Genomic structure of the horse major histocompatibility complex class II region resolved using PacBio long-read sequencing technology

Abstract

The mammalian Major Histocompatibility Complex (MHC) region contains several gene families characterized by highly polymorphic loci with extensive nucleotide diversity, copy number variation of paralogous genes, and long repetitive sequences. This structural complexity has made it difficult to construct a reliable reference sequence of the horse MHC region. In this study, we used long-read single molecule, real-time (SMRT) sequencing technology from Pacific Biosciences (PacBio) to sequence eight Bacterial Artificial Chromosome (BAC) clones spanning the horse MHC class II region. The final assembly resulted in a 1,165,328 bp continuous gap free sequence with 35 manually curated genomic loci of which 23 were considered to be functional and 12 to be pseudogenes. In comparison to the MHC class II region in other mammals, the corresponding region in horse shows extraordinary copy number variation and different relative location and directionality of the Eqca-DRB, -DQA, -DQB and -DOB loci. This is the first long-read sequence assembly of the horse MHC class II region with rigorous manual gene annotation, and it will serve as an important resource for association studies of immune-mediated equine diseases and for evolutionary analysis of genetic diversity in this region.

Figure 1. Physical map of the horse MHC class II region.

Sequence assembly of eight BAC clones constituting a minimum-tiling path. Consecutively numbered junctions are indicated with grey shading and Arabic numerals. The BAC clone assembly forms two contigs, Contig 1 and 2, joined by Amplicons 1 and 2. Blue and red colours indicates two the homologous chromosomes 20 of the horse Bravo, Chromosome A and Chromosome B. An asterisk indicates the partially sequenced BAC clone. Gene content illustrates the location and orientation of all detected genes (blue) and pseudogenes (green) with transcriptional directionality indicated by arrows. Probe density describes the variation in the probe coverage over the MHC class II region. Probe mapping locations of the Axiom Equine Genotyping array are displayed as black lines. Broad Institute SNP collection probe locations are displayed as grey lines and Illumina beadchip positions are highlighted in black. GC% describes the percentage of G and C nucleotides in 1 kbp windows.

Figure 2. MHC class II pseudogenes.

The schematic structure of the horse MHC class II pseudogenes with the disrupting mutations indicated in red.

Figure 3. The sequence comparison of the long-read sequenced Bravo MHC class II and Twilight reference assembly EquCab2.

Genomic coordinates used in the comparisons are displayed on the corresponding axes of the dot-plot and the MAUVE alignment. The homologous blocks in the MAUVE alignment are displayed in similar colours and connected with a line and no homology regions are showed as white areas. (A) A dot-plot and MAUVE alignment illustrating the overall comparison of the two genomic sequences. The region with potential structural discrepancies is highlighted with a black dashed square on the MAUVE alignment. The SNP and INDEL density is displayed below the alignment together with the gene content. (B) Detailed illustration of the region harbouring the potential structural discrepancies. Locations of the two structural discrepancies are indicated with dashed circles in the MAUVE alignment and in the dot-plot. The assembly details are displayed on the top of the MAUVE alignment and UCSC genome browser repetitive element annotation for EquCab2 is displayed below.

Figure 4. Comparison of MHC class II structure in mammals.

The schematic (not to scale) gene order comparison of the horse MHC class II region to house mouse and human, illustrated above the horse, and to cattle, pig, domestic cat and dog, illustrated below the horse. Genes coding for DR, DQ, DOB and DP molecules are coloured in blue, green, orange and purple, respectively. The highly conserved genes are coloured in black. Gene Eqca-G was not included in this comparison, because it does not belong to MHC class II region. For better overview, the cattle MHC class IIb sequence was inverted. The nomenclature of the classical MHC class II genes was adjusted based on the MHC Nomenclature report and the allele nomenclature in MHC-IPD database. Note that the nomenclature does not reflect orthology.