JOURNAL OF BACTERIOLOGY, Sept. 1971, p. 806-814Copyright ( 1971 American Society for Microbiology

Vol. 107, No. 3Printed in U.S.A.

Microbial Assimilation of Hydrocarbons:Phospholipid Metabolism

R. A. MAKULA AND W. R. FINNERTY

Department of Microbiology, University of Georgia, Athens, Georgia 30601

Received for publication 29 March 1971

An analysis of the turnover of the major phospholipids of Micrococcus cerificansgrowing or nongrowing cultures. The turnover rates of "4C-PE and "4C-PE were

61.5% of the total phospholipid, exhibited no significant rate of turnover in eithergrowing or nongrowing cultures. The turnover rates of PE-"4C and PE-32P were3.2% per hr and 1.2% per hr, respectively. Phosphatidylglycerol (PG) exhibited aturnover rate of 11% and 7.7% per hr for "4C and 32p, respectively, indicating anextremely slow metabolism. PG metabolism was examined in greater detail, andthe data indicated a preferential 75% incorporation of glycerol-1,3-"4C into theunacylated portion of the PG molecule. The turnover of cardiolipin (CL) was ex-tremely slow in growing cells whereas nongrowing cells exhibited a 30% and 36%increase per hr for "4C-CL and "4C-CL, respectively. Glycerol-1,3-"4C was notconverted to phospholipid fatty acid carbon; all radioactivity appeared only in thewater-soluble backbone of the phospholipids. The kinetics of assimilation of hexa-decane-J-"4C into cellular lipids is presented. Radioactivity in neutral lipid increasedapproximately sevenfold over the growth cycle, whereas radioactivity in phospho-lipid increased 50-fold during the same time period. The incorporation of radioac-tive fatty acids derived from the direct oxidation of hexadecane-I-"4C demon-strated differential kinetics of assimilation into PE, PG, and CL. The results indi-cated a rapid turnover of phospholipid fatty acids in M. ceriflcans growing at theexpense of hexadecane.

Little information is available concerningphospholipid metabolism in those microorga-nisms capable of hydrocarbon assimilation. Aprevious report characterized the phospholipidcomposition of Micrococcus cerificans grown atthe expense of long-chain n-alkanes and de-scribed the tentative biosynthetic pathway forthese phospholipids (II).

In current membrane models, lipid forms anintegral part of a protein-lipid network for whicha fundamental structural role is attributed tophospholipids (10). Natural membranes havebeen observed to be metabolically dynamic enti-ties with the phospholipids contributing to thedynamic character of such processes as activetransport of cations (1, 5, 14) and mitochondrialfunction (3, 4, 7). Phospholipids as metabolicallyactive constituents of membranes have beenstudied with respect to turnover of phospholipids(18), polar end-group conversions (8, 12, 17),and phospholipid fatty acyltransferase activities(6, 17).Most proponents believe that interactions of

lipids and proteins within membranes give thesebiointerphases their characteristic and vital prop-

erties. These lipid-lipid and lipid-protein interac-tions provide a hydrophobic environment for cer-tain processes that would not occur under morehydrophilic conditions. The role of lipid metabo-lism, therefore, becomes particularly germane tohydrocarbon assimilation. A relationship be-tween phospholipid biosynthesis and hydro-carbon oxidation has already been suggested (I 1).Two basic questions proposed at the onset ofthis work were as follows: (i) Do membranephospholipids exhibit specific turnover propertiesthat correlate with alkane oxidation? (ii) Whatrelationships exist between cellular lipid compo-sition and hydrocarbon assimilatory processes?These studies detail answers to certain aspects ofthe problem in addition to implicating heretoforeunrecognized relationships.

MATERIALS AND METHODSA description of culture conditions, extraction of lip-

ids, thin-layer chromatography, and silicic acid columnchromatography has been reported (I1).

Recovery of phospbolipids. Phospholipids were re-covered from silica gel thin-layer plates after one-di-mensional development in chloroform-methanol-glacial

806

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

11

Janu

ary

2022

by

119.

128.

112.

126.

PHOSPHOLIPID METABOLISM IN M. CERIFICANS

acetic acid (65:25:8, v/v/v). The silica gel area con-taining each phospholipid was removed and succes-sively extracted with five volumes of chloroform-meth-anol (1:1, v/v) and five volumes of methanol, eachsolvent system containing 2% water. The solutions werefiltered through glass wool and partitioned with 0.3%NaCl, and the sample was reduced to dryness undernitrogen. Purified phospholipids were dissolved in chlo-roform for subsequent analyses.

Acid hydrolysis of lipids and paper chromatographyof water-soluble products. Phosphatidylethanolamine(PE) was hydrolyzed for 12 hr in 3 N HCI at 100 C insealed glass ampoules. The water-soluble products,ethanolamine-hydrochloride and a-glycerophosphate,were obtained by evaporation of the aqueous phaseafter ether extraction. Ascending paper chromatog-raphy on Whatman no. I filter paper employingethanol-acetic acid-water (90:5:5, v/v/v) as the sol-vent system was used to separate these hydrolysis prod-ucts. Co-chromatography with known standards estab-lished chromatographic identity. Ethanolamine-hydro-chloride was revealed with 0.2% ninhydrin in acetone,and a-glycerophosphate was revealed by the phosphatereagent of Wade and Morgan (as described in refer-ence I1).

Deacylation of phospbolipids. Mild alkaline methan-olysis was performed by a modification of the proce-dure of White (16). The phospholipid was dissolved inI ml of anhydrous methanol-toluene (1:1, v/v), and Iml of freshly prepared 0.2 M methanolic KOH wasadded. The sample was held for 2 hr at 4 C, afterwhich the solvents were removed by nitrogen evapora-tion. The sample was suspended in acidified distilledwater (15% HCI) and exhaustively extracted withether.

Materials. Carrier-free 32P-orthophosphate and glyc-erol-1,3-14C (specific activity, 12.5 mCi/mmole) werepurchased from New England Nuclear Corp. N-hex-adecane-1-14C (specific activity, 47.2 mCi/mmole)was purchased from Amersham/Searle Co. and dilutedwith carrier hexadecane to specific activity 24.71iCi/mmole. All experiments with hexadecane-1- 14Cused 2.05 mmoles containing 3 x 107 total counts permin.

RESULTSPbospholipid turnover-growing cells. The de-

crease of radioactivity from phospholipids after aperiod of growth in 14C and 32P offers a measureof phospholipid metabolism during bacterialgrowth. M. cerificans was grown for 2 hr in thepresence of I mCi of H332P04 and 25 ACi ofglycerol-l,3-14C with hexadecane as the solecarbon and energy source. The cells were har-vested by centrifugation, washed with basalgrowth medium minus requisite salt amendments(MgSO4, CaCl2, FeSO4), and suspended incomplete growth medium with hexadecane as thecarbon source. Samples (60 ml) were removed attime zero and at hourly intervals for 5 hr, theywere extracted with chloroform-methanol at aratio of 2:1:0.2 (v/v/v) yielding a monophasic

system, and the phospholipids were separated aspreviously described (11). Phosphorus analyseson the crude lipid samples verified that phospho-lipid synthesis occurred during growth (Fig. IA).The turnover of crude lipid radioactivity (Fig.

IB) indicated a slow metabolism of the cellularphospholipids. The turnover of "4C-glycerol isfaster than the turnover of 32P (30% for 14C and20% for 32p) over the 5-hr incubation. Althoughthe amount of phospholipid tripled during thisperiod, the slow turnover rate of crude lipid ra-dioactivity indicates that the cellular phospho-lipids are stable.The phospholipids present as predominant spe-

cies, i.e., PE, phosphatidylglycerol (PG), andcardiolipin (CL), were isolated by thin-layerchromatography, and each sample was analyzedfor 14C and 32P. Figure IC shows a 16% and 6%decrease of 14C and 32P, respectively, in PE overthe 5-hr growth period. Turnover data for PG(Fig. ID) show a decrease of 14C and 32P of55% and 38.5%, respectively. CL exhibited aslight decrease in 14C (10%), with 32P remainingconstant for 4 hr. A 35% increased incorporationof both 14C and 32P into CL was noted duringthe last hour.

Phospholipid turover-nogrowing cells. Phos-pholipid turnover in a nongrowing culture wasanalyzed under previously described conditions,with the exception that cells were suspended in0.05 M phosphate buffer (pH 7.8). Figure 2Ashows that neither growth nor an increase inlipid phosphorus occurred during the 5-hr incu-bation period in nonradioactive phosphate buffer.The conditions, therefore, enable an assessmentof phospholipid metabolism under nongrowingconditions for direct comparison to growing con-ditions.The turnover of 14C and 32p from the crude

lipid fractions is shown in Fig. 2B. An initialdecrease in 14C and 32P was observed for thefirst 2 hr, followed by an increase to the initiallevel. The individual phospholipids, PE, PG, andCL, were purified and their turnover kineticswere examined. PE did not exhibit turnover ofeither 14C and 32P throughout the 5-hr incuba-tion, indicating this phospholipid to be stableunder these conditions (Fig. 2C). PG, however,exhibited turnover kinetics similar to crude lipid,showing an initial decrease in both 14C and 32Pand an increase to initial levels at later timepoints (Fig. 2D). Analysis of CL showed a 150%increase of 14C in CL over the 5-hr incubation,whereas 32P increased 110% over 3 hr and thenremained constant (Fig. 2E). Two major phos-pholipids of M. cerificans, PG and CL, appear topossess differential turnover rates in nongrowingcells.

VOL. 1 07, 1 971 807

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

11

Janu

ary

2022

by

119.

128.

112.

126.

MAKULA AND FINNERTY

4 * 2

X I .e t I~~~~~~A)

i4 G~~~~~~~ROWTHXgn )g/ OH*~~CUVE

2

2345601234

TIME OMPS

-4

II

0) (A)a: 6 GROTh1JL X CURVE

9 JI T23 4 5 6 2 )4 5

o ;L TIME (HRS)

(SI

CRUDELIPID

(a) '4c()SlIP

e

21p0

.i

0aI.-

(C)

PE10§86

2 3 4 5

4

I-1IL

{D)

Pe

so,

6

4

2 3 4 5

(a)CRUDL IPIDta) '4C

(C)PE

(DOPe

'o8 -a

2O2 34

10'

4

6

40a

6-

(E)CL

TIME IM

FIG. 1. Phospholipid turnover in a growing cultureof Micrococus cerificans. M. ceificans was grown withhexadecane as carbon source with the phosphate con-

centration in the medium reduced to S x 10-3 M.When the culture reached mid-exponential growthphase, I mCi of H,33 P04 and 25 ,Ci of glycerol-i ,3-14C were added (arrow), and the culture was allowedto grow an additional 2 hr.

Metabolism of PG. The lipids of M. cerificanswere labeled with 25 gCi of glycerol-1,3-14Cadded to the medium at mid-exponential growth,and samples were removed at 0.5, 1, 5, 10, 20,30, and 60 min. All samples were pipetted intochloroform-methanol (2:1, v/v) to yield a mono-

1 2 3 4 5

TIME IMS)

FIG. 2. Phospholipid turnover in a nongrowing cul-ture of Micrococcus cerificans. Conditions are thesame as for Fig. 1.

phasic system from which the crude lipid extractwas obtained. A sample of crude lipid from eachsample was assayed for radioactivity. Figure 3shows the kinetics of incorporation of glycerol-1,3-14C label into crude lipid. A constant rate ofincorporation was observed for the initial 10min. This represents assimilation of 3.4% ofavailable "4C-glycerol into phospholipid. Thecrude lipid extracts were fractionated into neu-tral lipid and phospholipid by silicic acid column

808 J. BACTERIOL.

2

6

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

11

Janu

ary

2022

by

119.

128.

112.

126.

PHOSPHOLIPID METABOLISM IN M. CERIFICANS

chromatography. Less than 5% of the radio-activity was found in neutral lipids whereas theremainder of the 14C label from glycerol-1,3-14Cwas in the phospholipid fraction. Since the phos-pholipid fraction contained essentially all the"4C-label, further analysis was performed onlywith the phospholipids.

Analysis for incorporation kinetics of glycerol-1,3-14C into each individual phospholipid isshown in Fig. 4. The curve designated PG1 repre-sents the total incorporation of glycerol-1,3-14Cinto PG without reference to either diglycerideglycerol or headgroup glycerol. PG exhibited the

U ZO

FIG. 3. Incorporation of glycerol-i,3-14C into thechloroform-soluble lipids of Micrococcus cerificans.

highest labeling of all phospholipids analyzedand was, therefore, subject to further analysis.Two dimensional thin-layer chromatography wasemployed to determine the contribution of phos-phatidylglycerol phosphate (PGP) to total "C-glycerol incorporation in PG ,. PG and PGPwere separated on Silica Gel G thin-layer plates(0.25 mm) with the following solvent systems:first dimension, chloroform-methanol-5 NNH,OH (65:30:5, v/v/v); second dimension,chloroform-methanol-glacial acetic acid (70:20:5, v/v/v). The contribution of 14C-PGP to PG1varied between 1.5 and 3% during the course ofthe reaction, so that subsequent analyses to bediscussed do not involve specific procedures forthe removal of PGP from purified PG.PG1 was purified from all samples by prepara-

tive thin-layer chromatography in an acidic sol-vent system (chloroform-methanol-glacial aceticacid; 70:20:5, v/v/v) and hydrolyzed in 90%acetic acid for 20 min. The products, diglycerideand a-glycerophosphate, were separated by etherextraction, and radioactivity was determined ineach fraction. The distribution of total radioac-tivity was 25% in the ether-soluble diglyceridemoiety and 75% in a-glycerophosphate. This factsuggests a difference in the metabolism of theunacylated and diacyl glycerols of PG. The curvedesignated PG2 represents a plot of glycerol-1,3-14C incorporation into the diacylglycerol portion

13 PG1

12

I I

I0

9

8

ro-1---4 ~~~~~~~~~PE

-,

43 PG2

2

1 ~~~~~~~~~~~~CL

01 5 10 20 30 40 50 60TIME ( m;n)

FIG. 4. Glycerol-1,3-14C incorporation into the phospholipids of Micrococcus cerificans. PG1 represents thetotal incorporation of glycerol-1 ,3-14C into PG. PG2 represents the incorporation of glycerol-I,3- 14C into thediglyceride backbone ofPG.

VOL. 107, 1971 809

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

11

Janu

ary

2022

by

119.

128.

112.

126.

MAKULA AND FINNERTY

of PG based on the radioactivity distributionanalysis of 25%. Assuming this 25% representsde novo synthesis of phospholipid from requisiteprecursors, it is possible to calculate percentagecomposition of phospholipid. Table 1 comparesthe percentage composition of M. cerificansphospholipids by two independent methods. Anextremely close approximation of percentagecomposition is obtained.A portion of PG1 from all samples was treated

by mild alkaline hydrolysis to determine theamount of 4C-glycerol converted to fatty acidcarbon. These results demonstrated no radioac-tivity in fatty acids, indicating that all glycerolincorporated into phospholipid was localized inthe water-soluble backbone, glycerylphos-phorylglycerol. The data further indicate the un-acylated glycerol of PG to be metabolically activefor reasons unknown.PE was purified by thin-layer chromatography

and hydrolyzed in 3 N HCI. The water-solubleproducts, ethanolamine-hydrochloride and a-glycerophosphate, were analyzed by paper chro-matography in ethanol-acetic acid-water (90:5:5, v/v/v) and compared to authentic standards.All radioactivity in the sample co-chromato-graphed with a-glycerophosphate. Mild alkaline

TABLE 1. Comparison ofphospholipid composition bytwo independent criteria

Percentage composi- Percentage com-

Phospholipida tion based on position based on14C-glycerol phosphorusincorporation analysisb

PE 60 61.5PG 23 19.9CL 10 12.9

a PE, phosphatidylethanolamine;dylglycerol; CL, cardiolipin.

bSee reference 11.

PG, phosphati-

hydrolysis of PE demonstrated no conversion ofglycerol carbon to fatty acid carbon.CL was not appreciably labeled with 14C

during the course of the experiment. This resultwas unexpected, since CL contains three glycerolmoieties per molecule, and PG, which is thoughtto be a precursor of CL, was highly labeled with14C. A possible explanation for the PG-CL la-beling pattern relates to an alternate biosyntheticpathway for CL. The biosynthetic pathway de-scribed in E. coli involves a condensation reac-tion between cytidine-diphosphodiglyceride andPG to yield CL (15). If this reaction mechanismwere operative in M. cerificans, a highly labeledCL molecule would be expected. An alternativepossibility is that, under conditions of the experi-ment, CL biosynthesis is limited.

14C-Hexadecane assimilation into cellular lipid.The assimilation of hexadecane carbon into cel-

lular lipid was assessed by growing M. cerificansin the presence of 49.4 ACi of hexadecane-J-"4C.Samples (25 ml) were removed at 2-, 4-, 8-, and12-hr intervals, and sodium azide (I ml of 10%solution) was added immediately to each sample.The cells were collected by centrifugation,washed twice with basal medium, lyophilized,and extracted for total lipid. The crude lipid wasfractionated by silicic acid chromatography intononmetabolized hexadecane, neutral lipid, andphospholipid by hexane, chloroform, and meth-anol elution to obtain the respective lipid frac-tions.A 7.92% increase of radioactivity from hexa-

decane-1-14C or products of hexadecane-l-14C as-

similation was observed over the time course(Table 2, A). Extraction of total lipid (Table 2,B) by chloroform-methanol (2:1, v/v) yieldsapproximately 50% of the total radioactivity inthe chloroform-soluble fraction. Since it has beenimpossible to remove hexadecane from cells byaqueous washing procedures, these determinations

TABLE 2. Distribution ofhexadecane-l-14C assimilation Products in the linids of Micrococcus cerifi,ans

J A B C D E F

Time (hr) | Cell massw Total0 (lO | Chloroform- Hexadecane Neutrals Phospholipids Recovery(mgry t) couts/mm)10 soluble (103 (103 (103 (103 eovr

counts/min counts/min) counts/min) counts/min) counts/min) (%)

2 150 250.0 (0.83)b 100.0 (40)t 80.0 (80)t 7.5 (37.5)" 2.2 ( I )b 90c4 150 407.0 (1.36) 210.0 (52) 172.0 (82) 10.8 (28.4) 6.4 (17) 908 150 975.0 (3.25) 485.0 (50) 338.0 (70) 18.0 (12.2) 31.0 (21) 8012 150 2,378.0 (7.92) 665.0 (28) 394.0 (59) 52.0 (19) 104.0 (38.4) 82

a i otal counts per minute were determined by solubilizing 5 mg of dry cells in hyamine hydroxide.b A, Total counts per minute/counts per minute of hexadecane-I-"4C added x 100; B, B/A x 100; C, C/B x

100; D, D/B - C x 100; E, E/B - C x 100.c F, C + D + E/B x 100.

810 J. BACTERIOL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

11

Janu

ary

2022

by

119.

128.

112.

126.

PHOSPHOLIPID METABOLISM IN M. CERIFICANS

are influenced by the presence of hexadecane-l-14C adsorbed to cell surfaces in an unspecifiedmanner. Therefore, fractionation of the totalcrude lipid was necessary to determine theamount of hexadecane-1-14C in each sample.Hexadecane was removed from the crude lipidsamples by silicic acid chromatography elutingwith 20-column volumes of hexane. Thin-layerchromatography and gas chromatography of thehexane fraction indicated only hexadecane-1-14Cto be present. Hexadecane-l-14C accounted for80% of the total chloroform-soluble radioactivityat 2 hr and decreased to 59% at 12 hr (Table 2,C). The neutral lipid was removed by elution ofthe silicic acid column with 10-column volumesof chloroform with total radioactivity increasingthroughout the time course (Table 2, D). Thepercentage of radioactivity derived from hexade-cane-l-14C decreased during the time samplings.Neutral lipid samples were analyzed by thin-layer chromatography for free fatty acid and freefatty alcohol. Commercial thin-layer chromatog-raphy silica gel sheets (Eastman chromograms)employing two independent solvent systems gave

identical results [(i) petroleum ether-diethylether (85:15, v/v); (ii) petroleum ether-diethylether-glacial acetic acid (90: 10: 1, v/v/v)]. Freefatty acid contributed 8.0, 9,0, 12.0, and 10.0%of the total radioactivity to the neutral lipidsamples whereas free fatty alcohol accounted for9.0, 8.3, 8.0, and 6.0% in the 2-, 4-, 8-, and 12-hrsamples, respectively. The ratios of total phos-pholipid radioactivity to total neutral lipid ra-

dioactivity for the samples were 0.29, 0.59, 1.72,and 2.0, indicating progressively greater labelingof phospholipids through the growth cycle. Theresults demonstrate that hexadecane-)-_4C was

converted to radioactive intermediates which canbe isolated in the neutral lipid fraction. Sinceonly I hexadecane carbon atom in 16 containsradioactivity, interpretation becomes tenuouswith respect to total neutral lipid and conversionefficiency. The phospholipid fraction was elutedwith 20-column volumes of methanol and con-

centrated to a known small volume. The phos-

pholipid fraction reflects an increasing per-centage of radioactivity derived from "4C-hexa-decane metabolism over the time course (Table 2,E). Radioactivity appearing in this fraction rep-resents assimilation of metabolic intermediatesderived from the substrate by direct incorpora-tion (i.e., fatty acids) or indirect incorporation ofcarbon moieties derived from hexadecane metab-olism (i.e., glycerophosphoryl esters).

Phospbolipid labeling. The phospholipid frac-tion was analyzed by thin-layer chromatographyto determine the radioactivity in each phospho-lipid species throughout the time course (Table3). Phosphatidylserine (PS) never exceeded 1.6%of the total phospholipid radioactivity, whereasradioactivity contained in PE was 41% of the totalat 2 hr and increased to 63% at 8 hr. PE exhib-ited the greatest change among the phospholip-ids, with the exception of CL, which decreasedfrom 25.9% (2 hr) to 6.4% (12 hr). Radioactivityin PG and PGP remained relatively constant inall samples. PG and PGP are combined sinceseparate analyses showed PGP to contribute only4% to the total radioactivity for all samples.PE, PG, and CL were purified, hydrolyzed by

mild alkaline hydrolysis, and separated intoether-soluble fatty acids and water-soluble glyc-erolphosphoryl esters. Figure 5 shows the distri-bution of radioactivity into each of these derivedmoieties for PE, PG, and CL. PE contained 75%of its radioactivity in the fatty acids at 2 hr, in-creasing to 90% at 4 hr. The water-soluble glyc-erolphosphoryl ester did not exhibit significantradioactivity at any time point throughout theexperiment. For radioactivity to appear in thewater-soluble portion of the phospholipid mole-cule, extensive degradation and resynthesis ofhexadecane carbon into requisite intermediatesmust occur. In experiments using terminally la-beled substrates, the factor of metabolic dilutionof carbon becomes of critical importance. There-fore, low labeling patterns may reflect either ex-

tensive turnover or minimal incorporation of rel-evant precursors into specific phospholipids.The comparison of radioactivity distribution

TABLE 3. Distribution of radioactivity between major phospholipids

Total counts/minPhospholipida

2 hr 4 hr 8 hr 12 hr

PS 0 (0)b 100 (1.6) 400 (1.4) 1,200 (1.3)PE 864 (41) 2,960 (48) 18,240 (63) 61,600 (64)PG-PGP 736 (36) 2,560 (42) 9,760 (34) 32,000 (33)CL 560 (25.9) 920 (14.3) 2,920 (9.4) 6,400 (6.4)

a PS, phosphatidylserine; PE, phosphatidylethanolamine; PG-PGP, phosphatidylglycerol-phosphatidylglycerolphosphate; CL, cardiolipin.bNumbers in parentheses represent percentage of radioactivity for each designated phospholipid of total phos-

pholipid radioactivity recovered. All recoveries of phospholipid radioactivity exceeded 95%.

811VOL. 107, 1971

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

11

Janu

ary

2022

by

119.

128.

112.

126.

MAKULA AND FINNERTY

PE PG CL

100

90

80

70

60

50

40

30

20

10

2 4 8 120 2 4TIME (HI

FIG. 5. Distribution of phospholipid radioactivity dencerificans. Symbols: ether soluble (x); water soluble (0).

patterns of PE with PG and CL reveals a distinctdifference. PG and CL exhibit extensive radioac-tivity labeling of the water-soluble backboneearly in the time course, with radioactivity de-creasing with time. The fatty acids comprisingthe makeup of PG and CL show an inverse la-beling pattern to labeling of the water-solubleglycerolphosphoryl esters.

DISCUSSIONThe purpose of these studies was to investigate

a possible relationship between the metabolismof M. ceraficans phospholipids and hydrocarbonassimilation. The turnover kinetics of the majorphospholipids were examined in context of de-fining a functional rather than structural role forindividual phospholipids. Since total phospho-lipid was 50% higher in hydrocarbon-grown cells(I 1), it seemed reasonable that one or morephospholipid species would exhibit rapid turn-over properties.PE is the major component; it represents

61.5% of the total phospholipid and exhibited nosignificant rate of turnover in either growing ornongrowing cultures. The turnover rate of 14C-PE was 3.2% per hr and for 32P-PE it was 1.2%per hr (approximate doubling time, 60 min).Ames showed similar results with Salmonellatyphimurium (2), and Kanfer and Kennedy withEscherichia coli B (9). White and Tucker re-ported that PE of Haemophilus parainfluenzaeexhibited no turnover of the glycerol moiety,whereas the ethanolamine and phosphate moie-ties turn over more rapidly (17).PG exhibited a turnover rate of 11 and 7.7%

per hr for "IC and 32P, respectively, indicatingan extremely slow metabolism of PG in growing

120 2 4 8 12

hexadecane-l-14C oxidation by Micrococcus

cells. These data differ significantly from thosereported by Ames (2) and Kanfer and Kennedy(9), where approximately 50% of the PG turnedover in excess of one generation. White andTucker detailed the turnover of PG from H. par-ainfluenzae, demonstrating disproportionate turn-over kinetics for the diacyl glycerol, unacylatedglycerol, and the phosphate moieties of PG (17).These workers found that the unacylated glyceroland phosphate of PG lose radioactivity at iden-tical rates, whereas the loss of radioactivity fromthe diacyl glycerol moiety of PG is 2.7 timesslower. The turnover data for PG in nongrowingM. cerificans cultures was characterized by aninitial decrease in the amount of radioactivityfollowed by an increased incorporation of ra-dioactivity back into PG. This pattern might re-sult from the action of phospholipase D on PGyielding phosphatidic acid and glycerol, whichare reincorporated into PG without dilutionthrough a pool. Since cardiolipin exhibited anincrease in incorporation of radioactivity, it issuggestive that PG is a precursor of CL and, fur-ther, that CL synthesis is stimulated under non-growing conditions. Evidence concerning the bio-synthetic pathway of CL has indicated that intactPG is converted to CL by addition of a moleculeof phosphatidic acid (15). Our data concerningmetabolism of PG indicate a preferential 75%incorporation of glycerol-i,3-14C into the unacy-lated portion of the PG molecule over incorpora-tion into the acylated portion. The significanceremains obscure but suggests a role for specificphospholipases in the metabolism of membranephospholipids. A phospholipase D specific forCL has been described in H. parainfluenzaeyielding phosphatidic acid (PA) and PG as prod-

812 J. BACTERIOL.

i----x

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

11

Janu

ary

2022

by

119.

128.

112.

126.

PHOSPHOLIPID METABOLISM IN M. CERIFICANS

ucts (13). This CL-specific phospholipase D hasbeen implicated in the active degradation of CLto PA and PG with a corresponding resynthesisof CL from the hydrolysis products.The turnover of CL was extremely slow in

growing M. cerificans, whereas nongrowing cellsexhibited a 30 and 36% increase per hr for 14C-CL and 32P-CL, respectively, for the initial 3 hr.Since CL represents 12.9% of the total phospho-lipids and PG contains 62% of the glycerol-1,3-14C incorporated into all phospholipids, it was

surprising that CL contained only approximately5% of all radioactivity incorporated into thephospholipids. If PG constitutes the cellular pre-cursor for CL biosynthesis, more extensively la-beled CL would be expected. This conclusion isfurther supported by consideration of the turn-over data from growing cells indicating slowmetabolism of CL. Alternatively, PG could beindependently turning over exclusive of its con-version to CL. Our data indicate that the head-group glycerol of PG is metabolically labile andthat a significant percentage of the PG content iscontinuously changing with respect to specificactivity.These studies have revealed that glycerol-i,3-

14C is not converted to phospholipid fatty acid.Mild alkaline hydrolyses of PE, PG, and CLshowed all radioactivity in the water-solublebackbone of the phospholipid. A comparisonwith H. parainfluenzae (17), which converts 63%of the total "4C-glycerol incorporated into lipidto phospholipid fatty acid, demonstrates a

striking difference between these two microorga-nisms. Of several compounds tested for conver-sion to lipid carbon, only acetate, palmitate,cetyl alcohol, and hexadecane have demonstratedsignificant results with hexadecane-grown cells.Negative compounds tested were methionine-methyl-5 4C, "4C-serine, "4C-ethanolamine and4C-glycerol.The analysis of hexadecane-1- 4C conversion

to cellular lipids represents an attempt to definefunctional lipid metabolic patterns in M. cerifi-cans. Owing to inherent and obvious difficultiesassociated with studying the assimilation of a

terminally labeled alkane, the data point to sev-eral pertinent observations concerning hydro-carbon metabolism in this microorganism. Totalradioactivity associated with cells increased line-arly, paralleling cellular growth. Fractionation oftotal cellular lipid indicated that the percentageof available hexadecane, specifically or nonspe-cifically bound to cells, decreased throughout thegrowth period. Total neutral lipid radioactivityincreased approximately sevenfold, although theproportion of the radioactivity in total neutrallipid decreased when compared to chloroform-

soluble radioactivity minus hexadecane radioac-tivity (see Table 2). The percentage compositionof free fatty acid in neutral lipid remained rela-tively constant in all samples. Phospholipid ra-dioactivity increased 50-fold over the growth cy-cle, in addition to increasing in percentage ofradioactivity. These results indirectly support ourconclusions concerning a direct role for fattyacid metabolism in M. cenficans (Makula andFinnerty, submitted for publication).Phospholipid fatty acid incorporation of ra-

dioactivity derived from hexadecane-1-54C oxida-tion demonstrated distinct differences betweenPE and PG-CL. PE was extensively labeled withfatty acids early in the growth cycle (2 hr). PGand CL fatty acid labeling patterns were initiallylow, increasing with growth. These data suggestrapid turnover properties for fatty acyl residuesin M. ceraficans phospholipids. Data presenteddemonstrated that 98% of the phospholipid fattyacids from hexadecane-grown cells were 16 car-bons. Since no lysophospholipids have been de-tected in M. cenficans (11), an active acylase isimplicated.

Initial objectives imposed at the outset of thisstudy have been partially accomplished. Thequestion of relevancy for membrane phospho-lipids in hydrocarbon oxidation remains unan-swered. Structural relationships of fatty acids tohydrocarbon chain length suggest a functionalcorrelation (submitted for publication). Currentdata presented demonstrate that the polar por-tions of M. cerificans are relatively stablewhereas the apolar side chains appear to be met-abolically labile. Further analyses are in progressdetailing the turnover parameters of fatty acidsin individual phospholipids with respect to posi-tional specificity and structural changes.Novel membrane properties appear character-

istic of this microorganism. Enzymatic propertiesassociated with these membranes become rele-vant to structure-function relationships. Mem-brane-bound enzymes demonstrated in M. cerifi-cans are cytidine triphosphate: phosphatidic acidcytidyl transferase, a-glycerol phosphate-cytidinemonophosphate-phosphatidyl transferase, L-serine-cytidine monophosphate-phosphatidyltransferase, phosphatidylserine decarboxylase,and phosphatidic acid phosphatase (W. R. Fin-nerty, American Chemical Society SE-SW Meet-ing, Dec. 1970, p. 31). Membrane fractionationplus further characterization of these enzymesoffer probes into membrane structure and lipidinvolvement in membrane function.

ACKNOWLEDGMENTSThis investigation was supported by grant GB 8208 from the

National Science Foundation to W.R.F. Public Health Servicepredoctoral fellowship GM-41274-01 was awarded to R.A.M.

813VOL. 107 , 1971

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

11

Janu

ary

2022

by

119.

128.

112.

126.

MAKULA AND FINNERTY

LITERATURE CITED

1. Abood, L. G., and D. S. Rushmer. Surface interaction ofcalcium and ATP with phospholipids and other surfac-tants, p. 169-188. In E. D. Goddard (ed.), Molecularassociation in biological and related systems. Advancesin chemistry series, no. 84. American Chemical Society,Washington, D.C.

2. Ames, G. F. 1968. Lipids of Salmonella typhimurium andEscherichia coli: structure and metabolism. J. Bacteriol.95:833-843.

3. Ansell, G. B., and J. N. Hawthorne. 1964. The role ofphospholipids. p. 256-258. In Phospholipids - chemistry,metabolism, and function. B.B.A. Library, vol. 3. El-sevier Publishing Co., Amsterdam.

4. Azzi, A., S. Fleischer, and B. Chance. 1969. Cytochrome cand phospholipid interaction: structural transitions asso-ciated with valency changes. Biochem. Biophys. Res.Commun. 36:322-327.

5. Bangham, A. D., M. N. Standish, J. C. Watkins, and G.Weismann. 1967. The diffusion of ions from a phospho-lipid model membrane system, p. 183-187. In L. Bolis(ed.), Symposium on biological transport. Springer-Verlag Co., New York.

6. Eibl, H., E. E. Hill, and W. E. M. Lands. 1969. The sub-cellular distribution of acyltransferases which catalyzethe synthesis of phosphoglycerides. Eur. J. Biochem. 9:250-258.

7. Gel'man, N. S., M. A. Lukoyanova, and D. N. Ostrovski.1967. Respiration and phosphorylation of bacteria.Plenum Press, New York.

8. Gould, R. M., M. P. Thornton, V. Liepkalns, and W. J.Lennarz. 1968. Participation of aminoacyl transfer ribo-nucleic acid in aminoacyl phosphatidylglycerol synthesis.II. Specificity of alanyl phosphatidylglycerol synthesis.J. Biol. Chem. 243:3096-3104.

9. Kanfer, J., and E. P. Kennedy. 1963. Metabolism andfunction of bacterial lipids. I. Metabolism of phospho-lipids in Escherichia coli B. J. Biol. Chem. 238:2919-2922.

10. Leslie, R. B. 1968. Membranes and bioenergetics, p. 289-346. In D. Chapman (ed.), Biological membranes. Aca-demic Press Inc., New York.

11. Makula, R. A., and W. R. Finnerty. 1970. Microbial as-

similation of hydrocarbons: identification of phospho-lipids. J. Bacteriol. 103:348-355.

12. Nesbitt, J. A., and W. J. Lennarz. 1968. Participation ofaminoacyl transfer ribonucleic acid in aminoacyl phos-phatidylglycerol synthetase. I. Specificity of lysyl phos-phatidylglycerol synthetase. J. Biol. Chem. 243: 30883095.

13. Ono, Y., and D. C. White. 1970. Cardiolipin-specific phos-pholipase D of Haemophilus parainfluenzae. II. Charac-teristics and possible significance. J. Bacteriol. 104:712--718.

14. Redwood, W. R., H. Mulder, and T. E. Thompson. 1969.Interaction of a bacterial adenosine triphosphatase withphospholipid bilayers. Proc. Nat. Acad. Sci. U.S.A. 64:989-996.

15. Stanacev, N. Z., Y. Y. Chang, and E. P. Kennedy. 1967.Biosynthesis of cardiolipin in Escherichia coli. J. Biol.Chem. 242:3018-3019.

16. White, D. C. 1968. Lipid composition of the electron trans-port membrane of Haemophilus parainfluenzae. J. Bac-teriol. 96:1 159-1170.

17. White, D. C., and A. N. Tucker. 1969. Phospholipid me-tabolism during bacterial growth. J. Lipid Res. 10:220233.

18. White, D. C., and A. N. Tucker. 1969. Phospholipid me-

tabolism during changes in the proportions of mem-

brane-bound respiratory pigments in Haemophilus par-

ainfluenzae. J. Bacteriol. 97:199-209.

814 J. BACTERIOL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

11

Janu

ary

2022

by

119.

128.

112.

126.