The case for defined protein folding pathways

Foldons and Defined Pathways.

To investigate protein folding mechanisms, experimentalists have exploited every available methodology to define intermediate forms between the unfolded and native states. The major problem is that kinetic folding intermediates have short lifetimes and cannot be isolated for structural studies. Widely used methods that can follow the folding process in real time, mostly spectroscopic in nature, provide kinetic but little structural information.

Hydrogen exchange (HX) and related methods (21, 27) have made it possible to obtain informative structural information on intermediate states and their formation and progression in folding and unfolding pathways. Early HX work found that cytochrome c (cyt c) (28) and other proteins (29) reversibly unfold and refold at a low level under native conditions. HX-NMR studies were able to divide the equilibrium unfolding of cyt c into an energy ladder of reversible partially unfolded forms (PUFs; Fig. 1B) (15). Mutational and stability modification experiments in both kinetic (30) and equilibrium (31–34) modes established the identity of the high-energy PUFs, as shown in the ribbon diagrams in Fig. 1B. Reading upward in free energy from the native state in Fig. 1B, the identity of the cyt c PUFs is infrared alone unfolded (large bottom loop), then infrared + red unfolded, then those two + yellow unfolded, then those three + green, and finally the additional unfolding of blue, producing the fully unfolded protein (cyt c color coding as in Fig. 1B). Each energetically upward step unfolds one more foldon unit of the native protein. In light of microscopic reversibility, the sequential unfolding ladder defined under native conditions suggested that cyt c might normally fold in a similar stepwise sequence in reverse order down the free energy ladder (Fig. 1B).

A recently advanced HX mass spectrometry experiment (HX MS) (35–38) is uniquely able to define the structure of populated intermediates during kinetic folding. In this experiment a denaturant-unfolded protein is diluted into folding conditions. Folding synchronously commences. After various test folding times the sample is subjected to a brief D-to-H HX pulse to label, in a structure-sensitive way, partially folded transient intermediates that may be present (39). The protein sample is quenched (low pH, cold) to halt further exchange, proteolytically fragmented into many peptides, and the many peptides are roughly separated by HPLC and then further separated and analyzed at high resolution by MS. One result is that the same sequence of cyt c foldon-dependent intermediate forms seen in the equilibrium-uphill unfolding experiments form in reverse order as on-pathway intermediates in downhill kinetic folding, as illustrated in Fig. 1B (40).

The example in Fig. 2 shows that RNase H similarly folds through an ordered sequence of native-like foldon units (41). Each of the 156 overlapping peptides obtained in the HX MS experiment provides a time- and structure-resolved series of snapshots of the fraction of the protein population that is already protected (folded) against the D-to-H labeling pulse (heavier) and not yet protected (lighter) at each segmental position at the time of the HX labeling pulse. Fig. 2 A–D show time-dependent HX MS data for the folding of four peptides that represent known protein segments within four distinct RNase H foldon units. The bimodal MS envelopes for each peptide show that the protein segment that it represents transits from an unfolded (unprotected) to a folded (protected) condition in a concerted reaction. The same behavior is seen throughout the protein. The entire protein population folds energetically downhill by forming a first structural unit (blue) in much less than 9 ms, then a second (green) with ∼5-ms lifetime, and so on for yellow and red (Fig. 2 E and F). Each segment once folded remains folded as later foldons add on, tracing a sequential folding process.

As for cyt c, these results document a well-defined, linear, stepwise folding pathway. In the case of heterogeneous pathways expected for many-pathway behavior, different molecules would fold any given segment in different orders at different rates. Rather, these experiments reveal concerted population-wide intermediate formation in a given sequence, each at the same given rate. The entire protein population folds the same segment first, all at the same rate, and all experience the same subsequent steps in the same order at the same rate, as suggested in Fig. 1B and detailed in Fig. 2 A–D.

Much experimental work using HX and other methods indicates that other proteins all share the same behavior. Maltose binding protein first forms a native-like on-pathway intermediate that brings together sequentially distant segments from its two domains and then more slowly folds the rest of the protein (42). Bai and coworkers (43) used HX-NMR and mutational methods to define a foldon-dependent sequential folding pathway for apocytochrome b₅₆₂. Georgescauld et al. (44) used HX MS to determine a structurally defined folding pathway for a TIM barrel protein (DapA) when it is encapsulated inside the GroEL chaperonin. Similar results have been found for other proteins including apomyoglobin (45, 46), apoflavodoxin (47), OspA from Borrelia burgdorferi (48), staphylococcal nuclease (49), and a β-sandwich FHA domain (50). Silverman and Harbury used a sulfhydryl reactivity method to demonstrate foldons and their role in the sequential equilibrium unfolding of triose phosphate isomerase (51). Kay used relaxation dispersion NMR to define foldons in several small proteins and infer their on-pathway nature (52).

In summary, a quantity of direct experimental evidence for different proteins using various methods authenticates the reality and the ubiquity of cooperative foldon units as building blocks of native proteins and the stepwise formation of the same foldons in fairly well-defined folding pathways. These and other experiments would have detected multiple structure-formation pathways if they existed.

The Many-Pathway Model.

Although various observations have been thought to be consistent with the many-pathway model, there has been no clear experimental demonstration that characterizes or even detects the many trajectories. The one case to do so is analyzed here.

Ritchie and Woodside (18) and Woodside and Block (53), using high time-resolution single-molecule force spectroscopy, were able to detect many transitions between the unfolded and folded states of a monomeric prion (54) and also between the unfolded state and the first folded state on the (mis)folding pathway of a dimerized prion (19, 20). The prion protein was held in a dual-trap optical tweezers instrument under a constant pulling force, poised at the midpoint of a folding–unfolding transition so that many dynamic transitions in both directions could be observed. The experiment was able to time-resolve not only unfolding and refolding rates (Fig. 3 A and B) but also the more fleeting transition paths as individual molecules climb the transition barrier (Fig. 3 C and D). Fig. 3A is a recording of dwell times in the unfolded and folded states of the monomeric prion (54), which folds in a fast two-state manner rate-limited by an initial barrier-crossing event. The initial pathway step is rate-limiting because subsequent steps on the folding pathway are even faster. As expected, folding times are stochastically distributed, some faster and some slower, and the number distribution of dwell times fits a single exponential decay (Fig. 3B), indicating the two-state folding rate constant.

The tethered dimer prion protein folds more slowly to some nonnative state through a distinct pathway with three detectable intermediates (19). Apparently, the two adjacent prion molecules are easily entangled, perhaps by domain swapping, and new kinetic barriers for structure formation are inserted. Neupane et al. (20) focused on the first step in this pathway and were able to measure the transition path time (i.e., the time spent in mounting the initial free energy barrier). Fig. 3C illustrates one unfolding transit and one folding transit. The transition path time is the time required to jump between the two blue dashed lines, which mark the dual trap extensions taken to represent the unfolded condition of the dimeric prion and its first partially folded state. The time for many such transits was measured. As before, some are faster and some slower. Their time distribution is in Fig. 3D. Other details are not distinguishable.

In Fig. 3D, the absence of measured transits at zero transit time and the subsequent upsweep of the distribution curve are due to instrument response time, the finite time needed to accomplish even the fastest start-to-end crossing between the two positions marked in Fig. 3C, and other factors. The later time bins (downsweep) capture the distribution of transition path times that are not limited by the exigencies of measurement and the initial built-in delay function. Calculation predicts this kind of up-and-down transit-time distribution (55). The entire measurable downsweep is well fit by a single exponential (Fig. 3E), indicating a given population rate constant rather than the heterogeneous population expected for the many-pathway model. The same experiment performed with a short DNA hairpin found multiple transition path trajectories with a time distribution like Fig. 3D and a single population rate constant (20).

In a commentary, P. G. Wolynes (56) equates the varied transition paths and path times with the many disparate folding pathways proposed in ELT and supplies a figure to emphasize the concept of pathway heterogeneity. He states that the multiplicity and heterogeneity of the trajectories confirms some of the most basic notions of ELT, and that there is no evidence for an obligate single folding pathway. This view ignores that the trajectories observed proceed only to formation of an initial distinct intermediate on a reproducible stepwise pathway through later distinct intermediates (19).

Further, the variability that does exist among the many trajectories seems to have no functional impact. For heterogeneity of the kind that is integral to meaningful ELT considerations (Fig. 1A), one expects a range of distinguishable folding rates and transition times as if for a mixture of different proteins, as implied in the commentary. In fact, the distribution of transit times measured for the dimeric prion, as for the folding times of the monomeric prion, shows a simple stochastic distribution as expected for a homogeneous population. All molecules reach the same target structure within the maximally narrow time distribution expected for a given population rate constant. Their different trajectories undoubtedly explore many different bond rotational configurations in different time order, but these differences have no physically meaningful effect in respect to rate constants, pathways, or intermediate states.

In summary, the single-molecule experiments of Ritchie and Woodside (18) and Woodside and Block (53) were finally able to measure the multiple parallel trajectories that anchor energy landscape considerations. The results show that trajectory multiplicity and variability do exist but the distribution of folding rates (Fig. 3B) and transition path rates (Fig. 3E) is as expected for a kinetically homogeneous population. Especially interesting, the multiple trajectories carry all of the dimer prion molecules only to an initial distinct intermediate state, all with the same homogeneous population rate constant, which is then followed by other distinct intermediates in a distinct pathway (19). These results do not match the basic characteristics expected for the heterogeneous many-pathway model but are fully consistent with the defined-pathway model.

Energy Considerations.

A critical feature of the funneled ELT model is that the many-pathway residue-level conformational search must be biased toward native-like interactions. Otherwise, as noted by Levinthal (57), an unguided random search would require a very long time. How this bias might be implemented in terms of real protein interactions has never been discovered. One simply asserts that natural evolution has made it so, formulates this view as a so-called principle of minimal frustration, and attributes it to the shape of the funneled energy landscape. Proteins in some unknown way “know” how to make the correct choices.

A calculation by Zwanzig et al. (58) at the most primary level quantifies the energy bias that would be required. In order for proteins to fold on a reasonable time scale, the free energy bias toward correct as opposed to incorrect interactions, whatever the folding units might be, must reach 2 kT (1.2 kcal/mol). The enthalpic bias between correct and incorrect interactions must be even greater, well over 2 kcal/mol, because competition with the large entropic sea of incorrect options is so unfavorable. Known amino acid interaction energies, less than 1 kcal/mol (59), seem to make this degree of selectivity impossible at the residue–residue level.

In contrast, the foldon hypothesis is able to satisfy the requirement for a large energetic bias toward native-like interactions (58) because it operates at a more macroscopic level that naturally adds the energies of many native interactions at each important step. Each foldon formation step is definitively demarcated not by individual weak residue-level interactions but by the concerted sum of the energies of multiple cooperatively organized stereochemically native-like intrafoldon interactions. Each sequential pathway step is determined by collectively summed stereochemically native-like interfoldon interactions.

The natural tendency of partially folded forms toward energy minimization seems likely to additionally entrain not-yet-folded residues in stabilizing nonnative ways. For example, in cyt c folding, the first-formed PUF (blue foldon folded, Fig. 1B) is more stable than the unfolded state by almost 3 kcal/mol in free energy (60), even though the isolated blue foldon is much less stable (61, 62). In RNase H folding the first two foldons seem unlikely to be stable in isolation (see model in Fig. 2).

In summary, many microscopic-level folding trajectories undoubtedly exist, especially at the earliest stage of folding, but conformational searching at this level seems unable to account for the energy bias or the structural selectivity that is needed to support native-directed choices, either at this stage or in subsequent structure formation. The more macroscopic foldon-dependent model combines cooperative sets of interactions that are stereochemically native-like. They naturally recognize native choices and can account for the degree of energy bias required to choose them.

Foldon Evolution.

Experiment shows that proteins unfold and refold by stepping through the cooperative subglobal units that compose them. This behavior seems definitive and widespread. How can one rationalize this situation? One general possibility is that folding in steps is an unavoidable consequence of protein structural cooperativity, a manifestation of the same kind of cooperative interactions that give rise to secondary structure. Another is that this folding strategy is such an effective way to overcome the difficulties of the folding process that biological evolution has over time discovered and widely incorporated it.

A more specific suggestion comes from recent work on repeat proteins. Over the last 15 y, one has become aware of families of repeat proteins that make up as much as 5% of the global proteome. These proteins have a nonglobular body plan made of small repeated motifs in the 20–40 residue range that are assembled in a linear array (63–65). The repeat units are individually cooperative, have low stability, and are further stabilized by their interfacial interactions. Repeat proteins have been found to fold in repeat-sized steps, one or several units at a time, in a distinct pathway order (66, 67). Thus, the repeat units look like and act like known foldons in contemporary globular proteins.

The different families of repeat proteins are very different in detailed structure but within each family the repeats are topologically nearly identical. These observations suggest that repeat proteins arose through repeated duplication at an early stage in the evolution of larger proteins from smaller fragments (68–70). Available examples show that globular organization can arise from continued repetitive growth that closes the linear geometry, and by the fusion of nonidentical units (69, 71), and so would carry forward their foldon-like properties.

The utility of foldons for the efficient folding of proteins might be seen as a dominant cause for the development and retention of a foldon-based body plan through protein evolution. In this view, contemporary proteins came so consistently to their modular foldon-based design and their foldon-based folding strategy because these linked characteristics coevolved. However, the fact that many known foldons bring together sequentially remote segments requires, at the least, some additional mechanism.

Computer Simulations.

One has long looked forward to the time when computer simulations would illuminate the intimate details of protein folding processes. The effort has been obstructed by the lack of formal equations that can capture the delicate balance of multiple interlocking structural interactions and by the immense computer power needed to simulate protein folding in atomistic detail (72). Much of the theoretical folding literature describes efforts to overcome these limitations. Crucial advances have been accomplished with the design of the Anton generation of computers (23), and massively parallel computing (26), and other promising strategies (73). Microlevel dynamics continue to be seen but they do not energetically drive or conformationally direct the course of folding. A key finding is that proteins fold by putting their structural elements into place over and over again in the same reproducible sequence (24, 74). Whether or not these elements formally qualify as typical foldons, the very same physical chemical principles that govern the defined stepwise folding model seem to be in play.

Theoretical efforts to search for foldons have been few. Soon after their discovery, Panchenko et al. (75) scanned the sequence database to search for putative foldons using a criterion of uncertain provenance (ratio of a contiguous segment’s energetic stability gap to the energy variance of that segment’s molten globule states). It may be useful to note that half of known foldons incorporate different segments that are not contiguous. A partially successful effort used a course-grained Gō model simulation with no side chains to compute some of the known foldons in cyt c (76). These approaches have not been revisited.

The Energy Landscape Model.

When the many-pathway model was formulated in the period 1988–1995, the accepted view was that conformational searching for native structure occurs at the microscopic amino acid level. Levinthal (57, 77) had contributed the seminal observation that a random search could not account for known folding rates. The funneled landscape proposal accepted the microscopic search view and so was forced to adopt the suggestion that the conformational search is not random. The search must be biased to select correct native-like interactions in a way that, it was expected, would soon be explained. However, how this propensity might be encoded in the physical chemistry of protein structure has never been discovered. One simply asserts the general proposition that it is encoded in the shape of the landscape and to an ad hoc principle named minimal frustration imposed by natural evolution (78).

Energy landscapes are commonly drawn to illustrate the physical chemistry and quantum mechanics that are built into molecules and govern their reaction rates and behaviors (79). Atypically, the funneled landscape emblematic of energy landscape theory does not deal with molecular properties that would serve to guide interactions. It portrays some external thermodynamic constraints that are valid for the folding of proteins, RNA, or any other polymer. It contains in itself no molecular information or molecule-based constraints or predictions.

One does not doubt that the different proteins in a refolding population will explore many microscopically different configurations. This is characteristic for any thermally driven diffusional process when viewed at a sufficiently microscopic level. The experimental measurement of microscopic trajectories early in folding, now achieved in single-molecule studies (20), displays their reality but shows that the initial folding step exhibits a single homogeneous population rate constant. The trajectory diversity has no influence. Further, residue-level searching does not account for the subsequent structure-formation pathway. The HX MS experiment applied to several proteins (e.g., Fig. 2) would have detected the presence of many alternative pathways during structure formation, as would other folding studies, but they have not. An analogous situation seems to hold even for small apparently two-state folding proteins (24, 52). (Suggestions in the experimental literature for two or so alternative pathways do not constitute a great multiplicity and seem to require some other interpretation.) Quantitative evaluation described above shows that individual residue–residue interaction energies are inadequate for selecting native-like interactions in competition with the large number of competing nonnative alternatives. The assertion that the needed degree of energetic bias is supplied by the shape of an indefinite energy landscape because nature has made it so is—plainly said—not a useful physical–chemical explanation.

These considerations indicate the lack of an influential role for the heterogeneous microscopic search imagined in Fig. 1A. The question is what kind of conformational searching can explain the processes and pathways that carry unfolded proteins to their native state. The foldon-dependent defined-pathway model directly answers each of these challenges.

The Defined-Pathway Model.

Experimental results described before provide clear evidence for the existence and ubiquity of protein foldons and their determining role in constructing well-defined protein folding pathways. Cooperative foldons have been identified as integral structural units in over a dozen proteins of different sizes and topologies. They have been shown to determine intermediates and pathways in experiments on equilibrium unfolding, on kinetic folding, under varied conditions, and using varied methodologies. The selection of native vs. nonnative interactions as folding proceeds, and thus the construction and order of formation of pathway intermediates, is determined by cooperative intra- and interfoldon interactions that are encoded in the collective energies and stereochemistry of the same sets of interactions that cooperatively stabilize the native protein.

The defined-pathway model requires no new physical chemistry but uses familiar cooperativity principles to solve a demanding problem in a previously unsuspected way. Whole-molecule cooperativity is subdivided into smaller but still cooperative units, and kinetic folding recapitulates this foldon-based construction. The same stereochemical information and collective energetics that determine the equilibrium native structure of any given protein also specify the kinetic folding pathway for getting there. These are the factors that make fast definitive protein folding possible. Evolutionary considerations credibly tie together the early codevelopment of foldon-based equilibrium structure and foldon-based kinetic folding.

Does the defined-pathway model guarantee an absolutely single pathway? The MS envelopes in Fig. 2 A–D and other proteins so far determined in this way rule out the presence of a minor pathway that would carry as much as 5% of the folding flux. However, because protein structure and folding pathways are energetically determined, they can be manipulated in obvious ways (mutation, solution conditions, etc.), foldon structures must be somewhat malleable, on-pathway foldon formation may adopt an alternative order, and more than one pathway order may coexist depending on relative stabilities and kinetic barriers.

It has been suggested that the different pathway views discussed here might dominate at different stages in folding. Microscopic many-trajectory behavior is most likely at the earliest prenucleation stage. It may also contribute to subsequent structure-formation steps, although assisted folding guided by prior structure, so-called sequential stabilization (32), must play a dominant role.