LRRC8 complexes are ATP release channels that regulate platelet activation and arterial thrombosis

Abstract

Platelet shape and volume changes are early mechanical events contributing to platelet activation and thrombosis. Here, we identify single-nucleotide polymorphisms in leucine-rich repeat-containing 8 (LRRC8) protein subunits that form the volume-regulated anion channel (VRAC), which are independently associated with altered mean platelet volume. LRRC8A is required for functional VRAC in megakaryocytes (MKs) and regulates platelet volume; adhesion; and agonist-stimulated activation, aggregation, adenosine triphosphate (ATP) secretion, and calcium mobilization. MK-specific LRRC8A conditional knockout mice have reduced laser injury-induced cremaster arteriolar thrombus formation and prolonged FeCl3 induced carotid arterial thrombosis without prolonged bleeding times. Mechanistically, platelet LRRC8A mediates swell-induced cytosolic ATP release to amplify agonist-stimulated calcium-phosphoinositide 3-kinase-protein kinase B signaling. Small-molecule LRRC8 channel inhibitors recapitulate defects observed in LRRC8A-null platelets in vitro and in vivo. These studies identify the mechanoresponsive LRRC8 channel complex as an ATP release channel in platelets, which positively regulates platelet function and thrombosis, providing a proof of concept for a novel antithrombotic drug target.

Graphical abstract

Figure 1.

LRRC8 proteins are highly expressed in platelets, functionally encode VRAC in MKs, and regulate platelet volume. (A,D) LRRC8 mRNA transcript expression in human (A) and mouse (D) platelets. (B,E) LRRC8 mRNA transcript expression compared with transcripts for ANO6 (TMEM16F), TRPV4, Piezo1, and Piezo2 in human (B) and mouse (E) platelets (data in panels A-B,D-E are from Rowley et al⁴⁸; supplemental Table 4). (C,F) Western blot showing LRRC8 subunit protein expression of human (n = 3) (C) and WT C57BL/6N mouse (n = 3) (F) platelets. (G-H) Western blot detecting P-selectin, LRRC8A, integrin β3, and GAPDH in platelets isolated from Lrrc8a^fl/fl (WT, n = 4) and Pf4-Cre;Lrrc8a^fl/fl (cKO, n = 4) mice (G), with densitometric quantification (H). (I-J) MPV in whole blood from WT (n = 8), cKO (n = 8), and cHets (Pf4-Cre;Lrrc8a^fl/+, n = 3) (I), with corresponding platelet counts (J). (K) MPVs of washed platelets isolated from WT (n = 35) and cKO (n = 31) mice. (L) Representative images of perforated patch-clamped freshly isolated MKs from WT or cKO mice, under ISO and HYPO conditions. (M) Cell capacitances of MKs isolated from WT (n = 14) and cKO mice (n = 10). (N) VRAC current-time relationship of WT MK cell induced by HYPO (210 mOsm) swelling, and subsequent inhibition upon application of 10 μM SN-401 (DCPIB). (O-P) VRAC current-voltage relationship after HYPO swelling during voltage ramps from −100 mV to +100 mV in MKs isolated from WT ± SN-401 (O) and cKO (P) mice. (Q) Mean outward (+100 mV) and inward (−100 mV) current densities measured in WT MKs after HYPO stimulation in the presence (green, n = 6) or absence (gray, n = 6) of 10 μM SN-401, or in cKO MKs after HYPO stimulation (red, n = 9). Data are represented as mean ± standard error of the mean (SEM). Statistical significance was determined by unpaired t test for panels H,K,M, Q; and by ordinary 1-way analysis of variance (ANOVA) with the Tukey multiple comparisons test for panels I-J. ∗P < .05; ∗∗P < .01; ∗∗∗P < .001. cHet, heterozygous control; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; HYPO, hypotonic; ISO, isotonic; ns, not significant; RPKM, reads per kilobase million.

Figure 2.

LRRC8A deletion impairs agonist-induced platelet adhesion, activation, aggregation, Ca²⁺ signaling, and PI3K-AKT-GSK3β signaling. (A-B) Adhesion of platelets to a fibrillar collagen-coated surface under shear (A, 650 s⁻¹; B, 2600 s⁻¹) conditions in whole blood isolated from Lrrc8a^fl/fl (WT, n = 4-6) and Pf4-Cre;Lrrc8a^fl/fl (cKO, n = 4-5) mice. (C-D) P-selectin exposure (C) and JonA binding to activated αIIbb3 integrin (D) as measured by flow cytometry of platelets isolated from WT and cKO mice and stimulated with thrombin (Thr; 0.03 U/mL; n = 6 per group), CRP (0.2 μg/mL; n = 3), or the calcium ionophore A23187 (0.5 μM; n = 6 per group). (E-I) Aggregation of platelets isolated from WT and cKO mice stimulated with Thr (0.05 U/mL, n = 7 per group) (E), ADP (ADP, 10-20 μM; fibrinogen, 30 μg/mL; n = 6 per group) (F), CRP (0.05-0.15 μg/mL; n = 7 per group) (G), U46619 (0.5-0.8 μM; n = 6 per group) (H), and calcium ionophore A23187 (1 μM; n = 3 per group) (I). Concurrent ATP release from platelets treated with Thr (J), CRP (K), and U46619 (L). AUC of ATP release (M). Representative traces for aggregation and ATP release are provided in supplemental Figure 5. (N) Intracellular ATP content in platelets isolated from WT (n = 7) and cKO (n = 8) mice. (O-Q) Cytosolic Ca²⁺ measured using the ratiometric dye Fura-2 in platelets isolated from WT (n = 11) and cKO (n = 9) mice stimulated with Thr (0.02 U/mL) for 10 minutes (O), with corresponding measurements of peak Ca²⁺ (P) and rate of Ca²⁺ rise (Q). (R-S) Western blots detecting LRRC8A; pPI3K^Tyr458; pAKT2^Ser474; AKT2; pGSK3β^Ser9; GSK3β; and GAPDH in platelets from WT and cKO mice after stimulation with PAR4-AP (400 μM) for 5 minutes during aggregometry (R), with densitometric quantification (S). Data are represented as mean ± SEM. Statistical significance was determined by unpaired t test for panels A-L,N,P,Q,S; and by 2-way ANOVA for panel O. For panel M, statistical significance was determined by unpaired t test (Thr) or Mann-Whitney U test (CRP and U46619). ∗P < .05; ∗∗P < .01; ∗∗∗P < .001. Thr used in panels C-E,J,M was sourced from Sigma-Aldrich (St. Louis, MO), whereas Thr used in panels O-Q was sourced from the Chrono-Log Corporation (Havertown, PA). a.u., arbitrary units; AUC, area under curve; GSK3β, glycogen synthase kinase-3β; pAKT2, phosphorylated AKT2; pGSK3β, phosphorylated glycogen synthase kinase-3β; pPI3K, phosphorylated PI3K; ROI, region of interest.

Figure 2.

LRRC8A deletion impairs agonist-induced platelet adhesion, activation, aggregation, Ca²⁺ signaling, and PI3K-AKT-GSK3β signaling. (A-B) Adhesion of platelets to a fibrillar collagen-coated surface under shear (A, 650 s⁻¹; B, 2600 s⁻¹) conditions in whole blood isolated from Lrrc8a^fl/fl (WT, n = 4-6) and Pf4-Cre;Lrrc8a^fl/fl (cKO, n = 4-5) mice. (C-D) P-selectin exposure (C) and JonA binding to activated αIIbb3 integrin (D) as measured by flow cytometry of platelets isolated from WT and cKO mice and stimulated with thrombin (Thr; 0.03 U/mL; n = 6 per group), CRP (0.2 μg/mL; n = 3), or the calcium ionophore A23187 (0.5 μM; n = 6 per group). (E-I) Aggregation of platelets isolated from WT and cKO mice stimulated with Thr (0.05 U/mL, n = 7 per group) (E), ADP (ADP, 10-20 μM; fibrinogen, 30 μg/mL; n = 6 per group) (F), CRP (0.05-0.15 μg/mL; n = 7 per group) (G), U46619 (0.5-0.8 μM; n = 6 per group) (H), and calcium ionophore A23187 (1 μM; n = 3 per group) (I). Concurrent ATP release from platelets treated with Thr (J), CRP (K), and U46619 (L). AUC of ATP release (M). Representative traces for aggregation and ATP release are provided in supplemental Figure 5. (N) Intracellular ATP content in platelets isolated from WT (n = 7) and cKO (n = 8) mice. (O-Q) Cytosolic Ca²⁺ measured using the ratiometric dye Fura-2 in platelets isolated from WT (n = 11) and cKO (n = 9) mice stimulated with Thr (0.02 U/mL) for 10 minutes (O), with corresponding measurements of peak Ca²⁺ (P) and rate of Ca²⁺ rise (Q). (R-S) Western blots detecting LRRC8A; pPI3K^Tyr458; pAKT2^Ser474; AKT2; pGSK3β^Ser9; GSK3β; and GAPDH in platelets from WT and cKO mice after stimulation with PAR4-AP (400 μM) for 5 minutes during aggregometry (R), with densitometric quantification (S). Data are represented as mean ± SEM. Statistical significance was determined by unpaired t test for panels A-L,N,P,Q,S; and by 2-way ANOVA for panel O. For panel M, statistical significance was determined by unpaired t test (Thr) or Mann-Whitney U test (CRP and U46619). ∗P < .05; ∗∗P < .01; ∗∗∗P < .001. Thr used in panels C-E,J,M was sourced from Sigma-Aldrich (St. Louis, MO), whereas Thr used in panels O-Q was sourced from the Chrono-Log Corporation (Havertown, PA). a.u., arbitrary units; AUC, area under curve; GSK3β, glycogen synthase kinase-3β; pAKT2, phosphorylated AKT2; pGSK3β, phosphorylated glycogen synthase kinase-3β; pPI3K, phosphorylated PI3K; ROI, region of interest.

Figure 3.

MK LRRC8 channels permeate ATP. (A-C) Current-time relationship of inward I_ATP (ATP efflux) and outward VRAC in MKs isolated from Lrrc8a^fl/fl (WT) mice induced by HYPO (210 mOsm) swelling with an intracellular ATP concentration of 1 mM (A) or 50 mM (B) followed by application of 10 μM SN-401 (DCPIB), and in MKs isolated from Pf4-Cre;Lrrc8a^fl/fl (cKO) mice (C) with an intracellular ATP concentration of 50 mM. The inward component of the current represents I_ATP generated from ATP efflux. (D-F) Respective current-voltage relationships of inward I_ATP (ATP efflux) and outward VRAC elicited from voltage ramps from −140 mV to +80 mV. (G) Mean current densities of inward I_ATP (ATP efflux) at −140 mV in MKs isolated from WT mice after HYPO swelling, with an intracellular concentration of 1 mM ATP (n = 7) or 50 mM ATP (n = 6), and inhibition of ATP efflux by 10 μM SN-401 (n = 7 for 1 mM ATP + SN-401; n = 6 for 50 mM ATP + SN-401), and mean current densities of inward I_ATP in MKs isolated from cKO mice after HYPO swelling, with an intracellular concentration of 50 mM ATP (n = 7). (H-I) Current-voltage relationship of inward I_ATP (ATP efflux) and outward VRAC in MEG-01 cells elicited from voltage ramps from −140 mV to +80 mV before and after HYPO swelling with an intracellular ATP concentration of 1 mM (H) or 50 mM (I), followed by application of 10 μM SN-401 (DCPIB). The inward component of the current represents I_ATP generated from ATP efflux. (J) Mean current densities of inward I_ATP (ATP efflux) in MEG-01 cells at −140 mV after HYPO swelling, with an intracellular concentration of 1 mM ATP (n = 5) or 50 mM ATP (n = 4), and inhibition of ATP efflux by 10 μM SN-401 (n = 5 for 1 mM ATP + SN-401; n = 4 for 50 mM ATP + SN-401). (K-L) Current-time relationship of inward I_ATP and outward VRAC induced by HYPO (210 mOsm) swelling in MEG-01 cells transduced with adenoviral shSCR (K) or shLRRC8A (L). (M) Current-voltage relationship of inward I_ATP (ATP efflux) and outward VRAC during voltage ramps from −140 mV to +80 mV after HYPO swelling in MEG-01 cells treated with either shSCR or shLRRC8A. (N) Mean current densities of inward I_ATP (ATP efflux) in MEG-01 cells treated with shSCR (n = 7) or shLRRC8A (n = 7) at −140 mV after HYPO swelling. Data are represented as mean ± SEM. Statistical significance was determined by unpaired t test for panels G,J,N. ∗P < .05; ∗∗P < .01; ∗∗∗P < .001. shLRRC8A, short hairpin RNA–targeting LRRC8A; shSCR, short hairpin control.

Figure 4.

LRRC8A deletion impairs ATP release despite preserved dense granule secretion. (A-L) Aggregation (A,E,I), P-selectin exposure (B,F,J), ATP release (C,G,K), and CD63 exposure (D,H,L) in platelets isolated from WT (n = 4) and cKO (n = 4) mice stimulated with 0.02 U/mL (A-D), 0.06 U/mL (E-H) and 0.08 U/mL (I-L) Thr. (M) Serotonin secretion from platelets isolated from WT (n = 4) and cKO (n = 4) mice stimulated with 0.08 U/mL Thr. (N) Summary of ATP release from platelets isolated from WT and cKO mice stimulated with 0.02 to 0.08 U/mL Thr (n = 4 for all groups). (O) Summary of P-selectin exposure in platelets isolated from WT and cKO mice stimulated with 0.02 to 0.08 U/mL Thr (n = 4 for all groups). (P) Summary of CD63 expression in platelets isolated from WT and cKO mice stimulated with 0.02 to 0.08 U/mL Thr (n = 4 for all groups). (Q) Summary of serotonin secretion from platelets isolated from WT and cKO mice stimulated with 0.02 to 0.08 U/mL Thr (n = 4 for all groups). (R) ATP-to-serotonin ratios for platelets isolated from WT and cKO mice stimulated with 0.02 to 0.08 U/mL Thr (n = 4 for all groups). (S) Inhibition of ATP and serotonin secretion from platelets isolated from cKO mice stimulated with 0.08 U/mL Thr (n = 4 for all groups) relative to ATP and serotonin secretion from platelets isolated from WT mice also treated with 0.08 U/mL Thr. (T) Aggregometry of platelets isolated from WT and cKO mice stimulated with 400 μM PAR4-AP only (n = 5 for WT; n = 5 for cKO), or with 400 μM PAR4-AP in the presence of 40 μM ATP (n = 4 for WT; n = 4 for cKO). Data are represented as mean ± SEM. Statistical significance was determined by unpaired t test for panels B,D,F,H,J,L-Q,S; and by 2-way ANOVA for panels A,C,E,G,I,K,R,T. ∗P < .05; ∗∗P < .01; ∗∗∗P < .001. Thr used in this experiment was sourced from the Chrono-Log Corporation. Data for panels G,K were calculated based on extrapolation as described in supplemental Methods and shown in supplemental Figure 7. Representative aggregometry traces for panels A,E,I, and ATP traces for panels C,G,K are also provided in supplemental Figure 7, and for panel T is provided in supplemental Figure 8. A.U., arbitrary units; ctrl, control; MFI, mean fluorescence intensity; ns, not significant.

Figure 5.

Platelet-targeted LRRC8A deletion impairs injury-induced arterial thrombosis, without prolonging tail bleeding. (A-C) Thrombosis in the cremasteric arterioles of Lrrc8a^fl/fl (WT) and Pf4-Cre;Lrrc8a^fl/fl (cKO) mice after laser-induced injury as observed by real-time confocal intravital microscopy (red, DyLight 649 anti-CD42C) (A) with median time course (B) and quantification of the fluorescent reporter at 60, 120, and 180 seconds after injury (n = 37 for both WT and cKO) (C). (D) Representative traces of arterial blood flow through the carotid artery of WT (top) and cKO (bottom) mice after FeCl₃-induced injury. (E) Median time taken for the carotid artery of WT (n = 8) and cKO (n = 7) mice to occlude after FeCl₃-induced injury. (F-G) Total bleeding time after tail tip amputation of WT (n = 10) and cKO (n = 10) mice (F), with corresponding Hb concentrations of red blood cell lysates collected during the assay (G). Data in panel C are represented as mean only, whereas data in panels B,E are represented as median. Data in panels F,G are represented as mean ± SEM. Statistical significance was determined by the Mann-Whitney U test for panels C,E, and by unpaired t test for panels F-G. ∗P < .05; ∗∗P < .01; ∗∗∗P < .001. Hb, hemoglobin.

Figure 6.

LRRC8 channel small-molecule inhibitors suppress agonist-induced aggregation, ATP release, and calcium influx. (A) Molecular structure of the small-molecule LRRC8 channel complex modulator SN-401 (DCPIB) and SN-406, a derivative of SN-401. (B) Current-time relationship of inward I_ATP (ATP efflux) and outward VRAC induced by HYPO swelling in MEG-01 cells and subsequent inhibition by application of 10 μM SN-406. (C-D) Current-voltage relationship of inward I_ATP (ATP efflux) and outward VRAC during voltage ramps from −140 mV to +80 mV in MEG-01 cells after HYPO swelling in the absence or presence of 10 μM SN-406 (C), with mean current densities of inward I_ATP at −140 mV (n = 7-15) (D). (E-H) Aggregometry of platelets isolated from WT C57BL/6J mice stimulated with Thr (0.02 U/mL) in the presence of vehicle (0.02% dimethyl sulfoxide [DMSO]; n = 37), SN-401 (10 μM; n = 7), or SN-406 (10 μM; n = 5) (E), with concurrent ATP release (n = 5-30) (F-H). (I) Cytosolic calcium measured using the ratiometric dye Fura-2 in platelets isolated from WT C57BL/6J mice stimulated with PAR4-AP (400 μM; n = 8) in the presence of vehicle only (0.02% DMSO) or SN-406 (10 μM). (J-K) Aggregometry of human platelets stimulated with PAR1-AP and PAR4-AP (10 μM and 100 μM, respectively) in the presence of vehicle (0.02% DMSO; n = 10 from 3 healthy volunteers) or SN-406 (10 μM; n = 11 from 3 healthy volunteers) (J), with measurements of cytosolic calcium (K). (L) Aggregometry of PAR4-AP (400 μM) stimulated mouse platelets in the presence of vehicle (0.02% DMSO; n = 35), and SN-406 at 0.1 μM (n = 5), 1 μM (n = 5), or 10 μM (n = 6). (M) SN-406 dose-dependent inhibition of PAR4-AP (400 μM) stimulated mouse platelet aggregation (SN-406: 0-10 μM; n = 5-6). (N) Aggregometry of platelets isolated from Pf4-Cre;Lrrc8a^fl/fl (cKO) and Lrrc8a^fl/fl (WT) littermate controls stimulated by PAR4-AP (400 μM) in the presence of vehicle (0.02% DMSO; n = 5 for WT; n = 7 for cKO) or 1.1 μM SN-406 (n = 6 for WT; n = 8 for cKO). (O) Current-time relationship of inward I_ATP and outward VRAC induced by HYPO (210 mOsm) swelling in MEG-01 cells and subsequent inhibition by application of 10 μM dicoumarol. (P-Q) Current-voltage relationship of inward I_ATP and outward VRAC during voltage ramps from −140 mV to +80 mV in MEG-01 cells after HYPO swelling in the absence or presence of 10 μM dicoumarol (P), with mean current densities of inward I_ATP –140 mV (Q) (n = 4 for both groups). (R) Aggregometry of platelets isolated from WT C57BL/6J mice stimulated with PAR4-AP (400 μM) in the presence of vehicle (0.1% DMSO; n = 4) or dicoumarol (20 μM; n = 3), with subsequent application of ATP (40 μM). Data are represented as mean ± SEM. Statistical significance was determined by unpaired t test for panels D,G,H,Q; and by 2-way ANOVA for E, F, I-K,N,R. Thr used in this experiment was sourced from the Chrono-Log Corporation. Representative traces for panels E-F,J,N,R are provided in supplemental Figure 10. ∗P < .05; ∗∗P < .01; ∗∗∗P < .001. a.u., arbitrary unit.

Figure 7.

LRRC8 channel inhibitor SN-89B suppresses agonist-induced platelet activation, aggregation, PI3K-AKT-GSK3β signaling, and in vivo thrombosis. (A-B) Current-voltage relationship of inward I_ATP (ATP efflux) and outward VRAC during voltage ramps from −140 mV to +80 mV in MEG-01 cells after HYPO swelling in the absence or presence of 10 μM SN-89B (A); mean current densities of inward I_ATP at −140 mV (n = 4) (B). (C-D) P-selectin exposure (C) and PAC-1 binding to activated αIIbb3 integrin (D) as measured by flow cytometry of human platelets stimulated with PAR1-AP and PAR4-AP (10 μM and 100 μM, respectively) in the presence of vehicle (0.02% DMSO) or SN-89B (10 μM). Each sample was normalized to the average of unstimulated controls (n = 8 for all groups, isolated from 2 healthy volunteers). (E) Adhesion of platelets in reconstituted blood (platelets + red blood cells) to a von Willebrand factor–coated surface under shear (1800 s⁻¹) conditions in the presence of vehicle (0.02% DMSO) or 10 μM SN-89B (n = 3 for both groups, isolated from 3 healthy volunteers), as determined by fluorescence microscopy. (F) Aggregometry of human platelets stimulated with PAR1-AP and PAR4-AP (10 μM and 100 μM, respectively) in the presence of vehicle (0.02% DMSO; n = 6; isolated from 1 healthy volunteer) or SN-89B (10 μM; n = 4; isolated from 1 healthy volunteer). Representative traces are provided in supplemental Figure 14. (G-H) Cytosolic calcium measured using the ratiometric dye Fura-2 in human platelets stimulated with Thr (0.01 U/mL) in the presence of vehicle (0.02% DMSO) or 10 μM SN-89B (n = 6 for both groups, isolated from 1 healthy volunteer) (G), with corresponding measurements of peak Ca²⁺ and rate of Ca²⁺ rise (H). (I-J) Western blots detecting pAKT1^Ser473; AKT1; pAKT2^Ser474; AKT2; pGSK3β^Ser9; GSK3β; and GAPDH in human platelets stimulated with PAR1-AP and PAR4-AP (10 μM and 100 μM, respectively), in the presence of vehicle (0.02% DMSO; n = 3; isolated from 1 healthy volunteer) or SN-89B (10 μM, n = 3, isolated from 1 healthy volunteer) (I), with densitometric quantification of pAKT1, pAKT2 and pGSK-3β normalized to their respective unphosphorylated forms and GAPDH (J). (K-M) Thrombosis in the cremasteric arterioles of WT mice treated with 50 mg/kg SN-89B (n = 5) or vehicle (n = 5) after laser-induced injury as observed by real-time confocal intravital microscopy (red, DyLight 649 anti-CD42C) (K) with median time course (L) and quantification of the fluorescent reporter at 30, 60, 120, 180, and 240 seconds after injury (M) (n = 36 for vehicle; n = 38 for SN-89B). (N) Time to occlusion of the carotid artery of WT C57BL/6J mice treated with 50 mg/kg SN-89B (n = 8) or vehicle (n = 5) after FeCl₃-induced injury. (O-Q) Cumulative bleeding times (O) and times taken for bleeding to cease for the first time (before any rebleeding) (P) after tail tip amputation of WT C57BL/6J mice treated with vehicle (n = 13), 50 mg/kg SN-89B (n = 13), or 30 mg/kg ticagrelor (n = 10), with corresponding hemoglobin concentrations of red blood cell lysates collected during the assay (Q). Data in panel L are represented as median, whereas data in panel M are represented as mean only. All other data are represented as mean ± SEM. Statistical significance was determined by the unpaired t test for panels B-D,H,J,N,P-Q; the Mann-Whitney U test for panel M; and by the 2-way ANOVA for panels E-G. Statistical significance for panel O was determined by the unpaired t test (vehicle compared with SN-89B) or the Mann-Whitney U test (ticagrelor compared to vehicle and SN-89B). ∗P < .05; ∗∗P < .01; ∗∗∗P < .001. Thr used in this experiment was sourced from the Chrono-Log Corporation. F platelets, platelet fluorescence; ns, not significant; ROI, region of interest.