Emergence and spread of a B.1.1.28-derived lineage with Q675H and Q677H Spike mutations in Uruguay

Uruguay was able to control the viral dissemination during the first nine months of the SARS-CoV-2 pandemic. Unfortunately, towards the end of 2020, the number of daily new cases exponentially increased. We previously identified a B.1.1.28 sublineage carrying mutations Q675H+Q677H in the viral Spike, with local transmission in Rocha, a department bordering Brazil. To understand whether these B.1.1.28+Q675H+Q677H sequences were part of an emergent SARS-CoV-2 lineage broadly disseminated in Uruguay, herein we analyzed the country-wide genetic diversity of viruses between November, 2020 and April, 2021. Our findings support that B.1.1.28+Q675H+Q677H probably arose around November 2020, in Montevideo, Uruguay's capital department. This clade spread to other Uruguayan departments, with evidence of further local transmission clusters. It also spread to the USA and Spain. The Q675H and Q677H mutations are in the proximity of the polybasic cleavage site at the S1/S2 boundary and also arose independently in many SARS-CoV-2 lineages circulating worldwide. Although in Uruguay the B.1.1.28+Q675H+Q677H lineage was dominated by the VOC P.1 since April 2021, the monitoring of the concurrent emergence of Q675H+Q677H in VOIs and/or VOCs should be of worldwide interest.


Introduction
Uruguay was able to control the early viral dissemination during the first nine months of the SARS-CoV-2 pandemic. Unfortunately, towards the end of 2020, the number of newly daily cases exponentially increased, from 60 cases per day on average during October and November to over 400 during December [1,2]. This coincided with the loss of the TETRIS (Test, Trace and Isolation strategy) safety zone [3,4]. With a 1,068 km long Uruguayan-Brazilian dry border, multiple introductions and successful dissemination of Brazilian lineages B.1.1.28, B.1.1.33, P.1 (Gamma) and P.2 have occurred during 2020 and 2021 [5][6][7]. Little is known, however, regarding the SARS-CoV-2 strains that circulated during the first exponential increase of COVID-19 cases by the end of 2020, before the arrival of the VOI P.2 and the VOC P.1 (which earliest samples are from January and February 2021, respectively) [6,7].
In a previous study we surveyed patients diagnosed between November 2020 and February 2021 in Rocha, an eastern Uruguayan department bordering Brazil, where we found that lineage B.1.1.28 was the most prevalent during November-December, 2020 in that location [7]. Many B.1.1.28 sequences branched in a clade highly prevalent in southern Brazil, while others branched in a second monophyletic cluster harboring several lineage-defining mutations including two non-synonymous changes in the Spike protein: Q675H and Q677H, so far not concurrently reported. The convergent appearance of S:Q677H in different viral lineages and its proximity to the S1/S2 cleavage site raised concerns about its functional relevance [8,9]. To understand the SARS-CoV-2 strains associated with the first COVID-19 epidemic wave in Uruguay, herein we analyzed the genetic diversity of viruses circulating in different country localities by the end of 2020 and first months of 2021.

Ethics statement
This work was done by the Inter-Institutional Working Group (IiWG) for SARS-CoV-2 genomic surveillance in Uruguay, which involves a diagnostic network, expertise and resources to handle large-scale sequencing, computational scientists for genomic analysis, and an affordable and decentralized "in house" qPCR test designed to detect known VOCs [6]. Residual de-identified RNA samples from SARS-CoV-2 positive patients were remitted to the Institut Pasteur de Montevideo (IPMon). IPMon was validated by the Ministry of Health of Uruguay as an approved center providing diagnostic testing for COVID-19. All samples were de-identified before receipt by the study investigators. All relevant ethical guidelines have been appropriately followed. Additionally, the project was approved by the Ethics Committee of the Sanatorio Americano SASA (Uruguay). Ethical approval was given and signed informed consent was obtained from the participants. SARS-CoV-2 samples SARS-CoV-2 RNA samples were recovered from nasopharyngeal-throat combined swabs collected from clinically ill or asymptomatic individuals that reside in different Uruguayan departments and were diagnosed from December 2020 to April 2021 in Uruguay. While our target period of study is November 2020 to April 2021, residual patient samples from November were not available at any laboratory of the IiGW diagnostic network (the IiWG started working as such in March 2021). Positive RNA samples were reverse transcribed using SuperScript™ II Reverse Transcriptase (Thermo Fisher Scientific Inc., MA, USA) or the LunaScript® RT SuperMix Kit (New England Biolabs, Ipswich, MA, USA). A negative control was included at this point and carried throughout the protocol. Samples were initially screened by the VOC-qPCR assay, and those VOC-negative were further processed for whole-genome sequencing.
Genome sequencing Sequencing libraries were prepared according to the classic ARTIC protocol described by Quick J. [10,11] or the SARS-CoV2 genome sequencing protocol (1200bp amplicon "midnight" primer set, using Nanopore Rapid kit) described by Freed N. and Silander O. [12,13] with minor modifications (Table S1). Final library was eluted in EB buffer (ONT) and quantified using a fluorometric assay. Recommended amounts of library were loaded into a FLO-MIN106D R9.4.1 flowcell and sequenced on the GridION X5 sequencing platform (ONT). Basecalling and demultiplexing was performed with Guppy 4.3.2 or higher [14] using the high or super accuracy mode. Consensus genomes were generated using the poreCov pipeline [15][16][17][18][19][20][21][22][23][24] and Nanopolish was used for consensus generation. Complete sequences with up to 15% of Ns were kept for further analysis. All genomes obtained in this study were uploaded to the EpiCoV database in the GISAID initiative under the accession numbers EPI_ISL_XX to EPI_ISL_XX (Table S1).
Phylogenetic and phylogeographic analysis SARS-CoV-2 full-length consensus sequences were manually curated in specific genome positions, such as clade defining mutations. Genotyping was performed according to Rambaut et al. [25] using the Pangolin application [26] and later confirmed using maximum likelihood (ML) analysis. Uruguayan B.1.1.28 sequences (n=174) were next analyzed in the context of additional B.1.1.28 sequences from Uruguay and Brazil, downloaded from the EpiCoV database of the GISAID initiative [27] (Table S3 with GISAID Acknowledgements). Downloaded B.1.1.28 sequences from Uruguay (n=184) were complete, with full collection date information and sampled before May 31, 2021. Sequences from Brazil (n=1,428) were complete, high quality, with full collection date information and sampled before May 31, 2021 (Table S2). Additionally, we downloaded four B.1.1.28 from USA (n=2), Spain (n=1) and Belgium (n=1) which also harbor both S:Q675H and S:Q677H mutations (Table S2 and Table  S3). Finally, we added to the B.1.1.28 dataset six available sequences collected by the IiWG out of the target period of this work (Table S1). Alignment was performed with MAFFT v7.471 [28]. Maximum likelihood (ML) phylogenetic analysis of the 1,796 B.1.1.28 sequences was performed with IQ-TREE version 1.6.12 under the model GTR+F+R3 of nucleotide substitution selected by the built-in ModelFinder option [29]. Branch support was assessed by the approximate likelihood-ratio test based on a Shimodaira-Hasegawa-like procedure (SH-aLRT) with 1,000 replicates [30]. The tree root was established with the sequence EPI_ISL_416036 with the earliest collection date 2020-03-05. This tree was time-scaled using TreeTime [31] applying a fixed clock rate of 8x10 -4 substitutions/site/year [32,33], keeping polytomies. The time-scaled tree was then employed for the ancestral character state reconstruction (ACR) of epidemic locations with PastML [34], using the Marginal Posterior Probabilites Approximation (MPPA) method with an F81-like model. Brazilian sequences were grouped according to the region: South, Southeast, Central West, North and Northeast. A time-scaled Bayesian phylogeographic analysis was next performed to infer the geographical source and dissemination pattern of the Uruguayan B.1.1.28+Q675H+Q677H samples and to estimate the time of their most recent common ancestors (TMRCA). Phylogenetic trees were estimated in BEAST v1.10 [35] using the GTR+F+I nucleotide substitution model, the nonparametric Bayesian skyline model as the coalescent tree prior [36], a strict molecular clock model with a uniform substitution rate prior (8-10 x10 -4 substitutions/site/year) and a reversible discrete phylogeographic model (using Uruguayan departments as epidemic locations) [37] with a continuous-time Markov chain (CTMC) rate reference prior [38]. MCMC chains were run for 100 million generations and convergence (Effective Sample Size > 200) in parameter estimates was assessed using Tracer v1.7 [39]. Maximum clade credibility (MCC) tree was summarized with TreeAnnotator v1.10 [40] and visualized using FigTree v1.4.4 [41]. Additional visualizations were implemented in the R environment with treeio and ggtree Bioconductor packages [42].
Determination of prevalence of Q675H+Q677H To assess the prevalence of co-occurring S:Q675H and S:Q677H in worldwide SARS-CoV-2 genomes, we downloaded from GISAID (accessed on July 7, 2021) 129 complete genomes, with high quality and full collection date information (Table S2-S3).
Structural representation of the Spike protein The molecular model of SARS-CoV-2 spike glycoprotein was taken from D. E. Shaw Research database (DESRES-ANTON-11021566) [43] . The visual rendering was done with VMD 1.9.3 [44].

Results
The Uruguayan inter-institutional working group (IiWG) for SARS-CoV-2 genomic surveillance sequenced 567 SARS-CoV-2 positive samples detected in Uruguay between December 2020 and April 2021, that were classified in the following lineages: 180 B.1.1.28 (Table S1) Figure 1A). Analysis of all SARS-CoV Uruguayan sequences determined by the IiWG between November 2020 and April 2021 plus additional Uruguayan sequences obtained from GISAID database revealed that the relative frequency of B.1.1.28+Q675H+Q677H sequences increase from 26% in December 2020 up to 76% in February 2021 ( Figure 1B) and was later replaced by the VOC P.1 as the most prevalent lineage through the country as previously described [6].  Figure 1C and Figure S1). Fourteen introductions resulted in singletons or dyads with no evidence of further dissemination. Six introductions resulted in highly supported clades (SH-aLRT>0.79) with at least three Uruguayan samples. All 314 Uruguayan sequences carrying mutations Q675H+Q677H branched in a highly supported (SH-aLRT>0.98) monophyletic clade, now designated as the new P.6 pango lineage [25]. This clade also comprises two sequences collected in the USA and one sequence sampled in Spain. According to the ML-based ACR of epidemic locations, the lineage P.6 was most likely introduced from the southeastern Brazilian region (ACR location marginal probability = 0.99) and was disseminated from Uruguay to the USA (two independent times, ACR location marginal probability = 0.99 and 1.00, respectively) and Spain (ACR location marginal probability = 0.99) (Figure 1C and Figure S2). The second largest monophyletic clade designated as UY-N:P13L (Figure 1C and Figure S1) is composed of 27 Uruguayan sequences and five Brazilian ones and is part of a B.1.1.28 clade characterized by the amino acid change P13L in the Nucleocapsid viral protein that is widely distributed in southern Brazil and already detected circulating in Rocha department by the end of 2020 [7,45,46]. We next performed a time-scaled phylogeographic analysis of all lineage P.6 Uruguayan sequences produced in this work (n=163) and six Brazilian basal sequences. The spatiotemporal reconstruction suggests that lineage P.6 probably arose around November 9, 2020 (HPD: October 20, 2020 to November 26, 2020) in Montevideo, the capital department of Uruguay (Figure 2A). Lineage P.6 spread from Montevideo to other Uruguayan departments, with evidence of local transmission clusters in Rocha, Salto, San Jose and Tacuarembo. The TMRCA of local transmission clusters in Rocha and Salto was traced to December 23, 2020 (HPD: December 12, 2020 to December 30, 2020) and December 30, 2020 (HPD: December 19, 2020 to January 09, 2021), respectively (Figure 2A-B). The introduction and dispersion of this lineage in each department coincided with the increase in new COVID-19 cases reported daily (Figure 2A).
This new lineage P.6 spreading in Uruguay is characterized by 10 lineage-defining genetic changes, including five non-synonymous mutations, two of them in the Spike viral protein (Figure 2B). The amino acid change S:Q677H has been reported as a recurrent mutation arising independently in many SARS-CoV-2 lineages circulating worldwide by the end of 2020 [8,9]. In fact, a search in the EpiCoV database (accessed on July 7, 2021) reported 85 non-Uruguayan samples with high quality SARS-CoV-2 genome sequences with both Q675H and Q677H present. The pair Q675H+Q677H appeared distributed in 12 different countries (in descendent frequency order: Uruguay, England, USA, Belgium, India, Australia, Switzerland, Spain, Netherlands, Japan, Germany and France) and classified as 13 different pango lineages (in descendent frequency order: B. 1.1.28, B.1.36, B.1.2, C.36, B.1.538, B.1.1.316,  B.1.526, B.1.525, B.1.243, B.1.1.70, B.1.1.7, B.1.1.63 and B.1) (Figure S3). Structural analysis of the SARS-CoV-2 Spike glycoprotein shows that residues Q675H and Q677H are within the subdomain SD2 of each protomer constituting the homotrimer (Figure 2C). The non-synonymous mutations Q675H and Q677H lay at the beginning of a very flexible loop (residues 675-690) [47], which embraces the solvent accessible polybasic furin cleavage site (Figure 2C, inset). These mutations are close to two experimentally observed O-glycosylation sites at T676 and T678 [48,49] and at the same domain of D614G mutation [50]. Particularly, D614G mutation is known to impact structural and thermodynamic aspects of the spike [51][52][53][54], and to enhance the protease cleavage [55], likely by allosterically increasing the binding to furin [56]. Recently, the increase in protease-mediated cleavage and the promotion of syncytium formation and virus infectivity has been shown for G614 viruses [57]. Knowing that histidine residues function as pH-sensors in other viruses [58], Q675H and Q677H mutations might also provide some synergic structural changes in the dynamics of the subdomain SD2. All in all, both Q675H and Q677H amino acid changes might trigger some effects on viral transmissibility by enhancing D614G effects.

Discussion
Since March 2020, Uruguay had been successful at keeping the COVID-19 pandemic at check. Closed international borders and an aggressive contact-tracing system, among other government measures, were able to avoid virus transmission growing exponentially [1,2]. A few outbreaks occurred in departments bordering Brazil or in the southern departments, but related to those happening close to the Brazilian frontier [5]. Brazil has been a COVID-19 hotspot in South America and the 1,068 km long Uruguayan-Brazilian dry border allowed the rapid local establishment of SARS-CoV-2 Brazilian lineages B.1.1.28 and B.1.1.33. However, by the end of 2020, the pandemics worsened with a clear increase of daily cases in December. Summer-related social gatherings and relaxed vigilance are some of the proposed reasons to explain the epidemic growth that led to the loss of the TETRIS strategy [2]. However, we suggest that the emergence and spread of a B.1.1.28-derived lineage might have been involved.
In this study, we described a new B.1.1.28 sub-lineage designated P.6 that arose in Montevideo by November 2020 and spread throughout the country. Lineage P.6 comprises most Uruguayan virus genomes sequenced in the late 2020 and early 2021 and its emergence and local spread coincided with the national and local (as shown for Rocha and Salto departments) increase in daily SARS-CoV-2 cases. Lineage dispersion and prevalence changes can occur purely by chance and human behaviour, as was demonstrated for variant 20E(EU1) which emerged in Spain and spread through Europe in the boreal summer of 2020 [59]. While the human behaviour in Uruguay could have changed by the end of the austral spring and there was an ambiance of relaxed restrictions, we consider that the amino acid changes Q675H and Q677H in the Spike protein might alter SARS-CoV-2 fitness, contributing All rights reserved. No reuse allowed without permission.
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
We have not conducted any experimental study on the effect on the viral fitness produced by these Q675H and Q677H Spike mutations and we are not aware of any experimental assay that assessed them. However, independent data might indicate that these amino acid changes (either one or both of them) could facilitate viral transmissibility. First, as explained above, they are in close proximity to the polybasic cleavage site at the S1/S2 boundary and the mutations are consistent with its potential functional relevance during cell entry. Second, a recent study that developed an innovative model on epidemiological variables integrating the effect of Spike amino acid changes in viral fitness, forecasted that both Q675H and Q677H could appear in emerging SARS-CoV-2 VOCs in the following months [60]. Third, and most importantly, convergent evolution is a hallmark of positive selection, and we identified the independent appearance of both Q675H and Q677H in 12 additional SARS-CoV-2 lineages. Moreover, Q677H alone has also been identified before as a recurrently appearing mutation in different lineages [8,9].
In summary, this study describes the emergence and local spread of lineage P.6, a new B.1.1.28-derived lineage carrying Spike mutations Q675H+Q677H, in Uruguay that coincided with the first exponential growth phase of the COVID-19 epidemic in the country that started by November, 2020 and lasted until mid-February 2021. The ancestral B.1.1.28 virus carrying mutation Q675H was probably introduced from southeastern Brazil into Montevideo, Uruguay's capital city, by November 2020 and this virus rapidly fixed mutation Q677H and spread across the whole country. We propose that lineage P.6 has some altered biological properties which contributed to its swift dissemination in Uruguay and to the first COVID-19 wave that the country suffered by the end of 2020 and beginning of 2021. Although the lineage P.6 was rapidly substituted by the VOC P.1 as the most prevalent lineage in Uruguay since April 2021 [6], the monitoring of the concurrent emergence of Spike mutations Q675H+Q677H in VOIs and/or VOCs should be of worldwide interest.

Acknowledgments
The National Ministry of Health (Uruguay) is the main health Institution in our country. It is a dedicated ethics oversight body and has granted us the ethical approval for this work. All necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived. The authors wish to thank all the health care workers and scientists who have worked hard to deal with this pandemic threat, the GISAID team, and all the EpiCoV database's submitters (GISAID acknowledgment table containing the sequences used in this study is in supplementary Table S3). This work was supported by FOCEM-Fondo para la Convergencia Estructural del Mercosur (COF03/11). We thank Christian Brandt from Institute for Infectious Diseases and Infection Control (Jena University Hospital), for adapting their poreCov Nextflow pipeline to our requirements and quickly fixing reported bugs.  Tacuarembó  (orange). Daily new cases for Rocha, Salto and Tacuarembó were multiplied by a factor proportional to the population of that department in comparison to Montevideo (times 24,13 and 16 for Rocha, Salto and Tacuarembó, respectively) for visualization purposes. Confidence intervals of TMRCA of Montevideo (fuchsia), Rocha (pink), Salto (green) and Tacuarembó (orange) clades are shown as shaded areas. B. Bayesian phylogeographic analysis of P.6 clade in Uruguay, implemented in BEAST. Uruguayan sequences generated by the IiWG (n=163) were combined with six additional basal sequences from southeastern Brazil. Tips and branches of the time-scaled Bayesian tree are colored according to sampling location and the most probable location state of their descendent nodes, respectively, as indicated in the legend. Posterior probability support values and estimated TMRCAs are indicated at key nodes. Additionally, a heatmap represents the presence or absence of synapomorphic sites. The color scheme indicates the different mutations, as indicated in the legend. In each case, genomic position, nucleotide substitution, viral protein and amino acid is shown. C. One Spike protomer of the homotrimer is shown in thick ribbons, while the others are represented as transparent solvent accessible surfaces. The subdomain SD2 (residues 590-700) is indicated in blue, and the polybasic furin cleavage site (PRRAR, residues 681-685) is in red. The inset shows a zoom into the structural context of SD2, representing residues D614, Q675 and Q677 in balls and sticks, and nearby residues in sticks.   (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted July 31, 2021. ; https://doi.org/10.1101/2021.07.27.21261150 doi: medRxiv preprint sequences were obtained from GISAID on July 7, 2021. B.1.1.28 corresponds to the newly designated P.6 pango lineage. Table S1. Information on the B.1.1.28 samples from Uruguay used in this study. # indicate 11 samples already analyzed in Rego et al. [7], ^ show six samples here sequenced although they were not collected in our target period (November 2020 -April 2021) and * shows the samples processed using the "midnight" primer set, using Nanopore Rapid kit. Table S2. Information about Uruguayan, Brazilian and worldwide samples obtained from GISAID (accessed on July 7, 2021). Table S3. GISAID Acknowledgements Table.