Post by Admin on Jul 7, 2021 3:47:29 GMT
Validation of direct physical interactions between selected COV-VIPs and SARS-CoV-2 proteins
To further validate that an ancient viral epidemic was responsible for the observed selection signals, next we test whether the 35 out of 42 selected CoV-VIPs that interact with SARS-CoV-2 (as opposed to other coronaviruses in our dataset) are indeed CoV-VIPs and directly interact with SARS CoV-2 viral proteins. While these interactions were originally identified by high-throughput mass spectrometry,19 high-throughput mass spectrometry can sometimes identify indirect interactions in a larger protein complex or false positives altogether.41 We co-express the candidate CoV-VIPs:SARS CoV-2 protein pairs in a cell-free protein expression system and test their interactions using an AlphaLISA protein:protein interaction assay (STAR Methods). This approach (Figure S6A) was previously used for rapid analysis of intra-viral PPI network of Zika virus.42 The assay is expected to detect ∼70% of protein interactions with human proteins (30% false negative rate; STAR Methods). Out of 35 selected SARS-CoV-2 CoV-VIPs, 33 interacting protein pairs can be tested with the assay (STAR Methods). Figure 6 highlights the results for six of the 33 CoV-VIPs, while Figure S6 presents the results for the remaining CoV-VIPs. Among the 33 interactions tested, we confirm 24 or 73%, the expected confirmation rate (taking the false negative rate into account) if 100% or close to 100% of the selected CoV-VIPs are indeed CoV-VIPs (Figures 6A–6C and S6B; Data S1J). This very high validation rate further strengthens the evidence for an ancient viral epidemic in East Asia.
Figure 6
Validation of selected CoV-VIPs/SARS-CoV-2 protein interactions using cell-free expressed proteins
Selected CoV-VIPs are enriched for antiviral and proviral factors
To further clarify that a viral epidemic caused the strong burst of selection, and not another ecological pressure acting on the same set of genes, we test whether the 42 selected CoV-VIPs are enriched for genes with antiviral or proviral effects relative to other CoV-VIPs (i.e., loci that are known to have a detrimental or beneficial effect on the virus, respectively). Because the relevant literature for coronaviruses is currently limited, we extend our set of anti- and proviral loci to include loci reported for diverse viruses with high confidence from the general virology literature (STAR Methods; Data S1K and S1L). We find that 21 (50%) of the 42 CoV-VIPs that came under selection ∼900 generations ago have high-confidence anti- or proviral effects (versus 29% for all 420 CoV-VIPs), a significant inflation in such effects (hypergeometric test p = 6.10−4) that further supports our claim that the underlying selective pressure was most likely a viral epidemic.
To further validate that an ancient viral epidemic was responsible for the observed selection signals, next we test whether the 35 out of 42 selected CoV-VIPs that interact with SARS-CoV-2 (as opposed to other coronaviruses in our dataset) are indeed CoV-VIPs and directly interact with SARS CoV-2 viral proteins. While these interactions were originally identified by high-throughput mass spectrometry,19 high-throughput mass spectrometry can sometimes identify indirect interactions in a larger protein complex or false positives altogether.41 We co-express the candidate CoV-VIPs:SARS CoV-2 protein pairs in a cell-free protein expression system and test their interactions using an AlphaLISA protein:protein interaction assay (STAR Methods). This approach (Figure S6A) was previously used for rapid analysis of intra-viral PPI network of Zika virus.42 The assay is expected to detect ∼70% of protein interactions with human proteins (30% false negative rate; STAR Methods). Out of 35 selected SARS-CoV-2 CoV-VIPs, 33 interacting protein pairs can be tested with the assay (STAR Methods). Figure 6 highlights the results for six of the 33 CoV-VIPs, while Figure S6 presents the results for the remaining CoV-VIPs. Among the 33 interactions tested, we confirm 24 or 73%, the expected confirmation rate (taking the false negative rate into account) if 100% or close to 100% of the selected CoV-VIPs are indeed CoV-VIPs (Figures 6A–6C and S6B; Data S1J). This very high validation rate further strengthens the evidence for an ancient viral epidemic in East Asia.
Figure 6
Validation of selected CoV-VIPs/SARS-CoV-2 protein interactions using cell-free expressed proteins
Selected CoV-VIPs are enriched for antiviral and proviral factors
To further clarify that a viral epidemic caused the strong burst of selection, and not another ecological pressure acting on the same set of genes, we test whether the 42 selected CoV-VIPs are enriched for genes with antiviral or proviral effects relative to other CoV-VIPs (i.e., loci that are known to have a detrimental or beneficial effect on the virus, respectively). Because the relevant literature for coronaviruses is currently limited, we extend our set of anti- and proviral loci to include loci reported for diverse viruses with high confidence from the general virology literature (STAR Methods; Data S1K and S1L). We find that 21 (50%) of the 42 CoV-VIPs that came under selection ∼900 generations ago have high-confidence anti- or proviral effects (versus 29% for all 420 CoV-VIPs), a significant inflation in such effects (hypergeometric test p = 6.10−4) that further supports our claim that the underlying selective pressure was most likely a viral epidemic.