Lipoic acid synthesis: a new family of octanoyltransferases generally annotated as lipoate protein ligases

Abstract

Bacillus subtilis lacks a recognizable homologue of the LipB octanoyltransferase, an enzyme essential for lipoic acid synthesis in Escherichia coli. LipB transfers the octanoyl moiety from octanoyl-acyl carrier protein to the lipoyl domains of the 2-oxoacid dehydrogenases via a thioester-linked octanoyl-LipB intermediate. The octanoylated dehydrogenase is then converted to the enzymatically active lipoylated species by insertion of two sulfur atoms into the octanoyl moiety by the S-adenosyl-l-methionine radical enzyme, LipA (lipoate synthase). B. subtilis synthesizes lipoic acid and contains a LipA homologue that is fully functional in E. coli. Therefore, the lack of a LipB homologue presented the puzzle of how B. subtilis synthesizes the LipA substrate. We report that B. subtilis encodes an octanoyltransferase that has virtually no sequence resemblance to E. coli LipB but instead has a sequence that resembles that of the E. coli lipoate ligase, LplA. On the basis of this resemblance, these genes have generally been annotated as encoding a lipoate ligase, an enzyme that in E. coli scavenges lipoic acid from the environment but plays no role in de novo synthesis. We have named the B. subtilis octanoyltransferase LipM and find that, like LipB, the LipM reaction proceeds through a thioester-linked acyl enzyme intermediate. The LipM active site nucleophile was identified as C150 by the finding that this thiol becomes modified when LipM is expressed in E. coli. The level of the octanoyl-LipM intermediate can be significantly decreased by blocking fatty acid synthesis during LipM expression, and C150 was confirmed as an essential active site residue by site-directed mutagenesis. LipM homologues seem the sole type of octanoyltransferase present in the firmicutes and are also present in the cyanobacteria. LipM type octanoyltransferases represent a new clade of the PF03099 protein family, suggesting that octanoyl transfer activity has evolved at least twice within this superfamily.

Figure 1. Octanoylation of Lipoyl Domains

The model for octanoyl-LD synthesis by either octanoyltransfer or octanoyl ligation is shown. The reactions of the characterized octanoyltransferases proceed via an octanoyl-enzyme thioester intermediate whereas the characterized lipoyl ligases (which also function with octanoate) proceed via an enzyme bound acyl-adenylate. Octanoyl-LD is converted to lipoyl-LD by lipoyl synthase.

Figure 2

Complementation of E. coli Lipoic Acid Auxotrophs. Growth curves of lipoic acid auxotrophs carrying various plasmids are shown. Panel A. Growth tests for lipB complementation in minimal glycerol medium. Growth curves of the wild type (WT) and lipB lplA strains with the empty pCC1FOS vector are shown. Cosmid isolate 1 (Cos1) or pQC039 were also tested in the lipB lplA strain. Panel B. Growth tests for lipB complementation in minimal arabinose medium. Growth curves for the WT and lipB lplA strains carrying the empty vector pBAD322G are shown. Plasmids carrying the genes indicated were tested in a lipB lplA strain for their abilities to support growth. Panel C. Growth tests for lplA complementation in minimal arabinose medium containing lipoic acid. The WT and lipA lplA strains carrying the empty vector pBAD322G are shown. Plasmids carrying the genes indicated were tested in a lipA lplA strain for their abilities to support growth. Growth was measured by optical density. Panels B and C have the same color code.

Figure 3

Analysis of Purified Proteins. Panel A: SDS PAGE of 0.2 μmol each of purified LipM and GcvH. Molecular weight standards are indicated in kDa. Panel B: electrospray ionization mass spectrum of GcvH. Calculated masses are represented by black circles. Panel C: MALDI mass spectra of purified LipM preparations. LipM- denotes enzyme purified from cells of a wild type strain grown without triclosan or oleate whereas LipM+ denotes enzyme purified from cells of a ΔfabA strain grown with triclosan and oleate The peak mass values for the LipM- and LipM+ proteins were 33,300 and 33,179, respectively. Panel D: Size exclusion chromatogram of LipM (absorbance at 280 nm is plotted). The elution positions of chymotrypsinogen and ribonuclease A are designated by a triangle and a circle, respectively.

Figure 4

Liquid Chromatography Tandem Mass Spectrometric Analysis of LipM tryptic peptides. Panel A. Modification states of C150 detected. The theoretical peptides are shown with C150 in bold and underline type. The difference in mass from the modification is listed. The ion score is equal to -10log(P), where P is the probability the result is random chance. Panel B. Sequence coverage of LipM. The LipM peptide sequences detected are shown in bold type.

Figure 5

Gas Chromatography Mass Spectrometric Analysis of LipM Bound Octanoate. LipM-bound octanoyl moieties were assayed by transesterification to the butyl esters followed by GC-MS analysis as described in Experimental Procedures. Panel A: The values are the molar percentage of octanoic acid per LipM preparation. The grey bar indicates cultures without triclosan added (LipM) whereas the white bar indicates cultures with triclosan added (LipM+). The error bars represent one standard deviation for LipM preparations from three independent cultures. Both purifications were from a ΔfabA strain. Panel B: Representative gas chromatogram of a LipM preparation. The butyl heptanoate internal standard and the analyte, butyl octanoate, are indicated. Panel C: The mass spectrum of the butyl octanoate from a LipM preparation.

Figure 6

Octanoyltransferase Activities of Wild Type and Mutant LipM Proteins. Assay of octanoyl transfer from octanoyl-ACP to B. subtilis GcvH using either [1-¹⁴C]octanoyl-ACP or octanoyl-ACP as substrates as described in Experimental Procedures. Autoradiograms of dried SDS-PAGE gels are shown. Panel A: The [1-¹⁴C]octanoyl-ACP was synthesized from [1-¹⁴C]octanoate by AasS added to the reaction (ATP was also added as an AasS substrate). Synthesis of octanoyl-ACP required AasS, ATP and holo-ACP whereas formation of octanoyl-GcvH required apo-GcvH, LipM and octanoyl-ACP. Panel B: Transfer of octanoate from purified [1-¹⁴C]octanoyl-LipM to an equimolar amount of either GcvH or holo-ACP as indicated. Panel C: The same assay mixture from panel A with 10-fold more LipM or mutant LipM proteins added to enable detection of the octanoyl-LipM intermediate. The lanes are: NE, no enzyme; KR, LipM K165R; KA, LipM K165A; CA, LipM C150A; CS, LipM C150S and WT, LipM wild type. Panel D. A reaction as in lane 6 of panel A was run with octanoate in place of [1-¹⁴C]octanoate. After incubation GcvH was recovered from the reaction mixture by Ni^2+- cheleate chromatography and analyzed by electrospray ionization mass spectrometry. The mass obtained was 15,072.4, a value 128.2 amu greater than that of the apo protein (Figure 3) and in good agreement with the octanoyl moiety mass (126.2).

Figure 7

Comparison of the octanoyl thioesters of ACP and CoA as LipM Substrates. Assays of octanoyltransfer to B. subtilis GcvH using either octanoyl-CoA or octanoyl-ACP as described in Experimental Procedures. Panel A: Autoradiograms of dried SDS-PAGE gels are shown. [1-¹⁴C]octanoate was converted to [1-¹⁴C]octanoyl-ACP or [1-¹⁴C]octanoyl-CoA by AasS or acyl-CoA synthetase (AcsA), respectively. The presence (+) and absence (-) of LipM is indicated. Panel B: A double label experiment with a mixture of purified [1-¹⁴C]octanoate -CoA and purified ³H-octanoyl-ACP is shown. The substrates were column 1, [1-¹⁴C]octanoate -CoA; column 2, an equimolar mixture of both substrates and column 3, ³H-octanoyl-ACP. The error bars represent one standard deviation from six independent assays.

Figure 8

Alignment of LipM with Enzymatically Characterized LipBs and LplA Proteins. Alignments were performed as described in Experimental Procedures and displayed using Jalview (24) . The sequence name indicates the enzyme type, the Uniprot code indicates the organism of origin and the numbers indicate the amino acid residues displayed. Positions having 50% or greater similarity are highlighted in grey. The catalytic cysteine residues of the octanoyltransferases are boxed and highlighted in black, as is the conserved PF03099 lysine residue.

Figure 9

Phylogeny of LipM. The minimum evolution tree of selected PF03099 protein sequences with bootstrap percentage confidence values shown for each branch is given. Phylogenetic analyses were done as described in Experimental Procedures. The scale bar corresponds to a 20% difference in compared residues, on average, per branch length. Biotin protein ligase sequences were used as a related out-group to compare lipoate protein ligases, LipB octanoyltransferases and LipM octanoyltransferases.