Walter S. Zawalich, Marc Bonnet-Eymard, and Kathleen C. Zawalich

Yale University School of Nursing, 100 Church Street South, New Haven, CT 06536-0740 USA

Received 2/24/97; Accepted 2/27/97; On-line 3/15/97

4. Glucose -induced insulin secretion: Role of PLC activation

4.1 Effects of glucose on PLC-mediated phosphoinositide (PI) hydrolysis

The ability of glucose stimulation to increase insulin secretion has been the subject of intense experimental analysis. In large part, this is due to the realization that the emergence of NIDDM is due to ß-cell decompensation and their inability to respond to glucose stimulation (30, 38). Glucose must be metabolized to augment secretion and the consensus is that the increase in ATP levels or the ATP/ADP ratio results in the closure of ATP-sensitive K⁺ channels, cell depolarization and the influx into the ß-cell of calcium via voltage-regulated Ca²⁺ channels (1, 39). While glucose-dependent insulin secretion clearly depends on extracellular calcium influx into the ß-cell (40-42), this event alone is insufficient to explain the large and rising biphasic insulin secretory response which occurs in response to the hexose. For example, ß-cell depolarization with high levels of K⁺ and the ensuing sustained increase in intracellular calcium results in a large first phase insulin secretory response with little sustained stimulatory effect on release (43-45). Other signals dependent on glucose metabolism must be generated and, since the mitochondrial fuels and insulin secretagogues alpha-ketoisocaproate and monomethylsuccinate duplicate many of the actions of glucose (15, 20), a mitochondrial signal in addition to calcium has been postulated (46). Moreover, mitochondrial signals and calcium are also essential for the full activation of PLC by glucose (45) leading to the conclusion that the activation of PLC may play an important role in the regulation of glucose-induced insulin secretion.

A variety of techniques have been employed to assess the impact of glucose stimulation on PLC-mediated phosphoinositide hydrolysis. Most investigators incubate islets, usually isolated from rats, in myo-[2-³H]-inositol for various lengths of time. This label is exclusively incorporated into the family of phosphoinositide lipids. After washing to remove unincorporated label, the islets are then stimulated with various agonists and the accumulation of labeled inositol phosphates (IPs), derived from phosphoinositide precursors, serves as the index of PLC activation (Figure 2) (47, 48). Alternatively islets can be perifused after labeling and the efflux of free [2-³H]-inositol measured in the perifusate along with insulin (14, 42, 49). Similar results have been obtained with both approaches and the conclusion reached is that glucose activates PLC, a response dependent upon both glucose metabolism (16) and calcium influx (45). The precise identity of the metabolic signal has remained elusive but it can be generated by both glycolytic and mitochondrial fuels (50, 51).

Figure 2. Dose-response Effects of Glucose Stimulation on Phosphoinositide Hydrolysis. Groups of rat (closed circles) or mouse (open circles) islets were incubated for three hr in ³H-inositol, washed to remove unincorporated label and then stimulated with various levels of glucose. The inositol phosphates (IP) accumulating during a 20 minute (rat) or a 30 minute (mouse) stimulatory period with the hexose are shown here plotted against the glucose level. Also indicated is the IP response of mouse islets (open circle) stimulated with the combination of 15mM glucose plus 100µM carbachol demonstrating that with the proper agonist a significant IP response can be generated.

4.2 PLC isozymes in islets and their activation

An additional level of complexity has been added by nature to the activation of PLC. This has to do with the recent demonstration that rat islets (52) like many other tissues (53) express more than one type of PLC. ß-cells contain the three major isozymes of PLC (ß1, gamma1 and delta1) and the available evidence suggests that different classes of insulin secretagogues activate distinct isozymes of PLC (52, 54, 55). For example, nutrient secretagogues activate the enzyme by increasing intracellular calcium and generating an as yet to be identified mitochondrial signal (41, 45). Evidence for this concept derives from the fact that glucose must be metabolized to increase PI hydrolysis, that its effects on PLC activation can be duplicated by both glycolytic and mitochondrial substrates and that the full activation of PLC requires both metabolism and calcium influx into the ß-cell. While calcium contributes to the activation of a fuel-regulated PLC, sustained increases in intracellular calcium induced by depolarizing levels of potassium result in only small increments in insulin secretion and IP accumulation (45). Furthermore blocking calcium influx reduces but does not abolish glucose-induced IP accumulation (41). The simplest interpretation of these and other data is that at least two signals--calcium and metabolically-derived factor- -are involved in the activation of the nutrient-regulated PLC.

The concept that neurohumoral agonists such as acetylcholine or its nonhydrolyzable analogue carbachol activate an isozyme of PLC different from that activated by nutrients is supported by both the calcium and metabolic independence of neurohumoral agonists on IP accumulation (47, 56, 57). Most importantly maximal stimulatory levels of glucose and these neurohumoral agonists interact in at least an additive fashion to increase IP accumulation (54, 57). However, perhaps the most cogent data supporting the concept that different isozymes of PLC regulate the responses to fuel and neurohumoral agonists on PLC-mediated PI hydrolysis and insulin secretion comes from studies where the responses of rat islets were compared to those from mouse islets, the two most commonly employed species used to unravel the complex maze of signaling events which regulate insulin secretion (52). Most importantly the divergent effects of glucose on the activation of PLC in these species was paralleled by a marked divergence in the insulin secretory response (Figure 1) to the hexose as well (See below, Section 6).