The Effect of Fasting on Synthesis and 4’Phosphopantetheine Exchange in Rat Liver Fatty Acid Synthetase”

SUMMARY The level of rat liver fatty acid synthetase is greatly decreased upon fasting. This report presents several additional observations of the effect of fasting on that enzyme.

The level of rat liver fatty acid synthetase is greatly decreased upon fasting.
This report presents several additional observations of the effect of fasting on that enzyme. (a) The rate of synthesis is reduced during 48 hours of fasting and is 20% of the steady state rate after 12 hours. (c) The exchange of the prosthetic group, 4'-phosphopantetheine, is largely eliminated, perhaps by effects on the mechanism of attachment or hydrolysis. (d) Feeding a protein-free diet mimics the effect of starvation on the enzyme levels.
In higher animals fatty acids are synthesized de nouo in the liver by a multienzyme complex, fatty acid synthetase.
XIuch of the knowledge about the mechxnism and control of activit,y of the complex has been derived from in vitro studies (l-10). Some features which may have an influence on the control of fatty acid synthesis carmot be studied except by in vi?10 esperimeutstion, since they occur as a result of complex cellular organization.
In particular, regulation of the level of enzyme and controls effecting the availability of prosthetic group 4'-phosphopantetheine are of interest.
A previous paper (II) has demonstrated that the fatty acid syrrthetase complex turns over with a half-life of about 3 to 4 days, whereas the covalently bound prost.lletic group seems to be exchanged very rapidly.
.I similar observation of prosthetic group mobility has been made in the Exherich.in coli fatty acid syrlt,hesizing system (12).
It has been known for some t,ime that an effect of fasting on rat liver is to decrease the measured activity of enzyme and correspondingly to decrease the liver content of fatty acids (13). This correlation seems to reinforce the position that the level of fatty acid syuthetase is important, iu the cellular control of fatty acid production. This rrport rs:unilles the mea,ns whereby thca loss of enzyme act,ivity is cffec~tcd by \~arioua diet, c+r:lngrs, :lnd describes an effect of fasting on t,hc? ljrost.lret,ic group r~ck~~~gc phenomenon. In all esperiments, animals had access to drillking water and did not avoid food when the diet was changed.
Liver srtpernatant holutions and fatty acid synt,hetase preparationa were prepared :lccording to the procedure of Ihu%on et al. (15) with the exception that the Sepha.dex G-100 gel filt,rat,ion step was omitted.
Ai further purification step was necessary in order to rid the prel)arations of contaminating 7 S species (16). This purification was achieved either by sucrose densit,y gradient ceutrifugation (with a SW41 or SW56 rotor) in 5K; t.o 20% sucrose prepared in 0.5 M potassium phosphate, pH 7.0, and 2 mM dith'othreitol, or by gel filtration on a Sepharose 613 column 50 cm long equilibrated with the above buffer. In each case the enzyme from the l)EAE purification (15) was activated for 30 min ut 35" prior t,o :rpplicat,ion (17). Tubes corresponding to the leading edge oC the peak and contaiuing enzyme act.ivity were 1)ooled.
Use of the Sepharose 61% c:olumn revealed a macromolecular species traveling with the void volume (exclusion limit 4 x 106) which exhibited \-ariablr acstivity (50 tJo 2.50 nrnoles per min per mg), This fraction, when preci1)itatetl with ammonium sulfate a11tl rechromatographed, reformed a small amount oE 12 S mat,e rial, which appeared in the elu:lt,c. These results suggest the eskknce of an aggregated species of fatty acid synthetase.
The conclusion of ('oIlins el nl. (I 6) that the contaminating 7 S protein species is tlot related to I'atty :toid synthetase as a breakdown 1)roduct is here cxonfirmrd. This species does not contain pantothenate, which tends to c~liminate a large portion of the fatt.y acid synthetase molecule.
?cIore ronvincingly, however, a pulse of labeled amino acids given to rats in various stages of fasting yields, aft.er sucrose gradient. sedimentation, a 7 S species which is never lower in specific radioactivity than the enzyme and which reaches a specific radioactivity three times that of the enzyme. In addition, the pattern of 7 S formation as a function of time bears no resemblance to that of enzyme.
These observations are not conskteut with a precursor product relationship between fat.ty acid synthetase :ttld 7 d protein.
All enzymes assays were pcrlormetl by observation of the malonyl-CoA dependent oxidation of 'L'PNFT (18). Maximal specific activities of about 1300 nnloles per min per mg were opcasionally observed, but the usu:d range observed was 1000 to 1200 nmoles per min per mg.
Protein determhhations were made by measuring the spectrophotometric 260:280 ratio (19). Characteristic ratios of 0.55 to 0.60 for fatty acid synthetase and 0.9 for the "aggregate" species mentioned above were noted.
CoA Determination-When the specific radioactivity of CoA was required, the supernatant fruct.ion from the second ammonium sulfate precipitation (15) was boiled t,o remove protein and made 80%, iu methanol to pt'ecipitate salt, followed by evaporation under a stream of tlitrogen.
The concentrated CoA was then isolated by use of a I)EAE column (20) and assayed by the phosphotransncetylase reaction (22). d~G?ao d&d Specific ilctivity-One milliliter of liver super-na~ant solution (15) was precipitated with 10% trichloroacetic acid and the resulting pellet. washed with 10% trichloroacetic acid. The pooled supernates were diluted with 40 ml of water and applied to a 5-ml column of Dowex 50-X8 which had previously been washed with 1 M NaOH, 1 M HCI, and finally water.
Affer loading, the column WAS elut,ed with 0.02 M HCI followed by 15 ml of 4 M NH&H.
The ammonia fraction was evaporated to dryness over concentrated H?S04 in a vacuum, washed wit,h 1 M KOH, and re-evaporated.
The resulting salt was taken up in 1 ml of 0.2 M citrate, pH 5, titrated with citric acid to pH 5, and used in the quantitative ninhydrin procedure of Cocking and Yemm (22). An aliquot of the sample was also counted and from t,hese data the specific act,ivit,y of the amino acid pool was determiued.
The above procedure yielded the same values for specific activity as the unchromatographed trichloroacetic acidtreated supernutant solution for periods after fasting up t,o 24 hours. After this time, however, the amount of labeled nonamino acid material became significant,.
The above procedure failed to retain glut,amic* and aspnrtic acid on the Dowes resin.
Wheu the total liver proteiu was determined, :i I-ml aliquot of the liver supernatant solutiou was precipitat,ed by making it 10cy; in perchloric acid, the pellet washed :md counted.
The protein rolltrllt was determilled in this c~sr by the biuret method (23).
CounGng-All radioactivity counting w:~ performed in a Nuclexr Chi(~ago i20 statics liquid scintill:ltio~l sprc$romrtr~ cmployiilg eitlrclr Aquas01 or :L t~oliiri~e-et~l~:~~~ol-Ou~~~ifl~~or mixture as :I sc'intillat~ior~ fluid. All protjriil was oountjc>d after caomplete solution in Protosol. All counting was performed t,o bet,ter tlr:nl &:Z(;d (confidence level of t;wo st,nndard deviations), and :rll samples mere quench correct,ed by interih:rl stalldardiz:a.t.iolt The control of enzyme levels in higher animals has been reviewed by Schimke and Doyle (24). The question of how fnsting effects a change in fatt,y acid synthetase levels can be collvelliently considered in two parts: (a) Is the half-life of the enzyme altered?
(b) Is the rate of synthesis altered? Since fatty acid synthetase levels drop by 80 5; to 90"Tc after 2 days of fastilrg (13), this is the time period of interest.
During this period of fasting animals :lre not in the steady state and therefore the rate of synthesis of the enzyme must be determined separat,ely from t,he rate of degradat,ion. The simplest a.pproach would be t.o measure loss of enzyme :lfter synthesis has been stopped.
This can be approximated if the determination is made after 12 hours. Fig. 1 shows :L curve fitted by the method of least squares to a series of determinntions of enzyme nrtivit'y per liver as :i fullrtion of time after removal of all food. The d:k:t represent det,ermiilatiolls from four separate esperiments and are expressed as perc~entage of zero time activity in the high speed supernat:mt solution.
The srat,ter is partly due to the difficulty of obtaining uniform supernatant preparations.
The slope of this line corresponds to a half-life of 20 hours.
The half-life in the normal steady state is about 70 hours (1 l), so it appears that fa&,ting :Iffectx the degradation of the enzyme. This result is not unespert.ed since a 70-hour half-life could not account for almost total loss of activit,y in 48 hours. The approximate fit of the data to a single curve suggests a rapid change to a new half-life as opposed to a slower incremental change. However, it is not. clear whether this loss of enzyme activity represents total degradation of the enzyme complex or rather a change to some inactive form. For example, the prosthetic group could be removed upon fa.sting a.nd later replaced on refeeding, a possibility to be discussed later. During the a-day fast period the animal weight decreased only slightly. The liver, however, after a brief lag period, rapidly lost weight to For information about the means of obtaining the various data see 'LExperimental Procedure." Rats were fed the vitamin B complex test diet complete.
X-X, liver wet weight, maximum 12 g; O-.-O, enzyme level, maximum 8900 units per liver; A-A, specific radioactivity of total soluble protein, maximum 1600 dpm per mg; l ---0, total soluble protein per liver, maximum 900 mg. attain a new relatively constant weight after 16 hours (Fig. 2). Some of this weight loss can be attributed either to degradation or to transport of soluble protein.
The weight losses, as well as the probable production of amino acids from degraded protein, make it difficult to be certain that the concentration of amino acids in the in vivo amino acid pool is constant throughout the esperiment.
Since the specific activity of injected label is affected by the dilution of the label by metabolic pools, it is necessary to assess the effect of dilution in each animal. This was done by either of two methods: isolation of the total free amino acid pool and measurement, of its apecifrc :rct.ivity as described under "Experimental Procedure," or me:rsurement of the specific activity of the trichloroacetic 3c,id-l)recil)it.~t.ted total soluble protein.
Both of these methods produced comparable results.
A second difficulty arose when the rate of iucorpomtion of a short pulse of labeled amino acids into enzyme was used bo measure the rate of synthesis.
The data of Fig. 1 were interpreted to indicate that the pool of enzyme was dropping rapidly. Therefore enzyme synthesized during the period of the pulse was diluted by a dwindling pool of unlabeled enzyme as the period of fasting increased.
A correction for this pool dilution effect was therefore necessary. These two corrections tend to cancel each other.
A third difficulty considered was that the labeled molecules present in the pool are subject to degradation.
Siuce iu the steady state it can be calculated that a quantity equal to approximately 1% of the pool of fatty acid synthetase molecules is being synthesized in any hour and since the rnte of syuthexis seems to drop much more rapidly than the pool size upon fasting, it seems reasonable to neglect this objection.
The corrected curve iu Fig. 3 was obtained from t,he points of the uncorrected curve by multiplying the value of en,ch point by a factor: protein specific activity (cpm/mg)(t = 0) protein specific activity (cpm/mg)(l = x) X enzyme units/liver (t = s) enzyme units/liver (t = 0) From the corrected curve it can be seen that within 6 hours after food was removed the rate of synthesis dropped a.nd continued to fall until the end of the experiment. The ability of the liver t.o respond to the absence of food after such a short period is striking. This may be due in part to the timing of the starvation which was begun in all cases between 8:00 and 9:00 a.m., an hour after daylight in the programmed light cycle. In a preliminary effort to determine what diet component is the ca,use of this change in rate, the experiment represented in Fig. 4 was performed.
This experiment was different from those previously described in that the equilibrating diet was not the vitamin B complex t'est diet complete, but rather the normal protein test diet. The change was necessary to achieve a better matching of the control and experimental diets. The values for the rate of synthesis in Fig.  4 were corrected as indicated above. The sudden drop in the rate of synthesis indicated by all of the curves in that figure possibIy reflects a difficulty in "changing over" metabolically to the requirements of a new diet. While the rate of synthesis resulting from change to a fat free diet began to rise after 12 hours, that resulting from the protein-free diet continued to decrease. This could be an indication that the protein component in the diet was responsible for the effect of fasting on enzyme level observed after 12 hours. Further similarity between those conditions is indicated by a second experiment described in Table I. There it can be seen that over a longer period of time the effect of the protein-free diet on enzyme levels mimics that observed for fasting.
In addition, feeding a fasted animal with protein-free food did not change the level of enzyme, but feeding both fasted and protein-deprived animals with fat-free food had a similar enhancing effect. Since fatty acid synthetase represents a sizable portion of the total soluble protein (0.8oj,. as determined from a comparison of the specific activity of crude supernatant solution with purified enzyme) of the steady state liver, it is expected that the enzyme may play an important part in the adjustment to a negative nitrogen balance.
The results of experiments which utilized a fat-free diet are also expected in view of the effect of that diet on fasted animal and the similar effect recorded for mouse liver (25). The results of Table I indicate that fasting prior to fat-free feeding produced a change in the level of enzyme additional to that expected from the simple diet change. Thus there are additional phenomena to be considered in analyzing the effect of refeeding a fat-free diet. Part of the effect could be due to reactivation of an inactive enzyme species (mentioned below) as well as the de nova synthesis already verified (26). The levels of enzyme activity found in the supernatant fraction from animals fed the vitamin B complex test diet com-  We now turn to a discussion of the effects of fasting on the pantothenate incorporation into fatty acid synthetase. In a previous publication (11) it was concluded that pantothenate compounds in the cytosol are rapidly exchanged with 4'.phosphopantetheine bound to the enzyme. From Fig. 5 it can be seen that for animals in the steady state, a rapid rise in specific radioactivity of enzyme was found as previously observed (II), and therefore it can be concluded that, under these dietal-conditions also, the exchange phenomenon is occurring.
In the c:ise of animals fasted for 12 hours l)rior to administration of isotope, the exchange of prosthetic group was not as readily discerned.
Even after 48 hours, the specific radioactivity of the protein failed to reach the levels found in the steady state. This is not due to a failure of the label to enter tile pantothenate pool since in all cases the measured specific radioactivity of CoA from the fasted series was higher than that observed for the 20-hour steady state point.
If COG% is a precursor of pautothenate in fatty acid synthet,ase it can be concluded that fasting interferes with the erchange process somewhere bet.ween Co.% and the protein, possibly directly inhibiting the enzyme responsible for attaching or removing the prosthetic group. This conclusion is supported by the results shown in Fig. 5. 1Sy making the assumption that the turnover of enzyme reported earlier (11) is similar to that observed here in the steady state, and by making use of the pool size of enzyme units and CoA specific activity per liver as well as the changes in rate of synthesis alld degradation described above, a rough calculation of the amount of pantothenate expected to be incorporated solely as a result of de 12000 synthesis of fatty acid synthetase can be made. This calculated result, when conlpared with the actual measured incorporat,ion of labeled l)atltothenate, :~llows arl estitmttc of itlc*orporation due to Axchange. The disparity betwerlr the steady stnt.e experimental and calculated curves indicatt~s the high degree of exchange mentioned above. However, after 12 hours fast,ing the experi-men&l curve lies relatively close t.o the calculated curve indicatitlg that nlost of the incorl)or;rtiou of label is due to de novo sytithesis.
Thin result is corisistellt with the conclusion that in fnst.ed animals a precursor of the prot.eilk bound 4'-phosphopantetheine is l;lbelrd, but that this precursor is not being exrhanged with the facility achieved in the steady state.
The pnnt.etheine exchange phenomenon might involve the ovel,-all cellular control of Co.\ metabolism, but at present this relation is obscure. The eschange might have an effect on fatty acid metabolism by co&rolling the presence of prosthetic group on the enzyme comples.
Thus, if the enzyme which synthesizes holoellzyme is inhibited during fasting then one could account for the observed loss of enzyme activity in terms of removal of prosthetic group by a hydrolase.
Such loss would not be replaced from the precaursor pool until food is restored.
A study of the loss of enzyme during fasting as measured by an anti-fat,ty acid synthetase ant body should help clarify this point.
Studies by llajerus (27,28) 011 rat acetyl-CoA carboxylase bear a resemblance to those reported here. Acet,yl-CoA carboxylase is the enzyme responsible for the synthesis of malonyl-Co;\ from acetyl-Co-1 and COs. Fast,ed rats show an increased rate of degradnt.ion and a decrease in the rate of synthesis of of this enzyme. Changes in t.he level of enzyme were shown to be c%:mges in the level of enzyme content in the liver rather that1 activation or inhibition of l)reviously formed protein. Animals fed diets deficient in biotin led to an accumulation of ape-ncetyl Co.1 carbosylase in the rat epididymal fat pad. Subsequent inject,ion of labeled biotic led to a rapid appearance of labeled enzyme.
The analogous experiment cannot be accom-plished utilizing pantothenate-deficient animals:. Such animals show a normal level of fatty arid synthetnse in the epididymal fat pad up until the time of death due t,o vitamin deficiency (29).