Epigenetic mechanisms in stress-related memory formation

ARTICLE IN PRESS Psychoneuroendocrinology (2007) 32, S21–S25

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/psyneuen

Epigenetic mechanisms in stress-related memory formation Johannes M.H.M. Reul, Yalini Chandramohan Henry Wellcome Laboratories for Integrative Neuroscience and Endocrinology, University of Bristol, Whitson Street, Bristol BS1 3NY, UK Received 26 January 2007; received in revised form 6 March 2007; accepted 12 March 2007

KEYWORDS Chromatin; Histone; Phosphorylation; Acetylation; c-fos; Glucocorticoid; NMDA; MAPK; ERK; MSK; Behavior; Learning and memory; Cognition

Summary Coping with stressful events is part of everyone’s daily life. The organism’s response to stress is a complex array of physiological and behavioral changes aimed at the preservation/protection of the organism during the stressful event as well as at stimulating adaptive and mnemonic processes in case the event would re-occur in the future. The hippocampus including its ‘gate’, the dentate gyrus, is highly involved in these processes. We have been collecting evidence suggesting that the transcriptional activation seen in dentate gyrus neurons, which are involved in the encoding of memories of a psychologically stressful event, requires chromatin remodeling in these neurons driven by the phosphorylation (at Serine10) and acetylation (at Lysine14) of histone H3. These particular epigenetic mechanisms are potentially of special interest for neuronal functioning as they are associated with the induction of hitherto silent genes. The phospho-acetylation of histone H3 is brought about by the concurrent activation of two, possibly converging, signaling pathways, being the glucocorticoid receptor and the NMDA/ MAPK/ERK/MSK signaling pathways. Thus, we present a new model about how signaling to the chromatin can shape a specific gene transcriptional response in dentate granule neurons required for the encoding of memory of the stressful event. & 2007 Elsevier Ltd. All rights reserved.

1. Introduction: stress, glucocorticoids and emotional memory It is well-known that memories of emotionally stressful events are strong and sometimes lasting for life. Such Corresponding author. Tel.: +44 117 331 3137;

fax: +44 117 331 3139. E-mail address: [email protected] (J.M.H.M. Reul). 0306-4530/$ - see front matter & 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.psyneuen.2007.03.016

memories are established by a complex interplay between parts of the limbic system (e.g. amygdala, hippocampus) and regions of the neocortex. The amygdala processes the emotional aspects of the event, whereas the hippocampus plays a salient role in the organization of the stress response and in the storage process of memories of the event. Glucocorticoid hormones released as part of the stress response play a critical role in enhancing the formation of such memories. This effect of glucocorticoids on stressrelated memory has been shown in various behavioral

ARTICLE IN PRESS S22 models such as fear conditioning (Roozendaal et al., 2006), Morris water maze behavior (Oitzl and De Kloet, 1992) and the forced swim test (Bilang-Bleuel et al., 2005).

2. NMDA receptors, signaling and neuroplasticity At present, however, the identity of the neuronal population(s) and the neurocircuitry involved in stress-related memory formation as well as the molecular mechanisms steering the plasticity processes in these neurons are still not resolved. A role of the neurotransmitter glutamate has been in the focus of research for many years. Particularly with regard to neuroplasticity processes the glutamatebinding N-methyl-D-aspartate receptor (NMDA-R) has received much attention. Upon ligand binding and concomitant membrane depolarization, NMDA-Rs act as Na+/Ca2+permeable cation channels resulting in a rise of intracellular Ca2+ concentrations leading to the activation of Ca2+calmodulin kinase II (CAMKII), protein kinase C (PKC), protein kinase A (PKA, via Ca2+-mediated activation of adenylate cyclases 1 and 8 (ACy1, ACy8)) and the Ras/Raf/ MAPK/ERK signaling cascade (mitogen-activated protein kinase (MAPK), extracellular signal-regulated kinase (ERK)) (Impey et al., 1999). MAPK/ERK signaling is critical for synaptic plasticity (e.g. dendritic spine formation), longterm potentiation (LTP; Impey et al., 1999), and learning and memory (Impey et al., 1999). The transcription factor CREB (cAMP responsive element-binding protein) which, after phosphorylation by PKA, CAMKII, ERK and other kinases, activates CRE-mediated gene transcription is thought to play a significant role in these neuroplasticity processes and memory formation.

3. Chromatin remodeling through histone modifications During the last decade, it has become clear that gene transcription involves chromatin remodeling events to allow the RNA polymerase complex access to the DNA template and transcribe genes (Spencer and Davie, 1999; Jaskelioff and Peterson, 2003). Chromatin consists of DNA–histone protein complexes building together a so-called nucleosomal structure. Histone proteins such as histone H3 have evolutionary highly conserved N-terminal tails which protrude from the nucleosome and can be subjected to post-translational modifications such as acetylation, phosphorylation and others (Strahl and Allis, 2000). Importantly, these histone modifications determine the functional state of the chromatin. Particularly, the combination of phosphorylation of Serine-10 (Ser10) with the acetylation of Lysine-14 (Lys14) within the histone H3 tail (P(Ser10)-Ac(Lys14)-H3 or PAc-H3) is of interest as it is associated with the local opening of condensed, inactive chromatin allowing the transcriptional activation of hitherto dormant genes (Cheung et al., 2000; Clayton et al., 2000; Nowak and Corces, 2000). In contrast, the sole acetylation of lysine amino acids is seen in open, transcriptionally active chromatin (Strahl and Allis, 2000). Thus, the specific posttranslational histone modifications provide clues regarding

J.M.H.M. Reul, Y. Chandramohan the transcriptional activity of the cell whereby uniquely the phospho-acetylation of histone H3 is of special interest as it indicates that novel transcriptional events have been initiated. For instance, histone H3 phosphorylation and phospho-acetylation have been found associated with the induction of c-fos and c-jun gene expression in mouse fibroblast cells (Clayton et al., 2000) and heat shock proteins in Drosophila after heat treatment (Nowak and Corces, 2000). The mitogen- and stress-activated kinases 1 and 2 (MSK1/2) have been identified as the principal histone H3 kinases (Soloaga et al., 2003) but they are also potent CREB kinases (Wiggin et al., 2002). They are ubiquitously expressed and are activated by ERK (in response to growth factors, neurotransmitters, peptide hormones and others) and members of the p38 family of MAPKs (in response to various cellular stress stimuli and cytokines). It should be noted, however, that these insights are mainly based on in vitro work using cell cultures and cell-free systems.

4. Chromatin remodeling underlying psychological stress-induced transcriptional activation We discovered that the phospho-acetylation of histone H3 occurs in neurons of rats and mice in vivo under physiological conditions and in a functional context, i.e. in relation to stress-related memory formation. We found neurons showing phospho-acetylated histone H3 in their nuclei mainly in the dentate gyrus of the hippocampus and few neurons scattered in the amygdala, neocortex and striatum (Bilang-Bleuel et al., 2005; Chandramohan et al., 2007, in press). Our studies indicate that histone H3 if phosphorylated at Ser10 will be acetylated at Lys14 (Chandramohan et al., 2007, in press). Exposure of rats or mice to a stressor with a strong psychological component (which involves processing by neocortical and higher limbic (i.e. amygdala, hippocampus structures), such as forced swimming, a predator or a novel environment, resulted in a marked increase in the number of PAc-H3-positive neurons specifically in the dentate gyrus (Bilang-Bleuel et al., 2005; Chandramohan et al., 2007, in press). The increase was transient and lasted for up to 4 h (Chandramohan et al., 2007, in press). In view of the role of phospho-acetylation of histone H3 in chromatin remodeling and de novo transcriptional activation, our results suggest that in neurons showing histone H3 phospho-acetylation after psychological stress a distinct gene transcriptional response is evolving. Indeed, in dentate gyrus neurons the phospho-acetylation of histone H3 is closely associated with the induction of c-fos in those neurons (Chandramohan et al., 2007, in press). This was an extraordinary finding because this association was specific to the dentate gyrus and was not found elsewhere in the brain. For instance, forced swimming and novelty challenges evoke c-fos induction but no histone H3 phospho-acetylation, in the hypothalamic paraventricular nucleus (PVN), a key structure in generating the glucocorticoid and sympathetic components of the stress response (Bilang-Bleuel et al., 2002; Chandramohan et al., 2007, in press). Thus, our observations show that the increase in phospho-acetylation in dentate neurons after psychological stress results in a specific gene transcription response. Moreover, it appears

ARTICLE IN PRESS Epigenetics and stress-related memory that in dentate neurons the part of the chromatin containing the c-fos gene is normally inactive and requires PAc-H3driven remodeling to become accessible for transcription. Within the dentate gyrus the PAc-H3 and c-fos response was neuroanatomically very specific as it only occurred in (NeuN-positive, i.e. adult) neurons in the middle and superficial aspects of the granular cell layer of the dorsal blade (Bilang-Bleuel et al., 2005; Chandramohan et al., 2007, in press). According to Wang et al. these neurons are morphologically and electrophysiologically more mature than those in the deep aspects, close to the subgranular zone (Wang et al., 2000).

5. Signaling mechanisms underlying psychological stress-induced chromatin remodeling and transcriptional activation in the brain We were able to show that the enhanced histone H3 phospho-acetylation and c-fos induction after novelty and forced swim stress was elicited by concomitant activation of two signaling pathways being the glucocorticoid receptor (GR) and NMDA-R (Bilang-Bleuel et al., 2005; Chandramohan et al., 2007, in press). Inhibition of either signaling pathway blocked the forced swimming- and novelty-induced histone modifications and c-fos. The necessity for dual activation of pathways was further underscored by the observation that injection of rats with a GR-occupying dose of corticosterone

S23 was ineffective (Chandramohan et al., 2007, in press). Furthermore, inhibition of nitric oxide synthesis or blocking the mineralocorticoid receptor (MR) were also ineffective excluding a role of these signaling molecules (Chandramohan et al., 2007, in press). Recently, we studied the signaling pathway(s) downstream of the NMDA-R involved in the forced swimminginduced histone H3 phospho-acetylation response in the dentate gyrus (Chandramohan et al., 2006a, b) (see Fig. 1). Inhibition of ERK activation by using the MAPK-ERK kinase (MEK) inhibitor SL327 blocked the forced swimming-induced increase in PAc-H3. In a next step we investigated whether the ERK-activated histone H3 kinases MSK1/2 play a role in the forced swimming-induced histone modifications. Indeed, double genetic deletion of MSK1/2 in mutant mice (MSK1/ 2 / ; collaboration with Dr. S. Arthur, University of Dundee, UK) completely abolished the forced swimming-induced increase in histone H3 phospho-acetylation (Chandramohan et al., 2006a, b). No impairment was observed in dentate Ac-H3 levels indicating no defects in this aspect of gene transcription. Importantly, the MSK1/2 / mice also showed no increase in dentate c-fos expression after forced swimming. However, the c-fos expression pattern outside the dentate gyrus (e.g. PVN, neocortex) was comparable in MSK1/2 / and wild-type mice corresponding with our observation that c-fos induction in extra-dentate regions is not PAc-H3-driven and, based on our MSK1/2 deletion experiments, also does not require MSK activation (Chandramohan et al., 2006a, b).

Figure 1 Proposed dual—GR- and MAPK/ERK/MSK-mediated—signaling cascades activated by forced swimming leading to the phospho-acetylation of histone H3 and transcriptional induction of c-fos in a distinct population of mature dentate granule neurons, and the encoding of memory of the stressful event. The functional consequences of this cognitive process can be seen in the re-test when the animal presents enhanced immobility behavior which is regarded as an appropriate adaptive response. PKA and CAMKII are drawn in as they are well-known CREB kinases which are activated by Ca2+-dependent mechanisms (PKA indirectly via activation of ACy (adenylate cyclases 1 and 8)) and which are involved in neuroplasticity mechanisms underlying learning and memory. ‘P’ in star: phosphorylation (only shown for MSK), ‘Ac’ in plaque: acetylation. For other abbreviations, see text.

ARTICLE IN PRESS S24 In the light of these data, we postulate that the histone H3 phospho-acetylation response after psychological stress is evoked by concurrent activation of the GR and the NMDA/ MAPK/ERK/MSK signaling pathways (Fig. 1). This model corresponds with the model on the in vitro mouse mammary tumor virus (MMTV) promoter induction by progestins recently proposed by Vicent et al. (2006). They showed that the activated progesterone receptor (a steroid receptor very similar to GR) binds phosphorylated ERK (P-ERK) which then binds MSK. In a next step, MSK is phosphorylated (i.e. activated) by P-ERK and the whole complex is recruited to the promoter where P-MSK phosphorylates the histone H3 tails at Ser10 which facilitates the recruitment of an ATPdependent chromatin remodeling complex to prepare gene transcription (Vicent et al., 2006). Thus, our in vivo model (Fig. 1) on the interaction between GR and MAPK/ERK/MSK signaling in histone H3 phosphorylation and subsequent acetylation in dentate neurons is strongly supported by this molecular framework based on in vitro work. The signaling mechanisms underlying the acetylation of phosphorylated histone H3 are presently still unclear. It may be speculated, however, that given that MSK is a potent CREB kinase and CREB phosphorylation is indeed taking place in dentate neurons after psychological stressors such as forced swimming and novelty (Bilang-Bleuel et al., 2002), under such circumstances P-CREB may recruit one of the coactivator CREB-binding proteins (e.g. CBP, P300, pCAF) whose histone acetyl transferase (HAT) activity has been shown, at least in vitro, to acetylate histone H3 at Lys14 (Schiltz et al., 1999; Li et al., 2003; Fig. 1).

6. Significance of PAc-H3-driven transcriptional induction for encoding of stress-associated memory We have been able to accumulate evidence that the biochemical responses in dentate neurons after forced swimming are of functional significance. When rats or mice are forced to swim in a basin from which they cannot escape, they show in a re-test 24 h later a characteristic behavioral response: they quickly stop their attempts to escape from the water and retain an immobile posture for about 70% of the 5 min re-test. The animals show this acquired immobility response because they have learned from their experience 24 h earlier that escape from the water is not possible. Therefore, the immobility response shown 24 h after the initial test can be regarded as an adaptive learning and memory response with the experimental advantage that it fully develops overnight (De Pablo et al., 1989; Korte, 2001; Bilang-Bleuel et al., 2005). The immobility response observed in the re-test is strongly dependent of GR and NMDA-R signaling evoked by the initial forced swim test, thus for the acquisition (i.e. encoding of memory, but not retrieval) of the learned immobility response (Korte, 2001; Padovan and Guimaraes, 2004). Glucocorticoids have been shown to reverse the adrenalectomy-induced impairment of the immobility response in the re-test if the hormone was given between 15 min and 1 h, but not 4 h, after the initial swim test (for review, see Korte, 2001). It is of interest to note that the forced swimming-induced enhancement of histone H3 phospho-

J.M.H.M. Reul, Y. Chandramohan

Table 1 Role of chromatin remodeling in dentate granule neurons in stress-related learning and memory— correlation between forced swimming-induced histone H3 phospho-acetylation and the acquisition of behavioral immobility as measured in the re-test. Manipulation

FS-induced PAc-H3

Immobility in re-test

NMDA-R blockade GR blockade MR blockade MEK inhibition MSK1/2 genetic deletion Water temperature in test

Blocked Blocked No effect Blocked Blocked Dependency

Impaired Impaired No effect Impaired Impaired Dependency

Data are reported in Bilang-Bleuel et al. (2005) and Chandramohan et al. (2006a, b). Note: No effects on behavior were observed during the initial test. Regarding the manipulation ‘Water temperature in Test’, this reflects findings published in Bilang-Bleuel et al. (2005) in which it was reported that a lowering of the water temperature in the initial test results in a further enhancement of histone H3 phosphorylation and phospho-acetylation and an increased immobility in the retest (as compared to data obtained at the ‘normal’ water temperature (i.e. 25 1C)). FS, forced swimming; for other abbreviations, see text.

acetylation in the dentate gyrus develops transiently within 4 h after the swim test (Chandramohan et al., 2006a, b), thus within the critical time period of the encoding of memory of the stressful event. Moreover, in terms of neuroanatomical specificity, impairment of immobility has been observed after injection of GR antagonists or antisense against GR into the dentate gyrus but not in the nucleus parafascicularis or PVN, pointing to the dentate gyrus as a site of action of the forced swimming-induced glucocorticoid hormones (Korte et al., 1996). We found that not only blocking GRs and NMDA-Rs but also inhibition of ERK activation and genetic deletion of MSK1/2 resulted in impairment of immobility behavior in the re-test (Table 1) (Bilang-Bleuel et al., 2005; Chandramohan et al., 2006a, b). Importantly, none of these conditions affected behavior during the initial test. Thus, during the course of our studies, it has become increasingly clear that the chromatin remodeling response and c-fos induction observed in dentate neurons after the initial forced swim test is of critical importance for the acquisition of the behavioral immobility response which can be tested 24 h later in a retest (Fig. 1, Table 1) (Bilang-Bleuel et al., 2005; Chandramohan et al., 2006a, b).

7. Conclusion Psychologically stressful events evoke a distinct gene transcriptional response in a select population of mature dentate granule neurons which requires PAc-H3-driven chromatin remodeling. These epigenetic processes necessi-

ARTICLE IN PRESS Epigenetics and stress-related memory tate concurrent signaling via the GR and the NMDA/MAPK/ ERK/MSK pathways. As illustrated by the acquisition of forced swimming-induced immobility behavior, these molecular mechanisms seem to occur in neurons involved in the encoding of memory of the stressful event.

Role of the funding sources Funding for the authors’ studies described in this review was provided by the Medical Research Council (MRC; Grant G78/ 8083), the Max Planck Society (MPG; intramural funding) and the Neuroendocrinology Charitable Trust (NCT; Grant PMS/ MMS-06/07-10080); the MRC, MPG and NCT had no further role in the study design; in the collection, analysis and interpretation of data; in the writing of the report; and in the decision to submit the paper for publication. University of Lille 1, USTL (France) contributed with editorial assistance, reviewed drafts of the manuscript and contributed to the decision to submit the manuscript for publication. The authors retained full editorial control and responsibilities throughout the preparation of the manuscripts.

Conflict of interest None declared.

Acknowledgments This supplement is based upon the proceedings from Lille Summer School in Neurosciences—Brain Plasticity in Life Span held in France in 2006. The supplement is supported financially by University of Lille 1, USTL (France), Servier (France) and Nestle ´ (Switzerland).

References Bilang-Bleuel, A., Rech, J., Holsboer, F., Reul, J.M.H.M., 2002. Forced swimming evokes a biphasic response in CREB phosphorylation in extrahypothalamic limbic and neocortical brain structures. Eur. J. Neurosci. 15, 1048–1060. Bilang-Bleuel, A., Ulbricht, S., Chandramohan, Y., De Carli, S., Droste, S.K., Reul, J.M.H.M., 2005. Psychological stress increases histone H3 phosphorylation in adult dentate gyrus granule neurons: involvement in a glucocorticoid receptor-dependent behavioural response. Eur. J. Neurosci. 22, 1691–1700. Chandramohan, Y., Droste, S.K., Arthur, J.S., Reul, J.M.H.M., 2006a. Forced swimming-induced behavioural immobility involves chromatin remodelling in dentate granule neurons via recruitment of the NMDA/MAPK/ERK/MSK pathway. In: Proceedings of the Focused Meeting of The Physiological Society ‘New Developments in Stress Physiology—From Gene to Man’, Bristol, 4–5 December 2006, No. PC2. Chandramohan, Y., Droste, S.K., Arthur, J.S., Reul, J.M.H.M., 2006b. The forced swimming-induced behavioral immobility response involves phospho-acetylation of histone H3 in dentate gyrus neurons via activation of the NMDA/MAPK/ERK/MSK pathway. Society for Neuroscience Abstracts No. 59.3. Chandramohan, Y., Droste, S.K., Reul, J.M.H.M., 2007. Novelty stress induce phospho-acetylation of histone H3 in rat dentate gyrus granule neurons through coincident signaling via the NMDA

S25 receptor and the glucocorticoid receptor: relevance for c-fos induction. J. Neurochem. 101, 815–828. Cheung, P., Tanner, K.G., Cheung, W.L., Sassone-Corsi, P., Denu, J.M., Allis, C.D., 2000. Synergistic coupling of histone H3 phosphorylation and acetylation in response to epidermal growth factor stimulation. Mol. Cell 5, 905–915. Clayton, A.L., Rose, S., Barratt, M.J., Mahadevan, L.C., 2000. Phosphoacetylation of histone H3 on c-fos-and c-jun-associated nucleosomes upon gene activation. EMBO J. 19, 3714–3726. De Pablo, J.M., Parra, A., Segovia, S., Guillamo ´n, A., 1989. Learned immobility explains the behavior of rats in the forced swim test. Physiol. Behav. 46, 229–237. Impey, S., Obrietan, K., Storm, D.R., 1999. Making new connections: role of ERK/MAP kinase signaling in neuronal plasticity. Neuron 23, 11–14. Jaskelioff, M., Peterson, C.L., 2003. Chromatin and transcription: histones continue to make their marks. Nat. Cell Biol. 5, 395–399. Korte, S.M., 2001. Corticosteroids in relation to fear, anxiety and psychopathology. Neurosci. Biobehav. Rev. 25, 117–142. Korte, S.M., De Kloet, E.R., Buwalda, B., Bouman, S.D., Bohus, B., 1996. Antisense to the glucocorticoid receptor in hippocampal dentate gyrus reduces immobility in forced swim test. Eur. J. Pharmacol. 301, 19–25. Li, X., Wong, J., Tsai, S.Y., Tsai, M.J., O’Malley, B.W., 2003. Progesterone and glucocorticoid receptors recruit distinct coactivator complexes and promote distinct patterns of local chromatin modification. Mol. Cell Biol. 23, 3763–3773. Nowak, S.J., Corces, V.G., 2000. Phosphorylation of histone H3 correlates with transcriptionally active loci. Genes Dev. 14, 3003–3013. Oitzl, M.S., De Kloet, E.R., 1992. Selective corticosteroid antagonists modulate specific aspects of spatial orientation learning. Behav. Neurosci. 106, 62–71. Padovan, C.M., Guimaraes, F.S., 2004. Antidepressant-like effects of NMDA-receptor antagonist injected into the dorsal hippocampus of rats. Pharmacol. Biochem. Behav. 77, 15–19. Roozendaal, B., Hui, G.K., Hui, I.R., Berlau, D.J., McGaugh, J.L., Weinberger, N.M., 2006. Basolateral amygdala noradrenergic activity mediates corticosterone-induced enhancement of auditory fear conditioning. Neurobiol. Learn. Memory 86, 249–255. Schiltz, R.L., Mizzen, C.A., Vassilev, A., Cook, R.G., Allis, C.D., Nakatani, Y., 1999. Overlapping but distinct patterns of histone acetylation by the human coactivators p300 and PCAF within nucleosomal substrates. J. Biol. Chem. 274, 1189–1192. Soloaga, A., Thomson, S., Wiggin, G.R., Rampersaud, N., Dyson, M.H., Hazzalin, C.A., Mahadevan, L.C., Arthur, J.S., 2003. MSK2 and MSK1 mediate the mitogen- and stress-induced phosphorylation of histone H3 and HMG-14. EMBO J. 22, 2788–2797. Spencer, V.A., Davie, J.R., 1999. Role of covalent modifications of histones in regulating gene expression. Gene 240, 1–12. Strahl, B.D., Allis, C.D., 2000. The language of covalent histone modifications. Nature 403, 41–45. Vicent, G.P., Ballare, C., Silvina Nacht, A., Clausell, J., SubtilRodriquez, A., Quiles, I., Jordan, A., Beato, M., 2006. Induction of progesterone target genes requires activation of Erk and Msk kinases and phosphorylation of histone H3. Mol. Cell 24, 367–381. Wang, S., Scott, B.W., Wojtowicz, J.M., 2000. Heterogenous properties of dentate granule neurons in the adult rat. J. Neurobiol. 42, 248–257. Wiggin, G.R., Soloaga, A., Foster, J.M., Murray-Tait, V., Cohen, P., Arthur, J.S., 2002. MSK1 and MSK2 are required for the mitogenand stress-induced phosphorylation of CREB and ATF1 in fibroblasts. Mol. Cell Biol. 22, 2871–2881.