3.6. The effect of mutations to the encoded proteins
Some common mutations in Delta and Omicron variants were analyzed to examine their thermodynamic effect on the Spike protein (Supplementary Table 6). Mutations in the Spike gene of Delta variant, including T95I, G142D, A222V, Q613H, and D614G, demonstrated a stabilizng effect on the structure of the Spike protein. We then analyzed common mutations in the Spike gene of Omicron variant. We found that these mutations have either a stablizing and destabilizing effect on the structure of the Spike protein. In the nsp12 gene, amino acid substitutions P323L found in both Delta and Omicron variants with a frequency of 100% has a stabilizing effect on the RdRp protein.
In the nsp3 gene, we analyzed three mutations (G28R, P77L, and L120I) in Delta variant, which all have a stabilizng effect on the PLpro protein. In Omicron variant, T259I mutation has destabilizng effect. For the nsp5 gene, we analyzed L75F (Delta) as well as P132H and P241L (Omicron) mutations, which all showed a stabilizing effect on the 3CLpro protein.
Discussion
Our current study showed that by analyzing the Spike, nsp12, nsp3, and nsp5 genes, we clearly demonstrated a distinct evolutionary pattern of SARS-CoV-2 Delta and Omicron variants circulating in Yogyakarta and Central Java provinces, Indonesia, during May 2021 to February 2022. In Indonesia, the Omicron upsurge was occurred from late January until February 2022 and subsequently replaced the Delta variant.45 We have previously shown that Delta- and Omicron-infected patients had similar hospitalization and mortality rates.36 Our current study emphasized the need of continued and extensive SARS-CoV-2 sequencing surveillance as the primary method to quickly detect and respond the emergence of new variants.
A latest study showed that compared to the previous VOCs (Alpha, Beta, Gamma, and Delta), Omicron had the highest enrichment of amino acid substitutions within the Spike gene.46 Indeed, new strains with higher transmissibility and infectivity have emerged due to mutations in the Spike protein.2 Alterations of the amino acid in the Spike protein influenced the binding affinity and the viral fusion process.47 Our results showed that D614G mutation occurred in all (100%) Delta and Omicron variants identified in this study, indicating that it was fixed in the viral population. Indeed, all VOCs were generated from the G614 variant lineage. Three mutations always accompany the D614G variant: a C-to-T in the 5’ UTR; a silent C-to-T mutation at position 3.037; and a C-to-T mutation at position 14,408 that results in an amino acid change P323L in the nsp12 gene.48 The other mutations at position 477 (S477G, S477N, and S477R) of the Spike protein were prominent among monoclonal antibody (mAb)-escape mutation.2 Our results showed that S477N mutation was found only in the Omicron variant (46.8%). A computational analysis showed that N477 had an increased binding affinity to ACE2.49 Other studies reported that T19R, E156G, L452R, T478K, and P681R variants in the Spike protein exhibited a stabilizing effect on protein structure, facilitating the binding affinity for more stable interactions with the human ACE2. This stability effect may increase the transmission of the virus in human populations.50 Our studies found T19R, L452R, T478K, and P618R mutations in Delta with a high frequency above 70% and L452R in Omicron with 46.8% frequency.
The nsp12-P323L mutation first appeared in January 2020 and became the predominant globally (>90%) by late April. Other mutations were found in a low frequency (3.5-4.0%), including E254D, A423V, A656S, V720I, and V776L.51 It has been shown that P323L mutation may increase the mutation rate of SARS-CoV-218 and was associated with Covid-19 severity.52 Molecular dynamic simulations showed that P323L mutation led to tighter binding with antiviral drug remdesivir (RDV).53 Interestingly, it has been shown that the Spike D614G and RdRp P323L (G/L variant) have co-evolved to become more superior than the original D/P variant.54 In our study, the P323L mutation is similarly conserved between Delta and Omicron variants, indicating its beneficial effect for viral evolution.
An early study analyzing SARS-CoV-2 isolates circulating in Indonesia until September 2020 revealed mutations in the PLpro (P77L and V205I) and 3CLpro (M49I and L50F) genes.34 However, we did not find these mutations in our current study, suggesting that these mutations were not fixed in the viral population circulating in our region. Within the 3CLpro gene, several unique (signature) mutations are identified, which are different across SARS-CoV-2 lineages. P132H, K90R, and G15S are prevalent mutations found in Omicron (B.1.1.529), Beta (B.1.351), and Lambda (C.37) variants, respectively. In contrast, Delta variant had no unique mutations within this gene.55 Consistently, our finding showed that all Omicron variant circulating in our region harboring P132H mutation, while Delta variant had no prevalent mutation identified. One mutation (L75F) found in Delta variant was very low in frequency (1.4%). Notably, P132H mutation did not lead to nirmatrelvir resistance.55
Several mutations have been identified to confer resistance to nirmatrelvir, including Y54C, L167F, and E166V.26,56These mutations were not identified in our isolates. Interestingly,in vitro studies combining Y54C and L167F with P132H resulted in a functional protein.56 This finding provides an insight that additional mutations in the 3CLpro gene that may emerge in the Omicron variant may be clinically relevant in the future. In addition, mutations in PLpro and 3CLpro genes were shown to be associated with the clinical course of Covid-19 patients. P108S mutation in the 3CLpro gene resulted in reduced activity of the protease protein and was associated with milder clinical course.57 In the PLpro gene, it has been shown that P78L and K233Q mutations were associated with increased risk of death.58 Interestingly, our current study found the synonymous mutation F106F in the PLpro gene both in Delta (97.7%) and Omicron (99.3%) variants.
It has been estimated that the background nucleotide substitution rate of SARS-CoV-2 was about 1.1 x 10-3subs/site/year.59 The emergence of VOCs can be ascribed to a 4-fold increase of the substitution rate above its background level that may have lasted for several weeks or months.60 In addition to their defining (unique) mutations, each VOC may have a distinct evolutionary rate.60 A previous study analyzing the phylodynamic of Delta and Omicron variants also showed that a mean rate of nucleotide substitution was higher in Omicron compared to Delta variants, i.e. 3.898 x 10−3 subs/site/year (range: 2.686 x 10− 3 to 5.102 x 10− 3)  and 3.677 × 10− 4 subs/site/year (range: 1.311 x 10− 4 to 6.144 x 10− 4 ) for the Omicron and Delta variants, respectively.61
Alterations in amino acids that decrease viral fitness are often eliminated by negative selection, while changes that improve viral fitness are retained by positive selection. On the other hand, amino-acid alterations are regarded as “neutral” when they have no effects on viral fitness. Because the existence of negative or positive selection suggests that specific sites are functionally significant, it is crucial to determine which sites evolve under selective pressure, especially in the case of novel emerging pathogens such as SARS-CoV-2. In this study, several positively selected sites in the Spike, nsp12, and nsp3 genes were identified in Delta variant. However, in Omicron variant, we identified positively selected sites only in theSpike gene. This finding may imply a distinct genetic “hot-spot” in SARS-CoV-2 tropism, replication, and infectivity, and need to be further investigated.
Conclusion
During the study period, we identified 213 and 139 isolates of Omicron and Delta variants, respectively, co-circulating in our region. Particularly in the Spike and nsp5 genes, high frequency amino acid substitutions were significantly more abundant in Omicron than in Delta variants. Consistently, in all of four genes studied, the substitution rates of Omicron were higher than that of Delta variants, especially in the Spike and nsp12 genes. In addition, selective pressure analysis revealed several sites in particular genes that were positively selected, implying that these sites were functionally essential for virus evolution, including 95 (T), 142 (G), 222 (A), 452 (L), 614 (D), 1264 (V) in Delta and 213 (V), 339 (G), 375 (S), 417 (K), and 440 (N) in the Spike gene of Omicron variants. Our study demonstrated a distinct evolutionary pattern of SARS-CoV-2 Delta and Omicron variants circulating in Yogyakarta and Central Java provinces, Indonesia. Thus, our study emphasized the need of continued and extensive SARS-CoV-2 genomic surveillance to quickly detect and respond the emergence of new variants.