DNA methylation and associated gene expression in blood prior to lung cancer diagnosis in the Norwegian Women and Cancer cohort

Abstract

The majority of lung cancer is caused by tobacco smoking, and lung cancer-relevant epigenetic markers have been identified in relation to smoking exposure. Still, smoking-related markers appear to mediate little of the effect of smoking on lung cancer. Thus in order to identify disease-relevant markers and enhance our understanding of pathways, a wide search is warranted. Through an epigenome-wide search within a case-control study (131 cases, 129 controls) nested in a Norwegian prospective cohort of women, we found 25 CpG sites associated with lung cancer. Twenty-three were classified as associated with smoking (LC-AwS), and two were classified as unassociated with smoking (LC-non-AwS), as they remained associated with lung cancer after stringent adjustment for smoking exposure using the comprehensive smoking index (CSI): cg10151248 (PC, CSI-adjusted odds ratio (OR) = 0.34 [0.23–0.52] per standard deviation change in methylation) and cg13482620 (B3GNTL1, CSI-adjusted OR = 0.33 [0.22–0.50]). Analysis among never smokers and a cohort of smoking-discordant twins confirmed the classification of the two LC-non-AwS CpG sites. Gene expression profiles demonstrated that the LC-AwS CpG sites had different enriched pathways than LC-non-AwS sites. In conclusion, using blood-derived DNA methylation and gene expression profiles from a prospective lung cancer case-control study in women, we identified 25 CpG lung cancer markers prior to diagnosis, two of which were LC-non-AwS markers and related to distinct pathways.

Agnostic Cys34‐albumin adductomics and DNA methylation: Implication of N‐acetylcysteine in lung carcinogenesis years before diagnosis

“…As detailed in the Supporting Information, we ran a series of linear mixed models for each disease‐related adduct and each of the 2,670 smoking‐related CpG, setting the methylation level as the variable of interest. These models also accounted for technical variation in the methylation data using a two‐step strategy first estimating, and second removing technically‐induced shifts in measured methylation levels . Results from these analyses were visualized as a bipartite network where edges were selected based on statistical significance of the pairwise associations, correcting for 2,670 tests.…”

Lifetime Ultraviolet Radiation Exposure and DNA Methylation in Blood Leukocytes: The Norwegian Women and Cancer Study

“…Approximately 50 000 women in the NOWAC cohort donated blood samples and constitute the postgenome cohort 21 . The present paper includes controls from three data sets from the postgenome cohort, all cancer-free women at the time of blood sampling and selected as controls in case-control studies of melanoma (discovery set, n = 183 controls), breast cancer (replication set R 1 , n = 191 controls) 22 , and lung cancer (replication set R 2 , n = 125 controls) 5,23 . Matching factors were time since blood sampling and year of birth (1943-1947, 1948-1952, 1953-1957).…”

Integrative Analysis of Multi-omics Data Improves Model Predictions: An Application to Lung Cancer

“…Integrated PCA has been implemented in R and its code is available at [25]. Note that iPCA, unlike aJIVE, does not require the specification of the initial ranks, al- Both methylation and gene expression have been shown to associate with the occurrence and characteristics of lung cancer [26][27][28][29] but it is also known that the different data sources might contribute together and jointly relate to these biological outcomes [30,31]. We apply aJIVE and iPCA to estimate such joint and individual contributions, and use these components in prediction models for the occurrence of lung cancer and classification of tumor types.…”

“…Laboratory processing and microarray analyses for mRNA expression and DNA methylation are described in [31]. For miRNA, laboratory processing included miRNA isolation and purification from 100 µl plasma using the Qiagen miRNeasy The filtering of miRNA expressions was based on the counts per million, that is the total read counts of a miRNA divided by the total read counts of the sample and multiplied by 10 6 , and signals having less than one count per million were excluded.…”