Effects of dietary restriction on motor learning and cerebellar noradrenergic dysfunction in aged F344 rats

BRAIN RESEARCH ELSEVIER

Brain Research 684 (1995) 150-158

Research report

Effects of dietary restriction on motor learning and cerebellar noradrenergic dysfunction in aged F344 rats Thomas J. Gould a, Kathryn E. Bowenkamp a, Gaynor Larson a, Nancy R. Zahniser a, Paula C. Bickford a,b,* a Department of Pharmacology, University of Colorado Health Sciences Center, 4200 East Ninth Avenue, Denver, CO 80262, USA b Department of Veterans Affairs Medical Center, Research Service, 1055 Clermont Street, Denver, CO 80220, USA Accepted 14 March 1995

Abstract Fisher 344 rats were fed either ad libitum or with a diet containing a 40% reduction of calories beginning at 4 months of age. At 14 months and 22 months male rats were tested for their ability to learn a complex motor skill. At both ages the diet restricted rats reached criterion of performing 10 successful crosses in 10 min at an earlier time than ad libitum fed controls. At 22 months of age the diet restricted rats showed improved acquisition of running times for the task. Male rats at 14 and 22 months and female rats at 24 months were examined electrophysiologically for the ability of isoproterenol to augment the action of GABA in the cerebellum when both substances were applied iontophoretically from an extracellular multibarreled glass electrode. In all 3 age and sex groups there was an improvement in the /3-adrenergic receptor modulation of GABA responses in the dietary restricted vs. ad libitum rats. However, no difference was observed between dietary restricted and ad libitum rats when the number and affinity of cerebellar/3-adrenergic receptors was assessed with 125I-iodopindolol binding. Overall, there was a significant improvement in cerebellar noradrenergic function in the dietary restricted rats and this was accompanied by an improvement in motor learning.

Keywords: Cerebellum; Aging; Motor learning; Norepinephrine; Rat; fl-Adrenergic receptor

1. Introduction Aging leads to functional changes in both brain and behavior. Specifically, aged-related changes are seen in Purkinje cell noradrenergic function. In aged rats, the ability of norepinephrine (NE) to modulate GABAergic inhibition of Purkinje cells is decreased [3]. This alteration has been associated with a deficit in the /3-adrenergic receptor [34,35]. Moreover, in the cerebellum, NE function is related to the rate of acquisition during motor learning [2,45,46]. Norepinephrine may facilitate learning by enhancing the signal to noise ratio of cerebellar afferents during acquisition of a motor task. For example, when NE is absent in cerebellum, climbing fiber augmentation of movement evoked mossy fiber input to Purkinje cells is decreased [2,30]. During local application of NE, however, parallel fiber-driven basket cell inhibition of Purkinje cell

* Corresponding author. Dept. Pharmacology, UCHSC, 4200 East Ninth Ave, Denver, CO 80262, USA. Fax: (1) (303) 331-8324. 0006-8993/95/$09.50 © 1995 Elsevier Science B.V. All rights reserved SSDI 0 0 0 6 - 8 9 9 3 ( 9 5 ) 0 0 4 0 7 - 6

firing is increased [14,16]. Furthermore, aged rats have a deficit in motor learning that is correlated with decreased cerebellar fl-adrenergic receptor function [1]. In a task that required rats to learn to run on unevenly spaced pegs, aged rats acquire proficiency at the task significantly more slowly than young rats. The aged rats, however, learned the task to a level similar to the young rats. Thus, the age-related deficit in /3-adrenergic receptor function is correlated with a deficit in the learning rate but not the ability to learn the task. This is identical to what is observed following depletion of NE with the neurotoxin 6-hydroxydopamine or injection of a fl-adrenergic receptor antagonist [2,4]. Thus cerebellar fl-adrenergic receptors appear to be particularly sensitive to the effects of aging and decreased fl-adrenergic receptor function correlates with decreased rates of motor learning. The mechanism(s) producing the deficit in fl-adrenergic receptor function, however, remain unknown. In the field of aging research, one simple manipulation, diet restriction, has consistently produced beneficial results that counter some of the deleterious effects of aging. In

151

Thomas J. Gould et al. /Brain Research 684 (1995) 150-158

1935 McCay et al. showed that diet restriction extended both the mean and maximum life span of rodents [29]. Diets that restricted Calories nearly doubled the mean life span of male rats compared to ad lib fed males. Since that time a tremendous amount of research has confirmed that diet restriction increases life span [47]. Whereas it is possible to manipulate many dietary components, the most effective form of diet restriction is not malnutrition but instead is a reduction in calories. For example, restriction of fat alone or protein alone did not increase life span in hamsters [5,6]. In contrast, Leto, Kokkonen and Barrows reported that mice fed 22% less protein had increased life-span [25]. However, several labs reported that protein restriction was not nearly as beneficial as caloric or total diet restriction [10,28,53]. Thus, caloric restriction is the most beneficial manipulation. As an example, in mice fed a diet 40% lower in calories compared to ad lib fed control animals, life span was 43% greater [41]. Other factors also contribute to the effectiveness of dietary restriction. The percent of diet restriction and the time-course of implementation are important determining factors in the effectiveness of diet restriction. One study found that a 33% reduction was beneficial but a 50% reduction was harmful [23]. Similarly, another study found that rats fed with a diet reduced 40% live longer than rats fed a diet reduced by 20%. Finally, Cheney et al. found that whereas diet restriction during any part of the life-span had beneficial effects, diet restriction throughout the life-span had the most benefit [8]. In addition to increasing life span, diet restriction has been shown to have beneficial effects on other biological processes. Ross and Bras found that rats fed a diet restricted in calories had fewer tumors than ad lib fed rats [38]. In addition, diet restriction improved immune response by decreasing age-related declines in antigen presentation and T-cell proliferation [12]. Diet restriction also improved cardiovascular function. Heart rate and baroreflex responsiveness to hypertensive stress were improved in diet restricted animals [20,42]. Furthermore, caloric restriction reduced age-related deficits in DNA repair [48]. Thus, reduction of calories improves life span and reduces disease but the mechanism,~ through which diet restriction works and the entire beneficial extent of diet restriction are not completely understood. The purpose of this study was to examine the effects of caloric restriction on the age-related deficit in cerebellar noradrenergic function described above and to determine if caloric restriction was associated with improvement in motor learning in aged rats.

Rats were maintained on a 12 h / 1 2 h light-dark cycle. Ad lib fed animals were given free access to food and water whereas diet restriction animals had free access to water but were on a diet that had a 40% reduction in calories. Animals were fed at lights out. The diet was prepared by Teklad and for the ad lib animals consisted of 19.21% protein (casein 218 g / k g , DL-methionine 3.0 g / k g ) , 9.38% fat, 0.653% calcium, 0.5% phosphate and 4.023 kcal/g. The diet for the diet- restricted animals consisted of 32.0% protein (casein 366.3 g / k g , DL-methionine 5.0 g / k g ) , 12.54% fat, 1.088% calcium, 0.836% phosphate and 4.023 kcal/g. The diet restriction diet was fed at 60% of the grams consumed by the ad lib fed animals. This produced a 40% reduction in calories while maintaining comparable levels of nutrients. Diet restriction began at 4 months of age and continued throughout the experiment. Eight diet restricted males (DRM14) and 10 ad lib males (ALM14) were behaviorally tested at 14 months of age and one rat was lost from each group before electrophysiology. Eight diet restricted males (DRM22) and 7 ad lib males (ALM22) were behaviorally tested at 22 months of age and one rat was lost from the diet restricted group and 2 rats were lost from the ad lib fed group before electrophysiology. Five diet restricted females (DRF24) and 5 ad lib fed females (ALF24) were electrophysiologically tested at 24 months of age. 2.2. Motor learning task

Twelve hours prior to training, water was withdrawn from free access. Rats were maintained at 85% of initial body weight and they received 3 min of ad lib access to water after each training session. Training consisted of two stages. Initially rats were shaped and trained on a regular pattern of peg placement in the runway and then rats were tested on an irregular pattern of peg placement. Only 14-and 22-month-old male rats were behaviorally tested. The task required the rats to negotiate a narrow runway by walking on spaced aluminum pegs in order to receive a water reward (Fig. 1). Sessions were run and data was

A. REGULAR PATTERN (REG) .., ,, ,,I ,, I,,,, ,,,, I I,,,, m I

2.1. Subjects

Both male and female Fisher 344 rats were tested. Rats were housed in an AAALAC approved, barrier facility.

i

]3. I R R E G U L A R

PHOTOBEAM DETECTOR

J

•

WATER SPOUT

l

PATTERN

(IRR) END PLATFORM

I

2. Methods

l

,,,,,,,,,,,a,,,,i

I

T I

I I i It i II I II I I I II I I I I I I I I I I I I I I 11 I I I I I

~

,

HORIZONTALROD

~ l J i i Fig. 1. Overheadview of the regular pattern (A) and the irregularpattern (B) of rod placement used in the runway motor task. The 4 photocells placed 20 era apart with the beams 3 era above the tops of the rods measured running times from goal box to goal box.

152

Thomas J. Gould et al. / Brain Research 684 (1995) 150-158

collected by a computer. Infrared photo detectors placed along the runway were used to track the rats' movements. When a rat reached a goal box, 0.3 ml of water was delivered along with a loud sound that served as a conditioned reinforcer. A training session lasted until a rat made 20 crosses or the rat was in the runway for 30 min. Initially, the rods were covered with Plexiglas and the rats were trained to cross the runway until they consistently negotiated the runway or spent a week at this stage. During the next phase of shaping, sections of Plexiglas were removed until all the pegs in the regular pattern were exposed. Rats were tested on the completely exposed regular pattern for 5 days. After shaping, rats received a 2-week break from the runway during which they had ad lib access to water. Twelve hours before testing on the irregular pattern started, water was withdrawn. During the testing phase, rats were run on the irregular pattern until individual running times reached asymptotic levels. Group and individual rates of acquisition were determined by plotting average daily performance versus time. Each daily performance was based on a minimum of 6 crosses and a maximum of 20 trials. Only trials in which the rats did not pause were used for analysis. The running times per day were fit using a nonlinear mixed effects model of Hirst et al. [21]. The exponential decay model was of the form Y = A + ( 1 0 0 - A ) e C ( X - l ~ + S, where Y is percent initial run-time, X is the days of training, A is the asymptote as X approaches ~, C is the rate constant describing the rate of improvement in running time and S is the total standard deviation about the model. The decay constant is then an index of the rate of learning for further analysis. This parameter was equivalent to the rate of learning and was used to detect differences between diet restricted and ad lib fed groups. In addition, days to reach a learning criterion of 10 crosses or more in a session on the irregular pattern were compared. The 14-month-and 22-month-old animals were combined to provide a number large enough for a non-directional X2 test. The time spent in goal boxes, number of valid trials and actual running times were compared with an ANOVA for DRM14 months vs. ALM14 months and DRM22 months vs. ALM22 months. After testing was completed, all rats were tested on balance and strength measures. Rats were tested on round balance beams as adapted from Wallace et al. [44]. A 60-cm wooden dowel of either 2.5 or 5 cm diameter was suspended between two safety platforms 60 cm above a padded surface. A rat was placed in the center of the rod and the latency to reach a platform or fall was recorded. Two min was the maximum length of a trial and a rat that reached a platform was given the maximum score of 120 s. Each rat was given 3 trials and assigned the best score. An additional index was recorded using the following scale: 0 = fall, 1 = clasp, 2 = all paws on top, 3 = takes steps, 4 = reaches platform. Rats were also tested on a wire hang. The forepaws of a rat were placed on a horizontal

wire of 2 mm diameter suspended 50 cm above a padded surface. The rat was released and the latency to fall was recorded, The final test was the inclined screen. Rats were placed on a screen inclined at a 60 ° angle and the latency for the rat to climb down was recorded. A maximum of 2 min on the inclined screen was allowed. 2.3. Electrophysiology

Prior to electrophysiological testing, the male rats were 15 months and 24 months of age because of the length of the behavioral testing procedure. In addition, 24-month-old female rats were also electrophysiologically tested. Rats were anesthetized with urethane (0.75-1.25 g/kg), intubated and allowed to breath spontaneously. Aged rats required lower doses of anesthetic to induce an equivalent level of anesthesia. Corneal reflex and toe pinch were used to monitor anesthetic level. A heating pad was used to maintain body temperature at 37 ° C. Animals were placed in a stereotaxic frame and the skin and muscle over the posterior cerebellum was removed. The cistern was drained and the skull and dura over the cerebellum was removed. A solution of 2% agar in saline covered the brain. Recordings were made in either lobules VI and VII of cerebellar vermis (DRM14 and ALM14) or paramedian lobe (all others) from Purkinje cells as identified by anatomical location and the characteristic complex spiking of Purkinje cells [11]. Neuronal signals were amplified and filtered ( - 3 dB at 0.3 and 5 kHz) and displayed on a storage oscilloscope. Action potentials were isolated using a window discriminator and the output was displayed using a strip chart recorder. Single units had to have a signal to noise ratio of at least 2:1. Multibarrel glass micropipettes were used for single cell recording and drugs were applied locally via microiontophoresis. The resistance of the recording electrodes was 1.5-3.3 M O. In the multibarrel glass micropipettes, two barrels were filled with 3M NaC1 and the other 2 barrels were filled with GABA (0.25 M, pH = 4.0-4.5) and with the /3-adrenergic agonist isoproterenol (ISO) (0.25 M, pH = 4.0-4.5) respectively. A constant-current source provided ejection and retaining currents for the drug barrels and passed an equal current of opposite polarity through the balance barrel to neutralize the tip potential [40]. Uniform pulses of drug were applied at regular intervals [15]. Current was adjusted until GABA produced a 10-40% inhibition of Purkinje cell firing. The dose of GABA was recorded and four applications were given before ISO was co-administered. The level of ISO was adjusted until a change in GABA induced Purkinje cell inhibition was seen or the until baseline Purkinje cell firing rate altered. The ISO was then given continuously and four samples of GABA with ISO were taken. After the fourth sample, ISO was turned off and GABA was given until it could be determined if the pre-ISO level of GABAergic inhibition

Thomas J. Gould et al. / Brain Research 684 (1995) 150-158

would return. Only cells in which the post-ISO level of GABAergic inhibition matched the pre-ISO level of GABAergic inhibition were analyzed. Drug-induced responses were quantified by computer. The rate meter data were digitized and the percent inhibition of firing rate resulting from drug applications were calculated on a computer. ANOVAs were' used to detect differences in ISO-induced modulation of GABAergic Purkinje cell inhibition between groups.

"

z-=°

DRM

1OO.o

O:

80

"

7o

I

T

.

l-l Ul 0

80

=

so

T

/ 2

3

4

5

6

7

8

DAYS

Fig. 2. Running times of the 22-month-old diet restricted (circles DRM) and the ad lib fed rats (crosses ALM) on the irregular pattern. The data were fit to an exponential decay curve, described in the methods section, the curve for the diet restricted rats Y = 5 1 . 9 3 + ( 1 0 0 - 5 1 . 9 3 ) × e(-0.427×(x- 1)) ± 11.5) was significantly different from that for the ad lib rats Y = 26.96+(100-26.96)× e ( - ° ' l * r x ( x - 1)) ± 11.5). The decay constants or rate of learning (diet restricted - 0 . 4 2 7 and ad lib - 0 . 1 4 6 ) were also different between the groups. Error bars reflect standard error of the mean on individual days.

learning than aged ad lib fed male rats. In addition, both DRM14 and DRM22 spent less time in the goal boxes compare to their ALM controls ( P < 0.01, ANOVA). This could indicate motivational differences or a better comprehension of task requirements. Animals were not different in the amount of water consumed. Also, the DRM22 had more valid trials (i.e. trials in which rats did not pause) (mean= 18.91) then the ALM22 (mean= 17.51) ( P < 0.005, ANOVA) but no difference in valid trials was seen between 14 month old rat groups. Finally, DRM22 had significantly faster actual running times (mean = 1.95 s) than ALM22 (mean = 2.64 s) ( P < 0.05, ANOVA) but no difference existed between 14-month-old rat groups. Rats were also tested for differences in strength and coordination. For the 22-month-old male rats, no differDAY 1

~

DAY 2

~

DAY 3

~

DAY 4

100

3. Results

The DRM22 rats learned the motor task significantly faster than the ALM22 (P < 0.05, X 2 test) (Fig. 2). The overall curves were different with the main difference being in the slope of the curve ( P < 0.02) rather than the asymptote. No difference was found between the learning curves for the DRM14 and ALM14. However, overall the diet restricted males in both age groups were quicker to reach the criterion of at least 10 crosses per session compared to ALM ( P < 0.001, X 2 test) (Fig. 3). Thus aged diet restricted male rats perform better in motor

T T

1

3.1. Behavioral data

ALM

+

z_

2.4. Radioligand binding to fl-adrenerg!c receptors Following electrophysiological recording, the brains were rapidly removed from the 24-month-old male rats; and the cerebella were frozen. Cerebellar membranes were isolated by homogenization in 20 mM Hepes buffer (pH 7.5) containing 154 mM NaCI (tissue buffer) and centrifugation at 20,000 × g for 10 min. Duplicate samples (approximately 40 tzg protein) were incubated with 1251iodopindolol (IPIN, 2.2 Ci//xmol; Dupont New England Nuclear, Boston, MA) and drugs in tissue buffer in a total volume of 250 /xl for 30 min at 37 ° C. For saturation curves, nine concentrations of IPIN ranging from 10 to 900 pM were used. For competition curves, a single concentration of IPIN (100 pM) and 14 concentrations of the selective fll-adrenergic receptor antagonist ICI 89,406 (gift from Imperial Chemical Industries PLC, Cheshire, Great Britain) ranging from 46 nM to 10 /zM were used. Nonspecific binding was defined in the presence of 1 /~M /-propranolol (Research Biiochemicals, Natick, MA). The incubations were terminated by addition of 5 ml of 10 mM Tris-HCl buffer (pH 7.5) containing 154 mM NaC1 (room temperature), filtration over glass fiber filters (Schleicher and Shuell #30, Keene, NH) and washing with an additional 5 ml buffer. Radioactivity trapped on the filters was quantitated with a gamma counter. Proteins were determined by the method of Bradford using BSA as the standard [7]. The data were analyzed using iterative curve fitting (GraphPAD Software, San Diego, CA). Student's t-tests were used for statistical analysis•

153

90 <

8O

~

70

~ ~

6o so

m o

40

~

3o 2O

10 DRM14

I

ALM14

I DRM22

I ALM22

GROUPS

Fig. 3. Days to criterion for the motor learning task (10 or more crosses) for 14-month-old and 22-month-old diet restricted males versus ad lib male F344 rats. Overall, the diet restricted rats reached criterion signifi• cantly faster than ad lib rats ( P < 0.001, X2).

154

Thomas J. Gould et al. / Brain Research 684 (1995) 150-158

ences in performance on the wire hang, large balance beam or small balance beam were found. On the inclined screen, however, DRM22 spent more time then ALM22 ( P < 0.05, t-test). For the 14-month-old rats, no difference in performance between DRM and ALM was found for the wire hang, the large balance beam or the inclined screen. The DRM14, however, spent more time on the small balance beam than ALM14 ( P < 0.01, t-test). Overall, no consistent difference was seen in task performance between ALM and DRM animals. This suggests that the differences found in acquisition of the learned motor task were not simply due to differences in motor ability.

I A GABA0 nA DRF24

ls.6..~

B

18.9%

34.7%

20.8%

ISO 2S nA

3.2. Electrophysiological data Extracellular recordings were made from cerebellar Purkinje neurons. The ability of ISO to modulate the action of locally applied GABA was evaluated as an index of /3-adrenergic receptor function. Fig. 4 compares a cell recorded from a DRF24 to a cell recorded from an ALF24. In the diet restricted rat (Panel I) locally applied GABA produced a 22% inhibition of firing rate in the baseline condition. When ISO was concomitantly applied, the GABA response increased to an average 92% inhibition. The GABA response returned back to control levels following termination of the ISO. Thus in this cell ISO modulated GABA responses in a manner similar to that observed in young animals [1,50]. In contrast, the cell recorded from the ad lib animal (panel II) does not demonstrate a modulatory action of ISO similar to the pattern observed in young rats. In this case GABA produces a similar inhibition of firing rate prior to ISO. During ISO application the response to GABA is diminished from an average of 53% inhibition to an average of 23% inhibition, an effect opposite to that observed in young rats or in diet restricted rats of the same age. The ability of ISO to modulate GABAergic inhibitions was tested in all groups and results similar to those illustrated above for the female rats were observed. The ability of GABA to inhibit Purkinje cell firing was not different between the diet restricted groups and the ad libitum fed groups (average inhibition before ISO was 37.5 + 1.79%; no difference was noted between groups). The effects of ISO on the modulation of GABAergic inhibition of Purkinje cells, however, was significantly different between groups ( P < 0.0001, ANOVA; Fig. 5). Post-hoc comparison with the Tukey HSD test revealed that co-application of ISO significantly increased GABAergic inhibition in DRM14 compared to ALM14 ( P < 0.0005). Similarly, ISO significantly increased GABAergic inhibition in DRM22 compared to ALM22 ( P < 0.01, Tukey test) and ISO significantly increased GABAergic inhibition in DRF24 compared to ALF24 ( P < 0.0005, Tukey test). The data were also analyzed by looking at the percent of cells in which ISO either produced a minimum of a 15% increase in inhibition, no change in inhibition or a

C

20 Hz 20 see

II A

GABA 23 mA 54.3%

~__=~%

.~1.1%

59.4%

B lso 122 mA 14_"t,/,

z,.._s%

7-3._.s%

~o.4._%

C 62.~,z,

~.~,/,

.%.,%

42.A'/,

20HI Fig. 4. Continuous ratemeter records from a diet restricted 24-month-old female (I DRF24) and an ad lib fed 24-month-old female (II ALF24). For the DRF24 (I), GABA (0 hA) produced an average 22.5% inhibition of firing rate (A) and during co-application of ISO (25 nA) inhibition of firing rate increased to an average 92.0% (B). For the ALF24 (II), GABA (23 hA) produced an average of 53.8% inhibition of firing rate (B) but co-application of ISO (122 hA) decreased the firing rate to an average of 23.0%. For both rats, after ISO ejection ceased, GABA response returned to control levels (C). The Purkinje cell from the DRF24 shows a response similar to that observed in young rats, whereas the cell from the ALF24 is similar to what has previously been reported in aged F344 male rats. The bars over the ratemeter indicate when drug was being applied and the numbers directly over the bar indicate the percent inhibition. The dose of GABA remained constant throughout the trial. The vertical calibration bar indicates the firing rate of the Purkinje neurons in spikes per second; the horizontal calibration bar indicates time.

Thomas J. Gould et al. / Brain Research 684 (1995) 150-158 15% or greater decrease in inhibition. Using the above categorization, analysis with a non-directional X 2 test revealed that the DRM14 and A L M 1 4 groups were significantly different from each other ( P < 0.001) the DRM22 and A L M 2 2 groups were also significantly different from each other ( P < 0.001) as were the DRF24 and A L F 2 4 groups ( P < 0.05). For all three age and sex groups, diet restricted animals had more cells for which ISO increased G A B A e r g i c inhibition of Purkinje cells and less cells for which ISO decreased GASBAergic inhibition of Purkinje cells compared to ad lib fed animals. See Table 1 for the percent of cells in each category. Included in Table 1 are numbers from young male rats for comparison, as can be seen the data obtained from the diet restricted rats is very similar to that previously observed in young rats.

3.3. Radioligand binding data The affinities and densities of/3-adrenergic receptors in the cerebellar membranes from A L M and D R M rats were determined from IPIN saturation curves. These curves were best fit to a single site. Neither the affinities ( A L M Kd: 210 + 75 pM; D R M Kd: 220 + 59 pM) nor the densiDRMorDRF

~

ALMorALF

45

Table 1 Percent cells that ISO produced an increase of 15% or greater, a decrease 15% or less or produced no change in inhibition

DRM14 ALM14 DRM22 ALM22 DRF24 ALF24 Young male a

( + )15% or > change

( -)15% or < change

No change

65% 17% 60% 35% 78% 53% 71%

5% 17% 2% 25% 5% 12% 5.5%

30% 66% 38% 40% 17% 35% 24.5%

a data from Lin et al. [26]

ties ( A L M Bma~: 36 + 6.1 f m o l / m g protein; D R M Bma~: 32 + 4.5 f m o l / m g protein) differed between the two groups. In an attempt to investigate potential changes in /31-adrenergic receptors, competition curves were carried out with the selective /31 receptor antagonist ICI 89,406. These curves plateaued at 80% inhibition of IPIN binding, similar to the level of binding observed in the presence of 1 /zM L-propranolol. These curves were also best fit by a single site model. This site had similar affinities for ICI 89,406 in the A L M ( K i = 77 + 8.6 nM) and D R M ( K i = 80 + 7.6 nM) groups.

3.4. Survival analysis and body weights

40 35

30

20 15 10 5 o

155

T 14 Month Male

2:2 Month Male

2 4 Month Female

Fig. 5. Summary of the percent change of the GABA response during application of ISO in cerebellum. Data were collected as in Fig. 4. The effects of ISO to modulate GABAergic inhibition of Purkinje cells was significantly different between groups (P < 0.0001, ANOVA). Forty cells were recorded from the DRM14, 47 cells from the ALM14, 50 from the DRM22, 40 from the ALM22, 41 from the DRF24 and 40 from the ALF24. The ability of GABA to inhibit Purkinje cell firing was not different between DRM14 (39.95% average + 2.14% S.E.M. inhibition at an average GABA dose of 22.68 nA) and ALM14 (36.1+1.56% at 125.17 nA GABA), between DRM22 (37.34 + 1.71% at 26.36 nA GABA) and ALM22 (40.80 + 1.89% at 30.88 nA GABA) and between DRF24 (34.95 + 1.81% at 17.20 nA GABA) and ALF24 (35.99 + 1.67% at 30.03 nA GABA). Post-hoc comparison with the Tukey HSD test revealed that co-application of ISO significantly increased GABAergic inhibition in DRM14 (62.30+3.41% at an average ISO dose of 73.62 hA) compared to ALM14 (36.20+3.18% at 334.35 nA ISO) (P < 0.0005). Similarly, ISO significantly increased GABAergic inhibition in DRM22 (61.47+ 3.21% at 58.34 nA ISO) compared to ALM22 (48.34+3.84% at 69.25 nA ISO) (P <0.01, Tukey test) and ISO significantly increased GABAergic inhibition in DRF24 (69.58+4.10% at 73.23 nA ISO) compared to ALF24 (50.01 + 4.30% at 76.3 nA ISO) (P < 0.0005, Tukey test).

The rats studied in this manuscript were part of a larger group of rats on dietary restriction housed in the U C H S C animal care facility. A s part of this larger study 10 rats in each sex and diet group were set aside for survival analysis. A t the time of writing this paper several of diet restricted rats remained alive. A K a p l a n - M e i e r survival analysis such as described by Lee was used to analyze survival as this approach is appropriate for right censored survival data and is suited for a small sample size [24]. The median survival times ( + asymptotic standard error) were 109 __+1.58 weeks for A L F rats and 124 _ 5.8 weeks for D R F rats ( P < 0.05); for male rats median survival times were 103 + 3.69 weeks for A L M rats and 1 1 7 + 2.37 weeks for D R M rats ( P < 0.02). Thus both diet restricted male and female rats showed significantly longer median survival times. Throughout the experiment, b o d y weights were recorded. Because the standard deviations were non-Gaussian, a K r u s k a l - W a l l i s test was used to detect significant differences between groups. No difference in body weights was found between D R F and A L F rats but body weights were significantly different between A L M and D R M ( P < 0.05) (Fig. 6). A t the time of initiation of behavioral testing the mean body weight for A L M 1 4 was 461.40 g + 4.28 g, for A L M 2 2 was 439.85 g + 12.80 g, for DRM14 was 335.67 g + 1.66 g, and the mean body weight for D R M 2 2 was 347.73 g + 4.51 g. A t the time of initiation of electrophysiological testing the mean body weight

156

Thomas J. GouM et aL / Brain Research 684 (1995) 150-158 --e--

ALF

s°° I

--l--

--0--

ALM

--D--

DRF

DRM

~jmmrm.=.='u',~,~_._¢~i.i,i._ T

300

<

100

0 1

,

4

i 8

,

i 12

,

l 16

,

i

,

20

i 24

,

i

28

,

i

32

,

i

36

Months of Age

Fig. 6. Body weights were measured and recorded across time for ALF (filled circles), ALM (filled squares), DRF (open circles) and DRM (open squares) groups. Statistical analysis with a Kruskal-Wallis test found no difference between ALF and DRF groups but a difference existed between ALM and DRM groups (p < 0.05). The precipitous fall in weight at the end of the ALM and DRM curves reflect a small number of rats (n = 2) that lost weight prior to death.

for ALM14 was 466.60g + 5.20g, for ALM22 was 414.31 g ___14.78 g, for ALF24 was 319.68 g + 11.11 g, for DRM14 was 343.75 g -t- 2.74 g, for DRM22 was 330.60 g + 6.26 g and the mean body weight for DRF24 was 236.05 g -t- 3.01 g.

4. Discussion This present study showed that a regiment of 40% caloric restriction that began at 4 months of age and continued until conclusion of the experiment prevented or slowed aged-related deficits in neural function and in motor learning. Cerebellar /3-adrenergic receptor function assessed electrophysiologically was improved by 17-22% overall in diet restricted groups. Thus, for both male and female rats, diet restriction slowed or prevented loss of cerebellar /3-adrenergic receptor function. Also, it was found that diet restriction improved learning of a complex motor task. In 14-month-old male rats no difference was seen in the rate of acquisition of the complex motor task but a difference was observed in days to criterion with the DRM14 reaching criterion quicker. The 14-month-old rats may not have been old enough for differences in the rate of acquisition to be detected because a large deficit in learning was not observed in the ALM14 rats. In the 22-month-old rats differences existed in the learning curves. The DRM22 group had a significantly improved rate of acquisition when compared with the ALM22 group. In addition, the DRM22 rats also reached the criterion of 10 or more crosses in a session faster than the ALM22 rats. Thus, diet restriction delayed age-related declines in motor learning. One possible explanation for the differences in motor learning between groups could be a difference in strength

and coordination. In tests of strength and coordination, however, no consistent differences were observed. Because differences were not found on most of the tasks and differences seen were not the same across age groups, the improvement in motor learning in diet restricted rats was most likely not the result of a general motor deficit but instead primarily a learning deficit. In addition, prior studies have demonstrated that age-related deficits in learning the irregular pattern were separate and distinguishable from changes in strength and coordination [2,4]. The deficits in motor learning for the present task are associated with deficits in cerebellar/3-adrenergic receptor function. Although correlation does not necessarily mean causation, the data do support a possible connection between cerebellar/3-adrenergic function and learning. In studies comparing Purkinje cell function between young and old rats, a deficit was seen in /3-adrenergic receptor function [3,34,35]. One index of fl-adrenergic receptor function is to examine the modulatory effects of this receptor on other neurotransmitter systems. We have examined the ability of ISO to augment GABAergic inhibition of cerebellar Purkinje cell-firing rates. In aged rats there is a diminished effect of ISO to augment GABAergic transmission. The diet restricted animals had improved fl-adrenergic receptor function compared to aged-matched ad lib fed controls. In addition, it has been demonstrated that the age-related deficit in /3-adrenergic receptor function correlated with motor learning deficits [1]. Other studies have also shown fl-adrenergic involvement in motor learning. In rabbits, injection of the fl-adrenergic antagonist propranolol into the flocculus impaired plasticity of the vestibulo-ocular reflex [43]. Also, administration of propranolol to young rats delayed motor learning [4]. Furthermore, norepinephrine depletion in young rats impaired motor learning producing an average learning curve similar to that recorded from aged rats [2]. Enhancement of GABAergic inhibition by /3-adrenergic processes may facilitate learning and thus disruption of or deficits in /3adrenergic receptor function may retard learning. In the present study, the aged diet restricted rats were better at motor learning than the ad lib fed controls. Thus, diet restricted rats resemble young rats both in /3-adrenergic receptor function and in motor learning. Between the ages of 6 and 18 months, cerebellar /3adrenergic receptors in Fischer 344 rats decrease by 4 0 45% [31] (Curella and Zahniser, unpublished observations). The receptor density observed in the present study is consistent with this magnitude of loss. In cerebella of Sprague-Dawley rats, the majority (80-95%) of/3-adrenergic receptors have been identified as /32 receptors [32,37,49]. The single, relatively low affinity site observed here for ICI 89,406 indicates that /32 receptors were the only subtype of /3-adrenergic receptor detected in the cerebella of aged Fischer 344 rats [33]. We hypothesized that chronic dietary restriction might retard the aging-related decline in /3-adrenergic receptor density. However,

Thomas J. Gould et al. / Brain Research 684 (1995) 150-158

the results of the binding studies indicated that these receptors were not regulated in response to dietary restriction. Perhaps this lack of regulation of the fl2-adrenergic receptors is not surprising since it has been reported that in cerebella of young Sprague-Dawley rats, only the /31adrenergic receptors are up-regulated in response to denervation or chronic antagonist treatment [49]. In addition, it is the /31-adrenergic receptor which is associated with the electrophysiological effect reported here [34,50]. In the present study, we had thought that if the density of /31 receptors were greatly increased by dietary restriction, these receptors would be detected. No /31 receptors, however, were detected. Another difference between the electrophysiological studies and the binding studies is that the functional studies were done using agonists and the binding studies were performed with antagonists; a possible alteration in high affinity vs. low affinity binding would not be detected with antagonist binding. The lack of detectable binding to /31 receptors precluded addressing this issue. However, interestingly, food restriction has been reported to partially attenuate the 75% reduction in cerebellar /3-adrenergic receptors observed in 24-month-old female Wistar rats [36]. Even though motor learning and fl-adrenergic receptor function were improved wiith diet restriction, the mechanisms through which diet restriction deter neural and behavioral deficits are not well understood. One possible mechanism is through reduction or prevention of age-related free radical induced damage. While several theories of aging exist, to date one theory that continues to receive support is the free radical theory of aging (see Yu [52] for review). The free radical theory of aging postulates that it is the accumulation of free' radical reaction damage that contributes to or constitutes aging [19]. In turn, free radicals are the most reactive substance in biological systems and the instability of free radicals produces deterioration of function through alteration of biological molecules [51]. The body, however, has inborn defenses to combat free radicals. Antioxidants and enzymes such as superoxide dismutase, catalase, peroxidase, ascorbate and vitamin E, to name a few, reduce or ,;cavenge free radicals. During aging, however, the balance between effective antioxidants and free radicals shifts to an excess of free radical [38]. Thus, the inability of antioxidants to reduce free radicals may further contribute to the damaging effects of aging. Diet manipulations may retard the damaging effects of aging by reducing dietary levels of free radical reaction catalyst, by increasing levels of antioxidants and by reducing oxygen consumption and thus reducing mitochondrial free radical production [13,18,19]. In support, caloric restriction has been shown to reduce lipid peroxidation in mice [9]. Whereas, the rate of generation of mitochondrial superoxide and hydrogen peroxide increases with age, the rate of generation was higher in ad lib fed animals than in diet restricted animals [41]. Thus, caloric restriction decreases free radical production. Also, aged rats given

157

N-tert-butyl-a-phenylnitrone (PBN), a spin-trapping agent that may scavenge free radicals, had improved cerebellar noradrenergic receptor function compared to aged-matched control rats [17]. This result is similar to the results from the present study. Thus, dietary restriction and treatment with PBN produce similar effects on /3-receptor function. Although, different mechanisms could be involved for the two treatments, these data are consistent with the theory of a reduction of free radicals in the ameliorative effects of dietary restriction. Additional studies have also reported improved neural function and learning in diet-restricted animals. With aging, choline acetyltransferase activity has been shown to fall but diet restricted rats showed less decrease than controls [27]. Similarly, diet restriction decreased age-related dopamine receptor loss [39], although we did not observe any change in fl-adrenergic receptor number or affinity in this study. Finally, similar to our results, diet restriction in mice decrease age-related deficits in complex maze learning [22]. Aged diet restricted mice had a level of errors similar to young mice but aged control mice had significantly more errors. Thus, diet restriction can improve neural function which in turn may improve learning. In the present study, diet restricted rats showed superior cerebellar fl-adrenergic receptor function and superior motor learning compared to aged-matched ad lib fed rats.

Acknowledgements This work was supported by USPHS Grants AG04418 (P.B., N.Z.) and AG05686 (T.G.) and the VAMRS (P.B.).

References [1] Bickford, P.C., Motor learning deficits in aged rats are correlated with loss of cerebellar noradrenergic function, Brain Res., 620 (1993) 133-138. [2] Bickford, P., Heron, C., Young, D.A., Gerhardt, G.A. and de la Garza, R., Impaired acquisition of novel locomotor tasks in aged and norepinephrine-depleted F344 rats, Neurobiol. Aging, 13 (1992) 475-481. [3] Bickford, P.C., Hoffer, B.J. and Freedman, R., Interaction of norepinephrine with Purkinje cell responses to cerebellar afferent inputs in aged rats, Neurobiol. Aging, 6 (1985) 89-94. [4] Bickford, P.C., Aging and motor learning: a possible role for norepinephrine in cerebellar plasticity, Rev. Neurosci., in press. [5] Birt, D.F., Higginbotham, S.M., Patti, K. and Pour, P., Nutritional effects on the life-span of Syrian hamsters, Age, 5 (1982) 11-19. [6] Birt, D.F., Baker, P.Y. and Hruza, D.S., Nutritional evaluations of three dietary levels of lactalbumin throughout the life-span of two generations of Syrian hamsters, J. Nutr., 112 (1982) 2151-2160. [7] Bradford, M.M., A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding, Anal Biochem., 72 (1976) 248-254. [8] Cheney, K.E., Liu, R.K., Smith, G.S., Meredith, P.J., Mickey, M.R. and Waiford, R.L., The effect of dietary restriction of varying duration on survival, tumor patterns, immune function, and body temperature in B10C3F1 female mice, J. Gerontol., 38 (1983) 420-430.

158

Thomas J. Gould et al. / Brain Research 684 (1995) 150-158

[9] Davis, L.J., Tadolini, B., Biagi, P.L., Walford, R. and Licastro, F., Effect of age and extent of dietary restriction on hepatic microsomal lipid peroxidation potential in mice, Mech. Ageing Dec., 72 (1993) 155-163. [10] Davis, T.A., Bales, C.W. and Beanchene, R.E., Differential effects of dietary caloric and protein restriction in the aging rat, Exp. Gerontol., 18 (1983) 427-435. [11] Eccles, J.C., Ito, M. and Szentogothai, J., The Cerebellum as a Neuronal Machine, Springer, New York, 1967. [12] Effros, R.B., Walford, R.L, Weindruch, R. and Mitcheltree, C., Influences of dietary restriction on immunity to influenza in aged mice, J. Gerontol., 46 (1991) B142-B147, [13] Feuers, R.J., Weindruch, R. and Hart, R.W., Caloric restriction, aging, and antioxidant enzymes, Mutation Res., 295 (1993) 191-200. [14] Freedman, R., Hoffer, B.J., Puro, D. and Woodward, D.J., Noradrenaline modulation of responses of the cerebeUar Purkinje cell to afferent synaptic activity, Br. J. Pharmacol., 57 (1976) 603-605. [15] Freedman, R., Hoffer, B.J. and Woodward, D.J., A quantitative microiontophoretic analysis of the responses of central neurons to norepinephrine: interaction with cobalt, manganese, verapamil, and dichloroisoproterenol, Br. J. Pharmacol., 54 (1975) 529-539. [16] Freedman, R., Hoffer, B.J., Woodward, D.J. and Puro, D., Interaction of norepinephrine with cerebellar activity evoked by mossy and climbing fibers, Exp. Neurol., 55 (1977) 269-288. [17] Gould, T.J., and Bickford, P.C., The effects of chronic treatment with N-tert-butyl-a-phenyinitrone on cerebellar noradrenergic receptor function in aged F344 rats, Brain Res., 660 (1994) 333-336. [18] Harman, D., Free radical theory of aging: nutritional implications, Age, 1 (1978) 145-152. [19] Harman, D., The aging process, Proc. Natl. Acad. Sci., 78 (1981) 7124-7128. [20] Herlihy, J.T., Stacy, C. and Bertrand, H.A., Long-term calorie restriction enhances baroreflex responsiveness in Fischer 344 rats, Am. J. Physiol., 263 (1992) H1021-H1025. [21] Hirst, K., Zerbe, G.O., Boyle, D.S. and Wilkening, R.B., On nonrandom linear effects models for repeated measurements, Commun. Stat. B Simul. Comput., 20 (1991) 463-478. [22] Ingrain, D.K., Weindmch, R., Spangler, E.L., Freeman, J.R. and Walford, R.L., Dietary restriction benefits learning and motor performance of aged mice, J. Gerontol., 42 (1987) 78-81. [23] King, J.T. and Visscher, M.B., Longevity as a function of diet in the C3H mouse, Fed. Proc., Fed. Am. Soc. Exp. Biol., 9 (1950) 70, [24] Lee, E.T., Statistical Methods for Survival Data Analysis, Lifetime Learning Publications, Belmont, CA, 1980. [25] Leto, S., Kokonen, G.C. and Barrows, C.H., Dietary protein, lifespan, and biochemical variables in female mice, J. Gerontol., 31 (1976) 144-148. [26] Lin, A.M.Y, Bickford, P.C. and Palmer, M.R., The effects of ethanol on "y-amino acid-induced depressions of cerebellar Purkinje neurons: Influence of beta adrenergic receptor action in young and aged Fischer 344 rats, J. Pharm. and Exp. Ther., 264 (1993) 951-957. [27] London, E.D., Waller, S.B., Ellis, A.T. and Ingrain, D.K., Effects of intermittent feeding on neurochemical markers in aging rat brain, Neurobiol. Aging, 6 (1988) 199-204. [28] Masoro, E.J., Iwasaki, K., Gleiser, C.A., McMahan, C.A., Seo, E.J. and Yu, B.P., Dietary modulation of the progression of nephropathy in aging rats: an evaluation of the importance of protein, Am. J. Clin. Nutr., 49 (1989) 1217-1227. [29] McCay, C.M., Crowell, M.F. and Maynard, LA., The effect of retarded growth upon the length of life span and upon the ultimate body size, J. Nutr., 10 (1935) 63-79. [30] McElligott, J.G., Ebner, T.J. and BIoedel, J.R., Reduction of cerebellar norepinephrine alters climbing fiber enhancement of mossy fiber input to the Purkinje cell, Brain Res., 397 (1986) 245-252. [31] Miller, J.A. and Zahniser, N.R., Quantitative autoradiographic analy-

sis of 125I-pindolol binding in Fischer 344 rat brain: changes in fl-adrenergic receptor density with aging, Neurobiol. Aging, 9 (1988) 267-272. [32] Minneman, K.P. Hegstrand, L.R. and Molinoff, P.B., Simultaneous determination of beta-1 and beta-2-adrenergic receptors in tissues containing both receptor subtypes, Mol. Pharmacol., 16 (1979) 34-46. [33] Neve, K.A., McGonigle, P. and Molinoff, P.B., Quantitative analysis of the selectivity of radioligands for subtypes of beta adrenergic receptors, J. Pharmacol. Exp. Ther., 238 (1986) 46-53. [34] Parfitt, K.D. and Bickford-Wimer, P.C., Age-related subsensitivity of cerebellar Purkinje neurons to locally applied Betal-selective adrenergic agonist, Neurobiol. Aging, 11 (1990) 591-596. [35] Parfitt, K.D., Age-related electrophysiological changes in cerebellar noradrenergic receptors, Age, 11 (1988) 120-127. [36] Pieri, C., Food restriction slows down age-related changes in cell membrane parameters, Ann. NYAcad. Sci., 621 (1991) 353-362. [37] Rainbow, T.C., Parsons, B. and Wolfe, B.B., Quantitative autoradiography of ill-and fl2-adrenergic receptors in rat brain, Proc. Natl. Acad. Sci., 81 (1984) 1585-1589. [38] Ross, M.H. and Bras, G., Tumor incidence patterns and nutrition in the rat, J. Nutr., 87 (1965) 245-260. [39] Roth, G.S., Ingram, D.K. and Joseph, J.A., Delayed loss of striatal dopamine receptors during aging of dietarily restricted rats, Brain Res., 300 (1984) 27-32. [40] Salmoihragi, G.C. and Weight, F.F., Micromethods in neuropharmacology: an approach to the study of anesthetics, Anesthesiology, 28 (1967) 54-64. [41] Sohal, R.S., Ku, H.H., Agarwal, S., Forster, M.J. and Lal, H., Oxidative damage, rnitochondrial oxidant generation and antioxidant defenses during aging and in response to food restriction in the mouse, Mech. Ageing Dev., 74 (1994) 121-133. [42] Thomas, J., Bertrand, H., Stacy, C. and Herlihy, J.T., Long-term caloric restriction improves baroreflex sensitivity in aging Fischer 344 rats, J. Gerontol., 48 (1993) B151-B155. [43] van Neerven, J., Pompeiano, O., Collewijn, H. and van der Steen, J., Injections of /3-noradrenergic substances in the flocculus of rabbits affect adaptation of the VOR gain, Exp. Brain Res., 79 (1990) 249-260. [44] Wallace, J.E., Kranter, E.E. and Campbell, B.A., Motor and reflexive behavior in the aging rat, J. Gerontol., 35 (1980) 364-370. [45] Watson, M. and McElligott, J.G., 6-OHDA induced effects upon the acquisition and performance of specific locomotor tasks in rats, Pharmacol. Biochem. Behav., 18 (1983) 927-934. [46] Watson, M. and McEUigott, J.G., Cerebellar norepinephrine depletion and impaired acquisition of specific locomotor tasks in rats, Brain Res., 296 (1984) 129-138. [47] Weindmch, R. and Walford, R.L., The Retardation of Aging and Disease by Dietary Restriction, Thomas, Springfield, IL, 1988, [48] Weraarchakul, N., Strong, R., Wood, W.G. and Richardson, A., The effect of aging and dietary restriction on DNA repair, Exp. Cell Res., 181 (1989) 197-204. [49] Wolfe, B.B., Minneman, K.P. and Molinoff, P.B., Selective increases in the density of cerebellar /31-adrenergic receptors, Brain Res., 234 (1982) 474-479. [50] Yeh, H.H. and Woodward, D.J., Beta-1 adrenergic receptors mediate noradrenergic facilitation of Purkinje cell responses to gamma-amminobutyrie acid in cerebellum of rat, Neuropharmacology, 22 (1983) 629-639. [51] Yu, B.P., How diet influences the aging process of the rat, P.S.E.B.M., 205 (1994) 97-105. [52] Yu, B.P., Modulation of Aging Processes by Dietary Restriction, CRC Press, Boca Raton, FL, 1994. [53] Yu, B.P., Masoro, E.J. and McMahan, C.A., Nutritional influences on aging of Fischer 344 rats: I. Physical, metabolic, and longevity characteristics, J. Gerontol., 40 (1985) 657-670.