Enzyme Variants in Thermal Acclimation

Citrate synthase occurs in two kinetically distinguishable forms, termed C and W in cold and warm acclimated trout, respectively. Oxalacetate and acetyl-CoA saturation curves for both enzymes are hyperbolic and do not show any substrate inhibition. The apparent K,, values for the C form increase with temperature above 10”. At saturating (0.1 mM) levels of oxalacetate, the K,, of acetyl-CoA is 0.05 m&r at lo”, about one-fifth the value at 35”. At saturating levels of acetyl-CoA, the &,, of oxalacetate is 7 pM at lo”, about one-fourth the value at 35”. These values at low temperature are similar to the K,,, constants for the W form below 15”, but above 1.5”, the K,, values for the W form are distinctly lower than for the C form. In the upper biological temperature range for both cold and warm acclimated trout, the Q10 for citrate synthase is about 1.0 at low substrate concentrations because the increase in K,, compensates for the increasing thermal energy. At saturating substrate concentrations, the temperature characteristics of the two enzymes are the same, and the calculated activation energies are 8.8 kcal per mole. Both C and W forms of citrate synthase are strongly inhibited by ATP, which increases the Km of acetyl-CoA (5fold at pH 7.5 and 22”) but does not alter the calculated V,,,,. ATP inhibition is noncompetitive with respect to oxalacetate, and the oxalacetate saturation curve remains hyperbolic in the presence of ATP. The Ki of ATP for the C form increases dramatically with temperature above 15”, while the Ki of the W form is thermally insensitive to nearly the lethal limit for trout. The calculated V,,,,, values for both C and W forms increase 4-fold between pH 6.5 and 8.5, but the Km values are rather insensitive to pH changes. The Kc of ATP, on the other hand, increases 4-fold between pH 6.5 and 8.5. Since the pH in poikilotherms increases at low temperature, these effects of pH on citrate synthase may (a) reduce the Qlo of the reaction irrespective of substrate concentrations, and (b) reduce thermal effects on ATP inhibition.


SUMMARY
Citrate synthase occurs in two kinetically distinguishable forms, termed C and W in cold and warm acclimated trout, respectively.
Oxalacetate and acetyl-CoA saturation curves for both enzymes are hyperbolic and do not show any substrate inhibition.
The apparent K,, values for the C form increase with temperature above 10". At saturating (0.1 mM) levels of oxalacetate, the K,, of acetyl-CoA is 0.05 m&r at lo", about one-fifth the value at 35". At saturating levels of acetyl-CoA, the &,, of oxalacetate is 7 pM at lo", about one-fourth the value at 35". These values at low temperature are similar to the K,,, constants for the W form below 15", but above 1.5", the K,, values for the W form are distinctly lower than for the C form.
In the upper biological temperature range for both cold and warm acclimated trout, the Q10 for citrate synthase is about 1.0 at low substrate concentrations because the increase in K,, compensates for the increasing thermal energy. At saturating substrate concentrations, the temperature characteristics of the two enzymes are the same, and the calculated activation energies are 8.8 kcal per mole.
Both C and W forms of citrate synthase are strongly inhibited by ATP, which increases the Km of acetyl-CoA (5fold at pH 7.5 and 22") but does not alter the calculated V,,,,.
ATP inhibition is noncompetitive with respect to oxalacetate, and the oxalacetate saturation curve remains hyperbolic in the presence of ATP. The Ki of ATP for the C form increases dramatically with temperature above 15", while the Ki of the W form is thermally insensitive to nearly the lethal limit for trout.
The calculated V,,,,, values for both C and W forms increase 4-fold between pH 6.5 and 8.5, but the Km values are rather insensitive to pH changes. The Kc of ATP, on the other hand, increases 4-fold between pH 6.5 and 8.5. Since the pH in poikilotherms increases at low temperature, these effects of pH on citrate synthase may (a) reduce the Qlo of the reaction irrespective of substrate concentrations, and (b) reduce thermal effects on ATP inhibition. For lX~lll\ yc:uY, biologists hare known that the aerobic metabolism of l~oikilotl~ern~ic tissues depends critically upon the * This work was supported by National Rcsenrch Council of C:\ll:lda Gl.illlt, A-X%. acclimation temperature of the organism. Over acclimation periods of 1 to several weeks, t'lle rc+iration of many poikilothermic systems is adjusted, often in a mnm~cr that conrpcnsntes for the new thermal regime (see Reference 1 for a i,cccnt review). Short term or immediate coml)cnsution of aerobic nletabolism in poikilothermic tissues, less commonly ob~rvc~l, is also known. Thus, succinate, pyruvate, and acetate oxidation rates by fish tissues and by mitochondria prcparcd from intertidal organisms are stable over most of each species' thermal range (2)(3)(4).
Sinrilarly, 02 consumption by tuna muscle minces is insensitive to temperature betlveen 5 and 35" (5). Although the functiorlal advantages of such short term and long term t~hcrmal compensations are clear, underlying euzymic n~ccl~:rnisn~s arc not.
In this context, n-e initiated ;I study of the properties of citrate synthase (citrate oxalacctate-lynx (('oar-acctyl:ltillg), from liver tissue of cold and wmi accliniatcd trout. In other organisms examined, the enzyme occurs as a siirgle molecular species strongly modulated by the adenylatcs (6). In the trout, v-e found that liver citrate synthasc occurs in t1v-o kinetically dist~inguishable forms. One form occurs in cold (2') acclimated organisms, xvhile an alternate form of citrate synthase occurs in the liver of warm (18") acclimated organisms.
In the upper biological temperature range for both cold and warm nc-&mated fish, increases in the apparent Michaelis constants compensate for increasing temperatures; in consequence, at low substrate concentrations, the & of the reaction catalyzed by each variant is reduced.

Eqwimental
Animals-The salmonids are generally eurythermal, and their thermal relations as a group are relatively well documented at several levels of organization (7). Our esperimental organism, Salmo gairdneri, was obtained from a local hatchery.
The animals used lvere all from a small brood stock; hence, genetic variability was substantially lower than in wild stock rainbow trout.
The trout were acclimated to 2 and 18' by holding them at these temperatures for at least 4 weeks. Photoperiod conditions and feeding rates were similar for both acclimation groups.
Enzyme &soy-Citrate synthase activity was followed by measuring the increase in absorbance at 412 mp resulting from the reaction of CoA-SH, liberated in t,he enzymic reaction, rrith DTNBl (8). The assay cuvette (l-cm light path) contained 25 PM DTNB, varying concentrations of ncety-Coh and oxal- Inc., Westbury, Sew York). When necessary, care was taken to set the pII of Tris-IICl buffer at the assay temperatures, since the pK of Tris is temperature-dcpencent.
The homogenate was centrifuged at about 5000 x g for GO min, and the pellet, which contains only about 1055 of the total citrate synt.hase activity, was discarded.
The supernntnnt was brought to 35 5;. saturation n-it.h ammonium sulfate (209 g per liter), stirred for at least 1 hour, and centrifuged. The precipitate, which showed no activity, was discarded. The supernatant XIS brought to 70"; saturation with ammonium sulfate (472 g per liter) and again stirred for 1 hour. After cent,rifugntion, the precipitate, which shorted all of the remaining citrate synthase activity, was taken ~111 in 0.1 M Tris-HCl buffer, PI-I 7.5. After dialysis, the preparation was used directly as a source of citrate sSnt,hase activit'y.
The enzyme was stable upon freezing for nt least 3 to 4 weeks.
1soelect~o~ocusing-The electrofocusing column had a volume capacity of 110 ml and n-as equipped with double cooling jackets (LKB-Produckter X13, Stockholm, Sweden). The column was held at 3" by a refrigerated, conTt:rnt temperature bath and cir- culator. The p1-I 3 to 10 gradient was achieved with mnpholytc LKB 8141 (a mixture of polyaminopolycarbosylic :Icids). The cathode solution ~~1s routinely placed at the bottom of the column. The high speed supernatant solution of freshly prepared tissue homogenates, rrhich was used as a source of enzyme, was layered on the sucrose-ampholyte about halfway up the column. InSal voltage was usually set at 300 volts; initial current was about 5 ma. Upon completion of the run, the column was drained at about 10 to 15 drops per min and collected with :m LKB fraction collector.
Fractions were assayed for citrate synthnse activity as described above. The l&I of each fraction was measured at 20" \&h a Radiometer $1 meter (Copenhagen, Denmark).
Reagents forms, which are thought to represent oxidized and reduced states of a single protein species (9,10).
In the case of 2" acclimated trout, liver citrate synthnse usually migrat'es as a single recoverable activity peak during isoelectrofocusing in pH 3 to 10 gradients (Fig. 1). For convenience, we refer to this form of the enzyme as the C form or the 2" enzyme.
In 18" acclimated fish, citrate synthase also migrates as a single major peak, which we designate the W form of the enzyme. In short (10 hours) parallel runs, in which activity loss during electrofocusing is reduced, the C and W forms of the enzyme can be readily distinguished (Fig. 1 trofocusing indicates that the isoelectric points for both enzymes are below pH 5.0 but leads to about 905: loss of citrate synthase activity.
We hare no information on the nature of enzyme inactivation. Recorerable activity of both C and W forms is usually in single peaks, although minor shoulders are occasionally observed.
The major and minor peaks may represent oxidizedreduced states of the enzyme or alternate isozymes, ax suggested by the minor shoulder appearing in the 18" preparation in Fig. 1 is seen more clearly when the K,, values, determined by Linen-eaver-Burk plots, are plotted against temperature (Fig. 4). It is evident than the apparent K, values of both C and W variants increase with temperature over the upper t,hermal range. The apparent K, values for acetyl-Coil and oxalacetate of the C form appear to be more temperature-dependent than those of the W form; the minimum K, values for the C form are about onefifth the K, values at 35". In the case of the W form, the minimum K, values of acetyl-Co-4 and oxalacetate are only about one-half the values at 35". The minimum K, values occur at about 10".
1 consequence of the Km-temperature relationship is the reduction of &lo. At low concentrations of substrates, the decrease in apparent K, can compensate completely for reduced thermal energy. This is evident in Arrhenius plots for both the C and W forms of the enzyme under conditions of saturating (0.1 mM) oxalacetnte conceiitratioii~ aiitl T-arious acctyl-('oh concentrations (Fig. 5). TTnder these condition> above 1 O", the & for the C form is about 1 at 0.025 1~131 acetylM'o.\ and ran be less than 1 at 0.01 nihI ncetyl-%I.
This complex behavior is not sho\~n by the trout enzymes over the temperature range that we studied (Fig. 5), and the arrhenius plots appear to be linear to temperatures beyond the lethal limit (25-27") for salmonids (12).
Bden$ute Modulation-As pointed out above, the 1,e:rction catalyzed by trout liver citrate synthases follows uormnl hlichaelis kinetics with respect to both substrates.
*411\IP and ADP also inhibit trout liver citrate synthase. .\1IP is notably less effective; in :I typical experiment, S mar .\I\IP reduced the activity to only SO:'; of control.
IIDl', 011 the other hand, is nearly as effective an inhibitor as &YrP; sin&u. couccntrations of ADP or hTP yield comparable perceutages of inhibition. -4Dl' inhibition is also competitive with re$pcct to acetyl-Cob and noncompetitire with respect to oxalacetate. We have not examined the ADP inhibition in further detail and hare assumed that its behavior Tvould mimic ATP inhibit~ion.
X major kinetic distinction between the C and W forms of citrate synthase is the thermal sensitivity of ATP inhibition. Wibh the W variant, the Ki of ATE' is relatively insensitive to temperature up to at least 22" but rises sharply above this temperature (Fig. 8). At 35", the K; is about 2-fold higher than at 22". The Ki of ATP for the C form increases dramatically with temperature abore 15"; below 15", the Ki is insensitive to further temperature decrease, and the absolute value ia very similar to the Ki for the W form of the enzyme. Thus, over the normal thermal ranges for both cold and warm acclimated trout, ATP modulation of liver citrate synthasc is essentially independent of temperat'ure. These changes in Ki appear to parallel the changes in the K, of acetyl-CoA; since the Ka of ATP may represent binding, this similarity suggests that the K, of acetyl-CoA may accurately represent an affinity constant.

Efect of pH on Trout Liver Citrate
Synthases-Rnlm (14) has shown that poikilotherms regulate H+: OH-ratios in extracellular fluids rather than strictly regulating the pH as is commonly observed in the homeotherm.
In maintaining a constant H+:OH-ratio in the blood, an increase in blood 1'11 occurs at lower temperatures which parallels the increase in the ionization constant of water. Recent evidence appears to bear out Rahn's origiunl suggestion that a similar requirement for regulation of intracellular H+:OH-ratio confronts the poikilotherm (15). For these reasons, an examination of the role of pH in the regulation of trout citrate synthase seemed necessary.
Although trout liver citrate synthases do not show a pH optimum between pH 6.5 and 8.5, the maximum catalytic activity increases quite strikingly as the pH is raised. Thus, the calculated V,,, at pH 8.5 for both forms of the enzyme is about 4 times higher than at pH 6.5 (Fig. 9). Similar effects of pH upon V ,,,&X of mammalian citrate synthase have been observed (11). In the physiological pH range of blood in fishes (pH 7.3 to 8.0), the catalytic activity of the enzyme appears to be most sensitive to pH changes. The K, values of acetyl-CoA and oxalacetate do not vary between pH 7.5 and 8.5, but the Ii, of acetyl-CoA is decreased at pH 6.5 to two-thirds the value at pH 7.5.
An important effect of pH upon the Zii of ATP is seen wit#h citrate synthases of cold and warm acclimated fish (Fig. 10). Increasing the pH from 7.0 to 8.5 results in about a 4-fold increase in the K; of ATP. Between pH 7.5 and 8.0, the Ki for the W form of the enzyme appears slightly less stable to pH change than the C form. It is clear, however, that efficiency in tivo of ATP modulation of both the C and W forms of citrate synthase must depend critically upon intracellular pH. T, effect of temperatrue. The pH at any temperature is taken as the pH of blood at that temperature (14).

DISCUSSIOS
A number of recent studies indicate that citrate synthase is modulated by the ndcnylates. The enzyme from tissues of several animal species, from yeast, and from mitochondria of plants (16)(17)(18)(19) is strongly inhibited by ATP, which causes an increase in the apparent Michaelis constant for acetyl-CoA. ADP and AMP, when tested singly, may also be inhibitory, but these inhibitors are less effective than ATP. In Escherichiu co&, the situation is somewhat more complex.
Citrate synthase is inhibited by ATP only at pH values above 7.5; at lower pH values, ATP acts as a positive modulator (19). The catalytic rates in the presence of AMP, however, are greater than those with ATP at all pH values examined.
Moreover, in E. coli, other "products" of the Krebs citrate cycle, NADH and a-ketoglutarate (20, al), serve as negative modulators, and a similar situation apparently prevails generally in gram-negative bacteria (22). The inhibition of citrate synthase is presumed to be of the allosteric type because the enzyme can be sensitized to its specific inhibitors without loss of enzymic activity; also, from direct electron microscope observations, it appears that citrate synthase undergoes important allosteric transitions upon binding modulators (23). Our studies are consistent with this general picture and indicate that the trout enzymes display characteristics intermediate between those of the mammalian and the bacterial enzymes.
It is critical, in considering possible physiological consequences of observed enzyme-ligand interactions, to take into account the biological context of the enzyme reaction involved.
In this connect,ion, the unique problems of metabolic control in poikilotherms living under variable and often unpredictable thermal regimes have been largely overlooked.
The relationship between the apparent K, and temperature, for example, suggests that the temperature coefficient of the reaction is decreased at low substrate concentrations, and therefore that, in wivo, the Q10 of the reaction will depend on cellular substrate concentrations.
At low substrate concentrations, when the K, is important in determining the reaction velocity, the Q10 will be low; at higher substrate concentrat,ions, the Q10 will increase. These considerations would not hold at temperatures below about lo", at which levels the apparent K, values appear to be less thermally sensitive.
We have no information on cellular concentrations of acetyl-CoA and oxalacetate but, according to recent estimates, these may be assumed to be in the same range as, or lower than, the K, values for the trout enzymes (24,25).
We have already noted that the K i values of ATP for both the C and W forms of citrate synthase increase with pH ( Fig. 10) and with temperature (Fig. 8). Under normal biological conditions, these effects of pH and temperature may cancel each other because pH decreases with increasing temperature. This is shown for the C form of the enzyme in Fig. 12, in which the pH at any given temperature is taken as the blood pH at that temperature.
A possible deduction from these data is that, in vivo, the ATP-binding site of citrate synthase is modulated by pH to maintain XTP control equally effective over a larger part of the thermal range of the species. Another consequence of cellular pII varying with temperature relates to the pH optimum for enzyme function.
Because increasing pH activates trout citrate synthnse (Fig. 9), the temperature coefficient of the reaction is decreased from about 1.7 to about 1.45 when the pH is adjusted according to Rahn's pII-temperature relationship (14). This serves as an iutercsting mechanism by which the effective &lo of the reaction in viva can be decreased over a broad thermal range, irrespective of substrate concentrations.
This kind of mechanism, together with the I<,temperature relationship, which at low substrate levels also serves thermally to xt,abilize the reaction, may contribute to the low &lo values that have been reported for aerobic metabolism of liver preparations from rainbow trout (Z), as well as for tissues from other poikilotherms (3,5).