THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 265, No. 28, Issue of October 5, pp. 17215-17221. 1990 Q 1990 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A.

Characterization of Escherichia coli Cells Deficient in 1-Acyl-sn- glycerol-3-phosphate Acyltransferase Activity*

(Received for publication, February 23, 1990)

Jack Coleman From the Department of Biochemistry, College of Agricultural and Life Sciences, University of Wisconsin, Madison, Wisconsin 53706 and the $Department of Biochemistry and Molecular Biology, Louisiana State University Medical Center, New Orleans, Louisiana 701 I2

A mutant of Escherichia coli K-12 defective in l- acyl-sn-glycerol-3-phosphate acyltransferase has been isolated. At the permissive temperature for growth, 30 “C, 20% of the total cellular glycerophospholipids is 1-acyl-sn-glycerol-3-phosphate, as identified by mass spectral analysis and proton NMR. This percent- age of 1-acyl-sn-glycerol-3-phosphate rises to about 30% when the temperature of the culture is shifted to 42 “C!. This increase is primarily at the expense of phosphatidylethanolamine. Extracts from cells har- boring the plsC mutation have no detectable l-acyl-sn- glycerol-3-phosphate acyltransferase activity. The fatty acid composition of the accumulated l-acyl-sn- glycerol-3-phosphate is about 60% cis-vaccenate and 40% palmitate, with no detectable amounts of palmi- toleate or other fatty acids, consistent with the known fatty acid composition of the sn-1 position of glycero- phospholipids.

The isolation of this gene, plsC, completes the list of genes known to be required for the synthesis of the major glycerophospholipids in E. coli.

Escherchia coli phospholipid biosynthesis mutations have been isolated and characterized in nearly every step in the pathways. These mutations are proving invaluable in the study of phospholipid biosynthesis, including the regulation of head group composition and fatty acid composition.

Escherichia coli carefully regulates the composition of its membrane glycerophospholipids. To be a biologically func- tional membrane, the membrane phospholipids must be at the boundary of the crystalline-fluid phase transition. The membrane is nonfunctional if all the lipids are in the ordered state or if they are all in the disordered state. There is a wide variation in fluidity that is tolerated; however, there does seem to be an optimal fluidity at which cell growth is more rapid (Cronan and Rock, 1987).

The tight mutations in phospholipid biosynthetic enzymes, in most cases, have resulted in cell death or arrest of cell growth (Cronan and Rock, 1987). Significant changes in phos- pholipid composition, in most cases, cause marked changes in the physiology of the cell.

In the present study, I characterize the effects of a mutation in the gene thought to encode the enzyme responsible for the

*This investigation was supported in part by Grant DK-19551 from the National Institutes of Health (to C. R. H. Raetz). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “adver- tisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$ Address to which correspondence and reprint requests should be sent. Tel.: 504-568-4388.

second step in phospholipid biosynthesis, l-acyl-sn-glycerol- S-phosphate acyltransferase (1-acyl-G-3-P’ acyltransferase). This enzyme can utilize either acyl-CoA or acyl-acyl carrier protein to donate a fatty acid to 1-acyl-G-3-P (Rock et al., 1981). This enzyme is thought to determine the temperature- dependent incorporation of saturated and unsaturated fatty acids at the sn-2 position of glycerophospholipids. Unlike the sn-1 fatty acid which is used as a fatty acid donor for lipopro- teins (Jackowski and Rock, 1986) and lipopolysaccharide (Brozek et al., 1987), the sn-2 position fatty acid is stable, and there are no other known mechanisms for fatty acid addition to the sn-2 position, as there are for the sn-1 position (Homma et al., 1981). Although much work has been done to charac- terize the sn-glycerol-3-phosphate acyltransferase and other phospholipid biosynthetic enzymes (Scheideler and Bell, 1989; Ray et al., 1970; Cronan and Rock, 1987), very little work has been done to characterize 1-acyl-G-3-P acyltransferase. The isolation of a mutation in the gene for 1-acyl-G-3-P acyltrans- ferase, p=!.sC, will enable us to better understand the functions of this enzyme and its involvement in the control of fatty acid composition.

EXPERIMENTAL PROCEDURES

Muterials-[3’P]Orthophosphate was obtained from Du Pont-New England Nuclear. Palmitoyl-CoA, fatty acid-free bovine serum albu- min, and phospholipids were from Sigma. Silica Gel 60 plates (0.25 mm) were purchased from E. Merck, Darmstadt, Germany. Silica Gel was Bio-Sil A from Bio-Rad. Yeast extract and tryptone used in LB medium were obtained from Difco. Biosafe II liquid scintillation fluid was from Research Products International Corp. Bicinchoninic acid (BCA) was from Pierce Chemical Co.

Growth of Bacteria-Bacterial strains used are listed in Table I. Bacteria were grown in LB or M9 medium (Miller, 1972), supple- mented with tetracycline (30 pg/ml) or streptomycin (100 rg/ml).

Gene Mapping-PI transduction (Silhavy et al., 1984), Hfr and F’ matings (Miller, 1972), and ColEl matings (Raetz et al., 1977) were done as previously described.

Quantitution of Lipopolysaccharide and Phospholipid S.ynthesis- Lipopolysaccharide and glycerophospholipid biosyntheses were quan- titated as described (Galloway and Raetz, 1990).

Isolation of Unknown Lipid-To isolate sufficient quantities of the unknown lipid for NMR analysis and protein assays, strain 2-1 was grown at 30 “C in 1 liter of LB medium until late log phase of growth. This culture was added to an equal volume of LB at 30 “C. The culture was then transferred to a 42 “C water bath, and growth was continued for 2 h. Cells were harvested by centrifugation at 5000 x g., for 20 min. Total phospholipids were extracted under acidic Bligh- Dyer (Nishijima and Raetz, 1979) conditions as described earlier (Brozek et al., 1989). Pyridine was added at %o volume, and the solvents were removed by rotary evaporation at 30 “C. The residue was redissolved in 10 ml of chloroform:methanol:water (26:lO:l) and applied to a 20-ml silicic acid column (Bio-Sil A) equilibrated in the

1 The abbreviations used are: l-acyl-G-3-P, l-acyl-sn-glycerol-3- phosphate; PG, phosphatidylglycerol; PE, phosphatidylethanolamine.

17215

by guest on September 4, 2020

http://ww

w.jbc.org/

Dow

nloaded from

17216

Strain or plasmid

1 -Acyl-sn-glycerol-3-phosphate Acyltransferase Mutant

TABLE I Strains and plasmids

Genotype Reference or source

E. coli SM105

2-1 CS1562 SO1023 NK6027 2-1tet

SM2-1

JC201

JC200

JA200

Plasmids pLC37-24 pLC4-14

thr-1 ara-14 A(gal-atth) hisG4 rpsL136 ~~1-5 mtl-1 lacy1 k-78 eda- rfbD1 thi-1

(Galloway and Raetz, 1990)

SM105 nitrosoguanidine-treated (pl.sC serA) tolC::TnlO HfrPOl nupG5ll::TnlO reL4 serA6 spoTI thi-l HfrPOl metCl62::TnlO A(gpt-lac)5 rel41 spoT1 thi-1 plsC metCl6S::TnlO (derived from 2-1 by Pl transduction using

NK6027 as donor)

This work CGSC CGSC CGSC This work

pl.sC metCl6S::TnlO serA+ (derived from SM105 by Pl transduction using 2-ltet as donor)

p/sC (derived from SM2-1 by Pl transduction using SO1023 as donor, selecting met+ tets)

pl.0 (derived from SM2-1 by Pl transduction using SO1023 as do- nor, selecting met+ tets)

F+ recA thr-1 leuB6 trpE5 recA56 thi-1 ara-14 lacy1 galK2 galT22 xyl- 5 mtl-1 supE44 (host strain for Clarke and Carbon (1976) plasmids)

This work

This work

This work

CGSC

metC+ plsC- colicin Els CGSC metC+ plsC’ colicin Els CGSC

’ Coli Genetic Stock Center.

same buffer. The column was chromatographed using the same sol- vent. Fractions (5 ml) containing the unknown lipid (detected by thin layer chromatography and sulfuric acid charring) were pooled and dried by rotary evaporation. The residue was redissolved in chloro- form:methanol:(28%) ammonia (13:5:1) and applied to a lo-ml silicic acid column (Bio-Sil A) equilibrated in the same buffer. The column was washed with 75 ml of chloroform:methanol:(28%) ammonia. The unknown lipid was eluted with chloroform:methanol:acetic acidwater (65:25:4:2). Fractions (5 ml) containing the unknown lipid were pooled, pyridine was added at */LO volume, and the lipid was dried at 30 “C in a rotary evaporator. The lipid isolated was pure as judged by charring on TLC plates developed in many different solvent systems, NMR spectrometry, and mass spectral analysis.

“‘P-Unknown lipid was isolated by pulse-labeling cells for 5 min after 1 h at 42 “C. Lipids were isolated under acidic Bligh-Dyer conditions as described (Nishijima and Raetz, 1979). Lipids were separated on Silica Gel 60 TLC plates developed in chloro- form:methanol:acetic acid:water (65:25:4:2). The position of the un- known lipid was determined by autoradiography, and the portion of the plate containing the unknown lipid was scraped off and suspended in chloroform:methanol:acetic acidwater (65:25:4:2). Silica gel was removed by filtration through glass wool, and residual lipid was washed off with the same buffer. Pyridine was added at i/20 volume, and the lipids were dried under nitrogen.

Proton NMR--‘H NMR spectra were obtained using 1 mg of unknown lipid, prepared as the pyridinium salt and dissolved in CDCL:CD:,OD (2:l or 4:1, v/v). Suectra were recorded on a Bruker AM 500 spectrometer operating at 500.13 MHz at either 25 “C or 37 “C. The spectrometer was equipped with an Aspect 3000 computer and digital phase shifters. Each spectrum was an average of 75 scans.

FAB Mass Spectrometry-The sample was analyzed at the Univer- sity of Wisconsin-Madison Biochemistry Department Mass Spec- trometry Laboratory on a Kratos Ultrahigh Resolution MS50TC spectrometer, equipped with an Ion Tech FAB gun. Approximately 15 rg of the unknown lipid was applied to the probe in a glycerol matrix made slightly basic with NaOH. The sample was bombarded with xenon gas from the FAB gun, which was operated at about 6 kV. The computer system used to plot the spectrum drops the digits after the decimal point rather than rounding off to the nearest whole mass unit.

Preparation of Cell Extracts-Cells grown at 28 “C, were harvested in late log phase, resuspended in 50 mM Tris-HCl, pH 8.0, and disrupted using a French Pressure cell at 18,000 p.s.i. The suspension was centrifuged at 8,000 X g., for 15 min to remove unbroken cells. The low speed supernatant was centrifuged at 105,000 x g,, for 90 min. The particulate fraction (membranes) was resuspended in 50 mM Tris-HCl, pH 8.0. All steps were performed at 4 “C. Samples were quick frozen in liquid nitrogen and stored at -70 “C. Protein concentration was determined by the BCA method of Smith et al. (1985).

l-Acyl-G-3-P Acyltransferase Assay-The assay procedure for l-

acyl-G-3-P acyltransferase was essentially as described by Ray et al. (1970) with minor modifications. The reaction mixture consisted of [“‘PI-1-acyl-G-3-P, 10,000 cpm (concentration is negligible); l-acvl- - - G-3-P, 30 PM (isolated from strain 2-l as described above); palmitoyl- CoA. 40 uM: Tris-HCl. DH 9.0. 100 mM: M&l.,. 0.5 mM: fattv acid- free bovine serum album&, 1 m&ml; and’par&late enzyme fraction, 0.17 mg/ml. The reaction is initiated by the addition of the enzyme fraction and continued at 30 “C for 2 and 4 min. The reaction is terminated by spotting a portion of the reaction onto a silica gel TLC plate and developed in chloroform:methanol:acetic acidwater (25:15:4:2). The substrate, l-acyl-G-3-P, and the product, phospha- tidic acid, were located by autoradiography, scraped from the plate, and counted in scintillation fluid.

Phospholipid Analysis-Cells were labeled with 32Pi in LB medium as indicated. Incorporation of label was terminated by addition of 0.8 ml of culture to 3.0 ml of chloroform:methanol (1:2, v/v). The debris was removed by centrifugation, and the glycerophospholipids were extracted by the acidic two-phase partition of Bligh and Dryer (1959) as modified bv Nishiiima and Raetz (1979). A portion of the extract was analyzed by two-dimensional Silica Gel 60 thin layer chromatog- raphy (TLC) developed first in chloroform:methanol:water (65:25:4), dried, then developed in chloroform:methanol:acetic acid (65:25:10). The positions of the phospholipids were identified by autoradiogra- phy, the radioactive spots were scraped from the plate, and the radioactivity was determined by scintillation counting in Biosafe II scintillation mixture.

RESULTS

Isolation of Membrane Defect Mutants-In order to obtain a mutant defective in membrane biosynthesis, E. coli strain SM105 was treated with nitrosoguanidine, and membrane synthesis mutants were selected by the method of Tsuruoka et al. (1988), by sedimentation through sucrose. Of the colo- nies isolated, 7% (18 of 257) were temperature-sensitive for growth, and 72% (13 of 18) of the temperature-sensitive colonies were sensitive to 0.2% deoxycholate in the growth medium. A disruption of the outer membrane of Gram-nega- tive bacteria is often indicated by a sensitivity to deoxycholate (Raetz and Foulds, 1977; Misra and Benson, 1988; Hirota et al., 1977). One temperature-sensitive, deoxycholate-resistant strain, 2-1, was selected for further study.

Initial Characterization of the Mutant Strain-E. coli strain 2-l carries a mutation, designatedplsc, which causes the cells to accumulate on unidentified phospholipid, especially at 42 “C. Mutant (2-l) and wild-type (SM105) cells were pulse- labeled with [“‘Plorthophosphate to determine if this lipid is a normal precursor of the glycerophospholipids and not re-

by guest on September 4, 2020

http://ww

w.jbc.org/

Dow

nloaded from

1 -Acyl-sn-glycerol-3-phosphate Acyltransferase Mutant 17217

TABLE II Three-factor cross ofplsC, metC, and tolC or plsC, metC, and nupG

Phage lysates prepared on the donor strains were used to transduce the recipient strain, SM2-1 (&C, metC:TnlO). Selection was for met+.

Distribution of markers DOnOr

Phenotype Number Percent

SO1023 (nupG) met+ nupG pLsC 10 6 met+ nupGpkC+ 9 6 met+ nupG+ pl.sC 30 19 met+ nupG+ pl.0 113 70

CS1562 (tolC) met+ tolC plsC 0 0 met+ tolC pl.sC+ 14 33 met+ tolC+ pkC 14 33 met+ tolC+ plsC+ 14 33

nupC mew plsC suf1-pare tOIC

h4' 65' 66'

FIG. 1. Genetic map of E. coli K12 (64- to 66-min segment). Locations are calculated by the transduction frequencies shown in Table II and from Kato et al. (1988).

lated to lipopolysaccharide or other biosyntheses. Phospho- lipids and lipopolysaccharides were isolated after a 5-min pulse with 32Pi after 1 h of growth at 42 “C as described under “Experimental Procedures.” The ratio of the rate of synthesis of glycerophospholipids to the rate of synthesis of lipopoly- saccharide was 7.7 for both 2-l and SM105, if synthesis of the unknown lipid was included with the glycerophospho- lipids. If the unknown lipid is ignored, the ratio of the rate of synthesis of glycerophospholipids to the rate of synthesis of lipopolysaccharide in the p/.sC mutant strain 2-1 is 3.4, and the ratio in SM105 is still 7.7. These results suggest that the unknown lipid is either a precursor or a byproduct of glycer- ophospholipid biosynthesis.

Genetic Mapping of the plsC Mutation-Hfr mapping and F’ complementation revealed the mutation to map near 65 min on the chromosome (data not shown). A two-factor cross between 2-l (p&Z) and NK6027 (metC, 65 min) or SO1023 (nupC; 64 min)’ using Pl transduction indicates that plsC maps very close to metC (71% co-transduction).

Fine mapping of plsC was accomplished using three-factor crosses with p&Z’, metC (65 min), and tolC (66 min) or nupG (64 min). Table II shows the results of these three-factor crosses. In the cross between SM2-1 (plsC metC) and CS1562 (tolC), the lack of transductants that got metC+ and tolC-, but did not get pLsC’, indicates that pLsC is located between metC and tolC. Using the equation described by Wu (1966) relating co-transduction frequencies and map distance, I have local- ized the genes as described in Fig. 1, placing pl.sC slightly closer to metC than tolC. Two of the resulting strains from the SO1023 X SM2-1 cross were isolated for further study, JC201 (pl.sC metC+ nupG+), and JC200 (p&T+ metC+ nupG+).

Conformation of this localization was obtained by identifi- cation of plsC and metC on a single cloned fragment of DNA in the Clarke and Carbon (1976) DNA library. Three plasmids from the Clarke and Carbon (1976) collection have been identified to contain metC (Phillips et al., 1987), pLC4-13, pLC4-14, and pLC37-24. pLC4-14 and pLC37-24 were tested for their ability to complement the plsC mutation in JC201. The plasmids were introduced into JC201 in a 30-min mating selecting for resistance to colicin El and streptomycin. JC201 harboring pLC4-14, but not JC201 harboring pLC37-24, was able to grow at 42 “C. pLC4-14 and not pLC37-24 also restored the ability of JC201 to synthesize a normal complement of

’ B. Bachmann, persona1 communication.

lipids. The lipid composition of JC201 harboring pLC4-14 was indistinguishable from the isogenic wild-type strain JC200 (data not shown). Kato et al. (1988) have shown that pLC4-14 also complements pa& and sufl. pa& and sufr appear to be in an operon that is localized to one end of the chromosomal fragment of DNA in pLC4-14 (Kato et al., 1988). Taken together with the present data, these results indicate that pLsC is located between metC and sufl-pa&.

A comparison of the restriction map data and known gene locations of suf1 and metC (Kato et al., 1988; Belfaiza et al., 1986) with the physical map of the entire E. coli chromosome reveals that plsC is located between 3208 and 3220 kilobases on the Kohara (Kohara et al., 1987) map of the E. coli chromosome.

‘H NMR Spectroscopy of the Unknown Lipid-As a first step in determining the structure of the unknown lipid, a l- mg sample (prepared on a large scale as described under “Experimental Procedures”) was subjected to ‘H NMR analy- sis at 500 MHz (Fig. 2). The lipid was prepared as a pyridinium salt, and it was dissolved in CDC13:CD30D (2:1, v/v). The spectra were run at 20 “C. As expected, several characteristic features of a glycerophospholipid spectrum were identified in the spectrum of the unknown (Fig. 2). These features include the methylene groups of the fatty acyl chain at 1.25 ppm, the protons of the terminal methyl groups resonating at 0.8 ppm, and the protons at the sn-1 position of a glycerol of a l-acyl- glycerophospholipid at 4.1 ppm (if a fatty acid is attached to the sn-2 position, a peak would be expected at about 5.1 ppm (Kates, 1986)). The triplet at 2.3 ppm is characteristic of protons on carbon 2 of fatty acid, and the resonance at 1.5 ppm is a feature of fatty acyl protons on the third carbon. Indeed, if the resonance at 1.5 ppm is selectively decoupled, the triplet of 2.3 ppm collapses to a singlet, as expected for a fatty acid (data not shown).

The triplet at 5.3 ppm is characteristic of protons attached to unsaturated carbons in a fatty acid. The methylene protons adjacent to the double-bonded carbons would be expected to resonate at approximately 1.9 ppm (Kates, 1986). As can be

L? L L

F 2 G OS g *

ri 9

d0 5.0 4.0 3.0 21) 1.0 - -7.0 PPrn

FIG. 2. ‘H NMR analysis of the unknown lipid. The spectrum of 1 mg of the unknown lipid, prepared as described under “Experi- mental Procedures,” was recorded in 0.5 ml of CDC1,:CDsOD (2:1, v/ v). The chemical shifts are expressed as parts/million relative to an internal standard of tetramethylsilane. The assignments are based on published analysis of known lipids (Kates, 1986) and selective decoupling described in the text. C represents protons on cis-vaccen- ate, P denotes protons attached to palmitate, and G indicates protons attached to the glycerol backbone. Numbering of the carbons on the fatty acids initiates with the carbon attached to the glycerol.

by guest on September 4, 2020

http://ww

w.jbc.org/

Dow

nloaded from

17218 1 -Acyl-sn-glycerol-3-phosphate Acyltransferase Mutant

seen in Fig. 2, there is a resonance at this position. As expected, when the resonance at 1.9 ppm was selectively decoupled, the triplet at 5.3 collapsed to a singlet (data not shown).

To determine if the unknown compound resonates in the region of 4.3 ppm, which is obscured by the Ha0 peak in Fig. 2, the spectra of the unknown was analyzed in CDCls:CD30D (4:1, v/v) at 37 “C, which shifted the water peak to 4.0 ppm. No additional resonances were observed in this spectrum (data not shown).

FAB Mass Spectrometry of the Unknown Lipid-To better identify the unknown lipid produced by strain 2-1, the purified lipid (see “Experimental Procedures”) was subjected to fast atom bombardment mass spectrophotometry in the negative mode (Fig. 3). The two apparent mass peaks (M - H)-, differing by 26 mass units, at m/z 409 + 1 and 435 f 1 (Fig. 3) indicate that the sample contains two species. These values of (M - H)-, m/z of 435 f 1 and 409 f 1, are consistent with 1-acyl-G-3-P containing cis-vaccenate or palmitate, respec- tively. The expected atomic mass (M - H)- of l-cis-vaccenoyl- sn-glycerol-3-phosphate is 435.52, and the expected atomic mass (M - H)- of l-palmitoyl-an-glycerol-3-phosphate is 409.48.

Cronan and Vagelos (1972) have shown that in normal E. cob, that palmitic acid is found almost exclusively at position 1 of glycerophospholipids, cis-vaccenic acid is found at posi- tions 1 and 2, and palmitoleic acid is found almost exclusively at position 2. As expected for a strain of E. coli defective in the ability to add a fatty acid to position 2 of an-glycerol-3- phosphate, only trace amounts of l-palmitoleoyl-sn-glycerol- 3-phosphate, atomic mass (M - H)- of 407.47, is visible by mass spectral analysis (Fig. 3). Taken together with the results of the proton NMR, the unknown lipid can be identified as being primarily l-acyl-G-3-P containing palmitate or cis- vaccenate as the fatty acid. Both the mass spectral data and the NMR spectra indicate that about 60% of the l-acyl-G-3- P molecules contain cis-vaccenate.

In the pathway of synthesis of the major phospholipids in E. coli (Cronan and Rock, 1987), l-acyl-G-3-P is the substrate for the second reaction, the addition of a fatty acid to l-acyl- G-3-P. This reaction is catalyzed by 1-acyl-G-3-P acyltrans- ferase.

Determining the Actiuity of 1 -Acyl-G-3-P Acyltransferase- To demonstrate that the defect in strain 2-l is due at least in part to 1-acyl-G-3-P acyltransferase, the activity of the acyl- transferase was quantitated in vitro as described under “Ex- perimental Procedures.” Membrane extracts were isolated from strains grown at the permissive temperature (28 “C). For membrane extracts of wild-type strain SM105, the assay was linear with both time and protein concentration, in the con- centrations and times used. Fig. 4 (lane 3) demonstrates that even if the concentration of protein in the extract is increased

100 435

80

FIG. 3. Fast bombardment mass spectrum of the unknown lipid in the negative mode. The samples was suspended in glycerol made partially alkaline by the addition of a small amount of NaOH. Numbers refer to m/z of the fragments obtained.

123

PA- l

LPA- Ir 0 @ FIG. 4. In vitro assay of 1-acyl-G-3-P acyltransferase in

pZsC+ and plsC cells. The acyltransferase assay was performed as described under “Experimental Procedures,” except that the concen- tration of l-acyl-G-3-P was reduced to increase the specific activity of l-acyl-G-3-P 6-fold, to help identify any conversion using extracts from p/&I! mutant cells. The labeled lipids were separated on silica gel TLC developed with chloroform:methanol:acetic acidwater (25:15:4:2). The phospholipids were visualized by autoradiography, which was carried out for 12 h at -70 “C with Kodak XAR film. Lane 1, complete reaction mix using extract (1.7 mg/ml final protein concentration) from SM105 (p&Y). Lane 2, reaction mix using extract from SM105, in the absence of added acyl donor palmitoyl- CoA. Lane 3, complete reaction mix using extract (1.7 mg/ml final protein concentration) from 2-l (&C). LPA and PA denote the substrate 1-acyl-G-3-P and the product phosphatidic acid, respec- tively.

A.

B

3 ‘2 7

Ii 6

-50 0 on 100 Loo 200 2.50 300 350 100 Time (min after shift to 42’)

60 100 ml 204 a50 300 360 100 ‘kme (min after shift to 42’)

FIG. 5. Growth and viability of JC200 and JC201 after shift to 42 “C. Fresh overnight cultures of JC200 (&C+) and JC201 (plsC) were diluted into LB medium and grown at 30 “C for at least three generations. The cultures were then split, and half was grown at 30 ‘C and half was grown at 42 “C. A, the optical density of the culture was determined using a Klett-Summerson calorimeter. B, viability was determined by counting the number of colonies which were formed on LB agar after plating an aliquot of the culture at 30 “C. Each point represents an average of two determinations. Tri- angles represent cultures of JC200 and circles represent cultures of JC201. Filled in symbols and the solid lines indicate cultures grown at 42 “C. Open symbols and the dashed lines denote cultures grown at 30 “C.

by guest on September 4, 2020

http://ww

w.jbc.org/

Dow

nloaded from

I-Acyl-sn-glycerol-3-phosphate Acyltransferase Mutant 17219

a b

PE r

Qe

PE ,

a

PE ,

w

1 7

b

PE I

Qla,

FIG. 6. Analysis of phospholipids produced by SM105 and 2-l. The lipids were isolated as described under “Experimental Procedures” and separated using two-dimensional TLC. The first dimension (ascending) was developed using chloroform: methanol:water (65:25:4), and the second dimension (right to left) was developed using chloroform:methanol:acetic acid (65:25:10). The phospholipids were visualized by autoradiography, which was carried out for 12 h at -70 “C with Kodak XAR film. A, cells, growing in LB medium, were pulse-labeled with “‘Pi (50 &i/ml) for 5 min after a 1 h shift to 42 “C. Panel a, SM105 (pkC+). Panel b, 2-l (plsc). B, cells were continuously labeled with 82Pi. Cells were grown at 30 “C overnight with 32Pi in LB medium, then diluted in the same labeled medium and grown for several generations at 30 “C, followed by a 2-h shift to 42 “C. PE, phosphati- dylethanolamine; PC, phosphatidylglycerol; CL, cardiolipin; PA, phosphatidic acid; LPA, 1-acyl-G-3-P; ?, unknown hydrophobic lipid.

TABLE III

Phospholipid composition and rate of synthesis of SMl05 (plsC+) and 2-1 (pLsC)

Unidentified radioactive material at the origin was not counted in determining the total phospholipid, which was taken as the sum of PE, PG, 1-acyl-G-3-P, and all other minor phospholipid (including cardiolipin). The induced hydrophobic lipid is included only in the steady state compositional analysis, because it was not detected in the rate of synthesis analysis. The steady state compositional analysis was determined after growth at 30 “C, followed by a shift to 42 “C for 2 h.

Strain PE PG 1-Acyl-G-3-P Other Hydrophobic

% total Rate of synthesis

SM105 30 “C 62.0 19.5 0.2 18.3 SM105 42 “C 67.0 22.1 0.2 10.6 2-l 30 “C 36.3 9.9 37.5 16.3 2-l 42 “C 0.3 5.5 75.7 18.5

Steady state composition SM105 70.5 22.1 0.3 6.9 0.1 2-l 32.1 16.2 29.8 20.0 1.9

lo-fold and the specific activity of the 1-[32P]acyl-G-3-P is increased B-fold, extracts from 2-1 do not catalyze the addition of a palmitate from palmitoyl-CoA to l-acyl-G-3-P. This demonstrates a greater than loo-fold reduction in l-acyl-G- 3-P acyltransferase activity in strain 2-l as compared to SM105. No activity could be detected from extracts of strain 2-l under all conditions tested, including longer incubations and lower temperatures. The specific activity of l-acyl-G-3-P acyltransferase in the wild-type cells was 5 nmol/min/mg of protein, in good agreement with previously published values (Ray et al., 1970). At the concentration of extract shown in Fig. 4, greater than 80% of the l-acyl-G-3-P is converted to phosphatidic acid using the wild-type extract of SM105 (Fig. 4, lane I). There is no conversion in the absence of acyl donor

(Fig. 4, lane 2). Similar results were obtained with the isogenic pair JC200 (pkC+) and JC201 (pkC). In addition, if JC201 harbors the plasmid pLC4-14 (pkC+), the specific activity of the l-acyl-G-3-P acyltransferase increases to greater than 5- fold over wild-type activity (>26 nmol/min/mg of protein).

Growth and Viability of the plsC Mutant-To determine whether the plsC mutation caused cell death under nonper- missive conditions, or simply arrested growth of the cells, JC201 (p/&F) and its isogenic wild-type strain JC200, growing exponentially at 30 “C in LB medium, were shifted to 42 “C. The optical density of the culture, as measured by a Klett- Summerson calorimeter, plateaued 90 min after the shift and remained level for at least 7 additional h (Fig. 5A). However, cell viability immediately plateaued for about 5 h, at which point cell viability decreased dramatically, as judged by plat- ing efficiency on LB agar at 30 “C (Fig. 5B). At 30 “C, the growth rate of the pLsC mutant cells is approximately 60% the rate of the isogenic wild-type cells, It should be noted that even at 30 “C as the pkiC strain starts to reach stationary phase, the cell viability decreases.

When cells are observed under the light microscope, there is no visible change in cell morphology of the p.!sC cells (JC201) for 2 h after shift to 42 “C. After 2 h at 42 “C, some filaments do start to form. This phenotype of formation of filaments under nonpermissive growth conditions is common to phospholipid biosynthetic mutants (Raetz, 1976; Hawrot and Kennedy, 1978). At no time is there evidence of cell lysis in the pkC mutant cells (data not shown).

Phospholipid Analysis of pLsC Cells under Different Growth Conditions-The isolation of a severe mutation in l-acyl-G- 3-P acyltransferase raised several important questions. Are cells able to survive with significant amounts of l-acyl-G-3-P in their membrane? As the concentration of l-acyl-G-3-P in the membrane is increased, do cells compensate for this by a

by guest on September 4, 2020

http://ww

w.jbc.org/

Dow

nloaded from

17220 1 -Acyl-sn-glycerol-3-phosphate Acyltransferase Mutant

change in the composition of other phospholipids synthe- sized? Is the synthesis of other nonstandard phospholipids stimulated by the accumulation of 1-acyl-G-3-P?

TO answer these questions, the phospholipid composition of the cells was analyzed after separation of the lipids on two- dimensional thin layer chromatography (TLC). To determine the rate of synthesis of each of the phospholipids, cells were pulse-labeled with 32Pi for 5 min either at 30 “C or after 1 h at 42 “C. Phospholipids were extracted and separated as de- scribed under “Experimental Procedures.” Fig. 6A shows a typical aut.oradiogram of phospholipids labeled at 42 “C and separated on two-dimensional TLC. As can be seen in panel b, pkC mutant cells (2-l) produce a large amount of I-acyl- G-3-P, and the rate of incorporation of phosphate into phos- phatidylethanolamine (PE) is decreased dramatically when compared to phosphatidylglycerol (PG). Quantitation of the radioactive spots is shown in Table III. As can be seen in Table III, 37.5% of the phosphate in phospholipids synthe- sized at 30 “C in strain 2-l is 1-acyl-G-3-P. However, the ratio of PE to PG remains at near normal levels. This accumulation of 1-acyl-G-3-P at 30 “C reflects the lack of in vitro activity of the l-acyl-G-3-P acyltransferase even at low temperatures, mentioned above. At 42 “C, the amount of 1-acyl-G-3-P syn- thesized increases to 75.7% of the total phosphate incorpo- rated into phospholipid, with a dramatic decrease in PE biosynthesis. It should be noted that the rate of synthesis of PG also decreases, but not to the same extent.

The steady state phospholipid compositional analysis of the pkC strain 2-l was accomplished with a steady state labeling of cells with 32Pi. Cells were grown for several hours at 30 “C!, then shifted to 42 “C for 2 h. Phospholipid analysis was done as described under “Experimental Procedures.” Wild-type cells (SM105) accumulated nearly undetectable amounts of l-acyl-G-3-P; however, the mutant cells produced a large amount (29.8% of the total cellular phospholipid). The ratio of PE to PG also decreases when the steady state lipids are measured. As can be seen in Fig. 6B in addition to l-acyl-G- 3-P, a new very hydrophobic lipid is detected in the pLsC strain that is only found in very small amounts in the wild- type strain. This new lipid was not detected in pulse-labeled cells, either at 30 “C or 42 “C. Similar results were obtained using the isogenic pair JC200 (pLsC’) and JC201 (plsC-).

DISCUSSION

Previously, mutations had been obtained in each step of synthesis of the major phospholipids except for the second step, the acylation of I-acyl-G-3-P to form phosphatidic acid. Studies using these mutations have greatly increased our understanding of how cells regulate the composition of their membrane lipids. These mutant strains have also aided in our understanding of the effects of certain membrane perturba- tions. The identification of the gene for 1-acyl-G-3-P acyl- transferase in this work completes the list of genes known to be required for the synthesis of the major phospholipids of E. coli.

The mutation in the plsC gene slows down the synthesis of PG, PE, and cardiolipin, particularly at the nonpermissive temperature for growth (42 “C). As synthesis of the more common phospholipids ceases, there is an increase in concen- tration, up to 30% of the total lipids, of a new lipid identified by mass spectral and NMR analysis to be 1-acyl-G-3-P. This high concentration of a precursor lipid at the time of cell death is similar to the accumulation of phosphatidic acid in the CDP-diglyceride synthetase mutant (c&8; Ganong and Raetz, 1982). Also like the cds mutation, the pt.& mutation causes an increase in the apparent rate of synthesis of PG

relative to PE, although this effect is much more pronounced in the pZsC mutation. Unlike the cds mutant, this apparent increase in the concentration of PG relative to PE is observed in the final lipid composition at the time of cell death. This apparent increase in the rate of synthesis of PG may reflect the greater turnover of the glycerol head group, which is used in the synthesis of membrane-derived oligosaccharides (vanGolde et al., 1973) and lipoproteins (Wu et al., 1983). The cause of the chemical accumulation of PG is unknown, but may reflect a control by the cell, triggered by the large accumulation of the detergent-like molecule 1-acyl-G-3-P. Ballesta and Schaechter (1971) in their studies on the syn- thesis of phospholipids in E. coli after the removal of nutrients from the growth medium concluded that synthesis of PE is increased under conditions of increased membrane synthesis, and the apparent increased rate of synthesis of PG is not related to growth and cell division, but presumably to other processes that are carried out by nongrowing metabolically active cells. Because the pLsC cells are no longer growing at 42 “C, this may explain the increase in the ratio of PG to PE. However, Ballesta and Schaechter (1971) only examined the rate of incorporation of a precursor into the phospholipids, which does not necessarily reflect the actual lipid concentra- tion.

The loss in activity of 1-acyl-G-3-P acyltransferase appears to be immediate after the shift to 42 “C, as evidenced by the lack of increase in optical density within 1 h of the shift. Although the cells no longer continue to grow, they are still viable for 3 h after the shift to 42 “C. The fact that the cells are still viable 2 h after the shift to the nonpermissive tem- perature is unexpected because at this point 30% of the total phospholipids present is the detergent I-acyl-G-3-P.

The isolation of a conditional lethal mutation in. 1-acyl-G- 3-P acyltransferase confirms the assumption that this enzyme is distinct from the sn-glycerol-3-phosphate acyltransferase responsible for the addition of a fatty acid to the sn-1 position of a phospholipid. Cessation of growth in this mutant could result from the inability of the cell to make lipids, preventing membrane growth. Alternatively, a change in the composition of the membrane may render an essential membrane protein inactive. On the other hand, the accumulation of 1-acyl-G-3- P, a detergent-like molecule, may itself adversely affect mem- brane function and prove lethal to the cell. The resolution of such questions requires a detailed study of membrane function under conditions in which the lipid composition is carefully regulated. This mutation should prove useful in the study of the regulation of head group composition and fatty acid composition.

REFERENCES

Ballesta, J. P. G., and Schaechter, M. (1971) J. Bacterial. 107, 251- 258

Belfaiza, J., Parsot, C., Martel, A., Bouthier de la Tour, C., Margarita, D.. Cohen. G. N.. and Saint-Girons. I. (1986) Proc. Natl. Acad. Sci. U:S. A. 8&867&71

Bligh, E. G., and Dyer, W. J. (1959) Can. J. Biochem. Physiol. 8,911- 917

Brozek, K. A., Bulawa, C. E., and Raetz, C. R. H. (1987) J. Biol. Chem. 262,5170-5179

Brozek, K. A., Hosaka, K., Robertson, A. D., and Raetz, C. R. H. (1989) J. Biol. Chem. 264.6956-6966

Clarke, L., and Carbon, J. (1976) Cell 9,91-99 Cronan, J. E., Jr., and Rock, C. 0. (1987) in Escherichia coli and

Salmonella typhimurium: Cellular and Molecular Biology (Neid- hardt, F. C., ed) pp. 474-497, American Society for Microbiology, Washington, D.C.

Cronan, J. E., Jr., and Vagelos, P. R. (1972) Biochim. Biophys. Acta 265, 25-60

by guest on September 4, 2020

http://ww

w.jbc.org/

Dow

nloaded from

1 -Acyl-sn-glycerol-3-phosphate Acyltransferase Mutant 17221

Galloway, S. M., and Raetz, C. R. H. (1990) J. Biol. Chem. 265, 6394-6402

Ganong, B. R., and Raetz, C. R. H. (1982) J. Biol. Chem. 257, 389- 394

Hawrot, E., and Kennedy, E. P. (1978) J. Biol. Ckm. 253, 8213- 8220

Hirota, Y., Suzuki, H., Nishimura, Y., and Yasuda, S. (1977) Proc. Natl. Acad. Sci. U. S. A. 74,1417-1420

Homma, H., Nishiiima, M., Kobayashi, T., Okuyama. H., and Nojima, S. (1981) Biochim. kophys. Acta 663, 1-13

_

Jacowski. S.. and Rock. C. 0. (1986) J. Biol. Chem. 261. 11328- 11333

Kates, M. (1986) in Laboratory Techniques in Biochemistry and Molecular Biology: Techniaues of Lbidologv-Isolation. Analvsis. and Identificati& of Lipids- (Burden; R. Hz-and vanKnippenberg; P. H., eds) pp. 171-179, Elsevier Science Publishers B.V., Amster- dam

Kato, J.-I., Nishimura, Y., Yamada, M., Suzuki, H., and Hirota, Y. (1988) J. Bacterial. 170, 3967-3977

Kohara, Y., Akiyama, K., and Isono, K. (1987) Cell 50,495-508 Miller, J. H. (1972) Experiments in Molecular Genetics, Cold Spring

Harbor Laboratory, Cold Spring Harbor, NY Misra, R., and Benson, S. A. (1988) J. Bacterial. 170,3611-3617 Nishijima, M., and Raetz, C. R. H. (1979) J. Biol. C&em. 254,7837-

7844 Phillips, T. A., Vaughn, V., Bloch, P. L., and Neidhardt, F. C. (1987)

in Escherichia coli and Salmonella typhimurium: Cellular and Mo- lecular Biology (Neidhardt, F. C., ed) pp. 919-966, American Society for Microbiology, Washington, D.C.

Raetz, C. R. H. (1976) J. Biol. Chem. 251,3242-3249 Raetz. C. R. H.. and Foulds. J. (1977) J. Biol. Chem. 252.5911-5915 Raetz; C. R. H:, Larson, T: J.,‘ and ‘Dowhan, W. (1977)‘Proc. Natl.

Acad. Sci. U. S. A. 74, 1412-1416 Ray, T. K., Cronan, J. E., Jr., Mavis, R. D., and Vagelos, P. R. (1970)

J. Biol. Chem. 245,6442-6448 Rock, C. O., Goelz, S. E., and Cronan, J. E., Jr. (1981) J. Biol. Chem.

256,736-742 Scheideler, M. A., and Bell, R. M. (1989) J. Biol. Chem. 264, 12455-

12461 Silhavy, T. J., Berman, M. L., and Enquist, L. W. (1984) Experiments

with Gene Fusions, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY

Smith, P. K., Krohn, R. I., Hermanson, G. T., Mallia, A. K., Gartner, F. H., Provenzano, M. D., Fuiimoto, K. K., Isoeke, N. M., Olson, B. J., and Klenk, D. C. (1985jAnal. Biochem. 150, 76-85

Tsuruoka. T., Ito. M.. Tomioka. S.. Hirata. A.. and Matsuhashi. M. (1988) i. Bc-zcte;iol. '170, 5229;5235

vanGolde, L. M. G., Shulman, H., and Kennedy, E. P. (1973) Proc. Natl. Acad. Sci. U. S. A. 70. 1368-1372

Wu, H. C., Tokunaga, M., Tokunaga, H., Hayashi, S., and Giam, C.- Z. (1983) J. Cell. Biochem. 22, 161-171

Wu, T. T. (1966) Genetics 54, 405-410

by guest on September 4, 2020

http://ww

w.jbc.org/

Dow

nloaded from

J Colemanphosphate acyltransferase activity.

Characterization of Escherichia coli cells deficient in 1-acyl-sn-glycerol-3-

1990, 265:17215-17221.J. Biol. Chem. 

  http://www.jbc.org/content/265/28/17215Access 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/265/28/17215.full.html#ref-list-1

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

by guest on September 4, 2020

http://ww

w.jbc.org/

Dow

nloaded from