Abstract
We previously reported the association between DNA methylation (DNAm) of pro-inflammatory cytokine genes and age. In addition, neurotrophic factors are known to be associated with age and neurocognitive disorders. Therefore, we hypothesized that DNAm of neurotrophic genes change with age, especially in delirium patients. DNAm were analyzed using the Illumina HumanMethylation450 or HumanMethylationEPIC BeadChip Kit in 3 independent cohorts: blood from 383 Grady Trauma Project subjects, brain from 21 neurosurgery patients, and blood from 87 inpatients with and without delirium. Both blood and brain samples showed that most of the DNAm of neurotrophic genes were positively correlated with age. Furthermore, DNAm of neurotrophic genes were more positively correlated with age in delirium cases than in non-delirium controls. These findings support our hypothesis that the neurotrophic genes may be epigenetically modulated with age, and this process may be contributing to the pathophysiology of delirium.
Keywords: Delirium, Age, Genome-wide DNA methylation, Neurotrophic factor, BDNF, GDNF
1. Introduction
Delirium is underdiagnosed and undertreated (Spronk et al., 2009), although it is common among elderly patients (Inouye, 2006). Various screening methods have been developed to characterize the epidemiology and risk factors of delirium, e.g., the Confusion Assessment Method (CAM) (Inouye et al., 1990) and Confusion Assessment Method for Intensive Care Unit (CAM-ICU) (Ely et al., 2001a; Ely et al., 2001b). Although these methods have excellent sensitivity and specificity in research settings, they are found to have suboptimal sensitivity (38–47%) in the context of “real world” intensive care units (ICUs) (Nishimura et al., 2016; van Eijk et al., 2011). Thus, among the elderly at high risk for delirium, biomarkers of delirium would aid in the prediction and treatment of delirium.
Previously we reported the association of DNA methylation (DNAm) changes in pro-inflammatory cytokine genes with age, indicating their possible role in the pathophysiology of delirium (Shinozaki et al., 2018a). This emphasized past findings of the importance of neuroinflammation in the development of delirium with the additional link of DNAm changes. However, the etiology of delirium is complex, with many facets leading up to its manifestation, and it is important to look at additional factors involved. Given the evidence of neurotrophic factors being associated with age-related behavior (Gunstad et al., 2008; Li et al., 2009), it is possible that DNAm of these genes may be differentially regulated in the context of age with delirium.
Neurotrophic factors are signaling molecules that contribute to neural plasticity, learning, and memory (Ghitza et al., 2010). Decreased serum levels of the brain-derived neurotrophic factor (BDNF) were associated with cognitive impairment in elderly people (Shimada et al., 2014). Serum levels of the glial cell line–derived neurotrophic factor (GDNF) as well as BDNF were decreased in subjects with mild cognitive impairment and Alzheimer’s disease (AD) (Forlenza et al., 2015). A Neuronal Per-Arnt-Sim domain protein 4 (NPAS4) knockdown mouse model showed increased cell death in cortical neurons (Choy et al., 2016). This in vivo investigation also revealed an increased lesion size and greater neurodegeneration after a photochemically induced stroke (Choy et al., 2016). The nuclear receptor subfamily 4A2 (NR4A2) critically regulates AD-related pathophysiology (Moon et al., 2019). An NR4A2 knockdown mouse model exacerbates AD symptoms/pathology of neuroinflammation/degeneration and plaque accumulation. Also, administration of a synthetic NR4A2 agonist significantly reduced cognitive impairments and plaque numbers in a mouse model of AD (Moon et al., 2019).
Advanced age is associated with drastically altered gene expression in the brain (Berchtold et al., 2008; Kang et al., 2011), and such changes are regulated by a variety of epigenetic modifications including DNAm (Zampieri et al., 2015), histone modifications (Gong et al., 2015), and micro RNA-mediated transcriptional control (Beveridge et al., 2014; Danka Mohammed et al., 2017). Indeed, DNAm patterns have been shown to change dynamically throughout the human lifespan (Numata et al., 2012), and epigenetic processes involved in age have been studied in detail (Christensen et al., 2009; Day et al., 2013; Pal and Tyler, 2016; Sen et al., 2016; Tsai et al., 2012). It has previously been shown that DNAm levels of BDNF are significantly correlated with age in blood (Ihara et al., 2018), but whether this is exacerbated or dysregulated in age-related disorders, such as delirium, needs to be studied.
Therefore, in this study, we analyzed samples from three cohorts to examine the degree of correlation between DNAm and age in neurotrophic genes. We hypothesized that DNAm in neurotrophic genes in both brain and blood would increase with age, and that these associations would be more significant among delirium patients. Our first aim was to primarily examine whether DNAm of additional neurotrophic genes are significantly correlated with age in blood and brain tissues in non-delirium populations. The second aim was to compare the degree DNAm levels of the neurotrophic genes were correlated in delirium cases versus controls.
2. Material and Methods
2.1. Grady Trauma Project (GTP) Cohort
2.1.1. GTP Cohort Subjects
Three hundred eighty-three subjects (age average = 41.50 years, SD = 12.93, age range = 18–77, female = 273, race = 380 African American, 2 Native American, and 1 other) were analyzed from the Grady Trauma Project (GTP) cohort. Detailed information of this cohort has been previously described (Smith et al., 2015).
2.1.2. GTP Sample Processing
Details of the sample processing is described in a previous study (Smith et al., 2015). Briefly, blood was collected in ethylenediaminetetraacetic acid (EDTA) vacuum tubes, and DNA was extracted with the Puregene Genomic DNA kit (Invitrogen, Carlsbad, CA).
2.1.3. GTP Methylome Analysis
The Illumina HumanMethylation450 (450K) was used to process DNAm levels. The 450K array covers over 450,000 CpG sites across the genome. Details of the methylome analysis is described in a previous study (Ratanatharathorn et al., 2017). Briefly, a quality control protocol, the Psychiatric Genomics Consortium (PGC) pipeline, was used. In the PGC pipeline, samples and probes were filtered out as previously described (Ratanatharathorn et al., 2017).
2.2. Neurosurgery (NSG) Cohort
2.2.1. NSG Cohort Subjects
Twenty-one subjects (age average = 31.0 years, age SD = 16.4, age range = 5–61 years) were analyzed from the neurosurgery (NSG) cohort. Details of this cohort are described in a previous study (Braun et al., 2019). Subjects participated between March 2014 and April 2017 at the University of Iowa Hospitals and Clinics. Resected brain tissue was collected from subjects with intractable epilepsy who underwent neurosurgery. Written informed consent was obtained from all of the participants. This study was approved by the University of Iowa’s Human Subjects Research Institutional Review Board.
2.2.2. NSG Sample Processing
Details of the sample processing are described in a previous study (Braun et al., 2019). Briefly, resected brain tissue samples were stored at −80 °C. Genomic DNA was extracted with the MasterPure™ DNA Purification kit (Epicentre, MCD85201) and bisulfite-converted with the EZ DNA Methylation™ Kit (Zymo Research, D5002).
2.2.3. NSG Methylome Analysis
The Infinium HumanMethylationEPIC BeadChip™ Kit (Illumina, WG-317–1002) (EPIC) was used to measure DNAm levels. The EPIC array is an updated version of the 450K array, and it can test over 850,000 probes, more than 90% of the probes on the 450K array (Pidsley et al., 2016). DNAm levels between probes that overlapped both 450K and EPIC array were highly correlated (r2 > 0.98) (Braun et al., 2019). Details of the methylome analysis are described in a previous study (Braun et al., 2019). Briefly, the R packages RnBeads (Assenov et al., 2014) and minfi (Aryee et al., 2014; Fortin et al., 2017) were used to process the raw data and perform quality control checks, data filtering, and normalization techniques. Probes were filtered out if they 1) overlapped within 5 bp of a single nucleotide polymorphism (SNP) (21,358 probes), 2) had a detection p-value greater than 0.01 or were considered unreliable measures based on RnBeads’s greedy-cut algorithm (18,864 probes), or 3) were context-specific sites (2,873 probes). Samples were normalized with beta mixture quantile dilation (Teschendorff et al., 2013). The final dataset contained 822,996 probes.
2.3. Epigenetics of Delirium (EOD) Cohort
2.3.1. EOD Cohort Subjects
A separate ongoing study of delirium was performed at the University of Iowa Hospitals and Clinics. Further characteristics of this cohort were detailed previously (Shinozaki et al., 2019; Shinozaki et al., 2018a; Shinozaki et al., 2018b). Ninety-two subjects participated between November 2017 and October 2018. Among them, eighty-seven subjects (age average = 70.2 years, age SD = 10.2, age range = 42–101 years, male; n = 60) were analyzed as described below. Written informed consent was obtained from all of the participants, and this study was approved by the University of Iowa’s Human Subjects Research Institutional Review Board.
2.3.2. Definition of Delirium Status
Details of the EOD cohort inclusion/exclusion criteria have been previously described (Shinozaki et al., 2018b). Briefly, participants were screened for delirium by hospital records, the Confusion Assessment Method for Intensive Care Unit (CAM-ICU) (Ely et al., 2001a), the Delirium Rating Scale - Revised-98 (DRS-R-98) (Trzepacz et al., 2001), and the Delirium Observation Screening Scale (DOSS) (Schuurmans et al., 2003). Dementia was recorded based on chart review. A final decision of delirium and dementia phenotyping was conducted by trained psychiatrist (G.S.).
2.3.3. EOD Sample Processing
For the EOD cohort, whole blood samples were collected with EDTA tubes, and all samples were stored at −80 °C. Genomic DNA was extracted with the MasterPure ™ DNA Purification kit (Epicentre, MCD85201) and bisulfite-converted with the EZ DNA Methylation™ Kit (Zymo Research, D5002).
2.3.4. EOD Methylome Analysis
The Infinium HumanMethylationEPIC BeadChip™ Kit (Illumina, WG-317–1002) was used to process DNAm levels of 93 samples (two samples were from one subject). The R packages ChAMP (Morris et al., 2014) and minfi (Aryee et al., 2014; Fortin et al., 2017) were used to process the raw data. Using the ChAMP package, one sample was filtered out because it had above 10% of sites with detection p-value greater than 0.01, and CpG sites were filtered out as described below. This resulted in 92 samples and 701,196 probes. Additional outlier samples were found with the density and multidimensional scaling plots. This included five samples, two of which were the duplicate samples from the same individual. The remaining 87 samples (43 samples were cases) were reprocessed with the ChAMP package. CpG sites were filtered out if they 1) had a detection p-value greater than 0.01 (12,903 probes), 2) had less than three beads in at least 5% of samples per probe (8,280 probes), 3) were non-CpG probes (2,911 probes), 4) were SNP-related probes (Zhou et al., 2017) (94,425 probes), 5) were cross-reactive probes (Nordlund et al., 2013) (11 probes), or 6) were located on the X or Y chromosome (16,356 probes). Samples were normalized with beta mixture quantile dilation (Teschendorff et al., 2013), and the Combat method was used to correct for batch effect (Johnson et al., 2007; Leek et al., 2019). The final dataset contained 731,032 probes.
2.4. Neurotrophic genes
In the GTP cohort, seven genes—BDNF, GDNF, activity regulated cytoskeleton associated protein (ARC), Fos Proto-Oncogene, AP-1 Transcription Factor Subunit (FOS), NPAS4, nuclear receptor subfamily 4A1 (NR4A1), and NR4A2—were selected to examine the association between age and DNAm levels at those genes because of their well investigated roles in a neurocognitive process. Although this is not an exhaustive list of genes relevant to that domain, we chose to focus on these genes to start with because of a practical limitation. In the NSG and EOD cohorts, we focused on the top four genes that showed genome-wide significance signals (p < 5 × E-8) associated with age in the GTP cohort.
2.5. Statistical Analysis
All statistical analyses were performed with R (R Core Team, 2019). In the GTP and EOD cohorts, the correlations between age and DNAm levels of neurotrophic genes in each CpG were tested by Pearson’s correlation analysis. In the NSG cohort, the correlation between age and DNAm levels of neurotrophic genes in each CpG was tested by Spearman’s rank correlation analysis because of the small sample size. Continuous data was calculated with an unpaired t-test, and categorical data was calculated with the chi-square test, to compare between delirium cases and controls among the EOD cohort. A p-value of less than 0.05 was determined to be nominally significant.
3. Results
3.1. DNA methylation from GTP cohort blood samples
Table 1 shows the demographic characteristics of each cohort. In the processed GTP methylome dataset, there are 226 CpGs annotated to be in the seven neurotrophic genes. Of these, DNAm of 16 CpGs were significantly associated with age at genome-wide significance levels (p < 5 × E-8), including seven CpGs in BDNF and four CpGs in GDNF. The top hits were at a CpG in GDNF (cg02328239; p = 1.54 × E-20) and a CpG in BDNF (cg05733135; p = 4.49 × E-15) (Table 2, Figure 1). Among the top 53 CpGs (Bonferroni-corrected for 226 CpGs, significant level p < 0.05/226 = 2.21 × E-4), 49 CpGs (92.5%) were positively correlated with age (Table 2). One hundred nine CpGs were associated with age at nominal significance (p < 0.05). Correlations between age and DNAm levels at nominal significant 109 CpGs were shown in Supplementary Table 1.
Table 1:
Demographic characteristics of each cohort
| Cohort | GTP | NSG | EOD |
|
|---|---|---|---|---|
| Delirium | Control | |||
| Tissue | Blood | Brain | Blood | |
| Array | 450K | EPIC | EPIC | |
| N | 383 | 21 | 43 | 44 |
| Female N (%) | 273 (71.3) | 7 (33.3) | 13 (30.2) | 14 (31.8) |
| Age in years, mean (SD) | 41.5 (12.9) | 31.0 (16.0) | 70.5 (10.7) | 69.9 (9.8) |
| Age range, years | 18–77 | 5–61 | 42–89 | 53–101 |
| DRS-R-98, mean (SD) | 16.3 (6.3) * | 6.1 (3.8) | ||
| DOSS, mean (SD) | 4.2 (3.3) * | 0.4 (1.6) | ||
| Dementia N (%) | 11 (25.6) * | 2 (4.5) | ||
Notes:
Significant vs control (p < 0.01).
Abbreviations: 450K, Illumina HumanMethylation450; DOSS, Delirium Observation Screening Scale; DRS-R-98, Delirium Rating Scale—Revised-98; EOD, Epigenetics of Delirium; EPIC, Infinium HumanMethylationEPIC; GTP, Grady Trauma Project; NSG; neurosurgery.
Table 2:
Correlations between age and DNAm levels of neurotrophic genes in blood samples obtained from the GTP cohort
| Gene | CpG | r | p-value |
|---|---|---|---|
| GDNF1 | cg02328239**, † | 0.47 | 1.54E-20 |
| BDNF | cg05733135**, ‡ | 0.41 | 4.49E-15 |
| GDNF | cg00049047** | 0.34 | 7.74E-13 |
| BDNF | cg13974632** | 0.37 | 1.06E-12 |
| BDNF | cg06816235**, ‡‡ | 0.36 | 1.88E-12 |
| BDNF | cg22043168** | 0.35 | 7.92E-12 |
| NR4A2 | cg11358945** | 0.37 | 9.50E-12 |
| NPAS4 | cg14699728** | 0.34 | 1.86E-10 |
| GDNF | cg07442479** | 0.35 | 1.88E-10 |
| BDNF | cg23947039** | 0.32 | 1.33E-09 |
| BDNF | cg26949694**, † | 0.30 | 1.95E-09 |
| NR4A2 | cg00194126** | 0.30 | 1.14E-08 |
| NR4A1 | cg26446929** | −0.27 | 1.92E-08 |
| BDNF | cg03167496** | 0.31 | 2.25E-08 |
| GDNF | cg26789779** | 0.30 | 2.99E-08 |
| ARC | cg16922810** | −0.30 | 4.11E-08 |
| BDNF | cg01583131* | 0.29 | 8.73E-08 |
| BDNF | cg25412831* | 0.25 | 1.34E-07 |
| NR4A2 | cg14811105* | 0.29 | 1.47E-07 |
| GDNF | cg21590264* | 0.29 | 1.94E-07 |
| BDNF | cg25962210* | 0.30 | 3.58E-07 |
| NR4A2 | cg02964555* | 0.26 | 3.58E-07 |
| NR4A2 | cg21226516* | 0.28 | 4.47E-07 |
| BDNF | cg11718030* | 0.27 | 1.01E-06 |
| BDNF | cg17413943* | 0.23 | 1.41E-06 |
| GDNF | cg25602684* | 0.28 | 1.59E-06 |
| GDNF | cg07715201* | 0.24 | 2.40E-06 |
| GDNF | cg14492800* | 0.29 | 2.62E-06 |
| GDNF | cg08204023* | 0.26 | 2.73E-06 |
| NPAS4 | cg04768203* | 0.26 | 5.33E-06 |
| GDNF | cg26295057* | 0.24 | 9.56E-06 |
| GDNF | cg18182111* | 0.21 | 1.31E-05 |
| BDNF | cg04481212* | 0.22 | 1.39E-05 |
| BDNF | cg21010859* | 0.22 | 1.51E-05 |
| GDNF | cg21319053* | 0.27 | 2.06E-05 |
| BDNF | cg24249411* | 0.20 | 2.21E-05 |
| FOS | cg12061886* | 0.18 | 3.64E-05 |
| GDNF | cg04209913* | 0.19 | 3.66E-05 |
| GDNF | cg18725867* | 0.24 | 3.78E-05 |
| GDNF | cg04442426* | −0.26 | 4.82E-05 |
| BDNF | cg06684850* | 0.23 | 5.29E-05 |
| GDNF | cg05330056* | −0.44 | 5.73E-05 |
| BDNF | cg06991510* | 0.19 | 6.11E-05 |
| BDNF | cg01225698* | 0.19 | 6.57E-05 |
| GDNF | cg14590843* | 0.17 | 8.41E-05 |
| ARC | cg19438565* | 0.23 | 9.43E-05 |
| NR4A2 | cg00426720* | 0.21 | 1.03E-04 |
| NR4A2 | cg10089963* | 0.19 | 1.14E-04 |
| GDNF | cg07423205* | 0.19 | 1.40E-04 |
| BDNF | cg25328597* | 0.23 | 1.47E-04 |
| BDNF | cg18867480* | 0.21 | 1.52E-04 |
Notes: Blue highlights a positive correlation, and pink highlights a negative correlation. r = Pearson’s correlation coefficient.
Significant after Bonferroni-corrected genome-wide significance levels (p < 5 × E-8),
Significant after correction for multiple testing level (p < 0.05/226 = 2.21 × E-4),
Overlapped with the significant CpGs at multiple-testing significance level (p < 2.49 × E-4) in the NSG cohort,
Overlapped with the nominally significant (p < 0.05) CpGs in the NSG and EOD cohort,
Overlapped with the nominally significant (p < 0.05) CpGs in the EOD cohort.
Abbreviations: AP-1, Transcription Factor Subunit; ARC, activity-regulated cytoskeleton-associated protein; BDNF, brain-derived neurotrophic factor; FOS, Fos Proto-Oncogene; GDNF, glial cell–derived neurotrophic factor; GTP, Grady Trauma Project; NPAS4, neuronal Per-Arnt-Sim domain protein 4; NR4A1; nuclear receptor subfamily 4A1; NR4A2, nuclear receptor subfamily 4A2
Figure 1.
Correlation between age and DNA methylation level of the top two CpGs in the GTP cohort.
Abbreviations: BDNF, Brain-derived neurotrophic factor; GDNF, glial cell-derived neurotrophic factor; GTP, Grady Trauma Project.
3.2. NSG study samples, DNAm in brain
We next tested whether these associations of DNAm and age in blood samples in the GTP cohort were also present in brain tissue in our NSG cohort (N = 21). We focused on the top four genes based on the 16 CpGs significantly associated with age at genome-wide significant levels in the GTP cohort (BDNF, GDNF, NR4A2, and NPAS4). There were 201 CpGs annotated to be in these four genes. In BDNF, 18 CpGs were positively correlated with age at nominal significance (p < 0.05), whereas 4 CpGs were negatively correlated with age at nominal significance (Table 3). Similarly, 15 CpGs were positively correlated with age at nominal significance, whereas no CpGs were negatively correlated with age at nominal significance in GDNF; 20 CpGs were positively correlated with age at nominal significance, whereas 3 CpGs were negatively correlated with age at nominal significance in NR4A2; and 2 CpGs were positively correlated with age at nominal significance, whereas no CpGs were negatively correlated with age at nominal significance in NPAS4 (Table 3). Furthermore, the top hit CpG at cg02328239 in GDNF in the GTP study (p = 1.54 × E-20) was also positively correlated with age at multiple testing significance level (Bonferroni corrected for 201 CpGs, significant level p < 0.05/201 = 2.49 × E-4) in brain tissue in the NSG study (Table 3). In fact, among the significant CpGs at multiple testing significant level (p < 2.49 × E-4) in the NSG cohort, two CpGs (cg02328239 in GDNF and cg26949694 in BDNF) overlapped with the significant CpGs at multiple testing significant levels in the GTP cohort (Tables 2, 3).
Table 3:
Correlations between age and DNAm levels of neurotrophic genes in brain samples obtained from the NSG cohort
| Gene | CpG | ρ | p-value |
|---|---|---|---|
| NPAS4 | cg08715791* | 0.85 | 1.17E-06 |
| GDNF | cg02328239*, † | 0.85 | 1.26E-06 |
| NR4A2 | cg27074041* | 0.82 | 4.87E-06 |
| NR4A2 | cg00558219* | 0.75 | 9.09E-05 |
| BDNF | cg26949694*, † | 0.75 | 9.49E-05 |
| GDNF | cg03230469* | 0.73 | 1.67E-04 |
| NR4A2 | cg02964555 | 0.70 | 3.85E-04 |
| BDNF | cg12296752 | 0.70 | 4.28E-04 |
| GDNF | cg04407962 | 0.69 | 5.65E-04 |
| NR4A2 | cg20570611 | 0.68 | 6.67E-04 |
| NPAS4 | cg04768203 | 0.67 | 9.36E-04 |
| GDNF | cg12930882 | 0.64 | 0.002 |
| BDNF | cg20108357 | 0.63 | 0.002 |
| GDNF | cg04209913 | 0.62 | 0.003 |
| NR4A2 | cg14811105 | 0.62 | 0.003 |
| BDNF | cg07238832 | 0.61 | 0.003 |
| BDNF | cg06684850 | 0.60 | 0.004 |
| NR4A2 | cg11358945 | 0.60 | 0.004 |
| GDNF | cg21286419 | 0.60 | 0.004 |
| NR4A2 | cg16058600 | 0.58 | 0.006 |
| NR4A2 | cg11209121 | 0.58 | 0.006 |
| GDNF | cg01107142 | 0.57 | 0.007 |
| GDNF | cg26559974 | 0.57 | 0.007 |
| NR4A2 | cg01123282 | −0.57 | 0.007 |
| BDNF | cg04106006 | 0.57 | 0.007 |
| NR4A2 | cg00194126 | 0.56 | 0.008 |
| NR4A2 | cg17654050 | 0.56 | 0.008 |
| GDNF | cg15368455 | 0.56 | 0.008 |
| BDNF | cg14291693 | 0.56 | 0.009 |
| GDNF | cg14590843 | 0.56 | 0.009 |
| BDNF | cg08760147 | 0.55 | 0.010 |
| GDNF | cg00761985 | 0.55 | 0.010 |
| NR4A2 | cg14617996 | 0.54 | 0.011 |
| BDNF | cg18595174 | 0.54 | 0.012 |
| BDNF | cg05189570 | 0.53 | 0.013 |
| BDNF | cg15710245 | −0.53 | 0.013 |
| NR4A2 | cg13500877 | 0.53 | 0.014 |
| BDNF | cg01583131 | 0.53 | 0.014 |
| NR4A2 | cg00240195 | 0.52 | 0.015 |
| BDNF | cg25928860 | 0.52 | 0.016 |
| BDNF | cg10558494 | −0.52 | 0.016 |
| BDNF | cg03167496 | 0.51 | 0.018 |
| GDNF | cg23097534 | 0.51 | 0.018 |
| BDNF | cg06816235‡ | 0.51 | 0.019 |
| NR4A2 | cg06101180 | 0.51 | 0.019 |
| BDNF | cg06260077 | 0.50 | 0.021 |
| NR4A2 | cg16246410 | −0.50 | 0.021 |
| BDNF | cg17413943 | 0.49 | 0.024 |
| NR4A2 | cg11932911 | 0.49 | 0.024 |
| NR4A2 | cg11399967 | 0.49 | 0.024 |
| BDNF | cg09606766 | −0.49 | 0.025 |
| NR4A2 | cg10089963 | 0.49 | 0.025 |
| NR4A2 | cg21226516 | 0.48 | 0.027 |
| GDNF | cg02331025 | 0.48 | 0.027 |
| GDNF | cg23400942 | 0.47 | 0.030 |
| GDNF | cg19717018 | 0.47 | 0.031 |
| BDNF | cg10635145 | 0.46 | 0.035 |
| NR4A2 | cg15699971 | −0.46 | 0.036 |
| BDNF | cg07159484 | −0.44 | 0.044 |
| NR4A2 | cg03339537 | 0.44 | 0.045 |
| NR4A2 | cg18941818 | 0.44 | 0.048 |
| BDNF | cg01546433 | 0.43 | 0.049 |
Notes: Blue highlights a positive correlation, and pink highlights a negative correlation. ρ = Spearman’s rank correlation coefficient.
Significant after correction for multiple-testing level (p < 0.05/201 = 2.49 × E-4),
Overlapped with the significant CpGs at multiple testing significance level (p < 2.21 × E-4) in the GTP cohort,
Overlapped with the significant CpGs at Bonferroni-corrected genome-wide significance levels (p < 5 × E-8) in the GTP cohort and the nominally significant (p < 0.05) CpGs in the EOD cohort.
Abbreviations: BDNF, Brain-derived neurotrophic factor; GDNF, glial cell–derived neurotrophic factor; NPAS4, neuronal Per-Arnt-Sim domain protein 4; NR4A2, nuclear receptor subfamily 4A2; NSG, neurosurgery
3.3. Comparison of delirium cases vs non-delirium controls in EOD study samples
We next expanded upon our previous analyses in a population of delirium patients to determine if there were differences in the degrees of correlation for neurotrophic genes between DNAm and age among delirium cases and controls. We evaluated the degree of correlation between age and DNAm levels at CpGs in BDNF (Table 4), GDNF, NR4A2, and NPAS4 (Supplementary Tables 2–4). There were 192 CpGs annotated in the four genes. There were more nominal significant (p < 0.05) CpGs in delirium cases than in non-delirium controls with BDNF (5 in delirium cases vs. 3 in non-delirium controls), GDNF (5 vs. 2), and NR4A2 (3 vs. 0). In addition, there were more positively correlated CpGs in delirium cases than in non-delirium controls with BDNF, GDNF, and NPAS4. In the EOD study, the two top CpGs among delirium patients were located in the BDNF gene (cg06816235 and cg05733135). Of note, these CpGs were overlapping with significant CpGs in the GTP and/or NSG cohort as follows: cg06816235 was significant after Bonferroni-corrected genome-wide significance levels (p < 5 × E-8) in the GTP cohort and nominally significant (p < 0.05) in the NSG cohort, and cg05733135 was significant after Bonferroni-corrected genome-wide significance levels (p < 5 × E-8) in the GTP cohort whereas it was not significant in the NSG cohort.
Table 4:
Correlation of age and blood DNAm at 83 CpGs in the BDNF gene compared between delirium cases and non-delirium controls in the EOD cohort
| Gene | Delirium Cases (N=43) | Non Delirium Controls (N=44) | ||||
|---|---|---|---|---|---|---|
| CpG | r | p-value | CpG | r | p-value | |
| cg06816235** | 0.52 | 3.68E-04 | cg20340655 | 0.38 | 0.010 | |
| cg05733135* | 0.43 | 0.004 | cg26949694 | 0.30 | 0.047 | |
| cg13974632 | 0.39 | 0.010 | cg06046431 | 0.30 | 0.048 | |
| cg25328597 | 0.37 | 0.015 | cg04106006 | 0.26 | 0.088 | |
| cg10022526 | 0.36 | 0.017 | cg06816235** | 0.26 | 0.091 | |
| cg09505801 | 0.30 | 0.052 | cg18117895 | 0.25 | 0.098 | |
| cg11718030 | 0.30 | 0.053 | cg13974632 | 0.22 | 0.143 | |
| cg01642653 | 0.29 | 0.060 | cg06684850 | 0.22 | 0.155 | |
| cg27351358 | 0.26 | 0.087 | cg15710245 | 0.21 | 0.173 | |
| cg04106006 | 0.26 | 0.092 | cg25457956 | 0.20 | 0.186 | |
| cg07159484 | 0.25 | 0.100 | cg09505801 | 0.20 | 0.193 | |
| cg01583131 | 0.25 | 0.103 | cg10022526 | 0.20 | 0.204 | |
| cg17882499 | 0.24 | 0.124 | cg23947039 | 0.19 | 0.207 | |
| cg10558494 | 0.23 | 0.134 | cg17413943 | 0.19 | 0.221 | |
| cg15710245 | 0.23 | 0.145 | cg22043168 | 0.17 | 0.258 | |
| cg23619332 | 0.22 | 0.161 | cg05218375 | 0.17 | 0.264 | |
| cg25156688 | 0.21 | 0.177 | cg25928860 | 0.16 | 0.304 | |
| cg06991510 | 0.21 | 0.183 | cg01225698 | 0.15 | 0.334 | |
| BDNF | cg08362738 | 0.20 | 0.193 | cg21010859 | 0.15 | 0.345 |
| cg17413943 | 0.19 | 0.215 | cg08362738 | 0.13 | 0.414 | |
| cg09492354 | 0.19 | 0.217 | cg05733135* | 0.12 | 0.438 | |
| cg25457956 | 0.18 | 0.244 | cg06979684 | 0.12 | 0.454 | |
| cg04481212 | 0.17 | 0.280 | cg01583131 | 0.11 | 0.480 | |
| cg15313332 | 0.17 | 0.287 | cg05189570 | 0.10 | 0.503 | |
| cg02613510 | 0.16 | 0.298 | cg11806762 | 0.09 | 0.570 | |
| cg22128379 | 0.16 | 0.307 | cg07159484 | 0.09 | 0.575 | |
| cg11865360 | 0.16 | 0.315 | cg24650785 | 0.07 | 0.645 | |
| cg03984780 | 0.15 | 0.331 | cg20954537 | 0.07 | 0.648 | |
| cg27193031 | 0.15 | 0.337 | cg01546433 | 0.07 | 0.652 | |
| cg10635145 | 0.15 | 0.341 | cg15914769 | 0.07 | 0.655 | |
| cg26949694 | 0.15 | 0.342 | cg08760147 | 0.06 | 0.676 | |
| cg06979684 | 0.15 | 0.349 | cg12021170 | 0.06 | 0.682 | |
| cg11806762 | 0.14 | 0.361 | cg22288103 | 0.06 | 0.682 | |
| cg06025631 | 0.14 | 0.369 | cg03984780 | 0.06 | 0.719 | |
| cg01225698 | 0.14 | 0.387 | cg06025631 | 0.05 | 0.731 | |
| cg07238832 | 0.14 | 0.388 | cg27351358 | 0.05 | 0.734 | |
| cg24650785 | 0.12 | 0.458 | cg01642653 | 0.05 | 0.737 | |
| cg22043168 | 0.12 | 0.458 | cg03167496 | 0.05 | 0.746 | |
| cg16257091 | 0.12 | 0.459 | cg25156688 | 0.04 | 0.798 | |
| cg15462887 | 0.12 | 0.462 | cg11718030 | 0.04 | 0.819 | |
| cg06684850 | 0.11 | 0.468 | cg07704699 | 0.03 | 0.856 | |
| cg23947039 | 0.11 | 0.471 | cg20108357 | 0.03 | 0.860 | |
| cg05218375 | 0.11 | 0.488 | cg24249411 | 0.03 | 0.868 | |
| cg07704699 | 0.11 | 0.499 | cg01636003 | 0.02 | 0.891 | |
| cg06046431 | 0.10 | 0.510 | cg15462887 | 0.02 | 0.923 | |
| cg24065044 | 0.09 | 0.548 | cg05847680 | 0.01 | 0.924 | |
| cg21010859 | 0.09 | 0.550 | cg03747251 | 0.00 | 0.999 | |
| cg09606766 | 0.08 | 0.599 | cg00298481 | 0.00 | 0.990 | |
| cg11241206 | 0.08 | 0.615 | cg11865360 | −0.01 | 0.958 | |
| cg12067298 | 0.08 | 0.623 | cg10558494 | −0.01 | 0.951 | |
| cg20954537 | 0.07 | 0.637 | cg02386994 | −0.02 | 0.917 | |
| cg04672351 | 0.06 | 0.684 | cg12067298 | −0.02 | 0.908 | |
| cg03747251 | 0.06 | 0.692 | cg02527472 | −0.03 | 0.867 | |
| cg03167496 | 0.05 | 0.731 | cg04481212 | −0.03 | 0.867 | |
| cg12448003 | 0.05 | 0.749 | cg11241206 | −0.03 | 0.850 | |
| cg05818894 | 0.04 | 0.801 | cg22128379 | −0.03 | 0.832 | |
| cg22288103 | 0.03 | 0.860 | cg14291693 | −0.03 | 0.828 | |
| cg08388004 | 0.02 | 0.906 | cg17882499 | −0.04 | 0.808 | |
| cg08760147 | 0.01 | 0.938 | cg15313332 | −0.05 | 0.771 | |
| cg20340655 | −0.01 | 0.940 | cg26057780 | −0.05 | 0.739 | |
| cg02527472 | −0.02 | 0.902 | cg26840770 | −0.05 | 0.732 | |
| cg24249411 | −0.02 | 0.896 | cg25328597 | −0.06 | 0.709 | |
| cg02386994 | −0.03 | 0.844 | cg06991510 | −0.06 | 0.676 | |
| cg18117895 | −0.04 | 0.817 | cg24065044 | −0.07 | 0.637 | |
| cg05847680 | −0.04 | 0.807 | cg18354203 | −0.07 | 0.633 | |
| cg26057780 | −0.04 | 0.777 | cg04672351 | −0.08 | 0.621 | |
| cg12021170 | −0.06 | 0.720 | cg09606766 | −0.08 | 0.595 | |
| cg15914769 | −0.07 | 0.673 | cg07238832 | −0.10 | 0.529 | |
| cg18354203 | −0.08 | 0.627 | cg06260077 | −0.10 | 0.514 | |
| cg15014679 | −0.08 | 0.597 | cg05818894 | −0.11 | 0.494 | |
| cg26840770 | −0.12 | 0.429 | cg12448003 | −0.11 | 0.481 | |
| cg00298481 | −0.13 | 0.412 | cg10635145 | −0.13 | 0.409 | |
| cg01636003 | −0.14 | 0.385 | cg16257091 | −0.14 | 0.374 | |
| cg05189570 | −0.15 | 0.336 | cg09492354 | −0.15 | 0.331 | |
| cg06260077 | −0.15 | 0.336 | cg23143371 | −0.15 | 0.317 | |
| cg23143371 | −0.17 | 0.275 | cg23619332 | −0.20 | 0.200 | |
| cg23426002 | −0.18 | 0.243 | cg18595174 | −0.20 | 0.189 | |
| cg20108357 | −0.21 | 0.178 | cg23426002 | −0.21 | 0.181 | |
| cg18595174 | −0.22 | 0.162 | cg27193031 | −0.23 | 0.128 | |
| cg14291693 | −0.23 | 0.137 | cg02613510 | −0.26 | 0.084 | |
| cg01546433 | −0.25 | 0.105 | cg15688670 | −0.30 | 0.046 | |
| cg15688670 | −0.26 | 0.092 | cg08388004 | −0.31 | 0.041 | |
| cg25928860 | −0.30 | 0.047 | cg15014679 | −0.34 | 0.023 | |
Notes: Blue highlights a positive correlation, and pink highlights a negative correlation. r = Pearson’s correlation coefficient.
Overlapped with the significant CpGs at Bonferroni-corrected genome-wide significance levels (p < 5 × E-8) in the GTP cohort and nominally significant (p < 0.05) in the NSG cohort,
Overlapped with the significant CpGs at Bonferroni-corrected genome-wide significance levels (p < 5 × E-8) in the GTP cohort.
Abbreviations: BDNF, Brain-derived neurotrophic factor; DNAm, DNA methylation; EOD, epigenetics of delirium.
4. Discussion
In the present study, 16 CpGs among the genes investigated, including 7 CpGs in BDNF and 4 in GDNF, were significantly correlated with age in the large GTP cohort with blood samples, and 4 of these CpGs were significant in the NSG cohort of brain samples. Furthermore, there were more nominal significant CpGs in neurotrophic genes positively correlated with age in delirium cases than controls in the EOD cohort. These results support our hypothesis that DNAm in neurotrophic genes in both brain and blood increases with age, and such changes are more significant among delirium patients.
Data from GTP cohort blood samples showed that the majority (96 among 109 CpGs; 88.1%) of the nominally significant CpGs were positively correlated with age. With the NSG cohort brain samples, we also found that the majority (55 among 62 CpGs; 88.7%) of the nominally significant CpGs were positively correlated with age. Furthermore, the top hit CpG in the GTP study was also significantly positively correlated in the NSG study with brain samples. In total, 6 among the 16 significant CpGs after Bonferroni-corrected genome-wide significance levels in the GTP cohort were also nominally significant in the NSG cohort. These results indicated that some neurotrophic CpGs show similar associations between age and DNAm levels in brain and blood tissues.
In the EOD cohort, more CpGs in neurotrophic genes were positively correlated with age in delirium cases than in controls. This is consistent with our hypothesis that neurotrophic genes show greater DNAm changes with age among elderly patients, indicating these genes may be dysregulated with altered expression levels compared to controls. In general, DNAm is known to change with increasing age (Jones et al., 2015). In addition, because some specific sites of DNAm are highly correlated with aging, they are used for predicting chronological age (Hannum et al., 2013; Horvath, 2013; Jones et al., 2015). However, what is lacking in the literature is a better understanding of the biological implications of such changes over the lifespan. To address one aspect of biological factors through the aging process, we aimed to conduct this study investigating the association between DNAm among neurotrophic genes and age, as these genes have been well studied in their association with neurocognitive function domains.
In the present report, DNAm in neurotrophic genes were positively correlated with age. This is consistent with the result of a previous study that showed a significant positive correlation between DNAm of the BDNF gene and age among healthy female subjects (Ihara et al., 2018). These results suggest that DNAm in neurotrophic genes may increase with aging, and they provide supportive evidence that potentially DNAm is playing a role in the biological mechanisms of the aging process.
Regarding the role of neurotrophic factor related to delirium, we showed that DNAm of BDNF was more positively correlated with age in delirium cases than those in controls. Therefore, we speculated that BDNF levels in blood would be decreased in delirium cases; however, studies of the relationship between circulating BDNF levels and delirium reveal mixed results. Higher plasma BDNF levels were associated with emergence of agitation in elderly patients after gastrointestinal surgery (Mei and Tong, 2016). A study of ICU patients showed that serum level of BDNF was significantly higher in delirium patients than in non-delirious controls (Grandi et al., 2011). On the other hand, in patients undergoing spine surgery, an intraoperative decline in plasma BDNF was greater in those who developed delirium (Wyrobek et al., 2017). In addition, no significant differences between delirium cases and controls were found in a study of oncology inpatients (Brum et al., 2015) in a study of acute medical inpatients (Williams et al., 2017). Therefore, BDNF level itself does not seem to be a reliable marker for delirium, and further mechanistic investigations including an epigenetics approach is warranted as presented here.
There are several limitations to this study. First, the NSG cohort was limited in sample size due to the infrequent occurrence of neurosurgical cases. Second, both blood and brain tissues consist of heterogeneous cellular populations. In the present study, we did not measure cellular populations in both blood and brain tissues. DNAm is known to be different according to the blood cellular types (Houseman et al., 2012). In addition, associations between DNAm level and age are different among brain cell types (Shinozaki et al., 2018a). Therefore, potential heterogeneity of blood and brain cellular populations might have been a confounding factor in the present study. Third, a peripheral tissue, blood, had to be used to assess DNAm levels in the EOD cohort of delirium patients for obvious reasons, despite the brain being the main organ of interest for delirium. Our recent work investigating the level of correlation between DNAm of brain and peripheral tissues can be useful to shed light in this regard (Braun et al., 2019). Fourth, although our main interest was to investigate the role of aging on DNAm change, the dataset presented here was not able to address such question due to the cross-sectional nature of the study design. To obtain such information, future longitudinal study design is required where bio-samples can be collected over two different time points from the same individuals. Fifth, because gene expressions were not investigated in this study, we were not able to show associations between DNAm and gene expressions. Such additional investigations are of significant importance for future research. Finally, dementia status might affect the result between DNAm and age as a confounding factor in the EOD cohort. Although we did not have detailed information about the dementia status of our study population, future study should address the influence of dementia.
Based on a thorough literature search, this is the first study to analyze the degree of correlation between DNAm in neurotrophic genes and age in delirium patients. We found that DNAm of the neurotrophic genes were positively associated with age in both blood and brain samples. In addition, these DNAm were more positively correlated with age in delirious patients than controls. These findings provide initial evidence that neurotrophic genes may be epigenetically modulated with age, and in patients with delirium, this process may be dysregulated.
Supplementary Material
Highlights.
Neurotrophic genes’ DNAm were positively correlated with age in blood and brain
Neurotrophic genes’ DNAm were positively correlated with age in delirious patients
- Neurotrophic genes may be modulated with aging epigenetically
- Dr. Shinozaki G is co-founder of Predelix Medical LLC, and reports U.S. Provisional Patent Application No. 62/731599, titled “Epigenetic biomarker of delirium risk.” The other authors report no financial interests or potential conflicts of interest.
- This work was supported by research grants from National Science Foundation1664364; and the National Institute of Mental Health (K23 MH107654).
- This paper has not been previously published nor submitted for publication elsewhere except the preprint version for bioRχiv (doi:https://doi.org/10.1101/730382).
- This study was approved by the University of Iowa’s Human Subjects Research Institutional Review Board.
- All authors directly participated in this study and have read and approved the final version.
Acknowledgments
Funding
This work was supported by research grants from the National Science Foundation (1664364) and the National Institute of Mental Health (K23 MH107654).
Dr. Shinozaki is co-founder of Predelix Medical LLC, and reports U.S. Provisional Patent Application No. 62/731599, titled “Epigenetic biomarker of delirium risk.”
Abbreviations
- AD
Alzheimer’s disease
- ARC
activity regulated cytoskeleton associated protein
- BDNF
brain-derived neurotrophic factor
- CAM
Confusion Assessment Method
- CAM-ICU
Confusion Assessment Method for Intensive Care Unit
- DNAm
DNA methylation
- DOSS
Delirium Observation Screening Scale
- DRS-R-98
Delirium Rating Scale—Revised-98
- EDTA
ethylenediaminetetraacetic acid
- EOD
epigenetics of delirium
- FOS
Fos Proto-Oncogene
- AP-1
transcription factor subunit
- GDNF
glial cell line–derived neurotrophic factor
- GTP
Grady Trauma Project
- NPAS4
Neuronal Per-Arnt-Sim domain protein 4
- NR4A1
nuclear receptor subfamily 4A1
- NR4A2
nuclear receptor subfamily 4A2
- NSG
neurosurgery
- SNP
single nucleotide polymorphism
Footnotes
Disclosures
The other authors report no financial interests or potential conflicts of interest.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- Aryee MJ, Jaffe AE, Corrada-Bravo H, Ladd-Acosta C, Feinberg AP, Hansen KD, Irizarry RA, 2014. Minfi: a flexible and comprehensive Bioconductor package for the analysis of Infinium DNA methylation microarrays. Bioinformatics 30, 1363–1369. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Assenov Y, Muller F, Lutsik P, Walter J, Lengauer T, Bock C, 2014. Comprehensive analysis of DNA methylation data with RnBeads. Nat Methods 11, 1138–1140. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Berchtold NC, Cribbs DH, Coleman PD, Rogers J, Head E, Kim R, Beach T, Miller C, Troncoso J, Trojanowski JQ, Zielke HR, Cotman CW, 2008. Gene expression changes in the course of normal brain aging are sexually dimorphic. Proc. Natl. Acad. Sci. U. S. A 105, 15605–15610. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Beveridge NJ, Santarelli DM, Wang X, Tooney PA, Webster MJ, Weickert CS, Cairns MJ, 2014. Maturation of the human dorsolateral prefrontal cortex coincides with a dynamic shift in microRNA expression. Schizophr. Bull 40, 399–409. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Braun PR, Han S, Hing B, Nagahama Y, Gaul LN, Heinzman JT, Grossbach AJ, Close L, Dlouhy BJ, Howard MA, 2019. Genome-wide DNA methylation comparison between live human brain and peripheral tissues within individuals. Translational psychiatry 9, 47. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Brum C, Stertz L, Borba E, Rumi D, Kapczinski F, Camozzato A, 2015. Association of serum brain-derived neurotrophic factor (BDNF) and tumor necrosis factor-alpha (TNF-alpha) with diagnosis of delirium in oncology inpatients. Braz J Psychiatry 37, 197–202. [DOI] [PubMed] [Google Scholar]
- Choy FC, Klarić TS, Leong WK, Koblar SA, Lewis MD, 2016. Reduction of the neuroprotective transcription factor Npas4 results in increased neuronal necrosis, inflammation and brain lesion size following ischaemia. J. Cereb. Blood Flow Metab 36, 1449–1463. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Christensen BC, Houseman EA, Marsit CJ, Zheng S, Wrensch MR, Wiemels JL, Nelson HH, Karagas MR, Padbury JF, Bueno R, Sugarbaker DJ, Yeh RF, Wiencke JK, Kelsey KT, 2009. Aging and environmental exposures alter tissue-specific DNA methylation dependent upon CpG island context. PLoS Genet 5, e1000602. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Danka Mohammed CP, Park JS, Nam HG, Kim K, 2017. MicroRNAs in brain aging. Mech. Ageing Dev 168, 3–9. [DOI] [PubMed] [Google Scholar]
- Day K, Waite LL, Thalacker-Mercer A, West A, Bamman MM, Brooks JD, Myers RM, Absher D, 2013. Differential DNA methylation with age displays both common and dynamic features across human tissues that are influenced by CpG landscape. Genome Biol. 14, R102. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ely EW, Inouye SK, Bernard GR, Gordon S, Francis J, May L, Truman B, Speroff T, Gautam S, Margolin R, Hart RP, Dittus R, 2001a. Delirium in mechanically ventilated patients: validity and reliability of the confusion assessment method for the intensive care unit (CAM-ICU). JAMA 286, 2703–2710. [DOI] [PubMed] [Google Scholar]
- Ely EW, Margolin R, Francis J, May L, Truman B, Dittus R, Speroff T, Gautam S, Bernard GR, Inouye SK, 2001b. Evaluation of delirium in critically ill patients: validation of the Confusion Assessment Method for the Intensive Care Unit (CAM-ICU). Crit. Care Med 29, 1370–1379. [DOI] [PubMed] [Google Scholar]
- Forlenza OV, Miranda AS, Guimar I, Talib LL, Diniz BS, Gattaz WF, Teixeira AL, 2015. Decreased neurotrophic support is associated with cognitive decline in non-demented subjects. J. Alzheimers Dis 46, 423–429. [DOI] [PubMed] [Google Scholar]
- Fortin J-P, Triche TJ Jr, Hansen KD, 2017. Preprocessing, normalization and integration of the Illumina HumanMethylationEPIC array with minfi. Bioinformatics 33, 558–560. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ghitza UE, Zhai H, Wu P, Airavaara M, Shaham Y, Lu L, 2010. Role of BDNF and GDNF in drug reward and relapse: a review. Neurosci. Biobehav. Rev 35, 157–171. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Gong H, Qian H, Ertl R, Astle CM, Wang GG, Harrison DE, Xu X, 2015. Histone modifications change with age, dietary restriction and rapamycin treatment in mouse brain. Oncotarget 6, 15882–15890. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Grandi C, Tomasi CD, Fernandes K, Stertz L, Kapczinski F, Quevedo J, Dal-Pizzol F, Ritter C, 2011. Brain-derived neurotrophic factor and neuron-specific enolase, but not S100beta, levels are associated to the occurrence of delirium in intensive care unit patients. J. Crit. Care 26, 133–137. [DOI] [PubMed] [Google Scholar]
- Gunstad J, Benitez A, Smith J, Glickman E, Spitznagel MB, Alexander T, Juvancic-Heltzel J, Murray L, 2008. Serum brain-derived neurotrophic factor is associated with cognitive function in healthy older adults. J. Geriatr. Psychiatry Neurol 21, 166–170. [DOI] [PubMed] [Google Scholar]
- Hannum G, Guinney J, Zhao L, Zhang L, Hughes G, Sadda S, Klotzle B, Bibikova M, Fan JB, Gao Y, Deconde R, Chen M, Rajapakse I, Friend S, Ideker T, Zhang K, 2013. Genome-wide methylation profiles reveal quantitative views of human aging rates. Mol. Cell 49, 359–367. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Horvath S, 2013. DNA methylation age of human tissues and cell types. Genome Biol. 14, R115. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Houseman EA, Accomando WP, Koestler DC, Christensen BC, Marsit CJ, Nelson HH, Wiencke JK, Kelsey KT, 2012. DNA methylation arrays as surrogate measures of cell mixture distribution. BMC Bioinformatics 13, 86. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ihara K, Fuchikami M, Hashizume M, Okada S, Kawai H, Obuchi S, Hirano H, Fujiwara Y, Hachisu M, Hongyong K, Morinobu S, 2018. The influence of aging on the methylation status of brain-derived neurotrophic factor gene in blood. Int. J. Geriatr. Psychiatry 33, 1312–1318. [DOI] [PubMed] [Google Scholar]
- Inouye SK, 2006. Delirium in older persons. N. Engl. J. Med 354, 1157–1165. [DOI] [PubMed] [Google Scholar]
- Inouye SK, van Dyck CH, Alessi CA, Balkin S, Siegal AP, Horwitz RI, 1990. Clarifying confusion: the confusion assessment method. A new method for detection of delirium. Ann. Intern. Med 113, 941–948. [DOI] [PubMed] [Google Scholar]
- Johnson WE, Li C, Rabinovic A, 2007. Adjusting batch effects in microarray expression data using empirical Bayes methods. Biostatistics 8, 118–127. [DOI] [PubMed] [Google Scholar]
- Jones MJ, Goodman SJ, Kobor MS, 2015. DNA methylation and healthy human aging. Aging Cell 14, 924–932. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kang HJ, Kawasawa YI, Cheng F, Zhu Y, Xu X, Li M, Sousa AM, Pletikos M, Meyer KA, Sedmak G, Guennel T, Shin Y, Johnson MB, Krsnik Z, Mayer S, Fertuzinhos S, Umlauf S, Lisgo SN, Vortmeyer A, Weinberger DR, Mane S, Hyde TM, Huttner A, Reimers M, Kleinman JE, Sestan N, 2011. Spatio-temporal transcriptome of the human brain. Nature 478, 483–489. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Leek JT, Johnson WE, Parker HS, Fertig EJ, Jaffe AE, Storey JD, Zhang Y, Torres LC, 2019. sva: Surrogate variable analysis. R package version. [Google Scholar]
- Li G, Peskind ER, Millard SP, Chi P, Sokal I, Yu C-E, Bekris LM, Raskind MA, Galasko DR, Montine TJ, 2009. Cerebrospinal fluid concentration of brain-derived neurotrophic factor and cognitive function in non-demented subjects. PLoS One 4, e5424. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Mei X, Tong J, 2016. The plasma levels of brain-derived neurotrophic factor are positively associated with emergence agitation in the elderly after gastrointestinal surgery. J. Anesth 30, 811–816. [DOI] [PubMed] [Google Scholar]
- Moon M, Jung ES, Jeon SG, Cha MY, Jang Y, Kim W, Lopes C, Mook‐Jung I, Kim KS, 2019. Nurr1 (NR4A2) regulates Alzheimer’s disease‐related pathogenesis and cognitive function in the 5XFAD mouse model. Aging cell 18, e12866. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Morris TJ, Butcher LM, Feber A, Teschendorff AE, Chakravarthy AR, Wojdacz TK, Beck S, 2014. ChAMP: 450k Chip Analysis Methylation Pipeline. Bioinformatics 30, 428–430. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Nishimura K, Yokoyama K, Yamauchi N, Koizumi M, Harasawa N, Yasuda T, Mimura C, Igita H, Suzuki E, Uchiide Y, Seino Y, Nomura M, Yamazaki K, Ishigooka J, 2016. Sensitivity and specificity of the Confusion Assessment Method for the Intensive Care Unit (CAM-ICU) and the Intensive Care Delirium Screening Checklist (ICDSC) for detecting post-cardiac surgery delirium: A single-center study in Japan. Heart Lung 45, 15–20. [DOI] [PubMed] [Google Scholar]
- Nordlund J, Bäcklin CL, Wahlberg P, Busche S, Berglund EC, Eloranta M-L, Flaegstad T, Forestier E, Frost B-M, Harila-Saari A, 2013. Genome-wide signatures of differential DNA methylation in pediatric acute lymphoblastic leukemia. Genome Biol. 14, r105. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Numata S, Ye T, Hyde TM, Guitart-Navarro X, Tao R, Wininger M, Colantuoni C, Weinberger DR, Kleinman JE, Lipska BK, 2012. DNA methylation signatures in development and aging of the human prefrontal cortex. Am. J. Hum. Genet 90, 260–272. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Pal S, Tyler JK, 2016. Epigenetics and aging. Sci Adv 2, e1600584. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Pidsley R, Zotenko E, Peters TJ, Lawrence MG, Risbridger GP, Molloy P, Van Djik S, Muhlhausler B, Stirzaker C, Clark SJ, 2016. Critical evaluation of the Illumina MethylationEPIC BeadChip microarray for whole-genome DNA methylation profiling. Genome Biol. 17, 208. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Team R Core, 2019. R: A language and environment for statistical computing. https://www.R-project.org/ (Accessed February 13, 2019).
- Ratanatharathorn A, Boks MP, Maihofer AX, Aiello AE, Amstadter AB, Ashley‐Koch AE, Baker DG, Beckham JC, Bromet E, Dennis M, 2017. Epigenome‐wide association of PTSD from heterogeneous cohorts with a common multi‐site analysis pipeline. American Journal of Medical Genetics Part B: Neuropsychiatric Genetics 174, 619–630. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Schuurmans MJ, Shortridge-Baggett LM, Duursma SA, 2003. The Delirium Observation Screening Scale: a screening instrument for delirium. Res. Theory Nurs. Pract 17, 31–50. [DOI] [PubMed] [Google Scholar]
- Sen P, Shah PP, Nativio R, Berger SL, 2016. Epigenetic Mechanisms of Longevity and Aging. Cell 166, 822–839. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Shimada H, Makizako H, Yoshida D, Tsutsumimoto K, Anan Y, Uemura K, Lee S, Park H, Suzuki T, 2014. A large, cross-sectional observational study of serum BDNF, cognitive function, and mild cognitive impairment in the elderly. Front. Aging Neurosci 6, 69. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Shinozaki G, Bormann NL, Chan AC, Zarei K, Sparr NA, Klisares MJ, Jellison SS, Heinzman JT, Dahlstrom EB, Duncan GN, Gaul LN, Wanzek RJ, Cramer EM, Wimmel CG, Sabbagh S, Yuki K, Weckmann MT, Yamada T, Karam MD, Noiseux NO, Shinozaki E, Cho HR, Lee S, Cromwell JW, 2019. Identification of Patients With High Mortality Risk and Prediction of Outcomes in Delirium by Bispectral EEG. J. Clin. Psychiatry 80. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Shinozaki G, Braun PR, Hing BWQ, Ratanatharathorn A, Klisares MJ, Duncan GN, Jellison SS, Heinzman JT, Nagahama Y, Close L, Sabbagh S, Dlouhy BJ, Howard MA, Kawasaki H, Cho HR, 2018a. Epigenetics of Delirium and Aging: Potential Role of DNA Methylation Change on Cytokine Genes in Glia and Blood Along With Aging. Front. Aging Neurosci 10, 311. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Shinozaki G, Chan AC, Sparr NA, Zarei K, Gaul LN, Heinzman JT, Robles J, Yuki K, Chronis TJ, Ando T, Wong T, Sabbagh S, Weckmann MT, Lee S, Yamada T, Karam MD, Noiseux NO, Shinozaki E, Cromwell JW, 2018b. Delirium detection by a novel bispectral electroencephalography device in general hospital. Psychiatry Clin. Neurosci 72, 856–863. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Smith AK, Kilaru V, Klengel T, Mercer KB, Bradley B, Conneely KN, Ressler KJ, Binder EB, 2015. DNA extracted from saliva for methylation studies of psychiatric traits: evidence tissue specificity and relatedness to brain. Am. J. Med. Genet. B Neuropsychiatr. Genet 168b, 36–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Spronk PE, Riekerk B, Hofhuis J, Rommes JH, 2009. Occurrence of delirium is severely underestimated in the ICU during daily care. Intensive Care Med. 35, 1276–1280. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Teschendorff AE, Marabita F, Lechner M, Bartlett T, Tegner J, Gomez-Cabrero D, Beck S, 2013. A beta-mixture quantile normalization method for correcting probe design bias in Illumina Infinium 450 k DNA methylation data. Bioinformatics 29, 189–196. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Trzepacz PT, Mittal D, Torres R, Kanary K, Norton J, Jimerson N, 2001. Validation of the Delirium Rating Scale-revised-98: comparison with the delirium rating scale and the cognitive test for delirium. J. Neuropsychiatry Clin. Neurosci 13, 229–242. [DOI] [PubMed] [Google Scholar]
- Tsai PC, Spector TD, Bell JT, 2012. Using epigenome-wide association scans of DNA methylation in age-related complex human traits. Epigenomics 4, 511–526. [DOI] [PubMed] [Google Scholar]
- van Eijk MM, van den Boogaard M, van Marum RJ, Benner P, Eikelenboom P, Honing ML, van der Hoven B, Horn J, Izaks GJ, Kalf A, Karakus A, Klijn IA, Kuiper MA, de Leeuw FE, de Man T, van der Mast RC, Osse RJ, de Rooij SE, Spronk PE, van der Voort PH, van Gool WA, Slooter AJ, 2011. Routine use of the confusion assessment method for the intensive care unit: a multicenter study. Am. J. Respir. Crit. Care Med 184, 340–344. [DOI] [PubMed] [Google Scholar]
- Williams J, Finn K, Melvin V, Meagher D, McCarthy G, Adamis D, 2017. The Association of Serum Levels of Brain-Derived Neurotrophic Factor with the Occurrence of and Recovery from Delirium in Older Medical Inpatients. Biomed Res Int 2017, 5271395. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wyrobek J, LaFlam A, Max L, Tian J, Neufeld KJ, Kebaish KM, Walston JD, Hogue CW, Riley LH, Everett AD, Brown C.H.t., 2017. Association of intraoperative changes in brain-derived neurotrophic factor and postoperative delirium in older adults. Br. J. Anaesth 119, 324–332. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zampieri M, Ciccarone F, Calabrese R, Franceschi C, Burkle A, Caiafa P, 2015. Reconfiguration of DNA methylation in aging. Mech. Ageing Dev 151, 60–70. [DOI] [PubMed] [Google Scholar]
- Zhou W, Laird PW, Shen H, 2017. Comprehensive characterization, annotation and innovative use of Infinium DNA methylation BeadChip probes. Nucleic Acids Res. 45, e22–e22. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.