Initial experiments confirmed that RBC effectively
Initial experiments confirmed that RBC8 effectively inhibited both RalA and RalB activation in an identical, dose-dependent manner following platelet stimulation with the GPVI-specific ligand, CRP (Fig. 1Ai). Non-specific, upper bands were observed when immunoblotting for activated RalB, with the specific ‘GTP’ signal denoted by the arrow (Fig. 1 Ai). The half-maximum inhibitory value (IC50) of RBC8 for RalA and RalB was 2.2 μM and 2.3 μM, respectively (Fig. 1 Aii), which is relatively similar to reported IC50 values of 3.5 and 3.4 μM in H2122 and H358 cells, respectively . Having confirmed the inhibitory effect of RBC8 on Ral activity, subsequent experiments set out to assess the effects of RBC8 treatment on platelet functional responses. We specifically chose a threshold concentration of CRP (0.6 μg/mL) as we had previously observed a relatively weak, but statistically significant reduction in dense granule secretion (ATP release), but not aggregation, using this concentration in Ral DKO mouse platelets . With this, we observed a dose-dependent inhibitory effect of RBC8 on human platelet aggregation (Fig. 1B), with a concomitant decrease in dense granule secretion (Fig. 1C). Secretion of ADP from platelet dense granules is an important autocrine/paracrine signalling mediator of GPVI platelet responses and we therefore used ADP “rescue” experiments with exogenously added ADP (10 μM) to understand the mechanism through which RBC8 inhibits human platelet aggregation [33,34]. Notably, exogenous ADP fully recovered the aggregation defect in the 1 and 3 μM RBC8-treated platelet samples, but not completely in 10 μM RBC8 samples (Fig. 1B). This suggests that RBC8, particularly within the IC50 range (1–3 μM) reduces ADP secretion necessary for full aggregation responses, but at the 10 μM dose there is an ADP-independent component to RBC8-mediated reduction in platelet aggregation. Previously, we established that genetic loteprednol etabonate of Rals in mouse platelets causes a substantial reduction in P-selectin surface exposure, without a significant change in integrin αIIbβ3 activation . Using the same flow cytometry assays, we investigated the effect of RBC8 on human platelet responses. Here, RBC8 significantly decreased both readouts of activation in human platelets and these reductions were not agonist-dependent as significant decreases were observed in response to both CRP and PAR4-AP (Fig. 1D and E). While the reduction in P-selectin exposure with RBC8 is consistent with responses in mouse Ral DKO platelets, the decrease in integrin activation with RBC8 was a noticeable divergence in functional responses between RBC8-treated human platelets and mouse Ral DKO platelets. Similarly, we observed a pronounced defect in thrombus formation in vitro in RBC8-treated whole human blood perfused over a collagen-coated surface; a defect which was not apparent in whole blood from Ral DKO mice (Fig. 1F). However, considering the effect of RBC8 on human platelet aggregation, granule secretion and integrin activation, it was not entirely unexpected to observe defective platelet thrombus formation in RBC8-treated whole blood . This finding does also support the efficacy of using RBC8 in native environments such as whole blood, consistent with the seminal paper by Yan et al. reporting decreases in tumor growth from in vivo studies . Further experiments in RBC8-treated human platelets assessed the soluble release of the α-granule marker, PF4, Ca2+ mobilisation and phosphatidylserine (PS) exposure (Fig. 1G–I). Here, platelet responses following RBC8 treatment generally showed no effect compared with vehicle/DMSO-treated platelets. The lack of defect in PF4 secretion with RBC8 treatment is important, and aligned with our observations in Ral DKO mouse platelets that show a major defect in P selectin expression with no defect in PF4 release . Furthermore, the absence of altered Ca2+ signalling with RBC8 is consistent with previous reports demonstrating that Ral activity is downstream of Ca2+ signalling, as an increase in cytosolic Ca2+, either due to release from intracellular stores and/or cellular influx, is essential for Ral activation . These rises in cytosolic Ca2+ are also important for platelet procoagulant function, as measured by annexin V binding to exposed PS, and therefore the absence of altered PS responses with RBC8 is also unsurprising . Importantly, RBC8 did not alter basal/unstimulated annexin V binding values in unstimulated platelets, confirming that the compound (between 1 and 10 μM) does not non-specifically induce apoptosis in resting platelets (Fig. 1I).