Pathogen-driven selection shaped adaptive mutations in immunity genes, including those contributing to inflammatory disorders. Functional characterization of such adaptive variants can shed light on disease biology and past adaptations. This popular idea, however, was difficult to test due to challenges in pinpointing adaptive mutations in selection footprints. In this study, using a local-tree-based approach, we show that 28% of risk loci (153/535) in 21 inflammatory disorders bear footprints of moderate and weak selection, and part of them are population specific. Weak selection footprints allow partial fine-mapping, and we show that in 19% (29/153) of the risk loci under selection, candidate disease variants are hitchhikers, and only in 39% of cases they are likely selection targets. We predict function for a subset of these selected SNPs and highlight examples of antagonistic pleiotropy. We conclude by offering disease variants under selection that can be tested functionally using infectious agents and other stressors to decipher the poorly understood link between environmental stressors and genetic risk in inflammatory conditions.
The extant reptiles are one of the most diverse clades among terrestrial vertebrates and one of a few groups with instances of parthenogenesis. Due to the hybrid origin of parthenogenetic species, reference genomes of the parental species as well as of the parthenogenetic progeny are indispensable to explore the genetic foundations of parthenogenetic reproduction. Here, we report on the first genome assembly of rock lizard Darevskia valentini, a paternal species for several parthenogenetic lineages. The novel genome was used in the reconstruction of the comprehensive phylogeny of Squamata inferred independently from 7369 trees of single-copy orthologs and a supermatrix of 378 conserved proteins. We also investigated Hox clusters, the loci that are often regarded as playing an important role in the speciation of animal groups with drastically diverse morphology. We demonstrated that Hox clusters of D. valentini are invaded with transposons and contain the HoxC1 gene that has been considered to be lost in the amniote ancestor. This study provides confirmation for previous works and releases new genomic data that will contribute to future discoveries on the mechanisms of parthenogenesis as well as support comparative studies among reptiles.
Darevskia rock lizards include 29 sexual and seven parthenogenetic species of hybrid origin distributed in the Caucasus. All seven parthenogenetic species of the genus Darevskia were formed as a result of interspecific hybridization of only four sexual species. It remains unknown what are the main advantages of interspecific hybridization along with switching on parthenogenetic reproduction in evolution of reptiles. Data on whole transcriptome sequencing of parthenogens and their parental ancestors can provide value impact in solving this problem. Here we have sequenced ovary tissue transcriptomes from unisexual parthenogenetic lizard D. unisexualis and its parental bisexual ancestors to facilitate the subsequent annotation and to obtain the collinear characteristics for comparison with other lizard species. Here we report generated RNAseq data from total mRNA of ovary tissues of D. unisexualis, D. valentini and D. raddei with 58932755, 51634041 and 62788216 reads. Obtained RNA reads were assembled by Trinity assembler and 95141, 62123, 61836 contigs were identified with N50 values of 2409, 2801 and 2827 respectively. For further analysis top Gene Ontology terms were annotated for all species and transcript number was calculated. The raw data were deposited in the NCBI SRA database (BioProject PRJNA773939). The assemblies are available in Mendeley Data and can be accessed via doi:10.17632/rtd8cx7zc3.1.
Objectives: This study is performed in the frame of a bigger study dedicated to genomics and transcriptomics of parthenogenesis in vertebrates. Among vertebrates, obligate parthenogenesis was first described in the lizards of the genus Darevskia. In this genus, all found parthenogenetic species originated via interspecific hybridization. It remains unknown which genetic or genomic factors play a key role in the generation of parthenogenetic organisms. Comparative genomic and transcriptomic analysis of parthenogens and their parental species may elucidate this problem. Darevskia valentini is a paternal species for four (of seven) parthenogens of this genus, which we promote as a particularly important species for the generation of parthenogenetic forms. Data description: Total cellular RNA was isolated from kidney and liver tissues using the standard Trizol Tissue RNA Extraction protocol. Sequencing of transcriptome libraries prepared by random fragmentation of cDNA samples was performed on an Illumina HiSeq2500. Obtained raw sequences contained 117,6 million reads with the GC content of 47%. After preprocessing, raw data was assembled by Trinity and produced 491,482 contigs.
Most causal variants for complex diseases are expected to affect gene regulation in a cell- and context-specific manner. Hence, identification of such dynamically functioning variants requires functional readouts in disease-relevant tissues and context. In this study, we prioritized causal variants for psoriasis by adding functional annotations from disease-relevant cells. We demonstrate that disease-relevant immune cells, unlike most other tissues, possess functional annotations that match candidate causal SNPs. Specifically, we identified an eQTL, rs4672505, that reduces B3GNT2 gene expression only in Th1/Th17 cells with a memory phenotype, i.e., antigen-experienced T helper cells. This eQTL, a likely causal variant, also matched an enhancer chromatin mark exclusive to memory T helper cells and absent in other tissues. A disease-risk allele A at the eQTL correlates with reduced expression of the B3GNT2 glycosyltransferase. B3GNT2 deficiency in murine models reduces the glycosylation of the CD28 co-receptor involved in the CD28/B7 co-stimulation pathway and results in increased T cell activation upon antigen stimulation. We hypothesize that the risk allele A in patients increases the activation of memory Th1/Th17 cells upon re-exposure to antigens, which constitutes “signal 1”. Increased reactivity to antigens depends on “signal 2” via CD28/B7 co-stimulation from antigen-presenting cells that need to encounter microbial products. Hence, this genetic risk mechanism lies at the nexus of the response to specific antigens and microbial exposure, for instance, infection or vaccination, both of which are known to exacerbate psoriasis.
Most causal variants for complex diseases are expected to affect gene regulation in a cell- and context-specific manner. Hence, identification of such dynamically functioning variants requires functional readouts in disease-relevant tissues and context. In this study, we prioritized causal variants for psoriasis by enhancing functional annotations from disease-relevant cells. We demonstrate that disease-relevant immune cells, unlike most body tissues, possess functional annotations matching candidate causal SNPs. Specifically, we identified an eQTL rs4672505 affecting B3GNT2 gene expression only in Th1/Th17 cells with a memory phenotype, i.e., antigen-experienced cells. This eQTL also matched an enhancer chromatin mark exclusive to memory T cells and absent in other tissues. For this eQTL rs4672505, the risk allele A reduces B3GNT2 gene expression. B3GNT2 deficiency in murine models reduces the glycosylation of the CD28 co-receptor and increases T cell activation upon antigen stimulation. Notably, memory T cells also require CD28 co-stimulation, which occurs only when antigen-presenting cells encounter an infection. Thus, our findings suggest that the risk allele likely increases co- stimulation of memory Th1/Th17 cells by their cognate antigens under infection. This putative mechanism aligns with the observation that psoriasis can manifest, exacerbate, and recur after infection.