Diabesity: A Polygenic Model of Dietary-Induced Obesity from Ad Libitum Overfeeding of Sprague–Dawley Rats and Its Modulation by Moderate and Marked Dietary Restriction

Abstract

This study compared the effects of ad libitum (AL) overfeeding and moderate or marked dietary restriction (DR) on the pathogenesis of a metabolic syndrome of diabesity comprised of age-related degenerative diseases and obesity in a outbred stock of Sprague–Dawley (SD) rats [Crl:CD (SD) IGS BR]. SD rats were fed Purina Certified Rodent Diet AL (group 1), DR at 72–79% of AL (group 2), DR at 68–72% of AL (group 3) or DR at 47–48% of AL (group 4) for 106 weeks. Interim necropsies were performed at 13, 26, and 53 weeks, after a 7-day 5-bromo-2-deoxyuridine (BrdU)-filled minipump implantation. Body weights, organ weights, carcass analysis, in-life data including estrous cyclicity, and histopathology were determined. At 6–7 weeks of age SD rats had 6% body fat. AL-feeding resulted in hypertriglyceridemia, hypercholesterolemia, and dietary-induced obesity (DIO) by study week 14, with 25% body fat that progressed to 36–42% body fat by 106 weeks. As early as 14 weeks, key biomarkers developed for spontaneous nephropathy, cardiomyopathy, and degenerative changes in multiple organ systems. Early endocrine disruption was indicated by changes in metabolic and endocrine profiles and the early development and progression of lesions in the pituitary, pancreatic islets, adrenals, thyroids, parathyroids, liver, kidneys, and other tissues. Reproductive senescence was seen by 9 months with declines in estrous cyclicity and pathological changes in the reproductive organs of both sexes fed AL or moderate DR, but not marked DR. The diabesity syndrome in AL-fed, DIO SD rats was readily modulated or prevented by moderate to marked DR. Moderate DR of balanced diets resulted in a better toxicology model by significantly improving survival, controlling adult body weight and obesity, reducing the onset, severity, and morbidity of age-related renal, endocrine, metabolic, and cardiac diseases. Moderate DR feeding reduces study-to-study variability, increases treatment exposure time, and increases the ability to distinguish true treatment effects from spontaneous aging. The structural and metabolic differences between the phenotypes of DIO and DR SD rats indicated changes of polygenic expression over time in this outbred stock. AL-overfeeding of SD rats produces a needed model of DIO and diabesity that needs further study of its patterns of polygenic expression and phenotype.

Introduction

It is well established that moderate caloric or dietary restriction (DR) significantly improves 2-year survival, controls adult body weight, and delays the onset of diet- and age-related spontaneous diseases and tumors. This results in a better experimental toxicity model by reducing the “noise” of background diseases while allowing an increased duration of exposure to test substances for the evaluation of the potential carcinogenicity and toxicity in long-term studies. The adverse effects of ad libitum (AL)-overfeeding on the early development of many spontaneous tumors and degenerative diseases of this SD outbred stock (Gumprecht et al., 1993; Keenan et al., 1994a, 1994b, 1995a, 1995b, 1996, 1999, 2000a, 2000b; Dixit et al., 1996; Keenan et al., 1997; Hubert et al., 1997; Laroque et al., 1997; Hoe et al., 1998; Vermorel et al., 1998; Hubert et al., 2000; Kemi et al., 2000; Molon-Noblot et al., 2003) and other aged rat strains (McCay et al., 1935; Burek, 1978; Tucker, 1979; Ross et al., 1983; Kritchevsky et al., 1984; Maeda et al., 1985; Berry, 1986; Masoro et al., 1989; Laganiere and Yu, 1989a, 1989b; Yu et al., 1989; Mietes, 1990; Chapin et al., 1993; Grasl-Kraupp et al., 1994; Merry and Holehan, 1994; Sonntag and Yu 1994; Roe et al., 1995; Masoro et al., 1996; Masoro and Austad, 1996; McShane and Wise, 1996; Seki et al., 1997; Kritchevsky, 1999; Sonntag, 1999; Duffy et al., 2001; Haseman et al., 2003; Wan et al., 2003) have been reported. However, the role of AL-overfeeding in the pathogenesis of dietary-induced obesity (DIO) and the metabolic syndrome (syndrome X) associated with adult-onset diabetes, or “diabesity” (Levin et al., 1997; Leiter, 2002; Reifsnyder and Leiter, 2002; Axen et al., 2003) in SD rats has not been fully investigated or exploited as a model of the polygenic diabesity syndrome which is common in heterozygous human populations worldwide (Klinger et al., 1996; Weindruch and Sohal, 1997; Brunner et al., 2001; Eckel et al., 2002; Pasquale et al., 2003).

This paper describes the temporal, clinical, and pathological features of the “diabesity syndrome” as observed in AL-overfed SD rats and demonstrates the beneficial effects of moderate or marked DR in modulating the many co-morbidities associated with this syndrome. The SD rat stock used in this study is officially designated as the Charles River CD rat, and should not be confused with other “Sprague–Dawley” rat stocks that breeders have developed with very different phenotypes under similar housing and feeding conditions. We refer to the animals in our study as “SD,” but the breeder uses the designation of CD IGS or Crl:CD (SD) IGS BR to identify their albino outbred SD stock from others. This stock originated in 1925 at the University of Wisconsin through the efforts of Robert W. Dawley (the stock’s name was derived from his wife’s maiden name and his own). A docile, large hybrid hooded male was mated to female Wistar (albino) rats, and after 7 generations the rats were outbred. Obtained by the Charles River Company in 1950, the stock was rederived in 1955 and in 1991 the breeder selected 8 lines of this stock to form the “IGS” foundation colony that was rederived in isolators in 1997. This stock is very docile, with high fecundity and rapid growth when AL-fed commercial diets. It is considered one of the best outbred SD rat stocks and is commonly used in behavioral, nutritional, reproductive, teratological, toxicological, and carcinogenicity testing worldwide.

Adult-onset human or animal type 2 diabetes associated with obesity (diabesity) is induced by a complex set of genetic, dietary and environmental interactions (Weindruch and Walford, 1988; Klinger et al., 1996; Brunner et al., 2001; Eckel et al., 2002; Bray, 2002). For example, monogenic obesity mutations in rodents such as those in the leptin gene (Lep^ob, ob/ob mice) or its receptor gene (Lepr^db, db/db mice, Lepr^fa fa/fa Zucker rats or Lepr^cp cplcpJCR rats) have been extremely useful in the study of these processes and useful in efficacy testing of anti-obesity and anti-diabetic drugs (Harrison and Archer, 1987; Lee and Yu, 1990; Lu et al., 1991; Leiter, 2002; Inui et al., 2004; Park and Prolla, 2005). However, the monogenic basis of these mutated inbred rodent models does not reflect the more common forms of human obesity and adult-onset type 2 diabetes (diabesity) which are known to be a polygenic syndrome in heterozygous human populations (Klinger et al., 1996; Whitaker et al., 1997; Eckel et al., 2002; Hursting et al., 2003; Konstantinov, 2003; Rauser et al., 2003; Park and Prolla, 2005). Thus, there is a need for a polygenic outbred obesity rodent model in which disease trait loci interact with each other and can be modulated by the diet and environment to elicit DIO syndromes that are potentially diabetic and will better represent the most common human syndromes. While inbred and transgenic rats and mice with well-known quantitative trait loci for obesity and/or diabetes have been helpful in understanding the genetics of these processes, many diabetes-prone strains with different combinations of disease-associated loci do not develop obesity and conversely many strains that develop genetically driven obesity syndromes are not diabetes-prone (Leiter, 2002). Inbred strains with null mutations in specific genes are not truly representative of the human polygenic syndrome of obesity-driven type 2 diabetes or diabesity. For this reason, more polygenic outbred obesity rodent models are needed to determine the quantitative trait loci that interact with each other and could be modified by new therapeutic means. When overfed, the Charles River outbred SD rat stock has been shown to develop a phenotype of DIO that progresses to an adult-onset type-II diabetic syndrome. This syndrome, characterized by the development of hyperlipidemia, hyperinsulinemia, changes in glucose metabolism and the many other co-morbidities that model a polygenic adult-onset, obesity-induced diabetes (diabesity) in humans (Keenan, 1994a, 1994b, 1996, 1997, 1999, 2000a, 2000b; Levin et al., 1997; Molon-Noblot et al., 2001, 2003). This SD rat stock’s phenotype is reminiscent of the syndrome described in hybrid mice in which different quantitative trait loci are combined, leading to obesity and diabetes syndromes (Leiter, 2002; Reifsnyder and Leiter, 2002; Park and Prolla, 2005). This cross-sectional and longitudinal study describes the pathologic features of DIO adult-onset diabesity in SD rats induced by simple ad libitum overfeeding of a commercial rodent diet and the modulation of this syndrome by different degrees of moderate to marked dietary restriction. These data demonstrate the untapped potential of this outbred SD stock as a more appropriate model of the human polygenic adult-onset type 2 diabetes syndrome that is driven by dietary-induced obesity.

Materials and Methods

Animals

Three hundred-eighty male and 380 female Sprague–Dawley [Crl:CD (SD) IGS BR] rats were obtained from Charles River Laboratories, Raleigh, NC. The animals were 7 weeks old at the initiation of the study and weighed 172–266 g for males and 134–213 g for females. The rats were individually housed in stainless steel cages in an environmentally controlled room with a 12-hour light cycle (lights on at 0700 hours and off at 1900 hours). They were individually identified with implantable microchip identification devices (BioMedic Data Systems) and allocated to cages by a randomized columnar allocation scheme. The animals were assigned to 4 different treatment groups using a balanced random allocation scheme based on body weight. Each group consisted of 95 males and 95 females with 15 animals/sex/group allocated for the 13-, 26-, and 53-week interim necropsy, and with 50 animals/sex/group allocated to the 106-week final necropsy. The Institutional Animal Care and Use Committee at Merck Research Laboratories, West Point, PA reviewed and approved all procedures used in this study.

Diet and Dietary Regimen

Purina Certified Rodent Diet was provided in the morning (between 0730 to 0830 hours) either (1) AL (Group 1), (2) moderate DR (Group 2) fed 72–79% of the adult AL intake (17 and 24 g/day to females and males respectively), (3) moderate DR (Group 3) fed 68–72% of the adult AL intake (16 and 22 g/day for females and males respectively) or (4) marked DR, fed 47–48% of the adult AL amount (11 and 14.5 g/day to females and males respectively). The DR amounts were fixed, measured portions fed daily throughout the study to the respective groups. Purina Certified Rodent Diet contained 21% protein, 4.5% lipid, 55% carbohydrate and 4.1% fiber. The metabolizable energy content was 3.1 kcal/g of diet (Keenan et al., 1997, 2000). Drinking water was available AL. All rats were individually dosed with 0.5% aqueous methylcellulose by oral gavage daily at a dosing volume of 5 ml/kg to match control dosing conditions in typical toxicity and carcinogenicity studies in our laboratories.

Clinical Evaluations

All rats were observed daily for clinical signs and weighed before the start of the study, once in week 1, generally twice weekly through week 13, and once a week thereafter. At each necropsy, blood was collected from the vena cava in 15/rats/sex/group for determination of terminal serum biochemistry, hematology, and hormone levels as reported elsewhere. Food consumption was measured weekly and food wastage was measured on study weeks 6, 12, 24, 48, and 96. Water consumption was measured on study weeks 38, 49, 78, and 99. To determine the effects of aging and feeding methods on reproductive senescence, all females scheduled for the final necropsy (50/sex/group) were evaluated by daily vaginal lavage and cytology for a minimum of 20 consecutive days at 6 weeks, 6, 9, 12, 15, 18, 21, and 24 months of age (study weeks 1, 18, 31, 43, 60, 71, 83, and 96 respectively) to monitor estrous cyclicity patterns.

Osmotic Minipump Implantation

One week prior to each necropsy, 10 rats/sex/group selected by a stratified random allocation scheme were implanted with osmotic minipumps (model #2ML1, 2ML, Alza Corp., Palo Alto, CA) for a 7-day continuous delivery of 5-bromo-2′-deoxyuridine (BrdU; Sigma Chemical Co., St. Louis, MO). Prior to implantation, the minipumps were loaded with BrdU at a concentration of 50 mg/ml in a 0.5 N sodium bicarbonate solution. The loaded minipumps were surgically implanted in the subcutaneous tissue under isoflurane anesthesia, via a small dorsal midline skin incision. The incisions were closed with surgical staples, and the rats were returned to their cages until scheduled necropsy.

Necropsy, Carcass Analysis, and Histopathology

At the scheduled interim and final necropsies, following an overnight fast, all rats were euthanized by exsanguination under deep anesthesia and the minipumps were removed prior to obtaining terminal body weights, organ weights, and before tissue sampling. A complete gross examination was performed on all animals. Following sampling for histology the carcasses were frozen, ground and sampled for carcass analysis of total body protein (Kieldahl nitrogen), fat (Acid hydrolysis), moisture, and ash content as previously described (Keenan et al., 1994a, 1999). Numerous organs were weighed and tissue samples were taken for microscopic examination and fixed in 10% neutral buffered formalin. Testes and epididymides were fixed in Bouin’s solution. Five-μm-thick sections of paraplast embedded tissues were stained with hematoxylin and eosin, and examined microscopically. Additional histochemistry stains were performed as indicated for diagnosis and evaluation. Immunohistochemistry for BrdU staining was performed on adjacent sections of heart and liver as previously described (Keenan et al., 1994a, 1995a, 1995b).

Stereology

Hepatocyte Nuclear Number

The hepatocyte nuclear area (μm²) was analyzed using 5-μm liver sections stained for bromodeoxyuridine (BrdU). A total of 6 random fields per liver section were selected at a magnification of ×600 using an Olympus BH-2 microscope, a Sony 3 CCD video camera, and the Bioquant/TCW image analysis software. Using Bioquant/TCW software, each hepatocyte nucleus was traced automatically. The area (μm²) and perimeter (μm) were determined by the software and mean values were calculated for each animal.

For hepatocyte number, the numerical density was calculated from direct measurements to determine the number of hepatocyte nuclei per cubic centimeter of liver. To determine the volume of liver, the density of liver (0.93 cm³/gm) was multiplied by the weight (gm) of the liver. The corrected liver volume was calculated by multiplying the volume of liver by a measured shrinkage factor (0.118). The number of hepatocyte nuclei per liver was determined by multiplying the corrected liver volume by the number of hepatocyte nuclei per cubic centimeter of liver. The total number of BrdU labeled hepatocyte nuclei per liver was determined by multiplying the total number of nuclei per liver by the BrdU labeling index (%LI).

Hepatocyte BrdU Labeling Index (LI%)

The 7-day cumulative DNA synthesis (BrdU labeling) index of the hepatocyte nuclei was measured by scoring approximately 2,000 hepatocyte nuclei. To count 2,000 nuclei, 20 fields per liver were analyzed by the CHRIS (Cytology/Histology Recognition System) computer software system at a magnification of ×400. The percentage of hepatocyte BrdU nuclear labeling index (%LI) was calculated for each animal. The total number of BrdU labeled hepatocyte nuclei per liver was determined by multiplying the total number of nuclei per liver by the labeling index.

Myocardial Fibrotic Index (MFI%)

The myocardial fibrotic index was analyzed using 5-μm heart sections stained with Masson’s Trichrome. At least 20 random fields were evaluated throughout the left ventricular freewall and the interventricular septum for each animal at a magnification of ×100 using an Olympus BH-2 microscope, a Sony 3 CCD video camera, the Bioquant/TCW image analysis software. The valvular tissues and chordae tendineae were avoided. The myocardial fibrotic areas included subepicardial, subendocardial, interstitial, and perivascular fibrosis. The myocardial fibrotic index (MFI%) was obtained by using the total area (μm²) of cardiac tissue stained with the Biebrich Scarlet and Aniline blue portions of the Masson’s Trichrome stain within the field and then measuring the area (μm²) of fibrotic tissue stained with Aniline blue alone. The MFI% was calculated for each field and then a mean was calculated for each animal by dividing the area of Aniline blue-stained connective tissue by the total area of cardiac tissue measured and multiplying by 100, as previously described (Weibel, 1979; Keenan et al., 1995; Kemi et al., 2000).

Statistical Analysis

Statistical analyses were performed on organ weights, carcass analysis data, water consumption, food consumption, lipid profiles, estrus cycle and survival data. These analyses were done in a logarithmic scale in order to satisfy the assumptions for continuous parameters that follow a normal distribution. Each parameter was analyzed separately for each gender with necropsy intervals combined. Additional analysis for each necropsy time point was done for some parameters. Analysis for the combined time points was conducted by using area under the curve in serial sacrifice to characterize the time response pattern. Tukey’s multiple range test was used for the comparisons among the 4 dietary groups. A trend analysis was used to determine if there was a significant trend with decreasing diet across all dietary groups. When there was a significant trend, the data from the low dietary group was excluded and the trend test was repeated. This process was repeated until non-significance (p > 0.05) was achieved. The group averages were summarized by the geometric mean (Tukey et al., 1985; Snedecor et al., 1989; Hoe et al., 1998; Holder et al., 1999).

Results

Mortality and Survival

For both sexes survival was proportional to the degree of DR. In groups 1, 2, 3, and 4, 106-week survival rates were 18, 40, 56, and 82% in females, and 18, 44, 68, and 78% in males, respectively. The group 4 animals had the highest percentage and average weeks of survival on the study, and the AL-fed group 1 animals had the lowest values (approximately 18% survival for both sexes and 81 to 83 average weeks on study). Interestingly, there was a difference in survival that was significant between the two moderate DR-fed groups 2 and 3, in both sexes, by trend test. The average survival and weeks on study were statistically different between groups by both trend tests and pair-wise comparisons as shown in Table 1. The most common cause of death in both sexes of all groups was pituitary adenomas. The next most common cause of death in the AL-fed males was chronic nephropathy, followed by cardiomyopathy, and in AL-fed females it was mammary gland tumors. The tumor data will be presented in more detail in a separate paper.

Food and Water Consumption

The mean food consumption in the AL-fed group 1 over the course of the study ranged in females from 19.8 to 29.8 grams per day (mean value 23.6 grams per day) and in males from 26.4 to 33.3 grams per day (mean value 30.4 grams per day) (Figure 1). The mean measured food consumption was approximately equivalent to their daily allotment of ration of 17, 16, and 11 grams per day for females and 24, 22, and 14.5 grams per day for males in groups 2, 3, and 4, respectively., Females in groups 2, 3, and 4 consumed 72, 68 and 47% of AL intake and males in the same groups consumed 78, 72, and 48% of AL intake respectively. The absolute food intake corrected for food wastage was approximately 6 to 12% in all groups. There were no significant differences in food wastage between any of the 4 treatment groups over the course of the study. While absolute food intake was significantly different between the 4 groups, the mean relative food consumption per gram body weight was similar in all groups. These data indicate total body growth is proportional to the total amount of food provided or consumed in all groups. However, when the body weights were expressed as percent lean body weight, the percent food intake was not proportional to the lean body weight. As they aged the AL-fed animals partitioned more of their intake into body fat rather than lean body mass (Figure 9). The calculated intake of nutrients, including metabolizable energy, showed similar trends throughout the study, with absolute and percent lean body weight, nutritive and energy intake proportional to the total body weights achieved. However, relative food intake expressed as a percent of total body weight was remarkably similar across groups (Figures 2 and 3). These data indicate a significant portion of the food consumed by the AL groups was converted and stored as body fat.

Water consumption, measured throughout the study as an absolute value and a relative value adjusted for body weight, indicated absolute intake was proportional to body size. However, when expressed as ml. water per kg. body weight, the DR-fed males in groups 2 and 3 had approximately a 30% increase in relative water consumption and the females in the same groups had approximately 20 to 40% increase in relative water consumption compared to their AL-fed counterparts. Males in group 4, the marked DR-fed group, had approximately a 70% increase in water consumption, and females had 10 to 15% increase in water consumption compared to the average water intake per gram body weight in the AL-fed groups (Figures 4 and 5).

Glucose, Insulin, Triglyceride, and Cholesterol Levels

Fasting blood glucose and insulin data have been previously reported (Molon-Noblot et al., 2001). In summary, the mean insulin values were higher in AL-fed animals than other groups over the course of the study. In fasted rats, glucose values from vena cava samples were generally similar across groups, but group 4 had the lowest mean values. Blood glucose was also determined at intervals via tail stick methods on unfasted animals over the course of the day. In male rats, glucose values were generally comparable across groups in the morning prior to feeding when AL-fed rats’ glucose was higher than those of the restricted groups. In females, mean blood glucose values were generally higher in groups 1 and 2 than in groups 3 and 4 (Molon-Noblot et al., 2001). The AL-fed animals had higher serum levels of insulin, IGF-1 and glucose than the DR-fed animals as previously reported (Molon-Noblot et al., 2001, 2003).

Plasma total cholesterol levels were consistently higher in AL-fed animals of both sexes and the lowest in the Group 4 marked DR-fed rats throughout the study. By study weeks 25 through 51, cholesterol levels in the Group 1 AL-fed animals increased significantly above any of the food restricted groups (Figure 6). Serum triglycerides showed an even more dramatic pattern of change in the AL-fed animals so that by 24 weeks, a significant increase in triglycerides was seen in the AL-fed animals of both sexes. A proportional decrease in tryglyceride levels relative to food intake was observed in the 3 food restricted groups. These differences in lipid profiles were most evident in the females at the latter portions of the study (Figure 7).

Growth Curves and Body Weights at Interim and Terminal Necropsies

Differences in the total food intake resulted in not only proportional and increasing rates of body weight gain, but terminal body weights that were different between the groups as early as the first few weeks of the study. For the AL-fed and 2 moderate DR-fed groups, body weight gain and terminal body weights increased proportionally up to 1 year and then leveled off for the males, but increased in AL-fed females through 106 weeks. In the group 4 females, there was a slight decrement in body weight during the first 8 weeks of the study at which point this group began to gain body weight and maintained its average weight over the course of the study (Figure 8). Individual rats, particularly the AL-fed rats developed a terminal senescent body weight loss as they developed tumors, and cardiac or endocrine disease as noted by others (Black et al., 2003). These events and the early deaths of the largest rats tended to lower the average body weights of most groups from study week 80 onward (Figure 8 and Table 1).

Carcass Analysis at Interim and Terminal Necropsies

The whole body carcass analysis for percentages of protein, fat, moisture and ash content were determined on 10–15 rats/sex/group from each interim necropsy and the final necropsy. In addition, carcass analysis was performed on 10 rats/sex at 6 weeks of age to establish baseline values on AL-fed animals at the study initiation. The mean percent whole body carcass analysis values for AL-fed, 6-week-old male and female SD rats were: 72.4 and 72.3% moisture, 17.7 and 17.3% protein, 6.14 and 5.84% fat, 2.50 and 2.53% ash and 27.6 and 27.7% dry matter respectively. Therefore, 6-week old, AL-fed SD rats are typically lean animals at the study onset, with only 6% whole body fat. Differences in carcass composition between adult study groups generally reflected differences in total food consumption and their resultant body weight gains and terminal body weights. The AL-fed rats of both sexes rapidly developed DIO with the largest body fat content that became a significant portion of their total body weight from 6 months onward. On gross examination much of this body fat was white adipose tissue that was centrally located in the abdominal cavity and the subcutaneous tissues of the abdomen and lower thoracic area (Figure 10). The group 4 (50% DR) rats had the smallest body weight fat content. For both females and males a clear dose response pattern was noted for all components of body composition. Furthermore, the AL-fed groups were statistically significantly different in carcass composition from both the moderate and the marked DR-fed groups 2, 3, and 4 (Table 2). Thus, the AL-fed animals had the greatest gain in central body fat with age, the greatest percent of body fat, the greatest grams of body fat per animal, and when expressed as relative values of percent of body weight or brain weight, a clear dose-related increase in body fat was observed in the AL-fed animals relative to the three other measured-fed groups (Figure 9). Conversely, the highest percentage of carcass protein, ash and moisture content was seen in the marked DR-fed group 4 which was largely a reflection of their proportional decrease of total body fat content throughout the study. The moderate DR-fed groups were not lean and gained body fat with age, with males and females developing averages of 13 and 25% body fat, respectively, from 1 year onward (Figure 9 and Table 2).

Organ Weights at Interim and Terminal Necropsies

The AL-fed group 1 animals had the largest percent of body weight gain, largest terminal body weights and largest internal organs, compared to the other three DR-fed groups (Tables 3a, b, c). These changes reflected growth patterns similar to those observed in central body fat increases and somatic growth of non-fat tissues. These differences appear to reflect the effects of higher levels of energy intake and correlated with increased levels of growth promoting hormones (Growth hormone, IGF-1, prolactin, insulin) that were reported previously (Molon-Noblot et al., 2001, 2003). This resulted in a greater central body fat content, larger musculoskeletal growth and greater thoracic and abdominal organ sizes in the AL-fed animals. In contrast, no significant differences were seen between all the groups in the growth and absolute brain and testes weights of the AL-fed and moderately DR-fed animals. However, the 50% DR-fed group 4 animals did have slightly smaller brain and testes weights and growth than the other three groups. These observations indicate that the moderate measured feeding regimens (Groups 2 and 3) do not interfere with gross brain growth and development, but marked DR of group 4 does affect all organs’ growth (Tables 3a, b, c, 5, 6).

Compared to the AL-fed group, the absolute and relative weights of spleen, thymus, heart, kidneys, liver, adrenals, thyroids, ovaries, prostate, pituitary and pancreas were generally smaller and frequently statistically so by trend and pairwise statistical comparisons in the measured groups in a food intake or dose-proportional manner (Tables 3a, b, c, 5, 6). The degree of these differences between each group appeared generally proportional to their total food intake. Group 4 (50% DR) were the smallest animals and had the smallest organ weights. In most cases the lower thoracic and abdominal organ weights seen in the moderate and marked DR-fed groups correlated with a relative decreased incidence and/or severity of degenerative lesions in these animals.

Because of differences in body size, central obesity and thoracic and abdominal organ size observed across the groups, additional analyses were done on the relative organ weights expressed as a percent of body weights. Relative to the AL-fed animals, the measured groups had proportionally significant increases in the relative size of their hearts, kidneys, adrenals, livers, and other endocrine organ weights as a percent of body weights (data not shown). These changes generally reflected differences in metabolic and physiological stresses placed on these organs, the rate of body-weight gain and the differences in their carcass composition. As mentioned previously, differences in the absolute and relative sizes of brains and testes were not significantly different in the moderately DR-fed and AL-fed groups, but were lower in the 50% DR-fed group. However, these organs expressed as a percent of body weight were relatively increased in size proportional to the absolute size of their bodies and growth (data not shown).

Estrous Cyclicity

To determine the effects of aging and DR on estrous cyclicity and reproductive senescence, all females scheduled for the final necropsy, approximately 50 females per group, were evaluated throughout the two years of the study. The estrous cycles were monitored for approximately 20 consecutive days by vaginal lavage and microscopic examination beginning in study weeks 1, 18, 31, 43, 60, 71, 83, and 96. Because the dietary restrictions were implemented at the start of the study (at approximately 6 weeks of age), the effect of dietary restriction on the onset of puberty and ovarian cyclicity were not affected in this study. In study week 1, all groups of these 7–8-week-old females were cycling normally. They were also under “AL-feeding conditions” in both the AL-fed group 1 and in the moderate DR-fed groups 2 and 3, since the portion of food provided exceeded AL-food consumption at that age. Through the first six months of age, approximately 75% or greater of the animals in all groups exhibited patterns of regular estrus cyclicity. During study weeks 1 through 3, the average cycle length of marked DR-fed group 4 females was increased approximately 1 day relative to the average cycle lengths of the other 3 groups. Although the cycle lengths of the marked DR-fed group 4 females remained longer throughout 1 year of age, the magnitude of the difference lessened with time (Table 4 and Figure 11).

The onset of reproductive senescence in these unmated females was characterized by irregular estrous cyclicity in the AL-fed and moderate DR-fed groups, but was delayed in group 4 females under marked DR. The difference between cyclicity patterns in the 50% DR-fed group and those seen in the AL-fed and 2 moderate DR-fed groups was clearly established by 9 months of age, at which time regular estrous cyclicity was still evident in 70% of the marked DR-fed group 4 females compared with approximately 36% of the females in groups 1, 2, and 3. By 12 months of age this difference in regular estrous cycles was 43% in group 4 compared to approximately 12% in the other three groups. At 18, 21 and 24 months, approximately 25% of the females in group 4 still exhibited regular estrous cyclicity. None or very few of the females in the group 1 AL-fed animals or the group 2 and 3 moderate DR-fed females had regular estrous cycles at these ages. Throughout the 106 week study, the percentage of surviving females exhibiting patterns of regular estrous cyclicity in the AL-fed and moderate food restricted groups 2 and 3 were comparable (Table 4 and Figure 11).

Gross and Microscopic Degenerative Changes

Morphologic changes observed in all the tissues examined reflected the development of common tumors (to be reported separately) and proliferative and degenerative processes seen in control SD rats of this stock at comparable ages. However, there were clear differences between the AL-fed and the DR-fed groups fed the same diet. In general, there was a tendency for delay of onset and/or severity of common tumors and degenerative diseases in the DR-fed animals.

Morphologic Pathology

Because of the better survival observed in the food restricted animals, appropriate statistics were performed to reflect age adjusted differences between groups in mortality and lesion onset. As expected, the differences were most clear by pairwise and trend analysis for all the restricted groups in the common and fatal tumors, and will be reported separately. In addition, the AL-fed rats had the highest incidence and greatest severity of endocrine, renal, and cardiovascular lesions and tumors compared to the measured-fed groups. The lowest incidence and severity of degenerative lesions were seen in marked DR-fed animals. These rats had the least adverse effects observed in their hearts, kidneys, adrenals, pancreas, liver, musculoskeletal system and eyes. Females in group 4 did have the smallest ovaries and uteri, but these animals had structurally normal reproductive organs and had the highest proportion of females cycling normally beyond 1 year (Keenan et al., 1996). A brief summary of the major organ changes is discussed below along with stereologic data on selected organs.

Pituitary

The qualitative and quantitative changes observed in the pituitaries over time have been separately reported (Molon-Noblot et al., 2003). In summary, the AL-fed animals had the largest pituitaries, highest levels of prolactin and growth hormone secretion, and the highest incidence of focal hyperplasia and tumors of the anterior and intermediate lobes. The age-adjusted incidence of these tumors was statistically significantly decreased in the animals of each of the measured-fed groups. Pituitary tumors were the most common cause of death in all groups and both sexes.

Mammary Glands

The AL-fed animals had the highest incidence and earliest onset of degenerative lesions (i.e., galactoceles, etc.), hyperplastic lesions (lobular hyperplasias), and mammary gland tumors compared to the food restricted groups. These tumor data will be presented in detail separately, but in general the onset of mammary gland tumors and the number and size of masses in the mammary glands appeared highest in the AL-fed animals compared to their DR-fed counterparts. For DR-fed females a significant age-adjusted decreased incidence for mammary gland tumors was most obvious in group 4.

Pancreas

The detailed pancreas pathology data have been reported separately (Molon-Noblot et al., 2001). In summary, the DR-fed rats had an age-adjusted decrease in the incidence of islet cell adenomas. All of the DR-fed female and male groups had a lower incidence of focal islet cell hyperplasia than their AL-fed counterparts. The males in groups 2 and 3 had better survival than the AL-fed group, and had a higher incidence of smaller islet cell hyperplasias and adenomas, suggesting a delay in the onset of these lesions in the moderately restricted animals. Group 1, AL-fed animals had the highest incidence and severity of islet fibrosis, in addition to an increased age-adjusted incidence of islet cell tumors. These results and data on glucose and insulin levels were discussed in detail separately (Molon-Noblot et al., 2001).

Adrenals

The AL-fed animals had the largest adrenal glands (Table 3b) and the highest incidence and/or severity of cortical cystic degeneration, focal cortical hyperplasia and cortical hypertrophy. In general, there was an increase in the absolute size of the adrenal glands proportional to food intake and body size. Group 4 animals had the smallest absolute weight of adrenals, and the lowest incidence of degenerative and proliferative lesions in the cortex and medulla. However, relative adrenal weights expressed as a percent of body weight, demonstrate the DR-fed rats had larger adrenals than the AL-fed rats proportional to their body size. These observations, combined with the lower incidence and/or severity of degenerative proliferative lesions in the DR-fed animals indicate adrenal function was likely to be better preserved in DR-fed rats, as has been reported in other studies of the effects of DR on adrenal function (Masoro, 1996; Levin et al., 2000).

Heart

The AL-fed group, particularly the males, had the largest hearts and the highest incidence and severity of cardiomyopathy. This common cardiovascular disease was the third most common cause of death in the AL-fed male rats. The lesions comprising this process were graded separately into myocardial fibrosis, myocardial degeneration and cellular infiltrates. Stereological evaluation of the hearts of both sexes of all groups from the interim and final necropsies confirmed the subjective histological data and indicated an earlier onset and progression of the lesions from the first to the second year (Table 5). Statistical analysis of absolute and relative heart weights and stereological myocardial fibrotic index (%) demonstrated a clear dose response pattern for most of the parameters. Absolute and relative heart weights in group 1 AL-fed rats of both sexes were significantly different at all times from other groups. Pairwise comparisons distinguished group 1 and group 4 from the moderate restricted groups 2 and 3 at weeks 13, 26, 53, and 106 weeks of the study (Table 5). Group 4 rats were different from all other groups for most parameters at all time points. By the 106-week necropsy, the fibrotic index was significantly different between the group 1 AL-fed animals of both sexes from all other food restricted groups. These data were consistent with the overall trends and incidence and severity of cardiomyopathy seen in this study and comparable studies (Keenan et al., 1994a, 2000a; Roe et al., 1995; Kemi et al., 2000; Faine et al., 2002; Wan et al., 2003). Cardiomyopathy contributed to the co-morbidity seen in many animals when tumors or renal disease were determined to be the primary cause of death.

Kidneys

The AL-fed animals had the largest kidneys and the highest incidence and severity of renal disease. Chronic nephropathy was the most common cause of death in the AL-fed rats of both sexes following tumors of the pituitary and mammary glands. The lesions comprising chronic nephropathy were subjectively graded separately and included glomerulosclerosis, cellular infiltration, tubular basophilia and tubular dilatation. Stereological analyses of these lesions have been published separately (Keenan et al., 2000b). In summary, the marked DR-fed group 4 animals had the smallest and least adversely affected kidneys. The AL-fed rats had the largest and most diseased kidneys, and these findings correlated with glomerular and total nephron hypertrophy, the early development of glomerular sclerosis and other degenerative changes in the tubules and the interstitium (Keenan et al., 2000b). These data were consistent with large nephrons undergoing chronic persistent metabolic stress, injury and leading to chronic nephropathy. Renal disease was also an important co-morbidity factor that contributed to the morbidity and the overall mortality in the AL-fed group. In the AL-fed rats with severe or fatal nephropathy, secondary lesions such as uremic gastric mineralization, diffuse parathyroid hyperplasia, and fibrous osteodystrophy of the bones were frequently observed. The uremic lesions were not observed in the moderate and marked DR-fed groups. These data demonstrated that the early events in the development of chronic nephropathy occur as early as 14 study weeks, with measurable increases in kidney size, glomerular hypertrophy, hypertrophy of the entire nephron and a loss of renal function in response to the metabolic overload from increased food intake and rapid growth (Keenan et al., 2000b). Glomerular hypertrophy in the AL-fed rats appeared to peak by study week 26 and then was followed by increasing severity in the progression of glomerular sclerosis, tubular basophilia, interstitial inflammation and increased BrdU tubular and interstitial cell labeling (Keenan et al., 2000b). These changes were gradual and progressive, and correlated with a steady decline in renal function as measured by clinical biochemistry, creatinine clearance, urinalyses and urine protein electrophoresis, as reported (Keenan et al., 2000b).

Liver

The AL-fed animals had the largest livers and the highest incidence of hepatocellular degenerative lesions and proliferative changes. These changes reflect the high metabolic load on the livers of the AL-fed animals. The hepatic parameters measured were favorably changed proportional to the degree of food restriction. Cell proliferation and stereological evaluation indicated the density of hepatocyte nuclei (nuclei per cm³) was not statistically different among groups, indicating individual hepatocyte number and volume per unit area were similar between all groups. Therefore, the total hepatocyte nuclei per liver was proportional to absolute liver weight and body weight. The AL-fed animals of both sexes had the largest livers and, therefore, the largest total number of nuclei per liver. The marked DR-fed group 4 rats of both sexes had the smallest livers and the least total number of nuclei per liver. The 7-day cumulative Percent BrdU LI and total number of BrdU-labeled nuclei per liver were not statistically different among the different dietary groups by comparison to the AL-fed rats (Table 6). These data differ from those of other studies (Grasl-Kraupp et al., 1994) on the effects of DR on rodent liver cell proliferation, but are similar to our findings in earlier studies (Keenan et al., 1994a, 1995). Thus, decreased relative or absolute 7-day hepatocyte proliferation rate in DR-fed SD rats does not appear to occur after adult liver size is established.

While no differences were seen in liver tumor incidence between the different dietary groups compared to the AL-fed group (to be reported separately), degenerative changes, such as hepatocellular periportal vacuolation and telangiectasis were most evident and severe in the AL-fed groups compared to the DR-fed groups, although most of the DR-fed animals lived for a significantly longer time than their AL-fed counterparts. In animals with hepatocellular periportal vacuolation, particularly females, there was a relative increase in BrdU nuclear labeling of hepatocytes in that region, but not in the total BrdU labeling index. Other proliferative changes, such as bile duct hyperplasia, occurred at a similar incidence in the AL-fed and DR-fed groups but were of greater severity in the AL-fed animals. Basophilic and eosinophilic altered hepatocellular foci were seen in all groups with a similar incidence and grade in the AL-fed and moderate DR-fed groups, but a lower incidence and grade in the marked DR-fed group 4. These data were consistent with previously reported morphological studies and measures of metabolic and oxidative stress (Laganiere and Yu, 1989a, 1989b; Yu et al., 1989; Grasl-Kraupp et al., 1994; Keenan et al., 1995b; Hikita et al., 1999).

Male Reproductive Tract

Pathological and organ weight changes in the male reproductive tract indicated that there were little differences between the absolute and relative percent of brain weights of the testes across the 4 groups. The AL-fed males did have earlier onset and a higher incidence and severity of seminiferous tubular degeneration. These data indicate a preservation of the testicular histology and no observable adverse effect in the seminiferous tubules or the epididymis of the moderate and marked DR-fed males. These data are consistent with reproductive studies of breeding male SD rats that indicate that no adverse effect on fecundity is seen in moderately restricted SD male breeders (Mattson, unpublished data).

No difference was seen in the incidence of testicular tumors (to be reported separately), but the incidence of focal interstitial cell (Leydig cell) hyperplasia in AL-fed males was similar to each of the DR-fed groups, even though the DR-fed males lived for a considerably longer time.

While there were no differences in testes weight among the groups, the weight of the prostate glands was significantly decreased by trend and pairwise comparisons at each interim and at the final necropsies in all the DR-fed groups. In the AL-fed males the incidence of chronic prostatitis was higher and graded as more severe. The incidence and grades of epithelial hyperplasia in either the dorsolateral lobe or the ventral lobes of the prostate were of similar incidence across the AL-fed and DR-fed groups. However, the one adenoma of the dorsolateral lobes of the prostate was seen in this study.

The other male accessory sex glands (seminal vesicles, coagulating glands and bulbourethral glands, etc.) were generally observed as being much larger in the AL-fed males with a higher incidence of inflammatory or degenerative changes observed compared to the DR-fed male groups. These accessory sex glands in the moderate and marked DR males, while smaller, were not atrophic and had very few lesions.

Female Reproductive Tract

The group 4 (50% DR-fed) females had the statistically smallest ovaries and uteri throughout the study compared to other groups; however, at the 1-year time point, 43% of these animals were cycling normally while the other 3 groups had only 12% with normal estrous cyclicity. The ovaries of the group 4 females were morphologically smaller, and tended to have fewer observable follicles and fewer corpora lutei with a greater proportion of interstitial cells. However, this group had the lowest incidence of ovarian atrophy over the course of the study into the second year. The uterine body and horns of the group 4 50% DR-fed females, while smaller than all other groups, were histologically normal and had normal estrous cyclicity, that indicated they were functioning normally. While these females had the lowest percent body fat, about 88% were still cycling normally at 6 months and 43% were cycling normally at 1 year. In contrast, the AL-fed females and the other 2 moderate DR-fed female groups were nearing reproductive senescence at 12 months. By 18, 21, and 24 months approximately 25% of the Group 4 marked DR-fed females had regular cycles. This group had the lowest incidence and severity of degenerative or proliferative lesions in their ovaries (ovarian atrophy) and uteri (i.e., endometrial stromal hyperplasia and tumors). None or few of the females in groups 1, 2, and 3 had regular cycles during the second year and they all had a high incidence and severity of ovarian and uterine lesions consistent with reproductive senescence. These data indicate moderate DR does not delay reproductive senescence in the SD rat, but marked DR does extend the period of normal estrous cycles in unbred female SD rats.

Eyes

No ophthalmoscopic or histological differences were seen between the AL-fed and 2 moderate DR-fed groups over the course of this study. However, the group 4 (50% DR) rats of both sexes had the lowest incidence and least severe lesions of corneal dystrophy (multifocal linear or pinpoint areas of mineralization in the corneal basement membrane) that correlated with the significantly lower clinical incidence of corneal opacities observed ophthalmoscopically, and were consistent with our previous studies (Hubert et al., 1997). In addition, the marked DR-fed group 4 animals also had the highest incidence of retinal atrophy, which was seen ophthalmologically as retinal hyperreflexia. The marked DR-fed (50%) groups seemed uniquely susceptible to this change. This degree of retinal lesions was not seen in the AL-fed or moderate DR-fed groups, and could not be related to cage location or relative exposure to light in the animal rooms. Detailed analysis of the corneal opacities, retinal atrophies and other ocular lesions in all 760 rats observed in this study, and its modulation by moderate and marked food restriction will be reported separately.

Tails

No differences were seen in the incidence or severity of tail lesions between AL-fed and the moderate DR-fed groups. However, the marked DR-fed group 4 rats of both sexes had the highest incidence of terminal focal necrosis of the tail tip (1–2 cm). This lesion was observed over the course of study clinically at a high incidence (~ 40%) at the interim and final necropsies. This lesion was characterized by distal vascular thrombosis, infarction and necrosis seen in the terminal 1–2 cm of the tails of the marked DR-fed animals affected with this minor lesion. The lesion may reflect an attempt by the smaller rats with larger surface areas in well-ventilated steel wire cages to alter their thermoregulation and maintain their core temperature by peripheral vasoconstriction when under marked energy restriction.

Musculoskeletal System

Age-dependent degeneration and atrophy of the skeletal muscle, referred to as sarcopenia, and degeneration of the peripheral nerves were observed with increasing incidence and severity in the AL-fed rats, particularly the males. While the skeletal muscles and bones were smaller in the moderate and marked DR-fed animals, a much lower incidence and severity of skeletal muscle degeneration and atrophy was evident than that seen in their shorter-lived, larger AL-fed counterparts. The smaller skeletal muscle mass and bone size of the DR-fed groups developed fewer degenerative changes over time (Vermorel et al., 1998). Metabolic bone changes (fibrous osteodystrophy secondary to uremia and renal hyperparathyroidism) were only seen in AL-fed animals with severe renal disease (Table 7).

Another clinical observation was that the aging, obese AL-fed rats were much less active during the daily observation periods and dosing and during random observation period during the dark 12-hour cycle. All the DR-fed animals had much greater spontaneous activity levels than the AL-fed animals during the light and dark cycles.

Adult-onset Type 2 diabetes and obesity in humans is associated with a high incidence of plantar pressure ulcers and peripheral neuropathy. Therefore attention was given to changes in the feet and peripheral nerves that might have increased risk for plantar ulcers under the metatarsal bones due to increased pressure on insensitive skin on the wire cage floors. The incidence and severity of peripheral sciatic nerve degeneration were increased as they aged. However, if peripheral neuropathy was a factor in plantar ulcers, it was difficult to make a correlation due to the DR-fed rats’ longer life spans and the longer time they had contact with the wire cage bottoms. This resulted in differences between groups in lesion severity, but not in lesion incidence (Table 7).

Plantar ulcerative granulomas and dermatitis were observed in both sexes in groups 1, 2, and 3 and in only a few rats in group 4. These foot lesions were generally not found until the rats had been housed in the wire bottom cages for more than one year, as noted by others (Peace et al., 2001). The incidence of these lesions in the moderate DR-fed males in groups 2 and 3 was twice that of the group 1 AL-fed males. The severity of these lesions in the AL-fed animals correlated with underlying osteoarthritis of the tarsal joints in most of the affected animals. The higher incidence of plantar granulomas in the group 2 and 3 males appeared to be a function of their improved survival and duration on study compared to the AL-fed group 1 males. However, this trend was not seen in the females of groups 2 and 3 or both sexes in group 4. In spite of a higher incidence all of the DR-fed groups had a much lower severity of plantar chronic or ulcerative dermatitis and the less severe osteoarthritis of the tarsal joints (Gefen, 2002; Mueller et al., 2003).

In contrast, the highest incidence and/or severity of osteoarthritis of the stifle joints was seen in both sexes of group 1 AL-fed rats. Both sexes of groups 2, 3 and 4 had a reduced incidence and/or severity of stifle joint lesions, with group 4 having the mildest changes (Table 7). Therefore, a clear increased incidence and severity of osteoarthritis of the stifle joints of the male AL-fed rats was observed relative to all the DR-fed groups. This was likely to be biologically significant since the DR-fed groups had much longer life spans than the AL-fed animals, and thus were in contact with the wire bottom caging for the longest period.

Discussion

The results of this study demonstrate that the Charles River outbred SD rat stock [Crl:CD(SD) IGS BR] when AL-fed a commercial balanced diet such as Purina Rodent Diet, develops a profile of degenerative diseases and a syndrome of adult-onset obesity that progresses to a metabolic syndrome similar to that seen with human, polygenic adult-onset type 2 diabetes and obesity (diabesity). This syndrome is characterized by the chronic development of hyperlipidemia, hyperinsulinemia, changes in glucose metabolism, the development of chronic renal disease, cardiovascular disease, and degenerative changes in the weight-bearing joints, liver, pancreas, adrenals, thyroids and other endocrine organs. The multiple co-morbidities observed in these animals were readily manipulated by controlled DR feeding, and thus provide an excellent model to study experimental modulations of the diabesity syndrome by control of caloric intake in a manner similar to that observed in human beings undergoing dietary therapy for obesity and type 2 diabetes (Whitaker et al., 1997; Bray, 2002; Konstantinov, 2003).

A World Health Organization (WHO) 2002 report identified the main global risks affecting human disease, disability, and death rates (Eckel et al., 2002; Konstantinov, 2003). The WHO found that among the top 10 risks accounting for 40% of the worldwide deaths, excessive weight and obesity were listed 10th, and hypertension, elevated cholesterol, and inactivity were numbered 3rd, 7th, and 14th respectively, and all these conditions are associated with diabesity. It is estimated that 1.1 billion people globally are overweight or obese. In the USA, adult obesity rates rose from 14% in 1978 to 31% of the population in 2002. In the UK, adult obesity rates rose from 6% in men and 8% in women in 1980 to 21% of men and 24% of women in 2002. The WHO 2002 World Health Report estimated over 2.2 million deaths per year worldwide were over weight-related, with 220,000 per year in Europe and over 300,000 per year in the USA. In addition, obesity related health risks among Asians have been rising with an estimate that a significant portion of the 3.6 billion Asian population already has an excessive body mass index (BMI). Thus obesity is prevalent in both developed and developing countries, and is also reaching epidemic proportions in the children of these populations. The current public health epidemic of diabesity is related to excessive food (caloric) intake, and is due to behavioral patterns including decreased physical activity and over-consumption of high fat, energy-dense foods. As with many species of animals, many humans become obese because of a biological predisposition to readily gain weight when food is available, in preparation for unfavorable environmental conditions when food is less available or energy needs are extreme. The worldwide prevalence of persistent obesity in developed countries has resulted in many serious sequelae in human beings, including type 2 diabetes, heart disease, hypertension, stroke, osteoarthritis of weight-bearing joints, many forms of cancer, a poor quality of life and an excess of premature deaths. In an extensive study of American men and women, both increased obesity and reduced exercise were shown to be strong and independent predictors of early death (Hu et al., 2004). The WHO predicts that the economic burden and medical complications of diabesity in human beings threaten to overwhelm health services, and the impact on morbidity and mortality in people soon may overtake that of tobacco products. All of these unfortunate sequelae can be prevented or reduced by control of food intake resulting in weight loss, even by modest weight loss. Furthermore, in North America, losing weight is estimated to prevent 1 in 6 cancer deaths, or more than 90,000 per year. It is estimated that excessive weight and obesity in the United States could account for 14% of all cancer deaths in men and 20% in women (Brunner et al., 2001; Bray, 2002; Eckel et al., 2002; Konstantinov, 2003). Since the human populations suffering the sequelae of diabesity are heterozygous populations, this phenotype is clearly the result of polygenic trait loci that interact in complex ways with environmental factors, but can be readily controlled and managed by behavioral and exercise therapy and moderate dietary or caloric restriction of food intake.

Type 2 diabetes mellitus is strongly associated with obesity. Over 80% of human patients with Type 2 diabetes are overweight or obese. The risk of developing diabetes increases in an exponential manner with increasing BMI. Even moderate weight excess is considered a significant risk factor for multiple diabetes co-morbidities (Whitaker et al., 1997; Brunner et al., 2001; Eckel et al., 2002; Konstantinov, 2003; Pasquale et al., 2003; Rauser et al., 2003; Hu et al., 2004; Lazar, 2005). The adverse effects of such excess weight on blood glucose control are attributed to insulin resistance, and obese patients with diabetes have a higher prevalence of other risk factors, including cardiovascular and renal disease, hypertension and dyslipidemias. This segregation of risk factors known as metabolic syndrome, insulin-resistant syndrome, or syndrome X, explains the very high cardiovascular and renal morbidity and mortality rates in the population of humans and rodents with adult-onset type 2 diabetes and obesity, or diabesity (Whitaker et al., 1997; Brunner et al., 2001; Eckel et al., 2002; Leiter, 2002; Axen et al., 2003; Konstantinov, 2003). In humans it appears that obese individuals prone to develop defective blood glucose control and insulin secretion will develop overt type 2 diabetes due to a complex interaction of genetic predispositions and environmental factors of which uncontrolled food intake inducing excessive body weight gain is a key factor. Current management of obese diabetic patients emphasizes lifestyle modification as the cornerstone of both the prevention and the treatment of the disease process. In a large epidemiological study of over 80,000 American nurses, controlled dietary intake and lifestyle changes such as moderate exercise were associated with a lower risk of Type 2 diabetes, and increases in body weight were the strongest predictor for the development of the disease and its co-morbidities (Whitaker et al., 1997; Eckel et al., 2002; Konstantinov, 2003; Hu et al., 2004).

Obesity and Type 2 diabetes are known to be a polygenic syndrome in heterozygous human populations, and more polygenic rodent models are needed in which genetic trait loci can be identified for their interactions with each other, and with the environmental factors that elicit obesity syndromes and lead to diabetes (Masoro, 1996; Keenan et al., 2000a, 2000b; Leiter, 2002; Axen et al., 2003; O’Rahilly et al., 2005; Park and Prolla, 2005). Although monogenic obesity mutations in rodents such as those in leptin or its receptor have been extensively used to test anti-obesity and anti-diabetic drugs, the monogenic basis of these rodent obesity models is not reflective of the more common forms of human obesity that have been documented to be polygenic in origin (Whitaker et al., 1997; Bray, 2002; Konstantinov, 2003). Attempts to use recombinant congenic strains of rats and mice have been useful for analysis of the human polygenic syndromes (Leiter, 2002; Reifsnyder et al., 2002; Koubova and Guarente, 2003; O’Rahilly et al., 2005; Park and Prolla, 2005). In studies of single genes that contribute a major portion of the variants in a phenotypic trait, a congenic strain will be adequate. However, this approach is clearly inadequate when considering that a complex phenotype representing a variable collection of quantitative trait loci individually may make small contributions to the trait, but interact in complex ways with the diet and environment to bring about the disease phenotype. Thus, the approach in rodents of studying a single congenic region may not recreate the phenotype or even the subphenotype of a complex syndrome such as diabesity. Numerous obesity quantitative trait loci have been identified in mice, rats and humans that have aided in the understanding of the polygenic nature of this disease syndrome (Levin et al., 1997; Leiter, 2002; Reifsnyder et al., 2002; Konstantinov, 2003; Koubova et al., 2003; O’Rahilly et al., 2005; Park and Prolla, 2005). However, these reductionist approaches are unlikely to yield predictive outcomes for the development of anti-obesity and anti-diabetic drugs tested for efficacy or safety in susceptible human populations with the polygenic syndrome.

Approximately 1 in 5 U.S. adults have metabolic syndrome (syndrome X or diabesity) which is manifest as a cluster of metabolic abnormalities identified by three or more changes including abdominal obesity, hypertriglyceremia, low HDL cholesterol, high blood pressure and high fasting glucose (Whitaker et al., 1997; Brunner et al., 2001; Eckel et al., 2002; Konstantinov, 2003). The potential morbidity and mortality consequences of this polygenic syndrome in the general population cry out for a reliable polygenic rodent model to better define the key factors involved in the development of the disease, and importantly, predict the effectiveness of therapeutic interventions that may prevent or treat the syndrome of diabesity. Therefore, targeting the metabolic syndrome as a whole rather than specifically dissecting each one of its subcomponents should be considered as a key objective in the development of models to test treatments and management programs of this important human syndrome.

In this regard, the Charles River outbred SD rat appears to represent an excellent polygenic model that is only now beginning to be thoroughly phenotyped and understood from the perspective of an animal model of diabesity (Gumprecht et al., 1993; Klinger et al., 1996; Keenan et al., 1997, 1994a, 1994b, 1995, 1996, 1999, 2000a, 2000b; Hubert et al., 1997, 2000; Laroque et al., 1997; Levin et al., 1997, 2000; Hoe et al., 1998; Vermorel et al., 1998; Molon-Noblot et al., 2001, 2003; Kemi et al., 2000). While not all of the phenotypic and genetic characteristics of this stock have yet been well defined, the results of this study and previous work indicate that when AL-fed this animal clearly has a well-developed set of phenotypic characteristics that would be best described as diabesity. This stock has a well-developed predisposition to develop obesity when fed a variety of diets ad libitum (Keenan et al., 1995, 1997, 2000a; Kemi et al., 2000). It subsequently develops a dramatic, adult-onset dyslipidemia with high levels of triglycerides and total cholesterol, and glucose imbalances with high persistent insulin levels and other evidence suggesting insulin and leptin resistance (Gumprecht et al., 1993; Keenan et al., 1994a, 1994b; Levin et al., 2000; Molon-Noblot et al., 2001). The early dysfunction of its endocrine system due to AL-overfeeding, including functional and pathologic changes in the pancreatic islets, pituitary, adrenals, thyroid and other organs indicates an early onset of the endocrine disruption syndrome described as diabesity (Keenan et al., 1994a, 1994b; Capen, 1996, 2001a, 2001b; Keenan et al., 2000a, 2000b; Molon-Noblot et al., 2001; Axen et. al., 2003). With the development of new genetic technologies and other molecular methods to describe this syndrome at a genetic and molecular level, it would appear that this stock of rats is an ideal candidate to dissect the multifactorial genetic and phenotypic aspects of this syndrome and test the effects of pharmaceutical treatments on the whole animal syndrome.

The results of this and previous studies have demonstrated that ad libitum overfeeding of numerous commercial and experimental diets is the most significant uncontrolled variable affecting the outcomes of the current rodent bioassay used in modern safety assessment (Gries et al., 1982; Ross et al., 1983; Berry, 1986; Chapin et al., 1993a, 1993b; Gumprecht et al., 1993; Roe et al., 1995; Keenan et al., 1994a, 1994b, 1995a, 1995b, 1996, 1997, 1999, 2000a, 2000b; Turturro et al., 1995; Capen, 1996; Chiu et al., 1996; Dixit et al., 1996; Hubert et al., 2000; Capen, 2001a, 2001b; Faine et al., 2002; Haseman et al., 2003; Hursting et al., 2003). There is a high correlation between uncontrolled AL food consumption, rapid onset of body weight gain, endocrine disruption of the pituitary axis, obesity and its co-morbidities and low survival currently seen in control laboratory rodents, particularly this stock of SD rats (Seki et al., 1997; Simpkin et al., 1977; Ross et al., 1983; Mietes, 1990; Breese et al., 1991; Sonntag et al., 1994; Roe et al., 1995; Chiu et al., 1996; McShane et al., 1996; Sonntag et al., 1999; Levin et al., 2000; Small et al., 2002; Wan et al., 2003; Roy et al., 2003) Moreover, data on moderate DR-feeding indicate that initial weanling body weight is not the most important determinant factor in survival of this stock (Laroque et al., 1997). These data demonstrate the genetic predispositions of the SD rat can be managed with appropriate control of dietary and environmental conditions. It appears that the key extrinsic factor of excessive food intake (caloric intake) determines the adult body weight and the degree of obesity achieved, and the onset and progression of diet-related degenerative disease and tumors that determine the animal’s final longevity (Ross et al., 1983; Laroque et al., 1997). These data indicate that the genetic predispositions of these SD rats to the diabesity syndrome are not sufficient to explain all the adverse effects of AL-overfeeding or the beneficial effects of controlled DR-feeding on the SD rat’s longevity. Rather, it is the complex interactions of the environment and diet (total caloric intake) with the SD rat’s genome that produces or prevents the polygenic phenotype of diabesity from being expressed. It also appears that feeding modified diets with lowered protein, reduced metabolizable energy content, and increased fiber does not necessarily result in the control of diabesity or improved survival if the diets are provided ad libitum (Tucker, 1979; Iwasaki et al., 1988; Masoro et al., 1989; Keenan et al., 1994a, 1994b, 1995a, 1995b, 2000a, 2000b; Roe et al., 1995; Yu, 1995; Masoro, 1996; Whitaker et al., 1997; Masoro et al., 1996). Only dietary restriction (DR) of all diets tested in rodents consistently improves survival and delays the onset of the spontaneous development of obesity and the degenerative diseases referred to as “diabesity.”

While moderate DR does result in a reduced onset of spontaneous degenerative diseases and tumors, the method only seems to delay the onset, rather than reduce the incidence of diet-related tumors in SD rats (Keenan et al., 1994a, 1994b, 1996, 1997, 1999, 2000a, 2000b; Hubert et al., 2000; Kemi et al., 2000; Molon-Noblot et al., 2001, 2003). For example, there has been a well-demonstrated age-adjusted decreased incidence of pituitary and mammary gland tumors in moderate DR-fed SD rats, but the effect seems to be a delay in tumor onset and growth in the moderate DR-fed groups (Keenan et al., 1994a, 1994b, 1995a, 1996). Conversely, many of the degenerative kidney and heart disease processes can be delayed or completely prevented by the judicious application of dietary restriction to control growth and to modify in a healthful way the animal’s physiological status (Gumprecht et al., 1993; Hubert et al., 2000; Keenan et al., 2000b; Kemi et al., 2000; Faine et al., 2002,).

This study confirms previous reports that DR-fed SD rats eat approximately the same or slightly more food per gram per day body weight as the AL-fed animals (Keenan et al., 1997). Thus, the AL-fed rats are disproportionally partitioning the nutrients and energy consumption per gram lean body weight and are storing foodstuffs as excess visceral and subcutaneous adipose tissue without the normal utilization of lipid and protein foodstuffs for maintenance. Utilization of nutritional food, while essential for life, may have long-term, low-intensity negative consequences when not used efficiently or provided in excess. The use of oxygen and the oxidative metabolism of fuels results in free radical production, which in excess, can be very damaging (Laganiere et al., 1989a, 1989b; Dixit et al., 1996; Weindruch, 1996; Zainial et al., 2000; Duffy et al., 2001; Lowell et al., 2005; Park et al., 2005). Glucose, like other nonreduced sugars, undergoes a nonenzymatic reaction with amino groups of proteins called the glycation reaction, and thus fuels such as glucose can become reactive molecules in their own right (Yu, 1995; Masoro, 1996; Lowell et al., 2005). These basic metabolic processes occur during the utilization of nutritional food and have led to the major metabolic hypotheses that implicate free radicals, the glycation reaction, and/or the Maillard reactions. These metabolic changes in turn induce neuroendocrine per-tubations in hormones controlling these factors, and lead to the multi-organ adverse effects of AL-overfeeding that result in diabesity. The preventative effects seen in the DR-fed SD rat model provide the best support for these ideas, and demonstrate that controlling energy intake alone is in fact the most efficient and effective modulator of the adverse effects of both free radical metabolism, glycation processes and disruption of normal neuroendocrine mechanisms related to somatic growth and energy metabolism (Gries et al., 1982; Lee et al., 1990, 1999; Mietes, 1990; Breese et al., 1991; Merry et al., 1994; Sonntag et al., 1994; Chiu et al., 1996; Sonntag et al., 1999; Levin et al., 2000; Hursting et al., 2003; Rajala et al., 2003; Roy et al., 2003; Wan et al., 2003; Weindruch, 2003).

AL-fed and DR-fed rats appear to have similar rates of oxygen consumption per unit lean body weight mass because the metabolic body mass of the DR-fed animals is reduced. It appears that the DR-fed animals reset their energy utilization mechanisms to more effectively utilize glucose and control insulin secretion (Yu, 1995; Masoro, 1996; Keenan et al., 1997). However, when calculated on a lean body mass basis, there is a slight reduction in the averaged metabolic energy (ME) intake per gram of lean body mass in the DR-fed rats (Keenan et al., 1997, 1999, 2000a, 2000b). These data suggest that subtle alterations in metabolic rate may be a factor in energy utilization. These data do indicate that it is the total reduction of energy intake per animal that appears to be the main factor that prevents the development of diabesity rather than a general effect on the long-term rate of fuel use (Yu, 1995; Masoro, 1996; Weindruch, 1996; Weindruch et al., 1997; Lowell et al., 2005).

Comparisons of different levels of DR to AL food intake in this model indicate that the metabolizable energy (ME) intake was comparable to the predicted maintenance energy requirements calculated from standard equations (Keenan et al., 1997). The actual ME intake for SD rats fed AL or DR was only slightly greater than the predicted maintenance ME requirements calculated for all the groups except for the most marked DR-fed animals in group 4. In contrast, the actual ME intake of AL- and DR-fed male rats was actually slightly less than the predicted maintenance energy requirements, with the greatest differences occurring in the 50% AL-fed group (Keenan et al., 1999). These and other data indicate that the AL-fed and moderate DR-fed groups of SD rats have remarkably similar maintenance energy requirements although they partition energy very differently into fat mass and lean body mass.

DR-fed rats demonstrate increases in their metabolic efficiency. It has been reported that there are significant differences in feeding activity, drinking activity, water consumption, core body temperature, oxygen consumption, carbon dioxide production and motor activity as measured in several strains of rats fed 40% DR compared to AL-fed rats (Turturro et al., 1995). Food restriction can affect metabolic efficiency primarily by its shifting the utilization and synthesis of specific nutrients from one source to another. Significant shifts in respiratory quotient (RQ) in DR-fed rats, but not AL-fed rats suggests that DR-fed animals make rapid changes during the day in their metabolic pathways depending on food availability. The decrease in RQ observed in DR-fed rats prior to feeding indicates that glycogen reserves have been utilized and the DR-fed rats are metabolizing protein and fats. After feeding, the DR-fed rats demonstrate a rapid rise in RQ which indicates they shift to carbohydrate metabolism (Turturro et al., 1995; Keenan et al., 1997). In contrast, the AL-fed rats maintain a constant RQ throughout the day indicating a constant ratio of protein, lipid and carbohydrate metabolism (Turturro et al., 1995; Keenan et al., 1997). The gradual drop in RQ observed in DR-fed rats suggests that they have smaller glycogen reserves stored in the liver, and thus limit the amount of time that they are dependent on carbohydrates for energy (Turturro et al., 1995; Keenan et al., 1997). These events help explain why the DR-fed animals store less energy as white adipose tissue and why the AL-fed animals store large amounts of energy as excessive visceral body fat. It is apparent that DR-feeding provides the animals with a more diverse utilization of different nutrients (fat, protein, carbohydrates) throughout the day as they metabolize carbohydrates prior to feeding and metabolize proteins and fats following feeding. This daily shift from protein and fat metabolism to carbohydrate metabolism apparently maintains better glycemic control and prevents the development of diabesity in these otherwise healthy, reproductively effective SD rats.

The large adipose mass that develops in the AL-fed animals may be dysfunctional as a metabolic and endocrine tissue. Adipose tissue has evolved complex mechanisms to efficiently store energy in times of natural caloric restriction, a survival function negated by the caloric excesses of AL-feeding. The adipose tissue is responsive to both central and peripheral metabolic signals and can secrete a number of adipocyte-specific proteins, called adipokins (Konstantinov, 2003; Rajala and Scherer, 2003; Lazar, 2005; Lafontan, 2005; Schwartz and Porte, 2005). With obesity, adipocytes increase in cell number and size, but the metabolic and secretory profiles of the larger adipocytes in the obese fat mass alter and become disregulated as a metabolic and endocrine tissue (Rajala and Scherer, 2003; Yildiz et al., 2004; Lazar, 2005; Schwartz et al., 2005,).

The secretory products of adipocytes now include molecules regulating energy homeostasis (i.e., leptin, adiponectin, resistin), molecular regulators of acute phase reactant and inflammatory responses (i.e., α 1 acid glycoprotein, SAA), molecules regulating the innate immune system (i.e., TNF α, IL-6), molecules regulating vasculature (i.e., VEGF, monobutyrin), matrix components (type IV collagen), in addition to their metabolic enzyme systems for bioconversions (Rajala and Scherer, 2003). In obesity, leptin levels increase and act on the hypothalamic arcuate nucleus to regulate food intake, but are not able to prevent overconsumption or obesity in heterozygous populations of humans or animals (Harrison et al. 1987; Weindruch et al., 1997; Konstantinov, 2003; Rajala et al., 2003). Leptin also alters regulation of hormones in the hypothalamus-pituitary-adrenal axis and affects the secretion of growth hormone, prolactin and other hormones of the pars distalis (Mietes et al., 1990; Sonntag et al., 1999; Levin et al., 2000; Hursting et al., 2003; Konstantinov, 2003; Molon-Noblot et al., 2003; Rajala et al., 2003). In addition, obesity results in disruption of the anorexigenic (appetite-inhibiting) effects of adipocyte leptin on the neuroendocrine network. Ghrelin, a stomach derived hormone, is an agonist for the pituitary hormone secretagogue receptor and is an orexigenic (appetite-stimulating) hypothalamic signal that is antagonistic to leptin by its actions on neuropeptides that regulate feeding behavior and stimulate growth (Inui et al., 2004; Schwartz and Porte, 2005). The patterns of secretion of these hormones have been shown to differ between lean and obese humans in diurnal pariation and synchronicity (Yildiz et al., 2004). These patterns have not been determined in the SD rat, nor has their role in regulation of body weight and energy homeostasis been resolved.

The adipocyte is also a source and target for inflammation. Obesity results in increased adipocyte production of TNF α. Increases in adipocyte TNF α secretion result in increases in systemic C-reactive protein (CRP) that is associated with obesity, insulin resistance (diabetes type 2) and cardiovascular disease (Rajala and Scherer, 2003; Lafontan, 2005). In addition, adipocytes in obese animals can produce excessive acute phase reactants, including IL-6, α-1 acid glycoprotein and serum amyloid A (SAA) associated with degenerative processes of aging, obese humans and animals.

Obesity is also associated with increased risks of cancer in humans and animals. The increased secretion of adipocyte cytokines plays an active role in the inflammatory state associated with obesity, and adipocyte production of type IV collagen may be critical in tumor progression (Bray, 2002; Konstantinov, 2003; Rajala and Scherer, 2003). In addition, the increase in adipose tissue stroma production of aromatase results in conversion of adrenal gland androstenedione into estrone, a source of estrogenic compounds in reproductive senescence in obese women and animals (Kritchevsky, 1999; Bray, 2002; Small et al., 2002; Konstantinov, 2003; Lafontan, 2005). Since the rate of estrone production is related to the fat mass, it may contribute significantly to the cancer risks for endometrial and mammary gland cancers and early development of reproductive dysfunction due to the abnormal amounts of estrogenic compounds produced in obese individuals (Bray, 2002; Small et al., 2002). In addition, DR-fed rodents have an induction of the expression of SIRT1 deacetylase in numerous tissues, including visceral fat pads, and this causes the sequestration of the apoptotic factor BAX from mitochondria, preventing stress-induced apoptotic cell death and extending life (Lu et al., 1991; Cohen et al., 2004; Rhodes, 2005). Thus, obesity due to AL-overfeeding may be a causative factor in the large number of spontaneous metabolic, endocrine, reproductive, cardiovascular and neoplastic complications seen in diabesity, and DR-feeding induces healthful modulations.

The onset of dyslipidemia and the development of visceral obesity are dramatic in the AL-fed animals. As early as 14 study weeks and well established by 27 study weeks, significant differences in AL-fed males and females compared to the moderate and marked DR-fed animals are seen in circulating cholesterol, triglycerides and total carcass fat as a percent of body weight or as grams of fat per animal. Between 6 months and 1 year, a disproportional increase in total body fat and lipid profiles has occurred in the AL-fed animals, and from 1 year forward these parameters remain abnormal, and continue to rise to massive proportions by 2 years. The induction of extreme DIO in outbred SD rats appears somewhat unique to the Charles River SD stock. Comparative studies with other stocks of outbred SD rats have reported much lower body fat in AL-fed animals at 1 year relative to the commonly used Charles River SD stock (Klinger et al., 1996). Since the Charles River SD stock is an outbred line, it would seem that they would be an ideal DIO model of the rapid onset of both juvenile and adult obesity leading to multiple co-morbidities associated with obesity in human beings (Levin et al., 1997). It is also important to note that the distribution of body fat in these animals, particularly the AL-fed animals, is central and abdominal, though abundant subcutaneous body fat is observed as well. This suggests that the model would be ideal for imaging studies of the distribution and changes in body fat distribution and modulation with treatments of potential anti-obesity drugs and other therapies. The degree to which AL-fed SD rats add white adipose tissue to their body mass is dramatic. The change from a 6% body fat content of 6-week-old growing SD rats to that of 35 to 40% body fat content as 2-year-old adults is remarkable. This growth of adipose mass represents an average addition of 280 grams of fat in both females and males during that period. This is compared to approximately 20–25% adult body fat content in the 2 moderate DR-fed male groups and approximately 10% body fat maintained in the marked DR-fed group of both sexes. However, the 10% body fat of the marked DR-fed rats is similar to that reported in wild rats (Kritchevsky et al., 1984). Since the moderate DR-fed male animals are close to the range (25% total body fat) of what would be considered obese in humans, it is clear that moderate food restriction maintains the animals in a desirable moderate state of maintenance with abundant fat reserves as is typical of this stock’s young adult phenotype. In contrast, the marked (50%) DR-fed animals contain approximately 10% body fat throughout their lifetime and are much closer in phenotype of wild rats and lean stocks of laboratory rats under AL-feeding (Klinger et al., 1996). The modulation of adiposity by simple caloric restriction indicates this stock would be an ideal model to study the distribution, development, progression, and pathogenesis of both moderate and morbid obesity. This animal represents a relatively unexamined model of the metabolic syndrome with great potential to yield important insights using the developing molecular, genomic and imaging technologies (Lee et al., 1999; Zainial et al., 2000; Koubova and Guarente, 2003; Weindruch, 2003).

The metabolic and morphological details of the individual co-morbidities of chronic nephropathy, chronic cardiomyopathy, and the early development of spontaneous pancreatic and pituitary disease in this stock have been reported (Gumprecht et al., 1993; Keenan et al., 1994a, 1994b, 1995, 1996, 1997, 2000a, 2000b; Levin et al., 1997, 2000; Kemi et al., 2000; Molon-Noblot et al., 2001; Molon-Noblot et al., 2003). Our study on the effects of AL-overfeeding and moderate and marked DR-feeding on age-related pancreatic islet cell change, clearly demonstrated a phenotype that is consistent with the development of adult-onset type 2 diabetes as described in various inbred rat models (Molon-Noblot et al., 2001). In AL-fed rats, early changes in islet morphology occurred which resulted in a high incidence of islet fibrosis focal islet hyperplasia, and eventually insulinomas or islet adenomas by 2 years (Molon-Noblot et al., 2001). The animals under moderate to marked DR had dose-proportional decreases in glucose and serum insulin levels, a decreased incidence and severity of islet cell fibrosis, and hyperplasia, and a delay in the onset of these changes (Molon-Noblot et al., 2001). Results of stereology and cell proliferation studies demonstrated that, as compared to the AL-fed animals, DR-fed rats had proportionally smaller total pancreas weights, smaller pancreatic islets, smaller insulin-secreting cell volumes, a lower degree of islet fibrosis, and lower islet cell BrdU labeling indices, which eventually correlated with a lower incidence of islet cell adenoma and carcinoma at the 106-week period. Moderate and marked DR delayed the onset and severity of islet cell hyperplasia and fibrosis in a temporal and caloric dose-related manner. In contrast to marked DR that dramatically prevented most of these adverse changes, moderate DR delayed but did not prevent the onset of islet cell tumors (Molon-Noblot et al., 2001). The changes in structure and functional modifications observed in the pancreatic islets of old, obese AL-fed SD rats resembled age-related islet hyperplasia and fibrosis that develop in the hyperinsulinemic obese Zucker rat (Molon-Noblot et al., 2001; Rhodes, 2005). The diabesity syndrome in old rats and humans is characterized by a defect of beta cell adaptation to insulin resistance and is manifested clinically by hyperglycemia, hyperinsulinemia, obesity, and the co-morbidities associated with diabetes in both species (Molon-Noblot et al., 2001). In SD rats, beta cell hyperplasia and islet enlargement occur with aging, and have been attributed to an increased demand for insulin resulting from insulin resistance and/or glucose intolerance. In older SD animals, these changes can evolve into beta cell atrophy and islet cell fibrosis. In addition, the increase in beta islet cell proliferation in the damaged islets leads ultimately to increased serum insulin levels and islet cell adenomas (Molon-Noblot et al., 2001; Rhodes, 2005).

Age-related spontaneous pituitary gland pathology occurs in most strains and stocks of rats including the Charles River SD rat (Tucker, 1979; Gries et al., 1982; Ross et al., 1983; Berry, 1986; Mietes et al., 1990; Keenan et al., 1995; Roe et al., 1995; Capen, 1996; Chiu et al., 1996; Capen, 2001; Molon-Noblot et al., 2003). We documented the temporal effects of AL-overfeeding in moderate and marked DR on these lesions and have shown that multiple endocrine organs are damaged by AL-overfeeding, and SD rats maintained in this way are suffering from an early onset of severe endocrine disruption. The AL-fed SD rats developed hyperplastic and neoplastic pituitary changes before one year that progressed with age, affecting almost all of the animals by 106 weeks. These lesions were associated with abnormally high prolactin, growth hormone, and IGF-1 levels. The hormone data correlated with the early development of pituitary adenomas, the most common cause of death of both sexes in the AL-fed group. In the DR-fed rats, a delayed onset and decreased incidence of pituitary tumors were observed in association with decreased serum IGF-1, prolactin, estradiol, and LH levels. The pituitary gland stereological and cell proliferation studies demonstrated the DR-fed animals’ pituitaries contained lower prolactin and growth hormone secreting cell volumes, had a lower epithelial cell 7-day BrdU labeling index that, in turn, correlated with a lower incidence or delayed onset of pituitary tumors by 2 years. Although the AL-fed rats were endocrinologically more like the moderate DR-fed animals than the marked DR-fed rats, the temporal delays of endocrine lesions and dysfunctions indicate that moderate DR allows for far better control of these spontaneous changes, and results in a better model to understand the toxicity of compounds with endocrine disruption potential.

As would be expected with endocrine disruption, AL-overfeeding had adverse effects on estrous cyclicity and onset of reproduction senescence in the SD females (Chapin et al., 1993; Merry et al., 1994; Keenan et al., 1995; McShane et al., 1996; Small et al., 2002; Pasquale et al., 2003; Rauser et al., 2003; Roy et al., 2003; Carney et al., 2004; Chiu et al., 1996). Since the stages of the polyestrous cycle in the rat can be classified by cell types present in vaginal lavage, and the cycle length is normally approximately 4 to 5 days, we monitored these parameters to determine the reproductive status of the unbred females on this study. Vaginal epithelia exhibited cytological changes in response to changing estrogen and progesterone levels (Merry et al., 1994; Keenan et al., 1996; McShane et al., 1996; Molon-Noblot et al., 2003; Roy et al., 2003). With age and onset of reproductive senescence, female rats developed increasing irregularity in their estrous cycles. Their transition from regular 4 to 5 day cycles to acyclicity is not abrupt but follows several transient pathways. The variability in these patterns occurred not only between rodent species, but between strains and stocks of the same species. The effects of AL-overfeeding and moderate and marked DR on estrous cyclicity were similar. At the age of study onset (6 to 8 weeks of age) all female groups had regular estrous cycles. With increasing age, the patterns of estrous cyclicity among AL-fed females and moderate DR-fed female groups were comparable. In the marked DR-fed group 4 females, the onset of reproductive senescence appeared to be delayed. During the first 3 weeks of the study, the average cycle length was increased approximately 1 day in the marked DR-fed females compared to the other groups, but no other abnormalities were noted. The difference between the estrous cycles in the marked DR-fed group 4 and the AL-fed group 1 and moderate DR-fed groups 2 and 3 was clearly established by nine months of age and only 30–40% of the group 1, 2, and 3 females had regular estrous cycles. By 1 year, 45% of the females in the marked DR-fed group 4 exhibited normal cyclicity compared to approximately 12% in the AL-fed and moderate DR-fed SD females. Although the cycle lengths remained slightly longer in the 50% DR-fed group 4 animals, the magnitude of this difference lessened. Compared to the AL-fed females, moderate DR-fed groups 2 and 3 females did not show a significant difference in their pattern of reproductive senescence from 9 months onward. It was only in the marked DR-fed group 4 animals that delays in reproductive senescence were seen, with the largest number of these females still exhibiting normal patterns of estrous cyclicity at 2 years. These data indicate that AL-fed and moderate DR-fed SD rats are similar in the onset of reproductive senescence, though AL-fed rats have an earlier development of endocrine disruption.

The AL-fed group 1 animals of both sexes had the largest adrenal glands and the highest incidence and/or severity of cortical cystic degeneration, focal hyperplasia and hypertrophy. In general, there was a proportional increase in the absolute size of the adrenal glands, food intake, and body size and an increased incidence and severity of degenerative adrenal changes. Although all the DR-fed animals had smaller absolute adrenal weights, the relative weights as a percent of body weight were proportionally larger in the DR-fed groups, in addition to having a low incidence of degenerative and proliferative changes in this organ. These data indicated less endocrine disruption, more normal adrenal function in the anatomically intact adrenals over the course of the study in the DR-fed animals (Keenan et al., 1995a, 1995b; Levin et al., 1994, 2000; Masoro, 1996). These adrenal findings and data from other endocrine organs are consistent with the observation that AL-fed animals undergo early rapid endocrine disruption in numerous systems that may be controlled or prevented by varying degrees of DR.

The cardiovascular and renal systems were particular targets with the AL-fed group, especially the males, which had the largest hearts and kidneys, and highest incidence and severity of cardiomyopathy and nephropathy. These progressive degenerative diseases were the most common causes of death after fatal tumors in the AL-fed animals. There was an increased incidence of severity and earlier onset of both cardiovascular and renal changes in the AL-fed animals. Stereologic evaluation of the hearts and kidneys of both sexes of all groups confirms the histological finds and demonstrated the progression of the lesions from the first to second year. Analysis of the absolute and relative organ weights and the stereological data demonstrated a clear dose response pattern for most parameters. The marked DR-fed rats were different from all other groups for most parameters at all time points. By 106 weeks the data were statistically significant between the AL-fed animals and all of the moderate and marked DR-fed groups. These data were consistent with the overall trends and incidence and severity of cardiomyopathy and nephropathy seen in this and other studies and indicate the toxicity of AL-overfeeding on cardiac and renal tissues (Lu et al., 1991; Gumprecht et al., 1993; Keenan et al., 1994, 2000; Kemi et al., 2000; Wan et al., 2003). These data suggest that DR-fed rats maintain normal cardiovascular and renal morphology and function longer than their AL-fed counterparts.

In our previous studies with this stock of SD rats, examining two different types of diets resulted in similar cardiovascular and renal findings. The standard Purina PMI (Purina Rodent Chow 5002) was compared to a modified rodent chow (Purina Rodent Chow 5002–9) containing less protein, fat, total kilocalories and a marked increase in crude fiber content (Gumprecht et al., 1993; Keenan et al., 1994a, 1994b, 1995b, 2000b; Kemi et al., 2000). Serum lipids, carcass composition, heart, and kidney weights and morphology were evaluated and qualitatively and quantitatively examined in animals at one and two years. The animals fed the standard diet (5002) AL had the greatest weights and most severe lesions and stereology findings. However, regardless of the type of diet fed, both AL groups fed either the high protein, high energy standard rodent diet (5002) or the low protein, low energy, high fiber modified diet (5002–9) had the most severe cardiomyopathy and nephropathy throughout the study. Moderate DR-feeding allowed isocaloric food intake comparisons of the relative effects of the modified diets, and indicated only a slight improvement in the severity and progression of spontaneous cardiomyopathy and nephropathy when modifications were made in the protein, fiber, fat and energy content of the diet, if fed ad libitum. This was because the AL-fed groups consumed almost 30% more of the diet and thus were taking in a similar amount of calories as the animals AL-fed the standard diet (Keenan et al., 1997). Thus, moderate DR of either diet was more effective than a change in the diet composition in preventing and controlling the progression of cardiomyopathy in the male SD rats (Keenan et al., 1997; Kemi et al., 2000).

The results of these and other studies indicate that the most likely mechanisms involved with the renal, musculoskeletal and cardioprotective effects of DR appear to be related to the reduction in total caloric intake and thus modulation of the role of reactive oxygen species (ROS) over the life span of the rat (Yu et al., 1989; Lee et al., 1990; Yu, 1995; Keenan et al., 1995a, 1995b; Weindruch et al., 1997; Vermorel et al., 1998; Lee et al., 1999; Zainial et al., 2000; Faine et al., 2002; Wan et al., 2003; Lowell and Shulman, 2005; Park et al., 2005). Mitochondria and microsomes are the major sources of reactive oxygen species and are targets for ROS attack. Although ROS production may not increase uniformly with age, the damage such as that observed with lipid peroxidation does increase with age, especially in post-mitotic tissues such as the heart and brain (Lee et al., 1999; Weindruch, 2003; Park and Prolla, 2005). The increased susceptibility of aging rats to oxidative damage is also indicated by their increased susceptibility to premature renal and cardiovascular death. These observations indicate that this outbred SD rat model would make an excellent representative species for the analysis of the multi-factorial polygenic, nutritional and environmental factors known to influence renal and cardiovascular disease and as a testing system to evaluate the nephrotoxic and cardiotoxic potential of compounds. A current review of rat nephropathy questions the relevance of using rats with compromised kidneys as models for human risk assessment, and points out the need to control these diet-induced diseases in the animal model (Hard et al., 2004).

Conclusion

The environmental and nutritional conditions under which laboratory animals are maintained can powerfully influence the experimental results. Nutrition is of major importance in toxicological bioassays and research because diet composition and the conditions under which it is fed can affect the metabolism and activity of xenobiotic test substances and alter the results and reproducibility of long-term studies. It is known that AL-overfed, sedentary laboratory rodents suffer from an early onset of degenerative disease, metabolic and endocrine disruption, and diet-related tumors that lead to poor survival in chronic bioassays. AL-fed animals are not well-controlled subjects for any experimental studies. Examination of study-to-study variability in food consumption, body weight, organ weights and survival in carcinogenicity studies for the same strain or stock of rodents shows AL-feeding results in tremendous laboratory-to-laboratory variability (Keenan et al., 1994a, 1994b, 1996, 2000a; Turturro et al., 1995; Duffy et al., 2001). However, a significant correlation between average food (calorie) consumption, adult body weight and survival has been clearly established (Weindruch et al., 1988; Laroque et al., 1997; Weindruch et al., 1997; Keenan et al., 1999, 2000a, 2000b; Kritchevsky, 1999). The use of moderate dietary restriction (DR) of a nutritionally balanced diet results in a better controlled rodent model with a lower incidence or delayed onset of metabolic and endocrine disruption, spontaneous diseases and tumors (Ross et al., 1983; Kritchevsky et al., 1984; Iwasaki et al., 1988; Maeda et al., 1985; Weindruch et al., 1988; Masoro et al., 1989; Yu et al., 1989; Keenan et al., 1994a, 1994b, 1995a, 1995b, 1996, 1997, 1999, 2000a, 2000b; Roe et al., 1995; Turturro et al., 1995; Yu, 1995; Masoro, 1996; Masoro et al., 1996; Weindruch, 1996, 2003; Whitaker et al., 1997; Kritchevsky, 1999; Zainial et al., 2000; Kemi et al., 2000) Operationally simple and having been automated under industrial conditions (Petruska et al., 2001), moderate DR of balanced diets significantly improves survival, controls adult body weight and obesity, reduces age-related renal, endocrine, metabolic and cardiac diseases, reduces study-to-study variability, increases study and treatment exposure time, and increases the statistical sensitivity of these expensive, chronic bioassays to detect true treatment effects. A severe dietary restriction such as a 40–50% reduction of the maximum AL intake of a given diet is not recommended as an appropriate control method for any toxicological studies. However, a moderate DR regimen of 25% to 35% restriction of the maximum unrestricted adult AL food intake of either a semi-purified or natural-product well-balanced rodent diet is recommended as a nutritionally rational, well-established method in conducting well-controlled experimental studies with the SD laboratory rat. Conversely, the AL-overfed SD rat does appear to have a great untapped potential as a polygenic outbred disease model of adult-onset diabesity that should be exploited further using current molecular and genomic methods to understand these complex disease processes, identify the most appropriate molecular targets and thus better design and test interventions to treat and control the polygenic human syndrome of diabesity, and reduce its significant effects on morbidity and mortality in the human beings worldwide.