The Inhibition of Phosphatidylglycerol Synthesis in Escherichia coli by ...

2 downloads 30 Views 919KB Size Report
Sep 26, 1973 - trichloroacetic acid and cold carrier E. coli sonic extract .... tures were labeled with [33P]phosphate for five generations prior to the addition of ...
Vol.

THE JOURNAL OF BIOLOGICAL CHEMISTRY 249, No. 8, Issue of April 25, pp. 2473-2477,

Printed

1974

in U.S.A.

The Inhibition of Phosphatidylglycerol Escherichia coli by 3,4=Dihydroxybutyll-phosphonate”

Synthesis

(Received

CHARLES S. SHOPSIS,$

ROBERT

in

for

publication,

September

26,

1973)

ENGEL, AND BTTRTONE. TROPP

From the Department of Chemistry, QueensCollege, City liniversity

SUMMARY The effect of 3,4-dihydroxybutyl-1-phosphonate on phosphoglyceride metabolism of Escherichia coli strain 8 was examined. The compound inhibits the incorporation of [33P]phosphate into phospholipids. Phosphatidylglycerol accumulation is very strongly inhibited, while phosphatidylethanolamine and cardiolipin accumulation are less significantly affected. Pulse labeling studies using [32P]phosphate indicate that phospholipid synthesis is inhibited to a much greater extent by 3,4-dihydroxybutyl-1 -phosphonate than is the synthesis of DNA or RNA. These studies also indicate that 3,4dihydroxybutyl-1 -phosphonate severely inhibits phosphatidylglycerol synthesis and that phosphatidylethanolamine synthesis is inhibited less severely. Cardiolipin synthesis is only slightly aBected. Studies on the rate of disappearance of [32P]phosphate from the phosphoglycerides of prelabeled cells indicate that 3,4-dihydroxybutyl-1-phosphonate treatment does not alter the rate of turnover of the phospholipids. 3,4-Dihydroxy[3-3H]butyl-l-phosphonate was readily incorporated into a chloroform-extractable fraction of E. coli strain 8.

of New York, Flushing, New York 11367

of strains of E. coli that are constitutive for the glycerol S-phosphate transport system and that lack the membrane-bound glycerol 3-phosphate dehydrogenase (4). The growth inhibition caused by 3,4-dihydroxybutyl-1-phosphonate differs from that caused by the natural metabolite in that it is not offset by the presence of glucose or high concentrations of phosphate in the growth medium and occurs in the presence of the membranebound glycerol 3-phosphate dehydrogenase (2). Differences in the metabolic effects of glycerol 3-phosphate and 3,4-dihydroxybutyl-1-phosphonate have also been reported (3). Studies on the incorporation of labeled precursors into DNA, RNA, protein, and lipid revealed that the 4-carbon phosphonate inhibits phospholipid synthesis most effectively. Glycerol 3phosphate was found to have its strongest inhibitory effect on the incorporation of labeled uracil into RNA. Treatment of E. coli strain 8 with 3,4-dihydroxybutyl-l-phosphonate caused an alteration in the distribution of labeled acetate incorporated into the phospholipid fraction. The major change was a reduction in the percentage of radioactivity in the phosphatidylglycerol fraction. Treatment of this strain with glycerol 3-phosphate had little effect on the distribution of label into the phosphoglycerides. The present report describes further studies on the role played by 3,4-dihydroxybutyl-1-phosphonate on phospholipid metabolism in E. coli strain 8. MATERIALS

Two phosphonic acid analogues of glycerol S-phosphate, 3,4-dihydroxybutyl-1-phosphonate, and 2,3-dihydroxypropyl-lphosphonate were synthesized (1) with the hope that they might influence the regulation of phospholipid metabolism. Previous reports have described some of the effects of these compounds upon the growth and metabolism of Escherichia coli (2, 3). The 4-carbon phosphonate was discovered to inhibit the growth of E. coli strains possessing an active glycerol 3-phosphate transport system (2). The a-carbon phosphonate does not show this inhibitory property. Glycerol S-phosphate inhibits the growth * This investigation was supported by Grant GB33718 from the National Science Foundation and a grant from the City University of New York Faculty Research Award Program. Taken in part from a dissertation submitted to the Faculty of Biochemistry of the City University of New York in partial fulfillment of the requirements for the degree of Doctor of Philosophy (C.S.S.). $ Present address, Department of Medical Biophysics, University of Toronto, Toronto, Ontario. 2473

AND

METHODS

Chemicals-Carrier-free [33P]phosphate and carrier-free [JzP]phosphate were purchased from New England Nuclear Corp., Boston, Mass. 3,4-Dihydroxy[3-3H]butyl-l-phosphonate (31 mCi per mmole)’ was prepared by the reduction of the diethyl ester of 4-acetoxy-3-oxobutyl-I-phosphonic acid with NaB3H4. DL-Glycerol3-phosphate (Grade X) was purchased from the Sigma Chemical Co., St. Louis, MO. Silica Gel G thin layer plates were purchased from Analabs, Inc., New Haven, Conn. The polygram Sil-N-HR thin layer plates used in the chromatography of 3,4-dihydroxybutyl-1-phosphonate are manufactured by Brinkmann Instruments, Inc., Westbury, N. Y. The bacterial phospholipid standards, phosphatidylethanolamine, phosphatidylglycerol, and cardiolipin were purchased from Supelco, Inc., Bellefonte, Pa. A modification of the procedure of Rosenthal and Geyer (5) was used to synthesize the dilithium salt of 3,4-dihydroxybutyl-l-phosphonate (1). All comments concerning glycerol S-phosphate or one of its phosphonic acid analogues refer to the racemic mixtures unless a specific enantiomer is specified. All other chemicals were of reagent grade. 1 S. Goldstein, 1). manuscript submitted

Braksmayer, for publication.

B. E.

Tropp,

and

R.

Engel,

2474 Bacteria-E. coli strain 8 was kindly provided by J. Cronan, Jr. The genotype of this strain as expressed by (Yale University). the genetic symbols described by Taylor (6) and Taylor and Trotter T7) is as-follows: Hfr C glpD3, glp Rc2, phoA8j tonA22, T2a, rel-1 0,). This strain was orieinallv isolated in E. C. C. Lin’s laboratory (Harvard Medical School). The culture medium of Garen and Levinthal (8) supplemented with 0.6 mM phosphate and 0.5yo potassium succinate was used throughout the present studies. All cultures were incubated at 37”. Overnight cultures were diluted 30-fold into the same synthetic medium and the diluted cultures were then incubated in a New Brunswick Metabolyte water-bath shaker, model G77, at 200 cpm. Cell growth was determined at 660 nm in a Klett-Summerson calorimeter (2). The phosphonic acid analogue was added when the turbidity of the cultures reached 25 to 30 Klett units which corresponds to 1.5 X lo8 cells per ml. In all experiments in which the dilithium salt of 3,4-dihydroxybutyl-1-phosphonate was added to a set of cultures, an equivalent concentration of lithium ion in the form of lithium chloride was added to the parallel set of untreated cultures. The concentrations of lithium ion used do not affect the rate of bacterial growth. Phospholipid Accumulation-E. coli strain 8 was cultured in the presence of 0.2 PCi of [a3P]phosphate per ml for five generations prior to the addition of 3,4-dihydroxybutyl-1-phosphonate. At various times, prior and subsequent to the addition of the phosphonate, 1.6-ml aliquots were removed. To the aliquots, 6.0 ml of methanol-chloroform (2:1), 2.0 ml of chloroform, and 2.0 ml of The mixture was agitated after water were added sequentially. This is aceach addition and the chloroform fraction collected. cording to the procedure of Bligh and Dyer (9) as modified by Ames (10). The chloroform extracts were then washed three times with 2 M KC1 and once with water. A 0.5-ml portion of these chloroform soluble extracts was counted as previously described (3). The remainder of the chloroform soluble extract was retained for subsequent analysis. The second set of experiments used to investigate the synthesis of phospholipids involved the simultaneous addition of [a3P]phosphate to a final concentration of 4.0 &i per ml and 3,4-dihydroxybutyl-1-phosphonate to cultures of E. coli strain 8. At various time intervals 1.6-ml aliquots were removed. These samples were extracted, washed, and counted as described for the first set of experiments. Rate of Synthesis of RNA, DNA, and Phosphoglycerides-The rate of synthesis of RNA, DNA, and phosphoglycerides was determined by pulse labeling with [32P]phosphate. Cultures of E. coli strain 8 were treated with 3,4-dihydroxybutyl-l-phosphonate or lithium chloride. At various time intervals after treatment 2.0-ml samples were removed and incubated with 20 &i of [3”P]phosphate for 10 min. These samples were then analyzed by a modificati‘on of the procedure of Lusk and Kennedy (11). They were treated with 2.0 ml of cold 10% (w/v) trichloroacetic acid and cold carrier E. coli sonic extract was added. The precipitates were collected by centrifugation, washed two times with cold 5% trichloroacetic acid, and then suspended in a mixture of 5 ml of chloroform, 5 ml of methanol, and 1 ml of water for 30 min. The samples were then centrifuged and the supernatant fluid containing the phosphoglycerides was removed and subsequently washed, evaporated, and counted as described above. The pellets were washed with chloroform-methanol-water (5:5:1), the wash was discarded, and the pellets were suspended in 2.0 ml of 0.5 N KOH. These suspensions were incubated overnight at 37” to hydrolyze the RNA. The samples were chilled, 2.0 ml of 1 M perchloric acid added, and the samples centrifuged to sediment the DNA. A 1.0.ml aliquot of the supernatant fluid containing hydrolyzed RNA was counted in 10 ml of Patterson-Greene scinThe DNA pellets were dissolved in 2.0 ml of tillation fluid (12). 1 N KOH. neutralized with HCl. and nrecinitated with 12% trichloroacetic acid. The newapellets w&e washed once with 5yo trichloroacetic acid, dissolved in 1.0 ml of 0.2 N KOH, and counted in 10 ml of Patterson-Greene scintillation fluid (12). Turnover of Phosphoglycerides-Phospholipid turnover was studied b.y culturing E. coli strain 8 for three generations in media containing 1.3 &i per ml of [32P]phosphate.The bacteria were collected on membrane filters (Millinore Corn.. Bedford. Mass.). washed with culture medium at 37”,-and resuspended in the or&: Lithium inal volume of culture medium at the same temperature. chloride or 3,4-dihydroxybutyl-1-phosphonate was added to the

cultures. Immediately before the addition of these compounds and at several time intervals after their addition 2.0.ml samples were removed. Lipids were extracted from these samples by the Ames procedure (10) described previously. A 0.5.ml portion of these extracts was counted to determine total phospholipids and the remainder was used for the analysis of individual phosphoIn some experiments 2.0.ml samples were centrifuged glycerides. and the supernatant fluid and the pellet were extracted independently to determine whether lipids had been excreted into the medium. Analysis of Phosphoglycerides-Lipid extracts obtained in the experiments described above were analyzed by thin-layer chromatography. The solvent was evaporated and the samples were redissolved in a chloroform-methanol (1: 1) solution containing cold carrier lipid which had been extracted from E. coli K12 by the Ames procedure. The chromatographic procedure was the two-step developing system previously described (3, 13). Incorporation of S,Q-Dihydroxy[S-3H]butyl-l-Phosphonate into Chloroform-Extractable Fraction-The degree of incorporation of 3,4-dihydroxybutyl-1-phosphonate into a chloroform-extractable fraction was determined by adding the labeled analogue (0.03 mM, 31 mCi per mmole) to early log phase cultures of bacteria. At various time intervals 2.0.ml samples were removed. The phosphoglycerides were extracted from these samples by the Ames A portion of these extracts was procedure described previously. dried and counted in toluene-based scintillation fluid to determine the extent of incorporation into the chloroform-extractable fraction and the remainder was saved for thin layer chromatographic analysis. To determine whether 3,4-dihydroxy[3-3H]butyl-I-phosphonate remained intact after incorporation into the chloroform-soluble

material

the labeled phospholipid

fraction

was hydrolyzed

by re-

The fluxing with 1 M NaOH in 50yo aqueous methanol for 1 hour. sodium ions were removed bv treating with Dowex 5OW-X8 ion exchanger in the H+ form. ‘The concentrated hydrolysate was sootted on a Polveram Sil-N-HR thin-laver olate. The olate was developed in the solvent system of It. Tyhach2 which consists of butanol-HCl-Hz0 (12:2:1). Th is chromatographic system yields a discrete peak of label with an RF of approximately 0.65 when

authentic

3,4-dihydroxy

[3-3H]butyl-l-phosphonate

is chromato-

graphed. RESULTS

The

present

studies

were

designed

to obtain

further

informa-

tion about the perturbation of phosphoglyceride metabolism in E. coli caused by 3,4-dihydroxybutyl-1-phosphonate. Specifically, the effects of 0.03 mM 3,4-dihydroxybutyl-l-phosphonate This concentration was previously demonhave been examined. strated to strongly inhibit phospholipid synthesis but to have only a mild effect upon cell growth (3). Fig. 1 depicts the results of an experiment in which the cultures were labeled with [33P]phosphate for five generations prior to the addition of the phosphonate in order to achieve a constant Addition of the specific activity of label in the phospholipids. nhosnhonate to these cultures resulted in a decrease in the rate of accumulation of total phospholipids (Fig. 1A). An analysis of the fates of the individual phospholipids (Fig. 1, B to D) indicates that phosphatidylglycerol accumulation was very strongly inhibited while the accumulation of phosphatidylethanolamine and cardiolipin were much less significantly affected by the addition of 3,4-dihydroxybutyl-I-phosphonate. The amount of label in phosphatidylglycerol actually decreased as a result of the treatment of the bacteria with the phosphonate. Experiments in which the 3,4-dihydroxybutyl-1-phosphonate and [32P]phosphate were simultaneously added to the cultures also revealed that the rate of phosphatidylglycerol accumulation was the most sensitive to the presence of 0.03 mM phosphonic acid It should be noted that some inanalogue (data not shown). corporation of label into phosphatidylglycerol did take place in 2 Personal

communication.

2475

4

2 Hours

a;

2 Hours

FIG. 1. The effect of 3,4-dihydroxybutyl-1-phosphonate on the accumulation of [aaP]phosphate into the phospholipids of Escherichia coli strain 8. The bacteria were cultured in radioactive medium for five generations prior to the addition of the phosphowas measnate at the time indicated by the arrozu. Incorporation

n -m, unured as described under “Materials and Methods.” treated cells; l -•,0.03 mM 3,4-dihydroxybutyl-l-phosphonatetreated cells. A, total phospholipids; B, phosphatidylglycerol; C, phosphatidylethanolamine; D, cardiolipin.

the second type of experiment. The actual decrease of [““PIphosphate in phosphatidylglycerol revealed by Fig. 1 is due to The results obtained thus far the turnover of t,his component. might be due to the inhibition of phosphatidylglycerol synthesis, the increase in its rate of turnover, or a combination of both of these events taking place in the presence of 3,4-dihydroxybutyl1-phosphonate. Pulse-labeling studies were performed to determine if the data presented in Fig. 1 are a result of the inhibition of phosphoglyceride synthesis by the phosphonate and to compare the effects of the inhibitor on the synthesis of phospholipids, RNA, and Cultures of E. coli strain 8 were pulsed with [3zP]phosDNA. phate at various times after the addition of the phosphonate. Figs. 2, 3, and 4A indicate that 0.03 mM 3,4-dihydroxybutyl-lphosphonate inhibits the synthesis of total phospholipids much more effectively than it inhibits the synthesis of RNA or DNA. The synthesis of RNA and DNA appears to undergo a transient These stimulation as a result of treatment with the phosphonate. results have been reproduced several times. Fig. 4, B to D presents the results of an analysis of the phospholipids synthesized during the pulse. Phosphatidylglycerol synthesis is severely inhibited by the phosphonate at the earliest time point.

Phosphatidylethanolamine synthesis shows significant inhibition at later time points. The synthesis of cardiolipin was only slightly affected by the presence of 3,4-dihydroxybutyl-l-phosphonate. It should be emphasized that while the data have not been adjusted for the increase in cell number during the course of the experiment the turbidities of treated and untreated cultures increase in an identical fashion. The effects of 3,4-dihydroxybutyl-1-phosphonate on the turnover of previously synthesized phospholipids was examined next. E. coli strain 8 was cultured for three generations in the presence of [32P]phosphate. The bacteria were collected on membrane filters, washed, and resuspended in unlabeled medium. Treatment of these prelabeled cultures with 3,4-dihydroxybutyl-lphosphonate has virtually no effect upon the rates of turnover Crowfoot et al. (14) have of the phospholipids (data not shown). obtained evidence for the excretion of phospholipids into the medium under conditions in which phospholipid metabolism is perturbed. To examine this possibility cultures were prelabeled with [a2P]phosphate and then treated with 3,4-dihydroxybutyl-lphosphonate. Aliquots of the culture were removed at various times and centrifuged. The bacterial pellet and the medium were extracted independently by the Ames method (10). These

2476 experiments failed to reveal phospholipid excretion by either treated or untreated cultures (data not shown). Experiments were performed to determine whether E. coli strain 8 could incorporate 3,4-dihydroxy[3-*H]butyl-l-phosphonateinto a chloroform-solublematerial. Fig. 5 indicatesthat 0.03 mM labeled phosphonateis linearly incorporated into the chloroform-solublefraction for at least90 min. Analysis of this fraction by thin-layer chromatography on Silica Gel G plates developed with either chloroform-methanol-water (615 :25: 3) (13) or chloroform-methanol-aceticacid (65:25:8) (10) revealed that almostall of the radioactivity appliedto the platesremained at or just above the origin of the chromatogram. This doesnot correspondto the Rp values of any of the major phospholipids of E. coli (10) and is indicative of a very polar molecule. Alkaline hydrolysis of the tritium-labeled phospholipidfraction followed by thin layer chromatography on Sil-N-HR plates revealed that approximately 70% of the radioactive material migrated asexpectedfor 3,4-dihydroxybutyl-1-phosphonate. The remainderof the labeledmaterial trailed from the origin to the main peak. There wasno label in the regionbetweenthe region containing

3,4-dihydroxybutyl-1-phosphonate

and the solvent

front. The trailing was diminishedbut not completely eliminatedby treating the hydrolysate with alkalinephosphataseprior

FIG. 2 (left). The effect of 3,4-dihydroxybutyl-1-phosphonate on the rate of RNA synthesis by Escherichia coli strain 8. RNA synthesis was measured by pulse labeling for 10 min with [32P]phosphate as described under “Materials and Methods.” Zero time indicates the time of addition of the inhibitor. WW, untreated cells; O----O, 0.03 mM 3,4-dihydroxybutyl-l-phosphonate-treated cells. FIG. 3 (right). The effect of 3,4-dihydroxybutyl-1-phosphonate on the rate of DNA synthesis by Escherichia coli strain 8. DNA svnthesis was measured bv nulse labeling for 10 min with [a*P]ghosphate as described uidkr “Mat,eriaG and Methods.” Zero time indicates the time of addition of the inhibitor. m-m, untreated cells; O----O, 0.03 mM 3,4-dihydroxybutyl-l-phosphonate treated cells.

to chromatography. Theseresultsindicate that 3,4-dihydroxy[3-aH]butyl-l-phosphonate

intact into the chloro-

DISCUSSION

The present studies indicate that phospholipidsynthesis is more sensitiveto inhibition by 3,4-dihydroxybutyl-l-phosphonate than is the synthesisof either RNA or DNA. The incorporation of isoleucineinto trichloroacetic acid-precipitablematerial waspreviously demonstratednot to be significantly inhibited by the 4-carbonphosphonate(3). Fig. 3 indicatesthat 3,4-dihydroxybutyl-l-phosphonate inhibits the rate of DNA synthesisat the later time points examined. This result which appearsto conflict with our previousreport showingvery little inhibition of DNA synthesisby the 4-carbon phosphonate(3) is probably due to one or more of the following differencesin experimental proceduresused. (a) The present study reports the results of pulse labeling with [S2P]phosphate, whereasthe previous one reported the incorporation of [3H]thymine into DNA. (6) The minimal mediumand carbonsourcesusedin the two experiments differed. (c) The strainsof E. coli useddiffered (strain 8 in the

FIG. 5. The incorporation of 0.03 mM 3,4-dihydroxy[3-*H]butylI-phosphonate into the chloroform-extractable fraction of Escherichia coli strain 8. Incorporation was measured as described under “Materials and Methods.” Zero time indicates the time of addition of the phosphonate.

20

FIG. 4. The effect of 3,4-dihydroxybutyl-1-phosphonate on the rate of phospholipid synthesis by Escherichia coli strain 8. Phospholipid synthesis was measured by pulse labeling for 10 min with [**P]phosphate as described under “Materials and Methods.”

is incorporated

form-soluble material. The experiment with alkaline phosphataseis suggestiveof a phosphatelinkage to the incorporated phosphonicacid analogue. Unfortunately the absenceof a second discretepeak of label in the alkaline phosphataseuntreated samplemakesthis suggestionquite tentative.

40 MlWJN8

60

W

Mln!4nl

40

60

&---¤, Zero time indicates the time of addition of the inhibitor. untreated cells; 0-0, 0.03 mM 3,4-dihydroxybutyl-l-phosphonate-treated cells. A, total phospholipids; B, phosphatidylglycerol; C, phosphatidylethanolamine; L), cardiolipin.

2477 present study and strain 1908 in the previous one). The inhibition observed in the present study is not immediate and may represent interfercncc with the initiation of a new round of rcplication. We have proposed that the phenethyl alcohol-sensitive component postulated by Lark and Lark (15) may be a phospholipid (13). Since 3,4-dihydroxybutyl-l-phosphonatc also perturbs phospholipid metaboiism it too may interfere with the initiation of a new round of replication. Further studies are required to establish what role, if any, the phosphonic acid analogue plays in DNA metabolism. The data from the present as well as the previous study clearly indicate that phosphatidylglycerol synthesis is more effectively inhibited by 3,4-dihydroxybutyl-1-phosphonate than is the synthesis of either phosphatidylethanolamine or cardiolipin. It is clear from Fig. 1 that 3,4-dihydroxybutyl-l-phosphonatc causes an actual decrease in the cellular content of phosphatidylglycerol. The decrease is due to the t,urnover of phosphatidylglycerol. The rate of this turnover is not affected by the 4-carbon phosphonate, whereas the rate of synthesis is affected (Fig. 4). The fact that the inhibition of phosphatidylethanolamine synthesis is milder and occurs later than the inhibition of phosphatidylglycerol synthesis suggests that this effect- of 3,4-dihydroxybutyl-lphosphonate is a secondary one. The continued synthesis of cardiolipin when phosphatidylglycerol synthesis is inhibited may be due to a preferential conversion of the phosphatidylglycerol synthesized into cardiolipin. In vitro studies appear to be consistent with the in vivo investigations. Thus while the 4-carbon analogue is an inhibitor of CDP-diglyceride : glycerol 3-phosphate phosphatidyl transferase it has no affect upon CDP-diglyceride :serine phcsphatidyl transferase activity.3 The 4-carbon phosphonate does not appear to be either a substrate or an inhibitor of glycerol 3-phosphate acyltransferaseP The lack of accumulation of phosphatidylserine in phosphonate-treated cells suggests that phosphatidylserine decarboxylase, the other enzyme specific to phosphatidylethanolamine synthesis, is not inhibited in cells treated with the phosphonate. 3,4-Dihydroxy[3-3H]butyl-l-phosphonate was readily incorporated into the lipid fraction of E. coli strain 8 (Fig. 5). The lipid(s) it was incorporated into has not yet been characterized. The low RF of the labeled material upon thin layer chromatography suggests a highly polar lipid. The chloroform-soluble tritium-labeled material accumulates in amounts roughly equivalent t.o the net increase in phosphatidylglycerol in untreated cells a Unpublished 4 Unpublished

data of W. D. Nunn, R. Engel, and B. E. Tropp. data of P.-J. Cheng, R. Engel, and B. E. Tropp.

during the same time period. In vitro experiments indicat.e that, 3 ,4-dihydrosy[3-311]but~yl-l -phosphonate is a substrat’e for CDPdiglyc!ride:glyterol 3-phosphate phosphatidyl t,ransferase.” We therefore believe that the phosphonic acid analogue of phosphatidylglycerol phosphate is synthesized in viva. Assuming the 4-carbon phosphonate is incorporated into the analogue of phosphatidylglycerol phosphate one of the following may account for its mode of cell growth inhibition: (a) the inhibition of phosphatidylglycerol synthesis, (b) the appearance of the phosphonic acid analogue of phosphatidylglycerol phosphate, or (c) the presence of a high intracellular level of 3,4-dihydroxybutyl-1-phosphonate. Other possibilities cannot be excluded at. this time. For example 3,4-dihydroxybutyl-1-phosphonate may interfere in some fashion with the synthesis or function of the new class of oligosaccharides reported by van Golde et al. (16). It is clear that 3,4-dihydroxybutyl-l-phosphonate has a profound effect upon the metabolism of E. coli strain 8 and that it inhibits phosphatidylglycerol synthesis in this strain. Acknowledgments-We wish to thank Mr. C.-T. Tang for his technical assistance. We are grateful to Dr. William Nunn, Dr. John Cronan, Jr., and Mr. Richard Tyhach for helpful discussions during the course of this investigation.

REFERENCES J., DXPILIPPI, L., ENGEL, R., AND TROPP, B. E. (1972) J. Med. Chem. 16, 1074 SHOPSIS, C. S., ENGEL, R., AND TROPP, B. E. (1972) J. Bacteriol. 112, 408 SHOPSIS, C. S., NUNN, W. D., ENGEL, R., AND TROPP, B. E. (1973) Antimicrob. Agents Chemother. 4, 467 COZZARELLI, N. R., KOCH, J. P., HAYASHI, S., AND LIN, E. C. C. (1965) J. Bacterial. 90, 1325 ROSENTHAL, A., AND GEYER, R. (1958) J. Amer. Chem. Sot. 80, 5240 TAYLOR, A. L. (1970) Bacterial. Rev. 34, 155 TAYLOR, A. L., AND TROTTER, C. D. (1972) Bacterial. Rev. 36, 504 GAREN, A., AND LEVINTHAL, C. (1960) Biochim. Biophys. Acta 38, 470 BLIGH, E. G., AND DYER, W. J. (1959) Can. J. Biochem. 37, 911 AMES, G. F. (1968) J. Bacterial. 96, 833 LUSK, J. E., AND KENNEDY, E. P. (1972) J. Bacterial. 109, 1034 PATTERSON, M. S., AND GREENE, R. C. (1965) Anal. Chem. 37, 854 NUNN, W. D., AND TROPP, B. E. (1972) J. Bacterial. 109, 162 CROWFOOT, P. D., OKA, T., ESFAHANI, M., AND WAKIL, S. J. (1972) J. Bacterial. 112, 1396 LARK, K. G., AND LARK, C. (1966) J. KoZ. BioZ. 20, 9 VAN GOLDE, L. M. G., SCHULMAN, H., AND KENNEDY, E. P. (1973) Proc. Nat. Acad. Sci. U. S. A. 70, 1368

1. KABAK, 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16.