Does glyceraldehyde enter pancreatic islet metabolism via both the triokinase and the glyceraldehyde phosphate dehydrogenase reactions?: A study of these enzymes in islets

Abstract

Glyceraldehyde has been known to be an insulin secretagogue for more than 15 years. It has been (reasonably) assumed that glyceraldehyde enters the glycolytic pathway via its phosphorylation by ATP to form glyceraldehyde phosphate, a reaction catalyzed by the enzyme triokinase, and that subsequent metabolism is identical to that of glucose. However, up to now there have been no studies verifying the presence of triokinase in the pancreatic beta cell. We report here that (1) the activity of triokinase in pancreatic islets is very low, indicating that the activity is intrinsically low and/or the enzyme was rapidly inactivated during the preparation of tissue for assay; (2) the activity is much lower than glucose phosphorylating activity (hexokinase plus glucokinase) in islets, even though glyceraldehyde is a more efficient insulin secretagogue than glucose; (3) glyceraldehyde phosphate dehydrogenase from pancreatic islets can use glyceraldehyde as a substrate in place of glyceraldehyde phosphate (the V max of glyceraldehyde phosphate dehydrogenase from islets when glyceraldehyde is the substrate is 20-fold that of triokinase when glyceraldehyde is the substrate); and (4) the K m of glyceraldehyde phosphate dehydrogenase with respect to glyceraldehyde (4.8 m m) is similar to the concentration of glyceraldehyde that gives one-half maximal rates of insulin release from pancreatic islets, whereas the K m of triokinase with respect to glyceraldehyde is much lower (< 50 μ m). These data suggest that besides stimulating insulin release in islets via its entering metabolism by phosphorylation to glyceraldehyde phosphate in the triokinase reaction, glyceraldehyde could be phosphorylated by P i in the glyceraldehyde phosphate dehydrogenase reaction to form glycerate 1-phosphate which is probably unmetabolizable in islets. The second reaction could drastically increase the NADH/NAD ratio in islets without providing substrates for hydrogen shuttles that reoxidize cytosolic NADH. Since an increased NAD(P)H/NAD(P) ratio is believed to be a key part of the signal for insulin release, such a mechanism would explain the potent insulinotropism of glyceraldehyde in short-term experiments. In addition, the formation of unmetabolizable acids may explain the toxic effects of long-term exposure of islets to glyceraldehyde and why glyceraldehyde causes the beta cell to become acidic, whereas glucose does not.