Voltar
Artigo Acesso aberto Revisado por pares
BIBTEX RIS

Capacitative Calcium Entry Induces Hippocampal Long Term Potentiation in the Absence of Presenilin-1

2003; Elsevier BV; Volume: 278; Issue: 45 Linguagem: Inglês

10.1074/jbc.m300971200

ISSN

1083-351X

Autores

Laurence Ris, Ilse Dewachter, Delphine Reversé, Emile Godaux, Fred Van Leuven,

Tópico(s)

Cholinesterase and Neurodegenerative Diseases

Resumo

Presenilins, whose mutant forms are the most common cause of early onset familial Alzheimer's disease, are involved in two very distinct processes: (i) proteolytic activity as γ-secretase acting on amyloid precursor protein to produce amyloid peptides and (ii) storage of Ca2+ in the endoplasmic reticulum (ER). In particular, absence of presenilin-1 (PS1) was claimed to potentiate capacitative calcium entry (CCE), i.e. the mechanism of replenishment of ER Ca2+ stores. However, until now, evidence in favor of the latter role has been obtained only in isolated or cultured cells and not on neurons in situ. Here, we studied the strength of the synapses between Schaffer's collaterals and CA1 neurons in hippocampal slices when they were submitted first to Ca2+-free medium containing thapsigargin and subsequently to normal artificial cerebrospinal fluid, a procedure known to trigger CCE. We demonstrate that Ca2+ influx via the CCE mechanism is sufficient to trigger robust long term potentiation of the synapses in hippocampal slices from transgenic mice with a postnatal, neuron-specific ablation of PS1, but remarkably not from wild-type mice. Our data establish for the first time in neurons confined in normal neuronal networks that PS1 acts on the refilling mechanism of ER Ca2+ stores. Presenilins, whose mutant forms are the most common cause of early onset familial Alzheimer's disease, are involved in two very distinct processes: (i) proteolytic activity as γ-secretase acting on amyloid precursor protein to produce amyloid peptides and (ii) storage of Ca2+ in the endoplasmic reticulum (ER). In particular, absence of presenilin-1 (PS1) was claimed to potentiate capacitative calcium entry (CCE), i.e. the mechanism of replenishment of ER Ca2+ stores. However, until now, evidence in favor of the latter role has been obtained only in isolated or cultured cells and not on neurons in situ. Here, we studied the strength of the synapses between Schaffer's collaterals and CA1 neurons in hippocampal slices when they were submitted first to Ca2+-free medium containing thapsigargin and subsequently to normal artificial cerebrospinal fluid, a procedure known to trigger CCE. We demonstrate that Ca2+ influx via the CCE mechanism is sufficient to trigger robust long term potentiation of the synapses in hippocampal slices from transgenic mice with a postnatal, neuron-specific ablation of PS1, but remarkably not from wild-type mice. Our data establish for the first time in neurons confined in normal neuronal networks that PS1 acts on the refilling mechanism of ER Ca2+ stores. Mutations in presenilin-1 (PS1) 1The abbreviations used are: PS1, presenilin-1; ACSF, artificial cerebrospinal fluid; APV, d(-)-2-amino-5-phosphonovaleric acid; CCE, capacitative calcium entry; ER, endoplasmic reticulum; SERCA, smooth ER Ca2+-ATPase pump; EPSP, excitatory postsynaptic potential; fEPSP, field EPSP; LTP, long term potentiation; TG, thapsigargin; PKA, cyclic AMP-dependent protein kinase; JNK, c-Jun NH2-terminal kinase; SAPK, stress-activated protein kinase; MAP, mitogen-activated protein; SOC, store-operated calcium channels; VDCC, voltage dependent calcium channels; NMDA, N-methyl-d-aspartate. and presenilin-2 (PS2) are the most common cause of early onset cases of familial Alzheimer's disease (1Levy-Lahad E. Wasco W. Poorkaj P. Romano D.M. Oshima J. Pettingell W.H. Yu C.E. Jondro P.D. Schmidt S.D. Wang K. Science. 1995; 269: 973-977Crossref PubMed Scopus (2230) Google Scholar, 2Rogaev E.I. Sherrington R. Rogaeva E.A. Levesque G. Ikeda M. Liang Y. Chi H. Lin C. Holman K. Tsuda T. Mar L. Sorbi S. Nacmias B. Placentini S. Amaolucci L. Chumakov I. Cohen D. Lannfelt L. Fraser P.E. Rommens J.M. St. George-Hyslop P.H. Nature. 1995; 376: 775-778Crossref PubMed Scopus (1791) Google Scholar, 3Sherrington R. Froelich S. Sorbi S. Campion D. Chi H. Rogaeva E.A. Levesque G. Rogaev E.I. Lin C. Liang Y. Ikeda M. Mar L. Brice A. Agid Y. Percy M.E. Clerget-Darpoux F. Piancentini S. Marcon G. Nacmias B. Amaducci L. Frebourg T. Lannfelt L. Rommens J.M. St George-Hyslop P.H. Hum. Mol. Genet. 1996; 5: 985-988Crossref PubMed Scopus (243) Google Scholar). Presenilins are serpentine proteins containing 6–8 putative transmembrane domains (4Doan A. Thinakaran G. Borchelt D.R. Slunt H.H. Ratovitsky T. Podlisny M. Selkoe D.J. Seeger M. Gandy S.E. Price D.L. Sisodia S.S. Neuron. 1996; 17: 1023-1030Abstract Full Text Full Text PDF PubMed Scopus (355) Google Scholar, 5Li X. Greenwald I. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 7109-7114Crossref PubMed Scopus (142) Google Scholar) that are localized predominantly in the membranes of the endoplasmic reticulum (ER) (6Cook D.G. Sung J.C. Golde T.E. Felsenstein K.M. Wojczk B.S. Tanzi R.E. Trojanowski J.Q. Lee V.M.-Y. Doms R.W. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 9223-9228Crossref PubMed Scopus (182) Google Scholar, 7Kovacs D.M. Fausett H.J. Page K.J. Kim T.-W. Moir R.D. Merriam D.E. Hollister R.D. Hallmark O.G. Mancini R. Felsenstein K.M. Hyman B.T. Tanzi R.E. Wasco W. Nat. Med. 1996; 2: 224-229Crossref PubMed Scopus (513) Google Scholar, 8Kim S.H. Lah J.S. Thinakaran G. Levey A. Sisodia S. Neurobiol. Dis. 2000; 7: 99-117Crossref PubMed Scopus (49) Google Scholar). Functional evidence demonstrates that presenilins are involved in two very distinct types of processes. Best documented is the proteolytic activity first discovered as acting on the amyloid precursor protein (APP) in producing the amyloid peptides (9De Strooper B. Saftig P. Craessaerts K. Vandersticheles H. Guhde G. Annaert W. Von Figura K. Van Leuven F. Nature. 1998; 391: 387-390Crossref PubMed Scopus (1552) Google Scholar), the major constituents of amyloid plaques in the brain of Alzheimer's disease patients. Presenilins are also involved in calcium signaling in neurons and other cells (10Yoo A.S. Cheng I. Chung S. Grenfell T.Z. Lee H. Pack-Chung E. Handler M. Shen J. Xia W. Tesco G. Saunders A.J. Ding K. Frosch M.P. Tanzi R.E. Kim T.W. Neuron. 2000; 27: 561-572Abstract Full Text Full Text PDF PubMed Scopus (273) Google Scholar, 11Leissring M.A. Akbari Y. Fanger C.M. Cahalan M.D. Mattson M.P. LaFerla F.M. J. Cell Biol. 2000; 149: 793-797Crossref PubMed Scopus (288) Google Scholar, 12Leissring M.A. Murphy M.P. Mead T.R. Akbari Y. Sugarman M.C. Jannatipour M. Anliker B. Müller U. Saftig P. De Strooper B. Wolfe M.S. Golde T.E. LaFerla F.M. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 4697-4702Crossref PubMed Scopus (254) Google Scholar, 13Schneider I. Reversé D. Dewachter I. Ris L. Caluwaerts N. Kuipéri C. Gilis M. Geerts H. Kretzschmar H. Godaux E. Moechars D. Van leuven F. Herms J. J. Biol. Chem. 2001; 276: 11539-11544Abstract Full Text Full Text PDF PubMed Scopus (119) Google Scholar). Cells store calcium ions in the ER (for a review, see Ref. 14Berridge M.J. Neuron. 1998; 21: 13-26Abstract Full Text Full Text PDF PubMed Scopus (1756) Google Scholar) by the action of Ca2+-transporting ATPases (i.e. SERCA), that are irreversibly blocked by thapsigargin (TG) (15Thastrup O. Cullen P.J. Drobak B. Hanley M.R. Dawson A.P. Proc. Natl. Acad. Sci. U. S. A. 1990; 87: 2466-2470Crossref PubMed Scopus (3001) Google Scholar). Declining Ca2+ stores in the ER produce a signal, the molecular nature of which is still a matter of debate (16Putney J.W. Nature. 2001; 410: 648-649Crossref PubMed Scopus (32) Google Scholar), to open store-operated Ca2+ channels (SOC) in the plasma membrane (17Parekh A.B. Penner R. Physiol. Rev. 1997; 77: 901-930Crossref PubMed Scopus (1291) Google Scholar). These activated channels serve to replenish the ER Ca2+ stores by the mechanism known as "capacitative calcium entry" (CCE) (18Putney Jr., J.W. Cell Calcium. 1990; 11: 611-624Crossref PubMed Scopus (1261) Google Scholar). From experiments carried out on isolated or cultured cells, it has been proposed that presenilins exert different effects on the Ca2+ store system. First, deficiency of PS1 potentiates CCE (10Yoo A.S. Cheng I. Chung S. Grenfell T.Z. Lee H. Pack-Chung E. Handler M. Shen J. Xia W. Tesco G. Saunders A.J. Ding K. Frosch M.P. Tanzi R.E. Kim T.W. Neuron. 2000; 27: 561-572Abstract Full Text Full Text PDF PubMed Scopus (273) Google Scholar), whereas mutant PS1 causes a deficit in CCE (10Yoo A.S. Cheng I. Chung S. Grenfell T.Z. Lee H. Pack-Chung E. Handler M. Shen J. Xia W. Tesco G. Saunders A.J. Ding K. Frosch M.P. Tanzi R.E. Kim T.W. Neuron. 2000; 27: 561-572Abstract Full Text Full Text PDF PubMed Scopus (273) Google Scholar, 11Leissring M.A. Akbari Y. Fanger C.M. Cahalan M.D. Mattson M.P. LaFerla F.M. J. Cell Biol. 2000; 149: 793-797Crossref PubMed Scopus (288) Google Scholar, 12Leissring M.A. Murphy M.P. Mead T.R. Akbari Y. Sugarman M.C. Jannatipour M. Anliker B. Müller U. Saftig P. De Strooper B. Wolfe M.S. Golde T.E. LaFerla F.M. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 4697-4702Crossref PubMed Scopus (254) Google Scholar). Secondly, the ER calcium content, assessed by the rise in cytosolic [Ca2+] induced by thapsigargin in Ca2+-free medium, is significantly decreased in cells lacking PS1 (12Leissring M.A. Murphy M.P. Mead T.R. Akbari Y. Sugarman M.C. Jannatipour M. Anliker B. Müller U. Saftig P. De Strooper B. Wolfe M.S. Golde T.E. LaFerla F.M. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 4697-4702Crossref PubMed Scopus (254) Google Scholar) and increased by a mutant PS1 (11Leissring M.A. Akbari Y. Fanger C.M. Cahalan M.D. Mattson M.P. LaFerla F.M. J. Cell Biol. 2000; 149: 793-797Crossref PubMed Scopus (288) Google Scholar). Thirdly, inositol triphosphate-mediated release of Ca2+ from the ER is decreased in PS1-deficient cells (12Leissring M.A. Murphy M.P. Mead T.R. Akbari Y. Sugarman M.C. Jannatipour M. Anliker B. Müller U. Saftig P. De Strooper B. Wolfe M.S. Golde T.E. LaFerla F.M. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 4697-4702Crossref PubMed Scopus (254) Google Scholar). However, until now, the modulation of the Ca2+ store system by PS1 has not been studied in neurons in situ, a more physiological preparation. Here, we aimed to study the proposed role of PS1 in calcium signaling in neurons confined within normal neuronal networks, not by direct Ca2+ measurements but by an indirect mean, more appropriate to studies in slices from adult animals. The synaptic efficiency between Schaffer's collaterals and CA1 neurons in hippocampal slices was assessed in circumstances that trigger entry of Ca2+ by the CCE mechanism. This provided a suitable system to define the effect of increased CCE in neurons that are deficient in PS1 as opposed to wild-type neurons. Based on two premises, i.e. (i) long term potentiation (LTP) induced by electrical tetanus in CA1 neurons is dependent on Ca2+ influx and (ii) Ca2+ influx by CCE is potentiated in cultured cells deficient in PS1, we hypothesized and tested that increased Ca2+ entry through CCE would be sufficient to induce LTP in CA1 neurons deficient in PS1 and not in wild-type neurons. This was achieved by carrying out experiments in hippocampal slices from transgenic mice with postnatal, neuron-specific ablation of PS1, denominated PS1(n–/–) (19Dewachter I. Reversé D. Caluwaerts N. Ris L. Kuipéri C. Van den Haute C. Spittaels K. Umans L. Sterneels L. Thiry E. Moechars D. Mercken M. Van Leuven F. J. Neurosci. 2002; 22: 3445-3453Crossref PubMed Google Scholar). Transgenic mice with neuron-specific deficiency in PS1, denoted PS1(n–/–), were generated by targeting the PS1 gene in embryonic stem cells to contain loxP sites. The resulting mice were crossed with transgenic mice that express Cre recombinase specifically and exclusively in their central neurons under the control of the mouse thy-1 gene promoter (19Dewachter I. Reversé D. Caluwaerts N. Ris L. Kuipéri C. Van den Haute C. Spittaels K. Umans L. Sterneels L. Thiry E. Moechars D. Mercken M. Van Leuven F. J. Neurosci. 2002; 22: 3445-3453Crossref PubMed Google Scholar). The combination of thy1-Cre recombinase and loxP-PS1 gene was maintained by selection based on four independent PCR reactions performed on DNA extracted from tail-tip cuttings from the offspring. Double transgenic mice were further bred to the homozygous condition for the loxP-modified PS1 gene, to result in a neuron-specific deficiency of PS1. The PS1(n–/–) mice used in the experiments presented here were viable and fertile and have normal brain morphology and normal behavior (19Dewachter I. Reversé D. Caluwaerts N. Ris L. Kuipéri C. Van den Haute C. Spittaels K. Umans L. Sterneels L. Thiry E. Moechars D. Mercken M. Van Leuven F. J. Neurosci. 2002; 22: 3445-3453Crossref PubMed Google Scholar). Some experiments were carried out on transgenic mice that express the FAD-mutant PS1(A246). These mice were generated using the thy-1 gene promoter and have been described and characterized extensively (20Moechars D. Dewachter I. Lorent K. Reversé D. Baekelandt V. Naidu A. Tesseur I. Spittaels K. Van Den Haute C. Checler F. Godaux E. Cordell B. Van Leuven F. J. Cell Biol. 1999; 274: 6483-6492Scopus (614) Google Scholar). Hippocampal slices were at all times bathing in artificial cerebrospinal fluid (ACSF) with the following composition: 124 mm NaCl, 5 mm KCl, 26 mm NaHCO3, 1.24 mm KH2PO4, 2.4 mm CaCl2, 1.3 mm MgSO4, 10 mm glucose, bubbled with a mixture of 95% 02 and 5% CO2. Mice were anesthetized with ether and decapitated. The brain was cut with a vibratome in transverse slices (400 μm thick) in cold ACSF, which were kept afterward at room temperature until placed in a submersion recording chamber at 30 °C. Electrophysiological recording was started not earlier than 3 h after dissection to allow recovery of the slices. Only a single slice from each tested animal was investigated. The measuring chamber was perfused with ACSF (3 ml/min). Bipolar tungsten microelectrodes (World Precision Instruments, Sarasota, FL) were used to stimulate Schaffer's collaterals, whereas evoked field excitatory postsynaptic potentials were recorded in the stratum radiatum of the CA1 region with low resistance (2 megohm) glass microelectrodes filled with 2 m NaCl. Test stimuli were 0.1-ms constant-voltage pulses delivered every 30 s at an intensity sufficient to evoke an approximate 33% maximal response. The slope of the field excitatory postsynaptic potential (fEPSP, mV/ms) was measured from the average wave from four consecutive responses, as described previously (13Schneider I. Reversé D. Dewachter I. Ris L. Caluwaerts N. Kuipéri C. Gilis M. Geerts H. Kretzschmar H. Godaux E. Moechars D. Van leuven F. Herms J. J. Biol. Chem. 2001; 276: 11539-11544Abstract Full Text Full Text PDF PubMed Scopus (119) Google Scholar, 19Dewachter I. Reversé D. Caluwaerts N. Ris L. Kuipéri C. Van den Haute C. Spittaels K. Umans L. Sterneels L. Thiry E. Moechars D. Mercken M. Van Leuven F. J. Neurosci. 2002; 22: 3445-3453Crossref PubMed Google Scholar, 20Moechars D. Dewachter I. Lorent K. Reversé D. Baekelandt V. Naidu A. Tesseur I. Spittaels K. Van Den Haute C. Checler F. Godaux E. Cordell B. Van Leuven F. J. Cell Biol. 1999; 274: 6483-6492Scopus (614) Google Scholar). The following commercially available drugs were used: thapsigargin (Sigma), cyclopiazonic acid (Tocris, Bristol, UK), d(-)-2-amino-5-phosphonovaleric acid (Acros), SKF96365 (Tocris), anisomycin (Sigma), and cycloheximide (Sigma). For each slice, the fEPSP slopes were normalized against the average slope over the 30 min before a specific treatment. To determine whether or not the normalized fEPSP of a group of slices submitted to the same experimental conditions was significantly potentiated (p < 0.05), the percentages of baseline measured just before the treatment and 2 h (unless otherwise specified) after its end were compared by a paired Student's t test. Statistical significance (p < 0.05) of the difference in increase of the fEPSP measured 2 h after a treatment applied to two distinct groups of slices was assessed by a Student's t test. CCE-induced LTP in Hippocampus of PS1(n–/–) Mice— Classically, LTP is induced by tetanic stimulation, and it is well known that entry of calcium ions is an absolute requirement. Here, we tested the hypothesis that LTP could be induced in PS1-deficient neurons by their overactive mechanism of CCE, i.e. the mechanism of replenishing ER Ca2+ stores. CCE was triggered by first incubating the brain slices in Ca2+-free medium containing 1 μm TG, which is known to decrease the Ca2+ stores of the ER, and subsequently exposing them to normal Ca2+-containing medium (Fig. 1). The efficacy of transmission at the synapses between Schaffer's collaterals and CA1 pyramidal cells of the hippocampus was measured with extracellular microelectrodes in the classical manner. In slices of wild-type mice, Ca2+-free perfusion medium resulted in a complete synaptic blockage, evidently by blocking all neurotransmitter release. Upon reperfusion of non-transgenic slices with Ca2+-containing medium, the slope of the fEPSP began to recover after about 10 min and subsequently stabilized at control levels over the entire observation period (Fig. 1). After 2 h, the slope of the fEPSP was 96.5 ± 9.7% (mean ± S.D., n = 6), i.e. practically unchanged relative to the resting values. The very small variation on these measurements demonstrated the extremely reproducible nature of the response in the six independent control mice analyzed (Fig. 1). In contrast, this protocol of induction of CCE caused a dramatic increase in the slope of the fEPSP in hippocampal slices from PS1(n–/–) mice (Fig. 1). Two h after stopping the perfusion with TG/Ca2+-free medium, the slope of the field EPSP was increased to 174.8 ± 18.7% (n = 6) of the basal value (p < 0.001). This effect was not unique for TG since another agent known to block ER-based calcium pumps and thereby to deplete the ER Ca2+ stores, i.e. cyclopiazonic acid, produced qualitatively the same effect (Fig. 2). Two h after stopping superfusion of the slices with 2 μm cyclopiazonic acid in calcium-free medium, the slope of the fEPSP was also markedly enhanced in CA1 neurons in sections from PS1(n–/–) mice (135.0 ± 12.6%, n = 4, p < 0.05) but did not increase at all in slices derived from non-transgenic mice (93.2 ± 6.0%, n = 4) (Fig. 2). These results demonstrated that a protocol known to induce an entry of Ca2+ through the CCE mechanism was able to trigger an LTP in CA1 neurons deficient in PS1, but not in normal ones. Origin of Ca2 + Influx—LTP induced by incubation in TG/Ca2+-free medium was completely blocked by the CCE inhibitor SKF96365 (10Yoo A.S. Cheng I. Chung S. Grenfell T.Z. Lee H. Pack-Chung E. Handler M. Shen J. Xia W. Tesco G. Saunders A.J. Ding K. Frosch M.P. Tanzi R.E. Kim T.W. Neuron. 2000; 27: 561-572Abstract Full Text Full Text PDF PubMed Scopus (273) Google Scholar) when the slices were pretreated for 3 h with 30 μm (relative change versus baseline: 107.8 ± 10.5%, n = 4, p = 0.12) (Fig. 3A). Because tetanus-induced LTP in the hippocampal CA1 neurons is triggered by an entry of Ca2+ through the NMDA receptors, we also investigated a potential role of these channels on the TG/Ca2+-free induced LTP (Fig. 3B). Brain slices from PS1(n–/–) mice were submitted to the TG/Ca2+-free protocol in the presence of 50 μm APV, a well known antagonist of the NMDA receptor. APV modified neither the profile of the CCE-induced LTP nor its amplitude measured 2 h after application of TG/Ca2+-free treatment. The values of the fEPSP were 158.9 ± 11.0% (n = 4) in the presence of APV (Fig. 3B) versus 174.8 ± 18.7% (n = 6) in the absence of APV (Fig. 1, p = 0.16). Relationship between CCE-induced LTP and Tetanus-induced LTP—Tetanus-evoked LTP occluded the induction of TG/Ca2+-free induced LTP (Fig. 4). Thirty min after the induction of strong LTP by three tetanic stimulations (1 s, 100 Hz) separated from each other by 10 min, the stimulation intensity was reduced to obtain fEPSP with amplitudes corresponding to the control values. Ten min later, the slice was submitted to the Ca2+-free medium containing 1 μm TG for 30 min. In this situation, subsequent TG/Ca2+-free treatment did not induce LTP since after 1 h, the fEPSP was measured to be 95.2 ± 4.1% (n = 4) relative to basal values (p = 0.50). However, the converse was not true since the LTP induced by TG/Ca2+-free medium did not occlude subsequent tetanus-evoked LTP (Fig. 5A). Two h after the induction of LTP by TG/Ca2+-free treatment, the stimulation intensity was reduced to obtain fEPSP with amplitudes corresponding to the control values for 15 min. A single tetanus (1 s, 100 Hz) applied at that moment evoked LTP, i.e. 30 min after the application of the tetanus, fEPSP was 145.3 ± 20.4% of control values (n = 4, p < 0.05). To rule out the possibility that the observed absence of occlusion of tetanus-induced LTP by CCE-induced LTP(Fig. 5A) was due to the fact that the CCE-induced LTP was not saturated yet at the moment of the application of the electrical tetanus, the following control experiment was performed (Fig. 5B). Two h after the induction of LTP by TG/Ca2+-free treatment, the stimulation intensity was reduced to obtain fEPSP amplitudes corresponding to baseline for 15 min. A second TG/Ca2+-free treatment applied at that time did not induce any LTP since 90 min after the second stimulus, the fEPSP was measured to be 97.9 ± 5.5% (n = 4) relative to basal values (p = 0.35) CCE-induced LTP and Protein Synthesis—To investigate whether CCE-induced LTP was dependent on protein synthesis, we applied anisomycin, an inhibitor of protein synthesis, for 30 min before, during (30 min), and for 30 min after perfusion of the slices with a Ca2+-free medium containing thapsigargin. Pretreatment with anisomycin completely abolished the synaptic enhancement induced by TG/Ca2+-free treatment in PS1(n–/–) mice (98.3 ± 15.6, n = 4, p = 0.94) (Fig. 6A). To demonstrate that the effect of anisomycin was related to inhibition of protein synthesis and not to another effect of the drug, which is also a potent activator of stress-activated protein kinases (JNK/SAPK) and p38 MAP kinase, we tested the effect of cycloheximide, an unrelated inhibitor of protein synthesis, at a concentration of 60 μm according to the same temporal pattern as anisomycin. Cycloheximide also nearly completely blocked CCE-induced LTP in PS1(n–/–) mice (Fig. 6B). Two h after the TG/Ca2+-free treatment, the fEPSP was as low as 113.7 ± 13.5% (n = 5) relative to basal values (p = 0.34). We conclude that CCE-induced LTP was dependent on protein synthesis. Mutant PS1 and CCE-induced LTP—Cultured neurons and fibroblasts harboring a mutant PS1 causing familial Alzheimer's disease have markedly reduced CCE (10Yoo A.S. Cheng I. Chung S. Grenfell T.Z. Lee H. Pack-Chung E. Handler M. Shen J. Xia W. Tesco G. Saunders A.J. Ding K. Frosch M.P. Tanzi R.E. Kim T.W. Neuron. 2000; 27: 561-572Abstract Full Text Full Text PDF PubMed Scopus (273) Google Scholar, 11Leissring M.A. Akbari Y. Fanger C.M. Cahalan M.D. Mattson M.P. LaFerla F.M. J. Cell Biol. 2000; 149: 793-797Crossref PubMed Scopus (288) Google Scholar). We therefore predicted that the protocol based on TG/Ca2+-free triggering would not produce LTP in CA1 neurons in brain slices from mutant PS1 mice. This prediction was proven in slices from the brain of transgenic mice that express the early onset cases of familial Alzheimer's disease mutant PS1(A246E) (13Schneider I. Reversé D. Dewachter I. Ris L. Caluwaerts N. Kuipéri C. Gilis M. Geerts H. Kretzschmar H. Godaux E. Moechars D. Van leuven F. Herms J. J. Biol. Chem. 2001; 276: 11539-11544Abstract Full Text Full Text PDF PubMed Scopus (119) Google Scholar). Indeed, the TG/Ca2+-free medium was unable to induce significant hippocampal LTP in PS1(A246E) slices since after 2 h of TG/Ca2+-free perfusion, the slope of the fEPSP was 99.2 ± 22.4% (n = 5, p = 0.49) (Fig. 7). The major finding of this work is the following. The treatment applied to the hippocampal slices, which is Ca2+-free/thapsigargin followed by normal ACSF, induced an LTP of the synaptic strength between Schaffer's collaterals and CA1 neurons in PS1(n–/–) mice but not in wild-type mice. Such an LTP was due to Ca2+ entry by the CCE mechanism, as it was blocked by an agent known to block CCE (SKF96365). This is thus experimental evidence obtained on brain slices where neuronal networks were intact, and not based on isolated or cultured neurons, for a role of presenilin 1 in the refilling mechanism of the intracellular Ca2+ stores. Before discussing the relationships between CCE-induced LTP and tetanus-induced LTP, it is worth remembering the characteristics of the classical tetanus-induced LTP whose major actors are shown below (see Fig. 8). LTP caused by multiple trains of high frequency electrical stimulation can be divided into two phases: (i) an early phase, consisting of an induction mechanism and an early maintenance phase; and (ii) a late phase, which is a maintenance phase (21Kandel E.R. Science. 2001; 294: 1030-1038Crossref PubMed Scopus (2692) Google Scholar). Induction is triggered by initial increase of cytosolic Ca2+ in dendritic spines due mainly to Ca2+ entry through NMDA receptors (22Malenka R.C. Nicoll R.A. Science. 1999; 285: 1870-1874Crossref PubMed Scopus (2240) Google Scholar). In this induction phase, activation of calcium/calmodulin-dependent protein kinase (23Giese K. Fedorov N. Filipkowski R. Silva A. Science. 1998; 279: 870-873Crossref PubMed Scopus (884) Google Scholar) but also of protein kinase C (24Huang Y.Q. Lu W.Y. Ali D.W. Pelkey K.A. Pitcher G.M. Lu Y.M. Aoto H. Roder J.C. Sasaki T. Salter M.W. MacDonald J.F. Neuron. 2001; 29: 485-496Abstract Full Text Full Text PDF PubMed Scopus (238) Google Scholar) are believed to play a pivotal role. In contrast to the early phase, the late maintenance phase of tetanus-induced LTP requires new protein synthesis (25Deadwyler S.A. Dunwiddie T. Lynch G. Synapse. 1987; 1: 90-95Crossref PubMed Scopus (75) Google Scholar, 26Huang Y.Y. Kandel E.R. Learn Mem. (Cold Spring Harb.). 1994; 1: 74-82PubMed Google Scholar, 27Frey U. Morris G.M. Nature. 1997; 385: 533-536Crossref PubMed Scopus (1284) Google Scholar). This is triggered by a cascade initiated by still another kinase, the cyclic AMP-dependent protein kinase (PKA), which is activated by the large Ca2+ influx provoked by repeated trains of action potentials (21Kandel E.R. Science. 2001; 294: 1030-1038Crossref PubMed Scopus (2692) Google Scholar, 28Impey S. Mark M. Villacres E.C. Poser S. Chavkin C. Storm D.R. Neuron. 1996; 16: 973-982Abstract Full Text Full Text PDF PubMed Scopus (508) Google Scholar). All the signaling pathways underlying the tetanus-induced LTP (both early and late phases) are unaffected by genetic ablation of PS1 (19Dewachter I. Reversé D. Caluwaerts N. Ris L. Kuipéri C. Van den Haute C. Spittaels K. Umans L. Sterneels L. Thiry E. Moechars D. Mercken M. Van Leuven F. J. Neurosci. 2002; 22: 3445-3453Crossref PubMed Google Scholar, 29Yu H. Saura C.A. Choi S.Y. Sun L.D. Yang X. Handler M. Kawarabayashi T. Younkin L. Fedeles B. Wilson M.A. Younkin S. Kandel E.R. Kirkwood A. Shen J. Neuron. 2001; 31: 713-726Abstract Full Text Full Text PDF PubMed Scopus (221) Google Scholar). In PS1(n–/–) mice, calcium entering the neurons through the CCE mechanism triggers long term potentiatiation by activating only pathway(s) involved in tetanus-induced LTP (since tetanus-induced LTP occludes CCE-induced LTP), although not all of them (since CCE-induced LTP does not occlude tetanus-induced LTP). One of the shared pathways is that leading to the late phase of tetanus-induced LTP. Indeed, CCE-induced LTP and the late phase of tetanus-induced LTP share the property of being protein synthesis-dependent. PS1 is a serpentine protein known to be localized mainly in the ER. At the postsynaptic level, it is not clear whether the changes of CCE induced by a PS1 deficiency take place in the dendritic spines, in the dendritic shafts, and/or in the cell soma. Whereas the presence of functional ER calcium stores in the spines of hippocampal CA1 is attested by the fact that ryanodine receptors are enriched in dendritic spines (30Sharp A.H. McPherson P.S. Dawson T.M. Aoki C. Campbell K.P. Snyder S.H. J. Neurosci. 1993; 13: 3051-3063Crossref PubMed Google Scholar), subsurface cisterns of ER are closely apposed to the plasma membrane of the soma (14Berridge M.J. Neuron. 1998; 21: 13-26Abstract Full Text Full Text PDF PubMed Scopus (1756) Google Scholar). It is clear that Ca2+ entering the dendritic spines by the CCE mechanism could have access to the signaling pathways of the tetanus-induced LTP. However, even if the increased CCE in PS1(n–/–) mice was confined to the soma, it could be sufficient to activate the late maintenance component of the tetanus-induced LTP. Indeed, APV blocks the early phase of LTP but not the late phase of LTP, whereas nifedipine, which blocks L-type voltage-dependent calcium channels (VDCC), attenuates the late phase of LTP (28Impey S. Mark M. Villacres E.C. Poser S. Chavkin C. Storm D.R. Neuron. 1996; 16: 973-982Abstract Full Text Full Text PDF PubMed Scopus (508) Google Scholar). This suggests that the late phase of tetanus-induced LTP is triggered by Ca2+ ions, which enter not only through NMDA receptors but also through L-type VDCC channels. In this regard, it is interesting to note that L-type VDCC are localized in the somatic membrane and proximal dendrites (31Westenbroek R.E. Ahlijanian M.K. Catterall W.A. Nature. 1990; 347: 281-284Crossref PubMed Scopus (398) Google Scholar). Thus, entry of Ca2+ through SOC located in the plasma membrane of the soma or of the dendritic spines could be responsible for the induced enhancement of the synaptic strength. Because functional calcium stores are present not only in the soma and dendrites, but also in synaptic boutons (32Emptage N.J. Reid C.A. Fine A. Neuron. 2001; 29: 197-208Abstract Full Text Full Text PDF PubMed Scopus (429) Google Scholar, 33Llano I. González J. Caputo C. Lai F.A. Blayney L.M. Tan Y.P. Marty A. Nat. Neurosci. 2000; 3: 1256-1265Crossref PubMed Scopus (329) Google Scholar), a possible presynaptic mechanism responsible for the CCE-induced LTP in the PS1(n–/–) mice deserves further discussion. In synaptic terminals of CA3 pyramidal cells, a presynaptic action potential causes a transient increase in Ca2+ due primarily to an entry of Ca2+ from the extracellular space through the VGCC but also to a calcium-induced release of calcium by the presynaptic calcium stores (32Emptage N.J. Reid C.A. Fine A. Neuron. 2001; 29: 197-208Abstract Full Text Full Text PDF PubMed Scopus (429) Google Scholar). It has been demonstrated that a depletion of the Ca2+ stores in presynaptic terminals by cyclopiazonic acid or TG induced a decrease in the calcium transients evoked by action potentials in presynaptic boutons. However, as this effect is a decrease and as it does not affect the amplitude of EPSPs caused by single stimuli, it cannot be responsible for the increase in synaptic strength observed in CCE-induced LTP. Pharmacologically induced depletion of Ca2+ stores in synaptic boutons also causes another effect, i.e. it activates the store-operated channels of the presynaptic membrane. The resulting continuous influx of calcium through the SOC then increases the frequency of the miniature EPSPs (32Emptage N.J. Reid C.A. Fine A. Neuron. 2001; 29: 197-208Abstract Full Text Full Text PDF PubMed Scopus (429) Google Scholar). This mechanism, however, is not involved in the CCE-induced LTP either since an increased rate of spontaneous transmitter release would act at the synapses through the NMDA receptors, whereas CCE-induced LTP is not blocked by APV (Fig. 3B) The next question is how CCE enhances synaptic strength in the absence of PS1 and not in its presence. The first possibility is that the amount of Ca2+ entering the cell by the CCE mechanism is increased by the absence of PS1 and so sufficient to reach the threshold level needed to trigger the signal cascades leading to an increased synaptic efficacy (Fig. 8, A and B). By measuring the concentration of Ca2+ in the cytosol of cultured cells, it has been demonstrated that genetic ablation of PS1 resulted in: (i) an increased entry of Ca2+ by the CCE mechanism (10Yoo A.S. Cheng I. Chung S. Grenfell T.Z. Lee H. Pack-Chung E. Handler M. Shen J. Xia W. Tesco G. Saunders A.J. Ding K. Frosch M.P. Tanzi R.E. Kim T.W. Neuron. 2000; 27: 561-572Abstract Full Text Full Text PDF PubMed Scopus (273) Google Scholar) and (ii) a decrease in their ER-Ca2+ stores (12Leissring M.A. Murphy M.P. Mead T.R. Akbari Y. Sugarman M.C. Jannatipour M. Anliker B. Müller U. Saftig P. De Strooper B. Wolfe M.S. Golde T.E. LaFerla F.M. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 4697-4702Crossref PubMed Scopus (254) Google Scholar). LaFerla's group (11Leissring M.A. Akbari Y. Fanger C.M. Cahalan M.D. Mattson M.P. LaFerla F.M. J. Cell Biol. 2000; 149: 793-797Crossref PubMed Scopus (288) Google Scholar) had proposed previously that CCE was activated when calcium levels within the lumen of the ER fell below a threshold level. According to these authors, CCE would be increased in PS1 knock-out cells because their ER-Ca2+ stores are abnormally low (12Leissring M.A. Murphy M.P. Mead T.R. Akbari Y. Sugarman M.C. Jannatipour M. Anliker B. Müller U. Saftig P. De Strooper B. Wolfe M.S. Golde T.E. LaFerla F.M. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 4697-4702Crossref PubMed Scopus (254) Google Scholar). A second possibility that could explain how CCE induces LTP in PS1 knock-out neurons and not in control neurons is that the absence of PS1 would not modify the amount of Ca2+ entering by the CCE mechanism, but instead its access to its targets. PS1 could structurally help to position and maintain SOC close to SERCA pumps at the level of the subsurface cisterns of the ER, thus organizing a local "signaling microdomain" (34Delmas P. Wanaverbecq N. Abogadie F.C. Mistry M. Brown D.A. Neuron. 2002; 14: 209-220Abstract Full Text Full Text PDF Scopus (231) Google Scholar). In this case, Ca2+ ions entering through SOC would either be readily pumped into the ER in the absence of thapsigargin or be locally "trapped" when SERCA is blocked irreversibly by thapsigargin (Fig. 9A). In the absence of PS1, clustering of SOC and their apposition in front of subsurface cisterns of ER would be lost, and Ca2+ entering through SOC would no longer be locally trapped between the membrane of the cell and that of the corresponding subsurface cistern. That would allow Ca2+ to have direct access to its target the adenylate cyclase and hence activate PKA and trigger enhanced synaptic strength (Fig. 9B). A final question, not addressed in this study but most interesting to raise, is whether the increase in CCE observed in PS1-deficient mice is related to their evident inhibition of γ-secretase activity (19Dewachter I. Reversé D. Caluwaerts N. Ris L. Kuipéri C. Van den Haute C. Spittaels K. Umans L. Sterneels L. Thiry E. Moechars D. Mercken M. Van Leuven F. J. Neurosci. 2002; 22: 3445-3453Crossref PubMed Google Scholar). Most recently, the first two effects were claimed to be mimicked by γ-secretase inhibitors (12Leissring M.A. Murphy M.P. Mead T.R. Akbari Y. Sugarman M.C. Jannatipour M. Anliker B. Müller U. Saftig P. De Strooper B. Wolfe M.S. Golde T.E. LaFerla F.M. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 4697-4702Crossref PubMed Scopus (254) Google Scholar). On the other hand, it is most interesting to note that PS1 acts on CCE directly, independently of APP, as observed in neurons from APP-null mice (35Van Leuven F. Dewachter I. Herms J. Godaux E. Neurobiol. Aging. 2002; 23: S243Google Scholar). It is clear that fundamentally we do not fully understand the diverse functions of PS1 and the context of its dual function in regulated proteolytic activity and in calcium signaling. We thank Christiane Busson for the realization of the figures and Ramona Shelby for correcting the English.

Referência(s)