Many mammals, nearing the end of life, spontaneously decrease their food intake and body weight, a stage we refer to as senescence. The spontaneous decrease in food intake and body weight is associated with attenuated responses to intracerebroventricular injections of neuropeptide Y (NPY) compared with old presenescent or with young adult rats. In the present study, we tested the hypothesis that this blunted responsiveness involves the number and expression of hypothalamic paraventricular nucleus (PVN) Y1 and/or Y5 NPY receptors, both of which are thought to mediate NPY-induced food intake. We found no significant difference in mRNA levels, via quantitative PCR, for Y1 and Y5 receptors in the PVN of senescent vs. presenescent rats. In contrast, immunohistochemistry indicated that the number of PVN neurons staining for Y1 receptor protein was greater in presenescent compared with senescent rats. We conclude that a decreased expression and number of Y1 or Y5 receptors in the PVN cannot explain the attenuated responsiveness of the senescent rats to exogenous NPY.
- anorexia of aging
- food intake control
- neuropeptide Y
we previously reported that old rats may exhibit two distinct body weight patterns (4, 22, 23). In some cases, they maintain relatively stable body weights, while in other cases, they have a slow gradual body weight loss. However, near the end of life, rats, like many other mammals, exhibit a rapid loss of body weight that reflects significant decreases in daily food intake. We refer to these rats as senescent, a widely accepted term in aging research (9, 21), which indicates that they are no longer thriving, but rather are progressing rapidly toward death (3, 4, 23). Notably, the onset of senescence does not occur at the same chronological age, even in inbred species. For example, in Fischer 344 (F344) rats, we have observed the onset of senescence in ages ranging from 22 to 33 mo, demonstrating a lack of close correlation with chronological aging.
Our studies on aging male F344 rats have also shown that in the early stages of senescence, these rats eat significantly smaller meals of shorter duration, resulting in an inadequate caloric consumption that is followed by death within 3–4 wk (4). The fact that these senescent animals do not increase their energy intake sufficiently to maintain body weight, as do young or old presenescent rats after food restriction, indicates dysfunction in one or more components of the pathways involved in food intake regulation. Although concentrations of serum leptin, an adipocyte hormone that plays a role in modulating food intake, decreased in senescent rats (a finding that is consistent with their loss of body fat), these rats did not respond to this fall in leptin by increasing their food intake, as do younger or old presenescent rats (10). The possibility that this failure to increase food intake reflects alterations in the responsiveness to neuropeptide Y (NPY), a potent stimulator of food intake (7, 33, 34), is suggested by our finding that intracerebroventricular injection of NPY did not elicit a rapid, robust feeding response in senescent rats as it did in old presenescent or young rats (3).
In this study, we have further explored this blunted responsiveness to exogenous NPY by evaluating the relative expression of NPY receptors in the hypothalamic paraventricular nucleus (PVN) as well as the number of PVN neurons that are immunoresponsive to NPY receptor antibodies. We focused on the PVN because of the extensive evidence that it is a site where exogenous NPY potently stimulates food intake, (17, 29, 33, 37, 38). In addition, injection of exogenous NPY, NPY immunoneutralization, and NPY receptor antagonist experiments have identified the PVN as a region of the brain that is important for NPY's orexigenic effects under normal physiological conditions (16, 33).
Although six NPY receptor subtypes have been described (12, 25), only Y1 and Y5 receptors appear to be involved in the control of food intake (6, 8, 28, 30, 31), with Y1 receptor expression possibly being more responsive to deprivation and fasting-induced changes in ingestive behavior (31, 35). The specific aim of this investigation was to test the hypothesis that PVN expression of NPY Y1 and/or Y5 receptors is attenuated in senescent male F344 rats compared with their presenescent counterparts. To this end, we utilized quantitative PCR (qPCR) to assess the expression levels of NPY Y1 and Y5 receptor mRNA. In addition, we used immunohistochemistry to determine the number of PVN neurons stained by antibodies against the Y1 receptor in ad libitum-fed young, old presenescent, and senescent rats, as well as food-restricted young and old presenescent animals.
MATERIALS AND METHODS
Animals and Animal Care
Male F344 rats, aged 6 and 24 mo (n = 20 and 30, respectively), were obtained from the National Institute on Aging colony maintained by Harlan Sprague Dawley Laboratory (Indianapolis IN). On arrival, rats were housed individually in hanging wire-bottom cages (20 × 25 × 18 cm) and maintained at 25–26°C and 50% humidity on a reversed 12:12-h light-dark cycle (lights off at 0700, on at 1900). Rats were provided with NIH-31 laboratory chow (Teklad Research Diets, Indianapolis, IN) and distilled water ad libitum. All rats were maintained in our facility for a minimum of 2 wk before experimentation. At the time the rats were killed, they were visually inspected for any evidence of gross pathology. None exhibited any such pathology, and thus none were excluded from the study. Every 2 mo, two sentinel rats, housed in the same facility, were tested for specific organ disease and serological abnormalities. All tests were negative. All procedures using animals were approved by the University of California Davis Animal Care and Use Committee.
Old rats (>25 mo) whose body weight remained stable or declined slowly during the experimental period were categorized as presenescent; older rats that exhibited rapid, spontaneous weight loss were classified as senescent (3, 4, 10, 22, 23). The senescent rats were killed when rapid weight loss had resulted in at least a 10%, but not >16%, reduction in body weight. The onset of senescence (as identified by rapid weight loss) occurred between 27 and 33 mo of age. The presenescent rats were killed between 27 and 30 mo of age to match the chronological age of the senescent rats. Young adult rats were killed between 8 and 10 mo of age. To determine the effect of weight loss per se, a group of young (8–10 mo; n = 10) and old presenescent (27–30 mo; n = 10) rats were food restricted to 70% of their ad libitum food consumption until they incurred a weight loss of approximately 10–12% of their prerestriction body weight (average time required was 10–14 days.) This food restriction regime resulted in a rate of weight loss that was similar to that for the senescent rats.
Perfusion and slicing.
Rats were anesthetized by injection with 0.1 ml/100 g body wt of a cocktail containing acepromazine (1 mg/ml), ketamine hydrochloride (71 mg/ ml), and xylazine (18 mg/ml). Animals were then killed by pneumothorax and perfused transcardially with 150 ml PBS (pH 7.3), followed by 150 ml of 4% paraformaldehyde in 0.1 M PBS. Each brain was immediately removed from the skull and fixed in 4% paraformaldehyde in 0.1 M PBS (pH 7.3) for 2 h. Brains were then cryoprotected via incubation in 25% sucrose solution for 24 h. Coronal sections were cut on a Leica CM1900 cryostat to 10 μm (internal temperature −16°C), mounted on polylysine-coated slides (Sigma Aldrich, Milwaukee, WI), and frozen at −70°C until analysis. Immunohistochemistry assays were begun within 7 days after sectioning.
Each animal was killed between 0900 and 1100 by carbon dioxide inhalation followed by decapitation. Brains were rapidly removed, frozen on powdered dry ice, placed into sterile vials (Fisher Scientific, Pittsburgh, PA) containing 70% ethanol in autoclaved, 0.1% diethylpyrocarbonate (DEPC; Sigma-Aldrich, Milwaukee, WI)-treated water, and stored at −20°C for up to 1 mo before use. Immediately before sectioning, brains were trimmed at the anterior and posterior ends using a Brain Blocker small animal stereotaxic instrument (Kopf Instruments, Tujunga, CA). The trimmed brains were then mounted on the vibratome tray with cyanoacrylic glue. Tissue, submerged in 70% ethanol-30% DEPC-treated water, was sliced to 300 μm (amplitude 8; speed 4) using a vibratome (Vibratome 1000 Classic CE; St. Louis, MO). The PVN was microdissected (punched) according to the method of Palkovits (26) and using the Paxinos and Watson (27) rat brain atlas to determine the location of the PVN. Briefly, microdissection was accomplished by placing the tissue slice onto a frozen plate on dry ice; PVN was then removed under a dissecting microscope using a 0.5-mm tissue microbore fitted with a stylet (Fine Science Tools). Each tissue slice was inspected and photographed after vibratome slicing and microdissection to verify accuracy and consistency of anatomic sampling. For each animal, all punches containing PVN were placed into a 1.5-ml centrifuge tube containing 250 μl of Trizol Reagent (Invitrogen, Carlsbad, CA) and frozen at −70°C for up to 2 wk before RNA isolation and reverse transcription (see below). Pilot studies in our laboratory have shown that RNA in such punches is stable for at least 3 yr at −70°C.
The PVN tissue punches were homogenized using a Kontes glass-glass homogenizer (size 20), with a total volume of 500 μl Trizol Reagent (Invitrogen), according to the provided directions for small quantities of tissue. For each sample, a series of centrifugation and extraction steps were sequentially completed until a pellet containing the majority of the RNA from the tissues remained. This pellet was then resuspended in 12 μl of DEPC-treated water, and the sample was used immediately for reverse transcription.
Reverse Transcription and cDNA Synthesis
Reverse transcription was carried out by the SuperScript II First Strand cDNA synthesis system for RT-PCR (GIBCO BRL, Rockville, MD). Instructions as well as all of the required reagents came with this kit except for the deoxynucleotide triphosphate (dNTP) mix and random primers. Briefly, in a 0.5-ml microfuge tube, 5 μl of resuspended RNA was combined with random primers, dNTP mix, and DEPC-treated water. This mixture was heated for 5 min at 65°C and then placed on ice for at least 1 min. The 5× First Strand Buffer and DTT were added to the tube, mixed gently, and spun quickly. Samples were reheated and cooled. Finally SuperScript II, RNase H, and RT were added to each tube for a final heating incubation (we extended heating at 42°C to 70 min). The reaction was terminated by incubation at 70°C for 15 min and then chilled on ice. Condensation was collected by quick centrifugation, and newly synthesized cDNA was frozen at −70°C until used for qPCR.
The sequence for the primers that were used for both the Y1 and Y5 receptors was taken from Beck et al. (2). The primers for S19 were designed by our laboratory using DNA strider 1.3 software for primer design (GenBank accession no. X56858). All primers were synthesized by Invitrogen. The sequences used were as follows: NPY Y1, forward TCTTCTCTGCCCTTCGTGATC and reverse TGAACGCCGCAAGTGATACA; NPY Y5, forward CAATACAGCTGCTGCTCGGA and reverse AAATCGTCTACGCTGCCTCTG; S19, forward CTTCGGCATGATCTATGATTCT and reverse TCCCCTGACCTTCTTCATTCTG.
Before amplification, first-strand cDNA was diluted 100-fold. For each cDNA, we prepared triplicate reactions containing 2× SYBR Green Master Mix [AmpliTaq Gold Taq DNA polymerase (Applied Biosystem, Foster City, CA), dNTPs, MgCl2], 0.3 U AmpErase Uracil N-glycosylase, and 800 nM of the previously described primer sets. Amplification and quantification of amplicon levels were performed using a GeneAmp 5700 Sequence Detection System (50°C for 10 min; 95°C for 10 min for 1 cycle; and 94°C, 62°C, 72°C each for 15 s for 40 cycles). Standard curves for each primer set, including S19, were determined by amplification of serial dilutions of cDNA from whole hypothalamus. Message levels for both the Y1 and Y5 receptors were expressed relative to S19 ribosomal protein expression for the same sample. Pilot studies in our laboratory (unpublished data) have shown that S19 mRNA expression remains relatively stable in food restriction.
Hypothalamic slices containing the PVN were identified using the fourth-edition stereotaxic atlas of Paxinos and Watson (27). The slices selected for each animal corresponded to bregma −0.40 mm to −1.80 mm (atlas plates 20–26). Alternate slices were stained with cresyl violet and microscopically examined to ensure consistent tissue-slice selection. Nine tissue slices (3 per slide) containing the PVN were selected for each rat for immunohistochemical analysis. A hydrophobic pen (Molecular Probes, Eugene, OR) was used to trace a barrier around each tissue slice, allowing reactions for each slice to occur separately. Sections were first made permeable with 0.5% Triton X-100 in PBS (Bio-Rad, Hercules, CA). Nonspecific binding sites were blocked with 10% normal goat serum (Molecular Probes) for 1 h before incubation with the primary antiserum containing affinity-purified polyclonal rabbit anti-NPY Y1 antibodies (diluted 1:1,000, final concentration). These antibodies detect a unique segment of the intracellular, COOH-terminal fragment of the NPY Y1 receptor (ImmunoStar, Hudson, WI). The secondary antiserum was goat anti-rabbit Alexa Fluor 488 (Molecular Probes). Due to the unavailability of specific antibodies against the NPY Y5 receptor, immunohistochemistry for this receptor was not done.
Tissue slices were incubated with the rabbit anti-NPY Y1 antiserum overnight at 4°C in a humid chamber. The following day, the sections were washed and reincubated with secondary antibodies (5 μg/ml; Molecular Probes) for 2 h at room temperature. Because others have shown that NPY Y1 receptors in the brain are found primarily on neuronal cells (5, 36), we used the counterstain Neuro Trace Fluorescent Nissl Stain at a dilution of 1:500 (Molecular Probes) to identify neurons. After staining, all slides were coverslipped using glycerol-based mounting media with or without the nuclear stain 4′,6-diamidino-2-phenylindole (DAPI) with antifade properties (Molecular Probes). These slides were then photographed, using a fluorescence microscope (Nikon E600) at 40× magnification, linked to a digital imaging system. Slides containing tissue slices were then cataloged and stored at −70°C.
To confirm that staining by the primary and secondary antibodies was specific for the NPY Y1 receptor, we included the following controls: 1) omitting the primary antiserum, 2) replacing the primary antiserum with normal rat serum, and 3) verifying reactivity and specificity of the antibody using protein analysis. These control experiments showed low levels of nonspecific immunoreactivity that were consistent across the groups.
Quantification of Immunohistochemically Labeled Neurons
All quantification was done from digital images to avoid fluorescence fading. Neurons in both the right and left regions of the PVN were counted, resulting in a total of 12 regions per animal (the bilateral regions on each of 6 tissue slices), which were then averaged. The microscope was fitted with an eyepiece (1 × 1 mm, 100 grid) that was positioned over the center of the PVN. This allowed for counting within areas of equal size. Image collection was completed within 3 h after staining although only minimal fluorescence fading was observed after 24 h. The digital images were assigned numbers so that those doing the counting were unaware of the age, feeding regimen, or physiological state of the animal. Each image was counted by at least two different individuals, with the results being averaged. In the event that the two counts differed by >10%, a third individual counted the image, and the two closest counts were averaged. Discrepancies requiring a third count occurred in only 4% of the samples (i.e., 2 of 50 rats). Tissue slices were critically evaluated for any morphological and/or size differences that appeared consistently throughout an age group and/or feeding regimen. No such differences were found.
ANOVA followed by Fisher's protected least significant difference test was used to evaluate the effects of age (young, old presenescent, old senescent) and feeding regimen (ad libitum and calorie restricted). Differences were considered significant at P ≤ 0.05. All means are presented ± SE of the mean.
The age at which spontaneous rapid weight loss began in the senescent rats varied from 810 to 990 days (27 to 33 mo) of age (Fig. 1). The senescent rats' loss of 10% of their presenescent body weight generally occurred within 7–10 days after entry into senescence. The average body weight of the old rats immediately before rapid weight loss was 318 ± 16 g. This differed from the group of rats that were still presenescent at death (365 ± 12 g) and from the young rats (379 ± 14 g). The senescent rats' weight loss averaged 12 ± 4% of their presenescent weight (with a range of 10–16%). The food-restricted young rats lost an average of 11 ± 4% of their prerestriction weight (276 ± 32 vs. 310 ± 36 g, respectively, with an average of 34 ± 8 g lost), while the presenescent rats lost 16 ± 6% of their prerestriction weight (308 ± 38 vs. 367 ± 41 g, respectively, with an average of 59 ± 44 g lost).
NPY Y1 and Y5 Receptor Expression in the PVN
As shown in Fig. 2, young ad libitum-fed rats had significantly higher levels of Y1 and Y5 receptor mRNA levels than did all other groups. Relative mRNA levels of Y1 receptors were similar in the presenescent and senescent rats as were those for the Y5 receptors. When we compared the ad libitum-fed groups with the food-restricted groups, differential responses to food restriction were observed. In the young rats, food restriction led to a substantial decrease in the expression levels of both Y1 and Y5 receptors, while in the presenescent animals, restricting food intake resulted in significant increases in Y1 and Y5 receptor mRNA in the PVN.
In all groups, Y5 receptor gene expression was significantly lower than that of the Y1 receptor, with the ratio of Y1/Y5 being significantly lower in the young (ad libitum and food restricted) than that in the old (presenescent, senescent, and presenescent food restricted) groups.
Quantification of Immunohistochemically Labeled Neurons
Number of neurons.
The number of neurons present in the PVN slices did not differ among the groups (Table 1).
Number of Y1 receptor-stained neurons.
Ad libitum-fed, presenescent, and senescent animals had significantly more cells staining for NPY Y1 receptors than did similarly fed young rats (Table 1, Fig. 3). Although significant differences were not found between the ad libitum-fed presenescent vs. senescent rats, the latter had significantly fewer immunoresponsive neurons than did the food-restricted presenescent rats. Similarly, while there were no significant differences in the numbers of stained neurons in the food-restricted presenescent vs. food-restricted young rats, both groups had significantly more cells that exhibited staining than did their age-matched ad libitum-fed counterparts. That is, food restriction caused an elevation in the number of neurons staining positive for Y1 receptors, but the reduced food intake of the senescent rats did not.
Number of Y1 receptor-stained neurons relative to the total number of neurons counted.
When comparing the ad libitum-fed groups, presenescent and senescent rats had the highest percentage of stained PVN neurons (∼53 and 48%, respectively, vs. 24% for the young rats). In both food-restricted groups, ∼56% of the neurons present in the PVN stained positive for the Y1 receptor.
The present investigation evaluated the possibility that decreased NPY Y1 and/or Y5 receptor gene expression and/or fewer neurons expressing Y1 receptors within the PVN could explain our previous finding of attenuated responsiveness of senescent rats to intracerebroventricular injection of NPY (3). Using qPCR, we found no significant differences in mRNA levels for Y1 or for Y5 receptors in the PVN of senescent vs. presenescent rats, although mRNA levels of these receptor subtypes were significantly lower than in the young rats (Fig. 2). In contrast, immunocytochemistry indicated that the number of PVN neurons staining for Y1 receptor protein was significantly greater in presenescent and senescent than in younger rats although there was no significant difference between the presenescent and senescent rats themselves. Thus the decreased responsiveness of senescent rats to intracerebroventricular injection of NPY cannot be explained by reduced mRNA level of Y1 and/or Y5 receptors or by fewer Y1 immunosensitive cells within the PVN. However, the fact that Y1 and Y5 receptor gene expression was lower in young vs. old rats (senescent and presenescent) may at least partially explain the gradual decline in food intake that has been observed in chronologically aged mammals (4, 22).
The hypothesis tested in this investigation was an outgrowth of results from several previous investigations. These included our findings that 1) senescent rats have significantly lower serum leptin concentrations than do their comparably aged presenescent counterparts, consistent with their diminished fat stores (10); 2) despite these lower concentrations of serum leptin, senescent rats had diminished rather than greater food intake (relative to presenescence) (10); and 3) food intake in response to intracerebroventricular injection of exogenous NPY was blunted in senescent rats (4). The results from the present study suggest that the blunted response of the senescent rats to intracerebroventricular injection of NPY may lie distal to the NPY-NPY receptor interaction within the neuron or may involve non-NPY pathways.
With respect to the latter, NPY acts in conjunction with a variety of interacting orexigenic and anorexigenic compounds and receptors (14, 15). These interacting pathways can enhance or suppress NPY's stimulatory effects on food intake by acting on the NPY pathway itself and/or on alternate pathways that can modulate or compete with NPY signaling. Our laboratory has begun investigations aimed at evaluating several such modulators of NPY signaling as well as the role of neural inhibition as a possible defect that may account for the blunted NPY response observed in the senescent rat.
Although our results do not explain the differences in NPY responsiveness between presenescent and senescent rats, we did find that NPY Y1 and Y5 mRNA levels were significantly lower in old (presenescent and senescent) vs. young rats. Thus fewer functional NPY receptors may account, at least in part, for the gradual loss in body weight and food intake observed in aging mammals (4, 22). Further investigation will be needed to test this hypothesis.
Despite the lower NPY Y1 receptor mRNA levels in the PVN of the older (both presenescent and senescent) rats, these older rats had more PVN neurons that stained for NPY Y1 receptors than did the younger rats. This apparent paradox suggests that the old rats have more neurons producing Y1 receptors but at lower levels than do the young rats. The recruitment of more neurons transcribing Y1 receptors may be a compensatory response to less efficient transcription in the older cells.
We also found that the number of Y1 immunoresponsive cells increased in both young and old presenescent rats with short-term food restriction. This finding is consistent with previous investigations describing increased Y1 in the PVN and other brain regions with food restriction (13, 35) and the view that more neurons producing Y1 receptors during food restriction would allow for greater NPY responsiveness once food became available. Interestingly, while the number of responsive cells increased in food-restricted young and old rats, mRNA levels for Y1 and Y5 receptors decreased in the PVN of the young but increased in the old rats. The explanation for this differential response to food restriction is unclear. One possibility is that synthesis of NPY Y1 and Y5 receptors in young ad libitum-fed rats is limited to a relatively few PVN neurons (as alluded to above), which make significant amounts of the receptors, but in times of food shortage more cells may be recruited to produce these receptors but only at low levels. This suggestion is consistent with our results showing lower overall quantities of mRNA with food restriction, yet higher numbers of cells staining for the receptors (Table 1 and Fig. 3). Similar observations of increased receptor staining with decreased mRNA production have recently been reported for orexin receptors in response to food restriction (19).
The fact that food-restricted presenescent rats that had lost a comparable proportion of body weight as did the senescent rats had higher levels of receptor expression as well as more cell staining than did the senescent rats is further evidence that the senescent rats do not behave physiologically as if they were food restricted. Thus, while appropriate feeding cues remain in the chronologically old presenescent rats, they appear to be significantly diminished (or absent) in the senescent animals.
In conclusion, the attenuated responsiveness of senescent rats to intracerebroventricular injections of NPY cannot be explained by decreased NPY Y1 and/or Y5 receptor gene expression or diminished Y1 receptor immunoreactivity in PVN neurons. Rather, our data suggest that components downstream of the NPY receptor and/or alterations of other modulators of food intake are likely to be responsible for the altered response to NPY in senescent rats. Our data also demonstrate that the responses of the senescent rats are not due simply to loss of body weight but rather reflect an altered functional state distinct from food-restricted presenescent animals. Finally, we found that the gene expression of NPY Y1 and Y5 receptors is significantly lower in old (presenescent and senescent) vs. young adult animals, a difference that may contribute to the gradual decrease in body weight and food intake often observed in chronologically old mammals.
An age-related loss in body weight, either slow and prolonged or rapid and spontaneous near the end of life, appears to be a normal consequence of aging in both humans and laboratory rodents (1, 11, 18, 20, 22, 24, 32). The rapid and spontaneous body weight loss, a phase that we refer to as senescence, is concomitant with similar declines in food intake, cold-induced thermoregulation, and circadian rhythmicity of body temperature (3, 4, 22, 23). The fact that all of these functions have significant points of regulation in the hypothalamus suggests the possibility of a generalized dysfunction within this brain region. Such wide-ranging dysfunction may indicate disruptions in critical physiological/biochemical pathways that are common to all of the affected systems. We believe that future investigations should focus on these basic mechanisms as a means to identify the commonalities of age-related dysfunction.
This work was supported by National Institute on Aging Grant AG-06664.
We thank T. Landerholm, E. Hernandez, S. M. Chau, S. Hamilton, E. Bautista, and C. Craig-Veit for technical assistance and insightful comments.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2004 the American Physiological Society