In answer to the comments of Islam and Berggren concerning our hypothesis on the CD38-cyclic ADPribose (cADPR) signalling system, we will present several lines of evidence that we believe can explain the discrepancies between their view and ours.
The Okamoto model and cADPRGlucose is the primary stimulus of insulin secretion and synthesis in pancreatic beta cells of the islets of Langerhans [1±3]. Increases in the intracellular Ca 2+ concentration mediate the biochemical events that couple glucose stimulation to insulin secretion, and mobilization of Ca 2+ from intracellular stores in the endoplasmic reticulum as well as Ca 2+ influx from extracellular sources are important in this process [4]. Concerning the mechanism of Ca 2+ influx from extracellular sources, it has been proposed that ATP generated in the process of glucose metabolism inhibits the ATP-sensitive K + channel, causing beta cell membrane depolarization, thereby opening the voltagedependent Ca 2+ channel and resulting in Ca
2+-influx from the extracellular space [5]. However, it was thought that inositol 1, 4, 5-trisphosphate (IP 3 ) is a second messenger for Ca 2+ mobilization from intracellular stores [6]. Since 1981, we have proposed a model for beta cell damage and its prevention as shown in Figure 1; that is, diabetogenic agents such as streptozotocin and alloxan damage DNA, activating poly(ADP-ribose) synthetase which uses NAD + as a substrate [2, 7±12]. Consequently, intracellular levels of NAD + fall dramatically, causing the inhibition of cellular functions such as insulin synthesis and secretion, and the beta cell ultimately dies. Although insulin-dependent diabetes mellitus (IDDM) can be caused by many different agents such as immunologic abnormalities, inflammatory tissue damage and betacytotoxic chemical substances, the final pathway for the toxic agents is the same, as shown in Figure 1. Therefore, IDDM is theoretically preventable by suppressing immune reactions, scavenging free radicals, and inhibiting the poly (ADP-ribose) synthetase by nicotinamide and 3-aminobenzamide. The question then arises as to why maintaining the cellular NAD + level is essential for beta cells to secrete and synthesize insulin. Since 1987, a metabolite of NAD + , cADPR, has been reported to be as potent and powerful a releaser of intracellular Ca 2+ as IP 3 in a variety of cells [13±29]. In 1993, we found that the cADPR level increased in pancreatic islets in response to glucose stimulation and that cADPR released Ca 2+ from the microsomes, suggesting that cADPR has a second messenger role in insulin secretion [15]. Furthermore, we found that the cADPR level was not increased by glucose stimulation in pancreatic islets pretreated with streptozotocin [15]. When poly(ADP-ribose) synthetase inhibitors such as nicotinamide and 3-aminobenzamide were present, the cADPR level increased with glucose stimulation even in the presence of streptozotocin. These results suggest that cADPR is synthesized from NAD + by glucose stimulation in beta cells [12,15]. We have t...