Whole-Genome Characterization of Rotavirus G9P[6] and G9P[4] Strains That Emerged after Rotavirus Vaccine Introduction in Mozambique

Mozambique introduced the Rotarix® vaccine into the National Immunization Program in September 2015. Following vaccine introduction, rotavirus A (RVA) genotypes, G9P[4] and G9P[6], were detected for the first time since rotavirus surveillance programs were implemented in the country. To understand the emergence of these strains, the whole genomes of 47 ELISA RVA positive strains detected between 2015 and 2018 were characterized using an Illumina MiSeq-based sequencing pipeline. Of the 29 G9 strains characterized, 14 exhibited a typical Wa-like genome constellation and 15 a DS-1-like genome constellation. Mostly, the G9P[4] and G9P[6] strains clustered consistently for most of the genome segments, except the G- and P-genotypes. For the G9 genotype, the strains formed three different conserved clades, separated by the P type (P[4], P[6] and P[8]), suggesting different origins for this genotype. Analysis of the VP6-encoding gene revealed that seven G9P[6] strains clustered close to antelope and bovine strains. A rare E6 NSP4 genotype was detected for strain RVA/Human-wt/MOZ/HCN1595/2017/G9P[4] and a genetically distinct lineage IV or OP354-like P[8] was identified for RVA/Human-wt/MOZ/HGJM0644/2015/G9P[8] strain. These results highlight the need for genomic surveillance of RVA strains detected in Mozambique and the importance of following a One Health approach to identify and characterize potential zoonotic strains causing acute gastroenteritis in Mozambican children.


Introduction
Rotavirus remains one of the primary causative agents of gastroenteritis in children under five years of age, exerting a substantial global health burden [1,2].It is estimated that rotavirus infections resulted in 128,500 deaths in 2016, of which 104,733 occurred in Sub-Saharan Africa [3].
Four rotavirus vaccines have received prequalification from the World Health Organization (WHO): Rotarix ® , RotaTeq ® , Rotavac ® and Rotasiil ® .These vaccines have demonstrated significant efficacy in reducing diarrheal morbidity and mortality on a global scale [8].In September 2015, Mozambique introduced the Rotarix ® vaccine into the National Immunization Program.Since then, the prevalence of rotavirus infection in Mozambique has decreased from 40.6% to 19.1% [14,15], the vaccine effectiveness has been estimated to be lower than what was reported in many other African countries, with an effectiveness of 30% against G1P [8] strains and 35% against non-G1P [8] strains [16].
Prior to vaccine implementation, G9P [8] and G1P [8] had been the most predominant genotypes in Mozambique [15,17].Whole-genome analyses (WGA) of human Mozambican RVA strains before vaccine introduction have suggested genetic diversity was partially driven by reassortment events between animal and human strains [18,19].However, postvaccine introduction, G1P [8] became the predominant genotype nationwide [15,17], despite the use of a G1P [8]-based vaccine, although WGA of these G1P [8] strains indicated no significant mutations in epitope regions that might lead to vaccine escape, and no distinct clustering was observed between pre-and post-vaccine strains [20].
During the post-vaccine period, G9P [4], G9P [6], G3P [8] and G3P [4] have emerged as predominant genotype combinations.However, the origin of these strains, as well as their relation to the G9P [8] strains reported before vaccine introduction, remains unclear [14].Therefore, the whole genomes of strains detected between 2015 and 2018 were determined and analyzed in the current study in order to elucidate the origin of the G9 genotype detected following vaccine introduction in Mozambique.

Sample Collection
Fifty-seven fecal samples collected between 2015 and 2018 as part of ongoing hospitalbased sampling within the National Diarrhea Surveillance System (ViNaDia) in Mozambique [15] were selected for WGA.The samples collected in 2015 represent the pre-vaccine period and those collected in 2016-2018, the post-vaccine period (Table S1).These samples had previously tested positive for RVA by ELISA (Prospect EIA rotavirus, Basingstoke, United Kingdom) and the binary genotype combination was determined by multiplex Reverse Transcriptase (RT)-PCR [21][22][23].

Viral Genomic dsRNA Extraction, cDNA Library Building and Illumina MiSeq Sequencing
Total RNA was extracted from stool samples with TRI-reagent (Sigma, Darmstadt, Germany), and single-stranded RNA was precipitated with lithium chloride.The self-priming Viruses 2024, 16, 1140 3 of 17 PC3-T7 loop primer (Integrated DNA Technologies, Coralville, IA, USA) was ligated to dsRNA to obtain full-length sequences.Complementary DNA (cDNA) was synthesized using the Maxima H Minus double-stranded cDNA kit (Thermo Fisher Scientific, Massachusetts, MA, USA) as previously described [19].The cDNA was synthesized at the Next Generation Sequencing Unit at the University of the Free State in Bloemfontein, South Africa.In brief, the cDNA library was made by NEBNext Ultra RNA Library Prep Kit for Illumina v1.2 (New England Biolabs, Ipswich, MA, USA), and NEBNext Multiplex Oligos for Illumina (New England Biolabs, Ipswich, MA, USA) according to the manufacturer's instructions and purified using Agencourt AMPure XP magnetic beads (Beckman Coulter, Brea, CA, USA).Nucleotide sequencing was performed using an Illumina MiSeq sequencer (Illumina, San Diego, CA, USA) using a MiSeq Reagent Kit V3 (Illumina, San Diego, CA, USA) [19].

Genome Assembly
A de novo assembly was performed for all samples using CLC Bio Genomics Workbench (12.0.3;Qiagen, Aarhus, Denmark); all contigs with an average coverage above 100 were identified on the Nucleotide Basic Local Alignment Search Tool (BLASTn at the National Center for Biotechnology Information-NCBI).Reference sequences were chosen based on the BLASTn results for reference mapping and extraction of consensus sequences for each segment [19].

Determination of RVA Genotypes
The genotype of each of the 11 genes for each strain was determined using the Virus Pathogenic database and analysis resource (ViPR) according to the guidelines proposed by the Rotavirus Classification Working Group [6,13].

Phylogenetic Analysis
Multiple sequence alignment of each gene was carried out using Multiple Sequence Comparison by Log Expectation (MUSCLE) alignment available in Molecular Evolutionary Genetic Analysis X (MEGA X) [24].

Nucleotide Sequence Accession Numbers
The nucleotide sequence data presented were deposited in GenBank under the following accession numbers: PP585813-PP586043 and PP848501-PP848786.
Viruses 2024, 16,1140 with the other 12 G9P [8] Mozambican strains (Table S3).This Mozambican strain cluste with a Japanese strain from 2016 (Figure 1).The tree was constructed based on the maximum likelihood method implemented in MEGA X [24], applying Tamura-3-parameter (T92+G+I) as the model.Bootstrap values (1000 replicates) ≥70% are shown with Wa-like strain serving as an out-group.The scale bar indicates genetic distance expressed as the number of nucleotide substitutions per site.G9P [4] Mozambican strains are indicated by blue circles, G9P [6] by red circles and G9P [8] by green circles.
All nine P [6] strains in combination with G9 clustered closely together in lineage I with the G2P [6] strains described in this study and with other G12P [6] and G2P [6] strains previously described from Mozambique and other Southern and Eastern African countries.The exception was RVA/Human-wt/MOZ/HCN1328/2016/G2P [6] that grouped separately from the rest of the study strains with G12P [6] Mozambican strains detected in 2012 (Figure 3).
Viruses 2024, 16, x FOR PEER REVIEW 9 of 1     nucleotide substitution model T92+G+I was used.The tree was constructed based on the maximum likelihood method implemented in MEGA X [24].Bootstrap values (1000 replicates) ≥70% are shown with Wa-like as out-group.The scale bar indicates genetic distance expressed as the number of nucleotide substitutions per site.G9P [4] Mozambican strains are indicated by blue circles, G9P [6] by red circles, G3P [4] by purple circles, G2P [4] by yellow circles and G2P [6] by brown circles.

Discussion
In the present study, WGS was performed for 47 strains with specific focus on strains identified as G9P [6], G9P [4] and G9P [8], obtained from Mozambican children with gastroenteritis between 2015 and 2018.
The report of G9P [8] strains as the most predominant genotype in the country before vaccine introduction and the first detection of G9P [4] and G9P [6] genotypes after the vaccine introduction in Mozambique [15], led to the need to monitor changes in strain diversity at gene level.Phylogenetic analysis of the G9 Mozambican strains showed that G9P [8] strains had a Wa-like constellation and the G9P [6] and G9P [4] strains, a DS-1-like constellation with the exception of strain RVA/Human-wt/MOZ/HCN1595/2017/G9P [4]

Discussion
In the present study, WGS was performed for 47 strains with specific focus on strains identified as G9P [6], G9P [4] and G9P [8], obtained from Mozambican children with gastroenteritis between 2015 and 2018.
The report of G9P [8] strains as the most predominant genotype in the country before vaccine introduction and the first detection of G9P [4] and G9P [6] genotypes after the vaccine introduction in Mozambique [15], led to the need to monitor changes in strain Viruses 2024, 16, 1140 14 of 17 diversity at gene level.Phylogenetic analysis of the G9 Mozambican strains showed that G9P [8] strains had a Wa-like constellation and the G9P [6] and G9P [4] strains, a DS-1-like constellation with the exception of strain RVA/Human-wt/MOZ/HCN1595/2017/G9P [4] which contained an E6 NSP4 gene.These results highlight the increase of the DS-1 backbone strains after vaccine introduction following the global trend at the time that the strains were detected [10,[34][35][36][37].
The 13 G9P [8] strains, detected before the introduction of the rotavirus vaccine, clustered together in lineage III.The P [8] lineage III is described as the most common globally [25].
Strain RVA/Human-wt/MOZ/HGJM0644/2015/G9P [8] clustered in a genetically distinct lineage known as lineage IV or OP354-like P [8].This lineage has been reported in different parts of Europe, Africa and Asia [25,27].In fact, the Mozambican P [8] lineage IV strain in this study was detected in a 9-month-old female child and could not be detected by RT-PCR, being non-typable for the P genotype.The same observation was reported in Ghana, where 10.4% of non-typeable rotavirus VP4 genes were identified as rare OP354-like P [8] by full-genome sequencing of this rare strain [27,38].
Interestingly, the RVA/Human-wt/MOZ/HGM1782/2017/G9P [6] strain clustered with the G3P [4] strains in segments encoding VP6, VP2, VP3 and NSP1.HGM1782 was detected in the Maputo province, in the same geographic location as the G3P [4] strains, which could explain the similar clustering for the four genome segments.The RVA/Humanwt/MOZ/HCN1598/2017/G9P [4] strain from the northern region of the country, clustered with G3P [4] Mozambican strains from the southern region of Mozambique, for the VP4 encoding gene.These results were confirmed in the Mvista analysis, where these strains were diverse in relation to the others, thus suggesting that they originated through reassortment events.Most of the G9P [4] study strains were detected in vaccinated children, unlike the G9P [6] study strains, which were mostly obtained from unvaccinated children.Regardless, no different clusters were observed between strains detected in vaccinated and unvaccinated children for both genotypes.
The segment encoding for VP6 had the most distinct clustering pattern among the strains.In this segment, seven G9P [6] strains had a higher genetic identity and formed a cluster with animal strains from South Africa and a human Mozambican strain, which had previously been reported as a mixed infection of animal origin [18].The antelope strain, which was similar to the G9P [6] study strains, was highlighted as having a common origin with the G6P [14] human strains in some segments, excluding the VP6 encoding gene [39].Three of these Mozambican strains were isolated from children who had contact with animals, including horses, sheep and cats.These results suggest that interspecies transmission occurred.
Unusual G9P [4] RVA strains have been reported in several countries, such as India, Italy, Japan, Benin and Ghana, during pre-and post-vaccine introduction periods [31,32,[34][35][36][37][40][41][42].It has been hypothesized that reassortment events among contemporary human rotavirus strains generated these unusual G9P [4] strains [43].The E6 NSP4 genotype was first identified in 2000 in India, and the analysis showed that the most common recent ancestor was likely to have been around 1981 in Asia [31].A recent study in Benin and Ghana described this genotype in Africa for the first time [34,40].The G9P [4] E6-NSP4 Mozambican strain was identified in the same period as the Benin and Ghanian strains; however, it was related to strains from India.Only one of the characterized strains exhibited the E6 genotype, which indicates a single inter-genotype reassortment event or a sporadic event in Mozambique [37].Further analyses are necessary to compare diarrhea severity and changes in the NSP4 protein; for example, an evaluation of the E6 genotype in an animal model could possibly determine if the observed chances are linked to increased pathogenesis of the genotype.In addition, continued genome surveillance is needed to monitor the occurrence of such unusual strains and the possibility of becoming predominant in the country causing severe acute gastroenteritis in children.
Limitations of the study include the short period analyzed (2015-2018) and the inability to calculate the Vesikari score to compare the severity of strains from the post-vaccine period.There is a need to expand the whole-genome analysis to strains detected after 2018 to fully comprehend the genetic diversity of rotavirus strains detected after vaccine introduction.

Conclusions
The study results indicate that G9P [4] and G9P [6] strains exhibited a DS-1-like genetic constellation after Rotarix ® vaccine introduction in Mozambique.The occurrence of unusual genotypes and close relationship with animal strains, suggesting inter-genotype reassortment and interspecies events, highlight the need for continuous genomic surveillance of RVA strains detected in Mozambique and the importance of following a One Health approach to identify and characterize potential zoonotic strains causing acute gastroenteritis in children.

Institutional Review Board Statement:
The study was conducted in accordance with the Declaration of Helsinki and approved by the National Ethics Committee of Mozambique (CNBS) under number IRB00002657, reference number 348/CNBS/13.Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
The data are available upon request from the corresponding author.

1 .
Figure 1.Phylogenetic tree based on the open reading frame (ORF) nucleotide sequence of the V G9 encoding gene of strains circulating in Mozambique compared to global strains obtained f GenBank.The tree was constructed based on the maximum likelihood method implemente MEGA X [24], applying Tamura-3-parameter (T92+G+I) as the model.Bootstrap values (1 replicates) ≥70% are shown with Wa-like strain serving as an out-group.The scale bar indic genetic distance expressed as the number of nucleotide substitutions per site.G9P[4] Mozamb

Figure 1 .
Figure 1.Phylogenetic tree based on the open reading frame (ORF) nucleotide sequence of the VP7-G9 encoding gene of strains circulating in Mozambique compared to global strains obtained from GenBank.

Figure 2 .
Figure 2. Phylogenetic tree based on the open reading frame (ORF) nucleotide sequence of the VP4 P[4] encoding gene of strains circulating in Mozambique compared to global strains obtained from GenBank.The tree was constructed based on the maximum likelihood method implemented i MEGA X[24], applying Tamura-3-parameter (T92+I) as the model.Bootstrap values (1000 replicates ≥70% are shown with Wa-like strain serving as an out-group.The scale bar indicates genetic distanc expressed as the number of nucleotide substitutions per site.G9P[4] Mozambican strains ar indicated by blue circles, G3P[4] by purple circles and G2P[4] by yellow circles.

Figure 2 .VP4-P[ 6 ]Figure 3 .
Figure 2. Phylogenetic tree based on the open reading frame (ORF) nucleotide sequence of the VP4-P[4] encoding gene of strains circulating in Mozambique compared to global strains obtained from GenBank.The tree was constructed based on the maximum likelihood method implemented in MEGA X[24], applying Tamura-3-parameter (T92+I) as the model.Bootstrap values (1000 replicates) ≥70% are shown with Wa-like strain serving as an out-group.The scale bar indicates genetic distance expressed as the number of nucleotide substitutions per site.G9P[4] Mozambican strains are indicated by blue circles, G3P[4] by purple circles and G2P[4] by yellow circles.

Figure 3 . 19 VP4-P[ 8 ]Figure 4 .
Figure3.Phylogenetic tree based on the ORF nucleotide sequence of the VP4-P[6] encoding gene of strains circulating in Mozambique compared to global strains obtained from GenBank.The tree was constructed based on the maximum likelihood method implemented in MEGA X[24], applying

Figure 4 . 19 VP6- I2 Figure 5 .
Figure 4. Phylogenetic tree based on the ORF nucleotide sequence of the VP4-encoding genes (P[8]) of strains circulating in Mozambique compared to global strains obtained from GenBank.The tree

Figure 5 .
Figure 5. Phylogenetic tree based on the ORF nucleotide sequence of the VP6-encoding I2 genes of strains circulating in Mozambique compared to global strains obtained from GenBank.The best-fit

19 Figure 6 .
Figure 6.Nucleotide sequence similarities of the G9P[4] and G9P[6] concatenated genomes using the RVA/Human-wt/MOZ/ HCN1347/2016/G9P[6] strain as reference.The name of the Mozambican strains is indicated on the left, and the positions of the 11 genes are indicated at the top.The scale indicates the distance in kb.

Figure 6 .
Figure 6.Nucleotide sequence similarities of the G9P[4] and G9P[6] concatenated genomes using the RVA/Human-wt/MOZ/HCN1347/2016/G9P[6] strain as reference.The name of the Mozambican strains is indicated on the left, and the positions of the 11 genes are indicated at the top.The scale indicates the distance in kb.