ARCHIVES
OF BIOCHEMISTRY
AND
BIOPHYSICS
Vol. 292, No. 1, January, pp. 34-41, 1992
Thermal Stabilities of Mutant Escherichia co/i Tryptophan Synthase a Subunits Woon Ki Lim,l
Christie
Brouillette,2
and John
K. Hardman*T3
Cellular and Developmental Molecular Biology Section, Department of Biological Sciences, University Alabama 35487; and *Department of Medicine, University of Alabama at Birmingham, Birmingham,
of Alabama, Tuscaloosa, Alabama 35294
Received June 10, 1991, and in revised form August 30, 1991
Random chemical mutagenesis, in vitro, of the 5’ portion of the Escherichia coli trpA gene has yielded 66 mutant a subunits containing single amino acid substitutions at 49 different residue sites within the first 121 residues of the protein; this portion of the a subunit contains four of the eight o helices and three of the eight fi strands in the protein. Sixty-two of the subunits were examined for their heat stabilities by sensitivity to enzymatic inactivation (52°C for 20 min) in crude extracts and by differential scanning calorimetry (DSC) with 29 purified proteins. The enzymatic activities of mutant a subunits that contained amino acid substitutions within the a and @ secondary structures were more heat labile than the wild-type a subunit. Alterations only in three regions, at or immediately C-terminal to the first three B strands, were stability neutral or stability enhancing with respect to enzymatic inactivation. Enzymatic thermal inactivation appears to be correlated with the relative accessibility of the substituted residues; stability-neutral mutations are found at accessible residual sites, stabilityenhancing mutations at buried sites. DSC analyses showed a similar pattern of stabilization/destabilization as indicated by inactivation studies. T, differences from the wild-type a subunit varied + 7.6”C. Eighteen mutant proteins containing alterations in helical and sheet structures had T,‘s significantly lower (-1.6 to -7.5%) than the wild-type T, (59.5”f.J). In contrast, 6 mutant a subunits with alterations in the regions following j3 strands 1 and 3 had increased T,‘s (+1.4 to f7.6”C). Because of incomplete thermal reversibilities for many of the mutant a subunits, most likely due to identifiable aggregated forms in the unfolded state, reliable differences in thermodynamic stability parameters are not possible. The availability of this group of mutant a subunits which clearly contain structural alterations should prove useful in defining the roles of certain residues or sequences in the unfolding/folding pathway for this protein when examined by ureajguaninidine denaturation kinetic analysis. 0 1992 Academic Press, Inc.
An increasingly common approach to understanding protein stability is the use of mutant proteins containing amino acid differences from the native protein [for reviews, see (l-3)]. Although substitutions that might be expected to result in a wide variety of differences in covalent and nonconvalent interactions can lead to instability, it appears that, most often, residues critical to stability are located at sites of low solvent accessibility and low mobility in the native structure. A potential bias in a number of the studies is that, although both site-specific and random mutagenesis were employed, there is often a selection for altered activity in addition to temperature sensitivity (4-8). However, the imposition of such a functional screening step can potentially qualify some of the conclusions that are drawn about stability alone. For example, in addition to intrinsic heat stability differences, a change in function may also reflect, for example, changes in sensitivity to proteases and catalytic efficiencies (9-11). This report describes the thermal stability of a number coli trpA encoded mutant tryptophan synof Escherichia thase (TSase)4 (Y subunits obtained by a random mutagenesis procedure with no functional or stability selection. The only selection applied is that there is a trpA gene sequence alteration (12, 13). The (Y subunit consists of a polypeptide chain of molecular weight 28,600 having no prosthetic groups or disulfide bonds. Physiologically, it exists as part of the TSase a& complex which catalyzes the terminal step in tryptophan biosynthesis (indole 3-glycerol phosphate + i This work is in partial fulfillment of the Ph.D. requirements, Department of Biological Sciences, University of Alabama. ’ Present address: Division of Biochemistry, Southern Research Institute, Birmingham, AL 35205. 3 To whom correspondence should be addressed at Department of Biological Sciences, University of Alabama, Tuscaloosa, AL 35487. 4 Abbreviations used: TSase, tryptophan synthase; SDS, sodium dodecyl sulfate; PAGE, polyacrylamide gel electrophoresis; DSC, differential scanning calorimetry.
34 All
0003.9861/92 $3.00 Copyright 0 1992 by Academic Press, Inc. rights of reproduction in any form reserved.
THERMAIL
STABILITIES
OF MUTANT
E. coli TRYPTOPHAN
L-Serine + L-Tryptophan + D-glyceraldehyde 3-phosphate). The three-dimensional structure of the (Y&, complex from Salmonella typh!imurium is known to 2.5 A resolution (14). Because of the high sequence identity (85%) between the CYsubunit from S. typhimurium and that from E. coli and their virtually identical enzymatic/folding properties, it is believed that their three-dimensional structures are very similar (15-M). A limited number of mutant (II subunits have been employed to study various aspects of heat stability (20-22). Here we describe the enzymatic heat stability of 62 mutant CYsubunits altered at 49 different sites within the Nterminal 121 residues. Of these, 29 mutant proteins were purified and examined also by differential scanning microcalorimetry. METHODS
AND
MATERIALS
Enzymes and chemicals. All reagents and enzymes used for the mutagenesis and enzyme analyses have been described before (12, 13). Bacterial strains and mutagenesis. All of the E. coli host strains and overexpression vectors and mutagenesis methods are described in detail elsewhere (12,13). The essential features of the mutagenesis procedure are as follows: (a) fragments of the @A gene are subcloned, mutagenized, separated from the corresponding wild-type (unmutagenized) fragment on urea/formamide gradient gels, and sequenced, (b) selected mutant fragments are inserted in place of the wild-type fragment into an otherwise intact trpA gene which is contained in an overexpression vector. Preparation of crude extracts o/mutant and wild-type a subunits. All proteins were obtained from cells containing overexpression vectors carrying the trpA gene (wild-type or mutant) under the control of the tat-promoter as described before (12,13). Following induction by lactose or isopropylthio P-D-galactoside, (Y subunit production was monitored on SDS-PAGE (23). Optimum conditions for production of each mutant protein were determined before crude extracts were prepared.
of wild-type and mutant a subunits and the p2 subPurification unit. The wild-type and mutant o( subunits (except mutant protein SL33) were purified by the method of Milton et al. (12). The SL33 mutant (Ysubunit was recovered from inclusion bodies using 1 M KSCN (24) and then purified as were the other (Ysubunits. The & subunit was purified by a method described by Lim et al. (24). Proteins were stored at -2O’C under 60% saturated ammonium sulfate or at -80°C in 0.1 M potassium phosphate buffer (pH 7.8) containing 5 mM dithioerythritol, fluoride. All purified 5 mM EDTA, and 0.2 mM phenylmethylsufonyl proteins appeared as single bancls on SDS-PAGE (23). Heat inactiuation. The thermal stability of enzyme activity in crude extracts was determined as previously described (13). The crude extracts (1 mg/ml) of 01subunit in 0.1 M potassium phosphate (pH 7.8) containing 1 mM EDTA and 0.1 mM phenylmethylsulfonyl fluoride were incubated at 52’C for 20 min. After treatment, they were cooled quickly on ice and assayed. The interval between heat treatment and enzymatic assay was kept constant (at 20 min) for all proteins. /3 reaction activity [indole + L-serine + L-tryptophan, assayed according to Kirschner et al. (25)] was utilized for several reasons. It is the most reliable among the three TSase reactions and is the TSase reaction that is least affected by the a subunit mutations. This assay is necessarily performed in the presence of an excess of the pz subunit; saturating p2 subunit levels were determined experimentally for each mutant a subunit before and after heat treatment. This was particularly relevant for mutant 01subunits PL28, GD51, PH(T)58, DG56, AV67, and YC102 which normally have reduced specific activities in the p reaction (26). The “activity remaining” (in Fig. 1B) is determined as specific activity after heat treatment/initial specific activity X 100 (13).
SYNTHASE
cy SUBUNITS
35
Calorimetry. The thermal stabilities of all proteins were determined with a Microcal MC-2 differential scanning microcalorimeter (North Hampton, MA) at a scanning rate of 60”/h, under nitrogen pressure of 26 psi. For several proteins the scan rate 30”/h was employed; no significant differences were noted. The protein concentrations were between 0.90 and 1.97 mg/ml in a solution of 1 mM sodium phosphate (pH 7.2) containing 0.1 mM dithioerythritol and 0.2 mM EDTA. Proteins were thoroughly dialyzed against the above buffer and centrifuged at 40,OOOg for 20 min. Samples were deaerated at approximately 20 mm Hg for 510 min under an atmosphere of argon. Soluent accessibility. The relative static solvent accessibilities of residues were calculated by the ACCESS program of Lee and Richards (27), utilizing the coordinates of the a subunit portion of TSase a.& structure. The assumption here is that the E. coli a subunit alone is similar to the (Y subunit in the S. typhimurium TSase a& complex. The static accessibility of an amino acid residue, X, in model tripeptides Ala-X-Ala as calculated in Lee and Richards (27) was used here as the reference accessibilities of fully extended polypeptide. When an amino acid had two sets of data from different conformations, the average was adopted, and the solvent accessibilities of Glu, Asp, and the LYcarbon of Gly were considered those of Gln, Asn, and the side chain of Gly, respectively. Protein concentration. The concentrations of purified 01and pz subunits were estimated from the extinction values at 278 nm, assuming E$$ = 4.2 for mutant 01 subunit YC(YH)102 and YC115, 4.4 for the other mutant and wild-type o( subunits (28), and 6.5 for the & subunit (29). The protein concentrations in crude extracts were determined by microbiuret method with bovine serum albumin as a standard (30).
RESULTS
Distribution and identity of mutational alterations in the a subunit. Sixty-six singly substituted mutants were obtained (12, 13). These are located at 49 sites in the Nterminal half (residues 1-121) of the LYsubunit (Fig. 1A; mutant designations are given as YC4, SP6, etc.; these refer, respectively, to mutant (Ysubunits in which a wildtype Tyr (Y) residue is changed to a Cys (C) residue at polypeptide position 4; a wild-type Ser (S) residue is changed to a Pro (P) residue at polypeptide position 6, etc.). Single substitutions are found at 34 sites, two substitutions at 13 sites, and three substitutions at 2 sites. Replacements of all residue types except lysine and tryptophan (which is absent in the (Y subunit) were obtained. The substitutions result in changes in charge, hydrophobicity, residue volume, and LYhelix and p strand forming tendencies. Alterations in six of the eight prolyl residues were also found in this portion of the (Y polypeptide. As indicated in Fig. 1,19 altered sites are located in cyhelices, 14 in p strands, and 16 in turns and random coil regions. The alterations are distributed approximately equally between exposed (44%) and buried (56%) residue sites. Heat inactiuation of mutant a subunits in crude extracts. Figure 1B (open bars) shows the results of enzymatic heat inactivation at pH 7.8 of wild-type and mutant proteins in crude extracts. The activity monitored in these experiments was the TSase 6 reaction. The wildtype Q!subunit undergoes a sharp decrease in activity in this reaction between 50 and 60°C; the approximate midpoint of this transition is 52°C. The wild-type (Y subunit retains 61% of its activity under the conditions used (52°C
36
LIM,
BROUILLETTE,
AND
HARDMAN
1 3dhl
QlIM
ELF
a
THERMAL
STABILITIES
OF MUTANT
REMAINING
(%)
SYNTHASE
37
(Y SUBUNITS
folding transition begins at about 5O”C, appears complete by 70°C and shows an apparent single peak with a AC:. The buffer composition adopted here is as described by Matthews et al. (31), except that the pH is slightly different (pH 7.2 vs 7.8), and a low salt concentration is maintained in order to avoid protein precipitation at high temperatures. Reversibility of the unfolding transition was ascertained by rescanning the samples that had been slowly cooled at the end of a heating cycle. Under these conditions reversibility ranged from 15% (when heated to 85°C prior to cooling in order to obtain reliable postdenaturational baselines for AC,d estimations.) to 85% (when the scan was stopped immediately after the denaturational transition). The loss of reversibility results to a large extent from intermolecular disulfide bond formation (despite the deaeration procedure and the low concentration of dithioerythreitol) as judged by the appearance of discrete higher molecular weight bands on SDS-PAGE in heat/cooled preparations. These bands disappear if such preparations are treated with a reducing agent prior to electrophoresis. The inclusion of higher concentrations of reducing agent during the heating phase was avoided because their presence results in erratic, nonreproducible baselines (32). Figure 3 also shows some features about the treatment of the DSC data. Values of T,,,, ACpd,and AH=,,, obtained at this pH for the wild-type cr subunit are T, = 59.5“C; ACE = 2.6 kcal/mol/K; AHHcal= 123 kcal/mol; AS = 371 Cal/mole/K. The use of sigmoidal transitional baselines do not significantly (
OF BIOCHEMISTRY
AND
BIOPHYSICS
Vol. 292, No. 1, January, pp. 34-41, 1992
Thermal Stabilities of Mutant Escherichia co/i Tryptophan Synthase a Subunits Woon Ki Lim,l
Christie
Brouillette,2
and John
K. Hardman*T3
Cellular and Developmental Molecular Biology Section, Department of Biological Sciences, University Alabama 35487; and *Department of Medicine, University of Alabama at Birmingham, Birmingham,
of Alabama, Tuscaloosa, Alabama 35294
Received June 10, 1991, and in revised form August 30, 1991
Random chemical mutagenesis, in vitro, of the 5’ portion of the Escherichia coli trpA gene has yielded 66 mutant a subunits containing single amino acid substitutions at 49 different residue sites within the first 121 residues of the protein; this portion of the a subunit contains four of the eight o helices and three of the eight fi strands in the protein. Sixty-two of the subunits were examined for their heat stabilities by sensitivity to enzymatic inactivation (52°C for 20 min) in crude extracts and by differential scanning calorimetry (DSC) with 29 purified proteins. The enzymatic activities of mutant a subunits that contained amino acid substitutions within the a and @ secondary structures were more heat labile than the wild-type a subunit. Alterations only in three regions, at or immediately C-terminal to the first three B strands, were stability neutral or stability enhancing with respect to enzymatic inactivation. Enzymatic thermal inactivation appears to be correlated with the relative accessibility of the substituted residues; stability-neutral mutations are found at accessible residual sites, stabilityenhancing mutations at buried sites. DSC analyses showed a similar pattern of stabilization/destabilization as indicated by inactivation studies. T, differences from the wild-type a subunit varied + 7.6”C. Eighteen mutant proteins containing alterations in helical and sheet structures had T,‘s significantly lower (-1.6 to -7.5%) than the wild-type T, (59.5”f.J). In contrast, 6 mutant a subunits with alterations in the regions following j3 strands 1 and 3 had increased T,‘s (+1.4 to f7.6”C). Because of incomplete thermal reversibilities for many of the mutant a subunits, most likely due to identifiable aggregated forms in the unfolded state, reliable differences in thermodynamic stability parameters are not possible. The availability of this group of mutant a subunits which clearly contain structural alterations should prove useful in defining the roles of certain residues or sequences in the unfolding/folding pathway for this protein when examined by ureajguaninidine denaturation kinetic analysis. 0 1992 Academic Press, Inc.
An increasingly common approach to understanding protein stability is the use of mutant proteins containing amino acid differences from the native protein [for reviews, see (l-3)]. Although substitutions that might be expected to result in a wide variety of differences in covalent and nonconvalent interactions can lead to instability, it appears that, most often, residues critical to stability are located at sites of low solvent accessibility and low mobility in the native structure. A potential bias in a number of the studies is that, although both site-specific and random mutagenesis were employed, there is often a selection for altered activity in addition to temperature sensitivity (4-8). However, the imposition of such a functional screening step can potentially qualify some of the conclusions that are drawn about stability alone. For example, in addition to intrinsic heat stability differences, a change in function may also reflect, for example, changes in sensitivity to proteases and catalytic efficiencies (9-11). This report describes the thermal stability of a number coli trpA encoded mutant tryptophan synof Escherichia thase (TSase)4 (Y subunits obtained by a random mutagenesis procedure with no functional or stability selection. The only selection applied is that there is a trpA gene sequence alteration (12, 13). The (Y subunit consists of a polypeptide chain of molecular weight 28,600 having no prosthetic groups or disulfide bonds. Physiologically, it exists as part of the TSase a& complex which catalyzes the terminal step in tryptophan biosynthesis (indole 3-glycerol phosphate + i This work is in partial fulfillment of the Ph.D. requirements, Department of Biological Sciences, University of Alabama. ’ Present address: Division of Biochemistry, Southern Research Institute, Birmingham, AL 35205. 3 To whom correspondence should be addressed at Department of Biological Sciences, University of Alabama, Tuscaloosa, AL 35487. 4 Abbreviations used: TSase, tryptophan synthase; SDS, sodium dodecyl sulfate; PAGE, polyacrylamide gel electrophoresis; DSC, differential scanning calorimetry.
34 All
0003.9861/92 $3.00 Copyright 0 1992 by Academic Press, Inc. rights of reproduction in any form reserved.
THERMAIL
STABILITIES
OF MUTANT
E. coli TRYPTOPHAN
L-Serine + L-Tryptophan + D-glyceraldehyde 3-phosphate). The three-dimensional structure of the (Y&, complex from Salmonella typh!imurium is known to 2.5 A resolution (14). Because of the high sequence identity (85%) between the CYsubunit from S. typhimurium and that from E. coli and their virtually identical enzymatic/folding properties, it is believed that their three-dimensional structures are very similar (15-M). A limited number of mutant (II subunits have been employed to study various aspects of heat stability (20-22). Here we describe the enzymatic heat stability of 62 mutant CYsubunits altered at 49 different sites within the Nterminal 121 residues. Of these, 29 mutant proteins were purified and examined also by differential scanning microcalorimetry. METHODS
AND
MATERIALS
Enzymes and chemicals. All reagents and enzymes used for the mutagenesis and enzyme analyses have been described before (12, 13). Bacterial strains and mutagenesis. All of the E. coli host strains and overexpression vectors and mutagenesis methods are described in detail elsewhere (12,13). The essential features of the mutagenesis procedure are as follows: (a) fragments of the @A gene are subcloned, mutagenized, separated from the corresponding wild-type (unmutagenized) fragment on urea/formamide gradient gels, and sequenced, (b) selected mutant fragments are inserted in place of the wild-type fragment into an otherwise intact trpA gene which is contained in an overexpression vector. Preparation of crude extracts o/mutant and wild-type a subunits. All proteins were obtained from cells containing overexpression vectors carrying the trpA gene (wild-type or mutant) under the control of the tat-promoter as described before (12,13). Following induction by lactose or isopropylthio P-D-galactoside, (Y subunit production was monitored on SDS-PAGE (23). Optimum conditions for production of each mutant protein were determined before crude extracts were prepared.
of wild-type and mutant a subunits and the p2 subPurification unit. The wild-type and mutant o( subunits (except mutant protein SL33) were purified by the method of Milton et al. (12). The SL33 mutant (Ysubunit was recovered from inclusion bodies using 1 M KSCN (24) and then purified as were the other (Ysubunits. The & subunit was purified by a method described by Lim et al. (24). Proteins were stored at -2O’C under 60% saturated ammonium sulfate or at -80°C in 0.1 M potassium phosphate buffer (pH 7.8) containing 5 mM dithioerythritol, fluoride. All purified 5 mM EDTA, and 0.2 mM phenylmethylsufonyl proteins appeared as single bancls on SDS-PAGE (23). Heat inactiuation. The thermal stability of enzyme activity in crude extracts was determined as previously described (13). The crude extracts (1 mg/ml) of 01subunit in 0.1 M potassium phosphate (pH 7.8) containing 1 mM EDTA and 0.1 mM phenylmethylsulfonyl fluoride were incubated at 52’C for 20 min. After treatment, they were cooled quickly on ice and assayed. The interval between heat treatment and enzymatic assay was kept constant (at 20 min) for all proteins. /3 reaction activity [indole + L-serine + L-tryptophan, assayed according to Kirschner et al. (25)] was utilized for several reasons. It is the most reliable among the three TSase reactions and is the TSase reaction that is least affected by the a subunit mutations. This assay is necessarily performed in the presence of an excess of the pz subunit; saturating p2 subunit levels were determined experimentally for each mutant a subunit before and after heat treatment. This was particularly relevant for mutant 01subunits PL28, GD51, PH(T)58, DG56, AV67, and YC102 which normally have reduced specific activities in the p reaction (26). The “activity remaining” (in Fig. 1B) is determined as specific activity after heat treatment/initial specific activity X 100 (13).
SYNTHASE
cy SUBUNITS
35
Calorimetry. The thermal stabilities of all proteins were determined with a Microcal MC-2 differential scanning microcalorimeter (North Hampton, MA) at a scanning rate of 60”/h, under nitrogen pressure of 26 psi. For several proteins the scan rate 30”/h was employed; no significant differences were noted. The protein concentrations were between 0.90 and 1.97 mg/ml in a solution of 1 mM sodium phosphate (pH 7.2) containing 0.1 mM dithioerythritol and 0.2 mM EDTA. Proteins were thoroughly dialyzed against the above buffer and centrifuged at 40,OOOg for 20 min. Samples were deaerated at approximately 20 mm Hg for 510 min under an atmosphere of argon. Soluent accessibility. The relative static solvent accessibilities of residues were calculated by the ACCESS program of Lee and Richards (27), utilizing the coordinates of the a subunit portion of TSase a.& structure. The assumption here is that the E. coli a subunit alone is similar to the (Y subunit in the S. typhimurium TSase a& complex. The static accessibility of an amino acid residue, X, in model tripeptides Ala-X-Ala as calculated in Lee and Richards (27) was used here as the reference accessibilities of fully extended polypeptide. When an amino acid had two sets of data from different conformations, the average was adopted, and the solvent accessibilities of Glu, Asp, and the LYcarbon of Gly were considered those of Gln, Asn, and the side chain of Gly, respectively. Protein concentration. The concentrations of purified 01and pz subunits were estimated from the extinction values at 278 nm, assuming E$$ = 4.2 for mutant 01 subunit YC(YH)102 and YC115, 4.4 for the other mutant and wild-type o( subunits (28), and 6.5 for the & subunit (29). The protein concentrations in crude extracts were determined by microbiuret method with bovine serum albumin as a standard (30).
RESULTS
Distribution and identity of mutational alterations in the a subunit. Sixty-six singly substituted mutants were obtained (12, 13). These are located at 49 sites in the Nterminal half (residues 1-121) of the LYsubunit (Fig. 1A; mutant designations are given as YC4, SP6, etc.; these refer, respectively, to mutant (Ysubunits in which a wildtype Tyr (Y) residue is changed to a Cys (C) residue at polypeptide position 4; a wild-type Ser (S) residue is changed to a Pro (P) residue at polypeptide position 6, etc.). Single substitutions are found at 34 sites, two substitutions at 13 sites, and three substitutions at 2 sites. Replacements of all residue types except lysine and tryptophan (which is absent in the (Y subunit) were obtained. The substitutions result in changes in charge, hydrophobicity, residue volume, and LYhelix and p strand forming tendencies. Alterations in six of the eight prolyl residues were also found in this portion of the (Y polypeptide. As indicated in Fig. 1,19 altered sites are located in cyhelices, 14 in p strands, and 16 in turns and random coil regions. The alterations are distributed approximately equally between exposed (44%) and buried (56%) residue sites. Heat inactiuation of mutant a subunits in crude extracts. Figure 1B (open bars) shows the results of enzymatic heat inactivation at pH 7.8 of wild-type and mutant proteins in crude extracts. The activity monitored in these experiments was the TSase 6 reaction. The wildtype Q!subunit undergoes a sharp decrease in activity in this reaction between 50 and 60°C; the approximate midpoint of this transition is 52°C. The wild-type (Y subunit retains 61% of its activity under the conditions used (52°C
36
LIM,
BROUILLETTE,
AND
HARDMAN
1 3dhl
QlIM
ELF
a
THERMAL
STABILITIES
OF MUTANT
REMAINING
(%)
SYNTHASE
37
(Y SUBUNITS
folding transition begins at about 5O”C, appears complete by 70°C and shows an apparent single peak with a AC:. The buffer composition adopted here is as described by Matthews et al. (31), except that the pH is slightly different (pH 7.2 vs 7.8), and a low salt concentration is maintained in order to avoid protein precipitation at high temperatures. Reversibility of the unfolding transition was ascertained by rescanning the samples that had been slowly cooled at the end of a heating cycle. Under these conditions reversibility ranged from 15% (when heated to 85°C prior to cooling in order to obtain reliable postdenaturational baselines for AC,d estimations.) to 85% (when the scan was stopped immediately after the denaturational transition). The loss of reversibility results to a large extent from intermolecular disulfide bond formation (despite the deaeration procedure and the low concentration of dithioerythreitol) as judged by the appearance of discrete higher molecular weight bands on SDS-PAGE in heat/cooled preparations. These bands disappear if such preparations are treated with a reducing agent prior to electrophoresis. The inclusion of higher concentrations of reducing agent during the heating phase was avoided because their presence results in erratic, nonreproducible baselines (32). Figure 3 also shows some features about the treatment of the DSC data. Values of T,,,, ACpd,and AH=,,, obtained at this pH for the wild-type cr subunit are T, = 59.5“C; ACE = 2.6 kcal/mol/K; AHHcal= 123 kcal/mol; AS = 371 Cal/mole/K. The use of sigmoidal transitional baselines do not significantly (
