INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

Borrelia burgdorferi, the spirochete that causes Lyme disease, is a remarkably adaptable bacterium. During its life cycle, it is able to survive in the midguts of vector ticks (of the Ixodes ricinus complex) both in the presence and absence of host blood and in the ticks' saliva. In addition, after withstanding the transit between a poikilothermal and a homeothermal host (usually a rodent) (16), it will colonize practically every organ of the latter. It stands to reason that numerous changes must take place on the spirochetal surface in anticipation of, or after transfer to, a new host or host environment. Several such changes have been observed, yet their functional significance remains unknown. For example, following a tick blood meal, there is a switch in the major surface protein from OspA to OspC (5, 6, 23). Another antigen, P21, appears to be selectively expressed in the mammalian host and not in ticks (3, 26). However, little else is known about other adaptive changes, especially the ones that do not involve antigens, primarily due to the difficulty in harvesting a sufficient number of spirochetes from their hosts for direct examination.

In a recent publication, we showed that the expression of two plasmid-borne genes that encode the lipoproteins P35 (7, 8) and P7.5 (11) of B. burgdorferi is upregulated in the post-logarithmic and stationary phases during growth of the spirochetes in culture (9). We demonstrated further that the upregulation of expression of these genes correlated with higher levels of their mRNA during the later stages of in vitro growth. The phenomenon of growth phase regulation of genes in B. burgdorferi is of biological relevance because the spirochetal population experiences alternating periods of rapid growth and quiescence during its transit through the ticks (4, 19). We therefore decided to investigate more thoroughly this phenomenon with the long-term objective of developing models for understanding the life cycle of the spirochete within a tick. As a result of this study, we have now identified additional candidate proteins that are differentially expressed or upregulated in B. burgdorferi during growth in vitro from logarithmic to stationary phase. We present the results of our study in this paper.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Bacterial strains. The JD1 strain of B. burgdorferi originally isolated from an Ixodes dammini (scapularis) nymph (18) was used for this study. The Escherichia coli strain used was XL1 Blue (Stratagene, La Jolla, Calif.).

In vitro culture conditions. B. burgdorferi JD1 (passage 5) was cultured in BSK-H medium (Sigma Chemical Company, St. Louis, Mo.) as described previously (20) with the following modifications. Briefly, 750 ml of fresh BSK-H medium was inoculated with 1.5 ml of a frozen stock of the JD1 strain of B. burgdorferi to yield a starting cell density of 10⁵ spirochetes/ml. The cell density was monitored daily by counting spirochetes under dark-field microscopy. Cells were harvested on days 3 (at a cell density of 4 × 10⁶ spirochetes per ml), 4 (2 × 10⁷ per ml), 6 (8 × 10⁷ per ml), and 8 (1 × 10⁸ per ml), and whole-cell lysates for protein analysis and RNA for Northern blot analysis were prepared. Sample volumes were adjusted for the various cell densities (see below). An aliquot of cells in the stationary phase (day 8) was reinoculated into fresh BSK-H medium at an initial density of 10⁵ spirochetes per ml as described above and allowed to progress to stationary phase, at which point the cells were subcultured for one final growth cycle. Cells in the log phase of this last subculture, i.e., of passage 8, hereinafter designated "log_rev" for log revertant, were also processed for protein and RNA analyses as mentioned above.

Antibodies. Mouse monoclonal antibodies H5332 and H9724, specific for the B. burgdorferi proteins OspA and flagellin, respectively, were purchased from the University of Texas Health Sciences Center (San Antonio). Additional mouse monoclonal antibodies that are specific to B. burgdorferi OspB, BmpA (P39), GroEL, and a 72-kDa protein were kindly provided by Barbara Johnson, Division of Vector-Borne Infectious Diseases, Centers for Disease Control and Prevention, Ft. Collins, Colo., and an anti-OspC monoclonal antibody (L221F8) (28) was provided by Bettina Wilske, Max von Pettenkofer Institut, Munich, Germany. Finally, the P35-specific monoclonal antibody has been described previously (9). A mouse monospecific polyclonal antibody directed against the BmpD protein of B. burgdorferi, was generated as follows. The mature BmpD polypeptide starting at Cys¹⁷ was expressed with a leader peptide, Met-Arg-Gly-Ser-His₆-Gly-Ser-Cys, in which the cysteine residue represents the first residue of the mature BmpD sequence. This His₆ affinity-peptide-tagged BmpD protein was purified with a nickel-nitrilotriacetic acid-agarose affinity column as per the instructions of the manufacturer (Stratagene). C57BL/10.j mice were intraperitoneally immunized with 15 µg of the fusion protein emulsified in complete Freund's adjuvant (Sigma). Thereafter, the mice received one of two more injections of the same antigen in Freund's incomplete adjuvant (Sigma) at 1- and 2-months intervals and were bled 2 weeks after the last immunization.

Preparation of protein samples, SDS-PAGE, and Western blotting. Cultures (100 ml for sample 1 [4 × 10⁶/ml] [see Fig. 1A] and 50 ml for the rest of the samples) were spun down at 5,000 × g at 4°C for 10 min in a tabletop centrifuge and washed with 1 ml of a wash buffer (20 mM HEPES [pH 7.6], 10 mM NaCl) (2). The washed spirochetes were then resuspended in 100 µl of the wash buffer, and the optical density at 600 nm (OD₆₀₀) of a 1:100 dilution of this was measured in a spectrophotometer. The remainder of the cells was adjusted with wash buffer and 3× sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) lysis buffer such that the final lysate concentration was equivalent to an OD₆₀₀ of 1, thus normalizing each sample by cell mass.

RNA isolation and Northern blotting. RNA isolation and Northern blotting of spirochetal samples taken at specific time points along the growth curve were performed essentially as described previously (21). Equal amounts (3 µg/lane) of the RNA samples were electrophoresed through a 1.4% agarose gel in the presence of 2.2 M formaldehyde. In addition to spectrophotometrically quantifying RNA, we stained one blot with methylene blue prior to hybridization to confirm equal loadings of the RNA samples that were obtained at the different time points.

Quantification of bands on Western and Northern blots. The blots were digitized with a scanner (Hewlett-Packard Scanjet 4c), and individual bands in the digitized images were quantified with Digital Science 1D software, version 2.02 (Eastman Kodak Company, Rochester, N.Y.).

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

In a recent paper we described the cell density-dependent expression of B. burgdorferi lipoproteins P7.5 and P35, during growth of the spirochetes in culture (9). In this study we explored the possibility that other B. burgdorferi proteins were similarly differentially expressed during growth in vitro. Samples collected at time points along the growth curve that corresponded to the logarithmic (days 3 and 4), postlogarithmic (day 6), and stationary (day 8) phases were processed for protein and RNA analysis (Fig. 1A). In order to distinguish between proteins that are truly differentially expressed versus those whose expression may be irreversibly affected for unknown reasons, we also analyzed spirochetes from the same batch after they had undergone two additional cycles of growth. Accordingly, we have designated this sample of spirochetes harvested at a cell density similar to that of the parental day 4 logarithmic phase culture "log_rev" to indicate the cells' phenotypic reversion to their predecessors' logarithmic-phase phenotype.

View larger version (42K): [in this window] [in a new window]   FIG. 1.   (A) Growth curve of B. burgdorferi JD1 (passage 6) in vitro. The arrows indicate the cell densities at which spirochetes were harvested and processed for Western and Northern blotting analyses. Numbers above the arrows designate each sample. The open arrow depicts the passage 8 logrev sampling point. (B) Silver-stained gel showing the profiles of proteins expressed at different times during culture. The filled arrowhead indicates OspC, and the open arrowhead indicates P28. The lane numbers correspond to the numbers in panel A and designate the time point at which each sample was collected. (C) Western blot developed with serum from a monkey with an acute infection of B. burgdorferi JD1 (22). Arrowheads indicate antigens whose levels of expression steadily increased with the increasing age of the culture. Lane numbers are as described for panel B.

Profile of protein expression in B. burgdorferi during growth from log to stationary phase in vitro. Initially, we analyzed the protein samples from the different time points along the growth curve by one-dimensional SDS-PAGE followed by silver staining. This method identified two proteins, 23 and 28 kDa in size, that were differentially expressed against a background of several bands that remained fairly constant in intensity (Fig. 1B). The level of expression of the 28-kDa protein was 2-fold higher in the experiment shown in Fig. 1B (Table 1) and 10-fold higher in a previous experiment (data not shown) early during the logarithmic phase than its level of expression in the stationary phase. In contrast, increasing amounts of the 23-kDa protein accumulated within the cell during growth from logarithmic to stationary phase (Table 1). The 23-kDa band corresponded to the lipoprotein OspC (discussed later). The other proteins that are also known to be differentially expressed, such as P35 and P7.5 (9) and the antigens newly identified in this study by Western blotting (see below), are not apparent in the silver-stained gel, perhaps due to their poor overall levels of expression (P35) or, in the case of P7.5, because they were run off the gel.

Examination by Western and Northern blotting of the expression of several well-characterized antigens during in vitro growth of B. burgdorferi. Next, we determined whether the expression of some of the well-characterized borrelial antigens was affected by the growth phase of the spirochetes. Western blots were analyzed with monoclonal antibodies to BmpD, GroEL, flagellin, OspC, and P35 (Fig. 2), as well as to OspA, OspB, BmpA (P39), and P72 (not shown). P35 was included as a control for a growth-phase-regulated protein because in a previous study we had demonstrated that its expression was upregulated during the postexponential stages of growth (9).

View larger version (31K): [in this window] [in a new window]   FIG. 2.   Western blots showing the patterns of expression of specific B. burgdorferi proteins during growth of the spirochetes in vitro. The blots were developed with monoclonal or monospecific polyclonal antibodies specific for the following proteins: OspA, OspB, OspC, GroEL, P72, flagellin, BmpA, BmpD, and P35. Lanes 1 to 5 contain samples of the initial culture taken on days 3, 4, 6, and 8 and of the second subculture taken on day 5, respectively.

View larger version (94K): [in this window] [in a new window]   FIG. 3.   Northern blots showing the mRNA profiles for ospAB, ospC, groEL, fla, bmpAB, bmpD, and P35. It should be noted that lanes 2 to 5 contain samples of the initial culture taken on days 4, 6, and 8 of the second subculture taken on day 5, respectively.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

In this paper we have extended our earlier study of the phenomenon of the regulated expression of genes in B. burgdorferi during growth in vitro. We generally searched for proteins that might be differentially expressed during in vitro growth of these spirochetes and also specifically evaluated some of these organisms' best-characterized antigens. We have so far identified 13 proteins whose expression levels are affected in cultured spirochetes during growth from the logarithmic to the stationary phase. These proteins include the previously described P35 and P7.5 (9) and, from this study, P15, P16.5, P22, P23 (OspC), P27, P28, P29, P31, and P43.5, based on an estimation of their apparent molecular weights as determined by SDS-PAGE. With the exception of P28, the expression of the other seven newly identified proteins, including OspC, increased steadily during growth from logarithmic to stationary phase, a pattern of expression similar to that of P35 and P7.5. P28, on the other hand, was expressed at a relatively higher level in the logarithmic phase than in the stationary phase.

Among the previously characterized proteins individually tested with specific monoclonal antibodies, OspC, GroEL, and BmpD displayed differential levels of expression during growth of the spirochetes in culture. The expression of OspA, OspB, BmpA, flagellin, and P72 remained largely constant throughout the culture period. There was an excellent correlation between the mRNA and protein levels for ospAB, bmpA, and fla. With groEL, the marginal increase in the protein level during the later stages of growth did not correspond to its mRNA levels at these time points. Nevertheless, the overall good correlation between the protein and the mRNA levels is strong evidence that the expression of these genes, including groEL, is predominantly controlled at the transcriptional level during in vitro growth of spirochetes in BSK-H medium. Furthermore, the close resemblance of the steady-state protein and RNA profiles of ospAB, bmpAB, groEL, and fla may mean that these genes are regulated (or unregulated) similarly at the transcriptional level. However, this may not always be true of ospAB and fla. ospAB expression has been found to be quite varied at the RNA level among strains and between the different passages of a given strain (10, 13). In addition, a DNA-binding activity that binds to a region upstream of the ospAB operon has been demonstrated recently in stationary-phase spirochetes (14). Based on the 5' flanking DNA sequence, it has been speculated that the expression of the fla gene may be under the control of an alternative sigma factor and hence may be regulated (17).

The expression of OspC increased steadily during the transition from logarithmic to stationary phase. However, the expression of OspC in the log_rev sample was not always as expected. In one experiment, there was very little expression of OspC in the log_rev sample (data not shown), whereas its level of expression in a repeat experiment was consistent with the pattern of a steady increase with the increasing age of the culture (Fig. 1B and C). Several other studies have also demonstrated the variable expression of OspC in cultured spirochetes (12, 15, 25, 28). More importantly, OspC has been shown to be upregulated by a shift in culture temperature from 24 to 37°C (24) and this induction occurs at the transcriptional level (27). It is interesting to speculate, based on the results from this study, that ospC expression may also be regulated transcriptionally and translationally during growth in culture. First, the level of ospC mRNA was relatively higher than those of ospAB, bmpAB, groEL, and fla throughout the culture period, suggesting that ospC transcription may be active during both the logarithmic and postlogarithmic phases, although the persistence of the ospC mRNA may also reflect increased messenger stability. Second, while the ospC mRNA level peaked on day 4 (late logarithmic phase) the protein level reached a maximum only on day 6 (early stationary phase) (Fig. 1B and 2). By contrast, as mentioned above, the levels of expression of other proteins, such as OspA, OspB, flagellin, BmpA, and GroEL, correlated very well with their levels of mRNA. This result hints at the possibility of additional translational control of ospC expression.

The expression of bmpD also appears to be regulated in vitro, but the pattern of RNA expression suggests an entirely different mechanism from that of ospC or P35 RNA expression. The levels of bmpD mRNA and protein peaked at the onset of the stationary phase, unlike those of the genes that are expressed early in logarithmic phase and of the 11 antigens, including P7.5, P35, and ospC, whose expression continued to increase late in growth. The notion of regulated expression of bmpD is supported by the presence of a 13-bp inverted repeat sequence upstream of the gene and overlapping the putative 35 region (21). Furthermore, this overlap of the inverted repeat with the 35 region and the lack of a consensus 35 sequence suggest that bmpD may be positively regulated, being turned on or induced by the binding of a transcriptional activator to the inverted-repeat region.

Interestingly, in all our Western blots developed either with infected-monkey serum or with specific monoclonal antibodies we saw no evidence for a decrease in the levels of antigens during the culture period. In other words, once these antigens were synthesized, their levels were maintained by the bacterial cell throughout the culture period. Included in this group of antigens are some of the well-characterized lipoproteins of B. burgdorferi such as OspA, OspB, OspC, BmpA, and BmpD and the nonlipidated flagellin. Therefore, the diminished expression of proteins such as OspA and OspB that has been observed in tick-borne spirochetes after a tick blood meal (23) may reflect a combination of transcriptional repression of the OspAB operon and selective loss of these proteins during rapid spirochetal multiplication.

The 18 proteins, including OspC, that are recognized by the serum of the B. burgdorferi-infected rhesus monkey correspond to antigens that are present in the spirochete during the early stages of infection. However, OspC expression is known to be turned on in the spirochetes within the tick gut during the course of a blood meal and to continue following infection of the mammalian host. It is tempting then to speculate that the other nine antigens that are upregulated like OspC during growth from logarithmic to stationary phase in vitro may also be similarly turned on within the tick during the course of the blood meal prior to invasion of the mammalian host. Surprisingly, the same serum failed to bind to the abundantly expressed and logarithmic-phase-specific protein P28. This may simply mean that P28 is not immunogenic or that it is absent in spirochetes during the early stages of invasion of the mammalian host.

To summarize, in our earlier paper we suggested that the lipoprotein P35, whose expression is transcriptionally regulated in vitro, might be used as a model to dissect mechanisms of gene expression in B. burgdorferi. We have now identified other proteins, including BmpD, OspC, and P28, all of which also may be used for this purpose.