AlphaFold2 Reveals Structural Patterns of Seasonal Haplotype Diversification in SARS-CoV-2 Nucleocapsid Protein Variants

“Structural alignments and visualizations were carried out using Chimera [45].”

“The CTD region, however, had lower GDT_TS and TM scores than the NTD but remained slightly below 70 and ~0.74, respectively, which are still adequate for use in research as both these scores indicate accurate topologies but less confidence with side-chain conformations.”

“Matplotlib [50], numpy [51], and pandas [52] in python were used to process and visualize the data retrieved from the tools used in this study.”

“Matplotlib [50], numpy [51], and pandas [52] in python were used to process and visualize the data retrieved from the tools used in this study.”

“The virus has a 79.6% sequence similarity to SARS-CoV-1, the causative agent of the first SARS outbreak [13].”

“Using an RMSD threshold of 3 Å (a standard for protein comparison and homologous proteins [56,57]), we observed that the N-protein exhibited li tt le change for each haplotype and VOC, aside from a ff ecting the four target regions.”

“Topological similarities of individual regions or entire molecules were evaluated with average template modeling scores (TM scores) using US-align [46,47].”

“The Coronavirus Disease 19 ( COVID-19 COVID-19) pandemic has led to a total of 775 , Viruses 2024 , 16 , 1358 2 of 20 experienced 641 COVID-19 infections out of 3711 passengers, with over half (328) of those infected being asymptomatic [11].”

“To quantify protein flexibility and other dynamic features across the pandemic, the Bio3D [48] package in R [49] was used to obtain fluctuation values which were then visualized in python.”

“Accuracy was measured with the predicted local-distance difference test (pLDDT) and the predicted aligned error (PAE). pLDDT provides a per-residue estimate of prediction confidence based on the LDDT-C α metric [44].”

“This is achieved through RNA-binding at various locations of the N-protein, which spans the N-arm [75], NTD [69], LKR [75], and CTD [20,76,77] regions.”

“AlphaFold2, the pLDDT IDRs are known to exhibit lower pLDDT values. score, we showed that these regions had lower In fact, AlphaFold2 now uses pLDDT scores to predict regions of disorder [60]. values (Figure 4).”

“…comprises five distinct regions, two of which are structured and are known as the N-Terminal Domain (NTD) (residues 44– 174 ) and the C-Terminal C-Terminal Domain ( CTD ) ( residues Variants of Concern (VOCs) have been replacing each other since the beginning of the COVID-19 pandemic [28].”

“The AlphaFold2 model of the NTD also remained under an RMSD of 1.5 Å for more than 96% of its structure (126/131 residues) as reported by the AS2S server.”

“In fact, known RNA-binding elements [22] traced on the surface In fact, known RNA-binding elements [22] traced on the surface of the molecule spanning the ‘basic finger’ of the NTD domain and the C-tail of the molecule (Figure 7b) matched the most salient electropositive regions of the molecular…”

“AlphaFold2 [31], the star of the last two biannual ab initio structure prediction experiments (CASP 14 and 15) [32,33], greatly broke records as its neural network (called the evoformer ) approach to modeling proteins without homologs was within the margin of error of experimental structure…”

“This information will be especially valuable for therapeutic interventions and predictive intelligence applications, which could facilitate understanding of protein reformation, de novo protein and construct design, and assembly of molecular complexes.”

“This strongly suggests that a co-operative activity exists in haplotype-mediated protein communication, which was already made explicit in previously described haplotype-delimited protein interaction networks [14].”

“The CTD region, however, had lower GDT_TS and TM scores than the NTD but remained slightly below 70 and ~0.74, respectively, which are still adequate for use in research as both these scores indicate accurate topologies but less confidence with side-chain conformations.”

“…known to bind to RNA [69]. and the backbone of the single-stranded RNA molecule was localized in a electropositive ‘canyon’ between the ‘basic fi nger’ and the ‘palm’ helical sheet of the NTD, which embeds three arginine residues (R92, R93, and R149) and a lysine (K102) known to bind to RNA [69].”

“The CTD region, however, had lower GDT_TS and TM scores than the NTD but remained slightly below 70 and ~0.74, respectively, which are still adequate for use in research as both these scores indicate accurate topologies but less confidence with side-chain conformations.”

“This reveals a difference in evolutionary pressures for different proteins and showcases how haplotypes can be more or less beneficial for their corresponding VOCs.”

“…that IDRs that have high pLDDT scores from AlphaFold2 can undergo conditional folding under certain conditions to interact with other proteins [63]. that IDRs that have high pLDDT scores from AlphaFold2 can undergo conditional folding under certain conditions to interact with other proteins [63].”

“The protein is necessary for viral assembly and RNA synthesis and participates in several cellular processes affecting immunological and cell cycle responses of the host [16].”

“There also exists a pattern of recruitment where different haplotypes and VOCs adopt similar conformations at specific sites across the pandemic indicating the importance of both structural and mutational entropy [17].”

“This reveals a difference in evolutionary pressures for different proteins and showcases how haplotypes can be more or less beneficial for their corresponding VOCs.”

“The lower pLDDT scores of R3 and R4 likely indicate the transient nature The three IDRs in the N-protein had notably lower pLDDT scores than the ordered NTD and CTD regions. helices, where they exist only brie fl y and could be part of a short linear motif [61,62].”

“This reveals a difference in evolutionary pressures for different proteins and showcases how haplotypes can be more or less beneficial for their corresponding VOCs.”

“AlphaFold2 [31], the star of the last two biannual ab initio structure prediction experiments (CASP 14 and 15) [32,33], greatly broke records as its neural network (called the evoformer ) approach to modeling proteins without homologs was within the margin of error of experimental structure…”

“This reveals a difference in evolutionary pressures for different proteins and showcases how haplotypes can be more or less beneficial for their corresponding VOCs.”

“…makes the personal responsibility of an individual the ‘number-one’ factor in preventing the spread of COVID-19 through masking and other precautionary measures [12] and poses an even greater challenge for policymakers to ensure the proper measures are in place to prevent the spread of COVID-19.”

“There also exists a pattern of recruitment where different haplotypes and VOCs adopt similar conformations at specific sites across the pandemic indicating the importance of both structural and mutational entropy [17].”

“Despite advances, a comprehensive study of structural changes induced by haplotypes on the N-protein is absent from the current literature with most research being focused on the VOCs [40,41].”

“In the absence of experimental protein structures from VOCs, numerous studies utilized AlphaFold2 to explore the differences between the wild-type (WT) Wuhan strain, which is used as reference, and the emerging variants [15,34–36].”

“H2 showed a seasonal pattern of accumulation, which later turned into complete prevalence during the pandemic.”

“The CTD region, however, had lower GDT_TS and TM scores than the NTD but remained slightly below 70 and ~0.74, respectively, which are still adequate for use in research as both these scores indicate accurate topologies but less confidence with side-chain conformations.”

“The lower pLDDT scores of R3 and R4 likely indicate the transient nature The three IDRs in the N-protein had notably lower pLDDT scores than the ordered NTD and CTD regions. helices, where they exist only brie fl y and could be part of a short linear motif [61,62].”

“In the absence of experimental protein structures from VOCs, numerous studies utilized AlphaFold2 to explore the differences between the wild-type (WT) Wuhan strain, which is used as reference, and the emerging variants [15,34–36].”

“Despite advances, a comprehensive study of structural changes induced by haplotypes on the N-protein is absent from the current literature with most research being focused on the VOCs [40,41].”

“These methods explore the 3D conformational space of a particular protein sequence with a set of rules [30].”

“In fact, COVID-19 epidemiological variables exhibit significant negative correlation with temperature and positive correlation with latitude, often peaking around winter months, further indicating the seasonal nature of the virus [37–39].”

“In the absence of experimental protein structures from VOCs, numerous studies utilized AlphaFold2 to explore the differences between the wild-type (WT) Wuhan strain, which is used as reference, and the emerging variants [15,34–36].”

“For instance, AF2 has been used to study patterns of haplotype diversification of spike and nucleocapsid proteins of SARS-CoV2 [46,47].”

“The domains are separated by a disordered linker, which together with the two domains hold crucial epitopes for vaccine development that have been actively mutated during the COVID-19 pandemic to change the structure of the molecule and tailor the seasonal behavior of the virus [50].”

“Notably, AF2 ’ s accuracy has reached the level where it can assist in interpreting X-ray crystallographic diffraction data and guide experimental model building (Ali and Caetano-Anollés, 2024; Gutnik et al., 2023).”