Chemogenetic manipulation of parasympathetic neurons (DMV) regulates feeding behavior and energy metabolism

Parasympathetic nervous system (PNS) innervates with several peripheral organs such as liver, pancreas and regulates energy metabolism. However, the direct role of PNS on food intake has been poorly understood. In the present study, we investigated the role of parasympathetic nervous system in regulation of feeding by chemogenetic methods. Adeno associated virus carrying DREADD (designer receptors exclusively activated by designer drugs) infused into the target brain region by stereotaxic surgery. The stimulatory hM3Dq or inhibitory hM4Di DREADD was over-expressed in selective population of dorsal motor nucleus of the vagus (DMV) neurons by Cre-recombinase-dependent manners. Activation of parasympathetic neuron by intraperitoneal injection of the M3-muscarinic receptor ligand clozapine-N-oxide (CNO) (1 mg/kg) suppressed food intake and resulted in body weight loss in ChAT-Cre mice. Parasympathetic neurons activation resulted in improved glucose tolerance while inhibition of the neurons resulted in impaired glucose tolerance. Stimulation of parasympathetic nervous system by injection of CNO (1 mg/kg) increased oxygen consumption and energy expenditure. Within the hypothalamus, in the arcuate nucleus (ARC) changed AGRP/POMC neurons. These results suggest that direct activation of parasympathetic nervous system decreases food intake and body weight with improved glucose tolerance.


Introduction
The number of people with obesity and diabetes has been increasing worldwide. Obesity and diabetes are recognized as common diseases and causing not only healthcare but also economic burdens. Although many various research and treatment methods have been suggested and tested for obesity and diabetes [8,14], it is still far from complete cure of those diseases. Autonomic nervous system plays a crucial role in the regulation of energy homeostasis [7,9,13]. To date, its role of appetite regulation is increasingly recognized because it malfunctions resulted in obese phenotypes. These mechanisms involve a complex interplay between central and peripheral nervous systems including both afferent and efferent vagus nerve fibers [2,21]. Previous works revealed several different hypothalamic regions such as arcuate nucleus (ARC), ventral medial hypothalamus (VMH), lateral hypothalamus (LH) modulate autonomic neural flow to peripheral organs such as liver, pancreas, and fat [10][11][12][13][14]24]. Both efferent central neuronal and afferent peripheral signals converged on dorsal motor nucleus of the vagus. DMV encompasses the nucleus tractus solitaries and modulates autonomic nervous system which regulates feeding behavior and energy metabolism [12,19]. Thus, DMV is a critical node in autonomic nervous system and its downstream VN fibers provide a potential therapeutic target for anti-obesity treatments. In fact, vagus nerve electric stimulator has been approved for human anti-obesity treatment option by the Food and Drug Administration (FDA) [4]. Compared to the role of sympathetic nervous system on feeding and energy metabolism, the role of parasympathetic nervous system on feeding and energy expenditure is unclear. To examine specifically in vivo effects of efferent parasympathetic nervous signals on feeding and energy expenditure, we have generated mouse models of temporal activation or inhibition of DMV regions by chemogenetic approach using DREADD techniques. DREADD techniques involve an artificial membrane receptor expression on target cells and receptor-specific artificial ligand administration for temporal modulation of the system [16][17][18][19]. Our hypothesis was that specific activation/inhibition of the vagus nerve can modulate feeding behavior and energy expenditure. We also used Cre recombinase expressing genetically modified mouse models such as ChAT-Cre to improve the specificity of DREADD receptor expression on parasympathetic motor neurons. ChAT-Cre mice show Cre expression in preganglionic parasympathetic neurons of the DMV, intermediolateral nucleus of the cholinergic neurons [27]. Here we investigated whether the specific changes in parasympathetic nervous system could be alternative treatment options for obesity and diabetes. Our results provide that possible therapeutic roles of parasympathetic modulation in anti-obesity and diabetes.

Animals
ChAT-Cre mice on a C57Bl/6 genetic background were generated as previously described [25,26]. This mouse line was Cre-recombinase expression is controlled via an IRES-Cre sequence was inserted in the genome downstream of the ChAT gene stop codon. The Cre allele was detected using the following primers: 5′-GTTTGCAGAAGCGGTGGG-3′ (M336), 5′-GATAGATAATGAGAGGCTC-3′ (M337), and 5′-AGATAGAT AATGAGAGGCTC -3′ (M338). All animal Procedures were conducted in accordance with The Institutional Animal Care and Use Committee of the Seoul National University and Seoul National University Hospital Institute of Biomedical Research, Seoul, Korea. Mice weighing 23∼25 g were maintained in individual cages under controlled temperature (21∼23°C) and light (light on 8:00, off at 20:00) with free ad libitum access to food and water.

AAV vectors for hM3dq or hM4di expression
The stimulatory DREADD, designated "hM3Dq", couples through the Gq pathway to depolarize neurons [1,4]. The hM3Dq and hM4Di coding sequences were cloned into a mCherry vector [11,15] upstream of the mCherry sequence to generate C-terminal mCherry fusion proteins. The hM3Dq-mCherry and hM4Di-mCherry coding sequences were amplified by PCR, and the amplicons and a Cre-inducible AAV vector with a human Synapsin 1 promoter [3]. We were hM3Dq and hM4Di AAV-virus purchased from addgene.

Body weight and food intake
Mice were housed in individual cages, and body weight and food intake were determined dark phase (8 A.M -8 P.M) and light phase (8 A.M -8 P.M). Preweighed food was placed in the food hoppers and measured on a per-cage basis. Food intake was determined as grams consumed per day.

Metabolic cage studies
Mice were placed into an 8-cage Comprehensive Laboratory Animal Monitoring System (CLAMS; Columbus Instruments, Columbus, OH, USA) on the experimental day, and given a 48 h to acclimatize. On experiment day after IP administration of CNO (1 mg/kg) measured both groups. O 2 consumption (VO 2 ), CO 2 consumption (VCO 2 ), Respiratory exchange ratio (RER), heat production and locomotor activity, food intake, water intake were monitored every 10 min during the 48 h period at room temperature. Cages were opened and calculations were stopped for 1 h between 9:00 and 10:00 am daily for replenishing food, measuring body weight, and performing injections. Energy expenditure calculated energy expenditure according to the following formula provided by the manufacturer: energy expenditure (kcal)=(3.815 + 1.232VO 2 /VCO 2 ) xVO 2 . On final day, Mice were removed from the cages.

Intraperitoneal glucose tolerance test (GTT)
Glucose tolerance tests were 12-14 wk old male ChAT-Cre mice that were being maintained on a normal chow diet were obtained from the Seoul National University and Seoul National University Hospital Institute of Biomedical Research. After an overnight fast (16-18 h), On the day of experimentation, Mice were intraperitoneal (i.p) glucose (1 mg/kg) administration. Blood samples were drawn from the tail vein immediately prior to CNO treatment and at 0, 15, 30, 60, and 120 min for plasma glucose measured using a Glucometer (Accucheck). For plasma insulin, blood samples (50 or 100ul) were collected from the tail vein into EDTA-coated tube 10-minute Glucose loading, immediately centrifuged, and the plasma was separated and stored at −20°C until assayed.

Immunohistochemistry
I.p. administration was performed 60 min before perfusion in overnight (16∼18 h) fasted mice. Mice were anesthetized by an intraperitoneal administration of sodium ketamine (80 mg/kg) with xylazine (10 mg/kg). The brain was removed and post-fixed overnight with 4% paraformaldehyde in phosphate buffer saline (PBS). And, the samples were subsequently cryo-protected in 0.1 M PB containing 20% sucrose. The brain embedded in Tissue-Tek OCT compound (Sakura Finetek Japan, Tokyo, Japan), were section into 14㎛ thickness and were prepared on a freezing microtome (Leica CM3050S; Leica, Nussloch, Germany) chilled at -20℃. For immunofluorescence staining, sections were blocked in 3% normal donkey serum (Sigma Aldrich, CA, USA) for 1 h incubated at room temperature, and then incubated 48 h at 4℃ with goat AGRP antibody (1:1000; Abcam, UK), rabbit β-endorphin (a product of POMC) antibody (1:3000; Phoenix Pharmaceuticals, Inc. CA). Slices were washed with PBS and incubated with Alexa Fluor 488labeled anti-rabbit or Alexa Flour 555-labeled anti-goat (1:500; Invitrogen, Carlsbad, CA) at room temperature for 1 h. Nuclei in brain sections were identified by staining four, 6-diamidino-2-phenylindole (DAPI) (Molecular Probes, Eugene, OR). For fluorescent section images, an Olympus fluorescence microscope (Olympus, Tokyo, Japan) was used. For quantitative histological analysis of POMC and AGRP neurons was manually counted using Image J software (NIH) at a magnification of 40X for three sections. For the immunofluorescence staining, 4 mice per group were analyzed.

Electrophysiology study
ChAT-Cre mice injected with pAAV-hSyn-hM3D(Gq)-mCherry viruses were used activation of DMV neruons upon CNO administration. Changes in DMV neuron firing rates were recorded from DMV-mCherry + neurons in the brain stem DMV were measured by silicon neural probe.

Statistical analysis
Data are expressed as means ± s.e.m. The level of statistical significance was determined using paired t-test when the difference between the means of two populations was considered or A two-way repeated ANOVA (Drug x Time as repeated measures) analysis of variance was used to study the effect of hM3Dq and hM4Di injection of CNO in ChAT-Cre mice.

Selective activation/inhibition of parasympathetic neurons
We used a Cre-recombinase-dependent adeno-associated virus (AAV) to express either hM3Dq, hM4Di receptor for inhibition of DMV complex neurons in ChAT-Cre mice (Fig. 1A, B). The receptors are fused to mCherry fluorescent protein so that virus-mediated the receptor expression could be monitored. We have confirmed that mCherry fluorescent protein expression was detected exclusively in the DMV regions in both ChAT-Cre (Fig. 1C). Additionally, I.p. injection of clozapine-N-oxide, a specific ligand for DREADD receptor, activates the receptor and initiates the neuronal firing rate and generate action potentials for hM3dq receptor and it has been recorded by multichannel recording silicon probe (Fig. 1D). The activation of parasympathetic neurons are confirmed by immunohistochemistry staining methods (Fig. 1E)

Effect of parasympathetic modulation on feeding, blood glucose levels
We investigated the effect of acute modulation of parasympathetic nervous system on food intake. The activation of parasympathetic nervous system decreased food intake in ChAT-Cre mice ( Fig. 2A-C). The reduction in food intake was significantly different compared to the control group and last up to 4 h after CNO i.p injection ( Fig. 2A). In contrast, the inhibition of parasympathetic nervous system significantly but only briefly increased food intake in ChAT-Cre (Fig. 2E). We next examined the effect of parasympathetic nervous system on glucose homeostasis. After activation or inhibition of parasympathetic neurons, we measured blood glucose levels during intraperitoneal glucose tolerance test in ChAT-Cre. The activation of parasympathetic nervous system in ChAT-Cre mice shows improved glucose tolerance (Fig. 2G) while the inhibition of parasympathetic nervous system results in impaired glucose tolerance (Fig. 2H).

Effect of parasympathetic modulation on energy expenditure
Food intake and energy expenditure is a critical factor for body weight maintenance. To examine metabolic phenotypes including energy expenditure, individual mouse is acclimatized and housed in a single metabolic cage from Comprehensive Laboratory Animal Monitoring System (CLAMS; Columbus Instruments, Columbus, OH, USA). We measured oxygen consumption (VO2), Energy expenditure (EE) to study the effect of parasympathetic nervous system (Fig. 3). Energy expenditure calculated energy expenditure according to the following formula provided by the manufacturer: energy expenditure (kcal)=(3.815 + 1.232VO 2 /VCO 2 ) xVO 2 . In Chat-Cre mouse models, modulation of parasympathetic nervous system significantly increased oxygen consumption and energy expenditure (Fig. 3C,D). Thus we concluded body weight reduction in Chat-Cre mouse model is mainly because of less food intake.

Effect of modulation of parasympathetic nervous system on AgRP/ POMC neurons in arcuate nucleus of the hypothalamus
We also investigated the cellular mechanisms underlying the regulation of food intake in the arcuate nucleus (ARC) of the hypothalamus. Co-staining with c-Fos, early neuron activation marker, and either POMC or AGRP revealed that chemogenetic activation of DMV (hM3Dq) increased POMC neurons in ChAT-Cre mice (Fig. 4A, B). However, inhibition of DMV (hM4Di) significantly decreased the number of not only POMC neurons but also POMC and c-fos doublepositive neurons in ChAT-Cre mouse (Fig. 4A, C). These results show that reduced food intake us due to increased POMC expressing neurons upon activation of parasympathetic nervous system. We next investigated AGRP expression in the ARC of the hypothalamus. Chemogenetic activation of DMV (hM3Dq) decreased AGRP neurons in ChATcre. In contrast, chemogenetic inhibition of DMV (hM4Di) highly increased AGRP neurons in ChAT-Cre mice (Fig. 4D, F). These results suggest that specific chemogenetic modulation of parasympathetic neurons mediates food intake through altering the number of AGRP and POMC expression neurons in ARC of the hypothalamus.

Discussion
The parasympathetic nervous system plays a key role in the control of both food intake and energy expenditure and results in body weight changes [19,20]. Our results suggest that DMV modulates the parasympathetic nervous system and modulation of DMV affects food intake and energy expenditure. Altered food intake and energy expenditure result in body weight changes. The efferent parasympathetic autonomic signal is conveyed via preganglionic cell in DMV brain stem [9]. These mechanisms play a role in the control of energy expenditure [11,22]. The preceding discussion provides two main evidences. Parasympathetic nervous system regulates 1) energy expenditure by innervating with peripheral tissue and 2) interplay with central neurons in ARC of the hypothalamus. Hypothalamic arcuate nucleus (ARC) contains a various population of neurons expressing the orexigenic factor neuropeptide Y (NPY) and AGRP and the anorexigenic factors POMC and cocaine-and amphetamine-regulated transcript (CART) [16]. However, not all neural networks that calibrate energy status require signaling from the hypothalamus, and a solely hypothalamus centric view offers only an incomplete picture of homeostatic energy regulation [23]. Possible mechanisms by which VNS affects energy expenditure have been suggested in several studies. From an anatomical viewpoint, it is interesting to note that the majority of fibers present within the vagus nerve are afferent fibers (74%) and only a minority (26%) are efferent, including that the vagus nerve is both an afferent nerve and efferent nerve as well [27]. Vagal sensory information plays a crucial role in the mechanism of satiation but the underlying circuitry in the caudal brainstem and higher up in the brain is not defined. In this study, specific modulation of the parasympathetic nervous system by chemogenetic methods causes changes in the number of proopiomelanocortin (POMC) and agouti-related protein (AGRP) neurons in ARC of the hypothalamus. Our results support that DMV is a node of parasympathetic nervous system including both afferent and efferent signals. The signal from DMV reaches to ARC of the hypothalamus and activates the number of AGRP and POMC expressing neuron. C-Fos and either POMC or AGRP double co-staining show that the parasympathetic nervous system activation affects both POMC and AGRP expression in ARC of the hypothalamus. Interestingly, while chemogenetic activation of DMV neurons simultaneously is increasing peripheral blood glucose, the serum insulin contents are not increased followed by increased blood glucose levels. These results suggest that modulation of DMV may cause an error in central glucose sensing or another common pathway. Previously studies showed that vagus nerve stimulation affects energy expenditure, demonstrated that cephalic phase of digestion induced gastric acid secretion and motility [28]. These cholinergic pre-ganglion also travels to the pancreas within the bulbar outflow tract and the hepatic and gastric nerves of the vagus. In addition to peripheral organs, arcuate nucleus is particularly reciprocal connected with the dorsal vagal complex and integration of endocrine and behavioral aspects of food intake satiety [2]. The activation of preganglionic parasympathetic neurons in the DMV generates action potentials (AP) and the AP travels through the vagus nerve. Parasympathetic neurons innervate with peripheral organs. When AP arrives on the target organs, acetylcholine is released into a synapse and it binds to the receptor expressed on the target organs [16]. For example, when the VN is stimulated, the terminals of the preganglionic nerves in the intra-pancreatic ganglia, release acetylcholine to the synapse with the acetylcholine, which in turn causes the release of acetylcholine from their terminals within the islet [29,30]. Our results support previous studies that the parasympathetic nervous system regulates blood glucose levels and pancreatic insulin secretion independently. We provided the evidence for these functional connectivity among the parasympathetic nervous system, hypothalamic neurons, and peripheral tissues. Chemognetic modulation of parasympathetic nervous system on feeding behavior and energy metabolism had changed feeding and glucose and energy expenditure in ChAT-Cre mice (Table 1). However, we were not able to define which neural pathway interplay to control parasympathetic nervous system. Therefore, more detailed working mechanism of the neurons and the researches for functional neural connections are still required. The present study suggest that chemogenetic acute activation or inhibition of the parasympathetic nervous system regulates feeding behavior and energy expenditure.