- Email: [email protected]
Spatial short-term memory in rats: Effects of learning trials on metabolic activity of limbic structures
Neuroscience Letters 483 (2010) 32–35
Contents lists available at ScienceDirect
Neuroscience Letters journal homepage: www.elsevier.com/locate/neulet
Spatial short-term memory in rats: Effects of learning trials on metabolic activity of limbic structures Magdalena Méndez-López ∗ , Marta Méndez, Azucena Begega, Jorge L. Arias Laboratorio de Neurociencias, Departamento de Psicología, Universidad de Oviedo, Plaza Feijoo s/n, 33003 Oviedo, Spain
a r t i c l e
i n f o
Article history: Received 16 June 2010 Received in revised form 21 July 2010 Accepted 21 July 2010 Keywords: Matching-to-position task Water maze Cytochrome oxidase Hippocampus Mammillary bodies Anterior thalamic nuclei
a b s t r a c t The oxidative metabolism was assessed in the hippocampus and related regions in rats that were trained in a spatial short-term memory task in the water maze following distinct training schedules. The cytochrome oxidase (COx) histochemistry was evaluated in groups of rats that received a daily session made up of either two or three learning trials. An untreated group was added to determine baseline levels of COx. We found that the 3-Trials group exhibited a better performance concerning the differences in latency between trials 1 and 2. Trained groups showed higher COx activity than the untreated group in the medial prefrontal cortex, dentate gyrus, CA1, and the mammillary region. However, a decrease in COx activity was found in the dentate gyrus, CA1, and supramammillary region of the 3-Trials group. In addition, COx activity levels found in this group were similar to those of the untreated group in some thalamic nuclei. Most of the regions that presented significant correlations between COx activity and behavioral scores were found in the 3-Trials group. These findings suggest an influence of task difficulty in the oxidative metabolism of brain regions involved in spatial learning. © 2010 Elsevier Ireland Ltd. All rights reserved.
The water maze is a learning paradigm for assessment of spatial memory in rodents [18] in which animals use visible cues from the environment to find a hidden platform. Within this context, the evaluation protocol for a short-term memory task involves the temporal storage of a spatial location to use within a training session, but not between sessions [5,16,17]. Some studies have shown that differences in the training schedule can influence learning of the water maze task [2,22,24,28]. For example, an improvement in spatial memory was found by separating trial blocks [2,24], reducing the retention interval [2,28], and increasing the number of training sessions [2,22]. Furthermore, some of these studies have shown changes in the number and morphology of hippocampal neurons as a consequence of modification in training variables [2,24]. Studies examining the neural substrates of memory have found that spatial learning is related to the functioning of a neuronal circuit composed of cortical, hippocampal, and diencephalic brain regions [1]. Hence, learning impairment in the water maze has been found after lesions of the dorsal hippocampus [19], medial prefrontal cortex [14], anterior thalamus [27] and mammillary region [23,25]. The cytochrome oxidase (COx) is a mitochondrial enzyme involved in the oxidative phosphorylation process in which ATP
∗ Corresponding author. Tel.: +34 985103212; fax: +34 985104144. E-mail addresses: [email protected], [email protected] (M. Méndez-López). 0304-3940/$ – see front matter © 2010 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.neulet.2010.07.054
is generated and this metabolic energy is necessary to sustain neuronal functions [27]. Histochemical labelling of COx [7] has been extensively used in studies to evaluate the metabolic activity of limbic structures associated with spatial training in the water maze [3,15–17]. Interestingly, it has been shown that this technique can detect significant changes in metabolic activity over training days [3]. In the present study, we examined the effect of changes in the training schedule on brain metabolic activity of subjects trained in a matching-to-position task developed for the assessment of shortterm spatial memory in the water maze. The COx activity of the dorsal hippocampus, medial prefrontal cortex, anterior thalamic nuclei, and mammillary body region was examined by quantitative COx histochemistry in rats submitted to a daily session made up of either two or three learning trials. In addition, correlations between COx activity and behavioral scores were also calculated. Twenty-one male Wistar rats (240–270 g; vivarium of the University) maintained under standard conditions (21 ± 1 ◦ C temperature, 65–70% relative humidity and 12 h light–dark cycle) with ad libitum access to food and water, were used. The experimental procedures were carried out according to the National legislation (Royal Decree no. 1201/2005), which is in compliance with the European Community Council Directive 86/609/EEC for the use and care of laboratory animals. Fourteen rats were trained in the water maze [18] and the remaining rats were used as a control group (CO, n = 7) to determine basal COx activity. These rats remained in their home cage. The maze consisted of a cylindrical tank (150 cm diameter, 40 cm
M. Méndez-López et al. / Neuroscience Letters 483 (2010) 32–35
deep) of water at 22 ± 1 ◦ C and a hidden platform. The pool was surrounded by panels on which the spatial clues were placed (a square, a horizontal line and a vertical line) and was divided into four quadrants to locate the platform and the start positions. Animals were trained on a matching-to-position task that was developed in three consecutive days (one session daily) and received a habituation session one day prior to the initial test [16,17]. Each daily session involved two identical trials for the “2-Trials group” (n = 7) and three identical trials for the “3-Trials group” (n = 7). The trials consisted of releasing the animal from one starting point and letting it swim until it reached the platform or 60 s had elapsed. If the animal had not reached the platform in this time, it was placed on the platform and kept there for 15 s. The intertrial interval was 5 s. The task demands a recall of the position occupied by the platform during the first trial. The locations of the start positions and of the hidden platform were the same during one session but varied on the different days in a pseudorandom order. The escape latencies were recorded. The mean escape latency was calculated statistically for each animal in the trials 1, 2 and 3 (in the case of the 3-Trials group), grouping all sessions carried out. Also, the difference in escape latency scores between trials 1 and 2 of the last session was used as performance index. This difference was defined as saving in escape latency to platform. Ninety minutes after the end of the behavioral task, the animals were decapitated. The CO group was decapitated at the same time as the trained groups. Brains were removed intact, frozen in isopentane (Sigma–Aldrich, Spain) and stored at −40 ◦ C. Coronal sections (30 m) of the brain were processed with quantitative COx histochemistry [7]. Briefly, sets of tissue homogenate standards from Wistar rat brain were cut at different thickness (10, 30, 40 and 60 m) and mounted on a slide. Previously, COx activity of the homogenate was spectrophotometrically assessed [8]. Sets of tissue homogenate standards of known COx activity were used as calibration standards in each staining batch. Sections were fixed with 0.5% (v/v) glutaraldehyde and 10% (w/v) sucrose in phosphate buffer (pH 7.6; 0.1 M). After this, the sections were rinsed in 0.1 M phosphate buffer with 10% (w/v) sucrose. Then 0.05 M Tris buffer, pH 7.6, with 275 mg/l cobalt chloride, 10% (w/v) sucrose, and 0.5% (v/v) dimethylsulfoxide, were applied. Subsequently, they were rinsed and incubated in a solution of 0.06 g cytochrome c (Sigma–Aldrich), 0.016 g catalase, 40 g sucrose, 2 ml dimethylsulfoxide and 0.4 g diaminobenzidine tetrahydrochloride in 800 ml of phosphate buffer (pH 7.6; 0.1 M). Finally, the tissue was fixed in 4% (v/v) buffered formalin with 10% (w/v) sucrose, dehydrated and coverslipped. Quantification of COx histochemical staining intensity was done by densitometric analysis using a computer-controlled microscope (Leica Q550IW) with a light source (Leica DM-RHC) connected to a camera CCD (Cohu, Japan) and an image analysis software (Leica Q-Win). The mean optical density (OD) of each structure was measured on the right side of bilateral structures using three consecutive sections in each animal [16]. In each section, four, nonoverlapping readings were taken using a square-shaped sampling window that was adjusted for each region size. A total of 12 measurements were taken in each region per animal and averaged to obtain a mean. To control staining variations between different staining baths, measurements were taken from stained brain homogenate standards. Regression curves between section thickness and known COx activity measured in each set of standards were calculated for each incubation bath [8]. OD values were then converted to COx activity units (mol cytochrome c oxidized/min/g wet tissue at 23 ◦ C) using the regression curves calculated. The regions of interest were the infralimbic cortex (IL) and prelimbic cortex (PL); the anterodorsal thalamic nucleus (ADT), anteroventral thalamic nucleus (AVT) and anteromedial thalamic nucleus (AMT); the dentate gyrus (DG), CA1 and CA3 subfields
33
Fig. 1. (A) Escape latencies to platform (mean ± S.E.M). Significance of differences between trials, grouping all sessions carried out, *p ≤ 0.001. (B) Difference in latency scores between trials (mean ± S.E.M) within the third session. Significant difference between groups in the 1–2 saving measure, *p = 0.005. Significantly different with respect to the 1–3 saving measure of the 3-Trials group, # p = 0.02.
of the hippocampus; and the medial mammillary nucleus (MM), lateral mammillary nucleus (LM) and supramammillary nucleus (SuM). The stained coronal sections corresponded to Bregma levels [20]: +3.24 mm for IL and PL; −1.92 mm for ADT, AVT and AMT; −3.84 mm for DG, CA1 and CA3; and −4.44 mm for MM, LM and SuM. A paired-sample t-test was used to describe differences between trials in escape latencies, grouping all training sessions (Fig. 1A). A reduction of escape latencies was showed for both trained groups in trial 2 (t6 ≥ 6.81, p ≤ 0.001). In addition, shorter escape latencies were found during trial 3 in the 3-Trials group (t6 = 9.08, p ≤ 0.001). Also, saving from trial 1 to trial 2 within the last session was compared between groups using a t-test. As shown in Fig. 1B, this saving was lower in the 2-Trials group (t12 = −3.47, p = 0.005). Similarly, this saving was significantly lower than the trials 1–3 saving of the 3-Trials group within the last session (t12 = −2.639, p = 0.02; Fig. 1B). The results of the COx activity measures obtained from the different brain regions studied are illustrated in Table 1. COx activity of the studied groups was compared using a one-way analysis of variance and Tukey’s test was applied as post hoc test. COx values taken in the IL, CA3 hippocampal subfield, and AMT were similar between groups (F2,18 ≤ 2.52, p ≥ 0.10). However, differences between groups were found in the PL, MM and LM (F2,18 ≥ 6.83, p < 0.001), where an increase in COx activity was found in the trained groups (p < 0.05). Also, an increase in COx activity was found in the trained groups in the DG, CA1 and SuM (F2,18 ≥ 20.79,
34
M. Méndez-López et al. / Neuroscience Letters 483 (2010) 32–35
Table 1 Cytochrome oxidase activity of the selected brain regions. CO PL IL DG CA3 CA1 ADT AVT AMT MM LM SuM
17.22 16.19 18.15 11.49 12.18 31.46 23.99 16.10 20.30 23.67 9.36
2-Trials ± ± ± ± ± ± ± ± ± ± ±
0.22 0.36 0.58 0.50 0.20 0.90 1.14 1.56 0.66 0.94 0.59
19.29 18.40 24.53 13.45 19.77 36.44 30.64 18.98 29.81 28.10 14.15
± ± ± ± ± ± ± ± ± ± ±
3-Trials 0.46* 0.77 0.72** 1.30 0.59** 1.07** 0.45** 0.70 1.71** 1.56* 0.61**
20.09 17.95 21.11 11.71 15.75 30.92 24.73 17.02 27.48 29.86 11.58
± ± ± ± ± ± ± ± ± ± ±
0.77** 0.94 0.50**,† 0.54 0.45**,† 0.30† 1.40† 0.73 1.63** 1.06** 0.32*,†
Mean ± S.E.M. of COx activity units (mol cytochrome c oxidized/min/g wet tissue at 23 ◦ C) measured in the prelimbic cortex (PL), infralimbic cortex (IL), dentate gyrus (DG), CA1 and CA3 subregions, anterodorsal thalamus (ADT), anteroventral thalamus (AVT), anteromedial thalamus (AMT), medial mammillary (MM), lateral mammillary (LM) and supramammillary (SuM). Tukey Test, p value in relation to: CO (control group) (*p < 0.05; **p < 0.01) and 2-Trials group († p < 0.01).
p < 0.001), but a significant decrease was shown in the 3-Trials group as compared to the 2-Trials group (p ≤ 0.02). Finally, differences among groups were found in the ADT and AVT (F2,18 ≥ 11.40, p < 0.001), where the 2-Trials group showed higher COx activity than all remaining groups (p ≤ 0.003). Pearson product-moment correlations between COx activity and behavioral data were calculated for all measured regions in each trained group, using as performance index the difference in escape latency between trials 1 and 2 of the last session (trials that were common to both trained groups). As shown in Fig. 2, significant negative correlations were detected in some measured regions. These significant correlations were more frequent in the 3-Trials group. This study shows that the neuronal metabolic activity was lower in some limbic system regions when animals were trained using one more daily trial. Also, these rats presented better performance shown by the saving increment in escape latency. Therefore, this
behavioral accuracy could suggest that the insertion of a single and additional trial in the training schedule reduces task difficulty because the animal had more chances to learn the spatial location. Supporting this, it has been shown that an increase of the number of training trials improved the retention capacity of long-term memory for spatial locations in the water maze [2,22]. In relation to the neuronal metabolism, an increase in COx activity was found in the hippocampus, prelimbic cortex, anterior thalamus and mammillary nuclei of the trained groups. In this regard, COx activity levels in these brain regions have previously been reported to increase in rats submitted to the same short-term memory task with two daily trials [15]. Furthermore, the contribution of these regions has been related to the spatial memory process in several studies [9,12,19,21,23,25,26]. Taking into account COx activity levels displayed by each trained group, the reduction shown by the 3-Trials group in the supramammillary nucleus, anterodorsal and anteroventral thalamic nuclei, and the dentate gyrus, and CA1 is of outstanding interest. Since this group received more trials, this reduction might be explained by changes in the contribution of these regions to the task. In relation to this, a time-dependent involvement of brain regions associated with spatial memory was described using the COx histochemistry [3]. In addition, as mentioned above, the insertion of a third learning trial led to a behavioral improvement. Hence, this could indicate that the task is easier to perform. Addressing the relationship between the difficulty of a learning task and brain COx activity, Hu et al. [10] have reported a functional involvement of the dentate gyrus and CA field of rats trained in a difficult visual discrimination task, whereas only the dentate gyrus was involved in the easy version of the same task. Therefore, collectively, these results are consistent with the hypothesis that brain activity is directly related to the difficulty of task performance. In support of these observations, several human studies have revealed greater brain activation as the degree of difficulty of cognitive tasks increased by apply-
Fig. 2. Correlations between COx activity and saving in escape latency (saving from trial 1 to trial 2 within the last session). Significant negative correlations were found in the dentate gyrus (DG), CA1 subfield of the hippocampus, supramammillary (SuM), anterodorsal thalamus (ADT), anteroventral thalamus (AVT) and lateral mammillary (LM).
M. Méndez-López et al. / Neuroscience Letters 483 (2010) 32–35
ing neuroimaging techniques that are also based on the metabolic capacity of the brain [6,11,13]. The neurobehavioral correlations have shown that the mastery of the spatial short-term memory task was most related to neuronal COx activity in the rats trained under the three trials condition. Indeed, most of the regions that presented negative correlations between brain activity and the saving in escape latency were those which showed a significant lower COx activity in this condition. Therefore, this suggests that training conditions were the key factor in modulating brain activity. Finally, the prelimbic cortex, medial mammillary and lateral mammillary were the only nuclei explored in which COx activity increased following learning but did not significantly differ between training conditions. This suggests that mammillary nuclei and prelimbic cortex have an essential involvement in this task that is independent of task difficulty. The contribution of these regions to spatial learning has been proposed previously. Thus, the prelimbic cortex has been related to some cognitive processes such as attention and behavioral flexibility required when the subject must learn a different daily location in the water maze [4,14]. In addition, the lesion of the mammillary nuclei impaired delayed-matchingto-position tasks in the water maze [23,25]. Furthermore, these experiments have shown that this region is involved in the acquisition of spatial information, and its role is independent of task difficulty and is not affected by the magnitude of the retention interval [23,25]. Taken together, the results of the present study suggest that COx histochemistry is sensitive to detect changes in the oxidative metabolism of the hippocampus and related regions involved in spatial short-term memory in response to variations in the training schedule that improve task performance.
[7]
[8]
[9]
[10]
[11] [12]
[13]
[14]
[15]
[16]
[17]
[18] [19]
[20]
Acknowledgements [21]
˜ Díaz for their assistance We thank Piedad Burgos and Begona and V. Navascués for reviewing the English text of this manuscript. This research was supported by grant SEJ2007-63506 and PSI201019348 (Spanish Ministry of Science and Innovation and FEDER).
[22]
[23]
References [1] J.P. Aggleton, M.W. Brown, Episodic memory, amnesia, and the hippocampal anterior thalamic axis, Behav. Brain Sci. 22 (1999) 425–489. [2] K. Bolding, J.W. Rudy, Place learning in the Morris water task. Making the memory stick, Learn. Mem. 13 (2006) 278–286. [3] N.M. Conejo, H. González-Pardo, F. Gonzalez-Lima, J.L. Arias, Spatial learning of the water maze: progression of brain circuits mapped with cytochrome oxidase histochemistry, Neurobiol. Learn. Mem. 93 (2010) 362–371. [4] B. Delatour, P. Gisquet-Verrier, Functional role of rat prelimbic-infralimbic cortices in spatial memory: evidence for their involvement in attention and behavioural flexibility, Behav. Brain Res. 109 (2000) 113–128. [5] P.A. Dudchenko, An overview of the tasks used to test working memory in rodents, Neurosci. Biobehav. Rev. 28 (2004) 699–709. [6] D.C. Glahn, J. Kim, M.S. Gohen, V.P. Poutanen, S. Therman, S. Bava, T.G.M. Van Erp, M. Manninen, M. Huttunen, J. Lönnqvist, C.G. StandertskjöldNordenstam, T.D. Cannon, Maintenance and manipulation in spatial working
[24]
[25]
[26]
[27] [28]
35
memory: dissociations in the prefrontal cortex, NeuroImage 17 (2002) 201–213. F. Gonzalez-Lima, A. Cada, Cytochrome oxidase activity in the auditory system of the mouse: a qualitative and quantitative histochemical study, Neuroscience 63 (1994) 559–578. F. Gonzalez-Lima, D. Jones, Quantitative mapping of cytochrome oxidase activity in the central auditory system of the gerbil: a study with calibrated activity standards and metal-intensified histochemistry, Brain Res. 660 (1994) 34–49. V. Hok, E. Save, P.P. Lenck-Santini, B. Poucet, Coding for spatial goals in the prelimbic/infralimbic area of the frontal cortex, Proc. Natl. Acad. Sci. U.S.A. 102 (2005) 4602–4607. D. Hu, X. Xu, F. Gonzalez-Lima, Hippocampal cytochrome oxidase activity of rats in easy and difficult visual discrimination learning, Int. J. Neurosci. 115 (2005) 595–611. J.M. Jansma, N.F. Ramsey, R. Coppola, R.S. Kahn, Specific versus nonspecific brain activity in a parametric n-back task, NeuroImage 12 (2000) 688–697. B. Kocsis, R.P. Vertes, Characterization of neurons of the supramammillary nucleus and mammillary body that discharge rhythmically with the hippocampal theta rhythm in the rat, J. Neurosci. 14 (1994) 7040–7052. E.D. Leshikar, A.H. Gutchess, A.C. Hebrank, B.P. Sutton, The impact of increased relational encoding demands on frontal and hippocampal function in older adults, Cortex 46 (2010) 507–521. R.J. McDonald, A.L. King, N. Foong, Z. Rizos, N.S. Hong, Neurotoxic lesions of the medial prefrontal cortex or medial striatum impair multiple-location place learning in the water task: evidence for neural structures with complementary roles in behavioural flexibility, Exp. Brain Res. 187 (2008) 419–427. M. Méndez, M. Méndez-López, L. López, M.A. Aller, J. Arias, J.L. Arias, Basal and learning task-related brain oxidative metabolism in cirrhotic rats, Brain Res. Bull. 78 (2009) 195–201. M. Méndez-López, M. Méndez, L. López, J.L. Arias, Spatial working memory in Wistar rats: brain sex differences in metabolic activity, Brain Res. Bull. 79 (2009) 187–192. M. Méndez-López, M. Méndez, L. López, J.M. Cimadevilla, J.L. Arias, Hippocampal heterogeneity in spatial memory revealed by cytochrome oxidase, Neurosci. Lett. 452 (2009) 162–166. R.G. Morris, Developments of a water-maze procedure for studying spatial learning in the rat, J. Neurosci. Methods 11 (1984) 47–60. M.B. Moser, E.I. Moser, E. Forrest, P. Andersen, R.G. Morris, Spatial learning with a minislab in the dorsal hippocampus, Proc. Natl. Acad. Sci. U.S.A. 92 (1995) 9697–9701. G. Paxinos, C. Watson (Eds.), The Rat Brain in Stereotaxic Coordinates—The New Coronal Set, 5th ed., Elsevier, London, 2005, 206 pp. M.E. Ragozzino, S. Adams, R.P. Kesner, Differential involvement of the dorsal anterior cingulate and prelimbic-infralimbic areas of the rodent prefrontal cortex in spatial working memory, Behav. Neurosci. 112 (1998) 293– 303. V. Ramírez-Amaya, I. Balderas, J. Sandoval, M.L. Escobar, F. Bermúdez-Rattoni, Spatial long-term memory is related to mossy fiber synaptogenesis, J. Neurosci. 21 (2001) 7340–7348. L.J. Santín, S. Rubio, A. Begega, J.L. Arias, Effects of mammillary body lesions on spatial reference and working memory tasks, Behav. Brain Res. 102 (1999) 137–150. H.M. Sisti, A.L. Glass, T.J. Shors, Neurogenesis and the spacing effect: learning over time enhances memory and the survival of new neurons, Learn. Mem. 14 (2007) 368–375. S.D. Vann, J.P. Aggleton, Evidence of a spatial encoding deficit in rats with lesions of the mammillary bodies or mammillothalamic tract, J. Neurosci. 23 (2003) 3506–3514. M. Wolff, S.J. Gibb, J.C. Cassel, J.C. Dalrymple-Alford, Anterior but not intralaminar thalamic nuclei support allocentric spatial memory, Neurobiol. Learn. Mem. 90 (2008) 71–80. M.T.T. Wong-Riley, Cytochrome oxidase: an endogenous metabolic marker for neuronal activity, Trends Neurosci. 12 (1989) 94–101. S.J. Zhou, M.E. Zhu, D. Shu, X.P. Du, X.H. Song, X.T. Wang, R.Y. Zheng, X.H. Cai, J.F. Chen, J.C. He, Preferential enhancement of working memory in mice lacking adenosine A(DA) receptors, Brain Res. 1303 (2009) 74–83.
Contents lists available at ScienceDirect
Neuroscience Letters journal homepage: www.elsevier.com/locate/neulet
Spatial short-term memory in rats: Effects of learning trials on metabolic activity of limbic structures Magdalena Méndez-López ∗ , Marta Méndez, Azucena Begega, Jorge L. Arias Laboratorio de Neurociencias, Departamento de Psicología, Universidad de Oviedo, Plaza Feijoo s/n, 33003 Oviedo, Spain
a r t i c l e
i n f o
Article history: Received 16 June 2010 Received in revised form 21 July 2010 Accepted 21 July 2010 Keywords: Matching-to-position task Water maze Cytochrome oxidase Hippocampus Mammillary bodies Anterior thalamic nuclei
a b s t r a c t The oxidative metabolism was assessed in the hippocampus and related regions in rats that were trained in a spatial short-term memory task in the water maze following distinct training schedules. The cytochrome oxidase (COx) histochemistry was evaluated in groups of rats that received a daily session made up of either two or three learning trials. An untreated group was added to determine baseline levels of COx. We found that the 3-Trials group exhibited a better performance concerning the differences in latency between trials 1 and 2. Trained groups showed higher COx activity than the untreated group in the medial prefrontal cortex, dentate gyrus, CA1, and the mammillary region. However, a decrease in COx activity was found in the dentate gyrus, CA1, and supramammillary region of the 3-Trials group. In addition, COx activity levels found in this group were similar to those of the untreated group in some thalamic nuclei. Most of the regions that presented significant correlations between COx activity and behavioral scores were found in the 3-Trials group. These findings suggest an influence of task difficulty in the oxidative metabolism of brain regions involved in spatial learning. © 2010 Elsevier Ireland Ltd. All rights reserved.
The water maze is a learning paradigm for assessment of spatial memory in rodents [18] in which animals use visible cues from the environment to find a hidden platform. Within this context, the evaluation protocol for a short-term memory task involves the temporal storage of a spatial location to use within a training session, but not between sessions [5,16,17]. Some studies have shown that differences in the training schedule can influence learning of the water maze task [2,22,24,28]. For example, an improvement in spatial memory was found by separating trial blocks [2,24], reducing the retention interval [2,28], and increasing the number of training sessions [2,22]. Furthermore, some of these studies have shown changes in the number and morphology of hippocampal neurons as a consequence of modification in training variables [2,24]. Studies examining the neural substrates of memory have found that spatial learning is related to the functioning of a neuronal circuit composed of cortical, hippocampal, and diencephalic brain regions [1]. Hence, learning impairment in the water maze has been found after lesions of the dorsal hippocampus [19], medial prefrontal cortex [14], anterior thalamus [27] and mammillary region [23,25]. The cytochrome oxidase (COx) is a mitochondrial enzyme involved in the oxidative phosphorylation process in which ATP
∗ Corresponding author. Tel.: +34 985103212; fax: +34 985104144. E-mail addresses: [email protected], [email protected] (M. Méndez-López). 0304-3940/$ – see front matter © 2010 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.neulet.2010.07.054
is generated and this metabolic energy is necessary to sustain neuronal functions [27]. Histochemical labelling of COx [7] has been extensively used in studies to evaluate the metabolic activity of limbic structures associated with spatial training in the water maze [3,15–17]. Interestingly, it has been shown that this technique can detect significant changes in metabolic activity over training days [3]. In the present study, we examined the effect of changes in the training schedule on brain metabolic activity of subjects trained in a matching-to-position task developed for the assessment of shortterm spatial memory in the water maze. The COx activity of the dorsal hippocampus, medial prefrontal cortex, anterior thalamic nuclei, and mammillary body region was examined by quantitative COx histochemistry in rats submitted to a daily session made up of either two or three learning trials. In addition, correlations between COx activity and behavioral scores were also calculated. Twenty-one male Wistar rats (240–270 g; vivarium of the University) maintained under standard conditions (21 ± 1 ◦ C temperature, 65–70% relative humidity and 12 h light–dark cycle) with ad libitum access to food and water, were used. The experimental procedures were carried out according to the National legislation (Royal Decree no. 1201/2005), which is in compliance with the European Community Council Directive 86/609/EEC for the use and care of laboratory animals. Fourteen rats were trained in the water maze [18] and the remaining rats were used as a control group (CO, n = 7) to determine basal COx activity. These rats remained in their home cage. The maze consisted of a cylindrical tank (150 cm diameter, 40 cm
M. Méndez-López et al. / Neuroscience Letters 483 (2010) 32–35
deep) of water at 22 ± 1 ◦ C and a hidden platform. The pool was surrounded by panels on which the spatial clues were placed (a square, a horizontal line and a vertical line) and was divided into four quadrants to locate the platform and the start positions. Animals were trained on a matching-to-position task that was developed in three consecutive days (one session daily) and received a habituation session one day prior to the initial test [16,17]. Each daily session involved two identical trials for the “2-Trials group” (n = 7) and three identical trials for the “3-Trials group” (n = 7). The trials consisted of releasing the animal from one starting point and letting it swim until it reached the platform or 60 s had elapsed. If the animal had not reached the platform in this time, it was placed on the platform and kept there for 15 s. The intertrial interval was 5 s. The task demands a recall of the position occupied by the platform during the first trial. The locations of the start positions and of the hidden platform were the same during one session but varied on the different days in a pseudorandom order. The escape latencies were recorded. The mean escape latency was calculated statistically for each animal in the trials 1, 2 and 3 (in the case of the 3-Trials group), grouping all sessions carried out. Also, the difference in escape latency scores between trials 1 and 2 of the last session was used as performance index. This difference was defined as saving in escape latency to platform. Ninety minutes after the end of the behavioral task, the animals were decapitated. The CO group was decapitated at the same time as the trained groups. Brains were removed intact, frozen in isopentane (Sigma–Aldrich, Spain) and stored at −40 ◦ C. Coronal sections (30 m) of the brain were processed with quantitative COx histochemistry [7]. Briefly, sets of tissue homogenate standards from Wistar rat brain were cut at different thickness (10, 30, 40 and 60 m) and mounted on a slide. Previously, COx activity of the homogenate was spectrophotometrically assessed [8]. Sets of tissue homogenate standards of known COx activity were used as calibration standards in each staining batch. Sections were fixed with 0.5% (v/v) glutaraldehyde and 10% (w/v) sucrose in phosphate buffer (pH 7.6; 0.1 M). After this, the sections were rinsed in 0.1 M phosphate buffer with 10% (w/v) sucrose. Then 0.05 M Tris buffer, pH 7.6, with 275 mg/l cobalt chloride, 10% (w/v) sucrose, and 0.5% (v/v) dimethylsulfoxide, were applied. Subsequently, they were rinsed and incubated in a solution of 0.06 g cytochrome c (Sigma–Aldrich), 0.016 g catalase, 40 g sucrose, 2 ml dimethylsulfoxide and 0.4 g diaminobenzidine tetrahydrochloride in 800 ml of phosphate buffer (pH 7.6; 0.1 M). Finally, the tissue was fixed in 4% (v/v) buffered formalin with 10% (w/v) sucrose, dehydrated and coverslipped. Quantification of COx histochemical staining intensity was done by densitometric analysis using a computer-controlled microscope (Leica Q550IW) with a light source (Leica DM-RHC) connected to a camera CCD (Cohu, Japan) and an image analysis software (Leica Q-Win). The mean optical density (OD) of each structure was measured on the right side of bilateral structures using three consecutive sections in each animal [16]. In each section, four, nonoverlapping readings were taken using a square-shaped sampling window that was adjusted for each region size. A total of 12 measurements were taken in each region per animal and averaged to obtain a mean. To control staining variations between different staining baths, measurements were taken from stained brain homogenate standards. Regression curves between section thickness and known COx activity measured in each set of standards were calculated for each incubation bath [8]. OD values were then converted to COx activity units (mol cytochrome c oxidized/min/g wet tissue at 23 ◦ C) using the regression curves calculated. The regions of interest were the infralimbic cortex (IL) and prelimbic cortex (PL); the anterodorsal thalamic nucleus (ADT), anteroventral thalamic nucleus (AVT) and anteromedial thalamic nucleus (AMT); the dentate gyrus (DG), CA1 and CA3 subfields
33
Fig. 1. (A) Escape latencies to platform (mean ± S.E.M). Significance of differences between trials, grouping all sessions carried out, *p ≤ 0.001. (B) Difference in latency scores between trials (mean ± S.E.M) within the third session. Significant difference between groups in the 1–2 saving measure, *p = 0.005. Significantly different with respect to the 1–3 saving measure of the 3-Trials group, # p = 0.02.
of the hippocampus; and the medial mammillary nucleus (MM), lateral mammillary nucleus (LM) and supramammillary nucleus (SuM). The stained coronal sections corresponded to Bregma levels [20]: +3.24 mm for IL and PL; −1.92 mm for ADT, AVT and AMT; −3.84 mm for DG, CA1 and CA3; and −4.44 mm for MM, LM and SuM. A paired-sample t-test was used to describe differences between trials in escape latencies, grouping all training sessions (Fig. 1A). A reduction of escape latencies was showed for both trained groups in trial 2 (t6 ≥ 6.81, p ≤ 0.001). In addition, shorter escape latencies were found during trial 3 in the 3-Trials group (t6 = 9.08, p ≤ 0.001). Also, saving from trial 1 to trial 2 within the last session was compared between groups using a t-test. As shown in Fig. 1B, this saving was lower in the 2-Trials group (t12 = −3.47, p = 0.005). Similarly, this saving was significantly lower than the trials 1–3 saving of the 3-Trials group within the last session (t12 = −2.639, p = 0.02; Fig. 1B). The results of the COx activity measures obtained from the different brain regions studied are illustrated in Table 1. COx activity of the studied groups was compared using a one-way analysis of variance and Tukey’s test was applied as post hoc test. COx values taken in the IL, CA3 hippocampal subfield, and AMT were similar between groups (F2,18 ≤ 2.52, p ≥ 0.10). However, differences between groups were found in the PL, MM and LM (F2,18 ≥ 6.83, p < 0.001), where an increase in COx activity was found in the trained groups (p < 0.05). Also, an increase in COx activity was found in the trained groups in the DG, CA1 and SuM (F2,18 ≥ 20.79,
34
M. Méndez-López et al. / Neuroscience Letters 483 (2010) 32–35
Table 1 Cytochrome oxidase activity of the selected brain regions. CO PL IL DG CA3 CA1 ADT AVT AMT MM LM SuM
17.22 16.19 18.15 11.49 12.18 31.46 23.99 16.10 20.30 23.67 9.36
2-Trials ± ± ± ± ± ± ± ± ± ± ±
0.22 0.36 0.58 0.50 0.20 0.90 1.14 1.56 0.66 0.94 0.59
19.29 18.40 24.53 13.45 19.77 36.44 30.64 18.98 29.81 28.10 14.15
± ± ± ± ± ± ± ± ± ± ±
3-Trials 0.46* 0.77 0.72** 1.30 0.59** 1.07** 0.45** 0.70 1.71** 1.56* 0.61**
20.09 17.95 21.11 11.71 15.75 30.92 24.73 17.02 27.48 29.86 11.58
± ± ± ± ± ± ± ± ± ± ±
0.77** 0.94 0.50**,† 0.54 0.45**,† 0.30† 1.40† 0.73 1.63** 1.06** 0.32*,†
Mean ± S.E.M. of COx activity units (mol cytochrome c oxidized/min/g wet tissue at 23 ◦ C) measured in the prelimbic cortex (PL), infralimbic cortex (IL), dentate gyrus (DG), CA1 and CA3 subregions, anterodorsal thalamus (ADT), anteroventral thalamus (AVT), anteromedial thalamus (AMT), medial mammillary (MM), lateral mammillary (LM) and supramammillary (SuM). Tukey Test, p value in relation to: CO (control group) (*p < 0.05; **p < 0.01) and 2-Trials group († p < 0.01).
p < 0.001), but a significant decrease was shown in the 3-Trials group as compared to the 2-Trials group (p ≤ 0.02). Finally, differences among groups were found in the ADT and AVT (F2,18 ≥ 11.40, p < 0.001), where the 2-Trials group showed higher COx activity than all remaining groups (p ≤ 0.003). Pearson product-moment correlations between COx activity and behavioral data were calculated for all measured regions in each trained group, using as performance index the difference in escape latency between trials 1 and 2 of the last session (trials that were common to both trained groups). As shown in Fig. 2, significant negative correlations were detected in some measured regions. These significant correlations were more frequent in the 3-Trials group. This study shows that the neuronal metabolic activity was lower in some limbic system regions when animals were trained using one more daily trial. Also, these rats presented better performance shown by the saving increment in escape latency. Therefore, this
behavioral accuracy could suggest that the insertion of a single and additional trial in the training schedule reduces task difficulty because the animal had more chances to learn the spatial location. Supporting this, it has been shown that an increase of the number of training trials improved the retention capacity of long-term memory for spatial locations in the water maze [2,22]. In relation to the neuronal metabolism, an increase in COx activity was found in the hippocampus, prelimbic cortex, anterior thalamus and mammillary nuclei of the trained groups. In this regard, COx activity levels in these brain regions have previously been reported to increase in rats submitted to the same short-term memory task with two daily trials [15]. Furthermore, the contribution of these regions has been related to the spatial memory process in several studies [9,12,19,21,23,25,26]. Taking into account COx activity levels displayed by each trained group, the reduction shown by the 3-Trials group in the supramammillary nucleus, anterodorsal and anteroventral thalamic nuclei, and the dentate gyrus, and CA1 is of outstanding interest. Since this group received more trials, this reduction might be explained by changes in the contribution of these regions to the task. In relation to this, a time-dependent involvement of brain regions associated with spatial memory was described using the COx histochemistry [3]. In addition, as mentioned above, the insertion of a third learning trial led to a behavioral improvement. Hence, this could indicate that the task is easier to perform. Addressing the relationship between the difficulty of a learning task and brain COx activity, Hu et al. [10] have reported a functional involvement of the dentate gyrus and CA field of rats trained in a difficult visual discrimination task, whereas only the dentate gyrus was involved in the easy version of the same task. Therefore, collectively, these results are consistent with the hypothesis that brain activity is directly related to the difficulty of task performance. In support of these observations, several human studies have revealed greater brain activation as the degree of difficulty of cognitive tasks increased by apply-
Fig. 2. Correlations between COx activity and saving in escape latency (saving from trial 1 to trial 2 within the last session). Significant negative correlations were found in the dentate gyrus (DG), CA1 subfield of the hippocampus, supramammillary (SuM), anterodorsal thalamus (ADT), anteroventral thalamus (AVT) and lateral mammillary (LM).
M. Méndez-López et al. / Neuroscience Letters 483 (2010) 32–35
ing neuroimaging techniques that are also based on the metabolic capacity of the brain [6,11,13]. The neurobehavioral correlations have shown that the mastery of the spatial short-term memory task was most related to neuronal COx activity in the rats trained under the three trials condition. Indeed, most of the regions that presented negative correlations between brain activity and the saving in escape latency were those which showed a significant lower COx activity in this condition. Therefore, this suggests that training conditions were the key factor in modulating brain activity. Finally, the prelimbic cortex, medial mammillary and lateral mammillary were the only nuclei explored in which COx activity increased following learning but did not significantly differ between training conditions. This suggests that mammillary nuclei and prelimbic cortex have an essential involvement in this task that is independent of task difficulty. The contribution of these regions to spatial learning has been proposed previously. Thus, the prelimbic cortex has been related to some cognitive processes such as attention and behavioral flexibility required when the subject must learn a different daily location in the water maze [4,14]. In addition, the lesion of the mammillary nuclei impaired delayed-matchingto-position tasks in the water maze [23,25]. Furthermore, these experiments have shown that this region is involved in the acquisition of spatial information, and its role is independent of task difficulty and is not affected by the magnitude of the retention interval [23,25]. Taken together, the results of the present study suggest that COx histochemistry is sensitive to detect changes in the oxidative metabolism of the hippocampus and related regions involved in spatial short-term memory in response to variations in the training schedule that improve task performance.
[7]
[8]
[9]
[10]
[11] [12]
[13]
[14]
[15]
[16]
[17]
[18] [19]
[20]
Acknowledgements [21]
˜ Díaz for their assistance We thank Piedad Burgos and Begona and V. Navascués for reviewing the English text of this manuscript. This research was supported by grant SEJ2007-63506 and PSI201019348 (Spanish Ministry of Science and Innovation and FEDER).
[22]
[23]
References [1] J.P. Aggleton, M.W. Brown, Episodic memory, amnesia, and the hippocampal anterior thalamic axis, Behav. Brain Sci. 22 (1999) 425–489. [2] K. Bolding, J.W. Rudy, Place learning in the Morris water task. Making the memory stick, Learn. Mem. 13 (2006) 278–286. [3] N.M. Conejo, H. González-Pardo, F. Gonzalez-Lima, J.L. Arias, Spatial learning of the water maze: progression of brain circuits mapped with cytochrome oxidase histochemistry, Neurobiol. Learn. Mem. 93 (2010) 362–371. [4] B. Delatour, P. Gisquet-Verrier, Functional role of rat prelimbic-infralimbic cortices in spatial memory: evidence for their involvement in attention and behavioural flexibility, Behav. Brain Res. 109 (2000) 113–128. [5] P.A. Dudchenko, An overview of the tasks used to test working memory in rodents, Neurosci. Biobehav. Rev. 28 (2004) 699–709. [6] D.C. Glahn, J. Kim, M.S. Gohen, V.P. Poutanen, S. Therman, S. Bava, T.G.M. Van Erp, M. Manninen, M. Huttunen, J. Lönnqvist, C.G. StandertskjöldNordenstam, T.D. Cannon, Maintenance and manipulation in spatial working
[24]
[25]
[26]
[27] [28]
35
memory: dissociations in the prefrontal cortex, NeuroImage 17 (2002) 201–213. F. Gonzalez-Lima, A. Cada, Cytochrome oxidase activity in the auditory system of the mouse: a qualitative and quantitative histochemical study, Neuroscience 63 (1994) 559–578. F. Gonzalez-Lima, D. Jones, Quantitative mapping of cytochrome oxidase activity in the central auditory system of the gerbil: a study with calibrated activity standards and metal-intensified histochemistry, Brain Res. 660 (1994) 34–49. V. Hok, E. Save, P.P. Lenck-Santini, B. Poucet, Coding for spatial goals in the prelimbic/infralimbic area of the frontal cortex, Proc. Natl. Acad. Sci. U.S.A. 102 (2005) 4602–4607. D. Hu, X. Xu, F. Gonzalez-Lima, Hippocampal cytochrome oxidase activity of rats in easy and difficult visual discrimination learning, Int. J. Neurosci. 115 (2005) 595–611. J.M. Jansma, N.F. Ramsey, R. Coppola, R.S. Kahn, Specific versus nonspecific brain activity in a parametric n-back task, NeuroImage 12 (2000) 688–697. B. Kocsis, R.P. Vertes, Characterization of neurons of the supramammillary nucleus and mammillary body that discharge rhythmically with the hippocampal theta rhythm in the rat, J. Neurosci. 14 (1994) 7040–7052. E.D. Leshikar, A.H. Gutchess, A.C. Hebrank, B.P. Sutton, The impact of increased relational encoding demands on frontal and hippocampal function in older adults, Cortex 46 (2010) 507–521. R.J. McDonald, A.L. King, N. Foong, Z. Rizos, N.S. Hong, Neurotoxic lesions of the medial prefrontal cortex or medial striatum impair multiple-location place learning in the water task: evidence for neural structures with complementary roles in behavioural flexibility, Exp. Brain Res. 187 (2008) 419–427. M. Méndez, M. Méndez-López, L. López, M.A. Aller, J. Arias, J.L. Arias, Basal and learning task-related brain oxidative metabolism in cirrhotic rats, Brain Res. Bull. 78 (2009) 195–201. M. Méndez-López, M. Méndez, L. López, J.L. Arias, Spatial working memory in Wistar rats: brain sex differences in metabolic activity, Brain Res. Bull. 79 (2009) 187–192. M. Méndez-López, M. Méndez, L. López, J.M. Cimadevilla, J.L. Arias, Hippocampal heterogeneity in spatial memory revealed by cytochrome oxidase, Neurosci. Lett. 452 (2009) 162–166. R.G. Morris, Developments of a water-maze procedure for studying spatial learning in the rat, J. Neurosci. Methods 11 (1984) 47–60. M.B. Moser, E.I. Moser, E. Forrest, P. Andersen, R.G. Morris, Spatial learning with a minislab in the dorsal hippocampus, Proc. Natl. Acad. Sci. U.S.A. 92 (1995) 9697–9701. G. Paxinos, C. Watson (Eds.), The Rat Brain in Stereotaxic Coordinates—The New Coronal Set, 5th ed., Elsevier, London, 2005, 206 pp. M.E. Ragozzino, S. Adams, R.P. Kesner, Differential involvement of the dorsal anterior cingulate and prelimbic-infralimbic areas of the rodent prefrontal cortex in spatial working memory, Behav. Neurosci. 112 (1998) 293– 303. V. Ramírez-Amaya, I. Balderas, J. Sandoval, M.L. Escobar, F. Bermúdez-Rattoni, Spatial long-term memory is related to mossy fiber synaptogenesis, J. Neurosci. 21 (2001) 7340–7348. L.J. Santín, S. Rubio, A. Begega, J.L. Arias, Effects of mammillary body lesions on spatial reference and working memory tasks, Behav. Brain Res. 102 (1999) 137–150. H.M. Sisti, A.L. Glass, T.J. Shors, Neurogenesis and the spacing effect: learning over time enhances memory and the survival of new neurons, Learn. Mem. 14 (2007) 368–375. S.D. Vann, J.P. Aggleton, Evidence of a spatial encoding deficit in rats with lesions of the mammillary bodies or mammillothalamic tract, J. Neurosci. 23 (2003) 3506–3514. M. Wolff, S.J. Gibb, J.C. Cassel, J.C. Dalrymple-Alford, Anterior but not intralaminar thalamic nuclei support allocentric spatial memory, Neurobiol. Learn. Mem. 90 (2008) 71–80. M.T.T. Wong-Riley, Cytochrome oxidase: an endogenous metabolic marker for neuronal activity, Trends Neurosci. 12 (1989) 94–101. S.J. Zhou, M.E. Zhu, D. Shu, X.P. Du, X.H. Song, X.T. Wang, R.Y. Zheng, X.H. Cai, J.F. Chen, J.C. He, Preferential enhancement of working memory in mice lacking adenosine A(DA) receptors, Brain Res. 1303 (2009) 74–83.