Studying the mechanism(s) of platelet hyperactivity in Diabetes Mellitus

Diabetes Mellitus (DM) is a rapidly growing major health problem worldwide. The main complication and cause of death with diabetes is artherothrombosis. Platelet hyperactivity has been thought to contribute to the increased arterial thrombotic events in diabetic individuals, however, there has been limited success in identifying specific mechanism(s) that cause or predict human diabetic platelet reactivity. The slow progress in understanding diabetic platelet function is largely due to variable confounding factors that can influence platelet responses in diabetic individuals. To avoid this issue, I began my study in Chapter III using a mouse model of type 1 diabetes and identified that diabetic platelets displayed a striking shear-dependent up-regulation in integrin αIIbβ3 adhesive function, leading to an enhanced (2.5-fold) platelet-fibrinogen interaction under flow. This phenotype was intrinsic to platelets, was not triggered by acute but prolonged hyperglycaemia in vivo. Surprisingly, soluble agonist induced integrin αIIbβ3 function remained unaltered in diabetic mouse platelets, resulting in normal integrin αIIbβ3 activation, platelet aggregation and spreading. These studies have identified for the first time that diabetic platelets are hyper-responsive to biomechanical forces, promoting integrin αIIbβ3-mediated platelet adhesion. The mechanism underlying the up-regulated integrin IIb3 adhesive function was then investigated in Chapter IV. Integrin IIb3 ligation by fibrinogen triggers outside-in signalling involving the activation of Src family kinases, PI 3-kinases and protein kinase C (PKC), leading to ADP release and/or thromboxane A2 (TXA2) generation, important to sustain integrin IIb3 activation. Diabetic platelet adhesion to fibrinogen was more sensitive (30-50%) to inhibition of TXA2 generation than ADP antagonism (20%), in contrast to non-diabetic platelets which were more sensitive to inhibition of ADP signalling (50%) relative to TXA2 inhibition (10%). A major role for Src kinases in integrin IIb3 outside-in signalling was demonstrated under flow, as blocking Src kinases inhibited platelet adhesion by ~ 60% in both diabetic and non-diabetic platelets. Interestingly, despite the previously proposed role of PKCβ in diabetic complications such as retinopathy, PKCβ inhibition resulted in mild inhibition of diabetic platelets (30%) relative to a much more potent inhibition of non-diabetic platelets (>90%). Surprisingly, diabetic platelets were more sensitive to PI 3-kinase inhibition (50%) relative to non-diabetic platelet adhesion (20%). The combination of inhibiting PI 3-kinase activity and TXA2 generation prevented greater than 90% diabetic platelet adhesion to fibrinogen under flow. These studies suggest a selective role for PI 3-kinase activation and TXA2 production in integrin IIb3 outside-in signalling in diabetic platelet adhesion under flow. The relevance of these findings from mouse studies was validated in Chapter V by performing similar studies on type 1 diabetic patients. Similar to mouse studies, human diabetic platelets also exhibited elevated (50%) adhesion to fibrinogen under flow conditions relative to non-diabetic platelets. Interestingly, this enhanced platelet adhesion occurred mainly in patients with poorly controlled hyperglycaemia (HbA1c >7.5%) and dyslipidemia (LDL >2.5mmol/l and TG >2.0mmol/l). Consistent with my finding in mouse studies, human diabetic platelets were also more sensitive to inhibition of TXA2 production (30-50%) relative to 10% in non-diabetic platelets, however inhibiting PI 3-kinase activity prevented 50-60% platelet adhesion in both diabetic and non-diabetic platelets. Combined inhibition of these two signaling processes more effectively inhibited diabetic human platelet adhesion to fibrinogen under flow (70%), compared with non-diabetic platelet adhesion (50%). These studies not only validated the use of the diabetic mouse model for human studies, but also highlighted this animal model as an extremely useful tool to dissect out the underlying mechanisms for diabetic platelet hyperactivity. In addition, they also highlight that both poorly controlled glycaemia and dyslipidaemia promote diabetic hyperadhesive function. Since hyperglycaemia and dyslipidaemia are associated with oxidative stress, multiple genetically manipulated mouse models that are known to regulate oxidative stress, including RAGE-/-, NOX-2-/-, GPx-1-/- and ApoE-/- mice, were used to investigate the role of oxidative stress in regulating diabetic platelet adhesion under flow. Strikingly, RAGE deficiency led to markedly reduced platelet adhesion in diabetic (95%) and non-diabetic (40%) mice compared with normal diabetic and normal non-diabetic controls, respectively. In contrast, the lack of the major ROS synthase NOX-2 did not have a significant impact on adhesion in either diabetic or non-diabetic platelets. The deletion of major intracellular antioxidant GPx-1 resulted in a dramatic further increase in platelet adhesion in both diabetic (4-fold) and non-diabetic platelets (2-fold) compared with normal non-diabetic controls. Interestingly, hyperlipidaemia due to ApoE deficiency enhanced diabetic and non-diabetic platelet adhesion by 3 and 1.3-fold, respectively, and was exacerbated 6-fold (diabetic ApoE-/-) and 3-fold (non-diabetic ApoE-/-) by a high-fat diet, relative to normal non-diabetic platelets. These studies demonstrated a major role for oxidative stress in promoting diabetic platelet adhesiveness under shear conditions, through RAGE and GPx-1. They also confirmed the findings from human studies in Chapter V that dyslipidaemia in combination with hyperglycaemia leads to more exaggerated platelet adhesion under flow. Overall, my studies in this thesis highlighted an important role for shear forces in diabetic platelet adhesive function. In addition, a selective sensitivity to anti-platelet agents such as aspirin and PI 3-kinase inhibitors was identified for diabetic platelets, revealing new insights into the antiplatelet therapy potentially tailored for diabetic patients in the future. Several key molecules involved in ROS regulation, including RAGE and GPx-1, have been demonstrated to be critical for mediating diabetic platelet adhesion, potentially identifying additional novel anti-platelet strategies. These studies also emphasize the importance of long term control of hyperglycaemia and dyslipidaemia on diabetic platelet hyperactivity. Together, these findings have provided a better understanding of diabetic platelet function that may lead to new approaches to reduce the risk of atherothrombosis in diabetic patients.