Back to search

University of Minnesota

Multiscale Dynamics of Host-Pathogen-Environment Interactions: Integrating Epigenomics, Genomics, and Ecology in Small Mammal Systems

Abstract

dc:description.abstract

Small mammals in the orders Chiroptera and Rodentia are essential to ecosystem functioning and services while simultaneously serving as key reservoirs for zoonotic pathogens. Effectively addressing the complex dynamics of host-pathogen-environment interactions requires an integrative, multiscale approach that bridges laboratory-based molecular methods and field-based ecology within a One Health framework. This dissertation investigates these dynamics across individual, population, and ecological scales using advanced molecular and bioinformatic workflows, including Oxford Nanopore long-read sequencing, low-coverage whole-genome skimming, and transcriptomics. At the individual scale, I first investigated the role of epigenetics in prion pathogenesis. Prion diseases are fatal, transmissible, neurodegenerative disorders characterized by abnormally folded prion proteins inducing misfolding of normal prion proteins, leading to neurotoxic fibrils and plaques. Although epigenetic mechanisms such as DNA methylation are increasingly implicated in prion-like disorders (e.g., Alzheimer’s disease), their precise role in prion disease remains unclear. I used nanopore sequencing and RNAseq to measure genome-wide methylation and gene expression in the brains of Syrian hamsters (Mesocricetus auratus) experimentally infected with a hamster-adapted murine synthetic prion strain (n = 9) alongside age-matched mock-infected controls (n = 9) at 80, 120, and 160 days post-infection (dpi). I identified 1,586, 1,692, and 2,429 differentially methylated regions (DMRs) at 80, 120, and 160 dpi, respectively, with early- and mid-stage disease (80 and 120 dpi) skewing toward hypermethylation and late-stage disease (160 dpi) skewing toward hypomethylation. Gene ontology (GO) analysis of DMR-associated genes at 160 dpi highlighted neuron regulation and signaling, neurodevelopment, and cellular stress pathways. Additionally, I identified 178 differentially expressed genes (DEGs) at 80 dpi, 90 at 120 dpi, and 616 at 160 dpi; the majority of DEGs were downregulated at 80 dpi, whereas most were upregulated at 120 and 160 dpi. Overlap in DEGs across timepoints was limited, with GO terms showing upregulation of disease/injury response and cell death pathways at later timepoints. Overall, I observed a stage-specific transcriptional shift from immune suppression to widespread immune and inflammatory activation, providing key time-resolved data on methylation and transcriptional dynamics associated with impaired neuronal structure, function, and communication during prion disease progression. Next, at the population scale, I investigated the population genomic structure of Desmodus rotundus across its range. While the common vampire bat (Desmodus rotundus) is ecologically significant and acts as a key pathogen vector across the Neotropics, traditional single-gene and mitochondrial markers offer limited resolution of its complex evolutionary history. Here, I employed nanopore-based whole-genome skimming (~0.35× coverage) of 67 D. rotundus individuals across nine countries to characterize nuclear and mitochondrial genomic diversity, using approximately 43 million mapped reads and 41 million loci for genotype likelihood analysis in ANGSD, as well as assembling 28 complete mitogenomes, 61 COI, and 60 CYTB consensus genes. My analyses revealed widespread mitonuclear discordance across Neotropical populations. Genome-wide FST values ranged from 0.114 to 0.358, with nuclear genome-based principal component analysis and ancestry population structure modeling clustering populations along a continuous geographic gradient with a latitudinal transition zone in Panama. In contrast, mitogenomic phylogenies and maternal structure models resolved highly structured, isolated regional lineages, including maternal divergence across the Ecuadorian Andes and localized matrilineal sorting in Guyana. This mitonuclear discordance suggests male-mediated gene flow, identifying Panama as a critical movement corridor and the Andes as a primary geographic barrier supporting population isolation in Pleistocene refugia. Ultimately, these findings refine the evolutionary history of Desmodus and demonstrate the utility of whole-genome skimming for resolving complex wildlife lineages. Finally, at the ecological scale, I evaluated how anthropogenic disturbance influences Bartonella infection and diversity in the Neotropics. I evaluated the prevalence and genetic diversity of Bartonella spp. bacterial infection in the short-tailed fruit bat (Carollia perspicillata) across gradients of human landscape disturbance in Guyana. The sites were composed of an intensive agricultural matrix (Mahaica), a fragmented forest-agriculture mosaic (Kwebana), and an intact primary forest within the Indigenous-managed Kanashen Amerindian Protected Area (Masakenari). Bat species (n = 63) were confirmed via mitochondrial CYTB sequencing, and spleen tissues were screened for Bartonella spp. via nested PCR targeting the gltA gene, followed by Oxford Nanopore sequencing, phylogenetic reconstruction, and multivariable logistic regression. Overall Bartonella prevalence was 52.54% (31/59) and was significantly higher in the low-disturbance primary forest (66.67%, 20/30) than in the combined human-modified landscapes (37.93%, 11/29; p = 0.038). Controlling for host sex, bats in the primary forest exhibited over three times higher odds of infection (Adjusted OR = 3.28, 95% CI: 1.12 - 10.22, p = 0.034). Sequence analysis of 31 Bartonella gltA amplicons from Carollia perspicillata identified 118 polymorphic sites and 11 unique haplotypes (Hd = 0.80, π = 0.12), with the primary forest displaying higher haplotype and nucleotide diversity (h = 8, Hd = 0.83, π = 0.13) than human-modified landscapes (h = 3, Hd = 0.62, π = 0.11). An Analysis of Molecular Variance revealed no significant genetic differentiation between disturbance regimes (Φₛₜ = 0.0326, p = 0.198), with 96.7% of total variation harbored within local host populations. While two generalist Bartonella clades circulated across all sites, the primary forest also had unique Neotropical lineages, though none at any site clustered with known human pathogens. These findings challenge the paradigm that anthropogenic disturbance uniformly elevates vector-borne bacterium prevalence and demonstrate that Indigenous-managed forests serve as crucial ecological baselines for One Health and conservation research. Altogether, these studies demonstrate the value of coupling multi-omics technology with landscape ecology to resolve host-pathogen-environment dynamics across generational, evolutionary, and spatial scales, advancing knowledge critical for biodiversity conservation, public health, and disease surveillance.

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Frank, Lexi E.

Subjects

dc:subject × 6

Rights

Language dc:language.iso
en

Identifiers

dc:identifier.*
Dc Identifier Citation
Frank, Lexi E.. (2026). Multiscale Dynamics of Host-Pathogen-Environment Interactions: Integrating Epigenomics, Genomics, and Ecology in Small Mammal Systems. Retrieved from the University Digital Conservancy, https://hdl.handle.net/11299/283089.

Chain of custody

source
Harvested from
University of Minnesota
Base URL
conservancy.umn.edu/server/oai/request
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
citation

Frank, Lexi E.. Multiscale Dynamics of Host-Pathogen-Environment Interactions: Integrating Epigenomics, Genomics, and Ecology in Small Mammal Systems. 2026. https://hdl.handle.net/11299/283089