Selective degradation of platelet BTK by PROTAC NX-5948 provides antithrombotic benefits without affecting haemostasis

Figures

• 

  Figure 1.. Concentration-dependent reduction of BTK expression by NX-2127 and NX-5948 in human platelets. (A-D) Concentration response of NX-2127 and NX-5948 in washed human platelets (A-B) and PRP (C-D). Platelets were treated with the indicated concentrations of NX-2127 (A,C) and NX-5948 (B,D) for 4 hours at 30°C. BTK levels were quantified by densitometry of western blot and normalized to vehicle (dimethyl sulfoxide [DMSO])–treated control. The DC50 in washed platelets for NX-2127 and NX-5948 was 12nM and 5.3nM, respectively (calculated using nonlinear regression in GraphPad Prism). The DC50 in PRP for NX-2127 and NX-5948 was 98nM and 12.8nM, respectively. (E-F) Time-dependent degradation of BTK by NX-2127 (E) and NX-5948 (F) in human PRP. Human platelets in PRP were treated with 3.0μM of NX-2127 and 100nM of NX-5948 for the indicated time points, ranging from 30 minutes to 270 minutes. BTK levels were quantified by densitometry of western blot and normalized to vehicle (DMSO)–treated control. The curves were generated from quantification of western blots (mean ± standard error of the mean [SEM]; n = 3 biological replicates), with representative blots shown above the respective curve.…

• 

  Figure 2.. Specificity of degradation by NX-2127 and NX-5948 in human platelets. (A-C) Human platelets in PRP were treated with 3.0μM NX-2127 (A) or 100nM NX-5948 (B) for 4 hours at 30°C before being washed, lysed, and analyzed by TMTpro-based quantitative proteomic analysis (A-B) or western blotting (C). Proteins significantly downregulated (P < .05) are shown in red. Proteins exhibiting >75% (chosen a priori) reduction relative to vehicle-treated levels and meeting the same significance threshold are shown in blue (n = 3 biological replicates) (A,B). NX-2127 reduced BTK levels and appeared to reduce the AAK1-annotated protein group (A0A096LP25), whereas NX-5948 reduced BTK alone (A,B). Western blotting confirmed that AAK1 expression levels were not affected by NX-2127 (supplemental Figure 2). Neither PROTAC had a significant effect on the levels of TEC (C). The bar graph represents quantification of the western blots, expressed as a percentage of the vehicle control (mean ± SEM; n = 3). NX-2127 degraded BTK by 94.3%, whereas NX-5948 degraded BTK by 98.3% (C). FC, fold change; NS, non-significant; Veh, vehicle.

• 

  Figure 3.. Effect of PROTAC mechanistic inhibitors. (A-B) Human platelets in PRP were preincubated with either vehicle (DMSO), 10μM pomalidomide, or 25μM MLN4924 for 1 hour at 30°C. After preincubation, platelets were further treated with either 3μM NX-2127 (A) or 100nM NX-5948 (B) for 4 hours at 30°C. The graph represents quantification of the western blots, expressed as a percentage of the vehicle control (mean ± SEM; n = 4). Representative blots are shown above the graph. Both PROTACs function as anticipated based on their mechanism of action. Pomalidomide, an IMiD-class CRBN binder, prevents BTK degradation by competing with both NX-2127 and NX-5948 for binding to the E3 ligase cereblon. MLN4924, a neddylation inhibitor of Cullin-RING ligases, blocks BTK degradation by inhibiting the transfer of ubiquitin, thereby preventing proteasomal targeting. ∗P < .05, ∗∗P < .01, ∗∗∗P < .001, and ∗∗∗∗P < .0001. CRBN, cereblon; IMiD, immunomodulatory drug; Pom, pomalidomide.

• 

  Figure 4.. BTK degradation impairs CRP-mediated, but not TRAP-6–mediated, platelet function. (A-C) Human platelets in PRP were treated with 3.0μM NX-2127 and 100nM NX-5948 for 4 hours at 30°C. Ca2+ mobilization was assessed in Fura-2–loaded platelets in response to various concentrations of TRAP-6 and CRP. (Data were expressed as the 340/380 nm fluorescence ratio) (A). Platelet activation was assessed by stimulation with either TRAP-6 or CRP, then measuring integrin αIIbβ3 activation and P-selectin expression by flow cytometry (B). Data are expressed as a percentage of vehicle control (mean ± SEM; n = 5). There were no significant changes in platelet activation in NX-2127– or NX-5948–treated platelets after stimulation with TRAP-6. However, CRP-mediated platelet activation was significantly reduced by both NX-2127 and NX-5948 treatment, with NX-5948 having a stronger effect (B). Platelet aggregation (optical aggregometry) was unaffected in platelets stimulated by TRAP-6 after NX-2127 and NX-5948 treatment. However, BTK degradation by NX-2127 and NX-5948 severely affected CRP-mediated platelet aggregation (C). Data are presented as percentage of maximum response to PPP (mean ± SEM; n = 5).…

• 

  Figure 5.. Effect of BTK degradation on platelet spreading and thrombus formation. (A-B) Human platelets in PRP were treated with 3.0μM NX-2127 and 100nM NX-5948 for 4 hours at 30°C. Platelets were washed and added to surfaces coated with 10 μg/mL collagen (A) or 20 μg/mL fibrinogen (B) and allowed to adhere for 75 minutes. Scale bar, 10 μm. Platelets were stained with ActinGreen and visualized by confocal microscopy. Five images were taken per coverslip for each condition, and platelet spreading stage (round/filopodia/lamellipodia) was analyzed using FIJI (ImageJ). The graphs show the lamellipodia-to-filopodia ratio for each condition. Bovine serum albumin (BSA) represents spreading on BSA-coated coverslips (negative control), and “collagen/fibrinogen (untreated)” indicates platelets spread on ECM without DMSO (positive control). mean ± SEM, n = 6 (A); mean ± SEM, n = 7 (B). (C) Human platelets in PRP were treated with 100nM NX-5948 for 4 hours at 30°C. Platelets were stained with DiOC6 and recombined with the red blood cell layer, and whole blood was passed over a 50 μg/mL collagen-coated surface under flow (3.07 mL/h [shear rate, 1000 s˗1] for 5 minutes). Representative images are shown…

• 

  Figure 6.. NX-5948 impairs ex vivo CRP-mediated platelet function in mice. (A-E) Mice were treated with 30 mg/kg of NX-5948 by oral gavage for 24 hours. Platelets were isolated and washed. Btk expression was measured by western blotting (A). The graph represents quantification of the western blots, expressed as a percentage of the vehicle control (mean ± SEM; n = 5). Btk levels in mouse platelets were significantly reduced by over 97% (unpaired t test with Welch’s correction; P = .0003) (B). NX-5948 treatment also led to a significant reduction of BTK in both B cells and megakaryocytes. There was a 99% reduction in B cells (unpaired t test with Welch’s correction; P = .0123) (C) and ∼95% in megakaryocytes (unpaired t test with Welch’s correction; P = .0172) (D). In platelets from mice treated with NX-5948, CRP-mediated platelet activation was significantly reduced when measured by both αIIbβ3 integrin activation and P-selectin expression (CRP and thrombin used at 5 μg/mL and 1 U/mL, respectively). Thrombin-mediated platelet activation was unaffected. Data are expressed as percentage of vehicle control (mean ± SEM; n = 5) (B). ∗P < .05, ∗∗P < .01, ∗∗∗P < .001, and ∗∗∗∗P < .0001.

• 

  Figure 7.. Impaired arterial thrombosis in NX-5948–treated mice. (A-B) Mice were treated for 24 hours with either vehicle (DMSO) or 30mg/kg NX-5948 by oral gavage. Mice were anaesthetized using ketamine. Ferric chloride was applied to the carotid artery of mice to inflict vascular injury and induce thrombus formation (12% for 3 minutes) (A). Thrombus formation was imaged by intravital microscopy in real time. The data are expressed as mean integrated fluorescence density of the thrombus (mean ± SEM; n = 6), indicative of thrombus size and density (A). Tail bleeding was assessed after a tail incision. The data are expressed as bleeding time in seconds (mean ± SEM; n = 9). There is no significant difference between the bleeding time of vehicle- and NX-5948–treated mice (B). ∗P < .05, ∗∗P < .01, ∗∗∗P < .001, and ∗∗∗∗P < .0001. AU, arbitrary units; ns, nonsignificant.