Borreliella burgdorferi factor H-binding proteins are not required for serum resistance and infection in mammals

ABSTRACT

The causative agent of Lyme disease (LD), Borreliella burgdorferi, binds factor H (FH) and other complement regulatory proteins to its surface. B. burgdorferi B31 (type strain) encodes five FH-binding proteins (FHBPs): CspZ, CspA, and the OspE paralogs OspE_BBN38, OspE_BBL39, and OspE_BBP38. This study assessed potential correlations between the production of individual FHBPs, FH-binding ability, and serum resistance using a panel of infectious B. burgdorferi clonal populations recovered from dogs. FHBP production was assessed in cultivated spirochetes and by antibody responses in naturally infected humans, dogs, and eastern coyotes (wild canids). FH binding specificity and sensitivity to dog and human serum were also assessed and compared. No correlation was observed between the production of individual FHBPs and FH binding with serum resistance, and CspA was determined to not be produced in animals. Notably, one or more clones isolated from dogs lacked CspZ or the OspE proteins (a finding confirmed by genome sequence determination) and did not bind FH derived from canines. The data presented do not support a correlation between FH binding and the production of individual FHBPs with serum resistance and infectivity. In addition, the limited number and polymorphic nature of cp32s in B. burgdorferi clone DRI85A that were identified through genome sequencing suggest no strict requirement for a defined set of these replicons for infectivity. This study reveals that the immune evasion mechanisms employed by B. burgdorferi are diverse, complex, and yet to be fully defined.

INTRODUCTION

Lyme disease (LD) is the most common vector-borne disease in the northern hemisphere (1–4). Borreliella burgdorferi, the primary causative agent of LD in North America (5, 6), is maintained in an enzootic cycle involving Ixodes ticks and diverse vertebrate hosts (7). LD is a growing health threat, with more than 470,000 clinician-diagnosed human cases each year in the United States (1). While routine screening for LD in humans is not practiced, yearly screening is recommended in veterinary medicine for dogs residing in or near endemic regions (8). The Companion Animal Parasite Council (CAPC) reported 465,721 B. burgdorferi antibody-positive tests in client-owned dogs in 2023 (https://capcvet.org). However, since the CAPC collects data for only 30% of the total number of United States Department of Agriculture (USDA)-approved B. burgdorferi antibody tests run yearly in the US, the actual number of positive Ab tests likely exceeds 1 million. While wildlife and wild canids have a high exposure risk to tick-transmitted pathogens (9), little is known regarding the impact of tick-borne diseases on wildlife health. A serological analysis of plasma samples from eastern coyotes (Canis lupus) from the Commonwealth of Pennsylvania collected from 2015 to 2018 revealed that 64.8% and 72.7% were antibody-positive for B. burgdorferi and Anaplasma phagocytophilum, respectively (10).

The LD spirochetes must evade innate immunity to establish an infection in mammals. The mechanisms associated with complement evasion have been the subject of intensive research [reviewed in reference (11)]. Negative regulators of complement, including factor H (FH), FH-like protein 1, FH-related proteins, and C1r- and C4b-binding proteins, have been demonstrated to bind to specific Borreliella surface proteins (12–16). B. burgdorferi B31 produces five FH-binding proteins: CspZ, CspA, and the OspE paralogs OspE_BBN38, OspE_BBL39, and OspE_BBP38. All are encoded by genes carried by plasmids or 32 kb prophage (cp32s) (17). The molecular interactions between recombinant FH-binding proteins (FHBPs) and FH are well characterized (18–22). While in vitro studies have revealed that FH bound to the surface of B. burgdorferi is competent to serve as a cofactor for factor I-mediated cleavage of C3b, an important opsonin [reviewed in reference (23)], the significance of FH binding in ticks and mammals remains an open question. In this study, we assessed 24 clonal populations of B. burgdorferi that were recovered from infected dogs for FHBP production, FH binding, and serum sensitivity. We demonstrate that CspA is not produced during infection in mammals and that the OspE and CspZ proteins (and FH binding in general) are not required for serum resistance. The data suggest that complement evasion may be mediated through FH-independent mechanisms.

MATERIALS AND METHODS

Bacterial strains and growth conditions

B. burgdorferi clones analyzed in this report were recovered from tissue biopsies from dogs that were experimentally infected using Ixodes scapularis or Ixodes pacificus ticks (24). The original non-clonal isolates were assigned the prefixes DRI, DCT, DWI, or DCA to indicate that they originated from dogs (D) that were infected using ticks collected in Rhode Island (RI), Connecticut (CT), Wisconsin (WI), or California (CA). An arbitrary identifier number follows these designations. A letter designation further differentiates the clones obtained by sub-surface plating. All clones were cultivated in BSK-H complete media supplemented with 6% rabbit serum at 34°C, and growth was monitored using wet mounts and dark-field microscopy. Cells were harvested by centrifugation, and cell lysates were generated for SDS-PAGE and immunoblot analyses, as described below. B. burgdorferi B31 (clone 5A4) was a control, as its genome sequence is known (17), and its FHBPs have been well characterized (25).

Serum samples

Serum samples from B. burgdorferi antibody-positive client-owned dogs were provided by the College of Veterinary Medicine, North Carolina State University. Blood samples collected from eastern coyotes (Canis lupus) harvested in PA between 2015 and 2020 were obtained under a Pennsylvania Game Commission Special Use Scientific Studies Permit (#48548). The coyote serum had been previously assessed for antibodies to B. burgdorferi, A. phagocytophilum (10), canine parvovirus, and distemper virus (26). The coyote samples were included in this study as they are wild canids with high exposure risk for tick-borne diseases. Their analysis is not complicated by previous vaccination or acaricide applications. Human serum samples were obtained from the Lyme disease Biobank (27). Note that except for negative control sera, all serum samples used in this study were confirmed to be antibody positive for B. burgdorferi by whole-cell lysate immunoblot and enzyme-linked immunosorbent assay (ELISA) using recombinant B. burgdorferi VlsE and other characterized proteins as the detection antigens.

Protein nomenclature and generation of recombinant proteins

Several nomenclature strategies have been used to differentiate the FHBPs of B. burgdorferi and related species. These proteins have been collectively referred to as complement regulator-acquiring surface proteins (28), Erps (OspEF-related proteins) (29), or Elps (30). Erp designations have been broadly applied to diverse and functionally distinct proteins belonging to different gene families (31). For example, the OspF proteins (BB039, BBM38, and BBR42), which do not bind to complement regulatory proteins (32), have been referred to as Erps in some studies. To be consistent with previous studies from our laboratory (19, 33–36), we employ the original protein and open reading frame (ORF) designations assigned to the FHBPs of B. burgdorferi B31 (CspA, CspZ, OspE_BBL39, OspE_BBP38, and OspE_BBN38). OspE_BBL39 and OspE_BBP38 are encoded by genes located on different cp32s but are identical in sequence, and hence, we collectively refer to them as OspE_BBL39/P38 (17).

Genes encoding recombinant proteins were PCR amplified using B. burgdorferi B31 DNA and standard conditions with Phusion polymerase (ThermoScientific). PCR primers were synthesized with restriction sites or ligase-independent cloning tails to allow for cloning into pET45b(+) and pET46, respectively (Novagen), as previously described (37). Some genes were codon optimized for expression in Escherichia coli, synthesized (Genscript), and provided by the supplier directly in pET45b(+). Recombinant plasmids were transformed into E. coli BL21/DE3 cells and protein production induced using isopropyl β-D-1-thiogalactopyranoside (IPTG) (0.1 mM) or autoinduction (overnight). The cells were harvested by centrifugation, lysed with a high-pressure cell homogenizer, and purified via their N-terminal hexa-His tags from the soluble fraction using nickel affinity chromatography on an AKTA FPLC purification platform (Cytiva). Protein purification was performed as previously described (38).

Generation of antisera

Antisera to recombinant proteins were generated in Sprague-Dawley rats (10). In brief, rats were anesthetized with isoflurane, injected intraperitoneally with 40 µg of each recombinant protein in Freund’s complete adjuvant (day 0), and then boosted with 40 µg of protein in Freund’s incomplete adjuvant (days 21 and 35). On day 42, the rats were euthanized, blood was collected by cardiac puncture, and serum was harvested using standard methods. All animal experiments were conducted following the Guide for the Care and Use of Laboratory Animals (eighth edition) and in accordance with protocols peer-reviewed and approved by Virginia Commonwealth University Institutional Animal Care and Use Committees.

SDS-PAGE and immunoblot analyses

Cell lysates were fractionated by SDS-PAGE in 15% Criterion AnyKDa Gels (Bio-Rad). The proteins were transferred to polyvinylidene fluoride (PVDF) membranes (0.22 µm) using the Trans-Blot Turbo system according to manufacturer instructions (Bio-Rad). The membranes were incubated with blocking solution (5% non-fat dried milk, 0.2% Tween-20 in phosphate buffered saline [PBS]; 1 h; room temperature) followed by the addition of anti-CspA (1:1,000), anti-CspZ (1:1,000), anti-OspE_BBL39/P39 (1:20,000), or anti-OspE_BBN38 antisera (1:1,000). The antisera were removed. The blots were washed three times (0.2% Tween-20 in PBS), followed by the addition of the appropriate secondary antibodies (1:40,000 in blocking solution). Bound IgG was detected using Clarity Western ECL substrate (5 min; Bio-Rad). Images were captured using a ChemiDoc Touch imaging system (Bio-Rad; auto-optimal settings). Images were cropped for presentation purposes.

FH-binding assays

FH binding was assessed as previously described (25). In brief, whole-cell lysate immunoblots were blocked, washed, and incubated with 10% dog, coyote, or human serum (human serum from Complement Tech). Bound FH was detected using sheep anti-FH antiserum (1:1,000; Invitrogen) with rabbit anti-sheep IgG as the secondary (Pierce). Bound IgG was detected as above.

ELISA

ELISA plate wells were coated with 500 ng of recombinant protein in bicarbonate buffer (overnight; 4°C) (39). Dog, coyote, human, or mouse sera were added (1:1,000 dilution; 1 h; room temperature), followed by three washes. A secondary anti-IgG antibody was added at a 1:15,000 dilution. After washing, the 2,2'-azino-bis[3-ethylbenzothiazoline-6-sulfonic acid (ABTS)] substrate was added (20 min), and absorbance was measured at 405 nm. Immobilized bovine serum albumin (BSA) served as an immobilized protein control for non-specific antibody binding. Serum samples were scored as antibody positive if the mean absorbance value was twofold greater than the mean absorbance of the serum with BSA or preimmune serum.

Serum sensitivity analyses

Mid-log phase B. burgdorferi B31 cells were incubated (34°C; 18 h) in 0%, 40%, and 80% (final concentrations) dog or human complement-preserved serum (Innovative Research). The assays were performed in triplicate. Live cells were counted (400×; an average of five fields of view) using dark-field microscopy. Percent killing was calculated by dividing the number of live cells in each tube by the number of live cells in the controls. All statistical analyses were conducted as previously described (40).

Mouse infectivity analyses

C3H/HeN mice were injected (subcutaneously) with 10⁵ mid-log phase B. burgdorferi B31 or B. burgdorferi DRI85A and sacrificed 17 days later. Ear tissue and urinary bladders were collected, macerated, and placed in media with a standard Borreliella antibiotic cocktail (41). Blood was harvested from each mouse by cardiac puncture. Seroconversion was assessed by ELISA, as detailed above, using the infection serum at a 1:1,000 dilution.

DNA extraction, preparation, and sequencing

Bacterial pellets were resuspended in 200 µL of 1× PBS and transferred to a 2-mL bead beating tube (Matrix E; MP Biomedicals). Proteinase K (20 µL; Qiagen) was added, and the cells homogenized (SPEX 1600 MiniG; 1 min; 1,500 Hz Fisher Scientific). DNA was extracted using the Qiagen DNeasy Blood & Tissue Kit according to the manufacturer’s instructions (Qiagen). The DNA was quantified using the 1× dsDNA HS kit (ThermoFisher Scientific on a Qubit). DNA was prepped with the SMRTbell Template Prep Kit 2.0 (Pacific Biosciences) to make PacBio SMRTbell libraries with barcodes sourced from the Barcoded Overhang Adaptor Kit 8A and 8B (Pacific Biosciences). The sequencing primers were annealed and bound to Polymerase 3.0 using the Sequel Binding Kit 3.0 (Pacific Biosciences). The bound complex was purified and sequenced on a PacBio Sequel I using an SMRT Cell M1 v3 tray (Pacific Biosciences). The spike-in controls for each PacBio Sequel I run were from the Internal Control Kit 3.0 (Pacific Biosciences).

Genome assembly, annotation, and pangenome construction

Genome sequences were processed using pbcromwell (v1.0.4) (Pacific Biosciences) and the barcoded data were demultiplexed using pb_demux_subreads and assembled using pb_assembly_microbial. The base modification motifs for each assembly were computed using pb_basemods. The species’ identity was verified using the taxator and GTDB-Tk (v1.7.0) with refpack (v r202). All reference and sequenced Borreliella assemblies were annotated with Prokka (v 1.11) (42) and EggNOG (43), and then homologous genes were clustered with Roary (v 3.5.1) (44). The BlastP threshold was calculated consecutively at 5% intervals to determine the final threshold (75%).

RESULTS

Comparative analysis of the in vitro expression of FHBPs by B. burgdorferi clonal populations

FHBP profiles of low-passage B. burgdorferi clonal populations isolated from infected dogs were determined by immunoblot analysis of cell lysates using anti-OspE_BBN38, anti-OspE_BBL39/P38 anti-CspA, and anti-CspZ antisera (Fig. 1). Anti-OspE_BBL39/P38 antisera and anti-OspE_BBN38 antisera reacted with proteins of the predicted molecular weight in most clones except B. burgdorferi DRI85A. Note that anti-OspE_BBN38 antibodies bound to one or more of the OspF proteins (26–28 kDa; indicated in Fig. 1) produced by some clones. The absence of immunoreactive OspE and OspF proteins in DRI85A suggests that this clone lacks one or more cp32s that carry these genes (discussed in detail below). Consistent with the universal distribution of linear plasmid 54 (lp54), which carries cspA, a protein consistent in size with CspA was detected in all clones. In contrast, CspZ was detected in only 17 of 24 clones. This observation is consistent with previous studies that demonstrated that the cspZ encoding plasmid (lp28-3) is polymorphic and (45) not carried by all isolates (46).

Fig 1.

FHBP production by B. burgdorferi clones derived from infected dogs. Cell lysates of B. burgdorferi clonal populations were fractionated by SDS-PAGE using AnyKda gels. Clone designations are indicated along the top of the figure. B. burgdorferi B31 served as a positive control. One gel was stained with Coomassie brilliant blue to demonstrate similar loading (top panel). Immunoblots were generated and screened with antisera to the FHBPs (indicated to the right). The molecular weight (MW) of OspE_BBN38 and OspE_BBL39 are 20.675 and 19.560 kDa, respectively. Variation in the MW of the OspE paralogs in some clones rendered definitive identification at the paralog level difficult. Hence, the proteins are labeled simply as OspE. Note that anti-OspE_BBN38 also reacts with OspF (indicated), which is not an FHBP. The migration position of MW standards is indicated to the left. The antisera were used at a 1:1,000 dilution. All methods were as described in the text, and the images were cropped for presentation purposes.

FH-binding analyses

To determine if the FHBPs of the B. burgdorferi clones bind FH from humans, dogs, and eastern coyotes, whole-cell lysate immunoblots were incubated with serum from each (Fig. 2). Human FH bound to CspZ, CspA, and one or more OspE paralogs in most clones. In contrast, dog and coyote FH bound only to OspE proteins in a subset of clones. Canine FH did not bind to proteins in the OspE size range in clone DRI85A. This is consistent with the lack of detectable OspE proteins in the immunoblot analyses described above. The fact that clone DRI85A does not produce proteins that bind FH is noteworthy, as this clone was derived from an infected dog and, as shown below, is infectious in mice.

Fig 2.

Comparative analysis of the FHBP profiles of B. burgdorferi clonal populations. Immunoblots of cell lysates were generated as described in the Materials and Methods section and in Fig. 1. The membranes were incubated with the dog, eastern coyote, or human sera as indicated. The sera, which served as the species-specific FH source, were confirmed to be antibody-negative for B. burgdorferi before use. Bound FH was detected using anti-FH antiserum, as detailed in the text. Clone designations are indicated along the top of the figure. The identities of the proteins that bound FH are indicated to the right.

Antibody responses to the FHBPs in B. burgdorferi antibody-positive dogs, coyotes, and humans

To determine if OspE_BBL39/P38, OspE_BBN38, CspZ, and CspA were produced in naturally infected humans, dogs, and eastern coyotes, we screened serum from B. burgdorferi antibody-positive animals against recombinant FHBPs. Antibodies to OspE_BBN38 and anti-OspE_BBL39/P38 were detected in 42% and 6% of human serum samples, respectively (Table 1). Antibodies to CspZ and CspA were detected in 14 of the 100 of the dog and coyote serum samples. While the absence of detectable antibodies to the FHBPs in most serum samples suggests that they are not universally produced during infection in humans and canines, we cannot exclude the possibility that differences in immune responses among animals, varying surface exposure or expression levels, are contributing factors.

TABLE 1.

Percentage of human, dog, and coyote serum samples positive for Borreliella proteins 610 by ELISA screening

Detection antigen	Human (N = 50)	Dog (N = 50)	Coyote (N = 50)
CspA	00.0%	00.0%	02.0%
CspZ	00.0%	00.0%	14.0%
OspEBBN38	42.0%	34.0%	12.0%
OspEBBL39/P38	06.0%	38.0%	24.0%
OspEBBN38 or OspEBBL39/P38	46.0%	48.0%	28.0%
VlsE	100.0%	100.0%	100.0%
BSA	00.0%	00.0%	00.0%

Serum sensitivity to human and canine serum

The sensitivity of 10 B. burgdorferi clones to increasing concentrations of complement-certified dog and human serum was assessed. Coyote serum was not tested since it is not commercially available at the “complement certified” grade. The percent killing among clones treated with 40% dog serum ranged from 11.4% to 100.0% (Table 2). The near complete killing of all clones (94.9% to 100.0%) was observed when exposed to 80% dog serum. The difference in percent killing of DWI14C and DWI14I in 40% canine serum (11.4% versus 100.0%) is noteworthy as these clones were isolated from the same tissue biopsy (24), and both produce CspA, CspZ, and at least one OspE protein. However, as detailed above, canine FH does not bind to CspA or CspZ. When the clones were treated with 40% and 80% human serum, cell killing ranged from 0.0% to 95.6% and 64.1% to 100.0% killing, respectively. Clones DRI85A and DRI03A were not killed by 40% human serum, whereas clones DWII4I and DWII4C were highly sensitive (>92%). Cells incubated in media alone or 80% heat-inactivated (HI) serum served as controls, and as expected, HI serum did not negatively impact growth. The fact that the CspA protein produced by all clones bound human FH but did not convey serum resistance indicates that CspA in and of itself is not sufficient to protect against human complement.

TABLE 2.

Sensitivity of diverse B. burgdorferi clones to canine and human serum

	BSK H media	Heat inactivated 80% serum	40% serum (complement certified)	80% serum (complement certified)
Canine serum assays
B31	0.0 ± 0.0	0.0 ± 0.0	72.1 ± 0.5	100.0 ± 0.0
DWI14I	0.0 ± 0.0	0.0 ± 0.0	100.0 ± 0.0	100.0 ± 0.0
DWI28J	0.0 ± 0.0	0.0 ± 0.0	39.2 ± 4.6	94.9 ± 0.6
DWI11G	0.0 ± 0.0	0.0 ± 0.0	100.0 ± 0.0	100.0 ± 0.0
DWI14C	0.0 ± 0.0	0.0 ± 0.0	11.4 ± 4.3	100.0 ± 0.0
DCT29B	0.0 ± 0.0	0.0 ± 0.0	91.8 ± 1.5	100.0 ± 0.0
DCT2J	0.0 ± 0.0	0.0 ± 0.0	93.8 ± 2.1	100.0 ± 0.0
DRI85A	0.0 ± 0.0	0.0 ± 0.0	37.8 ± 3.5	100.0 ± 0.0
DRI03A	0.0 ± 0.0	0.0 ± 0.0	78.9 ± 1.5	96.8 ± 1.1
DRI40I	0.0 ± 0.0	0.0 ± 0.0	100.0 ± 0.0	100.0 ± 0.0
Human serum assays
B31	0.0 ± 0.0	0.0 ± 0.0	34.1 ± 1.3	100.0 ± 0.0
DWI14I	0.0 ± 0.0	0.0 ± 0.0	92.6 ± 0.6	100.0 ± 0.0
DWI28J	0.0 ± 0.0	0.0 ± 0.0	89.5 ± 2.5	90.0 ± 1.3
DWI11G	0.0 ± 0.0	0.0 ± 0.0	95.6 ± 1.1	95.6 ± 0.8
DWI14C	0.0 ± 0.0	0.0 ± 0.0	69.8 ± 2.5	88.1 ± 2.6
DCT29B	0.0 ± 0.0	0.0 ± 0.0	19.1 ± 1.5	72.5 ± 3.2
DCT2J	0.0 ± 0.0	0.0 ± 0.0	77.8 ± 3.2	96.1 ± 0.6
DRI85A	0.0 ± 0.0	0.0 ± 0.0	0.0 ± 0.0	64.1 ± 4.7
DRI03A	0.0 ± 0.0	0.0 ± 0.0	0.0 ± 0.0	75.7 ± 5.5
DRI40I	0.0 ± 0.0	0.0 ± 0.0	68.2 ± 1.5	100.0 ± 0.0

Infectivity analysis of B. burgdorferi clone DRI85A

The aforementioned data suggest that FHBPs are not required for infectivity. To assess this directly, the ability of clone DRI85A to infect mice was tested with B31 serving as a positive control. DRI85A was cultured from ear biopsies and/or the bladder of four of six mice. Seroconversion was assessed by ELISA using DbpB, VlsE, and OspA as detection antigens (Fig. 3). Due to limited amino acid sequence identity (54%) between VlsE from DRI85A and B31, recombinant proteins were generated for both sequence variants and included in the ELISA. Serum from B31 and DRI85A infected mice displayed similar antibody responses to DbpB. However, the reaction to the VlsE proteins was variant-specific. Antibodies to the negative control proteins (OspA and BSA) were not detected. It can be concluded from the infectivity analyses that the OspE proteins and FH binding, in general, are not essential for the Borreliella infection of mice and dogs overall.

Fig 3.

Immunoblot analysis and ELISA of B. burgdorferi clones DRI85A and B31 using serum from infected mice. Mice were inoculated with B. burgdorferi clone DRI85A or B. burgdorferi B31. Seventeen days post-inoculation, blood was collected, and the serum was harvested. To assess seroconversion, the sera were screened by ELISA against recombinant DbpB and VlsE (B31 sequence) and the VlsE protein derived from clone DRI85A. OspA, a Borreliella protein that is not expressed during infection, and BSA, served as negative controls. Significance was assessed as detailed in the text (* indicates P < 0.05).

Genome data

The genome of DRI85A is 1,482,443 bases and consists of 18 plasmids and a linear chromosome of 919,128 bases. The complete genome sequence of DRI85A and other clones analyzed in this study were determined as part of a large-scale pangenome project (Bioproject PRJNA1026537) that will be described separately. Here, we present only information relevant to FH binding and immune evasion. B. burgdorferi DRI85A carries only four cp32-related replicons, none encoding ospE or ospF genes. By comparison, B. burgdorferi B31 carries 11 cp32s. The DRI85A cp32-related replicons consist of a 22,696 bp circular plasmid (contig 12) with 99.92% identity (100% query coverage) to cp32-4 of B. burgdorferi strain 297; a circular plasmid of 66,968 bp (contig 17) that is an imperfect “concatemer” of two cp32s; a 30,636 bp circular plasmid that shares 98.86% nt identity (100% query coverage) with cp32-6 from B. burgdorferi 297; and a 10,740 bp circular plasmid with 99.23% nt identity (98% query coverage) to a segment of cp32-8 of B. burgdorferi strain JD1. The absence of cp32s with ospE and ospF genes is consistent with the lack of detection of these proteins in the immunoblot and FH-binding analyses detailed above.

DISCUSSION

The relative contribution of individual B. burgdorferi FHBPs in complement evasion has been challenging to determine due to functional redundancy and the variability of some replicons that encode the FHBPs. This report focused on our analysis of FH binding and serum sensitivity, exclusively on low-passage B. burgdorferi clonal populations recovered from dogs that were infected using field-collected ticks (24). This approach eliminates the variables associated with using non-clonal and lab-attenuated strains.

The FHBP profile of each clone was determined by screening whole-cell lysate immunoblots with antiserum generated against each B. burgdorferi FHBP. Consistent with a recent comprehensive genome analysis that reported that plasmid lp54 is universal (47), CspA was detected in all clones. Although CspA is expressed during cultivation, anti-CspA antibodies to CspA were not detected in any human or canine serum samples (n = 100). Earlier studies demonstrated transcriptional downregulation of cspA by B. burgdorferi when cultivated in media supplemented with human blood (48) or when cultured in dialysis membrane chambers implanted in the peritoneal cavity of rats (49). In addition, cspA mRNA was not detected in spirochetes residing in the skin of infected mice (50), and CspA was not detected in the tissues of infected mice using advanced proteomic approaches (51). Collectively, data presented to date firmly indicate that natural isolates do not produce CspA during infection in mammals. CspA may facilitate complement evasion in the mid-gut of fed ticks. In support of this possibility, B. burgdorferi cspA gene deletion mutants are rapidly cleared from the midgut of feeding nymphal ticks (52).

In contrast to CspA, CspZ was only detected in a subset of clones (17 of 24). It has been demonstrated that the plasmid that encodes CspZ is not universal among LD isolates (47). Analyses of cspZ expression during infection have led to different conclusions. This study detected anti-CspZ antibodies in only 14 canine (wild and domestic) sera and none of the human sera. In contrast, Baum et al. reported detecting anti-CspZ antibodies in laboratory-infected purpose-bred beagles (53). Dulebohn et al. experimentally cured lp28-3 from B. burgdorferi and then compared the ability of the mutated strain and its isogenic parental strain to infect mice, disseminate, and be taken up by ticks (acquisition) (54). The lp28-3-deficient strain disseminated in mice, but its acquisition by feeding ticks was inefficient, suggesting that proteins encoded by lp28-3 increase fitness. Coleman et al. directly tested the potential requirement for CspZ by deleting the gene and testing the mutant for its ability to infect mice (55). The cspZ deletion mutant was fully infectious in mice. The data presented here, the non-universal distribution of lp28-3, and gene deletion studies (cited above) indicate that CspZ is not required for complement evasion or infection of mammals and ticks.

The immunoblot analyses revealed that all clones cultivated in vitro, except DRI85A, produced one or more OspE proteins. Significant variation in the MW and number of OspE paralogs among clones was noted. ELISAs were conducted to indirectly assess expression in vivo. Antibodies to one or more OspE proteins were detected in 46%, 24%, and 50% of the human, dog, and coyote serum. The apparent absence of OspE and OspF proteins, encoded by cp32s, prompted us to determine the genome sequence of DRI85A. The genome sequence was determined using PacBio sequencing. The atypical cp32 content of DRI85A is described above. In that, DRI85A is a natural and non-genetically manipulated strain, it can be concluded that not all cp32s, including those that carry ospE and ospF genes, are not required for a strain to circulate in nature.

Many FHBPs selectively bind FH derived from different mammalian species (11). Since all clones analyzed in this study were recovered from dogs, we sought to determine if they display enhanced or specific binding to canine FH. Serum from dogs and eastern coyotes served as the FH source, with human serum as the comparator. Surprisingly, the CspZ and CspA proteins did not bind to canine FH and only weakly bound to some OspE proteins. In contrast, human FH is bound to CspA, CspZ, and most OspE proteins. The lack of binding of canine FH to CspA and CspZ suggests that these proteins do not contribute to FH-mediated complement evasion in dogs and, most likely, in other mammals.

Based on variable FHBP protein profiles, we speculated that clonal populations may differ in their sensitivity to dog and human complement. The serum sensitivity of 10 clones to complement-certified dog and human serum (0%, 40%, and 80%) was assessed. As detailed in the Results section and Table 2, serum sensitivity differed significantly among clones. A matching set of cells was incubated with the heat-inactivated serum to determine if the killing was complement-dependent. All clones tested were unaffected by exposure to 80% heat-inactivated serum, indicating that the serum-associated killing was due to complement. A correlation between FHBP expression (inferred from the immunoblots; Fig. 1) and FH binding (overlay assays; Fig. 2) with serum sensitivity/resistance was not observed. Previous studies have classified B. burgdorferi, Borreliella afzelii, and Borreliella garinii (the three primary causative agents of LD) as having high, intermediate, or low resistance to complement (56). However, as demonstrated here, serum sensitivity varies significantly among clones of a given species. Hence, we suggest that generalizations about serum resistance at the species level (23) are invalid and that serum sensitivity must be assessed at the clonal level.

The ability of DRI85A to survive in human and canine serum suggests that ospE and ospF, as well as most of the cp32s, are not required for infection. While DRI85A was recovered from an infected dog and is a low passage, it is possible that cp32s could have been lost during post-isolation cultivation. Its ability to infect mice was tested to verify the infectious phenotype of DRI85A. Spirochetes (10⁵) were delivered by needle injection, and 17 days later, the mice were euthanized, and biopsies and blood were collected. The biopsies were placed in media with antibiotics in an attempt to recover spirochetes. Clone DRI85A was cultured from the ear and/or bladder of four of six mice and elicited seroconversion against recombinant VlsE and DbpB. Notably, the antibody response in mice infected with B31 and DRI85A to VlsE variants was strain-specific. Sequence divergence between the two VlsE variants is found throughout the protein, including the C6 domain. A polypeptide corresponding to the C6 region is widely used as a Lyme disease diagnostic peptide (57). Like our findings, Baum et al. (58) also reported on the variant-specific immune responses to VlsE. These findings have implications regarding the reliability of the C6 peptide in diagnosing infection with, or exposure to, Borreliella species.

In summary, the analyses conducted in this study do not support a correlation between the production of FHBPs and FH binding or serum resistance. In addition, the limited and polymorphic nature of the cp32s in DRI85A indicates no strict requirement for a defined set of these replicons for infectivity. This conclusion is supported by a recent study by Wachter et al. in which a B31-derived clone that was experimentally cured of all cp32s, infected mice, and ticks (59). The results of this study call into question assumptions that we and others have made about the biological functions of CspZ, OspE, and the OspF proteins, their potential utility as vaccine candidates, and the role of cp32-encoded proteins in the enzootic cycle. Lastly, it is likely that there are complement evasion mechanisms employed by the LD spirochetes that are yet to be defined.