Emergence of Omicron FN.1 a descendent of BQ.1.1 in Botswana

Abstract

Botswana, like the rest of the world, has been significantly impacted by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). In December 2022, we detected a monophyletic cluster of genomes comprising a sublineage of the Omicron variant of concern (VOC) designated as B.1.1.529.5.3.1.1.1.1.1.1.74.1 (alias FN.1, clade 22E). These genomes were sourced from both epidemiologically linked and unlinked samples collected in three close locations within the district of Greater Gaborone. In this study, we assessed the worldwide prevalence of the FN.1 lineage, evaluated its mutational profile, and conducted a phylogeographic analysis to reveal its global dispersal dynamics. Among approximately 16 million publicly available SARS-CoV-2 sequences generated by 30 September 2023, only 87 were of the FN.1 lineage, including 22 from Botswana, 6 from South Africa, and 59 from the UK. The estimated time to the most recent common ancestor of the 87 FN.1 sequences was 22 October 2022 [95% highest posterior density: 2 September 2022-24 November 2022], with the earliest of the 22 Botswana sequences having been sampled on 7 December 2022. Discrete trait reconstruction of FN.1 identified Botswana as the most probable place of origin. The FN.1 lineage is derived from the BQ.1.1 lineage and carries two missense variants in the spike protein, S:K182E in NTD and S:T478R in RDB. Among the over 90 SARS-CoV-2 lineages circulating in Botswana between September 2020 and July 2023, FN.1 was most closely related to BQ.1.1.74 based on maximum likelihood phylogenetic inference, differing only by the S:K182E mutation found in FN.1. Given the early detection of numerous novel variants from Botswana and its neighbouring countries, our study underscores the necessity of continuous surveillance to monitor the emergence of potential VOCs, integrating molecular and spatial data to identify dissemination patterns enhancing preparedness efforts.

Keywords: Africa; Botswana; Omicron FN.1; SARS-CoV-2; phylodynamics.

Figure 1.

Distribution of SARS-CoV-2 FN.1, a sublineage of Omicron VOC. (a) Map of Botswana partitioned in nine COVID-19 zones, uniquely colored, (b) FN.1 sequence counts plotted on Botswana map partitioned into nine COVID-19 zones over December 2022 to February 2023. The frequency of FN.1 sequences and colored based on gradient. Botswana reported 22 sequences including Boteti (2), Greater Gaborone (n = 19), and Greater Francistown (1). (c) Changes in the genomic prevalence of Omicron lineage FN.1 overtime. The UK reported 59 sequences including England (n = 58) and Scotland (1). South Africa reported six sequences: including Mpumalanga (n = 1), Gauteng (n = 4), Free State (n = 1). (d) Proportion graph showing the landscape of SARS-CoV-2 lineages circulating between May 2022 and end of February 2023. This period correspondence to the emergence of BQ.1.1.64, BQ.1.1.74, and FN.1 lineages.

Figure 2.

Mutation profiles and characteristics of FN.1. Spike protein changes in lineage BQ.1.1 and FN.1. The mutations common in BQ.1.1 relative to SARS-CoV-2 reference strain (NC_045512) encoded protein are represented in red. (a) The lineage defining mutations for FN.1 are indicated in black. The spike mutations in NTD S:K182E (nuc:22106 G), and RDB S:T478R (nuc:22995 G) for FN.1 are indicated in bold. The ACE2 affinity and immune escape scores of FN.1 relative to Omicron (BA.1) were 0.629 and 0.834, respectively. (b) Annotation of the S:T478R mutation on the 3D spike protein. (c) The boxplot comparing the binding affinities of spike proteins with T478R mutation and without (reference strain: NC_045512). (d) Bar plot showing the impact scores of key signature mutations of BQ.1.1.64, BQ.1.1.74, and FN.1 lineages. Properties assessed include antigenicity, pathogenicity, and immunogenicity. For these analyses, the VaxiJen tool was used to assess antigenicity, the IED tool for immunogenicity, and the MutPred2 software for pathogenicity. Based on the MutPred2 software, a score >0.5 indicates an increased likelihood of pathogenicity (Pejaver et al. 2020). An absolute change >0.0102 in antigenicity (three times the median across sites) is deemed significant, as is an absolute change exceeding 0.2754 for immunogenicity. In this context, antigens with a VaxiJen prediction score >0.4 are considered candidate antigens (Doytchinova and Flower 2007). An MHC I immunogenicity score >0 suggests a higher likelihood of eliciting an immune response.

Figure 2.

(continued).

Figure 3.

(a) MCC tree of 87 FN.1 global genomes including sequences generated in this study. These were generated in Botswana, South African, UK, and Scotland. The Discrete BSSVS SkyGrid model estimated the tRMCA on 22 October 2022 with Botswana at the origin seeding two introductions to South Africa and one to the UK. Both exports have caused onward transmission chains which is important for genomic surveillance. (b) Mapping inferred viral dissemination patterns of the SARS-CoV-2 Omicron sublineage FN.1 sequences from phylogeographic reconstructions based on discrete BSSVS SkyGrid model. Overall movement of the virus from Botswana to South Africa and UK shown.

Figure 3.

(continued).

Figure 4.

A ML tree of all the 5730 sequences from Botswana and root-to-tip plot showing SARS-CoV-2 lineage that is closely related to FN.1 lineage. (a) ML tree including all 99 SARS-CoV-2 lineages among the 5370 sequences from Botswana. The overall tree was rooted by the midpoint rooting. Among all the sequences, BQ.1.1.64, BQ.1.1.38, and BQ.1.1.74 were mostly on the basal of FN.1 (shown in Supplementary Fig. S2). Of these, BQ.1.1.74 sequences represented the most significant statistical support for the FN.1 cluster (P > .90). (b) Expanded view of the FN.1 and BQ.1.1.74 sublineages extracted from the ML tree of the SARS-CoV-2 whole genome sequences characterised in Botswana. The diversity is represented on the x-axis of the branched tree. (c) The root-to-tip regression obtained from TempEst analysis for the sampled clusters of BQ.1.1.74 and FN.1 lineages, showing a relatively strong clock-like behaviour, the regression line (representing the estimated mean evolutionary rate) is shown with error buffers that represent 95% CIs.