Table 1 A summary of selected articles in this systematic review
From: What do DNA methylation studies tell us about depression? A systematic review
ID | First author | Publication year | Country | Sample size | Sample characteristics | Study design | Diagnoses of depression | Biological samples | DNA methylation methods/kits | Targeted genetic locations | Markers found in genome-wide studies/ CpG sites for candidate-gene studies | Major findings |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
1 | Bostrom et al. | 2017 | Sweden | 223 | Population-based adolescent cohort | Case–control | Depression in general | Whole blood | Illumina 450k | Genome-wide | The promoter region of miRNA4646 and TSS of ZSWIM5 | Two CpG sites (cg13227623 and cg04102384) predicted depression in adolescents. cg04102384 was hypomethylated |
2 | Roy et al. | 2017 | USA | 34 | Hospital-based cohort | Case–control | MDD | Peripheral blood mononuclear cells | Immunoprecipitate the 5-methyl cytosine-enriched and qPCR | BDNF, FKBP5, CRHBP, CRHR1, NR3C1 | Promoters, CpG islands | BDNF, FKBP5, CRHBP, CRHR1, NR3C1 gene promoters were significantly hypermethylated in MDD |
3 | Meng et al. | 2017 | China | 162 | Hospital-based cohort | Case–control | MDD | White blood cells | Bisulfite conversion, pyrosequencing | NET or SLC6A2 | Promoters, other CpG sites | There were no significant differences in DNA methylation of the NET gene promoter between healthy controls and patients with MDD |
4 | Kaut et al. | 2017 | Germany | 12 | Senior cases and controls | Case–control | MDD | Brain tissue | Bisulfite conversion, pyrosequencing | PSD-95 and GLA-1 | Promoters, CpG islands | There were no significant differences in DNA methylation of PSD-95 and GJA-1 between controls and cases |
5 | Ryan et al. | 2017 | Australia | 380 | Late-life MDD and controls | Case–control | MDD | Buccal cells | Bisulfite conversion, pyrosequencing | IL-6 and treatment responses | Promoters, CpG islands | Individuals with depression (current MDD or high depressive symptoms) had lower IL6 methylation levels at one of the four sites investigated. Antidepressant use was independently associated with higher IL-6 methylation at the same site |
6 | Shi et al. | 2017 | China | 161 | Hospital-based cohort | Case–control | MDD | Whole blood | Bisulfite conversion, pyrosequencing | 5-HTT or SLC6A4 | Promoters, CpG islands | Methylation (hypo- and hyper-) at positions 4 and 5 was significantly associated with MDD |
7 | Han et al. | 2017 | South Korea | 145 | Hospital-based cohort | Case–control | MDD | Whole blood | Bisulfite conversion, pyrosequencing | TESC | Gene body, other CpG sites | MDD had significantly higher methylation on CpG2 position of TESC gene-regulating genetic variant (rs7294919) than controls |
8 | Takeuchi et al. | 2017 | Japan | 20 | Cases with best and worst treatment responses | Case–control | MDD | Whole blood | Bisulfite conversion, pyrosequencing | Genome-wide | PPF1A4, HS3ST1 | Patients’ DNA-methylation profile at specific genes such as PPF1A4 and HS3ST1 was associated with individual variations in therapeutic responses |
9 | Crueanu et al. | 2016 | Canada | 32 | White Caucasians, cases and controls | Case–control | MDD | PFC brain tissue | Bisulfite conversion, quantified with EpiTYPER | SYN2 | Promoters and gene body, CpG islands | Hypomethylation of synapsins (SYN2) was linked to depression |
10 | Won et al. | 2016 | South Korea | 74 | Antidepressant-free cases and controls | Case–control | MDD | Whole blood | Bisulfite conversion, pyrosequencing | SLC6A4 | Promoter region, other CpG sites | Significant inverse correlations were observed between SLC6A4 DNA methylation and fractional anisotropy. SLC6A4 DNA methylation was significantly higher at CpG2 in MDD |
11 | Walker et al. | 2016 | Scotland | 29 | Members of a large family multiply affected by BD and MDD | Case–control | MDD | Whole blood | Sodium bisulphite using the EZ-96 DNA Methylation Kit, bead array using the Infinium HumanMethylation450 BeadChip | Genome-wide | Three DMR regions (promoter region of HOXA5 for hypomethylation, 5’ end of RNF39 for hypermethylation, and promoter and first exon of AGPAT1 and RNF5 for hypo-methylation) | Nominally significant differences in DNA methylation were observed; altered DNA methylation was a potential mechanism for mood disorders |
12 | Osborne et al. | 2016 | USA | 291 | Derived from two prospective cohorts designed to study PPD and two cohorts from which DNA was taken long after pregnancy | Case–control | PPD | Whole blood | Illumina Human Methylation 450 (HM450) bead array for 51 women with mood disorders (existing data); bisulfite conversion pyrosequencing using PyroMark MD system for the rest of the samples | Genome-wide | No site identified | Epigenetic variation at PPD biomarker loci was likely to be associated with expression |
13 | Bustamante et al. | 2016 | USA | 147 | Lifetime MDD and controls | Case–control | MDD | Whole blood | Bisulfite conversion using EpiTect Bisulfite Kit, pyrosequencing using PyroMark Q24 Assay Design Software | NR3C1 | Promoters, CpG islands | DNA methylation was significantly lower over CpG sites 5–13 in those with vs without MDD |
14 | Na et al. | 2016 | South Korea | 117 | Recurrent MDD and controls | Case–control | MDD | Whole blood | Bisulfite conversion, pyrosequencing, using PyroMark ID system with the Pyro Gold reagents kit (Qiagen, Valencia, CA, USA) | BDNF and treatment response | Promoters, CpG islands | Patients with MDD had significantly higher rates of methylation at CpG2 and CpG4 than healthy controls. No difference was found in naive or on-medication patients |
15 | Kimmel et al. | 2016 | USA | 352 | Caucasian women | Cohort | PPD | Whole blood | Bisulfite conversion by EZ DNA Methylation-Gold Kit and pyrosequencing using PyroMark MD system | OXTR | 5’-UTR, CpG islands | CpG (cg12695586) positioned in the middle of SP1 transcription factor binding site. Its methylation had a negative correlation with PPD |
16 | Kahl et al. | 2016 | Germany | 70 | Treated MDD in patients and university announcements for controls | Case–control | MDD | Whole blood | Bisulfite conversion, PCR and sequencing. Sodium-bisulfite conversion using the EpiTect® Bisulfite Kit | GLU1, GLU4 | Promoters, CpG islands | Increased methylation of GLUT1 in MDD |
17 | Iga et al. | 2016 | Japan | 57 | Unmediated cases and controls | Case–control | MDD | Leukocytes | Bisulfite conversion, pyrosequencing, EpiTect Plus DNA Bisulfite Kit (Qiagen) | SLC6A4 | Promoters, CpG islands | Mean methylation level was significantly increased in patients compared with controls, p = 0.04. No significant difference was found in single CpG site |
18 | Oh et al. | 2015 | Peripheral blood samples from Australia, The Netherlands, and UK; prefrontal cortex and sperm samples from Canada | 260 | Cases and matched controls | Case–control | MDD | Peripheral blood, prefrontal cortex, and sperm | Bisulfite conversion, pyrosequencing using Gold Q96 reagents, and Pyromark Q24 | Genome-wide | No site identified | Hypermethylated loci were found in the white blood cells of MDD twins. The brain and the sperm showed higher proportions of hypomethylated regions in MDD patients compared with the controls |
19 | Nagy et al. | 2015 | Canada | 121 | Cases with MDD and died from suicide, and controls, not died from suicide and with no MDD | Case–control | MDD | Brain tissue | Bisulfite conversion using EpiTect Bisulfite kit from Qiagen, PCR, and sequencing | Genome-wide | 115 DMRs | Significant differences (decrease) in the methylation patterns specific to astrocytic dysfunction associated with depressive psychopathology |
20 | van der Knapp et al. | 2015 | The Netherlands | 954 | Adolescents cohort | Case–control | Depression in general | Whole blood | Methylation levels analyzed using EpiTYPER method; bisulfite conversion using EZ-96 DNA Methylation Kit, followed by PCR | NR3C1 and SLC6A4 | Promoters, CpG islands | NR3C1 methylation levels at NR3C1_1 were positively associated with the risk of a depressive disorder and were positively associated with depressive symptom scores at follow-up, but became non-significant when accounted for depressive symptom scores at the baseline |
21 | Melas et al. | 2015 | Sweden | 44 | Female cases and controls | Case–control | Depression in general | Saliva | Bisulfite conversion using EZ-96 DNA Methylation-Gold Kit, PCR, and sequencing | MAOA | Gene body, other CpG sites | Subjects with a history of depression were hypomethylated, compared to controls. Female individuals were hypermethylated at the MAOA region compared to males |
22 | Hohne et al. | 2015 | Germany | 116 | Remitted MDD and healthy controls | Case–control | MDD | Peripheral blood cells | Bisulfite conversion, PCR, and sequencing using EpiTYPER assay | FKBP5 | Gene body, other CpG sites | Subjects with TT genotype and a lifetime history of MD had a 10% higher DNA methylation rate than healthy controls with the same FKBP5 genotype |
23 | Choi et al. | 2015 | South Korea | 113 | MDD with a mixed history of treatment | Case–control | MDD | Whole blood | Bisulfite conversion, pyrosequencing was performed on a PyroMark ID system using the Pyro Gold reagent kit (Qiagen) | BDNF | Promoters, other CpG sites | There were no significant differences in the BDNF DNA methylation status at CpG1, CpG2, CpG3, and CpG4 between patients with MDD and healthy controls |
24 | Domschke et al. | 2015 | Germany | 94 | Caucasian case cohort with antidepressants | Cohort | MDD | Whole blood | Sodium bisulfite converted using EZ-96 DNA methylation kit, PCR, and sequencing using BigDye Terminator | MAOA and treatment response | Promoters and gene body, not mentioned for CpG sites | The study did not find a major influence of MAOA DNA methylation on antidepressant treatment response. However, the presently observed trend towards CpG-specific MAOA gene hypomethylation might potentially drive impaired antidepressant treatment response in females—larger pharmacoepigenetic studies are needed |
25 | Córdova-Palomera et al. | 2015 | Spain | 34 | Caucasian MZ twins | Twin study | Depression in general | Whole blood | Bisulfite conversion, bead array using The Illumina Infinium HumanMethylation450 (450K) BeadChip | DEPDC7 | Gene body, other CpG sites | A hypomethylation of cg09090376 in a co-twin would be associated with an increase in his/her depressive symptom score |
26 | Reiner et al. | 2015 | Germany | 85 | Female inpatients and controls | Case–control | Depression and/or dysthymia | Leukocytes | Bisulfite conversion using EpiTect Bisulfite Kit, PCR, and sequencing using BigDye Terminator v3.1 Cycle Sequencing Kit | OXTR | Gene body, other CpG sites | Depressed female patients had decreased OXTR exon 1 DNA methylation compared to non-depressed women. Exon 1 methylation appears to be associated with depressive phenotypes, whereas exon 2 methylation was influenced by genotype rs53576 |
27 | Haghighi et al. | 2015 | USA | 120 | Age- and sex-matched cases and controls | Case–control | MDD | Buffy coat of blood | Bisulfite conversion by EpiTect Bisulfite Kit, pyrosequencing using PyroMark Q96 MD | FADS1, FADS2, and ELOVL5 | 5’-UTR, CpG islands, and shores | MDD patients had a lower methylation in FADS2, but higher in ELOVL5 |
28 | Chagnon et al. | 2015 | Canada | 43 | Women aged 65 years and plus | Case–control | Depression (major and minor) and/or anxiety | Saliva | Bisulfite conversion, pyrosequencing using Pyromark 96, except for APOE analyzed on Illumina Beadchip | BDNF, OXTR, SLC6A4, and APOE | Gene body, other CpG sites | A higher BDNF and OXTR DNA methylation was observed in subjects with anxiety/depression compared to controls |
29 | Córdova-Palomera et al. | 2015 | Spain | 34 | Twin pairs with MDD and healthy controls | Case–control twin study | MDD | Whole blood | Bisulfite conversion using Illumina Infinium HumanMethylation450 Beadchip | Genome-wide | cg01122889 (WDR26) | Hypomethylation in WDR26 gene was associated with a lifetime diagnosis of depression |
30 | Bell et al. | 2015 | USA | 545 | Nested case–control study in a longitudinal cohort | Nested case–control | PPD | Whole blood | Bisulfite conversion, pyrosequencing using PyroMark Gold Q24 | OXTR | Gene body, other CpG sites | Methylation was not significantly associated with postpartum depression |
31 | Zhang et al. | 2015 | China | 125 | MDD only, with or without suicide attempts | Case–control | MDD | Whole blood | Bisulfite conversion, methylation-specific PCR | TPH2 | Promoters, other CpG sites | The TPH2 promoter was methylated in 36.0% (18/50) of MDD + suicide patients, as compared with that in 13.0% (10/75) of MDD patients |
32 | Nantharat et al. | 2015 | Thailand | 62 | Untreated MDD and controls | Case–control | MDD | Whole blood | Bisulfite pyrosequencing. PyroMark LINE-1 kit (Biotage-Qiagen, Uppsala, Sweden) | NR3C1 | Promoters, CpG islands | Hypermethylation levels at CpG7 were found in MDD in females but not in males |
33 | Kleimann et al. | 2015 | Germany | 11 | Treatment-resistant cases | Perspective cohort | MDD | Whole blood | Bisulfite conversion using EpiTect Bisulfite Kit, PCR, and sequencing using BigDye Terminator Cycle Sequencing Kit | Treatment responses on BDNF | Promoters, CpG islands | Remitters had a significantly lower mean promoter methylation rate than non-remitters, especially exon I |
34 | Kim et al. | 2015 | South Korea | 969 | Patients with recent acute coronary syndrome | Longitudinal | Mix of major and minor depression | Leukocytes | Bisulfite conversion using EpiTech Bisulfite Kit, pyrosequencing using PSQ 96M System | BDNF | Promoters, CpG islands | At baseline, a higher methylation percentage in MDD compared with no depression. Higher BDNF methylation independently associated with prevalent depressive disorder at baseline and follow-up |
35 | Kaut et al. | 2015 | The Netherlands | 12 | Recurrent MDD and controls | Pilot–replication | MDD | Postmortem brain, HIP, PFC tissue | Bisulfite conversion with a ZymoResearch bisulfite kit and Ininium Human Methylation 450K bead arrays | Genome-wide, selected genes for replication | three CpG sites on GRIN2A | 11 genes in the hypocampus and 20 genes in the prefrontal cortex revealed differential methyaltion. In replication, GRIN2A was found hypermethylated in both tissues and single CpG level |
36 | Kang et al. | 2015 | South Korea | 631 | Aged 65 years and plus for cases and controls | Longitudinal | Depression in general | Leukocytes | Bisulfite conversion using EpiTech Bisulfite Kit, pyrosequencing using the PSQ 96M System | BDNF | Promoters, CpG islands | Higher BDNF methylation was independently associated with depression and severe depressive symptoms |
37 | Kang et al. | 2015 | South Korea | 309 | Hospital-based, all women with breast cancer undergoing breast surgery | Longitudinal | Mix of major and minor depression | Leukocytes | Bisulfite conversion using EpiTech Bisulfite Kit, pyrosequencing using the PSQ 96M System | BDNF | Promoters, CpG islands | A higher methylation percentage at CpG9 with depression, both 1 week and 1 year after breast cancer |
38 | Januar et al. | 2015 | France | 1024 | Aged 65 years and plus for cases and controls | Case–control | MDD | Buccal cells | Bisulfite conversion, PCR, and sequencing. Sodium-bisulfite conversion using the EpiTect® Bisulfite Kit); sequencing was performed using a BigDye Terminator v3.1 Cycle Sequencing Kit | BDNF | Promoters, CpG islands | Depression at baseline and chronic late life was associated with higher BDNF methylation |
39 | Frodl et al. | 2015 | Ireland | 60 | Cases had experienced acute depressive episodes, matched on age and sex with controls | Case–control | MDD | Whole blood | Bisulfite conversion, pyrosequencing; PyroMark Q24 | SLC6A4 | Promoters, CpG islands | MDD was not significantly associated with methylation |
40 | Booij et al. | 2015 | Canada | 69 | Adults, matched on sex and gender between cases and controls, cases not taking antipsychotics or mood stabilizers | Case–control | MDD | Whole blood | Bisulfite conversion, pyrosequencing; PyroMark Q24 Software (Qiagen) for methylation percentage at each site. | SLC6A4, treatment response | Gene body, CpG islands | MDD diagnosis was not significantly associated with DNA methylation. Patients with SSRIs had greater methylation |
41 | Numata et al. | 2015 | Japan | 63 | Hospital-based cases and matched controls | Case–control | MDD | Whole blood | Bisulfite conversion using EZ DNA methylation Kit (ZYMO research), Infinium Human Methylation 450 Beadchips | Genome-wide | 363 (313 CGIs) | 363 CpG sites demonstrated lower DNA methylation in MDD patients than in controls. 18 MDD-associated DNA methylation markers to discriminate cases from controls |
42 | Haghighi et al. | 2015 | USA | 53 | MDD and suicide cases and controls | Case–control | MDD | Whole blood | Bisulfite conversion using Illumina Infinium HumanMethylation27 BeadChip | Genome-wide | Not mentioned | Increased age-related DNA methylation perturbations in the prefrontal cortex in major depression suicide compared with nonpsychiatric controls |
43 | Tadic et al. | 2014 | Germany | 39 | MDD inpatients | Cohort | MDD | Leukocytes | Bisulfite conversion, PCR, and sequencing using BigDye Terminator v3.1 Cycle Sequencing Kit | Treatment response on BDNF | Promoters, CpG islands | Antidepressant treatment did not significantly affect the methylation at BDNF promoter IV; thus, changes in the methylation status in this DNA region seem not to be involved in the response to antidepressant treatment |
44 | Khulan et al. | 2014 | Finland | 166 | Senior cases and controls | Case–control | Depressive symptoms | Whole blood | Bisulphite conversion using EZ DNA methylation kit, bead array using Illumina methylation 450k beadchip and Infinium chemistry | Genome-wide | CpG islands, shores, and TSS | Hypomethylation was associated with depressive symptoms. The results supported that DNA methylation differences may be important in the pathogenesis of psychiatric disease |
45 | Domschke et al. | 2014 | Germany | 94 | Caucasian cases with antidepressants | Cohort | MDD | Whole blood | Sodium bisulfite converted using EZ-96 DNA methylation Kit, PCR, and sequencing using BigDye Terminator | Treatment response on 5-HTT | Gene body, CpG islands | Hypomethylation of the 5-HTT transcriptional control region might impair antidepressant treatment response in Caucasian patients with MDD |
46 | Kaminsky et al. | 2014 | USA | Not mentioned | Not mentioned | Longitudinal | PPD | Whole blood | Not mentioned | HP1BP3 and TTC9B | Not mentioned | HP1BP3 and TTC9B (hypermethylation) predicted PPD with an area under the receiver operator characteristic curve (AUC) of 0.87 |
47 | Guintivano et al. | 2014 | USA | 93 | Caucasian women | Longitudinal | PPD | Whole blood | Illumina’s Infinium Human Methylation450 Beadchip Kit | Genome-wide | Two loci within the HP1BP3 and TTC9B genes | CpG methylation levels at two loci within the HP1BP3 and TTC9B genes were identified as biomarkers predictive of PPD |
48 | Tseng et al. | 2014 | China (Taiwan) | 74 | MDD cases and controls | Case–control | MDD | Leukocytes | ELISA-based for global DNA methylation profiling. MethylFlash methylated DNA quantification kit (for 5-mc), MethylFlash hydroxymethylated DNA quantification kit (for 5-hmc) | Genome-wide | Gobal methylation levels, no site mentioned | Lower levels of 5-hmc and 5-mc in severe MDD than controls, no difference among severe and remitted patients |
49 | Okada et al. | 2014 | Japan | 100 | Untreated cases or cases without a history of depressive episodes | Case–control | MDD | Whole blood | Bisulfite conversion using EZ DNA methylation kits; analyzed using a MassARRAY | SLC6A4 | Promoters, CpG islands | The pre-treatment-methylation rate(CpG3) of SLC6A4 is associated with therapeutic responses to antidepressants in un-medicated patients with MD |
50 | Na et al. | 2014 | South Korea | 117 | Untreated cases (no history of antidepressants) | Case–control | MDD | Whole blood | Bisulfite conversion, pyrosequencing using PyroMark ID system with the Pyro Gold reagent kit (Qiagen, Valencia, CA, USA) | NR3C1 | Promoters, CpG islands | MDD had significantly lower methylation than healthy controls at two CPG sites (CpG3, -4) |
51 | Davies et al. | 2014 | UK | 454 (50 twins, 354 case–control) | Monozygotic twins, discordant for depression | Twin study and case–control | MDD | Whole blood | Methylated DNA immunoprecipitation combined with ultra-deep sequencing (MeDIP-seq) (enrichment for methylated regions) | Genome-wide | Coding region of ZBTB20 gene | Both AU and UK did not identify DMR of genome-wide significance. MDD was associated with hypermethylation on the coding region of ZBTB20 |
52 | Carlberg et al. | 2014 | Austria | 554 | Unrelated in- and outpatients of White European origin | Case-Control | MDD | Peripheral blood mononuclear cells (PBMCs) | Bisulfite conversion using EZ-96 DNA Methylation Kit. Used methylation-specific quantitative PCR following the MethyLight protocol using SYBR green | BDNF, treatment response | Promoters, CpG islands | BDNF exon I promoter was significantly increased in MDD. Current antidepressant therapy was associated with increased methylation |
53 | Dell’Osso et al. | 2014 | Italy | 87 | Stable, pharmacological treated MDD and matched controls | Case–control | MDD | Peripheral blood mononuclear cells (PBMCs) | Bisulfite conversion, PCR, and sequencing | BDNF, treatment response | Promoters, CpG islands | Overall lithium and valproate tend to decrease the DNA methylation level at BDNF gene promoter, when compared to other classes of medications. However, within each different disorder, mood stabilizers did not seem to affect DNA methylation, suggesting that such an alteration was likely not due to treatment use |
54 | Zhao et al. | 2013 | USA | 84 | MZ twins (male veterans) for lifetime and concurrent MDD | Twin study | MDD | Leukocytes | Bisulfite conversion using EZ DNA methylation kit, pyrosequencing using PSQ 96 HS System | SLC6A4 | Promoters, CpG islands | Variation in methylation level within the promoter region of SLC6A4 was associated with variations in depressive symptoms. A 10% increase in the difference in mean DNA methylation level was associated with a 4.4-fold increase in the difference in BDI scores. The 5-HTTLPR genotype did not modulate this association. The use of antidepressants did not affect the relationship between SLC6A4 methylation and depressive symptoms |
55 | Melas et al. | 2013 | Sweden | 174 | Female cases and controls | Case–control | Depression in general | Saliva | Bisulfite conversion using EZ-96 DNA Methylation-Gold Kit, PCR, and sequencing, EpiTyper software | MAOA | Gene body, other CpG sites | Overall MAOA methylation levels were decreased in depressed females compared to controls |
56 | Byrne et al. | 2013 | Australia | 48 | Queenland twin study (discordant MDD and concordant no MDD) | Twin study | MDD | White blood cells | Bisulphite conversion, Illumina Human Methylation 450 BeadChip | Genome-wide | 17 sites (6 CpG islands) | The difference in mean methylation was significant in females within discordant pairs, but not in males |
57 | Kim et al. | 2013 | South Korea | 286 | Patients with a recent ischemic stroke | Longitudinal | Post-stroke depression (both major and minor) | Leukocytes | Bisulfite conversion using EpiTech Bisulfite Kit, pyrosequencing using PSQ 96M System | SLC6A4 | Promoters, CpG islands | Higher SLC6A4 methylation status was independently associated with a major post-stroke depression at both baseline and follow-up |
58 | Kim et al. | 2013 | South Korea | 286 | Patients with a recent ischemic stroke | Longitudinal | Post-stroke depression (both major and minor) | Leukocytes | Bisulfite conversion using EpiTech Bisulfite Kit, pyrosequencing using PSQ 96M System | BDNF | Promoters, CpG islands | Prevalent, persistent, and incident PSD had a higher BDNF methylation status. CpG site 6 was significantly associated with incident post-stroke depression |
59 | Kang et al. | 2013 | South Korea | 108 | Patients with MDD only | Longitudinal | MDD | Leukocytes | Bisulfite conversion using EpiTech Bisulfite Kit, pyrosequencing using PSQ 96M System | SLC6A4, treatment response | Promoters, CpG islands | SLC6A4 methylation status as a marker for childhood adversities among MDD, but was not associated with treatment outcomes |
60 | Bayles et al. | 2013 | Australia | 106 | Newly diagnosed or currently untreated and have not been receiving antidepressants for at least 4 weeks | Case–control | MDD | Leukocytes | Bisulfite conversion, PCR, and sequencing; EpiTYPER methylation analysis | SLC6A2 or NET | Promoters, CpG islands | There were no significant differences between MDD cases and controls in terms of the pattern of methylation of the SLC6A2 promoter. Antidepressant treatment did not change the result |
61 | Zill et al. | 2012 | Germany | 162 | Caucasian cases and controls | Case–control | MDD | Leukocytes | Bisulfite conversion, PCR, and sequencing, EpiTect Bisulfite Kit | ACE | Promoters, CpG islands | MDD patients showed a hypermethylation pattern at all the CpG sites compared to healthy controls |
62 | Sabunciyan et al. | 2012 | USA | 154 | MDD and controls | Replication | MDD | Postmortem frontal cortex, lymphoblastoid cell lines, postmortem brain | CHARM assay platform | Genome-wide | No site identified | PRIMA1 significantly increased the methylation in MDD in pilot, but not in replication |
63 | Uddin et al. | 2011 | USA | 100 | Lifetime depression cases and non-depressed controls | Case–control | Depression in general | Whole blood | Bisulfite conversion using EZ-96 DNA Methylation Kit, bead array using HumanMethylation27 (HM 27) DNA Analysis BeadChip | Genome-wide | 21 uniquely methylated and 107 uniquely unmethylated sites with depression | Uniquely unmethylated gene sets distinguished between those with versus without lifetime depression. In particular, some processes (e.g., brain development, tryptophan metabolism) showed patterns suggestive of increased methylation among individuals with depression whereas others (e.g., lipoprotein) showed patterns suggestive of decreased methylation among individuals with depression |
64 | Fuchikami et al. | 2011 | Japan | 38 | Japanese adults | Case–control | MDD | Whole blood | Bisulfite conversion using EZ DNA methylation kit | BDNF | Promoters, CpG islands | Significant methylation difference was found in CpGI, but not in -IV |
65 | Olsson et al. | 2010 | Australia | 150 | Australian adolescents (cases and controls) | Case–control | MDD | Buccal cells | Bisulfite conversion, Sequenom MassARRAY EpiTyping | SLC6A4 | Promoters, CpG islands | There was no association between depressive symptoms and either buccal cell 5-HTT methylation or 5-HTTLPR. Depressive symptoms were more common among those with elevated buccal cell 5-HTT methylation who carried a 5-HTTLPR short allele |
66 | Alt et al. | 2010 | The Netherlands | 12 | Depression and control groups matched for sex, age, brain weight, and postmortem delay | Case–control | MDD | Brain tissues | Bisulphite conversion, pyrosequencing using PyroMark ID | NR3C1 | Promoters, CpG islands | No significant difference in methylation pattern was found between case and control groups |
67 | Philibert et al. | 2008 | USA | 192 | Lifetime MDD and controls | Longitudinal | MDD | Lymphoblast cell lines | Bisulfite conversion, methylation ratios calculated by usingMassARRAY | SLC6A4 | Promoters, CpG islands | Greater amounts of methylation in females vs males, and a trend of higher methylation was associated with greater vulnerability of lifetime MDD |
