Temporal lineage replacements and dominance of imported variants of concern during the COVID-19 pandemic in Kenya.

Background: Kenya's COVID-19 epidemic was seeded early in March 2020 and did not peak until early August 2020 (wave 1), late-November 2020 (wave 2), mid-April 2021 (wave 3), late August 2021 (wave 4), and mid-January 2022 (wave 5).
Methods: Here, we present SARS-CoV-2 lineages associated with the five waves through analysis of 1034 genomes, which included 237 non-variants of concern and 797 variants of concern (VOC) that had increased transmissibility, disease severity or vaccine resistance.
Results: In total 40 lineages were identified. The early European lineages (B.1 and B.1.1) were the first to be seeded. The B.1 lineage continued to expand and remained dominant, accounting for 60% (72/120) and 57% (45/79) in waves 1 and 2 respectively. Waves three, four and five respectively were dominated by VOCs that were distributed as follows: Alpha 58.5% (166/285), Delta 92.4% (327/354), Omicron 95.4% (188/197) and Beta at 4.2% (12/284) during wave 3 and 0.3% (1/354) during wave 4. Phylogenetic analysis suggests multiple introductions of variants from outside Kenya, more so during the first, third, fourth and fifth waves, as well as subsequent lineage diversification. Conclusions: The data highlights the importance of genome surveillance in determining circulating variants to aid interpretation of phenotypes such as transmissibility, virulence and/or resistance to therapeutics/vaccines.

Introduction

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has impacted public health, social, political and economic spheres of life since its emergence in Wuhan, China and subsequent spread to the rest of the world. It reached every part of the globe in less than 9 months, and at the time of writing this report, it had infected over 535 million and killed over 6,309 million people globally. The virus is the etiological agent of coronavirus disease 2019 (COVID-19), a mysterious severe respiratory illness that first appeared in Wuhan, Hubei Province, China, in December 2019[1]. The origin of SARS-CoV-2 is controversial[2] but from genetic studies, the closest relatives are bat coronaviruses[3-6]. It remains to be proven whether the bat facilitated both the evolution of SARS-CoV-2 and its transmission to humans[7]. A recent study examining the evidence for a spillover event from Wuhan versus a Chinese Lab, found overwhelming evidence for a market in Wuhan as the most probable source of SARS-CoV-2, and not a Chinese government laboratory. According to that study, two independent zoonotic spillover events occurred two weeks apart, the first involving lineage B viruses while the second involved lineage A[8]. To date, COVID-19 has overshadowed all other human health calamities, ravaged global economies and disrupted human social interactions[9]. With the spread to different countries, the original A and B variants have diversified leading to the emergence of multiple variants, some with greater virulence than others[8].

In Kenya, the first confirmed case of COVID-19 was on 12th March 2020 from a Kenyan citizen returning home from the USA via London, UK[10]. Within two weeks, 31 cases traceable to the index case and other international travelers were identified. To try and curtail the spread, the government instituted a series of countermeasures that included border closures, mandatory quarantine on returning travelers, night curfews, ban on gatherings, and mandatory mask use while in public spaces[11]. While these measures slowed the spread of the disease, the virus still managed to infiltrate into the community, and new infections associated with local transmission events continued to drive the spread. To date, Kenya has reported over 327,000 cases and 5000 deaths[12], but the total number of cases are likely to be a gross underestimate considering the inadequacy of testing[13,14].

The earliest published description of SARS-CoV-2 in Kenya covered the period between February and March 2021 and traced the introduction and spread of European lineages in the coastal region of Kenya[15,16]. In the whole country, wave one peaked in early-August 2020, and by mid-September 2020, SARS-CoV-2 numbers had declined to very low levels, marking the “end” of the first wave. Buoyed by the reduction in COVID-19 numbers, most of the COVID-19 restrictions were lifted to ease pressure on a slumping economy[17]. The next one and half months were characterized by low infections, but this short-lived lull was interrupted by a spike in the number of cases that rose steadily, peaking at 1,554 by mid-November 2020. This triggered another round of lockdowns. The reopening of schools was postponed, and public gatherings, especially political rallies, stood banned. These measures, and probably because of other reasons progressively reduced infections, and by January 2021, the second wave burned out. Using population mobility models, it was theorized that the second wave was triggered by a population of Kenyans associated with higher socialeconomic status returning to pre-COVID-19 mobility patterns[18].

Between January 2021 and February 2022, Kenya experienced three COVID-19 waves, all of which were associated with SARS-CoV-2 variants of concern (VOCs) and variants of interest (VOIs) that emerged independently from different parts of the globe. The VOCs were associated with increased transmissibility, virulence, clinical presentation and decreased effectiveness to diagnostics and therapeutics, including vaccines[19]. Kenya has so far detected 4/5 currently classified VOCs, i.e. the Alpha variant (B.1.1.7/ 20I/V1), Beta (B.1.351, 20H/V2), Delta (B.1.617.2, 21 A/21I/21 J)[18,20] and Omicron (B.1.1.529, 21 K/21 L/21 M)[21]. SARS-CoV-2 genomes deposited in GISAID and Genbank from Kenya indicate the presence of two VOI, i.e. Eta (B.1.525, 20 A/S484K) and Kappa (B.1.617.1, 21B). The initial fear that some of the emerging mutants could negatively impact vaccine efficacy and constitute postvaccination “antigenic escape” has been witnessed with the Delta[22] and Omicron[23].

In this report, we use genomic surveillance to dissect the five COVID-19 waves that have occurred in Kenya since the beginning of the outbreak and provide data that show temporal lineage dominance, diversification and emergence of the more transmissible VOCs.

Methods

Ethics statement

This study was performed as part of public health surveillance approved by the Kenya government through the Ministry of Health (MOH) as part of response to the COVID-19 pandemic. The Scientific and Ethical Research Unit (SERU) of Kenya Medical Research Institute (KEMRI) approved a country-wide protocol to allow SARS-CoV-2 whole genome sequencing (SERU 4035) for all samples that were being collected as part of COVID-19 pandemic response, in order to allow tracking of virus evolution. Since the samples were obtained as part of public health surveillance, the IRB also waived the need to obtain prior informed consent.

Study sample acquisition

Multiple laboratories, including the Basic Science Laboratory (BSL) in Kisumu were designated by the MOH as COVID-19 testing centers. BSL started supporting COVID-19 mass testing and whole genome sequencing in March 2020, and by February 2022, the laboratory had screened 63,542 respiratory samples by RT-qPCR. Samples came from different parts of the country, and where indicated, a few came from other countries (Table 1). Nucleic acid isolation was performed using MagMAX Viral/Pathogen nucleic acid isolation Kits with the KingFisher Flex particle purification system (Thermo Fisher Scientific, CA, USA). Of the 63,542 samples tested, 8.5% (n = 5375) were positive for SARS-CoV-2 at varying cycle thresholds (Ct). Of these, 1089 with Cts <33 were selected for whole genome sequencing.

Table 1

Demographic data for subjects who contributed genome sequences used in this study.

		N	%
Age distribution (years)	<20	83	8.0
	21-30	268	25.9
	31-50	404	39.1
	51-60	83	8.0
	>61	89	8.6
	missing data	107	10.3
Median age	35
Gender	Male	574	55.5
	Female	357	34.5
	?	103	10.0
Nationality	Kenya	1017	98.4
	Uganda	9	0.9
	Congo	2	0.2
	India	1	0.1
	Zimbabwe	1	0.1
	Japan	1	0.1
	USA	2	0.2
	Rwanda	1	0.1
Infection waves	Wave 1	120	11.6
	Wave 2	79	7.6
	Wave 3	284	27.5
	Wave 4	354	34.2
	Wave 5	197	19.1

Whole genome sequencing and genome assembly

Complementary DNA (cDNA) was synthesized from RNA using random primers with either Superscript IV one-step reverse transcriptase kit (Thermo Fisher Scientific, CA, USA) or the LunaScript RT SuperMix Kit (New England Biolabs, MA, USA). The cDNA was then used for tiled multiplex PCR using the Q5 High-Fidelity 2X Master Mix (New England Biolabs) and ARTIC v3 primers (Supplementary Data 3 on Figshare) as described in the associated protocol[24]. The amplicons were cleaned with AMPureXP beads (Beckman Coulter, USA) and then used to create sequence libraries using the NexteraXT (Illumina, CA, USA) and Collibri ES (ThermoFisher Scientific, CA, USA) library preparation kits, as per the manufacturers’ instructions. The libraries were assessed on D1000 HS screen tape on a Tape Station 4200 (Agilent, CA, USA) for size distribution and concentration. A 12 pM library spiked with 10% Phix genome were sequenced on the MiSeq benchtop sequencer (Illumina, CA, USA) using 600 cycles V3 paired-end chemistry.

Read demultiplexing was conducted onboard the MiSeq using the MiSeq reporter v2.6. The reads were quality filtered to remove Illumina adapters and low-quality sequences using Trimmomatic v0.36[25]. Trimmed reads were assembled against SARS-CoV-2 Wuhan 1 as a reference (GenBank accession number: NC_045512) using bwa 0.7.5[26]. Samtools v0.1.19[27] was used to create pileups from the alignment, while ivar v1.3.1[28] was used to remove primers and build the consensus sequence. The consensus sequences were further curated using Nextclade Web v1.35.1[29] to screen out samples with too much missing information (Ns > 30% of the genome), mixed sites and private mutations.

Lineage and clade assignment

Lineage assignment was performed on each consensus sequence using PANGOLIN v3.1.17 and PANGOLEARN v2021-12-06 (Phylogenetic Assignment of named Global Outbreak LINeages)[30], which offers a hierarchical dynamic nomenclature describing a lineage as a cluster of sequences observed in a geographically distinct region with evidence of transmission in that region. Clades were assigned to each consensus sequence using Nextclade Web v 1.13.1. The Nextstrain clade system[31] uses a year-letter nomenclature on a clade exceeding 20% global representation and >2 positional differences from its parent clade while considering clade persistence with time as well as the extent of its geographical spread. Of the 1089 COVID-19 nasal specimens that passed the threshold for whole genome sequencing (Cts <33), 45 were dropped because they did not pass the threshold required for assigning Pango lineages. Ten additional samples were dropped because they lacked date of collection. The remaining 1034 genomes were used to monitor the evolution of SARS-CoV-2 lineages across the five COVID-19 waves.

Global data acquisition

To compare the genome sequences of the current study to global sequences, SARS-CoV-2 genomes were sampled from the Global Initiative on Sharing All Influenza Data (GISAID)[32]. Due to the huge number of genomes present in the GISAID, we opted to utilize the globally sampled genomes maintained by Nextstrain. These genomes are sub-sampled using a criterion that considers their spatial–temporal characteristics, as well as their genotypes, to yield a balanced and inclusive subsample of 2891 genomes. We also downloaded all SARS-CoV-2 genomes from Kenya deposited in GISAID that were outside this study, as well as very early reported genomes of each VOC. All the datasets were downloaded/sampled on 17th February 2022. For the Kenyan datasets and early reported genomes for VOCs, only sequences flagged as “complete (>29,000 bp)”, “high coverage only” (entries with <1% Ns and <0.05% unique amino acid mutations not seen in other sequence databases and with no unconfirmed insertion/deletions), and “low coverage excl” (excluded entries with >5%Ns) were downloaded from GISAID.

Phylogenetic analysis

Of the 1034 genomes that were used for lineage assignments, only 969 with genome lengths >27000 bp could be used for phylogenetic analysis.

Five phylogenies were constructed; one to determine the phylogenetic placement of the study genomes against a background of globally sampled genomes (context genomes). This tree consisted of 316 context genomes sampled from around the globe and 969 genomes from this study. The Alpha variants tree was constructed with genomes from Kenya (n = 381), against a global subsample (n = 164) that included some of the earliest reported Alpha variants (n = 43), mostly from England. The Beta variants tree was constructed with genomes from Kenya (n = 27), against a global subsample (n = 55) that included early Beta variants from Southern Africa (n = 30). The Delta variants tree was constructed with genomes from Kenya (n = 634) and a global subsample (n = 232) that included the earliest reported Delta variants from India (n = 22). The Omicron variants tree was reconstructed from the Kenyan genomes (n = 827) and a global subsample (n = 214) that included the earliest reported Omicron variants, mostly from South Africa (n = 8).

All trees were reconstructed with augur v14 as implemented in the Nextstrain pipeline version 3.0.6[31]. Within Nextstrain, a random subsampling method was used to cap the maximum number of context sequences from the rest of the world – to provide phylogenetic context, based on genomic proximity. Only genomes >2700 nt long were aligned with nextalign v1.11[29]. Phylogenies were reconstructed using IQTree[33] employing a General Time Reversible (GTR) model. Estimation of time-scaled phylogenies was done using Tree Time v0.8.6[34], assuming a nucleotide substitution rate of 8 × 10−4 per site per year, and a coalescent model. The resulting trees were visualized using auspice v2.29.1 and figtree v1.4.4[35].