Synthase in Escherichia coli

JOURNAL OF BACTERIOLOGY, Aug. 1981, p. 552-562 0021-9193/81/080552-1 1$02.00/0 Vol. 147, No. 2 Cloning of Genes Involved in Membrane Lipid Synthesis...
Author: Shanon Ami Bell
1 downloads 2 Views 2MB Size
JOURNAL OF BACTERIOLOGY, Aug. 1981, p. 552-562 0021-9193/81/080552-1 1$02.00/0

Vol. 147, No. 2

Cloning of Genes Involved in Membrane Lipid Synthesis: Effects of Amplification of Phosphatidylglycerophosphate Synthase in Escherichia coli AKINORI OHTA,t KAREN WAGGONER, ANNA RADOMINSKA-PYREK, AND WILLIAM DOWHAN* Department of Biochemistry and Molecular Biology, University of Texas Medical School and University of Texas Graduate School of Biomedical Sciences, Houston, Texas 77025 Received 16 April 1981/Accepted 20 May 1981

The structural gene (pgsA) for the CDP-diacylglycerol:sn-glycero-3-phosphate phosphatidyltransferase (EC 2.7.8.5, phosphatidylglycerophosphate synthase) from Escherichia coli has been cloned, using pSC101 as the vector. The resulting hybrid plasmids not only correct the lack of in vitro synthase activity in pg.sA strains but also cause an amplification (6- to 40-fold over wild-type levels) in enzymatic activity in direct proportion to the copy number of the plasmids found in vivo. The cloned gene also corrects the abnormally low level of polyglycerophosphatides found in pgsA strains and actually increases the level of phosphatidylglycerol to above that normally found in E. coli. The degree of alteration in phospholipid composition brought about by these hybrid plasmids is not of the order expected if fluctuations in enzyme levels in vivo were an important regulatory mechanism in phospholipid metabolism. The isolated hybrid plasmids have been mapped by restriction endonuclease analysis. The presence and location of other genetic markers have also been established. The above data, along with analysis of deletion derivatives of these plasmids and subcloning of appropriate restriction fragments, have established the position of the pgsA locus on the hybrid plasmids. From this data, the position of the pgsA locus has been determined to lie between flaI and uurC on the E. coli genetic map. The CDP-1,2-diacyl-sn-glycerol (CDP-diacylglycerol) :sn-glycero-3-phosphate phosphatidyltransferase (EC 2.7.8.5, phosphatidylglycerophosphate synthase) of Escherichia coli is tightly associated with the inner cytoplasmic membrane (5, 42). The purified enzyme from E. coli B (18) exhibits many of the properties of a membrane-associated enzyme. This enzyme catalyzes the committed step for the synthesis of polyglycerophosphatides in E. coli (19) and may therefore be involved in control of phospholipid metabolism. The genetic control of the synthesis of this enzyme, as well as the synthesis of phosphatidylglycerophosphate in E. coli, appears to be under the control of two genetic loci (26, 27, 33). Mutants in the structural gene (pgsA near min 42) for phosphatidylglycerophosphate synthase have been isolated which grow normally under all conditions tested, but show a significantly lower phosphatidylglycerol content; lack of in vitro enzymatic activity in these mutants is corrected by F'150. Mutants at the pgsB locus

(mapping near min 4) have no phenotype by themselves (26, 27), but in a pgsA background confer temperature sensitivity at 44°C for growth and polyglycerophosphatide synthesis. The expression of residual phosphatidylglycerophosphate synthase activity in these double mutants is temperature sensitive at 44°C, but activity expressed before a temperature shift appears to be unaffected by incubation at 44°C; this enzymatic activity is stable to treatment at 70°C in wild-type strains but sensitive in pgsA mutants. To do further studies on the structure and enzymology of this enzyme as well as the genetic aspects related to its synthesis and assembly into the membrane, we report the isolation of the pgsA gene locus carried by extra chromosomal hybrid plasmids. The availability of the cloned gene will aid studies both in vivo and in vitro on the relationship between the two loci affecting the synthesis of phosphatidylglycerol and the role they play in the synthesis and assembly of the phosphatidylglycerophosphate

synthase. (A preliminary report of these results by A. Ohta and W. Dowhan has appeared in the Ab-

t Present address: Department of Biochemistry, Saitama

University, Urawa, 338, Japan.

552

PHOSPHATIDYLGLYCEROPHOSPHATE SYNTHASE GENE

VOL. 147, 1981

TABLE 1. Strains of E. coli K-12

stracts of the XIth International Congress of

Biochemistry, abstr.

no.

376, 1979.)

MATERIALS AND METHODS Materials. All chemicals were reagent grade or better. Radiochemicals were obtained from Amersham Corp., Arlington Heights, Ill. Tryptone, yeast extract, and agar were purchased from Difco Laboratories, Detroit, Mich. Restriction endonucleases were purchased from either Bethesda Research Laboratories, Bethesda, Md., or New England Biolabs, Beverly, Md. New England Biolabs supplied the A c1857 and pX174 DNA and the E. coli polynucleotide ligase, whereas the T4 DNA ligase was purchased from Worthington Diagnostics, Freehold, N.J. Precoated analytical silica gel, thin-layer plates were purchased from E. Merck AG, Darmstadt, West Germany. Rohm and Haas, Philadelphia, Pa., supplied the Triton X-100. Lysozyme and all antibiotics used in this work were obtained from Sigma Chemical Co., St. Louis, Mo. Ethidium bromide, agarose powder, and Dowex AG50W-X8 were purchased from Bio-Rad Laboratories, Richmond, Calif. Cesium chloride was obtained from Schwarz/Mann, Orangeburg, N.Y. E. I. du Pont de Nemours & Co., Inc., Wilmington, Del., supplied the X-ray film (Cronex-4). Bacterial strains and growth conditions. E. coli strains used in this work are listed in Table 1. R477-10 and R477-100 are an isogenic pair obtained from C. R. H. Raetz (University of Wisconsin, Madison). AD1 and AD2 are spontaneous nalidixic acid-resistant (gyrA [24] based on resistance to 25 [Lg/ml) derivatives of R477-10 and R477-100, respectively, which were made thymine dependent by trimethoprim treatment as described by Miller (24). These strains were then made recA by mating with KL16-99 as the donor. JA200 was obtained from one of the plasmid-carrying colonies (colony 34-44, see reference 8) of the Clarke and Carbon collection as a clone sensitive to colicin El toxin and lacking plasmid DNA, as determined by agarose gel electrophoresis of lysates from single colonies, as previously described (30). The plasmid-containing collection was generously supplied to us by John Carbon (University of California, Santa Barbara). All of the MS strains are motility-minus derivatives of AB1884 received from M. Simon (University of California, San Diego). JC1553 carries plasmid F'150 (see Table 2) which covers the E. coli genetic map from eda to his and includes the pgsA gene locus (27). Bacteria were routinely grown in liquid cultures or on agar plates containing LB or enriched LB medium supplemented with glucose as previously described (30). Minimal (M9) salts (24) medium (6 g of Na2HPO4, 3 g of KH2PO4, 0.5 g of NaCl, 1 g of NH4Cl, 0.12 g of MgSO4, and 10 mg of CaCl2 per liter) supplemented with 40 jig of each required L-amino acid per ml, 1 jig of thiamine per ml, and 0.2% glucose was used for both agar plates and liquid cultures. Antibiotics (100 /g of streptomycin per ml, 50 jig of ampicillin per ml, 25 jig of tetracycline per ml, or 20 ,ig of nalidixic acid per ml) or colicin El toxin (50 to 100 U/ml) prepared as previously described (30) was added to cultures and agar plates when required. DNA. Plasmids pSCIOI and pBR322 (Table 2) were

553

Source Relevant markers C. Raetz pg.sA1O (lacking in vitro phosphatidylglycerophosphate synthase activity) his recA + rpsL C. Raetz R477-100 pgsA+ his recA+ rpsL CGSCa HfrKL16 recA thy' KL16-99 This work ADI pgsA-10 his recA+ thy gyrA rpsl This work AD2 pgsA + his recA + thy gyrA rpsL This work pgsA-10 his gyrA recA AD10 thy + rpsL (derived from AD1 by mating with KL16-99 donor) This work AD100 pgsA + his gyrA recA thy' rpsL (derived from AD2 by mating with KL16-99 donor) This work JA200 F+/pgsA + AtrpE5 recA thr leu thi CGSC uvrC rpsL AB1884 M. Simon uurC motA rpsL MS725 M. Simon uvrC flaD thyA rpsL MS691 M. Simon uvrC flaD thyA rpsL MS788 CGSC JC1553/F'150 F'150/argG metB his leu recA rpsL C. Raetz A324 pgsA + proC lacI rpsL Stock Yale a CGSC, E. coli Genetic Center, University, New Haven, Conn.

Strain R477-10

TABLE 2. Plasmids carried by E. coli K-12 Plasmid

Relevant markers

Source

Tet' A. Dugaiczyk Tet' Amp' A. Dugaiczyk R. McMacken Km' Amp' CGSC" eda+pgsA+ uurC+ his' J. Carbon ColElmm uvrC' pLC13-12 flaD+ This work pPG1 pgsA + uvrC' Tet' pPG2 pgsA+ uurC' Tetr This work This work pPG1-A1 ApgsA AuvrC Tet' This work pPG1-A2 ApgsA AuurC Tetr pPG1-L pgsA+ uurC+ Tet' This work This work pPGL2019 pgsA + Ampr CGSC, E. coli Genetic Stock Center, Yale University, New Haven, Conn.

pSC1O0 pBR322 pKC7 F'150

a

obtained from A. Dugaiczyk (Baylor College of Medicine, Houston, Tex.) and prepared from tetracyclineresistant transformants of JA200, as described previously (30). Plasmid pKC7 (36) was obtained from Roger McMacken (John Hopkins University, Baltimore, Md.). All plasmid DNAs listed in Table 2 were isolated from cell lysates by CsCl bouyant density centrifugation in the presence of ethidium bromide (10). Plasmids containing either pBR322 or ColEl as the cloning vector were amplified before isolation by incubation of cultures in the presence of chloramphenicol (11). Plasmids containing only pSC1O1 as the vector could not be amplified, so preparation of the plasmid was done from cells harvested in mid-log phase of growth.

554

OHTA ET AL.

Preparation of cell fractions. Crude extracts of cells were prepared by sonic disruption at 0°C after suspension in 100 mM Tris-hydrochloride, pH 7.4; unbroken cells were removed by centrifugation at 3,000 x g. The membrane and supernatant fractions were separated by centrifugation at 100,000 x g for 90 min.

Enzyme assay. Phosphatidylglycerophosphate synthase was assayed by the conversion of labeled glycerophosphate to chloroform-soluble material at 30°C, as previously described (18). The preparation of the lipid substrate CDP-diacylglycerol (1, 7) and radiolabeled glycerophosphate (18) has been described elsewhere. The enzyme assay mixture (0.1 ml) consisted of 0.1 M Tris-hydrochloride, pH 8.0, 0.1 M MgCl2, 1% (wt/vol) Triton X-100, 0.2 mM CDP-diacylglycerol, and 0.8 mM sn-[2-3H]glycero-3-phosphate. Phosphatidylserine synthase (30) and phosphatidylserine decarboxylase (13) were assayed as previously described. Specific activity is defined as nanomoles of product formed under the above conditions per minute per milligram of protein, as determined by the method of Lowry et al. (23). Labeling of phospholipids. Cells were labeled for at least five generations in LB broth containing [32p]pi (10 MCi/mnl) to determine phospholipid composition. Extraction of labeled lipids from cells and analysis by two-dimensional, thin-layer chromatography were carried out as previously described (27). Colony autoradiography. Single colonies were screened for their ability to synthesize phosphatidylglycerophosphate in vitro by the filter paper autoradiography technique described by Raetz (33). The method is dependent on the ability of individual lysed colonies to incorporate sn-[U- 4C]glycero-3-phosphate into trichloroacetic acid-insoluble material dependent on the presence of CDP-diacylglycerol. Agarose gel electrophoresis. Plasmids and DNA fragments were separated and analyzed by electrophoresis in agarose gel slabs (30). The sizes (in kilobase [kb] pairs) of linear DNA fragments were determined relative to known restriction fragments of A and OX174 DNA. The presence of plasmids was confirmed by direct analysis of lysates from single bacterial colonies as described by Barnes (3). Endonuclease digestion, ligation, and transformation. Endonuclease digestions and DNA ligations were carried out under the conditions recommended by the suppliers of the enzymes. Ligated DNA was used to transform competent cells prepared by CaCl2 treatment as previously described (21). Individual transformants were selected under the appropriate conditions on agar plates. Isolation of F'150 DNA. F'150 is susceptible to deletions (22), particularly in the region near the pgsA gene. The F'150 (his') used in this work was transferred from JC1553 to ADIO (pgsA his) by conjugation in LB broth at 370C as described by Miller (24). AD10 carrying F'150 was selected by its ability to grow on minimal agar plates in the presence of nalidixic acid (gyrA) and the absence of histidine. Four colonies selected from several hundred isolates were grown in minimal medium supplemented with thiamine, threonine, leucine, and glucose, but lacking histidine. Con-

J. BACTERIOL. trol cultures of ADIO and AD100 were grown with the appropriate supplements. Cell lysates of each were assayed for phosphatidylglycerophosphate synthase activity which showed the presence of the enzymatic activity in AD100, AD1O(F'150), but not in ADIO lacking F'150, confirming the presence ofpgsA (27) on the F'150 used in this work. The covalently closed circular DNA form of F'150 was prepared by the method used by Hansen and Olsen (15) to prepare large plasmids as briefly summarized below. Strain JC1553(F'150) was grown to mid-log phase at 37°C in 2 liters of minimal medium supplemented with glucose, thiamine, methionine, arginine, threonine, and leucine. The harvested cells were washed, lysed with lysozyme and sodium dodecyl sulfate, and treated with alkali before removal of the membrane-chromosome complex by centrifugation. Covalently closed circular DNA was separated from linear DNA by equilibrium sedimentation as described above. Construction and isolation of pgsA-containing plasmids. F'150 (5.4 ig in 50 Ml) and pSCIOI (1.1 Mg in 10 Ml) were digested separately with EcoRI endonuclease at 37°C, using a NaCl concentration (44 mM) lower than is normally employed, thereby broadening the specificity of EcoRI and making possible the restriction at EcoRI* sites (32). The reaction was stopped by heating at 65°C for 5 min. The digested DNA samples were then mixed and subjected to ligation at 15°C, using 2 Weiss units of T4 polynucleotide ligase. The reaction was terminated by bringing the solution to 20 mM in EDTA. Approximately 2 x 1011 competent cells (ADIO) were transformed with the above ligation mixture. Colonies resistant to tetracycline were selected by growth on LB agar plates containing tetracycline at 37°C. The resulting colonies were screened for the ability to synthesize labeled phosphatidylglycerophosphate in vitro after treatment at 70°C from CDP-diacylglycerol and sn-[U-14C]glycero-3-phosphate by the colony autoradiography technique described above. From the colony autoradiographic analysis of 12,000 transformants, six positive colonies were isolated. Each of the six positive colonies was checked for the genetic markers of ADIO, size of plasmid DNA, and presence of phosphatidylglycerophosphate synthase activity. Of the six original isolates, one contained only pSC101, three contained a plasmid of about 20 kb (pPG1), and two contained a plasmid of about 17 kb (pPG2). Strains (ADIO) carrying pPG1 and pPG2 when grown in LB broth exhibited a specific activity for phosphatidylglycerophosphate synthase much higher than that of ADl00; the pSC101-carrying strain showed no detectable phosphatidylglycerophosphate synthase activity.

RESULTS Characterization of pgsA-carrying plasmids. Plasmid DNA of both sizes (pPG1 and pPG2) was isolated and used to transform AD10 and AD 100 to tetracycline resistance. The transformation efficiency of both plasmids was about 104 transformants per yg of DNA with AD 100 as

PHOSPHATIDYLGLYCEROPHOSPHATE SYNTHASE GENE

VOL. 147? 1981

recipient and about 105 transformants per ug with ADlO as recipient; several hundred colonies were isolated in each case. Cell-free extracts of cultures derived from each size class of DNA were made after growth to stationary phase in LB broth containing tetracycline. The specific activity of phosphatidylglycerophosphate synthase in these cultures is shown in Table 3 and is compared to an AD100 and ADIO control. Therefore, correction of the pgsA phenotype (i.e., lack of in vitro enzymatic activity) and overproduction of enzymatic activity are coincident with tetracycline resistance. Mixing of extracts from plasmid-containing strains with plasmid-free extracts showed no evidence for an activating agent in the plasmid-containing extracts (data not shown). Restriction maps and genetic markers. Figure 1 shows the restriction endonuclease digestion products of pPG1, pPG2, and pSC101, using EcoRI under normal salt conditions and at a reduced NaCl concentration (EcoRL* conditions). EcoRI digestion of pPG1 results in a fragment the size of linear pSC101 (about 9.5 kb) and a 10.4-kb fragment which was presumably derived from E. coli DNA. EcoRI digestion of pPG2 yields one large very intense fragment (about 17 to 18 kb) and two smaller less intense fragments of near 9.5 kb (pSC101) and 8.1 kb (derived from E. coli), respectively. Since the size of the covalently closed circular form of pPG2 is only slightly smaller than the same form of pPG1 (data not shown), the large linear fragment of pPG2 (lane A, Fig. 1) results from a single cut of pPG2, and the two smaller fragments result from a second slower cut (at a strong EcoRI* site) within the large fragment. This conclusion is further supported by EcoRI* digestion of pPG2 (lane B) which results in the two smaller fragments of pPG2 (lane A) being the largest and most intense fragments, with many smaller fragments also being formed. TABLE 3. Phosphatidylglycerophosphate synthase levels induced by plasmids carrying pgsA Recipient strain

Plasmid

Sp act"

1.1 ADIOO 11.7 ± 1.1 pPG1 AD100 11.6 ± 1.4 pPG2 AD100