BHB generation are decreased in the colonic mucosa from active IBD patients
Given the evident significance of ketogenic enzyme in ketogenesis, we first analyzed the expression levels of genes encoding ketogenic enzymes in public datasets of IBD samples and found that the mRNA levels of ACAT1, HMGCS2, and BDH1 were significantly decreased in the colonic mucosa from active IBD patients compared to healthy controls, while there was not significant difference in the mRNA level of HMGCL [34]. Consistently, we found that the protein levels of ACAT1, HMGCS2, and BDH1 were significantly reduced in the colonic mucosa from active IBD patients compared to healthy controls (Fig. 1a, b). In addition, DSS mice had decreased expression of ACAT1, HMGCS2, and BDH1 in the colonic tissues (Additional file 1: Fig. S1a-c). These findings suggest that IBD patients may have reduced endogenous ketone metabolites in the colon.
Since BHB is the most prominent ketone metabolite, we subsequently detected the level of BHB by β-Hydroxybutyrate colorimetric assay kit and found that the level of BHB was significantly decreased in the colonic mucosa from active IBD patients compared to healthy controls (Fig. 1c). Furthermore, we explored the clinical impact of BHB in IBD patients and found that colonic BHB levels were negatively correlated with the IBD activity index (Mayo scores in the UC and CDAI in the CD) and ESR values (Fig. 1d). Taken together, these results suggest that decreased BHB production in the colon might be involved in the pathogenesis of IBD.
Exogenous BHB supplement attenuates DSS-induced colitis
To assess whether BHB participates into experimental colitis, we first conducted a study on the colonic level of BHB in mice challenged with 2.5% DSS in drinking water. BHB levels were markedly decreased in the colonic tissues after 6-day DSS treatment (Fig. S1d). This result suggests that BHB may be involved in the development of colitis.
To further confirm the role of BHB in DSS-induced colitis, we next rectally injected exogenous BHB or saline as a control to DSS-exposed mice, according to the methods shown in Fig. 2a. BHB-treated mice showed significantly reduced weight loss, disease activity index, and colon shortening, compared to control mice (Fig. 2b–e). Moreover, histological examination displayed less crypt loss, epithelium damage, and decreased inflammation in the colons of BHB-treated mice than in those of control mice (Figs. 1g and 2f). However, without DSS exposure, BHB administration had no effect on body weight, colonic length, and histology (Fig. 2b, e, g). These results indicate that BHB protects mice from DSS-induced colitis may through inhibiting colonic inflammation and promoting post-injury repair of colonic tissues.
BHB exerts protective effect against colitis dependently via colonic macrophages
Since intestinal epithelial barrier, gut microbiota, and macrophages are critical for the pathological processes associated IBD, we investigated whether the beneficial effect of BHB on colitis is dependent on these factors. First, we examined whether BHB affects intestinal epithelial integrity by measuring the levels of immunofluorescence staining of tight junction proteins ZO-1 and occluding. No difference was noted in ZO-1 and occluding immunostaining between BHB-treated and control groups (Additional file 1: Fig. S2a-d).
The 16S diversity analysis showed that there was no significant difference in alpha diversity between the two groups of mice (Additional file 1: Fig. S3a-b). However, the relative abundance of specific strains, such as Firmicutes at the phylum level and Lachnospiraceae at the family level, was significantly increased in the BHB group when compared to that in the saline group (Additional file 1: Fig. S3c-e); unweighted PCoA analysis revealed that clusters of gut microbiota between BHB and saline mice were entirely separated, suggesting that the gut flora exhibited a very different composition in two groups (Additional file 1: Fig. S3d).
Therefore, we explored whether BHB intervention attenuates DSS-induced colitis is dependent on gut microbiota. To this end, mice were treated with antibiotics (ABX) to ablate the impact of intestinal microbiota followed by rectally injecting exogenous BHB or saline as a control to DSS-exposed mice (Additional file 1: Fig. S4a). After ABX treatment, the mice received BHB enema still developed decreased colitis as reflected by less body weight loss, disease activity index, and colon shortening, as well as significantly reduced histology scores, compared to control mice (Additional file 1: Fig. S4b-g). These results indicate that BHB ameliorates DSS-induced colitis independently of intestinal microbiota.
Subsequently, we asked whether intestinal macrophages are involved in the protective effect of BHB during colitis development. To test it, mice were intraperitoneally injected with clodronate liposomes to accomplish macrophage depletion (Fig. 3a, b). Notably, ablation of macrophage largely abrogated the protective effect of BHB, which failed to improve the clinical parameters of DSS-induced colitis such as body weight loss, disease activity index, and colon shortening (Fig. 3c–f). Histological examination further demonstrated that the beneficial effect of BHB was largely diminished when deletion of intestinal macrophage (Fig. 3g, h). These results show that the protective effect of BHB against DSS-induced colitis is depend on intestinal macrophages.
The protective effect of BHB against colitis involves M2 macrophage polarization
A previous study reported that BHB reduces neuroinflammation through shaping microglia, a group of ramified brain-resident phagocytes, toward anti-inflammatory M2 polarization [35]. Thus, we proposed that BHB administration attenuates DSS-induced colitis may through regulating M2 macrophage polarization in the intestine. As expected, BHB administration resulted in significantly increased mRNA expression of M2 macrophage-associated genes including IL-4Ra, IL-10, arginase 1 (Arg-1), and chitinase-like protein 3 (Chil3), and their downstream effectors, such as FGF2, FGF7, TGFb, TGFbr1, PDGFra, and PDGFrb in the colons of DSS-exposed mice (Fig. 4a, b). Moreover, BHB significantly increased in the number of F4/80+CD206+ M2 macrophages in the colons of DSS-exposed mice as determined by immunofluorescence staining (Fig. 4c, d). Taken together, these results indicate that BHB protects mice from DSS-induced colitis through enhancing M2 macrophage polarization.
BHB promotes IL-4-induced M2 macrophage polarization through the JAK2-STAT6 signaling pathway
To confirm the function of BHB on the induction of M2 macrophage polarization, BHB was added to bone marrow-derived macrophages (BMDMs) isolated from mice in vitro. In agreement with the in vivo data, the addition of BHB significantly increased the mRNA expression of M2-associated genes, including Arg-1, IL-10, Chil3, and Retnla, in IL-4-stimulated BMDMs, but not in BMDMs (Fig. 5a). However, BHB addition did not affect the mRNA expression of M1-associated genes, such as NOS2, TNF-α, IL-6, and IL12p40, in BMDMs, nor in LPS-stimulated BMDMs (Fig. 5b). These results show that BHB promotes IL-4-induced M2 macrophage polarization.
To further dissect the underlying mechanism, BMDMs isolated from mice were stimulated with IL-4 alone or with IL-4 plus BHB, and the mRNA expression profiles were analyzed by RNA sequencing. We observed that a large number of genes, including M2 macrophage-specific genes Arg-1, Chil3, and Retnla, were upregulated in the IL-4 plus BHB group compared to the IL-4 group (Fig. 5c, d). With Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis, it was found that these upregulated genes were significantly enriched in regulation of the JAK-STAT signaling pathway (Fig. 5e). Moreover, protein–protein interaction network analysis of these upregulated genes identified Janus kinase 2 (JAK2) as a downstream signaling molecule of IL-4Ra (Fig. 5f). Binding of IL-4 to IL-4Rα results in tyrosine phosphorylation and activation of STAT6, which is a key signaling node for M2 macrophage polarization [36, 37]. Our findings combined with previous findings suggest that BHB promotes M2 macrophage polarization may through the JAK2-STAT6 signaling pathway.
We next performed western blot analysis to examine whether BHB influences the activation of the JAK2-STAT6 signaling pathway in BMDMs. STAT6 phosphorylation was undetected in unstimulated BMDMs or in stimulated cells with BHB alone. As expected, obvious STAT6 phosphorylation was observed in IL-4-stimulatedBMDMs (Fig. 6a). The addition of BHB increased the intensity of phosphorylated STAT6, but not total STAT6, in IL-4-stimulated BMDMs (Fig. 6a). This result was further confirmed by a significant increase in pSTAT6 when normalized to total STAT6 (Fig. 6b). Consistently, IL-4 and BHB together resulted in significantly increased JAK2 phosphorylation, but both alone cannot in BMDMs (Fig. 6a, b). Collectively, these results indicate that BHB enhances IL-4-induced activation of the JAK2-STAT6 signaling pathway in BMDMs.
We subsequently examined whether the upregulation effect of BHB on IL-4-induced M2 macrophage polarization is dependent on the JAK2-STAT6 signaling pathway. To this end, AS1517499 (AS), a potent STAT6 phosphorylation inhibitor, was administrated in IL-4-stimulated BMDMs with or without BHB, and the mRNA expression of M2-associated genes was measured by real-time PCR. We found that AS treatment effectively suppressed STAT6 phosphorylation in IL-4-stimulated BMDMs (Fig. 6c, d). Moreover, the upregulation effect of BHB on IL-4-induced STAT6 phosphorylation was obviously inhibited by AS (Fig. 6c, d). Consistently, BHB significantly increased the mRNA levels of Arg-1, IL-10, Chil3, and Retnla in IL-4-stimulated BMDMs, but this upregulation role was completely inhibited by AS (Fig. 6e). Taken together, these results show that BHB could directly promote IL-4-induced M2 macrophage polarization through enhancing the JAK2-STAT6 signaling pathway.
The protective effect of BHB against colitis is dependent on STAT6 activation
Because STAT6 activation is a critical signaling node for BHB’s regulating M2 macrophage polarization, we speculated that it is also required for BHB therapeutic role during colitis development. To test it, BHB enema or saline as a control was rectally injected to DSS-exposed mice, then these mice were received intraperitoneal injection of AS1517499 (Additional file 1: Fig. S5a). The mice received BHB treatment displayed increased p-STAT6 phosphorylation in F4/80+ macrophages compared to control mice (Additional file 1: Fig. S6a and c), which was effectively inhibited by AS administration in vivo (Additional file 1: Fig. S6b and d). Consistently, AS administration largely eliminated the protective effect of BHB against DSS-induced colitis as reflected by no significant difference in weight loss, disease activity index, colon shortening, and histology scores between the BHB and saline groups (Additional file 1: Fig. S5b-g). These results indicate that the protective effect of BHB against DSS-induced colitis is dependent on the STAT6 signaling pathway.
BHB-treated M2 macrophages promote intestinal epithelial regeneration
Previous studies have shown that M2 macrophage contributes to tissue repair and regeneration [38, 39]. In this study, we found that BHB treatment promotes intestinal epithelial proliferation during colitis development (Fig. 7a, b). Therefore, we evaluated whether BHB regulates M2 macrophage function in intestinal epithelial proliferation. To test it, BMDMs isolated from mice were stimulated with IL-4 alone or with IL-4 and BHB together to induce M2 macrophage polarization. Then, the cell culture supernatant purified by ultrafiltration was used to stimulate the normal intestinal epithelial cell line IEC-6 and cell proliferation was measured by CCK-8 assays. We found that the levels of cell proliferation were significantly higher in the IL-4 plus BHB group than in the IL-4 group (Fig. 7c). However, this effect was completely inhibited by AS1517499 (Fig. 7c).
To mimic the in vivo environment and investigate the effect of BHB-treated M2 macrophages on intestinal epithelium proliferation, mouse colonic crypts were isolated and cultured in vitro. Then, we co-cultured these colonic organoids with IL-4-stimulated BMDMs with or without BHB in a transwell system, which allows extracellular cytokines derived from the BMDMs to contact the organoids directly. We found that organoids grew without gross abnormalities in the presence of IL-4. Moreover, the number of colonic organoids were significantly higher in the IL-4 plus BHB group than in the IL-4 group (Fig. 7d, e). Moreover, in the IL-4 plus BHB group, organoids were larger in size and show higher percentage of buddings than that in the IL-4 group (Fig. 7f–h). However, these effects were obviously inhibited by AS1517499 (Fig. 7d–h). Taken together, these results indicate that BHB-treated M2 macrophages promote intestinal epithelial cells proliferation and organoids growth through the STAT6-dependent signaling pathway.