Differences in syncytia formation by SARS-CoV-2 variants modify host chromatin accessibility and cellular senescence via TP53

Abstract

Coronavirus disease 2019 (COVID-19) remains a significant public health threat due to the ability of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) variants to evade the immune system and cause breakthrough infections. Although pathogenic coronaviruses such as SARS-CoV-2 and Middle East respiratory syndrome (MERS)-CoV lead to severe respiratory infections, how these viruses affect the chromatin proteomic composition upon infection remains largely uncharacterized. Here, we use our recently developed integrative DNA And Protein Tagging methodology to identify changes in host chromatin accessibility states and chromatin proteomic composition upon infection with pathogenic coronaviruses. SARS-CoV-2 infection induces TP53 stabilization on chromatin, which contributes to its host cytopathic effect. We mapped this TP53 stabilization to the SARS-CoV-2 spike and its propensity to form syncytia, a consequence of cell-cell fusion. Differences in SARS-CoV-2 spike variant-induced syncytia formation modify chromatin accessibility, cellular senescence, and inflammatory cytokine release via TP53. Our findings suggest that differences in syncytia formation alter senescence-associated inflammation, which varies among SARS-CoV-2 variants.

Keywords: ATAC-seq; COVID-19; CP: Immunology; SARS-CoV-2; chromatin; iDAPT-MS; p53; proteomics; senescence; spike.

Figure 1.. TP53 protein is stabilized upon SARS-CoV-2 infection and contributes to the SARS-CoV-2 host cytopathic effect

(A) Experimental design and schematic of the iDAPT approach (made with Biorender). (B) ATAC-seq volcano plot demonstrating the changes in accessible chromatin loci upon SARS-CoV-2 infection vs. mock infection in VeroE6 cells 48 h post-infection. FDR, false discovery rate. n = 4 technical replicates/group. (C) Principal component analysis of ATAC-seq profiles upon infection with SARS-CoV-2, HKU5-SARS-CoV-1-S, or MERS-CoV and mock infection. PC, principal component. n = 4 technical replicates/group. (D) iDAPT-MS volcano plot of SARS-CoV-2 infection (MOI 0.5) vs. mock infection in VeroE6 cells 48 h post-infection. Black circles, significant transcription factors (FDR < 0.05 and |log2 fold change| > 0.5). n = 2 technical replicates/group. (E) Principal component analysis of iDAPT-MS profiles upon infection with SARS-CoV-2, HKU5-SARS-CoV-1-S, or MERS-CoV and mock infection. n = 2 technical replicates/group. (F) Heatmap of relative protein abundance levels of differentially abundant (FDR < 0.05 and |log2 fold change| > 0.5) transcription factors from SARS-CoV-2 vs. mock infection iDAPT-MS analyses. Specificity is defined as the Pearson correlation between iDAPT-MS protein abundance and SARS-CoV-2 infection status. Sig, significant differential chromatin abundance by iDAPT-MS (FDR < 0.05 and |log2 fold change| > 0.5) upon infection relative to mock-infected cells. CRISPR, VeroE6 SARS-CoV-2 CRISPR-Cas9 screening results from Wei et al. Positive, Z score > 2. Negative, Z score < −2. (G) TP53 single guide RNA (sgRNA) enrichment from CRISPR-Cas9 pooled screening in VeroE6 cells for the three coronaviruses assessed vs. mock infection (Wei et al.). Z-scores are quantile-normalized for improved comparison between distributions. (H) Top, western blotting analysis of the indicated proteins after CRISPR-Cas9-based targeting of two TP53-targeting sgRNAs (TP53-1 and TP53-2) or negative control (NT, nontargeting). Bottom, the corresponding VeroE6 cells were infected with VSV pseudotyped viral particles (VSVpp) expressing VSV-G or different coronaviral spike proteins. Luciferase relative to the VSV-G-expressing VSVpp control was measured 24 h after infection. The mean ± SEM are shown. The Welch two-sample t test with Holm p value correction was used to assess statistical significance relative to negative control: n.s., not significant. n = 4 technical replicates/group. (I) Western blotting analysis of the indicated proteins after infection with the corresponding coronaviruses (MOI 0.1, 24 h). Relative ratios of TP53/ACTB (normalized to mock) are annotated below TP53. (J) Top, TP53 sgRNA enrichment across published CRISPR-Cas9 genetic screens stratified by TP53 mutational status of the parental cell line (Rebendenne et al.). The mean Z score of each group is demarcated by a horizontal line. p value, Welch two-sample t test. Bottom, TP53 mutational status associated with each parental cell line and corresponding impact of TP53 point mutations on its transcriptional activity as determined in Ursu et al.

Figure 2.. The SARS-CoV-2 spike protein is a key determinant of host cell chromatin accessibility changes and TP53 stabilization

(A) Top, schematic of the SARS-CoV-2 viral genome with encoded proteins. Bottom, number of significant (FDR < 0.05) ATAC-seq peaks upon transfection of plasmid-encoded viral proteins relative to empty vector and EGFP (48 h) in VeroE6 cells. Black triangles, peaks with increasing accessibility. Red circles, peaks with decreasing accessibility. n = 2 technical replicates/group. (B) Western blotting analysis of VeroE6 cells 48 h after transfection with plasmids encoding the indicated proteins. EV, empty vector. Relative ratios of TP53/ACTB (normalized to the corresponding EV) are annotated below TP53. (C) Western blotting analysis of VeroE6 cells 48 h after transfection with plasmids encoding the indicated FLAG-tagged proteins. EV, empty vector. Relative ratios of TP53/ACTB (normalized to EV) are annotated below TP53. (D) Schematic of cell-cell fusion assay. A 1:1 mixture of VeroE6 cells stably transduced with either pLenti-HiBiT-Bsr or pLenti-LgBiT-P2A-Bsr is transfected with corresponding plasmids. After 48 h, the cell-cell fusion effect is measured by NanoBiT luciferase complementation. Made with Biorender. (E) Cell-cell fusion analysis of VeroE6 cells transfected with the corresponding spike proteins. S2P, 2P mutant of the spike, which is incapable of membrane fusion. The Welch two-sample t test with Holm p value correction was used to assess statistical significance relative to the S2P mutant as negative control: ***p < 0.001; *p < 0.05; n.s., not significant. n = 5 technical replicates/group. (F) Schematic of VSV-G acidification assay. Cells were transfected with plasmid encoding VSV-G. 24 h later, cells were treated with a 3-min pulse of DMEM buffered with 10 mM morpholine-ethanesulfonic acid at different pH levels. Cells were analyzed 24 h after acidification. (G) Cell-cell fusion analysis of VeroE6 cells transfected with VSV-G plasmid and treated with buffers at different pH levels. The Welch two-sample t test with Holm p value correction was used to assess statistical significance relative to pH 7.5 as control: ***p < 0.001; **p < 0.01; n.s., not significant. n = 4 technical replicates/group. (H) Western blotting analysis of VeroE6 cells transfected with VSV-G plasmid and treated with buffers at different pH levels. Relative ratios of TP53/ACTB (normalized to pH 7.5) are annotated below TP53.

Figure 3.. Increasingly fusogenic SARS-CoV-2 variants promote increased TP53 protein stabilization

(A) Cell-cell fusion analysis of VeroE6 cells transfected with the corresponding spike proteins. S2P, 2P mutant of the spike. The Welch two-sample t test with Holm p value correction was used to assess statistical significance relative to the S2P mutant as negative control: ***p < 0.001. n = 6 technical replicates/group. (B) Western blotting analysis of VeroE6 cells 48 h after transfection with plasmids encoding the indicated FLAG-tagged proteins. EV, empty vector. Relative ratios of TP53/ACTB (normalized to EV) are annotated below TP53. (C) Western blotting analysis of A549ACE2 pInducer20 cells transduced with the indicated gene constructs and treated with or without 1 μg/mL doxycycline (Dox) for 48 h. Relative ratios of TP53/ACTB (normalized to EGFP -Dox) are annotated below TP53. (D) Representative brightfield images of A549ACE2 pInducer20 cells transduced with the indicated gene constructs and treated with or without 1 μg/mL Dox for 48 h. Scale bars, 200 μm. (E) Western blotting analysis of the indicated proteins 48 h after 1 μg/mL Dox treatment of VeroE6 pInducer20 cells transduced with the indicated gene constructs. (F) RNA-seq analysis of TP53 target gene upon ancestral, Alpha, or Delta spike expression vs. EGFP expression. p value, gene set enrichment analysis. n = 2 technical replicates/group. Also see Figures S4B and S4G. (G) Pearson correlation of ATAC-seq profiles upon ancestral, Alpha, or Delta spike expression vs. EGFP expression compared to treatment with or without 1 μM nutlin-3a for 24 h. p value, Pearson correlation. n = 4 technical replicates/group. Also see Figures S4C–S4F. (H) iDAPT-MS volcano plots of ancestral, Alpha, or Delta spike expression vs. EGFP expression in VeroE6 pInducer20 cells 48 h after 1 μg/mL Dox induction. Black points, significant transcription factors (FDR < 0.05 and |log2 fold change| > 0.5). n = 2 technical replicates/group. (I) Western blotting analysis of the indicated proteins after infection with the corresponding SARS-CoV-2 viral variants and MOIs (24 h). Relative ratios of TP53/ACTB (normalized to mock) are annotated below TP53. (J) Percentage of TP53-positive cells in lung biopsies from patients infected with pre-Delta or Delta SARS-CoV-2 variants vs. patient without infection (negative). Immunohistochemistry quantification was performed using QuPath. p value, linear regression coefficient. n = 3 patients/group. (K) Representative immunohistochemistry images of TP53 protein, SARS-CoV-2 spike, and CXCL8/IL8 protein epitopes in the lungs of patients infected with pre-Delta and Delta variant strains. Scale bars, 100 μm.

Figure 4.. Inhibition of SARS-CoV-2 spike-induced syncytia formation attenuates TP53 stabilization

(A) Top, western blotting analysis of A549ACE2 cells infected with VSVpp-SARS-CoV-2-S (VSVpp-S) displaying the indicated SARS-CoV-2 spike sequences for 24 h. A549ACE2 pInducer20 cells transduced with the Delta spike sequence (pInd20-Delta) and treated with or without 1 μg/mL doxycycline (Dox) for 24 h are shown for comparison. Relative ratios of TP53/ACTB (normalized to Delta + Dox) are annotated below TP53. Bottom, representative brightfield images of A549ACE2 cells infected with the corresponding VSVpp-SARS-CoV-2-S viruses for 24 h. Scale bars, 200 μm. (B) Top, western blotting analysis of 1:1 mixtures of A549 and A549ACE2 pInducer20 cells transduced with the indicated gene constructs and treated with or without 1 μg/mL Dox for 48 h. Relative ratios of TP53/ACTB (normalized to Delta + Dox) are annotated below TP53. Bottom, representative brightfield images of A549/A549ACE2 pInducer20 cell mixtures as above treated with or without 1 μg/mL Dox for 48 h. Scale bars, 200 μm. (C) Top, western blotting analysis of A549ACE2 pInducer20 cells transduced with the indicated gene constructs and treated with or without 1 μg/mL Dox and with the corresponding amount of neutralizing antibody (NAb, μg/mL, Sino Biological #40592-R001) for 24 h. Relative ratios of TP53/ACTB (normalized to Delta + Dox) are annotated below TP53. Bottom, representative brightfield images of A549ACE2 pInducer20-Delta spike cells treated with or without 1 μg/mL Dox and with the corresponding amount of neutralizing antibody (μg/mL) for 24 h. Scale bars, 200 μm. (D) Top, western blotting analysis of A549ACE2 pInducer20 cells transduced with the indicated gene constructs and treated with or without 1 μg/mL Dox and with the corresponding amount of CMK (μM) for 24 h. Relative ratios of TP53/ACTB (normalized to Delta + Dox) are annotated below TP53. Bottom, representative brightfield images of A549ACE2 pInducer20-Delta spike cells treated with or without 1 μg/mL Dox and with the corresponding amount of CMK (μM) for 24 h. Scale bars, 200 μm. (E) Top, western blotting analysis of VeroE6 cells 48 h after transfection with plasmids encoding the indicated FLAG-tagged proteins. Relative ratios of TP53/ACTB (normalized to EV) are annotated below TP53. Bottom, cell-cell fusion analysis of VeroE6 cells transfected with the corresponding spike proteins. S-2P/D-2P, 2P mutant of the spike, which is incapable of membrane fusion. The Welch two-sample t test with Holm p value correction was used to assess statistical significance relative to the S-2P mutant as negative control: ***p < 0.001. n = 6 technical replicates/group. (F) Top, western blotting analysis of A549ACE2 pInducer20 cells transduced with the indicated gene constructs and treated with or without 1 μg/mL Dox and with the corresponding amount of 25-hydroxycholesterol (μM) for 24 h. Relative ratios of TP53/ACTB (normalized to Delta + Dox) are annotated below TP53. Bottom, representative brightfield images of A549ACE2 pInducer20-Delta spike cells treated with or without 1 μg/mL Dox and with the corresponding amount of 25-hydroxycholesterol (μM) for 24 h. Scale bars, 200 μm.

Figure 5.. Differences in spike-induced TP53 stabilization modify the activation of host chromatin accessibility, senescence, and inflammatory cytokine release

(A) Representative brightfield images of A549ACE2 TP53ko pInducer20 cells transduced with the indicated gene constructs and treated with or without 1 μg/mL doxycycline (Dox) for 48 h. Scale bars, 200 μm. (B) Top, western blotting analysis of whole cell lysates of A549ACE2 or A549ACE2 TP53ko pInducer20 cells transduced with the indicated gene constructs and treated with or without 1 μg/mL doxycycline (Dox) for 48 h. Relative ratios of TP53/ACTB (normalized to EGFP -Dox) are annotated below TP53. Bottom, western blotting analysis of supernatants (SUP) of the corresponding cell constructs and treatments. (C) MSigDB pathway enrichment analysis of differential RNA-seq profiles upon doxycycline-induced expression (48 h, 1 μg/mL) of SARS-CoV-2 ancestral or Delta spike relative to expression of EGFP in either A549ACE2 or A549ACE2 TP53ko pInducer20 cells. n = 2 technical replicates/group. (D) Cell viability as assessed by CellTiter-Glo in the indicated A549ACE2 or A549ACE2 TP53ko pInducer20 cells treated with or without 1 μg/mL doxycycline (Dox) for 72 h (mean ± SEM). The Welch two-sample t test with Holm p value correction was used to assess statistical significance: ***p < 0.001; **p < 0.01. n = 5 technical replicates/group. (E) RNA levels of CDKN1A by RT-qPCR in the indicated A549ACE2 or A549ACE2 TP53ko pInducer20 cells treated with or without 1 μg/mL doxycycline (Dox) for 48 h (mean ± SEM). The Welch two-sample t test with Benjamini-Hochberg p value correction was used to assess statistical significance relative to the EGFP -Dox condition: ***p < 0.001; **p < 0.01; *p < 0.05; n.s., not significant. n = 4 technical replicates/group. (F) Senescence-associated beta-galactosidase activity in the indicated A549ACE2 or A549ACE2 TP53ko pInducer20 cells treated with or without 1 μg/mL doxycycline (Dox) for 48 h (mean ± SEM). The Welch two-sample t test with Holm p value correction was used to assess statistical significance: **p < 0.01; n.s., not significant. n = 14 technical replicates/group. (G) Left, ATAC-seq volcano plots of ancestral or Delta spike vs. EGFP expression in A549ACE2 pInducer20 cells 48 h after 1 μg/mL doxycycline induction. Right, number of significant (FDR < 0.05) ATAC-seq peaks upon doxycycline-induced expression (48 h, 1 μg/mL) of SARS-CoV-2 ancestral or Delta spike relative to expression of EGFP in either A549ACE2 or A549ACE2 TP53ko pInducer20 cells. Black triangles, peaks with increasing accessibility. Red circles, peaks with decreasing accessibility. n = 2 technical replicates/group. (H) Left, heatmap of relative levels of differentially expressed microRNAs (FDR < 0.05 and base mean > 10) from ancestral or Delta spike vs. EGFP expression in A549ACE2 or A549ACE2 TP53ko pInducer20 cells 48 h after 1 μg/mL doxycycline induction. Right, number of significant (FDR < 0.05) differentially expressed microRNAs upon doxycycline-induced expression (48 h, 1 μg/mL) of ancestral or Delta spike vs. EGFP expression in A549ACE2 or A549ACE2 TP53ko pInducer20 cells. Black triangles, significantly upregulated microRNAs. Red circles, significantly downregulated microRNAs. n = 2 technical replicates/group. (I) Heatmap of relative levels of cytokines in supernatants of A549ACE2 or A549ACE2 TP53ko pInducer20 cells treated with 1 μg/mL doxycycline for 48 h. The Welch two-sample t test with Benjamini-Hochberg p value correction was used to assess statistical significance. Sig, FDR < 0.05. n = 4 technical replicates/group. (J) Representative brightfield images of HNEpC (human nasal epithelial cell)-ACE2 pInducer20 cells transduced with the indicated gene constructs and treated with or without 1 μg/mL doxycycline (Dox) for 48 h. Scale bars, 200 μm. (K) Western blotting analysis of supernatants of HNEpC-ACE2 pInducer20 cells transduced with the indicated gene constructs and treated with or without 1 μg/mL doxycycline (Dox) for 48 h.

Figure 6.. Differences in SARS-CoV-2 spike sequences modify TP53 stabilization and cell-cell fusion

(A) Evolution of SARS-CoV-2 variants in the USA and the U.K. over the course of the COVID-19 pandemic. Gray, other SARS-CoV-2 variants. Data were last accessed from covariants.org on 4/3/2023. (B) Top, annotated SARS-CoV-2 spike subdomains. Bottom, heatmap representation of SARS-CoV-2 variant spike sequences. White boxes, amino acid sequences corresponding with ancestral spike; colored boxes, amino acid sequences different from ancestral spike at the indicated amino acid position. CTD, C-terminal domain; FP, fusion peptide; HR1/HR2, heptapeptide repeat sequences 1 and 2; NTD, N-terminal domain; RBD, receptor-binding domain; PBCS, polybasic cleavage site; TA, transmembrane anchor. (C) Western blotting analysis of VeroE6 cells 48 h after transfection with plasmids encoding the indicated FLAG-tagged proteins. EV, empty vector. Relative ratios of TP53/ACTB (normalized to EV) are annotated below TP53. (D) Cell-cell fusion analysis of VeroE6 cells transfected with the corresponding spike proteins. EV, empty vector. The Welch two-sample t test with Holm p value correction was used to assess statistical significance relative to EV as negative control: ***p < 0.001; **p < 0.01; *p < 0.05; n.s., not significant. n = 4 technical replicates/group. (E) Scatterplot of fusion effect (Figure 6D) and TP53 protein level (Figure 6C) by SARS-CoV-2 spike variant. Red points, Omicron spike sequences. R, Pearson correlation. p value, Pearson correlation test. (F) Proposed model of findings.