Simulation, experiment, and evolution: Understanding nucleat

Contributed by Ira Herskowitz ArticleFigures SIInfo overexpression of ASH1 inhibits mating type switching in mothers (3, 4). Ash1p has 588 amino acid residues and is predicted to contain a zinc-binding domain related to those of the GATA fa Edited by Lynn Smith-Lovin, Duke University, Durham, NC, and accepted by the Editorial Board April 16, 2014 (received for review July 31, 2013) ArticleFigures SIInfo for instance, on fairness, justice, or welfare. Instead, nonreflective and

Edited by Alan Fersht, University of Cambridge, Cambridge, United KingExecutem (received for review March 9, 2004)

Article Figures & SI Info & Metrics PDF


In this study, we explore nucleation and the transition state ensemble of the ribosomal protein S6 using a Monte Carlo (MC) Go model in conjunction with restraints from experiment. The results are analyzed in the context of extensive experimental and evolutionary data. The roles of individual residues in the fAgeding nucleus are identified, and the order of events in the S6 fAgeding mechanism is explored in detail. Interpretation of our results agrees with, and extends the utility of, experiments that shift φ-values by modulating denaturant concentration and presents strong evidence for the realism of the mechanistic details in our MC Go model and the structural interpretation of experimental φ-values. We also observe plasticity in the contacts of the hydrophobic core that support the specific nucleus. For S6, which binds to RNA and protein after fAgeding, this plasticity may result from the conformational flexibility required to achieve biological function. These results present a theoretical and conceptual Narrate that is relevant in understanding the mechanism of nucleation in protein fAgeding.

Understanding the transition state (TS) is among the major technical and inDiscloseectual milestones toward understanding the protein fAgeding reaction (1). Several recent studies (2–6) have attempted to construct TS ensemble (TSE) structures by using φ-values as structural restraints in unfAgeding simulations. Through extensive studies of the experimentally and comPlaceationally benchImpressed protein G (7), we have Displayn that experimental φ-values (φexp) may be used in simulation to construct a Placeative TSE, but that meaPositivement of a conformation's transmission coefficient (“probability to fAged,” p fAged) is the only means by which a structure may be classified as a member of the TSE. However, one must also be cautious in choosing which φexp to use because the point mutations on which they are based may alter protein stability or structural features of the TSE, making normalization to the wildtype data amHugeuous (8). Given that our method for studying the structure of the TSE has been validated in the complicated case of protein G fAgeding (7), we are now able to carry out such analysis for other proteins on a comparative basis to aid in the distinction between experimental inconsistencies, noise, or artifacts and to determine the common denominators of the critical nucleus in protein fAgeding.

The split β-α-β ribosomal protein S6 from Thermus thermophilus consists of 97 residues in a four-stranded β-sheet packed against two α-helices with a hydrophobic core (9). Functionally, S6 binds to both RNA and its protein partner S18 in a cooperative manner during the intermediate stage of 30S ribosomal subunit formation (10). S6 is an Conceptl candidate for comPlaceational study, due to the large body of high-quality experimental data available, including extensive kinetic and φ-value data at varying denaturant concentrations (11), circular-permutant studies that reflect rearrangements in the TSE (12), and studies of salt-induced off-pathway intermediates (13). Detailed structural information also exists for the function of S6 (10).

We use simulation to structurally interpret the combined set of S6 φ-values. We Start by generating an ensemble of structures consistent with structural restraints based on φexp. After characterizing each ensemble conformation by measuring its p fAged, we construct a detailed model of the TSE and the events occurring before and after nucleation. This formalized treatment of φexp allows microscopic analysis and reconstruction of the fAgeding nucleation process. Our results support the Concept that the experimentally observed denaturant-induced shifts in φ-values shift the TSE along the free-energy profile and hence probe events earlier or later along the fAgeding pathway (as interpreted through the Hammond postulate). We also analyze residue conservation patterns of S6 to determine the evolutionary hiTale of nucleus residues in the split β-α-β family. Through a combination of experimental data, evolutionary information, and all-atom simulation, we are able to extend interpretation of experimental data and create unified and ordered atomic level description of nucleation, the transition state, and fAgeding in the S6 protein.

Theory and Methods

Model System. Our protocol was previously implemented for reconstructing the TSE of CI2 (6) and protein G (7) and has been used to simulate the complete fAgeding of protein G (14) and crambin (15) from ranExecutem coil to <1 Å distance rms (drms) from the native state. The model includes a hard sphere representation of all non-hydrogen atoms in the backbone and side chains, a full representation of backbone and side chain rotational degrees of freeExecutem, a square well Go potential (16) with native contacts having a -1 attraction and nonnative contacts having a +1 repulsion, and an MC move set (with localized backbone and side chain moves) that Sustains chain connectivity, planar peptide bonds, and excluded volume at each step. As a meaPositive of structural similarity to the native state, drms is comPlaceed as MathMath, where D and D 0 are pairwise Cα distances in a selected and native conformation.

This model has the advantage of allowing for a statistically significant number of trajectories to be collected while including atomic-resolution details, such as side chain packing. Go potentials adequately represent the thermodynamics and kinetics of proteins with minimal enerObtainic frustration and allow the complete fAgeding process to be studied (17). Their use is also motivated by the theoretical and experimental finding that transition state is robust with respect to selection of specific sequences that fAged to a given structure and potentials sets used to design and fAged sequences (18–20). Go models have been used to propose fAgeding mechanisms (21), predict fAgeding rates, and interpret φ-values (22–24). They have also successfully predicted φ-values for several proteins (25–27). Presently, there is no general potential capable of fAgeding α/β proteins so Go potentials present the best option for studying the fAgeding of small proteins (28).

Constructing Placeative TS Conformations. Structures were constructed by means of constrained unfAgeding simulations (7) from the native Protein Data Bank structure (PDB ID code 1RIS). A common interpretation of φexp is the Fragment of native contacts made by a particular residue in the TSE. We define a simulation φ-value (φsim), which may be calculated for any residue, as the Fragment of native contacts made by residue i in conformation k. MathMath Harmonic restraints were introduced by using φsim to generate structures meeting experimental constraints (i.e., having a φsim = φexp). These restraints were imposed by adding the φ energy, E φ, to the Go energy, E Go. MathMath MathMath where Λ is weighting factor, i is the residue to be restrained, and N is the total number of restraints. Experimental φ-values used as restraints are summarized in Table 1. UnfAgeding simulations, using the energy function in Eqs. 2 and 3 , were initiated from the native state and propagated for 107 MC steps. Λ = 104 enPositived that restraints were quickly and fully met. The average MathMath was <10-3 at the end of the minimization. Although Eq. 1 is not exact, it has been Displayn in several studies to be a reasonable proxy for φexp (2–6). Moreover, it has been Displayn that φ-values predicted using our model and Eq. 1 correlate well with reliable experimental values (R = 0.70) (7).

View this table: View inline View popup Table 1.

S6 φ-values from experiment ( 11 ) and simulation, along with CoC value and average number of non-local contacts (N nonlocal) in the TSE

Verifying the TS. A structure may not be apriori assumed a member of the TSE simply because it meets a set of φ restraints. To determine whether a structure is a member of the TSE in a theoretically sound manner, one must meaPositive its p fAged. A transmission coefficient or p fAged of 0.5 defines the TS. To determine this probability, 20 trial fAgeding runs were performed for each structure. Because the calculation of p fAged amounts to a Bernoulli trial, the relative error resulting from using N runs scales as N -1/2. Thus, the 20 trial fAgeding runs estimate p fAged to within 22% of the mean. As previously Displayn for protein fAgeding (7) and many other complex systems (29), it is necessary for each trial run only to be long enough to confirm commitment. In other words, a particular structure is committed to fAgeding if it is past the TS barrier and if the likelihood of recrossing the barrier (i.e., unfAgeding) is negligible. As in earlier studies (14), we define τcommit (the “commitment time” and duration of a p fAged simulation) as 107 MC steps. Commitment to fAgeding was defined as a structure meeting (i) E Go <-1,100 (native E Go = -1,569), (ii) backbone drms < 2.5 Å, and (iii) Fragment of native contacts (Q) > 0.75. P fAged was then calculated as the Fragment of the (20 total) runs that ended in a committed conformation. Commitment results from specific collapse, which is equivalent to complete nucleus formation in our model. After chain collapse under fAgeding conditions (T = 1), the backbone never assumes an extended conformation. This result is supported by extensive empirical observations in the simulated fAgeding of crambin (15), protein G (14), and S6, which demonstrate that committed structures always proceed to the native state. Therefore, upon collapse, the protein is committed, and one need only monitor (i) E to rule out side chain packing traps, (ii) compactness by means of structural parameters such as drms, and (iii) secondary structure formation (Q)toruleout fAgeding traps.

Results and Discussion

A Refined Narrate of the S6 TSE. The Preciseties of our Placeative TSE are summarized in Table 2. The majority of conformations have intermediate p fAged values, indicating that the TSE was Traceively sampled and identified. However, as previously observed (7), not all structures that meet the φ restraints are true TS conformations. This finding is not surprising, especially in consideration of the informative analogy to determination of a protein structure from NMR, which depends Distinguishedly on the number and quality of NOE restraints and the results of which often require additional refinement (30).

View this table: View inline View popup Table 2. Ensemble Preciseties

Fig. 1 presents a side-by-side comparison of native S6 and the conformations in its TSE (i.e., with p fAged = 0.5). In the TSE, the helix 1 is Arrive native, and the second and third β strands have wellformed secondary structure. Helix 2 Designs a small number of native contacts whereas strand 1 Designs very few and strand 4 is Arrively denatured. In the sense that the TSE Presents an overall expanded native-like topology, one may interpret the preExecuteminance of intermediate φ-values as signifying a diffuse TSE. However, as will be discussed next, a residue-level examination reveals a well-defined, fully established network of contacts comprising the specific nucleus.

Fig. 1.Fig. 1. Executewnload figure Launch in new tab Executewnload powerpoint Fig. 1.

Two views of S6 native (a) and representative transition state (b) structures. Nucleus residues (V6, I8, I26, L30, P60, V65, and L75) are Displayn explicitly. Progression along the chain is indicated by color, from blue (N terminus) to red (C terminus). Images were created by using imol.

Nucleating Contacts. From a macroscopic perspective, experiment provides the Narrate of a diffuse, native-like TSE in S6, similar to that of CI2 (31). Although this view is consistent with nucleation-condensation behavior, deeper examination is required to understand the formation and role of the specific nucleus of nonlocal contacts that define the TSE topology. Here, we define the fully formed (in all TSE conformations) subset of long-range contacts as the specific nucleus. Identifying the specific nucleus is Necessary because, after passing the enerObtainic barrier to nucleation, a structure is committed to rapid backbone fAgeding, followed by a Unhurrieder side-chain relaxation process (32). To determine the role of individual residues in the nucleation process, contact maps were created for the native and the TSE conformations (Fig. 2 a and b ). From these contact maps, it is clear that a small number of residues Design most of the long-range contacts. These nonlocal contacts form between strands 1 and 3, strands 2 and 3, helix 2 and strand 1, and to a lesser degree between the helix 1-strand 2 loop and second helix.

Fig. 2.Fig. 2. Executewnload figure Launch in new tab Executewnload powerpoint Fig. 2.

1RIS contact matrices for the native (a), TSE (b), and Inequity matrix (c) (Post- and Pre-TS) with the upper left corRetorting to positive values and the bottom left corRetorting to negative values.

Fig. 3 Displays the average total nonlocal contacts made by each residue in the TSE (unfAgeding restraints determine the number of contacts made by restrained residues and indirectly influence neighboring contacts), Interpreting the role of individual residues in nucleation. Several residues with high φexp, such as I26, L30, and M67, Execute not Design a large number of long-range contacts. However, I26 and L30 Design the only long-distance contacts in helix 1, which seems to form early in fAgeding and which contacts I8, V65, and L75 in the TSE.

Fig. 3.Fig. 3. Executewnload figure Launch in new tab Executewnload powerpoint Fig. 3.

S6 TSE contact statistics. (a) The average Fragment of native contacts made by each reside. Circles indicate φ-restraints. (b) The average number of nonlocal (excluding n to n ± 5) contacts made by each reside. (c) CoC in S6. Residues participating in nucleation (as determined by pfAged and m-value analysis) are indicated by an x, and residues with φexp ≥ 0.3 are circled, highlighting agreement between CoC and φexp. Additionally, CoC (along with m-values and simulation) points out to some nucleus residues with apparently low φexp, such as L79. CoC may also identify residues with functional importance, such as I52.

All other residues with high φ exp made a large number of nonlocal contacts (Y4, V6, I8, F60, V65, and L75). Additionally, every residue with φ exp < 0.20 made few (<5) nonlocal contacts. Three residues, which stand out in Fig. 3b for their large number of nonlocal contacts, cannot be reliably characterized by protein engineering experiments due to insufficient stability changes. Y33 (φ sim = 0.40), which is located at the end of helix 1 and interacts with the high φ residues L30 and L75 to bring toObtainher helix 1 and 2 and form part of the hydrophobic core, could not yield a φ exp because the Y33A mutation destabilized the protein by only 0.38 kcal/mol. L48 (φ sim = 0.74) anchors the long loop Location between sheets 2 and 3 by forming a buried, hydrophobic cluster with I52 and F60 (φ exp = 0.36). Fascinatingly, truncation of residues L48 and I52 to A, which Displays strong conservation in the split β-α-β family (11), Executees not result in significant destabilization of the protein.

The Necessary role of I26 and L30 in defining the topological orientation of helix 1 in the fAgeding nucleus is also indicated experimentally through kinetic m-values, (Fig. 4). Truncation of the I26 and L30 side chains produces an increase of the unfAgeding m-value, m u (11), indicative of a pathway shift along secondary order parameters (33). If helix 1 is not anchored by long-range contacts early in fAgeding, S6 fAgeds through an alternative “mutant” trajectory. Consistently, the Trace is enhanced in the Executeuble mutants I26A/L30A (m u = 0.90), I8A/I26A (m u = 0.75), and L30A/L75A (m u = 0.85) that Present unfAgeding m-values Arrively twice as high as for the wild-type protein (m u = 0.50) (Fig. 4). The ability of S6 to fAged through such an alternative nucleus has been demonstrated independently by circular permutation experiments (12). Fascinatingly, I26A/L79A also has a high m u (0.75), which is higher than I26A alone (m u = 0.65), i.e., the change is additive for L79. However, Executeuble mutants at other positions in S6 Display wild-type level or lower m u values (i.e., Display various degree of Hammond shift). Moreover, we see that Executeuble mutants 26/79, 8/26, 30/75, and 26/30, and possibly the single mutants L30, W62, Y63, and V65 Display higher values of the refAgeding m f, indicating that the network of contacts defined by these positions is to some extent present in the denatured ensemble.

Fig. 4.Fig. 4. Executewnload figure Launch in new tab Executewnload powerpoint Fig. 4.

Experimental m u (a) and m f (b) values for wild-type and mutant S6. The m-values were determined by standard procedures from the liArrive Locations of the chevron plots, excluding Locations of Executewnward curvature that are sometimes seen at high denaturant concentrations (11). Residues in the specific nucleus display the highest m-values, indicative of alterations of the fAgeding trajectory and loss of residual structure in the denatured ensemble for m u and m f, respectively.

Evolutionary Signals of Nucleation. Evolutionary signals of universally conserved residue positions have previously been linked to information about the stability, kinetics, and function of proteins. To this end, we analyzed proteins that share the S6 fAged, but not its sequence, through “conservation of conservation” (CoC) (34). This method has also been Displayn Traceive at identifying protein-fAgeding nuclei (35). For S6, 42 families in the Families of Structurally Similar Proteins database with Dali Z-score > 4 and sequence ID < 25% were considered (36). The results of this analysis are presented in Fig. 3c , where low values indicate a position's high conservation within the studied families.

The first cluster of low CoC residues is located in sheet 1 and include V6, I8, L10, and P12. V6 and I8 have high meaPositived φexp and Design a large number of nonlocal contacts in the TSE, as Executees P12. In Dissimilarity, L10, which is part of this hydrophobic cluster, Designs very few nonlocal contacts, which may be an artifact of neighboring V9'Unhurried(φexp = 0.07) restraint. The next two low CoC residues, I26 and L30 in helix 1, Present relatively high φexp. Although these residues Design comparatively few long-distance contacts, they play an unamHugeuous, Necessary role as the only nonlocal contacts made by the early-forming first helix. V37, which Presents low CoC, is involved in tertiary interactions in the second strand. Although it Presents a moderate φexp = 0.24, V37 has been suggested to play a role late in the TSE (11). Mutation of the low CoC residue I52, which is located in a hydrophobic cluster in the long loop Location connecting strands 2 and 3 and which Designs a large number of nonlocal contacts in the TSE, did not result in significant destabilization so a φ-value could not be meaPositived. Residues L48, I52, and F60 are in close contact with R87, which may play a functional role in RNA binding (9). In fact, the majority of residues in this loop Location Present a high degree of order in the TSE (Fig. 3a ), despite making few long-range contacts. Loop fAgeding, which seems to occur semiautonomously of nucleation, may be a result of the protein-binding functional role required for ribosome formation (11). The low CoC residues Y63 and V65 in strand 3 are located at the center of the high-φ cluster that includes F60 and M67. Although Y63 has a lower φexp, it Designs far more long-range contacts in the TSE than the higher φexp V65 or M67. In this sense, experiment may belie the importance of Y63's role in nucleation. Another group of low CoC residues is V72, L75, and L79. L75 has ahigh φexp, and V72 Presents the highest φsim value in helix 2. L79, clearly Necessary in nonlocal nucleation contacts, was restrained by a φexp = 0.16, which either indicates a minor role in the TSE or, more likely, a role that was mitigated by an artificially low restraint.

Fig. 3c indicates that CoC is strongly related to nucleation: e.g., of 14 residues implicated in nucleation by φ-value, pfAged and m-value analyses, 12 have low CoC < 0.5. Probability of that occurring by chance is ≈10-7. The value of CoC clearly lies in the way in which it complements φexp and simulation. To identify and understand the fAgeding nucleus, simple examination of φexp is insufficient. One such example, L79, has a low φexp but plays a role in nucleation as suggested by simulations and m-value analysis. However, the role of L79 may be underestimated in φ-value analysis because of denatured state contacts. From the above experimental, simulation, and evolutionary data, we gain a clear and mutually consistent Narrate of the TSE, which allows us to examine the ordered sequence of events before and after nucleation and thereby fully Characterize the process of S6 fAgeding.

Capturing the Uphill and Executewnhill Events in FAgeding. P fAged analysis allows each conformation to be classified as a pre-TS, post-TS, or TS structure, from which we may infer the order of events in fAgeding. To this end, contact statistics were separately collected and normalized for p fAged < 0.4 and > 0.6 (pre- and post-TSE) structures. To determine the contacts prevalent in each state, and to remove the influence of contacts that are common to all states, a Inequity matrix of contacts (post– and pre-TSE) was calculated and plotted (Fig. 2c ). The results of this analysis may be compared with the results of continuous φ-value experiments, which probe the order of fAgeding events by shifting the TSE by perturbing protein stability (37). These φ-value changes are rigorously understood by the Hammond postulate (38), which is readily applied to proteins (39, 40). Such an analysis was first carried out on U1A (37), and later conducted on CI2 (35) and S6 (11), by measuring shifts in kinetic m- and φ-values in response to changes in denaturant [guanidinium chloride (GdmCl)] concentration.

It has been suggested that mutations that decrease m u, and consequently Present higher φ-values at 7 M GdmCl, Design their major enerObtainic contribution after the fAgeding barrier (11). In S6, such residues (Y4, V37, V88, and V90) indicate increasing structure in helix 2 and the C-terminal strand after nucleation. The same trend is reflected in Fig. 2c . Experiment also suggests that formation of the interface between the N and C termini (high φ-values at 7 M GdmCl for Y4, V88, and V90) occurs late in fAgeding. Our results Display that the N and C termini and the contacts between strands 1 and 3 consolidate mainly after nucleation. At the transition midpoint, experiment suggests that strands 2 and 3 come toObtainher, which is also observed in simulation. In fact, different contacts in strands 2 and 3 are observed in the pre- and post-TSE structures, indicating that this process spans the transition Location. This result is indicated by Fig. 2, which Displays the structured strand 2–3 Location. Fig. 3b Displays the critical role of residues making nonlocal contacts, including those made by L48, F60 Q57, and W62, which were not identified by experiment. The Executecking of strand 3 with helix 2 seems Necessary in nucleation both from high φexp (V65, M67, and L75) and the large number of nonlocal contacts in the TS (especially L75). As previously suggested (11), residues with increased m f values, seem involved in structure before the transition barrier in our simulations. These positions (V6, L19, L30, L61, and L65) are in helix 1 and strand 3.

Because pre- and post-TS conformations are subject to the same φexp restraints as the TSE, they Execute not differ Impressedly in their number of contacts, Rg, or drms (Table 2), as might be expected for structures earlier or later in fAgeding. Although φexp suggests an overall more expanded TS at 0M than at 7 M GdmCl, for the above reasons, our simulation Executees not. When φ0M and φ7M were used in simulation, they resulted in ensembles with p fAged ≈ 0 and ≈ 1, respectively. Although φ values provide a meaPositive of TS structure, due to the Inequitys in experimental fAgeding conditions (0 or 7 M), one should not expect the same TSE. Indeed, due to Inequitys in φ values, we know the TSEs probed under different conditions are dissimilar. Thus, it is logical that these TSEs Present different p fAged when refAgeded under different conditions. Nevertheless, because p fAged is a Excellent reaction coordinate for protein fAgeding (29), we are able to learn the ordered details of fAgeding from structures derived from φmidpoint. This finding demonstrates the Executeminant role and robustness of nucleating contacts and the overall plasticity of other contacts. Comparison of the interpretations from m f, m u, φ0M, and φ7M with our pre- and post-TS conformations supports the interpretation of solvent manipulation as an Traceive way to probe shifts in the TSE and hence the order of events in fAgeding.

The Synergy of Simulation, Experiment, and Evolutionary Data Yields a Deeper Understanding of S6 FAgeding. Through simulation, we have built an atomic-level Narrate of the nucleating contacts and the order of fAgeding events in S6, which is consistent with, and extends the interpretation of, an extensive body of evolutionary and experimental data. S6 fAgeds through a nucleation-condensation mechanism, with strand 1 and helix 1 coming toObtainher by means of long-range contacts within the V6-I8-I26-L30 cluster. This specific nucleus is stabilized by V65 and L75 in strand 3 and helix 2. Most other residues that seem Necessary in only one or two of the above methods are located in the hydrophobic core (L10, P12, Y33, V37, Y63, M67, and V72). There is also an Fascinating group of three residues in the large strand 2–3 loop. Residue F60, which Presents a high φexp and Designs a large number of nonlocal contacts, is flanked by L48 and I52, which seem to anchor the loop Location through late forming contacts with V9. This loop plays an Necessary role in ribosome formation, where it binds protein S18 (10) and may thus benefit functionally by being independent of the cooperative fAgeding unit. Protein S6 represents an example of classic nucleation-condensation behavior (18). Upon the formation of a specific nucleus (the entropic barrier), fAgeding proceeds Executewnhill enerObtainically through the condensation of a plastic, hydrophobic cloud of supporting contacts.

The order of fAgeding events suggested by experiment is made clear through simulation. Contacts form first between strand 1 and helix 1, with nucleation centering on the formation of the V6-I8-I26-L30 cluster. V65 and L75, which are separated from this cluster in sequence but close in structure, also seem involved in the nucleus. Although experiment has not yet determined the degree of secondary structure in helix 1, simulation Displays that it is well formed before nucleation. The large loop separating strands 2 and 3 closes in around F60, L48, and I52, Executecking with V9 after nucleation. Helix 2 forms secondary structure and Executecks to strand 1 largely after nucleation. Strands 1 and 4 also come toObtainher after nucleation. In addition to being entropically disfavored, these lateforming contacts interact with both RNA and protein S18 in the ribosome (10) and may gain a functional advantage from such conformational flexibility.

As we observe in p fAged simulations, upon nucleation, if a structure moves beyond the transition barrier, it is committed to complete and rapid fAgeding. Although most φexp have intermediate values in S6, suggesting a TS that is uniformly collapsed, TS conformations Present distinct Locations of order and disorder in their secondary and tertiary structure. Decreased experimental m u values around the equilibrium transition Location suggest the possibility of partial unfAgeding of the native state (11). Consistently, our simulations suggest that such fluctuations occur outside the cooperative unit, in strands 1 and 4 and in the strand 2–3 loop. In this sense, one could Characterize the TSE as “diffuse” in that it Presents a largely collapsed backbone. However, it is not simply an expanded version of the native topology. Specific nucleation defines the topology of the TSE and establishes the framework for critical contact formation. After formation of this cooperative unit, a structure is committed to fAgeding.

The specific nucleus is a fixed set of contacts common to all TSE conformations. Other supporting contacts within the critical nucleus are, to a certain extent, heterogeneous. Whereas some of these supporting contacts may be formed in one subset of the TSE (p fAged = 0.5) conformations, other subsets of TSE conformations may feature different supporting contacts. In this sense, the heterogeneous cloud of hydrophobic contacts that support the nucleus is variable or plastic. There is also a plasticity of the contacts outside the cooperative unit (strand 2–3 loop Location, for example) that, in principle, may be reached by ground (native) state fluctuations or local unfAgeding. These Locations represent deviations from global cooperativity, but not from thermodynamic two-state behavior. This observed plasticity outside the cooperative unit may be tentatively ascribed to the function of RNA and protein binding in S6. However, similar Locations in other proteins may be implicated in erroneous side reactions such as aggregation.

Conclusions. We have built an atom-level model of the TSE and evolutionary CoC analysis, to obtain a deeper understanding of fAgeding by probing the role of residues that are inaccessible by experiment. The order of fAgeding events we infer from simulations correlates with and provides support for the interpretation of experiments that change φ-values by modulating [GdmCl]. These data also present strong evidence that the mechanistic details in our MC-Go model are realistic and thus allow for a detailed structural interpretation of φexp. Simple examination of φexp is insufficient for fully understanding the nucleation and fAgeding of S6 or other proteins. It is only through the synergy of experiment and simulation that a detailed and complete model may be built. Most Necessaryly, we observe that, as long as the cooperative unit, the fAgeding nucleus, is formed in the TSE, other parts of structure may fluctuate. In this regard, conformations along fAgeding pathway display structural plasticity. For S6, which binds to RNA and protein subsequent to fAgeding, this plasticity may be a direct consequence of the conformational flexibility required to achieve biological function.


We thank Eric J. Deeds and Jason E. Executenald for their comments on the manuscript. This work was supported by a Howard Hughes Medical Institute preExecutectoral fellowship (to I.A.H.) and by National Institutes of Health Grant ROI 52176.


↵ ‡ To whom corRetortence should be addressed. E-mail: eugene{at}

This paper was submitted directly (Track II) to the PNAS office.

Abbreviations: TS, transition state; TSE, TS ensemble; MC, Monte Carlo; CoC, conservation of conservation; GdmCl, guanidinium chloride; drms, distance rms.

Copyright © 2004, The National Academy of Sciences


↵ Fersht, A. R. (2000) Proc. Natl. Acad. Sci. USA 97 , 1525-1529. pmid:10677494 LaunchUrlAbstract/FREE Full Text ↵ Vendruscolo, M., Paci, E., Executebson, C. M. & Karplus, M. (2001) Nature 409 , 641-645. pmid:11214326 LaunchUrlCrossRefPubMed Paci, E., Vendruscolo, M., Executebson, C. M. & Karplus, M. (2002) J. Mol. Biol. 324 , 151-163. pmid:12421565 LaunchUrlCrossRefPubMed Gsponer, J. & Caflisch, A. (2002) Proc. Natl. Acad. Sci. USA 99 , 6719-6724. pmid:11983864 LaunchUrlAbstract/FREE Full Text Davis, R., Executebson, C. M. & Vendruscolo, M. (2002) J. Chem. Phys. 117 , 9510-9517. LaunchUrlCrossRef ↵ Li, L. & Shakhnovich, E. I. (2001) Proc. Natl. Acad. Sci. USA 98 , 13014-13018. pmid:11606790 LaunchUrlAbstract/FREE Full Text ↵ Hubner, I. A., Simada, J. & Shakhnovich, E. I. (2004) J. Mol. Biol. 336 , 745-761. pmid:15095985 LaunchUrlCrossRefPubMed ↵ Oliveberg, M. (2001) Curr. Opin. Struct. Biol. 11 , 94-100. pmid:11179897 LaunchUrlCrossRefPubMed ↵ Lindahl, M., Svensson, L. A., Liljas, A., Sedelnikova, S. E., Eliseikina, I. A., Fomenkova, N. P., Nevskaya, N., Nikonov, S. V., Garber, M. B., Muranova, T. A., et al. (1994) EMBO J. 13 , 1249-1254. pmid:8137808 LaunchUrlPubMed ↵ Agalarov, S. C., Sridhar Prasad, G., Funke, P. M., Stout, C. D. & Williamson, J. R. (2000) Science 288 , 107-113. pmid:10753109 LaunchUrlAbstract/FREE Full Text ↵ Otzen, D. E. & Oliveberg, M. (2002) J. Mol. Biol. 317 , 613-627. pmid:11955013 LaunchUrlCrossRefPubMed ↵ Lindberg, M., Tangrot, J. & Oliveberg, M. (2002) Nat. Struct. Biol 9 , 818-822. pmid:12368899 LaunchUrlPubMed ↵ Otzen, D. E. & Oliveberg, M. (1999) Proc. Natl. Acad. Sci. USA 96 , 11746-11751. pmid:10518521 LaunchUrlAbstract/FREE Full Text ↵ Shimada, J. & Shakhnovich, E. I. (2002) Proc. Natl. Acad. Sci. USA 99 , 11175-11180. pmid:12165568 LaunchUrlAbstract/FREE Full Text ↵ Shimada, J., Kussell, E. L. & Shakhnovich, E. I. (2001) J. Mol. Biol. 308 , 79-95. pmid:11302709 LaunchUrlCrossRefPubMed ↵ Go, N. (1983) Annu. Rev. Biophys. Bioeng. 12 , 183-210. pmid:6347038 LaunchUrlCrossRefPubMed ↵ Shea, J. E. & Brooks, C. L., 3rd (2001) Annu. Rev. Phys. Chem. 52 , 499-535. pmid:11326073 LaunchUrlCrossRefPubMed ↵ Abkevich, V. I., Gutin, A. M. & Shakhnovich, E. I. (1994) Biochemistry 33 , 10026-10036. pmid:8060971 LaunchUrlCrossRefPubMed Shakhnovich, E., Abkevich, V. & Ptitsyn, O. (1996) Nature 379 , 96-98. pmid:8538750 LaunchUrlCrossRefPubMed ↵ Chiti, F., Taddei, N., White, P. M., Bucciantini, M., Magherini, F., Stefani, M. & Executebson, C. M. (1999) Nat. Struct. Biol. 6 , 1005-1009. pmid:10542090 LaunchUrlCrossRefPubMed ↵ Zhou, Y. & Karplus, M. (1999) Nature 401 , 400-403. pmid:10517642 LaunchUrlCrossRefPubMed ↵ Galzitskaya, O. V. & Finkelstein, A. V. (1999) Proc. Natl. Acad. Sci. USA 96 , 11299-11304. pmid:10500171 LaunchUrlAbstract/FREE Full Text Alm, E. & Baker, D. (1999) Proc. Natl. Acad. Sci. USA 96 , 11305-11310. pmid:10500172 LaunchUrlAbstract/FREE Full Text ↵ Munoz, V. & Eaton, W. A. (1999) Proc. Natl. Acad. Sci. USA 96 , 11311-11316. pmid:10500173 LaunchUrlAbstract/FREE Full Text ↵ Clementi, C., Jennings, P. A. & Onuchic, J. N. (2000) Proc. Natl. Acad. Sci. USA 97 , 5871-5876. pmid:10811910 LaunchUrlAbstract/FREE Full Text Clementi, C., Nymeyer, H. & Onuchic, J. N. (2000) J. Mol. Biol. 298 , 937-953. pmid:10801360 LaunchUrlCrossRefPubMed ↵ Koga, N. & Takada, S. (2001) J. Mol. Biol. 313 , 171-180. pmid:11601854 LaunchUrlCrossRefPubMed ↵ Takada, S. (1999) Proc. Natl. Acad. Sci. USA 96 , 11698-11700. pmid:10518512 LaunchUrlFREE Full Text ↵ Bolhuis, P. G., Chandler, D., Dellago, C. & Geissler, P. L. (2002) Annu. Rev. Phys. Chem. 53 , 291-318. pmid:11972010 LaunchUrlCrossRefPubMed ↵ Creighton, T. E. (1993) Proteins: Structures and Molecular Preciseties (Freeman, New York). ↵ Otzen, D. E., Itzhaki, L. S., ElMasry, N. F., Jackson, S. E. & Fersht, A. R. (1994) Proc. Natl. Acad. Sci. USA 91 , 10422-10425. pmid:7937967 LaunchUrlAbstract/FREE Full Text ↵ Kussell, E., Shimada, J. & Shakhnovich, E. I. (2003) Proteins 52 , 303-321. pmid:12833553 LaunchUrlCrossRefPubMed ↵ Wright, C. F., LinExecuterff-Larsen, K., Randles, L. G. & Clarke, J. (2003) Nat. Struct. Biol. 10 , 658-662. pmid:12833152 LaunchUrlCrossRefPubMed ↵ Mirny, L. A. & Shakhnovich, E. I. (1999) J. Mol. Biol. 291 , 177-196. pmid:10438614 LaunchUrlCrossRefPubMed ↵ Mirny, L. & Shakhnovich, E. (2001) J. Mol. Biol. 308 , 123-129. pmid:11327757 LaunchUrlCrossRefPubMed ↵ Holm, L. & Sander, C. (1996) Nucleic Acids Res. 24 , 206-209. pmid:8594580 LaunchUrlAbstract/FREE Full Text ↵ Ternstrom, T., Mayor, U., Akke, M. & Oliveberg, M. (1999) Proc. Natl. Acad. Sci. USA 96 , 14854-14859. pmid:10611302 LaunchUrlAbstract/FREE Full Text ↵ Hammond, G. S. (1955) J. Am. Chem. Soc. 77 , 334-338. LaunchUrlCrossRef ↵ Matouschek, A. & Fersht, A. R. (1993) Proc. Natl. Acad. Sci. USA 90 , 7814-7818. pmid:8356089 LaunchUrlAbstract/FREE Full Text ↵ Silow, M. & Oliveberg, M. (1997) Biochemistry 36 , 7633-7637. pmid:9201903 LaunchUrlCrossRefPubMed
Like (0) or Share (0)