Genome-Wide Investigation of Regional Blood-Based DNA Methylation Adjusted for Complete Blood Counts Implicates BNC2 in Ovarian Cancer

Stacey J Winham, Sebastian M. Armasu, Mine Cicek, Melissa C. Larson, Julie M Cunningham, Kimberly R. Kalli, Brooke L. Fridley, Ellen L Goode

Research output: Contribution to journalArticle

10 Citations (Scopus)

Abstract

Due to its potential as a biomarker for early cancer detection, blood-based DNA methylation (DNAm) is of interest in cancer research. Specifically, highly predictive mechanisms for early detection of epithelial ovarian cancer (EOC) are desired, so previous studies have compared DNAm between EOC cases and controls. However, case-control studies are confounded by the distribution of white blood cell types through an immune response induced by the cancer. Rather than determining the distribution of the cell types manually or investigating isolated cell types, an alternative approach involves the use of complete blood count (CBC), which is routinely collected. In the analysis of an EOC case-control study of DNAm, we incorporate CBC measures to adjust for this confounding and compare DNAm between 242 EOC cases and 181 age-matched controls (assayed on the Illumina Infinium HumanMethylation27 or HumanMethylation450 Beadchips), at both the individual CpG and CpG island levels. We found that adjustment for leukocyte distribution using CBC measurements dramatically reduced confounding, with 62 single CpG sites found to be associated with EOC status after adjustment (P < 5E-8). Additionally, regional DNAm was assessed by applying principal components analysis to CpG islands. The top associated CpG island (P = 7E-6) was located in the promoter/transcription start site of the human basonuclin 2 gene (BNC2), a known susceptibility gene for EOC risk identified through GWAS. Follow-up studies are necessary to establish the role of BNC2 in blood-based DNA and EOC, including prospective studies to validate this region as a potential biomarker and predictor of EOC susceptibility.

Original languageEnglish (US)
Pages (from-to)457-466
Number of pages10
JournalGenetic Epidemiology
Volume38
Issue number5
DOIs
StatePublished - 2014

Fingerprint

Blood Cell Count
DNA Methylation
Ovarian Neoplasms
Genome
CpG Islands
Genes
Case-Control Studies
Leukocytes
Biomarkers
Ovarian epithelial cancer
Genome-Wide Association Study
Transcription Initiation Site
Principal Component Analysis
Early Detection of Cancer
Neoplasms
Prospective Studies
DNA
Research

Keywords

  • Complete blood count
  • CpG island
  • Epigenetics
  • Etiology
  • Principal components

ASJC Scopus subject areas

  • Genetics(clinical)
  • Epidemiology
  • Medicine(all)

Cite this

Genome-Wide Investigation of Regional Blood-Based DNA Methylation Adjusted for Complete Blood Counts Implicates BNC2 in Ovarian Cancer. / Winham, Stacey J; Armasu, Sebastian M.; Cicek, Mine; Larson, Melissa C.; Cunningham, Julie M; Kalli, Kimberly R.; Fridley, Brooke L.; Goode, Ellen L.

In: Genetic Epidemiology, Vol. 38, No. 5, 2014, p. 457-466.

Research output: Contribution to journalArticle

@article{89f5cd1190d74f39b3fbc22aebc462fa,
title = "Genome-Wide Investigation of Regional Blood-Based DNA Methylation Adjusted for Complete Blood Counts Implicates BNC2 in Ovarian Cancer",
abstract = "Due to its potential as a biomarker for early cancer detection, blood-based DNA methylation (DNAm) is of interest in cancer research. Specifically, highly predictive mechanisms for early detection of epithelial ovarian cancer (EOC) are desired, so previous studies have compared DNAm between EOC cases and controls. However, case-control studies are confounded by the distribution of white blood cell types through an immune response induced by the cancer. Rather than determining the distribution of the cell types manually or investigating isolated cell types, an alternative approach involves the use of complete blood count (CBC), which is routinely collected. In the analysis of an EOC case-control study of DNAm, we incorporate CBC measures to adjust for this confounding and compare DNAm between 242 EOC cases and 181 age-matched controls (assayed on the Illumina Infinium HumanMethylation27 or HumanMethylation450 Beadchips), at both the individual CpG and CpG island levels. We found that adjustment for leukocyte distribution using CBC measurements dramatically reduced confounding, with 62 single CpG sites found to be associated with EOC status after adjustment (P < 5E-8). Additionally, regional DNAm was assessed by applying principal components analysis to CpG islands. The top associated CpG island (P = 7E-6) was located in the promoter/transcription start site of the human basonuclin 2 gene (BNC2), a known susceptibility gene for EOC risk identified through GWAS. Follow-up studies are necessary to establish the role of BNC2 in blood-based DNA and EOC, including prospective studies to validate this region as a potential biomarker and predictor of EOC susceptibility.",
keywords = "Complete blood count, CpG island, Epigenetics, Etiology, Principal components",
author = "Winham, {Stacey J} and Armasu, {Sebastian M.} and Mine Cicek and Larson, {Melissa C.} and Cunningham, {Julie M} and Kalli, {Kimberly R.} and Fridley, {Brooke L.} and Goode, {Ellen L}",
year = "2014",
doi = "10.1002/gepi.21815",
language = "English (US)",
volume = "38",
pages = "457--466",
journal = "Genetic Epidemiology",
issn = "0741-0395",
publisher = "Wiley-Liss Inc.",
number = "5",

}

TY - JOUR

T1 - Genome-Wide Investigation of Regional Blood-Based DNA Methylation Adjusted for Complete Blood Counts Implicates BNC2 in Ovarian Cancer

AU - Winham, Stacey J

AU - Armasu, Sebastian M.

AU - Cicek, Mine

AU - Larson, Melissa C.

AU - Cunningham, Julie M

AU - Kalli, Kimberly R.

AU - Fridley, Brooke L.

AU - Goode, Ellen L

PY - 2014

Y1 - 2014

N2 - Due to its potential as a biomarker for early cancer detection, blood-based DNA methylation (DNAm) is of interest in cancer research. Specifically, highly predictive mechanisms for early detection of epithelial ovarian cancer (EOC) are desired, so previous studies have compared DNAm between EOC cases and controls. However, case-control studies are confounded by the distribution of white blood cell types through an immune response induced by the cancer. Rather than determining the distribution of the cell types manually or investigating isolated cell types, an alternative approach involves the use of complete blood count (CBC), which is routinely collected. In the analysis of an EOC case-control study of DNAm, we incorporate CBC measures to adjust for this confounding and compare DNAm between 242 EOC cases and 181 age-matched controls (assayed on the Illumina Infinium HumanMethylation27 or HumanMethylation450 Beadchips), at both the individual CpG and CpG island levels. We found that adjustment for leukocyte distribution using CBC measurements dramatically reduced confounding, with 62 single CpG sites found to be associated with EOC status after adjustment (P < 5E-8). Additionally, regional DNAm was assessed by applying principal components analysis to CpG islands. The top associated CpG island (P = 7E-6) was located in the promoter/transcription start site of the human basonuclin 2 gene (BNC2), a known susceptibility gene for EOC risk identified through GWAS. Follow-up studies are necessary to establish the role of BNC2 in blood-based DNA and EOC, including prospective studies to validate this region as a potential biomarker and predictor of EOC susceptibility.

AB - Due to its potential as a biomarker for early cancer detection, blood-based DNA methylation (DNAm) is of interest in cancer research. Specifically, highly predictive mechanisms for early detection of epithelial ovarian cancer (EOC) are desired, so previous studies have compared DNAm between EOC cases and controls. However, case-control studies are confounded by the distribution of white blood cell types through an immune response induced by the cancer. Rather than determining the distribution of the cell types manually or investigating isolated cell types, an alternative approach involves the use of complete blood count (CBC), which is routinely collected. In the analysis of an EOC case-control study of DNAm, we incorporate CBC measures to adjust for this confounding and compare DNAm between 242 EOC cases and 181 age-matched controls (assayed on the Illumina Infinium HumanMethylation27 or HumanMethylation450 Beadchips), at both the individual CpG and CpG island levels. We found that adjustment for leukocyte distribution using CBC measurements dramatically reduced confounding, with 62 single CpG sites found to be associated with EOC status after adjustment (P < 5E-8). Additionally, regional DNAm was assessed by applying principal components analysis to CpG islands. The top associated CpG island (P = 7E-6) was located in the promoter/transcription start site of the human basonuclin 2 gene (BNC2), a known susceptibility gene for EOC risk identified through GWAS. Follow-up studies are necessary to establish the role of BNC2 in blood-based DNA and EOC, including prospective studies to validate this region as a potential biomarker and predictor of EOC susceptibility.

KW - Complete blood count

KW - CpG island

KW - Epigenetics

KW - Etiology

KW - Principal components

UR - http://www.scopus.com/inward/record.url?scp=84902958088&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84902958088&partnerID=8YFLogxK

U2 - 10.1002/gepi.21815

DO - 10.1002/gepi.21815

M3 - Article

VL - 38

SP - 457

EP - 466

JO - Genetic Epidemiology

JF - Genetic Epidemiology

SN - 0741-0395

IS - 5

ER -