Assessment of decorin-binding protein A to the infectivity of Borrelia burgdorferi in the murine models of needle and tick infection

Background Decorin-binding proteins (Dbps) A and B of Borrelia burgdorferi, the agent of Lyme disease, are surface-exposed lipoproteins that presumably bind to the extracellular matrix proteoglycan, decorin. B. burgdorferi infects various tissues including the bladder, heart, joints, skin and the central nervous system, and the ability of B. burgdorferi to bind decorin has been hypothesized to be important for this disseminatory pathogenic strategy. Results To determine the role of DbpBA in the infectious lifecycle of B. burgdorferi, we created a DbpBA-deficient mutant of B. burgdorferi strain 297 and compared the infectious phenotype of the mutant to the wild-type strain in the experimental murine model of Lyme borreliosis. The mutant strain exhibited a 4-log decrease in infectivity, relative to the wild-type strain, when needle inoculated into mice. Upon complementation of the DbpBA-mutant strain with DbpA, the wild-type level of infectivity was restored. In addition, we demonstrated that the DbpBA-deficient mutant was able to colonize Ixodes scapularis larval ticks after feeding on infected mice and persist within the ticks during the molt to the nymphal state. Moreover, surprisingly, the DbpBA-mutant strain was capable of being transmitted to naïve mice via tick bite, giving rise to infected mice. Conclusion These results suggest that DbpBA is not required for the natural tick-transmission process to mammals, despite inferences from needle-inoculation experiments implying a requirement for DbpBA during mammalian infection. The combined findings also send a cautionary note regarding how results from needle-inoculation experiments with mice should be interpreted.


Background
The causative agent of Lyme disease, Borrelia burgdorferi, is introduced into a mammalian host via tick bite, whereupon the organisms enter the skin, disseminate hematogenously, and persist in the presence of a strong host immune response [1]. The dissemination and persistence of B. burgdorferi within a mammalian host is thought to be predicated, at least in part, on the organism's ability to bind molecules of the extracellular matrix (ECM) [2][3][4], inasmuch as these interactions have been shown to be important for other bacterial pathogens [5]. Among various ECM components, B. burgdorferi binds to type I collagen [4], fibronectin [6,7], integrins [8,9], the proteoglycan decorin [10], and glycosaminoglycans (GAGs) [11,12]. The B. burgdorferi proteins described as ECM-binding proteins include BBK32 [6,13], Bgp (borrelia-GAG binding protein) [12], P66 (Oms66) [14], decorin-binding protein (Dbp) A, and DbpB [10,15]. These proteins, and perhaps other as yet unidentified molecules, may play a significant role in the infectivity and pathogenesis phenotypes of B. burgdorferi.
Our laboratory has been interested in DbpA as a vaccine candidate for the prevention of Lyme disease and the contribution of DbpA and DbpB to B. burgdorferi pathogenesis and infectivity [16,17]. Since their in vitro characterization as decorin-binding proteins [10,15], the DbpA and DbpB lipoproteins have been implicated as potential contributors to adhesion and colonization of B. burgdorferi within mammalian hosts [15,18]. The genes that encode DbpA and DbpB reside in an operon, dbpBA, on linear plasmid 54 (lp54) and are found within many B. burgdorferi sensu lato isolates [19]. Neither protein is expressed by B. burgdorferi within the tick vector [17], however, expression of dbpA (and presumably dbpB) is upregulated in the mammalian host after ticks deposit spirochetes into the skin [20,21]. DbpA and DbpB expression likely remains high for the duration of mammalian infection, as inferred by the presence of antibodies against both antigens in the serum of mice as late as one year after infection (Hagman, unpublished data). The presence of antibodies against DbpA in the serum of patients with late-stage, disseminated Lyme disease also is well-documented [22], providing added support for DbpA expression by B. burgdorferi during chronic infection.
Although the combined data to date suggest an important role for the decorin-binding proteins of B. burgdorferi during mammalian infectivity, virtually all data inferring the importance of DbpA and DbpB thus far have been indirect. The first direct investigation into the role of the dbpBA operon in the infectious lifecycle of B. burgdorferi was carried out by Shi et al [23]. In this study, mutational analysis of dbpBA in B. burgdorferi strain B31 indicated that neither DbpA, nor DbpB was essential in the murine needle-challenge infection model of borreliosis. However, there was evidence suggesting that these mutants exhibited a modest level of attenuation in immunocompetent mice. Unfortunately, a comprehensive 50% infectious dose (ID 50 ) was not included in this report to further investigate this possible defect, nor was genetic complementation of the mutation performed. This latter point is of particular importance given the genetic plasticity of B. burgdorferi and the possibility that spontaneous loss of an endogenous borrelial plasmid might account for the apparent defect in this mutant. To more directly examine the role of both DbpA and DbpB in the murine/tick model of Lyme disease, a dbpBA-deficient mutant and a dbpA genetic complement of the mutant were generated in the infectious strain 297 of B. burgdorferi. Phenotypic assessment of mouse needle infectivity by ID 50 analysis and Ixodes scapularis tick colonization and tick-transmis-sion capacity by the Bb297 dbpBA-mutant also were performed.

Construction and characterization of a dbpBA-deficient mutant
To assess the roles that DbpA or DbpB play in the infectivity and pathogenesis of B. burgdorferi, we created a dbpBAdeficient mutant by allelic exchange of a PflgB-kan cassette for the majority of the dbpBA operon (Fig. 1A). Infectious Bb297 was chosen for construction of the dbpBA mutant because it is a human isolate [24] and prior infectious mutants of this strain have been readily created [25][26][27][28].
Bb297 was electroporated with the suicide vector, pKHd-bpAko, containing a PflgB-kan cassette flanked by 1.5 kb of DNA on the left side of the dbpBA operon and 890 bp of DNA on the right side of the dbpBA operon (Fig. 1A). This allelic exchange strategy relies on a double crossover, homologous recombination event to replace the dbpBA operon with the PflgB-kan cassette. Several kanamycinresistant transformants were obtained, and the presence (Fig. 1B) and orientation (Fig. 1C) of the PflgB-kan cassette within the dbpBA operon was confirmed by PCR analysis.
During genetic manipulation and in vitro cultivation, B. burgdorferi may spontaneously lose endogenous plasmids that are not required for growth in vitro, but are essential for mammalian infection. At least two plasmid-encoded genes, vlsE [29] and pncA (BBE22) [30], fit this criteria. One kanamycin-resistant transformant, BbKH500, retained both vlsE and pncA (Fig. 2) and was analyzed by immunoblotting to confirm the loss of DbpA and DbpB. As expected, DbpA and DbpB were absent from BbKH500 (Figs. 3A and 3B). BbKH500 exhibited identical doubling times when compared to wild-type Bb297 (data not shown) and PCR-based plasmid profiling revealed that the endogenous plasmid profile of BbKH500 matched that of the parent Bb297 (Fig. 2).

Loss of DbpBA significantly reduces the infectivity of B. burgdorferi in mice when introduced via needle inoculation
To determine the contribution of DbpA and DbpB to the infectivity of B. burgdorferi in mice, C3H/HeJ mice were challenged intradermally via needle inoculation with increasing numbers of BbKH500 (10 4 , 10 5 , 10 6 and 10 7 spirochetes). As a positive control, an additional five mice were inoculated with wild-type Bb297 at a dose of 10 3 spirochetes (ID 50 of approximately 50 spirochetes; [27]). The results are shown in Table 1. Whereas the mice infected with 10 3 of Bb297 showed culture positive ear punch biopsies at two weeks post-infection, needle challenge of naïve mice with 10 4 BbKH500 did not produce an infec-Construction of dbpBA-deletion mutant BbKH500 and Prom-dbpA complementation vector and PCR confirmation Figure 1 Construction of dbpBA-deletion mutant BbKH500 and Prom-dbpA complementation vector and PCR confirmation. (A) Strategy for the replacement of the dbpBA operon with the PflgB-kan and complementation with pKH2000. pKH-dbpBAko was the pGEM-T easy-based suicide plasmid used to transform Bb297 for the homologous recombination of the kanamycin-resistance gene into the dbpBA operon. "A" denotes AscI sites. PflgB-kan denotes the kanamycin-resistance marker expressed from the flgB promoter. The borrelial shuttle vector containing the dbpBA Prom fused to the dbpA ORF (pKH2000) was transformed into BbKH500 to restore DbpA expression. Oligonucleotide primers used for PCR are indicated with short arrows. (B) PCR using primers ko5 and ko4 (shown in panel A). The first two lanes are undigested PCR products from Bb297 and BbKH500, whereas the second two lanes are the corresponding PCR products digested with AscI. (C) Lanes 1, 3 and 5 are PCR products derived from BbKH500 template DNA and lanes 2, 4 and 6 are PCR products derived from Bb297 template DNA. Primer pairs used in PCR are indicated above the lanes.  dbpA PflgB-kan dbpA tion in any mouse (n = 15), even when the infection was allowed to progress for 14 weeks. Seventeen mice challenged with 10 5 and ten mice challenged with 10 6 BbKH500 also did not show signs of infection (negative ear-punch biopsy culture) at four weeks post-challenge (Table 1). However, after 10 weeks, three mice from the group infected with 10 5 BbKH500 and two mice from the group challenged with 10 6 BbKH500 were shown to be infected by positive ear-punch biopsy cultures. Additionally, mice were challenged with either 10 5 or 10 6 BbKH500 and ear-punch biopsies were harvested 14 weeks after infection (Table 1). From the group that received 10 5 spirochetes, two mice were infected (n = 10), and from the group that received 10 6 spirochetes, four mice were infected (n = 10). All of the mice challenged with 10 7 of BbKH500 had positive ear-punch cultures as early as five-weeks post-challenge, and as late as 14 weeks post-infection (n = 5). These spirochetes from ear punchpositive cultures were analyzed by diagnostic PCR as described above, and confirmed to be Bb297 or BbKH500 (data not shown). Aliquots of these cultures also were passed to media containing kanamycin to confirm resistance and sensitivity of BbKH500 and Bb297, respectively. Seroconversion analyses performed on a subset of mice also revealed that none of the mice infected with BbKH500 showed antibody reactivity against DbpA or DbpB (data not shown). Furthermore, only mice that exhibited culture-positive ear punch cultures showed significant serum antibody reactivity with P39 (data not shown). Analysis of all infection results yielded an ID 50 for BbKH500 of 1.2 × 10 6 (p <0.001) compared with ID 50 of approximately 50 bacteria for wild-type Bb297 [27].

Complementation of the dbpBA-deletion mutant with DbpA restores infectivity in needle-challenged mice
The results above suggested that the dbpBA operon is required for full infectivity of B. burgdorferi when mice are infected via needle inoculation. However, during genetic manipulation of B. burgdorferi, it is not uncommon to lose one or more plasmid(s) which potentially contribute to infection, therefore genetic complementation is necessary to definitely ascribe the attenuated phenotype in BbKH500 to the dbpBA lesion. Although BbKH500 carries a mutation in both DbpA and DbpB, we chose to complement with only dbpA because i) DbpA is better characterized that DbpB [31,32] and ii) experimentation suggests that DbpA is the prominent Dbp in B. burgdorferi [15,33]. To restore DbpA expression in BbKH500, a shuttle plas-PCR-based plasmid profiling to compare the plasmid contents of the strains employed in this study Figure 2 PCR-based plasmid profiling to compare the plasmid contents of the strains employed in this study. PCR amplification with primers specific for each of the known endogenous B. burgdorferi plasmids was used to compare the plasmid content of parent Bb297, BbKH500, and BbKH501. Sequence information for the primers utilized is provided in Table 4 mid carrying a copy of dbpA driven by its native promoter was constructed (Fig. 1A). Since dbpA is the second gene in the dbpBA operon, we chose to clone the promoter for this operon upstream of the dbpA gene and then insert the fusion product in the multiple-cloning site of pJD51. BbKH500 cells were transformed with this shuttle plasmid (pKH2000), yielding clones resistance to both kanamycin and streptomycin. These transformants were screened by PCR to verify the presence of the complementing plasmid (Fig. 3C). Clones shown to contain the complementing plasmid were checked for vlsE and pncA (Fig 2.), as described above. PCR-based plasmid profiling, performed on one of the transformants that contained both pncA and vlsE, revealed a profile identical to the parent strain, BbKH500 (Fig. 2). This clone, designated BbKH501, was selected for further analysis and DbpA expression was confirmed by both SDS-PAGE ( Fig. 3A) and immunoblot analysis (Fig. 3B). Proteinase K digestion of BbKH501 and Bb297, under conditions that left FlaB intact, demonstrated surface exposure of DbpA in both strains (Fig. 3D). Three groups of five C3H/HeJ mice were needle inoculated (i.d.) with BbKH501 at 10 2 , 10 3 or 10 4 borreliae per mouse. Ear punch biopsies were harvested at either two weeks (10 3 and 10 4 doses) or three weeks (10 2 dose); all 15 mice became infected, as confirmed by positive ear-punch biopsy cultures. Although a precise determination of the ID 50 of BbKH501 could not be calculated, the fact the 100% of the mice were infected with 10 2 bacteria suggests that the ID 50 is less than 10 2 spirochetes (4-logs lower than BbKH500). Aliquots of these cultures were passed to media containing streptomycin and kanamycin selection to confirm resistance. Diagnostic PCR also confirmed that the spirochetes growing out of the ear punch biopsy were BbKH501 (data not shown). Seroconversion analyses carried out on a subset of mice also revealed that the mice infected with BbKH501 showed antibody reactivity against DbpA, but not DbpB (data not shown). These data confirm that DbpA is necessary for wild-type levels of B. burgdorferi infectivity in mice when introduced via needle inoculation, and that DbpB is dispensable for this aspect of the infectious process.

Neither DbpA nor DbpB is required for acquisition of B. burgdorferi by ticks or infection of mice via tick bite
To determine whether I. scapularis ticks could acquire borreliae lacking dbpA and dbpB, naïve larvae were allowed to feed to repletion on mice infected with BbKH500 or on mice infected with Bb297. Larvae were collected after feeding, allowed to molt, and the unfed nymphs were examined for the presence of B. burgdorferi by IFA; approximately 11 weeks had elapsed since these ticks had fed to repletion as larvae. Microscopic examination of multiple fields revealed approximately one spirochete per field in the BbKH500-infected ticks (23 ticks examined), as opposed to 20-30 spirochetes per field in the Bb297infected ticks (10 ticks examined). All Bb297-infected ticks examined were positive for B. burgdorferi. Of the ticks that had fed on BbKH500-infected mice, 80% were positive for spirochetes. Characterization of BbKH500 and BbKH501 Because the ticks maintained strain BbKH500 through their molt, it was important to assess whether the loss of DbpA and DbpB affected B. burgdorferi transmission to mice via tick bite. To test this, five BbKH500-infected nymphs were placed on each of two naïve mice and ten BbKH500-infected nymphs were placed on each of six naïve mice. An additional two naïve mice received five Bb297-infected ticks each. The ticks were allowed to feed to repletion, and at three weeks post-infestation, earpunch biopsies were harvested from all mice. The earpunch biopsies were cultured in BSK medium and the cultures examined by dark-field microscopy for the presence of B. burgdorferi. These results are shown in Table 2. One of the two mice infested with five ticks harboring BbKH500 was infected as demonstrated by a positive earpunch culture, and four of the six mice infested with ten ticks harboring BbKH500 were infected. Both of the mice on which the Bb297-infected ticks fed had positive earpunch biopsy cultures. These spirochetes were analyzed by diagnostic PCR as described above, and confirmed to be Bb297 or BbKH500 (data not shown). Seroconversion analyses performed on a subset of the BbKH500-infected mice also revealed that none of the mice infected with BbKH500 showed antibody reactivity against DbpA or DbpB (data not shown). Furthermore, only mice that exhibited culture-positive ear punch cultures showed significant antibody reactivity against P39 (data not shown). Taken together, these data indicate that DbpA and DbpB are not essential for the transmission of B. burgdorferi from ticks to mice.

Discussion
The importance of ECM-binding proteins to the pathogenic strategy of B. burgdorferi is not currently known. It is presumed that borreliae bind mammalian host cells during infection because binding to various ECM molecules and to tissue culture cells has been demonstrated in vitro [4,34,35]. Of the known ECM-binding proteins of B. burgdorferi, Bgp [36] and BBK32 [37] have been deleted from infectious strains N40 and B31, respectively, and shown to be dispensable for infectivity in the mouse model of Lyme disease. A deletion of integrin-binding protein, P66, in a noninfectious strain of B. burgdorferi, HB19 [38], caused the spirochetes to lose their ability to bind b 3chain integrins [14] but the phenotype of a P66 mutant in an infectious strain of B. burgdorferi is currently unknown. The remaining two known ECM-binding proteins, DbpA and DbpB, were the focus of this work due to multiple lines of evidence suggesting a role for one or both proteins during infection of mammalian hosts [15,18,31]. To directly address the contribution of both DbpA and DbpB to B. burgdorferi infectivity, we created BbKH500, a dbpBAdeletion mutant of the human isolate Bb297, and used this strain to challenge mice with increasing numbers of spirochetes. We observed a 4-log reduction in infectivity of the mutant strain when the spirochetes were needle inoculated into mice. Complementation of the dbpBAmutant with DbpA alone in the dbpBA-mutant restored infectivity of B. burgdorferi to wild-type levels.
Prior to this work, Fischer et al. [18] had restored DbpA and DbpB expression (via shuttle plasmid) to B. burgdorferi strain B314, a non-adherent, noninfectious derivative of strain B31 that has lost several plasmids, including lp54 [39]. Although expression of the Dbp molecules restored binding of the spirochetes to purified decorin, dermatan sulphate, and human epithelial cells [18], experiments to test the infectivity of B314 expressing DbpA and DbpB in the murine model of Lyme disease were implausible due to the loss of plasmids that are required for mammalian infection [39]. A recent study by Shi et al. [23] assessed the infectivity of a DbpA,B mutant of B. burgdorferi strain 5A18NP1, a B31 derivative that lacks the BBE02 gene [40]. Based on needle-challenge results obtained from a single inoculation dose of 10 5 bacteria (80% infection), Shi et al. [23] ascertained that the dbpBA locus was not required for infectivity. However, in the present study, we observed a 4-log increase in the ID 50 of BbKH500 (>10 6 bacteria) and a 19% infection rate with a dose of 10 5 BbHK500 when mice were infected by needle inoculation. The reason for the difference in infectivity levels of the dbpBA-deficient mutants reported by Shi et al. and BbKH500 created in our laboratory is unknown at this time. One possible explanation for this disparity could be due to our use of Bb297 as the parental strain for the present mutational analysis, whereas Shi et al. [23] utilized a clonal derivative of B31 that is a BBE02-mutant. This variability also could be explained by spirochete enumeration differences prior to needle challenge. As mentioned in the Methods, significant attention was given to the enumeration of BbKH500 prior to needle inoculation of mice.
It should be noted that during the course of the study described herein, a subsequent report on the role of DbpBA was published by Shi et al. [41], which, unlike their previous study, included genetic complementation experiments. Our current findings agree with those of this most recent report in that Shi et al. also observed a 4-log a Varying number of ticks infected with BbKH500 were allowed to feed to repletion on naïve mice. b Infection rates were assessed at three weeks post-infestation by culturing BbKH500 from ear-punch biopsies increase in ID 50 values with their dbpBA mutant [41]. In addition, Shi et al. observed a defect in the ability of the dbpBA mutant to colonize heart, joint and skin tissues, suggesting an overall deficiency in dissemination [41]. The observation that mice infected with BbKH500 showed a delay in infectivity (10-to 14-weeks post-infection) also suggests that BbKH500 might be attenuated with respect to its capacity to disseminate through the host. However, whereas we were able to compensate fully for the loss infectivity in the dbpBA mutant by complementing with dbpA alone (as assessed by ear punch biopsy culture), Shi et al. reported that both dbpA and dbpB were required to restore the infectivity of their dbpBA mutant [41]. Although the precise reason for this disparity is unknown at this time, there are several differences between the experimental approaches of these studies that might account for these differing results. First, it is possible that some intrinsic differences(s) in the bacterial strains utilized might account for the variation observed between these two studies. Specifically, as noted above, strain 297 was the parental strain used in the current study, and Shi et al. utilized a highly-transformable clone (5A13) of strain B31 that lacks both lp56 and the virulence-associated plasmid lp25 as the background for their mutagenesis experiments [30,41]. Second, the disparity in our results might be due to the use of different strains of mice in these two studies; C3H/HeJ mice were used in our study, whereas Shi et al used BALB/c [41]. This may be relevant because numerous studies have reported that experimental infection of C3H strains of mice with B. burgdorferi results in significantly different disease pathologies [42,43], higher spirochetal loads in multiple tissues [44], and different cytokine responses [45,46] by comparison to similarly infected BALB/c mice. Although, it is difficult to precisely predict how the reported differences in pathogenesis observed in these two distinct murine backgrounds might be impacted by DbpA and/or DbpB during infection, it is reasonable to suspect that these differences might account for some of the disparity in the results of complementation experiments obtained in the aforementioned studies.
Even though DbpA was required for full infectivity by Bb297 in mice when introduced by needle inoculation (intradermally), neither DbpA, nor DbpB, was required for infection of larval I. scapularis ticks or for transmission of BbKH500 from infected ticks to naïve mice. That DbpA was not required for uptake of the spirochetes by larval ticks was not surprising due to our earlier work demonstrating that DbpA is not expressed by B. burgdorferi harbored within the tick vector [17]. However, based on the data we obtained from needle inoculation of mice with BbKH500, it was surprising that as few as five nymphal ticks could transmit BbKH500. This is especially surprising because semi-quantitative analysis of the total number of spirochetes observed in the midguts of unfed ticks infected with BbKH500 showed a lower spirochete density (approx. 1/field) by comparison to number of bacteria present in the dissected midguts of the Bb297-infected ticks (20-30/field). While these data might suggest a possible defect in either the efficiency of acquisition of the dbpBA mutant or the capacity of this mutant to persist in the tick midgut, there were still sufficient numbers of the mutant spirochetes in these ticks to infect naïve mice.
Since it is not known precisely how many spirochetes are transmitted by ticks during the feeding process [47], it was impossible to directly compare the number of B. burgdorferi transmitted by tick bite to our needle-challenge experiments. Spirochete numbers within salivary glands during tick feeding have been estimated between 20 [47] and 61 spirochetes [48] per salivary gland pair. These data suggest that an individual tick deposits far fewer than 10 4 spirochetes during the feeding process, and our results have shown that by the needle-inoculation route, 10 4 BbKH500 are not infectious.
In addition to the data presented by Fischer et al. [18], additional reports have attempted to address the role of DbpA, DbpB, and the ECM proteoglycan, decorin, with respect to the infectivity and pathogenesis of B. burgdorferi. Brown et al. described B. burgdorferi infection of decorin-deficient mice (Dcn -/-) [49]. Brown et al. [49] found that, by comparison to Dcn +/+ mice, Dcn -/mice challenged with a higher dose of B. burgdorferi (10 4 ) had i) fewer infected joints, ii) a reduction in the severity of arthritis, but iii) no significant defect in colonization of the other tissues. Whereas 10 4 wild-type B. burgdorferi were infectious for Dcn -/mice in the studies by Brown et al. [49], the Bb297 dbpBA-mutant was unable to infect mice at this dose. When considered together, these data support the hypothesis that DbpA, in addition to binding decorin, may have an additional ligand(s) or has another function critical for infectivity. At the present time, the only other known ligand recognized by DbpA is the GAG dermatan sulfate [18], but the contribution of this interaction to the pathogenesis of B. burgdorferi remains to be elucidated.

Conclusion
Despite the disparities between the results of the complementation experiments described in the current study and those obtained by Shi et al. [41], the overall results of our needle-inoculation experiments are in agreement with the most recent conclusion of Shi et al. that mutation of dbpBA results in significant attenuation of B. burgdorferi in the murine model of Lyme borreliosis. However, the observation that the dbpBA mutant showed a significant reduction in infectivity when the mice were needle inoculated is overshadowed by the finding that this same mutant was capable of infecting mice via tick challenge.
The fact that DbpA and DbpB are dispensable for infection via the tick-mediated route of infection suggests that B. burgdorferi transmitted via tick bite are in some way phenotypically different than their in vitro-cultivated counterparts, and/or that tick-derived salivary components, such as Salp15, may assist B. burgdorferi during the early infection process [50]. Taken together, these results emphasize the importance of characterizing the impact of a given gene in the infectious lifecycle of B. burgdorferi using the natural tick vector, as opposed to using only the artificial needle-challenge model.

Generation of DbpBA deletion mutant in Bb297
The dbpBA-deficient Bb297 strain was created by allelic exchange of the dbpBA operon with a kanamycin-resistance cassette, PflgB-kan [53], derived from pBSV2 [54]. The mutagenesis construct, pKHdbpBAko, was created by generating two PCR products that constituted the left and right flanking regions of the dbpBA operon which then were joined via an AscI restriction site, thus deleting a significant portion of the dbpBA operon. Takara EX Taq polymerase (Takara Bio Inc., Shiga, Japan) and oligonucleotide primers ko1 and ko2 (Table 3) were used to amplify the left arm and ko3 and ko4 (Table 3) were used to amplify the right arm; primers ko2 and ko3 were modified such that AscI restriction sites would be introduced into the "middle termini" of the two PCR fragments. The resulting PCR products then were digested with AscI and ligated together. The linear ligated product was used as the template in a second PCR amplification containing the primers ko1 and ko4. The resulting PCR product, representing the joined flanking regions, was cloned into pGEM-T Easy vector (Promega Corp., Madison, WI) to generate pGEMT-dbpBA-AscI. The PflgB-kan cassette was from pJD55, a derivative of pJD44 in which the original aph [3']-IIIa was replaced with the PflgB-kan cassette of pBSV2 [27,54]. In pJD55, the PflgB-kan cassette is flanked by AscI sites which facilitated the cloning of the marker into the unique AscI site within pGEMT-dbpBA-AscI to create pKHdbpBAko. Primers kan5' and kan3' were used with primers ko1 and ko4 to determine the orientation of the PflgB-kan cassette with respect to the dbpBA operon.
Bb297 were made electrocompetent and transformed with pKHdbpBAko as described by Yang et al [28]. After electroporation with pKHdbpBAko, spirochetes were allowed to recover overnight at 35°C without antibiotic selection in 20 ml BSK-H. The cells were diluted in BSK-H medium containing the appropriate concentration of antibiotic(s) and aliquoted in 96-well tissue culture plates (Corning, Lowell, MA). Transformants were recovered 7-21 days after plating from wells in which a red to orange color change of the medium was observed. The presence of viable spirochetes was confirmed visually by dark-field microscopy and clones expanded into BSK medium supplemented with kanamycin.
Transformants were verified as dbpBA-deficient mutants by diagnostic PCR using primers ko5 and ko4 (Table 3) followed by analytical restriction enzyme digestion of the PCR product with AscI. DNA for PCR analysis was extracted from borreliae harvested from the expansion cultures. The PCR product generated from amplification of Bb297 DNA (wild-type) is of similar size to that generated by amplification of DNA from a dbpBA::PflgB-kan mutant, but the former lacks AscI sites. Therefore, to verify that the dbpBA operon was replaced by the PflgB-kan cassette, PCR products from both Bb297 and mutant-derived were digested with AscI prior to agarose-gel electrophoresis. The presence of the vlsE and pncA (BBE22) genes in kanamycin-resistant transformants was confirmed by PCR using vlsE-5' and vlsE-3' for vlsE amplification or BBE22-5' and BBE22-3' for BBE22 amplification; refer to Table 3 for sequence information. FlaB5' and FlaB3' primers, which amplified the flaB gene of B. burgdorferi, were used as a control for DNA integrity. A single dbpBA-deficient clone, BbKH500, that retained vlsE and pncA, was chosen for additional PCR-based analyses to compare the endogenous plasmid content of this clone to Bb297. The sequences of the primers utilized for plasmid profiling are provided in Table 4. Nine of the primer pairs have been previously described by Eggers et al. [55]. The remaining primers utilized are unique from those cited by Eggers et al. and were designed primarily based on sequence data from strain 297.

Construction of shuttle plasmids for genetic complementation of BbKH500 with DbpA
Complementation of DbpA in BbKH500 was achieved by transforming electrocompetent BbKH500 with the B. burgdorferi-shuttle plasmid, pKH2000. Because dbpB precedes dbpA in the native operon, it first was necessary to clone the promoter for the dbpBA operon directly in front of the dbpA gene from Bb297, thereby removing the dbpB gene. The dbpBA promoter region (Prom) was amplified by PCR using primers comp1 and comp2 and the dbpA open reading frame (ORF) was amplified using primer comp3 and comp4. To facilitate fusion of the Prom and dbpA open reading frame fragment, an NdeI restriction site was introduced into the 3' end of oligonucleotide comp2 and the 5' end of primer comp3. Following PCR amplification, the resulting PCR fragments were digested with NdeI and ligated together. A second PCR amplification was performed using the ligation product as the template and the oligonucleotides comp1 and comp4 for primers. The resultant PCR product was digested with BglII (a unique BglII site is located in the PCR product, 11 bp downstream of the 3' end of the comp1 primer) and SphI (5' end of comp4 primer) then cloned into the BglII and SphI sites of pJD51, a derivative of pJD44 [27], that contains the aadA gene encoding streptomycin resistance in B. burgdorferi [56], to create pKH2000. Electroporation of BbKH500 was performed as described above and transformants were selected in the presence of kanamycin and streptomycin. Antibiotic-resistant clones first were checked for the presence of the Prom-dbpA construct by PCR using primers ko5 and comp5 (the latter anneals near the middle of the dbpA coding strand); Bb297 = 1.2 kb product; BbKH500 = no product; Prom-dbpA = 535 bp product. Next, DbpA expression in clones identified by PCR confirmation was assessed by immunoblot analysis as described below. Clones that expressed DbpA from the Prom-dbpA construct were assessed for the presence of the vlsE and BBE22 genes by PCR amplification using the primers described above. One clone, BbKH501, was chosen for further characterization. PCR-based plasmid profiling was performed on BbKH501, as described above, to compare the plasmid content of this clone to BbKH500 and Bb297.

Proteinase K digestion of B. burgdorferi
Intact, motile borreliae were exposed to 200 μg of proteinase K (40 mg/ml; Fisher Scientific, Pittsburgh, PA) or were sham treated for 40 min at room temperature. To stop the reaction, 10 μl of phenylmethylsulfonyl fluoride (50 mg/ml in isopropanol; Sigma) was added to each sample and the bacteria prepared for SDS-polyacrylamide gel electrophoresis (PAGE) and immunoblot analysis as described below. Nitrocellulose membranes were probed with either 6B3-DbpA, a mouse monoclonal antibody that specifically recognizes DbpA, chicken anti-B. burgdorferi FlaB IgY, or rat anti-DbpB anti-sera. The monoclonal antibody, 6B3-DbpA, was produced in collaboration with the Antibody Production Core facility at UT Southwestern and the pol-yclonal rat anti-DbpB antisera was described previously [16]. The chicken anti-FlaB antibody was produced in collaboration with Lampire Biological Laboratories (Pipersville, PA). Secondary antibodies were horseradish peroxidase-conjugated goat anti-mouse immunoglobulin G (IgG), donkey anti-chicken IgY, or goat anti-rat IgG (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA) diluted 1:10,000-1:30,000 for chemiluminescent detection. Immunoblots were developed using Immobilon Chemiluminescent Western HRP Substrate (Millipore, Billerica, MA) and exposed to X-ray film (Phenix Research Products, Hayward, CA).

Infection of mice by needle inoculation
Prior to use of the cultures in needle-inoculation experiments, the bacterial cell density in each culture was accurately determined by counting spirochetes in no fewer than 60 microscopic fields (400× magnification) using dark-field microscopy. Cultures exhibiting cell aggregation were not used for infections as the presence of clumps prevented accurate enumeration of spirochete density. Cultures with spirochetes exhibiting reduced motility also were not used for mouse infections. Three-to five-weekold female C3H/HeJ mice (The Jackson Laboratory, Bar Harbor, ME) were used for all studies. Mice were infected via intradermal injection with serial dilutions of B. burgdorferi in BSK medium as previously described [16]. At the appropriate time intervals, ear-punch biopsies were harvested using a 2 mm ear punch, placed in BSK-H medium supplemented with BAM, and cultures were examined by dark-field microscopy for the presence of spirochetes. Aliquots of each culture were passed to media containing antibiotics to confirm the antibiotic-resistance phenotype of the bacteria that were recovered from cultures of the ear punch biopsies. The ID 50 was calculated using the method described by Reed and Muench [57]. UT Southwestern is accredited by the International Association for Assessment and Accreditation of Laboratory Animals Care (AAALAC) and all animal protocols were approved by the Institutional Animal Care and Use Committee (IACUC) at UT Southwestern.

Colonization of Ixodes scapularis larvae with B. burgdorferi
Female C3H/HeJ mice were needle inoculated as described above with either Bb297 (10 4 spirochetes) or BbKH500 (100 μl of a post-exponential growth phase culture; 10 6 -10 7 spirochetes). Infection of mice was confirmed by ear-punch biopsy culture; 4-and 8-weeks postinoculation for Bb297 and BbKH500, respectively. Naïve, pathogen-free I. scapularis larvae, obtained from the Department of Entomology and Plant Pathology at Oklahoma State University (Stillwater, OK), were allowed to feed to repletion on the infected mice individually housed in cages with raised wire-bottoms above water to facilitate recovery of the ticks. Fed larvae were collected and washed sequentially with 70% ethanol, deionized water (dH 2 O), 1× Fungizone (Gemini Bio-Products, West Sacramento, CA), and dH 2 O before placing them in 100% cotton fabric-lined Petri dish molting chambers for storage until they molted to the nymphal stage. The molting chambers containing the ticks were housed in a humidified chamber (97-98% humidity) containing saturated potassium sulfate solution at 20°C with a 16 h light, 8 h dark cycle. Unfed nymphs were collected and stored in autoclaved glass vials containing approximately 1 cm of sand. The vials were closed with vented lids and the ticks housed as described above.

Direct immunofluorescence assay (IFA) on B. burgdorferiinfected ticks
Prior to placement of unfed (flat) nymphs on naïve mice, five nymphs from each of two individual mice were dissected on silylated slides (CEL Associates, Inc., Pearland, TX) in 50 μl phosphate-buffered saline containing 10 mM MgCl 2 . At the time testing was performed, approximately 11 weeks had elapsed since these ticks had fed to repletion as larvae. Midgut tissues were extracted from the ticks, allowed to dry on the slides, blocked with Tris-buffered saline with 0.1% Tween-20, then probed with a FITC-conjugated rabbit anti-B. burgdorferi antibody (Fitzgerald Industries International, Inc., Concord, MA) as described previously [27].

Transmission of B. burgdorferi from infected nymphal ticks to naïve mice
To assess the transmissibility of strain Bb297 and BbKH500 from flat nymphs to naïve mice, either five or ten ticks were allowed to feed to repletion on three-to five-week-old female C3H/HeJ mice. At three-weeks postinfestation, ear-punch biopsies were harvested from each mouse and cultured in BSK-H medium without antibiotics to determine infection status. Borreliae from the earpunch biopsy cultures were transferred to BSK-H medium supplemented with kanamycin and streptomycin to confirm their antibiotic-resistance phenotypes.

Statistical analysis
Statistical analysis was performed with assistance from the UT Southwestern Clinical Sciences Department. Both chisquare and Fisher's exact test were applied to the mouse infection data in pair-wise comparisons between experimental groups.