Oregon Sea Grant

Marine mammals are top predators that are essential for the health and function of our oceans. These top predators are often affected by various factors that can be detrimental to their populations. Therefore, there is a need to evaluate undetermined causes of deaths and to better understand known diseases in marine mammals to mitigate future marine mammal mortality events. In this dissertation, I examined a mortality event that occurred in 2009 and affected seven harbor seals that died from an unknown brain disease. In 2009, a cohort of harbor seals stranded along the California coastline and the necropsies of these animals showed necrosis in the cerebrum and cerebellum. However, the etiology of the disease could not be determined with conventional diagnostic procedures. The results from the necropsy reports suggested that a virus was the likely causative agent, although it was also noted that exposure to a toxin, nutrient depletion, or hypoxia could have also caused the death of these animals. To investigate the source of this stranding event I compared the brain tissues of these harbor seals, that I termed “unknown cause of death” (UCD), to the brain tissues of seven other harbor seals with known causes of death that I termed “comparative” samples. Given that UCD animals were hypothesized to have died from an unknown virus type, I used meta-​transcriptomics analysis of the brains to assess the presence of gene expression patterns from DNA/RNA viruses and opportunistic bacteria. Upon evaluation of the UCD animals, I found that there was no indication that viruses were present in the brain tissue of these animals. However, I did find the presence of a previously described Phocine herpesvirus-1 (PhV- 1) in 57% of comparative harbor seal samples. Interestingly, the microbiome analysis of the UCD animals showed two significantly abundant bacteria types, Burkholderia cepacia complex (BCC) and Coxiella burnetii. BCC was prevalent in all UCDs, which expressed a significant abundance of BCC virulence factors relative to comparative samples. Furthermore, only one UCD animal had a high abundance of C. burnetii, which represented ~94% of the microbiome community and showed a high abundance of transcripts for invasion and translation, suggesting that this was an active pathogen infection. From this evidence, I conclude that viruses were not the cause of death of UCD animals and that BCC and C. burnetii were potential opportunistic neurotropic infections in UCD animals. While our meta-​transcriptomic analysis showed interesting trends in the microbiome of UCD animals, the cause of death for UCDs remained undetermined. To further evaluate the death of these animals, I applied a transcriptomics analysis to evaluate the effects of this event on host gene expression. I compared UCD transcriptomes to only those harbor seals with a PhV-1 infection. Analysis of the host transcriptome showed that while PhV-1 infected seals had gene repertoires of animals undergoing a virus infection, UCD animals did not. UCD animals did, however, demonstrate a significantly high gene expression pattern associated with fatty acid metabolic pathways. Interestingly, fatty acid dysregulation is associated with nutrient depletion and exposure to toxins, thus I speculate that UCD animals died from these factors, and led to a high expression of fatty acid metabolism genes in the brain. Along with characterizing the transcriptome of UCD samples, I also characterized the harbor seal-virus response during a PhV-1 infection. The transcriptome results showed that all animals undergoing a PhV-1 infection were mounting both innate and adaptive immune system responses. Also, the gene expression patterns I uncovered suggests that PhV-1 hijacks the host DNA packaging and exocytosis pathways for viral replication. Lastly, to evaluate the genome and the gene expression patterns of PhV-1 during infection in the brain, I first used metagenomics to sequence and characterize the unique long (UL) and unique short (US) regions of the PhV-1 genome. I then used transcriptomics to identify the gene expression of PhV-1. In this study, we described 72 genes in PhV-1, the majority which have never been described in this virus. In all four seals, the gene expression pattern showed similar patterns of active regions across the genome, which may be important for the late phases on infection in the brain. Overall this work has applied a combination of ‘-omics’ analyses to understand an unknown disease in harbor seals and further characterized the host-virus response of a known marine mammal pathogen.