Introduction
Despite the global impact of the COVID-19 pandemic, the exact protein-coding capacity of the SARS-CoV-2 genome remained a subject of debate during the early stages of the crisis. The virus utilizes a compact RNA genome where genes often overlap in different reading frames, making it difficult to distinguish functional proteins from non-functional genetic noise. Resolving this "parts list" is critical for understanding viral biology and interpreting the functional significance of emerging mutations. This study employs comparative genomics to provide a high-confidence gene set and characterize the evolutionary pressures shaping the virus.
Methods and Datasets
To distinguish functional genes from accidental open reading frames (ORFs), researchers analyzed 44 complete Sarbecovirus genomes, including SARS-CoV-2, SARS-CoV, and 42 related bat-infecting strains. This evolutionary distance provided a "Goldilocks" window—diverse enough to show selective signals but conserved enough for accurate alignment.
The primary analytical tool used was PhyloCSF (Phylogenetic Codon Substitution Frequencies), which identifies protein-coding signatures by quantifying the preference for synonymous mutations over non-synonymous ones. Additionally, the FRESCo tool was used to detect "synonymous constraint," identifying regions where the RNA sequence is conserved for secondary functions, such as structural elements or overlapping coding frames. These findings were cross-referenced with 1,875 single-nucleotide variants (SNVs) from early pandemic isolates to compare long-term evolutionary trends with real-time adaptation.
The High-Confidence Gene Set
By applying these rigorous evolutionary metrics, the study finalized a set of 13 protein-coding genes: 1a, 1ab, S, 3a, 3c, E, M, 6, 7a, 7b, 8, N, and 9b.
Key findings regarding the gene set include:
- De-annotation of Suspected Genes: Several ORFs frequently cited in early literature, including ORF10, ORF9c, 3b, and 3d, were found to lack protein-coding signatures. ORF10, in particular, appears to be a highly conserved RNA pseudoknot structure involved in replication rather than a protein-coding gene.
- Confirmation of ORF3c: The study confirmed ORF3c as a functional, novel gene that overlaps with ORF3a. Its conservation across the Sarbecovirus lineage suggests it plays a vital role in the viral life cycle.
- Exhaustive Search: The analysis concluded that no other protein-coding genes remain to be discovered in the SARS-CoV-2 genome.
Evolutionary Dynamics and Pandemic Mutations
The study revealed a strong correlation (0.70) between long-term evolutionary constraints and the mutations observed during the pandemic. However, the deviations from this trend provide the most significant insights into how the virus is adapting to human hosts.
Adaptation and Immune Avoidance
One of the most notable findings involves the Spike D614G mutation. This variant disrupts an amino acid residue that has been perfectly conserved across Sarbecoviruses for millions of years, signaling a specific adaptation to optimize human-to-human transmission.
Similarly, the Nucleocapsid (N) protein exhibited significantly higher mutation rates (21%) than predicted by its historical conservation (11%). A specific 20-amino-acid hotspot within the N protein, overlapping with a predicted B-cell epitope, showed a cluster of mutations. This suggests the virus is under intense selective pressure to evade the human immune system.
Evolutionary Deceleration
In contrast, the S1 subunit of the Spike protein and nsp3 showed fewer mutations than expected. This "deceleration" might indicate that these regions reached a fitness peak early in the pandemic or adapted to humans prior to the first widely tracked outbreaks.
Conclusion
By resolving the SARS-CoV-2 gene set and mapping evolutionary constraints, this research provides a foundational resource for COVID-19 biology. The identification of 1,394 synonymously-constrained codons highlights regions where the RNA structure itself is vital, offering new targets for therapeutic intervention. Furthermore, the ability to distinguish between routine genetic drift and adaptive "rule-breaking" mutations allows researchers to better prioritize variants of concern as the virus continues to evolve.
Summary updated