This article is mentioned in:

Introduction

Alzheimer's disease (AD) is a chronic neurodegenerative disease characterized by a progressive decline in cognitive and memory function (1). The prevalence of AD was 35.6 million globally in 2010, and this number is expected to double by 2030. AD has become one of the most common forms of dementia in the elderly worldwide (2), exerting a huge burden on families and society.

AD is a complex disease that is influenced by environmental and genetic factors (3). Heritability studies have shown that ~70% of AD risk may be attributed to genetic factors (4). Epigenetic modifications are thought to link environmental and genetic factors (5–7). DNA methylation, a type of epigenetic modification, has been shown to have a significant role in the etiology of several diseases, including leukemia (8), type 2 diabetes (9,10), essential hypertension (9), coronary heart disease (11,12), schizophrenia (13) and AD (14,15). DNA methylation often occurs in a CpG dinucleotide context (16). The CpG islands of a promoter are CpG-rich regions that are predominantly hypomethylated (17). Alterations in promoter methylation often affect gene expression (7,18–20).

Dopamine receptor D4 (DRD4) encodes the D4 subtype of the dopamine receptor (21). An increasing amount of evidence supports a link between DRD4 and AD (22). DRD4 polymorphism has been observed to be significantly associated with AD (22) and DRD4 promoter methylation (23). DRD4 gene hypermethylation has been demonstrated to increase DRD4 gene expression and the risk of schizophrenia in males (13). A hypermethylated DRD4 promoter has also been identified in patients with alcohol addiction (24).

Although DNA methylation levels vary among tissues, independent studies have revealed that the DNA methylation patterns of multiple loci in peripheral blood were similar to those in brain tissues (25–28). In light of these previous findings, the goal of the present study was to evaluate the contribution of DRD4 promoter methylation to AD using peripheral blood as a surrogate of brain tissue.

Materials and methods

Sample collection

A total of 46 sporadic AD patients and 61 matched controls were selected from Ningbo No. 1 Hospital (Ningbo, China) and Ningbo Kangning Hospital (Ningbo, China). AD was diagnosed by two professional neurological physicians (CZ and ZQ) according to National Institute of Neurological and Communicative Disorders and Stroke-Alzheimer's Disease and Related Disorders Association criteria from ICD-10 (29), on the basis of medical and family histories, neurological examination, blood studies, brain imaging studies, neuropsychological testing and cognitive screening tests. All controls were free of any type of physical or mental disorder. In the current study, two kinds of drugs (Exelon and Aricept) were used for AD patients. Exelon (Novartis Farmaceutica S.A, Spain) was administered at 1.5 mg twice per day with morning and evening meals. Following a minimum of two weeks of treatment, if the initial dose was well-tolerated, it was increased to 3 mg twice per day. Initial treatment with Aricept (Eisai China Inc., Jiangsu, China) was 5 mg per day at bedtime. The dose was increased to 10 mg per day after 4 to 6 weeks if the response was not adequate. All individuals in the present study were Han Chinese originating from Ningbo city in Eastern China. Peripheral blood samples were collected in 3.2% sodium citrate-treated tubes (Jiangsu Kailijian Medical Device Co., Ltd., Jiangsu, China) and then stored at −80°C. The study protocol was approved by the Ethical Committee of Ningbo University (Ningbo, China), Ningbo No.1 Hospital and Ningbo Kangning Hospital. Written informed consent was obtained from all subjects or their guardians.

Detection of biochemical factors

The serum concentrations of total protein (TP) and albumin (ALB) were detected using the biuret (30) and bromocresol green methods (31), respectively. Globulin (GLB) was calculated as TP minus ALB. The concentrations of glutamic-pyruvic transaminase, alkaline phosphatase and glutamic oxalacetic transaminase were determined using the velocity method (32,33). The levels of total bile acid and homocysteine (Hcy) were measured using the cycling enzymatic method (34,35). Plasma concentrations of blood glucose, triglyceride, total cholesterol, carbamide (UREA), creatinine (CRE) and uric acid were determined using the enzymatic methods (36–41). The high-density lipoprotein cholesterol level was determined using the one-step detection method (42). The proportions of apolipoprotein-A (ApoA) and apolipoprotein-B were measured by the turbidimetry method (43,44). The concentrations of lipoprotein A (Lp (a)) and C Reactive Protein (CRP) were detected using the endpoint method (45) and latex agglutination assay (46), respectively. The apolipoprotein E levels were detected using the immunoturbidimetric assay (47).

Bisulfite pyrosequencing assay

DNA extraction and consequent bisulfite pyrosequencing assays were performed as described in our previous studies (8,9,12,13,48,49). PCR primers were designed using PyroMark Assay Design software version 2 (Qiagen China Co., Ltd., Shanghai, China). Primer sequences were 5′-biotin-GGGAGGTTTTGTTAGATATTAGGT-3′ for the forward primer; 5′-CCACCCTAAACCCAATATTTACTCATCTTA-3′ for the reverse primer; and 5′-ACCAAACCAAACCCT-3′ for the sequencing primer.

Statistical analyses

Statistical analyses were performed using SPSS software version 16.0 (SPSS, Inc., Chicago, IL, USA), and a P<0.05 was considered to indicate a statistically significant difference. The two independent samples t test was used to compare the differences in the mean values of continuous variables between AD patients and controls. Pearson's correlation test was used to assess the associations between DRD4 methylation and the metabolic characteristics of the AD subjects. Bonferroni correction was used to adjust the results. Power analysis was estimated with the Power and Sample Size Calculation software version 3.043 (http://biostat.mc.vanderbilt.edu/wiki/Main/PowerSampleSize). In the present study, according to the online power calculator, α was the type I error probability for a two-sided test; n was the sample number of AD patients; δ was the difference in population means, which was equal to the mean methylation levels in AD patients minus those in normal controls; σ was the within-group standard deviation; and m was the ratio of control to experimental patients.

Results

The selected promoter fragment in the current methylation assay

The bisulfite pyrosequencing assay was performed on the CpG island region (chr11:636877-637167) of the DRD4 promoter. As shown in Fig. 1, a total of four CpG sites were measured. As there was a significant correlation among the methylation levels of the four CpG sites (Fig. 1; r=0.442, P<0.001), the mean DNA methylation of the four CpGs was also evaluated in the subsequent analyses.

Figure 1 Significant correlation among the four CpGs of the DRD4 promoter. DRD4 (chr11: 637304-640705) encodes dopamine receptor D4. CGI (chr11: 636907-640628) represents CpG island, which is enriched with CpG sites; Four CpG sites from the CpG island of the upstream DRD4 gene were selected and tested by pyrosequencing. Significant associations were observed among these sites (P<0.001). A total of three primers were used in the present study. F, forward primer; R, reverse primer; S, sequencing primer.

Association tests between clinical phenotypes with AD

The present study involved a total of 46 AD patients and 61 controls. As shown in Table I, among the 22 phenotypes, the plasma levels of ApoA, Lp (a), Hcy and CRP were observed to significantly differ between AD patients and controls (Table I; ApoA, P=0.011; Lp (a), P<0.001; Hcy, P=0.046; CRP, P=0.016). A significant male-specific association was identified between the mean DRD4 methylation and ApoA and level (Fig. 2; ApoA, P=0.042). A significant female-specific association was observed between the average DRD4 methylation and several phenotypes, including ApoA and CRE levels (Fig. 2; ApoA, P<0.001; CRE, P=0.045). Age is a well-known risk factor for AD, therefore the association between age and DRD4 methylation was tested. The results revealed an association between age and DRD4 methylation (Fig. 2; males: r=0.281, P=0.021; females: r=0.222, P=0.169).

Figure 2 Correlation tests between DRD4 methylation and three phenotypes. Pearson correlation analyses were used between DRD4 methylation levels and phenotypes in two genders. Positive associations between DRD4 methylation with age and APOA were observed in males (Age, P=0.021; ApoA, P=0.042). Positive associations between the average DRD4 methylation with ApoA and CRE levels were observed in females (ApoA, P<0.001; CRE, P=0.045). DRD4, dopamine receptor D4; APOA, apolipoprotein A; CRE, creatinine.

Table I Characteristics of the 107 subjects.

Table I

Characteristics of the 107 subjects.

Characteristic	Patients (n=46), mean ± SD	Controls (n=61), mean ± SD	P-value
Age, years	80.67±9.20	79.54±7.87	0.495
TP, g/l	68.79±6.91	65.48±9.45	0.099
ALB, g/l	38.32±3.83	36.86±3.61	0.090
GLB, g/l	30.46±5.32	29.55±4.58	0.418
A/G	1.29±0.22	1.28±0.20	0.768
ALT, U/l	13.87±10.68	18.20±13.46	0.193a
ALP, U/l	78.00±24.30	96.82±63.34	0.113a
TBA, µmol/l	6.91±3.84	6.07±5.86	0.503
AST, U/l	20.52±7.22	23.58±11.70	0.258a
Glu, mmol/l	5.23±1.58	5.53±2.71	0.444b
TG, mmol/l	1.35±0.76	1.42±0.98	0.896b
TC, mmol/l	4.48±1.04	4.28±1.22	0.378
HDL-C, mmol/l	1.12±0.27	1.04±0.30	0.118
ApoA, g/l	1.06±0.21	0.94±0.18	0.011c
ApoB, g/l	0.66±0.19	0.73±0.25	0.194
Lp(a), g/l	184.86±233.63	34.86±27.32	<0.001a,c
ApoE, mg/l	37.73±17.44	36.69±10.37	0.800
UREA, mmol/l	7.77±10.00	6.45±3.45	0.804b
CRE, µmol/l	82.73±47.25	78.52±30.04	0.626
UA, µmol/l	309.93±106.30	308.88±112.75	0.967
Hcy, µmol/l	19.76±10.82	17.67±20.84	0.046a,c
CRP, mg/l	6.20±11.72	15.00±26.21	0.016a,c

a Log-transformation was used.

b Nonparametric rank test was applied.

c Significant difference between patients and controls. TP, total protein; ALB, serum albumin; GLB, serum globulin; A/G, ALB/GLB; ALT, glutamic-pyruvic transaminase; ALP, alkaline phosphatase; TBA, total bile acid; AST, glutamic oxalacetic transaminase; Glu, blood glucose; TG, triglyceride; TC, total cholesterol; HDL-C, high-density lipoprotein cholesterol; ApoA, apolipoprotein A; ApoB, apolipoprotein B; Lp(a), lipoprotein A; ApoE, apolipoprotein E; UREA, carbamide; CRE, creatinine; UA, uric acid; Hcy, homocysteine; CRP, C reactive protein; SD, standard deviation.

Association tests of DRD4 methylation with AD

Significantly increased DRD4 methylation levels were observed in AD patients compared with controls (Table II; CpG1, P=0.001; CpG2, P=0.013; CpG3, P=0.001; CpG4, P<0.001; mean CpG1-4 methylation, P<0.001). Among 23 male patients and 45 male controls, elevated methylation levels of four CpG sites of the DRD4 promoter were observed following breakdown analysis by gender (Table II; CpG1, P=0.013; CpG2, P=0.012; CpG3, P<0.001; CpG4, P<0.001; mean CpG1-4 methylation, P<0.001). As shown in Table II, the power was sufficient in the overall (power=0.978) and male-based subgroup (power=0.976) case-control comparisons; however, the power was only 0.405 in the female-based subgroup analysis, suggesting that the negative association in the female subgroup may be due to a lack of power. A consequent breakdown analysis by gender revealed that the above significant association of DRD4 methylation with AD existed only in males (Fig. 3A; males, P<0.001; females, P=0.080).

Figure 3 DRD4 methylation comparisons between AD patients and controls and between Exelon-treated and Aricept-treated patients. (A) Comparison of DRD4 methylation levels between patients and controls. DRD4 methylation levels were significantly higher in male AD patients than male controls. (B) Comparison of CpG3 methylation levels between Exelon-treated and Aricept-treated patients. The Aricept-treated patients had significantly higher CpG3 methylation levels compared with Aricept-treated patients. DRD4, dopamine receptor D4; AUC, area under curve.

Table II Comparison of DRD4 DNA methylation levels between AD patients and controls.

Table II

Comparison of DRD4 DNA methylation levels between AD patients and controls.

A, All patients
Characteristic, %	Patients, mean ± SD (n=46)	Controls, mean ± SD (n=62)	P-value	Power
CpG1	16.50±5.14	13.13±4.67	0.001a,b	0.916
CpG2	12.26±4.84	10.13±3.92	0.013a,b	0.610
CpG3	14.98±5.27	11.59±4.59	0.001a,b	0.906
CpG4	14.96±4.22	10.92±4.90	<0.001a,b	0.987
Mean methylation	14.67±4.12	11.44±3.72	<0.001a,b	0.978

B, Male patients
Characteristic, %	Patients, mean ± SD (n=46)	Controls, mean ± SD (n=62)	P-value	Power
CpG1	16.74±4.92	13.68±4.55	0.013a,b	0.666
CpG2	12.87±3.94	10.39±3.62	0.012a,b	0.676
CpG3	15.61±4.35	11.16±4.27	<0.001a,b	0.975
CpG4	15.65±4.60	10.36±4.42	<0.001a,b	0.992
Mean methylation	15.22±3.72	11.40±3.47	<0.001a,b	0.976

C, Female patients
Characteristic, %	Patients, mean ± SD (n=46)	Controls, mean ± SD (n=62)	P-value	Power
CpG1	16.26±5.45	11.71±4.82	0.009a,b	0.720
CpG2	11.65±5.63	9.47±4.67	0.201	0.212
CpG3	14.35±6.10	12.71±5.31	0.380	0.125
CpG4	14.26±3.78	12.35±5.88	0.220	0.162
Mean methylation	14.13±4.51	11.56±4.40	0.080	0.405

a Significant difference between patients and controls.

b P-values remained significant after multiple tests. SD, standard deviation.

DRD4 methylation levels in AD patients with various treatments

Exelon and Aricept are commonly used acetylcholinesterase inhibitors in the treatment of AD (48). Both drugs aim to enhance cholinergic neurotransmission in specific parts of the brain and to improve the clinical symptoms of AD (50). The results of the present study revealed that the patients using Exelon-treated patients had significantly higher CpG3 methylation levels compared with Aricept-treated patients Aricept (Fig. 3B; P=0.044).

Discussion

The present study evaluated the levels of DRD4 promoter methylation in AD patients and matched controls to clarify the contribution of DRD4 promoter methylation to AD risk. It was observed that the DRD4 promoter methylation levels in AD patients were significantly higher than those in controls. In a breakdown analysis by gender, there were varying associations of methylation status in males and females. Positive results were identified for all four CpG sites observed in males. Positive correlations between DRD4 methylation and age, as well as DRD4 methylation and ApoA level, were also observed in males. In addition, clear positive correlations were observed between DRD4 methylation and ApoA and between DRD4 methylation and CRE in females. Furthermore, varying methylation levels were observed in patients who used Exelon and those who used Aricept.

Anomalies in dopaminergic transmission may lead to the disturbance of synaptic plasticity and advanced cognitive behavior (22). A male-specific association between DRD4 methylation and schizophrenia has been reported in Han Chinese (13). As a significant receptor of dopamine, DRD4 is pivotal to the development of AD (51). In the present study, significantly hypermethylated DRD4 promoters were observed in AD patients compared with controls. These results indicated that DRD4 may be involved in the progression of AD. Understanding the DNA methylation changes in the DRD4 promoter may aid in understanding the pathological mechanisms of AD. This information may also provide insight into the function of AD-associated genes, including DRD4, and help identify novel targets for therapeutic strategies to reverse the promoter methylation of DRD4.

Gender differences in AD have been widely documented. Females have a higher risk of AD at all ages, and the age-adjusted odds ratio for females has been shown to be 3.1 between AD patients and controls (52). Gender-specific DNA methylation exists in mice (53) and humans (12,54). DRD4 methylation research in another nervous system disorder, schizophrenia, has also identified a male-specific significant association (13). Significant differences were reported in all CpG sites observed in the present study in males, but a significant difference was observed in only CpG1 in females. These phenomena of varying methylation statuses in the DRD4 promoter also provide insight into gender differences in AD. The results of the present study support the idea that gender differences should be considered when establishing a clinical treatment plan for AD.

Furthermore, a significant association between DNA methylation and age in males but not in females was observed in the present study. Age is considered a major risk factor for AD (54). A previous study reported that DNA methylation is dynamically regulated in the human cerebral cortex throughout life, involves differentiated neurons and affects a substantial proportion of genes predominantly by an age-associated increase (55). Detailed descriptions of associations between DNA methylation and age in various gender subgroups require additional study.

A total of 22 phenotypes were analyzed in the present study. It is well-known that ApoA1, an increase in which leads to an increased risk of AD, is the major apolipoprotein constituent of high-density lipoprotein, and Apo A1 has been observed to affect brain cholesterol metabolism and angiogenesis (56). In an earlier study, serum ApoA concentration was shown to have a high correlation with the severity of AD (57). Significantly increased levels of ApoA1 were observed in AD patients compared with controls in the present study. Based on previous findings, it was speculated that Lp (a) may participate in the progression of dementia (58) and AD (59). The present study showed that AD patients had higher levels of Lp (a), which was similar to the results of a previous cross-sectional study (59). Hcy levels, another AD factor that may induce amyloid β accumulation, synaptic dysfunction and memory impairment, were significantly different between AD patients and matched controls in the present study (60–63). It has been reported that CRP participates in the systemic response to inflammation (64). Additional neuropathological studies have shown that CRP is associated with neurofibrillary tangles (65) and senile plaques (66) in AD brain tissue. The subjects with AD had significantly lower levels of plasma CRP than control subjects in the present study, which was consistent with the results of a previous study (64).

Aricept is an acetylcholinesterase inhibitor used for the symptomatic treatment of AD (67). Furthermore, Aricept is a high-affinity sigma-1 receptor antagonist, which has been investigated as a potential disease-modifying agent for several CNS disorders (64). Exelon is an oral drug approved by the US Food and Drug Administration for the treatment of AD (67). Significantly higher levels of DNA methylation were reported at the CpG3 site in the patients who were treated with Exelon compared with those treated with Aricept. This finding indicates that varying DNA methylation effects were produced by Exelon and Aricept. However, additional research is required to uncover the detailed association between DNA methylation, Exelon and Aricept.

The current study has a number of limitations. Firstly, the sample size was small, which may have influenced the results, particularly for the gender-stratified association test of DRD4 methylation with AD. In addition, the levels of DNA methylation of the four CpGs that were tested cannot represent the entire influence of DRD4. Additional studies investigating the DRD4 promoter or gene body regions are required. Furthermore, samples were mainly from the elderly, who may have underlying diseases. In other words, there may be certain unknown or potential risk factors for AD present in this study. Also, the DNA methylation level of DRD4 was measured in peripheral blood only, which may not be an accurate reflection of the situation in the brain tissue. Additional comprehensive studies on the association of DRD4 methylation within brain tissues and peripheral blood are required. Finally, four CpG positions were assessed per pyrosequencing read. Certain P-values may not retain their significance after being corrected for this number of CpG sites. The possibility that the present positive findings arose by chance cannot be excluded.

In conclusion, the present study supports DRD4 promoter hypermethylation as a risk factor for AD in males. Additionally, positive associations between DRD4 methylation and age, as well as DRD4 methylation and ApoA levels, were observed in males. Disparate DRD4 methylation levels were reported for patients taking various drugs: Patients taking Exelon appear to have higher levels than those taking Aricept. The results of the present study may contribute to an improved understanding of the molecular mechanisms underlying the pathophysiology of AD.

Acknowledgments

The present research was supported by grants from the National Natural Science Foundation of China (grant nos. 81371469, 81471938 and U1503223), the Natural Science Foundation of Zhejiang Province (grant no. Y15H090032), Ningbo International Science and Technology Cooperation Project (grant no. 2014D10019), the Public Technology Research and Social Development Project of Zhejiang Province (grant no. 2015C33155), the Disciplinary Project of Ningbo University (grant no. B01350104900) and K. C. Wong Magna Fund in Ningbo University.

References