DNA methylation changes in a pharmaco-epigenomic EWAS in depression: comparing fixed and response-guided dosing paradigms for ketamine in the KADS trial

DNA methylation changes in a pharmaco-epigenomic EWAS in depression: comparing fixed and response-guided dosing paradigms for ketamine in the KADS trial | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article DNA methylation changes in a pharmaco-epigenomic EWAS in depression: comparing fixed and response-guided dosing paradigms for ketamine in the KADS trial Evelien Van Assche, Christa Hohoff, Victoria Statz, Sophia Wissing, and 23 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6778101/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 13 You are reading this latest preprint version Abstract DNA methylation is a dynamic biomarker suited to investigate processes, e.g., treatment response. We investigate DNA methylation changes in different dosing paradigms for ketamine (s.c.) vs. midazolam (s.c.) in treatment resistant depressed patients (the Ketamine for Adult Depression Study – KADS). After a 4-week randomized controlled trial (RCT) DNA methylation data (Illumina Infinium MethylationEPIC 850k BeadChip) were available for 87 patients, with 76 having DNA methylation data at baseline too. Patients received either ketamine or midazolam. Initially dosing was fix, however, following a protocol amendment, newly recruited patients received the clinically more effective response-guided ‘flexible’ dosing. We performed cross-sectional and paired longitudinal DNA methylation analyses, the latter with focus on differentially methylated regions (DMR), comparing treatment with ketamine and midazolam, as well as fixed and flexible dosing. We used R-packages RnBeads and comb-p for quality control and statistical analyses, and DMR identification, respectively. P-values were False-discovery-rate (FDR)-corrected (Benjamini Hochberg). Only the response-guided cohort in the ketamine vs. midazolam comparison returned one epigenome-wide significant CpG (cg15945600; PDCD6; p-valFDR < 0.05). Flexible dosing (midazolam + ketamine) vs. fixed dosing (midazolam + ketamine) returned a suggestive hit (cg20023762, NEDD9; p-valFDR < 0.10). The flexibly dosed ketamine condition returned the only statistically significant DMR in the longitudinal analyses (p-valFDR = 0.0045; CAPS2/GLIPR1L2). The clinically more effective response-guided paradigm has a correlate at the DNA methylation level with most pronounced DNA methylation changes in the flexibly dosed ketamine group. Pharmaco-epigenomics in Psychiatry, as a growing field, facilitates interpretation of DNA-methylation dynamics for treatment response in depression. Biological sciences/Molecular biology Health sciences/Biomarkers/Prognostic markers Figures Figure 1 Figure 2 1. Introduction Ketamine has established its place as a rapid-acting antidepressant at sub-anesthetic doses ( 1 ). Varying approaches to dosing have been used. The KADS-trial (Ketamine for Adult Depression Study) was a randomized controlled clinical trial (RCT) designed to compare ketamine with midazolam, as an active control condition ( 2 ). The study initially used a fixed dose of ketamine and midazolam (‘Cohort 1’), though after lack of remitters was observed in either group for the first 51 participants, a revision of the protocol occurred, with the introduction of flexible response-guided dosing ketamine and midazolam (‘Cohort 2’) ( 2 ). The study showed that flexibly dosed ketamine was more efficacious than midazolam in reducing depressive symptoms in patients with treatment resistant depression (TRD) ( 2 ). However, the molecular underpinnings of the relationship between clinical response and treatment paradigm are poorly understood. The two dosing cohorts in the KADS trial provide an excellent opportunity to investigate epigenetic characteristics of the dynamic response-guided pharmacological strategy for ketamine as an antidepressant. DNA methylation and epigenetic markers are biomarkers that are sensitive to environmental and pharmacological influences ( 3 ). Epigenome-wide DNA methylation may reflect changes that occur in the individual over the course of treatment, be it pharmacotherapy ( 4 , 5 ) or psychotherapy for depression ( 6 ). These changes are also reflected in gene expression, one of the mechanisms through which DNA methylation affects the individual ( 7 ). Prior research suggests that ketamine acts through altered DNA methylation profiles ( 8 ). Reports in rodent models have supported this hypothesis for the BDNF gene, which codes for the brain-derived neurotrophic factor (BDNF) protein, in models for Post-traumatic stress disorder (PTSD) ( 9 ). There are scant longitudinal DNA methylation data that allow an in-depth investigation of the effect of different dosing paradigms for an antidepressant treatment, as well as longitudinal epigenome-wide studies on ketamine for depression. DNA methylation has been shown to change alongside treatment in longitudinal designs and randomized controlled trials, in line with treatment response ( 6 ), though without the opportunity to investigate in-depth the impact of different dosing paradigms on DNA methylation. Therefore, we focus on the clinically observed benefit of the response-guided ‘flexible’ dosing paradigm and expect that DNA methylation changes to follow this pattern with a more pronounced molecular response at the epigenomic level for the flexibly dosed ketamine cohort. To investigate the correlation between treatment paradigm and DNA methylation in depth, it is advantageous to use longitudinal data, such as in our sample. The characteristics of a randomized controlled trial assist in controlling for phenotypic heterogeneity between groups, as is typical for depression. In addition, it is a rare opportunity to investigate two dosing scenarios of the same medication within a single clinical trial. Hence, we aimed to test the hypothesis that the clinical benefit of the flexible response-guided treatment paradigm is reflected in DNA methylation changes accordingly: we expect more pronounced DNA methylation changes in the flexibly dosed cohorts, particularly in individuals from the ketamine-group. Our research presents the first epigenome-wide association study (EWAS) for ketamine. We tackle the pharmaco-epigenomic question at hand from an innovative personalized dosing perspective, comparing the impact of dosing and treatment modalities on DNA methylation changes, with a combined cross-sectional and longitudinal approach. We investigated DNA methylation at the CpG dinucleotide and differentially methylated regions (DMR) with the hypothesis that response-guided dosing, which optimizes dosage based on individual response, shows a stronger effect at the molecular level, also independent of antidepressant treatment response per se . Hence, we expect to find the largest effect of treatment on DNA methylation at end of RCT treatment for the flexible dosed ketamine group. The combination of cross-sectional analysis at end of RCT treatment and longitudinal analyses spanning the treatment course allow for an in-depth exploration of the response of DNA methylation to dosing or treatment specific effects. 2. Methods 2.1. Sample description The KADS sample (Ketamine for Adult Depression Study) consisted of 184 individual patients with treatment resistant major depressive disorder. The study was a randomized controlled trial (RCT) with individuals receiving subcutaneous ketamine or midazolam as an active control condition. Doses were either administered according to a fixed paradigm (ketamine: 0.5 mg/kg; midazolam 0.025 mg/kg, identical volumes of injection; ‘Cohort 1’) or a flexible paradigm allowing for dose increases based on clinical assessment (ketamine: 0.5–0.9 mg/kg; midazolam 0.025–0.045 mg/kg; ‘Cohort 2’). For a detailed description of the KADS trial see Loo et al. ( 2 ). Along with phenotypic information, whole blood was collected at five time-points during the randomized and the open label extension phases. The randomized phase took place between baseline (T1) and end of RCT treatment (T3), 4 weeks apart, which is also the focus of this analysis. Blood collections T4 and T5 occurred during the open-label phase. In total, 388 blood samples were analyzed over the five time-points (T1-T5; Fig. S1 ). Quality control (QC) steps were performed for all 388 samples and resulted in a final dataset with 379 samples from 113 unique participants (see supplementary data). 87 participants had DNA methylation data available for end of RCT treatment (T3), and 76 of these also had DNA methylation data available for baseline (T1), allowing for longitudinal analyses. From the sample of 87 individuals, the mean age was 48.3 years and 32.2% were women. 41 individuals received ketamine, 46 received midazolam, with 17 and 19 individuals in the flexible cohort, i.e., higher average doses, respectively (Fig. S1 ). Mean MADRS score (Montgomery-Åsberg Rating Scale for Depression), a rating scale for depression severity ( 10 ), at baseline was 29.3 points ( SD = 5.9). The study was approved by the Sydney Local Health District (RPAH Zone) Human Research Ethics Committee (Australia; X16-0146 and HREC/16/RPAH/168) and the Southern Health and Disability Ethics Committee (New Zealand; 16/STH/104). All participants provided written informed consent. Trial registration: ACTRN12616001096448 at www.anzctr.org.au . 2.2. Epigenome-wide DNA methylation analysis DNA methylation data (Illumina Infinium MethylationEPIC 850k BeadChip) were available from 388 samples of 113 unique individuals over multiple time-points (Fig. S1 ). Though we solely focus on T1 (baseline) and T3 (end of RCT treatment), QC, pre-processing, and normalization were performed for all five time-points combined for methodological purposes. Following QC steps, 9 samples were discarded, resulting in 379 available samples. After finalizing QC, we extracted the samples available for the time-points of interest in line with our analysis strategy with cross-sectional and longitudinal analyses (i.e., 76 for T1 and 87 for T3). DNA was isolated from whole blood samples using standard procedures (QIAamp DNA Blood Midi-Kit, Qiagen, Hilden, Germany) followed by purification (Amicon 0,5ml 3K; Merck/Millipore, Darmstadt, Germany) and pipetted on 96-well plates for chip-based analyses. Bisulfite conversion and handling of the DNA methylation chips were performed at the Life&Brain Institute (Bonn, Germany). Samples were randomized on plates and chips based on patient’s sex and age. Samples from the same patient were analysed on the same chip to minimize batch-effects for within-individual analyses. Following analysis on HiScan array scanning systems (Illumina, San Diego, CA), data were transferred as .idat files. The further processing and quality control of the DNA methylation data was performed using R (version 4.4.2) and the ‘RnBeads’ pipeline (Package RnBeads 2.0 ( 11 , 12 )). Normalization was performed using “wm.dasen” as embedded in the RnBeads package. A detailed description of the QC steps can be found in the supplementary data (Table S1 ). 2.3. Statistical analyses 2.3.1. Cross-sectional analyses Differential methylation analyses for end of RCT treatment were performed with ‘limma’, as also embedded in the RnBeads package. In total, four cross-sectional differential methylation analyses were performed: ( 1 ) an overall analysis comparing ketamine vs. midazolam for the whole sample, ( 2 ) comparing flexibly dosed ketamine and flexibly dosed midazolam, ( 3 ) comparing fixed dose ketamine and fixed dose midazolam, and ( 4 ) comparing fixed versus flexible dosing, combined across both treatment conditions. Cross-sectional analyses were corrected for biological sex, age, site where the blood was collected, six methylation-based cell-type estimates, as well as ancestry, represented by the two first principal components of genome-wide single nucleotide polymorphism (SNP) data ( 13 ). The overall analysis was also corrected for dosing cohort (i.e., fixed or flexible), which was added as an additional covariate. Cell types were estimated as suggested by Salas et al. ( 14 ) with the GSE110554 6-cell-types reference dataset using the Houseman method ( 15 ) to estimate the most important cell type fractions for our samples: neutrophils, monocytes, B-lymphocytes, natural killer cells, and CD4 + and CD8 + T-cells. As we had body weight available, but no height, we did not have body mass index available and did not include weight in the model. Technical confounding factors (e.g., batch effects) were addressed using surrogate variables. For all four cross-sectional analyses the overall genomic inflation factor ( λ ; χ 2 -test) was between 0.97 and 1.01. Details on the surrogate variables and QQ-plots can be found in the supplementary data (Fig. S3-S6). 2.3.2. Paired longitudinal analyses We extended our analysis strategy with a paired longitudinal analysis comparing DNA methylation at baseline with end of RCT treatment. Analogous with the cross-sectional analyses, the paired longitudinal analyses started from the whole sample of individuals having DNA methylation available at baseline and end of RCT treatment ( N = 76), comparing DNA methylation at baseline with DNA methylation at end of RCT treatment. In a second step, analyses were stratified by treatment: ketamine-only and midazolam-only. Finally, analyses were stratified by treatment and dosing paradigm: ketamine flexible ( N = 17 pairs), ketamine fixed ( N = 16 pairs), and midazolam flexible ( N = 19 pairs), and midazolam fixed ( N = 24 pairs). For the paired longitudinal analysis, the Welch’s t-test was used in the RnBeads package. Due to the small numbers, results of the paired analyses were only interpreted as differentially methylated regions (DMRs). DMRs were calculated using the comb-p algorithm embedded in the Enmix R-package ( 16 ). Settings for the DMR identification were: dist.cutoff = 1000, bin.size = 500, seed = 0.01. Finally, in line with existing evidence on the polygenic nature of psychiatric disorders, we assumed a polygenic involvement of multiple small effects not reaching statistical significance, though relevant for the phenotype at hand, as shown through polygenic risk scores ( 17 ). Therefore, in analogy of enrichment tests ( 18 ), we compared the count of CpGs with p < 0.05 for each of the four treatment conditions as a proxy of the overall impact of treatment on longitudinal DNA methylation changes. The comparison of CpG-proportions with p 0.05 between treatment conditions (cf. enrichment analysis) was performed using a χ 2 -test. For all the longitudinal analyses the overall genomic inflation factor ( λ ; χ 2 -test) was between 0.89 and 1.00 (supplementary fig. S7-S13). In line with the available literature we considered a p -value below 9.42 × 10 − 8 to be significant for the EPIC array ( 19 ). For the DMR-analysis false-discovery-rate (FDR)-corrected p -values (Benjamini-Hochberg) below 0.05 were considered statistically significant. For the interpretation of results, the following online databases were used: UCSC genome browser ( 20 ), Alliance of Genome Resources ( 21 ), and GeneCards®( 22 ). Overall comparability of the midazolam and ketamine cohorts regarding confounding factors at baseline was tested using ANOVA and Fisher’s exact test. Secondary explorations and post-hoc sensitivity analyses of the significant results were performed using ANOVA or by calculating Pearson’s correlation coefficient. 3. Results 3.1. Sample characteristics Baseline characteristics of the end of RCT treatment-sample ( N = 87) showed comparable groups for each of both treatment conditions. No significant group differences were found for age at baseline ( F (1,85) = 2.48; p = 0.12), gender distribution (% female; p = 0.25), mean MADRS at baseline ( F (1,85) = 1.59; p = 0.21), weight at baseline ( F (1,85) = 0.14; p = 0.71), and years of education ( F (1,85) = 0.026; p = 0.87). 3.2. Cross-sectional epigenome-wide analyses at end of RCT treatment. 3.2.1. Overall analysis of Ketamine vs. Midazolam The overall analysis comparing all individuals treated with ketamine vs. individuals treated with midazolam independent of dosing scheme did not return any epigenome wide significant CpGs. The CpG with the smallest p -value was cg11159519, located in the KCNH1 gene body: a potassium channel gene (Chr1: 210857380; South Shore; p = 2.38x10 − 6 ; ketamine group relatively hypomethylated; DNA methylation difference = 0.0041). Other CpGs in the top 10 were in the vicinity of the following genes: MAML2 , HSBP1 , CRTC3 , NMU , PER1 , VAC14 , ATE1 , and NCKAP5 /RN7SKP154 pseudogene (see supplementary Table S2). 3.2.2. Flexible doses of Ketamine vs. flexible doses of Midazolam In line with the clinical observation, the stratified analysis comparing flexibly dosed ketamine with flexibly dosed midazolam showed one epigenome-wide significant CpG: cg15945600 ( p = 7.28x10 − 8 , FDR-corrected p = 0.049; Fig. 1 a). This CpG is in the gene-body of the PDCD6 gene. For this CpG the ketamine group is relatively hypomethylated with a DNA methylation difference of 2.3% (table 1). A secondary exploration of this result showed distinct DNA methylation values for each of the treatment cohorts ( F (3,83) = 5.10; p = 0.0028; Fig. 1 b), which was not present at baseline ( F (3,91) = 1.64; p = 0.19). 3.2.3. Fixed dose Ketamine vs. fixed dose Midazolam In line with the clinical observation that the antidepressant effect was more pronounced in the flexible dosing cohort ( 2 ), the analyses comparing fixed dosing groups showed no significant epigenetic differences. The most significant CpG was cg04584009 ( p = 3.09x10 − 6 ), located in a CpG island in the REXO1 gene-body. REXO1 has been associated with brain function in prior research, in particular in the context of schizophrenia ( 23 ). For this CpG the ketamine group was hypermethylated (DNA methylation difference: 1.3%). Within the top 10 of CpGs from this analysis, one region was represented three times (table S3). The region in question is particularly linked to regulatory aspects, e.g., long non-coding RNA and cis-regulatory elements (see also supplementary fig. S15). 3.2.4. Fixed doses vs. flexible doses Ketamine and Midazolam combined. In line with our hypothesis, we compared the lower fixed dose cohort 1 with the higher flexibly dosed cohort 2. For this analysis, one CpG came close to the pre-defined significance threshold (p = 1.11x10 − 7 ; FDR-corrected p = 0.074). This CpG, cg20023762, is located in the NEDD9 gene body, a gene both linked to depression and pharmacogenomics ( 24 , 25 ). The top 10 of CpGs for this analysis are in Table S4. 3.3. Paired longitudinal analyses: baseline versus end of RCT treatment. The whole-sample longitudinal analysis comparing DNA methylation at end of RCT treatment with baseline showed no significant differentially methylated region (DMR). This was also the case for the longitudinal analysis for the individuals treated with ketamine, and those treated with midazolam. A sensitivity analysis with the most-significant CpG of the whole-sample longitudinal analysis (cg10080497, p = 1.68x10 − 6 ), showed a significant correlation between change in DNA methylation over time and change in MADRS for the ketamine group only ( r ( 31 ) = 0.36; p = 0.042; supplementary fig. S14). The only analysis returning one significant DMR was the longitudinal analysis for patients receiving ketamine in flexible higher doses over the course of 4 weeks. This DMR consists of two CpGs (Chr 12; 75784541–75784618; p -value (FDR-corrected) = 0.0045) in the CAPS2 gene-region and close to the GLIPR1L2 promotor-linked CpG island (supplementary figure S16). Both CpGs showed a decrease in DNA methylation over time (cg20291222: mean difference over time = − 0.010; cg14292619: mean difference over time = -0.0062). The same analyses in each of the other sub-groups, i.e., ketamine fixed doses and both midazolam cohorts, did not return any significant DMR. Overall, the longitudinal cohort for ketamine in flexible doses, returned 32.305 CpGs with p -value < 0.05. This was the case for 29.947 CpGs for the ketamine fixed doses sample, and 28.666 CpGs and 28.680 CpGs for the fixed and flexible doses of Midazolam, respectively. Between both ketamine groups, p -value proportions were significantly different (χ 2 (1, N = 667 027) = 93.609, p 0.95). This was again consistent for the comparison of both cohorts with fixed doses (χ 2 (1, N = 667 027) = 29.237, p < 0.001) and both cohorts with flexible, i.e., higher doses (χ 2 (1, N = 667 027) = 225.67, p < 0.001). 4. Discussion Reflecting results from the randomized phase of the KADS clinical trial ( 2 ), our hypothesis was that the observed clinical response for the flexible, response-guided dose condition would also be evident at the level of DNA methylation. We assumed that response-guided dosing which optimizes dosing on an individual basis, would show a stronger effect at the molecular level, also independent of antidepressant treatment response per se . By combining cross-sectional and paired longitudinal analyses both on the whole sample and in a stratified manner, we demonstrated support for this hypothesis. In our cross-sectional analysis at the end of RCT treatment, we found that, despite the relatively small sample size, on comparing flexible high dose ketamine with flexible high dose midazolam there was an epigenome-wide significant hit. Also, in line with our hypothesis that the flexible dosed cohort would show a stronger effect at the DNA methylation level, the fixed-dose comparison did not return any significant signal and that the combined flexible-dose comparison independent of the treatment, suggested a potentially relevant CpG, notwithstanding the epigenome-wide cut-off or the FDR-correction. Our cross-sectional analyses hint at some genes and systems that may be involved in ketamine’s antidepressant action. The PDCD6 -gene is a calcium-binding protein, suggested as being involved in the regulation of cytoskeletal dynamics and innate immune responses ( 26 ), but also apoptosis ( 26 ). PDCD6 has also been suggested to play a role in the PI3K/mTOR/p70S6K pathway by interacting with VEGFR-2 ( 27 ). The mTor system has previously been linked to several psychiatric treatments, ketamine in particular ( 28 – 30 ). A recent genome-wide association study (GWAS) on esketamine and treatment response also highlighted a connection with the immune-system through the IRAK3 -gene (Interleukin 1 Receptor Associated Kinase 3) ( 31 ). The gene NEDD9 is another gene that has been cautiously highlighted by our cross-sectional analyses. NEDD9 is a stimulator of cell differentiation and has been shown to affect the glutamate pathway and depressive-like behavior in rodents ( 32 ). NEDD9 promoted neurite outgrowth of neuronal cells after ischemic damage ( 33 ). In the context of pharmacogenetics, NEDD9 was suggested as a selective serotonin reuptake inhibitor (SSRI) target for development and modulating synaptic connectivity and emotional behaviour in mice ( 34 ). The cross-sectional analysis including the full sample and comparing treatment with ketamine versus midazolam returned some CpGs located in or in the vicinity of biologically plausible genes too, particularly in the context of pharmaco-epigenomics. CpGs in the top 10 hinted at genes linked to potassium signaling ( KCNH1 ); often linked to epilepsy ( 35 )– for which midazolam is an effective treatment ( 36 ). Furthermore, genes linked to immune-mediated inflammation were also represented in this top 10 such as NMU (‘Neuromedin U’) with links to schizophrenia ( 37 ) and suggested as a treatment target for other psychiatric phenotypes ( 38 , 39 ), as well as genes known to play a role in depression-related white-matter changes and the circadian rhythm in depression (e.g., PER1 and VAC14 ; 42–44 ) . In addition, MAML2 and CRTC3 both act in the same CREB-pathway which is linked to synaptic plasticity and has already been suggested as a therapeutic target for depression ( 43 , 44 ). NCKAP5 has been linked to synaptic plasticity as well in bipolar disorder ( 45 ). However, none of these CpGs reached statistical significance. The QQ-plot suggested low statistical power, which might be linked to the relatively small sample size. However, we also expect a dilution effect in the overall sample, as flexible higher and fixed lower dosing schemes are combined, given the lack of clinical effect for the fixed doses ( 2 ). Therefore, the stratification probably led to more homogeneous groups, which might have been beneficial for the statistical power, despite the smaller sample size for the stratified analyses. Our paired longitudinal analysis also supported our hypothesis and highlights the connection between response-guided dosing and changes at the molecular level, as well as the benefit of stratification, by returning a significant DMR for the flexibly dosed ketamine group only. Despite the small sample size, the number of CpGs with a p -value below 0.05 was also substantially, and significantly, higher than for each of the other conditions. The CAPS2 gene is also calcium-dependent and involved in promoting the BDNF-pathway ( 46 , 47 ). The CAPS2 gene-expression is decreased in post-mortem samples of depressive patients that died from suicide ( 48 ), and may be upregulated in the context of symptom remission following a cognitive psychotherapeutic intervention ( 7 ). It also has been linked to despair-like behaviour in rodents ( 48 ). The observation that both CpGs that contribute to the DMR show decreased DNA methylation over the course of treatment is in line with these prior results. GLIPR1L2 as a gene and DMR has already been linked to depression in a prior epigenome-wide association study with elderly twins ( 49 ). Overall, the two genes that showed a link with our hypothesis that the response-guided dosing condition is linked to more prominent DNA methylation changes, PDCD6 and CASP2 , play a central role in pathways that have been linked to ketamine’s antidepressant effect in previous research. Both have been shown to play a role in the mTOR/VEGF/BDNF-connection, which has been proposed as a key-element in ketamine’s antidepressant mode of action. Therefore, these data suggest that DNA methylation is a dynamic biomarker that reflects the clinically observed relationship between dosing paradigm and response, both in the context of ketamine as an active antidepressant treatment, as well as independent of the therapeutic agent. Despite the availability of longitudinal analyses, causality remains unclear. The correlation suggested by our analysis does not answer the question of whether DNA methylation precedes recovery and is part of the recovery process or does recovery itself lead to changes in DNA methylation. As recovery itself is highly complex involving multiple biological systems, both might also be parallel processes, that occur independently, driven by a shared impulse, such as immunological changes ( 50 , 51 ). Our study has some limitations. One limitation is the relatively small sample size as mentioned above. We expect the relationship between DNA-methylation and response-guided dosing to be more complex than we could show with our current investigation. We also expect a polygenic contribution of DNA methylation on the relationship between dosing paradigm and response at a molecular and clinical level. A larger sample size could contribute to the signal by exposing multiple smaller effects. Hence, our sample size does not allow a more in-depth analysis or thorough mechanistic exploration. Nonetheless, by tackling the question at hand from multiple angles, we can argue that DNA methylation changes occur in parallel to response-guided dosing. The trade-off between the sample size and sample heterogeneity is a challenge for both depression research and the field of DNA methylation. With the prior clinical observation ( 2 ), we chose to analyse stratified by pharmacological modality in favour of homogeneity. Combining treatment categories, e.g., fixed and flexible doses, could have possibly diluted the effect of pharmacotherapy on DNA methylation. Reducing heterogeneity and confounding effects is another limitation, in part inherent to DNA methylation research. It is hard to control sensitivity to environment and treatment-related exposures with real-live patients suffering from depression. By restricting ourselves to biological sex, age, site of data collection, estimated cell-types and ancestry, we anticipated the most relevant variability relevant to our research question, without overfitting the model. Body mass index and smoking are often associated with depression, as well as with the DNA methylation and metabolic fitness, but these were unavailable in our study dataset and hence these may be unmeasured confounding factors. Hence, we did not include these variables in the model. However, it is likely that the highly regulated format of an RCT provides an optimal setting, as the study design implies that confounding factors are randomized too. Though, smaller sample sizes are more prone to sampling biases, these can best be overcome by independent replication to test the reliability of our results in different patient groups. We are not aware of any similar ketamine samples to date suitable for independent replication. We hope that this study contributes to the motivation to investigate the dosing of ketamine and its relationship with the level of DNA methylation. Lastly, as shown in the fixed-dose analysis, although not statistically significant by our pre-set cut-off, our results returned many CpGs that are expected to have a regulatory function. In our interpretation we focused on the genes that were highlighted by our analyses. However, we want to stress that, particularly in a treatment context the role of DNA methylation as an epigenetic marker is most likely regulatory in nature, as it is expected to play an orchestrating role ( 52 , 53 ). Many questions remain unanswered in our model of response-guided dosing and its response at the DNA methylation level. Overall, these data show that treatment with ketamine affects DNA methylation differently than the active control condition of midazolam. We could also show that flexible dosing of either medication can affect DNA methylation. The treatment-stratified longitudinal analyses also showed a significant DMR exclusively for the flexibly dosed ketamine group. For this condition more CpGs had a p -value below 0.05 than for each of the other conditions. Our research argues in favour of dose-dependent DNA methylation changes for ketamine over time in the treatment of depression. More research is needed to replicate these findings and investigate the matter of causality: do DNA methylation changes contribute to recovery or are they the result of improved depression symptomatology? By investigating DNA methylation changes in a response-guided dosing paradigm, this study takes an initial step toward a better understanding of the relationship between epigenetic modulation and ketamine treatment success. Declarations COI: In the past 36 months, N.G. has received speaker's bureau honoraria from Servier Laboratories, Janssen and Lundbeck, and served on advisory boards for Servier Laboratories, Esia, Seqirus and Lundbeck. D.B. is a director and part-owner of Neurotrials Victoria Pty Ltd, trading as Neurocentrix and Neurocentrix TMS Pty Ltd; he serves on the advisory board for Eli Lilly and Janssen, and is currently supported by grant funding from Praxis, Janssen, Eli Lilly, Biogen and NHMRC; he has served on speaker panels for Servier, Janssen and Eli Lilly in the past 12 months; he is an investigator on the Janssen Quality of Life Esketamine study. B.T.B. has received grants and served as consultant, advisor or CME speaker for AstraZeneca, Bristol-Myers Squibb, Janssen, Lundbeck, Otsuka, Servier, the NHMRC, the Fay Fuller Foundation and the James and Diana Ramsay Foundation. In the past 3 years, P.B.F. has received equipment for research from Neurosoft, Nexstim and Brainsway Ltd; he has served on scientific advisory boards for Magstim and LivaNova and received speaker fees from Otsuka; he is a founder and board member for TMS Clinics Australia and Resonance Therapeutics. Within the last 36 months, P.G. has attended a Janssen New Zealand advisory board, and is named on a patent for a controlled release ketamine tablet developed by Douglas Pharmaceuticals. In the past 36 months, D.M. has received research consulting fees from Douglas Pharmaceuticals for a clinical trial involving ketamine. P.B.M. has received remuneration from Janssen (Australia) and Sanofi (Hangzhou) for lectures or advisory board membership within the past 3 years. M.B. has received honoraria from EPA Warsaw, Lundbeck, Controversias Barcelona, Servier, Medisquire, HealthEd, ANZJP, European Psychiatric Association, Janssen, Medplan, Milken Institute, Abbott India, ASCP, Allori for Eisai, Otsuka, St Bio Pharma and Sandoz in the past 3 years. G.C. has received educational and travel support from Servier, Astra Zeneca, Otsuka Australia, Merck Sharp & Dohme and Janssen-Cilag in the past 5 years; he also served on an advisory board for the AFFINITY trial. A.A.S. is a director of the Australian Medicines Handbook Pty Ltd (unpaid) and has received funding support from the Australian and New Zealand College of Anaesthetists to investigate ketamine for chronic postsurgical pain. S.H. has received speaker and consultancy fees from Janssen and Servier and served on advisory boards for Janssen and Lundbeck. C.K.L. is on the Clinical Advisory Board for Douglas Pharmaceuticals and has received fees for the following: Janssen Cilag advisory board, Douglas Pharmaceuticals advisory board. E.V.A., Ch.H., V.S., S.M.W., N.T.M, S.N., S.S., A.R., D.H.-P., A.A., V.D., M.L.H., C.M., M.L.C.: None. Acknowledgements: Evelien Van Assche thanks the Faculty of Medicine of the University of Münster for the research support (flexible Forschungszeit). We sincerely thank all the participants who were involved in the study, and all investigators and staff at the study centers. The study was funded by a competitive research grant from the Australian National Health and Medical Research Council (APP1105089). M.B. is supported by a NHMRC Leadership 3 Investigator grant (2017131). C.K.L. is supported by a NHMRC Leadership Investigator grant (1195651). Supplementary information: Supplementary information is available at MP’s website. The supplementary information includes a more detailed sample description and outline of the RCT, as well as details on the QC. It also includes information on sensitivity analyses, as well as additional information to interpret the reported results. Tables that provide more details on reported results are also included. The supplementary information consists of one .pdf file. References McIntyre RS, Rosenblat JD, Nemeroff CB, Sanacora G, Murrough JW, Berk M, et al. Synthesizing the Evidence for Ketamine and Esketamine in Treatment-Resistant Depression: An International Expert Opinion on the Available Evidence and Implementation. AJP. 2021;178(5):383–99. Loo C, Glozier N, Barton D, Baune BT, Mills NT, Fitzgerald P, et al. Efficacy and safety of a 4-week course of repeated subcutaneous ketamine injections for treatment-resistant depression (KADS study): randomised double-blind active-controlled trial. Br J Psychiatry. 223(6):533–41. Moore LD, Le T, Fan G. DNA Methylation and Its Basic Function. Neuropsychopharmacol. 2013;38(1):23–38. Arčan IŠ, Kouter K, Paska AV. Depressive disorder and antidepressants from an epigenetic point of view. World Journal of Psychiatry. 2022;12(9):1150. Ju C, Fiori LM, Belzeaux R, Theroux JF, Chen GG, Aouabed Z, et al. Integrated genome-wide methylation and expression analyses reveal functional predictors of response to antidepressants. Transl Psychiatry. 2019;9(1):1–12. Van Assche E, Hohoff C, Zang J, Knight MJ, Baune BT. Longitudinal early epigenomic signatures inform molecular paths of therapy response and remission in depressed patients. Front Mol Neurosci. 2023;16:1223216. Zang JCS, Hohoff C, Van Assche E, Lange P, Kraft M, Sandmann S, et al. Immune gene co-expression signatures implicated in occurence and persistence of cognitive dysfunction in depression. Prog Neuropsychopharmacol Biol Psychiatry. 2023;127:110826. Flynn LT, Gao WJ. DNA methylation and the opposing NMDAR dysfunction in schizophrenia and major depression disorders: a converging model for the therapeutic effects of psychedelic compounds in the treatment of psychiatric illness. Mol Psychiatry. 2023;28(11):4553–67. Ju LS, Yang JJ, Lei L, Xia JY, Luo D, Ji MH, et al. The Combination of Long-term Ketamine and Extinction Training Contributes to Fear Erasure by Bdnf Methylation. Front Cell Neurosci. 2017;11. Available from: https://www.frontiersin.org/journals/cellular-neuroscience/articles/ 10.3389/fncel.2017.00100/full Montgomery SA, Åsberg M. A New Depression Scale Designed to be Sensitive to Change. The British Journal of Psychiatry. 1979;134(4):382–9. Assenov Y, Müller F, Lutsik P, Walter J, Lengauer T, Bock C. Comprehensive analysis of DNA methylation data with RnBeads. Nature Methods. 2014;11(11):1138–40. RnBeads 2.0: comprehensive analysis of DNA methylation data | Genome Biology | Full Text [Internet]. [cited 2023 Apr 13]. Available from: https://genomebiology.biomedcentral.com/articles/ 10.1186/s13059-019-1664-9 Peterson RE, Kuchenbaecker K, Walters RK, Chen CY, Popejoy AB, Periyasamy S, et al. Genome-wide Association Studies in Ancestrally Diverse Populations: Opportunities, Methods, Pitfalls, and Recommendations. Cell. 2019;179(3):589–603. Salas LA, Koestler DC, Butler RA, Hansen HM, Wiencke JK, Kelsey KT, et al. An optimized library for reference-based deconvolution of whole-blood biospecimens assayed using the Illumina HumanMethylationEPIC BeadArray. Genome Biology. 2018;19(1):64. Houseman EA, Molitor J, Marsit CJ. Reference-free cell mixture adjustments in analysis of DNA methylation data. Bioinformatics (Oxford, England). 2014;30(10):1431–9. Xu Z, Xie C, Taylor JA, Niu L. ipDMR: identification of differentially methylated regions with interval P-values. Bioinformatics. 2021;37(5):711–3. Purcell SM, Wray NR, Stone JL, Visscher PM, O’Donovan MC, Sullivan PF, et al. Common polygenic variation contributes to risk of schizophrenia and bipolar disorder. Nature. 2009;460(7256):748–52. Schork AJ, Thompson WK, Pham P, Torkamani A, Roddey JC, Sullivan PF, et al. All SNPs are not created equal: genome-wide association studies reveal a consistent pattern of enrichment among functionally annotated SNPs. Gibson G, editor. PLoS genetics. 2013;9(4):e1003449–e1003449. Campagna MP, Xavier A, Lechner-Scott J, Maltby V, Scott RJ, Butzkueven H, et al. Epigenome-wide association studies: current knowledge, strategies and recommendations. Clin Epigenet. 2021;13(1):214. Lee BT, Barber GP, Benet-Pagès A, Casper J, Clawson H, Diekhans M, et al. The UCSC Genome Browser database: 2022 update. Nucleic Acids Research. 2022;50(D1):D1115–22. Kishore R, Arnaboldi V, Van Slyke CE, Chan J, Nash RS, Urbano JM, et al. Automated generation of gene summaries at the Alliance of Genome Resources. Database. 2020;2020:baaa037. Safran M, Rosen N, Twik M, BarShir R, Stein TI, Dahary D, et al. The GeneCards Suite. In: Abugessaisa I, Kasukawa T, editors. Practical Guide to Life Science Databases. Singapore: Springer Nature; 2021. p. 27–56. Available from: https://doi.org/10.1007/978-981-16-5812-9_2 Cui LB, Zhao SW, Zhang YH, Chen K, Fu YF, Qi T, et al. Associated transcriptional, brain and clinical variations in schizophrenia. Nat Mental Health. 2024;2(10):1239–49. Han Y, Gu S, Li Y, Qian X, Wang F, Huang JH. Neuroendocrine pathogenesis of perimenopausal depression. Front Psychiatry [Internet]. 2023 Mar 30 [cited 2024 Nov 24];14. Available from: https://www.frontiersin.org/journals/psychiatry/articles/ 10.3389/fpsyt.2023.1162501/full Lauterbach EC. Psychotropic Drug Effects on Gene Transcriptomics Relevant to Alzheimer Disease. Alzheimer Disease & Associated Disorders. 2012;26(1):1. Zhu Y, Li Q. Multifaceted roles of PDCD6 both within and outside the cell. Journal of Cellular Physiology. 2024;239(5):e31235. Rho SB, Song YJ, Lim MC, Lee SH, Kim BR, Park SY. Programmed cell death 6 (PDCD6) inhibits angiogenesis through PI3K/mTOR/p70S6K pathway by interacting of VEGFR-2. Cellular Signalling. 2012;24(1):131–9. Chen Y, Guan W, Wang ML, Lin XY. PI3K-AKT/mTOR Signaling in Psychiatric Disorders: A Valuable Target to Stimulate or Suppress? International Journal of Neuropsychopharmacology. 2024;27(2):pyae010. Kato T. Role of mTOR1 signaling in the antidepressant effects of ketamine and the potential of mTORC1 activators as novel antidepressants. Neuropharmacology. 2023;223:109325. Cavalleri L, Merlo Pich E, Millan MJ, Chiamulera C, Kunath T, Spano PF, et al. Ketamine enhances structural plasticity in mouse mesencephalic and human iPSC-derived dopaminergic neurons via AMPAR-driven BDNF and mTOR signaling. Mol Psychiatry. 2018;23(4):812–23. Li QS, Wajs E, Ochs-Ross R, Singh J, Drevets WC. Genome-wide association study and polygenic risk score analysis of esketamine treatment response. Sci Rep. 2020;10(1):12649. Tordera RM, Garcia-García AL, Elizalde N, Segura V, Aso E, Venzala E, et al. Chronic stress and impaired glutamate function elicit a depressive-like phenotype and common changes in gene expression in the mouse frontal cortex. European Neuropsychopharmacology. 2011;21(1):23–32. Sasaki T, Iwata S, Okano HJ, Urasaki Y, Hamada J, Tanaka H, et al. Nedd9 Protein, a Cas-L Homologue, Is Upregulated After Transient Global Ischemia in Rats. Stroke. 2005;36(11):2457–62. Soiza-Reilly M, Meye FJ, Olusakin J, Telley L, Petit E, Chen X, et al. SSRIs target prefrontal to raphe circuits during development modulating synaptic connectivity and emotional behavior. Mol Psychiatry. 2019;24(5):726–45. Chen T, Giri M, Xia Z, Subedi YN, Li Y. Genetic and epigenetic mechanisms of epilepsy: a review. Neuropsychiatric Disease and Treatment. 2017;13:1841–59. Mula M. The safety and tolerability of intranasal midazolam in epilepsy. Expert Review of Neurotherapeutics. 2014;14(7):735–40. Zhang C, Dong N, Xu S, Ma H, Cheng M. Identification of hub genes and construction of diagnostic nomogram model in schizophrenia. Front Aging Neurosci [Internet]. 2022 Oct 14 [cited 2025 Jan 12];14. Available from: https://www.frontiersin.org/journals/aging-neuroscience/articles/ 10.3389/fnagi.2022.1032917/full Gajjar S, Patel BM. Neuromedin: An insight into its types, receptors and therapeutic opportunities. Pharmacological Reports. 2017;69(3):438–47. Pałasz A, Worthington JJ, Filipczyk Ł, Saganiak K. Pharmacomodulation of brain neuromedin U signaling as a potential therapeutic strategy. Journal of Neuroscience Research. 2023;101(11):1728–36. Zhao R, Sun JB, Deng H, Cheng C, Li X, Wang FM, et al. Per1 gene polymorphisms influence the relationship between brain white matter microstructure and depression risk. Front Psychiatry. 2022;13. Available from: https://www.frontiersin.org/journals/psychiatry/articles/ 10.3389/fpsyt.2022.1022442/full Bunney BG, Li JZ, Walsh DM, Stein R, Vawter MP, Cartagena P, et al. Circadian dysregulation of clock genes: clues to rapid treatments in major depressive disorder. Mol Psychiatry. 2015;20(1):48–55. Drange OK, Smeland OB, Shadrin AA, Finseth PI, Witoelar A, Frei O, et al. Genetic Overlap Between Alzheimer’s Disease and Bipolar Disorder Implicates the MARK2 and VAC14 Genes. Front Neurosci. 2019;13. Available from: https://www.frontiersin.org/journals/neuroscience/articles/ 10.3389/fnins.2019.00220/full Guan W, Ni MX, Gu HJ, Yang Y. CREB: A Promising Therapeutic Target for Treating Psychiatric Disorders. Current Neuropharmacology. 2024;22(14):2384–401. Blendy JA. The Role of CREB in Depression and Antidepressant Treatment. Biological Psychiatry. 2006;59(12):1144–50. Kohshour MO, Papiol S, Ching CRK, Schulze TG. Genomic and neuroimaging approaches to bipolar disorder. BJPsych Open. 2022;8(2):e36. Shinoda Y, Sadakata T, Nakao K, Katoh-Semba R, Kinameri E, Furuya A, et al. Calcium-dependent activator protein for secretion 2 (CAPS2) promotes BDNF secretion and is critical for the development of GABAergic interneuron network. Proceedings of the National Academy of Sciences. 2011;108(1):373–8. Arévalo JC, Deogracias R. Mechanisms Controlling the Expression and Secretion of BDNF. Biomolecules. 2023;13(5):789. Yoo H, Yang SH, Kim JY, Yang E, Park HS, Lee SJ, et al. Down-regulation of habenular calcium-dependent secretion activator 2 induces despair-like behavior. Scientific Reports. 2021;11:3700. Starnawska A, Tan Q, Soerensen M, McGue M, Mors O, Børglum AD, et al. Epigenome-wide association study of depression symptomatology in elderly monozygotic twins. Transl Psychiatry. 2019;9(1):1–14. Sampson E, Mills NT, Hori H, Cearns M, Schwarte K, Hohoff C, et al. Long-term characterisation of the relationship between change in depression severity and change in inflammatory markers following inflammation-stratified treatment with vortioxetine augmented with celecoxib or placebo. Brain, Behavior, and Immunity. 2025;123:43–56. Kavakbasi E, Van Assche E, Schwarte K, Hohoff C, Baune BT. Long-Term Immunomodulatory Impact of VNS on Peripheral Cytokine Profiles and Its Relationship with Clinical Response in Difficult-to-Treat Depression (DTD). International Journal of Molecular Sciences. 2024;25(8):4196. Leenen FAD, Muller CP, Turner JD. DNA methylation: conducting the orchestra from exposure to phenotype? Clin Epigenet. 2016;8(1):92. Mangiavacchi A, Morelli G, Orlando V. Behind the scenes: How RNA orchestrates the epigenetic regulation of gene expression. Front Cell Dev Biol. 2023;11. Available from: https://www.frontiersin.org/journals/cell-and-developmental-biology/articles/ 10.3389/fcell.2023.1123975/full Tables Table 1. Top 10 differentially methylated CpGs between flexibly dosed cohorts. CpG Position Genomic context (UCSC) Mean Mid. Mean Ket. Methylation difference p -value FDR p -value cg15945600 Chr5: 304075 North Shelf PDCD6 gene-body 0.84 0.82 0.023; Ketamine Hypomethylated 7.28x10 -8 0.049 cg13615030 Chr9: 130639783 Island AK1 (promotor linked 1 variant) 0.12 0.10 0.017; Ketamine Hypomethylated 2.24x10 -6 0.748 cg15323881 Chr2: 9768508 North Shore YWHAQ 0.82 0.79 0.033; Ketamine Hypomethylated 4.03x10 -6 0.845 cg08951301 Chr12: 19865160 Open Sea ENCODE Candidate Cis-Regulatory Element 0.22 0.18 0.041; Ketamine Hypomethylated 5.07 x10 -6 0.845 cg15980099 Chr6: 137166932 Open Sea PEX7 gene body 0.87 0.89 -0.023; Ketamine Hypermethylated 7.91 x10 -6 1.000 cg08056716 Chr19: 35397106 South Shore LINC01838/LINC00904 (long intergenic non-coding RNA) 0.78 0.76 0.020; Ketamine Hypomethylated 9.14 x10 -6 1.000 cg15686281 Chr1: 44097898 Open Sea ENCODE candidate Cis-Regulatory Element; nearest to PTPRF 0.87 0.89 -0.013; Ketamine Hypermethylated 1.34 x10 -5 1.000 cg00725221 Chr9: 72789212 Open Sea MAMDC2/MAMDC2-AS1 gene body 0.90 0.92 -0.012; Ketamine Hypermethylated 1.38 x10 -5 1.000 cg23063825 Chr12: 298151 North Shore nearest to SLC6A12 0.34 0.43 -0.093; Ketamine Hypermethylated 1.67 x10 -5 1.000 cg09648702 Chr11: 20184779 Island ENSG00000294773/ Regulatory Elements/ DBX1 0.085 0.074 0.011; Ketamine Hypomethylated 1.81 x10 -5 1.000 Additional Declarations Yes There is no direct conflict of interest to this work. All potential conflicts of interest are disclosed in the manuscript. Supplementary Files SupplementarymaterialsKADSforsubmission2.pdf DNA methylation changes in a pharmaco-epigenomic EWAS in depression: comparing fixed and response-guided dosing paradigms for ketamine in the KADS trial Cite Share Download PDF Status: Under Review Version 1 posted Unknown event 03 Oct, 2025 Editorial decision: Reject after peer review 01 Jul, 2025 Review # 3 received at journal 29 Jun, 2025 Review # 2 received at journal 18 Jun, 2025 Reviewer # 3 agreed at journal 15 Jun, 2025 Reviewer # 2 agreed at journal 10 Jun, 2025 Review # 1 received at journal 06 Jun, 2025 Reviewer # 1 agreed at journal 05 Jun, 2025 Reviewers invited by journal 04 Jun, 2025 Editor assigned by journal 02 Jun, 2025 Submission checks completed at journal 02 Jun, 2025 First submitted to journal 30 May, 2025 Unknown event 30 May, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6778101","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":466674340,"identity":"ebc659a4-27d7-4ada-8480-a669dfa98118","order_by":0,"name":"Evelien Van Assche","email":"","orcid":"https://orcid.org/0000-0003-3472-3031","institution":"University of Münster","correspondingAuthor":false,"prefix":"","firstName":"Evelien","middleName":"Van","lastName":"Assche","suffix":""},{"id":466674341,"identity":"f766cd61-b358-415d-88a9-914e1ed7a920","order_by":1,"name":"Christa Hohoff","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Christa","middleName":"","lastName":"Hohoff","suffix":""},{"id":466674342,"identity":"45b5fb5b-c090-4612-b458-5e9772da1dcb","order_by":2,"name":"Victoria Statz","email":"","orcid":"","institution":"University of Münster","correspondingAuthor":false,"prefix":"","firstName":"Victoria","middleName":"","lastName":"Statz","suffix":""},{"id":466674343,"identity":"065aa2d5-e5bf-4373-8788-7bf5c0acf6e4","order_by":3,"name":"Sophia Wissing","email":"","orcid":"https://orcid.org/0009-0007-3123-2233","institution":"University of Münster","correspondingAuthor":false,"prefix":"","firstName":"Sophia","middleName":"","lastName":"Wissing","suffix":""},{"id":466674344,"identity":"3d2dff8f-61f6-4098-b716-ff0f602b4b8e","order_by":4,"name":"Angelo Alonzo","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Angelo","middleName":"","lastName":"Alonzo","suffix":""},{"id":466674345,"identity":"caf37853-7e0b-4056-83d6-6fcc09edc9e3","order_by":5,"name":"David Barton","email":"","orcid":"","institution":"NeuroCentrix Research Institute","correspondingAuthor":false,"prefix":"","firstName":"David","middleName":"","lastName":"Barton","suffix":""},{"id":466674346,"identity":"8b9eae13-f544-4e2a-8261-c42cd83bf482","order_by":6,"name":"Michael Berk","email":"","orcid":"https://orcid.org/0000-0002-5554-6946","institution":"Deakin University","correspondingAuthor":false,"prefix":"","firstName":"Michael","middleName":"","lastName":"Berk","suffix":""},{"id":466674347,"identity":"076c4189-52bb-4e75-ac3d-e2ee4a3ab9f7","order_by":7,"name":"Gregory Carter","email":"","orcid":"https://orcid.org/0000-0001-8180-1798","institution":"Faculty of Medicine and Health, University of Newcastle, Australia","correspondingAuthor":false,"prefix":"","firstName":"Gregory","middleName":"","lastName":"Carter","suffix":""},{"id":466674348,"identity":"2ee52403-0d7b-4f08-9837-8dea04a4ad7c","order_by":8,"name":"Mary Lou Chatterton","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Mary","middleName":"Lou","lastName":"Chatterton","suffix":""},{"id":466674349,"identity":"4326feac-1976-46dd-bd55-f1796bc14074","order_by":9,"name":"Vanessa Dong","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Vanessa","middleName":"","lastName":"Dong","suffix":""},{"id":466674350,"identity":"e4d79793-09af-44a9-b47b-7717191b7aeb","order_by":10,"name":"Paul Fitzgerald","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Paul","middleName":"","lastName":"Fitzgerald","suffix":""},{"id":466674351,"identity":"56a5b8b3-0d39-4eb6-8764-0d1cf0fb28ae","order_by":11,"name":"Veronica Galvez Ortiz","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Veronica","middleName":"Galvez","lastName":"Ortiz","suffix":""},{"id":466674352,"identity":"36a8eb2b-0245-44b1-bde8-d6ec235ace25","order_by":12,"name":"Nick Glozier","email":"","orcid":"https://orcid.org/0000-0002-0476-9146","institution":"University of Sydney","correspondingAuthor":false,"prefix":"","firstName":"Nick","middleName":"","lastName":"Glozier","suffix":""},{"id":466674353,"identity":"a94ca294-d618-4e1b-bc66-231f39268f5b","order_by":13,"name":"Paul Glue","email":"","orcid":"","institution":"University of Otago","correspondingAuthor":false,"prefix":"","firstName":"Paul","middleName":"","lastName":"Glue","suffix":""},{"id":466674354,"identity":"a2bc00cd-6674-4fed-884d-1ce07dab3613","order_by":14,"name":"Maree Hackett","email":"","orcid":"","institution":"Faculty of Medicine, University of New South Wales, The George Institute for Global Health","correspondingAuthor":false,"prefix":"","firstName":"Maree","middleName":"","lastName":"Hackett","suffix":""},{"id":466674355,"identity":"5bf6490e-b90b-472e-98b0-d4ab965b061b","order_by":15,"name":"Dusan Hadzi-Pavlovic","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Dusan","middleName":"","lastName":"Hadzi-Pavlovic","suffix":""},{"id":466674356,"identity":"8d36f5bd-e8ce-4be0-bdfd-e99b58ad496b","order_by":16,"name":"Sean Hood","email":"","orcid":"https://orcid.org/0000-0003-2852-7923","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Sean","middleName":"","lastName":"Hood","suffix":""},{"id":466674357,"identity":"8f3cc736-89d3-4cdf-8a34-d7a5a7825668","order_by":17,"name":"Colleen Loo","email":"","orcid":"https://orcid.org/0000-0003-3267-0554","institution":"University of New South Wales \u0026 Black Dog Institute, Sydney, Australia","correspondingAuthor":false,"prefix":"","firstName":"Colleen","middleName":"","lastName":"Loo","suffix":""},{"id":466674358,"identity":"a53620a1-f477-4507-bd47-e2c5f323f33a","order_by":18,"name":"Donel Martin","email":"","orcid":"https://orcid.org/0000-0002-8452-0390","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Donel","middleName":"","lastName":"Martin","suffix":""},{"id":466674359,"identity":"eb78363b-4006-4cde-a0bd-89411bde1172","order_by":19,"name":"Cathrine Mihalopoulos","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Cathrine","middleName":"","lastName":"Mihalopoulos","suffix":""},{"id":466674360,"identity":"2a0b1cc7-2123-419f-87f8-943a22ea91b6","order_by":20,"name":"Natalie Mills","email":"","orcid":"https://orcid.org/0000-0003-3255-5118","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Natalie","middleName":"","lastName":"Mills","suffix":""},{"id":466674361,"identity":"8800791f-550a-4cb6-bfb7-4cbf77906f20","order_by":21,"name":"Philip Mitchell","email":"","orcid":"https://orcid.org/0000-0002-7954-5235","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Philip","middleName":"","lastName":"Mitchell","suffix":""},{"id":466674362,"identity":"cd4c48bd-5cda-41b2-8b93-b5ca21736eab","order_by":22,"name":"Stevan Nikolin","email":"","orcid":"https://orcid.org/0000-0003-4440-9273","institution":"University of New South Wales","correspondingAuthor":false,"prefix":"","firstName":"Stevan","middleName":"","lastName":"Nikolin","suffix":""},{"id":466674363,"identity":"342bfc5a-e549-4c22-aeb0-aa3057b94194","order_by":23,"name":"Anthony Rodgers","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Anthony","middleName":"","lastName":"Rodgers","suffix":""},{"id":466674364,"identity":"4fb4bbfc-787d-4f5b-8c07-024203243163","order_by":24,"name":"Shanthi Sarma","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Shanthi","middleName":"","lastName":"Sarma","suffix":""},{"id":466674365,"identity":"12238703-b705-4e77-b3ac-9177d6224a38","order_by":25,"name":"Andrew Somogyi","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Andrew","middleName":"","lastName":"Somogyi","suffix":""},{"id":466674339,"identity":"a49ed049-7355-4e88-9023-1caaf0feb35b","order_by":26,"name":"Bernhard Baune","email":"data:image/png;base64,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","orcid":"","institution":"University of Münster","correspondingAuthor":true,"prefix":"","firstName":"Bernhard","middleName":"","lastName":"Baune","suffix":""}],"badges":[],"createdAt":"2025-05-29 15:56:38","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6778101/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6778101/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":84325010,"identity":"435b061d-f545-4366-ae5f-fe613a7d3d0f","added_by":"auto","created_at":"2025-06-10 14:59:27","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":254530,"visible":true,"origin":"","legend":"\u003cp\u003eDNA methylation differences between flexible dosed cohorts.\u003c/p\u003e\n\u003cp\u003e1a. DNA methylation differences between both groups and \u003cem\u003ep\u003c/em\u003e-values. The most significant CpG (cg15945600; \u003cem\u003ep\u003c/em\u003e=7.28x10\u003csup\u003e-8\u003c/sup\u003e) is marked with the corresponding \u003cem\u003ep\u003c/em\u003e-value. 1b. Mean change of DNA methylation at cg15945600 per treatment group (mean and standard error). 1c. Density plot of DNA methylation at cg15945600 at baseline, before RCT treatment. 1d. DNA methylation at cg15945600 post RCT treatment by treatment condition.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-6778101/v1/1513a2baf3376ac8e2161055.png"},{"id":84325012,"identity":"39adb7e0-c39e-48ac-ba4e-db7c5ba34331","added_by":"auto","created_at":"2025-06-10 14:59:27","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":245837,"visible":true,"origin":"","legend":"\u003cp\u003eSummary of analyses comparing fixed and flexible treatment (ketamine and midazolam combined).\u003c/p\u003e\n\u003cp\u003e2a. DNA methylation differences between both groups and \u003cem\u003ep\u003c/em\u003e-values. The most significant CpG (cg20023762; \u003cem\u003ep\u003c/em\u003e=1.11x10\u003csup\u003e-7\u003c/sup\u003e) is marked with the corresponding \u003cem\u003ep\u003c/em\u003e-value. 2b. Mean change of DNA methylation at cg20023762 per treatment group (mean and standard error). 1c. Density plot of DNA methylation at cg20023762 at baseline, before RCT treatment. 1d. DNA methylation at cg20023762 post RCT treatment by treatment condition.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-6778101/v1/a2d9855c548e7648adb6ecf9.png"},{"id":85749706,"identity":"57e5fdf5-b26c-4003-92c7-283d39d26215","added_by":"auto","created_at":"2025-07-01 09:55:05","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1482792,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6778101/v1/2fed0582-34f4-4501-bca4-710aba8b431f.pdf"},{"id":84325013,"identity":"3814ecd6-2219-4893-9d5f-60b0f848444a","added_by":"auto","created_at":"2025-06-10 14:59:27","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":3206072,"visible":true,"origin":"","legend":"DNA methylation changes in a pharmaco-epigenomic EWAS in depression: comparing fixed and response-guided dosing paradigms for ketamine in the KADS trial","description":"","filename":"SupplementarymaterialsKADSforsubmission2.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6778101/v1/129cb7e325e2bb3c8b871daf.pdf"}],"financialInterests":"\u003cb\u003eYes\u003c/b\u003e\nThere is no direct conflict of interest to this work. \r\nAll potential conflicts of interest are disclosed in the manuscript.","formattedTitle":"DNA methylation changes in a pharmaco-epigenomic EWAS in depression: comparing fixed and response-guided dosing paradigms for ketamine in the KADS trial","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eKetamine has established its place as a rapid-acting antidepressant at sub-anesthetic doses (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e). Varying approaches to dosing have been used. The KADS-trial (Ketamine for Adult Depression Study) was a randomized controlled clinical trial (RCT) designed to compare ketamine with midazolam, as an active control condition (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e). The study initially used a fixed dose of ketamine and midazolam (\u0026lsquo;Cohort 1\u0026rsquo;), though after lack of remitters was observed in either group for the first 51 participants, a revision of the protocol occurred, with the introduction of flexible response-guided dosing ketamine and midazolam (\u0026lsquo;Cohort 2\u0026rsquo;) (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e). The study showed that flexibly dosed ketamine was more efficacious than midazolam in reducing depressive symptoms in patients with treatment resistant depression (TRD) (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e). However, the molecular underpinnings of the relationship between clinical response and treatment paradigm are poorly understood. The two dosing cohorts in the KADS trial provide an excellent opportunity to investigate epigenetic characteristics of the dynamic response-guided pharmacological strategy for ketamine as an antidepressant.\u003c/p\u003e \u003cp\u003eDNA methylation and epigenetic markers are biomarkers that are sensitive to environmental and pharmacological influences (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e). Epigenome-wide DNA methylation may reflect changes that occur in the individual over the course of treatment, be it pharmacotherapy (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e) or psychotherapy for depression (\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e). These changes are also reflected in gene expression, one of the mechanisms through which DNA methylation affects the individual (\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e). Prior research suggests that ketamine acts through altered DNA methylation profiles (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e). Reports in rodent models have supported this hypothesis for the \u003cem\u003eBDNF\u003c/em\u003e gene, which codes for the brain-derived neurotrophic factor (BDNF) protein, in models for Post-traumatic stress disorder (PTSD) (\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThere are scant longitudinal DNA methylation data that allow an in-depth investigation of the effect of different dosing paradigms for an antidepressant treatment, as well as longitudinal epigenome-wide studies on ketamine for depression. DNA methylation has been shown to change alongside treatment in longitudinal designs and randomized controlled trials, in line with treatment response (\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e), though without the opportunity to investigate in-depth the impact of different dosing paradigms on DNA methylation. Therefore, we focus on the clinically observed benefit of the response-guided \u0026lsquo;flexible\u0026rsquo; dosing paradigm and expect that DNA methylation changes to follow this pattern with a more pronounced molecular response at the epigenomic level for the flexibly dosed ketamine cohort.\u003c/p\u003e \u003cp\u003eTo investigate the correlation between treatment paradigm and DNA methylation in depth, it is advantageous to use longitudinal data, such as in our sample. The characteristics of a randomized controlled trial assist in controlling for phenotypic heterogeneity between groups, as is typical for depression. In addition, it is a rare opportunity to investigate two dosing scenarios of the same medication within a single clinical trial. Hence, we aimed to test the hypothesis that the clinical benefit of the flexible response-guided treatment paradigm is reflected in DNA methylation changes accordingly: we expect more pronounced DNA methylation changes in the flexibly dosed cohorts, particularly in individuals from the ketamine-group.\u003c/p\u003e \u003cp\u003eOur research presents the first epigenome-wide association study (EWAS) for ketamine. We tackle the pharmaco-epigenomic question at hand from an innovative personalized dosing perspective, comparing the impact of dosing and treatment modalities on DNA methylation changes, with a combined cross-sectional and longitudinal approach. We investigated DNA methylation at the CpG dinucleotide and differentially methylated regions (DMR) with the hypothesis that response-guided dosing, which optimizes dosage based on individual response, shows a stronger effect at the molecular level, also independent of antidepressant treatment response \u003cem\u003eper se\u003c/em\u003e. Hence, we expect to find the largest effect of treatment on DNA methylation at end of RCT treatment for the flexible dosed ketamine group. The combination of cross-sectional analysis at end of RCT treatment and longitudinal analyses spanning the treatment course allow for an in-depth exploration of the response of DNA methylation to dosing or treatment specific effects.\u003c/p\u003e"},{"header":"2. Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1. Sample description\u003c/h2\u003e \u003cp\u003eThe KADS sample (Ketamine for Adult Depression Study) consisted of 184 individual patients with treatment resistant major depressive disorder. The study was a randomized controlled trial (RCT) with individuals receiving subcutaneous ketamine or midazolam as an active control condition. Doses were either administered according to a fixed paradigm (ketamine: 0.5 mg/kg; midazolam 0.025 mg/kg, identical volumes of injection; \u0026lsquo;Cohort 1\u0026rsquo;) or a flexible paradigm allowing for dose increases based on clinical assessment (ketamine: 0.5\u0026ndash;0.9 mg/kg; midazolam 0.025\u0026ndash;0.045 mg/kg; \u0026lsquo;Cohort 2\u0026rsquo;). For a detailed description of the KADS trial see Loo et al. (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eAlong with phenotypic information, whole blood was collected at five time-points during the randomized and the open label extension phases. The randomized phase took place between baseline (T1) and end of RCT treatment (T3), 4 weeks apart, which is also the focus of this analysis. Blood collections T4 and T5 occurred during the open-label phase. In total, 388 blood samples were analyzed over the five time-points (T1-T5; Fig. \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). Quality control (QC) steps were performed for all 388 samples and resulted in a final dataset with 379 samples from 113 unique participants (see supplementary data). 87 participants had DNA methylation data available for end of RCT treatment (T3), and 76 of these also had DNA methylation data available for baseline (T1), allowing for longitudinal analyses. From the sample of 87 individuals, the mean age was 48.3 years and 32.2% were women. 41 individuals received ketamine, 46 received midazolam, with 17 and 19 individuals in the flexible cohort, i.e., higher average doses, respectively (Fig. \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). Mean MADRS score (Montgomery-\u0026Aring;sberg Rating Scale for Depression), a rating scale for depression severity (\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e), at baseline was 29.3 points (\u003cem\u003eSD\u003c/em\u003e\u0026thinsp;=\u0026thinsp;5.9).\u003c/p\u003e \u003cp\u003e The study was approved by the Sydney Local Health District (RPAH Zone) Human Research Ethics Committee (Australia; X16-0146 and HREC/16/RPAH/168) and the Southern Health and Disability Ethics Committee (New Zealand; 16/STH/104). All participants provided written informed consent. Trial registration: ACTRN12616001096448 at \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e\u003ca href=\"http://www.anzctr.org.au\" target=\"_blank\"\u003ewww.anzctr.org.au\u003c/a\u003e\u003c/span\u003e\u003cspan address=\"http://www.anzctr.org.au\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2. Epigenome-wide DNA methylation analysis\u003c/h2\u003e \u003cp\u003eDNA methylation data (Illumina Infinium MethylationEPIC 850k BeadChip) were available from 388 samples of 113 unique individuals over multiple time-points (Fig. \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). Though we solely focus on T1 (baseline) and T3 (end of RCT treatment), QC, pre-processing, and normalization were performed for all five time-points combined for methodological purposes. Following QC steps, 9 samples were discarded, resulting in 379 available samples. After finalizing QC, we extracted the samples available for the time-points of interest in line with our analysis strategy with cross-sectional and longitudinal analyses (i.e., 76 for T1 and 87 for T3). DNA was isolated from whole blood samples using standard procedures (QIAamp DNA Blood Midi-Kit, Qiagen, Hilden, Germany) followed by purification (Amicon 0,5ml 3K; Merck/Millipore, Darmstadt, Germany) and pipetted on 96-well plates for chip-based analyses. Bisulfite conversion and handling of the DNA methylation chips were performed at the Life\u0026amp;Brain Institute (Bonn, Germany). Samples were randomized on plates and chips based on patient\u0026rsquo;s sex and age. Samples from the same patient were analysed on the same chip to minimize batch-effects for within-individual analyses. Following analysis on HiScan array scanning systems (Illumina, San Diego, CA), data were transferred as .idat files. The further processing and quality control of the DNA methylation data was performed using R (version 4.4.2) and the \u0026lsquo;RnBeads\u0026rsquo; pipeline (Package RnBeads 2.0 (\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e)).\u003c/p\u003e \u003cp\u003eNormalization was performed using \u0026ldquo;wm.dasen\u0026rdquo; as embedded in the RnBeads package. A detailed description of the QC steps can be found in the supplementary data (Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3. Statistical analyses\u003c/h2\u003e \u003cdiv id=\"Sec6\" class=\"Section3\"\u003e \u003ch2\u003e2.3.1. Cross-sectional analyses\u003c/h2\u003e \u003cp\u003eDifferential methylation analyses for end of RCT treatment were performed with \u0026lsquo;limma\u0026rsquo;, as also embedded in the RnBeads package. In total, four cross-sectional differential methylation analyses were performed: (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e) an overall analysis comparing ketamine vs. midazolam for the whole sample, (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e) comparing flexibly dosed ketamine and flexibly dosed midazolam, (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e) comparing fixed dose ketamine and fixed dose midazolam, and (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e) comparing fixed versus flexible dosing, combined across both treatment conditions.\u003c/p\u003e \u003cp\u003eCross-sectional analyses were corrected for biological sex, age, site where the blood was collected, six methylation-based cell-type estimates, as well as ancestry, represented by the two first principal components of genome-wide single nucleotide polymorphism (SNP) data (\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e). The overall analysis was also corrected for dosing cohort (i.e., fixed or flexible), which was added as an additional covariate. Cell types were estimated as suggested by Salas et al. (\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e) with the GSE110554 6-cell-types reference dataset using the Houseman method (\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e) to estimate the most important cell type fractions for our samples: neutrophils, monocytes, B-lymphocytes, natural killer cells, and CD4\u0026thinsp;+\u0026thinsp;and CD8\u0026thinsp;+\u0026thinsp;T-cells. As we had body weight available, but no height, we did not have body mass index available and did not include weight in the model.\u003c/p\u003e \u003cp\u003eTechnical confounding factors (e.g., batch effects) were addressed using surrogate variables. For all four cross-sectional analyses the overall genomic inflation factor (\u003cem\u003eλ\u003c/em\u003e; \u003cem\u003eχ\u003c/em\u003e\u003csup\u003e\u003cem\u003e2\u003c/em\u003e\u003c/sup\u003e-test) was between 0.97 and 1.01. Details on the surrogate variables and QQ-plots can be found in the supplementary data (Fig. S3-S6).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section3\"\u003e \u003ch2\u003e2.3.2. Paired longitudinal analyses\u003c/h2\u003e \u003cp\u003eWe extended our analysis strategy with a paired longitudinal analysis comparing DNA methylation at baseline with end of RCT treatment. Analogous with the cross-sectional analyses, the paired longitudinal analyses started from the whole sample of individuals having DNA methylation available at baseline and end of RCT treatment (\u003cem\u003eN\u003c/em\u003e\u0026thinsp;=\u0026thinsp;76), comparing DNA methylation at baseline with DNA methylation at end of RCT treatment. In a second step, analyses were stratified by treatment: ketamine-only and midazolam-only. Finally, analyses were stratified by treatment and dosing paradigm: ketamine flexible (\u003cem\u003eN\u003c/em\u003e\u0026thinsp;=\u0026thinsp;17 pairs), ketamine fixed (\u003cem\u003eN\u003c/em\u003e\u0026thinsp;=\u0026thinsp;16 pairs), and midazolam flexible (\u003cem\u003eN\u003c/em\u003e\u0026thinsp;=\u0026thinsp;19 pairs), and midazolam fixed (\u003cem\u003eN\u003c/em\u003e\u0026thinsp;=\u0026thinsp;24 pairs). For the paired longitudinal analysis, the Welch\u0026rsquo;s t-test was used in the RnBeads package. Due to the small numbers, results of the paired analyses were only interpreted as differentially methylated regions (DMRs). DMRs were calculated using the comb-p algorithm embedded in the Enmix R-package (\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e). Settings for the DMR identification were: dist.cutoff\u0026thinsp;=\u0026thinsp;1000, bin.size\u0026thinsp;=\u0026thinsp;500, seed\u0026thinsp;=\u0026thinsp;0.01.\u003c/p\u003e \u003cp\u003eFinally, in line with existing evidence on the polygenic nature of psychiatric disorders, we assumed a polygenic involvement of multiple small effects not reaching statistical significance, though relevant for the phenotype at hand, as shown through polygenic risk scores (\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e). Therefore, in analogy of enrichment tests (\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e), we compared the count of CpGs with \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05 for each of the four treatment conditions as a proxy of the overall impact of treatment on longitudinal DNA methylation changes. The comparison of CpG-proportions with \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05 vs. \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026gt;\u0026thinsp;0.05 between treatment conditions (cf. enrichment analysis) was performed using a \u003cem\u003eχ\u003c/em\u003e\u003csup\u003e\u003cem\u003e2\u003c/em\u003e\u003c/sup\u003e-test. For all the longitudinal analyses the overall genomic inflation factor (\u003cem\u003eλ\u003c/em\u003e; \u003cem\u003eχ\u003c/em\u003e\u003csup\u003e\u003cem\u003e2\u003c/em\u003e\u003c/sup\u003e-test) was between 0.89 and 1.00 (supplementary fig. S7-S13).\u003c/p\u003e \u003cp\u003eIn line with the available literature we considered a \u003cem\u003ep\u003c/em\u003e-value below 9.42 \u0026times; 10\u003csup\u003e\u0026minus;\u0026thinsp;8\u003c/sup\u003e to be significant for the EPIC array (\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e). For the DMR-analysis false-discovery-rate (FDR)-corrected \u003cem\u003ep\u003c/em\u003e-values (Benjamini-Hochberg) below 0.05 were considered statistically significant. For the interpretation of results, the following online databases were used: UCSC genome browser (\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e), Alliance of Genome Resources (\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e), and GeneCards\u0026reg;(\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eOverall comparability of the midazolam and ketamine cohorts regarding confounding factors at baseline was tested using ANOVA and Fisher\u0026rsquo;s exact test.\u003c/p\u003e \u003cp\u003eSecondary explorations and post-hoc sensitivity analyses of the significant results were performed using ANOVA or by calculating Pearson\u0026rsquo;s correlation coefficient.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"3. Results","content":"\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e\n \u003ch2\u003e3.1. Sample characteristics\u003c/h2\u003e\n \u003cp\u003eBaseline characteristics of the end of RCT treatment-sample (\u003cem\u003eN\u003c/em\u003e\u0026thinsp;=\u0026thinsp;87) showed comparable groups for each of both treatment conditions. No significant group differences were found for age at baseline (\u003cem\u003eF\u003c/em\u003e(1,85)\u0026thinsp;=\u0026thinsp;2.48; \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.12), gender distribution (% female; \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.25), mean MADRS at baseline (\u003cem\u003eF\u003c/em\u003e(1,85)\u0026thinsp;=\u0026thinsp;1.59; \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.21), weight at baseline (\u003cem\u003eF\u003c/em\u003e(1,85)\u0026thinsp;=\u0026thinsp;0.14; \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.71), and years of education (\u003cem\u003eF\u003c/em\u003e(1,85)\u0026thinsp;=\u0026thinsp;0.026; \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.87).\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\n \u003ch2\u003e3.2. Cross-sectional epigenome-wide analyses at end of RCT treatment.\u003c/h2\u003e\n \u003cdiv id=\"Sec11\" class=\"Section3\"\u003e\n \u003ch2\u003e3.2.1. Overall analysis of Ketamine vs. Midazolam\u003c/h2\u003e\n \u003cp\u003eThe overall analysis comparing all individuals treated with ketamine vs. individuals treated with midazolam independent of dosing scheme did not return any epigenome wide significant CpGs. The CpG with the smallest \u003cem\u003ep\u003c/em\u003e-value was cg11159519, located in the \u003cem\u003eKCNH1\u003c/em\u003e gene body: a potassium channel gene (Chr1: 210857380; South Shore; \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;2.38x10\u003csup\u003e\u0026minus;\u0026thinsp;6\u003c/sup\u003e; ketamine group relatively hypomethylated; DNA methylation difference\u0026thinsp;=\u0026thinsp;0.0041). Other CpGs in the top 10 were in the vicinity of the following genes: \u003cem\u003eMAML2\u003c/em\u003e, \u003cem\u003eHSBP1\u003c/em\u003e, \u003cem\u003eCRTC3\u003c/em\u003e, \u003cem\u003eNMU\u003c/em\u003e, \u003cem\u003ePER1\u003c/em\u003e, \u003cem\u003eVAC14\u003c/em\u003e, \u003cem\u003eATE1\u003c/em\u003e, and \u003cem\u003eNCKAP5\u003c/em\u003e/RN7SKP154 pseudogene (see supplementary Table S2).\u003c/p\u003e\n \u003c/div\u003e\n \u003cdiv id=\"Sec12\" class=\"Section3\"\u003e\n \u003ch2\u003e3.2.2. Flexible doses of Ketamine vs. flexible doses of Midazolam\u003c/h2\u003e\n \u003cp\u003eIn line with the clinical observation, the stratified analysis comparing flexibly dosed ketamine with flexibly dosed midazolam showed one epigenome-wide significant CpG: cg15945600 (\u003cem\u003ep\u0026thinsp;=\u003c/em\u003e\u0026thinsp;7.28x10\u003csup\u003e\u0026minus;\u0026thinsp;8\u003c/sup\u003e, FDR-corrected \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.049; Fig. \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003ea). This CpG is in the gene-body of the \u003cem\u003ePDCD6\u003c/em\u003e gene. For this CpG the ketamine group is relatively hypomethylated with a DNA methylation difference of 2.3% (table 1).\u003c/p\u003e\n \u003cp\u003eA secondary exploration of this result showed distinct DNA methylation values for each of the treatment cohorts (\u003cem\u003eF\u003c/em\u003e(3,83)\u0026thinsp;=\u0026thinsp;5.10; \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.0028; Fig. \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003eb), which was not present at baseline (\u003cem\u003eF\u003c/em\u003e(3,91)\u0026thinsp;=\u0026thinsp;1.64; \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.19).\u003c/p\u003e\n \u003c/div\u003e\n \u003cdiv id=\"Sec13\" class=\"Section3\"\u003e\n \u003ch2\u003e3.2.3. Fixed dose Ketamine vs. fixed dose Midazolam\u003c/h2\u003e\n \u003cp\u003eIn line with the clinical observation that the antidepressant effect was more pronounced in the flexible dosing cohort (\u003cspan class=\"CitationRef\"\u003e2\u003c/span\u003e), the analyses comparing fixed dosing groups showed no significant epigenetic differences. The most significant CpG was cg04584009 (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;3.09x10\u003csup\u003e\u0026minus;\u0026thinsp;6\u003c/sup\u003e), located in a CpG island in the \u003cem\u003eREXO1\u003c/em\u003e gene-body. \u003cem\u003eREXO1\u003c/em\u003e has been associated with brain function in prior research, in particular in the context of schizophrenia (\u003cspan class=\"CitationRef\"\u003e23\u003c/span\u003e). For this CpG the ketamine group was hypermethylated (DNA methylation difference: 1.3%).\u003c/p\u003e\n \u003cp\u003eWithin the top 10 of CpGs from this analysis, one region was represented three times (table S3). The region in question is particularly linked to regulatory aspects, e.g., long non-coding RNA and cis-regulatory elements (see also supplementary fig. S15).\u003c/p\u003e\n \u003c/div\u003e\n \u003cdiv id=\"Sec14\" class=\"Section3\"\u003e\n \u003ch2\u003e3.2.4. Fixed doses vs. flexible doses Ketamine and Midazolam combined.\u003c/h2\u003e\n \u003cp\u003eIn line with our hypothesis, we compared the lower fixed dose cohort 1 with the higher flexibly dosed cohort 2. For this analysis, one CpG came close to the pre-defined significance threshold (p\u0026thinsp;=\u0026thinsp;1.11x10\u003csup\u003e\u0026minus;\u0026thinsp;7\u003c/sup\u003e; FDR-corrected \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.074). This CpG, cg20023762, is located in the \u003cem\u003eNEDD9\u003c/em\u003e gene body, a gene both linked to depression and pharmacogenomics (\u003cspan class=\"CitationRef\"\u003e24\u003c/span\u003e, \u003cspan class=\"CitationRef\"\u003e25\u003c/span\u003e). The top 10 of CpGs for this analysis are in Table S4.\u003c/p\u003e\n \u003c/div\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e\n \u003ch2\u003e3.3. Paired longitudinal analyses: baseline versus end of RCT treatment.\u003c/h2\u003e\n \u003cp\u003eThe whole-sample longitudinal analysis comparing DNA methylation at end of RCT treatment with baseline showed no significant differentially methylated region (DMR). This was also the case for the longitudinal analysis for the individuals treated with ketamine, and those treated with midazolam.\u003c/p\u003e\n \u003cp\u003eA sensitivity analysis with the most-significant CpG of the whole-sample longitudinal analysis (cg10080497, p\u0026thinsp;=\u0026thinsp;1.68x10\u003csup\u003e\u0026minus;\u0026thinsp;6\u003c/sup\u003e), showed a significant correlation between change in DNA methylation over time and change in MADRS for the ketamine group only (\u003cem\u003er\u003c/em\u003e(\u003cspan class=\"CitationRef\"\u003e31\u003c/span\u003e)\u0026thinsp;=\u0026thinsp;0.36; \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.042; supplementary fig. S14).\u003c/p\u003e\n \u003cp\u003eThe only analysis returning one significant DMR was the longitudinal analysis for patients receiving ketamine in flexible higher doses over the course of 4 weeks. This DMR consists of two CpGs (Chr 12; 75784541\u0026ndash;75784618; \u003cem\u003ep\u003c/em\u003e-value (FDR-corrected)\u0026thinsp;=\u0026thinsp;0.0045) in the \u003cem\u003eCAPS2\u003c/em\u003e gene-region and close to the \u003cem\u003eGLIPR1L2\u003c/em\u003e promotor-linked CpG island (supplementary figure S16). Both CpGs showed a decrease in DNA methylation over time (cg20291222: mean difference over time\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;0.010; cg14292619: mean difference over time = -0.0062). The same analyses in each of the other sub-groups, i.e., ketamine fixed doses and both midazolam cohorts, did not return any significant DMR.\u003c/p\u003e\n \u003cp\u003eOverall, the longitudinal cohort for ketamine in flexible doses, returned 32.305 CpGs with \u003cem\u003ep\u003c/em\u003e-value\u0026thinsp;\u0026lt;\u0026thinsp;0.05. This was the case for 29.947 CpGs for the ketamine fixed doses sample, and 28.666 CpGs and 28.680 CpGs for the fixed and flexible doses of Midazolam, respectively. Between both ketamine groups, \u003cem\u003ep\u003c/em\u003e-value proportions were significantly different (\u0026chi;\u003csup\u003e2\u003c/sup\u003e(1, \u003cem\u003eN\u003c/em\u003e\u0026thinsp;=\u0026thinsp;667 027)\u0026thinsp;=\u0026thinsp;93.609, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001). This was not the case for the comparison of both midazolam cohorts, despite the dose differences (\u0026chi;\u003csup\u003e2\u003c/sup\u003e(1, 667 027)\u0026thinsp;=\u0026thinsp;0.0031, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026gt;\u0026thinsp;0.95). This was again consistent for the comparison of both cohorts with fixed doses (\u0026chi;\u003csup\u003e2\u003c/sup\u003e(1, \u003cem\u003eN\u003c/em\u003e\u0026thinsp;=\u0026thinsp;667 027)\u0026thinsp;=\u0026thinsp;29.237, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001) and both cohorts with flexible, i.e., higher doses (\u0026chi;\u003csup\u003e2\u003c/sup\u003e(1, \u003cem\u003eN\u003c/em\u003e\u0026thinsp;=\u0026thinsp;667 027)\u0026thinsp;=\u0026thinsp;225.67, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001).\u003c/p\u003e\n\u003c/div\u003e"},{"header":"4. Discussion","content":"\u003cp\u003eReflecting results from the randomized phase of the KADS clinical trial (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e), our hypothesis was that the observed clinical response for the flexible, response-guided dose condition would also be evident at the level of DNA methylation. We assumed that response-guided dosing which optimizes dosing on an individual basis, would show a stronger effect at the molecular level, also independent of antidepressant treatment response \u003cem\u003eper se\u003c/em\u003e. By combining cross-sectional and paired longitudinal analyses both on the whole sample and in a stratified manner, we demonstrated support for this hypothesis.\u003c/p\u003e \u003cp\u003eIn our cross-sectional analysis at the end of RCT treatment, we found that, despite the relatively small sample size, on comparing flexible high dose ketamine with flexible high dose midazolam there was an epigenome-wide significant hit. Also, in line with our hypothesis that the flexible dosed cohort would show a stronger effect at the DNA methylation level, the fixed-dose comparison did not return any significant signal and that the combined flexible-dose comparison independent of the treatment, suggested a potentially relevant CpG, notwithstanding the epigenome-wide cut-off or the FDR-correction.\u003c/p\u003e \u003cp\u003eOur cross-sectional analyses hint at some genes and systems that may be involved in ketamine\u0026rsquo;s antidepressant action. The \u003cem\u003ePDCD6\u003c/em\u003e-gene is a calcium-binding protein, suggested as being involved in the regulation of cytoskeletal dynamics and innate immune responses (\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e), but also apoptosis (\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e). \u003cem\u003ePDCD6\u003c/em\u003e has also been suggested to play a role in the PI3K/mTOR/p70S6K pathway by interacting with VEGFR-2 (\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e). The mTor system has previously been linked to several psychiatric treatments, ketamine in particular (\u003cspan additionalcitationids=\"CR29\" citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e). A recent genome-wide association study (GWAS) on esketamine and treatment response also highlighted a connection with the immune-system through the \u003cem\u003eIRAK3\u003c/em\u003e-gene (Interleukin 1 Receptor Associated Kinase 3) (\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e). The gene \u003cem\u003eNEDD9\u003c/em\u003e is another gene that has been cautiously highlighted by our cross-sectional analyses. \u003cem\u003eNEDD9\u003c/em\u003e is a stimulator of cell differentiation and has been shown to affect the glutamate pathway and depressive-like behavior in rodents (\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e). NEDD9 promoted neurite outgrowth of neuronal cells after ischemic damage (\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e). In the context of pharmacogenetics, NEDD9 was suggested as a selective serotonin reuptake inhibitor (SSRI) target for development and modulating synaptic connectivity and emotional behaviour in mice (\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe cross-sectional analysis including the full sample and comparing treatment with ketamine versus midazolam returned some CpGs located in or in the vicinity of biologically plausible genes too, particularly in the context of pharmaco-epigenomics. CpGs in the top 10 hinted at genes linked to potassium signaling (\u003cem\u003eKCNH1\u003c/em\u003e); often linked to epilepsy (\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e)\u0026ndash; for which midazolam is an effective treatment (\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e). Furthermore, genes linked to immune-mediated inflammation were also represented in this top 10 such as \u003cem\u003eNMU\u003c/em\u003e (\u0026lsquo;Neuromedin U\u0026rsquo;) with links to schizophrenia (\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e) and suggested as a treatment target for other psychiatric phenotypes (\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e), as well as genes known to play a role in depression-related white-matter changes and the circadian rhythm in depression (e.g., \u003cem\u003ePER1\u003c/em\u003e and \u003cem\u003eVAC14\u003c/em\u003e; 42\u0026ndash;44\u003cb\u003e)\u003c/b\u003e. In addition, \u003cem\u003eMAML2\u003c/em\u003e and \u003cem\u003eCRTC3\u003c/em\u003e both act in the same CREB-pathway which is linked to synaptic plasticity and has already been suggested as a therapeutic target for depression (\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e, \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e). \u003cem\u003eNCKAP5\u003c/em\u003e has been linked to synaptic plasticity as well in bipolar disorder (\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e). However, none of these CpGs reached statistical significance. The QQ-plot suggested low statistical power, which might be linked to the relatively small sample size. However, we also expect a dilution effect in the overall sample, as flexible higher and fixed lower dosing schemes are combined, given the lack of clinical effect for the fixed doses (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e). Therefore, the stratification probably led to more homogeneous groups, which might have been beneficial for the statistical power, despite the smaller sample size for the stratified analyses.\u003c/p\u003e \u003cp\u003eOur paired longitudinal analysis also supported our hypothesis and highlights the connection between response-guided dosing and changes at the molecular level, as well as the benefit of stratification, by returning a significant DMR for the flexibly dosed ketamine group only. Despite the small sample size, the number of CpGs with a \u003cem\u003ep\u003c/em\u003e-value below 0.05 was also substantially, and significantly, higher than for each of the other conditions. The \u003cem\u003eCAPS2\u003c/em\u003e gene is also calcium-dependent and involved in promoting the BDNF-pathway (\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e, \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e). The \u003cem\u003eCAPS2\u003c/em\u003e gene-expression is decreased in post-mortem samples of depressive patients that died from suicide (\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e), and may be upregulated in the context of symptom remission following a cognitive psychotherapeutic intervention (\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e). It also has been linked to despair-like behaviour in rodents (\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e). The observation that both CpGs that contribute to the DMR show decreased DNA methylation over the course of treatment is in line with these prior results. \u003cem\u003eGLIPR1L2\u003c/em\u003e as a gene and DMR has already been linked to depression in a prior epigenome-wide association study with elderly twins (\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eOverall, the two genes that showed a link with our hypothesis that the response-guided dosing condition is linked to more prominent DNA methylation changes, \u003cem\u003ePDCD6\u003c/em\u003e and \u003cem\u003eCASP2\u003c/em\u003e, play a central role in pathways that have been linked to ketamine\u0026rsquo;s antidepressant effect in previous research. Both have been shown to play a role in the mTOR/VEGF/BDNF-connection, which has been proposed as a key-element in ketamine\u0026rsquo;s antidepressant mode of action.\u003c/p\u003e \u003cp\u003eTherefore, these data suggest that DNA methylation is a dynamic biomarker that reflects the clinically observed relationship between dosing paradigm and response, both in the context of ketamine as an active antidepressant treatment, as well as independent of the therapeutic agent. Despite the availability of longitudinal analyses, causality remains unclear. The correlation suggested by our analysis does not answer the question of whether DNA methylation precedes recovery and is part of the recovery process or does recovery itself lead to changes in DNA methylation. As recovery itself is highly complex involving multiple biological systems, both might also be parallel processes, that occur independently, driven by a shared impulse, such as immunological changes (\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e, \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eOur study has some limitations. One limitation is the relatively small sample size as mentioned above. We expect the relationship between DNA-methylation and response-guided dosing to be more complex than we could show with our current investigation. We also expect a polygenic contribution of DNA methylation on the relationship between dosing paradigm and response at a molecular and clinical level. A larger sample size could contribute to the signal by exposing multiple smaller effects. Hence, our sample size does not allow a more in-depth analysis or thorough mechanistic exploration. Nonetheless, by tackling the question at hand from multiple angles, we can argue that DNA methylation changes occur in parallel to response-guided dosing.\u003c/p\u003e \u003cp\u003eThe trade-off between the sample size and sample heterogeneity is a challenge for both depression research and the field of DNA methylation. With the prior clinical observation (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e), we chose to analyse stratified by pharmacological modality in favour of homogeneity. Combining treatment categories, e.g., fixed and flexible doses, could have possibly diluted the effect of pharmacotherapy on DNA methylation.\u003c/p\u003e \u003cp\u003eReducing heterogeneity and confounding effects is another limitation, in part inherent to DNA methylation research. It is hard to control sensitivity to environment and treatment-related exposures with real-live patients suffering from depression. By restricting ourselves to biological sex, age, site of data collection, estimated cell-types and ancestry, we anticipated the most relevant variability relevant to our research question, without overfitting the model. Body mass index and smoking are often associated with depression, as well as with the DNA methylation and metabolic fitness, but these were unavailable in our study dataset and hence these may be unmeasured confounding factors. Hence, we did not include these variables in the model. However, it is likely that the highly regulated format of an RCT provides an optimal setting, as the study design implies that confounding factors are randomized too. Though, smaller sample sizes are more prone to sampling biases, these can best be overcome by independent replication to test the reliability of our results in different patient groups. We are not aware of any similar ketamine samples to date suitable for independent replication. We hope that this study contributes to the motivation to investigate the dosing of ketamine and its relationship with the level of DNA methylation.\u003c/p\u003e \u003cp\u003eLastly, as shown in the fixed-dose analysis, although not statistically significant by our pre-set cut-off, our results returned many CpGs that are expected to have a regulatory function. In our interpretation we focused on the genes that were highlighted by our analyses. However, we want to stress that, particularly in a treatment context the role of DNA methylation as an epigenetic marker is most likely regulatory in nature, as it is expected to play an orchestrating role (\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e, \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e). Many questions remain unanswered in our model of response-guided dosing and its response at the DNA methylation level.\u003c/p\u003e \u003cp\u003eOverall, these data show that treatment with ketamine affects DNA methylation differently than the active control condition of midazolam. We could also show that flexible dosing of either medication can affect DNA methylation. The treatment-stratified longitudinal analyses also showed a significant DMR exclusively for the flexibly dosed ketamine group. For this condition more CpGs had a \u003cem\u003ep\u003c/em\u003e-value below 0.05 than for each of the other conditions. Our research argues in favour of dose-dependent DNA methylation changes for ketamine over time in the treatment of depression. More research is needed to replicate these findings and investigate the matter of causality: do DNA methylation changes contribute to recovery or are they the result of improved depression symptomatology?\u003c/p\u003e \u003cp\u003eBy investigating DNA methylation changes in a response-guided dosing paradigm, this study takes an initial step toward a better understanding of the relationship between epigenetic modulation and ketamine treatment success.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eCOI:\u003c/strong\u003e In the past 36 months, N.G. has received speaker's bureau honoraria from Servier Laboratories, Janssen and Lundbeck, and served on advisory boards for Servier Laboratories, Esia, Seqirus and Lundbeck. D.B. is a director and part-owner of Neurotrials Victoria Pty Ltd, trading as Neurocentrix and Neurocentrix TMS Pty Ltd; he serves on the advisory board for Eli Lilly and Janssen, and is currently supported by grant funding from Praxis, Janssen, Eli Lilly, Biogen and NHMRC; he has served on speaker panels for Servier, Janssen and Eli Lilly in the past 12 months; he is an investigator on the Janssen Quality of Life Esketamine study. B.T.B. has received grants and served as consultant, advisor or CME speaker for AstraZeneca, Bristol-Myers Squibb, Janssen, Lundbeck, Otsuka, Servier, the NHMRC, the Fay Fuller Foundation and the James and Diana Ramsay Foundation. In the past 3 years, P.B.F. has received equipment for research from Neurosoft, Nexstim and Brainsway Ltd; he has served on scientific advisory boards for Magstim and LivaNova and received speaker fees from Otsuka; he is a founder and board member for TMS Clinics Australia and Resonance Therapeutics. Within the last 36 months, P.G. has attended a Janssen New Zealand advisory board, and is named on a patent for a controlled release ketamine tablet developed by Douglas Pharmaceuticals. In the past 36 months, D.M. has received research consulting fees from Douglas Pharmaceuticals for a clinical trial involving ketamine. P.B.M. has received remuneration from Janssen (Australia) and Sanofi (Hangzhou) for lectures or advisory board membership within the past 3 years. M.B. has received honoraria from EPA Warsaw, Lundbeck, Controversias Barcelona, Servier, Medisquire, HealthEd, ANZJP, European Psychiatric Association, Janssen, Medplan, Milken Institute, Abbott India, ASCP, Allori for Eisai, Otsuka, St Bio Pharma and Sandoz in the past 3 years. G.C. has received educational and travel support from Servier, Astra Zeneca, Otsuka Australia, Merck Sharp \u0026amp; Dohme and Janssen-Cilag in the past 5 years; he also served on an advisory board for the AFFINITY trial. A.A.S. is a director of the Australian Medicines Handbook Pty Ltd (unpaid) and has received funding support from the Australian and New Zealand College of Anaesthetists to investigate ketamine for chronic postsurgical pain. S.H. has received speaker and consultancy fees from Janssen and Servier and served on advisory boards for Janssen and Lundbeck. C.K.L. is on the Clinical Advisory Board for Douglas Pharmaceuticals and has received fees for the following: Janssen Cilag advisory board, Douglas Pharmaceuticals advisory board. E.V.A., Ch.H., V.S., S.M.W., N.T.M, S.N., S.S., A.R., D.H.-P., A.A., V.D., M.L.H., C.M., M.L.C.: None.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements:\u003c/strong\u003e Evelien Van Assche thanks the Faculty of Medicine of the University of Münster for the research support (flexible Forschungszeit). We sincerely thank all the participants who were involved in the study, and all investigators and staff at the study centers.\u003c/p\u003e\n\u003cp\u003eThe study was funded by a competitive research grant from the Australian National Health and Medical Research Council (APP1105089). M.B. is supported by a NHMRC Leadership 3 Investigator grant (2017131). C.K.L. is supported by a NHMRC Leadership Investigator grant (1195651).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSupplementary information:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSupplementary information is available at MP’s website.\u003c/p\u003e\n\u003cp\u003eThe supplementary information includes a more detailed sample description and outline of the RCT, as well as details on the QC. It also includes information on sensitivity analyses, as well as additional information to interpret the reported results. Tables that provide more details on reported results are also included.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe supplementary information consists of one .pdf file.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eMcIntyre RS, Rosenblat JD, Nemeroff CB, Sanacora G, Murrough JW, Berk M, et al. Synthesizing the Evidence for Ketamine and Esketamine in Treatment-Resistant Depression: An International Expert Opinion on the Available Evidence and Implementation. AJP. 2021;178(5):383\u0026ndash;99.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLoo C, Glozier N, Barton D, Baune BT, Mills NT, Fitzgerald P, et al. Efficacy and safety of a 4-week course of repeated subcutaneous ketamine injections for treatment-resistant depression (KADS study): randomised double-blind active-controlled trial. Br J Psychiatry. 223(6):533\u0026ndash;41.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMoore LD, Le T, Fan G. DNA Methylation and Its Basic Function. Neuropsychopharmacol. 2013;38(1):23\u0026ndash;38.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eArčan IŠ, Kouter K, Paska AV. Depressive disorder and antidepressants from an epigenetic point of view. World Journal of Psychiatry. 2022;12(9):1150.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJu C, Fiori LM, Belzeaux R, Theroux JF, Chen GG, Aouabed Z, et al. Integrated genome-wide methylation and expression analyses reveal functional predictors of response to antidepressants. Transl Psychiatry. 2019;9(1):1\u0026ndash;12.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVan Assche E, Hohoff C, Zang J, Knight MJ, Baune BT. Longitudinal early epigenomic signatures inform molecular paths of therapy response and remission in depressed patients. Front Mol Neurosci. 2023;16:1223216.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZang JCS, Hohoff C, Van Assche E, Lange P, Kraft M, Sandmann S, et al. Immune gene co-expression signatures implicated in occurence and persistence of cognitive dysfunction in depression. Prog Neuropsychopharmacol Biol Psychiatry. 2023;127:110826.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFlynn LT, Gao WJ. DNA methylation and the opposing NMDAR dysfunction in schizophrenia and major depression disorders: a converging model for the therapeutic effects of psychedelic compounds in the treatment of psychiatric illness. Mol Psychiatry. 2023;28(11):4553\u0026ndash;67.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJu LS, Yang JJ, Lei L, Xia JY, Luo D, Ji MH, et al. The Combination of Long-term Ketamine and Extinction Training Contributes to Fear Erasure by Bdnf Methylation. Front Cell Neurosci. 2017;11. Available from: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.frontiersin.org/journals/cellular-neuroscience/articles/\u003c/span\u003e\u003cspan address=\"https://www.frontiersin.org/journals/cellular-neuroscience/articles/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fncel.2017.00100/full\u003c/span\u003e\u003cspan address=\"10.3389/fncel.2017.00100/full\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMontgomery SA, \u0026Aring;sberg M. A New Depression Scale Designed to be Sensitive to Change. The British Journal of Psychiatry. 1979;134(4):382\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAssenov Y, M\u0026uuml;ller F, Lutsik P, Walter J, Lengauer T, Bock C. Comprehensive analysis of DNA methylation data with RnBeads. Nature Methods. 2014;11(11):1138\u0026ndash;40.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRnBeads 2.0: comprehensive analysis of DNA methylation data | Genome Biology | Full Text [Internet]. [cited 2023 Apr 13]. Available from: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://genomebiology.biomedcentral.com/articles/\u003c/span\u003e\u003cspan address=\"https://genomebiology.biomedcentral.com/articles/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s13059-019-1664-9\u003c/span\u003e\u003cspan address=\"10.1186/s13059-019-1664-9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePeterson RE, Kuchenbaecker K, Walters RK, Chen CY, Popejoy AB, Periyasamy S, et al. Genome-wide Association Studies in Ancestrally Diverse Populations: Opportunities, Methods, Pitfalls, and Recommendations. Cell. 2019;179(3):589\u0026ndash;603.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSalas LA, Koestler DC, Butler RA, Hansen HM, Wiencke JK, Kelsey KT, et al. An optimized library for reference-based deconvolution of whole-blood biospecimens assayed using the Illumina HumanMethylationEPIC BeadArray. Genome Biology. 2018;19(1):64.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHouseman EA, Molitor J, Marsit CJ. Reference-free cell mixture adjustments in analysis of DNA methylation data. Bioinformatics (Oxford, England). 2014;30(10):1431\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXu Z, Xie C, Taylor JA, Niu L. ipDMR: identification of differentially methylated regions with interval P-values. Bioinformatics. 2021;37(5):711\u0026ndash;3.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePurcell SM, Wray NR, Stone JL, Visscher PM, O\u0026rsquo;Donovan MC, Sullivan PF, et al. Common polygenic variation contributes to risk of schizophrenia and bipolar disorder. Nature. 2009;460(7256):748\u0026ndash;52.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchork AJ, Thompson WK, Pham P, Torkamani A, Roddey JC, Sullivan PF, et al. All SNPs are not created equal: genome-wide association studies reveal a consistent pattern of enrichment among functionally annotated SNPs. Gibson G, editor. PLoS genetics. 2013;9(4):e1003449\u0026ndash;e1003449.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCampagna MP, Xavier A, Lechner-Scott J, Maltby V, Scott RJ, Butzkueven H, et al. Epigenome-wide association studies: current knowledge, strategies and recommendations. Clin Epigenet. 2021;13(1):214.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLee BT, Barber GP, Benet-Pag\u0026egrave;s A, Casper J, Clawson H, Diekhans M, et al. The UCSC Genome Browser database: 2022 update. Nucleic Acids Research. 2022;50(D1):D1115\u0026ndash;22.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKishore R, Arnaboldi V, Van Slyke CE, Chan J, Nash RS, Urbano JM, et al. Automated generation of gene summaries at the Alliance of Genome Resources. Database. 2020;2020:baaa037.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSafran M, Rosen N, Twik M, BarShir R, Stein TI, Dahary D, et al. The GeneCards Suite. In: Abugessaisa I, Kasukawa T, editors. Practical Guide to Life Science Databases. Singapore: Springer Nature; 2021. p. 27\u0026ndash;56. Available from: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/978-981-16-5812-9_2\u003c/span\u003e\u003cspan address=\"10.1007/978-981-16-5812-9_2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCui LB, Zhao SW, Zhang YH, Chen K, Fu YF, Qi T, et al. Associated transcriptional, brain and clinical variations in schizophrenia. Nat Mental Health. 2024;2(10):1239\u0026ndash;49.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHan Y, Gu S, Li Y, Qian X, Wang F, Huang JH. Neuroendocrine pathogenesis of perimenopausal depression. Front Psychiatry [Internet]. 2023 Mar 30 [cited 2024 Nov 24];14. Available from: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.frontiersin.org/journals/psychiatry/articles/\u003c/span\u003e\u003cspan address=\"https://www.frontiersin.org/journals/psychiatry/articles/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fpsyt.2023.1162501/full\u003c/span\u003e\u003cspan address=\"10.3389/fpsyt.2023.1162501/full\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLauterbach EC. Psychotropic Drug Effects on Gene Transcriptomics Relevant to Alzheimer Disease. Alzheimer Disease \u0026amp; Associated Disorders. 2012;26(1):1.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhu Y, Li Q. Multifaceted roles of PDCD6 both within and outside the cell. Journal of Cellular Physiology. 2024;239(5):e31235.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRho SB, Song YJ, Lim MC, Lee SH, Kim BR, Park SY. Programmed cell death 6 (PDCD6) inhibits angiogenesis through PI3K/mTOR/p70S6K pathway by interacting of VEGFR-2. Cellular Signalling. 2012;24(1):131\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen Y, Guan W, Wang ML, Lin XY. PI3K-AKT/mTOR Signaling in Psychiatric Disorders: A Valuable Target to Stimulate or Suppress? International Journal of Neuropsychopharmacology. 2024;27(2):pyae010.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKato T. Role of mTOR1 signaling in the antidepressant effects of ketamine and the potential of mTORC1 activators as novel antidepressants. Neuropharmacology. 2023;223:109325.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCavalleri L, Merlo Pich E, Millan MJ, Chiamulera C, Kunath T, Spano PF, et al. Ketamine enhances structural plasticity in mouse mesencephalic and human iPSC-derived dopaminergic neurons via AMPAR-driven BDNF and mTOR signaling. Mol Psychiatry. 2018;23(4):812\u0026ndash;23.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi QS, Wajs E, Ochs-Ross R, Singh J, Drevets WC. Genome-wide association study and polygenic risk score analysis of esketamine treatment response. Sci Rep. 2020;10(1):12649.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTordera RM, Garcia-Garc\u0026iacute;a AL, Elizalde N, Segura V, Aso E, Venzala E, et al. Chronic stress and impaired glutamate function elicit a depressive-like phenotype and common changes in gene expression in the mouse frontal cortex. European Neuropsychopharmacology. 2011;21(1):23\u0026ndash;32.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSasaki T, Iwata S, Okano HJ, Urasaki Y, Hamada J, Tanaka H, et al. Nedd9 Protein, a Cas-L Homologue, Is Upregulated After Transient Global Ischemia in Rats. Stroke. 2005;36(11):2457\u0026ndash;62.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSoiza-Reilly M, Meye FJ, Olusakin J, Telley L, Petit E, Chen X, et al. SSRIs target prefrontal to raphe circuits during development modulating synaptic connectivity and emotional behavior. Mol Psychiatry. 2019;24(5):726\u0026ndash;45.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen T, Giri M, Xia Z, Subedi YN, Li Y. Genetic and epigenetic mechanisms of epilepsy: a review. Neuropsychiatric Disease and Treatment. 2017;13:1841\u0026ndash;59.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMula M. The safety and tolerability of intranasal midazolam in epilepsy. Expert Review of Neurotherapeutics. 2014;14(7):735\u0026ndash;40.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang C, Dong N, Xu S, Ma H, Cheng M. Identification of hub genes and construction of diagnostic nomogram model in schizophrenia. Front Aging Neurosci [Internet]. 2022 Oct 14 [cited 2025 Jan 12];14. Available from: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.frontiersin.org/journals/aging-neuroscience/articles/\u003c/span\u003e\u003cspan address=\"https://www.frontiersin.org/journals/aging-neuroscience/articles/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fnagi.2022.1032917/full\u003c/span\u003e\u003cspan address=\"10.3389/fnagi.2022.1032917/full\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGajjar S, Patel BM. Neuromedin: An insight into its types, receptors and therapeutic opportunities. Pharmacological Reports. 2017;69(3):438\u0026ndash;47.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePałasz A, Worthington JJ, Filipczyk Ł, Saganiak K. Pharmacomodulation of brain neuromedin U signaling as a potential therapeutic strategy. Journal of Neuroscience Research. 2023;101(11):1728\u0026ndash;36.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhao R, Sun JB, Deng H, Cheng C, Li X, Wang FM, et al. Per1 gene polymorphisms influence the relationship between brain white matter microstructure and depression risk. Front Psychiatry. 2022;13. Available from: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.frontiersin.org/journals/psychiatry/articles/\u003c/span\u003e\u003cspan address=\"https://www.frontiersin.org/journals/psychiatry/articles/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fpsyt.2022.1022442/full\u003c/span\u003e\u003cspan address=\"10.3389/fpsyt.2022.1022442/full\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBunney BG, Li JZ, Walsh DM, Stein R, Vawter MP, Cartagena P, et al. Circadian dysregulation of clock genes: clues to rapid treatments in major depressive disorder. Mol Psychiatry. 2015;20(1):48\u0026ndash;55.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDrange OK, Smeland OB, Shadrin AA, Finseth PI, Witoelar A, Frei O, et al. Genetic Overlap Between Alzheimer\u0026rsquo;s Disease and Bipolar Disorder Implicates the MARK2 and VAC14 Genes. Front Neurosci. 2019;13. Available from: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.frontiersin.org/journals/neuroscience/articles/\u003c/span\u003e\u003cspan address=\"https://www.frontiersin.org/journals/neuroscience/articles/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fnins.2019.00220/full\u003c/span\u003e\u003cspan address=\"10.3389/fnins.2019.00220/full\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGuan W, Ni MX, Gu HJ, Yang Y. CREB: A Promising Therapeutic Target for Treating Psychiatric Disorders. Current Neuropharmacology. 2024;22(14):2384\u0026ndash;401.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBlendy JA. The Role of CREB in Depression and Antidepressant Treatment. Biological Psychiatry. 2006;59(12):1144\u0026ndash;50.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKohshour MO, Papiol S, Ching CRK, Schulze TG. Genomic and neuroimaging approaches to bipolar disorder. BJPsych Open. 2022;8(2):e36.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShinoda Y, Sadakata T, Nakao K, Katoh-Semba R, Kinameri E, Furuya A, et al. Calcium-dependent activator protein for secretion 2 (CAPS2) promotes BDNF secretion and is critical for the development of GABAergic interneuron network. Proceedings of the National Academy of Sciences. 2011;108(1):373\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAr\u0026eacute;valo JC, Deogracias R. Mechanisms Controlling the Expression and Secretion of BDNF. Biomolecules. 2023;13(5):789.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYoo H, Yang SH, Kim JY, Yang E, Park HS, Lee SJ, et al. Down-regulation of habenular calcium-dependent secretion activator 2 induces despair-like behavior. Scientific Reports. 2021;11:3700.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStarnawska A, Tan Q, Soerensen M, McGue M, Mors O, B\u0026oslash;rglum AD, et al. Epigenome-wide association study of depression symptomatology in elderly monozygotic twins. Transl Psychiatry. 2019;9(1):1\u0026ndash;14.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSampson E, Mills NT, Hori H, Cearns M, Schwarte K, Hohoff C, et al. Long-term characterisation of the relationship between change in depression severity and change in inflammatory markers following inflammation-stratified treatment with vortioxetine augmented with celecoxib or placebo. Brain, Behavior, and Immunity. 2025;123:43\u0026ndash;56.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKavakbasi E, Van Assche E, Schwarte K, Hohoff C, Baune BT. Long-Term Immunomodulatory Impact of VNS on Peripheral Cytokine Profiles and Its Relationship with Clinical Response in Difficult-to-Treat Depression (DTD). International Journal of Molecular Sciences. 2024;25(8):4196.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLeenen FAD, Muller CP, Turner JD. DNA methylation: conducting the orchestra from exposure to phenotype? Clin Epigenet. 2016;8(1):92.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMangiavacchi A, Morelli G, Orlando V. Behind the scenes: How RNA orchestrates the epigenetic regulation of gene expression. Front Cell Dev Biol. 2023;11. Available from: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.frontiersin.org/journals/cell-and-developmental-biology/articles/\u003c/span\u003e\u003cspan address=\"https://www.frontiersin.org/journals/cell-and-developmental-biology/articles/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fcell.2023.1123975/full\u003c/span\u003e\u003cspan address=\"10.3389/fcell.2023.1123975/full\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003eTable 1. Top 10 differentially methylated CpGs between flexibly dosed cohorts.\u003c/p\u003e\n\u003cdiv\u003e\n \u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"600\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 75px;\"\u003e\n \u003cp\u003eCpG\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 95px;\"\u003e\n \u003cp\u003ePosition\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 142px;\"\u003e\n \u003cp\u003eGenomic context (UCSC)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 47px;\"\u003e\n \u003cp\u003eMean Mid.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 47px;\"\u003e\n \u003cp\u003eMean Ket.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003eMethylation difference\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e\u003cem\u003ep\u003c/em\u003e-value\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 43px;\"\u003e\n \u003cp\u003eFDR \u003cem\u003ep\u003c/em\u003e-value\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 75px;\"\u003e\n \u003cp\u003e\u003cstrong\u003ecg15945600\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 95px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eChr5: 304075\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003eNorth Shelf\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 142px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003ePDCD6\u003c/em\u003e\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;gene-body\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 47px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.84\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 47px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.82\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.023; Ketamine Hypomethylated\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e7.28x10\u003csup\u003e-8\u003c/sup\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 43px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.049\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 75px;\"\u003e\n \u003cp\u003ecg13615030\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 95px;\"\u003e\n \u003cp\u003eChr9: 130639783\u003c/p\u003e\n \u003cp\u003eIsland\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 142px;\"\u003e\n \u003cp\u003e\u003cem\u003eAK1\u0026nbsp;\u003c/em\u003e(promotor linked 1 variant)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 47px;\"\u003e\n \u003cp\u003e0.12\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 47px;\"\u003e\n \u003cp\u003e0.10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003e0.017; Ketamine Hypomethylated\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e2.24x10\u003csup\u003e-6\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 43px;\"\u003e\n \u003cp\u003e0.748\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 75px;\"\u003e\n \u003cp\u003ecg15323881\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 95px;\"\u003e\n \u003cp\u003eChr2: 9768508\u003c/p\u003e\n \u003cp\u003eNorth Shore\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 142px;\"\u003e\n \u003cp\u003e\u003cem\u003eYWHAQ\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 47px;\"\u003e\n \u003cp\u003e0.82\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 47px;\"\u003e\n \u003cp\u003e0.79\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003e0.033; Ketamine Hypomethylated\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e4.03x10\u003csup\u003e-6\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 43px;\"\u003e\n \u003cp\u003e0.845\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 75px;\"\u003e\n \u003cp\u003ecg08951301\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 95px;\"\u003e\n \u003cp\u003eChr12: 19865160\u003c/p\u003e\n \u003cp\u003eOpen Sea\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 142px;\"\u003e\n \u003cp\u003eENCODE Candidate Cis-Regulatory Element\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 47px;\"\u003e\n \u003cp\u003e0.22\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 47px;\"\u003e\n \u003cp\u003e0.18\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003e0.041; Ketamine Hypomethylated\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e5.07 x10\u003csup\u003e-6\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 43px;\"\u003e\n \u003cp\u003e0.845\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 75px;\"\u003e\n \u003cp\u003ecg15980099\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 95px;\"\u003e\n \u003cp\u003eChr6: 137166932\u003c/p\u003e\n \u003cp\u003eOpen Sea\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 142px;\"\u003e\n \u003cp\u003e\u003cem\u003ePEX7\u003c/em\u003e gene body\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 47px;\"\u003e\n \u003cp\u003e0.87\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 47px;\"\u003e\n \u003cp\u003e0.89\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003e-0.023; Ketamine Hypermethylated\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e7.91 x10\u003csup\u003e-6\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 43px;\"\u003e\n \u003cp\u003e1.000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 75px;\"\u003e\n \u003cp\u003ecg08056716\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 95px;\"\u003e\n \u003cp\u003eChr19: 35397106\u003c/p\u003e\n \u003cp\u003eSouth Shore\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 142px;\"\u003e\n \u003cp\u003eLINC01838/LINC00904 (long intergenic non-coding RNA)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 47px;\"\u003e\n \u003cp\u003e0.78\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 47px;\"\u003e\n \u003cp\u003e0.76\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003e0.020; Ketamine Hypomethylated\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e9.14 x10\u003csup\u003e-6\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 43px;\"\u003e\n \u003cp\u003e1.000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 75px;\"\u003e\n \u003cp\u003ecg15686281\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 95px;\"\u003e\n \u003cp\u003eChr1: 44097898\u003c/p\u003e\n \u003cp\u003eOpen Sea\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 142px;\"\u003e\n \u003cp\u003eENCODE candidate Cis-Regulatory Element; nearest to \u003cem\u003ePTPRF\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 47px;\"\u003e\n \u003cp\u003e0.87\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 47px;\"\u003e\n \u003cp\u003e0.89\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003e-0.013; Ketamine Hypermethylated\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e1.34 x10\u003csup\u003e-5\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 43px;\"\u003e\n \u003cp\u003e1.000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 75px;\"\u003e\n \u003cp\u003ecg00725221\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 95px;\"\u003e\n \u003cp\u003eChr9: 72789212\u003c/p\u003e\n \u003cp\u003eOpen Sea\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 142px;\"\u003e\n \u003cp\u003e\u003cem\u003eMAMDC2/MAMDC2-AS1\u003c/em\u003e gene body\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 47px;\"\u003e\n \u003cp\u003e0.90\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 47px;\"\u003e\n \u003cp\u003e0.92\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003e-0.012; Ketamine Hypermethylated\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e1.38 x10\u003csup\u003e-5\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 43px;\"\u003e\n \u003cp\u003e1.000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 75px;\"\u003e\n \u003cp\u003ecg23063825\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 95px;\"\u003e\n \u003cp\u003eChr12: 298151\u003c/p\u003e\n \u003cp\u003eNorth Shore\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 142px;\"\u003e\n \u003cp\u003enearest to \u003cem\u003eSLC6A12\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 47px;\"\u003e\n \u003cp\u003e0.34\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 47px;\"\u003e\n \u003cp\u003e0.43\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003e-0.093; Ketamine Hypermethylated\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e1.67 x10\u003csup\u003e-5\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 43px;\"\u003e\n \u003cp\u003e1.000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 75px;\"\u003e\n \u003cp\u003ecg09648702\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 95px;\"\u003e\n \u003cp\u003eChr11: 20184779\u003c/p\u003e\n \u003cp\u003eIsland\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 142px;\"\u003e\n \u003cp\u003eENSG00000294773/ Regulatory Elements/ DBX1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 47px;\"\u003e\n \u003cp\u003e0.085\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 47px;\"\u003e\n \u003cp\u003e0.074\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003e0.011; Ketamine Hypomethylated\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e1.81 x10\u003csup\u003e-5\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 43px;\"\u003e\n \u003cp\u003e1.000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"translational-psychiatry","isNatureJournal":false,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"tp","sideBox":"Learn more about [Translational Psychiatry](http://www.nature.com/tp/)","snPcode":"41398","submissionUrl":"https://mts-tp.nature.com/cgi-bin/main.plex","title":"Translational Psychiatry","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-6778101/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6778101/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"DNA methylation is a dynamic biomarker suited to investigate processes, e.g., treatment response. We investigate DNA methylation changes in different dosing paradigms for ketamine (s.c.) vs. midazolam (s.c.) in treatment resistant depressed patients (the Ketamine for Adult Depression Study – KADS). After a 4-week randomized controlled trial (RCT) DNA methylation data (Illumina Infinium MethylationEPIC 850k BeadChip) were available for 87 patients, with 76 having DNA methylation data at baseline too. Patients received either ketamine or midazolam. Initially dosing was fix, however, following a protocol amendment, newly recruited patients received the clinically more effective response-guided ‘flexible’ dosing. We performed cross-sectional and paired longitudinal DNA methylation analyses, the latter with focus on differentially methylated regions (DMR), comparing treatment with ketamine and midazolam, as well as fixed and flexible dosing. We used R-packages RnBeads and comb-p for quality control and statistical analyses, and DMR identification, respectively. P-values were False-discovery-rate (FDR)-corrected (Benjamini Hochberg). Only the response-guided cohort in the ketamine vs. midazolam comparison returned one epigenome-wide significant CpG (cg15945600; PDCD6; p-valFDR \u003c 0.05). Flexible dosing (midazolam + ketamine) vs. fixed dosing (midazolam + ketamine) returned a suggestive hit (cg20023762, NEDD9; p-valFDR \u003c 0.10). The flexibly dosed ketamine condition returned the only statistically significant DMR in the longitudinal analyses (p-valFDR = 0.0045; CAPS2/GLIPR1L2). The clinically more effective response-guided paradigm has a correlate at the DNA methylation level with most pronounced DNA methylation changes in the flexibly dosed ketamine group. Pharmaco-epigenomics in Psychiatry, as a growing field, facilitates interpretation of DNA-methylation dynamics for treatment response in depression.","manuscriptTitle":"DNA methylation changes in a pharmaco-epigenomic EWAS in depression: comparing fixed and response-guided dosing paradigms for ketamine in the KADS trial","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-06-10 14:59:22","doi":"10.21203/rs.3.rs-6778101/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"transferred","content":"Translational Psychiatry","date":"2025-10-03T11:45:53+00:00","index":"","fulltext":""},{"type":"decision","content":"Reject after peer review","date":"2025-07-01T09:38:57+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"This content is not available.","date":"2025-06-30T03:37:50+00:00","index":3,"fulltext":"This content is not available."},{"type":"editorInvitedReview","content":"This content is not available.","date":"2025-06-18T13:33:50+00:00","index":2,"fulltext":"This content is not available."},{"type":"reviewerAgreed","content":"This content is not available.","date":"2025-06-15T10:03:21+00:00","index":3,"fulltext":"This content is not available."},{"type":"reviewerAgreed","content":"This content is not available.","date":"2025-06-10T08:02:08+00:00","index":2,"fulltext":"This content is not available."},{"type":"editorInvitedReview","content":"This content is not available.","date":"2025-06-06T15:45:21+00:00","index":1,"fulltext":"This content is not available."},{"type":"reviewerAgreed","content":"This content is not available.","date":"2025-06-05T16:36:18+00:00","index":1,"fulltext":"This content is not available."},{"type":"reviewersInvited","content":"","date":"2025-06-04T23:05:58+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-06-02T13:13:34+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-06-02T11:49:41+00:00","index":"","fulltext":""},{"type":"submitted","content":"Molecular Psychiatry","date":"2025-05-30T17:10:39+00:00","index":"","fulltext":""},{"type":"checksFailed","content":"","date":"2025-05-30T15:02:46+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"translational-psychiatry","isNatureJournal":false,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"tp","sideBox":"Learn more about [Translational Psychiatry](http://www.nature.com/tp/)","snPcode":"41398","submissionUrl":"https://mts-tp.nature.com/cgi-bin/main.plex","title":"Translational Psychiatry","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"63a77dfb-43e2-45c7-aa50-7ec50a1251d3","owner":[],"postedDate":"June 10th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":49548951,"name":"Biological sciences/Molecular biology"},{"id":49548952,"name":"Health sciences/Biomarkers/Prognostic markers"}],"tags":[],"updatedAt":"2025-11-28T14:41:10+00:00","versionOfRecord":[],"versionCreatedAt":"2025-06-10 14:59:22","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6778101","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6778101","identity":"rs-6778101","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}