Dietary self-selection and intake of protein and energy is altered by the form of the diets

What does self-selection of dietary proteins in rats tell us about protein requirements and body weight control?

Omnivores are able to correctly select adequate amounts of macronutrients from natural foods as well as purified macronutrients. In the rat model, the selected protein levels are often well above the requirements estimated from the nitrogen balance. These high intake levels were initially interpreted as reflecting poor control of protein intake, but the selected levels were later found to be precisely controlled for changes in dietary protein quality and adjusted for cold, exercise, pregnancy, lactation, age, etc. and therefore met physiological requirements. Several authors have also suggested that instead of a given level of protein intake, rodents regulate a ratio of protein to dietary carbohydrates in order to achieve metabolic benefits such as reduced insulin levels, improved blood glucose control, and, in the long term, reduced weight and fat gain. The objective of this review was to analyze the most significant results of studies carried out on rats and mice since the beginning of the 20th century, to consider what these results can bring us to interpret the current causes of the obesity pandemic and to anticipate the possible consequences of policies aimed at reducing the contribution of animal proteins in the human diet.

Effect of chronic treatment with clomipramine on food intake, macronutrient selection and body weight gain in rats

Long-term treatment with clomipramine (CMI), a tricyclic antidepressant, induces food craving and body weight gain in patients. The present study investigated the effects of chronic treatment with CMI on total food intake, macronutrient selection, and body weight gain in rats. Male Wistar rats were maintained on a dietary self-selection regime with separate sources of protein, fat and carbohydrate. Animals received i.p. injections of CMI (0, 3, 10, 30 mg/kg) during 27 consecutive days. Food consumption and body weight were recorded daily and results were calculated as average of three consecutive days, namely during pre-treatment (3 d before pharmacological treatment), treatment (7th-9th; 16th-18th and 25th-27th days), and post-treatment (28th-33rd days). Results showed that CMI (30 mg/kg) significantly decreased energy intake during all treatment period, an effect that was related to a decrease in both carbohydrate-rich diet intake and body weight gain. At dose of 3 mg/kg CMI increased the total energy intake in the 16th-18th days, suggesting an apparent biphasic effect of chronic treatment with CMI on caloric intake. Chronic administration with CMI (27 d) did not alter protein-rich or fat-rich diet consumption. The main result of this study indicated that chronic treatment with CMI decreases rather than increase food consumption and body weight gain in rats exposed to a macronutrient self-selection procedure.

Avoidance of hydrolyzed casein by mice

When casein, a milk protein, is hydrolyzed, it renders human foods that contain it (e.g., hypoallergenic infant formula, cheeses) distasteful to many people. This rejection of hydrolyzed casein (HC)-containing foods has recently been found to also occur in a non-human species (deer, Odocoileus spp.). Identifying other animals that avoid HC would facilitate understanding how and why HC-containing food is often rejected. This study determined whether HC-containing food is avoided by Mus musculus and whether consumption patterns were sensitive to testing conditions, specifically food form (powder, pellet or dough) and food access (ad libitum or 1.5 h/day following 6 h of food deprivation). Diets were offered in two-choice tests that paired an HC-containing food with an intact casein-containing alternative at seven protein concentrations (0%-50% w/w). Five experimental groups were tested under different combinations of food form and food access. Three groups (ad lib/powder, ad lib/pellet, and 1.5 h/pellet) avoided the HC diet starting at the 30% protein level. At the 40% and 50% protein levels, all groups showed strong avoidance of HC. Although testing conditions influenced total caloric intake and body weight gain, avoidance of HC at the highest concentrations was robust to the manipulations in experimental conditions. Our study suggests that mice may be a useful model for understanding the mechanisms of HC rejection.

Obesity by choice: the powerful influence of nutrient availability on nutrient intake

The consumption of nutrients by rodents is markedly influenced by the number of containers of each nutrient provided. Most rats given a choice from separate sources of protein, carbohydrate, and fat thrived if given one cup of each but half failed to thrive if given one cup of each and three extra cups of carbohydrate or fat. Rats given five bottles of sucrose solution and one bottle of water became fatter than rats given five bottles of water and one of sucrose. These studies in rats may point to a model for human obesity, in which the availability of food can override physiological controls of ingestion.

Divergence in proportional fat intake in AKR/J and SWR/J mice endures across diet paradigms

These experiments were designed to test the hypothesis that the contrasting patterns of macronutrient selection described previously in AKR/J (fat preference) and SWR/J (carbohydrate preference) mice are not dependent on a single diet paradigm. The effect of mouse strain on proportional fat intake was tested in naive mice by presenting two-choice diets possessing a variety of physical, sensory, and nutritive properties. In three separate experiments, AKR/J mice preferentially selected and consumed a higher proportion of energy from the high-fat diet than SWR/J mice. Specifically, this phenotypic difference was observed with 1) fat-protein vs. carbohydrate-protein diets, independent of fat type (vegetable shortening or lard), 2) isocaloric, high- vs. low-fat liquid diet preparations, and 3) high- vs. low-fat powdered-granular diets. These results confirm our previous observation of a higher proportional fat intake by AKR/J compared with SWR/J mice using the three-choice macronutrient selection diet and show that this strain difference generalizes across several diet paradigms. This strain difference is due largely to the robust and reliable fat preference of the AKR/J mice. In contrast, macronutrient preference in SWR/J mice varied across paradigms, suggesting a differential response by this strain to some orosensory or postingestive factor(s).

Macronutrient-specific dietary selection in rodents and its neural bases

The only evidence for nutrient selection comes from baseline or treatment effects on nutrient intakes that are qualitatively similar when sensorily contrasting forms of each macronutrient are investigated and/or dietary compositions and strains of rat or mouse are different within or between laboratories. By that criterion the only potential case of a treatment reliably altering macronutrient selection identified in the present review of the literature is d-norfenfluramine, fluoxetine and paraventricular serotonin (5-HT) reducing the intake of dextrin-containing diets at early dark. The only clear example of reverse effects of an agonist and an antagonist on dietary intake was found with serotonergic agents. Claims for catecholaminergic or opioid involvement in protein intake and peptidergic involvement in carbohydrate intake were not substantiated. There remain the issues of which learnt macronutrient-specific postgastric actions and sensory cues from the affected diet rely on the neural pathway(s) on which the drug is acting to alter dietary selection. Until experiments address these questions, the neural bases of nutrient-specific appetites will remain unknown. Drug effects must be consistent across differently textured and flavoured versions of each macronutrient tested.

Effect of diet textural characteristics on the temporal rhythms of feeding in rats

To examine whether the diurnal rhythms of protein-rich and carbohydrate-rich diet ingestion can be altered by presenting the diets in different textural forms, adult male Wistar rats were assigned to two dietary groups. One group received a two-way choice between high-protein powder and high-carbohydrate granular (HPP-HCG) diets. In the other group the textures were reversed [high-protein granular and high-carbohydrate powder (HPG-HCP) diets]. Rats fed HPP-HCG diets selected significantly less protein (kcal) vs. rats fed HPG-HCP diets, during the 24-h and 12-h dark phase and during the 4-h early and late dark phases. Carbohydrate intakes of the two dietary groups were not significantly different. Total caloric intake for the HPG-HCP dietary group was significantly higher than that of the HPP-HCG dietary group during the 24-h and 12-h dark phase. Body weight was significantly lower in rats fed HPP-HCG diets. In conclusion, macronutrient-rich diets presented in different textural forms alter protein-rich diet ingestion and total energy intake.

Self-selection of dietary protein and energy by broilers grown under a tropical climate: effect of feed particle size on the feed choice

Broilers, 2 wk of age, that had been previously adapted to energy: protein choice feeding, were offered corn (either ground, cracked, or presented as whole grains) and a protein concentrate (43.7% CP) in mash or pellet form. When corn was fed as whole grains, protein concentrate in the selected diet was significantly higher (35.1%) than with cracked corn (29.3%) or ground corn (29.1%). Presenting the concentrate as pellets resulted in a significantly higher concentration in the diet (32.7%) than when mash concentrate was fed (29.6%). Live BW at 4 and 6 wk of age were not significantly affected by feed texture. However, offering corn as whole grains or concentrate as pellets induced a significant improvement in feed efficiency. Total time to eat larger size particles (whole grains, pelleted concentrate) was significantly less than total time to eat ground corn or mash concentrate. Furthermore, the mean duration of the feeding bouts was two times shorter for whole grains (48 s) than for ground corn (98 s) and for pelleted concentrate (56 s) than for mash concentrate (114 s). Chickens ate whole grains or pellets at a significantly slower rate (number of pecks per second feeding time) than when eating ground corn or mash concentrate. There was a rejection during the first 24 h when the form of the concentrate (mash to pellets) was changed. Full adaptation to the new size of the concentrate required about 3 d.

Effects of dietary preference and galanin administration in the paraventricular or amygdaloid nucleus on diet self-selection

Microinjection of galanin into the paraventricular nucleus (PVN) of the hypothalamus has been reported to predominantly increase dietary fat consumption during the light or late dark periods. However, there have been no studies of the macronutrient-specific effects of galanin administered into the amygdala (AMY). We used male Sprague-Dawley rats to test the effect of galanin administered into the PVN or central nucleus of the AMY on diet selection using two different protocols: the two-choice composite diet (10% or 55% energy as fat) (Study 1) or the three-choice macronutrient diet (Study 2). In the first study, total food intake was significantly increased 60 min after galanin injection (0.3 nmol) into either the PVN or AMY, but there was no difference in intake between the low- and high-fat diets. In Study 2, using the three-choice macronutrient diet, the feeding stimulation produced by galanin in the PVN also was not diet-dependent. In the AMY, there was a significant effect of diet on the feeding response to galanin, due primarily to the lack of stimulation of protein intake, whereas the intakes of fat and carbohydrate were not significantly different. In summary, there was no difference in the effect of galanin on fat versus carbohydrate intake for either nuclei, whether animals were injected during the light or the late-dark phase. Because rats often display preferences when allowed to choose among individual macronutrient sources, we also examined the relationship between baseline macronutrient preference and the feeding response to galanin in groups of highly fat-preferring and carbohydrate-preferring rats (Study 3). After PVN galanin injection, fat-preferring rats demonstrated significant increases in their consumption of both carbohydrate and fat, whereas carbohydrate-preferring rats responded by eating carbohydrate almost exclusively. There were significant positive associations between baseline macronutrient intake and galanin-induced intake. The results of these studies combined suggest that the effect of galanin on macronutrient selection is not specific for fat and that underlying macronutrient preferences are important in determining the feeding response to galanin.

Compared effects of cold ambient temperature and cytokines on macronutrient intake in rats

To compare the effects of cold and cytokines on spontaneous dietary self-selection, rats (n = 14) were given free access to carbohydrate, protein and fat diets for 4 hours a day. After a 10-day period of habituation to this regimen, they were injected with physiological saline, IL-1 beta (4 micrograms/rat ip) or LPS (83 micrograms/rat ip) or exposed to cold (5 degrees C), the order of treatments being randomized. LPS- and IL-1 beta-treated rats ate less, but ingested relatively more carbohydrate and less protein whereas relative fat intake remained unchanged. In contrast, cold exposed rats slightly increased their food intake but in a non significant manner. They also increased their relative intake of fat but did not change their relative intake of carbohydrate and protein. These results are discussed with respect to the pyrogenic and metabolic effects of cytokines.

5-HT manipulation and dietary choice: variable carbohydrate (Polycose) suppression demonstrated only under specific experimental conditions

The effects of six 5-HT anorectic agents, d-fenfluramine (5-HT releaser and reuptake inhibitor), fluoxetine (5-HT reuptake inhibitor), mCPP (5-HT1B/5-HT1C receptor agonist), RU24969 (5-HT1A/5-HT1B receptor agonist), MK212 (5-HT1C receptor agonist) and DOI (5-HT2/5-HT1C receptor agonist), and two non-5-HT anorectic agents, salbutamol (beta 2-adrenergic agonist) and d-amphetamine (catecholaminergic agonist), were examined in an experimental procedure designed to disclose selective effects on carbohydrate consumption. In this procedure, a revised version of what we have termed "The Classic Sclafani Paradigm", animals are presented with powdered Polycose as an optional carbohydrate supplement to hydrated chow (nutritionally complete diet). All drugs produced significant reductions in total (hydrated chow plus powdered Polycose) intake. However, only the 5-HT drugs DOI and fluoxetine exerted significantly stronger anorectic effects on intake of powdered Polycose than on intake of hydrated chow. d-Fenfluramine also showed a tendency to selectively suppress Polycose intake but this effect marginally failed to reach significance. These results suggest that when experimental conditions are favourable, what appears to be selective carbohydrate (Polycose) suppression can be demonstrated with certain 5-HT drugs. They also suggest that a selective effect on carbohydrate intake is not the most prominent feeding response to 5-HT drugs.

Protein selection by rats adapted to high or moderately low levels of dietary protein

After preliminary studies on flavor acceptability, patterns and indices of subsequent feeding behavior were monitored by computer in young rats which were adapted to 15% or 70% casein diets before being offered, sequentially, choices between flavored diet pairs in which the proportions of percentage casein were 5/65, 5/55, 5/45, 5/35 and 5/25. Similarly adapted rats received these choices in the reverse sequence. Rats adapted to 15% casein usually ate randomly from the diet pairs and selected approximately 15-30% casein; individual behaviors were prominent. The 70% casein groups avoided the higher casein diet, often within minutes (except for the first-offered 5/25 choice), and seldom selected more than 10% casein; individual differences were infrequent. Such rats also distinguished between flavored 70% and 65% casein diets. Sizes and numbers of meals and rates of eating differed for the paired diets, especially for rats adapted to 70% casein. A flavor added to the 70% casein adaptation diet was not avoided when present only in the 5% casein diet of a 5/65 choice. Rats adapted to 70% soy protein before receiving flavored 5/65 to 5/25 choices selected 20-28% soy protein, a level far above those of casein selections by rats adapted to 70% casein. Dietary adaptation and type of protein thus affect subsequent diet selection and feeding patterns and indices.

Impact of hypothalamic d-norfenfluramine and peripheral d-fenfluramine injection on macronutrient intake in the rat

Previous research with hypothalamic injection of serotonin (5-HT) has suggested that this monoamine may act within the medial hypothalamus to suppress carbohydrate intake in a selective, phasic and circadian-related fashion. To explore further the action of 5-HT in the brain, the present studies tested the serotonergic stimulants, d-norfenfluramine (DNF) and d-fenfluramine (DF), in freely feeding, brain-cannulated animals maintained on pure macronutrient diets (protein, carbohydrate and fat) and tested at different times of the diurnal cycle. The results show that administration of DNF into the paraventricular nucleus (PVN) potently influences appetite for a specific nutrient at a particular time of the light-dark cycle. Specifically, DNF injection at the onset of the nocturnal (active) period selectively and dose-dependently suppresses carbohydrate consumption, while leaving protein and fat intake unchanged. This drug, however, has no effect, even at high doses, on macronutrient intake in the middle and late h of the dark phase, strongly implicating a function for hypothalamic 5-HT in the control of carbohydrate ingestion at the beginning of the nocturnal cycle. The possibility that peripherally injected DF may act, in part, through this endogenous serotonergic system is supported by the additional finding that, at low doses of 0.06-0.5 mg/kg, DF preferentially modulates carbohydrate ingestion exclusively at the onset of the nocturnal period. However, at doses above 0.5 mg/kg, this compound produces a potent and general suppression of feeding of all macronutrients. In animals with brain cannulas aimed at different hypothalamic nuclei, the feeding-suppressive effect of DNF is found to be site specific; it is localized to the medial hypothalamic nuclei, including the ventromedial, suprachiasmatic and dorsomedial nuclei as well as the PVN. Serotonin in these nuclei may function to produce satiety specific for carbohydrate and, through the suprachiasmatic nucleus, control energy intake in a circadian-related manner.

Medial hypothalamic serotonin: role in circadian patterns of feeding and macronutrient selection

Hypothalamic serotonin (5-HT) is believed to have an inhibitory effect on food intake in a variety of species. To define more precisely the nature of this effect, this study investigated the effects of medial hypothalamic 5-HT injection on natural patterns of macronutrient intake in freely feeding rats. Serotonin (5-20 nmol) was injected directly into the paraventricular nucleus (PVN) of brain-cannulated rats maintained ad libitum on pure macronutrient diets, protein, carbohydrate and fat, and measurements of nutrient intake were taken one hour later. To assess whether the action of 5-HT on macronutrient intake varies across the light-dark cycle, these tests were conducted at 3 different times in the nocturnal feeding period, during hours 1, 6 and 11 after lights out. The results demonstrate that the suppressive effect of PVN 5-HT on food intake is dose dependent, nutrient selective, as well as time dependent. Specifically, PVN injection of 5-HT, at all doses tested, was effective at only one time of the nocturnal cycle, namely, at the onset of the active, dark period. While no behavioral effect of 5-HT was detected in the middle and late phases of the dark, a strong, dose-dependent reduction of nutrient intake was revealed immediately after lights out. This suppressive effect was characterized by a highly selective decrease in carbohydrate intake, along with a significant enhancement in preference for protein, as well as for fat, and little change in total caloric intake.(ABSTRACT TRUNCATED AT 250 WORDS)

Do serotoninergic drugs decrease energy intake by reducing fat or carbohydrate intake? Effect of d-fenfluramine with supplemented weight-increasing diets

The capacity of serotoninergic drugs to selectively suppress carbohydrate (CHO) intake was investigated using a procedure sensitive to drug action. The drug d-fenfluramine was administered chronically to rats whose weight had been increased by exposure to either a fat or CHO-supplemented hyperphagia-inducing diet. The drug exerted a more potent anorexic effect and weight-reducing action in rats given the dietary supplements than in control chow-fed rats. Tolerance to the drug was not apparent even after 40 days of treatment. However, there was no evidence for a selective inhibition of CHO intake, nor was the drug more potent with the CHO-supplemented diet. d-Fenfluramine was equally effective against the hyperphagia and weight gain induced by either fat or CHO supplements.

Effect of phenylpropanolamine on diet selection in rats

Four doses of phenylpropanolamine (PPA; i.e., dl-norephedrine: 2.5, 5, 10, and 20 mg/kg) and saline were injected intraperitoneally in female rats maintained on a dietary self-selection paradigm. Intake of all three macronutrients (carbohydrate, fat, and protein) was equally affected by high doses of PPA. Lower doses decreased fat and protein more than carbohydrate.

Fenfluramine and amphetamine suppress dietary intake without affecting learned preferences for protein or carbohydrate cues

The effects of pharmacological modulation of monoamine transmitter activity on genuine nutrient selection were assessed: that is, drug-induced changes in nutrient-specific dietary choice behaviour were measured, using rats that had learned to select an odour cueing protein content or carbohydrate content of the diet. Anorexigenic doses of DL-fenfluramine-HCl (1.25 or 2.5 mg/kg) and D-amphetamine2 SO4 (0.5 or 1.0 mg/kg) did not affect the selection of protein-paired or carbohydrate-paired odours by trained rats. This weighs against the claim that 5-hydroxytryptamine or catecholamine transmitters are involved in the selection of macronutrients, as distinct from selection between diets in response to sensory differences functionally unrelated to nutrient composition.

Food intake and selection after peripheral tryptophan

Two studies investigated the effects of peripheral (IP) administration of the dietary indispensible amino acid tryptophan, on food intake and macronutrient selection in rats adapted to a 12 hr nocturnal feeding period and a choice of 10% and 60% casein diets. In a dose-response study (35, 55, 75, 95, 115 mg/kg), the threshold dose of 75 mg/kg produced a significant reduction in total food intake (3.6 to 2.3 g, p less than 0.05) during the first hour of feeding. The reduction in carbohydrate intake (2.1 vs. 1.2 g, p less than 0.05) was greater than that for protein intake (1.6 vs. 1.1 g, p less than 0.05). Twelve hr total food intake was also decreased (20.9 to 19.5 g, p less than 0.05) and this was attributable to decreased carbohydrate intake (13.2 to 11.8 g, p less than 0.05). In a second study designed to determine if tryptophan's effects were mediated by the central nervous system, brain tryptophan uptake was blocked by co-injecting valine with tryptophan. The significant reduction in first hour total food intake by tryptophan was not prevented by co-injection of an equal quantity of valine (3.5 to 1.8 g, p less than 0.05). Again the suppression of carbohydrate intake (2.0 to 0.9 g p less than 0.05) was greater than that for protein intake (1.5 to 0.9 g, p less than 0.05). This dose of valine significantly reduced brain tryptophan uptake by 16% (21.3 to 17.8 micrograms/g, p less than 0.05) and when administered alone did not affect first hour total food intake (3.1 vs 3.2 g).(ABSTRACT TRUNCATED AT 250 WORDS)