Phylogeography and genomic epidemiology of SARS-CoV-2 in Italy and Europe with newly characterized Italian genomes between February-June 2020

Abstract

The aims of this study were to characterize new SARS-CoV-2 genomes sampled all over Italy and to reconstruct the origin and the evolutionary dynamics in Italy and Europe between February and June 2020. The cluster analysis showed only small clusters including < 80 Italian isolates, while most of the Italian strains were intermixed in the whole tree. Pure Italian clusters were observed mainly after the lockdown and distancing measures were adopted. Lineage B and B.1 spread between late January and early February 2020, from China to Veneto and Lombardy, respectively. Lineage B.1.1 (20B) most probably evolved within Italy and spread from central to south Italian regions, and to European countries. The lineage B.1.1.1 (20D) developed most probably in other European countries entering Italy only in the second half of March and remained localized in Piedmont until June 2020. In conclusion, within the limitations of phylogeographical reconstruction, the estimated ancestral scenario suggests an important role of China and Italy in the widespread diffusion of the D614G variant in Europe in the early phase of the pandemic and more dispersed exchanges involving several European countries from the second half of March 2020.

Figure 1

Spatial distribution of lineages and clades. (a, b) Map of Italy reporting the lineage distribution (a) and the clade assignment (b) in every region.

Figure 2

SARS-CoV-2 Bayesian phylogeographic tree of 479 strains. Large red and purple circles indicate highest posterior probability ranging from 1 to 0.9. The branches are coloured based on the most probable lineage of the descendent nodes.

Figure 3

Ancestral reconstruction of SARS-CoV-2 lineages B.1 using the Italian dataset. The figure shows the compressed visualization produced by PastML using marginal posterior probability approximation (MPPA) with an F81-like model. Different colours correspond to different Italian geographical regions and lineages. Numbers inside (or next to) the circles indicate the number of strains assigned to the specific node.

Figure 4

Ancestral reconstruction of SARS-CoV-2 lineages B.1 using the European dataset. The figure shows the compressed visualization produced by PastML using marginal posterior probability approximation (MPPA) with an F81-like model. Different colours correspond to different European countries and lineages. Numbers inside (or next to) the circles indicate the number of strains assigned to the specific node. The joint ancestral scenario (Joint) and maximum a posteriori (MAP) predictions are shown for the uncertain nodes (shown as octagonal icons). CN, China; IT, Italy, EU, Europe.

Figure 5

Ancestral reconstruction of SARS-CoV-2 lineages B using the European dataset. The figure shows the compressed visualization produced by PastML using marginal posterior probability approximation (MPPA) with an F81-like model. Different colours correspond to different European countries and lineages. Numbers inside (or next to) the circles indicate the number of strains assigned to the specific node. The joint ancestral scenario (Joint) and maximum a posteriori (MAP) predictions are shown for the uncertain nodes (shown as octagonal icons). CN, China; IT, Italy, EU, Europe.