Skip to main content
NCBI home page
Springer logoLink to Springer
. 2026 Jan 13;42(1):19. doi: 10.1007/s10565-025-10132-5

BMSC-derived extracellular vesicles affect gluconeogenesis and lipogenesis by releasing 5'-tRF-GlyCCC to improve MAFLD insulin sensitivity

Chenyun Yang 1,2, Huiling Chen 1,2, Xiaojing Huang 1,2, Yanyan Li 1, Song Wen 1, Ligang Zhou 1, Xinlu Yuan 1,✉
PMCID: PMC12835049  PMID: 41530468

Abstract

TsRNAs (tRNA-derived small RNAs) also known as tRNA-derived small RNAs, are a relatively new type of non-coding RNAs that have demonstrated promising effect in treating various liver diseases. However, the function of small extracellular vesicles (sEVs) secreted by human bone marrow mesenchymal stem cells (BMSCs) in safeguarding against metabolic-associated fatty liver disease (MAFLD) is still uncertain. In this research, we explored the effects of BMSCs sEVs on lipid metabolism using Palmitic Acid (PA)-induced HepG2 cells, both in the presence and absence of the sEVs inhibitor GW4869. Pandora sequencing and RNA sequencing were utilized to identify differentially expressed genes in sEVs and hepatocytes in vitro. Furthermore, we carried out in vivo studies involving male C57BL/6J mice that fed with high-fat diet (HFD) and either treated with an AAV 5'-tRF-GlyCCC mimic or not, through tail vein injection. Our findings revealed that BMSC-sEVs can relieve lipid accumulation in PA-caused HepG2 cells by inhibiting the formation of de novo fatty acid. We found that 5'-tRF-GlyCCC forms a direct connection with the 3' UTR of FoxO3, thereby decreasing the level of gluconeogenic genes PEPCK and G6Pase. Tail vein administration of the 5'-tRF-GlyCCC AAV alleviated liver gluconeogenesis and lipid metabolism issues in MAFLD mice by enhancing hepatic insulin sensitivity. The results imply that the 5'-tRF-GlyCCC/FoxO3 gluconeogenesis-signaling pathway could be crucial in the therapeutic benefits of BMSC sEVs on MAFLD.

Graphical Abstract

1. BMSC-sEVs exert therapeutic effects on MAFLD by alleviating hepatic lipid accumulation and enhancing insulin sensitivity.

2. 5′-tRF-GlyCCC, the critical functional cargo in BMSC-sEVs, directly targets the 3′-UTR of FoxO3.

3. This targeting of FoxO3 by 5′-tRF-GlyCCC suppresses the expression of gluconeogenic genes (PEPCK, G6Pase) and lipogenic genes (SREBP1c, ACC, FASN), mediating BMSC-sEVs’ anti-MAFLD effects.

graphic file with name 10565_2025_10132_Figa_HTML.jpg

Supplementary Information

The online version contains supplementary material available at 10.1007/s10565-025-10132-5.

Keywords: BMSCs, Insulin sensitivity, MALFD, Pandora sequencing, 5'-tRF-GlyCCC

Introduction

The metabolic-associated fatty liver disease (MAFLD) is the prevalent chronic liver ailment, with global prevalence of 25.24%, the incidence rate stands at 30% among Western populations. In addition, over 88% of the obese population exhibits this characteristic (Browning et al. 2004; Chaney 2021). MAFLD is marked by irregular lipid metabolism within liver cells (Younossi et al. 2018). MAFLD initially manifests as simple steatosis, with the feature of fat accumulation in the liver without significant inflammation or hepatocellular injury. As the condition progresses,  metabolic associated fatty liver (MAFL) can evolve into  metabolic dysfunction-associated steatohepatitis (MASH), and may cause severe hepatocellular damage. MASH can progress to liver fibrosis, cirrhosis, hepatocellular carcinoma (Byrne and Targher 2015). The pathogenesis of MAFLD involves the "multiple-hit" hypothesis. Some scholars found that various elements play a role in the development and advancement of MAFLD (Wang et al. 2021). Key elements to consider are the disruption of lipid metabolism in liver cells, the presence of oxidative stress,and the damage to hepatocytes.Among these, fatty liver is the critical driving factor for MAFLD (Zheng et al. 2023; Teixeira et al. 2023). Hence, it is crucial to explore new treatment strategies that target to these pathways in order to effectively manage and alleviate the disease (Kitade et al. 2017).

The human bone marrow mesenchymal stem cells (BMSCs) are a category of versatile stem cells. Their remarkable ability to self-renew and differentiate has garnered significant interest, making them a promising therapeutic option for various diseases (Wang et al. 2024a; Chen et al. 2025). Recently, scholars have increasingly investigated the potential applications of BMSCs in MAFLD (Jiang, et al. 2024; Bi et al. 2021; Afarin et al. 2023; Meneses et al. 2023). Studies suggest that BMSCs achieve their positive impacts in MAFLD mainly by releasing small extracellular vesicles (sEVs). These carriers possess bioactive molecules like cytokines, microRNAs, and growth factors (Chen et al. 2023a; Han et al. 2025). These sEVs have a crucial impact on hepatic inflammation and lipid metabolism, which in turn enhances the pathological characteristics of MAFLD (Yang et al. 2025; Zhang et al. 2024). Moreover, studies have demonstrated that BMSCs contribute to hepatocyte restoration and decrease lipid buildup, potentially mitigating liver injury and decelerating the development of fibrosis (Moayedfard et al. 2022; Damania et al. 2018; Jiang et al. 2024; Ding et al. 2023). Although there are promising results, the exact ways in which BMSC-derived sEVs achieve these effects are still not well understood. Therefore, additional investigations are required to unravel the intricate processes responsible for their effects. This is very important for maximizing the therapeutic effect of BMSCs in treating MAFLD.

tRNA-derived small RNAs (tsRNAs), which encompass tRFs and stress-induced tiRNAs, are a category of molecules. These molecules are gaining recognition as crucial regulators of cellular processes like stress response, gene expression, and protein synthesis (Chen et al. 2024; Chen et al. 2021; Li et al. 2024a; Pan et al. 2021). In the context of liver ailments, tsRNAs have been found to be essential in inflammation, fibrosis, and cellular stress responses (Ma et al. 2025; Ying et al. 2023; Huang et al. 2021). They regulate the expression of genes related to liver injury and healing, influencing pathways that control oxidative stress, apoptosis, and metabolic equilibrium. In the context of MAFLD, tsRNAs play a key role in regulating lipid metabolism and inflammatory responses, thereby significantly influencing the progression of these diseases (Li et al. 2025). Due to their stability in bodily fluids like blood and urine, tsRNAs show promise as non-invasive indexes for ongoing monitoring of liver illness (Li et al. 2024b; Chen et al. 2023b; Zhang et al. 2025). Although these potential functions are intriguing, the specific pathways by which tsRNAs exert their influence are still not fully understood. There is a need for more research to clarify the functional roles of these entities and their potential as therapeutic targets or diagnostic tools in liver diseases.

Interestingly, diverse functional tsRNAs are abundant in sEVs (Zhao et al. 2025; Wang et al. 2024b), but it is uncertain whether BMSC-derived sEVs influence MAFLD via tsRNAs (Fang et al. 2023; Wang et al. 2020). In this research, we explored the impacts and potential mechanisms of tsRNAs carried by sEVs derived from BMSCs on hepatic cell gluconeogenesis and lipid metabolism in MAFLD through different experimental models.

Materials and Methods

Cell culture and treatment, transfection

Human BMSCs and HepG2 cells were collected from Procell Life Company. BMSCs were incubated in α-MEM; (Invitrogen, located in Grand Island, NY, added 10% FBS sourced from Gibco Company in the USA, along with 100 U/mL of penicillin also obtained from Gibco/Life Technologies. The temperature was set to 37 °C and the atmosphere contained 5% CO2. Every 2–3 days, the fluid was replaced, and cells from the third generation were utilized for subsequent experiments. HepG2 cells were cultivated in DMEM (C3110-0500;VivaCell, which consists of 10% FBS, is cultured in incubator at 37 °C with a 5% carbon dioxide (CO2) atmosphere. To create a model resembling MAFLD, hepG2 cells were treated with PA. The 5'-tRF-GlyCCC mimic and 5'-tRF-GlyCCC inhibitor sequences were synthesized and prepared by GenePharma Co, Ltd.The cells from Shanghai, China were then transfected into PA induce-HepG2 cells via liposomes. The respective sequences were as follows: 5'-tRF-GlyCCC mimic, GTGGTAGAATTCTCGCCT, 5'-tRF-GlyCCC inhibitor, TCCGCTCTTAAGATGGTG.

Osteogenic differentiation assay

BMSCs were seeded onto 6-well plates and cultured in osteogenic induction medium, and 10 mmol/L β-glycerophosphate was added to the medium. In the following two weeks, the cells were treated with solution containing 50 µg per milliliter of ascorbic acid and 100 nM per liter of dexamethasone. The medium was replenished every three days. Then, the cells were stabilized via 4% paraformaldehyde. The USA-based procedure as follow, 15 min duration,staining with Alizarin Red solution 30 min at 25℃.Then, the positive mineralized nodules were detected inverted microscope.

Adipogenic differentiation assay

The cells were seeded in 6-well plates and cultured in adipogenic induction medium (with 1 μmol/L dexamethasone, 10 µg per milliliter of insulin). The cells were treated with 200 micromolar indomethacin for 21 days. The medium was replaced every three days. And then stained with Oil Red O solution. Positive lipid droplets were detected via inverted microscope.

Chondrogenic differentiation assay

The cells were centrifugated with speed of 1500 rpm for 5 min, and then cultivated in a chondrogenic induction medium, with 10 ng/mL TGF-β3, ascorbic acid, and 100 nmol/L dexamethasone, treated for 4 weeks.After that, the pellets were embedded in paraffin, then sectioned, and stained via Alcian Blue solution (USA).

Flow cytometry for surface markers

The BMSCs were trypsinized, then washed via PBS,and resuspended. Subsequently, the cells were cultured with antibodies that were conjugated with fluorochromes, specifically targeting CD29 and CD44 (BD Biosciences, USA). The CD31 and CD34 (BioLegend, USA) were stored at 4 °C for 30 min without light. Then it was detected via a flow cytometer (BD FACSCanto II, USA). The obtained data were treated via FlowJo software and the positive rates of surface markers were calculated.

sEVs isolate and Identification

The cell supernatants were collected from BMSCs that were cultured in serum-free media produced by Umibio, so as to avoid contamination from serum-derived EVs. The process of ultracentrifugation was employed to eliminate dead cells and large cell fragments, specifically by subjecting them to a centrifugal force of 2,000 × g for a duration of 10 min, followed by a higher centrifugal force of 10,000 × g for a shorter period of 5 min.Subsequently, sEVs were acquired utilizing exoEasy Maxi Kit according to the guidelines. TEM was used to examine the morphology of sEVs. In summary, the sEVs were treated with a 1ml fixative solution and then kept at 4 °C for a duration of 4 to 6 h.Ultra-thin slices (measuring 65 nm) were prepared and subsequently stained using a solution containing 1% uranyl acetate. Before conducting the analysis via TEM (JEOL, USA), it is important to ensure that the sample is properly prepared. The Nanoparticle tracking analysis (NTA) was conducted with 1.2 × 1010 particles/mL of SEVs. To measure the protein level of sEVs related proteins CD81, CD63, Alix, a western blot assay was employed.

Glucose and lipid parameters

To investigate the glucose production in HepG2 cells, an ELISA kit was employed. To measure liver triglyceride (TG) and TC levels in the liver, the kits from Cayman Chemical in the USA were utilized following the manufacturer's guidelines.

Pandora sequence and RNA sequence

Researchers discovered that sncRNAs present in BMSCs derived sEVs exhibited a notable abundance relative to other RNA types. Briefly, sncRNAs were extracted using the Qiagen miRNeasy Mini Kit (Qiagen), followed by treatment with AlkB, and subsequently reverse-transcribed into cDNA. The abundance of sncRNAs relative to each other in small extracellular vesicles (sEVs) derived from BMSCs was assessed via cDNA for PCR amplification and sequencing, utilizing Thermo Fisher Scientific's tools from the USA. In the process of data analysis, raw reads were aligned to a variety of ncRNA databases, like miRBase for miRNA and GtsRNAdb for tsRNA. To complete the corresponding annotation, we can utilize the piRNA database known as piRNABank.RNA extraction was performed on HepG2 cells that had been subjected to treatment with PA or PA combined with BMSC-sEVs, followed by reverse transcription into cDNA for subsequent RNA-seq analysis. The original sequencing reads were mapped using Bowtie2 software, and similarly. The expression levels of genes were determined through RSEM, and the DeSeq2 was applied to determine the differential expression of the genes. Genes that had a FDR less than 0.05 and a FC greater than 2 were chosen as being differentially expressed.

RT-qPCR assay

Total RNA was obtained from hepG2 cells utilizing TRIzol Reagent, and subsequently, 2 µg of RNA was converted into complementary DNA (cDNA) employing the RevertAid First Strand cDNA Synthesis Kit. After the initial step, a qRT-PCR analysis was conducted employing SYBR Premix Ex Taq polymerase (manufactured by TaKaRa Biotechnology, located in Dalian, China). The relative sncRNA expression of 5'-tRF-GlyCCC was normalized to that of U6.

Western blotting

In this experimental process, radioimmunoprecipitation assay buffer (Beyotime, China) was utilized and treated with protease as well as phosphatase inhibitor. To determine the concentration of the isolated proteins, the BCA kit from the USA was utilized. After that, the proteins underwent standard SDS polyacrylamide gel electrophoresis (USA) for separation and were subsequently transferred onto polyvinyl difluoride membranes (Millipore, USA). The membranes were subjected to a thorough washing process using PBS (Gibco, USA) and subsequently blocked with milk (BD Difco, USA). The cells were subjected to probing using primary antibodies specifically targeting CD9 (1:1000, Abcam in the UK) and CD63 (1:800, Abcam).UK), ALIX (1:1000, USA), pFoxO3 (1:1000), FoxO3 (1:1000), PEPCK (1:1000), G6Pase (1:1000,Abcam (UK), SREBP1c-FL (1:1000, USA), and SREBP1c-N (1:1000, USA), FASN, SCD1 (1:1000), and GAPDH (1:5000, Proteintech, USA) for 12 h at 4℃. Then, the membranes that had been cultured via the aforementioned primary antibodies underwent washing, followed by secondary antibodies (Cell Signaling Technology). The USA) was subjected to incubation at 25℃ for one hour. The protein bands were identified using an improved chemiluminescence reagent (USA). The abundance of proteins was calculated via the ImageJ software (NIH, USA).

Chromatin immunoprecipitation (ChIP)

The protein-DNA interactions were identified through the use of a ChIP assay, employing the Chromatin IP kit (manufactured by Cell Signaling Technology in the USA) and making some modifications. In summary, the HepG2 cells were subjected to crosslinking using formaldehyde obtained from Sigma-Aldrich.USA) for 10 min and subsequently subjected to lysis using Cell lysis buffer (produced by Cell Signaling Technology, The chromatin solution containing DNA with an average size of 500 bp was obtained and immunoprecipitated with protein A/G-agarose beads at 4 °C for 15 min. The experiment involved conjugating the samples with normal IgG (produced by Cell Signaling Technology in the USA) or an anti-FoxO3 antibody (also obtained from Cell Signaling Technology in the USA) at a temperature of 4 °C for an extended period overnight. Next, the immunoprecipitated DNA was extracted and measured through PCR. The primers sequences: (PEPCK Region 1: F-5’-AGGCGCTTAAAAAGGAGGGG-3’ and R-5’-GGGCGGAGGAAAGCTAGTG-3’, Region 2: F-5’-GTGCCCAATGCTAAGGTCCA-3’ and R-5’-AGGAAGAGACCCAGGGGTAG-3’). Area 3: F-5’-CCAGCCGCACATGATGTAAC-3’ and R-5’-CTGGAAAGGCAATGCCCAAC-3’.

Luciferase reporter assay

The DNA sequences of the wild-type and mutant 3′UTR of FoxO3 were synthesized and acquired from GenePharma Co, Ltd.(Shanghai, China). The Target Expression Vector was utilized to insert the sequences. The vector and 5'-tRF-GlyCCC mimic/NC were co-transfected into cells via the transfection reagent Lipofectamine 2000 (Invitrogen, as previously mentioned).USA) at 25℃ for 20 min. Subsequently, luciferase activities at 560 nm were further measured following the transfection via the Reporter Gene System (USA).

sEVs uptake

The lipophilic dye Dil (produced by Thermo company in the USA) was applied to label stem cell-derived extracellular vesicles from BMSCs, which were then incubated at a temperature of 37 °C for a duration of 30 min. Following the removal of excess dye using PBS, the labeled sEVs underwent another round of ultracentrifugation to separate them. HepG2 cells were treated with BMSC-sEVs, then fixed using 4% paraformaldehyde, followed by washing with PBS, and finally stained with DAPI (Thermo Fisher Scientific, USA). Ultimately, Dil fluorescence was detected via a fluorescence microscope (Olympus, Japan) for the purpose of capturing photographs.

Animal experiments

Male C57BL/6J mice, 18–20 g, were bought from the Shanghai SLAC Laboratory Animal Company for a duration of 6 to 8 weeks.The specimen was cultivated under condition of 22 ± 2 ℃ and RH of 60 ± 5%, under a 12-h light cycle followed by a 12-h dark cycle. After 7 days, the animals were randomly allocated into four distinct groups, each with 10 mice. The control group was provided with a standard chow diet (NCD, D12450J, Trophic Diets, Nantong, China). The experimental group was provided with a high-fat diet (HFD, D12492, Trophic Diets, Nantong, China) for a continuous period of 20 weeks. Furthermore, AAV8 is specifically targeted towards the liver.VectorBuilder) 5'-tRF-GlyCCC Ctrl/OE (3 × 10^9/200uL) was administered via tail vein injection to both the control and MAFLD model group. For intraperitoneal insulin tolerance tests (ITT), following a 6-h fasting period, the mice were administered an injection of 1U/kg insulin (Sigma-Aldrich,The USA) was measured at intervals of 0, 30, 60, 90, and 120 min, and the concentration was determined. Intraperitoneal glucose tolerance tests (GTT) were conducted on mice following an overnight fast, during which they received a dose of 2 g/kg of glucose (Sigma-Aldrich, USA).Following the initial steps, blood glucose concentration was detected at time point of 15, 30, 60, 120 min via glucometer (Accu-Chek, Switzerland).The Animal Use and Care Committee gave approval to all animal experiments conducted.

Statistical analysis

All figures now note the quantity of repetitions ("n = 3 "In the in vitro experiment, n = 10, and in the in vivo experiment, statistical data were expressed as m ± SD. The T-test was applied to check the difference in different groups. Three or more sets of data were checked via ANOVA test. The GraphPad 8.0.2 software were used to treat the data, with statistical level of 0.05. *. **, ***, ****is corresponded to p < 0.05, 0.01, 0.001, 0.0001.

Results

Identification of BMSC-sEVs and Verification of Hepatocyte Uptake

The research employed BMSCs in Passage 3, and their identity and stemness were confirmed. The ability of multilineage differentiation was verified through Alizarin Red S staining (osteogenic), Oil Red O staining (adipogenic), and Alcian Blue staining (chondrogenic) (Supplementary Fig. S1A). Flow cytometry analysis revealed a significant upregulation of BMSC-specific markers CD29 and CD44. The cells exhibited low levels of endothelial/hematopoietic markers CD31 and CD34, thereby confirming that they fulfilled the standard criteria for BMSCs (Supplementary Fig. S1B).

In order to better understand the core intervention vector for future functional experiments, we initially isolated and identified sEVs produced by BMSCs. We examined if these vesicles could be internalized by the target liver cells. BMSC-sEVs were extracted from the culture supernatants of BMSCs through a process that involved differential centrifugation and utilized a kit-based technique. Observations using TEM indicated that the isolated vesicles exhibited a cup-shaped morphology, aligning with the ultrastructural characteristics of sEVs as depicted in Fig. 1A. The nanoparticle tracking analysis showed that the vesicle size predominantly fell within the range of 80–120 nm. The average size of these particles is around 120 nm, fulfilling the size criteria for sEVs (Fig. 1B). Western blotting (WB) confirmed that the vesicles exhibited high level of sEV-specific markers like CD9, CD63, and Alix, which indicated successful isolation of sEVs (Fig. 1C).

Fig. 1.

Fig. 1

Open in a new tab

Identification of BMSC-sEVs and uptake in HepG2 cells. A The morphology of sEVs was observed by transmission electron microscopy, Scale bar: 100 nm. B The size of sEVs was observed by NanoSight analysis. C sEVs marker proteins (CD9, CD63, and Alix) were analyzed via WB method. D The fluorescence of HepG2 cells treated with DIL dye-labeled sEVs was measured via immunofluorescence. E The uptake of sEVs in HepG2 cells was quantified

In Vitro Inhibition of Palmitic Acid (PA)-Induced Hepatocellular Lipid Accumulation by BMSC-sEVs

Once it was established that BMSC-sEVs could be internalized by hepatocytes, we further confirmed the regulatory impact of sEVs on hepatocellular lipid metabolism by employing a PA-induced HepG2 cell model (which simulates the pathological condition of MAFLD in vitro). HepG2 cells were subjected to 0.2 mmol/L PA for 24 h in order to create a model of lipid accumulation and were subsequently subjected to treatment using BMSC-sEVs, as well as the culture supernatant of BMSCs (which contains sEVs). Before collecting the supernatant, BMSCs were pretreated with the sEV inhibitor GW4869 (10 µM) for 24 h. The findings from the Oil Red O staining indicated that, in comparison to the control group, the PA-treated group exhibited a notable rise in red lipid droplets. Utilizing BMSC-sEVs or BMSC supernatant as a treatment method significantly diminished the quantity of lipid droplets. The addition of GW4869 had a significant effect in reversing the decrease in lipid droplet accumulation, as compared with the BMSC-sEVs group (Fig. 2A), implying that the inhibitory effect of BMSCs on lipid accumulation relies on sEVs.

Fig. 2.

Fig. 2

Open in a new tab

Identification of lipogenesis in PA-treated HepG2 cells. A The lipid accumulation was detected in different groups by Oil Red O staining. B The adipogenic protein expression levels of SREBP1c-FL (full-length), SREBP1c-N (N-terminal), FASN and SCD1 were detected by WB. C The quantification of WB data for the protein levels was normalized to GAPDH and expressed relative to the control group. D The secretion of triglyceride (TG) in HepG2 cells was measured using a TG detection kit. E The secretion of total cholesterol(TC) in HepG2 cells was measured using a TC detection kit. ** p <0.01, *** p <0.001, **** p <0.0001.

To better understand the impact of sEVs on lipid production, WB was employed to assess the levels of crucial lipid synthesis proteins. In hepatocytes treated with PA, the expression of sterol regulatory element-binding protein 1c full-length (SREBP1c-FL) was detected. The stearoyl-CoA desaturase 1 (SCD1), active N-terminal fragment of SREBP1c (SREBP1c-N), and fatty acid synthase (FASN) were obviously increased. Significant downregulation of these proteins was observed when intervention with BMSC-sEVs or BMSC supernatant was applied, but the downregulatory effect was reversed by GW4869 (Fig. 2B, C).Biochemical analysis findings strongly indicate that the application of BMSC-sEVs led to a substantial reduction in triglyceride (TG) and total cholesterol (TC) levels within PA-induced HepG2 cells, specifically with TG: p < 0.01; The effect was blocked by GW4869 (p < 0.001), as shown in Figs. 2D and E. The findings suggested that BMSC-sEVs have the potential to decrease PA-induced lipid accumulation by suppressing the expression of crucial proteins associated with hepatocellular lipid synthesis. The provision of functional evidence serves as a basis for further research into the fundamental molecular mechanism.

Molecular Mechanism of BMSC-sEVs Regulating Hepatocellular Glucose and Lipid Metabolism via the 5'-tRF-GlyCCC/FoxO3 Axis

After confirming that BMSC-sEVs could inhibit hepatocellular lipid accumulation, multi-omics sequencing combined with functional experiments were further performed to explore the key molecules and regulatory pathways involved. First, Pandora small RNA sequencing was used to determine the composition of their carried non-coding RNAs (Shi et al. 2021). It was found that the small RNAs enriched in sEVs were mainly rRNA-derived fragments (rsRNAs, about 50%), unannotated small RNAs (Unanno UMG) and tRNA-derived fragments (tsRNAs). The 5'-tRF had the highest rate (10.7%). According to the amino acid association analysis result, glycine (Gly)-associated tsRNAs have the highest abundance in sEVs (Figs. 3A–C), which reflects that 5'-tRF-Gly is a critical functional molecule in sEVs.

Fig. 3.

Fig. 3

Open in a new tab

Identification of small RNAs in BMSC-sEVs through Pandora analysis. A, B The abundance and length of different types of small RNAs were detected through Pandora analysis. C Detection of amino acid abundance in BMSC-sEVs. 

To confirm the alterations in expression of candidate tsRNAs in PA-induced HepG2 cells, RT-qPCR was employed. The levels of expression for 5'-tRF-GlyCCC, 5'-tRF-GluCTC, 3'-tRF-GluTTC, and 3'-tRF-AspGTC were notably reduced in the PA-treated group. The BMSC-sEVs intervention had a significant effect in reversing this trend, and among the various factors, 5'-tRF-GlyCCC exhibited the most substantial rise in expression (p < 0.001) (Fig. 4A). RNA sequencing and KEGG pathway enrichment analysis were conducted on HepG2 cells that had been exposed to BMSC-sEVs. The findings indicated that genes exhibiting differential expression were notably overrepresented in the PI3K-Akt pathway, which is closely related to insulin sensitivity and glucose-lipid metabolism. Heatmap analysis demonstrated that the expression of FoxO3 (a crucial downstream transcription factor within the PI3K-Akt pathway) in hepatocytes treated with PA was substantially elevated. PA-BMSC-sEVs intervention effectively reduced its expression levels, as shown in Figs. 4B and C.

Fig. 4.

Fig. 4

Open in a new tab

Identification of 5'-tRF-GlyCCC expression in PA-treated HepG2 cells after BMSC-sEVs treatment. A The tsRNA expression levels in different groups of HepG2 cells were detected by RT-qPCR assay. B Identification of KEGG pathway enrichment analysis for differentially expressed tsRNA. C Clustering heatmap of expression in MAFLD model cells treated with PBS and BMSC-sEVs. ** p <0.01, *** p <0.001.

In order to confirm the regulatory connection between 5'-tRF-GlyCCC and FoxO3, analysis utilizing the RNAhybrid database demonstrated that 5'-tRF-GlyCCC has the capability to specifically bind to the 3' UTR of FoxO3, as depicted in Fig. 5A. The dual-luciferase reporter assay demonstrated that the co-transfection of a 5'-tRF-GlyCCC mimic and a FoxO3-3' UTR wild-type vector caused an obvious decrease in luciferase activity (p < 0.01). Although the mutant vector transfection had no significant impact, it confirms that 5'-tRF-GlyCCC can directly target and bind to the 3' UTR of FoxO3 (Fig. 5B). Functional studies demonstrated that the introduction of a 5'-tRF-GlyCCC mimic resulted in a substantial reduction in the protein levels of FoxO3 and crucial gluconeogenic genes (PEPCK and G6Pase) within HepG2 cells (p < 0.001), as well as a decrease in hepatocellular glucose production (p < 0.001). In contrast, the transfection of the 5'-tRF-GlyCCC inhibitor produced an effect that was the exact opposite (Figs. 5C-E).

Fig. 5.

Fig. 5

Open in a new tab

5'-tRF-GlyCCC may bind to FoxO3 3′UTR to regulate glucose production. A The binding sites of in FoxO3 3′-UTR. B The transcriptional activity of FoxO3 regulated by 5'-tRF-GlyCCC was detected by dual-luciferase reporter gene assay. C The protein expressions of p-FoxO3, FoxO3, G6Pase, PEPCK were detected by western blotting. D The quantification of WB data were normalized to GAPDH and expressed as relative value. E The quantitative detection of glucose production through ELISA kit. ** p < 0.01, *** p <0.001, **** p <0.0001.

In order to further verify that 5'-tRF-GlyCCC controls glucose and lipid metabolism via FoxO3, rescue experiments were carried out. In PA-treated HepG2 cells,when co-transfecting a 5'-tRF-GlyCCC mimic and a FoxO3 overexpression plasmid, it was observed that the overexpression of FoxO3 significantly counteracted the downregulatory impact of the 5'-tRF-GlyCCC mimic on the levels of PEPCK and G6Pase (Fig. 6A, B) and restored hepatocellular glucose production (p < 0.001) (Fig. 6C). The Chromatin Immunoprecipitation (ChIP) assay result indicated that FoxO3 directly interacts with three distinct areas of the PEPCK gene promoter. The strongest binding appeared at Region 3 (Site 3: 5’-GTAAACA-3’) as anticipated by the JASPAR database. Moreover, when FoxO3 is overexpressed, it significantly enhances the binding enrichment of FoxO3 at the PEPCK promoter (Fig. 6D, E). The findings revealed that the 5'-tRF-GlyCCC transported by BMSC-sEVs can inhibit the expression of FoxO3 and the transcription of downstream gluconeogenic genes through direct interaction with the 3' UTR of FoxO3, as indicated in Fig. 5A-E. Therefore, it is a key factor in controlling the metabolic processes of glucose and lipids in hepatocytes.

Fig. 6.

Fig. 6

Open in a new tab

5'-tRF-GlyCCC mediates glucose production by downregulating the expression of FoxO3. A The protein expression of FoxO3, G6Pase, and PEPCK in different groups of HepG2 cells was detected by WB. B The quantification of WB data were normalized to GAPDH. C The quantitative detection of glucose production in different groups of HepG2 cells was detected ELISA kit. D Schematic diagram of the identified binding sequence in the promoter region of the PEPCK gene targeted by FoxO3. E The occupancy of FoxO3 on PEPCK promoter in PA-treated HepG2 cells was detected using ChIP assay. ** p < 0.01, *** p <0.001.

Verification of the Therapeutic Effect of AAV-Mediated 5'-tRF-GlyCCC Overexpression in a MAFLD Mouse Model

In light of the in vitro validation of the regulatory mechanism involving the 5'-tRF-GlyCCC/FoxO3 axis, the therapeutic potential of this axis in vivo was further confirmed through experimentation on a mouse model with MAFLD. Three groups were formed from male C57BL/6J mice. The experimental groups consist of a control group receiving a normal diet along with AAV-GFP, a MAFLD model group undergoing a high-fat diet combined with AAV-GFP, and an intervention group subjected to a high-fat diet supplemented with AAV-5'-tRF-GlyCCC mimic. Following a 20-week period of intervention, significant changes were observed in the relevant indicators: The study on body weight monitoring revealed that the mice in the experimental group exhibited an obviously higher body weight compared to those in the control group. The body weight of mice in the 5'-tRF-GlyCCC intervention group was obviously less than the model group (p < 0.01) as shown in Fig. 7A, B. The findings from the glucose tolerance test (GTT) and insulin tolerance test (ITT) indicated a notable decline in the model group's capacity to clear glucose, as compared to other groups.The 5'-tRF-GlyCCC intervention led to a substantial enhancement in these indicators (p < 0.01) (Figs. 7C, D).

Fig. 7.

Fig. 7

Open in a new tab

5'-tRF-GlyCCC mimic suppresses HFD-induced MAFLD. A, B Detection of curve and gain of body weight of mice (n=10). C Measurement of plasma glucose during glucose tolerance test (GTT) of mice (n=10). D Measurement of plasma glucose during insulin tolerance test (ITT) of mice (n=10). E Oil Red O staining and PAS staining of liver sections from mice (n=10). F,G Detection and of quantification protein expressions of the genes. * p < 0.05, ** p <0.01, *** p < 0.001.

Liver histopathological analysis revealed that the HFD group exhibited a notable rise in Oil Red O-positive lipid droplets within liver tissue. The PAS staining demonstrated a rise in hepatic glycogen buildup. On the other hand, the 5'-tRF-GlyCCC treatment group exhibited a notable decrease in lipid droplet quantity within liver tissue, along with a substantial reduction in glycogen buildup (Fig. 7E). WB analysis of liver tissue proteins revealed that the levels of phosphorylated FoxO3 (p-FoxO3 Ser253), the expression of FoxO3, gluconeogenic genes (G6Pase, PEPCK), and lipid synthesis genes (SREBP1c-FL, SREBP1c-N, FASN, SCD1) in the model group were obviously elevated. The intervention using 5'-tRF-GlyCCC notably reduced the expression of these proteins (Fig. 7F, G). The findings demonstrated that the excessive expression of 5'-tRF-GlyCCC in the in vivo MAFLD model has the potential to improve hepatic lipid accumulation and rectify abnormal gluconeogenesis. The study aimed to investigate the impact of 5'-tRF-GlyCCC, derived from BMSC-sEVs, on insulin resistance in mice by modulating FoxO3 and its downstream genes related to glucose-lipid metabolism, thereby validating the therapeutic potential of this molecule for MAFLD.

Discussion

Metabolic-associated fatty liver disease (MAFLD) has emerged as a significant global health issue, marked by an abnormal buildup of lipids in the liver and insulin resistance (Younossi et al. 2018; Byrne and Targher 2015). Despite significant progress in comprehending the development of this condition, there is still a dearth of effective treatment choices available. In this investigation, we provide evidence that BMSC-sEVs can reduce hepatic lipid buildup and improve insulin sensitivity in both laboratory and animal models of MAFLD. Importantly, we identify 5′-tRF-GlyCCC, a tRNA-derived small RNA (tsRNA) enriched in BMSC-sEVs, as a key mediator of these effects. Through the action of 5′-tRF-GlyCCC on FoxO3, hepatic gluconeogenesis and lipogenesis are suppressed, thereby enhancing metabolic homeostasis. This research reveals an innovative pathway by which stem cell-derived sEVs influence liver metabolism and suggests a promising treatment approach for MAFLD. The mechanism being discussed is depicted in a schematic diagram, which can be seen in Fig. 8.

Fig. 8.

Fig. 8

Open in a new tab

Graphical mechanism: BMSC-sEV-derived 5′-tRF-GlyCCC mitigates MAFLD via glucose and lipid metabolism modulation. BMSCs secrete sEVs enriched with 5′-tRF-GlyCCC. These sEVs are transferred to hepatocytes in MAFLD, where 5′-tRF-GlyCCC promotes the degradation of FoxO3 mRNA. Reduced FoxO3 subsequently suppresses the transcription of gluconeogenic genes (PEPCK, G6Pase), attenuating hepatic gluconeogenesis. Concurrently, 5′-tRF-GlyCCC-mediated downregulation of SREBP1c curtails the expression of lipogenic genes (ACC, FASN), inhibiting hepatic lipid synthesis. Collectively, BMSC-sEV-derived 5′-tRF-GlyCCC alleviates hepatic lipid accumulation and insulin resistance in MAFLD by modulating glucose and lipid metabolic pathways.

BMSC-sEVs Modulate Hepatic Lipid and Glucose Metabolism

Small extracellular vesicles (sEVs) are a key factor in intercellular communication by facilitating the transfer of functional RNA, proteins, and lipids between cells (Kim et al. 2024). In line with prior research showcasing the therapeutic promise of such sEVs in metabolic disorders (Damania et al. 2018), our results show that BMSC-sEVs effectively reduce lipid accumulation and improve insulin sensitivity in PA-treated hepatocytes. The reduction in intracellular triglycerides and cholesterol, alongside decreased expression of lipogenic markers (SREBP1c, FASN, SCD1) (Gong et al. 2023; Li et al. 2023), indicates that BMSC-sEVs have a dampening effect on hepatic lipogenesis. Additionally, increased insulin signaling, evidenced by enhanced glucose tolerance test (GTT) and insulin tolerance test (ITT) outcomes, highlights the importance of BMSC-sEVs in enhancing glucose absorption and utilization. These results emphasize the promising role of BMSC-sEVs as a therapeutic option for addressing MAFLD by targeting various facets of hepatic metabolic dysfunction.

Role of 5′-tRF-GlyCCC in Regulating Gluconeogenesis

The identification of tsRNAs as regulators of cellular metabolism introduces an additional dimension to our comprehension of RNA biology. In this research, we aim to provide a deeper understanding of the significance of sEV-derived tsRNAs in MAFLD. We performed Pandora sequencing and identified that 5′-tRF-GlyCCC is among the most prevalent tsRNAs found in BMSC-sEVs. Our findings indicate that the level of 5′-tRF-GlyCCC is notably decreased in hepatocytes exposed to PA, but this reduction can be reversed through treatment with BMSC-sEVs, implying a potential involvement in mitigating metabolic dysfunction. Functional tests demonstrated that the overexpression of 5′-tRF-GlyCCC suppresses gluconeogenesis by decreasing the levels of PEPCK and G6Pase. While the inhibition of this substance produces the opposite outcome, it establishes a direct connection between 5′-tRF-GlyCCC and the regulation of glucose metabolism (Sun et al. 2017; Zhang et al. 2023).

From a mechanistic standpoint, we show that 5′-tRF-GlyCCC has an impact by suppressing FoxO3, which is a crucial transcription factor responsible for controlling gluconeogenic gene expression (Hu et al. 2022; Dobson et al. 2011; Wang et al. 2019; Piccolo et al. 2021). The 5′-tRF-GlyCCC directly binds to the 3′UTR of FoxO3, thereby inhibiting its translation and subsequently preventing the activation of gluconeogenic pathways. The role of 5′-tRF-GlyCCC in suppressing FoxO3-mediated transcriptional activity is supported by these findings, which also emphasize the significance of tsRNAs in regulating hepatic glucose metabolism.

In Vivo Validation and Therapeutic Implications

Our in vitro findings are further supported by in vivo experiments conducted on an HFD-induced MAFLD mouse model. In a study involving mice, it was observed that those administered with AAV-5′-tRF-GlyCCC demonstrated substantial enhancements in body weight, glucose tolerance, and hepatic lipid content. Furthermore, there was a noticeable decrease in the expression of FoxO3, PEPCK, and G6Pase. These findings affirm that 5′-tRF-GlyCCC has the ability to enhance hepatic insulin sensitivity and decrease lipid accumulation in living organisms, thereby substantiating its therapeutic potential for metabolic liver diseases.

Future Directions and Clinical Implications

Our study provides a mechanistic framework in which BMSC-sEVs deliver 5′-tRF-GlyCCC to hepatocytes, where it suppresses FoxO3-mediated gluconeogenesis and lipogenesis. This pathway offers a novel strategy for therapeutic intervention in MAFLD. However, several questions remain. First, the upstream mechanisms regulating 5′-tRF-GlyCCC biogenesis within BMSC-sEVs warrant further investigation. Second, while we have demonstrated the efficacy of 5′-tRF-GlyCCC in preclinical models, its clinical translation requires additional studies to evaluate its stability, delivery methods, and potential off-target effects.

Additionally, the potential for combinatorial therapies, where BMSC-sEVs are used in conjunction with other interventions, could further enhance therapeutic efficacy in treating MAFLD. Future clinical trials should focus on optimizing delivery strategies for tsRNAs or other sEV cargo to maximize their therapeutic potential and expand the application area.

Conclusion and Limitations

In conclusion, our study identifies BMSC-sEV-derived 5′-tRF-GlyCCC as a novel regulator of hepatic metabolic homeostasis, improving insulin sensitivity and reducing hepatic lipid accumulation in MAFLD models. By targeting FoxO3, 5′-tRF-GlyCCC suppresses key metabolic pathways, offering a promising therapeutic approach for MAFLD and potentially other metabolic diseases. Nevertheless, this study still exists some limitations. 5′-tRF-GlyCCC may also bind to other genes, like SIRT3, which may cause off-target effects and impact the treatment outcomes (Cheng et al. 2016). Further investigations are necessary to confirm these possible targets. Furthermore, although this research primarily investigates the 5′-tRF-GlyCCC/FoxO3 pathway, there are other sEV components that exert key roles. Other small RNAs, proteins, and lipids may also impact the treatment effects of MAFLD and need to be analyzed in the future. These results broaden the understanding of tsRNAs' involvement in metabolic control and emphasize the promise of sEV-based therapies for liver disease management.

Supplementary Information

Below is the link to the electronic supplementary material.

ESM 1 (7.1MB, docx)

(DOCX 7.14 MB)

ESM 2 (2.1MB, pdf)

(PDF 2.06 MB)

Acknowledgements

None.

Authors’ contributions

All authors were involved in the conception and design of the study, with their contributions being significant. Chenyun Yang and Huiling Chen executed investigations and scrutinized data; Chenyun Yang and Xinlu Yuan were responsible for drafting the manuscript; Huiling Chen and Xiaojing Huang were involved in conducting both animal and in vitro experiments. Chenyun Yang, Yanyan Li, and Song Wen were responsible for verifying the accuracy of the data. Yanyan Li and Song Wen revised the manuscript; Ligang Zhou and Xinlu Yuan oversaw the research project and made substantial edits to the manuscript. Every author meticulously reviewed and made revisions to the manuscript.

Funding

This study has obtained support from the National Natural Science Foundation of China (82100850), the Project of Key Medical Specialty of Pudong Hospital of Fudan University (No Tszb2025-13) and Scientific Program of Shanghai Pudong Hospital (YJRCJJ201808).

Data availability

All the data involved in present study are available from the corresponding author with reasonable requirements.

Declarations

Ethical approval

Ethical clearance and consent for involvement The procedures for animal testing were granted approval by the Ethics Committee of Pudong Hospital (Approval No. 2021-DS-Q-08). All animal procedures were performed according to the institutional guidelines of Laboratory Animal Care.

Consent for publication

Not applicable.

Clinical trial number

Not applicable.

Competing interests

The authors declare no competing interests.

Footnotes

The original online version of this article was revised due to Figures1-8 were wrongly numbered.

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Change history

4/24/2026

The original online version of this article was revised: The upper blank control spectra in Figure 3K were inadvertently reused from a previous study. This reuse was not adequately clarified in the manuscript, potentially causing confusion. The authors now amended figure 3 and all following figures where they have systematically revised the axis labeling.

Change history

4/23/2026

A Correction to this paper has been published: 10.1007/s10565-026-10187-y

References

  1. Afarin R, et al. The effect of lipopolysaccharide-stimulated adipose-derived mesenchymal stem cells on NAFLD treatment in high-fat diet-fed rats. Iran J Pharm Res. 2023;22(1):e134807. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Bi Y, et al. Bone marrow derived-mesenchymal stem cell improves diabetes-associated fatty liver via mitochondria transformation in mice. Stem Cell Res Ther. 2021;12(1):602. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Browning JD, et al. Prevalence of hepatic steatosis in an urban population in the United States: impact of ethnicity. Hepatology. 2004;40(6):1387–95. [DOI] [PubMed] [Google Scholar]
  4. Byrne CD, Targher G. NAFLD: a multisystem disease. J Hepatol. 2015;62(1 Suppl):S47-64. [DOI] [PubMed] [Google Scholar]
  5. Chaney A. Obesity and Nonalcoholic Fatty Liver Disease. Nurs Clin North Am. 2021;56(4):543–52. [DOI] [PubMed] [Google Scholar]
  6. Chen K, et al. Adipocytes-derived exosomal miR-122 promotes non-alcoholic fat liver disease progression via targeting Sirt1. Gastroenterol Hepatol. 2023a;46(7):531–41. [DOI] [PubMed] [Google Scholar]
  7. Chen K, et al. Identification of ferroptosis/autophagy-related genes and potential underlying mechanisms involved in the effect of BMSC senescence on the osteogenic differentiation of aging BMSCs. Genes & Diseases. 2025;12(1):101259. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Chen Q, et al. Origins and evolving functionalities of tRNA-derived small RNAs. Trends Biochem Sci. 2021;46(10):790–804. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Chen Q, et al. Biological functions and clinical significance of tRNA-derived small fragment (tsRNA) in tumors: Current state and future perspectives. Cancer Lett. 2024;587:216701. [DOI] [PubMed] [Google Scholar]
  10. Chen S, et al. Urinary exosome tsRNAs as novel markers for diagnosis and prediction of lupus nephritis. Front Immunol. 2023b;14:1077645. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Cheng Y, et al. MicroRNA-421 induces hepatic mitochondrial dysfunction in non-alcoholic fatty liver disease mice by inhibiting sirtuin 3. Biochem Biophys Res Commun. 2016;474(1):57–63. [DOI] [PubMed] [Google Scholar]
  12. Damania A, et al. Mesenchymal stromal cell-derived exosome-rich fractionated secretome confers a hepatoprotective effect in liver injury. Stem Cell Res Ther. 2018;9(1):31. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Ding J, et al. Emerging role of engineered exosomes in nonalcoholic fatty liver disease. World J Hepatol. 2023;15(3):386–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Dobson M, et al. Bimodal regulation of FoxO3 by AKT and 14-3-3. Biochim Biophys Acta. 2011;1813(8):1453–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Fang S, et al. tsRNA-15797-modified BMSC-derived exosomes mediate LFNG to induce angiogenesis in osteonecrosis of the femoral head. Turk J Biol. 2023;47(3):186–98. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Gong Z, et al. Rhinacanthin C ameliorates insulin resistance and lipid accumulation in NAFLD mice via the AMPK/SIRT1 and SREBP-1c/FAS/ACC signaling pathways. Evid Based Complement Alternat Med. 2023;2023:6603522. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Han JJ, Li J, Huang DH. Mesenchymal stem cell-derived extracellular vesicles carrying Circ-Tulp4 attenuate diabetes mellitus with nonalcoholic fatty liver disease by inhibiting cell pyroptosis through the HNRNPC/ABHD6 axis. Tissue Eng Regen Med. 2025;22(1):23–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Hu Y, et al. MicroRNA-122-5p inhibition improves inflammation and oxidative stress damage in dietary-induced non-alcoholic fatty liver disease through targeting FOXO3. Front Physiol. 2022;13:803445. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Huang P, et al. Elevation of plasma tRNA fragments as a promising biomarker for liver fibrosis in nonalcoholic fatty liver disease. Sci Rep. 2021;11(1):5886. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Jiang HT, et al. Therapeutic Role of Bone Marrow Mesenchymal Stem Cells (BMSCs) in Nonalcoholic Steatohepatitis (NASH) Cirrhosis. Curr Stem Cell Res Ther. 2025;20(8):830–8. [DOI] [PubMed] [Google Scholar]
  21. Jiang Y, et al. Current perspectives on mesenchymal stem cells as a potential therapeutic strategy for non-alcoholic fatty liver disease. World J Stem Cells. 2024;16(7):760–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Kim HI, et al. Recent advances in extracellular vesicles for therapeutic cargo delivery. Exp Mol Med. 2024;56(4):836–49. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Kitade H, et al. Nonalcoholic fatty liver disease and insulin resistance: new insights and potential new treatments. Nutrients. 2017. 10.3390/nu9040387. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Li N, et al. SREBP regulation of lipid metabolism in liver disease, and therapeutic strategies. Biomedicines. 2023. 10.3390/biomedicines11123280. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Li Y, et al. tRNA and tsRNA: from heterogeneity to multifaceted regulators. Biomolecules. 2024a. 10.3390/biom14101340. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Li X, et al. Serum tRF-33-RZYQHQ9M739P0J as a novel biomarker for auxiliary diagnosis and disease course monitoring of hepatocellular carcinoma. Heliyon. 2024b;10(9):e30084. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Li P, et al. Nsun2-mediated tsRNAs alleviate liver fibrosis via FAK dephosphorylation. Cell Prolif. 2025;58(10):e70058. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Ma X, et al. Exploring the clinical application and biological function of novel tsRNA tsRNA-Asp-5-0002 in hepatocellular carcinoma. Gene. 2025;956:149459. [DOI] [PubMed] [Google Scholar]
  29. Meneses MJ, et al. Targeting NAFLD with umbilical cord-derived mesenchymal stromal cells: a preclinical study. Diabetologia. 2023;66(Suppl 1):S64–S64. [Google Scholar]
  30. Moayedfard Z, et al. The role of the immune system in the pathogenesis of NAFLD and potential therapeutic impacts of mesenchymal stem cell-derived extracellular vesicles. Stem Cell Res Ther. 2022;13(1):242. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Pan Q, Han T, Li G. Novel insights into the roles of tRNA-derived small RNAs. RNA Biol. 2021;18(12):2157–67. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Piccolo P, et al. Up-regulation of miR-34b/c by JNK and FOXO3 protects from liver fibrosis. Proc Natl Acad Sci U S A. 2021. 10.1073/pnas.2025242118. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Shi J, et al. PANDORA-seq expands the repertoire of regulatory small RNAs by overcoming RNA modifications. Nat Cell Biol. 2021;23(4):424–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Sun X, et al. Protective effects of glycyrrhizic acid against non-alcoholic fatty liver disease in mice. Eur J Pharmacol. 2017;806:75–82. [DOI] [PubMed] [Google Scholar]
  35. Teixeira FS, et al. Impaired hepatic lipid metabolism and biomarkers in fatty liver disease. Biochimie. 2023;215:69–74. [DOI] [PubMed] [Google Scholar]
  36. Wang L, et al. FoxO3 regulates hepatic triglyceride metabolism via modulation of the expression of sterol regulatory-element binding protein 1c. Lipids Health Dis. 2019;18(1):197. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Wang T, et al. A novel tsRNA-16902 regulating the adipogenic differentiation of human bone marrow mesenchymal stem cells. Stem Cell Res Ther. 2020;11(1):365. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Wang S, et al. Hyperoside attenuates non-alcoholic fatty liver disease in rats via cholesterol metabolism and bile acid metabolism. J Adv Res. 2021;34:109–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Wang Y, et al. The effect and mechanism of liraglutide on the biological functions of BMSCs in diabetic patients. Oral Dis. 2024a;30(8):5160–74. [DOI] [PubMed] [Google Scholar]
  40. Wang J, et al. The mechanism of MMP14-positive tumor-associated fibroblast subsets in inhibiting PD-1 immunotherapy for esophageal cancer through exosomal tsRNA-10522. Funct Integr Genomics. 2024b;24(5):186. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Yang S, et al. Emerging insights into mesenchymal stem cells and exosome-based therapies for liver injury. Biomol Biomed. 2025;25(8):1691–708. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Ying S, et al. tRF-Gln-CTG-026 ameliorates liver injury by alleviating global protein synthesis. Signal Transduct Target Ther. 2023;8(1):144. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Younossi Z, et al. Global burden of NAFLD and NASH: trends, predictions, risk factors and prevention. Nat Rev Gastroenterol Hepatol. 2018;15(1):11–20. [DOI] [PubMed] [Google Scholar]
  44. Zhang Z, Ji G, Li M. Glucokinase regulatory protein: a balancing act between glucose and lipid metabolism in NAFLD. Front Endocrinol (Lausanne). 2023;14:1247611. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Zhang J, et al. Extracellular vesicle therapy for obesity-induced NAFLD: a comprehensive review of current evidence. Cell Commun Signal. 2024;22(1):18. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Zhang Z, et al. Exosomal tsRNA-Gly-5-0007 may be used as a diagnostic marker for colorectal cancer. Sci Rep. 2025;15(1):26751. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Zhao GY, et al. Tumor-derived exosomal tsRNA 3′tiRNA-AlaCGC in promoting fibroblast senescence and Galectin-9 secretion to induce immune tolerance in lung adenocarcinoma. Cell Death Discov. 2025. 10.1038/s41420-025-02695-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Zheng Y, et al. Mitochondrial metabolic dysfunction and non-alcoholic fatty liver disease: new insights from pathogenic mechanisms to clinically targeted therapy. J Transl Med. 2023;21(1):510. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

ESM 1 (7.1MB, docx)

(DOCX 7.14 MB)

ESM 2 (2.1MB, pdf)

(PDF 2.06 MB)

Data Availability Statement

All the data involved in present study are available from the corresponding author with reasonable requirements.

Articles from Cell Biology and Toxicology are provided here courtesy of Springer

RESOURCES