Thermodependence of Basal and Stimulated Rat Liver Adenylate Cyclase A RE-EVALUATION*

cyclase activity from a purified rat liver plasma membrane preparation was assayed at 15 to 20 temperatures between 17” and 39”. Reaction conditions were such that: (a) the substrate concentration was saturating (ATP 2 mM; MgCl, 5 mM); (b) ATPH+, a possible adenylate cyclase inhibitor, was absent (pH: 8.2); and (c) linearity of reaction velocity was effective throughout the 5-min incubation. The Arrhenius plot of basal activity exhibited two slopes with a break at 25”. The apparent energy of activation (E,) was 29 kcallmol below 25” and 0.8 kcallmol above 25”. Irreversible thermodenaturation did not explain the low E, above 25”. When the experiment was performed with increasing concentrations of NaF (from 2.5 to 15 mM), the break at 25 persisted. However, for temperatures above 25”, E, increased hyperbolically Similar data were obtained in the presence

Adenylate cyclase activity from a purified rat liver plasma membrane preparation was assayed at 15 to 20 temperatures between 17" and 39". Reaction conditions were such that: (a) the substrate concentration was saturating (ATP 2 mM; MgCl, 5 mM); (b) ATPH+, a possible adenylate cyclase inhibitor, was absent (pH: 8.2); and (c) linearity of reaction velocity was effective throughout the 5-min incubation. The Arrhenius plot of basal activity exhibited two slopes with a break at 25". The apparent energy of activation (E,) was 29 kcallmol below 25" and 0.8 kcallmol above 25". Irreversible thermodenaturation did not explain the low E, above 25". When the experiment was performed with increasing concentrations of NaF (from 2.5 to 15 mM), the break at 25 persisted. Below 25", E, was always equal to 29 kcallmol. However, for temperatures above 25", E, increased hyperbolically from 0.8 to 18 kcallmol as a function of increasing NaF concentrations. Similar data were obtained in the presence of increasing concentrations of glucagon (from 0.5 nM to 1 FM), except that the break point shifted toward a higher temperature.
At a saturating concentration, guanyl-5'-yl imidodiphosphate, another adenylate cyclase activator, drastically modified the temperature-activity relationship so that the Arrhenius plot showed no break, and E, was equal to 24 kcallmol both above and below 25". We demonstrate here that the increase in E, produced by all the activators tested is related to an increase in entropy of activation.
In contrast to most enzyme systems, adenylate cyclase (ATP pyrophosphate lyase (cyclizing), EC 4.6.1.1) possesses the double privilege of being a membrane-bound and multiregulated enzyme system (1). However, despite extensive research, its mechanism remains obscure. The present study was undertaken to assess the functional relationship of the various membrane components during activation by hormonal or nonhormonal ligands in purified membranes from rat liver. It is based on the fact that membrane-bound enzymes are characterized by a particular thermodependence as evidenced by a typical, two-sloped Arrhenius plot. Changes in this thermodependence have been reported as a function of the nature of the surrounding fatty acids, or in the presence of various agonists (2-8). We, therefore, explored the influence of some effecters on the thermodependence of adenylate cyclase in rat liver plasma membrane.
The present study demonstrates that: (a) basal adenylate cyclase displays a two-sloped Arrhenius plot; (b) stimulation of basal adenylate cyclase activity by hormonal or nonhormonal ligands at temperatures above 25" is paralleled by an apparently paradoxical increase in the enzyme E,l; (c) this increase in E, is related to an increase in the entropy of activation; (d? fluoride, glucagon, and guanine nucleotides appear to transform the enzyme system into the same final state. Furthermore, in the presence of these activators, the thermodynamic state of the enzyme appears similar to that observed at temperatures below the critical point at which the break of the Arrhenius plot occurs.
The physical significance of the entropy of activation is difficult to assess and there is a tremendous gap going from the entropy of activation to molecular mechanism. However, we suggest as a working hypothesis that activation of adenylate cyclase is paralleled by the elimination of diffusionlinked hindrance to the enzymatic function caused by the domain surrounding the enzyme at temperatures above 25".

DISCUSSION2
In the absence of activators, hepatic adenylate cyclase exhibits an abrupt drop in the slope of the Arrhenius plot at 25". The abruptness of this change allows determination of well defined values for energy of activation below and above the break point (29 kcal/mol and 0.8 kcal/mol, respectively) (Fig. 2). The value of 29 kcal/mol represents a very high energy of activation for a soluble enzyme but is a common value for membrane-bound enzymes. On the other hand, 0.8 kcal/mol is a very low value. We verified that this low value could not be explained either by thermal irreversible denaturation (Fig. 3) or by a modification of phosphodiesterase or ATPase activities (data not shown).
Such a thermodependence pattern for membrane-bound enzymes has been related to the physical properties of the membrane, especially to a phase change in the lipid component of the membrane. This possibility is supported by the results of studies concerning: (a) differential scanning calorimetry (2); (b) the use of lipophilic spin probes (15, 16); and (cl modifications in the nature of membrane fatty acids (17). However, due to the large mole fraction of cholesterol in most plasma membranes and assuming that cholesterol is equally distributed in the membrane, it is unlikely that the change in E, is only linked to the lipid composition of the membrane. Moreover, two-sloped Arrhenius plots are not the specific attribute of membrane-bound enzymes; soluble enzymes, such as lipase, trypsin, and pepsin have been studied at very low temperatures and in each case there is a large change in activation energy at about 0". For lipase the values are 37 and 7.6 kcabmol below and above o", respectively. It does not appear that a phase change of the water from liquid to solid is the cause of the break, for the same results are obtained if freezing is prevented by the addition of high concentrations of glycerol (18). In fact, from an enzymatic point of view, this type of Arrhenius plot can be accounted for by (a) a transconformation between two forms of the enzyme in equilibrium, both of them characterized by different energies of activation as proposed by Massey et al. (19) for the enzyme n-amino acid oxidase; (6) the existence of two parallel reactions (18); or (cl by two successive reactions with different energies of activation (20). For theoretical reasons, our observation on adenylate cyclase cannot be explained by both latter assumptions: in the case of two parallel reactions, the Arrhenius plot should be concave upwards (18); in the second case, the difference between both activation energies would have to be at least 200 kcal/mol (18).
At all the temperatures studied, all the activators (NaF, Gpp(NH)p, glucagon) increased adenylate cyclase velocity. However they altered the thermodependence of the enzyme differently. In the presence of increasing concentrations of NaF, the slope of the Arrhenius plot above the break point (25") was increased hyperbolically from 0.8 to 18 kcal/mol. The extrapolated value of energy of activation, at maximally effective fluoride concentration, was 24 kcal/mol. The slope of the Arrhenius plot below the break point (25") remained unaltered and corresponded to an E, of 29 kcabmol at all NaF concentrations. Similarly, Gpp(NH)p brought about complete linearization of the Arrhenius plot by raising E, above 25 from 0.8 to 24 kcal/mol. Glucagon had a more complex effect since it not only increased the slope of the Arrhenius plot above the break point, but also shifted this point toward the higher temperatures as a function of its concentration.
At a maximally effective glucagon concentration, full linearization of the plot was not achieved.
The low activation energy of basal adenylate cyclase activity at temperatures above 25" has already been reported (6)(7)(8) and several authors also observed that stimulated activities are characterized by higher energy of activation (21-25). In particular, Harwood and Bodbell showed that activation of adipose tissue adenylate cyclase by fluoride was minimal below 25" (26). Bodbell et al. also reported that basal hepatic cyclase activity was relatively unaffected by temperature whereas Gpp(NH)p-stimulated activity displayed marked increase when the temperature was raised from 30 to 37". This was interpreted as a seemingly selective effect of temperature on stimulation by Gpp(NH)p (1). Orly and Schramm further assumed that the temperature-sensitive reaction of the adenylate cyclase in turkey erythrocytes involved the regulatory GTP site. They suggested that the break in the Arrhenius plot reflected the repeated functional introduction of GTP by the hormone into the regulatory site (3). More recently, Houslay et al. reported that adenylate cyclase in rat liver plasma membranes stimulated by fluoride or Gpp(NH)p yielded linear Arrhenius plots whereas activation by glucagon resulted in a biphasic Arrhenius plot with a well defined break at 28.5 2 1" (4, 27). Finally, Engelhard et al. reported that the basal adenylate cyclase activity of purified plasma membrane from fibroblasts (LM-cells) exhibited striking triphasic temperature dependence. Incubation in the presence of prostaglandin PGE 1 eliminated the break at 25" and linearized the Arrhenius plot (8). Our data are in good agreement with the results described above. However, all these data have an apparently paradoxical consequence, namely, that adenylate cyclase activation by any stimulating agent is paralleled by an unexpected increase in the energy of activation of the reaction. In our system, adenylate cyclase activation at temperatures above 25" is accompanied by an increase in the E, from 0.8 kcal/mol to 18, 24, and 12, in the presence of maximal concentrations of fluoride, Gpp(NH)p, and glucagon, respectively (Figs. 2, 5, 6).
It should be noted that the energy of activation comprises two terms, one related to the reaction rate constant, the other involving the entropy of activation.
Both terms may vary independently, so that if the entropy term increases, a rise in E, does not necessarily mean a drop in the reaction rate; according to Arrhenius, the empirical relationship between the rate constant and the E, is rendered by the following equation: Ink= -s+lnA where k is the rate constant, R the gas constant, T the absolute temperature, and A the velocity constant. However, A involves an entropy term, which appears in the following equation: where K is Boltzmann's constant, h Plan&s constant, ASS the entropy of activation and AH$ the enthalpy of activation, related to the experimental energy of activation as follows: AH$ = E, -RT. Equation 2 explains that a drop in E, (or in AH%) does not necessarily coincide with an increase in the reaction rate (k) if this drop is associated with a change in the entropy term (AS*). This equation further shows that an increase in E, associated with an increase in the reaction rate necessarily indicates a rise in entropy of activation.
When applied to our data, these considerations indicate that the various activators (fluoride, Gpp(NH)p, glucagon) which increase enzyme velocity at all temperatures, would do connected with a decrease in entropy of activation: A&$ -AS,* = '+ Thermodependence of Rut Liver Adenylate Cyclase so concomittantly with an increase of entropy of activation above 25", which would account for the observed rise in E,. In addition, when the incubation temperature was increased from low temperatures (~25") to temperatures higher than 25", we observed a drop in energy of activation (maximal in the basal state, and reduced dose-dependently in the presence of each different activator). It is simple to show that the change from low to high temperature (involving a drop in E,) is related to a reduction in the entropy of activation; the demonstration takes into account the continuity in the reaction rate at the temperature where the change in AHS and AS* occurs. At this break point CT,,), the velocity of the reaction is the same for both states. Let the suflix L indicate the thermodynamic constants of the reaction at the temperature below the break, and the suffix H, the constants at the temperature above the break. We can write that: In k, = In k, that is from Equation 2: or This pattern may fit in with the idea that raising the temperature above the 25" break point coincides with the appearance of an ordered process during the reaction. To the contrary, the activators, or a drop in temperature, would eliminate this additional process, It is tempting to assume that the latter is linked to diffusion; temperature, as well as effecters of the enzyme, would influence the conformation of the enzyme in such a way as to increase its maximal velocity and also modify its relationship to the surrounding environment. In fact, we wish to emphasize that adenylate cyclase, and probably most membrane enzymes possess kinetic behavior similar to that of immobilized enzyme systems which are known to be controlled by diffusional effects (28). For these immobilized enzyme systems also, at sufficiently low temperature, in the domain of kinetic control of the reaction, the true energy of activation is observed. At sufficiently high temperature, when the reaction become bulk diffusion controlled, the observed rate is essentially independent of temperature and the apparent energy of activation is close to zero (28).

AH,S -AH& = T,(AS,S -AS&) REFERENCES
The drop in E, occurring when the temperature is raised is References 1 to 28 are found on p. 841.