Paternal Smoking and DNA Methylation in Childhood Leukemia

Sponsor: University of California Tobacco-Related Disease Research Program (TRDRP)

Location(s): United States


Smoking causes a number of cancers and other health problems. One question that epidemiologists have asked for years is: does parental smoking cause leukemia in their children? The answer to many studies is: “not clear,” since there does appear to be some increased leukemia rate in children in families with smokers, but most studies are too small to reach statistical significance.

One of the issues is that this increased risk is more often linked to the father smoking rather than the mother. Three new studies have been published in the last year providing more weight of evidence to the link between father’s smoking and childhood leukemia. We have seen this same pattern in our Northern California Childhood Leukemia Study (NCCLS). We believe that we can only prove this link by finding the biological imprint left behind from smoking-related carcinogens on leukemia cells. The discovery of such imprints will confirm the link between parental smoking and childhood leukemia, which will in turn lead to very strong public health messages against smoking in the family home. We present preliminary data in this grant proposal that shows that the pattern of DNA methylation in children who have smoking fathers and mothers in the home is different than leukemias that arise from other mechanisms. This same phenomenon has been found in lung and head and neck cancers, in which early teen smoking seems to result in a DNA methylation imprint that impacts the risk and the biology of the cancers that develop decades later. The biological “imprint” from parental smoking is likely to happen very early, as paternal smoking before conception is strongly related to the methylation pattern and may be established during sperm development. We are able to assess DNA methylation in thousands of locations within the human genome at the same time, and we plan to sort the significant DNA methylation events from the bystanders by doing additional tests with different technologies and study subjects, focusing on paternal smoking’s effects on methylation within the human genome. Our leukemias come from our NCCLS, which is one of the largest population-based epidemiology studies in the world, with a superb set of biological materials such as diagnostic DNA, normal DNA from cheek cells, and Guthrie cards available for study. This study includes about 42% Hispanics, 42% Whites, and the remainder African and Asian Americans, so we are capturing the diversity of the California population. Our ultimate goal is to establish a firm link between parental smoking and childhood leukemia that bridges the two with a mechanistic framework. We will then be able to bring this message out firmly to young adults who are considering a family, certainly an important message for this target population.