The Infectivity Gene bbk13 Is Important for Multiple Phases of the Borrelia burgdorferi Enzootic Cycle

ABSTRACT Lyme disease is a multistage inflammatory disease caused by the spirochete Borrelia burgdorferi transmitted through the bite of an infected Ixodes scapularis tick. We previously discovered a B. burgdorferi infectivity gene, bbk13, that facilitates mammalian infection by promoting spirochete population expansion in the skin inoculation site. Initial characterization of bbk13 was carried out using an intradermal needle inoculation model of mouse infection, which does not capture the complex interplay of the pathogen-vector-host triad of natural transmission. Here, we aimed to understand the role of bbk13 in the enzootic cycle of B. burgdorferi. B. burgdorferi spirochetes lacking bbk13 were unable to be acquired by naive larvae fed on needle-inoculated mice. Using a capsule feeding approach to restrict tick feeding activity to a defined skin site, we determined that delivery by tick bite alleviated the population expansion defect in the skin observed after needle inoculation of Δbbk13 B. burgdorferi. Despite overcoming the early barrier in the skin, Δbbk13 B. burgdorferi remained attenuated for distal tissue colonization after tick transmission. Disseminated infection by Δbbk13 B. burgdorferi was improved in needle-inoculated immunocompromised mice. Together, we established that bbk13 is crucial to the maintenance of B. burgdorferi in the enzootic cycle and that bbk13 is necessary beyond early infection in the skin, likely contributing to host immune evasion. Moreover, our data highlight the critical interplay between the pathogen, vector, and host as well as the distinct molecular genetic requirements for B. burgdorferi to survive at the pathogen-vector-host interface and achieve productive disseminated infection.

nymphal ticks (7). Nymphs then molt into sexually mature adults, which mate and lay eggs to give rise to the next generation of uninfected Ixodes ticks.
The molecular mechanisms that promote B. burgdorferi infection of its mammalian host are not completely understood, and this has slowed the development of new strategies to curb the rising prevalence of Lyme disease. Genetic manipulation of B. burgdorferi coupled with a mouse model of infection has been central to the discovery of novel B. burgdorferi infectivity genes. Needle inoculation of mouse models of infection allows the precise control of both the inoculum dose and route of inoculation, enabling detailed characterization of infection kinetics. However, needle inoculation is an artificial route of infection that excludes the contribution of the tick vector to B. burgdorferi transmission-both its influence on B. burgdorferi itself and its influence on the host microenvironment at the site of feeding. B. burgdorferi adapts to the conditions present in the unfed tick and undergoes significant transcriptional changes during tick feeding, potentiating its infectivity (8,9). Moreover, the feeding activity of the ticks themselves, particularly the influence of the tick saliva on the skin feeding site, affects B. burgdorferi infection (10)(11)(12). The tick feeding site is unique in that it is the combined interaction of the pathogen, the vector, and the mammalian host, which is likely critical to the adaptation and mechanisms of pathogenesis of B. burgdorferi as well as other tick-borne pathogens.
We recently characterized the role of the novel B. burgdorferi infectivity gene bbk13 in the early phase of B. burgdorferi mammalian infection. Using an intradermal needle inoculation model of infection, we showed that bbk13 is important for spirochete population expansion in the skin, which then drives the downstream steps of disseminated infection, culminating in B. burgdorferi colonization of distal tissues (13). Consequently, the loss of bbk13 leads to reduced colonization of distal tissue sites such as the skin, heart, and joints (13). Furthermore, we have established that BBK13 forms large oligomers in the spirochete membrane and is highly immunogenic (13,14). In this work, we used Ixodes ticks and mice to study the role of bbk13 in the natural enzootic cycle of B. burgdorferi. We demonstrate that bbk13 is similarly required for productive disseminated infection when delivered by tick bite transmission. Interestingly, B. burgdorferi inoculation by tick feeding bypassed the early requirement for bbk13 to promote B. burgdorferi population expansion in the skin. Despite this, bbk13 remained critical for efficient colonization of distal tissues. In addition, we show that distal tissue colonization by Dbbk13 B. burgdorferi is restored in an immunocompromised mouse host. This work points toward a role of bbk13 in host immune evasion and uncovers a more complex model of bbk13 function in B. burgdorferi mammalian infection, emphasizing the impact of the natural tick vector on the establishment of B. burgdorferi infection as well as the distinct molecular genetic requirements for B. burgdorferi survival during early localized infection at the pathogen-vector-host interface and the establishment of productive disseminated infection of distal host tissues.

RESULTS
Naive larvae fail to acquire Dbbk13 B. burgdorferi from infected mice. B. burgdorferi's enzootic cycle depends not only on successful tick transmission but also on its acquisition from the animal host by naive larvae. Needle inoculation of Dbbk13 B. burgdorferi into mice leads to reduced, but not absent, colonization of distal tissues, including the skin (13). This allowed us to ask whether Dbbk13 B. burgdorferi can be acquired from infected mice by feeding Ixodes larvae. We needle inoculated cohorts of C3H/HeN mice with 10 4 wild-type (WT), Dbbk13, or Dbbk13/bbk13 1 B. burgdorferi spirochetes intradermally to generate infected mice for naive tick feeding. Consistent with what we have shown previously (13), Dbbk13 B. burgdorferi spirochetes were found at reduced levels during blood dissemination (Fig. 1A) and in distal tissues (Fig. 1B). More importantly, Dbbk13 B. burgdorferi was still present in the skin, as indicated by reisolation from the ear (Fig. 1B, top row). B. burgdorferi colonization of tissues was assessed via semiquantitative tissue reisolation assay (see Materials and Methods). Briefly, reisolation cultures were inspected daily for the presence of spirochetes by dark-field microscopy. A numerical score was assigned to each culture corresponding to the observed density of spirochetes (0 for no spirochetes and 1 to 3 depending on the number of spirochetes).
At 3 weeks postinoculation, naive Ixodes larvae were fed to repletion on the three groups of B. burgdorferi-infected mice. Individual fed larvae were assessed for B. burgdorferi acquisition via homogenization and plating in solid Barbour-Stoenner-Kelly (BSK)-agarose medium to quantify the number of B. burgdorferi spirochetes per tick. The rate of acquisition of wild-type B. burgdorferi after larval tick feeding was 96% (48/ 50 larvae), with an average of ;5 Â 10 3 B. burgdorferi spirochetes per tick (Fig. 1C). In stark contrast, only 4% of the Ixodes larvae (4/100 larvae) that fed on Dbbk13 B. burgdorferi-infected mice acquired spirochetes and with B. burgdorferi loads at dramatically reduced levels (Fig. 1C). The rate of acquisition of Dbbk13/bbk13 1 B. burgdorferi from infected mice was 86% (43/50 larvae), with the average number of B. burgdorferi spirochetes per tick being comparable to that of the wild-type B. burgdorferi group (Fig. 1C). These data indicated that there is a significant deficiency in the ability of Ixodes larvae to acquire spirochetes from Dbbk13 B. burgdorferi-infected mice.
Increasing the inoculum dose does not rescue low Dbbk13 B. burgdorferi numbers in distal tissues. For Ixodes ticks to acquire B. burgdorferi through feeding, the spirochetes must be present in sufficient numbers at the skin feeding site. Although both wild-type and Dbbk13 B. burgdorferi spirochetes were present at a skin site (Fig. 1B, FIG 1 Naive larval ticks fail to acquire Dbbk13 B. burgdorferi during feeding on infected mice. Groups of six C3H/HeN mice were needle inoculated intradermally with 10 4 WT, Dbbk13, or Dbbk13/bbk13 1 B. burgdorferi spirochetes. (A) At 6 days postinoculation, disseminating B. burgdorferi spirochetes were enumerated by plating blood in solid BSK-agarose medium and enumerating CFU. Each dot represents an individual mouse. The mean value is indicated by a black horizontal line. The Kruskal-Wallis test with Dunn's multiple comparisons was used to determine statistical significance (**, P # 0.01). (B) At 4 weeks postinoculation, the indicated tissues were assessed for B. burgdorferi by a reisolation assay. Semiquantitative scoring of reisolation cultures was performed ("0" indicates no spirochetes, and "1" to "3" indicate increasing spirochete densities). (C) Naive tick larvae fed to repletion on C3H/HeN mice 3 weeks after needle inoculation with 10 4 WT, Dbbk13, or Dbbk13/bbk13 1 B. burgdorferi spirochetes. Individual fed larvae were homogenized and plated in solid BSK-agarose medium, and B. burgdorferi loads were determined by enumerating CFU. Each data point represents an individual tick. The tick infection rate for each B. burgdorferi clone is indicated above the corresponding scatterplot. The mean is indicated by a horizontal black line. The Kruskal-Wallis test with Dunn's multiple comparisons was used to determine statistical significance (****, P , 0.0001; n.s., not significant).
bbk13 Is Critical to the B. burgdorferi Enzootic Cycle Infection and Immunity ear), Dbbk13 B. burgdorferi spirochetes were found at reduced numbers compared to wild-type B. burgdorferi during late infection (13). Reduced Dbbk13 B. burgdorferi numbers in the skin likely negatively influence acquisition via tick feeding, raising the question of whether low Dbbk13 B. burgdorferi acquisition was due to reduced spirochete numbers and/or a molecular function of BBK13. To address this, we attempted to increase the Dbbk13 B. burgdorferi load in the skin by increasing the inoculum dose. We needle inoculated 10 5 , 10 6 , or 10 7 Dbbk13 B. burgdorferi spirochetes intradermally into C3H/HeN mice. At 3 weeks postinoculation, distal tissues were collected and assessed for B. burgdorferi colonization via the semiquantitative tissue reisolation assay (13). After 5 days of incubation, nearly all of the reisolation cultures from mice inoculated with 10 4 wild-type B. burgdorferi spirochetes demonstrated a score of "3" (23/24 cultures) ( Fig. 2A). At the same time point, reisolation cultures from mice inoculated with increasing doses of Dbbk13 B. burgdorferi were mostly negative for spirochetes (10 5 dose, 9/24 positive; 10 6 dose, 6/24 positive; 10 7 dose, 8/24 positive), and the positive cultures demonstrated varying spirochete densities (scores ranging from 1 to 3) ( Fig. 2A). Thus, no observable increase in distal tissue colonization was detected upon increasing the inoculum dose of Dbbk13 B. burgdorferi. These trends were maintained at the 2-week endpoint of the reisolation assay ( Fig. 2A). Direct measurement of B. burgdorferi loads in the distal tissues by quantitative PCR (qPCR) supported the findings in the tissue reisolation assay. Despite the increased inoculum doses, the mean Dbbk13 B. burgdorferi loads in the ears and joints remained consistently lower than those of wild-type B. burgdorferi at a lower inoculum ( Fig. 2B and C). Moreover, Dbbk13 B. burgdorferi loads were above the level of detection in only 2 or 3 tissues out of 6 mice per inoculum group, further emphasizing the reduced efficiency of distal tissue colonization of Dbbk13 B. burgdorferi. bbk13 is dispensable for B. burgdorferi maintenance in Ixodes ticks. To further understand the role of bbk13 in the enzootic cycle of B. burgdorferi, we investigated the survival of Dbbk13 B. burgdorferi in ticks after the blood meal and through the molt. Given the inefficiency of Dbbk13 B. burgdorferi to be acquired by feeding ticks, we artificially infected unfed Ixodes larvae with wild-type, Dbbk13, or Dbbk13/bbk13 1 B. burgdorferi by immersion (15). After recovery from immersion, positive infection was confirmed and quantified by plating homogenates from groups of 10 larvae per infecting B. burgdorferi clone. No significant difference was detected between the numbers of B. burgdorferi spirochetes per 10 larvae across the three B. burgdorferi clones (Fig.  3A). The infected larvae were fed to repletion on naive C3H/HeN mice. Approximately 7 days after feeding, 20 fed larvae per infecting B. burgdorferi clone were individually homogenized and plated in solid BSK-agarose medium to measure spirochete loads. Robust multilog expansion was detected for all B. burgdorferi clones (Fig. 3B, left), with no statistical difference between the average numbers of Dbbk13 or Dbbk13/bbk13 1 B. burgdorferi spirochetes and those of wild-type B. burgdorferi (Fig. 3B, left). A portion of the fed larvae were allowed to molt to unfed nymphs, from which 20 nymphs per B. burgdorferi clone were then individually assayed for B. burgdorferi loads by homogenization and plating for CFU. B. burgdorferi loads were comparable across all three groups and displayed the expected decline in number following the molt (Fig. 3B, middle) (16). The nymphs were then allowed to feed to repletion on mice, and the number of B. burgdorferi spirochetes per fed nymph was determined. No bbk13-dependent decrease in the B. burgdorferi load per fed nymph was detected; rather, the number of Dbbk13 B. burgdorferi spirochetes per fed nymph was found to be significantly increased compared to that of wild-type B. burgdorferi (Fig. 3B, right). Although not statistically significant, the average number of Dbbk13/bbk13 1 B. burgdorferi spirochetes in all tick stages except for fed nymphs tended to be lower than those of wildtype and Dbbk13 B. burgdorferi (Fig. 3), perhaps due to dysregulated expression of bbk13 from the shuttle vector (13). In all, the loss of bbk13 did not result in a defect in the replication and survival of B. burgdorferi in ticks.  bbk13 is critical for B. burgdorferi infection of mice by Ixodes nymph transmission. Our previous study used needle inoculation to demonstrate the importance of bbk13 during mouse infection (13). In nature, B. burgdorferi is transmitted via the feeding activity of Ixodes ticks. Moreover, tick feeding influences both B. burgdorferi and the host to promote infection (6,17). Unfed nymphs carrying equivalent loads of wild-type, Dbbk13, or Dbbk13/bbk13 1 B. burgdorferi (Fig. 3B, middle) were used to test the requirement for bbk13 during natural B. burgdorferi infection of mice by tick feeding. Three weeks after free feeding of infected nymphs on mice, distal tissues were collected and tested for B. burgdorferi colonization via the semiquantitative tissue reisolation assay. After 5 days of incubation, a majority (17/20) of the cultures containing tissues from mice fed on by wild-type B. burgdorferi-infected nymphs had a reisolation score of 3 (Fig. 4A, 5-day culture). In contrast, nearly all (19/20) reisolation cultures from mice fed on by Dbbk13 B. burgdorferi-infected nymphs had no spirochetes upon microscopic inspection and were given a score of 0 (Fig. 4A, 5-day culture). Moreover, at the 2-week endpoint, only 4 out of 20 Dbbk13 B. burgdorferi reisolation cultures were positive for spirochetes, with three of the five mice having no detectable spirochetes in any distal tissue examined (Fig. 4A, endpoint). In contrast, all reisolation cultures from the wild-type and Dbbk13/bbk13 1 B. burgdorferi-infected mouse tissues were positive for spirochetes at the 2-week endpoint (Fig. 4A, endpoint). Consistent with the observed deficit in tissue reisolation, Dbbk13 B. burgdorferi was below the level of detection by qPCR in the hearts (0/5 mice) and joints (0/5 mice) (Fig. 4B) of mice after tick feeding. Spirochete numbers were measurable in two out of five ear tissues from Dbbk13 B. burgdorferi-infected mice but were reduced compared to those in wild-type and Dbbk13/bbk13 1 B. burgdorferi-infected mice (Fig. 4B). Moreover, in agreement with reduced disseminating spirochetes and reduced overall loads in distal tissues, mice fed on by Dbbk13 B. burgdorferi-infected ticks exhibited a serological response with a reduced signal intensity yet a banding pattern similar to that of wildtype B. burgdorferior Dbbk13/bbk13 1 B. burgdorferi-infected mice (Fig. 4C), as demonstrated previously for mice infected with spirochetes lacking bbk13 (13). The attenuated infection phenotype of Dbbk13 B. burgdorferi by nymph transmission was recapitulated by free-feeding artificially infected larvae (Fig. 4D), together supporting the importance of bbk13 for B. burgdorferi infection by tick bite transmission.
B. burgdorferi population expansion occurs at the tick feeding site. We previously showed that bbk13 is important early in B. burgdorferi mammalian infection to promote spirochete expansion in the skin within days after intradermal needle inoculation (13). The deficiency of Dbbk13 B. burgdorferi in distal tissue colonization was observed by both needle inoculation ( Fig. 1A and B) (13) and natural tick transmission (Fig. 4). We were interested in determining whether wild-type B. burgdorferi undergoes a similar population expansion in the skin after tick bite transmission and if bbk13 contributes to this process. To achieve this, we confined wild-type B. burgdorferi-infected nymphs to capsules affixed to the dorsal skin of mice and determined the kinetics of the B. burgdorferi population in the skin feeding site. The feeding site was harvested daily from cohorts of mice starting at day 3 out to day 12 after nymph application. The bbk13 Is Critical to the B. burgdorferi Enzootic Cycle Infection and Immunity spirochete load in the skin was assessed using the semiquantitative reisolation assay as well as quantitative PCR. Feeding sites collected on days 3 to 6 after nymph application showed inconsistent spirochete reisolation, with one or two of the three mice in the cohort showing an absence of spirochetes and peak scores being achieved only at the end of the scoring period in some cultures (Fig. 5A). However, from days 7 to 10 after nymph application, spirochetes were detected in all cultures as soon as day 1 of incubation and reached a peak score of 3 as early as day 2 of incubation ( Fig. 5A). Reisolation from skin on days 11 to 12 after nymph application resembled that at the early time points, with spirochetes being detected later in the scoring period (Fig. 5A). Quantification of B. burgdorferi loads from the skin feeding site by quantitative PCR corroborated the findings from the reisolation assay. There were few or no measurable B. burgdorferi spirochetes in the skin feeding site from days 3 to 7 after nymph application, followed by a dramatic increase in the B. burgdorferi load during days 8 to 10 after nymph application (Fig. 5B). A precipitous drop in B. burgdorferi loads was detected on days 11 to 12 (Fig. 5B). Blood was also collected from mice throughout the duration of the nymph capsule feeding experiment and assessed for the presence of disseminating spirochetes in circulation. Plating of whole blood in solid BSK-agarose medium and CFU enumeration revealed peak numbers of B. burgdorferi in the blood on days 8 to 10 after nymph application (Fig. 5C). Taken together, these results show that B. burgdorferi undergoes population expansion in the skin after tick transmission, and the peak of this expansion occurs on days 8 to 10 after nymph application. By tick bite transmission, bbk13 is dispensable for B. burgdorferi early population expansion but critical for disseminated infection. Having established the kinetics of wild-type B. burgdorferi population expansion in the skin following tick transmission, we used the nymph capsule feeding model to assess the contribution of bbk13 to this process. Unfed nymphs infected with comparable numbers of either wild-type B. burgdorferi or Dbbk13 B. burgdorferi spirochetes (Fig. 6A) were placed in feeding capsules on the dorsal skin of C3H/HeN mice to feed to repletion. The skin feeding site was collected on days 9 and 10 after nymph application, which corresponded to the peak of spirochete expansion of tick-transmitted wild-type B. burgdorferi (Fig. 5B). A reisolation assay of the skin site fed on by wild-type B. burgdorferi-infected nymphs collected on both days 9 and 10 demonstrated the detection of spirochetes within 1 day of culture incubation (Fig. 6B). Cultures containing the skin site fed on by Dbbk13 B. burgdorferiinfected nymphs showed little reisolation delay compared to wild-type B. burgdorferi, with spirochetes being detected on day 2 of incubation ( Fig. 6B). Surprisingly, no difference in B. burgdorferi loads measured by quantitative PCR was detected between wildtype and Dbbk13 B. burgdorferi-infected skin sites collected on either day (Fig. 6C). These results are in contrast to previous findings that, when delivered by intradermal needle inoculation, Dbbk13 B. burgdorferi failed to undergo efficient population expansion in the skin inoculation site compared to wild-type B. burgdorferi, thus negatively impacting distal tissue colonization at later time points of infection (13). Yet our data indicated that Dbbk13 B. burgdorferi is highly attenuated for disseminated mouse infection by free-feeding tick transmission (Fig. 4). Therefore, we examined the disseminated infection outcomes of Dbbk13 B. burgdorferi delivered by nymph capsule feeding. Three weeks after nymph application in feeding capsules, the feeding site and multiple distal tissues were dissected and assessed for spirochete colonization. Interestingly, despite the lack of B. burgdorferi load differences between wild-type and Dbbk13 B. burgdorferi at the feeding site early in infection, Dbbk13 B. burgdorferi demonstrated reduced numbers in all tissues examined at this later time point. Nearly all reisolation cultures containing tissues from mice fed on by wild-type B. burgdorferiinfected nymphs were positive for spirochetes, while only 5 out of 20 cultures were positive for spirochetes from mice fed on by Dbbk13 B. burgdorferi-infected nymphs (Fig. 6D). Furthermore, quantitative PCR analysis of spirochete loads consistently showed reduced or undetectable Dbbk13 B. burgdorferi spirochetes in tissues (Fig. 6E), indicating the critical requirement for bbk13 for productive disseminated infection.
Dbbk13 B. burgdorferi infectivity is restored in immunocompromised mice. The above-described results show that delivery by tick bite locally rescued the defect of Dbbk13 B. burgdorferi in spirochete expansion in the skin during early infection. It is known that the feeding activity of Ixodes ticks leads to immunomodulation at the bite site, promoting B. burgdorferi survival (9,17), which therefore prompted the question as to whether bbk13 is involved in immune evasion. To test this, we determined whether Dbbk13 B. burgdorferi mammalian infectivity was restored if the host immune response was reduced overall. We delivered 10 4 wild-type, Dbbk13, or Dbbk13/bbk13 1 B. burgdorferi spirochetes by intradermal needle inoculation into highly immunocompromised Nod-scid IL2Rg null (NSG) mice and the congenic control strain, Nod/ShiLtJ. At day 6 postinoculation, blood was collected from mice and plated in solid BSK-agarose medium to enumerate circulating B. burgdorferi spirochetes. No B. burgdorferi clonespecific differences were observed between the number of circulating B. burgdorferi spirochetes in immunocompromised NSG mice and that in the control mouse strain (Fig. 7A). At 3 weeks postinoculation, infection of the inoculation site and distal tissues was assessed by a semiquantitative reisolation assay. As expected, Dbbk13 B. burgdorferi spirochetes were highly attenuated for infection of the control mice, with only 6 out of 30 cultures having a score of 1 or 2 following 5 days of incubation. In contrast, 29 out of 30 and 24 out of 30 reisolation cultures from wild-type B. burgdorferiand Dbbk13/bbk13 1 B. burgdorferi-infected NOD mice, respectively, received a score of 1 to 3 at the same time point (Fig. 7B, top). Strikingly, in immunocompromised NSG mice, FIG 6 When delivered by tick transmission, bbk13 is not required for B. burgdorferi population expansion in the skin but is critical for disseminated infection. (A to C) Unfed nymphs infected with wild-type (WT) or Dbbk13 B. burgdorferi were placed in capsules adhered to the dorsal skin of groups of six naive C3H/HeN mice to feed. The tick feeding site was collected at days 9 and 10 after nymph application for analysis. (A) Prior to feeding, subsets of individual unfed B. burgdorferi-infected nymphs were homogenized and plated in solid BSK-agarose medium to assess spirochete loads by enumerating CFU. (B) The tick feeding site was assessed for B. burgdorferi by a tissue reisolation assay. Semiquantitative scoring of reisolation cultures was performed daily over 4 days of culture incubation (0 indicates no spirochetes, and 1 to 3 indicate increasing spirochete densities). (C) Total DNA was extracted from the tick feeding site. The B. burgdorferi load was measured by quantifying B. burgdorferi flaB copies normalized to 10 6 mouse nid copies using quantitative PCR. In cases where not all samples per group had detectable B. burgdorferi DNA, the numbers of positive samples out of 6 mice per group are indicated in the scatterplot. Data points below the level of detection are not shown. The mean is indicated by a horizontal black line. The Mann-Whitney test was used to determine statistical significance (n.s., not significant). (D and E) Unfed nymphs infected with WT or Dbbk13 B. burgdorferi were placed in capsules adhered to the dorsal skin of groups of 2 to 4 naive C3H/HeN mice to feed. Three weeks after nymph application, the tick feeding site and various distal tissues were collected for analysis. (D) Tissues were assessed for B. burgdorferi by a tissue reisolation assay. Semiquantitative scoring of reisolation cultures was performed on day 6 of culture incubation (0 indicates no spirochetes, and 1 to 3 indicate increasing spirochete densities). (E) Total DNA was extracted from the indicated tissues. B. burgdorferi loads were measured by quantifying B. burgdorferi flaB copies normalized to 10 6 mouse nid copies using quantitative PCR. In cases where not all samples per group have detectable B. burgdorferi DNA, the numbers of positive samples out of the total mice per group are indicated in the scatterplot. Data points below the level of detection are not shown. The mean is indicated by a horizontal black line.
Dbbk13 B. burgdorferi demonstrated robust infection. Twenty-one out of 30 reisolation cultures from Dbbk13 B. burgdorferi-inoculated NSG mice received a score of 3 following 5 days of incubation, congruent with those of wild-type B. burgdorferi (25/30 cultures) and Dbbk13/bbk13 1 B. burgdorferi (22/30 cultures) (Fig. 7B, bottom). These data indicated that without host immune pressure, the infection defect of Dbbk13 B. burgdorferi was overcome, resulting in productive disseminated infection of distal tissues and suggesting a role for BBK13 in immune evasion.

DISCUSSION
The maintenance of B. burgdorferi in the natural reservoir depends on its ability to colonize a vertebrate host, to position itself to be acquired during tick feeding, and to subsequently be transmitted to a new vertebrate host. Although B. burgdorferi colonizes several mammalian tissues that contribute to the pathogenesis of Lyme disease in humans, colonization of the skin of reservoir hosts is most pertinent to the upkeep of the enzootic cycle. Intravital imaging of the skin of infected mice revealed that B. burgdorferi undergoes directed migration toward a feeding tick (18). Furthermore, B. burgdorferi has shown positive chemotactic responses to components of the tick salivary gland and that these factors are important during spirochete acquisition by feeding ticks (19,20). This evidence supports the ability of B. burgdorferi to sense the presence of the feeding tick and to position itself to best be taken up by a feeding tick. However, these cues are likely restricted to the immediate vicinity of an active tick feeding site. Therefore, efficient B. burgdorferi acquisition likely depends on achieving the appropriate density of B. burgdorferi spirochetes in the skin such that enough B. burgdorferi spirochetes will encounter the cues from a feeding tick. B. burgdorferi spirochetes must then be present in the feeding site at a sufficient number for successful tick acquisition and continuation of the enzootic cycle. bbk13 Is Critical to the B. burgdorferi Enzootic Cycle Infection and Immunity We and others have previously described the population expansion of B. burgdorferi at the site of needle inoculation (13,(21)(22)(23)(24). We further showed that this population expansion is promoted by gene bbk13, and the inability to expand in numbers at the skin inoculation site leads to attenuated dissemination and distal tissue colonization (13). We now show that naive larvae fail to acquire Dbbk13 B. burgdorferi from needleinoculated mice. Our data suggest that the number of Dbbk13 B. burgdorferi spirochetes colonizing the skin is likely below the threshold required for successful tick acquisition but do not rule out the possibility of a specific function of bbk13 as a driver of the acquisition defect. Matching the wild-type and Dbbk13 B. burgdorferi numbers in the skin would allow any bbk13-specific requirement to be tested during tick acquisition. However, increasing the needle inoculum dose of the mutant did not result in increased spirochete loads in the skin.
Using the semiquantitative tissue reisolation assay as a measure of tissue burden and highly immunocompromised NSG mice, we discovered that when host immune pressures are relieved, Dbbk13 B. burgdorferi is able to colonize distal tissues similarly to wild-type B. burgdorferi. Thus, it appears that bbk13 is important for circumventing the host immune response, either directly or indirectly, to promote B. burgdorferi survival in the mammalian host. The restoration of Dbbk13 B. burgdorferi loads in the tissues of NSG mice to apparent wild-type levels suggests that this model could be used to test for a bbk13-specific requirement for tick acquisition. However, there are a number of limitations to this model in its ability to address this question. The semiquantitative nature of the tissue reisolation assay does not provide sufficient resolution at a score of 3 (.15 spirochetes in the field of view) to precisely distinguish quantitative differences in tissue loads between wild-type and Dbbk13 B. burgdorferi infections. Furthermore, assessment of spirochete loads in distal tissues of NSG mice using quantitative PCR enumeration of B. burgdorferi genome copies was unreliable. Tissues that had no reisolation of live spirochetes recorded high spirochete loads by qPCR (data not shown). This observation indicates that B. burgdorferi DNA is retained in the tissues despite the absence of viable spirochetes. The NSG mice and the congenic control strain used in our studies are of the Nod/ShiLtJ background, commonly called NOD (nonobese diabetic). NOD mice have been shown to have poor clearance of apoptotic cells and the generation of autoantibodies against DNA (25), which indicates poor clearance of DNA. Moreover, hyperglycemic mice are suspected to suffer from deficiencies in clearing B. burgdorferi debris and DNA from tissues (26). This is an important consideration for mouse infection studies since quantitative PCR measurement of B. burgdorferi genome copies in distal tissues, the current gold standard for assessing B. burgdorferi tissue loads, assumes a correlation between the amounts of B. burgdorferi genomic DNA and viable spirochetes. The caveats of precise quantitation of B. burgdorferi in the tissues of NSG mice have the potential to confound the interpretation of tick acquisition, or the lack thereof, of Dbbk13 B. burgdorferi. Therefore, NSG mice were deemed not to be the appropriate model to test this question. In addition, while the removal of immune pressures clearly allowed Dbbk13 B. burgdorferi to achieve disseminated infection, given the technical limitations of these mice, we cannot rule out the potential that BBK13 also contributes to a broader mechanism of dissemination not overcome by the immunodeficiency of the NSG background. Nonetheless, the broadly immunocompromised background of NSG mice is a useful tool to eliminate a large swath of the host immune response across both innate and adaptive arms. Future studies will focus on the identification of the specific aspect(s) of the host immune response that BBK13 contributes to providing protection against.
Consistent with previous findings for B. burgdorferi lacking the entire linear plasmid 36 (lp36) (27), on which the bbk13 gene resides (28), no bbk13-dependent defect in B. burgdorferi replication or survival was detected across the tick life stages. Tick bite transmission of B. burgdorferi induces significant changes in both the spirochete and the mammalian host to promote infection (8,9), which are not recapitulated by in vitro-grown B. burgdorferi delivered by needle inoculation. Therefore, we sought to assess the contribution of the bbk13 gene to mouse infection by tick bite transmission. B. burgdorferi spirochetes lacking bbk13 were highly attenuated for disseminated infection 3 weeks following free feeding of infected ticks on mice. Our previous findings indicated a role for bbk13 in B. burgdorferi population expansion in the skin site of infection and the importance of this population expansion for the establishment of productive disseminated infection (13). To investigate the contribution of bbk13 to this model of infection in the context of tick bite transmission, we first established the early kinetics of wild-type B. burgdorferi in the skin site of tick feeding using capsule-confined nymphs. Similar to the spirochete numbers in the skin delivered by needle inoculation, tick-delivered B. burgdorferi underwent population expansion in the feeding site, with peak numbers of spirochetes being detected on days 8 to 10 after nymph application. Given that B. burgdorferi transmission can occur 24 to 48 h after the start of tick feeding and that tick feeding is not synchronized, the observed kinetics of spirochete population expansion in the skin was comparable to that determined by needle inoculation (13). Using the nymph capsule feeding approach, analysis of the numbers of Dbbk13 B. burgdorferi spirochetes in the tick feeding site on days 9 and 10 after nymph application revealed the striking result that the loss of bbk13 had no impact on spirochete population expansion in the skin. It is important to note that peak numbers of wild-type B. burgdorferi spirochetes in the blood were also detected on days 8 to 10 after nymph application. As the skin is a highly vascularized tissue and the mice were not perfused prior to harvest of the tick feeding site, it remains a possibility that circulating B. burgdorferi spirochetes in the blood may have contributed to the numbers of wild-type and Dbbk13 B. burgdorferi spirochetes detected in the skin at these time points. Despite the amelioration of the early infection defect of Dbbk13 B. burgdorferi when delivered by tick feeding, assessment of mice fed on by capsule-confined nymphs 3 weeks after nymph application revealed that the bbk13 gene remained critical for disseminated infection. Together, these data suggest that one or more of the biological changes that the spirochetes and/or the host undergoes during tick feeding allowed Dbbk13 B. burgdorferi to overcome the barrier to population expansion in the skin. However, the relief of this early phenotype was not sufficient to overcome the need for bbk13 during disseminated infection. These findings emphasize that B. burgdorferi population expansion in the skin and dissemination to distal tissues are distinct events of mammalian infection that are uniquely influenced by the feeding tick and likely require different sets of B. burgdorferi genes. Furthermore, this work underscores the significance of the vector as a driver of vector-borne pathogen adaptation and pathogenesis.
Our studies have provided greater insight into the role of bbk13 in the biology of B. burgdorferi, yet the molecular function of BBK13 remains unknown. Recently, we established that BBK13 forms large oligomeric complexes in the spirochete membrane (14). Current work is focused on investigating how BBK13 oligomerization contributes to the mechanisms by which B. burgdorferi surmounts immune barriers in the mammalian host in order to establish a disseminated infection. In sum, this work cements the importance of bbk13 in promoting B. burgdorferi mammalian infectivity and highlights the influence of the tick vector on the pattern of infection, particularly at the skin-the interface between host, vector, and pathogen.

MATERIALS AND METHODS
B. burgdorferi clones and growth conditions. The Borrelia (Borreliella) burgdorferi clones used in this study were derived from the low-passage-number infectious clone B31 A3-68 Dbbe02, referred to here as the wild type, which lacks plasmids cp9 and lp56 as well as the bbe02 gene on lp25 (29). The Dbbk13 mutant clone and the Dbbk13/bbk13 1 complemented strain have been described previously (13). B. burgdorferi cultures were grown at 35°C in liquid Barbour-Stoenner-Kelly II (BSKII) medium containing gelatin and 6% rabbit serum. Alternatively, B. burgdorferi spirochetes were plated in solid BSKagarose medium and incubated at 35°C under 2.5% CO 2 . The following antibiotics were used, as needed: kanamycin (200 mg/ml), streptomycin (50 mg/ml), and gentamicin (40 mg/ml).
Ethics statement. The University of Central Florida is accredited by the International Association for Assessment and Accreditation of Laboratory Animal Care. Protocols for all animal experiments were bbk13 Is Critical to the B. burgdorferi Enzootic Cycle Infection and Immunity