key: cord-0845533-fmzzpdjy authors: Geng, Jiejie; Chen, Liang; Yuan, Yufeng; Chen, Ruo; Wang, Ke; Deng, Yongqiang; Du, Peng; Liu, Jiangning; Wu, Guizhen; Wang, Youchun; Xu, Ke; Sun, Xiuxuan; Guo, Ting; Yang, Xu; Wu, Jiao; Jiang, Jianli; Li, Ling; Zhang, Jun; Zhang, Kui; Zhu, Hua; Zheng, Zhaohui; Fu, Xianghui; Yang, Fengfan; Chen, Xiaochun; Tang, Hao; Zhang, Zheng; Wei, Ding; Zhang, Yang; Shi, Ying; Zhu, Yumeng; Pei, Zhuo; Huo, Fei; Chen, Shirui; Wang, Qingyi; Xie, Wen; Li, Yirong; Shi, Mingyan; Bian, Huijie; Zhu, Ping; Chen, Zhi-Nan title: CD147 antibody specifically and effectively inhibits infection and cytokine storm of SARS-CoV-2 variants date: 2021-05-14 journal: bioRxiv DOI: 10.1101/2021.05.14.444111 sha: ad3bd49d9f218f7154b7c383dd48859fd2abcb3a doc_id: 845533 cord_uid: fmzzpdjy SARS-CoV-2 and its variants are raging worldwide. Unfortunately, the global vaccination is not efficient enough to attain a vaccine-based herd-immunity and yet no special and effective drug is developed to contain the spread of the disease. Previously we have identified CD147 as a novel receptor for SARS-CoV-2 infection. Here, we demonstrated that CD147 antibody effectively inhibits infection and cytokine storm caused by SARS-CoV-2 variants. In CD147KO VeroE6 cells, infections of SARS-CoV-2, its variants (B.1.1.7, B.1.351) and pseudovirus mutants (B.1.1.7, B.1.351, B.1.525, B.1.526 (S477N), B.1.526 (E484K), P.1, P.2, B.1.617.1, B.1.617.2) were decreased. Meanwhile, CD147 antibody effectively blocked the entry of variants and pseudomutants in VeroE6 cells, and inhibited the expression of cytokines. A model of SARS-CoV-2-infected hCD147 transgenic mice was constructed, which recapitulated the features of exudative diffuse alveolar damage and dynamic immune responses of COVID-19. CD147 antibody could effectively clear the virus and alveolar exudation, resolving the pneumonia. We found the elevated level of cyclophilin A (CyPA) in plasma of severe/critical cases, and identified CyPA as the most important proinflammatory intermediate causing cytokine storm. Mechanistically, spike protein of SARS-CoV-2 bound to CD147 and initiated the JAK-STAT pathway, which induced expression of CyPA. CyPA reciprocally bound to CD147, triggered MAPK pathway and consequently mediated the expression of cytokine and chemokine. In conclusion, CD147 is a critical target for SARS-CoV-2 variants and CD147 antibody is a promising drug to control the new wave of COVID-19 epidemic. The rapid and continual mutation of SARS-CoV-2 is giving rise to a new wave of epidemic, with more than 150 million people infected globally, resulting in over 3 million deaths. Since the outbreak of the pandemic in 2020, virus variants have These mutant strains have extensive mutations in spikes, including N501Y, E484K, E484Q, D614G, which accelerate the spread of the virus or facilitate to escape immune response (1-3). At present, global vaccination is the hope of eliminating the epidemic, but researchers have found that the existing vaccine is not effective against the mutant strains (4). Therefore, there is a need to develop safe and effective broad-spectrum approach that can control the rapid spread of SARS-CoV-2. COVID-19 pneumonia patients, if not treated in a timely manner, deteriorate rapidly within 7 to 14 days, leading to acute respiratory distress syndrome, which represents the major cause of morbidity in COVID-19 patients (5, 6) . The main pathological features of COVID-19 pneumonia comprise exudative diffuse alveolar damage with massive capillary congestion and hemorrhage (7) (8) (9) . The pathology of these lung tissue damages, which has been described in up to 20% of COVID-19 patients, are thought to be related to cytokine storm syndrome (CSS). CSS induces exudative diffuse alveolar damage, fibrin exudation in the alveolar space, and infiltration of inflammatory cells such as macrophages and neutrophils, accompanied by diffuse hemorrhage and massive release of cytokines and chemokines. Immune analysis in COVID-19 patients highlighted profound impaired type I IFN responses characterized by a low level of IFN activity and downregulation of IFN stimulated genes, and an increased number of Th17 cells and their related inflammatory responses, leading to hyper-inflammation and acute lung injury (10) (11) (12) . However, the mechanistical cause of CSS in COVID-19 pneumonia patients remains unclear due to a lack of suitable preclinical mouse model that recapitulate the pathology of COVID-19 pneumonia following SARS-CoV-2 infection. To study the inflammatory responses in COVID-19, many laboratories have constructed mice expressed by human angiotensin-converting enzyme 2 (hACE2), which was introduced into the mice in different ways, such as transient introduction of hACE2 via adenoviral vectors (13) and expression of hACE2 driven with the mouse ACE2 (14) , HFH4 (15) and K18 (16) promoters. These mouse models all supported SARS-CoV-2 infection and presented either mild or severe pneumonia features, including infiltration of inflammatory cells in the lung, high levels of proinflammatory cytokines and chemokines in lung homogenates, and severe interstitial and consolidative pneumonia (17) . However, hACE2 mouse is not an ideal model to study the pathology of COVID-19 pneumonia, because SARS-CoV-2-infected hACE2 mice mostly exhibited interstitial pneumonia, and exudation was rarely observed, which cannot well simulate the pathological and immune characteristics of patients with COVID-19 pneumonia. Our previous study discovered that CD147 is a novel receptor for SARS-CoV-2 infection, and human CD147 transgenic (hCD147) mice can be successfully infected with the virus (18) . In this article, CD147 KO VeroE6 cells, SARS-CoV-2-infected hCD147 transgenic mice, and CD147 antibody were used to evaluate the variants infection and cytokine expression, and analyze the pathological phenotypes of pneumonia and immune characteristics. We found that the CD147 molecule was still a receptor for variants' entry into host cells and could mediate cytokine storm. Thus, CD147 is confirmed to be a target for broad-spectrum of SARS-CoV-2 variants. To determine whether CD147 is an infection target of SARS-CoV-2 variants, we inhibition rates of 68.7%, 75.7% and 52.1% respectively for 60 μg/mL antibody (Fig. 1C ). We further infected 9 strains of pseudoviruses mutants, including B. (Fig. 1D) , and CD147 antibody could effectively inhibit the infection of these pseudoviruses (Fig. 1E) 1F ) and SARS-CoV-2 (18) . These results indicate that SARS-CoV-2 and its variants can enter cells by CD147, and CD147 antibody can resist the infection of broad-spectrum variants. The pathological and immune characteristics of hCD147 transgenic mice infected with SARS-CoV-2 are similar to those of COVID-19 patients. A variety of cytokines and chemokines were detected on VeroE6 cells infected by SARS-CoV-2, B.1.1.7 and B.1.351 variants, and it was found that the expressions of cytokines and chemokines caused by the three strains were comparable (Fig. S1 ). Therefore, SARS-CoV-2 was used as the main strain to detect the pathogenicity in vivo. In order to detect the pathogenicity of SARS-CoV-2 via CD147, we constructed a human CD147 (hCD147) transgenic mouse model by replacing extracellular region of mouse CD147 with human CD147 in embryonic stem cells (Fig. S2A ). PCR and FACS analyses showed that human CD147 was successfully expressed in mouse cells ( Fig. S2B and C) . We then inoculated hCD147 mice via the intranasal route with 3 × 10 5 TCID50 of SARS-CoV-2. After 5 days post infection (d.p.i.), hCD147 mice demonstrated marked weight loss, and all mice lost approximately 10% of their body weight by 13 d.p.i. (Fig. 2A) . High levels of viral RNA were detected in lung homogenates at 2 d.p.i. (Fig. 2B ). In some of the mice, low levels of viral RNA were detected in organs, such as heart, spleen and kidney (Fig. 2B) . H&E-stained lung sections from virus-infected hCD147 mice showed distinctive pathological characteristics at three time points, with the most severe symptoms at 6 d.p.i., characterized by widened alveolar wall, massive inflammatory cell infiltration, obvious alveolar serous fluid exudation and local lung consolidation ( Fig. 2C and S2D ). Multiplex immunofluorescence staining showed that a large number of macrophages, T cells and neutrophils infiltrated the lung alveoli and interstitium at 6 d.p.i. (Fig. 2D and E). In addition, infiltrated macrophages and neutrophils were also detected in the alveolar cavities (Fig. 2D) . We found SARS-CoV-2 infection elevated several proinflammatory cytokines and chemokines, such as IL6, CXCL2, CXCL10, IFNγ and IL17a, in the lung tissues of hCD147 mice, compared with that of virus-infected C57BL/6J mice at 6 d.p.i. (Fig. 2F ). RNA-seq analysis of lung homogenates from C57BL/6J and hCD147 mice at 2 d.p.i. showed 354 upregulated and 175 downregulated genes (Fig. S2E) . KEGG analysis showed intensive inflammatory responses in hCD147 mice, including Th17 cell differentiation, Th1 and Th2 cells differentiation and MAPK pathway (Fig. 2G) , similar to the immune responses in COVID-19 patients. The levels of cytokines and chemokines (IL6, IL10, IL1b, CXCL1, CXCL2 and CCL2) increased in the lung tissues of hCD147 mice, which was also found elevated in COVID-19 patients (19) (20) (21) . In addition, inflammation-related transcription factors were analyzed increased in the lung tissues of virus-infected hCD147 mice (Fig. 2H) . These results suggest that CD147 mediates potent inflammatory responses to SARS-CoV-2 infection, and the pathological and immune characteristics of hCD147 mice infected with SARS-CoV-2 imitate those of COVID-19 patients, which is an ideal model for study of pathogenesis of COVID-19. the hACE2 and hCD147 mouse models. To understand the pathological process caused by SARS-CoV-2 infection through different receptors, we infected hACE2 mice (22) and hCD147 mice with SARS-CoV-2. At 2 d.p.i. hACE2 mice presented interstitial pneumonia, whereas hCD147 mice showed exudative alveolar inflammation with more exudation and alveolar wall damage (Fig. 3A) . Electron microscopy revealed infiltration of macrophages and neutrophils, and leakage of erythrocytes, in alveoli of hCD147 mice, whereas these cells were rarely observed in alveoli of hACE2 mice (Fig. 3B ). Virions presented in the alveolar type II cells, vascular endothelial cells and macrophages in both mouse models (Fig. 3B ). In addition, exfoliated alveolar type II epithelial cells were observed in the lung tissues of hCD147 mice (Fig. 3B ). More infiltration of macrophages and neutrophils were found in the lung tissues of hCD147 mice than hACE2 mice (Fig. 3C ). RNA-Seq analysis showed that the upregulated genes in SARS-CoV-2-infected hCD147 mice were enriched in many cytokines and chemokines, with IL-17 signaling related genes (Fig. 3D ). These results indicate that SARS-CoV-2 causes different pneumonia phenotypes and immune responses in the hACE2 and hCD147 mouse models, and that the proinflammatory responses mediated by CD147 is stronger, which may trigger the occurrence of CSS. In order to make clear the molecular mechanism by which SARS-CoV-2 mediated different pathogeny through CD147 and ACE2 infection, we analyzed the upregulated immune related genes in hCD147 infected mice. In addition to cytokines and chemokines, the highly expressed genes related to the immune system were immunoglobulin, coagulation factor, complement system and so on (Fig. 3E) . Notably, CyPA (PPIA), which is an upstream regulator of effectors such as IL6, IL1b, IL1a, CXCL1 and CXCL2 (23) , showed significant upregulation in hCD147 mice (Fig. 3E) . We verified CyPA expression in the lung tissues of infected mice by RT-qPCR and found that CyPA expression in hCD147 mice was considerably higher than that in hACE2 mice (Fig. 3F ). Compared with C57BL/6J mice, CyPA expression in the lung tissues of virus-infected hCD147 mice was significantly increased at 2 d.p.i., peaked at 6 d.p.i. and reached a low level at 13 d.p.i. (Fig. 3G) , which suggests the involvement of CyPA in inflammation caused by SARS-CoV-2 infection. Immunohistochemical staining further confirmed that the expression of IL6, CCL2 and CyPA was higher in the lung tissues of hCD147 mice than hACE2 mice (Fig. S3A , B and C). In addition, compared with hACE2 mice, total p-STAT3 expression was higher in lung tissues of hCD147 mice ( Fig. S3D and E) . The intensity of p-STAT3 was positively correlated with the CyPA expression in the CyPA-positive cells in hCD147 mice (Fig. S3F ). These results suggest that CyPA is involved in the inflammatory responses of SARS-CoV-2 infection through CD147 and is likely to be a driver in the development of COVID-19 CSS. To further determine whether CyPA is a key intermediate proinflammatory factor that promotes the pathological process of SARS-CoV-2 infection via CD147, we conducted experiments in cell lines. In VeroE6 cells, upregulation of CyPA was induced by different titer virus infection (Fig. 4A ). Although the virus content was reduced by CD147 or ACE2 knockdown (Fig. S4A) , CyPA was detected downregulation by CD147 knockdown, but not by ACE2 knockdown (Fig. 4B ). In addition, SARS-CoV-2 failed to infect VeroE6 cells with double knockout of CD147 and ACE2 (Fig. 4C) , suggesting ACE2 and CD147 are the main receptors found for virus entry at present. CyPA expression decreased significantly in double knockout cells (Fig. 4D) , and enhanced by CD147 overexpression (Fig. 4E) . Previously, CyPA was reported to be regulated by the JAK-STAT pathway (24) . Indeed, CyPA level could be inhibited by JAK inhibitor (Fig. 4E) , indicating CD147 induces CyPA expression via JAK pathway during virus infection. Furthermore, the expression of CD147 and CypA was measured by multiplex immunofluoresence staining in lung tissues from COVID-19 patients (Fig. S4B) , and the results showed that CyPA expression was significantly elevated in CD147 high regions than CD147 low regions ( Fig. S4C ). The expression of CD147 had a strong positive correlation with CyPA ( Fig. S4D) , which further verified that CD147 upregulated CyPA expression upon SARS-CoV-2 infection. CyPA is an intracellular protein that can also be secreted extracellular, and its most advantageous place to exert its role as a driver of COVID-19 CSS is extracellular. Therefore, CyPA antibody was used to block the extracellular CyPA to evaluate whether the expression of cytokines and chemokines would be affected. Evidently, the anti-CyPA antibody reduced the expression of IL6 and CCL2 ( CyPA induced an increase in p-p38 and c-FOS expression, which could be reversed by either anti-CyPA antibody, anti-CD147 antibody or MAPK inhibitor (Fig. 4I ). In addition, the expression of IL6 and CCL2 was effectively inhibited by the above three agents (Fig. 4J) . These results indicate that CyPA induces cytokine expression through the CD147-mediated MAPK pathway (Fig. 4K ). To confirm the clinical significance of CyPA in the inflammatory responses during COVID-19, a cytokine chip assay was performed and showed that 32 out of 40 cytokines were upregulated in plasma of COVID-19 patients (n = 200, including 41 severe/critical and 159 mild), compared with healthy individuals (n = 100) ( Fig. 5A and S5). Meanwhile, CyPA level in plasma analyzed by ELISA was significantly higher in severe/critical COVID-19 patients (n = 106, comprising the above 41 severe/critical cases) than in mild (n = 159) and healthy cohorts (n = 100) (Fig. 5B) . Further, CyPA level in clinical samples was significantly correlated with those above upregulated cytokine level, especially IL17, CCL2 and IL6 (Fig. 5C ). We then collected lung tissues from 13 cases of COVID-19 patients through puncture or autopsy and stained them with H&E staining (Fig. S6A ). In order to further clarify the correlation between CyPA and infiltrating cells and cytokines, multiplex immunofluorescence staining was carried out to detect the expression of CyPA, macrophage marker CD68, CCL2 and IL6 ( Fig. 5D and G). Each collected image was divided into nine process regions, and the MFI of each marker in each region was analyzed with Halo software and correlation analysis was performed ( CyPA promotes IL6 expression. These results indicate that CyPA is involved in inflammatory responses in COVID-19. In addition, there was a correlation between IL6 and macrophages, but the correlation coefficient was only 0.5 (Fig. 5I ), further indicating that not only the macrophages secrete IL6, but other cells in the lung tissues of COVID-19 could also highly express IL6. To verify whether CD147 functions as a therapeutic target for COVID-19, we treated SARS-CoV-2-infected hCD147 mice with CD147 antibody (Meplazumab) (Fig. S7A ). CD147 antibody effectively cleared the virus and alveolar exudation, resolving the pneumonia at 6 d.p.i. (Fig. 6A , B and S7B). CD147 antibody treatment had a favorable therapeutic effect on the downregulation of cytokines and chemokines at 6 d.p.i., such as IL6, IL10, IL1b, CCL2, CXCL1, CXCL2, IFN-γ, IL-17 and CyPA ( Fig. 6C ). Infiltration of macrophages and neutrophils was decreased after the CD147 antibody treatment (Fig. 6D) . These results indicate that CD147 antibody is effective against COVID-19, supportive of our exploratory phase II clinical trial results (28) . Therefore, CD147 can be a promising target for the treatment of COVID-19 patients, especially those with severe pneumonia. SARS-CoV-2 variants are raising global concerns not only because of their increased transmission but also because of their extensive mutations in spike protein that could lead to antigenic changes detrimental to neutralization of antibody therapy and vaccine protection. D614G, occurred at the initial stage of the pandemic, is the most common mutant, whichmakes the virus more infectious and transmissible (29) . N501Y, commonly appeared in B.1.1.7, B.1.351 and P.1 variants, allows the virus to more readily bind to ACE2 (30) . E484K, first found in B.1.351 variant, helps the virus slip past the body's immune defense (31) . The uncertainty of spike protein mutation considerably undermines the efficacy of therapeutics and vaccination, thus calling for new strategies against SARS-COV-2. Previous studies in our lab have found that the RBD of spike can bind with CD147 to help SARS-CoV-2 enter host cells (18) . Our study showed that the four mutations located in the RBD domain, including N501Y, E484K, Y453F and N439K, do not change the affinity of RBD to CD147, indicating that CD147 still served as a receptor for SARS-CoV-2 variants. More importantly, we confirmed that CD147 antibody can effectively block the infection of the prevalent variants, such as B.1.1.7, B.1.351, P.1, B.1.526 and B.1.617, at an inhibition rate comparable to that of SARS-CoV-2. As such, CD147 is a shared receptor for SARS-CoV-2 and its variants, and CD147 antibody is a specific and effective receptor antagonist against COVID-19. At present, another constraint for COVID-19 control is the lack of sufficient understanding of its pathogenesis. We investigated the pathogenesis of SARS-CoV-2 using hCD147 transgenic mice. Consistent with reports by others (14, 32) , we observed interstitial pneumonia in virus-infected hACE2 mice, while hCD147 mice Cytokine storm caused by SARS-CoV-2 is a main cause of death in severe COVID-19 patients, and Th17 cell response contributes to the cytokine storm during viral infection, which results in tissue damage and pulmonary edema (35) . The infection of SARS-CoV-2 via CD147 can induce Th17 cell response, suggesting that CD147 is the key factor to cytokine storm. In this process, CyPA is an important intermediate product, which is considered to be a critical proinflammatory factor (36) . The spike protein of SARS-CoV-2 binding to CD147 activates JAK-STAT3 pathway, thus promoting the expression of CyPA. A high level of CyPA not only induces chemotaxis in numerous monocytes and neutrophils to the inflammatory site, but also activates the MAPK signaling pathway by binding with CD147 to promote the massive production of cytokines, such as CCL2 and IL6. Hence, in the process of SARS-CoV-2 infection, CyPA is a key mediator that enhances the inflammatory response, and CD147, as a regulatory molecule of JAK-STAT3 upstream, triggers the cytokine storm. Taken together, CD147 plays a central role in SARS-CoV-2 and its variants infection and pathogenesis of COVID-19. CD147 is not only a viral entry receptor, but also a signal initiator for cytokine storm. Accordingly, blocking CD147 inhibits SARS-CoV-2 infection and suppresses cytokine storm. In conclusion, CD147 is a core target for SARS-CoV-2 variants and CD147 antibody is a specific and effective drug to control ever-spreading COVID-19 epidemic. The authors have no competing interests to declare. Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Zhi-Nan Chen (znchen@fmmu.edu.cn). All materials and reagents generated in this study are available from the Lead Contact with a completed Materials Transfer Agreement. The data that support the findings of this study are available from the corresponding author upon request, and has been uploaded to GEO (GSE167403). The SARS-CoV-2 strain (BetaCov/human/CHN/Beijing_IME-BJ05/2020) used for in VeroE6 cells were transfected with supernatants containing lentiviruses encoding the shCD147 and shACE2 constructs (GeneChem Co. Ltd.) using Lipofectamine 2000 reagent (Invitrogen) to generate the CD147 and ACE2 knockdown cell lines, respectively. After 48 h, the infected cells were selected with 3 μg/ml puromycin, and monoclonal cells were selected for further study. CD147-/-ACE2-/-VeroE6 cells were generated using the CRISPR/Cas9 system (GeneChem Co. Ltd). After being anaesthetized, each mouse was intranasally inoculated with SARS-CoV-2 at a TCID50 of 3 × 10 5 . For CD147 antibody treatment, 3 mg/kg CD147 antibody (Meplazumab) was administered via the tail vein at 1 d.p.i. Lung tissues were collected for RNA extraction or fixed with paraformaldehyde and PLP Fixing Solution for H&E staining, IHC, immunofluorescence staining and TEM. Mouse lung samples were fixed with 4% paraformaldehyde, embedded in paraffin and cut into 3-µm-thick sections. Formalin-fixed paraffin-embedded mouse lung tissue sections were deparaffinized with xylene and alcohol. Fixed tissue samples were used for haematoxylin-eosin (H&E) staining and indirect IHC. For H&E staining, the slides were counterstained with haematoxylin for 15 min and eosin for 10 min. For IHC, the slides were deparaffinized and rehydrated, followed by 15 min of blocking with 5% goat serum in phosphate-buffered saline (PBS) and staining with primary antibodies for 2 h. Immunoperoxidase staining was performed using a General SP kit (SP-9000, ZSGB-BIO), and the sections were then treated with 3,3'-diaminobenzidine (ZLI-9019, ZSGB-BIO) to detect the target proteins, followed by counterstaining of nuclei with haematoxylin. Multiplex immunofluorescence analyses were performed using 3-μm-thick sections of formalin-fixed and paraffin-embedded tissues. Briefly, slides were deparaffinized in xylene and hydrated in ethanol at gradient concentrations. After heat-induced antigen retrieval in either citrate (pH = 6) or Tris-EDTA (pH = 9) buffer, the samples were permeabilized with 0.5% Triton X-100, blocked with 5% goat serum-PBS and stained with primary antibodies. Nuclei were stained with DAPI. A TSA-indirect kit was used according to the manufacturer's instructions (PerkinElmer). Image analysis was performed using HALO software (Akoya Biosciences). Forty human cytokines were detected by Quantibody® array kits (QAH-INF-3, RayBiotech) according to the manufacturer's protocol. Then, 100 µl of standard cytokines or samples was added to each well after blocking, followed by incubation with the arrays at room temperature for 2 h. The wells were washed, and 80 µl of the detection antibody cocktail was added to each well and incubated at room temperature for 1 h. Eighty microlitres of Cy3 equivalent dye-conjugated streptavidin was added to each well and incubated in the dark at room temperature for 1 h. The signals were visualized by Axon GenePix. The data were analysed by microarray analysis software (GenePix, ScanArray Express, ArrayVision and MicroVigene). The content of CyPA in plasma was detected using ELISA kits according to the manufacturer's instructions (Eiaab Science, Inc., 0979 h). In brief, the microtiter plate provided in the kit was precoated with a biotin-conjugated antibody specific to the target CyPA. Standards were established at 6 different dilutions ranging from 0 ng/ml to 10 ng/ml, and nonimmune serum was used as a negative control. Next, 100 µl of the plasma, standard or negative control was added to each well of the plate and incubated for 2 h at 37°C. Then, avidin conjugated to horseradish peroxidase (HRP) was added to each well and incubated. The reaction was developed by adding 100 µl of 3,3',5,5'-tetramethylbenzidine (TMB) to the wells for 12 min at room temperature. Finally, 200 M sulfuric acid was added to stop the reaction, and the absorbance at 450 nm was determined using a BioTeck Epoch microplate reader. RNA was isolated using MyOne Silane Dynabeads (Thermo Fisher Scientific, Inc.). Illumina adaptors. The primers were removed using Agencourt AMPure XP (Beckman Coulter/Agencourt), and the samples were amplified by 14 PCR cycles. The libraries were gel purified and quantified using a Qubit high-sensitivity DNA kit (Invitrogen), and the library quality was confirmed using Tapestation High-Sensitivity DNA tapes (Agilent Technologies). The RNA-seq reactions were sequenced (50-base pair reads) on an Illumina NextSeq platform (Illumina) according to the manufacturer's instructions, and analysis was performed using the CLC Genomics Workbench v.8.0.1 RNA-seq analysis software package (Qiagen). Briefly, the reads were aligned (mismatch cost = 2, insertion cost = 3, deletion cost = 3, length fraction = 0.8, similarity fraction = 0.8) to the mouse genome, and differential expression analysis was performed (total count filter cut-off = 5.0). The results were normalized to reads per million, and Gene-e (Broad Institute) was used to generate heat maps. The datasets generated in the current study are available from the corresponding author upon reasonable request. Total RNA was extracted using Total RNA Kit II (Omega Bio-tek) according to the manufacturer's instructions. cDNA was synthesized from 1 μg of total RNA using an RNA reverse transcriptase kit (Takara). Real-time qPCR was performed using the ABI Ex Taq II (Takara) was used for amplification according to the manufacturer's instructions. The cDNA inputs were standardized, and PCR was performed for 40 cycles. The mouse organ tissues were homogenized to detect the virus gene copy number by qPCR (TaqMan Universal Master Mix II with UNG, 4440044, Applied Biosystems). The sequences of the primers are listed in Table 2 . SPR analysis was performed on a Biacore 3000 system (Biacore). In brief, His-CyPA (produced by our laboratory) was fixed to the surface of CM5 sensor chips (GE Healthcare Bio-Sciences AB) by an Amine Coupling Kit (GE Healthcare, BR-1000-50). The interaction between CyPA and CD147 was detected using the kinetic analysis/concentration series/direct binding mode; the flow rate was set to 15 μl/min, and the binding and dissociation time was 3 min. The same methods were applied to detect the interactions between CD147 and the SARS-CoV-2 spike proteins with mutations (N501Y, E484K, Y453F and N439K) (SinoBio, 40592-V08H8, 40592-V08H84, 40592-V08H80, 40592-V08H14) except that His-CD147 (produced by our laboratory) was fixed to the CM5 sensor chip surface. The results were analysed by BIAevaluation software (Biacore) to determine the affinity constants. Co-IP assays were performed using a Pierce® Co-Immunoprecipitation Kit (26149, Thermo Fisher Scientific, Inc.) according to the manufacturer's protocol. A mouse anti-human CyPA antibody (50 μg) and mouse anti-human CD147 antibody (Jiangsu Pacific Meinuoke Biopharmaceutical Co. Ltd, China, 50 μg) were used for antibody immobilization. The eluted proteins were detected by Western blot. Having boiled for 5 min, the eluted proteins were loaded onto a 12% SDS-PAGE gel and then transferred to a PVDF membrane (Millipore). After blocking with 5% nonfat milk for 1 h, the membrane was incubated with the corresponding primary antibodies at 4°C overnight. The images were developed following incubation with the secondary antibody (goat anti-mouse IgG (H+L), 31430, Thermo Fisher Scientific, 1:5000 dilution) at room temperature for 1 h. The lung tissues were cut into 1 mm 3 pieces and fixed with PLP Fixing Solution (G2220, Solarbio) for at least 24 h. After washing with PBS, the sections were treated with osmic acid for 1.5 h. Then, the tissues were dehydrated with alcohol at gradient concentrations and soaked in acetone for 15 min. After embedding and polymerization overnight with epoxy resin, the slices were cut with an ultrathin slicer and glued to a perforated copper grid. Finally, each section was stained with lead citrate and uranium solution for 10 min, washed with water and dried. The ultrastructure was observed by transmission electron microscopy (JEM-1230, JEOLLTD). The cells were cultured at 37°C under 5% CO2 overnight, and the cell medium was then replaced by medium containing the virus. After the cells were infected for 1 h at 37°C, the virus supernatant was discarded, and the cells were washed twice with PBS. Finally, the cells were cultured with 2% FBS maintenance medium with or without CD147 antibody (Meplazumab, 15 or 60 μg/ml). At 48 h after infection, viral and cellular RNA were extracted together and detected by RT-qPCR. The SARS-CoV-2 pseudovirus and its variants (N501Y, N501Y-D614G, B.1.1.7 1.7, B.1.351, B.1.525, B.1.526(S477N) The cytokines detected by the protein chip were analysed by moderated t-statistics. ELISA data were measured using a parameter logistic curve. All multiplex fluorescence and immunohistochemical staining images were analysed by HALO Image Analysis Software, and significant differences were calculated by unpaired t tests. Correlations were analysed by Pearson correlation coefficients. Significant differences in cell and mouse tests were analysed by unpaired t tests with a two-tailed distribution. p < 0.05 was considered to be statistically significant. Statistical analyses were performed using GraphPad Prism, version 8.0. The heat map of 32 differential cytokines. 40 human cytokines were detected in G-CSF (pg/ml) S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 S12 S13 S14 S15 S16 S17 S18 S19 S20 S21 S22 S23 S24 S25 S26 S27 S28 S29 S30 S31 S32 S33 S34 S35 S36 S37 S38 S39 S40 S41 S42 S43 S44 S45 S46 S47 S48 S49 S50 S51 S52 S53 S54 S55 S56 S57 S58 S59 S60 S61 S62 S63 S64 S65 S66 S67 S68 S69 S70 S71 S72 S73 S74 S75 S76 S77 S78 S79 S80 S81 S82 S83 S84 S85 S86 S87 S88 S89 S90 S91 S92 S93 S94 S95 S96 S97 S98 S99 S100 S102 S103 S104 S105 S106 S107 S108 S109 S110 S112 S113 S114 S115 S117 S118 S119 S120 S121 S122 S124 S125 S126 S127 S128 S129 S130 S131 S132 S133 S135 S136 S137 S138 S139 S141 S142 S143 S144 S145 S147 S149 S150 S151 S152 S153 S154 S155 S156 S159 S160 S162 S163 S164 S165 S167 S168 S169 S171 S173 S174 S176 S179 S180 S183 S184 S186 S189 S190 S191 S194 S195 S196 S197 S198 S200 S201 S202 S204 S205 S206 S207 S209 S210 S211 S212 S213 S214 S217 S218 S219 S220 S222 S224 S225 S226 S227 S228 S229 S231 S232 S233 S234 S236 S237 S238 S239 S240 S241 S242 S243 S244 S245 S247 S248 S249 S250 S251 S252 S253 S254 S256 S257 S258 S259 S260 S261 S262 S264 S266 S267 S268 S270 S271 S272 S273 S274 S275 S277 S278 S279 S280 S283 S284 S285 S286 S287 S288 S289 S290 S291 S292 S293 S294 S295 S296 S297 S298 S299 S300 S101 S111 S116 S123 S134 S140 S146 S148 S157 S158 S161 S166 S170 S172 S175 S177 S178 S181 S182 S185 S187 S188 S192 S193 S199 Nanobody cocktails potently neutralize SARS-CoV-2 D614G N501Y variant and protect mice The E484K mutation in the SARS-CoV-2 spike protein reduces but does not abolish neutralizing activity of human convalescent and post-vaccination sera. medRxiv Identification of a SARS-CoV-2 Variant with L452R and E484Q Neutralization Resistance Mutations Resistance of SARS-CoV-2 variants to neutralization by monoclonal and serum-derived polyclonal antibodies Clinical characteristics and risk factors associated with COVID-19 disease severity in patients with cancer in Wuhan, China: a multicentre, retrospective, cohort study Clinical Characteristics of Coronavirus Disease 2019 in China Fibrotic progression and radiologic correlation in matched lung samples from COVID-19 post-mortems Tracking the time course of pathological patterns of lung injury in severe COVID-19 Histopathology features of the lung in COVID-19 patients Impaired type I interferon activity and inflammatory responses in severe COVID-19 patients Immunopathological characteristics of coronavirus disease 2019 cases in Guangzhou Characteristics of Peripheral Lymphocyte Subset Alteration in COVID-19 Pneumonia A SARS-CoV-2 Infection Model in Mice Demonstrates Protection by Neutralizing Antibodies The pathogenicity of SARS-CoV-2 in hACE2 transgenic mice Role of the GTNGTKR motif in the N-terminal receptor-binding domain of the SARS-CoV-2 spike protein Lethality of SARS-CoV-2 infection in K18 human angiotensin-converting enzyme 2 transgenic mice Comparison of transgenic and adenovirus hACE2 mouse models for SARS-CoV-2 infection CD147-spike protein is a novel route for SARS-CoV-2 infection to host cells Systems biological assessment of immunity to mild versus severe COVID-19 infection in humans Cytokine and Chemokine Levels in Coronavirus Disease 2019 Convalescent Plasma. Open Forum Infect Dis Systems biological assessment of immunity to mild versus severe COVID-19 infection in humans A Mouse Model of SARS-CoV-2 Infection and Pathogenesis Potential role of cyclophilin A in regulating cytokine secretion IL-37) Reduces High Glucose-Induced Inflammation, Oxidative Stress, and Apoptosis of Podocytes by Inhibiting the STAT3-Cyclophilin A (CypA) Signaling Pathway A genomic regulatory element that directs assembly and function of immune-specific AP-1-IRF complexes Mitogen-activated protein kinase pathways mediated by ERK, JNK, and p38 protein kinases Oxidative stress and vascular smooth muscle cell growth: a mechanistic linkage by cyclophilin A Meplazumab treats COVID-19 pneumonia: an open-labelled, concurrent controlled add-on clinical trial. medRxiv Human neutralising antibodies elicited by SARS-CoV-2 non-D614G variants offer cross-protection against the SARS-CoV-2 D614G variant The new SARS-CoV-2 strain shows a stronger binding affinity to ACE2 due to N501Y mutant SARS-CoV-2 spike E484K mutation reduces antibody neutralisation K18-hACE2 mice develop respiratory disease resembling severe COVID-19 TH17 responses in cytokine storm of COVID-19: An emerging target of JAK2 inhibitor Fedratinib Repurposing Treatment of Wernicke-Korsakoff Syndrome for Th-17 Cell Immune Storm Syndrome and Neurological Symptoms in COVID-19: Thiamine Efficacy and Safety, In-Vitro Evidence and Pharmacokinetic Profile Inflammatory Cytokine: IL-17A Signaling Pathway in Patients Present with COVID-19 and Current Treatment Strategy Potential role of cyclophilin A in regulating cytokine secretion We thank Prof. Chuan Qin from the Chinese Academy of