The early SARS-CoV-2 epidemic in Senegal was driven by the local emergence of B.1.416 and the introduction of B.1.1.420 from Europe

Abstract

Molecular surveillance of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is growing in west Africa, especially in the Republic of Senegal. Here, we present a molecular epidemiology study of the early waves of SARS-CoV-2 infections in this country based on Bayesian phylogeographic approaches. Whereas the first wave in mid-2020 was characterized by a significant diversification of lineages and predominance of B.1.416, the second wave in late 2020 was composed primarily of B.1.1.420. Our results indicate that B.1.416 originated in Senegal and was exported mainly to Europe. In contrast, B.1.1.420 was introduced from Italy, gained fitness in Senegal, and then spread worldwide. Since both B.1.416 and B.1.1.420 lineages carry several positive selected mutations in the spike and nucleocapsid genes, each of which may explain their local dominance, their mutation profiles should be carefully monitored. As the pandemic continues to evolve, molecular surveillance in all regions of Africa will play a key role in stemming its spread.

Keywords: Bayesian inference of phylogeny; SARS-CoV-2; Senegal; phylogeography; positive selection analysis; variants of interest; viral surveillance.

Figure 1.

Dynamics and lineage diversity of the first two SARS-CoV-2 epidemic waves in Senegal. (A) The number of weekly cases of SARS-CoV-2 infection (purple bars) and associated deaths (black bars). The two epidemic waves are denoted. (B) changes in R_e estimations over time (purple line) and the progression of vaccination efforts (black bars). (C) Geographic representation of SARS-CoV-2 genome sampling in Senegal. (D) Frequency and distribution of the most common PANGOLIN lineages from March 2020 to February 2021. (E) The estimated time-resolved maximum clade credibility tree from 5,230 genomes containing a global representation (white dots), an African overrepresentation (blue dots), and all data available from Senegal (purple dots). The clades formed by variants of concern or interest are denoted. Data for Panels A and B are publicly available from Our World in Data (www.github.com/owid/covid-19-data/blob/master/public/data/). The column of panels containing A, B, and D are on the same time axis.

Figure 2.

Evolutionary history, demographic dynamics, and effective rate of infection of SARS-CoV-2 strains circulating in Senegal. (A) The estimated time-resolved maximum clade credibility tree from 231 SARS-CoV-2 sequences collected in Senegal (plus Wuhan reference) with tip shapes colored by PANGOLIN lineage. The topological reconstruction revealed five major clades; the tMRCA for each clade is denoted (confidence intervals are noted in Supplementary Information, Table S1). (B) Demographic history of SARS-CoV-2 strains collected in Senegal inferred via a BSP with coalescent tree prior and an exponential, uncorrelated clock model. The shading represents the 95 per cent HPD of the product of generation time and effective population size (N_eτ). The line tracks the inferred median of N_eτ. (C) The deconvolved BSP (with 95 per cent HPD shading) estimated for each of the five major clades. (D) The reconstructed Re, also deconvolved for each of the five major clades. The median R_e profile for each clade (line), computed from the posterior birth–death rates and SIR trajectories, are framed by the 95 per cent HPD interval (shading).

Figure 3.

Discrete phylogeographical evaluations of SARS-CoV-2 PANGO lineages (A) B.1.416 and (B) B.1.1.420. Both A and B follow the same format from left to right. Left panel: Markov jump trajectory plot representing the ancestral transition history for the most ancestral genome in Senegal for each lineage. The trajectories are summarized from a posterior tree distribution with Markov jump history annotation using a sampling location and travel history model. Lines in a horizontal trajectory represents the time during which a particular location state is maintained in the spatiotemporal ancestry of the virus, and vertical lines represent Markov jumps between two locations in the trajectory. The most ancestral genomes for each lineage in Senegal location are denoted. Central panel: maximum clade credibility tree from Bayesian inference, colors of the tip shapes represent the sampling locations in the legend. Right panel: circular migration flow plot representing the origin and destination of international spreading events. The width of the links indicates the frequency of viral movements as estimated using post hoc summarized posterior expectations of the Markov jump events. The direction of the migration flow starts from the outer ring to the destination location denoted by an arrowhead.

Figure 4.

Spatiotemporal dynamics of the dispersal history of SARS-CoV-2 within Senegal during (A) Wave 1 and (B) Wave 2. Continuous phylogeographic reconstructions are based on 10³ post-burn-in posterior trees. Both panels represent the maximal wavefront distance from the epidemic origin in dark blue (A) and dark red (B) lines. The 95 per cent HPD intervals are represented in light blue (A) and light red (B) bracketing the median. The map borders represent Senegal at the Département level and display the maximal spread distance of the virus at various points in time.

Figure 5.

Positively selected mutation events in Senegalese SARS-CoV-2 strains. (A) Violin plots depicting the mutation counts in each of the five Senegalese clades at the nucleotide or amino acid level in either the whole genome or spike protein alone. Plots depict the full probability density of counting a certain number of mutations in the strains; black dots indicate the mutation count median; black vertical lines depict 1 SD. (B) The maximum clade credibility tree reproduced from Fig. 2A with emphasis on lineages B.1.416 and B.1.1.420, as well as positive selection events called out at the ancestral node of each clade or lineage of interest. Gray-dotted nodes are the clade ancestors and call out positively selected sites identified by PAML’s branch-site model, while red- or pink-dotted nodes are the B.1.416 or B.1.1.420 lineage ancestors and call out positively selected sites identified by PAML’s site model. (C) The three-dimensional structure of SARS-CoV-2 spike glycoprotein complexed with the hACE2 receptor, with positively selected amino acid sites and regions of interest identified. Mutation N501Y (shown with an asterisk in panel B) is a positively selected mutation; however, it does not occur in the lineages of interest, B.1.416 or B.1.1.420.