Eating behavior and body weight regulation have evolved in mammals into highly complex processes, with multiple feedback loops involving both peripheral and central systems. This complexity helps to explain the notable failure of unidimensional approaches to dieting and treating disturbances in eating and weight control. Over the past 30 years, this laboratory has pursued a multidisciplinary research program to identify and characterize neurobiological substrates in rodents that mediate these complex processes, under both normal and pathological conditions. We have demonstrated that brain mechanisms and diet consumption function in a highly interactive manner. For example, with injections of neurochemicals into specific brain sites, we have observed profound changes in eating patterns and preferences for the macronutrients, carbohydrate or fat, along with changes in circulating hormones and metabolites. Also, individual differences in these behavioral and physiological patterns, which arise at different developmental stages and in relation to gender, are paralleled by corresponding shifts in endogenous expression and levels of these neurochemicals in the brain. In fact, individual differences expressed around puberty or produced by early exposure to fat- or carbohydrate-rich diets are found to be predictive of later pathologies. An important concept revealed by these results is that specific neurobiological systems and diet intake are functionally linked within a positive feedback loop, whereby a specific diet stimulates particular brain neurochemicals that in turn stimulate further consumption of that same diet. This diet-neurochemical-diet feedback process, while appropriate for producing overeating and gorging under conditions when food is scarce, helps to explain the eating and body weight disorders that develop when sugar- and fat- rich foods are abundant. Surprising recent findings show that similar positive feedback systems exist in the brain for controlling the consumption of alcohol and nicotine and that in utero exposure to these substances can produce permanent changes in brain neurochemical systems of the offspring that cause behavioral disturbances throughout life.
Early injection and lesion studies:
In 1970, little was known about brain neurochemical mechanisms controlling food intake and energy balance. In our early investigations, we discovered that acute injections into specific brain sites of the monoamines (norepinephrine, dopamine and serotonin) and neuropeptides (neuropeptide Y, galanin, and the opioids, enkephalin and dynorphin) have marked short-term effects on eating as well as hormones and metabolism. Chronic injections even produce long-term changes resulting in obesity or anorexia. Moreover, we demonstrated that the paraventricular nucleus of the hypothalamus and its projections, previously with little known function in energy balance, play a critical role in controlling food intake and body weight. The neurochemicals and their receptors in this nucleus were found be highly responsive to hormones and metabolites circulating in the blood, supporting the existence of closely linked neurochemical and neuroendocrine systems.
We then recognized at an early stage that diet also plays an important role in the functioning of these neural mechanisms. In the 1980s, we expanded our investigations to include a variety of different diets, either mixed or pure macronutrient diets, which allow animals to exhibit their natural eating patterns and nutrient preferences. This novel approach generated a range of important discoveries about the integral relationships that exist between brain neurochemicals, dietary nutrients, and even circulating nutrients. These studies demonstrated that imbalanced diets rich in carbohydrate or fat can, in themselves, create pernicious neurobiological feedback processes. These feedback processes, in turn, can have profound effects on eating patterns and subsequent weight gain. We first demonstrated that norepinephrine and neuropeptide Y, which stimulate food intake, have a close and specific relationship to dietary carbohydrate and sugar. The feeding response induced by hypothalamic injection of these neurochemicals is much stronger with diets rich in these macronutrients, and subpopulations of animals with a strong carbohydrate preference exhibit high levels of endogenous neuropeptide Y in the hypothalamus. We then discovered that other orexigenic peptides, galanin, the opioids enkephalin and dynorphin, and orexin, exhibit a close relation, not to carbohydrate but to dietary fat. Once again, the feeding responses induced by injections of these peptides are much stronger on diets rich in fat, and those rats with a natural preference for fat have higher levels of these particular peptides. These studies provided the first evidence supporting the existence of specific neurochemical systems differentially related to carbohydrate and fat. Further examination of the monoamines, serotonin and dopamine, which reduce feeding and produce satiety, has differentiated these neurotransmitters in the hypothalamus in terms of their inhibitory effects on macronutrient intake. These studies have helped to elucidate the actions of weight-reducing drugs, such as fenfluramine and amphetamine.
Brain systems, biological rhythms and developmental stages:
Investigations in the 1990s of biological rhythms and developmental stages confirmed these relationships between specific macronutrients and particular neurochemical systems in the brain. Spontaneous shifts in endogenous gene expression and production of these neurochemicals in the hypothalamus are found to coincide with natural shifts in the animals' ingestion of carbohydrate or fat. Specifically, endogenous norepinephrine and neuropeptide Y increase in physiological states when carbohydrate is the preferred macronutrient. This occurs at the start of the natural feeding cycle and early in development around the time of weaning, in association with elevated levels of the adrenal steroid, corticosterone. Further, galanin increases in physiological states when there was a spontaneous surge in preference for fat. This occurs during the middle of the natural feeding cycle, towards the onset of puberty, and during the proestrous stage of the female cycle, in association with elevated levels of the ovarian steroids. Studies of gender differences additionally reveal in females a stronger preference for fat and higher galanin expression in the hypothalamus as well as pituitary.
Effects of diet on brain peptides:
The success of these studies, showing close associations between the orexigenic peptides and specific macronutrients, led us to examine next whether specific diets and circulating nutrients can have direct impact on these neurochemical systems in the brain. These investigations provided a result that was all the more exciting because it generated an unexpected principle: that the macronutrient to which a specific peptide is known to be closely associated and to have a stronger behavioral effect does, in turn, directly stimulate the expression and production of this same peptide. These results suggested the existence of positive feedback loops that may have possible consequences of promoting overeating of this macronutrient. Specifically, the ingestion of a carbohydrate- or sugar-rich diet increases the production of neuropeptide Y, an effect that should further augment the animal's preference for this macronutrient. Moreover, the ingestion of a fat-rich diet, particularly one high in saturated fat, stimulates endogenous galanin, opioid peptides and orexin, which should promote the over-consumption of fat-rich foods. These effects of diet on the hypothalamus are found to be amazingly rapid, occurring within the context of a single day — or even a single meal. They are also shown to be anatomically localized, with the "carbohydrate-responsive" peptides expressed in the basal hypothalamus and the "fat-responsive" peptides expressed in nuclei of the dorsal hypothalamus. This provides new information on how the peptide systems in the hypothalamus may be functionally organized in relation to the specific macronutrients.
Physiological mechanisms mediating diet effects:
This strong evidence for causal effects of diet on neuropeptide systems encouraged us to explore possible physiological mechanisms mediating these effects. While our work and that of others had suggested that hormones may have some role in controlling these neurochemical systems, there were few studies of nutrients, e.g., glucose or lipids, circulating in the blood. Our recent investigations provided the first evidence that these nutrients, which shift dramatically in relation to carbohydrate and fat ingestion, have marked effects on hypothalamic systems. They show that the carbohydrate- and sugar-related neurochemicals, norepinephrine and neuropeptide Y as well as their receptors, respond specifically to shifts in glucose levels or glucose utilization and also that the fat-responsive peptides, galanin, opioids and orexin, are stimulated by a rise in circulating lipids, notably triglycerides, and suppressed by a reduction in fat metabolism. Thus, circulating nutrients and their metabolism may have an important role in determining the activity of hypothalamic neurochemicals within their positive feedback loops. Our evidence additionally demonstrates that, after an imbalanced meal rich in sugar or fat, elevated levels of glucose or triglycerides can predict and subsequently promote non-homeostatic overeating, ultimately leading to obesity. This behavioral effect may be a consequence of the nutrients over-riding of negative feedback signals. These findings relating blood-borne nutrients to neurochemicals provide insight into disturbances within the brain that may result from the hyperglycemia and hypertriglyceridemia commonly seen in obesity and diabetes and that, in turn, may further exacerbate these pathological conditions.
Predictive studies and diet exposure early in life:
In recent years, these investigations have encouraged us to pursue predictive studies in outbred strains of rats. These studies have identified subpopulations, at normal body weight, that have a differential propensity toward overeating and dietary obesity. For example, in adult rats, brief periods of access to a fat-rich diet reveal neuroendocrine and neurochemical disturbances that are predictive of a strong predisposition (or resistance) to later dietary obesity. Also, specific behavioral patterns before puberty are also found to differentiate rats, which exhibit marked disturbances in specific neurochemical systems that may contribute to their increased propensity for adult obesity. In a particularly exciting animal model developed and characterized in our laboratory, we have demonstrated that sugar- or fat-rich diets introduced early in life, even during pregnancy, can produce profound changes in circulating nutrients and brain neurochemical systems. These changes persist over time, even after the diet is returned to a balanced mixture, and they result from an increase in cell proliferation and neurogenesis in specific peptide systems that ultimately cause the offspring to overeat and become obese as adults.
In a separate series of investigations (in collaboration with Dr. Bart Hoebel at Princeton University), the peptides linked specifically to dietary fat have also been found to be closely related to the consumption of alcohol, which in turn is closely linked to fat intake. Voluntary drinking of 10% alcohol is significantly greater in rats previously shown to naturally overeat on a chronic high-fat diet or have elevated levels of circulating lipids, and it is higher in rats after consumption of a single high-fat meal or injection of a fat emulsion. Further, alcohol consumption is stimulated by hypothalamic injection of the fat-related peptides, galanin, opioids and orexin, while reduced by their receptor antagonists or a mutation of the peptide gene, and the intake or injection of alcohol stimulates the expression and production of these specific peptides but not those related to carbohydrate. Thus, there exists a positive feedback loop between these peptides and alcohol intake, which is similar to that seen with dietary fat and may be involved in promoting the over-consumption of alcohol as well as a fat-rich diet. These feed-forward peptide systems are further linked by their similar relation to circulating lipids and fat metabolism. Together, this evidence suggests for the first time that alcohol and fat intake during a meal may synergize to produce larger meals and greater alcohol consumption. Further studies using microdialysis indicate a possible role for forebrain dopamine in mediating the effects of these orexigenic peptides and the reinforcing properties of dietary fat and alcohol.
Current Research Projects
These investigations to date have allowed us, first, to develop a variety of natural animal models with precisely defined behavioral and physiological characteristics and, second, to define potential functions for known peptidergic and monoaminergic systems in both natural and pathological states. This work has laid a strong foundation for this laboratory's current research program. The goal of this research is to apply a variety of molecular biological techniques, including subtractive hybridization, microarrays, real-time quantitative PCR and in situ hybridization, to examine in these different animal models a broader range of neurobiological systems involved in eating behavior and alcohol intake. We are screening and functionally characterizing novel systems in the brain that respond to dietary and drug manipulations and circulating nutrients and that may contribute to natural or pathological states involving excess consummatory behavior and ultimate weight gain. In these studies, we are examining small brain areas and distinct cell groups in the hypothalamus as they relate to forebrain and hindbrain systems, with each animal model precisely characterized in terms of its hormonal and metabolic state, developmental stage, and dietary history.
Some initial work along these lines, as described in recent publications, demonstrates for the first time that: a) two proteins in the brain, apolipoprotein D and diacylglycerol kinase zeta, are stimulated by consumption of a fat-rich diet and constitute a downstream component of the signal transduction pathway for the adipocyte hormone, leptin, in the brain; b) a third protein, huntingtin-associated protein 1, is reduced by a high-carbohydrate diet and, through its suppression, mediates the feeding-inhibitory action of insulin in the brain; and c) imbalanced diets and drugs, when consumed during pregnancy, have profound impact on the development of the brain and specific neural systems in the embryo, permanently altering the behavior of the offspring throughout life. Most interestingly, the effects of fat, alcohol and nicotine are found to be very similar, involving changes in proteins that control consummatory behavior. We are investigating these neurochemical mechanisms to understand the pathologies that are causally related to eating disorders, obesity, and addiction to alcohol and drugs, in hopes of facilitating the design of more precise approaches for treating these disorders.