Our study is the first report on chow-fed Ghsr−/−
mice during normal aging up to 26-months of age. Old Ghsr−/−
mice fed on regular chow have reduced fat mass, improved lipid profile (), and markedly improved insulin sensitivity (), similar to high fat-fed young adult Ghsr−/−
mice reported by others (Zigman et al. 2005
; Longo et al. 2008
). Interestingly, during hyperinsulinemic-euglycemic clamps, only young Ghsr−/−
mice exhibit increased glucose infusion, but not old Ghsr−/−
mice ( and Fig. S2.B
). At the same time, old Ghsr−/−
mice clearly showed improved insulin sensitivity during hyperglycemic and hypoglycemic clamps (), suggesting that the insulin-sensitive phenotype of old Ghsr−/−
mice is more pronounced under challenging conditions. We previously reported that ghrelin ablation increases insulin secretion as a result of lowering UCP2 (Sun et al. 2006
). Here, we demonstrated that ablation of GHS-R attenuates age-associated obesity and insulin resistance, maintaining a healthier metabolic profile. These data suggest that the ghrelin signaling pathway plays an important role in glucose homeostasis by regulating both insulin secretion and insulin actions.
With no difference in food consumption (), the body weight and fat mass of old Ghsr−/−
mice were lower compared to WT mice (), suggesting old Ghsr−/−
mice might have increased energy expenditure. Indeed, the old Ghsr−/−
mice exhibited higher energy expenditure (). Increased energy expenditure could result from increased physical (locomotor) activity, higher resting metabolic rate (RMR), and/or increased thermogenesis. Our results show that the locomotor activity is unchanged (). RMR is an important parameter in determining energy balance, which accounts for ~60% of total energy expended (Tentolouris et al. 2006
); and thyroid hormones are important regulators of RMR (Roti et al. 2000
mice have a higher RMR (), but total serum T3 and T4 levels were no different between old WT and Ghsr−/−
mice (Fig. S6
). This suggests that the elevated RMR of old Ghsr−/−
mice is not due to changes in circulating thyroid hormones, but we cannot preclude enhanced thyroid hormone activity at hypothalamus in the Ghsr−/−
mice (Lopez et al. 2010
). Thus, the increased energy expenditure observed in old Ghsr−/−
mice is likely due to increased thermogenesis.
RQ is an indicator of fuel preference. Previously, it was reported that ghrelin increases RQ in rats and mice (Tschop et al. 2000
; Theander-Carrillo et al. 2006
), and Ghsr−/−
mice showed reduced RQ under HFD feeding (Zigman et al. 2005
). Interestingly, in the old Ghsr−/−
mice fed normal chow, we detected higher RQ (), which indicates that they favor carbohydrate as a fuel substrate. The elevated RQ may be due to reduced lipid supply in the lean old Ghsr−/−
mice, thus forcing the mice to use more carbohydrate instead of fat. We detected increased GLUT4 in BAT, and increased glucose uptake in the muscle of the null mice (). Increased glucose uptake in BAT and muscle of Ghsr−/−
mice may contribute to the higher RQ. Old Ghsr−/−
mice have greater RCF (), which indicates the null mice have greater metabolic flexibility, consistent with the insulin-sensitive phenotype. Collectively, our data support that Ghsr
ablation increases energy expenditure, RMR, and metabolic flexibility during aging, thereby maintaining a youthful, insulin-sensitive metabolic state.
During the preparation of this manuscript, another group reported that suppression of GHS-R using anti-sense RNA activates BAT and decreases fat storage in rats, and the effects were more pronounced under high fat feeding (Mano-Otagiri et al. 2010
). Similar to our study, their GHS-R suppressed rats were lean and had increased energy expenditure. In contrast to our study, calorie intake and locomotor activity were increased in those rats. Our old Ghsr−/−
mice exhibited a lean and insulin-sensitive phenotype with increased energy expenditure, but neither their food intake nor their activity levels were changed (). Our data provide the first evidence that GHS-R regulates fat metabolism without affecting appetite or activity. Calorie restriction is the only proven intervention which prolongs lifespan. Low blood glucose, low triglycerides, and low insulin are the characteristic metabolic hallmarks of calorie-restricted subjects and centenarians (Barzilai et al. 1998
; Barzilai & Gabriely 2001
). Our Ghsr−/−
mice have all these youthful health traits. GHS-R antagonists may offer great promise for novel interventions to mimic calorie restriction without diet or exercise.
Increasing evidence has shown that adipocyte cell size, but not adipocyte cell number, is correlated with insulin resistance (O’Connell et al. 2010
). Larger adipocytes exhibit elevated lipolysis and secrete more FFA into the circulation, resulting in fatty acid toxicity in insulin-responsive organs (Raz et al. 2005
). Histological data revealed that decreased fat mass in old Ghsr−/−
mice was mainly due to the reduction of cell size, rather than decreased cell number (), which is consistent with the insulin-sensitive phenotype. We previously showed that Ghsr−/−
mice have reduced circulating glucose levels under fasting condition (Sun et al. 2008
). In the current study, the levels of circulating cholesterol, triglycerides, and fasting FFA levels were lower in the old Ghsr−/−
mice (). As shown in of epididymal fat and the Supplemental Table
of inguinal fat, the expression levels of glucose/lipid uptake genes (GLUT4, lipoprotein lipase and CD36) were dramatically down-regulated. Similarly, expression levels of lipogenic genes (aP2, FAS, Lipin 1) were also lower. These data suggest that reduced glucose/lipid uptake and lipogenesis could contribute to the reduced WAT mass of Ghsr−/−
mice. Also as shown in , GHS-R ablation also appeared to decrease adiposity by promoting lipid export (ABCG) and/or inhibiting lipid recycling (PEPCK). Thus, GHS-R ablation may reduce fat mass of WAT by regulating lipid uptake, lipogenesis, lipid storage, lipid recycling, and/or lipid export.
BAT plays an important role in energy expenditure by regulating thermogenesis; BAT mass and activity show severe impairment during aging (Pfannenberg et al. 2010
). It is intriguing that GHS-R expression was only detectable in BAT of old WT mice, but not young WT mice (); this suggests that GHS-R may play a unique role in thermogenic regulation during aging. Our results showed that while the weight of BAT was significantly reduced in old mice as expected, there was no difference in weight of BAT between WT and Ghsr−/−
mice at either young or old age (). UCP1 expression decreases with age, and correlates with age-associated thermogenic impairment. Remarkably, BAT of old GHS-R mice maintained a level of UCP1 similar to that of young mice (). The higher mRNA and protein expression levels of UCP1 and higher mitochondrial density detected in old Ghsr−/−
mice suggest that GHS-R ablation improves thermogenic function and protects against aging-associated decline of thermogenesis (). Indeed, we detected significantly higher core temperatures in the old Ghsr−/−
mice (). Body temperature is typically tightly regulated. The increased body temperature of old Ghsr−/−
mice is significant, because it has been shown that increased body temperature of about 1°C could thermodynamically correspond to a 10% increase in metabolic rate (Cannon & Nedergaard 2010
). Furthermore, old Ghsr−/−
mice have significantly reduced subcutaneous fat; reduced subcutaneous fat can reduce insulating capacity and lead to increased heat loss. Thus, the difference we detected may be understated due to increased radiated heat loss; the amount of heat generated in GHS-R null mice is likely to be much higher than that reflected by core body temperature.
In direct contrast to WAT, all lipid metabolic genes in BAT were up-regulated (). Glucose and lipid uptake (GLUT4, LPL and CD36) and lipogenesis (aP2) genes were significantly increased in the BAT of Ghsr−/−
mice, suggesting increased lipid uptake/synthesis. These results suggest that deletion of GHS-R may activate lipid synthesis machinery and increase lipid anabolic capacity in BAT, thus enhancing mitochondrial thermogenesis. Since elevated heat generation in BAT requires a huge amount of energy, lipid in BAT of the null mice may be quickly mobilized to generate heat. Thereby, even though lipid anabolic capacity is elevated in BAT of Ghsr−/−
mice, no increased lipid accumulation could be detected (Fig. S5
). Triglyceride and glucose are oxidizable substrates for thermogenesis. The elevated thermogenesis (heat production) in BAT of the null mice may subsequently diminish the lipid substrates in circulation, which may further promote fat mobilization in WAT, and then lead to reduced fat mass and improved insulin sensitivity.
Since our Ghsr−/−
mice are global knockout, the effects in adipose tissues we detected could be direct or indirect, central or peripheral. We detected down-regulation of the lipid metabolic genes in subcutaneous fat of old GHS-R null mice (Supplemental Table
), but we failed to detect GHS-R expression in subcutaneous fat of old WT mice (). This suggests that the lipid metabolic effect of GHS-R in subcutaneous fat is likely an indirect effect. On another note, we detected more pronounced body weight difference as mice aged when GHS-R is expressed in adipose tissues (); in the same time, we also detected modest body weight difference in young mice when GHS-R is hardly detectable in adipose tissues (Fig. and ). The body weight difference in young mice may not be explained by GHS-R expression in adipose tissues. GHS-R is expressed at a high level in the brain (hypothalamus) starting from a young age, and hypothalamus is also known to play a role in thermogenesis and energy homeostasis. It is possible that the weight difference in the young mice reflects the central effect of GHS-R.
Our data in the old mice showed that low levels of GHS-R were detectable in epididymal WAT and interscapular BAT ( and ). The expression of GHS-R1a in white and brown adipose tissues is low, which invites the question of whether the expression is real and physiologically meaningful. We believe that a low-level of expression could still have significant impacts on biological functions, depending upon the binding partners and signaling cascades. To determine whether the expression of GHS-R is age-dependent but not obesity-dependent, we studied the expression of GHS-R in epididymal fat and BAT of diet-induced obese mice. While our data clearly showed an elevated diet-induced thermogenesis in the BAT of (insert it) obese mice, the GHS-R expression in WAT and BAT remained unchanged (Fig. S4
). This further supports our conclusion that GHS-R expression is correlated with age but not obesity. Our result of GHS-R1a expression in visceral WAT of aging mice is consistent with reports in adipose tissues of old rats (Choi et al. 2003
; Davies et al. 2009
), and GHS-R is expressed at higher levels in rat epididymal fat than in subcutaneous fat (Davies et al. 2009
). We have shown that circulating ghrelin levels and mRNA expression of brain and pituitary GHS-R increase with age (Sun et al. 2007b
), which may suggest that there is an increase in “ghrelin resistance” during aging. GHS-R may be “turned on” in those fat depots during aging, and play a causative role in age-associated obesity and insulin resistance. In line with our observation, chronic i.v.
infusion of ghrelin into rats has been shown to induce a depot-specific increase in WAT mass, and ghrelin’s effect on adiposity has been shown to be attenuated in Ghsr−/−
mice (Choi et al. 2003
; Davies et al. 2009
). An animal model of GHS-R inducible systems turning GHS-R on or off during aging may provide further direct evidence as to whether the dysregulation of GHS-R plays a key role in fat metabolism during aging. Increased obesity and insulin resistance in mice during aging may be explained by the activation of the GHS-R pathway in the white and/or brown fat depots.
To determine whether GHS-R has direct effects on lipid metabolism in white and brown adipocytes, we studied expression of GHS-R1a in white adipose-derived cell line 3T3-L1. We were not able to detect GHS-R1a expression in either undifferentiated or differentiated 3T3-L1 cells (data not shown). Interestingly, GHS-R mRNA expression was readily detectable in both undifferentiated and differentiated brown adipose HIB1B cells (). Ghrelin strongly suppresses the expression of differentiation regulator PPARγ and thermogenic regulator UCP1 in HIB1B cells (). In contrast, GHS-R antagonist [D-Lys3]-GHRP-6 has opposite effects on PPARγ and UCP1, and abolishes ghrelin’s inhibitory effects (). This data is in line with our gene expression data of BAT in , and further supports our in vivo observation that GHS-R antagonism increases differentiation and thermogenic capacity of brown adipocytes. The data more importantly suggests that ghrelin, via GHS-R, may directly regulate lipid metabolism and thermogenesis in BAT.
Our data provide the first evidence that GHS-R is an important regulator of lipid metabolism during normal aging, and Ghsr ablation increases energy expenditure through up-regulation of thermogenic function. Increased GHS-R expression in fat during aging may promote impairment of thermogenesis, and contribute to aging-associated obesity and insulin resistance. Ghsr ablation prevents age-associated decline of thermogenic gene expression in BAT, and ghrelin/GHS-R has direct thermogenic effects in brown adipocytes, suggesting that GHS-R may directly regulate thermogenesis in BAT. Our data showed that GHS-R ablation has effects on both energy-storing WAT and energy-burning BAT: reducing adiposity in WAT to reduce energy substrates, and activating heat production in BAT to increase energy expenditure (). This unique property of GHS-R suggests that GHS-R antagonists may serve as new anti-obesity and anti-insulin resistance drugs shifting metabolic states from obesogenic to a thermogenic.
The schematic diagram of the role of GHS-R ablation in fat metabolism during aging