PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of medscimonISLHome PageSearch ArticleTable of ContentsSubmit ManuscriptSubscribe
 
Med Sci Monit. 2017; 23: 5851–5861.
Published online 2017 December 10. doi:  10.12659/MSM.905524
PMCID: PMC5733562

ATP Binding Cassette Subfamily A Member 2 (ABCA2) Expression and Methylation are Associated with Alzheimer’s Disease

Wanhua Hu,1,A,B,C,E Xiaodong Lin,2,A,B,C,D,E,F,G Huihe Zhang,1,A,B,D,F and Na Zhao1,B,C,D,F

Abstract

Background

ABCA2 has been genetically linked to Alzheimer’s disease (AD) risk, but its mRNA expression and epigenetics in AD have not been investigated.

Material/Method

To explore the diagnosis value of ABCA2 mRNA expression in AD, 2 datasets GES15222 and GSE33000 containing expression profile of brain cortex tissues and 2 datasets GSE63063 (Cohort 1) and GSE63063 (Cohort 2) containing expression profile of blood were downloaded from the NCBI GEO database and analyzed by receiver operating characteristic curve (ROC) analyses and logistic regression. The ABCA2 co-expressed genes were also analyzed by GO annotation to investigate the potential molecular mechanisms.

Results

The analyses results suggested ABCA2 mRNA expression was upregulated significantly in AD compared with controls in all datasets. ROC analysis suggested that ABCA2 was associated with AD in all datasets, which were also proved by univariate and multivariate analyses. Next, the dataset GSE80970 containing methylation profiles of prefrontal cortex tissues from AD patients were downloaded and analyzed. Methylation of 2 of 36 CpG islands in ABCA2 gene with high diagnostic accuracy of AD from controls in ROC analyses were found to be negatively associated with AD risk in univariate analysis. One was still associated with AD risk after adjustment of confounding factors. Additional analyses indicated that ACBA2 mRNA expression could be used to diagnose mild cognitive impairment (MCI) and Huntington’s disease (HD) from controls and to distinguish HD from AD, but not AD from MCI. Furthermore, the genes involved in AD during ABCA2 alteration were analyzed by GO analysis.

Conclusions

ABCA2 mRNA expression and methylation is associated AD risk. ABCA2 may be used as a biomarker for AD diagnosis and may be a potential therapeutic target of AD.

MeSH Keywords: Alzheimer Disease, ATP-Binding Cassette Transporters, Gene Expression Profiling, Methylation

Background

Alzheimer’s disease (AD), the most common neurodegenerative disease, is defined by short-term memory loss progressing into dementia and neuropathology [13]. It has been estimated that 46 million people are affected by dementia syndromes in 2015 and about 60~70% of them are diagnosed as AD [4]. The hippocampus and the surrounding cortical regions are the major sites of AD-related pathology [5]. The early and accurate diagnosis of AD is still a challenge and many patients are misclassified or diagnosed late [6]. AD is irreversible and progressive and there is currently no drug that halts or cures it [7]. Therefore, early and accurate diagnosis of AD, identification of people with risk factors, and early invention are crucial to decrease the socioeconomic burden of AD. However, currently available diagnostic approaches for AD, including brain imaging, are neither suited for mass population screening nor sufficiently cost-effective to be practical [8].

ATP-binding cassette (ABC) transporters are one of the largest protein superfamilies. There are 7 subfamilies, from A to G, of ABC transporters and a total of 49 members have been identified in humans [9,10]. These transporters are located throughout the brain, with a focus at the blood-brain barriers; they facilitate the strictly regulated influx/efflux of various substances, metabolites, and peptides, and hence protect the brain from toxic and harmful compounds [1113]. ABCA2, a member of the ABCA subfamily, is abundant in the frontal cortex of human AD brains compared to normal controls but was detected at lower concentrations in the parietal, occipital, and cerebellar regions [14]. In vitro studies reveal that it regulates the lipid metabolism via modulating the expression and turnover rate of low-density lipoprotein receptor (LDLR) [15,16]. Overexpression of ABCA2 can also enhance the levels of amyloid beta precursor protein (APP) and beta-amyloid, which is a key player in AD progression [17]. ABCA2 has been genetically linked to AD risk. The synonymous SNP rs908832, a C-T polymorphism in exon 14 of the ABCA2 gene, is associated with early-onset and late-onset or sporadic AD, as reported in 2 independent studies [18,19]. A meta-analysis also suggested a significant association between rs908832 and susceptibility to AD (OR=1.55, 95% CI=1.12–2.16) [20].

There have been no studies to date that investigated the mRNA expression and epigenetics of ABCA2 in AD. Here, we compared the mRNA expression of ABCA2 in brain tissues and blood in AD patients with the expression in controls by data mining of datasets from the NCBI GEO databases. The associations of ABCA2 mRNA expression and methylation with AD risk were estimated by receiver operating characteristics curves and logistic regression. In addition, the diagnostic value of ABCA2 mRNA level in distinguishing other neurodegenerative disease, such as Huntington’s disease and mild cognitive impairment, from AD or controls were also explored.

Material and Methods

To explore the diagnostic value of ABCA2 in differentiating Alzheimer’s disease patients from controls without related disease, AD gene expression profile data with corresponding clinical data were searched and downloaded from the publicly available GEO database. After reviewing the search results, GES15222 [21], GSE33000 [22], and GSE63063 [23] were finally chosen because these datasets consisted of many subjects. For further analyses, ABCA2 expression data was first normalized with Z-score method. Difference in ABCA2 expression levels between patients with Alzheimer’s disease and controls in each dataset was then compared using the Mann-Whitney test. Receiver operating characteristic curve (ROC) analyses were performed to investigate the relationship between ABCA2 expression levels and Alzheimer’s disease risk in every dataset. Area under the curves and the corresponding 95% confidence interval (CI) were calculated and compared with 0.5. The optimal cut-off value was calculated by Youden index [24]. ABCA2 expression levels were then divided into low expression and high expression groups by the cut-off values. Associations of ABCA2 with clinical features, age, and sex were determined by Fisher’s exact test or chi-square test. Univariate and multiple logistic regression analyses were also performed to explore the correlations of ABCA2 with Alzheimer’s disease. In addition, the GSE33000 and GSE63063 datasets contained the expression profile data of patients with other neurodegenerative diseases and Huntington’s disease, and mild cognitive impairment, respectively. ABCA2 expression levels among patients with different neurodegenerative diseases and controls were compared using the Kruskal-Wallis test. The associations of ABCA2 with the risk of Huntington’s disease or mild cognitive impairment and the diagnostic value of ABCA2 in distinguishing Alzheimer’s disease from Huntington’s disease or mild cognitive impairment were analyzed by ROC curves and logistic regression. Fisher’s exact test, chi-square test, and Kruskal-Wallis test were conducted via GraphPad Prism 6 software. Univariate and multivariate logistic regression were performed using SPSS 19.0 software. All tests were two-sided. P<0.05 was considered to be statistically significant.

To investigate the related genes during ABCA2 alteration that might be involved in Alzheimer’s disease, GSE33000, which contains the gene expression profile of brain tissue, was chosen for deep analyses. Differently expressed genes (DEGs) between Alzheimer’s disease and controls were obtained by the web tool GEO2R in NCBI. The expression correlations of the genes with ABCA2 in GSE33000 were examined by Pearson’s correlation analysis. Then, ABCA2 co-expressed genes with Pearson’s correlation coefficients above 0.5 (P value was less than 1E-20) were selected and overlapped with the DEGs whose P value was less than 1E-20. The final obtained genes were subjected to GO analysis in the GeneCoDis3 online database (http://genecodis.cnb.csic.es).

Results

ABCA2 was upregulated in Alzheimer’s disease

Four microarray datasets (GES15222 [21], GSE33000 [22], GSE63063 (Cohort 1) [23], and GSE63063 (Cohort 2) [23]) were downloaded from NCBI GEO databases. The characteristics of these datasets are shown in Table 1. Studies of GSE15222 and GSE33000 analyzed the brain transcriptome in postmortem samples, whereas GSE6306 used patient blood as the source of RNA. GSE63063 consisted of 2 independent cohorts. In addition to the Alzheimer’s disease patients, GSE3300 also included patients with Huntington’s disease and GSE6306 also included patients with mild cognitive impairment. In total, there were 582 controls, 770 cases with Alzheimer’s disease, 189 cases with mild cognitive impairment, and 157 cases with Huntington’s disease. ABCA2 mRNA profile data were collected from all the datasets and normalized with Z-score method. The difference in ABCA2 mRNA levels was first compared between Alzheimer’s disease patients and controls, and the statistical analyses suggested ABCA2 was significantly upregulated in Alzheimer’s disease patients in all datasets.

Table 1
Characteristics of the included datasets.

Association of ABCA2 with clinical features, age, and sex, and risk of Alzheimer’s disease

Receiver operating characteristics (ROC) curves were used to investigate the diagnostic value of ABCA2 in Alzheimer’s (Figure 1). The results suggested that ABCA2 was associated with Alzheimer’s disease in all datasets (GSE15222, AUC=0.6575, 95%CI=0.6014–0.7136, P<0.0001; GSE33000, AUC=0.7243, 95%CI=0.6760–0.7727, P<0.0001; GSE63063, Cohort 1, AUC=0.7146, 95%CI=0.6545–0.7750, P<0.0001; GSE63063, Cohort 2, AUC=0.6562, 95%CI=0.5866–0.7258, P<0.0001; GSE63063, Cohort 1+2, AUC=0.6841, 95%CI=0.6388–0.7295, P<0.0001). Then, cut-off values were obtained by calculating the Youden index [24] from the ROC curves and used to divide the ABCA2 expression into low expression and high expression groups. Fisher’s test analyses indicated that ABCA2 was significantly associated with Alzheimer’s disease (Table 2). The associations of ABCA2 with the age and sex of the patients were next estimated, and age was found to be correlated to ABCA2 expression in all datasets except for the combination of both cohorts of GSE63063 (Table 3). No significant association was observed between sex and ABCA2 expression. Then, we performed further analyses via logistic regression to investigate the associations of Alzheimer’s disease with age, sex, and especially ABCA2 expression (Table 4). Univariate analysis revealed that ABCA2 was associated with AD risk in all datasets. Age was also found to be associated with AD risk in all datasets and sex was only associated with AD risk in the GSE33000 dataset. After adjusting for the confounding factors and age, we found a significant association of ABCA2 with AD risk in each dataset (GSE15000, OR=2.446, 95%CI=1.359–4.404, P=0.003; GSE33000, OR=2.774, 95%CI=1.770–4.347, P<0.001, GSE63063, Cohort 1, OR=2.391, 95%CI=1.408–4.060, P=0.001; GSE63063, Cohort 2, OR=3.748, 95%CI=2.148–6.542, P<0.001; GSE63063, Cohort 1+2, OR=3.157, 95%CI=2.180–4.572, P<0.001).

Figure 1
Diagnostic value of ABCA2 in differentiating patients with Alzheimer’s disease from control subjects. NCBI GEO datasets GSE15222 (A), GSE330000 (B), GSE63063 (Cohort 1, C), and GSE63063 (Cohort 2, D) were downloaded from the online database. ABCA2 ...
Table 2
ABCA2 mRNA level comparison between patients with Alzheimer’s disease and control subjects in included datasets.
Table 3
Associations of ABCA2 expression with AD risk and clinical parameters, age and gender.
Table 4
Associations of ABCA2 with AD risk determined by Univariate and multivariate analysis.

As methylation might influence expression of ABCA2, we then investigated the associations of ABCA2 methylation with AD risk. The GSE80970 dataset was downloaded from the GEO database, which contained the methylation profile of prefrontal cortex tissues from 74 AD patients and 68 controls. All the 36 CpG islands within ABCA2 gene were first analyzed by ROC to explore their diagnostic value in AD. However, the AUC of all CpG islands methylation was not significantly different in the ROC analyses (Table 5). Two of the CpG islands, cg13930557 and cg03349123, were significantly different (P values 0.086 and 0.079, respectively). The univariate logistic regression analysis revealed that high methylation level of both the CpG islands was negatively and significantly associated with AD risk (cg13930557, OR=0.465, 95%CI=0.229–0.944, P=0.034; cg03349123, OR=0.393, 95%CI=0.195–0.791, P=0.009; Table 6). Only cg03349123 was found to be a potential negative independent indicator of AD risk (OR=0.436, 95%CI=0.203–0.936, P=0.033).

Table 5
Diagnosis value of ABCA2 methylaitons in AD demonstrated by receiver operation characteristics curve analyses.
Table 6
Associations of ABCA2 methylation with AD risk determined by Univariate and multivariate analysis.

Diagnostic value of ABCA2 in differentiating AD from Huntington’s disease or mild cognitive impairment

The GSE33000 and GSE63063 datasets also contained the data of patients with Huntington’s disease and mild cognitive impairment, respectively, and were used to determine the association of ABCA2 with HD risk or MCI risk and the diagnostic value of ABCA2 in distinguishing AD from other neurodegenerative diseases. The differences in ABCA2 mRNA levels among the patients with neurodegenerative diseases and controls were first compared. In GSE33000, the expression of ABCA2 was increased sequentially and significantly in controls, AD patients, and HD patients (Figure 2A). In GSE63063 Cohort 1, Cohort 2, and integrated analyses of both cohorts, ABCA2 expression was higher in MCI patients compared with controls (Figure 2B–2D). However, there was no significant difference in ABCA2 expression between MCI patients and controls (Figure 2B–2D). Then, ROC curve analyses were performed and the results suggested that ABCA2 could be used to distinguish HD patients from controls, HD from AD, and MCI from controls, but not AD from MCI (Figure 3). As shown in Table 7, univariate and multivariate logistic regression analyses also indicated that ABCA2 was associated with HD compared to controls (OR=20.630, 95%CI=11.273–37.753, P<0.001), with HD compared to AD (OR=6.139, 95%CI=3.421–11.016, P<0.001), and with MCI compared to controls (OR=2828, 95%CI=1.878–4.257, P<0.001), but not with AD compared to MCI.

Figure 2
Comparison of ABCA2 levels in AD patients with that in patients with other neurodegenerative diseases, Huntington’s disease (HD), and mild cognitive impairment (MCI). Besides AD patients, GSE33000 also included the ABCA2 expression profile of ...
Figure 3
Diagnostic value of ABCA2 in differentiating the patients with Alzheimer’s disease from those with other neurodegenerative diseases. Receiver operating characteristics curves (ROC) analyses were performed between HD and controls (A), and HD and ...
Table 7
Associations of ABCA2 with HD risk and MCI risk determined by Univariate and multivariate analysis.

Analyses of the genes involved in Alzheimer’s disease during ABCA2 alteration

GSE33000, consisting of ABCA2 mRNA expression profiles in brain tissues, was chosen to use in analyzing the related genes. First, GEO2R, an online tool in NCBI, was used to obtain the differently expressed genes (DEGs) between Alzheimer’s disease patients and controls. Then, Pearson correlation analyses were carried out to find the genes co-expressed with ABCA2. The genes derived from these processes overlapped. There were 757 genes remaining, and these were subjected to GO analysis in the GeneCoDis3 online database (http://genecodis.cnb.csic.es). The most enriched biological processes of these genes were signal transduction, respiratory electron transport chain, intracellular protein transport, and regulation of transcription. The most enriched KEGG pathways were Huntington’s disease, Parkinson’s disease, oxidative phosphorylation, cancer, and Alzheimer’s disease.

Discussion

ABCA2 has been shown to be genetically linked to AD susceptibility, but whether its expression and methylation are associated with AD risk was not investigated. In the present study, the NCBI GEO database for the datasets containing expression profile of AD disease and controls and 4 datasets – GEO15222, GSE33000, GSE63063 (cohort 1), and GSE63063 (cohort 2) – were derived and downloaded. The sample source of GEO15222 and GE33000 was brain prefrontal cortex tissue and that of GES63063 was blood. Overexpression of ABCA2 was observed in AD compared with controls in all dataset. ROC curves and univariate and multivariate analyses suggested ABCA2 overexpression was associated with AD risk. A methylation profile dataset of AD and controls, GSE80970, was selected to explore the association of ABCA2 methylation with AD risk. Methylation of 2 CpG islands with high diagnostic accuracy in ROC curve analysis was found to be negatively correlated with AD risk. Additional analyses indicated that ACBA2 mRNA expression could be used for diagnosis of other neurodegenerative diseases, mild cognitive impairment (MCI), and Huntington’s disease (HD) and to distinguish HD from AD, but not AD from MCI.

The major sites of AD-related pathology in the brain were the hippocampus and the surrounding cortical regions, which were characterized by the increasing accumulation of amyloid beta plaques and tau-related neurofibrillary tangles [5]. The changes in genes in disease sites reflected the “real” regulation of the genes during AD progression; however, sampling from brain tissue is impossible for AD diagnosis. Cerebrospinal fluid (CSF) and blood are alternative sample sources. CSF has contact with extracellular spaces and reflects molecular events; it has been used in several clinical trials for treating AD and Parkinson’s disease [25,26]. Compared with CSF, blood samples are easy, non-invasive, and economical to obtain. Although blood RNA regulation has the same mechanisms as the brain, some inconsistencies may exist. Here, the datasets GSE15000 and GSE3300 used brain cortex from postmortem exams as the sample source and GSE63063 used peripheral blood as the sample source. Both studies of GSE15222 and GSE33000 analyzed the brain transcriptome in postmortem samples, whereas GSE6306 used patient blood as the source of RNA. The associations of ABCA2 mRNA expression with AD in brain tissues and blood were consistent, indicating ABCA2 in blood reflected ABCA2 alteration in brains and could be used as a biomarker for diagnosis of AD.

RNA expression is under genetic, epigenetic, and environmental control [2730]. Although ABCA2 SNP rs908832 was identified to be associated with AD susceptibility, the mutation is a synonymous mutation, meaning that the transition of U to C does not change. We speculate that ABCA2 expression is correlated with its methylation. We found the CpG cg03349123 methylation in ABCA2 was negatively associated with AD risk, consistent with ABCA2 mRNA results, because the transcriptional activity is usually downregulated by the methylation of the gene. This speculation needs to be verified in further studies.

Several shortcomings exist in the present study. First, the patients included in the datasets were mainly white and the effects of ethnicity could not be analyzed. Second, although the methylation of ABCA2 was associated with susceptibility to AD, its correlations with transcriptional activity of ABCA2 were not verified. Third, the sample sizes of the patients with Huntington’s disease and mild cognitive impairment were relative small. Further well-designed studies with larger sample sizes and more diverse ethnic composition should be performed to verify our conclusions.

Conclusions

ABCA2 mRNA expression and methylation is associated with AD risk. The results of ABCA2 mRNA expression in brain cortex in diagnosis of AD were consistent with that in blood. ABCA2 may be used for early diagnosis of AD and is as a potential therapeutic target of AD.

Footnotes

Conflict of interest

None.

Source of support: The study was support by the Studio of Ouyue Famous doctor Wanhua Hu

References

1. Mayeux R, Stern Y. Epidemiology of Alzheimer disease. Cold Spring Harb Perspect Med. 2012;2(8):a006239. [PMC free article] [PubMed]
2. Perl DP. Neuropathology of Alzheimer’s disease. Mt Sinai J Med. 2010;77(1):32–42. [PMC free article] [PubMed]
3. Schellenberg GD, Montine TJ. The genetics and neuropathology of Alzheimer’s disease. Acta Neuropathol. 2012;124(3):305–23. [PMC free article] [PubMed]
4. Scarano S, Lisi S, Ravelet C, et al. Detecting Alzheimer’s disease biomarkers: From antibodies to new bio-mimetic receptors and their application to established and emerging bioanalytical platforms – A critical review. Anal Chim Acta. 2016;940:21–37. [PubMed]
5. Armstrong RA. The molecular biology of senile plaques and neurofibrillary tangles in Alzheimer’s disease. Folia Neuropathol. 2009;47(4):289–99. [PubMed]
6. Humpel C. Identifying and validating biomarkers for Alzheimer’s disease. Trends Biotechnol. 2011;29(1):26–32. [PMC free article] [PubMed]
7. Salloway S, Sperling R, Fox NC, et al. Two phase 3 trials of bapineuzumab in mild-to-moderate Alzheimer’s disease. New Engl J Med. 2014;370(4):322–33. [PMC free article] [PubMed]
8. Biasutti M, Dufour N, Ferroud C, et al. Cost-effectiveness of magnetic resonance imaging with a new contrast agent for the early diagnosis of Alzheimer’s disease. PLoS One. 2012;7(4):e35559. [PMC free article] [PubMed]
9. Pahnke J, Langer O, Krohn M. Alzheimer’s and ABC transporters – new opportunities for diagnostics and treatment. Neurobiol Dis. 2014;72(Pt A):54–60. [PMC free article] [PubMed]
10. Cascorbi I. Role of pharmacogenetics of ATP-binding cassette transporters in the pharmacokinetics of drugs. Pharmacol Ther. 2006;112(2):457–73. [PubMed]
11. Bernstein HG, Holzl G, Dobrowolny H, et al. Vascular and extravascular distribution of the ATP-binding cassette transporters ABCB1 and ABCC1 in aged human brain and pituitary. Mech Ageing Dev. 2014;141–142:12–21. [PMC free article] [PubMed]
12. Begley DJ. ABC transporters and the blood-brain barrier. Curr Pharm Des. 2004;10(12):1295–312. [PubMed]
13. Schinkel AH, Jonker JW. Mammalian drug efflux transporters of the ATP binding cassette (ABC) family: An overview. Adv Drug Deliv Rev. 2003;55(1):3–29. [PubMed]
14. Chen ZJ, Vulevic B, Ile KE, et al. Association of ABCA2 expression with determinants of Alzheimer’s disease. FASEB J. 2004;18(10):1129–31. [PubMed]
15. Smith JD, Le Goff W, Settle M, et al. ABCA1 mediates concurrent cholesterol and phospholipid efflux to apolipoprotein A-I. J Lipid Res. 2004;45(4):635–44. [PubMed]
16. Davis W., Jr The ATP-binding cassette transporter-2 (ABCA2) regulates cholesterol homeostasis and low-density lipoprotein receptor metabolism in N2a neuroblastoma cells. Biochim Biophys Acta. 2011;1811(12):1152–64. [PMC free article] [PubMed]
17. Davis W., Jr The ATP-binding cassette transporter-2 (ABCA2) increases endogenous amyloid precursor protein expression and Abeta fragment generation. Curr Alzheimer Res. 2010;7(7):566–77. [PMC free article] [PubMed]
18. Mace S, Cousin E, Ricard S, et al. ABCA2 is a strong genetic risk factor for early-onset Alzheimer’s disease. Neurobiol Dis. 2005;18(1):119–25. [PubMed]
19. Wollmer MA, Kapaki E, Hersberger M, et al. Ethnicity-dependent genetic association of ABCA2 with sporadic Alzheimer’s disease. Am J Med Genet B Neuropsychiatr Genet. 2006;141B(5):534–36. [PubMed]
20. Davis W., Jr The ATP-Binding Cassette Transporter-2 (ABCA2) overexpression modulates sphingosine levels and transcription of the Amyloid Precursor Protein (APP) gene. Curr Alzheimer Res. 2015;12(9):847–59. [PMC free article] [PubMed]
21. Webster JA, Gibbs JR, Clarke J, et al. Genetic control of human brain transcript expression in Alzheimer disease. Am J Hum Genet. 2009;84(4):445–58. [PubMed]
22. Narayanan M, Huynh JL, Wang K, et al. Common dysregulation network in the human prefrontal cortex underlies two neurodegenerative diseases. Mol Syst Biol. 2014;10:743. [PMC free article] [PubMed]
23. Sood S, Gallagher IJ, Lunnon K, et al. A novel multi-tissue RNA diagnostic of healthy ageing relates to cognitive health status. Genome Biol. 2015;16:185. [PMC free article] [PubMed]
24. Youden WJ. Index for rating diagnostic tests. Cancer. 1950;3(1):32–35. [PubMed]
25. Peskind ER, Riekse R, Quinn JF, et al. Safety and acceptability of the research lumbar puncture. Alzheimer Dis Assoc Disord. 2005;19(4):220–25. [PubMed]
26. Kim D, Kim YS, Shin DW, et al. Harnessing cerebrospinal fluid biomarkers in clinical trials for treating Alzheimer’s and Parkinson’s diseases: Potential and challenges. J Clin Neurol. 2016;12(4):381–92. [PMC free article] [PubMed]
27. Horvath S, Zhang Y, Langfelder P, et al. Aging effects on DNA methylation modules in human brain and blood tissue. Genome Biol. 2012;13(10):R97. [PMC free article] [PubMed]
28. Schwanhausser B, Busse D, Li N, et al. Global quantification of mammalian gene expression control. Nature. 2011;473(7347):337–42. [PubMed]
29. Westra HJ, Franke L. From genome to function by studying eQTLs. Biochim Biophys Acta. 2014;1842(10):1896–902. [PubMed]
30. Keller P, Vollaard NB, Gustafsson T, et al. A transcriptional map of the impact of endurance exercise training on skeletal muscle phenotype. J Applied Physiol. 2011;110(1):46–59. [PubMed]

Articles from Medical Science Monitor : International Medical Journal of Experimental and Clinical Research are provided here courtesy of International Scientific Literature, Inc.