Molecular Biology: Protein Structure and Functions - Stepanov V.M. 2005

Tertiary protein structure
Protein renaturation

Protein Denaturation under certain in vitro conditions can be reversed by transitioning from an extended polypeptide chain to a compact globule with a well-defined Spatial Structure. This process, known as renaturation, simulates—albeit partially—the folding of a polypeptide chain into a globule during Translation in METABOLISM/35.html">Protein Biosynthesis.

As is well known, the Spatial structure of a protein is determined by its Primary Structure. The familiar one-Gene-one-protein paradigm is essentially equivalent to the statement that a genetically determined Amino Acid Sequence is sufficient to uniquely direct its folding into the tertiary structure characteristic of a given protein. Naturally, this Conclusion generally holds true only under conditions close to those existing within The Cell during The biosynthesis of that specific protein, which may include a specific pH range, the presence of certain ions (such as Calcium Ions), and protein-specific Cofactors such as Coenzymes, heme, etc.

Provided these conditions are met, the aforementioned balance of factors governing protein globule stability will be achieved, thereby creating the thermodynamic Prerequisites for the Formation of the native structure—namely, protein renaturation.

This principle was confirmed by successful in vitro protein renaturation experiments conducted for the first time by C. Anfinsen and coworkers in the 1960s using pancreatic Ribonuclease and Lysozyme, followed by a series of other objects.

Many Proteins, including those possessing intramolecular Disulfide Bonds, undergo complete denaturation in the presence of high concentrations of urea or guanidine hydrochloride, coupled with the reduction of disulfide bonds by reducing agents (such as mercaptoethanol, mercaptoacetic acid, or dithiothreitol). Naturally, this is accompanied by their complete inactivation.

Upon the slow removal of the denaturing agent and reducing agent (e.g., via dialysis or Gel filtration), the native structure forms, disulfide bonds are re-formed (for which oxidation by atmospheric oxygen is often sufficient), and full protein activity is restored. Special experiments confirmed the correct formation of disulfide bonds, which, In addition to reactivation, indicates the restoration of the native structure. Taking lysozyme as an example, this conclusion was directly confirmed by X-Ray Diffraction Analysis. Table 6.4 illustrates the renaturation efficiency for A number of proteins.

It might seem that the conclusion regarding the predetermination of spatial structure by The amino acid sequence—supported by the aforementioned experiments—is not entirely rigorous. The point is that experimentally proving the complete disruption of all non-covalent interactions during protein denaturation is difficult; therefore, the retention of some core portion of the native structure cannot be entirely ruled out. However, this argument is refuted by the fact that a number of (albeit small) proteins have been successfully obtained through the Chemical synthesis of their polypeptide chains from Amino Acid Derivatives. Such peptide chains invariably yielded a high recovery of active (and thus native) protein upon renaturation (see Chapter 2).

Class="center">Table 6.4 Renaturation of proteins denatured by Treatment with 8 M urea or 6 M guanidine hydrochloride with simultaneous reduction of disulfide bonds

* The theoretical yield is calculated assuming that disulfide bonds form randomly.

** a-Amylase from Aspergillus oryzae.

Despite the successful formation of the native structure, in vitro protein renaturation proceeds quite slowly, which is inconsistent with the high rate of tertiary structure folding under physiological conditions. Apparently, The most significant challenge lies in selecting the optimal, kinetically favorable pathway for the transition from the extended polypeptide chain to the compact spatial structure. The number of all possible Folding Pathways for a sufficiently long polypeptide chain is so vast (including numerous unproductive, dead-end paths) that transitioning to the correct structure through a simple trial-and-error search of various Conformations would require an immense amount of time.

Actual protein folding evidently proceeds via a single pathway or, at any rate, very few pathways, which dramatically reduces the time required to form the native structure. In vivo folding of the polypeptide chain is necessarily synchronized with its biosynthesis, which inherently allows for the Selection of a specific pathway since spatial structure formation can occur cotranslationally. Moreover, during protein folding under physiological conditions, both the conformation in which the polypeptide chain exits the peptidyl transferase center of The Ribosome and even The rate of chain elongation may be crucial—there is evidence of translation pauses at sequence regions that mark the boundaries between Secondary structure elements.

It is clear that during in vitro protein renaturation, neither of these two factors can operate, and although it still proceeds not randomly but via a restricted number of pathways, it requires significantly more time to find these pathways than it does in vivo.

The existence of a defined polypeptide folding pathway, considered a highly probable hypothesis until recently, is supported by a number of experimental facts. For instance, disulfide bonds in pancreatic Trypsin inhibitor and ribonuclease form in a specific order, and the initial bonds created are not necessarily the "correct" ones—that is, those present in the final native protein form. Temporary disulfide bonds are sometimes observed to form between Cysteine residues that presumably come into close proximity in an intermediate complex. Subsequently, such a disulfide bond engages in an exchange reaction with other unoxidized cysteine residues and is cleaved, making way for the bonds characteristic of the native conformation.

The application of Site-Directed Mutagenesis (Protein Engineering) has shown that alongside amino acid residues critical for forming a stable protein spatial structure, there are others whose substitution does not affect the Stability of the protein globule but severely impacts The kinetics of polypeptide chain folding. Consequently, these amino acid residues are important for generating intermediate structures along the pathway toward the stable native globule.

For example, during the folding of the a-subunit of Tryptophan synthase (Fig. 6.11), an intermediate stage is observed in which the large N-terminal domain (comprising amino acid residues 1–188 and formed by six parallel ß-Structure segments and five a-helices) is already folded, whereas the C-terminal fragment (beginning at residue 189 and forming two ß-structure segments and three a-helices in the native protein; 80 amino acid residues in total) remains denatured and disordered. Note that this protein lacks spatially separated domains—its structure is rather compact; rather, these domains are considered intermediate folding forms of the a-subunit.

Fig. 6.11. Tertiary Structure of the tryptophan synthase a-subunit.

Like Triosephosphate isomerase, it is formed by alternating a-helices (cylinders) and ß-structure segments (arrows). The small circles in the center indicate the inhibitor binding site. The arrow marks the boundary between the N- and C-terminal domains; in this region, trypsin Hydrolysis at the Arg-188 residue leads to domain Separation. Numbers indicate the sequential arrangement of secondary structure elements

Among the mutants of this enzyme, variants were selected that affect the contacts between the N- and C-terminal domains. Phe-22 of the N-terminal domain, which contacts the 8th ß-structure segment in the C-terminal domain, was replaced with Leu. This resulted in slowed rates of denaturation and renaturation without affecting protein stability. This is evidently due to a reduction in the stability of the Transition State during folding. The replacement of Gly-234 in the C-terminal domain with aspartic acid slowed unfolding to a greater extent than folding. Since the first and eighth ß-structure segments—incorporating phenylalanine and Glycine, respectively—are brought close together in the tertiary STRUCTURE OF THE folded subunit, their mutual "docking" appears to play an important role in forming the folding intermediate, thereby accelerating renaturation.

Investigations into the renaturation kinetics of various proteins allow this process to be described in a first approximation as follows. First, short segments of secondary structure—a-helices, ß-sheets, ß-turns—rapidly form. Naturally, each such region on its own, prior to The formation of a compact globule, is insufficiently stable; therefore, the corresponding secondary structure elements constantly form and break apart, or "fluctuate" (Fig. 6.12).

Next, these regions of relatively organized structure—which collectively make up a polypeptide fragment of 50–150 Amino Acids (the future protein domain)—effectively "collapse" driven by non-covalent interactions. This produces the so-called molten globule, which closely resembles the native spatial structure overall while differing in being less compact and having a different arrangement of amino acid side chains. This is followed by structural refinement at the domain level, after which contact between the domains is established. These relatively slow stages complete the formation of the protein globule, rendering it active.

Fig. 6.12. Conformational states of the 69–87 fragment corresponding to the C-terminal a-helix of myohemerythrin (based on NMR data).

a — set of unstable conformations adopted by the peptide in Water; b — preferred conformation of the peptide in a less polar medium (a water-trifluoroethanol mixture), where the C-terminal region forms an α-Helix; c — conformation of the peptide within the compact structure of myohemerythrin, where the α-helix spans the entire sequence. Numbers indicate the positions of residues in the peptide (a, b) and the corresponding fragment of myohemerythrin (c).

In light of the discussed data on in vitro protein renaturation, it is evident that protein folding, predetermined by its primary structure, can in many cases be reproduced with a sufficiently high yield. In addition to the kinetic limitations considered above, one must also take into account that in vivo, the spatial structure is normally formed by each newly synthesized molecule individually, without the accumulation of unfolded, denatured structures.

The Role of the latter factor has become particularly apparent during the overproduction of foreign proteins in microbial Cells. Such systems frequently exhibit the accumulation within producer cells of so-called inclusion bodies, which are aggregates of denatured peptide chains. If the protein contains disulfide bonds, they form haphazardly within such aggregates, often in an intermolecular fashion. The formation of these aggregates can be explained by the fact that during intensive overproduction, structured-free peptide chains are simultaneously generated on numerous Ribosomes at a rate that far outpaces the folding process. As a result, the molecule fails to 'sample' intermediate conformations and find its folding pathway in time. Instead, it aggregates with neighboring, still-unfolded polypeptide chains.

As a rule, the separation of inclusion bodies from other cellular components proceeds smoothly, which facilitates their purification. However, this is followed by the formidable challenge of protein renaturation, for which a satisfactory solution is often elusive. A standard approach involves first dissolving the aggregated protein and achieving complete aggregate disruption using concentrated urea or guanidine hydrochloride in the presence of disulfide-reducing compounds (e.g., mercaptoethanol). This is followed by a more or less gradual removal of the denaturing agent and reducing agent to establish conditions for progressive protein renaturation, while maintaining a low protein concentration. To promote the formation of disulfide bonds, a mild redox system is employed, such as a mixture of 10 mM oxidized and 1 mM reduced Glutathione. In a number of cases, this method yields excellent results; for instance, the renaturation of interleukin-2—whose polypeptide chain consists of 133 Amino Acids and contains a single disulfide bond and one 'unpaired' cysteine residue—proceeds with nearly quantitative yield.

The renaturation of Membrane Proteins presents a distinct challenge. The transmembrane protein Bacteriorhodopsin, following extraction from the membrane and chromatographic purification in an organic solvent mixture, was successfully renatured by placing it into a mixture of phospholipid and a surfactant. Evidently, even in this case, the self-assembly of the protein in the presence of lipid guides the denatured molecule toward a structure folded identically to the native protein.

Thus, protein renaturation is fundamentally possible and practically feasible in many cases; however, a series of obstacles—predominantly kinetic—can severely hinder or occasionally completely halt this process. It must be emphasized that the intrinsic capacity of a protein polypeptide chain to spontaneously form a spatial structure does not imply that this process occurs under real biosynthetic conditions without the involvement of external factors, including various proteins that act as catalysts for polypeptide chain folding.

One reason for the delayed formation of the protein's spatial structure in vivo could be a lag in the establishment of disulfide bonds. There exists a specialized 'shuffling' enzyme—protein disulfide isomerase—that accelerates the sampling of possible disulfide bonds and thereby facilitates the search for correctly formed ones. In its presence, renaturation proceeds much faster. For instance, 70% renaturation of pancreatic ribonuclease is achieved within just 30 minutes under the action of this enzyme isolated from Liver, whereas spontaneous reactivation during the same timeframe is so minimal as to be undetectable.

In Eukaryotic cells, protein disulfide isomerase is localized in the Endoplasmic reticulum and is a dimeric protein whose subunits each contain two domains structurally similar to thioredoxin, a small redox-active protein. Each of these domains presumably houses the catalytic center, which features a pair of cysteine residues (in the reduced form) capable of cleaving a disulfide bond in the substrate protein while forming its own disulfide bond in the process (the oxidized form). Conversely, the oxidized form of protein disulfide isomerase, by reacting with free cysteine SH-groups in the substrate, induces the formation of a disulfide bond within that protein, thereby driving the rapid sampling of all potential disulfide bridges:

Presumably, thioredoxin itself can play a similar role, with the enzyme-catalyzed disulfide exchange in vivo being coupled to the formation of an increasing number of disulfide bonds in the substrate protein via the reduced glutathione – oxidized glutathione system According to the scheme:

In some cases, the rate-limiting step that restricts the speed of protein renaturation turns out to be the cis-trans isomerization of individual peptide bonds, particularly those formed by the imino group of Proline. As already noted in the Discussion of secondary structure (see Chapter 5), the energy difference between such cis- and trans-peptide bonds is small, making the transition from one isomer to the other during protein denaturation quite probable. The reverse transition from the cis-form—which is generally unfavorable for folding—to the trans-form (the state in which the overwhelming majority of peptide bonds exist in a protein) also requires a certain activation energy. Consequently, a molecule that has not yet completed its spatial structure is forced, as it were, to 'wait' for the favorable isomerization of the proline peptide bond.

There is a widely distributed enzyme known as peptidyl-prolyl cis-trans isomerase (in eukaryotic cells, this is cyclophilin, which specifically binds the immunosuppressant drug cyclosporine). It catalyzes the cis-trans isomerization of proline peptide bonds, thereby accelerating the folding of the chain into a compact structure.

It should be noted that the two aforementioned Enzymes catalyzing protein renaturation still fall far short of providing the folding speeds corresponding to the rate of translation. Research has established that during the in vivo formation of a protein's spatial structure, the polypeptide chain is not left to its own devices; rather, it interacts with a suite of specialized proteins known as chaperones, whose function is to ensure the rapid Discovery of the correct spatial structure.

Several chaperone families are known, being particularly abundant among the so-called heat Shock proteins. The latter are named because they are synthesized by cells in large quantities in response to stressors that hinder efficient tertiary structure folding—most notably, elevated temperatures. However, they are also produced and function under normal physiological conditions.

One chaperone family comprises the so-called 70-kDa stress proteins of eukaryotic cells, which have a molecular mass of approximately 70 kDa. They form complexes with as-yet-unfolded polypeptide chains, preventing their mutual interaction and undesired aggregation. These proteins consist of two interacting domains: one domain forms a complex with segments of the unfolded polypeptide chain, while the other binds ATP and is capable of cleaving it, acting as a slow-acting ATPase. Upon ATP hydrolysis, the protein undergoes a conformational transition, causing its complex with the polypeptide chain to dissociate.

Polypeptide chains destined for export from the Cytoplasm are maintained in an unfolded state—optimally adapted for membrane translocation—through complexation with 70-kDa stress proteins. The lifespan of such complexes is presumably determined by the time required for the intramolecular hydrolysis of ATP bound by the 70-kDa stress proteins (cf. Chapter 12). It remains unclear whether these proteins directly catalyze the Formation of secondary or tertiary structure itself. It is possible that their role is restricted to preventing intermolecular interactions and the aggregation of still-unfolded polypeptide chains, which would otherwise interfere with this process.

Another widely distributed group of proteins involved in tertiary structure formation comprises chaperonins—proteins encoded by the GroEL and GroES genes in E. coli. GroEL proteins are composed of subunits with a molecular mass of ~60 kDa, which assemble into a characteristic quaternary structure built from two stacked rings of seven subunits each. It is hypothesized that a partially folded polypeptide chain resides On the surface of such a ring. Its individual segments, bound by chaperonin subunits with varying affinities, can be released from the complex to form their characteristic secondary structure without Interference from neighboring segments or other polypeptide chains held in a bound state. Under such conditions, the formation of regular secondary structure elements occurs sequentially, so to speak. Upon completion of this process, the complex dissociates—a step dependent, as in the previously considered case, on the hydrolysis of ATP bound to the GroEL protein. GroES proteins, with a molecular mass of 10 kDa, associate with the GroEL protein and somehow regulate its ATPase activity, thereby controlling the lifetime of the complex.