Heterogeneity in White Blood Cells Has Potential to Confound DNA Methylation Measurements

Epigenetic studies are commonly conducted on DNA from tissue samples. However, tissues are ensembles of cells that may each have their own epigenetic profile, and therefore inter-individual cellular heterogeneity may compromise these studies. Here, we explore the potential for such confounding on DNA methylation measurement outcomes when using DNA from whole blood. DNA methylation was measured using pyrosequencing-based methodology in whole blood (n = 50–179) and in two white blood cell fractions (n = 20), isolated using density gradient centrifugation, in four CGIs (CpG Islands) located in genes HHEX (10 CpG sites assayed), KCNJ11 (8 CpGs), KCNQ1 (4 CpGs) and PM20D1 (7 CpGs). Cellular heterogeneity (variation in proportional white blood cell counts of neutrophils, lymphocytes, monocytes, eosinophils and basophils, counted by an automated cell counter) explained up to 40% (p<0.0001) of the inter-individual variation in whole blood DNA methylation levels in the HHEX CGI, but not a significant proportion of the variation in the other three CGIs tested. DNA methylation levels in the two cell fractions, polymorphonuclear and mononuclear cells, differed significantly in the HHEX CGI; specifically the average absolute difference ranged between 3.4–15.7 percentage points per CpG site. In the other three CGIs tested, methylation levels in the two fractions did not differ significantly, and/or the difference was more moderate. In the examined CGIs, methylation levels were highly correlated between cell fractions. In summary, our analysis detects region-specific differential DNA methylation between white blood cell subtypes, which can confound the outcome of whole blood DNA methylation measurements. Finally, by demonstrating the high correlation between methylation levels in cell fractions, our results suggest a possibility to use a proportional number of a single white blood cell type to correct for this confounding effect in analyses.

[1]  V. Gudnason,et al.  Age, Gene/Environment Susceptibility-Reykjavik Study: multidisciplinary applied phenomics. , 2007, American journal of epidemiology.

[2]  Devin C. Koestler,et al.  DNA methylation arrays as surrogate measures of cell mixture distribution , 2012, BMC Bioinformatics.

[3]  H. Tesch,et al.  DNA methylation profiles in the human genes for tumor necrosis factors alpha and beta in subpopulations of leukocytes and in leukemias. , 1991, Proceedings of the National Academy of Sciences of the United States of America.

[4]  G. Martin.  Epigenetic drift in aging identical twins. , 2005, Proceedings of the National Academy of Sciences of the United States of America.

[5]  R. Santella,et al.  Global methylation profiles in DNA from different blood cell types , 2011, Epigenetics.

[6]  N. Schork,et al.  Altered DNA Methylation in Leukocytes with Trisomy 21 , 2010, PLoS genetics.

[7]  A. Feinberg,et al.  Genome-wide methylation analysis of human colon cancer reveals similar hypo- and hypermethylation at conserved tissue-specific CpG island shores , 2008, Nature Genetics.

[8]  E. Buratti,et al.  Identification of a novel vertebrate homeobox gene expressed in haematopoietic cells. , 1992, Nucleic acids research.

[9]  A. Ashworth,et al.  HEX: a novel homeobox gene expressed during haematopoiesis and conserved between mouse and human. , 1993, Nucleic acids research.

[10]  Margaret R Karagas,et al.  Peripheral Blood Immune Cell Methylation Profiles Are Associated with Nonhematopoietic Cancers , 2012, Cancer Epidemiology, Biomarkers & Prevention.

[11]  S. Kochanek,et al.  Interindividual concordance of methylation profiles in human genes for tumor necrosis factors alpha and beta. , 1990, Proceedings of the National Academy of Sciences of the United States of America.

[12]  Shicai Fan,et al.  CpG island methylation pattern in different human tissues and its correlation with gene expression. , 2009, Biochemical and biophysical research communications.

[13]  A. Feinberg,et al.  Intra-individual change over time in DNA methylation with familial clustering. , 2008, JAMA.

[14]  Li Hsu,et al.  Tissue-specific variation in DNA methylation levels along human chromosome 1 , 2009, Epigenetics & Chromatin.

[15]  R Peto,et al.  Association of fibrinogen, C-reactive protein, albumin, or leukocyte count with coronary heart disease: meta-analyses of prospective studies. , 1998, JAMA.

[16]  Tom H. Pringle,et al.  The human genome browser at UCSC. , 2002, Genome research.

[17]  Martin J. Aryee,et al.  Personalized Epigenomic Signatures That Are Stable Over Time and Covary with Body Mass Index , 2010, Science Translational Medicine.

[18]  Hein Putter,et al.  Variation, patterns, and temporal stability of DNA methylation: considerations for epigenetic epidemiology , 2010, FASEB journal : official publication of the Federation of American Societies for Experimental Biology.

[19]  A. M. López,et al.  Prospective study of leukocyte count as a predictor of incident breast, colorectal, endometrial, and lung cancer and mortality in postmenopausal women. , 2007, Archives of internal medicine.

[20]  Kimberly D Siegmund,et al.  Epigenetic profiling of somatic tissues from human autopsy specimens identifies tissue- and individual-specific DNA methylation patterns. , 2009, Human molecular genetics.

[21]  A. Rustighi,et al.  Differential expression of a novel proline-rich homeobox gene (Prh) in human hematolymphopoietic cells. , 1995, Blood.

[22]  S. Kochanek,et al.  Interindividual concordance of methylation profiles in human genes for tumor necrosis factors a and fi , 2004 .

[23]  J. Rogers,et al.  DNA methylation profiling of human chromosomes 6, 20 and 22 , 2006, Nature Genetics.

[24]  Jim Stalker,et al.  A Novel CpG Island Set Identifies Tissue-Specific Methylation at Developmental Gene Loci , 2008, PLoS biology.