Key points are not available for this paper at this time.
Richard G. Peterson is professor of anatomy at Indiana University School of Medicine, Indianapolis. Walter N. Shaw is a senior research scientist at Lilly Research Laboratories, Indianapolis. Mary-Ann Neel is a senior research technician and Leah A. Little is a research technician; they carried out all breeding and animal care protocols. J. Eichberg is professor of biochemical and biophysical science, University of Houston, Texas. This project was partially supported by PHS Grant RO1DK30577. Richard G. Peterson is professor of anatomy at Indiana University School of Medicine, Indianapolis. Walter N. Shaw is a senior research scientist at Lilly Research Laboratories, Indianapolis. Mary-Ann Neel is a senior research technician and Leah A. Little is a research technician; they carried out all breeding and animal care protocols. J. Eichberg is professor of biochemical and biophysical science, University of Houston, Texas. This project was partially supported by PHS Grant RO1DK30577. Richard G. Peterson is professor of anatomy at Indiana University School of Medicine, Indianapolis. Walter N. Shaw is a senior research scientist at Lilly Research Laboratories, Indianapolis. Mary-Ann Neel is a senior research technician and Leah A. Little is a research technician; they carried out all breeding and animal care protocols. J. Eichberg is professor of biochemical and biophysical science, University of Houston, Texas. This project was partially supported by PHS Grant RO1DK30577. Richard G. Peterson is professor of anatomy at Indiana University School of Medicine, Indianapolis. Walter N. Shaw is a senior research scientist at Lilly Research Laboratories, Indianapolis. Mary-Ann Neel is a senior research technician and Leah A. Little is a research technician; they carried out all breeding and animal care protocols. J. Eichberg is professor of biochemical and biophysical science, University of Houston, Texas. This project was partially supported by PHS Grant RO1DK30577. Richard G. Peterson is professor of anatomy at Indiana University School of Medicine, Indianapolis. Walter N. Shaw is a senior research scientist at Lilly Research Laboratories, Indianapolis. Mary-Ann Neel is a senior research technician and Leah A. Little is a research technician; they carried out all breeding and animal care protocols. J. Eichberg is professor of biochemical and biophysical science, University of Houston, Texas. This project was partially supported by PHS Grant RO1DK30577. A number of animal models for non-insulin-dependent diabetes mellitus (NIDDM) have been described over the years. These include common laboratory animals such as mice ( Coleman, 1982a ), rats ( Greenhouse et al., 1988 ; Peterson et al., 1988a , b ), other rodents ( Coleman, 1982b ), dogs ( Engerman and Kramer, 1982 ), as well as more unusual animals such as Chinese hamsters ( Gerritson, 1982 ), swine ( Phillips et al., 1982 ), and nonhuman primates ( Howard, 1982 ). No single model can answer all the research questions related to a disease being investigated; however, models that resemble the disease under study in onset, course, and symptoms do have a very significant role. Two frequently used mouse models for NIDDM are the models containing the db and ob genes ( Coleman, 1982a ). These genes have been placed in several inbred strains, the result being a number of variations in the severity of diabetes ( Coleman, 1982a ). Rat models for NIDDM have been developed more recently using strains primarily involving the fa and the cp genes. The fa gene was first described in the Zucker rat ( Zucker and Zucker, 1961 ), and the cp gene appeared spontaneously in a colony produced as a result of a cross between the SHR/N strain and the SD rat ( Koletsky, 1973 ). These genes have been maintained and backcrossed into a number of strains ( Greenhouse et al., 1988 ; Peterson et al., 1988a ). Again, variations of the level of diabetes are evident. To date, the model that appears to be most consistently diabetic is the male Wistar diabetic fatty (WDF) rat ( Ikeda et al., 1981 ; Peterson et al., 1988a ). Most of these rat models can be described as having a moderate degree of hyperglycemia. As a general rule, Zucker fatty rats have not displayed hyperglycemia when fed ad libitum ( Bray, 1977 ; Herberg and Coleman, 1977 ), although they have demonstrated a reduced glucose tolerance and other indications of insulin resistance ( Rohner-Jeanrenaud et al., 1986 ). Several years ago, an unusually high level of blood glucose and intolerance to glucose were observed in a few animals in the colony of Zucker rats from which the rats described in this paper were originally derived ( Clark and Palmer, 1981 , 1982 ; Clark et al., 1983 ). Although this trait was difficult to follow in the intervening years, it continued to appear in a number of fatty male rats. About three years ago, we began selectively inbreeding for this trait and have been successful in maintaining a very high level of hyperglycemia in fatty male rats for the last eight generations. This paper describes the characteristics of the resulting model. Diabetic fatty rats and lean rats with a high degree of diabetes in their background were initially chosen for breeding. Male fatty-diabetic breeders were used at each generation when available. Inbreeding has continued for 10 generations. In all but one instance in which the mating was parent x offspring, inbreeding was by brother x sister mating. As a result of this inbreeding, every fatty male rat has been very hyperglycemic and has remained diabetic for as long as they have been followed (up to 42 weeks). The phenotype has been consistently expressed since the F2 generation. Fed hyperglycemia (blood glucose >175 mg percent) was observed in most fatty male rats at 7 weeks of age. All fatty male rats were above 200 mg percent by 9 weeks of age, and by 10 weeks the average level was above 400 mg percent. The average blood glucose levels then gradually increased to above 500 mg percent where they remained for the life of the rat. HbAl, free fatty acids, triglycerides, and cholesterol levels were also significantly higher than control. This partially inbred strain is being referred to as Zucker diabetic fatty (ZDF/Drt- fa ) and appears to be a very good model for NIDDM. The rats selected to start this inbreeding program were chosen either because they were diabetic or had a high incidence of diabetes in their background. Up to this point, there had been no inbreeding in our relatively small closed colony. Strict inbreeding practice has been followed at each subsequent generation. In all but one generation—where a fatty diabetic male ( fa/fa ) was bred with its daughter—the breeding has been by brother x sister mating. In all but two generations of inbreeding, fatty diabetic males ( fa/fa ) were bred with +/ fa females. Fatty males were produced in 9 out of the 10 inbred generations. Each of the fatty males was observed for the onset of diabetes from about 6 weeks of age. All rats were watered and fed (Purina 5008) ad libitum. Each week, animals were weighed, and post-feeding blood glucose levels were determined. In all cases, diabetes developed before the fatty male rats were mature enough to breed. Males and females were mated starting at about 9 weeks of age. Data were also collected from two other NIDDM rat models maintained in our laboratory. The first, WDF/Ta- fa (Wistar fatty), was first described by Ikeda ( Ikeda et al., 1981 ). We have published some additional data ( Berti-Mattera et al., 1989 ; Kava et al., 1990 ; Peterson et al., 1988a , b ). The WDF/TaDrt- fa rats we have used for comparison had been further inbred (Nil, F15). The second model is the WKY/NDrt- fa (N10, F3) rat, which was backcrossed in our facility ( Greenhouse et al., 1988 ; Peterson et al., 1988a , 1990 ). This latest model is being maintained at the National Institutes of Health by Dr. Carl Hansen. All animals in this study were maintained pathogen free in filter-top cages, and all cages were changed in a laminar-flow hood. Blood was collected and analyzed for HbAl (Quik-Sep ® , Fast Hemoglobin ® Test System, Isolab, Inc.), insulin (modified from Hedding, 1965 ), free fatty acids ( Falholt et al., 1973 ), triglyceride (BMC Enzymatic-UV-reagentset, triglycerides, Boehringer Mannheim Diagnostics) and cholesterol levels ( Sperry and Webb, 1950 ) either periodically from the tail vein or by exsanguination at the time rats were terminated. Data were analyzed using Student’s t-test. Figure 1 , the inbreeding diagram for the rats we have maintained and selectively bred for the diabetic trait, shows the breeding pairs for two generations before inbreeding began. Each generation’s fatty male rats were characterized as either diabetic or nondiabetic. One rat in the Fl generation was transiently diabetic and was not used for breeding. Birth dates and fa genotypes for all of the weaned rats in the litters are also shown. All inbreedings were brother-sister except for the F2 litter, which was produced by a father-daughter breeding. Fatty diabetic males were used for breeding in 8 of the 10 inbred generations. Fatty males were also used before inbreeding was started and are shown in an anticipated breeding at the right end of the figure. Inbreeding generations that have been maintained in our main line of Zucker diabetic fatty rats. Boxes on the left indicate the number of females that were fatty and lean in each litter, and the boxes on the right indicate the genotypes and phenotypes of the males. Many other inbreedings at each generation demonstrated the same phenotypes. The F3 litter was cesarean delivered. The phenotype of each parent and the delivery dates are shown down the center of the figure. Figure 2 shows the time course of weight gain for fatty diabetic male rats in this colony, which was greater than for control rats but less than for some other fatty rats we have followed. Table 1 shows a comparison of the weights at 12 weeks of age for the three fatty strains we have studied. ZDF/Drt- fa rats were significantly lighter than either WDF/TaDrt- fa or WKY/NDrt- fa rats. Weight gain with age (mean ® SD) for the Zucker diabetic fatty male rats followed from generation F4 to the present generation. The beginning of the curve includes data from up to 45 rats. At 35 weeks there were 8 rats. Only one rat was maintained beyond 35 weeks. Comparison of Weight and Blood Glucose Levels in Three Fatty Rat Models at 12 Weeks of Age Significant at p < 0.05 when compared to the two other groups. Figure 3 illustrates the changes in average blood glucose levels over time for ZDF/Drt- fa male rats followed for generations F4-F10. Because rats were kept for variable periods of time, the data from the younger animals at the beginning of the graph are from a much larger population of rats. Eight rats are represented at 35 weeks and only one rat at 36-37 weeks. Although fewer rats were followed in the last half of the graph, none of the rats beyond generation F2 lost their diabetic trait during the study period. Two rats in generation F6 and four rats in generation F8 did show a slower rise in blood glucose, and their blood glucose reached a plateau at a lower level (in the 300-400 mg percent range) than did that of the rest of the study group. Data from these rats are included in Figure 3 . If these data were not included, the error bars would be much smaller, and the average would be somewhat higher. This variability is probably related to the lack of homogenicity of other genes influencing the expression of diabetes at this point in inbreeding. Change in blood glucose levels (mean ® SD) for Zucker diabetic fatty male rats (data are from same rats as Figure 2 ). Table 1 also demonstrates the differences in blood glucose levels between 12-week-old ZDF/Drt- fa , WDF/TaDrt- fa , and WKY/NDrt- fa males. If the animals with lower blood glucose levels (mentioned above) are excluded from the analysis, the ZDF/Drt- fa males have a higher average blood glucose level with a much smaller standard deviation. Table 2 shows analytical data we have obtained from rats following euthanasia. There were significant differences between fatty and lean animals in the levels of glycosylated hemoglobin, free fatty acids, triglycerides, and cholesterol at both age ranges and in insulin levels in the younger age range. In all cases, the ZDF/Drt- fa rats had significantly higher values. Insulin levels were not significantly different in the older age ranges. Diabetes has been observed in a number of rat and mouse strains having genetic traits that make them obese. Some of these strains, such as most fatty Zucker rat colonies, are typically nondiabetic. We have isolated and described a genetic characteristic in our colony of male fatty Zucker rats that seems to cause a more severe form of diabetes than occurs in most other rat models for NIDDM. The diabetic trait in our colony was first reported in 1981. Since then, fatty rats in our colony periodically expressed the diabetic trait first described some years ago ( Clark and Palmer, 1981 , 1982 ; Clark et al., 1983 ). However, consistent demonstration of this characteristic was not seen until we selectively inbred for this trait. We are also attempting to selectively inbreed a totally nondiabetic Zucker model. The origin of this model is the same original breeding pair from which the ZDF/Drt- fa strain was derived. To date, we have seen diabetic, nondiabetic, and transiently diabetic rats with the fatty phenotype in this line. Based on the observation that there are three phenotypes in this group, we propose that the diabetic trait is dominant when it occurs with the fatty phenotype, but that it is incompletely expressed when only one gene for the trait is present. End-of-Experiment Data from Fatty and Lean Zucker Rats, Mean ± SD Significant at p < 0.05 when compared to lean at the same age.
Peterson et al. (Mon,) studied this question.