Insulin Biosynthesis in a Broken-Cell Preparation of Islets of Langerhans

There have been numerous attempts to study insulin biosynthesis by perfused pancreas slices and isolated islets of Langerhans, but so far insulin synthesis has rarely been successfully carried out with broken-cell preparation of islets or other subcellular fractions. Nevertheless, Wagle (1 965) showed that a pH 5 enzyme-microsomal fraction obtained from foetal dog pancreas incorporated *4C-labelled amino acids into insulin, which could be subsequently precipitated by an anti-insulin serum. However, in these experiments the labelled insulin was not further characterized. In the experiments described below with rabbit islets of Langerhans, labelled proinsulin was obtained in a broken-cell preparation in the presence of a ribonuclease inhibitor. Islets were prepared from male New Zealand White rabbits (2.5 kg) by collagenase digestion (Howell & Taylor, 1968). For each determination 100 islets were preincubated at 37°C for 20min in Gey & Gey bicarbonate-buffered medium, pH7.4, containing 2m~-glucose. After gentle centrifugation the medium was removed from the islets and replaced with 0.2ml of a maintenance medium containing ~ O ~ M K C I , 25m~-Tris, 4m~-magnesium acetate, ~ ~ M A T P , 1 PM-GTP, 3.3,u~-creatine phosphate, 1 unit of creatine phosphokinase and amino acids at a concentration of 1 PM for each of the amino acids present in porcine proinsulin but minus leucine. A ribonuclease inhibitor was added to the medium at a concentration of 2 Searle units/ml. The islets were then subjected to ultrasonic disruption for 2s (this was found to be sufficient time to disrupt the cell membrane in this preparation). Then 10,uCi of ~-[~H]leucine and either 2mM-gh1COSe, 16m~-glucose or 16 m~-glucose and cycloheximide (250,ug/ml) were added to the islet preparation. An incubation for 1 h was carried out at 37°C and stopped by the addition of acetic acid (pH3). The samples were then applied to a Sephadex G-50 (fine grade) column (1 cmx 55cm) equilibrated with 1 M-acetic acid. Peaks of radioactivity appearing in the proinsulin region of the elution profile were taken and freeze-dried in plastic tubes in the presence of 500,ug of albumin. The dried preparation was then taken up in 0.95 NaCI, and the proinsulin was purified by a double-antibody precipitation method previously described (Parry & Taylor, 1974). After extraction with acid ethanol the proinsulin was subjected to electrophoresis on paper at pH2. At this stage a sample was taken for assay purposes in an insulin assay system (Hales &Randle, 1963). Radioactivity on the electrophoretogram was assayed by cutting 1 cm strips and counting their radioactivity in a liquid-scintillation spectrometer. The material was further characterized as proinsulin by gel electrophoresis at pH8.9 and paper electrophoresis at pH8.6. Incorporation of ~-[~H]leucine into rabbit proinsulin was significantly increased when the glucose concentration was raised from 2 m ~ to 1 6 m ~ . Further, the presence of cycloheximide significantly lowered the rate of incorporation into proinsulin. The results shown in the present communication suggest that cellular integrity of the 8-cell is not essential for proinsulin synthesis to occur, provided that a ribonuclease inhibitor is present. Under conditions where glucose is present in the medium only proinsulin is produced, indicating that when cellular organization has been disrupted the conversion process to insulin can only occur at a specific locus within the fi-cell, probably in the Golgi region (Howell et al., 1969). Raising the glucose concentration of the medium from 2 to 1 6 m ~ increased leucine incorporation 3-fold, which compares favourably with values obtained for whole islets. A further indication that labelled proinsulin was being synthesized in a broken-cell preparation was obtained when it was shown that cycloheximide abolished the stimulatory effect of 16m~-glucose on incorporation. Theseexperiments suggest that glucose transport isunlikely to be involved in triggering the biosynthetic response to glucose in islets. Equally well, it seems unlikely that a mem-