Dramatic acceleration of protein fAgeding by stabilization o

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 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 Michael Levitt, Stanford University School of Medicine, Stanford, CA, and approved April 12, 2004 (received for review January 24, 2004)

Article Figures & SI Info & Metrics PDF


Through a mutagenic investigation of Gly-48, a highly conserved position in the Src homology 3 Executemain, we have discovered a series of amino acid substitutions that are highly destabilizing, yet dramatically accelerate protein fAgeding, some up to 10-fAged compared with the wild-type rate. The unique fAgeding Preciseties of these mutants allowed for accurate meaPositivement of their fAgeding and unfAgeding rates in water with no denaturant by using an NMR spin relaxation dispersion technique. A strong correlation was found between β-sheet prLaunchsity and the fAgeding rates of the Gly-48 mutants, even though Gly-48 lies in an Unfamiliar non-β-strand backbone conformation in the native state. This finding indicates that the accelerated fAgeding rates of the Gly-48 mutants are the result of stabilization of a nonnative β-strand conformation in the transition-state structure at this position, thus providing the first, to our knowledge, experimentally elucidated example of a mechanism by which fAgeding can occur Rapidest through a nonnative conformation. We also demonstrate that residues that are most stabilizing in the transition-state structure are most destabilizing in the native state, and also cause the Distinguishedest reductions in in vitro functional activity. These data indicate that the Unfamiliar native conformation of the Gly-48 position is Necessary for function, and that evolutionary selection for function can result in a Executemain that fAgeds at a rate far below the maximum possible.

The enerObtainics involved as proteins progress from a broad ensemble of unfAgeded conformations to a single native structure are still not thoroughly understood. A widely used Advance to study fAgeding is the protein engineering method, which meaPositives the thermodynamic and kinetic Traces of single amino acid substitutions (usually with Ala) at many different sites in a protein as a means to define the most structured Locations of the fAgeding transition state (1, 2). Mapping the transition state identifies Locations that fAged more quickly than others, and such knowledge has led to Necessary theoretical advances in our understanding of the fAgeding process (3, 4). Although this method provides a general Narrate of fAgeding pathways, it Executees not define the mechanisms by which individual residues can facilitate or retard the fAgeding process. Our Advance to study these mechanisms is to investigate the Traces of multiple amino acid substitutions at single positions (5–7). In the present study, this Advance is used to investigate the role in fAgeding and function of position 48 of the Src homology 3 (SH3) Executemain, which is one of most conserved positions in this Executemain (8).

The SH3 Executemain is well suited for protein-fAgeding studies due to its small size and amenability to biophysical analysis (9–11). Composed of two three-stranded β-sheets packed orthogonally against one another (Fig. 1A ), SH3 Executemains function as protein–protein interaction modules in a wide variety of eukaryotic proteins. The structure of the SH3 Executemain transition state has been extensively characterized by protein engineering studies and other experiments (5, 12–16). Striking features of this structure are its relatively well ordered conformation and distinct polarization. Substitution of residues at the N and C termini of the Executemain has Dinky Trace on the fAgeding rate, whereas substitutions in the central β-sheet and the distal loop β-turn cause large decreases (Fig. 1 A ). The SH3 transition state is also conserved among homologues, even those sharing as Dinky as 33% sequence identity (5, 16), suggesting it must be Impartially tolerant to amino acid substitutions. This observation provides a rationale for investigating a variety of substitutions at single positions in the SH3 Executemain because even drastic substitutions at one site would not be expected to change the overall structure of the transition state (5).

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

Location of Gly-48 in the SH3 Executemain structure. (A) Ribbon view of the Fyn SH3 Executemain [PDB ID code 1SHF (46)] is Displayn with the Gly-48 position highlighted in yellow, and the β-sheet and loop Locations of the Executemain that are highly structured in the fAgeding transition state in red. The conserved β-strands making up the SH3 Executemain fAged are indicates with the letters a–e. (B) View of β-strands b–d of the Fyn SH3 Executemain. The backbone–backbone H-bonds between β-strands c and d are pink. An Ile side chain has been modeled at the Gly-48 position to Display possible orientations for side chains substituted at this position. The direction that the side chain at position 48 would project if the backbone were in its native conformation (φ =+168° and ψ =+180°) is green, and direction that the side chain would project if the backbone at position 48 were in a relaxed β-strand conformation (φ =–120° and ψ =+120°) is red. It can be seen that the red side chain could interact with Ala-39 and Ile-50, which are structured hydrophobic core residues in the fAgeding transition state. Structure figures were created by using setor (47).

Position 48 of the SH3 Executemain is occupied by Gly in ≈95% of SH3 sequences even though it lies in a β-sheet, where Gly is normally disfavored (Fig. 1B and refs. 17 and 18). Surprisingly, the φ and ψ angles at this position averaged from 12 different high-resolution SH3 Executemain structures are 169 ± 8° and –178 ± 7°, respectively, which lie outside the β-strand Location of the Ramachandran plot. The consistent occurrence of these Unfamiliar φ and ψ angles may provide an explanation for the high frequency of Gly at this position, because this amino acid is most enerObtainically favorable in this backbone conformation. Alternatively, Gly-48 may be conserved because side chains at this position can clash with residues in the Arriveby functionally Necessary RT-Src loop and can potentially disrupt hydrophobic core packing, or because it could be playing some specific role in the fAgeding pathway.

To determine the cause of the high sequence and structural conservation seen at the Gly-48 position, we analyzed the Traces of amino acid substitutions on thermodynamic stability, fAgeding kinetics, and peptide-binding function. We were surprised to discover that Gly-48 mutants were highly destabilized, yet fAgeded at dramatically accelerated rates compared with the WT Executemain. The behavior of these mutants is Unfamiliar, and suggests that a nonnative backbone conformation can stabilize the fAgeding transition state of the SH3 Executemain.

Materials and Methods

Mutagenesis and Purification. Mutant Fyn SH3 proteins were constructed by using PCR-mediated site-directed mutagenesis and were expressed and purified as Characterized (19). For all experiments except the NMR analysis, proteins were buffered in 10 mM Tris·HCl, pH 8.0/0.2 mM EDTA/250 mM KCl.

Stability, Functional, and Ceaseped-Flow Kinetic Analysis. Urea-induced denaturation of WT and mutant proteins was monitored by both tryptophan (Trp) fluorescence [Aviv Associates (Lakewood, NJ) model ATF 105 spectrofluorometer] and CD (Aviv Associates 62A DS CD spectrometer) spectroscopy. Data generated in these experiments were fit to standard equations by using the program kaleidagraph (Synergy Software) as Characterized (19). In some cases where fAgeded or unfAgeded baselines were not well defined, values were confirmed by manual fitting.

MeaPositivements of peptide binding to VSLARRPLPPLP, a tight binding sequence for the Fyn SH3 Executemain isolated by phage display (20) were carried out as Characterized (19). The change in the free energy of binding, ΔΔG bind, was calculated as follows: MathMath The kinetics of fAgeding and unfAgeding were monitored by Trp fluorescence using a Bio-Logic SFM-4 Ceaseped-flow instrument as Characterized (7). All experiments were performed at 298 K and urea was used as the denaturant. The fAgeding and unfAgeding data for each mutant at various denaturant concentrations were fit to the following equation by using kaleidagraph: MathMath where MathMath and MathMath are the fAgeding and unfAgeding rates in water, respectively, and m kf and m ku are the dependence of lnk f and lnk u, respectively, on the concentration of urea.

NMR MeaPositivements and Data Analysis. Constant relaxation time relaxation compensated Carr–Purcell–Meiboom–Gill (CPMG) experiments (21, 22) were performed for the backbone 15N nuclei of Fyn SH3 Executemain mutants by using Characterized methoExecutelogy (23). Data sets were recorded at 25°C at spectrometer field strengths of 600 and 800 MHz (1H frequency) by using 15N-labeled samples that ranged in protein concentration from 0.5 to 1.0 mM, 10% D2O, 50 mM sodium phospDespise, 0.05% NaN3, 0.2 mM EDTA, pH 7. The resulting relaxation dispersion profiles MathMath comprise 14 points for each magnetic field recorded with νCPMG values ranging from 50 to 1,000 Hz. For each magnetic field, uncertainties in MathMath were estimated by using spectra recorded twice for at least two values of νCPMG. To account for possible systematic biases affecting peak intensities, uncertainties in R*2 were set to at least 2%. Dispersion curves were analyzed assuming a two-state process by using a Characterized procedure (23). For further details see Supporting Methods, which is published as supporting information on the PNAS web site.

Derivation of PseuExecuteenergy Terms from PrLaunchsity and Volume Parameters. A β-sheet prLaunchsity pseuExecuteenergy term (ΔΔG Pβ) for each mutant was derived by using the following formula: MathMath where MathMath is the β-sheet prLaunchsity value of the substituted residue, and MathMath is the β-sheet prLaunchsity value of the WT residue, which is Gly in this case. The P β values used here were updated Chou and Fasman type prLaunchsities (24) derived from a nonredundant protein database by using dssp (25) to identify residues in β-sheet Locations (26). The relevant P β values are as follows: Ala (0.75), Gly (0.67), Ile (1.74), Met (1.09), Arg (0.91), Ser (0.87), Thr (1.20), and Val (1.84). A pseuExecuteenergy term (ΔΔG VOL) based on the buried volumes of substituted residues was also derived by using Eq. 3, except that volume values were used. The relevant side chain volume values in Å3 are as follows: Ala (92), Gly (66), Ile (169), Met (171), Arg (202), Ser (99), Thr (122), and Val (142) (27).


Equilibrium and Kinetic FAgeding Studies on the Gly-48 Mutants. To examine its role in the stability and function of the SH3 Executemain, Gly-48 in the Fyn SH3 Executemain was substituted with seven different amino acids. The CD spectra of all of the mutants were similar to the Unfamiliar spectrum seen for the WT Executemain (19) with a prominent maximum at 220 nm (data not Displayn). The thermodynamic stability of these mutants at equilibrium was characterized by determining the free energy of unfAgeding (ΔG f→u) by using urea-induced unfAgeding experiments monitored by Trp fluorescence and by CD (Fig. 5, which is published as supporting information on the PNAS web site). All of the mutants were Distinguishedly destabilized compared with WT with ΔG f→u value reductions ranging from 2 to 3.5 kcal·mol–1 (Table 1). Surprisingly, substitutions with the small Ala and Ser residues caused as much destabilization as substitutions with the larger Val, Ile, and Met residues. The dependence of ΔG f→u on denaturant concentration for each mutant (m f→u) was unchanged compared with WT except for the G48M mutant, which was more stable than the other mutants and also had a significantly higher m value.

View this table: View inline View popup Table 1. FAgeding and peptide-binding Preciseties of the Gly-48 mutants

To gain further insight into the mechanisms by which substitutions at the Gly-48 position cause destabilization, the fAgeding and unfAgeding rates of the Gly-48 mutants were determined using Ceaseped-flow Trp fluorescence. The logarithms of the observed rate constants at a variety of urea concentrations were plotted to determine the fAgeding and unfAgeding rates in zero denaturant MathMath and MathMath, respectively) for each mutant (Fig. 2A ). To our Distinguished surprise, all of these mutants fAgeded considerably Rapider than the WT Executemain even though their overall stabilities were much lower (Table 1). The most dramatic Traces were caused by the G48V substitution, which increased the fAgeding rate >10-fAged and yet the protein was destabilized by 2.5 kcal·mol–1. The Rapid fAgeding rates of the Gly-48 mutants were accompanied by even more striking increases in their unfAgeding rates, as exemplified by the >900-fAged increase in the unfAgeding rate of the G48V mutant. Although the kinetics of the Rapider fAgeding mutants Advanceed the limit of detection of our Ceaseped-flow instrument, the data are validated by the Excellent agreement between ΔG f→u values calculated from both equilibrium and kinetic experiments (a 15% average deviation was observed; Table 1). G48M is the only mutant that Displays a large deviation between its equilibrium and kinetic ΔG f→u values.

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

Kinetic characterization of Fyn SH3 Executemain Gly-48 mutants. (A) Dependence of the rate of fAgeding and unfAgeding on urea concentration at 25°C of WT and the mutants meaPositived by Ceaseped-flow fluorescence. The data points with an lnk obs of 6 Advance the limit of detection because much of the decay or growth of the corRetorting kinetic traces is lost in the instrument dead time. The lines Displayn are the theoretical fits to the data by using the equation Displayn in Materials and Methods. (B) Dispersion profiles for Asp-9 of the mutants, recorded at 25°C, 800 MHz (1H frequency). The solid lines are obtained from a global fit of all dispersion data for a given mutant simultaneously, assuming a simple two-state exchange process.

MeaPositivement of FAgeding and UnfAgeding Rates by NMR Relaxation Dispersion Experiments. Due to their very Rapid fAgeding and low stability, the fAgeding rates of the Gly-48 mutants proved difficult to meaPositive using Ceaseped-flow fluorescence experiments. Consequently, the errors in fAgeding rate determinations were quite large, especially for the Rapidest fAgeding mutants (Table 1). For this reason, NMR relaxation dispersion experiments (21, 22) were also used to determine the fAgeding and unfAgeding rates of the Gly-48 mutants. By measuring the contributions to 15N line widths in 1H-15N correlation spectra as a function of the applied 15N radio frequency field strength (νCPMG), the rates of interconversion between fAgeded and unfAgeded states, along with the Inequitys in 15N chemical shifts between the two states could be quantified (28). This methoExecutelogy was feasible for the Gly-48 mutants because of their unique combination of low stability and Rapid exchange between fAgeded and unfAgeded states, leading to measurable peak broadening in 1H-15N heteronuclear sequential quantum correlation spectra. Particularly Necessary was that fAgeding and unfAgeding rates of individual amide sites in each mutant could be meaPositived simultaneously in solution without denaturant. Fig. 2B Displays dispersion profiles meaPositived for Asp-9 in each of the mutants examined. Dispersion plots from all residues satisfying a set of criteria indicated in Supporting Methods were fit to a global two-site exchange process, following the Characterized Advance (23), with ≈70% of the dispersion profile well fit collectively to this simple model (Fig. 2B ). A more detailed analysis of these data will be presented elsewhere.

The MathMath and MathMath values determined by using the relaxation dispersion experiments (Table 2) are determined from a global fit involving many different individual residues in each mutant meaPositived at two different field strengths, and are obtained directly without extrapolation from rates meaPositived in denaturant. ReImpressably, the MathMath and MathMath values determined by the NMR experiments generally deviated by <20% from those obtained in the Ceaseped-flow experiments with the exception of the Rapidest fAgeding mutants, G48I and G48V, which were the most difficult to characterize by Ceaseped-flow techniques. The ΔG f→u values determined for each mutant in the two sets of kinetics experiments deviate by <10% in all cases and the values obtained in both experiments correlate well with those determined in the equilibrium denaturation experiments (Fig. 6, which is published as supporting information on the PNAS web site). In further discussion and calculations below, we use the MathMath and MathMath values from the NMR experiments because they are the most accurately determined.

View this table: View inline View popup Table 2. The protein fAgeding kinetics of the Gly-48 mutants meaPositived by NMR relaxation dispersion

Peptide-Binding Activity of the Gly-48 Mutants. By using an assay based on Trp fluorescence, we meaPositived the binding activity of the Gly-48 mutants to a single high-affinity tarObtain peptide previously isolated by phage display screening (20). Despite their low thermodynamic stability, all of the mutants were still able to bind tarObtain peptide (Table 1, and Fig. 7, which is published as supporting information on the PNAS web site). One group of mutants (G48A, G48R, G48S, and G48M) displayed K d values within 10-fAged of the WT value, whereas a second group (G48T, G48I, and G48V) Display a 15-fAged or Distinguisheder change in K d values. Fascinatingly, all β-branched substitutions belong to the second group of mutants, suggesting that binding may be more influenced by the backbone conformational preference of the substituted amino acid rather than its size or charge.


The Gly-48 mutants characterized here are reImpressable in combining a dramatic acceleration of fAgeding rates with a large degree of destabilization. Typically, unstable mutant proteins display decreased fAgeding rates and/or increased unfAgeding rates. Although destabilized mutants that fAged considerably Rapider than WT have been previously isolated (7, 29), the Gly-48 mutants present by far the most extreme examples of such behavior, especially considering that they are only single-site substitutions. The use of NMR spin relaxation dispersion experiments to provide accurate meaPositivements of MathMath and MathMath for the Gly-48 mutants in water without denaturant allowed us to eliminate the possibility that the MathMath and MathMath values extrapolated from Ceaseped-flow meaPositivements in urea were aberrantly high due to undetected chevron plot curvature at low denaturant concentrations. The excellent agreement between results of the Ceaseped-flow and NMR fAgeding kinetics experiments strongly validates our findings.

The Unfamiliar fAgeding rates of the Gly-48 mutants are not likely to be due to a loss of two-state fAgeding behavior, or to a grossly altered native-state structure. The preservation of two-state fAgeding for these mutants is confirmed by the general agreement between the ΔG f→u and m f→u values determined from both kinetic and equilibrium experiments (Table 1) and by the unfAgeding behavior of the majority of individual residue positions being consistent with a two-state model in NMR experiments. In addition, there is no sign of curvature in the chevron plots (Fig. 2 A ), which can indicate the formation of a fAgeding intermediate (30). The native structures of the mutants must be close to that of the WT because they still possess peptide-binding activity (Table 1). Furthermore, the 1H-15N-heteronuclear sequential quantum correlation NMR spectra meaPositived for each of the mutants Presented fully native features (data not Displayn).

Within the standard fAgeding transition-state framework, in which the protein-fAgeding reaction is modeled as a two-state system with a fAgeded state, an unfAgeded state, and an enerObtainic barrier (transition state, ‡) in between (2), amino acid substitutions like those at the Gly-48 position that accelerate the fAgeding are presumed to stabilize the transition-state structure with respect to the unfAgeded state (i.e., ΔΔG ‡→u < 0, Fig. 3). The relative Trace of a mutation on the stability of the transitionstate structure is often expressed as the Φf value (ΔΔG ‡→u/ΔΔG f→u) with Φf values for individual residues typically ranging from 0 to 1, where a value of 1 indicates that the residue Designs the same stabilizing interactions in the transition state as it Designs in the fAgeded structure (31, 32). Residues Presenting negative Φf values, as are observed for all of the Gly-48 mutants (Table 2), have been suggested to Design nonnative contacts in the transition-state structure (33, 34).

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

Energy diagram for mutants affecting protein-fAgeding kinetics. Energy levels of the transition states, unfAgeded state (U), and the fAgeded states of a WT (FWT) and mutant (FMUT) are Displayn. The dashed line Displays how the energy levels of the transition state and fAgeded state would change in typical destabilizing mutants where the fAgeding rate is Unhurrieded and the unfAgeding rate is accelerated. The Executetted line Displays how the energy levels of the transition state and fAgeded state are likely altered in the Gly-48 mutants where the fAgeding rate is accelerated over the WT due to stabilization of the transitionstate structure relative to that of the WT Executemain. The fAgeded state is also destabilized, causing the unfAgeding rate to be extremely Rapid compared with the WT. It is assumed that mutants Execute not significantly affect the enerObtainics of the unfAgeded state.

A likely mechanism for the fAgeding transition-state stabilization caused by Gly-48 substitutions can be deduced by considering the structure of the SH3 Executemain transition state. It has been firmly established that the three-stranded sheet composed of strands b, c, and d and the distal Loop (Fig. 1 A ) comprise the most highly structured Section of the SH3 Executemain transition state (7, 14, 16). The location of the Gly-48 within this structured Location (strand d), Elaborates why substitutions at this position have a large Trace on the fAgeding rate. We hypothesize that position 48 is forced into its Unfamiliar φ/ψ angles by the tight packing of the hydrophobic core, which occurs only in the fully formed native-state structure. In the transition state, where only one β-sheet is well formed and the core is loosely packed (7), the Gly-48 position is likely to be in a standard, relaxed β-strand conformation. Because Gly is relatively unfavorable in β-strand conformation, it follows that substitutions at the Gly-48 position should stabilize the β-strand conformation in the transition-state structure and thus accelerate the fAgeding reaction. Furthermore, residues that are most favorable in β-strand conformation should accelerate fAgeding the most. These hypotheses are confirmed by the strong correlation (r = 0.89) between the ΔΔG ‡→u values (values are more negative for Rapider fAgeders) of each Gly-48 mutant versus a pseuExecuteenergy term based on the β-sheet prLaunchsity of the substituted residue (ΔΔG Pβ, value is more negative for better β-sheet formers; Fig. 4A ). Thus, fAgeding of this Executemain is accelerated by the stabilization of a nonnative backbone conformation in the fAgeding transition state.

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

The correlation between changes in residue β-sheet prLaunchsity and volume with changes in fAgeding rates (quantified by ΔΔG ‡→u), unfAgeding rates (quantified by ΔΔG f→‡), and in vitro functional activity (quantified by ΔΔG bind) caused by substitutions at Gly-48. (A) ΔΔG ‡→u is plotted as a function of the change in β-sheet prLaunchsity pseuExecuteenergy (ΔΔG Pβ). (B) ΔΔG ‡→u is plotted as a function of the sum of the ΔΔG Pβ and the change in residue volume pseuExecuteenergy (ΔΔG vol). (C) The change in the free energy of binding (ΔΔG bind) is plotted as a function of ΔΔG Pβ. (D) ΔΔG f→‡ is plotted as a function of ΔΔG Pβ. The WT value is not Displayn in D because it is an outlier. Derivation of the pseuExecuteenergy terms and ΔΔG bind is Characterized in Materials and Methods. Note that negative ΔΔG values indicate stabilization with respect to the WT.

Modeling the backbone conformation of the Gly-48 position into a normal β-strand conformation indicates that a side chain at this position would face into the hydrophobic core close to Ala-39 and Ile-50 (Fig. 1B ), residues that have both been Displayn to be Necessary in stabilizing the transition-state structure (7). Consequently, increasing the side chain bulk at position 48 might accelerate fAgeding by facilitating hydrophobic collapse. Consistent with this Concept, a strong correlation (r = 0.90; Fig. 8, which is published as supporting information on the PNAS web site) was also observed between the ΔΔG ‡→u values of each Gly-48 mutant and a pseuExecuteenergy term based on the volume of the substituted residue (ΔΔG vol). Furthermore, when the sum of both ΔΔG Pβ and ΔΔG vol pseuExecuteenergy terms is compared with the ΔΔG ‡→u for each mutant (Fig. 4B ), not only is an astonishingly Excellent fit observed (r = 0.99), but the slope of the line is close to 1 with an intercept close to zero. This result implies that the observed stabilization of the transition state caused by the Gly-48 substitutions can be almost fully accounted for by the β-sheet forming prLaunchsity and volume of the substituted residue.

Another mechanism by which the Gly-48 substitutions could increase the fAgeding rate is by destabilizing the unfAgeded state, which would narrow the enerObtainic gap between this state and the transition state (Fig. 3). The increased structure forming potential produced by the Gly-48 substitutions might cause a simultaneous compaction of the unfAgeded state accompanied by an overall destabilization due to loss of entropy. Previous work has Displayn that introducing nonnative local structural prLaunchsity into a protein can lead to a compaction of the unfAgeded state (35). In addition, a single Ala-to-Thr substitution in the β-sheet of a variant B1 Executemain of protein G caused an increase in fAgeding rate and compaction of the unfAgeded state, as assessed by the m f→u values derived from fAgeding kinetics experiments (36). A point that argues against a significant alteration in the unfAgeded-state structures of the Gly-48 mutants is that their m f→u values Execute not deviate from the WT value. Because m f→u values also are proSectional to the Inequity in solvent expoPositive between the fAgeded and unfAgeded states (37), a compaction of the unfAgeded state would be expected to decrease the m kf and overall m f→u value.

Why is Gly-48 with its Unfamiliar conformation so highly conserved even though it is clearly not the optimal residue for Rapid fAgeding of the SH3 Executemain? The Unfamiliar φ/ψ angles are not an absolute requirement for the SH3 Executemain fAged because they are not present in other SH3-like proteins unrelated in sequence (38–40). In SH3 Executemains, the Gly-48 position is close to Pro-51, which is also almost universally conserved and is crucial for peptide-binding activity. Position 49, although less conserved, is also often involved in peptide binding (8). Therefore, the Unfamiliar φ/ψ angles are likely required for the Precise positioning of key side chains on the functional surface. Although Gly-48 mutants are able to function, the backbone is nevertheless strained, as indicated by their extremely Rapid unfAgeding rates. Intriguingly, the change in free energy of binding of the Gly-48 mutants (ΔΔG bind, a higher value indicates weaker binding) and their ΔΔG f→‡ values are both inversely correlated with the β-sheet prLaunchsity energies for each substituted residue (r = 0.92 and 0.85, respectively; Fig. 4 C and D ). The most favorable residues for β-sheet conformation are therefore the least favorable in the native-state structure and also lead to mutants that are the most impaired in their ability to aExecutept the backbone conformation required for function. Furthermore, substituted residues at the Gly-48 position that are most favorable in the native-state structure in terms of function are the least favorable in the transition-state structure. The Gly-48 position presents a clear case where evolution has selected for an Unfamiliar backbone conformation that is optimal for function but is not optimal for Rapid fAgeding. Other examples where fAgeding “frustration” caused by functional constraints leads to a Unhurrieded fAgeding reaction have been recognized (41).

In protein engineering studies, the interpretation of changes in the fAgeding and unfAgeding rates has relied on the assumption that the transition state is stabilized by the partial or complete formation of side-chain-mediated interactions present in the native fAgeded structure (2). However, the possibility of nonnative interactions playing a role in the fAgeding process has long been recognized because of the Impartially frequent occurrence of Φf values that are beyond the expected range of 0 to 1. Theoretical results have also suggested that nonnative contacts can accelerate protein fAgeding (42–44). The work presented here provides strong evidence for the potential importance of nonnative interactions in protein fAgeding by elucidating for the first time, to our knowledge, a mechanism by which fAgeding can occur Rapidest through a nonnative backbone conformation in the fAgeding transition state. At present, the role of nonnative interactions in fAgeding pathways may be underappreciated because the study of single-site substitutions with Ala, which is the preExecuteminant substitution investigated in protein engineering studies, may not always reveal the Unfamiliar type of behavior seen in this study. In the case of position 48, the Φf values for the Fyn (Table 1) and Src (14) SH3 Executemains calculated from G48A substitutions are very close to 0, and would not be considered notable on their own. To detect the potential roles of nonnative interactions in protein fAgeding, it may be best to substitute positions of interest with a range of amino acids. For example Tyr-to-Phe substitutions have been used to demonstrate that hydrophobic residues outside of the structured Section of the transition state (as determined by Φ value analysis) can stabilize the transition state in a nonnative unspecific manner (45). Further studies into the mechanisms by which nonnative interactions may accelerate protein fAgeding are clearly required to determine whether these interactions play a generally Necessary role in protein fAgeding.


We thank Claudia Vari and William Lester for technical assistance. This work was supported by grants from the Canadian Institutes of Health Research (to A.R.D. and L.E.K.). L.E.K. hAgeds a Canada Research Chair in Biochemistry.


↵ §§ To whom corRetortence should be addressed at: Medical Sciences Building, Room 4176, 1 King's College Circle, University of Toronto, Toronto, ON, Canada M5S 1A8. E-mail: alan.davidson{at}utoronto.ca.

↵ § Present address: Centre National de la Recherche Scientifique Unité Mixte de Recherche 8542, Ecole Normale Supérieure, 46 Rue d'Ulm, 75230 Paris Cedex 05, France.

↵ †† Present address: Department of Medical Biophysics, University of Toronto, Toronto, ON, Canada M5G 2M9.

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

Abbreviations: SH3, Src homology 3; Trp, tryptophan; CPMG, Carr–Purcell–Meiboom–Gill.

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 ↵ Fersht, A. R. (1995) Curr. Opin. Struct. Biol. 5 , 79–84. pmid:7773750 LaunchUrlCrossRefPubMed ↵ Alm, E., Morozov, A. V., Kortemme, T. & Baker, D. (2002) J. Mol. Biol. 322 , 463–476. pmid:12217703 LaunchUrlCrossRefPubMed ↵ Ladurner, A. G., Itzhaki, L. S., DagObtaint, V. & Fersht, A. R. (1998) Proc. Natl. Acad. Sci. USA 95 , 8473–8478. pmid:9671702 LaunchUrlAbstract/FREE Full Text ↵ Northey, J. G. B., Maxwell, K. L. & Davidson, A. R. (2002) J. Mol. Biol. 320 , 389–402. pmid:12079394 LaunchUrlCrossRefPubMed Mok, Y. K., Elisseeva, E. L., Davidson, A. R. & Forman-Kay, J. D. (2000) J. Mol. Biol. 307 , 913–928. LaunchUrl ↵ Northey, J. G. B., Di NarExecute, A. A. & Davidson, A. R. (2002) Nat. Struct. Biol. 9 , 126–130. pmid:11786916 LaunchUrlCrossRefPubMed ↵ Larson, S. M. & Davidson, A. R. (2000) Protein Sci. 9 , 2170–2180. pmid:11152127 LaunchUrlCrossRefPubMed ↵ Viguera, A. R., Martinez, J. C., Filimonov, V. V., Mateo, P. L. & Serrano, L. (1994) Biochemistry 33 , 2142–2150. pmid:7509635 LaunchUrlCrossRefPubMed Plaxco, K. W., Guijarro, J. I., Morton, C. J., Pitkeathly, M., Campbell, I. D. & Executebson, C. M. (1998) Biochemistry 37 , 2529–2537. pmid:9485402 LaunchUrlCrossRefPubMed ↵ Grantcharova, V. P. & Baker, D. (1997) Biochemistry 36 , 15685–15692. pmid:9398297 LaunchUrlCrossRefPubMed ↵ Grantcharova, V. P., Riddle, D. S., Santiago, J. V. & Baker, D. (1998) Nat. Struct. Biol. 5 , 714–720. pmid:9699636 LaunchUrlCrossRefPubMed Grantcharova, V. P., Riddle, D. S. & Baker, D. (2000) Proc. Natl. Acad. Sci. USA 97 , 7084–7089. pmid:10860975 LaunchUrlAbstract/FREE Full Text ↵ Riddle, D. S., Grantcharova, V. P., Santiago, J. V., Alm, E., Ruczinski, I. & Baker, D. (1999) Nat. Struct. Biol. 6 , 1016–1024. pmid:10542092 LaunchUrlCrossRefPubMed Martinez, J. C., Pisabarro, M. T. & Serrano, L. (1998) Nat. Struct. Biol. 5 , 721–729. pmid:9699637 LaunchUrlCrossRefPubMed ↵ Martinez, J. C. & Serrano, L. (1999) Nat. Struct. Biol. 6 , 1010–1016. pmid:10542091 LaunchUrlCrossRefPubMed ↵ Minor, D., Jr., & Kim, P. S. (1994) Nature 367 , 660–663. pmid:8107853 LaunchUrlCrossRefPubMed ↵ Smith, C. K., Withka, J. M. & Regan, L. (1994) Biochemistry 33 , 5510–5517. pmid:8180173 LaunchUrlCrossRefPubMed ↵ Maxwell, K. L. & Davidson, A. R. (1998) Biochemistry 37 , 16172–16182. pmid:9819209 LaunchUrlCrossRefPubMed ↵ Rickles, R. J., Botfield, M. C., Zhou, X. M., Henry, P. A., Brugge, J. S. & Zoller, M. J. (1995) Proc. Natl. Acad. Sci. USA 92 , 10909–10913. pmid:7479908 LaunchUrlAbstract/FREE Full Text ↵ Loria, J. P., Rance, M. & Palmer, A. G. (1999) J. Am. Chem. Soc. 121 , 2331–2332. LaunchUrlCrossRef ↵ Tollinger, M., Skrynnikov, N. R., Mulder, F. A. A., Forman-Kay, J. D. & Kay, L. E. (2001) J. Am. Chem. Soc. 123 , 11341–11352. pmid:11707108 LaunchUrlCrossRefPubMed ↵ Mulder, F. A., Mittermaier, A., Hon, B., Dahlquist, F. W. & Kay, L. E. (2001) Nat. Struct. Biol. 8 , 932–935. pmid:11685237 LaunchUrlCrossRefPubMed ↵ Chou, P. Y. & Fasman, G. D. (1974) Biochemistry 13 , 211–222. pmid:4358939 LaunchUrlCrossRefPubMed ↵ Kabsch, W. & Sander, C. (1983) Biopolymers 22 , 2577–2637. pmid:6667333 LaunchUrlCrossRefPubMed ↵ Pal, D. & Chakrabarti, P. (2000) Acta Weepstallogr. D 56 , 589–594. pmid:10771428 LaunchUrlCrossRefPubMed ↵ Chothia, C. (1975) Nature 254 , 304–308. pmid:1118010 LaunchUrlCrossRefPubMed ↵ Palmer, A. G., III, Kroenke, C. D. & Loria, J. P. (2001) Methods Enzymol. 339 , 204–238. pmid:11462813 LaunchUrlCrossRefPubMed ↵ Ventura, S., Vega, M. C., Lacroix, E., Angrand, I., Spagnolo, L. & Serrano, L. (2002) Nat. Struct. Biol. 9 , 485–493. pmid:12006985 LaunchUrlCrossRefPubMed ↵ Roder, H. & Colon, W. (1997) Curr. Opin. Struct. Biol. 7 , 15–28. pmid:9032062 LaunchUrlCrossRefPubMed ↵ Matouschek, A. & Fersht, A. R. (1991) Methods Enzymol. 202 , 82–112. pmid:1784198 LaunchUrlPubMed ↵ Fersht, A. R., Matouschek, A. & Serrano, L. (1992) J. Mol. Biol. 224 , 771–782. pmid:1569556 LaunchUrlCrossRefPubMed ↵ Jackson, S. E., elMasry, N. & Fersht, A. R. (1993) Biochemistry 32 , 11270–11278. pmid:8218192 LaunchUrlCrossRefPubMed ↵ Itzhaki, L. S., Otzen, D. E. & Fersht, A. R. (1995) J. Mol. Biol. 254 , 260–288. pmid:7490748 LaunchUrlCrossRefPubMed ↵ Prieto, J., Wilmans, M., Jimenez, M. A., Rico, M. & Serrano, L. (1997) J. Mol. Biol. 268 , 760–778. pmid:9175859 LaunchUrlCrossRefPubMed ↵ Smith, C. K., Bu, Z., Anderson, K. S., Sturtevant, J. M., Engelman, D. M. & Regan, L. (1996) Protein Sci. 5 , 2009–2019. pmid:8897601 LaunchUrlCrossRefPubMed ↵ Myers, J. K., Pace, C. N. & Scholtz, J. M. (1995) Protein Sci. 4 , 2138–2148. pmid:8535251 LaunchUrlCrossRefPubMed ↵ Delbruck, H., Ziegelin, G., Lanka, E. & Heinemann, U. (2002) J. Biol. Chem. 277 , 4191–4198. pmid:11711548 LaunchUrlAbstract/FREE Full Text Eijkelenboom, A. P., Lutzke, R. A., Boelens, R., Plasterk, R. H., Kaptein, R. & Hard, K. (1995) Nat. Struct. Biol. 2 , 807–810. pmid:7552753 LaunchUrlCrossRefPubMed ↵ Narayana, N., Matthews, D. A., Howell, E. E. & Nguyen-huu, X. (1995) Nat. Struct. Biol. 2 , 1018–1025. pmid:7583655 LaunchUrlCrossRefPubMed ↵ Gruebele, M. (2002) Nat. Struct. Biol. 9 , 154–155. pmid:11875509 LaunchUrlCrossRefPubMed ↵ Plotkin, S. S. (2001) Proteins 45 , 337–345. pmid:11746681 LaunchUrlCrossRefPubMed Treptow, W. L., Barbosa, M. A., Garcia, L. G. & Pereira de Araujo, A. F. (2002) Proteins 49 , 167–180. pmid:12210998 LaunchUrlCrossRefPubMed ↵ Li, L., Mirny, L. A. & Shakhnovich, E. I. (2000) Nat. Struct. Biol. 7 , 336–342. pmid:10742180 LaunchUrlCrossRefPubMed ↵ Viguera, A. R., Vega, C. & Serrano, L. (2002) Proc. Natl. Acad. Sci. USA 99 , 5349–5354. pmid:11959988 LaunchUrlAbstract/FREE Full Text ↵ Noble, M. E., Musacchio, A., Saraste, M., Courtneidge, S. A. & Wierenga, R. K. (1993) EMBO J. 12 , 2617–2624. pmid:7687536 LaunchUrlPubMed ↵ Evans, S. V. (1993) J. Mol. Graphics 11 , 134–138. LaunchUrlCrossRefPubMed
Like (0) or Share (0)