Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Apr 1.
Published in final edited form as: Arthritis Rheumatol. 2024 Feb 18;76(4):647–659. doi: 10.1002/art.42765

Short chain fatty acids mitigate osteoclast-mediated arthritic bone remodeling

Katharine Lu Yang 1,*, Briana J Mullins 1,*, Alannah Lejeune 3, Ellie Ivanova 1, Jong Shin 1, Sofia Bajwa 1, Richard Possemato 1, Ken Cadwell 1,2, Jose U Scher 3,4, Sergei B Koralov 1
PMCID: PMC10965381  NIHMSID: NIHMS1947151  PMID: 37994265

Abstract

Objective

To study the effects of Short Chain Fatty Acids (SCFAs) on arthritic bone remodeling.

Methods

We treated a recently described preclinical murine model of psoriatic arthritis (PsA), R26STAT3Cstopfl/fl CD4Cre mice, with SCFA supplemented water. We also performed in vitro osteoclast differentiation assays in the presence of serum-level SCFAs to evaluate the direct impact of these microbial metabolites on maturation and function of osteoclasts. We further characterized the molecular mechanism of SCFAs by transcriptional analysis.

Results

The osteoporosis condition in R26STAT3Cstopfl/fl CD4Cre animals is attributed primarily to robust osteoclast differentiation driven by an expansion of osteoclast progenitor cells (OCPs), accompanied by an impaired osteoblast development. We show that SCFA supplementation can rescue the osteoporosis phenotype in this model of PsA. Our in vitro experiments revealed an inhibitory effect of the SCFAs on osteoclast differentiation, even at very low serum concentrations. This suppression of osteoclast differentiation enabled SCFAs to impede osteoporosis development in R26STAT3Cstopfl/fl CD4Cre mice. Further interrogation revealed that bone marrow derived OCPs from diseased mice expressed a higher level of SCFA receptors than that of control mice and that the progenitor cells in the bone marrow of SCFA-treated mice presented a modified transcriptomic landscape, suggesting a direct impact of SCFAs on bone marrow progenitors in the context of osteoporosis.

Conclusion

We demonstrated how gut microbiota-derived SCFAs can regulate distal pathology, i.e., osteoporosis, and identified a potential therapeutic option for restoring bone density in rheumatic disease, further highlighting the critical role of the gut-bone axis in these disorders.

Graphical Abstract

graphic file with name nihms-1947151-f0001.jpg

INTRODUCTION

Short-chain fatty acids (SCFAs) are microbial metabolites that can robustly influence host physiology. These are volatile free fatty acids with fewer than 6 carbon groups attached to the carboxylic acid moiety, generated as byproducts after the fermentation of dietary fibers by some anaerobic bacteria species such as the Bacteroidetes and the Firmicutes. Acetate, propionate, and butyrate, which contain 1, 2, and 3 carbon group(s) respectively, are the three most well-studied SCFAs. They have a myriad of functional effects on host cells, especially those of the immune system [1–6]. SCFAs act on host cells through several mechanisms. They can bind to membrane G protein-coupled receptors, such as GPR109A, GPR43, and GPR41 The latter two receptors are also known as free fatty acid receptors 2 and 3, respectively [7]. SCFAs can also activate downstream signaling pathways including MAPK, NF-κβ and PI3K [8]. In addition, SCFAs can exert intracellular effects upon crossing the cell membrane with the help of monocarboxylate transporters [9]. Once inside the cell, SCFAs can regulate cell-cycle progression, modify cellular metabolism, and change the transcriptomic landscape by directly inhibiting histone deacetylases (HDACs) [8]. For instance, in regulatory T (Treg) cells, inhibition of HDACs by butyrate leads to the acetylation of histone H3 in the promoter and the conserved non-coding regions of the FoxP3 gene, hence upregulating the expression of this key Treg transcriptional factor [2].

Bone homeostasis is a dynamic process regulated by multi-nucleated myeloid-derived osteoclasts that promote bone resorption and single-nucleated mesenchymal lineage-derived osteoblasts, which promote bone mineralization. Imbalance of these two cell types can lead to pathologic outcomes such as osteoporosis, when the rate of bone resorption exceeds the rate of bone formation. This skeletal disorder is characterized by a reduction of bone mineral density and a collapse of the defined bone microarchitecture. The skeletal system harbors key immune stem cells for hematopoiesis and the immune system can play a pivotal role in regulating bone homeostasis via the cytokines secreted by immune cells, demonstrating the interdependency between the skeletal and immune systems [10, 11].

We have previously established and characterized a novel murine model of psoriatic arthritis (PsA), R26STAT3Cstopfl/fl CD4Cre mice. In this model, a hyperactive mutant allele of a transcription factor STAT3 (STAT3C) is expressed selectively in T cells, thereby promoting augmented Th17 differentiation in the animals [12, 13]. The Th17-driven inflammation in the R26STAT3Cstopfl/fl CD4Cre mice results in dramatic enhancement of osteoclastogenesis, a phenotype that is diminished upon blocking of IL-17 and IL-22 cytokine functions [13]. In the present study, we queried whether SCFAs can be used to ameliorate the osteoporosis phenotype in our preclinical model of PsA. We demonstrate that SCFAs influence the differentiation of osteoclast progenitor cells, revealing a critical gut-bone axis mechanism that may have significant therapeutic implications for skeletal and rheumatic diseases.

MATERIALS AND METHODS

Mice

All experimental mice were housed in NYU Langone Medical Center Skirball animal facility and Jackson laboratory under specific-pathogen-free conditions. All experimental procedures and protocols were approved by the institutional Animal Care and Use Committee (IACUC). R26STAT3Cstopfl/fl mice were generated via gene targeting of Bruce4 ES cells, backcrossed to C57BL/6J mice and subsequently crossed with CD4Cre mice [12]. R26STAT3Cstopfl/fl and R26STAT3Cstopfl/+ CD4Cre mice were crossed to obtain desired experimental genotype and littermate controls (R26STAT3Cstopfl/+, R26STAT3Cfl/fl or CD4Cre mice). Due to the severe disease phenotype, R26STAT3Cstopfl/fl CD4Cre mice are not fit to be used as breeders. All genotypes are determined by PCR and further confirmed by flow cytometric analysis on GFP expression in T cells. Primers used for genotyping are as followed: STAT3C fwd 5’-GATCATGGATTGTACCTGC-3’; STAT3C rvs 5’-AGGTAGCACACTCCGAGG-3’; Rosa26wt fwd 5’-TCCCAAAGTCGCTCTGAGT-3’; Rosa26wt rvs 5’- TGAGCATGTCTTTAATCTACC-3’; CD4Cre fwd 5’-ACGACCAAGTGACAGCAATG-3’; CD4Cre rvs 5’- CTCGACCAGTTTAGTTACCC-3’. Both male and female mice of 6 to 10 weeks old were used in the experiments. No difference in disease phenotype between sexes was observed. Unless stated otherwise, control mice in this study were either R26STAT3Cstopfl/+ or R26STAT3Cfl/fl mice. However, we have also used CD4Cre mice as control in a previous study and found no difference between the different groups of control mice [12].

Short chain fatty acid supplementation in vivo

Acetate (sodium acetate, Sigma), propionate (sodium propionate, Sigma) and butyrate (sodium acetate, Sigma) in powder form were weighed and added to the drinking water for the mice to achieve a final concentration of 67μM acetate, 26μM propionate and 40μM butyrate. Pregnant female mice were given short chain fatty acids (SCFAs) supplemented water to allow fetal exposure to SCFAs. After the pups were weaned, SCFAs supplementation continued until the pups reached around 8 weeks old or ready for analysis. SCFA supplemented water was changed on a weekly basis.

MicroCT

Intact tibia, femur, hind paws, and vertebra were isolated from the mice. Muscles and all other soft tissues were removed. The bones were fixed in 10% formalin for 24 to 48 hours and stored in 70% ethanol. MicroCT scan was performed on the bones with the following parameters: 60 kV, 167 uA, 9.7 um, 2K, 750 ms exposure, 10 movements. Images were reconstructed using the Skyscan NRECON software (histogram range 0–0.06, beam hardening correction of 40, Gaussian smoothing (factor 1), ring artifact correction of 7). For qualitative analysis, 3-D images of the samples were then reconstructed from the cross-sectional slices using CTNa and CTVox software provided by Skyscan.

Gas Chromatography-Mass Spectrometry

Norvaline and 100% methanol were added to the samples and the standards before performing chloroform-based extraction. Samples were then dehydrated using vacuum drying. Samples were incubated with 20mg/ml of methoxyamine hydrochloride at 37°C for 1 hour and subsequently 1x TBDMS at 37°C for 1 hour for derivatization. Samples were then used to run the automated gas chromatography-mass spectrometry analysis.

Flow Cytometry

Bone marrow was isolated from both the tibia and femur by flushing the cells out using a syringe with a 25G needle. Cells were filtered through a 70-μM cell strainer (Falcon) and were ready for antibody staining. The following antibodies from eBioscience (except those stated otherwise), were used to stain cell surface markers: CD3e (clone 500A2), CD3e (145–2C11), CD4 (RM4–5), CD19 (1D3), CD11b (M1/70), TER-119, Ly-6C (BioLegend, HK1.4), CD25 (PC61.5), CD45 (BioLegend, 30-F11), CD45 (30-F11). Cells stained for intracellular cytokines were stimulated and treated with a PMA, ionomycin, brefeldin A, and monensin mix (eBioscience, San Jose, California) and incubated for 4 hours at 37°C. Cells were surface stained, fixed with 2% PFA and permeabilized with 0.5% saponin, and intracellularly stained for cytokine expression. For the Treg staining, Foxp3 Staining Buffer kit (eBioscience) was used according to the manufacturer’s protocol. The following fluorochrome conjugated antibodies were used: IL-17A (clone TC11–18H10.1, Biolegend, San Diego, California), IL-22 (clone 1H8PWSR), IFNγ (XMG1.2), IL-4 (11B11), and Foxp3 (FJK-16s) (all from eBioscience, San Jose, California). Samples were run on a BD Fortessa and analyzed using FlowJo software. For data analysis, cells were gated on the lymphocyte or appropriate cell population and then doublet cells were excluded.

Quantitative-PCR

The qPCR experiments were performed on StepOnePlus real-time PCR system machine (AB Applied Biosystems) using power SYBR green PCR master mix following manufacturer’s instruction. The primers were:

HPRT forward 5’- GTCATGCCGACCCGCAGTC-3’,

HPRT reverse 5’- GTCCTGTCCATAATCAGTCCATGAGGAATAAAC-3’.

GPR41/FFAR3 forward 5’- TCCTGCCGTTTCGCATGGTGG-3’,

GPR41/FFAR3 reverse 5’- ACCGCCGTCAGGAAGAGGGAG-3’,

GPR43/FFAR2 forward 5’- CCCTGTGCACATCCTCCTGC-3’,

GPR43/FFAR2 reverse 5’- GCGTTCCATGCTGATGCCCG-3’,

GPR109A/HCA forward 5’- GTTACAACTTCAGGTGGCACGAT-3’,

GPR109A/HCA reverse 5’- CTCCACACTAGTGCTTCGGTTATT-3’,

C-fos forward 5’- AGAATCCGAAGGGAAAGGAA-3’,

C-fos reverse 5’- ATTGAGAAGAGGCAGGGTGA -3’,

NFATc1 forward 5’- TGCCTTTTGCGAGCAGTATCT-3’,

NFATc1 reverse 5’- CAGGCAAGGATGGGCTCATAT-3’,

TRAP forward 5’- GCTGTCCTGGCTCAAAAAGC-3’,

TRAP reverse 5’- CACACCGTTCTCGTCCTGAA-3’,

In vitro osteoclast differentiation assays

The in vitro culture assays were performed as described previously [14, 15]. Bone marrow cells were flushed out from both the tibia and femurs of mice using a 25G syringe. Cells were plated onto a 10cm petri dish overnight in alpha-MEM media (supplemented with 10% FBS and penicillin/streptomycin). After the overnight culture, non-adherent cells were collected for osteoclast differentiation. Cells were plated in alpha-MEM media with 30ng/μl recombinant M-CSF (R&D) for 2 days. Then, additional 30ng/μl recombinant RANKL (R&D) was added to the culture and this media (30ng/μl of M-CSF and RANKL) was replaced every 3 days. After 10 days of culture, TRAP staining was performed to identify osteoclasts. Osteoclasts are cells that stained positive and present as multinucleated (more than 3 nucleus) light purple cells and the number of osteoclasts were quantified by double blind manual counting of osteoclasts in each well of the culture plate.

For short term osteoclast differentiation using osteoclast progenitor cells (OCPs), 10,000 to 40,000 sorted OCPs were culture in 96-well-plate with the alpha-MEM with 10% FBS, penicillin/streptomycin and 30ng/μl murine recombinant M-CSF and 30ng/μl murine recombinant RANKL. Fresh media was added on day 3 to achieve the same final concentrations of M-CSF and RANKL. Cells were fixed with 10% formalin and stained for TRAP staining on day 4.

Short chain fatty acid treatment in vitro

For short chain fatty acid (SCFA) treatments, acetate (sodium acetate, Sigma), propionate (sodium propionate, Sigma) and butyrate (sodium acetate, Sigma) were first dissolved in water to make 1000X stocks. Acetate, propionate, butyrate, and a combination of all three SCFAs were added to the media at 5μM final concentration. For the conventional osteoclast differentiation assay, SCFAs were added to the culture on day 1, day 3, day 6 and day 9 while for the short-term osteoclast differentiation assay, SCFAs were added to the culture on day 1 and day 3.

ELISAs

Mouse osteocalcin, P1NP, and CTX-1 ELISA kits were ordered from Novus Biologicals. Catalog Nos. NBP2–68151, NBP2–76466, and NBP2–69074. All mouse serum samples were tested on either first or second thaw and kit protocol was followed. Serum was diluted 1:16–1:60 for optimal quantification within range of the standard curve. Prism Graphpad was used to graph the data and perform statistical tests.

Bulk RNA-sequencing

Bone marrow cells were flushed from SCFA-treated and untreated R26STAT3Cfl/fl CD4Cre and control mice around 8 weeks of age. Cells were then plated onto a 10 cm petri dish overnight in alpha-MEM media (supplemented with 10% FBS and penicillin/streptomycin). After the overnight culture, non-adherent cells were collected for RNA extraction using the Qiagen RNeasy Plus Micro Kit. RNA libraries were then prepared by the NYU Genome Technology Center using the Cloentech PolyA protocol. Catalog No. 634848. Samples were then sequenced using an sp100 Illumina NovaSeq. Reads were cleaned, trimmed, and then aligned using the STAR program [16]. featureCounts was then used to generate count data for mapped reads [17]. (n=1 mouse per condition, cultured and sequenced in triplicate).

Statistics

Statistics were analyzed using unpaired nonparametric Mann-Whitney U test or Wilcoxon matched pair signed rank test. For all figures, more than 3 independent experiments have been performed, unless stated otherwise. Data are generally expressed as mean ± s.e.m. To generate volcano plots and heatmaps of differential gene expression from RNA-seq, we used DESeq2 to statistically make our comparisons and Complex Heatmap packages in R to visualize our data [18, 19]. For Pathway analysis we used Gene Ontology Enrichment from the ClusterProfiler package in R [20, 21].

Acknowledgements

We thank the following NYU Medical Center Core Facilities for the expert assistance: Flow Cytometry, Genome Technology Center, and NYU College of Dentistry micro–CT Core. Additionally, we would like to thank Cathy Shufro for assisting with the editing of this manuscript.

RESULTS

SCFA supplementation alleviates the osteoporosis phenotype observed in a spondyloarthritis mouse model.

In our mouse model, R26STAT3Cstopfl/fl CD4Cre, we observed a prominent osteoporosis phenotype compared to littermate controls [13]; which we quantified using micro-computed tomography (micro-CT). See (Figure 1A). Active bone remodeling can also be observed in R26STAT3Cstopfl/fl CD4Cre mice as identified by the serum elevation of type I collagen cross-linked c-telopeptide (CTX-1), procollagen 1 intact N-terminal pro-peptide (P1NP), and osteocalcin (Supplementary Figure 1A).

Figure 1. SCFAs alleviate the osteoporosis phenotype in R26STAT3Cstopfl/fl CD4Cre mice by inhibiting osteoclastogenesis.

Figure 1.

Open in a new tab

(A) Representative microCT images of femurs harvested from untreated control mice, untreated R26STAT3Cstopfl/fl CD4Cre mice, SCFA-treated control mice and SCFA-treated R26STAT3Cstopfl/fl CD4Cre mice. (B-C) Different microCT measurements of trabecular bones (top) and cortical bones (bottom) of untreated control mice, untreated R26STAT3Cstopfl/fl CD4Cre mice, SCFA-treated control mice and SCFA-treated R26STAT3Cstopfl/fl CD4Cre mice. (D) Representative images of TRAP assay and number of TRAP+ osteoclast counted on day 10 of the in vitro osteoclast differentiation assay using bone marrow progenitors of untreated control mice, untreated R26STAT3Cstopfl/fl CD4Cre mice, SCFA-treated control mice and SCFA-treated R26STAT3Cstopfl/fl CD4Cre mice. (BV/TV – bone volume over tissue volume, Th – thickness, Trabecular Sp – average space between two trabecular bones, BMD – bone mineral density, Cortical B. ar – two-dimensional cortical bone area. Each point represents an individual mouse).

We hypothesized that the imbalance of Th17/Treg cells in the bone marrow of R26STAT3Cstopfl/fl CD4Cre mice could be rectified by SCFA supplementation and that this would reduce the severity of osteoporosis in these animals. Pregnant mice received a combination of 67μM acetate, 26μM propionate, and 40μM butyrate in drinking water during the gestational period and subsequently the pups continued to receive this combination in drinking water until they reached adulthood ~ 8 weeks of life.

When micro-CT scans were performed on the femurs of the experimental animals, we observed a significant difference between treated and untreated R26STAT3Cstopfl/fl CD4Cre mice (Figure 1A), most notably a well-defined trabecular bone microarchitecture and thickened cortical bone in R26STAT3Cstopfl/fl CD4Cre mice treated with SCFA-supplemented water (Figure 1A–C). Other parameters, such as bone mineral density and bone volume/tissue volume ratio also revealed the amelioration of osteoporosis in R26STAT3Cstopfl/fl CD4Cre mice that received SCFA supplementation, compared to the untreated mice of the same genotype (Figure 1B, Figure 1C). These trabecular bone measurements in treated R26STAT3Cstopfl/fl CD4Cre mice reached levels comparable to those of untreated healthy control animals, indicating a restoration of normal trabecular bone development.

Improvements in the cortical bones were apparent in SCFA-treated R26STAT3Cstopfl/fl CD4Cre mice, but they did not attain the bone density and thickness observed in wild-type littermate controls (Figure 1C). Additionally, when comparing SCFA-treated control mice with untreated control mice, we observed a higher trabecular bone volume/tissue volume ratio and an increase in trabecular bone thickness in animals that were given SCFAs. This suggested that SCFAs could exert some effects in mice under homeostatic conditions.

We also assessed the effects of SCFAs on bone formation and resorption by measuring serum levels of osteocalcin, Procollagen type 1 N-terminal propeptide (P1NP), and C-terminal telopeptide of type I collagen (CTX-1), in our mice. Our measurements revealed that all three of these bone turnover makers were upregulated in the serum of our mutant mice, as compared to that of their littermate controls. These results indicate active bone remodeling taking place in R26STAT3Cstopfl/fl CD4Cre mice. Intriguingly serum from R26STAT3Cstopfl/fl CD4Cre mice showed a trend in reduction of osteocalcin and P1NP in response to SCFA treatment. However, no trend was seen with CTX-1 (Supplementary Figure 1A).

SCFAs influence differentiation of osteoclasts.

Because enhanced osteoclastogenesis could contribute to dysregulated bone remodeling and a higher osteoclast differentiation capacity in our R26STAT3Cstopfl/fl CD4Cre mice, we hypothesized that SCFAs ameliorated the osteoporosis in these model mice by inhibiting aberrant osteoclast differentiation. To test this hypothesis, we used control and R26STAT3Cstopfl/fl CD4Cre mice that were untreated or received SCFA supplementation and performed an osteoclast differentiation assay using total bone marrow. This 10-day assay requires M-CSF and RANKL to differentiate bone marrow cells into multinucleated osteoclasts. We performed a tartrate-resistant acid phosphatase (TRAP) assay on day 10 to stain osteoclasts light purple. We determined the number of osteoclasts by double-blind manual counting. Bone marrow from R26STAT3Cstopfl/fl CD4Cre mice treated with SCFA water failed to differentiate into the large number of osteoclasts as observed in culture using bone marrow from untreated R26STAT3Cstopfl/fl CD4Cre mice (Figure 1D). These data suggests that enhanced osteoclastogenesis underlies the osteoporosis phenotype in R26STAT3Cstopfl/fl CD4Cre mice and that SCFAs can correct this dysregulation by inhibiting osteoclast differentiation.

SCFAs suppress osteoclastogenesis despite the lack of change in BM Th17/Treg ratio in R26STAT3Cstopfl/fl CD4Cre mice.

We have previously shown that the enhanced osteoclast differentiation using total bone marrow from untreated R26STAT3Cstopfl/fl CD4Cre mice was due to an elevated number of osteoclast progenitor cells (OCPs), which can be identified as CD3-, CD19-, Ter119-, CD11blow, Ly6Chigh cells in the bone marrow (Figure 2A). More recent studies have demonstrated that the use of CD115 and anti-C-kit antibodies enables a more refined identification of OCPs within a monocyte-derived Ly6C+ population (Figure 2A) [22]. Indeed, our analysis of a CD115+ C-kit+ OCP population revealed a significant increase in this cell subset within the bone marrow of R26STAT3Cstopfl/fl CD4Cre mice (Figure 2B). However, we decided to continue using the broader gating for the isolation experiments to ensure sufficient cell recovery for downstream experiments. Analysis of bone marrow from SCFA-treated R26STAT3Cstopfl/fl CD4Cre mice revealed that the expansion in OCPs remains (Figure 2C), suggesting that the underlying processes that contributed to the overabundance of OCPs were still present.

Figure 2. Bone marrow T cell compartment of R26STAT3Cstopfl/fl CD4Cre mice remains proinflammatory despite the SCFAs treatment.

Figure 2.

Open in a new tab

(A) Representative flow plots of the osteoclast progenitors. Singlets, CD3-, CD19- and Ter119- cells were selected in the upstream gating (top). Representative flow plots showing the further refining of the osteoclast progenitor cell population using CD115 and C-kit (bottom). (B) Summarized refined OCPs as a percentage of total bone marrow cells. (C) The percentage of osteoclast progenitors in the bone marrow of untreated control mice, untreated R26STAT3Cstopfl/fl CD4Cre mice, SCFA-treated control mice and SCFA-treated R26STAT3Cstopfl/fl CD4Cre mice. (D) The percentage of IL-17+ Th17 cells, IL-22+ Th22 cells, IFN-gamma+ Th1 cells, and IL-4+ Th2 cells in the bone marrow T cell compartment of untreated control mice, untreated R26STAT3Cstopfl/fl CD4Cre mice, SCFA-treated control mice and SCFA-treated R26STAT3Cstopfl/fl CD4Cre mice. (E) Representative flow plots and proportion of regulatory T cells in the bone marrow T cell compartment of untreated control mice, untreated R26STAT3Cstopfl/fl CD4Cre mice, SCFA-treated control mice and SCFA-treated R26STAT3Cstopfl/fl CD4Cre mice. Singlets, CD3+ and CD4+ cells were selected in the upstream gating. (F) The ratio of Th17 cells over regulatory T cells in the bone marrow T cell compartment of untreated control mice, untreated R26STAT3Cstopfl/fl CD4Cre mice, SCFA-treated control mice and SCFA-treated R26STAT3Cstopfl/fl CD4Cre mice. (Each point represents an individual mouse).

Considering the known impact of IL-17 and IL-22 on the severity of osteoporosis in arthritic R26STAT3Cstopfl/fl CD4Cre mice, and the critical role Th17 cells can play in driving rheumatic disease in these animals, we evaluated the T cell compartment in the bone marrow of SCFA-treated mice. We then compared the T cell immune subsets with those from untreated controls and R26STAT3Cstopfl/fl CD4Cre mice. SCFA supplementation did not alter the percentage of IL-17- and IL-22-producing Th17 cells or the IFN- γ-producing Th1. Similarly, the fraction of IL-4- producing Th2 cells in the bone marrow of these SCFA-supplemented mice also remained steady (Figure 2D). For T cell subset gating see Supplementary Figure 1C. In addition, the percentage of T regulatory cells (Tregs), whose differentiation can be promoted by butyrate in the gut, remained low in the BM of treated R26STAT3Cstopfl/fl CD4Cre mice, suggesting that SCFAs did not affect the T-cell compartment in the bones of the treated mice (Figure 2E). There were more T cells in the bone marrow of mice that received SCFA supplementation, hence the absolute number of Th1, Th2, Th17, and Treg had increased with the treatment. However, the ratio of Th17/Treg had not been altered by SCFA treatment, therefore, the microenvironment remained skewed in favor of Th17 cells (Figure 2F).

SCFAs directly inhibit osteoclast differentiation in vitro.

Amelioration of the osteoporosis phenotype by SCFAs occurs without resetting of the Th17/Treg imbalance in the bone marrow of treated R26STAT3Cstopfl/fl CD4Cre mice, suggesting that these microbially derived metabolites are acting directly on osteoclasts and osteoblasts. To characterize the effects of SCFAs on bone cells under biologically relevant concentrations, we performed gas chromatography-mass spectrometry (GC-MS) to determine changes in the concentrations of SCFAs in different tissues after their exogenous administration. We observed an elevation in the concentration of individual SCFAs in the plasma of SCFA-treated mice compared to plasma from untreated animals, with the most significant changes observed in plasma butyrate concentration (Supplementary Figure 1B). This difference in SCFA concentrations demonstrates that SCFAs taken up by the gut can be distributed to and exert their functions in distal tissues. Critically, these measurements of absolute concentrations of SCFAs enabled us to identify physiologically relevant concentrations for in vitro work. Each of the SCFAs was found to be present at ~5μΜ concentration in the circulation of mice that had undergone SCFAs treatment.

When we added a combination of acetate, propionate and butyrate at 5μΜ concentration each to the osteoclast differentiation culture, along with the cytokines necessary for osteoclast differentiation (i.e., M-CSF and RANKL), we observed a significant reduction in differentiation of TRAP+ osteoclasts from the bone marrow of R26STAT3Cstopfl/fl CD4Cre mice (Figure 3A, Figure 3B). Higher concentrations of SCFAs (10x and 100x of the concentration we found in the serum of animals) even inhibited osteoclast differentiation more strongly (Figure 3A, Figure 3B). To further characterize the effects of individual SCFAs on osteoclastogenesis, we performed osteoclast differentiation in the presence of 5 μΜ acetate, 5 μΜ propionate, and 5 μΜ butyrate, as well as a combination of all three SCFAs at 5 μΜ concentration each. Bone marrow harvested from R26STAT3Cstopfl/fl CD4Cre mice presented a lower osteoclastogenesis potential in the presence of both individual SCFAs and the cocktail of SCFAs (Figure 3C, Figure 3D). In contrast, significant reduction in osteoclast differentiation of control bone marrow occurred only in the presence of the full complement of SCFAs (Figure 3D). This finding suggests that cells from R26STAT3Cstopfl/fl CD4Cre are more sensitive compared to control, to SCFA-mediated inhibition in osteoclastogenesis.

Figure 3. SCFAs inhibit osteoclastogenesis in vitro and OCPs from R26STAT3Cstopfl/fl CD4Cre mice upregulate SCFA receptors.

Figure 3.

Open in a new tab

(A) Representative images of TRAP staining on day 10 of osteoclast differentiation performed using bone marrow cells harvested from control mice (top) and R26STAT3Cstopfl/fl CD4Cre mice (bottom), exposed to either no SCFAs, serum concentration of SCFAs, or 100x serum concentration of SCFAs. (B) Number of TRAP+ osteoclasts after 10 days of osteoclast differentiation +/− SCFA added to culture. (C) Representative images of TRAP staining on day 10 of osteoclast differentiation culture in the absence of SCFAs and presence of 5μM acetate, propionate, or butyrate using bone marrow harvested from R26STAT3Cstopfl/fl CD4Cre mice. (D) Summarized number of TRAP+ osteoclasts differentiated in vitro under varying SCFA exposures using bone marrow harvested from control and R26STAT3Cstopfl/fl CD4Cre mice. (E) Relative expression of SCFA receptors (GPR43, GPR41 and GPCR109A) in sorted osteoclast progenitors from bone marrow of control and R26STAT3Cstopfl/fl CD4Cre mice. (F) Relative expression of SCFA receptor GPR43 in osteoclasts differentiated from osteoclast progenitors +/− SCFA exposure. (G) Relative expression of SCFA receptor GPR41 in osteoclasts differentiated from osteoclast progenitors +/− SCFAs exposure. (H) Relative expression of SCFA receptor GPR109A in osteoclasts differentiated from osteoclast progenitors +/− SCFA exposure. (A=Acetate, P=Propionate, B=Butyrate, Combo =All SCFAs. E-H: n=4 mice per condition).

Differential expression of SCFA receptors in OCPs from mice predisposed to spondyloarthritis

Since spondyloarthritis in R26STAT3Cstopfl/fl CD4Cre animals was accompanied by expansion of OCPs and was amenable to treatment with SCFA supplementation, we next examined whether OCPs in these animals had differential expression of SCFA-associated receptors. To elucidate if the expression of GPR43 (FFAR2), GPR41 (FFAR3) and GPR109A was different in OCPs from control mice compared to R26STAT3Cstopfl/fl CD4Cre mice, we performed real-time qPCR. We observed an increased expression of all three SCFA receptors in OCPs sorted from R26STAT3Cstopfl/fl CD4Cre mice compared to control littermates (Figure 3E). Interestingly, the sorted OCPs of R26STAT3Cstopfl/fl CD4Cre mice downregulated their expression of these SCFA receptors as they underwent differentiation into osteoclasts during a short 4-day culture with M-CSF and RANKL. This occurred irrespective of whether SCFAs were present in the culture (Figure 3F–H). This finding suggests that early osteoclast progenitors are likely the most responsive to SCFAs.

Consistent with lack of expression of GPR43 and GPR41 in differentiating osteoclasts, the addition of acetate and propionate to the short-term culture had minimal impact on osteoclast differentiation (Supplementary Figure 2A and 2B). By contrast, butyrate and a combination of the three SCFAs exerted a notable effect on the osteoclast differentiation capacity, albeit not a significant effect, likely due to persistent low-level of expression of GPR109A (Supplementary Figure 2A and 2B). This observed reduction in osteoclastogenesis did not appear to be due to downregulation of known key transcription factors for osteoclast differentiation, NFATc1 and c-Fos (Supplementary Figure 2C and 2D) This suggests that other signaling pathways important for differentiation of these bone resorbing cells were impacted.

SCFAs affect the transcriptomics landscape of bone marrow-derived progenitor cells.

To elucidate the underlying molecular mechanism driving the aberrant bone phenotype in our R26STAT3Cstopfl/fl CD4Cre mice and to gain insight into the therapeutic activity of SCFAs, we performed RNA sequencing on bone marrow-derived progenitor cells. We collected femur-derived bone marrow cells from untreated and SCFA-treated R26STAT3Cstopfl/fl CD4Cre and control mice. Following overnight culture (~24 hours), we collected non-adherent progenitor cells for each genotype and condition for RNA extraction and sequencing. We first compared gene expression in the progenitor cells of untreated R26STAT3Cstopfl/fl CD4Cre and control mice in order to understand the transcriptional differences between our sick animals and controls. This analysis revealed upregulation of interferon-stimulated genes, the most significant genes being Ifit3, Iigp1, and Ifit3b. In addition to these interferon-stimulated genes, Stat1 was also significantly upregulated in progenitor cells from R26STAT3Cstopfl/fl CD4Cre mice compared to progenitor cells from controls. In contrast, the expression of Col5a3, a collagen type 5 gene, was significantly diminished in comparison to expression in controls (Figure 4A).

Figure 4. RNA-seq shows reduced interferon signaling in bone marrow progenitor cells from SCFA-treated R26STAT3Cstopfl/fl CD4Cre mice compared to their untreated counterparts.

Figure 4.

Open in a new tab

(A) Volcano plot showing significantly downregulated and upregulated genes in bone marrow progenitor cells from R26STAT3Cstopfl/fl CD4Cre mice compared to cells from control mice. (B) Volcano plot showing differentially expressed genes in cells from SCFA-treated vs. untreated R26STAT3Cstopfl/fl CD4Cre mice. (C) Gene Ontology pathway analysis of the top 15 significantly upregulated pathways in R26STAT3Cstopfl/fl CD4Cre bone marrow progenitor cells compared to control. (D) Heatmap of the top differentially expressed interferon genes from pathway analysis in C, showing unsupervised hierarchical clustering of boned marrow-derived progenitor cells by genotype and treatment. (n=1 mouse per condition, cultured and sequenced in triplicate).

We next compared the gene expression profiles of SCFA-treated and untreated R26STAT3Cstopfl/fl CD4Cre mice. We observed a striking decrease in expression of interferon-stimulated genes in our SCFA-treated mice compared to untreated mice; Ifit3 was the most significantly downregulated gene. Notably, Car2, a positive regulator of osteoclastogenesis, was downregulated in the SCFA-treated group compared to the untreated group [23]. Furthermore, we found that SCFA treatment significantly increased Dusp1, a negative regulator of osteoclast formation and an anti-inflammatory mediator (Figure 4B) [24, 25]. Unexpectedly, we observed a large increase in the expression of Egr1 and Fos, genes linked with early differentiation of progenitor cells. While expression of Fos is specifically associated with osteoclast differentiation [26], Egr1 upregulation may be more related to osteoblast activity [27]. Additional differential gene expression for these comparisons can be found in Supplementary Figure 2E and 2F.

To gain deeper insights into the differences in myeloid progenitor cells across our two genotypes we conducted pathway analysis comparing untreated R26STAT3Cstopfl/fl CD4Cre mice with control. Consistent with the observed changes in genes related to interferon signaling, the most upregulated pathways were defense/antiviral response pathways and innate immune system activation pathways (Figure 4C). Indeed, the most significant genes which contributed to these pathways revealed numerous interferon-stimulated genes. We therefore decided to compare these genes across our different genotypes and treatment groups using unsupervised clustering. Strikingly we found that this interferon gene signature is significantly dampened upon SCFA treatment in R26STAT3Cstopfl/fl CD4Cre mice (Figure 4D).

DISCUSSION

Augmented Th17 leads to a dramatic rheumatic phenotype in R26STAT3Cstopfl/fl CD4Cre mice, punctuated by increased osteoclastogenesis and bone erosion. In the current study, we examined the impact of microbial metabolites known to modulate T cell response on the osteoporosis phenotype found in the R26STAT3Cstopfl/fl CD4Cre mice. While the combination of SCFAs, namely acetate, propionate, and butyrate, was able to significantly improve the bone volume/tissue volume, bone thickness and bone mineral density in these mice, their Th17/Treg imbalance and the downstream expansion of osteoclast progenitors (OCPs) persisted in the bone marrow of R26STAT3Cstopfl/fl CD4Cre animals that received SCFA supplementation. These progenitor cells are annotated as CD11b-/low Ly6C high in the bone marrow of inflammatory arthritic mice [22]. This prompted us to examine other mechanisms by which SCFAs may impact osteoclast development. To test if SCFAs could regulate bone cells directly and influence their development in vivo, we supplemented cultures during osteoclast differentiation with either individual SCFAs or a cocktail of SCFAs at concentrations found in the circulation (based on serum concentration established by mass spectrometry, which we found to be consistent with concentrations established by Lucas et al [28]). The serum concentrations determined by our measurements and used in the assays were markedly lower than the concentrations of these SCFAs found in the gut of treated animals or those commonly used in similar assays. These concentrations are also consistent with human serum concentrations of SCFAs, especially for propionate and butyrate [29, 30]. Even at a low concentration, SCFAs significantly reduced murine osteoclast differentiation, indicating that these microbial metabolites (which are generated as part of fermentation in the intestinal lumen) have the potential to act at distal sites to regulate bone development and homeostasis. This demonstrates a direct interaction between the gut and the bone, forming the gut-bone axis.

Through our in vitro osteoclast differentiation assay, we observed a distinct sensitivity to SCFAs in cell cultures of the control mice and the R26STAT3Cstopfl/fl CD4Cre mice. The pattern of expression of SCFA receptors in the OCPs likely contributed to the observed differences in mice predisposed to arthritic phenotype. GPR43, GPR41, and GPR109A were upregulated in OCPs of R26STAT3Cstopfl/fl CD4Cre mice compared to that of littermate controls. Overexpression of SCFA receptors could be responsible for more robust signal transduction downstream, although this needs to be validated. Intensified Th17 responses led to the enhanced osteoclastogenesis in the bones of R26STAT3Cstopfl/fl CD4Cre mice, but at the same time rendered the cells more prone to SCFA driven inhibition. Further studies are warranted to reveal the signaling pathways that promote the upregulation of these SCFA receptors and to explore whether patients with Th17-driven spondyloarthritis would be responsive to SCFA-based therapies.

The expression of GPR43 and GPR41 were downregulated significantly when OCPs differentiated into osteoclasts, while the expression of butyrate specific receptor GPR109A was downregulated to a lesser extent during osteoclast differentiation. This could explain the consistent effects of butyrate in suppressing osteoclastogenesis whether it was given in vivo, supplemented long term in vitro or added for just 4 days to the short-term osteoclast differentiation culture. Alternatively, the effects of acetate and propionate, which bind and act through GPR43 and GPR41, are both cell-specific and time-dependent. During the first few days of the 10-day osteoclast differentiation assay, when the OCPs proliferated in the presence of M-CSF alone, exposure to either acetate, propionate or butyrate effectively inhibited the differentiation of osteoclasts. However, during the 4-day differentiation assay when M-CSF and RANKL were added to sorted OCPs, osteoclast differentiation occurred right away, and the effect of acetate and propionate was lost. The expression patterns of GPR43 and GPR41 might explain the disparity in the effects of acetate and propionate observed in the 10-day and the 4-day osteoclast differentiation assays.

To better understand SCFA’s impact on osteoclast differentiation in vivo, we performed RNA-seq on myeloid progenitors derived from the non-adherent bone marrow cells of both SCFA-treated and untreated R26STAT3Cstopfl/fl CD4Cre and control mice. Unexpectedly, we observed no significant upregulation of osteoclast differentiation genes in our R26STAT3Cstopfl/fl CD4Cre mice compared to the control group. Instead, we observed a large increase in interferon-stimulated genes, suggesting that the difference in osteoclastogenesis between our R26STAT3Cstopfl/fl CD4Cre mice and control mice may be due to inflammatory signaling (Figure 4A). Our pathway analysis of this comparison revealed significant upregulation in viral response and immune activation pathways, with the most prominent pathways driven by enrichment of interferon-response genes and STAT1 signaling (Figure 4C).

When we compared our SCFA-treated and untreated R26STAT3Cstopfl/fl CD4Cre mice we found an increase in the osteoclast differentiation gene Fos [26], however we also observed a decrease in Car2 (a positive regulator of osteoclast differentiation [23]) (Figure 4B). Interestingly, while inflammatory genes were up in osteoclast progenitor cells from R26STAT3Cstopfl/fl CD4Cre, we saw a significant decrease in interferon-stimulated genes in response to SCFA treatment and an increase in Dusp1, a negative regulator of osteoclast differentiation and a primary mediator of the resolution of inflammation [24–25]. A comprehensive comparison across all our experimental groups of the most significant interferon-stimulated genes from our pathway analysis, revealed a remarkable decrease in gene expression among SCFA-treated R26STAT3Cstopfl/fl CD4Cre mice, similar to the levels seen in our control groups (Figure 4D). Our findings indicate that SCFAs could modulate osteoclastogenesis by diminishing inflammation through IFN and STAT1-driven immune responses. Yet, the bulk-sequencing method used has its constraints; it doesn’t pinpoint the exact progenitor cell subsets affected by SCFAs. Although the literature confirms that type I IFNs act as negative regulators of osteoclast formation and bone resorption [31–34], IFN-γ has been shown to augment osteoclast differentiation via activation of antigen-specific T cells which secrete RANKL [35]. However, our transcriptional analysis does not distinguish type I from type II IFN-induced genes and merely shows upregulation of IFN-induced inflammatory genes in bone marrow progenitor cells in our arthritis murine model.

The IFN response in R26STAT3Cstopfl/fl CD4Cre mice’s bone marrow progenitor cells was moderated by SCFA treatment. While microbial-derived acetate has been found to elevate IFN-ß in mouse lungs—evidenced by increased upregulation of IFN-stimulated genes during viral infections [36] —butyrate tends to downregulate IFN-stimulated genes [37] and reduce IFN-γ production. This is observed in the context of cultured lymphocytes [38] as well as colonic mucosa, and has been attributed to reduced STAT1 signaling [39, 40]. We hypothesize that butyrate’s effects are more pronounced in our model, possibly due to the enhanced expression of its high-affinity receptors (GPR109A and GPR43) in bone marrow derived OCPs. This in turn leads to a marked decrease in IFN-stimulated gene expression in the progenitor populations in our SCFA-treated R26STAT3Cstopfl/fl CD4Cre mice.

In the future it will be important to identify the effects of SCFAs on other cell types as well as intracellular signaling pathways downstream of SCFA receptors. These future studies may reveal more details about how SCFAs modulate inflammation and osteoclast development. Lucas et al previously suggested a change in cellular metabolism after SCFAs exposure [28]. Other known mechanisms of SCFAs, such as regulation of cell cycle and direct inhibition of HDACs, can also be tested to determine the mode(s) of action by which SCFAs impact osteoclast differentiation.

Our data also demonstrated a direct impact of SCFAs on OCPs, but this does not exclude other indirect influences of these metabolites on osteoclasts, or even bone building osteoblasts and T lymphocytes. SCFAs, which are usually produced by the distal gut microbiome, can influence the composition of microbial species in the gut. Preliminary studies from our group have demonstrated microbiome changes in wild-type mice that were treated with SCFAs supplemented water for merely 2 weeks. An expansion of Akkermansia was observed in fecal samples from SCFA-treated mice [41]. Multiple pieces of evidence support the role of dysbiosis or changes in the gut microbial community in host health and disease. To decouple the effects of microbiome changes in response to SCFA supplementation, gnotobiotic experiments with SCFA treatment will need to be performed in the future. This could help us to distinguish microbiome dependent and microbiome independent functions of SCFAs.

Recent studies suggest that SCFA may benefit patients with rheumatoid arthritis (RA). Dietary supplementation with high-fiber bars, which promotes higher SCFA generation by the microbiome, has been shown to reduce serum level of pro-arthritogenic cytokines, such as CCL2, IL-18 and IL33 in RA patients [42]. However, the study did not evaluate bone conditions of these patients to ascertain any functional changes driven by SCFAs. Another study following a group of at-risk individuals for 6 years demonstrated that those who did not progress into full blown RA had a higher baseline serum level of SCFAs, particularly acetate and butyrate, than those who progressed into clinical arthritis [43]. This suggests that SCFA may act to delay the progression of inflammation into active joint swelling. Detailed examination of murine collagen induced arthritis models corroborated the findings on the beneficial effects of SCFAs [44]. Notably, the authors also did not observe any increase in peripheral regulatory T cells upon SCFA treatment, similar to the results from our study [44]. These data suggest SCFAs’ potential clinical significance in arthritis treatment, highlighting the need for further studies.

Overall, our work adds to the current understanding of the influential contribution of microbial derived metabolites and their pleiotropic functions in host health and disease. Together, our study highlights the role of the gut-bone axis in mediating skeletal diseases and opens new avenues for further mechanistic studies in the field.

Limitations of the study

The experiments and the observations in this study were performed in a murine model of PsA characterized by Th17 driven pathology. Our findings suggest reduced IFN signaling in bone marrow progenitors in response to SCFA treatment. Importantly, this population included all non-adherent bone marrow progenitor cells and therefore includes both osteoclast progenitors as well as other hematopoietic cells. In the future, it will be crucial to define the cell populations being directly affected by SCFAs. Critically, the translational relevance of these results needs to be validated by examining the impact of SCFAs on human derived OCPs. We focused on the osteoclasts, which play a critical role in driving the osteoporosis phenotype in these mice and have not explored the potential effects of the SCFAs on osteoblasts, which also contribute to bone homeostasis and are reduced in our mouse model. Additionally, it will be important to explore the impact of SCFAs on bone remodeling in models driven by other underlying pathologies, such as in collagen-induced arthritis.

Supplementary Material

fig S1

Supplementary Figure 1. Serum bone remodeling markers and concentration of individual SCFAs in mice after SCFA treatment. (A) Serum levels of bone remodeling markers Osteocalcin, P1NP, and CTX-1 by genotype and treatment. (B) Relative concentration of the individual short chain fatty acids (SCFAs) in the fecal samples, liver samples and plasma samples of WT mice treated or untreated with SCFAs. (C) Representative flow gating strategy to identify Th17, Th22, Th1, and Th2 cells after doublet exclusion. (CTX-1 – type I collagen cross-linked c-telopeptide, P1NP – procollagen 1 intact N-terminal propeptide. A-B: each point represents an individual mouse).

fig S2

Supplementary Figure 2. SCFA supplementation in terminal osteoclast differentiation from OCPs and differential gene expression of bone marrow progenitor cells of R26STAT3Cstopfl/fl CD4Cre mice. (A) Representative images of TRAP+ osteoclasts after 4 days of differentiation from sorted osteoclast progenitor cells of control and R26STAT3Cstopfl/fl CD4Cre mice. (B) Quantification of TRAP+ osteoclasts differentiated from osteoclast progenitors in the absence and presence of individual short chain fatty acids (acetate, propionate and butyrate). (C) Relative expression of transcription factor NFATc1 in osteoclasts differentiated from the osteoclast progenitors in the absence and presence of short chain fatty acids. (D) Relative expression of transcription factor c-Fos in osteoclasts differentiated from the osteoclast progenitors in the absence and presence of short chain fatty acid. (E) Heatmap of the top differentially expressed genes in bone marrow progenitor cells across R26STAT3Cstopfl/fl CD4Cre and control mice, showing genes with base mean > 50, p-value < 0.005, & LogFC >2.0). (F) Heatmap of the top differentially expressed genes in bone marrow progenitor cells across SCFA-treated and untreated R26STAT3Cstopfl/fl CD4Cre mice, showing genes with base mean > 100, p-value < 0.005, & LogFC >2.0). (A-D n=4 mice per condition).

Financial Support

This work was supported by grants to SBK and JUS: a National Psoriasis Foundation Diagnostic Test Grant and a Colton Center for Autoimmunity Pilot Grant. Work in the Koralov lab was further supported by the LEO Foundation and the Drs. Martin and Dorothy Spatz Charitable Foundation and NIH R01 (CA271245). BJM was supported by I&I T32 (T32AI100853-11). KLY was supported by an A*STAR National Science Scholarship from Singapore. JUS was supported by MOTI T32 (T32CA009161). The Cadwell Laboratory was supported by DK093668 and AI140754. The Scher lab was further supported by the Beatrice Snyder foundation and NIH grant R01 (AR074500).

Disclosures

KC has received research support from Pfizer, Takeda, Pacific Biosciences, Genentech, and Abbvie. KC has consulted for or received an honorarium from Puretech Health, Genentech, and Abbvie. KC is named as an inventor for U.S. patent 10,722,600 and provisional patents 62/935,035 and 63/157,225. JUS and SBK are named as inventors for U.S. patent 10,226,443. SBK has previously received research support from Micreos, BMS, Agios, Dracen and KymeraTx – none of it related to the topic of this manuscript.

REFERENCES

  • 1.Qiu J, et al. , Acetate Promotes T Cell Effector Function during Glucose Restriction. Cell Rep, 2019. 27(7): p. 2063–2074.e5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Arpaia N, et al. , Metabolites produced by commensal bacteria promote peripheral regulatory T-cell generation. Nature, 2013. 504(7480): p. 451–455. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Smith PM, et al. , The Microbial Metabolites, Short-Chain Fatty Acids, Regulate Colonic Treg Cell Homeostasis. Science, 2013. 341(6145): p. 569–573. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Chen G, et al. , Sodium Butyrate Inhibits Inflammation and Maintains Epithelium Barrier Integrity in a TNBS-induced Inflammatory Bowel Disease Mice Model. EBioMedicine, 2018. 30: p. 317–325. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Chen L, et al. , Microbiota Metabolite Butyrate Differentially Regulates Th1 and Th17’Cells’ Differentiation and Function in Induction of Colitis. Inflamm Bowel Dis, 2019. 25(9): p. 1450–1461. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Nastasi C, et al. , Butyrate and propionate inhibit antigen-specific CD8+ T cell activation by suppressing IL-12 production by antigen-presenting cells. Scientific Reports, 2017. 7(1): p. 14516. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Nøhr MK, et al. , GPR41/FFAR3 and GPR43/FFAR2 as cosensors for short-chain fatty acids in enteroendocrine cells vs FFAR3 in enteric neurons and FFAR2 in enteric leukocytes. Endocrinology, 2013. 154(10): p. 3552–64. [DOI] [PubMed] [Google Scholar]
  • 8.Li M, et al. , Pro- and anti-inflammatory effects of short chain fatty acids on immune and endothelial cells. Eur J Pharmacol, 2018. 831: p. 52–59. [DOI] [PubMed] [Google Scholar]
  • 9.Sivaprakasam S, et al. , Short-Chain Fatty Acid Transporters: Role in Colonic Homeostasis. Compr Physiol, 2017. 8(1): p. 299–314. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Amarasekara DS, et al. , Regulation of Osteoclast Differentiation by Cytokine Networks. Immune Netw, 2018. 18(1): p. e8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Tyagi AM, et al. , The Microbial Metabolite Butyrate Stimulates Bone Formation via T Regulatory Cell-Mediated Regulation of WNT10B Expression. Immunity, 2018. 49(6): p. 1116–1131.e7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Fogli LK, et al. , T cell-derived IL-17 mediates epithelial changes in the airway and drives pulmonary neutrophilia. J Immunol, 2013. 191(6): p. 3100–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Yang L, et al. , Augmented Th17 Differentiation Leads to Cutaneous and Synovio-Entheseal Inflammation in a Novel Model of Psoriatic Arthritis. Arthritis Rheumatol, 2018. 70(6): p. 855–867. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Mediero A, et al. , Netrin-1 is a critical autocrine/paracrine factor for osteoclast differentiation. J Bone Miner Res, 2015. 30(5): p. 837–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Mediero A, et al. , Adenosine A(2A) receptor ligation inhibits osteoclast formation. Am J Pathol, 2012. 180(2): p. 775–86. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Dobin A, Davis CA, Schlesinger F, Drenkow J, Zaleski C, Jha S, Batut P, Chaisson M, Gingeras TR. STAR: ultrafast universal RNA-seq aligner. Bioinformatics. 2013. 29(1): p. 15–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Liao Y, Smyth GK, Shi W. featureCounts: an efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics. 2014. 30(7): p. 923–30. [DOI] [PubMed] [Google Scholar]
  • 18.Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome biology. 2014. 15(12): p. 1–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Gu Z, Eils R, Schlesner M. Complex heatmaps reveal patterns and correlations in multidimensional genomic data. Bioinformatics. 2016. 32(18): p. 2847–9. [DOI] [PubMed] [Google Scholar]
  • 20.Wu T, Hu E, Xu S, Chen M, Guo P, Dai Z, Feng T, Zhou L, Tang W, Zhan L, Fu x, Liu S, Bo X, Yu G. “clusterProfiler 4.0: A universal enrichment tool for interpreting omics data.” The Innovation, 2021. 2(3): p. 100141. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Yu G, Wang L, Han Y, He Q. clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS: A Journal of Integrative Biology, 2012. 16(5): p. 284–287. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Charles JF, et al. , Inflammatory arthritis increases mouse osteoclast precursors with myeloid suppressor function. J Clin Invest, 2012. 122(12): p. 4592–605. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Lehenkari P, Hentunen TA, Laitala-Leinonen T, Tuukkanen J, Väänänen HK. Carbonic anhydrase II plays a major role in osteoclast differentiation and bone resorption by effecting the steady state intracellular pH and Ca2+. Experimental cell research. 1998. Jul 10;242(1):128–37. [DOI] [PubMed] [Google Scholar]
  • 24.Choi Y, et al. , Calcium-Phosphate Crystals Promote RANKL Expression via the Downregulation of DUSP1. Mol Cells, 2019. 42(2): p. 183–188. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Hoppstädter J, Ammit AJ. Role of dual-specificity phosphatase 1 in glucocorticoid-driven anti-inflammatory responses. Frontiers in immunology, 2019. 10: p. 1446. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Wagner E, Eferl R, Hoebertz A, Kenner L, Matsuo K. Bone cell differentiation and the role of Fos/AP-1 proteins. Arthritis Res Ther. 2003. 5: p. 1–54. [Google Scholar]
  • 27.Chaudhary LR, Cheng SL, Avioli LV. Induction of early growth response-1 gene by interleukin-1β and tumor necrosis factor-α in normal human bone marrow stromal and osteoblastic cells: regulation by a protein kinase C inhibitor. Molecular and cellular biochemistry, 1996. 156: p. 69–77. [DOI] [PubMed] [Google Scholar]
  • 28.Lucas S, et al. , Short-chain fatty acids regulate systemic bone mass and protect from pathological bone loss. Nat Commun, 2018. 9(1): p. 55. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Bertges M, van Helden J, and Weiskirchen R, Quantification of Short Chain Fatty Acids (acetate, butyrate, propionate) in human blood with ion exclusion chromatography. Practical Laboratory Medicine, 2021. 26: p. e00244. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Wolever TM, Fernandes J, and Rao AV, Serum acetate:propionate ratio is related to serum cholesterol in men but not women. J Nutr, 1996. 126(11): p. 2790–7. [DOI] [PubMed] [Google Scholar]
  • 31.Takayanagi H, Kim S, Taniguchi T. Signaling crosstalk between RANKL and interferons in osteoclast differentiation. Arthritis Research & Therapy. 2002. 4: p. 1–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Xiong Q, Zhang L, Ge W, Tang P. The roles of interferons in osteoclasts and osteoclastogenesis. Joint Bone Spine. 2016. 83(3): p. 276–81. [DOI] [PubMed] [Google Scholar]
  • 33.Zhao B and Ivashkiv LB, Negative regulation of osteoclastogenesis and bone resorption by cytokines and transcriptional repressors. Arthritis Research & Therapy, 2011. 13(4): p. 234. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Place DE, et al. , Osteoclast fusion and bone loss are restricted by interferon inducible guanylate binding proteins. Nat Commun, 2021. 12(1): p. 496. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Gao Y, et al. , IFN-gamma stimulates osteoclast formation and bone loss in vivo via antigen-driven T cell activation. J Clin Invest, 2007. 117(1): p. 122–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Antunes KH, Fachi JL, de Paula R, da Silva EF, Pral LP, Dos Santos AÁ, Dias GB, Vargas JE, Puga R, Mayer FQ, Maito F. Microbiota-derived acetate protects against respiratory syncytial virus infection through a GPR43-type 1 interferon response. Nature Communications. 2019. Jul 22;10(1):3273. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Chemudupati M, Kenney AD, Smith AC, Fillinger RJ, Zhang L, Zani A, Liu SL, Anderson MZ, Sharma A, Yount JS. Butyrate reprograms expression of specific interferon-stimulated genes. Journal of virology. 2020. Jul 30;94(16):10–128. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Cavaglieri CR, Nishiyama A, Fernandes LC, Curi R, Miles EA, Calder PC. Differential effects of short-chain fatty acids on proliferation and production of pro-and anti-inflammatory cytokines by cultured lymphocytes. Life sciences. 2003. Aug 15;73(13):1683–90. [DOI] [PubMed] [Google Scholar]
  • 39.Zimmerman MA, Singh N, Martin PM, Thangaraju M, Ganapathy V, Waller JL, Shi H, Robertson KD, Munn DH, Liu K. Butyrate suppresses colonic inflammation through HDAC1-dependent Fas upregulation and Fas-mediated apoptosis of T cells. American Journal of Physiology-Gastrointestinal and Liver Physiology. 2012. Jun 15;302(12):G1405–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Klampfer L, Huang J, Sasazuki T, Shirasawa S, Augenlicht L. Inhibition of interferon γ signaling by the short chain fatty acid butyrate. Molecular Cancer Research. 2003. Sep 1;1(11):855–62. [PubMed] [Google Scholar]
  • 41.Malik F MJ, Herrera A, Attur M, Reddy SM, Yang L, Koralov S, Scher JU., Effects of Fatty Acid Supplementation in Modulation of Gut Microbiome and T-Regulatory Cells in Health and Psoriatic Disease [abstract].. Arthritis Rheumatol., 2018. 70 (suppl 9). p. https://acrabstracts.org/abstract/effects-of-fatty-acid-supplementation-in-modulation-of-gut-microbiome-and-t-regulatory-cells-in-health-and-psoriatic-disease/. [Google Scholar]
  • 42.Dürholz K, Hofmann J, Iljazovic A, Häger J, Lucas S, Sarter K, Strowig T, Bang H, Rech J, Schett G, Zaiss MM. Dietary short-term fiber interventions in arthritis patients increase systemic SCFA levels and regulate inflammation. Nutrients. 2020. 12(10): p. 3207. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Martinsson K, Dürholz K, Schett G, Zaiss MM, Kastbom A. Higher serum levels of short-chain fatty acids are associated with non-progression to arthritis in individuals at increased risk of RA. Annals of the Rheumatic Diseases. 2022. Mar 81(3): p. 445–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Yao Y, Cai X, Zheng Y, Zhang M, Fei W, Sun D, Zhao M, Ye Y, Zheng C. Short‐chain fatty acids regulate B cells differentiation via the FFA2 receptor to alleviate rheumatoid arthritis. British Journal of Pharmacology. 2022. (17): p. 4315–29. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

fig S1

Supplementary Figure 1. Serum bone remodeling markers and concentration of individual SCFAs in mice after SCFA treatment. (A) Serum levels of bone remodeling markers Osteocalcin, P1NP, and CTX-1 by genotype and treatment. (B) Relative concentration of the individual short chain fatty acids (SCFAs) in the fecal samples, liver samples and plasma samples of WT mice treated or untreated with SCFAs. (C) Representative flow gating strategy to identify Th17, Th22, Th1, and Th2 cells after doublet exclusion. (CTX-1 – type I collagen cross-linked c-telopeptide, P1NP – procollagen 1 intact N-terminal propeptide. A-B: each point represents an individual mouse).

NIHMS1947151-supplement-fig_S1.pdf (543.2KB, pdf)
fig S2

Supplementary Figure 2. SCFA supplementation in terminal osteoclast differentiation from OCPs and differential gene expression of bone marrow progenitor cells of R26STAT3Cstopfl/fl CD4Cre mice. (A) Representative images of TRAP+ osteoclasts after 4 days of differentiation from sorted osteoclast progenitor cells of control and R26STAT3Cstopfl/fl CD4Cre mice. (B) Quantification of TRAP+ osteoclasts differentiated from osteoclast progenitors in the absence and presence of individual short chain fatty acids (acetate, propionate and butyrate). (C) Relative expression of transcription factor NFATc1 in osteoclasts differentiated from the osteoclast progenitors in the absence and presence of short chain fatty acids. (D) Relative expression of transcription factor c-Fos in osteoclasts differentiated from the osteoclast progenitors in the absence and presence of short chain fatty acid. (E) Heatmap of the top differentially expressed genes in bone marrow progenitor cells across R26STAT3Cstopfl/fl CD4Cre and control mice, showing genes with base mean > 50, p-value < 0.005, & LogFC >2.0). (F) Heatmap of the top differentially expressed genes in bone marrow progenitor cells across SCFA-treated and untreated R26STAT3Cstopfl/fl CD4Cre mice, showing genes with base mean > 100, p-value < 0.005, & LogFC >2.0). (A-D n=4 mice per condition).

NIHMS1947151-supplement-fig_S2.pdf (1.5MB, pdf)

RESOURCES