Transcriptomic analysis of sheep macrophages and their response to Lipopolysaccharide
McCulloch, Mary Elizabeth Blanche
Sheep are ruminant animals, highly adapted to exploit pastures of low biological value and an economically important livestock species. They also provide a useful animal model for multiple areas of human medicine. Their productivity is compromised by many viral, bacterial, fungal, protozoan and helminthic pathogens. The innate immune system evolves under stringent selective pressure from pathogens. In this project, I aimed to identify genes involved in innate immunity in sheep by dissecting the transcriptome of sheep macrophages. The analysis focussed on identification of genes that may be associated with either disease susceptibility or resistance traits. The project formed part of the broader transcriptomic atlas for sheep. A method was established and validated for the production of sheep bone marrow derived macrophages (BMDMs) grown in macrophage colony‐stimulating factor (CSF1). These cells responded to bacterial Lipopolysaccharide (LPS), an archetypal agonist of the pattern recognition receptor, TLR4, with induction of inflammatory cytokines, but unlike rodent macrophages, sheep produced no nitric oxide. Bone marrow‐derived macrophages were produced from male (n=3) and female (n=3) Scottish Blackface X Texel animals used in The Sheep Atlas project, and mRNA was isolated from the cells at 0, 2, 4, 7 and 24 hours following stimulation. Two different protocols of mRNA preparation were used. For 0 and 7 hour samples, samples were depleted of rRNA, and RNA‐Seq was carried out at a depth of 100 million reads. For all samples in the time course (including 0 and 7), mRNA was prepared by polyA selection, and RNA‐Seq was performed at lower depth (25 million reads) and a detailed analysis of the different outcomes is presented. Two pipelines, Kallisto and StringTie were used to identify and quantify transcripts in the LPS time course transcriptomic data, along with other subsets of innate immune cells from the wider atlas. The former pipeline provides quantitative known transcript expression estimates, the latter generates novel transcript and gene models. Analysing the transcriptional signatures of these samples provided insight into the metabolic, regulatory and innate response pathways that sheep share with other animals, with both Myeloid differentiation primary response 88 protein (MYD88) dependent and independent pathways being activated following LPS stimulation, with hundreds of the same downstream cytokines being highly expressed in response, such as TNF and many interleukins as seen in other species innate responses. It has also highlighted aspects of the response that separate sheep from other animals, such as their metabolism and biosynthesis of steroid and tryptophan as well as demonstrating differences in specific expression of certain genes. Two highly regulated and expressed genes noted during the peak response at 7hrs (ENSOARG00000005159 and ENSOARG00000006889), were both assigned a functional annotation as being the protease inhibitor Serpin family B member 2 gene (SERPINB2). This appears to be a gene duplication. This sheep expression profile of SERPINB2 is shared with mice but distinct from humans and pigs. Similar to other animals, individual animals were found to vary markedly in their transcriptional response to LPS, demonstrating hyper, early and delayed responses between the individuals. The current reference transcriptome OarV3.1 contains 28757 transcripts, of which only 18488 are functionally annotated. More than 85% of all reference transcripts were detected in sheep macrophages and over 300 candidate annotations for genes identified only by Ensembl sheep (Ovis aries) gene ID (ENSOARG) numbers were manually assigned a functional annotation by clustering co‐expression estimates using the network analysis tool, Miru and inferring function by ‘guilt by association’ of these unknown genes which demonstrated similar expression profiles to those with known function. The nearest orthologues and synteny with other species were used to validate these suspected annotations. A select few examples include PYD and CARD domain containing gene (PYCARD), all the early growth response factors and many of the C‐X‐C Motif chemokines. In summary, this project identified multiple sheep‐specific aspects of the innate immune response, whilst assigning hundreds of genes a functional annotation and detecting new transcript models for multiple known genes. Many regulated, highly expressed, multi exoned, novel gene models have also been identified which are worthy of future investigation.