes to the control of a large number of cellular processes:

Apoptosis signal-regulating kinase 1 (ASK1)

Apoptosis signal-regulating kinase 1 (ASK1) is part of the mitogen-activated protein kinase (MAPK) signalling toolkit (Module 2: Table MAPK signalling toolkit) where it functions as one of the MAPK kinase kinases (MAPKKKs) that operates within the c-Jun N-terminal kinase (JNK) pathway (Module 2: Figure MAPK signalling). One of the stimuli capable of activating ASK1 is tumour necrosis factor α (TNFα) (see step 2 in Module 2: Figure NF-κB activation).

The activity of ASK1 can also be modulated by at least two other signalling mechanisms. First, the reduced form of the redox buffer thioredoxin (Trx) binds to ASK1 and inhibits its activity. Following ROS formation, the Trx is oxidized and dissociates thus removing its inhibitory effect and allowing ADK1 to stimulate apoptosis. Secondly, ASK1 activity is inhibited by binding to calcium and integrin-binding protein 1 (CIB1). An increase in the level of Ca2+ removes this inhibition thus allowing ASK1 to induce apoptosis.

Sprouty (SPRY)

There are two families of sprouty proteins that function as inhibitors of the MAP kinase signalling pathway. The founding sprouty family has four members (SPRY1–4) while the sprouty-related, EVH1 domain-containing protein (SPRED) family contains three members (SPRED1–3) together with EVE-3, which is a splice variant of SPRED3. Members of this family have a cysteine-rich domain (CRD).

The SPRY proteins seem to function during development to control the formation of a number of tissues and organs. The level of SPRY1 is regulated by miR-21 to increase MAP kinase signalling during cardiac development.

Like the SPRY family, the SPREDs act by binding to either Ras or Raf1 to inhibit the MAP kinase signalling pathway. SPRED1 can act either as a homodimer or as a heterodimer with SPRED2.

Mutations in the gene that encodes SPRED1, which will result in hyperactivation of the Ras/MAP kinase signalling pathway, causes neurofibromatosis type 1-like syndrome (NFLS).

p38 pathway

The p38 pathway is one of the major signalling cassettes of the mitogen-activated protein kinase (MAPK) signalling pathway. It functions in the control of apoptosis and the release of cytokines by macrophages and neutrophils. The pathway takes its name from the family of p38 kinases, which are the MAPKs that interact with the final effectors (Module 2: Figure MAPK signalling).

The p38 pathway can be activated either by different receptor mechanisms or by various environmental stresses such as osmotic, redox or radiation stress. For example, one of the targets of the p38 pathway that is activated by UV irradiation is one of the Cdc25 enzymes that control cell cycle progression. Phosphorylation of Ser-323 on Cdc25B results in the binding of 14-3-3 protein, which then prevents this enzyme from initiating entry into mitosis.

The Toll receptor signalling pathway provides an example of how the p38 pathway is activated by a Toll receptor for lipopolysaccharide (LPS) (Module 2: Figure Toll receptor signalling).

Mitogen-activated protein kinase (MAPK) signalling properties

The different mitogen-activated protein kinase (MAPK) signalling pathways have a number of important properties that greatly enhance their signalling efficiency.

Temporal aspects of mitogen-activated protein kinase (MAPK) signalling

The outcome of these pathways is very dependent on temporal aspects, particularly signal duration. For example, prolonged stimulation is necessary to induce cell proliferation.

Fidelity

Another important property is the fidelity of these signalling pathways. The three major signalling pathways share many common features: they often share signalling components, they are often activated by similar inputs [especially in the case of the c-Jun N-terminal kinase (JNK) and p38 pathways], and they can also regulate similar cellular processes. A question therefore arises as to how the fidelity of these signalling pathways is achieved in order to reduce cross-talk and to ensure that they carry out their particular functions. It seems that much of this fidelity is achieved by using molecular scaffolds to hold together all the components of each signalling pathway in a multimolecular complex. In this way, information can be passed from one component to the next without interference from other signalling pathways (Module 6: Figure signalling hierarchies). Examples of such scaffolds are the scaffolding proteins kinase suppressor of Ras 1 (KSR1) (Module 2: Figure ERK signalling) and JNK-interacting protein 1 (JIP1) (Module 2: Figure JNK signalling).

Other determinants of fidelity are the docking sites that enable the different mitogen-activated protein kinases (MAPKs) to bind to their specific downstream effectors.

Phenotypic remodelling of the mitogen-activated protein kinase (MAPK) signalling pathway

The mitogen-activated protein kinase (MAPK) signalling pathway operates autoregulatory loops in that the different signalling pathways can regulate the expression of their own signalling components. For example, the extracellular-signal-regulated kinase (ERK) pathway can regulate the expression of MAPK phosphatase 1 (MKP1) (Module 2: Figure ERK signalling), whereas the c-Jun N-terminal kinase (JNK) pathway induces the expression of JNK-interacting protein 1 (JIP1) (Module 2: Figure JNK signalling). This induction of MKP1 and JIP1 effectively set up negative-feedback loops that limit the activity of the MAPK signalling pathway, and is an example of signalsome stability.

Phenotypic remodelling of the MAPK signalsome may result in the abnormal cell proliferation observed in polycystic kidney disease (Module 12: Figure polycystins and polycystic kidney disease).

Nuclear factor κB (NF-κB) signalling pathway

The transcription factor nuclear factor κB (NF-κB) is activated by a large number of external stimuli such as the tumour necrosis factors (TNFs), interleukin-1 (IL-1) and the pathogen-associated molecular patterns (PAMPs), which are responsible for controlling processes such as inflammation, cell proliferation and apoptosis. NF-κB belongs to the group of transcription factors that lie latent in the cytoplasm and then translocate into the nucleus upon activation (mechanism 2 in Module 4: Figure transcription factor activation). This diversity of downstream effector processes indicates that there must be separate signalling pathways, and this is evident from the nuclear factor κB (NF-κB) signalling toolkit, which contains multiple isoforms both of the transcription factors (the NF-κB/Rel family) and of the activation components. This complexity is carried through to the nuclear factor κB (NF-κB) signalling pathway, where there are many variations on the basic theme of NF-κB activation (i.e. Mechanism 2 in Module 4: Figure transcription factor activation).

There are two remarkable aspects of the NF-κB signalling pathway. Firstly, it can control a very large number of genes that are often activated as large cohorts in specific cells by different stimuli. Secondly, it is used by a number of different signalling systems with subtle variations in the mechanism and the components that are used to convey information into the nucleus. The tumour necrosis factor α (TNFα) signalling pathway and the Toll receptor signalling pathway will be described to illustrate the main features of the NF-κB signalling pathway. The receptor activator of nuclear factor κB (NF-κB) ligand (RANKL), which is a transmembrane protein that belongs to the TNF family of cytokines, activates the RANKL receptor (RANK) that uses the NF-κB signalling pathway to control osteoclastogenesis (Module 8: Figure osteoclast differentiation).

Nuclear factor κB (NF-κB) signalling toolkit

The NF-κB signalling toolkit is composed of four main classes of signalling components (Module 2: Table NF-κB signalling toolkit).

Nuclear factor κB (NF-κB)/Rel family

The nuclear factor κB (NF-κB)/Rel family consists of five members (Module 2: Table NF-κB signalling toolkit). There is some confusion concerning the nomenclature of this family. The notations shown at the beginning of the table (p50, p52, p65, RelB and c-Rel) will be used here. All members of the family share an N-terminal Rel homology domain (RHD), which binds to DNA (Module 2: Figure NF-κB, IκB and IKK structure). The RHD is also used to associate NF-κB with the inhibitor of nuclear factor κB (NF-κB) (IκB) proteins. The NF-κB/Rel family of transcription factors normally function as heterodimers, and the p50/p65 complex was the first to be discovered. Some of the family members can form homodimers, such as p50/p50 and p52/p52, and these act as repressors of NF-κB-sensitive genes.

Inhibitor of nuclear factor κB (NF-κB) (IκB)

The inhibitor of nuclear factor κB (NF-κB) (IκB) family are characterized by an ankyrin repeat domain (Module 2: Figure NF-κB, IκB and IKK structure), which functions in its interaction with NF-κB to form the inactive complex that resides in the cytoplasm. The IκB inhibits transcription by masking the nuclear localization signal (NLS) on NF-κB, which thus prevent it from entering the nucleus. The NF-κB/IκB complex remains inactive within the cytoplasm until the IκB is removed following activation of the nuclear factor κB (NF-κB) signalling pathway. Much of the specificity within the NF-κB signalling pathway depends on these IκB isoforms being able to bind to different dimers of the NF-κB/Rel family.

Inhibitor of nuclear factor κB (IκB) kinases (IKKs)

The inhibitor of nuclear factor κB (IκB) kinases (IKKs) function as heterodimers and are responsible for phosphorylating IκB to mark it for subsequent degradation by the proteasome. The kinase domain is located in the N-terminus, whereas the C-terminus has leucine zipper and helix–loop–helix domains (Module 2: Figure NF-κB, IκB and IKK structure). These two protein-association domains enable the enzyme to associate with a large multisubunit complex in the cytoplasm. Another important component of this complex is nuclear factor κB (NF-κB) essential modulator (NEMO) (IKKγ), which is a regulatory/structural subunit responsible for the interaction with upstream kinases on the receptor as part of the tumour necrosis factor α (TNFα) signalling pathway (Module 2: Figure NF-κB activation). In the case of NF-κB activation in T-cells, the IKK is stimulated by protein kinase θ (PKCθ) using the scaffolding proteins CARMA1, Bcl10 and MALT1 (Module 9: Figure TCR signalling).

The tumour suppressor CYLD binds to NEMO and inhibits the subsequent phosphorylation of IκB as part of the TNFα signalling pathway.


Module 2: Table NF-κB signalling toolkit The nuclear factor κB (NF-κB) signalling toolkit

Nuclear factor κB (NF-κB) transcription system Comments
NF-κB/Rel family of transcription factors
 p50 (NF-κB1) p50 has a p105 precursor protein
 p52 (NF-κB2) p52 has a p100 precursor protein
 p65 (NF-κB3, also known as RelA)
 RelB
 c-Rel
Inhibitor of NF-κB (IκB) family
 IκBα Bind to p65 and c-Rel
 IκBβ
 IκBγ
 IκBε
 IκBζ Binds preferentially to p50
 Bcl3 B cell lymphoma 3 binds to p50/p50 and p53/p53 homodimers.
Inhibitor of NF-κB (IκB) kinases (IKKs)
 IKKα
 IKKβ
 NEMO (IKKγ) A regulatory subunit responsible for the interaction with upstream kinases on the receptor complex (Module 2: Figure NF-κB activation)
Associated proteins
 Cdc37
 Hsp90



Tumour necrosis factor α (TNFα) signalling pathway

The basic operation of the nuclear factor κB (NF-κB)/Rel family of transcription factors is that they are activated in the cytoplasm and then translocate into the nucleus to activate transcription (mechanism 2 in Module 4: Figure transcription factor activation). There are a number of variations in the way that this translocation process is initiated, depending on the nature of the incoming signals and the receptors that are being activated. In addition to the activation of NF-κB, these non-enzyme-containing receptors can activate other signalling pathways (Module 1: Figure cytokines). In this section, attention will focus on the NF-κB signalling pathway, which is used by the tumour necrosis factor α (TNFα). It is considered to be one of the ‘classical’ mechanisms as illustrated by the following sequence of events (Module 2: Figure NF-κB activation):

  1. Occupation of the TNF receptor (TNF-R) by TNF induces receptor oligomerization to form a complex that attracts the adaptors TRADD and the TNFR-associated factor 2 (TRAF2), which is an adaptor that belongs to the TNF-receptor-associated factor (TRAF) family.
  2. TRAF2 is a RING domain E3 ubiquitin ligase that associates with the heterotrimeric ubiquitin-conjugating (E2) complex that contains Ubc13 and Uev1A. This is a K63 ubiquitinating complex that adds ubiquitin chains to TRAF2 and this then helps to recruit the receptor-interacting protein 1 (RIP1), which is also ubiquitinated. These ubiquitin chains provide an important scaffolding role in the assembly of additional elements of the signal transduction pathway. For example TRAF2 interacts with apoptosis signal-regulating kinase 1 (ASK1) that then relays information to both the JNK signalling and p38 signalling pathways
  3. The developing receptor complex attracts additional components such as the transforming growth factor β activated kinase-1 (TAK1), the TAK1-binding (TAB) proteins 1 to 3 (TAB1-3) and the multisubunit cytoplasmic complex containing the inhibitor of NF-κB (IκB) kinase (IKK) α/IKKβ dimer and the regulatory NF-κB essential modifier (NEMO) subunit, which is a central player in this translocation sequence.
  4. Once this complex is complete, the TAK1 phosphorylates and activates IKKβ.
  5. The IKKβ then phosphorylates IκBα on two sites (Ser-32 and Ser-36).
  6. The phosphorylated IκBα is then susceptible to ubiquitination by the Skp1/cullin/F-box (SCF) ubiquitin ligase.
  7. The polyubiquitinated IκBα is then sent to the proteasome, where it is degraded, resulting in the liberation of the NF-κB heterodimer p50/p65.
  8. The NF-κB is imported into the nucleus, where it binds to the κB promoter elements to activate expression of many different genes (Module 4: Figure NF-κB activation and function). One of these genes is hypoxia-inducible factor 1α (HIF-1α), which thus serves to link the innate immune response to the hypoxic response.
  9. Transcription ceases when NF-κB is exported from the nucleus. One of the genes activated by NF-κB is IκBα, which thus sets up a negative-feedback loop, because it binds to the NF-κB that is exported from the nucleus to reconstitute the inactive cytoplasmic complex.
  10. The genotoxic stress activation of NF-κB signalling represents an alternative signalling pathway that initiates in the nucleus. DNA damage activates the phosphorylation of ataxia-telangiectasia mutated (ATM) (Module 9: Figure G1 checkpoint signalling) that then binds to NEMO and exports it into the cytoplasm where it is polyubiquitinated by the linear ubiquitin chain assembly complex (LUBAC) E3 ligase, which consists of haem-oxidized IRP2 ubiquitin ligase-1 (HOIL-1L), HOIL-1P interacting protein (HOIP), and shank-associated RH domain interacting protein (SHARPIN). The activity of LUBAC can be enhanced by the ubiquitin E3 ligase Parkin. The polyubiquitinated NEMO then combines with the IKKα/IKKβ dimer to initiate the activation of NF-κB as described above.
  11. The ubiquitin signalling pathway contributes to the recovery of this signalling cascade after TNFα is withdrawn by removing the ubiquitin scaffolds that hold together the transducing complex. The deubiquitinating enzymes A20 and CYLD are particularly active in removing the ubiquitin chains. The reversible ubiquitination of signal transducing components is thus an essential part of the processing of information by this TNF signalling pathway.

Tumour necrosis factor (TNF)-receptor-associated factor (TRAF) family

The tumour necrosis factor (TNF)-receptor-associated factor (TRAF) family has seven members, which function in the signalling pathways of various TNF receptor superfamily and Toll/interleukin-1 receptor mechanisms. TRAF2 functions as an adaptor for the tumour necrosis factor receptor (Module 2: Figure NF-κB activation). In the case of osteoclasts, it is TRAF6 that couples the receptor activator of nuclear factor κB (NF-κB) ligand (RANKL) receptor (RANK) to downstream signalling pathways (Module 8: Figure osteoclast differentiation).

Toll receptor signalling pathway

The Toll receptor signalling pathway plays a central role in inflammatory responses (Module 11: Figure inflammation). This pathway is activated by different types of stimuli: cytokines, represented by interleukin-1 (IL-1), stimuli derived from pathogens that are known as pathogen-associated molecular patterns (PAMPs) (Module 11: Figure formation and action of PAMPs) and endogenous damage-associated molecular patterns (DAMPS). The DAMPS and PAMPs act on Toll-like receptors (TLRs) and the IL-1 receptor (IL-1R) and these form a TLR/IL-1R family that act through a similar Toll receptor signalling pathway (Module 1: Figure cytokines). The primary function of this signalling pathway is to stimulate the transcriptional processes that result in the expression of a wide range of inflammatory cytokines and immunoregulators (Module 2: Figure Toll receptor signalling). As such, the Toll receptor signalling pathway has to transmit information from the TLRs and the IL-1R on the cell surface to the transcriptional factors such as nuclear factor κB (NF-κB) and activating protein 1 (AP-1) (c-Jun/Fos) that act within the nucleus. This information transfer system will be illustrated by reference to the response of the receptor TLR4 to lipopolysaccharide (LPS) (Module 11: Figure formation and action of PAMPs). As for many other signalling pathways, information is transferred through both protein–protein interactions and protein phosphorylation reactions. The ubiquitin signalling system also has an important role in orchestrating this Toll receptor signalling pathway as shown in the following sequence (Module 2: Figure Toll receptor signalling):

  1. The TLR4 is a transmembrane protein that has leucine-rich repeats in its ectodomain, while the cytoplasmic domain has a Toll/interleukin 1 (IL-1) receptor (TIR) domain. The lipopolysaccharide (LPS) that initiates this signalling pathway binds first to an extracellular LPS-binding protein (LBP) and to CD14, which is a glycosylphosphatidylinositol-anchored membrane glycoprotein, and this complex carries the LPS to the Toll-like receptor 4 (TLR4).
  2. The TIR domain on the TLR4 receptor forms homophilic interaction with the TIR domain of the adaptor proteins TIR adaptor protein (TIRAP) and MyD88. The latter is specific for certain TLRs, such as TLR4, but not others. The other end of these adaptors have a death domain that draws in the IL-1 receptor-associated kinases 1 and 4 (IRAK-1 and IRAK-4), which undergoes an autophosporylation reaction that enables them to bind to another adaptor called TRAF6, which belongs to the tumour necrosis factor (TNF)-receptor-associated factor (TRAF) family that has a critical role to play in the next series of reactions.
  3. The IRAK-1 and TRAF6 dissociate from the receptor and move in to the cytoplasm.
  4. The TRAF 6 is a RING domain E3 ubiquitin ligase that associates with the heterotrimeric ubiquitin-conjugating (E2) complex that contains Ubc13 and Uev1A. This is a K63 ubiquitinating complex that results in the autoubiquitination of TRAF6. The ubiquitinated TRAF6 then binds the transforming growth factor β activated kinase-1 (TAK1) and the TAK1-binding (TAB) proteins 1 to 3 (TAB1-3). The multisubunit cytoplasmic complex containing the inhibitor of NF-κB (IκB) kinase (IKK) α/IKKβ dimer and the regulatory NF-κB essential modifier (NEMO) subunit are also drawn into the complex. The resulting activation of TAK1 is then responsible for relaying information out to other components of the signalling pathways.
  5. The TAK1 activates both the c-Jun N-terminal kinase (JNK) pathway and the p38 pathway (Module 2: Figure MAPK signalling). The JNK and p38 function to phosphorylate transcription factors such as AP-1, which binds to the cyclic AMP response element (CRE) site.
  6. The other action of TAK1 is to phosphorylate the IKKβ component of the inhibitor of NF-κB (IκB) kinase (IKK) α/IKKβ dimer.
  7. The phosphorylated IKKβ then acts to phosphorylate IκBα to liberate the NF-κB.
  8. The phosphorylated IκBα is then susceptible to ubiquitination by the Skp1/cullin/F-box (SCF) ubiquitin ligase. The polyubiquitinated IκBα is then sent to the proteasome, where it is degraded,
  9. The NF-κB enters the nucleus, binds to the κB site and activates the genes that code for the inflammatory cytokines and immunoregulators that contribute to inflammatory responses (Module 11: Figure inflammation).
  10. An important aspect of the recovery of this signalling cascade after LPS is withdrawn, is the removal of the ubiquitin scaffolds that hold together the transducing complex. The deubiquitinating enzymes A20 and CYLD are particularly active in removing the ubiquitin chains. This ubiquitin signalling system thus plays an essential role in the processing of information by this Toll receptor signalling pathway.
  11. In addition to functioning as a coreceptor to transfer LPS to TLR4, CD14 may also be capable of activating a Ca2+ signalling pathway. The CD14 activates Src that is coupled to phospholipase Cγ1 (PLCγ1), which is then capable of triggering an influx of external Ca2+ through an unknown mechanism. The increase in Ca2+ stimulates calcineurin (CaN) to dephosphorylate the transcription factor NFAT that enters the nucleus to activate expression of genes such as interleukin-2 and various pro-apoptotic genes (nur77, Gadd45g, Ddit3 and Tia1) (Module 2: Figure Toll receptor signalling). This Ca2+ pathway has been described in dendritic cells, but appears to be absent in macrophages.

This signalling pathway is particularly evident on resident macrophages (Module 11: Figure macrophage signalling) and mast cells (Module 11: Figure mast cell signalling) that respond to invading pathogens. Mutations in MyD88 have been linked to diffuse large B cell lymphoma (DLBCL).

Virus recognition and antiviral responses

The recognition and initiation of antiviral responses, which is a part of the innate immune system, depend upon the activation of a number of signalling pathways such as the Toll receptor signalling pathway. When viruses enter cells they are broken down into fragments such as double-stranded RNA (dsRNA), single-stranded RNA (ssRNA), 5' triphosphate ssRNA or CpG DNA, which are located either in the cytoplasm or within the endosomal compartment (Module 2: Figure virus recognition). These fragments are examples of the pathogen-associated molecular patterns (PAMPs) that are responsible for triggering a variety of inflammatory responses (Module 11: Figure formation and action of PAMPs). The following sequence of reactions describes how three main groups of receptors detect these viral PAMPs to initiate a number of antiviral signalling pathways such as Toll-like receptor (TLR) signalling, Retinoic acid-inducible gene I (RIG-I)-like receptor (RLRs) signalling and Nucleotide oligomerization domain (NOD) protein-like receptor (NLR) signalling (Module 2: Figure virus recognition):

Toll-like receptor (TLR) signalling

  1. Those viral PAMPs, which are directed into the endosome, interact with Toll-like receptors (TLR3, TLR7/8 and TLR9) that are located in the endosomal membrane (see green box on the left of Module 2: Figure virus recognition).
  2. The TLR9, which responds to CpG DNA, and the TLR7/8 that bind ssRNA both recruit the adaptor protein MyD88. The latter then acts through TRAF6, which is one of the tumour necrosis factor (TNF)-receptor-associated factor (TRAF) family, to stimulate the IKKαβ. The latter then activates the transcription factor NF-κB through the Toll receptor signalling pathway described in more detail earlier (Module 2: Figure Toll receptor signalling).
  3. The NF-κB is imported into the nucleus where it functions to induce the transcription of both the inflammatory cytokines and the genes for type I interferon (IFN).
  4. When activated by dsRNA, the TLR3 receptor interacts with the TIR-domain-containing adaptor protein inducing IFN-β (TRIF). The latter then activates the two IKK-related proteins TBK1 [TRAF-family member-associated NF-κB activator (TANK)-binding protein] and IKKε, which is also known as inducible IKK (iIKK).
  5. The activated TBK1/IKKε complex then phosphorylates the interferon-regulatory factors (IFRs) IRF3 and IRF7. When phosphorylated, these two IRFs form homodimers that are then imported into the nucleus where they contribute to the transcription of the type I interferons (IFN-α and IFN-β).

Retinoic acid-inducible gene I (RIG-I)-like receptor (RLRs) signalling

  1. Those viral PAMPs that are directed into the cytoplasm interact with cytoplasmic receptors such as the RIG-I-like receptors (RLRs) (see yellow box in the middle of Module 2: Figure virus recognition).
  2. The two main RLRs are retinoic acid-inducible gene I (RIG-I) itself and melanoma-associated gene 5 (MDA5). These two receptors have a characteristic DExD/H box helicase domain that is responsible for binding double-stranded RNA (dsRNA). RIG-I can also respond to 5’ triphosphate ssRNA. These two receptors also have a pair of N-terminal caspase recruitment domains (CARD), which are the functional transducing components that signal to downstream elements. The LGP2 (laboratory of genetics and physiology-2), which also has a helicase domain, can also bind dsRNA, but it lacks the CARD domains and thus fails to transduce cellular signals but may act as a negative regulator of MDA5 and RIG-I.
  3. The CARD domains on MDA5 and RIG-I interact with the IFN-β promoter stimulator-1 (IPS-1), which also contains an N-terminal CARD domain and is attached to the outer mitochondrial membrane. IPS1 is also known as MAVS (mitochondrial antiviral signalling), VISA (virus-inducing signalling adaptor) or CARDIF (CARD adaptor inducing IFN-β). The IPS1 protein then stimulates the TBK1/IKKε complex that then phosphorylates the interferon-regulatory factors (IFRs) IRF3 and IRF7 as described above (see step 5).

Nucleotide oligomerization domain (NOD) protein-like receptor (NLR) signalling

  1. The double-stranded RNA (dsRNA) in the cytoplasm can also interact with the nucleotide oligomerization domain (NOD) protein-like receptors (NLRs) (see pink box on the right of Module 2: Figure viral recognition). One of the NLRs is the Nacht domain-, Leucine-rich repeat-, and PYD-containing protein 3 (NALP3), which is also known as cryopyrin that responds to dsRNA to activate caspase-1. The interaction between NALP3 and caspase-1 is facilitated by an adaptor called apoptosis-associated speck-like protein containing a CARD (ASC). The interaction between these three proteins takes place in a macromolecular complex known as the inflammasome.
  2. The activated caspase-1 contributes to the development of an inflammatory response by cleaving pro-IL-1β to form the inflammatory cytokine interleukin 1β (IL-1β).

Inflammasome

The inflammasome is a key component of the innate immune system with important antiviral and antibacterial functions. A key component of the inflammosome is Nacht domain-, Leucine-rich repeat-, and PYD-containing protein 3 (NALP3), which is a pattern recognition receptor. When NALP3 is activated it forms an inflammatory complex containing an adaptor called apoptosis-associated speck-like protein containing a CARD (ASC) and the pro-caspase-1 that functions to activate caspase-1 (Module 2: Figure viral recognition).

The expression of NALP3 is regulated by nuclear factor κB (NF-κB) activated through the Toll receptor signalling pathway. The activity of NLRP3 is also regulated through ubiquitination in that its stability is maintained by BRCC, which is one of the JAMM-domain DUBs.

Mutations in NALP3 have been linked to various inflammatory diseases such as gout, septic shock, diabetes and Alzheimer's disease (Module 12: Figure Inflammation and Alzheimer's disease).

Phospholipase D (PLD) signalling pathway

The phospholipase D (PLD) signalling pathway functions by generating phosphatidic acid (PA), which acts to regulate a wide range of cellular processes. Phospholipase D (PLD) activation depends upon a number of mechanisms, and these may vary depending on where the signalling mechanism is located within the cell. The primary messenger produced by this signalling pathway is PA, and phosphatidic acid (PA) action is carried out through a number of downstream effectors. Phosphatidic acid (PA) metabolism occurs through different pathways that generate further signalling molecules such as diacylglycerol (DAG) and lysophosphatidic acid (LPA).

Phospholipase D (PLD) activation

Mammals have two phospholipase D (PLD) genes (PLD1 and PLD2), both of which have splice variants. There is a non-classical phosphoplipase D3 (PLD3) that has come to prominence because it has been identified as a risk factor for Alzheimer's disease (AD). Most attention has focused on PLD1 as a signal transducer because it has a low basal activity that increases markedly in response to external stimuli. On the other hand, PLD2 has a high basal activity and its role in signalling is uncertain. While most of the PLD2 is located on the plasma membrane, PLD1 is found predominantly on intracellular membranes (e.g. Golgi, endoplasmic reticulum and endosomes), but has also been located at the plasma membrane, particularly at caveolae. The structure of the enzyme contains motifs responsible for its membrane location and catalytic activity (Module 2: Figure PLD isoforms). The activity of PLD1 increases following stimulation of both protein tyrosine kinase-linked receptors (PTKRs) and G protein-coupled receptors (GPCRs) (Module 2: Figure PLD signalling). One of the problems with trying to understand the signalling function of this pathway is to determine just how these different receptors act to stimulate PLD1. The fact that the activation of PLD1 seems to depend on the prior activation of other signalling pathways such as the diacylglycerol (DAG)/protein kinase C (PKC) signalling cassette or the PtdIns 3-kinase signalling cassette (Module 2: Figure PLD signalling) suggests that the PLD signalling pathway is not an autonomous signalling system, but should be considered more as a downstream effector of these other signalling pathways.

A specific N-arachidonylphosphatidylethanolamine (NAPE) phospholipase D (PLD) (NAPE PLD) functions to generate anandamide, which is one of the endocannabinoids (Module 1: Figure anandamide).

The Arf signalling pathway plays an important role in the activation of PLD (Module 2: Figure Arf signalling).

Phosphatidic acid (PA) action

The primary messenger of the phospholipase D (PLD) signalling pathway is the lipid phosphatidic acid (PA), which has a number of actions within the cell (Module 2: Figure PLD signalling). The signalling function of PA is mainly directed towards the regulation of various enzymes such as stimulation of the target of rapamycin (TOR) and sphingosine kinase or inhibition of protein phosphatase 1. It also plays an important role in regulating phagocytosis (Module 4: Figure phagosome maturation).

In addition to such signalling functions, a local accumulation of PA may also alter the physical properties of the membrane by creating curvatures to facilitate the formation of vesicles for intracellular trafficking as occurs during the COPII-mediated transport from the ER to the Golgi (Module 4: Figure COPII-coated vesicles) and the COPI-mediated transport from the Golgi to the ER (Module 4: Figure COPI-coated vesicles).

Phosphatidic acid (PA) metabolism

Two separate enzymes carry out the metabolism of phosphatidic acid (PA). It can be dephosphorylated to diacylglycerol (DAG) by a PA phosphohydrolase or it can be partially deacylated by a phospholipase A2 (PLA2) to form lysophosphatidic acid (LPA) (Module 2: Figure PLD signalling). The LPA released from the cell is a potent agonist on receptors of the endothelial differentiation gene (EDG) family (Module 2: Figure sphingomyelin signalling).

Sphingomyelin signalling pathway

The sphingomyelin signalling pathway that has been implicated in the control of a whole host of cellular processes through the generation and function of ceramide and sphingosine 1-phosphate (S1P), which are the main messengers operating in this signalling pathway. The action of sphingosine 1-phosphate (S1P) is complicated because, in addition to acting internally, S1P is released from cells to function as an external ligand acting on cell-surface receptors. This sphingomyelinase signalling pathway also produces ceramide, another messenger that plays a significant part in processes such as cell proliferation, apoptosis and the response of the cell to stress and injury. One of the difficulties in understanding this pathway is its pleiotropic effect on cells with responses that are often diametrically opposed, such as proliferation and apoptosis. A clue to this complexity lies in the fact that the pathway can spawn a number of messengers, and the action of these is very dependent on the current state of the cell, especially with regard to what other signalling pathways are active.

Generation and function of ceramide and sphingosine 1-phosphate (S1P)

The generation of ceramide and sphingosine 1-phosphate (S1P), which are the two main bioactive ‘messenger’ lipids of the shingomyelin signalling pathway, begins with the hydrolysis of sphingomyelin (SM) by various sphingomyelinases (SMases) to form ceramide (Module 2: Figure sphingolipid metabolism). A ceramidase (CDase) then cleaves off one of the fatty acid chains to form sphingosine (Sph), which is then phosphorylated by sphingosine kinase (SPHK). As illustrated by the figures in Module 2: Figure sphingolipid metabolism, the relative cellular levels of these sphingolipids vary enormously. The precursor SM is present at the highest level and this then declines during the different metabolic transformations.

An important aspect of this sphingomyelin signalling pathway is the cellular location and function of these different metabolic steps (Module 2: Figure sphingomyelin signalling):

  1. Sphingomyelin synthesis is carried out first in the endoplasmic reticulum (ER), where ceramide is formed, and is completed by the Golgi. The first step is carried out by serine palmitoyl transferase (SPT) that combines serine and palmitate to form dihydrosphingosine (dhSph), which is then converted into dihydroceramide (dhCer) by dihydroceramide synthase (CerS). Finally, a desaturase converts dhCer into ceramide (Cer). Mutations in serine palmitoyl transferase (SPT) are responsible for type I hereditary sensory and autonomic neuropathies (HSAN).
  2. The ceramide formed at the ER is then transported to the Golgi by a ceramide transfer protein (CERT). A sphingomyelin synthase (SMS) located in the Golgi converts ceramide into sphingomyelin (SM), which is transferred to the membrane through a vesicle transport mechanism.
  3. The hydrolysis of SM in the plasma membrane is activated by receptors sensitive to stimuli such as tumour necrosis factor α (TNFα), interleukin 1, CD28 and Fas. The coupling between receptors and the neutral sphingomyelinases (SMases) has been worked out in some detail for the TNFα receptor. A factor associated with neutral SMase (FAN) functions to couple the enzyme to the neutral SMase activation domain (NSD) of the TNFα receptor. These components of the sphingomyelinase signalling pathway are often highly concentrated in lipid rafts and caveolae, which represent the site where some of the sphingomyelinases function to hydrolyse sphingomyelin.
  4. The activation of neutral SMase at the plasma membrane provides one of the major sources of ceramide (Cer), which can then be metabolized to other signalling sphingolipids (Module 2: Figure sphingolipid metabolism). A ceramidase (CDase) hydrolyses Cer to sphingosine (Sph), which can then be phosphorylated by sphingosine kinase (SPHK) to form sphingosine 1-phosphate (S1P). Since the latter is soluble, it leaves the membrane and is free to diffuse both into and out of the cell (see later). The SPHK is activated by a number of signalling pathways using messengers such as Ca2+, DAG, cyclic AMP and ERK1/2 that are produced by signalling pathways activated by growth factors and hormones. Ceramide can also be phosphorylated by a ceramide kinase (CERK) to form ceramide 1-phosphate (C-1-P), which may act to regulate cell survival and proliferation.
  5. A variety of stress stimuli, such as ionizing radiation, UV irradiation and reactive oxygen species (ROS) act by stimulating the acidic sphingomyelinases (SMases) localized in lysosomes to hydrolyse sphingomyelin (SM) to form ceramide (Cer) (Module 2: Figure sphingomyelin signalling).
  6. S1P is metabolized via two separate pathways: it can be converted back into sphingosine by a sphingosine 1-phosphate phosphatase (SGPP) or it can be cleaved by an S1P lyase to give palmitaldehyde and phosphoethanolamine. The lyase has a 20-amino-acid transmembrane domain that positions the enzymes in cellular membranes.

The sphingomyelinase signalling pathway is thus sensitive to external signals at a number of discrete steps. The production of ceramide is accelerated by receptors sensitive to cytokines such as TNFα or to various stress signals such as ionizing radiation and reactive oxygen species (ROS). The formation of S1P can also be enhanced by growth factor and hormone receptors that elevate second messengers such as Ca2+, DAG, ERK1/2 or cyclic AMP that act by stimulating SPHK. In the absence of these signals, the main messenger will be ceramide, but this will switch over to S1P if the SPHK is activated. This interrelationship between ceramide and S1P has led to a ceramide/S1P rheostat model, where a balance between these two messengers is thought to determine cell fate. In general, ceramide seems to tip the balance in favour of cell cycle arrest, senescence and apoptosis, whereas S1P promotes survival and proliferation (Module 2: Figure sphingomyelin signalling). The different outcomes controlled by these two messengers may depend on their ability to activate separate signalling cascades.

  1. The ceramide formed at the plasma membrane or in the lysosomes acts on a number of targets such as ceramide-activated protein kinase (CAPK), protein kinase Cζ (PKCζ), cathepsin D and ceramide-activated protein phosphatases (CAPP), such as PP1 and PP2A. Most of these ceramide-sensitive targets act to promote cell cycle arrest and apoptosis either directly or indirectly. For example, the cathepsin D converts Bid into tBid to promote apoptosis (see Step 3 in Module 11: Figure apoptosis). On the other hand, the CAPPs such as PP2A can dephosphorylate various components that are used by the EDG receptors to drive cell survival and proliferation. For example the dephosphorylation of protein kinase B (PKB) not only prevents it from inhibiting apoptosis but it will also reduce its ability to enhance proliferation.
  2. One of the actions of the soluble sphingosine 1-phosphate (S1P) is to diffuse out of the cell passing through the ABCC1 transporter, which is a family member of the ATP-binding cassette (ABC) transporters (Module 3: Table ABC transporters). The extracellular S1P then activates EDG receptors, which belong to the family of G protein-coupled receptors (GPCRs) (Module 1: Table G protein-coupled receptors) that are capable of relaying information down a number of signalling pathways.
  3. These signalling pathways control cellular processes, such as cell survival, proliferation and anti-inflammatory responses, which are opposite to those controlled by ceramide.

This sphingomyelin signalling pathway has been implicated in a number of cellular processes:

Sphingosine 1-phosphate (S1P)

The key enzyme in the formation of sphingosine 1-phosphate (S1P) is sphingosine kinase (SPHK). The messenger function of S1P is complicated by the fact that it may have both intra- and extra-cellular actions.

S1P can exert its extracellular action either in an autocrine mode (activate the same cell from which it is released) (see Step 8 in Module 2: Figure sphingomyelin signalling) or a paracrine mode (diffuse away to activate neighbouring cells). This extracellular action is mediated by endothelial differentiation gene (EDG) receptors, so-called because they were first described in human umbilical vein endothelial cells induced to differentiate by phorbol esters. There is a family of these EDG receptors that are all G protein-coupled receptors (GPCRs) capable of activating most of the conventional signalling pathways (Module 1: Table G protein-coupled receptors). The EDG 1, 3, 5, 6 and 8 receptors mediate the action of S1P, whereas the EDG 2, 4 and 7 receptors appear to respond to lysophosphatidic acid (LPA), a related lysolipid. The EDG receptors that function to induce a mobilization of internal Ca2+ act by stimulating the formation of inositol 1,4,5-trisphosphate (InsP3). One function of the EDG receptors is to regulate cell motility and directional migration. An example of the latter is preosteoclast chemotaxis where S1P directs preosteoclasts form the bone marrow back into the blood (Module 8: Figure preosteoclast chemotaxis).

The intracellular action of S1P is not as well defined, mainly because it is often difficult to separate an intracellular action from that induced by its extracellular action. An intracellular action of S1P has been proposed for phagosome maturation, where it releases the Ca2+ necessary to activate the formation of PtdIns3P on endosomes (Module 4: Figure phagosome maturation). One of the problems with trying to establish such an intracellular action for S1P is that it could also mobilize internal Ca2+ by acting through the EDG receptors to generate InsP3. However, there are indications that S1P can act directly to release Ca2+ from an internal store. Unlike InsP3, however, release occurs without the appearance of elementary events. One of the difficulties with this hypothesis is that the channel on the endoplasmic reticulum (ER) that is opened by S1P remains to be identified. An earlier proposal that S1P might act through Scamper has not been substantiated. Another suggestion is that S1P functions as the Ca2+ influx factor (CIF) that is proposed to link store depletion to Ca2+ entry. Another proposed action of S1P is to activate the extracellular-signal-regulated kinase (ERK) cascade to promote proliferation. Overexpression of the SPHK can induce tumour formation. Therefore the precise role of S1P and its relationship to ER signalling require further clarification.

Ceramide

This is an enigmatic messenger in that it has been linked to both proliferation and apoptosis, depending very much on the background activity of other messenger systems. In keeping with its different functions, ceramide can activate a number of signalling components such as the ceramide-activated protein kinase (CAPK), protein kinase Cζ (PKCζ) and ceramide-activated protein phosphatases (CAPP), such as PP1 and PP2A. Some of these ceramide targets enable it to interact with other signalling systems such as the c-Jun N-terminal kinase (JNK) and mitochondrial systems. With regard to the latter, ceramide may activate mitochondrial reactive oxygen species (ROS) formation through ceramide-activated protein kinase (CAPK). It is important to stress that apoptosis can occur without the need for ceramide. However, there is overwhelming evidence that the formation of ceramide can strongly tip the balance in favour of apoptosis. It contributes to this activation of the cell death programme by activating the intrinsic pathway at the level of the mitochondria (Module 11: Figure apoptosis). One way of interacting with the mitochondria depends on its activation of the JNK pathway, which is known to induce apoptosis (Module 2: Figure JNK signalling). Ceramide can also activate cathepsin D, which converts Bid into tBid, which is one of the pro-apoptotic members of the Bcl-2 superfamily.

Ceramide is phosphorylated by ceramide kinase (CERK) to form ceramide 1-phosphate (C-1-P), which may act to enhance proliferation and cell survival.

Sphingomyelinases (SMases)

Sphingomyelinases (SMases) exist in different isoforms that can be distinguished by their pH optima: acidic, neutral or alkaline. The last is found in bile and plays a role in digestion. Sphingomyelin signalling seems to depend on the acidic and neutral isoforms (Module 2: Figure sphingomyelin signalling). The neutral SMase functions to generate ceramide in response to the activation of receptors sensitive to stimuli such as tumour necrosis factor α (TNFα) and interleukin 1 (see Step 3 in Module 2: Figure sphingomyelin signalling). The acidic form was originally identified in lysosomes where it generates ceramide in response to stress stimuli (see Step 5). There also are indications that the acidic SMase may act on sphingomyelin located in the outer leaflet. The acidic SMase is a component of lipid rafts and caveolae, which are microdomains of the plasma membrane containing high levels of the precursor sphingomyelin. It is this acidic isoform that is defective in patients with Niemann–Pick disease.

Ceramide-activated protein kinase (CAPK)

Ceramide-activated protein kinase (CAPK) is a membrane-bound proline-directed protein kinase that acts by phosphorylating Raf-1, thereby enabling ceramide to plug into the mitogen-activated protein kinase (MAPK) cascade and could account for those cases where ceramide promotes both inflammation and proliferation.

Sphingosine kinase (SPHK)

Sphingosine kinase (SPHK) appears as two isoforms (SPHK1 and SPHK2), which have different tissue distributions. SPHK1 is found at high levels in the lung and spleen, whereas SPHK2 is expressed mainly in liver and heart. Both have five highly conserved domains, with the ATP-binding site and catalytic site located in the C2 domain.

The enzyme is both membrane-bound and free in the cytosol, and there is some uncertainty as to how it is activated by cell-surface receptors. It seems likely that it responds to various downstream signals emanating from these receptors such as Ca2+/calmodulin, diacylglycerol (DAG)/protein kinase C (PKC), cyclic AMP/protein kinase A (PKA) or extracellular-signal-regulated kinase 1/2 (ERK1/2) (see Step 4 in Module 2: Figure sphingomyelin signalling). Another activator is oxidized low-density lipoprotein (LDL), which may play a role in cell proliferation. Consistent with the activation by various messenger pathways is the fact that SPHK contains consensus sequences for Ca2+/calmodulin binding and phosphorylation sites for PKA, casein kinase II and PKC.

Sphingosine 1-phosphate phosphatase (SGPP)

Mammals have two sphingosine 1-phosphate phosphatases (SGPP). Both SGPP1 and SGPP2 have eight to ten transmembrane domains that locate the enzymes within the endoplasmic reticulum (ER). SGPP1 is located mainly in the placenta and kidneys, whereas SGPP2 is found in the brain, heart, colon, kidney, small intestine and lung. These enzymes convert S1P back into ceramide and can diminish survival and promote apoptosis (see Step 6 in Module 2: Figure sphingomyelin signalling).

Ceramide-activated protein phosphatase (CAPP)

Ceramide-activated protein phosphatase (CAPP) is a member of the protein phosphatase 2A (PP2A) family of serine/threonine phosphatases that consist of three subunits: A and B are regulatory, whereas C is the catalytic subunit. In the case of CAPP, ceramide acts to stimulate phosphatase activity by binding to one of the regulatory B subunits. The ability of ceramide to inhibit growth is probably mediated through PP2A.

Scamper

A sphingolipid Ca2+-release-mediating protein of the endoplasmic reticulum (ER) (Scamper) was originally proposed to be the Ca2+ channel on the ER that responds to signals from the sphingomyelin signalling pathway, but more recent studies have shown that it has a single membrane-spanning segment that seems to remodel the actin cytoskeleton.

Janus kinase (JAK)/signal transducer and activator of transcription (STAT) signalling pathway

The Janus kinase (JAK) and the signal transducer and activator of transcription (STAT) function in the JAK/STAT signalling pathway provides a mechanism for rapidly activating gene transcription in response to a large number (>35) of external ligands. This signalling pathway is mainly activated by cytokines such as interferon, but is also used by receptor tyrosine kinases [epidermal growth factor receptor (EGFR), platelet-derived growth factor receptor (PDGFR)], non-receptor tyrosine kinases and G protein-coupled receptors (GPCRs). The Janus kinase (JAK)/signal transducer and activator of transcription (STAT) structure reveals the major features of these two components and how they are linked during the signal transducer and activator of transcription (STAT) activation cascade. Cell-surface receptors act by phosphorylating the STATs, which are latent transcription factors. Once phosphorylated, these STATs leave the membrane and then dimerize before migrating into the nucleus where they bind to specific DNA-binding elements to activate transcription. There is considerable evidence for a Janus kinase (JAK)/signal transducer and activator of transcription (STAT) function in growth and development.

Mutations in STAT3 have been linked to hyper-IgE syndrome (HIES).

Janus kinase (JAK)/signal transducer and activator of transcription (STAT) structure

The two major components of the signalling pathway are the Janus kinases (JAKs) and their substrates the signal transducers and activators of transcription (STATs) (Module 2: Figure JAK and STAT structure). Of the four mammalian JAKs, three are expressed fairly ubiquitously, whereas JAK3 is restricted to natural killer (NK) cells and thymocytes, with some expression in vascular smooth muscle cells and endothelium. The main structural component of the JAKs is the kinase domain that functions to phosphorylate the STATs at a key tyrosine in the region of residue 700 during the signal transducer and activator of transcription (STAT) activation cascade. The STATs have a number of functional domains whose three-dimensional structure reveals how the STAT dimers are formed and how they bind to DNA (Module 2: Figure STAT1/DNA complex).

Signal transducer and activator of transcription (STAT) activation cascade

The signal transducers and activators of transcription (STATs) are a family of latent cytoplasmic receptors that are activated by a phosphorylation cascade that can be induced by many different receptors (e.g. cytokine receptors, tyrosine kinase-linked receptors, non-receptor tyrosine kinases and G protein-linked receptors). The versatility of this signalling mechanism is greatly enhanced by the way the Janus kinases (JAKs) and STATs can be mixed and matched to generate an enormous number of combinations as illustrated by the different cytokine receptors (Module 2: Figure JAK/STAT heterogeneity).

A typical activation cascade for cytokine receptors, which illustrates the essential role of the JAKs in a sequence of tyrosine phosphorylations that culminates in the phosphorylation and activation of the STATs, is outlined in the following sequence (Module 2: Figure JAK/STAT function):

  1. Agonists induce dimerization by binding to the extracellular domains of the receptor subunits.
  2. The JAKs, which are associated with the cytoplasmic domain of these receptors, phosphorylate each other.
  3. The activated JAKs then phosphorylate tyrosine residues on the receptors to provide docking sites for the Src homology 2 (SH2) domains on the STATs.
  4. Once the STATs have docked, they are then phosphorylated by the JAKs on the tyrosine residue located in the C-terminal region (Module 2: Figure JAK and STAT structure).
  5. The phosphorylated STATs, which are dimerized through an intermolecular SH2–phosphotyrosine interaction, leave the receptor.
  6. These dimers are imported into the nucleus using importin-α (Imp-α) and attach to promoter regions through their DNA-binding domains (Module 2: Figure STAT1/DNA complex).
  7. STATs activate the transcription of a number of target genes. One group of these genes code for the suppressor of cytokine signalling proteins (SOCS), which thus function as part of a negative-feedback loop to limit the action of the signalling pathway (Module 2: Figure JAK/STAT function).
  8. STAT activity is terminated by a nuclear protein tyrosine phosphatase (N-PTP) that removes the tyrosyl phosphate groups. The inactive STAT is then exported from the nucleus by chromosome region maintenance 1 (CRM1), thus completing the cycle.

For some of the other receptors, these tyrosine phosphorylations are carried out by kinases other than the JAKs. After the STATs are phosphorylated, they leave the receptor and then dimerize before translocating into the nucleus. This activation cascade can occur quickly, with the activated STATs appearing in the nucleus within minutes. Transcriptional activity can be modulated by phosphorylation of the serine residue in the transcriptional activation domain (TAD) of STAT 1, 3, 4, 5A and 5B (Module 2: Figure JAK and STAT structure). In the case of STAT 1 and STAT4, phosphorylation enhances transcriptional activity, whereas the binding of STAT5a to DNA is greatly prolonged.

Suppressor of cytokine signalling proteins (SOCS)

The suppressor of cytokine signalling proteins (SOCSs) are induced during the activation of the Janus kinase (JAK)/signal transducer and activator of transcription (STAT) signalling pathway (Module 2: Figure JAK/STAT function). The SOCS act by inhibiting the Janus kinases (JAKs) and thus operate a negative-feedback loop to limit the action of cytokines.

There is a marked increase in expression of the SOCS-3 isoform in hypothalamic neurons following the action of leptin during the control of food intake and body weight (Module 7: control of food intake).

Janus kinase (JAK)/signal transducer and activator of transcription (STAT) function in growth and development

The Janus kinase (JAK)/signal transducer and activator of transcription (STAT) signalling pathway has a primary role in the regulation of growth and development, particularly of haematopoietic cells. For example, it functions in the interleukin-2 signalling pathway responsible for driving the final signalling steps in lymphocyte activation (Module 9: Figure T cell signalling map).

A mutation in the γc component of cytokine receptors results in X-linked severe combined immune deficiency (X-SCID). Since JAK3 associates with γc, it seemed likely that X-SCID may result from an alteration in the JAK/STAT signalling pathway. This was confirmed when SCID was found in patients carrying mutations in JAK3. SCID resulted from a dramatic reduction in the number of T cells, highlighting the important role for this signalling pathway in lymphocyte development.

One of the functions of STAT3 is to stimulate the transcription of genes that code for anti-apoptotic factors such as Bcl-2 and the inhibitor of apoptosis (IAP) family of proteins.

JAKs have also been implicated in certain forms of malignant transformation such as Sezary's syndrome, v-Abl-transformed cells and in some leukaemias.

Smad signalling pathway

This signalling pathway takes its name from the Smads, which are a collection of intracellular signalling molecules that act collectively to transfer information from cell-surface receptors into the nucleus. As such, some of the Smads function as transcription factors, whereas others either facilitate or inhibit this transcriptional activity. These Smads mediate the action of the transforming growth factor β (TGF-β) superfamily, which are cytokines that regulate many cellular functions such as proliferation, apoptosis, extracellular matrix formation and angiogenesis. In addition, they play a critical role in controlling events during early development and cell differentiation. There is an extensive Smad signalling toolkit, and the many components can be mixed and matched to assemble a large variety of Smad signalling pathways. The domain structure of the Smad family illustrates their multifunctional roles in the Smad signalling mechanism that functions to transfer information from activated receptors on the cell surface to gene targets in the nucleus. This signalling mechanism has two main components. Firstly, there is a process of transforming growth factor β (TGF-β) receptor activation, which is responsible for activating the Smads. Secondly, there is the Smad activation of transcription.

There are a number of mechanisms for the modulation of Smad signalling. Transforming growth factor β (TGF-β) inhibition of cell proliferation is one of the major functions of the Smad signalling pathway. It plays an important role in the differentiation of intestinal cells. Alterations in the signalling pathways controlled by the TGF-β superfamily have been associated with many cancers of epithelial and lymphoid origins, which fail to respond to the normal anti-proliferative effects of TGF-β. The TGF-β receptor is one of the tumour suppressors that are switched off in many cancers and particularly in colorectal cancer (CRC).

An increase in the expression of TGF-β1 may play a critical role in the transition from stable hypertrophy to congestive heart failure (CHF).

Smad signalling toolkit

The Smad signalling pathway is made up from a large number of components (Module 2: Table Smad signalling toolkit). Specific cell types express different combinations of these components so that there is considerable variability in the nature of Smad signalling pathways. However, the overall organization is fairly similar and is best exemplified by the action of transforming growth factor β (TGF-β) itself. The different components of the toolkit have been grouped together on the basis of their function in the signalling pathway (Module 2: Table Smad signalling toolkit):

Ligand traps

These are a large group of proteins that control access of the ligands to their cell-surface receptors. These ligand traps can have an important inhibitory action since they can prevent the ligands from reaching their cell-surface receptors. For example, follistatin inhibits the action of activin on gonadotrophs (Module 10: Figure gonadotroph regulation).

Ligands

The Smad signalling pathway is activated by a number of closely related ligands. The flagship ligand is transforming growth factor β (TGF-β), but there are a number of other ligands such as activin and nodal that also act through the Smads (Module 2: Table Smad signalling toolkit). Nodal has a role to play in establishing left–right asymmetry in the developing embryo.

Accessory receptors

These are cell-surface receptors that function as co-receptors in that they promote the binding of the ligands to their signalling receptors. For example, the accessory receptor endoglin is essential for the action of activin receptor-like kinase 1 (ALK1) during the activation of angiogenesis.

TGF-β signalling receptors

There are two groups of signalling receptors (Type I and II) (Module 2: Table Smad signalling toolkit). These receptors contain a single membrane-spanning region that separates the extracellular ligand-binding domain from the cytosolic region that contains a serine/threonine kinase domain. The Type I receptor also contains a glycine/serine-rich (GS) domain, which is phosphorylated by the Type II receptor during the signal transduction process (as occurs during the response to TGF-β Module 2: Figure TGF-βR activation). Ligands act to complex these receptors so that the Type II can then activate the Type I, which in turn activates the Smads (Module 2: Figure Smad signalling).

The activin receptor-like kinase 1 (ALK1), which responds to TGF-β, is strongly expressed in endothelial cells where it functions in the formation of blood vessels. Loss-in-function mutations in the ACVRL1 gene have been linked to hemorrhagic telangiectasia-2 (HHT2).

Smads

The Smads are the intracellular transducers of the Smad signalling pathway. There are three Smad types. Receptor-regulated Smads (R-Smads), which are activated by the signalling receptors, carry information into the nucleus. A single co-mediator Smad (Co-Smad) acts together with the R-Smads. Inhibitory Smads (I-Smads) set up a negative-feedback loop to limit the activity of the R-Smads. These different activities are reflected in the domain structure of the Smad family.

Smad regulatory factors

There are a number of proteins that contribute to the Smad signalling pathway.


Module 2: Table Smad signalling toolkit Components of the Smad signalling toolkit.

Component Comment
Ligand traps
 LAP Latency-associated polypeptide that binds TGF-β
 Decorin Binds TGF-β
 α2-Macroglobulin Binds TGF-β
 Noggin Binds to BMPs (Module 8: Figure epidermal stem cell)
 Follistatin Binds to activins and BMPs. It inhibits activin action on gonadotrophs (Module 10: Figure gonadotroph regulation)
 Chordin/SOG Binds to BMPs
 DAN/Cerberus Binds to BMPs
Ligands
 TGF-β1 Transforming growth factor β (Module 1: Figure enzyme-linked receptors)
 TGF-β2
 TGF-β3
 Activin One of its functions is to increase the transcription of FSH in gonadotrophs (Module 10: Figure gonadotroph regulation)
 Myostatin Growth and differentiation factor 8 (GDF8) (see Module 8: Figure satellite cell activation)
 Nodal
 BMP Bone morphogenetic factor functions in the control of epidermal stem cells (Module 8: Figure epidermal stem cell), proliferation of SNO cells (Module 8: Figure HSC regulation) and differentiation of white fat cells (Module 8: Figure white fat cell differentiation)
 Inhibin Inhibits activin action in gonadotrophs (Module 10: Figure gonadotroph regulation)
Accessory receptors
 Betaglycan Facilitates the binding of TGF-β to the Type II receptors. Mediates the action of inhibins in gonadotrophs (Module 10: Figure gonadotroph regulation)
 Cripto
 Endoglin This accessory receptor is essential for the action of ALK1 during the activation of angiogenesis
Signalling receptors
 Type I receptors
  ALK1 Activin receptor-like kinase 1 is also known as ACVRL1 (activin receptor-like 1) expressed mainly in endothelial cells
  ALK2 Activin receptor-like kinase 2 is also known as activin A receptor type 1 (ACVR1), usually responds to BMP
  ALK3 Activin receptor-like 3
  ALK4 Activin receptor-like 4
  ALK5 Activin receptor-like kinase 5 is also known as TGFβ type I (TβRI)
  ALK6 Activin receptor-like 6
  ALK7 Activin receptor-like 7
  ACVRI Activin receptor type I (ACVRI)
 Type II receptors
  ACVRIIA Activin receptor type IIA binds to activin and interacts with activin receptor type I (ACVRI)
  ACVRIIB
  BMPRII
  TβRII
Receptor-regulated Smads (R-Smads)
 Smad1
 Smad2
 Smad3
 Smad5
 Smad8
Co-mediator Smad (Co-Smad)
 Smad4 A germ-line mutation in Smad4 has been linked to juvenile polyposis syndromes (JPSs).
Inhibitory Smads (I-Smad)
 Smad6
 Smad7
Smad regulatory factors
 SARA Smad anchor for receptor activation
 Smurf1 Smad ubiquitin-regulatory factor 1. Smirf1 also regulates the degradation of Rho during neutrophil chemotaxis (Module 11: Figure neutrophil chemotactic signalling)
 Smurf2 Smad ubiquitin-regulatory factor 2



Domain structure of the Smad family

The main structural feature of the Smads are the two MAD homology domains (MH1 and MH2) that function both in the protein–protein and protein–DNA interactions that occur during the process of signal transduction (Module 2: Figure Smad domain structure). The regulatory Smads also have a C-terminal SSXS motif, which is critical for the transduction process because two of the serine residues are phosphorylated by the type I receptors as part of the Smad signalling mechanism. There also are numerous other sites that are phosphorylated by various kinases from other signalling pathways. The linker region between the MH1 and MH2 domains contains a phosphotyrosine (PY) motif, which is an interaction site for the binding of the Smad ubiquitin-regulatory factors 1 and 2 (Smurf1/2), which is an ubiquitin ligase that controls the selective proteolysis of the Smads. The single co-mediator Smad (Co-Smad, i.e. Smad4) resembles the R-Smads in some aspects. It also has MH1 and MH2 domains, but here the latter is split. In addition to the NLS, it also has a nuclear export signal (NES). Smad4 lacks the C-terminal phosphorylation motif, but it does contain a number of phosphorylation sites. The two inhibitory Smads (I-Smads) lack an MH1 domain, but they have the Smurf1/2-binding PY motif.

Smad signalling mechanism

The Smad signalling mechanism can be divided into two parts. Firstly, there is the process of transforming growth factor β (TGF-β) receptor activation, which concerns the way in which ligands such as TGF-β interact with the signalling receptors to trigger Smad activation (Module 2: Figure TGF-βR activation). The critical aspects of this activation process are the phosphorylation reactions that occur within the receptor complex. The Type II receptors are constitutively active and phosphorylate the Type I receptors. These activated Type I receptors then act to phosphorylate the Smads.

The second part is the Smad activation of transcription, during which the phosphorylated receptor-regulated Smads (R-Smads), together with their partner Smad4, translocate into the nucleus to induce gene transcription (Module 2: Figure Smad signalling).

Transforming growth factor β (TGF-β) receptor activation

Activation of the transforming growth factor β (TGF-β) receptor depends upon a series of reactions as illustrated by the following steps shown in Module 2: Figure TGF-βR activation.

  1. The ligand, in this case TGF-β, is often held in a latent ligand complex by being bound to one of the ligand trap proteins such as latency-associated polypeptide (LAP) or decorin. When it dissociates from the ligand trap, it can be taken up by one of the accessory receptors such as betaglycan.
  2. The TGF-β then associates with the two TGF-β receptor components to assemble an agonist/receptor complex. In the absence of ligand, the Type I and II receptor components exist as homodimers which are then brought together by TGF-β.
  3. When the two receptor types have been complexed by TGF-β, the serine/threonine kinase domain on the Type II receptors phosphorylates the serine residues on the glycine/serine-rich (GS) region of the Type I receptors.
  4. These phosphorylated GS regions on the Type I receptors provide a docking site for the MAD homology domain 1 (MH1) domain of Smad2 or Smad3. This recruitment of the Smads to the membrane is facilitated by Smad anchor for receptor activation (SARA). Once attached to the receptor, the SSXS motif is brought into contact with the serine/threonine kinase domain, and two of the serine residues are phosphorylated.
  5. Once Smad2/3 have been phosphorylated, their affinity for both the receptor complex and for SARA is reduced, and the two proteins pass into the cytoplasm. The activated Smads then translocate into the nucleus where they activate transcription (Module 2: Figure Smad signalling).

Smad activation of transcription

The receptor-regulated Smads (R-Smads) function as transcription factors responsible for activating a large number of target genes. Once they have been activated by the cell-surface receptors, the activated Smads translocate into the nucleus (Module 2: Figure Smad signalling). Some of the ligands, such as transforming growth factor β (TGF-β), activin and Nodal activate receptors that are coupled to Smad2 and Smad3, whereas bone morphogenetic protein (BMP) acts through Smad1, Smad5 or Smad8. Once these Smads have been phosphorylated on their C-terminal SSXS motif (Module 2: Figure TGF-βR activation) they leave the receptor, where they combine with Smad4 to form a dimer that then translocates into the nucleus (Module 2: Figure Smad signalling). Once the dimer enters the nucleus, it recognizes and binds to a specific DNA motif.

The transcriptional activity of the Smads is facilitated by associating with other site-specific transcription factors (TFs). Their activity is also modulated by coactivators and repressors. For example, Smad4 binds directly to the coactivator p300, which has an important role in activating transcription. The pathway on the right, which carries out the action of receptors that bind ligands such as bone morphogenetic protein (BMP), use Smad1, 5 or 8 to transmit information into the nucleus. As for the pathway described above, Smad4 is again used as a partner for the translocation process, and the transcriptional processes are also similar.

The Smads stimulate transcription of the collagen type I during the fibrogenesis induced in activated hepatic stellate cells in the liver (Module 7: Figure hepatic stellate cell).

Modulation of Smad signalling

There are a number of ways that the Smad signalling pathway can be modulated. One mechanism is by an increase in the expression of inhibitory Smads (I-Smads), which target the cell-surface receptors for degradation. This process is mediated by the Smad ubiquitin-regulatory factors (Smurfs), which are a family of ubiquitin E3 ligases that bind to the PY motif on the linker region of I-Smads (Module 2: Figure Smad domain structure). The E3 ligase functions in the ubiquitin–proteasome system (Module 1: Figure ubiquitin–proteasome system).

The mitogen-activated protein kinase (MAPK) signalling pathway can also modulate Smad signalling by phosphorylating sites located in the linker regions of Smad1 and 2 (Module 2: Figure Smad domain structure). This is an example of signalling cross-talk in that it provides a mechanism for the tyrosine kinase-linked receptors to antagonize the action of the transforming growth factor β (TGF-β) superfamily.

Transforming growth factor β (TGF-β) inhibition of cell proliferation

One of the major functions of transforming growth factor β (TGF-β) is to inhibit cell proliferation by altering expression of some of the key regulators of the cell cycle signalling pathway. One of its actions is to increase the transcription of the p15INK4B gene that codes for p15, which is one of the cyclin-dependent kinase (CDK) inhibitors that inhibits cyclin D/cyclin-dependent kinase 4 (CDK4) complex that is one of the early events of the cell cycle signalling cascade (Module 9: Figure cell cycle signalling mechanisms). Another important action is to repress c-Myc, which is one of the major transcriptional activators of genes that function in cell proliferation. Such an inhibition of Myc occurs during skin development when Smad2/3 induces the transcription of the Myc repressors Mad1 and Ovol1 to inhibit keratinocyte proliferation.

TGF-β plays an important role in regulating the proliferation of mesangial cells (Module 7: Figure mesangial cells). It also regulates the proliferation of stem cells such as the skeletal muscle satellite cells (Module 8: Figure Satellite cell activation) and the epidermal stem cells (Module 8: Figure epidermal stem cell).

Smad2 and 4 are tumour suppressors. A germline mutation in Smad4 has been linked to juvenile polyposis syndromes (JPSs).

Wnt signalling pathways

The Wnt signalling pathways play a critical role in the control of cell proliferation, cell fate specification and differentiation. These different pathways are activated by extracellular lipoprotein signalling molecules called Wnts responsible for transmitting information between cells over relatively short distances. There are three main pathways: the canonical and two non-canonical pathways (Module 2: Figure Wnt signalling pathways).

The primary function of the canonical Wnt/β-catenin pathway is to activate gene transcription to control processes during both development and in the adult organism. The functions of the non-canonical Wnt pathways are still somewhat of an enigma, since their precise functions have not been clearly identified, but there are clear indications that they may function in planar cell polarity (PCP). The Wnt/planar cell polarity (PCP) pathway, which has been characterized in Drosophila, appears to act through various GTP-binding proteins. A closely related Wnt/Ca2+ signalling pathway, which has been described in vertebrates, has a number of similar signalling components but also has additional features such as the hydrolysis of PtdIns4,5P2 to activate signalling through InsP3/Ca2+ and diacylglycerol (DAG)/protein kinase C (PKC).

Wnts

The term Wnt results from the fusion of the names of two orthologous genes, the Drosophila segment polarity gene Wingless (Wg) and a mouse proto-oncogene Int-1. The human genome has 19 Wnt genes, many of which appear to have distinct functions. They contain numerous cysteine residues, one of which is palmitoylated, thus making the Wnts somewhat insoluble, which may help to explain why they are short-range ligands. A large number of Wnt-binding proteins, such as secreted frizzled-related protein (SFRP) and Wnt inhibitory factor (WIF) also restrict their sphere of influence by acting as Wnt buffers.

Canonical Wnt/β-catenin pathway

The defining feature of the canonical Wnt/β-catenin pathway is the transcription factor β-catenin, which is responsible for regulating the transcription of Wnt target genes (Module 2: Figure Wnt canonical pathway). However, this is not the sole function of β-catenin, which also functions as a scaffolding protein in cell adhesion by providing a link between cadherin and the actin cytoskeleton. Here we consider how β-catenin functions in Wnt signalling to regulate gene transcription. There are a number of transcription factor activation mechanisms, and β-catenin belongs to those that depend on cell-surface receptors that generate cytosolic signals to activate latent transcription factors in the cytoplasm, which are then imported into the nucleus (Mechanism 2 in Module 4: Figure transcription factor activation). In the case of β-catenin, its cytosolic level is kept low because it is constantly being degraded by the proteasome as shown in the sequence of events shown in the left-hand panel in Module 2: Figure Wnt canonical pathway:

  1. Formation of β-catenin by protein synthesis occurs continuously.
  2. β-Catenin is drawn into a β-catenin degradation complex, which consists of the scaffolding protein axin, which binds the adenomatous polyposis coli (APC) tumour suppressor, the protein phosphatase PP2A, casein kinase Iα (CKIα), glycogen synthase kinase-3β (GSK-3β) and β-catenin. Within this multiprotein degradation complex, the GSK-3β phosphorylates β-catenin and thus targets it for destruction by the proteasome. Before β-catenin can be phosphorylated by GSK-3β it must first be primed by phosphorylating Ser-45 by casein kinase Iα (CKIα). PP2A, which consists of a scaffolding A subunit and a regulatory B subunit (Module 5: Figure PP2A holoenzyme), may inactivate the complex by dephosporylating GSK-3β.
  3. The phosphorylated β-catenin is recognized by the F-box/β-TrCP/ubiquitin ligase complex, which targets it for destruction by the proteasome.
  4. In the absence of β-catenin, the monomeric high-mobility group (HMG) DNA-binding proteins lymphocyte enhancer factor-1 (LEF-1) and T cell factor (TCF) inhibit the transcription of Wnt genes. An additional component of this repressor complex is histone deacetylase (HDAC), which prevents chromatin remodelling through histone deacetylation. Wnt acts to inhibit the destruction of β-catenin, which then accumulates in the cytoplasm and can enter the nucleus to promote transcription as shown in the right-hand panel in Module 2: Figure Wnt canonical pathway.
  5. Wnt initiates the signalling process by binding together two cell-surface proteins: it binds to the cysteine-rich domain (CRD) of the frizzled (Fz) receptor and it also draws in the Fz co-receptor LRP5 and LRP6 (LRP5/6), which are members of the low-density lipoprotein (LDL) receptor superfamily.
  6. The LRP5/6 co-receptor is then phosphorylated by membrane-bound casein kinase Iγ (CKIγ) isoform, which adds phosphates to multiple sites that have PPPSP motifs. Some of these motifs are also phosphorylated by a membrane-associated GSK-3β. These phosphorylated motifs then provide binding sites for the attachment of the scaffolding protein axin. Another key event is the binding of disheveled (Dsh), which is another scaffolding protein containing various signal transduction domains (e.g. DIX, PDZ and DEP). Dsh becomes hyperphosphorylated by an unknown mechanism and this contributes to its role in inhibiting the degradation complex. Axin and Dsh also bind to each other through their DIX domains. As the multiprotein complex associates with the membrane, the organization of the subunits is altered so that the activity of GSK-3β is inhibited, thus reducing the degradation of β-catenin. Loss-of-function mutations of LRP5 have been linked to osteoporosis pseudoglioma (OPPG).
  7. When GSK-3β is inhibited, the newly synthesized β-catenin is stabilized and accumulates within the cytoplasm, from where it can enter the nucleus to activate transcription.
  8. β-Catenin binds to LEF-1 and TCF to reduce their repressor activity to initiate the transcription of the Wnt genes. β-Catenin replaces HDAC with p300, which facilitates transcription by acetylating histones to remodel chromatin. This transcriptional activity of β-catenin is repressed by the chromodomain helicase DNA-binding protein 8 (CHD8) (Module 2: Figure Wnt canonical pathway). Mutations in CHD8 have been linked to autism spectrum disorders (ASD).
  9. Activation of transcription of the Wnt target genes results in the activation of the developmental and proliferative effects that characterize the operation of the Wnt signalling pathway.

In summary, the Wnt signalling pathway acts by switching off the GSK-3β-dependent degradation pathway, thus enabling β-catenin to accumulate in the cytosol and to enter the nucleus to activate transcription of the Wnt target genes.

Dishevelled (Dsh)

Dishevelled (Dsh) is a scaffolding protein that contains various transduction domains (DIX, PDZ and DEP), which functions in the Wnt signalling pathway by interacting with the frizzled (Fz) receptor (Module 2: Figure Wnt canonical pathway). One of its actions is to stimulate dishevelled-associated activator of morphogenesis 1 (DAAM1), which acts to stimulate actin formation and is particularly important in the regulation of tissue morphogenesis.

Low density lipoprotein (LDL) receptor-related protein family (LRP)

Members of the low-density lipoprotein (LDL) receptor superfamily are cell-surface proteins with multiple functions. The LDL receptor-related protein (LRP) family has multiple members (LRP1–6, LRP8, LRP10–12) with multiple functions:

Inhibitors of Wnt signalling

There are various extracellular molecules that can inhibit Wnt signalling. Some of these, such as the Wnt inhibitory factor 1 (WIF-1) and secreted frizzled-related protein (sFRP), bind to Wnt and thus prevent it from activating the Fz receptor. There is another group, such as Dickkopf (Dkk) and sclerostin (SOST) that interfere with the activity of the lipoprotein receptor-related protein (LRP) co-receptor LRP5/6. These inhibitors play an important role in preventing the proliferation of stem cells until they are required for tissue repair as occurs for the epidermal stem cells used for the hair follicle cycle (Module 8: Figure epidermal stem cell).

Wnt inhibitory factor 1 (WIF-1)

The Wnt inhibitory factor 1 (WIF-1) is a soluble extracellular factor that binds to Wnt and prevents it from interacting with the frizzled receptor (Module 2: Figure Wnt canonical pathway).

Secreted frizzled-related protein (sFRP)

The secreted frizzled-related protein (sFRP) is a soluble inhibitor that binds to Wnt and prevents it from interacting with the frizzled receptor (Module 2: Figure Wnt canonical pathway).

Dickkopf (Dkk)

Dickkopf (Dkk) binds to Kremen and the lipoprotein receptor-related protein (LRP) co-receptor LRP5/6 and this causes the complex to internalize and will thus inactivate Wnt signalling.

Sclerostin (SOST)

Sclerostin (SOST) inhibits Wnt signalling by binding to the lipoprotein receptor-related protein (LRP) co-receptor LRP5/6 (Module 2: Figure Wnt canonical pathway). SOST is produced by the osteoclasts to inhibit the proliferation and differentiation of the osteoblasts (Module 8: Figure bone cell differentiation).

Van Buchem disease is caused by a mutation in SOST that results in excessive activation of the Wnt signalling pathway.

Function of canonical Wnt signalling

The canonical Wnt pathway has an important role to play in regulating processes such as development and proliferation:

The adenomatous polyposis coli (APC) protein, which contributes to the cytoplasmic complex that degrades β-catenin in the cytoplasm, is a potent tumour suppressor that is frequently mutated in cancer cells and particularly in those that develop within the intestine such as colorectal cancer (CRC). Germline mutations of APC are responsible for familial adenomatous polyposis (FAP). Mutation in the LRP5 gene, which binds to the Wnt antagonist Dickkopf1 (DKK1), causes osteoporosis pseudoglioma (OPPG) syndrome.

Wnt/planar cell polarity (PCP) pathway

The Wnt/planar cell polarity (PCP) pathway has been characterized in insects, where it functions to establish planar cell polarity (PCP) during development (Module 2: Figure Wnt signalling pathways). Just how this pathway functions is still unclear, but there are indications that it relays information through the Rac signalling mechanisms and the Rho signalling mechanism, both of which function in the remodelling of actin. The Rho pathway acts through the Rho kinase (ROK) to activate contraction of the actin/myosin system (Module 2: Figure Rho signalling). These effects on actin remodelling and contraction may play an important role in planar cell polarity.

In vertebrates, there are similar PCP processes such as neural tube closure, the orientation of hair cell stereociliary bundles in the ear, mammalian hair follicle orientation and convergent extension (CE), during which there are large-scale movements of cells that occur during gastrulation. Components of the insect planar cell polarity pathway have also been described in vertebrates, where it includes signalling through Ca2+ and has thus been referred to as the Wnt/Ca2+ signalling pathway.

Wnt/Ca2+ signalling pathway

The Wnt/Ca2+ signalling system (Module 2: Figure Wnt signalling pathways) has been implicated in a number of planar cell polarity (PCP) processes in vertebrates. There are a number of similarities between this signalling pathway and the PCP pathway in Drosophila. They both can activate Rho and contraction of the actin/myosin system. The main difference is that the vertebrate system has an additional signalling component in that it can activate Ca2+ signalling. The activation of this Wnt/Ca2+ signalling pathway depends upon the frizzled receptor plugging into the inositol 1,4,5-trisphosphate (InsP3)/Ca2+ signalling cassette (Module 2: Figure InsP3 and DAG formation). There are reports that the Ca2+ signalling events associated with this pathway may act to inhibit the canonical Wnt/catenin pathway during dorsal/ventral axis specification (Module 8: Figure dorsoventral specification).

Mutations of the Fz4 receptor, which activates this Wnt/Ca2+ pathway, have been linked to familial exudative vitreoretinopathy (FEVR).

Casein kinase I (CKI)

The casein kinase I (CKI) family has seven members (CKIα, α2, δ, ε, γ1, γ2 and γ3), which can have very different functions in cells.

One of the functions of CK1 is to phosphorylate and activate IRBIT, which is an antagonist of the inositol 1,4,5-trisphosphate receptor (InsP3R) (Module 3: Figure InsP3R regulation).

Casein kinase Iα and α2

CKIα functions as a priming kinase for glycogen synthase kinase-3β (GSK-3β) during the operation of the Wnt signalling pathway (Module 2: Figure Wnt canonical pathway).

Casein kinase Iδ

Casein kinase Iδ (CKIδ), which is 97% homologous with casein kinase Iε (CKIε), contributes to the operation of the circadian clock by phosphorylating PER1 and PER2 to control the nuclear entry and stability of these clock proteins (Module 6: Figure circadian clock molecular mechanism). A mutation in the CKIδ gene has been linked to the rare sleep disorder familial advanced sleep phase syndrome (FASPS) and it also contributes to severe migraines.

Casein kinase Iε

CKIε participates in circadian rhythmicity by phosphorylating the PER and CRY proteins to regulate their nuclear entry and stability during the operation of the PER regulatory loop (Module 6: Figure circadian clock molecular mechanism). Mutation of CKIε, which results in a decrease in the ability of this kinase to phosphorylate the PER proteins of the circadian clock, is responsible for familial advanced sleep phase syndrome (FASPS).

Casein kinase Iγ1, γ2 and γ3

CKIγ is an unusual member of the family in that it has a palmitoylation site at its C-terminus and this fatty acid anchor attaches it to the plasma membrane. This membrane location is critical for one of its main functions, which is to phosphorylate the frizzled lipoprotein receptor-related protein (LRP) co-receptor LRP5/6 during activation of the Wnt signalling pathway (Module 2: Figure Wnt canonical pathway).

Casein kinase II (CK2)

Casein kinase II (CK2) is a serine/threonine protein kinase. It functions as a heterotetramer consisting of two 44 kDa catalytic α-subunits and two regulatory β-subunits. It has a unique ability to use GTP as well as ATP as a phosphate donor. CK2 is also known as phosvitin kinase, glycogen synthase kinase 5, troponin T kinase and casein kinase G, which reflects the fact that CK2 can phosphorylate many different substrates and thus contributes to many control mechanisms. This multifunctional kinase has been implicated in many cellular processes and seems to be particularly active in controlling cell proliferation and has also been implicated in cell transformation and tumorigenesis. Many of its actions depend on its ability to phosphorylate transcription factors such as Myc, p53, Rb, activating protein 1 (AP-1) and the vitamin D receptor (VDR). Although CK2 is constitutively active, it can also be activated by certain growth factors (insulin, IGF-I and EGF).

Hedgehog signalling pathway

The Hedgehog signalling pathway in mammals closely resembles that originally discovered and characterized in Drosophila. The mammalian Hedgehog signalling toolkit has many of the components found in insects. In comparison with the latter, however, much less is known about the mechanism of Hedgehog activation of transcription in mammals. There are multiple Hedgehog signalling functions that operate during both development and adult life.

Hedgehog signalling toolkit

The Hedgehog signalling pathway has a number of components (Module 2: Table Hedgehog signalling toolkit) whose names resemble those given to similar components in Drosophila, where this signalling system was originally identified and characterized. While many of the mammalian components have similar functions to those found in insects, it is clear that there are important differences. Since there are mammalian components that are not found in insects, the mechanism of Hedgehog activation of transcription has to be worked out separately for vertebrates.


Module 2: Table Hedgehog signalling toolkit The Hedgehog signalling toolkit.

Component Comment
Hedgehog (Hh) ligands
 Shh Sonic Hedgehog
 Ihh Indian Hedgehog
 Dhh Desert Hedgehog
Hedgehog-binding proteins
 HIP Hh-interacting protein
 Megalin A member of the low-density lipoprotein (LDL)-receptor-related family
Hedgehog receptors
 PTC1 Patched 1
 PTC2 Patched 2
Transducer
 SMO Smoothened; a seven-membrane-spanning protein
Cytoplasmic regulators
 KIF7 Functions to repress Sh responses
 FU The fused serine/threonine protein kinase
 SUFU Suppressor of fused; a negative regulator of GLI transcription factors
 MIM Missing in metastasis; an actin-binding protein that potentiates transcription
 Iguana A zinc-finger protein that promotes localization of GLI1
 IFT88 Ciliary protein that regulates GLI function
 IFT172 Ciliary protein that regulates GLI function
 FKBP8 Antagonizes Shh action
 SIL Cytoplasmic protein that acts downstream of PTC
 Rab23 This regulator of vesicular traffic is a negative regulator of Hh signalling
Transcription factors
 The multifunctional GLI transcription factors function to either activate or repress transcription
  GLI1 A zinc-finger transcriptional activator
  GLI2 A zinc-finger transcriptional activator/repressor
  GLI3 A zinc-finger transcriptional repressor


The function of some of these toolkit components are illustrated in Module 2: Figure Hedgehog signalling pathway.



Hedgehog activation of transcription

Hedgehog mediates its effects by activating gene transcription. Hedgehog uses mechanism 2 of the different transcription factor activation mechanisms found in cells (Module 4: Figure transcription factor activation). There are three Hedgehog transcription factors (GLI 1–3), which are held in an inactive state within the cytoplasm in resting cells (Module 2: Figure Hedgehog signalling pathway). This inactive state is maintained by the Hedgehog receptor patched (PTC), which inhibits the seven-membrane-spanning protein smoothened (SMO) that acts as the Hedgehog transducer. PTC is also associated with megalin, which is a member of the lipoprotein receptor-related protein family that may play a role in the trafficking of the hedgehog signalling components. In the absence of a signal from SMO, the GLI transcription factors are maintained in a latent state by interacting with a large number of cytoplasmic factors (Module 2: Table Hedgehog signalling toolkit). The precise function of all these factors is still being worked out. Hedgehog arriving at the cell surface induces a train of events that activate these transcription factors so that they translocate into the nucleus to induce gene transcription (Module 2: Figure Hedgehog signalling pathway). First of all, Hedgehog binds to its receptor PTC and this removes the inhibitory effect of PTC on SMO. The latter is then able to activate GLI by removing it from the inhibitory constraints of the cytoplasmic factors so that it is now free to translocate into the nucleus to activate transcription. Some of the genes that are activated are components of the Hedgehog signalling pathway and thus set up both positive and negative feedback loops. Many of the other genes contribute to Hedgehog signalling functions.

Expression of the GLI1 transcription factor is suppressed by miR-324-5p. Mutations in miR-324-5p result in the development of medulloblastomas.

Hedgehog signalling functions

The Hedgehog signalling pathway functions both during development and during adult life. Its function during development is wide-ranging in that it can control cell proliferation, cell determination and pattern formation of the developing embryo and the final processes of cell differentiation. Hedgehog signalling seems to play an important role in controlling the development of organs such as the skin, brain, digestive tract, pancreas and prostate. Mutations of components of this signalling pathway result in congenital malformations such as holoprosencephaly, which is a cranial defect that occurs when the midline structures of the brain and face fail to separate. The most dramatic phenotype of such malformation is cyclopia where only one eye is formed.

The function of Hedgehog continues in the adult organism, where it plays a major role in the formation and maintenance of the stem cell population. Since Hedgehog signalling plays such a major role in regulating cell proliferation, it is not surprising that alterations in this signalling pathway have been detected in many cancers such as skin cancer and some brain cancers

For example, patched (PTC) is a tumour suppressor that is inactivated in basal cell carcinomas (BCCs), medulloblastomas, gliomas in the brain and prostate cancers.

Mutations of the PTC gene are responsible for Gorlin's syndrome, which is also known as nevoid basal cell carcinoma syndrome (NBCC). The Hedgehog signalling pathway may also contribute to the growth of tumours by enhancing the activity of stromal cells that provide the tumour cell microenvironment that supports cancer cell survival and growth. Cancer cells release hedgehog that then uses a paracrine mechanism to stimulate neighbouring stromal cells such as the blood vessels, fibroblasts, immune cells and epithelial cells. These stromal cells assist cancer cell growth by providing both an extracellular matrix and essential growth factors such as insulin-like growth factor (IGF) and Wnt.

Hippo signalling pathway

The hippo signalling pathway was discovered initially in Drosophila where it functions to control proliferation, cell growth and apoptosis. This signalling pathway is highly conserved and mammalian homologues of the Drosophila hippo signalling components, which connect transduction events at the plasma membrane to alterations in gene transduction, have now been identified. (Module 2: Figure hippo signalling pathway).

In the case of the mammalian pathway, it seems likely that there are cell-surface receptors that activate the core kinase cascade, but these remain to be identified. One putative receptor is CD44, which is a receptor for hyaluronic acid and can also respond to osteopontin and collagen. The downstream target for such cell-surface receptors is likely to be the tumour suppressor merlin (moesinezrin-radixin-like protein), which is coded for by the neurofibromatosis type 2 (Nf2) gene and is sometimes referred to as NF2. Merlin then acts to regulate the activity of mammalian sterile-20-like kinase type 1 and type 2 (MST1/2), which is the first protein kinase of the core kinase cascade.

In the next of part of the cascade, MST1/2 phosphorylates /1transcriptional coactivator with PDZ-binding motif (TAZ). Once they are phosphorylated, YAP and TAZ are inactivated and leave the nucleus (Module 2: Figure hippo signalling pathway). In the case of TAZ, export from the nucleus is carried out by protein 14-3-3. YAP can also be inactivated by a ubiquitination pathway that begins with its phosphorylation by casein kinase Iδ (CKIδ) or casein kinase Iε (CKIε) that marks it out for ubiquitin ligase by SCFβ-TRCP.

The closely related transcription factors YAP and TAZ, which have overlapping and distinct functions, operate together with other co-activators of which the TEA domain (TEAD) proteins are particulary important. This transcriptional complex activates a number of target genes including amphiregulin (AR), cysteine-rich protein connective tissue factor (CTGF) and Baculoviral inhibitor of apoptosis repeat-containing 5 (BIRC5, also known as survivin). BIRC5 is a member of the inhibitor of apoptosis (IAP) family.

Hippo signalling appears to have multiple functions in mammalian cells. Many of these functions seem to be related to controlling cell proliferation and cell growth. For example, it can control the size of organs such as the liver. Contact inhibition in cultured cells seems to require activation of this signalling cascade. Its role in cell proliferation is apparent in numerous examples where uncontrolled growth occurs in many cancers when various components of the pathway are deleted. Hepatocellular carcinoma (HCC) can be induced by the overexpression of YAP.

Mammalian sterile-20-like kinase type 1 (MST1)

The mammalian sterile-20-like kinase type 1 and type 2(MST1/2) function in the hippo signalling pathway where they function by phosphorylating the large tumour suppressor-1 and -2 (Lats1/2), which is a serine/threonine protein kinase that acts as a tumour suppressor (Module 2: Figure hippo signalling pathway). The mammalian homologue of the Drosophila protein Salvador (Sav) is the cofactor WW45 (also known as Sav1) that has an important role in the activation of MST1/2.

Large tumour suppressor (Lats)

The Large tumour suppressor-1 and -2 (Lats1/2) is a serine/threonine protein kinase that acts as a tumour suppressor by inhibiting the G1/S transition of the cell cycle (Module 8: Figure ES cell miRNAs). It functions by phosphorylating and inactivating the transcription factor Yes-associated protein (YAP), which operates in the hippo signalling pathway (Module 2: Figure hippo signalling pathway). The expression of Lats1/2 is reduced my miR372 and miR-373.

Yes-associated protein (YAP)

The Yes-associated protein (YAP), which is the mammalian homologue of the Drosophila protein Yorkie (Yki), is a multifunction transcription factor that operates in the hippo signalling pathway (Module 2: Figure hippo signalling pathway). YAP shares almost 50% sequence identity with TAZ and they both have two central WW domains and a transactivation domain at the C-terminus. The transcriptional activity of YAP functions together with other co-activators such as the TEA domain (TEAD) family, which has four closely related members TEAD1–4, peroxisome-proliferator-activated receptor γ (PPARγ), thyroid transcription factor-1 (TTF-1), Runx2 and some of the SMADs. The transactivation with TEAD seems to be particularly important.

Transcriptional coactivator with PDZ binding motif (TAZ)

The transcriptional coactivator with PDZ-binding motif (TAZ) operates in the /1hippo signalling pathway (/1Module 2: Figure hippo signalling pathway). TAZ shares almost 50% sequence identity with /1YAP and they both have two central WW domains and a transactivation domain at the C-terminus. The TAZ/TEAD transcription complex may increase the expression of Zeb1, which plays a role in embryonic development by inducing the /1epithelial-to-mesenchymal transition (EMT).


Module 2: Table Notch signalling components Notch signalling toolkit.

Component Comment
Notch stimuli These Notch stimuli contain an N-terminal DSL (Delta, Serrate and LAG-2) domain that interacts with the Notch receptor (Module 2: Figure Notch signalling)
 Delta 1
 Delta 3
 Delta 4 Responsible for lateral inhibition during angiogenesis (Module 9: Figure angiogenesis signalling)
 Delta-like-4 (DLL4)
 Jagged 1
 Jagged 2
Receptors
 Notch 1
 Notch 2
 Notch 3
 Notch 4
Notch modulators
 Numb A major inhibitor of Notch. It facilitates the endocytic removal of Notch from the plasma membrane (Module 2: Figure Notch modulation)
 Crumbs May function to inhibit γ-secretase activity thereby reducing the activation of Notch
Proteases
 ADAM-10 Also known as Kuzbanian or MADM. Functions in neural and cardiac development
 ADAM-17 Also known as tumour necrosis factor-α (TACE)
 Furin A protein convertase that cleaves site 1 (S1) during Notch receptor maturation
 γ-secretase An enzyme complex containing presenilin, nicastrin, PEN2 and APH1
Glycosyltransferases
 Fringe family A family of glycosyltransferase in the Golgi that glycosylates Notch to alter its binding affinities to Delta and Jagged
  Lunatic fringe
  Radical fringe
  Manic fringe
Transcriptional regulators
 CSL CSL [CBF-1, Su(H), Lag-1] represses Notch target genes (Module 2: Figure Notch signalling)
 MAM Mastermind
 SKIP Ski-interacting protein interacts with CSL and the ankyrin repeats of NICD
Notch target genes
 HES1 Hairy and enhancer of split 1 (HES1) represses transcription of tissue specific genes responsible for differentiation
 HES2
 HES7
 HEY1
 HEY2
 HEYL
 ESR
Ubiquitin ligases
 Itch A Hect domain E3 ligase that functions in Notch trafficking (Module 2: Figure Notch modulation)
 Deltex A Ring finger E3 ligase that functions in Notch trafficking (Module 2: Figure Notch modulation)
 Mind bomb (Mib) Ring finger E3 ubiquitin ligases that initiate the trafficking of Delta and Jagged (Module 2: Figure Notch modulation)
 Neutralized (Neur) Ring finger E3 ubiquitin ligases that initiate the trafficking of Delta and Jagged (Module 2: Figure Notch modulation)
 SEL10 A nuclear ubiquitin ligase that interacts with the phosphorylated form of NICD (see Step 8 in Module 2: Figure Notch signalling)



Notch signalling pathway

The Notch signalling pathway is a highly conserved signalling system that functions in both development and adulthood. Many of its functions relate to cell-fate determination and this is particularly the case in its control of binary cell-fate decisions in stem cells. During the assymetrical divisions of stem cells, one cell retains its stem cell fate (self-renewal) while the daughter cell (progenitor cell) adopts a different state that will drive it towards proliferation and differentiation into a specific cell type (Module 8: Figure stem cell function). As these progenitor cells adopt their new cell fate, they use the Notch signalling pathway to feed information back to suppress their neighbour from adopting a similar cell fate. This is a short-range information transfer mechanism that depends upon direct contact between the cells, which is a hallmark of this signalling pathway. For example, the stimulus Jagged is an integral membrane protein located on the surface of communicating cells, whereas the Notch receptor that responds to Jagged is embedded in the surface of the receiving cell (Module 2: Figure Notch signalling).

The main feature of the transduction mechanism is deceptively simple. When Jagged binds to Notch, the Notch intracellular domain (NICD) is released into the cytoplasm and then diffuses into the nucleus where it induces the transcription of multiple Notch target genes. Despite its simplicity, this signalling system has an extensive number of interacting components (Module 2: Table Notch signalling components). Some of these function during the signal transduction process itself, whereas others play a role in the modulation of Notch signalling, which regulates the expression of the stimuli (Delta and Jagged) and the Notch receptor at the cell surface. The operation of the Notch signalling pathway depends upon the following steps (Module 2: Figure Notch signalling):

  1. Delta or Jagged located on one cell interacts with the Notch receptor on a neighbouring cell. The N-terminal DSL (Delta, Serrate and LAG-2) domain (yellow bar) on Delta/Jagged binds to the EGF-repeats 11 and 12 (red bar) on Notch. By itself, this interaction with Notch seems to have little effect and only initiates the signal transduction sequence as a result of Delta/Jagged physically pulling on Notch.
  2. The endocytosis of Delta/Jagged seems to be a critical step for receptor activation. As Delta/Jagged are drawn into the endosome, they pull on Notch, which undergoes a conformational change to expose site-2 (S2).
  3. S2 is cleaved by ADAM proteases such as ADAM-10 and ADAM-17 (Module 1: Table ADAMs proteases). The external domain is shed leaving behind a short transmembrane region and the intracellular domain, which then becomes a substrate for γ-secretase.
  4. The γ-secretase complex is an enzyme complex made up of presenilin, nicastrin, PEN2 and APH1 that cleaves the S3 site to release the Notch intracellular domain (NICD). The activity of γ-secretase may be inhibited by Crumbs, which is an apical polarity protein.
  5. NICD released from the membrane diffuses into the nucleus and binds to CSL (CBF-1, Suppressor of Hairless, Lag-1).
  6. CSL normally acts to repress Notch target genes. Repression is enhanced by CSL providing a framework to recruit co-repressors such as SMRT, SHARP, SKIP and CIR. This repression complex also recruits histone deacetylase (HDAC) that deacetylates chromatin further shutting down transcriptional activity. When NICD binds to CSL, the repression complex is disassembled and this paves the way for the assembly of an activation complex. An important part of the complex is the co-activator Mastermind (MAM) and the histone acetyl transferase p300. The latter is responsible for protein acetylation of histones to open up the chromatin to facilitate gene transcription.
  7. Gene activation results in an increase in the transcription of Notch target genes. Some of these such as the hairy and enhancer of split (HES) family encode transcriptional repressors responsible for suppressing the expression of tissue specific proteins, which accounts for the ability of the Notch signalling pathway to inhibit differentiation.
  8. The inactivation of target gene transcription begins when NICD is phosphorylated in the N-terminal PEST domain by kinases such as cyclin-dependent kinase-8 (CDK-8).
  9. The phosphorylated NICD becomes a substrate for nuclear ubiquitin ligases such as SEL-10 and is then exported to the cytoplasm where it is degraded by the proteasome.
  10. Once NICD is removed from the activation complex and degraded, the different co-activators fall away and are replaced by the repressors that result in gene repression. Gene repression also depends on the recruitment of histone deacetylase (HDAC) that deacetylates chromatin further shutting down transcriptional activity.

Notch signalling controls a number of developmental events and continues to have a role during adulthood particular in maintaining the stem cell population:

Alterations in Notch signalling have been implicated in a number of cancers. For example, overexpression of Notch has been identified in both ovarian and medullablastomas. An up-regulation of Notch and a decrease in the inhibitor Numb has been observed in breast cancer. Mutations in Notch have been linked to T cell acute lymphoblastic leukaemia.

Modulation of Notch signalling

The primary mechanism for modulating Notch signalling is to adjust the mechanisms responsible for regulating both ligand (Delta and Jagged) and receptor (Notch) maturation (Module 2: Figure Notch modulation). In both cases, the ligands and the receptor on the cell surface are taken up into the endosomes and are then either recycled back to the cell surface or are degraded. The balance between recycling and degradation thus determines their level of membrane expression and thus sets the sensitivity of the Notch signalling pathway. The ubiquitin–proteasome system plays a critical role in regulating the trafficking of these two signalling components.

The trafficking of Delta and Jagged follows the following series of events (Module 2: Figure Notch modulation):

  1. The cytoplasmic domain of Delta and Jagged are ubiquitinylated by Ring finger E3 ubiquitin ligases such as Mind bomb (Mib) and Neutralized (Neur).
  2. Endocytosis of these ubiquitylated proteins is facilitated by the protein epsin that binds to ubiquitin-containing cargo proteins. In addition, epsin binds to both clathrin and to phosphatidylinositol 4,5-bisphosphate (PtdIns4,5P2), which functions in the PtdIns4,5P2 regulation of membrane trafficking and endocytosis.
  3. The endosome then has two paths: it can move down the lysosomal path resulting in degradation of Delta and Jagged. Alternatively, it can recycle back to the plasma membrane where the ligands are then able to activate Notch.

Different players function in the regulation of Notch trafficking as shown in the following steps (Module 2: Figure Notch modulation):

  1. Entry of Notch into the endosome is dependent on the protein Numb, which is a major inhibitor of Notch signalling. The Notch receptor interacts with Numb, which also binds to α-adaptin. The latter is a component of the adaptor protein-2 (AP2) complex responsible for endocytosis. A Numb-associated kinase (NAK) is also a component of this complex. The Notch receptors in the endosome have two fates.
  2. Some of the Notch receptors are ubiquitinated by ubiquitin ligases such as Itch (a Hect domain E3 ligase) and Deltex (a Ring finger E3 ligase).
  3. Ubiquitinylated notch receptors are transferred to the multivesicular body (MVB) en route to degradation by the lysosome.
  4. Some of the Notch receptors are recycled back to the plasma membrane.
  5. These Notch receptors on the cell surface are then able to interact with ligands such as Delta and Jagged to initiate the Notch signalling pathway (Module 2: Figure Notch signalling).

Delta

Delta is a transmembrane protein that functions as one of the stimuli that activate the /1Notch receptor (/1Module 2: Table Notch signalling components). The extracellular domain is made up of EGF-like repeats and there is an N-terminal DSL (Delta, Serrate and LAG-2) domain that interacts with the Notch receptor (/1Module 2: Figure Notch signalling). In order for Delta to function as a stimulus, it has to undergo a maturation process, which depends upon its passage through the endosome (/1Module 2: Figure Notch modulation). This trafficking through the endosomal compartment determines the amount of Delta in the membrane and this is an important aspect of the /1modulation of Notch signalling. The expression of Delta-1 is increased on the surface of myofibres during muscle damage and this activates the /1Notch signalling pathway to stimulate the proliferation of /1satellite cells (/1Module 8: Figure Satellite cell activation).

Jagged

Jagged is a transmembrane protein that functions as a one of the stimuli that activate the Notch receptor (Module 2: Table Notch signalling components). The extracellular domain is made up of a cysteine-rich domain, EGF-like repeats and there is an N-terminal DSL (Delta, Serrate and LAG-2) domain that interacts with the Notch receptor. In order for Jagged to function as a stimulus, it has to undergo a maturation process, which depends upon its passage through the endosome (Module 2: Figure Notch modulation). This trafficking through the endosomal compartment determines the amount of Delta in the membrane and this is an important aspect of the modulation of Notch signalling.

Notch

Notch is a transmembrane protein that functions as the receptor for the Notch signalling pathway by responding to stimuli such as Delta and Jagged (Module 2: Figure Notch signalling). The large extracellular domain of Notch is made up of a variable number (29–36) of EGF-like repeats and a Lin21/Notch repeat region. EGF-repeats 11 and 12 (red bar) on Notch provides the binding site that interacts with the N-terminal DSL (Delta, Serrate and LAG-2) domain on Delta and Jagged. The cytoplasmic domain has a series of six ankyrin repeats and a terminal PEST domain.

Notch undergoes a number of post-translational modifications as it moves through the endoplasmic reticulum (ER) and the Golgi. While in the ER, certain sites on the EGF repeats are fucosylated by the chaperone O-fut. After the addition of these fucose groups, further extension of the carbohydrate chains is carried out by the Fringe family (Module 2: Table Notch signalling components). The degree of glycosylation can markedly influence the affinity of the Notch receptor for its ligands. The last modification, which occurs in the Golgi, is the cleavage of the molecule at site 1 (S1) by the protease; the heterodimeric receptor is then inserted into the membrane.

Trafficking of Notch through the endosomal compartment determines the amount of receptor in the membrane and this is an important aspect of the modulation of Notch signalling (Module 2: Figure Notch modulation).

Endoplasmic reticulum (ER) stress signalling

The endoplasmic reticulum (ER) has sophisticated stress signalling pathways that enable it to adapt to a whole host of stress factors mainly concerned with the way in which proteins are synthesized and packaged. Maintenance of a constant level of Ca2+ within the lumen of the ER is essential for the post-translational processing, folding and export of proteins. This protein processing is carried out by a number of Ca2+-sensitive chaperones such as 78 kDa glucose-regulatory protein (GRP78) [also known as immunoglobulin heavy-chain-binding protein (BiP)], GRP94 (endoplasmin) and calnexin. GRP78 may function as a modulator of the InsP3 receptor by regulating how this release channel is assembled (Module 3: Figure InsP3R regulation).

Any decline in the luminal level of Ca2+ results in the accumulation of misfolded proteins and the activation of the ER stress signalling pathways (Module 2: Figure ER stress signalling):

  1. After proteins are synthesized and enter the endoplasmic reticulum (ER), they undergo protein folding and are then in a state to be exported. An excessive accumulation of proteins within the ER results in the activation of the transcription factor nuclear factor κB (NF-κB), which acts to increase the production of interferons and cytokines, so contributing to an inflammatory response. In response to ER stress, the chaperones responsible for protein folding become dysfunctional resulting in an accumulation of misfolded proteins that act to induce a number of responses.
  2. Caspase 12, which is associated with the ER membrane, is activated and contributes to ER stress-induced apoptosis.
  3. Oligomerization and autophosphorylation of the inositol-requiring protein-1 (IRE1) initiates a splicing reaction of the XBP1 mRNA through a process known as regulated IRE1-α-dependent decay (RIDD) and this may contribute to the decline in translation. The transcription factor X-box binding protein 1 (XBP-1) is spliced to yield the active protein sXBP-1 that regulates transcription of ER chaperones as well as ER-associated degradation (ERAD) components such as HRD1, PDI or ER-localized DnaJ (Erdj4).
  4. Another of the transcriptional pathways depends upon the activation of the ER membrane-bound activating transcription factor 6 (ATF6). Under normal conditions, the ATF6 levels are kept low by being constantly degraded by the ubiquitin-proteasome system. The transmembrane wolframin (WFS1) protein located in the ER membrane provides a scaffold to bring together ATF6, HMG-CoA reductase degradation protein 1 (HRD1), which is an E3 ligase that ubiquitinylates ATF6 resulting in its hydrolysis by the proteasome. During ER stress, the ATF6 is released from wolframin and is then hydrolysed enabling it to enter the nucleus to activate the stress signalling targets such as BiP, C/EBP (CCAAT/enhancer-binding protein)-homologous protein 10 (CHOP) and X-Box binding protein 1 (XBP-1).
  5. The presence of misfolded proteins acts on the protein kinase RNA–like endoplasmic reticulum kinase (PERK) monomers in the ER membrane resulting in oligomerization and autophosphorylation of this resident protein kinase. The activated PERK then has a number of actions. Firstly, PERK can phosphorylate and activate nuclear factor erythroid 2 related factor 2 (NRF-2) that enters the nucleus to enhance the expression of antioxidants (Module 4: Figure NRF-2 antioxidant function). Secondly, PERK phosphorylates and inactivates the eukaryotic initiation factor eIF-2α. This phosphorylation and inhibition of eIF-2α results in protein synthesis being switched off (Module 9: Figure regulation of eIF-2α cycling). The phosphorylated eIF-2α is dephosphorylated by the serine/threonine-protein phosphatase 1 (PP1), which is recruited by GADD34 thus reversing the shut-off of protein synthesis enabling cells to recover from stress.

The PERK-induced decline in protein synthesis has two important consequences. Firstly, it reduces the protein burden within the ER by reducing the formation of misfolded proteins. Secondly, it results in the expression of the activating transcription factor 4 (ATF4) (Module 2: Figure ER stress signalling), which regulates the expression of a large number of genes that have different consequences depending on the pervading cellular conditions. One of the consequences is to stimulate the transcription of CHOP that promotes cell death by activating the expression of PUMA. ATF4 can also induce cell death by increasing the expression of Noxa.

These stress pathways are then responsible for switching off on-going protein synthesis, for up-regulating the production of new chaperones, for inducing apoptosis and for activating inflammatory responses. The degree to which these different responses are activated depends upon the nature of the stress. The fact that the ER can up-regulate chaperone levels results in the phenomenon of tolerance, whereby treatment of cells with low levels of stress stimuli can make cells much more tolerant to subsequent stressful stimuli.

There is increasing evidence for a role of the ER stress response in neurodegenerative diseases.

ER stress response in neurodegenerative diseases

A characteristic feature of many neurodegenerative diseases is an increase in the activity of the endoplasmic reticulum (ER) stress signalling pathway (Module 2: Figure ER stress signalling) that might arise through different mechanisms such as the accumulation of misfolded proteins or an alteration in the Ca2+ concentration or the redox status of the ER. There is an increase in the level of phosphorylated eukaryotic initiation factor eIF-2α in the brains of Alzheimer's disease (AD) patients. Such phosphorylation of eIF-2α by PERK would result in a decline in the neuronal protein synthesis that contributes to the calcium-induced memory loss in Alzheimer's disease (AD). A spine-specific Ca2+ signal is responsible for triggering protein synthesis at the polyribosome located at the base of each spine (step 8 in Module 10: Figure Ca2+induced synaptic plasticity).

Protein kinase RNA–like endoplasmic reticulum kinase (PERK)

Protein kinase RNA–like endoplasmic reticulum kinase (PERK) is a protein kinase that resides within the membrane of the endoplasmic reticulum where it functions in the endoplasmic reticulum (ER) stress response (Module 2: Figure ER stress signalling). Once activated by misfolded proteins it can have a number of actions. Firstly, PERK can phosphorylate and activate nuclear factor erythroid 2 related factor 2 (NRF-2) that enters the nucleus to enhance the expression of antioxidants (Module 4: Figure NRF-2 antioxidant function). Secondly, PERK phosphorylates and inactivates the eukaryotic initiation factor eIF-2α. The eIF-2α is also inactivated by other protein kinases such as haem-regulated inhibitor (HRI), double-stranded (ds) RNA-activated protein kinase (PKR) and general control non-derepressible-2 (GCN2). This phosphorylation and inhibition of eIF-2α results in protein synthesis being switched off (Module 9: Figure regulation of eIF-2α cycling). The phosphorylated eIF-2α is dephosphorylated by the serine/threonine-protein phosphatase 1 (PP1), which is recruited by GADD34 thus reversing the shut-off of protein synthesis enabling cells to recover from stress.

The PERK-induced decline in protein synthesis has two important consequences. Firstly, it reduces the protein burden within the ER by reducing the formation of misfolded proteins. Secondly, it results in the expression of the activating transcription factor 4 (ATF4) (Module 2: Figure ER stress signalling), which regulates the expression of a large number of genes that have different consequences depending on the pervading cellular conditions. One of the consequences is to stimulate the transcription of CHOP that promotes cell death by activating the expression of PUMA. ATF4 can also induce cell death by increasing the expression of Noxa.

GADD34

The growth arrest and DNA damage-inducible protein 34 (GADD34), which is also known as protein phosphatase 1 regulatory subunit 15A, is encoded by the PPP1R15A gene. GADD34 is a regulatory subunit of protein phosphatase 1 (PP1) and functions to recruit the enzyme to dephosphorylate eIF2α to remove the inhibition of protein synthesis that is induced by PERK during endoplasmic reticulum (ER) stress signalling (Module 2: Figure ER stress signalling).

Unfolded protein response (UPR)

An accumulation of misfolded proteins induces an unfolded protein response (UPR), which switches off ongoing protein synthesis and also activates various transcriptional cascades that result in the up-regulation of many of the key chaperones in an attempt to improve the defective protein packaging machinery (Module 2: Figure ER stress signalling). Activating transcription factor 6 (ATF6) is one of the transcription factors activated by the UPR.

Endoplasmic reticulum (ER) overload response (EOR)

An excessive build-up of proteins, as occurs during viral infections, switches on an endoplasmic reticulum (ER) overload response (EOR) that acts through the nuclear factor κB (NF-κB) signalling cascade to stimulate the release of interferons and cytokines as part of an inflammatory response (Module 2: Figure ER stress signalling).

Activation of apoptosis

Endoplasmic reticulum (ER) stress signalling pathways can also contribute to apoptosis (Module 11: Figure apoptosis). For example, one of the UPR pathways depends upon the release of the transcription factor activating transcription factor 6 (ATF6), which acts to switch on C/EBP (CCAAT/enhancer-binding protein)-homologous protein 10 (CHOP). Another mechanism depends upon the ER directly activating a subset of caspases during ER stress. A critical component is caspase 12, which is associated with the ER membrane and is released by proteolytic cleavage following ER stress. Several mechanisms have been proposed for this activation process. One suggestion is that the stress sensor molecule IRE1 recruits tumour-necrosis-factor-receptor-associated factor (TRAF) which then binds to caspase 12, making it sensitive to the Ca2+-responsive cysteine protease m-calpain. Another suggestion is that the hydrolysis of caspase 12 is mediated by caspase 7, which is recruited to the membrane during ER stress. An interesting aspect of this mechanism is that glucose-regulated protein 78 kDa (GRP78) appears to inhibit this activation process by forming a complex with caspase 7 and caspase 12. Once caspase 12 is released into the cytosol, it activates a specific cascade involving caspase 9 and caspase 3 in a cytochrome c-independent manner.

Metabolic messengers

There are a number of cellular metabolites that function as metabolic messengers to integrate the activities of cellular metabolism and cell signalling (/1Module 2: Figure metabolic messengers). In this context, a metabolic messenger is defined fairly widely: it includes components that are either a part of, or are derived from, cellular metabolism. Metabolism is regulated at many different levels. The most direct control depends upon feedback processes where certain substrates or products function as positive or negative regulators of the enzymes that synthesize or metabolize them. They are not signalling mechanisms in the strict sense, but are based on relatively straightforward mass action reactions that ensure that metabolism proceeds in an orderly and regulated manner. However, some of the metabolic components can be considered to be metabolic messengers because they activate or modulate clearly defined signalling pathways. There are a number of such metabolic messengers:

  • /1Adenosine triphosphate (ATP)
  • /1Adenosine monophosphate (AMP)
  • /1Fatty acids
  • /1Bicarbonate (HCO3)
  • /1Glycolytic signalling
  • /1NAD+ signalling pathways

The AMP signalling pathway is of major importance with regard to metabolic signalling systems. Components of various redox signalling systems, such as the nitric oxide signalling pathway, are also related to such metabolic signalling pathways.

Signalling through metabolites is a highly integrated system because some of the signalling pathways activated by the metabolic messengers can feed back to regulate metabolism (/1Module 2: Figure metabolic signalling). An example of how a change in metabolism can affect a signalling pathway has been described in cardiac myocytes, where the addition of pyruvate to increase metabolism has a marked effect on Ca2+ signalling (/1Module 2: Figure pyruvate and Ca2+ signalling).

AMP signalling pathway

Adenosine monophosphate (AMP)

Cells have an AMP signalling pathway that is activated by an increase in the AMP/ATP ratio, which results in the activation of an /1AMP-activated protein kinase (AMPK). AMP thus functions as a second messenger, since it is responsible for activating the signalling pathway. The AMPK that responds to AMP has been referred to as the “fuel gauge” in that it responds to a decrease in the level of ATP. This signalling cascade is sensitive to many stimuli, such as cell stress, oxidative damage, hypoxia and glucose deprivation. Once activated, AMPK induces an up-regulation of ATP-generating systems (fatty acid oxidation, glycolysis and mitochondrial biogenesis) while simultaneously down-regulating processes that consume energy (fatty acid synthesis and gluconeogenesis). One of its important actions is to reduce protein synthesis when energy levels are low by regulating the activity of the /1target of rapamycin (TOR) (/1Module 9: Figure target of rapamycin signalling). Another of its actions is to regulate the transcription of genes that function in the metabolism of glucose, fatty acids and cholesterol. The AMPK signalling pathway may also play a role in stimulating mitochondrial biosynthesis.

AMPK has also been implicated in cell growth control, where it functions to regulate the protein kinase TOR that controls protein synthesis.

AMP-activated protein kinase (AMPK)

AMP-activated protein kinase (AMPK) plays a critical role in regulating the usage of fuels such as glucose and fatty acids. However, it can have additional functions such as the regulation of insulin secretion by pancreatic β-cells, and may also play a role in controlling satiety centres in the hypothalamus. AMPK is a trimeric protein (Module 2: Figure AMPK structure) made up of multiple isoforms of a catalytic α subunit (α1 and α2) associated with β-subunits (β1 and β2) and γ-subunits (γ1–γ3). Cells express different combinations of these different isoforms.

AMPK carries out its function as a pleiotropic regulator of cell metabolism by regulating a large number of processes (Module 2: Figure AMPK control of metabolism):

  1. One of the primary regulators of AMPK is AMP, which is elevated when the level of ATP falls. Glycogen can inhibit the activity of AMPK by interacting with the glycogen-binding domain (GBD) of the β subunit (Module 2: Figure AMPK structure).
  2. Phosphorylation also plays a key role in regulating the activity of AMPK. There are two major AMPK kinases- LKB1 and Ca2+/calmodulin-dependent protein kinase kinase β (CaMKKβ). The LKB1 is thought to be constitutively active, but there are indications that its activity might be regulated by acetylation. The activation through CaMKKβ may explain how adiponectin exerts its effects on metabolism. The adiponectin receptors (AdipoR1 and AdipoR2) are thought to act by promoting the entry of Ca2+ that then stimulates CaMKKβ to phosphorylate and activate AMPK (Module 2: Figure AMPK control of metabolism).
  3. One of the important actions of AMPK is to regulate the activity of peroxisome-proliferator-activated receptor γ (PPARγ) coactivator-1α (PGC-1α), which is a key transcriptional regulator of genes involved in metabolism and particularly those that control mitochondrial biogenesis. PGC-1α activity is controlled both by its expression level and by post-translational mechanisms of which phosphorylation and acetylation are key processes. Phosphorylation of PGC-1α by AMPK acts to prime it for subsequent deacylation by the deacetylase SIRT1. AMPK also increases SIRT1 activity in two ways. Firstly, it removes deleted in breast cancer 1 (DBC1) leaving SIRT1 free to deacylate PGC-1α. Secondly, AMPK acts to increase the level of NAD+, which enhances the activity of SIRT1. The activity of PGC-1α is inhibited following its acetylation by the general control of amino-acide synthesis 5 (GCN5).
  4. One of the primary actions of PGC-1α is to stimulate the expression of genes that contribute to ATP generation, such as those that function in fatty acid oxidation, glycolysis and mitochondrial biogenesis.
  5. AMPK phosphorylates TORC2 to prevent it from acting as a cofactor to activate genes responsible for gluconeogenesis. For example, in the case of liver cells, AMPK can phosphorylate transducer of regulated cyclic AMP response element-binding protein 2 (TORC2), which is then prevented from acting as a cofactor for the cyclic AMP response element-binding protein (CREB) transcription factor, which can activate genes responsible for gluconeogenesis (Module 7: Figure liver cell signalling). AMPK can also inhibit the activity of various transcription factors such as the sterol regulatory element-binding protein 1c (SREBp1c) and hepatocyte nuclear factor 4α (HNF4α), which regulate lipogenic and glycolytic genes respectively.
  6. Lipid metabolism is strongly influenced by AMPK, which acts to divert fatty acids away from lipid synthesis and directs them towards the mitochondrial oxidative pathway to produce ATP. The AMPK-dependent phosphorylation of acetyl-CoA carboxylase (ACC) reduces the conversion of acetyl-CoA into malonyl-CoA.
  7. The fall in malonyl-CoA levels has two import consequences. First, fatty acid synthesis is reduced because malonyl-CoA is an important precursor for lipid synthesis. Secondly, there is an increase in mitochondrial ATP formation because malonyl-CoA normally inhibits fatty acid oxidation.
  8. AMPK can influence the balance between glycolysis and gluconeogenesis by stimulating the formation of fructose 2,6-bisphosphate (F-2,6-P2), which is a potent regulator of glycolysis through its ability to activate 6-phosphofructo-1-kinase (PFK-1) and to inhibit fructose-1,6-bisphosphate 1-phosphatase. The formation of F-2,6-P2 is regulated by two separate signalling pathways. AMPK promotes glycolysis by stimulating the phosphofructokinase (PFK-2) of the bifunctional enzyme, which has both kinase and phosphatase activities. Cyclic AMP acting through the fructose-2,6-bisphosphate 2-phosphatase component lowers the level of F-2,6-P2, which reduces glycolysis and promotes gluconeogenesis.
  9. AMPK stimulates the translocation of the glucose transporter (GLUT4) to the plasma membrane, where it facilitates the entry of glucose in skeletal muscle and heart cells.
  10. AMPK can also be activated by endogenous hormones such as glucagon or by drugs such as Resveratol. Glucagon, which is produced in the α-cells in the Islets of Langerhans (Module 7: Figure pancreas), is released when blood glucose levels fall too low as occurs during calorie restriction (CR). Glucagon acts through the cyclic AMP signalling pathway that uses protein kinase A (PKA) to stimulate LKB1 that then activates AMPK. Cyclic AMP may also act through the exchange proteins activated by cAMP (EPACs) (Module 2: Figure cyclic AMP signalling). Many of the effects of CR can be reproduced by Resveratol, which may have many beneficial effects with regard to age-related diseases. The mode of action of this drug is not absolutely clear, but there are indications that it may act by inhibiting phosphodiesterase (PDE) resulting in an increase in cyclic AMP and activation of AMPK.

In summary, AMPK switches the cell away from energy-requiring processes towards energy conservation. It does this by exerting rapid effects on processes such as glucose entry and glycolysis, as well as longer-term effects, by regulating the transcription of genes for mitochondrial biogensis and glycolytic and lipogenic hormones (Module 2: Figure AMPK control of metabolism). The AMPK signalling pathway functions in many different cell types:

The fact that AMPK plays such a central role in regulating energy metabolism has important implications for diabetes.

/1Glycogen storage disease in humans is caused by a mutation of the AMPK γ3-subunit.

LKB1

LKB1 is a serine/threonine protein kinase that functions to activate AMP-activated protein kinase (AMPK) by phosphorylating Thr-172 on the α-subunit (Module 2: Figure AMPK structure). The activation of AMPK is very dependent upon LKB1. Inactivation of the latter is responsible for Peutz–Jeghers syndrome. As such, LKB1 is considered to be one of the tumour suppressors.

Adenosine triphosphate (ATP)

ATP is an important metabolic messenger in that it has a number of different actions both as an internal and an external signal (Module 2: Figure metabolic messengers). With regard to the former, one of its main functions is to regulate the activity of the ATP-sensitive K+ (KATP) channel, which is particularly important in regulating the release of insulin (Module 7: Figure β-cell signalling). It also plays a role in regulating the activity of the ADP-ribosyl cyclase that generates the Ca2+-mobilizing messengers cyclic ADP-ribose (cADPR) and nicotinic acid–dinucleotide phosphate (NAADP) (Module 2: Figure cADPR/NAADP function). ATP is also released from the cell and can activate P2X receptors that activate Ca2+ entry or P2Y receptors that are coupled to phospholipase C (PLC) to generate inositol 1,4,5-trisphosphate (InsP3) and diacylglycerol (DAG) (Module 2: Figure metabolic messengers).

Bicarbonate (HCO3)

The CO2 produced during cellular metabolism is rapidly converted into bicarbonate (HCO3), which acts as a messenger to report the current state of metabolism. The HCO3 activates soluble adenylyl cyclase (Module 2: Figure cyclic AMP signalling) to generate cyclic AMP that has many signalling functions, including an effect on cellular metabolism through its activation of glycogen metabolism.

Fatty acids

Fatty acids come in two main forms: saturated and polyunsaturated. There are two main types of polyunsaturated fatty acid: omega-6 and omega-3 fatty acids. These are referred to as essential fatty acids since they are not made by the body but are components of the diet: omega-6 fatty acids are derived mainly from vegetable oils, whereas the omega-3 fatty acids are found in certain plants (linolenic acid) or from marine fish and shellfish (eicosapentanoic acid and docosahexanoic acid). Fatty acids not only provide an important energy source but can also be considered as metabolic messengers that contribute to a number of metabolic control mechanisms. For example, they control gluconeogenesis in liver cells by activating the peroxisome-proliferator-activated receptor α (PPARα) (Module 7: Figure liver cell signalling). Free fatty acids are also responsible for activating the uncoupling protein 1 (UCP1) that provides the proton leak during noradrenaline-induced heat production by brown fat cell mitochondria (Module 7: Figure brown fat cell). The omega-3 fatty acids have a number of specific signalling functions.

Arachidonic acid (AA) and other polyunsaturated fatty acids (PUFAs) such as docosahexenoic acid, linolenic and linoleic acid can activate the TREK subfamily of two-pore domain K+ (K2P) channels.

The positive-feedback loop that free fatty acids (FFAs) exert on the process of lipogenesis (i.e. Step 7 in Module 7: Figure metabolic energy network) may exacerbate the onset of obesity by enhancing fat storage.

The build-up of FFAs that occur during obesity play a critical role in inducing the insulin resistance (Module 12: Figure insulin resistance) that leads to the development of diabetes.

Omega 3 fatty acids

The omega-3 fatty acids such as α-linolenic acid (ALA), eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) are typical polyunsaturated fatty acids (PUFAs). These PUFAs are essential for normal cellular function. ALA and α-linolenic acid, which are the precursors of arachidonic acid and DHA respectively, are not synthesized de novo but are obtained from the diet. Arachidonic acid is the precursor of the endocannabinoids, such as anandamide and 2-arachidonoylglycerol (2-AG), which have important signalling functions (Module 1: Figure anandamide).

The first double bond of these omega-3 fatty acids is located at the third carbon atom from the methyl (CH3) end of the molecule (see white arrow in Module 2: Figure omega-3 fatty acids). These omega-3 fatty acids have a number of specific signalling functions as outlined in Module 2: Figure omega-3 fatty acids:

  1. When GPR120 was first identified, it was classed as an orphan receptor as there was no known ligand. Subsequently, the omega-3 fatty acids were identified as stimuli for this typical G-protein-coupled receptor (GPCR).
  2. GPR120 is coupled through Gq/11 and phospholipase C to activate the Inositol 1,4,5-trisphosphate (InsP3) and Ca2+ release signalling pathway.
  3. One of the functions of the increase in Ca2+ might be to stimulate the release of glucagon-like peptide 1 (GLP-1).
  4. Another function of the GPR120 receptors is to activate the multifunctional adaptor protein arrestin-2.
  5. The anti-inflammatory responses of omega-3 fatty acids might be carried out by arrestin-2 inactivating the TAK1-binding protein 1 (TAB1), which can inhibit the signalling pathways used by the Toll receptor signalling pathway (Module 2: Figure Toll receptor signalling) or by the tumour necrosis factor α (TNFα) signalling pathway (Module 2: Figure NF-κB activation).
  6. Apart from the intracellular signalling pathways described above, the omega-3 fatty acids are metabolized by cyclooxygenase 2 (COX2) to form the resolvins and the protectins, which are potent anti-inflammatory and immunoregulatory agents.
  7. The anti-inflammatory action of the omega-3 fatty acids might have beneficial effects for coronary artery disease, cardiac disease and for the insulin resistance that results in metabolic syndrome.

A decrease in dietary omega-3 fatty acids results in a decline in the ability of anandamide to activate long-term depression (LTD) in neurons and this might be the cause of various neuropsychiatric diseases such as depression. In the case of inflammation and Alzheimer's disease, omega-3 fatty acids may act to curb the build up of amyloid β (Aβ) by stimulating its phagocytosis by the microglia (Module 12: Figure Inflammation and Alzheimer's disease).

Glycolytic signalling

The primary role of glycolysis is to extract energy from glucose. Glycolysis operates in three different modes. In all three modes, glucose proceeds through a number of metabolic steps that culminates in the formation of phosphoenolpyruvate (PEP), which is then converted into pyruvate or lactate depending on the activity of two different pyruvate kinase isoforms, either M1 or M2 (PKM1, PKM2) (Module 2: Figure glycolytic signalling pathways):

  1. Pyruvate kinase M1 (PKM1), which is constitutively active, converts PEP into pyruvate that is then converted into acetyl-CoA by pyruvate dehydrogenase (PDH). Acetyl-CoA enters the mitochondria to fuel the generation of ATP.
  2. Similar to pathway 1 except that PEP is converted into pyruvate by pyruvate kinase M2 (PKM2), which is allosterically regulated by fructose 1,6-bisphosphate (F-1,6P2).
  3. Similar to pathway 2 except that the activity of PKM2 is inhibited, either by phosphorylation through tyrosine kinase signalling pathways or through oxidation by reactive oxygen species (ROS), that result in the homotetramer being converted into homodimers that have a number of signalling consequences:

Pyruvate kinase

Pyruvate kinase is a rate-limiting glycolytic enzyme that plays a key role in glycolysis by converting phosphoenolpyruvate (PEP) and ADP into pyruvate and ATP (Module 2: Figure glycolytic signaling pathways). There are two pyruvate kinase genes, PKLR and PKM2. The PKLR gene expresses the PKL and PKR isoforms whereas the PKM2 codes for the PKM1 and PKM2 pyruvate kinase isoforms. These four isoforms have different properties and cellular locations:

  • PKL is expressed in the kidney and liver.
  • PKR is found in erythrocytes.

Pyruvate kinase M1 (PKM1) functions as a tetramer to regulate glycolysis in bladder, brain, fibroblasts and muscle. Unlike the other three isoforms, PKM1 is not allosterically activated by fructose 1,6-bisphosphate (FBP) (see pathway 1 in Module 2: Figure glycolytic signaling pathways).

Pyruvate kinase M2 (PKM2) is expressed in most cells except in adult muscle. In its homotetramer form it is allosterically regulated by fructose 1,6-bisphosphate (FBP) to rapidly transfer metabolites to the mitochondria (see pathway 2 in Module 2: Figure glycolytic signaling pathways). However, this active tetramer is disrupted and converted into homodimers by either tyrosine kinase signalling pathways or through an increase in reactive oxygen species (ROS). These PKM2 homodimers play an important role in glycolytic signalling in that they can enter the nucleus to activate the transcription of a number of different genes (see pathway 3 in Module 2: Figure glycolytic signaling pathways). In some of these signalling functions, PKM2 is acting as a protein kinase.

NAD signalling pathways

As the name implies, the NAD signalling pathways encompass signalling systems that depend on nicotinamide–adenine dinucleotide (NAD+) and nicotinamide–adenine dinucleotide phosphate (NADP+). These two metabolic cofactors exist in the cell as NAD+/NADH and NADP+/NADPH redox couples and have a number of important functions both as messengers and as precursors of other metabolic messengers (Module 2: Figure NAD-dependent signalling pathways).

NAD and NADP as metabolic messengers

  • NAD+ functions in /1protein acetylation and protein ribosylation (/1Module 2: Figure NAD-dependent signalling pathways) to regulate a number of cellular processes including energy metabolism, gene transcription, DNA repair and perhaps /1ageing as well.
  • NADH activates the /1C-terminal binding protein (CtBP), which is a transcriptional corepressor that functions during growth and development.
  • NADH and NADPH regulate the transcription factors Clock/BMAL1 and NPAS2/BMAL1 that control gene expression during the operation of the circadian clock.
  • NADPH appears to regulate the ADP-ribosyl cyclase that is responsible for producing both cyclic ADP-ribose (cADPR) and nicotinic acid–adenine dinucleotide phosphate (NAADP) (Module 2: Figure cADPR/NAADP function).

NAD and NADP as precursors of metabolic messengers

  • NAD+ is a precursor of /1cyclic ADP-ribose (cADPR) signalling pathway that functions in Ca2+ signalling (/1Module 2: Figure NAD-dependent signalling pathways).
  • Metabolism of cADPR generates ADPR, which regulates transient receptor potential melastatin 2 (/1TRPM2), a member of the transient receptor potential (TRP) ion channel family that controls the entry of external Ca2+ (/1Module 2: Figure cADPR/NAADP function).
  • NADP+ is a precursor of /1nicotinic acid–adenine dinucleotide phosphate (NAADP) signalling pathway that functions in Ca2+ signalling.
  • NADPH is the substrate used to generate /1reactive oxygen species (ROS), which function as second messengers to regulate a number of cellular proteins (/1Module 2: Figure plasma membrane ROS formation).

Sterol sensing and cholesterol biosynthesis

The level of cholesterol in cell membranes is regulated by a signalling system that can sense the level of sterols and then relay information to the nucleus to adjust the transcription of the genes responsible for cholesterol biosynthesis. The signalling system is based on membrane-bound transcription factors: the sterol regulatory element-binding proteins (SREBPs), which are integral membrane proteins located in the endoplasmic reticulum (ER). They have two membrane-spanning domains with the free ends projecting into the cytosol. The N-terminal region is the latent transcriptional regulator, which is cleaved by a sterol-regulated system of proteases (Module 2: Figure sterol sensing). Once released into the cytosol, these transcription factors dimerize through a basic loop–helix–leucine zipper and enter the nucleus, where they bind to sterol regulatory elements on the genes that code for the enzymes of lipid biosynthesis (cholesterol, unsaturated fatty acids and triacylglycerols) and lipid-uptake mechanisms. The sterol sensor is a SREBP cleavage-activating protein (SCAP). When sterols build up in the membranes, they bind to SCAP to inhibit the proteolytic cleavage of the SREBPs. The ER thus plays a central role both in sensing the level of sterols and then providing the output signals that make the necessary adjustments to lipid synthesis to maintain a constant level of membrane cholesterol.

REFERENCES

Cullen, P.J. and Lockyer, P.J. (2002) Integration of calcium and Ras signalling. Nat. Rev. Mol. Cell Biol. 3, 339-348

Manser, E. (2003) Cytoskeletal regulation: small G-protein–kinase interactions. In Handbook of Cell Signaling, Vol. 1 (Bradshaw R.A. and Dennis, E.A., eds), pp. 499-503, Academic Press, San Diego

Mitlin, N., Rossman, K.L. and Der, C.J. (2005) Signalling interplay in Ras superfamily function. Curr. Biol. 15, R563-R574

Rossman, K.L., Der, C.J. and Sondek, J. (2005) GEF means go: turning on Rho GTPases with guanine nucleotide-exchange factors. Nat. Rev. Mol. Cell. Biol. 6, 167-180

Somlyo, A.P. and Somlyo, A.V. (2003) Ca2+ sensitivity of smooth muscle and nonmuscle myosin II: modulated by G proteins, kinases, and myosin phosphatase. Physiol. Rev. 83, 1325-1358

Tybulewicz, V.L.J. (2005) Vav-family proteins in T-cell signalling. Curr. Opin. Immunol. 17, 267-274

Welch, H.C.E., Coadwell, W.J., Stephens, L.R. and Hawkins, P.T. (2003) Phosphoinositide 3-kinase-dependent activation of Rac. FEBS Lett. 546, 93-97

Cooper, D.M. (2003) Regulation and organization of adenylyl cyclases and cAMP. Biochem. J. 375, 517-529

Whorton, M.R. and Sunahara, R.K. (2003) Adenylyl cyclases. In Handbook of Cell Signaling, Vol. 2 (Bradshaw R.A. and Dennis, E.A., eds), pp. 419-426, Academic Press, San Diego

Feig, L.A. (2003) Ral-GTPase: approaching their 15 minutes of fame. Trends Cell Biol. 13, 419-425

Myers, K.R. and Casanova, J.E. (2008) Regulation of actin cytoskeleton dynamics by Arf-family GTPases. Trend Cell Biol. 18, 184-192

Stenmark, H. (2009) Rab GTPases as coordinators of vesicle traffic. Nat. Rev. Mol. Cell Biol. 10, 513-525

Berridge, M.J., Bootman, M.D. and Roderick, H.L. (2003) Calcium signalling: dynamics, homeostasis and remodelling. Nat. Rev. Mol. Cell Biol. 4, 517-529

Maravall, M., Mainen, Z.F., Sabatini, B.L. and Svoboda, K. (2000) Estimating intracellular calcium concentration and buffering without wavelength ratioing. Biophys. J. 78, 2655-2667

Galione, A. and White, A. (1994) Ca2+ release induced by cyclic ADP-ribose. Trends Cell Biol. 4, 431-436

Hua, S.-Y., Tokimasa, T., Takasawa, S., Furuya, S., Nohmi, M., Okamoto, H. and Kuba, K. (1994) Cyclic ADP-ribose modulates Ca2+ release channels for activation by physiological Ca2+ entry in bullfrog sympathetic neurons. Neuron 12, 1073-1709

Lee, H.C. (1997) Mechanisms of calcium signalling by cyclic ADP-ribose and NAADP. Physiol. Rev. 77, 1133-1164

Lukyanenko, V., Györke, I., Wiesner, T.F. and Györke, S. (2001) Potentiation of Ca2+ release by cADP-ribose in the heart is mediated by enhanced SR Ca2+ uptake into the sarcoplasmic reticulum. Circ. Res. 89, 614-622

Lee, H.C. (2003) Ca2+ signalling: NAADP ascends as a new messenger. Curr. Biol. 13, R186-R188

Patel, S., Churchill, G.C. and Galione, A. (2001) Coordination of Ca2+ signalling by NAADP. Trends Biochem. Sci. 26, 482-489

Delmas, P., Crest, M. and Brown, D.A. (2004) Functional organization of PLC signalling microdomains in neurons. Trends Neurosci. 27, 41-47

Katan, M. (1998) Families of phosphoinositide-specific phospholipase C: structure and function. Biochim. Biophys. Acta 1436, 5-17

Kelley, G.G., Reks, S.E., Ondrako, J.M. and Smrcka, A.V. (2001) Phospholipase Ce: a novel Ras effector. EMBO J. 20, 743-754

Rebecchi, M.J. and Pentyala, S.N. (2000) Structure, function, and control of phosphoinositide-specific phospholipase C. Physiol. Rev. 80, 1291-1335

Rhee, S.G. and Bae, Y.S. (1997) Regulation of phosphoinositide-specific phospholipase C isozymes. J. Biol. Chem. 272, 15045-15048

Singer, W.D., Brown, H.A. and Sternweis, P.C. (1997) Regulation of eukaryotic phosphatidylinositol-specific phospholipase C and phospholipase D. Annu. Rev. Biochem. 66, 475-509

Irvine, R.F. (2003) Nuclear lipid signalling. Nat. Rev. Mol. Cell Biol. 4, 349-361

Martelli, A.M., Manzoli, L., Faenza, I., Bortul, R., Billi, A.M. and Cocco, L. (2002) Nuclear inositol lipid signalling and its potential involvement in malignant transformation. Biochim. Biophys. Acta 1603, 11-17

Czech, M.P. (2000) PIP2 and PIP3: complex roles at the cell surface. Cell 100, 603-606

Yin, H.Y. and Janmey, P.A. (2003) Phosphoinositide regulation of the actin cytoskeleton. Annu. Rev. Physiol. 65, 761-789

Cockcroft, S. and De Matteis, M.A. (2001) Inositol lipids as spatial regulators of membrane traffic. J. Membr. Biol. 180, 187-194

Di Paolo, G. and De Camilli, P. (2006) Phosphoinositide in cell regulation and membrane dynamics. Nature 443, 651-657

Ikonomov, O.C., Sbrissa, D., Fenner, H. and Shisheva, A. (2009) PIKfyve-ArPIKfyve-Sac3 core complex. J. Biol. Chem. 284, 35794-35806

Suh, B.-C. and Hille, B. (2005) Regulation of ion channels by phosphatidylinositol 4,5-bisphosphate. Curr. Opin. Neurobiol. 15, 370-378

McCrea, H.J. and De Camilli, P. (2009) Mutations in phosphoinositide metabolizing enzymes and human disease. Physiology 24, 8-16

McLaughlin, S., Wang, J., Gambhir, A. and Murray, D. (2002) PIP2 and proteins: interactions, organization, and information flow. Annu. Rev. Biophys. Biomol. Struct. 31, 151-175

Pendaries, C., Tronchère, H., Racaud-Sultan, C., Gaits-Iaconvani, F., Coronas, S., Manenti, S., Gratacap, M.-P., Plantavid, M. and Payrastre, B. (2005) Emerging roles of phosphatidylinositol monophosphates in cellular signalling and trafficking. Adv. Enzyme Regul. 45, 201-214

Takenawa, T. and Itoh, T. (2001) Phosphoinositides, key regulatory molecules for regulation of actin cytoskeleton and membrane traffic from the plasma membrane. Biochim. Biophys. Acta 1533, 190-206

Vicinanza, M., D'Angelo, G., Di Campli, A. and De Matteis, M.A. (2008) Function and dysfunction of the PI system in membrane trafficking. EMBO J. 27, 2457-2470

Hawkins, P.T., Anderson, K.E., Davidson, K. and Stephens, L.R. (2006) Signalling through Class I PI3Ks in mammalian cells. Biochem Soc. Trans. 34, 647-662

Leslie, N.R. and Downes, C.P. (2004) PTEN function: how normal cells control it and tumour cells lose it. Biochem. J. 382, 1-11

Woodgett, J.R. (2005) Recent advances in the protein kinase B signalling pathway. Curr. Opin. Cell Biol. 17, 150-157

Wymann, M.P. and Marone, R. (2005) Phosphoinositide 3-kinase in disease: timing, location, and scaffolding. Curr. Opin. Cell Biol. 17, 141-149

Shears, S.B. (2004) How versatile are inositol phosphate kinases? Biochem. J. 377, 265-280

Bogdan, C. (2001) Nitric oxide and the regulation of gene expression. Trends Cell Biol. 11, 66-75

Moncada, S., Palmer, R.M.J. and Higgs, E.A. (1991) Nitric oxide: physiology, pathophysiology, and pharmacology. Pharmacol. Rev. 43, 109-142

Pacher, P., Beckman, J.S. and Liaudet, L. (2007) Nitric oxide and peroxynitrite in health and disease. Physiol. Rev. 87, 315-424

Shaul, P.W. (2002) Regulation of endothelial nitric oxide synthase: location, location, location. Annu. Rev. Physiol. 64, 749-774

Stamler, J.S. (1994) Redox signalling: nitrosylation and related target interactions of nitric oxide. Cell 78, 931-936

Friebe, A. and Koesling, D. (2003) Regulation of nitric oxide-sensitive guanylyl cyclase. Circ. Res. 93, 96-105

Hofmann, F. (2005) The biology of cyclic GMP-dependent protein kinases. J. Biol. Chem. 280, 1-4

Hofmann, F., Ammendola, A. and Schlossmann, J. (2000) Rising behind NO: cGMP-dependent protein kinase. J. Cell Sci. 113, 1671-1676

Lucas, K.A., Pitari, G.M., Kazerounian, S., Ruiz-Stewart, I., Park, J., Schulz, S., Chepenik, K.P. and Waldman, S.A. (2000) Guanylyl cyclases and signalling by cyclic GMP. Pharmacol. Rev. 52, 375-413

Foster, M.W., McMahon, T.J. and Stamler, J.S. (2003) S-nitrosylation in health and disease. Trends Mol. Med. 9, 160-168

Marshall, H.E., Hess, D.T. and Stamler, J.S. (2004) S-nitrosylation: physiological regulation of NF-κB. Proc. Natl. Acad. Sci. U.S.A. 101, 8841-8842

Bokoch, G.M. and Knaus, U.G. (2003) NADPH oxidases: not just for leukocytes anymore. Trends Biochem. Sci. 28, 502-508

Cheek, T.R., McGuinness, O.M., Vincent, C., Moreton, R.B., Berridge, M.J. and Johnson, M.H. (1993) Fertilisation and thimerosal stimulate similar calcium spiking patterns in mouse oocytes but by separate mechanisms. Development 119, 179-189

Chiarugi, P. and Cirri, P. (2003) Redox regulation of tyrosine phosphatases during receptor tyrosine kinase signal transduction. Trends Biochem. Sci. 28, 509-514

Dröge, W. (2002) Free radicals in the physiological control of cell function. Physiol. Rev. 82, 47-95

Lambeth, J.D. (2004) NOX enzymes and the biology of reactive oxygen. Nat. Rev. Immunol. 4, 181-189

Mikkelsen, R.B. and Wardman, P. (2003) Biological chemistry of reactive oxygen and nitrogen and radiation-induced signal transduction mechanisms. Oncogene 22, 5734-5754

Rhee, S.G., Kang, S.W., Jeong, W., Chang, T.-S., Yang, K.-S. and Woo, H.A. (2005) Intracellular messenger function of hydrogen peroxide and its regulation by peroxyredoxins. Curr. Opin. Cell Biol. 17, 183-189

Singh, D.K., Kumar, D., Siddiqui, Z., Basu, S.K., Kumar, V. and Rao, K.V.S. (2005) The strength of receptor signaling is centrally controlled through a cooperative loop between Ca2+ and an oxidant signal. Cell 121, 281-293

Castro-Piedras, I. and Perez-Zoghbi, J.F. (2013) Hydrogen sulphide inhibits Ca2+ release through InsP3 receptors and relaxes airway smooth muscle. J. Physiol. 591, 5999-6015

Kimura, H., Shibuya, N. and Kimura, Y. (2012) Hydrogen sulfide is a signaling molecule and a cytoprotectant. Antioxid. Redox Signal. 17, 45-57

Paul, B.D. and Snyder, S.H. (2012) H2S signalling through protein sulfhydration and beyond. Nat. Rev. Mol. Cell Biol. 13, 499-507

Wang, R. (2012) Physiological implications of hydrogen sulfide: a whiff exploration that blossomed. Physiol. Rev. 92, 791-896

Davis, R.J. (2000) Signal transduction by the JNK group of MAP kinases. Cell 103, 239-252

Hattori, K., Naguro, I., Runchel, C. and Ichijo, H. (2009) The roles of ASK family proteins in stress responses and disease. Cell Commun. Signal. 7, 9

Johnson, G.L. and Lapadat, R. (2002) Mitogen-activated protein kinase pathways mediated by ERK, JNK, and p38 protein kinases. Science 298, 1911-1912

Morrison, D.K. and Davis, R.J. (2003) Regulation of MAP kinase signalling modules by scaffold proteins in mammals. Annu. Rev. Cell Dev. Biol. 19, 91-118

Pouysségur, J. and Lenormand, P. (2003) Fidelity and spatio-temporal control of MAP kinase (ERKs) signalling. Eur. J. Biochem. 270, 3291-3299

Wellbrock, C., Karasarides, M. and Marais, R. (2004) The Raf proteins take centre stage. Nat. Rev. Mol. Cell Biol. 5, 875-885

Adhikari, A., Xu, M. and Chen, ZJ. (2007) Ubiquitin-mediated activation of TAK1 and IKK. Oncogene 26, 3214-3226

Akira, S. (2003) Toll-like receptor signalling. J. Biol. Chem. 278, 38105-38108

Akira, S. and Takeda, K. (2004) Toll-like receptor signalling. Nat. Rev. Immunol. 4, 499-511

Hayden, M.S. and Gosh, S. (2008) Shared principles in NF-κB signalling. Cell 132, 344-362

Nakanishi, C. and Toi, M. (2005) Nuclear factor-κB inhibitors and sensitizers to anticancer drugs. Nat. Rev. Cancer 5, 297-309

Perkins, N.D. (2007) Integrating cell-signalling pathways with NF-κB and IKK function. Nat. Rev. Mol. Cell Biol. 8, 49-62

Viatour, P., Merville, M.-P., Bours, V. and Chariot, A. (2004) Phosphorylation of NF-κB and IκB proteins: implications in cancer and inflammation. Trends Biochem. Sci. 30, 43-52

Westwick, J.K., Schwamborn, K. and Mercurio, F. (2003) NFκB: a key integrator of cell signalling. In Handbook of Cell Signaling, Vol. 3 (Bradshaw R.A. and Dennis, E.A., eds), pp. 107-114, Academic Press, San Diego

Martinon, F., Mayor, A. and Tschopp, J. (2009) The inflammasomes: guardians of the body. Annu. Rev. Immunol. 27, 229-265

McCartney, S.A. and Colonna, M. (2009) Viral sensors: diversity in pathogen recognition. Immunol. Rev. 227, 87-94

Takeuchi, O. and Akira, S. (2009) Innnate immunity to virus infection. Immunol. Rev. 227, 75-86

Liscovitch, M., Czarny, M., Fiucci, G. and Tang, X. (2000) Phospholipase D: molecular and cell biology of a novel gene family. Biochem. J. 345, 401-415

Singer, W.D., Brown, H.A. and Sternweis, P.C. (1997) Regulation of eukaryotic phosphatidylinositol-specific phospholipase C and phospholipase D. Annu. Rev. Biochem. 66, 475-509

Futerman, A.H. and Hannun, Y.A. (2004) The complex life of simple sphingolipids. EMBO Rep. 5, 777-782

Hla, T., Lee, M.-J., Ancellin, N., Paik, J.H. and Kluk, M.J. (2001) Lysophospholipids: receptor revelations. Science 294, 1875-1878

Hannun, Y.A. and Obeid, L.M. (2008) Principles of bioactive lipid signalling: lessons from sphingolipids. Nat. Rev. Mol. Cell Biol. 9, 139-150

Le Stunff, H., Peterson, C., Liu, H., Milstien, S. and Spiegel, S. (2002) Sphingosine-1-phosphate and lipid phosphohydrolases. Biochim. Biophys. Acta 1582, 8-17

Spiegel, S., English, D. and Milstien, S. (2002) Sphingosine-1-phosphate signaling: providing cells with a sense of direction. Trends Cell Biol. 12, 236-242

Chen, X., Vinkemeier, U., Zhao, Y., Jeruzalmi, D., Darnell, J.E. and Kuriyan, J. (1998) Crystal structure of a tyrosine phosphorylated STAT-1 dimer bound to DNA. Cell 93, 827-839

Derynck, R. and Zhang, Y.E. (2003) Smad-dependent and Smad-independent pathways in TGF-β family signalling. Nature 425, 577-584

Shi, Y. and Massagué, J. (2003) Mechanisms of TGF-β signalling from cell membrane to the nucleus. Cell 113, 685-700

ten Dijke, P. and Hill, C.S. (2004) New insights into TGF-β-Smad signalling. Trends Biochem. Sci. 29, 265-273

Giles, R.H., van Es, J.H. and Clevers, H. (2003) Caught up in a Wnt storm: Wnt signalling and cancer. Biochim. Biophys. Acta 1653, 1-24

Logan, C.Y. and Nusse, R. (2004) The Wnt signalling pathway in development and disease. Annu. Rev. Cell Dev. Biol. 20, 781-810

Moon, R.T., Kohn, A.D., De Ferrari, G.V. and Kaykas, A. (2004) Wnt and β-catenin signalling: diseases and therapies. Nat. Rev. Genet. 5, 691-701

Nelson, W.J. and Nusse, R. (2004) Convergence of Wnt, β-catenin, and cadherin pathways. Science 303, 1483-1487

Reya, T. and Clevers, H. (2005) Wnt signalling in stem cells and cancer. Nature 434, 843-850

Strutt, D. (2003) Frizzled signalling and cell polarization in Drosophila and vertebrates. Development 130, 4501-4513

Veeman, M.T., Axelrod, J.D. and Moon, R.T. (2003) A second canon: functions and mechanisms of β-catenin-independent Wnt signalling. Dev. Cell 5, 367-377

Daya-Grosjean, L. and Couvé-Privat, S. (2005) Sonic hedgehog signalling in basal cell carcinomas. Cancer Lett. 225, 181-192

Hooper, J.E. and Scott, M.P. (2005) Communicating with hedgehogs. Nat. Rev. Mol. Cell Biol. 6, 306-317

Ingham, P.W. and Kim, H.R. (2005) Hedgehog signalling and the specification of muscle cell identity in the zebrafish embryo. Exp. Cell Res. 306, 336-342

Lum, L. and Beachy, P.A. (2004) The hedgehog response network: sensors, switches, and routers. Science 304, 1755-1759

Stecca, B., Mas, C. and Altaba, A.R. (2005) Interference with HH-GLI signalling inhibits prostate cancer. Trends Mol. Med. 11, 199-203

Bray, S.J. (2006) Notch signalling: a simple pathway becomes complex. Nat. Rev. Mol. Cell Biol. 7, 678-689

Doyle, K.M., Kennedy, D., Gorman, A.M., Gupta, S., Healy, S.J. and Samali, A. (2011) Unfolded proteins and endoplasmic reticulum stress in neurodegenerative disorders. J. Cell. Mol. Med. 15, 2025-2039

Ma, T., Trinh, M.A., Wexler, A.J., Bourbon, C., Gatti, E., Pierre, P., Cavener, D.R. and Klann, E. (2013) Suppression of eIF2α kinases alleviates Alzheimer's disease-related plasticity and memory deficits. Nat. Neurosci. 16, 1299-305

Rutkowski, D.T. and Kaufman, R.J. (2004) A trip to the ER: coping with stress. Trends Cell Biol. 14, 20-28

Zima, A.V., Kockskämper, J., Meijia-Alvarez, R. and Blatter, L.A. (2003) Pyruvate modulates cardiac sarcoplasmic reticulum Ca2+ release in rats via mitochondria-dependent and -independent mechanisms. J. Physiol. 550, 765-783

Anastasiou, D., Poulogiannis, G., Asara, J.M., Boxer, M.B., Jiang, J.K., Shen, M., Bellinger, G., Sasaki, A.T., Locasale, J.W. and Auld, D.S. (2011) Inhibition of pyruvate kinase M2 by reactive oxygen species contributes to cellular antioxidant responses. Science 334, 1278-1283

Luo, W. and Semenza, G.L. (2012) Emerging roles of PKM2 in cell metabolism and cancer progression. Trends Endocrinol. Metab. 23, 560-566

Vander Heiden, M.G., Lunt, S.Y., Dayton, T.L., Fiske, B.P., Israelsen, W.J., Mattaini, K.R., Vokes, N.I., Stephanopoulos, G., Cantley, L.C., Metallo, C.M. and Locasale, J.W. (2011) Metabolic pathway alterations that support cell proliferation. Cold Spring Harb. Symp. Quant. Biol. 76, 325-334

Hardie, D.G. (2005) New roles for the LKB1→AMPK pathway. Curr. Opin. Cell Biol. 17, 167-173

Kahn, B.B., Alquier, T., Carling, D. and Hardie, D.G. (2005) AMP-activated protein kinase: ancient energy gauge provides clues to modern understanding of metabolism. Cell Metab. 1, 15-25

Rutter, G.A., da Silva Xavier, G. and Leclerc, I. (2003) Roles of 5'-AMP-activated protein kinase (AMPK) in mammalian glucose homoeostasis. Biochem. J. 375, 1-16

Brown, M. S. and Goldstein, J.L. (1999) A proteolytic pathway that controls the cholesterol content of membranes, cells, and blood. Proc. Natl. Acad. Sci. U.S.A. 96, 11041-11048


Published online 1 October 2014


©2014 Portland Press Limited