Upregulation of cAMP prevents antibody-mediated thrombus formation in COVID-19

Abstract

Thromboembolic events are frequently reported in patients infected with the SARS-CoV-2 virus. The exact mechanisms of COVID-19-associated hypercoagulopathy, however, remain elusive. Recently, we observed that platelets (PLTs) from patients with severe COVID-19 infection express high levels of procoagulant markers, which were found to be associated with increased risk for thrombosis. In the current study, we investigated the time course as well as the mechanisms leading to procoagulant PLTs in COVID-19. Our study demonstrates the presence of PLT-reactive IgG antibodies that induce marked changes in PLTs in terms of increased inner-mitochondrial transmembrane potential (Δψ) depolarization, phosphatidylserine (PS) externalization, and P-selectin expression. The IgG-induced procoagulant PLTs and increased thrombus formation were mediated by ligation of PLT Fc-γ RIIA (FcγRIIA). In addition, contents of calcium and cyclic-adenosine-monophosphate (cAMP) in PLTs were identified to play a central role in antibody-induced procoagulant PLT formation. Most importantly, antibody-induced procoagulant events, as well as increased thrombus formation in severe COVID-19, were inhibited by Iloprost, a clinically approved therapeutic agent that increases the intracellular cAMP levels in PLTs. Our data indicate that upregulation of cAMP could be a potential therapeutic target to prevent antibody-mediated coagulopathy in COVID-19 disease.

Graphical abstract

Figure 1.

ICU COVID-19 patient serum-induced effects on PLTs during disease. (A-B) ICU COVID-19 (n = 26) or ICU non-COVID-19 control (n = 5) patient serum-induced changes in Δψ as well as PS externalization in wPLTs were analyzed by FC. (C-D) Sera of 4 ICU COVID-19 patients were collected for up to 14 days during hospitalization and analyzed for their ability to induce changes in Δψ as well as PS externalization in wPLTs via FC. Data are presented as mean ± SEM of the measured fold increase (FI) compared with control. The number of patient sera tested is reported in each graph. Dot lines in (A,B) represent the calculated cutoffs determined by testing sera from healthy donors as mean of FI + 2 × SEM. *P < .05, **P < .01, ***P < .001, and ****P < .0001. ns, not significant; FC, flow cytometry; HC, healthy control; Δψ, inner mitochondrial transmembrane potential; N, number of HCs or patients; PS, phosphatidylserine.

Figure 2.

ICU COVID-19 IgGs induce procoagulant PLTs, GPVI cleavage, increased thrombin generation and thrombus formation on collagen. (A) HC, ICU non-COVID-19, and ICU COVID-19 IgG-induced changes in wPLTs externalization of PS and expression of CD62p were analyzed via annexin V-FITC and CD62p-APC antibody staining, respectively. Data are shown as mean percentage ± SEM of annexin V-FITC and or CD62p-APC-positive-labeled wPLTs. (B) ICU-COVID-19 IgG-induced reductions in the expression of GPVI on the PLT surface were analyzed by GPVI-PE antibody staining and compared as percentage of GPVI-negative PLTs ± SEM to the controls. (C) PRP from healthy individuals was preincubated with HC or ICU COVID-19 IgG and analyzed for thrombin generation using CAT. Each curve represents the amounts of generated thrombin over time induced by HC (dotted blue line) or IgG from different ICU COVID-19 patients (red line) (D) PRP from healthy individuals with the blood group O was incubated with HC, ICU non-COVID-19 control, or ICU COVID-19 IgG, labeled with FITC conjugated calcein and perfused through microfluidic channels at a shear rate of 1500⁻¹ (60 dyne) for 5 minutes after reconstitution into autologous whole blood. Images were acquired at x20 magnification in the fluorescent (upper panel) as well as in the BF channel (lower panel). Scale bar 50 µm. (E) Mean percentage of surface are covered (SAC) by thrombus ± SEM in the presence of HC, ICU non-COVID-19 control, and ICU COVID-19 IgG after 5 minutes perfusion time. The number of patients and healthy donors tested is reported in each graph. See Figure 1 for P values and abbreviation definitions.

Figure 3.

ICU COVID-19 IgG-induced formation of procoagulant PLTs is FcγRIIA dependent. (A) ICU COVID-19 patient serum-induced changes of Δψ as well as ICU COVID-19 IgG-induced formation (B) of procoagulant PLTs were analyzed in the presence or absence of moAb IV.3 via FC. Data are presented as mean percentage of (A) Δψ depolarized PLTs and (B) CD62p/PS-positive PLTs ± SEM. The number of patients and healthy donors tested is reported in each graph. See Figure 1 for P values and abbreviation definitions.

Figure 4.

FcγRIIA inhibition prevents ICU COVID-19 IgG-induced thrombus formation. (A) PRP from healthy individuals with the blood group O was incubated with HC or ICU COVID-19 IgG in the presence of moAb IV.3 or isotype control (moAb) and perfused through microfluidic channels at a shear rate of 1500⁻¹ (60 dyne) for 5 minutes. Images were acquired at x20 magnification in fluorescent (upper panel) as well as in the BF channel (lower panel). Scale bar 50 µm. (B) Mean percent of SAC ± SEM induced by HC or ICU COVID-19 IgG in the presence or absence of moAb IV.3 or isotype control (moAb). The number of patients and healthy donors tested is reported in each graph. See Figure 1 for P values and abbreviation definitions.

Figure 5.

ICU COVID-19 IgG-induced formation of procoagulant PLTs is dependent on calcium. Panel (A-B) shows ICU COVID-19 IgG-induced changes of different PLT markers in the presence (vehicle) or absence of extracellular calcium (EGTA 1 mM). FC detected changes of (A) PLT Δψ and (B) formation of CD62p/PS-positive PLTs after ICU COVID-19 IgG incubation in vehicle or EGTA pretreated wPLTs, respectively. (C-D) ICU COVID-19 IgG-induced PLT changes in the presence (vehicle) or in intracellular calcium depleted (BAPTA 20 µM) wPLTs. FC detected changes of (C) PLT Δψ and (D) formation of CD62p/PS-positive PLTs induced by ICU COVID-19 IgG in vehicle or BAPTA preloaded wPLTs, respectively. Data are presented as mean percentage ± SEM of (A,C) Δψ depolarized PLTs and mean percentage ± SEM of (B,D) PS (Lactadherin-FITC) and CD62p-APC-positive wPLTs. Note that lactadherine is a calcium-independent marker of PS externalization. The number of patients and healthy donors tested is reported in each graph. See Figure 1 for P values and abbreviation definitions.

Figure 6.

cAMP elevation prevents ICU COVID-19 IgG-induced formation of procoagulant PLTs. Panels (A-B) show ICU COVID-19 IgG-induced changes of procoagulant PLT markers in vehicle or Forskolin (2.25 µM) preincubated wPLTs. (A-B) ICU COVID-19 IgG-induced changes of (A) Δψ and (B) formation of CD62p/PS-positive PLTs were analyzed in vehicle or Forskolin preincubated wPLTs by FC. Panels (C-D) show ICU COVID-19 IgG-induced PLT changes in the presence of vehicle or Iloprost (20 nM). FC detected changes of (C) Δψ and (D) formation of CD62p/PS-positive PLTs induced by ICU COVID-19 IgG in vehicle or Iloprost preincubated wPLTs, respectively. (A,C) Data are presented as mean percentage ± SEM of Δψ depolarized PLTs and (B,D) mean percentage ± SEM of PS and CD62p-APC-positive-labeled wPLTs. The number of patients and healthy donors tested is reported in each graph. See Figure 1 for P values and abbreviation definitions.

Figure 7.

Iloprost inhibits ICU COVID-19 IgG-induced increased thrombus formation. (A) PRP from healthy individuals with the blood group O was incubated with HC or ICU COVID-19 IgG in the presence of vehicle or Iloprost (20 nM). After reconstitution into autologous whole blood, samples were perfused through microfluidic channels at a shear rate of 1500^-1 (60 dyne) for 5 minutes. Pictures were acquired at x20 magnification in fluorescent (upper panel) as well as in the BF channel (lower panel). Scale bar 50 µm. (B) Mean percent of SAC by thrombus ± SEM induced by HC or ICU COVID-19 IgG in the presence of vehicle or Iloprost (20 nM). The number of patients and healthy donors tested is reported in each graph. See Figure 1 for P values and abbreviation definitions.