| dc.description.abstract | Cattle are economically important animals, providing a valuable food source for
people around the world. There are two main subspecies of cattle, Bos taurus taurus
(taurine) and Bos taurus indicus (indicine), which began to diverge from each other
over 210,000 years ago. Today, there are hundreds of established breeds that each
display unique phenotypes. Understanding the variation between cattle breeds that
are adapted to different local environments is essential for breeding animals with
improved climate and disease resistance. Nevertheless, bovine research is
dominated by European taurine breeds. In addition, studies of cattle diversity have
concentrated on genetic variation, with alternative mechanisms largely overlooked.
Variation at the epigenetic level is of particular interest because of its role in the
development and function of the immune system. However, the tools and reference
resources to study this variation in cattle are almost entirely lacking. This research
focussed on building and interrogating genome-wide maps of chromatin accessibility
and DNA methylation for seven immune cell types across three genetically distinct
cattle breeds. Holstein Friesian, N’Dama and Nelore cattle were selected to represent
the European taurine, African taurine and indicine cattle lineages, respectively. Gene
expression data for Holstein Friesian was also analysed.
Chapter 2 involved the development, validation, and selection of methods to best
study chromatin accessibility and DNA methylation in cattle. Based on its performance
using bovine monocytes, a modified version of the original assay for transposase-accessible chromatin using sequencing (ATAC-seq) protocol was chosen to profile
regions of open chromatin. Reduced representation bisulfite sequencing (RRBS) and
whole genome bisulfite sequencing were compared at similar sequencing costs, with
the former selected due to its better coverage of likely functional regions.
Chapter 3 used RRBS to generate DNA methylation reference profiles for seven
purified immune cell populations (monocytes, neutrophils, B cells, CD4 T cells, CD8
T cells, γδ T cells and NK cells) from the three cattle breeds. RNA-seq was also used
to profile the transcriptome of corresponding purified cell types for Holstein Friesian
cattle. These reference panels were demonstrated to be able to successfully estimate
cell type abundances in mixed cell populations using a computational tool called
CIBERSORTX. The cell type proportions estimated by CIBERSORTX, using gene
expression data, were highly correlated to the known compositions of ten in silico
cellular mixtures. Cell type proportions were also accurately imputed for six in vitro
and nine lysed blood mixtures using RRBS data, often even when the reference
panels and mixtures were derived from different breeds.
Finally, chapter 4 examined the chromatin accessibility, DNA methylation and
transcriptional maps generated for each cell type and breed. Extensive epigenetic
divergence was found between the taurine and indicine cattle lineages across cell
types, with most variation at regions distal to the transcription start site. Furthermore,
distinct sub-categories of CpG islands based on their chromatin and methylation
profiles were identified that discriminate between classes of distal and gene proximal
islands linked to discrete transcriptional states.
The data generated in this research provide a comprehensive resource to help exploit
the diversity among cattle breeds and improve cattle productivity for farming
communities in low and middle income countries. | en |
