Male CD rats weighing 90–120 grams (Charles River, Wilmington, Massachusett) were anesthetized with intraperitoneal (IP) amobarbital sodium 100 mg/kg and either if needed. A midline abdominal incision was made and the splenic portion of the pancreas mobilized by partially breaking mesenteric connections to the stomach, small bowel, and retroperitoneum. Pancreatic tissue was then removed according to the method of Folgia⁷⁾ by gently scrapping it away with cotton applicator being careful to leave major blood vessels intact. The portion excised was that bordered by the spleen and fundus of the stomach up to the junction with the antrum where the pancreas begins to narrow; it comprised 40–50% of total pancreatic weight (data not shown). Control rats (shams) underwent laparotomy and mobilization of the splenic lobe with gentle rubbing of it between the fingers. Postoperatively all rats received standard laboratory chow ab libitum.

Two weeks later, dexamethasone sodium phosphate 0.125 mg/kg (Elkins-Simm, Cherry Hill, New Jersey) or sterile saline was given daily intramusculary for 5 days (day 14–18 at 9 AM). Body weight was measured before the first and last injections; blood for plasma glucose and insulin measurements was obtained after tail snipping by collection 0.6 cc in a heparin-coated 6×50 mm glass tube which was kept on ice until centrifugation and storage of the plasma at −20°C. Twenty-four hours after the last injection (day 19), rats were studied with either an intraperitoneal glucose tolerance test or using the in vitro isolated perfused rat pancreas.

Following an overnight fast, blood for plasma glucose and insulin was obtained as described above, and then 20% glucose (2 grams/kg IP) was given with the sampling repeated 30 and 50 minutes later. The animals were then sacrificed by decapitation and their pancreases removed for quantitative morphometrics.

Following the IPGTT and sacrifice, pancreases were excised and cleared of fat and lympth nodes; shams were divided into two portions while Px were left whole. Each was weighed, placed in Bouin’s fixative overnight, and embedded in paraffin. Sections (5–7 μm) were then stained with hematoxylin, and the relative volume of islets determined using point-counting morphometries.⁸⁾ a minimum of 5000 points in 108 fields (systematically chosen throughout the tissue using the markings of the stage micrometer) were counted in each block with the relative islet volume being the number of intercepts over islet tissue divided by the number of intercepts over pancreatic tissue. Assuming that volume and mass are synonymous, the absolute mass of islet tissue was then calculated as the relative islet volume times pancreatic weight.

This technique has been described previously.⁹⁾ The rats were anesthetized with amobarbital sodium 100 mg/kg IP. The perfusate was a modified Krebs-Ringer bicarbonate buffer which contained 4% dextran (D-4751, Sigma, St. Louis, Missouri), 2mM calcium, 1.2 mM magnesium, and 0.2% bovine serum albumin fraction V (Sigma). It was bubbled for 20 minuted with 95% O₂ and 5% CO₂, and then 2.8 mM glucose was added, the pH was adjusted to 7.4, and it was stored in a reservoir maintained at 38°C by water bath; 10 mM arginine was added to perfusate in a second reservoir. The higher glucose concentration (16.7 mM) was produced with a sidearm syrings driven by syringe pump which added 0.3 cc to the usual flow rate of 3.0 cc/min. Prior to delivery, the perfusate passed through and artificial lung in order to insure adequate oxygenation.¹⁰⁾ Following cannulation of the aorta and portal vein, the body cavity was covered with gauze moistened with saline, and the rat was placed under a heat lamp with the temperature constantly monitored and maintained at 30–39°C. The baseline perfusate (2.8 mM glucose) was delivered for 20 minutes, and then 30 second samples were collected according to the protocol shown at the top of Figure 2 in chilled tubes containing 4 mg EDTA which were then kept on ice pending storage at −20°C.

Following study with the in vitro isolated perfused pancreas, pancreases were removed, cleared of lymph nodes, blotted, weighed, and stored at −20°C in acid ethanol. Later, on a single day, they were homogenized with an Ultra-Turrax SDT (Tekmar, Cincinnati, Ohio), diluted to a volume of 8 cc, and refrozen pending assay.

Plasma glucose was measured with a Beckman glucose Analyser II (Beckman, Brea, Califorina). Insulin concentration was determined by radioimmunoassay using charcoal separation¹¹⁾ and rat insulin standards (supplied by Eli Lilly, Indianapolis, Indians).

The data in the figures and tables are expressed as mean ± standard error. Statistical significance was determined using the unpaired two-tailed Student’s test.¹²⁾

There was no difference in body weight, plasma glucose, or plasma insulin values in Px and sham rats at the end of the saline treatment (Table 1). Similar results were also obtained at the beginning of the treatment period (day 14, data not shown).

Dexamethasone caused Px and sham rats to be lighter than their saline counterparts primarily because of failure to gain weight during the 5 day treatment period; the saline groups gained approximately 4 grams per day. Dexamethasone caused mild hyperglycemia only in Px. This was associated with a plasma insulin value that was lower than that in the shams although the difference did not quite reach statistical significance (2.70±0.23 versus 3.63±0.41 ng/ml, 2p<0.07).

Insulin contents are shown in Table 2. It was reduced 60% in Px-saline rats compared to shams given saline. Dexamethasone had no significant effect on either result.

At the end of the study (day 19), the Px remnant still weighed considerably less than the whole pancreas in shams (0.546 ± 0.066 versus 0.837 ± 0.030 g, 2p<0.005). In contrast, islet mass had returned to a value which was not statistically different from controls although in absolute terms it was still reduced about 20% (0.615 ± 0.65 versus 0.734 ± 0.042 mg).

Dexamethasone caused relative islet volume in both Px and shams to increase, but interpretation of this finding is made difficult by the slightly lowered pancreatic weight (Table 3). Nonetheless, absolute islet mass also rose 30% in both treated groups although the increase narrowly missed statistical significance in shams due in part to variability within the group (4 animals did not have values greater than the mean of controls while 3 were markedly increased). Variability was also present in Px with 3 animals given dexamethasone having raised islet mass while 2 did not. Morphologically, many markedly enlarged islet were found in both dexamethasone-treated groups as were a seemingly increased number of small ones.

Plasma glucose values both pre and post glucose were the same in all groups (Fig. 1). Dexamethasone caused the insulin response to the glucose challenge to be exaggerated in both Px and shams; the levels attained at all time points were equal in the two groups. The plasma insulin values in the 2 saline groups were also equal.

The protocol and results are shown in Fig. 2. The baseline perfusate contained 2.8 mM glucose and then 10 mM arginine was added. After reequilibration at 2.8 mM glucose, the glucose concentration was increased to 16.7 mM and arginine was again added.

As expected, the shams given saline had a much larger insulin response to arginine at the high glucose background with the mean insulin concentration being 14 times that at 2.8 mM (Table 3). The results in the Px-saline group were very similar; the mean insulin concentration attained with 16.7 mM glucose alone was only half that of shams, but the values were not significantly different (Table 3).

In shams, dexamethasone caused arginine induced secretion to be considerably increased at both glucose concentrations. In Px. the mean insulin concentration attained with arginine at 2.8 glucose was 4 times that in the Px-saline group. On the other hand, the results at the high glucose background were much more variable so that it was not clear whether this response was also increased (29.4 ± 7.11 Px-dexamethasone versus 20.4 ± 3.24 ng/ml Px-saline).