Address correspondence and reprint requests to Vania F. Prado and Robert Gros, Robarts Research Institute, Schulich School of Medicine & Dentistry, University of Western Ontario, London, ON N6A5K8, Canada. E-mails: email@example.com; firstname.lastname@example.org
Cholinergic neurons are known to regulate striatal circuits; however, striatal-dependent physiological outcomes influenced by acetylcholine (ACh) are still poorly under;?>stood. Here, we used vesicular acetylcholine transporter (VAChT)D2-Cre-flox/flox mice, in which we selectively ablated the vesicular acetylcholine transporter in the striatum to dissect the specific roles of striatal ACh in metabolic homeostasis. We report that VAChTD2-Cre-flox/flox mice are lean at a young age and maintain this lean phenotype with time. The reduced body weight observed in these mutant mice is not attributable to reduced food intake or to a decrease in growth rate. In addition, changed activity could not completely explain the lean phenotype, as only young VAChTD2-Cre-flox/flox mice showed increased physical activity. Interestingly, VAChTD2-Cre-flox/flox mice show several metabolic changes, including increased plasma levels of insulin and leptin. They also show increased periods of wakefulness when compared with littermate controls. Taken together, our data suggest that striatal ACh has an important role in the modulation of metabolism and highlight the importance of striatum cholinergic tone in the regulation of energy expenditure. These new insights on how cholinergic neurons influence homeostasis open new avenues for the search of drug targets to treat obesity.
Acetylcholine (ACh) is important for multiple physiological functions in the brain. In the striatum, cholinergic interneurons provide a continuous source of extracellular ACh that regulates dopaminergic and glutamatergic tone and the output of GABAergic medium-spiny neurons (MSN). Striatal cholinergic neurons are implicated in regulating several striatal-related behaviors including reward-related learning (Hikida et al. 2003), motor behavior (Pisani et al. 2007), and feeding (Hoebel et al. 2007). Indeed, increased ACh release in the striatum is observed during the last phase of feeding, suggesting that ACh is important in the onset of satiety (Mark et al. 1992; Avena et al. 2006). In addition, infusion of cholinesterase inhibitor in rats decreases feeding (Mark et al. 2011), which supports a role for accumbens ACh in satiety (Hoebel et al. 2007). Moreover, ablation of accumbens cholinergic neurons in rats increases feeding (Hajnal et al. 2000); and rats with striatal injection of a cholinergic toxin show increased lever pressing to receive food reward (Mark et al. 2011). However, contradictory findings have also been reported. Striatal infusion of muscarinic antagonists decreases feeding (Pratt and Kelley 2005), suggesting that ACh might be required for feeding behavior instead of satiety (Pratt and Kelley 2005; Pratt and Blackstone 2009). Moreover, M3 knockout mice show decreased feeding and substantial weight loss (Yamada et al. 2001).
Understanding the specific roles of ACh in brain regions is challenging given the multiplicity of receptors available and widespread cholinergic innervation. Although toxin-based approaches that ablate cholinergic neurons in specific brain areas have contributed to our understanding of possible functions of cholinergic neurons in behavior (Kaneko et al. 2000), they present limitations, including variable ablation of cholinergic neurons among individuals (Hanin 1996; McGaughy et al. 2000). Importantly, manipulations that ablate cholinergic neurons cannot separate the contributions of cotransmission of ACh and glutamate for behavior. Striatal cholinergic neurons express vesicular glutamate transporter 3 (VGLUT3) (Gras et al. 2008) and can release glutamate when stimulated, which activates NMDA receptors in MSNs (Higley et al. 2011).
We have developed a genetic approach to selectively interfere with ACh release by disturbing the expression of the vesicular acetylcholine transporter gene (VAChT) (Prado et al. 2006; de Castro et al. 2009b; Lima et al. 2010; Martins-Silva et al. 2011). VAChT provides the only mechanism for active storage of ACh in synaptic vesicles and therefore changes in VAChT expression have a direct impact on ACh release (Prado et al. 2006; de Castro et al. 2009a). We have used the Cre/lox system to selectively eliminate VAChT in the striatum, without affecting peripheral cholinergic tone (Guzman et al. 2011). Our previous experiments suggested that ACh and glutamate originating from cholinergic interneurons have separate roles in regulating striatal circuitry and behavioral function. Here, we investigated whether chronic decrease in striatal cholinergic tone could alter feeding behavior. We found that decreased striatal cholinergic tone does not modify food intake, but interestingly, mutant mice have reduced body weight. Furthermore, mutant mice show changes in different aspects of energy homeostasis and altered sleep, suggesting novel roles of ACh in regulating metabolism.
Materials and methods
All experiments were carried out in compliance with the Canadian Council of Animal Care (CCAC) guidelines for the care and use of animals. The protocol was approved by the University of Western Ontario Institutional Animal Care and Use Committee (protocol # 2008-127). All efforts were made to minimize the suffering of animals. Generation of VAChTflox/flox mice has been previously described (Martins-Silva et al. 2011). VAChTflox/flox were generated in a mixed background (B6/129/SvEv/NMRI) and then backcrossed to C57BL/6J for five generations. D2-Cre mice (Tg(Drd2-cre)44Gsat Stock# 017263-UCD) were obtained from the GENSAT project in a mixed background (FVB/B6/129/Swiss) and backcrossed to C57BL/6J for two generations. VAChTD2-Cre-flox/flox mice were generated by crossing VAChTflox/flox with the D2-Cre mouse line (Guzman et al. 2011). We then intercrossed VAChTD2-Cre-flox/wt to obtain VAChTD2-Cre-flox/flox mice. Mice used in this study were produced by breeding VAChTD2-Cre-flox/flox mice and VAChTflox/flox; they have a mixed background, although the C57BL/6J predominates. Only male mice were used in these studies. Animals were housed in groups of two to four per cage in a temperature-controlled room with a 14:10 light–dark cycle. Food and water were provided ad libitum.
Total RNA was extracted using the Aurum Total RNA for fatty and fibrous tissue kit (Biorad, Hercules, CA, USA) according to the kit manual. Quality analysis of extracted RNA was done by microfluidic analysis (Agilent 2100 Bioanalyzer, Agilent Technologies, Santa Clara, CA, USA). cDNA synthesis and qPCR analysis was performed as described before (Guzman et al. 2011). For each experiment, a non-template reaction was included as a negative control. Relative quantification of gene expression was done with the 2−ΔΔCT method using β-actin as the reference gene. Sequences for qPCR primers are available upon request.
VAChTD2-Cre-flox/flox mice and their control littermates were anesthetized, perfused transcardially with buffered 4% paraformaldehyde, and their brains removed and post-fixed overnight. Brains were sectioned in the coronal plane in a vibratome (40-μm thick). For immunohistochemistry, slices were washed with Tris-buffered saline (TBS) and peroxidase activity was quenched with 0.3% H2O2 in methanol. Section blocking was done by incubation for 1 h at 20°C with TBS with 0.3% Triton X-100, 3% bovine serum albumin, and 5% goat serum [Vectastain elite ABC kit (Vector Laboratories Inc., Burlington, ON, Canada)]. Sections were then incubated overnight at 4°C with rabbit anti-VAChT (1 : 200, Synaptic Systems Inc., Kitchener, ON, Canada). After several washes, brain slices were incubated with biotinylated goat anti-rabbit IgG. Immunoperoxidase staining was done using the Vectastain elite ABC kit and Vector SG substrate kit (Vector Laboratories) according to manufacturer instructions. Sections were mounted and counter-stained with cresyl-violet.
Mice body weight and length
The length of the animals from nose tip to tail base was measured at the end of the food-intake evaluation (day 7). Mouse body weight was measured once a week from day 33 to day 80. We also determined the weight of older mice (90- to 103-days old) and, as tibia length remains constant after maturity (Yin et al. 1982), the length of the two tibias was measured as a more accurate measurement of the animal size.
Assessment of indirect calorimetry, activity, and inactivity
Oxygen consumption, carbon dioxide production, respiratory exchange ratio, food and water intake, and physical activity were simultaneously measured for young and adult mice by using the Comprehensive Lab Animal Monitoring System (CLAMS) interfaced using Oxymax software (Columbus Instruments, Columbus, OH, USA). Mice were individually housed in the metabolic chambers maintained at 24 ± 1°C, and given free access to powdered standard rodent chow and water with airflow of 0.5 L/min. All the measurements were taken every 10 min for 24 h (12-h light/ 12-h dark) following a 16-h habituation period in the individual metabolic chambers. Volume of oxygen consumption (VO2) and carbon dioxide production (VCO2) measurements were normalized to body mass (mL/kg/h). Respiratory exchange ratio (RER), a measure of metabolism substrate choice (carbohydrate vs. fat), was calculated from VCO2/VO2. Pure carbohydrate oxidation is indicated by an RER of 1.0, and an RER of 0.7 indicates pure fat oxidation. Energy expenditure/heat (EE) (kcal/h) was calculated from RER and VO2 and corrected for body weight using the equation VO2 × (3.815 + 1.232 × RER).
Total activity, ambulatory activity, and sleep (periods of inactivity) were obtained using the Opto-M3 Activity Monitor and Oxymax software algorithms (Columbus Instruments, Columbus, OH, USA). Infrared photo beam breaks were monitored as counts on the X- and Y-axis and consecutive beam breaks were assessed as ambulatory movements. Periods of inactivity (sleep) were obtained using the sleep-detection algorithms within the Oxymax software and was based on 24 h (12-h light/12-h dark) of recorded activity in the X- and Y-axis. A bout of sleep was defined as 180-s epochs with a threshold of 10 counts of activity within each epoch. Total inactivity (sleep) time was calculated for the 12-h light and 12-h dark using the parameters described.
Glucose tolerance test
Animals were fasted for 5 h and then received an IP injection of glucose 2 g/Kg. Blood glucose levels were measured at 0 (base line), 30, 60, 90, 120, and 150 min after glucose injection.
Glucose levels were determined in blood samples obtained from a tail snip using ACCU-CHEK Advantage (Roche Diagnostics, Mannheim, Germany).
Serum levels of leptin and insulin
Blood samples were collected from the saphenous vein after the animals were fasted for 5 h. Blood samples were allowed to clot at 20°C. The clot was removed by centrifugation for 10 min at 2000 g. Serum was collected, aliquoted, and stored at −20°C until used. Leptin and insulin levels were measured by ELISA (Alpco diagnostics, Salem, NH, USA) following the manufacturer instructions.
Body composition analysis
Whole-body composition analysis was conducted using micro computed tomography (micro-CT) (Granton et al. 2010). Mice were anesthetized with 1.5% isoflurane in O2 and scanned using a Locus Ultra micro-CT scanner (GE Healthcare, London, ON, USA) in a single 16-s scan (80 kVp, 55.0 mA, 1000 views). Following reconstruction, the images (150-μm isotropic voxels) were scaled into Hounsfield Units (HU). The volume of adipose tissue was calculated as the total volume of voxels with intensity between −200 and −30 HU, while lean tissue was classified as tissue with intensities between −30 and 190 HU.
Data are expressed as mean ± SEM. Statistical analysis was performed using SigmaStat software (v. 3.5; Systat Software, Inc., Point Richmond, CA, USA). Comparisons between two groups were done with Student's t-test. If the data did not follow a normal distribution, a Mann–Whitney rank sum test was used. To analyze several experimental groups, we used a two-way anova with repeated measures, and when appropriated, a Tukey's post hoc comparison test was done.
VAChTD2-Cre-flox/flox mice are lean
Mice with selective elimination of ACh release in the striatum, VAChTD2-Cre-flox/flox mice, were obtained by tissue-specific deletion of the VAChT gene using the Cre-lox approach (Guzman et al. 2011). We previously showed that these mice have no release of ACh, whereas glutamate release is preserved in the striatum. VAChTD2-Cre-flox/flox mice displayed decreased body weight when compared with littermate controls (VAChTflox/flox). Young VAChTD2-Cre-flox/flox male mice (post-natal day 25) weighed about 12% less than their littermate controls [VAChTD2-Cre-flox/flox (n = 8): 10.10 ± 0.35, controls (n = 9): 11.50 ± 0.31; t(15) = 2.967; p < 0.05, Fig. 1a]. Adult VAChTD2-Cre-flox/flox mice also displayed reductions in body weight of about 12% [110-days old VAChTD2-Cre-flox/flox (n = 6): 23.09 ± 0.34, controls (n = 7): 26.82 ± 1.35; t(11) = 2.482; p < 0.05, Fig. 1a], indicating that decreased body weight was maintained throughout the entire observation period (~4 months; Two-way anova Genotype F(1, 227) = 56.16, p < 0.001; Time: F(13,227) = 63.62, p < 0.001; Interactions: F(13, 227) = 0.4700, p = 0.9396). Differences in weight were not attributed to changes in growth rate as VAChTD2-Cre-flox/flox mice and littermate controls did not show differences in body length (nose–anus), or tibia length (Fig. 1b and c). Moreover, the pronounced reduction in body weight displayed by the VAChTD2-Cre-flox/flox mice was not because of the expression of the Cre-enzyme, as this phenotype was not observed in D2-Cre transgenic mice (Figure S1).
VAChTD2-Cre-flox/flox mice show unaltered food intake
As striatal ACh has been suggested to modulate feeding behavior (Mark et al. 1992; Hajnal et al. 2000; Pratt and Kelley 2005), we investigated whether the reduction in body weight observed with VAChTD2-Cre-flox/flox mice was caused by altered food intake. Interestingly, we found that their slimness was not because of reduced food consumption, as both young and adult VAChTD2-Cre-flox/flox mice consumed as much chow daily as littermate controls (Table 1). Similarly, both young and adult VAChTD2-Cre-flox/flox mice showed comparable water consumption to littermate controls during the light and dark cycles (Table 1).
Table 1. Metabolic parameters of mutant mice and littermate controls
Ambulatory activity (Amb. act.) and Total activity (Total act.) were measured in counts/h. EE, energy expenditure; RER, respiratory exchange ratio.
Striatal cholinergic influence on feeding behavior is suggested to involve regulation of opioid peptides expression in MSN (Pratt and Kelley 2005). Thus, we compared mRNA expression levels of the striatal opioid peptides preproenkephalin (PPE) and preprodynorphin (PPD) in adult VAChTD2-Cre-flox/flox mice and littermate controls. Mutant mice showed unchanged levels of PPE and PPD mRNAs expression in the striatum (Figure S2), which is in agreement with our findings that VAChTD2-Cre-flox/flox mice show normal food intake.
Young VAChTD2-Cre-flox/flox mice show hyperactivity, hyperleptinemia, and hyperinsulinemia
Unchanged food intake coupled with decreased body weight suggested that metabolic rate may be altered in VAChTD2-Cre-flox/flox mice. We then evaluated energy balance in young and adult VAChTD2-Cre-flox/flox mice and littermate controls using metabolic cages. Mice were first habituated to the metabolic cages for 16 h and data were collected over the following 24 h.
We observed that young mutant mice (25–33 days) showed increased daily locomotor activity (24 h) when compared with littermate controls (t(13) = 3.235; p < 0.05), mainly because of increased locomotion during the dark cycle (Table 1; t(13) = 2.731; p < 0.05). In addition, young VAChTD2-Cre-flox/flox mice showed a significant increase in additional movements including grooming, sniffing, tail flicking, rearing, etc, both during the light (Mann–Whitney, T(13) = 4.0; p < 0.05) and dark cycles (Mann–Whitney, T(13) = 8.0; p < 0.05) (Total activity; Table 1). Accordingly, measurements of oxygen consumption, one of the parameters to evaluate energy expenditure, showed that young VAChTD2-Cre-flox/flox mice consumed significantly higher volume of oxygen during the dark period when compared with littermate controls (Table 1; t(13) = 2.839; p < 0.05). Thus, the augmented metabolic rate might be a consequence of the increased physical activity observed. In addition, the increased physical activity could also explain, at least in part, the decreased body weight observed in these young mutants. Their elevated metabolic rate was not a consequence of increased production of heat [Energy expenditure (EE); Table 1] or increased body temperature (VAChTflox/flox: 33.7 ± 0.5; VAChTD2-Cre-flox/flox: 33.3 ± 0.5). Moreover, young VAChTD2-Cre-flox/flox mice showed significant differences in respiratory exchange rate (RER) measurements throughout the light and dark cycles. Analyses of RER reflect the relative contributions of carbohydrate and fat oxidation to total energy expenditure. During the light cycle, when mice are mostly resting and have reduced feeding (basal metabolism), young mutants showed higher RER than control mice (Table 1; M = 0.9057 ± 0.005; t(13) = 4.585; p < 0.001), indicating an increased percentage of carbohydrate utilization in their energy production. During the dark period, when mice are mainly active and have increased feeding, young VAChTD2-Cre-flox/flox mice showed a higher fat utilization than littermate controls (Table 1, M = 0.9600 ± 0.009; t(13) = 2.619; p < 0.05).
To get further insight into the metabolic changes observed, we next examined whether the differences in energy balance in young mutants were accompanied by changes in plasma levels of glucose, glucose tolerance, leptin, or insulin (Fig. 2). Despite differences in both weight and in carbohydrate utilization during basal metabolism, differences in fasting plasma glucose levels were not detected between young mutants and littermate controls (Fig. 2a). However, we observed that young VAChTD2-Cre-flox/flox mice showed a mild, but significant improvement in glucose tolerance after 30 min of IP glucose injection (Fig. 2b,, repeated measures two-way anova showed differences in, genotype: F(1,45) = 5.093; p < 0.05, time: F(5,45) = 18.42; p < 0.001, and interactions: F(5,45) = 2.922; p < 0.05. Tukey's post hoc test showed a significant difference at 30 min, p < 0.001). Interestingly, young VAChTD2-Cre-flox/flox mice showed significantly increased levels of serum insulin (Fig. 2, t(8) = 3.085; p < 0.05) and leptin (Fig. 2d, t(7) = 3.083; p < 0.05). Increased serum insulin probably explains the improvement in glucose tolerance and could also explain the increased utilization of carbohydrate in their energy production during the light cycle, when their activity level is low. However at night, when mice are mostly active, these mutants used more lipids in their energy production, suggesting that they are more responsive to the increased plasma leptin level. Moreover, the increased levels of these two hormones could explain the observed increase in energy expenditure. These data indicate that striatal ACh is important for the regulation of different aspects of energy generation/expenditure.
Adult VAChTD2-Cre-flox/flox mice show hyperinsulinemia, but not hyperleptinemia and hyperactivity
Interestingly, not all the metabolic changes observed in young mutants were also observed in adult VAChTD2-Cre-flox/flox mice. One striking difference is the fact that adult VAChTD2-Cre-flox/flox mice showed no significant difference in locomotor or total activity when compared to littermate controls (Table 1). Importantly, adult VAChTD2-Cre-flox/flox mice also do not show increased locomotion when tested in a novel environment context using locomotor activity boxes (Guzman et al. 2011). Even though adult mutants and controls showed similar physical activity, adult VAChTD2-Cre-flox/flox mice showed increased metabolic rate measured as significantly higher oxygen consumption during the dark period (Table 1, t(14) = 2.444; p < 0.05). The increased metabolic rate did not seem to have been used to generate additional heat, as adult mutants and controls do not show difference in body temperature (VAChTflox/flox: 34.9 ± 0.2; VAChTD2-Cre-flox/flox: 34.5 ± 0.3) or in energy expenditure (Table 1). These results suggest that different factors contribute to the decreased body weight observed in young and adult VAChTD2-Cre-flox/flox mice. In addition, VAChTD2-Cre-flox/flox mice showed increased RER both during the dark (M = 0.9263 ± 0.010, t(14) = 2.375; p < 0.05) and light (M = 0.8713 ± 0.017, t(14) = 2.435; p < 0.05) cycles, suggesting that these mutants show a higher percentage of carbohydrate utilization in their energy production (Table 1).
To further investigate the origin of reduced weight of VAChTD2-Cre-flox/flox mice, body composition was studied using a high-resolution X-ray computed tomography (micro CT) system (Granton et al. 2010; Detombe et al. 2012). Micro-CT analysis was performed in pairs of adult littermates that were housed together so that they experienced identical living conditions. The analysis indicated that the decreased body weight in VAChTD2-Cre-flox/flox mice resulted from decreased body fat volume, while lean tissue volume did not differ between mutant and littermate controls (Fig. 3a and b; one-tailed t-test; p < 0.05). In addition, adult mutants did not show any difference in either blood glucose levels or glucose tolerance when compared to littermate controls (Fig. 3c and d). Serum insulin levels were significantly higher in adult mutant mice (Fig. 3e, t(9) = 3.009; p < 0.05); however, leptin levels did not differ between adult VAChTD2-Cre-flox/flox mice and littermate controls (Fig. 3f). These results suggest that the increased utilization of carbohydrate in their energy production, both during the light and dark cycle, results from elevated serum insulin levels.
Young and Adult VAChTD2-Cre-flox/flox mice show increased wakefulness periods
Because young VAChTD2-Cre-flox/flox mice showed increased total activity during their inactive period, we investigated their sleep/wake pattern based on inactivity/activity assessments obtained with the metabolic cages, an analysis shown to highly agree with electroencephalogram (EEG) recordings (Pack et al. 2007). Interestingly, mean values for the amount of sleep time differed markedly between VAChTD2-Cre-flox/flox mice and littermate controls, independently of their age (Fig. 4a and b, M(VAChTD2-Cre-flox/flox) = 295.4 ± 27.17, and M(controls) = 535.5 ± 31.71, t(13) = 5.660; p < 0.001, for young mice. For adult mice, M(VAChTD2-Cre-flox/flox) = 458.4 ± 20.33 and M(controls) = 644.1 ± 24.57, t(14) = 5.824; p < 0.001), suggesting that both young and adult VAChTD2-Cre-flox/flox mice sleep less than littermate controls.
VAChT expression is preserved in the hypothalamus of VAChTD2-Cre-flox/flox mice
Because VAChTD2-Cre-flox/flox mice show a number of metabolic changes, we investigated whether VAChT expression in this line was preserved in the hypothalamus, a brain region important for the regulation of energy homeostasis (Elmquist et al. 1999; Schwartz et al. 2000). qPCR analysis showed that levels of VAChT mRNA expression in the hypothalamus of adult VAChTD2-Cre-flox/flox mice is not changed when compared to littermate controls (Fig. 5a). In addition, immunostaining analysis of hypothalamic sections from adult VAChTD2-Cre-flox/flox mice and littermate controls using a VAChT antibody showed a very similar pattern of labeling (Fig. 5b and d), further indicating that hypothalamic expression of the VAChT protein is not changed. Staining with this VAChT antibody is specific, as it is suppressed in striatal sections from VAChTD2-Cre-flox/flox mice (Fig. 5c and e). These results demonstrate that the metabolic changes observed in VAChTD2-Cre-flox/flox mice result from elimination of cholinergic tone in the striatum and are not because of decreased cholinergic tone in the hypothalamus.
Hypothalamic neuropeptide mRNA expression in VAChTD2-Cre-flox/flox mice
Because VAChT mutants showed altered plasma levels of insulin (young and adult mutants) and leptin (young mutants), we investigated the expression of the hypothalamic neuropeptides proopiomelanocortin (POMC), agouti-related peptide (AGRP), prepro-orexin, and melanin-concentrating hormone (MCH); known regulators of feeding and energy balance that act downstream of the leptin/insulin system (Elmquist et al. 1999; Schwartz et al. 2000; Yeo and Heisler 2012). Hence, as a measurement of insulin and leptin signaling, we determined the mRNA levels for these peptides. Surprisingly, despite increased circulating levels of leptin and insulin, young mutants showed a trend to express decreased levels of POMC (t(9) = 1.759; p = 0.1125) and increased levels of orexin (t(8) = 1.736; p = 0.1207) (Fig. 6), a pattern more similar to what is observed in animals with low levels of circulating leptin (Bi et al. 2003). In adult VAChTD2-Cre-flox/flox mice, despite increased plasma level of insulin, mRNA expression for all four peptides was identical to that of controls (Fig. 6). These results may suggest that leptin/insulin signaling is altered in the hypothalamus of VAChTD2-Cre-flox/flox mice.
Here, we report that mice with selective elimination of VAChT in the striatum are lean at a young age and maintain this lean phenotype over time. The reduced body weight observed in these mutant mice resulted from decreased body fat content, but was not attributed to reduced food intake or because of a decrease in growth rate. Moreover, altered activity of the mutant mice could not completely explain the lean phenotype, as only young VAChTD2-Cre-flox/flox mice showed increased physical activity. Interestingly, plasma levels of insulin and leptin were elevated in VAChTD2-Cre-flox/flox mice. In addition, both young and adult mutants showed significantly increased wakefulness periods.
Elimination of striatal ACh does not change food intake
Although a number of studies in the literature suggest that striatal ACh is important for feeding behavior, the exact role of ACh in the modulation of feeding is not completely understood (Avena and Rada 2012). Our data suggest that striatal ACh is not essential for feeding as mice with no ACh release in the striatum do not show significant changes in food intake. Paradoxically, VAChTD2-Cre-flox/flox mice show decreased body weight. Similarly, rats that had striatal cholinergic neurons ablated by injection of a cholinergic neurotoxin (Hajnal et al. 2000) showed rapid decrease in body weight immediately after surgery and kept the lean phenotype during the entire investigation (10 weeks).
Unexpectedly, young mutants showed increased circulating leptin and insulin, and adult mutants showed increased insulin. These two hormones are known to circulate at levels proportional to carbohydrate and body fat abundance, respectively. They provide signals to the brain to activate neural pathways that reduce food intake and enhance energy expenditure (Schwartz et al. 2000). How elimination of striatal ACh influences release of insulin and leptin needs to be determined. It might involve changes in autonomic tone as activity of central cholinergic pathways seems to regulate central autonomic outflow (Nonogaki 2000; Rossi et al. 2011).
Leptin and insulin trigger inhibition of food intake by stimulating POMC neurons (Schwartz et al. 1997; Benoit et al. 2002) and inhibiting AGRP neurons in the hypothalamus (Schwartz et al. 2000; Plum et al. 2009; Yeo and Heisler 2012). Increased POMC signaling stimulates post-synaptic cells, including MCH neurons, and causes inhibition of appetite. Inactivation of AGRP neurons decreases stimulation of post-synaptic orexin neurons, further contributing to decreased appetite (Schwartz et al. 2000; Yeo and Heisler 2012). The unaffected food intake associated with hyperleptinemia and/or hyperinsulinemia observed in VAChTD2-Cre-flox/flox mice suggest that leptin/insulin pathways controlling food intake in these mutants might be defective. Indeed, our observation that expression levels of hypothalamic neuropeptides POMC, AGRP, prepro-orexin, and MCH do not change in both young and adult mutants suggests that hypothalamic neurons are not responsive to elevated plasma levels of leptin and/or insulin. In addition, despite the high plasma insulin levels, adult VAChTD2-Cre-flox/flox mice do not show differences in fasting glucose levels or glucose tolerance when compared to littermate controls, suggesting that these mutants might also show reduced peripheral sensitivity to insulin. In fact, evidences suggest that hyperinsulinemia, especially in the basal state, can lead to insulin resistance (Shanik et al. 2008; Dankner et al. 2009). This possibility is intriguing and we are currently starting a series of experiments to fully test this phenotype, by performing a thorough investigation on insulin signaling in these mutant mice.
Striatal ACh is important to control energy expenditure
A key finding in our study is that elimination of ACh release in the striatum of VAChTD2-Cre-flox/flox mice leads to increased energy expenditure, which might be mediated by elevated circulating levels of insulin and leptin observed in these mutants. Interestingly, it has been suggested that brain pathways mediating leptin and insulin control of food intake and energy expenditure are dissociated; paraventricular hypothalamus and/or amygdala are important to control food intake (Balthasar et al. 2005), while cholinergic sympathetic pre-ganglionic neurons of the spinal cord seem to contribute to energy expenditure control (Rossi et al. 2011). Our data suggest that striatal cholinergic interneurons may also take part of the energy expenditure control center.
Striatal ACh is important to control sleep
Young and adult mutant mice showed significantly increased wakefulness periods than littermate controls when assessed based on inactivity/activity. Importantly, in mice, sleep/wakefulness analysis based on inactivity/activity assessments obtained with metabolic cages highly agree with electroencephalogram (EEG) and electromyogram (EMG) recordings (Pack et al. 2007). The involvement of the striatum in the control of sleep/wakefulness has received very little attention (Qiu et al. 2010). Some old studies investigated the involvement of ACh in the process (Hernandez-Peon et al. 1967; Gadea-Ciria et al. 1973). A study using parallel measurements of ACh release within the cat striatum with EEG analysis showed that striatal ACh release is greatly dependent on the state of the sleep cycle, being elevated during slow wave sleep (SWS) (Gadea-Ciria et al. 1973). Moreover, cats that had ACh applied to the striatum fell asleep 30s to 5 min after the application (Hernandez-Peon et al. 1967). These results suggest that striatal ACh is an important regulator of sleep/wakefulness cycles. The observation that VAChTD2-Cre-flox/flox mice sleep less than littermate controls further supports a role for ACh in sleep. A systematic investigation of the sleep behavior in these mice is necessary to completely understand how the sleep/wake cycle is disrupted. Sleep disturbances are frequently observed in patients suffering from Huntington's as well as Parkinson's disease (Mena-Segovia et al. 2002; Comella 2007). Moreover, in Huntington's disease patients, where loss of ACh and choline acetyltransferase in the striatum has been reported, changes in sleep include diminished sleep efficiency and decreased SWS (Mena-Segovia et al. 2002). It is tempting to speculate that changes in ACh release contribute to the altered sleep pattern observed in these disorders.
Striatal ACh is important to control movements at a young age
Regulation of striatal circuits is highly influenced by levels of ACh released by cholinergic interneurons, dopamine (DA) from midbrain dopaminergic projections, and glutamate from cortical and thalamus neurons (Palmiter 2007). However, the specific role of each of these neurotransmitters in the regulation of the different physiological outcomes influenced by the striatal circuits is not clear. Recently, using the same VAChTD2-Cre-flox/flox mouse line used in this study, we suggested that in adult mutants, glutamate released from striatal cholinergic interneurons may be more important to regulate locomotor responses than ACh (Guzman et al. 2011).
Interestingly, absence of striatal ACh caused hyperactivity in young mutants, suggesting that ACh is important to control movements at a young age. After birth, the striatum of rodents undergo maturation processes that continue for several weeks (Tepper et al. 1998). Noteworthy, electrophysiological studies indicate that functional maturation of glutamatergic excitatory mechanisms on GABAergic projection neurons occurs only 5 weeks after birth (Tepper et al. 1998). These data suggest that at a young age, the influence of glutamate (either from glutamatergic neurons or from cholinergic neurons) on the striatum is limited. Our data suggest that prior to complete development of striatal circuits, ACh plays an important role in movement control. Taken together, our results suggest that the same physiological output may be regulated by different neurotransmitters depending on the developmental stage of the brain.
The comprehensive analysis of VAChTD2-Cre-flox/flox mice presented here revealed that striatal ACh has a much more important role in the modulation of homeostasis than previously anticipated, and highlights the importance of striatum on the regulation of energy homeostasis. Furthermore, our data indicated that striatal ACh is an important regulator of sleep. These new insights on how the central nervous system regulates homeostasis might open new avenues for the search of novel targets to treat obesity.
The authors declare no competing financial interests. We thank Joy Dunmore-Buyze and Sarah Detombe for assistance with image acquisition. This work was supported by the Canadian Institutes of Health Research (CIHR), the Heart and Stroke Foundation of Ontario (Grant NA 6656), Canadian Foundation for Innovation (CFI), and the Ontario Research Fund (ORF). M.S.G. received a post-doctoral fellowship from the Heart and Stroke Foundation of Canada. R.G. is supported by a New Investigator Award from the Heart and Stroke Foundation of Canada. M.D. is a Career Investigator of the Heart and Stroke Foundation of Ontario.