Growth Hormone Receptor Antagonist Transgenic Mice Are Protected From Hyperinsulinemia and Glucose Intolerance Despite Obesity When Placed on a HF Diet

Abstract

Reduced GH levels have been associated with improved glucose metabolism and increased longevity despite obesity in multiple mouse lines. However, one mouse line, the GH receptor antagonist (GHA) transgenic mouse, defies this trend because it has reduced GH action and increased adiposity, but glucose metabolism and life span are similar to controls. Slight differences in glucose metabolism and adiposity profiles can become exaggerated on a high-fat (HF) diet. Thus, in this study, male and female GHA and wild-type (WT) mice in a C57BL/6 background were placed on HF and low-fat (LF) diets for 11 weeks, starting at 10 weeks of age, to assess how GHA mice respond to additional metabolic stress of HF feeding. On a HF diet, all mice showed significant weight gain, although GHA gained weight more dramatically than WT mice, with males gaining more than females. Most of this weight gain was due to an increase in fat mass with WT mice increasing primarily in the white adipose tissue perigonadal depots, whereas GHA mice gained in both the sc and perigonadal white adipose tissue regions. Notably, GHA mice were somewhat protected from detrimental glucose metabolism changes on a HF diet because they had only modest increases in serum glucose levels, remained glucose tolerant, and did not develop hyperinsulinemia. Sex differences were observed in many measures with males reacting more dramatically to both a reduction in GH action and HF diet. In conclusion, our findings show that GHA mice, which are already obese, are susceptible to further adipose tissue expansion with HF feeding while remaining resilient to alterations in glucose homeostasis.

Decreased GH and IGF-1 induced intracellular signaling have been shown to consistently increase life span despite adiposity in multiple dwarf mouse lines including GH receptor gene disrupted (GHR−/−), Snell, Ames, and lit/lit mice (1). However, the dwarf GH receptor antagonist (GHA) line, defies this trend in that it has a normal life span (2). GHA mice express a mutant bovine GH gene in which the codon for glycine at position 119 is replaced with lysine (3–5). The mutant GH competitively inhibits the binding of endogenous GH to the GH receptor (GHR), thus blocking its function (6).

Some phenotypic differences among GHA and other mouse lines with decreased GH action have been reported. Like other dwarf mouse lines, GHA mice have decreased body weight; however, male GHA mice (not females) catch up in body weight to their wild-type (WT) controls as they age (2, 7), whereas other lines remain significantly smaller than controls throughout life (8). The catch up in body weight is exclusively due to fat mass gains, not to changes in lean mass (7). In fact, when depot weights are examined, GHA gain significantly in the sc depot, similar to GHR−/− mice, but also gain in the intraabdominal depots (7, 9). Additionally, they have alterations in glucose metabolic parameters and adipokine levels as they age. Specifically, in young GHA mice, serum levels of both glucose and insulin are low to normal; however, insulin levels spike at older ages, although glucose levels remain normal (2, 7, 9). Adipokine levels are also significantly altered because both adiponectin and leptin levels are higher than WT controls, similar to other mouse lines with a reduction in GH action, but leptin surges even higher with age only in the GHA mice (7, 9, 10). Thus, as GHA mice age and continue to accumulate fat mass, they begin to exhibit a metabolic phenotype more consistent with obesity than GH deficiency. It is important to note that some of the GHA phenotypes could be moderated by IGF-1, which is reduced in the line but not to the extent seen in dwarf, GHR−/− mice (2).

Often phenotypic differences when comparing mouse lines are either too slight or do not occur on standard chow diets. Previous studies on GHA mice have been conducted using only a low-fat (LF)/chow diet, but other mouse lines of GH resistance have been challenged with a high-fat (HF) diet. For example, male GHR−/− mice, which are normally obese on a LF diet, develop extreme obesity on a HF diet, with percent gains similar to or greater than controls (11, 12). Importantly, GHR−/− mice remain protected from high glucose and insulin levels, despite their extreme adiposity on a HF diet (11, 12). Similarly, male muscle-specific GHR−/− mice have modest improvements to metabolism with LF feeding that can be exaggerated with HF feeding (13). Thus, mouse lines with a reduction in GH action still experience weight gain on a HF diet but seem to be protected from the consequences normally associated with adiposity.

Whereas other mouse lines with decreased GH action experience protection from obesity-related metabolic dysfunctions, even on a HF diet, there are indications that GHA mice may not share the same fortune. Specifically, we hypothesized that GHA mice would have more dramatic gains in fat mass, due to the reduction in GH action, but no protection in glucose metabolism or metabolic dysfunction despite decreased GH action. Furthermore, no previous HF diet study using mice with altered GH action has included both sexes despite the well-documented sex differences in these and other GH-modified mouse lines (7, 8, 12). Therefore, the metabolic and body composition profiles of male and female GHA and WT mice on micronutrient-matched HF and LF diets were compared in this study, similar to a previous report using GHR−/− mice (12).

Materials and Methods

Animals

The GHA mice used in this study have been previously described (3–5, 12) and have been bred onto a C57BL/6J background (2). A cohort of 20 male and 20 female GHA mice and 44 age-matched controls were generated at the Edison Biotechnology Institute of Ohio University. Two to four littermates were housed per cage with controlled light (14 h light, 10 h dark cycle) and temperature (22 ± 2°C). The Ohio University Institutional Animal Care and Use Committee approved all procedures.

Dietary manipulation

The dietary manipulation used was previously described for other mouse lines (12). Briefly, all mice were fed a normal chow diet until 10 weeks of age. Then the mice were maintained on experimental diets until the conclusion of the study at 21 weeks of age. Male and female GHA (male: n = 11, female: n = 9) and WT (male: n = 12, female: n = 9) mice were provided a HF diet, consisting of 20 g fat per 100 g (19 g butter oil and 1 g soybean oil) and providing 4.54 kcal/g of energy. Another group of male and female GHA (male: n = 9, female: n = 11) and WT (male: n = 12, female n = 11) mice were given a micronutrient-matched LF diet, consisting of 4 g fat per 100 g (3 g butter oil and 1 g soybean oil) and providing 3.8 kcal/g energy. Both experimental diets were pelleted and provided by Dyets (AIN-93M).

Weight gain, body composition, and food intake measurements

Throughout the dietary manipulation, body weight was measured weekly using a Mettler Toledo PL 202-S balance. A Minispec mq benchtop nuclear magnetic resonance analyzer (Bruker Instruments) was used to measure body composition weekly in live, conscious mice as previously described (8, 15). Measurements began at 10 weeks of age and continued until the end of the study at 21 weeks of age. The amount of food consumed was calculated weekly by subtracting the remaining food weight from the food provided at the beginning of the week.

Fasting glucose levels and glucose tolerance testing

At 20 weeks of age (wk 10 of the diet intervention) and after a 12-hour fast, a glucose tolerance test (GTT) was performed. Glucose levels were measured using a One Touch LifeScan glucometer and test strips (LifeScan). After a baseline reading, each mouse was injected ip with 0.01 mL/g body weight of 15% glucose diluted in PBS (1.5 g/kg body weight glucose). Blood glucose was measured 20, 60, 90, and 120 minutes after the injection. Blood glucose levels were reported as the area under the curve (AUC).

Tissue weights

All mice were dissected at the end of the 11-week feeding study. The mice were fasted for 13 hours. Before dissection, mice were placed in a CO₂ chamber until unconscious, after which blood was collected from the orbital sinus. Blood samples were immediately stored on ice and then incubated at room temperature for 20 minutes and centrifuged at 7000 × g for 10 minutes at 4°C to separate the serum. The serum was collected and stored at −80°C until testing. Five distinct adipose tissue (AT) depots [inguinal sc (SQ), perigonadal (peri), retroperitoneal (retro), mesenteric (mes), and interscapular brown adipose tissue (BAT)] and six organs (kidney, liver, heart, spleen, muscle, and lung) were collected and weighed. All tissues were then preserved by flash freezing in liquid nitrogen and stored at −80°C for later use. A portion of the SQ and peri samples was placed in formalin and later paraffin embedded and processed for histological analysis.

Adipocyte sizing

Adipocyte size was obtained for both the SQ and peri depots in all groups. Formalin-fixed samples were sent to AML Labs (Baltimore, Maryland) where they were paraffin embedded, sliced (5 μm), and stained with hematoxylin and eosin. Slides were examined using a Nikon Eclipse E600 microscope under ×200 magnification, and five nonoverlapping images were obtained for each sample with a SPOT RT digital camera. All cells in each field were sized using Nikon Elements computer software.

Serum insulin and leptin levels

Insulin levels were measured using the Mercodia mouse ultrasensitive insulin ELISA kit with a sensitivity of 0.025 μg/L and intraassay coefficients of variations of 2.5% (ALPCO Diagnostics). Leptin concentrations were measured using a mouse leptin ELISA kit with a sensitivity of 200 pg/mL and coefficients of variations of 3.6% (Crystal Chem, Inc).

Liver triglyceride measurements

Liver triacylglycerol (TAG) concentrations were measured using a triglyceride-GPO kit (Pointe Scientific), with calculations assuming the molecular mass of TAG as 0.885 kDa (885 g/mol) (8, 15).

Statistical analysis

All measurement data are presented as the mean ± SEM. Males and females were analyzed separately. A two-way ANOVA (2 × 2) was used to identify differences dependent on genotype and diet for measurements taken at a single time point. A two-way repeated ANOVA was used to identify differences in measurements taken over time. The Pearson correlation was used to analyze the relationship between percentage fat mass and AUC/fasting glucose levels. A three-way ANOVA (2 × 2 × 2) was used to measure statistical differences in adipocyte size due to genotype, diet, and depot. All data analysis was performed using SPSS (version 17.0). Differences were considered statistically significant at P ≤ .05.

Results

Body weight and weight gain

As expected, at the beginning of the study (10 wk of age), both male and female GHA mice had lower body weights than their WT controls, with GHA body weights ranging from 60% to 62% the weight of WT males and from 67% to 68% of WT females (Figure 1, A and B). Prior to diet manipulation, mice of the same sex and genotype had approximately equal body weights. All groups gained body mass throughout the study on both diets, although to varying degrees due to diet and genotype. Strikingly, GHA males on a HF diet increased their body weight by 94% after 11 weeks, nearly doubling their weight, whereas HF WT males increased their body weight by 53%. GHA males also gained more than their WT counterparts on a LF diet, increasing their weight by 42% compared with the WT gain of 17%. However, the trends for females were similar although less drastic than males; on the HF diet, GHA females gained 52% of their starting weight and WT gained 42%. GHA females on a LF diet also gained more than their WT controls, 27% vs 13%, respectively.

Figure 1.

Total body weight, fat mass, and lean mass comparison of male and female WT and GHA mice on HF and LF diets. Body weight and composition of GHA and WT mice over time. Shown is body weight (A and B), fat mass (C and D), and lean mass (E and F) of male and female GHA and WT mice on HF and LF diets over time. Data are expressed as mean ± SEM. Two-way repeated measures ANOVA test for body weight (A and B) found significant effects of genotype and diet in both males and females. For fat mass (C and D), there was a significant effect of diet in both sexes but not of genotype. Analysis of lean mass (E and F) revealed a significant effect of both genotype and diet in both sexes. Geno, genotype.

Body composition

All mice gained significant fat mass on a HF diet, although differences between genotypes were apparent (Figure 1, C and D). At the start of the study, GHA mice of both sexes had more than double the absolute fat mass of their WT counterparts. However, WT mice gained more fat on a HF diet, and at the end of the study, there was no difference in absolute fat mass between WT and GHA mice. The effect of diet on fat mass was not equivalent between sexes because male mice had more dramatic increases than females. Notably, on a HF diet, male WT mice increased their fat mass by 1373% (∼14 times their starting fat mass), whereas GHA males increased by 554%. In contrast, female WT and GHA mice on a HF diet increased their fat mass by 489% and 323%, respectively. Mice on the LF diet also gained fat mass over the course of the study but to a much lesser degree. Importantly, on a LF diet, genotype differences remained significant throughout the study, with GHA mice maintaining greater absolute fat mass than their WT controls.

Because GHA mice are dwarf compared with their controls, fat mass was also analyzed relative to total body weight. GHA mice of both sexes maintained greater relative fat mass than their WT controls throughout the entire study in both sexes. Thus, although both genotypes on a HF diet had equivalent absolute fat mass at the end of the study, GHA mice had a greater relative fat mass on both diets.

Significant effects on lean mass (Figure 1, E and F) were seen for both genotype and diet. As expected, WT mice had greater lean mass than GHA mice in both sexes. Additionally, all mice on the HF diet had a greater lean mass over the course of the study than those on a LF diet.

Food consumption

When food consumption (grams per mouse per week) was analyzed (Table 1), significant effects were found due to genotype and diet in both sexes. Specifically, GHA males ate less than WT males on a LF diet, but on a HF diet, there was no difference due to genotype. In contrast to males, WT and GHA females had similar food consumption on a LF diet; however, on a HF diet, WT female consumption was significantly higher than GHA females. All groups, except for GHA females, significantly increased their consumption on a HF diet compared with their LF controls. Results were similar when data were analyzed as kilocalories and when normalized as kilocalories per gram body weight.

Table 1.

Food Consumption and Liver TAG Levels for Male and Female GHA and WT Mice on HF and LF Diets

Sex	Genotype	Diet	Food Consumption, g/Mouse/wk	Liver TAG, mg/g
Male	WT	LF	19.42 ± 0.22a	7.635 ± 0.897a
		HF	21.41 ±. 0.92b	14.895 ± 2.928b
	GHA	LF	16.74 ± 0.22c	20.759 ± 2.802c
		HF	21.45 ± 0.88b	25.771 ± 3.823c
Female	WT	LF	15.54 ± 0.20a	14.724 ± 1.468b
		HF	19.0 ± 0.58b	16.946 ± 2.667b
	GHA	LF	15.52 ± 0.19a	17.295 ± 1.797b
		HF	15.55 ± 0.50a	16.130 ± 2.172b

Means in the same column without a common superscript letter are significantly different (P < .05).

Glucose metabolism and serum hormone levels

Fasting blood glucose levels were measured at 20 weeks of age (Figure 2, A and B). Mice of both sexes on the HF diet had higher glucose levels than those on the LF diet. Additionally, there was a significant effect of genotype in both sexes, with WT mice having higher glucose levels than GHA. In addition, WT mice had a blood greater glucose increase on a HF diet with WT males increasing 13%, whereas GHA males increased only 9%. Surprisingly, females of both genotypes had greater increases than their male counterparts, with WT and GHA females increasing 26% and 15% on a HF diet, respectively.

Figure 2.

Measures of glucose homeostasis and serum hormone levels in GHA and WT male and female mice on HF and LF diets. Shown are fasting blood glucose levels (A and B) and AUC (C and D) from a GTT at 20 weeks of age (wk 10 of the study). Also shown are fasting insulin levels (E and F) and fasting leptin levels (G and H) from blood taken prior to dissection at week 11 of the study. Data are expressed as mean ± SEM. For blood glucose (A and B), there was a significant effect of genotype and diet in males and females. For AUC (C and D), the only significant effect was the genotype × diet interaction in males. Concerning fasting blood insulin levels (E and F), significant effects were found only in males in which both genotype and diet differences were significant. With respect to fasting blood leptin levels (G and H), significant effects were found due to both genotype and diet in males but only to diet in females. Geno, genotype; NS, not significant.

A GTT was performed during the 10th week of feeding (Figure 2, C and D). No statistical differences were found due to genotype or diet, although males saw a significant effect of the genotype × diet interaction. However, an analysis of the difference between mice on HF and LF diets in genotype- and sex-matched groups revealed that WT mice specifically had an increase in the AUC. For this measure, males saw a far greater increase than females, increasing 38% vs 3% on a HF diet. GHA animals had a slightly lower AUC than WT mice with males and females 13% and 2% lower on a HF diet, respectively. In addition, using the entire data set, a positive relationship was found between fat mass and both fasting glucose and AUC (fat mass × glucose: r = 0.447, P = 2.26 E-5; fat mass × AUC: r = 0.286, P = .009).

Serum collected at dissection was used to analyze insulin and leptin levels (Figure 2, E–H). Both hormones saw significant effects due to genotype and diet in some groups. Specifically, only males saw significant differences in their insulin level due to both genotype and diet, with GHA male mice having lower insulin levels on both diets. Notably, insulin levels in GHA males on a HF diet were 170% greater than their LF counterparts, whereas HF WT males were 257% greater than LF WT males. Insulin levels in females were not significantly affected by diet or genotype. For leptin, the effect of genotype was seen only in males, with GHA mice having higher levels on a LF diet, whereas males and females of both genotypes had significantly increased leptin on a HF diet.

Tissue weights

As expected, the weight of each AT depot collected was significantly increased in all groups on a HF diet (Figure 3, A and B; significance is listed in Supplemental Table 1). Male GHA mice had significantly greater sc depot weights than WT males on both diets. All other depots were either smaller than WT (peri) on a LF or HF diet or not significantly different (retro, mes, and BAT). Females shared similar trends, although significant genotype differences were not found in any depot. Differences in body size could confound these data; thus, the AT depot weights were also normalized to total body weight. Interestingly, normalized sc weights were significantly greater for GHA mice in both males and females, following the trend seen in absolute weights. In contrast, no differences were found due to genotype in normalized peri depot weights, indicating that the differences seen in absolute weight could be attributed to disparate body sizes. A comparison of sc to peri weight ratios revealed a depot-dependent trend in adipose deposition between genotypes. For GHA males, the peri to sc weight ratio was nearly 1.0 on both diets (HF: 1.018; LF: 0.995), whereas the WT males had much larger peri depots and therefore higher ratios (HF: 1.797; LF: 2.181). Thus, the male GHA mice stored fat equally in the sc and peri depot, whereas the WT males preferentially stored in the peri depot. An analysis of this ratio in females revealed that GHA females had larger sc stores compared with the perigonadal depots (HF: 0.7; LF: 0.6), whereas the WT females shared the same trend as their male counterparts (HF: 1.5; LF: 1.7).

Figure 3.

Absolute fat depot and organ weights of male and female GHA and WT mice on HF and LF diets at 21 weeks of age. Absolute fat depot weights are shown for four WAT depots [SQ, perigondal (peri), retroperitoneal (retro), mesenteric (mes), A and B] as well as BAT. Also shown are organ weights of the liver, kidney, spleen, heart, muscle (quadriceps), and lung (C and D). Data are expressed as mean ± SEM. Univariate ANOVA tests showed that all tissues saw significant effects of diet except for lung in females. Significant effects due to genotype were found in all tissues except for retro, mes, and BAT in males and SQ, peri, retro, mes, and BAT in females. All fat depot and organ weight P values are listed in Supplemental Table 1.

The absolute weights of all organs collected were significantly greater on a HF diet except for female lung (Figure 3, C and D; significance is listed in Supplemental Table 1). In addition, all organ weights were significantly greater in WT mice as would be expected due to the differences in body size between genotypes (Supplemental Table 1).

Adipocyte size

The adipocyte cross-sectional area was measured in the sc and peri depots and analyzed by three-way ANOVA (Figure 4). Both sexes saw a significant increase in adipocyte size on a HF diet in both depots. Males also saw significant effects of genotype as well as depot, with GHA mice and the peri depot having greater adipocyte area. When the ratio of peri to sc adipocyte area was considered within each group, WT males had the largest disparity in size with peri adipocytes being 1.8 and 1.6 times the size of sc adipocytes in the LF and HF groups, respectively. Female WT mice also had larger peri adipocytes with both LF (1.6 times) and HF (1.5 times) feeding. Both male and female GHA mice had only slight differences in adipocytes size between depots on either diet regimen (Male GHA 1.1 for HF and LF, female GHA 1.0 on LF, 0.9 on HF).

Figure 4.

Male and female adipocyte sizing in GHA and WT mice on HF and LF diets. A three-way ANOVA in males revealed a significant effect of genotype [F (1, 79) = 16.65; P = 1.15E-4), diet (F (1, 79) = 69.81; P = 3.33E-12], depot [F (1, 79) = 13.86; P = 3.88E-4], and the interaction of genotype × depot [F (1, 79) = 4.47; P = .038]. For females, significant effect of diet [F (1, 57) = 9.74; P = .003] was seen.

Liver triglycerides

At dissection, GHA mice had either equal or significantly greater liver triglyceride levels than their WT controls, regardless of diet (Table 1). Particularly striking, the triglyceride concentration in LF-fed GHA male mice was 2.7 times that of the LF-fed WT males. Although the levels in both genotypes of male mice tended to increase on the HF diet, only the WT males were significantly different from their genotype-matched LF controls. Thus, the male GHA mice did not have a significant increase in triglycerides in response to the HF feeding. The GHA male levels of triglycerides remained significantly higher than their WT counterparts on both diet regimens. Female mice did not experience an effect on liver triglyceride levels due to either genotype or diet. However, females of both genotypes had higher baseline triglyceride levels than males.

Discussion

GHA mice are unique in that they do not share the well-established connection between decreased GH action and longevity. We hypothesized that a HF diet may be useful in exaggerating their obese phenotype and perhaps identify reasons for their normal lifespan. Our results show several interesting characteristics unique to GHA mice on a HF diet. Specifically, they experience increased white adipose tissue (WAT) accumulation in both the healthy sc and unhealthy perigonadal depots. Importantly, despite further increases to adiposity when placed on a HF diet, GHA mice were protected from poor glucose metabolism, indicating they continue to experience beneficial effects of reduced GH-induced signaling. Furthermore, males responded more dramatically to both decreased GH action and HF diet, with females having more modest reactions across all measures. Of note, we have unpublished data that the GHA molecule in our mice do not bind to the prolactin receptor (PRLR) and neither bovine GH nor pegvisomant, a GH antagonist drug, is reported to interact with the PRLR (16). Thus, our results are due to altered GH action and not due to interaction with PRLR.

The reduction or absence of GH action consistently results in smaller body size, reduced lean mass, and increased fat mass (7, 12, 15, 17). GHA mice follow this trend, although at older ages their body weights equal those of their WT counterparts due to fat mass accumulation (7). Thus, we expected that GHA mice on a HF diet would have dramatic body weight gains and that increases in their body weight would be due primarily to gains in fat mass. Although GHA HF male mice in this study did not reach the same body weight as their diet-matched controls, they did have significantly greater total body weight gains than controls: that is, GHA HF males almost doubled their body weight, whereas WT males saw body weight gains of approximately 50% over their starting weight. In contrast, GHR−/− male mice do not have a greater percentage body weight gains than their WT controls on a HF diet (12). Thus, GHA mice appear more susceptible to weight gain with HF feeding than GHR−/− mice.

In the current study, body composition analysis shows that the body weight gains in HF male mice were largely due to fat mass accumulation in both genotypes. In fact, although WT HF males began the study with lower absolute fat mass, by the end of the study, they had reached the same fat mass as GHA HF males. However, GHA mice maintain significantly greater relative fat mass on both diets, whereas the relative fat mass of GHR−/− mice compared with WT controls achieves significance only on a LF diet, although a positive trend was noted in HF feeding (12). That GHA mice maintain greater relative fat mass over controls, whereas GHR−/− mice do not, reinforces the conclusion that GHA mice are more sensitive to a HF diet than other mouse lines with a reduction or absence of GH action.

GH has been shown to have a depot-specific effect on AT. Specifically, mice with lowered or absent GH action preferentially accumulate fat mass in the sc depot, although, uniquely, GHA mice also gain in the intraabdominal depots with advancing age (7, 8, 12, 15, 17). GHR−/− mice on a HF diet favor fat accumulation in the sc depot, although they also gain significantly in the intraabdominal depots as well (12). Our current results followed this trend because GHA mice had greater sc fat mass and less perigonadal fat mass than WT mice but displayed no differences in other adipose depots in both sexes on a LF diet. Importantly, when comparing depot weights within the same groups, GHA male mice stored fat almost equally between the sc and perigonadal depots, whereas female GHA mice stored greater amounts in the sc depot. For WT mice, both sexes had greater perigonadal than sc depot weights. On a HF diet, WT mice increased preferentially in the perigonadal depot, whereas the GHA mice increased fat accumulation equally in both the sc and perigonadal depots. The trend was similar when comparing adipocyte size between depots.

Subcutaneous and intraabdominal fat depots have been shown to have different metabolic and inflammatory profiles, with intraabdominal, specifically visceral depots, having greater association with obesity-related dysfunctions (18). In fact, it has been argued that, in humans, the distribution of AT has a greater effect on disease risk than the amount of AT (19). Specifically, intraabdominal AT accumulation has been associated with greater risk of disease (20). GHA mice have greater intraabdominal fat accumulation than GHR−/− mice on both HF and LF diets. Thus, what may seem to be a subtle difference could play a key role in the differences in health status between these two mouse lines. In particular, the greater intraabdominal fat mass in GHA mice may produce an adverse metabolic environment and begin to explain the differences in life span between the two lines.

Previous studies on mice with reduced or absent GH action, such as GHR−/− and Ames dwarf mice, have shown that decreased GH induced signaling results in increased insulin sensitivity and low plasma glucose levels (2, 21–23). However, although GHA mice have decreased GH action, they do not enjoy the same protective benefits as other mouse lines as they age. That is, measures of glucose homeostasis are reported as either improved or no different from controls at younger ages (2, 7), but they become hyperinsulinemic at older ages (7). The results of the current study agree with previous findings in young mice in that significant differences were not found when comparing plasma fasting glucose, insulin, and glucose tolerance for GHA and WT mice on the LF diet (Figure 2, A–F). However, GHA mice are somewhat protected from the negative effects of the HF diet, although results are different for males and females, with measures of glucose homeostasis in males more dramatically altered on a HF diet than in females. Males of both genotypes saw a rise in serum glucose levels on a HF diet, although the WT mice had a greater increase than GHA.

More dramatic differences between genotypes were seen for the AUC and serum insulin levels because both increased greatly in WT males, whereas neither was significantly altered in GHA males. Thus, although GHA males do have some elevation in glucose with HF feeding, it does not result in impaired glucose tolerance or increased insulin levels. Similar to the GHA data presented here, male GHR−/− previously have been shown to maintain both low glucose and insulin levels compared with controls on a HF diet (12). Thus, male GHA mice also seem to have some protection from altered glucose metabolism on a HF diet but not to the same extent as GHR−/− mice. Both GHA and WT females seem to be more protected than males from the effect of diet because they saw only a small increase in glucose levels on a HF diet, whereas neither the AUC or insulin levels were altered. Unfortunately, females have not been included in previous diet studies of mice with reduced or absent GH action. Thus, sex differences in the glucose metabolism of other GH modified lines on a HF diet have not been established.

Leptin is positively correlated with fat mass and previous studies have shown that GHA, GHR−/−, and Snell mice have increased leptin levels (7, 9, 24, 25). Also, although GHA mice have high leptin at young ages like other lines of decreased GH action, uniquely, their leptin levels spike even higher at older ages (7). Our results show that leptin levels in both WT and GHA mice of both sexes increase on a HF diet, presumably due to increased fat mass. Importantly, GHA mice have higher levels than WT mice, regardless of diet. The high levels of leptin in GHA mice and other mouse lines with reduced GH action may be significant for the overall health of the animals. For example, leptin has been shown to provide a neuroprotective effect, which could confer protection from neurodegenerative diseases (26) as well as a having a prosurvival effect on neurons (27) and other cells such as pancreatic β-cells (28). On the other hand, hyperleptinemia also has a pathological role in the development of atherosclerosis (29). Finally, in obesity, high levels of leptin are generally concomitant with leptin resistance (30). Thus, although our mice experience high levels of leptin, they may not have effective leptin action. Further investigation is required to fully understand the role of leptin in GHA mice and other models of altered GH action.

Hepatic steatosis is often associated with insulin resistance. Counterintuitively, male GHR−/− mice have higher liver triglyceride levels than their WT littermates at younger ages (31) but are not different from WT mice at older ages (8). Conversely, Ames dwarf mice have decreased liver triglyceride levels at a young age (31). In GHA mice, a slight but insignificant decrease is observed in the liver triglyceride levels when compared with WT controls at advanced ages (7). Our results were not consistent with the previous GHA study in that GHA mice on either diet had higher triglyceride levels than all WT controls. Additionally, the HF diet induced only a significant increase in triglyceride levels in WT male mice, whereas the increase in GHA males between HF and LF did not reach statistical significance. Further studies are warranted to determine whether the reduction in GHR signaling in this line results in liver physiology, consistent with that of other lines with altered GH action.

The inclusion of females in this study is unique because previous diet manipulation studies on GH altered mice have been conducted only on males. Importantly, this study has shown that genotype and diet alterations do not have the same effect on males and females. In general, males had more dramatic responses to both variables or, perhaps a more interesting perspective, females respond more weakly to HF diet stimulus. The diet response reported here is supported by research in nontransgenic animals in which sex differences in response to diet intervention are observed across many species, including mice. For example, when C57BL/6J mice are fed a HF diet, both sexes experience weight gain and altered glucose metabolism, but males do so to a greater extent than females (32–34). Thus, males are not only more susceptible to weight gain but perhaps also to the complications of obesity (33). In addition to the well-documented differences in the WT males and females on a HF diet, it is important to note that the differences observed in this study could be attributed to the interaction of GH and gonadal steroids. Sex hormones are able to alter the secretion of GH centrally as well as GH responsiveness peripherally. Importantly, estrogen diminishes the actions of GH in the liver, resulting in decreased IGF-1 (35), whereas T amplifies GH actions by promoting increased secretion as well as IGF-1 production (36, 37). Thus, male mice should see a greater effect of GH action through IGF-1, whereas female mice should see less dramatic effects. This pattern is clearly seen in the results of our study. In addition, GH is known to alter the expression of sex-specific genes in the mouse liver, which may also contribute to altered responses between males and females (14). Overall, alterations due to both genotype and diet differences were less extreme in the female mice, potentially due to decreased GH action as a result of GH and estrogen interactions as well as differential liver gene expression.

In summary, this is the first study to investigate the impact of a HF diet on GHA mice and the only HF feeding study using mice with altered GH action that used both sexes. Our results demonstrated significant increases in total body weight, due primarily to fat mass accumulation, in all HF groups. Notably, WT mice accumulated fat preferentially in the WAT perigonadal depot, whereas GHA increased in both the WAT sc and perigonadal depots to a similar degree. The large increase in the less healthy intraabdominal depots may begin to explain GHA's lack of increased life span. However, GHA mice have lower fasting glucose and insulin levels as well as better glucose tolerance on a HF diet compared with WT mice, indicating that they may be somewhat protected from the effects of the HF feeding, despite fat mass gain. Thus, GHA mice are equally susceptible to fat mass gains on a HF diet but without a significant impairment in glucose homeostasis. Thus, our study provides valuable insight into how GH deficiency regulates adiposity and glucose metabolism in the context of HF diet. Importantly, the study also allowed for comparison between males and females. Generally, males had a more severe response to both decreased GH action and diet manipulation, highlighting the importance of including both sexes in studies on mice with GH signaling alterations.