 |
INTRODUCTION |
Rats rapidly reduce their food intake when, after appropriate pretreatment, they are offered an amino acid-imbalanced (IMB)5 diet (reviewed in Harper et al. 1970
). This anorectic response is believed to be under the control of the brain (reviewed in Gietzen 1993, Leung and Rogers 1987), and considerable effort has gone into elucidating the neural mechanisms underlying the various aspects of this response. The roles of the anterior piriform cortex (APC) and amygdala have been worked out in some detail (Leung and Rogers 1971, Meliza et al. 1981), and lesion studies have been done throughout the brain (reviewed in Gietzen 1993 and Leung and Rogers 1987). Still, the neural mechanisms, particularly the neurochemical factors in other brain areas likely to be involved in the early feeding responses to IMB diets, have not yet been subjected to thorough evaluation.
The influence of IMB diets on the concentrations of amino acids has been examined in whole brain (Peng et al. 1972), with uptake into slices (Lutz et al. 1975) and via uptake studies after a variety of nutritional protocols (Tews et al. 1987), but regional differences among the various brain areas that could be associated with the responses to such diets have not been examined fully. Regional monoamine concentrations after IMB diet ingestion have received even less attention. Therefore a survey of several brain areas likely to have a role in the responses of rats fed IMB, basal (BAS) or corrected (COR) diets was undertaken in three studies, with the use of either threonine or isoleucine as the limiting amino acid.
The two limiting amino acids were selected because they are not precursors for the monoamine neurotransmitters and their metabolism differs, one from the other. In this way, we have an indication of whether the response may generalize to deficiencies of other essential amino acids. In addition, the three isoleucine IMB preparations, containing differing imbalanced proportions, were used to determine whether there is a dose-related response to the increasing imbalance in amino acid pattern. From previous work (Harper et al. 1970, Leung and Rogers 1987), we also anticipated that the time frame for observing a significant anorectic response to a mild by IMB diet would be longer than that for the more severely imbalanced diets. This gave us an opportunity to study the neurochemical responses both before and after a significant decrease in IMB diet intake took place. Our goals were to provide a description of the neurochemical systems that may be involved in the initial responses to such feeding and to postulate a neural network that may mediate these responses.
In our first two papers from this survey (Gietzen et al. 1986 and 1989), we reported the results of threonine and mildly isoleucine IMB diets on concentrations of monoamines and amino acids in the following five brain areas: 1) the APC (formerly known as the anterior prepyriform cortex, or PPC), an area known to be essential for the initial anorectic response (Leung and Rogers 1971); 2) the anterior cingulate cortex, an area associated with the later adaptation phase of the responses to IMB diets (Meliza et al. 1983); and three hypothalamic areas that may have roles in feeding, i.e., 3) the ventromedial hypothalamus, 4) the lateral hypothalamus, and 5) the paraventricular nucleus. Tissues were taken from rats killed 2.5 or 3.5 h after IMB diet introduction, i.e., around the time of the initial reduction in food intake. The samples from the remaining nine brain areas studied in the first two surveys were analyzed previously, but the data have not yet been reported. The third, more recent study, used two additional increasing levels of isoleucine IMB diets in a dose-response paradigm. This report continues our description of the neurochemical changes in the brain under amino acid limitation and provides a summary of the findings in the several brain areas from all three studies.
The results yield an antomical pattern of neurochemical responses to dietary amino acid limitation that should aid in developing hypotheses for the neural network(s) subserving the brain's responses to the challenge of amino acid limitation.
| |
MATERIALS AND METHODS |
Animals.
Male Sprague-Dawley rats (n = 78) weighing ~150 g at the beginning of each experiment were housed individually in hanging wire cages at 22 ± 2°C with lights on from 2400 to 1200 h. They were allowed to adapt to the cages, light cycle and the appropriate BAS diet for 2 wk before each experiment. The protocol was approved by the UC Davis Animal Use and Care Committee, and complied with NIH guidelines for laboratory animal care.
Diets.
There were three studies, based on two different growth limiting amino acids, threonine in the first experiment and isoleucine in the second and third experiments. In Experiment 2, the IMB diet was a mildly isoleucine-imbalanced diet (MILD), whereas in Experiment 3, the IMB proportions were raised, yielding moderately (MOD) and severely IMB (SEV) formulations in a dose-response pattern. Powdered diets with L-amino acids as the protein source and isoleucine as the limiting amino acid have been described previously (Hammer et al. 1990b); diets in which threonine was the limiting amino acid, as used in Experiment 1, are summarized in Table 1. Briefly, the threonine BAS diet contained 12.7 g/100 g amino acids, and the isoleucine BAS diet contained 11.7 g/100 g amino acids. The imbalanced diets were made by adding a mixture of all of the indispensable amino acids, except the one to be limiting, to the appropriate BAS diet. The indispensable amino acid mixtures added to cause the imbalance contained 4.93% (% of the diet as amino acids) for the threonine IMB and 9.86% for the MOD and SEV isoleucine IMB diets. The MOD was the same as SEV, but the MOD contained an added 0.5% of isoleucine such that the imbalance was partially, but not fully, corrected. The COR diets used in Experiments 1 and 2 consisted of the IMB diet plus enough of the limiting amino acid (0.4 and 0.5% for the threonine and MILD preparations, respectively) to correct the imbalance. When amino acids were added to the diet, the carbohydrate fraction was reduced proportionately. All diets contained the necessary vitamins and minerals with starch and sucrose (2:l, wt/wt) as the carbohydrate and 5% corn oil as the fat source.
Brain areas.
The anatomical strategy for these experiments was to examine all of the major brain areas that may be involved in the feeding responses to IMB diets. The regions fall into the following categories: 1) all of the areas found in earlier lesion studies to have a role in the feeding responses to IMB or amino acid-deficient diets (reviewed in Gietzen 1993, Leung and Rogers 1987), i.e., APC, anterior cingulate cortex, amygdala, hippocampus, septum and ventral tegmental area; 2) an additional hypothalamic area revealed recently in c-fos mapping studies (Wang et al. 1996a), i.e., the dorsomedial nucleus of the hypothalamus; 3) areas associated with the taste pathway, i.e., the nucleus of the solitary tract, parabrachial nucleus and central nucleus of the amygdala; 4) an area that has been implicated in conditioned taste aversion, along with the amygdala and parabrachial nucleus, the area postrema; 5) the classical hypothalamic areas associated with feeding, i.e., lateral hypothalamus, ventromedial hypothalamus and paraventricular nucleus, as before; and 6) the nuclei containing the major cell bodies of the norepinephrine and serotonin systems, i.e., the locus ceruleus and raphe nuclei, respectively.
Analyses.
The brains were removed within 45 s, frozen in liquid nitrogen and stored at 
80°C until analysis for the concentrations of monoamines, metabolites and amino acids. Transverse sections, ~1 mm thick, were cut from frozen brains and placed on slides over dry ice. The anterior cingulate cortex, APC, lateral hypothalamus and ventromedial hypothalamus were dissected with a scalpel as described (Gietzen et al. 1986 and 1989). We used the micropunch technique of Palkovitz (1973) to dissect the remaining brain areas. Monoamines and metabolites were assayed within 3-6 mo of tissue collection. We have found that values for these substances are stable for 6-9 mo, if the preparations are kept at 80°C. Sample preparation and HPLC with electrochemical detection (HPLC-EC) for analysis of monoamines and metabolites were as described previously (Gietzen et al. 1986). Briefly, the samples were sonicated in 0.1 mol/L perchloric acid medium containing an internal standard. They were centrifuged at 15,000 × g in a microcentrifuge housed in a cold room, filtered and injected onto the HPLC column. Identification of monoamines and metabolites was made in the chromatographic profile by peak conformation and retention time compared with an external standard containing norepinephrine, dopamine, serotonin and the serotonin metabolite, 5-hydroxyindoleacetic acid (5-HIAA). The levels of free amino acids were determined in aliquots of the same samples by automated amino acid analysis (Beckman 7300, Palo Alto, CA). Blood was collected into chilled heparinized tubes, centrifuged at 15,000 × g, and the plasma was analyzed for amino acids by automated amino acid analysis (Gietzen et al. 1986). There were six rats per group in Experiment 1 and seven rats per group in Experiments 2 and 3. However, a sample was occasionally lost in processing, particularly when there were small amounts of tissue, as in the micropunched brain areas. This resulted in smaller sample sizes for some values, as indicated in the footnotes to the tables and figures.
Experiment 1. The threonine study was designed to determine the neurochemical changes in discrete brain areas taken shortly after a measurable depression in food intake by the IMB group, from rats fed either a threonine IMB, BAS or COR diet. For 3 d before the experiment, the BAS diet was removed at 2400 h, the time the lights were turned on. The preweighed food cups were returned at 1200 h, (the time of lights-off), giving the rats access to the diet only during the dark phase. Thus, the rats were food restricted during the light phase, when they normally eat very little. This was done to synchronize the first meal of the dark phase. Food intake was measured at 1300, 1400 and 1500 h on the 3 d before the experiment by taking the difference in weight of the food cup before and after each interval and correcting for spillage. On the day of the experiment, the rats (n = 6 rats per group) were given either fresh threonine BAS, threonine IMB or COR diet. Food intake was again measured hourly. After we found that the rats in the IMB group had reduced their intake significantly by 2 h after presentation of the diets, the rats were decapitated at 2.5 h. At that time, cumulative food intake was 2.1 ± 0.3 g in the IMB group, compared with 4.7 ± 0.8 g in the COR group and 3.0 ± 0.3 g in the BAS group (IMB intake was less than that of the other two groups, P 
0.05).
Experiment 2. To collect tissues before the onset of the feeding depression, we used a less severely imbalanced isoleucine diet, MILD (Hammer et al. 1990b). The purpose of this study was to evaluate changes preceding the anorectic response, which may be associated with the induction of the anorexia, rather than resulting from it. Food intake of the rats offered the isoleucine BAS, MILD or COR diet was similar in all three groups (n = 7 rats per group) at the time they were killed, 3.5 h after diet presentation (intakes were as follows: IMB group, 5.9 ± 0.5 g; BAS group, 6.3 ± 0.8 g; and COR group: 5.4 ± 0.8 g, P > 0.05). A parallel group of seven rats reduced their intake of the MILD IMB to 52.4% of control (P <0.05) by 6 h after presentation of the diet. Thus, in the second experiment, the brains were taken approximately midway between diet presentation and the onset of the food intake depression; thus the intake of nutrients, including amino acids, differed only according to the diet composition. Otherwise, the animal treatment, tissue preparation and analysis were the same as in Experiment 1. An additional group of 11 rats was treated identically and blood was taken from the tail vein at 3.5 h after diet presentation. Plasma amino acid concentrations were determined in aliquots of plasma treated with sulfosalicylic acid (30 g/L, 50:50 v/v) and centrifuged at 15,00 × g in a microcentrifuge housed in a cold room before automated amino acid analysis of the supernatants (see above). Plasma values are expressed as micromoles per liter.
Experiment 3. In this study, we again used isoleucine as the limiting amino acid, but the severity of the imbalance was increased to determine whether there would be a dose-response pattern in the neurochemical measurements. The three diet groups (n = 7 rats per group) were BAS, MOD and SEV. The rats were decapitated 2.5 h after diet presentation. At that time, food intake was lower in both IMB groups relative to the basal group, as expected (BAS, 5.23 ± 0.43 g, MOD, 3.23 ± 0.36 g and SEV, 3.00 ± 0.45 g, P < 0.05 between BAS and either IMB group); food intake did not differ between the two IMB groups.
The same brain areas were collected as in Experiments 1 and 2, except that the dorsomedial hypothalamus was included, and the dissection of the amygdala was restricted to a punch in the area lesioned previously by Leung and Rogers (1973). These areas were punched, along with the remainder listed above. There were 14 brain areas used in the first two experiments, and 15 brain areas used in Experiment 3. Also, because plasma amino acid concentrations had not been published previously for rats fed these isoleucine diet formulations, we collected blood from the cervical wound into heparinized tubes during tissue collection in Experiment 3 and measured amino acids in plasma as described above for Experiment 2.
Statistical analysis.
Statistical significance was determined for levels of neurotransmitters or amino acids, within each brain area, with the use of a General Linear Models procedure (SAS/STAT Version 6.04, SAS Institute, Cary, NC) for ANOVA, Bonferroni, Bartlett's and least significant differences tests. If the variances were heterogeneous according to Bartlett's test, the data were log transformed before using the ANOVA procedure (Snedecor and Cochran 1967). For the neurotransmitter and food intake data, post-hoc evaluation of differences between group means was done, after a significant ANOVA, using the least significant differences test (Snedecor and Cochran 1967). The amino acid values for plasma and brain were also subjected to the Bonferroni correction for multiple comparisons (Miller 1981). Significance was assumed at P 0.05.
| |
RESULTS |
Plasma amino acids.
Values for plasma amino acids after feeding the threonine-limited diets used in Experiment 1 have been described in detail previously (Leung et al. 1968). After rats consumed the isoleucine IMB or COR diets used in Experiments 2 and 3, the concentrations of many of the amino acids in plasma differed, reflecting dietary amino acid levels (Tables 2 and 3), as expected from previous reports for IMB (e.g., Leung et al. 1968). In Experiment 2, all of the essential amino acids, except isoleucine and valine, were higher in the groups fed IMB or COR diets than in the BAS group, reflecting differences in the dietary amino acid composition. Valine was also higher in the COR group than in the BAS-fed group, but in the MILD group, the concentration of valine did not differ from that in the BAS group. The dispensable amino acids, taurine and glycine, did not differ among the diet groups, whereas aspartate, alanine, glutamate and ornithine were lower, and glutamine and serine were higher, in the IMB and COR groups than in the BAS group (P < 0.05, Table 2). The only plasma amino acids that differed significantly among all three groups were histidine, in which BAS < IMB < COR, and isoleucine, with IMB < BAS < COR (P < 0.05, Table 2). This pattern for isoleucine, the limiting amino acid, was expected because the biochemical changes that occur with ingestion of an IMB diet cause a reliable and pronounced reduction in the concentration of the limiting amino acid (Harper et al. 1970), and the concentration of isoleucine in the COR diet was higher than that in the BAS diet (Hammer et al. 1990b). Isoleucine in the MILD group was 42% of its value in the BAS group (P < 0.003) and 14% of the value in the COR group (P 0.0001) by 3.5 h after the initial introduction of the diets.
View this table:
[in this window]
[in a new window]
|
Table 2.
Concentrations of amino acids in plasma of rats after consuming isoleucine basal (BAS), mildly imbalanced
(MILD) or corrected (COR) diets for 3.5 h1
|
|
View this table:
[in this window]
[in a new window]
|
Table 3.
Concentrations of amino acids in plasma of rats after consuming isoleucine basal (BAS), moderately imbalanced (MOD) or severe imbalanced (SEV) diets for 2.5 h1
|
|
In Experiment 3, plasma concentrations of aspartate and glutamine did not differ significantly among the diet groups, nor did methionine, valine and leucine, which were among the amino acids added to the BAS diet to form the IMB diet (Hammer et al. 1990b). The other 15 amino acids differed significantly among the treatments (P < 0.05), as indicated in the table. Although the indispensable amino acids were, or tended to be, higher in the IMB groups than in the BAS group, the only essential amino acid that was significantly lower in the IMB groups again was isoleucine, the limiting amino acid, which was present at only 19% of the BAS value in the SEV group (P 0.0001) and at 30% of the BAS value in the MOD group (P 0.04). The other amino acids that were lower in these IMB plasma samples were the dispensable amino acids serine, glutamate, glycine, alanine, ornithine and arginine.
To evaluate the relative concentrations of the amino acids that compete for isoleucine transport into the brain, we calculated the ratios of each of the individual large neutral amino acids (LNAA) to the sums of the LNAA. As can be seen in Tables 2 and 3, threonine, tyrosine and phenylalanine were higher relative to the sum of the LNAA in the groups, with the ratio of threonine:LNAA being highest. In the MILD group, threonine:LNAA was twice as high in both COR and IMB groups, and the combined MOD and SEV groups showed a threefold greater threonine:LNAA ratio than the BAS group. Again, the MOD group did not differ from the SEV group. The amino acids for which this ratio was lower in both experiments included the branched-chain amino acids, valine, leucine and isoleucine. As expected, given its role as the limiting amino acid, the isoleucine:LNAA ratio was the lowest of all of the amino acid ratios. In Experiment 2, the isoleucine:LNAA ratio was 19% of BAS, followed by valine, at 75% of BAS. In Experiment 3, the ratio of isoleucine:LNAA was 19% of BAS in the MOD group and 12% of BAS in the SEV group, followed by the other branched-chain amino acids, valine (41 and 33% of BAS, respectively) and leucine (60 and 38% of BAS, respectively). Because threonine can also use the small neutral amino acid transport system, we also calculated the threonine:small neutral amino acid ratios in Experiments 2 and 3. This value was significantly elevated in the IMB groups as follows: in Experiment 2, BAS, 0.05 ± 0.003, MILD and 0.27 ± 0.01, COR, 0.22 ± 0.03, P < 0.001; in Experiment 3, BAS, 0.03 ± 0.002, MOD, 0.44 ± 0.07 and SEV 0.36 ± 0.05, P < 0.0001.
The tryptophan:LNAA ratio was higher in rats that consumed MILD than in controls (Table 2) where food intake did not differ between groups; this ratio was lower in the IMB groups of Experiment 3 (Table 3) where the hypophagic response was seen.
The limiting amino acid in brain.
The limiting amino acid concentrations for the various brain areas from Experiment 1 are given in Table 4, and those from Experiments 2 and 3 are listed in Table 5. An earlier report of concentrations of threonine in the brain areas from Experiment 1 was included in Gietzen (1993). However, additional rats, inclusion of results for the dorsomedial hypothalamus and more comprehensive data analyses have made a re-evaluation of these results necessary; the most recent results for this variable in threonine-limited diets are listed in Table 4.
View this table:
[in this window]
[in a new window]
|
Table 4.
Concentrations of L-threonine in brain areas of rats after consuming a threonine basal (BAS), imbalanced (IMB) or corrected (COR) diet for 2.5 h1
|
|
View this table:
[in this window]
[in a new window]
|
Table 5.
Concentrations of L-isoleucine in brain areas of rats after consuming isoleucine basal (Ile BAS), mildly imbalanced (MILD) or corrected (COR) diet for 3.5 h or Ile BAS, moderately imbalanced (MOD) or severely imbalanced (SEV) diet for 2.5 h1
|
|
Clearly, the limiting amino acid was not uniformly reduced throughout the brain, consistent with the observations of Hawkins et al. (1982). Previous observations have reported that the limiting amino acid is decreased in whole brain (Peng et al. 1972). Although amino acids in whole brain were not measured in these studies, the sums of the limiting amino acid values in the brain areas investigated here showed a lowering of the level of the limiting amino acid in the IMB groups to 13.4% of that in the BAS group in Experiment 1, in which threonine was limiting, and to 4.1, 42 and 52% of BAS in Experiments 2 and 3, in which isoleucine was mildly, moderately or severely limiting, respectively. Fernstrom and Fernstrom (1995) used a modified IMB protocol with mixtures devoid of tyrosine and phenylalanine (very severely imbalanced) and saw decreased tyrosine in the two brain areas they studied, hypothalamus and cerebral cortex.
In Experiment 1, threonine was lower in the IMB group than in the BAS group, in the anterior cingulate cortex and APC, as previously reported (Gietzen et al. 1986). In addition, threonine was significantly lower in the parabrachial nucleus in the IMB group relative to the COR group. These were the only brain areas in which threonine was lower in the IMB-fed rats than in either BAS- or COR-fed rats (Table 4). In the dorsomedial hypothalamic nucleus, examined in a similar but separate study with threonine as the limiting amino acid, threonine was slightly, but not significantly, less in the IMB group 2.5 h after the beginning of the feeding period.
In Experiments 2 and 3, the limiting amino acid, isoleucine, was significantly lower again in the APC from the SEV group, although not in the MOD group, compared with the BAS group (Table 5). Isoleucine was also lower in the dorsomedial hypothalamic nucleus, but only in the SEV group. In the anterior cingulate cortex, the limiting amino acid tended to be lower, but in contrast to Experiment 1, these reductions were not significant in either Experiment 2 or 3.
Thus, the brain area most consistently showing a significant reduction of the limiting amino acid in the IMB groups, throughout all three studies, was the APC.
Nonlimiting amino acids in the brain.
In contrast to the very few reductions in the limiting amino acid, six of the nonlimiting amino acids as well as the sum of the excitatory amino acid neurotransmitters, glutamate and aspartate, were altered in 10 different brain areas of IMB diet-fed rats, P < 0.05 according to the Bonferroni test. There was just one observation each of a change for glutamine, which was only 69% of BAS in the parabrachial nucleus in the threonine IMB group, glycine, which was 59% of BAS in the ventromedial hypothalamus of the SEV group, and arginine, which was 45% of BAS in the threonine IMB group of Experiment 1. In addition, lysine was found at 306% of BAS only in the ventral tegmental area in the threonine COR group. We assume that only those amino acids for which more than two consistent findings were made across the three studies have potential importance in the mechanisms underlying the responses to IMB diets. Therefore, these four amino acids will not be discussed further. More likely to have physiologic relevance are the remaining amino acids for which several significant differences were seen, i.e., aspartate, glutamate (Table 6), tyrosine (Fig. 1) and threonine (Fig. 2).
View this table:
[in this window]
[in a new window]
|
Table 6.
Excitatory amino acid neurotransmitters in brain areas of rats in the imbalanced (IMB) diet groups that differed significantly from the respective basal (BAS) diet group1
|
|
View larger version (25K):
[in this window]
[in a new window]
| Fig 1.
Concentrations of tyrosine (tyr) in brain areas after rats consumed (A) BAS, threonine basal diet; THR IMB, threonine-imbalanced diet; and COR, threonine-corrected diet in Experiment 1; (B) BAS, isoleucine basal diet; MILD, mildly isoleucine-imbalanced diet; and COR, isoleucine-corrected diet in Experiment 2; (C) BAS, isoleucine basal diet; MOD, moderately isoleucine-imbalanced diet; SEV, severely isoleucine-imbalanced diet in Experiment 3. Tissues were collected at 2.5 h (A and C) or 3.5 h (B). Brain areas were as follows: AMY, amygdala; APC, anterior piriform cortex; LH, lateral hypothalamus; LC, locus ceruleus; VMH, ventromedial hypothalamus. Bars and error markers indicate means and SEM, n = 4-7 rats. Significant differences from the appropriate BAS diet group, P < 0.05, indicated by an asterisk over the bar, were determined after a significant overall ANOVA with post-hoc least significant means tests, corrected for multiple comparisons using the Bonferroni correction.
|
|
View larger version (42K):
[in this window]
[in a new window]
| Fig 2.
Concentrations of threonine (thr) in rat brain areas for Experiments 2 and 3. (A) BAS, isoleucine basal diet; MILD, mildly isoleucine-imbalanced diet; COR, isoleucine-corrected diet in Experiment 2; (B) BAS, isoleucine basal diet; MOD, moderately isoleucine-imbalanced diet; SEV, severely isoleucine-imbalanced diet in Experiment 3. Tissues were collected at 2.5 (B) or 3.5 h (A). Brain areas were as follows: AMY, amygdala; APC, anterior piriform cortex; DMH, dorsomedial hypothalamus; LH, lateral hypothalamus; LC, locus ceruleus; NTS, nucleus of the solitary tract; RAPH, raphe nucleus. Bars and error markers indicate means ± SEM , n = 4-7 rats. Significant differences from the appropriate BAS diet group, P < 0.05, indicated by an asterisk over the bar, were determined after a significant overall ANOVA with post-hoc least significant means tests, corrected for multiple comparisons using the Bonferroni correction.
|
|
Glutamic acid and aspartate are excitatory amino acid neurotransmitters, but are not nutritionally essential amino acids. Glutamate, aspartate or their sum was significantly lower than BAS in the dorsomedial hypothalamus, parabrachial nucleus and ventromedial hypothalamus in a total of five instances in the threonine IMB, and the MOD and SEV isoleucine IMB groups (Table 6). Because these are not dietary essential amino acids (the dietary precursor is glucose), they may simply be reduced as a secondary result of the decreased food intake in these diet groups. However, there may be implications for neurotransmission in these data (see discussion below).
Tyrosine, the precursor of the catecholamines, was greater, relative to its competitor amino acids in the plasma and in five brain areas, from IMB groups only; the lateral hypothalamus was the only area in which the increase was consistent across all IMB groups (Fig. 1). There were no significant increases in tyrosine in any brain areas of the COR groups. There were no consistent changes in tyrosine together with norepinephrine in the various brain areas (compare Figs. 1 and 3) that suggest a general role for this amino acid as affecting overall norepinephrine concentrations in the brain in the responses to IMB diets.
View larger version (21K):
[in this window]
[in a new window]
| Fig 3.
Norepinephrine (NE) percentage of change from corresponding value from the respective control basal diet (BAS) group) in brain areas from rats fed one of the following diets: THR, threonine-imbalanced diet; MILD ILE, mildly isoleucine-imbalanced diet; MOD ILE, moderately isoleucine-imbalanced diet; or SEV ILE, severely isoleucine-imbalanced diet for 2.5 (THR, MOD ILE and SEV ILE) or 3.5 h (MILD ILE) after 2 wk prefeeding the BAS diet. Brain areas were as follows: LC, locus ceruleus; VMH, ventromedial hypothalamus; TEG, ventral tegmental area; APC, anterior piriform cortex. Bars and error markers indicate group means ± SEM, n = 4-7 rats. Significance of differences from the respective BAS diet group, indicated by an asterisk over the bars, was determined, after a significant (P < 0.05) ANOVA, by least significant difference testing.
|
|
Finally, threonine was lowest in the threonine IMB group, as described above, when it was the limiting amino acid. By contrast, in the isoleucine-limited diets, threonine was markedly elevated in seven brain areas after IMB diet ingestion, but also in the locus ceruleus after rats ate the COR diet (Fig. 2). Threonine was significantly higher in the IMB than BAS groups in a total of 14 instances; in four other areas, a similar pattern was seen, although with our stringent statistical criteria, significance was not reached in those areas. These results are consistent with the dramatic increase of threonine in plasma (Tables 2 and 3) and its relative increases in both the transport groups for the LNAA (doubled, Tables 2 and 3) and small neutral amino acids (three- to fourfold higher ratio in Experiment 2, over tenfold in Experiment 3, values given above). The present observations demonstrating higher plasma and brain levels of threonine and tyrosine after IMB feeding are consistent with the work of Tews et al. (1987), showing that amino acid transport is markedly affected by the level of amino acid nutriture from the prefeeding diet and the other amino acids provided by the IMB diet. Increased levels of threonine and tyrosine have been reported in whole brain after IMB diets (Sanahuja and Harper 1963). The effects of such large increases in plasma and brain threonine in the isoleucine IMB groups are unknown, although they do not seem to correlate with the feeding responses.
Threonine was not increased in the threonine IMB diet study, as expected, given that it was the dietary limiting amino acid. Yet, the feeding depression with threonine-IMB was similar to that seen in the MOD and SEV isoleucine IMB study. Also, threonine was increased in rats fed the COR diet (see locus ceruleus (LC) in Fig. 2), which animals consume readily (Harper et al. 1970, Leung and Rogers 1987, see results for food intake for the COR group in Experiment 1). Therefore, we suggest that these increases, although significant, may not be helpful in determining the mechanisms underlying the feeding responses of interest in this model.
Monoamine neurotransmitters in brain.
Among the monoamines and metabolites that were measured, we found significant changes for the following: norepinephrine; the ratio of the dopamine metabolite, dihydroxyphenylacetic acid, to dopamine; serotonin; and the serotonin metabolite, 5-hydroxyindole acetic acid (5-HIAA).
Norepinephrine was lower in the APC and higher in the ventromedial hypothalamus in both IMB groups, compared with the BAS groups in Experiment 3, replicating our previous reports (Gietzen et al. 1986 and 1989). In addition, we noted higher concentrations of norepinephrine in the locus ceruleus and lower norepinephrine in the ventral tegmental area in the IMB groups (Fig. 3). The locus ceruleus contains the cell bodies for norepinephrine within the brain stem. The ventral tegmental area, which we took in our dissection, was the same as that lesioned by Leung and Rogers (1980); it contains the ascending noradrenergic fibers of the central tegmental tract (Lindvall and Bjorklund 1978).
After threonine IMB was fed, the dopamine turnover ratio, dihydroxyphenylacetic acid:dopamine, was <10% of the value for the BAS group in the dorsomedial hypothalamus (9.4 ± 1.5% of BAS, P < 0.008). This ratio tended to be less in the IMB groups from Experiment 3 as well, but did not reach significance. There were no other indications of dopaminergic activity in either IMB or COR groups in any brain area.
In the threonine IMB study, the serotonergic indicators, serotonin itself and its metabolite, 5-HIAA, were affected only in the nucleus of the solitary tract. In this well-known relay in the taste pathway, 5-HIAA was lower by almost 50% in both the IMB and COR groups (P = 0.05), indicating a similar response to both new diets, the balanced COR and the IMB. In the isoleucine studies (Experiments 2 and 3), serotonin was affected in the following four other brain areas: raphe nucleus, the site of serotonin cell bodies, and hippocampus, parabrachial nucleus and anterior cingulate cortex (Table 7). In the raphe, serotonin was higher in all IMB groups (P = 0.05 vs. COR group in Experiment 2 and P = 0.02 in Experiment 3), and 5-HIAA was higher in the MILD group (P = 0.0006) and in the SEV group (P < 0.03). In hippocampus, in the MOD and SEV groups, serotonin did not differ, but the metabolite, 5-HIAA, was higher in both MILD (P 0.03) and SEV (P 0.04) groups (Table 7). In parabrachial nucleus, the MOD and SEV diets were associated with a dose-related increase in serotonin concentrations (P 0.0001). Finally, in the anterior cingulate cortex, there was a dose-response pattern of increases in 5-HIAA in Experiment 3, which reached significance in the SEV group (P 0.004).
View this table:
[in this window]
[in a new window]
|
Table 7.
Serotonin and its metabolite, 5-hydroxyindoleacetic acid (5-HIAA), in rat brain areas after isoleucine-limited diets1
|
|
| |
DISCUSSION |
These results confirm and extend our previous findings of neurochemical responses to IMB feeding. Taken together with findings using a variety of techniques from this and other laboratories, the results of these neurochemical surveys suggest that several brain areas may form a pair of circuits to mediate the recognition and learned aversion phases of the responses to acute amino acid deficiency.
The first phase, recognition.
The first "initial recognition" phase appears to be mediated by the APC (Leung and Rogers 1971). The APC may be the most chemically sensitive area of the brain (Piredda and Gale 1985); neural activity in the APC is stimulated by IMB and amino acid-devoid diets to such an extent that kindling (hyperexcitability resulting from repeated stimulation) occurs (Gietzen et al. 1996). In the piriform cortex of IMB-fed rats, we consistently saw a significant decrease in the concentration of the limiting amino acid (Gietzen et al. 1986, the present results). A review of the results in Tables 3 and 4 shows that there were no other brain areas in which the limiting amino acid was so consistently reduced after IMB ingestion. Moreover, 0.5-µL injections of 1 or 2 nmol of the appropriate limiting amino acid made directly into the APC significantly increased intake of an IMB diet (Beverly et al. 1990, Monda et al. 1997). To our knowledge, this effect has been seen only in the APC; injections into the amygdala or 2-mm posterior to the APC were ineffective (Beverly et al. 1990). It would thus appear that the primary neural stimulus provided by IMB diets is a decrease of the limiting amino acid in a susceptible brain area.
Just how this event stimulates neural activity in the APC is not clear, but subsequent activity in the catecholamine system(s) appears to play an important role. A decrease in norepinephrine in Experiment 3, confirming our earlier observation (Gietzen et al. 1986), was associated with downregulation of the 
2 noradrenergic receptor (Gietzen and Jhanwar-Uniyal 1996), decreased concentrations of cAMP (Gietzen 1993) and an apparent increase in dopamine metabolites in the interstitial spaces (Wang et al. 1997). In addition, intra-APC injections of the 2 adrenergic autoreceptor agonist, clonidine, increased early intake of an IMB (Gietzen and Beverly 1992).
Hindbrain areas 
the ascending noradrenergic system.
In contrast to the decreases seen in the APC, the concentrations of both norepinephrine and tyrosine were increased in the locus ceruleus, which contains the cell bodies of origin for the norepinephrine system. Thus, the precursor of norepinephrine clearly was not limiting in the cell bodies. These observations may be consistent with increased uptake and synthesis of the transmitter at the level of the cell bodies in response to activity in the terminal fields, including the APC. Both long and short loop feedback mechanisms exist in the control of the locus ceruleus, as reviewed recently (Cooper et al. 1996); thus a precedent for such a concept has already been set. Whether the increased concentration of norepinephrine seen in the locus ceruleus may be due to decreased release or increased synthesis in that region remains to be determined. Of course, there are well-described limitations to any mechanisms that may be suggested solely on the basis of concentrations of substances measured in homogenates of brain tissue. Nonetheless, correlative studies using several different approaches, such as those listed above for the APC, along with tissue concentrations of transmitters, precursors and/or metabolites, can lend support for mechanistic hypotheses. At the least, differences in neurochemical parameters in a brain area point to altered activity in that area.
Along the ascending pathways, noradrenergic fibers in the ventral ascending system at the level of the rostral pons were the target of electrolytic lesions made by Leung and Rogers (1980). This area contains the major ascending pathway through which the noradrenergic fibers travel, as part of the central tegmental tract, caudal to the dopaminergic (A8-A10) cell bodies and rostral to the locus ceruleus (Lindvall and Bjorklund 1978). The lesions caused hyperphagia in rats fed balanced diets, but significantly decreased intake of threonine-imbalanced or -devoid diets (Leung and Rogers 1980). In addition, the ventral tegmental area was the only brain region, other than the APC, in which the concentration of norepinephrine was lower after ingestion of isoleucine IMB (Fig. 3). A lower level of norepinephrine in the ventral tegmental area suggests that release from the somata may have been inhibited, and content in the pathway subsequently decreased. Neither of the precursors, tyrosine or dopamine, was altered in the tegmental samples after IMB diets; thus these changes are not likely to be the result of altered precursor availability.
Medial hypothalamic areas.
The hypothalamus contains terminals of the norepinephrine neurons that originate in the locus ceruleus and travel in the central tegmental tract, as noted above (Lindvall and Bjorklund 1978). In addition, both dorsomedial and ventromedial areas of the hypothalamus have reciprocal projections from the APC as shown by retrograde tracing techniques (Haberly and Price 1978). Rats with lesions of the ventromedial hypothalamus ate more of a mildly isoleucine IMB diet (Leung and Rogers 1970) or a histidine IMB diet (Nassett et al. 1967), but not of a severely isoleucine IMB diet (Leung and Rogers 1970) and also not of a threonine IMB diet (Scharrer et al., 1970). Among the hypothalamic areas tested, we again saw increased concentrations of norepinephrine and its precursor, tyrosine, in the ventromedial area (Gietzen et al. 1989, the present results). This is consistent with inhibition of norepinephrine release in the ventromedial hypothalamic area (historically considered the "satiety area") such that satiety might not be seen after ingesting IMB.
Recent c-fos (Wang et al. 1996a) and lesion (Bellinger and Gietzen 1995) studies also implicate the dorsomedial hypothalamic nucleus in the earliest phases of the response. Electrolytic lesions of the dorsomedial hypothalamus resulted in full restoration of feeding during the first 3 h of IMB exposure (Bellinger and Gietzen 1995). In addition, the concentration of the limiting amino acid in the dorsomedial hypothalamus was significantly decreased in the SEV group of Experiment 3; the dorsomedial nucleus of the hypothalamus was one of the very few brain areas in which the limiting amino acid was decreased after feeding any of the IMB diets (Tables 3, 4). Also, the lower levels of glutamate and aspartate seen in the dorsomedial hypothalamus in IMB groups (Table 6) may indicate increased release of excitatory amino acids in this area; both the threonine IMB group in Experiment 1 and the isoleucine IMB groups of Experiment 3 had lower levels of excitatory amino acids than the BAS group. Our measurements did not differentiate between intra- and extracellular levels of substances, nor between neural and glial tissue. Moreover, because these are not dietary essential amino acids (their precursor is glucose), they may simply be reduced as a secondary result of the decreased energy intake in the IMB groups. The level of energy depletion that can occur in animals in 2.5-3.5 h is not known, but we saw no differences in plasma glucose levels among similar groups in a previous study (Gietzen and Jhanwar-Uniyal 1996). Nevertheless, the role of excitatory amino acids in the dorsomedial hypothalamus has not yet been explored in this model and may deserve attention; utilization of amino acid neurotransmitters was altered in several brain areas in a conditioned emotional response paradigm (Lane et al. 1982). Also likely to be important in the role of the dorsomedial hypothalamus in the responses to IMB diets is the dopamine system, in which we saw a 90% lower metabolite/transmitter ratio. Because this ratio is an indicator of activity in a transmitter system, such a large decrease could suggest a reduction of dopaminergic activity in the dorsomedial hypothalamus. Clearly, from the electrolytic lesion studies, the dorsomedial nucleus contains neural fibers that are important in the responses to IMB diets (Bellinger and Gietzen 1995). Further studies of the dorsomedial hypothalamus are in progress.
View larger version (29K):
[in this window]
[in a new window]
| Fig 4.
Postulated neural circuit for recognition of amino acid deficiency. APC, anterior piriform cortex, the amino acid chemosensor (Leung and Rogers 1971); CeA, central nucleus of the amygdala; HYPO, hypothalamus; LC, locus ceruleus, site of norepinephrine cell bodies; PBN, parabrachial nucleus, relay in the taste pathway; RAPHE, raphe nucleus, site of serotonin cell bodies; and CTA, conditioned taste aversion, a secondary response, seen after recognition of the amino acid deficiency and hypophagia. Solid lines indicate responses mediated by the APC, directly to the hypothalamic area (lower left drawing) or via the locus ceruleus (lower right). Dotted lines indicate circuit mediated by serotonin, via the raphe nucleus. Filled stars indicate neural relays. Filled circles (lower left drawing) indicate brain areas that may mediate the responses listed underneath, CTA or hypophagia. Sketches adapted from Cuello (1983).
|
|
The second phase, conditioned taste aversion the taste pathway and serotonin.
A secondary phase in the responses is the development of a conditioned taste aversion to the IMB diet (Booth and Simson 1971, Leung and Rogers 1987, Meliza et al. 1981). The brain regions that have been associated with this phase of the response by lesion and c-fos mapping studies are the central nucleus of the amygdala (Meliza et al. 1981, Wang et al. 1996b) and the parabrachial nucleus (Norgren et al. 1996). In both of these regions, which are integral to the taste pathway (Norgren 1983), lesions render the animals unable to reject IMB diets, but they also fail to form a conditioned taste aversion when tested using lithium chloride.
In the amygdala, there was just one significant neurochemical change, i.e., a dose-response pattern in the increases in threonine in Experiments 2 and 3, in which isoleucine was the limiting amino acid. Our recent c-fos study showed increased neural activity in the central nucleus of the amygdala, but not until 3 h after introduction of the IMB diet (Wang et al. 1996b). Clearly, the amygdala appears to be involved in the responses to IMB diets later than the APC and the hypothalamic areas.
In contrast to the paucity of findings in the amygdala, there were several neurochemical changes in the parabrachial nucleus. The limiting amino acid was significantly decreased in Experiment 1, and we saw a dose-response pattern of increases in serotonin for this area in Experiment 3 (Table 7). We also noted increases in the excitatory amino acid glutamate and in the sum of the excitatory amino acids, glutamate plus aspartate, in Experiment 1 (Table 6). Note that we found no neurochemical changes in the parabrachial nucleus in Experiment 2 in which the tissues were taken before the reduction in food intake occurred. Taken together, the results support the concept that the parabrachial nucleus may be involved in the development of an aversion to IMB diets (Phase 2) earlier in the sequence of these responses than the amygdala. The present observations of increased serotonin in the parabrachial nucleus, along with blockade of the development of aversion to a diet devoid of amino acids by lesions in this area (Norgren et al. 1996), are consistent with our previous reports of a role for serotonin, perhaps at the serotonin3 receptor in the anorectic responses to IMB diets (Hammer et al. 1990a) and of blockade of aversion to an isoleucine IMB diet after treatment with serotonin3 receptor antagonists (Terry-Nathan et al. 1995).
We have no consistent evidence for a role of the nucleus of the solitary tract in selective responses to IMB diets from any of the present experiments, in spite of its important role in the taste pathway. Our c-fos results (Wang et al. 1996b) suggested that the nucleus of the solitary tract was activated similarly in both IMB and COR groups. It is important in this context to note that both the IMB and COR diets were new to the rats on the first experimental day. Behavioral tests have shown that rats cannot distinguish IMB from COR on the basis of taste or mouth feel (reviewed in Gietzen 1993). Thus we conclude that the nucleus of the solitary tract does not have a discriminating role in amino acid deficiency, apart from responding to novel diets. Rather, the parabrachial nucleus appears to be the first among these taste relays to respond selectively to the IMB diets.
The lateral hypothalamus.
The lateral hypothalamus was shown to be sensitive to iontophoretically applied amino acids in the work of Wayner et al. (1975), although lateral hypothalamic lesions do not interfere with the behavioral (anorectic) responses to IMB diets (Scharrer et al. 1970). Amino acids injected into the perifornical area (near the lateral hypothalamus) decreased food intake (Panksepp and Booth 1971). Neurons in this area also showed an increase in firing rate 30 min after injection of threonine into the APC in threonine-deficient rats (Monda et al. 1997). In addition, Ono and colleagues (reviewed in Torii et al. 1996) noted increased neuronal activity in ~5% of the cells in the lateral hypothalamic area, when a deficient animal was exposed to cues it had previously associated with repletion of the limiting amino acid. They also showed activation of the lateral hypothalamus with the use of magnetic resonance imaging 30-50 min after injecting the limiting amino acid peripherally (Torii et al. 1996). In the present results, we found higher concentrations of both tyrosine and threonine in the lateral hypothalamus in IMB groups. The APC projects directly to the lateral hypothalamus (Price et al. 1991), as does the locus ceruleus (Lindvall and Bjorklund 1978); future studies may clarify interactions of the lateral hypothalamus with other hypothalamic and extrahypothalamic areas involved in the responses to IMB diets.
In summary, the results presented here confirm and extend previous observations (Gietzen et al. 1986, Leung and Rogers 1971, Meliza et al. 1981, Wang et al. 1996a and 1996b), showing the importance of the APC and the amygdala in the initial and aversive phases of the responses to IMB diets, respectively. In addition, roles for several more brain areas are suggested. These findings appear to be consistent with other results for brain areas known to be associated with the behavioral (anorectic and aversive) responses to IMB diets, as reviewed above. We further postulate a pair of neural circuits (outlined in Fig. 4), one for norepinephrine, the other using serotonin; both would be affected at the level of their cell bodies after activation of the APC. Although likely to be an oversimplification, it is tempting to suggest that, in Phase 1, activation of the norepinephrine system by projections from the APC to the locus ceruleus could lead to increased neophobia, resulting from activity in the ventral tegmental area, along with inhibition of satiety systems in the ventromedial hypothalamus. Similar activation of the raphe nuclei could result in increased serotonergic activity in the parabrachial nucleus to initiate the conditioned taste aversion of Phase 2. Serotonergic activity in limbic areas associated with learning, such as the hippocampus, may contribute to the conditioned responses as well. These postulated circuits provide working hypotheses for further studies to elucidate just how the brain controls the feeding responses to dietary amino acid deficiency.
| |
ACKNOWLEDGMENTS |
The technical assistance of Dan Wong with the amino acid analyses and Blaine Lee with the monoamine analyses is gratefully appreciated.
![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
, **
Abstract
Introduction
Abstract
