- Email: [email protected]
Intrasteric regulation of MDM2
Update
346
TRENDS in Biochemical Sciences
bisubstrate adduct – S-carboxymethyl-CoA-carnitine – is formed in the active site, indicating the proximity of the CoA sulphur atom to either of the acetyl carbons. What remains to be addressed? A structure with both CoA and carnitine or a transition-state analogue inhibitor in the active site would be useful. Minor conformational adjustment might occur upon substrate binding to CrAT, as suggested by the slight alteration of KM values with different acyl groups bound to the other substrate [7], and by conformational changes measured by circular dichroism [16]. However, the real interest will be how the other members of the family (such as CPT-II, for which the conformational change is large) differ in the active site. Of particular interest is the malonyl-CoA binding-site in carnitine palmitoyltransferase 1 (CPT-I), which has been the subject of much debate [4]. The kinetic data suggest some overlap with the active site. From the structure at the solvent side of the CoA binding-site, the location of residues known to affect malonyl-CoA inhibition in CPT-I and COT led to speculation that binding there obstructed the CoA binding-site [3]. It will be important now to determine the structure of the extra N-terminal residues in CPT-1 that are essential for its malonyl-CoA sensitivity. Also contained in the additional N-terminal residues are two transmembrane helices that confer sensitivity to membrane fluidity [17]. How the active site visualized in the current acetyltransferase alters as a result of the interactions with the N-terminal extension is of great importance for the design of antilipidemic (lipid lowering) and antidiabetic drugs [18], where the aim is to inhibit mitochondrial fatty-acid oxidation but leave the other members of the family active to fulfil their function of buffering the acyl-CoA pools. CrAT itself is important in cell function and it might play a role in the efficacy of clinical therapy with acetyl-L carnitine, which improves neural function in diabetic neuropathy and in Alzheimer’s disease [19]. However, the determination of the structure of CrAT has made advances towards understanding the regulation of the long-chain carnitine acyltransferases possible. Acknowledgements The authors are grateful to Dr L. Tong for providing the originals of fig. 1.
References 1 Chase, J.F.A. et al. (1965) The preparation of crystalline carnitine acetyltransferase. Biochim. Biophys. Acta 96, 162 – 165
Vol.28 No.7 July 2003
2 Jogl, G. and Tong, L. (2003) Crystal structure of carnitine acetyltransferase and implications for the catalytic mechanism and fatty acid transport. Cell 112, 113 – 122 3 Wu, D. et al. (2003) Structure of human carnitine acetyltransferase: molecular basis for fatty acyl transfer. J. Biol. Chem. 278, 13159–13165 4 Ramsay, R.R. et al. (2001) Molecular enzymology of carnitine transfer and transport. Biochim. Biophys. Acta 1546, 21 – 43 5 Zammit, V.A. (1999) Carnitine acyltransferases: functional significance of subcellular distribution and membrane topology. Prog. Lipid Res. 38, 199– 224 6 Brass, E.P. (2002) Pivalate-generating prodrugs and carnitine homeostasis in man. Pharmacol. Rev. 54, 589 – 598 7 Colucci, W.J. and Gandour, R.D. (1988) Carnitine acetyltransferase – a review of its biology, enzymology, and bioorganic chemistry. Bioorg. Chem. 16, 307– 334 8 van der Leij, F.R. et al. (2000) Genomics of the human carnitine acyltransferase genes. Mol. Genet. Metab. 71, 139– 153 9 Saeed, A. et al. (1993) Carnitine acyltransferase enzymatic catalysis requires a positive charge on the carnitine cofactor. Arch. Biochem. Biophys. 305, 307 – 312 10 Lewendon, A. et al. (1990) Evidence for transition-state stabilization by serine-148 in the catalytic mechanism of chloramphenicol acetyltransferase. Biochemistry 29, 2075– 2080 11 Cronin, C.N. (1997) The conserved serine-threonine-serine motif of the carnitine acyltransferases is involved in carnitine binding and transition-state stabilization: a site-directed mutagenesis study. Biochem. Biophys. Res. Commun. 238, 784 – 789 12 Nicolas, A. et al. (1996) Contribution of cutinase serine 42 side chain to the stabilization of the oxyanion transition state. Biochemistry 35, 398– 410 13 Saeed, A. et al. (1994) 3-amino-5,5-dimethylhexanoic acid – synthesis, resolution, and effects on carnitine acyltransferases. J. Med. Chem. 37, 3247– 3251 14 Sulzenbacher, G. et al. (2001) Crystal structure of Streptococcus pneumoniae N-acetylglucosamine-1-phosphate uridyltransferase bound to acetyl-coenzyme A reveals a novel active site architecture. J. Biol. Chem. 276, 11844 – 11851 15 Chase, J.F.A. and Tubbs, P.K. (1969) Conditions for the self-catalysed inactivation of carnitine acetyltransferase. Biochem. J. 111, 225– 235 16 Yan, B. et al. (1995) Effects of substrate-binding and pH on the secondary structure of carnitine acetyltransferase. Biochim. Biophys. Acta 1253, 175 – 180 17 Jackson, V.N. et al. (2000) Use of six chimeric proteins to investigate the role of intramolecular interactions in determining the kinetics of carnitine palmitoyltransferase I isoforms. J. Biol. Chem. 275, 19560 – 19566 18 Anderson, R.C. (1998) Carnitine palmitoyltransferase: a viable target for the treatment of NIDDM? Curr. Pharm. Des. 4, 1 – 16 19 Pettegrew, J.W. et al. (2000) Acetyl-L -carnitine physical-chemical, metabolic, and therapeutic properties: relevance for its mode of action in Alzheimer’s disease and geriatric depression. Mol. Psychiatry 5, 616– 632 0968-0004/03/$ - see front matter q 2003 Elsevier Science Ltd. All rights reserved. doi:10.1016/S0968-0004(03)00137-3
Intrasteric regulation of MDM2 Harumi Shimizu and Ted R. Hupp Cancer Research UK laboratories, Department of Molecular and Cellular Pathology, University of Dundee, Dundee, Scotland, UK DD1 9SY
The oncoprotein murine double minute 2 (MDM2) binds to the tumour suppressor protein p53 and negatively regulates growth suppression by catalyzing the ubiquitination and degradation of p53. Antagonizing MDM2 is Corresponding author: Ted R. Hupp ([email protected]). http://tibs.trends.com
an ataxia telangiectasia mutated (ATM) kinase- and checkpoint kinase-2 (CHK2)-dependent DNA-damageinducible pathway that phosphorylates p53, releases p53 from MDM2 control and stabilizes the p300–p53 transcription complex. A recent report has identified a pseudo-substrate domain in MDM2 that can act as a lid
Update
TRENDS in Biochemical Sciences
to cover the hydrophobic MDM2-pocket and suggests a novel mechanism for control of the stability of the MDM2 – p53 protein complex in vivo. The p53 gene product is a tumour suppressor that functions at a key point in the response of cells to environmental change. p53-protein activation induces its transcription-factor activity that, in turn, induces mediators to carry out tumour-suppressor functions, including apoptosis or growth arrest and repair [1]. Control of the tumour-suppressor activity of p53 is posttranslational. Negative regulation of p53 activity is driven by the oncogene murine double minute 2 (MDM2), which is a component of a negative-feedback loop [2] that promotes the ubiquitination and degradation of p53 protein [3]. Positive regulation of p53-dependent tumour suppression is the transcriptional co-activator protein p300, which is required for p53 protein acetylation [4] and stabilization [5] in response to DNA damage. Intriguingly, p300 also promotes the poly-ubiquitination of mono-ubiquitinated p53 [6], identifying a paradoxically negative role for p300 in the p53 pathway. Therefore, p300 can promote p53 acetylation and transactivation or stimulate poly-ubiquitination and degradation of p53, which places p300 at the centre of the decision fork to switch p53 on or off. The balance between p300-mediated co-activation of p53 and MDM2-mediated degradation of p53 controls the transcriptional and degradation flux of p53, thus tipping the scale towards either tumour suppression or cancer progression. A recent study has identified a pseudosubstrate, negative regulatory domain in the oncogene MDM2 that has further implications for control of the p53 – MDM2 protein complex [7]. How P53 phosphorylation modulates its affinity for p300 and MDM2 Both p300 and MDM2 bind to an N-terminal motif in the transactivation domain of p53, but they both have distinct amino-acid contacts within p53 (Fig. 1). There are also three different protein-kinase sites (Ser15, Thr18 and Ser20) that overlap the p300 and MDM2-binding sites, suggesting a secondary mechanism for regulating protein– protein interactions. Phosphorylation of this BOX-I ATM CHK2
Human p53 (aa10–26) VEPPLSQETFSDLWKLL Fish p53 MEPPLSQETFEDLWSLL MDM2 Lid(aa12–25) GAVTTSQIPASEQE MDM2 contact p300 contact Ti BS
Fig. 1. The activation domain of p53 is targeted by multiple enzymes. The conservation of the transactivation domain between human p53 and fish p53 is indicated. The ataxia telangiectasia mutated (ATM) kinase or DNA-dependent protein kinase (DNA-PK) phosphorylation at Ser15 [8] and the checkpoint kinase-2 (CHK2) or casein kinase 1 site at Thr18 [9,13] are highly conserved between species, whereas the CHK2 site at Ser20 [8] is the least conserved. The key murine double minute 2 (MDM2) contact sites on human p53 are underlined and the p300 contact sites are in red [16]. The homology between the MDM2 lid and the MDM2-binding domain on p53 is highlighted by the green box. http://tibs.trends.com
Vol.28 No.7 July 2003
347
transactivation domain of p53 by tumour-suppressing protein kinases such as ataxia telangiectasia mutated (ATM) kinase or checkpoint kinase-2 (CHK2) plays a role, in part, in activating the p53 pathway [8,9]. The phosphatidylinositol (PtdIns)-3 kinase homologues such as ATM or DNA-dependent kinase (DNA-PK) can modify p53 at Ser15 [8,10]. The most likely kinases that modify p53 at Ser20 and Thr18 are CHK1 and/or CHK2 [9]. However, other kinases can modify the Ser20 site of p53 in undamaged cells [11] or in X-irradiated tumour cells [12]. Furthermore, casein kinase 1 can modify Thr18 in damaged cells [13]. Although original results suggested that ATM or DNA-PK phosphorylation of p53 at Ser15 disrupted the p53– MDM2 complex [10], more recent NMR studies have shown no effect of Ser15 phosphorylation of p53 on MDM2 binding [14]. CHK2 phosphorylation at Ser20 also does not effect MDM2 – p53-protein complex stability [14], but Ser20 phosphorylation stabilizes significantly the p300– p53 complex [15]. Ser20 phosphorylation stabilizes p300 binding to p53 via the interferon-3 binding domain (IBiD) and IBiD homology domain (IHD) phospho-peptide binding domains of p300 and promotes sequence-specific DNA-dependent acetylation of p53 by p300 [16] (Fig. 2). In contrast to Ser20 phosphorylation, phosphorylation at Thr18 by CHK2 or casein kinase 1 has two notable effects: completely destabilizing the p53 – MDM2 complex [14,17] and stabilizing the p300 – p53 complex [15]. The control of MDM2 by phosphorylation Dynamic conformational changes are induced in MDM2 upon ligand binding P53 induces the oncogene MDM2, creating a negativefeedback loop that maintains the steady-state levels of p53 [2]. However, this control is not a strict genetic negativefeedback loop because dosage compensation does not regulate p53 activity: p53-heterozygotes have higher cancer incidence than diploid wild-type p53 animals [18] and p53-knock-ins have reduced tumour incidence compared with diploid wild-type p53 [19]. Thus, the MDM2 – p53 feedback loop is not a closed system, and it is exposed to alternate controls. Defining the factors that control the conformational flexibility of MDM2 could reveal these other regulatory pathways. The biochemical effects of MDM2-binding to p53 result in multiple sites of mono-ubiquitination of p53 [3,20]. Although the mechanistic details of the ubiquitination and degradation reaction are lacking, important regulatory stages are now emerging from in vitro studies. For example, after p53 mono-ubiquitination by MDM2, p300 can catalyze poly-ubiquitination of mono-ubiquitinated p53 and promote p53-protein degradation [6]. The initial binding of MDM2 to p53 requires an N-terminal domain in MDM2 (, 110 amino acids) containing a deep hydrophobic pocket that interacts with the BOX-I activation domain of p53 [21] (Fig. 1). The structure of this MDM2 – BOX-I polypeptide complex has been analyzed, and data acquired recently using NMR has demonstrated that the type of peptide fragment bound to the MDM2 pocket can induce significantly different conformational changes in MDM2 [14]. This conformational change in MDM2 upon ligand
Update
348
LID
TRENDS in Biochemical Sciences
MDM2
Vol.28 No.7 July 2003
p53
(b)
(c)
p300 p300
p53 AC
(a)
(d)
MDM2 LID P Ti BS
Fig. 2. Concerted recruitment of protein kinases modifies the murine double minute 2 (MDM2)– p53 protein complex. Protein-kinase activation of the p53 pathway can now be divided into different biochemical steps involving: (a) phosphorylation (yellow) of the MDM2 lid (red) at Ser17 by a phosphatidylinositol (PtdIns)-3-like kinase-dependent pathway [23], which stabilizes the MDM2 pseudo-substrate lid over the MDM2 pocket [7]; (b) the release of p53 protein from MDM2 control; (c) phosphorylation of p53 at Thr18 and/or Ser20 by enzymes such as checkpoint kinase-2 (CHK2) [8] or other Ser20 kinases [11,12] that stabilize the docking of p300 onto the p53 tetramer through the interferon-3 binding domain (IBiD) and IBiD homology domain (IHD) of p300 [16] (white arrows); and (d) subsequent DNA-dependent acetylation (red arrow) of the p53 tetramer [16]. Permutations in the ataxia telangiectasia mutated (ATM) and/or CHK2 kinase pathways could influence the magnitude of the p53 response. The phosphorylation of p53 at Ser15 or at Ser20 has little effect on MDM2 binding [14], whereas phosphorylation at Ser20 stabilizes p300 docking and promotes DNA-dependent acetylation of p53 [16]. By contrast, phosphorylation at Thr18 can disrupt the MDM2– p53 complex [14,17] and simultaneously stabilize the p300–p53 complex [15] leading to p53 activation without MDM2-lid phosphorylation. Furthermore, MDM2-lid phosphorylation by PtdIns-3 kinases such as ATM and DNA-PK without p53 phosphorylation at Thr18 and Ser20 could also stabilize p53 protein. Therefore, the concerted action of enzymes such as ATM and CHK2 targeting both MDM2 and p53, respectively, gives maximal activation of the p53 response. Although p300 is required for p53 acetylation, co-activator recruitment and associated transactivation [28,29], p300 can also work with MDM2 to catalyze p53 poly-ubiquitination and degradation [6]. Presumably, this stage of p300 function occurs when the p53 response needs to be turned off.
binding is probably transmitted to other domains on the full-length MDM2 protein and might affect other activities of MDM2, including binding to p14ARF, p300, Zinc and RNA. In fact, RNA binding induces a conformational change in full-length MDM2 protein and can switch the substrate affinity of MDM2 from the original BOX-I activation domain to a flexible loop in the core domain of p53 [22]. Moreover, zinc binding by full-length MDM2 can destabilize MDM2 – p53-BOX-I-peptide complexes [22], suggesting a mechanism for allosteric regulation of the MDM2 – p53 complex. This intrinsic conformational flexibility of the MDM2 protein can be used to predict that it will be the focus of multiple post-translational control pathways. A novel control of the MDM2 – p53 protein complex: a flexible lid fits over the MDM2 pocket A recent report has highlighted a novel biological regulation of the MDM2 –p53 protein complex involving a flexible lid that can close over the hydrophobic pocket of MDM2 [7]. Previous work has identified the minimal domain in MDM2 that can bind to the p53 activation domain and the MDM2 hydrophobic pocket is lined with over ten aromatic and hydrophobic amino acids that make van der Waals contacts to p53. This hydrophobic pocket is maintained, in part, by b sheets that can form a scaffold for the ligand and, despite original claims that this structure is rigid [21], is very flexible [14]. This flexibility is the key to understanding the dynamic nature of MDM2 protein regulation. By focusing on the detailed NMR structure of MDM2 containing additional amino acids in the N terminus of the protein, assignments can be placed for a peptide fragment that fits over the hydrophobic pocket of MDM2 [7]. This peptide lid has a significant degree of homology to the p53-activation domain fragment (Fig. 1), suggesting that it might be able to compete in cis for binding with p53 and act like a pseudo-substrate domain. Furthermore, because the MDM2-lid also has a classic DNA-PK or ATM consensus site (SQ motif) whose http://tibs.trends.com
phosphorylation disrupts the p53– MDM2 complex [23], modification of the affinity of the lid for the MDM2 pocket is possible. In fact, a phospho-mimetic substitution at Ser17 (Fig. 1) in the MDM2-lid stabilizes its interaction with MDM2 and alters the conformation of MDM2 [7]. This is consistent with independent NMR studies that show substantial conformational shifts in MDM2 upon ligand binding [14]. These data provide an additional step for a PtdIns-3 kinase such as ATM to disrupt the MDM2 – p53 complex (Fig. 2). In summary, the conformational flexibility of MDM2 protein, the presence of a pseudosubstrate lid that might control MDM2-binding proteins, the existence of multiple ligand-binding domains and the contribution of RNA binding to ubiquitination activity all highlight the complex intradomain communication that exists within MDM2. Further research aimed at integrating p53-specific kinase signalling pathways to intradomain interactions of MDM2 will enhance details on how the p53 pathway is controlled post-translationally, and assist in developing assays to identify modifying loci of the p53 –MDM2 negative-feedback loop. Concluding remarks: intrasteric regulation in enzyme control Protein function in cells can be regulated in many ways including ligand binding, covalent modification by acetylation or phosphorylation and proteolytic processing. In addition, many regulatory proteins have intrinsic conformational control mechanisms as an additional layer of post-translational regulation. Such built-in conformational restraints are often called allosteric, as they simply describe the presence of cis-acting, conformationally flexible regulatory motifs that are outside the active site of the protein. A subclass of the allosteric model of enzyme regulation is the more recently highlighted intrasteric regulation by a pseudo-substrate domain [24,25]. This concept was highlighted as an auto-regulatory control mechanism in protein kinases in which an internal sequence of the enzyme has high homology to the substrate
Update
TRENDS in Biochemical Sciences
of the kinase, and disruption of this pseudo-substrate domain is an important component in kinase activation [26]. This mechanism is not confined to protein kinases [27]. The identification of a pseudo-substrate domain in the oncogene MDM2, whose affinity for MDM2 can be modified by phosphorylation, expands on the class of proteins that can be regulated by intrasteric mechanisms and has interesting implications for control of other MDM2-binding proteins like p53, p300, E2F, p73 and Rb. References 1 Vogelstein, B. et al. (2000) Surfing the p53 network. Nature 408, 307 – 310 2 Freedman, A. et al. (1999) Functions of the MDM2 oncoprotein. Cell. Mol. Life Sci. 55, 96 – 107 3 Honda, R. et al. (1997) Oncoprotein MDM2 is a ubiquitin ligase E3 for tumor suppressor p53. FEBS Lett. 420, 25 – 27 4 Sakaguchi, K. et al. (1998) DNA damage activates p53 through a phosphorylation – acetylation cascade. Gene Dev. 12, 2831– 2841 5 Yuan, Z.M. et al. (1999) Role for p300 in stabilization of p53 in the response to DNA damage. J. Biol. Chem. 274, 1883 – 1886 6 Grossman, S.R. et al. (2003) Polyubiquitination of p53 by a ubiquitin ligase activity of p300. Science 300, 342– 344 7 McCoy, M.A. et al. (2003) Flexible lid to the p53-binding domain of human Mdm2: Implications for p53 regulation. Proc. Natl. Acad. Sci. U. S. A. 100, 1645– 1648 8 Banin, S. et al. (1998) Enhanced phosphorylation of p53 by ATM in response to DNA damage. Science 281, 1674 – 1677 9 Shieh, S.Y. et al. (2000) The human homologs of checkpoint kinases Chk1 and Cds1 (Chk2) phosphorylate p53 at multiple DNA damageinducible sites. Genes Dev. 14, 289 – 300 10 Shieh, S.Y. et al. (1997) DNA damage-induced phosphorylation of p53 alleviates inhibition by MDM2. Cell 91, 325 – 334 11 Craig, A.L. et al. (1999) Dephosphorylation of p53 at Ser20 after cellular exposure to low levels of non-ionizing radiation. Oncogene 18, 6305 – 6312 12 Ahn, A. et al. Questioning the role of Chk2 in the p53 DNA damage response. J. Biol. Chem. (in press) 13 Sakaguchi, K. et al. (2000) Damage-mediated phosphorylation of human p53 threonine 18 through a cascade mediated by a casein 1-like kinase. Effect on Mdm2 binding. J. Biol. Chem. 275, 9278 – 9283 14 Schon, O. et al. (2002) Molecular mechanism of the interaction between MDM2 and p53. J. Mol. Biol. 323, 491 – 501
Vol.28 No.7 July 2003
349
15 Dornan, D. and Hupp, T.R. (2001) Inhibition of p53-dependent transcription by BOX-I phospho-peptide mimetics that bind to p300. EMBO Rep. 2, 139 – 144 16 Dornan, D. et al. (2003) DNA-dependent acetylation of p53 by the transcription coactivator p300. J. Biol. Chem. 278, 13431 – 13441 17 Craig, A.L. et al. (1999) Novel phosphorylation sites of human tumour suppressor protein p53 at Ser20 and Thr18 that disrupt the binding of mdm2 protein are modified in human cancers. Biochem. J. 342, 133– 141 18 Clarke, A.R. et al. (1993) Thymocyte apoptosis induced by p53dependent and independent pathways. Nature 362, 849– 852 19 Garcia-Cao, I. et al. (2002) Super p53 mice exhibit enhanced DNA damage response, are tumor resistant and age normally. EMBO J. 21, 6225– 6235 20 Lai, Z. et al. (2001) Human mdm2 mediates multiple monoubiquitination of p53 by a mechanism requiring enzyme isomerization. J. Biol. Chem. 276, 31357 – 31367 21 Kussie, P.H. et al. (1996) Structure of the MDM2 oncoprotein bound to the p53 tumor suppressor transactivation domain. Science 274, 948– 953 22 Shimizu, H. et al. (2002) The conformationally flexible S9 – S10 linker region in the core domain of p53 contains a novel MDM2 binding site whose mutation increases ubiquitination of p53 in vivo. J. Biol. Chem. 277, 28446 – 28458 23 Mayo, L.D. et al. (1997) Mdm-2 phosphorylation by DNA-dependent protein kinase prevents interaction with p53. Cancer Res. 57, 5013– 5016 24 Kobe, B. and Kemp, B.E. (1999) Active site-directed protein regulation. Nature 402, 373 – 376 25 Kobe, B. et al. (1997) Intrasteric regulation of protein kinases. Adv. Second Messenger Phosphoprotein Res. 31, 29 – 40 26 Kemp, B.E. et al. (1996) Intrasteric regulation of calmodulindependent protein kinases. Adv. Pharmacol. 36, 221 – 249 27 Kobe, B. et al. (1999) Structural basis of autoregulation of phenylalanine hydroxylase. Nat. Struct. Biol. 6, 442 – 448 28 Espinosa, J.M. and Emerson, B.M. (2001) Transcriptional regulation by p53 through intrinsic DNA/chromatin binding and site-directed cofactor recruitment. Mol. Cell 8, 57 – 69 29 Barlev, N.A. et al. (2001) Acetylation of p53 activates transcription through recruitment of coactivators/histone acetyltransferases. Mol. Cell 8, 1243– 1254
0968-0004/03/$ - see front matter q 2003 Elsevier Ltd. All rights reserved. doi:10.1016/S0968-0004(03)00140-3
| Protein Sequence Motif
The KIND module: a putative signalling domain evolved from the C lobe of the protein kinase fold Francesca D. Ciccarelli1,2, Peer Bork1,2 and Eugen Kerkhoff3 1
European Molecular Biology Laboratory, Meyerhofstr. 1, D-69012 Heidelberg, Germany Max-Delbrueck-Centrum, PO Box 740238, D-13092 Berlin, Germany 3 Institut fu¨r Medizinische Strahlenkunde und Zellforschung (MSZ), Universita¨t Wu¨rzburg, Versbacher Str. 5, D-97078 Wu¨rzburg, Germany 2
A novel putative interaction domain – KIND (kinase non-catalytic C-lobe domain) – has been identified as being similar to the C-terminal protein kinase catalytic fold (C lobe). Its presence at the N terminus of signalling proteins and the absence of the active-site residues in the catalytic and activation loops suggest that it folds Corresponding author: Francesca D. Ciccarelli ([email protected]). http://tibs.trends.com
independently and is likely to be non-catalytic. The occurrence of the novel domain only in metazoa implies that it has evolved from the catalytic protein kinase domain into an interaction domain possibly by keeping the substrate-binding features. The Drosophila melanogaster protein p150-Spir (CAB62901), which is phosphorylated by the c-Jun
346
TRENDS in Biochemical Sciences
bisubstrate adduct – S-carboxymethyl-CoA-carnitine – is formed in the active site, indicating the proximity of the CoA sulphur atom to either of the acetyl carbons. What remains to be addressed? A structure with both CoA and carnitine or a transition-state analogue inhibitor in the active site would be useful. Minor conformational adjustment might occur upon substrate binding to CrAT, as suggested by the slight alteration of KM values with different acyl groups bound to the other substrate [7], and by conformational changes measured by circular dichroism [16]. However, the real interest will be how the other members of the family (such as CPT-II, for which the conformational change is large) differ in the active site. Of particular interest is the malonyl-CoA binding-site in carnitine palmitoyltransferase 1 (CPT-I), which has been the subject of much debate [4]. The kinetic data suggest some overlap with the active site. From the structure at the solvent side of the CoA binding-site, the location of residues known to affect malonyl-CoA inhibition in CPT-I and COT led to speculation that binding there obstructed the CoA binding-site [3]. It will be important now to determine the structure of the extra N-terminal residues in CPT-1 that are essential for its malonyl-CoA sensitivity. Also contained in the additional N-terminal residues are two transmembrane helices that confer sensitivity to membrane fluidity [17]. How the active site visualized in the current acetyltransferase alters as a result of the interactions with the N-terminal extension is of great importance for the design of antilipidemic (lipid lowering) and antidiabetic drugs [18], where the aim is to inhibit mitochondrial fatty-acid oxidation but leave the other members of the family active to fulfil their function of buffering the acyl-CoA pools. CrAT itself is important in cell function and it might play a role in the efficacy of clinical therapy with acetyl-L carnitine, which improves neural function in diabetic neuropathy and in Alzheimer’s disease [19]. However, the determination of the structure of CrAT has made advances towards understanding the regulation of the long-chain carnitine acyltransferases possible. Acknowledgements The authors are grateful to Dr L. Tong for providing the originals of fig. 1.
References 1 Chase, J.F.A. et al. (1965) The preparation of crystalline carnitine acetyltransferase. Biochim. Biophys. Acta 96, 162 – 165
Vol.28 No.7 July 2003
2 Jogl, G. and Tong, L. (2003) Crystal structure of carnitine acetyltransferase and implications for the catalytic mechanism and fatty acid transport. Cell 112, 113 – 122 3 Wu, D. et al. (2003) Structure of human carnitine acetyltransferase: molecular basis for fatty acyl transfer. J. Biol. Chem. 278, 13159–13165 4 Ramsay, R.R. et al. (2001) Molecular enzymology of carnitine transfer and transport. Biochim. Biophys. Acta 1546, 21 – 43 5 Zammit, V.A. (1999) Carnitine acyltransferases: functional significance of subcellular distribution and membrane topology. Prog. Lipid Res. 38, 199– 224 6 Brass, E.P. (2002) Pivalate-generating prodrugs and carnitine homeostasis in man. Pharmacol. Rev. 54, 589 – 598 7 Colucci, W.J. and Gandour, R.D. (1988) Carnitine acetyltransferase – a review of its biology, enzymology, and bioorganic chemistry. Bioorg. Chem. 16, 307– 334 8 van der Leij, F.R. et al. (2000) Genomics of the human carnitine acyltransferase genes. Mol. Genet. Metab. 71, 139– 153 9 Saeed, A. et al. (1993) Carnitine acyltransferase enzymatic catalysis requires a positive charge on the carnitine cofactor. Arch. Biochem. Biophys. 305, 307 – 312 10 Lewendon, A. et al. (1990) Evidence for transition-state stabilization by serine-148 in the catalytic mechanism of chloramphenicol acetyltransferase. Biochemistry 29, 2075– 2080 11 Cronin, C.N. (1997) The conserved serine-threonine-serine motif of the carnitine acyltransferases is involved in carnitine binding and transition-state stabilization: a site-directed mutagenesis study. Biochem. Biophys. Res. Commun. 238, 784 – 789 12 Nicolas, A. et al. (1996) Contribution of cutinase serine 42 side chain to the stabilization of the oxyanion transition state. Biochemistry 35, 398– 410 13 Saeed, A. et al. (1994) 3-amino-5,5-dimethylhexanoic acid – synthesis, resolution, and effects on carnitine acyltransferases. J. Med. Chem. 37, 3247– 3251 14 Sulzenbacher, G. et al. (2001) Crystal structure of Streptococcus pneumoniae N-acetylglucosamine-1-phosphate uridyltransferase bound to acetyl-coenzyme A reveals a novel active site architecture. J. Biol. Chem. 276, 11844 – 11851 15 Chase, J.F.A. and Tubbs, P.K. (1969) Conditions for the self-catalysed inactivation of carnitine acetyltransferase. Biochem. J. 111, 225– 235 16 Yan, B. et al. (1995) Effects of substrate-binding and pH on the secondary structure of carnitine acetyltransferase. Biochim. Biophys. Acta 1253, 175 – 180 17 Jackson, V.N. et al. (2000) Use of six chimeric proteins to investigate the role of intramolecular interactions in determining the kinetics of carnitine palmitoyltransferase I isoforms. J. Biol. Chem. 275, 19560 – 19566 18 Anderson, R.C. (1998) Carnitine palmitoyltransferase: a viable target for the treatment of NIDDM? Curr. Pharm. Des. 4, 1 – 16 19 Pettegrew, J.W. et al. (2000) Acetyl-L -carnitine physical-chemical, metabolic, and therapeutic properties: relevance for its mode of action in Alzheimer’s disease and geriatric depression. Mol. Psychiatry 5, 616– 632 0968-0004/03/$ - see front matter q 2003 Elsevier Science Ltd. All rights reserved. doi:10.1016/S0968-0004(03)00137-3
Intrasteric regulation of MDM2 Harumi Shimizu and Ted R. Hupp Cancer Research UK laboratories, Department of Molecular and Cellular Pathology, University of Dundee, Dundee, Scotland, UK DD1 9SY
The oncoprotein murine double minute 2 (MDM2) binds to the tumour suppressor protein p53 and negatively regulates growth suppression by catalyzing the ubiquitination and degradation of p53. Antagonizing MDM2 is Corresponding author: Ted R. Hupp ([email protected]). http://tibs.trends.com
an ataxia telangiectasia mutated (ATM) kinase- and checkpoint kinase-2 (CHK2)-dependent DNA-damageinducible pathway that phosphorylates p53, releases p53 from MDM2 control and stabilizes the p300–p53 transcription complex. A recent report has identified a pseudo-substrate domain in MDM2 that can act as a lid
Update
TRENDS in Biochemical Sciences
to cover the hydrophobic MDM2-pocket and suggests a novel mechanism for control of the stability of the MDM2 – p53 protein complex in vivo. The p53 gene product is a tumour suppressor that functions at a key point in the response of cells to environmental change. p53-protein activation induces its transcription-factor activity that, in turn, induces mediators to carry out tumour-suppressor functions, including apoptosis or growth arrest and repair [1]. Control of the tumour-suppressor activity of p53 is posttranslational. Negative regulation of p53 activity is driven by the oncogene murine double minute 2 (MDM2), which is a component of a negative-feedback loop [2] that promotes the ubiquitination and degradation of p53 protein [3]. Positive regulation of p53-dependent tumour suppression is the transcriptional co-activator protein p300, which is required for p53 protein acetylation [4] and stabilization [5] in response to DNA damage. Intriguingly, p300 also promotes the poly-ubiquitination of mono-ubiquitinated p53 [6], identifying a paradoxically negative role for p300 in the p53 pathway. Therefore, p300 can promote p53 acetylation and transactivation or stimulate poly-ubiquitination and degradation of p53, which places p300 at the centre of the decision fork to switch p53 on or off. The balance between p300-mediated co-activation of p53 and MDM2-mediated degradation of p53 controls the transcriptional and degradation flux of p53, thus tipping the scale towards either tumour suppression or cancer progression. A recent study has identified a pseudosubstrate, negative regulatory domain in the oncogene MDM2 that has further implications for control of the p53 – MDM2 protein complex [7]. How P53 phosphorylation modulates its affinity for p300 and MDM2 Both p300 and MDM2 bind to an N-terminal motif in the transactivation domain of p53, but they both have distinct amino-acid contacts within p53 (Fig. 1). There are also three different protein-kinase sites (Ser15, Thr18 and Ser20) that overlap the p300 and MDM2-binding sites, suggesting a secondary mechanism for regulating protein– protein interactions. Phosphorylation of this BOX-I ATM CHK2
Human p53 (aa10–26) VEPPLSQETFSDLWKLL Fish p53 MEPPLSQETFEDLWSLL MDM2 Lid(aa12–25) GAVTTSQIPASEQE MDM2 contact p300 contact Ti BS
Fig. 1. The activation domain of p53 is targeted by multiple enzymes. The conservation of the transactivation domain between human p53 and fish p53 is indicated. The ataxia telangiectasia mutated (ATM) kinase or DNA-dependent protein kinase (DNA-PK) phosphorylation at Ser15 [8] and the checkpoint kinase-2 (CHK2) or casein kinase 1 site at Thr18 [9,13] are highly conserved between species, whereas the CHK2 site at Ser20 [8] is the least conserved. The key murine double minute 2 (MDM2) contact sites on human p53 are underlined and the p300 contact sites are in red [16]. The homology between the MDM2 lid and the MDM2-binding domain on p53 is highlighted by the green box. http://tibs.trends.com
Vol.28 No.7 July 2003
347
transactivation domain of p53 by tumour-suppressing protein kinases such as ataxia telangiectasia mutated (ATM) kinase or checkpoint kinase-2 (CHK2) plays a role, in part, in activating the p53 pathway [8,9]. The phosphatidylinositol (PtdIns)-3 kinase homologues such as ATM or DNA-dependent kinase (DNA-PK) can modify p53 at Ser15 [8,10]. The most likely kinases that modify p53 at Ser20 and Thr18 are CHK1 and/or CHK2 [9]. However, other kinases can modify the Ser20 site of p53 in undamaged cells [11] or in X-irradiated tumour cells [12]. Furthermore, casein kinase 1 can modify Thr18 in damaged cells [13]. Although original results suggested that ATM or DNA-PK phosphorylation of p53 at Ser15 disrupted the p53– MDM2 complex [10], more recent NMR studies have shown no effect of Ser15 phosphorylation of p53 on MDM2 binding [14]. CHK2 phosphorylation at Ser20 also does not effect MDM2 – p53-protein complex stability [14], but Ser20 phosphorylation stabilizes significantly the p300– p53 complex [15]. Ser20 phosphorylation stabilizes p300 binding to p53 via the interferon-3 binding domain (IBiD) and IBiD homology domain (IHD) phospho-peptide binding domains of p300 and promotes sequence-specific DNA-dependent acetylation of p53 by p300 [16] (Fig. 2). In contrast to Ser20 phosphorylation, phosphorylation at Thr18 by CHK2 or casein kinase 1 has two notable effects: completely destabilizing the p53 – MDM2 complex [14,17] and stabilizing the p300 – p53 complex [15]. The control of MDM2 by phosphorylation Dynamic conformational changes are induced in MDM2 upon ligand binding P53 induces the oncogene MDM2, creating a negativefeedback loop that maintains the steady-state levels of p53 [2]. However, this control is not a strict genetic negativefeedback loop because dosage compensation does not regulate p53 activity: p53-heterozygotes have higher cancer incidence than diploid wild-type p53 animals [18] and p53-knock-ins have reduced tumour incidence compared with diploid wild-type p53 [19]. Thus, the MDM2 – p53 feedback loop is not a closed system, and it is exposed to alternate controls. Defining the factors that control the conformational flexibility of MDM2 could reveal these other regulatory pathways. The biochemical effects of MDM2-binding to p53 result in multiple sites of mono-ubiquitination of p53 [3,20]. Although the mechanistic details of the ubiquitination and degradation reaction are lacking, important regulatory stages are now emerging from in vitro studies. For example, after p53 mono-ubiquitination by MDM2, p300 can catalyze poly-ubiquitination of mono-ubiquitinated p53 and promote p53-protein degradation [6]. The initial binding of MDM2 to p53 requires an N-terminal domain in MDM2 (, 110 amino acids) containing a deep hydrophobic pocket that interacts with the BOX-I activation domain of p53 [21] (Fig. 1). The structure of this MDM2 – BOX-I polypeptide complex has been analyzed, and data acquired recently using NMR has demonstrated that the type of peptide fragment bound to the MDM2 pocket can induce significantly different conformational changes in MDM2 [14]. This conformational change in MDM2 upon ligand
Update
348
LID
TRENDS in Biochemical Sciences
MDM2
Vol.28 No.7 July 2003
p53
(b)
(c)
p300 p300
p53 AC
(a)
(d)
MDM2 LID P Ti BS
Fig. 2. Concerted recruitment of protein kinases modifies the murine double minute 2 (MDM2)– p53 protein complex. Protein-kinase activation of the p53 pathway can now be divided into different biochemical steps involving: (a) phosphorylation (yellow) of the MDM2 lid (red) at Ser17 by a phosphatidylinositol (PtdIns)-3-like kinase-dependent pathway [23], which stabilizes the MDM2 pseudo-substrate lid over the MDM2 pocket [7]; (b) the release of p53 protein from MDM2 control; (c) phosphorylation of p53 at Thr18 and/or Ser20 by enzymes such as checkpoint kinase-2 (CHK2) [8] or other Ser20 kinases [11,12] that stabilize the docking of p300 onto the p53 tetramer through the interferon-3 binding domain (IBiD) and IBiD homology domain (IHD) of p300 [16] (white arrows); and (d) subsequent DNA-dependent acetylation (red arrow) of the p53 tetramer [16]. Permutations in the ataxia telangiectasia mutated (ATM) and/or CHK2 kinase pathways could influence the magnitude of the p53 response. The phosphorylation of p53 at Ser15 or at Ser20 has little effect on MDM2 binding [14], whereas phosphorylation at Ser20 stabilizes p300 docking and promotes DNA-dependent acetylation of p53 [16]. By contrast, phosphorylation at Thr18 can disrupt the MDM2– p53 complex [14,17] and simultaneously stabilize the p300–p53 complex [15] leading to p53 activation without MDM2-lid phosphorylation. Furthermore, MDM2-lid phosphorylation by PtdIns-3 kinases such as ATM and DNA-PK without p53 phosphorylation at Thr18 and Ser20 could also stabilize p53 protein. Therefore, the concerted action of enzymes such as ATM and CHK2 targeting both MDM2 and p53, respectively, gives maximal activation of the p53 response. Although p300 is required for p53 acetylation, co-activator recruitment and associated transactivation [28,29], p300 can also work with MDM2 to catalyze p53 poly-ubiquitination and degradation [6]. Presumably, this stage of p300 function occurs when the p53 response needs to be turned off.
binding is probably transmitted to other domains on the full-length MDM2 protein and might affect other activities of MDM2, including binding to p14ARF, p300, Zinc and RNA. In fact, RNA binding induces a conformational change in full-length MDM2 protein and can switch the substrate affinity of MDM2 from the original BOX-I activation domain to a flexible loop in the core domain of p53 [22]. Moreover, zinc binding by full-length MDM2 can destabilize MDM2 – p53-BOX-I-peptide complexes [22], suggesting a mechanism for allosteric regulation of the MDM2 – p53 complex. This intrinsic conformational flexibility of the MDM2 protein can be used to predict that it will be the focus of multiple post-translational control pathways. A novel control of the MDM2 – p53 protein complex: a flexible lid fits over the MDM2 pocket A recent report has highlighted a novel biological regulation of the MDM2 –p53 protein complex involving a flexible lid that can close over the hydrophobic pocket of MDM2 [7]. Previous work has identified the minimal domain in MDM2 that can bind to the p53 activation domain and the MDM2 hydrophobic pocket is lined with over ten aromatic and hydrophobic amino acids that make van der Waals contacts to p53. This hydrophobic pocket is maintained, in part, by b sheets that can form a scaffold for the ligand and, despite original claims that this structure is rigid [21], is very flexible [14]. This flexibility is the key to understanding the dynamic nature of MDM2 protein regulation. By focusing on the detailed NMR structure of MDM2 containing additional amino acids in the N terminus of the protein, assignments can be placed for a peptide fragment that fits over the hydrophobic pocket of MDM2 [7]. This peptide lid has a significant degree of homology to the p53-activation domain fragment (Fig. 1), suggesting that it might be able to compete in cis for binding with p53 and act like a pseudo-substrate domain. Furthermore, because the MDM2-lid also has a classic DNA-PK or ATM consensus site (SQ motif) whose http://tibs.trends.com
phosphorylation disrupts the p53– MDM2 complex [23], modification of the affinity of the lid for the MDM2 pocket is possible. In fact, a phospho-mimetic substitution at Ser17 (Fig. 1) in the MDM2-lid stabilizes its interaction with MDM2 and alters the conformation of MDM2 [7]. This is consistent with independent NMR studies that show substantial conformational shifts in MDM2 upon ligand binding [14]. These data provide an additional step for a PtdIns-3 kinase such as ATM to disrupt the MDM2 – p53 complex (Fig. 2). In summary, the conformational flexibility of MDM2 protein, the presence of a pseudosubstrate lid that might control MDM2-binding proteins, the existence of multiple ligand-binding domains and the contribution of RNA binding to ubiquitination activity all highlight the complex intradomain communication that exists within MDM2. Further research aimed at integrating p53-specific kinase signalling pathways to intradomain interactions of MDM2 will enhance details on how the p53 pathway is controlled post-translationally, and assist in developing assays to identify modifying loci of the p53 –MDM2 negative-feedback loop. Concluding remarks: intrasteric regulation in enzyme control Protein function in cells can be regulated in many ways including ligand binding, covalent modification by acetylation or phosphorylation and proteolytic processing. In addition, many regulatory proteins have intrinsic conformational control mechanisms as an additional layer of post-translational regulation. Such built-in conformational restraints are often called allosteric, as they simply describe the presence of cis-acting, conformationally flexible regulatory motifs that are outside the active site of the protein. A subclass of the allosteric model of enzyme regulation is the more recently highlighted intrasteric regulation by a pseudo-substrate domain [24,25]. This concept was highlighted as an auto-regulatory control mechanism in protein kinases in which an internal sequence of the enzyme has high homology to the substrate
Update
TRENDS in Biochemical Sciences
of the kinase, and disruption of this pseudo-substrate domain is an important component in kinase activation [26]. This mechanism is not confined to protein kinases [27]. The identification of a pseudo-substrate domain in the oncogene MDM2, whose affinity for MDM2 can be modified by phosphorylation, expands on the class of proteins that can be regulated by intrasteric mechanisms and has interesting implications for control of other MDM2-binding proteins like p53, p300, E2F, p73 and Rb. References 1 Vogelstein, B. et al. (2000) Surfing the p53 network. Nature 408, 307 – 310 2 Freedman, A. et al. (1999) Functions of the MDM2 oncoprotein. Cell. Mol. Life Sci. 55, 96 – 107 3 Honda, R. et al. (1997) Oncoprotein MDM2 is a ubiquitin ligase E3 for tumor suppressor p53. FEBS Lett. 420, 25 – 27 4 Sakaguchi, K. et al. (1998) DNA damage activates p53 through a phosphorylation – acetylation cascade. Gene Dev. 12, 2831– 2841 5 Yuan, Z.M. et al. (1999) Role for p300 in stabilization of p53 in the response to DNA damage. J. Biol. Chem. 274, 1883 – 1886 6 Grossman, S.R. et al. (2003) Polyubiquitination of p53 by a ubiquitin ligase activity of p300. Science 300, 342– 344 7 McCoy, M.A. et al. (2003) Flexible lid to the p53-binding domain of human Mdm2: Implications for p53 regulation. Proc. Natl. Acad. Sci. U. S. A. 100, 1645– 1648 8 Banin, S. et al. (1998) Enhanced phosphorylation of p53 by ATM in response to DNA damage. Science 281, 1674 – 1677 9 Shieh, S.Y. et al. (2000) The human homologs of checkpoint kinases Chk1 and Cds1 (Chk2) phosphorylate p53 at multiple DNA damageinducible sites. Genes Dev. 14, 289 – 300 10 Shieh, S.Y. et al. (1997) DNA damage-induced phosphorylation of p53 alleviates inhibition by MDM2. Cell 91, 325 – 334 11 Craig, A.L. et al. (1999) Dephosphorylation of p53 at Ser20 after cellular exposure to low levels of non-ionizing radiation. Oncogene 18, 6305 – 6312 12 Ahn, A. et al. Questioning the role of Chk2 in the p53 DNA damage response. J. Biol. Chem. (in press) 13 Sakaguchi, K. et al. (2000) Damage-mediated phosphorylation of human p53 threonine 18 through a cascade mediated by a casein 1-like kinase. Effect on Mdm2 binding. J. Biol. Chem. 275, 9278 – 9283 14 Schon, O. et al. (2002) Molecular mechanism of the interaction between MDM2 and p53. J. Mol. Biol. 323, 491 – 501
Vol.28 No.7 July 2003
349
15 Dornan, D. and Hupp, T.R. (2001) Inhibition of p53-dependent transcription by BOX-I phospho-peptide mimetics that bind to p300. EMBO Rep. 2, 139 – 144 16 Dornan, D. et al. (2003) DNA-dependent acetylation of p53 by the transcription coactivator p300. J. Biol. Chem. 278, 13431 – 13441 17 Craig, A.L. et al. (1999) Novel phosphorylation sites of human tumour suppressor protein p53 at Ser20 and Thr18 that disrupt the binding of mdm2 protein are modified in human cancers. Biochem. J. 342, 133– 141 18 Clarke, A.R. et al. (1993) Thymocyte apoptosis induced by p53dependent and independent pathways. Nature 362, 849– 852 19 Garcia-Cao, I. et al. (2002) Super p53 mice exhibit enhanced DNA damage response, are tumor resistant and age normally. EMBO J. 21, 6225– 6235 20 Lai, Z. et al. (2001) Human mdm2 mediates multiple monoubiquitination of p53 by a mechanism requiring enzyme isomerization. J. Biol. Chem. 276, 31357 – 31367 21 Kussie, P.H. et al. (1996) Structure of the MDM2 oncoprotein bound to the p53 tumor suppressor transactivation domain. Science 274, 948– 953 22 Shimizu, H. et al. (2002) The conformationally flexible S9 – S10 linker region in the core domain of p53 contains a novel MDM2 binding site whose mutation increases ubiquitination of p53 in vivo. J. Biol. Chem. 277, 28446 – 28458 23 Mayo, L.D. et al. (1997) Mdm-2 phosphorylation by DNA-dependent protein kinase prevents interaction with p53. Cancer Res. 57, 5013– 5016 24 Kobe, B. and Kemp, B.E. (1999) Active site-directed protein regulation. Nature 402, 373 – 376 25 Kobe, B. et al. (1997) Intrasteric regulation of protein kinases. Adv. Second Messenger Phosphoprotein Res. 31, 29 – 40 26 Kemp, B.E. et al. (1996) Intrasteric regulation of calmodulindependent protein kinases. Adv. Pharmacol. 36, 221 – 249 27 Kobe, B. et al. (1999) Structural basis of autoregulation of phenylalanine hydroxylase. Nat. Struct. Biol. 6, 442 – 448 28 Espinosa, J.M. and Emerson, B.M. (2001) Transcriptional regulation by p53 through intrinsic DNA/chromatin binding and site-directed cofactor recruitment. Mol. Cell 8, 57 – 69 29 Barlev, N.A. et al. (2001) Acetylation of p53 activates transcription through recruitment of coactivators/histone acetyltransferases. Mol. Cell 8, 1243– 1254
0968-0004/03/$ - see front matter q 2003 Elsevier Ltd. All rights reserved. doi:10.1016/S0968-0004(03)00140-3
| Protein Sequence Motif
The KIND module: a putative signalling domain evolved from the C lobe of the protein kinase fold Francesca D. Ciccarelli1,2, Peer Bork1,2 and Eugen Kerkhoff3 1
European Molecular Biology Laboratory, Meyerhofstr. 1, D-69012 Heidelberg, Germany Max-Delbrueck-Centrum, PO Box 740238, D-13092 Berlin, Germany 3 Institut fu¨r Medizinische Strahlenkunde und Zellforschung (MSZ), Universita¨t Wu¨rzburg, Versbacher Str. 5, D-97078 Wu¨rzburg, Germany 2
A novel putative interaction domain – KIND (kinase non-catalytic C-lobe domain) – has been identified as being similar to the C-terminal protein kinase catalytic fold (C lobe). Its presence at the N terminus of signalling proteins and the absence of the active-site residues in the catalytic and activation loops suggest that it folds Corresponding author: Francesca D. Ciccarelli ([email protected]). http://tibs.trends.com
independently and is likely to be non-catalytic. The occurrence of the novel domain only in metazoa implies that it has evolved from the catalytic protein kinase domain into an interaction domain possibly by keeping the substrate-binding features. The Drosophila melanogaster protein p150-Spir (CAB62901), which is phosphorylated by the c-Jun