JOURNAL OF BACTERIOLOGY, Nov. 1977, p. 497-504Copyright C 1977 American Society for Microbiology

Vol. 132, No. 2Printed in U.S.A.

Membrane Lipid Biosynthesis in Acholeplasma laidlawii B:De Novo Biosynthesis of Saturated Fatty Acids by

Growing CellsYUJI SAITO,t JOHN R. SILVIUS, AND RONALD N. McELHANEY*

Department of Biochemistry, University of Alberta, Edmonton, Alberta, Canada T6G 2H7

Received for publication 19 April 1977

The de novo biosynthesis of fatty acids of 12 to 18 carbons from precursors of5 carbons or fewer has been demonstrated in Acholeplasma laidlawii B.Radiolabeling experiments indicated that the normal primers for the synthesisof the even- and odd-chain fatty acids are acetate and propionate or valerate,respectively. Saturated straight-chain monomethyl-branched fatty acids of upto five carbons were readily utilized as primers, whereas more highly branchedspecies and those possessing halogen substituents or unsaturation were notutilized. At primer concentrations of 1 to 3 mM, up to 80% of the total cellularlipid fatty acids were derived from exogenous primer. The mean chain lengthof the exogenous primer-derived fatty acids rose with increasing primer incor-poration for methyl-branched short-chain fatty acids but was invariant forpropionate. The products of de novo biosynthesis varied only slightly withtemperature or cholesterol supplementation, suggesting that de novo biosyn-thesis is not directly influenced by membrane fluidity. Cerulenin inhibited denovo biosynthesis in a fashion that suggests the presence of two j3-ketoacylthioester synthetases, which differ in substrate chain length specificity and insusceptibility to inhibition by the antibiotic.

The incorporation of exogenous fatty acidsinto the membrane lipids (which comprise es-sentially all of the total cellular lipids [12]) ofAcholeplasma laidlawii B is by now a well-documented phenomenon which has provenquite useful in studies of the functional role oflipids in the membrane of this organism (6,14). However, A. laidlawii B can also be cul-tured indefinitely in a medium containing onlytraces of long-chain fatty acids, in contrast toalmost all other members of the order Myco-plasmatales (20). Cells cultured under suchconditions contain mainly saturated fatty acidsof 12 to 18 carbons, indicating that A. laidlawiiB either can biosynthesize saturated fatty acidsfrom short-chain primers (of two to five car-bons) or can elongate medium-chain fatty acidswhich are unsuitable for direct utilization incomplex lipid biosynthesis. Pollack and Tour-tellotte (13) found that cells cultured in thepresence of [1-'4C]acetate incorporated thelabel into the saturated but not the unsaturatedfatty acids of cellular lipids; however, this ex-periment could not clearly discriminate betweentrue de novo fatty acid biosynthesis and theelongation of medium-chain fatty acids to long-

t Present address: Department of Biological Chemistry,Washington University School of Medicine, St. Louis, MO63110.

chain homologues. Rottem and Panos (15)showed that a cell-free extract from A. laidlawiiA could synthesize long-chain fatty acids(mainly palmitate, stearate, and some arachi-date) in the presence of malonyl-coenzyme A,reduced nicotinamide adenine dinucleotide phos-phate, adenosine 5-triphosphate, and mag-nesium ion. However, as the A and B strains ofA. laidlawii differ in their growth requirementsfor fatty acids (10, 11), it is uncertain whetherthe lipid-metabolic capabilities of the twostrains are entirely comparable. Although awealth of information is now available regard-ing the functional properties of fatty acids ascomponents of glycerolipids in the cell mem-brane ofA. laidlawii B (14), very little informa-tion is available regarding the normal metab-olism of endogenously derived fatty acids inthis organism. In an effort to further our under-standing of membrane lipid function andbiogenesis in A. laidlawii B, we have con-clusively demonstrated the existence of a denovo fatty acid biosynthetic pathway in thisorganism, and we have characterized several ofits important properties in vivo.

MATERIALS AND METHODS

Chemicals. Straight-chain fatty acids were ob-tained from Analabs (New Haven, Conn.) and were

497

on April 21, 2019 by guest

http://jb.asm.org/

Dow

nloaded from

498 SAITO, SILVIUS, McELHANEY

of >99% purity. Cholesterol (Ultrex grade, J. T.Baker Co.) was recrystallized once from ethanolbefore use. Methyl-branched volatile fatty acidswere obtained from Sigma (St. Louis, Mo.) or K &K Laboratories (Plainview, N.Y.); unsaturated andhalogenated short-chain aliphatic acids were ob-tained from K & K Laboratories and were usedwithout further purification. Cerulenin, a fungalantibiotic that inhibits de novo fatty acid synthesisin a number of organisms (8, 9, 21), was the gener-ous gift of Satoshi Omura (Kitasato Institute, To-kyo, Japan). '4C-labeled acetic acid (uniformly la-beled) and 1- 4C-labeled propionic, butyric, hexan-oic, and octanoic acids were obtained as the sodiumsalts in ethanolic solution from New England Nu-clear (Boston, Mass.). Bovine serum albumin (fattyacid poor, grade B) was obtained from Calbiochem(San Diego, Calif.) or Miles Laboratories (Elkhart,Ind.). Bio-Sil A, 200/325 mesh (Bio-Rad Labora-tories, Mississauga, Ontario), was washed withchloroform and methanol and then activated over-night at 120°C before use. Chloroform (J. T. Baker)was routinely redistilled before use. Other solventswere checked for purity by concentrating a substan-tial sample in vacuo and analyzing the residue bygas-liquid chromatography (GLC) for contaminationand were redistilled if any impurities were found.All common laboratory chemicals were of reagentgrade.

Culture conditions. Cells were grown staticallyto late log phase at 35'C, except where otherwiseindicated, in samples of 125 or 250 ml of an unde-fined but lipid-poor medium, as previously described(16).

Cell growth was monitored turbidimetrically ona Bausch and Lomb Spectronic-20 spectrophotome-ter, reading absorbance at 450 nm, which is propor-tional to cell titer during log phase but not duringstationary phase (7).

Lipid analysis. Lipid extraction, purification, andtransesterification, and the gas-chromatographicanalysis of methyl esters of lipid fatty acids, werecarried out exactly as previously described (16).Analysis of lipids excreted to the growth mediumwas carried out as described elsewhere (18). Radio-activity was monitored on-line by a Nuclear-Chi-cago model S190 radioactivity monitor system, withrate meter connected directly to the recorder usedfor the mass response (flame ionization detector)input from the gas chromatograph, permitting si-multaneous display of mass and '4C radioactivitytraces. The total radioactivity was quantified bytriangulation; the rate meter was calibrated wherenecessary by measuring the response to a sample ofknown activity.

Butyl esters of fatty acids were prepared by thefollowing method: to a dried lipid sample (0.2 to 2mg) in a test tube (15 by 150 mm) with a tight-fitting screw cap, 2 ml of n-butanol (redistilled) and2 drops of concentrated sulfuric acid were addedand the mixture was blended in a Vortex mixer. Thecapped tube was then heated to 80 C for 3 h,cooled, and shaken with 2 ml of water and 1 ml ofhexane. After phase separation, the lower phase(containing mainly water and acid) was discarded,

3.0 ml of distilled water and 3 ml of methanol wereadded, and the tube was shaken vigorously. Afterphase separation, the upper hexane layer was trans-ferred to a 12-ml conical tube, and the lower phasewas re-extracted with 1 ml of hexane. The combinedhexane extracts were then washed with 1 ml ofwater to remove traces of acid and most of theresidual n-butanol. The residual quantity of n-bu-tanol was generally less than 10 1l upon hexaneevaporation, and the extract could either be directlyanalyzed by GLC or entirely freed of butanol bylow-temperature thin-layer chromatography accord-ing to Varanasi and Malins (22). Tracer analysishas shown that recovery of fatty acids of six ormore carbons is greater than 96%, considerablybetter than recoveries of shorter-chain species withmethyl ester preparations (4).

RESULTSIncorporation of [4C]acetate into cellular

fatty acids. Our initial attempts to determinethe origin of the long-chain fatty acids of A.laidlawii B grown in lipid-extracted mediawere focused on a determination of the chainlength of the precursor(s) of the even-chainsaturated fatty acids. To this end, cells werecultured without bovine serum albumin in thepresence of [U-14C]acetate (40 ACi/liter) plussufficient unlabeled acetate to bring the totalconcentration to 1 mM, a level that had noeffect on the cellular fatty acid composition.The cellular lipids were extracted, and thefatty acids were analyzed for mass and radio-activity by GLC. Simultaneous mass and radio-activity traces for a typical analysis are shownin Fig. 1; it can be seen that as reported byPollack and Tourtellotte (13), only saturatedfatty acids are labeled. As the ratio of radioac-tivity to mass (radioactivity monitor/flame ion-ization detector) responses is directly propor-tional to the specific activity of a fatty acidspecies, a plot of this ratio versus carbon num-ber for all even-chain species found in a sampleof cellular lipids should extrapolate to give thecarbon number of the unlabeled precursor asthe x-intercept. A typical plot of the radioactiv-ity monitor/flame ionization response ratio("relative specific acitivity") versus chainlength is roughly linear, as would be expectedif an unlabeled precursor is elongated by ace-tate units of uniform specific activity, andyields an estimated precursor chain length verynear zero, corresponding to labeled acetate(Fig. 2). Stearic acid invariably showed a lowerspecific activity than the other three even-chain fatty acids, probably due to its low levelsin vivo, which made this species particularlysusceptible to isotopic dilution by traces ofunlabeled stearate in the medium. Whereasresidual traces of (unlabeled) long-chain fatty

J. BACTERIOL.

on April 21, 2019 by guest

http://jb.asm.org/

Dow

nloaded from

FATTY ACID BIOSYNTHESIS IN A. LAIDLAWII B 499

MASSIJ \JV\

RAD10- J_ACTIVITY 12 14 15 16 17 18

FIG. 1. Simultaneous traces of the mass and ra-dioactivity responses for a sample of fatty acidsisolated from cells cultured with [U-'4C]acetate (5,Ci) and separated by GLC. Fatty acid chain lengthsare indicated below the radioactivity trace. GLCconditions: column, 10% diethyleneglycol succinate;carrier gas, helium (flow rate, 60 ml/min); tempera-ture, 160 to 200°C (programmed, rising 2 CImin).

acids in the growth medium occasionally di-luted the activity of endogenously synthesizedspecies to a small extent, our results clearlyindicate that the labeled even-chain fatty acidsare derived from very short-chain primers(probably acetate).To determine the chain length of the precur-

sor(s) of the cellular odd-chain fatty acids, theabove analysis was applied to the odd-chainspecies (tri-, penta-, and heptadecanoic acids).Although the incomplete resolution of the odd-chain fatty acid peaks from the larger even-chain acid peaks (Fig. 1) hindered the accuratedetermination of the radioactivities of the for-mer, the radioactivities of the odd-chain specieswere determined with fair accuracy in severalexperiments. The results of five separate exper-iments indicate that either propionate or val-erate, or both, is the normal primer for odd-chain fatty acid biosynthesis; our results arenot consistent with the utilization of longer-chain species (e.g., heptanoate) as primers invivo.Incorporation of other unbranched

short-chain primers. As an alternative dem-onstration of de novo fatty acid biosynthesis inA. laidlawii B, we investigated the incorpora-tion of other short straight-chain fatty acidsinto the cellular lipids of cells cultured withno other fatty acid supplement. Cells werecultured in the presence of [1-14C]propionate,

CHAIN LENGTH

FIG. 2. Relative specific activity ofthe three majoreven-chain fatty acids of cells grown in the presenceof [U-'4C]acetate. Cells were harvested in late logphase, and the lipids were isolated and analyzed asdescribed in the text.

-butyrate, -caproate, or -caprylate (40 ,uCi/liter)plus unlabeled carrier fatty acid to make thefinal concentration up to 0.12 mM, or withunlabeled valeric or heptanoic acid at this sameconcentration. Fatty acid incorporation wasquantified in the former cases as the incorpo-ration of radiolabeled fatty acid of known spe-cific activity and in the latter cases as theenhancement of the level of odd-chain fattyacids in the total lipids. Gas chromatographicanalysis showed that only odd-chain fatty acidsare labeled when [14C]propionate is added tothe growth medium, whereas only even-chainfatty acids are labeled when [14C]butyrate isadded. This observation indicates that thesetwo fatty acids are incorporated directly intocellular fatty acids without prior degradationto [14C]acetate, in agreement with previousreports that this organism lacks a fatty acid /-oxidation acitivity (5). The distributions ofmass and radioactivity among the fatty acidspecies derived from a given primer are quitesimilar, as is demonstrated for propionate-grown cells in Table 1. This result suggeststhat the bulk of the cellular fatty acid is synthe-sized by the same system that incorporates[14C]propionate or -butyrate into long-chainfatty acids. In Fig. 3, the incorporation ofstraight-chain fatty acids of three to sevencarbons (quantified as the mole fraction of fattyacids derived from exogenous primer in thetotal cellular lipid) at a 0.12 mM concentrationis plotted versus primer chain length. Exoge-nous butyrate is incorporated most efficientlyby the de novo biosynthetic pathway; the sharprise in incorporation of fatty acids of seven ormore carbons is due to an additional fatty acidincorporation activity specific for longer-chainfatty acids (Y. Saito, J. R. Silvius, and R. N.McElhaney, J. Bacteriol., in press).Uptake of branched-chain primers and

VOL. 132, 1977

on April 21, 2019 by guest

http://jb.asm.org/

Dow

nloaded from

500 SAITO, SILVIUS, McELHANEY

TABLE 1. Comparison of percentage of mass andradioactivity of various homologous odd-chained

fatty acids in the presence of [C4Cpropionate

% of totalFatty acida

Mass Radioactivity

13:0 31.8 31.115:0 58.4 55.417:0 9.8 13.6

a Fatty acids are designatedcarbon atoms followed by thebonds present in the molecule.

40

Z 300

200

z

10

by the number ofnumber of double

3 4 5 6 7PRIMER CHAIN LENGTH

FIG. 3. Mole percent cellular fdtty acids deriuedfrom various short-chain n-alkyl fatty acids (percentincorporation) as a function of chain length at a

standard primer concentration of 0.12 mM. Detailsof analysis of cell lipids are giuen in the text.

primer analogues. To determine more thor-oughly the primer specificity of de novo biosyn-thesis, the incorporation of a number of othershort-chain fatty acids was determined. Theincorporation of methyl-branched fatty acidswas easily monitored by GLC without recourseto radiolabeling, since cells grown withoutthese primers have negligible levels ofbranched-chain fatty acids (6). The molar per-centages of branched-chain fatty acids in thetotal fatty acids of cells cultured with variousiso- or anteiso-branched fatty acids at a fixedconcentration (0.12 mM) are shown in Fig. 4.For both classes of branched-chain fatty acids,primer incorporation is high for the shortesthomologue but drops off sharply with increas-

ing chain length. The incorporation of any ofthese primers, even at high concentrations (-2mM), had no detectable effect on cell growthrates or yields, or on total fatty acid levels(data not shown).The fatty acid compositions of cells grown in

the presence of propionate, isobutyrate, isoval-erate, or anteisovalerate at high concentra-tions, where a large fraction of the cellulartotal fatty acid is derived from the exogenousprimer, are given in Table 2. These primers allgive rise to fatty acids of 12 to 18 carbons.Since we detect no significant excretion of bio-synthesized fatty acids, labeled with either

0 31

2

z

40 -

20 -

10 -

PRIMER CHAIN LENGTH

FIG. 4. Mole percent branched-chain fatty acids(percent incorporati.on) as a function of primer car-bon number for cells grown with iso (!7 and anteiso-branched (A) short-chain fatty acids at a standardconcentration of 0.12 mM.

TABLE 2. Molar percentages of the various fattyacids of cellular lipids of cells cultured in the

presence of high concentrations of exogenous short-chain fatty acids

mol% (exogenous primer concn)

Fatty acida3:0 4:0i 5:0i 5:Oai

(2 mMI (1 mMI 1 mMI (1 mM

12:0i -b 0.7 - -12:0 2.6 1.6 0.2 0.613:Oai - - - 4.213:0i - - 17.4 -13:0 14.0 0.2 - -14:0i - 32.1 - -14:0 13.7 11.8 13.1 6.215:0ai - - - 35.215:0i - - 48.3 -15:0 31.1 2.0 - -16:0i - 4.8 - -16:0 23.8 36.3 8.3 35.617:0ia - - - 4.017:0i - - 14.6 -17:0 6.8 0.5 - 2.518:0i - 0.1 - -18:0 2.3 2.9 1.8 5.218:1 4.2 5.6 5.2 5.518:2 1.3 1.3 1.1 0.9

a Fatty acids are designated by the number ofcarbon atoms followed by the number of doublebonds present in the molecule. The letters i and aindicate a methyl branch attached to the penulti-mate (iso-branched) and ante-penultimate (anteiso-branched) carbon atoms, respectively.b, <0. 1%.

['4C]acetate or [4C]propionate, to the growthmedium in the presence of acetate, propionate,or isovalerate, the fatty acid spectrum of thecellular lipids is truly representative of the denovo biosynthetic output in vivo. Exogenousprimers are incorporated into long-chain fattyacids without prior metabolic alteration of thecarbon chain; isobutyrate and isovalerate, forexample, yield only even- or odd-chain iso-branched fatty acids, respectively.As both straight- and branched-chain fatty

acids can be utilized as primers, we tested theability of the de novo biosynthetic system to

J. BACTERIOL.

l"l/8

on April 21, 2019 by guest

http://jb.asm.org/

Dow

nloaded from

FATTY ACID BIOSYNTHESIS IN A. LAIDLAWII B 501

incorporate a number of primer analogues con-taining more extensive chain branching (3,3-dimethyl- and 2-ethylbutyric acids), halogensubstituents (3-bromopropanoic or 3-butyricand 3-chloropropanoic acids), or unsaturation(acrylic, cis- and trans-crotonic, and /-methyl-crotonic acids). The incorporation of any ofthese species could be readily detected as achange in the normal cellular fatty acid spec-trum determined by GLC. No incorporation ofany of the species listed above was observed,even at high concentrations (-2 mM).

Variation of the end products of de novobiosynthesis with the concentration of var-ious primers. To determine the concentrationdependence of primer incorporation, cells werecultured in the presence of varying concentra-tions of propionate, isobutyrate, isovalerate, oranteisovalerate, and the ratio of exogenousprimer-derived to acetate-derived fatty acids inthe cellular lipid was determined by GLC. Thisratio rises linearly with primer concentrationfor all primers tested, as is shown in Fig. 5 forpropionate. This result suggests that the exog-enous primers compete directly with acetatefor utilization as de novo biosynthetic primers.In the presence of 1 mM propionate, isobutyr-ate, isovalerate, or anteisovalerate, 39, 57, 80,and 59%, respectively, of the total cellular fattyacids are derived from the exogenous primer.The cellular lipids can thus be highly enrichedin branched-chain fatty acids by simply cultur-ing the organism in the presence of methyl-branched primers, a finding that may proveuseful in studies of the effects of branched-chain fatty acids on membrane function.

Since certain primers were incorporated tohigh levels into the fatty acids of membranelipids, we investigated the changes in the spec-trum of fatty acids derived from a given primerin the cellular lipids as primer incorporationincreased. The most convenient quantitativemeasure of primer elongation is the molar ratio

2.0-

1.0 -

DZ0>X0 2 3

PROPIONATE ADDED. mM

FIG. 5. Ratio ofodd (propionate-derivedj- to even-(acetate-derived)-chain fatty acids as a function ofconcentration of exogenously supplied propionate.Cellular fatty acids were analyzed as described inthe text.

of the two major fatty acids derived from theprimer in vivo. This ratio, symbolized by (C7, + 2/C,,), represents the molar ratios (pentadeca-noate/tridecanoate), (isopalmitate/isomyris-tate), (isopentadecanoate/isotridecanoate), and(anteisopentadecanoate/anteisotridecanoate)when propionate, isobutyrate, isovalerate,and anteisovalerate, respectively, are usedas primers. In Fig. 6, the ratios (C,, + 2/C,n) foreach primer at various primer concentrations,divided by the maximal observed (C,, + 2/C,,)value for each species (to adjust all data to acommon scale), are plotted versus the molarpercentage of primer-derived fatty acids in thetotal cellular fatty acid ("percent incorpora-tion"). Increasing incorporation of branched-chain primers increases the extent of primerelongation, especially at high levels of incorpo-ration. In contrast, the product spectrum ofpropionate does not vary with primer incorpo-ration. Sodium acetate at concentrations up to5 mM had no effect on cellular fatty acidcomposition; since acetate is a normal catabo-lite in this organism (20), exogenous acetatemay have relatively little effect on the size ofits intracellular acetate pool.

Effects of temperature and cholesterol onde novo biosynthesis. Since it has been re-ported that the product spectrum of de novobiosynthesis of membrane lipid fatty acids isdependent on the phase state of the membranebilayer in certain procaryotes (1, 19), we inves-tigated this possibility in A. laidlawii B. Whenthe cells were cultured at various temperaturesfrom 22 to 40°C in the presence of isobutyricacid (0.08 mM) or isovaleric acid (0.25 mM), nosystematic effect of temperature on primer in-corporation or elongation was observed. When

LI10 20 30 0 50

o~~~~~

1.0

0.8

(Cn+2)Cn

(Cn+2)FCn Ia0.D6

Q2

0 10 20 30 40 50 60 70 80 90INCORPORATION, %

FIG. 6. Ratio of the levels of the two major fattyacid species (C,IC,, + J derived from a short-chainprimer as a function of primer concentration. Sym-bols: (O) propionate primer, 15:0/13:0 ratio; (A)anteisovalerate primer, 15:0ai1l3:0ai ratio; (0) iso-butyrate primer, 16:0i114:0i ratio; (x) isovalerateprimer, 15:0i113:Oi ratio. All ratios are scaled asfractions of the maximal (Cn + Cn + 2) ratio measuredfor each primer to allow direct graphical comparisonof all data.

VOL. 132, 1977

on April 21, 2019 by guest

http://jb.asm.org/

Dow

nloaded from

502 SAITO, SILVIUS, McELHANEY

cells were cultured in the presence of choles-terol (0 to 25 ,ug/ml) at various temperatures,some small variations in membrane fatty acidcomposition with temperature and cholesterollevel were noted, but these were not of a natureand magnitude to justify the conclusion thatde novo biosynthesis is influenced directly bythe membrane phase state.

Inhibition of fatty acid de novo synthesisby cerulenin. Cerulenin, an antibiotic synthe-sized by the fungus Cephalosporum cerulinin,is a potent and apparently noncompetitive in-hibitor of fatty acid synthesis in a number oforganisms, apparently acting on the ,B-ketoacylfatty acid synthetase or "condensing enzyme"(2). We therefore investigated the effect of thisantibiotic on de novo biosynthesis of fatty acidsin A. laidlawii B. Preliminary experimentssuggested that de novo biosynthesis was in-hibited by cerulenin. At a concentration of 10mg/liter, cerulenin strongly inhibited cellgrowth, and this inhibition was relieved by0.12 mM elaidate; cultures grown with ceru-lenin and an exogenous fatty acid incorporatedmore of the exogenous acid than did culturesgrown without cerulenin. To further character-ize the effects of the antibiotic, cells were cul-tured with 0.02 mM elaidate (which main-tained cell viability in the presence of ceru-lenin) at various cerulenin concentrations, with[1-'4C]propionate or -butyrate (5 .tCi) added tolabel de novo-biosynthesized fatty acids. InTables 3 and 4, the distribution of radioactivity(percentages of total counts) among the various

TABLE 3. Distribution of radioactivity amongcellular fatty acids derived from ['4Clpropionate atvarious leuels of cerulenin in the culture medium

% Total fatty acid radioactivity at cerulenin

Fatty acid level of:0 0.5 mg/ml 1 mg/ml 2 mg/ml

11:0 1.6 3.1 5.0 6.013:0 26.4 46.3 61.1 64.815:0 52.6 45.3 33.9 29.217:0 19.4 5.3 <0.3 <0.3

TABLE 4. Distribution of radioactivity amongcellular fatty acids derived from ["'Cbutyrate atvarious levels of cerulenin in the growth medium

% Total fatty acid radioactivity at cerulenin

Fatty acid level of:

0 0.5 mg/ml 1 mg/ml 2 mg/ml

10:0 <0.2 0.2 0.3 0.712:0 7.3 17.3 24.0 30.814:0 29.0 44.4 52.4 51.816:0 59.1 37.8 23.2 16.718:0 4.6 <0.3 <0.3 <0.3

labeled long-chain fatty acids of the cellularlipids is tabulated as a function of ceruleninconcentration. Cerulenin sharply reduces theaverage chain length of the de novo fatty acidoutput at concentrations well below the mini-mal concentration for full growth inhibition.In addition, elaidate incorporation is virtuallyconstant over the cerulenin concentrationrange examined (0 to 2 mg/liter), suggestingthat total de novo output is not reduced atlevels sufficient to sharply diminish the extentof elongation of de novo biosynthetic products.Analysis of butyl esters derived from the cellu-lar lipids showed that no fatty acids of lessthan 10 carbons appear as the mean de novooutput chain length is diminished. It seemslikely that the inhibition of total de novo fattyacid biosynthetic output at high cerulenin lev-els is an effect distinct from the reduction ofthe average de novo output chain length atlower concentrations. This point is further de-veloped in the Discussion.

DISCUSSIONThe findings reported in this paper indicate

that A. laidlawii B can synthesize saturatedfatty acids from primers of two to five carbons;the results of our studies on ["4C]acetate incor-poration indicate that the bulk of the fattyacids of cells grown without exogenous fattyacid supplementation are de novo biosyntheticproducts. The de novo biosynthetic system pro-duces fatty acids of 12 to 18 carbons from allprimers tested; the bulk of the de novo output(>70%) is usually composed of one or two majorspecies of between 13 and 16 carbons. Short-chain fatty acids are apparently neither de-graded nor isomerized before utilization as denovo biosynthetic primers.The system responsible for the incorporation

of short-chain fatty acids is specific for straight-chain or monomethyl-branched saturated fattyacids of five carbons or fewer. Halogenated andunsaturated short-chain fatty acids are notincorporated into long-chain fatty acids, sug-gesting that the utilization of a primer is deter-mined by its electron distribution as well as byits steric properties. The unsuitability of theneo- and diethyl-branched species for utiliza-tion as primers is particularly fortuitous, forthe extensive branching of the long-chain fattyacids derived from them would tend to drasti-cally reduce the strength of van der Waals'interactions in the terminal regions of the fattyacyl chains of membrane lipids (17), therebygreatly increasing membrane fluidity and pos-sibly resulting in the impairment of membranefunctions (2, 3). The primer specificity of the

J. BACTERIOL.

on April 21, 2019 by guest

http://jb.asm.org/

Dow

nloaded from

FATTY ACID BIOSYNTHESIS IN A. LAIDLAWII B 503

de novo biosynthetic pathway could conceivablyoriginate at the level of transport into the cell(via a specific membrane permease), the thioes-terification of the free fatty acid with acylcarrier protein, or the actual sequence of chainelongation reactions. The first alternative isunlikely, since certain species (e.g., anteisono-nanoate) that are unsuitable as primers arequite hydrophobic and should enter the cellsrapidly by passive permeation. The third possi-bility is likewise unlikely, as the thioesters ofcertain species that cannot function as primers(e.g., anteisoheptanoate or -nonanoate) arepresumably intermediates in primer elonga-tion. Therefore, primer selectivity probablyoriginates at the level of the "activating en-zyme," which condenses the free fatty acid witha reactive thiol group of coenzyme A or acylcarrier protein before chain elongation.The very high observed molar percentages of

membrane lipid fatty acids derived from exog-enous primers at high primer concentrationsand the linear relationship found between ex-ogenous primer concentration and the molarratio of exogenous primer-derived to acetate-derived fatty acids in the cellular lipids suggestthat exogenous primers compete directly withthe primers normally utilized by cells grownon unsupplemented media (i.e., acetate andlesser amounts of propionate or valerate) andare not metabolized by a parallel but distinctpathway. It is worthy of note that the averagechain length of fatty acids derived from methyl-branched primers rises with increasing primerincorporation, whereas that of propionate-de-rived fatty acids does not. Studies on the rela-tionship of growth rate to membrane fatty acidcomposition and physical state (6; unpublishedobservations) have suggested that the optimalchain lengths of straight-chain and iso- andanteiso-branched fatty acids for supporting cellgrowth are roughly 15, 16, and 17 carbons,respectively. Cells cultured with either low orhigh levels of propionate show an averagechain length of odd-chain fatty acids near theideal 15 carbons, but the average chain lengthsof branched-chain fatty acids in cells grownwith low concentrations of branched-chainprimers fall well below the optimal values forsustaining growth. It is therefore of considera-ble interest that the average chain lengths ofbranched fatty acids approach the optimal val-ues more closely as branched-chain primer in-corporation rises and the branched-chain acidsthus become more important in determiningthe physical state of the membrane; this obser-vation suggests that a rudimentary mechanismfor membrane fluidity control may operate in

this organism. The relatively minor effects oftemperature and cholesterol on de novo biosyn-thesis indicate that such a control mechanismcannot directly monitor the phase state of thebilayer but rests entirely on metabolic factorsand intrinsic enzyme specificities. We havecharacterized a control mechanism of this typethat appears to act primarily at the level offatty acid incorporation into complex lipids andcan influence the elongation of both de novobiosynthetic primers and longer-chain exoge-nous fatty acids; details of this process havebeen described elsewhere (18).The inhibition of fatty acid biosynthesis by

the antibiotic cerulenin strongly suggests thatthe pathway for biosynthesis of saturated fattyacids in A. laidlawii B is basically similar tothat in numerous other systems studied (8, 9,21). Although the average chain length of themembrane fatty acids steadily decreases withincreasing cerulenin concentration, no appre-ciable quantities of C, or shorter-chain fattyacids are detected in cells cultured with radio-labeled butyrate or propionate, even when thecellular fatty acids are recovered as the lessvolatile butyl esters. This finding suggests thatde novo synthesis may proceed in two stages,the first yielding fatty acids of intermediatechain length (CI) to C13) and the second elon-gating these species further to generate thefatty acids normally found in the cellular lipids.If the ,3-ketoacyl thioester synthetase compo-nent of the latter system were appreciablymore sensitive to cerulenin than that of theformer system, one would predict that at lowcerulenin concentrations, the mean chainlength, but not the total production, of de novo-biosynthesized fatty acids would decrease. Highcerulenin levels would also inhibit the [3-keto-acyl thioester synthetase responsible for the ini-tial elongation of primers, reducing the level ofde novo biosynthesis and inhibiting cell growthin the absence of exogenous long-chain fattyacids. These predictions agree completely withour experimental observations. It is temptingto suggest that the elongation of exogenousmedium-chain fatty acids (Saito et al., in press)is catalyzed by the component of the de novobiosynthetic system that is responible for theterminal elongation cycles of this pathway.

ACKNOWLEDGMENTSWe thank Stephen H. J. Cook for his excellent technical

assistance.This work was supported by research grant MT-4261

from the Medical Research Council of Canada.

LITERATURE CITED

1. Cronan, J. E., Jr., and E. P. Gelmann. 1975. Physical

VOL. 132, 1977

on April 21, 2019 by guest

http://jb.asm.org/

Dow

nloaded from

504 SAITO, SILVIUS, McELHANEY

properties of membrane lipids: biological relevanceand regulation. Bacteriol. Rev. 39:232-256.

2. D'Agnolo, G., I. S. Rosenfield, J. Awaya, S. Omura,and P. R. Vagelos. 1973. Inhibition of fatty acidsynthesis by the antibiotic cerulenin. Biochim. Bio-phys. Acta 326:155-166.

3. Davis, M.-T., and D. Silbert. 1974. Changes in cellpermeability following a marked reduction of satu-rated fatty acid content of Escherichia coli K12.Biochim. Biophys. Acta 373:224-241.

4. Gander, G. W., R. G. Jensen, and J. Sampugna. 1962.Analysis of milk fatty acids by gas-liquid chromatog-raphy. J. Dairy Sci. 45:323-328.

5. Lynn, R. J. 1960. Oxidative metabolism of pleuropneu-monia-like organisms. Ann. N.Y. Acad. Sci. 79:538-542.

6. McElhaney, R. N. 1974. The effect of alterations in thephysical state of the membrane lipids on the abilityof Acholeplasma laidlawii B to grow at various tem-peratures. J. Mol. Biol. 84:145-157.

7. Maniloff, J. 1969. Use of blood agar plates to studyMycoplasma physiology. Microbios 2:125-135.

8. Nomura, S., T. Horiuchi. S. Omura, and T. Hata.1972. The action mechanism of cerulenin. I. Effect ofcerulenin on sterol and fatty acid biosynthesis inyeast. J. Biochem. (Tokyo) 71:783-796.

9. Omura, S. 1976. The antibiotic cerulenin, a novel toolfor biochemistry as an inhibitor of fatty acid synthe-sis. Bacteriol. Rev. 40:681-697.

10. Panos, C., and 0. Leon. 1974. Replacement of theoctadecanoic acid growth-requirement for Achole-plasma laidlawii A by cis-9, 10-methylene-hexadeca-noic acid, a cyclopropane fatty acid. J. Gen. Micro-biol. 80:93-100.

11. Panos, C., and S. Rottem. 1970. Incorporation andelongation of fatty acid isomers by Mycoplasma laid-lawii A. Biochemistry 9:407-412.

12. Pollack, J. D., S. Razin, M. E. Pollack, and R. C.Cleverdon. 1965. Fractionation of Mycoplasma cellsfor enzyme localization. Life Sci. 4:973-984.

13. Pollack, J. D., and M. E. Tourtellotte. 1967. Synthesisof saturated long chain fatty acids from sodium ace-tate-1 -'-C by Mycoplasma. J. Bacteriol. 93:636-641.

14. Razin, S. 1975. The Mycoplasma membrane, p. 257-312. In D. A. Cadenhead, J. F. Danielli, and M. D.Rosenberg (ed.), Progress in surface and membranescience, vol. 9. Academic Press Inc., New York.

15. Rottem, S., and C. Panos. 1970. The synthesis of long-chain fatty acids by a cell-free system from Myco-plasma laidlawii A. Biochemistry 9:57-63.

16. Saito, Y., and R. N. McElhaney. 1977. Membrane lipidbiosynthesis in Acholeplasma laidlawii B: incorpora-tion of exogenous fatty acids into membrane glyco-and phospholipids by growing cells. J. Bacteriol.132:485-496.

17. Salem, L. 1962. The role of long-range forces in thecohesion of lipoproteins. Can. J. Biochem. Physiol.40:1287-1298.

18. Silvius, J. R., Y. Saito, and R. N. McElhaney. 1977.Membrane lipid biosynthesis in Acholeplasma laid-lawii B. Investigations into the in Lilo regulation ofthe quantity and hydrocarbon chain lengths of denoLo biosynthesized fatty acids in response to exoge-nously supplied fatty acids. Arch. Biochem. Biophys.182:455-464.

19. Sinensky, M. 1974. Homeoviscous adaptation-a ho-meostatic process that regulates the viscosity of mem-brane lipids in Escherichia coli. Proc. Natl. Acad.Sci. U.S.A. 71:522-525.

20. Smith, P. F. 1971. The biology of Mycoplasmas. Aca-demic Press Inc., New York.

21. Vance, D., I. Goldberg, 0. Mitsuhashi, K. Block, S.Omura, and S. Nomura. 1972. Inhibition of fattyacid synthetases by the antibiotic cerulenin. Bio-chem. Biophys. Res. Commun. 48:649-656.

22. Varanasi, V., and D. C. Malins. 1970. Ester and ether-linked lipids in the mandibular canal of a porpoise(Phocoena phocoena>. Occurrence of isovaleric acid inglycerolipids. Biochemistry 9:4576-4579.

J. BACTERIOL.

on April 21, 2019 by guest

http://jb.asm.org/

Dow

nloaded from