Behavioural Brain Research 192 (2008) 12–19

Contents lists available at ScienceDirect

Behavioural Brain Research journal homepage: www.elsevier.com/locate/bbr

Review

A role for the spine apparatus in LTP and spatial learning Peter Jedlicka 1 , Andreas Vlachos 1 , Stephan W. Schwarzacher, Thomas Deller ∗ Institute of Clinical Neuroanatomy, J.W. Goethe-University of Frankfurt, Theodor-Stern-Kai 7, D-60590 Frankfurt am Main, Germany

a r t i c l e

i n f o

Article history: Received 23 November 2007 Received in revised form 16 February 2008 Accepted 18 February 2008 Available online 26 February 2008 Keywords: Endoplasmic reticulum Ca2+ -store Protein synthesis Plasticity Hippocampus Synaptopodin

a b s t r a c t Long-term potentiation (LTP) of synaptic strength is a long-lasting form of synaptic plasticity that has been linked to information storage. Although the molecular and cellular events underlying LTP are not yet fully understood, it is generally accepted that changes in dendritic spine calcium levels as well as local protein synthesis play a central role. These two processes may be influenced by the presence of a spine apparatus, a distinct neuronal organelle found in a subpopulation of telencephalic spines. Mice lacking spine apparatuses (synaptopodin-deficient mice) show deficits in LTP and impaired spatial learning supporting the involvement of the spine apparatus in synaptic plasticity. In our review, we consider the possible roles of the spine apparatus in LTP1 (protein synthesis-independent), LTP2 (translation-dependent and transcription-independent) and LTP3 (translation- and transcription-dependent) and discuss the effects of the spine apparatus on learning and memory. © 2008 Elsevier B.V. All rights reserved.

Contents 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The actin-binding protein synaptopodin (SP) is localized to the spine apparatus and essential for its formation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The lamina-specific distribution of SP in the hippocampus depends on afferent activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A role for the spine apparatus in hippocampal synaptic plasticity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mice devoid of the spine apparatus show deficits in spatial learning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The spine apparatus and calcium kinetics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Calcium, the spine apparatus and LTP induction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dendritic protein synthesis, LTP and the spine apparatus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Long-lasting forms of LTP—a role for SP and the spine apparatus? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1. Introduction Long-term potentiation (LTP) is a long-lasting form of synaptic change that has been linked to learning and memory. Although LTP itself has been described more than 30 years ago [13], direct evidence for its association with memory-related mechanisms has been provided only recently (see commentary in Ref. [14]). By the same token, experiments in the hippocampus have demonstrated that plastic changes of synaptic transmission underlie both LTP and hippocampus-dependent memory [31,52,77].

∗ Corresponding author. Tel.: +49 69 6301 6900; fax: +49 69 6301 6425. E-mail address: [email protected] (T. Deller). 1 These authors contributed equally to this work. 0166-4328/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.bbr.2008.02.033

12 13 13 13 15 15 16 17 17 17 18 18

Based on duration and biochemical mechanisms, LTP has been subdivided into three distinct phases that have been termed LTP1, 2, and 3, or early (E)-, intermediate (I)-, and late (L)-LTP [1,12,57,60]. LTP1 is a short-lasting (1 h) early form of LTP that requires post-translational modification of synaptic proteins but is protein synthesis-independent. LTP2 is slowly decaying (1–3 h) and dependent on protein translation but does not require gene transcription. Finally, LTP3 represents the long-lasting phase of LTP (hours, weeks) and is both translation- and transcription-dependent. Since LTP causes functional as well as structural changes, it can be regarded as a plastic process par excellence [51]. In the central nervous system, excitatory inputs typically terminate on dendritic spines. It is well known that afferent synaptic activity regulates the morphology of spines and that morphological plasticity of spines, in turn, contributes to the changes in synaptic

P. Jedlicka et al. / Behavioural Brain Research 192 (2008) 12–19

transmission (for recent reviews see Refs. [45,63,73]). Therefore, spines are seen as a major site of functional and structural plasticity in the brain [23,32,33]. Interestingly, a subset of dendritic spines in the cerebral cortex contains a distinct organelle known as the spine apparatus [17,19,30,66]. The spine apparatus belongs to the continuous endoplasmic reticulum network in telencephalic neurons [10] and is composed of stacks of smooth endoplasmic reticulum interdigitated by densely stained material. Since its first description in 1959 by Gray [30] it has been suggested to play a role in processes linked to synaptic plasticity, although its mechanistic role in this context remained elusive. More recent data have implicated the spine apparatus in the regulation of spine calcium kinetics [24,37,64] and in the post-translational modification and transport of locally synthesized proteins [54,55,68,72]. Since both of these processes have strong influences on LTP, they provide possible links between the spine apparatus and LTP. In this review we will, therefore, consider the possible roles of the spine apparatus in molecular mechanisms underlying LTP1, 2 and 3 and will discuss its effect on learning and memory.

2. The actin-binding protein synaptopodin (SP) is localized to the spine apparatus and essential for its formation The spine apparatus organelle can be identified using immunolabeling for the actin-associated protein synaptopodin (SP). This molecule has been characterized a decade ago by Mundel and coworkers [3,46] who described it in kidney and in brain. In brain, SP is found in a 100 kD form and expressed by neurons located in the telencephalon. Although initially believed to be linked with the postsynaptic density [46], later studies demonstrated that SP is tightly associated with the spine apparatus organelle and essential for its formation [17,19]. Using SP antibodies to label the spine apparatus, it thus became possible to study the distribution of this enigmatic organelle in brain and in particular in the hippocampus with light- and electron microscopic techniques [6,17,18,19]. Light microscopic analyses revealed SP-immunoreactive puncta primarily in the dendritic layers of the hippocampus (but see also Ref. [7], who described an association of SP with the cisternal organelle of the axon initial segment). These puncta were inhomogeneously distributed and followed a characteristic region- and layer-specific distribution pattern; even along the septo-temporal axis of the hippocampus, where higher SP-densities were found in the ventral hippocampus compared to its dorsal counterpart (Fig. 1a; [6,17,18,75]). Electron microscopy, in turn, demonstrated that SP puncta are primarily located within dendritic spines and only rarely associated with dendritic shafts of hippocampal neurons [17–19]. Within spines, SP was most abundant in the spine neck, where it was regularly associated with the spine apparatus organelle (Fig. 2a–c). Neither spines lacking a spine apparatus nor axon terminals forming contacts with spines were immunoreactive for SP [17]. To study the percentage of SP-positive puncta in dendrites and spines and to determine the percentage of spines containing a SP-positive spine apparatus organelle, SPimmunostained hippocampal sections of eGFP-mice were used for three-dimensional confocal reconstructions of identified hippocampal neurons (Fig. 1b and c; [6]). As suggested in the electron microscopic preparations, SP puncta were primarily found in dendritic spines (>95%) and only rarely in dendritic shafts. In line with the earlier observations suggesting a lamina-specific distribution of SP in the hippocampus, the analysis of eGFP-labeled and SP-immunostained dendritic segments also revealed a laminaspecific sorting of SP in identified hippocampal neurons [6]. Thus, SP-positive spine apparatuses are distributed in a region- and layerspecific fashion in the rodent hippocampus [6].

13

To demonstrate that SP is not only associated with but also critical for the formation of a spine apparatus organelle, mice homozygous for a targeted deletion of the SP gene were generated and analyzed [19]. SP-deficient mice were viable, did not show a pathological kidney phenotype under physiological conditions [3], and exhibited normal brains [19]. No pathology was observed at the light microscopic level. An extensive electron microscopic analysis performed in the striatum, cortex, and hippocampus, however, revealed that telencephalic neurons of these mice do not form spine apparatuses (Fig. 2d). In these animals, spine morphology and density appeared to be normal, suggesting that SP may act within spines to form a cytoskeletal scaffold for the spine apparatus organelle. Although this mechanism is yet hypothetical and requires further analysis, it is clear that SP is localized to the spine apparatus and plays an essential role for its formation [19]. 3. The lamina-specific distribution of SP in the hippocampus depends on afferent activity The major glutamatergic afferents to the hippocampus terminate in a highly laminated manner on the dendrites of the principal neurons. The observation that the distribution of the postsynaptic molecule SP respects the laminar organization of the hippocampus indicates that SP is sorted within hippocampal dendrites in response to afferent synaptic activity. This interpretation is in line with several studies that showed activity-dependent alterations in the distribution of SP in the hippocampus: Following induction of LTP in vivo SP-immunolabeling was increased in the stimulated laminae [28,80] and after partial denervation of granule cells or excitotoxic lesion of the hippocampus a loss of SP-immunostaining was reported [20,62]. Interestingly, upregulation of SP in stimulated fiber layers coincided with the upregulation of other cytoskeletal molecules, notably actin, in the same layers, suggesting that actin and SP play an important role in spine plasticity under conditions of synaptic strengthening [28]. 4. A role for the spine apparatus in hippocampal synaptic plasticity SP-deficient mice represent a suitable animal model for studying the involvement of the spine apparatus and SP in the modulation of activity-dependent synaptic plasticity. Indeed, LTP has been studied at Schaffer collateral-CA1 synapses in SP-deficient mice in vitro [19]. Acute slices derived from SP-deficient mice exhibited a decrease in LTP of the field excitatory postsynaptic potential (fEPSP) slope with respect to wildtype mice. Within 40 min LTP became significantly impaired, indicating that LTP1 was altered (Fig. 3a and b). Three hours after stimulation, LTP was still clearly disrupted in mutant mice, suggesting that LTP2 and LTP3 are also affected. The impairment of LTP in SP-deficient mice implies that SP and the spine apparatus are involved in the regulation of long-term synaptic plasticity and memory. However, while the association between SP, the spine apparatus and synaptic plasticity is well established, a recent paper demonstrated that there is no simple interrelationship between SP-densities and LTP in the hippocampus [75]: In this study rats were exposed to acute swim stress and LTP was measured in the ventral and dorsal hippocampus. In this behavioural situation, LTP was enhanced in the ventral hippocampus and suppressed in the dorsal hippocampus (cf. Refs. [41,42]). Although changes in SP-density were induced (see below), these changes did not mirror the functional changes, suggesting that the functional role of SP in the behaving intact animal is more complex than hitherto assumed.

14

P. Jedlicka et al. / Behavioural Brain Research 192 (2008) 12–19

Fig. 1. Distribution of synaptopodin in the hippocampus. Synaptopodin (SP) is an actin-binding protein found in a subset of telencephalic dendritic spines. It is an essential component of the spine apparatus organelle. (a) Frontal section of the rat hippocampus stained for SP (red) demonstrating the layer- and region-specific distribution of SP. The connectivity between DG, CA3 and CA1 is depicted in a schematic manner. DG, dentate gyrus; gcl, granule cell layer; iml, inner molecular layer; l-m, stratum lacunosummoleculare; o, stratum oriens; oml, outer molecular layer; pcl, pyramidal cell layer; rad, stratum radiatum. Scale bar: 500 ␮m. (b and c) Dendritic segments in area CA1 of eGFP-transgenic mice stained for SP (red; b, stratum radiatum; c, stratum lacunosum-moleculare). The arrows and arrowheads indicate SP-positive structures located within a spine head or a spine neck, respectively. SP puncta can be also found at the base of spines (asterisk) suggesting that SP might enter and leave spines and that the exact positioning within dendritic spines could play a functional role (asterisk). Scale bars: 2 ␮m.

Fig. 2. SP as an essential component of the spine apparatus organelle. (a) An electron micrograph of a spine apparatus (arrow) within the neck of a dendritic spine. The spine apparatus is a complex membranous structure extending from the endoplasmic reticulum of the dendrite into the spine. It is composed of stacks of smooth endoplasmic reticulum interdigitated by densely stained material. A schematic representation of structures seen in the electron micrograph is illustrated in the inset (top right; dendrite = gray, dendritic spine = yellow, mitochondrium = blue, presynaptic terminal = orange, postsynaptic density = black, spine apparatus = green). Scale bar: 1 ␮m. (b) An electron micrograph of SP-immunoreactive spines located in the outer molecular layer of the normal mouse dentate gyrus. SP immunoreactivity is strongest in the spine neck and closely associated with the spine apparatus (arrow). The smooth endoplasmic reticulum cisternae of the spine apparatus are heavily labeled with DABimmunoprecipitate. (c) Preembedding-immunogold labeling for SP confirms the close association of SP with the spine apparatus organelle. The arrowhead points to the synaptic cleft. (d) Absence of the spine apparatus in a hippocampal neuron from a SP-deficient mouse. Scale bars in b–d: 0.2 ␮m. Adapted from Ref. [18] and Ref. [19].

P. Jedlicka et al. / Behavioural Brain Research 192 (2008) 12–19

15

Fig. 3. Changes in LTP and learning behaviour in SP-deficient mice. (a) Group data for field EPSP recordings before and after theta-burst (TBS; 100 Hz) application. The difference between mutant and wild-type mice is significant (p < 0.01; T-test, two-sided). Error bars: S.E.M., n = number of slices. (b) Group data for field EPSP recordings before and 3 hours after TBS (100 Hz) application. Late LTP is also affected in SP-deficient mice. The difference between mutant and wild-type mice is significant (p < 0.05; T-test). Error bars: S.E.M., n = number of slices. Only slices that showed E-LTP were included in the analysis. (c) Locomotor activity in the open field. Bars indicate mean numbers (+S.E.M.) of line crossings. Mutant mice (n = 9) show less locomotor activity than wild-type controls (n = 10). (d–f) Anxiety related behavior in the elevated plus maze. Frequency of closed (d) and all (e) arm entries (mean + S.E.M.). The ratio of entries into closed arms versus all arms is illustrated in (f). Note that mutant mice (n = 9) are less anxious than wild-type controls (n = 10), and that this difference is not caused by reduced locomotor activity. (g) Spatial learning in the radial arm maze. Mean numbers (±S.E.M.) of spatial working memory errors of wild-type (n = 16; dots) and mutant mice (n = 14; triangles) during the 5-day training period. From day 3 on, mutant mice show significantly more failures than wild-type mice (* p < 0.05). Adapted from Ref. [19].

5. Mice devoid of the spine apparatus show deficits in spatial learning The impairment of CA3-CA1 LTP in SP-deficient mice suggested that the lack of SP and the spine apparatus might lead to changes on the behavioural level. Therefore, a variety of behavioural tests were performed with SP-deficient animals [19]. These mice exhibited a decreased horizontal locomotor activity in the open field test (Fig. 3c). Furthermore, knockout mice were less anxious than wildtype controls as demonstrated by monitoring their behaviour in the elevated plus maze (Fig. 3d–f). Finally and most relevant for the correlation with electrophysiological data from the hippocampus, the radial arm maze test revealed that SP-mice have a decreased ability to acquire LTP3-related spatial memory (Fig. 3g [19]). Spatial learning of transgenic mice was impaired after the third day of training as shown by a significantly increased error rate. The differences between genotypes were most prominent on the fifth training day with the wild-type mice making twice less error than SP-deficient mice. Thus, the behavioural phenotype of mutants correlated with the electrophysiological data indicating that SP and the spine apparatus are involved in the regulation of long-term synaptic plasticity and memory.

Although the above data provided evidence for a link between SP, LTP and learning and memory, it remained to be shown that behaviour could, in turn, influence the cellular distribution of SP. This was recently demonstrated in the hippocampus of rats exposed to acute swim stress [75]. These animals showed increased SPdensities in the dorsal hippocampus within 60 min, indicating that the adaptive response of hippocampal neurons to this stress situation includes changes in SP-expression or distribution. As has been pointed out in the previous paragraph, however, the relationship between SP-changes and behaviour is complex, since the behaviourally induced changes in SP-densities did not correspond to the electrophysiological changes recorded in the same brains. Combined behavioural and electrophysiological studies in awake animals will be required to unravel the link between behaviour, neuronal activity and changes in SP-expression in the hippocampus.

6. The spine apparatus and calcium kinetics What might be the biological significance of the spine apparatus for synaptic plasticity? Studies analyzing the role of calcium

16

P. Jedlicka et al. / Behavioural Brain Research 192 (2008) 12–19

in spines have implicated the spine apparatus in local calcium trafficking [24,37,64]. Neurons contain a complex endoplasmic reticulum network that consists of a continuous membrane system and is involved in neuronal calcium signalling [10]. This network of tubules and cisterns extends from the soma throughout the dendritic tree, reaching into necks of many large spines [67]. Thus, the spine apparatus can be considered a regional specialization of the smooth endoplasmic reticulum [11]. One of the primary functions of the endoplasmic reticulum organelle is to release calcium through either inositol 1,4,5trisphosphate (IP3 ) or ryanodine receptors (RyRs) in response to input signals [74]. Both IP3 Rs and RyRs are calcium-sensitive and therefore capable of regenerative calcium-induced calcium release (CICR). In addition to functioning as an internal calcium source, the endoplasmic reticulum can serve as a calcium buffer as well. Sarco(endo)plasmic reticulum calcium-ATPases (SERCA) mediate rapid calcium uptake into the lumen of the endoplasmic reticulum, thereby both maintaining the internal calcium stores and sequestering free cytosolic calcium [78]. As a sub-compartment of the endoplasmic reticulum, the spine apparatus may act as a calcium source and/or a calcium sink in dendritic spines [24,25]. However, no functional experiments have been provided yet to test this hypothesis and it remains unclear in which respect the spine apparatus differs from “regular” smooth endoplasmic reticulum which has been shown to act as a calcium store within dendritic spines.

7. Calcium, the spine apparatus and LTP induction Spine calcium plays a critical role in the induction of synaptic plasticity [12]. The primary source of calcium influx during hippocampal LTP-induction is glutamate-mediated opening of postsynaptic NMDA receptors (NMDARs). Now, an important question to ask is: do other sources of calcium play a significant role, too? Particularly, does the release of calcium from internal stores in spines contribute to plasticity-related calcium pulses evoked by synaptic stimulation? Interestingly, in the hippocampus RyRs are distributed throughout the endoplasmic reticulum in spines and dendrites while IP3 Rs exist mainly in dendritic shafts [37,64]. Calcium imaging experiments in cultured hippocampal neurons and organotypic slices suggested that CICR from internal calcium stores was responsible for a large part of synaptically induced calcium transients [22,37]. In these studies, spine calcium signals could be reduced by antagonists of CICR: Ryanodine which blocks RyRs; cyclopiazonic acid (CPA) and thapsigargin (TG) which inhibit SERCA pumps and deplete calcium stores. However, other experiments in acute hippocampal slices indicated that the contribution of RyR-mediated calcium release to synaptic calcium increase was not significant [38,43, see also 36,82]. These studies were hard to reconcile and could not solve the controversy about the source of synaptic calcium (see commentaries in Refs. [26,61,71,74]). Experiments in RyR-deficient mice were also not able to provide conclusive data to this debate [5,29,65]. Recent research has shed new light on the role of internal calcium stores in long-term synaptic plasticity. Pharmacological experiments in the CA1 region of the hippocampus have demonstrated that inhibition of RyRs, IP3 Rs and L-type voltage-dependent calcium channels (L-VDCCs) causes selective inhibition of LTP1, LTP2 and LTP3, respectively [58, see also 8,34,76]. Most importantly, the remarkable specificity of different calcium sources in induction of the three LTP forms could be attributed to spatial compartmentalization of calcium signals in a recent study using two-photon calcium imaging [59]. Weak tetanic stimulation (1 train of theta-

Fig. 4. A schematic illustrating the role of the spine apparatus in LTP. The spine apparatus is likely involved in two LTP-related processes: regulation of local calcium kinetics via store-operated calcium release and post-translational modification and transport of locally synthesized proteins. Spine apparatuses can be found mainly in spine necks, but also in the base of spines and at the proximal end of the spine heads. Their strategic position within the dendritic spine neck suggests a functional impact on segregating and/or propagating of calcium waves which enter the spine head via NMDA receptors (NMDARs). Thus, the exact positioning of the spine apparatus might be crucial for its function (indicated by the black double-headed arrow). While the endoplasmic reticulum in dendritic shafts is covered with both ryanodine receptors (RyRs) and inositol 1,4,5-trisphosphate receptors (IP3 Rs), the spine apparatus contains mainly ryanodine receptors (RyRs). Sarco(endo)plasmic reticulum calcium-ATPases (SERCA) might mediate rapid calcium uptake into the spine apparatus, thereby both maintaining the calcium store and sequestering free cytosolic calcium. Circumstantial evidence exists that the spine apparatus may be important for local modification and translation of proteins (see text for details). Inset: the eGFP-labeled dendritic spine (green; marked with an asterisk) immunostained for SP (red) served as a template for the artwork; scale bar: 1 ␮m. AMPARs, AMPA receptors; mGluRs, metabotropic glutamate receptors; L-VDCC, L-type voltage-dependent calcium channels.

burst stimulation, TBS) of Schaffer collaterals induced rapidly decaying RyRs-dependent LTP1 and generated RyRs-mediated calcium signals in dendritic spines of CA1 pyramidal neurons. These data suggest that TBS activates NMDAR-mediated calcium influx which triggers CICR. Stronger stimulation (4 TBS trains) resulted in LTP of intermediate duration (LTP2) and selectively increased calcium concentration in dendrites. These effects could be inhibited by blockage of IP3 Rs. Finally, strong stimulation (eight TBS trains) induced persistent form of LTP (LTP3) depending on somatic calcium influx through L-VDCCs. From these findings an elegant spatial encoding (‘compartmental’) model for the induction of LTP emerges [59,60]: spatially discrete calcium sources seem to provide the molecular basis for mechanistically distinct forms of LTP with different durations. Calcium as a single second messenger is capable of selectively triggering multiple downstream signalling cascades due to compartmentalization of calcium sources and possibly also calcium sensors. It has been suggested that calcium sensors underlying various LTP forms are co-localized with respective calcium sources. As already mentioned, in CA1 the spine apparatus is covered mainly with RyRs whereas the endoplasmic reticulum in dendritic shafts

P. Jedlicka et al. / Behavioural Brain Research 192 (2008) 12–19

contains both RyRs and IP3 Rs ([64], see Fig. 4). Intriguingly, the RyRs-bearing spine apparatus in CA1 spines is in close proximity with NMDARs and the postsynaptic density and therefore may constitute a functional calcium microdomain similar to ‘calcium release units’ of cardiac muscle [59,71]. Given the immunohistological distribution of calcium releaserelated receptors and the early LTP impairment in CA1 of SP-deficient mice [19], it seems that the spine apparatus is predominantly involved in LTP1 shaping mainly the calcium transients confined to dendritic spines. (We discuss its possible role in modulating LTP2 and LTP3 below.) The release of calcium from the spine apparatus might provide highly compartmentalized calcium signals for the induction of input-specific synaptic changes [10]. The spine apparatus could function as a modular calcium signalling unit [11] regulating synaptically evoked spine calcium transients independently of calcium pulses at neighbouring synapses and spines. On the other hand, IP3 Rs on the endoplasmic reticulum in dendritic shafts might act as coincident receptors to integrate information arriving from separate synaptic inputs [9]. In other words, in hippocampal neurons RyRs on the spine apparatus seem to modulate calcium signals within the spine whereas IP3 Rs on the dendritic endoplasmic reticulum appear to be involved in the transfer of information from one spine to adjacent spines [11,47]. Obviously, the distribution pattern of the spine apparatus and its exact location within individual spines would play an important role in this context. From a functional and anatomical point of view, the strategic positioning of spine apparatuses within spine necks makes the idea of compartmentalization favourable. Since the exact placement of the organelle can vary between different spines (base of spines, neck of spines, proximal part of spine head, see Fig. 1b and c), shifts in the location of the spine apparatus might have a functional impact on segregating and shaping of calcium waves, thus influencing the efficacy of synaptic transmission. Of interest, clear evidence exists that the expression of SP and the spine apparatus depends on synaptic activity [20,28]. Studies visualizing activitydependent rearrangements of the spine apparatus at single spine resolution with fluorescent-tagged SP as a marker might be a tool to discover rules governing the formation and positioning of the spine apparatus within dendritic spines, thus clarifying the role of exact location of the spine apparatus.

8. Dendritic protein synthesis, LTP and the spine apparatus It is now well established that local dendritic protein synthesis and degradation play an important role in long-term synaptic plasticity [16,27,53,70]. LTP1 expression is mediated by phosphorylation of pre-existing proteins (e.g. AMPARs) by calcium/calmodulin-dependent kinase II (CaMKII, and possibly other kinases) which is activated by calcium entry through NMDARs and CICR through RyRs. In contrast, LTP2 expression seems to require translation of pre-existing dendritic mRNAs [59]. In addition to NMDARs, LTP2-inducing stimulation activates Group I metabotropic glutamate receptors leading to IP3 Rs-mediated calcium release in dendrites. This dendritic signal transduction cascade has been proposed to trigger translational machinery in the dendrites via the protein kinase C (PKC) and extracellular signalregulated kinase (ERK)–mitogen-activated protein kinase (MAPK) pathway (for details see Ref. [53]). Resulting newly synthesized proteins contribute to LTP maintenance by regulation of glutamate receptor trafficking [53]. The spine apparatus has recently been hypothesized to play a role in postsynaptic receptor trafficking and/or receptor synthesis and modification [21,68]. It is considered part of the dendritic secre-

17

tory pathway that is required for the local synthesis, processing and transport of membrane-bound proteins to the spine compartment [32,68,69]. Such a role appears to be plausible since the spine apparatus contains molecular markers for a satellite secretory pathway that might be involved in the translocation and processing of locally synthesized integral membrane proteins (e.g. receptors [54,55]). Markers for protein translocation sites (Sec61␣) and markers for Golgi cisternae (giantin and ␣-mannosidase II) have been found in association with spine endoplasmic reticulum [54,55]. Thus, changes in the expression of SP following LTP may reflect the remodelling of components of the secretory pathway by the stimulated neurons. Indeed, such a remodelling has been previously described for polyribosomes, which are normally found at the base of dendritic spines [69] and which are translocated into the spine compartment under conditions of LTP [15,49]. Taken together, these data indicate that the spine apparatus could be a component of the dendritic secretory pathway. In this context it could be involved in mechanisms of LTP2 and LTP1 which are dependent on local translation and post-translational modification of synaptic proteins, respectively. Given that the spine apparatus plays a role in local protein synthesis, modification or trafficking under conditions of LTP, it remains quite unclear which molecules and processes depend on its presence within a stimulated spine. One possibility is that the spine apparatus modulates the spatially specific delivery of glutamate receptors to synapses [35,72], since both AMPARs and NMDARs have been localized to the spine apparatus [48,56] and SP could bind via alpha-actinin [3,40] to synaptic NMDARs [79]. These considerations make it highly attractive to compare glutamatergic neurotransmission at spines containing a spine apparatus with those lacking this organelle and to investigate whether these two subpopulations of spines show differences in their GluR composition and assembly. 9. Long-lasting forms of LTP—a role for SP and the spine apparatus? Activity-dependent changes of the actin cytoskeleton and its associated proteins play a crucial role in structural synaptic plasticity [28,39,50]. In particular long-lasting forms of LTP (e.g. LTP3) are regularly accompanied by a profound reorganization of the spine cytoskeleton [39,44]. Since SP interacts with alpha-actinin and bundles and elongates actin filaments ([3,4,40], see also Ref. [81]), it is conceivable that SP could affect LTP3-related synaptic remodeling via an actin-dependent mechanism. At present, however, data on SP and SP-actin interactions in neurons are scarce and most of our knowledge is based on observations from cell lines [3,4,40] which makes it difficult to draw robust conclusions with regard to its role in LTP3. Another way how SP and the spine apparatus could influence LTP3 is via calcium. The endoplasmic reticulum could provide the cellular basis for the IP3 Rs-dependent calcium waves that travel from the synapse to the nucleus [10,74; but see 2]. In this scenario a spine apparatus-mediated elevation of calcium at synapses would trigger the dendritic calcium wave which eventually reaches the nucleus and starts the gene transcription required for LTP3. 10. Summary In spines, the actin-binding protein SP is tightly associated with the spine apparatus. SP-immunolabeling can thus be employed to visualize this organelle in neurons. SP-deficient mice lack a spine apparatus, demonstrating that SP is required for its formation. Importantly, SP-deficient mice showed deficits in LTP at the

18

P. Jedlicka et al. / Behavioural Brain Research 192 (2008) 12–19

CA3–CA1 synapse and in the radial arm maze, implicating the spine apparatus in learning and memory. Studies analyzing the role of calcium in spines suggest that the spine apparatus is involved in local calcium trafficking. In addition, the spine apparatus contains markers for a satellite secretory pathway which might be involved in the translocation and processing of locally synthesized integral membrane proteins. Thus, available data support two roles for the spine apparatus in long-term synaptic plasticity: 1. calcium regulation and 2. dendritic protein synthesis and postranslational processing. Acknowledgements This work was supported by the Deutsche Forschungsgemeinschaft (JE 528/1-1; DE 551/8) and the German Israeli Foundation (827/2004). References [1] Abraham WC, Otani S. Macromolecules and the maintenance of long-term potentiation. In: Morrell F, editor. Kindling and synaptic plasticity. The legacy of Graham Goddard. Birkhausen; 1991 p. 92–109. [2] Adams JP, Dudek SM. Late-phase long-term potentiation: getting to the nucleus. Nat Rev Neurosci 2005;6:737–43. [3] Asanuma K, Kim K, Oh J, Giardino L, Chabanis S, Faul C, et al. Synaptopodin regulates the actin-bundling activity of alpha-actinin in an isoform-specific manner. J Clin Invest 2005;115:1188–98. [4] Asanuma K, Yanagida-Asanuma E, Faul C, Tomino Y, Kim K, Mundel P. Synaptopodin orchestrates actin organization and cell motility via regulation of RhoA signalling. Nat Cell Biol 2006;8:485–91. [5] Balschun D, Wolfer DP, Bertocchini F, Barone V, Conti A, Zuschratter W, et al. Deletion of the ryanodine receptor type 3 (RyR3) impairs forms of synaptic plasticity and spatial learning. EMBO J 1999;18:5264–73. [6] Bas Orth C, Vlachos A, Del Turco D, Burbach GJ, Haas CA, Mundel P, et al. Laminaspecific distribution of synaptopodin, an actin-associated molecule essential for the spine apparatus, in identified principal cell dendrites of the mouse hippocampus. J Comp Neurol 2005;487:227–39. [7] Bas Orth C, Schultz C, Muller CM, Frotscher M, Deller T. Loss of the cisternal organelle in the axon initial segment of cortical neurons in synaptopodindeficient mice. J Comp Neurol 2007;504:441–9. [8] Behnisch T, Reymann KG. Thapsigargin blocks long-term potentiation induced by weak, but not strong tetanisation in rat hippocampal CA1 neurons. Neurosci Lett 1995;192:185–8. [9] Berridge MJ. Inositol trisphosphate and calcium signalling. Nature 1993;361:315–25. [10] Berridge MJ. Neuronal calcium signaling. Neuron 1998;21:13–26. [11] Berridge MJ. The endoplasmic reticulum: a multifunctional signaling organelle. Cell Calcium 2002;32:235–49. [12] Bliss TV, Collingridge GL. A synaptic model of memory: long-term potentiation in the hippocampus. Nature 1993;361:31–9. [13] Bliss TVP, Lomo T. Long-lasting potentiation of synaptic transmission in dentate area of anesthetized rabbit following stimulation of perforant path. J PhysiolLondon 1973;232:331–56. [14] Bliss TVP, Collingridge GL, Laroche S. ZAP and ZIP, a story to forget. Science 2006;313:1058–9. [15] Bourne JN, Sorra KE, Hurlburt J, Harris KM. Polyribosomes are increased in spines of CA1 dendrites 2 h after the induction of LTP in mature rat hippocampal slices. Hippocampus 2007;17:1–4. [16] Bramham CR, Wells DG. Dendritic mRNA: transport, translation and function. Nat Rev Neurosci 2007;8:776–89. [17] Deller T, Merten T, Roth SU, Mundel P, Frotscher M. Actin-associated protein synaptopodin in the rat hippocampal formation: Localization in the spine neck and close association with the spine apparatus of principal neurons. J Comp Neurol 2000;418:164–81. [18] Deller T, Haas CA, Deissenrieder K, Del Turco D, Coulin C, Gebhardt C, et al. Laminar distribution of synaptopodin in normal and reeler mouse brain depends on the position of spine-bearing neurons. J Comp Neurol 2002;453:33–44. [19] Deller T, Korte M, Chabanis S, Drakew A, Schwegler H, Stefani GG, et al. Synaptopodin-deficient mice lack a spine apparatus and show deficits in synaptic plasticity. Proc Natl Acad Sci USA 2003;100:10494–9. [20] Deller T, Orth CB, Vlachos A, Merten T, Del Turco D, Dehn D, et al. Plasticity of synaptopodin and the spine apparatus organelle in the rat fascia dentata following entorhinal cortex lesion. J Comp Neurol 2006;499:471–84. [21] Deller T, Orth CB, Del Turco D, Vlachos A, Burbach GJ, Drakew A, et al. A role for synaptopodin and the spine apparatus in hippocampal synaptic plasticity. Ann Anat 2007;189:5–16. [22] Emptage N, Bliss TV, Fine A. Single synaptic events evoke NMDA receptormediated release of calcium from internal stores in hippocampal dendritic spines. Neuron 1999;22:115–24.

[23] Fiala JC, Allwardt B, Harris KM. Dendritic spines do not split during hippocampal LTP or maturation. Nat Neurosci 2002;5:297–8. [24] Fifkova E, Markham JA, Delay RJ. Calcium in the spine apparatus of dendritic spines in the dentate molecular layer. Brain Res 1983;266:163–8. [25] Fifkova E. A possible mechanism of morphometric changes in dendritic spines induced by stimulation. Cell Mol Neurobiol 1985;5:47–63. [26] Fitzjohn SM, Collingridge GL. Calcium stores and synaptic plasticity. Cell Calcium 2002;32:405–11. [27] Fonseca R, Vabulas RM, Hartl FU, Bonhoeffer T, Nagerl UV. A balance of protein synthesis and proteasome-dependent degradation determines the maintenance of LTP. Neuron 2006;52:239–45. [28] Fukazawa Y, Saitoh Y, Ozawa F, Ohta Y, Mizuno K, Inokuchi K. Hippocampal LTP is accompanied by enhanced F-actin content within the dendritic spine that is essential for late LTP maintenance in vivo. Neuron 2003;38:447–60. [29] Futatsugi A, Kato K, Ogura H, Li ST, Nagata E, Kuwajima G, et al. Facilitation of NMDAR-independent LTP and spatial learning in mutant mice lacking ryanodine receptor type 3. Neuron 1999;24:701–13. [30] Gray EG. Electron microscopy of synaptic contacts on dendrite spines of the cerebral cortex. Nature 1959;183:1592–3. [31] Gruart A, Munoz MD, Delgado-Garcia JM. Involvement of the CA3-CA1 synapse in the acquisition of associative learning in behaving mice. J Neurosci 2006;26:1077–87. [32] Harris KM, Fiala JC, Ostroff L. Structural changes at dendritic spine synapses during long-term potentiation. Philos Trans R Soc Lond B Biol Sci 2003;358: 745–8. [33] Harris KM, Jensen FE, Tsao B. Three-dimensional structure of dendritic spines and synapses in rat hippocampus (CA1) at postnatal day 15 and adult ages: implications for the maturation of synaptic physiology and long-term potentiation. J Neurosci 1992;12:2685–705. [34] Harvey J, Collingridge GL. Thapsigargin blocks the induction of long-term potentiation in rat hippocampal slices. Neurosci Lett 1992;139:197–200. [35] Kennedy MJ, Ehlers MD. Organelles and trafficking machinery for postsynaptic plasticity. Annu Rev Neurosci 2006;29:325–62. [36] Koester HJ, Sakmann B. Calcium dynamics in single spines during coincident pre- and postsynaptic activity depend on relative timing of back-propagating action potentials and subthreshold excitatory postsynaptic potentials. Proc Natl Acad Sci USA 1998;95:9596–601. [37] Korkotian E, Segal M. Fast confocal imaging of calcium released from stores in dendritic spines. Eur J Neurosci 1998;10:2076–84. [38] Kovalchuk Y, Eilers J, Lisman J, Konnerth A. NMDA receptor-mediated subthreshold Ca(2+) signals in spines of hippocampal neurons. J Neurosci 2000;20:1791–9. [39] Kramar EA, Lin B, Rex CS, Gall CM, Lynch G. Integrin-driven actin polymerization consolidates long-term potentiation. Proc Natl Acad Sci USA 2006;103:5579–84. [40] Kremerskothen J, Plaas C, Kindler S, Frotscher M, Barnekow A. Synaptopodin, a molecule involved in the formation of the dendritic spine apparatus, is a dual actin/alpha-actinin binding protein. J Neurochem 2005;92:597–606. [41] Maggio N, Segal M. Striking variations in corticosteroid modulation of longterm potentiation along the septotemporal axis of the hippocampus. J Neurosci 2007;27:5757–65. [42] Maggio N, Segal M. Unique regulation of long term potentiation in the rat ventral hippocampus. Hippocampus 2007;17:10–25. [43] Mainen ZF, Malinow R, Svoboda K. Synaptic calcium transients in single spines indicate that NMDA receptors are not saturated. Nature 1999;399:151–5. [44] Matsuzaki M, Honkura N, Ellis-Davies GC, Kasai H. Structural basis of long-term potentiation in single dendritic spines. Nature 2004;429:761–6. [45] Matsuzaki M. Factors critical for the plasticity of dendritic spines and memory storage. Neurosci Res 2007;57:1–9. [46] Mundel P, Heid HW, Mundel TM, Kruger M, Reiser J, Kriz W. Synaptopodin: an actin-associated protein in telencephalic dendrites and renal podocytes. J Cell Biol 1997;139:193–204. [47] Nishiyama M, Hong K, Mikoshiba K, Poo MM, Kato K. Calcium stores regulate the polarity and input specificity of synaptic modification. Nature 2000;408:584–8. [48] Nusser Z, Lujan R, Laube G, Roberts JD, Molnar E, Somogyi P. Cell type and pathway dependence of synaptic AMPA receptor number and variability in the hippocampus. Neuron 1998;21:545–59. [49] Ostroff LE, Fiala JC, Allwardt B, Harris KM. Polyribosomes redistribute from dendritic shafts into spines with enlarged synapses during LTP in developing rat hippocampal slices. Neuron 2002;35:535–45. [50] Ouyang Y, Wong M, Capani F, Rensing N, Lee CS, Liu Q, et al. Transient decrease in F-actin may be necessary for translocation of proteins into dendritic spines. Eur J Neurosci 2005;22:2995–3005. ´ [51] Paillard J. Reflexions sur l’usage du concept de plasticite´ en neurobiology. J Psychol 1976;1:33–47. [52] Pastalkova E, Serrano P, Pinkhasova D, Wallace E, Fenton AA, Sacktor TC. Storage of spatial information by the maintenance mechanism of LTP. Science 2006;313:1141–4. [53] Pfeiffer BE, Huber KM. Current advances in local protein synthesis and synaptic plasticity. J Neurosci 2006;26:7147–50. [54] Pierce JP, van Leyen K, McCarthy JB. Translocation machinery for synthesis of integral membrane and secretory proteins in dendritic spines. Nat Neurosci 2000;3:311–3. [55] Pierce JP, Mayer T, McCarthy JB. Evidence for a satellite secretory pathway in neuronal dendritic spines. Curr Biol 2001;11:351–5.

P. Jedlicka et al. / Behavioural Brain Research 192 (2008) 12–19 [56] Racca C, Stephenson FA, Streit P, Roberts JD, Somogyi P. NMDA receptor content of synapses in stratum radiatum of the hippocampal CA1 area. J Neurosci 2000;20:2512–22. [57] Racine RJ, Milgram NW, Hafner S. Long-term potentiation phenomena in the rat limbic forebrain. Brain Res 1983;260:217–31. [58] Raymond CR, Redman SJ. Different calcium sources are narrowly tuned to the induction of different forms of LTP. J Neurophysiol 2002;88:249–55. [59] Raymond CR, Redman SJ. Spatial segregation of neuronal calcium signals encodes different forms of LTP in rat hippocampus. J Physiol 2006;570:97–111. [60] Raymond CR. LTP forms 1, 2 and 3: different mechanisms for the ’long’ in longterm potentiation. Trends Neurosci 2007;30:167–75. [61] Rose CR, Konnerth A. Stores not just for storage. intracellular calcium release and synaptic plasticity. Neuron 2001;31:519–22. [62] Roth SU, Sommer C, Mundel P, Kiessling M. Expression of synaptopodin, an actin-associated protein, in the rat hippocampus after limbic epilepsy. Brain Pathol 2001;11:169–81. [63] Segal M. Dendritic spines and long-term plasticity. Nat Rev Neurosci 2005;6:277–84. [64] Sharp AH, McPherson PS, Dawson TM, Aoki C, Campbell KP, Snyder SH. Differential immunohistochemical localization of inositol 1,4, 5-trisphosphateand ryanodine-sensitive Ca2+ release channels in rat brain. J Neurosci 1993;13:3051–63. [65] Shimuta M, Yoshikawa M, Fukaya M, Watanabe M, Takeshima H, Manabe T. Postsynaptic modulation of AMPA receptor-mediated synaptic responses and LTP by the type 3 ryanodine receptor. Mol Cell Neurosci 2001;17:921–30. [66] Spacek J. 3-Dimensional analysis of dendritic spines. II. Spine apparatus and other cytoplasmic components. Anat Embryol 1985;171:235–43. [67] Spacek J, Harris KM. Three-dimensional organization of smooth endoplasmic reticulum in hippocampal CA1 dendrites and dendritic spines of the immature and mature rat. J Neurosci 1997;17:190–203. [68] Steward O, Schuman EM. Protein synthesis at synaptic sites on dendrites. Annu Rev Neurosci 2001;24:299–325. [69] Steward O, Reeves TM. Protein-synthetic machinery beneath postsynaptic sites on CNS neurons: association between polyribosomes and other organelles at the synaptic site. J Neurosci 1988;8:176–84.

19

[70] Sutton MA, Schuman EM. Dendritic protein synthesis, synaptic plasticity, and memory. Cell 2006;127:49–58. [71] Svoboda K, Mainen ZF. Synaptic [Ca2+ ]: intracellular stores spill their guts. Neuron 1999;22:427–30. [72] Sytnyk V, Leshchyns’ka I, Dityatev A, Schachner M. Trans-Golgi network delivery of synaptic proteins in synaptogenesis. J Cell Sci 2004;117:381–8. [73] Tada T, Sheng M. Molecular mechanisms of dendritic spine morphogenesis. Curr Opin Neurobiol 2006;16:95–101. [74] Verkhratsky A. Physiology and pathophysiology of the calcium store in the endoplasmic reticulum of neurons. Physiol Rev 2005;85:201–79. [75] Vlachos A, Maggio N, Segal M. Lack of correlation between synaptopodin expression and the ability to induce LTP in the rat dorsal and ventral hippocampus. Hippocampus 2008;18:1–4. [76] Wang Y, Wu J, Rowan MJ, Anwyl R. Ryanodine produces a low frequency stimulation-induced NMDA receptor-independent long-term potentiation in the rat dentate gyrus in vitro. J Physiol 1996;495(Pt 3):755– 67. [77] Whitlock JR, Heynen AJ, Shuler MG, Bear MF. Learning induces long-term potentiation in the hippocampus. Science 2006;313:1093–7. [78] Wuytack F, Raeymaekers L, Missiaen L. Molecular physiology of the SERCA and SPCA pumps. Cell Calcium 2002;32:279–305. [79] Wyszynski M, Lin J, Rao A, Nigh E, Beggs AH, Craig AM, et al. Competitive binding of alpha-actinin and calmodulin to the NMDA receptor. Nature 1997;385:439–42. [80] Yamazaki M, Matsuo R, Fukazawa Y, Ozawa F, Inokuchi K. Regulated expression of an actin-associated protein, synaptopodin, during long-term potentiation. J Neurochem 2001;79:192–9. [81] Yanagida-Asanuma E, Asanuma K, Kim K, Donnelly M, Choi HY, Chang JH, et al. Synaptopodin protects against proteinuria by disrupting Cdc42: IRSp53: Mena signaling complexes in kidney podocytes. Am J Pathol 2007;171:415– 27. [82] Yuste R, Majewska A, Cash SS, Denk W. Mechanisms of calcium influx into hippocampal spines: heterogeneity among spines, coincidence detection by NMDA receptors, and optical quantal analysis. J Neurosci 1999;19:1976– 87.