Meal characteristics of melanocortin 4 receptor knockout (MC4RKO) mice

Material Information

Meal characteristics of melanocortin 4 receptor knockout (MC4RKO) mice
VAUGHAN, CHERYL H. ( Author, Primary )
Copyright Date:


Subjects / Keywords:
Dosage ( jstor )
Food ( jstor )
Food intake ( jstor )
Genotypes ( jstor )
Obesity ( jstor )
Physiology ( jstor )
Rats ( jstor )
Receptors ( jstor )
Satiety ( jstor )
Test meals ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Cheryl H. Vaughan. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
Resource Identifier:
658216354 ( OCLC )


This item is only available as the following downloads:

Full Text




Copyright 2006 by Cheryl H. Vaughan


This document is dedicated to my father, Arthur.


iv ACKNOWLEDGMENTS One of the main forces in the accomplis hment of this dissertation has been my mentor, Dr. Neil Rowland. His support and gu idance over the past six years have been steadfast. Working in the lab has give n me experience in troubleshooting and independent thinking. I would like to extend appreciation to Drs. Darragh Devine, Alan Spector and Joanna Peris for the advice they have given regarding my dissertation as well as future endeavors. I would like to thank Dr. Haskell-Luevano for graciously providing all of the mice in these studies and offering he r lab space for me to learn new techniques. Acknowledgment is also due to Marcus Moor e, who was instrumental in teaching and helping with mouse genotyping. Fellow graduate students past and pres ent have supported me along the years, namely Connie Grobe, Staci Kies, Anaya Mi tra, Jonathan Pinkston, and Jaime Tartar. Their well wishes and sympathetic words have been very encouraging. Additionally, Kim Robertson has been integral in helping me to organize and carry out all of my experiments here at UF. Kim’s countless, se lfless contributions ha ve been invaluable. Lastly, I would like to express gratitude to my parents for th eir unwavering confidence in me.


v TABLE OF CONTENTS page ACKNOWLEDGMENTS...................................................................................................4 LIST OF TABLES............................................................................................................vii LIST OF FIGURES.........................................................................................................viii ABSTRACT.......................................................................................................................ix CHAPTER 1 INTRODUCTION........................................................................................................1 Homeostasis..................................................................................................................1 Parameters of a Meal....................................................................................................4 Neurobiological Controls of Feeding...........................................................................6 Central Mechanisms..............................................................................................6 Postingestive Feedback.......................................................................................11 Experimental Methods Used to Study Feeding..........................................................15 Models of Obesity.......................................................................................................18 Mouse Models.....................................................................................................19 The Melanocortin 4-rece ptor Knockout Model...................................................21 Aims of Study.............................................................................................................25 2 MEAL PATTERNS AND FORAGING IN MELANOCORTIN RECEPTOR KNOCKOUT MICE...................................................................................................26 Introduction.................................................................................................................26 Materials and Methods...............................................................................................27 Animals and Housing Environment....................................................................27 Food Procurement Schedule................................................................................28 Data Analysis.......................................................................................................29 Results........................................................................................................................ .30 Discussion...................................................................................................................33 3 FOOD MOTIVATED BEHAVIOR OF MC4RKO MICE UNDER A PROGRESSIVE RATIO SCHEDULE......................................................................37 Introduction.................................................................................................................37


vi Materials and Methods...............................................................................................39 Animals and Housing Environment....................................................................39 Progressive ratio sessions.............................................................................41 Data Analysis.......................................................................................................42 Results........................................................................................................................ .42 Discussion...................................................................................................................46 4 SHORT TERM FOOD INTAKE AND ME AL SATIETY IN MC4RKO MICE......51 Introduction.................................................................................................................51 Materials and Methods...............................................................................................54 Animals................................................................................................................54 Food Deprivation Study......................................................................................55 Preload Study.......................................................................................................55 Data Analysis.......................................................................................................56 Results........................................................................................................................ .57 Food Deprivation Study......................................................................................57 Preload Study.......................................................................................................58 Discussion...................................................................................................................61 5 SHORT TERM FOOD INTAKE IN MC4RKO MICE AFTER CCK AND BOMBESIN ADMINISTRATION............................................................................66 Introduction.................................................................................................................66 Materials.....................................................................................................................71 Animals................................................................................................................71 Drugs...................................................................................................................72 Procedure....................................................................................................................72 Adaptation Trials.................................................................................................72 Satiety Peptide Responses...................................................................................72 Data Analysis.......................................................................................................73 Results........................................................................................................................ .73 Discussion...................................................................................................................76 6 GENERAL DISCUSSION.........................................................................................81 Summary of Results....................................................................................................81 Effects of a Fixed vs. Pr ogressive Ratio Schedule.....................................................85 Post Ingestive Feedback.............................................................................................89 Known and Unknown Mechanisms of Feeding in MC4RKO Mice...........................92 Conclusion..................................................................................................................94 LIST OF REFERENCES...................................................................................................93 BIOGRAPHICAL SKETCH...........................................................................................112


vii LIST OF TABLES Table page 3-1. Mean daily food intake (g) during phases of experiment..........................................42 4-1. Experimental design for preload study......................................................................56 4-2. Mean intake (ml) after deprivation............................................................................58 4-3. Evaluation of randomized block design on conditioning..........................................59 5-1. Mean Ensure intake (ml) during adaptation trials..................................................73


viii LIST OF FIGURES Figure page 2-1. Group mean SE pellets per meal..........................................................................31 2-2. Group mean SE meals per day..............................................................................31 2-3. Group mean SE of pellets per day.........................................................................32 2-4. Mean body weights at the start, mi ddle and end of each phase in the operant chamber....................................................................................................................32 3-1. Breakpoints for WT, HET and MC4RKO mice.......................................................43 3-2. Mean SE number of initiated small feeding bouts................................................44 3-3. Twelve day average of tota l responses made per daily session................................45 3-4. Cumulative responses of i ndividual mice across a daily session.............................45 3-5. Weights of mice during experimental phases...........................................................46 4-1. Absolute Ensure intake at different time points....................................................57 4-2. Ensure intake separated by preload condition and genotype.................................60 4-3. Mean weights of mice...............................................................................................61 5-1. Ensure intake during 30 min te st meal after CCK administration.........................74 5-2. Ensure intake during 30 min test meal after Bombesin administration.................75 5-3. Weights of mice throughout testing..........................................................................75 6-1. A comparison of PR and FR cumulative responses.................................................88


ix Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy MEAL CHARACTERISTICS OF MELANOCORTIN RECEPTOR 4 KNOCKOUT (MC4RKO) MICE By Cheryl H. Vaughan May 2006 Chair: Neil E. Rowland Major Department: Psychology Animal models utilizing genetic manipula tions are beneficial to the study of food intake. We used the melanocortin 4 type receptor knockout (MC4RKO) murine model because of its relevance to humans. Our mice were derived from a stock originating at Millenium Pharmaceuticals. In humans iden tified with a monogenic cause of obesity, 6 % of cases are attributed to heterozyg ous mutations of the MC4R. Firstly, we investigated meal strategies of mice as a f unction of cost for food. In our protocol, mice lived in a two lever operant chamber where completing responses on two levers, one for “foraging” (procurement cost or PFR) and one for eating (consumatory cost or CFR), produced delivery of a 20-mg food pellet. Mice showed welldefined changes in meal taking strategy as a function of imposed procur ement cost, as reported previously in other rodents, and these adjustments were not dependent on MC4 receptors. A progressive ratio (PR) schedule was employed next on MC4RKO mice as an indicator of motivation for reward because PR schedules require an increasing number of


x responses to gain access to a reinforcer. Our results s how that MC4RKO mice exhibit high levels of motivation under a PR schedul e. MC4RKO mice had higher breakpoints, low meal frequencies, large meals and a greate r number of responses emitted than either heterozygous or wild type mice. To further describe meal characteristics, we investigated processes that sustain intermeal intervals (satiety ) and processes that terminate ongoing episodes of eating (satiation). Satiety was tested by administ ering an oral preloa d and then monitoring intake of a test meal. A preload is a small amount of food given to a subject that gives information about the individual’s appraisa l of hunger and satiety. KO mice also were able to sense differences in volume in the gut and therefore have normal appraisal of hunger and satiety. Satiation was tested by examining the effect of exogenous administration of cholecystokinin (CCK) and bombesin (BBS) on the size of a liquid diet meal. Our results show that KO mice are fully responsive to CCK and BBS. These experiments add significantly to our understanding of th e role of MC4R in normal feeding behavior.


1 CHAPTER 1 INTRODUCTION The obesity epidemic has been a topic of in terest recently due to the rise in obesityrelated illnesses. An estimated 300,000 deaths per year in the U.S. are associated with being overweight or obese (Mokdad et al., 2003). Funds necessary to address this rise in healthcare geared towards remedying these illnesses have been a national burden. In 2000, the economic cost of obesity in the Unite d States was more than $115 billion. This is more than the cost of tobacco-related illnesses (Center fo r Disease Control and Prevention [CDCP], 2000). A more effective a pproach to trying to decrease cases of obesity is investing in avenues aimed at discovering preventative measures to the development of obesity. The sharp increase in the incidence of humans identified as overweight in the past 10-20 years is directly attributable to envir onmental and lifestyle ch anges, however there has been evidence of predisposing genes. Physiologic and genomic research using animal models have greatly helped in treati ng obesity related illnesses at their onset. The understanding of factors that contribute to an obese phenotype will be an important contribution to benefit many i ndividuals in the long run. Homeostasis Homeostasis is defined as maintaining internal equilibrium by adjusting physiological processes. Maintenance of the body can be considered to be evolutionarily adaptive. Organisms evolve to propagate th e species, thus developi ng internal regulation


2 strategies. Canon (1929) stated that external cha nges incite reactions in the internal system. Mayer (1955) showed that wh en rabbits were challenged by differing temperatures, they adjusted their body weight and food intake accordingly, suggesting there was a direct relationship between energy input and output. The fact that their body weights hovered around a central point indicat ed that organisms do have an optimal equilibrium that is reflected by normal physiological function. What has been shown in subsequent y ears is that the body has two mechanisms for maintenance: short term and long term en ergy providers. Mayer (1955) proposed five criteria for both types of regulation. Regulation has to be integrative of fats, carbohydrates and proteins to re flect the real time intake exhibited by organisms. Regulatory processes should be based on known ne urological structures in order to have bases for practical metabolic changes. Regulation needs to have an accompanying sensory acuity to pinpoint the cause of the s hort term change; in other words regulating energy expenditure in proportion to the environm ental constraint or metabolic disorder. Lastly, regulation has to address the physical symptoms of hunger. Today, reviews are still attempting to establish rules by which maintenance should be defined (Smith, 1996; Schwartz, 2000; Moran, 2004). Canon (1929) acknowledged that the term “hom eostasis” is best applied to closed systems where known factors are balanced. Th e body is by no means always in balance and is not an entirely closed system; the envi ronment has an influence. External stimuli play a role in defining homeostasis for indi vidual organisms. The commonality of all organisms is the effort to impose reliable regulation. Kennedy (1967) defines regulation in terms of a system’s set point and the variable co ntrol to maintain this system. Bolles


3 (1980) describes regulation as a limit imposed by activation of a new factor after some critical value has been reached. He argue d that “real regulation” was only found in animals exhibiting negative feedback on feedi ng via energy stores a nd satiation (Bolles, 1980). The set point’s function is to relay positive or negative “error” signals (Bolles, 1980). There are four plausible states indivi duals can exist in. Two states lead to overeating: they involve in creased appetite coupled with a diminished perception of satiety. These characteristics are accompanied by large meal size and/or small intermeal intervals. The other two states involve a state of reduced ap petite in conjunction with an enhanced perception of satiety. These stat es are accompanied by small meal size and long intermeal intervals (Brobeck, 1955). If th ese four states were placed on a spectrum, the set point would have to lie in the middle. However as we ight is lost or gained the states on the spectrum become relative and the set point can shift. The idea of set point has incited various cr iticisms due to the rigid nature of the theory (e.g., Mrosovsky, 1990) but it has also provided a founda tion to base theories on regarding peripheral energy regul ation. Most of the current theories in the neurobiology of feeding stem from the lipostatic and gluc ostatic theories. Both hypotheses of Kennedy (1953) and Mayer (1955) attributed desce nding control of feedi ng to two peripheral factors that reported to the brain about the body’s metabo lic state. Kennedy (1953) focused on the hypothalamus as the locus by wh ich energy input and output regulated fat depots. He proposed that circulating fat metabolites exerted action on certain hypothalamic centers and today the “metabolite ” is known as leptin (Campfield et al., 1995). Mayer (1955) postulated that glucoreceptors must exist in the periphery to report


4 blood glucose levels to relevant hypothalami c areas. Glucoreceptor channels are found in the pancreas and do contribute to relaying nut ritional value to the LH (reviewed in Berthoud, 2002). The more global process of homeostasis cannot be fully appreciated without considering the local event of the meal. Parameters of a Meal Le Magnen (1981) pointed out that most of an organism’s physiological mechanisms are continuous whereas feeding is episodic. Discovering ways to accurately measure food intake in the laboratory was necessary in order to develop template characteristics for normal feeding behavior . The parameters of food intake can be categorized into two main categories: within meal behavior and betw een meal behavior. These two parameters give us the useful measures of meal size and meal frequency, respectively (Blundell et al., 1989). Documenting normal para meters of feeding give a reference point of comparison for genetically or anatomically altered models. Oropharyngeal stimulation happens throughout a meal and serves to perpetuate or inhibit further meal initiating behavior (Dav is, 1999). Negative postingestive feedback via sensory signals originating in the periphery is necessary for meal termination (Smith, 1996). Sensory input that take s place during all phases of ingestion contributes to the controls of feeding. Smith (1996) has further delineated controls of feeding into direct and indirect influences. Direct control of meal size is triggere d by nutrient contact with preabsorptive receptors (Smith, 1996). For example, CCK wo uld be considered a direct inhibitory control of meal size; it is re leased as a result of preabso rptive contact with nutrients (Moran, 2004). There are three things that can occur to reduce meal size. First, the lumen of the GI tract becomes distended and begins to contract. Second, ch emical properties of


5 the ingested material are released, and last ly peptides/neurotransmitters from the gut are released in response to the presence of ingested material (Schwartz, 2000). Direct excitatory controls maintain food intake until negative feedback or an indirect factor becomes sufficient. Indirect controls of meal size fall into a few categories and work as intervening variables in relation to the direct controls; included in these categories are the influences of timing, conditioned preferences and ecologi cal availability. Indirect controls ultimately work to modify the elements invol ved in direct controls. Some indirect controls include effort, metabolic changes in the individual and thermal characteristics of the environment (Smith, 1996). It has been shown that effort required to gain access to meals can indirectly affect parameters of m eals (Collier et al., 1986). Palatability plays an initial role in food intake; if food is not perceived as palatable, ingestion will end (Rolls & Hetherington, 1989). Since informati on from taste receptors and viscera travel to the amygdala and hippocampus, strong taste associations are infl uential in modulating food intake (reviewed in Berthoud, 2002). There is evidence that the pairing of oral cues and gut feedback can condition changes in meal size (Davis & Campbell, 1973; Mook et al., 1983). These studies of i ndirect controls give a more faceted view of feeding behavior. Satiety and satiation in the public arena ar e sometimes used interchangeably. In food intake research, satiety and satiation have two distinct implicati ons. Satiation refers to the collective processes (cognitive, physiologi cal, etc.) that end a me al. Satiety refers to the collective processes that inhibit further eating once a feeding episode has concluded (Blundell et al., 1989).


6 One of the overall objectives of the upco ming chapters is to describe how these parameters change when hyperphagia occurs due to changes in genes or the environment. Many rodent models of hyperphagia and result ant obesity are associ ated with increased meal size combined with shortened durations of satiety following those meals. It is necessary to understand how thes e mechanisms interact neurally in order to make sense of the study of feeding behavior. Neurobiological Controls of Feeding A highly conserved trait in mo st species is the ability to optimize feeding behavior in unpredictable environments. Early studies localized these feeding mechanisms to the hypothalamus and named it the feeding cente r (e.g., Stellar, 1994). These early explanations were accompanied by constructs such as hunger and motivation, which were thought to be involved in one area of the br ain, which at the time was poorly understood. Many other areas that interact with the hypothalamus have been identified and, in some cases, specific functions linked prominently to those regions. There is crosstalk to and from the hypothalamus that outlines positive or negative feedback for feeding and there are both central and peripheral infl uences of ingestive behavior. Central Mechanisms Nuclei in the hypothalamus have been locali zed as key players in feeding behavior. Hetherington and Ranson (1940) first reported th at bilateral electrol ytic lesions of the ventromedial area of the hypothalamus (VMH) resulted in hyperphagia in rats. The VMH was thereafter thought to be intrinsically involved in satiety and if compromised, rats could not detect a satiety signal thus ceas ing to eat. Bilateral lesions of the lateral hypothalamic (LH) nuclei resulted in aphagia. The LH was thereafte r thought to be a


7 hunger sensing center. Due to th e lesions, animals could not be cognizant of th e fact that they were in metabolic de ficit (Anand & Brobeck, 1951). Subsequent work has identified other areas including the arcuat e, paraventricular and dorsomedial nuclei of the hypothalamu s. The arcuate nucleus (ARC) is a circumventricular organ which allows for greater permeability of medium sized blood borne molecules into the brain parenchyma. The ARC is responsible for the entry of peripheral signals, like leptin and insulin, to the brain (reviewed in Williams et al., 2000). The paraventricular nucleus (PVN) is an ar ea of convergence for pathways related to energy balance. The PVN is an important site for the release of orexigenic (increases food intake) peptides, namely neuropeptide Y (NPY). Additional evidence shows that the PVN is essential for incorporation of si gnals, exogenous administration of orexigenic peptides have resulted in increased feedi ng. The dorsomedial nucleus (DMN) is also involved in NPY release and play s an integrative role in proc essing central and peripheral signals involved in food intake (r eviewed in Kalra et al., 1999). Neurons involved in feeding are presen t in an important region in the caudal brainstem, the nucleus of the solitary tract (N TS). The NTS is intrin sic in the relaying of messages between the forebrain and the vi scera. POMC neurons travel through the periaqueductal gray and the tegmentum to en ter the rostral NTS (E llacott & Cone, 2004). The NTS receives messages from cranial nerv e X, more commonly known as the vagus nerve, which innervates the organs of th e digestive tract making it intrinsic for post ingestive feedback. Vagal afferent fibers arise from the uppe r gastrointestinal tr act, go through the nodose ganglion and terminate in the NTS. No dose ganglia are inferior ganglia of the


8 vagus nerve and carry information centrally to the medulla and peripherally to other branches of the vagus (Fox et al., 2001). Va gal efferent fibers originate in the dorsal motor nucleus of the vagus nerve (DMX) exit the medulla and terminate in epithelia of various thoracic and ab dominal organs. Informational crosstalk between gut and brai n is also referred to as the gut-brain axis. The enteric nervous system (ENS) consis ts of the myenteric and submucosal ganglia and plexuses which are thought to mediate betw een mucosal, muscular and neural events (Schwartz, 2000; Johnson, 2003). Motility of the ENS is controlled by the two branches of the autonomic nervous system. Sympathetic neurons project to the gut from the spinal cord in three major categories: celiac ganglia , superior mesenteric ganglia and interior mesenteric ganglia. Parasympathetic neur ons project from the medulla oblongata via branches of the vagus nerve a nd the sacral region of the sp inal cord to the gut. Food intake is regulated at multiple levels of the nervous system. A system that has had some recent focus is the melanocortin system, partly due the availability of genetically e ngineered mice and the discovery of human analogs to this model. The melanocortin system include s AgRP (agouti related protein) neurons, proopiomelanocortin (POMC) neurons and downstream target sites (i.e. neurons containing melanocortin receptors). -MSH is the endogenous agonist of melanocortin type 3 and 4 receptors (MC3/4Rs) (revie wed in Hadley & Haskell-Luevano, 1999). MSH is released from POMC neurons after acti vation by leptin that results in attenuated food intake (Fan et al., 1997; Cowley et al ., 2001). AgRP is the endogenous antagonist of melanocortin type 3 and 4 receptors (MC3/4 Rs). Expression of the protein agouti is


9 related primarily to hair fo llicle pigmentation but when overexpressed it was found that it directly increased feeding (Bultman et al., 1992). Agouti related transcript, now known as Ag RP, was isolated some years later and its RNA was found in the hypothalamus. AgRP is similar in size and structure to the agouti protein (Ollman et al., 1997). AgRP is produced and released from neurons in the arcuate nucleus, which projects to the PVN, DMN, and posterior hypothalamus. Endogenous release of AgRP antagonizes MC3Rs and MC4Rs and results in an increase in feeding (reviewed in Cone, 1999). There is little expression of AgRP in the brainstem in comparison to the robust expression seen with POMC neurons (B agnol et al., 1999). Thus, the melanocortin system is involved in feeding via different peptides originating from the midbrain and hindbrain. POMC is produced in the arcuate nucl eus of the hypothalamus. POMC is a precursor protein that is cleaved to produ ce three classes of peptides: melanocortins, corticotropins and opioids (for review s ee Hadley & Haskell-Luevano, 1999). The posttranslational modification of POMC is regulated by cort icotropin releasing hormone (CRH), also produced in the hypothalamus. Deletion of the POMC gene and resulting effects on food intake occur due to the lack of melanocortins and their receptors and subsequent dysregulation of energy balance (reviewed in Cone, 1999). Excitation by leptin, transduced via leptin receptors, on POMC neurons projects to target sites to modulate feeding (Cowley et al., 2001). POMC terminals have been located in ARC, DMN and LH (Bagnol et al., 1999). Central melanocortin receptors have been the focus of studying the control of energy balance and mainly transduce metabo lic signals. Five melanocortin receptors


10 have been identified. The MC1R is the endogenous receptor for -MSH and has a role in mediating pigmentation; the re ceptor is antagonized by the agou ti protein. The MC2R is the endogenous receptor for ACTH and is found in the adrenal gland (reviewed in Vergoni et al., 2000). The MC5R binds MSH/ACTH and has roles in regulating exocrine gland function and aggression suppre ssion in mice (Morgan et al., 2004). The two MC receptors most pertinent to feeding are the MC3R and the MC4R (reviewed in Butler & Cone , 2003). The MC3R is found in adipocytes, stomach, duodenum, placenta, skeletal muscle, and brain. Specifically in brain, MC3R has been located in the hypothalamus (Gantz et al., 1993 a). Functionally, the MC3R is related to fat metabolism. Females with the MC3R de letion exhibit normal food intake while males show decreased food intake. In males and fe males, fat mass increases at 26 wks old and hyperleptinaemia develops. Despite these findings, MC3RKO mice are not overweight (Chen et al., 2000a). MC4R mRNA is found in rela tively high concentrations in brain regions such as the PVN, LHA, ARC and DMN in the hypothalamus and the NTS (Mountjoy et al., 1994). Activation of the MC4R reduces f ood intake and body weight. Mice lacking functional MC4Rs become obese and are hyperphagic, hype rleptinaemic and hypertensive with onset of thes e characteristics in early adol escence. They also become leptin resistant and show altered metabolic rate (Huszar et al., 1997) further contributing to their obese phenotype. Thus the MC4R app ears to be related to regulation of feeding and energy homeostasis. MC 4R knockout (KO) mice pair fed to wild type mice had lower body fat content, body weight and serum leptin content than ad lib fed MC4RKO


11 mice (Ste. Marie et al., 2000). This result illustrates how regulation of food intake in MC4RKO mice can be altered when external control by the experimenter is imposed. The MC4R is found in the medial nucleus accumbens (nAcc) shell, LH and amygdala. The amygdala to LH projection is critical for opioid mediated feeding. The nAcc shell and central nucleus of the am ygdala control downstream feeding motor pattern generators. There are projections to mo tor circuits from these structures that are involved in visceral and motivated movement (Will et al., 2004). The role of endogenous opioid release has been a well documented occu rrence that is implicated in motivational aspects of the maintenance of feeding, esp ecially when palatabl e foods are available (reviewed in Kelley et al., 2002). Inactiva tion of the MC4R in LH may alter reward circuitry and / or transmission that w ould affect parameters of food intake. Postingestive Feedback Control and regulation of ga strointestinal (GI) function starts with the cephalic phase then is followed by the gastric and inte stinal phases. The cephalic phase begins with the sensory experience of food followed by an increase of parasympathetic outflow to the GI tract. This initial phase of feeding begins the stomach’s preparation for incoming food. Organs in the peritoneal cavity are re cruited during the gastric phase. Once food has been ingested there are mechanoreceptors in the viscera and neuropeptides released that signal to the brain whethe r to stop or continue feedi ng. Distention of the stomach triggers the gastric phase response which is ex ecuted through vagal e fferent signals. The main signaling by neuropeptides to the brain is done via receptor s on the postganglionic afferents of the vagus nerve (CNX). CNX innerv ates most of the organs in the digestive tract. After food passes th rough the esophagus into the stomach, chemical breakdown


12 begins to occur. One of the physiological fa ctors that influence stomach emptying is the vago-vagal reflex (Johnson, 2003). Contents of the stomach that have not been utilized are passed to the small intestine. The small intestine is compartmentalized into 3 sections: the duodenum, ileum and jejunum. The duodenum secretes the peptide cholecystokinin (CCK) which serves as a satiety signal to the brain. Food intake is positively correlate d with endogenous CCK release (Liddle, 1997). CCK communicates a satiety signal to the brain through CCKA and CCKB receptors found on afferent ganglia of the vagus nerve (Weatherford et al., 1992; Moran & Ladenheim, 1998). There is considerable evidence supporting the contribution of CNX to meal termination. If the CNX is cut (vagotomy), an organism will exhibit abnormalities in feed ing because there is not su fficient visceral feedback. Denervating both the hepatic a nd gastric branches of the va gus, both of which innervate the stomach, significantly weakens the ability to suppress meal size after a gastric load during a short term (30-min) intake test ther efore signifying the role of vagal afferent feedback in CNS descending control of meal terminati on (Phillips & Powley, 1998). CNX relays this sensation of satiety to the caudal NTS, which mediates ascending visceral information. The caudal NTS projects to the lateral para brachial nucleus (PBN) and the PBN sends axons to nuclei in th e hypothalamus (lateral, ventromedial and arcuate) involved in inhibiting furthe r food intake (reviewed in Berthoud, 2002). Researchers can artificially create the feeling of fullness by administering CCK (Shilabeer & Davison, 1987; Corwin et al ., 1991). Also, inflating a balloon in the stomach is another means by which subjects will experience the feeling of fullness and thus decrease food intake. In comparison to gastric load and CCK infusion done


13 independently, Schwartz and colleagues (1991) have found that combining an intragastric load and CCK infusions significantly increases th e discharge rate of vagal afferent fibers. These results imply that while both load a nd CCK release will decr ease food intake, the effects of both administered simultaneously are additive. In addition to postingestive feedback, there are long-term and short-term energy stores that play a role in energy balance. Leptin is a long-term indicator of energy stores and contributes to the control of feeding. Though produced peripherally, leptin functions centrally as an anorexigenic signal to the brain. Leptin receptors are located on POMC producing neurons in the hypothalamus. When leptin binds to its recepto r, the POMC gene is activated, leading to increased POMC-related peptides and an ev entual decrease in food intake. POMC neurons can also decrease feeding by leptin's inhibitory effects on NPY neurons in the arcuate nucleus (reviewed in Schwartz et al., 2000). Leptin acts on the long form of the leptin receptor (Ob-Rb) in the arcuate nuc leus (Niswender & Schwartz, 2003). After leptin binds, intracellular cascades result in reduction of inhibitory GABA input to POMC neurons (Cowley et al., 2001). The leptin receptor is a cyt okine receptor that signal s through the janus kinase signal transducers and activ ators of transcription (J aK-STAT) pathway. The JaK phosphorylates the intracellular portion of the long form receptor and allows for binding of STAT molecules. STAT then travels to the nucleus of the ce ll where it is induces synthesis of transcription factors. Suppressors of cytokine signaling (SOCS) can turn off leptin receptor signaling (Niswender & Schwartz, 2003).


14 Short-term stores of energy contain glucos e; glucose is a main source of fuel for the brain and the body. Glucose av ailability is regulated by pa ncreatic release of insulin or glucagon. Insulin is released from the islet cells of Langerhans and glucagon is released from islet cells of Langerhans in the pa ncreas. Regulation of glucose is localized in the liver where insulin’s function is to convert glucose to glycogen for storage and glucagon’s function is to convert glycogen to glucose for use. Insulin is modulated also by food intake and is the main signal to the brain about the body’s metabolic state (Rosenzweig et al., 1999). Endocri ne cells in the intest inal mucosa secrete glucagon-like peptide -1 (GLP -1) and glucose-dependent in sulinotropic peptide (GIP; a.k.a. gastric inhibitory peptide) to enhance in sulin secretion (reviewe d in de Graaf et al., 2004). Insulin levels rise during a meal a nd thus insulin signaling contributes to controlling the subsequent parameters of a meal. Insulin travels via the blood across the blood brain barrier (BBB) th rough circumventricular orga ns, (CVO) areas where the protective BBB is selectively permeable, to insulin receptors in the brain (Niswender & Schwartz, 2003). There is a high level of in sulin receptor mRNA in the olfactory bulb, cerebellum, hippocampus and arcuate nucleus of the hypothalamus. Concordantly, the brain has higher insulin cont ent than circulating plas ma (Havrankova et al., 1978a; Havrankova et al., 1978b; Ma rks et al., 1990). Insulin signaling occurs through a tyrosine kinase receptor (Niswender & Schw artz, 2003). Disruption of this receptor illustrates the effects insulin has on feeding, obe sity and leptin levels. Mice engineered without insulin receptors have increased ad ipose tissue mass and thus show increased plasma leptin levels when on normal chow (Brning et al., 2000).


15 Insulin release is tied to meal presenta tion and / or meal anticipation (Strubbe, 1992). If a rat is food deprived, blood sugar dr ops slowly during the fast but climbs back up to normal upon access to food. Possibly due to gluconeogenesis, food deprived rats will eat more than non deprived rats despit e glucose levels being the same. Exogenous insulin can increase or decrea se food intake in humans wh ile endogenous insulin release generally suppresses appetite but differs in individuals in re lation to body weight (reviewed in de Graaf et al., 2004). Periphe ral feedback signals, whether exogenously or endogenously produced, work to coordinate beha vioral output which is then carried out via autonomic motor systems after cortical, lim bic and somatic system input (reviewed in Berthoud, 2002). Experimental Methods Used to Study Feeding Feeding has often been studied in an open economy or session paradigm. In this configuration, the experimenter initiates and terminates the meals of the subject being studied. Small portions of food are used as reinforcement, so the amount of food and the intertrial interval are solely in the hands of the experimenter . In a closed economy or freefeeding paradigm, the animal can be placed in an operant chamber where initiation and termination of meals are left up to the animal. This allows for repeatable, recordable acts and a solid measure of behavior. In this se t-up, the animal can lever press or perform any other convenient operant at any time and fo r as long as the subj ect is housed in the chamber. The free feeding paradigm is animal initiated and terminated and focuses on the meal. Optimal foraging is maximizing nutrient intake in the most efficient way so as to avoid possible predators. The study of fora ging behavior across species compares the significance of search strate gies given a particular ci rcumstance. The functional


16 relevance to studying foraging is analyzing th e requirements and stimuli that affect the animal’s drive to work for their reinforcement. Collier’s foraging paradigm modifies the free feeding paradigm by extending sessions and measures added dimensions of the meal. His paradigm focuses on how the animal discovers, evaluates and earns f ood (Collier, 1987). Using the two types of economies researchers are able to tease out th e components of a meal’s significance to a test subject. The lever pressing sequence dur ing a session attempts to replicate foraging in the wild by providing the animal with choi ces. Procurement cost, a type of foraging cost, is the price of meal initiation. The cons umatory cost is within meal behavior or the price per unit reinforcement. In an operant chamber, both parameters can be measured by fixed ratios (FRs): procurement (PFR) and consumatory (CFR). Presumably, then, the choices about meal size and frequency made under these conditions reflect cost/benefit computations relevant to feeding opt imally (Collier et al., 1986). The analysis of data coll ected in the chamber group the animal’s responses, which then can be used to create criteria for m eal definitions. The comm on finding of a typical presentation of PFRs and CFRs has been that as the price of the pellets increase, the frequency of meal initiation decreases. Meal frequency decreases as meal size (and associated duration) increases (Collier, 1987). Previous studies of meal patterns in mice ha ve yielded inconsistent results, but they do provide a context for the present work a nd so merit brief review. Petersen and McCarthy (1981), using mice inbred for small (S) and large (L) body size, devised a free feeding meal pattern monitoring system in which an overhead door was pushed open to gain access to a jar that c ontained powdered rodent food. Using a 5-min door closed


17 criterion to define the end of a meal, a nd with the assumption that meal size was proportional to the time the door was open, they found that both L and S mice took about 12 well-defined meals per day, mostly nocturnal . They also found that L mice ate more than S exclusively due to larger meal size; additionally, meal size was increased but frequency unaffected when total intake was increased by lowering ambient temperature. Gannon and colleagues (1992) used a cage in which powdered rodent food was freely and easily accessible in a recess (with an infra-red beam to record entries, but no door) to male SWR/J mice. They found a m ean of 36 meals per day (13 day, 23 night) using 5 min of no beam breaks to define the end of a meal. Thus, both the form of food and the criterion for a meal seem to be identi cal in these two studies, yet there is a 3-fold discrepancy in meal frequency. It would be su rprising if this magnit ude of effect was due to the differences in strain of mouse or br and of chow, and so the discrepancy cannot be understood at this time. Strohmayer and Smith’s (1987) report rema ins the most comprehe nsive analysis of meals in mice, using lean and leptin deficient ( ob/ob) C57BL/6J mice and a liquid diet (EC116). They found that both genotypes took ~75% of meals at ni ght, that meals in ob/ob were larger (by ~50%) than in lean mi ce, and that males took more meals per day (~50) than females (~30). Thus, like Gannon and colleagues (1992), their procedure led to many small meals but, similar to Pe tersen and McCarthy (1981), a hyperphagic phenotype was manifest by primary change in meal size, not frequency. However, their result differs from that of Anliker and Maye r (1956), who found day/night rhythmicity in lean but not ob/ob mice in a CFR protocol for solid 20mg food pellets. These results paint a very inconsistent picture, with the reasons for discrepancie s not at all clear.


18 We have shown that leptin-deficient ( ob/ob ) obese mice took characteristically large meals compared to lean mice when th e foraging (food procurement) cost was low, but like lean mice increased their meal size further and decreased number of meals per day as foraging cost increased (Vaughan & Rowland, 2003). This adaptability of meal parameters in the face of a minor foraging co st will be beneficial to assess how other genes affect meal taking strategies. Models of Obesity In obese models, homeostasis is modified. If there is a set point for organisms, why is it that the organism allows itself to become obese? Obesity is thought to be an eating disorder of its own (Gilbert, 1989). Ov ereating can occur if the cost of acquiring food is low or the diet is highly palatable (Bol les, 1980). Some theories offer the idea that there is some new set point that is designate d and animals keep challenging that set point as commodities become more or less avai lable (Friedman & Stricker, 1976; Bolles, 1980). Additionally, food choice in the envir onment is a contributor to regulation of energy intake. It has been widely shown that varying one’s diet can increase intake and result in obesity (reviewed in Rolls & Hetherington, 1989). Recent innovations in genetics have allowe d us to decipher what genes contribute to a hyperphagic phenotype. It has been esta blished that obesity develops due to the interaction between genes and the surroundi ng environment (Friedman, 2004). Variables such as postnatal nutrition and diet have been shown to influence adult obesity and physiology though the direction is still unclear (Ravelli et al ., 1976; Hofman et al., 1997). These results have been repl icated in a rodent model (Bush & Leathwood, 1975; Bergen et al., 1999). Since the envi ronment cannot be controlled for most of the human population, the main goal of the experiments in this dissertation is to assess the


19 contribution of environment, both external a nd internal, on feeding behavior in a mouse model known for developing metabolic disorders. Mouse models The completion of the mouse genome and wo rk on the rat genome has contributed tremendously to identifying gene s involved in particular beha viors. Leptin production, which is coded for by the lep gene, formerly known as the ob gene, has been deleted from the mouse genome to create a murine model of obesity (Zhang et al., 1994). Mice deficient in leptin production ( lep-/or ob/ob) are hyperphagic, hyperglycemic, hyperinsulinemic and insulin re sistant (Bray & York, 1979). ob/ob mice have decreased glucose oxidation in adipose tissue (Yen et al., 1968). Th ey respond to exogenous leptin by decreasing food intake and body weight. These results are seen after central and peripheral administration of le ptin and the results are also seen after chronic leptin administration. Another murine mutation ( db/db) is associated with a defective long form of the leptin receptor and in this m odel exogenous administration of leptin has no effect on food intake and body weight (Campfie ld et al, 1995). Both of these models confirm the importance of the pr esence and full function of the lep gene and its endogenous receptor. Another mouse model of obesity involv es the targeted mutagenesis of the proopiomelanocortin (POMC) gene. POMC co des for proteins involved in transducing messages about various parameters, includi ng food intake and fat metabolism (Hadley & Haskell-Luevano, 1999). Humans identified with polymorphisms on the POMC gene exhibit morbid obesity, binge eating, and hi gh body fat percentages (Lembertas et al., 1997). Mice engineered to lack POMC show increased fat mass levels, food intake and weight gain rates compared with wild t ype mice, much like their human counterparts


20 (Yaswen et al., 1999; Challis et al., 2004). These physiologi cal similarities allow for a foundation for discovering behavioral factors that contribute to this obese phenotype. Ay/a mice have a dominant mutation of the agouti gene. The agouti protein has pleiotropic effects; it has diffe ring effects that are receptor subtype dependent. In the brain, it antagonizes MC4Rs and in melanocyt es, agouti blocks MC1Rs so there is decreased cAMP production therefore disruptin g melanin production in the skin. cAMP normally aids in and promotes eumelani n production, which codes for black/brown pigmentation. Without cAMP, pheomelanin is made, which codes for red/yellow pigmentation of the hair (Bultman et al., 1992). The constant inhibition of MC4Rs in Ay mice is the reason for the obese and hyperphagic phenotype. Ay mice gain weight faster than thei r wild type counterparts. The genotype has also been linked to macronutri ent intake (fat vs. carbohydrates). When given a choice, Ay mice ate more calories from fat than carbohydrates. This effect was seen across two different age groups: old (pd 46-51) and young (pd 1-8) mice (Koegler et al., 1999). Expression of agouti in rat brain via a vira l vector results in increased food intake when injected into PVN and DMH but not when injected in LH. Effects of agouti were augmented in certain areas when animals that received agouti in the LH were put on a high fat diet, signifying the important c ontribution of food availability in the environment. This evidence suggests body wei ght is affected according to inhibition of specific receptors on neurons in specific nuc lei in the hypothalamus (Kas et al., 2004). We have chosen to use the melanocor tin 4 type receptor knockout (MC4RKO) model because of its relevance to humans. Heterozygous mutations of MC4R in humans


21 account for about 6% of severe cases of humans with monog enic obesity (Farooqi et al., 2003). In one study, Raffin-Sanson and Berthe rat (2001) reported 4% of obese children in their sample population had a heterozygous mutation on MC4R. To date there are 58 different types of mutations that have been documented that affect human MC4R function (MacKenzie, 2005). The Melanocortin 4-receptor Knockout Model Interest in the development of obesity in agouti yellow (Ay) and agouti lethal yellow (Avy) mutants led to the cloning of MC4R . The agouti yellow mouse exhibits mature onset of obesity due to ectopic expre ssion, expression of a gene where it is not normally expressed, of the agouti protein (Bultm an et al., 1992). It was reported that the hyperphagic phenotype seen in these mutants was a result of antagonism of MC4Rs by the agouti protein (Huszar et al ., 1997). Shortly after this di scovery, an initial report of two POMC deficient children, aged 3 and 5 years, was desc ribed. These children did not produce -MSH, the endogenous agonist to the MC 4R, and developed obesity along with other health conditions (Krude et al., 1998). These landmark studies demonstrated the contribution of the melanocortin syst em in homeostatic regulation. Subsequent reports have provided func tional verification that the MC4R is involved in energy regulation. The MC4R ge ne encodes for a 332 amino acid G protein coupled receptor (Gantz et al., 1993b). MC4R mRNA is present in many nuclei (i.e. PVN, LH, ARC and DMN) of the hypothalamus and in the NTS (Mountjoy et al., 1994). Particularly, the LH has been related to a propensity for diet induced obesity. Agouti expression in the LH resulted in increase of caloric intake and weight gain on a high fat diet (Kas et al., 2004).


22 MC4R deficiency has been linked specifica lly to fat consumption. When MTII (a MC3/4R agonist) was administered to WT mice, they ate less of a high fat diet compared to a high carbohydrate or protein diet; fat in take was reduced by 70% (Samama et al., 2003). Additionally, MC4RKO mice are known to respond to certain peripheral and central metabolic signals but not much is known about their sa tiety signaling after individual meals (Mar sh et al., 1999). MC4RKOs have intact leptin signaling in the brain so that downstream activation of NPY (orexigen) and cocaine-amphetamine related transcript (CART; anorexigen) can occur in PVN and LH, respectively. So MC 4RKOs theoretically can neurochemically modulate feeding but do so differently than WT, suggesting the MC4R is necessary for modulation of feeding. Albarado and colle agues (2004) have reported that MC4RKO mice have increased respiratory exchange ratio (RER), an indicator of metabolic efficiency, after being maintained on a hi gh fat diet for three days. The MC4RKO RER was higher than wild type mice and ob/ob mice. MC4RKO showed development of hepatic steatosis, excessive fat in the liver; all of this evidence adds up to altered metabolism which is ultimately responsible for the obese phenotype. However, the development was dependent on background strain of the knockout (Albarado et al., 2004). Most of the studies cited above have compared MC4RKOs to “wildtype” mice on a B6/129 background. We recognize that this is a not a strict “wildtype” strain and that not all strains are without phenotypi c traits specific for a par ticular strain. However for generality to previous studies, mention of wildtype mice in this dissertation refers to the background B6/129 strain. In a comparison of food intake for various mice, the same B6


23 (C57BL/6J) background strain for MC4RKO mice and a 129 variant (129P3/J) strain reportedly ingested ~4g of food per day (Bachmanov et al., 2002). There are limited studies describing a natural f ood intake pattern in different mouse strains without the addition of other independent variables, so we and others (e.g ., Albarado et al., 2004) acknowledge that a knockout’s b ackground strain is an importa nt factor in interpreting behavior. To date, mainly physiological parame ters of MC4RKO feeding have been documented (i.e. Huszar et al., 1997; Ste Ma rie et al 2000). Food choice studies have been one approach that has tried to address the role of environmental influence on MC4RKO feeding behavior (e.g., Ste Marie et al 2000; Samama et al., 2003). When MC4RKO mice were given a running wheel they did not wheel turn to expend energy after being on a diet with a moderate fat cont ent (Butler et al., 2001). In contrast, another group reported that presence of a running wh eel kept weight low in MC4RKO mice and changing to conventional single housing wit hout a running wheel resulted in rapid increase in body weight gain and food intake in KO mice (Irani et al., 2005). Collectively, these reports illustrate the ro le environment can play in weight gain. However, studies of MC4RKOs show limited i nvestigation of environmental influences that contribute to weight gain. The majority of work has been done on MC4RKO mice and not MC4R heterozygous mice, so in most of the experiments in this dissertation we have included HETs. Animals that are partial carriers of the MC4R gene become obese and have metabolic deficits, however to a lesser de gree. MC4R heterozygous mice (HETs), of both sexes, have been reported to weigh up to ~55 g in comparison to MC4RKOs that


24 weigh 60 to 70 g, both at 7 months of age, and both genotypes are markedly heavier that WT controls (mean: ~30 g). HETs ar e hyperglycemic and hyperinsulinemic in comparison to wild type mice. Developmen tally, HETs are intermediate in phenotype between WT and KOs beginning at 2 months ol d (56 days). Phenot ypic differences start to emerge around 5 weeks old (35 days); the degree of MC4R dominance increases with age. Fat deposition in HETs, similarly to KOs, is thought to occur due to hyperphagia not hypometabolism (Weide et al., 2003). There is great potential for environmental influences to be explored in human studies now that polymorphisms of the MC4R gene are being identified and published. A recent study (Branson et al., 2003) in humans has looked at en vironmental influences and the MC4R. The study focused on the incide nce of binge eating disorder in obese individuals. Obese indivi duals were genotyped, given a self-report measure and evaluated by a trained research er for binge eating tendencies. A portion of the subjects were found to have null variants of the MC4R . The individuals with the null variants were highly correlated with having a binge eating disorder, suggesting environment may exacerbate tendencies to overeat. However, with limited in-depth human studi es it is difficult to draw conclusions regarding environmental roles in the human model of MC4R re lated obesity until more is known. Using the animal model allows us to ask why, how and when these animals overeat. By utilizing HETs in most of th e following experiments, we can add to the literature by describing a phenot ype that is most commonly found in humans. The theme of the proposed experiments is to investigate the environmental constraints and the satiety feedback mechanisms that govern hyperphagia in MC4RKO mice.


25 Aims of Study There have been few investigations into th e parameters of the meal as influenced by the melanocortin system. Due to the lack of detailed descriptions of how the environment affects individual meals across th e day and across weeks, this dissertation includes two main sets of experiments. The first set of experiments explores the MC4R involvement in association of and motiva tion for reward by using fixed ratio and progressive ratio schedules, respectively. The second set of experiments examines the effect of satiety feedback on single meals.


26 CHAPTER 2 MEAL PATTERNS AND FORAGING IN MELANOCORTIN RECEPTOR KNOCKOUT MICE Introduction Proopiomelanocortin (POMC) gene-derived peptides in the hypothalamus are believed to be endogenous inhibitors of feeding (Hadley & Haskell-Luevano, 1999). POMC-derived peptides act at five known melanocortin recep tor subtypes (Gantz et al., 1993; Mountjoy et al., 1994; Gantz & Fong, 2003) . Of these, subtypes 3 and 4 (MC3R, MC4R) have been linked to feeding and fa t metabolism in both humans and mice (Huszar et al., 1997; Lembertas et al., 1997; Chen et al., 2004). Central administration of the MC3/4R agonist, melanotan (MTI I) or the endogenous agonist -MSH reduce food intake in animals (Fan et al., 1997; Chen et al., 2004). Conversely, either central administration or genetic overexpression of the endogenous MC4R antagonist, agouti related peptide (AGRP) produces hyperphagia (Ollman et al., 1997). Studies with targeted deletion or knockout (KO) of the MC4R gene have further refined our understanding of these systems. MC4RKO in mice results in overeating and obesity (Huszar et al., 1997; Chen et al., 2000b; Weide et al ., 2003). In humans, several examples of MC4R haploinsuffi ciency are associated with se vere obesity (Farooqi et al., 2003). Large spontaneous meals and rela tively short inter-meal intervals often characterize rodent models of genetic obe sity, suggesting impaired mechanisms of satiation and/or satiety (Castonguay et al., 1982; Vaughan & Rowland, 2003).


27 Previously, using a simulated foraging protocol adapted from rat studies (Collier et al., 1986), we showed that leptin-deficient ( ob/ob ) obese mice took characteristically large meals compared to lean mice when th e foraging (food procurement) cost was low, but like lean mice increased their meal size further and decreased number of meals per day as foraging cost increased (Vaughan & Rowland, 2003). The purpose of the present experiment is to describe meal patter ns of MC4RKO mice under low and high cost foraging conditions. Meal patt erns have not been reporte d in MC4RKO mice. Materials and Methods Animals and Housing Environment Five MC4RKO and five wild type (WT; mixed 129/B6 background) adult male mice were obtained from colonies maintained by Dr. Haskell-Luevano at the University of Florida. Males were used primarily so that estrous cycles would not impart higher variance to the intake and pattern data. MC4RKO mice homozygous for the deletion originated from stock generously provided by Millenium Pharmaceuticals (Huszar et al., 1997). Mice used were offspring of mating two MC4R+/derived originally from 129/B6 heterozygotes. All mice were bred, born a nd housed in a colony in the UF Health Sciences Center. They were genotyped using PCR analysis of DNA from a tail snip. At ~21 weeks of age mice were moved to th e Psychology department where they were housed in a vivarium with lights on 0600-1800 hr and ambient temperature 23+ 2oC. Initially, they were housed individually in standard shoebox cages with water and food (Purina 5001 Chow) available ad libitum. Two to three days before the start of th e experimental phase the food in the home cage was changed to a jar of 20-mg nutritionally complete food pellets (Noyes Precision


28 pellets, Research Diets Inc., New Brunswi ck NJ; 25% protein, 64% carbohydrate, 11% fat) that they were fed for the remainder of the study. After this adaptation to the 20-mg pellets, which occurred readily in all mice, they were hous ed individually in operant chambers measuring 13x13x12 cm with a steel rod floor (Med Associates, St. Albans, VT). Two levers protruded through one wall of the chamber, arranged symmetrically on either side of a recessed food trough. Water was available from a spout in the middle of the wall opposite the levers. The chambers were contained inside ventilated, sound attenuating cubicles, each w ith a 15-watt light providing the same 12:12 cycle as the vivarium. Mice lived in the chambers during the e xperimental phases of the study with the exception of 30 min in the middle of each day when they were removed to a holding cage without food while the chambers were cleaned and serviced. Body weights were recorded at the beginning and end of each block of fi xed ratio (FR) testing (see below). Four identical chambers were available, so mi ce were run in squads with at least one representative from each genotype in each squa d. Squads were run at one FR, and then had a 3-4 week period off from the study when they had free food in their home cages while the other squads were run at that FR. Then, the cycle restarted using a new FR. One mouse from each genotype completed only part of the study and were removed either because they failed to progress to th e next FR (MC4KO) or started to self-injure (WT). However, since their behavior in th e early phases was unex ceptional, their data were included in the analysis until their removal. Food Procurement Schedule The protocol was a two-lever tandem schedul e modified for application to foraging. One lever was designated as the procuremen t lever on which a procurement fixed ratio


29 (PFR) was to be performed, and the other lever was the consumption lever on which a consumption fixed ratio (CFR) was to be perf ormed. Small cue lights above each lever indicated which lever was active at any give n time. Initially, the light was on above the procurement lever. Completion of the imposed PFR caused that light to be extinguished and the light above the consumption lever to be illuminated signifying that presses here would deliver food. Completion of a CFR cause d the delivery of a 20-mg pellet into a small trough from a low noise dispenser. Mice could perf orm as many CFRs as they chose to constitute a meal but, if 10 mi nutes elapsed without any presses on the consumption lever, then the meal was declared finished and the cue lights reverted to the initial configuration. The PFR and CFR training schedules are de scribed in detail in our previous study (Vaughan & Rowland, 2003). Mice we re pressing reliably and for stable amounts of food within a few days. The CFR was set at a value of 5 or 10 because at lower CFRs we have found excessive waste of food. The mi ce were progressed through the following sequence of PFR-CFR: 15-5, 60-10, 120-10, 240-10 and 480-10. Each schedule was imposed for 7-10 days. Cue lights and pelle t delivery were controlled by Med-PC software (Med Associates, St. Albans, VT) th at also recorded the number and timing of presses on each lever. Data Analysis Frequency histograms of inter-p ellet intervals were used to determine a breakpoint in performance. The breakpoint for the lowest FR in WT mice occurred after an interval of two minutes and many intervals of 60 minutes or more when no leve r pressing occurred. Although our protocol terminated access afte r 10 minutes of no responding, we defined meals by intervals of 60 minutes without pressing. Af ter meal sizes and frequencies


30 were determined, values were collapsed a nd calculated for each animal. The individual mean total intake per day in each FR schedul e was calculated using 7 consecutive days of stable performance. Food spillage was counted as the number of pellets under the floor, and daily intake was corrected by that amount. Data used for analyses were the group means derived from individual total intakes. Pa rameters of food intake were analyzed using univariate ANOVAs and post hoc Newman Keuls pairwi se comparisons. Body weights were analyzed by t tests. Significance level was p<0.05. Results At the lower FRs, mean meal size was rela tively constant (~35 pellets or ~0.7g) across PFRs and genotype but at the highest PFR, meal size increased (Figure 2-1). Thus, ANOVA showed a main ef fect of PFR on meal size [ F (5, 66) = 4.9, p <.001], but no significant main or interactive effects of genotype. Post hoc analyses showed this effect was due exclusively to an increase at PFR 480. Mean meal fr equency (Figure 2-2) varied significantly as a function of PFR [ F (5, 66) =8.2, p <.001] but with no significant main or interactive effects of genotype. The function was not monotonic: meal frequency was highest (6-7/day) at PFR 15 a nd lowest at PFR 480 (2-3/day). Daily food intake (Figure 2-3) varied significantly as a function of PFR [ F (5, 48) =6.1, p <.001] but with no significant main or interactive effects of genotype. Daily intake declined slightly from 180 pellets (3.6 grams) at the lowest PFRs to 160 pellets at intermediate PFRs, but for both genotypes food intake was significantly lower at PFR 480 (~120 pellets) than at any other PFR (Ps<.05). Body weights of WT mice were stable ex cept for some modest weight cycling during the PFR 240 and 480 schedules (F igure 2-4). MC4RKO mice weighed


31 Fig 2-1. Group mean SE pellets per meal. Performan ce was not significantly different for the two genotypes. Meal size was increased significantly at PFR480 above all other ratios. Fig 2-2. Group mean SE meals per day. Performance was similar for both genotypes. At PFR 15 both groups had the highest meal frequency (p<.05) and at PFR 480 both groups had the lowe st meal frequency (p<.05). 5-5 15-5 60-10 120-10 240-10 480-10 PFR-CFR Mean pellets per meal 0 20 40 60 80 100 Wild type, n=4-5 MC4RKO, n=4-5 5-5 15-5 60-10 120-10 240-10 480-10 PFR-CFR Mean meals per day 0 1 2 3 4 5 6 7 8 9 10 Wild type, n=4-5 MC4RKO, n=4-5


32 Fig 2-3. Group mean SE of pellets per day. There we re no differences in total intake across genotype. Intake was lo west at the highest PFR, 480-10. Fig 2-4. Mean body weights at the start, mi ddle and end of each phase in the operant chamber. MC4RKO mice weighed signi ficantly more than wild type mice throughout the study (p< .001). Weights we re generally stable for the WT mice and significant weight loss occurred in the MC4RKO during each phase. PFR-CFR Mean total pellets consumed per day 50 100 150 200 250 Wild type (n=4-5) MC4RKO (n=4-5) 5-5 15-5 60-10 120-10 240-10 480-10 PFR CFR Grams 10 20 30 40 50 60 Wild type (n=4-5) MC4RKO (n=4-5) 5-5 15-5 60-10 120-10 240-10 480-10 *


33 significantly more than the WT mice at the beginning of each experimental phase (p<.05) but KO mice lost considerable weight (up to 20%) during each operant phase then regained that weight during the intervening weeks in the home cage. There was no tendency for the MC4RKO mice to eat more w ithin a phase as their weight declined. Discussion In our previous study (Vaughan & Rowland, 2003), using a similar two lever operant procedure, we repor ted that female C57BL/6 lep-/mice ate larger meals (~40 pellets or 0.8 g) at th e lowest PFR (5) than lep+/+ or +/controls (~25 pellets or 0.5 g), but slightly fewer meals (6 vs. 8/day). Fu rther, as PFR was increased to 480, the lep-/mice approximately doubled meal size and decreased frequency (to ~3/day) while the controls almost quadrupled mean meal size and decrea sed frequency to ~3 meals/day. The lep-/mice also ate substantially more than l ean mice throughout the study. Thus, like rats (Collier et al., 1986), mice showed a shift to larger and fewer meals as PFR increased but the larger meals of lep-/mice remained evident at the lower PFRs. The present results, obtained under almost identical experimental conditions, are somewhat different. First, at the lowest PFR, the male WT mice (129/B6 background) in this study took larger but fewer meals (~40 pe llets, 5 meals) than did the female B6 controls ( lep+/or +/+) in the previous study (~25 pellets, 8 meals), wh ile the total intake was comparable. Second, while the lep-/mice in the previous study remained hyperphagic and maintained their body wei ghts during the operan t phases, the MC4RKO mice were not hyperphagic and did not maintain their elevated weight, even at the lowest PFR. No significant differences were obs erved between WT and MC4RKO in total intake, meal size or meal frequency. We thi nk it is unlikely that these disparities can be ascribed to differences in sex or the older starting age (2 vs. 5 mo ) in the present study,


34 although direct comparison of control strains se ems to be warranted on the basis of these results and of the disparate meal size literat ure in mice that we cited before (Vaughan & Rowland, 2003). In WT mice, as expected from our pr evious result (Vaughan & Rowland, 2003), as the PFR was increased the meal frequency d ecreased from ~5 meal s/day at low PFRs to ~2 meals/day at high PFRs. Meal size doubled across the range of PFRs, but this was insufficient to maintain a normal level of inta ke at the highest PFR, and some weight loss occurred during this phase. The present WT mice took fewer but larger meals at low PFR than did the controls in our previous study, and this may account for their inability to fully compensate at the highest PFR becau se further increases in meal size may be imposed by stomach or other physiological capac ity. It is important to note that the genetic background on which a null mutation is generated may make a difference in feeding or metabolic studie s (Albarado et al., 2004). MC4RKO mice have been reported to be hyperphagic beginning at about 6 weeks of age (Huszar et al., 1997; Chen et al., 2000b; Weide et al., 2003), so we expected to find a large meal size consistent with ot her rodent models of obesity including lep-/mice. This prediction was not borne out: all of the meal paramete rs in the MC4RKO mice were similar to WT, even though the MC4RKO mice us ed are older than the ages reported for the onset of hyperphagia (Huszar et al., 1997) . They were not hyperphagic and did not maintain their body weight even at the lowe st PFR. Hyperphagia presumably occurred when these mice were returned to their home cages and regained all of the lost weight between PFR schedules, but intake was not me asured during these times. Thus, the meal parameters during hyperphagia in MC4RKO remain to be documented.


35 A key question then becomes why th e MC4RKO mice tolerated weight loss uncompensated by increasing food intake with a phase, and even at the lowest PFR. The fact that these mice are hyper phagic in their home cage suggests that they are exquisitely sensitive to novel or other aspect s of the test environment; the fact that they are willing to work for normal amounts of food, and work more as the PFR increases, suggests that it is not simply an unwillingness to emit operant responses. In free access conditions, MC4RKO mice have been shown to resume hype rphagia after a period of food restriction (Ste Marie et al., 2000). Furt her, administration of anorect ic and orexigenic peptides decrease and increase feeding, respectiv ely, in MC4RKO mice (Marsh et al., 1999). These and other observations (Butler et al., 2001; Butler et al., 2003) indicate that a functional MC4R is not necessary for sens ing and responding to nutritional deficits. The question of the circumstances u nder which rodents would exhibit a hyperphagia was first raised in connection with the effects of VMH le sions. Rats with these lesions exhibit hyperphagia only under co nditions of low cost (Teitelbaum, 1957); these and other reports suggest VMH lesion ra ts will work and eat to maintain a basal weight, but that hyperphagia and obesity en gage a separate mechanism (Sclafani & Kluge, 1974). The present findings in MC4RKO mice are strikingly similar. This is not simply a characteristic of all hyperphagic mice in our apparatus, because leptin deficient mice were able to sustain hyperphagia and weight under similar conditions (Vaughan & Rowland, 2003). Further, the fact that MC4RKO mice have lower locomotor activity than WT (Chen et al., 2000b; Ste Mari e et al., 2000) cannot account for these observations because leptin deficient mice ar e also hypoactive (Segal-Lieberman et al., 2003). Our results suggest that while decr eased MC4R signaling may be consistent with


36 increased food intake, then expression of that behavioral phenotype is highly dependent on the environment. The conditions under which MC4RKO mice are hyperphagic thus remain to be addressed.


37 CHAPTER 3 FOOD MOTIVATED BEHAVIOR OF MC4RKO MICE UNDER A PROGRESSIVE RATIO SCHEDULE Introduction In the previous chapter, we used a fixe d ratio (FR) schedule to document feeding patterns in mice and found that mice with the MC4R deletion were not hyperphagic in the operant chamber. This led us to hypothe size the MC4RKO mice did not overeat due to lack of motivation when required to work fo r food. The MC4R is found in areas of the brain believed to be involved in motivationa l aspects of feeding a nd it is possible that these KO mice may have impairments coding the reward value of food. This impairment could be why MC4RKO mice are only hyper phagic when no effort is required. The previous experiment showed that MC 4RKO could learn to lever press for food and in the present experiment we applied a di fferent operant schedule that is specifically designed to assess reward value. Progressive ratio (PR) sc hedules require a subject to emit an increasing number of responses to ga in access to successive reinforcers (Hodos, 1961). This type of schedule has been mo stly used in pharmacological studies to measure reward value of drugs but feeding lite rature has adapted the protocol to test the rewarding aspects of food (e.g., Staffo rd et al., 1998; Low et al., 2003). In a PR, the experimenter may choose to have the required responses increase in one of a large number of step functions, but arithmetic or geometric progressions are the most commonly used. At some point, animals will cease responding under such schedules. This in turn requires the expe rimenter to set an elapsed time with no


38 responding as a reset criterion or session e nding criterion. Under th e reset criterion, the schedule will reset to the initial value and the animal will have to start a new run of responses. The highest ratio completed be fore the reset is cal led the breakpoint in responding. A high breakpoint is interprete d as a high level of motivation for the reinforcer (Findley, 1958; Hodos, 1961). Previous protocols have us ed open economies with short, daily sessions. Short sessions are often equivalent to an open economy in which the experimenter determines the window in which the animal could possibly e ngage in feeding. In short sessions there are differences between performance of ad libitum fed rodents and performance of rodents kept at 80% of their free feeding we ight. In prior reports, food deprivation motivated rats to work through high ratios a nd emit large numbers of responses for only modest reinforcement (Hodos & Kalman, 1963; Jewett et al., 1995 ) . For our protocol, we wanted to use a clos ed economy or a paradigm of continuous food availability such that f ood intake is initiated and term inated by the subject and the focus is on the meal. In contrast to othe r studies, our mice were not food deprived and thus we expected to see low breakpoints, at least for wild type mice. In open economy studies, breakpoints usually end a session. In our protocol, the br eakpoint was a reset criterion at which the response requirements re verted to the initial value. Hence, the animal has complete control over the initiation and termination of meals. We hypothesize that the PR will lead to many small feeding bouts b ecause the first pellets in any meal are the cheapest. Thus, a stra tegy of many small feeding bouts will minimize work output. Each session was 23.5-hr in duration.


39 The PR protocol is designed to assess th e rewarding nature of the pellets and the motivation of the animal to work for food. Therefore, if animals are more motivated within a meal, the meal size (work toleran ce) should increase. Applied to MC4RKO mice, there are at least three possible outco mes. First, MC4RKO mice could perform comparably to wild type mice and show no overeating, as seen under FR schedules (Vaughan et al., 2005). Secondly, along those sa me lines if indeed the MC4RKO mice are less motivated to eat, then they will eat less than the WT mice under the PR schedule. Thirdly, we could see overeating represented by larger meals, more frequent meals or both. Larger meals would be reflected in high breakpoints, indicating that KO mice were more motivated to work for food within a meal. Materials and Methods Animals and Housing Environment Untimed pregnant heterozygous mothers were donated from a colony maintained by Dr. Haskell-Luevano at the University of Fl orida. Mothers gave birth 3-19 days after being moved to the Psychology department colony. All offspring born in the Psychology Department were used in a previous experiment in which they were reared in litters of different size or with diet of different fat content available. This previous experiment was conducted from age 1-77 days, and the mice were allowed back to standard maintenance conditions for at least 12 week s before the present work. Five MC4RKO, five heterozygotes (HET) and five wild type (WT) adult male mice were used in the current experiment. Only males were used in this experiment so that estrous cycles would not affect daily intake and meal patt ern data. Mice were genotyped using PCR to amplify DNA from a tail snip ta ken between ~8-10 wks of age.


40 The vivarium was illuminated from 0800-2000 h and was maintained at an ambient temperature of 23+ 2oC. Mice were housed individually in standard shoebox cages with water and food (Purina 5001 Chow) available ad libitum. For the first portion of the experiment, food intake was monitored daily for seven days for a baseline measure of free feeding. Food was measured by weighing the contents in the wire cage lid and bedding was checked for crumbs. Mice were then familiarized with the e xperimental food by presenting a jar of 20mg nutritionally complete food pellets (Noyes Precision pellets, Research Diets Inc., New Brunswick NJ; 25% protein, 64% carbohydrate, 11% fat) for a day. Next, mice were placed into the operant chamber (MED Associates Inc., St. Albans, VT) for a magazine training session for one day. During the se ssion, a pellet was automatically delivered every 8 minutes for 24 hours. Water was availa ble from a spout in the middle of the wall, opposite the levers. The chambers were cont ained inside ventilate d, sound attenuating cubicles, each with a 15-watt light providing th e same 12:12 cycle as the vivarium. Over the course of next 3-5 days, mice we re placed on fixed ratio (FR) schedules which increased (FR1 FR2 FR3) when mice showed sufficient lever pressing to maintain normal food intake. During FR tr aining mice had access to one lever. Located above the lever was a cue light that was on throughout the session. Pellet delivery was not additionally cued by any other com ponent except for the relatively noiseless advancing of the pellet dispenser. The FR pha se of this experiment established whether the mice would work for food. Four identical ch ambers were available, so mice were run in squads. Squads were run through FR sche dules and after stable performance (~7 days) at FR3, mice were moved on to the PR1 schedule.


41 Progressive ratio sessions On the PR1 schedule, the required responses for each successive pellet within a defined meal increased arithmetically (i.e. re quired responses = 1 pre ss, then 2 presses, then 3 presses). Each session ran for 23.5 hr . The session began with the left cue and house lights on. Pressing could begin immedi ately and there was no additional cue for pellet delivery (i.e. no time out period). Whenever there was a 3-minute stretch of no responding, the program reset. Resetting retu rned the lever press requirement to 1 and the animal was able to initiate a new period of pressing when desired. Using this PR, we observed that presses and food intake occurred virtually throughout the day and that mice were initiating many small f eeding bouts and avoiding work. To prevent such a fragmentation of meal patte rns, we then modified the PR schedule by imposing a 20minute reset criterion, but with the same PR1 incremental structure within meals. The PR1 schedule with the 20-minute reset also utilized cue lights. The session began with a one minute time out with the le ft cue and house lights off. After the one minute, the left cue light was illuminated and rats could perform their PR1 run on this lever. During pellet delivery the house light came on briefly and was otherwise off. Whenever 20 minutes elapsed with no respondi ng, the program reset. At the 20-minute reset marker, a new run was signaled to th e animal by a 1 minute time out without cue light illumination. The first PR schedule, with the 3-minut e reset conditions ran for 12-13 days for each mouse. Mice were then out of the chambers and back on ad libitum feed for 9-14 days. Next, the second PR schedule with a 20-minute reset was run for 12-13 days. The number of responses made by the animal was recorded. Those values were compiled to calculate an average breakpoi nt for the daily session. Response frequency and food


42 intake throughout the sessions were measur ed. Body weights were recorded at the beginning, during and end of each block of PR testing. Cue lights, real time records of pellet deli very and lever presses were controlled by Med-PC software (MED Associates Inc., St . Albans, VT). During all operant sessions, mice lived in the chambers with the excepti on of 30 minutes in the middle of each day when they were removed to a holding cag e without food while the chambers were cleaned and serviced. Data Analysis The individual mean total intake per day in each PR schedule was calculated using 12 consecutive days of stable performance w ith correction for food spillage. Food spillage was counted as the number of pellets found be neath the grill floor, and this amount was subtracted from the daily intake. Genotype was used as the grouping variable, whereby the genotype means used for analyses were deri ved from individual total intakes. Student t tests were used to analyze data. The significance criterion was set at p<0.05. Results During the FR 3 condition where mice had to lever press in the chamber for all of their food, MC4RKO mice ate significantly more than the other two genotypes (p<.001). Based on the last 12 sessions, the compilation of daily session means are shown for each genotype in Figure 3-1. Du ring the 3-minute reset criterion, breakpoints for all mice were relatively low, ranging from 16-35 pr esses. During the 20-minute reset criterion, MC4RKO mice had significantly hi gher breaking points than the other genotypes; their breakpoints ranged from 49-112 presses. Duri ng the 20-minute reset phase of the PR1 sessions, all genotypes significantly di ffered from each other (p< .001).


43 Table 3-1. Mean daily food intake (g) during phases of experiment. Wild type (n=5) Heterozygous (n=5) Knockout (n=5) FR 3 4.2 0.1 4.6 0.1 5.6 0.1** PR 3-min reset 4.0 0.1 4.5 0.1* 5.2 0.1** PR 20-min reset 4.2 0.1 4.7 0.1 * 5.7 0.1** Values in table are the mean SE. ** Intake greater than HET and WT mice within phase (p<.001). *Intake greater than WT within phase (p<.001). Intake greater than WT within phase (p<.05). Mean breakpoint per session 0 20 40 60 80 100 120 140 Wild type n=5 Knockout n=5 Heterozygous n=5A#* Mean breakpoint per session 0 20 40 60 80 100 120 140 Hetorozygous n=5 Wild type n=5 Knockout n=5B*# Figure 3-1. Breakpoints for WT , HET and MC4RKO mice. Bars represent the mean SE breakpoint values for all mice. Graph A shows breakpoints under the 3minute reset criterion. *MC4RKO mi ce had the highest breakpoints in comparison to WT (p< .001) and HET (p<.05). # HET and WT mice also significantly differed (p<.001). Grap h B shows breakpoints under the 20minute reset criterion. All genotypes significantly differed from each other (*KO> # HET> WT; p< .001). Food intake across PR schedules were ch aracteristic of genotype, with the MC4RKOs eating the most (see Table 3-1) . All the genotypes differed significantly (MC4RKO > HET > WT; p<.001). Increasing the reset criterion did gi ve the sessions more structure by reducing the “snacks” (s mall feeding bouts separa ted by at least 20 min) the mice appeared to be having thr oughout the day. Mice took a mean of 41-50 snacks during the 3-minute reset criterion (s ee Figure 3-2). The number of snacks was


44 reduced significantly for all genotypes to an average of 22-25 snacks per day during the 20-minute reset condition (p<.001). Number of small bouts 0 10 20 30 40 50 60 WT 3 min HET 3 min KO 3 minA* Number of small bouts 0 10 20 30 40 50 60 WT 20 min HET 20 min KO 20 minB*# Figure 3-2. Mean SE number of snacks or small feeding bouts initiated by the mice under both PR1 conditions as denoted on the y axis. Graph A represents data from the 3-minute reset condition. * Numb er of snacks were greater than HET and KO groups (p<.001). Graph B represents data from the 20-minute reset condition. *Number of snacks were greater than HET (p<.05) and KO (p<.001). #Number of snacks were greater than KO (p<.05). Response frequency differed among all three genotypes, as shown in Figure 3-3. MC4RKO mice pressed more than the other two genotypes in each reset condition. Wild type mice pressed the least and the performa nce of HET mice was intermediate in both reset conditions. Figure 3-4 shows the last day of each phase of the PR1 schedules for three mice, one from each genotype. Since mi ce were run in different squads, the start time (denoted as 0 on the x axis) was approxi mately the same. During the 3-min reset criterion all mice appeared to have similar pattern for the first few hours of the dark cycle. KOs took the fewest number of sma ll feeding bouts and seemed to be taking a longer period to eat in compar ison to WT and HET mice. During the 3 min reset PR1 phase the HET mice were accumulated the most responses over the 23.5-hr session, whereas during the 20-min reset phase KO mice had the highest number of responses across the sess ion. This is consistent with their high


45 Total responses per session 0 500 1000 1500 2000 2500 AHET n=5 WT n=5 MC4RKO n=5*# Total responses per session 0 500 1000 1500 2000 2500 BHET n=5 WT n=5 MC4RKO n=5#* Figure 3-3. Twelve day average of total re sponses made per daily session. Graph A shows data from the 3-min reset crit erion condition. *KO mice had a higher amount of responses daily than HET (P<.01) and WT (P<.001). #HET mice pressed more than WT (p<.001). Graph B shows data from the 20-min reset criterion condition. *KO mice had a hi gher number of responses daily than HET and WT (P<.001). #HET mice pressed more than WT (p<.001) during this condition as well. ATime (min) 0200400600800100012001400Bins (5 presses/bin) 0 100 200 300 400 500 600 het3; 0 = 3:19 p.m. ko3; 0 = 3:26 p.m. wt1; 0 = 4:34 p.m. lights off BTime (min) 0200400600800100012001400Bins (5 presses/bin) 0 100 200 300 400 500 600 het 3; 0=2:58 p.m. ko3; 0=3:19 p.m. wt 1; 0=3:54 p.m. lights off Figure 3-4. Cumulative responses of individual mice across a daily session; the actual start time for each session is indi cated in the legend. Graph A shows cumulative responses for the 3-min reset condition. Graph B shows cumulative responses for the 20-min reset condition. Stepwise response increments are steeper during the 20-mi n reset condition for these 3 mice.


46 Weight 20 30 40 50 60pre ad lib measure dur i ng ad l ib pre FR 1 FR3 e n d FR3 pre PR1 3min du r i ng end pre PR120MIN du r i ng endgrams HET WT KO Figure 3-5. Weights of mice during experiment al phases. Group means are represented. KO > HET > WT, p < .001. breakpoints and high count of le ver presses. However, KOs took fewer feeding bouts. Weights of the mice stayed stable throughout th e study (see Figure 3-5). As seen in other reports, MC4RKO mice were significantly heav ier than the other two genotypes. Discussion The present experiment shows that MC4RKO mice are capable of exhibiting motivation under a PR schedule. This is the first study in which animals deficient in MC4R function have been subjected to a PR schedule, and we found significant differences in certain parameters as a f unction of genotype. MC4RKO mice had higher breakpoints and a greater number of respons es emitted than HET or WT mice. Food intake patterns of MC4RKO mice showed the least amount of feeding bouts but the largest amount of food consum ed. WT mice showed the opposite food intake profile of MC4RKOs and HET mice were intermediate between the other two genotypes. As mentioned before, previous studies with shorter sessions found differences between performance of ad libitum fed rodents and performance of rodents kept at 80%


47 of their free feeding weight (Hodos & Kalm an, 1963; Jewett et al., 1995). We used a closed economy such that food intake is in itiated and terminated by the subject and the focus is on the meal. Our mice were not food deprived and initiated their own pattern of feeding bouts. This is most likely why we initially saw low breakpoints for all mice during the 3-min reset phase. Deprivation has been cited in the literature as a reliable means to train an animal to respond for a food reward (Heffner & Heffner , 1995). We did not use an experimenter imposed variable (deprivation) and thus wh en given the option, mi ce clearly elected to not go very long without eating. PR schedules have been used predominantly in short sessions where breakpoints end the session, but in our case they define elective meal termination and so our interpretation of the br eakpoint is most releva nt to the assessment of the subjects’ ability to sense fullness and end a di screte bout of feeding. Breakpoints were higher for MC4RKO mi ce in the 20-minute reset condition in comparison to the 3-minute reset condition. At an average of 111 47 presses before taking a break, KO mice received an average of 14 pellets (0.28 g) per defined meal. During this period KO mice were initiating fe eding bouts on average 22 0.6 times per day. At the highest breakpoint of 158 pre sses, KO mice would have accumulated ~17 pellets (0.34 g) in that discrete period of pressing. These feeding bouts were not very large in size relative to total daily intake but th is suggests that they avoided repeated meal initiation by pressing for longer periods th an the HET or WT mice. Due to the hyperphagic phenotype, it could be said that the KO mice were internally driven to maximize food intake. MC4RKO mice paid a hi gher price per pellet than WT mice as


48 reflected in a high number of lever presses and breakpoints and thus can be considered more motivated than WT mice. In the foraging protocol, effort plays a la rge role in meal ta king. Hedonics, the evaluation of palatability, becomes a mediating variable in the relationship between effort and feeding. Mice with defi ciencies in motivation to work for food, such as dopamine receptor (D1R) deficient mice show decreased pressing in comparison to wild type mice (El-Ghundi et al., 2003; Low et al., 2003). So it is plau sible that MC4RKO mice do not have an impairment in hedonics as dem onstrated by breakpoints and response rates higher than wild type mice. KO mice ma y organize feeding to maximize metabolic benefit and caloric balance, in other words maintaining body weight in relation to energy expenditure. There was a reduction in small feeding bouts seen for all mice to an average of 2225 bouts per day during the 20-minute reset condi tion. This shows that time available for feeding plays an important role in feeding st rategy. The MC4R app ears to play no role, under these conditions, in adjusting effort to time constraints. Under the 3-minute reset criterion, WT mice showed the lowest numbe r of presses with the highest amount of small feeding bouts. It has been shown that when initial price of pellets is high, meals tend to be smaller in rats (Johnson et al., 1993 ). Regardless of our reset criterion, our initial pellet price was always 1. Mice of all three genotypes took a smaller number of feeding bouts during the 20-min condition vs. the 3-min condition. This difference not seen in Johnson’s study (1993) could be because they had a 10-min reset criterion and the PR stepwise increments were steeper. Add itionally, they had contingencies on two levers


49 (procurement and consumption) and we used only one. The parameters of our experiment allowed for less effort on the part of the animal by adopting a meal strategy. Compared with their baseline, food intake stayed the same for KO mice when in the operant chamber but unexpectedly decrease d for HET and WT mice. It is possible that though KO mice are missing an important receptor for the control of food intake we still do not fully know what effect this ha s on postingestive feedback. The HET and WT mice may have more accurate negative feedback to end meals in comparison to KO mice and this may be most evident when mice ha ve to work for food. HET and WT mice ate less than KOs but still maintained their body weights overall; this could be due to a number of external and/or internal factors. We examined the data for genotype-speci fic 24-hr feeding patterns by analyzing records from three representative mice (see Fig 3-4). The three mice showed different profiles under the 3 min vs. the 20 min crite rion. The representative WT mouse stopped eating in the early morning hours, consistent with previous work in mice (i.e. Anliker & Mayer, 1956). The time constraint impos ed on mice in the 20 min reset condition resulted in an increased rate of pressing for all genotypes. In an early study, Anliker and Mayer (1956) found that ob/ob mice on an FR25 schedule had very high response rate. ob/ob mice were hyperphagic on this FR schedule and showed no cyclical day/night pattern of eating. A similar consistent eating pattern characteristic of hyperphagia was also seen in the MC4RKO mouse shown in Figure 3-4. Collier (1986) has shown that under FR schedules in which work is inescapable, rats take fewer, larger meals thus dist ributing their effort in many small bouts. It is thus not surprising that this pattern is also found under PR conditions in which work is avoidable


50 by initiating small feeding bouts. We observed no decrease in weight when mice were in the chambers pressing for food. This sugge sts the effort of lever pressing did not contribute a significant metabolic demand. The deletion of the MC4R is intrinsica lly involved in increas ing food intake in mice (Huszar et al., 1997). Our previous experiment in Chapter 2 did not find MC4RKO mice to be hyperphagic in the chamber, howev er when tested in a 24-hr PR schedule MC4RKOs did exhibit signifi cant and showed evidence of planning feeding bouts to maximize the amount of effort that contributed to feeding. It appears that a PR schedule may more accurately allow expression of the free feeding schedule of food intake, although why that should be the case and w hy the foraging PFR schedule does not is not immediately clear. Through this study and others it is eviden t the MC4R gene positively influences descending information regarding feeding related behaviors. Though the MC4R is located in brain areas ascribed to be i nvolved in motivation, the involvement of the MC4R seems to have the most significant role in the quantitative aspe cts of meal taking, like how much is eaten and how many calories to expend to ga in access to food. This PR experiment has successful applied paramete rs of behavioral economics to MC4RKO feeding. Such an approach is useful tool in discovering how subjec ts adjust meal taking to different simulated environments.


51 CHAPTER 4 SHORT TERM FOOD INTAKE AND MEAL SATIETY IN MC4RKO MICE Introduction Changes in meal frequency are attributed to modulation of processes that sustain intermeal intervals (satiety) wh ereas changes in meal size are attributed to modulation in the processes that terminate ongoing episodes of eating (satiation). Satiation and satiety may have both common and distinct compone nt mechanisms. The study of single meals has been advantageous in the past to discover exactly how endogenous and exogenous agents affect both satiation and satiety. A preload is a small amount of food (liqui d or solid) given to a subject either via the mouth or directly infused into the gastro intestinal (GI) tract. The preload is followed by an interval, usually determined by the nature of the preload, then a test meal (Seeley et al., 1993). The magnitude of the test meal following the preload gives information about the individual’s appraisal of hunger and satiet y. For example, a larger test meal may be observed after a low calorie preload in compar ison to a high calorie preload (Warwick et al., 2000). Manipulating preload volume is a common wa y to test an individual’s ability to detect the sensation of “fullness” after a m eal as defined by a decr ease in a subsequent test meal. This interpretation holds true for oral, gastric and duodena l preloads (Seeley et al., 1993, 1994; Azzara et al., 2002) illustrati ng that multiple sensory dimensions are involved in interpreting signals of satiation and satiety.


52 MC4RKO mice exhibit robust hyperphagia (H uszar et al., 1997) and may overeat due to a possible role for the MC4R in gast ric feedback. Rats exhibited a dose dependent decrease in glucose drinking following admi nistration of MT II, a MC3/4R agonist, suggesting the opposite effect would be reached from MC4R receptor blockade. In that study, MT II was administered icv followed by an intraduoednal pr eload; immediately after the preload, rats were given a 1hr gl ucose drinking test (Azzara et al., 2002). Additional evidence comes from a recent study by Azzara and colleagues (2005) who examined preload administration of nutrients in MC4RKO mice. Nutrients of various caloric contents were given by means of intr aduodenal catheters a nd were less effective in decreasing the liquid test meal in MC4R KO mice vs. wild type mice. This suggests that MC4RKO mice have an impairment in satiation and/or satiety. Sham-feeding and repletion-depletion st udies are generally done to identify endogenous feedback signals. Sham-feeding st udies utilize surgical procedures to eliminate gastric distention after a feeding bout. Fistulas can be placed in the stomach so that food is emptied upon ingestion and is not absorbed, hence the term sham feeding. Sham meals are usually larger than normal meals because there are no post-gastric feedback signals to end the feedi ng bout (Davis & Campbell, 1973). In depletion studies, the endogenous ener gy stores are thought to function as pacemaker signals letting individuals sense when feeding should occur. During a fast, hypothalamic peptide and receptor regulation are critical for subsequent feeding behaviors. Food restricted rats show significant increases in 125I labeled NDPMSH binding to MC4R in comparison to cont rol rats on ad libitu m food intake. Food restriction was defined as a10-day period wher e rats were allowed to eat 60% of their


53 normal intake (Harrold et al., 1999). It has been recently shown that after a 17-hr fast, 125I-AgRP is upregulated in liver, adrenal glands and adipose tissue of mice in contrast to fed mice; uptake in these areas is mediated by MCRs (Pan et al., 2005). Fasting also increases activity of AgRP containing neurons suggesting an increased signal to increase food intake (Takahashi & Cone, 2005). In the absence of functi onal MC4R, we expect that KO mice will be as sensitive to a fa st as heterozygous or wild type mice. MC4RKO mice and wild type mice have di fferent metabolic rates and thus may sense hunger differently (Chen et al., 2000b; Al barado et al., 2004). Bu tler et al (2001) conducted a study where MC4RKOs were food re stricted for several days then were allowed to re-feed freely. Both wild type mice and KOs regained weight at similar rates. These results suggest WT and KOs can adjust to partial food restri ction. We conducted a preliminary study to determine if MC4RKO and wild type mice respond similarly to different deprivation periods by increasing their subsequent intake. The findings are discussed later in this chapter. When pair fed with wild type mice, MC4RKO mice increase body weight but at a slower rate than KOs with ad libitum access to food (Ste Marie et al., 2000). MC4RKO mice show decreased body weight when ther e is a running wheel present in the home cage or effort is required for food (Irani et al., 2005; Vaughan et al., 2005). However, when returned to ad libitum feed, MC4RKO mice reverted to hyperphagia and increased body weight gain (Ste Marie et al., 2000; Irani et al., 2005). The common underlying factor that causes the rebound hyperphagia in these studies may be that without any external constraints MC4RKO mice do not have the ability to regulate meal size and frequency based on chemical and physical ch aracteristics of food already consumed.


54 To build on previous findings, in this st udy we gave an oral preload of differing volumes and then measured the effect on test meal intake. The volume needed to relay a sense of fullness in MC4RKO mice may be larger than that needed in wild type mice. In our study, we chose to give an oral preload instead of a gastric or duodenal preload. One advantage of using elective oral intake rather than GI infusi ons (cf Azzara et al., 2005) is that flavor cues will trigger the cepha lic reflexes involved in digestion. Control and regulation of GI function st arts with the cephalic phase then is followed by the gastric and intestinal phases. The cephalic phase begins with the sensory experience of food followed by an increase of parasympathetic innervation to the GI tract. This initial phase of feeding begins the stomach’s preparation for incoming food. Allowing MC4RKO mice to ingest their pr eload naturally will recruit physiological mechanisms that mimic how humans consume food. This aim of the present experiment is to determine the effects of volume on sati ety feedback and thus indirectly assess the sensitivity of MC4RKO mice to gastric and postg astric feedback signals relating to meal termination. Materials and Methods Animals Sixty mice were used in this experime nt. Fifty were born in the Psychology Department from mothers donated by Dr. Hask ell-Luevano and were used in previous experiments. Six of the 50 aforementioned mi ce were used in the experiment in Chapter 3. Ten additional mice were donated by Dr. Ha skell-Luevano and were experimentally nave. Ages of mice ranged from 4-12 months old. We noticed no systematic differences in the results between the mice from thes e different origins. Mice were housed individually in standard shoebox cages with water and food (Purina 5001 Chow; 3.6


55 kcal/g) available ad libitum, unless othe rwise noted. The vivarium had lights on 07001900 hr with an ambient temperature of 23+ 2oC. For both of the following studies, the spacing of testing days was partially dependent on time constraints of fasting a nd weight maintenance of the mice. The criteria for initiating food depriva tion were either no greater than a 2% change of the animals’ free feeding body weight before the expe riment, or 4 days since the last test. Food Deprivation Study All mice were nave to Ensure (Ross Laboratories, Columbus, OH; 14.4 % protein, 64 % carbohydrate, 21.6 % fat) and were first acclimated by attaching 10-ml pipettes (pipettes were fitted with metal dr inking spouts and rubber st oppers) to the home cage for 30 minutes for two days. Mice were then food deprived for 6, 12, or 24 hr on three separate occasions spaced ~5-6 days apart. The first occasion occurred a day following the two day familiarization period. After each food depriv ation period, mice were given Ensure (1.1 kca l/g) and their intake was monitored for 60 minutes. Intake readings were taken every five minutes for the first 15 minutes and then every 15 minutes for the remaining 45 minutes. Preload Study First, mice were randomly assigned to groups . For a repeated measures design, all mice progressed through four different condi tions in a randomized order. The four conditions differed in presence or absence of pr eload and the preload to test meal interval (see Table 4-1). Mice ran through each of the four conditions three times to control for possible conditioning due to order effects.


56 Table 4-1. Experimental design for preload study Condition Preload Preload access time Preload to test meal interval Test meal access time 1 (P30) Yes 15 min 30 min 30 min 2 (N60) No 15 min 60 min 30 min 3 (N30) No 15 min 30 min 30 min 4 (P60) Yes 15 min 60 min 30 min The preload given was 1.13 ml, which was ~50% of the mean volume consumed by all mice after the 12-hr deprivation period in the food depriva tion study. The 1.13ml preload was measured using an adjustab le volume Eppendorf Reference Series 2000 Pipette. Ensure was pipetted into a 5-ml beaker and then transferred to a 2-ml Fisherbrand pipette and closed with a silicone stopper. Th e pipette was attached to the home cage during the preload access period. A 15 min preload access time was used to insure that the entire preload was consume d. Water was not availa ble during the preload and test meal access periods. Weights were recorded before and after the 12-hr deprivation periods. Data Analysis In the food deprivation study, univari ate ANOVAs were done comparing the effects of time point, genotype and deprivati on period on Ensure intake. If significant differences were found Bonferonni post hoc test s were done to determine at which level the independent variables had effects. In the preload study, one way ANOVAs were done to determine if there were differences between the sessions within a block. No differences were found so the sessions were pooled into blocks and block became a new grouping variable. Mice that did not drink the entire preloa d in any of the 3 consecutive sessions within a block were not included in analyses. A Student t test was used to compare the last session mean


57 within a block to the first session mean in th e next block to look for conditioning effects. Univariate ANOVAs were done to determine th e effects of gender, genotype, preload and deprivation period on Ensure intake during each block. Weights were compared using a univariate ANOVA to look at the effects of gender and ge notype; Bonferonni post hoc tests were done when significance was found. Fo r all analyses, the si gnificance level was set at p<0.05. Results Food Deprivation Study On the first day of the acclimation trials , HET mice drank more (1.2 .2 ml) than either WT (0.6 .1 ml) or KO (0.4 .1 ml) mice. On the second day of the acclimation trials, WT (1.6 .1 ml) and HET (1.6 .1 ml) increased their intake to be comparable 0 0.2 0.4 0.6 0.8 1 1.2 Time points Volume consumed (ml) HET 6hr HET 12 hr HET 24 hr WT 6hr WT 12 hr WT 24hr KO 6hr KO 12hr KO 24hr5 min60 min 30 min 15 min 10 min45 min Figure 4-1. Absolute Ensure intake at different time points. Each point represents the group mean. There was a significant eff ect of time point on intake (p < .001; 5min > 10min > 15 min > 30 min > 45 min = 60 min).


58 to that of the KOs (1.4 .1 ml). Most of th e intake occurred within the first 15 minutes thus the allotted time was sufficient for maxi mal intake, as expected. Intake changed significantly with time across th e test session (p<.01). Th e highest intake of Ensure (see Figure 4-1) occurred in the first 5 min. Post hoc tests revealed th at the 5 min and 10 min time points had the highest amounts of Ensure consumed. There was an effect of deprivation period (p<.05) but no effect of genotype; for all mice the most Ensure was consumed after a 24-hr fa st. However, MC4RKO mice drank less overall than the other two phenotypes afte r the 6-, 12-, and 24-hr periods (see Table 4-2). Since mice drank comparably after all the deprivation periods, we used the intermediate 12-hr period in the preload study. Table 4-2. Mean intake (ml) after deprivation 6hr dep 12hr dep 24hr dep WT 2.28 2.41 2.40 HET 2.24 2.32 2.24 MC4RKO 1.98 2.15 2.26 Preload Study Analyses were done to compare the three in dividual sessions w ithin a block. There were no significant differences between any of the sessions within any of the four blocks indicating that there were no c onditioning effects due to the re peated testing. There were no significant differences between the third se ssion of one block and the first session of the next block, with the exception of the last comparison (see Table 4-3). Though there is statistical significance between the sessions in the last comparison, the difference is rather small, 2.01 ml vs. 2.31 ml.


59 Table 4-3. Evaluation of rando mized block design on conditioning Within block comparison Between block comparison (3rd of 3 to 1st of next 3) Block P value P value 1 .456 .589 2 .305 .307 3 .349 .033 4 .305 n/a Individual sessions were then collapsed into blocks and these results are shown in Figure 4-2. The only consistent main effect acro ss blocks was of the preload manipulation (p<.001). Other main effects were occasionally significant as follows: gender had an effect in blocks 1 and 4 (p<.01), genotype ha d an effect in block 2 and the preload-test meal interval had an effect in blocks 3 a nd 4. To streamline interpretation and analyses, all data were then analyzed for main effects of preload, ge notype and preload-test meal interval. Effects of the three in dependent variables were found. Mice drank significantly less after a preload (p<.001) and correspondingly drank more after a longer preload-test meal interv al (p<.01). There was a genotype dependent effect in Ensure intake and post hoc tests revealed that KO mice drank more than WTs. The only interaction seen was between presen ce of a preload and the preload-test meal interval (p<.01). When there was a preload, th e longer preload-test meal interval resulted in more intake.


60 A. 30 min ml of Ensure 0.0 0.5 1.0 1.5 2.0 2.5 3.0 Wild type Heterozygotes Knockouts P NP P P NP NP * * * B. 60 min ml of Ensure 0.0 0.5 1.0 1.5 2.0 2.5 3.0 Wild type Heterozygotes Knockouts P NP P P NP NP * * * Figure 4-2. Ensure intake separated by preload c ondition and genotype. *No preload (NP) groups drank more than preload (P) groups (p< .001). Genotype had an effect (p< .05) on intake after the 30 mi n preload – test meal interval (Graph A) but not after the 60 min interval (Graph B). N=20 mice for each genotype.


61 Figure 4-3. Mean weights of mice. Genotype (p<.001) and gender (p<.001) had main effects. The average weights of mice are depicted in Figure 4-3. Males weighed more than females in all groups (p<.001). There was an interaction between gender and genotype on weight. Since there was a negligible di fference between KO males and females, the interaction was mainly driven by the WT and HET weights. Between genotype differences were found; post hoc tests show ed that all genotypes’ differences were significant (p<.05). Discussion The main goal of this experiment was to assess behaviorally th e quality of satiety mechanisms in MC4RKO mice. MC4RKO mice did not cons ume more after a fast than wild type mice. This result is similar to a previous report in which food restricted MC4RKO mice ate comparable amounts to wild type mice after 1 day of resumed ad libitum access to food (Butler et al., 2001). MC4RKO mice thus are capable of sensing metabolic need comparably to WT and HET mice. grams 0 10 20 30 40 50 60 WT KO HET n=10 n=10 n=10 n=13 n=7 n=10 Male Female


62 In humans, palatability serves as an i ndicator for meal size (Rolls & Hetherington, 1989). In a study of normal weight humans , three deprivation periods yielded the greatest subjective appetite ratings after the longest period (4.75 hr) of fasting (Hulshof et al., 1995). The palatability of the Ensure used in this experiment allowed for substantial intake at all three of our deprivation periods. Since MC 4RKO mice were found to have no deficiencies in affective responsivene ss to brief-access trials of sapid compounds (Eylam et al., 2005), taste cues most likely had an influence on intake and allowed for intake that was relatively indepe ndent of deprivation period. Liquid meals empty relatively fast within a 60 min period (Collins et al., 1991). This is a probable explanation for the larger intake observed in all of our mice, regardless of genotype, after the 60 min preload – test meal interval vs. the 30 min interval. Corroborating data show that rats exhibited larger intake after a preload – test meal interval of 60 min in comparison to shor ter (20or 40-min) intervals (Warwick & Weingarten, 1994). Regarding our volume manipulation we f ound results contrary to Azzara’s study (2005). After our oral preload MC4RKO mice di d not vary significantly from wild types in their intake of a test meal. Methodologi cal differences may account for the findings of our respective groups. They administered intraduodenal infusions of different commodities while volume was held constant and as a control, all mice received equivalent distention of the duodenum by salin e infusions (personal communication from A.A.). In contrast, in our study we used gastric distention instead of duodenal and there was a clear control c ondition in which there was no prel oad, and the latter was done to


63 determine whether the absence of food would interact with the amount of food consumed during the test meal. Our study was designed to establish whether MC4RKO mice are capable of identifying a fasting state by discriminating presence or absence of a preload; we found that KO mice were indeed capable of both. The fact that Azzara et al. (2005) found MC4RKO mice had a lowered sensitivity after nutri ent infusions could also be due to the differing gastric emptying rates possessed by th e different genotypes. We found a minor genotype dependent effect on intake but no interactions after all the blocks were collapsed so a future study of gastric emptyi ng rate between genotypes might be useful. The main difference between our procedure a nd that of Azzara et al. is that our preload was administered via the mouth in stead of the duodenum. The cephalic and digestive phases of feeding add many dimens ions to the quality of food intake. For example, orosensory and cognitive behavior s have been associated with the cephalic phase of feeding. Food ingested orally rather than into th e gut results in an increased intake when the commodities are high in fat or carbohydrates (Warwick & Weingarten, 1995). There has been little work done with MC4RKO mice assessing their postingestive feedback mechanisms. Different models have b een used in rats to assess their capacity to sense fullness from a meal. Chronic decerebrate rats are not able to interpret signals of fullness due to signals only reaching the brains tem. Chronic decerebrate rats are able to decrease intake after a preload but cannot discriminate between food deprivation and no food deprivation showing that metabolic and ga strointestinal factors can be dissociated (Seeley et al., 1994). MC4RKO mice unlike chronic decerebr ate rats have shown that


64 they are able to interpret metabolic and gastrointestinal feedback. Our results show that MC4Rs are not necessary for the transmission of fullness and metabolic need which are signals carried to the brain via the vagus nerve. Mechanisms for the initiation or cessation of meals, i.e. mechanoreceptors and peripheral CCK receptors, appear to be functionally intact in KO mice. Early sham feeding studies have shown that rats will drink more than normal; it is thought that this occurs because there is no postingestive distention of the stomach. Davis and Campbell (1973) reported that rats with a gastric fistula drank successively more across individual sham drinking sessions sugge sting a conditioned response to stimulus exposure. The report hypothesized that condi tioned drinking can be ascribed to an increased orosensory control that compensate d for a lack of gastric control (Davis & Campbell, 1973). An impairment of the MC4R alone does not affect the ability to develop a conditioned drinking response. MC4RKO and HET mice did not show any systematic differences in drinking between blocks impl ying that there were no conditioned drinking responses. There was a statistical difference be tween the last sessions in the third block and the first session in the fourth block but the actual volumes collapsed across genotypes, were rather sm all (2.01 vs. 2.31 ml). The increase in sham drinking that has been observed over successive exposures has been attributed to rep eated pairings of orosensory input with no postingestive consequences (Mook et al., 1983). Our results gi ve further evidence that gastric feedback is playing a significant part in maintaining liquid meal intake follo wing a fast, regardless of a preload. Our study shows that when a llowed to experience the cephalic phase of


65 feeding MC4RKO and HET mice can monitor th eir postingestive state. Many studies have acknowledged the importance of taste cu es and other oral factor on liquid meals (e.g., Booth, 1972; Davis & Campbell, 1973; Mook et al., 1983). This experiment was able to describe the short term meal characteristics of MC4RKO mice in response to preload. Most of th e intake after a fast occurred within the first 15 min of access to the test meal during a 30 min test. KO mice also were able to sense differences in volume in the gut. These results suggest that th ere are intact satiety peptide release and mechanoreceptor feedback mechanisms in KO and HET mice. What still remains to be reported is the degree of sensitivity with which KO and HET mice can respond to ingested food. Further experime nts are necessary to add a qualitative perspective regarding ho w MC4RKO mice terminate individual meals.


66 CHAPTER 5 SHORT TERM FOOD INTAKE IN MC4RKO MICE AFTER CCK AND BOMBESIN ADMINISTRATION Introduction Peripheral pathways are involved in the communication of metabolic state to the brain. Satiety signals arise from the gastrointestinal tract an d inform the brain to engage in control of food intake (revi ewed in Berthoud, 2002). The si gnal is first relayed to the NTS, where there are neurons involved in si gnaling digestive responses (Smith et al., 1984). Neurons in the NTS receive taste and vi sceral input, which then relay information to forebrain structures an d hypothalamic nuclei involved in regulating food intake (reviewed in Berthoud, 2002). CCK, an endogenous octapeptide, is a put ative satiety hormone that relays information about an ongoing meal to the brai n via receptors on vaga l efferents (Garlicki et al., 1990). CCK exists in many forms; the most important fragment of the peptide is the carboxy terminal end. CCK is found in various areas of the brain including cerebral cortex layers (Rehfeld, 1978). CCK found in th e gut is secreted from endocrine cells in the mucosal layer of the duodenum. CCK release in the gut is trigge red by ingestion of food and activation of gastrin releasing peptide receptors on the CCK cell (Snow et al., 1994; Liddle, 1997). In humans, the fasting plasma levels of CCK are low in comparison to a replete state (Liddle et al., 1985). In rats, various de grees of fasting results in decrease of CCK plasma levels, duodenal c oncentration and mRNA levels (Koop et al., 1987; Kanayama & Liddle, 1991).


67 CCK binds to two classes of receptors, CCKA and CCKB: both of which have high affinity for CCK. Receptors are found in the muscle layers of the stomach and intestine (Smith et al., 1984). mRNA for both receptors is also found centra lly in the cortex, hippocampal formation and septum. In speci fic nuclei of the hypothalamus, mRNA for both receptors is found in the PVN and LHA (H onda et al., 1993). It was originally thought that CCKA receptors were found predominantly in the periphery, while CCKB receptors were found mostly in the central ne rvous system. More recent evidence shows that both receptor types are found in brai n and periphery (Moran & Ladenheim, 1998). There is confirmation that the meal termin ation effects of CCK are mediated through CCKA receptors (Corwin et al., 1991; Weat herford et al., 1992). Endogenous CCK delays gastric emptying therefore providing a lo ng lasting sensation of fullness (Shilabeer & Davison, 1987). Exogenous CCK decreases meal size and can prolong intermeal intervals, even in animals that have not been fasted (Hsi ao & Wang, 1983; West et al., 1984). A CCKA receptor antagonist, devazepide, can reverse th ese effects (Corwin et al., 1991). It has been suggested that exogenous and endoge nous CCK work differently to reduce food intake. Exogenous CCK reduces food intake by disrupting the normal migrating motor complex, a component of gastrointestinal motili ty exhibited by the intestines (Shillabeer & Davison, 1987; Rodriguez-Membrilla et al ., 1995). The contri bution of endogenous CCK is usually assessed by administering an antagonist. Devazepide inhibits ingestion and proglumide, a nonspecific antagonist, induces a decrease in gastric pressure (Shillabeer & Davison, 1987; Welle r et al., 1997). The net result of both of these actions as well as others is fullness and decrease or cessation of food intake.


68 CCK and MC4 receptors coexist in similar ar eas in the brain; they are both located centrally in the dorsomedial nucleus (D MN) of the hypothalamus (Honda et al., 1993; Mountjoy et al., 1994). OLETF rats, a gene tic model of obesit y, that lack CCK-A receptors have down-regulated MC4Rs, sugge sting a positively correlated relationship between the expression of these rece ptors (Lindblom et al., 2000). Presence of MC4Rs in the br ainstem suggests that the receptors are likely to be involved in brainstem mediated maintenan ce of food intake (Mountjoy et al., 1994). Evidence has shown that a MC4R agonist when injected into the fourth ventricle will decrease glucose intake (Williams et al., 2002) . This suggests that activation of MC4Rs found in the brainstem, particularly the DMX, may be key in reduci ng intake due to the sensation of satiety. Vagal efferents with CCKRs carrying visceral info have incoming and outgoing projections from the NTS. Fourth ventricle injection of a MC4R antagonist, SHU9119, was shown to block CCK induced stimulation of intracellular processing in the NTS and CCK induced inhibiti on of feeding. It has been proposed that NTS neurons integrate MC4R ac tivation with incoming vagal afferent information (Fan et al., 2004; Sutton et al., 2005). Another class of satiety signal involves the bombesin-like peptide family. Bombesin (BBS) was first isolated from amphibian skin and upon administration to mammals was found to decrease food intake by enhancing the sensation of satiety, therefore shortening meals (Anastasi et al ., 1971; Gibbs et al., 1979). The bombesin family of peptides includes neuromedin B and gastrin rel easing peptide (GRP) (McDonald et al., 1979).


69 Neuromedin B (NMB) and GRP bind to NMB receptors (bombesin BB1 receptor) and GRP receptors (bombesin BB2 receptor), respectively. Th ere is a bombesin receptor subtype – 3 (BRS-3; bombesin BB3 receptor), but no endogenous ligand has been identified (reviewed in Yamada et al., 2002). A fourth receptor has been isolated from frogs (bombesin BB4 receptor) and has a 56% and 61% sequence homology to the mammalian BB1 and BB2 receptors, respectively. The mammalian ligand is unidentified (Nagalla et al., 1995). BBS binds to both BB1 and BB2 receptors, but has the highest affinity for the BB2 receptor (reviewed in Yamada et al., 2002). Th e dominant receptor in rat brain is the BB1 subtype (Moran & Ladenheim, 1998). BBS is not found in mammals, however when administered it acts very similarly to endogenous GRP or NMB release (reviewed in Yamada et al., 2002). GRP receptors are found in the fundus of the stomach, olfactory nucleus and neocortex of mice (Ladenheim et al., 2002). Whether given intraperitoneally (ip) or via the fourth ventricle (4V) c-fos like immunoreactivity occurs in a variety of ar eas suggesting that receptors may exist correspondingly. After ip bombesin admini stration, Fos-like immunoreactivity occurs exclusively in the internal subdivision of the lateral parabrachia l nucleus and the area postrema. After 4V bombesin (30ng/3 l/rat) administration, Fo s-like immunoreactivity occurs exclusively in the supraoptic nucleus, lateral PVN, and locus coeruleus structures (Li & Rowland, 1996). It has been proposed th at BBS-like peptides vary according to region and this variation corresponds to the meal parameters (i.e. pran dial or preprandial) being observed (Merali et al., 1999)


70 Physiologically, peripheral GRP mediat es the release of gastrin, CCK and pancreatic polypeptide in humans and rats (Gibbs et al ., 1979; Konturek, 1994). Behaviorally, food intake is inhibited within the first hour after BBS administration and returns to normal about 120 minutes postinj ection in normal mice (Gibbs et al., 1979; Silver et al., 1988). Exogenous BBS suppresses feeding dos e dependently and reduces meal size in rats when administered both peripherally and centrally (Gibbs et al., 1979; Gibbs, 1985; Flynn, 1991). Icv BBS affects meal size in rats primar ily by lengthening intermeal intervals (Stuckey et al., 1985; Thaw et al., 1998). Lieverse and colleagues (1998) studied obe se and lean humans before and after BBS infusions. After a banana shake preload, lean subjects ate less of a test meal than subjects receiving saline. Obese subjects showed no effect of BBS on test meal food intake. Plasma BBS levels of lean a nd obese subjects did not significantly differ throughout the study, however basal CCK levels we re higher in lean subjects (Lieverse et al., 1998). We tested both satiety signals for tw o purposes. First, MC4RKO mice may have shown differential sensitivity to one vs. the ot her, like Zucker rats, another obese model. Previous evidence has shown that lean rats were equally sensitive to doses of BBS but were differentially responsive to doses of CCK . Obese rats actually increased food intake with low doses of BBS, while showing d ecreased sensitivity to doses of CCK (McLaughlin & Baile, 1980). Second, the site of action of each peptide is different. BBS requires both vagal and spinal visceral input whereas CCK requires gast ric branches of the vagus for their effects to occur (Gibbs, 1985). Add itionally, BBS dose dependently signals the release of CCK


71 (Kanayama & Liddle, 1991; Snow et al., 1994; Liddle, 1994), so if there was a discrepancy in the two peptide levels we could pinpoint if sa tiety is impaired in bombesin release or CCK release. By looking at feed ing responses after the different peptides, results could reveal whether and at what level MC4RKO feedback may be impaired. The only study to date (Fan et al., 2004) that has examined MC4RKO response to CCK-8 (3 nmol/kg) found that the peptid e had no significant effect on meal size reduction in MCRKO mice. We know of no publ ished report that has examined if BBS would have an effect on M4RKO mice. Evidence suggests obese MC4RKO mice may not be affected by BBS but due to partial pr esence of MC4Rs in he terozygous mice there may be some effect of the peptide on decrea sing test meal size. Previous reports show that bombesin decreased food intake in obe se mice but not more so than lean mice (McLaughlin & Baile, 1981; Taylor & Garc ia, 1985). We hypothesized that MC4RKO mice would not reduce test meal intake in response to exogenous CCK and would reduce test meal intake slightly or not be responsive at all to BBS. Materials Animals Male and female mice were obtained fr om the breeding colony of Dr. HaskellLuevano at the University of Florida. A to tal of 40 mice were used (12 WT, 12 HET and 16 MC4RKO) and ages ranged from 14 wk to 23 wk old at the beginning of the study. Males and females were included in all the genotypes but the relativ e numbers were not identical. All animals were included in the testing because we did not anticipate any major sex differences in the effects of the peptides. Originally group housed, mice acclimated to the Department of Psychology’s facilities for at least 5 days before being singly housed in standard shoebox cages with


72 water and food (Purina 5001 Chow) availabl e ad libitum, unless otherwise noted. The vivarium had lights on 0700-1900 h with an ambient temperature of 23oC. Mice were singly housed for 5 days before the beginning of the experiment. Drugs Bombesin and sulfated CCK-8 (Sigma, St . Louis, MO) were dissolved in 0.9% physiological saline vehicle. Procedure Adaptation Trials All mice were first adapted to Ensure (Ross Laboratories; 1. 1 kcal/g) using two 30 min tests. Mice had no previous experience with food deprivation and were temporarily food and water deprived during the 30 min intake adaptation trials. Satiety Peptide Responses Dose response curves for CCK-8 and bomb esin were determined within subjects. Mice were food deprived for 12 hr, and then give n alternate intraperitone al (ip) injections of peptide or vehicle. Mi ce that received bombesin did not receive CCK and vice versa. Five minutes after the injections, mice were given access to Ensure in their home cage for 30 min. Readings were taken at the beginn ing of 30 min period (0 min) and at the end (30 min). The doses of CCK-8 used were 2, 6, and 18 g/kg; the doses of bombesin used were 2, 4 and 8 g/kg. Both CCK and bombesin have s hort half-lives and we anticipated no carryover effects between testing days. These dose ranges were based on previous work in mice (McLaughlin & Baile, 1981; La denheim, et al., 2002; Fan et al., 2004, Chi et al., 2004). Doses of the peptides re ceived by the mice were administered by a modified Latin square design where saline in jections were interspersed between drug


73 injections. Injection of ve hicle or peptide within a group was counterbalanced for primacy effects. Testing was done every 3 da ys to insure that mice would be back at approximately ad libitum weight before each 12-hr fast. Data Analysis One way ANOVAs were done to find main effects of gender, dose of each peptide and genotype. Univariate ANOV As were done to find intera ctions between any of the independent variables. Student t tests were used to compare i ndividual intake means after peptide administration vs. salin e administration. Weights we re analyzed using a one way ANOVA. Bonferonni post hoc te sts were done where appropri ate. Significance level was set at p<.05. Results All mice drank during the adaptation trials . There was some evidence of novelty on the first day. However, mice increased inta ke on the second day of trials. Mice drank similarly regardless of genotype (see Table 5-1). In all geno types there was a significant effect of CCK administration on Ensure intake (p<.001). Post hoc tests revealed that the 18 g/kg dose of CCK had the greatest effect on intake for the genotypes. Student t tests comparing each CCK dose to vehicle were significant for all comparisons except the Table 5-1. Mean Ensure intake (ml) during adaptation trials. WT HET KO Day 1 0.6 .1 0.5 .1 0.7 .1 Day 2 1.2 .1 1.2 .1 1.5 .1 Values in table are the mean SE. HET intake after 2 g/kg vs. vehicle (see Fi gure 5-1). Post hoc tests also found that within the KO group the 18 g/kg dose of CCK had a larg er effect than the 2 g/kg dose (p< .05).


74 2 6 18 A. WT ml of Ensure 0.0 0.5 1.0 1.5 2.0 2.5 CCK (n=6) Vehicle mean ** * ***Dose (microg/kg) B. HET 0.0 0.5 1.0 1.5 2.0 2.5 CCK (n = 6) Vehicle mean ml of Ensure2 6 18 * ***Dose (microg/kg) C. MC4RKO 0.0 0.5 1.0 1.5 2.0 2.5 CCK (n = 9) Vehicle mean ml of Ensure26 18 Dose (microg/kg)* *** *** Figure 5-1. Ensure intake during 30 min test meal af ter CCK administration. Horizontal lines in each graph represent the genotype vehicle mean (solid) and the standard error (dashed). Intake after CCK doses are shown as the mean SE. Significance from post ip vehicle intake at .05*, .01** and .001*** levels. The effects of BBS were less pronounced than CCK and significant re ductions in intake were observed only in HET and KO genotype s (see Figure 5-2). Analyses found a difference between HET intake after 8 g/kg BBS in comparison to intake after vehicle injection (p<.01) and the same for KOs (p<.05). There were minor influences of gender. HET males (n=5) drank more than HET females (n=2) after both BBS injections and saline injections. KO females (n=6) drank more than KO males (n=2) after both BBS


75 A. WT ml of Ensure 0.0 0.5 1.0 1.5 2.0 2.5 BBS (n = 6) Vehicle mean 2 4 8 Dose (microg/kg) B. HET ml of Ensure 0.0 0.5 1.0 1.5 2.0 2.5 BBS (n=6) Vehicle mean 248 Dose (microg/kg)** C. MC4RKODose (microg/kg) ml of Ensure 0.0 0.5 1.0 1.5 2.0 2.5 BBS (n = 9) Vehicle mean 2 4 8 * Figure 5-2. Ensure intake during 30 min test meal af ter Bombesin (BBS) administration. Horizontal lines in each graph represen t the genotype vehicle mean (solid) and the standard error (dashed). Intake af ter BBS doses are shown as the mean SE. Significance from post ip vehicle intake at .05* and .01** levels. Testing Day grams 20 30 40 WT HET KO 12 3 4 56 Post testing * * * * * * * Figure 5-3. Weights of mice throughout testing. All genotypes differ significantly from each other at the different time points (p<.01). * KO mice are significantly heavier than WT mice (p<.05).


76 injections and saline in jections. It is difficult to ma ke a strong conclusion regarding gender dependent responses to the peptides due to the small number of animals in the subgroups and thus, low statistical power. Intake measures post vehicle injections we re pooled into two groups according to peptide and then separated by genotype. Th e three genotypes drank comparably after vehicle injections regardless of being assi gned to CCK (p=.51) or BBS (p=.48) groups. There was a genotype dependent difference in weights (see Figure 5-3). Across all of the days when weights were recorded, KO mi ce were consistently heavier than WT counterparts. Discussion Previous evidence has shown that gastric filling is not the on ly peripheral signal that contributes to decreasing meal size during a meal (reviewed in Davis, 1999). Exogenous responses to peptides that are nor mally released after stomach filling were used as the main indicator of functioning sa tiety mechanisms in MC4R impaired mice. We discovered results converse to our hypothesis that MC4RKO mice would have diminished responsiveness to BBS and CCK. The doses used were comparable to doses used previously with mice (McLaughlin & Baile, 1981; Ladenheim, et al., 2002; Fan et al., 2004, Chi et al., 2004). As in other studies, decreases in food intake oc curred at a threshold dose of about 2 g/kg CCK (McLaughlin & Baile, 1981; Strohmayer & Sm ith, 1986; Weatherford et al., 1992; Chi et al., 2004). The only other study that has given CCK to MC4RKO had results that differ from ours. Fan et al. (2004) used only one dos e, so one of the aims of this experiment was to add to this line of literature by test ing a range of doses. Contrary to Fan et al. (2004) we found decreases in food intake fo r MC4RKO mice after CCK administration.


77 Their group used an ip injecti on of 3nmol/kg, which is approx imately equivalent to a 3.42 g/kg dose. Our lowest dose, 2 g/kg, which was less than theirs (3.42 g/kg) produced a significant suppression of intake in co mparison to vehicle. Though not explicitly stated, Fan’s group (2004) admi nistered CCK immediately before the test meal, whereas in our experiment we waited five minutes be fore giving mice access to a liquid test meal. A previous report (Antin et al., 1975) tested the effect of different CCK injection—test meal intervals in rats found that CCK ad ministered six minutes before, immediately before and six minutes after feeding onset were the periods where food intake was significantly suppressed. There are some methodological differences; namely, test meal composition, age of mice used and food deprivation period. We used Ensure, a palatable liquid, and Fan’s group used pelleted chow. Previous studies ha ve used comparable doses of CCK in mice and still found suppression of food intake re gardless of the test meal composition. For example, CCK was found to decrease inta ke whether pellets, 20% sucrose, or nutritionally complete liquid diets were used (Garlicki et al., 1990; Weatherford et al., 1992; Chi et al., 2004). Another methodological difference is that our group used mice aged between 14-23 wk old and Fan et al used 9 wk old mi ce. Another study reported age dependent responses to CCK between obese and lean mi ce. When observed at two different ages, obese mice decreased intake after CCK admini stration at 7-8 weeks old and not at 5-6 weeks old (McLaughlin & Baile, 1981), whic h may not be two vast developmentally different stages. It is plausible that a similar phe nomenon could be the case for MC4RKO mice because they show an age depende nt decrease in response to exogenous


78 leptin (Marsh, et al., 1999). So there is some evidence of developmental neuronal changes in regards to energy hom eostasis in these mice, but generalizing the leptin results to our CCK results leaves room for explanati on. Lastly, Fan et al. (2004) had a 16 h fast and our study used a 12h fast. It is un likely that this methodological discrepancy significantly affected results because CCK has been found to decrease food intake across a variety of deprivation leve ls (Mueller & Hsiao, 1979). Two issues that limit the interpretation of the results following repeated peptide administration are tolerance and decreased intake due to malaise. Though mice received three different doses of CCK, continuous admi nistration of CCK (using a s.c. minipump) in the past has shown that tolerance does not develop (Hsi ao & Wang, 1983). CCK-8 has a half life of 17 min in rat pl asma (Koulischer et al., 1982) an d thus should have had little carryover effects in the mice tested in the current experiment. CCK-8s administration in rats did not produce a taste aver sion when directly compared to LiCl (West et al., 1987). Our lab has found (unpublished results) that MC4RKO mice are capable of forming a taste aversion to LiCl and thus would be capable to form a taste aversion to another substance. MC4R HETs and KOs were both the most responsive to BBS at the highest dose given. In normal mice, food intake was d ecreased after BBS doses that range from 4 g/kg to 51.2 g/kg (McLaughlin & Baile, 1981; Ladenhe im et al., 2002). In rats, all doses of BBS tested were found to be less potent on a molar basis than CCK (Gibbs, 1985; Garlicki et al., 1990), so we see a simila r effect in this experiment. Taylor and Garcia (1985) found genotype dependent sensit ivity in response to BBS. Their obese mice responded at their 3 nmol/kg dose, but their lean mice did not respond until 27


79 nmol/kg. Our WTs showed no differences from vehicle at all 3 doses suggesting that our lean littermates needed a substantially highe r dose to exhibit a beha vioral effect and a physiological effect of triggered CCK release. The issue at hand with interpreting the eff ects of BBS is tied to the doses chosen and the receptors at which the peptide binds. BBS binds to a variety of areas and seems to be working via an interaction of both BB1 and BB2 receptors (Ladenheim et al., 1996; Moran & Ladenheim, 1998). In the peripher y, the peptide binds to stomach muscle, various layers of the intestines and there is a high affinity receptor type in the pancreas for BBS, which is most likely a GRP receptor. Since we saw an effect at the highest dose for MC4R HET and KO mice, it can be stated that BB1 and BB2 receptors are functional but a further characterization of receptor location, density and responsivity would be beneficial in gathering more conclusi ons about satiety in these mice. We studied short-term food intake of MC4RKO mice following exogenous satiety signals and found that MC4RKO mice re spond to high doses of BBS and CCK. In interpreting CCK results, we ha ve yet to answer if the dos es we used fell within a relevant physiological range. It is possible that effects were seen because the system was flooded with CCK and therefore a response was seen. After a 10% food restriction alone, normal rats show increased CCK plas ma levels (Chowdhury & Rayford, 2001); MC4RKO mice could have experienced a similar physiological profile after their fast. Documenting the endogenous levels of CCK would more definitively answer whether MC4RKO mice are overeating due to fau lty satiety signals. Due to the fact that BBS binds to four different types of receptors and is most functionally active at two of them it is difficult to pinpoint where the activ ation was most biological ly relevant in the


80 MC4R HET and KO mice. In conclusion, the main objective was to learn if MC4RKO mice are hyperphagic because of faulty post in gestive feedback. The most remarkable result was that the HET and KO mice were actually more responsive than the WTs to CCK and BBS administration. This experiment adds to the litera ture the finding that exogenous satiety peptides can be operative at least at normal levels in MC4R impaired mice.


81 CHAPTER 6 GENERAL DISCUSSION Summary of Results In Chapter 2 , our major aim was to describe daily feeding patterns of MC4RKO and WT mice using Collier’s closed economy foraging paradigm . This protocol is not commonly used in mice and our results were generally similar to those found in other species, namely that as the procurement cost increased, meal size increased while meal frequency decreased. The differences between the feeding patterns were not statistically different between WT and MC4RKO mice. The foraging experiment suggests that melanocortin type 4 receptors are not essen tial for planning meals when a mouse has to work for food. The experiment also shows that the different genotypes have similar adaptive mechanisms to alter procurement and consumption of meals in a closed economy. These adaptive mechanisms did not promote the hyperphagia seen normally in MC4RKO mice in a free feeding situation. While WT mice maintained food intake in the foraging situation similar to free feedi ng, and thus maintained body weight, MC4RKO mice ate less under foraging than free situati ons and progressively lo st weight while in the test chambers without any trend to incr ease food intake. However, they regained weight when given free food in their home cage, leading to a weight cy cling profile. It is unclear why weight loss by the MC4RKO mi ce was not accompanied by increased food intake. One possibility is that MC4RKO mi ce have a generally decreased motivation to


82 work for food, therefore a second study was carried out using a closed economy paradigm with a different schedule of reinforcement, a progressive ratio (PR). Using a PR schedule, and in contrast to the foregoing result, we found that MC4RKO mice worked and consumed more food than WT mice; HET mice showed intermediate performance. During traini ng, when all mice were on a FR 3 schedule, MC4RKO mice were hyperphagic. In the PR 3-min reset condition, WTs exhibited a larger number of small feeding bouts and the lowest breakpoints. An interpretation of this pattern is that WTs avoided large amount s of pressing in orde r to have numerous, cheap meals. The same profile was seen under the 20 min reset for WT mice. MC4RKO mice had high breakpoints, relative to other mice, for both reset criteria. MC4RKO mice did initiate fewer feeding bout s but lever pressed significantl y more than other mice in both reset conditions. Interpretation of the MC4RKO profile suggests that MC4RKO mice were willing to sustain a cost in excess of 100 presses for a pellet in the 20 min reset condition. With the increased work required under the PR sche dule, we saw a result similar to MC4RKO performance under the FR schedule: feeding bout initiation decreased. As a result of sustained hyperphagia in the operant condition, the weight cy cling seen in MC4RKOs in the procurement experiment did not occur in the PR experiment. In fact, MC4RKO mice weighed slightly more when they were in the chambers which correspond to the hyperphagia they exhibited. A remaining issu e is what peripheral changes, such as insulin levels, and neural changes, such as regulation of eating related peptides, contributed to the MC4RKO’s choice of one st rategy over another. In order to complete


83 a thorough characterization of m eal characteristics we chose to investigate individual meals and whether these differed between the three genotypes. Satiety can be defined as the sensation that prolongs intermeal intervals. The level of satiety is often dictated by the size of the previous meal (Blundell et al., 1989). Previous reports have indicated MC4RKO mice are capable of some degree of metabolic regulation (Chen et al., 2000; Ste Marie et al., 2000; Alba rado et al., 2004) and our findings support these results. After fasts of differing dur ations, all of the genotypes consumed comparable amounts of liquid diet. Next, the ability to compensate intake after an oral preload was asse ssed. Most of the intake afte r said fasts happened in the first 15 minutes of the test session, illustrati ng again that all genotypes sensed deprivation appropriately. The preload significantly reduced meal si ze: KOs, HETs and WTs drank less after receiving a preload than no pr eload. Further, the reduction in intake was greater when the test meal was presented 30 min after the preload than when presented 60 min after the preload. This suggests that the short term sa tiety effects may have started to dissipate after 60 min. Because of uneven or small numbers of males and/or females in each genotype, we cannot make a conclusive statem ent concerning the effects of gender. Thus, the genetic absence of a functional MC4R had no effect on intake after a fast or on the sensing of and responding to a nutritive gastric load. Our next aim was to assess satiati on mechanisms after a test meal upon administration of exogenous, putative satiati on inducing peptides CCK and BBS. These are both peptide hormones that when released result in a sensation of fullness in humans and reduce food intake in animals (e.g., West et al., 1984; Li ddle et al., 1985).


84 Endogenous CCK release delays gastric empt ying and promotes meal termination (Shilabeer & Davison, 1987). Exogneous BBS acts similarly to endogenous GRP and mediates the release of CCK (Gibbs et al., 1979; reviewed in Yamada et al., 2002). In Chapter 5 , a 12h fast was followed by CCK, BBS or vehicle administration then a test meal. The doses used were based on result s from previous studies done with mice (e.g., McLaughlin & Baile, 1981). We know of only one published study (Fan et al., 2004) that has examined the behavioral response to CCK in MC4RKO mice. They found that after a 16hr fast, an ip injection of CCK-8s did not significantly s uppress food intake at four different time points (30, 60, 120, 180 min post injection). Our result was very different from that of Fan et al (2004) in that WT mice were actu ally more responsive to CCK and BBS than WT or HETs. This discrepancy could be due to methodological differences. Fan and colleagues (2004) used a singl e dose of CCK (3 nmol/kg 3.42 g/kg) , a single test meal measurement and between subjects design while we used a within subjects design with three doses and multiple test meal measurements. Additionally, after Fan et al.’s (2004) single test meal of solid food, MC 4RKO mice that received CCK ate ~0.1 g less than C57 controls that received saline; we saw a more robust di fference between our KO and WT intake. Our WT mean after vehicl e was ~1.8 ml and the KO mean intake after the 8 g/kg dose of CCK was ~.05 ml. In our ha nds all three doses of CCK tested decreased intake, relative to vehicle, in MC4RKO mice. A robust finding which is a novel contribution to the field is the marked decrease of Ensure intake after an 8 g/kg dose of BBS by MC4R HET and KO mice.


85 Effects of a Fixed vs. Pr ogressive Ratio Schedule Different schedules of reinforcement ar e commonly used in food or drug selfadministration literature. The various schedul es whether used separately or tandem can identify nuances of behavior. Animals learn fixed ratio schedules ea sily but these do not provide a sensitive measure of changes in mo tivation. Progressive ratio schedules are thought to be more sensitive measures of motivation because they reveal the maximum price an individual is willing to pay for a fixed reinforcer; this is manifested in the breaking point (Richardson & R oberts, 1996). Researchers have been able to discover the internal and external asp ects of food intake through the use of FR and PR schedules (Johnson et al., 1993; Zhang et al., 2003). We have previously used the FR schedule to describe meal patterns in another mouse model of obesity (Vaughan & Rowland, 2003). We chose to use the MC4RKO model due to its clinical relevance, and because the meal patterns underlying the robust hyperphagia (Huszar et al ., 1997; Mackenzie, 2005) have not been reported. The differential eating profiles of MC4RKO mice seen in FR vs. PR schedules was surprising. An analogous finding in an obese mouse model was reported by Low et al (2003). Their mouse model of -endorphin deficiency exhibi ts increased white adipose tissue and a higher propensity to gain we ight after being on a high fat diet. The endorphin peptide is tied to the positive reinfo rcing quality of operant behavior, including working for food reinforcement. In an operant task, -endorphin knockouts performed the same as wild type mice under a FR5 traini ng schedule but ate less than WTs under their PR3 schedule (Low et al., 2003). Our MC4R KO mice showed similar behavior, but with an increase of food intake under our PR1 schedul e. In comparison to Low et al.’s (2003) study, one could hypothesize that -endorphin signaling is inta ct in MC4RKO mice and


86 that other mouse models can be subjected to different food intake behavior according to the schedule of reinforcement imposed. The fact that MC4RKOs showed hyperphagi a using PR schedules but not using FR schedules could be due to the faster rate of reward seen under the PR schedule. The mice studied in Chapter 3 engaged in many small, less cos tly meals when given the option. The two lever contingency in the foraging expe riment adds time and effort to the main operant. In Chapter 2 , the procurement cost that produ ced the largest shift in meal strategy/pattern was 480 bar pr esses. While conducting th e procurement cost mice are ready to initiate a meal and essentia lly hungry because during the PFR 480-CFR 10 schedule, mice engaged in less frequent meals. The fact that the mice were willing to complete the procurement cost represents a breaking point of sorts, because MC4RKO mice were not willing to go past that to rece ive a food reward. In the PR protocol during procurement, mice only had a contingency on one lever therefore allowing concomitant consumption. WTs had small, cheap meals s uggesting that when allowed to choose a procurement cost when partially satiated they stay with pellets less than or equal to ~17 presses each. Few studies have employed a 24-hr operant paradigm to study food intake patterns. Johnson et al. (1993) used a two-lever continge ncy with a PFR on the left lever and a PR on the right lever. The PFR and the stepwise increment of the PR were manipulated. It was shown that the initial PFR dictated the s ubsequent meal size; the higher the PFR, the smaller the meal. Utilizing an exponentia l progression, instead of an arithmetic progression, with a more steep PR increment than in our PR experiment may not have promoted hyperphagia in MC4R HET and KO mice. An exponential or logarithmic


87 increment would eliminate the option for cheap meals and may have better assessed true motivation to feed. Sclafani and Ackroff (2003) compared PR and FR performance for a sucrose reward with licks as the operant. Rats liv ed in cages for 23-hr/day, received food and water ad libitum but had the option to lick for increasing concentrations of a sucrose solution across separate daily sessions. Th e FR contingency was 20 licks (FR 20) for .08 ml of sucrose. The PR increment was an increase of one only after every second reward (PR-.05). Under an FR 20 schedule, rats in creased sucrose intake in grams until the highest concentration tested, 32%. Under the PR-.05 schedule, rats steadily increased intake in grams as the concentration increased. The same trends in intake were reflected in number of licks emitted per day. Howeve r, the overall amount drank was lower in the PR group vs. the FR group. The authors postu late that the reason could be due to minimized postingestive satiation that resu lted from the timing constraints imposed by the PR structure (Sclafani and Ackroff, 2003). Since timing affected intake in rats, it is possible it also played a part in MC4R KO and HET intake. In the foraging experiment, it appears the 10 min lapse was sufficient to impose discrete meals when our mice were faced with completing a two-lever contingency. A 20 min reset was more succ essful in preventing ‘snacking’ throughout the day but the time and effort spent on the one lever contingency was considerably less than the procurement experiment. To be tter elucidate the perfo rmance of mice under different reinforcement schedules, the graphs below (Figure 6-1) include a comparison of the FR where the most food intake occu rred for MC4RKO mice and the PR1 schedule with the 20 min reset criterion.


88 The cumulative responses emitted by the different genotypes under the different schedules of reinforcement show that ove rall effort was highe r during the FR 15-5 schedule in comparison to the PR schedul e (see Figure 6-1). In Figure 6-1A, the MC4RKO mouse emitted 2592 LL presses and rece ived 6.4 g of food. In Figure 6-1B, the KO mouse shown emitted 3059 right lever (consumption) presses and 97 left (procurement) lever presses during this par ticular daily session. KO10 received 4.9 g of food. Under the PR1, the WT mouse showed here emitted 729 presses resulting in 3.4 g of food. The WT mouse in the FR 15-5 schedule emitted 1015 right lever presses and 152 procurement lever presses resulting in 3.7 g of food. Results show that the average total price per pellet was higher in the proc urement protocol; therefore the PR schedule Time (min) 010020030040050060070080090010001100120013001400Bins (5 pellets / bin) -10 0 10 20 30 40 50 60 70 KO3; 0=3:19 p.m. WT1; 0=3:54 p.m. HET3; 0=2:58 p.m. lights off A Figure 6-1. A comparison of PR (Graph A) and FR (Graph B) cumulative responses. Responses across a daily session under the PR 1 – 20-min reset condition is depicted on the left. Responses across daily session under the FR 15-5 schedule are shown on the right. Each da ily session is the last day of that particular schedule may be a more accurate test for instrument al performance for food in KO mice because it allows for the hyperphagic phenotype seen in ad libitum feeding. BTime(min) 010020030040050060070080090010001100120013001400Bins (5 pellets/bin) -10 0 10 20 30 40 50 60 70 WT2; 0=2:10p.m. KO10; 0=1:24 p.m. lights off


89 During the FR schedules KO and WT mice performed similarly, the demand for the reinforcer remained constant though the indivi dual daily sessions have slightly different patterns. The PR schedules re sulted in more variable performance during daily sessions for the MC4RKO mice than the WT mice. In the representative day shown in Figure 61A the slope of the MC4RKO perf ormance is more steep than wild types. The individual meals across the daily session show that the value of the reinforcer varies more so throughout the day for MC4RKOs compared to WTs. MC4RKOs show many peaks in eating throughout the day whereas the WTs have more discre te meals. An in depth analysis of each day in all of the schedules of reinforcement would give a more complete assessment of whether KO demand for food is more prone to change than WT mice. Real time physiological and neural measurem ents would be a more direct way of deciphering satiety feedback mechanisms in operant protocols in both normal and genetically altered animals. Others have conducted in vivo microdialysis measurements in mice during behavioral responses (e.g., Kehr et al., 2001), so it w ould be plausible to collect real time correlates during feeding. The long-term studies done with these mice have described how the environment can infl uence meal patterns but further study into what dictates how these animals pl an meal initiation is necessary. Post Ingestive Feedback The techniques used in this dissertation to measure satiation and satiety have been used extensively in rodents. In the prel oad literature, research ers have manipulated various parameters such as volume, calorie or macronutrient content and route of delivery of a preload (Seeley et al., 1994; Warwick & Weingarten, 1994). The general results have been that the length of the fast, prev ious experience with the test meal and

PAGE 100

90 palatability of the preload can affect the satiating potency of a stimulus (i.e. a test meal) (Smith, 1998). We chose to give a preload by the natura l route (orally) and manipulate load and timing. In a study by another group, MC4RKO mice were given intraduodenal infusions of fat, glucose or peptone of increasing con centrations (Azzara et al., 2005). In their study, nutrients of various caloric contents were given by means of intraduodenal catheters. They found that MC4RKO mice were less effective in decreasing the liquid test meal than WT mice whereas we found that MC4RKO mice were equally as sensitive as WT in decreasing the liquid test meal. One of the reasons cited in Chapter 4 as the basis for the disparity between our results and those of Azzara et al. (2005) was the different routes of administration of the preload. Differing eff ects from duodenal and oral administration has been observed in other studies. For example, a large reduction in intake occurred after an intraduodenal infusion in rats rather than after an intragastric infusion (Snowdon, 1975). Activation of the dorsal vagal complex, as assessed by Fos immunoreactivity, is also different after ga stric distention vs. after an intraduodenal infusion (Berthoud et al., 2001). A more detailed comparison of test meal intake after duodenal and oral preloads in th e same mice would be beneficial in clarifying this issue. One possibility for MC4RKO mice being less sensitive to duodenal infusions of fat could be faulty vagal afferent fiber feedbac k. There are specific fibers that respond to long-chain fats and short-chain fats and gly cerol. It has been shown that intestinal innervation by the vagus is necessary for fa t induced satiation (reviewed in Greenberg, 1998). MC4RKO mice will overeat on a moderate fat diet (Butler et al., 2001), which

PAGE 101

91 suggests that if this differential vagal ac tivation is the case it may play a part in MC4RKO hyperphagia. The question of overall vagal fiber sensit ivity was the impetus to test CCK and BBS responses in MC4RKO mice. CCK requires gastric branches of the vagus for their effects to occur while BBS requires both vaga l and spinal visceral input (Gibbs, 1985). CCK1 receptors are found on all branches of the vagus below the diaphragm (reviewed in Moran & Ladenheim, 1998). MC4R HET a nd KO mice decreased food intake to exogenous CCK, thus the receptors on those affe rent fibers appear to have functional crosstalk to the NTS and then upstream to th e hypothalamus. In addition to sufficient vagal signaling, MC4RKO mice may have res ponded to CCK due to their high leptin levels. It has been demonstrated that rats wi th high leptin levels were more sensitive to the meal-reducing signal of CCK (Matson et al., 1997). A more extensive study employing CCK receptor specific antagonists and/or capsaicin to different levels of vagal afferent fibers would aid in teasing out how visceral feedback is modified by CCK in these mice. MC4R KO and HETs were more sensitive to BBS than WT mice under the dose range used in our experiment. It may be due to faulty spinal viscer al input (Gibbs, 1985) but further study of anatomical substrates is necessary before making a definitive statement. It has been shown that CCK-8 is more effective than BBS in suppressing liquid food intake but act comparably in s uppressing solid food intake (Gibbs et al., 1979). Therefore, replica ting the experiment in Chapter 5 using solid food would contribute to explaining the role of BBS in inhibiting food intake. Different routes of administration have been shown to affect satia ting potency of the pept ide. Injecting ip

PAGE 102

92 has larger behavioral effects th an via the hepatic-portal vein or the inferior vena cava. Our ip injections may have been the best route to administer CCK but perhaps not BBS (reviewed in Smith & Gibbs, 1998). Known and Unknown Mechanisms of Feeding in MC4RKO Mice Feeding behavior is a coordinated even t that involves numerous peptides and numerous resulting structural changes. By studying long a nd short term meal patterns our aim was to decipher neur al and peripheral mechanisms that promoted hyperphagia in MC4RKO mice. Normal signaling of anorexig enic and orexigenic peptides have been reported and we found CCK at certain doses successfully reduced food intake in MC4RKO mice. CCK exerts action via MC4R dependent effectors (e.g., Sutton et al., 2005), so it appears there has to be co mpensation done by the MC4RKOs somewhere before or after CCK activation. Without doing more direct analyses, it is difficult to pinpoint what compensatory mechanisms are at work in HET and KO mice. The NPY signaling system has been descri bed in KO mice. NPY gene expression was not found to be affected by MC4R de letion and NPY induced feeding occurs normally in MC4RKO mice (Kesterson et al., 1997; Marsh et al., 1999). The leptin signaling system has also been described and feeding abnormalities in MC4RKO mice were reported to not be due to leptin fee dback dysregulation (Marsh et al., 1999; Weide et al., 2003). The CCK signaling system involves the peptide and its receptors found both centrally and peripherally. In rat brain, CCK1 receptors are found in motor nuclei of the brainstem and CCK2 receptors, which have the highest affinity for sulfated CCK-8, are found in sensory nuclei (Honda et al., 1993; reviewed in Little et al., 2005). Though satiety is thought to be mediated via low affinity CCK1 receptors in the pylorus, receptors

PAGE 103

93 in other areas contribute to th e transduction of signals into behavior (Weatherford et al., 1992; reviewed in Smith & Gibbs, 1998). KO and HET mice were sensitive to CCK-8 therefore their CCK2 receptors appear to be functional at physiologica lly relevant levels. CCK induc es excitatory potentials in POMC neurons of the NTS (Appleyard et al ., 2005) exhibiting some evidence of the connection between CCK efficacy and the mela nocortin system. The use of antagonists and intracellular techniques has shown the contribution of brainstem MC4Rs to energy homeostasis (Williams et al., 2000, 2002, 2003; Sutton et al., 2004, 2005). These findings lead to a hypothesis that MC4RKO mi ce would have impaired response to vagal afferent feedback. Our results challenge that hypothesis and/or show that compensatory neurotransmitter mechanisms are the plausi ble means by which MC4RKO mice are able to significantly decrease f ood intake after a preload and CCK administration. Noradrenergic neurons in the caudal NT S have recently been shown to be instrumental in relaying CCK induced a norexia to the hypothalamus (Rinaman, 2003). Satiation via gastric distention and CCK are relayed to the NTS, signaling release of glutamate which acts at NMDA receptors on second order NTS neurons (Berthoud et al., 2001). Neurons in the NTS project to the hypot halamus to modify anorexigenic signals and receive information from visceral moto r areas in the medial prefrontal cortex (reviewed in Berthoud, 2002). CCK-8 administration could have affected intake of MC4R KO and HET mice via its interactions with dopamine transmission (Kombian et al ., 2004). In addition to the peripheral effects, CCK could ha ve acted centrally to influence the “liking” of the liquid test meal by acting on CCK2 receptors in the nucleus accumbens (Honda et al., 1993).

PAGE 104

94 CCK has been shown to excite GABA receptors in the nucleus accumbens in vitro which in turn depressed dopamine transmission (Kom bian et al., 2004). However, when given in conjunction with D-amphetamine, which in creases dopamine release, CCK increases responding in an operant task for sucrose re inforcement (Phillips et al., 1993). Though the specifics are still being debated, the circuitry exists for CCK and dopamine to have some altered interaction in HET and KO mice and not WT mice. Conclusion In behavioral economics, the price of the reinforcement in relation to how much is consumed comprises the concept of demand (for review see Hursh, 1991). Due to the hyperphagic MC4RKO phenotype, we hypothesized that th ere would be genotype dependent changes in consumption dependent of the price of the commodity. The FR and PR schedules we used produced differing profiles. Others have cited that an advantage to using a PR schedule vs. a FR sche dule would be to get an idea of a range of prices within a single session. However, it is common to use both sche dules to get a total picture of behavior; the FR to establish a baseline and the PR to measure motivation (Richardson & Roberts, 1996; Stafford et al., 1998). Both of our operant tasks were to the first to show how MC4R KO mice can auto regulate food intake in a simulated foraging and depleting patch environments. Ac cording to Collier (1987), effort and how it is imposed affects performan ce; the results gathered in Chapters 1 and 2 are testament to this point. The measures of satiety and satiation resu lted in no aberrant results for the MC4R HET and KO mice compared with WT. MC 4RKO mice exhibit hyperphagia and show a propensity to gain weight on a high fat di et. Others have shown that decreased metabolism is not a major contributor to the increased weight gain (Huszar et al., 1997;

PAGE 105

95 Chen et al., 2000b; Weide et al., 2003). The individual meals studied showed no differences between genotypes. MC4RKO mi ce can sense fullness, both mechanically and physiologically, and adjust meals accordingly. Clearly the environment plays a large role in how this genetic abnormality influe nces an overall profile of eating. To date 58 mutations of the human MC4R have been reported but not all obese individuals have these mutations (Vaisse et al., 2000; Macken zie, 2006). Agonists of the receptor would be beneficial drug therapies but environmental changes may be the first step in modulating the accompanying MC4R phenotype. Nationally ~77% of the population report not eating a diet rich in fr uits and vegetables suggesting that self selected changes could reverse unhealthy eati ng habits (CDCP, 2002). Further study with MC4R KO and HET mice will be useful in understanding humans with similar mutations and provide methods with which to treat them.

PAGE 106

96 LIST OF REFERENCES Albarado, D.C., McClaine, J., Stephens, J.M., Mynatt, R.L., Ye, J., Bannon, A.W., Richards, W.G., & Butler, A.A. (2004). Im paired coordination of nutrient intake and substrate oxidation in melanoc ortin-4 receptor knockout mice. Endocrinology, 145 , 243-252. Anand, B.K., & Brobeck, J.R. (1951). Loca lization of a "feedi ng center" in the hypothalamus of the rat. Proceedings for the Society for Experimental Biology and Medicine, 77 (2), 323-324. Anastasi, A., Erspamer, V., & Bucci, M. (1971) . Isolation and structure of bombesin and alytesin, 2 analogous active peptides from the skin of the European amphibians Bombina and Alytes. Experientia, 27 (2), 166-167. Anliker, J., & Mayer, J. (1956). An opera nt conditioning technique for studying feedingfasting patterns in normal and obese mice. Journal of Applied Physiology, 8 , 667670. Antin, J., Gibbs., J., Holt, J., Young, R., & Sm ith, G.P. (1975). Cholecystokinin elicits the complete sequence of satiety in rats. Journal of Comparative and Physiological Psychology, 89 , 784-790. Appleyard, S.M., Bailey, T.W., Doyle, M.W., Jin, Y.H., Smart, J.L., Low, M.J. et al. (2005). Proopiomelanocortin neurons in nuc leus tractus solitar ius are activated by visceral afferents: Regulation by cholecystokinin and opioids. Journal of Neuroscience, 25 (14), 3578-3585. Azzara, A.V., Sokolnicki, J.P., & Schwartz , G.J. (2002). Central melanocortin receptor agonist reduces spontaneous and scheduled meal size but does not augment duodenal preload-induced feeding inhibition. Physiology & Behavior, 77 (2-3), 411416. Azzara, A.V., Schuss, B., Hong, S., Chua, S. C., & Schwartz G.J. (2005). Melanocortin 4 receptor (MC4R) knockout mice have increased food intake and meal size, and decreased sensitivity to post-oral nutrient stimulation [Abstract]. Appetite, 44 (3), 332. Bachmanov, A.A., Reed, D.R., Beauchamp, G. K., & Tordoff, M.G. (2002). Food intake, water intake, and drinking spout si de preference of 28 mouse strains. Behavioral Genetics, 32 (6), 435-443.

PAGE 107

97 Bagnol, D., Lu, X.Y., Kaelin, C.B., Day, H.E ., Ollmann, M., Gantz, I., et al., (1999). Anatomy of an endogenous antagonist: Relationship between Agouti-related protein and proopiomel anocortin in brain. Journal of Neuroscience, 19 (18; RC26), 1-7. Bergen, H.T., Mizuno, T., Taylor, J., & Mobbs, C.V. (1999). Resistance to diet-induced obesity is associated with increase d proopiomelanocortin mRNA and decreased neuropeptide Y mRNA in the hypothalamus. Brain Research, 851 (1-2), 198-203. Berthoud, H.R. (2002). Multiple neural system s controlling food intake and body weight. Neuroscience and Biobehavioral Reviews, 26 (4), 393-428. Berthoud, H., Earle, T., Zheng, H., Patterson, L.M., & Phifer, C. (2001). Food-related gastrointestinal signals activate caudal brainstem neurons ex pressing both NMDA and AMPA receptors. Brain Research, 915 (2) , 143. Blundell, J.E., Hill, A.J., & Lawton, C.L. ( 1989). Neurochemical factors in normal and abnormal eating. In R. Shepherd (Ed.), Handbook of the Psychophysiology of Eating (pp. 85-112). New York: Wiley. Bolles, R.C. (1980). Some functionalistic t houghts about regulation. In F.M. Toates & T.R. Halliday (Eds.), Analysis of motivational processes (pp 63-75). Burlington, MA:Academic Press. Booth, D.A. (1972). Conditioned satiety in the rat. Journal of Comparitive and Physiological Psychology, 81 (3), 457-471. Boucher, N., Lanouette, C. M., Larose, M., Perusse, L., Bouchard, C., & Chagnon, Y.C. (2002). A +2138InsCAGACC polymorphism of th e melanocortin receptor 3 gene is associated in human with fat level a nd partitioning in interaction with body corpulence. Molecular Medicine, 8 (3), 158-165. Branson, R., Potoczna, N., Kral, J.G., Lentes , K.U., Hoehe, M.R., & Horber, F.F. (2003). Binge eating as a major phenotype of me lanocortin 4 receptor gene mutations. New England Journal of Medicine, 348 (12), 1096-1103. Bray, G.A., & York, D.A. (1979) Hypothalami c and genetic obesity in experimental animals: an autonomic and endocrine hypothesis. Physiological Reviews, 59 (3), 719-809. Brobeck, J.R. (1955). Neural regulation of food intake. Annals of the New York Academy of Sciences, 63 (1), 44-55. Brning, J.C., Gautam, D., Burks, D.J., Gille tte, J., Schubert, M., Orban, P.C., et al., (2000) Role of brain insulin receptor in control of body weight and reproduction. Nature, 289 (5487), 2122-2125.

PAGE 108

98 Bultman, S.J., Michaud, E.J., & Woychik, R.P. (1992). Molecular characterization of the mouse agouti locus. Cell, 71 (7), 1195-1204. Bush, M., & Leathwood, P.D. (1975). Effect of different regimens of early malnutrition on behavioural development and adult av oidance learning in Swiss white mice. The British Journal of Nutrition, 33 (3), 373-385. Butler, A.A., & Cone, R.D. (2003) Knockout studies defining different roles for melanocortin receptors in energy homeostasis. Annals of the New York Academy of Sciences, 994 , 240-245. Butler, A.A., Marks, D.L., Fan, W., Kuhn, C. M., Bartolome, M., & Cone, R.D. (2001), Melanocortin-4 receptor is required for acute homeostatic responses to increased dietary fat. Nature Neuroscience, 4 (6) , 605-611. Campfield L.A., Smith F.J., Guisez Y., Devos R., & Burn, P. (1995). Recombinant mouse OB protein: Evidence for a periphe ral signal linking ad iposity and central neural networks. Science, 269 (5223), 546-549. Canon, W.B. (1929). Organization for physiological homeostasis. Physiological Reviews, 9, 399-431. Castonguay, T.W., Upton, D.E., Leung, P.M.B., & Stern, J.S. (1982). Meal patterns in the genetically obese Zucker rat: A reexamination. Physiology & Behavior, 28 (5), 911-916. Centers for Disease Control & Prevention (C DCP). (2000). Available from Center for Disease Control & Prevention Web site, http://apps.nccd.cdc .gov/brfss/Trends/ TrendData.asp Centers for Disease Control & Prevention (CDCP). (2002). Behavi oral Risk Factor Surveillance System Trends data. Availa ble from Center for Disease Control & Prevention Web site, Challis, B.G., Coll, A.P., Yeo, G.S., Pinnock, S. B., Dickson, S.L., Thresher, R.R., et al. (2004). Mice lacking pro-opi omelanocortin are sensitiv e to high-fat feeding but respond normally to the acute anorectic effects of peptide-YY3-36. Proceedings of the National Academy of Sciences USA. 101 (13), 4695-4700. Chen, A.S., Marsh, D.J., Trumbauer M.E., Fr azier, E.G., Guan, X., Yu, H. et al., (2000a). Inactivation of the mouse melanocortin-3 receptor results in increased fat mass and reduced lean body mass. Nature Genetics, 26 (1), 97-102. Chen, A.S., Metzger, J.M., Trumbauer, M. E., Guan, X., Yu, H., Frazier, E.G. et al. (2000b). Role of the melanocortin-4 receptor in metabolic rate and food intake in mice. Transgenic Research, 9 (2) , 145-154.

PAGE 109

99 Chen, P., Williams, S.M., Grove, K.L., & Smith. M.S. (2004). Melanocortin 4 receptormediated hyperphagia and activation of neuropeptide Y expression in the dorsomedial hypothalamus during lactation. The Journal of Neuroscience, 24 (22) , 5091-5100. Chi, M.M., Fan, G., & Fox, E.A. (2004). Increased short-term food satiation and sensitivity to cholecystokinin in neurotrophin-4 knock-in mice. American Journal of Physiolgy: Regulatory, Integrative and Comparative Physiology, 287 (5):R1044R1053. Chowdhury, P., & Rayford, P.L. (2001). E ffect of food rest riction on plasma cholecystokinin levels and exocrine pancreatic function in rats. Annals of Clinical and Lab Science, 31 (4):376-82. Collier, G. (1987). Operant methodologies for studying feeding and drinking. In N.E Rowland & F.M. Toates (Eds.), Feeding and drinking (pp. 37-76). New York, NY: Elsevier Science Publishers B.V. Collier, G.H., Johnson, D.F., Hill, W.L., &. Kaufman, L.W. (1986). The economics of the law of effect. Journal of Experimental A nalysis of Behavior, 46, 113-136. Collins, P.J., Houghton, L.A., Read, N.W., Horo witz, M., Chatterton, B.E., & Heddle, R. (1991). Role of the proximal and distal stomach in mixed solid and liquid meal emptying. Gut, 32 (6), 615-9. Cone, R.D. (1999). The central melanocorti n system and energy homeostasis. Trends in Endocrinology and Metabolism, 10 (6), 211-216. Corwin, R.L., Gibbs, J., & Smith, G.P. (1991). Increased food intake after type A but not type B cholecystokinin receptor blockade. Physiology & Behavior, 50 (1), 255-258. Cowley, M.A., Smart, J.L., Rubinstein, M., Ce rdan, M.G., Diano, S., Horvath, T.L., et al. (2001). Leptin activates anorexigenic POMC neurons through a neural network in the arcuate nucleus. Nature, 411 (6836), 480-484. Davis, J.D. (1999). Some new developments in the understanding of oropharyngeal and postingestional contro ls of meal size. Nutrition, 15 (1), 32-39. Davis, J.D., & Campbell, C.S. (1973). Peripheral control of meal size in the rat: Effect of sham feeding on meal size and drinking rate. Journal of Comparative and Physiological Psychology, 83 (3), 379-387. de Graaf, C., Blom, W.A.M., Smeets, P.A.M ., Stafleu, A., & Hendriks, H.F.J. (2004). Biomarkers of satiation and satiety. The American Journal of Nutrition, 79 (6), 946-961.

PAGE 110

100 El-Ghundi, M., O'Dowd, B.F., Erclik, M., &G eorge, S.R. (2003). Attenuation of sucrose reinforcement in dopamine D1 receptor deficient mice. European Journal of Neuroscience, 17 (4), 851-862. Ellacott, K.L.J., & Cone, R.D. (2004). Th e central melanocortin system and the integration of shortand long-term regulators of energy homeostasis. Recent Progress in Hormone Research, 59 , 395-408. Eylam, S., Moore, M., Haskell-Luevano, C., & Spector, A.C. (2005). Melanocortin-4 receptor-null mice display normal affective licking responses to prototypical taste stimuli in a brief-access test. Peptides , 26 (10), 1712-1719. Farooqi, I.S., Keogh, J.M., Yeo, G.S., Lank, E.J., Cheetham, T., & O’Rahilly. S. (2003) Clinical spectrum of obesity and mutations in the melanocortin-4 receptor gene. New England Journal of Medicine, 348, 1096-1103. Fan, W., Boston, B.A., Kester son, R.A., Hruby, V.J., & Cone, R.D. (1997). Role of melanocortinergic neurons in feedin g and the agouti obesity syndrome. Nature, 385, 165-168 . Fan, W., Ellacott, K.L., Halatchev, I.G., Ta kahash,i K., Yu, P., & Cone, R.D. (2004). Cholecystokinin-mediated suppression of feeding involves the brainstem melanocortin system. Nature Neuroscience, 7 (4), 335-336. Findley, J.D. (1958). Preference and sw itching under concurrent scheduling. Journal of the Experimental Analysis of Behavior, 1 , 123-144. Flynn, F.W. (1991). Effects of fourth vent ricle bombesin injection on meal-related parameters and grooming behavior. Peptides, 12 (4), 761-765. Friedman, J.M. (2004). Modern scien ce versus the stigma of obesity. Nature Medicine, 10 (6), 563-569. Friedman, M.I., & Stricker, E.M. (1976). The physiological psychology of hunger: A physiological perspective. Psychological Review, 83 (6), 409-431. Fox, E.A., Phillips, R.J., Baronowsky, E.A., Byerly, M.S., Jones, S., & Powley T.L. (2001). Neurotrophin-4 deficient mice ha ve a loss of vagal intraganglionic mechanoreceptors from the small intestine and a disruption of short-term satiety. The Journal of Neuroscience, 21 (21), 8602-8615. Gannon, K.S., Smith, J.C., Henderson, R., & He ndrick, P. (1992). A system for studying the microstructure of ingestive behavior in mice. Physiology & Behavior, 5, 515521. Gantz, I., & Fong, T.M. (2003). The melanocortin system. American Journal of Physiology: Endocrinology and Metabolism, 284, E468-E474.

PAGE 111

101 Gantz, I., Konda, Y., Tashiro, T., Shimoto, Y ., Miwa, H., Munzert, G., et al. (1993a) Molecular cloning of a novel melanocortin receptor. The Journal of Biological Chemistry, 268 (11) , 8246-8250. Gantz, I., Miwa, H., Konda, Y., Shimoto, Y., Ta shiro, T., Watson, S.J., DelValle, J., et al. (1993b). Molecular cloning, expression, a nd gene localization of a fourth melanocortin receptor. The Journal of Biological Chemistry , 268 (20), 1517415179. Garlicki, J., Konturek, P.K., Majka, J ., Kwiecien, N., & Konturek, S.J. (1990). Cholecystokinin receptors and vagal nerves in control of food intake in rats. The American Journal of Physiology, 258 (1 Pt 1), E40-E45. Gibbs, J. (1985). Effect of bombesin on feeding behavior. Life Sciences, 37 , 147-153. Gibbs, J., Fauser, D.J., Rowe, E.A., Rolls, B.J., Rolls, E.T., & Maddison, S.P. (1979). Bombesin suppresses feeding in rats. Nature, 282 , 208-210. Gilbert, S. (1989). Psychologica l aspects of obesity and its treatment. In R. Shepherd (Ed.), Handbook of the psychophysiology of eating (pp. 301-319). New York: Wiley. Greenberg, D. (1998). Intestinal satiety. In G.P. Smith (Ed.), Satiation: From gut to brain (pp. 40-70). New York: Oxford University Press, Inc. Hadley, M.E., & Haskell-Luevano, C. ( 1999). The proopiomelanocortin system. Annals of the New York Academy of Sciences, 885 , 1-21. Harrold, J.A., Widdowson, P.S., & Williams, G. (1999). Altered energy balance causes selective changes in melanocortin-4 (MC4 -R), but not melanocortin-3 (MC3-R), receptors in specific hypothalamic regions : Further evidence that activation of MC4-R is a physiological inhibitor of feeding. Diabetes, 48 (2), 267-271. Havrankova, J., Roth, J., & Brownstein, M. (1978a). Insulin re ceptors are widely distributed in the central nervous system of the rat. Nature, 272 (5656), 827-829. Havrankova, J., Schmechel, D., Roth, J., & Brownstein, M. (1978b). Identification of insulin in rat brain. Proceedings of the National Academy of Sciences of the U S A., 75 (11), 5737-5741. Hefner, H.E., & Hefner, R.S. (1995). Cond itioned avoidance. In G.M. Klump, R. J. Dooling, R. R. Fay, & W. C. Stebbins (Eds.), Methods in comparative psychoacoustics (pp. 79-93). Switzerland:Bir khauser Verlag Basel. Hetherington, A.W., & Ranson, S.W. (1983). Nutrition Classi cs. The Anatomical Record, Volume 78, 1940: Hypothalamic lesions and adiposity in the rat. Nutrition Reviews, 41 (4), 124-127.

PAGE 112

102 Hodos, W. (1961). Progressive ratio as a measure of reward strength. Science, 134 , 943944. Hodos, W., & Kalman, G. (1963). Effects of increment size and reinforcer volume on progressive ration performance. Journal of the Experi mental Analysis of Behavior, 6 (3), 387-392. Hofman, P.L., Cutfield, W.S., Robinson, E.M., Bergman, R.N., Menon, R.K., Sperling, M.A., et al. (1997). Insulin resistance in short children with intrauterine growth retardation. The Journal of Clinical Endocrinology and Metabolism, 82( 2), 402406. Honda, T., Wada, E., Battey, J.F., & Wank, S.A. (1993). Differential gene expression of CCKA and CCKB receptors in rat brain. Molecular and Cellular Neurosciences, 4 , 143-154. Hsiao, S., & Wang, C.H. (1983). Continuous infusion of cholecystokinin and meal pattern in the rat. Peptides, 4 , 15-17. Hulshof, T., de Graaf, C., & Weststrate, J.A. (1995). Short-term effects of high-fat and low-fat/high-SPE croissants on appetite a nd energy intake at three deprivation periods. Physiology & Behavior, 57 (2), 377-383. Hursh, S.R. (1991). Behavioral economics of drug self-administration and drug abuse policy. Journal of the Experimental Analysis of Behavior, 56 (2), 377-393. Huszar, D., Lynch, C.A., Fairchild-Huntress, V., Dunmore, J.H., Fang, Q., Berkemeier, L.R., et al. (1997). Targeted disruption of the melanocortin-4 receptor results in obesity in mice. Cell, 88 , 131-141. Irani, B.G., Xiang, Z., Moore, M.C., Mande l, R.J., & Haskell-Luevano, C. (2005). Voluntary exercise delays monogenetic obesity and overcomes reproductive dysfunction of the melanocortin -4 receptor knockout mouse. Biochemical and Biophysical Research Communications, 326 (3), 638-644. Jewett, D.C., Cleary, J., Levine, A.S., Schaal , D.W., & Thompson, T. (1995). Effects of neuropeptide Y, insulin, 2-deoxyglucose, and food deprivation on food-motivated behavior. Psychopharmacology, 120 (3), 267-271. Johnson, D., Triblehorn, J., & Collier, G. ( 1993). The effect of patch depletion on meal patterns in rats. Animal Behaviour, 46 , 1, 55-62. Johnson, L. R. (2003). Part V -Gastrointe stinal Physiology: Regulation. In L.R. Johnson (Ed.). Essential medical physiology (pp 465-558). San Diego, CA.: Elsevier.

PAGE 113

103 Kalra, S.P., Dube, M.G., Pu, S., Xu, B., Horv ath, T.L., & Kalra, P.S. (1999). Interacting appetite-regulating pathways in th e hypothalamic regulation of body weight. Endocrine Reviews, 20 (1), 68-100. Kanayama, S., & Liddle, R.A. (1991).Influ ence of food deprivation on intestinal cholecystokinin and somatostatin. Gastroenterology, 100 (4), 909-915. Kas, M.J., Tiesjema, B., van Dijk, G., Garn er, K.M., Barsh, G.S., Brake, O.T., et al. (2004). Induction of brain-region-spe cific forms of obesity by agouti. Journal of Neuroscience, 24 (45), 10176-10181. Kehr, J., Yoshitake, T., Wang, F.H., Wynic k, D., Holmberg, K., Lendahl, U., et al. (2001). Microdialysis in freely moving mi ce: determination of acetylcholine, serotonin and noradrenaline releas e in galanin transgenic mice. Journal of Neuroscience Methods, 109 (1), 71-80. Kelley, A.E., Bakshi, V.P., Haber, S.N., Steininger, T.L., Will, M.J., & Zhang, M. (2002). Opioid modulation of taste he donics within the ventral striatum. Physiology & Behavior, 76 (3), 365-377. Kennedy, G.C. (1953). The role of depot fat in the hypothalamic control of food intake in the rat. Proceedings of the Royal Society of London Biological Series , 140 , 578592. Kennedy, G.C. (1967). Ontogeny of mechanisms controlling food and water intake. In W. Heidel (Exec Ed.) & C.F. Code (Sect Ed.), Handbook of physiology: Sect 6. Alimentary canal (Vol 1, pp. 337-351). Washington, D.C.: American Physiological Society. Kesterson, R.A., Huszar, D., Lynch, C.A., Si merly, R.B., & Cone, R.D. (1997). Induction of neuropeptide Y gene expression in th e dorsal medial hypothalamic nucleus in two models of the agouti obesity syndrome. Molecular Endocrinology, 11 (5), 630637. Koegler, F.H., Schaffhauser, R.O., Mynatt, R.L., York, D.A., & Bray, G.A. (1999). Macronutrient diet intake of the le thal yellow agou ti (Ay/a) mouse. Physiology & Behavior, 67 (5), 809-812. Kombian, S.B., Ananthalakshmi, K.V., Parvathy, S.S., & Matowe, W.C. (2004). Cholecystokinin activates CCKB receptors to excite cells and depress EPSCs in the rat rostral nucleus accumbens in vitro. The Journal of Physiology, 555 (Pt 1), 71-84. Konturek, J.W. (1994). Cholecystokinin in the control of gastric acid and plasma gastrin and somatostatin secretion in healthy s ubjects and duodenal ulcer patients before and after eradication of Helicobacter pylori. Journal of Physiology and Pharmacology, 45 (4 Suppl 1), 3-66.

PAGE 114

104 Koop, I., Kimmich, T., Koop, H., & Arnold, R. (1987). Effect of food deprivation on the function of the intestinal cholecyst okinin-producing cell in the rat. Digestion, 38 (2), 114-123. Koulischer, D., Moroder, L., & Deschodt -Lanckman, M. (1982). Degradation of cholecystokinin octapeptide, related fr agments and analogs by human and rat plasma in vitro. Regulatory Peptides, 4 (3), 127-139. Krude, H., Biebermann, H., Luck, W., Horn, R., Brabant, G., & Gruters, A. (1998). Severe early-onset obesity, adrenal insuffi ciency and red hair pigmentation caused by POMC mutations in humans. Nature Genetics, 19 (2), 155-157. Ladenheim, E.E., Wirth, K.E., & Moran, T.H. (1996). Receptor subtype mediation of feeding suppression by bombesin-like peptides. Pharmacology, Biochemistry, and Behavior 54 (4), 705-711. Ladenheim, E.E., Hampton, L.L., Whitney, A.C., White, W.O., Battey, J.F., & Moran, T.H. (2002). Disruptions in feeding and body weight control in gastrin-releasing peptide receptor deficient mice. The Journal of Endocrinology, 174 (2), 273-281. Le Magnen, J. (1981). The metabolic basis of dual periodicity of feeding in rats. The Behavioral and Brain Sciences, 4 (4), 561-575. Lembertas, A.V., Perusse, L., Chagnon Y.C., Fisler, J.S., Warden, C.H., Purcell-Huynh, D.A., et al., (1997). Identification of an obesity quantitative trait locus on mouse chromosome 2 and evidence of linkage to body fat and insulin on the human homologous region 20q. Journal of Clinical Investigation, 100 (5), 1240-1247. Li, B-H., & Rowland , N.E. (1996). Periphera lly and centrally administered bombesin induce Fos-like immunoreactivity in different brain regions in rats. Regulatory Peptides, 62 (2-3), 167-172. Liddle, R.A. (1994). Regulation of cholecy stokinin synthesis a nd secretion in rat intestine. Journal of Nutrition, 124 (8 Suppl), 1308S-1314S . Liddle, R.A. (1997). Cholecystokin cells. Annual Review of Physiology, 59 , 221-242. Liddle, R.A., Goldfine, I.D., Rosen, M.S., Taplitz, R.A., & Williams, J.A. (1985). Cholecystokinin bio activity in human plasma. Molecular forms, responses to feeding, and relationship to gallbladder contraction. Journal of Clinical Investigation , 75 (4), 1144-1152. Lieverse, R.J., Masclee, A.A., Jansen, J.B ., Lam, W.F., & Lamers, C.B. (1998). Obese women are less sensitive for the satiety effects of bombesin than lean women. European Journal of Clinical Nutrition, 52 (3), 207-212.

PAGE 115

105 Lindblom, J., Schioth, H.B., Watanobe, H., Suda , T., Wikberg, J.E., & Bergstrom, L. (2000). Downregulation of melanocortin re ceptors in brain areas involved in food intake and reward mechanisms in obese (OLETF) rats. Brain Research, 852 (1), 180-185. Little,T.J., Horowitz, M., & Fein le-Bisset, C. (2005). Role of cholecystokinin in appetite control and body weight regulation. Obesity Reviews, 6 (4), 297-306. Low, M.J., Hayward, M.D., Appleyard, S.M., & Rubinstein, M. (2003) State-dependent modulation of feeding behavior by proopi omelanocortin-derived beta-endorphin. Annals of the New York Academy of Sciences , 994 , 192-201. Mackenzie, R.G. (2005) Obesity-associate d mutations in the human melanocortin-4 receptor gene. Peptides , 27(2):395-403 Marks, J.L., Porte, D. Jr., Stahl, W.L., & Baskin, D.G. (1990) Localization of insulin receptor mRNA in rat brain by in situ hybridization. Endocrinology, 127 (6), 32343236. Marsh, D.J., Hollopeter, G. Huszar, D., Laufer , R., Yagaloff, K.A., Fisher S.L., et al. (1999). Response of melanocortin-4 recep tor-deficient mice to anorectic and orexigenic peptides. Nature Genetics, 21, 119-22. Matson, C.A., Wiater, M.F., Kuijper, J.L., & Weigle, D.S. (1997). Synergy between leptin and cholecystokinin (CCK) to control daily caloric intake. Peptides, 18 (8):1275-1278. Mayer, J. (1955). Regulation of energy in take and the body weight: the glucostatic theory and the lipostatic hypothesis. Annals of the New York Academy of Sciences, 63 (1), 15-43. McDonald, T.J., Jornvall, H., Nilsson, G., Va gne, M., Ghatei, M., Bloom, S.R., Mutt, V. (1979).Characterization of a gastrin releasing peptid e from porcine non-antral gastric tissue. Biochemical and Biophysical Research Communications, 90 (1), 227233. McLaughlin, C.L., & Baile, C.A. (1980). Feedin g response of weanling Zucker obese rats to cholecystokinin and bombesin. Physiology & Behavior, 25 (3), 341-346. McLaughlin, C.L., & Baile, C.A. (1981). Obese mice and the satiety effects of cholecystokinin, bombesin and pancreatic polypeptide. Physiology & Behavior, 26 (3), 433-437. Merali, Z., McIntosh, J., & Anisman, H. (1999). Role of bombesin-related peptides in the control of food intake. Neuropeptides, 33 (5), 376-386.

PAGE 116

106 Mokdad, A.H., Ford, E.S., Bowman, B.A., Dietz, W.H., Vinicor, F., Bales, V.S., et al. (2003). Prevalence of obesity, diabetes, and obesity related health risk factors, 2001. Journal of American Me dical Association, 289 , 76-79. Mook, D.G., Culberson, R., Gelbart, R.J ., & McDonald, K. (1983). Oropharyngeal control of ingestion in rats: acquisition of sham-drinking patterns. Behavioral Neuroscience, 97 (4), 574-584. Moran, T.H. (2004). Gut peptides in the control of food intake: 30 years of ideas. Physiology & Behavior, 82 , 175-200. Moran, T.H., & Ladenheim, E.E. (1998). Identi fication of receptor populations mediating the satiating actions of brain and gut peptides. In G.P. Smith (Ed.), Satiation: From gut to brain (pp. 126-163). New York: Oxford University Press, Inc. Morgan, C., Thomas, R.E., & Cone, R.D. (2004) . Melanocortin-5 receptor deficiency promotes defensive behavior in male mice. Hormones and Behavior, 45 (1), 58-63. Mountjoy, K.G., Mortrud, M.T., Low, M.J., Simerly, R.B., & Cone, R.D. (1994). Localization of the melanocortin-4 r eceptor (MC4-R) in neuroendocrine and autonomic control circuits in the brain. Molecular Endocrinology, 8 , 1298-1308. Mrosovsky, N. (1990). Rheostasis: The physiology of change . New York: Oxford University Press, Inc. Mueller, K., & Hsiao, S. (1979). Consistenc y of cholecystokinin satiety effect across deprivation levels and motivational states. Physiology & Behavior, 22 (5), 809-815 . Nagalla, S.R., Barry, B.J., Creswick, K.C., Ede n, P., Taylor, J.T., & Spindel, E.R. (1995). Cloning of a receptor for amphibian [Phe13]bo mbesin distinct from the receptor for gastrin-releasing peptide: identification of a fourth bombesin receptor subtype (BB4). Proceedings of the National Academy of Sciences U S A., 92 (13), 62056209. Niswender, K.D., & Schwartz, M.W. (2003). Insulin and leptin revisited: adiposity signals with overlapping physiological a nd intracellular signaling capabilities. Frontiers in Neuroendocrinology, 24 (1), 1-10. Ollmann, M.M., Wilson, B.D., Yang, Y.K., Kern s, J.A., Chen, Y., Gantz, I., & Barsh, G.S. (1997). Antagonism of central melanoc ortin receptors in vitro and in vivo by agouti-related protein. Science, 278 (5335) , 135-138. Pan, W., Kastin, A.J., Yu, Y., Cain, C.M., Fairburn, T., Stutz, A.M., et al. (2005) Selective tissue uptake of agouti-rela ted protein (82) and its modulation by fasting. Endocrinology, 146 (12), 5533-5539. Petersen, S., & McCarthy, J.C. (1981). Co rrelated changes in feeding behavior on selection for large and small body size in mice. Behavioral Genetics, 11 , 57-64.

PAGE 117

107 Phillips R.J., & Powley, T.L. (1998). Gastric volume detection after selective vagotomies in rats. American Journal of Physiology, 274 (6 Pt 2), R1626-R1638 . Phillips, G.D., Le Noury, J., Wolterink, G., Donselaar-Wolterink, I., Robbins, T.W., & Everitt, B.J. (1993). Cholecystokinin-dopa mine interactions within the nucleus accumbens in the control over behaviour by conditioned reinforcement. Behavioural Brain Research, 55 (2), 223-231. Raffin-Sanson, M.L., & Bertherat, J. (2001) Mc 3 and Mc4 receptors: complementary role in weight control. European Journal of Endocrinology, 144 (3), 207-208. Ravelli, G.P., Stein, Z.A., & Susser, M.W. (1976). Obesity in young men after famine exposure in utero and early infancy. The New England Journal of Medicine, 295 (7), 349-353. Rehfeld, J.F. (1978) Immunochemical studies on cholecystokinin. II. Distribution and molecular heterogeneity in the central ne rvous system and small intestine of man and hog. The Journal of Biological Chemistry, 253 (11), 4022-4030. Richardson, N.R., & Roberts, D.C.S. (1996). Progressive ratio schedules in drug selfadministration studies in rats: A met hod to evaluate reinforcing efficacy. Journal of Neuroscience Methods, 66, 1-11. Rinaman, L. (2003). Hindbrain nor adrenergic lesions attenuate anorexia and alter central cFos expression in rats after ga stric viscerosensory stimulation. Journal of Neuroscience, 23, 10084. Rodriguez-Membrilla, A., Martinez, V., & Verg ara, P. (1995). Peripheral and central cholecystokinin receptors regulate postpran dial intestinal motility in the rat. The Journal of Pharmacology and Experimental Therapeutics, 275 (1), 486-493. Rolls, B.J., & Hetherington, M. (1989). The role of variety in eating and body weight regulation. In R. Shepherd (Ed.), Handbook of the psychophysiology of eating (pp. 57-84). New York: Wiley. Rosenzweig, M.R., Leiman, A.L ., & Breedlove, S.M. (1999). Biological psychology: An introduction to behavioral, cogni tive and clinical neuroscience (2nd ed). Sunderland, MA: Sinauer Associates, Inc. Samama, P., Rumennik, L., & Grippo, J.F. (2003). The melanocortin receptor MCR4 controls fat consumption. Regulatory Peptides, 113 (1-3), 85-88. Schwartz, G. (2000). The role of gastrointe stinal vagal afferents in the control of food intake: Current prospects. Nutrition, 16 (10), 866-873. Schwartz, G.J., McHugh, P.R., & Moran, T.H. (1991). Integration of vagal afferent responses to gastric loads and cholecystokinin in rats. American Journal of Physiology, 261 (1 Pt 2), R64-R69.

PAGE 118

108 Schwartz, M.W., Woods, S. C., Porte, D., Jr ., Seeley, R. J., & Baskin, D. G. (2000). Central nervous system control of food intake. Nature, 404 , 661-671. Sclafani, A., & Ackroff, K. (2003). Rein forcement value of sucrose measured by progressive ratio operant licking in the rat. Physiology & Behavior, 79 , 663-670. Sclafani, A., & Kluge, L. (1974). Food motivation and body weight levels in hypothalamic hyperphagic rats: A dual li postat model of hunger and appetite. Journal of Comparative & Physiological Psychology, 86, 28-46. Seeley, R.J., Kaplan, J., & Grill, H. (1993). E ffects of interrupting an intraoral meal on meal size and meal duration in rats. Appetite, 20 (1), 13-20. Seeley, R.J., Grill, H., & Kaplan, J. (1994). Ne urological dissociation of gastrointestinal and metabolic contributions to meal size control. Behavioral Neuroscience, 108 (2), 347-352. Segal-Lieberman, G., Bradley, R.L., Kokkot ou, E., Carlson, M., Trombly, D.J., Wang, X., et al. (2003). Melanin-concentrating horm one is a critical medi ator of the leptindeficient phenotype. Proceedings of the National Academy of Sciences USA, 100, 10085-10090. Shillabeer, G., & Davison, J.S. (1987). Endoge nous and exogenous cholecystokinin may reduce food intake by different mechanisms. American Journal of Physiology, 253 (2 Pt 2), R379-R382. Silver, A.J., Flood, J.F., & Morley, J.E. (1988) . Effect of gastrointestinal peptides on ingestion in old and young mice. Peptides, 9 (2), 221-225. Smith, G.P. (1996). The direct and in direct controls of meal size. Neuroscience and Biobehavioral Reviews, 20 (1), 41-46. Smith, G.P. (1998). Pregastric and ga stric satiety. In G.P. Smith (Ed.), Satiation: From gut to brain (pp. 10-39). New York: Oxford University Press, Inc. Smith, G.P., & Gibbs, J. (1998). The satiating effects of cholecystokinin and bombesinlike peptides. In G.P. Smith (Ed.), Satiation: From gut to brain (pp. 97-125). New York: Oxford University Press, Inc. Smith, G.T., Moran, T.H., Coyle, J.T., Kuhar, M.J., O'Donahue, T.L., & McHugh, P.R. (1984). Anatomic localization of cholecystokin in receptors to the pyloric sphincter. American Journal of Physiology , 246 (1 Pt 2), R127-R130. Snow, N.D., Prpic, V., Mangel, A.W., Sharar a, A.I., McVey, D.C., Hurst, L.J., et al. (1994). Regulation of cholecystokinin secre tion by bombesin in STC-1 cells. American Journal of Physiology , 267(5 Pt 1), G859-G865.

PAGE 119

109 Snowdon, C.T. (1975). Production of satiety with small intraduodenal infu sions in the rat. Journal of Comparative & Physiological Psychology, 88 (1), 231-238. Stafford, D., LeSage, M.G., & Glowa, J.R. (1998). Progressive-ratio schedules of drug delivery in the analysis of drug self-administration: A review. Psychopharmacology, 139 (3), 169-182. Stellar, E. The physiology of motivation. 1954. (1994). Psychological Review, 101 (2), 301-311. Ste Marie, L., Miura, G.I., Marsh, D.J., Ya galoff, K., & Palmiter, R.D. (2000). A metabolic defect promotes obesity in mice lacking melanocortin-4 receptors. Proceedings of the National Academy of Sciences USA, 97 , 12339-12344. Strohmayer, A.J., & Smith, G.P. (1986). Obese male mice (ob/ob) are normally sensitive to the satiating effect of CCK-8. Brain Research Bulletin, 17 (4), 571-573. Strohmayer, A.J., & Smith, G.P. (1987). Th e meal pattern of genetically obese ( ob/ob ) mice. Appetite, 8 , 111-123. Strubbe, J.H. (1992). Parasympathetic involve ment in rapid meal-associated conditioned insulin secretion in the rat. American Journal of Physiology, 263 (3 Pt 2), R615R618 . Stuckey, J.A., Gibbs, J., & Smith, G.P. (1985). Neural disconnection of gut from brain blocks bombesin-induced satiety. Peptides, 6 (6), 1249-1252. Sutton, G.M., Patterson, L.M., & Berthoud, H.R. (2004). Extracellular signal-regulated kinase 1/2 signaling pathway in solitary nucleus mediates cholecystokinin-induced suppression of food intake in rats. Journal of Neuroscience, 24 (45), 10240-10247. Sutton, G.M., Duos, B., Patterson, L.M., & Berthoud, H.R. (2005). Melanocortinergic modulation of cholecystokinin-induc ed suppression of feeding through extracellular signal-regulated kinase signaling in rat solitary nucleus. Endocrinology, 146 (9), 3739-3747. Takahashi, K.A., & Cone, R.D. (2005). Fasting induces a large, leptin-dependent increase in the intrinsic action potential freque ncy of orexigenic arcuate nucleus neuropeptide Y/Agouti-related protein neurons. Endocrinology, 146 (3), 1043-1047. Taylor, I.L., & Garcia, R. (1985). Effects of pancreatic polypeptide, caerulein, and bombesin on satiety in obese mice. American Journal of Physiology, 248 (3 Pt 1), G277-G280. Teitelbaum, P. (1957). Random and food-dire cted activity in h yperphagic and normal rats. Journal of Comparative & Ph ysiological Psychology, 50, 486-490.

PAGE 120

110 Thaw, A.K., Smith, J.C., & Gibbs, J. (1998) Mammalian bombesin-like peptides extend the intermeal interval in freely feeding rats. Physiology & Behavior, 64 , 425– 428. Vaisse, C., Clement, K., Durand, E., Hercber g, S., Guy-Grand, B., & Froguel, P. (2000). Melanocortin-4 receptor mutations are a frequent and heterogeneous cause of morbid obesity. The Journal of Clinical Investigation, 106 (2), 253-262. Vaughan, C.H., & Rowland, N.E. (2003) Meal patt erns of lean and le ptin-deficient obese mice in a simulated foraging environment . Physiology & Behavior, 79 (2), 275279. Vaughan, C.H., Moore, M.C., Haskell-Luev ano, C., & Rowland, N.E. (2005). Meal patterns and foraging in melanocortin receptor knockout mice. Physiology & Behavior, 84 (1), 129-133. Vergoni, A.V., Schioth, H.B., & Bertolini, A. (2000). Melanocortins and feeding behavior. Biomedicine & Pharmacotherapy. 54 (3), 129-134. Warwick, Z.S., & Weingarten, H.P. (1995). De terminants of high-fat diet hyperphagia: experimental dissection of orosen sory and postingestive effects. American Journal of Physiology, 269 (1 Pt 2), R30-R37. Warwick, Z.S., McGuire, C.M., Bowen, K.J., & Synowski, S.J. (2000). Behavioral components of high-fat diet hyperphagia: meal size and postprandial satiety. American Journal of Physiolgy: Regulatory, Integrative and Comparative Physiology, 278 (1), R196-R200. Weatherford, S.C., Chiruzzo, F.Y., & Laughton, W.B. (1992) Satiety induced by endogenous and exogenous cholecystokinin is mediated by CCK-A receptors in mice. American Journal of Physiology, 262 (4 Pt 2), R574-R578. Weide, K., Christ, N., Moar, K.M., Arens, J., Hinney, A., Mercer, J.G., et al. (2003). Hyperphagia, not hypometabolism, causes early onset obesity in melanocortin-4 receptor knockout mice. Physiological Genomics, 13 (1), 47-56. Weller, A., Gispan, I.H., Armony-Sivan, R., R itter, R.C., & Smith, G.P. (1997). Preloads of corn oil inhibit independent inge stion on postnatal day 15 in rats. Physiology & Behavior, 62 (4), 871-874. West, D.B., Fey, D., & Woods, S.C. (1984). Cholecystokinin persistently suppresses meal size but not food intake in free-feeding rats. American Journal of Physiology, 246 (5 Pt 2), R776-R787. West, D.B., Greenwood, M.R., Marshall, K.A ., & Woods, S.C. (1987). Lithium chloride, cholecystokinin and meal patterns: eviden ce that cholecystokinin suppresses meal size in rats without causing malaise. Appetite, 8 (3), 221-227.

PAGE 121

111 Williams, D.L., Kaplan, J.M., & Grill, H.J. (2000). The role of the dorsal vagal complex and the vagus nerve in feeding effects of melanocortin-3/4 receptor stimulation. Endocrinology, 141 (11) , 1332. Williams, D.L., Grill, H.J., Weiss, S.M., Bair d, J.P., & Kaplan, J.M. (2002). Behavioral processes underlying the intake suppressive effects of melanocortin 3/4 receptor activation in the rat. Psychopharmacology, 161 (1):47-53 Williams, D.L., Bowers, R.R., Bartness, T. J., Kaplan, J.M., & Grill, H.J. (2003). Brainstem melanocortin 3/4 receptor stim ulation increases uncoupling protein gene expression in brown fat. Endocrinology, 144, 4692. Williams, G., Harrold, J.A., & Cutler, D.J. (2000). The hypothalamus and the regulation of energy homeostasis: Lif ting the lid on a black box. The Proceedings of the Nutrition Society, 59 (3), 385-396. Yamada, K., Wada, E., Santo-Yamada, Y., & Wada, K. (2002). Bombesin and its family of peptides: Prospects for the treatment of obesity. European Journal of Pharmacology, 440 (2-3), 281-290. Yaswen, L., Diehl, N., Brennan, M.B., & Hoc hgeschwender, U. (1999). Obesity in the mouse model of pro-opiomelanocortin deficiency responds to peripheral melanocortin. Nature Medicine, 5 (9), 1066-1070 Yen, T.T., Lowry, L., & Steinmetz, J. (1968) . Obese locus in Mus musculus: A gene dosage effect. Biochemical and Biophysical Research Communications, 33 (6), 883887. Zhang, M., Balmadrid, C., & Kelley, A.E. (2003). Nucleus accumbens opioid, GABaergic, and dopaminergic modulation of palatable food motivation: contrasting effects revealed by a pr ogressive ratio study in the rat. Behavioral Neuroscience, 117 (2), 202-211. Zhang, Y., Proenca, R., Maffei, M., Barone , M., Leopold, L., & Friedman, J.M. (1994). Positional cloning of the mouse obese gene and its human homologue. Nature, 372 (6505), 425-432.

PAGE 122

112 BIOGRAPHICAL SKETCH Cheryl Hope Vaughan was born in Clar endon, Jamaica, in 1980. She moved to Miami, FL, at age 8. She completed high school at Chaminade-Madonna College Preparatory in Hollywood, FL. Cheryl receiv ed her B.A. in psychology from St. Thomas University in May of 2000. She attended the Un iversity of Florida for graduate training in the behavioral neuroscience area of the psychology department. She will graduate in May 2006 with her Ph.D. and go on to a pos tdoctoral position at Georgia State University.