3-Hydroxybutyrate ameliorates insulin resistance by inhibiting PPARγ Ser273 phosphorylation in type 2 diabetic mice

3HB treatment reduced insulin resistance in T2D mice

To explore the effect of 3HB on T2D, leptin receptor deficiency mice (db/db) and streptozotocin (STZ) induced T2D mice were used.16 1,3-butanediol (1,3-BDO), a precursor of endogenous 3HB, could be efficiently converted to 3HB in the liver.17,18 Considering the effect of increasing blood 3HB concentration and the influence on fluid intake volume of mice, 10% 1,3-BDO aqueous solution was chosen as drinking fluid for the mice to maintain a blood 3HB level of over 1.5 mM based on the literatures and corresponding experimental results (Supplementary Fig. S5b).19,20

Vehicle-treated db/db mice (db/db-Ctrl) developed pronounced hyperglycemia, with fasting blood glucose (FBG) levels ~18 mM (Fig. 1b). 10% 1,3-BDO treated mice (db/db-1,3-BDO) showed a significant reduction in FBG, which was <12 mM (Fig. 1b). The improvement in glucose homeostasis was achieved without a significant change in the body weight (Fig. 1a). Intraperitoneal glucose tolerance test (IPGTT) was performed to evaluate the body’s functionality of controlling blood glucose. While db/db-1,3-BDO mice did not show improved glucose clearance compared to db/db-Ctrl mice (Fig. 1c, d). Meanwhile, 3HB treatment did not change the fasting serum insulin level (Fig. 1e). However, the calculated results of HOMA-IR, a homeostatic model of insulin resistance, showed that 3HB could significantly reduce HOMA-IR by 52% (Fig. 1f), indicating that 3HB effectively reduced insulin resistance levels in db/db mice.

Fig. 1figure 1

3HB Treatment Reduces Insulin Resistance in T2D Mice. a Body weights of db/db mice (n = 10). b Overnight fasted blood glucose level of db/db mice (n = 10). c The glucose tolerance test of db/db mice. Glucose (2 g/kg body weight) was intraperitoneally (i.p.) administered in overnight-fasted db/db mice. (n = 5) d The area under the curve (AUC) in c. e Fasting serum insulin levels of db/db mice after 8 weeks of indicated treatment. (n = 6) f A homeostasis model for assessment of the insulin resistance index (HOMA-IR) of the mice in e. g Body weights of STZ induced T2D mice. (n = 10) h Overnight fasted blood glucose level of STZ mice. (n = 10) i The glucose tolerance test of STZ mice. Glucose (2 g/kg body weight) was intraperitoneally (i.p.) administered in overnight-fasted STZ mice. (n = 5–9) j The area under the curve (AUC) in i. k Fasting serum insulin levels of STZ mice after 8 weeks of indicated treatment. (n = 5–9). l A homeostasis model for assessment of the insulin resistance index (HOMA-IR) of STZ mice. (n = 5–9) Data reported as mean ± SD, *p < 0.05, **p < 0.001, and ***p < 0.0001

C57BL/6 J and HCAR2-/- (C57BL/6 J background) male mice treated with high-fat diet combined with low doses of STZ progressively lost glucose control compared to normal diet mice (ND) (Fig. 1h). The FBG in the ND group was always maintained within the normal range of 3.9–6.1 mM, while the FBG of the STZ-Ctrl group gradually increased and tended to be stabilized above 16 mM. Compared with STZ-Ctrl, the FBG of STZ-1,3-BDO mice was consistently and significantly lower than that of STZ-Ctrl group, and the highest value was just 13.3 mM (Fig. 1h). STZ-HCAR2-/- mice consistently had lower FBG than wild-type T2D mice (Fig. 1h), which was attributed to the effect of HCAR2 deficiency. However, limited to no benefit on FBG was provided by 1,3-BDO treatment in STZ-HCAR2-/- mice. The mean FBG of STZ-HCAR2-/--1,3-BDO group seemed lower than that of STZ-HCAR2-/--Ctrl, but without statistical difference. Therefore, even though the knockout of HCAR2 resulted in the lower FBG in STZ-HCAR2-/- mice, it could still be found that HCAR2 may mediate the hypoglycemic effect of 3HB on FBG in STZ-induced T2D mice to some degree. When comparing the weight changes of mice in each group, the body weight of ND group increased constantly over time. Although fed on a high-fat diet, both STZ-Ctrl and STZ-1,3-BDO mice body weights did not change significantly during the studies (Fig. 1g). The body weights of HCAR2-/- mice were significantly higher than that of wild-type mice, which was consistent with the reported phenomenon of HCAR2-/- mice.21 Unlike in db/db mice, 1,3-BDO treatment showed improved glucose clearance compared to STZ-Ctrl mice in IPGTT (Fig. 1i, j). After the glucose injection, the blood glucose levels of mice in the STZ-1,3-BDO group were significantly lower than those in the STZ-Ctrl group. The blood glucose levels of both STZ-HCAR2-/- groups were consistently lower than those in STZ-Ctrl group, but the effect of 1,3-BDO was not reflected (Fig. 1i, j). Compared with the ND group, the fasting insulin of the STZ-Ctrl and STZ-1,3-BDO groups was significantly reduced by >55%, but there was no difference between the two groups, indicating that 1,3-BDO had no effect on the level of serum insulin (Fig. 1k). Surprisingly, the serum fasting insulin levels of the two STZ-HCAR2-/- groups were almost the same as that of ND group, and also with no inter-group difference (Fig. 1k). This suggested that HCAR2 deletion may cause an impact on insulin expression or secretion in mice, and also explained the reason why STZ-HCAR2-/- mice had lower blood glucose levels than wild-type STZ-induced T2D mice did (Fig. 1h). The calculated results of HOMA-IR showed that the level of insulin resistance in the STZ-Ctrl group was increased by about 48% compared to that of the ND group, while 3HB provided 55% reduction of HOMA-IR value in STZ-1,3-BDO group compared to STZ-Ctrl mice (Fig. 1l). Although STZ-HCAR2-/- mice had lower FBG levels (Fig. 1h), it was attributed to the effect of higher levels of insulin in STZ-HCAR2-/- mice. Therefore, the insulin resistance levels in STZ-HCAR2-/- mice were similar to that of STZ-Ctrl group, and 1,3-BDO had no significant effect on insulin resistance in STZ-HCAR2-/- mice (Fig. 1l).

Apart from the findings above, 1,3-BDO treatment reduced adipose tissue, liver and kidney damage caused by T2D, and improved blood lipids and liver functions (Supplementary Fig. S2-S4). The protective effect on adipose tissue and liver was mediated by HCAR2 (Supplementary Fig. S3). In combination, in vivo application of 1,3-BDO resulted in reduced FBG, insulin resistance level and tissue injury through HCAR2 in T2D mice, strongly supporting the potential therapeutic value of 3HB in T2D.

3HB regulated Ca2+, cAMP, ERK1/2 and PPARγ related biological processes or signaling pathways

To further determine the underlying molecular mechanism of 1,3-BDO treatment in T2D mice, we performed RNA-sequencing to detect the gene expression profiles of adipose tissue of each group, and cluster maps were drawn according to differentially expressed genes (Fig. 2a, e). There were 800 differentially expressed genes in adipose tissue samples of mice in db/db-Ctrl group and db/db-1,3-BDO group. The heatmap (red and blue colors indicate increased and decreased gene expressions, respectively) showed that 1,3-BDO treatment could significantly alter the gene expression pattern of adipose tissue in db/db mice (Fig. 2a). 2461 differentially expressed genes were detected in the adipose tissue of ND, STZ-Ctrl, and STZ-1,3-BDO groups. Compared with STZ-Ctrl group, 1,3-BDO could alter the gene expression pattern of adipose tissue in T2D mice and promote the reversal of it to that of ND mice (Fig. 2e).

Fig. 2figure 2

Transcriptomics Analysis for Adipose Tissue of T2D Mice. a After 8 weeks of indicated treatment, the adipose tissue of db/db mice was collected and studied the differential gene expression analysis (p < 0.05) by RNA sequencing. b Downstream genes of PPARγ in differentially expressed genes of db/db mice. c Gene ontology (GO) enrichment analysis of db/db mice. The related biological processes (BP), cellular component (CC) and molecular function (MF) were analyzed. d Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis of db/db mice. e After 8 weeks of indicated treatment, the adipose tissue of STZ induced T2D mice was collected and conducted the differential gene expression analysis (p < 0.05) by RNA sequencing. f Downstream genes of PPARγ in differentially expressed genes of STZ mice. g GO enrichment analysis of STZ mice. The related biological processes (BP), cellular component (CC) and molecular function (MF) were analyzed. h KEGG enrichment analysis of STZ mice

Related enriched Gene Ontology (GO) terms were shown in Fig. 2c, g. In terms of biological processes (BP), 3HB was involved in Ca2+ and cAMP related processes, glucose homeostasis, insulin signaling pathways and extracellular signal-regulated kinases 1/2 (ERK1/2) regulation of the MAPK signaling pathway in adipose tissue.

The KEGG enrichment results showed that the regulation of PPAR, MAPK, Ca2+, cAMP, adipokines, insulin and some other signaling pathways were affected by 1,3-BDO administration (Fig. 2d, h). PPARγ is the main type of PPAR in adipose tissue, and is a major target for drug development in T2D,22,23 suggesting that 3HB could act as an insulin resistance reduction agent through PPARγ signaling pathway. Meanwhile, cAMP and Ca2+ can be the two important second messengers in the possible cell signal transductions caused by 3HB.

PPARγ regulates the transcription of multiple genes involved in adipogenesis and glucose homeostasis.24 If 1,3-BDO treatment has a regulatory effect on PPARγ-related signaling pathways in adipose tissue, the expression of PPARγ regulated downstream genes should also be affected. We screened a total of 39 reported PPARγ downstream genes from all differentially expressed genes of db/db mice, and performed cluster analysis on the screening results. 1,3-BDO treatment significantly changed the expression of PPARγ downstream genes in adipose tissue of db/db mice (Fig. 2b). A total of 136 reported PPARγ regulated genes were found out from all differentially expressed genes of ND, STZ-Ctrl, and STZ-1,3-BDO groups, and the cluster analysis on the screening results showed that the occurrence of T2D caused great changes in the expression of PPARγ downstream genes. However, 1,3-BDO treatment changed the expression pattern of PPARγ downstream genes in adipose tissue of T2D mice, and made it closer to that of ND group (Fig. 2f). In HCAR2-/- groups, the difference of total differentially expressed genes and PPARγ downstream genes patterns between STZ-HCAR2-/--Ctrl and STZ-HCAR2-/--1,3-BDO groups was not significant (Supplementary Fig. S1a and S1b). In summary, 1,3-BDO treatment affects the expression of PPARγ controlling downstream genes via HCAR2 in adipose tissues.

3HB inhibits phosphorylation of PPARγ Ser273 in vitro and in vivo

1,3-BDO treatment affects the transcriptional level of PPARγ downstream target genes through HCAR2 (Fig. 2b and f, Supplementary Fig. S1a and S1b). PPARγ agonists are widely used in the treatment of T2D and related complications by enhancing insulin sensitivity.2,25 To investigate the exact effect of 3HB on PPARγ, the LanthaScreen TR-FRET PPARγ Competitive Binding Assay Kit was used to investigate whether 3HB is a PPARγ ligand. The IC50 of rosiglitazone (Rosi) in this assay was calculated to be 37 nM, which is consistent with reported literature,26 but 3HB could not competitively bind to PPARγ ligand binding domain, indicating that 3HB is not a ligand of PPARγ (Fig. 3a). Meanwhile, 3HB had no effect on PPARγ transcriptional activity (Fig. 3b). PPARγ regulates the transcription of genes involved in lipogenesis, and PPARγ agonists or inhibitors could promote or inhibit the differentiation of preadipocytes.2,24 However, 3HB could not affect the differentiation of 3T3-L1 preadipocytes (Fig. 3c, d). To sum up, 3HB has no direct action on PPARγ. Whereas, in 3T3-L1 adipocytes, 3HB could significantly promote insulin-dependent glucose uptake with a weak dose-dependent effect after 24 h treatment. While in the HCAR2 knockdown group, the promoting effect of 3HB disappeared (Fig. 3e), indicating that 3HB might enhance insulin sensitivity in adipocytes via HCAR2 mediation.

Fig. 3figure 3

3HB Inhibits Phosphorylation of PPARγ Ser273 in vitro and in vivo. a TR-FRET PPARγ ligand displacement assay of 3HB and positive control rosiglitazone. b Influence on PPARγ transcriptional activity of 3HB or 0.5 μM rosiglitazone was measured using a 293 T cells based on the luciferase reporter assay. c Oil red-O staining of mature 3T3-L1 adipocytes after 8 days of 3HB treatment. d Quantification of the staining results in c was presented relative to the control. e 3T3-L1 adipocytes with or without siRNA knockdown of HCAR2 were pretreated with 3HB or 0.5 μM rosiglitazone for 24 h. After 10 min stimulation of insulin, the glucose uptake of cells was measured by a bioluminescent assay. f 3T3-L1 adipocytes with or without siRNA knockdown of HCAR2 were treated with TNFα, followed by treatment with 3HB or rosiglitazone (Rosi) for 1 h. Phosphorylated PPARγ at Ser273 and total PPARγ were detected using anti-pPPARγ or anti-PPARγ antibodies, respectively. hi pPPARγ (Ser273) and total PPARγ of adipose tissue in 8-week treated db/db mice, STZ induced T2D mice and STZ induced HCAR2-/- T2D mice, respectively. jm The amounts of pPPARγ (Ser273) and total PPARγ in fi were quantified using ImageLab, and pPPARγ/ PPARγ ratio was calculated. The results (expressed as the mean ± SD, n ≥ 3) were presented relative to the control (3HB 0 mM, db/db-Ctrl, STZ-Ctrl or STZ-HCAR2-/--Ctrl), *p < 0.05, **p < 0.001, and ***p < 0.0001, **** p < 0.00001

In fact, ligands of PPARγ, such as TZDs, may cause overactivation of PPARγ, leading to obesity, bone loss, heart failure and other serious side effects.4 Posttranslational modification of PPARγ is a new drug research direction to optimize the function of PPARγ and avoid adverse reactions.27 For example, PPARγ serine 273 site can be phosphorylated by CDK5 or MEK/ERK, resulting in functional disorder of PPARγ, inhibiting the expression of genes related to insulin sensitivity, and finally leading to insulin resistance.8,28 Inhibition of phosphorylation at this site significantly improves insulin resistance while avoiding the side effects associated with over-activation of PPARγ.8,28 To explore the effect of 3HB on the phosphorylation of PPARγ Ser273 and the role of HCAR2 in this process, 3T3-L1 adipocytes transfected with si-Ctrl or si-HCAR2 interfering RNA were stimulated with TNFα for 1 h to increase the cellular phosphorylation level of PPARγ Ser273, and followed with 3HB or Rosi treatment for another 1 h. The phosphorylation level of PPARγ Ser273 was detected by Western blot analysis, and the ratio of phosphorylated PPARγ (p-PPARγ)/total PPARγ was calculated to reflect the phosphorylation level of PPARγ Ser273. Both 3HB and Rosi could significantly inhibit the phosphorylation of PPARγ Ser273. However, when HCAR2 was silenced, the inhibitory effect of 3HB on PPARγ Ser273 phosphorylation disappeared (Fig. 3f). The results of quantitative analysis also showed that 1 mM and 5 mM 3HB effectively inhibited the phosphorylation level of PPARγ Ser273 by 38 and 44%, respectively, which was in a dose-dependent manner and even lower than the non-TNFα group (Fig. 3j). Therefore, 3HB could inhibit PPARγ Ser273 phosphorylation through HCAR2 in 3T3-L1 adipocytes.

In order to investigate whether 1,3-BDO treatment inhibits the phosphorylation of PPARγ Ser273 in the adipose tissue of T2D mice, the total protein of the adipose tissue of db/db mice was extracted for Western blot detection. 1,3-BDO treatment indeed reduced the phosphorylation of PPARγ Ser273 in the adipose tissue of db/db mice, along with significantly increased total amount of PPARγ (Fig. 3g). Quantitative analysis showed that 1,3-BDO treatment reduced the phosphorylation level of PPARγ Ser273 in adipose tissue by 47% (Fig. 3k). Compared with healthy mice in the ND group, the total protein expression of PPARγ in the adipose tissue of STZ-Ctrl was decreased, yet the phosphorylation of PPARγ Ser273 was significantly increased. 1,3-BDO treatment promoted the expression of PPARγ in the adipose tissue of STZ-1,3-BDO mice, together with a 76% reduction in PPARγ Ser273 phosphorylation (Fig. 3h and 3l). In HCAR2 knockout T2D mice, the inhibitory effect of 3HB on PPARγ Ser273 phosphorylation was abolished (Fig. 3i, m), suggesting that HCAR2 was essential for 3HB to inhibit PPARγ Ser273 phosphorylation in adipose tissue of T2D mice.

3HB inhibited phosphorylation of PPARγ Ser273 via HCAR2/Ca2+/cAMP/PKA/Raf1/ERK1/2/PPARγ pathway

Previous studies had shown that ERK1/2 was involved in the phosphorylation of PPARγ Ser273, and its inhibitors can significantly decrease PPARγ Ser273 phosphorylation level in adipose tissue of obese mice.8 The activation of HCAR2 can increase intracellular Ca2+ concentration in macrophages and hippocampal neurons, which then activates AC, and in turn increases intracellular cAMP levels.13,29 The above results showed that 3HB inhibited PPARγ Ser273 phosphorylation via HCAR2 and enhanced insulin sensitivity. To verify if cAMP, Ca2+ and ERK1/2 were involved in this process, 3T3-L1 adipocytes transfected with si-Ctrl or si-HCAR2 interfering RNA were treated with different concentrations of 3HB or 100 μM forskolin for 1 h, and then the intracellular cAMP concentration was studied. 3HB could significantly increase the cAMP level to >6.5 nmol/L. When HCAR2 expression was knocked down, the effect of 3HB was abolished (Fig. 4a), suggesting that 3HB could increase intracellular cAMP level in adipocytes through HCAR2. 3HB could also significantly increase the levels of intracellular Ca2+ through HCAR2 in a dose-dependent manner after 1 h treatment (1 mM treated group could increase by 16.9%; 5 mM group by 27.1%) (Fig. 4b). In order to explore whether 3HB affects the activity of ERK1/2 or not, we treated 3T3-L1 adipocytes with 3HB or 10 μM ERK1/2 inhibitor SCH772984 for 1 h, and then the ERK1/2 activity was detected, which was reflected by the final NADH production efficiency of the ERK1/2 catalytic reaction system. 3HB, like the ERK1/2 inhibitor SCH772984, could effectively inhibit ERK1/2 activity in a dose-dependent manner (1 mM 3HB could reduce ERK1/2 activity by~26%; 5 mM by~35%). When the expression of HCAR2 was knocked down by siRNA, the inhibitory effect of 3HB on ERK1/2 activity also disappeared (Fig. 4C), demonstrating that 3HB could inhibit ERK1/2 activity in adipocytes via HCAR2.

Fig. 4figure 4

3HB Inhibits Phosphorylation of PPARγ Ser273 Through HCAR2/Ca2+/cAMP/PKA/Raf1/ERK1/2 Pathway. a 3T3-L1 adipocytes with or without siRNA knockdown of HCAR2 were treated with 3HB or 100 μM forskolin for 1 h. Intracellular cAMP concentration was detected using an ELISA kit. b Intracellular Ca2+ levels of 3T3-L1 adipocytes treated with 3HB or 1 μM ATP for 1 h were detected with Fura-2/AM probe kit. The results were presented relative to 3HB 0 mM. c The kinase activity of ERK1/2 of 3T3-L1 adipocytes treated with 3HB or 10 μM SCH779284 for 1 h was measured using an ERK1/2 kinase Activity Quantitative detection Kit. d 3T3-L1 adipocytes with or without siRNA knock down of HCAR2 were treated with 3HB or 0.5 μM rosiglitazone (Rosi) for 1 h. Phosphorylated PKA at Thr197, Raf1 at Ser259, ERK1/2 at Thr202/Tyr204, PPARγ at Ser273, and total PKA, Raf1, ERK1/2, PPARγ, HCAR2, GAPDH were studied using corresponding antibodies. eh The amounts of phosphorylated and total proteins in d were quantified using ImageLab, and pPKA/PKA, pRaf1/Raf1. pERK1/2/ERK1/2 and pPPARγ/ PPARγ ratio was calculated. The results (expressed as the mean ± SD, n = 3) were presented relative to 3HB 0 mM, *p < 0.05, **p < 0.001, and ***p < 0.0001

Studies have shown that protein kinase A (PKA) activated by cAMP can phosphorylate the Ser259 site of Raf1 proto-oncogene serine/threonine-protein kinase (Raf1), the upstream kinase of MEK and ERK1/2, and the phosphorylation of Ser259 will lead to the inactivation of Raf1. Inactive Raf1 cannot activate MEK, which eventually leads to the inhibition of ERK1/2 activity.30,31,32 3HB could increase the concentration of cAMP in 3T3-L1 adipocytes and inhibit the activity of ERK1/2. Thus, these two effects of 3HB are likely to be linked by a putative pathway: cAMP/PKA/Raf1/ERK1/2. Total protein of 3T3-L1 adipocytes treated with 3HB or 0.5 μM Rosi for 1 h was extracted for Western blot assays of phosphorylation of PKA (Thr197), ERK1/2 (Thr202/Tyr204), Raf1 (Ser259) and PPARγ (Ser273). 3HB significantly increased the phosphorylation levels of PKA (Thr197) and Raf1 (Ser259), but decreased ERK1/2 (Thr202/Tyr204) and PPARγ (Ser273) phosphorylation in adipocytes. However, these effects were all abolished in HCAR2 silenced adipocytes (Fig. 4d–h). In other words, 3HB effectively activated PKA, inhibited Raf1, reduced the activity of ERK1/2 and the phosphorylation level of PPARγ Ser273 via HCAR2.

To sum up, 3HB could increase intracellular Ca2+ and cAMP levels via HCAR2 to activate PKA, thereby inhibiting Raf1 activity, resulting in a decrease in ERK1/2 activity, and ultimately decreasing the phosphorylation level of PPARγ Ser273. The pathway of 3HB/HCAR2/Ca2+/cAMP/PKA/Raf1/ERK1/2/PPARγ might be the molecular mechanism by which 3HB ameliorates insulin resistance.

3HB regulated the expression of genes affected by PPARγ Ser273 phosphorylation in 3T3-L1 adipocytes

The expression levels of PPARγ regulated downstream genes in adipocytes was studied using the real-time quantitative PCR, and several genes affected by phosphorylation of PPARγ Ser273 were also chosen.28 Although 3HB could not change the transcriptional level of PPARγ, it had a significant regulatory effect on the expression of PPARγ regulated genes, which was similar to Rosi (Fig. 5a). HCAR2 siRNA could effectively inhibit the transcription of Hcar2 by ~86% (Fig. 5e), while 3HB had no effect on the transcription of PPARγ downstream genes in these adipocytes, the effect of Rosi still existed (Fig. 5b). When inhibiting ERK1/2 kinase activity or antagonizing PPARγ in adipocytes with SCH772984 or GW9662, the effects of 3HB and Rosi on the expression of most selected genes were abolished (Fig. 5c, d). The above results suggested that 3HB could regulate the transcription of PPARγ downstream genes affected by Ser273 phosphorylation via HCAR2 and ERK1/2 in 3T3-L1 adipocytes, further revealing that the effect of 3HB on the phosphorylation of PPARγ Ser273 was depending on HCAR2 and ERK1/2.

Fig. 5figure 5

3HB Affects the Expression of Genes Regulated by PPARγ Ser273 Phosphorylation in 3T3-L1 Adipocytes. a, b Gene expression of adipocytes with or without siRNA knockdown of HCAR2 after 6 h 5 mM 3HB or 0.5 μM rosiglitazone (Rosi) treatment. c, d Gene expression of adipocytes treated with 5 mM 3HB or 0.5 μM for 6 h in the presence of 10 μM SCH779284 or 20 μM GW9662. e Relative HCAR2 expression of adipocytes with or without siRNA knockdown of HCAR2. f 3T3-L1 adipocytes with or without siRNA knockdown of HCAR2 were treated with 3HB or 0.5 μM rosiglitazone (Rosi) for 48 h in the absence or presence of 10 μM SCH779284 or 20 μM GW9662. Adiponectin, GLUT4, HCAR2 and GAPDH were detected by corresponding antibodies. g Adiponectin and GLUT4 protein expression of db/db mice adipose tissue. h Schematic diagram of the mechanism by which 3HB ameliorates insulin resistance in type 2 diabetic mice. The results (expressed as the mean ± SD, n = 3) were presented relative to Ctrl groups for each gene, *p < 0.05, and ***p < 0.0001

At the protein expression level, 3HB also promoted the expression of PPARγ regulated downstream genes encoding adiponectin and glucose transporter 4 (GLUT4), while the deficiency or inhibition of HCAR2, ERK1/2 or PPARγ eliminated the effect of 3HB (Fig. 5f). The promotion effect of 3HB on adiponectin and GLUT4 protein expression was also verified in adipose tissue of db/db T2D mice (Fig. 5g).

Therefore, 3HB significantly regulated the expression of PPARγ downstream genes which were affected by Ser273 phosphorylation at both gene transcription and protein translation levels, resulting in reducing of insulin resistance level. This effect was dependent on HCAR2 and ERK1/2, further confirming the above molecular mechanism by which 3HB inhibited PPARγ Ser273 phosphorylation.

留言 (0)

沒有登入
gif