Download Uncoupling insulin signalling by serine/threonine phosphorylation: a

812
Biochemical Society Transactions (2004) Volume 32, part 5
Uncoupling insulin signalling by serine/threonine
phosphorylation: a molecular basis for
insulin resistance
Y. Zick1
Department of Molecular Cell Biology, The Weizmann Institute of Science, Rehovot 76100, Israel
Abstract
Insulin resistance refers to a decreased capacity of circulating insulin to regulate nutrient metabolism. Recent
studies reveal that agents that induce insulin resistance exploit phosphorylation-based negative feedback
control mechanisms otherwise utilized by insulin itself to uncouple the insulin receptor from its downstream
effectors and thereby terminate insulin signal transduction. This article focuses on the Ser/Thr protein kinases
which phosphorylate insulin receptor substrates and the major Ser sites that are phosphorylated, as key
elements in the uncoupling of insulin signalling and the induction of an insulin resistance state.
Insulin signalling
Insulin is the major anabolic hormone whose action is essential for growth, development, homoeostasis of glucose, fat and
protein metabolism, and for the proper function of pancreatic
beta cells [1]. At the molecular level, insulin binding to its
transmembrane receptor (IR) stimulates the intrinsic tyrosine
(Tyr) kinase activity of the receptor (IRK), which then phosphorylates selected Tyr residues of target proteins. IR substrates include IRS (IR substrate) proteins (IRS-1–6), Shc
proteins, p60dok, Cbl, APS and Gab-1 [1–3]. The Tyr-phosphorylated IRSs function as signalling scaffolds, providing
a docking interface for downstream effectors having Srchomology 2 (SH2) domains that further propagate the
metabolic and growth-promoting effects of insulin.
IRS proteins are considered as the major IR substrates.
They contain a conserved PH (pleckstrin homology) domain,
located at their N-terminus, that serves to anchor the IRS
proteins to membrane phosphoinositides, which helps to
localize the IRS proteins in close proximity to the IR [4].
The PH domain is flanked by a PTB (phospho-Tyr binding)
domain that functions as a binding site to the NPXY
motif of the juxta-membrane region of IR, IGF-IR (insulinlike growth factor-1 receptor) and other receptors [5]. The
C-terminal region of IRS proteins contains multiple Tyr
phosphorylation motifs that serve as docking sites for SH2domain-containing proteins, including the p85α regulatory
subunit of PI3K (phosphoinositide 3-kinase), Grb2, Nck,
Crk, Fyn and the SH2-domain-containing tyrosine phosphatase SHP-2, which mediate various aspects of insulin
action [1–3]. IRS proteins contain over 70 potential Ser/Thr
residues with homologies to consensus phosphorylation sites
of different kinases. In non-treated cells, IRS-1 is heavily
phosphorylated on Ser residues and weakly phosphorylated
on Tyr residues, whereas following insulin stimulation, Tyr
and Ser phosphorylation of IRS-1 is increased. Ser phosphorylation of IRS-1 reduces its ability to undergo Tyr
phosphorylation and, in general, serves to turn off insulin
signalling [2] (see below).
Insulin signalling propagates through three major pathways: the PI3K, the MAPK (mitogen-activated protein
kinase) and the Cbl/CAP (Cbl-associated protein) pathways,
each having some unique characteristics. Insulin-mediated activation of the MAPK cascade leads to enhanced cell growth,
while association of PI3K with IRS proteins following insulin
stimulation results in production of phosphatidylinositol
3,4,5-trisphosphate (PIP3 ) that activates PDK1 (PIP3 -dependent kinase 1) and its downstream effectors PKB (protein
kinase B; also named Akt), mTOR (mammalian target of
rapamycin), p70 S6 kinase, as well as the atypical isoforms
of protein kinase C (PKC ζ /λ), all leading to enhanced
glucose transport, glycogen and protein synthesis [6]. Tyr
phosphorylation of Cbl by IRK mediates glucose transport
in a PI3K-independent manner through activation of the
GTP-binding protein TC10 and its interaction with the exocyst complex and the Cdc42-interacting protein 4/2 [7,8].
Collectively, these signalling pathways mediate the metabolic
and growth-promoting functions of insulin.
Inhibition of insulin signalling
Key words: insulin resistance, insulin signalling, serine/threonine phosphorylation.
Abbreviations used: IR, insulin receptor; IRK, IR kinase; IRS, IR substrate; SH2, Src-homology 2;
PH, pleckstrin homology; PTB, phospho-Tyr binding; PI3K, phosphoinositide 3-kinase; MAPK,
mitogen-activated protein kinase; PIP3 , phosphatidylinositol 3,4,5-trisphosphate; PDK1, PIP3 dependent kinase 1; PKB, protein kinase B; PKC, protein kinase C; mTOR, mammalian target of
rapamycin; IGF-I(R), insulin-like growth factor-1 (receptor); IKKβ, I-kappa B kinase β; NF-κB,
nuclear factor-κB; JNK, c-Jun N-terminal kinase; TNFα, tumour necrosis factor α; FFA, free fatty
acid.
1
email [email protected]
C 2004
Biochemical Society
The duration and extent of insulin signalling is regulated
at several levels. Some of the control mechanisms are
set into motion immediately following insulin stimulation.
They act to terminate insulin’s effects through activation of
lipid or protein phosphatases and through the induction of
Ser/Thr kinases that phosphorylate and uncouple various
elements along the insulin signalling pathways. Other
Signalling Outwards and Inwards
Figure 1 The impact of Ser/Thr phosphorylation of IRS on insulin action
Ser/Thr phosphorylation of IRS protein can either enhance or terminate insulin’s signal. Ser residues at the PTB domain of
IRS-1, located within consensus PKB phosphorylation sites, presumably function as positive effectors of insulin signalling
by protecting IRS proteins from the action of PTPs. Other insulin-stimulated Ser/Thr kinases, including mTOR, JNK, PKCζ
and IKKβ, mediate phosphorylation of IRS proteins. Phosphorylation of IRS proteins by these kinases is part of a negativefeedback control mechanism, induced by insulin, that results in the termination of insulin signal. Inducers of insulin resistance
take advantage of this negative-feedback control mechanism by activating either the same or other Ser/Thr kinases that
phosphorylate the IRS proteins, inhibit their function, terminate insulin action and induce insulin resistance. pY, phospho-Tyr;
pS, phospho-Ser.
negative feedback control mechanisms operate on a longer
time scale, and involve a reduction in the cellular content of
the IR, its substrates and other signalling elements.
Ser phosphorylation of IRS proteins:
a physiological means to terminate
insulin signalling
Control over insulin signalling can be achieved by autoregulation, whereby downstream enzymes inhibit upstream
elements (homologous desensitization). Alternatively, signals
from apparently unrelated pathways can inhibit insulin
signalling by heterologous desensitization. The IR, IRS and
Shc proteins are subjected to such complex control. Studies
in our laboratory have focused on Ser/Thr phosphorylation
of IRS proteins as a key feedback control mechanism
that uncouples the IRS proteins from their upstream and
downstream effectors and terminates signal transduction in
response to insulin [2,9,10].
We have shown that insulin-stimulated Ser/Thr phosphorylation of IRS proteins has a dual function in serving
as either a positive or a negative modulator of insulin
signalling (Figure 1). Phosphorylation of Ser residues within
the PTB domain of IRS-1 by PKB, in response to insulin stimulation, protects IRS proteins from the rapid action
of PTPs (protein Tyr phosphatases), and enables the
IRS proteins to maintain their Tyr-phosphorylated active
conformation, thus implicating PKB as a positive regulator
of IRS-1 functions [11]. Indeed, mutation of Ser-302 of
IRS-1, a potential PKB phosphorylation site, reduces insulin/
IGF-I-stimulated DNA synthesis, indicating that IRS-1
phosphorylation at this site promotes mitogenesis and cell
growth [12]. By contrast, Ser/Thr phosphorylation of IRS
proteins by other insulin-stimulated Ser/Thr kinases, such
as PKCζ [10], serves as a physiological negative feedback
control mechanism that inhibits further Tyr phosphorylation
of IRS proteins.
Ser/Thr phosphorylation can: induce the dissociation of
IRS proteins from the IR [9,10]; hinder Tyr phosphorylation
sites of IRS proteins [13]; release the IRS proteins from
intracellular complexes that maintain them in close proximity
to the receptor [14]; induce IRS protein degradation [15];
or turn IRS proteins into inhibitors of the IRK [16]. These
multiple effects place the spotlight on the Ser/Thr kinases that
phosphorylate the IRS protein, and on the Ser/Thr sites
that are getting phosphorylated.
The activity of insulin-stimulated IRS kinases is blocked
by inhibitors of the PI3K pathway, implicating downstream
effectors of PI3K as negative regulators of IRS protein
C 2004
Biochemical Society
813
814
Biochemical Society Transactions (2004) Volume 32, part 5
function [10,11]. One potential candidate is mTOR, which
enhances phosphorylation of Ser residues at the C-terminus
of IRS-1. This phosphorylation inhibits insulin-stimulated
Tyr phosphorylation of IRS-1 and its ability to bind
PI3K [17]. Another candidate is PKCζ . PKCζ mediates
phosphorylation of IRS proteins [10,18], which leads to the
dissociation of the IR–IRS complexes [10,19]. This inhibits
the ability of IRS proteins to undergo further insulinstimulated Tyr phosphorylation, and as a result, it terminates
insulin signalling. IRS-1 serves as a substrate for PKCζ
in vitro, and endogenous IRS-1 forms complexes with
PKCζ in an insulin-dependent manner [18]. These findings
suggest that PKCζ can function as an insulin-stimulated IRS
kinase. Ser-318 is a potential target for PKCζ -mediated
phosphorylation of IRS-1 [20]. Because Ser-318 is adjacent to
the PTB domain of IRS-1, its phosphorylation presumably
disrupts the interaction between the juxtamembrane domain
of the IR and the PTB domain of IRS-1, and in such a way
inhibits insulin-stimulated Tyr phosphorylation of IRS-1.
Another potential IRS kinase is IKKβ (I-kappa B kinase β).
IKKβ is a Ser/Thr kinase that is part of the IKK complex that
phosphorylates the inhibitor of nuclear factor-κB (NF-κB),
IκB. This results in the degradation of IκB, allowing for the
nuclear translocation and activation of the transcription factor
NF-κB [21]. IKKβ is a downstream effector of PKCζ . It can
bind PKCζ both in vitro and in vivo; it serves as an in vitro
substrate for PKCζ , and it is activated by a functional PKCζ
[22]. IKKβ is activated by insulin in Fao rat hepatoma
cells; furthermore, insulin-stimulated Ser phosphorylation of
IRS-1 is inhibited by salicylates, implicating IKKβ as an
insulin-stimulated IRS kinase (Y.F. Liu and Y. Zick, unpublished work). This conclusion is supported by findings indicating that IRS-1 is a direct substrate for IKKβ that phosphorylates IRS-1 on Ser-312 (the human homologue of mouse
Ser-307) following cellular stress [23].
Another IRS kinase is JNK (c-Jun N-terminal kinase), a
member of the MAPK family of protein kinases, which is
activated by insulin by as yet unknown mechanism. Recent
studies have shown that JNK associates with IRS-1 and
phosphorylates it at Ser-307 in an insulin-dependent manner
[24]. Insulin stimulation of JNK activity requires PI3K and
Grb2 signalling, suggesting that JNK activation is mediated
by a number of signalling pathways. Direct binding of JNK to
IRS-1 is not required for JNK activation by insulin; however,
direct interactions between JNK and IRS-1 are required for
Ser-307 phosphorylation [24].
PKB, mTOR, PKCζ and IKKβ are downstream effectors
of PI3K along the insulin signalling pathway. This suggests
that their action should be orchestrated to allow the phosphorylation of IRS proteins by PKB (their positive regulator)
to take place prior to the phosphorylation of IRS proteins by
mTOR, JNK or PKCζ , the negative regulators of IRS protein
function. Of note, the negative-feedback control mechanism
induced by PKCζ (or mTOR) includes a self-attenuation
mode, whereby PI3K-mediated activation of mTOR and
PKCζ inhibits IRS-1 function; thereby inhibiting further
activation of mTOR and PKCζ themselves [2,10].
C 2004
Biochemical Society
Ser phosphorylation of IRS proteins
and insulin resistance
Insulin resistance is a common pathological state in which
target cells fail to respond to ordinary levels of circulating insulin [25]. Individuals with insulin resistance are predisposed to developing Type II diabetes, a 21st century
epidemic. Insulin resistance is frequently associated with
a number of other health disorders, including obesity,
hypertension, chronic infection and cardiovascular diseases,
making it the focus of interest for a large number of studies
[25]. Emerging data suggest that agents such as TNFα
(tumour necrosis factor α), FFAs (free fatty acids) and
cellular stress that inhibit insulin signalling and induce insulin
resistance activate Ser/Thr kinases that phosphorylate the
IRS proteins and inhibit their function. Hence Ser/Thr phosphorylation of IRS proteins that uncouples them from their
upstream and downstream effectors in response to insulin
under physiological conditions could also be the underlying
mode of action of inducers of insulin resistance [2]. This
means that Ser phosphorylation of IRS proteins represents an
important new mechanistic theme that should be considered
alongside other potential mechanisms as the underlying cause
of insulin resistance. Because Ser/Thr phosphorylation of IRS
proteins is stimulated by insulin treatment and by inducers of
insulin resistance, it suggests that the same Ser kinases might
be utilized to phosphorylate the IRS proteins under both
physiological and pathological conditions (Figure 1). This
conclusion is based, for example, upon the fact that TNFα
can induce activation of sphingomyelinase and production of
ceramide, which stimulates PKCζ and its downstream target
IKKβ. Indeed, the inhibitory effects of TNFα on insulinstimulated Tyr phosphorylation of IRS proteins are mimicked
by sphingomyelinase and ceramide analogues [9,26]. TNFα
can also induce complex formation between PKCζ , p62 and
RIP proteins that serve as adaptors of the TNF receptor
and link PKCζ to TNFα signalling [27]. ‘Novel’ PKC isoforms, such as PKCθ, can also act as IRS kinases. These
kinases were implicated as mediators of FFA-induced insulin
resistance in skeletal muscle, in a process that involves
phosphorylation of IRS-1 at Ser-307 [28]. In contrast,
‘conventional’ members of the PKC family, such as PKCα,
which are activated by phorbol esters or endothelin-1, utilize
members of the MAPK pathway to phosphorylate IRS-1
at Ser-612 (located in a consensus MAPK phosphorylation
site) and at additional sites in its C-terminal tail [29,30].
IKKβ is another potential IRS kinase, the activity of which is
stimulated by FFAs, pro-inflammatory cytokines and other
inducers of insulin resistance. Activation or overexpression of
IKKβ attenuates insulin signalling, whereas IKKβ inhibition
by high doses of salicylates or by a reduction in IKKβ gene
dose reverses obesity- and diet-induced insulin resistance
in animal models [31] and improves glucose metabolism in
Type II diabetic subjects [32]. The mechanism by which
FFAs might activate IKKβ presumably involves an increase
in FFA-derived metabolites, such as diacylglycerol and ceramide, which are potent activators of PKCθ and PKCζ , both
Signalling Outwards and Inwards
known to activate IKKβ. At the molecular level, inhibition
of IKKβ prevents Ser/Thr phosphorylation of IRS proteins
induced by a high-fat diet, TNFα or phosphatase inhibitors,
whereas it improves insulin-stimulated Tyr phosphorylation
of IRS proteins. Hence IKKβ might serve as a point of convergence, where Ser kinases downstream of insulin signalling
and Ser kinases activated by pro-inflammatory cytokines
activate IKKβ to inhibit insulin signalling both under
physiological and pathological conditions.
Another IRS kinase, JNK, is also activated by proinflammatory cytokines such as TNFα, and its activity is
abnormally elevated in obesity. Conversely, the absence of
JNK1 results in decreased adiposity, significantly improved
insulin sensitivity and enhanced IR signalling capacity in
two different models of mouse obesity [33]. The inhibitory
effects of JNK on insulin signalling can be attributed, at least
partially, to its ability to phosphorylate Ser-307 of IRS-1 that
uncouples it from the IR [34]. Glycogen synthase kinase-3
(GSK3) [35], casein kinase II [36] and a novel kinase that
phosphorylates IRS-1 at Ser-789 [37] are additional potential
IRS kinases. The Ser-789 kinase is presumably activated under
conditions of insulin resistance, because increased phosphorylation of this site is observed in livers of insulin-resistant
rodent models [37]. Hence a large number of Ser/Thr kinases
can phosphorylate IRS proteins as a result of a stimulus
initiated by inducers of insulin resistance.
In summary, current findings implicate IRS proteins as
major targets for insulin-induced, phosphorylation-based,
negative-feedback control mechanisms that uncouple the IR
from its downstream effectors and terminate insulin signalling under physiological conditions. The kinases involved are
still under investigation, with current focus on PKCζ , IKKβ,
JNK and mTOR as potential candidates. A variety of agents
and conditions that induce insulin resistance, such as TNFα,
FFAs and stress inducers, also activate IRS kinases. At
present, different PKC isoforms, IKKβ, MAPK, JNK and
mTOR appear to be potential candidates, but additional
kinases are likely to emerge. Another question still unresolved
is whether the same or different signalling pathways are
being utilized by insulin or agents that induce resistance to
regulate the activity of IRS kinases. Although the same kinases
(PKCζ , JNK and IKKβ) can be activated by insulin, FFAs
and TNFα, there is no evidence that these kinases are being
activated along the same pathways by insulin or by inducers
of insulin resistance. Because each of the potential IRS kinases
has a unique substrate specificity, the question remains as to
which Ser/Thr sites are being modified by each kinase and
what are the consequences of such phosphorylation. Recent
studies indicate that negative regulatory sites are found
both in close proximity to the PTB domain (e.g. Ser-307)
and at the C-terminal end of IRS-1 (e.g. Ser-612), and it
still remains to be determined whether phosphorylation of
Ser residues distributed along the IRS molecule contributes
to a given functional change (e.g. enhanced dissociation
of IR–IRS complexes), or whether phosphorylation of Ser
residues clustered within a confined region results in a unique
functional consequence.
Given the large number of stimuli, pathways, kinases
and potential sites involved, it appears that Ser/Thr
phosphorylation of IRS proteins represents a combinatorial
consequence of several kinases, activated by different
pathways, acting in concert to phosphorylate multiple sites.
Still, in spite of the complexity and dynamic of the system,
unravelling IRS kinases and their biological role is an exciting
opportunity to provide a novel viewpoint to the molecular
basis for insulin resistance. This should enable rational drug
design to selectively inhibit the activity of relevant kinases
and generate a novel class of therapeutic agents for treatment
of insulin resistance and Type II diabetes.
References
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
LeRoith, D. and Zick, Y. (2001) Diabetes Care 24, 588–597
Zick, Y. (2001) Trends Cell Biol. 11, 437–441
Khan, A.H. and Pessin, J.E. (2002) Diabetologia 45, 1475–1483
Voliovitch, H., Schindler, D., Hadari, Y.R., Taylor, S.I., Accili, D. and
Zick, Y. (1995) J. Biol. Chem. 270, 18083–18087
Wolf, G., Trüb, T., Ottinger, E., Groninga, L., Lynch, A., White, M.F.,
Miyazaki, M., Lee, J. and Shoelson, S.E. (1995) J. Biol. Chem. 270,
27407–27410
Cantley, L.C. (2002) Science 296, 1655–1657
Chiang, S.H., Baumann, C.A., Kanzaki, M., Thurmond, D.C., Watson, R.T.,
Neudauer, C.L., Macara, I.G., Pessin, J.E. and Saltiel, A.R. (2001)
Nature (London) 410, 944–948
Inoue, M., Chang, L., Hwang, J., Chiang, S.H. and Saltiel, A.R. (2003)
Nature (London) 422, 629–633
Paz, K., Hemi, R., LeRoith, R., Karasik, A., Elhanany, E., Kanety, H. and
Zick, Y. (1997) J. Biol. Chem. 272, 29911–29918
Liu, Y.F., Paz, K., Herschkovitz, A., Alt, A., Tennenbaum, T., Sampson, S.R.,
Ohba, M., Kuroki, T., LeRoith, D. and Zick, Y. (2001) J. Biol. Chem. 276,
14459–14465
Paz, K., Yan-Fang, L., Shorer, H., Hemi, R., LeRoith, D., Quon, M.,
Kanety, H., Seger, R. and Zick, Y. (1999) J. Biol. Chem. 274,
28816–28822
Giraud, J., Leshan, R., Lee, Y.H. and White, M.F. (2004) J. Biol. Chem. 279,
3447–3454
Mothe, I. and Van Obberghen, E. (1996) J. Biol. Chem. 271,
11222–11227
Tirosh, A., Potashnik, R., Bashan, N. and Rudich, A. (1999) J. Biol. Chem.
274, 10595–10602
Pederson, T.M., Kramer, D.L. and Rondinone, C.M. (2001) Diabetes 50,
24–31
Hotamisligil, G.S., Peraldi, P., Budavari, A., Ellis, R., White, M.F. and
Spiegelman, B.M. (1996) Science 271, 665–668
Li, J., DeFea, K. and Roth, R.A. (1999) J. Biol. Chem. 274, 9351–9356
Ravichandran, L.V., Esposito, D.L., Chen, J. and Quon, M.J. (2001)
J. Biol. Chem. 276, 3543–3549
Neid, M., Datta, K., Stephan, S., Khanna, I., Pal, S., Shaw, L, White, M. and
Mukhopadhyay, D. (2004) J. Biol. Chem. 279, 3941–3948
Moeschel, K., Beck, A., Weigert, C., Lammers, R., Kalbacher, H.,
Voelter, W., Schleicher, E.D., Haring, H.U. and Lehmann, R. (2004)
J. Biol. Chem. 279, 25157–25163
Karin, M. (1999) J. Biol. Chem. 274, 27339–27342
Lallena, M.-J., Diaz-Meco, M.T., Bren, G., Paya, C.V. and Moscat, J. (1999)
Mol. Cell. Biol. 19, 2180–2188
Gao, Z., Hwang, D., Bataille, F., Lefevre, M., York, D., Quon, M. and Ye, J.
(2002) J. Biol. Chem. 277, 48115–48121
Lee, Y.-H., Giraud, J., Davis, R.J. and White, M.F. (2003) J. Biol. Chem.
278, 2896–2902
Kahn, B.B. and Flier, J.S. (2000) J. Clin. Invest. 106, 473–481
Kanety, H., Hemi, R., Papa, M.Z. and Karasik, A. (1996) J. Biol. Chem.
271, 9895–9897
Sanz, L., Sanchez, P., Lallena, M.J., Diaz, M.M. and Moscat, J. (1999)
EMBO J. 18, 3044–3053
Yu, C., Chen, Y., Zong, H., Wang, Y., Bergeron, R., Kim, J.K., Cline, G.W.,
Cushman, S.W., Cooney, G.J., Atcheson, B. et al. (2002) J. Biol. Chem.
277, 50230–50236
De Fea, K. and Roth, R.A. (1997) J. Biol. Chem. 272, 31400–31406
C 2004
Biochemical Society
815
816
Biochemical Society Transactions (2004) Volume 32, part 5
30 Bouzakri, K., Roques, M., Gual, P., Espinosa, S., Guebre-Egziabher, F.,
Riou, J.P., Laville, M., Le Marchand-Brustel, Y., Tanti, J.F. and Vidal, H.
(2003) Diabetes 52, 1319–1325
31 Yuan, M., Konstantopoulos, N., Lee, J., Hansen, L., Li, Z.-W., Karin, M. and
Shoelson, S.E. (2001) Science 293, 1673–1677
32 Hundal, R.S., Petersen, K.F., Mayerson, A.B., Randhawa, P.S., Inzucchi, S.,
Shoelson, S.E. and Shulman, G.I. (2002) J. Clin. Invest. 109,
1321–1326
33 Hirosumi, J., Tuncman, G., Chang, L., Gorgun, C.Z., Uysal, K.T., Maeda, K.,
Karin, M. and Hotamisligil, G.S. (2002) Nature (London) 420,
333–336
C 2004
Biochemical Society
34 Aguirre, V., Uchida, T., Yenush, L., Davis, R. and White, M.F. (2000)
J. Biol. Chem. 275, 9047–9054
35 Eldar-Finkelman, H. and Krebs, E.G. (1997) Proc. Natl. Acad. Sci. U.S.A.
94, 9660–9664
36 Tanasijevic, M.J., Myers, M.J., Thoma, R.S., Crimmins, D.L., White, M.F.
and Sacks, D.B. (1993) J. Biol. Chem. 268, 18157–18166
37 Qiao, L.Y., Zhande, R., Jetton, T.L., Zhou, G. and Sun, X.J. (2002)
J. Biol. Chem. 277, 26530–26539
Received 15 June 2004