Expression of NF-κB associated lncRNAs in schizophrenia

NF-κB signaling pathway has important roles in the regulation of growth and development of nervous system. This pathway has also been shown to participate in the pathogenesis of schizophrenia. Meanwhile, activity of NF-κB signaling pathway is regulated by several factors including non-coding RNAs (lncRNAs). In the current study, we evaluated expression of nine NF-κB-related lncRNAs namely DILC, ANRIL, PACER, CHAST, ADINR, DICER1-AS1, HNF1A-AS1, H19 and NKILA as well as two mRNA coding genes namely ATG5 and CEBPA in the peripheral blood of patients with schizophrenia compared with matched healthy subjects. Expressions of these genes were assessed by real time PCR technique. Expression of PACER was lower in patients with schizophrenia compared with controls (Posterior beta = − 0.684, P value = 0.049). On the other hand, expressions of CHAST, CEBPA, H19, HNF1A-AS1 and DICER1-AS1 were higher in patients compared with controls (Posterior beta = 0.39, P value = 0.005; Posterior beta = 0.844, P value < 0.0001; Posterior beta = 0.467, P value < 0.0001; Posterior beta = 1.107, P value = 0.005; Posterior beta = 0.176, P value = 0.044, respectively). We also appraised the diagnostic power of transcript quantities of CHAST, CEBPA, DICER1-AS1, H19 and HNF1A-AS1 in distinguishing between patients with schizophrenia and controls through depicting ROC curves. Based on the area under curve (AUC) values, CEBPA had the best diagnostic power (AUC = 0.948, P < 0.0001), followed by H19 (AUC = 0.815, P < 0.0001). Taken together, our study demonstrated dysregulation of NF-κB-related lncRNAs and genes in the peripheral blood of patients with schizophrenia and their potential as peripheral markers for this psychiatric condition.

Scientific Reports | (2020) 10:18105 | https://doi.org/10.1038/s41598-020-75333-w www.nature.com/scientificreports/ The basis for selection of these lncRNAs and mRNA coding genes was their functional relevance with NF-κB signaling. Expression of some of these genes are regulated by NF-κB. For instance, expression of NKILA has been shown to be activated by NF-κB. This lncRNA binds with NF-κB/IκB, and conceals phosphorylation sites of IκB, therefore inhibiting NF-κB activation 7 . Others indirectly regulate the interaction between NF-κB and other signaling pathways or modulate the role of NF-κB in cellular functions. The lncRNA DILC participates in the regulation of interaction between TNF-α/NF-κB signaling and IL-6/STAT3 axis 8 . ANRIL has been identified as an important element in the NF-κB pathway that regulates inflammatory responses 9 . PACER has been shown to induce COX-2 gene expression through blocking suppressive NF-κB complexes 10 . CHAST is an lncRNA which influences the activity of Wnt signaling 11 , a pathway that is functionally linked with NF-κB signaling 12 . ADINR is an lncRNA transcribed from a genomic region near of the CEBPA gene and its expression is correlated with the expression its nearby gene 13 . CEBPA participates in the induction of NF-κB target genes through replacement of histone deacetylases from NF-κB p50 homodimers 14 . DICER1-AS1 has been shown to regulate autophagy via modulation of miR-30b/ATG5 axis 15 . Notably, autophagy contributes in several cellular functions related with NF-κB signaling 16 . HNF1A-AS1 is an lncRNA which is transcribed from the opposite direction of the HNF1A gene 17 , a transcription factor that interacts with the NF-κB signaling 18 . Finally, several studies have reported the interaction between H19 and NF-κB signaling in diverse situations 19,20 . Based on the prominent roles of these genes in the regulation of NF-κB and involvement of this pathway in neurodevelopment and inflammation, we hypothesized that expression of these geens have been dysregulated in the peripheral blood of patients with schizophrenia as a reflection of their expression in the central tissues. Expression assay. Five milliliters of the peripheral blood were gathered in EDTA-containing falcon tubes.

Materials and methods
These specimens were subjected to RNA extraction by the Hybrid-R blood RNA extraction Kit (GeneAll, Seoul, Korea). All steps were accomplished according to the manufacturer's protocol. The yielded RNA was subsequently converted to cDNA by using the High-Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scientific, Gent, Belgium). Expressions of mentioned lncRNAs were assessed in all enrolled individuals using the RealQ Plus 2 × PCR Master Mix Green Without ROX PCR Master Mix (Ampliqon, Odense, Denmark). Cycling reactions were carried out in Step One Plus Real-Time PCR equipment (Applied Biosystems, Foster city, CA, USA). Primers used for expression assays and PCR products sizes were identical to our previous studies 24, 25 . Table S1 summarizes the primers features. We have chosen B2M gene as the reference gene based on our previous observations regrading constant expression of this gene in peripheral blood of patients with schizophrenia 26 .

Statistical methods. Relative transcripts levels of lncRNAs were measured in all samples the Ln
[Efficiency^ΔCT] method considering the transcript levels of B2M as normalizer. Hierarchical Bayesian regression model was used for comparison of these values between cases and controls. The impacts of independent variables were adjusted. The asymmetric Laplace prior distribution was supposed with 4000 iteration and 1000 warm-ups for parameterization of expression ratio of lncRNAs/genes. P values were calculated via median regression model. Correlations between expressions of genes/lncRNAs were valued by calculation of Spearman correlation coefficients. Data were analyzed using the R v.4 software and pROC, qreg, and Stan and loo packages. The diagnostic power of the transcript levels of lncRNAs/genes was measured through depicting receiver operating characteristic (ROC) curves.

Results
General data of patients with schizophrenia and controls. A total of 50 patients with schizophrenia and 50 healthy subjects were recruited for the current case-control study. Table 1 summarizes the demographic data of cases and controls.
Expression assays. Expression levels of mentioned lncRNAs and mRNA coding genes were compared between cases and controls. As demonstrated in Fig. 1, there were significant differences in the expression levels of PACER, CHAST, CEBPA, H19, HNF1A-AS1 and DICER1-AS1 between cases and controls. Expression of PACER was lower in patients with schizophrenia compared with controls (Posterior beta = − 0.684, P value = 0.049). Group-gender interaction was significant for this lncRNA (P value = 0.005). On the other hand, expressions of CHAST, CEBPA, H19, HNF1A-AS1 and DICER1-AS1 were higher in patients  Table 2 shows the results of Bayesian Regression model for comparison of expression of lncRNAs/ genes between cases and controls. Next, we appraised the diagnostic power of transcript quantities of CHAST, CEBPA, DICER1-AS1, H19 and HNF1A-AS1 in distinguishing between patients with schizophrenia and controls through depicting ROC curves (Fig. 2). Based on the area under curve (AUC) values, CEBPA had the best diagnostic power (AUC = 0.948, P < 0.0001), followed by H19 (AUC = 0.815, P < 0.0001).
Assessment of pairwise correlation between lncRNAs/genes revealed significant correlation between all pairs. The most robust correlations were observed between NKILA/ADINR and between NKILA/HNF1A-AS1 (Corerlation coefficients = 0.80 and 0.78, respectively). Figure 3 shows the correlation coefficients and P values.

Discussion
NF-κB pathway has crucial roles in the pathophysiology of schizophrenia. A recent study in schizophrenia patients has reported up-regulation of the majority of NF-κB family members, the entire NF-κB activation receptors, numerous kinases and IκBα in patients with schizophrenia 27 . This aberrant activity of NF-κB-associated factors has been suggested to be associated with higher levels of cortical immune activation in these patients 27 . Meanwhile, activity and expression of NF-κB-associated genes have been shown to be regulated by lncRNAs 28 . The interaction between lncRNAs and NF-κB-related genes have implications in the pathogenesis of human disorders 28 . In the current study, we evaluated expression of nine lncRNAs and two mRNAs in the peripheral blood of schizophrenic patients and healthy subjects. These genes were previously reported to be linked with NF-κB pathway. Figure 4 depicts a summary of identified interactions between these lncRNAs and NF-κB.
We reported significant down-regulation of PACER in patients with schizophrenia compared with healthy subjects. Expression of this lncRNA has been shown to be induced by CTCF 10 . In turn, PACER increases expression of COX-2 through blocking suppressive NF-κB complexes 10 . CTCF is possibly involved in the pathogenesis of schizophrenia, since a number of genomic variants within CTCF coding genes have been associated with schizophrenia 29 . Besides, assessment of data provided from genome-wide association studies has revealed functional links between schizophrenia risk variants and CTCF binding 30 . Moreover, CTCF has been shown to regulate a primary fate decision step in the construction of cortical interneurons, thus participating in the development of a wide range of neurological conditions including schizophrenia 31 . On the other hand, COX-2 has interactions with a number of neurotransmitters and participates in the modulation of immune responses www.nature.com/scientificreports/ in the central nervous (CNS) through regulation of production of prostaglandins 32 . Therefore, COX-2 inhibitors have been suggested as therapeutic modalities for some neuropsychiatric conditions such as schizophrenia 32 . The observed down-regulation of PACER in patients with schizophrenia in the current study might be due to a possible negative feedback of COX-2 or decreased levels of CFCT. Concomitant assessments of CFTC, PACER and COX-2 levels in the peripheral blood as well as central tissues of patients with schizophrenia and functional studies are needed to elaborate the molecular mechanism of involvement of PACER in the pathophysiology of schizophrenia. An important finding about this lncRNA was the observed group × gender interaction. While the overall effect suggests that schizophrenia cases have reduced expression of PACER relative to controls, this effect appears to be inverted in females. It is possible that the direction of the overall group effect was driven by the higher proportion of males in each group. Therefore, we recommend conduction of studies in larger cohorts of both male and female cases to appraise the group x gender interaction. On the other hand, expressions of CHAST, CEBPA, H19, HNF1A-AS1 and DICER1-AS1 were higher in patients compared with controls. Expression of CHAST has been shown to be activated by nuclear factor of activated T cells (NFAT) signaling 11 . The calcineurin-NFAT pathway has acknowledged roles in the normal function of CNS and in the pathobiology of neurological disorders 33 . Particularly, genomic variants within the genes coding for calcineurin isoenzymes have been linked with the susceptibility to schizophrenia 34 . CEBPA interacts with the promoter of leptin coding gene. Receptors for leptin have been detected in several areas of the brain including the hippocampus and cerebral cortex, and are involved in the development of brain and neuroendocrine functions 35 . The role of leptin in the pathobiology of schizophrenia is obscure. While serum levels of this factor are inversely correlated with the severity of positive symptoms in schizophrenia, no correlation has been detected between its concentrations and negative symptoms or cognition 36 . On the other hand, a Taken together, the CEBPA-induced alterations in leptin levels have potential roles in at least some aspects of pathobiology of schizophrenia. Alterations in CEBPA levels might also be associated with expression of H19.
A previous study has shown correlation between DNA methylation in a genomic region near to an important CTCF-binding site in the imprinting control region (ICR) upstream of H19 and cerebellum weight. Based on these results, authors have suggested an epigenetic mechanism for the development of schizophrenia 39 . H19 is possibly involved in the pathogenesis of psychiatric disorders through enhancement of neuron apoptosis. An animal study has shown that H19 increases hippocampal neuronal apoptosis through Wnt signaling 40 . On the other hand, down-regulation of H19 and the H19-originated miRNA miR-675 has been associated with overexpression of insulin-like growth factor receptor type 1 throughout the course of neural-like differentiation of stem cells 41 , indicating a possible role for this lncRNA in the neurodevelopment. NF-κB signaling has been shown to inhibit expression of HNF-1α 18 , the transcription factor that is locally related with this lncRNA. Yet, the functional correlation between this transcription factor and HNF1A-AS1 has not validated 17 . Interestingly, HNF1A-AS1 has been shown to activate expression of H19 17 . Therefore, a possible mechanism for participation of HNF1A-AS1 in the pathobiology of schizophrenia is its role in induction of H19 expression. Finally, DICER1-AS1 has been shown to regulate autophagy via modulation of miR-30b/ATG5 axis 15 . Autophagy has been shown to participate in the physiology of CNS through modulation of neuronal homeostasis. Failure in this process has been associated with the neurologic dysfunction, neurodegenerative disorders and schizophrenia 42 . On the other hand, antipsychotic drugs might ameliorate the observed downregulation of autophagy genes in some parts of brain areas in the schizophrenia patients 42,43 . Therefore, the observed up-regulation of DICER1-AS1 in patients with schizophrenia might be a compensatory mechanism for enhancement of autophagy in these patients. We also appraised the diagnostic power of transcript quantities of CHAST, CEBPA, DICER1-AS1, H19 and HNF1A-AS1 in distinguishing between patients with schizophrenia and controls through depicting ROC curves. Such aanlyses implied possible use of CEBPA and H19 for this purpose. However, as patients were under treatment with antipsychotic medication, we cannot definitly propose these transcripts as biomarkers. The findings of the present study could help guide future researches, and replication in a sample of drug naive individuals with first-episode psychosis. Such studies would propose possible candidates as biomarkers in schizophrenia.  www.nature.com/scientificreports/ Finally, assessment of pairwise correlation between lncRNAs/genes revealed significant correlation between all pairs which further supports their participation in a certain signaling pathway namely NF-κB signaling pathway. The most robust correlations were observed between NKILA/ ADINR and between NKILA/ HNF1A-AS1 which suggest the presence of important functional links between these genes that warrants additional functional analyses.
In brief, the current study demonstrated dysregulation of NF-κB-related lncRNAs and genes in the peripheral blood of patients with schizophrenia and their potential as peripheral markers for this psychiatric condition.