Elevated expression of TfR in respiratory tract and lung tissue of monkey and mouse infected by SARS-CoV-2
TfR is ubiquitously expressed 23. Given that respiratory tract is a susceptible site for SARS-CoV-2 infection, we employed qRT-PCR and western blot techniques to detect the expression of TfR and ACE2 in several tissues including respiratory tract (nasal cavity, trachea, and lung) and liver. In both RNA and protein levels, the expression of both TfR and ACE2 were significantly elevated in trachea and lung as compared with other tissues (Fig. 1A-C). As illustrated in Fig. 1D and E, TfR was upregulated in lung tissue of SARS-CoV-2 infected monkey and humanized ACE2 (hACE2) mice by immunohistochemical analysis.
Direct interactions among virus spike protein, ACE2 and TfR
Surface plasmon resonance (SPR) was used to study the interaction between TfR and the virus spike protein. As illustrated in Fig. 2A and B, TfR directly interacted with the virus spike protein through enzyme-linked immunosorbent assay (ELISA) and SPR analysis, whereby, the association rate constant (Ka), dissociation rate constant (Kd), and equilibrium dissociation constant (KD) values for the interaction between TfR and spike were 2.69× 105 M−1s−1, 7.92 × 10−4 s−1and 2.95 nM, respectively. The effect of TfR on the SARS-CoV-2 spike RBD, which binds to the cell receptor ACE224,25, was also assayed using SPR (Fig. 2C). The KD value between TfR and SARS-CoV-2 spike RBD was~43 nM, which is a bit weaker compared with the binding affinity between TfR and SARS-CoV-2 spike.
Based on the TfR structures 26 and the virus spike protein 27, we made a docking model of TfR-spike interaction (Fig. S1). According to the model, two designed peptides (SL8: SKVEKLTL; QK8: QDSNWASK) were used to interfere with the binding of TfR to spike (Table S1). As illustrated in Fig. 2D, these peptides inhibited TfR-Spike interaction.
As illustrated in Fig. 2E and F, TfR also directly interacted with ACE2 through SPR and native polyacrylamidegel electrophoresis (PAGE) analysis, and the Ka, Kd, and KD values for the interaction between TfR and ACE2 were 6.33× 104 M−1s−1, 1.25 × 10−2 s−1 and 200 nM, respectively. Based on the TfR, ACE2, and spike protein structures, we made docking models of TfR-ACE2 (Fig. S2) and TfR-ACE2-Spike interactions (Fig. S3). According to these models, two inhibitory peptides (SL8 and FG8: FPFLAYSG) were designed to interfere with TfR-ACE2 complex formation (Table S2). As illustrated in Fig. 2G, these peptides inhibited TfR-ACE2 interaction determined by SPR. Notably, co-immunoprecipitation analysis revealed that SL8, but not scrambled peptide of SL8 (SL8-scr), interfered with TfR-ACE2-Spike complex formation (Fig. 2H), indicating that SL8 affected both interactions of TfR-ACE2 and TfR-Spike.
In vitro direct interaction between TfR and SARS-CoV-2 has been confirmed as reported above. We next investigated the interaction between TfR and SARS-CoV-2 on cell membranes. As illustrated in Fig. 2I, high density of TfR was found on the membranes of Vero E6 cells. Following the infection of SARS-CoV-2 to the Vero E6 cells, significant co-localization of TfR and SARS-CoV-2 was observed on the cell membranes and in the cytoplasma (Fig. 2I), suggesting that TfR is a membrane receptor for SARS-CoV-2. Further study indicated that TfR was also co-localized with ACE2 on the membranes of both infected and uninfected cells by the virus. Importantly, in the infected cells by the virus, the co-localization of TfR, ACE2, and virus was observed on cell membranes (Fig. 2J), but only TfR-virus co-localization was observed in the cytoplasma, suggesting that the virus is transported into cytoplasma by TfR.
Interferences of TfR-Spike interaction inhibit SARS-CoV-2 infection
Soluble TfR, Tf, anti-TfR antibody and the designed peptides as mentioned above were used to test their effects on SARS-CoV-2 infections by cytopathic effect (CPE)-based anti-viral assay. As expected, CPE inhibition and quantitative RT-PCR (qRT-PCR) assays indicated that all of them blocked the virus infections to Vero E6 cells (Fig. 3). The concentration to inhibit 50 % viral entry (EC50) determined by CPE assays was 80, 125 and 50 nM (Fig. 3B, D, and F) for soluble TfR, Tf and anti-TfR antibody, while that was 93, 160 and 16.6 nM (Fig. 3C, E, and G) determined by RT-PCR, respectively. There was no cytotoxicity even in their concentration up to 1000 nM (Fig. 3B-G). Notably, the anti-viral effect of 200 nM anti-TfR antibody was comparable to that of high concentration of Remdeivir (4 mM). In addition, the designed peptides showed strong ability to inhibit the viral entry (Fig. 3H-K). At the concentration of 80 μM, SL8, QK8, and FG8 inhibited 87, 99, and 75 % virus infection, respectively (Fig. 3H-K).
SARS-CoV-2 infects ACE2 knockout cells
As illustrated in Fig. 4A and B, ACE2 in Vero E6 and A549 cells were successfully knocked out. Importantly, ACE2-knockout Vero E6 and A549 cells were infected by SARS-CoV-2, and infections in ACE2-knockout Vero E6 and A549 cells were inhibited by anti-TfR antibody (Fig. 4C and D). As illustrated in Fig. S4, the TfR expression levels were first validated by western blot and TfR overexpression promoted virus infection, which was inhibited by TfR down-regulation (Fig. 4E and F).
Humanized TfR (hTfR) mice are sensitized for SARS-CoV-2 infection
The adenoviral vector (AD5) expressing human TfR was constructed as the methods described28. Mice were transduced intranasally with Ad5-hTfR, and human TfR expression in lung tissue was validated by western blot (Fig. S5). As illustrated in Fig. 5A, elevated viral replication was detected in lung tissue of hTfR mice at 1, 3, and 5 dpi. Viral infection caused a decrease in mouse body weight, and hTfR showed an obvious decrease in body weight than wild-type mice (Fig. 5B). As illustrated in Fig. 5C, hTfR mice showed more severe vascular congestion and hemorrhage than wild-type mice. In addition, histopathological examination of the lungs sections indicated that hTfR mice showed typical interstitial pneumonia, characterized by infiltration of significant macrophages and lymphocytes into the alveolar interstitium, and accumulation of macrophages in alveolar cavities (Fig. 5D and E).
Anti-TfR antibody shows promising anit-SARS-CoV-2 effects
As illustrated in Fig. 6A, anti-TfR antibody administration inhibited viral replication in mice lung tissue at 3 and 5 dpi, whereas the isotype control IgG administration showed no effects on it. Anti-TfR antibody administration inhibited the decrease in mouse body weight caused by viral infection (Fig. 6B). Histopathological examination of the lungs sections indicated that mice in the control group showed typical interstitial pneumonia (Fig. 6C and D). Anti-TfR antibody administration showed significant protective effects and prevented histopathological injuries caused by virus infection compared with control IgG (Fig. 6C and D).