Staphylococcus aureus strain UAMS-1 was grown on Luria broth (LB) agar plates overnight at 37°C in 5% CO₂ and was cultured in LB on an orbital rocker at 37°C in 5% CO₂ overnight. Prior to use in our in vivo and in vitro studies, bacteria were grown to mid-log phase in LB at 37°C with 5% CO₂ and the number of colony forming units (CFU) was determined by spectrophotometry using a Genespec3 spectrophotometer (MiraiBio Inc.).

In these studies, we have used a murine model of S. aureus osteomyelitis, previously employed by our laboratory (Marriott et al., 2004, 2005; Chauhan and Marriott, 2010), that reproduces the pathology of human post-traumatic osteomyelitis (Spagnolo et al., 1993). Mice were anesthetized with inhalant isoflurane and the left femur was surgically exposed and abraded using a high-speed drill with a burred bit. Femurs were inoculated with S. aureus containing agarose beads, a method that induces localized infection rather than establishing a systemic bacterial infection. Agarose beads containing S. aureus were prepared by combining 1.4% low melt agarose cooled to 40–42°C with 1 × 10⁹ CFU S. aureus, followed by mineral oil addition with vigorous shaking on ice. Agarose beads were washed, stored on ice, and then administered to the abraded left femurs. Following application, the muscle fascia was sutured and the surgical incision closed with staples. At 3 days following S. aureus challenge, the infected left femur and contralateral uninfected femurs were isolated, weighed, and total protein extracted using T-PER. The protein concentration of IFN-β was evaluated by specific capture ELISA and normalized to femur weight.

Whole calvaria were isolated from two to three-day old murine neonates essentially as we have described (Bost et al., 2001; Gasper et al., 2002; Madrazo et al., 2003) with the following modifications. Briefly, primary osteoblasts were isolated from whole calvaria through six sequential 15-min trypsin/collagenase P digestions and the cells were maintained in DMEM supplement with 10% FBS and 1% penicillin/streptomycin at 37°C in a 5% CO₂ atmosphere. At 24 h, primary osteoblasts were plated in 6-well plates at a density of 2 × 10⁵ cells per well and differentiated in αMEM supplemented with 10% FBS, 0.1 M ascorbic acid, 1 M β-glycerol phosphate, and 100 U/mL penicillin/100 μg/mL streptomycin at 37°C in a 5% CO₂ atmosphere. The differentiation media was changed every other day until the experiment/infection at day 10, and the presence of mature osteoblasts was confirmed using an alkaline phosphatase staining kit (Abcam) and positive staining assessed by light field microscopy as we have described (Johnson et al., 2022).

Cells were infected with S. aureus at multiplicities of infection (MOI) between 50:1 and 150:1 bacteria to each bone cell, as indicated, in antibiotic-free media at 37°C with 5% CO₂ for 2 h, and then the media was replaced with media containing 100 U/mL penicillin/100 μg/mL streptomycin to kill residual extracellular bacteria for the duration of the experiment. At the indicated time points following bacterial infection, cell supernatants and whole cell protein isolates were collected. In studies evaluating intracellular S. aureus viability, infected osteoblasts (2 × 10⁶ cells) were lysed using 1% saponin in antibiotic-free media and the number of liberated viable bacteria was determined by serial colony counts.

Primary murine osteoblasts were infected with S. aureus at a bacteria-to-osteoblast ratio of 75:1 and RNA was isolated using a PureLink RNA mini kit at 4 h post-infection. The Genomic Sequencing and Analysis Facility at the University of Texas at Austin prepared Tag-Seq libraries using an established method (Meyer et al., 2011) and sequenced libraries using an Illumina HiSeq 3,000. The 100 bp, single-end reads were trimmed and quality-filtered using the FastX-toolkit (Pearson et al., 1997). The trimmed SOLiD reads were mapped to the reference genome (GRCm38.p6) using BIOSCOPE V. 1.3.1 software. Differential expression analysis was performed using R package DESeq2 (Love et al., 2014). Outlier analysis was performed using the ArrayQualityMetrics R package (Kauffmann et al., 2009). Normalized and rlog-transformed counts were used in all downstream analysis (_R3.5.0, R Core Team, 2015), following generation of a read-counts-per-gene file retaining only transcripts mapped to a single gene (included as supplemental information for (Sipprell et al., n.d.)). Gene expression heatmaps were made using the PHEATMAP package clustering expression patterns hierarchically (Wickham, 2009). A log₂ ratio of read counts above 0.5 was considered statistically significant (padj <0.1). Pairwise comparisons between each treatment at each timepoint using Wald tests in DESeq identified significantly differentially expressed genes (DEGs) with p-values calculated using the Benjamini-Hochberg procedure. False discovery rate (FDR) adjustment for multiple testing resulted in 606 DEGs. Of the 606 differentially expressed genes identified, 135 genes displayed greater than a 2-fold difference in expression. Genes encoding type I IFN signaling components and IFN-stimulated genes (ISGs) were identified via gene ontology and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis conducted using ShinyGO 0.76. For this analysis, all differentially expressed genes were converted to ENSEMBL gene IDS or STRING-db protein IDs. An FDR cutoff value of 0.05 and a pathway size range of 2 to 2000 was applied. The top 10 pathways were first selected by FDR and then sorted by fold enrichment. The original datasets are available in the Gene Expression Omnibus (GEO) publicly accessible repository under the accession number GSE 217455.

Specific capture ELISAs were conducted to quantify cell release of IFN-β and production of IFIT3 in response to S. aureus infection. A polyclonal goat anti-mouse IFN-β capture antibody (BioLegend, Cat # 519202, Clone Poly5192; 1 μg/mL) and an Armenian hamster biotin anti-mouse IFN-β detection antibody (BioLegend, Cat # 508105, Clone MIB-5E9.1; 2 μg/mL), were used to quantify mouse IFN-β concentrations. Streptavidin-HRP (R&D Systems; Lot # P198628) was added prior to the addition of tetramethylbenzidine substrate. The colorimetric reaction was stopped with 1:30 H₂SO₄ and the absorbance was measured at 450 nm. Recombinant murine IFN-β (BioLegend) was utilized to generate standard curves and extrapolation of the absorbance to the standard curve was used to determine the concentration of IFN-β in the cell supernatants. IFIT3 production was analyzed using a commercially available kit (MyBiosource) according to manufacturer guidelines.

Total cell lysates were evaluated by immunoblot analysis for the presence of IFIT1, IFIT3, and PLSCR1. Immunoblots were incubated with a monoclonal antibody directed against IFIT1 (Novus Biologicals, clone OTI3G8), a polyclonal antibody directed against IFIT3 (Invitrogen), or a monoclonal antibody directed against PLSCR1 (Invitrogen, clone ARC2028), overnight at 4°C. Immunoblots were then washed and incubated with a horseradish peroxidase (HRP)-conjugated anti-mouse or anti-rabbit secondary antibody, and bound antibody was detected with a Pierce enhanced chemiluminescence (ECL) immunoblotting substrate (ThermoFisher Scientific). Immunoblots were re-probed with a mouse monoclonal antibody against β-actin (Abcam, Cat # 49900; 0.13 μg/mL) to assess total protein loading. The immunoblots shown are representative of at least three separate experiments. Densitometric analysis was conducted using ImageLab software (BioRad) and IFIT1, IFIT3, and PLSCR1 protein levels were normalized to β-actin expression.

Data is expressed as the mean ± standard error of the mean (SEM). Commercially available software (GraphPad Prism, GraphPad Software, La Jolla, CA, United States) was used to conduct statistical analyses including Student’s t test, one-way analysis of variance (ANOVA) with Dunnet’s post hoc test, or two-way ANOVA with Tukey’s post hoc test as appropriate. In our RNA Tag-Seq analyses, differential expression analysis was performed using R package DESeq2. A log₂ ratio of read counts above 0.5 was considered statistically significant (padj <0.1). Gene ontology analysis was conducted using ShinyGO 0.76. A false discovery rate (FDR) cutoff value of 0.05 and a pathway size range of 2 to 2000 was applied. The top 10 pathways were first selected by FDR and then sorted by fold enrichment. FDR was calculated based on nominal p-value from the hypergeometric test. Fold enrichment was defined as the percentage of genes in our list belonging to a pathway divided by the corresponding percentage in the background. In some analyses, Chi-squared and Student’s t-tests were used to determine whether differentially expressed genes have special characteristics when compared with genes in the reference genome. For all experiments, a value of p of less than 0.05 was considered statistically significant.

To begin to assess the ability of bone cells to produce type I IFNs at sites of S. aureus infection, we have employed an established mouse model that reproduces post-traumatic staphylococcal osteomyelitis (Spagnolo et al., 1993; Marriott et al., 2004, 2005) and assessed the expression of IFN-β in infected femur bone tissue. As shown in Figure 1A. S. aureus infection elicited a marked increase in the level of IFN-β in bone tissue 3 days post-infection in comparison to contralateral uninfected control bone tissue.

To determine the contribution made by resident bone cells to IFN-β production, we first assessed the changes in mRNA expression by primary murine osteoblasts following S. aureus challenge via RNA Tag-Seq analysis. Our recent gene ontology analysis demonstrated that osteoblasts displayed enrichment of gene products associated with type I IFNs, ISGs, and/or IFN-mediated cellular effects, following S. aureus challenge (Sipprell et al., n.d.). Specifically, osteoblasts showed a marked increase in the expression of mRNA encoding IFN-β as early as 4 h post infection (Figure 1B). Importantly, we have confirmed that such upregulation results in type I IFN protein production with the demonstration that S. aureus challenge elicits the rapid release of IFN-β by osteoblasts and does so in a dose-dependent manner (Figure 1C). Furthermore, KEGG pathway analysis has indicated the upregulated expression of mRNA encoding microbial pattern recognition receptor (PRR) components such as RNA polymerase III, retinoic acid inducible gene-I (RIG-I), myeloid differentiation primary response 88 (MyD88), and TANK-binding kinase 1 (TBK1), and IFN signaling components including IFNAR1 and STAT1/2 (Figure 1D).

Our RNA Taq-Seq analysis also indicated the upregulated expression of an array of ISGs in primary murine osteoblasts following S. aureus infection. Specifically, osteoblasts showed marked increases in the expression of mRNA encoding IFIT1, IFIT3, SLFN2, IRGM2, MX2, PLSCR1, IFI205, and IGTP, as early as 4 h post-infection (Figure 1B). We then determined whether such upregulation results in increased ISG protein product levels in infected osteoblasts. As shown in Figure 2A S. aureus challenge induced the production of IFIT1 by osteoblasts in a time and dose-dependent manner. Similarly, infection elicited an increase in IFIT3 protein expression by osteoblasts over constitutive levels as determined by immunoblot analysis (Figure 2B) and specific capture ELISA (Figure 2C). In contrast, S. aureus infection failed to elevate PLSCR1 expression in osteoblasts above the robust levels seen in uninfected cells (Figure 2D) despite an apparent increase in mRNA expression encoding this ISG product (Figure 1B).

To determine whether S. aureus infection directly upregulates the expression of ISGs in osteoblasts or does so indirectly via the production and release of IFN-β, we have assessed the effect of IFNAR blockade on S. aureus-induced IFIT1 expression. As shown in Figure 3A, the presence of IFNAR blocking antibody significantly attenuated the increase in IFIT1 protein expression associated with S. aureus infection compared to infected osteoblasts exposed to an isotype control antibody. This data indicates that elevated ISG expression in S. aureus challenged osteoblasts occurs, in large part, to the actions of IFN-β produced by these cells.

The ability of osteoblasts to produce IFN-β in sufficient quantities to induce ISG expression in an autocrine and/or paracrine manner following bacterial challenge raises the possibility that such responses serve to limit S. aureus infections. To begin to test this hypothesis and assess the relevance of osteoblast type I IFN responses to this bacterium, we have determined whether IFN-β impacts the ability of S. aureus to persist/replicate intracellularly in these bone cells. As shown in Figure 3B, treatment of S. aureus-infected osteoblasts with recombinant IFN-β results in rapid and time-dependent reductions in the number of viable intracellular bacteria. Together, these results support the contention that osteoblasts produce IFN-β in response to S. aureus to induce antibacterial responses via IFNAR that limit intracellular bacterial burden.