Voxelotor and Red Blood Cell Pyruvate Kinase Activator Affect Clot Strength in Sickle Cell Disease

Abstract

Introduction: Sickle cell disease (SCD) is a common genetic red blood cell (RBC) disorder that results from a single substitution of valine for glutamic acid in the β-globin gene, resulting in sickle hemoglobin (HbSS). Under hypoxia, HbS polymerizes and forms fibers within the RBCs, thus making the cells rigid, deformed, and sickle-shaped [1]. Deformed RBCs influence the compactness and dynamics of blood clots [2]. In addition, people with SCD are susceptible to thrombophilia [2,3,4]. Recently, there has been a heightened interest in ameliorating SCD using RBCs pyruvate kinase (PKR) activators, which increase ATP and decrease 2,3 DPG, resulting in increased oxygen affinity and improved RBCs deformability [5]. Voxelotor is also an oxygen affinity-modifying drug, recently FDA-approved. Voxelotor delays deoxy-HbSS formation, preventing in vitro HbSS polymerization and RBC sickling, thus improving RBC deformability [6]. The in vitro impact of PKR activators and Voxelotor on blood clotting in clinical samples from people with SCD has not been examined. Here we use a microfluidic dielectric sensor, termed ClotChip ® [7,8], to investigate the effects of Voxelotor and a PKR activator on the clotting time and clot strength of samples from people with SCD. Methods: Venous blood samples were collected in sodium-citrate from subjects with homozygous HbSS and healthy volunteers (HbAA) under an IRB-approved protocol. ClotChip ® microsensors were fabricated as previously described [7]. Blood samples were centrifuged at 200 g for 10 min. Plasma and buffy coats were removed and stored at 4 °Cfor 6 hours. Isolated, washed RBCs were then re-suspended in PBS at 20% hematocrit and then mixed with 67 mg/mL of Voxelotor (GBT440; Selleckchem) to a final concentration of 600 µM (in DMSO) or a PKR activator (PKR activator 3; MedChem Express) at 10 mM (in DMSO) and incubated at 37 °Cfor 6 hours. For controls, HbSS- or HbAA-containing RBCs at 20% hematocrit were mixed with PBS containing 0.5% v/v DMSO and incubated at 37 °C for 6 hours. After incubation, HbSS+Vox, HbSS+PKR, and control samples were centrifuged at 500 g for 10 min to remove PBS. Each sample's plasma was used to reconstitute RBCs at 20% hematocrit. CaCl 2 was added to the samples to induce coagulation, and the samples were injected into the ClotChip ® microfluidic sensors. An Agilent 4294A impedance analyzer was used to obtain the ClotChip ® readout curve defined as the temporal variation of the normalized real part of blood dielectric permittivity at 1 MHz (see Figure 1A). Based on our previous studies [7,8], the time to reach a permittivity peak (T peak) parameter was taken to indicate the clotting time, whereas the maximum change in permittivity after the peak (ΔƐ r,max) parameter was taken to indicate the clot strength. Data are reported as mean ± standard deviation (SD). Results: We analyzed the ClotChip ® T peak and ΔƐ r,max readout parameters for HbSS (n=11), HbSS+Vox (n=11), HbSS+PKR (n=11), and HbAA control (n=8) samples ( Figure 1B). The clot strength was ~50% higher in HbAA than HbSS (HbAA vs. HbSS: p= 0.001, Mann-Whitney test). However, when the SCD samples were treated with Voxelotor, the clot strength was observed to increase by ~10% (HbSS vs. HbSS+Vox, p=0.01, Paired t-test). Treatment with a PKR activator increased the clot strength by ~30% (HbSS vs. HbSS+PKR, p=0.003, Figure 1B). A significant difference was observed in the clotting time between HbAA vs. HbSS (p=0.001) but treatment with Voxelotor or a PKR activator did not change the clotting time for HbSS (HbSS vs. HbSS+Vox, p=0.43; HbSS vs. HbSS+ PKR, p=0.29) ( Figure 1B). (p values were determined using a non-parametric Mann-Whitney test and paired t-test). Discussion : We found that treatment of HbS-containing RBCs with a PKR activator or voxelotor improves clot strength but not clotting time. PKR activators improved the strength of the clot significantly more than Voxelotor. Patients with SCD had a faster clotting time than healthy participants, which agrees with earlier researchers on hypercoagulability in SCD [4]. These data suggest that HbS polymerization and RBC health affect clot characteristics but not clot kinetics in SCD. Further investigations may be required to relate RBC deformability, which is well-known to impact SCD pathophysiology, to blood clotting and venous thromboembolism.