| Disease causes large economic losses in the swine industry, not only through the cost of medical treatment, but also due to reduced production performance of sick pigs. Expanding our knowledge of the genetic basis of diseases will allow more conscientious breeding programs to alleviate some of the economic impact. The overall objective of this thesis was to identify the molecular and quantitative genetic basis of two diseases in pigs, severe combined immunodeficiency (SCID) and porcine reproductive and respiratory syndrome (PRRS). SCID is a naturally occuring primary immunodeficiency in humans, horses, dogs, and, as we discovered, also in pigs. Immunological assays determined that these pigs had low/no B or T cells, but NK cells were present, resulting in T- B- NK+ SCID. Genetic mapping of the lesion and molecular characterization revealed two independent mutations in the Artemis gene that cause SCID in these pigs. Each of these mutations were associated with a distinct haplotype of single nucleotide polymorphisms (SNPs), both of which were traced back to the founder generation of this pig population, which was sourced from commercial pig populations in the midwestern United States. This suggests that the deleterious mutations repsonsible for causing SCID in these pigs may be present in the swine industry; although, likely at a low frequency. The presence of SCID in commercial swine operations may go undetected, but a PRRS outbreak certainly will not. Thirteen trials, each with ~200 piglets from commercial breeding programs, were experimentally infected with one of two PRRS virus (PRRSV) isolates. Phenotypes analyzed were viral load (VL) in blood during the first 21 days post infection (dpi) and weight gain (WG) from 0 to 42 dpi. We utilized genotypes determined on this population using the Illumina Porcine SNP60 panel to perform genome wide association studies (GWAS) and found several genomic regions associated with each trait. None of these genomic regions were associated in both PRRSV isolate trials, except for the QTL on Sus Scrofa Chromosome (SSC) 4 that was previously identified, which is associated with VL in both isolates and with WG in the NVSL isolate. These results contradicted the previously estimated high genetic correlations of each trait between the two PRRSV isolates and lead us to believe that response to each PRRSV isolate is controlled by different loci, at least for loci with detectable associations with these traits. Gene ontology (GO) annotation information of genes in SNP-associated regions showed enrichment of immunology-related GO terms in VL-associated regions and metabolism-related GO terms in WG-associated regions. We then used these data to assess the accuracy of genomic prediction across PRRSV isolates and across breeding companies. Incorporation of annotation information enabled the identification of subsets of SNPs that could be used for genomic prediction of response to PRRSV in pigs that have not been exposed to the virus. However, predictions based on these SNP subsets were not as accurate as predictions using SNPs across the whole genome. The work described in this thesis presents opportunities for disease reduction through genetic testing for SCID carrier status and genomic selection for response to PRRSV infection. Preventing the mating of SCID carriers will reduce losses in of nursery piglets, and genomic selection of pigs that are predicted to have more desirable response to PRRSV infection would lessen the impact of this disease. |
