- Email: [email protected]
Interactions between SH2 and SH3 Domains
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS ARTICLE NO.
242, 351–356 (1998)
RC977909
Interactions between SH2 and SH3 Domains Mauno Vihinen* and C. I. Edvard Smith†,‡ *Department of Biosciences, Division of Biochemistry, University of Helsinki, P.O. Box 56, FIN-00014 Finland; †Center for BioTechnology, Department of Biosciences at Novum, Karolinska Institute, S-141 57 Huddinge, Sweden; and ‡Department of Immunology, Microbiology, Pathology, and Infectious Diseases (IMPI), Karolinska Institute, Huddinge University Hospital, S-141 86 Huddinge, Sweden
Received November 24, 1997
Src homology 2 (SH2) and SH3 domains are abundant protein and peptide binding modules in signalling molecules. Certain SH2 and SH3 domains have been shown to form functional and physical interactions. The structural basis of dimer formation was studied by docking three dimensional structures of the domains and by analysing structural and functional properties of the dimers. The experimentally verified dimers were noticed to have very large buried surfaces, extensive hydrogen bonding networks, and complementary surfaces, properties which are characteristic for protein-protein interactions. The number of hydrogen bonds between the domains is exceptionally high for interacting protein pairs. Also the buried accessible surface is large, especially when considering the small size of the domains. The dimer results were used to describe mutation information in structural terms and to discuss regulation of protein tyrosine kinases. q 1998 Academic Press
Signalling proteins generally consist of several domains each of which can form typical interactions with their partners. Src homology 2 (SH2) and SH3 domains are abundant signalling domains (1, 2). SH2 domains bind to phosphorylated tyrosine (pY) residues in peptides and proteins. Also some residues C-terminal to pY are important for recognition. Polyproline II (PPII) type helices interact with SH3 domains. Activation and regulation of protein kinases are multistep processes. In the Src family protein tyrosine kinases (PTKs), SH2 and SH3 domains are required for the regulation of enzyme activity. Phosphorylation of specific tyrosine residue(s) is necessary in activation loop. Many kinases have also other phosphorylation sites. Src family PTKs are repressed by binding of their SH2 domain to phosphorylated tyrosine (Y527 in Src) in the C-terminal tail (3, 4) in SH3 domain dependent manner (5–7). Recent three dimensional structures of Src fam-
ily members Hck (8) and c-Src (9) with SH2 and SH3 domains revealed that the SH2 domain binds to the regulatory tyrosine and SH3 domain has binding site in the linker connecting SH2 and kinase domains (Fig. 1a). The catalytic site and substrate binding region are accessible, but the enzyme is in inactive conformation due to the organization of kinase domain lobes. In the active form, ATP and Mg2/ ions are bound between the two lobes of the kinase, whereas the upper lobe is twisted relative to the lower lobe in the inactive form. Regulation by intramolecular binding of SH2 and SH3 domains to the kinase domain tail and the linker between the SH2 and kinase domains, respectively, affects organization of the kinase lobes and thereby its activity. If the kinase is dephosphorylated or the SH2 and SH3 domains are prevented from binding either due to mutations (10, 11) or competition from high affinity peptides or proteins (12, 13) the enzyme is derepressed (Fig. 1b). Signalling domains can form several interactions. The SH2 and SH3 domains of Src family kinases have been shown to interact (14). Because the linker between the domains is too short to facilitate intramolecular recognition, dimerization might play a role in the regulation and function of these kinases. The structural basis of the dimers was addressed by studying the interactions between the three dimensional structures of the dimer forming domains. Analysis of several three dimensional structures indicates that many SH3 domains have a sequence stretch related to optimal SH2 domain binding sequence. Docking of the domains suggests stereochemically and energetically highly favourable interactions, which might have general importance in cellular signalling. MATERIALS AND METHODS The coordinates of the domains were taken from Protein Data Bank; Lck SH2 and SH3 domains entry 1lck (15), Src SH2 1sps (16), Src SH3 1rlp (17), and Fyn SH3 1nyf (18). The structure of the
351
0006-291X/98 $25.00 Copyright q 1998 by Academic Press All rights of reproduction in any form reserved.
AID
BBRC 7909
/
6945$$$241
12-30-97 18:26:55
bbrcg
AP: BBRC
Vol. 242, No. 2, 1998
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 1. Different modes of regulation of protein tyrosine kinases. (a) Src family PTKs are repressed by the binding of the SH2 domain to phosphotyrosine (P) in the C-terminal tail of the kinase. SH3 domain recognizes a proline rich segment (triangle) in the linker between the SH2 and kinase domains. Unique N-terminal domain is indicated with a diamond. (b) The Src family PTKs can be derepressed by dephosphorylation of the C-terminal regulatory tyrosine, by mutations in the SH3 domain, or by binding of high affinity peptides or proteins to the SH2 and SH3 domains. The kinase domain can bind to its substrate (curved shape) and phosphorylate it. (c) Dimer formation between SH2 and SH3 domains. It is not known whether the full length protein can form one or two SH2-SH3 dimers. (d) Phosphorylation of a tyrosine close to the binding site of SH3 domain prevents intramolecular binding to the kinase domain. Btk is depicted with the PH domain (square) and TH domain (arrow). (e) Increased specificity of PTKs due to simultaneous recognition for the substrate by several domains. (f) Phosphorylation of the SH2 domain leads to derepression of the kinase. (g) Intramolecular binding of the proline rich region in the TH domain into the Btk SH3 domain.
Btk SH3 domain has been determined with NMR (Hansson et al., submitted) and SH3 domain has been modeled (19). The three dimensional structures of the SH3 domains were docked to the SH2 domains by fitting the putative binding residues to the high affinity peptide of Src SH2 domain (16). The backbone of the residues pY to /3 were superimposed with the residues in the SH3 domains and the root mean square (rms) deviation was determined. Other SH2 domains, for which peptide complex structure was not available, were superimposed on top of the Src SH2 domain complex. The peptide was removed before further analysis. Final adjustment of the do-
mains in the docked pairs were made either manually fixing torsion angles of side chains or running a short energy minimization with program Discover (Molecular Simulations Inc.). In the energy minimization only the side chain atoms of the clashing residues were allowed to move. The hydrogen bonds were analysed by using program InsightII (Molecular Simulations, Inc.) which employed distance and stereochemical criteria. Atom pairs having favourable angle and distance ˚ were considered as hydrogen bond forming. Beequal or below 3 A cause this method does not take into account too closely positioned
352
AID
BBRC 7909
/
6945$$$241
12-30-97 18:26:55
bbrcg
AP: BBRC
Vol. 242, No. 2, 1998
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS TABLE 1
Analysis of SH2-SH3 Domain Dimers SH3
SH2
Number of hydrogen bonds
Total buried ˚ 2) accessible surface (A
Lcka Src Fyn Fyn Btk
Lck Src Lck Src Btk
5 22 20 19 15
1057 1712 2027 1798 1748
a
This is the dimer from crystal structure (15).
residues, the numbers of clashing residues were analysed. The accessible surfaces of the free and bound domains were determined with program Grasp (20).
SH2-SH3 DOMAIN DIMER FORMATION Binding of the SH2 and SH3 domains to their partners is important for signalling cascades. These do-
mains can be essential for localization of kinases and their synergistic binding may increase specificity of recognition. Analysis of SH2 and SH3 domains of Fyn and several other signalling molecules verified direct interactions between these modules (14). The SH3-SH2 structure of Src family member, Lck, has been determined (15). In crystals, the tandem domains form a dimer, where an SH2 domain proline is in the polyproline binding pocket of the opposing SH3 domain. One face of the triangular PP-II type helices has extensive interactions with SH3 domains (1, 2). The proline containing region of Lck SH3 domain does not have PP-II fold and only part of the SH3 domain binding pocket is occupied. Despite rather large buried accessible surface (Table 1) and shape complementarity, the dimer in Lck crystals does not resemble protein-protein interfaces. Typical interfaces have large buried surfaces, numerous hydrogen bonds and hydrophobic and other interactions (21-26). The interface in Lck has weak interactions which might have arisen from crystal packing effects. There is only five hydrogen bonds between the
FIG. 2. Dimer formation between SH2 and SH3 domains. (a) Binding of Src SH3 domain on white (17) to Src SH2 domain (16) on blue. The domains were docked on top of the high affinity SH2 domain ligand. The domains have complementary surfaces, very large buried accessible surface, and an excessive hydrogen bond network. (b) Ribbon presentation of Src SH2 (magenta) and SH3 (cyan) domains. The termini of the domains are shown, as well as the binding residues in the SH3 domain (Y92 to R95) (in yellow). Tyrosine residues shown to be phosphorylated in some kinases and affecting regulation are in red; from top to bottom Y136, Y90, and Y213. 353
AID
BBRC 7909
/
6945$$$241
12-30-97 18:26:55
bbrcg
AP: BBRC
Vol. 242, No. 2, 1998
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
domains when protein pairs have on average 10 (21) and the other studied dimers form close to 20 bonds (Table 1). Another SH2-SH3 structure has been determined for adaptor protein Grb 2 (27), which has the domains in the order SH3-SH2-SH3. Also this structure is a dimer where SH2 and SH3 domains interact, but the binding sites of the domains are accessible. The long linkers between the domains provide freedom for different orientations of the domains and for different binding modes (27). SH2 domain interactions with partners does not affect binding to the SH3 domains and vice versa (28). There are only eight hydrogen bonds between the Grb2 molecules in the dimer, although the buried accessible surface is large. Therefore the dimer in the three dimensional structure may be the result of crystallization. The binding specificity of several SH2 domains has been studied by using phosphopeptide libraries (29, 30). Sequence analysis indicated that SH3 domains of Src PTKs contain in the RT loop a sequence related to the optimal SH2 domain binding pattern. In Src the sequence is YESR (residues 92 to 95), in Fyn YEAR and in Lck YEPS. The optimal binding sequence of all Src family SH2 domains is pYEEI (29). We used this information as a starting point for docking SH3 domains into SH2 domains. Several SH2 and SH3 domains have been shown to form both physical and functional interactions (14) (Fig. 1c). The structural features of these interactions were studied by superimposing the backbone atoms of the putative binding residues of SH3 domain to amino acids pY to /3 of the high affinity peptide complex of Src SH2 domain (16) and then analysing the interactions within the dimer (Figs. 2a and b). In the docking studies the residue binding to the pY pocket was not phosphorylated, because the observed dimers are formed between unphosphorylated domains (14). The Src SH2 and SH3 domains (17) fitted well to each other (Fig. 2b). The rms deviation for the backbone atoms of residues pY to /3 in the high affinity peptide and resi˚ indicating close dues 92 to 95 in Src is only 0.85 A structural relationship. The rms deviation can be compared to the deviations of three dimensional structures determined with NMR. In NMR structure determination several structures are obtained for the same protein and their rms deviations are of the same order (31) as difference between the binding peptide stretches. For example, in the rigid domain docking of Src SH2 and SH3 domains only two regions had minor steric clashes, which could be easily resolved by adjusting flexible side chain torsion angles. No backbone adjustments were necessary. The average number of hydrogen bonds and electrostatic interactions in protein-protein recognition sites is 10 (21). In the docked Src SH2-SH3 dimer there is altogether 22 interactions indicating strong recognition
and complementarity. Hydrogen bonds have free energies in the order of 0.5-1.8 kcal/mol for neutral bonds, whereas charged or ion pair hydrogen bonds can contribute 3-6 kcal/mol (32). Thus, hydrogen bonds provide substantial stabilizing forces for the SH2-SH3 dimers, especially since the number of interactions is about two times larger than in interacting proteins in general. Another feature typical for strong protein-protein interactions is large buried accessible surface. Analysis of high resolution structures of protein complexes indi˚ 2 (16). Similar values cated a value of 1600 { 350 A have been obtained in the analysis of subunit interfaces of oligomeric proteins (33). In Src dimer the value is ˚ 2 (Table 1), which is very large when we about 1700 A take into account that SH2 and SH3 domains are relatively small compared to full size proteins. The pep˚ 2, and tide residues pY to /3 provide only about 600 A thereby the buried surface is mainly contributed by other regions of the interacting domains. The major forces favouring protein associations are hydrophobic energy from buried accessible surfaces and electrostatic energy from hydrogen bonds (21). It has been suggested that non-polar surfaces could pro˚ 2 when buried (21). It is diffivide as much as 25 cal/A cult to estimate different forces contributing to the stability of the dimers, especially in relation to entropy. Anyhow, the extensive hydrogen bonding, large buried surface area and complementarity of the surfaces with the experimental evidence for interactions are suggesting energetically highly favourable binding interactions. The dimer described in here differs markedly of that in the Lck and Grb2 crystals (15, 27). Although the pY binding cannot be the driving force since the SH3 domains are not generally phosphorylated, also the aromatic ring of the pY has extensive binding, e.g. in Src SH2 domain three residues interact with the aromatic ring of the pY (16). For some SH2 domains non-pY dependent binding has been demonstrated (34–38). In addition to the specificity determining residues, of which amino acids at positions /1 and /3 are the most important, also other regions may affect binding. As an example, the affinity of the HIV Nef protein to Src family SH3 domains is more than 300 fold higher than that for a corresponding peptide (39) due to extensive interactions outside the polyproline recognition region (40). Domains from several Src family members were shown to interact and form both homo- and heterodimers (14). The docked Src family domains have an extensive hydrogen bond network (Table 1), which with structural complementarity and large buried interface suggests highly favourable dimer formation. Also the observed heterodimers (14) were found to have matching surfaces and numerous hydrogen bonds (Table 1). The unproductive dimers of Fyn and Lck SH3 domain with Lck and p85a subunit of PI3 kinase SH2 domain (41) could be tested by docking. They were char-
354
AID
BBRC 7909
/
6945$$$241
12-30-97 18:26:55
bbrcg
AP: BBRC
Vol. 242, No. 2, 1998
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
acterized by infavourable binding sequence in the RT loop, weak hydrogen bonding and most importantly severe structural clashes when docked. The docked Lck SH2 and SH3 domains have a large buried accessible ˚ 2). There are also numerous atoms at surface (1677 A hydrogen bonding distance, but even after energy minimization several clashes remained, suggesting that these domains cannot in fact form productive dimer. This is indeed the case as shown by the experimental results (14). Thus, the SH2-SH3 dimerization requires in addition to proper sequence, structural complementarity and complementarity also outside the actual recognition site. The binding sequence in the SH3 domains is not optimal for the studied SH2 domains, although in Src family the residue at position /1, glutamate, is conserved. Src SH3 domain has R95 at position /3, when the optimal residue at this position is isoleucine. Arginine has very flexible side chain, which can bend such that the strongly basic d-guanidino group comes to surface. In the Nef-SH3 domain structure Fyn has R96I mutation in the RT loop (40). The affinity of the wild type Fyn has Kd ú20 mM (35), while the highest affinities of SH3 domain binding peptides are in the order of 10 mM (2). The full length Nef has Kd of about 0.38 mM. Thus, remarkably nonideal residue (when compared to the consensus binding sequence) is allowed at position /3 in Src and Fyn. The reduction in the affinity caused by nonideal residue is presumably compensated e. g. by complementary surface and hydrogen bonding. MUTATIONAL ANALYSIS OF DIMER FORMATION The dimer formation suggested in here could be prevented by mutations in key residues either in the SH2 or SH3 domain. If the dimerization were dependent on the phosphorylation, mutations at phosphate binding residues would prevent recognition. This seems not to be the case, because mutation R176K at pY residue phosphate binding residue of Fyn SH2 domain does not prevent dimerization (14). Also SH2 domain mutations S186P and R190C show normal domain recognition, although both these mutations are partially defective in pY binding (14). S186 locates in the b-strand B and R190 in a loop connecting two b structures. Thus, the SH2 domain can be mutated in phosphate group binding region without affecting phoshotyrosine independent binding of SH3 domain. Also mutations in the binding region of SH3 domain would hinder dimerization. In c-Src SH3 domain, Y92 and Y138 (Fig. 2a) have been shown to be phosphorylated in certain signalling pathways and to affect ligand binding (42). Y92 lies just two residues before the putative dimerization region (Y94 to R97). The corresponding residue is important also for regulation of Tec family member, Bruton’s tyrosine kinase (Btk), in which autophosphorylation of Y223 inactivates SH3
domain function (43). Phosphorylation would prevent dimerization (Fig. 1d), which according to docking of the modelled Btk SH3 (Hansson et al., submitted) and SH2 domains (20) is possible between the unphosphorylated modules (Table 1). CONCLUSIONS The phosphorylated Src family kinases are regulated by intramolecular interactions of the SH2 and SH3 domains with the kinase domain and its extension (8, 9). Increased specificity in binding to partners can be obtained by simultaneous recognition by several domains (44) (Fig. 1e). Based on the dimerization results, PTKs could be regulated by dimer formation between SH2 and SH3 domains. This would lead to the activation of the kinase domain. Further clue for the dimerization is provided by the domain organization of kinases, adaptors and other signalling molecules. The SH3 domain precedes almost without an exception SH2 domain in close proximity when both are present. Although the affinity of unphosphorylated peptide can in some cases be 104 times lower than that of a corresponding phosphorylated peptide (45), the SH2-SH3 dimerization could be important because there are other interactions and extensive buried surface. Furthermore, Shc has recently been shown to bind to the Grb2 SH2 domain regardless of the state of the tyrosine phosphorylation (45). These authors speculate that carboxylic group of aspartate could mimic the phosphate group of phosphotyrosine. According to our docking results, it is more likely that aromatic residue at pY binding pocket is sufficient for binding. The structure of the PTKs is presumably flexible due to linkers between domains. Grb2 has been suggested to have several conformations and binding modes (27). Interacting PTKs could form two of the dimers described above and still the kinase domains could be accessible. In addition to processes discussed above, the domains are involved in several mechanisms. SH2 domain phosphorylation at Y192 in Lck (47) and Y213 in Src (48) decreases affinity to ligands (Fig. 1f, 2a). These residues are outside the immediate recognition site. Another regulatory variation is provided by a proline rich region in the TH domain of Tec family PTKs, which has been shown in Itk to bind intramolecularly to the SH3 domain (49) (Fig. 1g). The proline rich region in Tec family kinases is recognized also by SH3 domains of Src family members (50-52). Thus, proteins containing SH2 and SH3 domains can be regulated in several ways possibly allowing a number of conformations involved in both intra- and intermolecular associations. ACKNOWLEDGMENTS This work has been supported by Finnish Academy, Biocentrum Helsinki, Instrumentariumin tiedesa¨a¨tio¨, Swedish Cancer Society,
355
AID
BBRC 7909
/
6945$$$241
12-30-97 18:26:55
bbrcg
AP: BBRC
Vol. 242, No. 2, 1998
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
˚ ke Wiberg Foundation, and Swedish Medical Research Council, A Magn. Bergvall Foundation.
REFERENCES 1. Pawson, T. (1995) Nature 373, 573–580. 2. Cohen, G. B., Ren, R., and Baltimore, D. (1995) Cell 80, 237– 248. 3. Cantley, L. C., Auger, K. R., Carpenter, C., Duckworth, B., Graziani, A., Kapeller, R., and Soltoff, S. (1991) Cell 64, 281–302. 4. Roussel, R. R., Brodeur, S. R., Shalloway, D., and Laudano, A. (1991) Proc. Natl. Acad. Sci. USA 88, 10696–10700. 5. Murphy, S. M., Bergman, M., and Morgan, D. O. (1993) Mol. Cell. Biol. 13, 5290–5300. 6. Okada, M., Howell, B. W., Broome, M. A., and Cooper, J. A. (1993) J. Biol. Chem. 268, 18070–18075. 7. Superti-Furga, G., Fumagalli, S., Koegl, M., Courtneidge, S. A., and Draetta, G. (1993) EMBO J. 12, 2625–2634. 8. Sicheri, F., Moarefi, I., and Kuriyan, J. (1997) Nature 385, 602– 609. 9. Xu, W., Harrison, S. C., and Eck, M. J. (1997) Nature 385, 595– 602. 10. DeClue, J. E., and Martin, G. S. (1989) J. Virol. 63, 542–554. 11. O’Brien, M. C., Fukui, Y., and Hanafusa, H. (1990) Mol. Cell. Biol. 10, 2855–2862. 12. Liu, X., Brodeur, S. R., Gish, G., Songyang, Z., Cantlye, L. C., Laudnao, A. P., and Pawson, T. (1993) Oncogene 8, 1119–1126. 13. Moarefi, I., LaFevre-Bernt, M., Sicheri, F., Huse, M., Lee, C.-h., Kuriyan, J., and Miller, T. (1997) Nature 385, 560–653. 14. Panchamoorthy, G., Fukazawa, T., Stolz, L., Payne, G., Reedquist, K., Shoelson, S., Songyang, Z., Cantley, L., Walsh, C., and Band, H. (1994) Mol. Cell. Biol. 14, 6372–6385. 15. Eck, M. J., Atwell, S. K., Shoelson, S. E., and Harrison, S. C. (1994) Nature 368, 764–769. 16. Waksman, G., Shoelson, S. E., Pant, N., Cowburn, D., and Kuriyan, J. (1993) Cell 72, 779–790. 17. Feng, S., Chen, J. K., Yu, H., Simon, J. A., and Schreiber, S. L. (1994) Science 266, 1241–1247. 18. Morton, C. J., Pugh, D. J. R., Brown, E. L. J., Kahmann, J. D., Renzoni, D. A. C., and Campbell, I. D. (1996) Structure 4, 705– 714. 19. Vihinen, M., Nilsson, L., and Smith, C. I. E. (1994) Biochem. Biophys. Res. Commun. 205, 1270–1277. 20. Nicholls, A., Sharp, K. A., and Honig, B. (1991) Proteins Struct. Funct. Genet. 11, 281–296. 21. Janin, J., and Chothia, C. (1990) J. Biol. Chem. 265, 16027– 16030. 22. Janin, J., Miller, S., and Chothia, C. (1988) J. Mol. Biol. 204, 155–164. 23. Argos, P. (1988) Protein Eng. 2, 101–113. 24. Korn, A. P., and Burnett, R. M. (1991) Proteins Struct. Funct. Genet. 9, 37–55. 25. Jones, S., and Thornton, J. (1996) Proc. Natl. Acad. Sci. USA 93, 13–20. 26. Lijnzaad, P., and Argos, P. (1997) Proteins Struct. Funct. Genet. 28, 333–343. 27. Maignan, S., Guilloteau, J.-P., Fromage, N., Arnoux, B., Becquart, J., and Ducruix, A. (1995) Science 268, 291–293.
28. Cussac, D., Frech, M., and Chardin, P. (1994) EMBO J. 13, 4011–4021. 29. Songyang, Z., Shoelson, S. E., Chaudhuri, M., Gish, G., Pawson, T., Haser, W. G., King, F., Roberts, T., Ratnofsky, S., Lechleider, R. J., Neel, B. G., Birge, R. B., Fajardo, J. E., Chou, M. M., Hanafusa, H., Schaffhausen, B., and Cantley, L. C. (1993) Cell 72, 767–778. 30. Songyang, Z., Shoelson, S. E., McGlade, J., Olivier, P., Pawson, T., Bustelo, X. R., Barbacid, M., Sabe, H., Hanafusa, H., Yi, T., Ren, R., Baltimore, D., Ratnofsky, S., Feldman, R. A., and Cantley, L. C. (1994) Mol. Cell. Biol. 14, 2777–2785. 31. Evans, J. N. S. (1996) Biomolecular NMR Spectroscopy, Oxford Univ. Press, Oxford. 32. Fersht, A. R., Shi, J. P., Knill-Jones, J., Lowe, D. M., Wilkinson, A. J., Blow, D. M., Brick, P., Carter, P., Waye, M. M. Y., and Winter, G. (1985) Nature 314, 235–238. 33. Janin, J., Miller, S., and Chothia, C. (1988) J. Mol. Biol. 204, 155–164. 34. Pendergast, A. M., Muller, A. J., Havlik, M. H., Maru, Y., and Witte, O. N. (1991) Cell 66, 161–171. 35. Muller, A. J., Pendergast, A.-M., Havlik, M. H., Puil, L., Pawson, T., and Witte, O. N. (1992) Mol. Cell. Biol. 12, 5087–5093. 36. Bibbins, K. B., Boeuf, H., and Varmus, H. E. (1993) Mol. Cell. Biol. 13, 7278–7287. 37. Cleghon, V., and Morrison, D. K. (1994) J. Biol. Chem. 269, 17749–17755. 38. Ng, D. H. W., Watts, J. D., Aebersold, R., and Johnson, P. (1996) J. Biol. Chem. 271, 1295–1300. 39. Lee, C-.H., Leung, B., Lemmon, M. A., Zheng, J., Cowburn, D., Kuriyan, J., and Saksela, K. (1995) EMBO J. 14, 5006–5015. 40. Lee, C.-H., Saksela, K., Mirza, U. A., Chait, B. T., and Kuriyan, J. (1996) Cell 85, 931–942. 41. Booker, G. W., Breeze, A. L., Downing, A. K., Panaytou, G., Gout, I., Waterfield, M. D., and Campbell, I. D. (1992) Nature 358, 684–687. 42. Broome, M. A., and Hunter, T. (1997) Oncogene 14, 17–34. 43. Park, H., Wahl, M. I., Afar, D. E. H., Turck, C. W., Rawlings, D. J., Tam, C., Scharenberg, A. M., Kinet, J.-P., and Witte, O. N. (1996) Immunity 4, 515–525. 44. Kanemitsu, M. Y., Loo, L. W. M., Simon, S., Lau, A. F., and Eckhart, W. (1997) J. Biol. Chem. 272, 22824–22831. 45. Piccione, E., Case, R. D., Domchek, S. M., Hu, P., Chaudhuri, M., Backer, J. M., Schlessinger, J., and Shoelson, S. E. (1993) Biochemistry 192, 3197–3202. 46. Rojas, M., Yao, S., Donahue, J. P., and Lin, Y.-Z. (1997) Biochem. Biophys. Res. Commun. 234, 675–680. 47. Couture, C., Songyang, Z., Jascur, T., Williams, S., Tailor, P., Cantley, L. C., and Mustelin, T. (1996) J. Biol. Chem. 271, 24880–24884. 48. Stover, D. R., Furet, P., and Lydon, N. B. (1996) J. Biol. Chem. 271, 12481–12487. 49. Andreotti, A. H., Bunnell, S. C., Feng, S., Berg, L. J., and Schreiber, S. L. (1997) Nature 385, 93–97. 50. Cheng, C., Ye, Z.-S., and Baltimore, D. (1994) Proc. Natl. Acad. Sci. USA 91, 8152–8155. 51. Alexandropoulos, K., Cheng, G., and Baltimore, D. (1995) Proc. Natl. Acad. Sci. USA 92, 3110–3114. 52. Yang, W., Malek, S. N., and Desiderio, S. (1995) J. Biol. Chem. 270, 20832–20840.
356
AID
BBRC 7909
/
6945$$$241
12-30-97 18:26:55
bbrcg
AP: BBRC
242, 351–356 (1998)
RC977909
Interactions between SH2 and SH3 Domains Mauno Vihinen* and C. I. Edvard Smith†,‡ *Department of Biosciences, Division of Biochemistry, University of Helsinki, P.O. Box 56, FIN-00014 Finland; †Center for BioTechnology, Department of Biosciences at Novum, Karolinska Institute, S-141 57 Huddinge, Sweden; and ‡Department of Immunology, Microbiology, Pathology, and Infectious Diseases (IMPI), Karolinska Institute, Huddinge University Hospital, S-141 86 Huddinge, Sweden
Received November 24, 1997
Src homology 2 (SH2) and SH3 domains are abundant protein and peptide binding modules in signalling molecules. Certain SH2 and SH3 domains have been shown to form functional and physical interactions. The structural basis of dimer formation was studied by docking three dimensional structures of the domains and by analysing structural and functional properties of the dimers. The experimentally verified dimers were noticed to have very large buried surfaces, extensive hydrogen bonding networks, and complementary surfaces, properties which are characteristic for protein-protein interactions. The number of hydrogen bonds between the domains is exceptionally high for interacting protein pairs. Also the buried accessible surface is large, especially when considering the small size of the domains. The dimer results were used to describe mutation information in structural terms and to discuss regulation of protein tyrosine kinases. q 1998 Academic Press
Signalling proteins generally consist of several domains each of which can form typical interactions with their partners. Src homology 2 (SH2) and SH3 domains are abundant signalling domains (1, 2). SH2 domains bind to phosphorylated tyrosine (pY) residues in peptides and proteins. Also some residues C-terminal to pY are important for recognition. Polyproline II (PPII) type helices interact with SH3 domains. Activation and regulation of protein kinases are multistep processes. In the Src family protein tyrosine kinases (PTKs), SH2 and SH3 domains are required for the regulation of enzyme activity. Phosphorylation of specific tyrosine residue(s) is necessary in activation loop. Many kinases have also other phosphorylation sites. Src family PTKs are repressed by binding of their SH2 domain to phosphorylated tyrosine (Y527 in Src) in the C-terminal tail (3, 4) in SH3 domain dependent manner (5–7). Recent three dimensional structures of Src fam-
ily members Hck (8) and c-Src (9) with SH2 and SH3 domains revealed that the SH2 domain binds to the regulatory tyrosine and SH3 domain has binding site in the linker connecting SH2 and kinase domains (Fig. 1a). The catalytic site and substrate binding region are accessible, but the enzyme is in inactive conformation due to the organization of kinase domain lobes. In the active form, ATP and Mg2/ ions are bound between the two lobes of the kinase, whereas the upper lobe is twisted relative to the lower lobe in the inactive form. Regulation by intramolecular binding of SH2 and SH3 domains to the kinase domain tail and the linker between the SH2 and kinase domains, respectively, affects organization of the kinase lobes and thereby its activity. If the kinase is dephosphorylated or the SH2 and SH3 domains are prevented from binding either due to mutations (10, 11) or competition from high affinity peptides or proteins (12, 13) the enzyme is derepressed (Fig. 1b). Signalling domains can form several interactions. The SH2 and SH3 domains of Src family kinases have been shown to interact (14). Because the linker between the domains is too short to facilitate intramolecular recognition, dimerization might play a role in the regulation and function of these kinases. The structural basis of the dimers was addressed by studying the interactions between the three dimensional structures of the dimer forming domains. Analysis of several three dimensional structures indicates that many SH3 domains have a sequence stretch related to optimal SH2 domain binding sequence. Docking of the domains suggests stereochemically and energetically highly favourable interactions, which might have general importance in cellular signalling. MATERIALS AND METHODS The coordinates of the domains were taken from Protein Data Bank; Lck SH2 and SH3 domains entry 1lck (15), Src SH2 1sps (16), Src SH3 1rlp (17), and Fyn SH3 1nyf (18). The structure of the
351
0006-291X/98 $25.00 Copyright q 1998 by Academic Press All rights of reproduction in any form reserved.
AID
BBRC 7909
/
6945$$$241
12-30-97 18:26:55
bbrcg
AP: BBRC
Vol. 242, No. 2, 1998
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 1. Different modes of regulation of protein tyrosine kinases. (a) Src family PTKs are repressed by the binding of the SH2 domain to phosphotyrosine (P) in the C-terminal tail of the kinase. SH3 domain recognizes a proline rich segment (triangle) in the linker between the SH2 and kinase domains. Unique N-terminal domain is indicated with a diamond. (b) The Src family PTKs can be derepressed by dephosphorylation of the C-terminal regulatory tyrosine, by mutations in the SH3 domain, or by binding of high affinity peptides or proteins to the SH2 and SH3 domains. The kinase domain can bind to its substrate (curved shape) and phosphorylate it. (c) Dimer formation between SH2 and SH3 domains. It is not known whether the full length protein can form one or two SH2-SH3 dimers. (d) Phosphorylation of a tyrosine close to the binding site of SH3 domain prevents intramolecular binding to the kinase domain. Btk is depicted with the PH domain (square) and TH domain (arrow). (e) Increased specificity of PTKs due to simultaneous recognition for the substrate by several domains. (f) Phosphorylation of the SH2 domain leads to derepression of the kinase. (g) Intramolecular binding of the proline rich region in the TH domain into the Btk SH3 domain.
Btk SH3 domain has been determined with NMR (Hansson et al., submitted) and SH3 domain has been modeled (19). The three dimensional structures of the SH3 domains were docked to the SH2 domains by fitting the putative binding residues to the high affinity peptide of Src SH2 domain (16). The backbone of the residues pY to /3 were superimposed with the residues in the SH3 domains and the root mean square (rms) deviation was determined. Other SH2 domains, for which peptide complex structure was not available, were superimposed on top of the Src SH2 domain complex. The peptide was removed before further analysis. Final adjustment of the do-
mains in the docked pairs were made either manually fixing torsion angles of side chains or running a short energy minimization with program Discover (Molecular Simulations Inc.). In the energy minimization only the side chain atoms of the clashing residues were allowed to move. The hydrogen bonds were analysed by using program InsightII (Molecular Simulations, Inc.) which employed distance and stereochemical criteria. Atom pairs having favourable angle and distance ˚ were considered as hydrogen bond forming. Beequal or below 3 A cause this method does not take into account too closely positioned
352
AID
BBRC 7909
/
6945$$$241
12-30-97 18:26:55
bbrcg
AP: BBRC
Vol. 242, No. 2, 1998
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS TABLE 1
Analysis of SH2-SH3 Domain Dimers SH3
SH2
Number of hydrogen bonds
Total buried ˚ 2) accessible surface (A
Lcka Src Fyn Fyn Btk
Lck Src Lck Src Btk
5 22 20 19 15
1057 1712 2027 1798 1748
a
This is the dimer from crystal structure (15).
residues, the numbers of clashing residues were analysed. The accessible surfaces of the free and bound domains were determined with program Grasp (20).
SH2-SH3 DOMAIN DIMER FORMATION Binding of the SH2 and SH3 domains to their partners is important for signalling cascades. These do-
mains can be essential for localization of kinases and their synergistic binding may increase specificity of recognition. Analysis of SH2 and SH3 domains of Fyn and several other signalling molecules verified direct interactions between these modules (14). The SH3-SH2 structure of Src family member, Lck, has been determined (15). In crystals, the tandem domains form a dimer, where an SH2 domain proline is in the polyproline binding pocket of the opposing SH3 domain. One face of the triangular PP-II type helices has extensive interactions with SH3 domains (1, 2). The proline containing region of Lck SH3 domain does not have PP-II fold and only part of the SH3 domain binding pocket is occupied. Despite rather large buried accessible surface (Table 1) and shape complementarity, the dimer in Lck crystals does not resemble protein-protein interfaces. Typical interfaces have large buried surfaces, numerous hydrogen bonds and hydrophobic and other interactions (21-26). The interface in Lck has weak interactions which might have arisen from crystal packing effects. There is only five hydrogen bonds between the
FIG. 2. Dimer formation between SH2 and SH3 domains. (a) Binding of Src SH3 domain on white (17) to Src SH2 domain (16) on blue. The domains were docked on top of the high affinity SH2 domain ligand. The domains have complementary surfaces, very large buried accessible surface, and an excessive hydrogen bond network. (b) Ribbon presentation of Src SH2 (magenta) and SH3 (cyan) domains. The termini of the domains are shown, as well as the binding residues in the SH3 domain (Y92 to R95) (in yellow). Tyrosine residues shown to be phosphorylated in some kinases and affecting regulation are in red; from top to bottom Y136, Y90, and Y213. 353
AID
BBRC 7909
/
6945$$$241
12-30-97 18:26:55
bbrcg
AP: BBRC
Vol. 242, No. 2, 1998
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
domains when protein pairs have on average 10 (21) and the other studied dimers form close to 20 bonds (Table 1). Another SH2-SH3 structure has been determined for adaptor protein Grb 2 (27), which has the domains in the order SH3-SH2-SH3. Also this structure is a dimer where SH2 and SH3 domains interact, but the binding sites of the domains are accessible. The long linkers between the domains provide freedom for different orientations of the domains and for different binding modes (27). SH2 domain interactions with partners does not affect binding to the SH3 domains and vice versa (28). There are only eight hydrogen bonds between the Grb2 molecules in the dimer, although the buried accessible surface is large. Therefore the dimer in the three dimensional structure may be the result of crystallization. The binding specificity of several SH2 domains has been studied by using phosphopeptide libraries (29, 30). Sequence analysis indicated that SH3 domains of Src PTKs contain in the RT loop a sequence related to the optimal SH2 domain binding pattern. In Src the sequence is YESR (residues 92 to 95), in Fyn YEAR and in Lck YEPS. The optimal binding sequence of all Src family SH2 domains is pYEEI (29). We used this information as a starting point for docking SH3 domains into SH2 domains. Several SH2 and SH3 domains have been shown to form both physical and functional interactions (14) (Fig. 1c). The structural features of these interactions were studied by superimposing the backbone atoms of the putative binding residues of SH3 domain to amino acids pY to /3 of the high affinity peptide complex of Src SH2 domain (16) and then analysing the interactions within the dimer (Figs. 2a and b). In the docking studies the residue binding to the pY pocket was not phosphorylated, because the observed dimers are formed between unphosphorylated domains (14). The Src SH2 and SH3 domains (17) fitted well to each other (Fig. 2b). The rms deviation for the backbone atoms of residues pY to /3 in the high affinity peptide and resi˚ indicating close dues 92 to 95 in Src is only 0.85 A structural relationship. The rms deviation can be compared to the deviations of three dimensional structures determined with NMR. In NMR structure determination several structures are obtained for the same protein and their rms deviations are of the same order (31) as difference between the binding peptide stretches. For example, in the rigid domain docking of Src SH2 and SH3 domains only two regions had minor steric clashes, which could be easily resolved by adjusting flexible side chain torsion angles. No backbone adjustments were necessary. The average number of hydrogen bonds and electrostatic interactions in protein-protein recognition sites is 10 (21). In the docked Src SH2-SH3 dimer there is altogether 22 interactions indicating strong recognition
and complementarity. Hydrogen bonds have free energies in the order of 0.5-1.8 kcal/mol for neutral bonds, whereas charged or ion pair hydrogen bonds can contribute 3-6 kcal/mol (32). Thus, hydrogen bonds provide substantial stabilizing forces for the SH2-SH3 dimers, especially since the number of interactions is about two times larger than in interacting proteins in general. Another feature typical for strong protein-protein interactions is large buried accessible surface. Analysis of high resolution structures of protein complexes indi˚ 2 (16). Similar values cated a value of 1600 { 350 A have been obtained in the analysis of subunit interfaces of oligomeric proteins (33). In Src dimer the value is ˚ 2 (Table 1), which is very large when we about 1700 A take into account that SH2 and SH3 domains are relatively small compared to full size proteins. The pep˚ 2, and tide residues pY to /3 provide only about 600 A thereby the buried surface is mainly contributed by other regions of the interacting domains. The major forces favouring protein associations are hydrophobic energy from buried accessible surfaces and electrostatic energy from hydrogen bonds (21). It has been suggested that non-polar surfaces could pro˚ 2 when buried (21). It is diffivide as much as 25 cal/A cult to estimate different forces contributing to the stability of the dimers, especially in relation to entropy. Anyhow, the extensive hydrogen bonding, large buried surface area and complementarity of the surfaces with the experimental evidence for interactions are suggesting energetically highly favourable binding interactions. The dimer described in here differs markedly of that in the Lck and Grb2 crystals (15, 27). Although the pY binding cannot be the driving force since the SH3 domains are not generally phosphorylated, also the aromatic ring of the pY has extensive binding, e.g. in Src SH2 domain three residues interact with the aromatic ring of the pY (16). For some SH2 domains non-pY dependent binding has been demonstrated (34–38). In addition to the specificity determining residues, of which amino acids at positions /1 and /3 are the most important, also other regions may affect binding. As an example, the affinity of the HIV Nef protein to Src family SH3 domains is more than 300 fold higher than that for a corresponding peptide (39) due to extensive interactions outside the polyproline recognition region (40). Domains from several Src family members were shown to interact and form both homo- and heterodimers (14). The docked Src family domains have an extensive hydrogen bond network (Table 1), which with structural complementarity and large buried interface suggests highly favourable dimer formation. Also the observed heterodimers (14) were found to have matching surfaces and numerous hydrogen bonds (Table 1). The unproductive dimers of Fyn and Lck SH3 domain with Lck and p85a subunit of PI3 kinase SH2 domain (41) could be tested by docking. They were char-
354
AID
BBRC 7909
/
6945$$$241
12-30-97 18:26:55
bbrcg
AP: BBRC
Vol. 242, No. 2, 1998
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
acterized by infavourable binding sequence in the RT loop, weak hydrogen bonding and most importantly severe structural clashes when docked. The docked Lck SH2 and SH3 domains have a large buried accessible ˚ 2). There are also numerous atoms at surface (1677 A hydrogen bonding distance, but even after energy minimization several clashes remained, suggesting that these domains cannot in fact form productive dimer. This is indeed the case as shown by the experimental results (14). Thus, the SH2-SH3 dimerization requires in addition to proper sequence, structural complementarity and complementarity also outside the actual recognition site. The binding sequence in the SH3 domains is not optimal for the studied SH2 domains, although in Src family the residue at position /1, glutamate, is conserved. Src SH3 domain has R95 at position /3, when the optimal residue at this position is isoleucine. Arginine has very flexible side chain, which can bend such that the strongly basic d-guanidino group comes to surface. In the Nef-SH3 domain structure Fyn has R96I mutation in the RT loop (40). The affinity of the wild type Fyn has Kd ú20 mM (35), while the highest affinities of SH3 domain binding peptides are in the order of 10 mM (2). The full length Nef has Kd of about 0.38 mM. Thus, remarkably nonideal residue (when compared to the consensus binding sequence) is allowed at position /3 in Src and Fyn. The reduction in the affinity caused by nonideal residue is presumably compensated e. g. by complementary surface and hydrogen bonding. MUTATIONAL ANALYSIS OF DIMER FORMATION The dimer formation suggested in here could be prevented by mutations in key residues either in the SH2 or SH3 domain. If the dimerization were dependent on the phosphorylation, mutations at phosphate binding residues would prevent recognition. This seems not to be the case, because mutation R176K at pY residue phosphate binding residue of Fyn SH2 domain does not prevent dimerization (14). Also SH2 domain mutations S186P and R190C show normal domain recognition, although both these mutations are partially defective in pY binding (14). S186 locates in the b-strand B and R190 in a loop connecting two b structures. Thus, the SH2 domain can be mutated in phosphate group binding region without affecting phoshotyrosine independent binding of SH3 domain. Also mutations in the binding region of SH3 domain would hinder dimerization. In c-Src SH3 domain, Y92 and Y138 (Fig. 2a) have been shown to be phosphorylated in certain signalling pathways and to affect ligand binding (42). Y92 lies just two residues before the putative dimerization region (Y94 to R97). The corresponding residue is important also for regulation of Tec family member, Bruton’s tyrosine kinase (Btk), in which autophosphorylation of Y223 inactivates SH3
domain function (43). Phosphorylation would prevent dimerization (Fig. 1d), which according to docking of the modelled Btk SH3 (Hansson et al., submitted) and SH2 domains (20) is possible between the unphosphorylated modules (Table 1). CONCLUSIONS The phosphorylated Src family kinases are regulated by intramolecular interactions of the SH2 and SH3 domains with the kinase domain and its extension (8, 9). Increased specificity in binding to partners can be obtained by simultaneous recognition by several domains (44) (Fig. 1e). Based on the dimerization results, PTKs could be regulated by dimer formation between SH2 and SH3 domains. This would lead to the activation of the kinase domain. Further clue for the dimerization is provided by the domain organization of kinases, adaptors and other signalling molecules. The SH3 domain precedes almost without an exception SH2 domain in close proximity when both are present. Although the affinity of unphosphorylated peptide can in some cases be 104 times lower than that of a corresponding phosphorylated peptide (45), the SH2-SH3 dimerization could be important because there are other interactions and extensive buried surface. Furthermore, Shc has recently been shown to bind to the Grb2 SH2 domain regardless of the state of the tyrosine phosphorylation (45). These authors speculate that carboxylic group of aspartate could mimic the phosphate group of phosphotyrosine. According to our docking results, it is more likely that aromatic residue at pY binding pocket is sufficient for binding. The structure of the PTKs is presumably flexible due to linkers between domains. Grb2 has been suggested to have several conformations and binding modes (27). Interacting PTKs could form two of the dimers described above and still the kinase domains could be accessible. In addition to processes discussed above, the domains are involved in several mechanisms. SH2 domain phosphorylation at Y192 in Lck (47) and Y213 in Src (48) decreases affinity to ligands (Fig. 1f, 2a). These residues are outside the immediate recognition site. Another regulatory variation is provided by a proline rich region in the TH domain of Tec family PTKs, which has been shown in Itk to bind intramolecularly to the SH3 domain (49) (Fig. 1g). The proline rich region in Tec family kinases is recognized also by SH3 domains of Src family members (50-52). Thus, proteins containing SH2 and SH3 domains can be regulated in several ways possibly allowing a number of conformations involved in both intra- and intermolecular associations. ACKNOWLEDGMENTS This work has been supported by Finnish Academy, Biocentrum Helsinki, Instrumentariumin tiedesa¨a¨tio¨, Swedish Cancer Society,
355
AID
BBRC 7909
/
6945$$$241
12-30-97 18:26:55
bbrcg
AP: BBRC
Vol. 242, No. 2, 1998
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
˚ ke Wiberg Foundation, and Swedish Medical Research Council, A Magn. Bergvall Foundation.
REFERENCES 1. Pawson, T. (1995) Nature 373, 573–580. 2. Cohen, G. B., Ren, R., and Baltimore, D. (1995) Cell 80, 237– 248. 3. Cantley, L. C., Auger, K. R., Carpenter, C., Duckworth, B., Graziani, A., Kapeller, R., and Soltoff, S. (1991) Cell 64, 281–302. 4. Roussel, R. R., Brodeur, S. R., Shalloway, D., and Laudano, A. (1991) Proc. Natl. Acad. Sci. USA 88, 10696–10700. 5. Murphy, S. M., Bergman, M., and Morgan, D. O. (1993) Mol. Cell. Biol. 13, 5290–5300. 6. Okada, M., Howell, B. W., Broome, M. A., and Cooper, J. A. (1993) J. Biol. Chem. 268, 18070–18075. 7. Superti-Furga, G., Fumagalli, S., Koegl, M., Courtneidge, S. A., and Draetta, G. (1993) EMBO J. 12, 2625–2634. 8. Sicheri, F., Moarefi, I., and Kuriyan, J. (1997) Nature 385, 602– 609. 9. Xu, W., Harrison, S. C., and Eck, M. J. (1997) Nature 385, 595– 602. 10. DeClue, J. E., and Martin, G. S. (1989) J. Virol. 63, 542–554. 11. O’Brien, M. C., Fukui, Y., and Hanafusa, H. (1990) Mol. Cell. Biol. 10, 2855–2862. 12. Liu, X., Brodeur, S. R., Gish, G., Songyang, Z., Cantlye, L. C., Laudnao, A. P., and Pawson, T. (1993) Oncogene 8, 1119–1126. 13. Moarefi, I., LaFevre-Bernt, M., Sicheri, F., Huse, M., Lee, C.-h., Kuriyan, J., and Miller, T. (1997) Nature 385, 560–653. 14. Panchamoorthy, G., Fukazawa, T., Stolz, L., Payne, G., Reedquist, K., Shoelson, S., Songyang, Z., Cantley, L., Walsh, C., and Band, H. (1994) Mol. Cell. Biol. 14, 6372–6385. 15. Eck, M. J., Atwell, S. K., Shoelson, S. E., and Harrison, S. C. (1994) Nature 368, 764–769. 16. Waksman, G., Shoelson, S. E., Pant, N., Cowburn, D., and Kuriyan, J. (1993) Cell 72, 779–790. 17. Feng, S., Chen, J. K., Yu, H., Simon, J. A., and Schreiber, S. L. (1994) Science 266, 1241–1247. 18. Morton, C. J., Pugh, D. J. R., Brown, E. L. J., Kahmann, J. D., Renzoni, D. A. C., and Campbell, I. D. (1996) Structure 4, 705– 714. 19. Vihinen, M., Nilsson, L., and Smith, C. I. E. (1994) Biochem. Biophys. Res. Commun. 205, 1270–1277. 20. Nicholls, A., Sharp, K. A., and Honig, B. (1991) Proteins Struct. Funct. Genet. 11, 281–296. 21. Janin, J., and Chothia, C. (1990) J. Biol. Chem. 265, 16027– 16030. 22. Janin, J., Miller, S., and Chothia, C. (1988) J. Mol. Biol. 204, 155–164. 23. Argos, P. (1988) Protein Eng. 2, 101–113. 24. Korn, A. P., and Burnett, R. M. (1991) Proteins Struct. Funct. Genet. 9, 37–55. 25. Jones, S., and Thornton, J. (1996) Proc. Natl. Acad. Sci. USA 93, 13–20. 26. Lijnzaad, P., and Argos, P. (1997) Proteins Struct. Funct. Genet. 28, 333–343. 27. Maignan, S., Guilloteau, J.-P., Fromage, N., Arnoux, B., Becquart, J., and Ducruix, A. (1995) Science 268, 291–293.
28. Cussac, D., Frech, M., and Chardin, P. (1994) EMBO J. 13, 4011–4021. 29. Songyang, Z., Shoelson, S. E., Chaudhuri, M., Gish, G., Pawson, T., Haser, W. G., King, F., Roberts, T., Ratnofsky, S., Lechleider, R. J., Neel, B. G., Birge, R. B., Fajardo, J. E., Chou, M. M., Hanafusa, H., Schaffhausen, B., and Cantley, L. C. (1993) Cell 72, 767–778. 30. Songyang, Z., Shoelson, S. E., McGlade, J., Olivier, P., Pawson, T., Bustelo, X. R., Barbacid, M., Sabe, H., Hanafusa, H., Yi, T., Ren, R., Baltimore, D., Ratnofsky, S., Feldman, R. A., and Cantley, L. C. (1994) Mol. Cell. Biol. 14, 2777–2785. 31. Evans, J. N. S. (1996) Biomolecular NMR Spectroscopy, Oxford Univ. Press, Oxford. 32. Fersht, A. R., Shi, J. P., Knill-Jones, J., Lowe, D. M., Wilkinson, A. J., Blow, D. M., Brick, P., Carter, P., Waye, M. M. Y., and Winter, G. (1985) Nature 314, 235–238. 33. Janin, J., Miller, S., and Chothia, C. (1988) J. Mol. Biol. 204, 155–164. 34. Pendergast, A. M., Muller, A. J., Havlik, M. H., Maru, Y., and Witte, O. N. (1991) Cell 66, 161–171. 35. Muller, A. J., Pendergast, A.-M., Havlik, M. H., Puil, L., Pawson, T., and Witte, O. N. (1992) Mol. Cell. Biol. 12, 5087–5093. 36. Bibbins, K. B., Boeuf, H., and Varmus, H. E. (1993) Mol. Cell. Biol. 13, 7278–7287. 37. Cleghon, V., and Morrison, D. K. (1994) J. Biol. Chem. 269, 17749–17755. 38. Ng, D. H. W., Watts, J. D., Aebersold, R., and Johnson, P. (1996) J. Biol. Chem. 271, 1295–1300. 39. Lee, C-.H., Leung, B., Lemmon, M. A., Zheng, J., Cowburn, D., Kuriyan, J., and Saksela, K. (1995) EMBO J. 14, 5006–5015. 40. Lee, C.-H., Saksela, K., Mirza, U. A., Chait, B. T., and Kuriyan, J. (1996) Cell 85, 931–942. 41. Booker, G. W., Breeze, A. L., Downing, A. K., Panaytou, G., Gout, I., Waterfield, M. D., and Campbell, I. D. (1992) Nature 358, 684–687. 42. Broome, M. A., and Hunter, T. (1997) Oncogene 14, 17–34. 43. Park, H., Wahl, M. I., Afar, D. E. H., Turck, C. W., Rawlings, D. J., Tam, C., Scharenberg, A. M., Kinet, J.-P., and Witte, O. N. (1996) Immunity 4, 515–525. 44. Kanemitsu, M. Y., Loo, L. W. M., Simon, S., Lau, A. F., and Eckhart, W. (1997) J. Biol. Chem. 272, 22824–22831. 45. Piccione, E., Case, R. D., Domchek, S. M., Hu, P., Chaudhuri, M., Backer, J. M., Schlessinger, J., and Shoelson, S. E. (1993) Biochemistry 192, 3197–3202. 46. Rojas, M., Yao, S., Donahue, J. P., and Lin, Y.-Z. (1997) Biochem. Biophys. Res. Commun. 234, 675–680. 47. Couture, C., Songyang, Z., Jascur, T., Williams, S., Tailor, P., Cantley, L. C., and Mustelin, T. (1996) J. Biol. Chem. 271, 24880–24884. 48. Stover, D. R., Furet, P., and Lydon, N. B. (1996) J. Biol. Chem. 271, 12481–12487. 49. Andreotti, A. H., Bunnell, S. C., Feng, S., Berg, L. J., and Schreiber, S. L. (1997) Nature 385, 93–97. 50. Cheng, C., Ye, Z.-S., and Baltimore, D. (1994) Proc. Natl. Acad. Sci. USA 91, 8152–8155. 51. Alexandropoulos, K., Cheng, G., and Baltimore, D. (1995) Proc. Natl. Acad. Sci. USA 92, 3110–3114. 52. Yang, W., Malek, S. N., and Desiderio, S. (1995) J. Biol. Chem. 270, 20832–20840.
356
AID
BBRC 7909
/
6945$$$241
12-30-97 18:26:55
bbrcg
AP: BBRC