Molecular Chaperone

26

The contribution of the two major cytosolic chaperone systems, Hsp70 and the cylindrical chaperonins, to cellular protein folding has been clarified by a number of recent papers. These studies found that, in vivo, a significant fraction of newly synthesized polypeptides transit through these chaperone systems in both prokaryotic and eukaryotic cells. The identification and characterization of the cellular substrates of chaperones will be instrumental in understanding how proteins fold in vivo.

Addresses Department of Biological Sciences, Stanford University, Stanford, CA 94305-5020, USA *e-mail: [email protected]

Current Opinion in Structural Biology 2000, 10:26–33

0959-440X/00/$ — see front matter © 2000 Elsevier Science Ltd. All rights reserved.

Abbreviations CCT chaperonin-containing TCP-1 Hsc heat-shock cognate protein Hsp heat-shock protein NAC nascent-chain associated complex TCP-1 tailless complex polypeptide-1 TF trigger factor TRiC TCP-1 ring complex

Introduction As a newborn polypeptide emerges into the world, its first contacts with the cellular environment may be critical for determining its fate. Ribosome-bound nascent polypep- tides are confronted by a unique set of dangers that must be avoided on the way to achieving a mature, native con- formation. Fortunately, a remarkable mechanism involving molecular chaperones has evolved to safeguard the folding of nascent chains. While progress has been made in under- standing the basic mechanisms of chaperone action, the contribution of chaperones to de novo cellular folding has remained poorly understood and controversial. Although chaperones are clearly important for protein folding and cellular viability, it has been argued that only a few essen- tial proteins require chaperones to fold correctly, whereas the majority of proteins fold spontaneously. An alternative possibility stems from the broad specificity of chaperone binding in vitro: as nearly every unfolded polypeptide has the potential to bind chaperones, all newly translated polypeptides might transiently associate with chaperones. A number of new studies have now addressed this problem experimentally and have begun to define the role of chap- erones in the folding of newly translated polypeptides. This review summarizes their major findings.

The folding problems of newly translated polypeptides It is generally accepted that the information necessary to specify the native three-dimensional structure of a protein is

inherent in its complete amino acid sequence [1]; however, efficient, reversible folding and unfolding is generally observed only for small proteins. Refolding experiments often lead to the formation of kinetically trapped intermedi- ates that aggregate, even in dilute aqueous solutions and at low temperature [2]. As aggregation is at least partly driven by hydrophobic interactions, it is even more pronounced when folding is attempted under the physiological condi- tions prevalent in the cell. In particular, the very high concentration of macromolecules creates conditions of crowding that highly favor aggregation (reviewed in [3]).

The folding of newly translated polypeptides faces an additional constraint, as it must be accomplished in the context of the vectorial protein synthesis process. The N-terminal portion of a nascent polypeptide could, in prin- ciple, fold spontaneously as it emerges from the ribosome, however, the cooperative nature of the interactions that stabilize folded structures requires that a complete folding domain (50–200 amino acids) be available for productive folding. Furthermore, translation occurs on a timescale of seconds (in bacteria) to several minutes (in eukaryotes), much slower than the millisecond timescale of hydropho- bic collapse. Similar dangers exist for proteins during their vectorial import into mitochondria, chloroplasts or the endoplasmic reticulum, into which polypeptide chains are translocated in an extended conformation. Although recent studies indicate that co-translational domain folding sim- plifies the folding problems encountered by multidomain proteins [4,5,6•,7,8], a growing polypeptide must still be prevented from misfolding and aggregation until a chain length suitable for productive folding has been synthe- sized. Mounting evidence now indicates that molecular chaperones interact with and stabilize nascent and translo- cating polypeptides in vivo and prevent nonproductive reactions, such as aggregation. Two major classes of ATP- dependent chaperone, the Hsp70s and the chaperonins, have been implicated in de novo protein folding in the cytosol of eukaryotic and prokaryotic cells, as well as in organelles of endosymbiotic origin, such as mitochondria and chloroplasts [9,10•,11]. Although substrate binding by both of these chaperone systems is regulated by nucleotide binding and hydrolysis, Hsp70 and the chaperonins are structurally and functionally distinct, and represent radi- cally different principles of chaperone action. The extensive studies on mechanistic aspects of these chaper- one systems have recently been summarized in several excellent reviews [9,10•,11].

The contribution of Hsp70s to de novo folding The Hsp70s, in conjunction with co-chaperones of the DnaJ/Hsp40 family, bind and release short linear peptide segments with a net hydrophobic character; such hydrophobic regions are probably present in all unfolded

Protein folding in vivo: the importance of molecular chaperones Douglas E Feldman and Judith Frydman*

SBA112.QXD 02/17/2000 01:21 Page 26

Protein folding in vivo and molecular chaperones Feldman and Frydman 27

polypeptides [9,10•,11]. Association with an Hsp70 results in the stabilization of a polypeptide in an extended confor- mation, thereby preventing its aggregation. For some model substrates, such as firefly luciferase, this is sufficient to promote folding in vitro. In many instances, however, the Hsp70-bound substrate must be transferred to a chap- eronin complex for productive folding.

A role for Hsp70 proteins in de novo folding was originally suggested by several lines of evidence. The observation both that cytoplasmic Hsp70 associated with ribosome- bound nascent chains in eukaryotic cells [4,12–15] and that mitochondrial and endoplasmic reticulum Hsp70s bound to translocating polypeptides [16,17] led to the suggestion that Hsp70s play a general role in stabilizing a translating or translocating polypeptide to prevent its premature mis- folding. Supporting this idea, genetic and biochemical studies in Saccharomyces cerevisiae demonstrated that the yeast Hsp70 homologs SSA1–4 are essential for viability [18] and assist the in vivo folding of model proteins [19•]. Furthermore, another class of yeast Hsp70, the Ssb pro- teins, associates stably with ribosomes and can be cross-linked to nascent chains [12,20••].

These experiments did not, however, identify the overall contribution of Hsp70 to de novo protein folding in vivo. This question was initially addressed using pulse-chase experiments in mammalian cells, whereby the flux of newly translated polypeptides through Hsp70 was assessed by quantitative immunoprecipitation [21••]. These experiments demonstrated that Hsp70 associates transiently with a broad spectrum of polypeptides larger than 20 kDa. Interestingly, a large fraction of these polypeptides are greater than 50 kDa in size. The size of individual domains in cytosolic proteins is approximately 25–30 kDa; hence, the substrates of Hsp70 probably include multidomain proteins that fold co-translationally. In contrast, smaller proteins may have a more limited requirement or weaker affinity for Hsp70. Maximal associ- ation with Hsp70 was observed at early chase times and only a small fraction of labeled polypeptide remained asso- ciated after 30 min chase. Interestingly, the kinetics of dissociation varied for different substrates, implying that folding of some proteins may require multiple cycles of binding and release. Quantitative analysis indicated that approximately 15–20% of newly synthesized proteins tran- sit through Hsp70 during their biogenesis; however, this is probably an underestimate, as the stringency of the co- immunoprecipitation method does not allow detection of weakly bound or rapidly dissociated substrates.

Early studies of the major bacterial Hsp70, DnaK, did not support a direct role in chaperoning nascent chains. ∆dnaK strains are viable, albeit heat-sensitive, indicating that this chaperone is dispensable for normal growth [22]. Furthermore, their viability does not arise from a functional overlap with another bacterial Hsp70 homolog, HscA, as the doubly deleted strain is also viable [23•]. These findings

called into question the proposal that Hsp70s play an essen- tial and evolutionarily conserved role in the folding of newly synthesized proteins; however, a direct role for DnaK in chap- eroning bacterial nascent chains has now been established [24••,25••]. Pulse-chase analysis indicated that DnaK inter- acts transiently with newly synthesized polypeptides over a broad size range, from 14 kDa to well over 90 kDa, binding preferentially to chains ranging from 30 to 75 kDa. Overall, approximately 10% of all soluble polypeptides are associated with DnaK at the earliest chase times and the bulk of these proteins dissociated within 2 min. The association of DnaK with nascent chains was examined by taking advantage of the fact that puromycin-released nascent chains become C-termi- nally tagged with puromycin and, hence, may be co-immunoprecipitated with antipuromycin antibody. At least 20% of DnaK-bound polypeptides could be reprecipi- tated using antipuromycin antibody [24••]. This finding confirms the co-translational interaction of Hsp70 with nascent chains in Escherichia coli and argues for a general role of Hsp70 in preventing protein misfolding at the ribosome.

If DnaK does indeed associate with nascent chains, why are cells unaffected in its absence? Only one other chaper- one component, the trigger factor (TF) protein, is known to bind nascent chains in E. coli [26]. The functional sig- nificance of this interaction was also unclear, as cells lacking TF (∆tig) are also viable [27]. The absence of TF results in a 2–3-fold increase in the amount of polypeptide associated with DnaK, suggesting that TF and DnaK cooperate in chaperoning nascent chains [24••,25••]. This functional overlap resonates with results from genetic crosses indicating that ∆tig and ∆dnaK are synthetically lethal. In the double-mutant strains, both newly synthe- sized and pre-existing proteins aggregated, with cytosolic proteins appearing to be most susceptible [25••]. These studies indicate that, together, DnaK and TF constitute an essential system for ensuring the productive folding of a substantial fraction of proteins in bacteria. An interesting lesson provided by these studies is that the chaperone sys- tem that can functionally replace DnaK in vivo is not an Hsp70 homolog, but is an altogether different class of ‘small’ chaperone. Future studies comparing the substrate binding motifs recognized by both chaperones, as well as their mechanisms of release of bound substrates, may clar- ify how TF and DnaK can bind to and promote the folding of the same protein subset in vivo.

The role of Hsp70 in de novo folding appears to be con- served in evolution. However, a comparison of eukaryotic and prokaryotic Hsp70 function reveals that nascent chains in the eukaryotic system remain bound to Hsp70 for longer than in bacteria, with a half-time of dissociation of approxi- mately 10 min. The greater proportion of nascent polypeptides associated with Hsp70, coupled with the decreased dissociation time, implies a more prominent role for Hsp70s in eukaryotic protein folding. Although eukary- otic homologs of TF have not been described, it is, in principle, possible that yet-to-be-identified component(s),

SBA112.QXD 02/17/2000 01:21 Page 27

such as the nascent-chain associated complex (NAC) [28], can partially replace or cooperate with Hsp70 in stabilizing nascent chains in eukaryotes. The recently described pre- foldin/GimC complex [29•,30•] has been proposed to fulfill a similar function in stabilizing newly translated actin [31•]; however, another study indicates that this complex acts at a later, post-translational stage in the folding pathway and assists chaperonin-mediated folding [32••]. Thus, the exact function of GimC remains a subject for future investigation.

The contribution of chaperonin complexes to de novo folding The chaperonins are large cylindrical protein complexes consisting of two stacked rings of seven to nine subunits each [10•,11]. Group I chaperonins, such as GroEL from E. coli and Hsp60 in mitochondria and chloroplasts, func- tion in conjunction with a ring-shaped cofactor, GroES or Hsp10, respectively, which forms the lid on a cage in which polypeptide substrates are enclosed during folding [10•,11]. In contrast, such a cofactor has not been found for the distantly related group II chaperonins from archaea and eukarya. The chaperonin of the eukaryotic cytosol, termed TRiC or CCT (for TCP-1 ring complex or chaper- onin-containing TCP-1, respectively, where TCP-1 is tailless complex polypeptide-1), also forms a cage-like structure, but it is hetero-oligomeric, containing eight dif- ferent subunits per ring (reviewed in [33,34•]). Unlike Hsp70s, chaperonins appear to interact with nonlinear hydrophobic determinants exposed in compact folding intermediates [4,35•,36].

Early studies of Hsp60 function in mitochondria and chloro- plasts suggested that chaperonins play an important role in mediating protein folding and assembly. Estimates of the contribution of the bacterial chaperonin GroEL to folding have ranged from barely 2–4% of cellular proteins [37] to approximately 30% [38]. Experiments directly analyzing the flux of newly synthesized proteins through GroEL indicat- ed that it transiently associates with approximately 12% of all newly synthesized proteins; this figure increases 2–3-fold during heat shock [39]. The majority of these substrates range between 10 and 55 kDa and are enriched for a specif- ic subset of approximately 300 polypeptides [40••]. Given the size constraints estimated for the central cavity of GroEL, the upper-size limit observed for physiological sub- strates is remarkably consistent with polypeptide folding within the cavity. Perhaps most dramatically, overexpression of GroEL increases the fraction of chaperonin-bound polypeptides, but does not change the overall distribution of substrates. This implies that the cellular concentration of GroEL is normally limited to permit only a fraction of avail- able substrates to transit through the chaperonin [39]. Several associated proteins continue to interact with GroEL throughout the course of their lifetime, indicating that, in addition to folding, the chaperonin may also play an impor- tant role in the structural maintenance of mature cellular proteins. Interestingly, structural analysis of over 50 natural GroEL substrates revealed a significant preference for pro-

teins composed of multiple α/β domains [40••]. As β sheets are assembled from discontinuous regions of the polypep- tide, the binding of these hydrophobic surfaces to GroEL might facilitate the correct packing of strands within the β sheet, as well as the packing of α helices against neigh- boring β sheets.

The role of the yeast mitochondrial chaperonin system in protein folding was also recently examined, using temper- ature-sensitive alleles of both Hsp60 and the co-chaperonin Hsp10 [41•]. As previously observed for GroEL, loss of Hsp60 results in a pronounced increase in the aggregation of a wide range of mitochondrial compo- nents. Interestingly, the subsets of proteins aggregated in Hsp10 and Hsp60 mutants were not identical, suggesting that some polypeptides may only require the assistance of Hsp60 for folding.

Despite its similarity to bacterial chaperonins, the sub- strate spectrum of the eukaryotic cytosolic chaperonin TRiC/CCT has been a matter of controversy. Primarily on the basis of the analysis of TRiC/CCT mutants in S. cere- visiae, which exhibit cytoskeletal defects characteristic of defective actin and tubulin function, it has been suggested that TRiC is a specialized chaperone that folds only a few cytoskeletal proteins [42]. In contrast, direct examination of the substrate spectrum of TRiC/CCT using pulse-chase analysis in mammalian cells demonstrated that 9–15% of newly synthesized proteins transit through the chaperonin [21••]. As observed for Hsp70 and GroEL, the dissociation kinetics from TRiC varied for different proteins, suggest- ing a differential requirement for cycles of binding and release. Interestingly, most TRiC-bound proteins were between 30 and 60 kDa in size. The restricted size range observed for cellular TRiC substrates bears parallels to the studies of GroEL substrates and lends further support to the idea that chaperonin-mediated folding occurs within an enclosed central cavity [33,34•]. Nonetheless, several large proteins of 100–120 kDa also transit through the chaperonin, raising the possibility of domain-wise folding of larger proteins by TRiC. Analysis of TRiC-associated substrates on two-dimensional gels identified at least 70 distinct substrate polypeptides. The identity and structur- al features that characterize cellular TRiC substrates remain to be defined; however, studies using model pro- teins have expanded the list of known TRiC substrates to include, in addition to actin and tubulin-related proteins, luciferase [4], G alpha transducin [43], cyclin E [44] and myosin [45]. On the basis of the structure of these known examples, TRiC substrates may have a complex domain organization that results in folding intermediates with a higher tendency to aggregate; alternatively, they may share a requirement for binding to either a cofactor or an oligomeric partner in order to complete folding. Given that most of the heterogeneity among TRiC subunits resides in the putative substrate-binding domain [33,34•], it is possi- ble that different subunits in the complex have evolved to recognize different motifs in substrate proteins.

28 Folding and binding

SBA112.QXD 02/17/2000 01:21 Page 28

Networks, pathways and the organization of chaperone action in the cell Recent years have witnessed a spirited debate concerning the extent of functional integration among the various chaperone systems in the cell [46,47]. Two models have been proposed to describe how chaperones mediate de novo folding [43,48–50]. According to one, the folding of newly synthesized proteins is a highly coordinated process involving the sequential and processive interaction of dif- ferent chaperone systems with folding intermediates [49,50]. The alternative model holds that chaperones inter- act with substrate proteins in a stochastic manner and that non-native folding intermediates partition freely through the cytosol, cycling between a network of available chap- erones and the machinery for proteolytic degradation [43,51]. Because a small fraction of the polypeptides reach the native state in each cycle, a major difference between the models is that, according to the partitioning model, non-native folding intermediates are fully discharged into the bulk cytosol multiple times before reaching the native state [43]. In contrast, the processive model proposes that the newly translated polypeptide is released into the bulk cytosol once it has adopted a conformation that is commit- ted to fold. To discriminate between these models, the processivity of de novo folding was examined in both yeast and mammalian cells by introducing a GroEL mutant (D87K GroEL) that acts as a trap for non-native folding intermediates [21••,32••]. D87K GroEL binds promiscu- ously to non-native proteins and is unable to release them (reviewed in [10•,11]). Indeed, when expressed in the cytosol of yeast or mammalian cells, D87K GroEL was fully capable of binding stress-denatured proteins, as well as newly translated polypeptides that were unable to fold. However, the D87K GroEL trap was unable to bind to the folding intermediates generated during protein synthesis, which associated instead with the endogenous cytoplasmic chaperones. These experiments support the view that folding in vivo is mediated by a highly organized chaper- one machinery that is functionally coupled to translation. They also suggest that the mechanisms that determine the fate of misfolded or stress-denatured proteins involve the cycling of non-native forms between cellular components and the cytosol, as proposed by the partitioning model.

At a mechanistic level, the coupling of folding and transla- tion (or translocation) might be accomplished by the specific recruitment of chaperone components to either the translation machinery or the translocation machinery. For instance, TF is directly associated with bacterial ribo- somes [26]. Hsp70 binding to substrates appears to be governed by association with proteins carrying the charac- teristic ‘J-domain’, which functions as a recruitment site for Hsp70 [52]. DnaJ family chaperone proteins contain additional domains that serve as localization signals, which target the various DnaJ homologs to a particular subcellu- lar location or organelle. These include TIM44, a component of the mitochondrial import machinery [53], and Sec63, a component of the endoplasmic reticulum

translocon [17]. In the eukaryotic cytosol, potential candi- dates for recruiting Hsp70 to bind nascent chains include Hsp40 [4], the J-domain protein zuotin, which also con- tains a charged region essential for ribosome association [54•], and NAC [28]. Yet another recruitment mechanism appears to be functional in chloroplasts, in which IAP100, a component of the translocation machinery, directly recruits Hsp60 [55].

The sequential nature of chaperone interactions in vivo was originally suggested by experiments that examined the folding of model proteins either imported into mito- chondria or chloroplasts [16,56,57], or translated in cell-free extracts [4], as well as by experiments using puri- fied chaperone components [58,59]. In these systems, the polypeptide was initially bound and stabilized by Hsp70, and subsequently transferred to a chaperonin. The recent examination of chaperone–substrate interactions in vivo is consistent with the sequential interaction model [24••]. Analysis of the transit of newly made polypeptides through bacterial chaperones indicated that overexpression of GroEL increases the flux of substrates through DnaK, as expected if the chaperonin is downstream in the folding pathway. Notably, TF, which functionally replaces DnaK in ∆dnaK strains, also appears to cooperate with GroEL in substrate binding [60]. It is thus possible that the cell has evolved redundant pathways of polypeptide transfer from ‘small chaperones’ (i.e. Hsp70 and TF) to chaperonins. It is not clear how substrate polypeptides are transferred among chaperone systems. It is possible that different con- formations of the substrate occur along the folding pathway and are specifically recognized by different chap- erones; however, it is also possible that adaptor proteins or direct interactions among the chaperones themselves bridge the transfer reaction.

An emerging model for chaperone action in vivo Through these studies, a more coherent picture of how proteins fold in vivo is now beginning to emerge. Despite important differences between prokaryotic and eukaryotic protein folding, such as their ability to promote co-transla- tional folding [5], there are also striking parallels between the two kingdoms (Figure 1). Quantitative analysis of chaperone interactions revealed that a large fraction of newly translated proteins flux through the major chaper- one systems in the cell. Newly translated polypeptides interact first with so-called ‘small chaperones’, including Hsp70 and TF (Figure 1a). The ability of these chaper- ones to prevent aggregation is probably sufficient to promote the folding of a large subset of polypeptides; how- ever, a considerable number of polypeptides also require the protected folding environment provided by the central cavity of prokaryotic and eukaryotic chaperonin complexes (Figure 1b). Perhaps these proteins have a more complex, aggregation-prone domain structure that requires exten- sive interactions among noncontiguous regions. Most chaperonin substrates are medium-sized proteins, between 25 and 60 kDa. This observed size distribution suggests

Protein folding in vivo and molecular chaperones Feldman and Frydman 29

SBA112.QXD 02/17/2000 01:21 Page 29

30 Folding and binding

Figure 1

SBA112.QXD 02/17/2000 01:21 Page 30

that very small proteins do not need the protected envi- ronment of the chaperonin cavity to fold. Conversely, large proteins too large to fit are probably composed of smaller individual domains that can fold co-translationally. The lack of a GroES-like cofactor in eukaryotes might allow the co-translational binding of one domain to the chaperonin (Figure 1c); this might be the case for firefly luciferase, whose N-terminal domain folds co-translationally [4], and for myosin, whose N-terminal motor domain also associ- ates co-translationally with the chaperonin [45]. Although the sequential interaction of newly synthesized polypep- tides with small and large chaperones has been observed in both prokaryotes and eukaryotes, and is possibly mediated by direct interactions (Figure 1d), it is possible that some proteins bind directly to the chaperonins.

An interesting corollary of these studies is that a substan- tial fraction of cellular proteins appear to fold without the assistance of either Hsp70 or the chaperonins (Figure 1e). How do these proteins reach the native state? The folding of specific subsets of cytosolic proteins may occur in an unassisted manner or may be carried out by novel, unchar- acterized chaperone systems. For example, Hsp90 does not appear to play a general role in de novo folding [61], but is required for folding a restricted class of proteins that includes steroid hormone receptors and Src-like tyrosine kinases (reviewed in [62]). Interestingly, these substrates are also reported to require a sequential interaction with Hsp70 prior to transfer to Hsp90 (reviewed in [63]). In addition, the translational machinery itself may also possess some chaperone-like functions, such as prevention of aggregation [64,65].

What determines whether the folding of a certain protein requires chaperone assistance? The in vivo analysis of pro- tein folding indicates that intermediates with exposed hydrophobic surfaces are not released into the bulk cytosol, except under stress conditions. Thus, it is proba- ble that if a newly translated polypeptide exposes hydrophobic surfaces it will be targeted to the chaperone machinery. In contrast, small proteins with rapid folding kinetics, as well as proteins consisting of small domains that form co-translationally, may not engage in stable or detectable interactions with cytosolic chaperones.

Perspectives and future directions Studies defining the role of the Hsp70s and chaperonins in the folding of a large fraction of cellular proteins raise crit- ical questions stemming from the discrepancy between the

substrate repertoire observed in vivo and in vitro. In vitro, both GroEL and Hsp70 interact promiscuously with most unfolded proteins. The observation that only a discrete fraction of the large constellation of cellular polypeptides actually interact in vivo with either of these chaperones raises the question of how this specificity is achieved. Importantly, this may be determined, in part, by the con- formation adopted by nascent polypeptides emerging from the ribosome. Thus, an important area of research will be to understand how co-translational folding events influ- ence the folding pathway of proteins.

The identification and characterization of in vivo chaper- one substrates may be a prerequisite for a better understanding of folding processes in the cell. This will be a challenging task and will probably require the use of global proteomics approaches. In the answer, however, may lie the fundamental rules of protein folding in the cell, with their staggering implications for our understanding of protein regulation under normal conditions and in the gen- eration of disease.

Note added in proof Four recent publications [66••,67••,68•,69••] represent important advances in our understanding of chaperone function.

References and recommended reading Papers of particular interest, published within the annual period of review, have been highlighted as:

• of special interest •• of outstanding interest

1. Anfinsen CB: Principles that govern the folding of protein chains. Science 1973, 181:223-230.

2. Jaenicke R: Protein self-organization in-vitro and in-vivo: partitioning between physical biochemistry and cell biology. Biol Chem 1998, 379:237-243.

3. Ellis RJ: Molecular chaperones: avoiding the crowd. Curr Biol 1997, 7:531-533.

4. Frydman J, Nimmesgern E, Ohtsuka K, Hartl FU: Folding of nascent polypeptide chains in a high molecular mass assembly with molecular chaperones. Nature 1994, 370:111-117.

5. Netzer W, Hartl FU: Recombination of protein domains facilitated by co-translational folding in eukaryotes. Nature 1997, 388:343-349.

6. Frydman J, Erdjument-Bromage H, Tempst P, Hartl FU: Co- • translational domain folding as the structural basis for the rapid

de novo folding of firefly luciferase. Nat Struct Biol 1999, 6:697-705. This study used a multidomain protein to demonstrate that folding during trans- lation and refolding from denaturant occur through different folding pathways.

7. Nicola AV, Chen W, Helenius A: Co-translational folding of an alphavirus capsid protein in the cytosol of living cells. Nat Cell Biol 1999, 1:341-345.

Protein folding in vivo and molecular chaperones Feldman and Frydman 31

Figure 1 legend

Schematic representation of de novo protein folding in the cytosol of prokaryotic and eukaryotic cells. The model emphasizes the evolutionarily conserved characteristics of the folding process; however, some aspects are specific to either prokaryotic or eukaryotic cells. For instance, co-translational domain folding, as well as association of the

chaperonin complex with nascent chains, is favored in eukaryotes. Conversely, no homolog of TF has been identified in eukaryotes (although several candidates exist). For simplicity, cofactors of Hsp70 and chaperonin, and alternative folding pathways involving other chaperones (e.g. Hsp90) are not represented. See text for details.

SBA112.QXD 02/17/2000 01:21 Page 31

8. Fedorov AN, Baldwin TO: Contribution of cotranslational folding to the rate of formation of native protein structure. Proc Natl Acad Sci USA 1995, 92:1227-1231.

9. Gething MJ, Sambrook J: Protein folding in the cell. Nature 1992, 355:33-45.

10. Bukau B, Horwich AL: The Hsp70 and Hsp60 chaperone machines. • Cell 1998, 92:351-366. An excellent review of the mechanism and function of the Hsp70 and chap- eronin systems.

11. Hartl FU: Molecular chaperones in cellular protein folding. Nature 1996, 381:571-579.

12. Nelson RJ, Ziegelhoffer T, Nicolet C, Werner-Washburne M, Craig EA: The translation machinery and 70 kd heat shock protein cooperate in protein synthesis. Cell 1992, 71:97-105.

13. Beckmann RP, Mizzen LA, Welch WJ: Interaction of Hsp 70 with newly synthesized proteins: implications for protein folding and assembly. Science 1990, 248:850-854.

14. Beck SC, De Maio A: Stabilization of protein synthesis in thermotolerant cells during heat shock. Association of heat shock protein-72 with ribosomal subunits of polysomes. J Biol Chem 1994, 269:21803-21811.

15. Hansen WJ, Lingappa VR, Welch WJ: Complex environment of nascent polypeptide chains. J Biol Chem 1994, 269:26610-26613.

16. Manning-Krieg U, Scherer PE, Schatz G: Sequential action of mitochondrial chaperones in protein import into the matrix. EMBO J 1991, 10:3273-3280.

17. Brodsky JL, Schekman R: A Sec63p-BiP complex from yeast is required for protein translocation in a reconstituted proteoliposome. J Cell Biol 1993, 123:1355-1363.

18. Werner-Washburne M, Stone DE, Craig EA: Complex interactions among members of an essential subfamily of hsp70 genes in Saccharomyces cerevisiae. Mol Cell Biol 1987, 7:2568-2577.

19. Kim S, Schilke B, Craig EA, Horwich AL: Folding in vivo of a newly • translated yeast cytosolic enzyme is mediated by the SSA class

of cytosolic yeast Hsp70 proteins. Proc Natl Acad Sci USA 1998, 95:12860-12865.

This study confirms the important role of the yeast Hsp70 Ssa in the folding of at least some cytosolic proteins by showing that Ssa2 is required for ornithine transcarbamoylase biogenesis in vivo.

20. Pfund C, Lopezhoyo N, Ziegelhoffer T, Schilke BA, Lopezbuesa P, •• Walter WA, Wiedmann M, Craig EA: The molecular chaperone Ssb

from Saccharomyces-cerevisiae is a component of the ribosome nascent chain complex. EMBO J 1998, 17:3981-3989.

This study demonstrates, through the use of a photoactivatable cross-linker incorporated into the nascent polypeptide, that the yeast Hsp70 Ssb asso- ciates directly with ribosome-bound polypeptide chains.

21. Thulasiraman V, Yang CF, Frydman J: In vivo newly translated •• polypeptides are sequestered in a protected folding environment.

EMBO J 1999, 18:85-95. This study presents the first assessment of the contribution of the two major cytosolic chaperone systems, Hsc70 and the cytosolic chaperonin TRiC, to de novo folding in mammalian cells. The study also provides evidence of the processive nature of cellular folding. Thus, the overexpression of a GroEL trap in the mammalian cytosol has no effect on cell growth and fails to bind to newly synthesized polypeptides, leading to the idea that folding interme- diates are not partitioned freely in the bulk cytosol but, rather, undergo fold- ing in a sequestered environment. A similar conclusion was reached by Siegers et al. [32••] using yeast cells.

22. Paek KH, Walker GC: Escherichia coli dnaK null mutants are inviable at high temperature. J Bacteriol 1987, 169:283-290.

23. Hesterkamp T, Bukau B: Role of the DnaK and HscA homologs of • Hsp70 chaperones in protein folding in E. coli. EMBO J 1998,

17:4818-4828. This paper investigates the possibility that DnaK cooperates with a closely related protein, Hsc66, and finds that bacterial strains harboring deletions of both genes are viable.

24. Teter SA, Houry WA, Ang D, Tradler T, Rockabrand D, Fischer G, •• Blum P, Georgopoulos C, Hartl FU: Polypeptide flux through

bacterial Hsp70: DnaK cooperates with trigger factor in chaperoning nascent chains. Cell 1999, 97:755-765.

See annotation to [25••].

25. Deuerling E, Schulze-Specking A, Tomoyasu T, Mogk A, Bukau B: •• Trigger factor and DnaK cooperate in folding of newly synthesized

proteins. Nature 1999, 400:693-696. This important paper, together with [24••], describes the in vivo role of DnaK in chaperoning nascent chains and characterizes the substrate repertoire of DnaK. In addition, these studies reveal that DnaK cooperates with trigger factor (TF) in de novo protein folding. The overexpression of GroEL decreased the transit time of substrates on DnaK, implying directionality in the transfer of folding substrates from DnaK to GroEL

26. Stoller G, Rucknagel KP, Nierhaus KH, Schmid FX, Fischer G, Rahfeld JU: A ribosome-associated peptidyl-prolyl cis/trans isomerase identified as the trigger factor. EMBO J 1995, 14:4939-4948.

27. Guthrie B, Wickner W: Trigger factor depletion or overproduction causes defective cell division but does not block protein export. J Bacteriol 1990, 172:5555-5562.

28. Wang S, Sakai H, Wiedmann M: NAC covers ribosome-associated nascent chains thereby forming a protective environment for regions of nascent chains just emerging from the peptidyl transferase center. J Cell Biol 1995, 130:519-528.

29. Geissler S, Siegers K, Schiebel E: A novel protein complex • promoting formation of functional alpha- and gamma-tubulin.

EMBO J 1998, 17:952-966. This paper reports the identification of yeast GimC, a novel chaperone complex.

30. Vainberg IE, Lewis SA, Rommelaere H, Ampe C, Vandekerckhove J, • Klein HL, Cowan NJ: Prefoldin, a chaperone that delivers unfolded

proteins to cytosolic chaperonin. Cell 1998, 93:863-873. An important paper that describes the purification and characterization of prefoldin, a heterohexameric complex that binds to denatured actin and tubulin, and can deliver these proteins to the chaperonin TRiC for productive folding, even under conditions in which an excess of trap chaperonin is also present. Prefoldin is the mammalian homolog of yeast GimC, which was independently discovered by Geissler et al. [29•].

31. Hansen WJ, Cowan NJ, Welch WJ: Prefoldin-nascent chain • complexes in the folding of cytoskeletal proteins. J Cell Biol 1999,

145:265-277. An examination of the chaperone associations of actin during in vitro trans- lation. The authors observe the association of nascent actin chains of more than 145 amino acids with prefoldin/GimC, an interaction that occurs prior to transfer to the chaperonin complex. A similar maturation pathway is oper- ative for tubulin, suggesting that GimC/prefoldin delivers substrates to the chaperonin. This proposal contrasts with the findings of Siegers et al. [32••].

32. Siegers K, Waldmann T, Leroux MR, Grein K, Shevchenko A, •• Schiebel E, Hartl FU: Compartmentation of protein folding in vivo:

sequestration of non-native polypeptide by the chaperonin-GimC system. EMBO J 1999, 18:75-84.

A significant study that adopts a similar approach to that of Thulasiraman et al. [21••], assessing the effects of trap GroEL expression on growth and protein folding in the budding yeast S. cerevisiae. Again, cell growth is unaf- fected by trap GroEL and most newly made polypeptides are not detectably bound by the trap; however, mutants of TRiC or its putative co-chaperone GimC do cause many newly made polypeptides to be exposed to the trap. These important findings also suggest that GimC and TRiC make essential contributions to the compartmentation of protein folding in the eukaryotic cytosol. Kinetic analysis of actin folding and chaperone interactions indicate that GimC acts on or after TRiC in the folding pathway.

33. Kubota H, Hynes G, Willison K: The chaperonin containing T- complex polypeptide-1 (Tcp-1): multisubunit machinery assisting in protein-folding and assembly in the eukaryotic cytosol. Eur J Biochem 1995, 230:3-16.

34. Gutsche I, Essen LO, Baumeister W: Group II chaperonins: new • TRiC(k)s and turns of a protein folding machine. J Mol Biol 1999,

293:295-312. An up-to-date review of current knowledge of group II chaperonin function.

35. Rommelaere H, De Neve M, Melki R, Vandekerckhove J, Ampe C: The • cytosolic class II chaperonin CCT recognizes delineated

hydrophobic sequences in its target proteins. Biochemistry 1999, 38:3246-3257.

This paper describes a deletion analysis to identify chaperonin-binding determinants of actin. The authors found that three separate regions in actin are required for binding to TRiC/CCT. This study provides support for the idea the interaction of the actin substrate with this chaperonin is polyvalent and occurs through multiple contacts.

36. Hayer-Hartl MK, Ewbank JJ, Creighton TE, Hartl FU: Conformational specificity of the chaperonin GroEL for the compact folding intermediates of alpha-lactalbumin. EMBO J 1994, 13:3192-3202.

37. Lorimer GH: A quantitative assessment of the role of the chaperonin proteins in protein folding in vivo. FASEB J 1996, 10:5-9.

32 Folding and binding

SBA112.QXD 02/17/2000 01:21 Page 32

38. Horwich AL, Low KB, Fenton WA, Hirshfield IN, Furtak K: Folding in vivo of bacterial cytoplasmic proteins: role of GroEL. Cell 1993, 74:909-917.

39. Ewalt K, Hendrick JP, Houry WA, Hartl FU: In-vivo observation of polypeptide flux through the bacterial chaperonin system. Cell 1997, 90:491-500.

40. Houry WA, Frishman D, Eckerskorn C, Lottspeich F, Hartl FU: •• Identification of in vivo substrates of the chaperonin GroEL.

Nature 1999, 402:147-154. A remarkable and detailed examination of the in vivo substrates of GroEL, including the identification by mass spectroscopy of over 50 abundant endogenous substrates. These proteins do not share a common consensus sequence, but instead are significantly enriched for a specific multiple α/β/α domain architecture. Another intriguing finding of this report is the identifi- cation of a number of pre-existing proteins that return continually to GroEL throughout the course of their lifetime.

41. Dubaquie Y, Looser R, Funfschilling U, Jeno P, Rospert S: • Identification of in vivo substrates of the yeast mitochondrial

chaperonins reveals overlapping but non-identical requirement for hsp60 and hsp10. EMBO J 1998, 17:5868-5876.

An analysis of yeast mitochondrial Hsp60 and Hsp10 function in protein folding using loss-of-function alleles in each gene. Interestingly, similar, but nonidentical, proteins became aggregated owing to loss of function in either Hsp60 or Hsp10.

42. Lewis SA, Tian G, Vainberg IE, Cowan NJ: Chaperonin-mediated folding of actin and tubulin. J Cell Biol 1996, 132:1-4.

43. Farr GW, Scharl EC, Schumacher RJ, Sondek S, Horwich AL: Chaperonin-mediated folding in the eukaryotic cytosol proceeds through rounds of release of native and nonnative forms. Cell 1997, 89:927-937.

44. Won KA, Schumacher RJ, Farr GW, Horwich AL, Reed SI: Maturation of human cyclin E requires the function of eukaryotic chaperonin CCT. Mol Cell Biol 1998, 18:7584-7589.

45. Srikakulam R, Winkelmann DA: Myosin II folding is mediated by a molecular chaperonin. J Biol Chem 1999, 274:27265-27273.

46. Johnson JL, Craig EA: Protein folding in vivo: unraveling complex pathways. Cell 1997, 90:201-204.

47. Ellis RJ: Molecular chaperones: pathways and networks. Curr Biol 1999, 9:137-139.

48. Bukau B, Hesterkamp T, Luirink J: Growing-up in a dangerous environment: a network of multiple targeting and folding pathways for nascent polypeptides in the cytosol. Trends Cell Biol 1996, 6:480-486.

49. Ellis RJ, Hartl FU: Protein folding in the cell: competing models of chaperonin function. FASEB J 1996, 10:20-26.

50. Frydman J, Hartl FU: Principles of chaperone-assisted protein folding: differences between in vitro and in vivo mechanisms. Science 1996, 272:1497-1502.

51. Buchberger A, Schroder H, Hesterkamp T, Schonfeld HJ, Bukau B: Substrate shuttling between the DnaK and GroEL systems indicates a chaperone network promoting protein folding. J Mol Biol 1996, 261:328-333.

52. Kelley WL: The J-domain family and the recruitment of chaperone power. Trends Biochem Sci 1998, 23:222-227.

53. Voos W, von Ahsen O, Muller H, Guiard B, Rassow J, Pfanner N: Differential requirement for the mitochondrial Hsp70-Tim44 complex in unfolding and translocation of preproteins. EMBO J 1996, 15:2668-2677.

54. Yan W, Schilke B, Pfund C, Walter W, Kim S, Craig EA: Zuotin; a • ribosome-associated DnaJ molecular chaperone. EMBO J 1998,

17:4809-4817. This paper describes a new member of the J-domain class of proteins, zuotin, which is ribosome associated. As zuotin deletion strains display phe- notypes identical to mutants of Ssb, a ribosome-bound member of the Hsp70 family, the authors postulate that zuotin may play a role in modulating the activity or substrate binding of its putative partner Ssb proteins.

55. Kessler F, Blobel G: Interaction of the protein import and folding machineries of the chloroplast. Proc Natl Acad Sci USA 1996, 93:7684-7689.

56. Heyrovska N, Frydman J, Hohfeld J, Hartl FU: Directionality of polypeptide transfer in the mitochondrial pathway of chaperone- mediated protein-folding. Biol Chem 1998, 379:301-309.

57. Tsugeki R, Nishimura M: Interaction of homologues of Hsp70 and Cpn60 with ferredoxin-NADP+ reductase upon its import into chloroplasts. FEBS Lett 1993, 320:198-202.

58. Langer T, Lu C, Echols H, Flanagan J, Hayer MK, Hartl FU: Successive action of DnaK, DnaJ and GroEL along the pathway of chaperone-mediated protein folding. Nature 1992, 356:683-689.

59. Petit MA, Bedale W, Osipiuk J: Sequential folding of UmuC by the Hsp70 and Hsp60 chaperone complexes of Escherichia coli. J Biol Chem 1994, 269:23824-23829.

60. Kandror O, Sherman M, Moerschell R, Goldberg AL: Trigger factor associates with GroEL in vivo and promotes its binding to certain polypeptides. J Biol Chem 1997, 272:1730-1734.

61. Nathan DF, Vos MH, Lindquist S: In-vivo functions of the Saccharomyces cerevisiae Hsp90 chaperone. Proc Natl Acad Sci USA 1997, 94:12949-12956.

62. Caplan AJ: Hsp90’s secrets unfold: new insights from structural and functional studies. Trends Cell Biol 1999, 9:262-268.

63. Frydman J, Hohfeld J: Chaperones get in touch: the Hip-Hop connection. Trends Biochem Sci 1997, 22:87-92.

64. Kudlicki W, Coffman A, Kramer G, Hardesty B: Ribosomes and ribosomal-RNA as chaperones for folding of proteins. Fold Des 1997, 2:101-108.

65. Caldas TD, Elyaagoubi A, Richarme G: Chaperone properties of bacterial elongation-factor EF-Tu. J Biol Chem 1998, 273:11478-11482.

66. Van den Berg B, Ellis RJ, Dobson CM: Effects of macromolecular •• crowding on protein folding and aggregation. EMBO J 1999,

18:6927-6933. This exciting paper provides experimental evidence that the conditions of macro- molecular crowding prevalent in the cell may profoundly affect the folding process. By studying the effect of crowding agents on the refolding of the model protein lysozyme, the authors make two important observations. Firstly, that crowding essentially abolishes spontaneous folding by favoring aggregation and secondly, that crowded conditions can greatly enhance chaperone activity.

67. Feldman DE, Thulasiraman V, Ferreyra R, Frydman J: Formation of the •• VHL-elongin BC tumor suppressor complex is mediated by the

chaperonin TRiC. Mol Cell 1999, 4:1051-1061. This paper demonstrates that the chaperonin TRiC/CCT is required for the folding of the VHL tumor suppressor protein and its assembly into a func- tional complex with its partner proteins elongin B and elongin C. VHL inter- acts with TRiC through a 55 amino acid domain that is a target of tumor-causing mutations. Some of these mutations disrupt the interaction of VHL with TRiC, suggesting that loss of protein function may arise through mutations that disrupt the chaperone-substrate interaction.

68. Leroux MR, Fandrich M, Klunker D, Siegers K, Lupas AN, Brown JR, • Schiebel E, Dobson CM, Hartl FU: MtGimC, a novel archaeal

chaperone related to the eukaryotic chaperonin cofactor GimC/prefoldin. EMBO J 1999, 18:6730-6743.

This paper reports the identification and characterization of mtGimC, the homolog of GimC in archaebacteria. Using purified components, the authors show that mtGimC can maintain substrates in a folding-competent form in vitro and deliver them to a bacterial or eukaryotic chaperonin. These results are consistent with those described in [30•]. As this archaeum lacks an Hsp70 homolog, the authors suggest that mtGimC may provide a function- al replacement of the Hsp70 system.

69. Llorca O, McCormack EA, Hynes G, Grantham J, Cordell J, •• Carrascosa JL, Willison KR, Fernandez JJ, Valpuesta JM: Eukaryotic

type II chaperonin CCT interacts with actin through specific subunits. Nature 1999, 402:693-696.

This paper presents the first immuno-electron microscopy reconstruction of a complex between TRiC/CCT and a folding substrate, actin. Actin appears to interact with TRiC through multiple subunit-specific contacts. The con- clusions of this paper are in agreement with those of [35•].

Protein folding in vivo and molecular chaperones Feldman and Frydman 33

SBA112.QXD 02/17/2000 01:21 Page 33