Gut microbiota of miR‐30a‐5p‐deleted mice aggravate high‐fat diet‐induced hepatic steatosis by regulating arachidonic acid metabolic pathway

C57BL/6J mice (6–8 weeks old, male) were purchased from GemPharmatech (Nanjing, China). We used CRISPR/Cas9 technology to construct miR‐30a‐5p^−/− mice with C57BL/6J genetic background, whose genotypes were verified by PCR and using the following primers: 5′‐AGCTTCCCTACTTTGGTGTTT‐3′; 5′‐ TGGTGTGTGTGAATTGACCT‐3′. All mice were kept in the specific pathogen‐free laboratory of the Animal Experimental Center of Xiamen University (22°C ± 2°C, humidity 50−70%, 12 h light/dark cycle). All animal studies were conducted in compliance 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).

For the treatment with a high‐fat diet (HFD, PAD, Dyets Biotechnology (Wuxi), Jiangsu, China), C57BL/6J wild‐type and miR‐30a‐5p^−/– mice were randomly assigned to two groups: normal diet (WT group, KO group) and high‐fat diet (WT + HFD group, KO + HFD group) for a duration of 8 weeks. Liver, blood and heart tissues were collected after sacrifice by inhalation of carbon dioxide.

To assess the impact of fecal microbiota transplantation from HFD‐fed miR‐30a‐5p^−/− mice on the liver and lipid metabolism of recipient mice, WT and miR‐30a‐5p^−/– mice (6‐8 weeks old, male) were subjected to HFD for 8 weeks, with fecal samples collected accordingly. Fecal microbiota were collected every two days and pooled within each group in sterile PBS (150 mg feces/1 mL PBS) and centrifuged at 500 rpm/min for 5 min. Before transplantation, recipient mice were treated with the antibiotic mixture (ABX), comprising vancomycin (.5 g/L, Cat#: 1161, BioFroxx, Germany), cefixime (1 g/L, Cat#: HY‐B1381, MedChemExpress, Shanghai, China) and metronidazole (1 g/L, Cat#: HY‐B0318, MedChemExpress, Shanghai, China) in drinking water for 2 weeks. Subsequently, the fecal microbiota were intragastric administrated to the recipient mice (6‐8 weeks old, male) for 2 weeks. Following transplantation, mice were maintained on a normal diet for an additional 8 weeks, as depicted in Figure 2A. Liver, blood and heart tissues were collected for further analysis upon sacrifice by inhalation of carbon dioxide.

To further verify the regulation of miR‐30a‐5p on the liver function and arachidonic acid metabolism, we injected 100 µL/each HBAAV2/9‐cTNT‐LUC (AAV‐NC) or HBAAV2/9‐cTNT‐mmu‐mir‐30a‐LUC (AAV‐OE, HH20230815FJCJY, HANBIO, Shanghai, China) into miR‐30a‐5p^−/− mice through the tail vein. Two weeks after injection, the mice began HFD for 8 weeks. Liver tissues and blood were collected after sacrifice by inhalation of carbon dioxide.

Following collection, mouse liver samples were embedded in an optimal cutting temperature (OCT) compound (Cat#: G6059, Servicebio, Wuhan, China) and frozen at −80°C. Frozen sections of 6 µm thickness were obtained. For H&E staining, we used a hematoxylin‐eosin kit (Cat#: G1076, Servicebio, Wuhan, China) as the protocol of the manufacturer. Masson trichrome staining was carried out using a Masson staining solution (Cat#: G1006, Servicebio, Wuhan, China). Oil Red O staining was performed using an Oil Red O saturated solution (Cat#: G1015, Servicebio, Wuhan, China) for lipid staining. Photographs were captured using a Leica microscope.

After treatment, mouse blood samples were collected and centrifuged at 3000 rpm for 10 min to obtain serum. 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), 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) levels in mouse serum were quantified using the Beckman Coulter AU biochemical analysis system (Beckman Coulter, CA, USA).

The results of 16S rDNA amplicon sequencing were performed and analysed by Applied Protein Technology Co., Ltd (Order NO: P20220502318, Shanghai, China). The genomic DNA of the gut contents from WT + HFD mice (n = 7) and KO + HFD mice (n = 7) was extracted using the cetyltrimethylammonium bromide (CTAB) and sodium dodecyl sulfate (SDS) method. DNA purity and concentration were assessed, and the selected V3‐V4 variable regions were amplified by PCR using specific primers with barcodes and high‐fidelity DNA polymerase: 16S V3‐V4 (341F: CCTAYGGGRBGCASCAG; 806R: GGACTACNNGGGTATCTAAT). PCR products were separated using 2% agarose gel electrophoresis, and the target fragments were excised. An AxyPrepDNA gel recovery kit (AP‐GX‐250, AXYGEN, CA, USA) was used for gel recovery. According to the preliminary quantitative results of electrophoresis, the recovered products of PCR amplification were detected and quantified by QuantiFluor™‐ST blue fluorescence quantitative system (Promega, Wisconsin, USA), and mixed according to the requirements of sequencing quantity of each sample. NEB Next ®Ultra library DNA Library Prep kit (E7645, NEB, USA) was employed for library construction, with library quality assessed using Agilent Bioanalyzer 2100 and Qubit. Upon passing the quality test, libraries were sequenced, followed by Operational Taxonomic Units (OTU) clustering and species classification analysis based on valid data. Based on the results of the OTU cluster analysis, we analysed the diversity index of OTU and detected the sequencing depth. Based on taxonomic information, statistical analysis of community structure was carried out at various taxonomic levels.

Untargeted metabolomics detection and analysis were conducted by Applied Protein Technology Co., Ltd (Order NO: P20221006501, Shanghai, China). Gut contents from WT + HFD mice (n = 8) and KO + HFD mice (n = 8) were separated by an ultra‐performance liquid chromatography system (Agilent 1290 Infinity LC, CA, USA), detected using a mass spectrometer (AB Sciex Triple TOF 6600, CA, USA), and then analysed in both electrospray ionisation positive and negative ion modes. The XCMS analysis software (xcmsonline.scripps.edu) was used for analysis. The variable importance (VIP) value of each variable in the OPLS‐DA model in projection was calculated, and the significance of differences between the two independent samples was determined by the t‐test. Metabolites exhibiting significant changes were identified based on criteria of VIP > 1 and p < .05.

Targeted metabolomics detection and analysis were performed by Applied Protein Technology Co., Ltd (Order NO: YAS202306290056, Shanghai, China). Liver tissues from WT + HFD mice (n = 9) and KO + HFD mice (n = 9) were separated using an ultra‐performance liquid chromatography system (Agilent 1290 Infinity LC, CA, USA) and detected using a mass spectrometer (5500 QTRAP, SCIEX, CA, USA). Chromatographic peak areas and retention times were extracted using the Multiquant 3.0.2 software. Retention times were adjusted using standard substance correction for the target compounds, and the metabolites were identified.

After treatment, the livers of mice were collected. The total RNA of the liver tissues was then isolated using the TRIzol/chloroform/isopropanol method. The total RNA was reverse transcribed into cDNA using a reverse transcription kit (Cat#: RR047A, Takara Biomedical Technology, Beijing, China). RNA levels were quantified using a quantitative PCR kit (Cat#: 11204ES08, Yeasen, Shanghai, China) with the following primer sets: ALOX5 (F:5′‐AGGCGAGGGCGAGATTGTC‐3′; R: 5′‐ CGTCTGTGCTGCTTGAGGATG‐3′), ALOX12 (F:5′‐CAGGTAGTGAGCACAGGTGGAG‐3′; R: 5′‐GCCAGATCATCAGGAGGACAGAG‐3′), ALOX15 (F:5′‐AGTTCCTCACCTACTCTACACCATG‐3′; R: 5′‐CCTCCTCCACCTGTGCTCATC‐3′), COX1 (F:5′‐CCAGAACCAGGGTGTCTGTGTC‐3′; R: 5′‐ACAGTTGGGGCCTGAGTAGC‐3′), and COX2 (F:5′‐ACCCATCAGTTTTTCAAGACAGAT‐3′; R: 5′‐GCGCAGTTTATGTTGTCTGT‐3′).

After treatment, mouse liver samples were embedded in the OCT agent (Cat#: G6059, Servicebio, Wuhan, China) and frozen at −80°C. Frozen sections of 6 µm thickness were obtained using the freezing microtome (Leica 1950, Germany). Sections were incubated with primary antibodies against ALOX5 (Cat#: A2158, Abclonal, Wuhan, China), ALOX12 (Cat#: A14703, Abclonal, Wuhan, China), ALOX15 (Cat#: A6864, Abclonal, Wuhan, China), COX1 (Cat#: A7341, Abclonal, Wuhan, China), COX2 (Cat#: A3560, Abclonal, Wuhan, China). The sections were then incubated with the corresponding secondary antibody (HRP labelled goat anti‐rabbit IgG, Cat#: GB23303, Servicebio, Wuhan, China) for 1 h. Staining was visualised using a DAB solution (Cat#: SW1020, Solarbio, Beijing, China) and counterstained with hematoxylin. The images were obtained using a microscope (Leica, Germany). Positive staining appeared as brownish yellow, while nuclei stained with hematoxylin appeared blue. The average optical density (AOD) was measured and analysed by Image J (IHC toolbox plugin) software.

Data are expressed as means ± standard deviations (Mean ± SD) and analysed using GraphPad Prism 9. Normally distributed data (2 groups) were analysed by Student's t‐test. One‐way analysis of variance (ANOVA) followed by either the Tukey test or Dunnett test was employed for multiple comparisons (> 2 groups) as deemed appropriate. Statistical significance was defined as a two‐sided p < .05.

To investigate the effect of miR‐30a‐5p deletion on liver function and lipid metabolism, we fed the WT and miR‐30a‐5p knockout (KO) mice with a high‐fat diet (HFD) or normal diet for 2 months and tested the liver histology, serum lipid levels, liver functions, and aortic root plaque formation. The deletion of miR‐30a‐5p in KO mice was confirmed by PCR and qPCR, as shown in Figure S1A and B. After treatment, there was increased liver adipose in WT mice after HFD treatment (Figure 1A). Conversely, HFD‐treated KO mice displayed augmented adipocyte size and quantity, accompanied by heightened inflammatory cell infiltration compared to HFD‐treated WT mice (Figure 1A). Consistent with H&E staining, Oil Red O staining revealed intensified red and dark colour in HFD‐treated KO mice compared to HFD‐treated WT mice (Figure 1B). We further evaluated blood lipid metabolism, including total cholesterol, triglyceride, high‐density lipoprotein‐cholesterol (HDL‐c), and low‐density lipoprotein‐cholesterol (LDL‐c). Results indicated that HFD treatment in WT mice led to elevated levels of total cholesterol and LDL‐c, which were further elevated in the HFD‐treated KO mice. The level of triglyceride had no significant change among the four groups, while the HDL‐C level demonstrated a sole reduction in the KO + HFD group (Figure 1C). Alanine aminotransferase (ALT) and aspartate aminotransferase (AST) are the key indicators for liver damage. Serum levels of ALT and AST in both WT and KO mice increased significantly after HFD treatment, particularly evident in the KO + HFD group (Figure 1D). The change in total bile acids was consistent with ALT and AST, but the level of cholinesterase remained largely unchanged across the four groups (Figure 1D). In a previous study,15 we have proved that the plaque in the aortic root of ApoE^−/− mice intervened by miR‐30a‐5p knockout adenovirus increased significantly compared with that of ApoE^−/− mice after high‐fat feeding. Thus, we assessed the aortic plaque formation of the above‐treated mice (Figure 1E and F). However, no significant differences in plaque formation were observed, potentially attributed to the lower sensitivity of C57BL/6J mice to plaque development. In brief, the deletion of miR‐30a‐5p in mice exacerbates hepatic injury and steatosis, along with elevated blood lipid levels following HFD.

To further ascertain the regulatory effect of intestinal microorganisms of KO mice on hepatic metabolism, after ABX treatment for 2 weeks, we transplanted fecal microorganisms from HFD‐treated WT or KO mice to the recipient WT or KO mice. Following transplantation for 2 weeks, a normal diet was fed to the mice for an additional 2 months (Figure 2A). The results showed that the levels of total cholesterol and LDL‐c in the serum of mice receiving FMT from the KO + HFD group were significantly higher compared to those receiving FMT from the WT + HFD group. The levels of total cholesterol and LDL‐c in the KO mice were consistently higher than those in the WT mice under the same treatment conditions (Figure 2B). However, there were no significant differences in the levels of triglyceride and HDL‐c in serum between the two groups, both in WT and KO mice (Figure 2B). Histological examination of liver tissue from mice receiving FMT from the KO + HFD group revealed increased liver adipose, inflammatory cells, and a looser arrangement of hepatocytes compared to the WT + HFD group, observed in both WT mice and in KO mice (Figure 2C and D). In addition, the indicators of liver function aligned with these observations that intestinal microbes of the KO + HFD group promoted levels of ALT, AST and total bile acids in the serum of both WT and KO mice (Figure 2E). These results indicated that irrespective of mouse genotype, liver structural and functional changes following transplantation of intestinal microorganisms from the KO + HFD group are more pronounced than those from the HFD group, with KO mice exhibiting a more significant response to these alterations. The results of H&E, Oil Red O, and Masson staining revealed no significant changes and no plaque in the aortic root of mice in each group (Figure 2F), consistent with the findings in Figure 1E and F. Therefore, the intestinal microorganisms of WT and KO mice exhibit differentiation after HFD treatment, and the phenotypic disparity in hepatic metabolism between WT and KO mice may be related to the heterogenisation of intestinal microorganisms.

To investigate the role of fecal microbiota of HFD‐treated ApoE^−/− mice on liver function and atherosclerosis, fecal microbiota from both ND‐ and HFD‐treated ApoE^−/− mice were collected and transplanted fecal microbiota to ApoE^−/− mice for 2 weeks (Figure S2). The result showed that plaques appeared in the aorta of ApoE^−/− mice after transplanting intestinal microorganisms of the HFD‐treated mice ApoE^−/− mice [FMT (ApoE^−/− + HFD)], but no plaques were observed after transplanting intestinal microorganisms in the ND‐treated mice [FMT (ApoE^−/− + ND)], as shown in Figure S2B and C. Consistent with these findings, histopathological examination in the aortic roots and liver using H&E, Oil Red O, and Masson staining supported plaque induction by FMT (ApoE^−/− + HFD) in ApoE^−/− mice (Figure S2). The levels of total cholesterol, triglyceride, LDL‐c, and HDL‐c were also increased after the FMT (ApoE^−/− + HFD) treatment (Figure S2H). These results demonstrate the dependency of plaque formation and hepatic metabolic function in ApoE^−/− mice on the activities of intestinal microorganisms.

Based on the preceding results, we hypothesised that the intestinal microorganisms of miR‐30a‐5p KO may influence liver function. To investigate this, we applied 16S rDNA sequencing to examine the microbial composition differences between WT and KO mice after HFD induction. Analysis of the intestinal microbiota of the HFD‐treated WT and KO mice revealed substantial disparities in species composition and distribution between the two groups (Figure 3A). LDA Effect Size analysis further delineated distinct species of fecal microbiota between the two groups (Figure S3). Notably, the top 10 abundant species exhibiting significant differences between these groups comprised s_Bacteroides.dorei, s_Escherichia.coli, s_Lactobacillus.reuteri, s_Bacteroides.fragilis, s_uncultured.Bacterioides.sp, s_Lactobacillus.murinus, s_Lactobacillus.gasseri, s_uncultured.organism, s_uncultured.Bacteroidales.bacterium, s_uncultured.Bacterium and others (Figure 3B). Further analysis at the genus level revealed g_Akkermansia as the most significantly altered taxon between the WT + HFD and the KO + HFD groups (Figure 3C). Additionally, Kyoto Encyclopedia of Genes and Genomes (KEGG) and Clusters of Orthologous Groups of proteins (COG) analyses for function prediction indicated a significant impact on energy metabolism and energy production and conversion orchestrated by KO (Figure S4A and B). Studies have shown that Akkermansia can help mitigate NAFLD, obesity and cardiovascular diseases by regulating lipid metabolism and reducing inflammation.16 As such, Akkermansia has been identified as a promising ‘next‐generation beneficial microorganism’ for its health‐promoting properties.17 Our current findings, in conjunction with earlier research, suggest that the proportion of Akkermansia may be a significant factor for KO‐regulated liver dysfunction.

The metabolites of intestinal microbiota play a crucial role in mediating intestinal‐liver communication, we further analysed the intestinal microbiota‐related metabolites to explore the potential metabolites for KO‐regulated function. We applied metabolomics and analysed the intestinal contents of HFD‐treated WT and KO mice. A significant number of metabolites were found to differ between the two groups in both positive ion mode (Figures S5A and S6) and negative ion mode (Figures S5B and S7). The hierarchical clustering heat map in positive ion mode is shown in Figure S8. After establishing the relationship model between the expression of metabolites and sample categories by partial least square discrimination analysis (PLS‐DA), we found that the samples in the HFD and KO + HFD groups were significantly separated, confirming differences between them (Figure 4A). We then performed a hierarchical cluster analysis of significantly different metabolites (VIP > 1, p‐value < .05), and the metabolites clustered in the same cluster could share similar functions or be involved in the same metabolic process/pathway. The specific metabolites were displayed by the hierarchical clustering heat map, as shown in Figure 4B. To identify key pathways with notable impact, the KEGG pathway enrichment analysis was performed. Among the nine pathways with significant differences, ‘Arachidonic acid metabolism’ stood out as the most significantly altered pathway (Figure 4C). In addition to arachidonic acid metabolism, there are other pathways, such as ‘Protein digestion and absorption’, ‘Central carbon metabolism in cancer’, ‘Valine, leucine and isoleucine degradation/biosynthesis’, which may also play an important role in the normal function of the body. Further exploration of the differential metabolite associated with this pathway highlighted 3 upregulated metabolites (thromboxane b2, 11‐dehydrothromboxane b2, 5s‐14r‐lipoxin b4) and 4 downregulated ones (12‐oxo‐5z,8z,10e,14z‐eicosatetraenoic acid, arachidonic acid, leukotriene e4 and leukotriene f4) (Figure 4D).

Arachidonic acid, crucial for human growth and development, possesses various physiological functions such as lowering cholesterol, increasing vascular elasticity, inhibiting platelet aggregation, and reducing blood viscosity. The cyclooxygenase (COX) and lipoxygenase (LOX) pathways are critical in arachidonic acid metabolism. Employing targeted metabolomics, we precisely measured the levels of arachidonic acid, along with COX‐ and LOX‐related metabolites in liver tissues of HFD‐treated WT and KO mice. Our findings revealed that the arachidonic acid level in the KO + HFD group exceeded that in the WT + HFD group (Figure 5A). Regarding the COX pathway, levels of prostaglandin F2a (PGF2a, Figure 5B), 8‐iso‐prostaglandin F2a (8‐iso‐PGF2a, Figure 5C), and prostaglandin E2 (PGE2, Figure 5F) in the liver of KO + HFD mice were higher than those of WT + HFD mice, the level of 6‐keto‐prostaglandin F1a (6‐keto‐PGF1a, Figure 5D) thromboxane B2 (TXB2, Figure 5E), and prostaglandin D2 (PGD2, Figure 5G) was not affected. In the LOX pathway, the levels of leukotriene B4 (LTB4, Figure 5H), leukotriene D4 (LTDI, Figure 5I), 12s‐hydroxy eicosatetraenoic acid (12S‐HETE, Figure 5J), 15s‐hydroxy eicosatetraenoic acid (15S‐HETE, Figure 5K) in the liver in KO + HFD group were higher than those in HFD group. Nevertheless, no significant changes were observed in 13s‐hydroxy octadecadienoic acid (13S‐HODE, Figure 5L) and 9s‐hydroxy octadecadienoic acid (9S‐HODE, Figure 5M). The above results demonstrated that KO mice, after HFD treatment, activated the LOX and COX pathways of arachidonic acid metabolism.

According to the results of targeted metabolomics, LOX/COX pathways were considered as an important research direction. In LOX/COX pathways, some enzymes, including ALOX5, ALOX12, ALOX15, COX1 and COX2, affect the metabolic process of many factors and play a key role in the process of arachidonic acid metabolism. To further verify these findings, we detected the expression of key enzymes (ALOX5, ALOX12, ALOX15, COX1 and COX2) in liver tissues of WT/KO mice with ND/HFD‐fed. Immunohistochemistry staining indicated that in both the KO group and the WT + HFD group, the expression of ALOX5 (Figure 6A and F), ALOX12 (Figure 6B and F), and COX2 (Figure 6E and F) in liver tissues was significantly increased, with an even greater increase in the KO + HFD group. However, the expression of ALOX15 (Figure 6C and F) and COX1 (Figure 6D and F) showed no significant changes. Furthermore, we assessed the transcriptional levels by quantifying mRNA expression through qPCR, which exhibited similar patterns as observed at the protein level (Figure 6G). Based on the above results, we make a summary that miR‐30a‐5p KO mice, after HFD treatment, experienced accelerated hepatic metabolic disorder due to the COX/LOX‐based arachidonic acid metabolic pathway, and this process was closely related to the imbalance of intestinal microbial homeostasis.

To further investigate the role of miR‐30a‐5p on liver function and arachidonic acid metabolism, KO mice were injected AAV‐NC or AAV‐OE through the tail vein, and the expression of miR‐30a‐5p in mouse liver increased after AAV‐OE treatment (Figure S9). Two weeks later, these mice were HFD‐treatment for 8 weeks (Figure 7A). The results of HE and Oil Red O staining in mouse liver showed that AAV‐OE could reverse the hepatic steatosis caused by deletion of miR‐30a‐5p (Figure 7B). The levels of total cholesterol, triglyceride, LDL‐c, and HDL‐c were also proved the results (Figure 7C). Additionally, serum levels of ALT and AST decreased after AAV‐OE‐treatment, total bile acids and cholinesterase did not show significant changes (Figure 7D). Above results showed that reintroducing miR‐30a‐5p reversed significantly hepatic steatosis from HFD‐treated KO mice. The results of immunohistochemistry directly proved that the expression of ALOX5, ALOX12, and COX2 decreased significantly after restoring the expression of miR‐30a‐5p (Figure 7E–G). The expression of ALOX5, ALOX12 (Figure 7H), and COX2 (Figure 7I) at the transcription level also proved this conclusion. Therefore, reintroducing miR‐30a‐5p could reverse hepatic steatosis and arachidonic acid metabolism disorder caused by HFD and miR‐30a‐5p deletion.

The 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.