Studies on the Mechanism of Fatty Acid Synthesis ACETYL-AND MALONYLTRANSACYLASE THE PIGEON LIVER FATTY ACID

SUMMARY Evidence is presented to show that the transacylase components of the pigeon liver fatty acid synthetase complex catalyze the transacylations of acetyl and malonyl groups from their CoA thioesters to the thioesters of Escherichia coli acyl carrier protein (ACP). The acetyl-and malonyltransacylase activities are inhibited by N-ethylmaleimide, although it appears that a sulfhydryl group is not directly involved in the transacylations of acetyl and malonyl groups to ACP. Protection of both transacylase activities against N-ethylmaleimide inhibition can be achieved by previously incubating the synthetase with either acetyl-CoA or malonyl-CoA.


SUMMARY
Evidence is presented to show that the transacylase components of the pigeon liver fatty acid synthetase complex catalyze the transacylations of acetyl and malonyl groups from their CoA thioesters to the thioesters of Escherichia coli acyl carrier protein (ACP).
The acetyl-and malonyltransacylase activities are inhibited by N-ethylmaleimide, although it appears that a sulfhydryl group is not directly involved in the transacylations of acetyl and malonyl groups to ACP.
Protection of both transacylase activities against N-ethylmaleimide inhibition can be achieved by previously incubating the synthetase with either acetyl-CoA or malonyl-CoA. Indoacetamide, while totally inhibiting triacetic acid lactone formation and fatty acid synthesis, has no effect on the acetyl-and malonyltransacylase activities. Acetyl-CoA competitively inhibits the malonyltransacylase activity and malonyl-CoA competitively inhibits the acetyltransacylase activity.
CoA is an inhibitor of both transacylase activities, inhibiting competitively with respect to E. co2i ACP.
Treatment of the synthetase with 0.5 M guanidine hydrochloride followed by diethylaminoethyl cellulose chromatography results in the separation of the acetyl-and malonyltransacylase activities from the bulk of the synthetase protein.
Further separation from synthetase protein is achieved by Sephadex G-ZOO chromatography, but very low yields of the purified transacylase activities are obtained. It has not been possible to separate the acetyl-and malonyltransacylase activities from each other.
The initial reactions of fatty acid synthesis as catalyzed by the pigeon liver fatty acid synthetase are the transacylations of * This investigation was supported in part by Grant GM-06242- 11  acetyl and malonyl groups from their CoA thioesters to the acyl binding sites of the synthetase (2)(3)(4).
One of these acyl binding sites has been shown to be the sulfhydryl group of the 4'-phosphopantetheine moiety of the fatty acid synthetase (2,4). Preliminary evidence indicates that the 4'-phosphopantetheine of the pigeon synthetase is part of an "acyl carrier protein-like" protein component of this multienzyme complex (5). We have found that in addition to catalyzing the transacylations of acetyl and malonyl groups to the 4'-phosphopantetheine of the pigeon synthetase, the transacylase activities of the complex will also transacylate acetyl and malonyl groups from their CoA thioesters to Escherichia coli acyl carrier protein.
By use of this transfer to E. coli ACPl to assay the acetyl-and malonyltransacylase activities we have investigated the general and kinetic properties of these enzymes as they exist in the intact synthetase complex. We have also attempted to fmctionate the a.cet,yl-and malonyltransacylase activities from the pigeon synthetase.
In this paper we report the results of these studies and relate our findings to the integrated functioning of the pigeon liver fatty acid synthetase (2).

Materials
ACP was purified from a crude acid precipitate of E. coli as previously described (6). Acetyl-ACP and malonyl-ACP were synthesized as described previously (7). All other reagents used were obtained as presented in the previous paper (2).

Methods
Enzyme Preparations-The pigeon liver fatty acid synthetase was purified by the method of Wakil et al. (8). The subfractions of the E. coli fatty acid synthetase, EI1 and EIII, were prepared as previously described (Q), with the following modifications.   for the acetyl-and malonyltransacylase activities of the pigeon synthetase were incubated at room temperature for 10 min. The reactions were stopped with the addition of 0.5 ml of 10% trichloracetic acid and the resulting acid-precipitable material collected by centrifugation.
The precipitates were washed twice with 2.5% trichloracetic acid, once with diethyl ether, and were suspended in water.
The pH was adjusted to 6.5 with sodium bicarbonate and the resulting solutions were stored at -20".
Formation and Chromatography of Hydroxamates-The acidprecipitable products from the transacylase reaction mixtures were reacted with neutral hydroxylamine according to the method of Stadtman and Barker (12). Chromatography of the resulting hydroxamates was carried out on Eastman Chromagram thin layer silica gel sheets (13) in n-butyl alcohol-water (100:18). E.

Preparation of WAcetyl and W-Ma&y1
Synthetase-W-Acetyl and W-malonyl synthetase were prepared and isolated as described in the preceding paper (2). Separation of Transaylase Activities from Pigeon Synthetase-Fatty acid synthetase (100 mg), routinely stored in 0.5 M potassium phosphate (pH 6.5) containing 1 mM dithiothreitol, was desalted by passage through a Sephadex G-25 column (1.2 X 19 cm) equilibrated in 0.01 M Tris-HCl (pH 7.5; hereafter referred to as Tris buffer).
The resulting enzyme solution (16 mg per ml) was stored at 4" for 45 to 60 hours. At the end of this period the enzyme solution was diluted to 10 mg per ml with Tris buffer, made 0.5 M in guanidine hydrochloride, and incubated for 30 min at room temperature.
The enzyme solution was then dialyzed against 2 liters of Tris buffer for 2 hours. The dialyzed enzyme was applied to a DEAE-cellulose column (1.2 x 20 cm; Whatman DE-52) equilibrated in Tris buffer. The column was eluted with a 400.ml linear gradient of 0 to 0.5 M NaCl in Tris buffer.
Fractions containing the transacylase activities which had separated from the bulk of the synthetase protein were pooled and concentrated by ultrafiltration (Amicon model 50 ultrafiltration cell; Diaflo UM-2 membrane). Determination of Protein-Protein was determined by the method of Lowry et al. (14).  We have observed that the acetyl and malonyl derivatives of E. coli ACP can serve as acyl donors for fatty acid synthesis as catalyzed by the pigeon liver fatty acid synthetase.
As can be seen from Fig. 1, identical rates of fatty acid synthesis were obtained when either the acyl-CoA or acyl-ACP substrates were used. This finding prompted us to determine whether the synthetase might also catalyze the transacylation of acetyl and malonyl units from their CoA thioesters to E. coki ACP. Such a transacylation could form the basis of a convenient assay for the acetyl-and malonyltransacylase activities of the pigeon synthetase complex.
When 14C-acetyl-CoA or 14C-malonyl-CoA was incubated with E. coli ACP in the presence of the pigeon liver fatty acid synthetase, radioactivity was transferred from an acid-soluble form to an acid-insoluble form. This transfer of radioactivity was dependent upon the amount of fatty acid synthetase protein added (Fig. 2). The acid-precipitable radioactivity obtained from incubations containing no ACP was not significantly above background levels. These results suggested that the acetyl-and malonyltransacylase activities of the pigeon synthetase could catalyze the transacylation of acetyl or malonyl groups from their CoA thioesters to E. coli ACP.
To establish that such a transacylation was occurring, several properties of the acidprecipitable radioactivity were examined.
Ident$cation of Transacylation Products-When the acidprecipitable radioactivity, recovered following incubation of '4C-malonyl-CoA, ACP, and the fatty acid synthetase, was mixed with carrier ACP and chromatographed on DEAE-cellulose, the elution pattern shown in Fig. 3  The elution pattern of the ACP was determined enzymically with the ACP-dependent, E. coli fatty acid synthesis assay. Identical results were obtained with the acid-precipitable radioactivity recovered from an incubation of 14C-acetyl-CoA, ACP, and pigeon synthetase.
If 14C-acetyl-CoA and 14C-malonyl-CoA were mixed with ACP in the absence of pigeon synthetase and this mixture chromatographed on DEAE-cellulose, the radioactivity eluted considerably earlier in the NaCl gradient than did the ACP (as indicated by the broken line in Fig. 3).
To establish that the acetyl and malonyl groups were not modified during their incubation with the pigeon synthetase, the acid-precipitable reaction products were reacted with neutral hydroxylamine and the resulting hydroxamates chromatographed on silica gel. The product of the acetyltransacylase activity yielded a radioactive hydroxamate which cochromatographed (RF = 0.39), with an authentic acetyl hydroxamate, and the product of the malonyltransacylase activity yielded a radioact,ive hydroxamate which cochromatographed (RF = 0.03)) with an authentic malonyl hydroxamate.
Further proof that the acetyl and malonyl groups retained their identity throughout their incubation with the synthetase was obtained in the following manner.
When the radioactive acid-precipitable product of the acetyltransacylase activity was incubated with nonradioactive malonyl-CoA, TPNH, and the E. coli fatty acid synthetase, the radioactivity was incorporated into fatty acids (Table I). Similar results were obtained when the radioactive product of the malonyltransacylase activity was incubated with nonradioactive acetyl-CoA, TPNH, and the E. coli fatty acid synthetase.
Finally, to show that the acetyl or malonyl group was attached to the ACP through a thioester linkage, performic acid oxidations of the acid-precipitable products were carried out. Performic acid oxidation of either the acid-precipitable product of Studies  Table II). 811 of these results are consistent with the pigeon synthetase's having transacylase activities capable of catalyzing the transacylation of both the acetyl and malonyl moieties from the CoA thioesters to thioesters of E. coli ACP.

General
Properties of Transacylase Activities-The relationship between the acetyl-and malonyltransacylase activities and pH is shown in Fig. 4. The optimum pH for the acetyltransacylase activity was around 6.5 while that for the malonyltransncylase activity was around 7.
Both the ace&l-and malonyltransacylase activities were relatively sensitive to heat. A 2-min treatment of the fatty acid synthetase at 50" reduced both activities by SOY,.
We have found, as did Joshi, Plate, and Wakil (2) and Lirodie, Wasson, and Porter (3) that acetyl and malonyl groups are transacylated from their CoA thioesters to covalent binding sites on the pigeon synthetase.
Moreover, we have shown this process to be reversible, as free CoA can remove all of the synthetase-bound acetyl or malonyl units (2). That ilCP can also elicit this reverse reaction is shown in Fig. 5. Fig. 5A shows the radioactive elution pattern that was obtained when 14C-acetyl synthetase was passed through a Sephadex G-100 column.
The synthetase emerged from this column in the void volume and had radioactivity associated with it. If the 14C-acetyl synthetase was previously incubated with ACP prior to being chromatographed on Sephadex G-100, the elution patt'ern shown in Fig. 5B was obtained.
The radioactivity which previously had emerged in the void volume disappeared and a new peak of included radioactivity appeared. Assaying this new peak for ACP with the E. coli fatty acid synthetase showed that the radioactivity was cochromatographing with ACP, suggesting that the r4Cacetyl group had been transferred from a synthetase-binding site to the ACP.
The effects of the sulfhydryl-alkylating agent N-ethylmaleimide on the acetyl-and malonyltransacylase activities are given in Table III. An N-ethylmaleimide concentration of 0.8 rnnc resulted in approximately 80% inhibition of both transacylase activities of the native synthetase.
However, iodoacetamide over the concentration range of 0.5 to 10 mM had no effect on either transacylase activities.
It was further found that the acetyltransacylase activity could be protected against N-ethylmaleimide inhibition by preliminary incubation with either acetyl-CoA or malonyl-CoA (Table IV). Protection of the malonyltransacylase activity against N-ethylmaleimide could also be effected with either acetyl-Co4 or malonyl-CoA. Bt all concentrations of acetyl-CoA and malonyl-CoA tested, acetyl-CoA proved to be the better protecting agent for both of the transacylase activities.
It has been shown previously by Bressler and Wakil (16)  Assays for acetyl-and malonyltransscylase activity were carried orlt as described under "Experimental Procedllre," except that the indicated concentrations of E. coli ACP and CoA were llsed. Units on the ordinate are (nanomoles of substrate transacylated per min)-I. fatty acid synthesis, as catalyzed by the pigeon liver synthetase, is sensitive to a variety of sulfhydryl inhibitors.
The insensitivity of the transacylases to iodoacetamide, under the incubation conditions used, prompted us to test the iodoacetamidet.reated synthetase for its ability to catalyze the synthesis of fatty acids. As can be seen from Fig. BA, synthetase treated with iodoacetamide, that was fully active for transacylation, was totally inhibited with respect to its ability to catalyze fatty acid synthesis.
Nixon, Putz, and Porter (17) have shown that in the absence of TPNH, the pigeon liver synthetase catalyzes the formation of triacetic acid. This is a B-carbon compound, two carbons deriving from acetyl-CoA and four carbons deriving from malonyl-Cob, which is released from the synthetase as a lactone. The condensation of these three Cs units is presumably catalyzed by the condensing enzyme unit of the pigeon synthetase complex. With the synthesis of triacetic acid lactone (TAL) as a measure of the condensing activity of the synthetase, it was possible to show that in the iodoacetamide-treated synthetase, condensation activity was completely inhibited (Fig. 6B). It was possible, therefore, to inhibit condensing activity without affecting transacylation by initially treating the synthetase with iodoacetamide. This finding allowed us to examine kinetically the effects of malonyl-CoA on acetyltransacylation and the effects of acetyl-CoA on malonyltransacylation without having the kinetic complication of condensation of these substrates with the subsequent release of triacetic acid lactone from the synthetase.
Kinetic Studies of Transacylase ilctivities-The rate of acetyltransacylation as a function of acetyl-CoA concentration, with iodoacetamide-treated synthetase, is shown in Fig. 7. The K, of acetyl-CoA was found to be 42.8 ~,LM and the V,,, obt,ained was 1.4 (nanomoles of acetyltransacylated per min). As is also evident from Fig. 7, malonyl-CoA served as a competitive inhibitor with respect to acetyl-CoA in acetyltransacylation, having a Ki of 6.6 PM.
Very similar values for the kinetic parameters of the acetyltransacylase activity were obtained with synthetase which had not been initially treated with iodoacetamide.
The relationship between malongltransacylation and malonyl-<'oh concentration is shown in Fig. 8. The K, value obtained for malonyl-CoA was 13.1 PM, a value considerably lower than The Sephadex G-200 column (1.2 X 43 cm) was equilibrated in 0.1 M potassium phosphate (pH 7.5), containing 0.2 M NaCl. Fraction volume was 1.0 ml. Aliquots of the fractions were assayed for acetyl-and malonyltransacylase activity as described under "Experimental Procedure." the K, for acetyl-CoA in acetyltransacylation. The V,,,, obtained for malonyltransacylation was 0.6 (nanomoles of malonyltransacylated per min) as compared to the value of 1.4 obtained for acetyltransacylation.
Acetyl-CoA was also found to inhibit the malonyltransacylase activity competitively with respect to malonyl-CoA, and had a Ri value of 39.2 PM.
As in the case of acetyltransacylation, prior treatment of the synthetase with iodoacetamide had lit,tle effect on the kinetic parameters of malonyltransacylation.
Acetyl-and malonyltransacylation as a function of ACP concentration is shown in Fig. 9. The K, for ACP in acetgltransacylation was 66 PM while its I', in malonyltransacylation was 58 PM.
CoA was found to be a competitive inhibitor with respect to ACP for both transacylase activities, having a Kc of 22 PM in acetyltransacylation and 16 PM in malonyltransacylation.
Separation of Transacylase Activities from Xynthetase

Complex-
If the pigeon liver fatty acid synthetase was treated with 0.5 M guanidine hydrochloride, as described under "Experimental Procedure," and then chromatographed on DEAE-cellulose, the elution pattern shown in Fig. 10  bulk of the synthetase protein.
If the treatment with guanidine hydrochloride was omitted, a much more symmetrical activity versus protein pattern from the DEAE-cellulose column was obtained.
The elution pattern for the malonyltransacylase activity as compared to the protein pattern from the DEAEcellulose column was very similar to that obtained for the acetyltransacylase activity.
Chromatography of the DEAE-pool on a Sephadex G-200 column gave the elution pattern shown in Fig. 11. Most of the protein emerged from the column shortly after the void volume and was presumably a mixture of undissociated fatty acid synthetase and denatured or aggregated material. Two peaks of transacylase activity were obtained, each peak containing both acetyl-and malonyltransacylase activity. One peak of transacylase activity emerged from the column with the protein peak shortly after the void volume and was most likely the transacylase activity of undissociated synthetase. There was also a second peak of acetyl-and malonyltransacylase activity which emerged from the Sephadex G-200 column in the included volume.
Application of Squire's equation, which is used for the estimation of molecular weights of globular proteins from their elution from Sephadex gels (18), gave an approximate molecular weight of 62,000 for the included peak. The very low yields of this included transacylase activity have made it impossible for us to further characterize this material at this time.
The susceptibility of the acetyltransacylase activity in the DE,4E-pool to inhibition by N-ethylmaleimide is given in Table  III.
As compared to the acetyltransacylase activity of the fatty acid synthetase complex, the acetyltransacylase activity in the DEAE-pool was less sensitive to inhibition by N-ethylmaleimide.
The malonyltransacylase activity in the DEAE-pool was also found to be less sensitive to inhibition by N-ethylmaleimide than was the same activity when assayed in the untreated synthetase complex. DISCUSSION The pigeon liver fatty acid synthetase complex contains, in a tightly bound form, all of the enzymes necessary to catalyze the biosynthesis of long chain fatty acids. In the preceding paper (2), evidence is presented which establishes that the acetyl and malonyl moieties, transacylated from their CoA thioesters, bind to thiol and nonthiol acyl binding sites on the pigeon synthetase. Both the acetyl and malonyl moieties bind to the sulfhydryl group of the 4'-phosphopantetheine of the synthetase, in addition to binding at nonthiol sites. The nonthiol binding site has been tentatively identified as the hydroxyl group of a serine residue.
The nonthiol site appears to be the initial acceptor site on the synthetase for the acetyl and malonyl groups.
Prior treatment of the pigeon synthetase with 0.25 mM N-ethylmaleimide almost totally inhibits the thiol binding of acetyl and malonyl groups without affecting their binding at the nonthiol site (2). Furthermore, CoA was able to remove the acetyl and malonyl groups bound at the nonthiol site under these conditions. Consistent with these findings, 0.25 mM N-ethylmaleimide inhibited the acetyl-and malonyltransacylase activities of the pigeon synthetase, as measured with E. coli ACP, by only 30% (Table   III). It is reasonable to suggest, therefore, that the acetyl-and malonyltransacylase activities of the pigeon synthetase catalyze t,he transacylations of acetyl and malonyl moieties from their CoA thioesters to the 4'-phosphopantetheine of the synthetase complex through the nonthiol binding site. A similar scheme has been proposed for acyl transfers to the yeast fatty acid synthetase (19).
Although several of the findings obtained in this study raise the possibility that the transacylation of both acetyl and malonyl groups is being catalyzed by a single transacylase component of the synthetase complex, the existence of specific acetyl-and malonyltransacylases cannot at present be dismissed.  Fig. 9). The fact that both acetyl-and malonyltransacylase activities separate together from the complex could be the result of our having isolated a portion of the complex containing two specific transacylase components.
The pigeon synthetases having one transacylase for both acetyl and malonyl groups would be in marked contrast to the E. coZi fatty acid synthetase, which has clearly been shown to have distinct acetyl-and malonyltransacylases (10,20). Further clarification of this point for the pigeon synthetase-may come from additional studies on the transacylase activities which have been fractionated from the complex.
To carry out such studies, we are currently attempting to modify our present methods such that our yields of the fractionated transacylases will be significantly improved over what we are presently able to obtain.
The use of E. coli ACP to assay the acetyltransacylase activity of the pigeon synthetase appears to involve the following reaction sequence: Acetyl-CoA + E-OH = acetyl-O-E + CoA (1) Acetyl-O-E + ACP ti acetyl-ACP + E-OH ('4 where E-OH represents the nonthiol binding site (serine hydroxyl) of the pigeon synthetase.
The assay for malonyltransacylase activity presumably follows the same reaction sequence. That Reaction 1 is reversible is indicated by the ability of CoA to remove acetyl and malonyl groups bound both at the thiol and nonthiol sites of the synthetase (2). E. coli ACP is also able to promote this reverse reaction (Fig. 5). That the CoA and ACP might be using the same site in these reactions is suggested by the finding that CoA is a competitive inhibitor, with respect to ACP, of both the acetyl-and malonyltransacylase activities (Fig. 9).
The inhibition of the acetyl-and malonyltransacylase activities by N-ethylmaleimide (Table III) is possibly the result of the N-ethylmaleimide reacting with a sulfhydryl group not involved in either transacylation or binding and sterically preventing binding of acetyl or malonyl groups at the nonthiol site. The protection of transacylase activity against N-ethylmaleimide inhibition by either acetyl-CoA or malonyl-CoA (Table IV) Issue of June 10, 1970 C. A. Plate, V. C. Joshi, and S. J. Walcil cannot be explained by the acetyl or malonyl groups binding and protecting a sulfhydryl group that is required for transacylation. It has been shown that the only sulfhydryl group to which both of these moieties bind is the sulfhydryl of the 4'-phosphopantetheine (2, 21) and, as discussed above, this sulfhydryl group is not necessary for measuring transacylase activity.
The protection probably stems from the inability of N-ethylmaleimide to bind this extra sulfhydryl group when either acetyl or malonyl groups are bound at the nonthiol site. That a sulfhydryl group of the synthetase is not intimately involved in the assay of transacylase activity using E. coli ACP is further supported by the decreased sensitivity to N-ethylmaleimide of the transacylase activities when they have been fractionated from the synthetase complex (Table III).
This decrease in N-ethylmaleimide sensitivity of the fractionated transacylase activities could either be caused by their separation from this extraneous sulfhydryl group or to some slight conformational change which has moved this sulfhydryl from the immediate vicinity of the nonthiol site.