- Email: [email protected]
Hypomethylation of LINE-1 elements in schizophrenia and bipolar disorder
Journal of Psychiatric Research 107 (2018) 68–72
Contents lists available at ScienceDirect
Journal of Psychiatric Research journal homepage: www.elsevier.com/locate/jpsychires
Hypomethylation of LINE-1 elements in schizophrenia and bipolar disorder a,1
a,b,1
c
d
a
a
T
e
Shufen Li , Qiong Yang , Yu Hou , Tingyun Jiang , Lu Zong , Zhongju Wang , Xia Luo , Wenquan Lianga, Hu Zhaoe,∗, Yuping Ninga,b,∗∗, Cunyou Zhaoa,f,∗∗∗ a
Department of Medical Genetics, School of Basic Medical Sciences, Southern Medical University, Guangzhou, Guangdong, China Department of Psychiatry, The Affiliated Brain Hospital of Guangzhou Medical University (Guangzhou Huiai Hospital), Guangzhou, Guangdong, China Department of Pediatric Neurology, Affiliated BaYi Children's Hospital, PLA Army General Hospital, Beijing, China d The Third People's Hospital of Zhongshan, Zhongshan, Guangdong, China e Faculty of Forensic Medicine, Zhongshan School of Medicine, Sun Yat-sen University, Guangzhou, China f Key Laboratory of Mental Health of the Ministry of Education, Southern Medical University, Guangzhou, Guangdong, China b c
A R T I C LE I N FO
A B S T R A C T
Keywords: Psychiatric disorder LINE-1 Global DNA methylation Epigenetics Retrotransposon
Schizophrenia (SCZ) and bipolar disorder (BPD) are severe mental illnesses with evidence of significant genetic and environmental etiological elements in their complex etiologies. 5‘-Methylcytosine is the main epigenetic DNA modification that mediates the interplay between genetic and environmental components. In humans, most 5‘-methylcytosine modifications are observed in CpG-rich regions within the long interspersed nuclear element (LINE-1). LINE-1 is a mobile retrotransposon that comprises ∼17% of the human genome, and its methylation levels are highly correlated with global DNA methylation levels. LINE-1 insertions are also reported to be mental illnesses-associated genomic risk factors. To examine the LINE-1 methylation levels in SCZ and BPD, this study employed a bisulfite conversion-specific one-label extension (BS-OLE) method to detect the methylation levels at three CpG sites (S1, S2 and S3) of LINE-1 in peripheral blood DNA from a Han Chinese cohort composed of 92 SCZ patients, 99 BPD patients and 92 controls (CON). The results showed a decreased S1 methylation level in SCZ, decreased S2 methylation level in BPD and decreased S3 methylation levels in both SCZ and BPD relative to those of the CON. A female-dependent positive correlation of the S3 methylation level with age in CON became non-significant in both SCZ and BPD. These findings demonstrated that LINE-1 methylation varied with development and disease status. The roles of LINE-1 methylation in the pathogenesis of SCZ and BPD remain to be elucidated.
1. Introduction Schizophrenia (SCZ) and bipolar disorder (BPD) are severe mental illnesses with evidence of significant genetic and environmental components in their complex etiologies (Auta et al., 2013). Epigenetic modifications play a seminal role in mediating the interplay between the genetic and environmental components underlying the development of these diseases (Teroganova et al., 2016). DNA methylation at a cytosine in CpG dinucleotide sites is the main epigenetic DNA methylation in the mammalian genome; this modification is found in CpG-rich regions within some retroviral repeat sequences, such as the long interspersed nuclear element (LINE-1), short interspersed nuclear element
(SINE) (He et al., 2014). Hypomethylation of these elements is involved in pathological processes, including tumorigenesis, development, cell differentiation, and aging (Babushok and Kazazian, 2007). LINE-1 is the active autonomous non-LTR retrotransposon in mammalian genomes and accounts for approximately 17% of the human genome (Penzkofer et al., 2017). The full-length mammalian LINE-1 consists of the 5‘-untranslated region (5‘UTR), two open reading frames and the 3‘-UTR encoding an RNA-binding protein and an endonuclease/reverse transcriptase fusion protein. LINE-1 retrotransposon activity can ‘copy’ and ‘paste’ itself into multiple genomic locations and has a significant impact on genomic structure and stability through processes such as gene disruption, transcriptional
∗
Corresponding author. Faculty of Forensic Medicine, Zhongshan School of Medicine, Sun Yat-sen University, Guangzhou, China. Corresponding author. Department of Psychiatry, the Affiliated Brain Hospital of Guangzhou Medical University (Guangzhou Huiai Hospital), Guangzhou, Guangdong, China. ∗∗∗ Corresponding author. Department of Medical Genetics, School of Basic Medical Sciences, Southern Medical University, Guangzhou, Guangdong, 510515, China. E-mail addresses: [email protected] (H. Zhao), [email protected] (Y. Ning), [email protected] (C. Zhao). 1 These authors contributed equally to this study. ∗∗
https://doi.org/10.1016/j.jpsychires.2018.10.009 Received 31 August 2018; Received in revised form 1 October 2018; Accepted 8 October 2018 0022-3956/ © 2018 Elsevier Ltd. All rights reserved.
Journal of Psychiatric Research 107 (2018) 68–72
S. Li et al.
with ethanol, and air-drying before dissolving in Tris-EDTA buffer (pH 8.0).
regulation and mobilization of Alu sequences (Penzkofer et al., 2017). Recent studies have shown that LINE-1 insertion is a genomic risk factor underlying the development of SCZ (Doyle et al., 2017; Guffanti et al., 2016), and LINE-1 expression has been examined in various tissues and may be related to cell proliferations and differentiation (Steinhoff and Schulz, 2004). DNA methylation is considered an important epigenetic mechanism in the regulation of LINE-1 expression. The LINE-1 promoters are CpG-rich regions and are highly methylated and silenced under normal conditions (Steinhoff and Schulz, 2004). Since LINE-1 elements are widely spread in the human genome, their methylation levels have been reported to be highly correlated with the global DNA methylation levels (Hoffmann and Schulz, 2005). Global DNA methylation alterations and increased LINE-1 genetic burdens have been observed in peripheral leukocytes from patients with SCZ or major depression (Bundo et al., 2014; Doyle et al., 2017; Jiang et al., 2017; Liu et al., 2016; Tseng et al., 2014). However, the methylation levels of LINE-1 elements in SCZ and BPD patients remain to be examined. In this study, we employ a bisulfite conversion-specific one-label extension (BS-OLE) method to examine the methylation levels of LINE-1 elements in peripheral blood DNA from a Han Chinese cohort composed of SCZ, BPD patients and controls to find differences in LINE-1 methylation pattern between different diagnostic status and to evaluate their roles in the development of these severe mental illnesses.
2.2. Bisulfite modification and BS-OLE assay Bisulfite conversion of genomic DNA was performed by denaturing 1 μg of genomic DNA with 0.3 mol/L of NaOH at 42 °C for 20 min, followed by incubation at 95 °C for 3 min and 0 °C for 1 min. Next, the samples were incubated at pH 5.0 with sodium metabisulfite (2.0 mol/ L) and hydroquinone (0.5 mmol/L) at 55 °C for 16 h in the dark and then preserved by overlaying with mineral oil (Li et al., 2015). The converted DNA was purified with the Promega Wizard DNA Clean Up System (Promega, USA). The eluted DNA was incubated with NaOH (0.3 mol/L) at 37 °C for 30 min and neutralized with 0.3 mol/L of NH4acetate to pH 7.0. The neutralized DNA was precipitated by 75% ethanol and recovered in 30 μL of TE buffer. The bisulfite-converted DNA was stored at −20 °C prior to the BS-PCR. The BS-PCR was employed to amplify target DNA segment of bisulfite-converted DNA using the LineF and LineR primers (Fig. 1). Amplification consisted of initial denaturation at 95 °C for 2 min, 40 cycles each of 30 s at 95 °C, 1 min at 60 °C, and 1 min at 72 °C, plus a final extension step at 72 °C for 5 min. The BS-PCR products were stored at −20 °C prior to the BS-OLE assay. The DNA methylation levels at the S1, S2 and S3 CpG sites of the LINE-1 5‘UTR region (Fig. 1) were examined with the BS-OLE method as previously described (Li et al., 2015). Briefly, the BS-OLE assay was performed in two separate reactions to quantify the methylated (viz. C) or unmethylated (viz. T) nucleotide levels at the target site. Each reaction was performed in a 20-μL reaction mixture comprising 100 ng of BS-PCR products, 250 nM extension primer (LineF1 for site S1, LineF2 for site S2 or LineF3 for site S3), 1 unit of polymerase (TaKaRa, Dalian, China), polymerase buffer (TaKaRa) and 100 nM of TAMRA-dCTP (5propargylamino-dCTP-5/6-TAMRA, Jena Bioscience, USA) or 50 nM of TAMRA-dUTP (Jena Bioscience). After an initial denaturation step for 3 min at 95 °C, primer extension was conducted by 30 thermal cycles consisting of 30 s at 95 °C and 30 s at 55 °C, plus a final extension step at 72 °C for 5 min. The fluorescence polarization (FP) value of each BS-OLE reaction was measured in 96- or 384-well plates using the Multimode Plate Reader Victor X5 (Perkin Elmer, USA) with an excitation wavelength of 544 nm (Filter FP544/15) and an emission wavelength of 595 nm (Filter 595/60) without the need for purification or separation of the reaction products. After correction for the blank FP reading, the FP readings of the dCTP- and dUTP-incorporated samples yielded the
2. Materials and methods 2.1. Subjects All studies were approved by the local Medical Ethics Committee and conducted in accordance with the Declaration of Helsinki. Informed consent was obtained from the participants after the nature of the procedures had been fully explained. Peripheral blood samples were obtained from Han Chinese patients with SCZ or BPD who met the diagnostic criteria of SCZ or BPD according to the Diagnostic and Statistical Manual of Mental Disorders, 4th edition (American Psychiatric Association). Unrelated non-psychiatric controls were used. The characteristics from a total of 283 participants, including patients with SCZ [n = 92, male/female = 62/30; age (mean ± sd) = 40.54 ± 10.50 years] or BPD [n = 99, male/female = 62/37; age = 31.57 ± 10.99 years] and the controls [n = 92, male/female = 62/30; age = 40.28 ± 9.84 years), are presented in Table 1. Genomic DNA was extracted from peripheral blood cells by phenol/ chloroform extraction followed by isopropanol precipitation, washing Table 1 Contributions of clinical and demographic characteristics to the methylation levels. Clinical characteristics Adjusted R2
Demographic characteristics Standardized coefficient
Adjusted R2
Clozapine S1 S2 S3
0.045* 0.061* 0.067*
0.238* NS NS
NS 0.075*** 0.098***
Standardized coefficient Diagnosis
Age
Gender
NS −0.297*** −0.169**
NS NS 0.186**
NS NS −0.149*
Antipsychotic medications (N, %) of the schizophrenic cohort [duration of illness (mean ± sd) = 13.43 ± 8.79 years] are shown as follows: Risperidone (34, 37.0), Clozapine (36, 39.1), Olanzapine (17, 18.5), Quetiapine (16, 17.4), Aripiprazole (7, 7.6), Sodium valproate (13,14.1), Sulpiride (2, 2.2), Chlorpromazine (2, 2.2), Ziprasidone (1, 1.1), Magnesium valproate (7, 7.6), Paliperidone (2, 2.2), Oxcarbazepine (2, 2.2), Lithium carbonate (3, 3.3), Trazodone (1, 1.1), Loratadine (1, 1.1), Oxiracetam (1, 1.1), Haloperidol (1, 1.1), and Unknown (5, 5.4). The use of multiple antipsychotic medications (N, %) are shown as follows: 1 drug (37, 40.2), 2 drugs (41, 44.6), and 3 drugs (9, 9.8). Antipsychotic medications (N, %) of the bipolar disorder patients [duration of illness (mean ± sd) = 9.08 ± 8.07 years)] are shown as follows: Risperidone (7, 7.1), Clozapine (7, 7.1), Olanzapine (4, 4.0), Quetiapine (5, 5.1), Aripiprazole (4, 4.0), Sodium valproate (13,13.1), Ziprasidone (2, 2.0), Magnesium valproate (8, 8.1), Lithium carbonate (6, 6.1), Sertraline (5, 5.1), Mirtazapine (1, 1.0), Bupropion (1, 1.0), Benzhexol (7, 7.1), Venlafaxine (4, 4.0), Flupenthixol Melitracen (2, 2.0), Haloperidol (1, 1.0), Lamotrigine (2, 2.0), Escitalopram (1, 1.0), and Unknown (71, 71.7). The use of multiple antipsychotic medications (N, %) are shown as follows: 2 drugs (13, 13.1), 3 drugs (8, 8.1), 4 drugs (5, 5.1) and 5 drugs (2, 2.0). Contributions of clinical or demographic characteristics to the LINE-1 S1, S2, or S3 methylation levels were assessed with a linear regression and shown with * P < 0.05, **P < 0.01 and ***P < 0.001. NS, no significance. 69
Journal of Psychiatric Research 107 (2018) 68–72
S. Li et al.
Fig. 1. The LINE-1 5‘UTR DNA fragment. Reverse complementary sequences of the human LINE-1 transposon (L1Hs) (gi|34196|emb|X58075.1) 5‘UTR DNA are shown. Upper lines are shown for original sequences (Ref-seq.) and lower for bisulfite-converted sequences (BS-seq.). The positions of primers employed in this study are underlined and the three CpG sites examined in this study are marked with S1, S2 and S3.
a covariate, significant differences in the S3 methylation levels were found between the controls and the patients with SCZ (Pcovariate < 0.001) or BPD (Pcovariate = 0.008) only in the males (Fig. S1). With respect to age, we only observed a positive correlation of S3 CpG site methylation with age for the combined male and female cohort (R2 = 0.07; Fig. 2D) or the individual female cohort (R2 = 0.20; Fig. 2F) of the controls, and for the combined male and female cohort (R2 = 0.05; Fig. 2G) or the individual male cohort (R2 = 0.13; Fig. 2H) of the BPD patients; a trend line was observed for the SCZ (Fig. 2J-L). We also observed a significant difference between the slopes of the control and the SCZ regression lines relating S3 CpG site methylation to age in the female cohorts (P = 0.02), which displayed positive regression in the controls (Fig. 2F) but a negative trend line in the SCZ (Fig. 2L).
extent of methylation in the sample: Methylation = 100 × dCTP/(dCTP + dUTP)% where dCTP or dUTP represents the blank-corrected FP readings from the TAMRA-dCTP or -dUTP incorporations. 2.3. Statistical analysis A stepwise linear regression is employed to evaluate the relationship between the LINE-1 methylation levels and the clinical characteristics, including the illness duration and use of antipsychotic medications in the patients with SCZ and BPD, or to examine the relationship between the LINE-1 methylation level and the diagnosis, age and gender among the patients and controls. The differences in methylation levels between the patients with SCZ or BPD and the control group were analyzed with Student's t-test or ANCOVA with age as a covariate.
4. Discussion 3. Results Accumulating evidence supports the involvement of DNA methylation in the development of SCZ and BPD (Teroganova et al., 2016). The major findings of the present study demonstrated hypomethylation of LINE-1 in peripheral blood samples from patients with SCZ or BPD. Consistent hypomethylation of LINE-1 S3 CpG sites in both SCZ and BPD patients reflects a shared epigenetic variant between these two diseases, whereas hypomethylation of the S1 site in SCZ patients and the S2 site in BPD patients suggest the existence of disease-specific epigenetic variants in SCZ or BPD. The shared epigenetic variants between the two disorders observed in this study, together with shared symptoms and genetic variants (Cross-Disorder Group of the Psychiatric Genomics Consortium, 2013), further suggest there is a biological overlap between them and encourage investigations into shared pathophysiologies across these two disorders. LINE-1 methylation levels have been reported to be associated with the global DNA methylation levels, and the decreased LINE-1 methylation levels observed in the present study are also in agreement with previous observation of global hypomethylation in SCZ or BPD patients (Huzayyin et al., 2014; Melas et al., 2012; Shimabukuro et al., 2007). However, there are reports of increased or unchanged methylation levels in these mental illnesses (Ceylan et al., 2018; Jiang et al., 2017). This conflict may be due to differences in the methods used to detect the methylation profiles or the demographic and clinical variables of the patients. Recently, hypomethylation of LINE-1 was only observed in first-episode SCZ patients with a history of childhood trauma (Misiak et al., 2015), and LINE-1 copy number was increased in the patients with major depressive disorder possibly due to its hypomethylation (Liu et al., 2016). LINE-1 is the active autonomous retrotransposon in mammalian genomes, and hypomethylation of LINE-1 may be associated with its retrotransposon activity. This activity has a significant impact on genomic structure and
To examine the methylation levels of LINE-1, we employed bisulfite (BS)-PCR to amplify the BS-converted DNA of LINE-1 and to determine the DNA methylation levels of three CpG sites (S1, S2 and S3; Fig. 1) of the LINE-1 5‘UTR using a BS-OLE method (Li et al., 2015) from peripheral blood DNA of 92 controls and 92 patients with SCZ or 99 patients with BPD. The effects of the clinical characteristics on the LINE-1 methylation levels among the SCZ or BPD patients were first assessed with a linear regression to generate a model in which administration of Clozapine explained 4.5% of the variance in the S1 methylation level among the SCZ patients (Table 1). Then, the demographic characteristic effects among the combined patient and control cohort were assessed with a linear regression, which showed that 7.5% of the variance in the S2 methylation level was explained by the diagnostic status and 9.8% of the variance in the S3 methylation level was explained by the diagnostic status, age and gender. Therefore, age was employed as a covariate in the S3 methylation level analysis by ANCOVA, which was performed in a combined male and female cohort or separate male and female cohorts. Compared with the controls, the LINE-1 methylation levels were significantly decreased by 5% for the S1 CpG site (P < 0.001; Fig. 2A) and by 3% for the S3 site (P = 0.002; Fig. 2C) in the SCZ cohort and were decreased by 8% for the S2 site (P < 0.001; Fig. 2B) and by 5% for the S3 site (P = 0.002; Fig. 2C) in the BPD cohort. We also observed significant differences in the S1 and S2 methylation levels between the SCZ and BPD patients, with an increased S1 methylation level (by 5%, P < 0.001; Fig. 2A) and a decreased S2 methylation level (by 12%, P < 0.001; Fig. 2B) in the BPD patients compared to those of the SCZ. When the males were analyzed separately from the females with age as 70
Journal of Psychiatric Research 107 (2018) 68–72
S. Li et al.
Fig. 2. LINE-1 methylation levels in the controls and patients with schizophrenia or bipolar disorder. (A–C) LINE-1 S1 (A), S2 (B) and S3 (C) site methylation levels in the control (CON), schizophrenia (SCZ) and bipolar disorder (BPD) groups were determined by the BS-OLE method (Fig. S1). The horizontal bars represent the mean methylation levels. The methylation level of each subject was normalized to the mean level of the control cohort. The y-axis represents the percentage of methylation relative to the mean level of the control cohort. △ = controls; ▼ = patients with schizophrenia; ▲ = patients with bipolar disorder. Significant differences between two groups were determined using Student's t-test analysis (*P < 0.05, **P < 0.01 and ***P < 0.001). (D-L) Relationships between the LINE1 S3 methylation level and age in the combined or separate female and male subgroups of the CON (D-F), BPD (G-I) and SCZ (J-L) cohort are represented by solid lines (where the P value of the regression coefficient was < 0.05) or dashed lines (where P ≥ 0.05).
for the development of SCZ. The present study has some limitation that should be addressed. First, the present results were based on a relatively small sample size, and larger independent samples are required. Second, the majority of patients were not drug-free, and antipsychotic treatment might have impacts on the DNA methylation levels. Third, the BS-OLE method employed in the present study cannot distinguish hydroxymethylation from methylation. Finally, DNA methylation analysis was performed for only three CpG sites of LINE-1, which only account for a small portion of the CpG sites in LINE-1. In conclusion, the present study revealed hypomethylation of LINE1 in SCZ and BPD patients and a female-dependent increase in the LINE1 methylation level with age in the controls that lost its significance in patients with SCZ or BPD. The role of demethylation of LINE-1 in the
stability and is involved in pathological processes, development, cell differentiation and aging. Moreover, the female-dependent increase in the LINE-1 methylation level with age in the controls lost its significance in the patients with SCZ or BPD in the present study, suggesting an important role of LINE-1 methylation in the development of these diseases and sex-specific endocrine mechanism underlying the epigenetic alteration. In fact, LINE-1 retrotransposition has been observed during neurogenesis in the hippocampus (Muotri et al., 2009) as well as in neurons from the prefrontal cortex of patients with SCZ or in induced pluripotent stem cells from patients with 22q11 deletion syndrome (Bundo et al., 2014). The hypomethylation of LINE-1 observed in the present study together with the increase in LINE-1 polymorphic insertions associated with SCZ (Doyle et al., 2017; Guffanti et al., 2016) indicated an important role of LINE-1 insertions as a genomic risk factor 71
Journal of Psychiatric Research 107 (2018) 68–72
S. Li et al.
pathogenesis of SCZ and BPD remains to be elucidated.
analysis. Lancet 381, 1371–1379. https://doi.org/10.1016/S0140-6736(12)62129-1. Doyle, G.A., et al., 2017. Analysis of LINE-1 elements in DNA from postmortem brains of individuals with schizophrenia. Neuropsychopharmacology 42, 2602–2611. https:// doi.org/10.1038/npp.2017.115. Guffanti, G., et al., 2016. LINE1 insertions as a genomic risk factor for schizophrenia: preliminary evidence from an affected family. Am J Med Genet B Neuropsychiatr Genet 171, 534–545. https://doi.org/10.1002/ajmg.b.32437. He, Z.M., et al., 2014. Transition of LINE-1 DNA methylation status and altered expression in first and third trimester placentas. PloS One 9, e96994. https://doi.org/10.1371/ journal.pone.0096994. Hoffmann, M.J., Schulz, W.A., 2005. Causes and consequences of DNA hypomethylation in human cancer. Biochem. Cell Biol. 83, 296–321. https://doi.org/10.1139/o05036. Huzayyin, A.A., et al., 2014. Decreased global methylation in patients with bipolar disorder who respond to lithium. Int. J. Neuropsychopharmacol. 17, 561–569. https:// doi.org/10.1017/S1461145713001569. Jiang, T., et al., 2017. Variation in global DNA hydroxymethylation with age associated with schizophrenia. Psychiatr. Res. 257, 497–500. https://doi.org/10.1016/j. psychres.2017.08.022. Li, S., et al., 2015. Fluorescence polarization-based method with bisulfite conversionspecific one-label extension for quantification of single CpG dinucleotide methylation. Genome 58, 357–363. https://doi.org/10.1139/gen-2014-0185. Liu, S., et al., 2016. Inverse changes in L1 retrotransposons between blood and brain in major depressive disorder. Sci. Rep. 6, 37530. https://doi.org/10.1038/srep37530. Melas, P.A., et al., 2012. Epigenetic aberrations in leukocytes of patients with schizophrenia: association of global DNA methylation with antipsychotic drug treatment and disease onset. Faseb. J. 26, 2712–2718. https://doi.org/10.1096/fj.11-202069. Misiak, B., et al., 2015. Lower LINE-1 methylation in first-episode schizophrenia patients with the history of childhood trauma. Epigenomics 7, 1275–1285. https://doi.org/ 10.2217/epi.15.68. Muotri, A.R., et al., 2009. Environmental influence on L1 retrotransposons in the adult hippocampus. Hippocampus 19, 1002–1007. https://doi.org/10.1002/hipo.20564. Penzkofer, T., et al., 2017. L1Base 2: more retrotransposition-active LINE-1s, more mammalian genomes. Nucleic Acids Res. 45, D68–D73. https://doi.org/10.1093/ nar/gkw925. Shimabukuro, M., et al., 2007. Global hypomethylation of peripheral leukocyte DNA in male patients with schizophrenia: a potential link between epigenetics and schizophrenia. J. Psychiatr. Res. 41, 1042–1046. https://doi.org/10.1016/j.jpsychires. 2006.08.006. Steinhoff, C., Schulz, W.A., 2004. Transcriptional regulation of the human LINE-1 retrotransposon L1.2B. Mol. Genet. Genom. 270, 394–402. https://doi.org/10.1007/ s00438-003-0931-2. Teroganova, N., et al., 2016. DNA methylation in peripheral tissue of schizophrenia and bipolar disorder: a systematic review. BMC Genet. 17, 27. https://doi.org/10.1186/ s12863-016-0332-2. Tseng, P.T., et al., 2014. Age-associated decrease in global DNA methylation in patients with major depression. Neuropsychiatric Dis. Treat. 10, 2105–2114. https://doi.org/ 10.2147/NDT.S71997.
Role of the funding source This work was supported by the Guangdong-Hong Kong Technology Cooperation Funding Scheme of the Guangdong Science and Technology Foundation [grant number 2017A050506026], the National Natural Science Foundation of China [grant number 8167133, 81371475], the National Key Research and Development Program of China [grant number 2016YFC1201800] and the Program for Changjiang Scholars and Innovative Research Team in University [grant number IRT_16R37]. The funding had no further role in study design; in the collection, analysis and interpretation of data; in the writing of the report; or in the decision to submit the paper for publication. Conflicts of interest The authors declare no conflict of interest. Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.jpsychires.2018.10.009. References Auta, J., et al., 2013. DNA-methylation gene network dysregulation in peripheral blood lymphocytes of schizophrenia patients. Schizophr. Res. 150, 312–318. https://doi. org/10.1016/j.schres.2013.07.030. Babushok, D.V., Kazazian Jr., H.H., 2007. Progress in understanding the biology of the human mutagen LINE-1. Hum. Mutat. 28, 527–539. https://doi.org/10.1002/humu. 20486. Bundo, M., et al., 2014. Increased l1 retrotransposition in the neuronal genome in schizophrenia. Neuron 81, 306–313. https://doi.org/10.1016/j.neuron.2013.10.053. Ceylan, D., et al., 2018. DNA redox modulations and global DNA methylation in bipolar disorder: effects of sex, smoking and illness state. Psychiatr. Res. 261, 589–596. https://doi.org/10.1016/j.psychres.2017.12.051. Cross-disorder Group of the psychiatric genomics consortium, C., 2013. Identification of risk loci with shared effects on five major psychiatric disorders: a genome-wide
72
Contents lists available at ScienceDirect
Journal of Psychiatric Research journal homepage: www.elsevier.com/locate/jpsychires
Hypomethylation of LINE-1 elements in schizophrenia and bipolar disorder a,1
a,b,1
c
d
a
a
T
e
Shufen Li , Qiong Yang , Yu Hou , Tingyun Jiang , Lu Zong , Zhongju Wang , Xia Luo , Wenquan Lianga, Hu Zhaoe,∗, Yuping Ninga,b,∗∗, Cunyou Zhaoa,f,∗∗∗ a
Department of Medical Genetics, School of Basic Medical Sciences, Southern Medical University, Guangzhou, Guangdong, China Department of Psychiatry, The Affiliated Brain Hospital of Guangzhou Medical University (Guangzhou Huiai Hospital), Guangzhou, Guangdong, China Department of Pediatric Neurology, Affiliated BaYi Children's Hospital, PLA Army General Hospital, Beijing, China d The Third People's Hospital of Zhongshan, Zhongshan, Guangdong, China e Faculty of Forensic Medicine, Zhongshan School of Medicine, Sun Yat-sen University, Guangzhou, China f Key Laboratory of Mental Health of the Ministry of Education, Southern Medical University, Guangzhou, Guangdong, China b c
A R T I C LE I N FO
A B S T R A C T
Keywords: Psychiatric disorder LINE-1 Global DNA methylation Epigenetics Retrotransposon
Schizophrenia (SCZ) and bipolar disorder (BPD) are severe mental illnesses with evidence of significant genetic and environmental etiological elements in their complex etiologies. 5‘-Methylcytosine is the main epigenetic DNA modification that mediates the interplay between genetic and environmental components. In humans, most 5‘-methylcytosine modifications are observed in CpG-rich regions within the long interspersed nuclear element (LINE-1). LINE-1 is a mobile retrotransposon that comprises ∼17% of the human genome, and its methylation levels are highly correlated with global DNA methylation levels. LINE-1 insertions are also reported to be mental illnesses-associated genomic risk factors. To examine the LINE-1 methylation levels in SCZ and BPD, this study employed a bisulfite conversion-specific one-label extension (BS-OLE) method to detect the methylation levels at three CpG sites (S1, S2 and S3) of LINE-1 in peripheral blood DNA from a Han Chinese cohort composed of 92 SCZ patients, 99 BPD patients and 92 controls (CON). The results showed a decreased S1 methylation level in SCZ, decreased S2 methylation level in BPD and decreased S3 methylation levels in both SCZ and BPD relative to those of the CON. A female-dependent positive correlation of the S3 methylation level with age in CON became non-significant in both SCZ and BPD. These findings demonstrated that LINE-1 methylation varied with development and disease status. The roles of LINE-1 methylation in the pathogenesis of SCZ and BPD remain to be elucidated.
1. Introduction Schizophrenia (SCZ) and bipolar disorder (BPD) are severe mental illnesses with evidence of significant genetic and environmental components in their complex etiologies (Auta et al., 2013). Epigenetic modifications play a seminal role in mediating the interplay between the genetic and environmental components underlying the development of these diseases (Teroganova et al., 2016). DNA methylation at a cytosine in CpG dinucleotide sites is the main epigenetic DNA methylation in the mammalian genome; this modification is found in CpG-rich regions within some retroviral repeat sequences, such as the long interspersed nuclear element (LINE-1), short interspersed nuclear element
(SINE) (He et al., 2014). Hypomethylation of these elements is involved in pathological processes, including tumorigenesis, development, cell differentiation, and aging (Babushok and Kazazian, 2007). LINE-1 is the active autonomous non-LTR retrotransposon in mammalian genomes and accounts for approximately 17% of the human genome (Penzkofer et al., 2017). The full-length mammalian LINE-1 consists of the 5‘-untranslated region (5‘UTR), two open reading frames and the 3‘-UTR encoding an RNA-binding protein and an endonuclease/reverse transcriptase fusion protein. LINE-1 retrotransposon activity can ‘copy’ and ‘paste’ itself into multiple genomic locations and has a significant impact on genomic structure and stability through processes such as gene disruption, transcriptional
∗
Corresponding author. Faculty of Forensic Medicine, Zhongshan School of Medicine, Sun Yat-sen University, Guangzhou, China. Corresponding author. Department of Psychiatry, the Affiliated Brain Hospital of Guangzhou Medical University (Guangzhou Huiai Hospital), Guangzhou, Guangdong, China. ∗∗∗ Corresponding author. Department of Medical Genetics, School of Basic Medical Sciences, Southern Medical University, Guangzhou, Guangdong, 510515, China. E-mail addresses: [email protected] (H. Zhao), [email protected] (Y. Ning), [email protected] (C. Zhao). 1 These authors contributed equally to this study. ∗∗
https://doi.org/10.1016/j.jpsychires.2018.10.009 Received 31 August 2018; Received in revised form 1 October 2018; Accepted 8 October 2018 0022-3956/ © 2018 Elsevier Ltd. All rights reserved.
Journal of Psychiatric Research 107 (2018) 68–72
S. Li et al.
with ethanol, and air-drying before dissolving in Tris-EDTA buffer (pH 8.0).
regulation and mobilization of Alu sequences (Penzkofer et al., 2017). Recent studies have shown that LINE-1 insertion is a genomic risk factor underlying the development of SCZ (Doyle et al., 2017; Guffanti et al., 2016), and LINE-1 expression has been examined in various tissues and may be related to cell proliferations and differentiation (Steinhoff and Schulz, 2004). DNA methylation is considered an important epigenetic mechanism in the regulation of LINE-1 expression. The LINE-1 promoters are CpG-rich regions and are highly methylated and silenced under normal conditions (Steinhoff and Schulz, 2004). Since LINE-1 elements are widely spread in the human genome, their methylation levels have been reported to be highly correlated with the global DNA methylation levels (Hoffmann and Schulz, 2005). Global DNA methylation alterations and increased LINE-1 genetic burdens have been observed in peripheral leukocytes from patients with SCZ or major depression (Bundo et al., 2014; Doyle et al., 2017; Jiang et al., 2017; Liu et al., 2016; Tseng et al., 2014). However, the methylation levels of LINE-1 elements in SCZ and BPD patients remain to be examined. In this study, we employ a bisulfite conversion-specific one-label extension (BS-OLE) method to examine the methylation levels of LINE-1 elements in peripheral blood DNA from a Han Chinese cohort composed of SCZ, BPD patients and controls to find differences in LINE-1 methylation pattern between different diagnostic status and to evaluate their roles in the development of these severe mental illnesses.
2.2. Bisulfite modification and BS-OLE assay Bisulfite conversion of genomic DNA was performed by denaturing 1 μg of genomic DNA with 0.3 mol/L of NaOH at 42 °C for 20 min, followed by incubation at 95 °C for 3 min and 0 °C for 1 min. Next, the samples were incubated at pH 5.0 with sodium metabisulfite (2.0 mol/ L) and hydroquinone (0.5 mmol/L) at 55 °C for 16 h in the dark and then preserved by overlaying with mineral oil (Li et al., 2015). The converted DNA was purified with the Promega Wizard DNA Clean Up System (Promega, USA). The eluted DNA was incubated with NaOH (0.3 mol/L) at 37 °C for 30 min and neutralized with 0.3 mol/L of NH4acetate to pH 7.0. The neutralized DNA was precipitated by 75% ethanol and recovered in 30 μL of TE buffer. The bisulfite-converted DNA was stored at −20 °C prior to the BS-PCR. The BS-PCR was employed to amplify target DNA segment of bisulfite-converted DNA using the LineF and LineR primers (Fig. 1). Amplification consisted of initial denaturation at 95 °C for 2 min, 40 cycles each of 30 s at 95 °C, 1 min at 60 °C, and 1 min at 72 °C, plus a final extension step at 72 °C for 5 min. The BS-PCR products were stored at −20 °C prior to the BS-OLE assay. The DNA methylation levels at the S1, S2 and S3 CpG sites of the LINE-1 5‘UTR region (Fig. 1) were examined with the BS-OLE method as previously described (Li et al., 2015). Briefly, the BS-OLE assay was performed in two separate reactions to quantify the methylated (viz. C) or unmethylated (viz. T) nucleotide levels at the target site. Each reaction was performed in a 20-μL reaction mixture comprising 100 ng of BS-PCR products, 250 nM extension primer (LineF1 for site S1, LineF2 for site S2 or LineF3 for site S3), 1 unit of polymerase (TaKaRa, Dalian, China), polymerase buffer (TaKaRa) and 100 nM of TAMRA-dCTP (5propargylamino-dCTP-5/6-TAMRA, Jena Bioscience, USA) or 50 nM of TAMRA-dUTP (Jena Bioscience). After an initial denaturation step for 3 min at 95 °C, primer extension was conducted by 30 thermal cycles consisting of 30 s at 95 °C and 30 s at 55 °C, plus a final extension step at 72 °C for 5 min. The fluorescence polarization (FP) value of each BS-OLE reaction was measured in 96- or 384-well plates using the Multimode Plate Reader Victor X5 (Perkin Elmer, USA) with an excitation wavelength of 544 nm (Filter FP544/15) and an emission wavelength of 595 nm (Filter 595/60) without the need for purification or separation of the reaction products. After correction for the blank FP reading, the FP readings of the dCTP- and dUTP-incorporated samples yielded the
2. Materials and methods 2.1. Subjects All studies were approved by the local Medical Ethics Committee and conducted in accordance with the Declaration of Helsinki. Informed consent was obtained from the participants after the nature of the procedures had been fully explained. Peripheral blood samples were obtained from Han Chinese patients with SCZ or BPD who met the diagnostic criteria of SCZ or BPD according to the Diagnostic and Statistical Manual of Mental Disorders, 4th edition (American Psychiatric Association). Unrelated non-psychiatric controls were used. The characteristics from a total of 283 participants, including patients with SCZ [n = 92, male/female = 62/30; age (mean ± sd) = 40.54 ± 10.50 years] or BPD [n = 99, male/female = 62/37; age = 31.57 ± 10.99 years] and the controls [n = 92, male/female = 62/30; age = 40.28 ± 9.84 years), are presented in Table 1. Genomic DNA was extracted from peripheral blood cells by phenol/ chloroform extraction followed by isopropanol precipitation, washing Table 1 Contributions of clinical and demographic characteristics to the methylation levels. Clinical characteristics Adjusted R2
Demographic characteristics Standardized coefficient
Adjusted R2
Clozapine S1 S2 S3
0.045* 0.061* 0.067*
0.238* NS NS
NS 0.075*** 0.098***
Standardized coefficient Diagnosis
Age
Gender
NS −0.297*** −0.169**
NS NS 0.186**
NS NS −0.149*
Antipsychotic medications (N, %) of the schizophrenic cohort [duration of illness (mean ± sd) = 13.43 ± 8.79 years] are shown as follows: Risperidone (34, 37.0), Clozapine (36, 39.1), Olanzapine (17, 18.5), Quetiapine (16, 17.4), Aripiprazole (7, 7.6), Sodium valproate (13,14.1), Sulpiride (2, 2.2), Chlorpromazine (2, 2.2), Ziprasidone (1, 1.1), Magnesium valproate (7, 7.6), Paliperidone (2, 2.2), Oxcarbazepine (2, 2.2), Lithium carbonate (3, 3.3), Trazodone (1, 1.1), Loratadine (1, 1.1), Oxiracetam (1, 1.1), Haloperidol (1, 1.1), and Unknown (5, 5.4). The use of multiple antipsychotic medications (N, %) are shown as follows: 1 drug (37, 40.2), 2 drugs (41, 44.6), and 3 drugs (9, 9.8). Antipsychotic medications (N, %) of the bipolar disorder patients [duration of illness (mean ± sd) = 9.08 ± 8.07 years)] are shown as follows: Risperidone (7, 7.1), Clozapine (7, 7.1), Olanzapine (4, 4.0), Quetiapine (5, 5.1), Aripiprazole (4, 4.0), Sodium valproate (13,13.1), Ziprasidone (2, 2.0), Magnesium valproate (8, 8.1), Lithium carbonate (6, 6.1), Sertraline (5, 5.1), Mirtazapine (1, 1.0), Bupropion (1, 1.0), Benzhexol (7, 7.1), Venlafaxine (4, 4.0), Flupenthixol Melitracen (2, 2.0), Haloperidol (1, 1.0), Lamotrigine (2, 2.0), Escitalopram (1, 1.0), and Unknown (71, 71.7). The use of multiple antipsychotic medications (N, %) are shown as follows: 2 drugs (13, 13.1), 3 drugs (8, 8.1), 4 drugs (5, 5.1) and 5 drugs (2, 2.0). Contributions of clinical or demographic characteristics to the LINE-1 S1, S2, or S3 methylation levels were assessed with a linear regression and shown with * P < 0.05, **P < 0.01 and ***P < 0.001. NS, no significance. 69
Journal of Psychiatric Research 107 (2018) 68–72
S. Li et al.
Fig. 1. The LINE-1 5‘UTR DNA fragment. Reverse complementary sequences of the human LINE-1 transposon (L1Hs) (gi|34196|emb|X58075.1) 5‘UTR DNA are shown. Upper lines are shown for original sequences (Ref-seq.) and lower for bisulfite-converted sequences (BS-seq.). The positions of primers employed in this study are underlined and the three CpG sites examined in this study are marked with S1, S2 and S3.
a covariate, significant differences in the S3 methylation levels were found between the controls and the patients with SCZ (Pcovariate < 0.001) or BPD (Pcovariate = 0.008) only in the males (Fig. S1). With respect to age, we only observed a positive correlation of S3 CpG site methylation with age for the combined male and female cohort (R2 = 0.07; Fig. 2D) or the individual female cohort (R2 = 0.20; Fig. 2F) of the controls, and for the combined male and female cohort (R2 = 0.05; Fig. 2G) or the individual male cohort (R2 = 0.13; Fig. 2H) of the BPD patients; a trend line was observed for the SCZ (Fig. 2J-L). We also observed a significant difference between the slopes of the control and the SCZ regression lines relating S3 CpG site methylation to age in the female cohorts (P = 0.02), which displayed positive regression in the controls (Fig. 2F) but a negative trend line in the SCZ (Fig. 2L).
extent of methylation in the sample: Methylation = 100 × dCTP/(dCTP + dUTP)% where dCTP or dUTP represents the blank-corrected FP readings from the TAMRA-dCTP or -dUTP incorporations. 2.3. Statistical analysis A stepwise linear regression is employed to evaluate the relationship between the LINE-1 methylation levels and the clinical characteristics, including the illness duration and use of antipsychotic medications in the patients with SCZ and BPD, or to examine the relationship between the LINE-1 methylation level and the diagnosis, age and gender among the patients and controls. The differences in methylation levels between the patients with SCZ or BPD and the control group were analyzed with Student's t-test or ANCOVA with age as a covariate.
4. Discussion 3. Results Accumulating evidence supports the involvement of DNA methylation in the development of SCZ and BPD (Teroganova et al., 2016). The major findings of the present study demonstrated hypomethylation of LINE-1 in peripheral blood samples from patients with SCZ or BPD. Consistent hypomethylation of LINE-1 S3 CpG sites in both SCZ and BPD patients reflects a shared epigenetic variant between these two diseases, whereas hypomethylation of the S1 site in SCZ patients and the S2 site in BPD patients suggest the existence of disease-specific epigenetic variants in SCZ or BPD. The shared epigenetic variants between the two disorders observed in this study, together with shared symptoms and genetic variants (Cross-Disorder Group of the Psychiatric Genomics Consortium, 2013), further suggest there is a biological overlap between them and encourage investigations into shared pathophysiologies across these two disorders. LINE-1 methylation levels have been reported to be associated with the global DNA methylation levels, and the decreased LINE-1 methylation levels observed in the present study are also in agreement with previous observation of global hypomethylation in SCZ or BPD patients (Huzayyin et al., 2014; Melas et al., 2012; Shimabukuro et al., 2007). However, there are reports of increased or unchanged methylation levels in these mental illnesses (Ceylan et al., 2018; Jiang et al., 2017). This conflict may be due to differences in the methods used to detect the methylation profiles or the demographic and clinical variables of the patients. Recently, hypomethylation of LINE-1 was only observed in first-episode SCZ patients with a history of childhood trauma (Misiak et al., 2015), and LINE-1 copy number was increased in the patients with major depressive disorder possibly due to its hypomethylation (Liu et al., 2016). LINE-1 is the active autonomous retrotransposon in mammalian genomes, and hypomethylation of LINE-1 may be associated with its retrotransposon activity. This activity has a significant impact on genomic structure and
To examine the methylation levels of LINE-1, we employed bisulfite (BS)-PCR to amplify the BS-converted DNA of LINE-1 and to determine the DNA methylation levels of three CpG sites (S1, S2 and S3; Fig. 1) of the LINE-1 5‘UTR using a BS-OLE method (Li et al., 2015) from peripheral blood DNA of 92 controls and 92 patients with SCZ or 99 patients with BPD. The effects of the clinical characteristics on the LINE-1 methylation levels among the SCZ or BPD patients were first assessed with a linear regression to generate a model in which administration of Clozapine explained 4.5% of the variance in the S1 methylation level among the SCZ patients (Table 1). Then, the demographic characteristic effects among the combined patient and control cohort were assessed with a linear regression, which showed that 7.5% of the variance in the S2 methylation level was explained by the diagnostic status and 9.8% of the variance in the S3 methylation level was explained by the diagnostic status, age and gender. Therefore, age was employed as a covariate in the S3 methylation level analysis by ANCOVA, which was performed in a combined male and female cohort or separate male and female cohorts. Compared with the controls, the LINE-1 methylation levels were significantly decreased by 5% for the S1 CpG site (P < 0.001; Fig. 2A) and by 3% for the S3 site (P = 0.002; Fig. 2C) in the SCZ cohort and were decreased by 8% for the S2 site (P < 0.001; Fig. 2B) and by 5% for the S3 site (P = 0.002; Fig. 2C) in the BPD cohort. We also observed significant differences in the S1 and S2 methylation levels between the SCZ and BPD patients, with an increased S1 methylation level (by 5%, P < 0.001; Fig. 2A) and a decreased S2 methylation level (by 12%, P < 0.001; Fig. 2B) in the BPD patients compared to those of the SCZ. When the males were analyzed separately from the females with age as 70
Journal of Psychiatric Research 107 (2018) 68–72
S. Li et al.
Fig. 2. LINE-1 methylation levels in the controls and patients with schizophrenia or bipolar disorder. (A–C) LINE-1 S1 (A), S2 (B) and S3 (C) site methylation levels in the control (CON), schizophrenia (SCZ) and bipolar disorder (BPD) groups were determined by the BS-OLE method (Fig. S1). The horizontal bars represent the mean methylation levels. The methylation level of each subject was normalized to the mean level of the control cohort. The y-axis represents the percentage of methylation relative to the mean level of the control cohort. △ = controls; ▼ = patients with schizophrenia; ▲ = patients with bipolar disorder. Significant differences between two groups were determined using Student's t-test analysis (*P < 0.05, **P < 0.01 and ***P < 0.001). (D-L) Relationships between the LINE1 S3 methylation level and age in the combined or separate female and male subgroups of the CON (D-F), BPD (G-I) and SCZ (J-L) cohort are represented by solid lines (where the P value of the regression coefficient was < 0.05) or dashed lines (where P ≥ 0.05).
for the development of SCZ. The present study has some limitation that should be addressed. First, the present results were based on a relatively small sample size, and larger independent samples are required. Second, the majority of patients were not drug-free, and antipsychotic treatment might have impacts on the DNA methylation levels. Third, the BS-OLE method employed in the present study cannot distinguish hydroxymethylation from methylation. Finally, DNA methylation analysis was performed for only three CpG sites of LINE-1, which only account for a small portion of the CpG sites in LINE-1. In conclusion, the present study revealed hypomethylation of LINE1 in SCZ and BPD patients and a female-dependent increase in the LINE1 methylation level with age in the controls that lost its significance in patients with SCZ or BPD. The role of demethylation of LINE-1 in the
stability and is involved in pathological processes, development, cell differentiation and aging. Moreover, the female-dependent increase in the LINE-1 methylation level with age in the controls lost its significance in the patients with SCZ or BPD in the present study, suggesting an important role of LINE-1 methylation in the development of these diseases and sex-specific endocrine mechanism underlying the epigenetic alteration. In fact, LINE-1 retrotransposition has been observed during neurogenesis in the hippocampus (Muotri et al., 2009) as well as in neurons from the prefrontal cortex of patients with SCZ or in induced pluripotent stem cells from patients with 22q11 deletion syndrome (Bundo et al., 2014). The hypomethylation of LINE-1 observed in the present study together with the increase in LINE-1 polymorphic insertions associated with SCZ (Doyle et al., 2017; Guffanti et al., 2016) indicated an important role of LINE-1 insertions as a genomic risk factor 71
Journal of Psychiatric Research 107 (2018) 68–72
S. Li et al.
pathogenesis of SCZ and BPD remains to be elucidated.
analysis. Lancet 381, 1371–1379. https://doi.org/10.1016/S0140-6736(12)62129-1. Doyle, G.A., et al., 2017. Analysis of LINE-1 elements in DNA from postmortem brains of individuals with schizophrenia. Neuropsychopharmacology 42, 2602–2611. https:// doi.org/10.1038/npp.2017.115. Guffanti, G., et al., 2016. LINE1 insertions as a genomic risk factor for schizophrenia: preliminary evidence from an affected family. Am J Med Genet B Neuropsychiatr Genet 171, 534–545. https://doi.org/10.1002/ajmg.b.32437. He, Z.M., et al., 2014. Transition of LINE-1 DNA methylation status and altered expression in first and third trimester placentas. PloS One 9, e96994. https://doi.org/10.1371/ journal.pone.0096994. Hoffmann, M.J., Schulz, W.A., 2005. Causes and consequences of DNA hypomethylation in human cancer. Biochem. Cell Biol. 83, 296–321. https://doi.org/10.1139/o05036. Huzayyin, A.A., et al., 2014. Decreased global methylation in patients with bipolar disorder who respond to lithium. Int. J. Neuropsychopharmacol. 17, 561–569. https:// doi.org/10.1017/S1461145713001569. Jiang, T., et al., 2017. Variation in global DNA hydroxymethylation with age associated with schizophrenia. Psychiatr. Res. 257, 497–500. https://doi.org/10.1016/j. psychres.2017.08.022. Li, S., et al., 2015. Fluorescence polarization-based method with bisulfite conversionspecific one-label extension for quantification of single CpG dinucleotide methylation. Genome 58, 357–363. https://doi.org/10.1139/gen-2014-0185. Liu, S., et al., 2016. Inverse changes in L1 retrotransposons between blood and brain in major depressive disorder. Sci. Rep. 6, 37530. https://doi.org/10.1038/srep37530. Melas, P.A., et al., 2012. Epigenetic aberrations in leukocytes of patients with schizophrenia: association of global DNA methylation with antipsychotic drug treatment and disease onset. Faseb. J. 26, 2712–2718. https://doi.org/10.1096/fj.11-202069. Misiak, B., et al., 2015. Lower LINE-1 methylation in first-episode schizophrenia patients with the history of childhood trauma. Epigenomics 7, 1275–1285. https://doi.org/ 10.2217/epi.15.68. Muotri, A.R., et al., 2009. Environmental influence on L1 retrotransposons in the adult hippocampus. Hippocampus 19, 1002–1007. https://doi.org/10.1002/hipo.20564. Penzkofer, T., et al., 2017. L1Base 2: more retrotransposition-active LINE-1s, more mammalian genomes. Nucleic Acids Res. 45, D68–D73. https://doi.org/10.1093/ nar/gkw925. Shimabukuro, M., et al., 2007. Global hypomethylation of peripheral leukocyte DNA in male patients with schizophrenia: a potential link between epigenetics and schizophrenia. J. Psychiatr. Res. 41, 1042–1046. https://doi.org/10.1016/j.jpsychires. 2006.08.006. Steinhoff, C., Schulz, W.A., 2004. Transcriptional regulation of the human LINE-1 retrotransposon L1.2B. Mol. Genet. Genom. 270, 394–402. https://doi.org/10.1007/ s00438-003-0931-2. Teroganova, N., et al., 2016. DNA methylation in peripheral tissue of schizophrenia and bipolar disorder: a systematic review. BMC Genet. 17, 27. https://doi.org/10.1186/ s12863-016-0332-2. Tseng, P.T., et al., 2014. Age-associated decrease in global DNA methylation in patients with major depression. Neuropsychiatric Dis. Treat. 10, 2105–2114. https://doi.org/ 10.2147/NDT.S71997.
Role of the funding source This work was supported by the Guangdong-Hong Kong Technology Cooperation Funding Scheme of the Guangdong Science and Technology Foundation [grant number 2017A050506026], the National Natural Science Foundation of China [grant number 8167133, 81371475], the National Key Research and Development Program of China [grant number 2016YFC1201800] and the Program for Changjiang Scholars and Innovative Research Team in University [grant number IRT_16R37]. The funding had no further role in study design; in the collection, analysis and interpretation of data; in the writing of the report; or in the decision to submit the paper for publication. Conflicts of interest The authors declare no conflict of interest. Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.jpsychires.2018.10.009. References Auta, J., et al., 2013. DNA-methylation gene network dysregulation in peripheral blood lymphocytes of schizophrenia patients. Schizophr. Res. 150, 312–318. https://doi. org/10.1016/j.schres.2013.07.030. Babushok, D.V., Kazazian Jr., H.H., 2007. Progress in understanding the biology of the human mutagen LINE-1. Hum. Mutat. 28, 527–539. https://doi.org/10.1002/humu. 20486. Bundo, M., et al., 2014. Increased l1 retrotransposition in the neuronal genome in schizophrenia. Neuron 81, 306–313. https://doi.org/10.1016/j.neuron.2013.10.053. Ceylan, D., et al., 2018. DNA redox modulations and global DNA methylation in bipolar disorder: effects of sex, smoking and illness state. Psychiatr. Res. 261, 589–596. https://doi.org/10.1016/j.psychres.2017.12.051. Cross-disorder Group of the psychiatric genomics consortium, C., 2013. Identification of risk loci with shared effects on five major psychiatric disorders: a genome-wide
72