INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

NEUROPEPTIDE Y (NPY) is recognized as an important component of the energy homeostasis system. Because of its powerful effects to initiate and sustain eating (5), NPY has been investigated to elucidate normal and abnormal regulation of food intake and body weight. One example of altered energy regulation is the "anorexia of aging," a syndrome seen in the elderly and characterized by spontaneous decrease in food intake and body weight. Efforts to identify the etiology of age-related anorexia/hypophagia include studies testing NPY function in aged animals. For example, Pich et al. (15) showed that old (24 mo old) rats responded less robustly to the feeding-stimulatory action of intrahypothalamic NPY injection than did young (3 mo old) rats. Gruenewald et al. (7) demonstrated lower fasting-induced increases in hypothalamic arcuate nucleus NPY mRNA with age in 3-, 12-, and 24-mo-old rats. Additionally, the old rats were slower to recover food intake and body weight after refeeding than the young animals. From these results, a causal relationship was suggested between reductions in activation of the NPY pathway and decreased feeding during aging.

We previously studied the feeding behavior of an animal model for age-related anorexia/hypophagia that was characterized by transition to a physiological state in which there was a rapid spontaneous decline in food intake and body weight (2). To this end, we monitored food intake continuously in 24-mo-old male Fischer 344 (F344) rats as they passed from their period of weight stability (presenescence) into spontaneous rapid weight loss (senescence). Regardless of age at which senescence began, all rats ate significantly smaller meals of shorter duration during their early stages of senescence than during presenescence. These findings, combined with the fact that meal frequency did not differ between presenescence and senescence, clearly demonstrated altered feeding behavior as one cause of age-related anorexia. The consumption of shorter, smaller meals, despite progressively declining body weight (which averaged 5% at the end of the experiment), implicated dysfunction in short- and long-term regulation of energy balance. Moreover, the senescent male F344 rat exhibits significant functional deterioration in a number of systems regulated by the hypothalamus, e.g., food intake, body weight, temperature homeostasis, and circadian rhythmicity of body temperature (2, 11, 12). These findings, together with data suggesting that paraventricluar hypothalamic concentrations of NPY do not differ significantly between presenescent and senescent animals, led us to postulate that attenuated hypothalamic NPY responsiveness may contribute to the altered feeding behavior of the senescent rat.

The specific aim of this investigation was to test the hypothesis that the decline in food intake observed in senescent male F344 rats near the end of their lives is accompanied by a significant reduction in their feeding response to NPY. To test this hypothesis, NPY was injected into the lateral ventricle in old (24 mo old) rats before and after the onset of senescence. Feeding behavior was monitored continuously for 24 h after injections, and comparisons were made between the presenescent and senescent periods. In addition, we measured serum levels of leptin and corticosterone, hormones influencing food intake and body weight (4, 14).

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Animals and animal care. Male F344 rats aged 24 mo (n = 11) and 8 mo (n = 9) were obtained from the National Institute on Aging colony maintained by Harlan Sprague Dawley Laboratory (Indianapolis, IN). Animals were housed individually and maintained at 25-26°C and 50% humidity on a reverse 12:12-h light-dark cycle (lights on at 1900, lights off at 0700). Rats were provided with ground NIH-31 laboratory chow (Teklad Research Diets, Indianapolis, IN) and distilled water ad libitum, except when indicated otherwise.

Experimental protocol. On arrival at our facility, rats were placed in polycarbonate metabolic cages designed for automated food intake measurement, as previously described (2). Briefly, digital scale measurements of the food cup and food spillage were transmitted to a computer every 15 s, and changes of 0.1 g were time stamped and recorded onto a spreadsheet (Microsoft Excel) using SoftwareWedge (T. A. L. Technologies, Philadelphia, PA). Food and water were replenished, cage bedding was changed, and rats were weighed and examined daily between 0800 and 1000 under low-intensity red light (15-20 lux).

Food restriction. To determine the effect of weight loss per se on NPY-induced feeding, presenescent rats were tested with NPY after weight loss due to food restriction. Rats were given a daily amount of food equal to two-thirds of their ad libitum consumption until a weight loss of ~10% of their prerestriction weight was observed (6-7 days). The 24-h allotment of food was divided into four equal parts that were sequentially dispensed into the food cup at 6-h intervals by an adapted aquarium feeder set to a timer (Rainbow Lifegard Aquarium Products, El Monte, CA). The purpose of distributing the food four times over 24 h was to minimize the possible effects on endogenous levels of NPY caused by rats eating their food at one time (19). Seventy minutes before testing, rats were allowed access to 0.5 g of food for 10 min to simulate the active feeding status that rats displayed at this time during ad libitum feeding. Sixty minutes before injection, all food and water were removed. The injection protocol was the same as that for AL experiments, except animals were food restricted during the 24- to 48-h period after each injection.

Food intake analysis. Feeding responses were analyzed using data collected during the 24-h postinjection period. Feeding variables included size and duration of the first meal after injection, latency to feed, and meal frequency. Meals were defined as the consumption of 0.2 g within 15 min, separated from other periods of feeding by 15 min. Quantitative feeding variables are presented as body mass-independent values (g intake/kg^0.67) (8) to account for differences in body weight among the animal groups. The patterns of feeding responses among and within animal groups were similar for the different NPY doses. For ease of presentation, feeding measures in Table 2 and Figs. 2 and 3 are displayed only for the maximally effective dose of NPY (10 µg).

Serum analysis. Serum corticosterone was measured by RIA using a diagnostic kit from ICN Pharmaceuticals (Orangeburg, NY). Serum leptin was measured by RIA using a diagnostic kit from Linco Research (St. Charles, MO).

Statistical analysis. ANOVA and post hoc Fisher's protected least significant difference test were used to evaluate the effects of age (young, old presenescent, and old senescent), feeding state (AL and FR), and treatment (aCSF and NPY) on feeding variables. Hourly food intake measurements were analyzed by repeated-measures ANOVA followed by Bonferroni/Dunn post hoc tests. Differences were considered significant at P 0.05. Values are means ±SE.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Body weight. Average preoperative body weights for the old and young rats were 409.8 ± 8.7 and 375.5 ± 0.4 g, respectively (P < 0.0001). Postoperative body weight loss averaged 11.3 ± 0.8% (old rats) and 11.1 ± 0.3% (young rats) and did not differ significantly. Neither group returned to preoperative weights. In the old senescent rats, mean body weight loss was 9.8 ± 1.1% at the first injection (8 µg of NPY) and 18.0 ± 1.5% at the final injection. The age at which spontaneous rapid weight loss began in the senescent rats varied from 809 to 917 days, and average duration of senescence before euthanasia was 19 ± 2 days. Temporal patterns of body weight and food intake for three representative old rats (of a total of 11) are presented in Fig. 1. Although there was variation in the shape of the curves for each animal, all exhibited simultaneous declines in food intake and body weight. These spontaneous losses occurred in senescent rats regardless of age (note differences in the experimental day at which the decline occurred in Fig. 1).

View larger version (29K): [in this window] [in a new window]   Fig. 1.   Temporal relationship of changes in body weight and food intake at 1st day after neuropeptide Y (NPY) injection in 24- to 30-mo-old rats. Final measurements were taken immediately before euthanasia at 934, 885, and 825 days of age for animals 1, 2, and 3 (A, B, and C, respectively). AL, presenescent ad libitum fed; FR, presenescent food restricted; S, senescent ad libitum fed.

Food intake analysis. Postoperative daily food intake before injections was significantly greater in old presenescent than in young rats: 27.91 ± 0.31 vs. 23.50 ± 0.44 g · day¹ · kg^0.67. Twenty-four-hour food intake was significantly reduced on aCSF injection days compared with prior injection days in the old presenescent and senescent, but not young, rats: 17.50 ± 3.13, 10.22 ± 2.85, and 21.05 ± 4.02 g · day¹ · kg^0.67, respectively. Comparisons of daily intakes between aCSF injection days and interinjection days (the 24-h periods separating injection treatments) showed no differences within animal groups (P = 0.24).

Response to NPY in young and old presenescent and senescent rats. As shown in Fig. 2, the ability of NPY to stimulate food intake at the first meal after injection in young and old presenescent rats was significantly attenuated after the onset of senescence. Increased duration and net change of the first meal after NPY injection were also blunted in senescent vs. presenescent AL and FR rats (P < 0.0001; Table 2). When size and duration of the first meal after injection were expressed as percentages of aCSF values, senescent and presenescent FR rats showed smaller increases than presenescent AL rats (P < 0.0001). In response to NPY, 24-h food intake increased and latency to feed decreased in senescent and presenescent AL, but not presenescent FR, rats. Although differences were not found among any of the older groups in eating rate (g intake/min), senescent rats consumed more meals during aCSF and NPY treatments than did old presenescent and young groups (P = 0.002).

View larger version (47K): [in this window] [in a new window]   Fig. 2.   Food intake at the 1st meal after injection in young (n = 8) and old (24-30 mo, n = 11) presenescent rats during ad libitum feeding, old presenescent rats during food restriction, and senescent rats during ad libitum feeding. Animals were treated with intracerebroventricular injection of artificial cerebrospinal fluid (aCSF) or NPY (2, 4, 8, 10, and 12 µg). * Significantly different from senescent ad libitum fed. ** Significantly different from senescent and presenescent ad libitum fed.  Significantly different from old presenescent ad libitum fed and food restricted and senescent ad libitum fed.  Significantly different from old presenescent and senescent ad libitum fed.

View larger version (15K): [in this window] [in a new window]   Fig. 3.   Average hourly food intake after intracerebroventricular injection of aCSF and NPY (10 µg) in old (24-30 mo, n = 11) rats during presenescent ad libitum feeding and senescent ad libitum feeding.

Response to NPY in young and old presenescent rats. A main effect of age was found on baseline (i.e., response to aCSF injection) feeding (g/kg^0.67) at the first postinjection meal that became less pronounced, yet remained significant, after treatment with NPY (Table 2). During ad libitum feeding and food restriction, young rats ate more (g/kg^0.67) than did old presenescent rats at the first meal (P < 0.008) and 24 h (P < 0.001) after injection. Rates of eating were faster in young than in old animals in both feeding states (P < 0.0001). No differences between age groups in NPY-induced feeding responses (calculated relative to aCSF values) were found, except for a greater decrease in latency to feed in young AL rats than in their old counterparts (P < 0.014). A main effect of NPY was seen on most measurements of feeding in young and old presenescent animals during ad libitum and restricted feeding. Repeated-measures analysis revealed that, within both feeding states, old rats ate significantly less than did young rats during postinjection hour 1 (P < 0.0001); after this time, hourly intakes were similar between the age groups.

Water intake. ANG II elicited significant increases compared with aCSF and NPY in 20-min and 24-h postinjection water intake in old presenescent and senescent rats. As shown in Fig. 4, the ability of ANG II to stimulate immediate drinking was significantly different among old rats: 2.4 ± 1.9, 14.5 ± 1.0, and 8.0 ± 2.5 ml in senescent, presenescent AL, and presenescent FR rats, respectively. Differences were also found among old rats in 24-h water intake: 27.66 ± 1.45, 42.83 ± 4.89, and 49.12 ± 7.89 ml in senescent, presenescent AL, and presenescent FR rats, respectively. NPY caused an increase in 24-h postinjection drinking in presenescent AL, but not in senescent or presenescent FR, rats.

View larger version (55K): [in this window] [in a new window]   Fig. 4.   Water intake during the 20-min period after intracerebroventricular injection of 150 ng of ANG II in old (24-30 mo) rats during presenescent ad libitum (AL) feeding (n = 11), presenescent food restriction (FR, n = 11), and senescent ad libitum feeding (n = 5). Values not sharing superscript differ significantly (P < 0.05).

Serum leptin. As shown in Table 3, senescent rats had significantly lower serum leptin levels than did old presenescent AL rats (P = 0.005). The 72% reduction in serum leptin was seen in association with an average body weight loss of 18% in senescent rats. No differences were found between the old presenescent and young animals (P = 0.839).

Serum corticosterone. Serum corticosterone concentrations did not differ between old presenescent and senescent rats (Table 3). During presenescence, age had a significant effect on corticosterone concentrations; levels were lower in old than in young rats (P = 0.009).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

We report here that, after transition from their weight-stable (presenescent) to rapid-weight-loss (senescent) period, old rats were markedly less sensitive to NPY's action to stimulate feeding at the first postinjection meal. The increases in first postinjection meal size and duration during senescence were only 25 and 33%, respectively, of their presenescent levels. Differences between presenescent and senescent rats in NPY-stimulated feeding were still apparent at hour 2, i.e., when NPY's effects are normally most potent (1, 10, 15). In addition to confirming our previous finding of altered feeding behavior of the rapid-weight-loss animals, this investigation supports our hypothesis of an alteration in NPY's action to acutely induce eating in senescent rats.

The mechanisms underlying the attenuation of the NPY-induced acute feeding response in senescent rats are unknown, and this investigation was not designed to fully elucidate any particular possibility. Given that NPY was injected directly into the brain, it is logical to assume that the defect occurs along the NPY pathway. Many elements are involved in the signaling cascade mediating the final effects of NPY, including NPY-receptor number and affinity, cAMP and calcium second-messenger activation of gene transcription factors, and modulation of genes responsible for initiating feeding (21). The physiological state of senescence may be associated with changes in one or more of these elements. Previous observations of decrements in binding characteristics for other neurochemicals in older animals as well as an age-related decrease in NPY binding in tissue other than brain suggest that altered NPY-receptor number/affinity might be involved in the senescent rats' blunted responses. For example, the degrees of binding of cerebellar cortical dopamine and hippocampal galanin to their respective receptors are reduced significantly in old vs. young animals (16, 17). We are unaware of similar studies evaluating hypothalamic NPY receptors in senescent rats. As a first step in determining the precise location of a possible defect in the NPY-signaling pathway, we have begun experiments to test the hypothesis that alterations in hypothalamic NPY-receptor binding and/or number occur concomitantly with attenuated feeding responsiveness in the senescent rat.

In reaching our conclusions, we considered other factors that might contribute to the reduced NPY-induced acute feeding response observed in senescent rats. Among these are altered levels of endogenous NPY, serum hormone abnormalities, and incomplete delivery of injected NPY. Endogenous levels of hypothalamic NPY are relatively low in AL rats, and NPY administration elicits a strong feeding response in these animals (18, 19). In contrast, food restriction causes endogenous hypothalamic NPY levels to rise and saturate binding sites, thereby blunting the effects of exogenous peptide (15, 22). The possibility that the blunted response to NPY administration in senescent rats is the result of elevated endogenous hypothalamic NPY levels is unlikely in light of our finding that there is no significant difference in paraventricular nucleus NPY content in senescent vs. AL presenescent male rats (6). Furthermore, senescent rats were less responsive to exogenous NPY than were FR presenescent rats that should have had much higher endogenous NPY levels. Thus it is unlikely that the decreased responsiveness of senescent animals to exogenous NPY can be accounted for by changes in levels and/or availability of endogenous NPY.

A second potential cause of senescence-related decrements in NPY function involves altered levels of circulating hormones that can modulate, either directly or indirectly, NPY. We measured two such hormones, leptin and corticosterone. The former inhibits NPY transcription (10); the latter not only modulates leptin synthesis (4), but also potentates NPY action on feeding (23). The lower serum leptin levels in senescent than in presenescent rats is consistent with their weight (adipose) loss and our previous results (2, 6). Even with significantly lower serum leptin levels, i.e., levels that should stimulate eating, the senescent rats failed to respond as robustly to NPY as they did in presenescence. Measurements of serum corticosterone showed no difference in levels between old presenescent and senescent rats. Nonetheless, presenescent and senescent rats (that were comparably handled on a daily basis) had significantly different responses to NPY. Thus it is unlikely that altered leptin or corticosterone levels can account for the blunted NPY-induced feeding.

Young animals had higher serum corticosterone values than did old presenescent rats. These results differ from those of Sapolsky (20), who showed age-related increases in plasma corticosterone in rats. However, disagreement exists in the literature on glucocorticoids and aging (see Ref. 20 for review). In most investigations, as in ours, measurements were made at only one time of day and thus do not assess the effect of age on the diurnal pattern of corticosterone. Also, differences in sampling procedures and species contribute to the lack of consensus among reports.

A third potential reason for reduced responsiveness to exogenous NPY in senescent animals is incomplete delivery of NPY due to cannula obstruction, resulting primarily from cell adhesion at the base of the guide cannula, especially in the old rats, in which the cannula was in place for up to 6 mo. However, several lines of evidence argue against such cannula obstruction: 1) on the basis of previous reports alerting us to this possibility (9), we designed our injector needles to extend 1 mm beyond the tip of the permanent guide cannula, thus ensuring penetration of any ependymal growth and entry into the ventricle. 2) Results presented here for old presenescent and young rats, as well as data from pilot studies, show that injection of ANG II elicited a strong drinking response, even after repeated (up to 13) treatment injections or in animals with the cannula in place for 160 days (see MATERIALS AND METHODS). 3) Histology showed extensive ventricular diffusion of ink injected at the time of euthanasia. 4) Investigations by one of us (J. E. Blevins) involving 15 hypothalamic injections of norepinephrine in young rats over a 30-day period showed no reduction in feeding responsiveness over a 30-day period (3). Therefore, we believe that all rats in this study reliably received the NPY dose administered and that the blunted NPY-induced feeding seen during the first few hours after injection is a function of physiological alterations associated with senescence.

Although the NPY-induced, mass-independent food intake of the presenescent rats was significantly greater during the first postinjection meal and during hours 1-3 after injection than that of the senescent animals (Fig. 3), 24-h food intake (g/body mass^0.67) did not differ (Table 2; non-body-mass-adjusted food intake remained significantly greater in the presenescent animals). This finding was surprising, and the mechanism underlying this difference is not apparent at this time. We offer two possibilities. First, the normal 2- to 3-h half-life of NPY may be extended in the senescent rat through reduced transport of NPY from the site of injection to NPY-sensitive tissues or a decreased rate of NPY degradation. This suggestion is supported by previous observations in aging humans and rodents of significantly reduced cerebral blood flow (for review see Ref. 13). Second, differences in the pattern of 24-h, postinjection food intake observed in the presenescent vs. senescent rats may reflect alterations in the leptin-NPY axis. The pattern observed in presenescent rats was expected, i.e., compensation for large initial food intake by reducing feeding later in the day. Serum leptin levels remaining elevated for an extended period because the filling of adipose tissue from excess energy intake may mediate this compensation. Alternatively, it is possible that the senescent animals do not have a "normal" leptin response, and thus they maintain food intake at higher levels than would be expected.

Although not pertinent to the discussion of NPY responsiveness and senescence, the significant postoperative losses in body weight in the old and young rats merit a brief discussion. Young and old presenescent rats experienced postoperative reductions in body weight that were nearly identical in severity (~11% loss of preoperative weight), and neither group of rats returned to preoperative body weights. These findings were surprising, since we previously reported that experiments involving intraperitoneal implantation, a procedure that is performed within 15 min of anesthesia induction (11), and other surgical procedures lasting longer than those used here (13) did not result in permanent postoperative losses in body weight. Thus the procedures associated with brain cannulation, rather than abdominal surgery and/or the effects of anesthesia, may be the cause of this failure of the young and old rats to recover their body weight. Moreover, FR rats, which were forced to lose ~10% of their postoperative body weight, returned to their postoperative, rather than preoperative, weight after refeeding; this is consistent with the establishment of a new energy set point. The basis for this does not appear to be due to age and warrants further investigation.

In conclusion, we find that the decline in food intake in senescent rats is associated with a significantly reduced response to intracerebroventricular injection of NPY. The suppressed stimulation of first postinjection meal size and duration in senescent rats is consistent with impairment in NPY's normal action to increase food intake. These data provide strong evidence implicating altered NPY responsiveness in the mechanism(s) underlying senescence-associated anorexia.

Perspectives