Complement therapeutic Factor H-IgG proteins as pre-exposure prophylaxes against Lyme borreliae infections

Abstract

Lyme disease (LD) is the most common vector-borne disease in the northern hemisphere and is caused by the bacteria Borrelia burgdorferi sensu lato (also known as Lyme borreliae) with no effective prevention available. Lyme borreliae evade complement killing, a critical arm of host immune defense, by producing outer surface proteins that bind to a host complement inhibitor, factor H (FH). These outer surface proteins include CspA and CspZ, which bind to the 6^th and 7^th short consensus repeats of FH (SCR(6–7)), and the OspE family of proteins (OspE), which bind to the 19^th and 20^th SCR (SCR(19–20)). In this study, we produced two chimeric proteins, FH-Fc, containing the Fc region of immunoglobulin G (Fc) with SCR(6–7) or SCR(19–20). We found that both FH-Fc constructs killed B. burgdorferi via bacterial lysis and phagocytosis and reduced bacterial colonization and LD-associated joint inflammation in vivo. While SCR(6–7)-Fc displayed Lyme borreliae species-specific bacterial killing, SCR(19–20)-Fc versatilely eradicated all tested bacterial species/strains. This correlated with SCR(6–7)-Fc binding to select variants of CspA and CspZ, but SCR(19–20)-Fc binding to all tested OspE variants. Overall, we demonstrated the concept of using FH-Fc constructs to kill Lyme borreliae and defined underlying mechanisms, highlighting the potential of FH-Fc as a pre-exposure prophylaxis against LD infection.

INTRODUCTION

Lyme disease is the most common vector-borne disease in the northern hemisphere, and the disease incidence is escalating; the CDC estimates more than 476,000 annual cases in the United States, and approximately 10,000 cases are reported each year in Europe (1–3). Transmitted by Ixodes ticks, Lyme disease is caused by more than 21 species of spirochete bacteria, collectively named Borrelia burgdorferi sensu lato (also known as Borreliella burgdorferi, B. burgdorferi s.l., or Lyme borreliae)(4). Humans can be infected by selected B. burgdorferi s.l. species, including B. burgdorferi sensu stricto (hereafter B. burgdorferi) prevalent in both North America and Eurasia, and B. afzelii, as well as B. bavariensis, and B. garinii isolated in Eurasia (5). Each of these species have evolved into multiple genetically distinct “strains” defined by different genotyping methods, such as the sequences of a Lyme borreliae polymorphic gene, ospC (6). These Lyme borreliae species and strains differ in their associated manifestations and the incidence of human cases (5). Following a tick bite, the bacteria disseminate through the bloodstream from the bite site on the skin to multiple tissues and organs, causing manifestations such as arthritis, neuroborreliosis, carditis, and acrodermatitis (7–11). Despite the continuing rising cases, geographical expansion of prevalence, and diversity of causative agents, no effective human LD preventive is currently available.

The survival of Lyme borreliae in humans and reservoir animals requires the ability to overcome different arms of the vertebrate host immune response in the blood, among which the first-line defense is the complement system (12–14). Complement can be activated through three canonical pathways: the classical and lectin pathways are initiated by the localization of antigen-antibody complexes and mannose-binding lectin (MBL) on the surface of microbes, respectively; the alternative pathway is initiated by C3b binding to the microbial surface (Fig. 1A) (15). Activation results in the formation of either of two C3 convertase enzymatic complexes: (1) C4b2a, whose assembly is triggered via the classical or lectin pathways, and (2) C3bBb, enzymatic complexes, the assembly of which is triggered via the alternative pathway (Fig. 1A). Both C3 convertases induce the release of proinflammatory peptides (C3a, C5a), the deposition of opsonins (iC3b) on the microbial surface, and, by recruiting other complement proteins, generate C5 convertases. The latter catalyzes the formation of the membrane attack complex, C5b-9, for pathogen lysis (Fig. 1A). The proinflammatory nature of the complement cascade necessitates tight control of this potentially destructive immune defense. Indeed, the host encodes regulators of complement activation (RCA), proteins that modulate each of the three activation pathways (16). For example, factor H (FH), which contains 20 short consensus repeats (SCRs), binds to and triggers the degradation of C3b to inhibit C3b-containing convertases generated by the alternative pathway (17, 18) (Fig. 1A).

Figure 1. Schematic diagram showing the composition of the FH-Fc constructs used in this study and their proposed mechanisms of action to eliminate Lyme borreliae.

(A) FH-Fc is proposed to promote B. burgdorferi s.l. killing by binding to the FH-binding proteins of Lyme borreliae via the FH region to prevent the pathogens from escaping alternative complement pathway-mediated pathogen killing. Simultaneously, the Fc region of FH-Fc may recruit the complement C1 components (C1) required for classical complement pathway-mediated pathogen killing. Lyme borreliae produce CspA, CspZ, and OspE to bind to FH in promoting alternative pathway evasion. Depicted are two FH-Fc constructs, S2635 and S2782, binding to CspA, CspZ, and/or OspE to prevent bacterial alternative pathway evasion and interacting with C1 to promote classical pathway activation. This panel was adopted from our previous publication (85). (B) Shown are the diagrams for the composition of S2635 and S2782: S2635 contains the SCR(6–7) region of human FH, a linker, the hinge region of the human IgG1, and the CH2-CH3 region of the Fc region from human IgG3 in sequential order from N to C terminus. S2782 contains the hinge region of the human IgG1, the CH2-CH3 region of the Fc region from human IgG3, and the SCR(6–7) region of human FH in sequential order from N to C terminus. The mutated amino acids at the C-terminal Fc region have been highlighted in bold. The amino acid sequences of S2635 and 2782 are indicated in Text S1 and S2, respectively.

To survive in the blood where complement is mainly located, Lyme borreliae produce RCA-binding proteins to bind and recruit complement regulatory proteins on the spirochete surface to inactivate complement (13, 19). One group of these proteins binds to FH(20). Among these FH-binding proteins, CspA (also known as Complement Regulator Acquiring Surface Protein 1 or CRASP-1) and CspZ (also known as CRASP-2) both bind to SCR(6–7) of FH (21, 22). The other proteins, belonging to an OspE protein family (also known as CRASP3–5), bind to SCR(19–20) of FH (21, 23, 24). During the enzootic cycle, CspA is expressed mostly when bacteria are in ticks prior to and during tick-to-vertebrate host transmission, whereas CspZ and OspE are largely produced while bacteria are in the vertebrate hosts after transmission (25). While the role of OspE in the enzootic cycle remains undefined, CspA and CspZ confer bacterial evasion to complement in ticks’ bloodmeal to facilitate tick-to-host transmission and in the host bloodstream for efficient dissemination, respectively (26–28). These findings thus led to the concept that targeting these FH-binding functions might serve as an intervention against Lyme disease.

In fact, efforts have been made to test this concept by generating several complement-targeted therapeutic candidates, including FH-Fc (29). FH-Fc are recombinant fusion proteins containing SCR(6–7) or SCR(19–20) that bind to the FH-binding proteins from pathogens, as most pathogens’ FH-binding proteins target those FH regions. By displacing FH on the pathogen surface, FH-Fc are intended to prevent alternative complement pathway evasion. FH-Fc also contains the Fc region of immunoglobulins, allowing the activation of classical pathway-mediated pathogen killing (29). In support of this concept, FH-Fc has been demonstrated to efficiently eliminate multiple bacterial or parasite species, such as Neisseria gonorrhoeae, Neisseria meningitidis, Streptococcus pyogenes, Haemophilus influenza, Trypanosoma cruzi (30–32, 33, 34).

In this study, we produced two FH-Fc constructs using Nicotiana benthamiana (tobacco plant expression system), a platform that can rapidly produce large amounts of foreign proteins for pharmaceutical use (35). We tested the ability of these FH-Fc constructs to kill different species or strains of B. burgdorferi s.l. in vitro and to prevent Lyme-associated bacterial colonization and manifestations using murine models. We also attempted to determine the mechanisms underlying the FH-Fc-mediated Lyme borreliae killing by measuring the binding affinity of FH-Fc constructs to bacterial FH-binding proteins to investigate the potential of FH-Fc as a Lyme disease prophylaxis.

MATERIALS AND METHODS

Ethics Statement.

All mouse experiments were performed in strict accordance with all provisions of the Animal Welfare Act, the Guide for the Care and Use of Laboratory Animals, and the PHS Policy on Humane Care and Use of Laboratory Animals. The protocol (Docket Number 22-451) was approved by the Institutional Animal Care and Use Committee of Wadsworth Center, New York State Department of Health. All efforts were made to minimize animal suffering.

Mouse, ticks, bacterial strains, and human serum.

Four-week-old female C3H/HeN mice were purchased from Charles River (Wilmington, MA). BALB/c C3-deficient mice were from in-house breeding colonies (26) and Ixodes scapularis tick larvae were obtained from BEI Resources (Manassas, VA). Escherichia coli strains BL21(DE3), M15 or DH5α and their derivatives were grown at 37°C or other appropriate temperatures in Luria-Bertani broth or agar, supplemented with kanamycin (50μg/mL) or ampicillin (100μg/mL) (Table S1). B. burgdorferi strain B31-5A4 (Table S1) was grown at 33°C in BSK II complete medium (36). Cultures of B. burgdorferi B31-5A4 were tested with PCR to ensure a full plasmid profile before use (37, 38). The remaining strains used in this study were kept within ten passages to prevent potential plasmid loss. A. tumefaciens GV3101 (pMP90RK) containing the binary vector pTRAkc-P19, encoding the post-transcriptional silencing suppressor P19, and each of the FH-Fc constructs was used for transient expression using an N. benthamiana expression system as indicated (Table S1). Uninfected human serum (CompTech, Tyler, TX) was confirmed as seronegative for Lyme disease infection as described (28).

Expression and purification of FH/Fc fusion proteins in tobacco plants

Nucleotide sequences encoding human FH SCR6–7 (aa residues 321–443 (Genbank #: NP_000177)) and human FH SCR19–20 (aa residues 1048–1231 (Genbank #: NP_000177)), incorporating the D1119G mutation (39)), designed to employ optimal codon usage for expression in N. benthamiana, were synthesized by GENEWIZ (South Plainfield, NJ). Similarly, nucleotide sequences encoding human CH2-CH3 domains from IgG3 (aa residues 130–346 (Genbank #: CAA67886.1)) were also synthesized for optimal codon usage for expression in N. benthamiana, by GENEWIZ. In S2635, the SCR6–7 and the human CH2-CH3 domains from IgG3 were placed at N- and C-terminus, respectively, connected with a linker (GGGGSGGGGSGGGGSS), followed by the IgG1 hinge sequence (EPKSCDKTHTCPPCP) (Fig. 1B, Text S1). To generate S2782, a portion of the IgG1 hinge sequence (DKTHTCPPCP), followed by the human CH2-CH3 domains from IgG3, was placed at the N-terminus, whereas the SCR19–20 was placed at the C-terminus. We did not add a flexible linker sequence in S2782. Note that in S2782, the C-terminal three residues of Fc, PGK, were replaced by the residues GQC to facilitate construct stability and resistance to protease cleavage (Fig. 1B, Text S2) (40). These synthetic sequences of S2635 and S2782 were then placed downstream of the signal peptide of the murine mAb24 heavy-chain (lph) (Fig. 1B, Text S1 and S2) (41). The entire synthetic sequences were cloned into the plant binary expression vector pTRAkc (PMID: 17412974).

These recombinant proteins were then produced via transient expression by whole-plant vacuum infiltration of N. benthamiana ΔXT/FT using A. tumefaciens GV3101 and pMP90RK vector, as described previously (42–44). We then purified and concentrated S2635 and S2782 using Protein A-MabSelect SuRe or PrismA affinity columns (GE HealthCare) as described (45). Protein concentrations were quantified using a UV spectrophotometer for the absorption at 280 nm and extinction coefficients predicted from the mature amino acid sequences (excluding signal peptides).

Cloning, transfection, expression and purification of CspA, CspZ, and OspE variants from Lyme borreliae.

DNA encoding histidine or glutathione-S-transferase (GST) tagged CspA, CspZ, and OspE variants from different Lyme borreliae strains or species (Table S2) was used to express and purify these proteins using an E. coli expression system as described previously (22, 26–28, 46–48). The DNA encoding both N-terminal histidine- and GST-tagged proteins were synthesized by Synbio Technologies (Monmouth Junction, NJ), followed by subcloning into the pET28a (Millipore Sigma, Burlington, MA), pQE30Xa (Qiagen, Germantown MD), and pGEX4T2 vectors (Cytiva, Marlborough, MA), respectively, via BglII/BamHI restriction sites, using the service from Synbio Technologies (Monmouth Junction, NJ) (Table. S2). After transforming the pET28a- or pGEX4T2-associated plasmids into E. coli strain B21(DE3) or the pQE30Xa-associated plasmids into E. coli strain M15, as described (26), the protein was purified as in our previous work (26).

Mouse infection.

Flat I. scapularis nymphs carrying B. burgdorferi strain B31-5A4 were generated as described previously using BALB/c C3-deficient mice (26, 49). These C3H/HeN mice were intramuscularly injected with PBS buffer (control) or 0.2 mg/kg of S2635 or S2782. Twenty-four hours later, they were infected by placing nymphs on the mice. The ticks were allowed to feed until repletion. At 21 dpf, the above-mentioned flat and replete ticks and mouse tissues were then used to determine bacterial burdens, and the tibiotarsus joints were used to determine the severity of arthritis in the section “Quantification of spirochete burdens and histological analysis of arthritis.” Mouse sera were utilized to define the Lyme disease bacterial seropositivity as described in the section “ELISAs.”

Quantification of spirochete burdens and histological analysis of arthritis.

DNA was extracted from the indicated mouse tissues to determine bacterial burdens using quantitative PCR analysis as described (50). Note that spirochete burdens were quantified based on the amplification of recA using forward (GTGGATCTATTGTATTAGATGAGGCTCTCG) and reverse (GCCAAAGTTCTGCAACATTAACACCTAAAG) primers. The number of recA copies was calculated by establishing a threshold cycle (Cq) standard curve of a known number of recA genes extracted from strain B31-5A4, and burdens were normalized to 100 ng of total DNA. For the ankles that were applied to histological analysis of Lyme disease-associated arthritis, the analysis was performed as described (50). Images were scored based on the severity of inflammation on a scale of 0 (no inflammation), 1 (mild inflammation with less than two small foci of infiltration), 2 (moderate inflammation with two or more foci of infiltration), or 3 (severe inflammation with focal and diffuse infiltration covering a large area).

ELISAs.

Seropositivity of the mice after infection with B. burgdorferi was determined by detecting the presence or absence of IgG recognizing C6 peptides, as described previously (51), as this methodology has been commonly used for human Lyme disease diagnosis (52). The maximum slope of optical density/minute of all the dilutions was multiplied by the respective dilution factor, and the greatest value was used as representative of anti-C6 IgG titers (arbitrary unit (A.U.)). Seropositive mice were defined as mice with sera yielding a value greater than the threshold, the mean plus three-fold standard deviation of IgG values derived from uninfected mice.

Borreliacidal assays.

The ability of S2635, S2782, or the combination of both variants to kill different Lyme borreliae species or strains was determined as described with modifications (47, 50). Briefly, S2635, S2782, or one to one ratio of S2635 and S2782 were incubated, at different concentrations, with the Lyme borreliae. We then mixed the FH-Fc-Lyme borreliae with complement-preserved human serum (CompTech, Tyler TX) at a final concentration of 40% (20% for B. garinii strains ZQ1 and PBr, as that was the maximal concentration that did not result in bacterial killing in the absence of FH-Fc (Fig. S1)). The mixture was incubated at 33°C for 24 hours. Surviving (motile) spirochetes were quantified by direct counting using dark-field microscopy and expressed as the proportion of S2635- or S2782-treated to untreated Lyme borreliae (those exposed to complement-preserved human serum only). The concentration of S2635 or S2782 that killed 50% of spirochetes (EC₅₀) was calculated using dose-response stimulation fitting in GraphPad Prism 9.3.1. Note that this approach to measure bactericidal activity is based on counting the bacteria visually from three fields per replicate, per condition. The variation of the number of bacteria would vary in different fields (i.e., more bacteria in one field and less bacteria in other fields). This may result in the mean percentage slightly greater or less than 100% in some data points of the BSA treatment groups.

Phagocytosis assays

The phagocytosis assays were performed as described previously with modifications (53). Lyme borreliae were labeled with carboxyfluorescein diacetate succinimidyl ester (CFSE, ThermoFisher Scientific) as described in the vendor’s manual. The suspension of spirochetes (10⁷) in BSK II media without rabbit sera, gelatin, and BSA was incubated with 3.3uM of CFSE at room temperature for 10 minutes. Untreated or heat-inactivated human sera were incubated with different strains of CFSE-labeled spirochetes (10⁷ bacteria) in the presence of S2635 or S2782 (2μM in PBS) at room temperature for 10 minutes. Such spirochete suspension was then mixed with freshly isolated human Polymorphonuclear neutrophils (PMNs) from a blood donor from iQBioscience (Alameda, CA) at the ratio of 25 to 1 in the presence of human sera (5%). After shaking at 37°C, 50 rpm for 10 minutes, the bacteria-PMNs mixture was incubated on ice for 10 minutes. The bacteria-PMN mixture that was immediately placed on ice after mixing was included as a control. Phagocytosis was stopped by transferring the bacteria-PMN mixtures to ice-cold Fluorescence-Activated Cell Sorting (FACS) buffer (PBS supplemented with 0.5% bovine serum albumin (BSA), 0.01% NaN3 and 0.35 mM EDTA) and stored at 4°C. Samples continually kept at 4°C were used as a control. PMNs were then washed and suspended with ice-cold FACS buffer before being applied to a Beckman Coulter Cytoflex Flow Cytometer (Beckman Coulter, Indianapolis, IN) and analyzed by FlowJo software (FlowJo, Llc, Ashland, OR). The phagocytosis index of each sample was calculated as mean fluorescence intensity (MFI)×percentage (%) positive cells) at 37°C minus (MFI×% positive cells) at 4°C. Each sample was tested in four independent experiments.

FH-Fc-binding assay of Lyme borreliae detected by Flow Cytometry.

The quantitative determination of Lyme borreliae binding to FH-Fc constructs using flow cytometry as described with modifications (26). S2635, S2782, or human IgG3 (control; ThermoFisher Scientific) (2μM in PBS) was incubated with tested Lyme borreliae species or strains (1 × 10⁷ bacteria per reaction) for one hour. After the bacteria were washed using PBS three times, the bacteria were suspended in HBSC buffer containing DB (25 mM Hepes acid, 150 mM sodium chloride, 1 mM MnCl₂, 1 mM MgCl₂, 0.25 mM CaCl₂, 0.1% glucose, and 0.2% BSA, final concentration and 100 μl per reaction). An FITC conjugated Goat F(ab’)2 anti-Human IgG Fc (ThermoFisher) (1: 250x) was added for staining the FH-Fc bound bacteria, followed by the treatment of 300μl formalin (0.1%) for fixing. The MFI values of Lyme borreliae were measured, and the FH-Fc- vs. IgG control-derived MFI values were compared to determine the FH-Fc binding capability of Lyme borreliae using flow cytometry as described in the section “Phagocytosis assays.”

Surface plasmon resonance (SPR) analyses.

The interactions of recombinant CspA, CspZ, or OspE proteins with S2635 or S2782 were determined using a Biacore T200 (Cytiva), similar to the work in our previous studies (54). Basically, 10 micrograms of human S2635 or S2782 were conjugated to a Protein A chip (Cytiva). Quantitative SPR experiments were used to determine the binding kinetics of the CspA, CspZ, and OspE variants that display FH-Fc binding activity. Basically, 10 μl of increasing concentrations (0.0625, 0.125, 0.25, 0.5, 1 μM, and/or 2 μM) of CspA, CspZ, or OspE proteins were injected into the control cell, and the flow cell immobilized with S2635 or S2782 at 30 μl/min in PBS at 25°C. To obtain the kinetic parameters of the interaction, sonogram data were fitted using BIAevaluation software version 3.0 (GE Healthcare), employing the one-step biomolecular association reaction model (1:1 Langmuir model), resulting in optimum mathematical fit with the lowest Chi-square values.

Statistical analyses.

Significant differences were determined with a Kruskal-Wallis test with the two-stage step-up method of Benjamini, Krieger, and Yekutieli (55) and a two-tailed Fisher test (for seropositivity) (56) using GraphPad Prism 9.3.1. A p-value < 0.05 was used to determine significance.

RESULTS

S2635 and S2782 were constructed and produced in Nicotiana benthamiana.

To generate the SCR6–7 and SCR19–20 versions of FH-Fc constructs (S2635 and S2782, respectively), we obtained the plant codon-optimized DNA of those FH domains. The SCR6–7 sequence in S2635 was then connected at the N-terminal end of the CH2-CH3 domains (Fc) of human IgG3, with SCR6–7 and Fc separated by a flexible linker (GGGGSGGGGSGGGGSS), followed by a portion of the IgG1 hinge sequence (EPKSCDKTHTCPPCP) (Fig. 1B, Text S1). IgG3 was used because IgG3 is documented to have superior classical pathway-mediated antibody killing, compared to other IgG isotypes(57). We introduced a R435H mutation into the Fc region of IgG3 (allotype IGHG1*04) in both S2635 and S2782, as this mutation both increases the half-life and enhances the ability of Fc to bind to protein A, facilitating purification of FH-Fc (58). The N-terminus of S2782 starts with a portion of IgG1 hinge sequence (DKTHTCPPCP) and the human CH2-CH3 domains from IgG3, with a small change in which the C-terminal three residues, PGK, were replaced by the GQC (Fig. 1B, Text S2). This change was intended to promote the formation of an inter-chain disulfide bond between paired CH3 domains to stabilize the construct against aggregation by low pH (40). The plant codon-optimized DNA sequence encoding human SCR19–20 was then appended to the C-terminal end of the IgG3 Fc (Fig. 1B, Text S2). Both S2635 and S2782 were then produced using a rapid N. benthamiana expression system and purified by protein A affinity chromatography (32). The resulting yields of purified S2635 and S2782 were 296 ±23 and 522 ±172 mg/kg, respectively.

S2635 and S2782 differ in their ability to eliminate B. burgdorferi in fed nymphs, but both significantly reduced bacterial colonization and joint inflammation.

We tested the ability of S2635 or S2782 to impact bacterial colonization and Lyme disease-associated manifestations in mice. One day prior to nymphal tick feeding (−1 dpf), we injected mice with 0.2 mg/kg of S2635, S2782 or PBS (control) intramuscularly, as immunoglobulins and Fc-fusions are highly bioavailable when introduced by this route (59)) (Fig. 2A). At 24-h after the injection, we allowed I. scapularis nymphal ticks carrying B. burgdorferi strain B31-5A4 to feed on the mice. Uninfected mice (PBS-pre-treated mice without tick feeding) were also included as a control (Fig. 2A). We first measured the bacterial burdens in replete nymphs after engorgement and found that the nymphs feeding on S2635- but not S2782-pre-treated mice had significantly lower burdens than those feeding on PBS-pre-treated mice. These results suggest the ability of S2635 to uniquely eliminate the bacteria in feeding ticks during tick-to-host transmission (Fig. 2B).

Figure 2. S2635 and S2782 protected mice from Lyme borreliae-associated colonization, seropositivity, and arthritis but differed in the ability to eliminate spirochetes in fed nymphs.

(A) Shown is the timeframe of FH-Fc inoculation and Lyme borreliae infection in this study. (B to H) Five C3H/HeN mice were intramuscularly injected with S2635 or S2782 at the dose of 0.2 mg/kg or PBS (control). At 24 hours after inoculation (1 day prior to tick feeding (dpf)), these mice were fed on by I. scapularis nymphs carrying B. burgdorferi B31-5A4 (Bb B31-5A4). An additional five mice injected with PBS but not fed on by ticks were included as the control (Uninfect.). (B) The engorged fed nymphs were collected from those mice at 4 dpf. (C) At 21 days, sera were collected from these mice to determine the seropositivity to Lyme disease infection by evaluating the IgG levels of C6 antigens. Spirochete burdens at (D) the tick feeding site (“Bite Site”), (E) bladder, (F) heart, and (G) knees were quantitatively measured at 21 dpf, shown as the number of spirochetes per 100ng total DNA. Data shown are the geometric mean ± geometric standard deviation of the spirochete burdens from five mice per group. Statistical significances (p < 0.05, Kruskal-Wallis test with the two-stage step-up method of Benjamini, Krieger, and Yekutieli) of differences in bacterial burdens relative to (*) uninfected mice are presented. (H) Tibiotarsus joints at 21dpf were collected to assess inflammation by staining these tissues using hematoxylin and eosin. Representative images from one mouse per group are shown. The top panels are lower-resolution images (joint, ×10 [bar, 160 μm]); the bottom panels are higher-resolution images (joint, 2×20 [bar, 80 μm]) of selected areas (highlighted in top panels). Arrows indicate infiltration of immune cells. (Inset figure) To quantify inflammation of joint tissues, at least ten random sections of tibiotarsus joints from each mouse were scored on a scale of 0–3 for the severity of arthritis. Data shown are the mean inflammation score ± standard deviation of the arthritis scores from each group of mice. Asterisks indicate the statistical significance (p < 0.05, Kruskal Wallis test with the two-stage step-up method of Benjamini, Krieger, and Yekutieli) of differences in inflammation relative to uninfected mice.

At 21-dpf we measured the spirochete burdens in mouse tissues and Lyme borreliae seropositivity (i.e., the IgG levels against a C6 peptide derived from a Lyme borreliae VlsE antigen, a commonly used biomarker for Lyme disease serodiagnosis (52)) (Fig. 2A). PBS-pre-treated mice yielded significantly greater levels of seropositivity (five out of five turning seropositive) than uninfected mice (Fig. 2C). In contrast, only one out of five S2635- or S2782-pre-treated mice turned Lyme borreliae seropositive, which is statistically indistinguishable from uninfected mice (Fig. 2C). Additionally, compared to uninfected mice, all 5 PBS-pre-treated mice had significantly higher bacterial burdens at tick bite sites (skin) and in bladder, heart, and knee joints, whereas the same tissues from only 1 out of 5 mice pre-treated with S2635 or S2782 had detectable bacterial burdens (Fig. 2D to G). These results indicate the ability of S2635 and S2782 to reduce bacterial colonization after B. burgdorferi tick-to-host transmission.

We further determined the Lyme disease-associated manifestations in those mice at 21-dpf by histological analysis of the ankle joints. In PBS-pre-treated mice, we observed elevated levels of infiltrations of granulocytes and mononuclear cells, such as neutrophils and monocytes in the connective tissues, tendons, and muscles (Fig. 2H). That resulted in significantly greater levels of inflammation scores in PBS-pre-treated mice than those in uninfected mice (Fig. 2H, inset figure). However, we did not observe such noticeable cell infiltrations in the joints of S2635- and S2782-pre-treated mice, agreeing with their indistinguishable inflammatory scores, compared to the scores of uninfected mice (Fig. 2H). Overall, these findings indicated that S2635 or S2782 pre-treatment prior to nymph transmission of B. burgdorferi B31-5A4 efficiently decreased spirochete colonization and Lyme disease-associated joint inflammation.

In vitro lysis of Lyme borreliae by S2635 and S2782.

The bactericidal mechanisms of FH-Fc constructs have been proposed through their ability to 1) prevent Lyme borreliae escape from FH-binding-dependent, alternative pathway-mediated complement evasion and 2) promote Fc-mediated bacterial killing. Both abilities can lead to Lyme borreliae lysis and phagocytosis in vitro. We thus first tested whether S2635 and S2782 can promote Lyme borreliae lysis by incubating human serum (the source of complement) with strain B31-5A4 in the presence of different concentrations of S2635, S2782, or BSA (control). FH-Fc constructs are proposed to eradicate spirochetes by promoting classical pathway-mediated killing via their Fc regions and preventing alternative pathway evasion by displacing FH from the bacteria. Therefore, 40% human serum was used as the final concentration in the reaction because this percentage allows for the observation of Lyme borreliae lysis by both pathways. However, B. garinii strains ZQ1 and PBr were incapable of surviving this concentration of human serum (Fig. S1). We thus titrated the serum and found 20% as the minimal concentration that these two strains can survive and thus incubated these strains in 20% human serum for this experiment (Fig. S1). We then quantified the bacteria present microscopically after 4-h of incubation and then normalized those numbers to those before incubation. Such normalization permitted us to measure the percent survival of bacteria, calculating the EC₅₀ values (the concentration of FH-Fc constructs that lead to 50% bacterial survival). While BSA-incubated B. burgdorferi B31-5A4 yielded close to 100% bacterial survivability, S2635 and S2782 incubations resulted in spirochete lysis, with S2635 having significantly more robust lysis than S2782 based on their EC₅₀ values (Fig. 3A and Table S3).

Figure 3. S2635 and S2782 varied in their breadth of killing to different Lyme borreliae species and strains.

S2635, S2782, or BSA (control) or PBS (control, data not shown) were serially diluted as indicated, mixed with human serum, and incubated with the following Lyme borreliae species and strains (5 × 10⁵ cells ml⁻¹): B. burgdorferi (Bb) strains (A) B31-5A4 and (B) 297, B. afzelii (Ba) strains (C) VS461 and (D) PKo, B. bavariensis (Bbav) strain (E) PBi, and B. garinii (Bg) strains (F) ZQ1 and (G) PBr. The final concentration of the human serum was 40%, except that of B. garinii strains, which was 20%. After incubating for 24 hours, surviving spirochetes were quantified from three fields of view for each sample using dark-field microscopy. The work was performed on three independent experiments. The survival percentage was derived from the proportion of FH-Fc-treated to PBS-treated spirochetes. Shown is one representative experiment, and in that experiment, the data points are the mean ± SEM of the survival percentage from three replicates. The 50% borreliacidal activity of each FH-Fc (EC₅₀), representing the FH-Fc concentrations that effectively killed 50% of spirochetes, was obtained and extrapolated from curve-fitting and shown in Table S3. The EC₅₀ values are shown as the mean ± SD from three experiments.

The most prevalent human-caused Lyme borreliae species in Europe are B. afzelii, followed by B. garinii and B. bavariensis (5), whereas the most commonly identified ospC genotype in Lyme disease patients in North America after ospC type A (i.e., B. burgdorferi B31-5A4) is ospC type K (i.e., B. burgdorferi 297)(6, 60, 61). We thus also tested whether such bactericidal activity of S2635 and S2782 can be extended to the abovementioned Lyme borreliae species and strains and obtained EC₅₀ values for those bacteria in the same fashion. Our results showed three patterns of bacterial lysis depending on the Lyme borreliae species or strains incubated with S2635 or S2782: 1) Similar to B. burgdorferi B31-5A4, when incubated with B. burgdorferi strain 297 or B. afzelii strains VS461 or PKo, both S2635 and S2782 efficiently killed those strains. However, S2635 eradicated these strains more efficiently than S2782 (Fig. 3B to D and Table S3); 2) When incubated with B. bavariensis strain PBi, although both S2635 and S2782 can eliminate this strain, S2782 showed more significantly robust lysis than S2635 (Fig. 3E and Table S3); 3) When incubated with B. garinii strains ZQ1 and PBr, S2635 did not show bacterial lysis, but S2782 was found to kill those strains (Fig. 3F and G). These results support the versatility of S2782-mediated spirochete lysis in comparison to unique species-specific Lyme borreliae lysis by S2635.

In vitro phagocytosis of Lyme borreliae by S2635 and S2782.

We next examined the ability of S2635 and S2782 to promote bacterial phagocytosis by incubating carboxyfluorescein diacetate succinimidyl ester (CFSE)-labeled Lyme borreliae species and strains with human polymorphonuclear neutrophils (PMNs) in the absence or presence of S2635 or S2782. After resulting PMNs were applied to flow cytometry, we measured the mean fluorescence intensity (MFI) values of those cells that were converted to phagocytosis index as described in Materials and Methods for the determination of phagocytotic levels. We found that either S2635- or S2782-treatment resulted in significantly greater levels of phagocytosis than BSA (control)-treatment for B. burgdorferi B31-5A4 and 297, B. afzelii VS461 and PKo, and B. bavariensis PBi (Fig. 4A to E). However, for B. garinii ZQ1 and PBr, we found that S2782- but not S2635-treatment led to significantly higher levels of phagocytosis than BSA-treatment (Fig. 3F and G). These results indicate while S2782 facilitated phagocytosis universally among Lyme borreliae species/strains, the S2635-promoted phagocytosis is Lyme borreliae species-specific.

Figure 4. S2635 and S2782 promote phagocytosis of Lyme borreliae in a bacterial species-specific manner.

Human sera in the presence of BSA (control), S2635, or S2782 (2μM) were incubated with PMNs (2 × 10⁶ cells ml⁻¹) and indicated CFSE-labeled Lyme borreliae species or strains (5 × 10⁷ cells ml⁻¹). These species strains include: B. burgdorferi (Bb) strains (A) B31-5A4 and (B) 297, B. afzelii (Ba) strains (C) VS461 and (D) PKo, B. barvariensis (Bbav) strain (E) PBi, and B. garinii (Bg) strains (F) ZQ1 and (G) PBr. The resulting Lyme borreliae-PMN mixtures were applied to flow cytometry after 10 min of incubation at 37°C to obtain the mean fluorescence intensity (MFI) values. The MFI values derived from Lyme borreliae-PMN mixtures incubated at 4°C throughout the experiment are included as a control. The MFI values of each of the Lyme borreliae-PMN mixtures in 37°C and 4°C were used to calculate the “Phagocytic index” as described in “Materials and Methods” to quantitatively show the levels of phagocytosis. This assay was performed in four independent determinations. One representative experiment is shown. Inset figures represent the geometric mean of the phagocytic index from four independent determinations ± geometric SD. Asterisks indicate the statistical significance (p < 0.05, Kruskal Wallis test with the two-stage step-up method of Benjamini, Krieger, and Yekutieli) of differences in phagocytic index relative to BSA-treated Lyme borreliae.

Binding of S2635 and S2782 to Lyme borreliae.

We hypothesized that the in vitro killing potency of S2635 and S2782 would depend on their binding ability to pathogens. We thus applied S2635-, S2782-, or human IgG (control)-treated Lyme borreliae to flow cytometry and compared resulting MFI values on the spirochete surface to determine the ability of Lyme borreliae in binding to those FH-Fc constructs. We found that the S2782 treatment resulted in significantly greater levels of MFI values than IgG treatment for all tested Lyme borreliae species of strains (Fig. 5A to G), showing the universal binding ability of S2782 to all tested spirochetes. However, S2635 treatment led to significantly higher levels of MFI values than IgG treatment only for B. burgdorferi B31 and 297, as well as B. afzelii VS461 and PKo, and B. bavariensis PBi (Fig. 5A to E). The treatment of S2635 yielded indistinguishable levels of MFI values from IgG treatment for B. garinii ZQ1 and PBr (Fig. 5F and G), indicating the Lyme borreliae species-specific and S2635-binding activity.

Figure 5. S2635 and S2782 bound to Lyme borreliae in a bacterial species-specific manner.

B. burgdorferi (Bb) strains (A) B31-5A4 and (B) 297, B. afzelii (Ba) strains (C) VS461 and (D) PKo, B. bavariensis (Bbav) strain (E) PBi, and B. garinii (Bg) strains (F) ZQ1 and (G) PBr were incubated with either irrelevant human IgG (hIgG, control), S2635 or S2782 (2μM). The bacteria were stained with a FITC-conjugated goat F(ab’)2 anti-human IgG Fc before being applied to flow cytometry analysis. Representative histograms of flow cytometry analysis showing the levels of hIgG, S2635, or S2782 binding to indicated Lyme borreliae species or strains. (Inset figures) hIgG, S2635, or S2782 binding to Lyme borreliae was measured by flow cytometry and presented as MFI. Each bar represents the geometric mean of four independent determinations ± geometric SD. Asterisks indicate the statistical significance (p < 0.05, Kruskal Wallis test with the two-stage step-up method of Benjamini, Krieger, and Yekutieli) of differences in MFI values relative to hIgG-treated Lyme borreliae.

Binding of S2635 and S2782 to purified Lyme borreliae FH-binding proteins.

We next tested if S2635 and S2782 bind to documented Lyme borreliae FH-binding proteins (CspA, CspZ, and OspE (20)) in similar patterns as their ability to bind to Lyme borreliae. We thus produced recombinant forms of these Lyme borreliae FH-binding proteins from four different species, B. burgdorferi, B. afzelii, B. bavariensis, and B. garinii (20). We produced one variant per FH-binding protein from each of the tested species based on sequences that are available in GenBank, with the exception that variants from two strains of B. burgdorferi (B31 and 297) were produced. Note that two CspA paralogs from B. bavariensis (Bga66 and Bga71) (62) and two OspE paralogs from B. burgdorferi strain B31 (ErpA and ErpP) have been shown to bind human FH (63, 64). Therefore, those variants were also included in the study. S2635 and S2782 were conjugated to SPR chips, which were used to measure binding affinity of each of the Lyme borreliae FH-binding proteins. We did not detect binding of any of the OspE variants to S2635 (Fig. 6A, top panel, Table S2), in agreement with prior findings that OspE does not bind to human SCR6–7 (65). However, we found that S2635 binds strongly to the CspA variants from B. burgdorferi strains B31 and 297 and B afzelii strain PKo (K_D = 1.1 to 2.2×10⁻⁷ M, Fig. 6B top panel, Table S4) but less efficiently to B. bavariensis strain PBi (K_D = 1.1×10⁻⁶ M, Fig. 6B top panel, Table S4). No binding of S2635 to the CspA variant of B. garinii strain ZQ1 was detected (Fig. 6B top panel, Table S4). Similarly, binding of CspZ to S2635 was species- and strain-specific; S2635 bound robustly to the CspZ variants from B. burgdorferi strain B31 and B. afzelii (K_D = 1.5 to 2×10⁻⁷ M, Fig. 6C top panel, Table S5), but less efficiently to the CspZ variant from B. bavariensis (K_D = 9.8×10⁻⁷ M, Fig. 6C top panel, Table S5). Additionally, S2635 did not bind to the CspZ variants from B. burgdorferi strain 297 or B. garinii (Fig. 6C top panel, Table S5). We also applied each of these CspA, CspZ, and OspE variants to the S2782-conjugated SPR chips. S2782 did not bind to any tested CspA or CspZ variants (Fig. 6B and C bottom panels, Table S1 and S4), consistent with the inability of these FH-binding proteins to bind to human SCR19–20 (22, 46). However, we detected the binding of all tested OspE variants to S2782, indicating the versatility of the S2782-binding ability of OspE (Fig. 6A bottom panel, Table S2) Taken together, these results indicate the S2782-binding versatility of SCR(19–20)-binding proteins (i.e., OspE) from all tested Lyme borreliae species and strains in contrast to S2635-binding specificity to SCR(6–7)-binding proteins (i.e., CspA and CspZ) of some spirochete species or strains.

Figure 6. S2635 and S2782 bound to Lyme borreliae factor H-binding protein variants in a variant-specific manner.

The indicated concentrations of the variants of (A) OspE, (B) CspA, or (C) CspZ from indicated strains B. burgdorferi (Bb), B. afzelii (Ba), B. garinii (Bg), or B. bavariensis (Bbav) were flowed in PBS buffer over the chip surface, conjugated with S2635 or S2782. Binding was measured in response units (R.U.) by SPR. The k_on, k_off, and K_D values were determined from the average of three experiments (Table S2, S4, and S5). Shown is one representative experiment.

DISCUSSION

As one of the major host defense mechanisms against infection, complement not only lyses pathogens but cross talks with different arms of the host immune response (66–68). Dysregulation of this defense mechanism thus often causes various autoimmune diseases and exacerbates infectious diseases. Manipulating complement regulation by targeting complement components or regulators has frequently been employed in developing therapies for such diseases (69). Specifically, to combat complement dysregulation-mediated infectious diseases, one of the most commonly used strategies is to target pathogens’ anti-complement proteins either by monoclonal antibodies or small molecules (70). However, targeting single pathogen proteins may risk pathogens developing “neutralization escape mutants”, leading to ineffective therapeutics (71). Additionally, the functional redundancy seen in pathogen anti-complement mechanisms makes this a complex strategy (72). An alternative approach is to target the host complement components or regulators that are involved in pathogen complement evasion to skew the complement responses toward pathogen elimination (73). Such therapeutics would be mimetics of the pathogen-binding domains of host complement regulators, preventing pathogens from hijacking host complement or eliciting complement regulators to inhibit complement activation (73). Some of these “complement-based” therapeutics have been produced as fusion proteins with the Fc of immunoglobulin, such as FH-Fc, to further enhance classical pathway-mediated pathogen killing (30–33). It should be more difficult for pathogens to evolve neutralization escape mutations to FH-Fc, as any mutations that reduce FH-Fc binding should have reduced FH binding, leading to more robust complement-mediated killing (30–33, 74). Moreover, as binding to complement regulator is a common approach shared by many pathogens to evade complement, complement modulation-targeted therapeutics (e.g., FH-Fc) have the potential to be applied as broad-spectrum anti-infectives.

In this study, both FH-Fc constructs (i.e., S2635 and S2782) efficiently killed B. burgdorferi in vitro and reduced B. burgdorferi dissemination and the development of Lyme disease-associated manifestations in the murine model when administered prior to tick challenge. However, pre-treatment of mice with S2635 but not S2782 eliminated bacterial burdens in fed nymphs, suggesting that the bacterial killing mechanisms for these FH-Fc constructs differ. One contributor to such differences could be the infection stages when the bacterial FH-binding proteins are produced. As a SCR(6–7)-based FH-Fc construct, S2635 was shown in this study to have physiological ranges of binding affinity (K_D = ~10⁻⁷ M) selective to CspA and CspZ from B. burgdorferi strain B31-5A4. CspA, but not CspZ, is produced in the spirochetes residing in feeding nymphs, which is required for bacteria to evade the complement in tick bloodmeal (25, 26). After B. burgdorferi invades a vertebrate host, CspZ, but not CspA, is produced to promote the systemic spread of bacteria (25, 27, 28). These findings thus support the possibility of S2635-mediated bacterial killing in the enzootic cycle by binding to CspA to decrease the levels of tick-to-host transmission of Lyme borreliae and to CspZ to reduce the extent of bacterial dissemination. Further, we found S2782 to bind uniquely to OspE variants from B. burgdorferi strain B31-5A4, which are documented SCR19–20 binders. OspE variants are produced on the surface of B. burgdorferi residing in fed nymphs and vertebrate hosts (25). OspE-targeted antibodies were reported to not reduce the bacterial loads in fed nymphs but significantly decrease bacterial burdens in hosts after tick-to-host transmission (75). Such phenotypes are similar to our findings in the S2782-treated mice and, therefore, support the protective mechanisms underlying S2782 by binding to pathogens via OspE after Lyme borreliae transmission. Additionally, CspA and OspE are produced when Lyme disease bacteria are in ticks, but our results showed that CspA-targeted FH-Fc (i.e., S2635) but not the OspE targeted FH-Fc (i.e., S2782) eliminated bacteria in fed nymphs. One possibility to address this distinction could be the lower expression of levels of OspE, compared to those of CspA, when bacteria are in fed nymphs (76). Moreover, because S2635 and S2782 bind to different Lyme borreliae proteins (S2635 for CspA/CspZ and S2782 for OspE), simultaneously using both S2635 and S2782 would ideally more efficaciously eliminate Lyme disease bacteria. We do not have the evidence in vivo to support this possibility because the tested condition (0.2mg/kg, Intramuscular injection, and infection at 24-hr after FH-Fc inoculations) in this study for either FH-Fc construct is sufficient to result in the in vivo readouts (i.e., bacterial burdens, C6 seropositivity, and joint inflammatory scores), indistinguishable from the uninfected mice. However, B. burgdorferi B31-5A4 or 297, B. afzelii VS461 or PKo, or B. bavariensis PBi treated with both S2635 and S2781 had BA₅₀ values significantly greater than those spirochetes treated with each of these FH-Fc constructs (Fig. S2, Table S3). These results support further work to test the efficacy of using both FH-Fc constructs simultaneously for more efficient Lyme disease prevention.

When the work was extended to different Lyme borreliae species or strains, we found that S2635 showed bacterial killing ability to selected bacterial strains and species but S2782 displaying a broader ability to eradicate all tested species and strains. These results suggest the potential of developing S2782 as an anti-tickborne therapeutic with great breadth. As the binding activity of FH-Fc constructs to their FH-binding partners from pathogens is one of the determining factors for their efficacy, our findings raise the possibility that S2635 and S2782 differ in their capability to bind to different spirochete FH-binding protein variants. CspA shares approximately 40% sequence identity among different Lyme borreliae species, and OspE variants from different strains within the same species display greater than 85% sequence identity (77–79). Such protein polymorphism is consistent with documented CspA and OspE variant-to-variant different levels of human FH-binding activity (26, 80, 81). Although CspZ is highly conserved (~98%) among different spirochete strains within the same species and moderately conserved (~80%) among different Lyme borreliae species, CspZ variants also display variant-specific FH-binding activity (28, 82). Here, we found that S2782 bound to all tested Lyme borreliae OspE variants at similar levels (K_D = 4 × 10⁻⁷ to 8 × 10⁻⁸ M). S2635 binds strongly to CspA and CspZ variants from B. afzelii and the CspA variant from B. burgdorferi 297 (K_D = ~10⁻⁷ M) but weakly to both CspA and CspZ variants from B. bavariensis (K_D = ~10⁻⁶ M). However, this FH-Fc neither bind to the CspZ variant from B. burgdorferi 297 nor to the CspA and CspZ variants from B. garinii. Therefore, these results suggest that the sequence differences between tested CspA and CspZ variants may impact S2635-binding activity, but the sequence variation of tested OspE variants here appears to not impact the S2782-binding activity. Further, such differences in the S2635- and S2782-binding affinities among the variants of spirochete FH-binding proteins are correlated with the extent of their ability to kill tested Lyme borreliae or strains. Our results suggest that S2782 has the potential as a broad spectrum Lyme borreliae anti-infective.

B. burgdorferi deficient in producing human FH-binding proteins (i.e., CspA, CspZ, and OspE), if available, could be useful in delineating which of these proteins are the targets of our FH-Fc constructs. However, we have reported that the cspZ mutant has merely partial survival defects at the early stages of bacterial dissemination after tick-borne transmission (28). Additionally, OspE proteins belong to a family that has multiple sequentially different variants and copies in individual Lyme borreliae strains (24), resulting in significant hurdles in generating a spirochete mutant strain deficient in producing all OspE proteins. Therefore, although using mutants deficient in each of these Lyme bacterial FH-binding proteins to explore the protective mechanisms of FH-Fc is interesting, this technical difficulty would need to be resolved before testing this concept using B. burgdorferi mutant strains. Further, we have shown the ability of S2635 and S2782 to impact Lyme borreliae survivability via in vitro bacterial lysis and phagocytosis. However, these bacterial killing activities can be 1) mediated by preventing Lyme borreliae from escaping from FH-binding-dependent, alternative pathway-mediated complement evasion (15) and/or 2) promoting Fc-mediated bacterial killing (83). As the key amino acids that confer Fc-mediated bacterial killing of antibodies have been identified (84), mutating these amino acids on S2635 or S2782 could help to delineate the mechanism(s) by which these proteins promote both in vitro and in vivo bacterial killing. Overall, this study demonstrated the concept of targeting Lyme borreliae FH-binding proteins with the enhancement of classical pathway killing by Fc for the prophylaxis of Lyme disease. Such work expands the use of FH-Fc to an additional pathogen species, Lyme disease bacteria, and provides an alternative option for Lyme disease prevention and prophylaxis, suggesting the possibility of developing a broad-spectrum preventive platform for multiple pathogens.

Glossary

LD

Lyme disease

FH

Factor H

Fc

The Fc region of the Immunoglobulin G

OspE

OspE family of proteins

MBL

Mannose-binding lectin

RCA

Regulators of complement activation

SCR

Short consensus repeat

GST

Glutathione-S-transferase

A.U.

Arbitrary unit

EC₅₀

The concentration of S2635 or S2782 that killed 50% of spirochetes

CFSE

Carboxyfluorescein diacetate succinimidyl ester

FACS

Fluorescence-Activated Cell Sorting

BSA

Bovine serum albumin

MFI

Mean fluorescence intensity

SPR

Surface plasmon resonance

PMN

Polymorphonuclear neutrophils

DATA AVAILABILITY

The amino acid sequences for S2635 and S2782 have been deposited in the GenBank under accession codes 2968033 and 2968034, respectively.