Graduation Year

2022

Document Type

Thesis

Degree

M.S.P.H.

Degree Name

MS in Public Health (M.S.P.H.)

Degree Granting Department

Public Health

Major Professor

Ryan McMinds, Ph.D.

Co-Major Professor

Mark Margres, Ph.D.

Committee Member

Lynn Martin, Ph.D.

Keywords

cancer, genetic population structure, genotype-phenotype, GWAS, joint modeling

Abstract

Coevolution is a driving force of rapid evolution, yet the complexity of coevolutionary interactions has made it difficult to characterize the genomic basis of traits mediating such relationships. Coevolutionary dynamics are especially important in host-pathogen systems where the host and pathogen must constantly adapt to one another. The Tasmanian devil and its species-specific transmissible cancer, devil facial tumor disease (DFTD), provide the rare opportunity to study host-pathogen coevolution in a complex natural system. Extensive spatiotemporal devil sampling, high linkage disequilibrium in devils, and a large selective pressure imposed by DFTD facilitate a system tractable for study. Here, we characterized devil and DFTD coevolution by looking at genetic population structures, genome architecture underlying force of infection and virulence, and the contribution of devil-DFTD genome interaction to explaining force of infection. A probe-capture sequence approach was used to sequence 456 devils and 504 tumors at ~197k loci. The genetic structures of devils and DFTD were then identified via clustering and compared. Associative modeling was used to determine genome architecture, and a joint host-pathogen model was used to assess the contribution of genome interactions. Devil and DFTD genetic clustering revealed a decoupled genetic structure, suggesting little evidence of coevolution. Variance in force of infection was attributable primarily to devil genomes (61.1%) and had both large-effect variants (~3 SNPs explained 22.8% total variance) and a polygenic component. Tumor genomes explained a large proportion of virulence (69.8%) and a few large-effect loci contributed to most of this explanatory power (~6 SNPs explained 51.2% total variance). Significant devil-DFTD genomic interactions for force of infection were detectable through joint modeling (40.3%), and genotype-by-genotype interaction tests revealed devil and tumor genes implicated in cancer. Despite the decoupled genetic structure between devils and DFTD, the significant genome interaction indicates potential coevolution. The identified devil-DFTD genome interaction represents the first finding providing evidence of coevolution between devils and DFTD, and the framework used here may be applied to various host-pathogen systems.

Download

Share

COinS