| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Articles |
Departments of Medicine (L.J., J.A.M.) and Radiology (D.Z.), Harvard Medical School, Boston, Massachusetts 02115
Address all correspondence and requests for reprints to: Lauren Jacobson, Ph.D., Division of Endocrinology, Enders 4, Children’s Hospital, 300 Longwood Avenue, Boston, Massachusetts 02115. E-mail: Jacobson{at}A1.TCH.Harvard.Edu
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Because protein malnutrition may arise from a variety of conditions, including illness and starvation, that can themselves affect adrenocortical activity, the neuroendocrine and metabolic signals for malnutrition-associated increases in glucocorticoids are poorly defined. Numerous clinical studies of victims of malnutrition in underdeveloped countries have suggested that cortisol production increases, but in many cases the effects of malnutrition may have been confounded by concomitant bacterial or parasitic infections, and the relative contribution of calorie vs. protein malnutrition to increased cortisol production was unclear (5, 6). Glucocorticoid levels and secretion have also been reported to increase in malnourished experimental animals (7). However, most of these studies did not control for the reduced food intake displayed by many species given low protein or protein-free diets (8). Thus, it is also not clear whether these results are specifically attributable to protein or to general calorie deficiency. Furthermore, there is little information, except for one report that protein malnutrition increases corticotrope number (9), on the potential role of the brain and pituitary in driving increased glucocorticoid production during malnutrition. To address these issues, we have investigated the regulation of ACTH and the messenger RNA (mRNA) encoding its precursor, POMC, in mature rats subjected to dietary protein deprivation or caloric restriction (pair-feeding).
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Diet groups
The following diet groups were studied.
21% protein. Rats had free access to a nutritionally complete, powdered diet (Harlan-Teklad, Madison, WI) that supplied 21% of calories as protein (casein), 66% as carbohydrate (corn starch and sucrose), and 13% as fat (corn oil).
0% protein. Rats had free access to an isocaloric powdered diet (Harlan-Teklad, Madison, WI) in which carbohydrate was substituted for protein, supplying 87% of calories from carbohydrate and 13% from fat. Vitamin and mineral levels were equivalent in normal and protein-free diets.
Pair-fed. Rats were individually paired to a rat in the 0% protein group and given a restricted amount of normal (21% protein) diet equivalent to the weight of food eaten by the protein-deprived counterpart. Body weights were matched to within 5 g before assignment to diet groups. In preliminary experiments, rats given protein-free chow ate 25% fewer calories within the first day than their ad libitum intake of normal diet the day before. To minimize the lag time in matching caloric intake between protein-deprived and pair-fed groups, pair-fed rats in subsequent experiments were restricted to 75% of their ad libitum intake on day 1 and were pair-fed thereafter. Food was presented within 1 h of lights-off each day to avoid potential disruption of circadian hormone rhythms due to food restriction (10) in the pair-fed group. All rats were killed in the morning, 5 days after the beginning of experimental diets.
Endocrine groups
To control for changes in endogenous corticosterone secretion
during dietary manipulation, some rats were adrenalectomized and were
either left untreated or replaced with constant levels of
corticosterone via a sc pellet as previously described (11). In
preliminary experiments, a 100-mg 30% corticosterone-cholesterol
pellet was determined to provide approximately physiological
replacement in normally nourished 225-g rats, in that this dose
inhibited adrenalectomy-induced increases in morning ACTH without
decreasing thymus weight (12). As the physical presence of sc pellets
has not been found to affect ACTH regulation (Jacobson, L., unpublished
observations), sham-adrenalectomized and adrenalectomized nonreplaced
rats were not given placebo pellets. Rats had free access to normal
diet for 2 days of postsurgical recovery before beginning experimental
diet regimens. All adrenalectomized groups were given 0.9% saline to
drink after surgery. At death, adrenalectomy was assessed by visual
inspection of the abdominal cavity; one adrenalectomized rat given
normal diet was excluded from data analysis in Exp 2B on this
basis.
Experimental design
All experiments were completed within 4 h of lights on.
Exp 1: effect of dietary protein deprivation on basal and stress-induced plasma ACTH and corticosterone. Two days before dietary manipulations, rats were either sham-adrenalectomized (sham) or adrenalectomized and replaced with a 100 mg sc pellet of 30% corticosterone-cholesterol (ADX+30% B). They were allowed free access to 21% protein diet for 2 days to recover from surgery, and then underwent either normal, protein-free, or pair-feeding dietary regimens for 5 days. On the morning of day 5, rats were decapitated for blood and tissue collection, either without stress or immediately after a 30-min restraint stress in wire mesh tubes.
Exp 2: effect of controlling changes in corticosterone levels during dietary protein deprivation on plasma ACTH and anterior pituitary POMC mRNA. To control for the increased plasma corticosterone observed after protein deprivation in ADX+30% B rats, rats were adrenalectomized and given either no replacement (ADX+0) or a 100-mg pellet containing different amounts of corticosterone (ADX+B). Initially (Exp 2A), rats were given 0%, 15%, or 30% corticosterone pellets. Sham-adrenalectomized rats that were given the 21% protein diet throughout served as controls for ACTH regulation under normal diet conditions. Anterior pituitary RNA preparations from these rats were unsatisfactory, probably due to poor quality isopropanol used in precipitation. Therefore, to assess POMC mRNA levels and to extend the dose range of corticosterone replacement, the experiment was repeated (Exp 2B), including an additional group of ADX rats given 60% corticosterone pellets and normal diet to match the elevated corticosterone levels in the protein-deprived, ADX+30% B group. Diets were the same as in Exp 1; rats were killed without stress for plasma and tissue collection in the morning after 5 days of experimental diets.
Assays and statistics
Plasma chemistry. Plasma ACTH was measured with a commercial
RIA kit (Incstar, Stillwater, MN) as previously described (13), except
that kit assay buffer was substituted for the previous diluent of 5%
human albumin. In the assay buffer currently supplied with the kit,
dilution curves of rat plasma exhibited parallelism across the entire
range of the standard curve (data not shown).
Plasma corticosterone was measured with a kit from ICN (Costa Mesa, CA), halving all reagent and sample volumes. Assay sensitivity (12.5 pg) and variability (7–10%, interassay) were not affected by this modification. The assay exhibited parallelism at rat plasma concentrations from 20 times higher to 5 times lower than the manufacturer’s recommended 1:200 working dilution. As necessary, samples were diluted within these limits to bring them within the standard curve. According to the manufacturer’s specifications, the maximum cross-reactivity of related steroids is 0.34% (deoxycorticosterone). We also found that 11-dehydrocorticosterone exhibits less than 0.022% cross-reactivity (data not shown).
Plasma glucose was measured by the glucose oxidase technique using an APEC analyzer (Danvers, MA).
Anterior pituitary POMC mRNA. Anterior pituitaries were dissected free of the neurointermediate lobe, snap-frozen in liquid nitrogen, and stored at -80 C. Total RNA was extracted using a commercial modification of the acid-phenol technique by homogenizing (Polytron, Brinkmann, Westbury, NY) each pituitary in 1 ml TRI reagent (Molecular Research Center, Cincinnati, OH), followed by chloroform extraction and isopropanol precipitation. Total anterior pituitary RNA (3 µg) was separated, blotted, and hybridized as previously described (14). Hybridization was performed for 16–20 h at 65 or 60 C with complementary RNA probes for mouse POMC or rat ß-actin, respectively. Probes were synthesized by in vitro transcription with either T3 or T7 polymerase (Stratagene, La Jolla, CA) from appropriately linearized plasmids and were used at 106 cpm/ml hybridization solution. The plasmid containing the 900-bp mouse POMC fragment has been previously described (15). The rat ß-actin plasmid was constructed by subcloning a 150-bp XbaI-KpnI fragment of a commercially available plasmid (Ambion, Austin, TX) into Bluescript II KS+ (Stratagene). POMC and ß-actin band intensities were quantitated on a PhosphorImager (Molecular Dynamics, Sunnyvale, CA). POMC was normalized to the ß-actin signal for each rat, and this ratio was expressed as the fold change over the average POMC/actin ratio for sham-adrenalectomized controls, which were included on every blot.
Statistical analysis
The data in
Figs. 1–6
and Tables 1 and 2 are presented as the
mean ± SEM. These data were analyzed by one- or
two-way ANOVA for independent measures, followed by the Newman-Keuls
test where overall main effects were significant (Sigmastat, Jandel
Corp., San Rafael, CA) (16). Departures from normality were assessed by
the Wilk-Shapiro test (17). Where required to establish normal
distribution and equal variance, data were log transformed before
analysis. Significance was defined as P < 0.05.
|
|
|
|
|
|
|
|
, the relationships between plasma ACTH or anterior pituitary
POMC mRNA vs. plasma corticosterone were analyzed by linear
regression of logarithmically transformed ACTH or normalized POMC
values against plasma corticosterone, followed by comparison of line
slopes and y-intercepts by analysis of covariance. Data
analysis was performed with the SPSS/PC statistical package, version
6.1 (SPSS, Chicago, IL). Separate regression equations were derived for
each diet group. The Pearson correlation coefficient (r) was used as a
measure of linear association between the dependent [ln(ACTH) and
ln(POMC)] and independent (corticosterone) variables. The coefficient
of determination (r2) was computed to indicate the
proportion of variance in ln(ACTH) or ln(POMC) accounted for by the
fitted regression line. Equality of slopes was compared by an
interaction F test between the diet group factor and the covariate
(plasma corticosterone). Differences in y-intercept were
assessed by the F test in the analysis of covariance. A two-tailed
P < 0.05 was considered significant.
|
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
), as
described in Materials and Methods. By the day before death,
the respective food intake in the three groups resulted in weight loss
in the protein-deprived group, whereas pair-fed rats maintained and
normal diet controls gained weight (Table 1). Although
weight gain lagged slightly, but significantly, in ADX+30% B rats fed
a normal diet, body weight in protein-deprived and pair-fed diet rats
did not differ significantly between Sham and ADX+30% B treatment
groups. Reduced caloric intake in both the 0% protein and pair-fed
groups was reflected in significantly lower morning plasma glucose
levels relative to those in rats fed a normal diet ad
libitum., but glucose levels did not differ significantly between
protein-deprived and pair-fed rats (Table 1).
Basal morning plasma corticosterone levels were significantly elevated
after 5 days of protein deficiency in sham-adrenalectomized rats.
However, pair-fed rats subjected to caloric restriction equivalent to
that in the protein-deprived groups did not exhibit significantly
increased corticosterone (Fig. 2, bottom
left). Despite this increase in corticosterone, no inhibition of
either basal or restraint-induced plasma ACTH was evident (Fig. 2, top); plasma corticosterone responses to restraint stress
were also not significantly different among the three
sham-adrenalectomized groups (Fig. 2, bottom right).
To determine whether diet-induced increases in corticosterone secretion
might have inhibited further increases in plasma ACTH in Exp 1, we
clamped plasma corticosterone at a constant physiological level by
adrenalectomy and corticosterone pellet replacement (12) 2 days before
diet manipulation. Plasma corticosterone in protein-deprived, ADX+30%
B rats was also significantly higher than that in their normal diet or
pair-fed counterparts (Fig. 2, bottom left). Consistent with
visual assessment of adrenalectomy at death, corticosterone
concentrations did not increase after restraint in ADX+30% B rats
(Fig. 2, bottom right). As in sham-adrenalectomized rats, a
30-min restraint induced comparable ACTH responses among normal,
protein-deprived, and pair-fed rats (Fig. 2, top right)
despite the differences in basal corticosterone levels (Fig. 2, bottom left).
Exp 2: effect of controlling changes in corticosterone levels
during dietary protein deprivation on plasma ACTH and anterior
pituitary POMC mRNA
Although adrenalectomy and low level corticosterone replacement in
Exp 1 prevented corticosterone secretion, it did not prevent increases
in plasma corticosterone levels in protein-malnourished rats. To assess
the effect of protein deprivation on basal ACTH secretion in the
absence of changes in corticosterone, three corticosterone replacement
groups were initially studied: 1) adrenalectomized rats without
corticosterone replacement (ADX+0), 2) adrenalectomized rats replaced
with a 15% corticosterone sc pellet (ADX+15% B), and 3)
adrenalectomized rats given a 30% corticosterone sc pellet (ADX+30%
B). The ADX+15% B and ADX+30% B groups were included as controls for
changes in corticosterone in case the ADX+0 rats did not tolerate
protein malnutrition. ADX+15% B rats were used to match, after the
increase associated with protein deprivation, the levels produced
by physiological corticosterone replacement (ADX+30% B) in normally
nourished rats.
Consumption of the normal, 21% protein diet was relatively stable and
comparable among endocrine groups throughout the 5-day experiment (Fig. 3A). Rats given the protein-free diet decreased intake
by about 25% on day 1, with further decreases to a final level
approximately 50% of the original intake of normal diet. Intakes of
the 0% protein diet were similar among endocrine groups, and although
ADX+0 rats tended, nonsignificantly, to eat less at all times (Fig. 3B), all rats in this group survived. Pair-fed controls were treated as
in Exp 1 (data omitted for clarity). Despite slight differences in food
consumption, protein-deprived rats all lost weight to a similar extent,
regardless of endocrine group (Table 2). In parallel
with the more pronounced reduction in caloric intake in the
protein-deprived ADX+0 group, pair-fed ADX+0 rats tended to lose,
rather than maintain, body weight, but this tendency was not
significant (Table 2). As in Exp 1, protein-deprived and pair-fed
groups were hypoglycemic relative to normal diet controls, but glucose
levels were similar between protein-deprived and pair-fed rats in each
endocrine group and across endocrine groups in each diet group (Table 2).
Plasma corticosterone changed as a function of replacement and diet
group, increasing with corticosterone replacement and increasing
further in the weight-losing, protein-deprived rats of each endocrine
group. As in Exp 1, adrenalectomy was verified at autopsy. Plasma
corticosterone in protein-deficient ADX+15% B rats was also
significantly higher than that in ADX+30% B rats given a normal diet
(Fig. 4, bottom). Plasma corticosterone
tended to be greater in pair-fed vs. normally nourished
ADX+15 and 30% B rats; however, this increase was neither significant
nor of the magnitude seen in protein-deprived rats.
Within a given diet group, plasma ACTH exhibited the expected increases
with adrenalectomy and inhibition with graded corticosterone
replacement. However, ACTH levels were augmented after protein
deprivation in ADX+0 and ADX+15% B rats. This increase was
significantly different from that in 21% protein diet controls in both
endocrine groups, but only statistically distinguishable from levels in
the corresponding pair-fed group in ADX+15% B rats (Fig. 4, top). Thus, the elevated ACTH levels in ADX+15% B rats were
not only significantly greater than plasma ACTH in the normal diet or
pair-fed controls in the same endocrine group, but occurred despite
circulating corticosterone concentrations that were as high or even
higher than those in the same diet groups at the 30% B replacement
level (Fig. 4, bottom).
Experiment 2 was repeated (Exp 2B) to evaluate the effects of dietary manipulation on anterior pituitary mRNA levels of POMC, the precursor to ACTH. In addition, a group of ADX+60% B rats fed a normal diet was included as a comparison for the high plasma corticosterone levels observed in protein-deprived, ADX+30% B rats. Consumption of the experimental diets and changes in body weight were not different from those in sham, ADX+0 B, ADX+15% B, and ADX+30% B rats from Exp 2A; ADX+60% B rats displayed food intake and weight gain comparable to those in other ADX groups given the normal diet (data not shown).
Changes in plasma corticosterone exhibited a similar pattern between
repetitions of Exp 2, although the tendency for corticosterone to
increase between pair-fed and normal diet ADX+15% B rats in Exp 2A
became significant in Exp 2B (Fig. 5, bottom). Unlike Exp
2A, plasma ACTH did not show marked changes with diet within any
endocrine group except in ADX+0% B rats, in whom significant
differences were found only between normal and protein-deprived groups
(Fig. 5, middle).
Despite differences in the pattern of ACTH between repetitions of Exp
2, the response of anterior pituitary POMC mRNA to manipulation of
corticosterone and diet in Exp 2B closely resembled that of plasma ACTH
in Exp 2A. Quantitative results of Northern analysis are shown at the
top of Fig. 5, with hybridization signals of
representative samples depicted in Fig. 6. Normalized
POMC mRNA expression was significantly elevated in 0% vs.
21% protein diet groups for both ADX+0% B and ADX+15% B rats. Like
plasma ACTH in Exp 2A, POMC expression was significantly higher in
protein-deprived ADX+15% B rats than in pair-fed controls.
Furthermore, POMC expression was significantly increased in
protein-deprived ADX+15% B and ADX+30% B rats compared to that in
normal diet or pair-fed rats with similar plasma corticosterone levels
from the next higher corticosterone group (Fig. 5, top).
Changes in POMC expression were not due to inverse variations in the
level of ß-actin mRNA that was used as an internal control, as diet
did not significantly affect anterior pituitary ß-actin expression
(Fig. 6 and data not shown).
Because the foregoing data suggested that inhibition of ACTH and POMC
required higher corticosteroid levels in protein-deprived rats, we
performed regression analysis against plasma corticosterone in ADX,
corticosterone-replaced rats for ACTH and POMC. Our goal was to
determine 1) if the slope, or rate of change in ACTH or POMC against
corticosterone, would reveal differences in sensitivity to
corticosteroid feedback inhibition, and 2) if the
y-intercept, or level of ACTH and POMC in the absence of
corticosterone, would indicate increased drive to the
hypothalamic-pituitary-adrenocortical (HPA) axis independent of changes
in feedback. Because this analysis addressed the relationship between
plasma ACTH and plasma corticosterone in individual rats without regard
to endocrine group, and because group hormone values differed only in
relative pattern and not absolute level between repetitions of Exp 2,
plasma ACTH data from Exp 2A and 2B were pooled. Over the range of data
obtained, a significant linear relationship was found when the natural
logarithm (ln) of the dependent variables ACTH and POMC was regressed
on the original plasma corticosterone data. In each analysis, there was
a strong inverse relationship between plasma corticosterone and the
dependent variable (P < 0.0001 for each). Details of
the regression modeling for each diet group are presented in Fig. 7. Tests of slopes and y-intercepts for ACTH
and POMC among the three diet groups are shown in Table 3.
|
), there were no differences in slope among the
regression lines for normal, protein-deprived, and pair-fed rats. The
difference in y-intercept between the normal and pair-fed
groups was only marginally significant (P = 0.07)
before and was not significant (P = 0.21) after
Bonferroni correction, indicating that plasma ACTH could not be
statistically distinguished between normal and pair-fed groups across
the entire range of plasma corticosterone levels. However, there were
significant differences in intercept for 0% protein vs.
both the normal and pair-fed diet groups (Table 3), indicating that
plasma ACTH in the protein-deprived group differed significantly from
that in the other two groups at all values of corticosterone.
For POMC mRNA (Fig. 7B), as for plasma ACTH, no statistical differences
in slope were found when the regression line of either the 21% or 0%
protein group was compared to that of the pair-fed group. However, the
regression line slope of the 0% protein group was significantly less
than that of the 21% protein group. In addition,
y-intercepts differed significantly among all three diet
groups. These differences remained significant after Bonferroni
correction, except for the comparison between protein-deprived and
pair-fed rats, which was marginally significant (P =
0.059) after Bonferroni correction (Table 3). As regression lines for
POMC data from the 21% and 0% protein groups were nonparallel, as
indicated by significant differences in slope, we tested for
differences in intercept at defined corticosterone values. By this
approach, the lines for normal and protein-deprived rats could be
statistically discriminated at plasma corticosterone levels equal to or
greater than 0.6 µg/dl.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In intact rats, despite equivalent caloric restriction between protein-deprived and pair-fed rats, only protein-depleted rats had significantly higher plasma corticosterone than normally nourished controls. Numerous studies have reported elevated glucocorticoid levels in protein-malnourished subjects (4, 5, 6, 19, 20, 21), but quantitative information on food intake has been lacking. We have shown that the increase in plasma corticosterone levels after protein-free feeding is not solely due to hypophagia. Although reduced food intake was associated with hypoglycemia in both protein-deprived and pair-fed groups, there was no difference in plasma glucose that could account for higher corticosterone levels in the 0% protein group. Although pair-fed rats may have also been protein deficient by virtue of food restriction, their protein intake was higher than that of rats in preliminary experiments given 4% protein, in whom plasma ACTH and corticosterone did not increase (data not shown). Furthermore, protein deficiency in pair-fed rats would only have minimized differences from the 0% protein group. Thus, although it is conceivable that there is an interaction between the stimuli of inadequate protein and insufficient caloric intake in voluntarily hypophagic, protein-deprived rats, caloric restriction alone, at least over the time span of these studies, does not account for the apparent increases in pituitary or adrenocortical activity.
Besides protein intake, only body weight differed clearly between the 0% protein and pair-fed groups. Pair-fed rats maintained weight, whereas protein-deprived rats lost approximately 11% of their initial weight within 4 days. Changes in the adipocyte hormone leptin, which have been implicated in fasting-induced adrenocortical activity (22), or in other systemic or central factors responsive to weight loss could augment glucocorticoid production in protein malnutrition.
Elevated basal corticosterone levels in protein-deprived rats did not
inhibit ACTH responses to restraint stress. In contrast, increasing
morning corticosterone to equivalent levels (
3 µg/dl) by low dose
pellet implantation in normal rats significantly impairs stress-induced
ACTH secretion (23). Our data are consistent with reports in humans
with protein malnutrition of high basal cortisol levels that neither
reduce adrenocortical responses to hypoglycemia or infection stress nor
suppress normally after dexamethasone treatment (5, 24, 25).
To identify the source of this apparent drive for glucocorticoid secretion during protein deficiency, we have investigated indexes of ACTH synthesis and secretion after controlling corticosterone levels by adrenalectomy and corticosterone replacement. In Exp 1, corticosterone pellet doses providing physiological levels of 3–5 µg/dl in normal adrenalectomized rats produced levels almost twice that after 5 days of protein deprivation. It is improbable that this increment was due to secretion from residual adrenal tissue, as adrenalectomy was confirmed by visual inspection at autopsy and the lack of significant increases in corticosterone after restraint stress. Instead, the increase was more likely due to weight loss in the protein-depleted group, which would decrease the distribution volume of the implanted corticosterone, and to decreased glucocorticoid clearance, which has been reported in protein malnutrition (5, 24).
When increases in corticosterone due to metabolic and body weight changes were prevented by adrenalectomy without replacement, increases in both plasma ACTH and anterior pituitary POMC mRNA were evident in protein-deprived relative to normally nourished rats. When corticosterone was controlled by replacement at half-physiological levels, plasma ACTH also differed significantly between protein-deprived and pair-fed groups in ADX+15% B rats in Exp 2A. However, although we cannot explain differences in the pattern of the ACTH response to protein deprivation in ADX+15% B rats between repetitions of Exp 2, POMC data from Exp 2B recapitulated the pattern of plasma ACTH from Exp 2A. Thus, data analysis according to endocrine and diet group, without regard to plasma corticosterone levels, indicated a consistent underlying stimulatory effect of protein deprivation on ACTH synthesis and secretion.
To determine whether such an effect was due to increased drive for or decreased glucocorticoid inhibition of ACTH production, we regressed plasma ACTH and anterior pituitary POMC mRNA expression on plasma corticosterone for each diet group. The lack of slope differences between regressions of ACTH on corticosterone indicated a similar relationship between ACTH and total plasma corticosterone in all diet groups; i.e. for a given linear increase in corticosterone, there was a similar log order magnitude decrease in ACTH for each group. This suggests that the increase in HPA activity with protein deprivation is less likely to be due to a decrease in corticosteroid feedback inhibition of plasma ACTH. Our finding of a significant linear relationship between ln(ACTH) and corticosterone during protein malnutrition stress, with little apparent change in feedback inhibition, is highly consistent with the relationship demonstrated between acute stress-induced ACTH and corticosteroids in dogs (26).
However, the significant differences in y-intercept between the 0% protein and both the normal diet and pair-fed groups imply that plasma ACTH was significantly higher in protein-deprived rats not only at any measurable level of corticosterone, but when corticosterone was extrapolated to zero. Thus, elevated plasma ACTH in protein-deprived rats is primarily due to increased stimulatory drive, evident even in the absence of corticosteroid negative feedback.
Regression analysis of POMC mRNA data yielded qualitatively similar results. Similar regression line slopes between normal diet and pair-fed groups as well as between pair-fed and protein-deprived rats suggested comparable sensitivity to total corticosterone levels. Although the slope of the regression for protein-deprived rats was significantly lower, implying lower sensitivity to corticosteroid feedback, than that of normally nourished rats, this difference is difficult to interpret in light of the similarity of either group’s slope to that of pair-fed rats. Analysis of the plasma corticosterone values at which the intersecting regressions for normal diet and protein-deprived rats diverge indicated that the two groups could be statistically discriminated at plasma corticosterone levels of 0.6 µg/dl and above. We infer from this low level of corticosterone and the parallelism of both groups relative to pair-fed rats, that POMC mRNA expression differs significantly between normally nourished and protein-deprived rats at any corticosterone concentration.
In contrast to those for ACTH, the y-intercepts for POMC differed significantly between normal diet and pair-fed groups, suggesting that food restriction in the latter group significantly increased the drive for POMC expression. Preliminary data (not shown) indicate that afternoon plasma ACTH levels are increased in pair-fed rats, probably in anticipation of food presentation (10). This diurnal stimulus for ACTH release may be reflected by residual elevations in POMC expression the next morning. Elevated morning anterior pituitary POMC mRNA levels are also associated with increased circadian drive for ACTH secretion in adequately nourished rats replaced with low levels of corticosterone that normalize morning, but not afternoon, plasma ACTH after adrenalectomy (27).
The difference in intercept distinguishing the protein-deprived from the pair-fed group was marginally significant after Bonferroni correction. Nevertheless, the adjusted P value (0.059) is sufficiently low and the correction so conservative for the number of data points (28), that we believe the difference in intercept reflects an additional drive for POMC expression specifically attributable to protein deprivation.
Because the current analyses only address total and not free corticosterone, it is possible that the free, biologically active fraction is normal in protein-deprived rats. However, as both corticosteroid-binding globulin and albumin decrease in protein malnutrition (25), there is likely to be an even greater discrepancy between the levels of plasma ACTH and POMC mRNA observed and those predicted from the presumed higher level of free corticosterone in protein-deprived rats. Although it is also conceivable that our assay is detecting an inactive metabolite of corticosterone, the most likely unconjugated metabolite, 11-dehydrocorticosterone, shows negligible cross-reactivity (see Materials and Methods).
The elevated plasma ACTH and anterior pituitary POMC mRNA expression induced by protein deprivation could also result from decreased clearance, rather than stimulated production, of these molecules. The circulating half-life of ACTH could increase during protein malnutrition, as has been shown for glucocorticoids (5, 7, 24). Generalized decreases in protein synthesis with dietary protein deprivation (29) could potentially increase POMC mRNA stability by decreasing ribonuclease levels. However, in contrast to our findings with ACTH and POMC mRNA, half-lives of other circulating proteins, such as IGF-I, (30), and levels of other anterior pituitary hormone mRNAs, such as TSH mRNA (31), decrease with protein restriction. Thus, even changes in clearance would involve selectivity for corticotrope functions or for the ACTH and POMC molecules themselves.
The factors mediating increased ACTH synthesis and secretion during
protein malnutrition are unknown. Of the main positive regulators of
corticotrope function, CRH is reported to be decreased by food
restriction or fasting (32, 33), whereas vasopressin may be decreased
or unaffected (33). However, these studies did not control for possible
feedback inhibition from increased plasma glucocorticoids induced by
food deprivation; one study in which glucocorticoid levels were
controlled demonstrated increased vasopressin mRNA expression in the
hypothalamic paraventricular nucleus (34). Furthermore, as our data
indicate that at the same level of caloric intake, protein deficiency
exerts different effects from food restriction on anterior POMC mRNA
and ACTH production, protein malnutrition could also differ from food
restriction in potentially stimulating CRH or vasopressin. In possible
support of a stimulatory effect on CRH, 
-helical CRH has been shown
to attenuate decreases in the consumption of protein-free diet by rats
(35), although it remains to be determined whether this effect is
specifically attributable to hypothalamic CRH or even to CRH-related
molecules such as urocortin (36).
Alternatively, protein deficiency may reduce levels of a nonglucocorticoid inhibitory factor. One such factor could be prepro-TRH-(178–199), which is immunologically detectable in the median eminence and has been shown to decrease the synthesis and secretion of ACTH in vitro (37). As protein deprivation, but not equivalent caloric restriction, decreases hypothalamic TRH mRNA expression (31), parallel reductions in the release of the inhibitory TRH-(178–199) fragment could account for the increases we observed in POMC mRNA and ACTH secretion.
Our results indicate that protein malnutrition augments basal plasma glucocorticoids at least in part by increasing plasma ACTH and anterior pituitary POMC mRNA levels. This stimulatory effect on ACTH and POMC is specific for protein vs. calorie deprivation, can be distinguished from reductions in sensitivity to corticosteroid feedback, and suggests that protein deprivation is a chronic stimulus to HPA activity that preserves responsiveness to other stressors while elevating basal glucocorticoid levels. Under these conditions, the catabolic actions of glucocorticoids may assist in liberating body protein stores as malnutrition progresses.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received September 19, 1996.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  H. Raff and L. Jacobson Glucocorticoid feedback control of corticotropin in the hypoxic neonatal rat J. Endocrinol., February 1, 2007; 192(2): 453 - 458. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. I. Lewis, X. Da, H. Li, and M. Fournier Tumor necrosis factor-{alpha} and malnutrition-induced inhibition of diaphragm fiber growth in young rats J Appl Physiol, November 1, 2005; 99(5): 1649 - 1657. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. D Bruder, L. Jacobson, and H. Raff Plasma leptin and ghrelin in the neonatal rat: interaction of dexamethasone and hypoxia J. Endocrinol., June 1, 2005; 185(3): 477 - 484. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Raff, L. Jacobson, and W. E. Cullinan Elevated corticosterone and inhibition of ACTH responses to CRH and ether in the neonatal rat: effect of hypoxia from birth Am J Physiol Regulatory Integrative Comp Physiol, November 1, 2003; 285(5): R1224 - R1230. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. M. Johnson, S. K. Vogt, M. W. Burney, and L. J. Muglia COX-2 inhibition attenuates anorexia during systemic inflammation without impairing cytokine production Am J Physiol Endocrinol Metab, March 1, 2002; 282(3): E650 - E656. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Lesage, B. Blondeau, M. Grino, B. Bréant, and J. P. Dupouy Maternal Undernutrition during Late Gestation Induces Fetal Overexposure to Glucocorticoids and Intrauterine Growth Retardation, and Disturbs the Hypothalamo-Pituitary Adrenal Axis in the Newborn Rat Endocrinology, May 1, 2001; 142(5): 1692 - 1702. [Abstract] [Full Text]
|
|
|
|
 |
|
 I. Castagliuolo, K. Karalis, L. Valenick, A. Pasha, S. Nikulasson, M. Wlk, and C. Pothoulakis Endogenous corticosteroids modulate Clostridium difficile toxin A-induced enteritis in rats Am J Physiol Gastrointest Liver Physiol, April 1, 2001; 280(4): G539 - G545. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. SOLOMON POMC-Derived Peptides and Their Biological Action Ann. N.Y. Acad. Sci., October 20, 1999; 885(1): 22 - 40. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. M. Solano and L. Jacobson Glucocorticoids reverse leptin effects on food intake and body fat in mice without increasing NPY mRNA Am J Physiol Endocrinol Metab, October 1, 1999; 277(4): E708 - E716. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Jacobson Lower Weight Loss and Food Intake in Protein-Deprived, Corticotropin Releasing Hormone-Deficient Mice Correlate with Glucocorticoid Insufficiency Endocrinology, August 1, 1999; 140(8): 3543 - 3551. [Abstract] [Full Text]
|
|
|
|
|
|
L. Jacobson Glucocorticoid Replacement, but not Corticotropin-Releasing Hormone Deficiency, Prevents Adrenalectomy-Induced Anorexia in Mice Endocrinology, January 1, 1999; 140(1): 310 - 317. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||
