Childhood Adversity Increases Methylation in the GRIN2B Gene

Journal of Psychiatric Research 132 (2021) 38–43

Contents lists available at ScienceDirect

Journal of Psychiatric Research

journal homepage: www.elsevier.com/locate/jpsychires

Short communication Childhood adversity increases methylation in the GRIN2B gene

Elin Engdahl a,*,1, Ali Alavian-Ghavanini b,2, Yvonne Forsell c, Catharina Lavebratt d,e, Jo¨elle Rüegg f a Karolinska Institutet, Institute of Environmental Medicine (IMM), Unit of Integrative Toxicology, Stockholm, Sweden b Swetox, Karolinska Institutet, Unit of Toxicology Sciences, Sodert¨ alje,¨ Sweden c Karolinska Institutet, Department of Global Public Health, Stockholm, Sweden d Karolinska Institutet, Department of Molecular Medicine and Surgery, Stockholm, Sweden e Center for Molecular Medicine (CMM), Karolinska University Hospital, Stockholm, Sweden f Uppsala University, Department of Organismal Biology, Uppsala, Sweden

ARTICLE INFO ABSTRACT

Keywords: Childhood adversity is an early life stressor associated with increased risk of several psychiatric disorders such as GRIN2B depression. Epigenetic changes, primarily DNA methylation, can be affected by early life stress, which in turn DNA methylation might contribute to altered disease susceptibility later in life. One plausible biomarker of early life stress is Childhood adversity methylation of the ionotropic glutamate receptor NMDA type subunit 2B (GRIN2B) gene, which has been pre Early life stress viously shown to be epigenetically affected by prenatal environmental stressors. Here, we set out to investigate if Depression Epigenetics stress-inducing adversity during childhood is associated with changes in methylation of GRIN2B in adulthood. We studied 186 individuals from a Swedish naturalistic population-based cohort who had provided saliva samples (DNA) as well as information regarding both childhood adversity (CA) and depressive symptoms (dep) (nCA,dep = 41, nCA,no-dep = 56, nno-CA,dep = 40, Nno-CA,no-dep = 49). Methylation at four CpG sites in a regulatory region of GRIN2B was analysed using bisulfitepyrosequencing. Associations for methylation status to childhood adversity and to depression status were investigated using linear regression models. Our study shows that childhood adversity is associated with increased methylation levels of GRIN2B in adulthood, for three of the measured CpGs (p = 0.007, 0.006 and 5 × 10 14). This indicates that GRIN2B methylation is susceptible to early life stress, and that methylation at this gene is persistent over time. No association was found between GRIN2B methylation and depression status. Yet, this does not rule out a role for alterations in GRIN2B methylation for other neuropsychological outcomes not studied here.

1. Introduction Childhood adversity (CA), such as parental death, neglect, sexual, physical or emotional abuse, and friction within the family, is an early On top of the genetic information, environmental factors, especially life stressor that has been associated with an increased risk of psychiatric early in life, shape an organism’s phenotype and disease susceptibility disorders like depression (Collishaw et al., 2007; Widom et al., 2007; later in life (Gluckman and Hanson, 2004; Suzuki, 2018). During Liu, 2017), psychosis (Varese et al., 2012) and schizophrenia (Wells development, epigenetic processes play a crucial role in organising the et al., 2020), but also with cognitive impairments (Pechtel and Pizza functional genome (Champagne, 2013). Changes in epigenetic patterns, galli, 2011; Martins et al., 2019; Wells et al., 2020). On a molecular such as DNA methylation, induced during development can be stable level, CA can induce long lasting changes in DNA methylation patterns throughout life and are thought to be one link between early life expo (Provencal and Binder, 2015; Burns et al., 2018; Park et al., 2019). sures and health outcomes later in life (Marsit, 2015; Provencal and Primarily, methylation of genes involved in glucocorticoid signalling (e. Binder, 2015; Tran and Miyake, 2017; Burns et al., 2018; Goyal et al., g. NR3C1, FKBP5 (Tyrka et al., 2012; Klengel et al., 2013; Turecki and 2019; Park et al., 2019). Meaney, 2016)) and cortisol stress reactivity (e.g. KITLG (Houtepen

* Corresponding author. E-mail address: [email protected] (E. Engdahl). 1 Present address: Uppsala University, Department of Organismal Biology, Uppsala, Sweden. 2 Present address: Departments of Physiology, University of Toronto, Canada. https://doi.org/10.1016/j.jpsychires.2020.09.022 Received 31 January 2020; Received in revised form 14 September 2020; Accepted 25 September 2020 Available online 3 October 2020 0022-3956/© 2020 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). E. Engdahl et al. Journal of Psychiatric Research 132 (2021) 38–43 et al., 2016)) has been reported to be affected by CA. Yet, methylation of study participants (controls) did not show any pathological symptoms genes with other cellular functions such as cellular/neuronal plasticity of depression, anxiety, obsessive compulsive disorder, agoraphobia, has also been reported to be affected by CA (Labonte et al., 2012; Yang eating disorder or social disability due to a psychological problem. In et al., 2013; Checknita et al., 2018), indicating that several molecular addition, controls had never received health care for a psychiatric dis mechanisms may underlie associations between CA and mental and order or nervous discomfort. cognitive health outcomes later in life. Alcohol consumption was assessed using the Alcohol Use Disorders One gene important for mental and cognitive development is Identification Test (AUDIT) consisting of 10 questions on alcohol use GRIN2B, encoding the NR2B subunit of the N-methyl-D-aspartate re (Saunders et al., 1993), and the average score between wave 1 and 2 was ceptor (NMDAR). NMDARs are receptors for the excitatory neurotrans used. Smoking (never or stopped/sometimes/regularly), education mitter glutamate and are involved in synaptic plasticity (reviewed in (9y/12y/>12y), occupation (no/yes), and financial stability (ability to (Liu and Zhao, 2013; Baez et al., 2018)). Mutations in GRIN2B have been pay 14 000 Swedish Crowns within a week in case of an unexpected associated with neurodevelopmental and neuropsychiatric disorders situation: No/No, probably not/Yes, probably/Yes, certainly) such as intellectual disability, developmental dyslexia, autism spectrum (Hallstr¨ om¨ et al., 2003) were self-reported in wave 2. disorders and schizophrenia (Platzer and Lemke, 1993; Mascheretti In total, 3018 DNA samples were collected. Of these, all individuals et al., 2015; Guo et al., 2016; Hu et al., 2016; Myers et al., 2019). In with CA among the depression cases and the non-depressed controls addition, specific genetic variants in GRIN2B have also been associated were selected (ndep = 46, nnon-dep = 57). Additionally, participants with major depressive disorder (MDD), specifically treatment resistant without CA were included, trying to match for depression status, sex and depression (Zhang et al., 2014), and suicide attempts (Sokolowski et al., smoking (ndep = 46, nnon-dep = 56). Demographic characteristics for all 2013). In post mortem brain samples from females, but not males, higher selected participants is presented in Table S1. GRIN2B gene expression has been observed in MDD patients compared to controls, where GRIN2B expression was even higher in MDD-suicide 2.2. DNA methylation of GRIN2B patients (Gray et al., 2015). These reports indicate that GRIN2B may play a role in depression, possibly in a sex-dependent way. Methylation status of 4 CpG sites in the GRIN2B regulatory region We have previously shown that methylation in a GRIN2B regulatory (figure S1) were assessed by targeted bisulfite pyrosequencing. First, region is affected by chemical exposure during development (Alavian- genomic DNA obtained from saliva samples was bisulfite converted Ghavanini et al., 2018), and others have shown that GRIN2B methyl using the EZ-96 DNA Methylation-Gold MagPrep kit (Zymo Research ation is altered in schizophrenia patients (Fachim et al., 2019). In the Corporation). After that, a 119 base pair (bp) fragment was amplified present study we hypothesised that not only chemical exposure, but also using the PyroMark PCR kit (Qiagen) together with 0.2 μM forward ′ ′ the psychological stressor CA affects GRIN2B methylation in humans. In primer 5 -TGATTTAGGGGGGAGGAGAAATT-3 and biotinylated reverse ′ ′ addition, we set out to investigate if GRIN2B methylation is associated primer 5 -AAACTACCTCCCCCAAAATCTTAACA-3 and with the addi with depressive disorders, which are commonly stress-induced and tion of 1 μl MgCl2 (25 mM) per 25 μl reaction, according to manufac associated with impaired cognitive function (Airaksinen et al., 2004). turer’s protocol. Pyrosequencing was performed on a PyroMark Q96 ID instrument (Qiagen) using PyroMark Gold Q96 reagents (Qiagen) and ′ ′ 2. Material and method 0.4 μM sequencing primer 5 -GGAAGATATTGTTTTTGTTTTTAG-3 ac cording to the manufacturer’s protocol. Methylation level at the four 2.1. PART cohort sequenced CpG sites were obtained using the PyroMark Q96 software. The first CpG site in this fragment (hereafter referred to as CpG1) is The samples included in this work were derived from the PART study identical with the Illumina 450K probe ID cg10091102. The assay was (Lundberg et al., 2005; Liu et al., 2015), a longitudinal population based validated using standard curves of commercial DNA with known study in Stockholm, Sweden, with the aim to identify risk and protective methylation status. factors for mental health. The PART study was approved by the ethical Samples from 205 individuals (Table S1) were analysed for GRIN2B review board at Karolinska Institutet and performed according to the methylation. 18 samples did not pass pyrosequencing quality check latest version of the Helsinki Declaration. All study participants pro (“ok”) at all analysed CpG sites and were excluded from the analysis. In vided written informed consent after the study had been fully explained. addition, one individual was regarded as an outlier (CpG1 methylation Swedish nationals aged 20–64 years were randomly selected to be = 76%) and was thus excluded. invited to participate in the PART study. Information regarding de mographics, childhood conditions, occupation and financial status, so 2.3. Statistics matic disorders, use of drugs and psychiatric symptoms was collected through self-reported questionnaires two times (waves) for each indi Differences in characteristics at sampling between those with CA and vidual, first in 1998–2000 and then 3 years later (2001–2003). Saliva those without CA were assessed using either Mann-Whitney-Wilcoxon was collected 2006–2007, with a nested case-control design, of those Test (age, AUDIT score) or chi-squared tests (sex, depression status, participating in wave 1 and 2 using self-administered kits (Sjoholm smoking, education, occupation, financial stability). Correlation be et al., 2009; Melas et al., 2010), from which DNA was extracted using the tween variables were measured with Spearman rank correlation Oragene Purifier (DNA Genotek Inc., Canada) (Liu et al., 2015). (Table S2) and based on the results of these analyses, the variables that In our study, CA is defined as, before the age of 18, having at least correlated significantly with any CpG methylation level were added as two of the following adverse events: a parental death, major or severe covariates in the linear regression models to be used. Residuals of re financial problems, major or severe disturbances within the family. In ported linear regression models were confirmed to be normally dividuals without CA did not experience any of the above mentioned distributed. adversities. To assess if there were any associations between CA and GRIN2B Out of the self-reported questionnaires, depression status (including methylation, linear regression models adjusted for age and sex were major depression, mixed anxiety depression and dysthymia) was used. In addition, since the measured CpG sites were correlated assessed according to the DSM-IV criteria using the Major Depression (Table S2), a second model was used where all measured CpG sites Inventory (MDI (Bech et al., 2001; Forsell, 2005)), the Sheehan (except the dependent CpG variable in the specificmodel) were added as Patient-Related (Panic) Anxiety Scale (Sheehan, 1983) and a question on covariates with the aim to pick up the effect of CA on solely one specific duration of depressive symptoms. Depression cases were those with a CpG site eliminating the influence of the nearby region. Since smoking depression status in at least one of the two waves. The non-depressed habits, level of education, occupational status and financial stability

39 E. Engdahl et al. Journal of Psychiatric Research 132 (2021) 38–43 differed between CA and non-CA individuals, additional analyses were individuals who had experienced CA and those who had not (Fig. 1), performed where these variables were added as covariates to the where CA individuals displayed higher methylation levels. To statisti models. The analyses were also stratified for sex based on our previous cally investigate the association between CA and GRIN2B methylation observation that bisphenol A (BPA) only associates with GRIN2B DNA levels, linear regression models on GRIN2B CpG methylation were used. methylation in females (Alavian-Ghavanini et al., 2018). The models were adjusted for age and sex, which were identified as Association between GRIN2B methylation and age, sex, depression, potential covariates as they correlated with CpG methylation in alcohol use and smoking was also investigated using linear regression Spearman rank correlation tests (Table S2). CA before the age of 18 models adjusted for age, sex (when possible) and CA. years inferred higher methylation at 3 CpG sites in the GRIN2B gene All statistics were calculated using R version 3.5.1 and 4.0.2. Sta (CpG1 β = 0.95 p = 0.007; CpG2 β = 0.70 p = 0.006; CpG3 β = 2.12 p = tistical significance was set at α = 0.05 since methylation at the 4 CpG 5 × 10 14; Table 2). These results did not notably change when smoking, sites correlated with each other (p < 0.01). level of education, occupational status and financial stability, which differed between CA and non-CA, were added as covariates to the model 3. Results (Table S3). Since the methylation levels of the different CpG sites were correlated (Table S2), these CpG sites were added as covariates in a We set out to investigate if severe adversity during childhood was second regression model. In this model, CA was only significantly associated with methylation changes in GRIN2B, a gene involved in associated with CpG3 methylation (β = 1.7 p = 2 × 10 12, Table 2), neurodevelopment. A case-control approach was used where CA cases suggesting that CpG3 is the main driver for the association between were selected based on having experienced at least two of the following GRIN2B methylation and CA. CAs: major financialproblems, major conflictswithin the family and/or Although sex was not significantly associated with GRIN2B methyl parental death before the age of 18 years. At sampling, the study cohort ation in the linear regression models (p ≥ 0.2 for all CpG sites), the test had a mean age of 56 years and consisted of 65% women and 44% for association between CA and methylation level was stratifiedfor sex. depressed individuals (Table 1). Age, sex, depression status and alcohol Association between CA and methylation at all 4 CpG sites was more consumption did not differ significantly between the CA and no CA pronounced (higher beta values) in females compared to males. In fe group. On the other hand, smoking habits, level of education, occupa males, CpG1-3 methylation was significantlyassociated with CA, while tional status and financial stability differed between the two groups only CpG3 methylation was significantly associated with CA in males (Table 1), where smoking, as well as lower level of education, occupa (Table S4). tion and financial stability, was more common among individuals who In line with the Spearman rank correlation results (Table S2), had experienced CA. methylation at CpG3 and CpG4 increased significantly with age in GRIN2B CpG1-4 methylation level distributions differed between adjusted linear regression models (CpG3 β = 0.03, p = 0.008, CpG4 β = 0.02, p = 0.03, Figure S2). Also in line with the Spearman rank corre lation results, neither smoking nor alcohol intake was associated with Table 1 GRIN2B methylation levels in adjusted linear regression models. Demographics of the study groups. In addition to investigate what may influence GRIN2B methylation, f No childhood Childhood p we also investigated if GRIN2B methylation levels associated with adversity (n = 89) adversity (n = 97) depression. However, none of the analysed CpG sites were associated Characteristics at sampling with depression status in the study cohort (Fig. 2). Mean age (SD) 55.0 (11.9) 57.7 (10.6) 0.13 Female sex 66% (59/89) 64% (62/97) 0.85 Depressed 45% (40/89) 42% (41/97) 0.83 4. Discussion Smokersa 10% (9/87) 27% (26/95) < 0.01 Harmful alcohol 8% (7/88) 5% (5/96) 0.35 In the present study we report, for the firsttime, that methylation in a consumptionb GRIN2B regulatory region is higher in individuals who experienced CA. High school education 87% (77/89) 62% (60/96) < 0.01 c This was seen for 3 out of 4 CpG sites studied, in particular the CpG we Occupationd 88% (78/89) 70% (68/96) < 0.01 refer to as CpG3. These findingsconfirm that methylation at this region e Financial stability 95% (84/89) 78% (76/96) < 0.01 of GRIN2B is sensitive to environmental exposures early in life. While Events during childhood (<18 years) previous studies have shown an effect of physical environmental factors Major financial 0% (0/89) 91% (88/97) < 0.01 (chemical exposure and diet) in utero on Grin2b methylation in rodents problems (Yan et al., 2017; Alavian-Ghavanini et al., 2018) and a human cohort Major conflicts within 0% (0/89) 84% (81/97) < 0.01 (Alavian-Ghavanini et al., 2018), our results here suggest that GRIN2B the family Parental death 0% (0/89) 35% (34/97) < 0.01 methylation act as a more general environmental “sensor”, affected also Parents divorced 0% (0/89) 31% (30/97) < 0.01 by psychological stress in childhood. This finding can increase the un All individuals included in the study analyses had passed the quality control derstanding of how early life stress may influence neural development (QC) for all four investigated CpG sites. Demographics for all individuals, on a molecular level, possibly being one underlying link between CA and including also the 19 who did not pass the QC, are presented in Table S1. Data psychiatric and/or cognitive adversities later in life. was missing for smoking (n = 4), alcohol consumption (n = 2), education (n = The specificGRIN2B region analysed in the present study was chosen 1), occupation (n = 1) and financial stability (n = 1). based on our previous experimental findings showing that in rat hip a % individuals who reported regular cigarette smoking, compared to in pocampus, increased methylation in a well-conserved region of this dividuals who reported no, stopped or sometimes smoking. specificGRIN2B sequence was correlated with lower Grin2b expression b > AUDIT score 8 (Saunders et al., 1993). (Alavian-Ghavanini et al., 2018). This suggests that methylation in this c At least 12 years of education. region affects gene function, in the case of higher methylation resulting d Having an occupation includes full time and part time employment, self- in decreased GRIN2B expression. employment, studying and maternity/paternity leave. We could not find an association between GRIN2B methylation and e Financial stability is here defined to be able to, or probably able to, obtain 14 000 Swedish Crowns within a week in case of an unforeseen and sudden depression, and therefore the association between CA and depression situation. reported by us (Lavebratt et al., 2010; Melas et al., 2013) and others f p-values calculated using either Mann-Whitney-Wilcoxon Test (age, AUDIT (Collishaw et al., 2007; Widom et al., 2007; Liu, 2017) is likely to be score) or chi-squared tests (sex, depression status, smoking, education, occu mediated through other pathways. This negative finding may not be pation, financial status and childhood events). surprising since genetic variations in GRIN2B have been more often

40 E. Engdahl et al. Journal of Psychiatric Research 132 (2021) 38–43

Fig. 1. Distribution of DNA methylation at four CpG sites in a regulatory region of the GRIN2B gene. CA = Persons who experienced childhood adversity, No CA = Persons without reported childhood adversity. The dashed lines indicate the median values of the two groups (red = CA, blue = No CA). Reported p-values represent the associations between CA and CpG methylation levels, obtained from linear regression models on GRIN2B methylation adjusted for age and sex. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

4.1. Limitations Table 2 Association between childhood adversity (CA) and GRIN2B methylation. Firstly, childhood adversity was retrospectively self-reported, hence Model 1 Model 2 Model 3 recall as well as subjectivity might be a limitation in this study. Also, the Unadjusteda Age & sexb Age, sex & CpGs c PART study provides only data on selected CA events, not including beta (SE) p beta (SE) p beta (SE) p physical and sexual abuse. In addition, information regarding other exposures, like chemical exposures, is lacking in this study. Secondly, CpG1 0.96 6.0E- 0.95 7.4E- 0.14 0.72 (0.35) 03 (0.35) 03 (0.37) information about the exact age at which the adversity took place is CpG2 0.74 3.2E- 0.70 5.8E- 0.06 (0.27) 0.83 missing in our data. This may decrease sensitivity of the analyses as it (0.25) 03 (0.25) 03 has been shown that effects on DNA methylation are more pronounced if CpG3 2.20 8.3E- 2.12 4.6E- 1.71 (0.23) 2.1E- the adversity happens before the age of 3 compared to later in life (Dunn (0.26) 15 (0.26) 14 12 et al., 2019). Also, the age at sample collection range from 20 to 64 CpG4 0.35 0.13 0.28 0.22 0.48 0.05 (0.23) (0.23) (0.24) years, which introduces a variation in time interval between a CA event and sampling. Thirdly, our data is obtained in the surrogate tissue saliva, Results from three different linear regression models on GRIN2B CpG methyl while GRIN2B methylation in brain would be of foremost interest ation. Bold font indicates statistical significant (p < 0.05) association between considering its functions. Finally, information regarding age at depres CA and methylation level. Beta = unstandardized beta values, SE = std error of the unstandardized beta. N = 186. sion onset is lacking in our data, and the depression status was deter a Model 1: Unadjusted. mined based on validated self-reported questionnaires. b Model 2 Adjusted for age & sex. c Model 3: Adjusted for age, sex & the methylation level of the 3 other CpG 5. Conclusion sites. In conclusion, our study shows for the first time that childhood associated with cognitive impairments like intellectual disability as well adversity is associated with increased methylation levels at a regulatory as language and learning disorders (Hu et al., 2016; Myers et al., 2019). region of GRIN2B, a gene important for neurodevelopment. This con If the CA associated increase in GRIN2B methylation has a mediating firms the notion that the investigated region is sensitive to environ function in the association between CA and impaired cognitive func mental stressors (physical and psychological) in early life. No tioning reported by others (Gould et al., 2012; Petkus et al., 2018) will association was found between GRIN2B methylation at this region and be subject of future investigation. depression status, which suggests that this GRIN2B methylation does not affect susceptibility to depressive disorders.

41 E. Engdahl et al. Journal of Psychiatric Research 132 (2021) 38–43

Fig. 2. GRIN2B methylation levels in depressed and non-depressed individuals, with and without childhood adversity (CA). Each grey dot represents CpG methylation of one individual, and the overlying coloured box plots indicate median value and the interquartile range (IQR) plus whiskers up to the lowest/highest value < 1.5 x IQR. The association between CpG methylation and depression was investigated in linear regression models on depression status adjusted for age, sex and CA. ns = not significant (p > 0.05), No CA = Persons without reported childhood adversity, CA = Persons who experienced childhood adversity. (For inter pretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Author statement Grin2b in female rats and humans. Sci. Rep. 8 (1), 11315. https://doi.org/10.1038/ s41598-018-29732-9. Baez, M.V., Cercato, M.C., Jerusalinsky, D.A., 2018. NMDA receptor subunits change EE: Data curation, Formal analysis, Funding acquisition, Writing - after synaptic plasticity induction and learning and memory acquisition. Neural original draft. AAG: Investigation, Writing - review & editing. YF: Plast. 2018, 5093048. https://doi.org/10.1155/2018/5093048. Funding acquisition, Resources, Writing - review & editing. CL: Bech, P., Rasmussen, N.A., Olsen, L.R., Noerholm, V., Abildgaard, W., 2001. The sensitivity and specificityof the Major Depression Inventory, using the Present State Conceptualization, Funding acquisition, Project administration (PART), Examination as the index of diagnostic validity. J. Affect. Disord. 66 (2–3), 159–164. Methodology, Resources, Writing - review & editing. JR: Conceptuali https://doi.org/10.1016/s0165-0327(00)00309-8. zation, Resources, Writing - review & editing. Burns, S.B., Szyszkowicz, J.K., Luheshi, G.N., Lutz, P.E., Turecki, G., 2018. Plasticity of the epigenome during early-life stress. Semin. Cell Dev. Biol. 77, 115–132. https:// doi.org/10.1016/j.semcdb.2017.09.033. Funding statement Champagne, F.A., 2013. Epigenetics and developmental plasticity across species. Dev. Psychobiol. 55 (1), 33–41. https://doi.org/10.1002/dev.21036. Checknita, D., Ekstrom, T.J., Comasco, E., Nilsson, K.W., Tiihonen, J., Hodgins, S., 2018. This work was supported by the Swedish Research Council [www.vr. Associations of monoamine oxidase A gene firstexon methylation with sexual abuse se; grant numbers 2009-5546, 2010-3631, 2014-10171], the Karolinska and current depression in women. J. Neural. Transm. 125 (7), 1053–1064. https:// Institutet’s Faculty Funds and Research Foundation Grants [https://fon doi.org/10.1007/s00702-018-1875-3. Collishaw, S., Pickles, A., Messer, J., Rutter, M., Shearer, C., Maughan, B., 2007. der.ki.se; grant number 2018-01508], the regional agreement on med Resilience to adult psychopathology following childhood maltreatment: evidence ical training and clinical research between Stockholm County Council from a community sample. Child Abuse Negl. 31 (3), 211–229. https://doi.org/ and Karolinska Institutet [www.forskningsstod.sll.se; grant numbers 10.1016/j.chiabu.2007.02.004. SLL20170292, SLL20140484, SLL20110560, 20090281], the Swedish Dunn, E.C., Soare, T.W., Zhu, Y., Simpkin, A.J., Suderman, M.J., Klengel, T., Smith, A., Ressler, K.J., Relton, C.L., 2019. Sensitive periods for the effect of childhood Brain Foundation [grant numbers FO2017-0129, FO2018-0141]. The adversity on DNA methylation: results from a prospective, longitudinal study. Biol. funders had no role in study design, data collection and analysis, deci Psychiatr. 85 (10), 838–849. https://doi.org/10.1016/j.biopsych.2018.12.023. sion to publish, or preparation of the manuscript. Fachim, H.A., Loureiro, C.M., Corsi-Zuelli, F., Shuhama, R., Louzada-Junior, P., Menezes, P.R., Dalton, C.F., Del-Ben, C.M., Reynolds, G.P., 2019. GRIN2B promoter methylation deficits in early-onset schizophrenia and its association with cognitive Declaration of competing interest function. Epigenomics 11 (4), 401–410. https://doi.org/10.2217/epi-2018-0127. Forsell, Y., 2005. The major depression inventory versus schedules for clinical assessment in neuropsychiatry in a population sample. Soc. Psychiatr. Psychiatr. The authors declare no conflict of interest. Epidemiol. 40 (3), 209–213. https://doi.org/10.1007/s00127-005-0876-3. Gluckman, P.D., Hanson, M.A., 2004. Living with the past: evolution, development, and – Acknowledgement patterns of disease. Science 305 (5691), 1733 1736. https://doi.org/10.1126/ science.1095292. Gould, F., Clarke, J., Heim, C., Harvey, P.D., Majer, M., Nemeroff, C.B., 2012. The effects We want to thank Mandy Tang for technical assistance on the of child abuse and neglect on cognitive functioning in adulthood. J. Psychiatr. Res. – bisulfite pyrosequencing. 46 (4), 500 506. https://doi.org/10.1016/j.jpsychires.2012.01.005. Goyal, D., Limesand, S.W., Goyal, R., 2019. Epigenetic responses and the developmental origins of health and disease. J. Endocrinol. 242 (1), T105–T119. https://doi.org/ Appendix A. Supplementary data 10.1530/JOE-19-0009. Gray, A.L., Hyde, T.M., Deep-Soboslay, A., Kleinman, J.E., Sodhi, M.S., 2015. Sex differences in glutamate receptor gene expression in major depression and suicide. Supplementary data to this article can be found online at https://doi. Mol. Psychiatr. 20 (9), 1139. https://doi.org/10.1038/mp.2015.114. org/10.1016/j.jpsychires.2020.09.022. Guo, Z., Niu, W., Bi, Y., Zhang, R., Ren, D., Hu, J., Huang, X., Wu, X., Cao, Y., Yang, F., Wang, L., Li, W., Li, X., Xu, Y., He, L., Yu, T., He, G., 2016. A study of single nucleotide polymorphisms of GRIN2B in schizophrenia from Chinese Han References population. Neurosci. Lett. 630, 132–135. https://doi.org/10.1016/j. neulet.2016.07.038. Airaksinen, E., Larsson, M., Lundberg, I., Forsell, Y., 2004. Cognitive functions in Houtepen, L.C., Vinkers, C.H., Carrillo-Roa, T., Hiemstra, M., van Lier, P.A., Meeus, W., depressive disorders: evidence from a population-based study. Psychol. Med. 34 (1), Branje, S., Heim, C.M., Nemeroff, C.B., Mill, J., Schalkwyk, L.C., Creyghton, M.P., 83–91. https://doi.org/10.1017/s0033291703008559. Kahn, R.S., Joels, M., Binder, E.B., Boks, M.P., 2016. Genome-wide DNA methylation Alavian-Ghavanini, A., Lin, P.I., Lind, P.M., Risen Rimfors, S., Halin Lejonklou, M., levels and altered cortisol stress reactivity following childhood trauma in humans. Dunder, L., Tang, M., Lindh, C., Bornehag, C.G., Ruegg, J., 2018. Prenatal bisphenol Nat. Commun. 7, 10967. https://doi.org/10.1038/ncomms10967. A exposure is linked to epigenetic changes in glutamate receptor subunit gene

42 E. Engdahl et al. Journal of Psychiatric Research 132 (2021) 38–43

Hu, C., Chen, W., Myers, S.J., Yuan, H., Traynelis, S.F., 2016. Human GRIN2B variants in Pechtel, P., Pizzagalli, D.A., 2011. Effects of early life stress on cognitive and affective neurodevelopmental disorders. J. Pharmacol. Sci. 132 (2), 115–121. https://doi.org/ function: an integrated review of human literature. Psychopharmacology (Berl) 214 10.1016/j.jphs.2016.10.002. (1), 55–70. https://doi.org/10.1007/s00213-010-2009-2. Hallstr¨ om,¨ T., Damstrom¨ Thakker, K., Forsell, Y., Lundberg, I., Tinghog,¨ P., 2003. The Petkus, A.J., Lenze, E.J., Butters, M.A., Twamley, E.W., Wetherell, J.L., 2018. Childhood PART Study : A Population Based Study of Mental Health in the Stockholm County: trauma is associated with poorer cognitive performance in older adults. J. Clin. Study Design. Phase l (1998-2000). Psychiatr. 79 (1) https://doi.org/10.4088/JCP.16m11021. Klengel, T., Mehta, D., Anacker, C., Rex-Haffner, M., Pruessner, J.C., Pariante, C.M., Platzer, K., Lemke, J.R., 1993. GRIN2B-Related Neurodevelopmental Disorder. Pace, T.W., Mercer, K.B., Mayberg, H.S., Bradley, B., Nemeroff, C.B., Holsboer, F., GeneReviews((R)). M.P. Adam, H.H. Ardinger, R.A. Pagon et al. Seattle (WA). Heim, C.M., Ressler, K.J., Rein, T., Binder, E.B., 2013. Allele-specific FKBP5 DNA Provencal, N., Binder, E.B., 2015. The effects of early life stress on the epigenome: from demethylation mediates gene-childhood trauma interactions. Nat. Neurosci. 16 (1), the womb to adulthood and even before. Exp. Neurol. 268, 10–20. https://doi.org/ 33–41. https://doi.org/10.1038/nn.3275. 10.1016/j.expneurol.2014.09.001. Labonte, B., Suderman, M., Maussion, G., Navaro, L., Yerko, V., Mahar, I., Bureau, A., Saunders, J.B., Aasland, O.G., Babor, T.F., de la Fuente, J.R., Grant, M., 1993. Mechawar, N., Szyf, M., Meaney, M.J., Turecki, G., 2012. Genome-wide epigenetic Development of the alcohol use disorders identification test (AUDIT): WHO regulation by early-life trauma. Arch. Gen. Psychiatr. 69 (7), 722–731. https://doi. collaborative Project on early detection of Persons with harmful alcohol org/10.1001/archgenpsychiatry.2011.2287. consumption–II. Addiction 88 (6), 791–804. https://doi.org/10.1111/j.1360- Lavebratt, C., Aberg, E., Sjoholm, L.K., Forsell, Y., 2010. Variations in FKBP5 and BDNF 0443.1993.tb02093.x. genes are suggestively associated with depression in a Swedish population-based Sheehan, D.V., 1983. The Anxiety Disease. Charles Scribner’s Sons, New York. cohort. J. Affect. Disord. 125 (1–3), 249–255. https://doi.org/10.1016/j. Sjoholm, L.K., Melas, P.A., Forsell, Y., Lavebratt, C., 2009. PreproNPY Pro 7 protects jad.2010.02.113. against depression despite exposure to environmental risk factors. J. Affect. Disord. Liu, J.J., Lou, F., Lavebratt, C., Forsell, Y., 2015. Impact of childhood adversity and 118 (1–3), 124–130. https://doi.org/10.1016/j.jad.2009.02.009. vasopressin receptor 1a variation on social interaction in adulthood: a cross- Sokolowski, M., Ben-Efraim, Y.J., Wasserman, J., Wasserman, D., 2013. Glutamatergic sectional study. PLoS One 10 (8), e0136436. https://doi.org/10.1371/journal. GRIN2B and polyaminergic ODC1 genes in suicide attempts: associations and gene- pone.0136436. environment interactions with childhood/adolescent physical assault. Mol. Liu, R.T., 2017. Childhood adversities and depression in adulthood: current findingsand Psychiatr. 18 (9), 985–992. https://doi.org/10.1038/mp.2012.112. future directions. Clin. Psychol. 24 (2), 140–153. https://doi.org/10.1111/ Suzuki, K., 2018. The developing world of DOHaD. J Dev Orig Health Dis 9 (3), 266–269. cpsp.12190. https://doi.org/10.1017/S2040174417000691. Liu, S.B., Zhao, M.G., 2013. Neuroprotective effect of estrogen: role of nonsynaptic Tran, N.Q.V., Miyake, K., 2017. Neurodevelopmental disorders and environmental NR2B-containing NMDA receptors. Brain Res. Bull. 93, 27–31. https://doi.org/ toxicants: epigenetics as an underlying mechanism. Int J Genomics 2017, 7526592. 10.1016/j.brainresbull.2012.10.004. https://doi.org/10.1155/2017/7526592. Lundberg, I., Damstrom Thakker, K., Hallstrom, T., Forsell, Y., 2005. Determinants of Turecki, G., Meaney, M.J., 2016. Effects of the social environment and stress on non-participation, and the effects of non-participation on potential cause-effect glucocorticoid receptor gene methylation: a systematic review. Biol. Psychiatr. 79 relationships, in the PART study on mental disorders. Soc. Psychiatr. Psychiatr. (2), 87–96. https://doi.org/10.1016/j.biopsych.2014.11.022. Epidemiol. 40 (6), 475–483. https://doi.org/10.1007/s00127-005-0911-4. Tyrka, A.R., Price, L.H., Marsit, C., Walters, O.C., Carpenter, L.L., 2012. Childhood Marsit, C.J., 2015. Influenceof environmental exposure on human epigenetic regulation. adversity and epigenetic modulation of the leukocyte glucocorticoid receptor: J. Exp. Biol. 218 (Pt 1), 71–79. https://doi.org/10.1242/jeb.106971. preliminary findings in healthy adults. PLoS One 7 (1), e30148. https://doi.org/ Martins, D.S., Hasse-Sousa, M., Petry-Perin, C., Arrial-Cordeiro, R.T., Rabelo-da-Ponte, F. 10.1371/journal.pone.0030148. D., Lima, F.M., Rosa, A.R., Bucker, J., Gama, C.S., Czepielewski, L.S., 2019. Varese, F., Smeets, F., Drukker, M., Lieverse, R., Lataster, T., Viechtbauer, W., Read, J., Perceived childhood adversities: impact of childhood trauma to estimated van Os, J., Bentall, R.P., 2012. Childhood adversities increase the risk of psychosis: a intellectual functioning of individuals with bipolar disorder. Psychiatr. Res. 274, meta-analysis of patient-control, prospective- and cross-sectional cohort studies. 345–351. https://doi.org/10.1016/j.psychres.2019.02.046. Schizophr. Bull. 38 (4), 661–671. https://doi.org/10.1093/schbul/sbs050. Mascheretti, S., Facoetti, A., Giorda, R., Beri, S., Riva, V., Trezzi, V., Cellino, M.R., Wells, R., Jacomb, I., Swaminathan, V., Sundram, S., Weinberg, D., Bruggemann, J., Marino, C., 2015. GRIN2B mediates susceptibility to intelligence quotient and Cropley, V., Lenroot, R.K., Pereira, A.M., Zalesky, A., Bousman, C., Pantelis, C., cognitive impairments in developmental dyslexia. Psychiatr. Genet. 25 (1), 9–20. Weickert, C.S., Weickert, T.W., 2020. The impact of childhood adversity on cognitive https://doi.org/10.1097/YPG.0000000000000068. development in schizophrenia. Schizophr. Bull. 46 (1), 140–153. https://doi.org/ Melas, P.A., Sjoholm, L.K., Forsner, T., Edhborg, M., Juth, N., Forsell, Y., Lavebratt, C., 10.1093/schbul/sbz033. 2010. Examining the public refusal to consent to DNA biobanking: empirical data Widom, C.S., DuMont, K., Czaja, S.J., 2007. A prospective investigation of major from a Swedish population-based study. J. Med. Ethics 36 (2), 93–98. https://doi. depressive disorder and comorbidity in abused and neglected children grown up. org/10.1136/jme.2009.032367. Arch. Gen. Psychiatr. 64 (1), 49–56. https://doi.org/10.1001/archpsyc.64.1.49. Melas, P.A., Wei, Y., Wong, C.C., Sjoholm, L.K., Aberg, E., Mill, J., Schalling, M., Yan, Z., Jiao, F., Yan, X., Ou, H., 2017. Maternal chronic folate supplementation Forsell, Y., Lavebratt, C., 2013. Genetic and epigenetic associations of MAOA and ameliorates behavior disorders induced by prenatal high-fat diet through NR3C1 with depression and childhood adversities. Int. J. Neuropsychopharmacol. 16 methylation alteration of BDNF and Grin2b in offspring Hippocampus. Mol. Nutr. (7), 1513–1528. https://doi.org/10.1017/S1461145713000102. Food Res. 61 (12) https://doi.org/10.1002/mnfr.201700461. Myers, S.J., Yuan, H., Kang, J.Q., Tan, F.C.K., Traynelis, S.F., Low, C.M., 2019. Distinct Yang, B.Z., Zhang, H., Ge, W., Weder, N., Douglas-Palumberi, H., Perepletchikova, F., roles of GRIN2A and GRIN2B variants in neurological conditions. F1000Res 8. Gelernter, J., Kaufman, J., 2013. Child abuse and epigenetic mechanisms of disease https://doi.org/10.12688/f1000research.18949.1. risk. Am. J. Prev. Med. 44 (2), 101–107. https://doi.org/10.1016/j. Park, C., Rosenblat, J.D., Brietzke, E., Pan, Z., Lee, Y., Cao, B., Zuckerman, H., amepre.2012.10.012. Kalantarova, A., McIntyre, R.S., 2019. Stress, epigenetics and depression: a Zhang, C., Li, Z., Wu, Z., Chen, J., Wang, Z., Peng, D., Hong, W., Yuan, C., Wang, Z., systematic review. Neurosci. Biobehav. Rev. 102, 139–152. https://doi.org/ Yu, S., Xu, Y., Xu, L., Xiao, Z., Fang, Y., 2014. A study of N-methyl-D-aspartate 10.1016/j.neubiorev.2019.04.010. receptor gene (GRIN2B) variants as predictors of treatment-resistant major depression. Psychopharmacology (Berl) 231 (4), 685–693. https://doi.org/10.1007/ s00213-013-3297-0.