Emergence and spread of two SARS-CoV-2 variants of interest in Nigeria

Nigeria, the most populous African country, reported its first SARS-CoV-2 case on February 27, 2020. Despite over 265,000 confirmed cases by mid-September 2022, incidence and mortality per capita were far lower than in many countries. Explanations include underascertainment, demographics, mobility patterns, host genetics, environmental factors, and potential pre-existing immunity. Prior to the dominance of globally sweeping variants (Delta, Omicron), the role of regionally circulating genetic diversity in Africa in shaping epidemiology was unclear. This study aims to characterize SARS-CoV-2 genomic epidemiology in Nigeria across the first three epidemic waves (March 2020–September 2021), focusing on the emergence, timing, origins, and transmission dynamics of variants of interest B.1.525 (Eta) and B.1.1.318, and to quantify Nigeria’s bidirectional transmission dynamics within Africa and internationally using integrated genomic, mobility, and epidemiological data.

The study situates Nigeria’s epidemic within broader African SARS-CoV-2 dynamics, referencing heterogeneous incidence, underascertainment, and factors like age structure and mobility. It cites prior African genomic surveillance and lineage studies (e.g., A.23.1 in Uganda; Rwanda’s traveler-linked lineage diversity; continent-wide surveillance trends), and global analyses of variants (Alpha, Delta, Omicron) and spike mutations (e.g., E484K associated with immune escape). Reports linking B.1.525 to Nigeria and B.1.1.318 outbreaks (e.g., Mauritius) are noted, as well as methods papers for phylogenetics, phylogeography, and mobility-informed import/export risk estimation (Du Plessis et al.). This literature underscores the need for integrated genomic and mobility analyses to resolve introductions, exports, and regional connectivity effects in under-sampled settings.

Study design and sampling: Nationwide genomic surveillance approved by the Nigeria NHREC (protocol NHREC/01/01/20017-08/08/2020; approval NHREC/01/01/2007-30/11/2021B). Samples were collected from community testing centers (including travelers) and hospitalized individuals from February 2020 to October 2021 across Nigeria (25 of 36 states plus FCT). Sampling fraction was computed as assembled genomes per confirmed cases per day. Sample processing and RT-qPCR: RNA extracted from NP swabs or saliva/sputum using QiAmp Viral RNA Mini Kit (Qiagen) and MagMax kit (ThermoFisher) on Kingfisher Flex. RT-qPCR targeting N, ORF1ab, and RdRP genes using commercial kits (genesig, DaAn-Gene, Liferiver, Genefinder, Sansure). Samples with Ct<30 were selected for sequencing. Sequencing: Illumina COVIDseq protocol with ARTIC v3 400 bp tiling amplicons. Libraries prepared with Nextera DNA Flex and sequenced on Illumina MiSeq, NextSeq 2000, and NovaSeq 6000 at ACEGID. Genome assembly and quality control: viral-ngs v2.1.19 used for demultiplexing (demux_plus), QC, and reference-based assembly against NC_045512.2 to generate coverage plots and consensus FASTA. 1,577 genomes with length ≥20,930 nt (≥70% of reference) were retained. Lineage assignment and visualization: PANGOLIN v3.1.12 for lineage calls; Nextclade v1.3.0 for clade assignment and mutation assessment. Lineage dynamics visualized with Microreact and Nigeria GeoJSON mapping. Phylogenetic analyses: Sequences aligned using MAFFT v7.490. Maximum-likelihood trees built with IQ-TREE v2.1.2 under GTR and 1000 bootstraps. Time-aware analysis and ancestral state reconstruction via TreeTime v0.92 within Nextstrain v3.0.3. Bayesian phylogenetics and phylogeography: All available B.1.1.318 (n=3858) and B.1.525 (n=8278) from GISAID (as of 18 Aug 2021) were filtered (≤5% Ns, ≥95% length, complete dates), aligned with minimap2 with UTRs and problematic sites masked, and downsampled using a phylogenetically informed scheme to 1118 (B.1.1.318) and 1746 (B.1.525) genomes (all Nigerian sequences retained: 73 for B.1.1.318; 256 for B.1.525). BEAST v1.10.5 used with HKY+Γ, relaxed log-normal molecular clock, exponential growth coalescent prior; two independent MCMC chains of 200M states each (sampling every 20,000; 20% burn-in) combined, with ESS>200 verified in Tracer v1.7. Discrete trait phylogeography at regional level with asymmetric rates and BSSVS; Markov jump counts extracted to estimate timing and origins of transitions into/out of Nigeria. MCC trees via TreeAnnotator; visualization with baltic. Mobility and travel data: International air travel volumes (monthly, by origin/destination country) from IATA via BlueDot for May 2020–April 2021; domestic human mobility from Google Mobility (Jan 2020–Nov 2021) across six categories (retail & recreation, grocery & pharmacy, parks, transit stations, workplaces, residential). Import/export intensity indices: Introduction Intensity Index (III) and Exportation Intensity Index (EII) computed per Du Plessis et al., combining estimated asymptomatic infections (back-extrapolated from deaths via outbreak.info data) with monthly air travel volumes (assumed uniformly distributed across days) for May 2020–April 2021. Land travel data were unavailable, potentially underestimating regional connectivity impacts.

- Sequencing and epidemiology: 1,577 SARS-CoV-2 genomes generated from March 2020 to September 2021 across 25 states and FCT, spanning three epidemic waves. Over 35 PANGO lineages detected. Lockdown measures during the first wave (ending August 2020) led to ~50% declines in workplace/retail/transit activity and ~25% increases in residential activity. - Emergence and introductions of B.1.1.318: First detected in Lagos in December 2020. Bayesian phylogeography indicates emergence in Africa in early August 2020 (mean tMRCA 5 Aug 2020; 95% HPD 25 Jun–20 Sep). At least 53 independent introductions into Nigeria (mean; 95% HPD 50–59) began in November 2020 after lifting travel restrictions, with most introductions from other African nations. - Emergence and exports of B.1.525 (Eta): Strongly supported emergence in Nigeria (root state posterior support = 0.998) in late July 2020 (mean tMRCA 23 Jul 2020; 95% HPD 7 Jun–4 Sep). Estimated lower bound of 295 exports from Nigeria (95% HPD 259–335), predominantly to Europe, followed by Africa and North America between December 2020 and March 2021; at least 20 reintroductions from Europe (95% HPD 6–37). - Air travel and risk indices: During travel restrictions (May–Oct 2020), inbound/outbound volumes were low and dominated by African routes; introduction/export risks were negligible due to low incidence and restricted travel. After reopening (October 2020), air travel surged (~30-fold increase in inbound passengers by December 2020), with increased flows from Europe, the Middle East, and North America. III peaked in Dec 2020–Jan 2021, driven by high incidence in the UK and USA; however, III likely underestimated regional (African) risk due to absence of land travel data. EII peaked during Nigeria’s second wave (Dec 2020–Feb 2021), highest toward Europe (notably the UK) and North America (USA); Middle East export risk was high by EII but underrepresented in genomic reconstructions, reflecting surveillance biases. - Variant dynamics: B.1.525 and B.1.1.318 co-circulated with Alpha during Nigeria’s second wave (Dec 2020–Mar 2021), with Delta initiating the third wave from June 2021. Both VOIs harbor E484K in Spike RBD, associated with reduced neutralization and experimental increases in infectivity. - Sampling fraction and representativeness: Overall genomic sampling represented ~0.026% of reported cases.

The integrated genomic and mobility analyses reveal that Nigeria’s regional connectivity was a key driver of SARS-CoV-2 introductions during the second wave, particularly for B.1.1.318, while B.1.525 (Eta) likely emerged within Nigeria and was heavily exported internationally, especially to Europe. Disparities between genomic estimates (which indicated many African-origin introductions) and travel-based risk indices (which highlighted Europe/North America post-reopening) underscore biases from uneven global sampling, low incidence reporting, and the absence of land travel data. The concordance between export peaks for B.1.525 and the EII, alongside high European air traffic and incidence, supports substantial Nigerian contributions to international spread. The findings highlight the utility of combining phylogeography with travel and epidemiological data to infer bidirectional transmission in under-sampled regions and stress that true within-Africa transmission is likely underestimated.

This study characterizes the emergence and spread of two SARS-CoV-2 VOIs in Nigeria across the first three waves, demonstrating: (1) B.1.1.318 emerged in Africa and entered Nigeria multiple times post-reopening, and (2) B.1.525 (Eta) likely emerged in Nigeria and was extensively exported, especially to Europe. Integrating genomic data with air travel and incidence metrics provided a more complete view of Nigeria’s role in regional and global transmission than genomics alone. The work underscores the need for strengthened, equitable genomic surveillance in Africa, improved data sharing, and inclusion of land-based mobility data. Future research should expand sampling coverage (including northern Nigeria), incorporate land border mobility data, refine underascertainment adjustments (e.g., excess mortality), and enhance real-time early warning systems for pandemic preparedness in the region.

- Undersampling: Overall sampling fraction was ~0.026% of reported cases; northern states were undersampled, particularly during the third wave, limiting within-country inference. - Metadata constraints: Inability to distinguish traveler-derived versus community-derived samples due to insufficient metadata prevented sensitivity analyses excluding travel-associated samples. - Mobility data gaps: Lack of land-based travel data likely led to underestimation of regional (within-Africa) introductions/exports in III/EII. - Reporting biases: III/EII rely on back-extrapolated infections from death time series, which are affected by reporting delays and underascertainment; underreporting in Africa (including Nigeria) likely biases estimates downward. - Global surveillance biases: Uneven international sequencing capacity leads to underestimation of transmission to/from undersampled regions (e.g., Middle East, neighboring African countries). - Temporal and geographic sampling unevenness: Sampling was not uniform across states or time, especially in the third wave, partly due to logistical and data-sharing challenges.