The BB0646 protein demonstrates lipase and haemolytic activity associated with Borrelia burgdorferi, the aetiological agent of Lyme disease

Abstract

The etiological agent of Lyme disease, Borrelia burgdorferi, is transmitted by ticks of the Ixodes genus and, if untreated, can cause significant morbidity in affected individuals. Recent reports have shown that polyunsaturated fatty acids in the B. burgdorferi cell envelope are potential targets for oxidative damage, which can be lethal. How B. burgdorferi responds to this assault is not known. Herein we report evidence that bb0646 codes for a lipase that is located within the bosR operon and that has specificity for both saturated and polyunsaturated fatty acids. Specifically, strains harbouring mutated copies of the lipase, either in the form of an insertionally inactivated construct or site-directed mutations within the active site, demonstrated attenuated lipolytic and haemolytic phenotypes when compared with the isogenic parent and trans-complements. In vivo analysis showed that while the bb0646 mutant remains infectious, the spirochaetal load is significantly lower than both the isogenic parent and the complemented mutant strains. Taken together, these data demonstrate that BB0646 is a broad substrate specific lipase that contributes to lipolytic and haemolytic activity in vitro and is required for optimal B. burgdorferi infection.

Figure 1

Isolation and confirmation of bb0646::gent^R mutants in B. burgdorferi B31 derivative, ML23. (A) Schematic diagram of the bb0646 insertional mutation of DS102. bb0646 was interrupted using a P_flgB-gent^R cassette at a unique NsiI restriction site within the gene. Constructs were transformed into ML23 and putative positive clones were screened by PCR using the indicated primers, 1–4, in panel B. (B) Putative transformants were screened with 3 sets of primers: DS111F/R (1 and 2), DS111-NdeI-F/gent_int- R (1 and 4), and DS111-NdeI-R/gent_int-F (2 and 3). (see Table 2 for primers used). Values shown on the left represent markers (in bp). (C) Southern blot confirmed the presence of the bb0646::gent^R allele via hybridization with a gent^R probe. DNA from B. burgdorferi strains ML23 and DS102 was digested with EcoRV and RsaI as indicated (restriction enzyme sites shown in panel A). Values shown on the left represent markers (in bp). (D) Western blot analysis demonstrated that putative transformants made no detectable BB0646 protein compared to the isogenic parent strain ML23 pBBE22 when B. burgdorferi lysates were analyzed by immunoblot analysis with polyclonal antibody specific for BB0646. An asterisk (*) denotes a non-specific, cross-reactive band that anti-BB0646 recognizes in all isolates. Values shown on the left represent markers (in kDa).

Figure 2

Complementation of the bb0646::gent^R insertional mutant, DS102, with the intact bb0646 gene in trans. (A) Transcriptional fusions of bb0646 to either its native promoter (P_bosR; pDS113) or the strong constitutive flagellar promoter (P_flgB; pDS126) were cloned into the borrelial shuttle vector pBBE22Gate. (B) Primer pair pncA_nested-F/pDS113_R-confirm (1 and 2; Table 2) were used to confirm the presence of pDS113 in DS102 isolates by PCR. Likewise, primer pair pncA_nested-F/pDS126_R-confirm (1 and 3; Table 2) was used to validate whether pDS126 was present in DS102. Values shown on the left represent markers (in bp). (C) Putative transformants were screened for the presence of BB0646 by Western immunoblot analysis using polyclonal anti-BB0646.

Figure 3

BB0646 functions as a lipase with specificity for both saturated and polyunsaturated fatty acid substrates. (A) Whole cell lysates from ML23/pBBE22, DS102/pBBE22, DS102/pDS113 and DS102/pDS126 were assayed using p-nitrophenyl palmitic acid as a substrate. Note the statistically significant decrease in lipase activity for DS102/pBBE22 relative to the isogenic parent and the complement DS102/pDS126. The dual asterisks (**) denote a P value of < 0.01. (B) Whole cell lysates of either ML23/pBBE22, DS102/pBBE22, DS102/pDS113 or DS102/pDS126 were incubated with the fluorogenic substrate, 7-hydroxycoumarinyl linolenic acid over a period of 90 min. and read at 10 min. intervals. Note that the bb0646::gent^R mutant (DS102/pBBE22; open circles) was nearly devoid of activity whereas the parent (ML23/pBBE22; closed circles) both complements (DS102/pDS113; inverted open triangles and DS102/pDS126; closed triangles) showed restoration of lipase activity with highest activity seen for DS102/pDS126. Bars indicate standard error. Time points from each strain were tested for significance; **, and *** denote P values < 0.001, and 0.0001 respectively.

Figure 4

Mutagenesis of the putative active site serine of BB0646 abrogates lipase activity in B. burgdorferi, but not BB0646 protein production. (A) Production of BB0646 in point mutants. The levels of BB0646 are compared between the parent (ML23/pBBE22), mutant (DS102/pBBE22), complement (DS102/pDS126), and the point mutants within the GXSXG conserved motif (DS102/pDS127; contains the bb0646-S152A allele; and DS102/pDS128; contains the bb0646-S152T allele). Whole cell equivalents were resolved by SDS-PAGE, immunoblotted and probed with antisera to BB0646. (B) Mutagenesis at serine 152 of bb0646 reduces recognition of saturated fatty acid substrate. The B. burgdorferi strains indicated above were tested for lipolytic activity against p-nitrophenyl palmitate as indicated in Fig. 3A. Note the statistically significant decrease in lipase activity for strains carrying the S152A (DS102/pDS127) and S152T (DS102/pDS128) alleles of bb0646 relative to the parent (ML23/pBBE22) and complement (DS102/pDS126). The activity observed for the bb0646 point mutants is commensurate with that observed for the bb0646 mutant (DS102/pBBE22). The single and dual asterisks denote a P value of < 0.05 and 0.01, respectively. (C) bb0646 S152 mutants do not recognize polyunsaturated fatty acid substrates. Whole cell lysates from the B. burgdorferi strains were tested for lipolytic activity against 7-HC linolenate as indicated in Fig. 3B. Samples from the parent (ML23/pBBE22; closed circles), mutant (DS102/pBBE22; open circles), the complement (DS102/pDS126; closed triangles), and the bb0646-S152 mutants (DS102/pDS127 [open triangles] and DS102/pDS128 [open squares]) were tested for their ability to cleave 7-HC linolenate over time. Cleavage of the substrate releases a fluorogenic reporter and the increased fluorescence is plotted as relative fluorescent units (RFU). Significance was measured at each time point between the bb0646 mutant and S152 mutant complements relative to the parent strain and the native complement. Bars indicate standard error. Single, dual, and tri asterisks denote P values of < 0.01, 0.001, and 0.0001 respectively.

Figure 5

BB0646 is required for maximal hemolytic activity in B. burgdorferi. (A) The parent (ML23), bb0646 mutant (DS102), both complements (DS102/pDS113 and DS102/pDS126), and both bb0646 S152 point mutation complements (DS102 carrying the bb0646-S152A and bb0646-S152T alleles, DS102/pDS127 and DS102/pDS128, respectively) were evaluated for hemolytic activity. The bb0646 mutant and point mutation complements exhibited attenuated β–hemolysis while both complement strains restored the deficiency to that observed in the parent strain. (B) Semi-quantitative assessment of borrelial-mediated hemolysis. Colonies from each strain were scored for partial β–hemolysis and α–hemolysis and with values listed as a percentage of the total hemolysis observed. Note that the S152 mutant complements exhibit hemolytic activity indistinguishable from the bb0646 mutant alone (DS102). Both are significantly different from that observed for the parent and P_flgB-bb0646 complement (P value < 0.05). Bars indicate standard error. Each strain represents an n of 6.

Figure 6

BB0646 is a detergent phase associated protein that is not surface localized. (A) Phase partitioning analysis of BB0646. Whole cell lysates from ML23 and DS102 were subjected to Triton X-114 phase partitioning and following exposure to high salt (i.e., 1M NaCl, 1M Tris, or 0.1 M Na₂CO₃). Detergent and aqueous phases were compared to whole cell lysates by Western immunoblotting with BB0646 antiserum. Detergent and aqueous phase proteins from 10⁸ whole cell equivalents were loaded into each lane. An arrow indicates the position of BB0646. Numbers on the left refer to the molecular mass of protein makers (in kDa). (B) Localization of BB0646. Strain ML23 was incubated with Proteinase K to assess whether BB0646 was surface exposed. Controls were either incubated in PBS alone (not treated; NT), permeabilized with Triton X-100 (TX-100), treated with Proteinase K alone (PK), or incubated with Proteinase K following exposure to Triton X-100 (TX-100 + PK). Following SDS-PAGE, samples were immunoblotted and probed with antisera against FlaB, P66, and BB0646 (as indicated on the right). Numbers on the left refer to the molecular mass of protein makers (in kDa).

Figure 7

Loss of BB0646 partially attenuates B. burgdorferi infectivity. C3H/HeN mice were infected with B. burgdorferi strains at either 10³ or 10⁵ inoculum doses. Tissues were aseptically removed after 3 weeks of infection and cultivated in BSK-II media. Cultures were evaluated for the presence of B. burgdorferi after 6–14 days. Data is expressed as percentages of culture-positive samples for each strain and dose tested. Each data set represents data obtained from 4–5 mice (depending on the strain).

Figure 8

Quantitative PCR shows a lowered infectious load for the bb0646 mutant at a low inoculum dose. Quantitative PCR was performed on (A) joint (B) lymph node and (C) skin tissues removed from mice that were infected with 10³ organisms to determine the absolute number of B. burgdorferi genomic equivalents in each tissue sample. Copies of recA were quantified and represent the number of total B. burgdorferi genomes. To normalize the data to host tissue, copies of β-actin were also enumerated and data points are expressed as copies of B. burgdorferi genomes per 10⁶ copies of β-actin. Bars indicate average value of the sample tested. The asterisks denote a P value < 0.05.