Sex-specific cardioprotective role of miR-30a-5p through estrogen-dependent mechanisms in a mouse model of heart failure

Abstract

Background

In recent years, except for the well-known heart failure with reduced ejection fraction (HFrEF), the incidence of heart failure with preserved ejection fraction (HFpEF) and heart failure with mildly reduced ejection fraction (HFmrEF) among the classification of heart failure (HF) has been increasing. However, due to their complex mechanisms, current research remains insufficient to address clinical needs.

Methods and results

Utilizing wild-type (WT), miR-30a-5p knockout (KO), and overexpression (OE) murine models combined with estrogen modulation and ovariectomy (OVX), this study delineates sex-specific regulatory networks in HF pathogenesis. Female KO mice lost the inherent resistance of WT females to HFpEF induction via 24-week HFD/L-NAME, whereas males exhibited comparable HFpEF susceptibility regardless of genotype, developing hallmark phenotypes including diastolic dysfunction (E/E′), myocardial hypertrophy (heart weight/tibia length), cardiac fibrosis, and hepatic steatosis. Particularly, due to the reduced ejection fraction in KO mice, combined with HFD/L-NAME, the HF phenotype was ultimately manifested as impaired diastolic function and slightly reduced ejection fraction (with the characteristics of HFpEF and HFmrEF). Mechanistically, KO-HF females displayed significant estrogen axis disruption (plasma estradiol and the expression of ERα, ERβ, ESRRA, and PELP1 expression). OVX in WT females validated the importance of estrogen for HFpEF resistance. Transcriptomic profiling identified convergent targets across cardiac (ITGAD, ITGAM, FGA, and FGB) and hepatic tissues (APOA1 and APOB), revealing miR-30a-5p’s orchestration of extracellular matrix remodeling (via ITGAD/ITGAM mechanotransduction),fibrinogen-mediated microvascular homeostasis, and APOB-driven metabolic dysregulation. Notably, OE intervention failed to mitigate OVX-induced cardiac/hepatic pathology, implicating estrogen-dependent miR-30a-5p functionality.

Conclusions

These findings establish miR-30a-5p as a crucial sex-specific regulator of HF (mainly HFpEF), operating through estrogen signaling to balance cardiac compliance and metabolic adaptation.

Graphical abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s12933-026-03090-7.

Background

Based on ejection fraction (EF), heart failure (HF) can be classified into three subtypes: heart failure with reduced ejection fraction (HFrEF; LVEF ≤ 40%), heart failure with mildly reduced ejection fraction (HFmrEF; LVEF 41% to 49%), and heart failure with preserved ejection fraction (HFpEF; LVEF ≥ 50%) [1]. Essentially, HFmrEF represents an intermediate clinical entity between HFrEF and HFpEF. Both HFpEF and HFmrEF are considered milder forms of heart failure, associated with a lower risk of cardiovascular and HF-related events but with a similar or greater risk of non-cardiovascular adverse events [2].

HFpEF constitutes nearly 50% of all heart failure (HF) cases, with its incidence, prevalence, and mortality rates continuing to increase globally [3]. However, the poorly understood pathophysiology of HFpEF limits the efficacy of therapies developed for HFrEF, resulting in suboptimal clinical outcomes [4]. Therefore, elucidating the pathophysiology of HFpEF and identifying novel therapeutic targets have become critical priorities in cardiovascular research. Current evidence characterizes HFpEF as a multisystem disorder involving the heart, lungs, kidneys, adipose tissue, immune system, and vasculature, clinically manifested by hypertension, cardiac hypertrophy, myocardial fibrosis, and diastolic dysfunction [5]. Furthermore, emerging studies highlight a bidirectional relationship between HFpEF and non-alcoholic fatty liver disease (NAFLD), with proposed phenotypic classifications including obstructive HFpEF, metabolic HFpEF, and HFpEF with advanced liver disease/cirrhosis [6]. These findings advance our understanding of HFpEF heterogeneity and lay the groundwork for developing phenotype-specific therapeutic strategies.

Epidemiological studies reveal that the incidence of HFpEF rises markedly after age 55, with a higher prevalence in females [7]. This sex disparity suggests that estrogen imbalance may contribute to age-related comorbidities such as obesity, hypertension, and diabetes—key accelerators of HFpEF pathogenesis [7]. Estrogen exerts cardioprotective effects, including ameliorating left ventricular diastolic dysfunction and enhancing endothelial function [8]. Clinically, women exhibit higher HFpEF prevalence and more severe symptoms than men, driving intense research into sex-specific mechanisms and therapeutic outcomes [9].

Estrogen modulates sex-biased microRNAs (miRNAs), many of which exhibit cardioprotective properties, potentially explaining sexual dimorphism in HFpEF and other cardiovascular diseases [10]. miR-30a-5p is a conserved member of the miR-30 family implicated in regulating cellular processes such as fibrosis, apoptosis, and autophagy by targeting genes like Sirt1 and TP53INP1 [11–13]. Its downregulation has been reported in models of atherosclerosis, myocardial infarction, and cardiac hypertrophy, where it facilitates macrophage M1/M2 polarization, pathological remodeling and fibrosis [14, 15]. Furthermore, emerging evidence suggests a potential interaction between miR-30a and estrogen signaling, hinting at a role in sex-specific differences in cardiovascular disease [16]. However, despite its recognized importance in general cardiac pathology, the specific expression pattern, functional role, and regulatory mechanisms of miR-30a-5p in the distinct context of HFpEF remain completely unknown. This critical gap in knowledge motivated our present study to investigate whether miR-30a-5p serves as a key modulator in HFpEF pathogenesis.

We established an HFpEF murine model using a multi-hit strategy combining high-fat diet and nitric oxide synthase inhibition (Nω-nitro-L-arginine methyl ester, L-NAME), which is accepted in the field to recapitulate the systemic metabolic and cardiovascular inflammatory stressors central to human HFpEF pathophysiology, particularly the obese/metabolic phenotype [17]. This combination effectively mimics key clinical features of HFpEF, including preserved ejection fraction with diastolic dysfunction, endothelial dysfunction, cardiac hypertrophy, myocardial fibrosis, and systemic metabolic disturbances. Striking sex differences emerged in both disease progression and miR-30a-5p knockout (KO) responses, suggesting estrogen-mediated regulation. This study aims to elucidate miR-30a-5p’s mechanistic role in HF sexual dimorphism, offering novel diagnostic and therapeutic targets for personalized HF management.

Methods

Animals

We applied CRISPR/Cas9 technology to construct miR-30a-5p^−/− (KO) and wild-type (WT) mice with the C57BL/6J genetic background, as previously described [13, 15, 18, 19]. Genotypes were verified by PCR using the following primers: 5′-AGCTTCCCTACTTTGGTGTTT-3′; 5′- TGGTGTGTGTGAATTGACCT − 3′. All mice were housed in a specific pathogen-free facility at the Animal Experimental Center of Xiamen University under a 12 h light/dark cycle, 22 °C ± 2 °C, and 50–70% humidity. All animal studies complied with the guidelines of the Institutional Animal Care and Use Committee of Xiamen University and were approved by the Laboratory Animal Management and Ethics Committee of Xiamen University (Approval NO. XMULAC20190120).

Animal models and groups

Mice were randomly assigned to experimental groups. No animals were excluded from the analysis post-allocation. All animals that entered the study completed the protocol, and their data are included in the results.

To assess sex-specific susceptibility to HFpEF, KO and WT mice (6-8-week-old males and females, n = 8/group) were fed a high-fat diet (HFD, 60 kcal% fat, D12492, Research Diets) and provided drinking water containing N^ω-nitro-L-arginine methyl ester hydrochloride (L-NAME, 0.5 g/L, N5751, Sigma-Aldrich) for 24 weeks to induce HFpEF [20]. Control groups received a normal diet (NC, 10 kcal% fat, D12450J, Research Diets) and plain water. After 24 weeks, cardiac systolic/diastolic function, blood pressure, exercise endurance, and other parameters were measured.

To investigate estrogen’s protective role, WT female mice (6-8-week-old female mice, n = 8/group) underwent ovariectomy (OVX) or sham surgery [21, 22]. After anesthesia using 1.5% isoflurane (R510-22-10, REAWORLD, Shenzhen, China), the female mice were exposed to identify bilateral ovaries, which were then excised to complete OVX. OVX modeling was completed when the mice woke up. One-week post-operation, both groups received HFD and L-NAME for 24 weeks. Follow-up experiments were conducted after model establishment.

To explore miR-30a-5p–estrogen interactions, OVX WT female mice (6–8 weeks old, n = 8/group) received tail vein injections of 100 µL AAV2/9-cTNT-mmu-mir-30a-Null-LUC (AAV-OE, 1.7 × 10¹² vg/mL, HH20210818FJCJY-AAV02, HANBIO) or AAV2/9-cTNT-EGFP (AAV-NC, 1.4 × 10¹² vg/mL, HH20210818FJCJY-AAV02, HANBIO) at weeks 1 and 12 of HFpEF modeling. Phenotypic analyses were performed after 24 weeks.

To confirm miR-30a-5p’s role in HFpEF, female KO mice (6–8 weeks old, n = 8/group) on HFD/L-NAME were injected with AAV-OE or AAV-NC at weeks 1 and 12. Cardiac and hepatic phenotypes were evaluated post-modeling. When conducting experiments such as echocardiography and pathological staining, the laboratory technicians adopted the blind method to the group identities throughout the process.

Histological analysis

Liver and heart samples were embedded in an Optimal Cutting Temperature (OCT) compound (Cat# G6059, Servicebio, Wuhan, China) and stored at – 80 ℃. Section (6 μm thickness) were prepared for staining. For hematoxylin-eosin (H&E) staining, we used an H&E kit (Cat# G1120, Solarbio Life Science, Beijing, China) to observe the pathological morphology following the manufacturer’s protocol. For Masson trichrome staining, a Masson staining kit (Cat#: G1340, Solarbio Life Science, Beijing, China) was used to differentiated collagen and muscle fibers. For Sirius red staining, Sirius red solution (Cat# GC307014, Servicebio, Wuhan, China) was used to stain collagen fibers with red, while other components appeared yellow. Images were acquired using a TissueFAXS scanning system (TissueGnostics, Austria).

Immunofluorescence and immunohistochemistry

Sections were dewaxed, underwent antigen retrieval, and blocked in 3% BSA. For immunofluorescence, sections were incubated with primary antibodies, including APOA1 (Cat#: A24291, Abclonal, Wuhan, China), APOA2 (Cat#: A3321, Abclonal, Wuhan, China), APOB (Cat#: A25420, Abclonal, Wuhan, China), FGB (Cat#: A1401, Abclonal, Wuhan, China), ESRRA (Cat#: A14184, Abclonal, Wuhan, China), and PELP1 (Cat#: A3189, Abclonal, Wuhan, China) respectively. The corresponding fluorescent secondary antibodies, including Alexa Fluor 488 labeled goat anti-rabbit IgG (Cat#: GB25303, Servicebio, Wuhan, China) and CY5 labeled goat anti-rabbit IgG (Cat#: GB27303, Servicebio, Wuhan, China) were applied to mark the target protein. The nucleus was re-stained with DAPI (C0065, Solarbio, Beijing, China). The results were obtained using a fluorescence microscope (Leica, Germany). For immunohistochemistry, sections were incubated in the primary antibodies, including ESRRA (Cat#: A14184, Abclonal, Wuhan, China), PELP1 (Cat#: A3189, Abclonal, Wuhan, China), FGA (Cat#: A1453, Abclonal, Wuhan, China), ITGAM (Cat#: A23254, Abclonal, Wuhan, China), ITGAL (Cat#: A23960, Abclonal, Wuhan, China), ITGAD (Cat#: A4054, Abclonal, Wuhan, China), ITGB3 (Cat#: A19073, Abclonal, Wuhan, China), Estrogen Receptor alpha (Cat#: GB111843, Servicebio, Wuhan, China), and Estrogen Receptor beta (Cat#: GB11268, Servicebio, Wuhan, China). DAB solution (SW1020, Solarbio, Beijing, China) and hematoxylin were finally used. Images were obtained under a microscope (Leica, Germany). The positive area was brownish yellow, and the nucleus stained with hematoxylin was blue. For sections of liver and heart, the images were randomly obtained after immunofluorescence and immunohistochemical staining. Notably, the heart sections were selected from the left ventricle.

Serum indicator detection

Murine blood samples were centrifuged at 3, 000 rpm for 10 min to isolate serum. Biochemical parameters were analyzed using a Beckman Coulter AU biochemical analysis system (Beckman Coulter, CA, USA). The reagents involved in this study included Lactate dehydrogenase (LDH, Cat#: AUZ0376, Beckman Coulter, Brea, CA, USA), creatine kinase (CK, Cat#: AUZ0451, Beckman Coulter, Brea, CA, USA), CK-MB (Cat#: OSR6179, Beckman Coulter, Brea, CA, USA), cardiac troponin T (cTnT, Cat#: 64100801, Guangzhou Hongen Medical diagnosis, Guangzhou, China), total cholesterol (Cat#: AUZ0167, Beckman Coulter, Brea, CA, USA), triglyceride (Cat#: AUZ0161, Beckman Coulter, Brea, CA, USA), high-density lipoprotein-cholesterol (HDL-c, Cat#: AUZ0195, Beckman Coulter, Brea, CA, USA ), low-density lipoprotein-cholesterol (LDL-c, Cat#: AUZ0386, Beckman Coulter, Brea, CA, USA), glucose (GLU, Cat#: AUZ0171, Beckman Coulter, Brea, CA, USA), alanine aminotransferase (ALT, Cat#: AUZ0528, Beckman Coulter, Brea, CA, USA), aspartate aminotransferase (AST, Cat#: AUZ0263, Beckman Coulter, Brea, CA, USA), cholinesterase (Cat#: AUZ0163, Beckman Coulter, Brea, CA, USA), and total bile acids (Cat#: AUZ0266, Beckman Coulter, Brea, CA, USA).

Enzyme-linked immunosorbent assay (ELISA)

Mouse serum samples or heart tissue homogenates from each group were incubated with horseradish peroxidase (HRP)-conjugated antibodies in 96-well plates at 37 °C for 60 min. After five washes with buffer, substrate solution was added and plates were incubated in the dark at 37 °C for 15 min. The reaction was stopped with stop solution, and absorbance (OD) was measured at 450 nm. Sample concentrations were calculated using standard curves. The following commercial ELISA kits were used: estradiol (E2, Cat# ER13121), estrogen receptor (ER, Cat# ER13120), testosterone (Cat# ER13034) (from Zancheng Technology, Tianjin, China), and insulin (D721159-0048, BBI). The homeostasis model assessment of insulin resistance (HOMA-IR) index was calculated as follows: HOMA-IR = [Fasting blood glucose (mmol/L)×Fasting insulin (µU/mL)] / 22.5.

RNA sequencing

RNA sequencing and analysis were performed by Gene Denovo Co., Ltd (Guangzhou, China; Order No. GHRL23080141_std_1). Total RNA was extracted from cardiac tissue using TRIzol reagent. RNA integrity was assessed by RNA integrity number (RIN) using an Agilent 2100 Bioanalyzer with the RNA 6000 Nano Kit (Agilent Technologies, CA, USA), and quantified using the Qubit RNA HS Assay Kit (Thermo Fisher Scientific, MA, USA). RNA-seq libraries were prepared using the Hieff NGS Ultima Dual-mode mRNA Library Prep Kit (Yeasen, Shanghai, China), with quality control performed using the Agilent High Sensitivity DNA Kit (5067 − 4626). Libraries were amplified by PCR and sequenced on an Illumina NovaSeq X Plus platform (Illumina, CA, USA). Bioinformatics analyses included Gene Ontology (GO) and Kyoto encyclopedia of Genes and genomes (KEGG) pathway enrichment.

RNA immunoprecipitation (RIP) analysis

HEK-293T cells were transfected with either AAV-miR-30a-5p OE (Lot# 34082017, HANBIO) or AAV-NC (Lot# 34082016, HANBIO). Cell lysates were incubated with magnetic beads conjugated with anti-Argonaute2 antibody, followed by RNA extraction from the immunoprecipitated complexes (Millipore, Billerica, MA, USA). And qRT-PCR analysis was used for determining the relative expression of miR-30a-5p and Itgad, Itgam, Fga, Fgb, Apoa1, and ApoB. IgG was used as the control group.

Statistical analysis

Data are presented as mean ± standard deviations (mean ± SD). Statistical analyses were performed using GraphPad Prism 9. Two-group comparisons of normally distributed data used Student’s t-test. For comparisons among three or more groups, one-way ANOVA was performed, followed by Tukey’s or Dunnett’s post hoc tests. Statistical significance was defined as p < 0.05, with p-values versus controls indicated in the corresponding figures.

Results

miR-30a-5p knockout abolished sexual dimorphism in HF phenotype development observed in wild-type mice

We established HFpEF models in both wild-type (WT) and miR-30a-5p knockout (KO) mice (6–8 weeks old) through 24 weeks of high-fat diet combined with L-NAME administration (Fig. 1A and D), with control groups receiving standard diet (genotype validation shown in Supplementary Fig. 1). Results demonstrated that female KO + HF mice exhibited significantly increased lung wet/dry weight ratio and body weight/tibia length compared to other groups (Fig. 1B), while WT + HF females only showed mild weight gain without pulmonary congestion. Blood pressure measurements revealed significantly elevated systolic (SBP) and diastolic blood pressure (DBP) in KO females versus WT females treated with either L-NAME or HFD + L-NAME (Fig. 1C, Supplementary Fig. 2). Cardiac function assessment showed markedly reduced ejection fraction (Fig. 1G, H) and impaired diastolic function (Fig. 1I, J, K), with exercise tolerance tests demonstrating a reduction in running distance and time (Fig. 1Q) in KO + HF females. In contrast, male WT + HF and KO + HF groups developed comparable pulmonary congestion, weight gain (Fig. 1E) and hypertension (Fig. 1F), with similar diastolic dysfunction though additional systolic impairment was observed in KO controls (Fig. 1L-P), and no intergroup differences in exercise capacity (Fig. 1R). These findings indicate that after HFD and L-NAME treatment only WT male mice developed HFpEF phenotypes (EF ≥ 50%), and miR-30a-5p KO females and males exhibited HFmrEF (40% ≤ EF ≤ 49%) and HFpEF characteristics with impaired diastolic function and slightly reduced ejection fraction. We also provided a table (Supplementary Table 1) summarizing HF phenotypes of all experimental groups. WT females maintained normal phenotypes through enhanced diastolic function and compensatory capacity under identical modeling conditions, suggesting an estrogen-dependent cardioprotective mechanism mediated by miR-30a-5p (Tables 1 and 2).

Fig. 1.

Female KO mice aggravated the phenotype of HF. A Experimental design for female cohorts: 6-8-week-old WT or KO mice received either the normal diet (ND) or HFD + L-NAME for 24 weeks (n = 8). B Lung wet/dry ratio and body weight/tibia length (BW/TL) ratio in the WT + NC, KO + NC, WT + HF and KO + HF groups of female mice (n = 8). C Systolic blood pressure (SBP) and diastolic blood pressure (DBP) were detected in the WT + NC, KO + NC, WT + HF and KO + HF groups of female mice (n = 8). D Experimental design for female cohorts: 6-8-week-old WT or KO mice received either ND or HFD + L-NAME for 24 weeks (n = 8). E Lung wet/dry ratio and body weight/tibia length (BW/TL) ratio in the WT + NC, KO + NC, WT + HF and KO + HF groups of male mice (n = 8). F Systolic blood pressure (SBP) and diastolic blood pressure (DBP) were detected in the WT + NC, KO + NC, WT + HF and KO + HF groups of male mice (n = 8). G–H Representative left ventricular M-mode echocardiographic tracings in long-axis view and left ventricular ejection fraction (LVEF) and left ventricular fractional shortening (LVFS) analysis in the WT + NC, KO + NC, WT + HF and KO + HF groups of female mice (n = 8). I–K Representative pulsed-wave Doppler tracings of PW-mode and Tissue-mode, and analysis of E/E′ ratio and E/A ratio in the WT + NC, KO + NC, WT + HF and KO + HF groups of female mice (n = 8). L–M, representative left ventricular M-mode echocardiographic tracings in long-axis view and left ventricular ejection fraction (LVEF) and left ventricular fractional shortening (LVFS) analysis in the WT + NC, KO + NC, WT + HF and KO + HF groups of male mice (n = 8). N–P Representative pulsed-wave Doppler tracings of PW-mode and Tissue-mode, and analysis of E/E′ ratio and E/A ratio in the WT + NC, KO + NC, WT + HF and KO + HF groups of male mice (n = 8). Q Running distance and running time from exercise exhaustion tests in the WT + NC, KO + NC, WT + HF, and KO + HF groups of female mice (n = 8). R Running distance and running time from exercise exhaustion tests of the WT + NC, KO + NC, WT + HF and KO + HF groups of male mice (n = 8). Data are expressed as mean ± SD and one-way ANOVA with Tukey’s test was used for multiple comparisons; ns indicated no significant difference, ^*P < 0.05, ^**P < 0.01, ^***P < 0.001, and ^****P < 0.0001

Table 1.

Summary of HF phenotypes in experimental group

	Group	EF	E/E’	Blood pressure	Exercise capacity	HF category
♂	WT + NC	≥ 50%	Normal	Normal	Normal	–
	KO + NC	40–49%	Mild increase	Normal	Normal	–
	WT + HF	≥ 50%	Increase	Increase	Decrease	HFpEF
	KO + HF	40–49%	Increase	Increase	Decrease	HFmrEF/ HFpEF
♀	WT + NC	≥ 50%	Normal	Normal	Normal	–
	KO + NC	≥ 50%	Normal	Normal	Normal	–
	WT + HF	≥ 50%	Normal	Mild increase	Mild decrease	–
	KO + HF	40–49%	Increase	Increase	Decrease	HFmrEF/ HFpEF

Table 2.

Expression of miR-30a-5p in the experimental group

Sex	Group	miR-30a-5p expression (fold of WT + NC ♂)
♂	WT + NC	1‌ ± 0.075
	KO + NC	0.109‌ ± 0.017
	WT + HF	0.778 ‌± 0.042
	KO + HF	0.076‌ ± 0.002
♀	WT + NC	1.206 ‌± 0.122
	KO + NC	0.094‌ ± 0.016
	WT + HF	‌0.941 ± 0.033
	KO + HF	‌0.065 ± 0.003
♀	HF	‌0.937 ± 0.094
	OVX + HF	‌0.882 ± 0.077
♀	AAV-NC-KO + HF	‌0.093 ± 0.007
	AAV-OE-KO + HF	‌0.471 ± 0.035
♀	AAV-NC-OVX + HF	‌0.855 ± 0.063
	AAV-OE-OVX + HF	‌2.728 ± 0.112

Comprehensive analysis of cardiac structure and function revealed striking sexual dimorphism in response to HF induction. Female KO + HF mice exhibited significant cardiac hypertrophy (Fig. 2A, C), with larger cardiomyocytes (WGA staining, Fig. 2E, F) and greater fibrosis (Masson/Sirius Red, Fig. 2I, J) compared to WT + HF females, while male WT + HF and KO + HF groups showed comparable cardiac remodeling (Fig. 2B, D, G, H, K, L). This phenotypic divergence was mirrored in metabolic parameters, where female KO + HF mice developed severe dyslipidemia (LDL-c) and liver steatosis (Oil Red O staining) compared to their WT + HF counterparts (Supplementary Fig. 3A, C, D), whereas males exhibited similar metabolic dysfunction regardless of genotype (Supplementary Fig. 3B, E, F). Serum biomarkers confirmed these findings, with female KO + HF mice showing higher cardiac troponin T (cTnT) and elevated liver enzymes versus WT + HF (Fig. 2M, Supplementary Fig. 3G), while male groups displayed equivalent biomarker elevations (Fig. 2N, Supplementary Fig. 3H). In addition, we found that fasting blood glucose, fasting insulin, and the homeostasis model assessment of insulin resistance (HOMA-IR) index were significantly elevated in female KO + HF mice relative to all other groups (Supplementary Fig. 4), thereby demonstrating systemic insulin resistance in our female HFpEF model.

Fig. 2.

Female KO mice promoted the development of heart injury and myocardial hypertrophy. A Representative images of hematoxylin and eosin (H&E) staining in longitudinal heart sections in female mice. Scale bars = 2 mm. B Representative images of H&E staining in longitudinal heart sections in male mice. Scale bars = 2 mm. C Ratio of heart/body weight (HW/BW), and ratio of heart weight/tibia length (HW/TL) in female mice (n = 8). D Ratio of heart/body weight (HW/BW), and ratio between heart weight/tibia length (HW/TL) in male mice (n = 8). E–F Quantification and representative images of wheat germ agglutinin (WGA) staining in female mice (n = 6). Scale bar = 50 μm. G–H Quantification and representative images of cardiomyocyte cross-sectional area based on WGA staining in male mice (n = 6). Scale bars = 50 μm. I–J Representative images and quantification of Masson’s Trichrome and Sirius red staining of the left ventricle section in female mice (n = 6). Scale bars = 50 μm. K–L Representative images and quantification of Masson’s Trichrome and Sirius red staining of the left ventricle section in male mice (n = 6). Scale bars = 50 μm. M The levels of serum lactate dehydrogenase (LDH), creatine kinase (CK), CK-MB and cardiac troponin T (CTNT) in female mice (n = 8). N The levels of serum LDH, CK, CK-MB, and CTNT in male mice (n = 8). Data are expressed as mean ± SD and one-way ANOVA with Tukey’s test was used for multiple comparisons; ns indicated no significant difference, ^*P < 0.05, ^**P < 0.01, ^***P < 0.001, and ^****P < 0.0001

In summary, these results demonstrate that intact miR-30a-5p signaling confers protection specifically in female mice, maintaining cardiac and metabolic homeostasis likely through estrogen-dependent mechanisms, as WT females resisted HF development despite identical HFD/L-NAME challenges that uniformly induced pathology in males.

Estrogen signaling disruption mediates sex-specific susceptibility to HF in miR-30a-5p knockout mice

Hormonal analyses revealed profound disruptions in estrogen signaling specifically underlying the differential HF susceptibility in female mice. KO + HF females exhibited lower serum estradiol levels compared to WT + HF counterparts (Supplementary Fig. 5A-B), with reciprocal testosterone elevation, mirroring their severe HF phenotype, whereas there was no difference between the WT + HF and KO + HF of males (Supplementary Fig. 5C-D).

However, short-term measurements may not fully capture dynamic hormonal changes. In order to make up for the deficiency, we further analyzed the expression levels of other estrogen-related factors through immunohistochemistry and immunofluorescence. Immunohistochemical analysis demonstrated nuclear-localized ERα and ERβ expression was diminished in KO + HF cardiac tissue versus WT + HF (Fig. 3A–D), while the estrogen-related factors ESRRA and PELP1 showed inverse regulation-a decrease in nuclear/cytoplasmic ESRRA but PELP1 accumulation (Fig. 3E–H). This dysregulated pattern was conserved in hepatic tissue, with immunofluorescence revealing reduced ESRRA but elevated PELP1 in KO + HF livers (Fig. 3I–H). These results establish that miR-30a-5p deficiency disrupts multiple tiers of estrogen signaling-from circulating hormone levels to receptor expression and downstream nuclear co-regulators-creating a permissive environment for HF pathogenesis specifically in females. The preserved estrogen signaling in WT females (despite HFD/L-NAME stress) correlates with their resistance to cardiac remodeling, while identical metabolic challenges uniformly induced pathology in males regardless of estrogen status, highlighting the crucial sex-specific protection conferred by intact miR-30a-5p/estrogen axis functionality.

Fig. 3.

The levels of estrogen-related factors in circulation, heart, and liver tissues of female mice. Representative images and quantitative results of immunohistochemistry staining of estrogen receptor alpha (ERα) (A–B), estrogen receptor beta (ERβ) (C–D), estrogen-related receptor-α (ESRRA) (E–F), and Proline-, glutamic acid-, and leucine-rich protein 1 (PELP1) (G–H), in heart tissues of the WT + NC, KO + NC, WT + HF and KO + HF groups of female mice (n = 6). Scale bars = 50 μm. I–J Representative images and quantitative results of immunofluorescence of ESRRA in liver tissues of the WT + NC, KO + NC, WT + HF and KO + HF groups of female mice (n = 6), Scale bars = 50 μm. K–L Representative images and quantitative results of immunofluorescence of PELP1 in liver tissues of the WT + NC, KO + NC, WT + HF and KO + HF groups of female mice (n = 6). Scale bars = 50 μm. Data are expressed as mean ± SD and one-way ANOVA with Tukey’s test was used for multiple comparisons; ns indicated no significant difference, ^*P < 0.05, ^**P < 0.01, ^***P < 0.001, and ^****P < 0.0001

OVX significantly reduced female mice’s resistance to HF induction

To further investigate estrogen’s role in HFpEF pathogenesis, we compared sham-operated (HF) and ovariectomized (OVX + HF) female mice subjected to HFD/L-NAME treatment (Fig. 4A). OVX + HF mice exhibited significantly increased BW/TL ratios, lung wet/dry weights, and cardiac hypertrophy indices (HW/BW, HW/TL) compared to HF controls (Fig. 4B), along with elevated systolic and diastolic blood pressures (Fig. 4C) and markedly reduced exercise capacity (Fig. 4D). Comprehensive cardiac analysis revealed: higher cardiac injury markers (LDH, CK-MB, cTnT; Fig. 4E), increased fibrosis (Masson/Sirius Red; Fig. 4F, G), and larger cardiomyocytes (WGA staining; Fig. 4H, I). While ejection fraction remained unchanged, OVX + HF mice showed significant diastolic dysfunction with higher E/E’ ratios (Fig. 4L, M). Concurrently, OVX induced severe hepatic metabolic disturbances including: more lipid droplets (H&E/Oil Red O; Fig. 4J, K), elevated liver enzymes (ALT/AST; Fig. 4N), and higher LDL-c levels (Fig. 4O). These findings demonstrate that estrogen deficiency exacerbates all key HFpEF pathological features-from cardiac remodeling and diastolic impairment to metabolic dysfunction—establishing estrogen’s crucial protective role in mitigating disease progression in female mice. The mirrored phenotypes between OVX-induced and genetic (miR-30a-5p KO) estrogen deficiency models confirm the central role of intact estrogen signaling in female-specific HFpEF resistance.

Fig. 4.

OVX in mice with HFD and L-NAME treatment was helpful to develop HF. A WT female mice (6–8 weeks) were treated with HFD and L-NAME for 24 weeks after ovariectomy (OVX) or sham operation. B Lung weight (wet/dry), BW/TL, HW/BW, and HW/TL in the HF and OVX + HF groups of female mice (n = 8). C SBP and DBP were detected in the HF and OVX + HF groups of female mice (n = 8). D Running distance and running time from exercise exhaustion tests in HF and OVX + HF groups of female mice (n = 8). E The levels of serum LDH, CK, CK-MB, and CTNT in HF and OVX + HF groups of female mice (n = 8). F–G Representative images and quantification of Masson’s Trichrome and Sirius red staining of the left ventricle section in the HF and OVX + HF groups of female mice (n = 6). Scale bars = 50 μm. H Representative images of WGA staining of heart tissues in female mice. Scale bars = 50 μm. I Quantification of cardiomyocyte cross-sectional area based on WGA staining in HF and OVX + HF groups of female mice (n = 6). J–K Representative images and quantification of H&E and oil red O of the liver in the HF and OVX + HF groups of female mice (n = 6). Scale bars = 50 μm. L Representative images of pulsed-wave Doppler tracings of PW-mode and Tissue-mode in HF and OVX + HF groups of female mice. M LVEF, and E/E′ were measured in the HF and OVX + HF groups of female mice (n = 8). N The levels of serum ALT and AST in HF and OVX + HF groups of female mice (n = 8). O The levels of serum total cholesterol, triglyceride, HDL-c, and LDL-c in the HF and OVX + HF groups of female mice (n = 8). Data are expressed as mean ± SD and Student’s t-test was used for two groups comparison; ns indicated no significant difference, ^*P < 0.05, and ^**P < 0.01

Female mice subjected to HF modeling exhibited distinct transcriptional-level differential gene expression profiles in the presence versus absence of miR-30a-5p

To elucidate the molecular mechanisms underlying the differential HF susceptibility between genotypes, we performed transcriptomic analysis of cardiac tissues from female WT + HF and KO + HF mice. Principal Component Analysis (PCA) revealed distinct clustering patterns between groups (Fig. 5A), with volcano plots and heat maps confirming significant differential gene expression (Fig. 5B, C). Pathway enrichment analysis identified the top 20 most dysregulated pathways (Fig. 5D, E), particularly highlighting: integrin-mediated processes (adhesion of integrin αIIbβ3 to fibrin network and Interaction of integrin αDβ2 with fibrin), fibrin-related pathways (formation of fibrin clot and fibrin monomers-fibrin multimer assembly), and inflammatory regulation (TLR signaling by endogenous ligands and platelet aggregation). Key differentially expressed genes mediating these pathways included Itgad, Itga2b, fibrinogen subunits (Fga, Fgb, Fgg), and apolipoproteins (Apoa1, Apoa2, Apob) (Supplementary Fig. 6). Notably, male mice exhibited fundamentally different transcriptional profiles, with KO vs. WT comparisons showing enrichment in apoptosis pathways (Supplementary Fig. 7), WT + HF vs. WT comparisons showing enrichment in metabolic pathways like cysteine formation from homocysteine (Supplementary Fig. 8), while WT + HF vs. KO + HF displayed fewer differential genes (Supplementary Fig. 9)-consistent with their similar phenotypic responses to HF induction. These findings reveal a female-specific transcriptional signature centered on integrin—fibrinogen interactions and vascular inflammation that underlies the protective effects of intact miR-30a-5p/estrogen signaling against HF pathogenesis.

Fig. 5.

Transcriptome sequencing between WT + HF and KO + HF female mice. A Principal component analysis (PCA) between the WT + HF and KO + HF groups (n = 5). B The volcano plot showed the difference of genes between the groups of the WT + HF and KO + HF. C The heat map was hierarchical clustering of gene expression differences between the WT + HF and KO + HF groups. D Top 20 enrichment bar plot between the WT + HF and KO + HF groups by Reactome enrichment. E Top 20 enrichment bubble chart between the WT + HF and KO + HF groups by Reactome enrichment

Transcriptome-identified molecular targets potentially mediate the successful establishment of HF in miR-30a-5p-deficient female mice

Based on the transcriptome sequencing results, we verified the expression of key genes in the heart and liver of female mice. Initially, we used immunohistochemistry to examine the expression of integrin-related proteins, including ITGAD, ITGA2B, ITGB3, ITGB2, ITGAL, and ITGAM. The results showed that ITGA2B and ITGB2 were almost undetectable in the heart tissue of female mice (Supplementary Fig. 10A), while ITGAL, ITGAM, ITGAD, and ITGB3 existed in the heart tissue (Fig. 6A–H). Among these, ITGAL and ITGB3 showed almost no variation across different groups (Fig. 6A, B, G, H). The expression levels of ITGAM and ITGAD increased after HF modeling, with the highest expression observed in the KO + HF group (Fig. 6C, D, E, F).

Fig. 6.

Verification of transcriptome sequencing results. The representative images and quantitative results of immunohistochemistry staining of integrin subunit alpha L (ITGAL) (A–B), integrin subunit alpha M (ITGAM) (C–D), integrin subunit alpha D (ITGAD) (E–F), integrin subunit beta 3 (ITGB3) (G–H), and fibrinogen alpha chain (FGA) (I–J) in heart tissues of WT + NC, KO + NC, WT + HF and KO + HF groups of female mice (n = 6), Scale bars = 50 μm. K–L Representative images and quantitative results of immunofluorescence of fibrinogen beta chain (FGB) in heart tissues of WT + NC, KO + NC, WT + HF and KO + HF groups of female mice (n = 6), Scale bars = 50 μm. Representative images and quantitative results of immunofluorescence of apolipoprotein A-1 (ApoA1) (M–N), apolipoprotein A-2 (ApoA2) (O–P), and apolipoprotein B (ApoB) (Q–R) in liver tissues of WT + NC, KO + NC, WT + HF and KO + HF groups of female mice (n = 6), Scale bar = 50 μm. Data are expressed as mean ± SD and one-way ANOVA with Tukey’s test was used for multiple comparisons; ns indicated no significant difference, ^**P < 0.01, ^***P < 0.001, and ^****P < 0.0001

We then examined the expression of the fibrinogen alpha chain (FGA) and beta chain (FGB) in the heart tissue of female mice (Fig. 6I, K). The results indicated that the expression of both FGA and FGB followed a similar pattern, with significantly lower levels in the KO + HF group compared to the WT + HF group (Fig. 6J, L). Furthermore, we verified that the expression of FGA and FGB in the liver of female mice was consistent with the findings in the heart tissue (Supplementary Fig. 10B, C). Since APOA1, APOA2, and APOB were minimally expressed in cardiac tissue, we assessed their expression in the liver of female mice through immunofluorescence (Fig. 6M, O, Q). The results showed that the expression levels of APOA1 and APOB in the liver tissue of the KO + HF group were significantly higher than in the other three groups (Fig. 6N, R). There was no significant difference in APOA2 expression among the groups, and its expression remained consistently high (Fig. 6P). Therefore, the cardiac damage of female mice caused by miR-30a-5p deletion may be closely associated with ITGAM, ITGAD, FGA, and FGB, while the liver damage might be linked to alterations in FGA, FGB, APOA1, and APOB. However, there were no significant differences in the candidate target gene (ITGAM, ITGAD, FGA, FGB, APOA1, and APOB) expression in WT + HF and KO + HF groups of male mice (Supplementary Fig. 11), which precisely reflects the sexual dimorphism of miR-30a-5p in HFpEF development.

Ago2 RNA immunoprecipitation (RIP) assays were performed to validate the direct binding between miR-30a-5p and the candidate target genes (Itgam, Itgad, Fga, Fgb, Apoa1, and Apob) (Supplementary Fig. 12). Unfortunately, the absence of enrichment for ITGAM, ITGAD, FGA, FGB, APOA1, and APOB suggests that these genes are indirectly regulated by miR-30a-5p under the tested conditions.

miR-30a-5p overexpression mitigates cardiac and hepatic damage in KO female mice via estrogen-regulated targets

Rescue experiment is very indispensable. So, we systemically administered either AAV-mediated miR-30a-5p overexpression vector (AAV-OE) or negative control vector (AAV-NC) via tail vein injection (Fig. 7A), and the expression of miR-30a-5p was verified after AAV infection for 4 weeks (Supplementary Table 2). To confirm the specificity and effects of knockout as well as overexpression, we verify the expression of SIRT1 [15], TP53INP1 [13], and RRM2 [23] that are well-established downstream targets of miR-30a-5p based on previously validated literature (Supplementary Fig. 13). As expected, results demonstrate that KO significantly upregulates the expression of Rrm2, Tp53inp1, and Sirt1, and OE downregulates them. AAV-OE treatment reduced pathological indices, including lung wet/dry weight, BW/TL, HW/BW, and HW/TL (Fig. 7B). Echocardiography revealed improved diastolic function (reduced E/E’ ratios) without altered systolic parameters (LVEF) (Fig. 7C, D). Blood pressure (SBP/DBP), exercise tolerance, and cardiac injury markers (LDH, CK, CK-MB, cTnT), were also improved post-AAV-OE (Fig. 7E, F, G). Histological analyses confirmed reduced hypertrophy (WGA staining), attenuated cardiac fibrosis (Masson’s trichrome, Sirius red), and diminished hepatic lipid accumulation (H&E, Oil Red O) (Fig. 7H, I, J, K; Supplementary Fig. 14A, B). The expression of miR-30a-5p in the liver was also significantly increased in the AAV-OE group (Supplementary Fig. 15), which might be due to expression leakage during AAV treatment. Serum metabolic markers (ALT, AST, cholesterol, triglycerides, LDL-c, glucose) declined significantly (Supplementary Fig. 14C, D). Mechanistically, AAV-OE downregulated cardiac ITGAD and ITGAM while elevating FGA and FGB expression (Fig. 7L, M, N, O), and suppressed hepatic APOA1 and APOB levels (Fig. 7P, Q). These findings demonstrate that miR-30a-5p restoration mitigates HF progression in KO females by modulating integrin/fibrinogen pathways (ITGAD, ITGAM, FGA, FGB) and lipid metabolism regulators (APOA1, APOB), with estrogen serving as a key upstream mediator.

Fig. 7.

MiR-30a-5p overexpression rescued HF phenotype of KO mice with HFD and L-NAME. A Experimental design of KO female mice (6–8 weeks) treatment with AAV-NC or AAV-OE under the induction of HFD and L-NAME. B Lung weight (wet/dry), BW/TL, HW/BW, and HW/TL in the AAV-NC + KO + HF and AAV-OE + KO + HF groups of female mice (n = 8). C Representative images of pulsed-wave Doppler tracings of PW-mode and Tissue-mode in AAV-NC + KO + HF and AAV-OE + KO + HF groups of female mice. The heart beats of echocardiographic analyses were stabilized to 450–550 beats/min. D LVEF and E/E′ were measured in AAV-NC + KO + HF and AAV-OE + KO + HF groups of female mice (n = 8). E SBP and DBP were detected in AAV-NC + KO + HF and AAV-OE + KO + HF groups of female mice (n = 8). F Running distance and running time from exercise exhaustion tests in AAV-NC + KO + HF and AAV-OE + KO + HF groups of female mice (n = 8). G Serum level of LDH, CK, CK-MB, and CTNT in AAV-NC + KO + HF and AAV-OE + KO + HF groups of female mice (n = 8). H Quantification of cardiomyocyte cross-sectional area based on WGA staining in AAV-NC + KO + HF and AAV-OE + KO + HF groups of female mice (n = 6). I Representative images of WGA staining of heart tissues in AAV-NC + KO + HF and AAV-OE + KO + HF groups of female mice (n = 6). Scale bars = 50 μm. J Representative images of Masson’s Trichrome and Sirius red staining of the left ventricle section in AAV-NC + KO + HF and AAV-OE + KO + HF groups of female mice (n = 6). Scale bars = 50 μm. K Quantification of Masson’s Trichrome and Sirius red staining of the left ventricle section in AAV-NC + KO + HF and AAV-OE + KO + HF groups of female mice (n = 6). L–M Representative images and quantification of immunohistochemistry staining of ITGAD and ITGAM of heart tissues in AAV-NC + KO + HF and AAV-OE + KO + HF groups of female mice (n = 4). Scale bars = 50 μm. N–O, representative images and quantification of FGA (immunohistochemistry) and FGB (immunofluorescence) of heart tissues in AAV-NC + KO + HF and AAV-OE + KO + HF groups of female mice (n = 4). Scale bars = 50 μm. P–Q Representative images and quantification of immunofluorescence staining of APOA1 and APOB of liver tissues in AAV-NC + KO + HF and AAV-OE + KO + HF groups of female mice (n = 4). Scale bars = 50 μm. Data are expressed as mean ± SD and Student’s t-test was used for two groups comparison; ns indicated no significant difference, ^*P < 0.05, ^**P < 0.01, ^***P < 0.001, and ^****P < 0.0001

miR-30a-5p overexpression has no protective effect against HF in WT OVX mice

Our study establishes that miR-30a-5p mitigates heart failure pathogenesis in female mice through estrogen-mediated regulation of integrin, fibrinogen, and lipid metabolism pathways. To dissect this estrogen-dependent mechanism, we employed ovariectomized (OVX) mice subjected to a multi-hit HF model (HFD/L-NAME) with systemic AAV-mediated miR-30a-5p overexpression (AAV-OE) or control (AAV-NC) (Fig. 8A). After OVX, the estradiol level of both groups of mice maintained a relatively low level (Supplementary Fig. 16). Despite successful miR-30a-5p restoration, OVX + HF mice exhibited unaltered pulmonary edema (lung wet/dry ratio, Fig. 8B), unchanged exercise tolerance (running distance/duration, Fig. 8C), and persistent diastolic dysfunction (elevated E/E’ ratios despite preserved LVEF, Fig. 8D, E). Hypertension (Fig. 8F), myocardial injury (Fig. 8G, H), hepatic steatosis (Supplementary Fig. 17A, B), and unimproved serum markers of cardiac damage (LDH, CK, CK-MB, cTnT) or dyslipidemia (Fig. 8I; Supplementary Fig. 17C, D) further confirmed the inefficacy of miR-30a-5p in estrogen-deficient conditions. Crucially, expression of key targets—ITGAD, ITGAM (integrins), FGA, FGB (fibrinogen), and APOA1, APOB (lipid regulators)—remained unchanged in AAV-OE + OVX + HF versus AAV-NC + OVX + HF groups (Fig. 8J, K, L, M, N, O, P, Q). As the control, we integrated the data of OVX + AAV-OE and OVX + AAV-NC (Supplementary Fig. 18). And there was no significant difference between the two groups and HF phenotype didn’t appear in both groups. These results conclusively demonstrate that miR-30a-5p’s protective effects in HF are strictly contingent on intact estrogen signaling, positioning estrogen as an essential mediator of its regulatory control over cardiac remodeling and metabolic homeostasis.

Fig. 8.

Effect of OE treatment on female mice with OVX under HFD and L-NAME induction. A WT female mice (6–8 weeks) were treated with HFD and L-NAME for 24 weeks after OVX or sham operation, and also treated 100 µL AAV-NC and AAV-OE intravenously at week 12 and 24 after OVX. B lung weight (wet/dry) and HW/BW in AAV-NC + OVX + HF and AAV-OE-OVX + HF groups of female mice (n = 8). C Running distance and running time from exercise exhaustion tests in AAV-NC + OVX + HF and AAV-OE-OVX + HF groups of female mice (n = 8). D representative images of pulsed-wave Doppler tracings of PW-mode and Tissue-mode in AAV-NC + OVX + HF and AAV-OE-OVX + HF groups of female mice. The heart beats of echocardiographic analyses were stabilized to 450–550 beats/min. E LVEF and E/E′ were measured in AAV-NC + OVX + HF and AAV-OE-OVX + HF groups of female mice (n = 8). F SBP and DBP were detected in AAV-NC + OVX + HF and AAV-OE-OVX + HF groups of female mice (n = 8). G–H Representative images and quantification of Masson’s Trichrome and Sirius red staining of the left ventricle section in AAV-NC + OVX + HF and AAV-OE-OVX + HF groups of female mice (n = 6). Scale bars = 50 μm. I The levels of serum LDH and CTNT in AAV-NC + OVX + HF and AAV-OE-OVX + HF groups of female mice (n = 8). J–M representative images and quantification of immunohistochemistry staining of ITGAD (J), ITGAM (K) and FGA (L) of heart tissues of female mice. Scale bars = 50 μm. N Representative images of immunofluorescence staining of FGB of heart tissues in female mice. Scale bars = 50 μm. O–P Representative images of immunofluorescence staining of Apoa1 (O) and ApoB (P) of liver tissues in female mice. Scale bars = 50 μm. Q Quantification of immunofluorescence staining of FGB, Apoa1 and ApoB in female mice (n = 4). Data are expressed as mean ± SD and Student’s t-test was used for two groups comparison, ns indicated no significant difference

Discussion

Our study revealed significant sex-specific differences in response to high-fat diet (HFD) and L-NAME treatment during HF model establishment in miR-30a-5p knockout (KO) versus wild-type (WT) mice. Building upon established evidence that estrogen exerts protective effects in various disease models [24–26], we observed that miR-30a-5p KO female mice lost this estrogen-mediated protection during HF development. We have previously confirmed that the ejection fraction of KO mice over 6 months old is significantly lower than that of WT mice [15]. Therefore, KO mice (female or male) showed the characteristics of HFmrEF and HFpEF after HFD/L-NAME treatment. Through comprehensive analysis, we identified that miR-30a-5p regulates estrogen production in female mice, which subsequently modulates HF pathogenesis through downstream effects on key molecular targets Itgam, Itgad, Fga, Fgb, Apoa1, and Apob in mice. This regulatory network provides a molecular basis for the observed sexual dimorphism in HF progression and suggests miR-30a-5p as a potential therapeutic target for sex-specific HF treatment strategies.

HFpEF is a complex clinical syndrome involving multiple interrelated pathological features including diastolic dysfunction, metabolic disturbances (particularly obesity), chronic inflammation, and microvascular dysfunction, though current mechanistic research remains in its early stages with primary focus on oxidative stress, microRNA networks, energy metabolism, and inflammatory responses [5, 27–33]. The significant clinical heterogeneity and mechanistic complexity of HFpEF demand interdisciplinary research approaches, particularly given the current lack of substantive clinical treatment guidelines. In our study, we employed a well-established murine model combining the high-fat diet and L-NAME (nitric oxide synthase inhibitor) treatment to recapitulate human HFpEF pathophysiology [20], which successfully induced characteristic phenotypes including preserved ejection fraction with diastolic dysfunction, hypertension, and metabolic disorders in male wild-type (WT) [34]. Interestingly, while female WT mice demonstrated expected resistance to HF development consistent with known estrogen-mediated protection [24, 35], this protective effect was abolished in female KO mice (characterized by reduced ejection fraction and disordered diastolic function), suggesting that miR-30a-5p may play a crucial regulatory role in estrogen signaling pathways during HF pathogenesis. These findings highlight the importance of sexual dimorphism in HF development and point to miR-30a-5p as a potential key mediator in the estrogen-related protective mechanisms against HF.

Estrogen plays dual physiological roles in female animals, promoting both the development of secondary sexual characteristics and providing cardiovascular protection. Substantial evidence demonstrates estrogen’s cardioprotective effects through multiple mechanisms: upregulation of nitric oxide synthase to enhance endothelial function [36], maintenance of DNA methylation homeostasis to mitigate age-related cardiovascular decline [37], and prevention of hyperlipidemia and atherosclerosis in estrogen-deficient models [38]. Our study reveals that estrogen deficiency—whether induced by ovariectomy or miR-30a-5p knockout—similarly facilitated HFpEF development in female mice, with KO mice showing significantly reduced estrogen levels following HFD/L-NAME treatment. To investigate the miR-30a-5p/estrogen axis, we employed AAV-mediated miR-30a-5p overexpression in ovariectomized mice. While existing literature documents bidirectional miRNA-estrogen interactions—including miRNA regulation of estrogen signaling [39, 40] and estrogen-mediated miRNA expression control [41, 42] (e.g., the estrogen/miR-122/TEAD1 pathway in hepatic steatosis [43]) —our findings uniquely position miR-30a-5p upstream of estrogen production. This conclusion is supported by the observation that miR-30a-5p overexpression failed to rescue HF phenotypes in ovariectomized mice, indicating its primary role in modulating estrogen synthesis.

To further investigate the molecular mechanisms underlying miR-30a-5p-mediated regulation of HF in female mice, we performed RNA sequencing analysis and identified several key candidate genes (Itgad, Itgam, Fga, Fgb, Apoa1, and Apob) potentially involved in this regulatory pathway. Among these, ITGAD and ITGAM represent α subunits of the integrin family—transmembrane cell adhesion molecules formed by non-covalent α/β heterodimers. Our findings suggest these integrins may contribute to diastolic dysfunction and cardiac hypertrophy development in female mice, consistent with previous reports demonstrating: elevated ITGAM expression in myocardial infarction patients, where it potentially mediates pathogenesis through immune cell infiltration [44]; ITGAM’s critical role in tumor cell invasion and adhesion via H3K4me3 modification [45]; and the pro-inflammatory actions of integrin αdβ2 (comprising ITGAD/ITGB2) expressed on macrophages, which promotes atherosclerosis and diabetes progression by retaining macrophages in inflamed tissues [46, 47]. Notably, we observed significant inflammatory infiltration in cardiac and hepatic tissues of KO mice following HFD/L-NAME treatment, suggesting integrin αdβ2-mediated macrophage retention may drive these pathological inflammatory responses in HF development.

Estrogen has been well-documented to enhance hepatic fibrinogen synthesis through ERα-mediated upregulation of FGA, FGB and FGG expression [48], with similar effects observed via ERRγ induction by cannabinoid receptor agonists [49]. In our study, miR-30a-5p knockout female mice showed significantly reduced cardiac and hepatic FGA/FGB expression following HFD/L-NAME treatment, paralleling the observed estrogen deficiency. While miR-30a-5p overexpression reversed the aberrant expression of ITGAD, ITGAM, FGA, FGB, APOA1 and APOB induced by miR-30a-5p deletion, it failed to compensate for ovariectomy effects on these targets, consistent with previous findings that miR-877-5p regulates cancer cell invasion through FGB targeting [50]. Integrating our results, we propose that miR-30a-5p regulated FGA and FGB via estrogen-dependent mechanisms. This pathway, functioning upstream of estrogen production, modulates tissue inflammatory responses and damage repair processes, thereby providing a crucial mechanistic insight into the sexual dimorphism of HF pathogenesis.

The high-fat diet not only induces obesity and systemic insulin resistance but also promotes hepatic lipid accumulation and dyslipidemia—key features of cardiometabolic disease [6, 51]. Emerging evidence indicates that postmenopausal women exhibit elevated APOA1 levels (the primary structural protein of HDL), while APOB (the main component of atherogenic lipoproteins including LDL and VLDL) progressively increases with age to approach male levels, leading to an elevated APOB/APOA1 ratio [52]. This altered ratio has been clinically associated with type 2 diabetes mellitus prevalence, fasting blood glucose levels, and insulin resistance, with estrogen deficiency potentially exacerbating these metabolic disturbances [53]. Prior work has firmly established that miR-30d/e were elevated in and contribute to diabetic coronary microvascular dysfunction [54]. Our findings revealed significant upregulation of both APOA1 and APOB (but not APOA2) in the livers of miR-30a-5p knockout female mice following HFD/L-NAME. The increased hepatic APOA1 expression correlated with elevated serum HDL levels, which we speculate represents a compensatory response to pathological lipid accumulation. This imbalance especially APOB likely exacerbates systemic inflammation, oxidative stress, and vascular stiffness, thereby accelerating diastolic dysfunction and cardiac remodeling [55]. This observation aligns with Ansari et al.‘s discovery that certain miRNAs can coordinately regulate both APOB-containing LDL and APOA1-containing HDL levels, suggesting the existence of a unified miRNA-mediated mechanism for bidirectional lipoprotein metabolism regulation [56]. The presence of hepatic steatosis remains a widely recognized indicator of insulin resistance and its metabolic complications [57], and the results of fasting insulin and HOMA-IR demonstrated systemic insulin resistance in our female HFpEF model of cardiometabolic HF. Thus, we believe that insulin resistance may serve as a critical pathophysiological link between the metabolic disturbances observed and the ensuing cardiac dysfunction. It was previously discovered that ERα directly binds to the Apoa1 gene’s estrogen response element, promoting transcription and establishing an active chromatin environment [58]. Our data also demonstrate that miR-30a-5p’s modulation of APOA1 in lipid metabolism is estrogen-dependent, providing novel insights into the sexual dimorphism of metabolic dysregulation in HF pathogenesis. These insights align with clinical observations that HFpEF often coexists with obesity, diabetes, and metabolic syndrome, and reinforce the importance of targeting metabolic pathways in the management of cardiometabolic heart failure [8, 59].

The limitation of this study is as follow: The systemic AAV9 delivery method leads to detectable transgene expression in both the heart and the liver. Consequently, while our data demonstrate a functional heart-liver interplay, we cannot definitively partition the observed hepatic phenotype between a direct effect of hepatic miR-30a-5p overexpression and an indirect effect mediated by cardiac-derived signals or systemic consequences of heart failure. The causal pathway for the hepatic changes is therefore likely multifactorial. Besides, while our study provides novel mechanistic insights into sex-specific regulation of HFpEF, we acknowledge that these findings are derived from a murine model and require direct validation in human HFpEF settings. Our study contrasts with a seminal study using a shorter-term HFpEF model, which found female protection to be hormone-independent [60]. This discrepancy underscores that the role of sex hormones may be critically dependent on the stage and chronicity of the disease process modeled. Future studies systematically comparing phenotypes and molecular signatures across multiple time points in both sexes are warranted to resolve these temporal dynamics.

Conclusion

In conclusion, our study establishes that miR-30a-5p exerts protective effects against HF pathogenesis—including cardiac systolic and diastolic dysfunction, hypertrophy, and hepatic metabolic disorders—through estrogen-mediated regulation of the molecular targets (ITGAM, ITGAD, FGA, FGB, APOA1, and APOB). By elucidating this multi-level regulatory mechanism within the complex pathophysiology of HF, our findings provide significant clinical insights into the sexual dimorphism of this condition. While these advances represent important progress in understanding HF mechanisms, substantial further research is needed to fully unravel the disease’s complexity and facilitate the translation of these discoveries into novel therapeutic strategies and clinical applications.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

HF

Heart failure

HFpEF

Heart failure with preserved ejection fraction

HFrEF

Heart failure with reduced ejection fraction

WT

Wild-type

KO

miR-30a-5p knockout

OE

Overexpression

OVX

Ovariectomy

NAFLD

Non-alcoholic fatty liver disease

MiRNA

microRNA

L-NAME

Nω-nitro-l-arginine methyl ester

HFD

High-fat diet

OCT

Optimal cutting temperature

H&E

Hematoxylin–eosin

WGA

Wheat germ agglutinin

LDH

Lactate dehydrogenase

CK

Creatine kinase

cTnT

Cardiac troponin T

HDL-c

High-density lipoprotein-cholesterol

LDL-c

Low-density lipoprotein-cholesterol

ALT

Alanine aminotransferase

AST

Aspartate aminotransferase

ELISA

Enzyme-linked immunosorbent assay

E2

Estradiol

ER

Estrogen receptor

SBP

Systolic blood pressure

DBP

Diastolic blood pressure

PCA

Principal component analysis

FGA

Fibrinogen alpha chain

FGB

Fibrinogen beta chain

BW/TL

Body weight-to-tibia length ratios

HW/BW

Heart weight-to-body weight ratios

HW/TL

Heart weight-to- tibia length ratios

Author contributions

G. L. and R.W. designed the experiments. R.W. and L.Z. performed most of the experiments. R.W. and G.L. wrote the manuscript. G.L. supervised the study. G.L. and G.F revised the manuscript. X-C.Z. prepared the KO mice. X-C.Z., M.Q., S-Y.Z., J-F.C., L-Q.Y., and X-X.Z. performed the animal experiments. G.L., L.Z. and R.W. provided funding. All authors contributed to the writing and editing of the manuscript.

Funding

This study was supported by grants from the National Natural Science Foundation of China (Grant no. 82200373) and the Natural Science Foundation of Fujian Province (Grant Nos. 2024J011423, 2023J011677 and 2025J011489).

Data availability

All the other data supporting the findings of this study are available within the article and its Supplementary Information files. All original data for this study can be obtained from the corresponding author.

Declarations

Ethics approval and consent to participate

All experiments of animals on this study were approved by the Ethics Committee of the Laboratory Animal Management and Ethics Committee of Xiamen University (Approval NO. XMULAC20190120).

Competing interests

The authors declare no competing interests.