Cooked meat-derived extracellular vesicles ssc-miR-1 induces metabolic disorders in the mice liver via PI3k/AKT pathway

Abstract

Diet derived extracellular vesicles (EV) can be absorbed and influence physiological processes. Herein, this study investigates the potential health impacts of extracellular vesicles in meat products. Experimental results showed the extracellular vesicles derived from cooked red meat (RM-EV) and white meat (WM-EV) were successfully isolated via ultra-high speed centrifugation. The median particle size of RM-EV is 137.5 nm and that of WM-EV is 116.1 nm. RM-EV and WM-EV are added to the drinking water of mice for ten weeks. The mice developed insulin resistance and abnormal lipid metabolism in the liver. The influences of RM-EV are more pronounced for mice than WM-EV. High-throughput sequencing indicated that ssc-miR-1 (52.05%) and scc-miR-133a-3p (14.21%) are the most abundant microRNA in RM-EV. However, the highest content of microRNA is gga-miR-133a-3p (44.16%) in WM-EV, followed by gga-miR-1 (21.58%). Therefore, the most abundant ssc-miR-1 in red meat EV (RM-EV-miR-1) was selected for in vitro studies. In vitro experiments revealed that RM-EV-miR-1 inhibited cell proliferation by targeting IGF1 in NCTC1469 liver cell. Besides, RM-EV-miR-1 exacerbates insulin resistance in NCTC1469 insulin resistance cell model by targeting PI3K/AKT signaling pathway. Our findings show that the differential health impacts of red and white meat may be partially attributed to the presence of EV. These results provide novel insights into dietary health from the perspective of EV.

Introduction

Meat is a main source of protein in food consumption and contributes important nutrients for humans. Meat is commonly categorized into two major types: red meat (e.g., pork and beef) and white meat (e.g., chicken and fish). Over the past few decades, the global consumption of meat has increased steadily. There has been prolonged controversy as to which meat is the more healthily, red meat or white meat. Some researchers suggested that there is no substantial difference in nutritional value between red meat and white meat¹. However, numerous studies have shown a connection between the consumption of red or white meat and an increased risk of disease^2,3. The consumption of red meat is increasing globally and it is a widely popular food group⁴. According to the World Health Organization (WHO) cancer agency report of 2015, red meat has been classified as “Likely to be Carcinogenic to Humans”⁵. Available studies reported that high intake of red meat was associated with higher risk of gastric cancer⁶, bladder cancer⁷, colorectal cancer⁸, breast cancer, lung cancer⁹, etc. Conversely, the increase of white meat consumption may reduce the risk of cancer⁶. In overall, most of studies involved meat consumption show red meat are at greater health risk than white meat^6,10,11. Although these studies reach similar conclusions, most of these are based on meta‑analysis methods and lack of direct experimental evidence.

Meat contains a variety of necessary nutrients for humans such as high-quality protein, vitamins, amino acids, minerals, etc. Some investigators have devoted efforts to find which cargo in red or white meat is unhealthy for human. Red meat is rich in high levels of saturated fatty acids have been reported, which is related to increased incidence of cancers and diabetes¹². Besides, red meat also contains a significant amount of heme iron, which is more easily absorbed than non-heme iron¹³. Numerous studies proved that heme iron of red meat is the key factor driving gastrointestinal tumorigenesis. It can causes hyper-proliferation of intestinal cells and changes in the gut microbiota^14,15. In addition, red meat is also a particularly rich carnitine source¹⁶. The carnitine exhibits anti-inflammatory and antioxidant effect. However, consumption of red meat has been associated with increased cardiovascular disease risk, which may be related to blood levels of L-carnitine¹⁷. Carnitine-rich red meat enhances atherosclerosis risk was also found in mice models¹⁶. Further, our research group conducted a interesting study in previous. We extracted extracellular vesicles (EV) from cooked pork and administered it to mice by drinking¹⁸. Our preliminary experimental results demonstrated that EV from red meat (prok) may cause metabolic disorder in mice. However, it is not known whether difference between red meat and white meat in EV.

EVs are nanoscale vesicles with a bilayer lipid membrane structure secreted by various cells and play an important role in cell communication^19,20. EV contain proteins, lipids, nucleic acids and other biologically active substances and has been reported to be involved in a diverse biological processes such as apoptosis²¹, inflammatory responses²², cancer development²³, tissue repair²⁴. Recently, the dietary derived EV have attracted much attention. Several studies have reported bovine milk EV participates in multiple biological processes for consumers. EVs in bovine milk can alter the intestinal microbial community of mice, indicated participate in the interactions between bacteria and animal hosts²⁵. In addition, bovine milk EV are associated with the purine metabolism in mice liver²⁶ and intestinal inflammation²⁷. Muscle, as important organ for animals and also a significant source of food for human, are able to produce EV²⁸. The isolation and identification of exosomes from porcine extensor digitorum longus (EDL-EXO) and soleus (SOL-EXO) muscles have been reported²⁹. SOL-EVs promoted, while EDL-EVs inhibited, lipid accumulation in LD intramuscular adipocytes in vitro²⁹. However, there is still a lack of research on exosome in meat derived from food sources.

In this study, we identified a EV in red meat (pork) and white meat (chiken). This study reported distinct microRNA transcriptional landscape in EVs derived fromm red meat coompared to white meat.The health effects of connsuming exosomes derived from red or white meat were evaluated in mic. Furthermore, red meat-derived exosomal microRNA was shown to inhibit proliferation and insulin signaling pathway in NCTC1469 liver cells in vitro. Our findings contribute to the understanding of the role of EV in diet.

Results

Extraction and identification of the extracellular vesicles derived from red and white meat

Extracellular vesicles (EV) were isolated use ultrahigh-speed centrifugation method to extract the from red and white meat respectively, according to the protocol previously established by our research group (Fig. 1A). Transmission electron microscopy (TEM) revealed that the nanovesicles with EV features were extracted from cooked red meat (RM-EV) and white meat (WM-EV) (Fig. 1B, C). Western blot analysis also verified successful isolation of EV as confirmed by the EV markers CD63, CD81 and TSG101 (Fig. 1D). The media size of RM-EV particles (137.5 nm) and WM-EV particles (116.1 nm) were determined by nanoparticle tracking analysis (NAT).

Fig. 1. Isolation and identification of exosomes from cooked red and white meat.

A Flow chart for extraction of exosomes from cooked red (RM-EV) and white (WM-EV) meat, respectively. B Transmission electron microscopy image of RM-EV. C Transmission electron microscopy image of WM-EV. D Western blot of exosome marker protein. E Nanosight tracking and particle size distribution of RM-EV. F Nanosight tracking and particle size distribution of WM-EV. The figure was created using Adobe Illustrator.

Supplementation of cooked meat-derived EV cause liver metabolic disorder in mice

On the basis of our previous findings¹⁸, EV extracted from red and white meat were added to drinking water for mice to take. The body weight of mice supplemented with EV (RME & WME group) increased significantly from 1 week to 10 weeks (Fig. 2A). The increase of boy weight was more obvious in RME mice compared with the WME mice. Mean BMI of RMW mice was significantly higher than in WME and CON mice (Fig. 2B). Meanwhile, the number of running laps per minute in RME mice was also significantly reduced (Fig. 2C). We further examined glucose metabolism function in mice. The results showed the blood glucose of RME and WME mice were elevated significantly after intraperitoneal injection of glucose (Fig. 2D). Insulin tolerance test showed blood glucose level of RME and WME mice decreased slowly, indicated a reduction in insulin sensitivity inRME and WME mice (Fig. 2E). Organ weights were assessed after the feeding period. The results showed there was a slightly rise in the liver index in RME and WME group, as compared to the CON group (Fig. 2F). Oil-Red O staining revealed the liver of RMW and WME mice accumulated lipid droplets, RMW mice was particularly apparent (Fig. 2G, H).

Fig. 2. Metabolic disorders in mice induced by supplementation with exosomes derived from cooked meat.

A Relative body weight changes of mice at different time points. B Body mass index (BMI) of mice. C The number of laps per minute the mice ran on the wheel. D Glucose tolerance test (GTT, n = 4). E Insulin tolerance test (ITT, n = 4). F Relative organ index (organ weight/body weight). G Relative quantification of the Oil Red O staining images (n = 3). H Oil red O staining of liver sections. Results are expressed as mean ± standard deviation. n = 8 in each group unless specified. *P < 0.05, **P < 0.01 between indicated groups. ^#P < 0.05 between WME and CON, ^$P < 0.05 between RME and WME. The figure was created using Adobe Illustrator.

EV supplementation changes liver transcriptome

The transcriptome changes of the mice liver were investigated. There were 833 upregulated genes and 412 downregulated genes observed in the RME group as compared to the CON group (Fig. 3A). In WME group, 801 genes were upregulated and 621 genes were downregulated compared to the CON group (Fig. 3C). The Venn diagram show the intersection of these differentially expressed genes (DEGs) (Fig. 3B). Among these DEGs, 521 upregulated and 217 downregulated genes were common to RME and WME. The heat map presented the expression level of the DEGs between groups (Fig. 3D, E). We performed the KEGG enrichment analysis of the above DEGs. In RME group, the upregulated genes were mainly enriched in Rap1 signaling pathway, cell adhesion molecules, Ras signaling pathway, whereas surprisingly the downregulated genes were enriched in oxidative phosphorylation, metabolic pathways, diabetic cardiomyopathy, non-alcoholic fatty liver disease (Fig. 3F, G). In WME group, the upregulated genes were also enriched in Rap1 signaling pathway, Ras signaling pathway, metabolic pathway and insulin resistance, and the downregulated genes were enriched in glycine, serine and threonine metabolism, Notch signaling pathway, Wnt signaling pathway (Fig. 3H, I).

Fig. 3. Changes of transcriptome in the mice liver.

A Volcano plot showing differential expressed genes (DEGs) between RME and CON. B Venn diagram of unique and shared DEGs between RME and WME groups. C Volcano plot showing differential expressed genes between WME and CON. Down, downregulated genes; Up, upregulated genes; No diff, not differentially expressed genes. D, E Heat map of differential gene expression profiles. KEGG enrichment analysis for upregulated genes in RME F, downregulated genes in RME G, downregulated genes in WME H, upregulated genes in WME I. The figure was created using Adobe Illustrator.

Characteristics of microRNA expression profiling of EV

The EV contain a diverse cargo composed of proteins, lipids, and nucleic acids, in which microRNA (miRNA) has now been reported to be involved in the multiple functions^28,30. We used high-throughput sequencing to establish microRNA expression profiles in RM-EV and WM-EV. We identified 315 microRNA in RM-EV and 462 microRNA in WM-EV (Fig. 4A). The sequences of 59 microRNA were completely conserved between RM-EV and WM-EV. ssc-miR-1/gga-miR-1a-3p was highest abundance conserved miRNA for both. The top 7 conserved miRNA were same between RM-EV and WM-EV (Fig. 4B). Ranked dot plots revealed most conserved miRNA were in top-ranked (Fig. 4C, D).

Fig. 4. MicroRNA profiles of RM-EV and WM-EV.

A Venn diagram of the number of miRNAs, and intersection part between circles represent the number of conserved miRNAs. B The abundance of the top 15 conserved miRNA. C, D The ordination diagram based on the abundances of miRNAs. The top 10 miRNAs was shown in the box. E Abundance and sequence comparison of top 2 miRNAs between RM-EV and WM-EV. F, G KEGG enrichment analysis of top 2 miRNAs. The figure was created using Adobe Illustrator.

The top-two miRNAs with the highest relative abundance in RM-EV were ssc-miR-1 and ssc-miR-133a-3p, which accounted for 52.05% and 14.21%, respectively. While the top-two miRNAs in WM-EV were gga-miR-133a-3p and gga-miR-1a-3p, which accounted for 44.16% and 21.58%, respectively (Fig. 4E). The ssc-miR-1 sequence was completely consistent with the gga-miR-1a-3p sequence. While ssc-miR-133a-3p and gga-miR-133a-3p has only one base difference (Fig. 2E). The KEGG functional enrichment analysis results showed ssc-miR-1/gga-miR-1a-3p was mainly enriched in PI3K-Akt signaling pathway, Ras signaling pathway, MAPK signaling pathway (Fig. 4F). Ssc-miR-133a-3p/gga-miR-133a-3p was mainly enriched in c-type lectin receptor signaling pathway, Rap1 signaling pathway, metabolic pathway (Fig. 4G).

3.5 MicroRNAs derived from cooked meat EV inhibited NCTC1469 cell proliferation

The NCTC1469 liver cells could were able to ingest red meat derived EV (Fig. 5A). RM-EV-miR-1 was the most abundant microRNA in red meat derived EV. RM-EV-miR-1 was predicted to potentially target IGF1 mRNA in NCTC1469 cells (Fig. 5B). Dual-luciferase reporter assay showed that RM-EV-miR-1 mimics significantly reduced luciferase activity of IGF1-3’UTR-WT group (Fig. 5C). The expression of RM-EV-miR-1 in the cells was significantly upregulated after NCTC14699 cell lines were transfected with mimics, and IGF1 mRNA expression was significantly reduced (Fig. 5D). The CCK-8 assay showed that transfection with RM-EV-miR-1 mimic significantly inhibited NCTC1469 cell proliferation (Fig. 5E). EdU staining revealed that compared with the control group, the EdU-positive rate in the RM-EV-miR-1 group was significantly reduced (Fig. 5F, G). The result of RT-qPCR showed RM-EV-miR-1 can significantly inhibit the expression of proliferation-related gene CCNB1, CCND1, CDK2, CDK4 and PCNA mRNA (Fig. 5H).

Fig. 5. RM-EV-miR-1 inhibited NCTC1469 liver cell proliferation.

A Exosome uptake was observed by fluorescence microscopy. RM-EV-PKH67 represents red meat derived-exosomes connected by PKH67, bar = 10 μm. B Predicted binding site of RM-EV-miR-1 in the 3′-UTR of IGF1. C Relative luciferase activity detected by dual luciferase assay. WT, wild-type. MT, mutant. D Transfection efficiency was measured using RT-qPCR. NC, negative control, untransfected cells. MC, mimics control. RM-EV-miR-1, ssc-miR-1 mimics. E CCK8 to detect cell proliferation. F, G EdU to detecte cell proliferation and analysis of EdU-positive cells, bar = 50 μm. H The mRNA expression levels was detected using RT-qPCR. The results were expressed as the mean ± SEM, n = 3. *P < 0.05, **P < 0.01. The figure was created using Adobe Illustrator.

RM-EV-miR-1 exacerbates NCTC1469 liver cell insulin resistance by targeting PI3K

RM-EV-miR-1 was predicted to bind to the PI3K 3′UTR. GSEA results showed that changes in PI3K-AKT and insulin signaling pathways (Fig. 6A). The targeted binding of RM-EV-miR-1 and PI3K was verified by dual-luciferase reporter assay (Fig. 6B, C). Through palmitic acid treatment of NCTC1469 cells, an in vitro insulin resistance cell model was constructed and further used to explore the function of RM-EV-miR-1 (Fig. 6D). RM-EV-miR-1 were transferred into insulin resistance NCTC1469 cell model (IR-NCTC1469) (Fig. 6E). RM-EV-miR-1 could inhibit the expression of RI3K, AKT and GLUT4 mRNA level in IR-NCTC1469 (Fig. 6F–I). The residual glucose concentration in the medium was increased after RM-EV-miR-1 treatment (Fig. 6J). We also tested 2-NBDG uptake in IR-NCTC1469 cell, and result showed a reduction in RM-EV-miR-1 treated cell (Fig. 6K). The expression of PI3K/AKT pathway-related proteins was further examined (Fig. 6M). It was found that after RM-EV-miR-1 treatment, the levels of p-PI3K, p- AKT, and p-GSK-3 were significantly decreased (Fig. 6N–P). In addition, GLUT4 expression was decreased in RM-EV-miR-1 treated-IR NCTC1469 cells and immunofluorescence showed the membrane translocation of GLUT4 was inhibited (Fig. 6Q, R). PAS staining showed reduction of glycogen deposition after RM-EV-miR-1 treatment (Fig. 6L).

Fig. 6. RM-EV-miR-1 exacerbates insulin resistance of NCTC1469 liver cell.

A The gene set enrichment analysis (GSEA) for mouse liver transcriptome data. B Predicted binding site of RM-EV-miR-1 in the 3′ UTR of PI3K. C Relative luciferase activity detected by dual luciferase assay. D The detailed grouping strategies. E The RM-EV-miR-1 qPCR products were checked by gel electrophoresis. F–I RT-q-PCR result of PI3K, AKT, PI3K, GSK-3β and GLUT4 mRNA expression. J The content of remaining glucose in culture media. K 2-NBDG uptake in NCTC1469 cells, bar = 50 μm. L Periodic acid-Schiff (PAS) staining for NCTC1469 cells, bar = 20 μm. M Western blot images. N–Q The relative protein levels were quantified. R Imunofluorescence detectionof GLUT4 expression. Red fluorescence represents cell membrane with DiI staining, bar = 10 μm. The results were expressed as the mean ± SEM, n = 3. *P < 0.05, **P < 0.01. The figure was created using Adobe Illustrator.

Activation of PI3K relieved RM-EV-miR-1 exacerbated NCTC1469 cells insulin resistance

To determine whether the RM-EV-miR-1 was mediated through PI3K, we further tested the effects of PI3K agonist 740Y-P (Fig. 7A). The residual glucose level in the medium was reduced by 740 Y-P treatment (Fig. 7B). The 2-NBDG uptake assay demonstrated that 740 Y-P enhanced the glucose uptake capacity of RM-EV-miR-1 treated NCTC1469 cells (Fig. 7C). Addition of 740 Y-P reversed RM-EV-miR-1-induced inhabit of mRNA of PI3K/AKT signaling pathway in varying degrees (Fig. 7D–G). The PI3K agonists 740 Y-P partially restored the suppression of the protein expression levels of p-PI3K, p- AKT, and p-GSK-3 (Fig. 7H–K). Furthermore, translocation GLUT4 protein in IR NCTC1469 cell membrane were partial recovery after 740 Y-P treatement (Fig. 7L, M).

Fig. 7. NCTC1469 cells were treated with the PI3K activator 740Y-P.

A The RM-EV-miR-1 qPCR products were checked by gel electrophoresis. B The content of remaining glucose in culture media. C 2-NBDG uptake in NCTC1469 cells, bar = 50 μm. D–G RT-q-PCR result of PI3K, AKT, PI3K, GSK-3β and GLUT4 mRNA expression. H Western blot images. I–L The relative protein levels were quantified. M Imunofluorescence detectionof GLUT4 expression. Red fluorescence represents cell membrane with DiI staining, bar = 10 μm. The results were expressed as the mean ± SEM, n = 3. *P < 0.05, **P < 0.01. The figure was created using Adobe Illustrator.

Discussion

Extracellular vesicles (EV) can be released from cell types from almost all eukaryotes and prokaryotes³¹. EV contain a variety of bioactive components and were one of the important mechanisms for intercellular communication^20,31. Food composition was the important factor regulating the metabolism of the body and thus affecting the health. Recent studies have reported dietary-derived EV could be directly absorbed by gut cells and travel through the bloodstream. To date, many researchers have successfully isolated EV have successfully from daily food. For example, the study group of Cho et al. isolated and identified a similar EV from apple³², Yam³³, Plum³⁴. In addition, milk as an important source of everyday nutrition has been extensively studied for its EV³⁵. However, EV extracted from meat have not been reported before. Our group have identified for the first time EV from cooked meat¹⁸. In this study, meat-derived EV were extracted by ultra-high speed centrifugation under 16,000 × g. At present, ultra-high speed centrifugation was still the mainstream method for EV extraction³⁶. We also identified the characteristic marker CD63,CD81 and TSG101 of EV by Western blot. This demonstrated the cooked meat-derived nanovesicles were EV.

We previously extracted EV from pig muscle, fat, and liver separately. These were common meat for Chinese consumer. Recently, it aroused our interest to explore whether or not there were differences in EV derived from different meat. Which was healthier to eat red meat or white meat was a concern for people. A lot of researches have indicated that consumption of red meat were closely related to cancer⁹, cardiovascular disease², nonalcoholic fatty liver disease³⁷. In addition to essential nutrients such as protein, vitamins, minerals, red meat contain saturated fatty acid, cholesterol and heme iron, which were closely related to the occurrence of diseases. Here, we attempted the isolation as EV from pork and chicken breast. Pork and chickens were the most consumed red meat and white meat in the world, respectively^38,39. The morphology of EV extracted was visualized through TEM analysis. The results of particle size analysis showed that the media size of red maet-derived EV (RM-EV) (137.5 nm) larger than white meat-derived EV (WM-EV) (116.1 nm). Bae et al. reported the EV extracted from milk were found to have a diameter of 80–190 nm⁴⁰. While the size of plum-derived EV was 211 nm³⁴ and of strawberry was from 30 to 191 nm⁴¹. The size of the EV can vary between different species. Moreover, size of EV was also associated with isolation method and sample size.

The cargo in EV consisted of various bioactive components including proteins, lipids and nucleic acids⁴². EV nucleic acids have been studied, including mRNA and non-coding RNA such as miRNAs, lncRNAs, circRNAs. Among them, miRNAs were short non-coding RNA approximately 22 nt, negatively regulating expression of targeted genes at the post-transcriptional level. EV miRNA was involved in the regulation of cellular functions had been widely demonstrated⁴³. Thus, we evaluated the miRNA expression profile in RM-EV and WM-EV through miRNA sequencing. Six of the top ten miRNAs have completely conserved sequences between RM-EV and WM-EV. This implied those miRNAs might participate in EV-mediated functional regulation. Interestingly, we found that miR-1 and miR-133a-3p was abundantly expressed in RM-EV and WM-EV. The microRNA miR-133a-3p previously named miR-133a⁴⁴. The miR-133a and miR-1 sequence were highly conserved throughout multiple species^45,46. MiR-1 and miR-133a also were abundantly expressed in muscle tissue and were co-transcribed miRNAs during myogenesis⁴⁷. This suggested that miRNAs in RM-EV and WM-EV were produced by muscle cells. ssc-miR-1 was the most abundant in RM-EV, accounted for 52.05%, while gga-miR-133a-3p was the most abundant in WM-EV, accounted for 44.16%. Functional enrichment analysis showed that ssc-miR-1 was mainly enriched in PI3K/AKT signaling pathway and gga-miR-133a-3p was mainly enriched in C-type lectin receptor signaling pathway. This suggested that RM-EV and WM-EV may exert different biological functions.

We were curious about whether the difference in miRNAs in meat-derived EV was part of the reason why red meat and white meat have different health effects. Then, RM-EV and WM-EV were added to the drinking water of ICR mice, according to the our previous strategy¹⁸. By using the everted gut sac method, we have previously demonstrated that the miRNA in meat-derived EV can penetrate the intestine¹⁸. It has been reported diet-derived plant miRNAs also can be taken up by the intestine cells and enter the circulation⁴⁸. In this study, mice supplemented with RM-EV and WM-EV have increased weight and BMI. Moreover, experimental group mice exhibit insulin resistance and increased hepatic lipid deposition. This indicates that meat-derived EV lead to metabolic disorders for mice. Interestingly, we found that the effects of RM-EV were more evident than WM-EV on the mice metabolic disorders. Zhu et al. reported pork proteins caused increased malondialdehyde productionand higher hepatic oxidative stress for rat than fish, chicken proteins⁴⁹. In addition, there was also study that suggest individuals with high intake of red meat have an increased risk of developing nonalcoholic fatty liver disease (NAFLD)³⁷. Similar results have been reported by Kita et al.⁵⁰, high consumption of red meat was associated with NAFLD and insulin resistance. Heterocyclic amines in red meat were associated with insulin resistance⁵⁰. In this study, the results of the liver transcriptome analysis showed there were 833 up-regulated genes and 412 down-regulated genes in mice supplemented RM-EV. These genes were significantly involved in lipid and atherosclerosis, PPAR signaling pathway, metabolic pathways and nonalcoholic fatty liver disease. These results all illustrate the red meat have detrimental effects on liver health. Our results proved that EV in red meat were one of the harmful factors.

In vitro experiments, we demonstrated that RM-EV labeled with PKH67 can be effectively internalized by NCTC1469 liver cells. Based on the differences in miRNAs between RM-EV and WM-EV, we selected RM-EV-miR-1 for further study in NCTC1469 liver cell in vitro. It is well known that miRNA work by combining seed sequences with target genes mRNA⁵⁰. We identified potential binding sites between RM-EV-miR-1 and the 3’UTR region of IGF1 mRNA. Furthermore, subsequent dual-luciferase reporter assays validated their binding relationship. IGF1 serves as a key mediator of growth hormone action, thereby exerting significant influence on cell proliferation and differentiation⁵¹. Previous studies have reported that miR-1 regulates the proliferation of vascular smooth muscle cells⁵², L-O2 hepatocytes⁵³, and colorectal cancer cells⁵⁴ by targeting IGF1. Consistent with these findings, our results also revealed that RM-EV-miR-1 downregulated the mRNA levels of IGF1 in NCTC1469 cells, accompanied by a decrease in the mRNA levels of cell proliferation-related marker genes, including CCNB1, CCND1, CDK2, CDK4, and PCNA. Based on the results of previous animal experiments, we observed that RM-EV caused abnormal liver metabolism in mice. Analysis of the mouse liver transcriptome data revealed that differentially expressed genes were enriched in insulin signaling and PI3K signaling pathways. Interestingly, our predictive analysis suggested a potential binding relationship between RM-EV-miR-1 and PI3K. Subsequently, dual-luciferase reporter assays confirmed that RM-EV-miR-1 can target PI3K. Wu et al. reported that miR-1 inhibits the PI3K/AKT/mTOR signaling pathway, effectively improving myocardial injury⁵⁵. Additionally, miR-1 plays a crucial role in the pathogenesis of non-small cell lung cancer through the PI3K/Akt pathway⁵⁶. The PI3K signaling pathway serves as one of the crucial pathways for insulin to regulate cellular metabolism and function⁵⁷. The insulin signaling pathway plays a significant role in regulating glucose metabolism, lipid metabolism, and protein metabolism in the liver⁵⁸. Consequently, we established an in vitro insulin resistance model based on previous reports^59,60 to investigate the regulatory role of RM-EV-miR-1.Our results showed that RM-EV-miR-1 suppressed the protein expression levels of p-PI3K, p-AKT, p-GSK-3β and GLUT4 by targeting PI3K, thereby exacerbating insulin resistance in NCTC1469 cells. Consequently, the glucose metabolism level of NCTC1469 cells was significantly reduced. Subsequently, we treated NCTC1469 cells with RM-EV-miR-1 in combination with the PI3K agonist 740 Y-P and found that 740 Y-P partially restored the impairment of the PI3K/AKT signaling pathway in NCTC149 cells caused by RM-EV-miR-1. This suggests that RM-EV-miR-1 affects insulin signaling transduction in NCTC1469 cells through PI3K.

Combining the results of animal experiments and in vitro cell experiments, we suggest that exosomal microRNAs derived from red meat may be involved in the metabolic regulation (Fig. 8). It is noteworthy that a limitation of this study lies in the fact that EV contain various bioactive substances. Although microRNAs play a significant role among them, other factors may also contribute to their effects. Furthermore, the diet is a complex and dynamic system, and its impact on health is a comprehensive effect of multiple factors. Our study provides a novel perspective for understanding this crucial system of diet.

Fig. 8. Graphic abstract of this study.

The figure was created using Adobe Illustrator.

In summary, this work successfully isolated EV from red meat (pork) and white meat (chicken). The diferences of miRNAs were detected between RM-EV and WM-EV. The myogenic miR-1 and miR-133a-3p were the main miRNAs of RM-EV and WM-EV. The supplement of RM-EV and WM-EV caused insulin resistance and increased lipid deposition in mice liver. The influence introduced by RM-EV was more pronounced compared to WM-EV. In vitro, RM-EV-miR-1 suppressed cell proliferation bytargeting IGF1 in NCTC1469 liver cells. In the insulin-resistant NCTC1469 cell model, RM-EV-miR-1 exacerbates cellular insulin resistance by suppressing the PI3K/AKT signaling pathway. This work suggested that red meat-derived EV may be one of the unfavorable environmental factors influencing health. Moreover, this study was worthy of further exploration. The meat-derived EV comprise a variety of substance (such as lipids, proteins, mRNAs and various noncoding RNA). The EV like a Pandoras box and which of them are detrimental and which are beneficial are worthy of further research.

Methods

Extracellular vesicles isolation and observation

We attempted to extracte extracellular vesicles (EV) from red meat and white meat according to an established method developed by our group¹⁸. The main steps are described below. Fresh pork loin and chicken breast were purchased form a local market. The meat was minced to a paste in a blender. Add 250 g of minced meat and 250 g of water to the pot. Meat boiled for 1 h until cooked completely. The sample was filtered with a filter paper and cooled to room temperature. Sample was centrifuged at 2000 × g for 30 min and the supernatant was collected and centrifuged again (12,000 × g, 45 min). Then, the sample was filtered with 0.22 μm filter. The samples were centrifuged at 160,000 × g for 2 h by an ultra-high-speed centrifuge to collect precipitatesand resuspended whit phosphate-bufered saline (PBS). The centrifugation was repeated at 160,000 × g for 1 h and resuspended to get EV suspension. EV samples were viewed using a transmission electron microscope (TEM, HT7800, Hitachi, Japan). Samples were processed as described previously⁶¹. EV size was measured on the ZetaView (Particle Metrix, Germany) with nanoparticle tracking analysis (NTA).

Animal feeding and management

ICR male mice were purchased from DaShuo biotechnology company (Chengdu, China). Mice were given access to water and food ad libitum. All experiments in the present study were conducted in accordance with the Regulations on the Administration of Laboratory Animals (Ministry of Science and Technology of China). Nutritional composition of the diets were as follows: 9.3% water + 13.4% protein + 14.3% fat + 23.5% sugar + 2.7% fiber + 4.4% ash + 0.8% Ca + 0.7% P. Mice were assigned to three groups. One group received red meat-derived EV in drinking water (RME, n = 8). One group received white meat-derived EV in drinking water (WME, n = 8). In addition, control group was treated with saline solution (CON, n = 8). A quantity of EV equivalent to the quantity contained in 5 g of meat were provided daily to each mice. The mice were deprived from water for 2 h before EV supplementation. EV isolated from every 5 g of meat were resuspended in 1 mL of drinking water and administered to the mice to ensure complete ingestion. EV supplementation per day from 9.00 AM to 10.00 AM. The experiment lasted for 10 weeks. After the feeding experiment, mice were placed in wheels and the number of laps per minute was recorded. The body mass index (BMI) was calculated as body weight (kg)/body surface area (m²). Body surface area was calculated using the DuBois equation⁶². Glucose tolerance tests (GTT) and insulin tolerance tests (ITT) was performed according to previous study⁶³. All mice were anesthetized by inhaling 3% isoflurane and then euthanized by neck breaking. Organs were collected immediately and weighed. The samples were frozen and stored at −80 °C until used.The paraffin-embedded liver tissues were cut into 4 μm paraffin sections and stained with hematoxylin and Oil Red O for microscopic observation. Image-Pro Plus 6.0 software was used to to analyze the image.

RNA extraction and library construction

Total RNA of EV and liver were extracted using Trizol LS (Ambion, Carlsbad, CA, USA) reagent according to the manufacturer’s instructions. Small RNA sequencing libraries were prepared using TruSeq Small RNA Sample Prep Kits (Illumina, San Diego, USA). Liver transcriptome library preparation was performed using TruSeq stranded mRNA library preparation kit with poly A selection (Illumina, San Diego, USA). Libraries were sequenced using Illumina HiSeq2000/2500 platform.

RNA sequencing data analysis

RNA sequence data processing and analysis were referred to in previous study^64,65. Raw data were first filtered to remove low quality reads. The clean reads were aligned to the pig reference genome (Sscrofa 11.1) using the bowtie2 mapping tool. An analysis of differential expression was performed using DESeq2.

microRNA target gene prediction and functional enrichment analysis

microRNA target gene prediction was performed using the OmicStudio (https://www.omicstudio.cn/analysis) based on TargetScan (5.0) and Miranda (3.3a). We used the Mus musculus (Ensembl_v107) as species background. The screening threshold was set as TargetScan_score ≥ 50 and miranda_Energy < -10. Functional enrichment analysis was performed using DAVID (v2022q3, https://david.ncifcrf.gov) based on Kyoto Gene and Encyclopedia of Genomes (KEGG) enrichment analysis.

Cell culture and treatment

The NCTC1469 murine liver cell line were purchased from Procell Biotechnology Co., Ltd (Wuhan, China). NCTC1469 cells were cultured in DMEM (Gbico) supplemented with 10% FBS (BI) and 100 U/mL penicillin/streptomycin at 37 °C with 5% CO₂. Palmitic acid (PA, 0.2 mM, Shanghai Siduorui Biotechnology Service Co., Ltd, China) were used to treat the NCTC1469 cells to establish insulin-resistant cell model according to the previous method^59,60. 740 Y-P (1 µM, Y413859, Aladdin) was added into NCTC1469 cells for activation of the PI3K pathway.

EV uptake

The cellular uptake of cooked meat derived EV were observed using fluorescence microscopy. EV solution was mixed with PKH67 (Umibio, Shanghai, China) dye solution according to the manufacturer’s instructions. Labeled EV were coincubated with NCTC1469 cells for 3 h. The cells were fixated with 4% paraformaldehyde. Nucleus was stained with DAPI (TargetMol, USA) and observed by microscope.

Cell transfection

Synthetic oligonucleotides RM-EV-miR-1 mimics and mimics control were purchased from Genepharma (Shanghai, China). Transfection of miRNA mimics was conducted using Lipo 3000 transfection agent (Invitrogen) according to the manufacturers instruction. After 24 h of mimics transfection, transfection efficiency was measured by RT-qPCR and nucleic acid electrophoresis.

Luciferase reporter assay

HEK293T cells were transfected with pmirGLO dual luciferase vectors containing wild-type IGF1 3′ UTR and mutant IGF1 3′ UTR, respectively. The pSiCheck2 dual luciferase vectors containing wild-type PI3K 3′ UTR and mutant PI3K 3′ UTR were transfected to HEK293T cells. Cells were transfected with RM-EV-miR-1 mimics or mimics control using Lipofectamine 3000 transfection reagent. Cells were harvested 36 h post transfection. The luciferase reporter assay was performed using a dual luciferase reporter assay kit (11402ES, Yeasen, Shanghai, China).

Cell proliferation assay

NCTC1469 cells were seeded on 96-well plates and they were used for transfection when cell fusion reached up to 30%. CCK8 reagent (Beyotime, China) was added to the plates at 24, 48, 72 h post-transfection. After 1 h incubation, the absorbance was measured at 450 nm. The EdU (CA1174, Solarbio, China) assay was performed 48 h after NCTC1469 cells were transfected with mimics according to the manufacturers instructions.

Measurement of glucose level

The medium was replaced with new medium 48 h after PA treatment and RM-EV-miR-1 transfection. Media were collected at 6 h and measured for glucose content. Glucose kit (hexokinase method) purchased from Nanjing Jiancheng Bioengineering Institute (Nanjing, China) was applied for experiments.

2-NBDG uptake assay

After treatment, cells were incubated with 100 nm insulin (Ye-Yuan, Shanghai, China) for 30 min. Next, the cells were further incubated with 20 μM 2-NBDG (Macklin, Shanghai, China) for 30 min. The cells were treated with precooling PBS to terminate the reaction. Cells were counterstained with Hoechst to detect cell nuclei. The 2-NBDG uptake was observed under a fluorescence microscopy.

Periodic acid-Schiff (PAS) and immunofluorescence staining

After 48 h of treatments, cells were fixed with fresh 4% paraformaldehyde. For PAS staining, glycogen was detected by PAS assay using PAS Kit (Nanjing Jiancheng Bioengineering Institute, Nanjing, China). Cells were observed with microscope under bright field. For immunofluorescence staining, cells were blocked with 10% goat serum in PBS for 1 h. Primary antibodies were incubated for 2 h at room temperature. Then cells were incubated with goat anti-rabbit FITC secondaryantibody (KGC6214, KeygenBioTECH) for 1 h. Cell membrane were stained with Dil (abs47048140, Absin) after nucleus staining with DAPI. The following antibodies were used for immunofluorescence: GLUT4 (1:200, Wanleibio, WL02425), TRITC Conjugated Goat Anti-Rabbit IgG H&L Goat Polyclonal Antibody (1:200, HUABIO, HA1016).

Western blotting

Western blotting was performed using standard methods. In brief, cells were washed with phosphate-buffered saline and lysed using RAPI lysis buffer (Beyotime, China) supplemented with a 5% 100 mM PSMF (AP02L015, Life-iLab, China) and phosphatase inhibitor (Wanleibio, China). Protein was separated on 10% SDS PAGE (M00666, GenScript, China) and transferred onto PVDF membrane. The membrane was blocked in 5% skim milk and incubated with primary antibody at 4 °C overnight followed by incubation with secondary antibody for 2 h. Finally, enhanced chemiluminescence (P10060, New Cell & Molecular Biotech) systems were used to detect the bands. The following antibodies were used: PI3K (1:1000, Wanleibio, WL02240), AKT (1:1000, Wanleibio, WL0003b), p-AKT (1:1000, Wanleibio, WLP001a), GSK-3β (1:1000, Wanleibio, WL01456), p-GSK-3β (1:1000, Wanleibio, WL03518), GLUT4 (1:1000, Wanleibio, WL02425), p-PI3K (1:1000, Beyotime, #AF5905), β-actin (1:3000, Zenbio, #380624), Anti-rabbit IgG, HRP-linked Antibody (1:2000, Cell Signaling, #7074).

Reverse transcription quantitative real-time PCR

Total RNA was extracted with the Trizol reagent from NCTC1469 cells. Then, the total RNA was reverse transcribed into cDNA using a Takara PrimeScript II 1st strand cDNA Synthesis Kit (Takara). microRNA reverse transcription was carried out using MiR-XTM miRNA First-Strand Synthesis kit (Takara). qPCR was performed using SYBR Premix Ex Taq kit (TaKaRa). The relative gene expressions were calculated using the 2^−ΔΔCt method. ACTB and U6 were used as internal reference. RT-qPCR primers sequences were shown in a supplementary file.

Statistical analysis

Data were visualized in GraphPad Prism 8 software. All results were shown as mean ± SEM. Statistical significance of the mean values were evaluated by Students t-test or one-way ANOVA. A value of P < 0.05 was considered significant and a P value < 0.01 was considered highly significant.

Author contributions

L.S., J.M., and S.L. designed the overall experiment and were the major contributors to writing the manuscript. Y.Y. and T.L. completed bioinformatics data analysis. S.L., Y.L., and L.C. conducted animal experiments and testing. J.M., Y.W., and L.N. conducted cell validation experiments. Y.Z. and L.S. prepare the main figures in the manuscript. M.G. and L.Z. provide project funding and revise the original manuscript. All authors reviewed the manuscript.

Data availability

All data generated or analysed during this study are included in this published article.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41538-026-00709-7.