A SARS-CoV-2-specific CAR-T-cell model identifies felodipine, fasudil, imatinib, and caspofungin as potential treatments for lethal COVID-19

Abstract

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)-induced cytokine storm is closely associated with coronavirus disease 2019 (COVID-19) severity and lethality. However, drugs that are effective against inflammation to treat lethal COVID-19 are still urgently needed. Here, we constructed a SARS-CoV-2 spike protein-specific CAR, and human T cells infected with this CAR (SARS-CoV-2-S CAR-T) and stimulated with spike protein mimicked the T-cell responses seen in COVID-19 patients, causing cytokine storm and displaying a distinct memory, exhausted, and regulatory T-cell phenotype. THP1 remarkably augmented cytokine release in SARS-CoV-2-S CAR-T cells when they were in coculture. Based on this "two-cell" (CAR-T and THP1 cells) model, we screened an FDA-approved drug library and found that felodipine, fasudil, imatinib, and caspofungin were effective in suppressing the release of cytokines, which was likely due to their ability to suppress the NF-κB pathway in vitro. Felodipine, fasudil, imatinib, and caspofungin were further demonstrated, although to different extents, to attenuate lethal inflammation, ameliorate severe pneumonia, and prevent mortality in a SARS-CoV-2-infected Syrian hamster model, which were also linked to their suppressive role in inflammation. In summary, we established a SARS-CoV-2-specific CAR-T-cell model that can be utilized as a tool for anti-inflammatory drug screening in a fast and high-throughput manner. The drugs identified herein have great potential for early treatment to prevent COVID-19 patients from cytokine storm-induced lethality in the clinic because they are safe, inexpensive, and easily accessible for immediate use in most countries.

Keywords: CAR-T; COVID-19; NF-κB pathway; SARS-CoV-2; anti-inflammation.

Fig. 1

The expression, activation, and function of SARS-CoV-2 S protein-targeted CAR-T cells (SARS-CoV-2-S CAR-T). A Schematic illustration of two second-generation SARS-CoV-2-S CAR constructs. The CAR is composed of a signal peptide of the interleukin (IL)2 receptor (Sp1), an anti-RBD scFv, a spacer (IgG1 Fc or IgG1 hinge), a CD28 transmembrane domain (CD28 TM), and intracellular signaling domains (CD28 and CD3ζ). The spacers IgG1 Fc and IgG1 hinge were used in Fc-SARS-CoV-2-S and Hinge-SARS-CoV-2-S CAR constructs, respectively. B 293 T cells transfected with a control lentiviral vector (CTL vector) or SARS-CoV-2-S CAR expression vectors as described in A were stained with eGFP-tagged S protein followed by flow cytometry analysis. Light gray, CTL vector; dark gray, SARS-CoV-2-S CAR. C 293 T cells transfected with a CTL vector or Fc-SARS-CoV-2-S CAR expression vector as described in A were stained with an APC-conjugated anti-human IgG-Fc antibody followed by flow cytometry analysis. Light gray, CTL vector; dark gray, Fc-SARS-CoV-2-S CAR. D Primary T lymphocytes from a healthy donor were expanded and stained with an anti-CD3 antibody conjugated with FITC (CD3-FITC) (left panel) or an anti-CD8 antibody conjugated with APC (CD8-APC) (right panel) followed by flow cytometry analysis. Light gray, blank; dark gray, CD3 or CD8 staining. E Primary T lymphocytes infected with control lentivirus (CTL lentivirus) or Fc-SARS-CoV-2-S CAR lentivirus as described in A were stained with an APC-conjugated anti-human IgG-Fc antibody followed by flow cytometry analysis. Light gray, CTL lentivirus; dark gray, Fc-SARS-CoV-2 CAR-S lentivirus. F SARS-CoV-2-S CAR-T cells were incubated with 293 T cells or 293 T cells transfected with S protein (S-293T cells) at a ratio of 3:1 for two days, and T cells in suspension were separated from adherent 293 T cells and costained with CD3-APC and CD69-PE or CD25-FITC followed by flow cytometry analysis. G SARS-CoV-2-S CAR-T cells were incubated with 293 T or S-293T cells and maintained in culture medium for the indicated durations. The number of viable cells was counted at the indicated time points (mean ± s.e.m, ***P < 0.001). H SARS-CoV-2-S CAR-T cells were incubated with 293 T or S-293T cells at different ratios (1:1, 3:1, 6:1, 10:1, 20:1 or 30:1) for three days before measuring the secretion of cytokines, including IFNγ, TNFα, IL2, granzyme B, perforin, GM-CSF, IL6, and IL10. CTL T cells were used as a negative control (mean ± s.e.m). I SARS-CoV-2-S CAR-T cells were incubated with 293 T or S-293T cells at different ratios for the indicated durations, followed by a cytotoxicity assay (mean ± s.e.m). J SARS-CoV-2-S CAR-T cells were incubated with 293 T cells fused with eGFP (eGFP-293T) or 293 T cells expressing spike fused with eGFP (eGFP-S-293T) at different ratios for the indicated duration, followed by GFP fluorescence detection. The number of GFP-positive cells was counted (mean ± s.e.m)

Fig. 2

Phenotypical characterization of SARS-CoV-2-S CAR-T cells upon antigen stimulation. SARS-CoV-2-S CAR-T cells were incubated with 293 T or S-293T cells at a ratio of 3:1 for two days, and T cells in suspension were separated from adherent 293 T cells and collected, followed by RNA extraction and RT‒qPCR analysis to examine the expression of representative genes associated with T-cell proliferation, activation, IFNγ response and cytotoxicity (A), naive and effector T-cell gene signatures (B), and check points and transcription factors (E). The significance test is shown in Fig. S2 (*P < 0.05; **P < 0.01; ***P < 0.001). C CTL T or SARS-CoV-2-S CAR-T cells were incubated with or without 293 T or S-293T cells at a ratio of 3:1 for two days, and T cells in suspension were separated from adherent 293 T cells and stained with an anti-CD8 antibody conjugated with APC (CD8-APC), an anti-CD62L antibody conjugated with PE (CD62L-PE), and an anti-CCR7 antibody conjugated with FITC (CCR7-FITC) followed by flow cytometry analysis. D CAR-T cells as described in A were subjected to staining with an anti-CD45RA antibody conjugated with APC (CD45RA-APC), an anti-CD45RO antibody conjugated with FITC (CD45RO-FITC), and an anti-CD62L antibody conjugated with PE (CD62L-PE). Coexpression of CD62L and CD45RA was analyzed by flow cytometry. The expression of CD45RO and CD45RA was further measured on gated CD62L-negative cells. F CAR-T cells as described in A were subjected to staining with an anti-PD1 antibody conjugated with FITC (PD1-FITC), an anti-TIM3 antibody conjugated with APC (TIM3-APC) and an anti-LAG3 antibody conjugated with PE (LAG3-PE) followed by flow cytometry analysis. G CAR-T cells as described in A were subjected to staining with an anti-CD25 antibody conjugated with APC (CD25-APC), an anti-CD4 antibody conjugated with PE (CD4-PE) and an anti-FOXP3 antibody conjugated with Alexa Fluor® 488 (FOXP3-488). Coexpression of CD4 and FOXP3 was analyzed by flow cytometry. The expression of CD25 and FOXP3 was further measured on gated CD4-positive cells. All data are representative of three independent experiments

Fig. 3

Monocytes enhance the cytokine release of SARS-CoV-2-S CAR-T cells. A THP1, S-293T, and SARS-CoV-2-S CAR-T cells were mixed at the indicated ratios for three days before measuring the secretion of IFNγ and IL6 (mean ± s.e.m, ***P < 0.001). Data presented are the mean (± s.e.m) from three repeats. B Cells as described in A were subjected to measurement of the secretion of multiple cytokines by a multiplex bead array (mean ± s.e.m). C The violin plot shows the expression of IL8 (CXCL8) in the PBMCs of COVID-19 patients and healthy controls (***P < 0.001)

Fig. 4

Felodipine, fasudil, imatinib, and caspofungin, screened by the SARS-CoV-2-S CAR-T-cell model, suppress cytokine release. A The flowchart for drug screening that can suppress SARS-CoV-2-induced inflammation based on the SARS-CoV-2-S CAR-T-cell model is shown. B S-293T, SARS-CoV-2-S CAR-T, and THP1 cells (1:10:10) were coincubated and treated with or without individual drugs (10 μM) from an FDA-approved library (~1049) before measuring the secretion of IL8 in the culture supernatants by ELISA. C The culture supernatants from drugs with more than 70% inhibition of IL8 secretion screened from B were further subjected to measurement of IFNγ by ELISA. D T cells were treated with drugs with more than 70% inhibition of IFNγ secretion screened from C for three days followed by toxicity assay (mean ± s.e.m). E S-293T, SARS-CoV-2-S CAR-T, and THP1 cells (1:10:10) were coincubated and treated with or without felodipine (#1), fasudil (#2), imatinib (#3), or caspofungin (#4) (10 μM) for three days before measuring the secretion of cytokines by a multiplex bead array. The inhibition rate is presented by a heatmap. F SARS-CoV-2-S CAR-T cells coincubated with S-293T cells (5:1) were treated with or without felodipine (#1), fasudil (#2), imatinib (#3), or caspofungin (#4) (10 μM) for the indicated duration, followed by a cytotoxicity assay (mean ± s.e.m). G SARS-CoV-2-S CAR-T cells were incubated with control 293 T or S-293T cells at a ratio of 3:1 for two days, and cells in suspension were subjected to RNA-seq analysis. Genes up- and downregulated by S-293T cells are shown in a pie chart (FDR < 0.01, FC > 1.5). H SARS-CoV-2-S CAR-T cells incubated with S-293T cells were treated with or without felodipine, fasudil, imatinib, or caspofungin (10 μM) for two days, and cells in suspension were subjected to RNA-seq analysis. Genes that were upregulated by S-293T cells but suppressed by felodipine, fasudil, imatinib, and caspofungin are shown. I Hallmark gene set enrichment analysis for overlapping genes (n = 865) as described in H. J, K SARS-CoV-2-S CAR-T cells were incubated with S-293T cells and treated with or without felodipine, fasudil, imatinib, or caspofungin (10 μM) for two days. T cells in suspension were subjected to RT‒qPCR analysis (J) to examine the expression of genes as indicated or immunoblotting analysis (K) using antibodies as indicated (Student’s t test, unpaired, two-tailed, *P < 0.05; **P < 0.01; ***P < 0.001)

Fig. 5

Felodipine, fasudil, imatinib, and caspofungin attenuate lethal inflammation, ameliorate severe pneumonia, and prevent mortality in SARS-CoV-2-infected Syrian hamsters. A Schematic representation of SARS-CoV-2 infection and drug treatment. Hamsters were intranasally inoculated with SARS-CoV-2 (1 × 10⁴ PFU) and then treated with or without felodipine (i.p.), fasudil (i.p.), imatinib (i.g.), or caspofungin (i.p.) on the indicated days (n = 6). Hamsters intranasally inoculated with SARS-CoV-2 without treatment were used as controls. Body weight was measured daily. Animals were euthanized to collect tissue samples at Day 7 post infection for virological and histological analysis. B The body weight of hamsters in each group from Day 0 to 7 post infection is shown. Significance was calculated using two-way ANOVA and is shown in the table (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, ns: nonsignificant). C The survival curve for hamsters in each group from Day 0 to 7 post infection is shown. Significance was calculated using the log-rank (Mantel‒Cox) test (**P < 0.01). D Gross images of lung tissues collected from hamsters at Day 7 post infection are shown. E Representative images after H&E staining for lung lobe sections in hamsters at Day 7 post infection are shown. Scale bars, 50 μm. F Comprehensive pathological scores based on the severity and percentage of injured areas for each lung lobe for lung sections shown in Fig. S3A are shown. Significance was calculated using one-way ANOVA (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, ns: nonsignificant). G The expression of inflammatory genes in lung tissues collected from hamsters at Day 7 post infection was detected by RT‒qPCR and represented by a heatmap. Data presented are the normalized value to the mock group after normalization to the expression of β-actin. The significance test is shown in Fig. S3B (*P < 0.05; **P < 0.01; ***P < 0.001). H The levels of cytokines, including IL6, IL10, IFNγ, and IFNβ in the lung tissues collected from hamsters at Day 7 post infection were measured by ELISA. Significance was calculated using one-way ANOVA (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001). I RNA was extracted from the lungs of hamsters as described in A at Day 7 post infection followed by RNA-seq analysis. The similarity in the gene expression profile for the mock and SARS-CoV-2-infected groups is shown by a PCA plot. J Genes regulated by SARS-CoV-2 infection are shown in a volcano plot. Blue and red dots represent down- and upregulated genes, respectively (FDR < 0.01, FC > 1.5). K Genes that were upregulated by SARS-CoV-2 infection as shown in J but repressed by felodipine, fasudil, imatinib, and caspofungin are shown in a Venn diagram. L KEGG analysis for those 706 genes that were upregulated by SARS-CoV-2 infection but commonly suppressed by felodipine, fasudil, imatinib, and caspofungin as described in K is shown

Fig. 6

A SARS-CoV-2-specific CAR-T-cell model identifies felodipine, fasudil, imatinib, and caspofungin as effective treatments for SARS-CoV-2-infected Syrian hamsters. We demonstrated the feasibility of producing SARS-CoV-2-S CAR-T cells that have immunological profiles similar to those of T cells from COVID-19 patients in the presence of the S protein. This CAR-T-cell model can be used to screen drugs to suppress SARS-CoV-2-induced cytokine release. Among the drugs screened, felodipine, fasudil, imatinib, and caspofungin are effective in protecting against SARS-CoV-2-induced lethal inflammation in Syrian hamsters