Genomic epidemiology and emergence of SARS-CoV-2 variants of concern in the United Arab Emirates

For the first time, this study demonstrates the entry of the new SARS-CoV-2 variants of concern and interests, and the outbreak dynamics in the UAE. Global massive ongoing transmission and the continuous evolution of new strains demonstrates that strict mitigation measures are important to effectively control the spread of the virus. To do so, a better understanding of the phylogenomic spread and transmission dynamics could contribute to more targeted and effective responses to the pandemic.

The analysis of 1274 viral genomes collected in the UAE, indicates the presence of 11 major clades. The occurrence of the root clades A and East Asian B was clearly seen in the early months of 2020 suggesting early spatiotemporal introduction into the UAE. Distribution of B.1 and B.1.1, which are descendants containing the spike mutation D614G, began in early May 2020 despite the vigilant health measures, which could suggest the multiple independent entry from Europe, Asia, and Middle East prior to the national lockdown . As the nation-wide public health measures were implemented, B.1.1 distributed locally until late July 2020. Despite lockdown and strict measures, we have observed a substantial local transmission within Abu Dhabi and Dubai, in addition to a low frequent spread of sub- pangolin clade lineages such as UK lineages (B.1.1.74, B.1.1.220, B.1.1.296, B.1.1.220 and B.1.1.190), European lineage (B.1.398 and B.1.1. 10) and UAE lineage (B.1.1.263) between June and July 2020.

In summer 2020, the expansion of B.1.177 and B.1 in western Europe was notable due to the presence of mutation in RBD (S:S477) which is crucial for ACE2 binding and antibody recognition^15,39. We reported the presence of these clades in UAE as the international flight resumed in late July 2020, in addition to B.1.1 and B.1.2, which were globally distributed. Strikingly, we observed the resurgence of A clade in early 2021, specifically Pango lineages A, A.20, A.28, A23.1, explaining the possible convergent spike mutations that could confer fitness advantages⁴⁴. We reported the first A.23.1 case in September 2020, and 32 cases in early 2021. Sub-lineages A.22 and A.28 can be explained by independent entry and reported only in home quarantine cases.

The spread of VOCs (B.1.1.7, B.1.351 & P.1) and VOIs (B.1.525) followed the gradual reopening of borders and worldwide travels. Our results indicate that the proportion of VOCs was greater in males than females, and significantly presented in patients aged > 48 years. Concordant to our results^45,46, B.1.1.7 (p < 0.001), B.1.351 (p < 0.001) and A.23.1 (p = 0.009) were associated with increased severity based on hospitalization rates. Deceased cases are more likely to occur when the patient was infected with B.1.351 (p < 0.001) and A.23.1(p = 0.022), whereas B.1.1.7 (p = 0.183) was not significantly associated with deceased cases. We reported the association between 10 out of 12 defining spike protein substitutions of B.1.1.7 and the hospitalized cases. Similarly, we reported the association between six defining mutations of B.1.351 and hospitalization. For A.23.1, four defining mutations were associated with hospitalized cases. The spread of B.1.351 and B.1.525 lineages have not been associated with hospitalization possibly due to the small size sample.

We have performed mutation analysis to define any significant correlation between patient severity and mutations resulting in amino acids sequence changes. A total of 37 structural and accessory proteins of SARS-CoV-2 are significantly associated with hospitalized COVID-19 cases after adjustment for age and gender, and Bonferroni correction. Overall, we have observed more mutation in the structural spike protein (n=13). We identified four major mutations of concerns in spike region that are associated with hospitalized cases in our study. N501Y that presents in B.1.1.7 and B.1.351 lineages has been reported to increase ACE2-binding affinity⁴⁷ and as a mean of immune escape³⁹. Other mutations such as A570D, D1118H, P681H, S982A, T716I, two deletions H69-V70 and Y144 in spike protein, in addition to D3L, S194L and S235F in nucleocapsid protein were found in B.1.1.7 lineage are in accordance with the studies indicating the high risk of hospital admission and severe disease in B.1.1.7 patients compared to wild-type variant^45,48. B.1.351 lineages mutations found in this study such as A701V were reported by Campbell et al.⁴⁹ to increase transmissibility by 25% and death in the hospitalized patients by 20%. Other mutations reported in spike and nucleocapsid regions (S: A243; N: M1X and S2Y) in this study have been associated with hospitalized cases, yet no studies have shown any association between these mutations and severity. The importance of ORF1a and ORF1b have been reported in viral replication, transcription, morphogenesis, and evasion of the host of the immune response. Concordant to our results, A1708D, I2230T, and T1001I mutations in ORF1a found in alpha lineages have been associated with hospital admission⁴⁵. The remaining mutations in ORF1ab (L730F, M372I, T350N, A1708D, I2230T, T1001I, K3353, F3677, G3676, L3667F, S3675, T239I, K1383R) have not been reported to correlate to the severity in other studies.

Other significant correlations were reported between hospitalized outcome and accessory proteins such as ORF8 and ORF9b. Although accessory proteins are not involved in virus replication, accumulating evidence demonstrated their critical roles in viral pathogenesis. Most mutations in accessory proteins were at ORF8 which were not identified in other studies. ORF8 was found to induce major histocompatibility class I (MHC1) down-regulation, thus providing protection against cytotoxic T cells (CTLs)⁵⁰. In addition, ORF8 expressing cell and SARS-CoV-2 infected cells are resistant to CLT lysis, which was restored with knockdown of ORF8 expression^50,51. It is suggested that SARS-CoV-2 could potentially benefit from missense mutations in ORF8 protein to evade immune surveillance⁵¹. We also identified K68, Q27, R52I, and Y73C mutation in ORF8, and R32P mutation in ORF9b in hospitalized patient. Mutations in ORF9b has been reported to interact with the mitochondria outer membrane protein (TOM70), thus suppresses interferon response⁴³.

Limitation of the study should be addressed. At the beginning of the pandemic, most patients (asymptomatic and symptomatic) were admitted to the hospital or quarantine areas which could not necessarily reflect the severity of the patient. Therefore, due to the complex nature of the COVID-19 phenotype presentation, statistical and methodological heterogeneity may be present. Also, the admission of patients may be influenced by other factors such as immediate status, comorbidities, and age. Second, the classification of ethnicity might be impression due to using nationality recorded from official passport as a surrogate for ethnicity. Epidemiological features such as travel-related, comorbidities, treatments and severe admission were limited in this study, which impacted post-hoc adjustment analysis. It is clearly noted that 95% of the patients were from the Emirate of Abu Dhabi which indicates the necessity of including further samples from other Emirates. Our mutation analysis may have sampling bias, since only 17% of patients were hospitalized, whereas the remaining were non-hospitalized.

Our study provides an epidemiological insight into the emergence of VOCs and VOIs following borders reopen and worldwide travels. It provides reassurance that hospitalization is markedly more associated with the presence of VOCs. The major strength of this study was the comprehensive longitudinal analysis which covered the early months of COVID-19 in UAE, until the peak of the 3rd wave in February 2021. However, the collection of good quality data such as vaccine status, severity, and travel history in combination with rapid genome sequence are imperative in understanding the behavior and role of variants related to clinical outcomes. This study can contribute to understanding the global transmission of SARS-CoV-2 variants.