Proc. Natd. Acad. Sci. USA Vol. 88, pp. 8116-8120, September 1991 Biochemistry pKa shifts accompanying the inactivating Asp121 Asn substitution in a semisynthetic bovine pancreatic ribonuclease (NMR/dectrtac efects/Poison-otzman cammlons) MARK T. CEDERHOLMt, JEANNE A. STUCKEYt, MARILYNN S. DOSCHERt, AND LANA LEEt§ tDepartment of Biochemistry, Wayne State University School of Medicine, Detroit, MI 48201; and tDepartment of Chemistry and Biochemistry, University of Windsor, Windsor, ON N9B 3P4, Canada Communicated by Frederic M. Richards, June 10, 1991 (receivedfor review March 13, 1991)

ABSTRACT A senisynthetic RNase, RNase-(1-118)-(111- combined with the corresponding in which no amino 124), consising of a noncovalent complex between residues acid changes have been introduced, the full enzymatic ac- 1-118 of RNase (obtained from the proteolytic digestion of tivity of RNase-(1-118)-(111-124) is generated (6). The over- RNase A), and a synthetic 14-residue peptide containing resi- lap between the peptide and RNase-(1-118), at residues dues 111-124 of RNase, exhibits 98% of the enzymatic activity 111-118, is required to achieve both good binding and precise of bovine pancreatic ribonuclease A (EC 3.1.27.5). The re- alignment ofthe two chains (7). A refined crystal structure at placement of -121 by in this semisyn- 1.8-A resolution of RNase-(1-118)-(111-124), the fully active thetic RNase to form the "D121N" analog reduces kd,/K. to parent complex, has been determined (8). 2.7% of the value for RNase A. In the present work, lH NMR The assignments of the C2 proton NMR resonances for spectroscopy has been used to probe the ionization sates of each ofthe four in bovine pancreatic RNase A and Pis12, His 9, and His"' in this catalytically defective semisyn- their pKa values have been made in several laboratories thetic RNase. A comparison of the observed resonances of (9-13). For a review, see ref. 14. The C2 proton NMR D121N with those previously determined by others for RNase spectrum of semisynthetic RNase-(1-118).(111-124) and its A enabled us to assign the C2 proton NMR resonances to pH dependence also have been obtained (15). The titration individual residues; the asignment of Hisll9 was confirmed by behavior of the four histidine residues in this semisynthetic titrating D121N with the fully deuterated peptide, [Asnl2l]- derivative was indistinguishable from that found by others for RNase-(111-124). The observed pKa values of His'2, HiS"M5, RNase A. We report here the pH titration behavior of the and His"' decrease 0.18, 0.16, and 0.02 pH unit, respectively, histidine residues in D121N. as a result of the D121N replacement. Values calculated by Using the solution to the Poisson-Boltzmann equation using a finite difference algorithm to solve the Poisson- provided in the electrostatics program DELPHI (16, 17) and Bodtzmann equation (the DELPH program, version 3.0) and a the coordinate sets for the crystal structures of both RNase- refined 2.0-A coordinate set for the crystal structure of D121N (1-118) (111-124) (8) and the asparagine analog (18), we have differ sWficantly for active site residues Hisl2 (ApK. = found substantial differences between our experimentally -0.58) and Hsl"' (ApK, = -0.55) but not for Hisl"' (ApKa = determined values for pKa of D121N minus pKa of RNase A -0.10). The elimination of bound water from the calulations (ApKa) and the ApKa values predicted for the D121N replace- reduced, but did not reconcile, these discrepancies (His2, ment. ApK. = -0.36; His"9, ApK. = -0.41). Several lines of evidence indicate that Asp"21, which is MATERIALS AND METHODS invariant throughout 40 species of mammalian pancreatic Materials. RNase A (RAF grade, salt-free, lot 54P6915) RNase (1), functions as part of the active site of bovine used in the NMR experiments was purchased from Cooper pancreatic RNase A (EC 3.1.27.5). Neutron diffraction anal- Biomedical. RNase A (type XII-A, lot 13F-8100) used in the ysis of single crystals of RNase A has revealed the existence preparation of RNase-(1-118) was purchased from Sigma, as of a between the carboxyl Q81 of Asp12' and were carboxypeptidase A (type I-DFP, lot 13F-8100) and ring N,2 of His119, a critical active site residue (2). The pepsin (P-6887, lot 57F-8105, 4000 units/mg). 2H20,2HCH, replacement of Asp'21 by asparagine in a semisynthetic NaO2H, and sodium 2,2-dimethyl-2-silapentane-5-sulfonate derivative of RNase reduces kcat for the small substrate were purchased from Merck Sharp & Dohme. cytidine 2',3'-(cyclic)phosphate at pH 6.0 to 12% ofthe value Preparation of RNase-(1-118). RNase-(1-118) was pre- for RNase A and increases the value ofKm 4-fold (refs. 3 and pared by the successive digestion ofRNase A with pepsin and 4; M. L. Ram and M.S.D., unpublished data). To delineate carboxypeptidase A (15), except that the gel-filtered prepa- further the role of Asp'21 in the function of RNase, we now rations were further purified by isocratic ion-exchange chro- have determined the apparent pKa values ofthree ofthe four matography at 50C on SP-Sephadex G-25 (40- to 120-,um histidine residues in the molecule by the measurement of the particles; Pharmacia) in 0.13 M sodium phosphate, pH 6.65. pH dependence of the C2 proton NMR resonances of the Synthesis of RNase-(111-124) and [Asp'21jRNase-(111-124). semisynthetic derivative containing the Asn12' replacement. RNase-(111-124) and [Asp121]RNase-(111-124) were prepared This derivative, "D121N," is prepared by combining RNase- (1-118), a totally inactive entity obtained by successively Abbreviations: RNase-(1-118), polypeptide consisting of residues digesting RNase A with pepsin and carboxypeptidase A (5), 1-118 of RNase A; RNase-(111-124), tetradecapeptide consisting of with a synthetic peptide composed of the 14 carboxyl- residues 111-124 of RNase A; [Asp"12]RNase-(111-124), RNase- terminal residues of RNase, except that Asp121 has been (111-124) in which Asp'2' has been replaced by asparagine; RNase- replaced by asparagine (3, 4). If, instead, RNase-(1-118) is (1-118)-(111-124), noncovalent complex of RNase-(1-118) and RNase-(111-124); D121N, noncovalent complex of RNase-(1-118) and RNase-(111-124)(D121N); C2, C2 atom of histidine (39); ApKa, The publication costs of this article were defrayed in part by page charge PKa of D121N minus pKa of RNase A, unless otherwise noted; pH*, payment. This article must therefore be hereby marked "advertisement" uncorrected pH of a 2H-containing solution. in accordance with 18 U.S.C. §1734 solely to indicate this fact. §To whom reprint requests should be addressed.

8116 Downloaded by guest on September 30, 2021 Biochemistry: Cederholm et al. Proc. NatL. Acad. Sci. USA 88 (1991) 8117 by solid-phase synthetic methods (19, 20) and purified by using a water probe radius of 1.8 A. One calculation using a methods previously described (15). probe radius of 1.4 A reduced the ApKa values further by NMR Experiments. NMR samples were prepared from 0.01-0.03 pH unit (see Table 2). stock solutions of known concentration as deter- In separate computations, the crystallographically bound mined by analysis. Lyophilized protein derived waters ofthe parent complex and the asparagine analog were from aliquot samples of these stock solutions was dissolved eliminated and their corresponding ApKa values were calcu- in 2H20, adjusted to pH* 3.0 with 1 M 2HC1, and then heated lated. to 60°C for 1 hr to exchange the backbone amide protons (10). The samples were then made 0.3 M in NaCl and 0.5 mM in sodium 2,2-dimethyl-2-silapentane-5-sulfonate and adjusted RESULTS to the desired pH*. All pH* measurements given are those Histidine 'H NMR Resonances of D121N. Spectrum A of observed directly and are not corrected for the deuterium Fig. 1 illustrates the 300-MHz proton NMR resonances ofthe isotope effect. pH* measurements were made at room tem- four histidines of native RNase A in 0.3 M NaCl, pH* 4.0, at perature with an Ingold 6030-02 microelectrode fitted to a 300C. This spectrum is in excellent agreement with previously Corning 240 pH meter. The proton NMR spectra were published data obtained under identical conditions (10). acquired on a Bruker AC-300 NMR spectrometer at 30°C These four resonances have previously been assigned to with a spectral width of 4500 Hz, 16,384 data points, and His'2, His"19, His'05, and His' in the order of decreasing quadrature phase detection. The 1H2HO resonance was re- chemical shift at pH* 4.0 (9-13). The analogous spectrum for duced by homonuclear decoupling. The chemical shifts are reported with respect to the principal resonance of sodium the parent semisynthetic complex, RNase-(1-118)-(111-124), 2,2-dimethyl-2-silapentane-5-sulfonate. spectrum B in Fig. 1, reveals a direct correspondence with Deuteration of [Asp'21JRNase_(111-124). The C2 proton of the four histidine resonances found in RNase A (15). How- His"9 in the synthetic peptide [Asp'21]RNase-(111-124) was ever, in this semisynthetic complex, there is a fifth resonance fully deuterated by incubating a 16.8 mM solution of the (stippled resonance in spectra B-D of Fig. 1) at 8.6 ppm, peptide at pH* 8.0 at 40°C in 2H20 for 8 days (10). which is the same chemical shift as seen in the tripeptide pH Titrations. The pKa values of the C2 protons of RNase Gly-His-Gly and in RNase-(111-124) (spectrum C) (15). This A and the semisynthetic RNase were calculated from a resonance has been attributed to "unstructured" histidine, four-parameter nonlinear least-squares curve-fitting program which is in slow exchange with those histidines in a native based upon the following function (21, 22): conformation. Spectrum E in Fig. 1 contains the NMR spectrum ofRNase-(1-118); the resonances at 8.88, 8.73, and 8obs = 8A + (SAH- 8A) ([Hr)/(Kn + [H]n), 8.34 ppm are due to His'2, His105, and His' by analogy with previous studies (15). The additional resonance observed at where 6A is the chemical shift of the unprotonated species, 8.6 ppm has again been ascribed to "unstructured" histidine; 8AH is the chemical shift of the protonated species, K is the apparent acid dissociation constant, and n is the Hill coeffi- the reason for the broadness of this resonance, with two cient. distinct chemical shifts evident, is not clear. Calculation of Electrostatic Potentials. The predicted ApKa The corresponding spectrum for the asparagine analog, values ofthe histidines in the semisynthetic RNase due to the D121N, in spectrum D in Fig. 1, contains two resonances D121N substitution were calculated by the application of a with chemical shifts previously attributed to His'2 and His48 finite difference solution to a combination of the linearized in both RNase A and RNase-(1-118)-(111-124); these reso- and nonlinearized Poisson-Boltzmann equations using the nances, therefore, have been tentatively so assigned in this program DELPHI, version 3.0, on a Silicon Graphics 4D/70GT analog as well. computer (23, 24). When the refined coordinates of D121N (2.0-A resolution, R = 0.187) (18) (Protein Data Bank, reference 2SRN) were used, the N82 of Asn'2' was replaced by an O', and a charge E of - /2 was introduced at both 0o1 and O' of this newly introduced aspartic acid; the electrostatic potentials of each of the histidines were then calculated based upon the loss of D these two - Y2 charges at this residue. The sulfate anion, which is in the active site in the crystal structures, was not included in the calculations, but all crystallographically bound water molecules were included. The following param- I .C eters were used: ionic strength 0.3 M; protein interior and bound waters, dielectric constant of 2; solvent, dielectric constant of 78.6; grid size, 60 x 60 x 60; focusing boundary conditions and rotational averaging. B When the refined coordinates of RNase-(1-118)-(111-124) (1.8-A resolution, R = 0.204) (8) (Protein Data Bank, refer- 19 105 48 ence 1SRN) were used, the effect of the loss of the two -1/2 A charges, assumed to be on O81 and O02 ofAsp'1 , on the pKa values for the histidines in the protein was calculated. The sulfate anion was again eliminated, and all crystallographi- 9.0 8.8 8.6 8.4 8.2 8.0 cally bound water molecules were included. The calculations 8, PPm included the following parameters: ionic strength 0.3 M; FIG. 1. The 300-MHz proton NMR spectra of 2.8 mM RNase A protein interior and bound waters, dielectric constant of 2; (A), 2.8 mM RNase-(1-118)(111-124) (B), 2.8 mM RNase-(111-124) solvent, dielectric constant of 80; grid size, 65 x 65 x 65; (C), 2.8 mM D121N (D), and 2.8 mM RNase-(1-118) (E). All samples rotational averaging and focusing boundary conditions. For were at pH* 4.0. Spectra B, C, and D are fully relaxed; spectra A and both coordinate sets, the protein solvent boundary was E are partially relaxed. Shadings are explained in the text. Numbers defined by measuring the solvent-accessible surface (25, 26), in spectrum A are histidine positions. Downloaded by guest on September 30, 2021 8118 Biochemistry: Cederholm et al. Proc. NaM Acad Sci. USA 88 (1991) The resonance due to His"9 was assigned by titrating reduced when asparagine replaces Asp'21; kinetic evidence D121N with a fully deuterated preparation of the tetrade- supports this postulate (see Discussion). These three reso- capeptide [Asp121]RNase-(111-124) at pH* 7.0. In a separate nances became progressively smaller and disappeared as the experiment (data not shown), the addition of 1.5 equivalents pH* was raised from 4.0 to 5.1, so their presence did not of the fully deuterated peptide to 2.8 mM D121N at pH* 7.08 interfere significantly with the tracing of the C2 resonances caused the resonance at 7.94 ppm, previously attributed to during the pH titrations (see below). Undeuterated His119, to decrease in intensity whereas those resonances [Asp'21]RNase-(111-124) (data not shown) exhibited the assigned to His12 (7.82 ppm) and His"05 (8.04 ppm) were same chemical shift as was seen in "unstructured" histidine unchanged, confirming that the resonance at 7.94 ppm is due at 8.6 ppm in RNase-(111-124) (spectrum C in Fig. 1). to His"9. Histdne Titration Curves. NMR spectra over the range of Three ancillary resonances (cross-hatched) appear in the 7.5-9.0 ppm at selected pH* values are shown for RNase A spectrum of D121N (spectrum D of Fig. 1) that are not and D121N in Fig. 2. The resonance of His"9 in the aspar- observed in the spectrum of RNase A (spectrum A) or in the agine analog has been assigned as discussed above, while parent complex (spectrum B), but which do appear in the His'2 and His'05 have been assigned by comparison with the spectrum of RNase-(1-118) alone (spectrum E). Their pres- previously published proton NMR spectra of RNase A. Fig. ence in the spectrum of D121N suggests that the strength of 3 A and B shows plots ofthe chemical shifts ofthe C2 protons of His12, His105, and His"19 of RNase A and of D121N, binding between RNase-(1-118) and RNase-(111-124) may be respectively, as a function ofpH. Analysis ofthe pH titration behavior of His' was not possible due to the broadening of this resonance over the pH range and under the conditions used in these experiments (27), a phenomenon that has been A observed previously for RNase A (27, 28) and for RNase- J (1-118).(111-124) as well (15). a The pKa values and Hill coefficients for the remaining three accessible histidines, calculated from the nonlinear least- squares analysis described in Materials and Methods, are - ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~Ipresented in Table 1. In comparison with the pKa values of RNase A, the pKa values of His12 and His 05 of D121N decrease 0.18 and 0.16 U H by pH unit, respectively, when 0 asparagine replaces aspartic acid at position 121. In contrast, the corresponding pKa value of His"9 is not altered by this G substitution. 0 Modeing the Electrostatic Effect of the ASp'21 -+ Asn U Substitution on the Histidine pK. Values. A refined coordinate set for D121N (18) was used in conjunction with a finite -%...F" 8:8 8'6 8.4 8:2 8.0 i.8 76 9.0 8, ppm 8.5 E ce E 8.0 A. A_ J D 7.5 3.5 4.5 5.5 6.5 7.5 8.5 pH C C 9.0

8.5 B aE A~~~~~~~ CO 8.0

8.8 8.6 84 8.2 8.0 7.8 7.6 7.5 . . 8, ppm 3.5 4.5 5.5 6.5 7.5 8.5 pH FIG. 2. Spectra A-E, 300-MHz proton NMR spectra of 2.8 mM RNase at selected pH* values: A, 4.00; B, 5.04; C, 6.00; D, 7.01; and FIG. 3. Chemical shifts of the C2 protons of His12 (-), His'05 (-), E, 7.98. Spectra F-J, 2.8 mM D121N at selected pH* values: F, 4.00; and His"9 (A) of RNase A (A) and D121N (B) as a function of pH. G, 5.09; H, 6.03; I, 7.07; and J, 7.99. The C2 protons of His'2 (U), Both samples were 0.3 M NaCl at 30°C in 2H20. The solid lines His'05 (e), and His"19 (A) have been labeled. Both samples were in 0.3 represent the calculated chemical shifts determined by the nonlinear M NaCl at 30°C in 2H20. least-squares analysis described in Materials and Methods. Downloaded by guest on September 30, 2021 Biochemistry: Cederholm et al. Proc. Natl. Acad. Sci. USA 88 (1991) 8119 Table 1. Least-squares analysis of histidine titration profiles Residue* System PA6 pKa n His'2 RNase-(1-118).(111-124)t 7.67 (0.01) 9.02 (0.01) 5.94 (0.02) 0.69 (0.02) RNase At 7.64 (0.03) 8.96 (0.03) 6.03 (0.02) 0.74 (0.01) D121Nt 7.70 (0.02) 8.97 (0.01) 5.85 (0.01) 0.79 (0.01) His'05 RNase-(1-118).(111-124)t 7.69 (0.02) 8.76 (0.01) 6.78 (0.02) 0.89 (0.03) RNase At 7.69 (0.01) 8.75 (0.02) 6.82 (0.01) 0.94 (0.01) D121Nt 7.70 (0.02) 8.78 (0.02) 6.66 (0.01) 0.88 (0.01) His"19 RNase-(1-118)-(111-124)t 7.76 (0.01) 8.83 (0.01) 6.26 (0.02) 0.77 (0.03) RNase At 7.76 (0.02) 8.80 (0.02) 6.33 (0.01) 0.86 (0.01) D121N§ 7.75 (0.02) 8.77 (0.01) 6.31 (0.01) 0.87 (0.01) Parameters are based on ref. 21. Numbers in parentheses represent standard deviations from the least-squares fits. *Assignments based on refs. 9-13. tValues from published measurements (15). tPutative assignment based on correlation of chemical shifts with RNase A; values from measurements of spectra shown in Fig. 2. §Assignment made by deuteration (see Results).

difference solution to the Poisson-Boltzmann equation (DEL- DISCUSSION PHI, version 3.0) (23, 24) to calculate the expected changes in the values ofthe three histidine residues as a result ofthe The similarity of the chemical shifts of fully protonated and pKa fully deprotonated His'2, His105, and His"19 in RNase A, asparagine substitution. The observed ApKa value of -0.16 RNase-(1-118).(111-124), and D121N listed in Table 1 sug- for His'05 is in good agreement with the predicted value of gests that the environments of these three histidine residues -0.10 for this residue (Table 2, rows 1 and 4). In contrast, the are similar in all three molecules at very low pH and at very predicted ApKa values for His'2 and His'19 of -0.58 and high pH. Even at pH* 4.0, the proton NMR spectrum of -0.55, respectively, are significantly greater than those of D121N contains resonances that correspond well with those -0.18 and -0.02 found experimentally (Table 2, rows 1 and ofHis'2 and His' in the parent complex and in RNase A (Fig. 4). 1). At this pH* value, however, the chemical shifts of His'05 If the numerous small structural changes with respect to and His"19 in the asparagine analog are significantly different. protein and crystallographically bound water that accompany The substitution of asparagine at position 121 has also the substitution of asparagine for Asp12' (18) are ignored by resulted in a decrease in the observed pKa values ofHis12 and calculating electrostatic potentials using the coordinate set His'05 of 0.18 and 0.16 pH unit, respectively (Table 1). Thus, for RNase-(1-118) (111-124) (8), the discrepancy between the the environments of these three histidine residues at inter- experimental and the predicted ApKa values for His12 and mediate pH values have evidently all been disturbed by this His"9 resulting from the loss of two -1/2 charges at o01 and mutation. O02 of Asp12' is still greater (Table 2, rows 1 and 3). Again, Some decrease in the pKa values of the histidine residues the agreement between the experimental (-0.16) and pre- was anticipated, as the replacement of aspartic acid by dicted (-0.15) pKa shift for His'05 is excellent. asparagine removes a negative charge from the molecule; this When crystallographically bound water molecules were change would be expected to destabilize the positively eliminated in the ApKa calculations, all of the values pre- charged protonated form of a histidine residue and concom- dicted using the RNase-(1-118)-(111-124) coordinate set (Ta- itantly decrease its pKa value. Such an electrostatic effect is ble 2, row 5) or the D121N coordinate set (Table 2, row 6) sharply dependent upon distance and ionic strength, but it decreased in magnitude. Significant discrepancies between has been shown experimentally to remain detectable at the experimental and theoretical values, nevertheless, re- considerable distances and substantial ionic strengths. In an main. extracellular subtilisin from Bacillus amyloliquefaciens, Rus- sell and coworkers (29, 30) have observed that the replace- ment of Asp" with reduced the pKa of the active site Table 2. Comparison of the observed and predicted histidine His' by 0.29 pH unit at an ionic strength of 0.1 M (ApKa = PKa changes in the semisynthetic RNases -0.29). These residues are separated by 12-13 A. At an ionic ApKa, pH units strength of 0.5 M, a corresponding decrease of 0.10 pH unit Row His'2 His'05 His"19 Comments (APKa = -0.10) could still be detected. The calculation ofthe electrostatic potentials in this subtilisin by the finite differ- 1* -0.18 -0.16 -0.02 ence Poisson-Boltzmann method (17) resulted in excellent 2t -0.09 -0.12 +0.05 agreement between the experimentally determined and pre- 3* -0.85 -0.15 -1.1 Asp'2'§, +H201 dicted ApKa values for His' (23). In a second example, a 4O -0.58 -0.10 -0.55 Asn'21i, +H20¶ dramatic decrease of 1.5 pH units occurs in the pKa of His57 51t -0.39 -0.09 -0.41 Asp'21§, -H201 in bovine pancreatic trypsin when Asp102, to which His57 is 6t -0.36 -0.07 -0.41 Asnl2l**, -H2O0** hydrogen bonded, is replaced by an asparagine (31). No *Experimentally determined pKa of D121N minus pKa of RNase A. major structural rearrangements result from this substitution tExperimentally determined pKa of D121N minus pKa of RNase- (32). (1-118)-(111-124) (15). In RNase A, His"9 is found in a conformation that brings fCalculated by computer simulation (DELPHI, version 3.0) as de- the side chains of and within scribed in Materials and Methods. His"9 (NW2) Asp'12 (Q81) §Based upon the coordinates of RNase-(1-118)-(111-124) (8). hydrogen bonding distance (2.74 A) (33-35), whereas in both fThe + and - signs indicate the presence and absence of crystallo- RNase-(1-118)-(111-124) and the asparagine analog, His"19 graphically bound water. occupies predominantly a second conformation that is I"Based upon the coordinates of D121N (18). achieved by rotation around the Ca-C0 bond (8, 36). In this **Use of a 1.4A water probe radius provided ApKa values of -0.35, conformation, the distance between these two residues is -0.06, and -0.38 for His'2, His'05, and His"19, respectively. considerably greater (9.9 and 8.8 A, respectively) (8, 18). Downloaded by guest on September 30, 2021 8120 Biochemistry: Cederholm et al. Proc. Natl. Acad. Sci. USA 88 (1991) Regardless ofits positioning, however, a sizeable decrease in data). The present study has eliminated models in which this the pK. value for His119 was expected, and, indeed, the inactivation is associated with a drastic decrease in the results from the application of the Poisson-Boltzmann equa- ground state pKa value ofactive site His12 or His119, a distinct tion confirmed this expectation. In addition, these calcula- possibility a priori. Further measurements in the presence of tions have revealed that the pK. shift for His12 is also active-site ligands may reveal differences in pKa values that substantially muted. would help to clarify the basis for the inactivation. The discrepancies between the observed and predicted pK5 values for His12 and His119 may be the result ofa number We thank Dr. Brian F. Pi Edwards, Dr. Philip D. Martin, and Dr. of factors. First, we have used the coordinates for crystal V. Srini J. deMel for providing the coordinates ofD121N. Amino acid structures in 3 M ammonium sulfate to model the titration analyses were performed by the Wayne State University Macromo- behavior of a protein in solution in 0.3 M NaCl. Second, with lecular Core Facility (supported in part by the Wayne State Univer- regard to His 19, the asparagine substitution has resulted in sity Center for Molecular Biology). This work was supported in part the imidazole ring ofthis residue undergoing a 1800 flip so that by the National Science and Engineering Research Council of the N81 of the ring now forms a strong hydrogen bond to a Canada, the J. P. Bickell Foundation, The University of Windsor water molecule (18). A third factor may be the effect of Research Board, and National Institutes of Health Grant GM 40630. changes in the arrangement of bound water molecules. A 1. Beintema, J. J., Schuller, C., hie, M. & Carsana, A. (1988) Prog. comparison of the structures of RNase-(1-118)*(111-124) and Biophys. Mol. Biol. 51, 165-192. D121N reveals numerous differences in the location and 2. Wlodawer, A. & Sjolin, L. (1981) Proc. Natl. Acad. Sci. USA 78, structure of crystallographically bound water networks (18). 2853-2855. 3. Merrifield, R. B. & Hodges, R. S. (1975) in Proceedings ofthe Interna- Such rearrangements may have resulted in significant tional Symposium on Macromolecules, ed. Mano, E. B. (Elsevier, changes in local dielectric constant. For RNase-(1-118).(111- Amsterdam), pp. 417-431. 124) and D121N, the initial ApKa calculations included crys- 4. Stern, M. S. & Doscher, M. S. (1984) FEBS Lett. 171, 253-256. tallographically bound water molecules, which were assigned 5. Lin, M. C. (1970) J. Biol. Chem. 245, 6726-6731. 6. Lin, M. C., Gutte, B., Moore, S. & Merrifield, R. B. (1970) J. Biol. a conventional dielectric constant of 2 (37). In both cases, Chem. 245, 5169-5170. large differences between the experimental (Table 2, rows 1 7. Gutte, B., Lin, M. C., Caldi, D. C. & Merrifield, R. B. (1972) J. Biol. and 2) and theoretical (Table 2, rows 3 and 4) ApKa values Chem. 247, 4763-4767. were observed. However, when the crystallographically 8. Martin, P. D., Doscher, M. S. & Edwards, B. F. P. (1987)J. Biol. Chem. 262, 15930-15938. bound water molecules of the parent complex and- of the 9. Meadows, D. H., Jardetzky, O., Epand, R. M., Ruterjans, H. H. & asparagine derivative were eliminated, there was a reduction Scheraga, H. A. (1968) Proc. Natl. Acad. Sci. USA 60, 766-772. in these discrepancies (Table 2, rows 5 and 6). This obser- 10. Markley, J. L. (1975) Biochemistry 14, 3546-3554. vation suggests that the dielectric constant of bound water 11. Patel, D. J., Canuel, L. L. & Bovey, F. A. (1975) Biopolymers 4, 987-997. molecules may be closer to that of bulk solvent. In the case 12. Shindo, H., Hayes, M. B. & Cohen, J. S. (1976) J. Biol. Chem. 251, of lysozyme, better results were also obtained after removal 2644-2647. ofbound water molecules from the crystallographic structure 13. Bradbury, J. H., Crompton, M. W. & Teh, J. S. (1977) Eur. J. Biochem. 81, 411-422. (38). 14. Eftink, M. R. & Biltonen, R. L. (1987) in Hydrolytic Enzymes, eds. When the coordinates for D121N are used in the presence Neuberger, A. & Brocklehurst, K. (Elsevier, Amsterdam), pp. 333-376. of crystallographically bound water, the discrepancy be- 15. Doscher, M. S., Martin, P. D. & Edwards, B. F. P. (1983) Biochemistry tween the observed and the predicted ApK. values (Table 2, 22, 4125-4131. rows 1 and 4) is moderated compared with the values 16. Gilson, M. K., Sharp, K. A. & Honig, B. H. (1987) J. Comp. Chem. 9, 327-335. obtained by using the coordinate set for the parent complex 17. Sharp, K. A. & Honig, B. (1990) Annu. Rev. Biophys. Biophys. Chem. (Table 2, rows 1 and 3). This moderation suggests that the 19, 301-332. structure of the protein as a whole is accommodating (or 18. deMel, V. S. J., Martin, P. D., Doscher, M. S. & Edwards, B. F. P. attempting to accommodate) to the change in charge distri- (1990) FASEB J. 4, A1760 (abstr.). 19. Merrifield, R. B. (1963) J. Am. Chem. Soc. 85, 2149-2154. bution resulting from the asparagine substitution. Such an 20. Merrifield, R. B. (1986) Science 232, 341-348. accommodation results, therefore, in a multitude of small, 21. Markley, J. L. (1973) Biochemistry 12, 2245-2250. but significant changes in structure throughout the molecule 22. Markley, J. L. (1975) Acc. Chem. Res. 8, 70-80. (18). 23. Klapper, I., Hagstrom, R., Fine, R., Sharp, K. & Honig, B. (1986) 1, 47-59. Three ancillary resonances are seen in the proton NMR 24. Gilson, M. K. & Honig, B. H. (1987) Nature (London) 330, 84-86. spectra of D121N over the pH* range of 4.0-5.1 that are not 25. Lee, B. & Richards, F. M. (1971) J. Mol. Biol. 55, 379-400. seen with RNase A or with the parent complex; they do 26. Richards, F. M. (1985) Methods Enzymol. 115, 440-464. appear in the spectrum of free RNase-(1-118X, however (Fig. 27. Roberts, G. C. K., Meadows, D. H. & Jardetzky,O. (1969) Biochemistry 8, 2053-2056. 1, spectrum E). The reduced binding energy between the 28. Markley, J. L. (1975) Biochemistry 14, 3554-3561. asparagine-containing peptide and RNase-(1-118) indicated 29. Russell, A. J., Thomas, P. G. & Fersht, A. R. (1987) J. Mol. Biol. 193, by this observation is supported by kinetic measurements at 803-813. pH 6.0: Kd = 33 (vs. 1 FxM for the parent complex) (M. L. 30. Sternberg, M., Hayes, F., Russell, A., Thomas, P. & Fersht, A. (1987) ,uM Nature (London) 330, 86-88. Ram and M.S.D., unpublished data). It is not likely that the 31. Craik, C. S., Roczniak, S., Largman, C. & Rutter, W. J. (1987) Science presence of significant amounts of free RNase-(1-118) and 237, 909-913. [Asp121]RNase-(111-124) in the pH range 4.0-5.1 has seri- 32. Sprang, S., Standing, T., Fletterick, R. J., Stroud, R. M., Finer-Moore, ously perturbed the titration curves of the histidine residues J., Xuong, N.-H., Hamlin, R., Rutter, W. J. & Craik, C. S. (1987) Science 237, 905-909. in the asparagine analog. If the pKa of the controlling group 33. Wlodawer, A., Bott, R. &Sjolin, L. (1982)J. Biol. Chem. 257,1325-1332. is 4.0 (as indicated by the presence of essentially equal 34. Wlodower, A. & Sjolin, L. (1983) Biochemistry 22, 2720-2728. concentrations of the complex and its two components at this 35. Bernstein, F. C., Koetzle, T. F., Williams, G. J. B., Meyer, E. F., pH value), only 7% of the chains remain dissociated at pH Brice, M. D., Rodgers, 1. R., Kennard, O., Shimanouchi, T. & Tasumi, M. (1977) J. Mol. Biol. 112, 535-542. 5.1. Moreover, at pH 5.1, only 6% of His119 (pKa = 6.3) and 36. Borkakoti, N., Moss, D. S. & Palmer, R. A. (1982) Acta Crystallogr. 17% of His12 (pKa = 5.8) will have been titrated. Sect. B Crystallogr. Cryst. Chem. 38, 2210-2217. The increase in Km and the reduction in kcat that accom- 37. Gilson, M. K. & Honig, B. H. (1988) Proteins Struct. Funct. Genet. 3, pany the replacement of Asp121 by asparagine in RNase result 32-52. 38. Dao-Pin, S., Liao, D.-I. & Remington, S. J. (1989) Proc. Natd. Acad. Sci. in an enzyme that exhibits 6% activity against cytidine USA 86, 5361-5365. 2',3'-(cyclic)phosphate at pH 6.0 under standard assay con- 39. IUPAC-IUB Joint Commission on Biochemical Nomenclature (1985) J. ditions (refs. 3 and 4; M. L. Ram and M.S.D., unpublished Biol. Chem. 26, 14-42. Downloaded by guest on September 30, 2021