The wild-type strain S. suis 05ZYH33 is a clinical isolate that caused an STSLS outbreak in Sichuan Province of China. S. suis was grown in Todd–Hewitt Broth (THB) medium at 37°C and were harvested for experimentation at the mid-log growth phase bacteria. 6 to 8 weeks old C57BL/6J female mice were purchased from SPF (Beijing, China) Biotechnology Co., Ltd. C57BL/6J Fpr2 ^-/- mice was generated as indicated in previous studies (13). In brief, Fpr2-deficient mice were generated using the Cre/LoxP system. Mice harboring a floxed allele of Fpr2 (Fpr2^loxP/⁺ ) were obtained from Cyagen Biosciences (Guangzhou, China); these mice were interbred to generate Fpr2 ^loxP/loxP mice. Fpr2 ^loxP/loxP mice were then mated with EIIa-cre transgenic mice to generate a null allele of Fpr2 (Fpr2 ^+/- ). Fpr2 ^+/- mice were identified by PCR genotyping using multiple primers (mFpr2_F1 [CTCATACGCATTTGCTGTCTTCACAC], mFpr2_R1 [TCCAATTATATCCCTTTCATGGCAAAC], and mFpr2_F3 [ACAAGGGCCTGCATGTGCCCTCTG]). Finally, Fpr2 ^+/- mice were interbred to generate Fpr2^-/- mice.

For the STSLS model, a standard bacterial dose (2 × 10⁷ CFU) of bacterial suspensions or a vehicle solution (THB) were infused intraperitoneally in WT or Fpr2^-/- mice. In survival experiments, high dose (5 × 10⁷ CFU) of S. suis were used. Peritoneal lavage fluid (PLF), blood, and tissues were obtained for analysis at the indicated time point during infection. Bacterial viability was monitored by plating serial dilutions of blood or PLF. Cytokine and chemokine levels and hematoxylin and eosin (H&E) stain were also performed at 12 h post-infection. All the mice were monitored up to 5 d and clinical score were assigned according to the scoring criteria developed in a previous study (14): 4 = dead; 3 = non-responsive or walking in circles; 2 = responds only to repeated stimuli; 1 = ruffled coat and slow response to stimuli, and 0 = normal response to stimuli. For the meningitis model, mice were inoculated with 10 ul of bacterial suspension (1.25 × 10⁵ CFU) suspension intracisternally after general anesthesia (pentobarbital sodium, 50 mg/kg). Mice were monitored up to 72 h to record deaths. Bacterial load in the brain and blood were evaluated at 14 h, when most animals developed clinical symptoms.

For the interference test, Fpr2/ALX antagonist, WRW4 (Sangon, China) and BOC-2 (N-t-Boc-Phe-Leu-Phe-Leu-Phe; Sangon, China,10 μg/kg; i.p.) were administered after 1 h of infection. In the remaining experiments, animals were treated with AnxA1 (Biobry, UK, 50 μg/kg; i.p.) or LXA4 (Cayman Chemical Company, USA; 2.5 μg/kg; i.p.) 1 h after infection (15).

26 mg/kg of anti-mouse Ly6G (BioXcell, USA, clone:1A8), InVivoMAb, or IgG2a isotype control (BioXcell, USA, clone:2A3) were injected into mice intraperitoneally (i.p.) before infection (16).

To determine the activity of neutrophils or macrophage recruitment, the PLF of mice was collected and analysis by FACS. At each indicated time point, animals were euthanized by carbon dioxide asphyxiation, and the abdomen gently massaged. 10 ml sterile harvest solution (PBS+EDTA) was injected intraperitoneally, then the peritoneal fluid was withdrawn and centrifuged (10 min, 200 × g) to collect recruited leukocytes. After quantifying the absolute number, cells were treated with an FcR-blocking reagent (CD16/CD32, BD Pharmingen) for 15 min, then cells incubated with mixed antibodies: FITC-conjugated anti-mouse CD45 (N418, Biolegend, USA); PerCP-Cy5.5-conjugated anti-mouse CD11b (M1/70 Biolegend, USA) APC-conjµgated anti-mouse F4/80 (BM8, Biolegend, USA), PE-conjugated anti-mouse Ly6G (RB6-8C5 Biolegend, USA), PE-Cy7-conjµgated anti-mouse Ly6C (HK1.4 Biolegend, USA), and APC-Cy7-conjµgated fixable viability dye eFluor780 (Invitrogen). After washing, cells were suspended in sorting buffer for FACS analysis. Data analysis was performed using BD Verse software.

The ProcartaPlex™ Multiplex Immunoassay (EPXR360-26092-901, Invitrogen) was used for measuring cytokine and chemokine levels in the mouse serum after infection. The test was performed according to the manufacturer’s instructions

Livers and spleen from infected mice were removed and fixed in formalin overnight before embedding in paraffin. Slides were stained with hematoxylin and eosin, and the histopathology of tissues was analyzed using microscope.

The cells in PLF of infected mice were collected and pipetted into the wells of Lab-Tek Chamber Slides (ThermoFisher Scientific). After a 5 min incubation, detached cells were removed from each well. The cells were incubated with 4% paraformaldehyde for 10 min at room temperature, then permeabilized (0.25% Triton X-100) and blocked (1% BSA, 1 h, RT). Cells were stained with anti-mouse elastase (Bioss, dilution 1:200) and detected with Alexafluor 488-labeled secondary goat-anti-rabbit Fab antibody fragments (Life Technologies, dilution 1:1000). The slides were counterstained with DAPI by Prolong Diamond mounting medium (Invitrogen). Analysis was performed by confocal microscopy (Nikon Eclipse Ti).

Cell isolation and identification were conducted based on previous research (17). In brief, mice were injected with 1.0 ml casein solution into the peritoneal cavity, and a second injection was performed after one night. The peritoneal fluid was collected 3 h after the second injection. Peritoneal exudate cells were mixed with 9 ml of Percoll gradient solution in a 10-ml ultracentrifuge tube. After ultracentrifuging (20 min at 60,650 × g 4°C), neutrophils in the second opaque layer were collected. Cell purity and viability were determined by FACS.

Neutrophils (2×10⁶ cells) were incubated with S. suis (MOI=5) for 1 or 3 h at 37˚C. A portion of the supernatants were taken and cultured overnight on THB plates at certain points to quantify non-engulfed extracellular bacteria. Part of the mixture was washed to remove the extracellular non-engulfed S. suis, then cells were laid onto 3 microscopic slides. Liu’s Stain was used to visualize cell morphology and 200 neutrophils per slide were counted for bacterial clearance rate. Bactericidal activity of neutrophils was determined by counting the number of live extracellular bacterial colonies, and the bacterial clearance rate was determined using Romanowsky Stain. Three replicate experiments were done for each experiment.

The data in this study were statistically analyzed by GraphPad Prism 8 software and all data were present as the mean ± standard deviation (mean ± SD) unless otherwise noted. A two-way analysis of variance was used for the clinical scoring of mice. For the survival experiment, data was analyzed using the Kaplan-Meier method and Log-Rank test. For bacterial load, cytokine expression, and immune cells detection, data were analyzed with the Mann-Whitney U test; P <0.05, was considered significant.

To analyze the role of Fpr2 in the pathogenesis of STSLS, the expression of Fpr2 was detected in mice inoculated with S. suis intraperitoneally. Survival was monitored in WRW4 and BOC-2 (Fpr2 antagonist) intervention experiments. The increased expression of Fpr2 gene in the spleen implied a participatory role of Fpr2 during infection ( Figure 1A ). In the intervention trial, inhibiting Fpr2 with an antagonist significantly reduced the bacterial load and mortality of mice with STSLS ( Figures 1B, C ). Fpr2^-/- mice were used for a more detailed study of S. suis infection. After monitoring the bacterial load and mortality in WT and Fpr2^-/- STSLS mice, lower mortality rates and bacterial load were observed in the Fpr2^-/- mice ( Figures 1D-F ). A significant difference in bacterial load was observed in 6 h and 12 h post-infection both in blood and PLF ( Figures 1E, F ). These results indicate that Fpr2 exacerbates the acute symptoms of STSLS.Cytokine production, pathological changes, body temperature changes, and clinical evaluation were also monitored in infected mice. Mice deficient in Fpr2 had a lower inflammatory response supported by decreased soluble mediators and alleviated tissue injury at 12 h post-infection ( Figures 2A, B ). The soluble mediators of PLF in infected mice at 3 and 6 h were also detected, the deceased cytokines were significantly observed at 6 h in Fpr2^-/- mice, especially in CXCL1, CXCL2 and IL-6 ( Figure S1 ). Hepatocyte apoptosis and splenocyte necrosis were only observed in WT mice ( Figure 2B ). Moreover, in contrast with Fpr2^-/- mice, induction of STSLS yielded worse clinical scores and prolonged hypothermia for WT mice. ( Figures 2C, D ). Collectively, these data indicate a potential contributing role of Fpr2 in STSLS pathogenesis.

We focused on leukocyte activity during infection to understand the role of Fpr2 in the STSLS-induced inflammatory response. WT and Fpr2^-/- mice were inoculated with standard dose of S. suis (2 ×10⁷ CFU) intraperitoneally. Counts of leukocyte in the PLF were determined during infection. Macrophages/monocytes and neutrophils are the two main immune cells in the PLF after infection ( Figure 3A ). Flow cytometry analysis revealed striking neutrophilia in Fpr2^-/- mice at 3 and 6 h ( Figure 3B ). No significant difference was observed in macrophage counts between the two mice genotypes ( Figure 3C ). Higher level of NETs was also found in the plasma and PLF of Fpr2^-/- mice at 6 h post-infection ( Figures 3D, E ), which was further confirmed by immunofluorescence results. The filamentous chromatin fibers, areas of decondensed extracellular thread-like DNA colocalizing with neutrophil elastase, as visualized by confocal microscopy depict the NETs. More NETs construction was observed in the PLF of Fpr2^-/- mice ( Figure 3F ). Together, these data suggest that Fpr2 impaired the recruitment of neutrophils and production of NETs during STSLS.

Having established active neutrophil recruitment and higher levels of NETs structure in Fpr2^-/- mice suffering from STSLS, we further explored how Fpr2 affects the morphology or function of neutrophils. The bactericidal ability and oxidative stress properties of WT and Fpr2^-/- neutrophils were evaluated after cells were treated by S. suis (MOI=5) in vitro. To determine bactericidal ability, the bacterial clearance rate was evaluated at 1 and 3 h ( Figure 4A ). The extracellular bacterial load, which reflects the bactericidal capacity of WT and Fpr2^-/- neutrophils, was also evaluated ( Figure 4B ). Bacterial clearance rate calculation of stimulated neutrophils revealed that Fpr2^-/- cells exhibited lower bacterial clearance rate than WT mice at 1 h post treatment. However, this had no influence on the bactericidal ability of neutrophils, as the phagocytic bacteria and extracellular bacteria between the cells of the two genotypes showed no significant differences at 3 h. Regarding oxidative stress properties, ROS production was monitored when WT and Fpr2^-/- cells were treated with S. suis at different MOIs. Although there was higher light intensity in the WT group before 25 min, as the infection progressed no differences were observed between the cells of the two genotypes at 40 min ( Figure 4C ). No significant difference was observed in the NETs production experiment when the two genotype neutrophils were treated with S. suis for 3 h in vitro. ( Figure 4D ). These results indicate that Fpr2 did not directly affect ROS and NETs production of neutrophils in vitro, and that the improved leukocyte infiltration of Fpr2 deficient mice may account for the higher levels of NETs observed in vivo.

Administration of LXA4 or AnxA1 efficiently alleviates neutrophil accumulation, reduces inflammation, and attenuates many diseases. Our previous study elucidated the potential therapeutic effect of AnxA1 in S. suis meningitis. In this study, LXA4 and AnxA1 were used as an intervention in STSLS infection to explore the potential effects of LXA4 and AnxA1 in STSLS treatment. Analysis of cytometer results showed the potential anti-recruitment function of LXA4 and AnxA1 in WT mice at 3 and 6 h post-infection, although no effect was observed in Fpr2^-/- mice. ( Figures 5A-C ). The bacterial load in the blood and PLF at 6 h was monitored. High bacterial loads were observed in LXA4 or AnxA1 treated WT mice, but this effect was not observed in Fpr2^-/- mice ( Figures 5D, E ). In order to confirm that Fpr2 primarily affects the activity of neutrophils in STSLS, we used anti-Ly6G antibody to block the neutrophils. Treatment with anti-Ly6G eliminated observed differences between the WT and Fpr2^-/- mice ( Figure 5F ). Compared to the isotype group, pretreatment with anti-Ly6G increased mortality and accelerated death from STSLS. Neutrophil recruitment during the early stages of STSLS plays an indispensable role in bacterial clearance, which is a key step regulated by Fpr2 during the inflammation response.

Clinical symptoms score, lethality test and histopathology changes in STSLS, were also assessed when LXA4 and AnxA1 were used. Treatment with LXA4 or AnxA1 in WT mice increased the clinical symptom score and lethality in these mice, indicating a detrimental role of anti-recruitment drugs in STSLS ( Figures 6A-C ). Exaggerated tissue damage was also observed in WT but not Fpr2^-/- mice ( Figure 6D ). Those data further demonstrate that neutrophil recruitment during the early stages of infection plays a crucial role in controlling bacterial proliferation and dissemination. The anti-recruitment mediators LXA4 and AnxA1 aggravated tissue damage and increased the inflammatory response in STSLS via Fpr2.

Previous studies from our laboratory have reported the beneficial role of Fpr2 in bacterial clearance and inflammation resolution during S. suis meningitis. Fpr2^-/- mice were highly susceptible to S. suis meningitis, and displayed increased bacterial dissemination and neutrophil migration. AnxA1 attenuated inflammatory responses and neutrophil invasion through Fpr2 during S. suis meningitis. When compared with the current study on the STSLS model, both studies revealed an anti-recruitment function of Fpr2 during infection, but opposite effects on disease development in the different infection models. We speculated that neutrophil recruitment may play a different role in bacterial dissemination and disease treatment between meningitis and STSLS. Dexamethasone (dex) is a powerful anti-inflammatory compound inhibiting inflammatory cell recruitment and production of proinflammatory cytokines. Thus, dexamethasone was used to pretreat mice suffering from S. suis induced STSLS or meningitis to investigate the potential role of early inflammation at different infection sites. Injection of dexamethasone decreased the mortality of Fpr2^-/- mice in the meningitis model, but accelerated the death in Fpr2^-/- mice with STSLS ( Figures 7A, B ). Similar results were also obtained in WT mice ( Figure S2 ). These results indicate that inflammation during the early stages of disease development is detrimental to host survival in the meningitis model, but beneficial to survival in the STSLS model. This finding was aligned with the results from a DNA intervention experiment which clarified the dual role of NETs during STSLS and meningitis. DNase I (DNase), which effectively degrades NET-associated DNA, was infused at the time of infection to prevent the accumulation of NETs in both the meningitis and STSLS models. Survival and bacterial load of infected mice in both infected models were evaluated. DNase significantly increased the mortality of the STSLS mice, but reduced the mortality of the meningitis mice ( Figures 7C, D ). Additionally, DNase treatment significantly aggrandized the bacterial load in the STSLS mice, but had an opposite effect on the meningitis group ( Figures 7E, F ). This suggests that NETs play protective roles in bacterial cleaning and host individual survival in the initial stages of STSLS, whereas it interferes with bacterial clearance in the CNS and aggravates the severity of meningitis. These results highlight a dual role of neutrophils and NETs on S. suis proliferation and host protection in meningitis and STSLS models.