HEK293T (ATCC, CRL-3216), A549 (ATCC, CCL-185), Huh7 (JCRB, JCRB0403), Calu-3 (ATCC, HTB-55), and Vero E6 (ATCC, CRL-1586) cells were maintained in DMEM (Thermo Fisher Scientific, United States) supplemented with 10% FBS (Thermo Fisher Scientific, United States), 100 units/ml penicillin, and 100 μg/ml streptomycin (Thermo Fisher Scientific, United States) at 37°C and 5% CO₂. These cells expressed considerable amounts of human angiotensin-converting enzyme 2 (hACE2). The HEK293T-hACE2^high cell line was generated by infecting wildtype HEK293T cells with lentiviruses which expressed hACE2. The hACE2-positive cells were sorted by fluorescence activated cell sorting (FACS) and confirmed by western blot with antibodies against hACE2. The HEK293T-hACE2^high cells were maintained as wildtype HEK293T cells. The HEK293T-hACE2^high/TMPRSS2^high cells and the Calu-3-CTSL^high cells were constructed similarly. All cells have been tested for mycoplasma by PCR-based assay and confirmed to be mycoplasma-free (Mycoplasma-F: 5′-GGGAGCAAACAGGATTAGTATCCCT-3′; Mycoplasma-R:5′-TGCACCATCTGTCACTCTGTTACCCTC-3′).

The plasmid expressing the S of SARS-CoV-2 (D614; Wuhan-Hu-1, GISAID: EPI_ISL_402125) was purchased from Generay Biotech company (Shanghai, China) and inserted into the pcDNA3.1 vector. SARS-CoV-2 S/HIV-1 pseudotyped viruses were packaged by co-transfecting a lentiviral construct pHIV-Luciferase (Addgene plasmid # 21375), a packaging construct psPAX2 (Addgene plasmid # 12260) and plasmid expressing S proteins into HEK293T cells. The luciferase gene within the pHIV-Luciferase vector is under the control of the EF-1α promoter and can be expressed after pseudotyped virus infection. Thus, the pseudotyped virus system can be used to indicate the infection of SARS-CoV-2 S/HIV-1 pseudotyped viruses. The culture medium was replaced with fresh DMEM at 6 h post transfection. The pseudotyped viruses-containing supernatant was collected 48 h post transfection and filtered through 0.45 μm filters. The amounts of pseudotyped viruses were quantified by RT-qPCR assays with primers against the long-term repeat (LTR) sequences of viral RNAs (HIVTotal-F: 5′-CTGGCTAACTAGGGAACCCACTGCT-LTR) sequences of viral RNAs (HIVTotal-F: 5′-CTGGCTAACTAGGGAACCCACTGCT-3′; HIVTotal-R: 5′-GCTTCAGCAAGCCGAGTCCTGCGTC-3′). These RT-qPCR experiments were performed as biological triplicate. Pseudotyped viruses including SARS-CoV S/HIV-1 and VSV-G/HIV-1 were packaged and quantified as pseudotyped SARS-CoV-2 S/HIV-1 viruses. Different pseudotyped SARS-CoV-2 S/HIV-1 mutants were also packaged and quantified as above, the S proteins of which included those of G614 virus (SYSU-IHV, EPI_ISL_444969), B.1.1.7 (Alpha, GISAID: EPI_ISL_581117), B.1.351 (Beta, EPI_ISL_678597), P.1 (Gamma, EPI_ISL_792683), B.1.429 (Epsilon, EPI_ISL_1675148), B.1.525 (Eta, EPI_ISL_1093465), B.1.526 (Iota, EPI_ISL_1080752), B.1.617.1 (Kappa, EPI_ISL_1372093), B.1.617.2 (Delta, EPI_ISL_1337507), B.1.621 (Mu, EPI_ISL_1220045) and C.37 (Lambda, EPI_ISL_1534645).

Patient-derived SARS-CoV-2 D614 virus (Wuhan-Hu-1) was obtained from Guangdong Provincial Center for Disease Control and Prevention. The SARS-CoV-2 G614 virus (SYSU-IHV) was isolated from the sputum sample of a female admitted in Guangzhou Eighth People’s Hospital who was infected at Guangzhou by an African traveler in April 2020. Vero E6 cells were utilized to propagate these viruses.

Pseudotyped viruses including SARS-CoV S/HIV-1, SARS-CoV-2 S/HIV-1, VSV-G/HIV-1, and 10 different SARS-CoV-2 mutants were packaged as above. For viral infection in siRNA-mediated gene knock-down experiment, targeted genes in HEK293T cells were firstly knocked down by a mixture of three different siRNAs targeting each gene, followed by the infection of pseudotyped SARS-CoV-2 S/HIV-1 viruses 24 h post transfection. Another 24 h later, siRNA- and virus-treated cells were lysed and measured for the amounts of luciferase proteins by luciferase reporter assays which could represent the percentages of virus-infected cells.

For virus infection in protein overexpression experiment, wildtype HEK293T cells were firstly transfected with CTSL-expressing or hACE2-expressing plasmids, followed by infection with pseudotyped SARS-CoV-2 S/HIV-1 viruses. The amounts of luciferase within each group were measured utilizing luminometer (Promega, United States) 48 h post infection and represented as relative luminescence units.

For virus infection in drug treatment experiment, HEK293T-hACE2^high cells were incubated with serially diluted drugs including teicoplanin (Selleck, S1399, United States), dalbavancin (Selleck, S4848, United States), vancomycin (Selleck, S2575, United States), and camostat (Selleck, S2874, United States), and different pseudotyped viruses. The amounts of luciferase within each group were measured 48 h post infection.

HEK293T-hACE2^high cells were seeded in 12-well plates. Twenty-four hours post seeding, cells were co-incubated with SARS-CoV-2 D614 (Wuhan-Hu-1) virus and twofold serially diluted teicoplanin. Another 48 h post incubation, the supernatant in each group was collected and proceeded to RNA extraction with RNeasy Mini Kit (QIAGEN, 74104, Germany) according to the manufacturer’s instruction. SARS-CoV-2 viral RNA copies were determined by one-stepSARS-CoV-2 RNA detection kit (PCR-Fluorescence Probing) (Da An Gene Co., DA0931, China) with the following primers and probes: N-F (5′-CAGTAGGGGAACTTCTCCTGCT-3′), N-R (5′-CTTTGCTGCTGCTTGACAGA-3′), and N-P (5′-FAM-CTGGCAATGGCGGTGATGCTGC-BHQ1-3′). These RT-PCR experiments were performed as biological triplicate. The half maximal inhibitory concentration (IC50) of teicoplanin against SARS-CoV-2 D614 virus was calculated by GraphPad software (San Diego, United States) according to these viral RNA copies within each group. The IC50 of teicoplanin against SARS-CoV-2 G614 virus (SYSU-IHV) was determined similarly. SARS-CoV-2 virus infection assays were conducted in the BSL-3 facility of Sun Yat-sen University.

To determine whether teicoplanin targeted the virus directly or targeted the host cell indirectly, HEK293T-hACE2^high cells were treated with drug or virus in three different ways. In the first group, HEK293T-hACE2^high cells were pre-treated with 0, 12.5, 25, and 50 μM teicoplanin, respectively. Four hours post drug treatment, cells were infected with pseudotyped SARS-CoV-2 S/HIV-1 virus. Another 48 h later, the amounts of luciferase within each group were monitored and represented as relative luminescence units. In the second group, HEK293T-hACE2^high cells were co-treated with different concentrations of teicoplanin and pseudotyped SARS-CoV-2 S/HIV-1 virus simultaneously. The amounts of luciferase were measured 48 h post co-treatment. In the third group, cells were pre-infected with pseudotyped SARS-CoV-2 S/HIV-1 virus. Four hours post infection, cells were treated with different concentrations of teicoplanin. Another 48 h later, cells were lysed and measured for the amounts of luciferase.

To evaluate the CTSL enzymatic activity upon teicoplanin treatment in vitro, the purified 250 ng CTSL proteins (Sino Biological Inc., 10486-H08H, China) were added into the CTSL assay buffer (400 mM NaAc, 4 mM EDTA, 8 mM DTT, and pH 5.5) and incubated in ice for 15 min for CTSL activation. To evaluate the inhibition of teicoplanin on CTSL activity, the purified 250 ng CTSL proteins were also co-incubated with 50 μM teicoplanin (Selleck, S1399, United States). After activation, about 2 μg in vitro purified SARS-CoV S (Sino Biological Inc., 40634-V08B, China) or SARS-CoV-2 S (Sino Biological Inc., 40589-V08B1, China) were added into each group. S protein only group (without CTSL and teicoplanin) was set as control group. The S-CTSL-teicoplanin mixtures were incubated at 37°C for 1.5 h. The enzymatic reaction was stopped by adding with SDS-PAGE loading buffer and followed by boiling at 100°C for 10 min. Digested proteins were proceeded to SDS-PAGE and analyzed by silver staining (Sigma-Aldrich, PROTSIL2-1KT, United States).

Eight-week-old specific-pathogen-free (SPF) transgenic hACE2 mice (C57BL/6) were purchased from GemPharmatech Co., Ltd. (Cat No.: T037657). All mice were housed in SPF facilities at the Laboratory Animal Center of Sun Yat-sen University. Animal experiments were conducted in strict compliance with the guidelines and regulations of the Laboratory Monitoring Committee of Guangdong Province of China. The Ethics Committees of Guangdong Provincial People’s Hospital and Sun Yat-sen University approved the animal experiments.

Six hours before the viral challenge, four hACE2 mice were intraperitoneally administrated with 100 mg/kg body weight teicoplanin (dissolved in saline). Four hACE2 mice in the control group were intraperitoneally administrated with equal volume of saline. Six hours later, all the mice were intranasally challenged with 1 × 10⁵ focus-forming units (FFU) of SARS-CoV-2 D614 virus. Another 5 days later, mice were euthanized to harvest lung tissues. Viral RNA copies in lung tissues were quantified by one-step SARS-CoV-2 RNA detection kit (PCR-Fluorescence Probing) (Da An Gene Co., DA0931, China). These RT-PCR experiments were performed as biological quadruplicate. SARS-CoV-2 challenge studies were approved by the Ethics Committee of Zhongshan School of Medicine of Sun Yat-sen University on Laboratory Animal Care (Assurance Number: SYSU-IACUC-2021-B0020).

Lung tissues of SARS-CoV-2-infected mice were fixed in 4% paraformaldehyde for at least 2 days. These lung tissues were embedded in paraffin and proceeded to histopathology and immunohistochemistry analysis (Nanjing FreeThinking Biotechnology Co., Ltd., China). For histopathology analysis, sections (3–4 μm) of lung tissues were stained with H&E. For immunohistochemistry analysis, sections of lung tissues were deparaffinized and rehydrated with xylene and gradient alcohol. Antigens were retrieved in citric acid buffer (pH 6.0) and quenched with 3% H₂O₂. After blocking with BSA, sections were incubated with rabbit anti-SARS-CoV-2 Nucleoprotein (N) for 24 h at 4°C, followed by incubating with goat anti-rabbit IgG secondary antibody (HRP-conjugated) and staining with 3,3′-diaminobenzidine. Antibody-conjugated sections were stained with hematoxylin, followed by dehydrating with gradient ethanol. Samples were covered by neutral balsam and imaged with HS6 microscope (Sunny Optical Technology Co., Ltd., China).

The genome sequences of SARS-CoV and SARS-CoV-2 were collected from the GenBank database¹ and the GISAID’s EpiCoV database.² The sequences of SARS-CoV circulating in 2003 contain 6 strains (accession numbers: AY278488, AY545918, AY545917, AY394977, AY394978, and AY394979). The sequences of SARS-CoV-2 include 13 variants: D614 (Wuhan-Hu-1, GISAID: EPI_ISL_402125), G614 (SYSU-IHV, EPI_ISL_444969), B.1.1.7 (Alpha, GISAID: EPI_ISL_581117), B.1.351 (Beta, EPI_ISL_678597), P.1 (Gamma, EPI_ISL_792683), B.1.429 (Epsilon, EPI_ISL_1675148), B.1.525 (Eta, EPI_ISL_1093465), B.1.526 (Iota, EPI_ISL_1080752), B.1.617.1 (Kappa, EPI_ISL_1372093), B.1.617.2 (Delta, EPI_ISL_1337507), B.1.621 (Mu, EPI_ISL_1220045), C.37 (Lambda, EPI_ISL_1534645), and B.1.1.529 (Omicron, EPI_ISL_6752027). The S gene sequences were obtained from the genome of SARS-CoV and SARS-CoV-2 according to the annotation in the GenBank database. The sequence datasets were aligned using the ClustalW program implemented in MEGA X software. Consensus sequences were created using BioEdit software³ based on the multiple alignment of SARS-CoV and SARS-CoV-2. The amino acid sequence logos were generated by WebLogo.

All the measurements in this study have been performed for at least three biological replicates by at least two lab technicians or students. Detailed statistical information including statistical tests, sample numbers, mean values, standard errors of the mean (SEM) and p-values have been shown in the main text and figure legends. Statistical analysis was conducted with Graphpad Prism 8.0 or Microsoft Excel. Biological triplicate and quadruplicate data were presented as mean ± SEM. A value of p ≥ 0.05 was not statistically significant and represented as “ns.” A value of p < 0.05 was statistically significant and represented as asterisk (*). A value of p < 0.01 was more statistically significant and represented as double asterisks (^**). A value of p < 0.001 was the most statistically significant and represented as triple asterisks (^***). When comparing mean differences between groups which were split by one independent variable, one-way ANOVA with Tukey’s multiple comparison test or Dunnett’s multiple comparison test was conducted. When comparing mean differences between groups which were split by two independent variables, two-way ANOVA with Tukey’s multiple comparisons test or Dunnett’s multiple comparisons test was conducted. For data with a normal distribution, we used Student’s t-test.

The proteolytic processing of the S protein is essential for SARS-CoV entry and fusion. Many host proteases, including TMPRSS2 and CTSL, are involved in the priming and activation of the SARS-CoV S protein, and some of which also have been identified and experimentally validated in SARS-CoV-2 infection (Bosch et al., 2008; Glowacka et al., 2011; Simmons et al., 2013; Hoffmann et al., 2020; Johnson et al., 2021; Wei et al., 2021; Zhao et al., 2021; Zhu et al., 2021). To systematically identify cellular proteases and receptors which mediated the entry and fusion of SARS-CoV-2 to target cells, we knocked down ten major proteases and receptors in HEK293T cells, which included CTSL, CTSB, CTSK, TMPRSS2, TMPRSS11A, TMPRSS11D, Furin, PLG, DPP4, and ACE2. Subsequently, these cells were infected with pseudotyped SARS-CoV-2 S/HIV-1 viruses which harbored an integrated luciferase gene. The expression of luciferase indicated the entry and expression of the pseudotyped virus. We found that the absence of CTSL, TMPRSS2, Furin, or ACE2 significantly decreased the pseudotyped SARS-CoV-2 virus infection (Figure 1A). The ACE2 protein has been identified as the major receptor of SARS-CoV-2 (Zhou et al., 2020). While the HEK293T cell line is a highly transfectable derivate of HEK293 cells, a cell line isolated from the kidney of a human embryo (HEK) that constitutively expresses the ACE2. Both TMPRSS2 and Furin also have been found to be essential for the efficient infection of SARS-CoV-2 (Hoffmann et al., 2020; Johnson et al., 2021). To determine whether CTSL is also involved in SARS-CoV-2 S protein activation, we first compared the cleavage sites of CTSL in the gene sequences encoding the SARS-CoV and SARS-CoV-2 S proteins. After alignment, we found that the cleavage sites of CTSL, the motif of which was AYT| MSL, were well-conserved between SARS-CoV and SARS-CoV-2 S proteins (Figure 1B). Moreover, the cleavage sites of CTSL on the S proteins were also highly conserved among all the major epidemic SARS-CoV-2 variants including D614 (Wuhan-Hu-1), G614 (SYSU-IHV), B.1.1.7 (Alpha), B.1.351 (Beta), P.1 (Gamma), B.1.429 (Epsilon), B.1.525 (Eta), B.1.526 (Iota), B.1.617.1 (Kappa), B.1.617.2 (Delta), B.1.621 (Mu), C.37 (Lambda), and B.1.1.529 (Omicron) (Candido et al., 2020; Cherian et al., 2021; He et al., 2021; Kimura et al., 2021; Laiton-Donato et al., 2021; McCallum et al., 2021; Meng et al., 2021; Ozer et al., 2021; Tegally et al., 2021; Zhou et al., 2021; Figure 1C). Previously, CTSL has been found to play pivotal roles in SARS-CoV infection by cleaving AYT| MSL motif within S proteins, which resulted in the activation of S proteins (Simmons et al., 2005; Bosch et al., 2008; Belouzard et al., 2012). We speculated that CTSL might also participate in SARS-CoV-2 entry and fusion utilizing the same proteolytic mechanism. Thus, we overexpressed CTSL proteins in HEK293T cells which were subsequently infected with pseudotyped SARS-CoV-2 S/HIV-1 viruses. We found that the infectivity of pseudotyped virus to HEK293T cells was linearly and positively correlated with the expression level of CTSL (Figure 1D). We also co-overexpressed ACE2 with CTSL in HEK293T cells. We found that the co-overexpression of CTSL significantly increased the infectivity of pseudotyped virus compared with ACE2-overexpression only (Figure 1E). These results indicated that the effective infection of SARS-CoV-2 to host cells depended on the proteolytic activity of CTSL.

Previously, we have found that teicoplanin, a glycopeptide antibiotic which inhibited CTSL activity, suppressed the entry of SARS-CoV, MERS-CoV, and Ebola viruses (Zhou et al., 2016). We speculated that teicoplanin might also be able to block the entry of SARS-CoV-2. Thus, we conducted the pseudotyped virus entry upon drug treatment assay. We generated a highly sensitive HEK293T-hACE2^high cell line which constitutively expressed high level of hACE2 receptors. HEK293T-hACE2^high cells were co-incubated with teicoplanin and pseudotyped SARS-CoV-2 S/HIV-1 virus. The infectivity of pseudotyped virus, which was represented by the amounts of luciferase within HEK293T-hACE2^high cells, was measured 48 h post infection (Figure 2A). To exclude the possibility that teicoplanin inhibited the early events of the pseudotyped HIV-1 life cycle, pseudotyped VSV-G/HIV-1 viruses bearing vesicular stomatitis virus (VSV) glycoproteins were also packaged and treated as the negative control. Pseudotyped SARS-CoV S/HIV-1 viruses bearing SARS-CoV S were packaged and treated as the positive control. The results showed that teicoplanin effectively inhibited the entry of both SARS-CoV-2 and SARS-CoV pseudotyped viruses in a dose-dependent manner, whereas teicoplanin treatment did not affect the infection of pseudotyped VSV-G/HIV-1 viruses (Figures 2B,C).

Teicoplanin homologs including dalbavancin also have specific inhibitory effects on CTSL based on our previous study (Zhou et al., 2016). While vancomycin, another glycopeptide antibiotic which was clinically used for Gram-positive bacterial infections, did not show inhibitory activity on CTSL. Therefore, we further tested whether dalbavancin and vancomycin could inhibit the entry of SARS-CoV-2. Similar to teicoplanin, dalbavancin effectively inhibited both SARS-CoV-2 and SARS-CoV pseudotyped viruses entering into HEK293T-hACE2^high cells in a dose-dependent manner, but it did not affect pseudotyped VSV-G/HIV-1 virus infection (Figures 2D,E). In contrast, vancomycin did not show any inhibitory activity on the infection of pseudotyped SARS-CoV-2, SARS-CoV, or VSV-G viruses (Figures 2F,G). Taken together, these results indicated that CTSL inhibitors teicoplanin and its homolog dalbavancin could suppress the entry of SARS-CoV-2.

To evaluate whether teicoplanin also inhibited SARS-CoV-2 infection in other susceptible cell lines, we infected A549 cells and Huh7 cells with pseudotyped SARS-CoV-2 S/HIV-1 viruses, accompanied by the treatment of serially diluted teicoplanin. The results showed that teicoplanin also effectively inhibited pseudotyped SARS-CoV-2 S/HIV-1 virus infection in both cell lines with a half maximal inhibitory concentration (IC₅₀) of 3.164 μM for the A549 cell line and an IC₅₀ of 1.885 μM for the Huh7 cell lines (Figures 3A,B). To further confirm that teicoplanin inhibited the entry of SARS-CoV-2, we investigated the antiviral activity of teicoplanin on authentic (live) SARS-CoV-2. We obtained two SARS-CoV-2 strains. One was SARS-CoV-2 D614 virus (Wuhan-Hu-1) which was provided by Guangdong Provincial Center for Disease Control and Prevention (GDCDC). The other was SARS-CoV-2 G614 virus (SYSU-IHV) which was isolated by us from the sputum sample of an infected patient (Ma et al., 2020). We found that teicoplanin effectively inhibited the entry of both strains with an IC₅₀ of 2.038 μM for the Wuhan-Hu-1 reference strain and an IC₅₀ of 2.116 μM for the SARS-CoV-2 (D614G) variant (Figures 3C,D). Given that the serum concentrations of teicoplanin in patients are at least 15 mg/L (8.78 μM) after the loading dose treatment for most Gram-positive bacterial infections, our data indicated that teicoplanin was able to potently suppress the entry of SARS-CoV-2 of both the original Wuhan-Hu-1 strain and the D614G mutation strain at a relatively low and safe dose (Zhou et al., 2016).

To elucidate whether the target of teicoplanin was the virus itself, or the host cell, or both, we conducted drug/virus pre-treatment assay. In the first group, HEK293T-hACE2^high cells were pre-treated with different concentrations of teicoplanin followed by infection with pseudotyped SARS-CoV-2 (drug pre-treatment group). In the second group, cells were co-incubated with both teicoplanin and pseudotyped virus (drug-virus co-treatment group). In the third group, cells were pre-infected with pseudotyped virus followed by treatment with teicoplanin (virus pre-treatment group). We found that the infectivity of pseudotyped SARS-CoV-2 in both drug pre-treatment group and drug-virus co-treatment group was negatively correlated with the concentration of teicoplanin, whereas the infectivity of pseudotyped virus in the virus pre-treatment group was almost unchanged upon the treatment of different concentrations of teicoplanin (Figure 3E). These results indicated that teicoplanin targeted host cells rather than viral particles.

Previous reports have shown that TMPRSS2 can activate S proteins for complete fusion through the cell surface entry route (Laporte et al., 2021). While CTSL activates S proteins to drive the fusion of viral membranes and endosomal membranes through the endosomal entry route. To identify whether TMPRSS2 and CTSL contributed to SARS-CoV-2 infection independently, we knocked down CTSL and TMPRSS2 in two different cell lines. One was HEK293T-hACE2^high cell line, which expressed high level of CTSL and low level of TMPRSS2 (Hoffmann et al., 2020). The other was Calu-3 cell line, which expressed high level of TMPRSS2 and low level of CTSL (Laporte et al., 2021). We found that the depletion of CTSL completely aborted the infection of pseudotyped SARS-CoV-2 S/HIV-1 viruses in CTSL⁺/TMPRSS2^– HEK293T-hACE2^high cells, whereas the depletion of TMPRSS2 did not abort the infection of pseudotyped SARS-CoV-2 S/HIV-1 viruses in the same cell line (Supplementary Figure 1A). The combination treatment of siCTSL and siTMPRSS2 also aborted the infection in HEK293T-hACE2^high cells, similar to the sole treatment of siCTSL. However, the depletion of CTSL in CTSL^–/TMPRSS2⁺ Calu-3 cells only slightly decreased the infection of pseudotyped SARS-CoV-2 S/HIV-1 viruses (Supplementary Figure 1B). The depletion of TMPRSS2 as well as the co-treatment of siCTSL and siTMPRSS2 could significantly abort the infection of pseudotyped SARS-CoV-2 in Calu-3 cells (Supplementary Figure 1B). These results indicated that CTSL and TMPRSS2 contributed to the SARS-CoV-2 infection in different cell types.

Our previous report has revealed that teicoplanin targets on CTSL directly within host cells (Zhou et al., 2016). While camostat inhibits the activity of TMPRSS2 (Kawase et al., 2012). To evaluate whether teicoplanin and camostat inhibited SARS-CoV-2 infection in different cell types, we firstly treated CTSL⁺/TMPRSS2^– HEK293T-hACE2^high cells with teicoplanin, camostat, or both. We found that the treatment of teicoplanin or the co-treatment of teicoplanin and camostat could significantly abort pseudotyped SARS-CoV-2 S/HIV-1 virus infection (Figure 4A). However, the sole treatment of camostat was unable to prevent the infection of pseudotyped SARS-CoV-2 in HEK293T-hACE2^high cells (Figure 4A). On the contrary, the sole treatment of camostat or the co-treatment of teicoplanin and camostat significantly inhibited pseudotyped SARS-CoV-2 S/HIV-1 virus infection in CTSL^–/TMPRSS2⁺ Calu-3 cells (Figure 4B). While the sole treatment of teicoplanin was unable to inhibit the infection of pseudotyped SARS-CoV-2 in Calu-3 cells (Figure 4B). To further confirm that teicoplanin and camostat inhibited SARS-CoV-2 infection by targeting CTSL and TMPRSS2, respectively, we constructed an HEK293T-hACE2^high/TMPRSS2^high cell line which expressed a high level of TMPRSS2. We found that neither teicoplanin nor camostat could totally abort pseudotyped SARS-CoV-2 S/HIV-1 virus infection in HEK293T-hACE2^high/TMPRSS2^high cells (Figure 4C). While the combination of teicoplanin and camostat was able to significantly inhibit the infection of pseudotyped SARS-CoV-2 in the same cell line (Figure 4C). We also constructed a Calu-3-CTSL^high cell line which expressed a high level of CTSL. Similarly, only when the Calu-3-CTSL^high cells were co-treated with both teicoplanin and camostat could totally abort pseudotyped SARS-CoV-2 S/HIV-1 virus infection (Figure 4D). The sole use of teicoplanin or camostat could only partially inhibit the infection of pseudotyped SARS-CoV-2 (Figure 4D). Overall, teicoplanin and camostat inhibited SARS-CoV-2 infection by inhibiting the activity of CTSL and TMPRSS2, respectively.

To provide direct evidence that teicoplanin inhibiting SARS-CoV-2 entry via inhibiting the activity of CTSL, we conducted in vitro CTSL enzymatic inhibition assay. The in vitro purified CTSL proteins were firstly activated in CTSL assay buffer, followed by incubating with SARS-CoV S or SARS-CoV-2 S proteins (Supplementary Figures 2A–C). In another group, pre-activated CTSL proteins were co-incubated with different S proteins and teicoplanin. Then the CTSL- and teicoplanin-treated S proteins were proceeded to SDS-PAGE and silver staining. We found that CTSL proteins were able to effectively cleave both SARS-CoV S and SARS-CoV-2 S (Figures 4E,F). However, the co-treatment of teicoplanin with CTSL proteins inhibited the enzymatic activity of CTSL on both S proteins, resulting in the presence of more full-length S proteins (Figures 4E,F). Our above results further confirmed that teicoplanin inhibited the entry of SARS-CoV-2 by directly inhibiting the proteolytic activity of CTSL within host cells.

Since December 2019, many SARS-CoV-2 mutants have emerged locally and spread worldwide, such as B.1.1.7 (Alpha), B.1.351 (Beta), P.1 (Gamma), B.1.429 (Epsilon), B.1.525 (Eta), B.1.526 (Iota), B.1.617.1 (Kappa), B.1.617.2 (Delta), B.1.621 (Mu), C.37 (Lambda), and B.1.1.529 (Omicron) (Candido et al., 2020; Cherian et al., 2021; He et al., 2021; Kimura et al., 2021; Laiton-Donato et al., 2021; McCallum et al., 2021; Meng et al., 2021; Ozer et al., 2021; Tegally et al., 2021; Zhou et al., 2021). We have found that the cleavage sites of CTSL on S proteins of these mutants were highly conserved (Figure 1C). Thus, we speculated that teicoplanin-mediated inhibition of CTSL activity might also cripple the cell entry of SARS-CoV-2 mutants. To evaluate whether teicoplanin still was able to inhibit the infection of these variants, we constructed different S-expressing plasmids which were derived from various SARS-CoV-2 mutants (Figure 5A). Like the package of pseudotyped SARS-CoV-2 (D614) S/HIV-1 viruses, we packaged ten different pseudotyped SARS-CoV-2 viruses based on the above S mutants. HEK293T-hACE2^high cells were co-incubated with different pseudotyped viruses and twofold serially diluted teicoplanin, followed by the measurement of the luciferase activity which could represent viral infectivity. The IC₅₀ of teicoplanin against different pseudotyped viruses was calculated based on the percentages of viral inhibition. We found that all the IC₅₀ of teicoplanin against the entry of these viruses were below 5 μM (3.002 μM for B.1.351/Beta, 3.117 μM for P.1/Gamma, 3.056 μM for B.1.429/Epsilon, 2.041 μM for B.1.525/Eta, 1.963 μM for B.1.526/Iota, 2.188 μM for B.1.617.1/Kappa, 2.300 μM for B.1.617.2/Delta, 1.998 μM for B.1.621/Mu, 2.306 μM for C.37/Lambda), except that the IC₅₀ of teicoplanin against pseudotyped SARS-CoV-2 (B.1.1.7/Alpha) S/HIV-1 viruses was 5.423 μM (Figures 5B–K). Taken together, our above results indicated that the CTSL inhibitor teicoplanin was able to inhibit the entry of different SARS-CoV-2 variants.

To evaluate whether the treatment of teicoplanin could protect individuals from SARS-CoV-2 infection, we conducted mice infection experiments upon teicoplanin treatment. We utilized K18-hACE2 mice which were generated by knocking in the human K18 promoter-driven human ACE2 within the mouse Hipp11 (H11) “safe-harbor” locus. hACE2 mice were intraperitoneally administrated with 100 mg/kg body weight teicoplanin or equal volume of saline, followed by intranasally challenging with 1 × 10⁵ focus-forming units (FFU) of SARS-CoV-2 D614 virus (n = 4 in each group). These mice were euthanized 5 days post infection (Figure 6A). Lung tissues of each mouse were proceeded to SARS-CoV-2 viral RNA quantification, H&E and immunohistochemistry (IHC) analysis. We found that lung tissues of hACE2 mice in saline group harbored large amounts of viral RNA copies (1.9 × 10⁴, 2.3 × 10⁵, 1.2 × 10⁵, and 8.8 × 10⁴ copies per ml for each infected mouse) (Figure 6B). While lung tissues of hACE2 mice in teicoplanin group harbored only few numbers of viral RNA copies (less than 10 copies per ml in average) (Figure 6B). HE and IHC assays also revealed that lung tissues in mice from the saline group were severely damaged upon SARS-CoV-2 challenge, which were interspersed with thickened alveolar septa, collapsed alveoli, and Nucleoprotein (N) protein-expressing cells (Figure 6C). Whereas no pathological changes and N-expressing cells were observed in teicoplanin treatment group (Figure 6C). These results indicated that teicoplanin treatment was able to prevent the infection of SARS-CoV-2 virus in hACE2 mice.