Pyroptosis of syncytia formed by fusion of SARS-CoV-2 spike and ACE2-expressing cells

Coronavirus Disease 2019 (COVID-19) is an infectious disease associated with systematical multi-organ failure caused by SARS-CoV-2, which mainly infects the lung and upper respiratory system 1,2 . Massive multinucleated syncytia are commonly observed in autopsy of severe COVID-19 patients 3 . It has been reported that the interaction between Spike (S) protein and ACE2 not only mediated the fusion of virus with host cells, but also multinucleated giant cells formation 4 – 7 . However, the underlying molecular mechanisms of syncytia death are poorly understood. To better observe the formation of syncytia


Lentivirus Production and Infection
HEK293T cells were co-transfected with lentiviral vector carrying interested cDNA and lentivirus-packing plasmids (PMDL/REV/VSVG) by the calcium phosphate precipitation method, GFP as control; 8 hours later, the cell culture medium was replaced with fresh medium and the virus-containing medium was collected 48 h later, then centrifuged at 1000g for 3 min, kept at 4℃. For infection, virus containing medium in the presence of 10 ug/ml Polybrene was added to cells plated in 12-well plates and then centrifuged at 1450g for 30 min at 37℃.The cell culture medium was replaced with fresh medium 8 h later. After 48h, the expression of indicated proteins were analyzed by western blot.

Confocal microscopy
All the images were obtained in live-cell scanning model. The cells harboring GFP or RFP tagged plasmid were cultured in 35mm Glass Bottom Culture Dishes (Nest).
Imaging was carried out using the Zeiss LSM 780 with a 100x/1.49 NA oil objective in a 37°C incubator containing 5% CO2. GFP was excited under a 488-nm argon laser, and RFP was excited under a 568-nm argon laser. Nuclei were stained using Hoechst 33342 (1:10,000) and excited under a 405-nm argon laser.

Western blot
The indicated cells were harvested and immediately lysed with 1.2x SDS sample buffer.
Total cell lysates were separated by SDS-PAGE and transferred to polyvinylidene fluoride membranes (EMD Millipore, 0.22μm) for Western blot analysis with the appropriate antibodies. The proteins were visualized by enhanced chemiluminescence in accordance with the manufacturer's instructions (NcmECL Ultra, Enhanced Chemiluminescent, ECL).

Cell-cell fusion assay
ACE2-expressing cells and Spike-expressing cells were co-culture (2 x 10 5 cells/well cells mixed at a 1:1 ratio in 24-well plates); then images of 5 fields (20x objective lens) were taken at indicated time by ZEISS Axio Observer; Nucleus was stained by Hoechst, counted by image J. The fusion index (FI) was calculated as "% of nuclei in fused cells" 2 .

Analysis of scRNA-seq data
We analyzed a dataset (Gene Expression Omnibus GSE122960) freely available from 5 public database. We used the R (version 3.6.3) and Seurat R package (version 3.9.9.9010). Firstly, Low-quality cells with less than 200 detected genes and genes found to be expressed in less than three cells were removed. Then, according to the median number of genes and the percentage of mitochondrial genes in the lung samples, we filtered cells that have unique feature counts over 5000 and cells that have >10% mitochondrial counts. After QC (Quality control), 42225 high quality lung cells were obtained.
For each individual sample, we employed a global-scaling normalization method "LogNormalize" that normalizes the feature expression measurements for each cell by the total expression, multiplies this by a scale factor (10,000 by default), and logtransforms the result. Identification of highly variable features was implemented in the FindVariableFeatures function. We detected the batch effect between eight different lung samples. So, we used the functions (FindIntegrationAnchors and IntegrateData) to integrate all eight individual samples and mitigate batch effects. Mitochondrial genes and differences in cell cycle stages between proliferating cells were regressed out by specifying the vars.to.regress argument in Seurat function ScaleData. We performed PCA on the scaled data, only the previously determined variable features were used as input. Identification of the true dimensionality of the dataset for further analysis was constructed by generating an "Elbow plot".
To cluster the cells, we first used the FindNeighbors function to construct a KNN graph based on the euclidean distance in PCA space, and refine the edge weights between any two cells based on the shared overlap in their local neighborhoods. We next iteratively

Fig. S2 Syncytia formation was dependent on the expression of Spike and ACE2
A549-ACE2-GFP cells was co-cultured with A549-Spike (or L929-Spike) cells (a and c), and A549-GFP cells was co-cultured with A549-Spike (or L929-Spike) cells (b and d), four hours later, syncytia formation in both ACE2 and Spike-expressing cells, nucleus was stained by Hoechst, bar, 50 μm.  :1 ratio, 4 hours later (a, b, c), or 12 hours later (d and e), image was scanned by LSM780, the nucleus(blue) was stained by Hoechst; Bar, 20 μm.
b. Western blot analyzed these knock-out cell clones.
c. To exclude the clonal effect, another individual clone was used to repeat the experiments, which showed the cleavage of Spike (refer to S2') induced by syncytia formation disappeared in C9-deficient cells.