Efferocytosis of apoptotic bodies drives SARS-CoV-2 infection and macrophage inflammation Authors: Thanh Kha Phan, Dylan Sheerin, Bo Shi, Ngee K. Chua, Sriram Gummadi, Merle Dayton, Liana Mackiewicz, Ana Maluenda, Dilara C. Ozkocak, Georgia K. Atkin-Smith, Nashied Peton, Ching-Seng Ang, Omar Audi, Quan Thinh Le, Thu Uyen Tran, Tuong-Khanh My Tu, Rochelle Tixeira, George W. Ashdown, Kathryn C. Davidson, Pamali Fonseka, Rebecca Feltham, Marcel Doerflinger, Mark D. Hulett, Anna K. Coussens, Ivan K. H. Poon Published: 2026 ABSTRACT SARS-CoV-2 typically utilises host receptor angiotensin-converting enzyme 2 (ACE2) for viral entry. Despite low ACE2 expression, monocyte-derived macrophages, the predominant lung macrophage during severe COVID-19, are often found with SARS-CoV-2 in infected lungs. As macrophage inflammation and cytokine storm are key immunopathological events that drive severe COVID-19, insights into mechanisms underlying viral entry into macrophages are critical to devise novel COVID-19 therapies. Mounting evidence supports that COVID-19 pathogenesis is associated with apoptosis, a type of programmed cell death which releases large extracellular vesicles called apoptotic bodies (ApoBDs). Here, we show that ApoBDs from SARS-CoV-2-infected cells carried infectious virions. Macrophages efferocytose these ApoBDs, enabling SARS-CoV-2 entry and pro-inflammatory responses including inflammasome and NF-κB signalling. To demonstrate targetability of this ApoBD efferocytosis-mediated viral entry, we screen for inhibitors of SARS-CoV-2-induced ApoBD formation and identified T-type voltage-gated calcium channel (T-channel) blockers. Mechanistically, T-channel blockers impair the extracellular calcium influxes required for ApoBD biogenesis. Importantly, blockade of ApoBD formation by T-channel blockers is able to limit cell-to-cell viral transmission, macrophage inflammation and lung immunopathology. Our discovery reveals a novel route for SARS-CoV-2 infection and cytokine storm induction, expanding our understanding of COVID-19 pathogenesis and demonstrating a therapeutic target for infectious diseases. Introduction SARS-CoV-2, the causative agent of COVID-19, typically requires the initial binding of its surface spike protein to the surface receptor angiotensin-converting enzyme 2 (ACE2) on host cells. Following ACE2 engagement, the spike protein folds into a virus-host membrane fusion-promoting state through proteolytic cleavage by trypsin-like proteases such as TMPRSS2 (cell surface entry) or cathepsin L (endosomal entry)1. Severe SARS-CoV-2 infection is typified by an exacerbated pro-inflammatory ‘cytokine storm’, also known as hypercytokinemia (e.g., elevated IL-1β, IL-6, TNF), which develops after peak viral titre2–5. Macrophages, rapid cytokine producers upon pathogen infection6, are a major driver of this COVID-19-associated hypercytokinemia7,8. SARS-CoV-2-infected macrophages and macrophage-associated inflammation have been frequently observed in patients’ lung samples4,9,10. Although ACE2 is detectable on certain lung macrophage subsets and permits canonical ACE2-dependent entry in those cells, ACE2 expression is highly heterogeneous and may be low or transient in other macrophage populations. For instance, infiltrating monocyte-derived macrophages, the predominant lung macrophage during inflammatory stages of severe COVID-1911, are not permissive to SARS-CoV-2 invasion and replication in vitro due to scarce ACE2/TMPRSS2 expression12–15. Notably, lack of viral internalisation and/or low viral replication in macrophages was observed upon SARS-CoV-2 exposure in vitro, resulting in limited pro-inflammatory cytokine release12. Thus, it remains largely unresolved how SARS-CoV-2 gains entry into low ACE2-expressing macrophage subsets and triggers a cytokine storm in vivo5,12,16. Apoptotic cell death underpins many important physiological and pathophysiological processes, including infection. Dying cells have an active role in their own clearance through the release of ‘find-me’ signals (e.g., ATP, UTP) to recruit phagocytes to the sites of cell death and the exposure of ‘eat-me’ signals (e.g., phosphatidylserine) that engage efferocytic receptors17. To further aid apoptotic cell clearance, apoptosing cells often break into ‘bite-sized’ (1–5 μm) membrane-bound extracellular vesicles (EVs) called apoptotic bodies (ApoBDs) to facilitate efficient dead cell removal via the process known as efferocytosis by professional phagocytes (e.g., macrophages) or by neighbouring non-professional phagocytes (e.g., epithelial cells)18–20. Recently, we have shown that ApoBD formation is not stochastic as traditionally thought, but rather a highly coordinated process termed apoptotic cell disassembly with three key morphological steps. Firstly, driven by Rho-associated kinase 1 (ROCK1), apoptotic cells form membrane blebs on the cell surface19,21,22. The cell membrane then extends, resulting in the generation of long protrusions that radiate from the blebs. These protrusions are known as apoptopodia, and their formation is negatively regulated by the membrane channel pannexin 1 (PANX1)18,23. ApoBDs are then released via the severing of apoptopodia24. Notably, ApoBDs can also mediate intercellular communication and elicit a wide range of potent cellular responses in recipient cells25,26. Moreover, certain pathogens such as influenza A virus, African swine fever virus, and hepatitis C virus actively induce apoptosis in infected host cells to propagate viral infection to neighbouring cells through ApoBD-associated virions27–29. However, the role of ApoBD formation in other apoptosis-associated infection settings, particularly during SARS-CoV-2 infection30–32, remains largely understudied. SARS-CoV-2 is known to trigger apoptotic pathways through its accessory viral proteins (ORFs 3a, 6, 7a, 7b, 8a and 9b) and structural proteins (M and N)33. In this study, we report that induction of apoptosis by SARS-CoV-2 infection leads to robust formation of ApoBDs that contain infectious virions. Importantly, ApoBDs released upon SARS-CoV-2 infection can facilitate viral entry into macrophages via an efferocytosis-dependent mechanism. Importantly, SARS-CoV-2-ApoBD-engulfing macrophages but not SARS-CoV-2-exposed macrophages secreted high levels of pro-inflammatory cytokines. Through a drug library screen, we identified the ability of small molecule inhibitors of voltage-gated T-type calcium channels (T channels) to impair ApoBD formation through blockade of calcium influxes. Of the greatest clinical significance, we further demonstrated that mibefradil, a previously FDA-approved T channel inhibitor, effectively dampened ApoBD-efferocytosis-mediated viral entry and inflammation in vitro in human macrophages and in the lung in vivo in a pre-clinical mouse model of severe COVID-19. Results SARS-CoV-2 infection promotes the formation of ApoBDs that contain infectious virions SARS-CoV-2-induced apoptosis has been observed in multiple tissues and in circulating T cells in COVID-19 patients and verified in vitro in 2D cell culture and organoids, as well as in vivo within mouse models32–37. Therefore, we first aimed to determine whether ApoBDs are formed upon SARS-CoV-2 infection. In concordance with previous studies, we confirmed that SARS-CoV-2-infected epithelial cells, Vero E6 (Fig. 1a) and Calu-3 (Supplementary Fig. 1a), underwent apoptosis as measured by staining for caspase 3/7 activation. Flow cytometry analysis38 using FLICA (to detect caspase 3/7 activity) and annexin 5 (A5) staining (indicative of apoptotic phosphatidylserine (PS) exposure) further detected apoptosis occurring in around 40% of Spike+ cells upon SARS-CoV-2 with MOI as low as 0.03125 (Fig. 1b). Importantly, we observed a large production of ApoBDs from SARS-CoV-2-infected Vero E6 and Calu-3 epithelial cells at 48 h post-infection, comparable to that achieved with the pro-apoptotic BH3 mimetic drug ABT-737 (Fig. 1c and Supplementary Fig. 1b), versus uninfected controls.Fig. 1SARS-CoV-2 infection induces the formation of infectious virion-containing ApoBDs.a Detection of apoptotic caspase 3/7 activity in SARS-CoV-2-infected Vero E6 cells by Caspase-Glo®3/7 assay. ABT-737 (1 μM) was included as an apoptosis control. b Flow cytometry analysis of SARS-CoV-2-infected Vero E6 using FLICA (to detect activated caspase 3/7), annexin 5 (A5) and anti-Spike antibody, showing levels of infected cells (left panel) and apoptosis among the infected (right panel) at 48 h post-infection. c Detection of ApoBD formation in viable control, ABT-737-treated and SARS-CoV-2-infected Vero E6 by flow cytometry according to previous studies38,39. d Viral Spike protein, e viral nucleocapsid protein and f viral N RNA (LOD limit of detection) were detected using flow cytometry and qPCR. g The infectivity of ApoBDs was determined by the TCID50 assay. h Apoptotic cells (FLICA staining), i ApoBD formation (TO-PRO-3 and A5 staining) and j spike+ ApoBDs (with anti-spike protein antibody) were measured in murine bronchoalveolar lavage using flow cytometry. Bronchoalveolar lavage was collected from mice inoculated intranasally with the mouse-adapted SARS-CoV-2 strain at 3 d.p.i. Data are mean ± SEM of n = 3. P values were determined by one-way ANOVA with a Tukey post hoc test or Mann–Whitney U test (for data in (j–l)). Source data are provided as a Source Data file. SARS-CoV-2 infection induces the formation of infectious virion-containing ApoBDs. a Detection of apoptotic caspase 3/7 activity in SARS-CoV-2-infected Vero E6 cells by Caspase-Glo®3/7 assay. ABT-737 (1 μM) was included as an apoptosis control. b Flow cytometry analysis of SARS-CoV-2-infected Vero E6 using FLICA (to detect activated caspase 3/7), annexin 5 (A5) and anti-Spike antibody, showing levels of infected cells (left panel) and apoptosis among the infected (right panel) at 48 h post-infection. c Detection of ApoBD formation in viable control, ABT-737-treated and SARS-CoV-2-infected Vero E6 by flow cytometry according to previous studies38,39. d Viral Spike protein, e viral nucleocapsid protein and f viral N RNA (LOD limit of detection) were detected using flow cytometry and qPCR. g The infectivity of ApoBDs was determined by the TCID50 assay. h Apoptotic cells (FLICA staining), i ApoBD formation (TO-PRO-3 and A5 staining) and j spike+ ApoBDs (with anti-spike protein antibody) were measured in murine bronchoalveolar lavage using flow cytometry. Bronchoalveolar lavage was collected from mice inoculated intranasally with the mouse-adapted SARS-CoV-2 strain at 3 d.p.i. Data are mean ± SEM of n = 3. P values were determined by one-way ANOVA with a Tukey post hoc test or Mann–Whitney U test (for data in (j–l)). Source data are provided as a Source Data file. To examine the presence of viral materials and/or intact virions, we isolated ApoBDs derived from SARS-CoV-2-infected Vero E6 supernatant (denoted ‘SARS-ApoBD’) using our previously established differential centrifugation approach with identity confirmed by flow cytometry39 (Supplementary Fig. 1c, d) to determine its biomolecular contents. Of note, the differential centrifugation approach consistently enabled ApoBD isolation to 90–95% purity (Supplementary Fig. 1d). For comparison, we also isolated non-infectious control ApoBDs generated from ABT-737 treatment (denoted ‘ABT-ApoBD’), and to control for any residual SARS-CoV-2 virions present in the media of the SARS-ApoBD preparation, we also incubated ABT-ApoBD with the leftover supernatant following SARS-ApoBD isolation (denoted ‘SN-ApoBD’). The ApoBD identity of each of these preparations were further verified by the presence of ApoBD-specific protein markers40 such as caspase-cleaved PANX1 (Supplementary Fig. 1e). In contrast to both ApoBD controls, we confirmed SARS-ApoBDs carried SARS-CoV-2 spike protein (Fig. 1d), nucleocapsid protein (Fig. 1e) and viral nucleocapsid RNA (Fig. 1f). Using transmission electron microscopy (TEM), revealed the size of these ApoBDs consistently ranged within the expected range of 1–5 μm (Supplementary Fig. 1f, g) The presence of viral materials and whole virions in ApoBDs generated from SARS-CoV-2-infected cells suggests their ability to traffic infectious virions. Through viral load estimation by 50% tissue culture infectious dose (TCID50) assay (Fig. 1g), comparable cytopathic effect was observed for Vero E6 cells incubated with SARS-ApoBDs and those infected directly with SARS-CoV-2, but to a much lesser extent with SN-ApoBD and ABT-ApoBD controls. Together, these data confirm ApoBDs generated from SARS-infected cells can harbour infectious virions. To examine whether virus-carrying ApoBDs are generated in vivo, we infected aged mice with a mouse-adapted SARS-CoV-2 strain that elicits a severe form of COVID-1941. In bronchoalveolar lavage (BALF) samples collected at day 3 post-infection by flow cytometry (Supplementary Fig. 1h, the levels of apoptotic cells were shown to be significantly elevated compared to mock-infected controls, as indicated by an elevated number of cells with active caspase 3/7 (Fig. 1h). Concomitantly, a significantly elevated ApoBD level was also detected in the infected mouse BALF, as compared to mock control (Fig. 1i). Strikingly, approximately 30% those ApoBDs contained viral spike protein (Fig. 1j). Of note, the majority (~75%) of ApoBDs from the infected mouse BALF were derived from CD45- non-hemapoeitic cells including epithelial cells, whilst the remaining were of neutrophils (1.8%), monocytes (3.2%), macrophages (0.3%) and other leukocytes (20.8%) (Supplementary Fig. 1i). SARS-CoV-2 infection-derived ApoBDs facilitate canonical inflammasome-mediated and NF-κB-dependent inflammation in macrophages via efferocytosis-mediated cell-to-cell viral transmission Despite relatively low ACE2 and TMPRRSS2 expression compared to other cells in the lungs7,8 (Supplementary Fig. 2a), macrophages are key drivers of COVID-19 hypercytokinemia and are identified in vivo to contain virus5,12,16. As monocyte-derived macrophages (MDM) predominate in the lungs of severe COVID-19 patients, whilst alveolar macrophages are depleted42, particularly after initial infection once epithelial cells undergoing apoptosis, we therefore sought to examine the potential role of virus-carrying ApoBDs in facilitating viral entry and hyperinflammatory processes in MDM. Firstly, co-culturing human MDMs with SARS-ApoBD for 48 h, we detected viral spike proteins in macrophages (~15% spike+) by fluorescence microscopy and flow cytometry analysis, but not when co-cultured with either control ApoBDs (ABT-ApoBDs or SN-ApoBDs) or SARS-CoV-2 alone with background level of ~2.5% (Fig. 2a, b). The qPCR analysis also revealed significant detection of nucleocapsid (N) RNA in SARS-ApoBD-co-cultured macrophages (Fig. 2c). Furthermore, SARS-ApoBD-co-cultured macrophages were the only condition that resulted in a significant decrease in macrophage viability, further indicating viral entry (Fig. 2d, e). Of note, a TCID50 assay on Vero E6 cells using the resulting macrophage-derived conditioned media revealed limited cytopathic effect, suggestive of aborted SARS-CoV-2 infection such that limited infectious virions are released by the SARS-ApoBD infected macrophages (Supplementary Fig. 2b). Similarly, we detected viral Spike protein and N RNA in SARS-ApoBD-co-cultured human small airway epithelial cells in submerged culture, which were also refractory to direct SARS-CoV-2 infection due to low ACE2 and TMPRRS2 expression43 (Supplementary Fig. 3a, b).Fig. 2SARS-ApoBDs facilitates viral entry and induces inflammation in macrophages.a Immunofluorescence imaging showing nucleus staining (light blue), F-actin (red) and viral Spike protein (green); scale bar, 20 μm. b Viral spike-staining flow cytometry and c N RNA-targeting qPCR analyses for ApoBD-co-incubated human MDMs (MDM:ApoBD ratio of 1:3, equivalent to SARS-CoV-2 MOI of 3); LOD: limit of detection. Cytopathic effect on MDMs at 48 h post-co-incubation was determined by d bright-field imaging (scale bar, 100 μm) and e cell viability assay. f Inflammatory cytokines released by ApoBD-incubated MDM, detected using bead-based cytokine assay. g Gene ontology enrichment analysis of biological processes commonly enriched in SARS-ApoBD relative to untreated, SARS-CoV-2-infected, and ABT-ApoBD-treated samples. Dot size indicates fold enrichment, while colour gradient represents statistical significance. h Heatmap depicting the profile of differentially abundant proteins in SARS-ApoBD compared to untreated, SARS-CoV-2-infected, and ABT-ApoBD-treated samples. Data are mean ± SEM of n = 3. Data in (a, d) are representative of three independent experiments. All P values (except for (g)) were determined by one-way ANOVA with a Tukey post hoc test. P values in (g) are from each comparison and were calculated using a hypergeometric test as obtained from the Gene Ontology resource. Source data are provided as a Source Data file. SARS-ApoBDs facilitates viral entry and induces inflammation in macrophages. a Immunofluorescence imaging showing nucleus staining (light blue), F-actin (red) and viral Spike protein (green); scale bar, 20 μm. b Viral spike-staining flow cytometry and c N RNA-targeting qPCR analyses for ApoBD-co-incubated human MDMs (MDM:ApoBD ratio of 1:3, equivalent to SARS-CoV-2 MOI of 3); LOD: limit of detection. Cytopathic effect on MDMs at 48 h post-co-incubation was determined by d bright-field imaging (scale bar, 100 μm) and e cell viability assay. f Inflammatory cytokines released by ApoBD-incubated MDM, detected using bead-based cytokine assay. g Gene ontology enrichment analysis of biological processes commonly enriched in SARS-ApoBD relative to untreated, SARS-CoV-2-infected, and ABT-ApoBD-treated samples. Dot size indicates fold enrichment, while colour gradient represents statistical significance. h Heatmap depicting the profile of differentially abundant proteins in SARS-ApoBD compared to untreated, SARS-CoV-2-infected, and ABT-ApoBD-treated samples. Data are mean ± SEM of n = 3. Data in (a, d) are representative of three independent experiments. All P values (except for (g)) were determined by one-way ANOVA with a Tukey post hoc test. P values in (g) are from each comparison and were calculated using a hypergeometric test as obtained from the Gene Ontology resource. Source data are provided as a Source Data file. Critically, SARS-ApoBD-treated MDMs did, however, secrete significant levels of cytokines associated with severe COVID-19; TNF, IL-6, IL-1β, IL-10 and IL-8 (Fig. 2f). These effects were not observed with direct SARS-CoV-2 or control ApoBDs co-culture (Fig. 2f), further suggesting a specific attribution to viron-carrying ApoBDs, rather than free/co-pelleted viruses or ApoBDs alone. Furthermore, proteomic analyses of MDM lysates revealed enrichment of proteins associated with viral entry/transport and endocytosis (e.g., ICAM1, ITGB3, MYH10), interferon antiviral defence (e.g., IFIT1/2, MX1/2) and inflammation/immune regulation (e.g., IL-18, IL-10RB, SIGLEC1, NFKB2, TNFAIP6, SERPINB2, RELB, IFI44) 48 h post-co-incubation with SARS-ApoBDs, compared to those treated with SARS-CoV-2 alone, non-infectious ApoBDs or untreated control (Fig. 2g, h). To determine the signalling pathways that mediate the observed macrophage inflammation, we performed immunoblot analysis and detected the oligomerisation of ASC protein, caspase-1 cleavage and IL-1β induction/processing, suggesting canonical inflammasome activation as early as at 6 h post-treatment with SARS-ApoBDs (Fig. 3a). We also detected p65-S536 phosphorylation and IκBα upregulation, indicative of NF-κB signalling induction from 18 h and more pronouncedly at 30 h post-treatment (Fig. 3a). It appears that MAPK did not play a role as no p38-T180/Y182 phosphorylation was seen at the indicated timepoints.Fig. 3SARS-ApoBDs-induced macrophage inflammation occurs via canonical inflammasome and NF-κB signalling and is mediated by ApoBD efferocytosis.a Immunoblotting analyses, detecting canonical inflammasome, NF-κB and MAPK signalling in untreated, ABT-ApoBD and SARS-ApoBD-treated macrophages at 6 h, 18 h, 30 h post-co-incubation. Untreated and Pam2CSK4/nigericin-treated THP-1 lysates were included as technical/positive controls. b Efferocytosis of SARS-CoV-2 infection-derived ApoBDs by MDMs and inhibition thereof by A5 and cytochalasin D, indicated by CypHer5E staining. c Reduced spike protein detection and d impaired cytokine release in SARS-ApoBD-treated MDM by efferocytosis inhibitors. e Schematic for the FACS of ApoBD engulfing MDMs (CypHer5E+) and non-engulfing MDMs (CypHer5E-), followed by detection of viral uptake, as determined by f spike protein signal and g N RNA ((LOD: limit of detection), as well as h cytokine release. Data are mean ± SEM of n = 3–4. P values were determined by one-way ANOVA with a Tukey post hoc test. Source data are provided as a Source Data file. SARS-ApoBDs-induced macrophage inflammation occurs via canonical inflammasome and NF-κB signalling and is mediated by ApoBD efferocytosis. a Immunoblotting analyses, detecting canonical inflammasome, NF-κB and MAPK signalling in untreated, ABT-ApoBD and SARS-ApoBD-treated macrophages at 6 h, 18 h, 30 h post-co-incubation. Untreated and Pam2CSK4/nigericin-treated THP-1 lysates were included as technical/positive controls. b Efferocytosis of SARS-CoV-2 infection-derived ApoBDs by MDMs and inhibition thereof by A5 and cytochalasin D, indicated by CypHer5E staining. c Reduced spike protein detection and d impaired cytokine release in SARS-ApoBD-treated MDM by efferocytosis inhibitors. e Schematic for the FACS of ApoBD engulfing MDMs (CypHer5E+) and non-engulfing MDMs (CypHer5E-), followed by detection of viral uptake, as determined by f spike protein signal and g N RNA ((LOD: limit of detection), as well as h cytokine release. Data are mean ± SEM of n = 3–4. P values were determined by one-way ANOVA with a Tukey post hoc test. Source data are provided as a Source Data file. We next sought to investigate whether ApoBD-mediated SARS-CoV-2 entry into macrophages was mediated through efferocytosis of ApoBDs. An engulfment assay using the pH-dependent CypHer5E stain as an engulfment indicator showed that MDMs readily efferocytosed SARS-ApoBDs (Fig. 3b). By contrast, sequestering the ‘eat-me’ signal phosphatidylserine on ApoBDs by A5 treatment or inhibiting actin polymerisation required for efferocytosis with cytochalasin D (CTD), significantly impaired SARS-ApoBD engulfment (Fig. 3b). Blockade of SARS-ApoBD uptake by A5 or CTD treatment also reduced the proportion of spike+ MDMs (Fig. 3c) and markedly reduced levels of macrophage cytokine release (Fig. 3d). Confirming the ACE-2 independence of SARS-ApoBD efferocytosis, antibodies against ACE2 on MDM or viral Spike did not cause any significant inhibitory effect on SARS-ApoBD-induced viral entry and inflammation in macrophages (Supplementary Fig. 4). Next, we isolated SARS-ApoBD engulfing and non-engulfing MDMs by fluorescence-activated cell sorting (FACS) based on high and low CypHer5E signal, respectively (Fig. 3e), with SARS-ApoBD-engulfing MDMs (CypHer5E+) showing viral uptake, with an increase in the percentage of spike+ cells (Fig. 3f) and high levels of viral N RNA (Fig. 3g), as well as substantial cytokine production (Fig. 3h). In contrast, their non-engulfing (CypHer5E-) counterparts and MDMs that engulfed ABT-ApoBD or SN-ApoBD did not show these effects (Fig. 2f–h). These data collectively demonstrate that efferocytosis of ApoBDs is necessary for ApoBD-mediated SARS-CoV-2 entry and subsequent triggering of macrophage hypercytokinemia. Mibefradil and other voltage-gated T-type calcium channel blockers impair SARS-CoV-2-induced ApoBD formation Disassembly of apoptotic cells into ApoBDs promotes clearance of apoptotic materials, and genetic disruption of ApoBD biogenesis leads to a noticeable reduction in efferocytosis efficiency19. Thus, we reasoned that identification of small molecules that could reduce infection-induced ApoBD formation could subsequently limit ApoBD-efferocytosis-mediated SARS-CoV-2 entry into macrophages and associated inflammation. To this end, we tested inhibitors of ApoBD formation based on our previous drug screening with apoptotic Jurkat T cells44, a well-established model for apoptotic cell disassembly19,38,45, prior to testing on an infection model19 (Fig. 4a). In this model, trovafloxacin (TVF) was added to inhibit the key negative regulator of apoptotic cell disassembly, PANX1, maximising Jurkat T cells generation of ApoBD’s19,38,45, and allowing identification of potent ApoBD formation inhibitors. From this screen, several calcium channel blockers in the drug library showed inhibitory effects on apoptotic cell disassembly (Supplementary Fig. 5a). In a secondary, validating screen, we found the calcium channel blocker mibefradil (MBF) substantially impaired ApoBD formation in a dose-dependent manner (Fig. 4b). As shown by time-lapse differential interference contrast (DIC) microscopy, MBF treatment attenuated the protrusion and fragmentation steps of ApoBD biogenesis, but not membrane blebbing, during the progression of apoptosis (Fig. 4c and Supplementary Movies 1 and 2).Fig. 4Voltage-gated T-type calcium channel blockers, novel inhibitors of apoptotic cell disassembly, impair SARS-CoV-2-induced ApoBD formation.a Schematic for the LOPAC® 1280 drug library screening to identify novel inhibitors of apoptotic cell disassembly. Trovafloxacin, a known apoptotic cell disassembly enhancer, was added to maximise ApoBD formation, hence allowing the screening for potent, robust blockers. b Detection of ApoBD formation following MBF treatment by flow cytometry using TO-PRO-3 and A5 staining. c Time-lapse DIC microscopy of anti-Fas (1:1000)-treated Jurkat T cells with or without 10 μM mibefradil (MBF). The level of apoptotic cell disassembly was quantified as the percentage of cells undergoing fragmentation. Scale bar, 10 μm. d Time-lapse Fluo-4 signal intensity (left panel) and peak counts (right panel), indicative of calcium influxes, of 1 mM EGTA-, 10 μM mibefradil- or vehicle-treated THP-1 monocytes post-UV irradiation, as measured by confocal imaging. e Detection of ApoBD formation following EGTA treatment by flow cytometry. f Time-lapse DIC microscopy to apoptotic cell disassembly in the presence of 1 mM EGTA or vehicle control. Scale bar, 10 μm. g Calcium influxes, indicated by Fluo-4 peak counts, of CACNA1H-deficient cells and Cas9 control. h Flow cytometry analysis and i time-lapse DIC microscopy to quantify and monitor ApoBD formation in CACNA1H-knockdown cells or Cas9 control following apoptotic induction using anti-Fas. Scale bar, 10 μm. j Chemical structures of T-channel blockers used in these studies. k Flow cytometry analysis of anti-Fas-treated Jurkat T cells treated with other T-channel inhibitors, including ML218, NNC 55-0396 and penfluridol. Detection of SARS-CoV-2 infection-induced ApoBD formation following treatment with l mibefradil and m other T-channel inhibitors. Data are mean ± SEM of n = 3. Data in (c, f–h, j, i) are representative of three independent experiments. P values were determined by one-way ANOVA with a Tukey post hoc test or Mann–Whitney U test (for data in (c, f–h, j)). Source data are provided as a Source Data file. Voltage-gated T-type calcium channel blockers, novel inhibitors of apoptotic cell disassembly, impair SARS-CoV-2-induced ApoBD formation. a Schematic for the LOPAC® 1280 drug library screening to identify novel inhibitors of apoptotic cell disassembly. Trovafloxacin, a known apoptotic cell disassembly enhancer, was added to maximise ApoBD formation, hence allowing the screening for potent, robust blockers. b Detection of ApoBD formation following MBF treatment by flow cytometry using TO-PRO-3 and A5 staining. c Time-lapse DIC microscopy of anti-Fas (1:1000)-treated Jurkat T cells with or without 10 μM mibefradil (MBF). The level of apoptotic cell disassembly was quantified as the percentage of cells undergoing fragmentation. Scale bar, 10 μm. d Time-lapse Fluo-4 signal intensity (left panel) and peak counts (right panel), indicative of calcium influxes, of 1 mM EGTA-, 10 μM mibefradil- or vehicle-treated THP-1 monocytes post-UV irradiation, as measured by confocal imaging. e Detection of ApoBD formation following EGTA treatment by flow cytometry. f Time-lapse DIC microscopy to apoptotic cell disassembly in the presence of 1 mM EGTA or vehicle control. Scale bar, 10 μm. g Calcium influxes, indicated by Fluo-4 peak counts, of CACNA1H-deficient cells and Cas9 control. h Flow cytometry analysis and i time-lapse DIC microscopy to quantify and monitor ApoBD formation in CACNA1H-knockdown cells or Cas9 control following apoptotic induction using anti-Fas. Scale bar, 10 μm. j Chemical structures of T-channel blockers used in these studies. k Flow cytometry analysis of anti-Fas-treated Jurkat T cells treated with other T-channel inhibitors, including ML218, NNC 55-0396 and penfluridol. Detection of SARS-CoV-2 infection-induced ApoBD formation following treatment with l mibefradil and m other T-channel inhibitors. Data are mean ± SEM of n = 3. Data in (c, f–h, j, i) are representative of three independent experiments. P values were determined by one-way ANOVA with a Tukey post hoc test or Mann–Whitney U test (for data in (c, f–h, j)). Source data are provided as a Source Data file. MBF is a T-type voltage-gated calcium channel (T-channel) blocker that inhibits T-channel-mediated calcium influx from the extracellular space into the cytosol46, previously FDA-approved for the treatment of hypertension and angina. To better understand the mechanism underpinning the inhibition of ApoBD formation by MBF, we performed calcium imaging by confocal microscopy using the calcium probe Fluo-4. MBF was found to impair both the intensity and the number of calcium influxes detected post-apoptosis induction and prior to apoptotic cell fragmentation (Fig. 4d and Supplementary Movies 3–5). A similar inhibitory effect was observed upon extracellular calcium chelation using EGTA (Fig. 4d). Remarkably, EGTA treatment also inhibited ApoBD formation (Fig. 4e), and the formation of membrane protrusion and cell fragmentation during progression of apoptosis (Fig. 4f). It is worth noting that the observed reduction in ApoBD formation following MBF and EGTA treatment was not due to impaired apoptosis induction as neither MBF nor EGTA altered the levels of cell death induced by anti-Fas (Supplementary Fig. 5b). These observations suggest MBF acts on calcium influxes that drive ApoBD biogenesis. Among the three T-channel isoforms (CaV3.1, CaV3.2 and CaV3.3 encoded by CACNA1G, CACNA1H and CACNA1I, respectively), our model Jurkat T cells express predominantly CaV3.2 (Supplementary Fig. 5c). Of note, T-channel isoforms such as CaV3.2 are expressed in many tissues and cell lines, including epithelial Calu-3 and Vero E6 cells (Supplementary Fig. 5d), suggesting their druggability to block SARS-CoV-2 infection-induced ApoBD formation. Next, to deduce the relevance of T-channels in apoptotic cell disassembly, we attempted to generate a CaV3.2-deficient cell line using CRISPR/Cas9. Although we were unable to generate a stable knockout cell line, the resultant CaV3.2 knockdown cells (Supplementary Fig. 5e) exhibited impaired calcium influxes post-apoptosis induction (Fig. 4g and Supplementary Movies 6 and 7) as well as decreased ApoBD formation due to impaired protrusion formation and fragmentation (Fig. 4h, i and Supplementary Movies 8 and 9). In addition, mibefradil-derivative NNC 55-0396 (NNC) (identified in the initial screen, Supplementary Fig. 5a) and other structurally unrelated T-channel specific inhibitors including ML218, and penfluridol (PFR) exerted similar dose-dependent inhibitory effects on ApoBD formation (Fig. 4j, k). By contrast, blockers of other types of calcium channels, including L-type calcium channel (nifedipine and nicardipine), N-type calcium channel (PD173212), calcium-release-activated calcium channel protein 1 Orai1 (AnCoA4), calcium storage release ryanodine receptor (dantrolene) or inositol trisphosphate receptor (2-aminoethoxydiphenylborane) did not inhibit ApoBD formation (Supplementary Fig. 5f). Therefore, these data indicate that T-channels, the targets of MBF, are positive regulators of apoptotic cell disassembly. Comparable to the Jurkat T cell model, MBF treatment markedly inhibited the formation of ApoBDs by SARS-CoV-2-infected Vero E6 cells (Fig. 4l) without affecting the levels of Vero cell apoptosis or viral production (Supplementary Fig. 5g, h). Similarly, the other T-channel inhibitors NNC, ML218 and PFR also reduced ApoBD release from SARS-CoV-2-infected apoptotic cells (Fig. 4m). Mibefradil reduces cell-to-cell SARS-CoV-2 transmission and inflammation in macrophages in vitro and in vivo Inhibition of SARS-ApoBD formation by T-channel inhibitors such as MBF entails their potential to limit ApoBD-efferocytosis-mediated SARS-CoV-2 entry into macrophages and dampen subsequent inflammation. To this end, having also confirmed MBF did not affect Vero cell viral levels (Supplementary Fig. 5h), we treated SARS-CoV-2-infected Vero E6 cells with or without MBF for 48 h and then transferred the conditioned media (CM) onto MDMs (Fig. 5a). CM from MBF-treated cells resulted in significantly reduced levels of ApoBD efferocytosis as measured by CypHer5E+ as well as viral transmission in MDMs as indicated by viral N RNA (Fig. 5b, c) and in human small airway epithelial cells as shown by flow cytometry (Supplementary Fig. 3c). Significantly, MDMs cultured in MBF-treated CM also showed enhanced cell viability (Fig. 5d, e) and reduced cytokine secretion (Fig. 5f). Of note, MBF, at the same concentration, did not significantly inhibit LPS-induced inflammation in macrophages (Supplementary Fig. 6), suggesting that MBF is likely to act on ApoBD formation, hence efferocytosis, rather than directly suppressing inflammation signalling. In addition, Z944, a selective T-type calcium channel blocker which is currently undergoing clinical trials47, similarly inhibited Vero cells SARS-CoV-2 infection-induced ApoBD formation, reduced CM SARS-CoV-2 entry and macrophage inflammation, though at a higher concentration (Supplementary Fig. 7).Fig. 5Mibefradil reduces SARS-CoV-2 entry and inflammation in vitro and in vivo.a Schematic for the transfer of conditioned media (CM) to elucidate the effect of MBF on viral uptake, cytopathic effect and inflammatory responses by MDMs. b Flow cytometry analysis of MDMs efferocytosing apoptotic materials in conditioned media, as measured by CypHer5E positivity. c qPCR analysis, d cell viability assay as well as e bright-field imaging (scale bar, 100 μm) and f cytokine bead array to determine viral N RNA (LOD limit of detection), cytopathic effect and inflammation, respectively, in MDMs upon condition media treatment. g Schematic for mibefradil administration in SARS-CoV-2-infected mice and post-treatment analysis to detect h ApoBD formation (by flow cytometry of the bronchoalveolar lavage fluid (BALF)), i spike+ macrophages (BALF), j viral load (by TCID50 assay on lung samples), k inflammatory cytokines (by cytokine bead array of BALF), l immune cell infiltration (BALF), and m lung histology (H&E staining and histopathology scoring, see Supplementary Fig. 9). Scale bar, 100 μm. Data in (b–d, f) are mean ± SEM of n = 3. Data in (e) are representative of three independent experiments. Data in (h−l) are mean ± SEM of n = 5–10. Histology data in (m) are representative of at least five mice per group, scoring is mean ± SEM of n = 5. P values were determined by one-way ANOVA with a Tukey post hoc test. Source data are provided as a Source Data file. Mibefradil reduces SARS-CoV-2 entry and inflammation in vitro and in vivo. a Schematic for the transfer of conditioned media (CM) to elucidate the effect of MBF on viral uptake, cytopathic effect and inflammatory responses by MDMs. b Flow cytometry analysis of MDMs efferocytosing apoptotic materials in conditioned media, as measured by CypHer5E positivity. c qPCR analysis, d cell viability assay as well as e bright-field imaging (scale bar, 100 μm) and f cytokine bead array to determine viral N RNA (LOD limit of detection), cytopathic effect and inflammation, respectively, in MDMs upon condition media treatment. g Schematic for mibefradil administration in SARS-CoV-2-infected mice and post-treatment analysis to detect h ApoBD formation (by flow cytometry of the bronchoalveolar lavage fluid (BALF)), i spike+ macrophages (BALF), j viral load (by TCID50 assay on lung samples), k inflammatory cytokines (by cytokine bead array of BALF), l immune cell infiltration (BALF), and m lung histology (H&E staining and histopathology scoring, see Supplementary Fig. 9). Scale bar, 100 μm. Data in (b–d, f) are mean ± SEM of n = 3. Data in (e) are representative of three independent experiments. Data in (h−l) are mean ± SEM of n = 5–10. Histology data in (m) are representative of at least five mice per group, scoring is mean ± SEM of n = 5. P values were determined by one-way ANOVA with a Tukey post hoc test. Source data are provided as a Source Data file. Finally, we tested the effect of MBF on aged mice infected with SARS-CoV-2 which were previously shown to develop a severe form of COVID-1941. MBF was intranasally administered on day 1 and day 2 post aerosol-infection, with viral peak occurring on day 2 in this model and cytokine peak on day 3 (Fig. 5g). Analogous to the in vitro findings (Fig. 4l, m), the number of ApoBDs detected in BALF samples was significantly lower in MBF-treated mice on day 3 post-infection compared to vehicle control (Fig. 5h, Supplementary Fig. 8). This effect also aligned with a reduction in spike+ macrophages in the BALF, suggesting less viral entry into lung macrophages in vivo upon MBF administration (Fig. 5i), as well as significantly reduced lung viral load as shown by the lung homogenate TCID50 assay (Fig. 5j) and reduced TNF and IL-6 secretion in BALF (Fig. 5k). In terms of innate immune cell infiltration into the lung, MBF treatment dramatically reduced the levels of infiltrating macrophages, neutrophils and, monocytes (Fig. 5l), with marked amelioration of the pulmonary pathology indicative of severe COVID-19 upon histological examination (Fig. 5m and Supplementary Fig. 9). Discussion The intricate mechanisms by which SARS-CoV-2 enters and interacts with host cells have been the subject of extensive research since the start of the COVID-19 pandemic, contributing to the growing understanding of viral pathogenesis. Nevertheless, the various ways SARS-CoV-2 enters macrophages and drives detrimental inflammation in COVID-19 patients are yet to be fully defined. Macrophages are heterogeneous and the expression of key host factors for SARS-CoV-2 entry, ACE2 and TMPRSS2, are low to undetectable in human MDMs and BALF macrophages investigated herein12–15,48. Moreover, immunohistochemistry studies showed a strong cytosolic staining pattern for SARS-CoV-2 specific to ACE2+ bronchial and alveolar epithelial cells, whereas alveolar macrophages had only punctate staining for viral protein or RNA, suggesting viral uptake but not productive infection (DOI: 10.1183/13993003.02725-2021). Of note, overexpression of ACE2 in macrophages permitted SARS-CoV-2 entry, replication, and virion release49,50. Furthermore, the major driver of alveolar macrophage inflammation and alveolar injury was reportedly not viral replication, but instead, the detection of viral materials upon viral uptake48,51,52. Consistent with this notion, recent in vitro studies showed that viral internalisation and/or viral replication in human MDMs upon SARS-CoV-2 exposure is limited even at the high MOIs and was unable to trigger inflammation12,49. It has been shown that in the presence of antiviral spike antibodies, ACE2-negative macrophages can take up antibody-opsonised SARS-CoV-2 through FcγR receptors and cause systemic inflammatory responses despite non-productive abortive infection53. Further elucidation of ACE2-independent viral uptake in macrophages and COVID-19-associated macrophage inflammation are therefore warranted. Here, we describe that in the absence of opsonising antibodies and at a relatively low MOI, SARS-CoV-2 can also enter ACE2-negative/low human primary macrophages through the efferocytosis of ApoBDs released by apoptotic SARS-CoV-2-infected epithelial cells. Unlike SARS-CoV-2 exposure alone, human MDMs became infected and were able to elicit secretion of TNF, IL-6, IL-1β, IL-8 and IL-10, indicative of the cytokine storm in severe COVID-19 patients54,55. This infection is also abortive, and MDM underwent only a modest cytopathic effect. Viral entry/endocytosis, viral sensing and inflammation pathways were also significantly enriched in MDMs upon efferocytosis of SARS-ApoBDs. Intriguingly, many proteins associated with these pathways, including MX1/256, IL-1857,58, IL-1059, SIGLEC160 and TNFAIP661, have been detected in severe COVID-19 patients. Furthermore, the efferocytosis-mediated macrophage inflammation is evidently triggered via the activation of host viral defence signalling, particularly the canonical inflammasome and, at a later stage, NF-κB pathways. Therefore, our findings suggest a potential alternative mechanism for viral entry and exacerbated macrophage-driven inflammation during COVID-19. While our work builds upon the broader concept that viruses can exploit apoptosis for viral dissemination, mechanism(s) of which hitherto remained largely unresolved, to our knowledge, it is the first to demonstrate the novel mechanism of ApoBD-mediated cell-to-cell transmission and for SARS-CoV-2 immunopathogenesis. As ApoBDs are readily circulating and efficiently engulfed by non-professional phagocytes such as endothelial and epithelial cells, our findings herein have the potential to be extended into other pathophysiological hallmarks of SARS-CoV-2 infection including, but not limited to, multi-organ injuries, cardiovascular damage, and neurological complications. Nevertheless, we emphasise that the proposed ApoBD efferocytosis-mediated viral entry does not preclude ACE2-dependent infection. Instead, it provides a complementary pathway that may be particularly important in ACE2-low cell subsets, after waves of epithelial apoptosis, and when infiltrating monocyte-derived macrophages expand in the lung. Moreover, although we deduced that efferocytic machineries underpin this viral entry pathway, exact efferocytic receptors and details of the downstream viral escape pathway warrant further research. Small EVs can exert a variety of effects on recipient cells, often mediated through their accompanying cargoes26. Notably, circulating EVs of 50−500 nm (classified as small EVs62) in COVID-19 patients can be endocytosed and cause apoptosis in recipient pulmonary endothelial cells63. Whether such EVs can facilitate viral entry and impact inflammation has not been determined. In line with the literature, our study confirmed that SARS-CoV-2-induced apoptosis in various host cells, as well as in the murine infection model, generates SARS-CoV-2-carrying ApoBDs (classified as large EVs) that are released exclusively during apoptosis. Importantly, we demonstrated that the engulfment of SARS-CoV-2-carrying ApoBDs can subsequently aid viral entry into macrophages and promote inflammation despite abortive infection. Our observation raises several opportunities to target this pathway of viral entry into macrophages, including: (i) targeting the induction of apoptosis, (ii) the formation of ApoBDs, and (iii) the efferocytosis of ApoBDs. In fact, blockade of apoptosis through caspase-6 inhibition arrests viral replication and substantially attenuates lung pathology in golden Syrian hamsters infected with SARS-CoV-264. Alternatively, targeting the efferocytosis machinery may also be another approach to block viral entry into macrophages and prevent hyperinflammation. However, it is worth noting that cell clearance itself is also an important mechanism to limit inflammation65. As the molecular basis of ApoBD formation is yet to be fully elucidated it represents novel opportunities for therapeutic targeting. We identified T-channels as regulators of apoptotic cell disassembly and demonstrated that the blockade of T-channels with MBF can limit SARS-CoV-2-induced ApoBD formation through the blockade of calcium influxes. These inhibitors were shown to mitigate ApoBD-mediated viral entry and not only dampen hypercytokinemia in primary human macrophages and mouse BALF macrophages but also ameliorate overall lung inflammation. As only a few enzyme-inhibiting antivirals have been approved with efficacy prior to viral peak and are prescribed sparingly for COVID-19, new/repurposed drugs with different mode of mechanisms like T-channel blockers (highly tolerable, topically applicable, efficacious when given during viral peak) have the potential to not only reduce disease burden in severe patients, but also lower the risk of drug resistance and have a longer efficacy window. Notably, inhibition of ApoBD formation by MBF may also limit secondary viral transmission to macrophages and, speculatively, other phagocytic cells, thereby restricting infection amplification. Furthermore, attenuation of macrophage-driven inflammation by mibefradil may reduce inflammation-mediated epithelial damage and secondary viral spread, collectively contributing to the observed reduction in viral burden (Fig. 5j). A limitation of this study is that all in vivo experiments were conducted in aged female mice, which were selected due to their availability, and therefore potential sex-specific effects of T-channel blockade cannot be excluded and warrant validation in aged male mice. It is also important to note that MBF has been withdrawn from the market due to safety concerns66. Further drug design and drug optimisation to reduce side effects, as well as long-term investigation of the drug treatment, are warranted before any clinical use of T-channel blockers for COVID-19. In addition, future testing of other ApoBD formation inhibitors could also be beneficial. Together, our findings not only provide further insights into SARS-CoV-2 pathogenesis but also rationalise the development of ApoBD formation-targeting drugs to tackle severe COVID-19 as well as other apoptosis-inducing newly emerging viral threats. Methods Ethics approval Human monocyte-derived macrophages (MDMs) were prepared from fresh buffy coat obtained from the Australian Red Cross Lifeblood Service with donor consent as previously described67,68. Appropriate ethics approval was granted by the La Trobe University and the Walter and Eliza Hall Institute of Medical Research (WEHI) Human Ethics Committees (WEHI HREC #18_09LR). All procedures involving animals and SARS-CoV-2 strains were conducted in an OGTR-approved Physical Containment Level 3 (PC3) facility at WEHI (Cert-3621). All mouse strains and procedures were reviewed and approved by the WEHI Animal Ethics Committee and were conducted in accordance with the Prevention of Cruelty to Animals Act (1986) and the Australian National Health and Medical Research Council Code of Practice for the Care and Use of Animals for Scientific Purposes (1997). Mammalian cell culture Cell lines Monkey kidney epithelial cells (Vero E6, clone CCL81) and human lung epithelial cells (Calu-3) were respectively cultured in DMEM medium and EMEM medium (ThermoFisher) supplemented with 10% heat-inactivated foetal bovine serum (FBS). Human Jurkat T cells and human THP-1 monocytes were cultured in RPMI 1640 medium (ThermoFisher), supplemented with 10% FBS, penicillin (50 U/mL), streptomycin (50 μg/mL) and MycoZap (Lonza, Switzerland). All cells were cultured at 37 °C in a humidified atmosphere with 5% CO2. Human monocyte-derived macrophage preparation Peripheral blood mononuclear cells were isolated from buffy coats by density gradient centrifugation with Lymphoprep (StemCell Technologies). CD14+ monocytes were then obtained by positive magnetic isolation using human CD14 microbeads (Miltenyi Biotec) and a MACS LS column with a QuadroMacsTM separator (Miltenyi Biotech). Monocytes were differentiated into GM-CSF-polarised (M1) MDMs over seven days in RPMI 1640 media supplemented with 1 mM L-glutamine (Gibco), 2 mM sodium pyruvate, 10% human adult AB serum (Sigma-Aldrich) and 5 ng/mL GM-CSF (Miltenyi Biotec) at 37 °C in a humidified atmosphere with 5% CO2. Human neutrophil preparation Human neutrophils were prepared from fresh Sodium Heparin blood obtained from consenting donors through the WEHI volunteer blood donor registry as previously described69. Ethics approval was granted by WEHI HREC #18_09LR. Whole blood was diluted 1:1 with 1× PBS (ThermoFisher) and layered over Lymphoprep followed by density gradient centrifugation. Neutrophils were isolated from the cell pellet following red blood cell (RBC) lysis (8% Ammonium chloride [NH4Cl], 0.8% Sodium bicarbonate [NaHCO3] and 0.4% Ethylenediaminetetraacetic acid [EDTA]). Following two washes in PBS, neutrophils were resuspended in OptiMEM (ThermoFisher), and 1 × 107 cells were seeded evenly into three wells of a six-well cell culture plate (Nest Scientific). Cells were incubated for 6 h at 37 °C in a humidified atmosphere with 5% CO2. SN was removed, and 1 mL TRIzol was added to each well prior to being stored at −80 °C before RNA extraction. Mice Aged (20-50 weeks old), female wild-type C57BL/6J mice were bred and maintained in the Specific Pathogen Free Physical Containment Level 2 Bioresources Facility at WEHI. Mice were transferred to the PC3 laboratory for all SARS-CoV-2 infection experiments at least 4 days prior to experiments. SARS-CoV-2 strains, viral infection and mibefradil treatment Infection of mammalian cells Mammalian cells were infected with a clinical isolate of the originating Wuhan strain of SARS-CoV-2 (hCoV-19/Australia/VIC001/2020, obtained from the Victorian Infectious Disease Reference Laboratory). To prepare viral stocks, viruses were grown in Vero E6 cells in serum-free DMEM media supplemented with 1 μg/mL TPCK-treated trypsin (ThermoFisher). Conditioned media were collected after 48 h and centrifuged at 3000×g for 10 min before passing through a 0.22-μm Millex® syringe filter (Merck Millipore). For viral infection, virus stocks were diluted in serum-free media (with 1 μg/mL TPCK-treated trypsin) to the indicated multiplicity of infection (MOI) and added onto mammalian cells. After 1 h of incubation, the infection media were removed and replaced with the appropriate culture media supplemented with 2% FBS and further incubated for 48 h. Infection of mouse models For SARS-CoV-2 murine infection, the mouse-adapted SARS-CoV-2 VIC2089 P21 isolate (generated through serial passaging of the clinical isolate hCoV-19/Australia/VIC2089/2020)41 was used. Briefly, virus stocks were diluted in serum-free DMEM to a final concentration of 1 × 104 TCID50/mouse and inoculated intranasally (30 μL) into methoxyflurane-anesthetised mice. After infection, animals were visually checked and weighed daily before being euthanised at 3 days post-infection (d.p.i.). For intranasal administration of mibefradil, female mice (20–50 weeks old, age-matched) were infected with SARS-CoV-2 (1 × 104 TCID50/mouse) or mock. MBF (reconstituted in 1× PBS, 20 mg/kg) or PBS vehicle was administered intranasally on 1 and 2 d.p.i. After infection, animals were visually checked and weighed daily before being euthanised at 3 d.p.i. ApoBD detection ApoBD detection in conditioned media ABT-737-treated or SARS-CoV-2-infected cells (detached using Trypsin/EGTA) and their respective conditioned media were combined and stained with fluorescently labelled annexin A5 (BD Biosciences, 1:2000) and 0.2 μM TO-PRO-3 (Life Technologies) in 1× A5 binding buffer (BD Biosciences) at room temperature for 10 min. Samples were immediately placed on ice, and SPHEROTM AccuCount beads (Spherotech) were added prior to analysis using a BD FACSAriaTM Fusion flow cytometer and FACSDiva 6.1.1 software (BD Bioscience). Data were analysed using FlowJo 10.7.1 software (Tree Star), as previously described70. ApoBD detection in BALF samples Mock or SARS-CoV-2-infected mice were humanely euthanised at 3 d.p.i. To collect bronchoalveolar lavage fluid (BALF), the trachea was cannulated and 4× 0.5 mL aliquots of 1× PBS were delivered to and captured from the lungs. BALF samples were then centrifuged at 3000×g for 20 min to collect cells and ApoBDs for staining with a combination of SR-DEVD-FMK (i.e., FLICA, a caspase 3/7 activity detection dye, ThermoFisher), annexin A5-V450 and TO-PRO-3 in 1× A5 binding buffer at room temperature for 10 min. Following fixation (with 2% paraformaldehyde) and permeabilisation (using 0.5% saponin), cell/ApoBD samples were incubated with FITC-conjugated antiviral spike antibodies (1:250, RnD Systems). SPHEROTM AccuCount beads (Spherotech) were also added prior to flow cytometry analysis using a BD FACSAriaTM Fusion flow cytometer. Cellular origins of ApoBDs from the infected mouse BALF were determined using a surface marker cocktail containing Ly6G-PE-Cy7 (1:300), F4/80-PerCP-Cy5.5 (1:150), CD45.2-AF700 (1:200), CD11C-APC-Cy7 (1:200), Ly6C-BV421 (1:300) and CD11b-V500 (1:300) as described below in the 'BALF flow cytometry analysis'. ApoBD detection in drug treatments To induce apoptosis and ApoBD formation in Jurkat T cells, cells at 1.5 × 106 cells/mL in serum-free RMPI 1640 supplemented with 1% BSA were treated with 1:1000 anti-Fas (human, activating, clone CH11) and 40 μM trova. Different concentrations of MBF, EGTA, ML218, NNC 55-0396 or penfluridol were also added. Anti-Fas, Trova, MBF, EGTA and penfluridol were purchased from Sigma-Aldrich. ML218 and NNC 55-0396 were purchased from Tocris Bioscience. Cell samples were incubated at 37 °C in a humidified atmosphere with 5% CO2 for 3 h prior to A5 and TO-PRO-3 staining for flow cytometry analysis. For in vitro infection experiments, 10 μM MBF, 10 μM ML218, 4 μM NNC 55-0396 or 4 μM penfluridol was added to Vero E6 cells following SARS-CoV-2 infection (MOI 0.5) and 1 μg/mL TPCK-treated trypsin. Cells were incubated for 48 h, and cells and conditioned media were harvested for ApoBD analysis by flow cytometry using a BD FACSAriaTM Fusion flow cytometer. ApoBD purification ApoBDs were purified from conditioned media of ABT-737-treated or SARS-CoV-2-infected Vero E6 cells (thus denoted ABT-ApoBDs and SARS-ApoBDs, respectively) using differential centrifugation as previously described39. In brief, apoptotic cell samples were centrifuged at 500× g for 10 min to remove cells. The resultant supernatant was then centrifuged at 3000 ×g for 20 min to pellet ApoBDs from smaller EVs. To control for residual extra-ApoBD virus in the SN and/or sticky viruses adhered to SARS-ApoBD, a separate ABT-ApoBD preparation was incubated with ApoBD-depleted SARS-CoV-2 supernatant, henceforth referred to as SN-ApoBDs. All ApoBD samples were washed three times with 1× PBS (Thermo Fisher) and subjected to flow cytometry analysis using ApoBD markers (FLICA activated caspase 3/7, annexin 5 (A5), TOPRO-3) for quality control prior to further assays. Caspase-Glo® caspase 3/7 assay Vero E6 or Calu-3 cells were seeded in 96-well plates at a concentration of 1 × 105 cells/mL for 24 h, followed by infection with 1:2 titrations of SARS-CoV-2 MOIs (0.5-0.03125). ABT-737 (1 μM) was included as an apoptosis control. After 48 h, Caspase-Glo® 3/7 reagent (Promega) was added to each well as per the manufacturer’s instructions. Cells were incubated for 3 h at 37 °C in a humidified atmosphere with 5% CO2 prior to luminescence reading using a spectrophotometer. Microscopy Transmission electron microscopy Purified ApoBDs were fixed in primary fixative 3% glutaraldehyde/PBS for 2 hrs at room temperature, embedded in 2.0% low melting point agarose and allowed to set on ice for 30 min. The agarose pellet was unstuck and incubated overnight in fresh primary fixative at 4 °C. The following day, samples were treated with 1% OsO4 (in 0.1 M cacodylate buffer) using the Pelco BioWave at 100 watts, 3 ×2 min. Following a buffer wash, sample dehydration was performed with 20%, 50%, 70%, 90% and 100% ethanol at 150 W. Next, resin infiltration was performed as follows at 250 W: 2 ×3 min with 75% ethanol/25% LR white acrylic resin, 2 ×3 min 50% ethanol/50% resin, 25% ethanol/75% resin, and 3 ×3 min 100% resin. Fresh LR-white acrylic resin was added, and samples were embedded at 60 °C for 72 h. Resin blocks were sliced into 50-nm-thick sections using DiATOME 45° Ultra Diamond Knife on a Leica EM UC7 ultramicrotome. These sections were mounted on EM-copper grids with formvar/carbon coating, post-stained in 4% aqueous uranyl acetate and Reynolds’ lead citrate before imaging with a JEOL JEM 2100 transmission electron microscope. Bright-field imaging Vero E6 or Calu-3 cells were seeded overnight in 96-well plates at 1 × 105 cells/mL. On the next day, freshly purified ApoBDs were added into each well (cell:ApoBD ratio of 1:3). After 48 h, cytopathic effect was visualised under the bright-field of a ZOETM Cell Imager (Bio-Rad). Fluorescence microscopy At 48 h post-SARS-CoV-2 infection or ApoBD co-incubation, MDMs were washed three times with 1× PBS, followed by fixation (with 2% paraformaldehyde), permeabilisation (with 0.5% saponin) and stained with Hoechst 34580 stain (1 μg/μL; ThermoFisher), Alexa Fluor™ 647 Phalloidin (165 nM; ThermoFisher) and Alexa Fluor® 488-conjugated anti-SARS-CoV-2 Spike S1 subunit antibody (1:200, RnD Systems). Fluorescence imaging was performed using a Zeiss LSM800 laser scanning confocal microscope. Differential interference contrast microscopy Cells were seeded in an 8-well chambered NuncTM Lab-TekTM II cover glass (Nunc) prior to anti-Fas (1:1000) and drug treatment. The chambers were pre-treated with 0.01% aqueous poly-L-lysine (Sigma-Aldrich) to enhance cell adherence. Time-lapse differential interference contrast imaging was performed at 37 °C with 5% CO2 using a spinning disc confocal microscope (Zeiss) with ×63 oil immersion objective. Image processing and analysis were performed using Zeiss imaging software. Calcium imaging Cells were stained with 2.5 μM cell-permeant Fluo-4 AM (Thermo Fisher)/0.02% pluronic acid F0127 in indicator-free media supplemented with 1% BSA at 37 °C for 30 min. Cells were then seeded in a poly-L-lysine-coated NuncTM Lab-TekTM II glass chamber, followed by apoptosis induction and drug treatment. Calcium influxes post-apoptosis induction were detected at 30 s interval using a spinning disc confocal microscope. Immunoblotting Cells or purified ApoBDs were lysed at 4 °C using ice-cold 1× cell lysis buffer (BD Biosciences). Protein concentration of the ApoBD lysates was determined using PierceTM bicinchoninic acid assay (Thermo Fisher). Subsequently, 30 μg of total protein per sample was subjected to reducing and denaturing SDS-PAGE prior to Western transfer onto nitrocellulose membrane (Bio-Rad). After blocking with 5% skim milk in PBS containing 0.1% (v/v) Tween®-20 (PBST), the membrane was probed with antibodies as followed: anti-human PANX1 (1:1,000), anti-human CaV3.2 (1:500), anti-human β-actin (1:1000) (all rabbit antibodies from Abcam) and horseradish peroxidase-conjugated donkey-anti-rabbit IgG (1:10,000; Cell Signaling Technology) in PBST containing 1% (w/v) BSA. Chemiluminescence detection was achieved using ECL Primer reagents (GE Healthcare, UK). For inflammation signalling detection, whole cell lysates of human MDMs (from three donors) were collected at indicated timepoints and subjected to immunoblotting using anti-ASC, anti-caspase-1, anti-GSDMD, anti-p65, anti-phospho-p65 (S536), anti-IκBα, anti-p38, anti-phospho-p38 (T180/Y182), anti-IL-1β and anti-GAPDH antibodies were used. For ASC cross-linking, BMDM lysates were treated with BS3 (bis(sulfosuccinimidyl)suberate) crosslinker (Thermo Fisher) prior to immunoblotting using anti-ASC. RT-qPCR RNA was isolated from TRIzol samples using Phasemaker tubes (Life Technologies) with chloroform:isoamyl alcohol (49:1 v/v, Sigma-Aldrich) and precipitated with isopropanol (Sigma-Aldrich), 0.3 M sodium acetate (Sigma-Aldrich) and 10 mg/mL linear polyacrylamide (Life Technologies) at −20 °C overnight. The precipitated RNA was washed twice with 80% ethanol and then resuspended in distilled H2O. RNA sample concentrations were measured using the Qubit 4 fluorometer (ThermoFisher) and the RNA Broad Range assay kit (ThermoFisher). SARS-CoV-2 nucleocapsid (N) positive control plasmid (250,000 copies per μL, IDT 10006625) was used to create a standard curve by performing seven two-fold dilutions (up to 1:128). iTaq Universal Probes One-Step Kit (Bio-Rad 1725141) master mix was made up according to the manufacturer’s instructions. Master mix was vortexed, and 16 µL was added to each well of a 96-well Light cycler plate, followed by 4 μL of RNA (standardised to 40 ng per sample), positive control (standards) or negative control (H2O). The plate was briefly pulsed in the centrifuge and sealed with a clear PCR plate seal (Roche Life Science). The LightCycler® instrument (Roche) was run using FAM dye as the reporter and the following cycling parameters:StepNo. of cyclesTemperatureTimePre-incubation150 °C10 min195 °C2 minTwo-step amplification5095 °C60 s60 °C30 sCooling137 °C30 s StepNo. of cyclesTemperatureTimePre-incubation150 °C10 min195 °C2 minTwo-step amplification5095 °C60 s60 °C30 sCooling137 °C30 s Data were analysed by performing an absolute quantification against the standard curve using the LightCycler® 96 analysis software (Roche). ACE2, RPL13A and TMPRSS2 expression were quantified as previously described67. TCID50 assay Purified ApoBDs were resuspended in infection media (DMEM containing 1 μg/mL trypsin-TPCK) to 1 × 106 ApoBDs/mL. Infectious viral titre was measured using TCID50 assay by plating 1:7 serially diluted ApoBD suspension onto confluent layers of Vero E6 cells in replicates of six on 96-well plates. Plates were incubated at 37 °C supplied with 5% CO2 for 4 days before scoring cytopathic effect under light microscope. The TCID50 calculation was performed using the Spearman Karber method71. For the murine lung TCID50 assay, left lungs were harvested and stored at −80 °C in serum-free DMEM until further processing. Lung samples were defrosted and homogenised in a Bullet Blender (Next Advance Inc) in 1 mL DME media (ThermoFisher) containing five steel homogenisation beads (Next Advance Inc). Lung homogenates were then centrifuged at 10,000×g for 5 min at 4 °C. Supernatant was retained for TCID50 assay to determine viral load. ApoBD and macrophage co-incubation Human MDMs were prepared by seeding 2.5 × 105 CD14+ monocytes onto a 48-well plate in RPMI 1640 media supplemented with 1 mM L-glutamine, 2 mM sodium pyruvate, 10% human adult AB serum and 5 ng/mL GM-CSF. After seven days, MDMs were washed with 1× PBS before co-incubation with freshly purified ApoBDs (MDM:ApoBD ratio of 1:3, equivalent to SARS-CoV-2 MOI of 3) or infection with SARS-CoV-2 (in the presence of 1 μg/mL trypsin-TPCK, MOI of 3) for 4 h. Infection media or excess ApoBDs were then removed and replaced with fresh RPMI 1640 media supplemented with 5% human adult AB serum, without GM-CSF. Cell viability assay The viability of SARS-CoV-2-infected or ApoBD-treated MDMs after 48 h were determined using a tetrazolium-based calorimetric assay as previously described72. Cytokine bead array Cytokines released by the SARS-CoV-2-infected or ApoBD-treated MDMs were measured using a human cytokine bead array as per the manufacturer’s instructions (BD Biosciences). Briefly, conditioned media from MDMs were collected at 48 h. A mixture of 2.5 μL each of captured cytokine beads and 15 μL of PE detection reagent was prepared per sample. This bead cocktail (30 μL) was incubated with an equal volume of cytokine standards or conditioned media in a 96-well plate and incubated for 3 h. After washing by centrifugation (200×g, 5 min, room temperature), samples were resuspended in 80 μL of wash buffer, then analysed using a BD FACSAriaTM Fusion flow cytometer. Data analysis was performed using FCAP Array software v.3 (BD Biosciences). In murine BALF experiments, BALF supernatant after 3000×g centrifugation was retained and subjected to cytokine bead array analysis as described. In vitro engulfment assay The ability of MDMs to efferocytose ApoBDs was quantified using an engulfment assay as previously described19. Briefly, macrophages and ApoBDs, pre-stained with CypHer5E NHS Ester (a pH-sensitive engulfment indicator, Sigma-Aldrich) were co-incubated at 1:3 ratio. At 4 h post-incubation, MDMs were harvested using Accutase® cell detachment solution (STEMCELL Technologies) for flow cytometry analysis vi BD FACSAriaTM Fusion. For inhibition of efferocytosis, CypHer5E-stained ApoBDs were blocked with BD PharmingenTM purified recombinant A5 in 1× A5 binding buffer (BD Biosciences; 50 μg for every 1 × 105 ApoBDs). Alternatively, MDMs were pre-treated with 5 μM cytochalasin D (Thermo Fisher) for 30 min before co-incubation. Proteomic analyses Whole cell lysates of human MDMs (from 5 donors), incubated with SARS-CoV-2 alone, ABT-ApoBDs, SARS-ApoBD or left untreated, were collected at 48 h post-treatment. An equal amount (30 μg) of protein samples were separated in SDS-PAGE. The separated protein bands were visualised using Coomassie Brilliant Blue stain, and the gel bands were then extracted using scalpel blades into five gel bands. The bands were then reduced using 10 mM DTT (Bio-Rad), alkylated with 25 mM iodoacetamide (Sigma) and trypsinised using 150 ng of trypsin (Promega). The tryptic peptides were extracted using 50% (w/v) acetonitrile and 0.1% (v/v) trifluoroacetic acid. LC MS/MS analysis was performed on an Orbitrap Ascend Mass Spectrometer (ThermoFisher Scientific). The LC systems were equipped with an Acclaim Pepmap Nano-trap column (Dinoex-C18, 100 Å, 75 µm × 2 cm) and an Acclaim Pepmap RSLC analytical column (Dinoex-C18, 100 Å, 75 µm × 50 cm). The tryptic peptides were injected into the enrichment column at an isocratic flow of 5 μL/min of 2% (v/v) CH3CN containing 0.05% (v/v) TFA for 6 min, applied before the enrichment column was switched in-line with the analytical column. The eluents were 0.1% (v/ v) FA (solvent A) in water and 100% (v/v) CH3CN in 0.1% (v/v) FA (solvent B), both supplemented with 5% DMSO. The gradient was at 300 nL/min from (i) 0–6 min, 3% B; (ii) 6–7 min, 3–4% B; (iii) 7–82 min, 4–25% B; (iv) 82–86 min, 25–40% B; (v) 86–87 min, 40–80% B; (vi) 87–90 min, 80–3% B; (vii) 90–90.1 min, 80–3% B and equilibrated at 3% B for 10 min before injecting the next sample. For DIA experiments, full MS scans were acquired at a resolution of 120,000 (at m/z 200), with a scan range of 350–1400 m/z in profile mode. The full MS AGC target was set to 250% with an injection time (IT) of 50 ms. The AGC target for fragment spectra was 2000%. Fragmentation was performed across 50 windows of 13.7 Da width with a 1 Da overlap. MS2 resolution was set to 30,000 with a maximum IT of 55 ms. Normalised collision energy (NCE) was set to 30%, and data were acquired in centroid mode with positive ion polarity. Data-independent acquisition (DIA) data was analysed using MaxDIA predicted algorithm within the MaxQuant software environment (version 2.6.2.0)73,74 was used for database search against the Homo Sapiens reference proteome (Uniprot downloaded March 2025). False discovery rate (FDR) was set to 1% at both peptide and protein levels with match between runs enabled. Enzyme specificity was set to Trypsin/P and to allow a maximum of two missed cleavages at C-termini of arginine and lysine, also when followed by proline. Cysteine carbamidomethylation was set as fixed modification. N-termini acetylation and methionine oxidation were considered as variable modifications. Other parameters were set to the software default. Label-free quantitation was achieved using the built-in MaxLFQ algorithm. Fluorescence-activated cell sorting FACS experiments were performed as per the engulfment assay but using 6-well plates (CD14+ monocytes seeded at 2 × 106 cells/well). After co-incubation with CypHer5E-stained ApoBDs, MDMs were detached and resuspended in FACS buffer (1× PBS, 1% FBS, 2 mM EGTA). Cell suspensions were strained through 70-μm nylon mesh filters into FACS tubes. ApoBD engulfers and non-engulfers (2 × 105 each) were separately sorted into complete media-containing collection tubes based on CypHer5E+ and CypHer5E−, respectively, using a BD FACSAriaTM Fusion flow cytometer and FACSDiva 6.1.1 software (BD Bioscience). Sorted MDMs were then pelleted at 500×g for 5 min and plated onto a 48-well plate. After 48 h, MDMs were harvested for flow cytometry and qRT-PCR analyses, whilst condition media were collected for human cytokine bead array. CRISPR/Cas9 gene editing Jurkat T cells deficient of CaV3.2 were generated using a CRISPR/Cas9 system as previously described75, using gRNA targeting CACNA1H exon 4: 5’-CAGGCCTTTGACGCCTTCAT-3’. Mibefradil conditioned media transfer assay Vehicle control or MBF (10 μM) was added onto Vero E6 cells following SARS-CoV-2 infection (MOI 0.5) and 1 μg/mL TPCK-treated trypsin. After 48 h incubation, conditioned media were collected and centrifuged at 3000×g for 20 min. After decanting supernatant containing small EVs and free viruses, pellets were resuspended in complete RPMI 1640 and transferred onto MDMs for 4 h. Subsequently, MDMs were washed with 1× PBS and further cultured in fresh RPMI 1640 media supplemented with 1 mM L-glutamine, 2 mM sodium pyruvate, 10% human adult AB serum for 48 h. BALF flow cytometry analysis BALF samples were collected from mice upon culling and centrifuged at 3000×g for 20 min to collect cells and ApoBDs for staining with a combination of LiveDead-APC stain (Thermo Fisher), SR-DEVD-FMK (i.e., FLICA) and A5-BV605 in staining buffer (1× A5 binding buffer, 2 mM EDTA 2% FCS, 1× PBS). After washing with staining buffer, samples were incubated with Fc block (BD Biosciences) and stained with a surface marker cocktail containing Ly6G-PE-Cy7 (1:300), F4/80-PerCP-Cy5.5 (1:150), CD45.2-AF700 (1:200), CD11C-APC-Cy7 (1:200), Ly6C-BV421 (1:300) and CD11b-V500 (1:300). All antibodies were purchased from BD Biosciences. Following fixation (with 2% paraformaldehyde) and permeabilisation (using 0.5% saponin), samples were incubated with FITC-conjugated antiviral spike antibodies (1:250, RnD Systems). At the end, SPHEROTM AccuCount beads (Spherotech) were added prior to flow cytometry analysis using a BD LSRFortessaTM Cell Analyser. Lung histological analysis The right lungs were harvested and fixed in 4% paraformaldehyde for 24 h before transferring to 70% ethanol for dehydration. Lung samples were then paraffin-embedded and sectioned for routine histology conducted by the WEHI Histology Facility. Slides were scanned using a 3D Histech Brightfield ´20 with CaseCenter software at ×20 magnification. Blinded lung histopathology scoring was performed, taking into consideration whole tissue distribution and applying none/normal (score of 0), minimal/mild (score of 1), moderate (score of 2), and severe (score of 3) to observed pathological changes, including inflammatory foci, proteinaceous debirs and fibrin strands, induction of bronchial associated lymphoid tissue, pleural mesothelial cell hyperplasia/subpleural inflammation, perivasculitis and perivascular marginating inflammatory cells. For the overall tissue damage score, a score of 0 (within normal ranges), 1 (<25% affected area), 2 (26 to 50% affected area) or 3 (>51% affected area) was applied. For alveolar thickness assessment, mock/vehicle block was used as a baseline sample when assessing the alveolar thickness of other samples. For the histopathological description, the presence of any pathological changes was described regardless if the changes were only focal. However, for scoring, the whole represented tissue was taken into consideration for the final score. Total lung histopathology scores were then calculated accordingly. Statistical analysis Data represent mean ± SEM. Statistical significance was determined using one-way ANOVA with a Tukey post hoc test or Mann–Whitney U test where appropriate. A P value of <0.05 was considered statistically significant. Sample size and number of replicates are specified within the figure legends. Reporting summary Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article. Supplementary information Supplementary Information Description Of Additional Supplementary File Supplementary Movie 1 Supplementary Movie 2 Supplementary Movie 3 Supplementary Movie 4 Supplementary Movie 5 Supplementary Movie 6 Supplementary Movie 7 Supplementary Movie 8 Supplementary Movie 9 Reporting summary Transparent Peer Review file Supplementary Information Description Of Additional Supplementary File Supplementary Movie 1 Supplementary Movie 2 Supplementary Movie 3 Supplementary Movie 4 Supplementary Movie 5 Supplementary Movie 6 Supplementary Movie 7 Supplementary Movie 8 Supplementary Movie 9 Reporting summary Transparent Peer Review file Source data Source data Source data