We describe the discovery of a heterohexameric chaperone protein, prefoldin, based on its ability to capture unfolded actin. Prefoldin binds specifically to cytosolic chaperonin (c-cpn) and transfers target proteins to it. Deletion of the gene encoding a prefoldin subunit in S. cerevisiae results in a phenotype similar to those found when c-cpn is mutated, namely impaired functions of the actin and tubulin-based cytoskeleton. Consistent with prefoldin having a general role in chaperonin-mediated folding, we identify homologs in archaea, which have a class II chaperonin but contain neither actin nor tubulin. We show that by directing target proteins to chaperonin, prefoldin promotes folding in an environment in which there are many competing pathways for nonnative proteins.

We describe the complete beta-tubulin folding pathway. Folding intermediates produced via ATP-dependent interaction with cytosolic chaperonin undergo a sequence of interactions with four proteins (cofactors A, D, E, and C). The postchaperonin steps in the reaction cascade do not depend on ATP or GTP hydrolysis, although GTP plays a structural role in tubulin folding. Cofactors A and D function by capturing and stabilizing beta-tubulin in a quasi-native conformation. Cofactor E binds to the cofactor D-beta-tubulin complex; interaction with cofactor C then causes the release of beta-tubulin polypeptides that are committed to the native state. Sequence analysis identifies yeast homologs of cofactors D (cin1) and E (pac2), characterized by mutations that affect microtubule function.

The production of native α/β tubulin heterodimer in vitro depends on the action of cytosolic chaperonin and several protein cofactors. We previously showed that four such cofactors (termed A, C, D, and E) together with native tubulin act on β-tubulin folding intermediates generated by the chaperonin to produce polymerizable tubulin heterodimers. However, this set of cofactors generates native heterodimers only very inefficiently from α-tubulin folding intermediates produced by the same chaperonin. Here we describe the isolation, characterization, and genetic analysis of a novel tubulin folding cofactor (cofactor B) that greatly enhances the efficiency of α-tubulin folding in vitro. This enabled an integrated study of α- and β-tubulin folding: we find that the pathways leading to the formation of native α- and β-tubulin converge in that the folding of the α subunit requires the participation of cofactor complexes containing the β subunit and vice versa. We also show that sequestration of native α-or β-tubulins by complex formation with cofactors results in the destabilization and decay of the remaining free subunit. These data demonstrate that tubulin folding cofactors function by placing and/or maintaining α-and β-tubulin polypeptides in an activated conformational state required for the formation of native α/β heterodimers, and imply that each subunit provides information necessary for the proper folding of the other.

SummaryMyopathy mutations in α-skeletal-muscle actin cause a range of molecular defects

We have characterized the cytosolic chaperonin from both rabbit reticulocyte lysate and bovine testis. The heteromeric complex contains eight subunits. Partial amino acid sequence data reveal that one of these is t-complex polypeptide 1 (TCP-1), while the other seven are TCP-1-related polypeptides, implicating the existence of a multigene family of TCP-1 homologues. We provide evidence that TCP-1 ring complex from bovine testis can facilitate the folding of both actin and tubulin, although, as in the case of chaperonin from reticulocyte lysate, two cofactors are required for the generation of properly folded tubulin. An additional molecule of TCP-1 may associate with the chaperonin depending on the purification procedure used. We propose that a highly conserved region in these polypeptides and in other chaperonins of the cpn6O chaperone family participates in ATP binding. of the chaperonin from reticulocyte lysate shows the presence of at least seven polypeptides. Both complexes contain a subunit that reacts with a monoclonal antibody to TCP-1 and amino acid sequence data confirm the presence ofTCP-1 in TRiC. A difference between the two chaperonins is that TRiC seems to contain two TCP-1 subunits, while the chaperonin from reticulocyte lysate contains only one (3, 5).We conducted this study to further characterize and compare the chaperonin from rabbit reticulocyte lysate and bovine testis. We demonstrate that the chaperonin from testis (TRiC), like the one from reticulocyte lysate, can facilitate the folding of actin and requires the presence of two additional protein cofactors for the generation of properly folded tubulin. Therefore, TRiC probably represents the bovine homologue of the chaperonin isolated from rabbit reticulocyte lysate. We show that the chaperonin from reticulocyte lysate contains eight different polypeptides one of which is TCP-1. All seven other polypeptides show various degrees of similarity to TCP-1. MATERIALS AND METHODSReagents and Enzymes. Untreated rabbit reticulocyte lysate was purchased from Promega. Bovine testes were obtained from the local slaughterhouse. Trypsin was from Sigma and endoproteinase Lys-C was from Boehringer Mannheim. 35S-labeled methionine and cysteine were from ICN Biochemicals. All other reagents were of analytical or HPLC grade.Chaperonin Purification. Chaperonin from rabbit reticulocyte lysate was purified as described by Gao et al. (3). Chaperonin was also purified from bovine testis following the procedure described by Gao et al. The main differences between the two purification procedures are the use of MgCl2 in 20 mM Tris (pH 7.2) (3) instead of NaCl in 10% (vol/vol) glycerol/50 mM Hepes, pH 7.6 (5), as an eluent on the Mono-Q ion-exchange column during the first step, and the use of size-exclusion chromatography (3) rather than sucrose density ultracentrifugation (5).Preparation of M"S-Labeled 1Actin. The entire coding region of human ,B-actin cDNA (provided by H. Leffers and J. Celis, Aarhus University, Denmark) was amplified by PCR using primers...

The nonhomologous proteins actin and alpha- and beta-tubulin need the assistance of the cytosolic chaperonin containing TCP-1 (CCT) to reach their correct native state, and their folding requires a transient binary complex formation with CCT. We show that separate or combined deletion of three delineated hydrophobic sequences in actin disturbs the interaction with CCT. These sites are situated between residues 125-179, 244-285, and 340-375. Also, alpha- and beta-tubulin contain at least one recognition region, and intriguingly, it has a similar distribution of hydrophobic residues as region 244-285 in actin. Internal deletion of the sites in actin favor a model for cooperative binding of target proteins to CCT. Peptide mimetics, representing the binding regions, inhibit target polypeptide binding to CCT, suggesting that actin and tubulin contact similar CCT subunits. In addition, we show that actin recognition by class II chaperonins is different from that by class I.

Abstract. The folding of or-and/3-tubulin requires three proteins: the heteromeric TCP-l-containing cytoplasmic chaperonin and two additional protein cofactors (A and B). We show that these cofactors participate in the folding process and do not merely trigger release, since in the presence of Mg-ATP alone, u-and B-tubulin target proteins are discharged from cytoplasmic chaperonin in a nonnative form. Like the prokaryotic cochaperonin GroES, which interacts with the prototypical Escherichia coli chaperonin GroEL and regulates its ATPase activity, cofactor A modulates the ATPase activity of its cognate chaperonin. However, the sequence of cofactor A derived from a cloned cDNA defines a 13-kD polypeptide with no significant homology to other known proteins. Moreover, while GroES functions as a heptameric ring, cofactor A behaves as a dimer. Thus, cofactor A is a novel cochaperonin that is structurally unrelated to GroES.

Eukaryotic prefoldin (PFD) is a heterohexameric chaperone with a jellyfish-like structure whose function is to deliver nonnative target proteins, principally actins and tubulins, to the eukaryotic cytosolic chaperonin for facilitated folding. Here we demonstrate that functional PFD can spontaneously assemble from its six constituent individual subunits (PFD1-PFD6), each expressed as a recombinant protein. Using engineered forms of PFD assembled in vitro, we show that the tips of the PFD tentacles are required to form binary complexes with authentic target proteins. We show that PFD uses the distal ends of different but overlapping sets of subunits to form stable binary complexes with different target proteins, namely actin and alpha- and beta-tubulin. We also present data that suggest a model for the order of these six subunits within the hexamer. Our data are consistent with the hypothesis that PFD, like the eukaryotic cytosolic chaperonin, has co-evolved specifically to facilitate the folding of its target proteins.