Ageing alters behavioural function and brain arginine metabolism in male Sprague–Dawley rats

Neuroscience 226 (2012) 178–196

AGEING ALTERS BEHAVIOURAL FUNCTION AND BRAIN ARGININE METABOLISM IN MALE SPRAGUE–DAWLEY RATS N. GUPTA, a  Y. JING, a  N. D. COLLIE, a H. ZHANG b AND P. LIU a*

INTRODUCTION Cognitive decline is common in aged individuals. It has been documented that age-related decline in learning and memory ability is associated with dysfunctions of the medial temporal lobe structures and prefrontal cortex (for reviews see Gallagher and Rapp, 1997; Tisserand and Jolles, 2003; Burke and Barnes, 2006; Greenwood, 2007). The medial temporal lobe memory system consists of the hippocampus and its adjacent entorhinal, perirhinal and parahippocampal (postrhinal in rodents) cortices (the parahippocampal region). The hippocampus contains three major sub-regions, CA1, CA3 and dentate gyrus (DG), and receives both spatial and non-spatial information through the parahippocampal region (Witter et al., 2000; Eichenbaum and Lipton, 2008). A growing body of evidence suggests the functional dissociation among the three sub-regions of the hippocampus (Kesner et al., 2004; Kesner, 2009). Adult neurogenesis exists in the DG, and the normal rate of neurogenesis is important in maintaining hippocampal function (for reviews see Deng et al., 2010; Aimone et al., 2011). It has been well documented that hippocampal neurogenesis is dramatically impaired during ageing (Klempin and Kempermann, 2007; Drapeau and Nora Abrous, 2008; Fabel and Kempermann, 2008). The prefrontal cortex is involved in a variety of memory functions, and is vulnerable to age-related deterioration (Shimamura, 1995; Greenwood, 2007). L-Arginine is one of the most metabolically versatile amino acids. It can be metabolized, for example, by nitric oxide synthase (NOS) to form nitric oxide (NO) and L-citrulline, by arginase to generate L-ornithine and urea, and by arginine decarboxylase to produce agmatine and carbon dioxide (Zhang and Snyder, 1995; Wu and Morris, 1998; Wiesinger, 2001). NO is a water- and lipid-soluble gas, and plays an important role in synaptic plasticity and learning and memory at physiological concentration (Holscher, 1997; Huang, 1997; Feil and Kleppisch, 2008). However, it can be neurotoxic when present in excessive amount due to its properties as a free radical (Calabrese et al., 2007). It has, therefore, been proposed that NO is critically involved in the ageing process, the so-called NO hypothesis of ageing (McCann, 1997; McCann et al., 1998, 2005), as well as the neurodegenerative process in Alzheimer’s disease (Law et al., 2001; Malinski, 2007). Previous research has demonstrated altered NOS activity and expression during ageing in memory-related brain structures and the correlations with age-related cognitive

a

Department of Anatomy, Brain Health Research Centre, University of Otago, P.O. Box 913, Dunedin 9054, New Zealand

b

School of Pharmacy, Brain Health Research Centre, University of Otago, P.O. Box 913, Dunedin 9054, New Zealand

Abstract—A growing body of evidence suggests the involvement of L-arginine and its metabolites in the ageing and neurodegenerative processes. The present study assessed behavioural performance in 4- (young), 12- (middle-aged) and 24- (aged) month-old male Sprague–Dawley rats, and investigated age-related changes in the activity of two key arginine metabolic enzymes, nitric oxide synthase (NOS) and arginase, and the levels of L-arginine and its downstream metabolites in a number of memoryrelated brain structures. Aged rats were less anxious and performed poorly in the water maze task relative to the young and middle-aged rats, and both middle-aged and aged rats displayed reduced exploratory activity relative to the young ones. There were significant age-related changes in NOS and arginase activities, and the levels of L-arginine, L-citrulline, L-ornithine, agmatine, putrescine, spermidine, spermine and glutamate, but not c-aminobutyric acid, in the CA1, CA2/3 and dentate gyrus sub-regions of the hippocampus and the prefrontal, entorhinal, perirhinal, postrhinal and temporal (an auditory cortex) cortices in a regionspecific manner. Cluster analyses revealed that the nine related neurochemical variables formed distinct groups, which changed as a function of ageing. Multiple regression analyses revealed a number of significant correlations between the neurochemical and behavioural variables. The present study further supports the involvement of arginine metabolism in the ageing process, and provides further evidence of the effects of animals’ behavioural experience on arginine metabolism. Ó 2012 IBRO. Published by Elsevier Ltd. All rights reserved.

Key words: ageing, L-arginine metabolites, hippocampus, prefrontal cortex, parahippocampal region, memory.

*Corresponding author. Tel: +64-03-4797536; fax: +64-03-4797254. E-mail address: [email protected] (P. Liu).   Joint first authorship due to equal contribution to the study. Abbreviations: ANOVA, analysis of variance; DG, dentate gyrus; EC, entorhinal cortex; GABA, aminobutyric acid; HPLC, high performance liquid chromatography; LC/MS, liquid chromatography/mass spectrometry; NMDA, N-methyl-D-aspartate; NO, nitric oxide; NOS, nitric oxide synthase; ODC, ornithine decarboxylase; PFC, prefrontal cortex; PRC, perirhinal cortex; POR, postrhinal cortex; TE, temporal cortex.

0306-4522/12 $36.00 Ó 2012 IBRO. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.neuroscience.2012.09.013 178

N. Gupta et al. / Neuroscience 226 (2012) 178–196

impairments (Sugaya et al., 1996; Law et al., 2000, 2002; Necchi et al., 2002; Liu et al., 2003a,b, 2004a,b, 2005, 2009b). L-ornithine is the main precursor of polyamines putrescine, spermidine and spermine. Putrescine is mainly formed from L-ornithine by ornithine decarboxylase (ODC). It combines with decarboxylated S-adenosylmethionine to produce spermidine and spermine via spermidine synthase and spermine synthase, respectively. It has been well documented that the physiological concentrations of polyamines are essential in maintaining normal cellular function (Williams, 1997; Wallace, 2000; Oredsson, 2003). There is evidence suggesting an important role of polyamines, putrescine in particular, in hippocampal neurogenesis (Malaterre et al., 2004). We have demonstrated that ageing affects polyamine concentrations in the hippocampus, parahippocampal region and prefrontal cortex significantly in a region-specific manner (Liu et al., 2008c). The presence of agmatine in mammalian brains was discovered in 1994 (Li et al., 1994). Agmatine is a putative neurotransmitter and interacts with a number of receptor subtypes, including N-methyl-D-aspartate (NMDA) receptors. It regulates the production of NO and polyamines by influencing the activities of NOS and ODC (for reviews see Reis and Regunathan, 2000; Satriano, 2003; Halaris and Plietz, 2007). Since agmatine can be metabolized by agmatinase to form putrescine, it is considered an alternative precursor of polyamines and a member of the polyamine family (Moinard et al., 2005). Recent research suggests that endogenous agmatine may directly participate in the processes of learning and memory as a neurotransmitter (Liu et al., 2008b, 2009a; Leitch et al., 2011; Seo et al., 2011), and that ageing affects agmatine levels in memory-related structures dramatically in a regionspecific manner (Liu et al., 2008a). As described above, L-arginine can be metabolized to generate a number of bioactive molecules, including glutamate and c-aminobutyric acid (GABA) due to the alternative source of generation from L-ornithine (Wu and Morris, 1998). Hence, it is important to investigate how the ageing process affects arginine metabolic profile in a single study. Since cognitive decline starts in middle age and continues throughout the ageing process (Finch, 2009; Salthouse, 2009), the present study assessed behavioural performance and arginine metabolic profile changes in memory-related structures in young, middleaged and aged rats. Specifically, animals were tested in the elevated plus maze, open field and water maze task, and age-related changes in NOS and arginase activity and the levels of L-arginine and its metabolites in the sub-regions of the hippocampus and the prefrontal, entorhinal, perirhinal, postrhinal and temporal (an auditory area as a control) cortices were investigated.

179

light–dark cycle (lights on at 7 a.m.) and provided ad lib access to food and water. The health condition (e.g., body weight, eyes, teeth, fur, skin, feet, urine and general behaviour) of aged and middle-aged animals was regularly monitored by animal technicians and a consultant veterinarian. Only animals showing good health were used for the study. All experimental procedures were carried out in accordance with the regulations of the University of Otago Committee on Ethics in the Care and Use of Laboratory Animals. Every attempt was made to limit the number of animals used and to minimize their suffering.

Behavioural procedures All behavioural experiments were conducted in a windowless room with four 60 W bulbs mounted on the ceiling. A video camera was mounted at ceiling height in the centre of the room and used for recording the performance during the experimental period. A radio speaker was located adjacent to the video camera at ceiling height to provide background masking noise. The extra maze cues (the laboratory furniture, lights and several prominent visual features on the walls, as well as the location of the experimenter) were held constant throughout the entire study. Elevated plus maze. The elevated plus maze was shaped like a plus sign in black-painted wood, with two unwalled (open) arms (50  13.5 cm2) surrounded by clear Pleixglass of 4 cm and two walled (closed) arms (50  13.5  29 cm3). The central area of the maze measured 13.5  13.5 cm2. The arm locations were kept constant with north and south being the closed arms. The maze was elevated approximately 60 cm above the floor. Animals were placed at the centre of the elevated plus maze facing one closed arm, and left in the maze for a period of 5 min. Animal behaviour was videotaped and analysed offline by a computerized tracking system (HVS 2020). The total time spent on each arm during the testing period was analysed. An entry was defined by placing all four paws into an arm, and no time was recorded when the animal was in the centre of the maze (Liu and Bergin, 2009; Liu and Collie, 2009; Bergin and Liu, 2010). Open field. The open field was an experimental chamber consisting of a 60  60 cm black hardboard box with walls 20 cm high and 36 equal-sized squares on the floor. The box was set up immediately after completion of the elevated plus maze test and elevated approximately 60 cm above the floor. Animals were placed into the chamber, and allowed to explore the apparatus freely for 5 min. Animal behaviour was videotaped and analysed offline by HVS 2020. The duration of wall-supported rearings, the path length generated during the entire 5-min testing period and the percentage of time spent in the outer zone (10 cm from the wall) and the centre (central four 10-cm squares) were analysed (Liu et al., 2008d; Liu and Bergin, 2009; Bergin and Liu, 2010).

EXPERIMENTAL PROCEDURES

Water maze task. The water maze pool was a black circular tank measuring 150 cm in diameter and 45 cm in height. It was filled with water to a depth of approximately 25 cm and maintained at a temperature of 25 ± 1 °C. Four points around the edge of the pool were designated as north (N), south (S), east (E) and west (W), which allowed the apparatus to be divided into four corresponding quadrants (i.e. NE, SW, NW and SE).

Male Sprague–Dawley rats, 4 (young, n = 10), 12 (middle-aged, n = 10) and 24 (aged, n = 10) months old, were housed three to five per cage (53  33  26 cm3), maintained on a 12-h

Place navigation (days 1–6). The day after the elevated plus maze and open field tests, all rats were trained in the water maze task. During the place navigation trials, a black platform (10 cm in diameter) was located in the centre of the SE quadrant and submerged 2 cm below the water surface. All rats received six

Subjects

180

N. Gupta et al. / Neuroscience 226 (2012) 178–196

trials per day with 120-s intervals between trials for six consecutive days. On each trial, the rat was placed into the pool facing towards the wall and was allowed to swim in search of the hidden platform. The rat was permitted to remain on the located platform for 10 s before being removed and placed in a holding box. If the rat did not find the platform within 90 s of being placed into the pool, it was immediately placed on the platform for 10 s before being returned to the holding box. During the 120-s intertrial interval, the rat was dried off, warmed and kept in the holding box. Starting locations (N, S, W, and E) were pseudorandomly selected. Probe tests (days 6 and 7). A probe trial was conducted 120 s after completion of the final training trial on day 6 (Probe 1). The platform was removed from the SE quadrant, and rats were placed into the pool from a fixed starting point and were allowed to swim freely for a duration of 60 s. The second probe trial was conducted 120 s before the start of the cued navigation trials on day 7 (Probe 2), which was approximately 24 h after completion of the final training trial on day 6. Cued Navigation (days 7 and 8). During the cued navigation trials, the platform was shifted to the centre of the NW quadrant and raised 2 cm above the water surface. The edge of the platform was marked by yellow tape to make it more visible. All rats received six trials per day for two consecutive days with 120-s intertrial intervals. Again, the maximum searching time allowed was 90 s and the rat was remained on the located platform for 10 s. Following completion of the water maze experiments, several performance variables were analysed from HVS 2020 (Gupta et al., 2009; Liu and Collie, 2009; Bergin and Liu, 2010; Gupta et al., 2012). Rat swimming speed was obtained across the 6 days of training during the place navigation. The distance the rats swam from the starting point to reach the platform (path length), the degree of thigmotaxic swimming (i.e., swimming close to the wall) and the accuracy of the initial heading angle were measured during the place and cued navigation. Thigmotaxic swimming was quantified by dividing the maze into two circles and measuring the time spent in the outer circle of the pool (15 cm wide). The absolute initial heading error was determined by the angle formed by a straight line from the start point to the platform and a line from the start point to the location of the animal after it had travelled 20 cm. During the probe trials, the percentage of path length in the target quadrant and the number of crossings over the previous platform location were determined.

Neurochemical procedures

Tissue preparation. All rats were decapitated without anaesthesia. The brains were rapidly removed and left in cold saline (4 °C) for at least 45 s. The sub-regions of the hippocampus (CA1, CA2/3 and DG), and the prefrontal (PFC), entorhinal (EC), perirhinal (PRC), postrhinal (POR) and temporal (TE) cortices were dissected freshly on ice (Gupta et al., 2009; Liu et al., 2008a,b,c, 2009a,b). The tissues harvested from one hemisphere were frozen immediately and stored at 80 °C until the NOS and arginase assays. The tissues harvested from other hemisphere were weighed, homogenized in ice-cold 10% perchloric acid (50 mg wet weight per millilitre) and centrifuged at 10,000 rpm for 10 min to precipitate protein. The supernatants (the perchloric acid extracts) were frozen immediately and stored at 80 °C until the high performance liquid chromatography (HPLC) and liquid chromatography/mass spectrometric (LC/MS) assays.

NOS and arginase assays. At the time of the assay, protease-inhibitory buffer containing 50 mM Tris–HCl (pH 7.4), 10 lM phenylmethylsulfonyl fluoride, 15 lM pepstatin A and 2 lM leupeptin (1:10 w/v) was added to the samples on ice. Brain tissues were then homogenized using ultrasonification (Branson Sonifier 150D, Branson Ultrasonics Corporation, Connecticut, USA) and centrifuged at 12,000g for 10 min at 4 °C. Protein concentrations in the supernatant were measured based on the Bradford method (Bradford, 1976) using a Bio-Rad protein assay dye reagent concentrate and Bio-Rad Benchmark Plus microplate spectrophotometer (Bio-Rad Laboratories Inc., California, USA). Each supernatant was then separated into two parts and used for the NOS and arginase assays, respectively. We employed a radioenzymatic assay technique to analyse NOS activity by measuring the ability of tissue homogenates to convert [3H] L-arginine to [3H] L-citrulline in the presence of co-factors, and a spectrophotometric assay method to determine arginase activity by measuring the amount of newly formed urea from L-arginine, as described previously (Liu et al., 2003a,b, 2004a,b, 2005, 2009b, 2011; Knox et al., 2011). All assays were performed in duplicate. For each brain region, the tissues from the three groups were processed at the same time and the order was counterbalanced. NOS and arginase activities were expressed as pmol [3H] L-citrulline/30 min/mg protein and lg urea/mg protein, respectively.

HPLC procedures Determination of L-arginine, L-citrulline, L-ornithine, glutamate, GABA, spermidine and spermine were carried out by the HPLC methods as detailed in our recent publications (Liu et al., 2008c, 2009a,b, 2010, 2011; Gupta et al., 2009, 2012; Knox et al., 2011). High purity external and internal standards were used (Sigma, Sydney, Australia). All other chemicals were of analytical grade. For each brain region, samples from all three age groups were assayed under one experimental condition at the same time, and the order was counterbalanced between groups. For L-arginine, L-citrulline, L-ornithine, glutamate and GABA, after adding internal standard (trazodone) to 30 ll of the perchloric acid extracts, the samples were alkalized with potassium hydrogen carbonate solution (pH 9.8) and derivatized with dansyl chloride in dark at 80 °C for 20 min. The reaction was stopped by adding 10 ll of acetic acid followed by centrifugation at 10,000 rpm for 10 min. Forty microlitres of the supernatant was injected onto the HPLC system, which consisted of a programmed solvent delivery system at a flow rate of 1.0 ml/min, an autosampler, a reversed-phase C18 column, and a UV detector set at a wavelength of 218 nm. Identifications of L-arginine, L-citrulline, L-ornithine, glutamate and GABA were accomplished by comparing the retention times of samples with the known standards. Assay validation showed that the analytical method was sensitive and reliable with acceptable accuracy (92–107% of true values) and precision (intra- and inter-assay CV < 15%). The concentrations of L-arginine, L-citrulline, L-ornithine, glutamate and GABA were calculated with reference to the peak area of external standards, and values were expressed as lg/g wet tissue. For spermidine and spermine, after adding internal standard (1, 7-diaminoheptane) to 20 ll of the perchloric acid extracts, the samples were alkalized with saturated sodium carbonate and derivatized with dansyl chloride. Spermidine, spermine and internal standard were extracted with toluene. The toluene phase was evaporated to dryness, reconstituted and injected onto the HPLC system, which consisted of a programmed solvent delivery system at a flow rate of 1.5 ml/min, an autosampler, a reversed-phase C18 column, and a fluorescence detector set at the excitation wavelength of 252 nm and emission wavelength of 515 nm. Identifications of spermidine

181

N. Gupta et al. / Neuroscience 226 (2012) 178–196 and spermine were accomplished by comparing the retention times of samples with the known standard. Assay validation showed that the analytical method was sensitive and reliable with acceptable accuracy (88–112% of true values) and precision (intraand inter-assay CV ± 15%). The concentrations of spermidine and spermine were calculated with reference to the peak area of external standards and values were expressed as lg/g wet tissue.

LC/MS procedure Agmatine and putrescine concentrations were measured by a highly sensitive LC/MS method (Liu et al., 2008a,b,c, 2009a, 2010, 2011; Gupta et al., 2009, 2012; Knox et al., 2011). High purity external and internal standards were used (Sigma, Sydney, Australia). All other chemicals were of analytical grade. For each brain region, samples from all three age groups were assayed under one experimental condition at the same time, and the order was counterbalanced between groups. After adding the internal standard (1,7-diaminoheptane) to 20 ll of the perchloric acid extracts, the samples were alkalized with saturated sodium carbonate and derivatized with dansyl chloride. Agmatine, putrescine and internal standard were extracted with toluene. The toluene phase was evaporated to dryness, reconstituted and injected onto the LC/MS/MS system. The samples were analysed by a reversed-phase C18 column (150  2.0 mm, 5 mm, Phenomenex) with 80% acetonitrile: 20% water containing 0.1% formic acid as mobile phase at a flow rate of 0.2 ml/min. The retention times of agmatine, putrescine and the internal standard were 1.7, 4.0 and 4.8 min, respectively. The total run time was 15 min. Detection by MS/ MS used an electrospray interface in positive ion mode. The standard curves for agmatine and putrescine were linear up to 1000 ng/ml (R2 > 0.99). The intra- and inter-day coefficients of variation were <15%. The concentrations of agmatine and putrescine in tissue were calculated with reference to the peak area of the external standards and values were expressed as lg/g wet tissue.

Statistical analysis The behavioural and neurochemical data were analysed using a one-way or two-way repeated measures analysis of variance (ANOVA) followed by Bonferroni post hoc tests (Zolman, 1993). The significance level was set at 0.05 for all comparisons. All calculations were performed with the Graphpad Prism program. Cluster analysis is an exploratory data analysis tool to sort different variables into groups in a way that the degree of association between two variables is maximal if they belong to the same group. As we measured the levels of nine interrelated neurochemical variables (L-arginine, L-citrulline, L-ornithine, agmatine, putrescine, spermidine, spermine, glutamate and GABA) in the hippocampus, parahippocampal region, prefrontal cortex and temporal cortex, cluster analyses were performed for each group and each brain region to determine which neurochemical variables varied closely with one another using Minitab 15 (Liu et al., 2010, 2011). Prior to the analysis, the data were standardized to obtain z values. Agglomerative methods were then used on the correlation coefficient distance, which started with each observation as a cluster and with each step combine observations to form clusters until there was only one large cluster. Comparisons of the different kinds of cluster analyses suggested that Complete linkage, McQuitty linkage, Average linkage and Ward linkage all produced similar results (data not shown). Therefore, Ward linkage (based on the sum of squares between the two clusters, summed over all variables) was used for all of the analyses. Dendograms were generated using Minitab 15 to display the groups formed by clustering of variables and their

similarity levels (defined as percent of the minimum distance at each step relative to the maximum inter-variable distance) (Liu et al., 2010, 2011).

RESULTS Body weight Animals were weighed before the start of the behavioural tests. One-way ANOVA revealed a significant difference between groups (F(2, 27) = 39.3, p < 0.0001). Post-hoc test indicated the differences between the young (501 ±10.8 g) and both middle-aged (734.6 ± 35.07 g) and aged (aged: 827 ± 28.44 g) groups (all p < 0.001), and the difference between the latter two (p < 0.05). Behavioural data Elevated plus maze. Fig. 1 illustrates the performance of the three age groups in the elevated plus maze. A one-way ANOVA revealed a significant difference between groups in the total time spent in the open arms (F(2, 27) = 6.19, p = 0.006), with the higher value in the aged group relative to the middle-aged (p < 0.05) and young (p < 0.01) ones and no difference between the latter two (Fig. 1A). There was also a significant difference between groups in the total time spent in the enclosed arm (F(2, 27) = 3.74, p = 0.037), with the lower value in the aged group relative to the young (p < 0.05), but not the middle-aged one and no difference between the latter two (Fig. 1B). When the ratio of the time spent in the open and enclosed arms was analysed, a significant difference between groups was observed (F(2, 27) = 5.66, p = 0.009), with the higher value in the aged group as compared to the middle-aged (p < 0.05) and young (p < 0.01) ones and no difference between the latter two (Fig. 1C). Open field. Fig. 2 presents animals’ performance in the open field. A one-way ANOVA revealed a significant difference between groups in terms of the duration of wall-supported rearings (F(2, 27) = 5.52, p = 0.01), with the lower values in the aged (p < 0.01) and middleaged (p < 0.05) groups as compared to the young one and no difference between the former two (Fig. 2A). There was no significant difference between groups for the path length measurement (F(2, 27) = 1.41, p = 0.26; Fig. 2B). When the percentage of time spent in the outer zone was analysed, one-way ANOVA revealed a significant difference between groups (F(2, 27) = 3.93, p = 0.03) with less time in the aged group as compared to the middle-aged and young ones (all p < 0.05) and no difference between the latter two (Fig. 2C). All three age groups spent less than 1% of time in the inner zone (data not shown). Water maze task. Place navigation. The swimming speed was averaged across 6 days of training. One-way ANOVA revealed a significant difference between groups (F(2, 27) = 6.42, p = 0.005) with slower swimming speed in the aged group relative to the middle-aged (p < 0.05) and young (p < 0.01) ones and

182

N. Gupta et al. / Neuroscience 226 (2012) 178–196

Fig. 1. Animals’ performance in the elevated plus maze. Mean (±SEM) time(s) spent in the open (A) and enclosed (B) arms and ratio (C) in the aged, middle-aged and young groups (n = 10 in each group). Aged rats spent significantly more time in the open arms relative to the middle-aged and young rats and less time in the enclosed arms relative to the young rats. Aged rats also had significantly higher ratio of time spent in the open and enclosed arms when compared to the middle-aged and young rats. ⁄ indicates a significant difference between groups: ⁄p < 0.05; ⁄⁄p < 0.01.

no difference between the latter two (Fig. 3A). For both the path length (Fig. 3B) and thigmotaxic swimming (Fig. 3C) measurements, two-way repeated measures ANOVA revealed significant effects of group (path length: F(2, 27) = 17.87, p = 0.0001; thigmotaxic swimming: F(2, 27) = 9.56, p = 0.0007) and day (path length: F(5, 135) = 42.44, p < 0.0001; thigmotaxic swimming: F(5, 135) = 9.84, p < 0.0001), but not group  day interaction (path length: F(10, 135) = 1.07, p = 0.38; thigmotaxic swimming: F(10, 135) = 1.58, p = 0.12), with greater path length in the aged group relative to the middle-aged and young ones on days 2–6 and no marked differences between the latter two in both measurements. In terms of the heading angle, there were significant effects of group (F(2, 27) = 3.67,

Fig. 2. Animals’’ performance in the open field. Mean (±SEM) duration (s) of wall-supported rearings (A), path length generated during the 5-min testing period (B) and percentage of time spent in the outer zone of the apparatus in the aged, middle-aged and young groups (n = 10 in each group). Both aged and middle-aged groups reared significantly shorter relative to the young group, with no difference between groups in path length. The aged group spent significantly less time in the outer zone of the apparatus relative to other two groups. ⁄ indicates a significant difference between groups: ⁄ p < 0.05; ⁄⁄p < 0.01.

p = 0.039) and day (F(5, 135) = 2.74, p = 0.022), but not group  day interaction (F(10, 135) = 1.89, p = 0.052) with greater heading angles in the aged and middle-aged groups relative to the young one during the last day of training (Fig. 3D). Probe tests. Probe 1 was conducted 120 s after the final training trial of day 6. One-way ANOVA revealed a significant difference between groups in the number of platform crossings (F(2, 27) = 4.72, p = 0.017; Fig. 3F), but not the percentage of path length in the target quadrant (F(2, 27) = 1.01, p = 0.38; Fig. 3E), with fewer number of platform crossings in the aged group relative to the middle-aged and young ones (all p < 0.05) and

N. Gupta et al. / Neuroscience 226 (2012) 178–196

183

Fig. 3. Animals’ performance in the water maze task. Mean (± SEM) swimming speed (cm/s, A), path length (cm) to reach the platform (B), percentage of path length spent in the outer zone (thigmotaxic swimming; C) and heading angle (degree, D) during the place (days 1–6) and cued (days 7 and 9) navigation, and percentage of path length in the target quadrant (E) and number of platform crossings (F) in Probes 1 and 2 (conducted 120 s and 24 h after the final training trial on day 6, respectively) in the aged, middle-aged and young groups (n = 10 in each group). The aged group swam slowly, generated longer path length to reach the platform and had greater percentage of path length in the outer zone of the swimming pool during the place and cued navigation relative to the middle-aged and young groups. The aged group made fewer platform crossings in Probe 1 relative to the middle-aged and young groups, and generated shorter path length in the target quadrant in Probe 2 as compared to the young group. ⁄ indicates a significant difference between groups: ⁄p < 0.05; ⁄⁄p < 0.01.

no difference between the latter two. During Probe 2 (conducted approximately 24 h after the final training trial of day 6), there was a significant difference between groups in the percentage of path length in the target quadrant (F(2, 27) = 3.39, p = 0.049; Fig. 3E), but not the number of platform crossings (F(2, 27) = 1.32, p = 0.28; Fig. 3F), with the shorter path length in the aged group as compared to the young one (p < 0.05). Cued ANOVA

navigation. Two-way revealed significant

repeated measures effects of group

(F(2, 27) = 11.67, p < 0.0002) and group  day interaction (F(2, 27) = 3.69, p = 0.038), but not day (F(1, 27) = 3.02, p = 0.094), with markedly longer path length in the aged group relative to the middle-aged and young ones on both days and no difference between the latter two (Fig. 3B). When thigmotaxic swimming was analysed, there was a significant effect of group (F(2, 27) = 5.82, p = 0.0079), but not day (F(1, 27) = 1.54, p = 0.23) or group  day interaction (F(2, 27) = 2.69, p = 0.086), with higher values in the aged group relative to the middle-aged and young ones on both days (Fig. 3C). In terms of the heading angle

184

N. Gupta et al. / Neuroscience 226 (2012) 178–196

Fig. 4. Mean (SEM) total NOS (A and B) and arginase (C and D) activities in the CA1, CA2/3 and dentate gyrus (DG) sub-regions of the hippocampus and the prefrontal (PFC), entorhinal (EC), perirhinal (PRC), postrhinal (POR) and temporal (TE) cortices in the aged, middle-aged and young groups (n = 10 in each group). There were increased total NOS activities with age in the CA2/3, PFC, EC, POR and TE, and decreased and increased arginase activities with age in the PFC and PRC respectively. ⁄ indicates a significant difference between groups at p < 0.05; ⁄⁄p < 0.01; ⁄⁄⁄ p < 0.001.

measurement, there was a significant effect of group  day interaction (F(2, 27) = 4.30, p = 0.024), but not day (F < 1) or group (F < 1) (Fig. 3D). Neurochemical data NOS and arginase activity. Fig. 4A, B illustrate total NOS activities in the hippocampus, prefrontal cortex, parahippocampal region and temporal cortex across the three age groups. A one-way ANOVA revealed significant differences between groups in CA2/3 (F(2, 27) = 5.35, p = 0.01), PFC (F(2, 27) = 14.82, p < 0.0001), EC (F(2, 27) = 5.64, p = 0.009), POR (F(2, 27) = 7.16, p = 0.003) and TE (F(2, 27) = 20.12, p < 0.0001), but not CA1 (F < 1), DG (F < 1) and PRC (F(2, 27) = 1.08, p = 0.35). Post-hoc test indicated significantly increased NOS activity in the aged group relative to the middle-aged (CA2/3 and EC: all p < 0.05; PFC and TE: all p < 0.001; POR: p < 0.01) and young (CA2/3, EC and POR: all p < 0.01; PFC and TE: all p < 0.001) ones, and in the middle-aged group when compared to the young one in TE (p < 0.05). Total arginase activities in the hippocampus, prefrontal cortex, parahippocampal region and temporal cortex across the three age groups are presented in Fig. 4C, D. There were significant differences between groups in PFC (F(2, 27) = 5.31, p = 0.01) and PRC (F(2, 27) = 7.03, p = 0.0035), but not CA1 (F < 1), CA2/3 (F < 1), DG (F(2, 27) = 3.10, p = 0.06), EC (F(2, 27) = 1.66, p = 0.21), POR (F(2, 27) = 1.41, p = 0.26) and TE (F(2, 27) = 2.87, p = 0.07), with decreased arginase activities in the aged (p < 0.01) and middle-aged (p < 0.05) groups relative to the young one

in PFC, but increased arginase activity in the aged group relative to the middle-aged (p < 0.05) and young (p < 0.01) ones in PRC. Amino acids and polyamines. Table 1 presents the levels of L-arginine, L-citrulline, L-ornithine, glutamate, GABA, agmatine, putrescine, spermidine and spermine in the sub-regions of the hippocampus, prefrontal cortex, parahippocampal region and temporal cortex across the three age groups. For L-arginine, a one-way ANOVA revealed significant differences between groups in CA1 (F(2, 27) = 8.57, p = 0.0013), CA2/3 (F(2, 27) = 11.97, p = 0.0002), PRC (F(2, 27) = 5.37, p = 0.01) and POR (F(2, 27) = 25.63, p < 0.0001), but not DG (F(2, 27) = 2.82, p = 0.08), PFC (F(2, 27) = 1.01, p = 0.38), EC (F < 1) and TE (F(2, 27) = 2.70, p = 0.085). In CA1, increased level of L-arginine was found in the middle-aged group relative to the aged (p < 0.05) and young (p < 0.01) ones. For CA2/3 and PRC, there were higher levels of L-arginine in the aged (CA2/3: p < 0.001; PRC: p < 0.05) and middle-aged (CA2/3: p < 0.01; PRC: p < 0.05) groups as compared to the young one. There was decreased L-arginine level in the aged group as compared to the middle-aged and young ones (all p < 0.001) in POR, with no difference between the latter two. For L-citrulline, there was a significant difference between groups in TE (F(2, 27) = 6.59, p = 0.005), but not CA1 (F < 1), CA2/3 (F < 1), DG (F < 1), PFC (F(2, 27) = 1.85, p = 0.18), EC (F(2, 27) = 3.16, p = 0.06), PRC (F < 1), or POR (F(2, 27) = 1.21, p = 0.32), with decreased L-citrulline level in the aged

Table 1. Mean (±SEM) L-arginine, L-citrulline, L-ornithine, glutamate, GABA, agmatine, putrescine, spermidine and spermine levels (ug/g wet tissue) in the CA1, CA2/3 and dentate gyrus (DG) of the hippocampus and the prefrontal (PFC), entorhinal (EC), perirhinal (PRC), postrhinal (POR) and temporal (TE) cortices in the young (Y), middle-aged (MA) and aged (A) groups (n = 10 in each group) Region Group

L-Arginine

L-Citrulline ⁄

Glutamate

L-Ornithine ⁄⁄$$

GABA ⁄⁄$

Agmatine

Putrescine

Spermidine ⁄

Spermine ⁄⁄⁄$$$

CA1

A MA Y

32.35 ± 1.62 35.33 ± 1.17## 27.94 ± 0.92

145.00 ± 4.97 154.40 ± 5.52 150.10 ± 4.69

40.88 ± 2.76 31.55 ± 2.43 28.60 ± 1.60

3978.00 ± 103.30 4458.00 ± 83.58 4421.00 ± 141.00

CA3

A MA Y

41.81 ± 1.89⁄⁄⁄ 37.65 ± 1.78## 29.94 ± 1.53

161.00 ± 4.48 167.50 ± 3.89 161.00 ± 2.75

50.56 ± 2.53⁄⁄⁄ 45.07 ± 1.78### 32.22 ± 1.41

3573.00 ± 126.70⁄⁄$$$ 4257.00 ± 60.06 4008.00 ± 97.25

327.40 ± 15.79 0.41 ± 0.03⁄⁄⁄$ 1.40 ± 0.06 354.70 ± 8.97 0.31 ± 0.03 1.40 ± 0.08 323.20 ± 8.07 0.23 ± 0.03 1.39 ± 0.18

DG

A MA Y

39.25 ± 1.21 44.67 ± 2.57 45.69 ± 2.15

149.00 ± 3.15 136.50 ± 9.35 144.10 ± 4.94

23.03 ± 2.44⁄ 22.24 ± 1.70# 16.50 ± 0.67

4682.00 ± 132.90 4699.00 ± 292.70 4593.00 ± 107.60

469.00 ± 14.27 0.41 ± 0.05 453.40 ± 22.90 0.38 ± 0.07 411.00 ± 10.94 0.46 ± 0.08

PFC

A MA Y

39.18 ± 1.94 43.51 ± 2.40 41.99 ± 2.21

166.50 ± 4.87 184.30 ± 10.72 182.70 ± 4.34

25.95 ± 1.57⁄⁄ 24.16 ± 0.76# 20.41 ± 0.75

5808.00 ± 114.50 6246.00 ± 250.60 6254.00 ± 125.60

277.5 ± 12.03 0.28 ± 0.05⁄⁄$$ 0.62 ± 0.07⁄⁄ 32.52 ± 0.79⁄⁄⁄$$$ 74.24 ± 2.82⁄⁄⁄$$$ 283.60 ± 8.93 0.71 ± 0.09 0.80 ± 0.05 27.06 ± 0.55### 60.52 ± 1.45### 277.70 ± 7.72 0.66 ± 0.11 0.91 ± 0.06 22.31 ± 0.89 51.63 ± 3.61

EC

A MA Y

23.19 ± 1.32 24.64 ± 0.59 24.47 ± 0.75

220.30 ± 6.71 246.80 ± 9.48 229.20 ± 6.15

29.23 ± 1.77⁄⁄$$ 35.25 ± 1.05### 23.64 ± 1.27

3551.00 ± 137.90⁄⁄⁄$$$ 299.50 ± 14.90 0.29 ± 0.04 5676.00 ± 236.90 289.20 ± 10.84 0.25 ± 0.03 5909.00 ± 144.10 288.90 ± 13.50 0.20 ± 0.01

PRC

A MA Y

32.20 ± 1.56⁄ 33.81 ± 1.98# 26.69 ± 1.19

197.90 ± 8.05 183.80 ± 10.00 180.60 ± 10.79

31.77 ± 1.87⁄⁄⁄ 27.31 ± 2.84## 16.90 ± 1.19

2648.00 ± 101.00 2878.00 ± 157.60 3085.00 ± 144.80

217.20 ± 13.08 0.57 ± 0.05⁄$ 212.60 ± 9.52 0.37 ± 0.02 215.40 ± 8.09 0.42 ± 0.07

POR

A MA Y

26.98 ± 1.57⁄⁄⁄$$$ 204.30 ± 22.11 40.65 ± 1.48 181.60 ± 6.66 39.15 ± 1.39 210.20 ± 5.86

25.05 ± 1.40 29.04 ± 1.36# 23.69 ± 1.46

4687.00 ± 239.10 5149.00 ± 176.90 5259.00 ± 191.20

TE

A MA Y

41.35 ± 2.50 35.50 ± 1.90 35.45 ± 1.71

0.23 ± 0.03 0.19 ± 0.04# 0.32 ± 0.02

1.06 ± 0.08 1.20 ± 0.09 1.45 ± 0.11

0.98 ± 0.10⁄ 1.18 ± 0.10 1.41 ± 0.13

58.50 ± 1.23 52.54 ± 1.59### 44.89 ± 0.86

41.35 ± 1.45 41.37 ± 0.82 40.81 ± 0.83

89.29 ± 1.44⁄⁄⁄$$$ 39.00 ± 1.19 70.27 ± 1.75### 41.93 ± 0.92 60.71 ± 2.04 42.19 ± 1.52 36.97 ± 1.06 37.08 ± 2.18 35.14 ± 1.91

66.75 ± 2.43 70.86 ± 4.74 71.28 ± 4.84

0.89 ± 0.10 0.91 ± 0.11 0.81 ± 0.08

34.72 ± 1.40 32.11 ± 1.88 29.12 ± 1.51

30.03 ± 1.68 31.50 ± 2.55 30.26 ± 2.69

0.84 ± 0.09 0.90 ± 0.03 0.87 ± 0.06

42.90 ± 1.34 41.99 ± 1.50 40.78 ± 1.15

46.93 ± 1.88⁄ 49.33 ± 1.52# 53.86 ± 1.18

241.00 ± 15.02 0.53 ± 0.05⁄⁄⁄$ 0.56 ± 0.05 263.70 ± 7.66 0.40 ± 0.04# 0.59 ± 0.06 250.10 ± 4.92 0.25 ± 0.05 0.63 ± 0.09

26.73 ± 0.84 27.13 ± 1.01 24.02 ± 1.46

48.64 ± 1.54 50.03 ± 2.63 44.23 ± 3.91

184.10 ± 10.71 0.53 ± 0.07⁄⁄$ 182.90 ± 7.45 0.34 ± 0.04 172.80 ± 7.90 0.28 ± 0.04

73.10 ± 1.91⁄⁄$$$ 60.27 ± 1.41# 65.73 ± 1.01

37.27 ± 1.81 36.34 ± 2.50 41.28 ± 1.61

0.66 ± 0.07⁄ 0.71 ± 0.07# 0.96 ± 0.09

N. Gupta et al. / Neuroscience 226 (2012) 178–196

130.10 ± 3.64⁄⁄ 67.15 ± 3.10⁄⁄⁄$$ 4428.00 ± 225.10 147.40 ± 4.29# 53.72 ± 3.60 4255.00 ± 99.67 154.90 ± 6.48 46.03 ± 1.67 4698.00 ± 210.70

281.80 ± 8.01 270.30 ± 7.06 267.10 ± 9.21

⁄

Indicates a significant difference between the aged and young groups: ⁄P < 0.05; ⁄⁄P < 0.01; ⁄⁄⁄P < 0.001. Indicates a significant difference between the middle-aged and young groups: #P < 0.05; ##P < 0.01; ###P < 0.001. $ Indicates a significant difference between the aged and middle-aged groups: $P < 0.05; $$P < 0.01; $$$P < 0.001. #

185

186

N. Gupta et al. / Neuroscience 226 (2012) 178–196

group relative to the middle-aged (p < 0.05) and young (p < 0.01) ones in TE. For L-ornithine, a one-way ANOVA revealed significant differences between groups in CA1 (F(2, 27) = 7.66, p = 0.0023), CA2/3 (F(2, 27) = 23.04, p < 0.0001), DG (F(2, 27) = 4.11, p = 0.028), PFC (F(2, 27) = 6.71, p = 0.0043), EC (F(2, 27) = 17.27, p < 0.0001), PRC (F(2, 27) = 13.45, p < 0.0001), POR (F(2, 27) = 3.91, p = 0.03) and TE (F(2, 27) = 13.53, p < 0.0001). In CA1 and TE, increased L-ornithine levels were found in the aged group relative to the middle-aged (all p < 0.01) and young (CA1: p < 0.01; TE: p < 0.001) ones with no differences between the latter two. In CA2/3, DG, PFC and PRC, there were significantly increased L-ornithine levels in the aged (CA2/3 and PRC: all p < 0.001; DG: p < 0.05; PFC: p < 0.01) and middle-aged (CA2/3: p < 0.001; DG and PFC: all p < 0.05; PRC: p < 0.01) groups as compared to the young one with no difference between the former two. In EC, increased L-ornithine levels were found in the aged (p < 0.01) and middleaged (p < 0.001) groups relative to the young one, and in the middle-aged group when compared to the aged one (p < 0.01). In POR, there was increased L-ornithine level in the middle-aged group relative to the young one (p < 0.05). For glutamate, there were significant differences between groups in CA1 (F(2, 27) = 5.69, p = 0.0086), CA2/3 (F(2, 27) = 12.35, p = 0.0002) and EC (F(2, 27) = 52.8, p < 0.0001), but not DG (F < 1), PFC (F(2, 27) = 2.13, p = 0.14), PRC (F(2, 27) = 2.56, p = 0.10), POR (F(2, 27) = 2.21, p = 0.13) and TE (F(2, 27) = 1.42, p = 0.26). Post-hoc test indicated decreased glutamate levels in the aged group relative to the middle aged (CA1: p < 0.05; CA2/3 and EC: all p < 0.001) and young (CA1 and CA2/3: all p < 0.01; EC: p < 0.001) ones. For GABA, no significant difference between groups was found in CA1 (F < 1), CA2/3 (F(2, 27) = 2.22, p = 0.13), DG (F(2, 27) = 3.19, p = 0.06), PFC (F < 1), EC (F < 1), PRC (F < 1), POR (F(2, 27) = 1.27, p = 0.30), or TE (F < 1). For agmatine, a one-way ANOVA revealed significant differences between groups in CA1 (F(2, 27) = 4.46, p = 0.02), CA2/3 (F(2, 27) = 9.31, p = 0.0008), PFC (F(2, 27) = 7.12, p = 0.003), PRC (F(2, 27) = 4.32, p = 0.023), POR (F(2, 27) = 9.45, p = 0.0008) and TE (F(2, 27) = 6.66, p = 0.0045), but not DG (F < 1) and EC (F(2, 27) = 1.74, p = 0.19). In CA1, although agmatine levels in both aged and middle-aged groups were lower relative to the young one, post hoc test indicated that the only statistical difference was between the middle-aged and young groups (p < 0.05). In CA2/ 3, PRC, POR and TE, increased agmatine levels were found in the aged group as compared to the middleaged (all p < 0.05) and young (CA2/3 and POR: all p < 0.001; PRC: p < 0.05; TE: p < 0.01) ones, and in the middle-aged group as compared to the young one (p < 0.05) in POR only. In PFC, there was dramatically decreased agmatine level in the aged group relative to the middle-aged and young ones (all p < 0.01) with no difference between the latter two.

For putrescine, there were significant differences between groups in CA1 (F(2, 27) = 4.38, p = 0.01), DG (F(2, 27) = 3.68, p = 0.038), PFC (F(2, 27) = 5.26, p = 0.01) and TE (F(2, 27) = 4.49, p = 0.02), but not CA2/3, EC, PRC and POR (all F < 1). Post-hoc test revealed significantly decreased putrescine levels in the aged (CA1, DG and TE: all p < 0.05; PFC: p < 0.01) and middle-aged (TE only, p < 0.05) groups relative to the young one. For spermidine, a one-way ANOVA revealed significant differences between groups in CA1 (F(2, 27) = 29.19, p < 0.0001), CA2/3 (F(2, 27) = 68.45, p < 0.0001), PFC (F(2, 27) = 45.37, p < 0.0001) and TE (F(2, 27) = 18.66, p < 0.0001), but not DG (F < 1), EC (F(2, 27) = 3.04, p = 0.065), PRC (F < 1) and POR (F(2, 27) = 2.22, p = 0.13). Post-hoc test indicated higher levels of spermidine in the aged group relative to the middle-aged (CA1, CA2/3, PFC and TE: all p < 0.001) and young (CA1, CA2/3 and PFC: all p < 0.001; TE: p < 0,01) ones, and in the middle-aged group when compared to the young one (CA1, CA2/3 and PFC: all p < 0.001). However, a lower level of spermidine was found in the middle-aged group relative to the young one (p < 0.05) in TE. For spermine, there were significant differences between groups in PFC (F(2, 27) = 16.84, p < 0.0001) and PRC (F(2, 27) = 5.13, p = 0.013), but not CA1 (F < 1), CA2/3 (F(2, 27) = 2.07, p = 0.15), DG (F < 1), EC (F < 1), POR (F(2, 27) = 1.12, p = 0.34) and TE (F(2, 27) = 1.71, p = 0.20). In PFC, post hoc test revealed significantly increased spermine levels in the aged and middle-aged groups relative to the young one (all p < 0.001), and in the aged group when compared to the middle-aged one (p < 0.001). In PRC, there were significantly decreased spermine levels in the aged and middle-aged groups relative to the young one (all p < 0.05) with no difference between the former two.

Cluster analyses. Figs. 5 and 6 illustrate the clusters of nine inter-related neurochemicals in the sub-regions of the hippocampus, and the prefrontal, entorhinal, perirhinal, postrhinal and temporal cortices across the three age groups. In young rats, L-ornithine, putrescine, spermidine, spermine and agmatine tended to vary together, whereas L-arginine, L-citrulline, glutamate and GABA tended to co-vary. The most discrete clusters consisted of: (1) L-arginine and L-citrulline (EC and PRC), spermidine (POR), glutamate (DG) or GABA (PFC and TE); (2) L-citrulline and L-ornithine (CA2/3), agmatine (PFC and TE) or glutamate (CA1 and POR); (3) L-ornithine and agmatine (PRC), glutamate (PFC) or GABA (EC); (4) agmatine and putrescine (CA1, CA2/3, DG, EC, and POR); (5) putrescine and spermidine (PRC); (6) spermidine and spermine (PFC and EC), and (7) glutamate and GABA (CA2/3 and POR). Cluster analyses of the middle-aged and aged groups revealed different patterns of the clusters in each brain region when compared to the young group. In CA1, for example, one cluster in the young, middle-aged or aged group was arginine–citrulline–glutamate–spermine,

187

N. Gupta et al. / Neuroscience 226 (2012) 178–196

Young

Middle-aged

Aged

A (CA1)

B

C

D (CA2/3)

E

F

G (DG)

H

I

J (PFC)

K

L

Fig. 5. Dendograms showing the similarities in the degree of expression of the nine neurochemical variables (arginine, citrulline, ornithine, agmatine, putrescine, spermidine, spermine, glutamate and GABA) in the CA1 (A, B, C), CA2/3 (D, E, F), dentate gyrus (DG; G, H, I) sub-regions of the hippocampus and the prefrontal cortex (PFC; J, K, L) in the young, middle-aged and aged groups (n = 10 in each group).

arginine–citrulline–glutamate–putrescine–GABA or arginine–ornithine–agmatine respectively, with the remaining variables forming the second one (Fig. 5A, C). In CA2/3, there were spermine–glutamate–GABA, agmatine–putrescine and agmatine–putrescine– spermidine as one cluster in the young, middle-aged and aged groups respectively, with the remaining variables forming the other (Fig. 5D, F). In DG, one cluster was arginine–glutamate–spermine–GABA, agmatine alone or arginine–agmatine in the young, middle-aged or aged group respectively, with the remaining variables forming another one (Fig. 5G–I). In PFC, spermidine and spermine (in the middle-aged

group), as well as putrescine (the young group), formed one cluster with the remaining formed the second, whereas agmatine and putrescine became one cluster with the rest forming the second in the aged group (Fig. 5J, L). These results suggest that ageing has major effects on the formation of neurochemical clusters and the co-variation of these variables. The clusters of nine inter-related neurochemicals also changed with age in the parahippocampal region and temporal cortex. In EC, for example, one cluster in the young, middle-aged or aged group was spermidine– spermine, citrulline–putrescine–ornithine–agmatine or agmatine–putrescine respectively, with the remaining

188

N. Gupta et al. / Neuroscience 226 (2012) 178–196

Young

Middle-aged

Aged

A (EC)

B

C

D (PRC)

E

F

G (POR)

H

I

J (TE)

K

L

Fig. 6. Dendograms showing the similarities in the degree of expression of the 9 neurochemical variables (arginine, citrulline, ornithine, agmatine, putrescine, spermidine, spermine, glutamate and GABA) in the entorhinal (EC; A, B, C), perirhinal (PRC; D, E, F), postrhinal (POR; G, H, I) and temporal (TE; J, K, L) cortices in the young, middle-aged and aged groups (n = 10 in each group).

variables forming the second one (Fig. 6A, C). There was one cluster of ornithine–agmatine, agmatine–glutamate or putrescine–spermine–spermidine in PRC (Fig. 6D, F), or agmatine–putrescine, putrescine–spermidine–spermine– glutamate–GABA or agmatine–putrescine in POR (Fig. 6G, I), in the young, middle-aged or aged group, respectively. In TE, one cluster was citrulline–agmatine– spermine–putrescine–spermidine in the young group, or agmatine–putrescine in the middle-aged and aged groups, with the remaining formed the second one (Fig. 6J, L).

As L-arginine can be metabolized by NOS, arginase and ADC to form L-citrulline, L-ornithine and agmatine respectively (Zhang and Snyder, 1995; Wu and Morris, 1998), we further compared the distance levels between L-arginine and its three main metabolites in each brain region across the three age groups during the first eight steps of clustering (Table 2). In CA1, L-arginine had the shortest distance to L-citrulline, was far away from L-ornithine and did not join agmatine in both the young and middle-aged groups. In the aged group, however, it became very close to L-ornithine followed by agmatine,

189

N. Gupta et al. / Neuroscience 226 (2012) 178–196

Table 2. Summary of the distance levels (arbitrary unit) between L-arginine and its three main metabolites (L-citrulline, L-ornithine and agmatine) in the CA1, CA2/3 and dentate gyrus (DG) of the hippocampus and the prefrontal (PFC), entorhinal (EC), perirhinal (PRC), postrhinal (POR) and temporal (TE) cortices in the young (Y), middle-aged (MA) and aged (A) groups (n = 10 in each group) Region

Group

Arginine-citrulline

Arginine-ornithine

Arginine-agmatine

CA1

A MA Y

2.44 0.71 0.46

0.38 1.33 1.85

0.87 – –

CA2/3

A MA Y

0.93 1.56 1.13

0.47 – –

2.15 1.98 0.77

DG

A MA Y

1.96 0.34 0.58

– 0.96 2.37

1.14 1.37 –

PFC

A MA Y

0.17 1.21 1.79

0.86 – 0.95

2.85 0.17 –

EC

A MA Y

1.61 2.03 0.40

0.46 – 1.94

2.46 – –

PRC

A MA Y

0.65 0.47 0.21

0.25 – 1.85

0.34 1.38 –

POR

A MA Y

1.05 1.10 2.45

0.86 – –

1.72 – 2.98

TE

A MA Y

0.88 1.22 2.34

0.96 0.12 0.51

2.36 2.25 –

Note: ‘–‘ indicates no clusters joined during the first eight steps of clustering.

and was far away from L-citrulline. In CA2/3, the shortest distance was found in arginine–agmatine followed by arginine–citrulline with no arginine–ornithine cluster present in the young group. However, the pattern was changed to the shortest distance in arginine–ornithine followed by arginine–citrulline and arginine–agmatine in the aged group. In DG, L-arginine had the shortest distance to L-citrulline followed by L-ornithine and agmatine in both the young and middle-aged groups, but such pattern did not exist in the aged group. In PFC, L-arginine was close to L-ornithine in the young group, and then agmatine and L-citrulline in the middle-aged and aged groups, respectively. In EC, L-arginine was very close to L-citrulline in the young group, and then became very close to L-ornithine in the aged group. In PRC, L-arginine had short distance to L-citrulline in all three groups, but also became closer to L-ornithine and agmatine in the aged group. In POR, L-arginine was far away from L-citrulline and agmatine and did not join L-ornithine in the young group. However, it became closer to L-ornithine in the aged group. In TE, L-arginine was close to L-ornithine in all three age groups, but also became closer to L-citrulline in the aged group. These findings demonstrate the interplay between L-arginine and its three main metabolites in a region-specific manner in young rats, which is altered dramatically by ageing.

Correlations between behavioural and neurochemical variables We further analysed the relationships between animals’ behavioural performance in the elevated plus maze (time spent in the open arms), open field (duration of wall-supported rearings) and water maze (mean path length generated during place navigation, percentage path length in the target quadrant and platform crossings during probe tests) and neurochemical levels in the eight brain regions using multiple regression analysis. As there were repeated measures in the behavioural and neurochemical data, the level of significance was set at p 6 0.01 (equivalent to a Geisser-Greenhouse correction for potential violation of the assumption of sphericity; Zolman, 1993). Fig. 7 illustrates the significant correlations in each age group. In the young group, the L-ornithine level in CA2/3 was positively correlated with the duration of rearings in the open field (r = 0.79, p = 0.006, Fig. 7A) and the number of platform crossings during Probe 1 (r = 0.76, p = 0.01; Fig. 7B). The CA2/3 GABA level was positively correlated with the mean path length generated during place navigation (r = 0.86, p = 0.001, Fig. 7C). In the middle-aged group, there were positive correlations between the number of platform crossings during Probe 1 and the CA1 glutamate (r = 0.82,

190

N. Gupta et al. / Neuroscience 226 (2012) 178–196

Fig. 7. Scattergrams showing the significant correlations between behavioural measures and neurochemical variables in the CA1, CA2/3 and dentate gyrus (DG) sub-regions of the hippocampus and the perirhinal (PRC) and postrhinal (POR) cortices in the young, middle-aged and aged groups (n = 10 in each group; see text for detailed description). The dimension of concentrations of L-arginine, L-ornithine, agmatine, putrescine, glutamate and GABA at the x-axis was lg/g wet tissue.

N. Gupta et al. / Neuroscience 226 (2012) 178–196

p = 0.004; Fig. 7D) and PRC putrescine (r = 0.88, p = 0.0008; Fig. 7I) levels, and negative correlations between the mean path length generated during place navigation and the CA2/3 L-arginine (r = 0.76, p = 0.01; Fig. 7E) and DG GABA (r = 0.76, p = 0.01; Fig. 7F) levels. The duration of wall-supported rearings was positively associated with the levels of L-arginine (r = 0.82, p = 0.004; Fig. 7G) and L-aornithine (r = 0.88, p = 0.0008; Fig. 7H) in PRC. In the aged group, the time spent in the open arms was positively correlated with the CA1 agmatine (r = 0.79, p = 0.009; Fig. 7J), PRC L-arginine (r = 0.80, p = 0.006; Fig. 7L) and POR L-arginine (r = 0.79, p = 0.006; Fig. 7P) and L-ornithine (r = 0.79, p = 0.007; Fig. 7Q) levels, but was negatively correlated with the CA2/3 putrescine level (r = 0.76, p = 0.01; Fig. 7K). The PRC putrescine level was negatively correlated with the duration of wall-supported rearings (r = 0.76, p = 0.01; Fig. 7M) in the open field, and the percentage path length in the target quadrant (r = 0.78, p = 0.008; Fig. 7N) and number of platform crossings (r = 0.83, p = 0.003; Fig. 7O) during Probe 1. There was also a negative correlation between the POR putrescine level and number of platform crossings during Probe 1 (r = 0.78, p = 0.01; Fig. 7R).

DISCUSSION Age-related behavioural deficits Ageing, a multi-factorial process, leads to altered behavioural function, including learning and memory. The present study assessed behavioural performance of young, middle-aged and aged rats in the elevated plus maze, open field and water maze tasks. The elevated plus maze is a commonly used test to assess the anxiety level in rodents based on their natural aversion for open and elevated sites (Pellow et al., 1985; Rodgers and Dalvi, 1997). In this task, aged rats spent significantly more time in the open arms relative to the middle-aged and young ones, indicating reduced anxiety level in the aged group (Rowe et al., 1998; Pisarska et al., 2000; Boguszewski and Zagrodzka; 2002; TorrasGarcia et al., 2005). The open field is a test to analyse rodents’ exploratory and locomotion behaviour based on their natural conflict between exploration of and aversion against bright open areas in a novel environment (Schmitt and Hiemke, 1998). We found that both aged and middle-aged rats made significantly less wall-supported rearings, but not path length, when compared to the young ones, suggesting reduced exploratory activity with age. Exploratory behaviour has often been used to assess animal’s ability to integrate spatial features into a representation of a novel environment (for a review see Thinus-Blanc, 1996) and is important for both allothetic and idiothetic navigation (Whishaw and Brooks, 1999). Hence, reduced exploratory activity may contribute to declined spatial learning and memory during ageing. It is of interest to note that the aged group spent significantly less time in the outer zone of the apparatus relative to other two groups. Sine time spent in the

191

centre and/or peripheral of the open field apparatus is an index of anxiety (Schmitt and Hiemke, 1998), this finding appears to be consistent with reduced anxiety in aged rats as suggested by the elevated plus maze data. We have previously reported decreased exploratory and locomotor activities in the open field with age in the same strain of animals that were fed with geriatric diet and under food restriction from 8 months of age (Liu et al., 2004b, 2005). The water maze is a commonly used task to assess rodents’ spatial learning and memory ability, and normal function of the hippocampus is essential for this task (Morris et al., 1982; Moser and Moser, 1998). In the present study, animals were trained to escape from water by swimming to a hidden platform using distal extra-maze cues (place navigation) and a visible platform (cued navigation). Aged rats generated significantly longer path length to reach the platform and swam close to the wall (thigmotaxic swimming) during the place navigation, and showed poor performance in the probe tests relative to the middle-aged and young ones, which are consistent with previous findings (Rowe et al., 1998; Liu et al., 2004b, 2005). It should be noted that in the present study two probe trials were conducted at 120 s (Probe 1) and 24 h (Probe 2) after the final training trial during place navigation, respectively. As animals experienced no platform presented during Probe 1, extinction (suppressed preference to the target quadrant) might occur during Probe 2. Since the young group generated similar percentage of path length in the target quadrant during the two probes, the difference between the aged and young groups in this measurement in Probe 2 may reflect age-related deficit in memory retrieval at a longer retention time, rather than extinction. It has been documented that more than two trials of probe testing likely result in extinction, and that four or more consecutive probe trials are often conducted to assess animals’ ability to decrease or extinguish its spatial bias for the previous platform location (Terry, 2009). It should also be pointed out that the aged group displayed markedly slower swimming speed and poor performance during cued navigation. Hence, age-related motor and/or sensory deficits may contribute to the performance impairments during the place navigation in these animals to a certain extent. Age-related changes in arginine metabolism Accumulating evidence suggests the involvement of arginine metabolism in the ageing and neurodegenerative processes (Law et al., 2001; McCann et al., 2005; Malinski, 2007). The present study systematically investigated how ageing affected arginine metabolic profiles in the hippocampus, parahippocampal region, prefrontal cortex and temporal cortex (an auditory cortex). We found increased total NOS activities with age in CA2/3, PFC, EC, POR and TE, and decreased and increased arginase activities with age in PFC and PRC respectively. It is of interest to mention the overall reverse patterns of age-related changes in NOS and arginase activities in PFC, which

192

N. Gupta et al. / Neuroscience 226 (2012) 178–196

may reflect the competitive relationship between the two enzymes in sharing the substrate L-arginine. We previously reported increased total NOS activities in DG, PFC, PRC, POR and TE and increased arginase activity in PRC in 24–month- old male SD rats that were under food restriction from 8 months of age with geriatric diet and were behaviourally tested (Liu et al., 2004a,b, 2005). Liu et al. (2009b) investigated the effect of ageing on NOS and arginase activities in young, middleaged and aged male SD rats with no behavioural experience, and found age-related increases in total NOS activities in CA1, CA2/3, PFC, EC, POR and TE, and in arginase activities in CA1, PFC, PRC, POR and TE. The differences in animals’ experience and experimental procedures may contribute to the discrepancies. It should be noted that iNOS activity was not measured in the present study, as both the activity and protein expression were not detectable in the same strain of animals housed in the similar environment (Liu et al., 2003a,b, 2004a,b, 2005). The present study also quantified the tissue concentrations of L-arginine and its metabolites in the 8 brain regions across the three age groups. For L-arginine, there were age-related increases in CA1, CA2/3 and PRC and decrease in POR, which have no consistent relationships with the NOS and/or arginase data. While L-citrulline levels were largely unchanged with age (except for age-related decrease in TE), L-ornithine showed age-related increases in all eight regions examined. It is obvious that the tissue concentrations of L-citrulline and L-ornithine do not correspond to the activities of NOS and arginase respectively. It has been shown that L-citrulline can also be formed from L-ornithine by ornithine transcarbamylase and recycled to generate L-arginine by argininosuccinate synthetase and argininosuccinate lyase, and that L-ornithine can be metabolized to form polyamine putrescine by ODC and to L-glutamyl-csemialdehyde, which is further converted to glutamate by P5C dehydrogenase (Wu and Morris, 1998; Wiesinger, 2001). The present study also found decreased glutamate levels with age in CA1, CA2/3 and EC, with no age-related changes in GABA in all eight regions examined. It should be noted that the patterns of age-related changes in L-arginine, L-citrulline, L-ornithine and glutamate observed in the present study are quite different from that of Liu et al. (2009b). The difference in animals’ experience (with and without behavioural testing) may contribute to the discrepancies. The levels of agmatine (decarboxylated arginine) were significantly decreased with age in the CA1 and PFC, but increased with age in the CA2/3, PRC, POR and TE. It should be noted that agmatine has an important role in regulating the production of NO by inhibiting neuronal and inducible NOS and stimulating endothelial NOS (Satriano, 2003; Halaris and Plietz, 2007; Santhanam et al., 2007). Interestingly, there were parallel changes in agmatine and total NOS activity in the CA2/3 PRC, POR and TE, however with the reversed patterns in the CA1 and PFC. As illustrated in Fig. 8, for example, agmatine was negatively correlated with NOS in the

Fig. 8. Scattergrams showing the correlations between agmatine and total NOS activity in the prefrontal (PFC; A), postrhinal (POR; B) and temporal (TE; C) cortices across the young, middle-aged and aged rats (see text for detailed description). The dimension of concentration of agmatine was lg/g wet tissue. NOS activity was expressed as pmol [3H] L-citrulline/30 min/mg protein.

PFC (r = 0.61, p = 0.0004), but was positively correlated with NOS in the POR (r = 0.49, p = 0.006) or TE (r = 0.57, p = 0.0009). These findings suggest that increased level of agmatine in a given region may be a compensatory mechanism to control age-related elevation in NOS activity, whereas lower level of agmatine may fail to do so. Since endogenous agmatine is involved in the processes of learning and memory as a neurotransmitter (Liu et al., 2008b, 2009a; Leitch et al., 2011; Seo et al., 2011), decreased level of agmatine may also lead to altered neurotransmission. Liu et al. (2008a) reported age-related changes in

N. Gupta et al. / Neuroscience 226 (2012) 178–196

agmatine in the sub-regions of the hippocampus, parahippocampal region and the prefrontal and temporal cortices in male SD rats without behavioural testing, and the overall results were similar to the present study. It has been shown that putrescine, spermidine and spermine are critical for cell proliferation and differentiation, synthesis of DNA, RNA and proteins, protein phosphorylation, signal transduction, as well as the regulation of neurotransmitter receptors (for reviews see Williams, 1997; Wallace, 2000; Oredsson, 2003). Therefore, the physiological concentrations of polyamines are essential for cells to grow and to function in an optimal manner. The present study found significantly decreased putrescine levels with age in CA1, DG, PFC and TE, increased spermidine levels with age in CA1, CA2/3, PFC and TE, increased spermine level with age in PFC but decreased spermine level with age in PRC. These findings suggest that ageing affects brain polyamines in a region-specific manner. Previous research has shown that polyamines modulate learning and memory by interacting with the polyamine binding site at the NMDA receptors, with spermidine and spermine being the positive modulators and putrescine as a negative influencer (Williams et al., 1991; Rock and Macdonald, 1995; Williams, 1997). Hence, increased spermidine level with age in CA2/3 may partially contribute to age-related hyperfunction in the CA3 subregion (for a review see Wilson et al., 2006). Since polyamines can be regulated tightly by NMDA receptors (Fage et al., 1992), altered polyamine levels could also be a compensatory mechanism to maintain the normal function of the receptors. Malaterre et al. (2004) demonstrated a novel role of putrescine in neurogenesis. Hence, decreased putrescine level with age in the DG may contribute to age-related impairments in hippocampal neurogenesis to a certain extent. It is of interest to mention that Virgili et al. (2001) compared the putrescine, spermidine and spermine levels in the whole hippocampus between aged and young rats, and failed to detect age-related alterations for all three polyamines. Given the functional dissociation across the three subregions of the hippocampus (Kesner et al., 2004; Kesner, 2009), the examination of the whole hippocampus might have masked age-related changes in polyamines in the hippocampus at the sub-regional level. Liu et al. (2008c) reported the effects of ageing on polyamines in the sub-regions of the hippocampus, parahippocampal region and prefrontal cortex in male SD rats without behavioural testing, and the results were slightly different when compared to the present study. It is of interest to mention that multiple regression analysis revealed a number of significant correlations between the behavioural and neurochemical variables. In the young group, higher levels of L-ornithine in CA2/3 appear to be associated with better performance in the open field (higher exploratory activity) and water maze probe test (better memory for the platform location), whereas higher levels of GABA in CA2/3 correlate with poor spatial learning in the water maze (greater path length to reach the hidden platform during place navigation). In the middle-aged group, higher levels of

193

glutamate (CA1), L-arginine (CA2/3 and PRC, GABA in DG), L-ornithine (PRC) and putrescine (PRC) are associated with better performance in the open field and water maze. In the aged group, animals with high levels of agmatine (CA1), L-arginine (PRC and POR) and L-ornithine (POR), but lower levels of putrescine (PRC), appear to be less anxious (more time spent in the open arms in the elevated plus maze). Moreover, aged rats with higher levels of putrescine in both PRC and POR seem to perform poorly in the open field (reduced exploratory activity) and water maze (poor memory for the platform location). These results clearly demonstrate a link between L-arginine and its metabolites and the behavioural functions (including learning and memory), although more studies need to be carried out in the future to fully understand the functional significance of these correlations. The present study reported the effects of ageing on L-arginine and its eight metabolites in a single study, which allowed us to perform the cluster analysis to determine whether the nine inter-related neurochemical variables clustered into distinct groups and whether the pattern changed as a function of age. There were two distinct clusters in each brain region, suggesting the interplay among the nine neurochemical variables in a region-specific pattern. However, the clusters were quite different across the three age groups, suggesting the effects of age on the way that these variables co-vary. In terms of the known neurobiology, chemically related agmatine, putrescine, spermidine and spermine varied together, albeit in different clusters. L-arginine, L-citrulline, L-ornithine, glutamate and GABA also co-varied, which would be expected given that NOS and arginase produces L-citrulline and L-ornithine from L-arginine, respectively, and that glutamate and GABA can be derived from L-ornithine. Table 2 presents the distance levels between L-arginine and its three main metabolites (L-citrulline, L-ornithine and agmatine by NOS, arginase and ADC, respectively) in each brain region across the three age groups. In the young group, L-arginine was close to either L-citrulline (CA1, DG, EC and PRC), L-ornithine (PFC and TE) or agmatine (CA2/ 3). In the aged group, however, L-arginine was close to L-ornithine in most of the brain regions examined. These findings suggest that the interplay among the NOS, arginase and ADC pathways in young rats appears to be region-specific, and that the ageing process alters their relationship and make the arginase pathway more predominant. It should be pointed out that L-arginine is one of the most metabolically versatile amino acids. Hence future research is required to better understand the mechanisms underlying these changes and their functional significance. Effects of behavioural experience on arginine metabolism As described above, we have previously compared NOS and arginase activity (Liu et al., 2009b), and the levels of L-arginine, L-citrulline, L-ornithine, glutamate and GABA (Liu et al., 2009b), agmatine (Liu et al., 2008a),

194

N. Gupta et al. / Neuroscience 226 (2012) 178–196

polyamines putrescine, spermidine and spermine (Liu et al., 2008c) in the sub-regions of the hippocampus, parahippocampal region and the prefrontal and temporal cortices in young, middle-aged and aged male SD rats without behavioural testing. The neurochemical data reported in the three earlier studies were obtained from the same set of animals that were actually housed along with the set of animals used in the present study. Animals in both sets (with and without behavioural testing) were then sacrificed together to collect the brain tissue samples with a counterbalanced order. For each brain region, the tissues from the two sets of animals were processed under one experimental condition at the same time and the order was counterbalanced. Hence, a direct comparison between the present and earlier studies was made to determine the effects of animals’ behavioural experience on arginine metabolism across the three age groups. Table 3 summarizes the differences in each neurochemical measure in 8 brain regions across the young, middle-aged and aged groups between the present and earlier studies (Liu et al., 2008a,c, 2009b). A difference was regarded as statistically significant (p < 0.05) if the ratio of the absolute value of the mean difference between the studies to the standard error of their difference was greater than 2.11 (based on the t-distribution using 17 degrees of freedom). We found

significant differences in NOS activity (CA1: young group; DG: all groups; PRC and TE: aged group; POR: middle-aged group), arginase activity (CA2/3, PFC and POR: middle-aged and aged groups; DG: aged group; EC: all groups; PRC: young and middle-aged groups; TE: young group), L-arginine (DG: aged group; PFC, EC and PRC: middle-aged group), L-citrulline (PFC and TE: aged group; EC: middle-aged group), L-ornithine (PRC: middle-aged group), glutamate (PFC: aged group), agmatine (TE: young group), putrescine (PRC: young group), spermidine (CA2/3 and TE: aged group; EC: young group) and spermine (PRC: young group). These findings suggest that animals’ behavioural experience does affect the activities of the enzymes involved in arginine metabolism and the levels of L-arginine and its metabolites in the region- and/or age-specific manner. In conjunction with previous studies showing different patterns of age-related changes in NOS and arginase activity (Liu et al., 2004b, 2005), the present study further demonstrates the importance of behavioural control in arginine metabolism research.

SUMMARY The present study demonstrated that ageing resulted in altered anxiety level, reduced exploratory activity and impaired spatial learning and memory. In addition,

Table 3. Differences in NOS and arginase activity and the levels of L-arginine and its metabolites in the CA1, CA2/3 and dentate gyrus (DG) of the hippocampus and the prefrontal (PFC), entorhinal (EC), perirhinal (PRC), postrhinal (POR) and temporal (TE) cortices in young (Y), middle-aged (MA) and aged (A) rats with (the present study) and without (Liu et al., 2008a,c, 2009b) behavioural tests Region Group NOS activity

Arginase activity

L-Arginine

L-Citrulline

L-Ornithine

Glutamate GABA Agmatine Putrescine Spermidine Spermine

CA1

A MA Y

0.10 0.24 2.44⁄

2.04 1.48 0.76

0.96 1.54 0.16

0.52 0.53 0.31

0.70 1.74 1.14

1.91 0.31 0.22

CA2/3

A MA Y

1.47 0.18 1.41

2.79⁄ 3.70⁄⁄ 2.08

1.63 0.14 1.35

0.42 0.38 0.94

0.57 0.70 1.08

DG

A MA Y

2.53⁄ 2.39⁄ 2.32⁄

1.18 0.95 1.51

2.44⁄ 0.40 0.55

0.03 0.13 0.95

PFC

A MA Y

0.20 1.17 1.47

0.30 3.55⁄⁄ 7.77⁄⁄⁄ 2.18⁄ 0.11 0.26

EC

A MA Y

0.40 0.80 1.80

PRC

A MA Y

4.33⁄⁄⁄ 0.97 1.29

POR

A MA Y

2.01 3.81⁄⁄ 0.95

3.15⁄⁄ 2.92⁄⁄ 0.00

1.90 1.96 1.61

1.41 0.53 0.12

TE

A MA Y

7.29⁄⁄⁄ 0.17 1.49

1.91 1.41 2.48⁄

0.01 0.03 0.33

2.67⁄ 1.34 1.46

0.12 0.27 1.36 1.40 1.61 0.49

0.93 1.22 1.30

0.33 1.13 0.18

0.67 0.54 1.03

1.40 0.14 1.00

0.26 0.28 1.44 0.29 0.88 1.67

0.44 0.56 0.87

2.24⁄ 1.21 0.71

0.37 0.60 0.92

1.58 0.17 1.00

1.04 0.31 0.96

1.11 0.74 0.01

0.43 0.93 1.67

0.15 0.14 0.11

1.25 1.10 0.33

2.31⁄ 1.79 1.77

0.81 0.80 1.13

2.74⁄ 1.04 1.71

1.83 0.02 0.86

1.21 0.84 0.43

0.82 0.29 1.11

0.63 1.71 0.70

3.95⁄⁄ 1.12 6.33⁄⁄⁄ 2.69⁄ 3.93⁄⁄ 1.65

0.71 2.37⁄ 1.08

0.35 1.56 1.39

0.85 0.38 0.95

1.15 1.26 0.97

0.39 0.03 0.23

1.33 0.54 1.31

0.32 0.58 2.75⁄

1.55 0.40 0.59

1.31 0.57 4.93⁄⁄⁄ 3.67⁄⁄ 3.58⁄⁄ 0.31

0.76 1.96 0.88

1.86 2.63⁄ 0.94

0.65 1.38 0.49

0.43 0.03 0.95

1.04 0.39 1.36

0.46 0.64 2.19⁄

0.67 1.12 0.19

0.48 0.98 2.56⁄

1.62 0.92 2.03

1.47 0.47 1.25

0.23 0.42 1.60

0.64 2.10 1.31

Note: ⁄ indicates the significant difference at ⁄p < 0.05; indicates lower levels in the present study.

⁄⁄

p < 0.01;

⁄⁄⁄

0.75 2.00 0.81 0.79 0.21 1.05

1.06 1.03 0.12 0.44 0.84 1.00

1.30 0.54 1.06

1.45 1.13 0.78

0.39 0.88 0.70

1.16 1.22 0.82 0.71 0.78 2.93⁄⁄

1.29 0.06 1.58

2.56⁄ 0.16 1.58

0.10 0.61 1.06

p < 0.001 (greater than 2.11, 2.9 and 3.96 t-distribution values at 17 df, respectively). Negative sign

N. Gupta et al. / Neuroscience 226 (2012) 178–196

ageing altered the activities of NOS and arginase and the levels of L-arginine and its metabolites, as well as their interplay, in the sub-regions of the hippocampus, and the prefrontal, entorhinal, perirhinal, postrhinal and temporal cortices in a region-specific manner. The present study further supported the involvement of arginine metabolism in the ageing process, and provided further evidence of the effects of animals’ behavioural experience on arginine metabolism. It has been documented that the decline in cognitive function starts in middle-age and continues throughout the ageing process (Finch, 2009; Salthouse, 2009). Although the middle-aged rats displayed no or very mild behavioural deficits, altered NOS activity and levels of L-arginine, L-ornithine, agmatine, putrescine, spermidine and spermine were clearly noticed in some of the brain regions in these animals. Hence, targeting neurochemical changes during the middle-age may be a good strategy to prevent further cognitive decline. Acknowledgements—This work was supported by New Zealand Neurological Foundation and Lottery Health Board. The authors would like to thank Renuka Devaraj, Sree Chary, and the technical staff in the Department of Anatomy and School of Pharmacy for their assistance. Neeraj Gupta was a recipient of the University of Otago Postgraduate Scholarship, and the University of Otago Postgraduate Publishing Bursary (PhD).

REFERENCES Aimone JB, Deng W, Gage FH (2011) Resolving new memories: a critical look at the dentate gyrus, adult neurogenesis, and pattern separation. Neuron 70:589–596. Bergin DH, Liu P (2010) Agmatine protects against b-amyloid25–35induced memory impairments in the rat. Neuroscience 169:794–811. Boguszewski P, Zagrodzka J (2002) Emotional changes related to age in rats – a behavioral analysis. Behav Brain Res 133:323–332. Bradford MM (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of proteindye binding. Anal Biochem 72:248–254. Burke SN, Barnes CA (2006) Neural plasticity in the ageing brain. Nat Rev Neurosci 7:30–40. Calabrese V, Mancuso C, Calvani M, Rizzarelli E, Butterfield DA, Stella AM (2007) Nitric oxide in the central nervous system: neuroprotection versus neurotoxicity. Nat Rev 8:766–775. Deng W, Aimone JB, Gage FH (2010) New neurons and new memories: how does adult hippocampal neurogenesis affect learning and memory? Nat Rev 11:339–350. Drapeau E, Nora Abrous D (2008) Stem cell review series: role of neurogenesis in age-related memory disorders. Aging Cell 7:569–589. Eichenbaum H, Lipton PA (2008) Towards a functional organization of the medial temporal lobe memory system: role of the parahippocampal and medial entorhinal cortical areas. Hippocampus 18:1314–1324. Fabel K, Kempermann G (2008) Physical activity and the regulation of neurogenesis in the adult and aging Brain. Neuromol Med 10:59–66. Fage D, Voltz C, Scatton B, Carter C (1992) Selective release of spermine and spermidine from the rat striatum by N-methyl-Daspartate receptor activation in vivo. J Neurochem 58:2170–2175. Feil R, Kleppisch T (2008) NO/cGMP-dependent modulation of synaptic transmission. Handb Exp Pharmacol:529–560.

195

Finch C (2009) The neurobiology of middle-age has arrived. Neurobiol Aging 30:515–520. Gallagher M, Rapp PR (1997) The use of animal models to study the effects of aging on cognition. Annu Rev Psychol 48:339–370. Greenwood PM (2007) Functional plasticity in cognitive aging: review and hypothesis. Neuropsychology 21:657–673. Gupta N, Zhang H, Liu P (2009) Behavioral and neurochemical effects of acute putrescine depletion by difluoromethylornithine in rats. Neuroscience 161:691–706. Gupta N, Zhang H, Liu P (2012) Chronic difluoromethylornithine treatment impairs spatial learning and memory in rats. Pharmacol Biochem Behav 100:464–473. Halaris A, Plietz J (2007) Agmatine: metabolic pathway and spectrum of activity in brain. CNS Drugs 21:885–900. Holscher C (1997) Nitric oxide, the enigmatic neuronal messenger: its role in synaptic plasticity. Trends Neurosci 20:298–303. Huang EP (1997) Synaptic plasticity: a role for nitric oxide in LTP. Curr Biol 7:R141–R143. Kesner RP (2009) Tapestry of memory. Behav Neurosci 123: 1–13. Kesner RP, Lee I, Gilbert P (2004) A behavioral assessment of hippocampal function based on a subregional analysis. Rev Neurosci 15:333–351. Klempin F, Kempermann G (2007) Adult hippocampal neurogenesis and aging. Eur Arch Psychiatry Clin Neurosci 257:271–280. Knox LT, Jing Y, Fleete MS, Collie ND, Zhang H, Liu P (2011) Scopolamine impairs behavioural function and arginine metabolism in the rat dentate gyrus. Neuropharmacology 61:1452–1462. Law A, Dore S, Blackshaw S, Gauthier S, Quirion R (2000) Alteration of expression levels of neuronal nitric oxide synthase and haem oxygenase-2 messenger RNA in the hippocampi and cortices of young adult and aged cognitively unimpaired and impaired Long– Evans rats. Neuroscience 100:769–775. Law A, Gauthier S, Quirion R (2001) Say NO to Alzheimer’s disease: the putative links between nitric oxide and dementia of the Alzheimer’s type. Brain Res Brain Res Rev 35:73–96. Law A, O’Donnell J, Gauthier S, Quirion R (2002) Neuronal and inducible nitric oxide synthase expressions and activities in the hippocampi and cortices of young adult, aged cognitively unimpaired, and impaired Long-Evans rats. Neuroscience 112:267–275. Leitch B, Shevtsova O, Reusch K, Bergin DH, Liu P (2011) Spatial learning induced increase in agmatine levels at hippocampal CA1 synapses. Synapse 65:146–153. Li G, Regunathan S, Barrow CJ, Eshraghi J, Cooper R, Reis DJ (1994) Agmatine: an endogenous clonidine-displacing substance in the brain. Science 263:966–969. Liu P, Bergin D (2009) Differential effects of icv microinfusion of agmatine on spatial working and reference memory in the rat. Neuroscience 159:951–961. Liu P, Collie ND (2009) Behavioral effects of agmatine in naive rats are task- and delay-dependent. Neuroscience 163:82–96. Liu P, Smith PF, Appleton I, Darlington CL, Bilkey DK (2003a) Nitric oxide synthase and arginase in the rat hippocampus and the entorhinal, perirhinal, postrhinal, and temporal cortices: regional variations and age-related changes. Hippocampus 13: 859–867. Liu P, Smith PF, Appleton I, Darlington CL, Bilkey DK (2003b) Regional variations and age-related changes in nitric oxide synthase and arginase in the sub-regions of the hippocampus. Neuroscience 119:679–687. Liu P, Smith PF, Appleton I, Darlington CL, Bilkey DK (2004a) Agerelated changes in nitric oxide synthase and arginase in the rat prefrontal cortex. Neurobiol Aging 25:547–552. Liu P, Smith PF, Appleton I, Darlington CL, Bilkey DK (2004b) Potential involvement of NOS and arginase in age-related behavioural impairments. Exp Gerontol 39:1207–1222. Liu P, Smith PF, Appleton I, Darlington CL, Bilkey DK (2005) Hippocampal nitric oxide synthase and arginase and ageassociated behavioral deficits. Hippocampus 15:642–655.

196

N. Gupta et al. / Neuroscience 226 (2012) 178–196

Liu P, Chary S, Devaraj R, Jing Y, Darlington CL, Smith PF, Tucker IG, Zhang H (2008a) Effects of aging on agmatine levels in memory-associated brain structures. Hippocampus 18:853–856. Liu P, Collie N, Chary S, Jing Y, Zhang H (2008b) Spatial learning results in elevated agmatine levels in the rat brain. Hippocampus 18:1094–1098. Liu P, Gupta N, Jing Y, Zhang H (2008c) Age-related changes in polyamines in memory-associated brain structures in rats. Neuroscience 155:789–796. Liu P, Rushaidhi M, Collie ND, Leong MT, Zhang H (2008d) Behavioral effects of intracerebroventricular microinfusion of agmatine in adult rats. Behav Neurosci 122:557–569. Liu P, Jing Y, Collie ND, Chary S, Zhang H (2009a) Memory-related changes in L-citrulline and agmatine in the rat brain. Hippocampus 19:597–602. Liu P, Jing Y, Zhang H (2009b) Age-related changes in arginine and its metabolites in memory-associated brain structures. Neuroscience 164:611–628. Liu P, Zhang H, Devaraj R, Ganesalingam GS, Smith PF (2010) A multivariate analysis of the effects of aging on glutamate, GABA and arginine metabolites in the rat vestibular nucleus. Hear Res 269:122–133. Liu P, Jing Y, Collie ND, Campbell SA, Zhang H (2011) Preaggregated Ab(25–35) alters arginine metabolism in the rat hippocampus and prefrontal cortex. Neuroscience 193:269–282. Malaterre J, Strambi C, Aouane A, Strambi A, Rougon G, Cayre M (2004) A novel role for polyamines in adult neurogenesis in rodent brain. Eur J Neurosci 20:317–330. Malinski T (2007) Nitric oxide and nitroxidative stress in Alzheimer’s disease. J Alzheimers Dis 11:207–218. McCann SM (1997) The nitric oxide hypothesis of brain aging. Exp Gerontol 32:431–440. McCann S, Licinio J, Wong ML, Yu W, Karanth S, Rettorri V (1998) The nitric oxide hypothesis of aging. Exp Gerontol 33:813–826. McCann SM, Mastronardi C, de Laurentiis A, Rettori V (2005) The nitric oxide theory of aging revisited. Ann NY Acad Sci 1057:64–84. Moinard C, Cynober L, Bandt JP (2005) Polyamines: metabolism and implications in human diseases. Clin Nutr 24:184–197. Morris R, Garrud P, Rawlins J, O’Keefe J (1982) Place navigation impaired in rats with hippocampal lesions. Nature 297:681–683. Moser MB, Moser EI (1998) Functional differentiation in the hippocampus. Hippocampus 8:608–619. Necchi D, Virgili M, Monti B, Contestabile A, Scherini E (2002) Regional alterations of the NO/NOS system in the aging brain: a biochemical, histochemical and immunochemical study in the rat. Brain Res 933:31–41. Oredsson SM (2003) Polyamine dependence of normal cell-cycle progression. Biochem Soc Trans 31:366–370. Pellow S, Chopin P, File SE, Briley M (1985) Validation of open: closed arm entries in an elevated plus-maze as a measure of anxiety in the rat. J Neurosci Methods 14:149–167. Pisarska M, Mulchahey JJ, Welge JA, Geracioti Jr TD, Kasckow JW (2000) Age-related alterations in emotional behaviors and amygdalar corticotropin-releasing factor (CRF) and CRF-binding protein expression in aged Fischer 344 rats. Brain Res 877:184–190. Reis DJ, Regunathan S (2000) Is agmatine a novel neurotransmitter in brain? Trends Pharmacol Sci 21:187–193. Rock D, Macdonald R (1995) Polyamine regulation of N-methyl-Daspartate receptor channels. Annu Rev Pharmacol Toxicol 35:463–482. Rodgers R, Dalvi A (1997) Anxiety, defence and the elevated plusmaze. Neurosci Biobehav Rev 21:801–810.

Rowe WB, Spreekmeester E, Meaney MJ, Quirion R, Rochford J (1998) Reactivity to novelty in cognitively-impaired and cognitively-unimpaired aged rats and young rats. Neuroscience 83:669–680. Salthouse T (2009) When does age-related cognitive decline begin? Neurobiol Aging 30:507–514. Santhanam AV, Viswanathan S, Dikshit M (2007) Activation of protein kinase B/Akt and endothelial nitric oxide synthase mediates agmatine-induced endothelium-dependent relaxation. Eur J Pharmacol 572:189–196. Satriano J (2003) Agmatine: at the crossroads of the arginine pathways. Ann NY Acad Sci 1009:34–43. Schmitt U, Hiemke C (1998) Combination of open field and elevated plus-maze: a suitable test battery to assess strain as well as treatment differences in rat behavior. Prog Neuropsychopharmacol Biol Psychiatry 22:1197–1215. Seo S, Liu P, Leitch B (2011) Spatial learning-induced accumulation of agmatine and glutamate at hippocampal CA1 synaptic terminals. Neuroscience 192:28–36. Shimamura AP (1995) Memory and the prefrontal cortex. Ann NY Acad Sci 769:151–160. Sugaya K, Chouinard M, Greene R, Robbins M, Personett D, Kent C, Gallagher M, McKinney M (1996) Molecular indices of neuronal and glial plasticity in the hippocampal formation in a rodent model of age-induced spatial learning impairment. J Neurosci 16:3427–3443. Terry Jr AV (2009) Spatial navigation (water maze) tasks. In: Buccafusco JJ, editor. Methods of behavior analysis in neuroscience. Boca Raton, FL: CRC Press. Chapter 13. Thinus-Blanc C (1996) Animal spatial cognition: behavioural and brain approach. Singapore: World Scientific. Tisserand D, Jolles J (2003) On the involvement of prefrontal networks in cognitive aging. Cortex 39:1107–1128. Torras-Garcia M, Costa-Miserachs D, Coll-Andreu M, Portell-Cortes I (2005) Decreased anxiety levels related to aging. Exp Brain Res 164:177–184. Virgili M, Necchi D, Scherini E, Contestabile A (2001) Increase of the ornithine decarboxylase/polyamine system and transglutaminase upregulation in the spinal cord of aged rats. Neurosci Lett 309:62–66. Wallace HM (2000) The physiological role of the polyamines. Eur J Clin Invest 30:72–78. Whishaw IQ, Brooks BL (1999) Calibrating space. exploration is important for allothetic and idiothetic navigation. Hippocampus 9:659–667. Wiesinger H (2001) Arginine metabolism and the synthesis of nitric oxide in the nervous system. Prog Neurobiol 64:365–391. Williams K (1997) Interactions of polyamines with ion channels. Biochem J 325:289–297. Williams K, Romano C, Dichter MA, Molinoff PB (1991) Modulation of the NMDA receptor by polyamines. Life Sci 48:469–498. Wilson IA, Gallagher M, Eichenbaum H, Tanila H (2006) Neurocognitive aging: prior memories hinder new hippocampal encoding. Trends Neurosci 29:662–670. Witter MP, Wouterlood FG, Naber PA, Van Haeften T (2000) Anatomical organization of the parahippocampal-hippocampal network. Ann NY Acad Sci 911:1–24. Wu G, Morris Jr SM (1998) Arginine metabolism: nitric oxide and beyond. Biochem J 336:1–17. Zhang J, Snyder SH (1995) Nitric oxide in the nervous system. Annu Rev Pharmacol Toxicol 35:213–233. Zolman JF (1993) Biostatistics: experimental design and statistical inference. New York: Oxford University Press.

(Accepted 7 September 2012) (Available online 16 September 2012)