TMPRSS2-mediated SARS-CoV-2 uptake boosts innate immune activation, enhances cytopathology, and drives convergent virus evolution

Abstract

The accessory protease transmembrane protease serine 2 (TMPRSS2) enhances severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) uptake into ACE2-expressing cells, although how increased entry impacts downstream viral and host processes remains unclear. To investigate this in more detail, we performed infection assays in engineered cells promoting ACE2-mediated entry with and without TMPRSS2 coexpression. Electron microscopy and inhibitor experiments indicated TMPRSS2-mediated cell entry was associated with increased virion internalization into endosomes, and partially dependent upon clathrin-mediated endocytosis. TMPRSS2 increased panvariant uptake efficiency and enhanced early rates of virus replication, transcription, and secretion, with variant-specific profiles observed. On the host side, transcriptional profiling confirmed the magnitude of infection-induced antiviral and proinflammatory responses were linked to uptake efficiency, with TMPRSS2-assisted entry boosting early antiviral responses. In addition, TMPRSS2-enhanced infections increased rates of cytopathology, apoptosis, and necrosis and modulated virus secretion kinetics in a variant-specific manner. On the virus side, convergent signatures of cell-uptake-dependent innate immune induction were recorded in viral genomes, manifesting as switches in dominant coupled Nsp3 residues whose frequencies were correlated to the magnitude of the cellular response to infection. Experimentally, we demonstrated that selected Nsp3 mutations conferred enhanced interferon antagonism. More broadly, we show that TMPRSS2 orthologues from evolutionarily diverse mammals facilitate panvariant enhancement of cell uptake. In summary, our study uncovers previously unreported associations, linking cell entry efficiency to innate immune activation kinetics, cell death rates, virus secretion dynamics, and convergent selection of viral mutations. These data expand our understanding of TMPRSS2's role in the SARS-CoV-2 life cycle and confirm its broader significance in zoonotic reservoirs and animal models.

Keywords: SARS-CoV-2 entry; cytolytic responses; host-species range; innate immunity; viral evolution.

Fig. 1.

TMPRSS2 enhances SARS-CoV-2 internalization into endosomes. (A) Conservation of S2’ site between variants. (B) Cartoon of human TMPRSS2 protein domains. Location of catalytic triad residues and autocleavage site (scissors) are highlighted above. (C) SARS-CoV-2 pp cell entry of the indicated S variants. Left panel: RLU counts at in HEK-293T-ACE2 cells at 48 hpi. RLU: Relative light units. LOQ: limit of quantification. (N = 4 mean ± SEM). Right panel: Western blot quantification of S protein expression in SARS-CoV-2 containing supernatants. (D) Uptake of the indicated SARS-CoV-2 pps in the presence of TMPRSS2 or TMPRSS2 (ΔHDS). (N = 4 mean ± SEM. **P < 0.01, ***P < 0.001, ****P < 0.0001). (E) Human ACE2 and TMPRSS2 mRNA (Top, N = 3) and protein expression (Bottom) in the indicated cell lines. (F) Authentic B.1 vRNA levels in the indicated cell lines at 4 hpi (Left) and 8 hpi (Right) in the presence of the indicated inhibitors, normalized to vRNA levels in DMSO control treated cells. MOI = 0.01. N = 3 (mean ± SEM). (G) EM visualization of authentic B.1 virus entry into A549-A and A549-AT cells. Virus uptake by endocytosis is observed at 5 and 10 min in both cell lines, with B.1 binding to CCPs (Left and Middle panels; black arrowheads). In both cell lines at 10 min, B.1 is observed in CCVs (Middle and Right panels; white arrowheads) and endocytic compartments (Right panels; white arrowheads). In A549-A cells at 10 min (Top Right), the black arrow highlights a virus–host membrane fusion event. In A549-AT cells at 10 min (Bottom Right) clusters of virions accumulate in endosomal compartments (white arrowheads). PM: plasma membrane; CCV: clathrin-coated vesicle; CCP: clathrin-coated pit. (H) Quantification of extracellular and internalized virions. Normalized virion counts per µm/membrane for extracellular plasma membrane–associated virions (blue) or internalized endosomal virions (orange). Infections were performed as in (G), in the presence of either DMSO or Pitstop 2 (100 µM), and image quantification was performed at 10 min post temperature shift. MOI = 500. (N = 24 mean ± SEM. *P < 0.05, ****P < 0.0001).

Fig. 2.

TMPRSS2 expression increases panvariant replication and modulates virus secretion. (A) TMPRSS2 expression increases virus replication and secretion rates in a variant-dependent manner. Infections were performed at MOI 0.01 with the indicated SARS-CoV-2 strains, and supernatants and cellular RNAs harvested at 1, 2, 4, 8, and 24 hpi. Top panels: Intracellular N gene copies determined by RT-qPCR. (N = 3 mean ± SEM). Bottom panels: secreted virus in supernatants. TCID₅₀. Tissue culture infectious dose 50. (N = 3 mean ± SEM. *P < 0.05, **P < 0.01). LOQ: limit of quantification. n.d.: not detectable. (B) Virus mapped reads as a percentage of total reads. Infections were performed with the indicated SARS-CoV-2 strains at low (0.01) and high (1) MOI, with cellular RNAs harvested at 24 hpi for RNA-seq. (C) Mean SARS-CoV-2 vRNA sequencing depth. (D) Direct comparison of mean viral read coverage across the respective SARS-CoV-2 genomes in A549-A and A549-AT cells. Cartoon directly below represents the relative locations of genome-encoded ORFs.

Fig. 3.

TMPRSS2-mediated entry increases the magnitude of early antiviral responses. (A) Principal component analysis (PCA) of the indicated transcriptomes. (B) Dot-plot summarizes cell processes and signaling pathways targeted upon infection. FDR: false discovery rate. N = 3. (C) Heat map visualizes fold change (log₂) in mRNA expression for selected host transcripts associated with different cellular processes after SARS-CoV-2 infection of A549-A and A549-AT cells. Fold change is relative to expression levels in the respective uninfected control cell line and represents mean values from N = 3 biological replicates. (D) Virus production rates in the indicated cell lines (MOI 0.01). Left panel: Virion production kinetics 0 to 96 hpi. Right panel: Cumulative virus production. (N = 3 mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001. n.s.: not significant). LOQ: limit of quantification. n.d.: not detectable. (E) Cell viability of the indicated cell lines at 72 hpi, determined by flow cytometry. (N = 3 mean ± SEM).

Fig. 4.

Convergent mutational switching at two Nsp3 residues confers enhanced interferon antagonism. (A) Codon switching in Orf1a/Orf1ab. Sequence logos depict codon-triplet base frequencies at codons 930 and 936 for B.1 infections at low MOI. Top plots: A549-A infections. Bottom plots: A549-AT infections. (B) Cartoon depicting the protein domains of Nsp3. UL: ubiquitin-like domain, HVR: hypervariable region, ADRP: ADP-ribose phosphatase, SUD: SARS-unique domain, PLpro: papain-like protease, NAB: nucleic acid binding domain, TM: multipass transmembrane domain, and ZF: zinc finger motif. Zoom of an 11 amino acid stretch of the Nsp3 acid HVR highlighting positions D112 and C118 is positioned above. (C) Amino acid frequency plots for the indicated viruses and cell lines highlight shifts in dominant Nsp3 residues 112 (Left) and 118 (Middle), compared to cellular gene dysregulation under the same conditions (Right). (D) Residue frequencies at Nsp3 112 (Top) and 118 (Bottom) are significantly correlated to cell gene dysregulation. (E) Phenotypic characterization of Nsp3 mutants. Left panel. Normalized luciferase induction in the indicated cell lines after Nsp3-EGFP fusion protein encoding plasmid transfection and subsequent treatment with PolyI:C or IFNα2a (N = 3 mean ± SEM. **P < 0.01, ***P < 0.001). Right panel. Quantification of EGFP expression in HEK-293 T and A549-dual cells (N = 3 mean ± SEM. n.s.: not significant).

Fig. 5.

TMPRSS2 proteins from diverse mammals enhance SARS-CoV-2 uptake. (A) Phylogenetic tree depicts the evolutionary relationships of TMPRSS2 coding sequences. Branch lengths are proportional to nucleotide substitutions per site, as determined by the scale bar. Significantly supported groupings (>70%) are labeled at internal nodes and species selected for further investigation are highlighted. (B) Left panels. Amino acid variability in mammalian TMPRSS2 orthologues is superimposed onto the predicted human TMPRSS2 structure. Right panel. Amino acid differences (red) and conservation (blue) between mammalian and zebrafish TMPRSS2 orthologues are superimposed onto the predicted zebrafish TMPRSS2 structure. (C) Conservation of transmembrane domains and catalytic triad residues in humans and nonhuman species. Top panel. Hydrophobicity plotting identifies a single transmembrane domain in humans and nonhuman species. Bottom Left panel. Transmembrane domain (blue) and catalytic triad residues (red) highlighted on the predicted human TMPRSS2 structure. TM. Transmembrane. Bottom Right panel. Amino acid alignment confirms conservation of serine protease catalytic triad residues in humans and nonhuman species. (D) SARS-CoV-2 pp entry into HEK-293T-ACE2 cells expressing the indicated TMPRSS2 orthologues, normalized to infection rates in empty vector transduced cells. T2: TMPRSS2. (N = 4 mean ± SEM. *P < 0.05, **P < 0.01, n.s.: not significant) (E) Viral replicative kinetics in A549-A cells transduced with TMPRSS2 orthologues from 10 species. A549-A cells were transduced with equal amounts of p24-enveloped pps encoding each TMPRSS2 orthologue, prior to infections with B.1, B.1.617.2, and B.1.1.529 viruses. For all viruses, N gene copies were determined at time points 4, 8, and 24 hpi.