Cyclic AMP-Responsive Element Binding ... - Wiley Online Library

Journal of Neurochemistry Lippincott Williams & Wilkins, Inc., Philadelphia © 1999 International Society for Neurochemistry

Cyclic AMP-Responsive Element Binding Protein in Brain Mitochondria *Martı´n Cammarota, *Gustavo Paratcha, *†Lia R. M. Bevilaqua, *Miguelina Levi de Stein, *Margarita Lopez, *Amanda Pellegrino de Iraldi, †Iva´n Izquierdo, and *Jorge H. Medina *Instituto de Biologia Celular y Neurociencias “Prof. Dr. Eduardo de Robertis,” Facultad de Medicina, Universidad de Buenos Aires, Buenos Aires, Argentina; and †Centro de Memoria, Departamento de Bioquimica, Instituto de Biociencias, Universidade Federal do Rio Grande do Sul, Porto Alegre, Brasil

Abstract: Cyclic AMP-responsive element binding protein (CREB) is critically involved in many important brain functions, including the formation of long-term memory. CREB is the best characterized member of a family of transcription factors (CREB/ATF family) recognized to be important nuclear targets for intracellular signal transduction systems. Here we show, by using different approaches, that CREB is unexpectedly localized to mitochondria of the rat brain. Controlled subcellular fractionation of hippocampus and cerebral cortex showed that both synaptic and nonsynaptic mitochondria exhibited immunoreactivity to the phosphorylated form of CREB (pCREB). Moreover, CREB extracted from synaptic mitochondria is able to be phosphorylated by the catalytic subunit of protein kinase A and dephosphorylated by protein phosphatase 1 or 2B. DNA mobility shift assays showed the presence of binding activity to the calcium– cyclic AMP-responsive element in mitochondrial extracts from hippocampus; this binding complex was specifically supershifted by an anti-CREB antibody. Immunoelectron microscopic analysis of hippocampal subcellular fractions revealed that pCREB immunoreactivity is localized in close association with the inner mitochondrial membrane. These results, together with recent findings describing the presence and phosphorylation of CREB in developing dendrites, suggest that CREB may participate in different mechanisms involved in the communication between extracellular signals and the expression of genes. Key Words: Cyclic AMP-responsive element binding protein—Brain mitochondria. J. Neurochem. 72, 2272–2277 (1999).

demonstrated that both the mRNA encoding cyclic AMP-responsive element binding protein (CREB) and the CREB protein are present in developing dendrites of hippocampal neurons and that phosphorylation of CREB on Ser133 occurred in isolated developing dendrites independent of the nucleus (Crino et al., 1998). In addition, CREB microperfused into dendrites is transported to and concentrated in the nucleus of cultured hippocampal neurons (Crino et al., 1998). Therefore, it was suggested that retrograde transport of dendritically located transcription factors might be important in modulating gene expression, bypassing different intracellular signaling pathways that converge to the nucleus (Meberg et al., 1996; Crino et al., 1998). CREB is a nuclear protein that regulates the transcription of genes with Ca2⫹ and/or cyclic AMP-responsive elements in their promoters (Ginty et al., 1993; Ghosh and Greenberg, 1995; Montminy, 1997). It is the first cloned and best characterized member of a family of transcription factors (CREB/ATF) that contains a leucine zipper DNA binding motif (Ghosh and Greenberg, 1995; Montminy, 1997). CREB phosphorylation on Ser133 by either cyclic AMP-, Ca2⫹-, or Ras-dependent protein kinases promotes the activation of gene expression (Ghosh and Greenberg, 1995). Received December 23, 1998; revised manuscript received January 14, 1999; accepted January 15, 1999. Address correspondence and reprint requests to Dr. J. H. Medina at Instituto de Biologia Celular y Neurociencias, Facultad de Medicina, Paraguay 2155, 3 piso, (1121) Buenos Aires, Argentina. Drs. G. Paratcha and L. R. M. Bevilaqua contributed equally to the work and could be cited in interchangeable order. Abbreviations used: CA-CRE, calcium– cyclic AMP-responsive element; ␣CAMKII, ␣-Ca2⫹/calmodulin-dependent kinase II; CREB, cyclic AMP-responsive element binding protein; Cyt, cytosolic; EMSA, electrophoretic mobility shift assay; FGF, fibroblast growth factor; GluR2/3, glutamate receptor 2/3; GSK3, glycogen synthase kinase 3; NF-␬B, nuclear factor-␬B; NMDAR2B, NMDA receptor 2B; nsMit, nonsynaptic mitochondria; Nuc, nuclear; pCREB, Ser133 phosphorylated form of CREB; PKA, protein kinase A; PP, protein phosphatase; Psd, postsynaptic densities; sMit, synaptic mitochondria; Spm, synaptic plasma membranes; Syn, synaptosomes.

A growing body of evidence indicates that transcription factors may participate in a novel signaling system that communicates between the synapse and nucleus (Meberg et al., 1996; O’Neill and Kaltschmidt, 1997). The inducible transcription factor nuclear factor-␬B (NF␬B) has been found in the synapses of cerebral cortex and hippocampus (Meberg et al., 1996; Suzuki et al., 1997), localized mainly to postsynaptic densities (Psd) (Suzuki et al., 1997), and exhibited activity-dependent changes (Meberg et al., 1996). Quite recently, it has been 2272

CREB IN BRAIN MITOCHONDRIA CREB is a crucial component of several intracellular signal pathways involved in the establishment of longterm memory in both vertebrates and invertebrates (Bourtchuladze et al., 1994; Yin et al., 1994; Bernabeu et al., 1997; Gusowski and McGaugh, 1997) and the setting of biological clocks (Ginty et al., 1993). Recently, we have used a specific antibody against the activated Ser133 phosphorylated form of CREB (pCREB) to show that memory formation of an inhibitory avoidance learning in rats is associated with an increase in the phosphorylation of CREB in CA1 hippocampal cells (Bernabeu et al., 1997). We also found that pCREB immunoreactivity was apparently not confined to the cell nucleus. Therefore, to determine whether CREB is present in different cellular compartments, here we performed electron microscopycontrolled subcellular fractionation of the rat hippocampus and immunoblotted the resulting fractions with three different CREB antibodies: one that recognizes CREB regardless of its phosphorylation state and two others that specifically detect pCREB. In addition, immunoelectron microscopy was used to localize pCREB in hippocampal synaptic mitochondria (sMit). MATERIALS AND METHODS Subcellular fractionation Throughout these studies, we used adult male Wistar rats weighing between 160 and 190 g. To obtain the different subcellular fractions from cerebral cortex or hippocampus, we performed the fractionation procedure essentially as described by Lai and Clark (1979). In brief, tissues were homogenized in ice-chilled isolation medium containing 0.32 M sucrose, 1 mM K⫹-EDTA, 1 mM EGTA, 10 mM Tris-HCl (pH 7.4), plus a cocktail of protease and protein phosphatase (PP) inhibitors to obtain the nuclear (Nuc), cytosolic (Cyt), and crude mitochondrial fractions. The latter was resuspended and layered onto a 7.5/10% Ficoll gradient and centrifuged to obtain the myelin, synaptosomes (Syn), and nonsynaptic mitochondria (nsMit). The Syn were lysed with 1 mM Tris-HCl (pH 6.8) buffer and layered onto a second 3/4.5/6% Ficoll gradient to obtain the sMit and the synaptic plasma membranes (Spm). Spm aliquots were incubated with Triton X-100, and the resulting Psd were isolated by a further centrifugation step. Mitochondria from liver, heart, kidney, and adrenal gland were obtained as described before (Guerra, 1974; Antkiewicz-Michaluk et al., 1988). To obtain soluble mitochondrial extracts, the sMit fraction was submitted to three freezing–thawing cycles, incubated during 30 min at 4°C in the presence of solubilization buffer (20 mM Tris-HCl, 12% glycerol, 0.5 M NaCl, 0.5 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride, 5 ␮g/ml aprotinin, 50 mM NaF, 0.5 Nonidet P-40, 0.2 mM EDTA, pH 7.9), and then centrifuged at 25,000 g during 20 min. The resulting supernatant was dialyzed against 20 mM Tris-HCl (pH 7.9) containing 12% glycerol and stored at ⫺70°C. Protein concentrations were determined by the Bradford method.

Immunoblotting analysis The experiments were performed using previously published methods (Towbin et al., 1979) with the following sources of primary antibodies: anti-CREB (BioLabs; directed against the CREB residues 123–137; 1:1,000), anti-pCREB (BioLabs; corresponding to pCREB and directed against residues 129 –137; 1:1,000), anti-pCREB (directed against CREB residues 123–

2273

136; 1:1,000; generously provided by D. Ginty), anti-c-fos (Calbiochem; 1:10,000), anti-NF-␬B (p65; Santa Cruz; 1:1,000), anti-GAP-43 (generously provided by B. Oestreicher; 1:10,000), anti-␣-Ca2⫹/calmodulin-dependent kinase II (anti␣CAMKII; Chemicon; 1:5,000), anti-NMDA receptor 2B (anti-NMDAR2B; Chemicon; 1:5,000), and anti-glutamate receptor 2/3 (anti-GluR2/3; Chemicon; 1:1,000). Antibody–antigen complexes were detected with a goat antibody to rabbit IgG conjugated to horseradish peroxidase (Bio-Rad) and visualized by the enhanced chemiluminescence method as described by the manufacturer (Amersham). PP2B was obtained from Sigma; PP1, recombinant protein kinase A (PKA) catalytic subunit, and glycogen synthase kinase 3 (GSK3) peptide substrate were obtained from New England Biolabs.

Electrophoretic mobility shift assay (EMSA) Double-stranded oligonucleotide (Promega; 5⬘-AGAGATTGCCTGACGTCAGAGAGCTAG-3⬘) containing the calcium– cyclic AMP-responsive element (CA-CRE) consensus sequence (indicated in bold letters) was labeled with [␥-32P]ATP using T4 polynucleotide kinase. The assays were performed as previously described (Kornhauser et al., 1992). In brief, sMit extracts from hippocampus were incubated with 0.5–1 ng of 32 P-labeled probe for 20 min at 4°C before electrophoresis on a 6% nondenaturing polyacrylamide gel in 0.5⫻ 50 mM Trisglycine (pH 7.5) and 1 mM EDTA. In the competition assays, a 30⫻ excess of either the unlabeled probe or a mutated CRE consensus sequence (AC3 TG) was added to the reaction mixture 10 min prior to the addition of the labeled probe. Supershift experiments were performed by preincubating the extracts during 30 min at room temperature with 1 ␮l of anti-CREB antibody. After electrophoresis, the gels were dried and exposed to x-ray film.

Electron microscopy and immunoelectron microscopy Electron microscopic analysis of the different subcellular fractions (De Robertis et al., 1962; Lai and Clark, 1979) obtained from the rat hippocampus and cerebral cortex was performed as previously described (De Robertis et al., 1962), utilizing a Zeiss EM 10 C or a Siemens Elmiskop 1 electron microscope. Preembedding immunoelectron microscopy was carried out as described by Liu and Jones (1996) with antipCREB (Ginty et al., 1993; 1:1,000, overnight). Control sections were stained by omitting the primary or the secondary antibodies.

RESULTS Figure 1a and b illustrates the immunoblots of different subcellular fractions of rat hippocampus, with total cell extracts from SK-N-MC cells prepared with or without treatment with forskolin and fibroblast growth factor (FGF) as controls of CREB and pCREB immunoreactivities, respectively. As expected, CREB immunoreactivity, evidenced as a single band migrating with a molecular mass of 43 kDa, was present largely in Nuc and Cyt fractions (Fig. 1a). Unexpectedly, CREB was also found in purified hippocampal sMit and nsMit (Fig. 1a). In agreement with previous reports (Ginty et al., 1993; Deisseroth et al., 1996; Montminy, 1997), basal levels of pCREB in Nuc extracts were very low (Fig. 1b). Although we did not detect pCREB immunoreactivity in Cyt (Fig. 1b) or in myelin or microsomal fractions (data J. Neurochem., Vol. 72, No. 6, 1999

2274

M. CAMMAROTA ET AL. FIG. 1. CREB and pCREB immunoreactivities in subcellular fractions of rat hippocampus and in mitochondria of other tissues, assessed by immunoblotting analysis. Thirty micrograms of protein was loaded in each lane. a: CREB (BioLabs) immunoreactivity in Syn, Cyt, Nuc, nsMit, and sMit fractions. We used an SK-N-MC cell extract without treatment with forskolin and FGF (BioLabs) as a positive control (n-pCon). b: The same blot as in (a), stripped out and reincubated with anti-pCREB (BioLabs). In this case, the positive control consisted of an SK-N-MC cellular extract treated with forskolin and FGF (pCon; BioLabs). c: pCREB immunoreactivity, using another anti-pCREB (Ginty et al., 1993), in Syn and their derived fractions: Spm and sMit. d: Presence of pCREB in mitochondria from heart, liver, kidney, and adrenal gland, detected with the same antibody as in (c). e: Electron micrograph showing purified mitochondria in the hippocampal sMit fraction. Bar ⫽ 0.2 ␮m.

not shown), we were able to detect pCREB immunoreactivity in Syn fractions (Fig. 1b). Further fractionation of hippocampal Syn showed that pCREB immunoreactivity was concentrated largely in the sMit fraction, as no consistent detectable pCREB immunoreactivity was found in the Spm fraction (Fig. 1c) or in the synaptosomal soluble extracts (data not shown). An intense pCREB immunoreactivity was also seen in the nsMit fraction (Fig. 1c). Quite similar results were obtained when rat cerebral cortex was used as starting material (data not shown). The presence of mitochondrial CREB is not restricted to brain structures. As can be seen in Fig. 1d, mitochondria from several peripheral tissues also exhibited pCREB immunoreactivity. The presence of CREB and pCREB in mitochondria appears not to be due to nuclear contamination, because only highly purified mitochondria were seen in mitochondrial fractions at the electron microscopic level (see Fig. 1e). To further explore the possibility that nuclear components were contaminating our sMit fraction, we carried out immunoblot experiments with specific antibodies against two other transcription factors. As shown in Fig. 2a, both the inducible transcription factor c-fos and the constitutive transcription factor NF-␬B (p65) were detected in Nuc but not in sMit extracts from hippocampus. These findings, together with those depicted in Fig. 1a– c showing a differential pattern of CREB and pCREB immunoreactivities detected in Nuc and mitochondrial extracts, rule out the possibility that the presence of CREB in mitochondria is the result of nuclear contamination. Moreover, no pCREB was extracted from sMit after 30-min incubation in the presence of 1 M NaCl or 0.5 M KSCN (Fig. 2b, lanes 2 and 3). In contrast, incubation of sMit with 0.5% Triton X-100 for 30 min markedly decreased pCREB levels in the resulting pellet (Fig. 2b, lane 4), suggesting that the presence of pCREB in hippocampal sMit is not due to a nonspecific adsorption or a loose attachment of pCREB to the outer mitochondrial membrane. J. Neurochem., Vol. 72, No. 6, 1999

The sMit preparations were not contaminated with other synaptic components, as we failed to detect immunoreactivity for NMDA2B receptor subunit, GluR2/3 subunit, or ␣CaMKII subunit, three well-known proteins localized mainly to the postsynaptic region, or GAP-43, a selective presynaptic marker (Fig. 2c). To further determine if CREB is present in mitochondrial extracts, EMSAs were performed. Similar specific binding to the CA-CRE sequence was detected in both sMit and Nuc fractions from hippocampus (Fig. 3, SHIFT lanes). This binding activity was specifically competed for by nonlabeled CRE probe but not by a mutated CRE probe (Fig. 3; 30 ⫻ CRE and 30 ⫻ CREm lanes, respectively). In addition, when anti-CREB anti-

FIG. 2. The sMit fraction seems not to be contaminated by nuclear or other synaptic components. a: Immunoblots for c-fos (80 ␮g/lane) and NF-␬B (p65; 50 ␮g/lane) in Nuc and sMit fractions. b: Immunoblots with anti-pCREB (Ginty et al., 1993), showing the failure to extract pCREB from sMit with 1 M NaCl (lane 2) or with 0.5 M KSCN (lane 3). pCREB immunoreactivity was significantly decreased when the sMit fraction was incubated with 0.5% Triton X-100 (lane 4). In brief, the sMit fraction was treated as indicated during 30 min at 4°C and then centrifuged at 10,000 g during 30 min; after this, the obtained pellets were resuspended and equal amounts of proteins analyzed by immunoblot. c: Immunoblots for typical presynaptic (GAP-43) or postsynaptic (NMDA2B, GluR2/3, and ␣CAMKII) markers in the Psd, sMit, and Spm fractions.

CREB IN BRAIN MITOCHONDRIA

2275

ting the primary or the secondary antibodies (Fig. 4C and D). The CREB extracted from sMit is able to be backphosphorylated in Ser133 by the catalytic subunit of PKA and dephosphorylated by purified PP2B or PP1 (Fig. 5). DISCUSSION

FIG. 3. Binding to the CA-CRE in sMit and Nuc extracts from hippocampus (single arrowheads point to the major shifted bands). The incubation mix (20 ␮l) contained the radioactive oligonucleotide plus sample (15 ␮g of protein) (SHIFT lanes). Several control reactions were performed: 30⫻ CRE, plus a 30-fold excess of unlabeled CRE oligonucleotide; 30⫻ CREm, plus a 30-fold excess of mutated CRE oligonucleotide; ⫹antiCREB, plus 1 ␮l of anti-CREB antibody; ⫹antiCREB⫹pept., plus 1 ␮l of anti-CREB and 0.1 ␮g/ml GSK3 substrate peptide (comprising amino acids 123–135 from CREB); ⫹pept., plus 0.1 ␮g/ml GSK3 substrate peptide alone; ⫹antiCREBi, plus 1 ␮l of heat-inactivated (56°C, 30 min) anti-CREB antibody. The same experiment is shown in the upper and lower panels; the film exposure time is shown on the right. Double arrowheads point to the major supershifted bands.

body was included in the reaction, a further reduction in the electrophoretic mobility of the shifted band was evident (Fig. 3; ⫹antiCREB lanes). The supershifted complexes were completely abolished by both the heat inactivation of the antibody (Fig. 3; ⫹antiCREBi lanes) and by the inclusion in the reaction mixture of a synthetic peptide corresponding to residues 123–135 of CREB (Fig. 3; ⫹antiCREB⫹pept. lanes); the sole inclusion of this peptide in the incubation medium had no effect on the formation of the complexes (Fig. 3; ⫹pept. lanes). The presence of pCREB in mitochondria was also confirmed by immunoelectron microscopic analysis of hippocampal crude mitochondrial fractions. As can be seen in Fig. 4B, there is an intense specific pCREB immunoreactivity that seems to be localized to the inner mitochondrial membrane. In synaptic terminals, pCREB immunoreactivity is confined to mitochondria (Fig. 4A). No immunoreactivity was found in control sections omit-

The main finding of the present study is that CREB and its Ser133 phosphorylated form are localized to brain mitochondria. This is based on three series of data. First, CREB and pCREB were detected in purified sMit and nsMit using three different specific antibodies (Fig. 1). Unexpectedly, basal pCREB immunoreactivity in brain mitochondria is higher than that detected in Nuc or Cyt fractions. Second, we detected in extracts from brain mitochondria the presence of a factor that binds the CA-CRE element. This complex is supershifted by a specific anti-CREB antibody (Fig. 3), indicating that CREB is a component of the complex. Finally, immunoelectron microscopy revealed that pCREB is detected in hippocampal sMit, probably localized to the inner mitochondrial membrane (Fig. 4). The phosphorylation state of CREB appears to be an important factor in long-lasting activity-dependent mechanisms of synaptic plasticity in a range of neural structures in both vertebrates and invertebrates, including the mammalian hippocampus (Bito et al., 1996; Deisseroth et al., 1996, 1998). It depends on intracellular signaling pathways triggered by the activation of NMDA receptors and L-type calcium channels and is engaged mainly in the maintenance phase of long-term potentiation. The activation of nuclear CREB via phosphorylation of Ser133 is very rapid in hippocampal neurons (Deisseroth et al., 1996). It has been recently demonstrated that brief synaptic activity produces a translocation of calmodulin to the nucleus and that this translocation, which is dependent on certain Ca2⫹ entry systems, is important for the rapid phosphorylation of CREB (Deisseroth et al., 1998). Our findings pose the intriguing question of whether the synaptic activity-dependent local rise in Ca2⫹ concentration may also participate in the phosphorylation of mitochondrial CREB. In this context, mitochondria have been postulated to act as a sort of Ca2⫹-signaling memory storage device (Miller, 1998), accumulating Ca2⫹ after brief tetanic stimulation but, more importantly, releasing Ca2⫹ to the cytoplasm long after the cessation of the original stimulus (Tang and Zucker, 1997; Miller, 1998). As mentioned in the introductory section, different signaling pathways have been shown to participate in the phosphorylation of nuclear CREB on Ser133 (Ginty et al., 1993; Ghosh and Greenberg, 1995; Montminy, 1997). In this regard, it has been recently demonstrated that phosphorylation of CREB on Ser133 occurred in isolated developing dendrites regulated by the activation of metabotropic glutamate receptors and that this phosphorylation is independent of the nucleus (Crino et al., 1998). In this context, it is noteworthy that the major forms of ␣CAMKII are mainly cytoplasmic (Bito et al., 1996), J. Neurochem., Vol. 72, No. 6, 1999

2276

M. CAMMAROTA ET AL.

FIG. 4. Preembedding immunostaining for pCREB in crude mitochondrial fractions. A: Synaptic terminals with dense immunoperoxidase reaction product in sMit (arrows). Bar ⫽ 0.1 ␮m. B: pCREB immunoreactivity associated with inner mitochondrial membranes (arrow). Bar ⫽ 0.025 ␮m. C and D: Control reactions without the addition of primary antibody. Bar in C ⫽ 0.1 ␮m; bar in D ⫽ 0.025 ␮m.

and it has been shown that ␣CAMKII is also localized to the outer mitochondrial membranes (Liu and Jones, 1996). Here we have shown that the CREB present in sMit can be “in vitro” phosphorylated by PKA and dephosphorylated by PP2B and PP1 (Fig. 5). Our findings confirm and extend recent data demonstrating the presence and phosphorylation of CREB in developing dendrites (Crino et al., 1998) and strongly suggest that dendritic CREB in the adult brain is localized mainly to mitochondria. The detailed molecular

FIG. 5. Immunoblots. A: Phosphorylation of hippocampal sMit CREB in Ser133 by means of the catalytic subunit of PKA. B and C: Dephosphorylation of hippocampal sMit Ser133 pCREB by means of PP2B or PP1, respectively. In brief, hippocampal sMit soluble extracts were heat inactivated (100°C, 5 min) and then incubated at 30°C for the indicated periods of time in the presence of 1 U of the above-mentioned enzymes in the following buffers: 50 mM Tris-HCl, 10 mM MgCl2, plus 200 ␮M ATP (pH 7.5) (A); 50 mM Tris-HCl, 0.5 ␮M calmodulin, and 0.5 mM CaCl2 (B); and 50 mM Tris-HCl, 0.1 mM EDTA, 1 mM MnCl2, 5 mM dithiothreitol, and 0.01% Brij 35 (pH 7.0) (C). The reactions were halted with hot 4⫻ Laemmli sample buffer, and the samples were submitted to sodium dodecyl sulfate–polyacrylamide gel electrophoresis and immunoblotted with anti-pCREB antibody.

J. Neurochem., Vol. 72, No. 6, 1999

characterization of CREB activation and its role in mitochondria will be a challenge. Acknowledgment: This work was supported by CONICET and the University of Buenos Aires (Argentina) and Pronex (Brazil).

REFERENCES Antkiewicz-Michaluk L., Guidotti A., and Krueger K. E. (1988) Molecular characterization and mitochondrial density of a recognition site for peripheral-type benzodiazepine ligands. Mol. Pharmacol. 34, 272–278. Bernabeu R., Bevilaqua L., Ardenghi P., Bromberg E., Schmitz P., Bianchin M., Izquierdo I., and Medina J. H. (1997) Involvement of hippocampal cAMP/cAMP-dependent protein kinase signaling pathways in a late memory consolidation phase of aversively motivated learning in rats. Proc. Natl. Acad. Sci. USA 94, 7041– 7046. Bito H., Deisseroth K., and Tsien R. W. (1996) CREB phosphorylation and dephosphorylation: a Ca2⫹- and stimulus duration-dependent switch for hippocampal gene expression. Cell 87, 1203–1214. Bourtchuladze R., Frenguelli B., Blendy J., Schutz G., and Silva A. (1994) Deficient long-term memory in mice with a targeted mutation of the cAMP-responsive element binding protein. Cell 79, 59 – 68. Crino P., Khodakhah K., Becker K., Ginsberg S., Hemby S., and Eberwine J. (1998) Presence and phosphorylation of transcription factors in developing dendrites. Proc. Natl. Acad. Sci. USA 95, 2313–2318. Deisseroth K., Bito H., and Tsien R. W. (1996) Signaling from synapse to nucleus: postsynaptic CREB phosphorylation during multiple forms of hippocampal synaptic plasticity. Neuron 16, 89 –101. Deisseroth K., Heist E. K., and Tsien R. W. (1998) Translocation of calmodulin to the nucleus supports CREB phosphorylation in hippocampal neurons. Nature 392, 198 –202. De Robertis E., Pellegrino de Iraldi A., Rodriguez de Lores Arnaiz G., and Salganicoff L. (1962) Cholinergic and non-cholinergic nerve endings in rat brain: I. Isolation and subcellular distribution of acetylcholinesterase. J. Neurochem. 9, 23–25.

CREB IN BRAIN MITOCHONDRIA Ghosh A. and Greenberg M. E. (1995) Calcium signaling in neurons: molecular mechanisms and cellular consequences. Science 268, 239 –247. Ginty D. D., Kornhauser J. M., Thompson M. A., Bading H., Mayo K. E., Takahashi J. S., and Greenberg M. E. (1993) Regulation of CREB phosphorylation in the suprachiasmatic nucleus by light and a circadian clock. Science 260, 238 –241. Guerra F. C. (1974) Rapid isolation technique for mitochondria: technique for rat liver mitochondria. Methods Enzymol. 31, 299 –305. Gusowski J. F. and McGaugh J. L. (1997) Antisense oligodeoxynucleotide-mediated disruption of hippocampal cAMP response element binding protein levels impairs consolidation of memory for water maze training. Proc. Natl. Acad. Sci. USA 94, 2693–2698. Kornhauser J. M., Nelson D. E., Mayo K. E., and Takahashi J. S. (1992) Regulation of jun-b messenger RNA and AP-1 activity by light and a circadian clock. Science 255, 1581–1584. Lai J. C. K. and Clark J. B. (1979) Preparation of synaptic and nonsynaptic mitochondria from mammalian brain. Methods Enzymol. 55, 51– 60. Liu X.-B. and Jones E. G. (1996) Localization of alpha type II calcium calmodulin-dependent protein kinase at glutamatergic but not aminobutyric acid (GABAergic) synapses in thalamus and cerebral cortex. Proc. Natl. Acad. Sci. USA 93, 7332–7336. Meberg P. J., Kinney W. R., Valcourt E. G., and Routtenberg A. (1996) Gene expression of the transcription factor NF␬B in

2277

hippocampus: regulation by synaptic activity. Mol. Brain Res. 38, 179 –190. Miller R. J. (1998) Mitochondria—the Kraken wakes! Trends Neurosci. 21, 95–97. Montminy M. (1997) Transcriptional regulation by cyclic AMP. Annu. Rev. Biochem. 66, 807– 822. O’Neill L. A. J. and Kaltschmidt C. (1997) NF-␬B: a crucial transcription factor for glial and neuronal cell function. Trends Neurosci. 20, 252–257. Suzuki T., Mitake S., Okumura-Noji K., Yang J., Fujii T., and Okamoto T. (1997) Presence of NF-␬B-like and I␬B-like immunoreactivity in postsynaptic densities. Neuroreport 8, 2931– 2935. Tang Y. G. and Zucker R. S. (1997) Mitochondrial involvement in post-tetanic potentiation of synaptic transmission. Neuron 18, 483– 491. Towbin H., Staehelin T., and Gordon J. (1979) Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc. Natl. Acad. Sci. USA 76, 4350 – 4354. Yin J. C. P., Wallach J. S., Del Vecchio M., Wilder E. L., Quinn W. G., and Tully T. (1994) Induction of a dominant negative CREB transgene specifically blocks long-term memory in Drosophila. Cell 79, 49 –58.

J. Neurochem., Vol. 72, No. 6, 1999