TMPRSS2 and ADAM17 cleave ACE2 differentially and only proteolysis by TMPRSS2 augments entry driven by the severe acute respiratory syndrome coronavirus spike protein

Abstract

The type II transmembrane serine proteases TMPRSS2 and HAT can cleave and activate the spike protein (S) of the severe acute respiratory syndrome coronavirus (SARS-CoV) for membrane fusion. In addition, these proteases cleave the viral receptor, the carboxypeptidase angiotensin-converting enzyme 2 (ACE2), and it was proposed that ACE2 cleavage augments viral infectivity. However, no mechanistic insights into this process were obtained and the relevance of ACE2 cleavage for SARS-CoV S protein (SARS-S) activation has not been determined. Here, we show that arginine and lysine residues within ACE2 amino acids 697 to 716 are essential for cleavage by TMPRSS2 and HAT and that ACE2 processing is required for augmentation of SARS-S-driven entry by these proteases. In contrast, ACE2 cleavage was dispensable for activation of the viral S protein. Expression of TMPRSS2 increased cellular uptake of soluble SARS-S, suggesting that protease-dependent augmentation of viral entry might be due to increased uptake of virions into target cells. Finally, TMPRSS2 was found to compete with the metalloprotease ADAM17 for ACE2 processing, but only cleavage by TMPRSS2 resulted in augmented SARS-S-driven entry. Collectively, our results in conjunction with those of previous studies indicate that TMPRSS2 and potentially related proteases promote SARS-CoV entry by two separate mechanisms: ACE2 cleavage, which might promote viral uptake, and SARS-S cleavage, which activates the S protein for membrane fusion. These observations have interesting implications for the development of novel therapeutics. In addition, they should spur efforts to determine whether receptor cleavage promotes entry of other coronaviruses, which use peptidases as entry receptors.

FIG 1

Cleavage of ACE2 by type II transmembrane serine proteases. (A) Plasmids encoding ACE2 and the indicated proteases were transiently cotransfected into 293T cells. Cells cotransfected with empty plasmid (pcDNA) or transfected with empty plasmid alone served as negative controls. The transfected cells were lysed and the lysates analyzed by Western blotting using an ACE2 monoclonal antibody directed against the ACE2 ectodomain (top panel) or a polyclonal antibody directed against the C terminus of ACE2 (middle panel). Detection of β-actin in cell lysates served as a loading control (bottom panel). (B) The experiment was carried out as described for panel A, but proteases from the indicated species were analyzed. The results of two gels run in parallel are shown. (C) The experiment was carried out as described for panel A, but different amounts of plasmid encoding TMPRSS2 and HAT wild types or catalytically inactive proteases (TMPRSS2 mut, HAT mut) were cotransfected. Ecto, ectodomain; cyto, cytoplasmic domain; h, human; m, mouse; ch, chicken; sw, swine.

FIG 2

Residue R621 is the HAT cleavage site in recombinant ACE2 but is not essential for cleavage of cellular ACE2 by TMPRSS2 and HAT. (A) Coomassie-stained SDS-PAGE gel showing that recombinant ACE2 (apparently migrating as a double band around 90 kDa in this gel system, bands 1 to 4) is cleaved into an 80-kDa fragment (band 5) upon incubation with HAT. Note that the 80-kDa fragment does not appear as a double band, suggesting that the differences in the electrophoretic mobility of intact ACE2 may be due to ragged C-terminal sequences (e.g., absence of the 10-His tag as stated in the product information for recombinant ACE2 [R&D Systems]). (B) Bands 1 to 5 were excised and subjected to mass spectrometric peptide mapping with trypsin and endoproteinase Asp-N, respectively. Visualization of peptide assignments to the ACE2 sequence (upper bars, trypsin; lower bars, Asp-N) shows the absence of the C-terminal part (amino acids 603 to 733) in the 80-kDa fragment. (C and D) Mass spectrometric analysis of Asp-N digests of ACE2 (bands 3 and 4) and its cleavage product (band 5). In the zoomed-in mass spectra in panel C, the mass signals at m/z 1,614.96 (highlighted in gray) represent the Asp-N cleavage product 598-DQSIKVRISLKSALG-612 ([M+H]⁺_calc = 1,614.954). Its presence in intact ACE2 (bands 3 and 4, upper and middle panels) together with its absence in the 80-kDa fragment (band 5, lower panel) indicated that the HAT cleavage site resides within this sequence. In the zoomed-in mass spectra in panel D, the mass signal at m/z 845.52 (highlighted in gray) represents the Asp-N cleavage product 598-DQSIKVR-604 ([M+H]⁺_calc = 845.484). Its absence in intact ACE2 (bands 3 and 4, upper and middle panels) together with its presence in the 80-kDa fragment (band 5, lower panel) revealed R604 as the HAT cleavage site. (E) Plasmids encoding the ACE2 wt or the indicated ACE2 mutants jointly with plasmids encoding TMPRSS2 or HAT were transiently cotransfected into 293T cells. The transfected cells were lysed and the lysates analyzed by Western blotting using an ACE2 monoclonal antibody directed against the ACE2 ectodomain (top panel) or a polyclonal antibody directed against the C terminus of ACE2 (middle panel). Detection of β-actin in cell lysates served as a loading control (bottom panel). Ecto, ectodomain; cyto, cytoplasmic domain.

FIG 3

Arginine and lysine residues within ACE2 amino acids 697 to 716 are essential for ACE2 cleavage by TMPRSS2 and HAT. (A) The domain organization of ACE2 is depicted schematically. The membrane-proximal region of ACE2 potentially harboring the TMPRSS2 and HAT cleavage site is highlighted, and its amino acid sequence is provided. Five clusters of arginine and lysine residues were identified in the membrane-proximal sequence, and their position is indicated by boxes. The residues were mutated to alanine, resulting in ACE2 mutants C0, C1, C2, C3, and C4. In addition, mutant C0 was combined with the remaining mutants, resulting in double mutants C0 + C1, C0 + C2, C0 + C3, and C0+ C4. (B and C) Plasmids encoding the ACE2 wt or the indicated ACE2 mutants jointly with plasmids encoding TMPRSS2 or HAT or no protease (pcDNA) were transiently cotransfected into 293T cells. The transfected cells were lysed and the lysates analyzed by Western blotting using an ACE2 monoclonal antibody directed against the ACE2 ectodomain (top panels) or a polyclonal antibody directed against the C terminus of ACE2 (bottom panels). (D) The intensity of the C-terminal cleavage fragment observed upon processing of the ACE2 wt or the indicated ACE2 mutants was quantified via ImageJ software. The average signals measured upon analysis of at least three independent Western blots are shown. Error bars indicate standard errors of the mean (SEM). The signal measured upon cleavage of the ACE2 wt was set as 100%. Ecto, ectodomain; cyto, cytoplasmic domain.

FIG 4

Arginine and lysine residues with ACE2 amino acids 697 to 716 are essential for TMPRSS2- and HAT-dependent augmentation of entry mediated by the SARS-CoV spike protein. (A) Plasmids encoding the ACE2 wt or the indicated ACE2 mutants were transiently transfected into 293T cells, and ACE2 surface expression was detected by FACS. Results of a single experiment are shown and were confirmed in two separate experiments. (B) Plasmids encoding the ACE2 wt or the indicated ACE2 mutants were transiently transfected into 293T cells and the cells transduced with a lentiviral vector pseudotyped with SARS-S. Cells transfected with empty plasmid (pcDNA) served as a negative control, while a vector pseudotyped with VSV-G served as a control for susceptibility to ACE2-independent transduction. Luciferase activities in cell lysates were determined at 72 h postransduction. The results of a representative experiment performed with triplicate samples are shown; error bars indicate standard deviations (SD). Similar results were obtained in two additional experiments. (C) The experiment was carried out as described for panel B, but transduction of target cells expressing the ACE2 wt or the indicated ACE2 mutants jointly with the indicated proteases was assessed. The results are shown as fold enhancement of transduction upon expression of the proteases TMPRSS2 and HAT. The results represent the averages from two to six independent experiments, and error bars indicate SEM.

FIG 5

ACE2 cleavage is dispensable for SARS-S activation by TMPRSS2. Plasmids encoding the ACE2 wt or ACE2 mutant C4 were transiently cotransfected into 293T cells with either plasmid encoding TMPRSS2 or empty plasmid. Subsequently, the cells were incubated with the cathepsin B/L inhibitor MDL28170 or an equal volume of dimethyl sulfoxide (DMSO) and transduced with pseudotypes bearing the indicated glycoproteins. Luciferase activities in cell lysates were determined at 72 h posttransfection. The results of a representative experiment performed with triplicate samples are shown, and error bars indicate SD. Comparable results were obtained in two separate experiments.

FIG 6

TMPRSS2 increases uptake of SARS-S into ACE2-expressing cells. (A) Plasmids encoding the ACE2 wt or TMPRSS2 or empty plasmid (pcDNA) were transiently transfected into 293T cells and the cells incubated with the Fc-tagged S1 subunit of SARS-S (SARS-S1-Fc) for 1 h at 4°C. As a negative control, cells were incubated with the Fc portion alone (Fc). Subsequently, the cells were washed, and the amount of bound proteins was detected by FACS analysis. The geometric mean channel fluorescence was measured, and the signal obtained for SARS-S1-Fc binding to ACE2-transfected cells was set as 100%. The averages from three independent experiments are shown, and error bars indicate SEM. (B) Plasmids encoding ACE2 and either TMPRSS2 or no protease were transiently cotransfected into 293T cells. The transfected cells were incubated with SARS-S1-Fc for 1 h at 4°C or 37°C. Subsequently, the cells were washed, and ACE2 and SARS-S1-Fc were detected by immunofluorescence staining and confocal microscopy. The results are representative of those of three separate experiments. α, anti.

FIG 7

TMPRSS2 suppresses ACE2 shedding by ADAM17. (A) 293T cells were transiently cotransfected with plasmid encoding ACE2 and either empty plasmid (left panels) or plasmid encoding TMPRSS2 (right panels). Subsequently, cells were incubated with medium or medium supplemented with the indicated amounts of PMA, and the presence of ACE2 in culture supernatants (sup; top panel) or cell pellets (pellet; bottom panel) was examined by Western blotting employing an antibody directed against the ACE2 ectodomain. (B) 293T cells were transiently cotransfected with plasmids encoding ACE2 and either the TMPRSS2 wt or catalytically inactive TMPRSS2 (mut) or empty plasmid (pcDNA). Subsequently, supernatants were harvested and proteins precipitated (sup). The cells were pelleted, washed with buffer containing different concentrations of NaCl, and pelleted again, and proteins present in supernatants were precipitated (sup + NaCl). The presence of ACE2 in cell lysates or supernatants was analyzed by Western blotting as described for panel A.

FIG 8

Arginine and lysine residues within ACE2 amino acids 652 to 659 are essential for ACE2 shedding by ADAM17. (A) 293T cells were transiently cotransfected with plasmids encoding the ACE2 wt or the indicated ACE2 mutants jointly with plasmid encoding TMPRSS2 or empty plasmid (pcDNA). The cells cotransfected with empty plasmid were cultivated in either regular medium or medium supplemented with PMA. Subsequently, ACE2 levels in cell lysates were detected by Western blotting using an ACE2 monoclonal antibody directed against the ACE2 ectodomain (top panel) or a polyclonal antibody directed against the C terminus of ACE2 (middle panel). In parallel, the presence of ACE2 in culture supernatants (Sup) was determined using an antibody against the ACE2 ectodomain (bottom panel). (B) A plasmid encoding the ACE2 wt was transiently transfected into 293T cells, and the cells were incubated with a lentiviral vector pseudotyped with SARS-S for 1 h at 4°C. A lentiviral vector bearing no glycoprotein (pcDNA) was included as a negative control. Thereafter, the cells were washed and incubated with medium containing 10 μM PMA or 50 μM TAPI-1 or an equal volume of DMSO at 37°C for 8 h. Cells exposed to bald vector (pcDNA) were incubated in medium alone. Subsequently, the media were replaced and the cells maintained in culture medium without inhibitor. Luciferase activities in cell lysates were determined at 72 h postransduction. The results of a representative experiment carried out with triplicate samples are shown; error bars indicate SD. Comparable results were obtained in two separate experiments.

FIG 9

Role of host cell proteases in the cellular entry of SARS-CoV. Host cell entry of SARS-CoV can proceed via two distinct routes, dependent on the availability of cellular proteases required for activation of SARS-S. The first route is taken if no SARS-S-activating proteases are expressed at the cell surface. Upon binding of virion-associated SARS-S to ACE2, virions are taken up into endosomes, where SARS-S is cleaved and activated by the pH-dependent cysteine protease cathepsin L. The second route of activation can be pursued if the SARS-S-activating protease TMPRSS2 is coexpressed with ACE2 on the surface of target cells. Binding to ACE2 and processing by TMPRSS2 are believed to allow fusion at the cell surface or upon uptake into cellular vesicles but before transport of virions into host cell endosomes. Uptake can be enhanced upon SARS-S activation of ADAM17, which cleaves ACE2, resulting in shedding of ACE2 in culture supernatants. Since normal expression of ACE2 protects from lung injury and ACE2 levels are known to be reduced in SARS-CoV infection, the ADAM17-dependent ACE2 shedding is believed to promote lung pathogenesis. The present study suggests that the cellular uptake of SARS-CoV can also be augmented upon ACE2 cleavage by TMPRSS2. Since the entry-promoting function of ADAM17 is not undisputed and increased uptake upon ACE2 cleavage by TMPRSS2 has only been demonstrated with soluble SARS-S1, the respective arrows are shown in dashed lines.