- Email: [email protected]
Towards a Complete Structure of Hsp90
Previews 501
Structure, Vol. 13, April, 2005, ©2005 Elsevier Ltd All rights reserved. DOI 10.1016/j.str.2005.03.004
Towards a Complete Structure of Hsp90 HtpG, the E. coli homolog of Hsp90, is emerging as a valuable structural model of eukaryotic Hsp90. A first insight into the domain arrangement of HtpG is presented by Huai et al. (2005) in this issue of Structure. Eukaryotic Hsp90 is a molecular chaperone of considerable medical relevance. Among its client proteins are a host of protooncogenic kinases, as well as various transcription factors involved in cell cycle control, endocrine function, apoptosis, and other signaling pathways. These proteins all depend on Hsp90 and its multiple cochaperones for proper folding and conformational regulation by mechanisms that are not yet well understood (Wegele et al., 2004). Hsp90 is a flexible, multidomain protein, and thus, obtaining its structure at atomic resolution presents a major challenge yet holds promise to provide long awaited mechanistic insights. A paper by Huai et al. (2005) in this issue of Structure reports significant progress toward this goal. Hsp90 is a dimer of elongated subunits, comprised of three major domains: an N-terminal ATP binding domain (N), a middle domain (M) that functions in substrate protein binding and in regulation of the ATPase activity, and a C-terminal dimerization domain (C). Client proteins are thought to be accommodated in a nonnative or near-native conformation between the two subunits. Upon binding of ATP, the N-terminal domains of the two subunits contact each other, causing the substrate protein to be transiently enclosed by the chaperone. In the eukaryotic system, this closed state of Hsp90 is stabilized by the accessory protein p23, and productive protein folding requires ATP hydrolysis. The N-terminal dimerization of Hsp90 is thought to resemble the dimer association of DNA Gyrase B (GyrB) and the DNA mismatch repair protein MutL, two other members of the GHKL (gyrase-Hsp90-histidine kinaseMutL) family of homodimeric ATPases (Dutta and Inouye, 2000). To date, the three separate domains of Hsp90 have been structurally characterized only by themselves (Harris et al., 2004; Meyer et al., 2003; Prodromou et al., 1997; Stebbins et al., 1997). In the case of the C-terminal domain, the E. coli Hsp90 homolog, HtpG, proved to be a suitable target for X-ray crystallography (Harris et al., 2004). HtpG appears to be somewhat less flexible than eukaryotic Hsp90 and therefore may be more accessible to crystallographic analysis. While solution of a crystal structure of full-length HtpG is already in progress (D. Agard, personal communication), Huai and colleagues report the crystal structure of a fragment of HtpG comprising the N and M domains (Huai et al., 2005). This NM fragment was analyzed in the nucleotide-free and in the ADP bound states and provides a first insight into the domain arrangement of Hsp90. Both states have very similar conformations, charac-
terized by tight interdomain interactions. The domain structures match closely those of the isolated N and M domains of eukaryotic Hsp90, except for a w50 residue long loop extension in the C terminus of the N domain, close to the interdomain linker, which is missing in HtpG. This loop segment is expendable for in vivo function in yeast (Louvion et al., 1996) and may have prevented the production of high-quality crystals of the eukaryotic NM fragment. Interestingly, the topology of the N-terminal part of the M domain is similar to the corresponding segments of GyrB and MutL, where the M domain is involved in ATP hydrolysis as well. However, the overall three-dimensional arrangement of the N and M domains varies in these structures, with different sections of the M domain approaching the nucleotide binding pocket. Differences between the HtpG structures in the apo and the ADP states are limited to the nucleotide binding site. In the presence of nucleotide, two helices flanking a previously identified “lid” segment partially unwind to avoid collision with the bound ligand. Movement of this segment upon ATP binding has been proposed to trigger the association of the N domains in the Hsp90 dimer (Prodromou et al., 1997). Importantly, the N domain by itself is not sufficient for ATP hydrolysis (Weikl et al., 2000). This has been shown for eukaryotic Hsp90 and the homologs GyrB and MutL, where residues in the N-terminal part of the M domain mediate the activation of ATP for nucleophilic attack (Ban et al., 1999; Brino et al., 2000). In the ADP bound structure of HtpG, the M domain does indeed approach the nucleotide, but a loop region that has been implicated in ATP hydrolysis in yeast Hsp90 remains well separated from the bound nucleotide (Meyer et al., 2003). Substantial rearrangements, possibly involving client protein binding, would be required for this segment to be involved in ATP hydrolysis. Clearly, complete structures of Hsp90 bound to different nucleotides will ultimately be necessary to understand the structural consequences of N domain dimerization and thus the catalytic mechanism of ATP hydrolysis. In assessing the value of HtpG as a structural model for Hsp90, it is necessary to consider the functional divergence of the two proteins that has occurred during evolution. Although the bacterial homolog shares a high degree of sequence similarity with eukaryotic Hsp90, HtpG does not depend on multiple accessory protein cofactors, such as p23 and Aha1, and it is not essential under normal growth conditions, in contrast to Hsp90. Comparative functional studies of both the bacterial and the eukaryotic Hsp90s will therefore be necessary to verify conclusions from the HtpG structure, but ultimately, complete structures of Hsp90 itself will be required. Achieving this goal may be facilitated by employing modified sequence constructs for crystallization that have been tailored based on the HtpG structure. We may then understand how client proteins bind to Hsp90 and how ATP hydrolysis and the multiple Hsp90 cofactors regulate these complex interactions.
Structure 502
Andreas Bracher and F. Ulrich Hartl Department of Cellular Biochemistry Max Planck Institute of Biochemistry Am Klopferspitz 18 82152 Martinsried Germany
Huai, Q., Wang, H., Liu, Y., Kim, H.-Y., Toft, D., and Ke, H. (2005). Structure 13, this issue, 579–590. Louvion, J.F., Warth, R., and Picard, D. (1996). Proc. Natl. Acad. Sci. USA 93, 13937–13942. Meyer, P., Prodromou, C., Hu, B., Vaughan, C., Roe, S.M., Panaretou, B., Piper, P.W., and Pearl, L.H. (2003). Mol. Cell 11, 647–658. Prodromou, C., Roe, S.M., O’Brien, R., Ladbury, J.E., Piper, P.W., and Pearl, L.H. (1997). Cell 90, 65–75.
Selected Reading Ban, C., Junop, M., and Yang, W. (1999). Cell 97, 85–97. Brino, L., Urzhumtsev, A., Mousli, M., Bronner, C., Mitschler, A., Oudet, P., and Moras, D. (2000). J. Biol. Chem. 275, 9468–9475. Dutta, R., and Inouye, M. (2000). Trends Biochem. Sci. 25, 24–28. Harris, S.F., Shiau, A.K., and Agard, D.A. (2004). Structure 12, 1087–1097.
Stebbins, C.E., Russo, A.A., Schneider, C., Rosen, N., Hartl, F.U., and Pavletich, N.P. (1997). Cell 89, 239–250. Wegele, H., Müller, L., and Buchner, J. (2004). Rev. Physiol. Biochem. Pharmacol. 151, 1–44. Weikl, T., Muschler, P., Richter, K., Veit, T., Reinstein, J., and Buchner, J. (2000). J. Mol. Biol. 303, 583–592.
Structure, Vol. 13, April, 2005, ©2005 Elsevier Ltd All rights reserved. DOI 10.1016/j.str.2005.03.004
Towards a Complete Structure of Hsp90 HtpG, the E. coli homolog of Hsp90, is emerging as a valuable structural model of eukaryotic Hsp90. A first insight into the domain arrangement of HtpG is presented by Huai et al. (2005) in this issue of Structure. Eukaryotic Hsp90 is a molecular chaperone of considerable medical relevance. Among its client proteins are a host of protooncogenic kinases, as well as various transcription factors involved in cell cycle control, endocrine function, apoptosis, and other signaling pathways. These proteins all depend on Hsp90 and its multiple cochaperones for proper folding and conformational regulation by mechanisms that are not yet well understood (Wegele et al., 2004). Hsp90 is a flexible, multidomain protein, and thus, obtaining its structure at atomic resolution presents a major challenge yet holds promise to provide long awaited mechanistic insights. A paper by Huai et al. (2005) in this issue of Structure reports significant progress toward this goal. Hsp90 is a dimer of elongated subunits, comprised of three major domains: an N-terminal ATP binding domain (N), a middle domain (M) that functions in substrate protein binding and in regulation of the ATPase activity, and a C-terminal dimerization domain (C). Client proteins are thought to be accommodated in a nonnative or near-native conformation between the two subunits. Upon binding of ATP, the N-terminal domains of the two subunits contact each other, causing the substrate protein to be transiently enclosed by the chaperone. In the eukaryotic system, this closed state of Hsp90 is stabilized by the accessory protein p23, and productive protein folding requires ATP hydrolysis. The N-terminal dimerization of Hsp90 is thought to resemble the dimer association of DNA Gyrase B (GyrB) and the DNA mismatch repair protein MutL, two other members of the GHKL (gyrase-Hsp90-histidine kinaseMutL) family of homodimeric ATPases (Dutta and Inouye, 2000). To date, the three separate domains of Hsp90 have been structurally characterized only by themselves (Harris et al., 2004; Meyer et al., 2003; Prodromou et al., 1997; Stebbins et al., 1997). In the case of the C-terminal domain, the E. coli Hsp90 homolog, HtpG, proved to be a suitable target for X-ray crystallography (Harris et al., 2004). HtpG appears to be somewhat less flexible than eukaryotic Hsp90 and therefore may be more accessible to crystallographic analysis. While solution of a crystal structure of full-length HtpG is already in progress (D. Agard, personal communication), Huai and colleagues report the crystal structure of a fragment of HtpG comprising the N and M domains (Huai et al., 2005). This NM fragment was analyzed in the nucleotide-free and in the ADP bound states and provides a first insight into the domain arrangement of Hsp90. Both states have very similar conformations, charac-
terized by tight interdomain interactions. The domain structures match closely those of the isolated N and M domains of eukaryotic Hsp90, except for a w50 residue long loop extension in the C terminus of the N domain, close to the interdomain linker, which is missing in HtpG. This loop segment is expendable for in vivo function in yeast (Louvion et al., 1996) and may have prevented the production of high-quality crystals of the eukaryotic NM fragment. Interestingly, the topology of the N-terminal part of the M domain is similar to the corresponding segments of GyrB and MutL, where the M domain is involved in ATP hydrolysis as well. However, the overall three-dimensional arrangement of the N and M domains varies in these structures, with different sections of the M domain approaching the nucleotide binding pocket. Differences between the HtpG structures in the apo and the ADP states are limited to the nucleotide binding site. In the presence of nucleotide, two helices flanking a previously identified “lid” segment partially unwind to avoid collision with the bound ligand. Movement of this segment upon ATP binding has been proposed to trigger the association of the N domains in the Hsp90 dimer (Prodromou et al., 1997). Importantly, the N domain by itself is not sufficient for ATP hydrolysis (Weikl et al., 2000). This has been shown for eukaryotic Hsp90 and the homologs GyrB and MutL, where residues in the N-terminal part of the M domain mediate the activation of ATP for nucleophilic attack (Ban et al., 1999; Brino et al., 2000). In the ADP bound structure of HtpG, the M domain does indeed approach the nucleotide, but a loop region that has been implicated in ATP hydrolysis in yeast Hsp90 remains well separated from the bound nucleotide (Meyer et al., 2003). Substantial rearrangements, possibly involving client protein binding, would be required for this segment to be involved in ATP hydrolysis. Clearly, complete structures of Hsp90 bound to different nucleotides will ultimately be necessary to understand the structural consequences of N domain dimerization and thus the catalytic mechanism of ATP hydrolysis. In assessing the value of HtpG as a structural model for Hsp90, it is necessary to consider the functional divergence of the two proteins that has occurred during evolution. Although the bacterial homolog shares a high degree of sequence similarity with eukaryotic Hsp90, HtpG does not depend on multiple accessory protein cofactors, such as p23 and Aha1, and it is not essential under normal growth conditions, in contrast to Hsp90. Comparative functional studies of both the bacterial and the eukaryotic Hsp90s will therefore be necessary to verify conclusions from the HtpG structure, but ultimately, complete structures of Hsp90 itself will be required. Achieving this goal may be facilitated by employing modified sequence constructs for crystallization that have been tailored based on the HtpG structure. We may then understand how client proteins bind to Hsp90 and how ATP hydrolysis and the multiple Hsp90 cofactors regulate these complex interactions.
Structure 502
Andreas Bracher and F. Ulrich Hartl Department of Cellular Biochemistry Max Planck Institute of Biochemistry Am Klopferspitz 18 82152 Martinsried Germany
Huai, Q., Wang, H., Liu, Y., Kim, H.-Y., Toft, D., and Ke, H. (2005). Structure 13, this issue, 579–590. Louvion, J.F., Warth, R., and Picard, D. (1996). Proc. Natl. Acad. Sci. USA 93, 13937–13942. Meyer, P., Prodromou, C., Hu, B., Vaughan, C., Roe, S.M., Panaretou, B., Piper, P.W., and Pearl, L.H. (2003). Mol. Cell 11, 647–658. Prodromou, C., Roe, S.M., O’Brien, R., Ladbury, J.E., Piper, P.W., and Pearl, L.H. (1997). Cell 90, 65–75.
Selected Reading Ban, C., Junop, M., and Yang, W. (1999). Cell 97, 85–97. Brino, L., Urzhumtsev, A., Mousli, M., Bronner, C., Mitschler, A., Oudet, P., and Moras, D. (2000). J. Biol. Chem. 275, 9468–9475. Dutta, R., and Inouye, M. (2000). Trends Biochem. Sci. 25, 24–28. Harris, S.F., Shiau, A.K., and Agard, D.A. (2004). Structure 12, 1087–1097.
Stebbins, C.E., Russo, A.A., Schneider, C., Rosen, N., Hartl, F.U., and Pavletich, N.P. (1997). Cell 89, 239–250. Wegele, H., Müller, L., and Buchner, J. (2004). Rev. Physiol. Biochem. Pharmacol. 151, 1–44. Weikl, T., Muschler, P., Richter, K., Veit, T., Reinstein, J., and Buchner, J. (2000). J. Mol. Biol. 303, 583–592.