ABSTRACT
The SpoVG protein of Borrelia burgdorferi, the Lyme disease spirochete, binds to specific sites of DNA and RNA. The bacterium regulates transcription of spoVG during the natural tick-mammal infectious cycle and in response to some changes in culture conditions. Bacterial levels of spoVG mRNA and SpoVG protein did not necessarily correlate, suggesting that posttranscriptional mechanisms also control protein levels. Consistent with this, SpoVG binds to its own mRNA, adjacent to the ribosome-binding site. SpoVG also binds to two DNA sites in the glpFKD operon and to two RNA sites in glpFKD mRNA; that operon encodes genes necessary for glycerol catabolism and is important for colonization in ticks. In addition, spirochetes engineered to dysregulate spoVG exhibited physiological alterations.
IMPORTANCE B. burgdorferi persists in nature by cycling between ticks and vertebrates. Little is known about how the bacterium senses and adapts to each niche of the cycle. The present studies indicate that B. burgdorferi controls production of SpoVG and that this protein binds to specific sites of DNA and RNA in the genome and transcriptome, respectively. Altered expression of spoVG exerts effects on bacterial replication and other aspects of the spirochete's physiology.
INTRODUCTION
Borrelia burgdorferi has evolved to survive within a defined enzootic cycle, alternately colonizing and transmitting between tick vectors and vertebrate hosts (1, 2). These two animal types represent drastically different environments, requiring B. burgdorferi to sense specific cues and to adapt its cellular programs in response to those cues (3). Appropriately executing such adaptations is critical to the success of B. burgdorferi as a pathogen. Very little is known, however, about the mechanisms by which B. burgdorferi accomplishes these feats.
The SpoVG protein was originally named for its role as a negative regulator of asymmetric division during Bacillus sp. sporulation (4). It has since been recognized as a highly conserved protein that is produced by many eubacteria, including B. burgdorferi. We previously reported that the SpoVG proteins of B. burgdorferi, Listeria monocytogenes, and Staphylococcus aureus are site-specific DNA-binding proteins (5). L. monocytogenes SpoVG has since been shown to also bind RNA (6).
When in a feeding tick or within a vertebrate, B. burgdorferi grows and divides rapidly, fed by the glucose and other nutrients in blood and serum. After a tick has digested the blood meal, however, nutrients available to colonizing B. burgdorferi are severely limited (2, 7–9). In response, the bacteria reduce or cease processes that require substantial energy, such as growth and division (2, 10, 11). Cells of postfeeding ticks secrete glycerol as a mechanism to survive cold stress (8). That glycerol serves as an important energy source for B. burgdorferi within unfed ticks. Mutants of the glpFKD operon are unable to use glycerol as a carbon source and are significantly impaired in their abilities to survive in ticks (12). Appropriate regulation of the glpFKD operon is necessary in order to traverse the enzootic cycle successfully (12, 13).
Herein we present characterization of spoVG expression throughout the mouse-tick infectious cycle. Those studies, and investigations of cultured B. burgdorferi, indicate that bacteria regulate spoVG transcript and SpoVG protein levels by both transcriptional and posttranscriptional mechanisms and spoVG affects the physiology of B. burgdorferi.
RESULTS
Borrelia burgdorferi controls expression of spoVG mRNA.Having demonstrated previously that the B. burgdorferi SpoVG protein is a site-specific DNA-binding protein (5), we hypothesized that it acts as a regulatory factor, which would require that SpoVG itself is regulated. To that end, we quantified spoVG expression levels during colonization of mice and ticks.
A cohort of seven mice were infected with wild-type B. burgdorferi. After 4 weeks, hearts were harvested, RNA extracted, and transcript levels analyzed by quantitative reverse transcription-PCR (qRT-PCR). Levels of spoVG transcripts were determined relative to those of the constitutively expressed flaB mRNA (14). Detectable levels of spoVG transcripts, ranging from 6% to <1% of those of flaB, were present for all of the mice (Fig. 1A).
Relative levels of spoVG expression by Borrelia burgdorferi in mice and ticks. (A) Cohorts of mice were infected with B. burgdorferi on two separate occasions. RNA was harvested from the heart of each mouse and analyzed by qRT-PCR. Shown are technical triplicates for each mouse heart. (B) Pools of 10 to 20 nymphal ticks were analyzed by qRT-PCR. Levels of spoVG transcripts were quantified and are depicted as a percentage of flaB transcript levels. Shown are biological replicates of tick pools for each group.
Ticks that were colonized with B. burgdorferi were similarly examined for spoVG mRNA levels. RNA was extracted from unfed tick nymphs and from nymphs that had attached to mice for 72 h (midway through the feeding process). Under both conditions, spoVG transcripts were nearly as abundant as flaB transcripts (Fig. 1B). The difference in spoVG transcript levels between unfed and feeding ticks was not significant. Adding further evidence that spoVG is abundantly expressed during tick colonization, an array-based study published while this manuscript was in preparation reported that spoVG was one of the 20 most abundant B. burgdorferi transcripts expressed in fed tick nymphs and larvae (15). From these results, it is apparent that B. burgdorferi naturally regulates spoVG transcript levels.
SpoVG expression in cultured Borrelia burgdorferi.To gain insight into the mechanisms controlling SpoVG expression, transcript and protein levels were assessed in cultured bacteria throughout different growth phases. Wild-type B. burgdorferi was cultured at 34°C and then harvested at either day 2 or day 3 and at days 5, 7, and 9. RNA was extracted and analyzed by qRT-PCR, and spoVG abundance was normalized to flaB transcript levels (Fig. 2A). There were no significant changes in spoVG transcript levels through the different growth stages. This experiment, using triplicate cultures, was repeated twice with similar results. Lysates for Western blotting were prepared from 34°C cultures harvested on days 3, 5, 7, and 9. SpoVG protein levels were readily detectable on day 3 but were nearly undetectable on day 9 (Fig. 2C). These disparities between protein and mRNA expression levels suggest that SpoVG protein levels are controlled through a posttranscriptional mechanism.
SpoVG transcript and protein expression in culture. (A) RNA extracted from cultured B. burgdorferi harvested on days 2, 5, 7, and 9 was analyzed by qRT-PCR. spoVG transcript levels were determined, relative to flaB transcripts, by the ΔΔCT method. (B) RNAs extracted from cultured wild-type (WT) and spoVG-ON B. burgdorferi grown at 34°C and 23°C were analyzed by qRT-PCR. Relative spoVG transcript abundance was normalized to flaB transcript levels by the ΔΔCT method. (C) Lysates from cultured B. burgdorferi harvested on days 3, 5, 7, and 9 were analyzed for SpoVG protein production. FlaB protein levels were used as a loading control. (D) Lysates from wild-type and spoVG-ON B. burgdorferi were harvested on day 3 from cultures grown at 34°C and on day 10 from cultures grown at 23°C and were analyzed by Western blotting. SpoVG abundance is shown with FlaB as the loading control.
B. burgdorferi differentially expresses numerous genes and proteins in response to changing culture temperatures (3, 16–21). Levels of spoVG mRNA were found to be 12 times greater in wild-type cultures grown at a constant 34°C than they were in bacteria maintained at a constant 23°C (Fig. 2B). Conversely, levels of SpoVG protein were not significantly different in the 34°C cultures than in the 23°C cultures, providing further evidence of posttranscriptional regulation (Fig. 2D).
Additional insight regarding borrelial control of SpoVG protein levels was serendipitously obtained from bacteria that were engineered to ectopically transcribe elevated levels of spoVG. We produced a plasmid that constitutively expressed spoVG from the borrelial flgB promoter, with the intent of producing greater amounts of the protein. This plasmid, named pCRS5, was transformed into B. burgdorferi on three separate occasions. Essentially identical results were obtained for all three transformants for all experiments; therefore, we collectively refer to the three mutant strains as spoVG-ON. Analyses of the naturally occurring small DNAs (plasmids) of the wild-type strain and two spoVG-ON strains revealed that the wild-type strain and one mutant had identical plasmid contents, while one of the mutants was lacking lp56. The identical results for the spoVG-ON strains and the identical plasmid contents of the wild-type strain and the mutant indicate that lp56 does not play a significant role in the effects observed in these studies.
Levels of spoVG mRNA were determined for the spoVG-ON and wild-type strains. Transcript levels were 5 times greater in the spoVG-ON strains than in the wild-type strain during cultivation at 34°C and were nearly 47 times greater during cultivation at 23°C (Fig. 2C). Both of those differences were statistically significant. However, increased expression of spoVG transcripts did not result in significantly elevated levels of SpoVG protein in the spoVG-ON strains at 34°C or 23°C (Fig. 2D). The pCR5 constructs were reisolated from the spoVG-ON strains and were found to have sequences that were identical to that of the initial plasmid, indicating that no mutations had occurred within B. burgdorferi. A control plasmid that produced elevated levels of revA mRNA yielded elevated levels of the RevA protein, indicating that transcripts could be translated from such constructs (22).
SpoVG binds its own DNA and RNA.As reported previously, SpoVG is a site-specific DNA-binding protein (5). To gain insight into whether SpoVG influences its own expression, we queried whether the protein binds to its DNA and RNA. Electrophoretic mobility assays (EMSAs) using recombinant SpoVG found specific binding to a 34-bp DNA region upstream of the translational start codon (Fig. 3A and 4A). The shifted band was partially ablated by addition of 100× specific unlabeled DNA but not by addition of 100× nonspecific DNA, indicating that the interaction is specific (Fig. 3A). As this region is 3′ of the transcriptional promoter and includes the ribosome-binding site, we examined whether SpoVG binds to RNA of the same sequence. EMSAs demonstrated that SpoVG bound to that RNA (Fig. 3B). Addition of 1,000× unlabeled DNA could not entirely compete away the shifted RNA, indicating that SpoVG binds to this RNA with a higher affinity than that for the template DNA sequence.
SpoVG binding of its own DNA and RNA. EMSAs were performed using either labeled DNA or RNA derived from the sequence 5′ of the spoVG translational start site or RNA derived from the sequence 5′ of the unrelated erpAB operon (23). The spoVG RNA probe formed a secondary structure in the absence of added protein (indicated by an asterisk). (A) Lane 1, 1 ng DNA probe; lane 2, 1 ng DNA probe with 2.4 μM rSpoVG; lane 3, 1 ng DNA probe with 4.8 μM rSpoVG; lane 4, 1 ng DNA probe with 4.8 μM rSpoVG and 100 ng unlabeled spoVG DNA; lane 5, 1 ng DNA probe with 4.8 μM rSpoVG and 100 ng unlabeled EMSA-pCR2.1 DNA. (B) Lane 1, 1 ng RNA probe (asterisk indicates a secondary structure of the RNA); lane 2, 1 ng RNA probe with 2.4 μM rSpoVG; lane 3, 1 ng RNA probe with 2.4 μM rSpoVG and 100 ng unlabeled spoVG DNA; lane 4, 1 ng RNA probe with 2.4 μM rSpoVG and 1,000 ng unlabeled spoVG DNA. (C) Lanes 1 to 5, 1 ng erp RNA; lanes 6 to 10, 1 ng spoVG RNA; lanes 1 and 6, no protein; lanes 2 and 7, 1.6 μM r SpoVG; lanes 3 and 8, 160 nM rSpoVG; lanes 4 and 9, 16 nM rSpoVG; lanes 5 and 10, 1.6 nM rSpoVG.
Nucleic acid sequences that contain SpoVG binding sites. The sequences are shown as DNA, although SpoVG also bound to all of these sites in single-stranded RNA. The DNA/RNA target regions used for EMSAs are underlined. (A) Intergenic region between the start of the spoVG open reading frame (ORF) and the gene located immediately upstream, rsf. The transcriptional start site of the spoVG ORF is highlighted in bold black type, and the stop codon of the rsf gene is highlighted in bold blue type. (B) Sequence 5′ of the glpFKD operon. The −35 and −10 sequences of the probable transcriptional promoter are boxed, and the translational initiation codon of glpF is in bold black type. The maximal sequence bound by SpoVG is underlined. The region previously identified as having a stimulatory effect on transcript levels is shown in blue type (28). (C) Sequence of the intergenic region between glpK and glpD. The maximal sequence bound by SpoVG is underlined. The −35 and −10 sequences of the predicted transcriptional promoter are boxed. The transcription start site mapped by Adams et al. (34) is indicated in red and marked with an asterisk. The translational initiation codon of glpD is in bold type. A small ORF of unknown function, bb0242, is indicated in blue.
The specificity of SpoVG binding to RNA was assessed by an EMSA using a sequence based on the operator region of the erpAB operon (23). When used at the same concentrations as used to generate the findings in Fig. 3B, recombinant SpoVG protein did not bind to the erpAB-derived RNA sequence (Fig. 3C).
Influence of spoVG mRNA on B. burgdorferi growth and division.Several analyses of cultured spoVG-ON bacteria were undertaken prior to the recognition that spoVG mRNA levels did not necessarily correspond directly to SpoVG protein levels. The following results indicate that substantial effects on B. burgdorferi physiology can be caused by changes in spoVG mRNA levels, even in the absence of changes in SpoVG protein levels. A possible explanation for these observations is that elevated levels of spoVG mRNA titrate SpoVG protein away from its other RNA and DNA targets, thereby altering transcription or translation of those targets.
Growth curves were determined for the wild-type and spoVG-ON B. burgdorferi strains cultured at either 34°C or 23°C. There were no differences between the two strains in growth rates at 34°C (Fig. 5A). As reported previously, wild-type B. burgdorferi grew more slowly at 23°C than at 34°C (17, 24). The spoVG-ON strains, however, consistently grew even more slowly at 23°C than did the wild-type strain. The exponential growth phase (days 0 to 20) for each strain was analyzed for exponential growth by nonlinear regression. The rate constant for the wild-type strain was calculated to be 0.2878 day−1, while that for the spoVG-ON strains was 0.1388 day−1; these values were significantly different, with P of 0.0001. During this period, the doubling time of the wild-type strain was 2.4 days, while the doubling time of the spoVG-ON strains was 5 days (Fig. 5B). This indicates that enhanced production of spoVG mRNA does not affect the growth rate at 34°C but spoVG mRNA in combination with some factor (or factors) that is present or absent in 23°C cultures has an appreciable impact on the growth rate of B. burgdorferi.
Effects of spoVG transcript levels on B. burgdorferi growth rates. (A and B) Cultured wild-type (WT) and spoVG-ON B. burgdorferi strains were grown in complete medium at 34°C and counted every 24 h (A) or were grown in complete medium at 23°C and counted every 3 days (B). (C and D) Cultured wild-type and spoVG-ON B. burgdorferi strains were also grown at 34°C in one of two deficient media, i.e., grown in 1.2% rabbit serum medium and counted every 3 days (C) or grown in 25% BSK medium and counted on days 2, 4, 8, and 16 (D).
An additional effect of enhanced spoVG mRNA and cultivation at 23°C was on bacterial survival. The 23°C cultures were left in incubators for 48 days after initial inoculation. At that time, each culture was diluted 1:100 into fresh Barbour-Stoenner-Kelly (BSK) medium and immediately observed with a dark-field microscope. At least 50 cells per culture were examined. In that time frame, none of the wild-type bacteria exhibited any form of motility. In contrast, at least 20% of the spoVG-ON cells had begun to move. The initial cultures were then returned to the incubator for an additional 8 days. On day 56, each culture was diluted 1:100 into fresh BSK medium and incubated at 34°C (final densities of approximately 1 × 106 bacteria per ml). By day 9, none of the wild-type cultures showed any evidence of bacterial growth, whereas all of the spoVG-ON cultures had reached densities of at least 1 × 108 bacteria per ml.
As noted above, B. burgdorferi normally grows more slowly at 23°C than 34°C. We previously demonstrated several instances in which the effects of those culture temperatures on gene expression were due to changes in borrelial growth rates and not to differences in temperature per se (24). Noting that culture of spoVG-ON strains at 23°C resulted in rates of division that were significantly lower than that of wild-type bacteria but spoVG-ON did not affect division rates at 34°C, we tested whether the effect of 23°C was due to temperature or growth rate. To that end, spoVG-ON and wild-type strains were cultured in one of two deficient media that were formulated to reduce the growth rate of wild-type B. burgdorferi by approximately 4-fold at 34°C (24). There were no differences in growth rates between wild-type and spoVG-ON B. burgdorferi strains cultured in either of the deficient media (Fig. 5C and D). This finding suggests that the growth-inhibiting effect of enhanced spoVG transcription at 23°C may be directly linked to the environmental temperature.
SpoVG binds DNA and RNA of the glpFKD operon.A preliminary transcriptome sequencing (RNA-Seq) analysis was performed early during these investigations, comparing a spoVG-ON strain with its wild-type parent. Analysis of the preliminary data suggested an impact on the glpFKD operon. Those genes encode three proteins that import glycerol into the bacterial cell and perform the first two enzymatic steps of its metabolism (12, 25). Subsequent qRT-PCR analyses of glpFKD transcript levels in either wild-type or SpoVG-ON strains yielded inconsistent results. During the time these studies were in progress, other research indicated that glpFKD transcript levels are affected by several factors, including the alternative sigma protein RpoS, cyclic di-GMP, and ppGpp (13, 26–28). Those factors are produced in response to growth rate and pH, an unknown external signal, and nutrient limitation, respectively, conditions that are difficult to reproduce accurately between experiments, which likely accounted for the inconsistent results (24, 29–32). More detailed RNA-Seq and qRT-PCR analyses will be performed in the future, after the posttranscriptional regulatory mechanisms that control SpoVG protein production have been elucidated. Nonetheless, because glycerol is a critical energy source for B. burgdorferi in unfed ticks (12) and because spoVG is highly transcribed during that stage, studies were undertaken to determine whether SpoVG could play a direct role in controlling glpFKD expression.
The first gene of the operon, glpF, is preceded by an approximately 200-bp 5′ untranslated transcript (28, 33, 34) (Fig. 4B). Previous work determined that a 42-bp sequence directly 3′ of the promoter had a positive effect on transcript levels (28). EMSAs with recombinant SpoVG indicated that this protein bound specifically to a 52-bp region of DNA that overlaps the start of transcription and the previously reported activator region (Fig. 4B and 6A). SpoVG also bound to RNA that contained the same sequence. As with the binding site in the spoVG locus, 1,000× unlabeled DNA of the same sequence was not sufficient to entirely compete away the protein-RNA complex, indicating a higher affinity of SpoVG for glpF 5′ untranslated region (UTR) RNA, relative to the corresponding DNA (Fig. 6B).
SpoVG binding to DNA and RNA at two sites in the glpFKD operon. (A and B) Sequence 5′ of glpF. (C and D) Region between glpK and glpD. EMSAs were performed using either labeled DNA or RNA. (A) Lane 1, 1 ng glpF DNA; lane 2, 1 ng glpF DNA with 2.4 μM rSpoVG; lane 3, 1 ng glpF DNA with 4.8 μM rSpoVG; lane 4, 1 ng glpF DNA with 4.8 μM rSpoVG and 100 ng unlabeled glpF DNA; lane 5, 1 ng glpF DNA with 4.8 μM rSpoVG and 100 ng unlabeled EMSA-pCR2.1 DNA. (B) Lane 1, 1 ng glpF RNA; lane 2, 1 ng glpF RNA with 2.4 μM rSpoVG; lane 3, 1 ng glpF RNA with 2.4 μM rSpoVG and 100 ng unlabeled glpF DNA; lane 4, 1 ng glpF RNA with 2.4 μM rSpoVG and 1,000 ng unlabeled glpF DNA. (C) Lane 1, 1 ng glpKD DNA; lane 2, 1 ng glpKD DNA with 0.73 μM rSpoVG; lane 3, 1 ng glpKD DNA with 1.47 μM rSpoVG; lane 4, 1 ng glpKD DNA with 3 μM rSpoVG; lane 5, 1 ng glpKD DNA with 3 μM rSpoVG and 100 ng unlabeled glpKD DNA; lane 6, 1 ng glpKD DNA with 3 μM rSpoVG and 100 ng unlabeled EMSA-pCR2.1 DNA. (D) Lane 1, 1 ng glpKD RNA; lane 2, 1 ng glpKD RNA with 2.4 μM rSpoVG; lane 3, 1 ng glpKD RNA with 2.4 μM rSpoVG and 100 ng unlabeled glpKD DNA; lane 4, 1 ng glpKD RNA with 2.4 μM rSpoVG and 1,000 ng unlabeled glpKD DNA. A graphic representation of SpoVG-binding sites and promoters, relative to the glpFKD operon, is shown at the bottom.
There is a 400-bp gap between the termination codon of glpK and the initiation codon of glpD (25) (Fig. 4C). Our in silico analysis of this intergenic region revealed a potential transcriptional promoter. Other studies demonstrated that transcripts do originate from this region, providing further evidence that this is a functional promoter (27, 34). EMSAs revealed that SpoVG protein bound specifically to a 38-bp sequence that is located 3′ of this putative promoter (Fig. 4B and 6C). Further analysis revealed that SpoVG also bound to this RNA sequence. As with other investigated sites, 1,000× unlabeled DNA with that sequence was unable to compete away the SpoVG-RNA complex (Fig. 6D).
DISCUSSION
One important consequence of these studies is the demonstration that B. burgdorferi SpoVG protein levels do not necessarily correlate directly with spoVG mRNA levels, indicating that bacterial protein levels are, in part, controlled through a posttranscriptional mechanism (or mechanisms). A likely contributor is SpoVG itself, since it binds to spoVG mRNA at a sequence near the ribosome-binding site. Such a feedback mechanism could prevent translation when levels of SpoVG reach a certain point. A similar autoregulatory mechanism appears to control bacterial levels of another B. burgdorferi site-specific RNA-binding protein, BpuR (35, 36). SpoVG also binds to its operon's DNA, at a site located 3′ of the transcriptional promoter; therefore, the protein could affect levels of its own transcription. It is also possible that additional factors influence spoVG transcription and/or translation or interact with SpoVG protein to modulate its nucleic acid-binding activities. The interplay between SpoVG protein, the spoVG transcript and gene, and other bacterial factors is an ongoing focus of our research. Noting the effects on B. burgdorferi physiology we observed, we predict that precise regulation of SpoVG protein and spoVG mRNA is critical to the Lyme disease spirochete. These findings also raise the possibility that other bacterial species control levels of their SpoVG proteins through similar means.
Another important finding is that situations were observed in which elevating the spoVG transcript levels resulted in phenotypic changes, even when there were no detectable changes in SpoVG protein levels. We observed that SpoVG binds not only its own mRNA but also two sites within the glpFKD mRNA and two DNA sites in the glpFKD operon. These findings suggest that there are likely to be additional SpoVG-binding sites in the borrelial transcriptome and genome. Analysis of the known SpoVG-binding sites has not revealed any obvious consensus sequence; therefore, we hypothesize that SpoVG may interact with certain nucleic acid structural motifs rather than a particular nucleotide sequence. A plausible explanation for how changes in spoVG transcript levels cause changes in bacterial physiology is that elevated spoVG message levels pull SpoVG protein away from other binding sites, thereby preventing SpoVG effects on those targets. This would be similar to the mechanism by which some bacteria control cellular levels of their CsrA RNA-binding proteins by producing regulatory noncoding RNAs that sequester the protein (37). If additional regulatory proteins bind to spoVG mRNA, then their sequestration could have further effects. It is also possible that spoVG mRNA can hybridize with other mRNAs or noncoding RNAs and thereby alter their activities.
The observed results could have effects far beyond B. burgdorferi. The observed posttranscriptional regulation of SpoVG protein levels indicates that it cannot be assumed that changes in bacterial mRNA levels necessarily equal changes in protein levels. Furthermore, changes in transcript levels can have effects on bacteria without altering levels of the encoded protein. These caveats need to be addressed in all transcriptomic studies, in B. burgdorferi and other organisms.
SpoVG bound to two sites within the glpFKD operon and to two sites in that operon's mRNA, suggesting that SpoVG could affect production of the GlpF transporter and/or GlpK and GlpD enzymes. At least three other factors influence levels of the glycerol catabolic proteins, i.e., RpoS, cyclic di-GMP, and ppGpp (13, 26–28). A concerted effort that focuses on all four of the known glp regulatory factors is needed to accurately address the mechanism by which B. burgdorferi controls glycerol import and catabolism.
Altogether, the current studies suggest a role for SpoVG in adaptation to different environments. The affinity of SpoVG for glpFKD DNA and mRNA suggest that it may affect that operon's transcription and/or translation and also may influence additional operons by such means. Since the expression of spoVG is not entirely bimodal between ticks and mice, it is conceivable that SpoVG functions as a modulator of information flow throughout the infectious cycle and helps to enact the appropriate responses to specific environmental cues.
MATERIALS AND METHODS
Bacterial strains.All studies used infectious clones of B. burgdorferi strain B31-MI, a derivative of the type strain, for which the complete genome sequence has been determined (25, 38). Mouse infection studies were performed using clone B31-MI-16 (39), and all other studies were performed using clone B31-A3 (40). Both clones are fully virulent, and they are genetically identical except for the absence of the small native plasmid cp9 in B31-A3.
Strains of B. burgdorferi that constitutively transcribe elevated levels of spoVG were generated as follows. The spoVG gene was cloned into the previously described plasmid pBLS715 (22), directly 3′ of the gentamicin resistance gene, to create pCRS5 (Fig. 7). The plasmid insert was sequenced to ensure that spoVG was cloned in-frame without mutations. The spoVG ribosomal binding site was modified to AGGAGG to ensure efficient translation of SpoVG protein (22). B. burgdorferi B31-A3 was transformed with pCRS5 on three separate occasions. To ensure that no mutations had arisen in B. burgdorferi, pCRS5 was purified from each transformant and the insert was sequenced. Bacterial contents of native plasmids were analyzed by multiplex PCR, as described by Bunikis et al. (41). As described above, growth curve, Western blot, and qRT-PCR analyses were performed for all three strains, with similar results being obtained for all strains; therefore, the strains are collectively referred to as spoVG-ON strains.
pCRS5 (spoVG-ON) plasmid construction. spoVG was cloned into the BamHI and KpnI sites of our previously described pBLS715 plasmid (22). The parental construct was created to contain a consensus ribosome-binding site (RBS) appropriately placed 5′ of the insert. The constitutively expressed borrelial flgB promoter drives expression of both the gentamicin resistance gene (aacC1) and spoVG. The identically produced pBLS715revA produces elevated levels of both revA mRNA and RevA protein (22).
Culture conditions.B. burgdorferi was cultured at 34°C in Barbour-Stoenner-Kelly II (BSK-II) medium (which contains 6% rabbit serum), unless otherwise noted (42). The deficient culture medium termed 25% BSK medium includes the components of BSK diluted 1:4 with phosphate-buffered saline (PBS), plus the standard 6% rabbit serum. The deficient culture medium termed 1.2% rabbit serum medium includes standard concentrations of BSK components plus 1.2% rabbit serum (24).
Bacterial growth curves were initiated by diluting mid-exponential-phase cultures of B. burgdorferi (approximately 107 bacteria/ml) 1:100 into fresh medium. Bacterial numbers were counted by use of a Petroff-Hausser counting chamber with dark-field microscopy. Cultures grown at 34°C in BSK-II medium were counted once every 24 h, and cultures grown at 23°C or in 25% BSK medium or 1.2% rabbit serum medium were counted every 48 h. Viability was assessed by counting motile cells and by passaging cultures into fresh medium and determining the growth rate. Differences in growth rates between strains and conditions during exponential-phase growth were calculated using Prism statistics. Graphs were analyzed by nonlinear regression for exponential growth, with the null hypothesis that the rate constants for the two strains were not statistically different (GraphPad Software, La Jolla, CA).
Quantitative reverse transcription-PCR analyses of mRNA expression.Relative spoVG expression levels during tick colonization and murine infection were determined by qRT-PCR using the oligonucleotide primers described in Table 1. Analyses of gene expression levels in cultured B. burgdorferi were performed as described previously (43), using oligonucleotide PCR primers specific for each gene being examined. Results were compared with those for the constitutively expressed flagellar gene flaB (14). Biological triplicates were assayed for all strains and conditions, all qRT-PCR assays were performed with technical triplicates, and runs were performed at least twice. Following each run, primer specificity and amplicon purity were assayed via melting curve analysis (43). qRT-PCR data on transcripts in culture were analyzed by one-way analysis of variance (ANOVA) (GraphPad Software).
Oligonucleotides used in this study
For analyses of transcript levels during tick colonization, uninfected Ixodes scapularis larvae were infected by feeding to repletion on B31-A3-infected mice. Ticks were allowed to molt to nymphs and then were placed on naive mice for feeding. Ticks were collected at specific time points, snap-frozen in liquid nitrogen, and stored at −80°C until use. RNA was isolated from pooled ticks, with 10 to 20 nymphs per pool. Pooled frozen ticks were crushed in 100 μl PBS using disposable mortars and pestles. RNA was isolated using the Nucleospin kit (Macherey-Nagel Co., Bethlehem, PA), according to the manufacturer's recommendations, or TRIzol reagent (Invitrogen, Carlsbad, CA). RNA was treated with DNase and converted to cDNA using a high-capacity cDNA reverse transcription kit (Applied Biosystems, Foster City, CA), as described previously (44). Three biological replicates (independent pools of ticks), with triplicate technical replicates, were assessed for gene expression by qRT-PCR, as described previously (45). A standard curve was generated with genomic DNA representing a range of 106 to 10 cells (in 10-fold serial dilutions), using the threshold cycle (CT) values from the flaB primer-probe set. This standard curve was then used to interpolate spoVG transcript copy numbers from the CT values generated with the spoVG primer-probe set. The spoVG transcript copy numbers were normalized to flaB transcript copy numbers. Negative controls lacking reverse transcriptase confirmed that all genomic DNA had been removed and did not contribute to the signal.
For analyses of transcript levels during mammalian infection, seven female, 4- to 6-week-old, C3H/HEN mice were infected by subcutaneous injection of 1 × 106 B31-MI-16 bacteria from a mid-exponential-phase culture. Two weeks postinfection, mouse blood was drawn from the saphenous vein and processed to serum, and the presence of B. burgdorferi-specific antibodies was determined with an enzyme-linked immunosorbent assay (ELISA), as described previously (46). Mice were euthanized after 4 weeks of infection, and hearts were collected, flash frozen in liquid nitrogen, and stored at −80°C until use. Hearts were chosen because they are colonized by B. burgdorferi and are substantial, uniformly sized organs (47). Frozen hearts were ground with mortars and pestles, followed by homogenization (PRO Scientific) in TRIzol reagent on ice. RNA was resuspended in RNAsecure reagent (Ambion) and treated with DNase I (Ambion) to remove contaminating DNA. The DNase was inactivated using DNase inactivation reagent (Ambion). A 1-μg aliquot of each DNA-free RNA preparation was reverse transcribed using first-strand cDNA synthesis kits (Life Technologies) with random hexamers. Controls containing all components except reverse transcriptase were prepared and treated similarly. Quantitative PCR was performed as described above. All cDNA samples were analyzed in triplicate. Every qRT-PCR series included negative-control samples of RNA processed without reverse transcriptase (see above), to test for DNA contamination, and samples that lacked a template, to test for DNA contamination of reagents. Tenfold serial dilutions of B31-MI-16 genomic DNA (100 ng to 100 fg) were included in every assay for each primer set. This enabled the generation of standard curves, from which the amount of transcript present in each cDNA sample could be calculated by using Bio-Rad myIQ software. Average expression values obtained from triplicate runs of each cDNA sample for all of the genes of interest were calculated relative to the average triplicate value for the constitutively expressed B. burgdorferi flaB transcript from the same cDNA preparation (43). All animal studies were performed with approval from and under the supervision of the institutional animal care and use committees of the University of Kentucky, the University of North Dakota, and Rocky Mountain Laboratories, NIH.
Protein analysis by Western blotting.Western blotting was performed as described previously (17). Polyclonal rabbit antisera specific for SpoVG were produced by NEO Group Inc. (Cambridge, MA). Unless otherwise noted, mid-exponential-phase B. burgdorferi cultures were passaged 1:100 into fresh BSK medium and harvested by centrifugation after 3 days. Cell pellets were frozen overnight at −80°C and resuspended in SDS loading buffer. Lysates were separated by SDS-PAGE, and proteins were transferred to nitrocellulose membranes. Membranes were blocked with 5% nonfat dry milk, incubated with primary antibody, then incubated with protein A-conjugated horseradish peroxidase. Blots were imaged with SuperSignal West Pico PLUS chemiluminescent substrate (Thermo Fisher).
Purification of recombinant SpoVG.SpoVG was purified as described previously (5). pBLJ132 containing spoVG cloned into pET101 was transformed into Rosetta II Escherichia coli, and expression was induced with 1 mM isopropyl-β-d-thiogalactopyranoside (IPTG). SpoVG was purified using MagneHis nickel particles (Promega, WI) and dialyzed into buffer containing 100 nM dithiothreitol, 50 mM Tris-HCl, 25 mM KCl, 10% glycerol (vol/vol), 0.01% Tween 20, and 1 mM phenylmethanesulfonyl fluoride. Protein purity and concentrations were assessed by SDS-PAGE and the bicinchoninic acid (BCA) assay (Thermo Fisher), respectively.
Electrophoretic mobility shift assays.EMSAs were conducted as described previously (48). Biotinylated DNA oligonucleotides (Integrated DNA Technologies, Coralville, IA) were annealed in order to generate specific probes. Biotinylated RNA oligonucleotides (IDT) contained the same nucleotide sequences as the sense strands of the DNA probes. Specific competitors were made from the same nucleic acid sequences, without biotin. Nonspecific competitor was made from pCR2.1 DNA amplified with M13-F and M13-R primers. Protein-binding sites were identified by the combined use of differently sized labeled probes and unlabeled competitors. Reaction mixtures containing EMSA buffer, 40 μg/ml poly(dI-dC), 1 nM DNA with the indicated amounts of protein, and unlabeled DNA were incubated for 20 min at room temperature. Reaction mixtures were separated by 6% native PAGE and transferred to positively charged nylon membranes. Blots were blocked, incubated with streptavidin-horseradish peroxidase, and imaged using the chemiluminescent nucleic acid detection module kit (Thermo Fisher).
In silico promoter analysis.The putative promoter between glpK and glpD was mapped using BPROM from SoftBerry (49).
ACKNOWLEDGMENTS
This work was supported by a grant from the National Research Fund for Tick-Borne Diseases to B.S., National Institutes of Health grant P20GM113123 to C.A.B., and funds from the Division of Intramural Research, National Institute of Allergy and Infectious Diseases, National Institutes of Health to P.A.R.
The funders had no role in study design, data collection and analysis, the decision to publish, or preparation of the manuscript.
We thank Will Arnold and Kathryn Lethbridge for helpful discussions during these studies.
FOOTNOTES
- Received 17 January 2018.
- Accepted 2 April 2018.
- Accepted manuscript posted online 9 April 2018.
- Address correspondence to Brian Stevenson, brian.stevenson{at}uky.edu.
↵* Present address: Brandon L. Jutras, Department of Biochemistry, Virginia Tech, Blacksburg, Virginia, USA.
Citation Savage CR, Jutras BL, Bestor A, Tilly K, Rosa PA, Tourand Y, Stewart PE, Brissette CA, Stevenson B. 2018. Borrelia burgdorferi SpoVG DNA- and RNA-binding protein modulates the physiology of the Lyme disease spirochete. J Bacteriol 200:e00033-18. https://doi.org/10.1128/JB.00033-18.
REFERENCES
- Copyright © 2018 American Society for Microbiology.
