Present address: Anton V. Bryksin, Department of Biochemistry, Rollins Research Center, Emory University School of Medicine, 1510 Clifton Road, Atlanta, GA 30322, USA.
Correspondence: Felipe C. Cabello, Department of Microbiology and Immunology, New York Medical College, Basic Science Building, Valhalla, NY 10595, USA Tel.: +1 914 594 4182; fax: +1 914 594 4176; e-mail: firstname.lastname@example.org
Bacterial pathogens face constant challenges from DNA-damaging agents generated by host phagocytes. Although Borrelia burgdorferi appears to have much fewer DNA repair enzymes than pathogens with larger genomes, it does contain homologues of uvrA and uvrB (subunits A and B of excinuclease ABC). As a first step to exploring the physiologic function of uvrABbu and its possible role in survival in the host in the face of DNA-damaging agents, a partially deleted uvrA mutant was isolated by targeted inactivation. While growth of this mutant was markedly inhibited by UV irradiation, mitomycin C (MMC) and hydrogen peroxide at doses that lacked effect on wild-type B. burgdorferi, its response to pH 6.0–6.8 and reactive nitrogen intermediates was similar to that of the wild-type parental strain. The sensitivity of the inactivation mutant to UV irradiation, MMC and peroxide was complemented by an extrachromosomal copy of uvrABbu. We conclude that uvrABbu is functional in B. burgdorferi.
Because many host defenses involve production of DNA-damaging reactive oxygen species (ROS) and reactive nitrogen species (RNS), the ability of pathogenic bacteria to repair damaged DNA is important to their survival in hosts. In Mycobacterium tuberculosis, uvrA mutants show decreased ability to survive within macrophages (Graham & Clark-Curtiss, 1999) and uvrB mutants are attenuated in mice (Darwin & Nathan, 2005). Similarly, in Helicobacter pylori and Yersinia sp., defects in uvrA are accompanied by attenuation in mice (Bijlsma et al., 2000; Garbom et al., 2004). These experimental results strongly suggest that lack of DNA repair mediated by the uvrA gene product attenuates bacterial pathogens because they cannot overcome the DNA-damaging systems of the host (Janssen et al., 2003).
The genome of Borrelia burgdorferi, the cause of Lyme disease, contains a minimal set of genes devoted to DNA repair and appears to lack an SOS response despite the presence of orthologues of uvrA, uvrB, uvrD, DNA polymerase I and DNA ligase (Fraser et al., 1997). It also lacks an orthologue for the repressor of the SOS response, lexA, and none of the genes potentially involved in DNA repair display consensus LexA binding boxes similar to those found in E. coli (Fraser et al., 1997). recA also does not appear to be involved in repair of UV-induced DNA damage in B. burgdorferi (Liveris et al., 2004; Putteet-Driver et al., 2004). Borrelia burgdorferi is exposed to antibacterial levels of ROS and RNS in infected ticks (Pereira et al., 2001) and mammals (Benach et al., 1984; Cinco et al., 1997; Hellwage et al., 2001) intracellularly, following phagocytosis, and extracellularly, by diffusion from intracellular sources or by production at the phagocyte plasma membrane (Putteet-Driver et al., 2004). Borrelia burgdorferi can also be exposed to solar UVB radiation in the erythema migrans skin lesion (Born & Born, 1987). Borrelia burgdorferi must therefore have functional DNA repair systems to overcome these exposures if it is to survive and proliferate in its hosts.
The B. burgdorferi uvrA homologue (BB0837) encodes a protein of 950 amino acids (UvrABbu) whose deduced amino acid sequence has 23–54% homology to UvrA of Treponema pallidum, Leptospira interrogans, Bacillus subtilus and E. coli, and, like the others, contains two zinc finger motifs and two ATP-binding sites (Savery, 2007). The function of BB0837 has not been experimentally verified, and study of its function, expression and regulation in B. burgdorferi is therefore likely to shed light on its role in DNA repair and bacterial survival. To this end, we inactivated uvrABbu and found that the resulting B. burgdorferi disruption mutant was more sensitive to UV radiation, mitomycin C (MMC) and ROS than the parental strain. This increased sensitivity was reversed by extrachromosomal complementation with a wild-type copy of uvrABbu.
Materials and methods
Strains and culture conditions
Low-passage infectious B. burgdorferi 297, clone BbAH130, was obtained from Dr M. V. Norgard, University of Texas Southwestern Medical Center. PCR analysis using appropriate primers (Iyer et al., 2003) indicated that this clone contained lp25, but lacked lp28-1. Cultures were routinely grown at 34 °C in Barbour–Stoenner–Kelly medium supplemented with 6% rabbit serum (BSK-H) (Sigma Chemical Co., St. Louis, MO). Escherichia coli DH5α (Gibco/Life Technologies, Grand Island, NY) was routinely used for cloning, and was grown and maintained in Luria–Bertani medium.
DNA and RNA manipulations
Genomic DNA was isolated from pelletted B. burgdorferi grown at 34 °C to 3 × 108 cells mL−1 with High Pure PCR Template Preparation Kit (Roche Diagnostics Corporation, Indianapolis, IN) and total RNA was isolated using TRizol Reagent (Invitrogen Life Technology, Carlsbad, CA), both according to the manufacturer's instructions. Traces of genomic DNA were removed from isolated RNA by treatment with RNase-free DNase. RNA was dissolved in RNase-free water (Ambion, Austin, TX) and stored in aliquots at −80 °C. cDNA was generated by AMV reverse transcriptase with random primers using the Access RT-PCR system (Promega Corporation, Madison, WI). Controls with the omission of reverse transcriptase were always included in each experiment. PCR reactions were performed using Taq polymerase (Denville Scientific Inc., Metuchen, NJ) or Expend Long Template DNA polymerase mix (Roche Applied Science) using parameters according to Tm of primers. All constructs were confirmed by restriction enzyme analysis, PCR and DNA sequencing using standard procedures (Sambrook & Russell, 2001). The primers used in this study are listed in Table 1.
The uvrABbu inactivation construct (Fig. 1a) was generated using overlap extension PCR fusion (Shevchuk et al., 2004). Flanking fragments of uvrA were amplified from B. burgdorferi 297 genomic DNA (Fraser et al., 1997) using target-specific primers. Briefly, the 544-bp upstream region of uvrABbu was amplified from B. burgdorferi genomic DNA using primers 12.4 and 12.3 (nt 889980–890523 in the B. burgdorferi chromosome) (Table 1). The 700-bp downstream region of uvrABbu was amplified using primers 12.2 and 12.1 (nt 891827–892526). The kanamycin resistance gene aph(3′)-IIIa from Enterococcus faecalis was amplified with its own promoter and stop codon from pBLS500 using primers III and IV (Shevchuk et al., 2004). Parameters for PCR reactions were denaturation at 94 °C for 2 min, 32 cycles of 94 °C for 15 s, 56 °C for 20 s, 68 °C for 2 min, and a final extension at 68 °C for 5 min. PCR fragments were fused by long PCR (Shevchuk et al., 2004), and the final 2279-kb PCR product containing the uvrABbu gene with a kanamycin resistance gene insertion was cloned into pGEM-T (Promega), a vector that cannot replicate in B. burgdorferi, to yield pBL12. Selection and maintenance of E. coli DH5α transformants with pBL12 was performed using solid and liquid Luria–Bertani medium containing 100 μg mL−1 of ampicillin.
Construction of pAB63 and pMS9 for complementation of uvrABbu inactivation mutant
To obtain pAB63 (Fig. 1b), a 3.4-kb PCR fragment containing uvrABbu and 504 bp 5′ to its translational start site (possible promoter region) were amplified from B. burgdorferi 297 genomic DNA using primers AVB3 (containing a SacI restriction site) (Table 1) and AVB4 (containing a PstI restriction site) (Table 1), and ligated into the multiple cloning site of pKFSS1 (Frank et al., 2003) digested with SacI and PstI. To obtain pMS9 (Fig. 1b), the flaBBbu promoter and uvrABbu were amplified from B. burgdorferi 297 genomic DNA using primers FflaB/RflaB (containing SacI and KpnI restriction sites) (Table 1) and FuvrA/RurvA (containing KpnI and PstI restriction sites) (Table 1), respectively, and cloned into pKFSS1, first the flaBBbu promoter, then the ORF for uvrABbu, using the appropriate restriction enzymes.
Transformation of B. burgdorferi
Spirochetes grown to mid-logarithmic phase were electroporated with 5–20 μg of plasmid DNA (Samuels, 1995). Individual clones were obtained by serial dilution of aliquots taken from antibiotic-resistant cultures in complete BSK-H containing antibiotics.
Sensitivity to MMC
Borrelia burgdorferi cells (1 × 105) (midlog phase) were inoculated into 0.5 mL of complete BSK-H containing 0.01, 0.1, 1, 5 or 10 μg of MMC (Sigma Chemical Co.) and cultured at 34 °C for 12–13 days, and spirochetes were counted in duplicate every 1–4 days by dark-field microscopy (Sicklinger et al., 2003). Bacteria were always kept in the dark during these experiments. Two independent experiments with each complementing plasmid were performed.
Sensitivity to UV radiation
Cells grown to a density of 3 × 107 cells mL−1 in complete BSK-H were harvested by centrifugation, resuspended in phosphate-buffered saline (PBS), pH 7.4, to 1 × 105 cells mL−1 and exposed to 800 or 1000 μJ cm−2 280-nm UV radiation (Spectrolinker XL-1000 UV crosslinker, Spectronics Corporation, Westbury, NY). Survival of cells after culture at 34 °C on semisolid BSK-H was determined at 14–18 days (Liveris et al., 2004). Borrelia burgdorferi not exposed to UV irradiation served as a control. Bacteria were always kept in the dark during these experiments. Results from two independent experiments with each complementing plasmid have been combined.
Sensitivity to ROS
Borrelia burgdorferi, 3 × 107 cells mL−1, were harvested by centrifugation, and diluted in triplicate to a density of 5 × 105 cells mL−1 in PBS containing 0, 0.1, 0.5, 1, 1.5, 2, 2.5, 3, 4 and 5 mM H2O2 (Sigma Chemical Co.). After incubation for 1 h at 34 °C, cells were washed with PBS, resuspended in complete BSK with appropriate antibiotics and cultured in capped 0.5-mL tubes or in 96-well plates in 3% CO2 at 34 °C for 12 days. End points were determined by the change of color of the medium, indicating bacterial growth (Terekhova et al., 2002). Results from two to four independent experiments have been combined and are reported as minimal inhibitory concentrations (MIC).
Sensitivity to NOS
NaNO2 (10, 25, 50, 100, 150 mM), (Z)-1-[N-(3-ammoniopropyl)-N-[4-(3-aminopropylammonio) butyl]-amino]-diazen-1-ium-1,2-diolate (0.01, 0.1, 1 mM) (SPER/NO, Sigma Chemical Co.) and S-nitroso-N-acetylpenicillamine (0.05, 0.1, 0.5, 1 mM) (SNAP, Sigma Chemical Co.) were used as sources of NOS. For treatment with NaNO2, 5 × 105 borrelia were inoculated into capped tubes containing 1 mL complete BSK-H and various concentrations of NaNO2 and cultured at 34 °C. For treatment with SPER/NO and SNAP, 5 × 105 cells were incubated in PBS with various concentrations of these reagents for 1 h at 37 °C, harvested by centrifugation, and resuspended and cultured at 34 °C in 1 mL complete BSK-H with appropriate antibiotics. Growth of B. burgdorferi was determined by counting under dark field microscopy every 2–3 days for 8 days. Results from two independent experiments have been combined.
Sensitivity to acid
Acidity of complete BSK-H (pH 7.5) was adjusted to pH 5.5, 6.0, 6.5 and 6.8 by addition of HCl. Borrelia burgdorferi, 5 × 105 cells, were inoculated into 1 mL of pH unadjusted and adjusted medium, and cultured at 34 °C for 9 days. Bacterial growth was assessed by counting under dark field microscopy. Results from two independent experiments have been combined.
Data were analyzed by one-way anova with a post hoc Bonferroni multiple comparisons test. The level of significance was set at P<0.05.
Results and discussion
To inactivate uvrABbu, a 2.3-kb DNA segment was constructed by long PCR (Shevchuk et al., 2004). This segment contained a small portion of the original uvrABbu gene lacking a domain necessary for function and an inserted kanamycin resistance gene (Fig. 1a). It was cloned into pGEM-T (a plasmid that cannot replicate in B. burgdorferi) to yield the suicide plasmid pBL12. After electroporation of pBL12 into low passage, infectious B. burgdorferi 297, multiple kanamycin-resistant clones were obtained; two were selected for genotyping. Genetic inactivation of uvrABbu in these clones was confirmed by PCR of genomic DNA using primers 12.1 and 12.4 (Supporting Information, Fig. S1a, compare lanes 1 and 2). Sequencing a 5.8-kb PCR fragment obtained with primers 12.5 (upstream gene BB0835) and 12.6 (downstream gene BB0838) confirmed homologous DNA exchange between the wild-type chromosomal uvrABbu gene in the chromosome and the disrupted uvrABbu allele in pBL12. Two plasmids were constructed to complement this mutant. Because promoters of B. burgdorferi often overlap the preceding ORF (Cabello et al., 2006), one of these, pAB63 (Fig. 1b), contained both the uvrABbu ORF and the 504 bp upstream of the translational start of uvrA. The other, pMS9 (Fig. 1b), contained the uvrABbu ORF under the control of the borrelial flaB promoter. Electroporation of these plasmids and the pKFSS1 vector control into B. burgdorferiΔuvrABbu, followed by selection and passaging yielded clones containing both full-length and disrupted uvrABbu (Fig. S1a) that expressed uvrA mRNA transcripts (Fig. S1a). Reactions performed without reverse transcriptase showed no amplicons and confirmed the lack of DNA contamination in total RNA samples (data not shown).
UV irradiation damages DNA by generating intrachain thymine dimers (Black et al., 1998; Aertsen et al., 2004; Fry et al., 2005; Maul & Sutton, 2005). Exposure of the parental strain to 800 and 1000 μJ cm−2 of UV radiation had little effect on its survival, while exposure of ΔuvrABbu or its derivative containing only the pKFSS1 cloning vector to these doses resulted in complete loss of viability (Fig. 2a and b). Significant complementation of the phenotypic defect of the inactivation mutant was obtained with both pAB63 and pMS9 (P<0.001). The inability of the inactivation mutant to survive UV radiation was partially corrected by pAB63 (uvrABbu and 504 bp 5′ up to the uvrABbu start codon, Fig. 2a) and fully corrected by pMS9 (Fig. 2b). This indicates that the uvrABbu gene product is involved in the ability of B. burgdorferi to repair intrachain DNA damage.
MMC, a nucleotide-akylating agent, cross-links DNA (Iyer & Szybalski, 1963). Bacterial mutants with various defects in DNA repair have been found to be more susceptible to growth inhibition by this agent than are wild type (Bijlsma et al., 2000; Liveris et al., 2004). In the absence of MMC, wild type, the ΔuvrABbu inactivation mutant and its pAB63 (not shown), pMS9 or pKFSS1 derivatives (Fig. 3a) grew equally well in complete BSK-H. All strains reached log-phase density (about 108 cells mL−1) by day 4 of culture. In the presence of MMC, the growth of ΔuvrABbu was significantly (P<0.001) inhibited [concentrations examined: 0.1 μg mL−1 (data not shown), 1 μg mL−1 (data not shown), 5 μg mL−1 (Fig. 3b), 10 μg mL−1 (Fig. 3c)]. This growth inhibition was reversed by extrachromosomal complementation of ΔuvrABbu with pMS9 (uvrABbu under the control of flaBp) but not with the cloning vector pKFSS1 (Fig. 3b and c). Similar results were obtained using pAB63 (uvrABbu under the control of 504 upstream nucleotides) to complement ΔuvrABbu (data not shown). This indicates that the uvrABbu gene product is involved in repair of interchain repair of DNA damage in B. burgdorferi, in striking difference to the situation in E. coli (Sancar, 1996; Savery, 2007). This functional difference between the uvrABbu and the uvrAEco gene products is not surprising given the evolutionary distance between E. coli and B. burgdorferi (Wu et al., 2009).
Borrelia burgdorferiΔuvrABbu was significantly more susceptible to H2O2 than the wild-type parental strain (Table 2), with the MIC of H2O2 for the wild-type B. burgdorferi being as much as fivefold higher than that of the ΔuvrABbu mutant. This increased sensitivity to ROS was partially reversed by complementation with either pAB63 or pMS9 (Table 2). Complementation was not affected by the presence of 3% CO2 (studies using pAB63), or its absence (studies using pMS9), during culture. Because the inserted kanamycin resistance gene contained its own stop codon, it seems unlikely that polarity effects on the downstream BB0838 gene contributed to the phenotype of the ΔuvrABbu mutant. However, homologues of BB0838 are present in other Borrelia as well as in Treponema, and because the function of this hypothetical protein is unknown, it is not possible to give a definitive answer.
Table 2. Growth of Borrelia burgdorferi and its ΔuvrA derivatives after exposure to DNA-damaging agents
For determination of MIC using pAB63 for complementation of Borrelia burgdorferi 297 ΔuvrA, cells were cultured on plates in 3% CO2. For determination of MIC using pMS9, cells were cultured in capped tubes in the absence of CO2. ND, not determined. See Materials and methods for details.
+, Growth at 10, 25, 50, 100 and 150 mM. Cells incubated and cultured in BSK-H with indicated concentration of NaNO2. See Materials and methods for details.
+, Growth at 0.01, 0.1 and 1 mM. Cells incubated for 1 h with SPER/NO in PBS, washed, then cultured in BSK-H. See Materials and methods for details.
+, Growth at 0.05, 0.1, 0.5 and 1 mM. Cells incubated for 1 h with SNAP in PBS, washed, then cultured in BSK-H. See Materials and methods for details.
+, Growth at pH 6.0, 6.5 and 6.8. Cells cultured in BSK-H at indicated pH. See Materials and methods for details.
B. burgdorferi 297 wild-type
B. burgdorferi 297 ΔuvrA
B. burgdorferi 297 ΔuvrA pAB63
B. burgdorferi 297 ΔuvrA pMS9
B. burgdorferi 297 ΔuvrA pKFSS1
In contrast to the sensitivity of the ΔuvrABbu mutant to ROS, its growth and that of its derivatives was not inhibited by exposure to NaNO2, SNAP or SPER/NO (Table 2). The lack of effect of exposure to any of these RNS generators on B. burgdorferi growth even though exposure to SNAP and SPER/NO was in PBS while exposure to NaNO2 was in BSK-H suggests that this lack of effect was not likely caused by the serum component of BSK-H (Sohaskey & Barbour, 1999). There were also no significant differences in growth of B. burgdorferi and its derivatives in complete BSK-H at pH 6.0, 6.5 or 6.8 (Table 2). None of the strains used in the study (wild-type, uvrA inactivation mutant, complemented mutants) were able to grow at pH 5.5 in complete BSK-H (data not shown).
The ability of pathogenic bacteria to repair challenges to their genomes from various DNA-damaging agents produced by host phagocytes is critical to their survival in their hosts (Fang, 2004; Steere et al., 2004). In the absence of DNA-damaging agents, the uvrABbu inactivation mutant grew as well as the wild-type strain but was markedly inhibited by exposure to UV radiation (Fig. 2), MMC (Fig. 3a) and ROS (Table 2). Extrachromosomal complementation with wild-type uvrABbu restored growth. In contrast, growth of the inactivation mutant was identical to that of the wild-type after exposure to RNS or decreased pH, conditions under which uvrA has been shown to be protective in other bacteria (Aertsen et al., 2004; Fang, 2004). The uvrABbu gene product is thus involved in the repair of DNA damage caused by UV radiation, ROS and MMC in B. burgdorferi, but not involved in damage due to RNS or decreased pH.
Repair of DNA damage caused by UV irradiation involves UvrA recognition of this damage and nucleotide excision (Sancar, 1996; Savery, 2007). Although both pAB63 and pMS9 restored UV resistance to the ΔuvrABbu mutant, they differed in the extent of this complementation (Fig. 2). Interestingly, the UV survival curve of the infectious B. burgdorferi 297 (clone BbAH130) wild-type strain used in the present study was likely similar to that of the infectious B. burgdorferi B31 clone (5A18NP1) used by Lin et al. (2009), but was distinctly different from that reported for the infectious B. burgdorferi B31M1 strain studied by Liveris et al. (2004, 2008). The reason for this difference is at present unclear, but may be strain-related, because the design of our experiments and those of Liveris and colleagues was otherwise identical.
In vitro growth of B. burgdorferi uvrA inactivation mutants was inhibited by ROS but not by RNS. Dissociation of in vitro susceptibility to ROS and RNS has been reported to occur in a M. tuberculosis uvrB mutant (Darwin & Nathan, 2005). In this case, the mutant was more susceptible to RNS than the wild-type parent but showed similar susceptibility to ROS. It was not possible to examine the in vivo function of B. burgdorferi uvrABbu because ΔuvrABbu and its derivatives, in contrast to the parental strain, lacked lp25 (Purser & Norris, 2000; Iyer et al., 2003) (data not shown) and were noninfectious. Studies are currently underway to develop an infectious uvrA inactivation mutant in order to examine its in vivo virulence.
Several lines of evidence suggest that the ability of B. burgdorferi to overcome DNA damage following phagocytosis is critical to its ability to survive and produce disease in the host. Mutants of mutS and mutS-II, genes whose products are involved in DNA mismatch repair, display decreased infectivity in immunocompetent mice (Lin et al., 2009). Furthermore, resistance of B. burgdorferi to rapid killing in vitro by phagocytes has been correlated with in vivo infectivity (Georgilis et al., 1991). Although the majority of phagocytosed borrelia are rapidly killed after ingestion, some remain viable and cultivable (Montgomery et al., 1993), and can stimulate a phagocytic oxidative burst (Georgilis et al., 1991). Plausibly, these few viable organisms could be sufficient to initiate infection of the mammalian host.
In summary, homologous recombination and extrachromosomal complementation have been used to show that uvrABbu is needed to repair DNA damage in B. burgdorferi exposed in vitro to UV, ROS and MMC but not in B. burgdorferi exposed to RNS or low pH.
M.S. and L.B.I. contributed equally to this work, which was supported by grant R01 AI 048856 to F.C.C. We would like to thank Dr M. Norgard, University of Texas Southwestern Medical Center, Dallas, TX, for providing B. burgdorferi 297 and clone BbAH130. We also thank Dr Julia Bugrysheva for advice.