跳到正文
原文
Frontiers in Psychiatry· Zhen Zheng·· 4 小时前AI 评分33

干细胞来源细胞外囊泡治疗孤独症谱系障碍动物模型：一项系统综述与元分析

Efficacy of stem cell-derived extracellular vesicles in animal models of autism spectrum disorder: a systematic review and meta-analysis

AI 导读

一项系统综述与元分析显示，干细胞来源细胞外囊泡（EVs）可改善ASD动物模型的重复行为与社交缺陷，并降低促炎细胞因子IL-6和IL-1β水平。该研究检索5个数据库至2026年4月20日，从1,165篇文献中纳入5项动物研究，采用Review Manager 5.4合并数据，结果均P<0.05。

正文

Abstract

Background:

Autism spectrum disorder (ASD) is a complicated neurodevelopmental condition that currently lacks a targeted therapy. Although stem cell-derived extracellular vesicles (EVs) represent a potential therapeutic approach for neurological disorders, their efficacy in ASD remains unclear. The objective of this study was to investigate the efficacy of stem cell-derived EVs on autistic-like behaviors in animal models of ASD and to explore the underlying mechanisms.

Methods:

Relevant studies published up to April 20, 2026 were searched across five databases. Data were pooled using Review Manager 5.4. Outcome measures included behavioral tests and inflammatory cytokine levels.

Results:

Out of 1,165 initially retrieved articles, five animal studies met the inclusion criteria. Pooled analyses demonstrated that EV treatment improved autistic-like behaviors in the self-grooming test, three-chamber social interaction test, and marble burying test, and decreased levels of proinflammatory cytokines IL-6 and IL-1β (all P < 0.05). No notable enhancements were observed for the open field test, Morris Water Maze, TNF-α or IL-10 (all P > 0.05).

Conclusion:

Stem cell-derived EV therapy may improve repetitive behaviors and social deficits in ASD animal models, with observed associations with reduced IL-6 and IL-1β levels. These preclinical findings suggest promising translational potential of EV-based therapy for ASD.

Systematic Review Registration:

https://osf.io/c78xm, identifier 10.17605/OSF.IO/C78XM.

1 Introduction

Autism spectrum disorder (ASD) is a heterogeneous neurodevelopmental condition marked by impairments in social communication, narrow interests, and repetitive behaviors. The global prevalence of ASD has been rising steadily, placing a heavy burden on patients’ families and the whole society (). Despite intensive research, the exact pathogenesis of ASD remains elusive. Current interventions mainly focus on behavioral and educational interventions (). No specific targeted treatment is available to cure the core symptoms of ASD. Therefore, there is an urgent demand for new therapeutic strategies targeting the fundamental pathological mechanisms.

Stem cell-derived extracellular vesicles (EVs) have recently shown therapeutic potential in treating various neurological disorders (–). EVs contain abundant bioactive molecules, including mRNA, microRNA, DNA, and proteins (). Compared with stem cell-based therapy, stem cell-derived EVs offer several benefits, including low immunogenicity, blood-brain barrier penetration, and simpler storage and operation. Stem cell-derived EVs exert protective effects through anti-inflammatory, anti-oxidative, anti-apoptotic, and neurotrophic mechanisms (–).

Recent findings indicate that neuroinflammation contributes significantly to the pathogenesis of ASD (–). Increased concentrations of pro-inflammatory cytokines, including interleukin-6 (IL-6) and interleukin-1β (IL-1β), have been reported in both individuals with ASD and animal models (–). Preclinical studies have reported that stem cell-derived EVs can improve behavioral deficits and attenuate neuroinflammation in animal models of ASD (–). However, these studies vary in study design, animal model types, stem cell sources, and intervention protocols, leading to inconclusive evidence.

To date, no systematic review or meta-analysis has summarized preclinical evidence of stem cell-derived EV therapy for ASD or elucidated its therapeutic mechanisms. Thus, we performed the first systematic review and meta-analysis of preclinical animal studies to determine the efficacy of stem cell-derived EVs on autistic-like behaviors and to explore the underlying mechanisms. Our study aimed to lay a foundation for the clinical translation of EV-based therapy for ASD.

2 Methods

2.1 Data sources and search strategy

We performed a systematic screening of literature published before April 20, 2026, in PubMed, Web of Science, Embase, Cochrane Library, and Scopus. We have retrospectively registered the protocol on the Open Science Framework (OSF; https://osf.io/c78xm). The search strategy employed a combination of terms associated with extracellular vesicles (e.g., “exosomes”, “exosome”, “extracellular vesicles”, “extracellular vesicle”, “vesicle”, “microparticle”, “secretory vesicles”, “exovesicles”, “exovesicle”, “endosomes”, “nanovesicle”) with terms related to autism spectrum disorder (e.g., “ASD”, “autism spectrum disorder”, “autistic disorder”, “autism”, “Asperger syndrome”, “pervasive developmental disorder”). Additionally, we reviewed the references of the relevant articles to uncover other eligible studies.

2.2 Inclusion and exclusion criteria

Studies meeting the following criteria were included: (1) conducted using animal models of ASD; (2) used stem cell-derived EVs as an intervention; (3) included a control group; (4) published in English; and (5) provided sufficient quantitative data for meta-analysis.

Studies meeting the following criteria were excluded: (1) lacking a control group; (2) duplicate literature; (3) reviews, conference abstracts, comments, or case reports; (4) clinical studies or in vitro experiments; and (5) incomplete data.

2.3 Study selection

Two investigators independently screened the titles and abstracts. Subsequently, the complete articles were assessed against the inclusion and exclusion criteria. If disagreement arose, a third investigator rendered the final judgment.

2.4 Data extraction

We obtained key data from eligible studies, including authors, publication year, author country, autism model, sources of stem cell-derived EVs, time of administration, route of administration, control group, dose of EVs, frequency, follow-up duration, and outcome assessments. When individual studies did not provide exact numerical results, we used WebPlotDigitizer to extract values from the published charts. Then we calculated the mean values and standard deviations (SDs). If a study only provided the standard error of the mean (SEM), we calculated the SD from the SEM using the equation SD = SEM × √n, where n stands for the sample size of the group.

2.5 Quality assessment

To evaluate the quality of studies, we applied the risk bias tool from the Systematic Review Centre for Laboratory Animal Experimentation (SYRCLE). Ten aspects were examined: random sequence generation, baseline characteristics, allocation concealment, random housing, blinding of participants and personnel, random outcome assessment, blinding of outcome assessment, incomplete outcome data, selective reporting, and other bias. Two independent reviewers assigned a rating of low, unclear, or high risk to each domain. Disagreements were resolved by consensus.

2.6 Outcome indicators

Primary outcome indicators: self-grooming test, three-chamber social interaction test, marble burying test, open field test, and Morris Water Maze.

Secondary outcome indicators: brain tissue levels of IL-6, IL-1β, TNF-α, and IL-10.

2.7 Statistical analysis

Review Manager 5.4 was used to perform data analysis. Between stem cell-derived EV treatment and control groups, the pooled effect size was derived as a standardized mean difference (SMD) with a 95% confidence interval (CI), and a forest plot was provided. Two-sided tests were employed, and P < 0.05 indicated statistical significance. Heterogeneity was quantified by I2. There was significant heterogeneity among the studies, including differences in animal species (mice vs. rats), ASD induction models (genetic vs. chemical/VPA), EV sources (bone marrow vs. umbilical cord), administration routes (intravenous vs. intranasal), and dose variations. Therefore, a random-effects model was adopted for all outcomes to uniformly account for the differences among the studies. For analyses involving at least three studies, sensitivity analysis was carried out through stepwise removal of individual studies. The SMD and 95% CI from the remaining studies were then compared with the overall estimate.

3 Results

3.1 Study selection process

Of the 1,165 articles found, 493 originated from PubMed, 263 from Web of Science, 394 from Embase, 9 from the Cochrane Library, 4 from Scopus, and 2 were retrieved from the reference lists of eligible studies. After excluding 396 duplicate articles, 769 articles remained for screening. After reviewing titles and abstracts, 644 articles were eliminated, and the remaining 125 articles underwent full-text evaluation. According to the established criteria, 120 articles were removed for the reasons below: 112 articles dealt with irrelevant topics, 4 articles lacked precise data of sample sizes, and 4 articles had no full text. Ultimately, 5 articles qualified for inclusion in this meta-analysis. Figure 1 presents the PRISMA flow diagram.

Figure 1

3.2 Study characteristics

Table 1 presents the baseline characteristics of five included studies. Published from 2018 to 2025, the studies were performed in various countries, including Egypt, China, and Israel. In terms of animal models, one study used rats (), three studies employed C57BL/6 mice (–), and one study utilized BTBR mice (). Regarding the source of exosomes, two studies used bone marrow-derived mesenchymal stem cells (, ), and three studies applied umbilical cord mesenchymal stem cells (–). In terms of administration routes, one study administered exosomes via intravenous injection (), and four studies administered intranasally (–). Regarding the dosage arrangement, three studies administered 100μg exosomes each time (, , ), one study administered 50μg exosomes each time (), and another study did not specify the exact dosage of exosomes ().

Table 1

First author, yearCountryAnimalsSources of extracellular vesiclesTime of administrationRoute of administrationControlDoseFrequencyFollow-upOutcomes
Elsherif, 2025EgyptMale Wistar ratsBone marrow derived mesenchymal stem cells from white
albino rat		3 weeksIntravenous injectionNot mention100μg39 weeksThree-chamber social interaction test, open field test, Morris Water Maze
Liang, 2020ChinaMale C57BL/6 miceUmbilical cord mesenchymal stem cells from human6 weeksIntranasal injectionSaline100μgNot mention10–11 weeksIL-6, IL-1β, TNF-α, IL-10
Perets, 2018IsraelMale BTBR miceBone marrow derived mesenchymal stem cells from human4 weeksIntranasal injectionSalineNot mention6Not mentionSelf-grooming test
Qin, 2025ChinaMale C57BL/6 miceUmbilical cord mesenchymal stem cells from human6–7 weeksIntranasal injectionPBS100μg7Not mentionSelf-grooming test, three-chamber social interaction test, marble burying test, open field test
Wu, 2025ChinaMale C57BL/6 miceUmbilical cord mesenchymal stem cells from humanpostnatal day 50Intranasal injectionNot mention50μg411 weeksSelf-grooming test, three-chamber social interaction test, marble burying test, open field test, Morris Water Maze, IL-6, IL-1β, TNF-α, IL-10

Characteristics of the included studies.

3.3 Study quality

Risk of bias assessment was performed with the SYRCLE tool (Figure 2). Random sequence generation was unclear in all studies. Baseline characteristics of animals, such as age and gender, were generally reported. However, the influence of other factors still cannot be ruled out. Allocation concealment methods were not adequately described in any study. All studies reported random housing. Blinding of participants or personnel and randomization in outcome assessment were not described in any study. Blinding in outcome assessment was reported in only three studies (, , ). Other potential sources of bias remain uncertain due to insufficient information. In summary, the included studies showed unclear risk in multiple domains, and the overall bias risk was not judged as high.

Figure 2

3.4 Stem cell-derived EV treatment ameliorated autistic-like behaviors in ASD animals

Multiple investigations have evaluated the effects of stem cell-derived EV treatment on autistic-like behaviors in ASD animals. Three studies (, , ) reported reduced self-grooming behavior in ASD animals subsequent to stem cell-derived EV treatment (SMD = -1.31, 95% CI: -1.88 to -0.74, I2 = 14%, P < 0.001) (Figure 3A). Sensitivity analysis revealed that omitting any single study did not affect the overall effect. Three studies (, , ) reported that after stem cell-derived EV treatment, the time spent with stranger 1 in the three-chamber social interaction test among ASD animals was increased (SMD = 2.40, 95% CI: 1.10 to 3.70, I2 = 69%, P< 0.001) (Figure 3B). Sensitivity analysis indicated that the overall finding was not dominated by any one study. Two studies (, ) assessed the effect of stem cell-derived EV intervention on the marble burying test, with significant reductions in the number of buried marbles observed in the ASD animals (SMD = -2.28, 95% CI: -4.17 to -0.40, I2 = 79%, P = 0.02) (Figure 3C). Three studies (, , ) reported that after stem cell-derived EV administration, there were no statistically significant variations in the time spent in the center of the open field test for ASD animals (SMD = 1.35, 95% CI: -1.28 to 3.98, I2 = 93%, P = 0.31) (Figure 3D). Sensitivity analysis revealed that omitting any single study did not affect the pooled effect. Two studies (, ) reported that after stem cell-derived EV treatment, ASD animals showed no statistically significant change in target quadrant time during the Morris Water Maze test (SMD = 0.92, 95% CI: -0.77 to 2.61, I2 = 84%, P = 0.29) (Figure 3E). An overview of these findings is presented in Table 2.

Figure 3

Table 2

MeasuresStudiesEffect estimateHeterogeneity (I2, P)
Self-grooming test3-1.31 [-1.88, -0.74]I2 = 14%, P < 0.001
Three-chamber social interaction test32.40 [1.10, 3.70]I2 = 69%, P < 0.001
Marble burying test2-2.28 [-4.17, -0.40]I2 = 79%, P = 0.02
Open field test31.35 [-1.28, 3.98]I2 = 93%, P = 0.31
Morris Water Maze20.92 [-0.77, 2.61]I2 = 84%, P = 0.29
IL-62-1.62 [-2.80, -0.44]I2 = 49%, P < 0.01
IL-1β2-1.67 [-2.65, -0.69]I2 = 27%, P < 0.001
TNF-α2-1.82 [-4.49, 0.85]I2 = 89%, P = 0.18
IL-1021.70 [-0.26, 3.66]I2 = 80%, P = 0.09

Summary table of meta-analysis results.

3.5 Stem cell-derived EV treatment modulated inflammatory cytokine levels in ASD animals

Several investigations have evaluated the effects of stem cell-derived EV treatment on inflammatory cytokines in brain tissue of ASD animals. The findings demonstrated that stem cell-derived EV treatment significantly reduced the levels of pro-inflammatory cytokines in brain tissue, including IL-6 (SMD = -1.62, 95% CI: -2.80 to -0.44, I2 = 49%, P < 0.01) (Figure 4A) and IL-1β (SMD = -1.67, 95% CI: -2.65 to -0.69, I2 = 27%, P < 0.001) (Figure 4B). In contrast, stem cell-derived EV treatment did not significantly change the levels of TNF-α (SMD = -1.82, 95% CI: -4.49 to 0.85, I2 = 89%, P = 0.18) (Figure 4C) and IL-10 (SMD = 1.70, 95% CI: -0.26 to 3.66, I2 = 80%, P = 0.09) (Figure 4D) in brain tissue of ASD animals.

Figure 4

4 Discussion

The present work represents the first systematic review and meta-analysis investigating the preclinical efficacy of stem cell-derived EVs in animal models of ASD. By pooling data from five studies, we obtained combined effect estimates showing that EV treatment significantly ameliorated the core behaviors of ASD, including self-grooming, sociability deficits in the three-chamber test, and marble-burying behavior. In contrast, pooled analyses revealed no significant effects on open field test or Morris Water Maze. Additionally, EV therapy significantly reduced brain levels of IL-6 and IL-1β, but not TNF-α or IL-10. These data imply that stem cell-derived EVs may alleviate the core symptoms of ASD through IL-6 and IL-1β-driven neuroinflammation, providing a preliminary foundation for further clinical translation.

The three-chamber social interaction test measures sociability (), while the self-grooming and marble burying tests assess repetitive behaviors (, 27), which are the core symptoms of ASD. The most compelling conclusion of this meta-analysis is that EV therapy significantly improved these core symptoms. Notably, despite heterogeneity in sources (bone marrow and umbilical cord), administration routes (intravenous injection and nasal administration), and animal models (C57BL/6, BTBR, rats), the meta-analytic estimates remained significant. However, moderate to high heterogeneity was observed for the three-chamber test (I2 = 69%) and marble burying test (I2 = 79%). We adopted random-effects models to adjust for this heterogeneity, and sensitivity analyses indicated that removal of any one study did not affect the overall effect. This robustness strengthens confidence in our pooled conclusions.

In contrast, the meta-analysis found no significant improvement in the open field test (reflecting locomotor activity and anxiety-like behavior) or the Morris Water Maze (reflecting spatial learning and memory abilities) (28, 29). The small number of studies limited statistical power to detect modest effect sizes. In addition, the high heterogeneity suggests substantial variability in baseline locomotor activity, anxiety levels, or cognitive testing conditions across studies. Furthermore, the lack of effect on open field behavior might imply that EV treatment does not broadly alter general activity and anxiety levels. Similarly, the absence of significant cognitive improvement in the Morris Water Maze might suggest that EV treatment does not robustly affect hippocampal-dependent spatial memory. Larger-scale studies with harmonized protocols are warranted to clarify the observed outcomes.

Regarding neuroinflammation, our pooled estimates reveal that stem cell-derived EVs significantly reduced IL-6 and IL-1β levels in brain tissue. Both cytokines are upstream drivers of neuroinflammation and are implicated in the pathogenesis of ASD (, , 30). Elevated IL-6 and IL-1β can impair synaptic plasticity and contribute to excessive neuronal excitation (31). The heterogeneity of these results is relatively low, supporting the consistency of the anti-inflammatory effects across different studies. In contrast, the meta-analysis did not detect significant changes in TNF-α or the anti-inflammatory cytokine IL-10. The high heterogeneity and wide confidence intervals suggest that the null results may be due to low statistical power and between-study variability. Alternatively, EVs may selectively inhibit specific IL-6 and IL-1β pathways rather than exerting a broad anti-inflammatory effect. Given that the limited number of studies contributed to these cytokine outcomes, no definitive conclusion can be drawn. Future mechanistic studies should directly compare EVs derived from different sources and their impact on specific inflammatory pathways in ASD models.

Beyond pure therapeutic applications, EVs also hold promising diagnostic potential for ASD and exhibit close therapeutic relevance. Circulating EVs can cross the blood-brain barrier and carry molecules from the brain, including microRNAs, proteins, and inflammatory mediators, which can reflect the pathological state of the central nervous system. Studies have shown that exosomal proteins and microRNAs differ in the peripheral blood of ASD patients compared to neurotypical controls (32, 33), suggesting that EVs may become candidate peripheral biomarkers for the diagnosis of ASD. Our previous study reported reduced miR-30b-5p in plasma exosomes and brains of VPA-exposed rats with elevated inflammatory cytokines, while overexpression of miR-30b-5p ameliorated ASD-like behaviors and suppressed inflammation (), highlighting that EV cargo may function both as a biomarker and a therapeutic target. In addition, evidence suggests that EV-encapsulated molecules can also be used to monitor therapeutic responses (34).

Although preclinical results are positive, there are still many obstacles when translating these findings to humans. Firstly, standards for EVs are not yet fully established. Different laboratories use different methods for isolation and characterization, which affects reproducibility of the research. Secondly, the biological distribution and targeting mechanisms of EVs are not clear. It is currently unknown what fraction of EVs can reach relevant brain regions after nasal or intravenous injection. Thirdly, the long-term safety has not been evaluated, including non-specific effects, immunogenicity, and the impact on neurodevelopment. Fourthly, the animal models have inherent limitations. Animal models cannot reproduce the heterogeneity of human genetic, environmental, and developmental processes. There are significant differences in behavior, cognition, and inflammatory dynamics among different species. Therefore, although this meta-analysis provides a foundation for subsequent research, caution is warranted when applying the evidence to human therapeutics.

Several limitations of this meta-analysis should be noted. First, the pool of studies was small in size (n = 5), with individual meta-analyses relying on only 2–3 studies. Such limited data yield imprecise effect estimates. Heterogeneity metrics such as I2 and τ2 are unstable and should be interpreted with considerable caution. Treatment effects may also be overestimated due to the limited number of included studies. Therefore, our findings are preliminary and require validation in larger, well-powered preclinical studies. Second, the studies exhibit suboptimal methodological standards. For most studies, the risk of bias was unclear across critical domains, including random sequence generation, allocation concealment, and blinding. This reflects a common defect in preclinical research and may lead to overestimation of treatment effects. Third, there was significant heterogeneity in some indicators, the sources of which could not be fully explored because of the few included studies. Fourth, the follow-up durations were relatively short, and it is still unclear whether the effects of EVs are durable and safe over the long term. Fifth, our meta-analysis only focused on several inflammatory cytokines, while other mechanisms (oxidative stress, mitochondrial function, synaptic plasticity, gut-brain axis) were not investigated.

Based on these limitations, we propose the following suggestions for future preclinical studies to improve the quality of evidence for meta-analyses. First, well-designed experiments following standardized protocols and with large sample sizes should be conducted. Second, standardized procedures for isolating, characterizing, dosing, and administering EVs are urgently needed to improve comparability across studies. Third, the continued efficacy and long-term safety of stem cell-derived EV treatment for ASD must be evaluated through extended follow-up studies. Fourth, the exact mechanisms by which EVs exert effects should be further explored.

5 Conclusion

This meta-analysis is the first to comprehensively analyze the preclinical evidence of stem cell-derived EV therapy in ASD. The pooled results show that EV treatment can improve repetitive behaviors and social deficits, with observed associations with reduced IL-6 and IL-1β levels. Stem cell-derived EVs represent a promising approach for ASD treatment, despite the limited number and high heterogeneity of available studies. Our results suggest clinical translation of EV-based therapy for ASD, which will require larger-scale, methodologically rigorous preclinical studies to verify efficacy and safety.

Statements

Data availability statement

The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding authors.

Author contributions

ZZ: Conceptualization, Methodology, Investigation, Data curation, Writing – original draft. MJ: Investigation, Data curation, Writing – original draft. YL: Writing – original draft, Formal analysis, Investigation, Software. SP: Writing – review & editing, Conceptualization.

Funding

The author(s) declared that financial support was received for this work and/or its publication. This work was supported by the National Science Foundation of China (81901523), the Guangzhou Science and Technology Planning Project (202201011476), the Fundamental Research Funds for the Central University (SCU2026D005, 2023KF02, 19ykpy38), and the Research Project on the Reform of Teaching Method in Sun Yat-sen University (82000-12263012).

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpsyt.2026.1956214/full#supplementary-material

References

  • 1

    GuoJChenMZhouYWangJHuangQDingJet al. The global burden of autism spectrum disorders in children aged 0-14 years from 1990 to 2021: a population-based study. Psychiatry Res. (2026) 362:117181. doi: 10.1016/j.psychres.2026.117181

  • 2

    HirotaTKingBH. Autism spectrum disorder: a review. JAMA. (2023) 329:157–68. doi: 10.1001/jama.2022.23661

  • 3

    De RosaGRussoCGaravelliSDi SilvestreDSpatoccoIMeleGet al. MicroRNA-142-3p shuttling in extracellular vesicles marks regulatory T cell dysfunction in multiple sclerosis. Sci Transl Med. (2025) 17:eadl1698. doi: 10.1126/scitranslmed.adl1698

  • 4

    LiJXuMCaiBLiXLiangZXiaXet al. Clinically inspired multimodal treatment using induced neural stem cells-derived exosomes promotes recovery of traumatic brain injury through microglial modulation. Adv Sci (Weinh). (2025) 12:e08574. doi: 10.1002/advs.202508574

  • 5

    LiYHuJChenJZhouMLiaoMYangYet al. RVG-modified BMSCs-derived small extracellular vesicles loaded with miR-21 alleviate neuronal injury resulted from excessive autophagy via targeting PTEN/Akt/mTOR pathway after cerebral ischaemia. J Extracell Vesicles. (2026) 15:e70295. doi: 10.1002/jev2.70295

  • 6

    MondalKGhantyRMahadevanAWaghmareGSanthoshkumarRBnNet al. Intranasal delivery of DPSC-derived small extracellular vesicles-encased phloroglucinol attenuates non-motor and motor deficits and promotes neurogenesis in an in vivo rat model of Parkinson's disease. Stem Cell Res Ther. (2025) 16:570. doi: 10.1186/s13287-025-04573-2

  • 7

    YangLChanMShengJQiSChanBShantaramDet al. Decoding adipose-brain crosstalk: distinct lipid cargo in human adipose-derived extracellular vesicles modulates amyloid aggregation in Alzheimer's disease. Alzheimers Dement. (2025) 21:e70603. doi: 10.1002/alz.70603

  • 8

    ZhangAZhangZLiuRZhaoZLiuL. Exosomes derived from human umbilical cord mesenchymal stem cells attenuate neuroinflammation in intracerebral hemorrhage rats by modulating lactylation modification. J Transl Med. (2026) 24:588. doi: 10.1186/s12967-025-07041-w

  • 9

    ParkJYJungSYYooDLeeSGoDBuiVDet al. Stem cell-derived extracellular vesicles in skin antiaging treatments. ACS Nano. (2025) 19:22504–27. doi: 10.1021/acsnano.4c17406

  • 10

    LiLZhangHSunLSuYXuYHuangJet al. Injectable porous microspheres loaded with biomimetic preconditioned bone marrow mesenchymal stem cell-derived exosomes for vascularized bone regeneration. Adv Sci (Weinh). (2026) 13:e74987. doi: 10.1002/advs.74987

  • 11

    WangQGuoWNiuLZhouYWangZChenJet al. 3D-hUMSCs exosomes ameliorate vitiligo by simultaneously potentiating treg cells-mediated immunosuppression and suppressing oxidative stress-induced melanocyte damage. Adv Sci (Weinh). (2024) 11:e2404064. doi: 10.1002/advs.202404064

  • 12

    XiongLHuoSYangYZhangQLiJBaiWet al. Exosomes derived from neural stem cells contribute to cerebral ischemia/reperfusion injury via inhibiting autophagy in rats. J Neuroimmunol. (2026) 413:578875. doi: 10.1016/j.jneuroim.2026.578875

  • 13

    YuRBaiHMoYLiuGZhouZXuLet al. Neural stem cell extracellular vesicle-mediated delivery of astragaloside IV attenuates hypoxic-ischemic brain damage by activating the mTOR pathway. Naunyn Schmiedebergs Arch Pharmacol. (2026) 399:17071–89. doi: 10.1007/s00210-026-05452-0

  • 14

    ZhangXJingFLiJBaiYGuoCJinHet al. Human neural stem cell-derived exosomes promote functional recovery in subarachnoid hemorrhage via bdnf/trkb pathway activation and astrocyte modulation. Acta Neuropathol Commun. (2025) 13:245. doi: 10.1186/s40478-025-02155-0

  • 15

    GuimarãesVQuagliatoL. Evaluation of IL-1β, 8-OHdG and TBARS in autistic children: the role of inflammatory biomarkers. Int J Dev Neurosci. (2026) 86:e70111. doi: 10.1002/jdn.70111

  • 16

    ZhengPZhaoHZhangXWuQZhengZLiuS. Oolong tea attenuates neuroinflammation by modulating the microbiota-gut-brain axis in a rat model of autism. Front Nutr. (2025) 12:1643147. doi: 10.3389/fnut.2025.1643147

  • 17

    ZhengZWuQZhangX. Plasma exosomal miR-30b-5p attenuates neuroinflammation in a rat model of autism spectrum disorder. Front Psychiatry. (2025) 16:1630782. doi: 10.3389/fpsyt.2025.1630782

  • 18

    ElsherifRMm Abdel-HafezAHusseinOASabryDAbdelzaherLABayoumyAA. The potential ameliorative effect of mesenchymal stem cells-derived exosomes on cerebellar histopathology and their modifying role on PI3k-mTOR signaling in rat model of autism spectrum disorder. J Mol Histol. (2025) 56:65. doi: 10.1007/s10735-024-10335-w

  • 19

    FuYZhangYLLiuRQXuMMXieJLZhangXLet al. Exosome lncRNA IFNG-AS1 derived from mesenchymal stem cells of human adipose ameliorates neurogenesis and ASD-like behavior in BTBR mice. J Nanobiotechnol. (2024) 22:66. doi: 10.1186/s12951-024-02338-2

  • 20

    QinQShanZXingLJiangYLiMFanLet al. Synergistic effect of mesenchymal stem cell-derived extracellular vesicle and miR-137 alleviates autism-like behaviors by modulating the NF-κB pathway. J Transl Med. (2024) 22:446. doi: 10.1186/s12967-024-05257-w

  • 21

    WuCLiXYangXWangHLinGLiuZ. Dual-target hUCMSCs/EVs therapy for autism spectrum disorder: remodeling gut microbiota and modulating neuroimmune crosstalk in a valproic acid-induced C57BL/6 mice model. Stem Cell Res Ther. (2025) 17:32. doi: 10.1186/s13287-025-04829-x

  • 22

    LiangYDuanLXuXLiXLiuMChenHet al. Mesenchymal stem cell-derived exosomes for treatment of autism spectrum disorder. ACS Appl Bio Mater. (2020) 3:6384–93. doi: 10.1021/acsabm.0c00831

  • 23

    QinQFanLZengXZhengDWangHLiMet al. Mesenchymal stem cell-derived extracellular vesicles alleviate autism by regulating microglial glucose metabolism reprogramming and neuroinflammation through PD-1/PD-L1 interaction. J Nanobiotechnol. (2025) 23:201. doi: 10.1186/s12951-025-03250-z

  • 24

    PeretsNHertzSLondonMOffenD. Intranasal administration of exosomes derived from mesenchymal stem cells ameliorates autistic-like behaviors of BTBR mice. Mol Autism. (2018) 9:57. doi: 10.1186/s13229-018-0240-6

  • 25

    SofyanaMPigginsHDJacksonMGRobinsonESJ. Beyond the barrier: an assessment of social behaviour in mice using a modified 3-chamber test. Physiol Behav. (2026) 25:115381. doi: 10.1016/j.physbeh.2026.115381

  • 26

    KawadaKMimoriSKuramotoNUnoK. Prenatal 4-phenylbutyric acid administration during mid-gestation ameliorates ASD-like behaviors by reducing cortical endoplasmic reticulum stress in VPA-induced ICR and BTBR mouse models. AIMS Neurosci. (2026) 13:29–49. doi: 10.3934/Neuroscience.2026002

  • 27

    MazaheriMRadahmadiMSharifiMR. Impacts of prolonged different social (equality and inequality) conditions on spatial learning and memory deficits, as well as obsessive-compulsive disorder-like behaviors in male rats. Physiol Behav. (2026) 305:115177. doi: 10.1016/j.physbeh.2025.115177

  • 28

    HanLXiongXFanMZhangLGaoSLiRet al. Early-life psychological stress impairs spatial learning and memory in rats by altering the hippocampal proteome. Behav Genet. (2026) 56:138–52. doi: 10.1007/s10519-026-10269-x

  • 29

    LinXHuangLLingXLuJLinLDingLet al. Intermittent fasting attenuates cognitive decline in D-galactose-induced aging rats in association with β-hydroxybutyrate and PI3K/AKT/GSK-3β signaling. Neurochem Res. (2026) 51:179. doi: 10.1007/s11064-026-04804-9

  • 30

    AlòROlivitoIMinerviniDAvolioEDe BartoloAPerrottaIet al. Ketogenic diet improves motor and anxiety-like behaviors plus cerebellar inflammation via axonal remyelination in the BTBR mouse model of autism. J Nutr Biochem. (2026) 157:110417. doi: 10.1016/j.jnutbio.2026.110417

  • 31

    AbdelazizAMFawzyMNFathyMKShokrMM. From insult to hyperexcitability: pharmacological targeting of MyD88 and JAK/STAT3 pathways in epilepsy. Inflammopharmacology. (2026) 34:3209–33. doi: 10.1007/s10787-026-02210-9

  • 32

    PhilosophKLYoffeCEl BirBBaryakhPSaarEGMayorovMet al. Distinct protein profiles in cord blood plasma of children with autism spectrum disorder. Sci Rep. (2025) 15:44947. doi: 10.1038/s41598-025-29420-5

  • 33

    ZhaoSZhongYShenFChengXQingXLiuJ. Comprehensive exosomal microRNA profile and construction of competing endogenous RNA network in autism spectrum disorder: a pilot study. Biomol BioMed. (2024) 24:292–301. doi: 10.17305/bb.2023.9552

  • 34

    NikanfarMAvciCBKhalilnezhadAShahriyariENourazarianA. Brain-derived exosomes in neurodevelopmental and neuropsychiatric disorders: molecular insights, therapeutic potential, and translational challenges. Mol Neurobiol. (2025) 62:16342–82. doi: 10.1007/s12035-025-05247-z

Keywords

animal models, autism spectrum disorder, meta-analysis, neuroinflammation, stem cell-derived extracellular vesicles

Citation

Zheng Z, Ji M, Liang Y and Pan S (2026) Efficacy of stem cell-derived extracellular vesicles in animal models of autism spectrum disorder: a systematic review and meta-analysis. Front. Psychiatry 17:1956214. doi: 10.3389/fpsyt.2026.1956214

Received

02 August 2026

Revised

14 September 2026

Accepted

23 September 2026

Published

12 October 2026

Volume

17 - 2026

Edited by

Gülay Tasci, Elazig Fethi Sekin City Hospital, Türkiye

Updates

Copyright

© 2026 Zheng, Ji, Liang and Pan.

This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

*Correspondence: Yayong Liang, lyyong@mail.sysu.edu.cn; Sinian Pan, pansn@mail.sysu.edu.cn

†These authors have contributed equally to this work

Disclaimer

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.

来源：Frontiers in Psychiatry · frontiersin.org

#研究发布#神经科学#儿童青少年#Frontiers

猜你喜欢