Separation of the Half -Molecular Weight Nonidentical Subunits of Pigeon Liver Fatty Acid Synthetase by Affinity Chromatography*

SUMMARY Two half-molecular weight subunits of 4’-phospho[14C]pan-tetheine-labeled pigeon liver fatty acid synthetase have been separated and shown to be nonidentical. An affinity chro-matographic column containing e-aminocaproylpantetheine bound to cyanogen bromide-activated Sepharose via the E-amino group was used to achieve this separation. The subunit containing 4’-phospho(14C]pantetheine and /3-keto-acyl thioester reductase activity is only slightly retarded on this column, whereas the subunit containing acetyl coenzyme A transacylase activity is strongly adsorbed. The latter subunit is eluted from the column with 0.1 M phosphate buffer at pH 10. Each subunit is then purified free of small amounts of the other subunit. This is accomplished by subjecting each to conditions which effect fatty acid synthetase forma-tion. Subsequent sucrose density gradient centrifugation then separates the pure half-molecular weight subunits containing transacylase or P-ketoacyl thioester reductase activity from the small amount of fatty acid synthetase complex. purified under appro-priate


SUMMARY
Two half-molecular weight subunits of 4'-phospho[14C]pantetheine-labeled pigeon liver fatty acid synthetase have been separated and shown to be nonidentical.
An affinity chromatographic column containing e-aminocaproylpantetheine bound to cyanogen bromide-activated Sepharose via the E-amino group was used to achieve this separation.
The subunit containing 4'-phospho(14C]pantetheine and /3-ketoacyl thioester reductase activity is only slightly retarded on this column, whereas the subunit containing acetyl coenzyme A transacylase activity is strongly adsorbed.
The latter subunit is eluted from the column with 0.1 M phosphate buffer at pH 10. Each subunit is then purified free of small amounts of the other subunit. This is accomplished by subjecting each to conditions which effect fatty acid synthetase formation.
Subsequent sucrose density gradient centrifugation then separates the pure half-molecular weight subunits containing transacylase or P-ketoacyl thioester reductase activity from the small amount of fatty acid synthetase complex. Recombination of the purified subunits under appropriate conditions yields enzymatically active fatty acid synthetase complex.
The pigeon liver fatty acid synthetase complex dissociates reversibly' into two subunits of approximately equal molecular weight (1,2). However, the separation of these subunits from one another had not been accomplished prior to this communication. Indeed, the similarity of the masses and the mass to charge ratios of the two half-molecular weight subunits is such that the question of their identity or nonidentity was unanswered until it was shown that only one phosphopantetheine group is present, per intact fatty acid synthet.ase molecule ( likely the subunits would have unequal affinity for an external analog of the internal substrate family of acyl 4'.phosphopantetheines.
If so, the separation of nonidentical subunits could then be proven by assay of each for the partial reactions of fatty acid synthesis.
Two partial reactions of fatty acid synthesis (Equations 1 and 2) were used in the present investigation to follow the separation of the half-molecular weight subunits of pigeon liver fatty acid synthetase. II 0 We chose as an affinity ligand in this investigation the substrate analog e-aminocaproylpantetheine bound to Sepharose by means of the amino group.
This ligand is analogous to the intermediate caproyl-4'.phosphopantetheine (4). Sepharose e-amino-n-caproic acid was prepared according to the method of Larsson and Xosbach (5) from cyanogen bromide-activated Sepharose (6). Crystalline pantetheine was obtained from Sigma Chemical Co., and it was reduced with sodium amaIgam to pantetheine prior to use. The pH of the reaction was kept, below 7 by the addition of Dowex 50-H+.
Sepharose e-aminocaproylpantetheine was prepared by the method of Cuatrccasas (7) but with the -SH compound in the solvent and the carbonyl compound in the stationary phase. An amount of ethyldi-methylamino~n-propylcarbodiimide sufficient to enable the binding of only 1.5 pmoles of pantetheine to 1 g of gel (wet weight) was used. This is less than the amount given by Cuatrecasas.
Larger amounts of the carbodiimide lead to excessive nonspecific (ionic) binding of fatty acid synthetase subunits. lJou1~1 pant.etheine was assayed by the Ellman method after hydrolyzing the washed Sepharose-e-aminocaproylpantetheine in 0.1 N KOH for 15 min at, room temperature.
The Sepharose e-aminocaproylpantethcinc was washed with 1 liter of dist.illed water and then with 150 ml of buffer (pH 8.7) cont,aining 0.1 M potassium phosphate, 0.1 hf Tris, and 0.0025 hf fi-mercaptoethanol or dithiothreitol. The final pII of the solution was 8.7. The Sepharose substrate analog was then packed in an ice-salt water-jacketed column (3 mm x 15 cm). ['4C]I'antetheit~e-labcled fatty acid synthetase (10 mg/0.5 ml) prepared as described previously (8) was dissociated by dialyzing 4 to 5 hours at 0" against 5 mM Tris, 35 rnhr glycine, 1 m&f EDTA, and 5 mM P-mercaptoethanol or dithiothreitol(1). The dialyzed, dissociated fatty acid synthetase was diluted with the same buffer to 2 ml and then loaded onto the Sepharose e-aminocaproylpantetheine column at O-1". Elution was carried out with at least 50 ml of a 0.1 M potassium phosphate-O.1 M Tris buffer (pH 8.5) containing 5 mM fl-mercaptoethanol at -1.0-O" and then at 25" with the same buffer adjusted to pH 10 with ammonia.
Loading and elution were carried out at a rate of 3 ml per hour and l-ml fractions were collected.
The fractions were monitored for ultraviolet light absorption at 280 nm. &Ketoacyl thioester reductase in each fraction was assayed spectrophotometrically, with N-acetyl-S-acetoacetyl cysteamine (Sigma) and TPNH as substrates by the method of Kumar et al. (9). Acetyl-CoA-pantetheine transacylase activity was measured radiochemically in a system consisting of 0.03 to 0.05 pg of protein, 3.2 nmoles of [Wlacetyl-CoA (4000 cpm per nmole), and 0.34 pmole of pantetheine in 0.1 ml of 0.2 M potassium phosphate buffer, pH 7.0, at 0". The reaction was carried out for 2 to 4 min and then stopped with 20 ~1 of 2 x acetic acid. The unchanged labeled starting material [14C]acetyl-CoA was removed from the product (acetylpantetheine) by passing the reaction mixture through a column (4 mm X 5 cm) of Dowes l-Cl-anion exchange resin in 0.2 N acetic acid. The product, 14C-labeled S-acetylpantetheine was washed from the column with 1.5 ml of 2 N acetic acid. One-half of the combined effluents and washes were added to 15 ml of dioxane-2,5-bis[2-(5.tertbutylbenzoxazolyl)]thiophene scintillation fhrid and then assays for radioactivity were carried out. by liquid scintillation spectrometry.
Protein and time of incubation for the transacylase reaction were adjusted so t,hat the quantity of product was linear with respect to each factor. W-Labeled pantetheine in the eluate fractions from the affinity column was determined by adding 200+1 aliquots to 15 ml of dioxane scintillation fluid and assaying for radioactivity.
The course of elution of labeled pantetheine is shown in Fig. 1. About 90% of the P-ketoacyl thioester reductase is eluted in fractions 20 to 35 with about 10 to 15y0 of the transacylase. The elution of radioactive pantetheine parallels that of reductase. Out of 6140 cpm recovered, 5299 cpm appeared in tubes 20 to 35. Fractions 60 to 70 which contained about 70% of the transacylase activity had only 750 cpm. Therefore, it is evident that reductase and transacylase activities are in separate subunits and that 4'.phosphopantetheine is associated with the subunit containing reductase activity.
In order to confirm the enzymatic nonidentity of the subunits and to complete their purification, the following procedure was used. The reductase Fractions 25 to 35 and the transacylase Fractions 62 to 70 were separately concentrated by means of a Diaflo PM-10 membrane filter. The concentrated rkductase fraction was then dialyzed 2 hours against 0.2 hf potassium phosphate buffer containing 5 mM P-mercaptoethanol at 25" and the transacylase fract.ion was dialyzed against 0.2 M potassium phosphate buffer containing 2 mM dithiot.hreitol. This procedure results in reassociation of unlike subunits to fatt,y acid synt.het.ase complex.
A t,hird dialysis was carried out with a mixture of equal amounts of reductase and transacylase protein. AS a control fatty acid synthetase complex was subjected to the same temperature (O-5") and time as the above reductase and transacylase fractions.
This enzyme complex was then dialyzed under the same conditions as used for the transacylase fraction.
Sucrose density gradient centrifugation was then carried out on each of the above fractions in a Spinco model L-350 ultra- centrifuge SW 27 rotor. Centrifugations were carried out at 58,000 x g for 44 hours at 4" in a 5 to 20% w/v sucrose gradient containing 0.1 M potassium phosphate buffer, pH 7.0, 1 rnbf &mercaptoethanol or dithiothreitol, and 1 to 2 mg of protein. Fatty acid synthetase, reductase, and transacylase fractions, and the reassociated mixture of each were loaded separately onto 38 ml of gradient.
After centrifugation the bottoms of the tubes were punctured and fractions were collect,ed dropwise. Thirty fractions were collected from each tube. The fractions were monit,ored for protein by ultraviolet light. absorption a.t 280 nm and for fatty acid synthctase, reductase, and transacylase activities The fractions, 400.~1 aliquots, containing W?labeled pantetheine and reductase activity were also monitored for radioactivity.
The profiles of fatty acid synthetase comples, reductasc, and transacylasc fractions, and comples reassociated from a mixture of reductase and transacylase obtained on sucrose density gradient centrifugation are shown in Fig. 2, A to D. The 9 S reductase peak (Fig. 2B) is free from transacylase activity and t,hc 9 S transacylase (Fig. 2C) peak is free from reductase activity.
The reassociation of the unlike subunits to complex (Fig. 2o) is virtually complete.
A significant amount. of fatt,y acid synthetase act,ivit,y, determined spectrophotomet.ricaIly as previously described (9)) was also obtained in this fraction on reassociation.
This is the first report of the separation of the half-molecular weight subunits of a fatty acid synthetasc complex from one another and the recombination of the separated halves with the recovery of fatty acid synthetase activity.
A number of factors are of critical importance in achieving the separations reported in this communication.
A low temperature is required to secure the complete dissociation of the fatty acid  (Fig. 1) were concentrated and dialyzed i n high ionic strength synthetase complex (9) and to prevent reassociation of the subunits.
Also, dialysis of the dissociated fatty acid synthetase must not be carried out for more than 12 hours, since some of the enzyme protein is converted to a form which cannot be eluted from the column.
Furthermore, the fatty acid synthetase must be frozen at -20" after purification on DEAE-cellulose. If dissociation is effected without freezing, the dissociated subunits are not eluted from the Sepharose substrate analog at 0". However, they are eluted at room temperature without separation. A slow flow rate on the affinity column is required to ensure binding of the transacylase-containing subunit at the temperature we use. Moderately high salt concentrations, along with a pH of 8.5 or higher, are required to dissociate the bound enzyme protein from the column.
At pH 8.5, only one-third of the transacylase activity is eluted from the column, whereas at pH 10 two-thirds or more of this fraction is eluted.
Pantetheine bound to e-aminocaproic acid is essential for the binding of the transacylase subunit.
Neither the binding of this subunit nor the separation of fatty acid synthetase subunits occur on the column when the Sepharose e-aminocaproic acid is allowed to react with ethyldiaminopropylcarbodiimide in the absence of pantetheine.
fl-Ketoacyl thioester reductase activity is relatively unstable. Half of the original enzyme activity is lost in the presence of dithiothreitol within 12 hours. Reductase activity can be restored, however, on dialysis for 3 hours in buffer containing freshly prepared dithiothreitol or fl-mercaptoethanol.
In the presence of P-mercaptoethanol the reductase fraction loses 10% of its activity in 24 hours. The transacylase-containing subunit is stable for 3 days in dithiothreitol, but loses 30% of its activity in 24 hours in /3-mercaptoethanol.
Further chemical and physical studies are being carried out on each subunit.