Department of Human Genetics Doctoral Candidate Jon Chernus defends his dissertation on "Methods and Genome-Wide Association Study for Meiotic Nondisjunction of Chromosome 21"

ADVISOR: Eleanor Feingold, PhD

COMMITTEE MEMBERS:

• Candace M. Kamerer, PhD

• John R. Shaffer, PhD

• George C. Tseng, ScD

ABSTRACT:

Up to one quarter of human conceptions may be aneuploid, having too many or too few chromosomes relative to the standard complement of 23 pairs. Most often this is the result of nondisjunction in maternal meiosis, making such errors a leading cause of pregnancy loss and congenital abnormalities. Prior research has established advancing age and altered patterns of meiotic recombination as risk factors for maternal nondisjunction in meiosis and has shown that meiosis I and II errors may involve different mechanisms, but potential genetic risk factors have not been systematically investigated. The goal of this dissertation is to advance our knowledge of aneuploidy by identifying and characterizing common genetic variants associated with maternal meiotic nondisjunction of chromosome 21, the most common aneuploid condition in conceptions that survive to term.

The first specific aim was to perform a candidate gene and genome-wide association study (GWAS) in which cases are mothers who have had a child with Down syndrome and controls are the fathers. We found plausible associations at variants at relevant loci. Stratifying by the stage of meiosis in which nondisjunction occurred (MI or MII), our results are consistent with the existence of general nondisjunction risk factors as well as some that could be specific for MI or MII.

In the second aim, we called recombination events on chromosome 21 in our data set in order to classify cases (mothers) according to their recombination profiles. We therefore developed and implemented novel methods for calling recombination events in both trios and dyads, finding that full-data trios can be used to successfully train the method for calling recombination in dyads, which contain less information.

The third aim was to further characterize the candidate gene and GWAS associations by performing stratified analyses in subgroups of mothers defined by recombination profile and maternal age. We interpret the associations in the context of possible meiotic error mechanisms.