0013-7227/06/$15.00/0 Printed in U.S.A.

Endocrinology 147(6):2801–2808 Copyright © 2006 by The Endocrine Society doi: 10.1210/en.2006-0086

Effect of Growth Hormone on Susceptibility to Diet-Induced Obesity Darlene E. Berryman, Edward O. List, Douglas T. Kohn, Karen T. Coschigano, Randy J. Seeley, and John J. Kopchick School of Human and Consumer Sciences (D.E.B.), College of Health and Human Services, Edison Biotechnology Institute (E.O.L., D.T.K., J.J.K.), and Department of Biomedical Sciences (K.T.C., J.J.K.), College of Osteopathic Medicine, Ohio University, Athens, Ohio 45701; and Department of Psychiatry and Genome Research Institute (R.J.S.), University of Cincinnati, Cincinnati, Ohio 45267 Mice with a deficiency in GH function due to disruption of the GH receptor/binding protein gene (GHRⴚ/ⴚ) are long lived, insulin sensitive, and obese, whereas mice with excess GH function due to expression of a bovine GH transgene (bGH) are short lived, glucose intolerant, and lean. When challenged with a high-fat (HF) diet, we hypothesized that these mice would be differentially susceptible to diet-induced obesity. To test this hypothesis, GHRⴚ/ⴚ, bGH, and littermate control (WT) mice were fed a HF diet (40% kcal) or a nutrient-matched low-fat diet (9% kcal) for 12 wk. On the HF diet, all mice, regardless of genotype, showed a similar percent weight gain and exhibited a significant increase in percent body fat and the mass of epididymal, retroperitoneal, and sc fat pads. For

G

H PLAYS A CRITICAL role in modulating a variety of pathways related to macronutrient storage and metabolism. Overall, GH has consistently been shown to promote growth of lean tissue while reducing accumulation of adipose tissue. Accordingly, GH-deficient states in children and adults are characterized by a decrease in lean body mass accompanied by increased adipose tissue (1–3). Treatment of GH-deficient states with exogenous GH can improve lean body mass while reducing fat mass (4 –7). These encouraging effects of GH have spawned recent interest in the efficacy of using GH or analogs in the treatment of obesity and lipodystrophic conditions (8 –10) despite the possibility that such treatments may exacerbate glucose intolerance and hyperinsulinemia (11). Given that GH affects a number of important metabolic tissues simultaneously, it is critical that in vitro data assessing GH action on isolated adipocytes, isolated adipose tissue, or clonal cell lines be complemented by in vivo approaches. Previous work using genetic mouse models to produce GH excess result in giant, short-lived, lean, and glucose-intolerant mice (12–15). Likewise, a deficiency in GH function, due to disruption of the GH receptor/binding protein gene (GHR⫺/⫺), results in dwarf, long-lived, and insulin-sensitive mice (16 –18). Other notable features of the GHR⫺/⫺ mice First Published Online March 23, 3006 Abbreviations: bGH, Bovine GH; GHR⫺/⫺, GH receptor/binding protein gene disruption; HF, high fat; LF, low fat; NMR, nuclear magnetic resonance; VO2, oxygen consumption; WT, wild type. Endocrinology is published monthly by The Endocrine Society (http:// www.endo-society.org), the foremost professional society serving the endocrine community.

bGH mice, the increase in adipose tissue was relatively small, compared with the WT or GHRⴚ/ⴚ mice, suggesting some resiliency, although not immunity, to diet-induced obesity. GHRⴚ/ⴚ mice, which are relatively obese on a low-fat diet, responded to the dietary challenge in a manner similar to WT controls. With HF feeding, all genotypes experienced an increase in insulin levels and depot-dependent effect of adipose tissue. Together, these results further support a role for GH in energy balance regulation and nutrient partitioning. More importantly, because there were genotype-specific effects of diet, these data stress the importance of diet selection and sampling multiple adipose depots in studies with these mouse models. (Endocrinology 147: 2801–2808, 2006)

include a profound increase in total fat mass with preferential accumulation of excess fat in the subcutaneous depots (13). A critical issue when examining the effect of such profound hormonal manipulations with in vivo models is the diet to which mice are exposed. This is particularly the case for GH because its effects are both tissue specific and profound with respect to lipid metabolism. A further concern is that the standard diet fed to mice is exceptionally low in fat. In fact, the standard diet is lower in fat than even the most austere human diet and is certainly not reflective of the diets consumed by most humans in developed countries in which rates of metabolic disease are on the rise (19, 20). Not surprisingly, humans and most strains of mice show increased energy intake and increased body fat when exposed to palatable and more calorically dense high-fat (HF) diets. Typically accompanying the increased fat mass is hyperinsulinemia and frank hyperglycemia. One strain of mice, C57BL/ 6J, has been extensively characterized on a calorie-dense diet. When mice from this strain are fed ad libitum a HF diet, they become obese and develop hyperinsulinemia and hyperglycemia, whereas on a low-fat (LF) diet, they remain lean and physically normal (21). In addition, the diet-induced obesity in C57BL/6J mice is characterized by selective deposition of fat in a region comparable with the visceral region in humans (22, 23). Thus, the development of obesity, insulin resistance, and hyperglycemia in the C57BL/6J mouse closely parallels the progression of obesity and diabetes in humans. Given the important effect of GH on lipid metabolism and body composition, we hypothesized that levels of wholebody GH signaling may play an important role in determining the sensitivity of an animal to weight gain when exposed

2801

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 January 2017. at 19:04 For personal use only. No other uses without permission. . All rights reserved.

2802

Endocrinology, June 2006, 147(6):2801–2808

to a HF diet. To test this hypothesis, we maintained C57BL/6J mice with targeted gene disruption of the GHR receptor (GHR⫺/⫺) and mice that express a bovine GH transgene (bGH) transgene on carefully matched LF and HF diets. The data show that excess in GH signaling results in some protection, but not complete resistance, to diet-induced obesity. On the other hand, whereas absence of GH signaling itself increases adiposity, the inclusion of a dietary challenge does not increase the susceptibility to diet-induced obesity beyond what was observed for the wild-type (WT) controls. Collectively, these data stress the importance of considering diet in measures of body weight and adiposity in whole animal models with altered GH function. Materials and Methods Animals The bGH transgenic mouse line used in this study has been previously described (13). In short, a gene fusion between a mouse metallothionein transcriptional regulatory element and a bGH cDNA containing the first intron was injected into the pronucleus of C57BL/6J embryos. The GHR⫺/⫺ mouse line, generated by homologous recombination that disrupted the fourth exon of the GHR/binding protein gene, has also been previously described (24). The specific GHR⫺/⫺ mice used in this study were backcrossed for eight generations into the C57BL/6J genetic background (17), resulting in mice that were 99.61% congenic for the C57BL/6J background. Because the GHR⫺/⫺ mice were not pure C57BL/6J, age-matched littermate controls from both the bGH and GHR⫺/⫺ mouse lines were used as controls. For all measurements, there were no statistically significant differences between the two groups of littermate controls; thus, data are shown for one group of control mice (GHR⫺/⫺ littermates). Male mice were housed in cages that were kept on a 12-h light, 12-h dark cycle with food and water provided ad libitum unless otherwise noted. Mice were bred and raised at the animal facility at Ohio University (Athens, OH). Animals were weaned onto standard rodent chow (ProLab RMH 3000; 14% of kilocalories from fat, 16% from protein, and 60% from carbohydrates) at 28 d of age. At 10 wk of age, male mice were shipped to the University of Cincinnati Mouse Metabolic Phenotyping Center for placement on a HF or LF diet and measurement of food intake, indirect calorimetry, and body composition, as described below. Mice were then returned to Ohio University for blood analysis and dissections. All procedures were approved by either the Ohio University or the University of Cincinnati Institutional Care and Use Committee and fully complied with federal, state and local policies.

Food intake At the University of Cincinnati Mouse Metabolic Phenotyping Center, 10-wk-old mice (n ⫽ 8 per each genotype and diet) were placed on one of two pelleted, semipurified, nutritionally complete, experimental diets provided by Dyets (Bethlehem, PA). These diets have been used previously with rodent models and are more fully described by Woods et al. (25). Briefly, the HF diet contained 20 g of fat per 100 g of diet (19 g butter oil and 1 g soybean oil to provide essential fatty acids) and provided 4.54 kcal/g of diet. The LF diet contained 3 g butter oil and 1 g soybean oil per 100 g of diet and provided 3.81 kcal/g. The diets were carefully matched for all other parameters including protein and micronutrient content per kilocalorie. Food intake was measured weekly along with body weight during a 10-wk period. All food intake measurements were collected on individually housed animals with data corrected for spillage. Whereas food intake was assessed for only 10 wk, mice remained on their respective diets (an additional 1.5 wk) until the mice were killed. During the additional diet period, body composition, indirect calorimetry, transport back to Ohio University, and blood collection were completed with body weights recorded before each manipulation.

Berryman et al. • GH Mouse Models and Diet-Induced Obesity

Body composition Body composition was assessed with a custom-designed rodent quantitative nuclear magnetic resonance (NMR) apparatus (Echo MRI whole body composition analyzer; Echo Medical Systems, Houston, TX), as described previously (26). Body composition measurements were performed on 20-wk-old mice after 10 wk of diet exposure.

Indirect calorimetry Indirect calorimetry was performed using the Columbus Instruments Oxymax 5.41 system as described previously (26). Twenty-one-week-old mice (n ⫽ 6 – 8 for each genotype) that had been maintained for 11 wk on their respective diets were placed in individual metabolic chambers. All animals were unrestrained in the sealed chamber and had access to water but not food. All measurements were performed during the light phase between 1000 and 1600 h and with an air flow of 0.6 l/min, with room air as the reference. Chamber air was sampled approximately every 30 min over a 6-h period, and the consumed oxygen concentration (VO2) was calculated. The decision to measure for 6 h was based on preliminary measures that suggested that the extremely small GHR⫺/⫺ mice would be unduly stressed in the metabolic chamber for an extended period of time; thus, a compromise of 6 h was chosen to measure all mice so that direct comparisons could be made between groups. For each time point, the samples for each group were averaged. Energy expenditure data were also adjusted for fat-free mass as measured by NMR. The respiratory exchange ratio, also known as the respiratory quotient, was calculated as the ratio of CO2 produced to oxygen consumed. Respiratory exchange ratio provides a measurement of relative carbohydrate and lipid oxidation and typically ranges between 1.0 when mice are oxidizing exclusively carbohydrate and 0.7 when mice are oxidizing exclusively lipids.

Glucose and insulin levels Two days after obtaining body composition and energy expenditure measurements, mice were returned to Ohio University and fasted for 4 h before blood collection. Blood collections occurred between 1800 and 1900 h. Whole blood was collected from the tail vein and blood glucose levels were determined using a Lifescan One Touch glucometer (Johnson & Johnson, New Brunswick, NJ). Additional whole blood was collected using heparinized capillary tubes and centrifuged at 7000 ⫻ g for 10 min at 4 C to collect plasma. Plasma was stored at ⫺70 C. Plasma insulin levels were determined using the Mercodia ultrasensitive mouse insulin ELISA kit (ALPCO Diagnostics, Salem, NY). The intraassay coefficients of variations for insulin were 2.5%.

Organ weights Mice were killed 1 d after blood collection. Two intraabdominal adipose depots (epididymal and retroperitoneal) and one sc depot (inguinal) were dissected and the wet mass weighed. Several other organs, including kidneys, liver, and brain, were also removed and weighed. Tissues were immediately flash frozen for future analysis.

Statistics Data are presented as mean ⫾ sem. Repeated-measures ANOVA was used to analyze body weights during the course of the 10-wk dietary study. Body composition, food intake, energy expenditure, fasting glucose and insulin levels, and organ weights were analyzed by two-way (genotype ⫻ diet) ANOVA with Tukey’s honestly significant difference post hoc test (SPSS 12.0 for Windows; SPSS Inc., Chicago, IL). Differences were considered significant at P ⬍ 0.05.

Results Weight gain and body composition

Initial body weights differed significantly between genotypes (Fig. 1A) with bGH mice having greater body weights and GHR⫺/⫺ mice having lower body weights than WT, as reported previously (13, 17). Specifically, at the start of the

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 January 2017. at 19:04 For personal use only. No other uses without permission. . All rights reserved.

Berryman et al. • GH Mouse Models and Diet-Induced Obesity

Endocrinology, June 2006, 147(6):2801–2808

2803

FIG. 1. Weight gain over time (A) and percent weight gain (B) for WT, GHR⫺/⫺, and bGH transgenic mice fed a LF and HF diet. Mice were provided food and water ad libitum and monitored for weight gain over 10 wk, starting at 10 wk of age. Note that starting and ending weights of each strained differed significantly; thus, the scale for weight gain over time is different for each mouse strain. Data are expressed as means ⫾ SEM, n ⫽ 8. For weight gain over time, asterisk (*) denotes statistically significant differences between LF and HF diet feedings (P ⬍ 0.05). For percent weight gain, means within a strain or diet group with a common letter do not differ; P ⬎ 0.05.

dietary study, the mean body weight of the bGH mice was 1.5 times greater than that of WT mice, whereas the GHR⫺/⫺ mice had a mean body weight that was approximately one third that of WT mice. All mice, regardless of genotype, gained weight when fed the HF diet, compared with the LF diet (Fig. 1A). Moreover, the weight gain was highly significant for all genotypes (WT, P ⬍ 0.0001; GHR⫺/⫺, P ⬍ 0.001; bGH, P ⬍ 0.002). Although the absolute weight gained differed, when expressed as a percent of initial body weight, normalized weight gain was not different across genotypes for either the LF or HF diet (Fig. 1B). On the LF diet, there was no significant difference in overall fat mass based on NMR data, despite the considerable size differences between these mouse groups (Fig. 2A). However, when normalized to body weight, GHR⫺/⫺ animals had significantly higher fat percentage, compared with the bGH or WT mice (Fig. 2B). These results are consistent with a previous study using GHR⫺/⫺ in the mixed Ola/BALBC genetic background (13). On the HF diet, the WT mice appeared to be the most vulnerable to an increase in fat mass and percent body fat, showing 3.1- and 2.7-fold increases, respectively. The GHR⫺/⫺ mice, which were already relatively obese, only experienced a 1.9-fold increase in fat mass and a 1.6-fold increase in percent body fat when fed the HF diet. The bGH mice, which are relatively lean, had a 2.1- and 1.8-fold increase in fat mass and percent body fat, respectively, due to a HF diet. Interestingly, virtually all of the weight gain observed upon consumption of a HF diet in WT and GHR⫺/⫺ mice (Fig. 1A) was attributed to an increase in the amount of fat rather than lean tissue mass. Conversely,

the amount of fat weight gained by the bGH mice on the HF diet accounted for only 54% of their observed weight gain. Thus, bGH mice, although gaining a similar amount of total weight on the HF diet, appeared to partition that energy more toward lean tissue than either WT or GHR⫺/⫺ mice. Food intake and energy expenditure

To assess whether weight gain was the result of increased caloric intake, decreased energy expenditure, or a combination of the two, we assessed both total caloric intake and energy expenditure in each set of mice. Although only bGH mice consumed more absolute food mass on the HF diet (Fig. 3A), all genotypes consumed significantly more kilocalories when fed the HF diet (Fig. 3B). Overall, energy intake was greatest with bGH animals, followed by WT and was lowest for GHR⫺/⫺ mice. However, when energy intake was adjusted according to body weight (Fig. 3C), GHR⫺/⫺ mice consumed significantly more energy and did not exhibit the same level of increased food intake on the HF diet as did the WT and bGH mice. Thus, the higher fat mass in GHR⫺/⫺ mice can be, at least in part, a result of greater caloric consumption relative to their body size. To assess the expenditure side of the energy balance equation, indirect calorimetry was used. Unlike food consumption, however, energy expenditure could not be monitored for the entire 10-wk period but rather was sampled at the end of the experiment for a 6-h period. During this assessment energy expenditure was not affected by the diet in any of the genotypes (Fig. 4). However, there was a significant main

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 January 2017. at 19:04 For personal use only. No other uses without permission. . All rights reserved.

2804

Endocrinology, June 2006, 147(6):2801–2808

Berryman et al. • GH Mouse Models and Diet-Induced Obesity

FIG. 2. Fat weight (A) and percent body fat (B) for WT, GHR⫺/⫺, and bGH mice fed a LF and a HF diet. Values are mean ⫾ SEM, n ⫽ 8. Means within a strain or diet group with a common letter do not differ; P ⬎ 0.05.

effect of genotype [F(4, 36) ⫽ 14.22, P ⬍ 0.001]. Tukey’s post hoc test reveals that GHR⫺/⫺ mice consume significantly more oxygen with no difference between bGH or WT animals. This significant increase in oxygen consumption by GHR⫺/⫺ mice was maintained, even when the data were expressed relative to lean body mass instead of whole body mass (data not shown). Dietary manipulation and genotype had no significant effect on respiratory quotient, an indicator of metabolic fuel preference (data not shown). Tissue weights

Because of the significant difference in body weights for bGH, GHR⫺/⫺, and WT mice, tissue weights are expressed both in absolute terms (Table 1, top, and Fig. 5, top) and normalized to body weight (Table 1, bottom; Figure 5, bottom). On a LF diet, the mouse models distributed adipose tissue in a distinct manner. The largest fat pad in LF-fed bGH and WT mice was the epididymal fat pad, an intraabdominal fat pad. In contrast, the inguinal (sc) fat pad was largest in the GHR⫺/⫺ mice. Interestingly, the inguinal depot in terms of absolute weight was not significantly different among the three genotypes, which is noteworthy, considering the extreme size differences of these animal models. This difference becomes apparent when inguinal fat pad weight is expressed as a percent of total body weight. These findings using

FIG. 3. Food intake of WT, GHR⫺/⫺, and bGH mice fed either a LF or HF diet. Data are presented based on the total grams of food consumed (A), total energy consumed (B), or total energy consumed per body weight (C). Mice were housed individually for all measurements. Energy was calculated by multiplying grams of food consumed by the digestible energy content of the diet. Values are mean ⫾ SEM, n ⫽ 8. Means within a strain or diet group with a common letter do not differ; P ⬎ 0.05.

C57BL/6J mice confirm earlier studies of GHR⫺/⫺ mice in the mixed Ola/BALBC genetic background (13). Liver, brain, and kidney sizes were also significantly impacted by genotype on the LF diet. When fed a HF diet, absolute and normalized weight of all fat pads (epididymal, retroperitoneal, and inguinal) increased significantly for every genotype. Thus, each mouse genotype used in this study was capable of increasing adipose stores when exposed to a HF diet. However, the response to diet was not the same for each genotype. Similar to the results with body composition, bGH mice appear to be somewhat protected from accumulating excess weight in each of the measured fat pads when fed the HF diet. More specifically, the bGH mice experienced an approximately 1.5-fold increase, whereas the WT and GHR⫺/⫺ mice experienced a 2- to 2.7-fold increase in fat pad mass when mice

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 January 2017. at 19:04 For personal use only. No other uses without permission. . All rights reserved.

Berryman et al. • GH Mouse Models and Diet-Induced Obesity

Endocrinology, June 2006, 147(6):2801–2808

2805

Discussion

FIG. 4. VO2 (milliliters per kilogram per hour) of male WT, bGH, and GHR⫺/⫺ mice fed either a LF or HF diet. Values were averaged over the course of the 6-h sampling and represent mean ⫾ SEM, n ⫽ 8. Means within a strain or diet group with a common letter do not differ; P ⬎ 0.05.

were fed a HF diet. Unlike previous reports in at least the mesenteric fat pad of the C57BL/6J strain (22, 23), there was no indication that other intraabdominal fat pads (epididymal and retroperitoneal) were the preferred site of depositing the excess energy on the caloric-rich HF diet. Regarding other tissues, the absolute weight of the liver and brain did not alter when mice were fed a HF diet. Interestingly, kidney weights did increase in WT and bGH mice fed a HF diet but did not change in the GHR⫺/⫺ mice. Blood glucose and insulin

Fasting glucose levels were significantly different between genotypes but not different between the dietary regimens (Table 2). As predicted from previous studies (17, 18), twoway ANOVA revealed a significant main effect of genotype for plasma insulin [F(1, 49) ⫽ 7.88, P ⬍ 0.001] with GHR⫺/⫺ mice having lower values and bGH having higher values than WT animals. Insulin levels were significantly increased with the HF diet feeding for WT, GHR⫺/⫺, and bGH animals. TABLE 1. Tissue weights of WT, GHR⫺/⫺, and bGH mice fed a LF or HF diet for 12 wk Liver

Brain

Absolute weight (g tissue) WT-LF 1.49 ⫾ 0.07a 0.44 ⫾ 0.007a WT-HF 1.58 ⫾ 0.21a 0.44 ⫾ 0.012a ⫺/⫺ b 0.41 ⫾ 0.03 GHR -LF 0.37 ⫾ 0.004b 0.37 ⫾ 0.02b GHR⫺/⫺-HF 0.36 ⫾ 0.006b bGH-LF 2.98 ⫾ 0.09c 0.50 ⫾ 0.09c bGH-HF 3.26 ⫾ 0.17c 0.50 ⫾ 0.005c Normalized weight (g tissue/g body wt) ⫻ 100 WT-LF 5.3 ⫾ 0.2a 1.6 ⫾ 0.04a WT-HF 4.3 ⫾ 0.3b 1.2 ⫾ 0.07b 3.7 ⫾ 0.2c GHR⫺/⫺-LF 3.4 ⫾ 0.1c ⫺/⫺ d 2.7 ⫾ 0.1 GHR -HF 2.6 ⫾ 0.1d bGH-LF 6.6 ⫾ 0.2e 1.1 ⫾ 0.03e bGH-HF 6.1 ⫾ 0.3e 0.9 ⫾ 0.02e

Kidney

0.35 ⫾ 0.009a 0.89 ⫾ 0.011b 0.11 ⫾ 0.006c 0.11 ⫾ 0.007c 0.51 ⫾ 0.017d 0.63 ⫾ 0.024e 1.2 ⫾ 0.1a 1.1 ⫾ 0.6a 1.0 ⫾ 0.05b 0.8 ⫾ 0.01c 1.1 ⫾ 0.03a,b 1.2 ⫾ 0.04a

Values are mean ⫾ SEM, n ⫽ 8. Means in a column for absolute or proportional weight without a common letter differ significantly; P ⬍ 0.05.

GH has long been known to exert potent metabolic effects on diverse tissues including adipose tissue. The two mouse models studied here, GHR⫺/⫺ and bGH mice, have been extensively studied when maintained on standard mouse chow. Largely the current data using a purified LF diet confirm previous reports that GHR⫺/⫺ mice are relatively obese (13, 27, 28), store a large fraction of their excess fat mass in the sc region (13), and are hyperphagic relative to their body size (17, 27). Similarly, data presented here for bGH animals are similar to data for several other rodent models with excessive GH function in that all have a lower percentage of body fat (13, 29 –31) and a food intake proportional to their body size (13, 14). The key issue for the present experiments, however, is the degree to which manipulations of the GH system result in increased or decreased susceptibility to weight gain when animals are placed on a HF diet. Both GHR⫺/⫺ and bGH mice are capable of gaining substantial amounts of weight when maintained on the HF diet. However, compared with wildtype controls, the bGH mice appear to be more resistant to excess weight gain in adipose tissue, most likely partitioning that excess energy to other tissues, as has been suggested by previous reports (32, 33). The bGH mice also showed a greater increase in consumption of the HF diet, compared with mice maintained on the LF diet, showing greater intake of absolute grams and absolute kilocalories as well as kilocalories per body weight. In contrast, for the GHR⫺/⫺ mice, consumption of the HF diet was similar to consumption of the LF diet when normalized to body weight. Although it is possible that the observed genotype difference in food consumption could be due to some physical limitation of the dwarf animals, this seems unlikely because the absolute mass of food consumed (Fig. 3A) did not significantly increase with HF feeding in either WT or GHR⫺/⫺ mice. On the energy expenditure side of the equation, there was no effect of the HF diet to reduce energy expenditure in any of the mice models tested. The absence of an effect for diet needs to be interpreted cautiously as we only monitored energy expenditure for a short time and only after extensive exposure to the HF diet. Thus, changes in energy expenditure that occurred early in the diet exposure could have contributed to the observed increased weight. Finally, insulin levels increased significantly for all three genotypes with the HF diet feeding despite not significant increase in glucose levels. Obesity results from an imbalance between food intake and energy expenditure. On a LF diet, GHR⫺/⫺ mice are obese relative to WT mice, whereas bGH mice are lean. Altered food intake and energy expenditure due to differing GH levels or GH signaling have been reported previously, although the results vary according to the model and methods used. Studies in rodents with excess levels of GH function report that either GH can suppress food intake (31) or food intake is proportional to body size (13, 34), as we show here. Interestingly, overexpression of GH specifically in the central nervous system appears to promote hyperphagia (34). Rodent models with GH deficiencies have shown either no effect on food consumption (13, 35) or, as with GHR⫺/⫺ mice on either the LF or HF diet in this study, an increase in

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 January 2017. at 19:04 For personal use only. No other uses without permission. . All rights reserved.

2806

Endocrinology, June 2006, 147(6):2801–2808

Berryman et al. • GH Mouse Models and Diet-Induced Obesity

FIG. 5. Comparison of intraabdominal (epididymal, retroperitoneal) and sc (inguinal) fat pad weights for WT, GHR⫺/⫺, and bGH mice fed either a LF (䡺, open bars) or HF (f, black bars) diet. Data are expressed in absolute weight (top) and relative to body weight (bottom) for each fat pad. Values are mean ⫾ SEM, n ⫽ 8. Means within a strain or diet group with a common letter do not differ; P ⬎ 0.05.

food intake relative to body weight (36, 37). Energy expenditure data have also been somewhat inconsistent, with excess GH function typically being associated with increases in energy expenditure in clinical studies (38). In a different mouse strain expressing bGH, Moura et al. (39) showed that these transgenic mice had a lower resting metabolic rate than littermate controls, whereas Olsson et al. (14) suggested a higher metabolic rate. Whereas GHR⫺/⫺ mice have been shown to have a reduction in core body temperature (40), this study is the first to use indirect calorimetry to assess energy expenditure in these mice and suggested, at least for the short time measured, that energy expenditure was elevated in these mice. However, a comprehensive study looking at another mouse model of GH deficiency, SMA1 mice with a missense mutation in the GH gene, systematically and more thoroughly addressed the impact of GH deficiency on basal and metabolic rates (41). They conclude that the metabolic alterations observed in the SMA1 mice reflect the reduced body size of these animals and provide no evidence for a direct impact of GH deficiency on in vivo RMR and promotion of positive energy balance. Overall, GH signaling appears to more consistently influence the energy intake half of the balance equation. The study presented here is the second to address the influence of a HF diet on obesity of bGH mice. The first study reported that bGH mice fed a HF diet are hyperphagic and diabetic but are completely resistant to diet-induced obesity in that the bGH mice on the HF diet did not show an increase in body weight or fat pad mass (14). The resistance to dietinduced obesity was attributed, in part, to an increase in energy expenditure. We have found that bGH mice, although somewhat protected from gaining fat mass on a HF diet, were

not completely resistant to diet-induced obesity (i.e. fat grams, percent body fat, fat pad mass, and proportional fat pad weight all increased when bGH mice were fed the HF diet). However, the increase in these parameters was relatively small, compared with the WT mice, suggesting some resiliency, although not immunity, to diet-induced obesity. Whereas there are differences in WT mice between this and the previous study with respect to feeding behavior, both studies show that bGH mice are relatively hyperphagic when fed the HF diet. However, energy expenditure data differed with ours showing no significant difference between bGH animals and littermate controls and no clear influence of diet. In the other study, bGH animals had elevated energy expenditure, compared with controls on the lower fat diet, and only bGH animals exhibited an increase in energy expenditure when given the HF diet. There are several differences between these studies that likely contributed to the dissimilar results. Although both HF diets contained a similar proportion of kilocalories from fat (⬃40%), our LF diet was controlled to match the HF diet in all respects except total amount of carbohydrate and fat, whereas chow was used as the control diet in the previous study. In addition, our study started with 10-wk-old male mice in a pure C57BL/6J background, and the mice were fed the diets for approximately 12 wk. The other study started with 5- to 6-month-old males in a mixed genetic background, with most of the results being obtained after only 8 wk on the diets. Whereas the difference in background strain alone could account for the reported differences, the older age of the bGH mice could be problematic because bGH mice have significant reductions in life span (13) accompanied by several notable health problems by 6 – 8 months of age (42– 44).

TABLE 2. Fasting plasma glucose and insulin concentrations in WT, GHR⫺/⫺, and bGH mice fed a LF or HF diet for 12 wk GHR⫺/⫺

WT

Glucose, mg/dl Insulin, ng/ml

bGH

LF

HF

LF

HF

LF

HF

148 ⫾ 6.3a 0.83 ⫾ 0.13a

148 ⫾ 5.9a 2.2 ⫾ 0.4c

84.5 ⫾ 8.8b 0.29 ⫾ 0.08b

90.6 ⫾ 8.9b 0.45 ⫾ 0.11d

111 ⫾ 3.7c 1.7 ⫾ 0.8a

122 ⫾ 5.6c 2.5 ⫾ 0.66c

Values are means ⫾ SEM, n ⫽ 8. Means within a row without a common letter differ; P ⱕ 0.05. Values are for the 12-wk time point after all food intake, body composition, and energy expenditure measurements.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 January 2017. at 19:04 For personal use only. No other uses without permission. . All rights reserved.

Berryman et al. • GH Mouse Models and Diet-Induced Obesity

Finally, indirect calorimetry data varied most likely because Olsson et al. (14) conducted VO2 measurements at thermoneutrality, thus accounting for inherent differences in body temperature. Visceral and sc adipose tissue have been shown to be metabolically distinct (45), which likely contributes to the association of visceral adipose tissue, and not total fat mass, with obesity-related metabolic alterations. Several studies in rodents have suggested that GH impacts adipose tissue in a depot-dependent manner (13, 28, 46). In this study, there were apparent differences between intraabdominal (epididymal and retroperitoneal) and sc (inguinal) fat pads in the various mouse models and in response to diet, demonstrating a depot-specific effect. For example, regardless of diet, the absolute weights of both intraabdominal pads were significantly reduced in GHR⫺/⫺ mice, compared with WT mice, but the absolute weights of the sc fat pad were similar in these same mice. In contrast, the bGH mice showed a diet-dependent effect with the absolute weight of every fat pad being comparable with WT when fed a LF diet but with all depots being markedly lower than WT when fed a HF diet. At this time, no single explanation can account for these genotype, diet, and depot-specific differences. This is likely because GH is thought to promote two somewhat opposing processes, cellular proliferation/differentiation and lipolysis. Accordingly, data expressed only in terms of absolute depot mass would not distinguish among these possibilities. Nevertheless, our data showing the ability of GH to uniquely impact a particular depot highlight the need to analyze more than overall body composition or a single adipose depot in studies with this hormone. In addition, these data suggest that these mouse models may represent a useful tool to investigate the genes and proteins involved in regulating or maintaining body fat distribution. Excess GH has been implicated in the development of renal complications (42, 47, 48). Accordingly, diabetic animals that have diminished GH signaling appear to be somewhat protected against renal enlargement, glomerular hypertrophy, and urinary albumin excretion (49 –52). Although diabetes and kidney function were not carefully assessed in this study, it is interesting to note that that the HF diet caused renal enlargement in bGH and WT mice, whereas the kidney weights of GHR⫺/⫺ mice were not altered. Because kidney hypertrophy is considered a sign of renal damage (53, 54), these data suggest that the kidneys of GHR⫺/⫺ mice might be protected against dietary insults. In summary, this study confirms previous reports with LF diets that the absence of GH signaling results in increased adiposity, preferentially localized to the sc region, whereas an excess in GH signaling results in an overall lean phenotype. On the LF diet, excess energy intake in the GHR⫺/⫺ mice relative to their dwarf size is, in part, responsible for the increased accumulation of adipose tissue. On a HF diet, both mouse models (GHR⫺/⫺ and GH) gained a similar proportion of total weight with some increase in fat mass. Thus, both mouse models are susceptible to diet-induced obesity. However, because the weight gain of bGH mice on a HF diet was not solely due to an increase in adipose tissue, unlike GHR⫺/⫺ and control animals, excess GH signaling does somewhat protect these mice from targeting excess kilocalo-

Endocrinology, June 2006, 147(6):2801–2808

2807

ries to adipose tissue, most likely redirecting this energy to lean tissues. Acknowledgments Received January 24, 2006. Accepted March 13, 2006. Address all correspondence and requests for reprints to: Darlene E. Berryman, W324 Grover Center, School of Human and Consumer Sciences, Ohio University, Athens, Ohio 45701. E-mail: berrymad@ ohio.edu. This work was supported in part by funds from the National Institute of Diabetes and Digestive and Kidney Disease (Grant DK064905; to D.E.B.) and the State of Ohio Eminent Scholar’s Program that includes a grant from Milton and Lawrence Goll (to J.J.K.). This work was also supported by the Metabolic Mouse Phenotype Center (DK059630) at the University of Cincinnati. Disclosures: D.E.B., E.O.L., D.T.K., K.T.C., and J.J.K. have nothing to declare. R.J.S. has received lecture fees from Abbott Laboratories and has previously received consulting fees from GlaxoSmithKline, Leptos Biomedical, Fulfillium, Inc., and Predix Pharmaceuticals.

References 1. Murray RD, Adams JE, Shalet SM 2004 Adults with partial growth hormone deficiency have an adverse body composition. J Clin Endocrinol Metab 89: 1586 –1591 2. Beshyah SA, Freemantle C, Thomas E, Rutherford O, Page B, Murphy M, Johnston DG 1995 Abnormal body composition and reduced bone mass in growth hormone deficient hypopituitary adults. Clin Endocrinol (Oxf) 42: 179 –189 3. De Boer H, Blok GJ, Voerman HJ, De Vries PM, van der Veen EA 1992 Body composition in adult growth hormone-deficient men, assessed by anthropometry and bioimpedance analysis. J Clin Endocrinol Metab 75:833– 837 4. Richelsen B, Pedersen SB, Borglum JD, Moller-Pedersen T, Jorgensen J, Jorgensen JO 1994 Growth hormone treatment of obese women for 5 wk: effect on body composition and adipose tissue LPL activity. Am J Physiol 266:E211– E216 5. Salomon F, Cuneo RC, Hesp R, Sonksen PH 1989 The effects of treatment with recombinant human growth hormone on body composition and metabolism in adults with growth hormone deficiency. N Engl J Med 321:1797–1803 6. Bengtsson BA, Eden S, Lonn L, Kvist H, Stokland A, Lindstedt G, Bosaeus I, Tolli J, Sjostrom L, Isaksson OG 1993 Treatment of adults with growth hormone (GH) deficiency with recombinant human GH. J Clin Endocrinol Metab 76:309 –317 7. Beshyah SA, Freemantle C, Shahi M, Anyaoku V, Merson S, Lynch S, Skinner E, Sharp P, Foale R, Johnston DG 1995 Replacement treatment with biosynthetic human growth hormone in growth hormone-deficient hypopituitary adults. Clin Endocrinol (Oxf) 42:73– 84 8. Vigano A, Mora S, Manzoni P, Schneider L, Beretta S, Molinaro M, di Natale B, Brambilla P 2005 Effects of recombinant growth hormone on visceral fat accumulation: pilot study in human immunodeficiency virus-infected adolescents. J Clin Endocrinol Metab 90:4075– 4080 9. Engelson ES, Glesby MJ, Mendez D, Albu JB, Wang J, Heymsfield SB, Kotler DP 2002 Effect of recombinant human growth hormone in the treatment of visceral fat accumulation in HIV infection. J Acquir Immune Defic Syndr 30:379 –391 10. Bays HE 2004 Current and investigational antiobesity agents and obesity therapeutic treatment targets. Obes Res 12:1197–1211 11. van der Lely AJ 2004 Justified and unjustified use of growth hormone. Postgrad Med J 80:577–580 12. Knapp JR, Chen WY, Turner ND, Byers FM, Kopchick JJ 1994 Growth patterns and body composition of transgenic mice expressing mutated bovine somatotropin genes. J Anim Sci 72:2812–2819 13. Berryman DE, List EO, Coschigano KT, Behar K, Kim JK, Kopchick JJ 2004 Comparing adiposity profiles in three mouse models with altered GH signaling. Growth Horm IGF Res 14:309 –318 14. Olsson B, Bohlooly YM, Fitzgerald SM, Frick F, Ljungberg A, Ahren B, Tornell J, Bergstrom G, Oscarsson J 2005 Bovine growth hormone transgenic mice are resistant to diet-induced obesity but develop hyperphagia, dyslipidemia, and diabetes on a high-fat diet. Endocrinology 146:920 –930 15. Quaife CJ, Mathews LS, Pinkert CA, Hammer RE, Brinster RL, Palmiter RD 1989 Histopathology associated with elevated levels of growth hormone and insulin-like growth factor I in transgenic mice. Endocrinology 124:40 – 48 16. Coschigano KT, Clemmons D, Bellush LL, Kopchick JJ 2000 Assessment of growth parameters and life span of GHR/BP gene-disrupted mice. Endocrinology 141:2608 –2613 17. Coschigano KT, Holland AN, Riders ME, List EO, Flyvbjerg A, Kopchick JJ 2003 Deletion, but not antagonism, of the mouse growth hormone receptor

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 January 2017. at 19:04 For personal use only. No other uses without permission. . All rights reserved.

2808

18.

19. 20. 21. 22. 23.

24.

25. 26. 27.

28. 29. 30. 31. 32. 33. 34.

35. 36.

Endocrinology, June 2006, 147(6):2801–2808

results in severely decreased body weights, insulin and IGF-1 levels and increased lifespan. Endocrinology 144:3799 –3810 Liu JL, Coschigano KT, Robertson K, Lipsett M, Guo Y, Kopchick JJ, Kumar U, Liu YL 2004 Disruption of growth hormone receptor gene causes diminished pancreatic islet size and increased insulin sensitivity in mice. Am J Physiol Endocrinol Metab 287:E405–E413 Lieberman LS 2003 Dietary, evolutionary, and modernizing influences on the prevalence of type 2 diabetes. Annu Rev Nutr 23:345–377 Popkin BM 2004 The nutrition transition: an overview of world patterns of change. Nutr Rev 62:S140 –S143 Surwit RS, Kuhn CM, Cochrane C, McCubbin JA, Feinglos MN 1988 Dietinduced type II diabetes in C57BL/6J mice. Diabetes 37:1163–1167 Rebuffe-Scrive M, Surwit R, Feinglos M, Kuhn C, Rodin J 1993 Regional fat distribution and metabolism in a new mouse model (C57BL/6J) of non-insulindependent diabetes mellitus. Metabolism 42:1405–1409 Surwit RS, Feinglos MN, Rodin J, Sutherland A, Petro AE, Opara EC, Kuhn CM, Rebuffe-Scrive M 1995 Differential effects of fat and sucrose on the development of obesity and diabetes in C57BL/6J and A/J mice. Metabolism 44:645– 651 Zhou Y, Xu BC, Maheshwari HG, He L, Reed M, Lozykowski M, Okada S, Cataldo L, Coschigamo K, Wagner TE, Baumann G, Kopchick JJ 1997 A mammalian model for Laron syndrome produced by targeted disruption of the mouse growth hormone receptor/binding protein gene (the Laron mouse). Proc Natl Acad Sci USA 94:13215–13220 Woods SC, Seeley RJ, Rushing PA, D’Alessio D, Tso P 2003 A controlled high-fat diet induces an obese syndrome in rats. J Nutr 133:1081–1087 Strader AD, Reizes O, Woods SC, Benoit SC, Seeley RJ 2004 Mice lacking the syndecan-3 gene are resistant to diet-induced obesity. J Clin Invest 114:1354 – 1360 Egecioglu E, Bjursell M, Ljungberg A, Dickson SL, Kopchick JJ, Bergstrom G, Svensson L, Oscarsson J, Tornell J, Bohlooly YM 2005 Growth hormone receptor deficiency results in blunted ghrelin feeding response, obesity and hypolipidemia in mice. Am J Physiol Endocrinol Metab 290:E317–E325 Flint DJ, Binart N, Kopchick J, Kelly P 2003 Effects of growth hormone and prolactin on adipose tissue development and function. Pituitary 6:97–102 Oberbauer AM, Murray JD 1998 Consequences of limited exposure to elevated growth hormone in the mature oMt1a-oGH transgenic mouse. Growth Dev Aging 62:87–93 Pomp D, Oberbauer AM, Murray JD 1996 Development of obesity following inactivation of a growth hormone transgene in mice. Transgenic Res 5:13–23 Malmlof K, Din N, Johansen T, Pedersen SB 2002 Growth hormone affects both adiposity and voluntary food intake in old and obese female rats. Eur J Endocrinol 146:121–128 Etherton TD 2001 Porcine growth hormone: a central metabolic hormone involved in the regulation of adipose tissue growth. Nutrition 17:789 –792 Turner ND, Knapp JR, Byers FM, Kopchick JJ 1998 Expression of mutant bovine growth hormone genes in mice perturbs age-related nutrient utilization patterns. J Nutr 128:520 –524 Bohlooly YM, Olsson B, Bruder CE, Linden D, Sjogren K, Bjursell M, Egecioglu E, Svensson L, Brodin P, Waterton JC, Isaksson OG, Sundler F, Ahren B, Ohlsson C, Oscarsson J, Tornell J 2005 Growth hormone overexpression in the central nervous system results in hyperphagia-induced obesity associated with insulin resistance and dyslipidemia. Diabetes 54:51– 62 Puche RC, Alloatti R, Chapo G 2002 Growth and development of male “little” mice assessed with Parks’ theory of feeding and growth. Growth Dev Aging 66:71–78 Furuhata Y, Kagaya R, Hirabayashi K, Ikeda A, Chang KT, Nishihara M, Takahashi M 2000 Development of obesity in transgenic rats with low cir-

Berryman et al. • GH Mouse Models and Diet-Induced Obesity

37. 38. 39. 40.

41. 42.

43.

44. 45. 46. 47.

48. 49. 50. 51.

52.

53. 54.

culating growth hormone levels: involvement of leptin resistance. Eur J Endocrinol 143:535–541 Furuhata Y, Hirabayashi K, Yonezawa T, Takahashi M, Nishihara M 2002 Effects of pair-feeding and growth hormone treatment on obese transgenic rats. Eur J Endocrinol 146:245–249 Berneis K, Keller U 1996 Metabolic actions of growth hormone: direct and indirect. Baillieres Clin Endocrinol Metab 10:337–352 Moura AS, Spiers DE, Lamberson WR 1998 Thermogenic activity of growth hormone transgenic mice. Growth Dev Aging 62:149 –159 Hauck SJ, Hunter WS, Danilovich N, Kopchick JJ, Bartke A 2001 Reduced levels of thyroid hormones, insulin, and glucose, and lower body core temperature in the growth hormone receptor/binding protein knockout mouse. Exp Biol Med (Maywood) 226:552–558 Meyer CW, Klingenspor M, Rozman J, Heldmaier G 2004 Gene or size: metabolic rate and body temperature in obese growth hormone-deficient dwarf mice. Obes Res 12:1509 –1518 Machado MO, Hirata RD, Sellitti DF, Iotti R, Iotti A, Cusumano AM, Riordan GP, Coschigano KT, Kopchick JJ, Zuhl I, Nguyen N, Hirata MH, Doi SQ 2005 Growth hormone promotes glomerular lipid accumulation in bGH mice. Kidney Int 68:2019 –2028 Frick F, Bohlooly YM, Linden D, Olsson B, Tornell J, Eden S, Oscarsson J 2001 Long-term growth hormone excess induces marked alterations in lipoprotein metabolism in mice. Am J Physiol Endocrinol Metab 281:E1230 – E1239 Bohlooly YM, Carlson L, Olsson B, Gustafsson H, Andersson IJ, Tornell J, Bergstrom G 2001 Vascular function and blood pressure in GH transgenic mice. Endocrinology 142:3317–3323 Klaus S, Keijer J 2004 Gene expression profiling of adipose tissue: individual, depot-dependent, and sex-dependent variabilities. Nutrition 20:115–120 Kadim IT, McCutcheon SN, Purchas RW, Wickham GA 1996 Manipulation of adult body composition by treatment of the neonatal rat with growth hormone and prolactin. Growth Regul 6:201–205 Yang CW, Striker LJ, Pesce C, Chen WY, Peten EP, Elliot S, Doi T, Kopchick JJ, Striker GE 1993 Glomerulosclerosis and body growth are mediated by different portions of bovine growth hormone. Studies in transgenic mice. Lab Invest 68:62–70 Yang CW, Striker LJ, Kopchick JJ, Chen WY, Pesce CM, Peten EP, Striker GE 1993 Glomerulosclerosis in mice transgenic for native or mutated bovine growth hormone gene. Kidney Int Suppl 39:S90 –S94 Chen NY, Chen WY, Bellush L, Yang CW, Striker LJ, Striker GE, Kopchick JJ 1995 Effects of streptozotocin treatment in growth hormone (GH) and GH antagonist transgenic mice. Endocrinology 136:660 – 667 Bellush LL, Doublier S, Holland AN, Striker LJ, Striker GE, Kopchick JJ 2000 Protection against diabetes-induced nephropathy in growth hormone receptor/binding protein gene-disrupted mice. Endocrinology 141:163–168 Flyvbjerg A, Bennet WF, Rasch R, Kopchick JJ, Scarlett JA 1999 Inhibitory effect of a growth hormone receptor antagonist (G120K-PEG) on renal enlargement, glomerular hypertrophy and urinary albumin excretion in experimental diabetes in mice. Diabetes 48:377–382 Li W, Zhu LH, Wang EB, Ye ZC, Lin J, Guo LH, Luo FH, Liu XH, Fang X, Shen SX 2004 Effect of two human growth hormone receptor antagonists on glomerulosclerosis in streptozotocin-induced diabetic rats. Acta Pharmacol Sin 25:490 – 495 Kriz W, LeHir M 2005 Pathways to nephron loss starting from glomerular diseases-insights from animal models. Kidney Int 67:404 – 419 Shea SM, Raskova J, Morrison AB 1978 A stereologic study of glomerular hypertrophy in the subtotally nephrectomized rat. Am J Pathol 90:201–210

Endocrinology is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving the endocrine community.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 January 2017. at 19:04 For personal use only. No other uses without permission. . All rights reserved.