Unique Variants of Avian Coronaviruses from Indigenous Chickens in Kenya

The avian gamma-coronavirus infectious bronchitis virus (AvCoV, IBV; Coronaviridae family) causes upper respiratory disease associated with severe economic losses in the poultry industry worldwide. Here, we report for the first time in Kenya and the Eastern African region two novel AvCoVs, designated IBV/ck/KE/1920/A374/2017 (A374/17) and AvCoV/ck/KE/1922/A376/2017 (A376/17), inadvertently discovered using random nontargeted next-generation sequencing (NGS) of cloacal swabs collected from indigenous chickens. Despite having genome organization (5′UTR-[Rep1a/1ab-S-3a-3b-E-M-4b-4c-5a-5b-N-6b]-3′UTR), canonical conservation of essential genes and size (~27.6 kb) typical of IBVs, the Kenyan isolates do not phylogenetically cluster with any genotypes of the 37 IBV lineages and 26 unique variants (UVs). Excluding the spike gene, genome sequences of A374/17 and A376/17 are only 93.1% similar to each other and 86.7–91.4% identical to genomes of other AvCoVs. All five non-spike genes of the two isolates phylogenetically cluster together and distinctly from other IBVs and turkey coronaviruses (TCoVs), including the indigenous African GI-26 viruses, suggesting a common origin of the genome backbone of the Kenyan isolates. However, isolate A376/17 contains a TCoV-like spike (S) protein coding sequence and is most similar to Asian TCoVs (84.5–85.1%) compared to other TCoVs (75.6–78.5%), whereas isolate A374/17 contains an S1 gene sequence most similar to the globally distributed lineage GI-16 (78.4–79.5%) and the Middle Eastern lineage GI-23 (79.8–80.2%) viruses. Unanswered questions include the actual origin of the Kenyan AvCoVs, the potential pathobiological significance of their genetic variations, whether they have indeed established themselves as independent variants and subsequently spread within Kenya and to the neighboring east/central African countries that have porous live poultry trade borders, and whether the live-attenuated Mass-type (lineage GI-1)-based vaccines currently used in Kenya and most of the African countries provide protection against these genetically divergent field variants.


Introduction
Infectious bronchitis virus (IBV) is an avian gamma-coronavirus (AvCoV; Coronaviridae family) that causes upper respiratory tract, enteric, renal, and reproductive disease (infectious bronchitis; IB) in domestic fowl and is associated with severe economic losses in the domestic poultry industry globally [1]. This virus was first reported as the causative agent of "respiratory disease of chicks" in the U.S. in the 1930s; however, it has been detected globally over the decades since and has also been reported in non-domestic Galliforms [2][3][4]. Available data from the U.S.A. and Europe have shown that effective control of clinical

Samples
The clinical samples used in the current study were collected in January 2017 from indigenous chickens traded at a live bird market (LBM) in Kariobangi North, Nairobi, an indoor market where the chickens are kept in wire-mesh cages (an average of 12 birds per cage). In the Kenyan context, indigenous chickens, which are kept under free-range or caged management systems in backyard farms, are defined as "non-descript crosses of Asiatic meat and game types, Mediterranean egg-types and Bantams of various origins" [30][31][32]. The sampled chickens, which originated from backyard rural farms in eastern (Machakos and Kitui counties) and central (Kiambu county) regions of Kenya, did not exhibit any overt clinical signs of respiratory or other diseases at the time of sampling. From each trader's consignment of chickens, a cloacal (CL) and an oropharyngeal (OP) swab were collected from each of three randomly selected adult chickens per cage and immediately placed in individual 2.0 mL cryogenic vials (Corning Inc., New York, NY, USA) that contained 1.5 mL of Difco™ brain-heart-infusion broth (Thermo Fisher, Waltham, MA, USA) according to standard procedures [27]. Swabs were immediately stored in liquid nitrogen and preserved at −80 • C until shipment to the Southeast Poultry Research Laboratory (SEPRL) of the United States Department of Agriculture, Agricultural Research Service (USDA-ARS) in Athens, GA, USA for analysis. Sampling was conducted in collaboration with the Directorate of Veterinary Services (DVS) and regional Central Veterinary Laboratories (CVLs) in Kenya.

RNA Extraction
Total RNA was extracted from 50 µL of each CL or OP sample (prepared separately for each bird) using the MagMAX™-96 AI/ND Viral RNA Isolation Kit (Thermo Fisher Scientific, Waltham, MA, USA) and eluted in 50 µL of elution buffer. To selectively deplete ribosomal RNAs (rRNAs) from the host (18S, 28S, and mitochondrial) and bacteria (16S/23S), 12 µL of the total RNAs were treated with an in-house RNaseH rRNA depletion protocol and purified using SPRI beads (Agencourt RNAClean XP Kit; Beckman Coulter Life Sciences, Indianapolis, IN, USA) as recently described [33].

Sanger Sequencing
Sanger sequencing was used to close short gaps in the consensus sequence obtained from one of the samples from this study (see results section). For this, the SuperScript IV One-Step RT-PCR kit (Thermo Fisher Scientific, Waltham, MA, USA) was used to convert and amplify missing regions using a set of primers specifically designed for the gaps. Amplicons were visualized in 1.5% agarose gel (0.5X TBE), followed by purification of excised DNA bands using the Zymoclean Gel DNA Recovery Kit (Zymo Research, Irvine, CA, USA) and Sanger sequencing using the BigDye™ Terminator v 1 Cycle Sequencing Kit performed on a 3730 xl DNA Analyzer (Thermo Fisher Scientific, MA, USA).

Phylogenetic Analysis
For classification, the S1 gene sequences of representative strains of the 32 IBV lineages, the proposed GVII and GVIII, UVs, and turkey coronaviruses (TCoVs) retrieved from GenBank (n = 256 sequences; full-length sequences selected based on classification by Valastro et al. [19]), together with those obtained from this study, were aligned using MAFFT v7.490 [40] executed in Geneious Prime and trimmed using trimAl tool v1.3 [41] to minimize the effects of poorly aligned regions. Phylogenetic analyses were performed using the maximum likelihood method in MEGA6 with 1000 bootstrap replicates of the original data; all positions with less than 95% site coverage were eliminated from phylogenetic tree reconstructions [42]. Phylogenetic analyses of the complete genome and all five non-spike genes were performed as described above using representatives of the IBV lineages and TCoVs. Comparative pairwise homology analyses of the genome and CDS of the Kenyan and other IBVs and TCoVs was performed based on sequence and phylogenetic analyses and a heatmap generated using ClustVis [43].

NGS-Based Nontargeted Virus Discovery
Forty-eight libraries (i.e., one CL and OP swab from each of 24 birds) derived from backyard farms in the eastern (Machakos and Kitui counties; n = 15 and 3 samples, respectively) and central (Kiambu county; n = 6 samples) regions of Kenya were successfully prepared, sequenced together (multiplexed) on an Illumina MiSeq sequencer, and processed as described above. The NGS pipeline detected IBV RNA in two CL samples, which contained 9148 and 17,811 IBV-specific read pairs. These two samples were obtained from the same trader, who had sourced the chickens from middle-men at an open-air LBMs in Machakos county. None of the other tested samples (including the OP counterparts of the IBV-positive CL samples) contained detectable quantities of IBV-specific RNAs. In addition to IBV, sample A374 contained RNAs of three other viral agents, i.e., virulent Newcastle disease virus (vNDV), avian rotavirus-G (AvRV-G), and pigeon picornavirus B (PiPV-B) and one avian pathogenic bacterium (Ornithobacterium rhinotracheale; ORT), whereas sample A376 contained NDV and two avian pathogenic bacteria (ORT and Riemerella anatipestifer). We have recently published complete genome sequences of the vNDVs that co-infect with the IBVs [29].
The complete genome sequences of A374 and A376 showed 87.2% nucleotide identity to each other, or 93.1% identity when the S gene is excluded. Comparative pairwise homologies of the CDS across the genomes of the two Kenyan AvCoVs showed that complete S-genes have the lowest identities (49.1% and 48.7% at the nucleotide and deduced amino acid sequences, respectively), whereas the highest identities are in M (96.4% for both nucleotide and amino acid levels) and 4b (96.1% and 98.9% at the nucleotide and amino acid levels, respectively) ( Table 3). Regarding the S1 gene, which is used for AvCoV classification, the two Kenyan isolates share low identities at the level of nucleotide (38.8%) and amino acid (21.6%) sequences.
Based on the S1 gene, the two Kenyan IBVs identified here do not phylogenetically group with any of the IBVs in lineages GI-GVIII or the UVs [19]; rather, A374/17 clusters non-monophyletically with Eurasian and Middle Eastern viruses (GI-16 and GI-23) and A376/17 with Asian TCoVs ( Figure S1). However, the bootstrap value supporting the separation of A374/17 from GI-16 and GI-23 strains is only 23-34% compared to 99% bootstrap support for A376/17 vs. the Asian TCoVs (see Figures 1 and S1). The complete S gene of A374/17 clusters with the globally distributed GI-13 and GI-16 viruses, the West African and European GI-14 and GI-26 viruses, and A376/17 with Asian TCoVs, but the Kenyan isolates branch out from these viruses with bootstrap support of 63% and 100%, respectively; the complete genome sequence tree shows similar clustering with 100% bootstrap support ( Figure 1). Phylogenetic analyses based on the other five genes, i.e., gene 1 (Rep1ab), gene 3 (3a, 3b, and envelope), gene 4 (membrane 4a and 4b), gene 5 (5a and 5b), and gene 6 (nucleocapsid and 6b (also referred to as putative 7 protein gene)), differ from the topologies of the S1, complete S, and genome trees in that the two Kenyan isolates cluster together in all five genes ( Figure S2). Table 3. Genomic organization and comparative homologies of isolates A374/17 and A376/17 identified in the current study. The position and nucleotide (nt) and amino acid (aa) lengths of the untranslated regions (5 -/3 -UTRs) and protein-coding sequences (CDS) of the genes are indicated.   Figure S1). Sequence names include GenBank accession numbers, avian species (where applicable), country abbreviation, strain/isolate, and year of sample collection. Lineages are named according to the current IBV classification system (UV indicates unique variant); geographical distribution of the lineages are color-coded [19]. The phylogenetic reconstruction involved 72, 37, and 34 nucleotide sequences (final datasets with 1377, 3233, and 26,625 positions) for the S1, complete S, and complete genome trees, respectively.

Analysis of the S Glycoprotein
Because the S gene is critical in IBV's pathogenesis (binding, entry, and tissue tropism within susceptible host cells and induction of host's immune responses; [17]), domain analysis of the gene was performed on the two Kenyan viruses compared to strains shown in the complete genome and S-gene trees in Figure 1 Structural features of isolate A374/17 S glycoprotein are illustrated in Figure S3, among which is the S1/S2 cleavage site with the furin recognition motif R-L-R-R↓S, which is located within a 20 amino acid residue region required for cleavage and fusion efficiency [44]. The auxiliary S2′ cleavage site in A374/17 (P-S-S-P-T-G-R↓S) has a mutation of the conserved R/K acid residue (consensus motif in IBVs is P-X-S-P-[R/K]-X-R↓S). The membrane fusion peptide of A374/17 has all four residues (CTAGPLGFVKDLVC), which

Analysis of the S Glycoprotein
Because the S gene is critical in IBV's pathogenesis (binding, entry, and tissue tropism within susceptible host cells and induction of host's immune responses; [17]), domain analysis of the gene was performed on the two Kenyan viruses compared to strains shown in the complete genome and S-gene trees in Figure 1 Structural features of isolate A374/17 S glycoprotein are illustrated in Figure S3, among which is the S1/S2 cleavage site with the furin recognition motif R-L-R-R↓S, which is located within a 20 amino acid residue region required for cleavage and fusion efficiency [44]. The auxiliary S2 cleavage site in A374/17 (P-S-S-P-T-G-R↓S) has a mutation of the conserved R/K acid residue (consensus motif in IBVs is P-X-S-P-[R/K]-X-R↓S). The membrane fusion peptide of A374/17 has all four residues (CTAGPLGFVKDLVC), which are conserved in CoVs [45], and the number of N-linked glycosylation sites (n = 23) is well within the range of 19-39 residues expected for CoVs [14].

Mutations in the S1 Subunit Hypervariable Regions (HVRs)
The three hypervariable regions (HVRs) of IBV's S1 subunit [46,47] in A374/17 are located at amino acid residues 60-88 (HVRI), 115-142 (HVRII), and 277-295 (HVRIII), all of which are heterogeneous when compared to other IBVs (Figure 3). Notable variations in the HVRI include the six residues described previously as critical for attachment to respiratory tract tissues of the GI-1 prototype M41 strain (residues N38, H43, S56, P63, I66, and T69; GenBank accession number AY851295; [48] [49], which is also present in some French Guinea fowl CoVs (GfCoVs). Although A376/17 and the Chinese TCoVs have the critical arginine (R) residue at the S2 site, the conserved asparagine (N) and glycine (G) residues are substituted with serine (S) residues (SQSR↓S), whereas TCoVs from France and Poland have proline (P) or S residues (P/SQGR↓S). The membrane-fusion peptide (FP) is conserved in the Polish and North American TCoVs, and in French GfCoVs (consensus motif CIASRGGSFTNLADLTC; conserved residues are underlined [45]) compared to substitutions in the Kenyan A376/17 (n = 2), Chinese (n = 1), and the French (n = 5) TCoVs. The antigen-binding region (FabR; 45 amino acid residues located at the S1 Cterminus [49,50]) of the Kenyan A376/17 and the Asian TCoVs are identical (except one amino acid difference; G vs. S) compared to high variations in the European TCoVs and GfCoVs. The numerous amino acid variations in the HVR are expected in TCoVs [51]. The S protein of A376 has 26 N-linked glycosylation sites, which is consistent with other CoVs [14].

Analysis of Recombination Events
Because recombination significantly contributes to the continuous emergence of IBV variants, RDP4 was used to investigate possible recombination events in the two Kenyan AvCoVs, and strains from the most widely distributed lineages (GI-1, GI-13, GI-16, and GI-19), North American/Asian lineages (GI-2 and GI-3), European/African lineages (GI-12 and GI-14), European lineage GI-21, indigenous Middle Eastern/African lineages (GI-23 and GI-26, respectively), and TCoVs (see Figure 1). One of the recombination signals was detected in Rep1ab (nsp11/12; RdRp) of both A374/17 and A376/17 with the Canadian strain TCoV/MG10 and Sudanese strain AR251-15 (lineage GI-19) as the major and minor parental strains, respectively (Table 4).  Figure 3. Heterogeneity of the HVRs in the S1 subunit of the Kenyan A374/17 (in red color) compared to other IBVs. Dots and dashes in the alignments indicate identical and missing (or gaps in alignment of) amino acid residues, respectively. Amino acid residue positions are numbered in reference to the consensus sequence. UV indicates a unique variant. The amino acid residues highlighted with asterisks in shaded boxes ("*"; i.e., N38, H43, S56, P63, I66, and T69) are reportedly critical for attachment of the GI-1 prototype strain AY851295/M41 (in blue color) to chicken respiratory tract tissues [48]. The classification is based on Valastro et al. [19]. UV indicates a unique variant. The amino acid residues highlighted with asterisks in shaded boxes ("*"; i.e., N38, H43, S56, P63, I66, and T69) are reportedly critical for attachment of the GI-1 prototype strain AY851295/M41 (in blue color) to chicken respiratory tract tissues [48]. The classification is based on Valastro et al. [19]. 3.75 × 10 −20 a "major parent" indicates a sequence in another strain most closely related to the sequence surrounding the fragment transferred to the recombinant strain through recombination. b "minor parent" indicates the sequence closely related to the fragment in the recombinant strain.
A recombination event was detected by six of the nine RDP4 methods [38] in the Kenyan isolate A376/17 with the Asian ahysx-1/16 and North American TX-1038/98 strains predicted as the minor and major parental sequences, respectively (Table 4 and Figure 4). Note that here, "minor parent" represents the strain (i.e., ahysx-1/16) closest to the sequence fragment that has potentially been transferred to the recombinant isolate (i.e., Kenyan isolate A376/17) via a recombination event from the "major parental" strain (i.e., TX-1038/98) closest to the sequence surrounding recombination breakpoints in the recombinant isolate. This recombinant signal was not detected in any other of the strains used in the analyses. It is of note that the Chinese TCoV strain ahysx-1/16 isolate was itself reported as being of recombinant origin between an IBV backbone and a TCoV-like S gene donor [52], but our analysis did not give evidence that strain ahysx-1/16 and the Kenyan isolates share the same parents. Nonetheless, the S sequence of A376/17 is more similar to ahysx-1/16 than to any other published sequence, and the recombinant fragment itself (3123 nucleotides in length) covers 39.5% of the C-terminus of nsp16, the full-length S1 gene, and approximately 75%% of the S2 region.
Four and six other recombination signals were detected in A374/17 and A376/17, respectively, but these were considered possible misidentifications of recombination because of one or more of the following reasons: (1) one or both breakpoints could not be identified; (2) the recombinant signal represented only trace evidence of a recombination event, i.e., the p-value was less than 10 −5 cutoff and/or was supported by fewer than five out of the nine RDP4 detection methods [38]; and (3) if there was more than 30% probability that one or both the major and minor parental strains are likely to be the actual recombinant. For instance, a recombination event signal was predicted (four methods; p-value of 4.6 × 10 −4 ) in Rep1ab (nsp14-15) of A376 with the Kenyan A374 and a GI-13 4/91-vaccine strain (GenBank accession number KF377577) as the major and minor parents with 93.9% and 86.9% identities, respectively, but A374 was suggested to be the actual recombinant. the analyses. It is of note that the Chinese TCoV strain ahysx-1/16 isolate was itself reported as being of recombinant origin between an IBV backbone and a TCoV-like S gene donor [52], but our analysis did not give evidence that strain ahysx-1/16 and the Kenyan isolates share the same parents. Nonetheless, the S sequence of A376/17 is more similar to ahysx-1/16 than to any other published sequence, and the recombinant fragment itself (3123 nucleotides in length) covers 39.5% of the C-terminus of nsp16, the full-length S1 gene, and approximately 75%% of the S2 region.  Table 4) with a North American major parent (strain TX-1038/98) and an Asian minor parent (strain ahysx−1/16 (see Table  4 footnotes)). The top panel illustrates the sequence in the recombinant Kenyan isolate A376/17 transferred by recombination event, which corresponds to the pink area flanked by recombination breakpoints on the left and right boundaries (genomic positions 19,995 and 23,117, respectively) in the recombination signal plot in the bottom panel. Also shown are structural/domain features in the recombinant sequence, which include N-/C-terminal receptor-binding domains (N-/C-RBD), hypervariable region (HVR), antigen-binding region (FabR), S1/S2 cleavage site, fusion peptide (FP), and heptad repeat region 1 (HR1); see Figure S4 for detailed features of A376/17. Definitions: "minor parent" represents the strain (i.e., ahysx-1/16) closest to the sequence fragment transferred to the recombinant strain (i.e., Kenyan isolate A376/17) via recombination event; "major parent" represents the strain (i.e., TX-1038/98) closest to the sequence surrounding recombination breakpoints in the recombinant isolate.  Table 4) with a North American major parent (strain TX-1038/98) and an Asian minor parent (strain ahysx−1/16 (see Table 4 footnotes)). The top panel illustrates the sequence in the recombinant Kenyan isolate A376/17 transferred by recombination event, which corresponds to the pink area flanked by recombination breakpoints on the left and right boundaries (genomic positions 19,995 and 23,117, respectively) in the recombination signal plot in the bottom panel. Also shown are structural/domain features in the recombinant sequence, which include N-/C-terminal receptor-binding domains (N-/C-RBD), hypervariable region (HVR), antigen-binding region (FabR), S1/S2 cleavage site, fusion peptide (FP), and heptad repeat region 1 (HR1); see Figure S4 for detailed features of A376/17. Definitions: "minor parent" represents the strain (i.e., ahysx-1/16) closest to the sequence fragment transferred to the recombinant strain (i.e., Kenyan isolate A376/17) via recombination event; "major parent" represents the strain (i.e., TX-1038/98) closest to the sequence surrounding recombination breakpoints in the recombinant isolate.

Discussion
Molecular and serological surveys in the backyard and commercial poultry flocks have reported IBV variants in the Mediterranean basin and Western and Northern African countries [27,53], but epidemiological and genomic data on IB and its impact on the poultry industry in Africa are scarce. This is despite IBV being one of the most common and important viral agents associated with respiratory disease and severe reduction in egg production in chickens, with IB outbreaks occurring in commercial and backyard flocks regardless of their vaccination status [54]. Contributing factors to this situation include a lack of surveillance initiatives for poultry viruses and inadequate genomic sequence data of detected isolates, most of which remain uncharacterized. The situation worsens in Central and Eastern Africa, where only Cameroon (in Central Africa) out of the 16 countries in these regions reported five GI-14, GI-16, and GI-19 strains during a 2013 surveillance of commercial poultry [25]. Towards the southernmost African region with 11 countries, only South Africa, Zimbabwe, and Botswana have reported serological detections of anti-IBV antibodies and a few GI-1, GI-13, and GI-19 variants [53,[55][56][57][58].
Here, we report for the first time in Kenya and the Eastern African region, two AvCoV variants identified using random untargeted NGS of samples collected from rural backyard indigenous chickens during a 2016-2018 surveillance for vNDVs in domestic and wild avian species in the country. The IBV RNAs were detectable in the CL samples and not in their counterpart OP samples, which suggests the possibility that the Kenyan isolates A374/17 and A376/17 are gastroenteric, but this remains to be further investigated because birds in general shed viruses for longer durations via the cloacal route compared to the respiratory routes. This is not unexpected, because the gastrointestinal tract is capable of supporting IBV replication [59]. Furthermore, the absence of overt signs of respiratory or other diseases in the chickens at the sampling time agrees with gastrointestinal tract IBV infections that are yet to be demonstrated to result in enteropathogenesis [60]. We have recently reported sequences of novel subgenotype V.3 vNDVs in A374 CL and A376 OP samples [29], and five low pathogenic influenza A subtype H9N2 viruses [61] from other samples from the same LBM as those in the current study (collected at the same time and originating from the same Machakos county). It is highly likely that these respiratory viruses have been disseminated to other Kenyan and East African regions, because the LBMs and associated trade routes form the backbone of the poultry industry in these regions [29,61].
The current IBV classification is predicated on the S1 gene being the most variable and immunogenic (epitomic) genomic region, and the most commonly sequenced [19], but the question arises whether one gene alone sufficiently reflects the actual variability of AvCoVs. The inconsistencies in the topologies of the full-length S1 gene, complete S CDS, and complete genome sequence trees vs. the trees of the other five viral genes observed in the current study reiterates the challenge in the classification and nomenclature of new AvCoVs that emerge continuously because of their high propensity for recombination. It is, therefore, not far-fetched to conclude that the evolution and epidemiology of these viruses could potentially be more reliably tracked by expanding genomic studies from using the S1 gene to using complete genome sequences, full-length of the S protein CDS, or by combining data from phylogenies of each gene.
The higher nucleotide identity among the two Kenyan isolates in the non-spike genomic regions (93.1%) and the phylogenetic grouping together of all five viral genes except the S-gene suggest that their genomic backbones have a common origin. These Kenyan isolates are closest to but group distinctly from the Middle Eastern IBVs (isolate A374/17) and Asian TCoVs (isolate A376/17) and are distantly related to the indigenously African GI-26 viruses (as described by Valastro et al., [19]), suggesting that they are indigenous (unique) to Kenya. It is common to refer to IBVs as indigenous to specific geographical regions where they were originally isolated, e.g., the North American Massachusetts (lineage GI-1) and Arkansas (lineage GI-9), the Dutch D274 and British UK/6/82 (lineage GI-12), and the Chinese QX (GI-19) viruses [19,63]. Whereas some variants remain restricted to circulating only within avian subpopulations in the geographical regions of their original discovery, other variants can persist long enough to adaptively evolve into independent strains that consequently spread and establish themselves in entirely different geographical areas where they can gain substantial economic importance in the poultry industry, even becoming responsible for disease outbreaks [19,64]. The questions as to what extent the Kenyan variants are established in eastern Kenya and whether they may have spread within the country or east Africa, remain to be determined. The possibility of any of these scenarios is of concern in view of the highly unregulated live poultry trade networks in the country and the region [65].
Together, data from the current study raise questions about the actual origin and evolution of the Kenyan AvCoVs, whether they are also circulating in commercial poultry flocks, and whether migratory wild birds and other non-domestic avian species frequently found scavenging together with rural backyard poultry play roles in the evolution and dissemination of the variants. From our sequence and phylogenetic analyses, the two Kenyan isolates are more similar to each other than to any other sequenced AvCoV sequences over the majority of the genomic backbone, suggesting a shared evolutionary history, and lack of similarities to other sequenced full-length AvCoVs. The S-protein coding sequence of isolate A376 appears to be of recombinant origin similar to but distinct from other sequenced TCoVs such as the Asian and North American TCoVs (at 85.9% and 79% nucleotide similarity, respectively), and only distantly related to other IBVs. This is a known phenomenon, in which the majority of TCoVs originate from a recombination of an IBV backbone with an S-gene fragment of often unknown origin [18,52,66]. The ability of AvCoVs to micro-evolve into novel variants via recombination of S1 sequence fragments from different backgrounds is consequential because they could confer on the variants the ability to infect new host tissues or naïve susceptible hosts (diversification of tissue tropism and host range), for instance, if the changes enhance the susceptibility of the S gene to proteolytic activation or increase affinity to receptor binding and efficiency to gain cellular entry (heightening of pathogenicity) [18,67]. Novel variants could also gain the ability for immune escape and unresponsiveness to vaccinations.
In most African countries, live-attenuated and inactivated Mass-type-based vaccines are the main control method [68]. Kenya has two official government-approved Mass-type vaccines, i.e., CEVAC ® BI L (strain B-48; used via eye-drop or spray method) and CEVAC ® Corymune 7K (strain M-41; used subcutaneously). However, even with the availability of these vaccines, vaccination rates are low outside commercial poultry farming, which is an issue of concern because~80% of the Kenyan poultry are under non-commercial settings (i.e., free-range birds kept in small-scale flocks of~30 birds). Some commercial farmers also tend to consult private veterinarians for vaccine options other than the governmentapproved ones. Based on the phylogenetic and sequence analyses presented in this study, the two Kenyan UVs presented in this study are clearly genetically different from the GI-1-based vaccine strains, which are currently used in Kenya and other African countries. A lack of evidence as to whether the Kenyan UVs cause clinical disease makes it necessary to review the protection of the Mass-type vaccines used in Kenya, because there is evidence that variations of as little as 5% in the S1 gene sequence can negatively affect vaccine efficacy [12].

Conclusions
We have reported for the first time in Kenya and the Eastern Africa region the presence of AvCoVs in backyard chickens. Our analyses have demonstrated that the two Kenyan AvCoVs are UVs that are distinct from each other but likely share a common origin not yet represented in the literature, and they do not group with any of the genotypes or strains belonging to the established lineages or other UVs. It is yet to be elucidated whether the genomic variations of the two Kenyan AvCoVs when compared to other previously reported strains are of any pathobiological significance. Outstanding questions include the actual origin of the Kenyan AvCoVs; whether they have indeed established themselves as independent variants in the eastern county of Machakos; and, if so, to what extent they may have spread to the rest of the country and the neighboring east African countries that have porous borders in terms of trade of live poultry. Nevertheless, data from this study underscore the need for active surveillance, identification, and characterization of the variants that are evolving and circulating in the country and the region, because effective vaccination against novel variants requires tailor-made vaccines.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/v15020264/s1, Table S1. Positions and sizes (nucleotides) of nsp2-nsp16 mature polypeptides proteolytically produced from Rep1a (nsp2-nsp11) and Rep1b (nsp12-nsp16) of the Kenyan AvCoVs reported in the current study. Abbreviations: DUF-protein domain of unknown function; ExoN-N-terminal exoribonuclease; Hel-helicase; holo-RdRp-RNA-dependent RNA polymerase (RdRp) holoenzyme; NendoU-nidoviral RNA uridylate-specific endoribonuclease; nsp-non-structural protein; PL pro-papain-like protease; RBP-RNA-binding protein; Rep-Replicase; 3CL pro-3C-like protease; TMD-transmembrane domain protein; 2 -O-MTase-(nucleoside-2 -O-)-methyltransferase; and ZF-MFP-zinc-finger multifunctional protein. Figure S1: Phylogenetic clustering of the two Kenyan IBVs identified in the current study (bold font) and representatives of other AvCoVs based on the nucleotide sequences of the full-length S1 gene. The lineages are named based on the current IBV classification system and are colorcoded based on their geographical distribution [19][20][21][22][23][24]. The phylogenetic reconstruction involved 256 sequences, with the final dataset having 1433 positions. Sequence names include GenBank accession numbers, avian species (where applicable), country, strain/isolate, and year of sample collection. Figure S2: Phylogenetic analyses of Kenyan AvCoVs (in red font) and other strains based on genes 1, 3, 4, 5, 6, and 6b. IBV lineages and TCoVs are highlighted in bold (in brackets) and blue text, respectively. Figure S3: Schematic representation of structural features of the Kenyan isolate A374/17 S glycoprotein compared to representatives of other IBV lineages (selected based on phylogenetics in Figure 1 and Figure S1). Dots and dashes indicate identical and missing (or gaps in alignment of) amino acid residues, respectively. Amino acid residues that are conserved across IBVs are underlined in the consensus sequence. Features of subunit S1 include N-terminal signal peptide (SP); N-and C-terminal domains (S1-NTD/CTD, which harbor the receptor-binding domains (RBD); and hypervariable regions (HVR I-III). Shown is a 20-amino acid residues S1/S2 cleavage site (running from positions P14 to P6 ), which consists of a core region (8-residues; position P6 to P2 with R-X-[K/R]-R↓S canonical motif; "X" is any amino acid residue; and "↓" indicates cleavage position illustrated by a red dotted vertical line) flanked by two solvent accessible regions (8-residue; P7 to P14, and a 4-residue; P6 to P2 ). Note the backward and forward numbering of P1-P14 and P1 -P6 , respectively, starting at the conserved R immediately upstream of the cleavage S1/S2 site. Subunit S2 domains include auxiliary S2 cleavage motif (P-X-S-P-R-X-R↓S) [49], fusion peptide (FP with consensus motif CIASRGGSFTNLADLTC), heptad repeat regions (HR1 and HR2), and transmembrane/non-cytoplasmic domains (TM/Non-CTD). Purple vertical bars represent predicted N-linked glycosylation sites (n = 23 in A374/17; positions of the asparagine (N) residues in N-X-S/T sequons are indicated; X is any amino acid residues). Figure S4: Schematic representation of structural features of the Kenyan isolate A376/17 S glycoprotein compared to Eurasian and North American TCoVs. Dots and dashes indicate identical and missing (or gaps in alignment of) amino acid residues, respectively. Conserved amino acid residues are underlined in the consensus sequence. Features of subunit S1 include N-terminal signal peptide (SP), N-and C-terminal receptorbinding domains (S1-N/C-RBD), a hypervariable region (HVR), and an antigen-binding region (FabR; n = 45 amino acid residues in length). The boxed amino acid residues in the FabR are reported to be critical for antibody neutralization in the North American TCoVs [50]. The S1/S2 cleavage site (R-X-R-R↓S; "X" is any amino acid residue and "↓" is the cleavage position illustrated by a red dotted vertical line) is shown. Subunit S2 features include an auxiliary S2 cleavage motif with NQGR↓S consensus observed in the North American and some French TCoVs [49]), fusion peptide (FP; consensus motif CIASRGGSFTNLADLTC), heptad repeat regions (HR1 and HR2), and transmembrane/non-cytoplasmic domains (TM/Non-CTD). Purple vertical bars represent predicted N-linked glycosylation sites (n = 26 in A376/17; positions of the asparagine (N) residues in N-X-S/T sequons are indicated; X is any amino acid residues).

Data Availability Statement:
The nucleotide sequence data of the Kenyan IBV isolates reported in this paper have been submitted to GenBank database and have been assigned the accession numbers OP899612 and OP899613.