Fueling platelets: Where does the glucose come from?

There has long been interest in the metabolic pathways occurring in platelets and their relationship to function. Despite their reliance on crude analytical tools and inhibitors, early studies showed that platelets must produce ATP to maintain functionality¹. Platelet activation leads to a significant increase in glucose uptake, glycolysis, and lactate production^2–5, suggesting that, in addition to oxidative phosphorylation, glycolysis is a significant producer of ATP to fuel platelet function. Electron microscopy studies clearly show that platelets contain glycogen granules, which could serve as a reservoir for energy production. Glucose can also be imported from the outside through three classes of facilitative transporters (Glucose Transporters, (GLUTs); Sodium-Glucose Transporters, (SGLTs); and Sugars Will Eventually be Exported Transporters, (SWEETs))^{6, 7}. These proteins facilitate the transmembrane movement of glucose, down a concentration gradient of either glucose or sodium⁶. Platelets contain at least two members of the GLUT family of uniporters, GLUT1 and GLUT3^{8, 9}. While GLUT1 may account for basal uptake of glucose in resting cells, GLUT3 has drawn attention because it is mobilized from α-granules to the plasma membrane (PM) upon platelet activation (Figure 1)^{4, 5}. In this issue, Fidler et al.⁹ used a platelet-specific knockout of GLUT3 to probe its role, and thus the roles of regulated glucose uptake, in platelet function.

Figure 1. Mobilization of GLUT3 in Activated Platelets.

While GLUT1 resides on the plasma membrane (PM), GLUT3 is present mainly in the α-granules of resting platelets and is mobilized to the PM upon activation. There, GLUT3 promotes increased glucose influx which fuels the generation of ATP, and thus platelet function, through oxidative phosphorylation in the mitochondria and glycolysis in the cytosol. Given the increase in glucose, much of the pyruvate produced by glycolysis in activated platelets appears to end up in lactate production.

GLUT3 was originally posited as a neuronal-specific glucose transporter that maintained cytoplasmic levels of the neuron’s preferred fuel¹⁰. Subsequently, it was detected on sperm and studies of GLUT3 knockout mice showed it is globally required for embryogenesis (reviewed in ¹¹). GLUT3 is unique, having the lowest K_m and highest V_max of the four major GLUTs (1–4); thus, it has a high transport capacity even at lower glucose concentrations¹². Localized to the PM in neurons and sperm, GLUT3 exists in an intracellular pool in certain blood cells, e.g., lymphocytes, macrophages, neutrophils, and platelets (reviewed in ¹¹). It is from this intracellular pool, that it is mobilized to the PM upon activation^{4, 5}, much like GLUT4 is mobilized in adipose and muscle cells upon insulin treatment.

Fidler et al.⁹ created mice lacking GLUT3 in platelets and megakaryocytes. Megakaryocytes from these mice showed no overt defects and platelet numbers were unaffected, suggesting that GLUT3 was not required for platelet production. While the mice had no defect in the tail bleeding assay of hemostasis, there were mild, but significant defects, in the autoimmune/inflammatory arthritis model and some protection in the collagen/epinephrine pulmonary embolism model. In vitro, GLUT3^−/− platelets produced normal microparticles in response to thrombin and collagen but were defective in spreading and clot retraction. Aggregation was also defective, but only at low agonist concentrations and the defects were resolved when agonist was increased. These data show a role for GLUT3 in platelet function but suggest that the transporter is not absolutely required since no assays showed a complete ablation of function.

Secretion from GLUT3^−/− platelets was only partially affected. ATP release and exposure of CD63 were unaffected but the release of PF4, IL1α, and exposure of CD62 were, suggesting that release of dense granule cargo did not require GLUT3 but that α-granule content release did. Since dense granules are thought to be “docked” to the PM, they may not require extra ATP generation for membrane fusion¹³. Ultrastructural analysis of α-granules showed that activated GLUT3^−/− platelets had more decondensed granules relative to wild-type. Thus, membrane fusion occurred but the cargo was not effectively dissolving and/or being extruded from the granules. This suggests that GLUT3 supplies the fuel for fusion pore expansion¹⁴ and/or an active extrusion of granule content. Alternatively, GLUT3 may transport polyol glucose into the concentrated granule cores to promote solubilization of α-granule contents. Future studies will be needed to address these points.

Secretion from the GLUT3^−/− platelets had a distinctive time course that suggested two sources of glucose may fuel different stages of α-granule secretion. Consistently, glycogen levels rapidly decreased at the initial time points, suggesting that glycogenolysis was important in the early stages of activation. To address this, Fidler et al.⁹ blocked glycogen phosphorylase and slowed the rate of secretion but did not affect its final extent in wild-type mouse and human platelets. Similar treatment of GLUT3^−/− platelets decreased both the initial rate and extent of P-selectin exposure. These data imply a role for glycogenolysis in the early stages of platelet activation and glucose influx, via GLUT3, at later stages. While these data show that differential mobilization of fuel sources is important for platelet function, a simple interpretation is complicated by the fact that GLUT3^−/− platelets had lower resting levels of glycogen.

Given that GLUT3 transports glucose bidirectionally in response to a glucose gradient, does GLUT3 on α-granules have a role in resting platelets? Heijnen and colleagues showed that ~85% of GLUT3 was on α-granules in resting cells⁴. While α-granules could simply be the depot for GLUT3, Fidler et al.⁹ suggest GLUT3 might facilitate glycolysis and ATP generation in α-granules (Figure 1). Consistent with this concept, proteomics studies have detected glycolytic enzymes in platelet releasates¹⁵. Fidler et al.⁹ show some association of hexokinase and lactate dehydrogenase with a α-granule fraction, though the activity of LDH in that fraction was low. Perhaps more compelling, the production of [¹³C] lactate in saponin-permeabilized, cytosol-depleted platelets decreased by 2.5 fold in the GLUT3^−/− platelets when fed [¹³C] glucose. However, due to difficulties in isolating intact α-granules, a true “smoking gun” demonstration that α-granules catalyze glycolysis is lacking. Future experiments in this area will assuredly be exciting. Does ATP generation truly occur inside α-granules and is that ATP contributing to platelet energetics? Alternatively, is that pool of ATP fueling intragranular phosphorylation or perhaps ecto-phosphorylation reactions outside the platelet¹⁶? However, given the ATP released from dense granules, it is unclear how much more α-granules can contribute to the microenvironment around an activated platelet.

How does GLUT3 affect platelet function? Based on the data presented in Fidler et al.⁹, the answer may be subtle. GLUT3 does account for the stimulation-dependent influx of glucose seen in activated platelets, confirming previous work^{4, 5}. This glucose influx is used to fuel spreading, clot retraction, and the sustained phase of α-granule secretion, but perhaps not its initial steps. The enzymatic properties of GLUT3 make it ideal for efficiently transporting glucose when in a thrombus core where diffusion is restricted¹⁷. Thus, its mobilization to the surface offers a mechanism to fuel clot retraction and other late stages of thrombosis. In a broader sense, Fidler et al.⁹ expand a new chapter in platelet energetics where the tools of metabolomics and genetically-modified mice are used to ask specific questions about how platelets power thrombosis and hemostasis. Such future studies will yield important insights into normal platelet function and perhaps an understanding of the dysfunctions associated with metabolic pathologies such as diabetes.