All experiments were conducted with the approval of Virginia Tech’s Institutional Animal Care & Use Committee (IACUC) under protocol number 24-060. Experiments using CHIKV and MAYV were performed in a BSL-3 or BSL-2 facility, respectively, in compliance with CDC and NIH guidelines and with approval from the Institutional Biosafety Committee (IBC) at Virginia Tech.

C57BL/6J mice (strain# 000664) and colony-stimulating factor 3 receptor gene knockout mice (B6.129X1 (Cg)-Csf3r^tm1Link/J; herein referred to as G-CSFR^-/-) (23) were purchased from The Jackson Laboratory at 6–8 weeks of age. Heterozygous G-CSFR^+/- mice were bred at Virginia Tech to obtain homozygous G-CSFR^-/- mice. Representative genotyping images and the primers used to confirm homozygous colonies are presented in Supplementary Figure S1 . Homozygous G-CSFR^-/- mice were used in all studies. Mice were kept in groups of four or five animals per cage at ambient temperature with ad libitum supply of food and water.

Vero cells were obtained from the American Type Culture Collection (ATCC; Manassas, VA) and grown in Dulbecco’s Modified Eagle’s Medium (DMEM, Gibco) with 5% fetal bovine serum (FBS, Genesee), 1 mg/mL gentamicin (Thermo Fisher), 1% non-essential amino acids (NEAA, Sigma) and 25 mM HEPES buffer (Genesee) at 37°C with 5% CO₂. CHIKV SL-15649 (ECSA lineage), rescued from an infectious clone, was a gift from Dr. Mark Heise (the University of North Carolina at Chapel Hill (33), and MAYV strain TRVL 4675 was derived from an infectious clone (34, 35). Virus titers were determined by plaque assay on Vero cells, as previously described (36). The limit of detection for plaque assays is defined as one plaque-forming unit at the lowest dilution tested; samples below the limit of detection were assigned an arbitrary value of nine plaques, one dilution below what was tested, which was typically an undiluted sample.

Mice were injected in both hind footpads with 10² PFU or 10⁵ PFU of CHIKV SL-15649 or 10⁴ PFU of MAYV in 50 μL viral diluent in each foot (37, 38). All virus dilutions were made in RPMI-1640 media with 10 mM HEPES and 2% FBS, hereafter referred to as viral diluent. Mice were monitored for disease development following infection through daily weighing, and footpad swelling was measured using a digital caliper. Blood was collected via submandibular bleed for serum isolation to determine viremia and cytokine levels. The limit of detection (LOD) for plaque assays was defined as one plaque-forming unit at the lowest dilution tested. Samples below the limit of detection were assigned an arbitrary value of nine plaques, corresponding to one dilution below the tested range, which was typically an undiluted sample. At sixteen or twenty-one-days post-infection (dpi), mice were euthanized, and blood was collected to isolate immune cells.

We quantified G-CSF levels in the serum of mock- and CHIKV/MAYV-infected animals using the mouse Luminex XL cytokine assay (Bio-Techne) and the G-CSF DuoSet ELISA kit (Catalog# DY414-05, R&D Systems) according to the manufacturer’s instructions. Serum IFNγ and TNFα levels were measured using DuoSet ELISA kits, also from R&D Systems. A standard curve was generated using the optical density values of the standards, which were used to estimate the cytokine levels in each sample.

For in vivo interferon-α/β receptor (IFNAR) blockade, C57BL/6J and G-CSFR^-/- mice were intraperitoneally (i.p.) inoculated with 0.1 mg of MAR1–5A3, anti-IFNAR antibody (Leinco Technologies, Fenton, MO) (39) one day prior to infection. Mice were then inoculated with 10² PFU of CHIKV SL-15649 in 50 μL of viral diluent in both hind feet and monitored for disease development until 21 dpi.

Mouse blood leukocytes were isolated using Mono-Poly resolving medium (M-P M, MP Bio, Cat. No. 091698049) according to the manufacturer’s instructions. Briefly, blood was mixed with an equal volume of PBS and layered slowly onto M-P M followed by centrifugation at 300 × g for 30 min in a swinging bucket rotor at room temperature (20–25°C). We collected cell layers between the plasma and M-P M to isolate leukocytes and added them to a 15 mL conical tube containing 10 mL of cold 10% FBS containing RPMI-1640 (RPMI-10). Cells were spun at 500 × g for 5 min at 4°C and used for flow cytometry.

Single cell suspensions were washed with PBS and resuspended in 100 μL Zombie aqua cell viability dye solution (1:400 prepared in PBS, Cat. No. 423101, BioLegend) and incubated at room temperature for 15–30 minutes. 200 μL flow cytometry staining (FACS) buffer (PBS containing 2% FBS) was added and centrifuged at 500 × g for 5 min at 4°C. The resulting cell pellet was resuspended in FACS buffer with 0.5 mg/mL rat anti-mouse CD16/CD32 Fc block (Cat. No. 553142, BD Biosciences) and incubated for 15 min on ice to block Fc receptors. For extracellular staining, a combined antibody solution was prepared in FACS buffer with fluorophore-conjugated antibodies: anti-mouse Alexa fluor 700 CD45 (Cat. No. 103128, BioLegend), anti-mouse PerCP/Cyanine 5.5 CD11b (Cat. No. 101227, BioLegend), anti-mouse APC Ly6G (Cat. No. 127614, BioLegend), anti-mouse PE Ly6C (Cat. No. 128007, BioLegend), anti-mouse PE CD3 (Cat. No. 100206, BioLegend), anti-mouse PerCP/Cyanine 5.5 CD4 (Cat. No. 116012, BioLegend), and anti-mouse FITC CD8a (Cat. No. 100706, BioLegend), and APC CD19 (Cat. No. 152410, BioLegend). 100 μL antibody cocktail was added to the single cell suspension, mixed, and incubated for 30 min on ice. Cells were washed with FACS buffer twice, and 100 μL 4% formalin (Thermo Fisher Scientific, Ref. No. 28908) was added to fix the cells. After 15 min incubation at room temperature, cells were washed with FACS buffer, resuspended in 100-200 μL PBS, and covered with aluminum foil before flow cytometry analysis. For each antibody, single color controls were run with Ultracomp ebeads (Cat. No. 01-2222-42, Thermo Fisher Scientific). The stained cells were analyzed using the FACSAria Fusion Flow cytometer (BD Biosciences).

Statistical comparisons were performed using GraphPad Prism version 10. Data are presented as mean ± standard deviation. The statistical tests used to analyze data are described in figure legends.

G-CSF is elevated during arthritogenic alphavirus infection in humans (25–28) and mice (29). To validate G-CSF levels following arthritogenic alphavirus infection, we infected mice with CHIKV and MAYV and collected serum at 2 and 7 dpi, the peak of viremia and footpad swelling, respectively (37). At 2 dpi, G-CSF levels increased 3-fold (p < 0.0001) following MAYV infection ( Figure 1A ) and 14-fold following CHIKV (p = 0.0079; Figure 1B ) compared to mock-inoculated controls. G-CSF levels remained higher at 7 dpi in MAYV (p=0.04) and CHIKV (p=0.01) infected mice compared to mock-infected groups ( Figures 1C, D ). These data validate that G-CSF levels increase following MAYV or CHIKV infection in mice.

G-CSFR is present on the surface of granulocyte precursors, initiating signal transduction and activation of cellular pathways that drive proliferation, differentiation, and maturation of granulocytes, especially neutrophils (22, 23). To assess the impact of G-CSF signaling on arthritogenic alphavirus disease outcomes, we inoculated mice genetically lacking the G-CSF receptor (G-CSFR^-/-) with MAYV or CHIKV. Following MAYV infection, we observed sustained weight loss in G-CSFR^-/- mice compared to WT mice until 16 dpi ( Figure 2A ). No differences were observed in footpad swelling ( Figure 2B ) and virus replication ( Figure 2C ). CHIKV-infected G-CSFR^-/- mice showed a similar sustained weight loss compared to the WT group until 21 dpi ( Figure 2D ). As with MAYV, no differences were observed in footpad swelling ( Figure 2E ) and virus replication ( Figure 2F ) between the two groups. These data suggest that G-CSFR plays a role in recovery from systemic arthritogenic alphavirus disease.

Next, we monitored the impact of the G-CSFR deficiency on the immune cell profile throughout infection. We isolated blood leukocytes at various time points following CHIKV and MAYV infection and profiled immune cell populations using flow cytometry (gating strategy presented in Supplementary Figure S2 ). As expected, G-CSFR deficiency resulted in a significant reduction in the proportion of neutrophils in the blood compared to wild-type (WT) mice during MAYV and CHIKV infection ( Figures 3A, D ). We also observed a significantly higher percentage of anti-inflammatory monocytes (Ly6C^lo) in the blood of MAYV-infected G-CSFR^-/- mice at 16 dpi ( Figure 3B ) and CHIKV-infected G-CSFR^-/- mice at 7 and 21 dpi ( Figure 3E ) compared to WT controls. Pro-inflammatory monocytes (Ly6C^hi) were higher in MAYV-infected G-CSFR^-/- mice at all time points tested ( Figure 3C ) and at 14 dpi in CHIKV-infected G-CSFR^-/- mice ( Figure 3F ); these cells trended higher in CHIKV-infected G-CSFR^-/- mice at 7 dpi, but the difference did not reach significance (p = 0.059). These data demonstrate that G-CSFR plays an important role in regulating myeloid populations in the blood during arthritogenic alphavirus infection.

We also determined the effect of G-CSFR^-/- deficiency on B and T cell dynamics during arthritogenic alphavirus infection. We observed no impact on CD4 and CD8 T cells in MAYV- and CHIKV-infected animals ( Supplementary Figures S3A, B, D, E ). In contrast, B cells were higher at 6 dpi in MAYV-infected G-CSFR^-/- mice and at 7 and 21 dpi in CHIKV-infected G-CSFR^-/- mice ( Supplementary Figures S3C, F ). To assess whether differences in B cell levels altered virus-specific antibody responses, we measured neutralization of MAYV at 13 dpi using a plaque reduction neutralization test with an 80% threshold (PRNT₈₀). No differences were observed between the groups ( Supplementary Figure S4 ). We also measured serum levels of IFNγ and TNFα at 6 dpi in MAYV-infected animals and observed no differences ( Supplementary Figure S5 ). These data demonstrate that G-CSFR^-/- deficiency has minimal impacts on the proportion of T cells during infection but leads to a significant increase in B cells following infection with both viruses.

Alphaviruses are highly sensitive to type I IFN restriction, and mice deficient in IFNAR-signaling are highly susceptible to disease (20, 40). Monocytes are increased in G-CSFR^-/- mice ( Figure 3 ) following CHIKV and MAYV infection and are known to produce type I IFNs in response to CHIKV or RRV infection (31, 32). Thus, we hypothesized that type I IFN might be responsible for the worse systemic disease in G-CSFR^-/- mice, and that blockade of type I IFN signaling might reverse disease outcomes in these mice. To test this, we injected MAR1-5A3 antibody (39) to block IFNAR signaling one day prior to CHIKV infection. We observed a similar pattern of weight loss in CHIKV-infected WT and G-CSFR^-/- groups, indicating that worse systemic disease observed in G-CSFR^-/- mice is mediated by type I IFN signaling ( Figure 4A ). We observed only minimal differences in footpad swelling ( Figure 4B ) and no difference in viremia at 3 dpi ( Figure 4C ) between the groups. These data suggest that G-CSF signaling contributes to the control of systemic arthritogenic alphavirus disease through type I IFN signaling.