Rhodobacter capsulatus OlsA is a bifunctional enzyme active in both ornithine lipid and phosphatidic acid biosynthesis

Abstract

The Rhodobacter capsulatus genome contains three genes (olsA [plsC138], plsC316, and plsC3498) that are annotated as lysophosphatidic acid (1-acyl-sn-glycerol-3-phosphate) acyltransferase (AGPAT). Of these genes, olsA was previously shown to be an O-acyltransferase in the second step of ornithine lipid biosynthesis, which is important for optimal steady-state levels of c-type cytochromes (S. Aygun-Sunar, S. Mandaci, H.-G. Koch, I. V. J. Murray, H. Goldfine, and F. Daldal. Mol. Microbiol. 61:418-435, 2006). The roles of the remaining plsC316 and plsC3498 genes remained unknown. In this work, these genes were cloned, and chromosomal insertion-deletion mutations inactivating them were obtained to define their function. Characterization of these mutants indicated that, unlike the Escherichia coli plsC, neither plsC316 nor plsC3498 was essential in R. capsulatus. In contrast, no plsC316 olsA double mutant could be isolated, indicating that an intact copy of either olsA or plsC316 was required for R. capsulatus growth under the conditions tested. Compared to OlsA null mutants, PlsC316 null mutants contained ornithine lipid and had no c-type cytochrome-related phenotype. However, they exhibited slight growth impairment and highly altered total fatty acid and phospholipid profiles. Heterologous expression in an E. coli plsC(Ts) mutant of either R. capsulatus plsC316 or olsA gene products supported growth at a nonpermissive temperature, exhibited AGPAT activity in vitro, and restored phosphatidic acid biosynthesis. The more vigorous AGPAT activity displayed by PlsC316 suggested that plsC316 encodes the main AGPAT required for glycerophospholipid synthesis in R. capsulatus, while olsA acts as an alternative AGPAT that is specific for ornithine lipid synthesis. This study therefore revealed for the first time that some OlsA enzymes, like the enzyme of R. capsulatus, are bifunctional and involved in both membrane ornithine lipid and glycerophospholipid biosynthesis.

FIG. 1.

PA and OL biosynthesis pathway in bacteria. (A) The first step for PA biosynthesis from G3P can be carried out by two different routes. In some bacteria, like E. coli, GPAT (PlsB) acylates the sn-1 position of G3P using either acyl-ACP or acyl-CoA to form LPA (step 1). In other bacteria, a recently identified route uses the soluble PlsX to convert acyl-ACP to acyl-phosphate (acyl-P), followed by the membrane-associated PlsY transferring the acyl chain to G3P (step 2). In all bacteria, the second step for PA biosynthesis is catalyzed by the membrane-associated AGPAT (PlsC) enzyme, which transfers an acyl chain from either acyl-ACP or acyl-CoA to LPA to yield PA. In R. capsulatus OlsA is alternative AGPAT enzyme for production of PA. (B) During OL biosynthesis, the first enzyme OlsB catalyzes the formation of an amide linkage (N-acyltransferase) between the α-amino group of ornithine and the carboxyl group of a 3-hydoxy fatty acid, forming LOL. The second enzyme, OlsA, catalyzes the formation of an ester linkage (O-acyltransferase) between the 3-hydroxy group of the fatty acyl group and the carboxyl of a second fatty acid, converting LOL to OL.

FIG. 2.

Comparison of various AGPAT homologues of R. capsulatus. The R. capsulatus (Rc) AGPAT homologues were aligned with the E. coli (Ec) and N. meningitidis (Nm) AGPAT sequences using the program ClustalW and presented using the BOXSHADE, version 3.21, software. Identical residues are shaded in black, and similar residues are shaded in gray. The catalytic (HX₄D) motif (24) and the substrate-binding (PEGTR) motif of GPATs and AGPATs are boxed and indicated by asterisks.

FIG. 3.

Characterization of plsC mutants. (A) Growth of wild-type (wt), olsA (SA4), plsC316 (SA13), and plsC3498 (SA11) null mutants on MPYE medium at 35°C under aerobic conditions after 2 days of incubation. (B) Growth of plsC316 mutant harboring a plasmid with (SA13/pMRC) or without (SA13/pRK415) olsA under the same conditions as described for panel A. (C) Comparison of the c-type cyt profiles of R. capsulatus plsC316 and olsA mutants. Membrane fractions were isolated from cells grown at 35°C in MPYE medium, proteins were separated by using 16.5% tricine-SDS-PAGE, and the c-type cyt were revealed using tetramethylbenzidine, as described in Materials and Methods. The c-type cyt subunits of the cbb₃-Cox (c_p and c_o), the cyt bc₁ complex (c₁), and the electron carrier cyt c_y (c_y) are indicated on the left together with the 32.5- and 25-kDa molecular size markers.

FIG. 4.

Total lipid composition of plsC316 and plsC3498 null mutants of R. capsulatus. In all cases, total polar lipids were extracted from [1-¹⁴C]acetate-labeled cells, similar amounts (60,000 cpm) were deposited on TLC plates, and 2D-TLC analyses were carried out as described in Materials and Methods. DGTS, diacylglyceryl trimethyl-homoserine; DMPE, phosphatidyl-N,N dimethylethanolamine. The vertical and horizontal arrows at the origin O refer to the first and second dimension of solvent migrations, respectively. The radioactivity associated with each spot was determined and is given in Table 2.

FIG. 5.

Expression of R. capsulatus olsA and plsC316 in E. coli. (A) The E. coli plsC(Ts) strain SM2-1 harboring plasmids carrying olsA, plsC316, or plsC3498 of R. capsulatus was grown on 2% l-arabinose-containing LB plates at 42°C to score heterologous complementation. SM2-1 cells carrying the cloning vector pBAD/Myc-His A were used as a control. (B) Expression of R. capsulatus olsA and plsC316 in E. coli plsC(Ts) mutant SM2-1 cells before (0) and after (2) induction with 2% l-arabinose for 4 h at 30°C. Following induction cells were resuspended in 2× SDS loading buffer, and expressed proteins were detected by SDS-PAGE and immunoblotting using anti-Myc antibody as described in Materials and Methods. The triangles point out the R. capsulatus OlsA and PlsC316 proteins (31 and 29.5 kDa, respectively) together with the 32.5- and 25-kDa molecular mass markers.

FIG. 6.

GPAT-AGPAT activities exhibited by appropriate E. coli plsC(Ts) mutants harboring R. capsulatus plsC homologues as well as R. capsulatus wild-type, olsA, and plsC316 mutants. (A) Time course assays of GPAT-AGPAT activities in E. coli plsC mutant harboring olsA or plsC316 were performed using radioactive G3P, vaccenyl-ACP, and membrane particles (prepared as described in the supplemental material) from SM2-1 cells grown at 30°C (SM2-1), SM2-1 cells grown at 30°C with a subsequent 30-min incubation at 42°C (SM2-1*), SM2-1 cells harboring olsA, and SM2-1 cells harboring plsC316, as described in Materials and Methods. The data shown are the means of two independent experiments with the standard errors, as indicated. (B) Assays similar to those shown in panel A were performed at 35°C for 5 min, and labeled lipids (approximately 7,000 cpm total) were extracted and separated by 1D-TLC, as described in Materials and Methods. LPA and PA produced using membranes from SM2-1 cells grown at 30°C (lane 1), SM2-1 grown at 30°C with a subsequent 30-min incubation at 42°C (lane 2), SM2-1 cells harboring olsA (lane 3), and SM2-1 cells harboring plsC316 (lane 4) are shown. Note the absence of PA production in lane 2 and PA overproduction in lane 4. (C) Time course assays of GPAT-AGPAT activities in wild-type (wt), ΔolsA (SA4), and ΔplsC316 (SA13) strains were performed as described for panel A. The data shown are the means of two independent experiments with the standard errors as indicated. (D) Labeled lipids (approximately 2,000 cpm total) were prepared and separated by 1D-TLC, as described for panel B. Note that the PA produced using membranes from the wild-type strain MT1131 and the ΔolsA mutant are readily seen while that produced by the ΔplsC316 mutant is barely detectable.