Lumpy Skin Disease Virus Genome Sequence Analysis: Putative Spatio-Temporal Epidemiology, Single Gene versus Whole Genome Phylogeny and Genomic Evolution

Abstract

Lumpy Skin Disease virus is a poxvirus from the genus Capripox that mainly affects bovines and it causes severe economic losses to livestock holders. The Lumpy Skin Disease virus is currently dispersing in Asia, but little is known about detailed phylogenetic relations between the strains and genome evolution. We reconstructed a whole-genome-sequence (WGS)-based phylogeny and compared it with single-gene-based phylogenies. To study population and spatiotemporal patterns in greater detail, we reconstructed networks. We determined that there are strains from multiple clades within the previously defined cluster 1.2 that correspond with recorded outbreaks across Eurasia and South Asia (Indian subcontinent), while strains from cluster 2.5 spread in Southeast Asia. We concluded that using only a single gene (cheap, fast and easy to routinely use) for sequencing lacks phylogenetic and spatiotemporal resolution and we recommend to create at least one WGS whenever possible. We also found that there are three gene regions, highly variable, across the genome of LSDV. These gene regions are located in the 5' and 3' flanking regions of the LSDV genome and they encode genes that are involved in immune evasion strategies of the virus. These may provide a starting point to further investigate the evolution of the virus.

Keywords: evolution; genome; lumpy skin disease; phylogeny.

Figure 1

(A,B) LSDV WGS-based phylogeny for the <2017 and the >2017 dataset. Nodes (clades and subclades) supported by significant values (SH/pp/BS-s >75/>97/75) are marked with *. The scale bare indicates substitutions per site.

Figure 2

(A) LSDV distribution and dispersal across the world based on WGS. Major strain clades are indicated: 1.1 (blue), 1.2 (red) and 2.5 (purple) across the world. The splash symbol indicates the first occurrence of the recombinant strain. The inverted triangles indicate vaccine sequences and the respective ancestral sequences from which they were derived. The thin arrows labelled with * indicate relationships between accession supported by the WGS–based phylogenies. The circles/ovals indicate clades. The large arrows indicate the subsequent direction of dispersal for which the exact trajectory is unknown or not yet included in this paper. The question mark indicates uncertainty about the exact origin which led to a new outbreak. Clade labels are added in the white outlined boxes. (B,C) Networks of the sequence-based datasets for Clades 1.2 and 2.5 decomposed into splits. Here, the 2.1 and 2.5 Subclades are indicated with different colors for different clusters.

Figure 3

Consensus network of 10,000 trees obtained from bootstrapping for the >2017 dataset. The network is based on a count of Edge–Weights with the threshold set at 10% for displaying conflicts. The numbers on the branches indicate the number of variable positions within the group.

Figure 4

Single gene trees for (A) GPCR; (B) RNA-dependent RNA polymerase subunit 30 (RPO30); (C) viral DNA polymerase (DNA_Pol); (D) P32 gene and (E) RNA-dependent RNA polymerase subunit 132 (RPO132 Nodes (clades and subclades) supported by significant values (SH/pp/BS-s >75/>97/75) are marked with *. The scale bare indicates substitutions per site.

Figure 4

Single gene trees for (A) GPCR; (B) RNA-dependent RNA polymerase subunit 30 (RPO30); (C) viral DNA polymerase (DNA_Pol); (D) P32 gene and (E) RNA-dependent RNA polymerase subunit 132 (RPO132 Nodes (clades and subclades) supported by significant values (SH/pp/BS-s >75/>97/75) are marked with *. The scale bare indicates substitutions per site.

Figure 5

Combined π-plot of wild-type LSDV WGS (red) and recombinant LSDV WGS (blue). The genomic regions displaying peaks with >average π- + 3 times standard deviation are indicated in the grey labels. Labels with * are only affected in wild-type strains. Identification is based on the annotation of the sequence belonging to the LSDV reference strain NC_003027. The 10 * average π-value (based on the >2017 dataset) is indicated by a solid line. The average π-value + 3 times standard deviation is indicated in the figure by a dashed line. The x-axis denotes the genomic position; the y-axis denotes the π-values.