Complete Genome Sequencing and Comparative Phylogenomics of Nine African Swine Fever Virus (ASFV) Isolates of the Virulent East African p72 Genotype IX without Viral Sequence Enrichment

Abstract

African swine fever virus (ASFV) is endemic to African wild pigs (Phacochoerus and Potamochoerus), in which viral infection is asymptomatic, and Ornithodoros soft ticks. However, ASFV causes a lethal disease in Eurasian domestic pigs (Sus scrofa). While Sub-Saharan Africa is believed to be the original home of ASFV, publicly available whole-genome ASFV sequences show a strong bias towards p72 Genotypes I and II, which are responsible for domestic pig pandemics outside Africa. To reduce this bias, we hereby describe nine novel East African complete genomes in p72 Genotype IX and present the phylogenetic analysis of all 16 available Genotype IX genomes compared with other ASFV p72 clades. We also document genome-level differences between one specific novel Genotype IX genome sequence (KE/2013/Busia.3) and a wild boar cell-passaged derivative. The Genotype IX genomes clustered with the five available Genotype X genomes. By contrast, Genotype IX and X genomes were strongly phylogenetically differentiated from all other ASFV genomes. The p72 gene region, on which the p72-based virus detection primers are derived, contains consistent SNPs in Genotype IX, potentially resulting in reduced sensitivity of detection. In addition to the abovementioned cell-adapted variant, eight novel ASFV Genotype IX genomes were determined: five from viruses passaged once in primary porcine peripheral blood monocytes and three generated from DNA isolated directly from field-sampled kidney tissues. Based on this methodological simplification, genome sequencing of ASFV field isolates should become increasingly routine and result in a rapid expansion of knowledge pertaining to the diversity of African ASFV at the whole-genome level.

Keywords: ASFV; African swine fever virus; East Africa; field isolates; p72 Genotype IX; phylogenomics; rapid whole-genome sequencing.

Figure 1

Geographical map of the sampling sites in Kenya and Uganda.

Figure 2

Maximum likelihood phylogenetic tree of ASFV isolates based on a whole-genome multiple alignment. Colored clades are highlighted according to: (i) historical p72 genotypes based on the 3′ end of the B646L ORF (Roman numerals); (ii) genotype groups as described in [42], derived from full-length p72 protein sequences (parenthesized numbers); (iii) biotypes as described in [43], from a full-proteome, ML-based analysis (Arabic numerals). The branch length values (scale at the bottom right of the figure) represent the mathematical expectation of the number of nucleotide substitution events in the pairwise alignment between the two sequences (ancestral or extant) present at the tips of any given branch. Branch length values lower than 0.5 × 10⁻³ are not displayed. This figure was prepared with iTOL (https://itol.embl.de/) before post-processing in Inkscape.

Figure 3

Heatmap of whole-genome sequence similarity across all available Genotype IX genomes.

Figure 4

The 3′ end of the coding sequences for B646L/p72 in a multiple sequence alignment including all published full-length genomes of Genotypes IX and X, among which are the nine Genotype IX genomes from this study. The 19 bp region framed with dashes corresponds to the complementary region for p72D. It harbors the two SNPs described in the text, the first of which (at Site 1924) is common to all viruses from Genotypes IX and X. The topmost, coloured sequence is a genotype II sequence chosen as a random reference so that the SNPs below are then highlithed with respect to that sequence. This image was prepared with UGENE [24].