Skip to main content
NCBI home page
Brain, Behavior, & Immunity - Health logoLink to Brain, Behavior, & Immunity - Health
. 2026 Feb 4;52:101190. doi: 10.1016/j.bbih.2026.101190

Dysregulated miRNAs and downstream gene expression associated with poor treatment response in first-episode psychosis

Shun-Chun Yu a,b,c,1, Yun-Chu Wang a,b,1, Hsiu-Ping Lin b, Ya-Wen Jen a,c, Tzung-Jeng Hwang d,e, Chih-Min Liu d, Hung-Yu Chan f, Chian-Jue Kuo g, Tsung-Tsair Yang h, Jen-Pang Wang i, Chen-Chung Liu d, Ming H Hsieh d, Yi-Ting Lin d, Yi-Ling Chien d, Po-Hsiu Kuo a,b, Ya-Wen Shih a, Sung-Liang Yu j, Hsuan-Yu Chen k, Charlotte Wang a,l, Wei J Chen a,b,c,d,⁎
PMCID: PMC12906215  PMID: 41694232

Abstract

Background

Treatment response in first-episode psychosis (FEP) is highly variable, and reliable biomarkers for poor outcomes remain limited. MicroRNAs (miRNAs), important post-transcriptional regulators, have been implicated in psychotic disorders. However, genome-wide miRNA profiling and analyses of their downstream gene networks related to treatment response in FEP remain insufficiently explored.

Methods

We analyzed baseline miRNA expression in peripheral blood mononuclear cells from 41 antipsychotic-naïve or minimally treated FEP patients with six-month follow-up. Patients were classified as good (n = 17) or poor responders (n = 24) based on ≥20% symptom improvement on the Positive and Negative Syndrome Scale. Differentially expressed miRNAs were identified by microarray. RNA sequencing was performed to detect candidate target genes, followed by differential expression and functional enrichment analyses.

Results

Hsa-miR-34a and hsa-miR-299 were significantly associated with 6-month treatment response. RNA sequencing identified candidate target genes regulated by these miRNAs (704 for hsa-miR-34a and 262 for hsa-miR-299). After multiple-testing correction, three hsa-miR-34a– and five hsa-miR-299–related genes were expressed at higher baseline levels in poor responders than in good responders; their expression levels after 6 months in poor responders remained similar or slightly decreased, whereas in good responders G3BP1, PARD6B, DDHD2, and SLC25A4 showed significant reduction and C14orf28, RSBN1, CDC16, and PPM1K showed little or no reduction. These genes are involved in neural development, neural maintenance, and immune response.

Conclusion

Our multi-omics approach helps identify miRNAs and their downstream target genes associated with FEP patients’ poor responses to antipsychotics, highlighting potential biomarkers for personalized therapy.

Keywords: microRNAs, RNA sequencing, Gene expression, Target genes, Treatment response

Highlights

  • •

    FEP patients with poor treatment response showed downregulation of miR-34a-5p and miR-299-5p.

  • •

    Eight genes exhibited differential expression between poor and good responders.

  • •

    G3BP1, PARD6B, DDHD2, and SLC25A4 remained overexpressed in poor responders.

  • •

    These four genes are involved in neural development, neural maintenance, and immune responses.

1. Introduction

Psychotic disorders are characterized by the loss of reality that results in delusions, hallucinations, and disorganized thoughts (Lieberman and First, 2018) and have a lifetime prevalence of approximately 3% worldwide (Perala et al., 2007). Antipsychotics are the first-line treatment for psychotic disorders (Keepers et al., 2020; National Collaborating Centre for Mental Health, 2014), but approximately 30% of patients show poor response (Demjaha et al., 2017; Gillespie et al., 2017). Although certain clinical features, such as a longer duration of untreated psychosis (Crespo-Facorro et al., 2013; Demjaha et al., 2017), family history (Crespo-Facorro et al., 2013), and early age at onset (Demjaha et al., 2017; Iasevoli et al., 2022), have been associated with poor treatment response, reliable biomarkers for predicting poor response in first-episode psychosis (FEP) remain limited, representing a major unmet medical need.

MicroRNAs (miRNAs) are post-transcriptional regulators of gene expression that modulate mRNA stability and translation (Bartel, 2004; O'Brien et al., 2018) and are enriched in the mammalian brain, where they play key roles in neurogenesis and neuronal differentiation. Following early evidence of dysregulated miRNAs in the postmortem prefrontal cortex of schizophrenia patients (Perkins et al., 2007), subsequent studies have reported aberrant miRNA in peripheral tissues, including peripheral blood mononuclear cells (PBMCs) (Chen et al., 2016; Gardiner et al., 2012; Lai et al., 2011, 2016), plasma (Song et al., 2014; Sun et al., 2015; Wei et al., 2015), and serum (He et al., 2019).

Several studies have further examined whether miRNA expression associated with psychotic disorders changes following antipsychotic treatment. These studies have either compared candidate miRNA levels before and after treatment (Chen et al., 2016; Lai et al., 2016; Liu et al., 2013; Song et al., 2014) or examined differential miRNA expression between patients with good versus poor treatment outcomes (Huang et al., 2021; Sun et al., 2015; Wei et al., 2015).

However, identifying miRNAs associated with treatment response does not directly translate into actionable biological insight, as individual miRNAs regulate multiple target genes and their temporal responses to environmental or pharmacological changes remains unclear. An alternative strategy is to leverage differentially expressed miRNAs to prioritize downstream target genes that may be more directly involved in treatment response. To date, few studies have conducted genome-wide miRNA profiling in FEP, and none have systematically examined the expression of downstream target genes associated with treatment response.

To address this gap, we recruited antipsychotic-naïve or minimally treated patients with FEP and classified them as poor responders if they showed <20% symptom reduction at the 6-month follow-up. We aimed to identify baseline miRNAs differentially expressed between poor and good responders using microarray profiling with reverse transcriptase quantitative real-time polymerase chain reaction (RT‒qPCR) validation, and to examine their downstream effects on target gene expression using RNA sequencing (RNA-seq), which enables sensitive detection of transcriptional changes in modest sample sizes.

2. Methods and materials

2.1. Participants

Patients were drawn from an ongoing prospective cohort study of FEP (Yu et al., 2024). Briefly, individuals meeting the Diagnosis and Statistical Manual of Mental Disorders, Fifth Edition criteria for a psychotic disorder were recruited from outpatient clinics and inpatient psychiatric wards in northern Taiwan, including National Taiwan University Hospital, Taipei Psychiatric Center, Taoyuan Psychiatric Center of Ministry of Health and Welfare, and two private clinics. Participants were aged 15–45 years, of Han Chinese ancestry, had experienced their first psychotic episode within one year, and were antipsychotic-naïve or minimally treated (<3 months). Exclusion criteria included organic brain disorders, major medical illness, intellectual disability, substance-induced psychosis, or steroid use. All participants provided written informed consent after receiving a complete description of the study. This study conformed to the principles embodied in the Declaration of Helsinki and was approved by the Research Ethics Committees of the participating institutions (National Taiwan University Hospital: 20150203RINC; Taipei City Hospital: TCHIRB-10501107; and Taoyang Psychiatric Center: B20151222).

Between January 2016 and May 2019, 80 out of 135 eligible patients were recruited. Participants with available 6-month follow-up data (n = 41) were included in the present analyses. These individuals underwent genome-wide miRNA profiling and transcriptome-wide sequencing of PBMCs, which were used as an accessible surrogate tissue for investigating molecular correlates of treatment response. Sociodemographic and clinical characteristics did not differ significantly between included and excluded participants, except that excluded participants were older (mean age, 27.4 years [SD, 6.1]) than included participants (mean age, 23.8 years [SD, 5.2]; P = 0.005) (Supplementary Table S1).

2.2. Measurements

Collected variables included sociodemographic characteristics (sex, age, education), clinical features (duration of untreated psychosis, family history of any psychiatric disorders in first-degree relatives, baseline antipsychotics dosage converted to chlorpromazine equivalents (Leucht et al., 2016), and symptom severity at baseline and 6 months. Duration of untreated psychosis was defined as the interval between the onset of first psychotic symptoms and initiation of antipsychotic treatment. This study recorded only medication as prescribed by the treating physicians, reflecting a treatment-as-prescribed approach. Symptom severity was assessed using the Chinese version (Cheng et al., 1996) of the Positive and Negative Syndrome Scale (PANSS) (Kay et al., 1987), administered by attending psychiatrists at both time points.

2.3. Definition of poor responders

Treatment response was defined based on longitudinal change in symptom severity. Patients were classified as poor responders if they failed to achieve a ≥20 % reduction in PANSS total score at the 6-month follow-up. This threshold was chosen to capture clinically meaningful improvement while maintaining sufficient group sizes. Use of a relative change-based criterion was supported by the heterogeneous distribution of PANSS change scores in the sample (Supplementary Fig. S1) and allowed assessment of treatment response across positive, negative, and general psychopathology domains. Although previous study has reported heterogeneous response patterns using the PANSS positive subscale over shorter follow-up periods (Troudet et al., 2020), our approach extended this concept to the PANSS total score to reflect treatment response across positive, negative and general psychopathology domains, which particularly relevant in a mixed inpatient-outpatient FEP cohort.

2.4. Laboratory procedures

PBMCs were selected to capture peripheral immune-related transcriptomic variation potentially relevant to psychosis and treatment response. Detailed laboratory procedures, including RNA isolation, miRNA microarray profiling, RT‒qPCR validation, RNA sequencing, and functional enrichment analyses, are provided in the Supplementary Information.

Venous blood samples (∼10 mL) were collected at baseline (n = 41) and 6 months (n = 33). Total RNA was extracted from PBMCs within 3 h of collection using TRIzol (Invitrogen, Grand Island, NY, USA) and stored at −80 °C. RNA quality was high (RNA integrity number >8).

Microarray-based miRNA profiling was performed using the GeneChip™ miRNA 4.0 Array (Thermo Fisher Scientific, Waltham, MA, USA), with standard preprocessing and normalization procedures. Criteria for miRNA expression detection were adjusted to enhance stringency (p < 0.01; >25% probes expressed). Differentially expressed miRNAs were validated using RT‒qPCR, with RNU48 used as an internal control. Relative expression levels were calculated using the 2ˆ(-ΔCt) method, where Ct denotes the cycle threshold.

RNA sequencing was performed using paired-end 150 bp reads on the Illumina NovaSeq 6000 platform (Mundelein, Illinois, USA). Unqualified sequencing reads (quality score <30) were removed using Trimmomatic software. Then, the sequence reads that passed QC were aligned to the reference genome (GRCH38) using HISAT2 (version 2.2.1) (D Kim et al., 2019). Normalization was performed using DESeq2 (Love et al., 2014). Selected mRNA expression levels were validated by RT‒qPCR using β-actin as the reference gene. RT‒qPCR was performed in triplicate using a StepOnePlus™ Real-Time PCR System (Thermo Fisher Scientific, Waltham, MA, USA).

Enrichment analyses were conducted using the GENE2FUNC function from the Functional Mapping and Annotation platform (https://fuma.ctglab.nl/) and STRING version 12.0 (https://version-12-0.string-db.org). For tissue-specific enrichment analysis, we used gene expression datasets from GTEx version 8 (The GTEx Consortium, 2020), which consists of tissue-specific samples. The Gene Ontology (GO) database was used to assess enrichment in biological processes and cellular components. Additionally, cell-type enrichment analysis was performed on target genes that passed multiple testing correction using the PsychSCREEN gene portal (https://psychscreen.beta.wenglab.org/psychscreen/gene).

2.5. Statistical analysis

Statistical analyses were performed using R, version 4.0.2 (R Core Team, 2020). Baseline characteristics were compared using t tests. chi-square tests, or Fisher's exact tests, as appropriate. Differential miRNA expression between poor and good responders was assessed using logistic regression adjusted for age and recruitment source. MiRNAs with fold change >1.55 and p < 0.05 were selected for RT‒qPCR validation.

Predictive performance of validated miRNAs was evaluated using receiver operating characteristic curves and area under the curve (AUC). Sensitivity analyses using alternative PANSS cutoffs were conducted to assess robustness.

Target genes of validated miRNAs were identified using TargetScan v7.1 (McGeary et al., 2019) and tested for differential expression between response groups. Multiple testing correction was performed using the false discovery rate (Benjamini et al., 2001). Multivariable logistic regression models adjusted for age and recruitment source were also fitted. Effect sizes (Cohen's d) were calculated for the comparison of mean gene expression change using group means and pooled standard deviations, and post-hoc power estimates were subsequently determined for each finding (alpha = 0.05) using G∗Power (Faul et al., 2009).

3. Results

3.1. Sample demographic characteristics

Among 41 patients, 17 (41.5%) were classified as good responders and 24 (58.5%) as poor responders. Baseline sociodemographic and clinical characteristics were largely comparable between groups (Table 1). The majority of participants had received antipsychotic medication (with treatment duration <3 months and none receiving clozapine), with only 2 (5%) being antipsychotic-naïve. However, the source of recruitment was associated with symptom improvement status (P = 0.021) and was therefore included as a covariate in subsequent models. Because the ages of these patients with first-episode psychosis were close to their age at onset, which has been inversely associated with poorer treatment response in the literature, and because participants excluded from the present analyses were older than those included, age was additionally included as a covariate in all models. Due to the limited number of participants with a family history of psychosis and the inclusion criteria restricting the sample to recent-onset cases, family history and duration of untreated psychosis were not included as covariates.

Table 1.

Demographic characteristics, source of recruitment, and clinical characteristics at baseline of patients with first-episode psychosis who had symptom ratings at 6-month follow-up in northern Taiwan recruited from 2016 to 2019, by their symptom improvement status.

Variable All patients (N = 41)
 Poor responders (N = 24)
 Good responders (N = 17)
 Group comparisona
n % n % n % P
Sex
 Male 15 37% 7 29% 8 47% 0.328
 Female 26 63% 17 71% 9 53%
Recruitment source
 Inpatient 15 37% 5 21% 10 59% 0.021∗
 Outpatient 26 63% 19 79% 7 41%
Diagnosis at baseline
 Schizophrenia 22 54% 12 50% 10 59% 0.678
 Other nonaffective psychosisb 14 34% 8 33% 6 35%
 Affective psychosis 5 12% 4 17% 1 6%
Family history 0.141
 Yes 5 12% 1 4% 4 24%
 No 36 88% 23 96% 13 76%
Antipsychotic medicationc 0.158
 Monotherapy (atypical) 23 56% 16 66% 7 41%
 Polytherapy (atypical) 6 15% 4 17% 2 12%
 Polytherapy (typical + atypical) 10 24% 4 17% 6 35%
 No antipsychotic medication 2 5% 0 0% 2 12%
Mean SD Mean SD Mean SD P
Age 23.8 5.2 24.2 5.2 23.2 5.2 0.549
Educational level, years 14.6 2.8 15.2 2.6 13.7 3.0 0.100
DUP, days 109.5 118.2 134.0 112.9 76.4 120.3 0.129
Chlorpromazine equivalentsd 180.5 122.9 153.4 98.0 218.7 146.0 0.094
Open in a new tab

∗p < 0.05.

a

Group comparisons were conducted using χ2 test or Fisher's exact test for categorical variables and t-test for continuous variables.

b

Including schizotypal personality disorder (n = 1), delusional disorders (n = 2), brief psychotic disorder (n = 3), schizoaffective disorder (n = 1), other psychotic disorder not due to a substance or known physiological condition (n = 1), and unspecified schizophrenia spectrum and other psychotic disorder (n = 6).

c

Prescribed antipsychotic medications during the period from baseline to 6-month follow-up.

d

Prescribed antipsychotic medications at baseline were collected and converted to chlorpromazine equivalents.

Baseline symptom severity is shown in Table 2. Overall, patients exhibited relatively mild psychotic symptoms at baseline, with a mean PANSS total score of 66.6. Inpatients showed significantly higher baseline positive symptom scores than outpatients. PANSS scores stratified by treatment response and recruitment source are present in Table 3. Compared with good responder, poor responders had significantly lower baseline positive symptom scores. This difference was primarily attributable to recruitment source, as poor responders were more likely to be recruited from outpatient settings, where earlier access to psychiatric care is common in Taiwan.

Table 2.

PANSS total score and three subscores at baseline and 6-month follow-up and changes in all FEP patients and stratified by the source of recruitment.

FEP patients
 Inpatients
 Outpatients
 Pa
(N = 41)
 (N = 15)
 (N = 26)
Mean (SD) Mean (SD) Mean (SD)
Baseline
 Total score 66.6 (21.8) 73.1 (21.6) 62.8 (21.4) 0.120
 Positive subscore 15.9 (6.8) 19.7 (7.2) 13.6 (5.6) 0.004∗
 Negative subscore 16.5 (7.4) 15.5 (7.6) 17.1 (7.3) 0.636
 General psychopathology subscore 34.2 (10.8) 37.9 (10.6) 32.1 (10.6) 0.082
6-month follow-up
 Total score 53.7 (20.5) 49.1 (20.1) 56.1 (20.6) 0.337
 Positive subscore 11.0 (5.1) 10.2 (5.3) 11.5 (5.1) 0.456
 Negative subscore 15.3 (7.0) 13.6 (6.2) 16.3 (7.3) 0.277
 General psychopathology subscore 27.4 (10.9) 25.7 (11.3) 28.4 (10.7) 0.453
Changes from baseline to 6-month follow-up
 Total score −12.9 (17.4)b −23.5 (21.6)b −6.7 (10.6)b 0.009∗∗
 Positive subscore −4.8 (6.9)b −9.5 (8.8)b −2.1 (3.5)b 0.006∗∗
 Negative subscore −1.2 (4.1) −1.9 (5.3) −0.8 (3.3) 0.389
 General psychopathology subscore −6.8 (10.2)b −12.1 (13.1)b −3.8 (6.6)b 0.026∗
Open in a new tab

PANSS, Positive and Negative Syndrome Scale.

ap-value for comparison between inpatients and outpatients (t-test).

bp-value <0.05 for comparison between baseline and 6-month follow-up (paired t-test).

∗p value < 0.05; ∗∗p value < 0.01.

Table 3.

PANSS total score and three subscale scores at baseline and 6-month follow-up and changes in all FEP patients and stratified by the source of recruitment and treatment response.

PANSS All FEP patients
 Pa Inpatients
 Pa Outpatients
 Pa
Poor responders
 Good responders
 Poor responders
 Good responders
 Poor responders
 Good responders
(N = 24)
 (N = 17)
 (N = 5)
 (N = 10)
 (N = 19)
 (N = 7)
Mean (SD) Mean (SD) Mean (SD) Mean (SD) Mean (SD) Mean (SD)
Baseline
 Total score 61.9 (21.9) 71.8 (21.6) 0.157 63.8 (28.4) 77.7 (17.3) 0.255 61.4 (20.7) 63.4 (25.7) 0.835
 Positive scale 14.0 (5.5) 18.5 (7.8) 0.036∗ 16.0 (5.6) 21.6 (7.4) 0.163 13.4 (5.5) 14.0 (6.4) 0.822
 Negative scale 16.3 (7.8) 16.1 (6.9) 0.922 16.4 (12.5) 15.0 (4.4) 0.819 16.3 (6.6) 17.6 (9.6) 0.695
 General psychopathology scale 31.6 (10.6) 37.3 (10.7) 0.101 31.4 (11.7) 41.1 (9.0) 0.097 31.7 (10.7) 31.9 (11.1) 0.971
6-month follow-up
 Total score 60.0 (22.8) 44.5 (11.4) 0.007∗∗ 62.4 (31.0) 43.1 (8.0) 0.239 59.4 (21.2) 46.6 (15.5) 0.159
 Positive scale 12.6 (6.0) 8.8 (2.4) 0.008∗∗ 14.4 (7.8) 8.1 (1.2) 0.147 12.1 (5.5) 9.7 (3.5) 0.299
 Negative scale 16.3 (7.1) 13.6 (6.5) 0.224 14.8 (7.3) 13.0 (6.0) 0.617 16.7 (7.3) 14.4 (7.7) 0.496
 General psychopathology scale 31.1 (12.3) 22.2 (4.4) 0.003∗∗ 33.2 (17.6) 22.0 (3.7) 0.229 30.6 (11.1) 22.4 (5.6) 0.079
Changes from baseline to 6-month follow-up
 Total score −1.9 (6.5) −27.3 (17.6)b <0.001∗∗∗ −1.4 (5.8) −34.6 (17.5)b <0.001∗∗∗ −2.0 (6.8) −16.9 (12.1)b <0.001∗∗∗
 Positive scale −1.4 (3.3)b −9.7 (8.0)b <0.001∗∗∗ −1.6 (3.1) −13.5 (8.0)b 0.007∗∗ −1.3 (3.4) −4.3 (4.0)b 0.072
 Negative scale 0.0 (3.9) −2.5 (4.0)b 0.053 −1.6 (6.9) −2.0 (4.8) 0.897 0.4 (2.8) −3.1 (2.4)b 0.006∗
 General psychopathology scale −0.5 (5.3) −15.1 (9.5)b <0.001∗∗∗ 1.8 (7.9) −19.1 (8.9)b <0.001∗∗∗ −1.1 (4.5) −9.4 (7.5)b 0.002∗∗
Open in a new tab

PANSS, the Positive and Negative Syndrome Scale.

ap-value for comparison between poor responders and good responders (t-test).

bp-value <0.05 for comparison between baseline and 6-month follow-up (paired t-test).

∗p-value <0.05; ∗∗p-value <0.01; ∗∗∗p-value <0.001.

3.2. Differentially expressed miRNAs at baseline

Adjusted logistic regression of microarray miRNA expression identified 49 miRNAs differentially expressed by response status (Supplementary Table S2). We then selected the top 15 miRNAs with the highest microarray-based fold change (either >0.6 or < −0.6) for validation by RT‒qPCR (Supplementary Fig. S2). However, three miRNAs that had an RT‒qPCR-based fold change that was in the opposite direction to the corresponding microarray-based expression levels (Supplementary Fig. S3) were removed from further analysis.

For the remaining 12 miRNAs, the RT‒qPCR-based expression levels for good responders versus poor responders are displayed in scatter plots in Fig. 1. In a multivariable logistic regression analysis of the RT‒qPCR-based expression level on responder status adjusted for the source of recruitment and age, two miRNAs, hsa-miR-34a-5p (p = 0.011) and hsa-miR-299-5p (p = 0.026), were confirmed to be expressed at a lower level at baseline in poor responders compared with good responders (Supplementary Table S3). Furthermore, the RT‒qPCR-based expression levels at baseline of the two miRNAs together led to the highest area under the ROC curve, i.e., 0.858 (95% CI: 0.747-0.968) in distinguishing good responders from poor responders at the 6-month follow-up (Fig. 2).

Fig. 1.

Fig. 1

Open in a new tab

Relative expression of 12 miRNAs in the poor responders (N = 24) and the good responders (N = 17) with the highest microarray-based fold changes and a direction of fold change consistent with the RT‒qPCR analysis. The comparison between the two groups was conducted using multivariable logistic regression analysis with adjustment for patient source and age. The solid lines denote the mean ± standard deviation.

Fig. 2.

Fig. 2

Open in a new tab

The accuracy of two miRNAs (miR-34a-5p and miR-299-5p), alone or in combination, in predicting the group membership of being good responders (N = 17) versus poor responders (N = 24) was evaluated using the area under the curve (AUC) of the receiver operating characteristics (ROC) in the multivariable logistic regression model with adjustment for patient source and age.

3.3. Differentially expressed target genes at baseline

RNA-seq analysis identified 704 and 262 predicted target genes for hsa-miR-34a-5p and hsa-miR-299-5p, respectively. Consistent with lower miRNA levels in poor responders, downstream analyses focused on target genes showing higher expression in poor responders (300 for hsa-miR-34a-5p; 120 for hsa-miR-299-5p; Supplementary Table S4–S5). After false discovery rate correction, 3 hsa-miR-34a-5p targets (G3BP1, PARD6B and C14orf28) and 5 hsa-miR-299-5p targets (DDHD2, SLC25A4, RSBN1, CDC16 and PPM1K) remained significantly higher in poor responders than in good responders (Table 4), with similar results in adjusted logistic regression models (Supplementary Table S6).

Table 4.

The 3 target genes of has-miR-34a-5p and the 5 target genes of has-miR-299-5p that had significantly higher expression levels at baseline in the poor responders than in the good responders after false discovery rate correction.

Target genes Patients included for miRNA analysis (N = 41)
 Poor responders (N = 24)
 Good responders (N = 17)
 Group comparisona
Mean SD Mean SD Mean SD P
hsa-miR-34a-5p
 G3BP1 26122 2955 27752 2100 23823 2429 <0.001∗
 PARD6B 608 118 668 88 524 103 <0.001∗
 C14orf28 577 90 615 75 524 84 <0.001∗
hsa-miR-299-5p
 DDHD2 5890 747 6240 564 5396 703 <0.001∗
 PPM1K 10568 1415 1123 1291 9645 1030 <0.001∗
 RSBN1 8442 961 8829 733 7894 734 <0.001∗
 SLC25A4 934 193 1005 184 832 160 <0.001∗
 CDC16 7998 969 8367 1012 7476 619 <0.001∗
Open in a new tab

∗Reaching significance (p < 0.05) after adjustment for multiple testing using the false discovery rate.

a

Group comparisons were conducted using t-test.

To validate the differential gene expression between good responders and poor responders identified via RNAseq, we conducted RT‒qPCR for each gene and derived the corresponding fold change. All the RT‒qPCR-based fold changes in the expression of 8 genes were in the same direction as the RNAseq-based fold changes and reached statistical significance, except SLC25A4 (with a P value of 0.17) (Supplementary Fig. S4). Hence, we report RNAseq-based gene expression in subsequent analyses.

3.4. Changes in target gene expression levels from baseline to follow-up at 6 months

Of the 41 patients, 33 provided venous blood samples at the 6-month follow-up. Patients who provided blood samples did not differ from those who did not in baseline sociodemographic characteristics, diagnostic composition, antipsychotic medication status, duration of untreated psychosis, symptom severity, responder status, or expression levels of the 8 target genes identified for hsa-miR-34a-5p and hsa-miR-299-5p (Table 5). Fig. 3 depicts the distribution of the expression levels at baseline and the 6-month follow-up and the changes from baseline to 6 months of the 8 target genes of the two miRNAs for the 33 patients (20 poor responders and 13 good responders). Based on the trends from baseline to the 6-month follow-up between poor and good responders, these 8 genes could be divided into two groups.

Table 5.

Comparing the 33 patients who provided blood sample at the 6-month follow-up with the remaining 8 patients who did not in baseline sociodemographic characteristics, symptom severity, responder status, and the expression levels of the 8 target genes identified for hsa-miR-34a-5p and hsa-miR-299-5p.

Variable Patients included for miRNA analysis (N = 41)
 Patients who provided blood sample at the 6-month follow-up (N = 33)
 Patients who did not provided blood sample at the 6-month follow-up (N = 8)
 Group comparisona
n % n % n % P
Sex
 Male 15 37% 12 36% 3 37% 1.000
 Female 26 63% 21 64% 5 63%
Recruitment source
 Inpatient 15 37% 13 39% 2 37% 0.687
 Outpatient 26 63% 20 61% 6 63%
Diagnosis at baseline
 Schizophrenia 22 54% 18 55% 4 54% 1.000
 Other non-affective psychosis 14 34% 11 33% 3 34%
 Affective psychosis 5 12% 4 12% 1 12%
Family historya 0.563
 Yes 5 12% 5 15% 0 0%
 No 36 88% 28 85% 8 100%
Antipsychotic medicationb 1.000
 Monotherapy (atypical) 23 56% 17 52% 6 56%
 Polytherapy (atypical) 6 15% 5 15% 1 15%
 Polytherapy (typical + atypical) 10 24% 10 30% 0 24%
 Without antipsychotic medication 2 5% 1 3% 1 5%
Mean SD Mean SD Mean SD P
Age 23.8 5.2 23.9 4.9 23.3 6.6 0.786
Education, year 14.6 2.8 14.7 2.4 13.9 4.1 0.621
Duration of untreated psychosis, day 109.5 118.2 113.6 116.3 92.9 132.2 0.663
Open in a new tab
a

Group comparisons were conducted using Fisher's exact test for categorical variables and t-test for continuous variables.

b

Prescribed antipsychotic medications during the period from baseline to 6-month follow-up.

Fig. 3.

Fig. 3

Open in a new tab

Scatter plots of target gene expression level at baseline and 6 months follow-up, respectively, and the percentage of the change in the poor responders (N = 20) and the good responders (N = 13) for the 3 genes regulated by miR34a, (a) G3BP1, (b) PARD6B, and (c) C14orf28, and the 5 genes regulated by miR-299, (d) DDHD2, (e) SLC25A4, (f) RSBN1, (g) CDC16, and (h) PPM1K. The error bar around the mean denotes one standard deviation.

The first group of genes, including G3BP1 and PARD6B for hsa-miR-34a-5p and DDHD2 and SLC25A4 for hsa-miR-299-5p, remained at a similarly high level of gene expression from baseline to follow-up in the poor responders but significantly improved in the good responders, with effect sizes ranging from 0.87 to 1.23 and the corresponding powers ranging from 0.77 to 0.96 at an alpha level of 0.05 (more detail in Supplementary Table S7).

The second group of genes, including C14orf28 for hsa-miR-34a-5p and RSBN1, CDC16, and PPM1K for hsa-miR-299-5p, decreased or remained at a similarly high level of gene expression from baseline to the follow-up in the poor responders, but the corresponding expression level did not improve in the good responders, with effect sizes ranging from 0.04 to 1.03 and the corresponding powers ranging from 0.06 to 0.88 at an alpha level of 0.05 (more detail in Supplementary Table S7).

3.5. Tissue-specific, cellular component, and cell-type enrichment analysis

We started our enrichment analysis with the 420 target genes (300 of has-miR-34a-5p and 120 of has-miR-299-5p) across 54 tissue types from the GTEx v8 dataset. Several brain areas were predicted to have the target gene set significantly enriched among both up-regulated and down-regulated differentially expressed genes (Supplementary Fig. S5). Next, the top enriched Gene Ontology terms in STRING indicate that those target genes are involved in biological process such as regulation of cell projection organization, synapse organization, neurogenesis, and neuron projection morphogenesis, and are associated with cellular components such as synapse, synaptic membrane, pre- and post-synapse, axon, dendrite, and cell junctions (Supplementary Fig. S6).

Finally, enrichment analysis was performed on the 8 target genes to explore single-cell sequencing datasets from PsychENCODE. Cell-type-specific gene expressions were detected in 5 genes, including G3BP1, PARD6B, DDHD2, RSBN1, and CDC16, in human brain excitatory and inhibitory neurons (Fig. 4).

Fig. 4.

Fig. 4

Open in a new tab

Cell-type-specific expression of miRNA target genes in human brain single-cell RNA sequencing data from PsychENCODE. Within each panel, rows correspond to individual PsychENCODE datasets and the x-axis lists major cortical cell types. Dot color intensity reflects the mean expression level of the gene in that cell type, and dot size indicates the percentage of cells expressing the gene. Only cell types with detectable expression are displayed.

3.6. Sensitivity analysis

We tested whether the results were robust by implementing different criteria for the classification of response status, i.e., the outcome variable. When the cutoff point for the reduction in the PANSS total score was set between 11% and 22%, the result of hsa-miR-34a-5p was robust (Supplementary Table S8). Similarly, when the cutoff point for the reduction in PANSS total score was set between 15% and 22%, hsa-miR-299-5p at base line was differentially expressed between the two groups of responders.

4. Discussion

To our knowledge, this is the first study to combine genome-wide miRNA profiling with RNA-seq–based assessment of downstream target gene expression to investigate antipsychotic treatment response in FEP. At baseline, poor responders showed lower expression of hsa-miR-34a-5p and hsa-miR-299-5p and higher expression of eight predicted target genes. Over 6 months, target gene expression remained persistently elevated in poor responders, whereas several targets decreased in good responders, consistent with differential molecular adaptation associated with treatment response.

Two of the top 15 microarray-identified miRNAs were validated by RT‒qPCR, which is consistent with previously reported validation rates (Li et al., 2020; Patil et al., 2019). The observed differences in variability among miRNAs were partly attributable to the nature of relative quantification using the 2ˆ(-ΔCt) method, in which ΔCt is calculated as the Ct value of the target miRNA minus that of the internal control. When the Ct value of a given miRNA is close to that of the internal control, small fluctuations in Ct values can result in relatively large changes in calculated relative expression, leading to greater apparent variability. In contrast, miRNAs with lower expression levels (i.e., higher Ct values) tend to show smaller relative changes after normalization. Nevertheless, we implemented several quality control measures, including requiring an RNA integrity number >8 and performing all qRT-PCR reactions in technical triplicates to minimize technical variability.

Because this was a case-only design, we could not determine whether response-associated miRNA differences also distinguish patients from healthy controls. Prior case‒control studies have consistently reported upregulation of hsa-miR-34a in the prefrontal cortex (Kim et al., 2010), amygdala (Liu et al., 2018), plasma (Song et al., 2014; Sun et al., 2015), and peripheral blood (Lai et al., 2011, 2016; Smigielski et al., 2020) of patients with schizophrenia, supporting its role in disease susceptibility. In contrast, hsa-miR-299-5p has not been previously linked to psychotic disorders, although it has been reported to be downregulated in Parkinson's disease (Tolosa et al., 2018) and relapsing-remitting multiple sclerosis (Golabi et al., 2022). Together, these observations suggest that the lower baseline expression of hsa-miR-34a-5p and hsa-miR-299-5p in poor responders may reflect treatment-response biology rather than disease risk per se.

A key strength of this study is the integration of miRNA profiling with longitudinal analysis of downstream target gene expression. Functional enrichment and single-cell expression data supported involvement of neuronal and synaptic pathways, with several targets expressed in human excitatory and inhibitory neurons, lending biological plausibility to their relevance for treatment response.

For the first group of 4 genes that showed persistent overexpression in poor responders but improvement in good responders, existing evidence suggests involvement in cellular and neuronal processes that may constrain adaptive responses to antipsychotic treatment:

  • •

    G3BP1 plays a central role in stress granule assembly and translational regulation (Anderson and Kedersha, 2008; Tourrière et al., 2003) and interacts with immune signaling pathways relevant to cellular stress responses (SS-Y Kim et al., 2019; Reineke and Lloyd, 2015; Yang et al., 2019). It has also been implicated in synaptic plasticity and neurodegenerative processes, suggesting that sustained overexpression may limit activity-dependent molecular adaptation required for optimal treatment response.

  • •

    PARD6B is involved in cell polarity, asymmetric cell division, and intracellular trafficking (Lin et al., 2000; Shi et al., 2003) and has been linked to neuropsychiatric and neurodegenerative disorders (Zhang and Wei, 2022). Dysregulation of PARD6B may therefore interfere with synaptic organization and neuronal signaling plasticity following antipsychotic exposure.

  • •

    DDHD2 has been associated with motor and cognitive impairments (Inloes et al., 2014) and prioritized as a schizophrenia-related gene (Trubetskoy et al., 2022; C Zhang et al., 2023). Its role in lipid metabolism and immune regulation further suggests potential downstream effects on neuronal function.

  • •

    SLC25A4 contributes to mitochondrial function and regulation of neurotransmitter availability in synaptic regions (Ayka and Şehirli, 2020; Batool et al., 2019) and has recently been implicated in immune regulation via mitochondrial mechanisms (Zhou et al., 2024). Given the importance of mitochondrial flexibility for synaptic plasticity, persistent overexpression of SLC25A4 may limit the metabolic adaptability required for effective antipsychotic response.

Taken together, these genes appear to converge on pathways involving neuronal maintenance, stress responses, and immune regulation, highlighting potential targets for prognostic assessment or therapeutic development.

For the second group of 4 genes, the relationship with treatment response remains less clear. One possibility is that their differential expression reflects factors unrelated to treatment response per se, such as environmental exposures influencing miRNA regulation; alternatively, a longer observation period may be required to detect meaningful expression changes in good responders:

Although their roles in antipsychotic response are not well established, they represent potential candidates for future mechanistic studies.

Collectively, these targets converge on neuronal, synaptic, metabolic, and immune-related processes previously implicated in psychosis, and our findings suggest that persistent dysregulation of these pathways may be associated with poor response to currently available antipsychotic medications in FEP.

5. Limitations and future directions

Although age was the only characteristic that differed between participants who completed the 6-month follow-up and those who did not, and no detectable differences in baseline sociodemographic or clinical characteristics were observed between patients who provided blood samples at the 6-month follow-up and those who did not, residual attrition bias due to unmeasured factors cannot be ruled out. Participant dropout and exclusion reduced the effective sample size, underscoring the need for replication in independent cohorts. Antipsychotic medications were often prescribed in combination, particularly in inpatient settings, precluding analyses of individual drug effects. In addition, the small number of antipsychotic-naïve patients limited subgroup analyses, despite evidence that prior medication exposure can influence miRNA and gene expression. Furthermore, target genes identified in peripheral blood require validation in brain tissue, as some expression differences may reflect pharmacological rather than disease-related mechanisms.

Treatment response was defined by longitudinal change in PANSS total score rather than baseline severity, and baseline symptom differences did not determine response status. A ≥20% PANSS reduction at 6 months was used as the primary threshold, supported by the response distribution in this cohort. Sensitivity analyses using alternative PANSS cutoffs yielded broadly consistent associations for hsa-miR-34a-5p and hsa-miR-299-5p, although differences in cohort composition, treatment regimens, and assessment intervals may limit generalizability.

PBMCs do not directly model molecular processes within neurons or glia, limiting inference regarding central nervous system (CNS) cell-type-specific mechanisms. However, immune dysregulation and inflammatory signaling are increasingly recognized as core features of psychosis (Capuzzi et al., 2017; Chen et al., 2024; de Bartolomeis et al., 2022; Pardo-de-Santayana et al., 2021; L Zhang et al., 2023), and transcriptomic studies suggest that PBMCs and brain tissue can share dysregulated pathways at the systems level (Goossens et al., 2021; Song et al., 2021). Consistent alterations of miRNA expression in peripheral tissues across schizophrenia studies (Chen et al., 2016; Gardiner et al., 2012; He et al., 2019; Lai et al., 2011, 2016; Song et al., 2014; Sun et al., 2015; Wei et al., 2015) support the utility of PBMC-based profiling for identifying biologically relevant molecular signatures associated with illness characteristics and treatment response.

Despite these limitations, this study demonstrates a multi-omics strategy for narrowing candidate pathways associated with poor antipsychotic response by integrating miRNA profiling with downstream target gene expression. The identified targets converge on neuronal, metabolic, and immune-related processes rather than direct neurotransmitter signaling, suggesting alternative biological mechanisms underlying treatment resistance. Future studies integrating peripheral miRNA data with additional molecular layers, such as lipidomics, metabolomics, and CNS-derived datasets, will be critical for advancing mechanistic understanding.

From a translational perspective, molecular expression trajectories observed during early treatment may provide preliminary indicators of antipsychotic response. If validated in independent cohorts, such biomarkers could support earlier identification of individuals at risk for poor response and inform the development of personalized therapeutic strategies in first-episode psychosis.

CRediT authorship contribution statement

Shun-Chun Yu: Writing – review & editing, Writing – original draft, Project administration, Methodology, Investigation, Formal analysis, Data curation. Yun-Chu Wang: Writing – review & editing, Writing – original draft, Validation, Methodology, Investigation, Formal analysis, Data curation. Hsiu-Ping Lin: Writing – review & editing, Validation, Methodology, Data curation. Ya-Wen Jen: Writing – review & editing, Methodology, Formal analysis. Tzung-Jeng Hwang: Writing – review & editing, Resources, Methodology, Investigation. Chih-Min Liu: Writing – review & editing, Resources, Methodology, Investigation. Hung-Yu Chan: Writing – review & editing, Resources, Methodology, Investigation. Chian-Jue Kuo: Writing – review & editing, Resources, Methodology, Investigation. Tsung-Tsair Yang: Writing – review & editing, Resources, Methodology, Investigation. Jen-Pang Wang: Writing – review & editing, Resources, Methodology, Investigation. Chen-Chung Liu: Writing – review & editing, Resources, Methodology, Investigation. Ming H. Hsieh: Writing – review & editing, Resources, Methodology, Investigation. Yi-Ting Lin: Writing – review & editing, Resources, Methodology, Investigation. Yi-Ling Chien: Writing – review & editing, Resources, Methodology, Investigation. Po-Hsiu Kuo: Writing – review & editing, Resources, Methodology, Conceptualization. Ya-Wen Shih: Writing – review & editing, Project administration. Sung-Liang Yu: Writing – review & editing, Resources, Methodology. Hsuan-Yu Chen: Writing – review & editing, Resources, Methodology. Charlotte Wang: Writing – review & editing, Resources, Methodology. Wei J. Chen: Writing – review & editing, Writing – original draft, Supervision, Resources, Methodology, Funding acquisition, Conceptualization.

Funding

This study was supported by the Taiwan Ministry of Science and Technology (MOST ,104-2314-B-002-070-MY3 107-2314-B-002-214-MY3, and 109-2314-B-002-172-MY3) and the National Health Research Institutes (09A1-SP07 and 10A1-SP02).

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

The authors thank Shih-Cheng Liao, Hsi-Chung Chen, Chi-Shin Wu, I-Ming Chen, Tsung-Yang Wang, and Hung-Kuang Su at National Taiwan University Hospital; Chun-Hung Pan, Lian-Yu Chen, Po-Yu Chen, Guan-Yu Chen, and Chih-Chiang Chiu at Taipei City Hospital; and Chia-Pin Huang, Ding-Lieh Liao, An-Sheng Lin, Yu-Yuan Hung, Zhen-Yang Wang, Ying-Chih Cheng, Cheng-Shian Sung, and Kuo-Ping Li at Taoyuan Psychiatric Center for assisting with patient recruitment. The authors thank Ching-Ing Tseng, Jia-Bei Chen, Wen-Hsuan Pan, Yi-Hsuan Lin, Ching-Hsuan Tseng, Yu-Chieh Huang, Shih-Chia Yang, Wan-Jung Lui, and An-Chi Chen for their assistance with data collection. The authors thank Melissa Stauffer for editorial assistance and the array services provided by Microarray Core Laboratory of National Health Research Institutes, Taiwan.

Footnotes

Appendix A

Supplementary data to this article can be found online at https://doi.org/10.1016/j.bbih.2026.101190.

Appendix A. Supplementary data

The following is the Supplementary data to this article:

Multimedia component 1
mmc1.docx (2.1MB, docx)

Data availability

Due to ethical and consent restrictions, raw sequencing reads and individual-level clinical data cannot be publicly shared. Processed and de-identified miRNA and mRNA expression matrices and summary-level results that support the findings of this study are available from the corresponding author upon reasonable request and subject to institutional approval. All analysis code and scripts used for data processing and statistical analyses are also available upon request to facilitate reproducibility.

References

  1. Anderson P., Kedersha N. Stress granules: the Tao of RNA triage. Trends Biochem. Sci. 2008;33(3):141–150. doi: 10.1016/j.tibs.2007.12.003. [DOI] [PubMed] [Google Scholar]
  2. Ayka A., Şehirli A. The role of the SLC transporters protein in the neurodegenerative disorders. Clin. Psychopharmacol. Neurosci. 2020;18(2):174–187. doi: 10.9758/cpn.2020.18.2.174. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bartel D.P. MicroRNAs: genomics, biogenesis, mechanism, and function. Cell. 2004;116(2):281–297. doi: 10.1016/s0092-8674(04)00045-5. [DOI] [PubMed] [Google Scholar]
  4. Batool S., Raza H., Zaidi J., Riaz S., Hasan S., Syed N.I. Synapse formation: from cellular and molecular mechanisms to neurodevelopmental and neurodegenerative disorders. J. Neurophysiol. 2019;121(4):1381–1397. doi: 10.1152/jn.00833.2018. [DOI] [PubMed] [Google Scholar]
  5. Benjamini Y., Drai D., Elmer G., Kafkafi N., Golani I. Controlling the false discovery rate in behavior genetics research. Behav. Brain Res. 2001;125(1-2):279–284. doi: 10.1016/s0166-4328(01)00297-2. [DOI] [PubMed] [Google Scholar]
  6. Bilmez Y., Ozturk S. Dynamic changes of histone methylation in male germ cells during spermatogenesis. FS Rev. 2023;4(3):187–205. [Google Scholar]
  7. Capuzzi E., Bartoli F., Crocamo C., Clerici M., Carrà G. Acute variations of cytokine levels after antipsychotic treatment in drug-naïve subjects with a first-episode psychosis: a meta-analysis. Neurosci. Biobehav. Rev. 2017;77:122–128. doi: 10.1016/j.neubiorev.2017.03.003. [DOI] [PubMed] [Google Scholar]
  8. Chen S., Tan Y., Tian L. Immunophenotypes in psychosis: is it a premature inflamm-aging disorder? Mol. Psychiatr. 2024;29(9):2834–2848. doi: 10.1038/s41380-024-02539-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Chen S.D., Sun X.Y., Niu W., Kong L.M., He M.J., Fan H.M.…Lu J. A preliminary analysis of microRNA-21 expression alteration after antipsychotic treatment in patients with schizophrenia. Psychiatry Res. 2016;244:324–332. doi: 10.1016/j.psychres.2016.04.087. [DOI] [PubMed] [Google Scholar]
  10. Cheng J.J., Ho H., Chang C.J., Lan S.Y., Hwu H.G. Positive and Negative Syndrome scale (PANSS): establishment and reliability study of a Mandarin Chinese language version. Chin. Psychiatry. 1996;10:251–258. [Google Scholar]
  11. Crespo-Facorro B., de la Foz V.O., Ayesa-Arriola R., Pérez-Iglesias R., Mata I., Suarez-Pinilla P.…Vázquez-Barquero J.L. Prediction of acute clinical response following a first episode of non affective psychosis: results of a cohort of 375 patients from the Spanish PAFIP study. Prog. Neuropsychopharmacol. Biol. Psychiatry. 2013;44:162–167. doi: 10.1016/j.pnpbp.2013.02.009. [DOI] [PubMed] [Google Scholar]
  12. de Bartolomeis A., Barone A., Vellucci L., Mazza B., Austin M.C., Iasevoli F.…Ciccarelli M. Linking inflammation, aberrant glutamate-dopamine interaction, and post-synaptic changes: translational relevance for schizophrenia and antipsychotic treatment: a systematic review. Mol. Neurobiol. 2022;59(10):6460–6501. doi: 10.1007/s12035-022-02976-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Demjaha A., Lappin J.M., Stahl D., Patel M.X., MacCabe J.H., Howes O.D.…Murray R.M. Antipsychotic treatment resistance in first-episode psychosis: prevalence, subtypes and predictors. Psychol. Med. 2017;47(11):1981–1989. doi: 10.1017/S0033291717000435. [DOI] [PubMed] [Google Scholar]
  14. Faul F., Erdfelder E., Buchner A., Lang A.-G. Statistical power analyses using G∗ power 3.1: tests for correlation and regression analyses. Behav. Res. Methods. 2009;41 doi: 10.3758/BRM.41.4.1149. [DOI] [PubMed] [Google Scholar]
  15. Gardiner E., Beveridge N.J., Wu J.Q., Carr V., Scott R.J., Tooney P.A.…Cairns M.J. Imprinted DLK1-DIO3 region of 14q32 defines a schizophrenia-associated miRNA signature in peripheral blood mononuclear cells. Mol. Psychiatr. 2012;17(8):827–840. doi: 10.1038/mp.2011.78. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Gillespie A.L., Samanaite R., Mill J., Egerton A., MacCabe J.H. Is treatment-resistant schizophrenia categorically distinct from treatment-responsive schizophrenia? A systematic review. BMC Psychiatry. 2017;17(1):12. doi: 10.1186/s12888-016-1177-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Golabi M., Fathi F., Samadi M., Hesamian M.S., Eskandari N. Identification of potential biomarkers in the peripheral blood mononuclear cells of relapsing-remitting multiple sclerosis patients. Inflammation. 2022;45(4):1815–1828. doi: 10.1007/s10753-022-01662-9. [DOI] [PubMed] [Google Scholar]
  18. Goossens J., Morrens M., Coppens V. The potential use of peripheral blood mononuclear cells as biomarkers for treatment response and outcome prediction in psychiatry: a systematic review. Mol. Diagn. Ther. 2021;25(3):283–299. doi: 10.1007/s40291-021-00516-8. [DOI] [PubMed] [Google Scholar]
  19. He K., Guo C., Guo M., Tong S., Zhang Q., Sun H.…Shi Y. Identification of serum microRNAs as diagnostic biomarkers for schizophrenia. Hereditas. 2019;156:23. doi: 10.1186/s41065-019-0099-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Huang X., Bao C., Lv Q., Zhao J., Hu G., Wu H.…Yi Z. MicroRNA-195 predicts olanzapine response in drug-free patients with schizophrenia: a prospective cohort study. J. Psychopharmacol. 2021;35(1):23–30. doi: 10.1177/0269881120959617. [DOI] [PubMed] [Google Scholar]
  21. Iasevoli F., Razzino E., Altavilla B., Avagliano C., Barone A., Ciccarelli M.…de Bartolomeis A. Relationships between early age at onset of psychotic symptoms and treatment resistant schizophrenia. Early Interv Psychiatry. 2022;16(4):352–362. doi: 10.1111/eip.13174. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Inloes J.M., Hsu K.L., Dix M.M., Viader A., Masuda K., Takei T.…Cravatt B.F. The hereditary spastic paraplegia-related enzyme DDHD2 is a principal brain triglyceride lipase. Proc. Natl. Acad. Sci. U. S. A. 2014;111(41):14924–14929. doi: 10.1073/pnas.1413706111. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Joshi M., Jeoung N.H., Popov K.M., Harris R.A. Identification of a novel PP2C-type mitochondrial phosphatase. Biochem. Biophys. Res. Commun. 2007;356(1):38–44. doi: 10.1016/j.bbrc.2007.02.108. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Kay S.R., Fiszbein A., Opler L.A. The positive and negative syndrome scale (PANSS) for schizophrenia. Schizophr. Bull. 1987;13(2):261–276. doi: 10.1093/schbul/13.2.261. [DOI] [PubMed] [Google Scholar]
  25. Keepers G.A., Fochtmann L.J., Anzia J.M., Benjamin S., Lyness J.M., Mojtabai R.…Hong S.H. The American Psychiatric association practice guideline for the treatment of patients with schizophrenia. Am. J. Psychiatr. 2020;177(9):868–872. doi: 10.1176/appi.ajp.2020.177901. [DOI] [PubMed] [Google Scholar]
  26. Kim A.H., Reimers M., Maher B., Williamson V., McMichael O., McClay J.L.…Vladimirov V.I. MicroRNA expression profiling in the prefrontal cortex of individuals affected with schizophrenia and bipolar disorders. Schizophr. Res. 2010;124:183–191. doi: 10.1016/j.schres.2010.07.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Kim D., Paggi J.M., Park C., Bennett C., Salzberg S.L. Graph-based genome alignment and genotyping with HISAT2 and HISAT-genotype. Nat. Biotechnol. 2019;37(8):907–915. doi: 10.1038/s41587-019-0201-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Kim S.S.-Y., Sze L., Lam K.-P. The stress granule protein G3BP1 binds viral dsRNA and RIG-I to enhance interferon-β response. J. Biol. Chem. 2019;294(16):6430–6438. doi: 10.1074/jbc.RA118.005868. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Lai C.-Y., Lee S.-Y., Scarr E., Yu Y.-H., Lin Y.-T., Liu C.-M.…Chen W.J. Aberrant expression of microRNAs as biomarker for schizophrenia: from acute state to partial remission, and from peripheral blood to cortical tissue. Transl. Psychiatry. 2016;6 doi: 10.1038/tp.2015.213. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Lai C.-Y., Yu S.-L., Hsieh M.H., Chen C.-H., Chen H.-Y., Wen C.-C.…Chen W.J. MicroRNA expression aberration as potential peripheral blood biomarkers for schizophrenia. PLoS One. 2011;6(6) doi: 10.1371/journal.pone.0021635. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Leucht S., Samara M., Heres S., Davis J.M. Dose equivalents for antipsychotic drugs: the DDD method. Schizophr. Bull. 2016;42(Suppl. 1):S90–S94. doi: 10.1093/schbul/sbv167. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Li S., Lu G., Wang D., He J.L., Zuo L., Wang H.…Deng Q.W. MicroRNA-4443 regulates monocyte activation by targeting tumor necrosis factor receptor associated factor 4 in stroke-induced immunosuppression. Eur. J. Neurol. 2020;27(8):1625–1637. doi: 10.1111/ene.14282. [DOI] [PubMed] [Google Scholar]
  33. Lieberman J.A., First M.B. Psychotic disorders. N. Engl. J. Med. 2018;379(3):270–280. doi: 10.1056/NEJMra1801490. [DOI] [PubMed] [Google Scholar]
  34. Lin D., Edwards A.S., Fawcett J.P., Mbamalu G., Scott J.D., Pawson T. A mammalian PAR-3-PAR-6 complex implicated in Cdc 42/Rac1 and aPKC signalling and cell polarity. Nat. Cell Biol. 2000;2(8):540–547. doi: 10.1038/35019582. [DOI] [PubMed] [Google Scholar]
  35. Liu S., Yuan Y.B., Guan L.L., Wei H., Cheng Z., Han X.…Yu X. MiRNA-365 and miRNA-520c-3p respond to risperidone treatment in first-episode schizophrenia after a 1 year remission. Chin Med. J. (Engl) 2013;126(14):2676–2680. [PubMed] [Google Scholar]
  36. Liu Y., Chang X., Hahn C.G., Gur R.E., Sleiman P.A.M., Hakonarson H. Non-coding RNA dysregulation in the amygdala region of schizophrenia patients contributes to the pathogenesis of the disease. Transl. Psychiatry. 2018;8(1):44. doi: 10.1038/s41398-017-0030-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Love M.I., Huber W., Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15(12):550. doi: 10.1186/s13059-014-0550-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. McGeary S.E., Lin K.S., Shi C.Y., Pham T.M., Bisaria N., Kelley G.M.…Bartel D.P. The biochemical basis of microRNA targeting efficacy. Science. 2019;366(6472) doi: 10.1126/science.aav1741. eaav1741. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. National Collaborating Centre for Mental Health . Psychosis and Schizophrenia in Adults: Treatment and Management: Updated Edition 2014. National Institute for Health and Care Excellence; (UK): 2014. National Clinical Guideline Number 178. London. [PubMed] [Google Scholar]
  40. O'Brien J., Hayder H., Zayed Y., Peng C. Overview of microRNA biogenesis, mechanisms of actions, and circulation. Front. Endocrinol. 2018;9:402. doi: 10.3389/fendo.2018.00402. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Oyarzabal A., Martínez-Pardo M., Merinero B., Navarrete R., Desviat L.R., Ugarte M.…Rodríguez-Pombo P. A novel regulatory defect in the branched-chain α-keto acid dehydrogenase complex due to a mutation in the PPM1K gene causes a mild variant phenotype of maple syrup urine disease. Hum. Mutat. 2013;34(2):355–362. doi: 10.1002/humu.22242. [DOI] [PubMed] [Google Scholar]
  42. Pardo-de-Santayana G., Juncal-Ruiz M., Vázquez-Bourgon J., Riesco-Dávila L., Ortiz-Garcia de la Foz V., Pelayo-Terán J.M.…Crespo-Facorro B. Active psychosis and pro-inflammatory cytokines in first-episode of psychosis. J. Psychiatr. Res. 2021;134:150–157. doi: 10.1016/j.jpsychires.2020.12.060. [DOI] [PubMed] [Google Scholar]
  43. Patil K.S., Basak I., Dalen I., Hoedt E., Lange J., Lunde K.A.…Møller S.G. Combinatory microRNA serum signatures as classifiers of Parkinson's disease. Park. Relat. Disord. 2019;64:202–210. doi: 10.1016/j.parkreldis.2019.04.010. [DOI] [PubMed] [Google Scholar]
  44. Perala J., Suvisaari J., Saarni S.I., Kuoppasalmi K., Isometsa E., Pirkola S.…Lonnqvist J. Lifetime prevalence of psychotic and bipolar I disorders in a general population. Arch. Gen. Psychiatry. 2007;64(1):19–28. doi: 10.1001/archpsyc.64.1.19. [DOI] [PubMed] [Google Scholar]
  45. Perkins D.O., Jeffries C.D., Jarskog L.F., Thomson J.M., Woods K., Newman M.A.…Hammond S.M. microRNA expression in the prefrontal cortex of individuals with schizophrenia and schizoaffective disorder. Genome Biol. 2007;8(2):R27. doi: 10.1186/gb-2007-8-2-r27. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. R Core Team . Austria R Foundation for Statistical Computing; Vienna: 2020. R: a Language and Environment for Statistical Computing.https://www.R-project.org/ URL: [Google Scholar]
  47. Reineke L.C., Lloyd R.E. The stress granule protein G3BP1 recruits protein kinase R to promote multiple innate immune antiviral responses. J. Virol. 2015;89(5):2575–2589. doi: 10.1128/JVI.02791-14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Shi S.H., Jan L.Y., Jan Y.N. Hippocampal neuronal polarity specified by spatially localized mPar3/mPar6 and PI 3-kinase activity. Cell. 2003;112(1):63–75. doi: 10.1016/s0092-8674(02)01249-7. [DOI] [PubMed] [Google Scholar]
  49. Smigielski L., Jagannath V., Rössler W., Walitza S., Grünblatt E. Epigenetic mechanisms in schizophrenia and other psychotic disorders: a systematic review of empirical human findings. Mol. Psychiatr. 2020;25(8):1718–1748. doi: 10.1038/s41380-019-0601-3. [DOI] [PubMed] [Google Scholar]
  50. Song H.T., Sun X.Y., Zhang L., Zhao L., Guo Z.M., Fan H.M.…Lu J. A preliminary analysis of association between the down-regulation of microRNA-181b expression and symptomatology improvement in schizophrenia patients before and after antipsychotic treatment. J. Psychiatr. Res. 2014;54:134–140. doi: 10.1016/j.jpsychires.2014.03.008. [DOI] [PubMed] [Google Scholar]
  51. Song X., Liu Y., Pu J., Gui S., Zhong X., Chen X.…Xie P. Transcriptomics analysis reveals shared pathways in peripheral blood mononuclear cells and brain tissues of patients with schizophrenia. Front. Psychiatr. 2021;12 doi: 10.3389/fpsyt.2021.716722. 2021. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Sun X.Y., Zhang J., Niu W., Guo W., Song H.T., Li H.Y.…Zhang Q.L. A preliminary analysis of microRNA as potential clinical biomarker for schizophrenia. Am. J. Med. Genet. B Neuropsychiatr. Genet. 2015;168b(3):170–178. doi: 10.1002/ajmg.b.32292. [DOI] [PubMed] [Google Scholar]
  53. The GTEx Consortium The GTEx Consortium atlas of genetic regulatory effects across human tissues. Science. 2020;369(6509):1318–1330. doi: 10.1126/science.aaz1776. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Tolosa E., Botta-Orfila T., Morató X., Calatayud C., Ferrer-Lorente R., Martí M.J.…Fernández-Santiago R. MicroRNA alterations in iPSC-derived dopaminergic neurons from Parkinson disease patients. Neurobiol. Aging. 2018;69:283–291. doi: 10.1016/j.neurobiolaging.2018.05.032. [DOI] [PubMed] [Google Scholar]
  55. Tourrière H., Chebli K., Zekri L., Courselaud B., Blanchard J.M., Bertrand E.…Tazi J. The RasGAP-associated endoribonuclease G3BP assembles stress granules. J. Cell Biol. 2003;160(6):823–831. doi: 10.1083/jcb.200212128. [DOI] [PMC free article] [PubMed] [Google Scholar] [Retracted]
  56. Troudet R., Ali W.B.H., Bacq-Daian D., Rossum I.W.v., Boland-Auge A., Battail C.…the O.s. g. Gene expression and response prediction to amisulpride in the OPTiMiSE first episode psychoses. Neuropsychopharmacology. 2020;45(10):1637–1644. doi: 10.1038/s41386-020-0703-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Trubetskoy V., Pardiñas A.F., Qi T., Panagiotaropoulou G., Awasthi S., Bigdeli T.B.…O'Donovan M.C. Mapping genomic loci implicates genes and synaptic biology in schizophrenia. Nature. 2022;604(7906):502–508. doi: 10.1038/s41586-022-04434-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Tugendreich S., Tomkiel J., Earnshaw W., Hieter P. CDC27Hs colocalizes with CDC16Hs to the centrosome and mitotic spindle and is essential for the metaphase to anaphase transition. Cell. 1995;81(2):261–268. doi: 10.1016/0092-8674(95)90336-4. [DOI] [PubMed] [Google Scholar]
  59. Wang Y., Iwamori T., Kaneko T., Iida H., Iwamori N. Comparative distributions of RSBN1 and methylated histone H4 Lysine 20 in the mouse spermatogenesis. PLoS One. 2021;16(6) doi: 10.1371/journal.pone.0253897. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Wei H., Yuan Y., Liu S., Wang C., Yang F., Lu Z.…Xu Q. Detection of circulating miRNA levels in schizophrenia. Am. J. Psychiatr. 2015;172(11):1141–1147. doi: 10.1176/appi.ajp.2015.14030273. [DOI] [PubMed] [Google Scholar]
  61. Yang W., Ru Y., Ren J., Bai J., Wei J., Fu S.…Zheng H. G3BP1 inhibits RNA virus replication by positively regulating RIG-I-mediated cellular antiviral response. Cell Death Dis. 2019;10(12):946. doi: 10.1038/s41419-019-2178-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Yu S.-C., Hwang T.J., Liu C.-M., Chan H.-Y., Kuo C.-J., Yang T.-T.…Chen W.J. Patients with first-episode psychosis in northern Taiwan: neurocognitive performance and niacin response profile in comparison with schizophrenia patients of different familial loadings and relationship with clinical features. BMC Psychiatry. 2024;24(1):155. doi: 10.1186/s12888-024-05598-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Zhan L., Kerr J.R., Lafuente M.J., Maclean A., Chibalina M.V., Liu B.…Nasir J. Altered expression and coregulation of dopamine signalling genes in schizophrenia and bipolar disorder. Neuropathol. Appl. Neurobiol. 2011;37(2):206–219. doi: 10.1111/j.1365-2990.2010.01128.x. [DOI] [PubMed] [Google Scholar]
  64. Zhang C., Li X., Zhao L., Guo W., Deng W., Wang Q.…Li T. Brain transcriptome-wide association study implicates novel risk genes underlying schizophrenia risk. Psychol. Med. 2023:1–11. doi: 10.1017/S0033291723000417. [DOI] [PubMed] [Google Scholar]
  65. Zhang L., Lizano P., Xu Y., Rubin L.H., Lee A.M., Lencer R.…Bishop J.R. Peripheral inflammation is associated with impairments of inhibitory behavioral control and visual sensorimotor function in psychotic disorders. Schizophr. Res. 2023;255:69–78. doi: 10.1016/j.schres.2023.03.030. [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Zhang L., Wei X. The roles of Par3, Par6, and aPKC polarity proteins in normal neurodevelopment and in neurodegenerative and neuropsychiatric disorders. J. Neurosci. 2022;42(24):4774–4793. doi: 10.1523/JNEUROSCI.0059-22.2022. [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Zhang Y., Geng R., Liu M., Deng S., Ding J., Zhong H.…Tu Q. Shared peripheral blood biomarkers for Alzheimer's disease, major depressive disorder, and type 2 diabetes and cognitive risk factor analysis. Heliyon. 2023;9(3) doi: 10.1016/j.heliyon.2023.e14653. [DOI] [PMC free article] [PubMed] [Google Scholar]
  68. Zhou T., Pan J., Xu K., Yan C., Yuan J., Song H.…Han Y. Single-cell transcriptomics in MI identify Slc25a4 as a new modulator of mitochondrial malfunction and apoptosis-associated cardiomyocyte subcluster. Sci. Rep. 2024;14(1):9274. doi: 10.1038/s41598-024-59975-8. [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

Multimedia component 1
mmc1.docx (2.1MB, docx)

Data Availability Statement

Due to ethical and consent restrictions, raw sequencing reads and individual-level clinical data cannot be publicly shared. Processed and de-identified miRNA and mRNA expression matrices and summary-level results that support the findings of this study are available from the corresponding author upon reasonable request and subject to institutional approval. All analysis code and scripts used for data processing and statistical analyses are also available upon request to facilitate reproducibility.

Articles from Brain, Behavior, & Immunity - Health are provided here courtesy of Elsevier

RESOURCES