Skip to main content
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2020 Jun 1.
Published in final edited form as: Psychiatry Clin Neurosci. 2019 Mar 28;73(6):323–330. doi: 10.1111/pcn.12835

Genome-wide DNA methylation investigation of glucocorticoid exposure within buccal samples

Patricia R Braun a,d, Mai Tanaka-Sahker a, Aubrey C Chan a, Sydney S Jellison a, Mason J Klisares a, Benjamin W Hing a, Yaseen Shabbir a, Lindsey N Gaul a, Yasunori Nagahama b, Julian Robles a, Jonathan T Heinzman a, Sayeh Sabbagh a, Ellyn M Cramer a, Gabrielle N Duncan a, Kumi Yuki a, Liesl N Close b, Brian J Dlouhy b, Matthew A Howard III b, Hiroto Kawasaki b, Kyle M Stein c, James B Potash d, Gen Shinozaki a,b,e,f
PMCID: PMC6561812  NIHMSID: NIHMS1018119  PMID: 30821055

Abstract

Aim:

Glucocorticoids play a major role in regulating the stress response, and an imbalance of glucocorticoids has been implicated in stress-related disorders. Within mouse models, CpGs across the genome have been shown to be differentially methylated in response to glucocorticoid treatment, and using the Infinium 27K array, it was shown that humans given synthetic glucocorticoids had DNA methylation (DNAm) changes in blood. However, further investigation of the extent to which glucocorticoids affect DNAm across a larger proportion of the genome is needed.

Methods:

Buccal samples were collected before and after synthetic glucocorticoid treatment in the context of dental procedure. This included 30 tooth extraction surgery patients who received 10 mg of dexamethasone. Genome-wide DNAm was assessed with the Infinium HumanMethylationEPIC array.

Results:

Five CpGs showed genome-wide significant DNAm changes that was >10%. These differentially methylated CpGs were in or nearest the following genes: ZNF438, KLHDC10, miR-544 or CRABP1, DPH5, and WDFY2. Using previously published datasets of human blood gene expression changes following dexamethasone exposure, a significant proportion of genes with FDR-adjusted significant CpGs were also differentially expressed. A pathway analysis of the genes with FDR-adjusted significant CpGs revealed significant enrichment of olfactory transduction, pentose and glucuronate interconversions, ascorbate and aldarate metabolism, and steroid hormone biosynthesis pathways.

Conclusion:

High-dose synthetic glucocorticoid administration in the setting of dental procedure was significantly associated with DNAm changes within buccal samples. These findings are consistent with prior findings of an influence of glucocorticoids on DNAm in humans.

Keywords: DNA methylation, Epigenomics, Glucocorticoids, Stress (Psychological), Dexamethasone

Introduction

Stress is a major contributing factor in the development of psychiatric disorders 1, 2. Major depressive disorder (MDD) has been linked to stress 1, 3, with the incidence of a depressive episode increased with a major stressful event 4. Approximately 17% of the population is affected by MDD and about 30% of patients do not undergo remission and suffer from chronic depression despite treatment 5, 6. Post-traumatic stress disorder (PTSD) is precipitated by a traumatic event, and it affects 5–10% of the population 7.

The hypothalamic-pituitary-adrenal (HPA) axis plays an essential role in mediating the stress response, which ultimately results in the release of glucocorticoids (GCs) from the adrenal cortex that regulate tissues throughout the body through the binding of GCs to the glucocorticoid receptor (GR) 8. Dysregulation of this system is often seen with chronic stress and in psychiatric disorders 9, 10. PTSD has been associated with suppressed cortisol levels, which can lead to hypersensitivity of the GR and enhanced negative feedback inhibition 11. In contrast, in MDD, blunted cortisol and adrenocorticotropic hormone responses are often present 12. These alterations in GC feedback on the HPA axis constitute one of the most consistent biomarkers for MDD and PTSD 13. The contribution of GCs to psychiatric disorders is further evidenced in a large epidemiological study of individuals given GC treatment for a range of inflammatory disorders 14. Individuals treated with GCs had an elevated risk of suicide or suicide attempt, depression, mania, delirium, and panic disorder 14. This is consistent with the hypothesis that the impact of stress on psychiatric disorders is mediated in part by GC exposure.

MDD, PTSD, and stressful life events have all been associated with changes in DNA methylation (DNAm) 1517. To tease out the role of GCs themselves, rodent models of GC exposure have been employed. One study reported DNAm changes in Fkbp5, the co-chaperone of the GR, within the hippocampus 18, which was later found to correlate partially with similar changes in blood 19. A genome-wide investigation of GC exposure in mice found differentially methylated regions (DMRs) across the genome, and 209 DMRs that mapped to the same genomic coordinates in the hippocampus and blood were significantly changed with GC exposure in the same direction within both tissues 20.

Only one study, to our knowledge, has previously investigated genome-wide DNAm changes in response to synthetic glucocorticoid administration in humans. In that study, investigators examined patients with chronic obstructive pulmonary disease (COPD) and made a DNAm comparison using blood samples from those who were given glucocorticoid treatment and those who were not. The study was performed with the Illumina 27K methylation array 21. A total of 511 sites were significantly differentially methylated, and an enrichment of corresponding genes involved in membrane components, hemostasis and coagulation, cellular ion homeostasis, leukocyte and lymphocyte activation, and protein transport was found 21.

To better understand the impact of GC exposure on genome-wide methylation within humans, we investigated DNAm changes by comparing buccal samples taken before and after individuals underwent tooth extraction and were given a synthetic GC, dexamethasone, prior to the procedure. We employed the Illumina EPIC array, which assays >850,000 CpG sites, and we thus substantially expanded the coverage of the genomic landscape compared to earlier work.

Methods

Participants and sample collection

Two cohorts of individuals exposed to synthetic glucocorticoids were used and both were approved by the University of Iowa’s Human Subjects Research Institutional Review Board. Written informed consent was obtained. The first cohort included individuals recruited from the dental clinic (DC). These subjects underwent tooth extraction procedure, and 10 mg of dexamethasone was administered prior to surgery. From 30 individuals, buccal tissue was collected with swabs (Puritan, 25–1506 1PF TT MC) before and after dental surgery. Buccal samples were chosen because they are more easily obtained than blood samples, and the dental procedure patients had difficulty producing saliva for sampling. Subject characteristics from the DC cohort are in Table 1. Depressive symptoms and childhood adversity were measured with the Quick Inventory of Depressive Symptomatology Self-Report (QIDS-SR-16) scale and the Adverse Childhood Experiences (ACE) questionnaires, respectively.

Table 1.

Clinical characteristics of the dental procedure study participants (n = 30)

Patient Sex Age GC duration (min) QIDS-SR_16 ACE
 2 female 31 112 3 1
 5 female 24 53 2 0
 7 female 22 109 5 3
 9 female 25 NA NA NA
13 male 26 118 5 1
14 female 19 106 4 0
16 male 22 118 15 5
17 female 21 170 15 3
18 female 36 134 NA NA
19 female 25 115 9 10
20 male 28 NA NA NA
21 female 23 107 9 4
22 female 24 244 16 3
24 female 19 128 NA NA
26 female 25 137 NA NA
28 female 29 199 9 1
29 male 25 135 4 0
30 male 18 110 13 2
33 female 24 112 NA NA
35 female 19 101 4 5
36 female 22 45 0 1
41 male 27 124 NA NA
43 male 19 44 6 2
44 female 36 35 2 0
45 male 22 58 5 0
52 female 22 NA 8 1
54 male 58 130 17 2
55 male 19 230 3 3
57 female 30 NA 10 1
58 female 27 110 11 4

The second cohort (NSG) included twenty-one subjects with medically intractable epilepsy undergoing neurosurgery who were recruited for a separate study between March 2014 and April 2017 at the University of Iowa Hospitals and Clinics. Before and after dexamethasone exposure, whole blood samples were collected in EDTA tubes, saliva samples with the Oragene DISCOVER™ kit (DNA Genotek Inc., OGR-500), and buccal tissue with swabs (Puritan, 25–1506 1PF TT MC). Resected brain tissue samples were taken within the operating room before dexamethasone exposure and between 30 minutes to 3 hours after. In total, both pre- and post-samples were collected for blood (n = 18), buccal tissue (n = 13), saliva (n = 21), and brain (n = 10). Samples were immediately stored and transported on dry ice and a portion of each brain region was sent to pathology to exclude tissues with malignancy. All samples were stored at −80°C. NSG sample characteristics are in Table 2.

Table 2. Subject and sample characteristics of the neurosurgery study participants.

Sample tissues available for each individual varied. Details of analyzed samples are shown. Timing of when peripheral tissues were obtained are listed as “days”. Day 0 is indicative of samples taken the day of surgery, after surgery is completed. For brain samples analyzed, the time (in minutes) between administration of the synthetic glucocorticoid and when the second (post) brain sample was resected was recorded. Abbreviations. GC: glucocorticoid, FCD: focal cortical dysplasia.

      Day(s) after surgery for peripheral sample collection    
Subject Sex Age Saliva Blood Buccal Brain region (pre) Brain region (post) GC duration (minutes) Pathology report (pre) Pathology report (post)
101 female 50 638        
102 female 46 309        
103 female 31 362 2      
104 female 51 263 0      
105 female 28 222 0      
106 male 61 2 0      
107 male 43 198 0      
108 male 40 241 0      
109 male 7 23 0 23 Occipital area Occipital area 180 FCD type Ic and gliosis Same as pre-GC sample
110 male 17 3 0 3 Inferior lateral temporal cortex Inferior lateral temporal cortex 38 Evidence of cortical dysplasia, focal meningitis, and cortical injury consistent with electrode placement Same as pre-GC sample
111 female 14 3 0 3  
112 male 31 2 0 2  
113 male 36 2 0 2 Temporal cortex Temporal cortex 100 Moderate, patchy gliosis, and multiple superficial subacute microinfarcts with adjacent leptomeningeal chronic inflammation consistent with history of previous dural grid placement Same as pre-GC sample
114 male 30 3 0 3 Lateral temporal cortex Lateral temporal cortex 130 Patchy mild gliosis Same as pre-GC sample
115 male 24 2 0 2 Lateral temporal cortex Hippocampus 90 Mild neuronal heterotopia and mild patchy gliosis Hippocampal sclerosis and moderate patchy gliosis
116 male 34 2 0 2 Lateral temporal cortex Lateral temporal cortex 130 Patchy moderate gliosis and focal meningitis secondary to electrode placement Same as pre-GC sample
117 female 49 2 0 2 Lateral temporal cortex Inferior lateral temporal cortex 100 FCD type Ic and white matter gliosis Subpial and white matter gliosis
118 male 5 4 0 4  
119 male 38 2 0 2 Lateral temporal cortex Hippocampus 64 White matter and cortical gliosis and subacute infarct Hippocampal sclerosis
120 male 11 122 0 14 Temporal operculum Temporal operculum 152 Mild neuronal disorganization with white matter and cortical gliosis Same as pre-GC sample
121 male 5 4 0 4 Temporal pole Temporal pole 32 FCD Same as pre-GC sample

Methylome assays

Genomic DNA was extracted from whole blood, buccal, saliva, and brain tissues with the MasterPure™ DNA extraction kit (Epicenter, MCD85201) following the respective protocols for each tissue type. DNA quality was assessed with NanoDrop spectrometry and quantified with the Qubit™ dsDNA Broad Range Assay Kit (ThermoFisher Scientific, Q32850). For each sample, 500 ng of DNA was bisulfite converted with the EZ DNA Methylation™ Kit (Zymo Research, D5002). The Infinium HumanMethylationEPIC BeadChip™ Kit (Illumina, WG-317–1002) was used to analyze genome-wide DNAm in all brain, blood, saliva, and buccal samples. The arrays were scanned with the Illumina iScan platform.

The methylation data were processed with the R packages Minfi and RnBeads 2224. The two cohorts were analyzed separately. Background correction was performed with the Noob method in Minfi 23, 25. Using RnBeads, probes were filtered out if they: 1) overlapped within 5 bp of a SNP (DC: 21,414; NSG: 21,358 probes), 2) had a detection P-value > 0.01 or were considered unreliable measures based on RnBeads’s greedy-cut algorithm (DC: 26,065; NSG: 18,864 probes), or 3) were context-specific sites (DC: 2,825; NSG: 2,873 probes). After filtering, 816,532 probes remained in the DC cohort and 822,996 in the NSG cohort. Samples were normalized with beta mixture quantile dilation (BMIQ) 26. Samples from the same individual were verified by generating a heatmap cluster of the 65 SNP probes on the array.

Statistical analysis

All statistical analyses were performed in R 27. Differential methylation was assessed with RnBeads using the limma method with a paired analysis by subject 24, 28. A hierarchical linear model was fitted with the empirical Bayes approach. Covariates included age, sex, and surrogate variables estimated with the “be” method 24, 29. Fisher’s exact test was performed for the enrichment of glucocorticoid expressed genes. The R package missMethyl was used for pathway analysis with KEGG-defined pathways, and it employs a hypergeometric test taking into account the number of probes per gene 30. For the comparison of DNAm from buccal and brain tissues, the correlation coefficient and associated p-value were calculated with Spearman’s test.

Results

Genome-wide analysis of buccal samples from dental procedure patients

Thirty individuals were recruited from the dental clinic, and buccal samples were taken before and after tooth extraction procedure. Subject characteristics can be found in Table 1. Average age was 25.6 ± 7.6 years old. The QIDS-SR depressive symptom score average was 7.6 ± 4.9, and the ACE average score was 2.3 ± 2.3, indicating a relatively young and healthy population. DNA samples were run on the Illumina EPIC array, and DNAm differences between pre- and post-dexamethasone samples were analyzed with the limma method in RnBeads, with a paired analysis adjusted for age, sex, and estimated surrogate variables.

For the analysis of individual CpGs, we set a Bonferroni genome-wide significance threshold for this experiment at p < 6.12 × 10−8 based on the 816,532 probes left after preprocessing, and we further only considered CpG changes meaningful if they showed DNAm differences greater than 10%. Figure 1 shows the distribution of individual CpG differences and their corresponding raw p-value. Five CpGs were genome-wide significant with differences greater than 10%, and all five also showed a decrease in methylation after dexamethasone exposure (Table 3). The top finding is an intergenic CpG closest to the zinc finger protein 438 (ZNF438; average DNAm: pre-steroid 68%, post-steroid 54%; p-value = 3.21 × 10−9). Subsequent top findings that surpassed the Bonferroni threshold included CpGs in or near the following genes: KLHDC10 (average DNAm: pre-steroid 85%, post-steroid 74%; p-value = 1.11 × 10−8), miR-544 or CRABP1 (average DNAm: pre-steroid 82%, post-steroid 70%; p-value = 1.56 × 10−8), DPH5 (average DNAm: pre-steroid 73%, post-steroid 58%; p-value = 3.66 × 10−8), and WDFY2 (average DNAm: pre-steroid 92%, post-steroid 81%; p-value = 4.36 × 10−8). For a more comprehensive list of the top differential methylation results, Supplemental Table 1 contains all the CpGs with the less conservative FDR-adjusted p-value of 0.01 and differences greater than 10%.

Figure 1.

Figure 1.

The degree of significance of all individual CpGs represented on the Illumina EPIC array and their methylation differences after dexamethasone exposure in buccal samples from individuals who underwent tooth extraction (n = 30).

Table 3.

Top differentially methylated CpGs associated with dexamethasone exposure in buccal samples from the dental procedure cohort.

cgid Chr Start mean.Pre mean.Post mean.diff p.val Gene Genic Location Distance in bps
cg14315118 chr10 31,426,697 0.68 0.54 −0.14 3.21E-09 ZNF438 upstream 105,831
cg01119284 chr7 129,735,795 0.85 0.74 −0.10 1.11E-08 KLHDC10 intron NA
cg26879918 chr15 78,617,501 0.82 0.70 −0.13 1.56E-08 miR-544; CRABP1 upstream; upstream 3,183; 23,071
cg16495532 chr1 101,472,258 0.73 0.58 −0.15 3.66E-08 DPH5 intron NA
cg06335209 chr13 52,247,531 0.92 0.81 −0.10 4.36E-08 WDFY2 intron NA

We further investigated whether psychiatric risk genes showed evidence for differential methylation from dexamethasone administration. In this analysis, CpGs were considered significant if their FDR-adjusted p-value was < 0.05 and they had differences greater than 10%. Among the genes considered (FK506 binding protein 5 [FKBP5], the glucocorticoid receptor [NR3C1], brain-derived neurotrophic factor [BDNF], the serotonin transporter [SLC6A4], and corticotropin-releasing hormone [CRH]), there were three significant CpGs in FKBP5, five in NR3C1, and one in BDNF (Table 4).

Table 4.

Methylation of pre- and post-dexamethasone buccal samples of CpGs within candidate genes from psychiatric disorders which were significant at FDR-adjusted p-value < 0.05 and with differences greater than 10% before and after dexamethasone exposure in the dental procedure cohort.

cgid Chr Start mean.Post mean.Pre mean.diff p.val fdr Gene
cg05039098 chr6 35,618,763 0.72 0.83 −0.11 1.16E-03 0.02 FKBP5
cg07696519 chr6 35,619,165 0.61 0.76 −0.15 8.77E-05 0.01 FKBP5
cg03546163 chr6 35,654,363 0.65 0.76 −0.11 4.73E-03 0.05 FKBP5
cg25708981 chr5 142,697,868 0.68 0.81 −0.13 1.77E-04 0.01 NR3C1
cg05900547 chr5 142,769,791 0.71 0.84 −0.13 3.39E-05 0.01 NR3C1
cg19432243 chr5 142,770,757 0.55 0.69 −0.14 2.95E-04 0.01 NR3C1
cg19820298 chr5 142,770,782 0.55 0.71 −0.16 3.96E-04 0.01 NR3C1
cg21979215 chr5 142,815,807 0.60 0.74 −0.14 2.85E-03 0.04 NR3C1
cg23143371 chr11 27,682,015 0.59 0.72 −0.13 3.59E-03 0.05 BDNF

Additionally, the genes implicated in GWASs of schizophrenia, bipolar disorder, and MDD (N=366) were analyzed (Supplemental Table 2)3133. 108 of the genes had at least one CpG that was differentially methylated, and 28 genes had at least one CpG that was differentially methylated and showed correlation between buccal and brain tissue DNAm. This included 240 individual CpGs that were differentially methylated, and 35 that were differentially methylated and correlated (Supplemental Table 3).

Pathway analysis

A pathway analysis was performed with KEGG-defined pathways using the missMethyl R package to correct for the different numbers of probes per gene on the array 30. Input for the analysis included CpGs with differences greater than 10% and FDR-adjusted p-values < 0.01 (N = 2,587). Top pathways included olfactory transduction (FDR-adjusted p-value = 1.15 × 10−3), pentose and glucuronate interconversions (FDR-adjusted p-value = 4.31 × 10−3), ascorbate and aldarate metabolism (FDR-adjusted p-value = 4.31 × 10−3), steroid hormone biosynthesis (FDR-adjusted p-value = 4.31 × 10−3), and retinol metabolism (FDR-adjusted p-value = 0.11; Table 5).

Table 5. Significance of top KEGG-defined pathways enriched for genes with CpGs differentially methylated after dexamethasone exposure in buccal samples from individuals who underwent tooth extraction.

N = number of genes in the pathway. DM = number of genes with differentially methylated CpGs within the associated pathway.

Pathway N DM p.val fdr
Olfactory transduction 406 50 3.50E-06 1.20E-03
Pentose and glucuronate interconversions 34 14 3.40E-05 4.30E-03
Ascorbate and aldarate metabolism 27 12 4.60E-05 4.30E-03
Steroid hormone biosynthesis 59 18 5.30E-05 4.30E-03
Retinol metabolism 66 15 2.00E-03 0.11
Drug metabolism - cytochrome P450 69 15 2.00E-03 0.11
Porphyrin and chlorophyll metabolism 42 12 2.80E-03 0.13
Axon guidance 174 46 5.20E-03 0.2
Chemical carcinogenesis 82 16 5.60E-03 0.2
Drug metabolism - other enzymes 79 17 6.70E-03 0.22

Overlap with genes differentially expressed following GC exposure

Differential gene expression within blood samples following administration of 1.5 mg dexamethasone has been investigated in two human studies 34, 35. These studies yielded 1,052 genes that were differentially expressed in response to dexamethasone across both studies and that are covered on the Illumina EPIC array. As with the pathway analysis, genes that were included had at least one CpG that was differentially methylated in our data at FDR-adjusted p-value < 0.01 and a DNAm difference > 10%. Of these, 154 genes (15%) also showed a significant proportion with differential methylation after dexamethasone exposure within our study (Fisher’s exact p-value = 8.63 × 10−4; Supplemental Table 4).

Analysis of dexamethasone exposure in neurosurgery cohort

We performed an additional study in a separate cohort of patients given dexamethasone. In this cohort, patients undergoing neurosurgical brain resection for treatment refractory epilepsy were administered dexamethasone during surgery. Blood, buccal, saliva, and brain samples were taken before and after dexamethasone exposure. No genome-wide significant changes in DNAm were found from any of the tissues, though the top differentially methylated CpGs are listed in Supplemental Table 5. Of the CpGs with FDR-adjusted p-value < 0.01 and differences > 10% from the dental procedure cohort, a subset replicated in additional tissues with differences greater than 10% and at nominal significance. This included 13 CpGs identified in samples from blood, 5 from buccal tissue, 58 from saliva, and 6 from brain (Supplemental Table 6). One CpG in the gene regulating synaptic membrane exocytosis 2 (RIMS2) was significant in all three peripheral tissues.

Correlation of buccal DNA methylation of top findings to brain tissue DNA methylation

With the NSG cohort in this paper, a separate analysis was performed of the five genome-side significant results, assessing the degree of correlation between DNAm of buccal and brain tissues (n = 21), using a web-based resource we created, IMAGE-CpG (https://han-lab.org/methylation/default/imageCpG) 36. We observed non-significant moderate correlation between brain and buccal DNAm for the CpG in WDFY2 (cg06335209; rho = 0.5, p-value = 0.07), and low-moderate correlation for the CpG in KLHDC10 (cg01119284; rho = 0.3, p-value = 0.27).

Discussion

This study assessed the impact of synthetic glucocorticoids on DNAm within humans. Studies in mice have demonstrated that chronic glucocorticoid exposure evokes DNAm changes across the genome in hippocampal and blood tissues 20. In a DNAm comparison of COPD patients with and without glucocorticoid treatment DNAm differences across 511 CpGs were found in blood samples using the Illumina 27K array 21. Individuals with Cushing’s syndrome have sustained elevated glucocorticoid levels, and in a case-control Illumina 450K array study, additional DNAm differences were found to be associated with this disorder 37. Our study builds upon these reports by demonstrating across a larger portion of the genome DNAm changes that occur within individuals exposed to a single high dose of glucocorticoids.

Among the top differentially methylated CpGs in this study was one whose nearest gene is ZNF438, a gene implicated in transcriptional inhibition and in regulation of the immune system 38, 39, and both transcriptional control and the immune system are regulated by glucocorticoids. A decrease in methylation was also significant in a CpG within the gene that encodes the cellular retinoic acid binding protein 1 (CRABP1), a protein involved in retinoic acid-mediated proliferation and differentiation 40. CRABP1 binds to and catalyzes the destruction of retinoic acid 41. Of interest, in rats exposed to chronic unpredictable mild stress, Crabp1 was decreased in the hypothalamus 42. Though not FDR-significant, retinol metabolism was also one of the top pathways enriched from the dexamethasone-induced DNAm findings in this study. Retinoic acid regulates gene transcription and is important in processes including the development of the central nervous system 43 and neuronal plasticity 4446. Hyperactivity of the HPA axis along with anxious and depressive behaviors have been seen in rats treated with retinoic acid 47.

Intriguingly, steroid hormone biosynthesis was also among the top pathways, indicating that high dose glucocorticoid treatment may influence epigenetic regulation of the HPA axis. This is also supported by the differential methylation seen in FKBP5 and NR3C1. In addition, steroid hormone biosynthesis involves the production of other cholesterol-derived hormones in addition to cortisol, including estradiol, aldosterone, and testosterone, all of which interact with and mediate the stress response 4854. The enrichment of steroid hormone biosynthesis genes suggests these genes may be regulated by dexamethasone-induced DNAm changes.

A major limitation of this study is the potential confounding of cell-specific DNAm effects. Buccal samples were used, and although they are primarily epithelial cells, contamination from saliva, which predominantly includes white blood cells, could impact methylation levels. Additionally, the sample size was limited (n = 30), and a replication of the findings within the same dental procedure context was not possible. Therefore, we included the additional NSG analysis that involves the analysis of three peripheral tissues and resected brain tissue. However, this cohort was also limited in sample size (blood (n = 18), buccal (n = 13), saliva (n = 21), and brain samples (n = 10), and differences with the DC cohort include different dexamethasone doses and timing of the post-sample collection. Furthermore, the brain regions resected were not uniform among all the subjects. Future studies will be needed to determine the functional impact of the differential methylation findings in response to GC exposure.

Our study investigated DNAm differences across the genome with the Illumina EPIC array in individuals given synthetic glucocorticoids. Significant DNAm changes were observed in genes involved in transcriptional inhibition, the immune system, cellular proliferation, and retinoic acid metabolism. High doses of synthetic glucocorticoid exposure may also impact the HPA axis through epigenetic modifications in steroid hormone biosynthesis genes.

Supplementary Material

Supp TableS1. Supplemental Table 1.

Dental procedure cohort differential methylation results of CpGs with FDR-adjusted p-values < 0.01 and differences > 10%.

Supp TableS2. Supplemental Table 2.

Psychiatric associated genes manually assembled from GWAS findings of major depressive disorder, schizophrenia, and bipolar disorder.

Supp TableS3. Supplemental Table 3.

Methylation and degree of correlation between buccal and brain DNA methylation of pre- and post-dexamethasone buccal samples of CpGs within psychiatric disorder risk genes which were significant at FDR-adjusted p-value < 0.05 and with differences greater than 10% before and after dexamethasone exposure in the dental procedure cohort.

Supp TableS4. Supplemental Table 4.

The characteristics of CpGs in the dental cohort that are in differentially expressed genes in blood from a separate study.

Supp TableS5. Supplemental Table 5.

Top 100 CpGs differentially methylated in the neurosurgery cohort within each peripheral tissue and the brain.

Supp TableS6. Supplemental Table 6.

CpGs that are differentially methylated in both the dental procedure cohort (FDR-adjusted p-value < 0.01 and a difference > 10%) and that are nominally significant in the neurosurgery cohort of either blood, buccal, saliva, or brain tissues (p-value < 0.05 and a difference > 10%).

Acknowledgements

Support for GS was through a National Institutes of Health Research Career Development K Award K23MH107654. PRB received training funding from the National Institutes of Health Predoctoral Training Grant T32GM008629, PI Daniel Eberl.

Footnotes

Within the Psychiatry and Clinical Neurosciences journal, this manuscript belongs in the field of molecular psychiatry and psychobiology.

Disclosure Statement

GS is a founder of Predelix Medical LLC, although that entity has no relevance or relationship to this manuscript. All other authors declare no conflicts of interest.

References

  • 1.Kessler R, Davis C, Kendler KS. Childhood adversity and adult psychiatric disorder in the US National Comorbidity Survey. Psychological Medicine 1997;27(5):1101–9. [DOI] [PubMed] [Google Scholar]
  • 2.Kendler KS, Karkowski LM, Prescott CA. Causal relationship between stressful life events and the onset of major depression. Am J Psychiatry 1999;156(6):837–41. [DOI] [PubMed] [Google Scholar]
  • 3.Hammen C Stress and depression. Annu Rev Clin Psychol 2005;1:293–319. [DOI] [PubMed] [Google Scholar]
  • 4.Paykel ES, Myers JK, Dienelt MN, Klerman GL, Lindenth JJ, Pepper MP. Life Events and Depression - a Controlled Study. Arch Gen Psychiat 1969;21(6):753. [DOI] [PubMed] [Google Scholar]
  • 5.Kessler RC, Berglund P, Demler O, Jin R, Koretz D, Merikangas KR, et al. The epidemiology of major depressive disorder: results from the National Comorbidity Survey Replication (NCS-R). JAMA 2003;289(23):3095–105. [DOI] [PubMed] [Google Scholar]
  • 6.Fava M Diagnosis and definition of treatment-resistant depression. Biological Psychiatry 2003;53(8):649–59. [DOI] [PubMed] [Google Scholar]
  • 7.Galea S, Nandi A, Vlahov D. The epidemiology of post-traumatic stress disorder after disasters. Epidemiol Rev 2005;27:78–91. [DOI] [PubMed] [Google Scholar]
  • 8.Chrousos GP, Gold PW. The concepts of stress and stress system disorders. Overview of physical and behavioral homeostasis. JAMA 1992;267(9):1244–52. [PubMed] [Google Scholar]
  • 9.de Kloet CS, Vermetten E, Geuze E, Kavelaars A, Heijnen CJ, Westenberg HG. Assessment of HPA-axis function in posttraumatic stress disorder: pharmacological and non-pharmacological challenge tests, a review. J Psychiatr Res 2006;40(6):550–67. [DOI] [PubMed] [Google Scholar]
  • 10.Heim C, Newport DJ, Mletzko T, Miller AH, Nemeroff CB. The link between childhood trauma and depression: insights from HPA axis studies in humans. Psychoneuroendocrinology 2008;33(6):693–710. [DOI] [PubMed] [Google Scholar]
  • 11.Yehuda R, Bierer LM, Andrew R, Schmeidler J, Seckl JR. Enduring effects of severe developmental adversity, including nutritional deprivation, on cortisol metabolism in aging Holocaust survivors. J Psychiatr Res 2009;43(9):877–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Dinan TG. Glucocorticoids and the genesis of depressive illness. A psychobiological mode. Br J Psychiatry 1994;164(3):365–71. [DOI] [PubMed] [Google Scholar]
  • 13.Griffiths BB, Hunter RG. Neuroepigenetics of stress. Neuroscience 2014;275:420–35. [DOI] [PubMed] [Google Scholar]
  • 14.Fardet L, Petersen I, Nazareth I. Suicidal behavior and severe neuropsychiatric disorders following glucocorticoid therapy in primary care. Am J Psychiatry 2012;169(5):491–7. [DOI] [PubMed] [Google Scholar]
  • 15.McGowan PO, Sasaki A, D’Alessio AC, Dymov S, Labonte B, Szyf M, et al. Epigenetic regulation of the glucocorticoid receptor in human brain associates with childhood abuse. Nat Neurosci 2009;12(3):342–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Davies MN, Krause L, Bell JT, Gao F, Ward KJ, Wu H, et al. Hypermethylation in the ZBTB20 gene is associated with major depressive disorder. Genome Biol 2014;15(4):R56. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Maddox SA, Kilaru V, Shin J, Jovanovic T, Almli LM, Dias BG, et al. Estrogen-dependent association of HDAC4 with fear in female mice and women with PTSD. Mol Psychiatry 2017;00:1–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Lee RS, Tamashiro KL, Yang X, Purcell RH, Harvey A, Willour VL, et al. Chronic corticosterone exposure increases expression and decreases deoxyribonucleic acid methylation of Fkbp5 in mice. Endocrinology 2010;151(9):4332–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Ewald ER, Wand GS, Seifuddin F, Yang X, Tamashiro KL, Potash JB, et al. Alterations in DNA methylation of Fkbp5 as a determinant of blood-brain correlation of glucocorticoid exposure. Psychoneuroendocrinology 2014;44:112–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Seifuddin F, Wand G, Cox O, Pirooznia M, Moody L, Yang X, et al. Genome-wide Methyl-Seq analysis of blood-brain targets of glucocorticoid exposure. Epigenetics 2017;8:637–52. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Wan ES, Qiu W, Baccarelli A, Carey VJ, Bacherman H, Rennard SI, et al. Systemic steroid exposure is associated with differential methylation in chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2012;186(12):1248–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Aryee MJ, Jaffe AE, Corrada-Bravo H, Ladd-Acosta C, Feinberg AP, Hansen KD, et al. Minfi: a flexible and comprehensive Bioconductor package for the analysis of Infinium DNA methylation microarrays. Bioinformatics 2014;30(10):1369-. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Fortin J, Triche TJ Jr, Hansen KD. Preprocessing, normalization and integration of the Illumina HumanMethylationEPIC array with minfi. Bioinformatics 2017;33(4):558–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Assenov Y, Muller F, Lutsik P, Walter J, Lengauer T, Bock C. Comprehensive analysis of DNA methylation data with RnBeads. Nat Methods 2014;11(11):1138–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Triche TJ Jr, Weisenberger DJ, Van Den Berg D, Laird PW, Siegmund KD. Low-level processing of Illumina Infinium DNA methylation beadarrays. Nucleic Acids Research 2013;41(7):e90. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Teschendorff AE, Marabita F, Lechner M, Bartlett T, Tegner J, Gomez-Cabrero D, et al. A beta-mixture quantile normalization method for correcting probe design bias in Illumina Infinium 450 k DNA methylation data. Bioinformatics 2013;29(2):189–96. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.R Core Team. R: A language and environment for statistical computing 2017. [Available from: https://www.R-project.org.
  • 28.Ritchie ME, Phipson B, Wu D, Hu Y, Law CW, Shi W, et al. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res 2015;43(7):e47. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Leek JT, Johnson WE, Parker HS, Jaffe AE, Storey JD. The sva package for removing batch effects and other unwanted variation in high-throughput experiments. Bioinformatics 2012;28(6):882–3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Phipson B, Maksimovic J, Oshlack A. missMethyl: an R package for analyzing data from Illumina’s HumanMethylation450 platform. Bioinformatics 2016;32(2):286–8. [DOI] [PubMed] [Google Scholar]
  • 31.Schizophrenia Working Group of the Psychiatric Genomics Consortium. Biological insights from 108 schizophrenia-associated genetic loci. Nature 2014;511(7510):421–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Stahl E, Forstner AJ, McQuillin A, Ripke S, Trubetskoy V, Mattheisen M, et al. Genome-wide association study identifies 30 Loci associated with bipolar disorder. bioRxiv 2017. [DOI] [PMC free article] [PubMed]
  • 33.Major Depressive Disorder Working Group, Wray NR, Sullivan PF. Genome-wide association analyses identify 44 risk variants and refine the genetic architecture of major depressive disorder. bioRxiv 2017. [DOI] [PMC free article] [PubMed]
  • 34.Menke A, Arloth J, Gerber M, Rex-Haffner M, Uhr M, Holsboer F, et al. Dexamethasone stimulated gene expression in peripheral blood indicates glucocorticoid-receptor hypersensitivity in job-related exhaustion. Psychoneuroendocrinology 2014;44:35–46. [DOI] [PubMed] [Google Scholar]
  • 35.Menke A, Arloth J, Putz B, Weber P, Klengel T, Mehta D, et al. Dexamethasone stimulated gene expression in peripheral blood is a sensitive marker for glucocorticoid receptor resistance in depressed patients. Neuropsychopharmacology 2012;37(6):1455–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Braun PR, Han S, Hing B, Nagahama Y, Gaul LN, Heinzman JT, et al. Genome-wide DNA methylation comparison between live human brain and peripheral tissues within individuals. Transl Psychiatry 2019;9(1):47. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Glad CA, Andersson-Assarsson JC, Berglund P, Bergthorsdottir R, Ragnarsson O, Johannsson G. Reduced DNA methylation and psychopathology following endogenous hypercortisolism - a genome-wide study. Sci Rep 2017;7:44445. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Zhong Z, Wan B, Qiu Y, Ni J, Tang W, Chen X, et al. Identification of a novel human zinc finger gene, ZNF438, with transcription inhibition activity. J Biochem Mol Biol 2007;40(4):517–24. [DOI] [PubMed] [Google Scholar]
  • 39.International Multiple Sclerosis Genetics C, Wellcome Trust Case Control C, Sawcer S, Hellenthal G, Pirinen M, Spencer CC, et al. Genetic risk and a primary role for cell-mediated immune mechanisms in multiple sclerosis. Nature 2011;476(7359):214–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Tang XH, Vivero M, Gudas LJ. Overexpression of CRABPI in suprabasal keratinocytes enhances the proliferation of epidermal basal keratinocytes in mouse skin topically treated with all-trans retinoic acid. Exp Cell Res 2008;314(1):38–51. [DOI] [PubMed] [Google Scholar]
  • 41.Fiorella PD, Giguere V, Napoli JL. Expression of cellular retinoic acid-binding protein (Type II) in Escherichia coli: Characterization and comparison to cellular retinoic acid-binding protein (Type I). J Biol Chem 1993:21545–52. [PubMed]
  • 42.Chen XN, Meng QY, Bao AM, Swaab DF, Wang GH, Zhou JN. The involvement of retinoic acid receptor-alpha in corticotropin-releasing hormone gene expression and affective disorders. Biol Psychiatry 2009;66(9):832–9. [DOI] [PubMed] [Google Scholar]
  • 43.Maden M, Holder N. The involvement of retinoic acid in the development of the vertebrate central nervous system. Dev Suppl 1991;Suppl:87–94. [PubMed] [Google Scholar]
  • 44.Aoto J, Nam CI, Poon MM, Ting P, Chen L. Synaptic signaling by all-trans retinoic acid in homeostatic synaptic plasticity. Neuron 2008;60(2):308–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Jacobs S, Lie DC, DeCicco KL, Shi Y, DeLuca LM, Gage FH, et al. Retinoic acid is required early during adult neurogenesis in the dentate gyrus. Proc Natl Acad Sci U S A 2006;103(10):3902–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Wang HL, Zhang Z, Hintze M, Chen L. Decrease in calcium concentration triggers neuronal retinoic acid synthesis during homeostatic synaptic plasticity. J Neurosci 2011;31(49):17764–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Cai L, Yan XB, Chen XN, Meng QY, Zhou JN. Chronic all-trans retinoic acid administration induced hyperactivity of HPA axis and behavioral changes in young rats. Eur Neuropsychopharmacol 2010;20(12):839–47. [DOI] [PubMed] [Google Scholar]
  • 48.Kubzansky LD, Adler GK. Aldosterone: a forgotten mediator of the relationship between psychological stress and heart disease. Neurosci Biobehav Rev 2010;34(1):80–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Hlavacova N, Jezova D. Chronic treatment with the mineralocorticoid hormone aldosterone results in increased anxiety-like behavior. Horm Behav 2008;54(1):90–7. [DOI] [PubMed] [Google Scholar]
  • 50.Weiser MJ, Handa RJ. Estrogen impairs glucocorticoid dependent negative feedback on the hypothalamic-pituitary-adrenal axis via estrogen receptor alpha within the hypothalamus. Neuroscience 2009;159(2):883–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Viau V, Meaney MJ. Variations in the hypothalamic-pituitary-adrenal response to stress during the estrous cycle in the rat. Endocrinology 1991;129(5):2503–11. [DOI] [PubMed] [Google Scholar]
  • 52.Lund TD, Munson DJ, Haldy ME, Handa RJ. Androgen inhibits, while oestrogen enhances, restraint-induced activation of neuropeptide neurones in the paraventricular nucleus of the hypothalamus. J Neuroendocrinol 2004;16(3):272–8. [DOI] [PubMed] [Google Scholar]
  • 53.Handa RJ, Kudwa AE, Donner NC, McGivern RF, Brown R. Central 5-alpha reduction of testosterone is required for testosterone’s inhibition of the hypothalamo-pituitary-adrenal axis response to restraint stress in adult male rats. Brain Res 2013;1529:74–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54.Handa RJ, Nunley KM, Lorens SA, Louie JP, McGivern RF, Bollnow MR. Androgen regulation of adrenocorticotropin and corticosterone secretion in the male rat following novelty and foot shock stressors. Physiol Behav 1994;55(1):117–24. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supp TableS1. Supplemental Table 1.

Dental procedure cohort differential methylation results of CpGs with FDR-adjusted p-values < 0.01 and differences > 10%.

Supp TableS2. Supplemental Table 2.

Psychiatric associated genes manually assembled from GWAS findings of major depressive disorder, schizophrenia, and bipolar disorder.

Supp TableS3. Supplemental Table 3.

Methylation and degree of correlation between buccal and brain DNA methylation of pre- and post-dexamethasone buccal samples of CpGs within psychiatric disorder risk genes which were significant at FDR-adjusted p-value < 0.05 and with differences greater than 10% before and after dexamethasone exposure in the dental procedure cohort.

Supp TableS4. Supplemental Table 4.

The characteristics of CpGs in the dental cohort that are in differentially expressed genes in blood from a separate study.

Supp TableS5. Supplemental Table 5.

Top 100 CpGs differentially methylated in the neurosurgery cohort within each peripheral tissue and the brain.

Supp TableS6. Supplemental Table 6.

CpGs that are differentially methylated in both the dental procedure cohort (FDR-adjusted p-value < 0.01 and a difference > 10%) and that are nominally significant in the neurosurgery cohort of either blood, buccal, saliva, or brain tissues (p-value < 0.05 and a difference > 10%).

RESOURCES