Female C57BL/6 (#000664) and CCR7^-/- mice (#006621) (31) were purchased from Jackson Laboratory. Mice were maintained under specific pathogen-free conditions and used at 8-10 weeks of age, following protocols approved by the Institutional Animal Care and Use Committee (protocol # 2101001) at the University of Texas Medical Branch (UTMB) in Galveston, TX. All mouse infection studies were performed in the ABSL3 facility in the Galveston National Laboratory located at UTMB; all tissue processing and analysis procedures were performed in the BSL3 or BSL2 facilities. All procedures were approved by the Institutional Animal Care and Use Committee (IACUC) and the Institutional Biosafety Committee, in accordance with Guidelines for Biosafety in Microbiological and Biomedical Laboratories. UTMB operates to comply with the United States Department of Agriculture (USDA) Animal Welfare Act (Public Law 89–544), the Health Research Extension Act of 1985 (Public Law 99–158), the Public Health Service Policy on Humane Care and Use of Laboratory Animals, and the NAS Guide for the Care and Use of Laboratory Animals (ISBN-13). UTMB is a registered Research Facility under the Animal Welfare Act and has a current assurance on file with the Office of Laboratory Animal Welfare, in compliance with NIH Policy.

Mice were inoculated with Ot Karp strain (4 × 10³ focus forming units (FFU), 10 -12 µL volume) or phosphate buffered saline (PBS, mock control) in the dermis of the lateral ears by using a 30 G-needle and 25-μl Hamilton syringe (Hamilton Company, Reno, NV) (28). Mice were monitored daily for weight loss, signs of disease, and survival. Some mice were intraperitoneally injected with anti-Ly6G, anti-Gr-1 or anti-Ly6G/6C (clone#1A8, RB6-8C5 and NIMP-R14, respectively; 250 µg/mouse, BioxCell, Lebanon, NH) at day -1, 0, 1, and 2 post-infection. Control mice were treated with Rat IgG (250 µg/mouse). For blocking local chemokine signaling, mice were i.d. injected with anti-CCL19 (2 µg, #AF880, R&D, Minneapolis, MN), anti-CCL21 (2 µg, #AF457, R&D, Minneapolis, MN) or both (2 µg/each, 4 µg total) at 1 day prior to infection. Goat IgG (#Ab-108, R&D) was used as a control or supplemented to the equal amount of antibody (4 µg total) per injection. At indicated time-points, mice were euthanized by CO₂ inhalation and tissues were harvested for further analysis.

Ot was prepared from heavily (80-100%) infected L929 cell cultures (22). Briefly, bacteria were inoculated onto confluent monolayers in T150 cell culture flasks and gently rocked for two hours at 37°C, at the end of which Minimum Essential Medium (MEM) with 10% fetal bovine serum, 100 units/mL of penicillin and 100 µg/mL of streptomycin were added. Cells were harvested by scraping at day 7 post-infection, re-suspended in MEM, and lysed using 0.5 mm glass beads by vortex for 1 min. The cell suspension was collected and centrifuged at 300×g for 10 min to pellet cell debris and glass beads. The supernatant from one T150 flask was further inoculated onto new monolayers of five T150 flasks. This process was repeated for a total of six passages. Cells from five flasks were pooled in a 50 mL conical tube with 20 mL medium and 5 mL glass beads. The conical tubes were gently vortexed at 10 sec intervals for 1 min to release the intracellular bacteria and placed on ice. The tubes were then centrifuged at 700×g to pellet cell debris, and the supernatant was collected in Oakridge high speed centrifugation bottles, followed by the centrifugation at 22,000×g for 45 min at 4°C to harvest bacteria. Sucrose-phosphate-glutamate buffer (0.218 M sucrose, 3.8 mM KH₂PO₄, 7.2 mM KH₂PO₄, 4.9 mM monosodium L-glutamic acid, pH 7.0) was used for preparing bacterial stocks, which were stored at -80°C (17, 32). The same lot of Ot Karp stocks was used for all experiments described in this study.

For inoculation of L929 by using DC lysates, we harvested dLN from 30 mice at day 2 post-infection and pooled them for the purification of CD11c⁺ population by using anti-CD11c magnetic beads (Miltenyi Biotec, San Jose, CA). CD11c^- cells were also collected as a control. We then lysed 2 × 10⁶ cells using glass beads, as we described above, and seeded cell lysates into L929 cells in T12.5 flasks. At day 7, L929 cells were harvested, and cell lysates were further seeded into new T12.5 flasks with L929 cells. At day 14, cells were collected for the measurement of bacterial burdens by qPCR.

The CFSE labelling of bacteria was performed according to a previous report (30). Briefly, bacterial stock was added into a 15 mL tube of 1 mL PBS. One microliter of CFSE (5 mM, Biolegend, San Diego, CA) was added into the tube, followed by a gentle vortex. Samples were incubated for 10 min at 4°C in the dark. After incubation, the remaining dye was quenched by adding 10-fold volume of MEM + 10% FBS and incubated for 5 mins. The labeled bacteria were pelleted at 22,000 × g for 15 min, washed once with cold PBS, and resuspended either by cold PBS for animal infection or MEM for in vitro cell infection.

Cells were lysed with RLT lysis buffer for 10 min, and RNA was extracted using RNeasy Mini kits (Qiagen, Germantown, MD), followed by the synthesis of cDNA using an iScript Reverse Transcription kit (Bio-Rad, Hercules, CA). cDNA was amplified in a 10 μL reaction mixture containing 5 μL of iTaq SYBR Green Supermix (Bio-Rad, Hercules, CA) and 5 μM each of gene-specific forward and reverse primers. The PCR assays were denatured for 30 s at 95°C, followed by 40 cycles of 15 s at 95°C, and 60 s at 60°C, by utilizing the CFX96 Touch real-time PCR detection system (Bio- Rad, Hercules, CA). Relative quantitation of mRNA expression was calculated using the 2^−ΔΔCt method. The primers are listed in Table S1 .

To determine bacterial burdens, tissues (e.g., skin, lung and brain) were collected and incubated with proteinase K at 56°C. DNA was then extracted using a DNeasy Blood & Tissue Kit (Qiagen, Germantown, MD) and used for qPCR assays, as previously described (23, 32). The 47 kDa gene was amplified using the primer pair OtsuF630 (5’-AACTGATTTTATTCAAACTAATGCTGCT-3’) and OtsuR747 (5’-TATGCCTGAGTAAGATACGTGAATGGAATT-3’) primers (IDT, Coralville, IA) and detected with the probe OtsuPr665 (5’-6FAM-TGGGTAGCTTTGGTGGACCGATGTTTAATCT-TAMRA) (Applied Biosystems, Foster City, CA) by SsoAdvanced Universal Probes Supermix (Bio-Rad, Hercules, CA). Bacterial loads were normalized to total nanogram (ng) of DNA per µL for the same samples. The copy number for the 47-kDa gene was determined by known concentrations of a control plasmid containing a single-copy insert of the gene. Gene copy numbers were determined via serial dilution (10-fold) of the Ot Karp 47-kDa plasmid.

Bone marrow cells were collected from the tibia and femur of B6 mice, followed by incubation with Red Cell Lysis Buffer (Sigma, St. Louis, MO) for 5 min at room temperature to remove red blood cells. Neutrophils were purified by anti-Ly6G magnetic beads in LS columns of positive selection (Miltenyi Biotec, San Jose, CA) (33). The purity of neutrophils was confirmed by flow cytometry as ~96% ( Figure S1 ). For cell infection, 5 × 10⁵ neutrophils were seeded in 12-well plates. After one-hour incubation in 37°C and 5% CO₂, bacteria were added to the plate at MOI 10, followed by a gentle mixing by pipette and continuous culture in the incubator. At 4, 10 and 18 h post-infection, cells were harvested in 1.5 mL Eppendorf tubes by centrifugation (300×g, 10 min, 4°C) and lysed for 5 min using RLT buffer (Qiagen, Germantown, MD). The lysed samples were stored in -80°C until the use of RNA isolation.

Bone marrow-derived DCs were generated from B6 mice by cultivation with rGM-CSF (20 ng/ml), as described previously (34). Fresh GM-CSF-containing medium was added at days 3 and 6. DCs were harvested at day 8 and infected with Ot (MOI 10). Cells were collected at 24 and 48 h post-infection and analyzed for MHCII, CD80 and CD86 by flow cytometry. Bacterial loads were measured at 24, 48, 72 and 96 h post-infection by qPCR.

For detection of infiltrated immune cells in the skin, each ear was collected in 1.5 mL Eppendorf tubes with 1 mL accutase (Biolegend, San Diego, CA) and cut with scissors for two min into small pieces. Minced tissues were loaded into Medicons and homogenized by a BD Mediamachine System (BD Biosciences, San Jose, CA) (23). Single-cell suspensions were prepared by passing cell homogenates through 70-µm cell strainers and treated with a Red Cell Lysis Buffer (Sigma, St. Louis, MO). Draining lymph nodes were also collected and passed through cell strainers for preparing single cell suspension. For surface marker analysis, leukocytes were stained with the Fixable Viability Dye (eFluor 506, Thermo Fisher Scientific, Waltham, MA) for live/dead cell discrimination, blocked with FcγR blocker, and incubated with fluorochrome-labeled antibodies (Abs) for 30 min. For apoptosis detection, cells were washed twice with Annexin V Binding Buffer (BD Bioscience, San Jose, CA), and stained with Annexin V. The fluorochrome-labeled antibodies were purchased from Thermo Fisher Scientific and Biolegend as below: APC-efluor 780 anti-CD45 (30-F11), Alexa Fluor 700 anti-CD11b (M1/70), APC anti-Ly6G (1A8), PE/Dazzle-594 anti-Ly6C (HK1.4), BV786 anti-CD24 (M1/69), BV650 anti-CCR2 (475301), PE anti-CD64 (X54-5/7.1), efluor450 anti-MHCII (M5/114.15.2), Percp-efluor710 anti-EpCAM (G8.8), PE-Cy7 anti-CD80 (16-10A1), APC anti-CD86 (GL-1), APC Annexin V, Fixable Viability Dye eFluor 506, and BUV395 anti-CD172a (P84). Cells were fixed in 2% paraformaldehyde overnight at 4°C before analysis. Data were collected by a BD LSR Fortessa and analyzed via FlowJo software version 10 (BD, Franklin Lakes, NJ).

Superficial cervical lymph nodes were removed and fixed in 4% paraformaldehyde (PFA) and embedded in Tissue-Tek OCT. Cryosections (6 µm) were fixed for 10 min in ice-cold acetone and rehydrated with PBS. For immunofluorescent staining, sections were permeabilized and blocked using 5% normal goat serum, 0.2% Triton X-100 in PBS for 1 h, followed by primary antibody incubation overnight at 4°C. The Alexa Fluor488 anti-mouse LYVE1 (ALY7, 1:500 dilution) was purchased from Thermo Fisher Scientific. The rabbit anti-Orientia Karp TSA56 polyclonal antibody (1:200 dilution) was generated and validated by Genscript (Piscataway, NJ). The Alexa Fluor555-conjugated anti-rabbit IgG (H+L), F(ab’)2 Fragment secondary antibody (1:500 dilution, Cell Signaling Technology, Danvers, MA) was used to stain the sections for 1 h at room temperature, followed by three washes with PBS. The TrueVIEW Autofluorescence Quenching Kit (Vector Laboratories, Newark, CA) was applied to diminish unwanted autofluorescence and the microscope slides were covered by using Fluoroshield Mounting Medium with DAPI (Abcam, Cambridge, MA). Images were acquired by an Olympus IX73 inverted microscope equipped with a color camera for light microscopy and with a monochrome camera to capture fluorescent images with standard wavelength filters. All images were captured under the same conditions of laser power, detector gain, and background offset.

Data are presented as mean ± standard error of the mean. Differences between individual treatment and control groups were determined using a two-tailed, unpaired Student’s t-test. One-way ANOVA was used for multiple group comparisons, with Tukey’s post hoc for multiple comparisons among group means. Statistically significant values are referred to as *, p <0.05; **, p <0.01; ***, p <0.001; ****, p <0.0001; NS, no significant difference, respectively.

Scrub typhus is an acute febrile disease that occurs due to the accidental transmission of Ot through human skin by the forest dwelling vector, Leptotrombidium larva (1). Most studies of scrub typhus in animal models have adopted either intraperitoneal or intravenous inoculation of bacterial stocks (17, 21, 24, 32, 35); however, studies of host innate responses and bacterial dissemination in a natural route of infection are limited. By using an intradermal murine model of Ot infection that we established (28), we examined local dynamics of immune cell populations in the skin within the first 24 h of infection ( Figure 1A ). At 4 h post-infection, infiltrated cells detected in the inoculation site were predominantly neutrophils (~35%) and ly6C^hi monocytes (~28%); the percentages of these cells were decreased to 20% at 24 h. The proportions of Ly6C^lo monocytes, Langerhans cells (LC), and conventional type 1 dendritic cells (cDC1) were gradually increased following infection. Limited conventional type 2 dendritic cells (cDC2), monocyte-derived proinflammatory cells (MCs) and double negative DCs were observed during 24 h of infection.

To investigate bacterial dissemination from the skin to visceral organs, we first determined which cells were infected by Ot in the inoculation site. We i.d. injected CFSE-labeled bacteria into mouse ears, isolated skin immune cells at various time-points, and identified Ot-infected cells by flow cytometry. Our results showed that neutrophils (MHCII^-Ly6G^hiCD11b^hiLy6C^intCCR2^loCD64^-) and macrophages/monocytes (MHCII-Ly6G-CD11b^intLy6C^hiCCR2^hiCD64⁺) accounted for 70% and 20% of the infected cells at 4 h post-infection, respectively ( Figures 1B, C ). The proportion of infected neutrophils gradually declined to 20% at 24 h. In contrast, Ot infected more macrophages/monocytes (~60%) and DCs (~15%) at 24 h ( Figure 1C ). These results suggest that these subsets of myeloid cells may facilitate Ot dissemination from the skin to internal organs.

Neutrophils are the most abundant leukocytes patrolling the body, and play a central role in host defense (36, 37); however, they can also contribute to pathogen dissemination, acting as a Trojan horse during infection with intracellular microbes (e.g., Leishmania major and Brucella abortus) (38–41). In the eschar biopsy of scrub typhus patients, neutrophils can form an intense zone of infiltration around the necrotic eschar center (10). Having demonstrated that neutrophils were the first responders and the major target of Ot infection at the inoculation site ( Figure 1 ), we then examined the fate and function of these infected neutrophils during Ot infection. We infected bone marrow-derived neutrophils with Ot in vitro (MOI 10) and assessed cell viability by flow cytometry. We found that Ot-infected neutrophils displayed more cell death (~7%) at 12 h compared with the mocks (~2%) ( Figures 2A, B ). Cell death increased gradually at 18 and 24 h in both mock and infected groups, but infected cells displayed a two-fold higher death rate than mocks. At 36 h, more than half of neutrophils died by Ot infection, while the death rate in the mock group was less than 40% ( Figures 2A, B ). To determine neutrophil activation, we analyzed transcript levels of pro-inflammatory genes, including TNF-α, CXCL10, ICAM-1 and iNOS. Our qRT-PCR results showed significantly increased expression of these genes at 4 h, with peak expression at 10 h ( Figure 2C ). High CXCL10 expression levels were sustained at 18 h, while the expression of other markers (TNF-α, ICAM-1 and iNOS) decreased. Arg-1, the enzyme for metabolizing arginine to ornithine and thereby reducing extracellular arginine at sites of infection (42), was slightly increased in infected cells at 4 h, followed by a 2.5-fold and one-fold upregulation at 10 and 18 h, respectively ( Figure 2C ). Consistent with our finding that Ot promoted neutrophil death ( Figures 2A, B ), we found elevated caspase-3 transcript levels in neutrophils with Ot infection at all examined time-points ( Figure 2C ). To evaluate neutrophil anti-bacterial function, we also analyzed neutrophil granule enzymes by qRT-PCR. We found that Ot infection resulted in a significant decrease of myeloperoxidase (MPO), which is known to be the key neutrophil component for bacterial killing (43, 44). In addition, the levels of proteinase-1 and elastase, which are neutrophil serine proteases for intracellular killing of pathogens, were comparable between mock and infected neutrophils ( Figure 2C ). Our in vitro data suggest that Ot may induce neutrophil activation and apoptosis, but inhibit MPO expression.

To further confirm that Ot can induce neutrophil apoptosis, we i.d. infected mice with CFSE-labeled bacteria in the ears and harvested the skin tissues for flow cytometric analysis at 18 h. As shown in Figure 2D , there was a significant increase in the percentages of CD45⁺Ly6G⁺ neutrophils that infiltrated the infected mouse ears as compared to the mock ears (10% vs. 5%, p<0.001). The result of apoptotic cell staining showed that neutrophils from infected mice exhibited more than a 5-fold increase of Annexin V levels (55% vs. 10%, p<0.001, Figure 2D ), indicating that Ot induced extensive cell apoptosis at the inoculation site. To examine whether neutrophil death was attributed to Ot infection, we analyzed the Annexin V expression on CFSE⁺ (infected) and CFSE^- (uninfected) neutrophils. As shown in the last panel of Figure 2D , more than 55% of neutrophils were CFSE⁺ and Annexin V⁺, while only 10% of cells were CFSE^- but Annexin V⁺, indicating that most of Ot-infected neutrophils were undergoing apoptosis. These lines of evidence collectively argue that Ot infection triggers neutrophil activation and apoptosis in the skin.

Having demonstrated neutrophil recruitment to the inoculation site to engulf bacteria and some neutrophil death ( Figures 1 , 2 ), we next examined whether neutrophils could promote Ot dissemination via antibody-mediated cell depletion. We i.d. infected mice with Ot or PBS in the ears and i.p. treated mice with anti-Ly6G, anti-Gr-1, or anti-Ly6G/6C Abs at day -1, 0, 1 and 2 post-infection. Flow cytometric result confirmed the complete depletion of neutrophils in the skin following the antibody treatment ( Figure S2 ). At day 3, we found comparable bacterial copy numbers in the skin and draining lymph nodes ( Figure 3A ), as well as comparable body weight changes ( Figure 3B ), between IgG control and antibody-treated groups. We also measured bacterial loads of mouse lung and brain tissues harvested at day 7 and 14 post-infection, and found comparable bacterial burdens in mice with or without neutrophil depletion ( Figure 3C ). In all, our data indicate a dispensable role of neutrophils for bacterial control and dissemination in our i.d. model of Ot infection.

To determine the cell types that were relevant to bacterial dissemination, we i.d. injected mouse ears with CFSE-labeled Ot and harvested the cervical lymph nodes at 4, 8, 16 and 24 h for flow cytometric analysis, using our gating strategy shown in Figure 4A (45). The Ot-infected immune cells were gated as CD45⁺CFSE⁺ first, followed by the cell subpopulation gating as Langerhans cells (MHCII^hiCD24^hiEpCAM⁺), macrophages (MHCII^hiCD24^hiCD64⁺EpCAM^-), cDC1 (MHCII^hiCD24^hiCD64^-EpCAM^-), neutrophils (MHCII^-CD24^-Ly6G⁺), and Ly6C^hi monocytes (MHCII^-CD24^-Ly6G^-Ly6C^hiCCR2^hi), respectively. We found that Ot-infected cells were detected in dLN as early as 4 h, and gradually increased during 24 h post infection ( Figure 4B ). Importantly, most Ot-infected immune cells in dLN were cDC1 and LC, as well as some macrophages, suggesting that DCs and macrophages were the Trojan horse for bacterial dissemination from the inoculation site to dLN. CFSE⁺ neutrophils and Ly6C^hi monocytes were nearly undetectable, indicating that these cells were not the main carriers of Ot for dissemination to dLN ( Figure 4B ). Instead, DCs may play a key role in bacterial dissemination via transmission of Ot into dLN. We also infected bone marrow-derived DCs with Ot in vitro and analyzed cell activation markers. Ot infection only slightly increased CD86 expression, but not MHCII and CD80 at 24 and 48 h ( Figure S3A ). Moreover, we found that Ot replicated in DCs as evidenced by significantly increased bacterial burdens in the cultures ( Figure S3B ). To confirm that DCs can carry live Ot for migration into dLN, we purified CD11c⁺ cells from mouse dLN at day 2 post-infection. Cells (2 × 10⁶) were homogenized by glass beads, and the cell lysates were inoculated into L929 cells. At day 14 of cultivation, we found 6 × 10⁴ and 2 × 10⁴ copy numbers of Ot in L929 cells by the inoculation of CD11c⁺ and CD11c^- cell lysates, respectively ( Figure S4 ). Bacteria were undetectable in uninfected L929 cells. Our results suggest that DCs might be the potential reservoirs of Ot and facilitate bacterial dissemination in vivo.

The chemokine receptor CCR7 is known to govern skin DC migration and play a critical role in the entry of dermal and epidermal DCs into lymphatic vessels (46, 47). Our findings of DCs as the major cell populations for Ot transmission into the dLN ( Figure 4 ) led us to speculate that the lack of CCR7 signaling may subsequently reduce early bacterial dissemination. To test this hypothesis, we i.d. infected WT and CCR7^-/- mice with CFSE-labeled Ot and harvested the dLN at days 1, 3 and 5. The CCR7 deficiency resulted in decreased size of the dLN at days 3 and 5 when compared to WT controls ( Figures 5A, B ). Flow cytometric results demonstrated less Ot-carrying DCs in the dLN in the absence of CCR7 as evidenced by decreased MHCII⁺CFSE⁺ cells in CCR7^-/- mice at days 1 and 3 ( Figures 5C, D ). Moreover, the bacterial copy numbers of the dLN were much lower in CCR7^-/- mice ( Figure 5E ), indicating that CCR7 mediated bacterial dissemination from the inoculation site into the dLN. Since CCL19 and CCL21 are the unique ligands for CCR7, contributing to DC migration into the dLN, we treated mice with CCL19- and CCL21-neutralizing Abs prior to infection and subsequently measured bacterial burdens in dLN. Our findings showed that the dLN of anti-CCL21-treated mice had significantly lower bacterial burdens as compared to mice received control IgG ( Figure 5F ). Given that Ot-infected DCs are known to have the ability to migrate from inoculation site into dLN via lymphatic vessels (14), we used fluorescent microscopy and found less Ot (red) in the dLN of CCR7^-/- mice at day 3 as compared to that of WT mice. Moreover, most bacterial staining was clustered in LYVE-1⁺ (green) lymphatic vessels ( Figure 5G ), indicative of early endothelial infection in dLN by Ot bacteria.