- Email: [email protected]
Relationship between learning, stress and hippocampal brain-derived neurotrophic factor
Neuroscience 121 (2003) 825– 828
LETTER TO NEUROSCIENCE RELATIONSHIP BETWEEN LEARNING, STRESS AND HIPPOCAMPAL BRAIN-DERIVED NEUROTROPHIC FACTOR S. SCACCIANOCE,* P. DEL BIANCO, A. CARICASOLE, F. NICOLETTI AND A. CATALANI
bilization stress (Smith et al., 1995). The adrenal hormone corticosterone might be involved in this suppression, because exogenous administration of the hormone is known to reduce BDNF mRNA (Schaaf et al., 1998). On these bases, BDNF has been suggested to be involved in stressinduced hippocampal adaptation as well as in the pathogenesis of depression (Duman et al., 1997). Interestingly, hippocampal BDNF expression is up-regulated by learning, although the majority of learning tasks incorporate a substantial degree of stress and are associated with elevated plasma corticosterone levels. An example is provided by the Morris water maze, in which hippocampal BDNF expression is increased despite of high plasma corticosterone levels that are expected to reduce BDNF mRNA (Schaaf et al., 1999). Hence, it becomes important to precisely dissect how learning and stress contribute to the overall changes in hippocampal BDNF levels during the execution of a specific behavioral task. To address this issue, we have set up an experimental model in which two groups of rats received the same amount of stress, but only one group could learn how to avoid it.
Department of Human Physiology and Pharmacology Vittorio Erspamer, University of Rome “La Sapienza,” Piazzale Aldo Moro, 5 00185 Rome, Italy
Abstract—Brain-derived neurotrophic factor (BDNF) expression in the hippocampus is reduced in response to acute, as well as repeated immobilization stress. This effect might be mediated by corticosterone, because corticosterone administration is known to reduce hippocampal BDNF. However, rats subjected to a learning paradigm showed an increased BDNF expression in the hippocampus despite the high corticosterone levels found during the test. To dissect the relative contributions of learning and stress to the overall changes in BDNF levels we set up an experimental model in which two groups of rats received the same amount of stress, but only one group had the possibility to learn how to avoid it. Using this model, we now report that learning and stress exert an opposite modulation on BDNF levels in the hippocampus, and that the increasing effect of learning predominates over the decreasing effect of stress. © 2003 IBRO. Published by Elsevier Ltd. All rights reserved. Key words: behavior, corticosterone, ELISA, hippocampus.
EXPERIMENTAL PROCEDURES Experimental animals
Brain-derived neurotrophic factor (BDNF), a member of the neurotrophin (NT) family (Hyman et al., 1991), is expressed by specific neuronal populations of the CNS (Hofer et al., 1990). Expression of BDNF and its receptor, tyrosine kinase B (trkB), is particularly high in the hippocampus (Hofer et al., 1990; Merlio et al., 1992), a region that is critically involved in learning and memory processes. There is evidence of learning-associated changes in hippocampal BDNF expression. In situ hybridization studies demonstrate a rapid and selective induction of BDNF mRNA in the hippocampus during contextual learning (Hall et al., 2000). Spatial learning is also associated with an increase in BDNF mRNA levels in the hippocampus (Mizuno et al., 2000), combined with an activation of trkB as shown by increased levels of phosphorylated trkB (Mizuno et al., 2003). Hippocampal BDNF expression is reduced after the application of single or repeated immo-
Adult (2.5 months) male Wistar rats (Harlan, Udine, Italy) weighing 275–300 g were housed two per cage in a temperature-controlled room (24⫾2 °C), with a 12-h light/dark period (lights on: 07:00 – 19:00 h) for at least 2 weeks before the experiments. Food (Standard Diet 4RF21; Charles River, Calco, Italy) and tap water were provided ad libitum. No female rats were present in the animal room. The week before the experiment, the rats were daily handled gently by the same operator between 09:30 and 10:30 h to minimize stress response to manipulation. All experiments were carried out in accordance to the European Communities Council Directive 86/609/EEC and to the Italian national rules laid down by the Ministry of Health. All efforts were made to minimize animal suffering during the experiments.
Experimental model The experimental model consisted of two 2-way shuttle boxes (U. Basile, Varese, Italy). One cage, the “escape cage,” was divided into two equally sized compartments (24⫻21 cm, 21 cm high) and was equipped with an oscillating grid floor which allowed the termination of the electric stimulus (0.4 mA) when the rat moved from one compartment to the other through a 7⫻7 cm gate. There was no cue present before or during the delivery of the stimulus. A second cage, the “yoked cage” (48⫻21 cm, 21 cm high) was not divided, had no oscillating floor, and was electrically connected to the escape cage in order to deliver an electric stimulus (0.4 mA) to
*Corresponding author. Tel: ⫹39-06-4991-2511; fax: ⫹39-06-49912511. E-mail address: [email protected] (S. Scaccianoce). Abbreviations: ANOVA, analysis of variance; BDNF, brain-derived neurotrophic factor; GL, good learner; LPT, long-term potentiation; NT, neurotrophin; PL, poor learner; trkB, tyrosine kinase B; YGL, animals yoked to good learners; YPL, animals yoked to poor learners.
0306-4522/03$30.00⫹0.00 © 2003 IBRO. Published by Elsevier Ltd. All rights reserved. doi:10.1016/S0306-4522(03)00514-1
825
826
S. Scaccianoce et al. / Neuroscience 121 (2003) 825– 828
the grid floor for the same duration and rhythmicity as the escape cage during the sessions.
Behavioral studies Rats were subjected to a training consisting of three sessions of 30 trials for 3 consecutive days; in each trial, an electric shock was delivered (maximal duration 4 s) followed by an interval of 16 s. Paired sets of rats were placed in the escape and yoked cages, and subjected to simultaneous sessions. In each session, the number of escapes, the latencies to the escapes, and the number of failures were recorded. Upon termination of the last trial of the third session, each pair of rats was immediately killed by decapitation, the blood was collected, and the hippocampus was dissected and frozen. Rats trained in the escape cage were arbitrarily divided into two groups depending on the learning ability. “Good learners” (GL) had a number of escapes in the third session greater than, or equal to, the mean value calculated from all rats in the same session. “Poor learners” (PL) were defined as those having a number of escapes lower than the mean value. Rats located in the “yoked cage” were also divided into two groups depending on the ability of the partner: yoked to GL and yoked to PL. The control group consisted of undisturbed rats removed directly from their home cages.
Hippocampal BDNF determination Hippocampi were processed to measure BDNF protein content by ELISA using commercially available kits (BDNF Emax Immunoassay System; Promega, Milan, Italy). Briefly, tissues were sonicated in a lysis buffer (137 mM NaCl, 20 mM Tris, 1% Nonidet P-40, 10% glycerol, 1 mM phenylmethylsulfonyl fluoride, 10 g/ml aprotinin, 1 g/ml leupeptin, 0.5 mM sodium orthovanadate, 5 l/mg tissue) and centrifuged at 12,000⫻g at 4 °C for 30 min. The antibodies used have negligible crossreactivity (⬍3%) with other growth factors including nerve growth factor, NT-3 and NT-4/5. Colorimetric detection of peroxidase activity was achieved by adding TMB solution and peroxidase substrate. The enzymatic reaction was stopped with HCl (1 M) and the optical density of each well was measured at 450 nM using an Emax plate reader.
Plasma corticosterone determination Blood was collected into tubes containing 0.13 M EDTA. After centrifugation at 1,900⫻g at 4 °C for 15 min plasma samples were separated and stored at ⫺20 °C prior to assay. Plasma corticosterone concentrations were determined by RIA (ICN Biomedicals, Milan, Italy). The cross-reactivity of the polyclonal corticosterone–antisera with related substances was negligible. The interand intra-assay coefficients of variation were 8% and 3%, respectively, with a detection limit of 0.0125 ng/tube. Plasma samples were diluted 1:100 with Steroid Diluent (ICN Biomedicals). All measurements were in the linear range of the standard curve (0.0125–3.0 ng/tube).
Statistical analysis Behavioral data were analyzed with Student’s t-test. Hippocampal BDNF and plasma corticosterone data were analyzed with two-way (Good/Poor; Operant/Yoked) analysis of variance (ANOVA) followed by Duncan’s multiple comparison test. Differences between control and experimental groups were analyzed with ANOVA followed by Dunnett’s multiple comparisons test.
Fig. 1. Number of escapes during the last session in GL and in PL rats (A). Cumulative duration of shocks received during the last session by GL and PL rats (B). Data are expressed as means⫾S.E.M., n⫽6. *** P⬍0.001 (Student’s t-test).
RESULTS Behavioral profile During the third session, GL had a number of escapes higher than PL (GL 21.2⫾1.4; PL 2.8⫾1.2; means⫾S.E.M., n⫽6, t⫽9.9, P⬍0.001, Student’s t-test; Fig. 1A). Hence, they, and their paired-yoked rats, received a shorter duration of shocks as compared with PL and their yoked partners (GL 68.1⫾6.7; PL 110.5⫾3.8; means⫾S.E.M., n⫽6, t⫽5.5, P⬍0.001, Student’s t-test; Fig. 1B). Hippocampal BDNF concentration in GL and PL and respective yoked partners Hippocampal BDNF concentrations were determined immediately after the completion of the last trial of the third session (Fig. 2). Two-way ANOVA indicated a significant difference between GL and PL (F⫽19.3, P⬍0.001) and between operant and yoked rats (F⫽25.4, P⬍0.001) and no significant interaction between these two factors (F⫽1.16, P⬎0.05). GL showed higher hippocampal BDNF
S. Scaccianoce et al. / Neuroscience 121 (2003) 825– 828
827
trations were significantly (P⬍0.001, Dunnett’s test) elevated in all animals subjected to the trials (controls: 1.4⫾0.4; GL: 21.1⫾1.1; YGL: 21.3⫾1.9; PL: 20.0⫾1.3; YPL: 28.4⫾1.0, micrograms of corticosterone/100 ml of plasma; means⫾S.E.M., n⫽6). Remarkably, corticosterone levels in the YPL rats were higher than those observed in the other experimental groups (P⬍0.01 versus GL; P⬍0.01 versus YGL; and P⬍0.01 versus PL, Duncan’s test).
DISCUSSION
Fig. 2. Hippocampal BDNF concentrations in GL and in YGL rats, and in PL and in YPL rats. Rats were killed at the end of the last trial of the third session. Values (mean⫾S.E.M., n⫽6) are expressed as percent of variation with respect to the control (CONT) group (BDNF concentration: 3215.5⫾205.4 pg/g of tissue, n⫽6). # P⬍0.05, ## P⬍0.01 versus CONT (Dunnett’s test). * P⬍0.05, ** P⬍0.01 (Duncan’s test).
levels as compared with the corresponding yoked animals (YGL; P⬍0.05, Duncan’s test) and to control animals (P⬍0.05, Dunnett’s test). BDNF levels did not differ between YGL and control animals (P⬎0.05, Dunnett’s test). PL showed no changes in hippocampal BDNF levels as compared with control animals (P⬎0.05, Dunnett’s test). However, animals yoked to PL (YPL) showed a significant decrease in hippocampal BDNF levels (P⬍0.01 versus control, Dunnett’s test; P⬍0.01 versus PL, Duncan’s test). Plasma corticosterone concentration in GL and PL and respective yoked partners Plasma corticosterone levels were measured in nonshocked controls and in the experimental rats at the end of the last trial (Fig. 3). Two-way ANOVA indicated a significant difference between GL and PL (F⫽5.1, P⬍0.05) and between operant and yoked rats (F⫽10.2, P⬍0.01) and a significant interaction between these two factors (F⫽9.2, P⬍0.01). Compared with controls, corticosterone concen-
Fig. 3. Plasma corticosterone concentrations in control (CONT), GL, YGL, PL and in YPL rats. Rats were killed at the end of the last trial of the third session. Mean⫾S.E.M., n⫽6. ### P⬍0.001 versus CONT (Dunnett’s test). ** P⬍0.01 versus GL, YGL and PL (Duncan’s test).
These data show that the same amount of a stressful stimulus (i.e. an electric footshock) produces different effects on hippocampal BDNF and plasma corticosterone levels: the variable that governs the outcome is the chance to learn. Animals that performed well in the learning paradigm showed an increase in hippocampal BDNF levels despite of the high levels of circulating corticosterone. On the other side, YPL (which received a greater amount of electrical stimulation than YGL) showed a reduction in hippocampal BDNF protein associated with the highest plasma levels of corticosterone. Our finding of the learningcorrelated BDNF increase is in agreement with previous studies which, on the basis of an induction of hippocampal BDNF mRNA, have hypothesized the implication of this NT in learning and in memory formation (Schaaf et al., 1999; Hall et al., 2000). Studies carried out in cultured hippocampal neurons show that BDNF is rapidly secreted during induction of long-term potentiation (LPT), an established electrophysiological substrate for associative learning (Gartner and Staiger, 2002). In addition, application of BDNF induces LTP at the medial perforant path-granule cell synapses in living animals (Ying et al., 2002) and evokes Ca2⫹ transients in dentate granule cells of mouse hippocampal slices (Kovalchuk et al., 2002), which are required for the induction of a non-decremental form of LTP. There are indications of a close relationship between learning, BDNF and neurogenesis at hippocampal level. It has been postulated a role for BDNF as a signal that controls neurogenesis (Cameron et al., 1998) and learning is found to be associated with neurogenesis (see Grassi Zucconi and Giuditta, 2002 for a review). However, an impaired neurogenesis in the dentate gyrus is also observed in chronically stressed rats (McEwen, 2001). Whether or not hippocampal neurogenesis plays a role in learning under our experimental conditions remains to be established. An additional interesting feature of our study was observed in the PL and in the YPL groups. PL rats had no variations in hippocampal BDNF, which is in line with the observation of Ma et al. (1998). Rats yoked to this group, however, had a reduction in BDNF protein in the hippocampus combined with a greater increase in plasma corticosterone levels as compared with the PL group. Thus, we speculate that YPL rats, which received a greater amount of electrical stimulation because of the poor behavioral performance of the partner, had a more robust adrenocortical activation responsible for the reduction in hippocampal BDNF content. We conclude that a dynamic
828
S. Scaccianoce et al. / Neuroscience 121 (2003) 825– 828
relationship between corticosterone, learning and hippocampal BDNF exists, and that learning may overcome a lowering effect of stress-induced adrenocortical activation on hippocampal BDNF levels. Acknowledgements—We wish to thank Dr. Alessandro Giuliani (Istituto Superiore di Sanita`) for the assistance with the statistical analysis of the data. This research was supported by a grant from Universita` La Sapienza (ricerche di facolta`) to S.S.
REFERENCES Cameron HA, Hazel TG, McKay RD (1998) Regulation of neurogenesis by growth factors and neurotransmitters. J Neurobiol 36:287– 306. Duman RS, Heninger GR, Nestler EJ (1997) A molecular and cellular theory of depression. Arch Gen Psychiatry 54:597–606. Gartner A, Staiger V (2002) Neurotrophin secretion from hippocampal neurons evoked by long-term-potentiation-inducing electrical stimulation patterns. Proc Natl Acad Sci USA 99:6386 –6391. Grassi Zucconi G, Giuditta A (2002) Is it only neurogenesis? Rev Neurosci 13:375–382. Hall J, Thomas KL, Everitt BJ (2000) Rapid and selective induction of BDNF expression in the hippocampus during contextual learning. Nat Neurosci 3:533–535. Hofer M, Pagliusi SR, Hohn A, Leibrock J, Barde YA (1990) Regional distribution of brain-derived neurotrophic factor mRNA in the adult mouse brain. EMBO J 9:2459 –2464. Hyman C, Hofer M, Barde YA, Juhasz M, Yancopoulos GD, Squinto SP, Lindsay RM (1991) BDNF is a neurotrophic factor for dopaminergic neurons of the substantia nigra. Nature 350:230 –232. Kovalchuk Y, Hanse E, Kafitz KW, Konnerth A (2002) Postsynaptic induction of BDNF-mediated long-term potentiation. Science 295: 1729 –1734.
Ma YL, Wang HL, Wu HC, Wei CL, Lee EH (1998) Brain-derived neurotrophic factor antisense oligonucleotide impairs memory retention and inhibits long-term potentiation in rats. Neuroscience 82:957–967. McEwen BS (2001) Plasticity of the hippocampus: adaptation to chronic stress and allostatic load. Ann NY Acad Sci 933:265–277. Merlio JP, Ernfors P, Jaber M, Persson H (1992) Molecular cloning of rat trkC and distribution of cells expressing messengers RNAs for members of the trk family in the rat central nervous system. Neuroscience 51:513–532. Mizuno M, Yamada K, Olariu A, Nawa H, Nabeshima T (2000) Involvement of brain-derived neurotrophic factor in spatial memory formation and maintenance in a radial arm maze test in rats. J Neurosci 20:7116 –121. Mizuno M, Yamada K, Takei N, Tran MH, He J, Nakajima A, Nawa H, Nabeshima T (2003) Phosphatidylinositol 3-kinase: a molecule mediating BDNF-dependent spatial memory formation. Mol Psychiatry 8:217–224. Schaaf MJ, de Jong J, de Kloet ER, Vreugdenhil E (1998) Downregulation of BDNF mRNA and protein in the rat hippocampus by corticosterone. Brain Res 813:112–120. Schaaf MJ, Sibug RM, Duurland R, Fluttert MF, Oitzl MS, De Kloet ER, Vreugdenhil E (1999) Corticosterone effects on BDNF mRNA expression in the rat hippocampus during Morris water maze training. Stress 3:173–183. Smith MA, Makino S, Kvetnansky R, Post RM (1995) Stress and glucocorticoids affect the expression of brain-derived neurotrophic factor and neurotrophin-3 mRNAs in the hippocampus. J Neurosci 15:1768 –1777. Ying SW, Futter M, Rosenblum K, Webber MJ, Hunt SP, Bliss TV, Bramham CR (2002) Brain-derived neurotrophic factor induces long-term potentiation in intact adult hippocampus: requirement for ERK activation coupled to CREB and upregulation of Arc synthesis. J Neurosci 22:1532–1540.
(Accepted 2 July 2003)
LETTER TO NEUROSCIENCE RELATIONSHIP BETWEEN LEARNING, STRESS AND HIPPOCAMPAL BRAIN-DERIVED NEUROTROPHIC FACTOR S. SCACCIANOCE,* P. DEL BIANCO, A. CARICASOLE, F. NICOLETTI AND A. CATALANI
bilization stress (Smith et al., 1995). The adrenal hormone corticosterone might be involved in this suppression, because exogenous administration of the hormone is known to reduce BDNF mRNA (Schaaf et al., 1998). On these bases, BDNF has been suggested to be involved in stressinduced hippocampal adaptation as well as in the pathogenesis of depression (Duman et al., 1997). Interestingly, hippocampal BDNF expression is up-regulated by learning, although the majority of learning tasks incorporate a substantial degree of stress and are associated with elevated plasma corticosterone levels. An example is provided by the Morris water maze, in which hippocampal BDNF expression is increased despite of high plasma corticosterone levels that are expected to reduce BDNF mRNA (Schaaf et al., 1999). Hence, it becomes important to precisely dissect how learning and stress contribute to the overall changes in hippocampal BDNF levels during the execution of a specific behavioral task. To address this issue, we have set up an experimental model in which two groups of rats received the same amount of stress, but only one group could learn how to avoid it.
Department of Human Physiology and Pharmacology Vittorio Erspamer, University of Rome “La Sapienza,” Piazzale Aldo Moro, 5 00185 Rome, Italy
Abstract—Brain-derived neurotrophic factor (BDNF) expression in the hippocampus is reduced in response to acute, as well as repeated immobilization stress. This effect might be mediated by corticosterone, because corticosterone administration is known to reduce hippocampal BDNF. However, rats subjected to a learning paradigm showed an increased BDNF expression in the hippocampus despite the high corticosterone levels found during the test. To dissect the relative contributions of learning and stress to the overall changes in BDNF levels we set up an experimental model in which two groups of rats received the same amount of stress, but only one group had the possibility to learn how to avoid it. Using this model, we now report that learning and stress exert an opposite modulation on BDNF levels in the hippocampus, and that the increasing effect of learning predominates over the decreasing effect of stress. © 2003 IBRO. Published by Elsevier Ltd. All rights reserved. Key words: behavior, corticosterone, ELISA, hippocampus.
EXPERIMENTAL PROCEDURES Experimental animals
Brain-derived neurotrophic factor (BDNF), a member of the neurotrophin (NT) family (Hyman et al., 1991), is expressed by specific neuronal populations of the CNS (Hofer et al., 1990). Expression of BDNF and its receptor, tyrosine kinase B (trkB), is particularly high in the hippocampus (Hofer et al., 1990; Merlio et al., 1992), a region that is critically involved in learning and memory processes. There is evidence of learning-associated changes in hippocampal BDNF expression. In situ hybridization studies demonstrate a rapid and selective induction of BDNF mRNA in the hippocampus during contextual learning (Hall et al., 2000). Spatial learning is also associated with an increase in BDNF mRNA levels in the hippocampus (Mizuno et al., 2000), combined with an activation of trkB as shown by increased levels of phosphorylated trkB (Mizuno et al., 2003). Hippocampal BDNF expression is reduced after the application of single or repeated immo-
Adult (2.5 months) male Wistar rats (Harlan, Udine, Italy) weighing 275–300 g were housed two per cage in a temperature-controlled room (24⫾2 °C), with a 12-h light/dark period (lights on: 07:00 – 19:00 h) for at least 2 weeks before the experiments. Food (Standard Diet 4RF21; Charles River, Calco, Italy) and tap water were provided ad libitum. No female rats were present in the animal room. The week before the experiment, the rats were daily handled gently by the same operator between 09:30 and 10:30 h to minimize stress response to manipulation. All experiments were carried out in accordance to the European Communities Council Directive 86/609/EEC and to the Italian national rules laid down by the Ministry of Health. All efforts were made to minimize animal suffering during the experiments.
Experimental model The experimental model consisted of two 2-way shuttle boxes (U. Basile, Varese, Italy). One cage, the “escape cage,” was divided into two equally sized compartments (24⫻21 cm, 21 cm high) and was equipped with an oscillating grid floor which allowed the termination of the electric stimulus (0.4 mA) when the rat moved from one compartment to the other through a 7⫻7 cm gate. There was no cue present before or during the delivery of the stimulus. A second cage, the “yoked cage” (48⫻21 cm, 21 cm high) was not divided, had no oscillating floor, and was electrically connected to the escape cage in order to deliver an electric stimulus (0.4 mA) to
*Corresponding author. Tel: ⫹39-06-4991-2511; fax: ⫹39-06-49912511. E-mail address: [email protected] (S. Scaccianoce). Abbreviations: ANOVA, analysis of variance; BDNF, brain-derived neurotrophic factor; GL, good learner; LPT, long-term potentiation; NT, neurotrophin; PL, poor learner; trkB, tyrosine kinase B; YGL, animals yoked to good learners; YPL, animals yoked to poor learners.
0306-4522/03$30.00⫹0.00 © 2003 IBRO. Published by Elsevier Ltd. All rights reserved. doi:10.1016/S0306-4522(03)00514-1
825
826
S. Scaccianoce et al. / Neuroscience 121 (2003) 825– 828
the grid floor for the same duration and rhythmicity as the escape cage during the sessions.
Behavioral studies Rats were subjected to a training consisting of three sessions of 30 trials for 3 consecutive days; in each trial, an electric shock was delivered (maximal duration 4 s) followed by an interval of 16 s. Paired sets of rats were placed in the escape and yoked cages, and subjected to simultaneous sessions. In each session, the number of escapes, the latencies to the escapes, and the number of failures were recorded. Upon termination of the last trial of the third session, each pair of rats was immediately killed by decapitation, the blood was collected, and the hippocampus was dissected and frozen. Rats trained in the escape cage were arbitrarily divided into two groups depending on the learning ability. “Good learners” (GL) had a number of escapes in the third session greater than, or equal to, the mean value calculated from all rats in the same session. “Poor learners” (PL) were defined as those having a number of escapes lower than the mean value. Rats located in the “yoked cage” were also divided into two groups depending on the ability of the partner: yoked to GL and yoked to PL. The control group consisted of undisturbed rats removed directly from their home cages.
Hippocampal BDNF determination Hippocampi were processed to measure BDNF protein content by ELISA using commercially available kits (BDNF Emax Immunoassay System; Promega, Milan, Italy). Briefly, tissues were sonicated in a lysis buffer (137 mM NaCl, 20 mM Tris, 1% Nonidet P-40, 10% glycerol, 1 mM phenylmethylsulfonyl fluoride, 10 g/ml aprotinin, 1 g/ml leupeptin, 0.5 mM sodium orthovanadate, 5 l/mg tissue) and centrifuged at 12,000⫻g at 4 °C for 30 min. The antibodies used have negligible crossreactivity (⬍3%) with other growth factors including nerve growth factor, NT-3 and NT-4/5. Colorimetric detection of peroxidase activity was achieved by adding TMB solution and peroxidase substrate. The enzymatic reaction was stopped with HCl (1 M) and the optical density of each well was measured at 450 nM using an Emax plate reader.
Plasma corticosterone determination Blood was collected into tubes containing 0.13 M EDTA. After centrifugation at 1,900⫻g at 4 °C for 15 min plasma samples were separated and stored at ⫺20 °C prior to assay. Plasma corticosterone concentrations were determined by RIA (ICN Biomedicals, Milan, Italy). The cross-reactivity of the polyclonal corticosterone–antisera with related substances was negligible. The interand intra-assay coefficients of variation were 8% and 3%, respectively, with a detection limit of 0.0125 ng/tube. Plasma samples were diluted 1:100 with Steroid Diluent (ICN Biomedicals). All measurements were in the linear range of the standard curve (0.0125–3.0 ng/tube).
Statistical analysis Behavioral data were analyzed with Student’s t-test. Hippocampal BDNF and plasma corticosterone data were analyzed with two-way (Good/Poor; Operant/Yoked) analysis of variance (ANOVA) followed by Duncan’s multiple comparison test. Differences between control and experimental groups were analyzed with ANOVA followed by Dunnett’s multiple comparisons test.
Fig. 1. Number of escapes during the last session in GL and in PL rats (A). Cumulative duration of shocks received during the last session by GL and PL rats (B). Data are expressed as means⫾S.E.M., n⫽6. *** P⬍0.001 (Student’s t-test).
RESULTS Behavioral profile During the third session, GL had a number of escapes higher than PL (GL 21.2⫾1.4; PL 2.8⫾1.2; means⫾S.E.M., n⫽6, t⫽9.9, P⬍0.001, Student’s t-test; Fig. 1A). Hence, they, and their paired-yoked rats, received a shorter duration of shocks as compared with PL and their yoked partners (GL 68.1⫾6.7; PL 110.5⫾3.8; means⫾S.E.M., n⫽6, t⫽5.5, P⬍0.001, Student’s t-test; Fig. 1B). Hippocampal BDNF concentration in GL and PL and respective yoked partners Hippocampal BDNF concentrations were determined immediately after the completion of the last trial of the third session (Fig. 2). Two-way ANOVA indicated a significant difference between GL and PL (F⫽19.3, P⬍0.001) and between operant and yoked rats (F⫽25.4, P⬍0.001) and no significant interaction between these two factors (F⫽1.16, P⬎0.05). GL showed higher hippocampal BDNF
S. Scaccianoce et al. / Neuroscience 121 (2003) 825– 828
827
trations were significantly (P⬍0.001, Dunnett’s test) elevated in all animals subjected to the trials (controls: 1.4⫾0.4; GL: 21.1⫾1.1; YGL: 21.3⫾1.9; PL: 20.0⫾1.3; YPL: 28.4⫾1.0, micrograms of corticosterone/100 ml of plasma; means⫾S.E.M., n⫽6). Remarkably, corticosterone levels in the YPL rats were higher than those observed in the other experimental groups (P⬍0.01 versus GL; P⬍0.01 versus YGL; and P⬍0.01 versus PL, Duncan’s test).
DISCUSSION
Fig. 2. Hippocampal BDNF concentrations in GL and in YGL rats, and in PL and in YPL rats. Rats were killed at the end of the last trial of the third session. Values (mean⫾S.E.M., n⫽6) are expressed as percent of variation with respect to the control (CONT) group (BDNF concentration: 3215.5⫾205.4 pg/g of tissue, n⫽6). # P⬍0.05, ## P⬍0.01 versus CONT (Dunnett’s test). * P⬍0.05, ** P⬍0.01 (Duncan’s test).
levels as compared with the corresponding yoked animals (YGL; P⬍0.05, Duncan’s test) and to control animals (P⬍0.05, Dunnett’s test). BDNF levels did not differ between YGL and control animals (P⬎0.05, Dunnett’s test). PL showed no changes in hippocampal BDNF levels as compared with control animals (P⬎0.05, Dunnett’s test). However, animals yoked to PL (YPL) showed a significant decrease in hippocampal BDNF levels (P⬍0.01 versus control, Dunnett’s test; P⬍0.01 versus PL, Duncan’s test). Plasma corticosterone concentration in GL and PL and respective yoked partners Plasma corticosterone levels were measured in nonshocked controls and in the experimental rats at the end of the last trial (Fig. 3). Two-way ANOVA indicated a significant difference between GL and PL (F⫽5.1, P⬍0.05) and between operant and yoked rats (F⫽10.2, P⬍0.01) and a significant interaction between these two factors (F⫽9.2, P⬍0.01). Compared with controls, corticosterone concen-
Fig. 3. Plasma corticosterone concentrations in control (CONT), GL, YGL, PL and in YPL rats. Rats were killed at the end of the last trial of the third session. Mean⫾S.E.M., n⫽6. ### P⬍0.001 versus CONT (Dunnett’s test). ** P⬍0.01 versus GL, YGL and PL (Duncan’s test).
These data show that the same amount of a stressful stimulus (i.e. an electric footshock) produces different effects on hippocampal BDNF and plasma corticosterone levels: the variable that governs the outcome is the chance to learn. Animals that performed well in the learning paradigm showed an increase in hippocampal BDNF levels despite of the high levels of circulating corticosterone. On the other side, YPL (which received a greater amount of electrical stimulation than YGL) showed a reduction in hippocampal BDNF protein associated with the highest plasma levels of corticosterone. Our finding of the learningcorrelated BDNF increase is in agreement with previous studies which, on the basis of an induction of hippocampal BDNF mRNA, have hypothesized the implication of this NT in learning and in memory formation (Schaaf et al., 1999; Hall et al., 2000). Studies carried out in cultured hippocampal neurons show that BDNF is rapidly secreted during induction of long-term potentiation (LPT), an established electrophysiological substrate for associative learning (Gartner and Staiger, 2002). In addition, application of BDNF induces LTP at the medial perforant path-granule cell synapses in living animals (Ying et al., 2002) and evokes Ca2⫹ transients in dentate granule cells of mouse hippocampal slices (Kovalchuk et al., 2002), which are required for the induction of a non-decremental form of LTP. There are indications of a close relationship between learning, BDNF and neurogenesis at hippocampal level. It has been postulated a role for BDNF as a signal that controls neurogenesis (Cameron et al., 1998) and learning is found to be associated with neurogenesis (see Grassi Zucconi and Giuditta, 2002 for a review). However, an impaired neurogenesis in the dentate gyrus is also observed in chronically stressed rats (McEwen, 2001). Whether or not hippocampal neurogenesis plays a role in learning under our experimental conditions remains to be established. An additional interesting feature of our study was observed in the PL and in the YPL groups. PL rats had no variations in hippocampal BDNF, which is in line with the observation of Ma et al. (1998). Rats yoked to this group, however, had a reduction in BDNF protein in the hippocampus combined with a greater increase in plasma corticosterone levels as compared with the PL group. Thus, we speculate that YPL rats, which received a greater amount of electrical stimulation because of the poor behavioral performance of the partner, had a more robust adrenocortical activation responsible for the reduction in hippocampal BDNF content. We conclude that a dynamic
828
S. Scaccianoce et al. / Neuroscience 121 (2003) 825– 828
relationship between corticosterone, learning and hippocampal BDNF exists, and that learning may overcome a lowering effect of stress-induced adrenocortical activation on hippocampal BDNF levels. Acknowledgements—We wish to thank Dr. Alessandro Giuliani (Istituto Superiore di Sanita`) for the assistance with the statistical analysis of the data. This research was supported by a grant from Universita` La Sapienza (ricerche di facolta`) to S.S.
REFERENCES Cameron HA, Hazel TG, McKay RD (1998) Regulation of neurogenesis by growth factors and neurotransmitters. J Neurobiol 36:287– 306. Duman RS, Heninger GR, Nestler EJ (1997) A molecular and cellular theory of depression. Arch Gen Psychiatry 54:597–606. Gartner A, Staiger V (2002) Neurotrophin secretion from hippocampal neurons evoked by long-term-potentiation-inducing electrical stimulation patterns. Proc Natl Acad Sci USA 99:6386 –6391. Grassi Zucconi G, Giuditta A (2002) Is it only neurogenesis? Rev Neurosci 13:375–382. Hall J, Thomas KL, Everitt BJ (2000) Rapid and selective induction of BDNF expression in the hippocampus during contextual learning. Nat Neurosci 3:533–535. Hofer M, Pagliusi SR, Hohn A, Leibrock J, Barde YA (1990) Regional distribution of brain-derived neurotrophic factor mRNA in the adult mouse brain. EMBO J 9:2459 –2464. Hyman C, Hofer M, Barde YA, Juhasz M, Yancopoulos GD, Squinto SP, Lindsay RM (1991) BDNF is a neurotrophic factor for dopaminergic neurons of the substantia nigra. Nature 350:230 –232. Kovalchuk Y, Hanse E, Kafitz KW, Konnerth A (2002) Postsynaptic induction of BDNF-mediated long-term potentiation. Science 295: 1729 –1734.
Ma YL, Wang HL, Wu HC, Wei CL, Lee EH (1998) Brain-derived neurotrophic factor antisense oligonucleotide impairs memory retention and inhibits long-term potentiation in rats. Neuroscience 82:957–967. McEwen BS (2001) Plasticity of the hippocampus: adaptation to chronic stress and allostatic load. Ann NY Acad Sci 933:265–277. Merlio JP, Ernfors P, Jaber M, Persson H (1992) Molecular cloning of rat trkC and distribution of cells expressing messengers RNAs for members of the trk family in the rat central nervous system. Neuroscience 51:513–532. Mizuno M, Yamada K, Olariu A, Nawa H, Nabeshima T (2000) Involvement of brain-derived neurotrophic factor in spatial memory formation and maintenance in a radial arm maze test in rats. J Neurosci 20:7116 –121. Mizuno M, Yamada K, Takei N, Tran MH, He J, Nakajima A, Nawa H, Nabeshima T (2003) Phosphatidylinositol 3-kinase: a molecule mediating BDNF-dependent spatial memory formation. Mol Psychiatry 8:217–224. Schaaf MJ, de Jong J, de Kloet ER, Vreugdenhil E (1998) Downregulation of BDNF mRNA and protein in the rat hippocampus by corticosterone. Brain Res 813:112–120. Schaaf MJ, Sibug RM, Duurland R, Fluttert MF, Oitzl MS, De Kloet ER, Vreugdenhil E (1999) Corticosterone effects on BDNF mRNA expression in the rat hippocampus during Morris water maze training. Stress 3:173–183. Smith MA, Makino S, Kvetnansky R, Post RM (1995) Stress and glucocorticoids affect the expression of brain-derived neurotrophic factor and neurotrophin-3 mRNAs in the hippocampus. J Neurosci 15:1768 –1777. Ying SW, Futter M, Rosenblum K, Webber MJ, Hunt SP, Bliss TV, Bramham CR (2002) Brain-derived neurotrophic factor induces long-term potentiation in intact adult hippocampus: requirement for ERK activation coupled to CREB and upregulation of Arc synthesis. J Neurosci 22:1532–1540.
(Accepted 2 July 2003)