Ca2+-dependent Regulation of TrkB Expression in Neurons
2003; Elsevier BV; Volume: 278; Issue: 42 Linguagem: Inglês
10.1074/jbc.m303082200
ISSN1083-351X
AutoresTami J. Kingsbury, Peter D. Murray, Linda L. Bambrick, Bruce K. Krueger,
Tópico(s)Neurobiology and Insect Physiology Research
ResumoThe neurotrophin brain-derived neurotrophic factor (BDNF), via activation of its receptor, tyrosine receptor kinase B (trkB), regulates a wide variety of cellular processes in the nervous system, including neuron survival and synaptic plasticity. Although the expression of BDNF is known to be Ca2+-dependent, the regulation of trkB expression has not been extensively studied. Here we report that depolarization of cultured mouse cortical neurons increased the expression of the full-length, catalytically active isoform of trkB without affecting expression of the truncated isoform. This increase in protein expression was accompanied by increased levels of transcripts encoding full-length, but not truncated, trkB. Depolarization also regulated transcription of the gene, TRKB, via entry of Ca2+ through voltage-gated Ca2+ channels and subsequent activation of Ca2+-responsive elements in the two TRKB promoters. Using transient transfection of neurons with TRKB promoter-luciferase constructs, we found that Ca2+ inhibited the upstream promoter P1 but activated the downstream promoter P2. Ca2+-dependent stimulation of TRKB expression requires two adjacent, non-identical CRE sites located within P2. The coordinated regulation of BDNF and trkB by Ca2+ may play a role in activity-dependent survival and synaptic plasticity by enhancing BDNF signaling in electrically active neurons. The neurotrophin brain-derived neurotrophic factor (BDNF), via activation of its receptor, tyrosine receptor kinase B (trkB), regulates a wide variety of cellular processes in the nervous system, including neuron survival and synaptic plasticity. Although the expression of BDNF is known to be Ca2+-dependent, the regulation of trkB expression has not been extensively studied. Here we report that depolarization of cultured mouse cortical neurons increased the expression of the full-length, catalytically active isoform of trkB without affecting expression of the truncated isoform. This increase in protein expression was accompanied by increased levels of transcripts encoding full-length, but not truncated, trkB. Depolarization also regulated transcription of the gene, TRKB, via entry of Ca2+ through voltage-gated Ca2+ channels and subsequent activation of Ca2+-responsive elements in the two TRKB promoters. Using transient transfection of neurons with TRKB promoter-luciferase constructs, we found that Ca2+ inhibited the upstream promoter P1 but activated the downstream promoter P2. Ca2+-dependent stimulation of TRKB expression requires two adjacent, non-identical CRE sites located within P2. The coordinated regulation of BDNF and trkB by Ca2+ may play a role in activity-dependent survival and synaptic plasticity by enhancing BDNF signaling in electrically active neurons. The neurotrophin, brain-derived neurotrophic factor (BDNF), 1The abbreviations used are: BDNF, brain-derived neurotrophic factor; APV, dl(–)-2-amino-5-phosphonopentanoic acid; CRE, cyclic AMP-response element; CREB, cyclic AMP-response element binding protein; DNQX, 6,7-dinitroquinoxaline-2,3-dione; trkB, tyrosine receptor kinase B; TRKB, gene coding for trkB; UTR, untranslated region; MOPS, 4-morpholinepropanesulfonic acid.1The abbreviations used are: BDNF, brain-derived neurotrophic factor; APV, dl(–)-2-amino-5-phosphonopentanoic acid; CRE, cyclic AMP-response element; CREB, cyclic AMP-response element binding protein; DNQX, 6,7-dinitroquinoxaline-2,3-dione; trkB, tyrosine receptor kinase B; TRKB, gene coding for trkB; UTR, untranslated region; MOPS, 4-morpholinepropanesulfonic acid.mediates numerous functions in both the developing and mature nervous systems, including the survival of postmitotic neurons, axon growth and guidance, and synaptic plasticity (1Huang E.J. Reichardt L.F. Annu. Rev. Neurosci. 2001; 24: 677-736Crossref PubMed Scopus (3258) Google Scholar). These effects of BDNF are mediated by the tyrosine receptor kinase, trkB. Binding of BDNF to trkB initiates dimerization and trans-autophosphorylation of tyrosine residues in the intracellular domain of trkB (2Patapoutian A. Reichardt L.F. Curr. Opin. Neurobiol. 2001; 3: 272-280Crossref Scopus (906) Google Scholar). These phospho-tyrosine residues act as docking sites for effector proteins that activate downstream signaling pathways, leading to the activation of protein kinase cascades, Ca2+ mobilization, and gene expression, which orchestrate the cellular responses to BDNF (3Kaplan D.R. Miller F.D. Curr. Opinion Neurobiol. 2000; 10: 381-391Crossref PubMed Scopus (1644) Google Scholar). Excitatory synaptic input and the resulting elevation in intracellular [Ca2+] have been shown to increase the synthesis and release of BDNF (4Tao X. Finkbeiner S. Arnold D.B. Shaywitz A.J. Greenberg M.E. Neuron. 1998; 20: 709-726Abstract Full Text Full Text PDF PubMed Scopus (1277) Google Scholar, 5Shieh P.B. Hu S.-C. Bobb K. Timmusk T. Ghosh A. Neuron. 1998; 20: 727-740Abstract Full Text Full Text PDF PubMed Scopus (601) Google Scholar, 6Tabuchi A. Sakaya H. Kisukeda T. Fushiki H. Tsuda M J. Biol. Chem. 2002; 277: 35920-35931Abstract Full Text Full Text PDF PubMed Scopus (173) Google Scholar, 7Balkowiec A. Katz D.M. J. Neurosci. 2002; 22: 10399-10407Crossref PubMed Google Scholar, 8Hartmann M. Heumann R. Lessmann V. EMBO J. 2001; 20: 5887-5897Crossref PubMed Scopus (411) Google Scholar, 9Gartner A. Staiger V. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 6386-6391Crossref PubMed Scopus (155) Google Scholar). This BDNF activates trkB receptors in the same or neighboring cells to promote their survival and may also enhance synaptic plasticity (1Huang E.J. Reichardt L.F. Annu. Rev. Neurosci. 2001; 24: 677-736Crossref PubMed Scopus (3258) Google Scholar, 10Schinder A.F. Poo M. Trends Neurosci. 2000; 23: 639-645Abstract Full Text Full Text PDF PubMed Scopus (586) Google Scholar). Although trkB levels change during development and exhibit cell-specific expression patterns (11Escandon E. Soppet D. Rosenthal A. Mendoza-Ramirez J.-L. Szonyi E. Burton L.E. Henderson C.E. Parada L.F. Nikolics K. J. Neurosci. 1994; 14: 2054-2068Crossref PubMed Google Scholar, 12Klein R. Parada L.F. Coulier F. Barbacid M. EMBO J. 1989; 8: 3701-3709Crossref PubMed Scopus (478) Google Scholar, 13Fryer R.H. Kaplan D.R. Feinstein S.C. Radeke M.J. Grayson D.R. Kromer L.F. J. Comp. Neurol. 1996; 374: 21-40Crossref PubMed Scopus (236) Google Scholar), very little is known about the mechanisms that regulate TRKB expression.At least four isoforms of trkB are produced by alternative splicing of the primary transcripts of the TRKB gene (14Middlemas D.S. Lindberg R.A. Hunter T. Mol. Cell. Biol. 1991; 11: 143-153Crossref PubMed Scopus (685) Google Scholar, 15Klein R. Conway D. Parada L.F. Barbacid M. Cell. 1990; 61: 647-656Abstract Full Text PDF PubMed Scopus (618) Google Scholar, 16Stoilov P. Castren E. Stamm S. Biochem. Biophys. Res. Comm. 2002; 290: 1054-1065Crossref PubMed Scopus (130) Google Scholar). Of these, only the full-length isoform, which contains an intracellular tyrosine kinase domain, is known to be capable of mediating BDNF signaling. Three truncated isoforms (T1, T2, and Tshc), which lack the intracellular kinase domain but possess the same extracellular BDNF binding domain as full-length receptors, can also be generated by alternative splicing. T1 is prominently expressed in the brain (14Middlemas D.S. Lindberg R.A. Hunter T. Mol. Cell. Biol. 1991; 11: 143-153Crossref PubMed Scopus (685) Google Scholar) and can act as a dominant negative inhibitor of BDNF signaling (17Eide F.F. Vining E.R. Eide B.L. Zang K. Wang X.-Y. Reichardt L.F. J. Neurosci. 1996; 16: 3123-3129Crossref PubMed Google Scholar, 18Li Y.-X. Xu Y. Ju D. Lester H.A. Davidson N. Schuman E.M. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 10884-10889Crossref PubMed Scopus (133) Google Scholar, 19Gonzalez M. Ruggiero F.P. Chang Q. Shi Y.J. Rich M.M. Kraner S. Balice-Gordon R.J. Neuron. 1999; 24: 567-583Abstract Full Text Full Text PDF PubMed Scopus (190) Google Scholar, 20Dorsey S.G. Bambrick L.L. Balice-Gordon R.J. Krueger B.K. J. Neurosci. 2002; 22: 2571-2578Crossref PubMed Google Scholar, 21Haapasalo A. Sipola I. Larsson K. Akerman K.E.O. Stoilov P. Stamm S. Wong G. Castren E. J. Biol. Chem. 2002; 277: 43160-43167Abstract Full Text Full Text PDF PubMed Scopus (122) Google Scholar) by forming heterodimers with full-length trkB (17Eide F.F. Vining E.R. Eide B.L. Zang K. Wang X.-Y. Reichardt L.F. J. Neurosci. 1996; 16: 3123-3129Crossref PubMed Google Scholar). These observations raise the possibility that the relative expression of full-length and truncated trkB isoforms in normal neurons can modulate cellular responsiveness to BDNF. Dysregulation of trkB isoform expression may also underlie some nervous system abnormalities. For example, overexpression of truncated trkB has been reported in cortical neurons in Alzheimer's disease brain (22Ferrer I. Marin C. Rey M.J. Ribalta T. Goutan E. Blanco R. Tolosa E. Marti E. J. Neuropathol. Exp. Neurol. 1999; 58: 729-739Crossref PubMed Scopus (313) Google Scholar), where it may contribute to neurodegeneration, and in the trisomy 16 mouse model of Down syndrome, where it results in failure of BDNF-mediated neuron survival (20Dorsey S.G. Bambrick L.L. Balice-Gordon R.J. Krueger B.K. J. Neurosci. 2002; 22: 2571-2578Crossref PubMed Google Scholar).The TRKB gene can be transcribed from two promoters, P1 and P2 (23Barettino D. Pombo P.M.G. Espliguero G. Rodriguez-Pena A. Biochim. Biophys. Acta. 1999; 1446: 24-34Crossref PubMed Scopus (26) Google Scholar). Within the TRKB upstream sequence are multiple potential regulatory elements, including several Ca2+/cAMP-response elements (CREs), suggesting that elevated Ca2+ and/or cAMP may regulate TRKB expression.In this report, we demonstrate that depolarization alters the relative expression of full-length and truncated trkB receptors in cultured cortical neurons and identify Ca2+-dependent regulatory elements in the TRKB promoters involved in this response.EXPERIMENTAL PROCEDURESCell Culture and Transfection—Cortical neurons were isolated from embryonic day-16 mouse embryos and plated at a density of 6 × 105/well in 24-well dishes for luciferase assays and 3 × 106/well in 35-mm dishes for RNA and protein analysis. Neuron cultures were maintained in Neurobasal medium supplemented with 2% B27, 2 mm glutamine, and penicillin-streptomycin and incubated in 5% CO2 at 37 °C. Neurons were transiently transfected 3–5 days after plating using a Ca2+-phosphate protocol (24Xia Z. Dudek H. Miranti C.K. Greenberg M.E. J. Neurosci. 1996; 16: 5425-5436Crossref PubMed Google Scholar). TRKB luciferase reporters were introduced at a concentration of 1 μg/well; cells were cotransfected with 0.5 μg/well TK Renilla plasmid (Promega). All cell culture reagents were purchased from Invitrogen.Plasmids—Luciferase reporter genes were generated by inserting TRKB promoter fragments upstream of the luciferase gene in pGL3Basic (Promega) as BglII-BamHI fragments into the BglII site of the vector. Introduction of the BamHI site to clone the P2 promoters into pGL3 Basic regenerates the AT of the ATG translation start codon, therefore placing the 3′-end of the luciferase constructs at +2. Multiple independent PCR promoter products were sequenced and found to contain differences from the previously published sequence (23Barettino D. Pombo P.M.G. Espliguero G. Rodriguez-Pena A. Biochim. Biophys. Acta. 1999; 1446: 24-34Crossref PubMed Scopus (26) Google Scholar), including a one-base deletion that shifts the transcription start site for P1 from -1800 to -1799. All but one base substitution could be independently verified from Trace sequences available at Ensembl. The sequence of the 2560 bp used in these studies is available at GenBank™, accession number AY307416.PCR primers used to generate promoter fragments were: upstream primers -2558, 5′-AAAGATCTCATCTATGTGAAAATCTTG-3′; -2258, 5′-AGATCTCGGTGGTAGCAATGGC-3′; -1429, 5′-AGATCTCCTATGAGCATGGTGAG-3′; -944, 5′-AAAGATCTGGAGTTTCTGCCCC-3′; and 899, 5′-AGATCTGCCAGCAGTAGCAGAG-3′. Downstream primers were: -1710, 5′-AAGGATCCTAAATGCTTTGCACCGACC-3′, and +2, 5′-AAGGATCCCGAGCTGCCAGTGCC-3′.CRE site mutations were generated using site-directed mutagenesis. Forward primer sequences were: cre1-, 5′-TGGAGTTTCTGCCCCTGCTCCACTGCAGCCCTCACGT-3′; cre2-, 5′-CCTGCTCTGCGTCAGCCCCAGCTGCACTTCGCCAGCAGTAG-3′; and cre1-2-, 5′-TGGAGTTTCTGCCCCTGCTCCACTGCAGCCCCAGCTGCACTTCGCCAGCAGTAG-3′. Dominant negative CREB plasmid was provided by Dr. Yibin Wang, University of Maryland School of Medicine.Protein Analysis—Cells were harvested directly into boiling 2× sample buffer and fractionated by SDS-PAGE on 4–12% gels run in MOPS buffer as previously described (20Dorsey S.G. Bambrick L.L. Balice-Gordon R.J. Krueger B.K. J. Neurosci. 2002; 22: 2571-2578Crossref PubMed Google Scholar). Western blotting was conducted with ECL (Amersham Biosciences) using antibodies to the extracellular domain of trkB (BD Transduction Laboratories, Fig. 1A, or Santa Cruz Biotechnology, H-181, Fig. 1B) at 1:500, anti-phospho-trkA (Tyr-490; Cell Signaling Technologies) 1:500, and anti-actin (Sigma) 1:5000.RNA Isolation and Analysis—Cells were stimulated for 5 h by addition of 50 mm KCl. Following stimulation, neurons were harvested and homogenized using Qiashredder (Qiagen), and RNA was isolated using RNeasy RNA extraction kit (Qiagen). RNA was quantified by absorbance at 260 nm, and 0.5 μg was used for reverse transcription with Superscript II (Invitrogen) for 50 min at 42 °C. PCR was conducted in an Opticon real-time PCR cycler (MJ Designs) using Platinum Taq (Invitrogen), 0.2 mm dNTP, 30 μm/primer, and 3 mm MgCl2. Each PCR cycle consisted of 1 min at 94 °C, 30 s at 62 °C, followed by 1 min at 72 °C. Amplification was monitored using the fluorescent dye, Sybergreen I (Roche Applied Science), diluted 1:100,000. For real-time PCR analysis, each cDNA sample (untreated versus KCl-treated) was diluted 1:100 and assayed in duplicate at 1, .5, .25, and .125× dilutions. RNA expression was computed from the slope of the C T versus ln [cDNA] relation and normalized to the concentration of β-actin as amplified using 5′-ATCGTGGGCCGCCCTAGGCA-3′ and 5′-TGGCCTTAGGGTTCAGAGGGG-3′ (25Pfaffl M.W. Nucleic Acids Res. (2001). 2001; 29: 2002-2009Crossref Scopus (24916) Google Scholar). Full-length-specific primers were: 5′-GACAATGCACGCAAGGACTT-3′ and 5′-AGTAGTCGGTGCTGTACACA-3′. T1-specific primers were: 5′-ATAAGATCCCACTGGATGGG-3′ and 5′-CGTATAGTCAAACAGCTCGC-3′. Data are reported as the normalized RNA concentration from KCl-stimulated neurons relative to that of unstimulated neurons.Conventional RT-PCR to detect short 5′-UTRs of P1 and P2 was conducted using upstream primers in the unique 5′-UTR sequences (23Barettino D. Pombo P.M.G. Espliguero G. Rodriguez-Pena A. Biochim. Biophys. Acta. 1999; 1446: 24-34Crossref PubMed Scopus (26) Google Scholar) as follows: P1-1, 5′-AGGGTCGGTGCAAAGCATTT-3′; P1-2, 5′-TTAGGGACCAAGGAAGCATC-3′; P1-3, 5′-AGTTTCTGCCCCTGCTCTG-3′; or P2-4, 5′-AGCGCGGAGGGACTGTGT-3′ with the common downstream primer 5′-TCTTGCTGCTTGGTGCTGG-3′. The PCR amplification protocol consisted of 40 cycles of 1 min at 94 °C, 30 s at 60 °C, followed by 30 s at 72 °C.Luciferase Assays—Two days after transfection with TRKB luciferase reporter constructs, cells were stimulated for 6 h by the addition of 50 mm KCl in the absence or presence of 2 mm EGTA. EGTA was added 5 min prior to KCl stimulation. Cells were washed with phosphate-buffered saline and harvested in 150 μlof Renilla luciferase lysis buffer (Promega). Twenty μl of extract were used to measure TRKB reporters (luciferase assay reagent; Promega) and Renilla luciferase activity (Renilla luciferase assay system; Promega). TRKB luciferase activity was then normalized by dividing by Renilla activity to allow comparison among wells and stimulation conditions. Cells transfected with luciferase vector lacking the TRKB promoter exhibited less than 3% of the activity of TRKB-luciferase-transfected neurons, and this activity was unaffected by depolarization.RESULTSDepolarization Increases the Level of Full-length trkB in Cortical Neurons—Because of the importance of BDNF/trkB signaling in activity-dependent changes in neurons, we investigated the ability of depolarization to regulate trkB expression. Embryonic mouse cortical neurons grown in culture for 5–7 days were depolarized with 50 mm added KCl to induce Ca2+ influx. Cells were then harvested for Western blot analysis. As previously reported for hippocampal neurons (20Dorsey S.G. Bambrick L.L. Balice-Gordon R.J. Krueger B.K. J. Neurosci. 2002; 22: 2571-2578Crossref PubMed Google Scholar), cortical neurons expressed primarily full-length trkB (Fig. 1). In the presence of 50 mm added KCl, the level of full-length trkB protein was elevated by 5 h and continued to increase up to 16 h, the longest time studied (Fig. 1A). The low level of truncated trkB did not change following KCl treatment. Depolarization also increased the amount of phospho-trkB observed in response to BDNF stimulation (Fig. 1B), demonstrating that the additional full-length trkB was functional.Depolarization Increases the Level of mRNA Encoding Full-length trkB in Cortical Neurons—In light of the effects of depolarization on full-length trkB expression, real-time PCR was conducted to determine the effect of depolarization on TRKB RNA expression. Following reverse transcription of total RNA, TRKB transcripts were analyzed using primer pairs specific for either full-length or truncated T1 trkB isoforms (Fig. 2A) (25Pfaffl M.W. Nucleic Acids Res. (2001). 2001; 29: 2002-2009Crossref Scopus (24916) Google Scholar). Five hours of stimulation with elevated KCl produced an ∼3-fold increase in full-length trkB transcripts, whereas there was no increase in T1 transcripts (Fig. 2B). The increase in full-length transcripts was sensitive to EGTA (data not shown).Fig. 2Depolarization induces full-length trkB mRNA expression. A, diagram illustrating the location of isoform-specific primers (arrows) for full-length and truncated (T1) trkB mRNAs. Both isoforms share a common extracellular BDNF binding domain and transmembrane domain (TM) but have different intracellular domains and their mRNAs have distinct 5′- and 3′-UTRs. B, cortical neurons were incubated for 5 h in the absence or presence of 50 mm added KCl. Levels of trkB RNA were quantified using real-time PCR as described under "Experimental Procedures" using primer pairs specific for RNA encoding either full-length (trkB.FL) or T1 truncated (trkB.TR) receptors (arrows in panel A). Data shown are mean ± S.E. (n = 4 experiments) mRNA levels in the presence of elevated KCl relative to unstimulated control.View Large Image Figure ViewerDownload Hi-res image Download (PPT)Ca2+ Regulates TRKB Expression—The presence of three CRE sites in the TRKB promoter region (23Barettino D. Pombo P.M.G. Espliguero G. Rodriguez-Pena A. Biochim. Biophys. Acta. 1999; 1446: 24-34Crossref PubMed Scopus (26) Google Scholar) suggested that Ca2+ and/or cAMP can modulate TRKB. Ca2+-dependent TRKB transcription was investigated using a TRKB-luciferase reporter gene. Approximately 2.5 kb of the TRKB promoter region, including both P1 and P2, was cloned upstream of the luciferase gene and transiently transfected into cortical neurons using a Ca2+-phosphate method (24Xia Z. Dudek H. Miranti C.K. Greenberg M.E. J. Neurosci. 1996; 16: 5425-5436Crossref PubMed Google Scholar). Depolarization of the neurons by the addition of 50 mm KCl resulted in a 2-fold increase in TRKB-dependent transcription as measured by luciferase activity (Fig. 3A). The presence of EGTA eliminated the response to KCl, indicating that Ca2+ influx was required to stimulate TRKB-luciferase expression.Fig. 3Ca2+ entry stimulates TRKB transcription. A, cortical neurons transiently transfected with -2558/+2 TRKB-luciferase plasmid were stimulated for 6 h with elevated KCl in the absence or presence of 2 mm extracellular EGTA. Luciferase activity was assayed as described under "Experimental Procedures" and is reported relative to unstimulated activity. B, transfected neurons were stimulated as in panel A in the absence or presence of 100 μm nifedipine or 80 μm APV plus 20 μm DNQX. Data shown are means ± S.E. (n = 4). The same results were obtained in five (A) or two (B) experiments.View Large Image Figure ViewerDownload Hi-res image Download (PPT)To determine the pathway of Ca2+ entry in response to depolarization, neurons transfected with the -2558/+2 luciferase plasmid were stimulated with KCl in the absence and presence of Ca2+ channel or glutamate receptor (GluR) blockers (Fig. 3B). The l-type Ca2+ channel blocker, nifedipine (100 μm), completely blocked depolarization-induced activation of TRKB luciferase activity, whereas the combination of an NMDA GluR blocker (APV, 80 μm) and a kainate/AMPA GluR blocker (DNQX, 20 μm) had no effect. Control luciferase activity was not affected by EGTA, nifedipine, or APV/DNQX. Thus, activation of TRKB results from depolarization-induced entry of Ca2+ through l-type, voltage-gated Ca2+ channels and does not require GluR activation.P1 and P2 TRKB Promoters Are Differentially Regulated by Ca2+—Because the upstream regulatory region of TRKB has been shown to contain two promoters, the ability of Ca2+ to separately activate P1 and P2 was tested using TRKB-luciferase reporter genes containing either P1 or P2 (Fig. 4A). Two constructs, -2558/-1710 and -2258/-1710 were generated to test the effects of Ca2+ on P1. The shorter construct lacked a potential Ca2+-dependent regulatory site (CRE) located at -2480 within the P1 domain of TRKB. In contrast to the -2558/+2 construct, which contains both P1 and P2 (Fig. 3), the luciferase activity of the P1 constructs was reduced by ∼50% following 6 h of depolarization (Fig. 4B). Inclusion of EGTA abolished the inhibition, consistent with a requirement for Ca2+ influx in the inhibition of P1-dependent transcription. The finding that Ca2+ inhibited the -2558/-1710 and -2258/-1710 constructs to the same extent indicates that the CRE in P1 is not required for inhibition.Fig. 4Ca2+ differentially regulates P1 and P2. A, schematic of TRKB promoters and corresponding TRKB reporter constructs. CRE sites are indicated by hatching. Transcription start sites for P1 (-1799) and P2 (-448) are indicated. B and C, cortical neurons transiently transfected with P1 (B) or P2 (C) TRKB-luciferase reporters were stimulated for 6 h with 50 mm added KCl in the absence or presence of EGTA. Luciferase activity was assayed as described under "Experimental Procedures" and is reported relative to unstimulated activity. Data shown are mean luciferase ± S.E. (n = 4). Similar results were obtained in at least three experiments. D, RT-PCR analysis of cultured cortical neurons to detect P1- and P2-specific 5′-UTR sequences. Reactions 1–3 selectively amplify the three P1 5′-UTR splice variants. Reaction 4 detects the P2-specific 5′-UTR.View Large Image Figure ViewerDownload Hi-res image Download (PPT)The regulation of P2 was investigated using three constructs, -1429/+2, -944/+2, and -899/+2, which were transiently transfected into cortical neurons. In contrast to the P1 constructs, expression of P2 luciferase constructs was stimulated by KCl depolarization. Depolarization caused ∼3- and 2-fold increases in activity of the -1429/+2 and -944/+2 luciferase reporter constructs, respectively (Fig. 4C). Increased P2 activity was blocked by extracellular EGTA, demonstrating the requirement for Ca2+ influx to stimulate P2 TRKB expression. Expression of the -899/+2 TRKB P2-luciferase construct was not stimulated by depolarization, indicating that promoter elements between -944 and -899, possibly the tandem CRE sites, are required for the Ca2+ dependence of TRKB expression (see below). The larger degree of Ca2+-dependent stimulation observed with the -1429/+2 TRKB-luciferase construct as compared with the -944/+2 construct revealed the presence of an additional regulatory element(s) between -1429 and -944 capable of enhancing Ca2+-induced TRKB expression.Both P1- and P2-derived Transcripts Are Present in Cortical Neurons—The transcriptional activity of endogenous TRKB was investigated using RT-PCR analysis of P1- and P2-specific 5′-UTR sequences (23Barettino D. Pombo P.M.G. Espliguero G. Rodriguez-Pena A. Biochim. Biophys. Acta. 1999; 1446: 24-34Crossref PubMed Scopus (26) Google Scholar). Alternative splicing of the P1-derived transcript produces three potential 5′-UTRs (Fig. 4D, lanes 1–3), whereas P2 generates a single 5′-UTR (Fig. 4D, lane 4). PCR products corresponding to both P1- and P2-derived transcripts were readily detectable, indicating that TRKB is transcribed from both promoters in mouse embryonic cortical neurons.Both CRE Sites Are Required for Ca2+-stimulated Expression of TRKB P2—The sequence between -944 and -899 contains a pair of CRE sites separated by 4 bp (Fig. 5A). CRE sites are binding sites for members of the CREB family of transcription factors (26Sheng M. McFadden G. Greenberg M.E. Neuron. 1990; 4: 571-582Abstract Full Text PDF PubMed Scopus (873) Google Scholar), which mediate Ca2+-dependent expression of a wide variety of genes, including BDNF. The role of these tandem CRE sites in mediating P2-initiated TRKB expression following depolarization was examined by introducing substitution mutations (Fig. 5A). Mutation of either CRE site or both in the -1429/+2 TRKB-luciferase construct completely blocked stimulation of luciferase activity (Fig. 5B). Ca2+-stimulated expression of TRKB, therefore, requires both CRE sites. Mutation of either CRE site in the -944/+2 TRKB-luciferase construct similarly abolished depolarization-stimulated expression (data not shown).Fig. 5Tandem CRE sites are required for Ca2+-stimulated P2 expression. A, sequence of the tandem CRE elements in P2 with corresponding point mutations used to inactivate them. B, cortical neurons transiently transfected with the -1429/+2 TRKB-luciferase reporter containing either wild type or mutated CRE sites were stimulated for 6 h in 50 mm added KCl in the absence or presence of 2 mm EGTA. Luciferase activity was assayed as described under "Experimental Procedures" and is plotted relative to unstimulated activity. C, cortical neurons cotransfected with either dominant negative (DN) CREB or control vector were stimulated and assayed as in panel B. Data shown are means ± S.E. (n = 4). Similar results were obtained in at least two additional experiments.View Large Image Figure ViewerDownload Hi-res image Download (PPT)Mutation analysis of the CRE sites suggested that CREB, or a closely related family member, is responsible for mediating the Ca2+-dependent induction of TRKB expression. Cotransfection of a dominant negative CREB construct together with the -944/+2 TRKB-luciferase construct inhibited the Ca2+-stimulated expression of the TRKB reporter (Fig. 5C), consistent with a role for a CREB-related transcription factor in the induction of TRKB P2 by Ca2+.cAMP Stimulates TRKB Gene Expression—Because CREB is also activated in response to cAMP, we investigated the ability of cAMP to activate the TRKB P2 promoter. Cortical neurons transfected with either the -1429/+2 or -944/+2 TRKB-luciferase reporter gene were treated with 50 μm forskolin, an adenylate cyclase activator, for 6 h in either the absence or presence of 50 mm added KCl. Forskolin stimulated the luciferase activity of each TRKB reporter ∼2-fold, demonstrating that P2 can be stimulated by cAMP signaling (Fig. 6). The effects of depolarization-induced Ca2+ signaling and increased cAMP were additive for both -944/+2 and -1429/+2 P2-dependent reporter genes when the neurons were simultaneously treated with elevated KCl and forskolin.Fig. 6Cyclic AMP stimulation of TRKB expression. Cortical neurons transiently transfected with -1429/+2 TRKB-luciferase construct were stimulated with elevated KCl and/or 50 μm forskolin as indicated. Luciferase activity was assayed as described under "Experimental Procedures" and is plotted relative to unstimulated activity. Data shown are mean luciferase ± S.E. (n = 4).View Large Image Figure ViewerDownload Hi-res image Download (PPT)DISCUSSIONDepolarization resulted in increased expression of full-length trkB protein and increased phosphorylation of trkB upon BDNF stimulation (Fig. 1). Depolarization also preferentially increased the level of endogenous full-length trkB mRNA without significantly affecting the level of truncated trkB message (Fig. 2B). These results led us to investigate the role of Ca2+ in the transcriptional regulation of TRKB. We show here that the two promoters of TRKB are differentially regulated by Ca2+. P1 reporter constructs were inhibited by Ca2+, whereas P2 reporters were stimulated by Ca2+ (Fig. 4, B and C). Although the quantitative contribution of P1- and P2-derived transcripts to specific trkB isoform expression has not yet been established, these observations suggest that Ca2+-regulated TRKB promoter selection can alter t
Referência(s)