Non-coding RNA in extracellular vesicles and the occurrence and progression of psychiatric disorders: A narrative review

Abstract

As the role of non-coding RNAs (ncRNAs) in psychiatric disorders gains increasing prominence, attention has shifted to their dysregulated expression in these conditions. Extracellular vesicles (EVs), particularly exosomes, play a crucial role as key mediators of cellular communication within the brain and between peripheral tissues. Their influence on neurogenesis, synaptic plasticity, and neuronal stress responses underscores their significance in psychopathophysiology. Due to their stability, disease-specific expression, and functional roles, EV-encapsulated ncRNAs have emerged as highly promising biomarkers. This review systematically explores the interactions between EV-derived ncRNAs and psychiatric disorders, evaluating their diagnostic potential and therapeutic applications in the field of psychiatry.

Key Words: Non-coding RNA; Extracellular vesicles; Psychiatric disorders; Relationship; Psychiatry

Core Tip: Non-coding RNAs encapsulated within extracellular vesicles, particularly exosomes, play a crucial role in the pathophysiology of psychiatric disorders and are involved in key processes such as neurogenesis, synaptic plasticity, and neuronal stress responses, making them promising biomarkers for diagnosing and treating psychiatric conditions. Their stability and disease-specific expression further enhance their potential in advancing psychiatric research and therapeutic strategies.

INTRODUCTION

Psychiatric disorders, also known as mental disorders or mental illnesses, represent a diverse group of conditions that affect an individual’s cognitive, emotional, and behavioral functioning. Psychiatric disorders affect millions worldwide, including over 6.4 million individuals diagnosed with schizophrenia (SCZ), approximately 1.1 million suffering from bipolar disorder (BD)[1,2], and an alarming 193 million cases of major depressive disorder (MDD)[3]. Despite the clinical differences, all three disorders have neurodevelopmental origins, and studies have shown overlapping relationships and commonalities in underlying biological mechanisms[4].

Currently, physicians have a misdiagnosis rate of up to one-quarter during clinical diagnosis based on epidemiologic studies, as well as subjective feelings and expressions of patients that tend to influence clinicians’ assessment, and a high degree of heterogeneity in patients’ responses to antipsychotic medications, which are clinical problems that lead to delays or inappropriate management during the treatment of psychiatric disorders. Therefore, there is an urgent need for a reliable and objective biomarker to help clinicians achieve accurate diagnoses and treatments. In addition, a significant number of psychiatric patients do not respond well to commercially available medications. The primary reason is the incomplete understanding of the underlying mechanisms of these disorders and the limited biologically based guidance available for clinicians. This shortage of guidance hinders their ability to predict the progression of psychiatric disorders in high-risk patients and prescribe drugs that can effectively treat them.

Non-coding RNAs (ncRNAs), including microRNAs (miRNAs), circular RNAs (circRNAs), and long ncRNAs (lncRNAs), have emerged as critical regulators of gene expression and cellular function. Unlike traditional coding RNAs, ncRNAs are not translated into proteins but play pivotal roles in modulating cellular processes, including neurodevelopment, synaptic plasticity, and immune regulation[5]. Growing evidence suggests that dysregulation of ncRNAs is intricately linked to the pathogenesis of neurodegenerative and psychiatric disorders, such as SCZ, MDD, and BD[6-8]. Notably, the dynamic changes in ncRNA expression levels during disease progression and treatment highlight their potential as biomarkers and therapeutic targets[9]. Among ncRNAs, miRNAs, circRNAs, and lncRNAs have garnered particular attention due to their high conservation, tissue specificity, stability, and disease-specific expression patterns.

NcRNAs are not only critical regulators of cellular processes but are also actively packaged into extracellular vesicles (EVs), where they mediate intercellular communication. EVs, including exosomes, are membrane-bound nanoparticles secreted by cells that carry bioactive molecules such as RNAs, proteins, and lipids. These vesicles serve as critical mediators of intercellular communication, facilitating the transfer of molecular cargo between cells and influencing processes such as immune modulation, synaptic plasticity, and neuronal survival[10]. EVs are readily detectable in accessible body fluids, including blood, urine, and cerebrospinal fluid (CSF), making them an attractive tool for liquid biopsy. Liquid biopsy is a non-invasive technique that analyzes biomarkers such as circulating tumor DNA, circulating tumor cells, or exosomes in bodily fluids like blood to detect cancer and other diseases[11]. NcRNAs encapsulated within EVs not only enhance the stability and bioavailability of ncRNAs but also allow them to exert systemic effects across different tissues and organs. Moreover, the ability to profile EV-derived ncRNAs from peripheral blood provides a minimally invasive approach to identifying disease-specific biomarkers and monitoring treatment responses.

Recent studies have demonstrated the potential of EV-associated ncRNAs in unraveling the molecular mechanisms underlying psychiatric disorders[12]. For instance, specific miRNAs and lncRNAs carried by EVs have been implicated in the dysregulation of neuroinflammatory pathways, synaptic dysfunction, and neurotransmitter imbalances observed in these conditions[12]. This emerging field not only provides novel insights into the pathogenesis of psychiatric disorders but also opens new avenues for developing targeted therapies. This review aims to comprehensively summarize the relationships between ncRNAs in EVs and major psychiatric disorders, as well as the current state of research in this field. Based on these findings, this review endeavors to provide insights and novel perspectives for the discovery of new diagnostic biomarkers and therapeutic targets for psychiatric disorders.

MAJOR PSYCHIATRIC DISORDERS

SCZ

SCZ is a chronic psychiatric disorder characterized by abnormalities in perception, emotion, and behavior. It has a lifetime prevalence of around 1%, onset typically occurs during adolescence and early adulthood, and persists throughout most of an individual’s life[13]. Moreover, it imposes a substantial economic burden on both society and families[14]. The pathogenesis of SCZ is currently unknown. It has been found that the development of SCZ is caused by the interaction of genetics, brain structure, pregnancy complications, and environmental factors[15]. Genetic predisposition plays a pivotal role in the etiology of SCZ, as evidenced by its remarkably high heritability estimate, which approaches 80%[15,16]. Additionally, environmental factors and stressors also substantially influence its development, including maternal immune system infections during pregnancy[17], advanced paternal age, childhood abuse[18], and exposure to substances like marijuana[19]. It is hypothesized that many genes undergo epigenetic modifications as a result of environmental exposures that have long-term effects, and that these modifications accumulate in the brain and other tissues during the development of individuals, thereby affecting genome function[15,20]. The diagnosis of SCZ is confirmed only if the patient meets the diagnostic criteria; however, this diagnosis is based on a subjective assessment by the clinician. To date, no reliable physical examination findings or biomarkers have been found to support this diagnostic process.

MDD

Depression refers to a common psychiatric disorder characterized by persistent sadness and a lack of interest or pleasure in activities. The core symptoms of MDD include hopelessness, social avoidance, insomnia, anxiety, cognitive and sexual dysfunction[21]. They are accompanied by a high degree of chronicity and a high rate of recurrence[22-24]. The pathogenesis of depression is intricate, involving neurotransmitters, hormonal mediation, and multilevel interactions between different regions of the brain. Neurotransmitters such as dopamine, norepinephrine, and serotonin play critical roles in this process, and dysfunction of these systems leads to altered neuroplasticity, enhanced inflammatory responses, and abnormal synaptic function, collectively contributing to the development and persistence of depressive symptoms[25].

Currently, medications such as ketamine can effectively control the symptoms of depression[26], but they are unable to cure the disorder. Patients experience relapse within a short period of time after discontinuing the medication. During relapses, patients may experience increased blood pressure, nausea, memory loss, gastrointestinal disorders, and recurrent fluctuations in mental status. Given the lack of clear, objective diagnostic criteria and biomarkers for MDD, three commonly used clinical symptom rating scales are available: The Hamilton Depression Rating Scale, the Beck Depression Inventory, and the Self-Depression Scale, all of which can be used to reliably assess the degree of depression. However, there are certain inaccuracies in using the symptom rating scale to determine if a patient has MDD. Moreover, currently, there is a lack of objective diagnostic tools for depression.

BD

BD is a group of highly clinically heterogeneous, multifactorial disorders characterized by recurrent cycles of depression (typical features), mania or hypomania, alternating episodes of common psychiatric disorders[27]. The etiology of BD remains elusive. A substantial body of research data indicates that genetic factors, psychosocial factors, biological factors, and an individual’s psychological state all exert a significant influence on the development of BD[28]. Moreover, the interactions among these factors contribute to the onset and progression of the disorder[28]. Among them, genetic factors account for a large proportion, and the heritability range of BD is between 70% and 90%[29]. However, due to the complexity of the disease’s etiology, there are no exact diagnostic criteria, resulting in a high rate of underdiagnosis and misdiagnosis.

NCRNAS

MiRNAs

MiRNAs represent a class of evolutionarily conserved, small ncRNA molecules, typically comprising approximately 22 nucleotides in length[30]. They are present in both animal and plant cells, playing a crucial role in regulating cellular signaling and biological processes[15]. The biogenesis of miRNAs is initiated by RNA polymerase II, which transcribes miRNA genes to produce primary miRNAs[15]. These primary miRNAs are then processed into mature miRNAs, which can be found in various subcellular structures and may even be released via exosomes[30]. Mature miRNAs can bind to the 3’-untranslated regions of target mRNAs, leading to their degradation or translational repression, thereby controlling gene expression[15]. Alterations in miRNA expression play a crucial role in regulating genetic networks and signaling cascades. Moreover, abnormal miRNA expression is frequently associated with the development of various diseases[9]. Beyond diagnostics, miRNAs hold therapeutic potential due to their ability to modulate cellular functions involved in some diseases[31]. Additionally, the co-regulation of genes and pathways by multiple miRNAs forms a complex regulatory network that may underpin many human pathologies, suggesting a role for miRNAs in epigenetic regulation[32].

LncRNAs

LncRNAs are a class of ncRNAs exceeding 200 nucleotides in length, lacking a significant open reading frame and thus minimal protein-coding capacity[33]. Their formation is multifaceted, arising from various pathways, including the independent transcription of introns, promoters, enhancer regions, and specific lncRNA genes in intergenic areas[34,35]. This diversity in generation mechanisms underpins their extensive biological functions. LncRNAs exert regulatory effects on gene expression at multiple levels, including chromatin remodeling, RNA modification, RNA stability regulation, transcriptional modulation, splicing control, and translation initiation. These regulatory actions of lncRNAs have a significant impact on cellular gene expression profiles[36-39]. LncRNAs are involved in key cellular processes related to physiology, and alterations in their expression are linked to diverse disease developments[40]. They participate in dosage compensation, transcriptional and post-transcriptional regulation, epigenetic control, and cellular differentiation[41-43]. Their subcellular localization correlates with their functional roles, particularly in nuclear functions, where they can regulate protein localization, abundance, and host gene expression[41-43]. Additionally, lncRNAs interact with DNA, RNA, or proteins to form complexes that modulate gene expression patterns[44]. Their expression is finely tuned and contributes to various gene expression regulatory mechanisms, including influencing the transcription of neighboring genes and participating in chromatin biology processes such as DNA replication and the response to DNA damage[33,45].

CircRNAs

CircRNAs are widely expressed RNAs that do not have structures such as a 5’ terminal cap and a 3’ terminal polyadenylation tail[46]. They are mainly formed through a specific RNA splicing process, known as “reverse splicing”, which produces alternative splicing products that enable circRNAs to form a covalently closed continuous loop structure[47]. Research indicates that circRNAs participate in diverse molecular functions associated with cellular metabolism, the cell cycle, transcriptional regulation, and the efficient initiation of translated protein products[48,49]. CircRNAs are essential components of gene expression mediation, and they are involved in and influence a variety of biological processes, including early growth, immune response, neurodevelopment, and cancer formation[50]. CircRNAs contain both exons and introns and can participate in regulating gene expression[51].

CircRNAs exhibit multiplicity through alternative splicing, generating diverse exon-intron structures from different genes[52]. With high efficiency, the circRNA structure can act as a template for roll-over replication, facilitating the efficient production of multiple RNA copies[53]. In human fibroblasts, 14.4% of actively transcribed genes generate circRNAs, which show high expression and are more abundant and stable than their linear counterparts[54]. Due to the lack of 5’-3’ polarity and polyadenylated tails, circRNAs exhibit high stability[55,56], thereby resisting exonucleases and accumulating in various cell types, including neural tissues. Additionally, neuronal circRNAs in the mammalian brain exhibit conservation, with higher sequence conservation and expression levels compared to other RNA types[55,56].

NCRNAS AND PSYCHIATRIC DISORDERS

NcRNA and SCZ

A multifaceted and intricate interplay across multiple molecular levels characterizes the pathogenesis of SCZ. Core contributors to its development include dysfunctional molecular pathways such as neuroinflammation, oxidative stress, and mitochondrial dysfunction, which collectively underpin the disorder’s etiology[57]. Furthermore, disruptions in synaptic function and abnormalities in neurodevelopmental processes, as well as genetic predisposition, significantly influence the progression of SCZ. Complementing these central factors, dysregulation of neurotransmitter systems, particularly dopamine and glutamate, plays a crucial role in the pathogenesis of SCZ, underscoring the integrative nature of its underlying mechanisms[25].

In recent years, epigenetic mechanisms, particularly miRNA-mediated post-transcriptional regulation, have made significant progress in understanding the molecular mechanisms of SCZ, providing an increasing wealth of evidence for elucidating the pathophysiological processes of the disease[58] (Table 1). For example, Zhao et al[59] found that miR-3653-3p was expressed at a low level in the peripheral blood of patients with SCZ. Perkins et al[60] detected miRNAs in the postmortem prefrontal cortex (PFC) of 13 SCZ patients by performing customized microarray analysis[60]. Comparing them with healthy controls (HCs), they identified 16 miRNAs that were differentially expressed. García-Cerro et al[4] observed differences in miR-132-3p transcript levels in plasma and post-mortem PFC tissues from a mouse model of SCZ. They found that SCZ patients had reduced levels of miR-21-5p in peripheral blood mononuclear cells (PBMCs). Beveridge et al[61] analyzed 17 pairs of SCZ patients, finding overall elevated expression of miRNAs associated with SCZ in postmortem tissues of the superior temporal gyrus and dorsolateral PFC. The significance analysis of microarrays reported that a total of 59 miRNAs were upregulated[61]. Elevated levels of HOXA-AS2, Linc-ROR, MEG3, SPRY4-IT1, and UCA1 were found in SCZ patients’ blood compared to HCs[62]. When gene expression was assessed by gender, these lncRNAs were found to be significantly expressed in females[62]. This study suggests that gender-based lncRNA dysregulation exists in patients with SCZ, and thus, they may be used as diagnostic biomarkers.

Table 1 The association between non-coding RNA and schizophrenia.

Ref.	Sample source	Samples	NcRNA	Method	Results
Miller et al[100], 2012	Prefrontal cortical tissue	SCZ: 35; HCs: 34; BD: 31	MiRNA	RT-qPCR	483 miRNAs up
Du et al[94], 2019	Exosome in blood	SCZ: 49; HCs: 46	MiRNA	MiRNA-seq NGS	18 miRNAs up
Zhao et al[59], 2023	Peripheral blood	SCZ: 8; HCs: 15	MiRNA	RT-qPCR	41 miRNAs down; 35 miRNAs up
Liu et al[107], 2019	PBMCs	SCZ: 32; HCs: 48	LncRNA	RT-qPCR	The expression levels of CSMD1 are correlated with SCZ
Fallah et al[65], 2019	Peripheral blood	SCZ: 60; HCs: 60	LncRNA	RT-qPCR	5 lncRNAs up
Teng et al[108], 2023	SCZ postmortem brains	SCZ: 559; HCs: 936	LncRNA	RNA-seq analysis; RT-qPCR	LncRNA-GOMAFU down
Zimmerman et al[79], 2020	Human postmortem brain total RNA	SCZ: 34; HCs: 34	CircRNA	RT-qPCR	CircHomer1a down
Mahmoudi et al[109], 2021	PBMCs	SCZ: 39; HCs: 20	CircRNA	RNA-seq analysis	8762 circRNAs up; 55 circRNAs down
Huang et al[110], 2023	Plasma	SCZ: 10; HCs: 8	CircRNA	RT-qPCR; RNA-seq analysis	1 circRNAs up; 234 circRNAs down

NcRNA: Non-coding RNA; PBMCs: Peripheral blood mononuclear cells; SCZ: Schizophrenia; HCs: Healthy controls; BD: Bipolar disorder; MiRNA: MicroRNA; CircRNA: Circular RNA; LncRNA: Long non-coding RNA; RT-qPCR: Reverse transcription-quantitative polymerase chain reaction; NGS: Next-generation sequencing; CSMD1: CUB and Sushi multiple domains 1.

NcRNA and MDD

Current research suggests that ncRNAs play a complex and critical role in the mechanisms of MDD, involving multiple biological processes, such as neuroinflammation, neuroplasticity, and apoptosis[63]. Wu et al[64] found that miR-144-5p expression was significantly reduced in the hippocampal dentate gyrus of a mouse model of chronic unpredictable stress. MiR-144-5p upregulation in the dentate gyrus ameliorated depressive behavior in chronic unpredictable stress mice[65]. MiR-144-5p was also found to be down-regulated in the serum of MDD patients and correlated with depressive symptoms[66]. Serum exosome-derived miR-144-5p levels were consistently reduced in MDD patients[64]. Ceylan et al[67] assessed the expression levels of miRNAs in the blood of 20 patients with MDD and 20 HCs by microarray technology. They found that five miRNAs (hsa-let-7a-5p, hsa-let-7d-5p, hsa-let-7f-5p, hsa-miR-24-3p, and hsa-miR-425-3p) were abnormally expressed in MDD patients. Camkurt et al[68] investigated miRNA expression levels in peripheral blood samples of 50 patients with MDD and 41 HCs. Four miRNAs exhibited significant alterations, and miR-451a may serve as a potential biomarker for depression[68]. Research has shown that miR-16 is associated with MDD by regulating the expression of the 5-hydroxytryptamine transporter protein gene[69]. The expression of miR-16 in patients’ CSF was significantly lower than that in the control group[69].

Cui et al[70] evaluated the association between lncRNA expression in PBMCs and the risk of suicide in patients with MDD using reverse transcription-quantitative polymerase chain reaction. The results showed that the expression of six lncRNAs was significantly down-regulated and negatively correlated with the risk of suicide in patients with MDD[70]. The expression of lncRNAs in PBMCs may be helpful for clinicians to determine the risk of suicide in patients with MDD, so that they can provide timely intervention treatment and suicide prevention[70].

Dysregulation of circRNAs can contribute to neurotrophic signaling deficits and synaptic remodeling in depression by influencing central inflammatory responses and neuronal damage[71]. Wang et al[72] examined blood samples from 50 patients with MDD and 40 HCs. They discovered that the expression of circRNA protein tyrosine kinase 2 in PBMCs of MDD patients was decreased, which could assist in differentiating MDD patients from healthy individuals. Therefore, circRNA-homeodomain-interacting protein kinase-2 (circHIPK2) may be used as a potential biomarker for predicting and diagnosing MDD. The association between ncRNA and MDD is shown in Table 2.

Table 2 The association between non-coding RNA and major depressive disorder.

Ref.	Sample source	Samples	NcRNA	Method	Results
Wei et al[111], 2020	Blood exosomes	MDD: 33; HCs: 46	MiRNA	RT-qPCR	24 miRNAs up; 14 miRNAs down
Ran et al[112], 2022	Serum EVs	MDD: 9; HCs: 8	MiRNA	RNA-seq analysis; western blotting	miR-450a-2-3p up
Burrows et al[113], 2024	Peripheral blood	MDD: 41; HCs: 35	MiRNA	RT-qPCR	MDD is associated with lower NEEV miR-93 expression
Liu et al[114], 2014	Peripheral blood	MDD: 10; HCs:10	LncRNA	RNA-seq analysis; RT-qPCR	1556 lncRNA up; 441 lncRNA down
Cui et al[115], 2016	Peripheral blood	MDD: 120; HCs: 63	LncRNA	RT-qPCR	6 lncRNAs down
Yrondi et al[116], 2021	Peripheral blood	MDD: 201; HCs: 104	LncRNA	RNA-seq analysis; genome-wide DNA methylation analysis	BASP1-AS1 up
Cui et al[117], 2016	PBMCs	MDD: 30; HCs: 0	CircRNA	RT-qPCR	4 circRNAs up
Bu et al[118], 2021	Peripheral blood	MDD: 83; HCs: 83	CircRNA	RT-qPCR	hsa_circ_0126218 up
Yu et al[102], 2024	Plasma	MDD: 81; HCs: 48	CircRNA	RT-qPCR	circHIPK2 up

NcRNA: Non-coding RNA; EVs: Extracellular vesicles; PBMCs: Peripheral blood mononuclear cells; HCs: Healthy controls; MDD: Major depressive disorder; MiRNA: MicroRNA; CircRNA: Circular RNA; LncRNA: Long non-coding RNA; RT-qPCR: Reverse transcription-quantitative polymerase chain reaction; circHIPK2: Circular RNA-homeodomain-interacting protein kinase-2.

NcRNA and BD

The specific etiology of BD, a chronic mood disorder, remains largely unsolved. NcRNAs play a key role in the pathomechanism of BD. Increasing evidence suggests that lncRNAs are involved in many neurodevelopmental conditions, serving as potential therapeutic targets or biomarkers[73]. Camkurt et al[74] detected significantly elevated miR-107 and miR-125a-3p in PBMCs of 58 patients with BD compared to HCs. MiR-106a-5p and miR-107 were significantly associated with manic episodes compared to other miRNAs detected[74]. Kim et al[75] observed 15 miRNAs differentially expressed in postmortem brain tissue from the dorsolateral PFC of BD patients. The overexpressed miRNAs were all moderately to highly expressed and showed significant correlation with each other. Maffioletti et al[76] assessed the expression levels of miRNAs in the blood of 20 BD patients and 20 HCs using microarray technology. They found that five miRNAs (hsa-miR-140-3p, hsa-miR-30d-5p, hsa-miR-330-5p, hsa-miR-378a-5p, and hsa-miR-21-3p) were altered explicitly in BD patients. Tsai et al[77] found that miR-221-5p/carbonic anhydrase 1/matrix metalloproteinase 9 and miR-370-3p/phenylalanyl-tRNA synthetase beta subunit/proprotein convertase subtilisin/kexin type 9 axes may play key roles in the pathomechanism of BD-II in a study of peripheral candidate miRNAs and protein biomarkers in 96 BD-II patients. Lee et al[78] found that miR7-5p, miR221-5p, and miR370-3p were significantly associated with brain-derived neurotrophic factor levels in all patients. Zimmerman et al[79] found that circHomer1a was a neuron-rich circRNA abundantly expressed in the frontal cortex and derived from HOMER1. CircRNAs expression analyses were carried out on post-mortem total brain RNA samples sourced from the orbital frontal cortex of 32 individuals diagnosed with BD[79]. The findings indicated that the expression levels of circHomer1a were significantly reduced in neuronal cultures generated from induced pluripotent stem cells[79]. The association between ncRNA and BD is shown in Table 3.

Table 3 The association between non-coding RNA and bipolar disorder.

Ref.	Sample source	Samples	NcRNA	Method	Results
Fiorentino et al[119], 2016	Peripheral blood	BD: 2078; HCs: 1355	MiRNA	RT-qPCR	A single recurrent variant in the vicinity of the miR-708 gene
Mola-Ali-Nejad et al[120], 2023	Peripheral blood	BD: 50; HCs: 50	LncRNA	RT-PCR	NKILA and COX2 down
Lin et al[121], 2021	Post-mortem human; brain samples	BD: 13; HCs: 13	CircRNA	RT-qPCR; RNA-seq analysis	26 circRNAs up
Hosseini and Mokhtari[122], 2024	Peripheral blood	BD: 50; HCs: 50	LncRNA	qRT-PCR	HOXA-AS2 and MEG3 up
Zamani et al[123], 2023	Peripheral blood	BD: 50; HCs: 50	LncRNA	RT-qPCR	FOXD3-AS1 and GAS5 up
Maffioletti et al[76], 2016	Peripheral blood	BD: 20; HCs: 20	MiRNA	Microarray	5 miRNAs up; 2 miRNAs down
Ceylan et al[67], 2020	Peripheral blood	BD: 69; HCs: 41	MiRNA	RT-qPCR	3 miRNAs down; miR-185-5p up
Zimmerman et al[79], 2020	Human postmortem brain; total RNA samples	BD: 32; HCs: 34	CircRNA	RT-qPCR	CircHomer1a down
Mahmoudi et al[109], 2021	PBMCs	BD: 39; HCs: 20	CircRNA	RNA-seq	55 circRNAs down

NcRNA: Non-coding RNA; PBMCs: Peripheral blood mononuclear cells; HCs: Healthy controls; BD: Bipolar disorder; MiRNA: MicroRNA; CircRNA: Circular RNA; LncRNA: Long non-coding RNA; RT-qPCR: Reverse transcription-quantitative polymerase chain reaction; qRT-PCR: Quantitative real-time polymerase chain reaction; RT-PCR: Real-time polymerase chain reaction.

EVS

EVs are a diverse group of subcellular, closed membrane structures with a phospholipid bilayer, which are secreted by cells and detectable in various body fluids, such as blood, urine, and saliva[80]. These vesicles, typically having diameters ranging from 30 nm to 5000 nm, play a crucial and irreplaceable role in mediating communication between the external environment and organisms, as well as interactions among diverse cell types[80]. They have now gained extensive recognition as pivotal carriers of cellular signaling. EVs can be categorized into subgroups, including exosomes (30-150 nm), microvesicles (150-1000 nm), and apoptotic vesicles (1000-5000 nm)[81]. In recent years, there has been increasing interest in the potential roles of EVs in various biological processes. The biological functions of EVs are primarily determined by their cell of origin. EVs contain a variety of biological molecules, highlighting their potential for use in medical diagnostics and therapeutic applications[82].

Biogenesis of EVs

EVs can be produced by a wide range of organisms, spanning from prokaryotes to higher eukaryotes[83]. These vesicles play essential physiological and pathological roles in organisms. Surprisingly, almost all cell types can produce EVs[83]. Bacteria in prokaryotes can release EVs by outgrowth, whereas cells in eukaryotes produce EVs by endocytosis and cytotoxicity in higher eukaryotes, such as mammals. EV production is closely linked to several physiological processes, including the immune response, cellular communication, and waste removal[84].

The process of EV development varies depending on the type, but generally involves the following steps. Initially, the cell membrane invaginates to form primary endosomes, which then develop into more mature secondary endosomes. These secondary endosomes then evolve into more mature secondary endosomes, which eventually develop into multivesicular bodies (MVBs) containing multiple small vesicles. During the formation of these vesicles, exosomes, which are primarily derived from MVBs, are released to the outside of the cell by fusion with the cytoplasmic membrane[85]. In addition, microvesicles are formed by budding and shedding directly from the cytoplasmic membrane. This process also enables MVB to bind to other organelles, such as lysosomes or autophagosomes, and participate in the degradation and recycling of substances[86]. This ubiquitous phenomenon suggests that EVs have a wide range of essential functions in the biological world. Due to their disease-specific and highly stable EVs and their contents, they may also serve as ideal biomarkers for various diseases[87,88].

NCRNA IN EVS AND MAJOR PSYCHIATRIC DISORDERS

EVs are involved in neurological disease processes by carrying a variety of biologically active molecules (e.g., ncRNAs) for intercellular delivery[12]. These ncRNAs can be selectively loaded and transported by exosomes to regulate processes such as cell growth, differentiation, development, and apoptosis. Due to the load diversity of exosomes and their ability to bi-directionally penetrate the blood-brain barrier, they have emerged as potential non-invasive biomarkers for the screening, diagnosis, and treatment of brain disorders[89,90]. Therefore, targeting strategies that specifically target ncRNAs in EVs may become a promising therapeutic option. Krylova and Feng[91] found that ncRNAs are present in blood, CSF, saliva, and urine, and that they can be released into the circulation in free form or packaged in exosomes. Although a large proportion of ncRNAs are composed of ribosomal RNAs and transfer RNAs involved in the translation process, it is increasingly recognized that non-coding products contain a wide range of species associated with core biological processes and diseases. To date, the pathology and action mechanisms of most psychiatric disorders are unclear. The pathogenic factors of psychiatric disorders primarily encompass neuroinflammation and dysregulation of the internal environmental homeostasis within the central nervous system. Remarkably, EVs have been shown to have a significant impact on these pathological processes (Figure 1). EVs are capable of mapping the immediate state of the brain and specific clinical phenomena, such as non-suicidal self-mutilation and cognitive dysfunctions[92].

Figure 1 Schematic diagram of extracellular vesicles-mediated psychiatric disorders. External stress influences the expression of intracellular non-coding RNAs, which are precisely packaged into exosomes. Due to the load diversity of exosomes and their ability to bi-directionally penetrate the blood-brain barrier, they have emerged as potential non-invasive biomarkers for the screening, diagnosis, and treatment of brain disorders. SCZ: Schizophrenia; MDD: Major depressive disorder; BD: Bipolar disorder; BBB: Blood-brain barrier; MiRNA: MicroRNA; CircRNA: Circular RNA; LncRNA: Long non-coding RNA; MVB: Multivesicular body.

Some studies found miRNA dysregulation in EVs of brain tissue[75], CSF[69], and peripheral blood[93] after death in patients with SCZ or other psychiatric disorders. For example, Du et al[94] were the first to screen 18 differentially expressed miRNAs from blood exosomes of SCZ patients, suggesting that EV-derived ncRNAs hold greater potential for clinical application. Some studies have suggested that specific miRNAs in exosomes from blood may contribute to the pathogenesis of SCZ, including abnormalities in protein glycosylation, neurotransmitter biosynthesis, and dendritic development, which may lead to the remodeling of the brain's structural and functional phenotype[95]. For example, miR-132 in EVs was decreased in SCZ and increased in MDD and BD[95], suggesting different pathological mechanisms between BD and SCZ. Additionally, miR-132 has a broad range of effects on the nervous system[96]. It not only controls neuronal differentiation, maturation, and function, but also profoundly intervenes in axon extension and neuroplasticity, which in turn protects neurogenesis and related memory functions in the hippocampal region from defects[96]. MiR-130b and miR-193a-3p were found to be upregulated in SCZ patients[63]. These miRNAs were involved in pathological mechanisms, including the dysregulation of the protein kinase B pathway and the production of cytokines in psychiatric patients[63]. This involvement provides a plausible explanation for the overlapping symptom profiles observed across various psychiatric disorders[63].

DISCUSSION

Over the past decade, there has been an increasing effort to develop exosome-based biomarkers to fill the gap in the early diagnosis of human psychiatric disorders. NcRNAs in EVs are stable in body fluids and can be used as biomarkers for disease diagnosis, prognosis, and response to therapy. Abnormally expressed miRNAs, lncRNAs, and circRNAs offer new ideas and approaches for the development of minimally invasive diagnostic tests for early detection and monitoring of brain diseases. In addition, the development of novel therapeutic interventions aimed at improving clinical outcomes is expected for aberrantly expressed ncRNAs[25].

Recent studies on three major psychiatric disorders have revealed key findings of abnormal ncRNA expression in patients with SCZ, MDD, and BD. Expression alterations of miRNAs such as miR-132, miR-144-5p, and miR-16 were detected in patients’ blood, CSF, or brain tissue[65,69,96]. Specific miRNAs, such as miR-132, are downregulated in SCZ but upregulated in MDD and BD, suggesting their involvement in disease-specific pathways. Furthermore, these miRNAs may serve as biomarkers for distinct subtypes within the same disorder[97]. The phenomenon of the same miRNA exhibiting differential expression across disorders reflects significant molecular heterogeneity and overlapping pathological mechanisms among psychiatric conditions. These discrepancies may stem from multiple factors, including differences in sample sources (brain tissue, blood) or methodological variations (reverse transcription-quantitative polymerase chain reaction, RNA-seq analysis).

The discovery of the roles of ncRNAs in psychiatric disorders such as SCZ, MDD, and BD has deepened our understanding of disease mechanisms and opened new avenues for therapeutic exploration[74,98]. NcRNAs mechanistically influence pathophysiological processes through diverse mechanisms. For instance, miR-132, a well-studied miRNA, regulates synaptic plasticity by modulating dendritic spine formation and neuronal connectivity, which are critical for cognitive function and are often impaired in psychiatric disorders[99-101]. Similarly, Yu et al[102] found that circHIPK2 was involved in depression by inducing hippocampal astrocyte dysfunction in depressed mice, which resulted in high plasma and hippocampal expression in depressed mice, and circHIPK2 levels decreased with the alleviation of depressive symptoms in mice. Further studies showed that the expression of circHIPK2 was significantly increased in the plasma of MDD patients and may be involved in a variety of signal pathways, including the P53 signal pathway, the AMP-activated protein kinase signal pathway, the tumor necrosis factor signal pathway, and the Ras signal pathway, which are thought to regulate glial cell function[103].

The levels of ncRNAs in EVs often reflect disease progression and severity. For example, specific miRNAs exhibit altered expression at distinct stages of SCZ, ranging from early mild cognitive deficits to later, more severe clinical manifestations[98]. Similarly, the up- or downregulation of ncRNAs during various pathological processes underscores their potential as therapeutic targets[92]. These findings suggest that ncRNAs could revolutionize the diagnosis, prognosis, and treatment of psychiatric disorders.

EVs play dual biological roles: They serve as carriers of disease-specific ncRNAs for early diagnosis and as low-immunogenicity, targeted delivery systems for therapeutic nucleic acids and small-molecule drugs[104]. Their ability to cross the blood-brain barrier and deliver therapeutic molecules, such as neurotrophic factors or CRISPR-Cas9 components, to specific brain regions overcomes the limitations of conventional drug delivery methods[89]. This unique property positions EVs as a promising tool for treating central nervous system diseases.

Currently, the research and application of ncRNAs in EVs face multiple challenges. For example, the diversity of exosome isolation methods results in significant differences in purity, concentration, and isoform composition, complicating direct comparisons across studies[105,106]. Additionally, complex body fluids like blood contain contaminants such as lipoproteins and protein aggregates, which are similar in size and density to exosomes and can interfere with ncRNA sequencing analyses, leading to false-positive or false-negative results. Therefore, replicated validation in larger, multi-ethnic cohorts is essential to advance clinical translation.

While existing studies have provided initial insights into the link between ncRNAs and psychiatric disorders, these findings require further exploration and validation to establish a robust scientific foundation. First, the heterogeneity of EVs and the absence of standardized isolation and characterization methods present significant challenges for reproducibility and cross-study comparisons. Second, the limited understanding of the specific functional roles of ncRNAs in psychiatric disorders underscores the need for more in-depth mechanistic investigations. Finally, the current findings necessitate validation in larger, longitudinal cohorts to confirm their translational potential and clinical applicability. Addressing these limitations will be critical for advancing the field and realizing the therapeutic and diagnostic promise of ncRNAs in psychiatric disorders.

CONCLUSION

NcRNAs play a pivotal role in psychiatric disorders by regulating gene expression and mediating key biological pathways, offering novel insights into the molecular mechanisms underlying these conditions. The analysis of small and long RNA profiles in serum and plasma-derived exosomes has identified specific miRNAs, circRNAs, and lncRNAs as potential biomarkers for major psychiatric disorders. Aberrant expression of EV-associated ncRNAs has been consistently observed across various psychiatric disorders, highlighting their utility as a promising source of disease biomarkers. These findings not only enhance our understanding of the molecular basis of psychiatric disorders but also pave the way for the development of innovative diagnostic tools and therapeutic strategies. Further research is essential to validate these biomarkers and explore their clinical applications, ultimately improving patient outcomes.