Fig 2: Cumulative hazard curves for infection-related hospitalization by half of the study population according to serum ACE2 (sACE2) levels.

Fig 3: Functionality of recombinant SARS-CoV-2 spike protein and levels of ACE2 expression in ALI cultures. (A) shows the levels of expression of ACE2 isoforms (glycosylated, non-glycosylated, and short), as detected by Western blot, in 21-day ALI cultures that were resubmerged for the indicated times. Data are shown as ACE2 expression normalized to β-actin, SEM. Statistical differences were assessed using Kruskal–Wallis tests for multiple comparisons. (B) shows the results of recombinant SARS-CoV-2 binding to ACE-2 assays done by ELISA, as described in the Methods section. Data are indicated as mean absorbance at 450 nm, SEM; each datapoint is represented with black circles. Statistical differences were calculated using Student’s t-Test in GraphPad Prism, p < 0.05 symbolized by * and p < 0.01 symbolized by **.

Fig 4: Cumulative hazard curves for respiratory infection-related hospitalization by half of the study population according to serum ACE2 (sACE2) levels.

Fig 5: Circulating evACE2 increased in the peripheral blood of COVID-19 patients.a ACE2+ EVs detected in human plasma samples of sero-negative controls (light blue), acute phase (dark green), and convalescent COVID-19 patients (green). One-tail t test (*p = 0.038, **p = 0.0061 and **p = 0.0016). Data are presented as mean values ± SEM. b Representative microflow vesiclometry (MFV) plots with gated ACE2+ EVs from sero-negative, acute phase and convalescent COVID-19 patients. c MFV detection of circulating ACE2+ EVs with CD63+ EVs in human plasma of convalescent COVID-19 patient samples (CSB-029 and CSB-023) (green line). Blue line is isotype IgG-negative control. d Flow profiles of ACE2 expression in HEK and HeLa parental control cells (Con, light blue line, ACE2-) and with ACE2 overexpression (ACE2, green line). e NanoSight NTA analysis of the sizes of HEK-derived ACE2- (ev1Con) and ACE2+ (ev1ACE2) and HeLa-derived ACE2- (ev2Con) and ACE2+ (ev2ACE2). f Immunoblots of HEK and HeLa (ACE2- and ACE2+) EVs and cell lysates for ACE2, TSG101, CD63, CD81, GRP94 and loading control of the membrane proteins upon Ponceau staining. RIPA buffer and Bradford protein assay were used for cells/EVs lysis and protein measurement, respectively (N = 1 experiment). g Cryo-EM images of HEK-derived EVs, ACE2- (evCon, left) and ACE2+ (evACE2, right), stained with ACE2 (top) and CD81 (bottom). Scale bars = 100 nm. h Quantified counts of Apogee MFV-based total extracellular vesicles (EVs) and ACE2+ EVs (N = 2 experiments with n = 6 technical replicates for total EV particles and n = 3 technical replicates for ACE2+ counts). Control EVs are in light blue and ACE2+ EVs in green. Data are presented as mean values +/- SD. i Overlay flow profiles of ACE2 positivity within CD63+ (left column) and CD81+ (right column) EVs isolated from HEK-ACE2 (top row) and HeLa-ACE2 (bottom row) cells, respectively (n = 3 technical replicates). Light blue line for Control EVs and green line for ACE2+ EVs.