Positive and Negative Regulation of Glycerol Utilization by the c-di-GMP Binding Protein PlzA in Borrelia burgdorferi

The Lyme disease pathogen, Borrelia burgdorferi, has a simple cyclic dimeric GMP (c-di-GMP) signaling system essential for adaptation of the pathogen to the complicated tick environment. The c-di-GMP effector of B. burgdorferi, PlzA, has been shown to regulate multiple cellular processes, including motility, osmolality sensing, and nutrient utilization. The findings of this study demonstrate that PlzA not only controls multiple targets but also has different functional modalities, allowing it to act as both positive and negative regulator of the glp operon expression. This work highlights how bacteria with a small genome can compensate for the limited regulatory repertoire by increasing the complexity of targets and modes of action in their regulatory proteins.

ABSTRACT

Borrelia burgdorferi, the causative agent of Lyme disease, encounters two disparate host environments during its enzootic life cycle, Ixodes ticks and mammalian hosts. B. burgdorferi has a small genome that encodes a streamlined cyclic dimeric GMP (c-di-GMP) signaling system comprising a single diguanylate cyclase, Rrp1, and two phosphodiesterases. This system is essential for spirochete survival in ticks, in part because it controls the expression of the glp operon involved in glycerol utilization. In this study, we showed that a B. burgdorferi c-di-GMP receptor, PlzA, functions as both a positive and a negative regulator for glp expression. Deletion of plzA or mutation in plzA that impaired c-di-GMP binding abolished glp expression. On the other hand, overexpression of plzA resulted in glp repression, which could be rescued by simultaneous overexpression of rrp1. plzA overexpression in the rrp1 mutant, which is devoid of c-di-GMP, or overexpression of a plzA mutant incapable of c-di-GMP binding further enhanced glp repression. Combined results suggest that c-di-GMP-bound PlzA functions as a positive regulator, whereas ligand-free PlzA acts as a negative regulator for glp expression. Thus, PlzA of B. burgdorferi with a streamlined c-di-GMP signaling system not only controls multiple targets, as previously envisioned, but has also evolved different modes of action.

IMPORTANCE The Lyme disease pathogen, Borrelia burgdorferi, has a simple cyclic dimeric GMP (c-di-GMP) signaling system essential for adaptation of the pathogen to the complicated tick environment. The c-di-GMP effector of B. burgdorferi, PlzA, has been shown to regulate multiple cellular processes, including motility, osmolality sensing, and nutrient utilization. The findings of this study demonstrate that PlzA not only controls multiple targets but also has different functional modalities, allowing it to act as both positive and negative regulator of the glp operon expression. This work highlights how bacteria with a small genome can compensate for the limited regulatory repertoire by increasing the complexity of targets and modes of action in their regulatory proteins.

INTRODUCTION

Bis-(3′-5′)-cyclic dimeric GMP (c-di-GMP) is a ubiquitous bacterial second messenger involved in the regulation of a plethora of biological processes in bacteria. It modulates multiple physiological functions, including bacterial transition from motility to sessility, metabolism, virulence, biofilm formation and dispersal, cell cycle, and production of secondary metabolites (1, 2). The turnover of c-di-GMP is controlled by diguanylate cyclases (DGCs), which synthesize c-di-GMP, and by phosphodiesterases (PDEs), which degrade it. Most free-living bacteria have multiple copies of DGCs and PDEs. Multiple c-di-GMP signaling pathways allow bacteria to sense and respond to diverse environmental signals.

Borrelia burgdorferi is a spirochetal pathogen that causes Lyme disease, the most prevalent vector-borne infection in the United States (3). In contrast to many other bacteria, B. burgdorferi has a streamlined c-di-GMP signaling cascade that involves a single DGC and two PDEs (4–11). Hk1-Rrp1, one of the two-component systems in B. burgdorferi, is responsible for the production of c-di-GMP. The response regulator Rrp1 contains an N-terminal response regulator receiver domain and a C-terminal GGDEF domain involved in c-di-GMP synthesis (7). The DGC activity strictly depends on the phosphorylation status of Rrp1, which is controlled by the cognate histidine sensor kinase Hk1 (4, 8, 12). Degradation of c-di-GMP in B. burgdorferi is carried out by two evolutionarily distinct PDEs, the EAL domain-containing PdeA (BB0363) and the HD-GYP domain-containing PdeB (BB0374) (9, 10). Inactivation of hk1 or rrp1 abolishes c-di-GMP synthesis (6, 13), while deletion of pdeA and pdeB leads to c-di-GMP accumulation (9, 10). Both the rrp1 and hk1 mutants have survival deficiency in feeding ticks and are unable to complete the B. burgdorferi enzootic life cycle, indicating that c-di-GMP is essential for spirochetal survival within the tick vector (5, 6, 8).

How c-di-GMP signaling modulates spirochetal survival in ticks remains poorly understood. We and other groups have shown that one of the defects of the rrp1 and hk1 mutants is associated with its inability to utilize glycerol, chitobiose, and N-acetylglucosamine (5, 13, 14). Glycerol is produced by certain insects, as well as by other arthropods, as a cryoprotective molecule. Expression of the glycerol uptake and metabolism operon, glpFKD (glp), is essential for the fitness of spirochetes in ticks (5, 15). Constitutive glp expression improves survival of the rrp1 mutant in ticks (5). Chitobiose is a major component of the tick cuticle and an important source of N-acetylglucosamine for cell wall synthesis in B. burgdorferi, and supplementing N-acetylglucosamine in the tick midguts also partially rescues the rrp1 mutant defect (14). In addition to nutritional requirement, the c-di-GMP signaling system plays multifunctional roles, including chemotaxis, motility, cell envelope synthesis, and osmolality sensing (6, 9–11, 13, 16). To be noted, besides c-di-GMP, multiple factors were involved in the adaptation of the spirochete to its tick vector (for a recent review, see reference 17).

Given the importance of glycerol metabolism to the tick cycle of B. burgdorferi, we focused on elucidating the mechanism underlying the regulation of the glp operon by c-di-GMP. We found that c-di-GMP positively regulates the glp operon via PlzA, a previously identified PilZ domain-containing c-di-GMP effector in B. burgdorferi (18, 19). Surprisingly, in the absence of c-di-GMP, PlzA functions as a negative regulator of glp expression. Our findings reveal the bimodal mechanism of regulation of glp expression in B. burgdorferi via PlzA. This study enhances our understanding of the mechanisms by which bacteria with streamlined genomes compensate for the limited regulatory repertoire by increasing the number of targets and modes of action in their regulatory proteins.

RESULTS

c-di-GMP-dependent control of glp operon expression.

Earlier, we showed that expression of the glycerol utilization glp operon is dependent on c-di-GMP levels (5). To identify conditions where c-di-GMP-dependent regulation of glycerol utilization is maximal in vitro, we examined the effects of growth phase on the level of glp operon transcription. Our results showed that the transcriptional level of glpF, the first gene in the glp operon, was highest during in vitro growth in the mid-logarithmic phase compared to that in the early or post-logarithmic phases (Fig. 1A). This growth phase-dependent regulation proved to require the DGC Rrp1, whose deletion dramatically reduced glp expression in all growth phases (Fig. 1A). Furthermore, the transcriptional level of glpF was 4-fold higher in bacteria grown in the medium in which glycerol was the primary carbon source (Barbour-Stoenner-Kelly [BSK]–glycerol medium) compared to those grown in the medium in which glucose was the primary carbon source (BSK-glucose medium) (Fig. 1B). Subsequent experiments were done in mid-logarithmic-phase cultures.

FIG 1.

Regulation of glp expression by c-di-GMP. (A) Wild-type strain 5A4NP1 (WT), the rrp1 mutant (Rrp1⁻), and the isogenic complementation strain (Rrp1^com) were cultivated in regular BSK-II (BSK-glucose) medium at 37°C and harvested at different growth phases. E, mid-logarithmic phase (5 × 10⁶ cells per ml); M, late-logarithmic phase (2 × 10⁷ cells per ml); S, stationary phase (5 × 10⁸ cells per ml). (B) The same set of strains was also cultivated in either BSK-glucose or BSK-glycerol medium to mid-logarithmic phase. (C) Wild-type B. burgdorferi strain B31 A3 (WT) and the isogenic pdeB overexpression strain (WT/P_flgB-pdeB) cells were cultivated in the BSK-II medium to mid-logarithmic phase at 37°C. RNA was isolated from all of the cultures and subjected to qRT-PCR analysis for glpF expression. In panels A to C, values represent the average numbers of glpF transcripts per 100 copies of flaB from three biological replicates. Error bars indicate standard deviations. (D) Growth curve of WT and WT/P_flgB-pdeB in the BSK-glycerol medium. *, P < 0.05; **, P < 0.001.

To gather further evidence of the importance of c-di-GMP, we analyzed glp expression in the low-c-di-GMP strain that overexpresses pdeB, the c-di-GMP PDE gene. glpF expression was significantly reduced in the pdeB overexpression strain (Fig. 1C), and growth of this strain in the BSK-glycerol medium was impaired (Fig. 1D), similarly to that of the phenotype reported previously in the rrp1 mutant (5).

In an additional experiment, we lowered intracellular c-di-GMP levels by expression in B. burgdorferi of a well-characterized heterologous c-di-GMP binding protein, YcgR, from Escherichia coli (20). This protein was expected to sequester c-di-GMP and thus effectively lower its intracellular levels. A shuttle vector containing a ycgR gene under the control of a strong and constitutive flgB promoter was transformed into the wild-type strain. The results showed that the glpF expression was significantly reduced compared to that in the parental strain (Fig. 2), while expressions of other genes, such as recA (Fig. 2B) and ospC (data not shown), were not affected. Consistent with the reduced glpF expression, this strain had reduced growth in the BSK-glycerol medium (Fig. 2C). These observations suggest that overexpression of the high-affinity c-di-GMP binder YcgR sequestered c-di-GMP and affected the availability of c-di-GMP, leading to reduced glp expression.

FIG 2.

Overexpressing a heterogeneous c-di-GMP sequesterer, YcgR, reduces glp expression. (A and B) Cells of wild-type B. burgdorferi strain B31 A3 (WT) and B31 A3 carrying a flgB-driven ycgR gene from E. coli (WT/P_flgB-ycgR) were cultivated in the BSK-II medium at 37°C, harvested at the logarithmic phase, and subjected to RNA isolation and qRT-PCR analysis for glpF (A) or recA (B) transcript levels. (C) Growth curve of WT and WT/P_flgB-YcgR in the BSK-glycerol medium. Values indicate the fold changes compared with the wild-type strain from three biological replicates (±standard deviations). *, P < 0.05; **, P < 0.001.

PlzA is a positive regulator for glp transcription.

We examined whether the c-di-GMP binding protein PlzA affects glp operon expression. Caimano et al. previously reported that the plzA mutant abolished glp expression (13). We confirmed this finding (Fig. 3A). Furthermore, we complemented the plzA mutant with a shuttle vector expressing the plzA gene from its native promoter, which was lacking in the earlier study (13). If PlzA played an important role in glp operon expression, the growth of the plzA mutant would also be expected to be impaired in the medium containing glycerol as the primary carbon source. Consistent with this expectation, growth of the plzA mutant in BSK-glycerol medium was impaired, similarly to that of the phenotype of the rrp1 mutant (Fig. 3B). Complementation fully restored glpF expression (Fig. 3A) and spirochete growth (data not shown).

FIG 3.

PlzA is a positive regulator for glp expression. (A) Wild-type Borrelia strain B31 A3 (WT), the plzA mutant (PlzA⁻), and its complemented strain (PlzA^com) were cultivated in the BSK-II medium to mid-logarithmic phase at 37°C. RNAs were isolated from these cells and subjected to qRT-PCR analysis for glpF transcript levels. Values indicate the fold changes (with the mean value of the wild-type strain as 1.0) from three biological replicates (±standard deviations). (B) Growth defect of the plzA mutant with glycerol as the carbon source. Wild-type strain (WT), the rrp1 mutant (Rrp1⁻), and the plzA mutant (PlzA⁻) were cultivated in the BSK-glycerol medium. *, P < 0.01; **, P < 0.001.

The role of c-di-GMP in PlzA-dependent control of glp expression.

If PlzA is a c-di-GMP-dependent positive regulator for glp, plzA overexpression might be expected to upregulate glp expression. To test this hypothesis, we introduced into the wild-type B. burgdorferi strain 5A4NP1 a shuttle vector, pMH105, which expresses the plzA gene under the constitutive flgB promoter. According to quantitative reverse transcription-PCR (qRT-PCR) analysis, plzA transcript in the plzA overexpression strain was increased by more than 75-fold, compared to wild-type levels (Fig. 4B). The overexpression levels of PlzA were also confirmed by Western blot using anti-FLAG antibodies (Fig. 4C). To our surprise, overexpression of plzA significantly reduced, rather than increased, glpF transcription in the wild-type strain (Fig. 4).

FIG 4.

Overexpressing plzA in wild-type B. burgdorferi represses glp expression, which can be rescued by cooverexpression of plzA and rrp1. Cells of wild-type strain B31 5A4NP1 (WT), WT overexpressing plzA (WT/P_flgB-plzA), and WT cooverexpressing plzA and rrp1 (WT/P_flgB-plzA+P_flaB-rrp1), were cultivated at 37°C in the BSK-II medium, harvested at mid-logarithmic phase, and subjected to RNA extraction and qRT-PCR analysis. (A and B) The levels of glpF (A) and plzA (B) expression were reported as fold changes, with the mean value of wild-type strain as 1.0. All data were calculated from three biological replicates (±standard deviations). *, P < 0.05; **, P < 0.01. (C) Western blot analysis of overexpression of PlzA using monoclonal antibody specific to FlaB or FLAG.

Based on the results from the plzA deletion and overexpression experiments, we speculate that the c-di-GMP-bound form of PlzA functions as a positive regulator, while the ligand-free PlzA protein functions as a negative regulator. When PlzA is overproduced in the wild type, the ligand-free PlzA form outcompetes the c-di-GMP-bound PlzA, leading to glp operon repression. If so, overexpression of both plzA and rrp1 should rescue glp expression. To explore this possibility, we coexpressed rrp1 from the flaB promoter along with plzA. We found that overexpression of both rrp1 with plzA rescued the glp expression deficiency (Fig. 4), thus suggesting that glp operon expression depends on the ratio of PlzA and c-di-GMP.

To gain further evidence that PlzA has dual roles in glp regulation, dependent on c-di-GMP levels, we overexpressed plzA in the rrp1 mutant, where PlzA is predicted to function solely as a negative regulator, since no c-di-GMP is present. Indeed, overexpression of PlzA in the rrp1 mutant lowered glpF expression (Fig. 5A). Furthermore, the expression level was lower than that of the wild-type strain overexpressing plzA (Table 1).

FIG 5.

Overexpressing plzA in the rrp1 mutant or overexpressing plzA_DXXXD enhances repression of glp expression. (A and B) Cells of wild-type strain B31 5A4NP1 (WT), the rrp1 mutant (Rrp1⁻), and Rrp1⁻ overexpressing PlzA (Rrp1⁻/P_flgB-plzA) (A) or WT, the plzA mutant overexpressing plzA (PlzA⁻/P_flgB-plzA), and the plzA mutant overexpressing the mutated plzA (PlzA⁻/P_flgB-plzA_DxxxD) (B) were cultivated at 37°C in the BSK-II medium, harvested at mid-logarithmic phase, and subjected to RNA extraction and qRT-PCR analysis. The levels of glpF expression were reported as fold changes, with the mean value of wild-type strain as 1.0. All data were calculated from three biological replicates (±standard deviations). *, P < 0.01; **, P < 0.001. (C) Western blot analysis of overexpression of PlzA or PlzA_DXXD using monoclonal antibody specific to FlaB or FLAG.

TABLE 1.

Relative glpF transcript levels among various strains of B. burgdorferi^a

Strain	Estimated activation level by PlzA	Estimated repression level by PlzA	Relative glpF transcript level
Wild-type B. burgdorferi (WT)	Normal	No	100.6 ± 7.7
WT overexpressing plzA (WT/PflgB-plzA)	Normal	High	37.0 ± 9.0*
WT cooverexpressing plzA and rrp1 (WT/PflgB-plzA +PflaB-rrp1)	High	High	78.3 ± 3.2
rrp1 mutant (Rrp1−)	No	Normal	15.3 ± 0.9**
rrp1 mutant overexpressing plzA (Rrp1−/PflgB-plzA)	No	High	2.5 ± 0.3**
plzA mutant overexpressing plzA (PlzA−/PflgB-plzA)	Normal	High	53.1 ± 0.5*
plzA mutant overexpressing plzADXXXD (PlzA−/PflgB-plzADxxxD)	No	High	5.7 ± 0.6**

^aRelative transcription levels of glpF were normalized with the level of flaB gene expression. The glpF expression level in the wild-type (WT) strain was set as 100. Standard deviations from three biological replicates are shown. Student's t test was performed for each group relative to that of the WT for the comparison of glp expression levels. *, P < 0.01; **, P < 0.001.

We also examined the impact of a PlzA mutant impaired in c-di-GMP binding. PlzA with double amino acid substitutions (R150D and R154D), designated PlzA_DxxxD, was demonstrated to be impaired in c-di-GMP binding (18). We generated a shuttle vector expressing this mutant and found that it did not rescue glpF expression in the plzA mutant, as expected (Fig. 5B). Expressions of wild-type and mutagenized PlzA were similar, as shown by Western blotting (Fig. 5C). More importantly, overexpressing plzA_DXXXD in the plzA mutant resulted in a lower level of glpF expression than that resulting from overexpression of plzA (Fig. 5B). These data suggest that PlzA with impaired c-di-GMP binding functions solely as a negative regulator for glp expression.

DISCUSSION

Bacteria with multiple c-di-GMP signaling components often dedicate specific c-di-GMP effectors for different functions. B. burgdorferi employs a simple c-di-GMP system involving one DGC, two PDEs and the effector protein, PlzA. Amazingly, this system is capable of modulating multiple physiological processes, including motility, nutrient utilization, and virulence (5, 6, 8, 13, 14, 19). The findings of this and previous studies demonstrate that PlzA not only is involved in regulating multiple spirochetal activities but also has several operational modalities. We showed that it functions as both a positive and a negative regulator of glp operon expression. This finding reveals how a bacterium with a streamlined genome, such as B. burgdorferi, increases complexity at the regulator level to compensate for its limited regulatory repertoire.

Rrp1 is the only DGC in B. burgdorferi, which has been supported not only by the genome sequence analysis but also by the measurement of intracellular c-di-GMP concentration of the wild type and the rrp1 mutant (6, 13). Caimano et al. first reported that the plzA mutant has reduced glp expression (13), suggesting that PlzA is the positive regulator of glp expression; however, no complementation was performed. In this study, we performed the complementation experiment, thus solidifying the conclusion that PlzA is a positive regulator for glp expression (Fig. 3). We provided several lines of evidence that the positive regulatory function of PlzA requires c-di-GMP. (i) The DGC, Rrp1, is a positive regulator of glp expression (Fig. 5) (21). (ii) Depletion of c-di-GMP, either by overproduction of the PDE, PdeB (Fig. 1C), or by the heterologous c-di-GMP effector YcgR reduces glp expression (Fig. 2). (iii) PlzA impaired in c-di-GMP binding, PlzA_DXXD, decreases glp expression (Fig. 5B). Of note, we used glpF expression as an indicator for glpFKD operon expression. However, several reports indicate that the transcription of the glp operon is complicated and that glpD also has its own promoter (22). In this study, we focused on the transcriptional regulation of glpF.

Several lines of evidence led us to the conclusion that PlzA functions as a negative regulator of glp expression in the absence of c-di-GMP. (i) Overexpression of plzA reduced glp expression, yet simultaneous overexpression of rrp1 and plzA rescued the defect of glp expression (Fig. 4). (ii) Overexpression of plzA in the rrp1 mutant further lowered glp expression compared to the level in the rrp1 mutant (Fig. 5) or compared to plzA overexpression in the wild type (Fig. 4 and 5; see quantitation in Table 1). (iii) Overexpression of PlzA_DXXD also further lowered glp expression compared to overexpression of the wild-type PlzA (Fig. 5B and Table 1).

How PlzA functions as both positive and negative regulator for glp expression remains unclear. One obvious possibility is that PlzA functions as a DNA-binding protein that directly regulates gene expression, as observed in some PliZ domain-containing c-di-GMP effectors, such as MrkH in Klebsiella pneumoniae (23). PlzA in B. burgdorferi is a “stand-alone” c-di-GMP effector comprising only the PilZ domain. PlzA lacks a predicted DNA-binding domain. We performed electrophoretic mobility shift assay (EMSA) experiments, and the results showed that PlzA failed to bind to the glp operon promoter region in the absence or presence of c-di-GMP (data not shown), suggesting that PlzA may not function as a direct transcriptional regulator for glp expression. In other bacteria, PilZ domain proteins have been shown to be involved in multiple cell functions, including biofilm formation (24, 25), cell motility (20), exopolysaccharide synthesis (26, 27), and the cell cycle (2). Crystal structures of the PilZ domain proteins exhibited distinct c-di-GMP binding stoichiometry, mechanistic principles, and conformational changes, which likely explains the evolutionary success of the PilZ domain as a multifunctional c-di-GMP signal transducer (28–31). The dual activator/repressor roles of c-di-GMP binding proteins are not unprecedented. For example, upon c-di-GMP binding, the transcriptional regulator from Pseudomonas aeruginosa, FleQ, can be converted from a repressor to an activator of the pel exopolysaccharide operon (32). Interestingly, we previously showed that PlzA, in addition to having both positive and negative roles in regulating glp expression, acts as a positive and negative regulator for yet another gene, bosR (21). We postulate that PlzA indirectly regulates glp and bosR expression by interacting with an as-yet-unidentified partner(s) and that conformational changes in response to the presence or absence of c-di-GMP switches between the positive and negative roles of PlzA. In this regard, conformational change of PlzA of B. burgdorferi and PlzC of Borrelia hermsii upon binding to c-di-GMP has been demonstrated recently by Mallory et al. (33). It is likely that PlzA interacts with more than a single partner, because it exerts its effect on a diverse range of downstream processes, such as glp and bosR expression, N-acetylglucosamine utilization, cell envelope synthesis, motility, and osmolarity (11, 13, 14, 19, 21, 28). Further studies are necessary to identify the PlzA partners and to decipher how it can differentially affect its partners in the c-di-GMP-free and c-di-GMP-bound states.

In summary, our study, along with previous findings by others, supports the following model, depicted in Fig. 6. When the spirochete is in the tick vector, where glycerol is abundantly present, the Hk1-Rrp1 pathway is activated, resulting in an elevated level of c-di-GMP within B. burgdorferi. PlzA binds to c-di-GMP, presumably interacts with a yet-to-be-identified regulator, and activates the glp operon for glycerol uptake and utilization. During mammalian infection, when glucose is abundant as a preferred carbon source for B. burgdorferi, Hk1-Rrp1 is not activated, and the c-di-GMP level within the spirochetes is low. Unbound PlzA plays a negative role in glp expression with an unknown mechanism. Thus, c-di-GMP appears to play a role in the catabolic switch of B. burgdorferi where a classical catabolic repression system is lacking. Of note, different functions in bound and unbound forms of PlzA in ticks and mammals have been proposed by Mallory et al. (33), based on the fact that the rrp1 mutant and the plzA mutant have different phenotypes in the enzootic cycle of B. burgdorferi (5, 6, 19). In some isolates of B. burgdorferi, there is an additional copy of PilZ-containing c-di-GMP receptor, PlzB, and a recent report suggests that it could not complement the function of PlzA (34). In addition to c-di-GMP and PlzA, RpoS and guanosine pentaphosphate [(p)ppGpp] are also involved in regulation of glp expression (12, 13, 22, 35–37). RpoS negatively regulates glp expression (37), whereas (p)ppGpp positively regulates glp expression (12, 22). Further elucidation of how these multiple layers of glp regulation are orchestrated and the mechanisms underlying these regulations is warranted.

FIG 6.

Model for the transcriptional regulation of glp operon by PlzA. Upon the histidine kinase Hk1 sensing a yet-to-be-identified periplasmic signal, Hk1 phosphorylates the receiver domain of Rrp1 by phosphorelay and activates the diguanylate cyclase activity of its GGDEF domain, leading to the production of c-di-GMP. The turnover of c-di-GMP is controlled by the phosphodiesterases PdeA and PdeB. When PlzA binds to c-di-GMP, it positively regulates glp expression, in an indirect fashion. When c-di-GMP concentration is low, unbound PlzA plays an inhibitory role in glp expression. Other factors, including RpoS and (p)ppGpp, are also involved in regulation of glp expression.

MATERIALS AND METHODS

Bacterial strains and culture conditions.

The Borrelia strains and plasmids used in this study are listed in Table 2. Low-passage-number, virulent B. burgdorferi B31-derived strains 5A4NP1 and A3 were used as wild-type strains in the current study. All Borrelia strains were cultivated in BSK-II medium (BSK-glucose) with 6% rabbit serum at 37°C with 5% CO₂. BSK-glycerol medium was prepared according to a previous report by replacing glucose and CMRL 1066 in the regular BSK-II medium with 0.6% glycerol and glucose-free CMRL 1066, respectively (5). Relevant antibiotics were added to the cultures with the following concentrations when necessary: 300 μg/ml kanamycin, 100 μg/ml streptomycin, 100 μg/ml gentamicin, and 50 ng/ml erythromycin. The shuttle vector was constructed and maintained in E. coli DH5α cells. The transformation of Borrelia and screening was conducted according to a previously report (38).

TABLE 2.

Borreliae strains and plasmids used in this study

Strain or plasmid	Phenotype	Source or reference
Strain
B31 A3	Wild type, a low-passage-no., virulent strain derived from B31-MI	41
B31 5A4NP1	Wild type, a low-passage-no., virulent strain derived from B31	42
Rrp1−	The rrp1 mutant in 5A4NP1	5
Rrp1com	The rrp1 mutant complemented with a wild-type rrp1 gene	5
WT/PflgB-plzA	5A4NP1 containing pMH105 for overexpressing plzA	This study
Rrp1−/PflgB-plzA	Rrp1− containing pMH105 for overexpressing plzA	This study
WT/PflgB-plzA+PflaB-rrp1	5A4NP1 containing pJJ054 for cooverexpressing plzA and rrp1	This study
PlzA− (A3)	The plzA mutant in B31 A3	20
PlzAcom	The plzA mutant complemented with a wild-type plzA gene	20
PlzA− (5A4NP1)	The plzA mutant in B31 5A4NP1	22
Plasmid
pJD48	pBSV2-based shuttle vector with a promoterless lucBb+ (Borrelia codon-optimized luciferase gene, Kanr)	43
pJD50	pBSV2-based shuttle vector with an erythromycin resistance gene (Ermr)	43
pJD55	pBSV2-derived shuttle vector with a kanamycin resistance gene (Kanr)	43
pBT50	pJD50 with a promoterless lucBb+, Ermr	This study
pJJ50	pBT50 with the replacement of Erm marker with a gentamicin resistance marker (Genr)	This study
pMH105	pBSV2-derived shuttle vector containing a flgB promoter-driven plzA	22
pJJ105	pMH105 with a mutated plzA, resulting in R150D and R154D substitutions	This study
pJJ054	A flaB promoter-driven rrp1 inserted into pMH105 for cooverexpressing rrp1 and plzA	This study
pJJ030	plzA cloned into pET100 for expressing recombinant PlzA	This study
pJJ067	pBSV2-derived shuttle vector containing a flgB promoter-driven ycgR; the kanamycin resistance marker was replaced by a streptomycin resistance marker	This study
pYY001	A flgB promoter-driven pdeB inserted into pJD55	This study

RNA extraction and qRT-PCR.

RNA samples were extracted from different phases of B. burgdorferi cultures (from around 10⁸ cells) using the RNeasy minikit (Qiagen, Valencia, CA) according to the protocol of the manufacturer. Contaminating DNA in the RNA samples was removed by DNase I (New England BioLabs, Ipswich, MA) digestions and confirmed by PCR amplification of the flaB gene of B. burgdorferi. Three independent culture samples were prepared for each strain. The cDNA was prepared from 1 μg RNA using Superscript III reverse transcriptase with random primers (Invitrogen, Carlsbad, CA). qRT-PCR was performed using SYBR green PCR master mix (Applied Biosystems, Foster City, CA) on an ABI 7000 sequence detection system. The flaB gene of B. burgdorferi was used as a reference. The quantity of the targeted genes and flaB in cDNA samples was calculated as in our previous report (5). The relative transcription level was determined by the threshold cycle (2^−ΔΔCT) method (39).

Construction of pdeB and ycgR overexpression Borrelia strains.

The open reading frame of pdeB was amplified from the Borrelia genome and digested with the restriction enzymes NdeI and PstI. The resulting fragment was cloned into pJD55 that had been digested with the same restriction enzymes to generate pYY001. C-terminal human influenza hemagglutinin (HA)-tagged PdeB expression is driven by a flgB promoter in this shuttle vector, which was verified by sequencing before transforming into B. burgdorferi. The same strategy was used for the overexpression of ycgR, except that the shuttle plasmid was modified by replacing the kanamycin resistance marker with a streptomycin resistance marker. The resulting plasmid was designated pJJ067. All transformants were verified by qRT-PCR and immunoblot analysis using anti-HA monoclonal antibody to confirm the constitutive overexpression of the target genes.

Cooverexpression of plzA (bb0733) and rrp1 (bb0419) in Borrelia.

A shuttle plasmid, pMH105, was used for the overexpression of plzA in B. burgdorferi, in which FLAG-tagged PlzA was constitutively expressed under a flgB promoter (21). For the cooverexpression of plzA and rrp1, the open reading frame of rrp1 was PCR amplified from the genome of 5A4NP1 with primer sets containing BamHI and KpnI sites in the forward and reverse primers, respectively. The PCR fragment was then inserted into the BamHI and KpnI sites of pMH105 to construct pJJ054. Expression of the rrp1 gene in the resulting plasmid was constitutive and driven by a flaB promoter. All plasmids were confirmed by sequencing before being transformed into Borrelia strains, and the transformants were identified by immunoblot analysis using anti-FLAG and anti-Rrp1 antibodies, respectively. The constitutive overexpression of rrp1 and plzA was confirmed by qRT-PCR analysis.

PlzA site-directed mutagenesis.

A double amino acid substitution mutation (R150D and R154D) was introduced into plzA on the plasmid of pMH105 using a QuikChange site-directed mutagenesis method (40). The resulting plasmid, pJJ105, was sequenced to confirm the mutagenesis of PlzA before it was transformed into the plzA mutant of B. burgdorferi 5A4NP1 strain.

SDS-PAGE and immunoblot analysis.

B. burgdorferi strains were inoculated into BSK-II medium and grown to mid-log phase before the cells were harvested by centrifugation at 7,000 × g and washed twice with phosphate-buffered saline (PBS; 50 mM, pH 7.4). The pellets were collected for sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and immune blotting analysis. Immunoblots were developed using the SuperSignal West Pico chemiluminescent substrate according to the manufacturer's instructions (Pierce, Rockford, IL) and subsequently detected by exposure to X-ray film.

Primers.

The primers used in this study were synthesized by Integrated DNA Technologies, Inc. The purposes and sequences of the primers are shown in Table 3.

TABLE 3.

Primers used in this study

Purpose	Oligonucleotide name	Sequence (5′–3′)
Overexpression of pdeB	HDGYP F	AGGAATGACATATGCAAAATTCTGAAAGC
	HDGYP-HA R	CATGCCTGCAGTTAAGCGTAATCTGGAACATCGTATGGGTATAT
Overexpression of ycgR	pHX55-ycgR NdeI HA F	CGTTCATATGTATCCTTATGATGTTCCTGATTATGCTAGTCATTACCATGAGCAGTTCCTG
	pHX55-ycgR PstI R	ATGCCTGCAGTCAGTCGCGCACTTTGTCCGCTTTTTC
PlzA site-direct mutagenesis	PlzA DxxxD F:	GATATTCATGAGGATATTATTATCGATAAAGATTCTATTAGAAAGC
	PlzA DxxxD R:	ATCCTCATGAATATCCTGATTTTGCCCAAGCTTTAAATCAAG
flaB qPCR	qPCR-flaB-XF F	GCTCCTTCCTGTTGAACACCC
	qPCR-flaB-XF R	CTTTTCTCTGGTGAGGGAGCTC
glpF qPCR	q240F	ACACAGCGACAATCCATTTA
	q240R	AGCATAACCGTTCATTCCTC
plzA qPCR	qBB0733F	ATGCTTTTATCTAGAAAAATAAGAGATTATGG
	qBB0733R	ATTGAAATAATCATGGATCAACATAG

Statistical analyses.

All statistical analyses were performed using Microsoft Excel and validated with Prism version 5 (GraphPad). Statistical significance was determined using an unpaired t test (two-tailed distribution with two-sample equal variance) when comparing two conditions. The P values are indicated in the figures, and P < 0.05 was considered statistically significant unless otherwise stated.