Fully Phased Population-Prevalent East African Cattle BoLA-I Alleles Determined Using PacBio HiFi Long-Read Sequencing Represent Five Novel Specificities With Distinctive Peptide Binding Potential

Abstract

Due to factors such as lower biosecurity, greater wildlife/farm animal interfaces, and environmental challenges, cattle in sub-Saharan Africa are exposed to more diverse and intensive bacterial, viral and protozoan pathogen challenges than cattle in Europe and other high-income regions of the world. Classical class I genes of the major histocompatibility complex (MHC) contribute to protection from diseases caused by these pathogens by refining a huge pool of potential pathogen-derived peptide ligands into a smaller ensemble for presentation to CD8+ T cells. Knowledge of population-prevalent MHC alleles is therefore critical for evidence-based approaches to vaccine design and improved understanding of pathogen resistance. Whereas variation in MHC molecules is understood in most detail for European Bos taurus, the alleles expressed by Africa's cattle remain poorly defined. We have leveraged recent improvements in the accuracy of PacBio high-fidelity (HiFi) circular consensus sequencing (CCS) and adapted stringent sequence filtering algorithms to identify hundreds of as yet uncharacterised fully phased BoLA-I alleles from multiple populations of African taurine (Ankole) and indicine (Zebu) cattle in East Africa. The analysis highlights a convergence of population-prevalent class I MHC allelic repertoires in taurine and indicine cattle, likely due to the similar pathogen-driven selective pressures. Our analysis of the anchor residue accommodating pockets of these prevalent alleles revealed extremely high levels of polymorphism, which contrast with Holstein alleles that exhibit a more limited repertoire of MHC specificity-determining pocket residues, potentially constraining the breadth of peptide presentation. However, in the context of considerable sequence and physicochemical variation in the pocket-forming residues, it was possible to discern overlaps in the predicted peptide binding spectrum. Interrogation of potential differences in peptide binding specificities with European B. taurus alleles revealed that the fully phased African cattle class I MHC alleles represent five novel specificities. We envisage that this novel finding will find broad application in assessing potentially achievable vaccination coverages of future pathogen-encoded vaccine candidates against important intracellular pathogens. One aim of future research should be to leverage recent improvements in the sensitivity of mass spectrometry combined with immunoprecipitation of peptides bound to African cattle MHC to search directly for T-cell epitopes in the context of the inferred 'supertype' diversity.

Keywords: MHC; cattle; high‐throughput sequencing.

FIGURE 1

Map of Kenya, Uganda and Tanzania showing the location of sampling sites.

FIGURE 2

Box plot displaying the distribution of retained CCS reads, grouped by breed, following a stepwise filtering criterion. ‘Trimmed’ refers to CCS reads where both primers were detected without any mismatches and trimmed. ‘Filtered’ represent retained CCS reads upon filtering based on the DADA2 algorithm using minQ (min quality base) 3 and maxEE (expected errors for the complete trimmed read) = 1 and 800 < = length < = 1200. CCS reads that could be assigned to the inferred amplicon sequence variants are marked as ‘Denoised’, while ‘nochim’ denotes reads retained following application of the DADA2 chimera detection and removal algorithm. 76%, 40%, 29% and 24% of reads were retained after the trimming, filtering, denoising and chimera removal steps, respectively.

FIGURE 3

A site graph of the inferred rates of synonymous (orange) and non‐synonymous (blue) substitutions for African cattle BoLA‐I alleles. In this figure, the defined BoLA‐I domains include alpha (α) 1 (residues at sites 16–106), α2 (107–198), α3 (199–290), transmembrane (291–324) and an intra‐cytoplasmic (325–342).

FIGURE 4

Population‐prevalent African cattle class I MHC alleles grouped by breed. The authors' provisional allelic designations and their accession numbers in parentheses for the 87 population‐prevalent alleles are shown on the y‐axis. The occurrence of each of these alleles, grouped by breed, is shown on the x‐axis. The number of filter‐pass reads supporting each allele is shown to the right.

FIGURE 5

The distribution of the percentage pairwise identities among pocket forming residues of African cattle BoLA‐I alleles compared to Holstein allelic sequences present in the IPD. White circles show the medians; box limits indicate the 25th and 75th percentiles. Whiskers extend 1.5 times the interquartile range from the 25th and 75th percentiles. Polygons represent densi‐ty estimates of the data and extend to extreme values. The violin plot was generated using the R software vioplot package.

FIGURE 6

The physicochemical properties of the range of amino acids that line the primary and secondary anchor residue‐accommodating pockets of the population‐prevalent class I MHC alleles in African cattle (panel A) compared to European (IPD‐MHC) cattle alleles (panel B).

FIGURE 7

Class I MHC distance tree comparing the predicted peptide‐binding specificities of African (black branch labels) and European (red branch labels) cattle alleles. Branch support and consensus tree calculations were based on 1000 bootstrap replicates. Alleles with similar predicted peptide binding specificities branch together in groups or clusters (supertypes) and the closer two alleles branch, the larger the overlap between their predicted peptide‐binding repertoires. The novel specificity clusters are circled. The African cattle BoLA‐I alleles representing the five predicted novel peptide binding specificities are circled in blue. Representative logos illustrating the predicted peptide‐binding motifs for the five novel specificities are shown in Figure 8.

FIGURE 8

Representative logos illustrating the predicted peptide‐binding motifs for the five novel specificities (panel A) and the typical peptide binding motif for each of the other supertypes represented by European cattle BoLA‐I alleles (panel B). The height of each stack of symbols (y‐axis) represents information content (the level of amino acid conservation) in each position; the relative frequency of a particular amino acid at that position is represented by the individual height of the amino acid symbol and underrepresented amino acids.