THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 268, NO. 4, Imue of February 5, pp. 2828-2835.1993 Printed in U. S. A.

Sarcina ventriculi Synthesizes Very Long Chain Dicarboxylic Acids in Response to Different Forms of Environmental Stress*

(Received for publication, July 17, 1992)

Seunho Jung$., Susan E. Lowe$., Rawle I. HollingsworthSGT, and J. Gregory Zeikus$.II From the Departments of $Biochemistry, $Chemistry, and IlMicrobiology and Public Health and the Nutionak Science Fou~dation Center for Micmbiai Ecotogy, Michigan State University, East Lansing, M i c h ~ a n 48824

Changes in the composition of membrane lipids in a strictly anaerobic, facultative acidophilic eubacterium, Sarcina ventriculi, were studied in response to various forms of environmental stress. Changes in lipid com- position and structure occurred in response to changes in environmental pH. At neutral pH, the predominant membrane fatty acids ranged in chain length from CM to Cls. However, when cells were grown at pH 3.0, a family of unique very long chain fatty acids containing 32-36 carbon atoms was synthesized and accounted for 50% of the total membrane fatty acids. These acids were identified as very long chain a,@-dicarboxylic acids ranging in length from 28 to 36 carbons by electron impact mass spectrometry of methyl and (per- deuterio) methyl ester derivatives. These methyl esters all bore a vicinal dimethyl group toward the center of the chain. The assignment of the structures was con- firmed by isolating one of the very long chain unusual fatty acids as the ester form after methanolysis and performing further analyses including 'H and "C NMR spectroscopy and Fourier transform infrared spectros- copy. Coupling this information with the data from gas chromatography/mass spectrometry analysis, the ex- act structure was confirmed as a,@- 15,16-dimethyltri- cotanedioate dimethyl ester.

Addition of alcohols, either metabolic (0.25 M ethanol) or nonmetabolie (0.05 M butanol) to cells grown at pH 7.0, or thermal stress (growth tempera- ture at pH 7.0 was raised from 37 to 46 or 55 "C) also resulted in the synthesis of these very long chain fatty acids. Synthesis of these very long chain qo-dicarbox- ylic acids was reversed by reducing the temperature back to 37 *e. S. ventriculi is also unusual in that the membrane components are not the usual phospholipid components but appear to be predominantly glyco- lipids.

Bacterial membranes are extremely dynamic, complex sys- tems that interface directly with the environment and carry out many important functions. They provide compartmental- ization and transport and act as a matrix for cytoplasmic macromolecules and membrane proteins.

Bacteria in their natural ecosystem may experience many

of Energy Grants DE-FGO2-87ER13719 ( t o J. G. Z.) and DE- * This work was supported in part by United States Department

FG0289ER14029 (to R. I. H.) and by fellowships from the Center for Microbial Ecology, a National Science Foundation Science and Tech- nology Center, Michigan State University (to S. E. L. and S. J.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

ll To whom correspondence should be addressed.

changes in environmental factors, including temperature, pH, solvent concentration, nutrient levels, and oxygen concentra- tion. In this respect, the membrane, which is directly exposed to the environment, is one of the most critical and vulnerable components of the cell and must adapt to and survive these changes.

Sarcina uentriculi is a strictly anaerobic bacterium that can grow on sugars over a wide pH range, from pH 2 to 10 (Canale- Parola, E., 1970, 1986). Detailed physiological studies have been done on the influence of environmental pH range on growth, fermentation product formation, and the proton mo- tive force in this organism (Goodwin and Zeikus, 1987; Lowe and Zeikus, 1991; Tilak, 1970).

S. ventriculi also undergoes morphological adaptations in response to changes in environmental pH. Regular tetrads are formed at low pH, whereas cells are irregular in shape with higher numbers of cells within each packet at neutral pH, and spores are formed at alkaline pH {Lowe et aL, 1989).

The purpose of this report is to examine the effect of perturbation of environmental parameters (specifically of pro- ton and solvent concentrations and temperature) on mem- brane composition and structure. We report that the lipid composition and fatty acid structure dramatically change in response to these perturbations, resulting in the synthesis of a family of long chain a,@-dicarboxylic fatty acids. These new lipids are believed to be important for maintaining membrane integrity under the new conditions.

MATERIALS AND METHODS

Chemicals, Gases, and Isotopes-All chemicals were reagent-grade or better and were obtained from Sigma or Mallinckrodt~ Inc., Paris, KY. All gases were at least 99.9% pure and were passed over copper- filled Vycor furnaces (Sargent Welch Scientific Co., Skokie, IL) to remove oxygen.

Organism and Culture Condition+"-. ventriculi JK was cultivated as described previously (Goodwin and Zeikus, 1987). For growth under pH control, 4-liter Kimax jars (Baxter Scientific Products, Romulus, MI) containing 3 liters of medium were used. The jars were equipped with a pH probe, and the culture was mixed by placing the jars on a magnetic stirrer. When the organism was grown at pH 3.0, the initial pH did not change during the fermentation; at pH 7.0, the pH was controlled by the addition of 5 M NaOH. Cultures were harvested as outlined previously (Lowe et al., 1989) under aerobic conditions.

For temperature shift experiments and growth in the presence of solvents, 750-ml fermentation vessels (New Brunswick Scientific, New Brunswick, NJ) containing 350 mf of medium were used. The vessels were agitated at 200 rpm and pH was maintained at either pH 3.0 or 7.0. To determine the effect of solvents on membrane composition, vessels containing medium with 0.25 M ethanol or 0.05 M butanol at pH 7.0 were inoculated with a 5% (v/v) inoculum of cells grown at pH 7.0 in the absence of solvent. The cells were harvested at midexponential phase, washed twice with distilled water, and stored at -70 "C for further analysis.

Membrane Preparations-Cells were disrupted by passage through a French pressure cell (American Instruments Go., Inc., Silver Spring, MD) at 20,000 lb/in2. The disrupted cells were centrifuged at 20,000

2828

by guest on March 30, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Long Chain Dicarbox~lic Acids in S. uentriculi 2829

x g to remove unbroken cells, and the supernatant was centrifuged a t 110,000 X g to sediment the membranes, which were washed twice with distilled water.

Total Fatty Acid Analysis-Fatty acid analyses were performed on whole cells or isolated membrane fractions by treatment with meth- anolic HCI using either of two procedures. Procedure a was employed for whole cells and procedure b for isolated membranes. (a) cells (1- 5 mg) suspended with 0.3 ml of chloroform and 1.5 ml of 5% metha- nolic HCl solution were sealed in a Teflon-lined screw-capped vial and heated in a water bath or oven at 72 "C for 24 h. Chloroform (3 ml) was added every 8 h, followed by mild sonication for 5 min. After concentration to dryness under nitrogen gas, samples were partitioned between water and chloroform, and the aqueous layer was washed several times with chloroform or hexane. The combined solutions were filtered through glass wool. (b) 3 ml of chloroform was added to 1 ml of membrane suspension, followed by 15 ml of 5% methanolic HCl solution. The flask was sealed and heated in an oven at 72 "C for 12 h. 3 ml of chloroform was added every 6 h, followed by mild sonication for 5 min. The mixture was then concentrated on the rotary evaporator to dryness and extracted with chloroform. The combined organic fraction was redissolved in 1 ml of hexane. The fatty acid methyl esters prepared by either procedure a or b were subjected to gas chromatography analysis on a 25 M J&W Scientific DB1 capillary column using helium as the carrier gas and a temper- ature program of 150 "C initial temperature, 0.00 min hold time, and 3.0 "C/min rate to a temperature of 200 "C. A second ramp of 4.0 "C/ min was then immediately started until the final temperature of 300 "C was obtained. This temperature was held for 30 min. The relative proportion of lipid components were calculated from the integrated peak areas. The fatty acid identification and molecular weight were determined using GC/MS' analysis using a Jeol JMS- AX505H spectrometer interfaced with a Hewlett-Packard 5890A gas chromatograph.

Extraction of Lipids-Lipids were extracted from the isolated mem- brane or whole cells using procedures a and b, respectively. (a) TO each 5-10-mi membrane suspension, 30 volumes of chloroform/meth- anol (5:1, v/v) was added and then mixed to produce a single phase. The mixture was shaken or stirred vigorously, with intermittent sonication (approximately 5 min every 30 rnin), for 2 h a t 45 "C. The combined extracts were taken to dryness in a rotary evaporator. The residue was partitioned between 10 ml of chloroform/methanol (5:1, v/v) and 2.5 mi of water. The lower phase was taken to dryness and redissolved in 1 ml of chloroform/methanol (9:1, v/v). (bf Cells of S. ventriculi from 50 liters of culture medium were harvested by centrif- ugation at 10,000 X g for 10 min. Lipids from approximately 50 g, wet weight, of cells were extracted at 45 "C with 400 mi of a mixture of chloroform/methanol/water (15:3:2, by volume) for 2 b, followed by 200 ml of chloroform/methanol (5:1, v/v). Extraction was performed with intermittent sonication for 2 h. After centrifugation at 20,000 X g, the pellet (cell debris) was extracted again with the same solvent systems. After centrifugation, the supernatant was taken to dryness in a rotary evaporator, dissolved in 10 ml of chloroform/methanol (5:1, v/vf and then shaken with 2.5 ml of water. The lower phase containing the lipids was taken to dryness and dissolved in 1 ml of chloroform.

Analysis of Lipids-As a preliminary step for comparing the total lipid profiles between cells of S. uentriculi growing at pH 3.0 uersus pH 7.0, the lipids were separated by two-dimensional TLC using chloroform/methanol/ammonia/water (3.3:1.0:0.1:0.05, by volume) for the first dimension and chloroform/methanol/water (71.60.2, by volumes) for the second dimension. Analyses were performed on silica gel plates (Merck). Spots were made visible either by spraying with 50% ethanolicsulfuric acid and heating at 250 "C to char the organic components or by spraying with a 0.1% solution of 2',7-dichloroflu- orescein in aqueous ethanol (1:l) and viewing under ultraviolet light (Morris, 1962). Standard phospholipids, such as phosphatidylcholine, phosphatidylethanolamine, phosphatidylserine, and cardiolipin, and neutral lipids were used as standards in addition to free fatty acids. Spraying agents (Holme and Peck, 1983) for the detection of com- ponents included ninhydrin for phosphatidylethanolamine or phos- phatidylserine, dragendorff agent for phosphatidylcholine, orcinol for glycolipid, and molybdenum blue for phosphate.

Isolation of the a,@-Dicarboxylic Acid Dimethyl Ester-The total lipids (100 mg) extracted from the cells as described before were methanolyzed with 5% (w/v) HC1 in methanol (5 ml) for 12 h at

The abbreviations used are: GC, gas chromatography; MS, mass spectrometry.

72 "C. Chloroform (1 ml) was added every 4 h followed by mild sonication for 5 min. The mixture was then concentrated on the rotary evaporator to dryness and extracted with chloroform. The combined organic fraction was redissolved in 1 ml of hexane. This fraction was applied to a silica flash chromatography column and eluted with chloroform/hexane (l:l, by volume). Fractions were as- sayed by gas chromatography. Fractions containing very long chain fatty acid methyl esters were then concentrated and rechromato- graphed on 5% AgN03 impregnated preparative thin layer silica chromatography plates that were eluted with petroleum ether/diethyl etherlacetone (101:0.5, by volume). Spots were made visible either by spraying with 50% ethanolicsulfuric acid and heating at 250 "C to char the organic components or by spraying with a 0.1% solution of 2',7-dichlorofluorescein in aqueous ethanol (1:l) and viewing under ultraviolet light (Morris, 1962). Bands were scraped from the plate into a column fitted with a sintered disc, and the material was eluted from the silica gel with methanol and chloroform. Each fraction was concentrated by evaporation and redissolved in chloroform for further analysis. The purity of each fraction was assayed by GC/MS. One band containing the pure c~,w-15,16-dimethyltricotanedioate dimethyl ester was subjected to NMR and Fourier transform infrared analysis.

Isotope Labeling-Isotope labeling was used to aid in deducing the structures of esterified lipid components using GC/MS. Methyl esters of fatty acids obtained by acid methanolysis were further treated with 5% deuterated methano~ic HC1 for 6 h at 72 "C. Deuterated methyl esters of fatty acids were extracted and analyzed as described for the natural abundance samples. 'H NMR and 13C NMR Spectroscopy-Proton NMR spectra were

recorded at 300 MHz on solutions in CDCla, Fourier transform I3C

NMR spectra were recorded at 125 MHz on solutions in CDC13. Chemical shifts are quoted relative to the chloroform resonances taken at 7.24 ppm for proton and 77 ppm for "C measurements, respectively.

Fourier Transform Infrared Spectroscopy-The spectrum was ob- tained with a Nicolet model 710 FT-IR spectrometer on a 10% (w/vf solution of dicarboxylic acid dimethyl ester in chloroform.

RESULTS AND DISCUSSION

Experiments were initiated to examine lipid composition changes when cells of S. ventricdi were grown at pH 7.0 versus 3.0. Two-dimensional TLC analyses on the lipids from intact cells or isolated membrane of cells are shown in Fig. 1. Analysis of the components for specific lipids such as phos- phatidylethanolamine, phosphatidylserine, and cardiolipin in- dicated that only traces of these are present and that the major components are glycolipids. There are markedly differ- ent proportions of phospholipids and glycolipids in the cells grown at the two pH values.

In order to ensure efficient extraction of membrane lipids from the cells of S. ventriculi grown at pH 3.0, a relatively nonpolar solvent system was selected and used to solubilize the strong hydrophobic lipid groups at an elevated tempera- ture of 45 "C. GC analyses of fatty acid methyl esters extracted from cells grown at pH 7.0 versus 3.0 are shown in Fig. 2, A and B, respectively. Most of the fatty acids in the membrane of S. ventriculi grown at pH 7.0 were between 14 and 18 carbon atoms long (Fig. 2 A ) , but when cells were grown at pH 3.0, a family of novel lipid components appeared (Fig. 2B). The relative proportion of the unusual and unique lipid components was more than 50% by mass at pH 3.0, and only trace levels ( ~ 7 % ) were present during growth at pH 7.0. Further analysis of these lipids from pH 3.0 cells using GC after methanolysis revealed that peaks a, f, i, and j of Fig. 1A contained unusual fatty acid groups. This is a major difference between the cells grown at acid uersus neutral pH. This indicated that the unusual fatty acids are integral lipid com- ponents and do not occur in free uncombined form.

Fig. 3, A and B, shows the mass spectra of one of the unusual lipid c o m ~ n e ~ t s that appeared in the membrane of S. ventriculi grown at high proton concentration (pH 3.0). The same lipid components were also present in cells grown

by guest on March 30, 2018

http://ww

w.jbc.org/

Dow

nloaded from

2830

A

Long Chain Dicarboxylic Acids in S. ventriculi

Ir,

FIG. 1. Two-dimensional TLC analyses of the lipids ex- tracted from the cell of S. ventriculi grown at pH 7 ( A ) uersus pH 3 ( B ) . Equal amounts (-0.8 mg) of total lipids of the membranes isolated from the cells of S. uentriculi grown at different pH values were compared in two-dimensional TLC plates. A (a, b, c, j ) and B ( a , b, k ) , phospholipids; A ( e , j , g, h, i) and B (e, d, j , g, h, i, j ) , glycolipids. Spots A (a, j , i, j ) contained unusual fatty acids with long retention times on GC/MS.

Retention time (effective boiling points)

FIG. 2. Gas chromatographic analyses of lipid components in the membrane of S. uentriculi grown at pH 7 ( A ) versus pH 3 ( B ) . Total fatty acids within the membrane were analyzed as fatty acid methyl ester derivatives after methanolysis. 1, Clk0, carbox- ylic acid methyl ester; 2, C17:l, fatty aldehyde; 3, GI,:,, carboxylic acid methyl ester; 4 , Cleo, carboxylic acid methyl ester; 5, unknown; 6, C,,:,, carboxylic acid methyl ester; 7, C,,:,, carboxylic methyl ester; 8, C,,:O, a,w-dicarboxylic dimethyl ester (cu,w-15, 16-dimethyltricotane- dioate dimethyl ester); 9, Cakl, a,w-dicarboxylic dimethyl ester; IO, C,,:,, a,w-dicarboxylic dimethyl ester.

FIG. 3. Electron impact mass spectrum (70 eV) of Q,W-

15,16-dimethyltricotanedioate dimethyl ester) without ( A ) and with isotope labeling ( B ) . A , The electron impact spectrum contained major ions a t 538, 506, and 475. These corresponded to molecular ion (M+) with the sequential losses of methanol and meth- oxy group, respectively. B, the electron impact spectrum of deuterium- labeled molecules obtained by deuteration with D-4 methanolic HC1 solution. The presence of two carboxylic and two branching methyl groups was confirmed by isotope labeling.

FIG. 4. Electron impact mass spectral fragmentation pat- tern of c~,w-15,16-dimethyltricotanedioate dimethyl ester. All numbers indicate the ionic fragments ( m l z ) . 31 mass units indicates the methoxy group (-CH,O), 32 the methanol (CHXOH), and 42 the ketene (CHZCO). The ions at m/z 74 and 87 are characteristic McLafferty fragments of aliphatic ester groups.

in the presence of high alcohol or higher temperatures at pH 7.0. The electron impact mass spectrum in Fig. 3A shows major ions at m/z 538, 506, and 475, including characteristic McLafferty fragment ions 74 and 87. These corresponded to the molecular ion of a C32-a,w-dicarboxylic dimethyl ester (M+) with the sequential losses of methanol (CH30H) and methoxy (CH30) groups, respectively. The ion 464 m/z rep- resents the structure obtained by sequential losses of metha- nol (CH30H) and ketene (CH2CO) (Fig. 4, Table I).

The mass spectrum also contained a series of ions the general structure of which was CH30CO-(CH2),, beginning at m/z 74, characteristic of a saturated methyl ester. This series of ions continued up to and included two prominent ions at m/z 269 and 297 (296 is due to loss of one hydrogen from the m/z 297 fragment). The intense clusters of ions 28 mass units apart centered at m/z 269 (n = 15) and m/z 297 ( n = 17) strongly indicated the presence of a vicinal dimethyl

by guest on March 30, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Long Chain Dicarboxylic Acids in S. ventriculi 2831

TABLE I Analysis of 70 eV electron impact mass spectral fragments of ol,o-Z5,16-dimethyltricotune dimethyl ester

Structure of the ionic fragments mlz m/z (deuterium labeling)

M (Molecular ion) 538 544 CH3OCO(CH,)13-(CH3)CHCH(CH3)-(CH2)12CHCO 506 509 OC(CH,)13-(CH3)CHCH(cH,)-(CH,),,CHCO 475 475 506 (m/z)-18 (H20) 488 491 CH30CO(CH,)13-(CH3)CHCH(CH3)-(CH,),oCH = CH, 464 467 CH3OCO(CH,)13-(CH3)CHCH(CH3)-CH, 311 313 CH,OCO(CH,)13-(CH3)CHCH(CH3) 297 300 CO(CH~)I,-(CH~)CHCH(CH~) 266 266 CH~OCO(CHS)~-(CHB)CH 269 272 CH30CO(CH,)u 241 243 C O ( C H ~ I ~ - ( C H ~ ) C H 238 238 CO(CHA3 210 210 CH,-O-CO-(CH& 143 146 CH,-O-CO-(CH,), 129 132 CH3-O-CO-(CH,)d 115 118 CH3-O-CO-(CH,)3 101 104 CH3-O-CO-(CH,)Z 87 90 CH3-O-CO-(CH,)2 74 77 CHs-(CH,),; n = 3-8 57, 71, 85, 99, 113, 127 57, 71, 85, 99, 113, 127 CH~=CH-(CHY)A = 2-7 55, 69, 83, 97, 111, 125 55, 69, 83, 97, 111, 125

~ " ~ . " ' . ' 1 ' ~ " 1 " " 1 " " ' " " ' " " ' " " I

e 7 6 5 4 3 2 1 oom

FIG. 5. The 300 MHz 'H NMR spectrum of a,w-15,16-dimethyltricotanedioate dimethyl ester. Signals at d 3.65 were assigned to methyl groups of a methoxycarbonyl function. The peak a t d 1.28 is due to the methylene groups in the saturated hydrocarbon chains. Note also three characteristic signals, at d 2.28 (t, J = 13.2 Hz), representing the methylene group adjacent to the methoxy carbonyl function, at d 1.58 due to the protons of the p-carbons of this molecule, and at d 0.74 (d, J = 7.2 Hz), corresponding to the protons of a vicinal methyl group. The singlet at d 7.24 is due to the chloroform. Peaks marked with asterisks are due to contaminants from the silica layer.

group (Fig. 4). The ions at m/z 237 and 265 represent the loss of methanol (CH30H) from the m/z 269 and 297 ion frag- ments, respectively. Two other major groups of ions a t m/z 238 and 266 were assigned to acylium ions (m/z 238; OC- (CHz),3-CH(CH3),m/z266;OC-(CH2)l3-CH(CH3)CH(CH3)), corresponding to the loss of methoxy groups from the ions of m/z 269 and 297, respectively.

Fig. 3B shows the electron impact mass spectrum of the deuterium-labeled molecule obtained by methanolysis with the D-4 methanol/HCl solution (to produce the trideuterated methyl ester). The most striking change was the shift of the molecular ion by 6 mass units, which confirmed methanol

(CD30H) and a methoxy group (CD3O) from the molecular ion (544 mlz) to give ions at m/z 509 and 510 were also observed in addition to the deuterated McLafferty fragments (77 and 90 mlz). The ions at m/z 272 ((CD30CO- (CH2)&H(CH3)) and 300 (CD30CO-(CH2)&H(CH3) CH(CH3)) corresponded to deuterated m/z 269 (CH30CO- (CH&) and m/z 297 (CH30CO-(CH&CH(CH3)CH(CH3)), respectively. The ions at m/z 237 and 265 corresponded to loss of trideuterated methanol (35 mass units) from the ions at m/z 272 and 300, respectively. Most of the other unusual long chain fatty acid components gave spectra with similar properties, indicating that they had structures consistent with

by guest on March 30, 2018

http://ww

w.jbc.org/

Dow

nloaded from

by guest on March 30, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Long Chain Dicarboxylic Acids in S. ventriculi 2833

c I

Retention time

FIG. 8. Gas chromatographic analyses of lipid components extracted from cells of S. ventriculi grown at pH 7.0 in the presence of various forms of environmental stress. A, 0.25 M ethanol; B, 0.05 M butanol; C, elevated temperature (55 "C). 1 , CMO, carboxylic methyl ester, 2, C1tl, fatty aldehyde; 3, Cleo, carboxylic acid methyl ester; 4 , unknown; 5, C,,:,, carboxylic acid methyl ester; 6, Cleo, carboxylic acid methyl ester; 7, C32:o, a,w-dicarboxylic dimethyl ester (a,w-15,16-dimethyltricotanedioate dimethyl ester); 8, C34:1, a,w- dicarboxylic dimethyl ester; 9, CSe2, a,w-dicarboxylic dimethyl ester.

these acids are able to permeate through the cytoplasmic membrane and accumulate inside the cell at large ApH values and decrease the internal pH (Kell et al., 1981).

The presence of solvents would increase fluidity of the membrane and could trigger production of longer chain fatty acids to increase membrane integrity. As growth of S. uentri- culi at low pH is inhibited in the presence of elevated levels of metabolic acids, but not solvent, experiments were con- ducted at neutral pH in the presence of high levels of solvents t o determine if any changes occurred in the lipid composition of the cell. S. ventriculi grew at pH 7.0 in solvent concentra- tions as high as 0.6 M methanol, 0.5 M ethanol, 0.3 M propanol, and 0.10 M butanol. To examine the effect of solvents on the

fatty acid composition of S. ventriculi at neutral pH, the organism was grown in the presence of 0.25 M ethanol or 0.05 M butanol (concentrations at which growth was unaffected), and the fatty acid composition was determined. If the ethanol levels produced in the cells of S. ventriculi grown at pH 3 can trigger the synthesis of the unusual long chain lipid compo- nent, then this mechanism would be supported by demon- strating that the addition of exogenous ethanol to cells grown at pH 7.0 has the same effect on membrane composition as the addition of exogenous acid (i.e. growth at pH 3.0). Cells of S. ventriculi grown at pH 7.0 in the presence of ethanol or butanol (Fig. 8, A and B ) contained a family of unusual lipid components similar to those observed in cells grown at pH 3.0. These findings suggest that high levels of ethanol, bu- tanol, or protons (organic acids) can trigger the synthesis of the very long chain fatty acids and that the phenomenon is not specific.

There are numerous reports on the adaptations bacteria make in lipid composition when grown in the presence of alcohols (Baer et al., 1987; Dombek and Ingram, 1984; Ingram, 1976, 1977, 1982; LePage et al., 1987). Membrane integrity is in many cases the primary site of alcohol, in particular ethanol, damage, although alcohol clearly affects the proper- ties of all biological macromolecules to some degree. Ethanol adaptive changes have been divided into two groups: those that strengthen the hydrophobic barrier and those that de- crease the lipid sites available for passive leakage (Ingram, 1990). The former include the production of longer chain fatty acids and an increase in the proportion of nonpolar lipids in the membrane. This mechanism involves the extension of the average chain length by a few carbon chains after the pertur- bation (Ingram and Vreeland, 1980). The other group of adaptive changes include a reduction of the lipid-to-protein ratio and a reduction of the lipid pitches on the membrane surface sites available for passive leakage (Foster and Hall, 1990; Ingram, 1977).

Our results indicate a much more dramatic mechanism because the average chain length of the normal fatty acids remain the same after the addition of alcohol, and an entirely new class of very long chain membrane lipids are synthesized. Since the effect of diverse alcohols on membrane composition seems to be similar in S. uentriculi, it became apparent that the synthesis of these unusual lipid components might be a general response to environmental factors that might disrupt the membrane and reduce its fluidity. If the fluidity of the membrane were the real controlling element for this adaptive response, heat as another external perturbation would be expected to have an effect similar to that of alcohol or protons. Temperature effects on the composition of the membrane have been thoroughly studied in Escherichia coli (Ingram 1982; deMendoza and Cronan, 1983; deMendoza et al., 1983). The configuration of unsaturated carbons (cis- or trans-) or the hydrocarbon chain length of fatty acids are known to be responsible for temperature adaptive changes because config- urations and increased chain length can influence the packing density and hence fluidity. The optimum temperature for growth of S. uentriculi is 30-37 "C (Carey and Ingram, 1983). Experiments were performed under pH 7.0 controlled condi- tions to determine the effect of prolonged incubation (4 h) at 45 or 55 "C on membrane composition. Fig. 8C shows the GC profile of membrane lipid components after methanolysis of the membrane lipids of the cells of S. ventriculi grown at pH 7 after a temperature shift from 37 to 55 "C. Incubation at an elevated temperature resulted in synthesis of the long chain fatty acids even though the organism became nonviable. Cells of S. ventriculi were viable after a shift in temperature to

by guest on March 30, 2018

http://ww

w.jbc.org/

Dow

nloaded from

2834 Long Chain Dicarboxylic Acids in S. ventriculi

TABLE I1 Effect of environmental stresses on chain length and the degree of saturation of the fatty acids in Sarcina ventriculi

Fatty acids" pH 7.0 (37 "C)

pH 7.0 (0.25 M pH 7.0 (0.05 M pH 7.0 ethanol, 37 'C)

pH 7.0 butanol, 37 'C) (45 "C) (55 "C)

pH 3.0 (37 'C)

Saturated 32.7 24.6 24.6 29.7 Unsaturated 67.3

29.4 35.7 75.4 75.4 70.3

Regular (<Czo) 70.6

>93 55.2 55.2 58.2 62.0 48.9 64.3

Very long chain ( X z 8 ) <7 44.8 44.8 41.8 38.0 Ratio of U/Sb

41.1

Ratio of V/R' 3.06 2.37 2.40 1.80 0.81 0.72 0.61 1.05

2.06 3.06 0.08 0.81

a Relative amounts of unsaturated, saturated, regular (less than 20-carbon chain) or very long chain (more than 28-carbon) fatty acids were determined bv the calculation of intearated aeak areas on GC analvsis at each condition.

Unsaturated fatty acid/saturatedlatty acid. Very long chain fatty acid/regular fatty acid.

45 "C, and synthesis of the a,w-dicarboxylic very long chain acids occurred at this temperature. For the reversibility of the synthesis of these unusual membrane components, the tem- perature was shifted to 45 from 37 "C at the mid-log phase of growing cells of S. ventriculi a t neutral pH. A small amount ranging from close to 0 and rarely exceeding 7% of long chain lipids were always found in cells of S. ventriculi grown at neutral pH and probably represents the basal level of these lipids. After shifting the temperature to 45 "C, the amounts of long chain lipid components dramatically increased and returning the temperature to 37 "C reduced their proportions. In order to demonstrate that the decrease in very long chain fatty acids was due to the decrease in temperature, rather than turnover or dilution before the temperature was returned t o 37 "C, part of the culture was removed, and the incubation at 45 "C was continued. The temperature of the remainder of the culture was brought down to 37 "C, and after 90 min, the fatty acid content in the 45 "C culture was much higher than the culture a t 37 "C (data not shown), demonstrating that synthesis of these fatty acids was reversible and dependent on temperature.

E. coli compensates for an increase in growth temperature or the presence of ethanol by increasing the relative abun- dance of saturated fatty acids (Ingram, 1976; Marr and Ingra- ham, 1962; Sullivan et al., 1979). The same changes are also observed when Clostridium acetobutylicum is grown in the presence of alcohol (Baer et al., 1987; LePage et al., 1987). Our findings suggest that the ratio of saturated to unsaturated fatty acids is not decreased by increasing temperature or solvent but rather that S. ventriculi overcomes the increase in membrane fluidity by increasing the ratio of the very long chain fatty acids to regular length fatty acids (Table 11). Such a response has also been observed in C. acetobutylicum with increasing growth temperature (Baer et al., 1987; LePage et al., 1987), but the increase in chain length is not as dramatic as that observed in S. ventriculi, whereby the chain length almost doubles. In Zymomonas mobilis, longer chain fatty acids appear to be beneficial for the high ethanol tolerance exhibited by the organism (Dombek and Ingram, 1984). In 2. mobilis, the mechanism by which ethanol tolerance is achieved appears to be a concerted shift in the relative amounts of phospholipids, hopanoids, and proteins in the cell envelope. The large ethanol-dependent shifts in the hopanoid content suggest a major function of these sterol-like substances for membrane stabilization (Schmidt et al., 1986).

There have been a few other reports of bacteria producing unusually long chain fatty acids. Lactobacillus heterohiochii is a spoilage organism of Japanese wine, growing in concentra- tions of alcohol greater than 20%, possessing short chain fatty acids and considerable amounts of saturated and monounsat- urated fatty acids ranging in length from C20 to c30 (Uchida,

1974a, 1974b). Sulfate-reducing bacteria, Desulfotomaculum sp., and three strains of D. ruminis contained a-hydroxy and a,w-dicarboxylic acids with more than 20 carbons, and, in addition, very long chain fatty acids up to C34 were identified (Reznak et al., 1990). A Butyriuibrio sp. (strain S2) occurring in the rumen has an absolute requirement for fatty acid, and a new series of long chain dicarboxylic acids (C20-36) were found to be major components of the lipids (Clarke et al., 1980; Hazelwood and Dawson, 1979; Klein et al., 1979). An- other unusual C30 fatty acid is present in Clostridium ther- mohydrosulfuricum and Clostridium thermosulfurogenes, and this dicarboxylic acid accounts for about 10 and 23%, respec- tively, of the total apolar chains of these organisms (Lang- worthy and Pond, 1986). Fatty acids such as 27-hydroxyocta- cosanoic acid bipolar long chain fatty acids are thought to be key membrane components for controlling the fluidity of bacterial membranes in other systems where oxygen diffusion has to be regulated or where the organism might have to survive intracellularly within eucaryotic organisms (Bhat et al., 1991; Hollingsworth and Carlson, 1989; Hollingsworth and Lill-Elghanian, 1989).

Recently, homeoviscous adaptations in the archaebacter- ium, Methanococcus junnashcii, have been reported in which changes in the proportions of diether, macrocyclic ether, and tetraether lipids of the membrane were found at different temperatures (Sprott et al., 1991). The relative amounts of tetraether lipid (Cd0) increased with increasing temperature relative to the level of diether lipid (C20). This indicates that the transmembrane lipid components function to modulate the reduced viscosity due to increased temperature, thus main- taining the optimum fluidity.

Increased temperature, solvents, or proton levels disrupt membrane order and lead to increased molecular motion, affecting the viscosity of the membrane. This may trigger the synthesis of the very long hydrophobic lipid components that would reduce the increased viscosity. There are several clas- sical mechanisms that are known to be activated in response to these changes. The first is the heat shock gene phenomenon (Lindquist, 1986), whereby proteins are synthesized in re- sponse to various forms of stress, including heat (Neidhardt et al., 1984), ethanol (Plesset et al., 1982; Terracciano et al., 1988; VanBogelen et al., 1987), puromycin (Goff and Goldberg, 1985), anaerobiosis (Spector et al., 1986), and pH (Foster and Hall, 1990). I t is not clear to what extent the phenomenon we see in S. ventriculi is related to heat shock or stress proteins.

Our results suggest that there is a common membrane response to various environmental stresses in S. ventriculi. This response involves the synthesis of a new family of very long a,w-dicarboxylic acids that may control membrane flu- idity. From the structural data, the two polar groups on the dicarboxylic groups are present at either end, which suggests

by guest on March 30, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Long Chain Dicarboxylic Acids in S. ventriculi 2835

that this unusual lipid component probably spans the mem- brane, a position that would be more energetically favorable. The mechanism of alterations in membrane components in S. ventriculi is under study.

REFERENCES Baer, S. H., Blaschek, H. P., and Smith, T. L. (1987) Appl. Enuiron. Microbiol.

Bhat, U. R., Maye!, H., Yokota, A., Hollingsworth, R. I., and Carlson, 12. W.

Canale-Parola, E. (1970) Bacteriol. Reo. 3 4 , 82-97 (19911 J. Bactenol. 173,2155-2159

Canale-Parola, E. (1986) in Bergey's Manual o Systemattc Bacteriology (Sneath, P. H. A,, Mair, N. S., Shaarpe, M. E., and Aolt, J. G.,'eds), Vol. 2, pp. 1100- 1103, Williams & Wilkins, Baltimore

53,2854-2861

Carey, V. C., and Ingram, L. 0. (1983) J. Bacteriol. 154 , 1291-1300 Clarke, N. G., Hazelwood, G. P., and Dawson, R. M. C. (1980) Biochem. J.

Dombek, K. M., and Ingram, L. 0. (1984) J . Bacteriol. 157,233-239 Foster, J. W., and Hall, H. K. (1990) J. Bacteriol. 1 7 2 , 771-778 Goff, S. A., and Goldberg, A. L. (1985) CeU 4 1,587-595 Goodwin, S., and Zeikus, J. G. (1987) J. Bacteriol. 169 , 2150-2157 Hazelwood, G. P., and Dawson, R. M. C. (1979) J. Gen. Microbiol. 112,15-27 Hollingsworth, R. I., and Carlson, R. W. (1989) J. Biol. Chem. 264,9300-9303 Hollingsworth, R. I., and Lill-Elghanian, D. A. (1989) J. Biol. Chem. 264 ,

Holme, J. D., and Peck, H. (1983) Analytical Chemistry, pp. 436-467, Longman,

Ingram, L. 0. (1976) J. Bacteriol. 125,670-648 Ingram, L. 0. (1977) Can. J . Microbiol. 2 3 , 779-789 Ingram, L. 0. (1982) J. Bacteriol. 1 4 9 , 166-172

Ingram L. O., and Vreeland, N. S. (1980) J. Bacteriol. 144,481-488 Ingram, L. 0. (1990) Crit. Reu. Biotechnol. 9,305-319

Kell, D.'B., Peck, M. W., Rodger, G., and Morris, J. G. (1981) Biochem. Biophys.

191,555-569

14039-14042

Inc., New York

Res. Commun. 99,81-88

Klein, R. A., Hazelwood, G. P., Kemp, P., and Dawson, R. M. C. (1979) Biochem. J. 183,691-700

Langworthy, T. A., and Pond, J. L. (1986) in Thermophiles, General, Molecular, and Applred Mtcrobloology (Brock, T. D., ed), pp. 107-135, Wiley & Sons, New

LePage, C., Fayolle, F., Hermann, M., and Vandecasteele, J.-P.. (1987) J. Gen. York

Lindquist, S. (1986) Annu. Reu. Biochem. 55,1151-1191 Microbiol. 1 3 3 , 103-110

Lowe S. E., and Zeikus, J. G. (1991) Arch. Microbiol. 155 , 325-329 Lowe: S. E., Pankratz, H. S., and Zeikus, J. G. (1989) J. Bacteriol. 171 , 3775-

Marr, A. G., and Ingraham, J. L. (1962) J. Bacteriol. 84,1260-1267 deMendoza, D., and Cronan, J. E., Jr. (1983) Trends Biochem. Sci. 8,49-52 deMendoza, D., Ulrich, A. K., and Cronan, J. E., Jr. (1983) J . Biol. Chem. 2 5 8 ,

Morris, J. L. (1962) Chem. Znd. (Lord.) 1238-1240 Neidhardt, F. C., VanBogelen, R. A., and Vaughn, V. (1984) Annu. Reu. Genet.

Plesset, J., Palm, C., and McLaughlin, C. S. (1982) Biophys. Res. Commun.

Reznaka. T.. Sokolov. M. Yu.. and Viden. I. (1990) FEMS Microbiol. Ecol. 73.

3781

2098-2101

18,295-329

108,1340-1345 , , I

231-238 '

biol. Biotechnol. 2 6 , 32-36 Schmidt, A., Bringer-Meyer, S., Porolla, K., and Sahm, H. (1986) Appl. Micro-

Spector, M. P., Aliabadi, Z., Gonzalez, T., and Foster, J. W. (1986) J. Bacteriol.

Sprott, G. D., Meloche, M., and Richards, J. C. (1991) J. Bacteriol. 173,3907- 173,3907-3910

391 0 Sullivan, K. H., Hegeman, G. D., and Cordes, E. H. (1979) J. Bacteriol. 138,

Terracciano, J. S. Rapaport, E., and Kashket, E. R. (1988) Appl. Enuiron.

Tilak, K. V. B. R. (1970) Sci. Cult. 36,399-400

Uchida, K. (1974b) Biochim. Bwphys. Acta 3 6 9 , 146-155 Uchida, K. (1974a) Biochim. B(ophys. Acta 348, 86-93

VanBogelen, R. A., Kelley, P. M., and Neidhardt, F. C. (1987) J. Bacteriol.

133-138

Microbiol. 54,1989-1995

169.26-32

by guest on March 30, 2018

http://ww

w.jbc.org/

Dow

nloaded from

S Jung, S E Lowe, R I Hollingsworth and J G Zeikusdifferent forms of environmental stress.

Sarcina ventriculi synthesizes very long chain dicarboxylic acids in response to

1993, 268:2828-2835.J. Biol. Chem. 

  http://www.jbc.org/content/268/4/2828Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/268/4/2828.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on March 30, 2018

http://ww

w.jbc.org/

Dow

nloaded from