Coenzyme A activation of acetyl-CoA carboxylase.

Acetyl-CoA carboxylase is activated by physiological concentrations of CoA. Activation of partially purified enzyme by CoA is accompanied by a decrease in the Km for acetyl-CoA from 0.2 mM to about 4 microM, which is the physiological concentration of acetyl-CoA in the cytosol. CoA activation of the purified enzyme is accompanied by an increase in the Vmax, without changing the Km for acetyl-CoA. The Km for acetyl-CoA of the purified enzyme is about 10 to 40 microM. The purification procedure results in a decrease in the Km for acetyl-CoA; under these conditions, CoA activation does not cause further lowering of the Km. CoA activation is accompanied by polymerization of the enzyme. However, CoA activation is not causally related to polymerization. There is one CoA binding site/subunit of acetyl-CoA carboxylase. CoA binding at that site is not affected by the presence of citrate, but palmityl-CoA inhibits CoA binding. CoA alone cannot reverse palmityl-CoA inhibition of the carboxylase. Bovine serum albumin and CoA together can activate the palmityl-CoA-inhibited enzyme. This indicates that the involvement of bovine serum albumin-like protein, CoA, and palmityl-CoA may play a physiologically significant role in the control of acetyl-CoA carboxylase.

sible physiological role for CoA in the activity of carboxylase. However, when the effect of CoA was examined with purified enzyme, the activation effect was minimal, indicating differences in the properties of partially purified and highly purified enzyme preparations ( 7 ) .
In this communication, we show that the purified enzyme does respond to CoA activation when it is treated with Dowex 1-X8, a step which is used to remove tightly bound citrate (8). There is one CoA binding site on the carboxylase. Binding of CoA to this site activates the enzyme by changing either the V,,, or the K , for acetyl-coA, depending on the purity of the enzyme preparation. Binding of CoA is inhibited by palmityl-CoA but not by citrate. Additional characteristics of the CoA activation of carboxylase are presented.
Wistar rats (230 to 280 g) from our departmental rat colony were maintained on a commercial rat diet and fed ad libitum Purification of Acetyl-coA Carboxylase-Three different procedures have been used in the purification of acetyl-CoA carboxylase. Each procedure yielded enzyme with somewhat different properties presumably due to proteolytic activities during the purification. First, acetyl-coA carboxylase was purified by the method of Nakanishi and Numa (9), except that all buffers contained 0.2 m M phenylmethylsulfonyl fluoride. This procedure routinely gave a preparation with a specific activity of 12 units/mg of protein and a subunit molecular weight of 220,000, as well as proteolytic products of molecular weights 120,000 and 118,000.
To minimize enzyme degradation, the purification procedure of Witters and Vogt was utilized (IO). In the absence of NaF, this procedure yielded a preparation with a specific activity of 3.4 units/ mg of rotein, a subunit molecular weight of 240,000, and essentially no proteolytic products. The low specific activity of this enzyme preparation has been explained on the basis of its high phosphate 2289 content (about 5 mol of phosphate/mol of enzyme (10). The molecular weight of the enzyme is significantly higher than that obtained by the method of Nakanishi and Numa (9). The high specific activity of the fKst preparation could thus be due to partial proteolysis since it is well known that such proteolytic action activates the carboxylase many fold (11,12).
Finally, to obtain carboxylase with minimal modification of its native state, an aftinity chromatography column was used as follows. The partially purified enzyme preparation, at the 30% ammonium sulfate precipitate stage was incubated with CoA-Sepharose (417 milliunits of acetyl-coA carboxylase/0.5 ml gel) for 30 min a t 37OC. The gel mixture was then placed in a small disposable Pasteur pipette and the column was washed extensively with phosphate buffer until no protein could be detected in the eluate. The column was then washed successively with 2 ml each of 0.2 mM AMP, 0.2 mM NAD', and 4 mM CoA. The carboxylase was eluted by the CoA treatment. Carboxylase prepared by this method was not completely pure since it showed traces of lower molecular weight contaminants. The specific activity of enzyme prepared by this method was 1 unit/mg of protein.
Assay of Acetyl-CoA Carboxylase-Acetyl-coA carboxylase was assayed by measuring the formation of [lJC]malonyl-CoA as described by Majerus et al. (13). The standard reaction mixture contained the following compounds in a final volume of 150 pl: 50 mM Tris-HC1; pH 7.5; BSA,' 150 pg; 10 mM citrate; 10 mM MgC12; 1 mM dithiothreitol; 0.2 mM phenylmethylsulfonyl fluoride; 4.1 mM ATP; 0.13 mM acetyl-CoA; and 13.3 mM KHI4CO.~ (1.5 X loti cpm). The reaction was started by the addition of an appropriate amount of enzyme which had been preincubated a t 37'C for 30 min. The reaction was carried out for 2 to 3 min and terminated by the addition of 25 p1 of 6 N HCl. One hundred pl of the reaction mixture was plated on a disk of glass fibre filter, dried under an infrared lamp, and the radioactivity was measured. One unit of activity was defined as 1 pmol of malonyl-CoA formed/min a t 37°C. In those experiments where CoA activation was examined, citrate was omitted from the reaction mixture unless otherwise specified.
Preparation of Citrate-free Enzyme-Dissociation of tightly bound cit.rate was carried out by the method of Hashimoto et al. (8). Dowex I-X8 was washed extensively and equilibrated with a large excess of 50 mM Tris-HC1, pH 7.5, containing 0.25 M sucrose, 1 mM EDTA, and 5 mM P-mercaptoethanol. Three-tenths ml of the resin was packed into a small column (a disposable Pasteur pipette). One hundred pl of the purified enzyme (100 pg) in the same buffer was then passed through the column. The column was washed with the same buffer and 100-pl fractions were collected. The recovery of the enzyme was quantitative and the specific activity was unchanged before and after the column treatment (I2 to 14 units/mg) when the activity was measured in the presence of 10 mM citrate. This treatment removes 99% of the residual citrate (8). r3H]CoA Binding to the Carboxylase-One hundred forty p1 of the Dowex-treated enzyme (0.1 mg/ml) was incubated with different concentrations of ["HICoA (32 X 10" cpm/pmol) in a total volume of 200 p1 for 20 min at 37°C. Following incubation, 170 p1 was placed on a Millipore fiter (HAWP02400 pore size 0.45 p), which had been presoaked in ice-cold 50 mM Tris-HCI, pH 7.5, containing 0.25 M sucrose, 1 mM EDTA, and 5 mM P-mercaptoethanol for 1 h and washed with 10 ml of the same buffer under vacuum just before the samples were applied. The samples on the filters were then quickly washed with 10 ml of the same buffer, air-dried, and the radioactivity was counted.
Sucrose Density Gradient Centrifugation-A 0.2-ml enzyme sample was applied to a 5-ml linear gradient of 8 to 20% sucrose containing 50 mM Tris-HC1 buffer, pH 7.5, 1 mM EDTA, and 5 mM 8-mercaptoethanol. The CoA and citrate concentrations used in the gradients is indicated in the appropriate figures. Samples were centrifuged for 90 min (or as specified) at 25°C in an SW 65 swinging bucket rotor a t 45,000 rpm.

Activation of Purified Acetyl-coA Carboxylase by CoA-
In contrast to the partially purified enzyme, the purified acetyl-coA carboxylase does not respond to CoA activation (7). This difference between purified and crude enzyme has been investigated in terms of the following hypotheses: ( a ) Conventional purification procedures expose the carboxylase I The abbreviation used is: BSA, bovine serum albumin. to high citrate concentrations for an extended period of time in order to facilitate the purification. Such treatment causes significant changes in the enzyme structure due to the presence of tightly bound citrate on the enzyme. When citrate is tightly bound to the enzyme, activation by CoA may be insignificant. ( b ) It is known that the carboxylase is very sensitive to partial proteolysis (11,12). Partial proteolysis may make the purified carboxylase unresponsive to CoA activation; or (c) in vivo the carboxylase normally exists in an inactive state as a result of association with some inhibitor which is dissociated during purification. CoA may function in the removal of this hypothetical inhibitor.
Previously, Hashimoto et al. (8) used Dowex 1 to remove most of the tightly bound citrate from the purified carboxylase. When the purified carboxylase was treated with Dowex 1 according to the procedure of Hashimoto et al. (8), the treated enzyme became very sensitive to CoA activation ( Fig.  1). As shown in Fig. 1, CoA activation is a time-dependent process, as is also the case with citrate activation; maximum activation is obtained at 20 min at 37°C with GOA. Although Dowex treatment is known to remove the tightly bound citrate, it is uncertain whether anything else is removed.
Our search for the presence of palmityl-CoA in the purified enzyme according to the procedure of Ogiwara et al. (17) yielded negative results.
Recently, Witters and Vogt (10) introduced a simplified purification procedure for acetyl-coA carboxylase using polyethylene glycol. This procedure also exposes the carboxylase to high citrate, but only for a brief period. Carboxylase prepared by this method has a higher molecular weight (240,000) than that of enzyme prepared by Nakanishi and Numa's procedure (220,000). As reported by Witters and Vogt (lo), no degraded enzyme species (M, = 118,000 and 120,000) were found following sodium dodecyl sulfate gel electrophoresis.  Nakanishi and Numa (9) and then treated with Dowex I-X8 to remove citrate was incubated with or without 0.15 m~ CoA in medium containing 1 mM dithiothreitol, 1 mM theophylline, and 100 pg of BSA a t 37OC. Omission of theophylline from the incubation has no effect on CoA action. Aliquots were withdrawn at the indicated times for the enzyme assay. Control no CoA, 0.15 mM CoA, M .
When this enzyme preparation was treated with CoA, it was activated as in the case with the partially purified enzyme preparations (7), without Dowex treatment (Fig. 2). As will be shown later, the nature of CoA activation of these two purified enzyme preparations is different from that of the partially purified enzyme preparations. Although these experiments do not illuminate the reasons for the unresponsiveness of purified enzyme to CoA activation, they do show that the purified enzymes do respond to GOA activation under appropriate conditions. Also, these experiments indicate that purification procedure causes enzyme modification. The enzyme prepared by the method of Witters and Vogt (10) has about one-third of the specific activity of the enzyme prepared by the method of Nakamishi and Numa (9). Although it has been suggested that the low specific activity might be due to the presence of the highly phosphorylated form of the carboxylase, the experiments presented here suggest the possibility that less proteolysis and less bound citrate may account for the difference in specific activity. Tests of the Dowex treatment was not possible with the Witters and Vogt enzyme preparation because the enzyme was tightly bound to the Dowex and could not be eluted from it, possibly due to the presence of excess phosphate groups on the enzyme.
The previous studies with partially purified enzyme showed that CoA activation of the carboxylase was accompanied by enzyme aggregation (7). Since enzyme aggregation is a function of protein concentration, the effect of CoA on different enzyme concentrations was examined, as shown in Fig. 3. The maximum activation of carboxylase was observed at 140 PM CoA and higher concentrations of CoA slightly decreased the degree of activation. The concentration of CoA required for the half-maximum activation with three different enzyme concentrations is the same (15 PM CoA) and the effect was independent of the enzyme concentration. These concentrations of the carboxylase undergo polymerization in the presence of CoA as will be seen later (Figs. 5 and 6). This observation suggests that the observed aggregationper se may not be directly related to the enzyme activation. This interpretation is further supported by results which will be discussed later.
As in the case of the partially purified enzyme, CoA acti-  -) or without CoA (o"-o) in a medium containing 50 mM Tris/CI (pH 7.5), BSA (1 mg/ml), 1 mM dithiothreitol, and 1 mM theophylline a t 37°C for 20 min before assaying for the activity at different concentration of acetyl-coA as indicated. mixture not only completely abolished the activation effect of CoA, but the addition of Mg'+ to the activated enzyme also results in its prompt deactivation. The exact mechanism of Mg'+ action has yet to be established. However, it has been reported that divalent cations such as Mn2+ do cause gross structural changes of the CoA molecule (14). This effect of Mg'+ on CoA action could alternatively be due to its binding to the enzyme rather than CoA. As will be shown later, Mg2+ inhibits CoA binding to the carboxylase. Relationship between CoA-mediated Aggregation a n d Activation ofthe Carboxylase-It has long been established that the protomers of the carboxylase are inactive and only the polymers are active (see Ref. 1). Although CoA activation of the partially purified enzyme was accompanied by enzyme aggregation and it has been suggested that these two phenomena are related ( 7 ) , the experiments described in the previous section suggest that CoA activation is independent of carboxylase concentration. Therefore, the relationship between the two events was investigated further.
When the sedimentation behavior of enzyme purified by the method of Nakanishi and Numa (9) was examined in the absence of free citrate (citrate was removed by dialysis), the carboxylase sedimented as a protomer. As shown in Fig. 5, when 0.15 mM CoA was added to the purified enzyme, about two-thirds of the enzyme was polymerized. However, the effect of CoA in the activation of this type of carboxylase was less than 208, showing virtually no activation effect in spite of the polymerization. On the other hand, when the same di- in a medium containing 50 mM Tris/Cl (pH 7.5), 1 mM dithiothreitol, 0.2 mM phenylmethylsulfonyl fluoride, BSA ( I mg/ml), 1 mM theophylline at 37°C for 20 min. Samples were placed on sucrose gradients and subjected to centrifugation as described under "Materials and Methods." Acetyl-CoA carboxylase was assayed in the presence of 10 mM citrate. Purified enzyme which had been dialyzed as described in Fig. 4 was used and 100 pg of the dialyzed enzyme was passed through 0.3 ml bed volume of Dowex 1-X8 column which has been equilibrated with the same buffer as dialysis. The treated enzyme was incubated with M or without 0.15 mM CoA (A-A) in a medium as described in Fig. 4. a t 37°C for 20 min. Samples were placed on sucrose gradients, subjected to centrifugation, and assayed as described in Fig. 4. alyzed enzyme was subjected to Dowex treatment, and the sedimentation pattern was examined, the carboxylase sedimented as a protomer in the absence of CoA activation. When CoA was added to the Dowex-treated enzyme, two-thirds of the enzyme was again polymerized as in the case of the dialyzed enzyme (Fig. 6). When the effect of CoA was examined with respect to enzyme activation, the Dowex-treated enzyme was activated &fold. In this experiment, the carboxylase assay of the fractions were assayed in the presence of 10 mM citrate and the effect of CoA in the figure is not apparent.
These experiments suggest that aggregation of carboxylase occurs in the presence of CoA once the enzyme is exposed to citrate irrespective of the tightly bound citrate. However, the CoA activation of the carboxylase occurs only when the enzyme is treated with Dowex. Further evidence that the two events, enzyme activation and polymerization, are not causally related comes from the following experiments. When the partially purified carboxylase (35% ammonium sulfate stage in the procedure of Nakanishi and Numa (9)) was prepared in buffer without citrate, the enzyme preparation sedimented as protomers, as shown in Fig. 7 . When this enzyme preparation was activated by CoA, the activation was about 2-fold, without the formation of polymers. This experiment indicates that in enzyme preparations which have not been exposed to citrate, CoA activation can occur without enzyme aggregation. Under the same conditions, the addition of citrate causes enzyme polymerization (data not shown).
CoA Binding Site on Acetyl-coA Carboxylase-The experiments discussed above indicate that CoA activation is not directly related to CoA-mediated polymerization of the carboxylase. T o elucidate the relationship between CoA and citrate action, we have examined the binding site of CoA on the carboxylase and the effect of various ligands on CoA  Nakanishi and Numa (9) in the absence of citrate at the first ammonium sulfate stage. The enzyme was incubated with ( D " 0 ) or without (M) 0.6 mM CoA in a buffer containing 50 mM Tris-HCI, pH 7.5, 1 mM dithiothreitol, 1 mM theophylline, and BSA (1 mg/ml) for 20 min a t 37°C. The enzymes were then subjected to the sucrose density gradient centrifugation (5 to 20%) and the activity in the fractions was assayed as described under "Materials and Methods." There was no enzyme activity from fractions 1 to 15. binding. The effects of different concentrations of CoA on the binding was examined and the results are presented as a Scatchard plot, in Fig. 8. The results indicate that the purified enzyme exhibits one binding site/subunit. The molecular weight of the subunit is taken to be 220,000 (15). The dissociation constant obtained from the plot is 0.26 mM. Using the Dowex I-X8-treated enzyme in the same experiment also gave one binding site/subunit with a K d of 0.17 mM. This K,I value is about IO-fold higher then K , value obtained from Fig. 3 . This point is further discussed under "Discussion." These experiments indicate that the lack of CoA activation of the enzyme before Dowex treatment is not due to the absence of a CoA site, but appears to be due to the presence ofcitrate, or an equivalent ligand, removably by Dowex.
The independence of CoA binding to the presence of citrate is supported by studies of the effects of various ligands on the binding of CoA to the purified enzyme (Table I). In this experiment, the Dowex-treated enzyme (30 pg) was incubated with 1 mM CoA in the presence of different ligands. The addition of 2 mM citrate did not affect CoA binding at all, indicating that the CoA binding site is different from that for citrate. This conclusion is in accordance with our findings concerning the mode of carboxylase activation by citrate and CoA using the partially purified enzyme (7). CoA binding to the carboxylase was inhibited by palmityl-CoA. The presence of 10 PM palmityl-CoA abolishes almost 85% of the CoA binding. Mg" also reduces the binding of CoA as shown in Table I. The same concentration of Mg2+ completely abolished the CoA activation, as shown in Fig. 4. However, the binding studies indicate only a 50% decrease in CoA binding (Table I). FIG. 8. Scatchard plot for CoA binding of acetyl-coA carboxylase. Purified carboxylase (1 mg/ml) which has been dialyzed against buffer containing 50 mM Tris/Cl (pH 7.5), 0.25 M sucrose, 1 rnM EDTA, 5 mM P-mercaptoethanol at 4°C for overnight were incubated with different concentration of ['HICoA (6400 cpm/nmol) a t 37°C for 20 min. The samples were then placed on Millipore filters which have been washed with 10 ml of ice-cold buffer. The filters were then dried under IH lamp for 5 min and the radioactivity on the filters were counted in a scintillation counter. Effect of CoA Activation on the K,, for Acetyl-CoA-The K, for acetyl-coA of the carboxylase varies extensively depending upon the purification procedure used and the degree of purification. As shown in column 1 of Table 11, the K,,, for acetyl-coA is relatively high at 0.2 mM in the partially purified enzyme preparations. Such high K,,, values are sustained until the enzyme solution is subjected to high citrate concentrations for 24 h during the purification procedure (9). Following this treatment, the K,,, for acetyl-coA is drastically lowered, ie.
about 10-fold (data are not shown).
The enzyme prepared by the procedure of Witters and Vogt (10) has a higher molecular weight than that prepared by the procedure of Nakanishi and Numa (9), but this procedure also involves a brief exposure of the enzyme to a high concentration of citrate. This enzyme has a K,, for acetyl-coA of 80 p~, which is much lower than that of the partially purified enzyme (0.2 mM). When these highly purified enzyme preparations are activated by CoA, the K,, values for acetyl-coA remain unchanged, as shown in Table 11. In these experiments, the Nakanishi-Numa enzyme preparation was subjected to Dowex treatment to show the CoA effect, whereas the Witters-Vogt

Effect of CoA on palmityl-CoA-inactiuated enzyme
Purified enzyme which had been dialyzed in the Tris buffer containing no citrate was incubated with 0.24 PM Palrnityl-CoA in the buffer containing 50 mM Tris-HC1, pH 7.5, 1 m M dithiothreitol, 1 mM theophylline for 15 min at 37°C. Following this incubation, additional ligands were added as indicated and the reaction mixtures were further incubated for 30 rnin at 37°C. The carboxylase was assayed for 3 min in the absence of citrate and BSA. preparation was simply dialyzed free of citrate. With the latter preparation, the CoA effect was pronounced without requiring the Dowex treatment.
When the effect of CoA-affinity chromatography was examined with partially purified enzyme (35% ammonium sulfate stage) or more highly purified enzyme preparations, the K,, for acetyl-coA was as high as 0.2 mM. When these enzymes were activated with 0.125 mM CoA, both enzyme preparations showed the presence of a species of enzyme with an extremely low K,, for acetyl-coA. As shown in Fig. 9, the exact determination of low K,,, values was difficult because of the gradual changes in the curve. However, one can estimate the lowest K,,, for acetyl-coA to be in the range of 2 to 5 PM.
Thus, it is clear that the purified enzymes which have a low K,,, for acetyl-coA are activated by CoA by a mechanism that only involves changes in the V,,,,,. On the other hand, CoA activation of the enzyme in its more native state is accompanied by a decrease in the K, for acetyl-coA.

Effect of CoA on the Reactivation of Palmityl-CoA-inactiuated
Carboxylase-As indicated in our studies of CoA binding to carboxylase, the binding of CoA was effectively inhibited by palmityl-CoA but not by citrate. Since it has long been known that palmityl CoA inactivates carboxylase (16-18), it was of interest to determine whether CoA could reactivate palmityl-CoA inactivated enzyme. As shown in Table  111 ,UM) inactivates the carboxylase about 80%. Addition of CoA (108 PM) alone could not reverse the palmityl-CoA inhibition. The addition of BSA (0.1 mg/ml) restored enzyme activity. However, when both BSA and CoA were added the enzyme was activated to about 3 times the original activity. This experiment indicates that CoA alone cannot reverse the palmityl-CoA effect. Considering the extremely low K,, of palmityl-CoA compared to that of CoA, it is not surprising to see the lack of CoA effect in this regard. However, BSA which is known to bind palmityl-CoA does restore the inactivated activity; thus, when BSA and CoA are both present, CoA can activate the palmityl-CoA-free enzyme.

DISCUSSION
As indicated in the introduction, the short term regulation of acetyl-coA carboxylase appears to involve both allosteric control, mediated by cellular metabolites such as citrate and palmityl-CoA, and covalent modification by a phosphorylation-dephosphorylation sequence. Experimental evidence increasingly suggests that in those systems where covalent modification is involved in the control of enzyme activity such modifications result in changes in the enzyme's properties toward various allosteric cellular metabolites. High concentrations of citrate have long been known to activate acetyl-coA carboxylase; thus, it has been proposed to be a physiological positive effector of this rate-limiting enzyme for long chain fatty acid synthesis (1). However, this theory has been criticized since the lack of a correlation between cellular concentrations of citrate and lipogenic activity made it difficult to understand how citrate alone could regulate carboxylase activity. When the covalent modification mechanism was discovered and the associated changes in the properties of the dephosphorylated carboxylase were examined (6) it was found that the active form of the enzyme required only 0.2 mM citrate for activation compared to 2.4 to 45 mM (6,19). Thus, when lipogenesis is stimulated by a hormone such as insulin, the active species of the carboxylase can function with cellular concentrations of citrate. From these considerations, it becomes obvious that one cannot reject the role of citrate in the stimulation of lipogenesis. Gross changes in citrate concentration are not necessarily required for enzyme activity when the covalent modification mechanism functions in conjunction with the allosteric molecules.
However, there was an additional problem in the short term regulation of acetyl-coA carboxylase which could not be answered in terms of the allosteric and covalent modification mechanisms. Although one cannot easily assess the native state of acetyl-coA carboxylase, earlier studies by Swanson et a2. (11) showed that the majority of the rat liver enzyme occurs in an inactive state which can be activated by partial proteolysis. Indeed, in the crude state, the rat liver carboxylase is customarily activated by citrate before the enzyme is assayed (20). When the partially purified carboxylase was examined for the K, of acetyl-coA, the K,,, was extremely high  (27). It is interesting to note that while citrate is constantly degraded in the process of acetyl-coA production during lipogenesis, CoA only serves as an acyl group carrier and is constantly regenerated. Also, it should be noted that CoA is not a competitive inhibitor of acetyl-coA carboxylase with respect to acetyl-CoA.
A significant role for CoA in the control of lipogenesis becomes more apparent when one considers recent reports that fatty acid synthase requires CoA for activity (28) and that the phosphorylated form of citrate lyase shows decreased activity compared to the dephosphorylated form only in the presence of low CoA concentrations (29). Our studies on the effect of CoA on acetyl-coA carboxylase (7), along with those of others on fatt,y acid synthase (28) and citrate lyase (29), indicate that cellular CoA may play a significant role in the control of lipogenesis at the three important steps catalyzed by these enzymes. The cytosolic concentration of CoA in normal cells has been reported to be 5 pM when liver cells were fractionated by lyophilization and homogenization in an organic solvent followed by density gradient centrifugation (21). However, when cells from fasted animals were fractionated by the modified digitonin method (30), the cytosolic concentration of CoA was about 23 p~. Since cells from fasted animals contain only about one-fifth of the normal concentration (24), it follows that the cytosolic concentration would be about 115 PM in normal cells. Direct homogenization of liver tissue yielded about 130 PM CoA in the cytosol (31). In this case, the breakage of mitochondria was assessed by mitochondrial marker enzymes. Siess et al. (23) reported a value of 0.1 mM for CoA in normal fed cells. However, they used a cytosolic water content value of 1.05 ml/g of dry cells (23) in their calculation compared to 2 ml used by others (24). The CoA content in the cell fluctuates depending upon the medium in which the cells are prepared (23). However, the cytosolic concentrations are high enough to affect carboxylase activity.
Similar confusion exists as to the cytosolic concentration of acetyl-coA (23). The cytosolic acetyl-coA concentration of perfused rat liver has been determined as 0.005 m~ (21, 22), whereas Siess et al. reported an acetyl-coA concentration of 0.086 mM (23). It is difficult to assess which value is a more realistic representation of the in vivo situation. The K , for acetyl-coA of the partially purified enzyme is 0.2 mM, although that of the purified enzyme is about 10 to 25 p~ in the presence of 10 mM citrate. Thus, during the purification of the carboxylase either a protein which affects affinity for acetyl-CoA is removed, or exposure to artificially high citrate concentrations or proteolysis affects the affinity. The purification procedure of Witters and Vogt (10) takes less time and yields a carboxylase with a higher molecular weight. This enzyme preparation, which appears to have experienced less proteolysis, has a much higher K, for acetyl-coA (80 p~) , and is readily activated by CoA. However, the K , for acetyl-coA was not affected.
In our preliminary report, we tentatively concluded that the site of CoA action is different from that of citrate because under saturating concentrations of citrate, the effect of CoA activation was apparent with the partially purified enzyme (7). In the present studies, binding experiments with CoA using purified enzyme before or after Dowex treatment showed one binding site/subunit. Citrate did not interfere with CoA binding, supporting our previous conclusion that the CoA binding site is not the same as that for citrate. However, CoA binding was inhibited by palmityl-CoA, which brings up an interesting argument for the physiological role of palmityl-CoA. It has been suggested that palmityl-CoA is the negative feedback allosteric molecule for acetyl-coA carboxylase (17). Using partially or highly purified chicken liver enzyme, it has been shown that palmityl-CoA inactivation is reversed by BSA (32). In the present studies, palmityl-CoA inactivation of the purified enzyme was reversed by BSA, which binds to palmityl-CoA. When CoA was added together with BSA, the palmityl-CoA-inhibited enzyme activity was restored by BSA, and the carboxylase is then activated by CoA. Since CoA alone cannot reverse the palmityl-CoA inhibition, even at a 50-fold higher concentration, it is very unlikely that CoA replaces palmityl-CoA which could then bind to BSA. It should be emphasized that rat liver acetyl-coA carboxylase occurs in a more or less inactive state (11) and that it has been customary to activate the partially purified enzyme with citrate before enzyme assays are performed. It is not clear whether the inactive state of the acetyl-coA carboxylase found normally occurring is due to palmityl-CoA or not. However, our attempt to show the presence of palmityl-CoA was negative. However, even in the event that some protein similar to BSA (33) could dissociate inhibitory molecules, such as palmityl-CoA, the de-inhibited enzyme still could not function in the presence of the very low concentration of acetyl-coA in the cytosol. Therefore the present studies showing that CoA can lower the K, for acetyl-coA to the 2 to 4 PM range is very significant as an observation that might explain the functioning of acetyl-coA carboxylase under i n vivo conditions.
In the present studies, it was observed that the purified enzymes are activated by changes in the V,,,,,, although the K, for acetyl-coA is not affected. Purification procedures for the carboxylase lower the K,, for acetyl-coA, most likely by the effect of proteolysis. This hypothesis is based on the observation that BSA, alone or with citrate, does not lower the K,, for acetyl-coA in the partially purified enzyme preparation. Finally, the relationship between the covalent modification mechanism and CoA activation should be mentioned.
Our preliminary experiments indicate that the phosphorylated species of carboxylase does not respond to CoA activation. It should be noted that the K,, value for CoA is about 10-fold higher than the K, for CoA (Figs. 3 and 8). Since the K , value is affected by the substrates of an enzyme (34), it is difficult to assess the significance of this large difference in Kc, and K<,.
It has been shown previously that the K,, value should not be directly compared to K,, (34).