In the current study, we further identified a regulatory pathway in EAT adipocytes consisting of miR-21-3p, FGFR1, FGF21 and PPARγ that control EAT browning and participates the process of hyperglycemia-induced atrial fibrosis. Modulation of this signaling pathway might provide a therapeutic option for the prevention and treatment of atrial fibrosis in DM.
Numerous evidence has suggested that diabetes is a strong, independent risk factor for AF[7, 27]. The underlying mechanisms by which diabetes increases the susceptibility to AF are unclear but are thought to be associated with electrical and structural remodeling of the atria[28, 29]. Extensive atrial fibrosis resulting from hyperglycemia is thought to play a key role in both initiating and perpetuation AF as an increase of collagen deposition in the atria can cause abnormal conduction of signals, rupture of the propagating waves, and even the occurrence of re-entry[30]. In diabetic conditions, fibroblasts are activated leading to inappropriate collagen production and deposition no matter in type I DM[31, 32] or type II DM[33–35]. Emerging evidence poses a unique challenge to understanding the pathogenesis of atrial fibrillation caused by diabetes, which is often consistent with obesity. Obesity leads to increased thickness of EAT and enhanced invasiveness, which in turn leads to interstitial fibrosis. Recent studies have reported a higher volume or thickness of EAT in patients with AF, particularly in those with non-paroxysmal AF[36–38]. Some studies have reported a relationship between left atrium size and EAT thickness or volume, which might contribute to atrial fibrosis or cardiomyocytes electrophysiological disorders leading to AF[10, 11, 37]. In particular, posterior left atrial adipose tissue is supposed to contribute to the atrial remodeling leading to the onset of AF[39]. Moreover, under diabetic or hyperglycemic conditions, high volume of thickness of EAT is also observed[40–42]. There are several possible mechanisms for the associations between EAT and the increased risks of AF. First, EAT is a rich source of adipokines and cytokines which have pro-fibrotic and pro-inflammatory effects, and the closed proximity of EAT to atrial cardiomyocytes might favor the paracrine activity of EAT secretome, which seems to play a role in the pathogenesis of AF[10]. Second, adipocyte infiltration within atrial cardiomyocytes might lead to the loss of side-to-side cell connection with consequent reduced and heterogeneous voltage[43–45]. Third, the fibrotic remodeling of EAT was also associated with atrial myocardial fibrosis[10].
On the other hand, EAT is also made up of WAT and BAT. BAT is the primary site of non-shivering thermogenesis and is, therefore, a relevant site for adaptive energy expenditure processes. Compared with WAT, BAT improves insulin sensitivity, glucose tolerance, lipid homeostasis, and protects against the pathogenesis of CVD[13, 46]. It has been recently shown that adipose tissues have remarkable plasticity in relation to their contents of white and brown adipocytes. Modulation of the cardiac and vascular adipose tissue to increase the proportion of thermogenic brown or beige adipocytes might be a viable way to improve local inflammation and reduce cardiovascular risk[13, 46]. However, under diabetic or hyperglycemic conditions, impaired WAT browning potential is observed[14], which might aggregate the pathogenesis of AF. Consistent with previous studies, our experiments indicated that hyperglycemia inhibited the biomarkers of BAT in mouse EAT as well as in vitro cultured adipocytes, suggesting that hyperglycemia might decrease the process of EAT browning.
To date, the mechanisms of EAT browning and its role in DM-induced atrial fibrosis remain to be elucidated. FGF21 has been shown to have a beneficial effect on metabolism and energy balance by enhancing fatty acid-oxidation during prolonged fasting and also by promoting WAT browning[47–50]. And a recent clinical study showed that cold exposure increased circulating levels of the fat browning activators FGF21 and irisin and that treatment with either of these endocrine regulators up-regulated browning genes and promoted thermogenesis[51]. Mechanistically, FGF21 activates cell signaling by binding to a heteromeric cell-surface receptor tyrosine kinase complex composed of β-Klotho and FGFR1[49]. Both β-Klotho and FGFR1 are abundantly expressed in WAT, where FGF21-regulated genes are involved in a variety of metabolic processes including lipogenesis, lipolysis, fatty acid oxidation, and WAT browning[49]. Furthermore, PPARγ, a member of the nuclear receptor family of ligand-activated transcription factors, is also required for adipocyte differentiation. PPARγ agonist has been shown to induce browning of the EAT that probably contributes to the increase in lipid turnover[52]. And FGF21 was thought to be a key mediator of the physiologic and pharmacologic actions of PPARγ in WAT[49]. FGF21 stimulates PPARγ transcriptional activity and FGF21 deficiency mice have decreased PPARγ activity in WAT and corresponding reductions in WAT mass and adipocyte size[49]. On the other hand, FGF21 was previously shown to be induced by PPARγ agonists in WAT and to cooperate with rosiglitazone in promoting differentiation in 3T3-L1 adipocytes[53, 54]. And obese, insulin-resistant mice lacking FGF21 are refractory to the actions of rosiglitazone, including both beneficial and adverse effects[49]. Therefore, we conclude that the actions of FGF21 and PPARγ are fundamentally intertwined, and propose a feed-forward regulatory model in WAT[55]. And FGFR1/FGF21/PPARγ might be an important signal pathway to precipitate the browning of WAT. In our mouse diabetic model as well as in vitro cellular model, we found that FGFR1/FGF21/PPARγ pathway was inhibited in EAT or under hyperglycemia conditions. Consistently, FGF21 was also shown to protect the blood-brain barrier through FGFR1/FGF21/PPARγ activation, which up-regulated tight junction and adhesion junction proteins[56].
Recent studies indicated that miRs function as important regulators that participates in DM-induced atrial fibrosis[35]. The role of miR-21 in the pathogenesis of atrial fibrosis has been illuminated, and increased miR-21 expression was correlated positively with atrial fibrosis or fibrotic gene expression[20–22]. Many target genes of miR-21 have been found to play a large role in DM-induced atrial fibrosis through different biological pathways. Tao H et al. found that miR-21 regulated atrial fibrosis via dysregulation of WW Domain-Containing Protein 1[57]. Cao W et al demonstrated that the tumor suppressor cell adhesion molecule 1 was the potential target of miR-21, and miR-21 promoted cardiac fibrosis via STAT3 signaling pathway by decrease CADM1 expression[58]. In our mice diabetic model, we also found that miR-21 KO manifested a decreased atrial fibrosis as well as fibrotic gene expression. However, miR-21 has several subtypes, such as miR-21-3p and miR-21-5p in human, and miR-21-3p and miR-21-5p in mice. We further investigated the expression of different miR-21 subtypes in various clinic conditions, including healthy control, DM, AF, DM combined AF. And the results indicated that hsa-miR-21-3p, instead of hsa-miR-21-5p, was obviously increased in patients with DM and/or AF, suggesting that hsa-miR-21-3p is more likely to be involved in the regulation of DM-induced atrial fibrosis. Besides, emerging evidence indicates that miRs also function as important regulators in brown remodeling of adipocytes. Muscle-enriched miR-133a directly down-regulated expression of the key transcriptional activator of brown fat differentiation, positive regulatory domain containing 16 (PRDM16), and cold exposure decreased miR-133a levels and promoted brown fat cell differentiation[59, 60]. Brown adipocyte-enriched miR-155 was also shown to inhibit brown fat cell differentiation by directly targeting the browning transcription factor C/EBP[61]. But whether miR-21-3p played a role in the browning of EAT under diabetes and thus affected atrial fibrosis has not been studied. In the present study, we further identify miR-21-3p as a key regulator that controls EAT browning in hyperglycemia condition. Further, we predicted target genes by Targetscan and found browning transcription factor FGFR1 as its potential target. Besides, by virtue of in vitro co-culture model, miR-21-3p regulated fibrotic gene expression in atrial fibroblasts through affecting the adipocytes browning.
A paracrine effect of EAT on the neighboring myocardium has been proposed[10]. EAT produces a number of inflammatory mediators and adipocytokines that can modulate the functional and structural properties of the myocardium[62, 63]. In this line, it has been shown that the secretome of EAT can induce atrial fibrosis, an important determinant of the substrate of AF. In the present study, we observed that miR-21-3p mimics for adipocytes could increase the levels of inflammatory factors in co-culture model in hyperglycemia conditions.
Our current study has some limitations that deserve to be mentioned. First, mouse adipocyte-specific miR-21-3p KO was not performed in this study. Second, we did not conduct in vivo intervention experiments to study the correlation between miR-21-3p and EAT browning. Third, we only determined the extent of atrial fibrosis, however, we did not perform the experiments of cardiac electrophysiology and programmed stimulation, such as inducibility of AF.