The Native State Conformational Ensemble of the SH3 Domain from R-Spectrin†

The folding/unfolding equilibrium of the R-spectrin SH3 domain has been measured by NMRdetected hydrogen/deuterium exchange and by differential scanning calorimetry. Protection factors against exchange have been obtained under native conditions for more than half of the residues in the domain. Most protected residues are located at the â-strands, the short 310 helix, and part of the long RT loop, whereas the loops connecting secondary structure elements show no measurable protection. Apparent stability constants per residue and their corresponding Gibbs energies have been calculated from the exchange experiments. The most stable region of the SH3 domain is defined by the central portions of the â-strands. The peptide binding region, on the other hand, is composed of a highly stable region (residues 53-57) and a highly unstable region, the loop between residues 34-41 (n-Src loop). All residues in the domain have apparent Gibbs energies lower than the global unfolding Gibbs energy measured by differential scanning calorimetry, indicating that under our experimental conditions the amide exchange of all residues in the SH3 domain occurs primarily via local unfolding reactions. A structure-based thermodynamic analysis has allowed us to predict correctly the thermodynamics of the global unfolding of the domain and to define the ensemble of conformational states that quantitatively accounts for the observed pattern of hydrogen exchange protection. These results demonstrate that under native conditions the SH3 domain needs to be considered as an ensemble of conformations and that the hydrogen exchange data obtained under those conditions cannot be interpreted by a two-state equilibrium. The observation that specific regions of a protein are able to undergo independent local folding/unfolding reactions indicates that under native conditions the scale of cooperative interactions is regional rather than global. For many years, our view of the folding equilibrium in proteins has been distorted by the so-called two-state model, in which proteins are assumed to be in equilibrium between two discrete states, the native and unfolded states. Today, we know that the two-state view is misleading and that the behavior observed under transition conditions cannot be extrapolated to native conditions. In particular, the heterogeneity in the pattern of hydrogen exchange protection measured by NMR for many proteins has demonstrated the existence of multiple conformations under equilibrium conditions and cannot be rationalized in terms of the two-state model (1-13). In hydrogen/deuterium exchange experiments performed under the so-called EX2 regime (see, for example, refs 8 and 14 and below), amide groups that are buried from the solvent in the native structure become able to undergo the exchange reaction as a result of some unfolding event or conformational change that render them exposed to the solvent. If only two states were accessible to a protein, all amide groups that are protected in the native state would exhibit the same protection factors since the same unfolding event will expose all of them to the solvent. Moreover, the Gibbs energy calculated from this event would be equal to the global Gibbs energy for global unfolding. This is not what is observed experimentally. For all proteins studied to date, the pattern of protection factors shows significant variations between residues, indicating that certain amide groups become exposed to the solvent as a result of local rather than global unfolding events (1-8, 12, 13, 15-18). These local unfolding events occur independently of each other and give rise to a large number of conformational states in which small regions of the protein fold and unfold in all possible combinations. Under equilibrium conditions the probabilities of those conformations are dictated by their Gibbs energies. In this paper we have studied the conformational equilibrium of the SH31 domain from R-spectrin. SH3 domains are small conserved modular domains (∼60 amino acids) that mediate protein-protein interactions in cellular signaling cascades and many other important biological processes (1923). The structure of several SH3 domains is known at high † Supported by grants from the DGICYT (PB96-1446) (Spain), the European Commission (PL95-0200 and PL96-2180), the National Science Foundation (MCB-9816661), and the National Institutes of Health (GM51362) (Baltimore). * Correspondence can be addressed to either author. 1 Abbreviations: SH3, Src homology region 3; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis; COSY, scalar coupling correlated spectroscopy; DSC, differential scanning calorimetry. 8899 Biochemistry 1999, 38, 8899-8906 10.1021/bi990413g CCC: $18.00 © 1999 American Chemical Society Published on Web 06/19/1999 resolution. In general, SH3 domains fold as compact â-barrels with five to six antiparallel â-strands. SH3 domains mediate protein-protein interactions and modulate enzyme activity by their ability to bind specific proline-rich peptides. Since SH3 domains receive and transmit information via these specific interactions, it is important to understand how the structural cooperativity is propagated through the native domain structure. The combined use of NMR-detected hydrogen-deuterium exchange and high-sensitivity differential scanning calorimetry provides a unique way of measuring residue-level and global stability parameters and, therefore, of exploring the nature of the native state conformational ensemble. MATERIALS AND METHODS Protein Expression and Purification. Wild-type SH3 was overexpressed from a pET3d plasmid in the Escherichia coli BL21 (DE3) strain. The plasmid was a kind gift from Dr. Luis Serrano (EMBL, Heidelberg). Collected cells were lysed in 5 mM sodium citrate, pH 3.5, in a French press, and the protein was recovered from the supernatant by precipitation in ammonium sulfate at 75%. Precipitated protein was solubilized in 50 mM sodium phosphate buffer and 100 mM NaCl, pH 7.0, containing 6 M urea, and dialyzed extensively against the same buffer. The protein was purified by a gel filtration step on a Hi-Load Superdex 75 column (Pharmacia), and purity was checked by SDS-PAGE electrophoresis. Hydrogen Exchange. Exchange samples were prepared from a stock solution of R-spectrin SH3 by extensively dialyzing it against deionized water and further lyophilization. H-D exchange was initiated by rapidly dissolving the lyophilized protein in deuterated buffer (20 mM d5-glycine for pH* 3.0 and pH* 2.5 or 20 mM d3-acetate for pH* 4.0). The sample was filtered to remove insoluble protein, and then immediately placed in a 5 mm NMR tube and into the magnet. The final protein concentration was approximately 5 mM when dissolved in 500 μL of deuterated buffer. The pH* of samples was checked after each experiment. Values of pH* reported in D2O solutions represent direct pH meter readings, without correction for isotope effects. The experiment dead times (i.e., time between dissolving the protein and the start of data acquisition) were between 15 and 20 min. The probe temperature was determined immediately before each experiment using a sample of 80% ethylene glycol and 20% d6-dimethyl sulfoxide. Magnet shims were preadjusted on a sample identical to the exchange one. All NMR experiments were performed at 24.7 °C on a Brüker AMX-500 spectrometer belonging to the instrumentation center of the University of Granada. A set of 20-30 twodimensional phase-sensitive COSY spectra were acquired during the exchange at progressively longer time intervals. Acquisition was performed using the time-proportional phase incrementation technique (TPPI) (24). Data sets comprised 256 × 2048 data points and the spectral width was 6024 Hz. A light presaturation of the residual water signal was done during the relaxation delay. The program package NMRpipe (25) was used to process the NMR data on a Silicon Graphics O2 workstation. Prior to Fourier transformation, the 2D data matrix was multiplied by a square-sine bell window function in both dimensions and then zero-filled to 4096 × 4096 words. The NMRview package (26) was used for computerized cross-peak identification and analysis. The assignment of the COSY NHCRH cross-peaks of our spectra was readapted to our experimental conditions, taking as a reference the published assignment at pH 3.5 and 35 °C (27). The intensity of each NH-CRH cross-peak was taken to be the average of the intensities (absolute values) of the four components of the COSY cross-peak and was normalized by comparison with the averaged intensity of the cross-peak CRΗ-CâΗ corresponding to the nonexchangeable protons of Ala55. For each residue, the normalized peak intensity was considered to be proportional to the proton occupancy of the corresponding amide group. The time assigned to each proton occupancy value was taken to be the middle of the acquisition time for each COSY spectrum, and “time zero” is the time when the protein was dissolved into deuterated buffer. The exchange kinetics of 33-39 amide protons could be measured, depending on the conditions. A single-exponential decay function, I ) I0 exp(-kext) + c, was fitted to the normalized cross-peak intensities in order to determine their exchange rate constants, kex. I is the normalized signal intensity at time t (in seconds), I0 is the amplitude of the exchange curve, kex is the observed hydrogen exchange rate constant, and the constant c is the peak intensity at the infinity time point. A home made program was used to calculate the sequence-specific intrinsic rate constants of exchange for each amide proton, kint, at the different pH* values using the exchange data of Bai et al. (28) for model peptides. Protection factors against exchange, PF, were determined as PF ) kint/kex. Under our experimental conditions, the intrinsic exchange rate constants for all amide protons are several orders of magnitude lower than the reported refolding rate constant of R-spectrin SH3 (20). Accordingly, an EX2 mechanism for the amide hydrogen exchange could be assumed (28). Under EX2 exchange, for a particular residue, the equilibrium between the “open” state, in which the amide proton is exchange competent, and the “closed” state, in which the amide proton is protected against exchange, is fully established and the opening equilibrium constant, Kop, can be obtained as follows: Kop ) kex/kint ) (PF)-1. For each residue, the free energy difference between the closed and the open states is given by Calorimetry. High-sensitivity differential scanning calorimetry (DSC) was performed in deuterium oxide (D2O) solutions using a VP-DSC microcalorimeter (Microcal) at a scan rate of 60 deg/min. Unfolding experiments were carried out in 20 mM glycine or 20 mM acetate buffers in D2O at different values of pH* between 1.0 and 4.5. The samples were prepared by dissolving the lyophilized protein into the D2O buffers and then filtering the solutions with a 0.2 μm Millex filter (Millipore). The sample concentration was measured by UV absorption at 280 nm using an extinction coefficient of 16 147 for the native protein (20). The concentration of protein in these experiments was about 1 mg/mL. The temperature dependence of the molar partial heat capacity (Cp) of SH3 was calculated from the DSC data and analyzed using Origin 5.0 (Microcal). Cp curves were fitted by a nonlinear least-squares method using a two-state ∆G ) -RT ln Kop ) -RT ln(1/PF) (1) 8900 Biochemistry, Vol. 38, No. 28, 1999 Sadqi et al.