Christopher M.Dobson and Giampietro Ramponi1. Oxford Centre for Molecular Sciences, New Chemistry Laboratory,. University of Oxford, South Parks Road, ...
Protein Engineering vol.11 no.7 pp.557–561, 1998
Studies on enzymatic activity and conformational stability of muscle acylphosphatase mutated at conserved lysine residues
Fabrizio Chiti, Francesca Magherini1, Niccolo` Taddei1, Claudia Ilardi1, Massimo Stefani1,2, Monica Bucciantini1, Christopher M.Dobson and Giampietro Ramponi1 Oxford Centre for Molecular Sciences, New Chemistry Laboratory, University of Oxford, South Parks Road, OX1 3QT Oxford, UK and 1Department of Biochemical Sciences, University of Florence, Viale Morgagni 50, 50134 Florence, Italy 2To
whom correspondence should be addressed
An oligonucleotide-directed mutagenesis study was carried out on the five acylphosphatase conserved lysine residues to assess their possible participation in enzyme active site formation and their contribution to the enzyme conformational stability. The study was designed to eliminate the ambiguity arising from the presence of a sulfate ion, an enzyme competitive inhibitor, bound to lysine 32 and 68 in the crystal structure of the erythrocyte isoenzyme. Furthermore, previous kinetic studies suggested the presence of residues with pKa J 7.9 and 11, tentatively identified as two lysines. The kinetic parameters for the mutants under investigation are not significantly different from those of the wild-type enzyme, demonstrating that none of the lysine residues are involved in catalysis or in substrate binding. In addition, thermal and urea denaturation experiments performed by circular dichroism indicate that the mutated lysine residues do not play a significant role in the enzyme structural stabilization, as the destabilizing energy averages 1.40 kJ/mol. Such results are in agreement with those obtained with other proteins indicating that lysine residues make little contribution to the stability of the native structure. Keywords: acylphosphatase/oligonucleotide-directed mutagenesis/catalytic mechanism/circular dichroism/conformational stability
Introduction Acylphosphatase (E.C. 3.6.1.7) is a small cytosolic protein (11 365 kDa, the horse muscle form) which catalyzes the hydrolysis of acylphosphates such as acetylphosphate, carbamoylphosphate, 1,3-bisphosphoglycerate, succinylphosphate, β-aspartylphosphate and benzoylphosphate (reviewed in Stefani et al., 1997). Recently, a significant nuclease activity on both DNA and RNA and the presence of the enzyme in cell nuclei during apoptosis have been observed (Chiarugi et al., 1996, 1997). Acylphosphatase is widely distributed in differing vertebrate organs and tissues, where two different isoenzymes named muscle type and erythrocyte (organ common) type sharing 55% sequence homology can be found (Stefani et al, 1997). In addition, a gene coding for an enzyme clearly homologous to both acylphosphatase isoenzymes has been isolated in Drosophila, and an openreading frame coding for an enzyme similar to acylphosphatase has been detected in the E.coli genome (unpublished data). © Oxford University Press
Much information is currently available about the structure as well as the catalytic residues and mechanism of both isoenzymes. Recently, an oligonucleotide-directed mutagenesis approach led to the identification of two catalytic residues in the muscle isoenzyme, Arg23, found to be the main substrate binding residue, and Asn41 which appears to be involved in binding and positioning of the catalytic water molecule (Taddei et al., 1994; 1996a). Furthermore, the importance of a number of residues for enzyme structural stabilization has also been reported (Taddei et al., 1996 a,b). The importance of such residues has been confirmed by the recent crystal structure of the erythrocyte isoenzyme at 1.8 Å resolution (Thunnissen et al., 1997) which is very similar to the previously solved 1H NMR solution structure of the muscle form (Pastore et al., 1992). Acylphosphatase is a very compact globular α/β protein composed of a five-stranded antiparallel twisted β-sheet facing two antiparallel α-helices. The structures of the two isoenzymes show only minor and local differences mainly involving the loop regions; such differences have been attributed to the presence of a sulfate and a chloride ion in the active site of the eryhrocyte isoenzyme crystal structure (Thunnissen et al., 1997). In this structure, the sulfate and chloride ions are present near the Arg23 and Asn41 side chains making extensive contacts to the last residues of the 15–21 loop, the most conserved part of all acylphosphatases. The detailed description of the environment surrounding the sulfate and chloride ions led to the identification of the enzyme active site and to a proposal for a substrate-assisted catalytic mechanism (Thunnissen et al., 1997). However, the X-ray structure of the erythrocyte isoenzyme appears somewhat distorted by the presence of the sulfate and chloride ions in the region corresponding to the active site; moreover, the presence of an additional sulfate ion stabilized by two lysine residues raised the question of a possible involvement of these residues in catalysis (Thunnissen et al., 1997). The first kinetic investigation concerned with the catalytic mechanism of acylphosphatase (Satchell et al., 1972) indicated the presence, in the active site, of a number of basic residues characterized by pKa values of 7.9, 11 and .12, tentatively identified with two lysines and an arginine. The possible involvement of lysine residues in the enzyme catalysis was further suggested by a pyrydoxal-59-phosphate (PLP) inactivation study performed by Ramponi et al. (1975). For these reasons, we prepared mutants of the muscular acylphosphatase in which the five conserved lysine residues were replaced by glutamine. The five mutants were investigated for catalytic efficiency and stability properties to assess whether any of the conserved lysines is specifically involved in the enzyme structure–function relationship. Materials and methods Materials Benzoylphosphate was synthesized and purified as described by Camici et al. (1976). Isopropyl thiogalactoside (IPTG) and 557
F.Chiti et al.
restriction enzymes were from Promega. Sequenase was from USB; glutathione, thrombin and glutathione affinity gels were from Sigma. The pGEX-2T plasmid was from Pharmacia. The pGEX-AP plasmid, which is derived from the pGEX-2T vector by the insertion of a chemically synthesized gene coding for the human muscle acylphosphatase, was prepared as previously described (Modesti et al., 1993). Plasmid propagation and protein expression were achieved in Escherichia coli strain DH5α. Oligonucleotides were from Pharmacia Biotech. All other reagents were analytical grade or the best commercially available. Site-directed mutagenesis The acylphosphatase mutants (K32Q, K57Q, K67Q, K84Q and K88Q) were obtained by oligonucleotide-directed mutagenesis using a mutagenesis kit based on the unique site elimination technique (USE) devised by Deng and Nickoloff (1992). Five target primers harbouring the desired mutated codons were synthesized by the cyanoethyldeoxy (CED)-phosphoramidite method. The target mutagenic primers were complementary to the gene coding for the muscle acylphosphatase and carried the desired mutations: AAG or AAA to CAG or CAA. The target mutagenic primer and a selection primer were used to introduce the site-specific mutation into the plasmid. Both mutations have been incorporated into the same strand by in vitro DNA synthesis. The selection primer eliminates the unique recognition sequence for ApaI in a nonessential region of the pGEX-2T plasmid. In our case, the mutated doublestranded DNAs were introduced by transformation of E.coli DH5α strain. Mutations were confirmed by DNA sequencing according to Sanger et al. (1977). Protein expression and purification The expression of the recombinant mutated and wild-type acylphosphatase genes was achieved after inducing the cell cultures with 0.2 mM IPTG under the conditions previously described (Taddei et al., 1996a). The presence of the mutated proteins in cell lysates was detected by SDS–PAGE analysis according to Laemmli (1970). The fusion proteins were purified and subsequently cleaved by thrombin, and the pure mutated and wild-type recombinant acylphosphatases were isolated as previously described (Modesti et al., 1995; Taddei et al., 1996a). The purity of the protein samples was checked by SDS–PAGE analysis. Protein concentration determination and acylphosphatase activity measurements Protein concentration was measured by UV absorption using an A1% 1cm,280 5 14.2 for all acylphosphatases. Acylphosphatase activity was measured by a continuous optical test at 283 nm and 25°C, using benzoylphosphate as a substrate freshly dissolved in 0.1 M acetate buffer, pH 5.3, as previously reported (Ramponi et al., 1966), unless otherwise indicated. Pyrydoxal-59-phosphate inactivation experiments These experiments were performed at a protein concentration of 0.02 mg/ml for each mutant and for the wild-type enzyme, in 50 mM 3,3-dimethylglutarate buffer, pH 7.5. Each sample was preincubated for 15 min in the presence of 1 mM PLP. The residual enzymatic activity was measured in the above preincubation mixture in the presence of 5 mM benzoylphosphate as a substrate, and expressed as a percentage of the activity in the absence of PLP. The enzymatic activity was also measured in a control experiment performed in the absence of PLP. 558
Circular dichroism experiments Circular dichroism (CD) experiments were performed on a Jasco Model J720 spectropolarimeter equipped with a thermostatically-controlled cell holder and a NesLab model RTE-111 water circulating bath. Far UV spectra of the wild-type and mutated recombinant acylphosphatases were recorded at 25°C in 50 mM acetate buffer, pH 5.5, at a protein concentration of 0.4 mg/ml using a 1 mm path-length quartz cell. Thermal-induced denaturation The far UV signal at 222 nm was used to monitor thermal unfolding of each mutant. The thermal unfolding curves were recorded from 20 to 80°C using a heating rate of 0.5°C/min, while monitoring the temperature course by positioning a thermocouple inside the sample. Protein concentration for all mutants was 0.4 mg/ml in 50 mM acetate buffer, pH 5.5. The reversibility of the unfolding transitions was checked by measuring the CD signal at room temperature upon cooling down immediately after the end of the transition. The apparent unfolding equilibrium constant (Keq) was calculated in the range of temperature where the transition occurred, using Keq 5 (yn – y)/(y – yd)
(1)
where y is the ellipticity observed at the relevant temperature, yn and yd are the ellipticities characteristic of the folded and unfolded protein, respectively, extrapolated from the pre- and post-transition baselines at the temperature under consideration. Plots of ln(Keq) versus 1/T relative to the transition region were linearly fitted by using the van’t Hoff analysis: ln(Keq) 5 ∆Sm/R – ∆Hm/R(1/T)
(2)
∆Hm and ∆Sm being the enthalpy and entropy change upon unfolding at the melting temperature Tm, respectively. Urea induced denaturation Far UV CD signal at 222 nm was also used to monitor the urea-induced unfolding of the wild type and mutated acylphosphatases. Urea denaturation curves were obtained at 25°C by recording the ellipticity of 30 preincubated samples containing 0.4 mg/ml protein in 50 mM acetate buffer, pH 5.5 and urea concentrations ranging from 0 to 9 M with 0.3 M increments, using a 1 mm path-length quartz cuvette. Plots of ellipticity as a function of urea concentration were analysed (Santoro and Bolen, 1988) by y 5 {(yf 1 mf[urea]) 1 (yu 1 mu[urea]) exp[–∆G(H2O)/RT 1 m[urea]/RT]}/ {1 1 exp[–∆G(H2O)/RT 1 m[urea]/RT]}
(3)
where yf, mf, yu and mu are the slopes and intercepts of the pre- and post-transition regions, respectively, ∆G(H2O) is the free energy change between the unfolded and folded states in the absence of denaturant and m is the free energy dependence on urea concentration. The determined m value was basically similar for all mutants and an averaged value was calculated in order to prevent the experimental error affecting the m value from being reflected on ∆G(H2O), as suggested by Matouschek and Fersht (1991). Thus, ∆G(H2O) was re-calculated for each mutant using ∆G(H2O) 5 Cm3,m.
(4)
where ,m. is the averaged m value and Cm the denaturation midpoint of the mutant under consideration.
The role of lysine residues in acylphosphatase
Results and discussion A mutagenesis study has been performed on the five lysines conserved in all acylphosphatases isolated so far. Five oligonucleotides carrying the mutated codon for lysine were synthesized and used for the production of five lysine to glutamine mutants of muscle acylphosphatase under the conditions described in Materials and methods (Table I). The lysine to glutamine substitution was chosen in order to remove the positive charge without a notable effect on the steric hindrance and hydrophobicity of the side chain. Each mutation was confirmed by DNA sequencing according to Sanger et al. (1977). Wild-type and mutated recombinant enzymes were purified from cell lysates as fusion proteins and subsequently cleaved by thrombin and further purified with a final yield of 2–3 mg of protein/l of culture. A previous kinetic analysis carried out on muscle acylphosphatase suggested the possible presence of two lysine residues at the active site (Satchell et al., 1972). Therefore, we have performed a study on the catalytic properties of the mutants carrying substitutions at the site of conserved lysines. Table II reports the main kinetic parameters of the five lysine mutants as compared with those of the wild-type enzyme. None of these parameters appears significantly different from those displayed by the wild-type enzyme. In particular, both the wild-type and the mutated enzymes display very similar specific activity values and pH optima, indicating that none of the lysine side-chains replaced in this mutagenesis study is essential for catalysis. Similarly, the Km and Ki values for benzoylphosphate and inorganic phosphate, an acylphosphatase substrate and a competitive inhibitor, respectively, are not substantially altered when compared with those determined for the wild-type enzyme, suggesting that each mutant is still able to efficiently bind the substrate. Another kinetic study showed that pyrydoxal-59-phosphate, whose effect on enzymes proceeds through the reversible formation of a Schiff-base with lysine side-chains, is an acylphosphatase inactivator (Ramponi et al., 1975). This evid-
ence led the authors to suggest the possible binding of PLP to a lysine residue in the acylphosphatase active site or in its proximity. We therefore performed PLP inactivation experiments on the lysine mutants. Table II reports the percentage enzyme inactivation after 15 min preincubation of each sample in the presence of 1 mM PLP, at pH 7.5; under these experimental conditions, 15 min is the time required for the maximal inactivation to be established. In the case of PLP binding to a specific lysine residue, we would have expected a significant protection against the inactivation in the mutant carrying the target-lysine substitution. However, we have not observed a significant effect of this type for any of the mutants
Table I. Sequences of the oligonucleotides used for the production of the mutated enzymesa K32Q K57Q K67Q K84Q K88Q aThe
59-GAAGCCCGTCAGATCGGCGTG-39 59-CCCGAAGATCAGGTCAATAGT-39 59-TGGCTCTCTCAAGTGGGCAGC-39 59-TAGCAATGAGCAAACCATCAGTA-39 59-GACCATCAGTCAGCTTGAATAC-39
Fig. 1. Schematic representation of erythrocyte acylphosphatase structure drawn by WebLab Viewer, version 1.1. The lysine residues, the catalytic residues Arg23 and Asn41 and the two sulfate ions are reported (Thunnissen et al., 1997). α-Helices and β-strands are represented as cylinders and arrows, respectively.
mutated codons are in boldface.
Table II. Main kinetic properties of lysine mutantsa
WT K32Q K57Q K67Q K84Q K88Q
Specific activity (IU/mg protein)
Km (mM)
6500 5200 6000 5500 5700 6000
0.36 0.35 0.29 0.35 0.32 0.29
6 6 6 6 6 6
Ki (mM) 0.02 0.03 0.04 0.03 0.02 0.02
0.75 0.73 0.64 0.78 0.65 0.69
6 6 6 6 6 6
0.13 0.10 0.08 0.10 0.05 0.08
pH Optimum
% PLP inactivation
4.8–5.8 4.9–5.9 4.5–5.5 4.9–5.9 4.8–5.8 4.7–5.7
62 41 57 43 56 47
is defined as the enzymatic activity which hydrolyzes 1 µmol benzoylphosphate in 1 min, at 25°C and pH 5.3. The pH optimum was calculated at 25°C, in 0.1 M acetate buffer, pH 3.7–6.0, in 50 mM 3,3-dimethylglutarate buffer, pH 6.0–7.5 and in 0.1 M Tris–acetate buffer, pH 7.5–9.0. Km and Ki for benzoylphosphate as substrate and inorganic phosphate as competitive inhibitor, respectively, were calculated at 25°C in 0.1 M acetate buffer, pH 5.3. Pyrydoxal-59-phosphate inactivation is expressed as percentage inactivation with respect to the activity of the enzyme measured in the absence of PLP.
aIU
559
F.Chiti et al.
studied. Mutations at positions 32, 67 and possibly 88 appear to generate a slight protection of the enzyme against PLP inactivation, suggesting that these lysines could be partially responsible for transient PLP binding. The kinetic data reported here demonstrate that none of the lysines present in acylphosphatase are directly involved in the catalytic mechanism nor are they spatially close to the active site residues, contrary to previous suggestions (Satchell et al., 1972; Ramponi et al., 1975). This finding also eliminates the ambiguity arising from the presence of two sulfate ions in the crystal structure of the erythrocyte acylphosphatase (Thunnissen et al., 1997). Such a structure (Figure 1) shows that the first sulfate ion is bound to the Arg23 side-chain and to several backbone amides of the 15–21 loop residues forming the substrate binding pocket. The presence of a neighbouring chloride ion seems to push the sulfate ion somewhat out of the active pocket (Thunnissen et al., 1997). The new position of the sulfate ion is stabilized by an additional interaction with the side-chain of Lys68 (corresponding to Lys67 in the muscle isoenzyme) belonging to an adjacent (crystallographically related) acylphosphatase molecule. Both the presence of a second sulfate ion bound to Lys32 and Lys68, and the binding of Lys68 to the sulfate ion at the active site of a neighbouring acylphosphatase molecule can now be definitely ascribed to crystal packing and deprived of any possible catalytic relevance. We have also investigated the structural role and the contribu-
Fig. 2. Far-UV CD spectra of acylphosphatase recorded in 50 mM acetate buffer, pH 5.5 at 25°C: wild-type (continuous line); K57Q mutant (dotted line); K88Q mutant (dashed line). Urea denatured wild-type enzyme is represented as (dotted-dashed line).
tion to the enzyme conformational stability of each mutated lysine residue. Figure 2 shows the far UV circular dichroism spectra of K57Q and K88Q mutants as compared with the spectra of the wild-type enzyme in the native and urea denatured states. All the mutants gave rise to closely similar
Fig. 3. (A) Urea denaturation transition monitored by CD signal at 222 nm in 50 mM acetate buffer, pH 5.5, 25°C. The data are reported as fraction unfolded. (B) Natural logarithms of equilibrium denaturation constants as a function of the reciprocal of temperature. (d), wild-type acylphosphatase; (s), K32Q mutant; (j), K57Q mutant.
Table III. Thermodynamic parameters for the thermal and urea denaturation transition as calculated from CD dataa
WT K32Q K57Q K67Q K84Q K88Q aT m
Tm (°C)
Cm (M)
∆Hm (kJ·mol–1)
∆Sm (kJ·mol–1·K–1)
∆G(H2O) (kJ·mol–1)
56.5 52.2 50.1 56.4 50.6 55.1
4.61 4.22 4.00 4.68 3.98 4.50
391 392 370 358 324 370
1.18 1.20 1.44 1.11 1.00 1.13
18.5 16.9 16.0 18.7 15.9 18.0
and Cm are the temperature and the urea concentration, respectively, required to unfold 50% of the enzyme molecules and the other thermodynamic parameters are described under Materials and methods. An averaged value of 4.0 6 0.3 was calculated for the parameter m defined in Equation 3 and this was used to calculate the ∆G(H2O) values reported in the table (see Materials and methods). The mean errors on Tm and Cm values are 61°C and 60.15 M, respectively. The mean errors of the other parameters are within 10%.
560
The role of lysine residues in acylphosphatase
spectra, indicating that any lysine replacement does not induce substantial changes in the enzyme structure. The contribution of each lysine to the conformational stability has been investigated by urea and thermal denaturation experiments followed by CD at 222 nm. Figure 3 reports the urea-induced denaturation curves and the van’t Hoff plots relative to the K32Q and K57Q mutants as compared with the wild-type enzyme. The main thermodynamic parameters are summarized in Table III. Neither thermal nor chemical denaturation show large destabilizing energies upon substitution of any lysine residue by glutamine. Replacement of lysine residues at positions 67 and 88 does not cause any significant destabilization as shown by both thermal and chemical denaturation, and the K32Q mutant is only slightly destabilized. The largest destabilizations occur when the lysines at positions 57 and 84 are replaced, but yet they are not as remarkable as compared with the large destabilizations found for residues critically contributing to the conformational stability of proteins (Meeker et al., 1996). A very detailed study on the effect of charged residue replacements on the stability of a protein has recently been reported by Meeker et al. (1996). In this study, all the 23 lysines of staphylococcal nuclease were mutated to both alanine and glycine and the effect on the conformational stability was analysed in terms of the location of the residues in the three-dimensional structure of the protein with particular regard to the residue environment. The authors concluded that neither the Coulombic effects due to the charges nor the solvation differences occurring upon denaturation represent a significant fraction of the contribution of a charged residue to conformational stability. Indeed, the loss of packing interactions involving the proximal end of lysine side chains seems to be a major determinant of the observed destabilizing energies. The very low contribution of salt bridges to the stability of protein native structure also emerges from a large number of other studies (Akke and Forsen, 1990; Horovitz et al., 1990; Serrano et al., 1990; Pace et al., 1990, 1992; Dao Pin et al., 1991a,b). Our data are consistent with such previous reports. The residue environment of the five lysine residues under study can be deduced from the high resolution three-dimensional structure of the erythrocyte acylphosphatase (Figure 1). The lysine at positions 88 is fully exposed to the solvent and apparently does not interact with other neighbouring residues. Lys68 (corresponding to Lys67 in the muscle isoenzyme) is also fully exposed to the solvent; however, this position can be somewhat perturbed by the presence of the second sulfate ion binding residues 68 and 32. Lys32 also forms a salt bridge with Glu29; however, in the K32Q mutant no significant destabilization is observed. By contrast, the amino group in the Lys57 side-chain seems exposed to the solvent whereas the CH2 groups of the proximal end are packed together with the hydrophobic Pro54 and Leu35 residues. A larger destabilization free energy is reported for this mutant with both thermal and chemical denaturation. The destabilization observed for the K84Q mutant could be due to the breakage of the hydrogen bond formed between the Lys84 amino group and the backbone oxygen of Glu83. An average destabilization energy ∆∆Gaverage can be calculated from the data reported in Table III by using: ∆∆Gaverage 5 {Σ [∆G(H2O)wild type – ∆G(H2O)mutant]}/n
staphylococcal nuclease Lys to Ala mutants provides a similar averaged destabilization energy (1.25 kJ/mol), further suggesting that lysine residues do not significantly contribute to the overall stability of the native conformation of proteins. Acknowledgements This study has been supported by grants from EC (contract ERB BIO4-CT960517) and from Italian CNR (target project Structural Biology). F.C. is the recipient of an EC bursary (contract no. BIO4-CT96-5113) and is on leave from the Department of Biochemical Sciences of the University of Florence. C.M.D. research is partly supported by an International Research Scholars award from the Howard Hughes Medical Institute.
References Akke,M. and Forsen,S. (1990) Proteins, 8, 23–29. Camici,G., Manao,G., Cappugi,G. and Ramponi,G. (1976) Experientia, 32, 535. Chiarugi,P., Raugei,G., Fiaschi,T., Taddei,L., Camici,G. and Ramponi,G. (1996) Biochem. Mol. Biol. Int., 40, 73–81. Chiarugi,P., Degl’Innocenti,D., Raugei,G., Fiaschi,T. and Ramponi,G. (1997) Biochem. Biophys. Res. Commun., 231, 717–721. Dao Pim,S., Soderlind,E., Baase,W.A., Wozniak,J.A., Sauer,U. and Matthews, B.W. (1991a) J. Mol. Biol., 221, 873–887. Dao Pim,S., Sauer,U., Nicholson,H. and Matthews,B.W. (1991b) Biochemistry, 30, 7142–7153. Deng,W.P. and Nickoloff,J.A. (1992) Anal. Biochem., 200, 81–88. Horovitz,A., Serrano,L., Avron,B., Bycroft,M. and Fersht,A.R. (1990) J. Mol. Biol., 216, 1031–1044. Laemmli,U.K. (1970) Nature, 227, 680–685. Matouschek,A. and Fersht,A.R. (1991) Methods Enzymol., 202, 82–112. Meeker,A.K., Garcia-Moreno,E.B. and Shortle,D. (1996) Biochemistry, 35, 6443–6449. Modesti,A., Raugei,G., Taddei,N., Marzocchini,R., Camici,G., Manao,G. and Ramponi,G. (1993) Biochim. Biophys. Acta, 1216, 369–374. Modesti,A., Taddei,N., Bucciantini,M., Stefani,M., Colombini,B., Raugei,G. and Ramponi,G. (1995) Protein Exp. Purif., 6, 799–805. Pace,C.N., Laurents,D.V. and Thomson,J.A. (1990) Biochemistry, 29, 2564– 2572. Pace,C.N., Laurents,D.V. and Erickson,R.E. (1992) Biochemistry, 31, 2728– 2734. Pastore,A., Saudek,V., Ramponi,G. and Williams,R.J.P. (1992) J. Mol. Biol., 224, 427–440. Ramponi,G., Treves,C. and Guerritore,A. (1966) Experientia, 22, 1019–1023. Ramponi,G., Manao,G., Camici,G. and White,G.F. (1975) Biochim. Biophys. Acta, 391, 486–493. Sanger,F., Nicklen,S. and Coulson,A.R. (1977) Proc. Natl Acad. Sci. USA, 74, 5463–5467. Santoro,M.M. and Bolen,D.W. (1988) Biochemistry, 27, 8063–8068. Satchell,D.P.N., Spencer,N. and White,G.F. (1972) Biochim. Biophys. Acta, 268, 233–248. Serrano,L., Horovitz,A., Avron,B., Bycroft,M. and Fersht,A.R. (1990) Biochemistry, 29, 9343–9352. Stefani,M., Taddei,N. and Ramponi,G. (1997) Cell. Mol. Life Sci., 53, 141–151. Taddei,N., Stefani,M., Vecchi,M., Modesti,A., Raugei,G., Bucciantini,M., Magherini,F. and Ramponi,G. (1994) Biochim. Biophys. Acta, 1208, 75–80. Taddei,N., Stefani,M., Magherini,F., Chiti,F., Modesti,A., Raugei,G. and Ramponi,G. (1996a). Biochemistry, 35, 7077–7083. Taddei,N., Magherini,F., Chiti,F., Bucciantini,M., Raugei,G., Stefani,M. and Ramponi,G. (1996b) FEBS Lett., 384, 172–176. Thunnissen,M.M.G.M., Taddei,N., Liguri,G., Ramponi,G. and Nordlund,P. (1997) Structure, 5, 69–79. Received October 22, 1997; revised February 23, 1998; accepted March 2, 1998
(5)
where n is the number of mutants under study (five in our case). We calculated a value of 1.40 kJ/mol for muscle acylphosphatase. The statistical analysis carried out on the 23 561
