J.S1 eep Res.(2011)20,259-266 Sleep and memory doi:10.1111i.1365-2869.2010.00895.x Sleep deprivation impairs contextual fear conditioning and attenuates subsequent behavioural,endocrine and neuronal responses ROELINA HAGEWOUD,LILLIAN J.BULTSMA,R.PAULIEN BARF, JAAP M.KOOLHAAS and PETER MEERLO Department of Behavioral Physiology,Center for Behavior and Neurosciences,University of Groningen,Haren,the Netherlands Accepted in revised form 11 September 2010;received 9 July 2010 SUMMARY Sleep deprivation (SD)affects hippocampus-dependent memory formation.Several studies in rodents have shown that brief SD immediately following a mild foot shock impairs consolidation of contextual fear memory as reflected in a reduced behavioural freezing response during re-exposure to the shock context later.In the first part of this study,we examined whether this reduced freezing response is accompanied by an attenuated fear-induced activation of the hypothalamic-pituitary-adrenal (HPA)axis. Results show that 6 h of SD immediately following the initial shock results in a diminished adrenal corticosterone(CORT)response upon re-exposure to the shock context the next day.In the second part,we established whether the attenuated freezing response in SD animals is associated with reduced activation of relevant brain areas known to be involved in the retrieval and expression of fear memory.Immunohisto- chemical analysis of brain slices showed that the normal increase in phosphorylation of the transcription factor 3,5'-cyclic AMP response-element binding protein (CREB) upon re-exposure to the shock context was reduced in SD animals in the CAl region of the hippocampus and in the amygdala.In conclusion,brief SD impairs the consolidation of contextual fear memory.Upon re-exposure to the context,this is reflected in a diminished behavioural freezing response,an attenuated HPA axis response and a reduction of the normal increase of phosphorylated CREB expression in the hippocampus and amygdala. KEYWORDS cAMP response-element binding protein,glucocorticoids,hippocampus, hypothalamic-pituitary-adrenal axis,learning,sleep restriction INTRODUCTION A commonly used learning task to study the effects of SD on memory consolidation in rodents is the fear conditioning Sleep loss is a serious problem in our society(Hublin et al.,2001; paradigm.In this task animals learn to associate a specific Rajaratnam and Arendt,2001).For the people affected,this context (the test environment)or a conditioned stimulus (for may have important consequences for cognitive function and example,a tone cue)with an aversive unconditioned stimulus performance (Ellenbogen.2005:Walker,2008).Various studies (foot shock).When the animals are exposed later to the same in both humans and animals have demonstrated that sleep context or cue they will exhibit a fear-related freezing response deprivation (SD)following learning impairs memory consoli- (Blanchard and Blanchard,1969;Fanselow,1980).Both dation (Karni et al.,1994;Mograss et al.,2009;Palchykova contextual and cued fear-learning involve the amygdala. et al.,2006;Smith and Rose,1997;Stickgold et al.,2000). However,contextual fear learning also depends upon the hippocampus (Chen et al.,1996;Kim and Fanselow,1992; Correspondence:Peter Meerlo.Department of Behavioral Physiology. Center for Behavior and Neurosciences,University of Groningen,PO Phillips and LeDoux,1992).Various studies in rats and mice Box 14.9750 AA Haren,the Netherlands.Tel.:+31 (0)50 363 2334; have shown that brief SD immediately following the initial fax:+31 (0)50 363 2331:e-mail:P.Meerlo@rug.nl shock exposure impairs memory consolidation for contextual 2010 European Sleep Research Society 259
doi: 10.1111/j.1365-2869.2010.00895.x Sleep deprivation impairs contextual fear conditioning and attenuates subsequent behavioural, endocrine and neuronal responses ROELINA HAGEWOUD, LILLIAN J. BULTSMA, R. PAULIEN BARF, JAAP M. KOOLHAAS and PETER MEERLO Department of Behavioral Physiology, Center for Behavior and Neurosciences, University of Groningen, Haren, the Netherlands Accepted in revised form 11 September 2010; received 9 July 2010 SUMMARY Sleep deprivation (SD) affects hippocampus-dependent memory formation. Several studies in rodents have shown that brief SD immediately following a mild foot shock impairs consolidation of contextual fear memory as reflected in a reduced behavioural freezing response during re-exposure to the shock context later. In the first part of this study, we examined whether this reduced freezing response is accompanied by an attenuated fear-induced activation of the hypothalamic–pituitary–adrenal (HPA) axis. Results show that 6 h of SD immediately following the initial shock results in a diminished adrenal corticosterone (CORT) response upon re-exposure to the shock context the next day. In the second part, we established whether the attenuated freezing response in SD animals is associated with reduced activation of relevant brain areas known to be involved in the retrieval and expression of fear memory. Immunohistochemical analysis of brain slices showed that the normal increase in phosphorylation of the transcription factor 3¢,5¢-cyclic AMP response-element binding protein (CREB) upon re-exposure to the shock context was reduced in SD animals in the CA1 region of the hippocampus and in the amygdala. In conclusion, brief SD impairs the consolidation of contextual fear memory. Upon re-exposure to the context, this is reflected in a diminished behavioural freezing response, an attenuated HPA axis response and a reduction of the normal increase of phosphorylated CREB expression in the hippocampus and amygdala. keywords cAMP response-element binding protein, glucocorticoids, hippocampus, hypothalamic–pituitary–adrenal axis, learning, sleep restriction INTRODUCTION Sleep loss is a serious problem in our society (Hublin et al., 2001; Rajaratnam and Arendt, 2001). For the people affected, this may have important consequences for cognitive function and performance (Ellenbogen, 2005; Walker, 2008). Various studies in both humans and animals have demonstrated that sleep deprivation (SD) following learning impairs memory consolidation (Karni et al., 1994; Mograss et al., 2009; Palchykova et al., 2006; Smith and Rose, 1997; Stickgold et al., 2000). A commonly used learning task to study the effects of SD on memory consolidation in rodents is the fear conditioning paradigm. In this task animals learn to associate a specific context (the test environment) or a conditioned stimulus (for example, a tone cue) with an aversive unconditioned stimulus (foot shock). When the animals are exposed later to the same context or cue they will exhibit a fear-related freezing response (Blanchard and Blanchard, 1969; Fanselow, 1980). Both contextual and cued fear-learning involve the amygdala. However, contextual fear learning also depends upon the hippocampus (Chen et al., 1996; Kim and Fanselow, 1992; Phillips and LeDoux, 1992). Various studies in rats and mice have shown that brief SD immediately following the initial shock exposure impairs memory consolidation for contextual Correspondence: Peter Meerlo, Department of Behavioral Physiology, Center for Behavior and Neurosciences, University of Groningen, PO Box 14, 9750 AA Haren, the Netherlands. Tel.: +31 (0) 50 363 2334; fax: +31 (0) 50 363 2331; e-mail: P.Meerlo@rug.nl J. Sleep Res. (2011) 20, 259–266 Sleep and memory 2010 European Sleep Research Society 259
260 R.Hagewoud et al. fear (Graves et al.,2003:Hagewoud et al.,2010c:Vecsey et al.. provided ad libitum.Animals were maintained on a 12-h 2009).When animals subjected to 5 or 6 h of SD following light/12-h dark cycle with lights on at 08:00 h.Light intensity training are re-exposed to the shock environment the next day. in the light phase was approximately 45 lux.All procedures they exhibit less freezing behaviour compared with trained. described in the present study were approved by the Animal non-sleep-deprived control animals.Importantly,memory Experiment Committee of the University of Groningen in consolidation for cued fear is not affected (Graves et al.. compliance with Dutch law and regulations. 2003;Vecsey et al.,2009).This demonstrates that SD has a negative effect on memory consolidation,particularly when it Experimental set-up involves the hippocampus.Thus,by affecting hippocampus function and interfering with the processing of contextual This study consisted of two experiments in order to examine information,SD leads to a weaker association between the whether a reduced freezing response during testing for contex- context and shock.which then leads to an attenuated fear tual fear memory in animals subjected to 6 h of SD immedi- response when the rats are re-exposed to this context later on. ately following training in the contextual fear conditioning The hippocampus and amygdala are not only important for paradigm is accompanied by:(1)a reduced neuroendocrine the formation of contextual fear memory,but are also involved response,assessed by measuring plasma levels of CORT:and in the retrieval and expression of these memories (Fanselow, (2)a reduced neuronal activation in relevant brain areas,as 2000:Fendt and Fanselow.1999).The recall of contextual fear examined by immunoreactivity for activated CREB.Both memories is associated with increased phosphorylation and experiments contained a non-sleep-deprived trained group activation of the transcription factor 3,5'-cyclic AMP (T,n=9)and a group subjected to 6 h of SD immediately response-element binding protein(CREB)in both the hippo- following training (SDT,n=10).In addition,Experiment 2 campus and amygdala (Mamiya et al.,2009).Therefore,one contained a home cage control (HCC)group (n=7). might expect that the attenuated fear response found in animals that were sleep-deprived after training is associated Contextual fear conditioning with an attenuated expression of phosphorylated CREB (pCREB)in these brain areas compared with trained,non- One week prior to the start of the fear conditioning experiments sleep-deprived control animals.However,such changes in all animals were handled daily for 2 min.Contextual fear regional brain activity have not yet been reported. conditioning was performed in a black plexiglass chamber Moreover,while freezing behaviour is the most commonly (40 x 40 x 40 cm),which was located in a separate experimen- used read-out of fear in most of the studies mentioned above.it tal room.During training an animal was placed in the chamber might be expected that an SD-induced weakening of condi- and exposed to the conditioning context for 3 min followed by a tioned contextual fear can also be measured on a physiological mild electric foot shock (0.7 mA,2 s)delivered through the level,particularly in the degree of activation of neuroendocrine stainless steel grid floor.The animal was removed from the stress systems such as the hypothalamic-pituitary-adrenal chamber and returned to its home cage 30 s after the shock. (HPA)axis. The chamber was cleaned thoroughly with 70%ethanol The first aim of the present study was to assess in rats between subjects.Twenty-four hours later the animal was whether brief SD following exposure to a mild foot shock leads placed in the same chamber for 5 min without receiving a not only to an attenuated freezing response,but also to a shock.Contextual memory was tested by assessing freezing reduced neuroendocrine response upon testing for contextual behaviour,defined as complete lack of movement except for fear memory the next day.Specifically.we hypothesized that in respiration.Behaviour was recorded on videotapes,which was SD animals the reduced freezing response upon re-exposure to analysed afterwards by an experimenter who was blind for the shock box is accompanied by an attenuated adrenal treatment of the animals.The amount of time the animals corticosterone (CORT)response.In the second part of our displayed freezing behaviour was expressed as a percentage of study we tested whether the reduced freezing response in SD total test time. animals is associated with a reduced activation of brain areas known to be involved in the retrieval and expression of fear memory,particularly a reduced expression of pCREB in the Sleep deprivation hippocampus and amygdala. Animals were sleep-deprived for 6 h immediately following training.SD was accomplished by mild stimulation,which involved tapping on the cage,gently shaking the cage or.when METHODS this was not sufficient to keep the animals awake,disturbing the sleeping nest (Hagewoud et al.,2010a:Van der Borght Animals and housing conditions et al..2006).Previous studies have shown that this procedure The experiments were performed with adult male Wistar rats is effective in keeping rodents awake for several hours,as (Harlan,Horst,the Netherlands),weighing 300-350 g at the established by electroencephalic recordings (Meerlo et al., start of the experiments.Animals were housed individually and 2001),without being a major stressor (Hagewoud et al., a layer of sawdust served as bedding.Food and water were 2010a,b,c) 2010 European Sleep Research Society,J.Sleep Res..20,259-266
fear (Graves et al., 2003; Hagewoud et al., 2010c; Vecsey et al., 2009). When animals subjected to 5 or 6 h of SD following training are re-exposed to the shock environment the next day, they exhibit less freezing behaviour compared with trained, non-sleep-deprived control animals. Importantly, memory consolidation for cued fear is not affected (Graves et al., 2003; Vecsey et al., 2009). This demonstrates that SD has a negative effect on memory consolidation, particularly when it involves the hippocampus. Thus, by affecting hippocampus function and interfering with the processing of contextual information, SD leads to a weaker association between the context and shock, which then leads to an attenuated fear response when the rats are re-exposed to this context later on. The hippocampus and amygdala are not only important for the formation of contextual fear memory, but are also involved in the retrieval and expression of these memories (Fanselow, 2000; Fendt and Fanselow, 1999). The recall of contextual fear memories is associated with increased phosphorylation and activation of the transcription factor 3¢,5¢-cyclic AMP response-element binding protein (CREB) in both the hippocampus and amygdala (Mamiya et al., 2009). Therefore, one might expect that the attenuated fear response found in animals that were sleep-deprived after training is associated with an attenuated expression of phosphorylated CREB (pCREB) in these brain areas compared with trained, nonsleep-deprived control animals. However, such changes in regional brain activity have not yet been reported. Moreover, while freezing behaviour is the most commonly used read-out of fear in most of the studies mentioned above, it might be expected that an SD-induced weakening of conditioned contextual fear can also be measured on a physiological level, particularly in the degree of activation of neuroendocrine stress systems such as the hypothalamic–pituitary–adrenal (HPA) axis. The first aim of the present study was to assess in rats whether brief SD following exposure to a mild foot shock leads not only to an attenuated freezing response, but also to a reduced neuroendocrine response upon testing for contextual fear memory the next day. Specifically, we hypothesized that in SD animals the reduced freezing response upon re-exposure to the shock box is accompanied by an attenuated adrenal corticosterone (CORT) response. In the second part of our study we tested whether the reduced freezing response in SD animals is associated with a reduced activation of brain areas known to be involved in the retrieval and expression of fear memory, particularly a reduced expression of pCREB in the hippocampus and amygdala. METHODS Animals and housing conditions The experiments were performed with adult male Wistar rats (Harlan, Horst, the Netherlands), weighing 300–350 g at the start of the experiments. Animals were housed individually and a layer of sawdust served as bedding. Food and water were provided ad libitum. Animals were maintained on a 12-h light ⁄ 12-h dark cycle with lights on at 08:00 h. Light intensity in the light phase was approximately 45 lux. All procedures described in the present study were approved by the Animal Experiment Committee of the University of Groningen in compliance with Dutch law and regulations. Experimental set-up This study consisted of two experiments in order to examine whether a reduced freezing response during testing for contextual fear memory in animals subjected to 6 h of SD immediately following training in the contextual fear conditioning paradigm is accompanied by: (1) a reduced neuroendocrine response, assessed by measuring plasma levels of CORT; and (2) a reduced neuronal activation in relevant brain areas, as examined by immunoreactivity for activated CREB. Both experiments contained a non-sleep-deprived trained group (T, n = 9) and a group subjected to 6 h of SD immediately following training (SDT, n = 10). In addition, Experiment 2 contained a home cage control (HCC) group (n = 7). Contextual fear conditioning One week prior to the start of the fear conditioning experiments all animals were handled daily for 2 min. Contextual fear conditioning was performed in a black plexiglass chamber (40 · 40 · 40 cm), which was located in a separate experimental room. During training an animal was placed in the chamber and exposed to the conditioning context for 3 min followed by a mild electric foot shock (0.7 mA, 2 s) delivered through the stainless steel grid floor. The animal was removed from the chamber and returned to its home cage 30 s after the shock. The chamber was cleaned thoroughly with 70% ethanol between subjects. Twenty-four hours later the animal was placed in the same chamber for 5 min without receiving a shock. Contextual memory was tested by assessing freezing behaviour, defined as complete lack of movement except for respiration. Behaviour was recorded on videotapes, which was analysed afterwards by an experimenter who was blind for treatment of the animals. The amount of time the animals displayed freezing behaviour was expressed as a percentage of total test time. Sleep deprivation Animals were sleep-deprived for 6 h immediately following training. SD was accomplished by mild stimulation, which involved tapping on the cage, gently shaking the cage or, when this was not sufficient to keep the animals awake, disturbing the sleeping nest (Hagewoud et al., 2010a; Van der Borght et al., 2006). Previous studies have shown that this procedure is effective in keeping rodents awake for several hours, as established by electroencephalic recordings (Meerlo et al., 2001), without being a major stressor (Hagewoud et al., 2010a,b,c). 260 R. Hagewoud et al. 2010 European Sleep Research Society, J. Sleep Res., 20, 259–266
Sleep deprivation and memory consolidation 261 Permanent heart cannulations,blood sampling and CORT assay with PBS the sections were incubated at room temperature for 3 h with biotinylated goat anti-rabbit immunoglobin G To assess conditioned HPA axis responses in Experiment 1. (IgG)(1 500:Jackson Immunoresearch Laboratories)in 1% permanent heart cathethers were used that permitted frequent NGS,0.1%Triton X-100 in PBS.After 5x 5 min rinsing blood sampling in unrestrained,freely moving rats (Steffens, with PBS.sections were incubated for 1.5 h at room 1969).The animals were provided with a polyethylene catheter temperature with avidin-biotin complex (1:500.ABC Elite in the right atrium of the heart.The catheter was inserted kit;Vector Laboratories,Burlingame,CA,USA),0.1% through the right jugular vein and externalized on the top of Triton X-100 in PBS.After this step the sections were rinsed the skull.The whole procedure was performed under isoflu- overnight in PBS at 4C.After rinsing the sections rane/O2 inhalation anaesthesia.Animals had at least 2 weeks 6 x 10 min with PBS.labelled cells were visualized with of recovery before the start of the experiment.During this diaminobenzidine (DAB,0.7 mg mL-in milli q water; period,animals were habituated to handling and blood Sigma-Aldrich.Steinheim.Germany)with 0.1%H2O,as a sampling procedures. reaction initiator.The reaction was stopped by rinsing with On the second day of the experiment,when animals were re- PBS. exposed to the shock context for contextual fear-testing,blood For each subject,three sections were selected at approxi- samples were taken at baseline,5 min.30 and 60 min.The mately bregma-2.80 to-3.60 mm for the dorsal hippocampus baseline sample was collected at least 1 h before testing.The and two sections were selected for the amygdala at bregma 5-min sample was taken in the shock box at the end of the -2.12 to -3.30 mm (Paxinos and Watson.1998).In the re-exposure.Blood samples were collected in precooled plastic amygdala,immunopositive cells were counted bilaterally in centrifuge tubes containing 0.01%ethylenediamine tetraacetic each section at a 50x magnification using a computerized acid (EDTA)as anticoagulant and antioxidant.Blood was image analysis system (Quantimet 550:Leica.Cambridge, centrifuged at 4C for 15 min at 2600g and plasma was UK).For the basolateral amygdala (BLA),a fixed sample stored at-80C until further processing.CORT levels were window of 0.37 mm2 was used and for the central nucleus of determined by radioimmunoassay (MP Biomedicals,Orange- the amygdala (CeN)the sample window was 0.20 mm. burg,NY,USA). A threshold was set that marked all cells to be included in the counting.In the hippocampus,the cell layers were densely Assessment of regional pCREB expression packed with pCREB immunopositive cells,which made it difficult to distinguish and count individual cells.Instead, In Experiment 2 we tested whether a reduced conditioned optical densities (OD)were measured for the granular cell freezing response in animals that were sleep-deprived after the layer of the dentate gyrus (DG)and for the pyramidal cell initial shock exposure would be associated with changes in layer of cornu ammonis (CA)areas I and 3 of the dorsal regional brain activation by performing immunohistochemis- hippocampus using a 50x magnification.The OD is expressed try for phosphorylation of the transcription factor CREB.One in arbitrary units corresponding to grey levels using the hour after testing for contextual fear memory,rats were Quantimet image analysis system (Leica).To correct for sacrificed for brain collection.A group of HCC rats was variability in background staining among sections,the back- sacrificed in parallel.Under deep pentobarbital anaesthesia, ground labelling was measured in the stratum radiatum and rats were perfused transcardially with 150 mL 0.9%NaCl,1E extracted from the OD of the area of interest.The experi- per mL heparin,followed by 300 mL 4%paraformaldehyde menter was blind to the treatment of individual animals during for fixation.Brains were collected,postfixated for 24 h in 4% all cell counting and OD measurements.Data on pCREB paraformaldehyde,rinsed for I day in 0.01 M phosphate- immunoreactivity are expressed as percentage of the mean buffered saline (PBS.pH 7.4)and then transferred to a 30% value of the HCC group sucrose in PBS cryoprotectant overnight at 4 C.Brains were stored at-80 C until further processing. Statistical analysis Thirty um coronal sections containing the amygdala and hippocampus were collected and stored in PBS containing Behavioural data were analysed using an independent-samples 0.1%sodium azide.The brain sections were rinsed 3 x 5 min 1-test.The CORT responses were analysed using a repeated- in PBS,followed by 30 min in 0.3%H2O2 in PBS.After measures analysis of variance (ANOvA)with a between-subject 4 x 5 min rinsing in PBS,sections were preincubated at room factor 'treatment'(T or SDT)and a within-subject factor 'time temperature for 30 min in 5%normal goat serum (NGS: (at baseline,5,30 and 60 min).A post-hoc t-test was used to Jackson Immunoresearch Laboratories,West Grove,PA. establish at which time-points the treatment groups differed. USA),0.1%Triton X-100 in PBS to block non-specific pCREB immunoreactivity was analysed using a one-way ANovA binding of immunoreagents.Subsequently,sections were with a between-subjects factor 'treatment'(HCC.T or SDT). incubated for 2 h at room temperature followed by overnight Post-hoc comparisons were made using a Tukey test.In all incubation at 4C with rabbit polyclonal anti-p-CREB cases P 0.05 was considered significant.All data in the text antibody (1:2000:Upstate,Temecula,CA,USA)in 0.3% and figures are expressed as mean+standard error of the Triton X-100.1%NGS in PBS.After rinsing 4x 10 min mean (SEM). 2010 European Sleep Research Society.J.Sleep Res.,20.259-266
Permanent heart cannulations, blood sampling and CORT assay To assess conditioned HPA axis responses in Experiment 1, permanent heart cathethers were used that permitted frequent blood sampling in unrestrained, freely moving rats (Steffens, 1969). The animals were provided with a polyethylene catheter in the right atrium of the heart. The catheter was inserted through the right jugular vein and externalized on the top of the skull. The whole procedure was performed under isoflurane ⁄ O2 inhalation anaesthesia. Animals had at least 2 weeks of recovery before the start of the experiment. During this period, animals were habituated to handling and blood sampling procedures. On the second day of the experiment, when animals were reexposed to the shock context for contextual fear-testing, blood samples were taken at baseline, 5 min, 30 and 60 min. The baseline sample was collected at least 1 h before testing. The 5-min sample was taken in the shock box at the end of the re-exposure. Blood samples were collected in precooled plastic centrifuge tubes containing 0.01% ethylenediamine tetraacetic acid (EDTA) as anticoagulant and antioxidant. Blood was centrifuged at 4 C for 15 min at 2600 g and plasma was stored at )80 C until further processing. CORT levels were determined by radioimmunoassay (MP Biomedicals, Orangeburg, NY, USA). Assessment of regional pCREB expression In Experiment 2 we tested whether a reduced conditioned freezing response in animals that were sleep-deprived after the initial shock exposure would be associated with changes in regional brain activation by performing immunohistochemistry for phosphorylation of the transcription factor CREB. One hour after testing for contextual fear memory, rats were sacrificed for brain collection. A group of HCC rats was sacrificed in parallel. Under deep pentobarbital anaesthesia, rats were perfused transcardially with 150 mL 0.9% NaCl, 1E per mL heparin, followed by 300 mL 4% paraformaldehyde for fixation. Brains were collected, postfixated for 24 h in 4% paraformaldehyde, rinsed for 1 day in 0.01 m phosphatebuffered saline (PBS, pH 7.4) and then transferred to a 30% sucrose in PBS cryoprotectant overnight at 4 C. Brains were stored at )80 C until further processing. Thirty lm coronal sections containing the amygdala and hippocampus were collected and stored in PBS containing 0.1% sodium azide. The brain sections were rinsed 3 · 5 min in PBS, followed by 30 min in 0.3% H2O2 in PBS. After 4 · 5 min rinsing in PBS, sections were preincubated at room temperature for 30 min in 5% normal goat serum (NGS; Jackson Immunoresearch Laboratories, West Grove, PA, USA), 0.1% Triton X-100 in PBS to block non-specific binding of immunoreagents. Subsequently, sections were incubated for 2 h at room temperature followed by overnight incubation at 4 C with rabbit polyclonal anti-p-CREB antibody (1 : 2000; Upstate, Temecula, CA, USA) in 0.3% Triton X-100, 1% NGS in PBS. After rinsing 4 · 10 min with PBS the sections were incubated at room temperature for 3 h with biotinylated goat anti-rabbit immunoglobin G (IgG) (1 : 500; Jackson Immunoresearch Laboratories) in 1% NGS, 0.1% Triton X-100 in PBS. After 5 · 5 min rinsing with PBS, sections were incubated for 1.5 h at room temperature with avidin–biotin complex (1: 500, ABC Elite kit; Vector Laboratories, Burlingame, CA, USA), 0.1% Triton X-100 in PBS. After this step the sections were rinsed overnight in PBS at 4 C. After rinsing the sections 6 · 10 min with PBS, labelled cells were visualized with diaminobenzidine (DAB, 0.7 mg mL)1 in milli q water; Sigma-Aldrich, Steinheim, Germany) with 0.1% H2O2 as a reaction initiator. The reaction was stopped by rinsing with PBS. For each subject, three sections were selected at approximately bregma )2.80 to )3.60 mm for the dorsal hippocampus and two sections were selected for the amygdala at bregma )2.12 to )3.30 mm (Paxinos and Watson, 1998). In the amygdala, immunopositive cells were counted bilaterally in each section at a 50· magnification using a computerized image analysis system (Quantimet 550; Leica, Cambridge, UK). For the basolateral amygdala (BLA), a fixed sample window of 0.37 mm2 was used and for the central nucleus of the amygdala (CeN) the sample window was 0.20 mm2 . A threshold was set that marked all cells to be included in the counting. In the hippocampus, the cell layers were densely packed with pCREB immunopositive cells, which made it difficult to distinguish and count individual cells. Instead, optical densities (OD) were measured for the granular cell layer of the dentate gyrus (DG) and for the pyramidal cell layer of cornu ammonis (CA) areas 1 and 3 of the dorsal hippocampus using a 50· magnification. The OD is expressed in arbitrary units corresponding to grey levels using the Quantimet image analysis system (Leica). To correct for variability in background staining among sections, the background labelling was measured in the stratum radiatum and extracted from the OD of the area of interest. The experimenter was blind to the treatment of individual animals during all cell counting and OD measurements. Data on pCREB immunoreactivity are expressed as percentage of the mean value of the HCC group. Statistical analysis Behavioural data were analysed using an independent-samples t-test. The CORT responses were analysed using a repeatedmeasures analysis of variance (anova) with a between-subject factor treatment (T or SDT) and a within-subject factor time (at baseline, 5, 30 and 60 min). A post-hoc t-test was used to establish at which time-points the treatment groups differed. pCREB immunoreactivity was analysed using a one-way anova with a between-subjects factor treatment (HCC, T or SDT). Post-hoc comparisons were made using a Tukey test. In all cases P < 0.05 was considered significant. All data in the text and figures are expressed as mean ± standard error of the mean (SEM). Sleep deprivation and memory consolidation 261 2010 European Sleep Research Society, J. Sleep Res., 20, 259–266
262 R.Hagewoud et al. RESULTS (a) 50 Experiment 1:freezing behaviour and CORT response upon re-exposure to shock context 40 As expected,rats that were sleep-deprived for 6 h immediately following training (SDT)showed reduced freezing upon re- 8 exposure to the shock context 24 h after training compared 30 with control animals(T)(17.3±3.9%versus33.6±6.4%， with n=10 and n=9,respectively,in the two groups; 20 117=2.194,P=0.046:Fig.1a). Re-exposure to the shock context induced a clear CORT 10 response in all animals (repeated-measures ANOvA,time effect: F3.42 =73.26,P <0.01).However,this response was reduced significantly in SDT animals compared to T animals (treat- 0 SDT ment x time interaction:F3.42 =3.42,P<0.05;see Fig.Ib). Post-hoc t-tests showed that the plasma levels of CORT in (b)251 the SDT animals were significantly lower than those of the -O-T T animals at t 30 and 60 min (P<0.05 in both cases). ◆SDT Due to missing samples,three animals (one T and two SDT) 20 ★ were excluded from the CORT analysis. 15 Experiment 2:neuronal activation upon re-exposure to shock 骂 context 10 Again,rats that were sleep-deprived for 6 h immediately following training (SDT)showed reduced freezing upon re- 5 exposure to the shock box 24 h after training compared with ● trained,non-sleep-deprived control animals(T)(22.5+3.4% versus 35.2+5.0%,with n 10 and n =9,respectively; 117=2.147,P=0.047. 10 2030 40 50 鸟 A representative picture of pCREB expression in the Time(min) hippocampus is shown in Fig.2a.In the hippocampal sections Figure 1.Effect of brief sleep deprivation(SD)following training on that were collected I h after testing for contextual fear,the freezing behaviour and corticosterone (CORT)response upon re- optical density of pCREB immunoreactivity was affected by exposure to the shock context.(a)Twenty-four hours after training all the experimental treatment in the CAl area(F223 12.665, animals were tested for contextual fear during a 5-min test phase. Animals sleep-deprived for 6 h immediately following training(SDT. P<0.01),but not in the CA3 area (F2.23 1.051,P>0.3) n=10)displayed significantly less freezing behaviour in response to and DG (F2.23 =0.295,P>0.3)(Fig.2b-d).Trained rats the shock context than control animals (T,n 9).(b)CORT response that were re-exposed to the shock context (T)had a significant upon contextual fear-testing in trained animals (T,n=8)and animals increase in pCREB expression in the CAl area compared to subjected to 6 h of SD(SDT,n=8)immediately after training in the HCC animals (P<0.01).This increase was still visible in the contextual fear conditioning paradigm.Plasma CORT levels were animals subjected to brief SD immediately following training significantly lower in SDT animals compared to T animals at t=30 and 60 min.Data are expressed as mean standard error of the (SDT versus HCC,P 0.05),but it was significantly smaller mean.*P 0.05 than in the trained,non-sleep-deprived control animals(SDT versus T,P 0.05). A representative picture of pCREB expression in the increase did not reach statistical significance when compared amygdala is shown in Fig.3a.Due to missing sections,two to the HCC group (P>0.4 for both subregions). animals (one T and one SDT animal)were excluded from analysis.ANOvA revealed a significant main treatment effect for DISCUSSION the number of pCREB-positive cells in the BLA as well as in the CeN(F2.21=3.901,P<0.05andf2.21=3.528, In the present study we confirm the negative effects of brief P 0.05,respectively)(Fig.3b,c).Specifically,the number SD on the consolidation of contextual fear memory. of pCREB-positive cells in the BLA and CeN was higher in We show that 6 h of SD immediately following a mild foot animals subjected to contextual fear-testing (T)compared with shock leads to a reduced freezing response upon re-exposure animals in the HCC group (post-hoc Tukey test,P<0.05 for to the shock box the next day.In addition to this attenuated both regions).Although the number of pCREB immunoreac- behavioural response,the SD animals also displayed a redu- tive cells was also elevated slightly in the SDT group,this ced neuroendocrine activation,as shown by an attenuated 2010 European Sleep Research Society,J.Sleep Res..20,259-266
RESULTS Experiment 1: freezing behaviour and CORT response upon re-exposure to shock context As expected, rats that were sleep-deprived for 6 h immediately following training (SDT) showed reduced freezing upon reexposure to the shock context 24 h after training compared with control animals (T) (17.3 ± 3.9% versus 33.6 ± 6.4%, with n = 10 and n = 9, respectively, in the two groups; t17 = 2.194, P = 0.046; Fig. 1a). Re-exposure to the shock context induced a clear CORT response in all animals (repeated-measures anova, time effect: F3,42 = 73.26, P < 0.01). However, this response was reduced significantly in SDT animals compared to T animals (treatment · time interaction: F3,42 = 3.42, P < 0.05; see Fig. 1b). Post-hoc t-tests showed that the plasma levels of CORT in the SDT animals were significantly lower than those of the T animals at t = 30 and 60 min (P < 0.05 in both cases). Due to missing samples, three animals (one T and two SDT) were excluded from the CORT analysis. Experiment 2: neuronal activation upon re-exposure to shock context Again, rats that were sleep-deprived for 6 h immediately following training (SDT) showed reduced freezing upon reexposure to the shock box 24 h after training compared with trained, non-sleep-deprived control animals (T) (22.5 ± 3.4% versus 35.2 ± 5.0%, with n = 10 and n = 9, respectively; t17 = 2.147, P = 0.047). A representative picture of pCREB expression in the hippocampus is shown in Fig. 2a. In the hippocampal sections that were collected 1 h after testing for contextual fear, the optical density of pCREB immunoreactivity was affected by the experimental treatment in the CA1 area (F2,23 = 12.665, P < 0.01), but not in the CA3 area (F2,23 = 1.051, P > 0.3) and DG (F2,23 = 0.295, P > 0.3) (Fig. 2b–d). Trained rats that were re-exposed to the shock context (T) had a significant increase in pCREB expression in the CA1 area compared to HCC animals (P < 0.01). This increase was still visible in the animals subjected to brief SD immediately following training (SDT versus HCC, P < 0.05), but it was significantly smaller than in the trained, non-sleep-deprived control animals (SDT versus T, P < 0.05). A representative picture of pCREB expression in the amygdala is shown in Fig. 3a. Due to missing sections, two animals (one T and one SDT animal) were excluded from analysis. anova revealed a significant main treatment effect for the number of pCREB-positive cells in the BLA as well as in the CeN (F2,21 = 3.901, P < 0.05 and F2,21 = 3.528, P < 0.05, respectively) (Fig. 3b,c). Specifically, the number of pCREB-positive cells in the BLA and CeN was higher in animals subjected to contextual fear-testing (T) compared with animals in the HCC group (post-hoc Tukey test, P < 0.05 for both regions). Although the number of pCREB immunoreactive cells was also elevated slightly in the SDT group, this increase did not reach statistical significance when compared to the HCC group (P > 0.4 for both subregions). DISCUSSION In the present study we confirm the negative effects of brief SD on the consolidation of contextual fear memory. We show that 6 h of SD immediately following a mild foot shock leads to a reduced freezing response upon re-exposure to the shock box the next day. In addition to this attenuated behavioural response, the SD animals also displayed a reduced neuroendocrine activation, as shown by an attenuated T SDT Freezing (%) 0 10 20 30 40 50 * Time (min) 0 10 20 30 40 50 60 CORT (µg dL–1) 0 5 10 15 20 25 T SDT * * (a) (b) Figure 1. Effect of brief sleep deprivation (SD) following training on freezing behaviour and corticosterone (CORT) response upon reexposure to the shock context. (a) Twenty-four hours after training all animals were tested for contextual fear during a 5-min test phase. Animals sleep-deprived for 6 h immediately following training (SDT, n = 10) displayed significantly less freezing behaviour in response to the shock context than control animals (T, n = 9). (b) CORT response upon contextual fear-testing in trained animals (T, n = 8) and animals subjected to 6 h of SD (SDT, n = 8) immediately after training in the contextual fear conditioning paradigm. Plasma CORT levels were significantly lower in SDT animals compared to T animals at t = 30 and 60 min. Data are expressed as mean ± standard error of the mean. *P < 0.05. 262 R. Hagewoud et al. 2010 European Sleep Research Society, J. Sleep Res., 20, 259–266
Sleep deprivation and memory consolidation 263 (a) (c CA3 120 GA 100 80 60 40 20 HCC SDT (b) d DG 120 120 100 (33HJO 100 80 8 80 60 20 0 HCC 人 SDT HCC SDT Figure 2.Phosphorylated 3.5'-cyclic AMP response-element binding protein (pCREB)expression in the hippocampus upon re-exposure to the shock context 24 h following training.(a)Representative photomicrograph of pCREB immunoreactivity in the hippocampus.Optical density of pCREB immunoreactivity was measured for the granular cell layer of the dentate gyrus(DG)and the pyramidal cell layer of the CA3 and CAl areas of the hippocampus.The scale bar represents 500 um.Home cage controls(HCC,n =7),animals trained in the contextual fear conditioning paradigm without any interference (T,n=9)and animals subjected to 6 h of sleep deprivation immediately following training(SDT,n=10)were sacrificed I h after testing.(b-d)Testing for contextual fear did not affect pCREB immunoreactivity in the DG and CA3 areas.However,it increased pCREB expression significantly in the CAl area compared with HCC animals;6 h of SD immediately following training also increased pCREB expression in the CAl area but significantly less than in the trained group without any interference.Data are expressed as mean±standard error of the mean..◆p<0.05. CORT response and a lower neuronal activation within response upon re-exposure to the shock context 24 h after brain areas mediating contextual fear,as demonstrated by training. an attenuated increase in pCREB immunoreactivity in the Upon re-exposure to the shock context,the weaker hippocampus and amygdala. contextual memory was also associated with a reduced Several studies have demonstrated previously that brief SD neuronal activation,as shown by an attenuated increase in after training impairs the formation of fear memory (Graves pCREB expression in the hippocampus.It is noteworthy that et al.,2003;Hagewoud et al.,2010c;Vecsey et al.,2009). in the non-sleep-deprived rats,testing for contextual fear Importantly.SD does not affect the consolidation of amyg- induced an increase in pCREB expression solely in the CAl dala-dependent cued fear memory but only impairs selectively area of the hippocampus.This is in agreement with other the consolidation of hippocampus-dependent contextual fear studies,showing that recall of contextual memories induces memory (Graves et al.,2003;Vecsey et al.,2009).In other pCREB-regulated immediate early genes c-fos and zif268 words,SD does not have a general non-specific effect on fear specifically in the CAl area of the hippocampus(Hall et al., memory but,rather,impairs selectively the formation of fear 2001a:Strekalova et al.,2003).This increase in pCREB memory when this process involves the hippocampus. expression in the CAl region,compared with HCC animals, By affecting hippocampus function and interfering with the may not simply reflect memory retrieval,but may also be processing of contextual information,SD leads to a weaker involved in memory reconsolidation or extinction following association between the context and shock.Therefore,when retrieval (Mamiya et al.,2009). the rats are re-exposed to this context later on,they do not The reduced fear response,caused by impaired contextual show the full-blown fear response displayed by non-sleep- fear memory,was also associated with an attenuated activa- deprived animals.In the present study this reduced fear tion of the amygdala.While trained non-sleep-deprived response of rats in the SD group was reflected in an attenuated animals displayed a significant increase in pCREB expression behavioural freezing response and a lower adrenal CORT in the BLA and CeN.the trained SD animals did not.Indeed. 2010 European Sleep Research Society.J.Sleep Res.,20.259-266
CORT response and a lower neuronal activation within brain areas mediating contextual fear, as demonstrated by an attenuated increase in pCREB immunoreactivity in the hippocampus and amygdala. Several studies have demonstrated previously that brief SD after training impairs the formation of fear memory (Graves et al., 2003; Hagewoud et al., 2010c; Vecsey et al., 2009). Importantly, SD does not affect the consolidation of amygdala-dependent cued fear memory but only impairs selectively the consolidation of hippocampus-dependent contextual fear memory (Graves et al., 2003; Vecsey et al., 2009). In other words, SD does not have a general non-specific effect on fear memory but, rather, impairs selectively the formation of fear memory when this process involves the hippocampus. By affecting hippocampus function and interfering with the processing of contextual information, SD leads to a weaker association between the context and shock. Therefore, when the rats are re-exposed to this context later on, they do not show the full-blown fear response displayed by non-sleepdeprived animals. In the present study this reduced fear response of rats in the SD group was reflected in an attenuated behavioural freezing response and a lower adrenal CORT response upon re-exposure to the shock context 24 h after training. Upon re-exposure to the shock context, the weaker contextual memory was also associated with a reduced neuronal activation, as shown by an attenuated increase in pCREB expression in the hippocampus. It is noteworthy that in the non-sleep-deprived rats, testing for contextual fear induced an increase in pCREB expression solely in the CA1 area of the hippocampus. This is in agreement with other studies, showing that recall of contextual memories induces pCREB-regulated immediate early genes c-fos and zif268 specifically in the CA1 area of the hippocampus (Hall et al., 2001a; Strekalova et al., 2003). This increase in pCREB expression in the CA1 region, compared with HCC animals, may not simply reflect memory retrieval, but may also be involved in memory reconsolidation or extinction following retrieval (Mamiya et al., 2009). The reduced fear response, caused by impaired contextual fear memory, was also associated with an attenuated activation of the amygdala. While trained non-sleep-deprived animals displayed a significant increase in pCREB expression in the BLA and CeN, the trained SD animals did not. Indeed, CA1 HCC T SDT 0 20 40 60 80 100 120 * * * HCC T SDT 0 20 40 60 80 100 120 DG HCC T SDT Optical density (% of HCC) Optical density (% of HCC) Optical density (% of HCC) 0 20 40 60 80 100 120 CA3 CA3 DG CA1 (a) (c) (b) (d) Figure 2. Phosphorylated 3¢,5¢-cyclic AMP response-element binding protein (pCREB) expression in the hippocampus upon re-exposure to the shock context 24 h following training. (a) Representative photomicrograph of pCREB immunoreactivity in the hippocampus. Optical density of pCREB immunoreactivity was measured for the granular cell layer of the dentate gyrus (DG) and the pyramidal cell layer of the CA3 and CA1 areas of the hippocampus. The scale bar represents 500 lm. Home cage controls (HCC, n = 7), animals trained in the contextual fear conditioning paradigm without any interference (T, n = 9) and animals subjected to 6 h of sleep deprivation immediately following training (SDT, n = 10) were sacrificed 1 h after testing. (b–d) Testing for contextual fear did not affect pCREB immunoreactivity in the DG and CA3 areas. However, it increased pCREB expression significantly in the CA1 area compared with HCC animals; 6 h of SD immediately following training also increased pCREB expression in the CA1 area but significantly less than in the trained group without any interference. Data are expressed as mean ± standard error of the mean. *P < 0.05. Sleep deprivation and memory consolidation 263 2010 European Sleep Research Society, J. Sleep Res., 20, 259–266