Protein Synthesis and Folding

Published on 01/03/2015 by admin

Last modified 22/04/2025

Print this page

This article have been viewed 3697 times

CHAPTER 17 Protein Synthesis and Folding^*

William E. Balch, Ann L. Hubbard, J. David Castle, Pat Shipman

The nuclear genome contains information specifying many thousands of proteins. Whatever their final destination—nucleus, cytoplasm, membrane-bound organelles, or extracellular space—these proteins are synthesized in the cytoplasm. The few proteins encoded by genes in mitochondria and chloroplasts are synthesized in those organelles. The biochemical synthesis of proteins is called translation, as the process translates sequences of nucleotides in a messenger RNA (mRNA) into the sequence of amino acids in a polypeptide chain. Translation of mRNA requires the concerted actions of small transfer RNAs (tRNAs) linked to amino acids, ribosomes (complexes of RNA and protein), and many soluble proteins. GTP binding and hydrolysis regulate several proteins that orchestrate the interactions of these components. Ultimately, RNA bases in the ribosome catalyze the formation of peptide bonds. Some newly synthesized polypeptides fold spontaneously into their native structure in the cellular environment, but many require assistance from proteins called chaperones. It has been proposed that the bulk of the evolution of the translation apparatus occurred after the basic mechanisms were established, to provide greater precision. This perspective seems to explain the extraordinary complexity of the process.

Protein Synthetic Machinery

Messenger RNA

mRNAs have three parts: Nucleotides at the 5′ end provide binding sites for proteins that initiate polypeptide synthesis; nucleotides in the middle specify the sequence of amino acids in the polypeptide; and nucleotides at the 3′ end regulate the stability of the mRNA (see Figs. 15-1 and 16-1). Within the protein-coding region, successive triplets of three nucleotides, called codons, specify the sequence of amino acids. The genetic code relating nucleotide triplets to amino acids is, with a few minor exceptions, universal. One to six different triplet codons encode each amino acid (Fig. 17-1). An initiation codon (AUG) specifies methionine, which begins all polypeptide chains. In addition, any one of three termination codons (UAA, UGA, UAG) stops peptide synthesis.

Figure 17-1 the genetic code. The location of the nucleotide in first, second, and third position defines the amino acid encrypted by the code.

Eukaryotic and bacterial mRNAs differ in three ways. First, eukaryotic mRNAs encode one protein, and bacterial mRNAs generally encode more than one protein. Second, most eukaryotic (and eukaryotic viral) mRNAs are capped by an inverted 7-methylguanosine residue joined onto the 5′ end of the mRNA by a 5′-triphosphate-5′ linkage (Fig. 17-2). This 5′ cap is stable throughout the life of the mRNA and protects the 5′ end against attack by nucleases. Third, most eukaryotic mRNAs have a tail of 50 to 200 adenine residues added posttranscriptionally to the 3′ end (see Fig. 16-3). The poly(A) tail may protect the mRNA from degradation in the cytoplasm and increase reinitiation of transcription. Bacterial mRNAs lack 5′ caps or 3′ poly(A) tails. Most eukaryotic mRNAs require processing to remove introns (see Fig. 16-4). Many single-stranded mRNAs have some secondary structure (see Fig. 3-19) stabilized by hydrogen bonding of complementary bases. This secondary structure must be disrupted during translation to allow reading of each codon.

Figure 17-2 mRNA cap structures. Prokaryotic mRNAs end with a 5′ triphosphate. The 5′ cap of eukaryotic mRNAs consists of a 7-methylguanosine residue (m7G) linked to the mRNA by three phosphates. The protein eIF4E binds the cap and protects against degradation by nucleases. (PDB file: 1EJ1.)

Transfer RNA

tRNAs are adapters that deliver amino acids to the translation machinery by matching mRNA codons with their corresponding amino acids as they are incorporated into a growing polypeptide (Fig. 17-3). One to four different tRNAs are specific for each amino acid, generally reflecting their abundance in proteins. Specialized tRNAs carrying methionine (formylmethionine in Bacteria) initiate protein synthesis. Transfer RNAs consist of about 76 nucleotides that base-pair to form four stems and three intervening loops. These elements of secondary structure fold to form an L-shaped molecule stabilized by base pairing. A “decoding” triplet (the anticodon) is at one end of the L (the anticodon arm), and the amino acid acceptor site is at the other end of the L (the acceptor arm).

Figure 17-3 tRNA structure. tRNAs match an amino acid attached at the 3′ end with the mRNA triplet coding for that amino acid. A, Ribbon model, space-filling model, and textbook icon showing base pairing of the anticodon to an mRNA codon. B, Backbone model. C, Planar model showing stem loops of a generic tRNA. Single-letter code for the bases: adenine (A), any purine (R), any pyrimidine (Y), cytosine (C), guanine (G), pseudouridine (ψ), thymine (T), and uracil (U). (PDB file: 6TNA.)

Enzymes called aminoacyl-tRNA (aa-tRNA) synthetases catalyze a two-step reaction that couples an amino acid covalently to its cognate tRNA but not to any other tRNA (Fig. 17-4). In the first step, adenosine triphosphate (ATP) and the amino acid react to form a high-energy aminoacyl adenosine monophosphate (AMP) intermediate with release of pyrophosphate. The second step transfers the amino acid to the 3′ adenine of tRNA, forming an aa-tRNA. This reaction is appropriately called charging, since the high-energy bond between the amino acid and the tRNA activates the amino acid in preparation for forming a peptide bond with an amino group in the growing polypeptide chain. Each of the 20 aa-tRNA synthetases couples a particular amino acid to all of its corresponding tRNAs.

Figure 17-4 Charging a tRNA with its correct amino acid. tRNA synthetases (shown schematically and as a space-filling atomic model in purple) provide a docking platform for a specific amino acid and its cognate tRNA (shown in orange as a schematic model and as a ribbon model bound to a synthetase). The amino acid is first activated by reaction with ATP. The carboxyl group of the amino acid is coupled to the a-phosphate of AMP with the release of pyrophosphate. The synthetase then transfers the amino acid from the aminoacyl AMP (aa-AMP) to a high-energy ester bond (red disk) with either the 2′ (illustrated here) or 3′ hydroxyl of the adenine at the 3′ end of the tRNA. (PDB file: 1QTQ.)

The fidelity of protein synthesis depends on near-perfect coupling of amino acids to the appropriate tRNAs. Synthetases make this selection by interacting with as many as three areas of their cognate tRNAs: anticodon, 3′ acceptor stem, and the surface between these sites (Fig. 17-4). To distinguish between appropriate and inappropriate amino acids, synthetases use proofreading steps, which remove incorrectly paired amino acids from tRNAs.

Ribosomes

Ribosomes are giant macromolecular machines that bring together an mRNA and aa-tRNAs to synthesize a polypeptide. Base pairing between mRNA codons and tRNA anticodons directs the synthesis of a polypeptide in the order specified by the mRNA codons. Ribosomes consist of a small subunit and a large subunit that bind together during translation of an mRNA (Fig. 17-5). Each subunit consists of one or more ribosomal RNA (rRNA) molecules and many distinct proteins (Fig. 17-6). The sizes of these subunits and rRNAs are traditionally given in units of S, the sedimentation coefficient measured in an ultracentrifuge.

Figure 17-5 model of the bacterial ribosome illustrating overall organization. Two subunits (30S and 50S) form a functional 70S ribosome. An mRNA threads between the subunits in association with the small subunit. tRNAs bind to two sites, designated the A site and P site, between the large and small subunits. Codons of the mRNA in the A and P sites on the small subunit specify which aa-tRNAs occupy these sites. The amino acids at the far end of the bound tRNAs are positioned for peptide bond formation by the peptidyl transferase site on the large subunit. The growing polypeptide chain (shown in blue) emerges from a tunnel in the large subunit.

Figure 17-6 molecular components of ribosomes. A–B, Inventories of rRNAs (middle) and proteins (bottom). Prokaryotic and eukaryotic rRNAs and ribosomal proteins differ in size and number but are related by evolution and form similar structures. C–D, Secondary structures of prokaryotic 16S rRNA and 18S eukaryotic rRNA illustrate their similarities despite divergent sequences.

Ribosomal RNAs constitute the structural core of each ribosomal subunit (Fig. 17-7). The 16S rRNA of the small subunit consists of 1500 bases, most of which are folded into base-paired helices. The large subunit contains two RNAs: 23 S rRNA consisting of 2900 bases and 5S rRNA of 121 bases. The rRNAs fold into many based-paired helices, as predicted by phylogenetic analysis of sequences (Fig. 17-6). These helices and their intervening loops pack into a compact structure, as is seen in both surface views and cross sections. Although eukaryotic rRNAs differ in size and sequence from prokaryote rRNAs, their predicted secondary structures are similar, and they are expected to fold in similar ways. Many features of rRNAs have been conserved during evolution, including the surfaces where subunits and elements of RNA structure interact; sites that are required for binding tRNA, mRNA, and protein cofactors; and the residues involved with peptide bond formation.

Figure 17-7 crystal structures of the ribosome small and large subunits. RNA is shown in gray, and proteins are gold, except in panel G, which features various colors. A–B, Two views of the model of the small subunit of Thermus thermophilus. C–D, Representative structures of individual ribosomal proteins and their locations on the small and large subunits. (PDB file: 1FJF.) E–J, Structure of the large subunit of the ribosome of Haloarcula marismortui. E, Crown view from the perspective of the small subunit. F, View in panel C rotated 180 degrees around a vertical axis. G, Crown view of the proteins minus RNA. H, View in panel E rotated 180 degrees around a horizontal axis to show the exit from the nascent polypeptide tunnel, the dark patch in the middle. I, Crown view with models of tRNA in the A, P, and E sites. J, Cross section showing the tunnel for the nascent polypeptide extending from the peptidyl transferase (PT) site to the exit.

(A–B, From Wimberly BT, Brodersen DE, Clemons WM, et al: Structure of the 30S ribosomal subunit. Nature 407:327–339, 2000. E–J, Courtesy of T. Steitz, Yale University, New Haven, Connecticut; adapted from the work of Ban N, Nissen P, Hansen J, et al: The complete atomic structure of the large ribosomal subunit at 2.4 Å resolution. Science 289:905–920, 2000; and Nissen P, Hansen J, Ban N, et al: The structural basis of ribosome activity in peptide bond synthesis. Science 289:920–930, 2000. A–C, PDB file: 1FJF. D–J, PDB file: 1FFK.)

Most ribosomal proteins associate with the surface of the rRNA core, although several extend peptide strands into the core (Fig. 17-7). Ribosomal proteins are generally small (10 to 30 kD) and basic, but each has a unique structure. With one exception, ribosomes have just one copy of each protein.

Decoding of the mRNA and synthesis of the polypeptide take place in the cavity between the subunits. The surfaces of this cavity are generally free of proteins, so rRNAs—not proteins—are largely responsible for mRNA binding, tRNA binding, and peptide bond formation. tRNAs move sequentially through three sites shared by the two subunits: the A site (aa-tRNA), the P site (for peptidyl-tRNA), and the E site (for exit). The growing polypeptide chain exits through a tunnel in the RNA core of the large subunit.

Soluble Protein Factors

Many soluble proteins cycle on and off ribosomes during protein synthesis, enhancing the rate or fidelity of the reactions. The following sections highlight the role(s) of these soluble factors.

Outline of Protein Synthesis

Organisms in all three domains of life use many homologous components and similar reactions for protein synthesis, but many of the details differ as is expected after 3 billion years of evolutionary divergence. In all three domains, protein synthesis takes place in four steps: initiation, elongation, termination, and subunit recycling (Fig. 17-8). Guanosine triphosphatase (GTPase) proteins regulate the progress and fidelity of many of the steps (see Fig. 4-6 for details on GTPase cycles). Initiation, elongation, and termination all depend on directed movement of molecular machinery along an mRNA and precise recognition between amino acids, tRNAs, adapter proteins, and the gene sequence encoded in the mRNA.

Figure 17-8 Overview of the translation cycle showing six ribosomes on a single mrna. 1, initiation. initiator tRNA^Met, mrna, and accessory soluble factors assemble on the small subunit, which then joins with a large subunit. 2, Elongation. The polypeptide chain is synthesized, in the order specified by the mRNA, in sequential steps by recruitment of new aa-tRNAs that match the mRNA-coding sequence, formation of peptide bonds, and dissociation of free tRNA. 3, Termination. Release factors recognize the stop codon (yellow) and terminate translation. The ribosome releases the polypeptide for folding in the cytoplasm. 4, Subunit recycling. The ribosomal subunits dissociate and are available for another round of translation.

During initiation, a complex composed of a small ribosomal subunit and an initiator tRNA (carrying methionine) binds the initiation codon (AUG) of an mRNA. This ternary complex then associates with a large subunit to form a 70S ribosome in Bacteria and an 80S ribosome in eukaryotes. Eukaryotes use many more components than prokaryotes to regulate initiation.

During elongation, tRNAs bring amino acids to the ribosome in the order specified by the sequence of codons in the mRNA. The ribosome catalyzes formation of a peptide bond between the amino group of each new amino acid and the carboxyl group at the C-terminus of the growing polypeptide chain and then moves on to the next codon. The mechanism of elongation is conserved across the phylogenetic tree. More than one ribosome is active on most mRNAs, as coding sequences are usually much longer than the 40 to 50 nucleotides associated with a single ribosome. Once a ribosome proceeds about 60 nucleotides beyond the initiation codon, another ribosome-tRNA complex can assemble on the mRNA and start translation. Messenger RNAs with multiple ribosomes are called polysomes. This multiple occupancy of mRNAs explains why ribosomes are more abundant than mRNAs and how one mRNA molecule guides the synthesis of several copies of its protein product simultaneously.

Termination occurs when the ribosome encounters a termination codon (UAA, UAG, or UGA) at the 3′ end of the coding sequence. At this point, a protein factor (not an aa-tRNA) binds to the mRNA, and the C-terminal amino acid of the polypeptide chain is hydrolyzed from its tRNA. After the polypeptide is released from the ribosome, the ribosomal subunits dissociate and are available for recycling to initiate translation of another mRNA.

Initiation Phase

The goal of initiation is to bring together the initiator tRNA carrying methionine (or N-formylmethionine, fMet, in Bacteria) and the AUG initiator codon of the mRNA on the ribosome (Fig. 17-9). In eukaryotes, more than 10 soluble protein factors (eukaryotic initiation factors, or eIF) coordinate the interactions of the RNA molecules. Fewer protein factors (designated IF) participate in prokaryotes. In eukaryotes, several steps occur in succession:

Figure 17-9 steps in initiation in eukaryotes. 1, Initiation factors (green) assemble with mRNA, eIF-2a (purple, activated with GTP), and tRNA^Met on a small ribosomal subunit to form the preinitiation complex. 2, Other initiation factors (blue) bind the 5′ cap of the mRNA. For some mRNAs, these 5′ cap-binding factors interact with poly(A)-binding proteins at the 3′ end of the mRNA. This circularization promotes initiation of some mRNAs and inhibits initiation of other mRNAs. 3, The preinitiation complex binds an mRNA. 4, The small subunit scans the mRNA for the AUG start codon (green). 5, When the initiator tRNA binds the start codon, eIF-2a hydrolyzes its bound GTP. 6, Phosphate, GDP, eIF-2a, and other initiation factors dissociate and recycle for further rounds of initiation. 7, The small subunit binds a large subunit. 8, Elongation begins.

Step 1 Initiator Met-tRNA and the GTPase eIF-2a (with bound guanosine triphosphate [GTP]) form a preinitiation complex on a small ribosomal subunit.

Step 2 Several protein initiation factors assemble on the 5′ cap of the mRNA. The RNA helicase eIF4A in this complex uses ATP hydrolysis to remove any secondary structure or bound proteins at the 5′ end of the mRNA. These cap recognition factors also interact with poly(A)-binding proteins on the other end of the mRNA, forming a circular complex that can either favor or inhibit initiation.

Step 3 The cap recognition complex targets the mRNA to a preinitiation complex. The order of these first three steps is still being investigated. mRNA may also bind to the small subunit before the initiation factors and Met-tRNA.

Step 4 The ribosome scans along the mRNA for the initiator AUG codon. This movement depends on ATP hydrolysis, but its role is not clear. Eukaryotic mRNAs tend to begin translation at the first AUG codon encountered, but the local sequence of the mRNA may also contribute to the specificity as in Bacteria.

Step 5 When Met-tRNA base-pairs with the initiator AUG codon, eIF-2a hydrolyzes its bound GTP.

Step 6 eIF-2a and the other initiation factors dissociate from the small subunit for recycling back to other preinitiation complexes and 5′ caps.

Step 7 A large ribosomal subunit binds the small subunit complexed with both the mRNA and Met-tRNA. Another GTPase called eIF5B hydrolyzes its bound GTP before elongation of the polypeptide begins.

Initiation is the most highly regulated step in protein synthesis, frequently involving phosphorylation of initiation factors. For example, phosphorylation increases the affinity of eIF-2a for its guanine nucleotide-exchange factor (eIF-2b). Strongly bound eIF-2b inhibits initiation by competing with initiator tRNA for binding eIF-2a. Cells that are subjected to various stresses utilize phosphorylation of eIF-2a to inhibit translation. In contrast, phosphorylation of eIF-4F favors translation by enhancing the interaction of this initiation factor with the 5′ cap of mRNAs. This modulation of affinity can even influence the selective translation of particular mRNAs, since the 5′ caps of mRNAs vary in affinity for eIF-4F.

Elongation Phase

Accurate protein synthesis depends on the fidelity of amino acid coupling to the correct tRNA and of codon-anticodon pairing between mRNA and tRNA. Both reactions occur in two steps by a mechanism that increases accuracy. Much of the energy invested in protein synthesis is used to achieve this accuracy, and elongation is the most expensive phase of translation in terms of energy expenditure.

Repetitive cycles of codon-directed incorporation of amino acids into the polypeptide chain begin once the two ribosomal subunits are joined with an initiator tRNA and mRNA properly in place (Fig. 17-10). Each cycle of elongation consists of four steps: (1) binding of an aa-tRNA to the A site on the ribosome; (2) proofreading to ensure that it is the correct aa-tRNA; (3) peptide bond formation; and (4) translocation, which advances the mRNA by one codon and moves the peptidyl-tRNA from the A site to the P site on the ribosome.

Figure 17-10 steps in elongation and termination in eukaryotes. Starting in the upper left, elongation factor eEF1A (EF-Tu in Bacteria) forms a ternary complex with GTP and an amino acyl-tRNA^aa for delivery of the tRNA^aa, matching the mRNA codon in the A site to the ribosome. This ternary complex dissociates rapidly if the anticodon-codon match is incorrect. If the anticodon-codon match is correct, the ternary complex remains bound to the A site long enough for eEF1A to hydrolyze its bound GTP and dissociate from the tRNA still bound to the A site. The ribosome catalyzes formation of a new peptide bond (inset). eEF2 (EF-G in Bacteria) binds the A site transiently after peptide bond formation to facilitate movement of the tRNAs and mRNA through the ribosome. Release factors (RF, green) recognize the stop codon and terminate the polypeptide chain (blue), allowing the mRNA and ribosomal subunits to dissociate. The guanine nucleotide-exchange factor eEFX promotes the exchange of GDP for GTP on eEF-1A.

Elongation reactions occur in a cavity between the two ribosomal subunits. mRNA is threaded, codon by codon, between the subunits. aa-tRNAs enter on one side of the cavity and bind successively to three sites between the two ribosomal subunits. Having half of each site on each of the two subunits allows ribosomes to maintain contact with one end of the tRNA as it moves, step by step, from the A site to the P site to the E site prior to dissociation. Codon-anticodon recognition takes place at both the A and P sites on the small subunit, where the anticodons of the two tRNAs base-pair with mRNA. Peptide bonds form at the other end of the tRNAs, which position the amino acid and peptidyl chain on the A and P sites of the large subunit. Elongation factors (EF; eEF for eukaryotic elongation factors) and movements of the subunits relative to each other facilitate the movements of the tRNAs along the sequence of three sites. The growing polypeptide exits through a 10-nm-long tunnel in the large subunit.

Step 1 aa-tRNA binding. The GTPase eEF1A (EF-Tu in Bacteria; see Fig. 25-7) with bound GTP delivers aa-tRNAs to ribosomes with empty A sites. The nucleotide-exchange factor eEF-X (EF-Ts in Bacteria) prepares eEF1A to bind aa-tRNA by promoting the exchange of GDP for GTP. Cells contain enough eEF1A-GTP to bind all of the aa-tRNA and protect the labile ester bond of the aa-tRNA.

Step 2 Proofreading. A proofreading mechanism in the A site checks each aa-tRNA to ensure that its anticodon matches the mRNA codon in the decoding site of the small subunit. Correct aa-tRNAs are retained; incorrect aa-tRNAs dissociate. A “kinetic proofreading mechanism” discriminates between correct and incorrect aa-tRNAs using two first-order reactions. First, eEF1A associated with the aa-tRNA hydrolyzes its bound GTP. Then GDP-eEF1A dissociates from the aminoacyl end of the aa-tRNA and the ribosome, allowing the aminoacyl end of the aa-tRNA to move into the peptidyl transfer site on the large subunit. Each reaction takes a few milliseconds. Correct base pairing between the aa-tRNA anticodon and the mRNA codon promotes GTP hydrolysis by eEF1A, so GDP-eEF1A can dissociate and allow the aa-tRNA to form a peptide bond. Those aa-rRNAs with weak, imperfect codon-anticodon pairs dissociate from the A site before eEF1A can hydrolyze GTP and dissociate from the aminoacyl end of the tRNA.

Step 3 Peptidyl transfer. The RNA of the large subunit forms the highly conserved active site that catalyzes the formation of peptide bonds (Fig. 17-10). This reaction eliminates water and transfers the carboxyl group esterified to the peptidyl-tRNA in the P site to the free amino group of the aa-tRNA in the A site. Catalysis of peptide bond formation depends on a combination of precise orientation of the substrates and stabilization of the transition state (just like protein enzymes). The chemistry is similar, but in reverse, to the hydrolysis of peptide bonds by proteolytic enzymes such as chymotrypsin. After formation of the new peptide bond, the tRNA in the A site has the polypeptide on one end and its anticodon arm still base-paired to its mRNA codon on the small subunit. The antibacterial agent puromycin can disrupt elongation by mimicking a tRNA^Phe or tRNA^Tyr (Fig. 17-11). Puromycin attacks the esterified carboxyl group of a peptidyl-tRNA in the P site, but lacking an appropriate acceptor site for further peptidyl transfer reactions, it terminates elongation, resulting in premature release of the polypeptide chain from the ribosome.

Step 4 Translocation. Three linked reactions, promoted by elongation factor eEF2 (and the homologous protein EF-G in Bacteria), complete each elongation cycle. eEF2 is a GTPase with domains similar to domains 1 and 2 of EF-Tu (see Fig. 25-7) plus three domains that mimic the size and shape of a tRNA. Domain 1 binds and hydrolyzes GTP. Domains 3 to 5 target GTP-eEF2 to an empty A site on the ribosome. Binding of GTP-eEF2 to an empty A site promotes the movement of peptidyl-tRNA from the A site to the P site on the small subunit together with sliding of the mRNA three bases forward on the small subunit. At the same time, the deacylated tRNA in the P site is moved to the exit (E) site, where it dissociates from the ribosome. Hydrolysis of the bound GTP releases eEF2 from the A site, initiating another round of elongation.

Figure 17-11 mechanism of the puromycin reaction shown in four steps. The antibiotic puromycin mimics the terminus of amino acyl-tRNA^Tyr or tRNA^Phe. It is incorporated on the C-terminus of the polypeptide and terminates translation prematurely, as it is not attached to a tRNA and lacks an activated carboxyl group.

The growing peptide threads through a 10-nm-long tunnel in the large subunit lined with RNA (Figs. 17-5, 17-7, and 17-8). The tunnel accommodates an extended polypeptide about 40 residues long. The distal parts of the tunnel are wide enough to pass an a-helix. The N-terminus of longer peptides exits from the large subunit.

Cells balance speed and accuracy during translation to achieve an error rate of about 1 in 10⁴ incorrect amino acids. As a result of this compromise, ribosomes add about 20 amino acids per second to a polypeptide at 37°C, so synthesis of a protein of average size (300 amino acids) takes only 15 seconds. Greater precision might be achieved by slowing translation, but slower cellular growth might be an evolutionary dis-advantage.

Termination Phase

The assembly of a protein stops when a termination codon (UAA, UAG, or UGA) moves into the A site on the small subunit of the ribosome (Fig. 17-10). A release factor called eRF1 (RF1 or RF2 in Bacteria) recognizes stop codons, binds to the A site, and induces the ribosome active site to hydrolyze the peptidyl-tRNA ester in the P site. The completed polypeptide chain threads through the ribosome and is released. The large subunit dissociates from the mRNA and the small subunit, leaving both subunits ready to initiate another round of synthesis. Protein factors might contribute to these recycling reactions, but the details are still being investigated.

Spontaneous Protein Folding

Termination is the final step in translation but just the beginning for a new protein. A polypeptide begins to experience its new environment while still being synthesized. When it is about 40 residues long, its N-terminus emerges from the protected tunnel of the large ribosomal subunit into cytoplasm, where it must fold into a three-dimensional structure (see Fig. 3-5) and find its correct cellular destination.

The structure of folded proteins and the folding mechanism are both encoded in the amino acid sequence, making folding spontaneous under suitable conditions. For the soluble proteins, these conditions are aqueous solvent at physiological temperature, neutral pH, and moderate ionic strength. Folding of transmembrane proteins in a lipid bilayer is quite different (see Chapter 20). In test tube experiments, small soluble proteins can be denatured with high temperature, extremes of pH, or high concentrations of urea or guanidine. Denatured proteins exist as ensembles of unfolded polymers with little residual secondary structure.

When denatured polypeptides of modest length are transferred to physiological conditions, many fold spontaneously into their native three-dimensional structures on a microsecond to millisecond time scale. (Proteins that require isomerization of prolines, such as collagen, fold much more slowly; see Fig. 29-4.) Starting from many initial denatured states, the polypeptides converge toward a single low-energy native state (Fig. 17-12). The number of possible pathways to the native state is so numerous that if they were sampled individually, proteins would never fold. Thus, both theory and experiment indicate that folding involves a subset of the potential pathways, including an ensemble of loosely folded transition states with elements of secondary structure, certain turns, and hydrophobic contacts found in the core of the native protein.

Figure 17-12 energy considerations in protein folding. As a protein matures from the unfolded state (U) through transition states (TS) to the native folded state (N), native-like contacts form, and the free energy of the system decreases. The two paths (folding trajectories) illustrate that fast protein folding (yellow line) is observed when more native-like contacts are made. When proteins become trapped in partially folded intermediate states, folding is slower (pink line) because energy barriers must be overcome.

(Adapted from Radford SE, Dobson CM: Computer simulations to human disease: Emerging themes in protein folding. Cell 97:291–298, 1999.)

Many proteins also fold spontaneously on their own during biosynthesis in vivo. Folding begins when the N-terminus of the nascent polypeptide emerges from the ribosome. The vectorial nature of this “cotranslational folding” has both advantages and liabilities. An advantage is that vectorial folding limits the routes to the folded state and might account for why many proteins fold more efficiently during biosynthesis than from the denatured state. On the other hand, vectorial folding precludes interactions between N-terminal sequences with C-terminal sequences until they have emerged from the ribosome. Such interactions are common in folded proteins.

Folding of larger proteins is more complicated, especially in the presence of other partially folded proteins with exposed hydrophobic segments that are buried in the core of native proteins. These exposed core elements are prone to aggregate irreversibly before completing folding. Many newly synthesized native proteins also need assistance to avoid irreversible denaturation, aggregation, or destruction by proteolysis during folding.

Misfolding of mutant proteins contributes to many human diseases. For example, the most common cause of cystic fibrosis is genetic deletion of a single amino acid in CFTR, resulting in failure of the protein to fold properly (see Fig. 11-4). Beyond lacking function, misfolded proteins also poison the assembly of native proteins in blistering skin diseases (see Fig. 35-6), hypertrophic cardiomyopathies (see Table 39-4), and other “dominant negative” conditions. Folding of proteins into nonnative states causes prion and amyloid diseases.

Chaperone-Assisted Protein Folding

Several families of molecular chaperones (Fig. 17-13) facilitate folding of newly synthesized and denatured proteins. These chaperones do not fold polypeptides by directing the formation of secondary or tertiary structure. Rather, by binding exposed hydrophobic segments of nonnative polypeptides or providing sequestered environments, chaperones inhibit aggregation. They release polypeptides in a folding-competent state for attempts at folding. If folding fails, the cycle of binding and release can be repeated. The following sections cover trigger factor (and other chaperones associated with ribosomes), Hsp70, Hsp90, and cylindrical chaperonins. In addition, specialized chaperones assist with the folding of particular proteins such as tubulin and actin. Mutations in several of these chaperones have been associated with human disease. See Fig. 20-10 for chaperones in the endoplasmic reticulum.

Figure 17-13 comparison of chaperone-assisted folding pathways. A, Bacteria. B, Eukaryotes. The percentages refer to estimates of the fraction of proteins using each pathway. Most proteins fold without the assistance of chaperones. NAC, nascent polypeptide-associated complex.

Trigger Factor

Hydrophobic segments of the nascent chain must be protected from aggregation until enough of the chain has emerged from the ribosome to participate in folding. Each growing polypeptide first encounters a chaperone bound next to the exit tunnel on the large ribosomal subunit. The chaperone associated with bacterial ribosomes is called trigger factor (Fig. 17-13). A structurally unrelated protein called nascent polypeptide-associated complex has a similar function in Archaea and eukaryotes. Trigger factor has a hydrophobic groove that is suggested to cradle the growing polypeptide. The signal recognition particle binds on the other side of the exit tunnel, positioned so that its methionine-rich groove (see Fig. 20-5) also interacts with the growing polypeptide. Most bacterial polypeptides fold successfully after being released from trigger factor, while most eukaryotic polypeptides require assistance from additional chaperones.

Hsp70 Chaperones

The most widespread chaperones are members of the heat shock protein 70 (Hsp70) family (Fig. 17-14). Their name came from the observation that cells subjected to stresses, such as elevated temperature, increase the synthesis of these proteins to protect against denatured proteins. Hsp70s are present in Archaea, Bacteria (called DnaK), and most compartments of eukaryotes. The family includes Hsp70 in mitochondria and BiP in endoplasmic reticulum (see Fig. 20-4). Budding yeasts have genes for 14 Hsp70s; vertebrates have more.

Figure 17-14 hsp70 structure and function. A, The Hsp70 folding cycle with bacterial DnaK as the example. B–C, Atomic structures of DnaK (blue) and GrpE (green). The ATPase domain and peptide-binding domain work together in a cyclical mechanism. DnaJ (Hsp40) delivers an unfolded peptide to the ATP-bound open state of DnaK and promotes ATP hydrolysis. The ADP-bound closed state of DnaK binds the peptide strongly. GrpE promotes dissociation of ADP. Rebinding of ATP dissociates GrpE and the peptide, which is free to attempt folding. Multiple Hsp70 cycles are usually required to complete protein folding.

(PDB files: 1DKX and 1DKG. References: Zhu X, Zhao X, Burkholder WF, et al: Structural analysis of substrate binding by the molecular chaperone DnaK. Science 272:1606–1614, 1996; and Harrison CJ, Hayer-Hartl M, Hartl F, et al: Crystal structure of the nucleotide exchange factor GrpE bound to the ATPase domain of the molecular chaperone DnaK. Science 276:431–435, 1997.)

Hsp70s bind and release peptides with 8 to 13 hydrophobic residues in a wide range of nascent or unfolded polypeptides. ATP binding and hydrolysis drive cycles of peptide binding and release, protecting hydrophobic peptides from aggregation during attempts at folding, delivery to mitochondria and chloroplasts, and import into these organelles (see Figs. 18-4 and 18-6).

Bacterial Hsp70 is a well-characterized model for other members of the family. A flexible hinge connects the N-terminal ATP-binding domain to the C-terminal peptide-binding domain. ATP binding favors release of the polypeptide, whereas ATP hydrolysis and phosphate release favors association with an unfolded polypeptide. DnaJ (Hsp40) delivers unfolded proteins to DnaK and promotes their binding by stimulating hydrolysis of ATP bound to DnaK. GrpE promotes exchange of ADP for ATP and release of the bound peptide. Animal Hsp70s have a mechanism of action similar to that of DnaJ except that they have intrinsic nucleotide-exchange activity and do not require a nucleotide-exchange protein such as GrpE.

Hsp90 Chaperones

Hsp90 cooperates with other chaperones to stabilize steroid-hormone receptors before they bind their ligands such as progesterone, glucocorticoids, estrogens, or androgens (Fig. 17-15). The chaperones use cycles of ATP hydrolysis to maintain receptors in an “open” state, ready to bind hydrophobic steroids. Steroid binding completes the folding of the receptors and displaces the Hsp90 complex. Then the receptors move to the nucleus to regulate gene expression (see Fig. 15-22). Hsp90 also interacts with other signaling proteins including protein kinases.

Figure 17-15 Stabilization of ligand-free steroid hormone receptors (SHRs) by Hsp70, Hsp90, and various accessory factors (HOP, HIP, P23, GA, and IP). hormone binding releases the chaperones and allows the receptor-steroid complex to move to the nucleus.

(Reference: Buchner J: Hsp90 & Co.: A holding for folding. Trends Biochem Sci 24:136–142, 1999.)

Chaperonins

The chaperonin family of barrel-shaped particles promotes efficient protein folding (Fig. 17-16). They allow nascent and denatured polypeptides to fold or refold while sequestered in a cylindrical cavity protected from the complex environment of the cytoplasm. Although 85% of newly synthesized bacterial proteins fold spontaneously or with the assistance of Hsp70s, the remainder require the more isolated folding environment provided by chaperonins (Fig. 17-13). The mechanism of chaperonins is best understood for Escherichia coli GroEL and its co-chaperonin GroES. These assist with folding of nascent polypeptides, which in bacteria occurs largely after translation is complete.

Figure 17-16 Chaperonin-mediated folding by GroEL AND GroES. A, One folding cycle. B, Crystal structure of GroEL with a GroES cap bound to the upper, ATP-bound ring of seven subunits. Unfolded polypeptides bind the rim of an uncapped ring. Cooperative binding of ATP to each of the seven GroEL subunits in one ring changes their conformation, favors GroES binding, and doubles the volume of the central cavity, where the protein folds. Following ATP hydrolysis, binding of ATP and GroES to the lower ring structure dissociates the upper GroES and discharges the folded protein.

(B, Based on the work of Xu Z, Horwich AL, Sigler PB: The crystal structure of the asymmetric GroEL-GroES-(ADP)7 chaperonin complex. Nature 388:741–750, 1997. PDB file: 1AON.)

The GroEL/GroES complex consists of a cylinder with a central cavity composed of GroEL and a cap structure made of GroES. GroEL forms two rings of seven identical subunits. Mitochondrial (Hsp60/Hsp10), chloroplast (Cpn60/Cpn10), and eukaryotic chaperonins (TriC) are similar in design but more elaborate than GroEL/GroES, containing up to eight different gene products. This complexity represents evolutionary diversification for regulation of chaperonin function.

ATP binding and hydrolysis set the tempo for fold-ing cycles. Unfolded polypeptides bind to hydrophobic patches on the inner wall of the GroEL cylinder. Cooperative binding of ATP to each of the subunits in one of the two rings of seven changes their conformation (compare the upper and lower rings in Fig. 17-16B), expanding the internal volume by twofold and favoring binding of a heptameric ring of 10-kD GroES subunits. This closes the top of the cylinder and creates a folding cavity for proteins up to about 70 kD. After ATP hydrolysis on the ring surrounding the folding protein and ATP binding to the opposite ring of seven GroEL subunits, the GroES cap releases, and the cage opens. Folded polypeptides escape into the bulk solution, whereas incompletely folded intermediates can rebind GroEL for another attempt at folding.