Johanna Westerlund

Pulsatile insulin release from pancreatic islets with nonoscillatory elevation of cytoplasmic Ca2

The relationship between insulin release and cytoplasmic Ca2+ concentration ([Ca2+]i) was studied in isolated pancreatic islets from ob/ob mice. Although [Ca2+]i was low and stable in the presence of 3 mM glucose, basal insulin release exhibited low amplitude pulsatility, with a frequency of 0.32 +/- 0.04 min-1. Depolarization by raising K+ from 5.9 to 30.9 mM or by the addition of 1 mM tolbutamide caused a pronounced initial insulin pulse followed by declining pulses, but there was no change in frequency. This decline in amplitude of the insulin pulses was prevented in similar experiments performed in the presence of 11 mM glucose. Corresponding measurements of [Ca2+]i in islets exposed to tolbutamide or the high K+ concentration revealed stable elevations without oscillations. Although the [Ca2+]i level is an important determinant for the rate of secretion, the results indicate that pulsatile insulin release does not always depend on [Ca2+]i oscillations. It is suggested that cyclic generation of ATP may fuel pulsatile release under conditions when [Ca2+]i remains stable.

Pulsatile insulin release from mouse islets occurs in the absence of stimulated entry of Ca2

Pancreatic islets are known to respond to a raise of the glucose concentration with Ca2+ -induced 2-3-min pulses of insulin release. The reports of cyclic variations of circulating insulin in the fasting state made it important to explore whether insulin release is also pulsatile in the absence of stimulated entry of Ca2+. Individual pancreatic islets were isolated from a local colony of ob/ob mice and perifused under conditions allowing dual wavelength recordings of the cytoplasmic Ca2+ concentration ([Ca2+]i) with fura-2 and measurements of insulin with ELISA technique. At 3 mM of glucose, [Ca2+]i remained at a stable low level, but insulin was released in pulses with a frequency of 0.41+/-0.02 min-1, determined by Fourier transformation of original and autocorrelated data. Pulses of basal insulin release were also seen when glucose was omitted and 1 microM clonidine or 400 microM diazoxide was added to a glucose-free medium. The results indicate that pulsatile insulin release can be generated in the absence of stimulated entry of Ca2+. A tentative explanation for this phenomenon is inherent fluctuations in the ATP production of the beta cells.

Signaling Underlying Pulsatile Insulin Secretion

Regular oscillations of the circulating insulin concentrations were discovered in the monkey [28] and subsequently found in normal human subjects [50]. The characteristic insulin pattern is deteriorated in patients with type 2 diabetes [49] as well as in their close relatives [6 11. Studies in non-diabetic subjects with suppressed endogenous insulin secretion and diabetic patients have indicated that less insulin is required to maintain normoglycaemia if the hormone is infused in a pulsatile manner compared to a constant rate [12, 57, 59, 63, 641. This difference is probably explained by higher expression of insulin receptors, when insulin is delivered in pulses [27]. It is easy to envision a scenario for the development of type 2 diabetes in which deteriorated oscillations leads to insulin resistance with a compensating hypersecretion of the hormone. In susceptible individuals the increased insulin demand may eventually exhaust the pancreatic p-cells with resulting development of overt diabetes. What is then the origin of the regular insulin oscillations? One possibility is that they result from a negative feedback loop between the liver and the pancreatic pcell [50]. However, later studies have indicated that the oscillations occur independent of changes in plasma glucose, reflecting a pacemaker function in the pancreas [49, 581. This conclusion is consistent with measurements of secretion from the isolated perfused dog pancreas [76]. Another fundamental aspect is the frequency of the insulin oscillations. Whereas the early studies on humans and monkeys indicated a periodicity of 10-15 min [28, 501, measurements in the dog showed 4-8 min. The latter estimate is similar to the periodicity observed from the perfused dog pancreas [75] and that based on blood sampling from the portal vein of dogs [66]. The portal insulin oscillations are very prominent, indicating that pulsatile secretion accounts for 70% of total secretion (Fig. 1). In the periphery the oscillations are less pronounced due to recirculation and the fact that the liver extracts almost 50% of the portal hormone [ 131. The report of a lower frequency of the insulin oscillations in the peripheral blood may simply reflect difficulties in detecting insulin peaks due to a low signal-to-noise ratio [66]. Indeed, measurements on portal blood from patients with liver cirrhosis indicated a periodicity of 4.1-6.5 min [77]. Mechanisms underlying pulsatile insulin release will now be discussed at different levels of integration, starting with isolated pancreatic p-cells.

OSCILLATORY SIGNALING AND INSULIN RELEASE IN HUMAN PANCREATIC BETA -CELLS EXPOSED TO STRONTIUM

Oscillatory signaling and insulin release were studied in isolated pancreatic islets and β-cells obtained from human cadaveric organ donors. Taking advantage of Sr2+ as an analog for Ca2+, it was possible to demonstrate glucose-induced rhythmic activity in individual β-cells identified by immunostaining. Glucose-induced slow oscillations of Sr2+ (frequency, 0.1–1.0/min) were sometimes seen at a sugar concentration as low as 3 mm. Addition of 20 nm glucagon resulted in a broadening of the oscillations or in their transformation into sustained elevation. Moreover, the presence of glucagon resulted in the appearance of short transients of Sr2+, which disappeared after exposure to the intracellular Ca2+-adenosine triphosphatase inhibitor thapsigargin. Digital image analyses indicated that slow oscillations can be synchronized among cells in small aggregates and intact islets. The rhythmic activity in the glucose-stimulated β-cell had its counterpart in pulsatile insulin release when single islets were perifus...