To assess the importance of SLPI during murine Lyme arthritis, we compared the outcomes of B. burgdorferi infection of C57BL/6 WT and Slpi-/- mice. The C57BL/6 WT and Slpi-/- mice were infected with 10⁵ spirochetes subcutaneously. Infection burdens in the skin were assessed by qPCR of B. burgdorferi DNA in ear punch biopsies at 7, 14, and between 21–24-day post-infection (dpi) (Figure 1A–C). Infection burdens in the heart (Figure 1D) and tibiotarsal joint (Figure 1E) tissues were assessed between 21 and 24 dpi. We did not observe any significant difference in infection burden in the skin between WT and Slpi-/- mice (n=24) at 7, 14, 21–24 dpi, or in the heart between 21–24 dpi (Figure 1A–D).

Strikingly, we observed a significantly higher spirochete burden in the ankle joints of infected Slpi-/- mice (n=24) between 21 and 24 dpi (Figure 1E). Furthermore, at around 24 dpi, significant swelling was also observed solely in the infected Slpi-/- mice (Figure 2A, red arrow). The level of swelling was first scored visually. While 14 out of 20 infected Slpi-/- mice displayed visible swelling (score ≥1), only 1 in 14 infected WT mice displayed mild swelling (score 1) at the ankle (Figure 2B). The tibiotarsal joints were then dissected, fixed, and stained with H&E for histopathological evaluation of the level of inflammation (Figure 2C and D). Inflammation of bursa and soft tissue adjacent to the tibiotarsal joint, but not in the tibiotarsal synovium, was consistently observed in the infected Slpi-/- mice (Figure 2C, black rectangle). In contrast, only 1 out of 14 infected WT mice displayed modest inflammation (score = 2) in these sites (Figure 2D). The above data indicate the importance of SLPI in modulating the development of periarticular inflammation associated with murine Lyme arthritis.

It has been established that SLPI exerts its anti-inflammatory effect by inhibiting neutrophil serine protease (NSP) and dampening NF-κB activation in macrophages (Moreau et al., 2008). Thus, to investigate the mechanism underlying the effect of SLPI on murine joint inflammation, we sought to identify the population of infiltrating cells in the periarticular tissues of infected WT and Slpi-/- mice. Between 21 and 24 dpi, the ankle joints were dissected. To obtain single-cell suspensions of infiltrating cells, bone marrow cells were removed and discarded and the joint and periarticular tissues were digested (Akitsu and Iwakura, 2016). The cells were stained for flow cytometry with CD45, CD11b, and LY6G to label neutrophils (Figure 3A), and CX3CR1, CD64, and LY6C to label macrophages (Figure 3B; Feng et al., 2023). After gating, we observed significantly higher percentages of infiltrating neutrophils and macrophages in the dissected tissues from infected Slpi-/- than WT mice (Figure 3A and B). To further eliminate the possibility of neutrophilic contamination within the macrophage population, we also implemented a Ly6G-negative gating strategy. The result showed a consistently higher percentage of macrophages in the infected Slpi-/- mice (Figure 3B, Figure 3—figure supplement 1). Using RT-qPCR on the tibiotarsal tissues extracted from B. burgdorferi-infected Slpi-/- mice, we detected increased gene expression of neutrophil chemoattractant receptor C-X-C motif chemokine receptor 2 (Cxcr2), monocyte chemoattractant protein 1 (Mcp-1), and its receptor C-C motif chemokine receptor 2 (Ccr2) (Figure 3C–E).

Furthermore, the serum cytokine/chemokine profile was assessed from uninfected and B. burgdorferi-infected WT and Slpi-/- mice between 21 and 24 dpi. We observed a significant increase in IL-6 in infected Slpi-/- mice (Figure 3F). IL-6 recruits and stimulates neutrophils, leading to the secretion of neutrophil-derived serine proteases including neutrophil elastase (NE) and cathepsin G (CTSG) (Srirangan and Choy, 2010). The lack of serine protease inhibitors, such as SLPI, can cause excessive protease activity and subsequent tissue damage and inflammation (Wilkinson et al., 2019). Indeed, using ELISA, we observed a significantly higher level of NE solely in the serum of infected Slpi-/- mice (Figure 3G). An increased serum level of NE was also observed in the B. burgdorferi-infected, arthritis-susceptible C3H/HeN mice at 21 dpi (Figure 3H). These data suggest that, in the absence of SLPI, excessive serine protease activity can exacerbate murine Lyme arthritis.

A correlation between IL-6, macrophages, metalloproteinases (MMPs), and articular cartilage destruction has been observed in the synovial tissue of RA patients (Murphy and Nagase, 2008). Elevated levels of host matrix metalloproteinases (MMPs) have also been found in the synovial fluid of Lyme arthritis patients and can cause excessive tissue damage (Hu et al., 2001). Thus, a mouse MMP 5-Plex Discovery Assay was used to explore the serum levels of different MMPs. We observed a significant increase in the levels of MMP-8 solely in the infected Slpi-/- mice (Figure 3I). Taken together, our data suggest that SLPI suppresses inflammation in B. burgdorferi-infected mouse joint tissues by potentially inhibiting neutrophil and macrophage infiltration and subsequent protease-mediated tissue destruction.

Despite numerous studies of serum, synovial fluid, and tissue from Lyme arthritis patients, the importance of anti-protease–protease balance in Lyme arthritis has not been investigated (Lochhead et al., 2021). Based on our data obtained from the Slpi-deficient mice, we assessed the serum SLPI level in Lyme disease patients (Figure 4). Due to the limited samples available from Lyme arthritis patients, we included samples from Lyme disease patients who presented with earlier manifestations of Lyme disease. The serum level of human SLPI assessed by ELISA showed a significant decrease in the SLPI level in Lyme disease patients comparing to healthy adult controls (Figure 4). Similar to our data from B. burgdorferi-infected mice, this result suggests a correlation between the lack of SLPI and humans exhibiting clinical manifestations of Lyme disease, including arthritis.

It has been demonstrated that B. burgdorferi interacts with various mammalian proteins to establish infection in the mammalian host (Gupta et al., 2020; Shi et al., 2008; Li et al., 2006; Koenigs et al., 2013). Thus, we postulated that B. burgdorferi could interact with SLPI to influence the progression of joint inflammation. To test this hypothesis, we probed a recently developed BASEHIT (BActerial Selection to Elucidate Host-microbe Interactions in high Throughput) library with B. burgdorferi (Gupta et al., 2020; Sonnert et al., 2024; Arora et al., 2022; Hart et al., 2024). BASEHIT utilizes a genetically barcoded yeast display library expressing 3324 human exoproteins, thus enabling a comprehensive screen of host–microbe interactions in a high-throughput fashion. Human SLPI is one of the exoproteins that passed the significance threshold, indicating B. burgdorferi-SLPI binding. To further establish that hSLPI directly binds to B. burgdorferi, we performed ELISA with whole-cell B. burgdorferi lysates. We observed strong binding between whole-cell B. burgdorferi lysates and hSLPI at a level as low as 10 nM (Figure 5A).

To extend these studies, flow cytometry was performed using intact B. burgdorferi and both human and murine SLPI. A significant increase in fluorescence intensity was observed when B. burgdorferi, cultivated at 33°C, was incubated with human SLPI at 10 nM and 1 µM level (Figure 5B). Though the binding to 10 nM rmSLPI was at background level, we observed a significant increase in fluorescence intensity when B. burgdorferi were incubated with 1 µM rmSLPI (Figure 5C). Flow cytometry also demonstrated increased binding of B. burgdorferi cultured at 37°C to 1 µM hSLPI or rmSLPI (Figure 5B and C). This indicates that the binding was more robust when performed at temperatures that B. burgdorferi encounter in the mammalian host. Immunofluorescent microscopy was an additional method that also demonstrated direct binding of B. burgdorferi with human or murine SLPI (Figure 5D).

In contrast to B. burgdorferi B31-A3, an infectious strain used throughout this study, we did not observe any binding between hSLPI and B. burgdorferi B31A, a high-passage non-infectious strain (Figure 5—figure supplement 1A; Caine and Coburn, 2015; Elias et al., 2002). The above observation further suggests that the direct interaction between SLPI and B. burgdorferi could impact the pathogenesis of murine Lyme arthritis. To further investigate the potential B. burgdorferi ligand that interacts with SLPI, we probed protease-treated B. burgdorferi lysates with hSLPI using ELISA. After treatment with proteinase K, we observed a marked decrease in binding of hSLPI to B. burgdorferi lysates (Figure 5—figure supplement 1B). This result suggests that hSLPI can directly interact with a B. burgdorferi protein.

It has been showed that SLPI has antimicrobial effects against multiple gram-negative and -positive bacteria (Wiedow et al., 1998; Nishimura et al., 2008; Hiemstra et al., 1996). However, using the BacTiter-Glow microbial cell viability assay, we did not observe any significant changes in B. burgdorferi viability in the presence of hSLPI (Figure 5—figure supplement 2A). A previous study also demonstrated that the tick salivary protein, Salp15, specifically interacted with B. burgdorferi outer surface protein C (OspC) (Ramamoorthi et al., 2005). The binding of Salp15 protected spirochetes from killing by polyclonal mouse or rabbit antisera in vitro (Ramamoorthi et al., 2005; Schuijt et al., 2008). However, again, using the BacTiter-Glow microbial cell viability assay, the pre-incubation of hSLPI did not protect spirochetes from killing by mouse B. burgdorferi antisera (Figure 5—figure supplement 2B). Thus, the importance of the SLPI-B. burgdorferi interaction and the direct effect of such an interaction on B. burgdorferi biology are likely independent of direct borreliacidal activity or any interference with the antibody-mediated B. burgdorferi killing.

Females Slpi-/- and WT C57BL/6 mice were used for the in vivo B. burgdorferi infection. We have examined B. burgdorferi infection in both male and female C57BL/6 mice, and no differences in the development of infection or disease have been noted. Both male and female Lyme disease patients were included in the study. Sex was not considered a biological variable.

This study used archived serum samples from adult Lyme disease subjects and healthy controls that were previously collected under NIH U19AI089992 with approval of the Yale University Institutional Review Board for human subjects research (IRB protocol# 1112009475). All the animal experiments in this study were performed in accordance with the Guidelines for the Care and Use of Laboratory Animals of the National Institutes of Health. The animal protocol (2023-07941) was approved by the Institutional Animal Care and Use Committee at the Yale University School of Medicine.

SLPI levels were measured in a total of 23 serum samples from seven subjects at the time of Lyme disease diagnosis (four with a single erythema migrans lesion and three with the late manifestation of Lyme arthritis) and from five healthy controls. Serum samples from Lyme disease subjects were available at up to three times points: (1) study entry, range 0–9 days after onset of symptoms; (2) 30 days post diagnosis; and (3) up to 3 months after the completion of antibiotic therapy (range 4.5–6 months after diagnosis). Additional details can be found in Supplementary file 1. The level of SLPI in the serum was measured using the Human SLPI DuoSet ELISA kit (R&D Systems, #DY1274-05).

B. burgdorferi B31-A3, an infectious clonal derivative of the sequenced strain B31, was a generous gift from Dr. Utpal Pal at the Department of Veterinary Medicine, University of Maryland, College Park (Bernard et al., 2018). B. burgdorferi B31-A3 and B. burgdorferi B31A were grown in Barbour-Stoenner-Kelly H (BSK-H) complete medium (Sigma-Aldrich, #B8291) in a 33°C setting incubator. The live cell density was determined by dark field microscopy and using a hemocytometer (INCYTO, #DHC-N01). Low-passage (p<3) B. burgdorferi B31-A3 was used throughout this study.

The Slpi-/- C57BL/6 mice have been described previously (Matsuba et al., 2017; Nakamura et al., 2003). The wild-type (WT) C57BL/6 mice were purchased from the Jackson Laboratory and used as the controls. 5–7-week-old female WT and Slpi-/- C57BL/6 mice were used for infection. 4–6-week-old female C3H/HeN mice were purchased from Charles River Laboratories and used for infection. Both C57BL/6 and C3H/HeN mice were infected with low-passage 10⁵ B. burgdorferi subcutaneously (5–9 mice/group). PBS sham-infected mice were used as controls. Mice were euthanized approximately 3 weeks post infection within a 3-day window (between 21 and 24 dpi) based on the feasibility and logistics of the laboratory. Ear punch biopsies were taken at 7, 14, and between 21–24 dpi to determine the infection burden in the skin. Between 21 and 24 dpi, mice were euthanized, and heart and joint tissues were collected to quantify the spirochete burden. The protocol for the use of mice was reviewed and approved by the Yale Animal Care and Use Committee.

DNA was extracted from the heart, tibiotarsal joint, and ear punch samples using QIAGEN DNeasy Blood & Tissue Kit, QIAGEN. Quantitative PCR was performed using iQ-SYBR Green Supermix (Bio-Rad). For quantitative detection of B. burgdorferi burden within mouse tissue samples, q-PCR was performed with DNA using flagellin (flaB), a marker gene for Borrelia detection. The mouse β-actin gene was used to normalize the amount of DNA in each sample. The nucleotide sequences of the primers used in specific PCR applications were described previously (Gupta et al., 2020).

Mice were euthanized by CO₂ asphyxiation, and one rear leg from each mouse was dissected, immersion-fixed in Bouin’s solution (Sigma-Aldrich, #HT10132). Fixed tissues were embedded, sectioned, and stained with H&E by routine methods (Comparative Pathology Research Core in the Department of Comparative Medicine, Yale School of Medicine). Periarticular and joint inflammation was scored in a blinded fashion in a graded manner from 0 (negative), 1 (minimal), 2 (moderate), to 3 (severe).

The WT and Slpi-/- C57BL/6 mice were infected with B. burgdorferi as described above. The mice were euthanized between 21 and 24 dpi. The ankle joints were cut out at around 0.7 cm proximal to the ankle joint. The portion distal to the midfoot was discarded, and the skin removed. The bone marrow cells were flushed out with RPMI 1640 (Gibco, #11875-093) using a 27-gauge needle. The bone marrow-depleted ankles were cut into 3–4-mm-sized tissue pieces and incubated with digestion media containing 2.4 mg/ml hyaluronidase (Sigma-Aldrich, #H3506), 1 mg/ml collagenase (Sigma-Aldrich, #C2139) in RPMI 1640 (Gibco, #11875-093) supplemented with 10% fetal bovine serum (FBS) for 1 h at 37°C with 5% CO₂. The digestion media containing the tissue pieces were passed through a 70 μm cell strainer (Thermo Scientific, #352350). The remaining tissue pieces were mashed using a 10 ml syringe plunger. The digestion media containing the isolated cells were neutralized with RPMI 1640 with 10% FBS (Akitsu and Iwakura, 2016). The red blood cells were lysed by ACK Lysing buffer (Gibco, #A1049201). The cells were rinsed and resuspended in FACS buffer and ready for staining for flow cytometry.

The cells were incubated with Fc receptor antibody TruStain FcX anti-mouse CD16/32; BioLegend, #101320, and antibodies including PerCP anti-mouse CD45 (BioLegend, #103130), BV711 anti-mouse Ly6G (BioLegend, #127643), PE anti-mouse CD11b (BioLegend, #101208), APC/CY7 anti-mouse CX3CR21 (BioLegend, #149047), FITC anti-mouse Ly6C (BioLegend, #128005), APC anti-mouse CD64 (BioLegend, #139305), and LIVE/DEAD fixable violet stain kit (Invitrogen, #L34955) on ice for 30 min. The samples were rinsed twice with FACS buffer and run through BD LSRII (BD Biosciences). The data was then analyzed using FlowJo (Feng et al., 2023).

Mice were euthanized between 21 and 24 dpi. The ankle joints were excised as described above, snap-frozen in liquid nitrogen, and stored at −80°C. The frozen tissue was pulverized in liquid nitrogen using a mortar and pestle (Wilson et al., 2021). The RNA was purified using Trizol (Invitrogen, #15596-018) following a published protocol (Zheng and McAlinden, 2021). cDNA was synthesized using the iScript cDNA Synthesis Kits (Bio-Rad, #1708891). qPCR was performed using iQ SYBR Green Supermix (Bio-Rad, #1725124). The relative expression of each target gene was normalized to the mouse β-actin gene. The target genes and corresponding primer sequences are shown in the Key Resources Table.

Blood samples from each group of mice was collected by cardiac puncture between 21 and 24 dpi, and sera were collected. The murine neutrophil elastase level was measured using the Mouse Neutrophil Elastase/ELA2 DuoSet ELISA (R&D Systems, #DY4517-05). Serum was sent for cytokine analysis by the Mouse Cytokine/Chemokine 32-Plex Discovery Assay Array (MD32) and the Mouse MMP 5-Plex Discovery Assay Array (MDMMP-S, P) performed by Eve Technologies. The cytokines and chemokines represented by MD32 are Eotaxin, G-CSF, GM-CSF, IFN-γ, IL-1α, IL-1β, IL-2, IL-3, IL-4, IL-5,IL-6, IL-7, IL-9, IL-10, IL-12 (p40), IL-12 (p70), IL-13, IL-15, IL-17A, IP-10, KC, LIF, LIX, MCP-1, M-CSF, MIG, MIP-1α, MIP-1β, MIP-2, RANTES, TNFα, and VEGF. The MMPs represented by MDMMP-S, P are MMP-2, MMP-3, MMP-8, proMMP-9, and MMP-12.

The murine Slpi cDNA ORF clone was purchased from GenScript (OMu22721). The coding sequence was subsequently cloned into pET22b(+) expression vector (Novagen) in frame with the pelB signal peptide using Gibson Assembly (Maffia et al., 2007). E. coli strain Rosetta-gami 2 (DE3) (Novagen, #71351) was transformed with the SLPI-pET22b+ and grown at 37°C with ampicillin (100 μg/ml), tetracycline (12.5 μg/ml), streptomycin (50 μg/ml), and chloramphenicol (34 μg/ml). Cells were induced with 1 mM IPTG (18°C, overnight), harvested, and lysed with BugBuster Protein Extraction Reagent (Novagen, #70921-3). Recombinant mSLPI was purified with a Ni-NTA resin column as described by the manufacturer (QIAGEN). To evaluate the activity of the purified rmSLPI, the trypsin inhibitory activity was assayed with the fluorescent substrate Mca-RPKPVE-Nval-WRK(Dnp)-NH2 Fluorogenic MMP Substrate (R&D Systems, #ES002) and the absorbance was monitored at 405 nm using a fluorescent plate reader (Tecan).

Actively growing low-passage B. burgdorferi was cultured to a density of 10⁶–10⁷ cells/ml and harvested at 10,000 × g for 10 min. Cells were rinsed twice with PBS and blocked in 1% BSA for 1 h at 4°C. The cells were pelleted, rinsed, resuspended, and incubated with 10 nM and 1 µM human SLPI (R&D Systems, #1274-PI-100) and murine SLPI (produced in lab as described above) at 4°C for 2 h. The binding was detected with goat anti-human or murine SLPI (R&D Systems, #AF1274 and AF1735) and Alexa Fluor 488 or Alexa Fluor 647 donkey anti-goat IgG (H+L) (Invitrogen, #A32814, A-21447). The samples were fixed with 2% PFA before running through BD LSRII Green (BD Biosciences). The data was then analyzed using FlowJo. The integrity of B. burgdorferi organism was confirmed by Hoechst 33324/propidium iodide double staining (Figure 5—figure supplement 3). The fixed B. burgdorferi sample was included as a positive control for the propidium iodide staining. Permeabilization was not performed during this protocol. Thus, the binding detected was on the bacterial outer surface.

B. burgdorferi was cultured to a density of 10⁶–10⁷ cells/ml and harvested at 10,000 × g for 10 min. To make the B. burgdorferi lysate, cells were rinsed twice with PBS, pelleted, and lysed using BugBuster Protein Extraction Reagent (Novagen, #70921-3). Protein concentration in the lysate was measured by absorbance at 280 nm using the nanodrop (Fisher Scientific). For the protease assay, the B. burgdorferi lysate was incubated in the presence or absence of proteinase K (0.2 mg/ml, Thermo Scientific, #EO0491) for 10 min. In an immuno 96-well plate (MaxiSorp), wells were coated with 200 ng of B. burgdorferi lysate. Samples were blocked with 1% BSA followed by incubation with human SLPI at varying concentrations (1–1000 ng) for 1 h at room temperature. The binding was probed with goat anti-human SLPI (R&D Systems, #AF1274) and rabbit anti-goat IgG (whole molecule)-peroxidases antibody (Sigma-Aldrich, #A8919-2ML). KPL Sureblue TMB Microwell Peroxidase substrate, 1-component (Seracare, #5120-0077) was used. The reaction was stopped with KPL TMB stop solution (Seracare, #5150-0021), and absorbance was read at 450 nm.

Actively growing low-passage B. burgdorferi was cultured to a density of 10⁶–10⁷ cells/ml, rinsed twice with PBS, and blocked with 1% BSA for 1 h at 4°C. B. burgdorferi was incubated with human or murine SLPI at 4°C for 2 h. The spirochetes were probed with goat anti-human or murine SLPI (R&D Systems, #AF1274 and AF1735) and Alexa Fluor 488 donkey anti-goat IgG (H+L) (Invitrogen, #A32814). B. burgdorferi were then stained with Hoechst 33342 (Invitrogen, #H1399). The samples were fixed with 2% PFA before imaged with Leica SP8. The integrity of B. burgdorferi organism is confirmed in Figure 5—figure supplement 3. Permeabilization was not performed during this protocol. Thus, the binding detected was on the bacterial outer surface.

The BacTiter Glo microbial cell viability assay quantifies the ATP present in the microbial culture by measuring luminescence. The amount of ATP is proportional to the number of viable cells in culture (Gupta et al., 2020; Arora et al., 2022; Schwendinger et al., 2013). To test the borreliacidal activity of human SLPI, 1 × 10⁵ spirochetes were treated with 0–10 μM hSLPI (R&D Systems, #1274-PI-100) at 33°C for 48 h. The luminescence was measured using a fluorescence plate reader (Tecan). The percent viability was normalized to the control spirochetes culture without hSLPI treatment. To test the effect of hSLPI on the antibody-mediated B. burgdorferi killing, 1 × 10⁵ spirochetes were pretreated with 0–5 μM hSLPI (R&D Systems, #1274-PI-100) at 33°C for 2 h. 20% mouse B. burgdorferi antisera were then added and incubated for 2 and 4 h. The mouse antisera were collected from B. burgdorferi-infected mice between 21 and 24 dpi. The luminescence was measured as described above. The percent viability was normalized to the control spirochetes culture without any treatment.

The analysis of all data was performed using the nonparametric Mann–Whitney or ANOVA using Prism 10 software (GraphPad Software, Inc, San Diego, CA). A p-value of <0.05 was considered statistically significant.