HERV-W ENV Induces Innate Immune Activation and Neuronal Apoptosis via linc01930/cGAS Axis in Recent-Onset Schizophrenia

Abstract

Schizophrenia is a severe neuropsychiatric disorder affecting about 1% of individuals worldwide. Increased innate immune activation and neuronal apoptosis are common findings in schizophrenia. Interferon beta (IFN-β), an essential cytokine in promoting and regulating innate immune responses, causes neuronal apoptosis in vitro. However, the precise pathogenesis of schizophrenia is unknown. Recent studies indicate that a domesticated endogenous retroviral envelope glycoprotein of the W family (HERV-W ENV, also called ERVWE1 or syncytin 1), derived from the endogenous retrovirus group W member 1 (ERVWE1) locus on chromosome 7q21.2, has a high level in schizophrenia. Here, we found an increased serum IFN-β level in schizophrenia and showed a positive correlation with HERV-W ENV. In addition, serum long intergenic non-protein coding RNA 1930 (linc01930), decreased in schizophrenia, was negatively correlated with HERV-W ENV and IFN-β. In vitro experiments showed that linc01930, mainly in the nucleus and with noncoding functions, was repressed by HERV-W ENV through promoter activity suppression. Further studies indicated that HERV-W ENV increased IFN-β expression and neuronal apoptosis by restraining the expression of linc01930. Furthermore, HERV-W ENV enhanced cyclic GMP-AMP synthase (cGAS) and stimulator of interferon genes protein (STING) expression and interferon regulatory factor 3 (IRF3) phosphorylation in neuronal cells. Notably, cGAS interacted with HERV-W ENV and triggered IFN-β expression and neuronal apoptosis caused by HERV-W ENV. Moreover, Linc01930 participated in the increased neuronal apoptosis and expression level of cGAS and IFN-β induced by HERV-W ENV. To summarize, our results suggested that linc01930 and IFN-β might be novel potential blood-based biomarkers in schizophrenia. The totality of these results also showed that HERV-W ENV facilitated antiviral innate immune response, resulting in neuronal apoptosis through the linc01930/cGAS/STING pathway in schizophrenia. Due to its monoclonal antibody GNbAC1 application in clinical trials, we considered HERV-W ENV might be a reliable therapeutic choice for schizophrenia.

Keywords: HERV-W ENV; IFN-β; apoptosis; cGAS; linc01930; schizophrenia.

Figure 1

Dysregulation of HERV-W ENV, linc01930, IFN-β and their correlation analyses in schizophrenia patients. (a) Linc01930 RNA levels in schizophrenia patients compared with healthy controls by qRT-PCR (F, DFn, Dfd, 6.832, 25, 20). (b) IFN-β concentration in schizophrenia patients compared with healthy controls by ELISA (F, DFn, Dfd, 2.789, 20, 25). (c) HERV-W ENV mRNA levels in schizophrenia patients compared with healthy controls by qRT-PCR (F, DFn, Dfd, 54.93, 20, 25). (d) Correlation between HERV-W ENV and linc01930 RNA levels in schizophrenia patients and healthy controls, where Y was the RNA expression for linc01930 and X was HERV-W ENV mRNA value for each sample (F, DFn, Dfd, 0.2180, 1.000, 45.00). (e) Correlation between HERV-W ENV mRNA levels and IFN-β protein levels in patients with schizophrenia patients and healthy controls, where Y was the protein expression for IFN-β and X was HERV-W ENV mRNA value for each sample (F, DFn, Dfd, 26.86, 1.000, 45.00). (f) Correlation between linc01930 RNA levels and IFN-β protein levels in patients with schizophrenia patients and healthy controls, where Y was the protein expression for IFN-β and X was linc01930 RNA value for each sample (F, DFn, Dfd, 0.4263, 1.000, 45.00 ). Subfigures (a–c) were analyzed with Student’s t-test. Subfigures (d–f) were analyzed with Spearman’s rank correlation analysis. * p < 0.05; ** p < 0.01.

Figure 1

Dysregulation of HERV-W ENV, linc01930, IFN-β and their correlation analyses in schizophrenia patients. (a) Linc01930 RNA levels in schizophrenia patients compared with healthy controls by qRT-PCR (F, DFn, Dfd, 6.832, 25, 20). (b) IFN-β concentration in schizophrenia patients compared with healthy controls by ELISA (F, DFn, Dfd, 2.789, 20, 25). (c) HERV-W ENV mRNA levels in schizophrenia patients compared with healthy controls by qRT-PCR (F, DFn, Dfd, 54.93, 20, 25). (d) Correlation between HERV-W ENV and linc01930 RNA levels in schizophrenia patients and healthy controls, where Y was the RNA expression for linc01930 and X was HERV-W ENV mRNA value for each sample (F, DFn, Dfd, 0.2180, 1.000, 45.00). (e) Correlation between HERV-W ENV mRNA levels and IFN-β protein levels in patients with schizophrenia patients and healthy controls, where Y was the protein expression for IFN-β and X was HERV-W ENV mRNA value for each sample (F, DFn, Dfd, 26.86, 1.000, 45.00). (f) Correlation between linc01930 RNA levels and IFN-β protein levels in patients with schizophrenia patients and healthy controls, where Y was the protein expression for IFN-β and X was linc01930 RNA value for each sample (F, DFn, Dfd, 0.4263, 1.000, 45.00 ). Subfigures (a–c) were analyzed with Student’s t-test. Subfigures (d–f) were analyzed with Spearman’s rank correlation analysis. * p < 0.05; ** p < 0.01.

Figure 2

HERV-W ENV stimulated antiviral innate immune responses and mediated neuronal apoptosis (a,b) Respectively represent IFN-β mRNA levels in HERV-W ENV (0.9 μg) transfected SH-SY5Y cell (F, DFn, Dfd, 4.019, 3, 3) and rat primary neurons (F, DFn, Dfd, 5.896, 2, 2) detected by qRT-PCR. (c,d) Respectively represent IFN-β protein expression in HERV-W ENV (0.9 μg) transfected SH-SY5Y cell (F, DFn, Dfd, 3.494, 2, 2) and rat primary neurons (F, DFn, Dfd, 9.123, 2, 2) by western blotting (48 h after transfection). (e) Luciferase assays of pGL3- IFN-β promoter (0.2 μg) co-transfected with pCMV-HERV-W ENV (0.4 μg) in SH-SY5Y cell (F, DFn, Dfd, 3.494, 2, 2). (f) Cell proliferation of SH-SY5Y cell transfected with pCMV-HERV-W ENV (0.9 μg) and control vector by Cell counting kit 8 (CCK8) assays (F, DFn, Dfd, 5.892, 3, 3). (g) Flow cytometry analyses of HERV-W ENV (0.9 μg) on cell apoptosis in SH-SY5Y cells (F, DFn, Dfd, 1.671, 4, 4). Statistical analysis was performed by one-way analysis of variance (ANOVA). * p < 0.05; ** p < 0.01.

Figure 2

HERV-W ENV stimulated antiviral innate immune responses and mediated neuronal apoptosis (a,b) Respectively represent IFN-β mRNA levels in HERV-W ENV (0.9 μg) transfected SH-SY5Y cell (F, DFn, Dfd, 4.019, 3, 3) and rat primary neurons (F, DFn, Dfd, 5.896, 2, 2) detected by qRT-PCR. (c,d) Respectively represent IFN-β protein expression in HERV-W ENV (0.9 μg) transfected SH-SY5Y cell (F, DFn, Dfd, 3.494, 2, 2) and rat primary neurons (F, DFn, Dfd, 9.123, 2, 2) by western blotting (48 h after transfection). (e) Luciferase assays of pGL3- IFN-β promoter (0.2 μg) co-transfected with pCMV-HERV-W ENV (0.4 μg) in SH-SY5Y cell (F, DFn, Dfd, 3.494, 2, 2). (f) Cell proliferation of SH-SY5Y cell transfected with pCMV-HERV-W ENV (0.9 μg) and control vector by Cell counting kit 8 (CCK8) assays (F, DFn, Dfd, 5.892, 3, 3). (g) Flow cytometry analyses of HERV-W ENV (0.9 μg) on cell apoptosis in SH-SY5Y cells (F, DFn, Dfd, 1.671, 4, 4). Statistical analysis was performed by one-way analysis of variance (ANOVA). * p < 0.05; ** p < 0.01.

Figure 3

HERV-W ENV repressed linc01930 expression in neuronal cells. (a,b) Respectively represent linc01930 RNA levels in HERV-W ENV-transfected (0.6 μg) SH-SY5Y cell (F, DFn, Dfd, 10.25, 2, 2) and rat primary neurons (F, DFn, Dfd, 1.455, 2, 2) by qRT-PCR. (c) Luciferase assays of the pGL3-linc01930 promoter (0.2 μg) co-transfected with pCMV-HERV-W ENV plasmid (0.4 μg) in SH-SY5Y cell for 24 h (F, DFn, Dfd, 2.416, 2, 2). (d) Open Reading Frame (ORF) of linc01930 predicted by NCBI ORF Finder with three fragment (+215 to +304, +465 to +557, +502 to +723). (e) pEGFP-N3 (215-301, 465-554, 502-720, 1-720) plasmid (1.0 μg) was separately transfected in SH-SY5Y cell for 48 h and tested by western blotting. (f–h) Cellular distribution of linc01930 (F, DFn, Dfd, 2.703, 2, 2) was mainly located at the nucleus in the SH-SY5Y cell. Nuclear and cytoplasmic separation effects were quantified to RPS14 (F, DFn, Dfd, 1.316, 2, 2) in the cytoplasmic part and U6 (F, DFn, Dfd, 2.941, 2, 2) in the nuclear part. Statistical analysis was performed by one-way analysis of variance (ANOVA). * p < 0.05; ** p < 0.01.

Figure 4

Linc01930 diminished antiviral innate immune reaction and attenuated neural cell apoptosis. (a,b) Respectively represent IFN-β mRNA levels in linc01930 (0.6 μg)-transfected SH-SY5Y cell (F, DFn, Dfd, 3.647, 11, 11) and rat primary neurons (F, DFn, Dfd, 1.305, 2, 2) by qRT-PCR. (c,d) Respectively represent IFN-β protein expression in linc01930 (0.6 μg)-transfected SH-SY5Y cell (F, DFn, Dfd, 1.196, 4, 4) and rat primary neurons (F, DFn, Dfd, 1.054, 2, 2) by western blotting. (e) Luciferase assays of the pGL3-IFN-β promoter (0.2 μg) co-transfected with a pcDNA3.1-linc01930 plasmid (0.2 μg) or control vector in SH-SY5Y cell (F, DFn, Dfd, 11.37, 2, 2). (f) Cell proliferation of SH-SY5Y cell transfected with a pcDNA3.1-linc01930 plasmid (0.6 μg) or control vector by CCK8 assays (F, DFn, Dfd, 1.555, 4, 4). (g) Flow cytometry analyses of 0.6 μg linc01930 on cell apoptosis in SH-SY5Y cells (F, DFn, Dfd, 13.68, 3, 3). Statistical analysis was performed by one-way analysis of variance (ANOVA). * p < 0.05; ** p < 0.01.

Figure 4

Linc01930 diminished antiviral innate immune reaction and attenuated neural cell apoptosis. (a,b) Respectively represent IFN-β mRNA levels in linc01930 (0.6 μg)-transfected SH-SY5Y cell (F, DFn, Dfd, 3.647, 11, 11) and rat primary neurons (F, DFn, Dfd, 1.305, 2, 2) by qRT-PCR. (c,d) Respectively represent IFN-β protein expression in linc01930 (0.6 μg)-transfected SH-SY5Y cell (F, DFn, Dfd, 1.196, 4, 4) and rat primary neurons (F, DFn, Dfd, 1.054, 2, 2) by western blotting. (e) Luciferase assays of the pGL3-IFN-β promoter (0.2 μg) co-transfected with a pcDNA3.1-linc01930 plasmid (0.2 μg) or control vector in SH-SY5Y cell (F, DFn, Dfd, 11.37, 2, 2). (f) Cell proliferation of SH-SY5Y cell transfected with a pcDNA3.1-linc01930 plasmid (0.6 μg) or control vector by CCK8 assays (F, DFn, Dfd, 1.555, 4, 4). (g) Flow cytometry analyses of 0.6 μg linc01930 on cell apoptosis in SH-SY5Y cells (F, DFn, Dfd, 13.68, 3, 3). Statistical analysis was performed by one-way analysis of variance (ANOVA). * p < 0.05; ** p < 0.01.

Figure 5

Linc01930 reversed antiviral innate immune dysfunction and neural apoptosis mediated by HERV-W ENV. (a,b) Respectively represent IFN-β protein expression after co-transfection with HERV-W ENV (0.8 μg) and linc01930 (0.4 μg) in SH-SY5Y cell (F, DFn, Dfd, 4.515, 2, 2) and rat primary neurons (F, DFn, Dfd, 1.538, 2, 2) with western blotting. (c,d) Respectively represent IFN-β expression levels after co-transfection with HERV-W ENV (0.8 μg) and linc01930 (0.4 μg) in SH-SY5Y cell (F, DFn, Dfd, 1.324, 2, 2) and rat primary neurons (F, DFn, Dfd, 4.356, 2, 2) with ELISA. (e) Cell proliferation was examined in SH-SY5Y cells with co-transfection of HERV-W ENV (0.8 μg) and linc01930 (0.4 μg) using the CCK8 assays (F, DFn, Dfd, 3.190, 2, 2). (f) The effect of co-transfection of HERV-W ENV and linc01930 on cell apoptosis ratios in SH-SY5Y cell were analyzed (F, DFn, Dfd, 3.641, 2, 2). (g) Cell apoptosis was detected in SH-SY5Y cell with co-transfection of HERV-W ENV (0.8 μg) and linc01930 (0.4 μg) using the flow cytometry. Statistical analysis was performed by one-way analysis of variance (ANOVA). * p < 0.05; ** p < 0.01.

Figure 5

Linc01930 reversed antiviral innate immune dysfunction and neural apoptosis mediated by HERV-W ENV. (a,b) Respectively represent IFN-β protein expression after co-transfection with HERV-W ENV (0.8 μg) and linc01930 (0.4 μg) in SH-SY5Y cell (F, DFn, Dfd, 4.515, 2, 2) and rat primary neurons (F, DFn, Dfd, 1.538, 2, 2) with western blotting. (c,d) Respectively represent IFN-β expression levels after co-transfection with HERV-W ENV (0.8 μg) and linc01930 (0.4 μg) in SH-SY5Y cell (F, DFn, Dfd, 1.324, 2, 2) and rat primary neurons (F, DFn, Dfd, 4.356, 2, 2) with ELISA. (e) Cell proliferation was examined in SH-SY5Y cells with co-transfection of HERV-W ENV (0.8 μg) and linc01930 (0.4 μg) using the CCK8 assays (F, DFn, Dfd, 3.190, 2, 2). (f) The effect of co-transfection of HERV-W ENV and linc01930 on cell apoptosis ratios in SH-SY5Y cell were analyzed (F, DFn, Dfd, 3.641, 2, 2). (g) Cell apoptosis was detected in SH-SY5Y cell with co-transfection of HERV-W ENV (0.8 μg) and linc01930 (0.4 μg) using the flow cytometry. Statistical analysis was performed by one-way analysis of variance (ANOVA). * p < 0.05; ** p < 0.01.

Figure 6

HERV-W ENV activated cGAS-mediated antiviral signaling pathway. (a,b) Respectively represent cGAS mRNA levels in pCMV-HERV-W ENV (0.9 μg) transfected SH-SY5Y cell (F, DFn, Dfd, 4.199, 2, 2) and primary neurons (F, DFn, Dfd, 1.502, 2, 2) using qRT-PCR. (c,d) Respectively represent cGAS protein expression in pCMV-HERV-W ENV (0.9 μg) transfected SH-SY5Y cell (F, DFn, Dfd, 2.168, 2, 2) and primary neurons (F, DFn, Dfd, 6.156, 2, 2) using western blotting. (e) HERV-W ENV (0.9 μg) effect on IRF3 phosphorylation at Ser 386 site in the SH-SY5Y cell with western blotting (F, DFn, Dfd, 5.029, 4, 4). (f) Co-immunoprecipitation assays (Co-IP) were performed between pXJ40-HA-HERV-W ENV (5.0 μg) and pENTER-N-FLAG-cGAS (5.0 μg) with anti-Flag and anti-HA magnetic beads by western blotting in HEK-293T cell. Statistical analysis was performed by one-way analysis of variance (ANOVA). * p < 0.05; ** p < 0.01.

Figure 7

Linc01930 impaired the cGAS-mediated antiviral signaling pathway mediated by HERV-W ENV in neuronal cells. (a,b) Respectively represent cGAS mRNA levels in pcDNA3.1-linc01930 (0.6 μg) transfected SH-SY5Y cell (F, DFn, Dfd, 4.174, 8, 8) and primary neurons (F, DFn, Dfd, 1.365, 2, 2) using qRT-PCR. (c,d) Respectively represent cGAS protein expression in pcDNA3.1-linc01930 (0.6 μg) transfected SH-SY5Y cell (F, DFn, Dfd, 5.492, 2, 2) and primary neurons (F, DFn, Dfd, 4.058, 3, 3) using western blotting. (e) cGAS protein levels after co-transfection of pCMV-HERV-W ENV (0.8 μg) and pcDNA3.1-linc01930 (0.4 μg) in SH-SY5Y cell with western blotting (F, DFn, Dfd, 7.053, 2, 2). (f) cGAS protein levels after co-transfection of pCMV-HERV-W ENV (0.8 μg) and pcDNA3.1-linc01930 (0.4 μg) in primary neurons using western blotting (F, DFn, Dfd, 2.227, 2, 2). Statistical analysis was performed by one-way analysis of variance (ANOVA). * p < 0.05; ** p < 0.01.

Figure 8

HERV-W ENV induced antiviral innate immune responses, and neural apoptosis depended on the cGAS-mediated antiviral signaling pathway. (a) IFN-β levels in a transfected sh-cGAS plasmid (0.9 μg) in SH-SY5Y cell using western blotting (F, DFn, Dfd, 16.70, 2, 2). (b) Cell proliferation of SH-SY5Y cell transfected with a sh-cGAS plasmid (0.9 μg) performed with CCK8 test (F, DFn, Dfd, 11.18, 4, 4). (c) Cell apoptosis of SH-SY5Y cell transfected with sh-cGAS plasmid performed with the flow cytometry (F, DFn, Dfd, 58.90, 3, 3). (d,e) Respectively represent IFN-β protein expression after co-transfection with HERV-W ENV (0.8 μg) and sh-cGAS (0.6 μg) in SH-SY5Y cell (F, DFn, Dfd, 7.053, 2, 2) and rat primary neurons (F, DFn, Dfd, 7.053, 2, 2) with western blotting. (f,g) Respectively represent IFN-β expression levels after co-transfection with HERV-W ENV(0.8 μg) and sh-cGAS (0.6 μg) in SH-SY5Y cell (F, DFn, Dfd, 13.28, 2, 2) and rat primary neurons (F, DFn, Dfd, 2.392, 2, 2) with ELISA. (h) Cell apoptosis was detected in SH-SY5Y cell with co-transfection of HERV-W ENV (0.8 μg) and sh-cGAS (0.6 μg) using the flow cytometry. (i) The effect of co-transfection of HERV-W ENV and sh-cGAS on cell apoptosis ratios in SH-SY5Y cell were analyzed (F, DFn, Dfd, 2.650, 2, 2). (j) Cell proliferation was examined in SH-SY5Y cells with co-transfection of HERV-W ENV (0.8 μg) and sh-cGAS (0.6 μg) using CCK8 (F, DFn, Dfd, 4.019, 2, 2). Statistical analysis was performed by one-way analysis of variance (ANOVA). * p < 0.05; ** p < 0.01.

Figure 8

HERV-W ENV induced antiviral innate immune responses, and neural apoptosis depended on the cGAS-mediated antiviral signaling pathway. (a) IFN-β levels in a transfected sh-cGAS plasmid (0.9 μg) in SH-SY5Y cell using western blotting (F, DFn, Dfd, 16.70, 2, 2). (b) Cell proliferation of SH-SY5Y cell transfected with a sh-cGAS plasmid (0.9 μg) performed with CCK8 test (F, DFn, Dfd, 11.18, 4, 4). (c) Cell apoptosis of SH-SY5Y cell transfected with sh-cGAS plasmid performed with the flow cytometry (F, DFn, Dfd, 58.90, 3, 3). (d,e) Respectively represent IFN-β protein expression after co-transfection with HERV-W ENV (0.8 μg) and sh-cGAS (0.6 μg) in SH-SY5Y cell (F, DFn, Dfd, 7.053, 2, 2) and rat primary neurons (F, DFn, Dfd, 7.053, 2, 2) with western blotting. (f,g) Respectively represent IFN-β expression levels after co-transfection with HERV-W ENV(0.8 μg) and sh-cGAS (0.6 μg) in SH-SY5Y cell (F, DFn, Dfd, 13.28, 2, 2) and rat primary neurons (F, DFn, Dfd, 2.392, 2, 2) with ELISA. (h) Cell apoptosis was detected in SH-SY5Y cell with co-transfection of HERV-W ENV (0.8 μg) and sh-cGAS (0.6 μg) using the flow cytometry. (i) The effect of co-transfection of HERV-W ENV and sh-cGAS on cell apoptosis ratios in SH-SY5Y cell were analyzed (F, DFn, Dfd, 2.650, 2, 2). (j) Cell proliferation was examined in SH-SY5Y cells with co-transfection of HERV-W ENV (0.8 μg) and sh-cGAS (0.6 μg) using CCK8 (F, DFn, Dfd, 4.019, 2, 2). Statistical analysis was performed by one-way analysis of variance (ANOVA). * p < 0.05; ** p < 0.01.

Figure 9

The potential role of HERV-W ENV to trigger neuronal apoptosis via innate immune activation in schizophrenia. The decreased linc01930 was negatively correlated with increased HERV-W ENV and IFN-β in schizophrenia. HERV-W ENV repressed linc01930 expression via its promoter activity. HERV-W ENV activated cGAS and STING expression and elevated IRF3 phosphorylation, while linc01930 functioned as a negative regulator to HERV-W ENV-induced cGAS and STING expression and IRF3 phosphorylation. In addition, linc01930 was involved in regulating the cGAS/STING signaling pathway induced by HERV-W ENV. Moreover, HERV-W ENV activated IFN-β expression via its promoter activity, while linc01930 inhibited linc01930 expression via its promoter activity. Furthermore, HERV-W ENV mediated the increased cGAS and IFN-β expression and neuronal apoptosis by regulating linc01930 expression. Thus, Innate immune activation might contribute to the etiology of schizophrenia.