Circulating extracellular vesicle-derived miR-1299 disrupts hepatic glucose homeostasis by targeting the STAT3/FAM3A axis in gestational diabetes mellitus

Abstract

Background: Extracellular vesicles (EVs) are membrane-enclosed structures containing lipids, proteins, and RNAs that play a crucial role in cell-to-cell communication. However, the precise mechanism through which circulating EVs disrupt hepatic glucose homeostasis in gestational diabetes mellitus (GDM) remains unclear.

Results: Circulating EVs isolated from human plasma were co-cultured with mammalian liver cells to investigate the potential induction of hepatic insulin resistance by GDM-EVs using glucose output assays, Seahorse assays, metabolomics, fluxomics, qRT-PCR, bioinformatics analyses, and luciferase assays. Our findings demonstrated that hepatocytes exposed to GDM-EVs exhibited increased gluconeogenesis, attenuated energy metabolism, and upregulated oxidative stress. Particularly noteworthy was the discovery of miR-1299 as the predominant miRNA in GDM-EVs, which directly targeting the 3'-untranslated regions (UTR) of STAT3. Our experiments involving loss- and gain-of-function revealed that miR-1299 inhibits the insulin signaling pathway by regulating the STAT3/FAM3A axis, resulting in increased insulin resistance through the modulation of mitochondrial function and oxidative stress in hepatocytes. Moreover, experiments conducted in vivo on mice inoculated with GDM-EVs confirmed the development of glucose intolerance, insulin resistance, and downregulation of STAT3 and FAM3A.

Conclusions: These results provide insights into the role of miR-1299 derived from circulating GDM-EVs in the progression of insulin resistance in hepatic cells via the STAT3/FAM3A axis and downstream metabolic reprogramming.

Keywords: Extracellular vesicles; Gestational diabetes mellitus; Insulin resistance; MiR-1299/STAT3/FAM3A.

Fig. 1

EVs isolated from the plasma of pregnant women and internalization of EVs by liver cells. (A) Representative transmission electron microscopy images of isolated EVs. EVs displayed a cup-shaped morphology. (B) The distribution of EVs size by NTA analysis. (C) The expressions of EVs markers (CD81, CD63, and TSG101), and cellular markers (Calnexin) were assessed by western blotting. Appearance of PKH67-stained EVs in THLE-2 cells (D), and liver of recipient mice (E) after 24 h administration of EVs. DAPI stained the nucleus blue, and PKH67 stained EVs green (bar, 50 μm)

Fig. 2

Effect of EVs from GDM plasma on hepatic glucose metabolism and energy metabolism. (A) Glucose production and (B) the mRNA levels of G6PC and PCK1 in THLE-2 cells after 24 h incubation with plasma EVs from GDM and normal pregnant women. (C) AKT phosphorylation levels in THLE-2 cells after treatment with GDM-EVs or Normal-EVs. (D) Glucose production and (E) the mRNA levels of G6pc and Pck1 in primary mouse hepatocytes after 24 h incubation with plasma EVs. (F) AKT phosphorylation levels in primary mouse hepatocytes after treatment with GDM-EVs or Normal-EVs. (G) Normalized OCR measurements of cells incubated with GDM-EVs and Normal-EVs. After recording three baseline OCR measurements, oligomycin (Oligo), Carbonyl cyanide 4-(trifluoromethoxy)phenylhydrazone (FCCP), and Rotenone/Antimycin A (Rot/AA) were sequentially added, and OCR measurements were taken thrice after each treatment. After EVs were separately injected into mice via tail vein, OGTT(H) and ITT (I) were assessed in each group at GD 11.5. HOMA-IR (J), and the hepatic mRNA levels of G6PC and PCK1(K) in each group at GD 18.5. Results are means ± SEM (n = 3–7). *p < 0.05, **p < 0.01, ***p < 0.001

Fig. 3

Cellular metabolic profiles after administration of GDM-EVs and Normal-EVs to THLE-2 cells. (A) Partial least square discriminant analysis (PLS-DA) of metabolite patterns in THLE-2 cells incubated with plasma EVs from GDM and normal pregnant women. (B) Heatmap of metabolites with differential (p < 0.05) levels between the GDM-EVs group and the Normal-EVs group. (C) Activities of metabolic pathways in THLE-2 cells incubated with plasma EVs from GDM and normal pregnant women. Black dots represent metabolic activities in the Normal-EVs group that were adjusted to 0. Red dots represent metabolic activities in the GDM-EVs group. The metabolic activities were visualized using a log₂ scale. The dot size indicates the number of metabolites of the pathway, and the dot color indicates the p value. (D) A Circos plot displaying the connectivity between significantly enriched pathways and metabolites. (E) GDM-EVs induced oxidative stress in THLE-2 cells. Results are means ± SEM (n = 5–6). *p < 0.05, **p < 0.01

Fig. 4

¹³C-labelled glucose metabolic flux and intracellular TCA cycle metabolism in EV-treated THLE-2 cells. TCA cycle metabolism showed a lower rate of biochemical conversion after the administration of GDM-EVs. The histograms show a percentage difference of ¹³C-labelling metabolites between the GDM-EVs and the Normal-EVs groups. M⁺ is the main molecular ion of an identified metabolite, + 1 is 1 m/z higher than the M⁺. Results are means ± SEM (n = 5). *p < 0.05, **p < 0.01

Fig. 5

Significant changes in the miRNA profile of circulating EVs in GDM. (A) The differential expression level of miRNAs between GDM-EVs and Normal-EVs. (B) Volcano plot showing four upregulated (red) and eleven downregulated (blue) miRNAs (log₂ fold-change > 1.5). qPCR was used to determine the differential expression of miR-1299 in human (C) and murine (D) GDM-EVs compared to Normal-EVs. Results are means ± SEM (n = 3–5). *p < 0.05. (E) A miRNA-target gene topology network for miR-1299. The grey nodes represent genes, and the red nodes represent miRNAs. Nineteen putative target genes of miR-1299 were highlighted by enlarged blue circles with labelled gene names

Fig. 6

GDM-EVs induced insulin resistance through STAT3/FAM3A. (A) After 24 h exposure of THLE-2 cells to GDM-EVs or Normal-EVs, the levels of miR-1299 in cells were measured by qPCR analysis. The effects of Colivelin on recovery of glucose output (B) and gluconeogenic gene (C) in liver cells with treatment of GDM-EVs. (D) The effects of Colivelin on recovery of STAT3, FAM3A, P-STAT3, and P-AKT expression levels in cells exposed to GDM-EVs. (E) Statistical analysis of western blot results. Representative immunofluorescent staining of STAT3, FAM3A (F), and P-STAT3 (G) localization in the liver from mice injected GDM-EVs or Normal-EVs. (bar, 50 μm). Results are means ± SEM (n = 3–6). *p < 0.05, **p < 0.01, ***p < 0.001

Fig. 7

MiR-1299 impairs insulin sensitivity through STAT3/FAM3A. (A-D) Glucose output and gluconeogenic gene in THLE-2 cells after transfection with miR-1299 mimics or inhibitor. (E) Expression of STAT3, FAM3A, P-STAT3, and P-AKT after overexpressing or inhibiting of miR-1299. (F) MiR-1299 directly targets STAT3. The binding site of the miRNA on its target was predicted by TargetScan. Results are means ± SEM (n = 3–6). *p < 0.05, **p < 0.01, ***p < 0.001. WT = wild type, MUT = mutation

Fig. 8

Schematic representation of the mechanisms of GDM EVs-derived miR-1299 induction of hepatic insulin resistance. In women with GDM, the levels of circulating EVs were elevated and miR-1299 was highly expressed in these EVs. Circulating GDM EVs were absorbed by hepatocytes, leading to an increase in the level of miR-1299 within these cells. Then, miR-1299 binding to the 3ʹUTR of STAT3 mRNA resulted in STAT3 translational arrest which consequently reduced the expression of FAM3A. The decreased levels of STAT3 and FAM3A can potentially affect mitochondrial function by slowing down TCA cycle flux and mitochondrial respiration, while increasing oxidative stress. Therefore, elevated miR-1299 in circulating GDM EVs is proposed to induce hepatic insulin resistance through impaired mitochondrial function via inhibiting the STAT3/FAM3A axis