Reciprocal Expression of ospA and ospC in Single Cells of Borrelia burgdorferi

Abstract

Outer surface proteins (Osp) A and C of the Lyme disease spirochete (Borrelia burgdorferi) are selectively produced and of functional significance in the tick vector and mammalian host, respectively. Some studies indicate a simple, reciprocal relationship where the signals and pathways that turn on ospC also turn off ospA. Other studies indicate a more complex regulation where many spirochetes produce both proteins and others produce one of the proteins or neither protein. Here, we have used flow cytometry to characterize ospA and ospC transcript and protein levels in individual bacterial cells grown in culture. The results support a simple, reciprocal model where, at the level of single cells, the transcription of ospC is linked to the repression of ospA. We also demonstrate that under conditions conducive for OspC production, spirochetes display an “all or none” response, with some cells displaying high levels of ospC transcription and others demonstrating little or no transcription. Despite the reciprocal regulation of ospA and ospC at the single-cell level, we propose that spirochetes display an array of phenotypes due to stochasticity in the pathways that regulate osp expression and the slow turnover of outer surface proteins.

Borrelia burgdorferi, the spirochete responsible for Lyme disease, is transmitted by Ixodes ticks (3). B. burgdorferi persistently colonizes the gut of ticks. When infected ticks feed, the spirochetes multiply within the gut, migrate to the tick's salivary glands, and infect the vertebrate host (1). Within the feeding tick, the spirochetes alter the expression of many genes in preparation for transmission and infection of the new host (1, 26). B. burgdorferi outer surface proteins (Osp) A and C have served as a paradigm for understanding the regulation of bacterial gene expression within feeding ticks. In ticks, ospA is predominantly expressed before the blood meal, whereas ospC is induced during the blood meal (9, 17, 22, 23). The functions of these two proteins are consistent with their pattern of expression, where OspA is required for colonizing the vector and OspC is required for infecting the host (12, 19, 29).

Temperature, pH, and cell density act as signals for regulating the expression of ospA and -C in culture, and these signals are likely to play a role in the feeding tick as well (26). Proteomic and microarray studies with cultured spirochetes grown in “tick-like” (low temperature, high pH) or “host-like” (high temperature, low pH) conditions have led to the identification of large subsets of Borrelia proteins and genes with “OspA-like” or “OspC-like” patterns of expression (18, 20). The bacterial signaling pathway regulating the expression of ospC and ospC-like genes has been characterized in some detail (6, 7, 14, 24, 27, 28). The pathway is activated by a two-component system consisting of a sensor with a histidine kinase domain (HK2) and a cytoplasmic response regulator protein (Rrp2) (14, 27). Activated Rrp2, together with the alternative sigma factor RpoN, induces the expression of many genes, including a second alternative sigma factor, RpoS (24). RpoS activates the transcription of ospC and ospC-like genes associated with tick transmission and host infection (6, 7, 10).

The pathways and signals regulating ospA and ospA-like genes expressed in the vector are not as well characterized as the ospC expression signaling pathway. Some studies indicate a simple reciprocal relationship where the signals and pathways that induce ospC expression within feeding ticks also repress ospA expression (21). Studies also indicate that even under ideal conditions for ospC expression, the bacterial population is heterogeneous, with many spirochetes producing both proteins and others producing either one or neither of the two proteins, indicating a more complex regulation of ospA and -C (13, 17). We have developed flow cytometry as a method for following the phenotypes of individual spirochetes and applied this method to better understand the regulation of ospA and ospC.

MATERIALS AND METHODS

Borrelia strains and culture conditions.

A low passage culture of B. burgdorferi strain B31 (originally isolated from a tick in Shelter Island, NY) was provided to us by the Centers for Disease Control and Prevention, Fort Collins, CO (4). The stock was cloned on solid Barbour-Stoenner-Kelly H (BSK-H) medium and named B31-C1. Strains A3ntrA-Gm (rpoN-null mutant) and A3-Gm (control strain for rpoN-null mutant) were obtained from Frank Gherardini, NIAID, Rocky Mountain Laboratories, Hamilton, MT (10). Strains B31-CGFP and B31-FGFP were obtained from James Carroll, University of Pittsburgh, Pittsburgh, PA (8). B31-CGFP is B31 clone A3 harboring the plasmid pBSVΦ(ospCp-gfp), and B31-FGFP is B31 clone A3 harboring the plasmid pBSVΦ(flaBp-gfp). Plasmid pBSVΦ(ospCp-gfp) has gfp under the control of the ospC promoter, whereas in plasmid pBSVΦ(flaBp-gfp), gfp is under the control of the flaB promoter. Cultures were grown in BSK-H complete medium (Sigma, St Louis, MO). The culture was grown at 35°C until the cells reached a density of 1 × 10⁵ cells/ml. Cells were then transferred into fresh BSK-H complete medium for experiments.

Direct fluorescence antibody staining and flow cytometry analysis of Borrelia.

To stain spirochetes, fluorescently labeled monoclonal antibodies (MAbs) against OspA (C3.78 MAb-Alexa488) and OspC (B5 MAb-Alexa647), provided by Fred Kantor (Yale University) and Lamine Mbow (CDC, Fort Collins, CO), respectively, were used (11, 15). B. burgdorferi cultures were harvested and incubated at room temperature for 30 min with anti-OspC and anti-OspA MAbs at room temperature. Labeled spirochetes were washed twice with phosphate-buffered saline (PBS) and analyzed on a MoFlo modular flow cytometer (Cytomation, Inc., Fort Collins, CO) with a 15-mW, 488-nm argon ion laser and a Coherent Innova-90 krypton laser at 350 to 360 nm. Data were acquired and analyzed using Summit V.3.1 (Cytomation, Inc., Fort Collins, CO).

Quantitative RT-PCR for ospA and ospC mRNA.

A culture was grown at 35°C to a density of 5 × 10⁷ bacteria/ml. Approximately 10⁶ Borrelia cells were sorted by flow cytometry into A⁺/C⁻, A⁺/C⁺, and A⁻/C⁺ populations. The sorted cells were added to a buffer containing PBS, 100 mM dithiothreitol, and RNase inhibitor (RNAsin; Promega). Cells were subjected to four freeze-thaw cycles. Genomic DNA in the cell lysates was removed using a Turbo DNA-free kit (Ambion). Reverse transcription (RT) was performed using random primers (Invitrogen). The cDNA was used for quantitative PCR using Sybr green master mix (Applied Biosystems, Foster City, CA). PCR was performed on an ABI Prism 7000 sequence detection system (Applied Biosystems, Foster City, CA), using the following specific primer pairs: flaB-F (TTTCAGGGTCTCAAGCGTCT) and flaB-R (TGTTGAGCTCCTTCCTGTTG), ospC-F (GAAAGAGGTTGAAGCTTGC) and ospC-R (ATTGCATAAGCTCCCGCTAA), and ospA-F (GCAGCCTTGACGAGAAAAAC) and ospA-R (GGATCTGGAGTACTTGAAGGC). The thermal conditions applied for amplification were 1 cycle at 95°C for 15 min and 50 cycles at 95°C for 15 s and 60°C for 1 min. Melting curves were generated by treating the amplified samples at 95°C for 15 s, 60°C for 20 s, and 95°C for 15 s. Transcript levels of ospA and ospC in these cell populations were calculated by the 2^−ΔΔCT method (16) (Tables 1 and 2).

TABLE 1.

Measurement of ospC transcript levels in a phenotypically distinct subpopulation of B. burgdorferi B31-C1^a

Cell phenotype	Avg CT of:		ΔCTospC-flaBb	ΔΔCTc	Normalized ospC amt relative to OspA+ OspC− cellsd
	ospC	flaB
OspA+/OspC+	30.59 ± 0.18	28.69 ± 0.47	1.90 ± 0.50	−4.74 ± 0.50	26.8 (18.8-38.1)
OspA−/OspC+	28.54 ± 0.22	29.07 ± 0.37	−0.53 ± 0.43	−7.18 ± 0.43	145.51 (107.68-196.63)
OspA+/OspC−	36.21 ± 0.29	29.56 ± 0.41	6.64 ± 0.50	0.00 ± 0.50	1.0 (0.70-1.42)

^aBorrelia cultures grown at 5 × 10⁷ cells/ml were used for sorting by flow cytometry and subsequent quantitative RT-PCR analysis.

^bThe ΔC_T value was determined by subtracting the average flaB threshold cycle (C_T) value from the average ospC C_T value. The standard deviation of the difference is calculated from the standard deviations of the ospC and flaB values.

^cThe calculation of ΔΔC_T involves subtraction of the ΔC_T calibrator value from the ΔC_T value [ΔC_T − ΔC_{T(OspA⁺ OspC⁻)}]. This is subtraction of an arbitrary constant, so the standard deviation of ΔΔC_T is the same as the standard deviation of ΔC_T.

^dRanges (in parentheses) were determined by evaluating the expression 2^−ΔΔC_T with ΔΔC_T + s and ΔΔC_T − s, where s is the standard deviation of the ΔΔC_T value.

TABLE 2.

Measurement of ospA transcript levels in a phenotypically distinct subpopulation of B. burgdorferi B31-C1^a

Cell phenotype	Avg CT of:		ΔCTospA-flaBb	ΔΔCTc	Normalized ospA amt relative to OspA− OspC+ cellsd
	ospA	flaB
OspA+/OspC+	27.29 ± 0.22	28.69 ± 0.47	−1.39 ± 0.52	−1.39 ± 0.52	2.6 (0.48-1.52)
OspA−/OspC+	29.07 ± 0.20	29.07 ± 0.37	−0.007 ± 0.42	0.00 ± 0.42	1.0 (0.74-1.33)
OspA+/OspC−	26.68 ± 0.12	29.56 ± 0.41	−2.88 ± 0.43	−2.87 ± 0.43	7.34 (5.44-9.92)

^aBorrelia cultures grown at 5 × 10⁷ cells/ml were used for sorting by flow cytometry and subsequent quantitative RT-PCR analysis.

^bThe ΔC_T value is determined by subtracting the average flaB threshold cycle (C_T) value from the average ospA C_T value. The standard deviation of the difference is calculated from the standard deviations of the ospA and flaB values.

^cThe calculation of ΔΔC_T involves subtraction of the ΔC_T calibrator value from the ΔC_T value [ΔC_T − ΔC_{T(OspA⁻ OspC⁺)}]. This is subtraction of an arbitrary constant, so the standard deviation of ΔΔC_T is the same as the standard deviation of ΔC_T.

^dRanges were determined by evaluating the expression 2^−ΔΔC_T with ΔΔC_T + s and ΔΔC_T − s, where s is the standard deviation of the ΔΔC_T value.

RESULTS

OspA and OspC phenotypes of individual spirochetes grown in culture.

Experiments were done to characterize OspA and OspC production by individual bacterial cells at different temperatures and cell densities. Strain B31-C1 cultures were started at a density of 1 × 10³ bacteria/ml and grown at 35°C until the cultures reached densities of 1 × 10⁶, 1 × 10⁷, and 5 × 10⁷ bacteria/ml. A control culture was grown at 23°C to a density of 5 × 10⁷ cells/ml. Spirochetes were stained with anti-OspA C3.78 MAb-Alexa488 and anti-OspC B5 MAb-Alexa647 and analyzed by flow cytometry. Borrelia cells were plotted, with OspA staining intensity on the x axis and OspC staining intensity on the y axis. The plot was divided into four quadrants, representing A⁺/C⁺, A⁺/C⁻, A⁻/C⁺, and A⁻/C⁻ cells (Fig. 1).

FIG. 1.

Outer membrane phenotypes of B. burgdorferi grown to different cell densities. Strain B31-C1 was grown at 23°C to a density of 5 ×10⁷ cells/ml (A) or at 35°C to densities of 1 × 10⁶ cells/ml (B), 1 × 10⁷ cells/ml (C), and 5 × 10⁷ cells/ml (D); stained with anti-OspA (αOspA) C3.78 MAb-Alexa488 and anti-OspC B5 MAb-Alexa647 antibodies; and analyzed by fluorescence-activated cell sorting (FACS). Results shown are representative of three independent experiments. The quadrants in each plot were drawn to represent subpopulations that are OspA⁻/OspC⁺ (R4), OspA⁺/OspC⁺ (R5), OspA⁻/OspC⁻ (R6), and OspA⁺/OspC⁻ (R7).

The control culture grown at 23°C and the culture grown at 35°C to a density of 1 × 10⁶ bacteria/ml consisted almost exclusively (99%) of A⁺/C⁻ bacteria (Fig. 1A and B). As the culture grew from a density of 1 × 10⁶ to 5 × 10⁷ cells/ml at 35°C, the number of bacteria producing only OspC increased from 0.1 to 24%, whereas the proportion of bacteria producing only OspA decreased from 99 to 62% (Fig. 1B to D). The shapes of the scatter plots indicate that the spirochetes first shifted to a population containing both OspA and OspC, and these double-positive bacteria further increased expression of OspC while decreasing levels of OspA (Fig. 1B to D). These results demonstrate that the cells that increase OspC are the same cells that decrease OspA.

ospC and ospA mRNA levels in spirochete populations with distinct phenotypes.

A Borrelia culture was grown to 5 × 10⁷ cells/ml and sorted by flow cytometry to isolate A⁺/C⁻, A⁺/C⁺, and A⁻/C⁺ populations. We used quantitative RT-PCR to estimate ospA and ospC transcript levels in each population. When ospC transcript levels were compared among the three populations, the double-positive spirochetes (A⁺/C⁺) and the spirochetes that produced only OspC (A⁻/C⁺) were found to have 26.8- and 145.5-fold more ospC transcripts, respectively, than the spirochetes that did not produce OspC (A⁺/C⁻) (Table 1). When ospA transcript levels were compared among the three populations, the double-positive spirochetes (A⁺/C⁺) and the spirochetes that produced only OspA (A⁺/C⁻) were found to have 2.6- and 7.34-fold more ospA transcripts, respectively, than the spirochetes that did not produce OspA (A⁻/C⁺) (Table 2). These results indicate that even the double-positive population consists of bacteria that are in the process of up-regulating ospC and down-regulating ospA. Based on these results, we interpret the double-positive phenotype as a transient phenotype most likely caused by the stability of preexisting OspA protein.

Role of RpoN in OspA down-regulation.

If ospA and ospC are reciprocally regulated in single spirochetes, it is logical to assume that the RpoN/RpoS pathway that induces ospC and ospC-like genes also represses ospA and ospA like genes (6, 7, 14, 24, 27, 28). To directly test the role of RpoN in ospA expression, we used wild-type (B31-A3 WT) and rpoN mutant (B31-A3 ΔrpoN) strains created by Fisher et al. (10) in flow cytometry experiments. The bacteria were cultured at 35°C to late log phase (5 × 10⁷ cells/ml), stained with anti-OspA MAb-Alexa488 and anti-OspC MAb-Alexa647, and then analyzed by flow cytometry. As expected, B31-A3 ΔrpoN did not produce any OspC, as RpoN is required for ospC expression (data not shown) (5, 14). When the wild-type and ΔrpoN mutant strains were examined for loss of OspA, 15.61% of B31-A3 WT and 1.21% of the ΔrpoN strain were OspA negative (Fig. 2). These results demonstrate that RpoN is required for the induction of ospC as well as for the repression of ospA observed in culture.

FIG. 2.

RpoN is required for the down-regulation of OspA in B. burgdorferi. Wild-type (B31-A3 WT) and rpoN mutant (B31-A3 ΔrpoN) strains were grown at 35°C to a density of 5 × 10⁷ cells/ml. Cultures were stained with anti-OspA (αOspA) C3.78 MAb-Alexa488 and analyzed by FACS.

Binary distribution of OspC-positive and -negative spirochetes.

The flow cytometry data indicate that even under optimal conditions for OspC production, only a subpopulation of spirochetes induces the protein (Fig. 1D). To further understand the nature of this phenomenon, we used flow cytometry to measure the mean fluorescence intensity of OspC in individual cells grown at 35°C to a density of 5 × 10⁷ cells/ml. The results were expressed by plotting mean fluorescent intensity on the x axis and bacterial cell count on the y axis (Fig. 3). Even under optimal conditions for OspC production, a small population of cells (19.11%) expressed a high level of OspC (mean fluorescent intensity = 306.06) while the majority of cells expressed low levels of OspC (Fig. 3C), comparable to expression levels found in spirochetes grown under conditions unfavorable for OspC production (mean fluorescent intensity = 11.05) (Fig. 3A and B). A similar “all or none” response was observed when we measured transcription from the ospC promoter by using B31-A3 strains created by Carroll et al. containing gfp fused to a flaB or ospC promoter (8). In Borrelia strains containing gfp fused to the ospC promoter, 5.14% of cells expressed high levels of green fluorescent protein (GFP) (mean fluorescent intensity = 63.13), while the rest of the cells in the population expressed negligible amounts of GFP (mean fluorescent intensity = 2.00) (Fig. 3D). Almost all the Borrelia cells containing GFP under the flaB promoter expressed high levels of GFP (mean fluorescent intensity = 31.12) (data not shown). These results demonstrate that ospC transcription is regulated by an “all or none” response at the level of individual bacterial cells.

FIG. 3.

ospC expression is binary. Strain B31-C1 was grown at 23°C to a density of 5 × 10⁷ cells/ml (A) or at 35°C to densities of 1 × 10⁶ cells/ml (B) and 5 × 10⁷ cells/ml (C). Strain B31A3 ospC-GFP, containing gfp fused to the ospC promoter, was grown at 35°C to a density of 5 × 10⁷ cells/ml (D). To measure OspC levels in individual spirochetes (A, B, C), the bacteria were stained with anti-OspC (αOspC) B5 MAb-Alexa647 and analyzed by FACS. Transcriptional activity from the ospC promoter (D) was determined by using FACS to measure GFP levels. The data are plotted to display the level of OspC or GFP on each spirochete on the x axis and the number of cells on the y axis. Note that under conditions conducive for ospC expression, individual cells display a binary “all or none” response instead of a graded response.

DISCUSSION

Flow cytometry is a powerful technique that can be used to study heterogeneity between cells in a population. Here, we have used this tool to understand the regulation of B. burgdorferi ospC and ospA. Most studies for characterizing ospA and ospC transcript and protein levels under different culture conditions have been done by analyzing the total protein and RNA levels in the culture. An assumption made in these experiments is that all bacteria in the culture respond in similar manners to signals that regulate gene expression. Our results demonstrate that even in a culture, where all spirochetes experience the same temperature, pH, and cell density, ospC expression is binary, with some bacteria producing high levels of the protein and others producing little or no protein. Even for in vitro conditions that enhance OspC production (high temperature and high cell density), our results demonstrate that the population consists of two subpopulations with or without OspC. The studies with the ospC promoter-GFP fusion establish that the binary response is regulated at the level of transcription.

Our results also demonstrate that, within individual spirochetes, ospA and ospC are reciprocally regulated by the RpoN/RpoS pathway. The spirochetes that increase OspC are the same ones that down-regulate OspA. In these cells, OspC protein levels increase rapidly, while OspA protein levels decrease more slowly, most likely because of the stability of preexisting OspA. This gives rise to an intermediate A⁺/C⁺ population from which the spirochetes that produce only OspC emerge as the OspA protein is degraded (Fig. 1). We conclude that heterogeneous populations of spirochetes are found due to the stability of presynthesized protein and the nonresponsiveness of some bacteria. However, individual spirochetes that change the expression patterns of ospC and ospA do so in a reciprocal manner where the RpoN/RpoS pathway activates ospC and represses ospA transcription. This conclusion is supported by a recent study (6) demonstrating that RpoS is required for ospC induction and ospA repression in vivo.

Many Borrelia genes are differentially expressed and function at a specific stage in the life cycle. There is much interest in defining the signals and mechanisms regulating gene expression in Borrelia. Our discovery about the binary “all or none” expression of ospC at the level of individual cells adds another layer of complexity to our current understanding of Borrelia gene expression. It is now well established that many microbes, especially vector-borne pathogens, display phenotypic heterogeneity even within a genetically homogenous culture (2). Stochasticity and epigenetic regulation have emerged as major mechanisms driving cell-to-cell variation in the absence of genetic variation (2). We propose that stochasticity in the pathway that regulates ospC expression accounts for the “all or none” binary response observed for OspC in this study. Once a cell enters the “on” state, a positive-feedback loop could maintain high-level expression (25). Since ospC expression involves two global gene expression regulators, RpoN and RpoS, we predict that many other genes will also display the binary pattern described here for ospC. This phenotypic heterogeneity is likely to help the Borrelia cell adapt to the different, ever-changing environments encountered during its complex life cycle in the vector and host.