250
Biochemical Society Transactions (2007) Volume 35, part 2
Phosphotyrosine/phosphoserine binary switches:
a new paradigm for the regulation of PI3K
signalling and growth factor pleiotropy?
M.A. Guthridge1 and A.F. Lopez
Cytokine Receptor Laboratory, Department of Human Immunology, Hanson Institute, Institute of Medical and Veterinary Science, Frome Rd, Adelaide,
SA 5000, Australia
Abstract
Cytokines and growth factors exert multiple biological activities through their ability to engage and activate
specific receptors displayed on the surface of cells. How these receptors are able to differentially (and sometimes independently) regulate cell survival, proliferation, differentiation and activation to control quite
specific and distinct cellular outcomes is unclear. Similarly, how a single growth factor or cytokine receptor
can specify alternate cellular responses and control very different cellular fates is also not known. We present
a new mechanism by which cytokines and growth factors are able to control these pleiotropic responses.
Introduction
Cytokine and growth factors control pleiotropic cellular
responses through the binding and activation of specific cellsurface receptors. Ligand binding triggers a series of events
that include receptor dimerization/oligomerization, the activation of tyrosine kinases, tyrosine phosphorylation of the
cytoplasmic tail of the receptor, the binding of specific SH2
domain (Src homology 2 domain) or PTB domain (phosphotyrosine-binding domain) proteins leading to the activation
of downstream signalling cascades and biological responses.
At the heart of cytokine and growth factor pleiotropy lie
a number of fundamental biological responses that include
cell survival, proliferation and differentiation. How cytokines
and growth factors are able to regulate and co-ordinate these
fundamental biological responses leading to pleiotropy is not
known. For example, many of the key components involved
in intracellular signal transduction downstream of cellsurface receptors have not only been identified, but also have
been biochemically and/or genetically ‘positioned’ within
known signalling pathways. However, despite these advances
in our understanding of signalling circuitry, it remains unclear
how specificity in signalling is achieved. Below, we briefly
describe some of the current paradigms by which pleiotropy
is proposed to be regulated (for more detailed reviews, see
[1–7]) and we also present a new mechanism by which at least
some cytokine and growth factor receptors control pleiotropic responses.
Key words: 14-3-3 protein, cytokine, growth factor, phosphoinositide 3-kinase (PI3K),
phosphorylation, pleiotropy.
Abbreviations used: GM-CSF, granulocyte/macrophage colony-stimulating factor; PI3K, phosphoinositide 3-kinase; PTB, phosphotyrosine-binding domain; SH2 domain, Src homology 2
domain; Shc, Src homology and collagen homology.
1
To whom correspondence should be addressed (email [email protected]).
C 2007
Biochemical Society
The specificity versus redundancy
conundrum
The finding that not only are many cell-surface receptors
phosphorylated on multiple tyrosine residues following
ligand activation, but also that SH2 and PTB domains in
different proteins are able to recognize specific phosphotyrosine motifs in the cytoplasmic domains of these receptors
provide a possible molecular explanation for how cytokines
and growth factors may regulate pleiotropic activities [7].
In this scenario, distinct receptor phosphotyrosine residues
couple with specific SH2 or PTB domain proteins to control
specific signalling cascades and biological responses. However, redundancy in terms of both receptor tyrosine phosphorylation and the regulation of downstream signalling
pathways has been observed. For example, one aspect of
pleiotropy that has been difficult to reconcile is that the
‘lock and key’ specificity that has co-evolved at the level of
ligand–receptor interactions appears to operate via a limited
repertoire of intracellular signalling pathways (Figure 1).
For example, there are over 200 type I cytokine and
growth factor receptors or type I transmembrane proteins
thought to exert cytokine/growth factor activity (http://
locate.imb.uq.edu.au). However, each receptor does not possess its own dedicated signalling pathway but rather shares
signalling components and pathways with other receptors.
One hypothesis that may explain pleiotropy and which takes
into account this redundancy is the ‘strength of signalling’
hypothesis whereby cell-surface receptors regulate signalling pathways in temporally different manners to mediate
specific biological responses [1]. In this manner, specificity
in signalling can be achieved through the utilization of redundant signalling pathways. However, while this mechanism
appears to be at work in some cases, its universality and
the molecular basis underpinning their action are much less
clear.
3rd Focused Meeting on PI3K Signalling and Disease
Figure 1 Growth factor and cytokine receptors
Diverse type I growth factor and cytokine receptors displayed on the surface of cells are able to regulate pleiotropic biological
responses through the activation of a limited repertoire of signalling pathways, some of which are indicated. CNTF, ciliary
neurotrophic factor; EGF, epidermal growth factor receptor; EPO-R, erythropoietin receptor; FGF, fibroblast growth factor;
G-CSFR, granulocyte colony-stimulating factor receptor; GH-R, growth hormone receptor; IGF-1, insulin-like growth factor 1;
INS-R, insulin receptor; IL-6, interleukin 6; JAK, Janus kinase; LIF-R, leukaemia inhibitory factor receptor; MAP kinase,
mitogen-activated protein kinase; OSM-R, oncostatin M receptor; PKC, protein kinase C; PLC, phospholipase C; PRL-R, prolactin
receptor; SCF, Skp1/cullin/F-box; STAT, signal transducer and activator of transcription; TPO-R, thrombopoietin receptor; TRK,
tropomyosin receptor kinase; VEGF, vascular endothelial growth factor.
The identification of a
phosphotyrosine/phosphoserine binary
switch in the GM-CSF
(granulocyte/macrophage
colony-stimulating factor) receptor
In addition to the above mechanisms, we would like to introduce a new paradigm by which growth factors and cytokines
are able to control pleiotropic biological responses. We have
shown that cytokine receptors can also be phosphorylated
on serine residues in a site-specific manner in the context of
a 14-3-3-binding site [8]. Importantly, we have now shown
that a 14-3-3-binding site together with an Shc (Src homology
and collagen homology)-binding site in the GM-CSF receptor represents a novel phosphotyrosine/phosphoserine
binary switch that specifies two alternative signals to independently control cell survival and proliferation through the
regulation of PI3K (phosphoinositide 3-kinase) [9]. The binary switch is regulated by cytokine concentrations and toggles
between two mutually exclusive positions: Ser585 is phosphorylated and signals via 14-3-3 in response to lower concentrations of cytokine (fM) to promote cell survival alone,
whereas Tyr577 is phosphorylated and signals via Shc in response to higher concentrations of cytokine (pM) to promote
cell proliferation as well as survival. Such a mechanism allows
the GM-CSF receptor to convert an analogue input (GMCSF concentration) into a binary output (either Ser585 or
Tyr577 phosphorylation) thus permitting the independent
regulation of cell survival and proliferation. Importantly, the
phosphotyrosine/phosphoserine binary switch allows two
different modes of PI3K recruitment and activation: one
that is phosphotyrosine-independent and occurs via Ser585
and the other that is dependent on the specific tyrosine
phosphorylation of the GM-CSF receptor.
Phosphotyrosine/phosphoserine binary
switches and the control of cytokine and
growth factor pleiotropy
We have identified a number of critical features that allow
phosphotyrosine/phosphoserine binary switches, such as
that identified in the GM-CSF receptor, to specify different
biological responses. First, phosphorylation of the serine
residue and the tyrosine residue is mutually exclusive and
enforced by proximity-based steric interference whereby
phosphorylation of one residue blocks the subsequent phosphorylation of the other residue. In addition, recruitment of a
serine/threonine phosphatase to the switch under conditions
that lead to tyrosine phosphorylation would further enforce
unidirectional binary phosphorylation. Secondly, the binary
switch would act as a docking platform for the direct
binding of phosphotyrosine- and phosphoserine/threonine-binding proteins. The close proximity of the serine and
C 2007
Biochemical Society
251
252
Biochemical Society Transactions (2007) Volume 35, part 2
Figure 2 Model representing the phosphoserine/phosphotyrosine binary switch in the GM-CSF receptor
Low concentrations of cytokine (fM) promote Ser585 phosphorylation, 14-3-3 binding and cell survival alone. High
concentrations of cytokine (pM) promote Tyr577 phosphorylation, Shc binding and cell proliferation as well as survival.
In this manner the binary switch allows the conversion of an analogue signal (cytokine concentration) into a digital output
to control two different biological responses.
tyrosine residues ensures that only one phosphotyrosineor phosphoserine/threonine-binding protein can dock with
the platform at one time. We have shown that each of these
features operates in the GM-CSF receptor binary switch [8]
(Figure 2).
A phosphotyrosine/phosphoserine binary switch configured in this manner has implications for how we understand growth factor and cytokine receptors are activated.
One such implication is that receptors may be subject to
two modes of activation: one that would be considered a
classical mode that involves the activation of tyrosine kinases
and receptor tyrosine phosphorylation and a different ‘nonclassical’ mode that leads to receptor serine phosphorylation
and does not involve the activation of tyrosine kinases. In
the case of the GM-CSF receptor, serine versus tyrosine
phosphorylation is determined by cytokine concentration;
however, the precise mechanisms that may lead to these
different modes of receptor activation are not understood.
Furthermore, while we have identified a phosphotyrosine/
phosphoserine binary switch that controls survival and proliferation in the GM-CSF receptor, similar binary switches
may be found in non-receptor signalling molecules and may
also be important for controlling other biological outcomes.
For example, we have identified putative binary switch
motifs in PDK1 (phosphoinositide-dependent kinase 1;
R
GenBank accession number NP_002604), PAK3 (p21
R
activated kinase 3; GenBank accession number O75914) and
R
GSK3α (glycogen synthase kinase 3α; GenBank accession
number NP_063937). Furthermore, while the binary switch
C 2007
Biochemical Society
in the GM-CSF receptor is composed of a PTB domainbinding site followed by a 14-3-3-binding site, variations of
this structure could possibly exist. For example, motifs in
which other possible phosphotyrosine- [e.g. SH2 or CH2
(collagen homology 2)] or phosphoserine/threonine-binding
modules [e.g. Trp-Trp, forkhead-associated, polo-box and
the BRCA1 (breast-cancer susceptibility gene 1) C-terminal
domains] could possibly constitute similar binary switches
[10], or ‘reverse’ motifs in which the phosphoserine precedes
the phosphotyrosine residue could also occur. Thus such
novel signalling devices may serve diverse roles in regulating
different biological outcomes downstream of cytokine and
growth factor receptors and represent a new mechanism by
which pleiotropy can be regulated.
References
1
2
3
4
5
6
7
8
Marshall, C.J. (1995) Cell 80, 179–185
Simon, M.A. (2000) Cell 103, 13–15
Ozaki, K. and Leonard, W.J. (2002) J. Biol. Chem. 277, 29355–29358
De Meyts, P. (2002) Growth Horm. IGF Res. 12, 81–83
Pawson, T. and Saxton, T.M. (1999) Cell 97, 675–678
Pawson, T. and Nash, P. (2000) Genes Dev. 14, 1027–1047
Pawson, T. (2004) Cell 116, 191–203
Guthridge, M.A., Stomski, F.C., Woodcock, J.M., Barry, E.F., Winnall, W.,
Berndt, M.C. and Lopez, A.F. (2000) Mol. Cell 6, 99–108
9 Guthridge, M.A., Powell, J.A., Barry, E.F., Stomski, F.C., McClure, B.J.,
Ramshaw, H., Felquer, F.A., Dottore, M., Thomas, D.T., To, B. et al. (2006)
EMBO J. 25, 479–489
10 Yaffe, M.B. and Elia, A.E.H. (2001) Curr. Opin. Cell Biol. 13, 131–138
Received 26 October 2006