Chlamydia trachomatis Serovar L2 Can Utilize Exogenous Lipoic Acid through the Action of the Lipoic Acid Ligase LplA1

Bacterial strains and growth conditions.The strains used in the present study are listed in Table 1. Strains were grown in Luria-Bertani (LB) medium or M9 minimal salts supplemented with 0.5% mannitol at 37°C. M9 minimal medium was supplemented with 0.5% mannitol as the carbon source, because gene expression under the ara promoter in pBAD18 used as the vector for complementation assays is repressed in the presence of glucose. Growth medium was supplemented with antibiotics for selection at the following concentrations: ampicillin (100 μg/ml), kanamycin (50 μg/ml), and chloramphenicol (25 μg/ml). When required, media were also supplemented with arabinose (0.2%), IPTG (isopropyl-β-d-thiogalactopyranoside; 1 mM), sodium acetate (10 mM), sodium succinate (10 mM), or lipoic acid (10 μM). Lipoic acid was obtained from Sigma. The lipoylated tripeptide DK_LA was purchased from AnaSpec.

Quantitative determination of plating efficiency of the ligases was done as follows. Strains from exponential-phase cultures were washed three times in 1× phosphate-buffered saline (PBS; 1.05 mM KH₂ PO₄, 154 mM NaCl, 5.6 mM Na₂HPO₄), serially diluted, and plated on M9 mannitol medium without exogenous lipoic acid or M9 mannitol with arabinose and lipoic acid (10 μM). Efficiency of plating was calculated as the percentage ratio between the number of colonies obtained on plates without lipoic acid and those growing on plates with lipoic acid.

For growth curves, single colonies from appropriate LB plates were inoculated into M9 mannitol minimal with 0.2% arabinose, 10 mM sodium acetate, and 10 mM sodium succinate and then cultured overnight at 37°C. About 1 ml of the overnight culture was then washed three times in PBS and inoculated into fresh M9 mannitol medium containing 0.2% arabinose with or without 10 μM lipoic acid. Growth curves were performed by using a Bioscreen growth curve analyzer. The standard deviation represents the growth measurements of five replicates.

Strain and plasmid construction.ATM967 (E. coli ΔlplA ΔlipB::Kan^r) and ATM1102 (E. coli ΔlplA ΔlipA::Kan^r) were constructed by allelic exchange as described previously (12). Briefly, the kanamycin resistance cassette region (kan) from pKD4 (Table 1) with 5′ and 3′ overhangs identical to the 5′ and 3′ regions of E. coli lplA, lipA, and lipB (Table 1) was amplified by PCR. All primers are listed in Table 2. E. coli BW25113 containing pKD46 was used as the recipient for linear transformation for mutant construction. Kanamycin-resistant (Kan^r) recombinants were purified and screened via PCR to ensure proper insertion of the cassette. P1L4 lysates grown on the ΔlplA::kan strain and ΔlipB::kan strain were used to transduce the standard prototroph, MG1655 with selection for Kan^r. To construct the double mutant, the ΔlplA::kan allele was transduced into MG1655 and the transductant was transformed with pCP20 which expresses the Flp recombinase and mediates recombination between FRT sequences to excise the kan cassette. The Kan^s strain thus obtained was transduced with the ΔlipB::kan allele and selected on M9 minimal medium supplemented with mannitol, sodium acetate, sodium succinate, and kanamycin. The ΔlplA ΔlipB::kan double mutant was then stocked as ATM967. A P1L4 lysate grown on the BW25113 ΔlipA strain was used to transduce the MG1655 Kan^s ΔlplA strain to generate the ΔlplA ΔlipA::kan double mutant, which was stocked as ATM1102 (ΔlplA ΔlipA::kan).

All plasmids used in the present study are described in Table 1. Plasmids pLplA_Ec, pLipA_Ec, and pLipB_Ec contain an optimized E. coli ribosome binding site (RBS) and complete open reading frames of E. coli lplA, lipB, and lipA, respectively, under the control of the ara promoter in pBAD18. Similarly, pLplA1_Ct, pLplA2_Ct, and pLipA_Ct contain the same optimized RBS and complete open reading frames of ctl0537, ctl0761, and ctl0821, respectively, from C. trachomatis serovar L2 under the control of the ara promoter in pBAD18. All constructs were verified by sequencing at the Biomedical Instrumentation Center of the Uniformed Services University.

Protein expression and purification.LplA1_Ct and LplA2_Ct were amplified by PCR from C. trachomatis serovar L2 genomic DNA using Vent (New England Biolabs) and ligated into the XhoI site of the expression vector pGEX-6-P1 (GE Life Sciences), verified by sequencing, and transformed into ATM967 (ΔlplA ΔlipB::kan). ATM967 containing pGEX-6-P1::LplA1_Ct or pGEX-6-P1::LplA2_Ct was induced with 1 mM IPTG at room temperature overnight. Glutathione S-transferase (GST)-tagged proteins were then purified from bacterial lysates by using GST beads (GE Life Sciences) according to the manufacturer's instructions for batch purification. The N-terminal GST tag was cleaved from the recombinant protein by using PreScission protease (GE Life Sciences) and dialyzed against buffer containing 50 mM sodium phosphate (pH 7.6), 2 mM EDTA, 100 mM NaCl, 10 mM MgCl₂, and 10% glycerol. The BCKDH-E2 subunit encoded by ctl0657 was amplified from C. trachomatis serovar L2 by PCR and also cloned into pGEX-6-P1 by using the XhoI restriction site. Recombinant BCKDH-E2 subunit was purified as described above, and the cleaved protein was dialyzed into 1× PBS plus 10% glycerol. LplA_Ec was purified as a His-tagged protein according to a previously described report (22). Briefly, E. coli BL21DE3 expressing LplA_Ec with an N-terminal His tag was grown at 37°C until reaching an optical density at 600 nm (OD₆₀₀) of ∼0.5 and induced with 1 mM IPTG for 3 h at the same temperature. LplA_Ec was then purified under native conditions using HisPur cobalt resin (Pierce) according to the manufacturer's recommendations. Purified LplA_Ec was then dialyzed into buffer containing 20 mM sodium phosphate (pH 7.6), 2 mM EDTA, and 5% glycerol. All purified enzymes were stored at −80°C.

The purification of apo-PDH from ATM967 (ΔlplA ΔlipB::kan) was adapted from that of an earlier study (23). Briefly, ATM967 was grown in 100 ml of M9 minimal medium supplemented with mannitol, acetate, and succinate at 37°C. The cells were pelleted and resuspended in 12 ml of 0.1 M potassium phosphate (pH 7), lysed by sonication, and centrifuged for 30 min at 39,000 × g at 4°C. The supernatant was then centrifuged at 236,000 × g in a Beckman 45 Ti rotor for 8 h. The resulting pellet was resuspended in 5 ml of 50 mM potassium phosphate buffer and centrifuged for an additional 8 h at 236,000 × g. The supernatant was discarded, and the pellet was resuspended in 1 ml of potassium phosphate buffer containing 10% glycerol and stored at −80°C. Concentrations of all recombinant proteins were measured by using a BCA protein assay kit (Pierce).

Western blot analysis.For Western blot analyses, samples were resolved by SDS-PAGE and blotted. Lipoylated proteins were detected using α-lipoic acid antibody (1:10,000; Calbiochem), followed by the addition of donkey anti-rabbit IgG antibody conjugated to horseradish peroxidase (1:1,000; GE Life Sciences). Western blots were developed using the Pico Super Signal system (Pierce) and visualized by using a Fuji Intelligent Dark Box and Image Reader LAS-3000 software. For equal loading of samples, the protein concentration (A ₂₈₀) was determined by using a NanoDrop ND-1000 spectrophotometer.

Lipoic acid ligase assays.For lipoic acid ligase activity, purified LplA1_Ct (1 μM) or LplA_Ec (0.3 μM) was incubated in 10 mM sodium phosphate (pH 7) containing 5 mM ATP, 0.3 mM dithiothreitol, and 5 mM MgCl₂. E. coli apo-PDH/2-OGDH (1 μg) and chlamydial BCKDH-E2 (2 μM) were used as substrates in the reaction, and lipoic acid was used at a concentration of 1 mM. The reaction was incubated at 37°C for 1 h and analyzed by SDS-PAGE, followed by Western blotting with anti-lipoic acid antibody as described above.

Sequence alignments of LipA_Ct, LplA1_Ct, and LplA2_Ct.Genes encoding for putative enzymes involved in de novo lipoic acid biosynthesis (lipA) and lipoic acid salvage (lplA) were identified by BLASTp using the E. coli homologs against the C. trachomatis serovar L2 genome (42). The LipA homolog (LipA_Ct) shares 43 and 42% identity with lipoyl synthases from E. coli and Bacillus subtilis, respectively. LipA_Ct possesses both sets of conserved cysteine residues involved in coordinating iron-sulfur (Fe-S) clusters, the first of which is common to all radical SAM superfamily enzymes and the second of which is unique to lipoic acid synthases (40) (Fig. 2A).

• Open in new tab
• Download powerpoint

FIG. 2.

CLUSTAL W sequence alignments of chlamydial proteins predicted to be involved in lipoate utilization and synthesis with representative proteins. (A) The putative chlamydial LipA (LipA_Ct) was aligned with lipoic acid synthases from Escherichia coli (E.c), Bacillus subtilis (B.s), Saccharomyces cerevisiae, (S.c), plant Arabidopsis thaliana (A.t), and Homo sapiens (H.s). Residues common to SAM radical super family of enzymes (I) and residues specific for lipoic acid synthases (II) are highlighted in gray. (B) Alignment of the putative lipoic acid ligases from C. trachomatis serovar L2 (LplA1_Ct and LplA2_Ct) with ligases from E. coli (E.c.) and Thermoplasma acidophilum (T.a.). Conserved regions are highlighted in gray, and residues involved in lipoyl-AMP binding are indicated by an inverted triangle (▾).

C. trachomatis serovar L2 also possesses two lipoic acid ligases, LplA1_Ct and LplA2_Ct (encoded by ctl0537 and ctl0761), that are only 30 and 27% identical to E. coli LplA (Fig. 2B) as determined by BLASTp analyses. These enzymes possess the typical BPL_LplA_LipB domain, which is characteristic to this family of enzymes (35). Although LplA1_Ct showed homology to ligases from only prokaryotic organisms, LplA2_Ct appeared to be more closely related to plant lipoic acid ligases (8). However, the two ligases did not show any similarity to the E. coli LipB transferase. Orthologs of both ligases have been identified in all of the sequenced chlamydiae, suggesting that they are a well-conserved family of genes (Table 3).

Evidence of lipoylation in C. trachomatis serovar L2 EBs and RBs.Chlamydiae possess at least three proteins that are predicted to undergo posttranslational modification with lipoic acid, including the E2 subunits of PDH, 2-OGDH, and BCKDH (encoded by ctl0499, ctl0311, and ctl0657). All of the enzyme subunits are predicted to possess a single lipoyl domain (13). In order to examine whether C. trachomatis serovar L2 possesses lipoylated proteins, renografin purified RBs and EBs were analyzed with a polyclonal antibody specific for lipoylated proteins (37). Western blot analysis showed that the anti-lipoic acid antibody recognized lipoylated proteins in both RB and in EB fractions (Fig. 3). The sizes of three of these lipoylated proteins are close to the predicted sizes of the chlamydial E2 subunits of PDH (∼46 kDa), 2-OGDH (∼40 kDa), and BCKDH (∼42 kDa). The lipoylated protein (∼14 kDa) present only in RBs is consistent with the predicted size of the GcsH subunit encoded by ctl0534.

• Open in new tab
• Download powerpoint

FIG. 3.

Detection of lipoylated proteins in C. trachomatis serovar L2 RBs and EBs. Equal concentrations of Renografin-purified EBs and RBs were resolved by SDS-PAGE blotted and analyzed by Western blotting with anti-lipoic acid antibody. The sizes of E2 subunits of pyruvate dehydrogenase (PDH-E2), 2-oxoglutarate dehydrogenase (2-OGDH-E2), branched-chain keto-acid dehydrogenase (BCKDH-E2) and that of the glycine cleavage system protein H (GcsH) are indicated on the blot.

Functionality of lipA_Ct , lplA1_Ct , and lplA2_Ct .Currently, the necessary tools to generate targeted knockout mutants in chlamydiae is lacking; thus, E. coli is used as the surrogate host for functional characterization of relevant genes. Toward this end, we first constructed a ΔlplA ΔlipA mutant of E. coli that was deficient in lipoic acid synthesis (ΔlipA) and utilization of exogenous lipoic acid (ΔlplA) but possesses a functional lipB. The resulting strain, ATM1102 (ΔlplA ΔlipA::kan) has an absolute requirement for acetate and succinate which provide metabolic bypasses to enzymes that require lipoic acid and served as the background for determining the functionality of the chlamydial lipoic acid synthase (lipA_Ct ). Because ATM1102 (ΔlplA ΔlipA::kan) possess a functional lipB gene, complementation with E. coli lipA (lipA_Ec ) or functional chlamydial lipoyl synthase (lipA_Ct ) should restore de novo lipoic acid biosynthesis, thereby allowing bacteria to grow on M9 mannitol minimal medium in the presence of arabinose. ATM1102 was transformed with pLipA_Ec or pLipA_Ct expressing lipA_Ec and lipA_Ct , respectively, under the control of the arabinose-inducible promoter in pBAD18. Transformants were selected on M9 mannitol medium supplemented with acetate, succinate, and ampicillin. Select transformants were subsequently examined for growth in the presence of arabinose but in the absence of exogenous lipoic acid. Our complementation studies showed that upon induction with arabinose, lipA_Ec was able to complement the mutant strain (100% by efficiency of plating), whereas no complementation was observed with lipA_Ct (<10⁻⁴% by efficiency of plating) under similar induction conditions. Similar results were obtained when complementation analysis was repeated with lipA_Ct cloned into a high-copy vector (data not shown).

Next, we sought to assess the functionality of the two chlamydial ligase genes, lplA1_Ct and lplA2_Ct . For this, a second double mutant of E. coli (ΔlplA ΔlipB) was generated, and it was designated ATM967. This strain lacks the ability to synthesize lipoic acid (ΔlipB), as well as utilize exogenous lipoic acid (ΔlplA), and provides a better background for assessing ligase activity. To assay for putative ligase activity, ATM967 (ΔlplA ΔlipB::kan) was transformed with pLplA_Ec, pLplA1_Ct, and pLplA2_Ct expressing lplA_Ec , lplA1_Ct , and lplA2_Ct , respectively, under the control of arabinose-inducible promoter in pBAD. Transformants were screened for their ability to grow on M9 minimal medium in the presence of mannitol, 10 μg of lipoic acid/ml, and 0.2% arabinose. LplA_Ec was able to restore the ability to utilize exogenous lipoic acid (100% by efficiency of plating), while LplA1_Ct complemented at an efficiency of 94% determined by the same method. However, no complementation was observed with LplA2_Ct (<10⁻⁵%).

In order to further examine the functionality of the ligases and to see whether there was a delayed growth of the LplA2_Ct transformed strain, the growth rates of ATM967 (ΔlplA ΔlipB::kan) expressing LplA_Ec, LplA1_Ct, and LplA2_Ct were measured in minimal medium with or without lipoic acid. As seen in Fig. 4A, in the absence of lipoic acid, the pLplA1_Ct and the pLplA_Ec transformed strains showed growth kinetics that were identical to each other and grew faster than MG1655 transformed with pBAD18, presumably due to overexpression of the ligases. The MG1655 transformed with pBAD18 exhibited a longer lag most likely due to the presence of only a single copy of the chromosomal ligase. However, the pLplA_Ec- and pLplA1_Ct-expressing ATM967 and the MG1655 control strains all had the same slope in exponential phase. A similar trend was seen for these strains even in the presence of lipoic acid (Fig. 4B), with growth kinetics of the LplA_Ec- and LplA1_Ct-expressing strains being slightly faster than that of the MG1655 control. The growth rate of the pLplA2_Ct transformed strain was similar in the presence or absence of lipoic acid, reaching an OD₆₀₀ of only ∼0.3 even after 46 h of growth. The double mutant, ATM967, transformed with pBAD18 showed no growth throughout the experiment with or without lipoic acid.

• Open in new tab
• Download powerpoint

FIG. 4.

Growth of lipoic acid ligase mutant and transformed strains in the presence or absence of lipoic acid. Strains (see inset) were grown in M9 mannitol medium containing 0.2% arabinose at 37°C in the absence (A) or presence of 10 μM lipoic acid (B). The absorbance (OD₆₀₀) was measured in a Bioscreen growth curve analyzer and plotted against time (hours). Error bars represent the standard deviation from five replicates.

In order to confirm protein lipoylation in vivo, lysates of ATM967 (ΔlplA ΔlipB::kan) expressing lplA_Ec , lplA1_Ct , and lplA2_Ct were blotted with the anti-lipoic acid antibody. Wild-type E. coli is known to possess three lipoylated proteins: PDH-E2 (65 kDa), 2-OGDH-E2 (45 kDa), and GcsH protein (∼15 kDa). As shown in Fig. 5, two lipoylated proteins corresponding to PDH-E2 and 2-OGDH-E2 were detected in the E. coli MG1655 lysates and in strains expressing lplA_Ec and lplA1_Ct . The glycine cleavage protein GcsH was not observed in these lysates because the cultures were not induced by glycine. Interestingly, in lysates expressing lplA_Ec and lplA1_Ct , lipoylation was also seen in the absence of exogenous lipoic acid, suggesting that the LplA1_Ct may also be capable of utilizing endogenous octanoic acid as a substrate. The octanoylated apo-domain may then be modified into lipoylated holo-domain by the functional LipA in ATM967 (ΔlplA ΔlipB::kan) (Fig. 1). This observation is consistent with our growth curve studies and previous reports that have shown that LplA_Ec can also utilize free octanoic acid as a substrate (21, 28, 30). Therefore, LplA1_Ct appears to function in a manner similar to that of LplA_Ec.

• Open in new tab
• Download powerpoint

FIG. 5.

Detection of lipoylated proteins in lipoic acid ligase mutant and transformed strains. Lysates of transformed strains (LplA_Ec, LplA1_Ct, and LplA2_Ct), MG1655 (positive control), and ATM967 (negative control) grown in minimal medium with or without lipoic acid were analyzed for the presence of lipoylated proteins. Equal concentrations of the lysates were separated by SDS-PAGE, blotted, and probed with anti-lipoic acid antibody. Lipoylated proteins corresponding to PDH-E2 and 2-OGDH-E2 of E. coli were observed in the strains expressing LplA_Ec and LplA1_Ct in the presence or absence of lipoic acid but not in the LplA2_Ct-expressing strain.

We also examined the ability of LplA2_Ct to utilize exogenous octanoic acid. ATM967 (ΔlplA ΔlipB::kan) was transformed with pLipB_Ec, pLplA_Ec, or pLplA2_Ct and examined for its ability to grow on M9 mannitol medium supplemented with 50 μM octanoic acid. Although pLipB_Ec (98% efficiency of plating) and pLplA_Ec (80% efficiency of plating) were able to complement the double mutant upon induction with arabinose, no complementation was observed with pLplA2_Ct (<10⁻⁵% efficiency of plating). In vivo lipoylation was not seen in cells expressing LplA2_Ct in the presence or absence of lipoic acid (Fig. 4) and confirms the negative complementation data.

In vitro lipoylation activity of LplA1_Ct.For in vitro assays, we used lysates of ATM967 (ΔlplA ΔlipB::kan) enriched for apo-PDH and 2-OGDH and recombinant chlamydial BCKDH-E2 as lipoate acceptors. The latter was overexpressed in ATM967 as a GST-tagged protein so that it could purified in the apo-form. Consistent with our in vivo complementation data, recombinant LplA1_Ct was able to lipoylate apo-PDH and 2-OGDH-E2 subunits purified from E. coli ATM967 (Fig. 6A, left) as detected by Western blotting with anti-lipoic acid antibody. LplA1_Ct was also able to lipoylate the recombinant chlamydial BCKDH-E2 subunit (Fig. 6B, left). Purified LplA_Ec served as a positive control for both reactions (Fig. 6A and B, right). Recombinant LplA2_Ct was unable to lipoylate either substrate, suggesting that this enzyme was inactive under the conditions in which it was tested (data not shown).

• Open in new tab
• Download powerpoint

FIG. 6.

In vitro lipoic acid ligase activity. Purified LplA_Ec (positive control) and LplA1_Ct were assayed for activity in the presence or absence of various components as indicated. Purified LplA1_Ct was assayed for activity using a mixture of E. coli apo-PDH/2-OGDH (panel A, left) and recombinant chlamydial BCKDH-E2 (panel B, left) as substrates at 37°C for 1 h. Reactions were analyzed by SDS-PAGE and blotted, and lipoylated proteins were detected using anti-lipoic acid antibody. Purified LplA_Ec served as a positive control and was also tested with apo-PDH/2-OGDH (panel A, right) and chlamydial BCKDH-E2 (panel B, right).

Previous reports have described the ability of lipoic acid ligases from Listeria monocytogenes to enable the utilization of lipoylated tripeptides as a source of lipoic acid (24). In order to determine whether the chlamydial ligases could function in a similar manner, we used a lipoylated tripeptide consisting of aspartate-lysine [lipoic acid]-alanine (DK_LipA) as a source for lipoic acid in the in vitro assay. However, LplA1_Ct and LplA2_Ct were not able to utilize DK_LipA to lipoylate the apo-enzyme substrates (data not shown).