Defective Fatty Acid-mediated -Cell Compensation in Zucker Diabetic Fatty Rats

Although obesity is associated with insulin resistance, most obese humans and rodents remain normoglycemic because of compensatory hyperinsulinemia. This has been attributed to β-cell hyperplasia and increased low K glucose metabolism of islets. Since free fatty acids (FFA) can induce these same β-cell changes in normal islets of Wistar rats and since plasma FFA are increased in obesity, FFA could be the signal from adipocytes that elicits β-cell compensation sufficient to prevent diabetes. To determine if FFA-induced compensation is impaired in islets of rats with a diabetogenic mutation, the Zucker diabetic fatty (ZDF) rat, we cultured islets from 6-week-old obese (fa/fa) rats that had compensated for obesity and apparently normal islets from lean ZDF rats (fa/+) in 0, 1, or 2 mM FFA. Low K glucose usage rose 2.5-fold in FFA-cultured control islets from age-matched Wistar rats, but failed to rise in either the precompensated islets of ZDF rats or in islets of lean ZDF rats. Bromodeoxyuridine incorporation increased 3.2-fold in Wistar islets but not in islets from obese or lean ZDF rats. Insulin secretion doubled in normal islets cultured in 2 mM FFA (p < 0.01) but increased only slightly in islets from lean ZDF rats (not significant) and declined in islets from obese ZDF rats (p < 0.05). We conclude that, unlike the islets of age-matched Wistar rats, islets of 6-week-old heterozygous and homozygous ZDF rats lack the capacity for FFA-induced enhancement of β-cell function.

Pancreatic islets from obese rodents are enlarged and exhibit a marked increase in low K m glucose metabolism (1), which may account for their high output of insulin even at low concentrations of glucose (1)(2)(3). The hyperinsulinemia is regarded as a compensatory response that prevents hyperglycemia despite the insulin resistance that invariably accompanies obesity.
There is evidence that free fatty acids (FFA) 1 may be the signal from adipocytes that mediates this compensatory insulin secretion (1). Plasma FFA are elevated in obesity (4,5) and have long been known to stimulate insulin secretion (6 -8).
Moreover, the compensatory triad observed in islets from obese rats can be induced in normal islets by culturing them for 7 days in the presence of 1 or 2 mM FFA; low K m glucose metabolism rises dramatically (1), there is evidence of increased ␤-cell replication (1), and insulin secretion increases (1), confirming the earlier description of FFA-induced hyperinsulinemia (9). If the FFA-induced compensation in obesity is, at least in part, responsible for preventing diabetes, it follows that FFAinduced compensation may be impaired at or before the onset of diabetes. In this study we assess the ability of FFA to induce the compensatory triad of enhanced low K m glucose metabolism, increased ␤-cell replication, and insulin hypersecretion in islets from obese rats with a diabetogenic mutation, the Zucker diabetic fatty rat (ZDF-drt). We observe that FFA induction of these compensatory changes is impaired, not only in islets from obese homozygous rats, but also in islets from lean heterozygous ZDF rats.

MATERIALS AND METHODS
Animals-Four groups of rats were studied at 6 weeks of age. Male Wistar rats were purchased from Charles River Laboratories (Wilmington, MA). Homozygous obese ZDF-drt rats (fa/fa) and heterozygous lean ZDF littermates (fa/ϩ) were bred in our laboratory from (ZDF/drtfa(F10)) rats purchased from Dr. Richard Peterson (University of Indianapolis School of Medicine, IN). Rats from our colony exhibit the previously described phenotype (10). Obesity is discernible at ϳ5 weeks of age. Since all obese (fa/fa) male ZDF rats develop hyperglycemia (Ͼ200 mg/dl) and glycosuria by 8 -10 weeks of age, all 6-week-old obese males were considered to be prediabetic; not only was their 9 a.m. plasma glucose normal (135 Ϯ 3 mg/dl) but other ␤-cell hallmarks of diabetes, such as impaired glucose-stimulated insulin secretion and reduced GLUT-2, had not yet appeared (11,12). By contrast, obese homozygous females and lean heterozygous males almost never become hyperglycemic (9 a.m. plasma glucose 123 Ϯ 7 mg/dl).
All rats received standard rat chow (Teklad F6 8664, Teklad, Madison, WI) ad libitum and had free access to water. All institutional guidelines for animal care and use were followed.
Islet Culture-Pancreatic islets of 6-week-old rats were isolated by the method of Naber et al. (13) as modified by Lee et al. (14) and maintained for 7 days in suspension culture in 60-mm glass Petri dishes at 37°C in a humidified atmosphere of 5% CO 2 and 95% air as described previously (1,14); however, the glucose concentration of the medium was reduced to 8 mM, the minimum glucose level in which more than 80% of the islets will be present at the end of 7 days of culture. The long chain fatty acid mixture added to the culture medium was oleate: palmitate ϭ 2:1 (sodium salts, Sigma).
Perifusion Experiments-For perifusion experiments 50 -100 islets were hand-picked under a stereoscopic microscope, washed twice with Krebs-Ringer bicarbonate-Hepes buffer (pH 7.4, 3 mM glucose) and loaded into 13 mm chambers containing an 8-m nylon membrane filter (Millipore, Bedford, MA). Perifusion was carried out as described previously (14). Immunoreactive insulin was determined by radioimmunoassay (15) using charcoal separation (16). Insulin Content of Islets-Twenty islets each were washed twice with PBS and put into 500 l of acid ethanol (0.18 M HCl in 95% ethanol). Insulin was extracted overnight at 4°C after sonication for 1 min (Laboratory Supplies Co. Inc., Hicksville, NY).
Measurements of Glucose Usage-Glucose usage in cultured islets was measured by the method of Zawalich and Matchinsky (17) and Zawalich et al. (18), as described previously in full detail (1), except that 50 -100 islets were counted rather than pipetted and transferred to small vials.
Measurements of Cell Viability-After the 7-day culture period, ϳ20 islets were randomly selected from culture dishes of 0, 1, and 2 mM FFA of Wistar and ZDF (fa/ϩ) rats. Islets were washed twice with PBS and stained with fluorescein diacetate and ethidium bromide for 1 min. Five to six hundred cells from each of the six groups were counted randomly under a fluorescence microscope (Nikon Optiphot UFX II-A, Garden City, NY). Green cells were counted as viable and red cells as dead, and the data were expressed as percent viability.
Bromodeoxyuridine (BrdUrd) Incorporation-After a 3-day culture period, islets were fixed in Bouin's solution, immobilized in 6% gelatin and embedded in paraffin. Five-m thick serial sections were processed for insulin staining (Dako, Carpinteria, CA) and BrdUrd quantitation using an anti-BrdUrd antibody (19) (Boehringer Mannheim). BrdUrdpositive nuclei of insulin-positive cells were counted using a fluorescence microscope. Results were expressed as BrdUrd ϩ nuclei/total nuclei ϫ 100.
Statistical Analyses-All results are expressed as mean Ϯ S.E. Statistical significance was evaluated using two-way analysis of variance followed by Scheffe's multiple comparison test.

Effects of FFA on Basal Hyperinsulinemia of Prediabetic
Islets-As originally reported by Zhou and Grill (9) and confirmed by our laboratory (1), exposure of islets from normal Wistar rats to 1 and 2 mM FFA resulted in a concentration-dependent increase in steady-state insulin secretion at subfasting glucose levels (3 mM) (Fig. 1A, Table IA). Insulin secretion was also greater during perifusion with 23 mM glucose in the FFAcultured islets and the rise in insulin occurred sooner, although its magnitude was less. By contrast, in the obese ZDF groups the presence of 1 or 2 mM FFA in the culture medium profoundly reduced both basal and glucose-stimulated insulin secretion ( Fig. 1, C and D, Table IA). However, insulin secretion was much higher in the islets from both prediabetic and nonprediabetic obese ZDF rats in the absence of FFA, suggesting that maximal ␤-cell compensation had already taken place in vivo prior to the isolation of islets.
To avoid the influence of precompensation, we examined the effects of FFA on the islets of lean heterozygous ZDF animals in which obesity, insulin resistance and compensatory hyperinsulinemia were absent and no prior compensatory changes in islets had occurred. As shown in Fig. 1B, the FFA-induced increase in basal insulin secretion after culture in 2 mM FFA was small and not statistically significant; glucose-stimulated insulin secretion was significantly reduced, suggesting an intrinsic resistance to the effects of FFA on islet function in rats with a single ZDF allele.
To exclude insulin depletion as the cause of the impaired FFA-mediated effects on insulin secretion, insulin content was measured in all islet groups (Table IB). There were no significant differences in change of insulin content between the lean Wistar and ZDF groups. Moreover, since the secreted insulin in all groups (Table IA) represented at most only 3.8% of the total insulin content, it is unlikely that differences in pancreatic insulin content account for the effect of FFA on insulin secretion.
To exclude the possibility that elevated concentrations of FFA in vitro might have killed a greater number of cells in the islets of obese and lean ZDF rats than in the lean Wistar controls, we tested cell viability using fluoroacetate incorporation. As indicated in Table II, exposure to 2 mM FFA reduced viability by 7% in Wistar islets and 12% in ZDF islets, not nearly enough to account for the large differences in insulin secretion associated with exposure to FFA.
Effect of FFA on BrdUrd Incorporation in Islet Cells of ZDF-We have previously observed a 3-fold increase in BrdUrd incorporation in islets of normal Wistar rats cultured for 7 days in 2 mM FFA (1). Islets from ZDF rats were isolated at the age of 6 weeks, approximately 3 weeks before the expected onset of diabetes, and cultured in 0 or 2 mM FFA. BrdUrd incorporation, which in the absence of FFA was 1.6 times higher than in the smaller islets of lean controls, did not increase further, perhaps because ␤-cells had already expanded maximally in vivo. However, in islets of lean heterozygous males, in which no prior increase in ␤-cell volume had taken place, BrdUrd incorporation was not enhanced by culture in 2 mM FFA (Fig. 2, Table IC).

Effect of FFA on Low K m Glucose Usage by Prediabetic
Islets-Previously we reported a 3-fold increase in low K m glucose usage in normal Wistar rat islets cultured with 2 mM FFA (1), which we interpreted as evidence that FFA contributes to the compensatory hyperinsulinemia of obesity. To determine if the increased low K m glucose usage in precompensated islets of obese prediabetic rats could be further enhanced by exposure to FFA, we cultured such islets for 1 week in the presence of 1 or 2 mM FFA. As shown in Table III, no further enhancement of low K m glucose usage was induced in these precompensated islets. In islets of lean heterozygous ZDF rats which are devoid of precompensatory changes, low K m glucose usage in the absence of FFA was no greater than in islets of Wistar rats. However, in contrast to the Wistar islets, no increase in low K m glucose usage had occurred after 7 days of culture at 1 and 2 mM FFA (Table III).

DISCUSSION
The results indicate that islets from homozygous and heterozygous ZDF rats do not develop any of the "compensatory changes" that are induced in islets from normal Wistar rats after 7 day of culture in 1 or 2 mM long chain fatty acids (FFA). These changes include increased basal and glucose-stimulated insulin secretion, increased BrdUrd incorporation, and increased low K m glucose metabolism. Since compensatory hyperinsulinemia is required to prevent hyperglycemia in the face of worsening insulin resistance, it follows that the propensity for obesity-dependent diabetes in ZDF rats could be the consequence of the failure of FFA to induce the changes in ␤-cells that result in the necessary degree of hyperinsulinemia.
In the case of islets from 6-week-old obese homozygous ZDF rats a measure of compensation had already occurred in vivo prior to their isolation for the culture experiments, making it impossible to differentiate between precompensation that had reached a maximum and intrinsic resistance of ZDF ␤-cells to actions of FFA. The islets of 6-week-old obese rats are 3.3 times larger than those of lean controls (12) and their insulin secretion is 3.2 times greater (12). Lean heterozygous ZDF rats, by contrast, do not differ from Wistar islets with respect to ␤-cell volume density, low K m glucose usage or insulin secretion. Nevertheless, despite the absence of any antecedent compensatory changes in vivo, they too were completely unresponsive to culture in FFA. BrdUrd incorporation, which increased 3.2fold in islets of Wistar males in the presence of 2 mM FFA, did not increase at all in the islets of lean ZDF rats. Similarly, low K m glucose usage, which increased 2.5-fold in the Wistar animals, did not increase in the lean ZDF group. Finally, insulin secretion at 3 mM glucose, which more than doubled in the Wistar islets, increased only slightly (NS) in the lean ZDF group, and glucose-stimulated insulin secretion was reduced by the 2 mM FFA. Thus, high FFA levels comparable to those observed in plasma of prediabetic ZDF rats (14) did not elicit a normal compensatory insulin response. The ZDF colony may thus have an intrinsic ␤-cell defect that interferes with FFAmediated compensation. In as much as islets from 6-week-old obese homozygous ZDF rats appear to have reached a fully compensated state in vivo prior to their isolation, either more than 7 days are required in vitro for FFA-mediated induction of compensatory events in ZDF rats or, more likely in ZDF rats a time window during which compensation can occur close before the age of 6 weeks.
The mechanism by which FFA induce the changes that in islets of normal rats result in compensatory hyperinsulinemia has not as yet been identified. In preadipocytes a fatty acidactivated receptor with homology to the peroxisome proliferator-activated receptor (PPAR) was recently cloned (20). An isoform of PPAR is expressed in islets. 2 It may be involved in up-regulation of glycolytic enzymes responsible for low K m glucose usage. However, changes in alternative fuel use could be the cause of the increase in low K m glycolysis (21).
The mechanism of the normal FFA-induced increase in Br-dUrd incorporation is also unknown. To our knowledge, a mitogenic effect of FFA in vitro has not been described previously in mammalian cells. However, intracellular levels of palmitoyl-CoA within the known physiologic range are able to potentiate protein kinase C activity in vitro (22,23) and stimulate protein kinase C-catalyzed phosphorylation of epidermal-growth factor receptor (24).
If the compensatory hyperinsulinemia in normal islets is the result of FFA-mediated induction of enzymes, what is the mechanism of the failed compensation in islets of ZDF rats? Given the fact that islets from obese prediabetic rats have an abnormally high lipid content in vivo (14), lipid overload seems plausible. It has long been known that increased long-chain fatty acyl-CoA impedes glucose metabolism at multiple levels (25)(26)(27)(28)(29), as recently reviewed by McGarry (30). A novel additional mechanism, excessive acylation, also warrants consideration; just as unregulated overglycation resulting from hyperglycemia can modify the function of certain proteins, excessively high FFA levels causing overacylation might similarly alter protein functions (31).
In summary, these results confirm our earlier report (1) that normal islets cultured in the presence of elevated FFA levels develop the same changes in DNA synthesis, glucose usage and insulin secretion that are present in vivo in nondiabetic obese rats. We further demonstrated that islets from obese prediabetic and nonprediabetic rats do not develop the foregoing compensatory changes when cultured under these same conditions, perhaps because they had already occurred earlier in  vivo. Consequently, if insulin resistance worsens, such islets would be incapable of further insulin secretion and hyperglycemia would supervene. However, islets from lean heterozygous ZDF rats without ␤-cell precompensation also failed to respond normally to FFA enriched culture. It is possible that the ␤-cell unresponsiveness to FFA present at 6 weeks of age and thereafter accounts for the inability to compensate fully for the insulin resistance and thereby prevent diabetes.