Neuronal κB-binding Factors Consist of Sp1-related Proteins
2002; Elsevier BV; Volume: 277; Issue: 47 Linguagem: Inglês
10.1074/jbc.m204292200
ISSN1083-351X
AutoresXianrong Mao, A. Moerman, Steven W. Barger,
Tópico(s)Neuroinflammation and Neurodegeneration Mechanisms
ResumoNeurons contain a protein factor capable of binding DNA elements normally bound by the transcription factor NF-κB. However, several lines of evidence suggest that this neuronal κB-binding factor (NKBF) is not bona fide NF-κB. We have identified NKBF from cultures of neocortical neurons as a complex containing proteins related to Sp1. This complex was bound by antibodies to Sp1, Sp3, and Sp4 and was competed from binding to an NF-κB element by an oligonucleotide containing an Sp1-binding site. This Sp1 oligonucleotide detected an abundant factor in neuronal nuclei that migrated in electrophoretic mobility shift assays at a position consistent with NKBF. Expression of transfected Sp1 stimulated transcription in a manner dependent upon a κBcis-element. Similar to our previous reports for NKBF (Mao, X., Moerman, A. M., Lucas, M. M., and Barger, S. W. (1999) J. Neurochem.73, 1851–1858 and Moerman, A. M., Mao, X., Lucas, M. M., and Barger, S. W. (1999) Mol. Brain Res.67, 303–315), the activity of the Sp1-related factor was reduced by activation of ionotropic glutamate receptors, consistent with proteolytic degradation of all three Sp1-related factors. Expression of the N-methyl-d-aspartate receptor-1 (NR1) subunit of glutamate receptors correlated with the activity of the Sp1-related factor, specifically through an Sp1 element in the NR1 promoter. These data provide the first evidence that Sp1 or related family members are responsible for κB-binding activity and are involved in a negative feedback for NR1 in central nervous system neurons. Neurons contain a protein factor capable of binding DNA elements normally bound by the transcription factor NF-κB. However, several lines of evidence suggest that this neuronal κB-binding factor (NKBF) is not bona fide NF-κB. We have identified NKBF from cultures of neocortical neurons as a complex containing proteins related to Sp1. This complex was bound by antibodies to Sp1, Sp3, and Sp4 and was competed from binding to an NF-κB element by an oligonucleotide containing an Sp1-binding site. This Sp1 oligonucleotide detected an abundant factor in neuronal nuclei that migrated in electrophoretic mobility shift assays at a position consistent with NKBF. Expression of transfected Sp1 stimulated transcription in a manner dependent upon a κBcis-element. Similar to our previous reports for NKBF (Mao, X., Moerman, A. M., Lucas, M. M., and Barger, S. W. (1999) J. Neurochem.73, 1851–1858 and Moerman, A. M., Mao, X., Lucas, M. M., and Barger, S. W. (1999) Mol. Brain Res.67, 303–315), the activity of the Sp1-related factor was reduced by activation of ionotropic glutamate receptors, consistent with proteolytic degradation of all three Sp1-related factors. Expression of the N-methyl-d-aspartate receptor-1 (NR1) subunit of glutamate receptors correlated with the activity of the Sp1-related factor, specifically through an Sp1 element in the NR1 promoter. These data provide the first evidence that Sp1 or related family members are responsible for κB-binding activity and are involved in a negative feedback for NR1 in central nervous system neurons. Agonists often stimulate down-regulation of their receptors as a mechanism providing negative feedback. Glutamate plays a major role in many fundamental aspects of neurotransmission in the central nervous system (1Lipton S.A. Rosenberg P.A. N. Engl. J. Med. 1994; 330: 613-622Google Scholar). On the other hand, it is also well known that glutamate contributes to many neuropathological states. Glutamate is able to kill many types of neurons through the phenomenon of excitotoxicity (2Choi D.W. J. Neurosci. 1987; 7: 369-379Google Scholar). By activating its ionotropic receptors, especially those of the NMDA 1The abbreviations used are: NMDA, N-methyl-d-aspartate; NKBF, neuronal κB-binding factor; NR1, N-methyl-d-aspartate receptor-1; HIV, human immunodeficiency virus; IL6, interleukin; RBP-Jκ, recombination signal binding protein Jκ; RT, reverse transcription; EMSA, electrophoretic mobility shift assay. class, glutamate can generate an ion flux and free radical production so robust that cellular homeostatic mechanisms are overwhelmed. In a population of cortical neurons challenged with glutamate, a fraction die by rapid osmotic stress; however, others die 18–24 h later (2Choi D.W. J. Neurosci. 1987; 7: 369-379Google Scholar). This later cell death bears signs of apoptosis (2Choi D.W. J. Neurosci. 1987; 7: 369-379Google Scholar, 3Ankarcrona M. Dypbukt J.M. Bonfoco E. Zhivotovsky B. Orrenius S. Lipton S.A. Nicotera P. Neuron. 1995; 15: 961-973Google Scholar) and is dependent on an early commitment phase. The commitment appears to involve calcium influx (4Choi D.W. Trends Neurosci. 1988; 11: 465-469Google Scholar) and a mitochondrial membrane permeability transition (5Schinder A.F. Olson E.C. Spitzer N.C. Montal M. J. Neurosci. 1996; 16: 6125-6133Google Scholar, 6Reynolds I.J. Ann. N. Y. Acad. Sci. 1999; 893: 33-41Google Scholar), but is otherwise incompletely understood. The delayed aspect of excitotoxicity permits consideration of changes in gene expression. The number of genes altered by glutamate receptor activation is also consistent with transcriptional contributions to the delayed component excitotoxicity. Unfortunately, the influence of abusive glutamate receptor stimulation on transcription is poorly characterized. Some reports have suggested that the transcription factor NF-κB is an important target for glutamate (7Guerrini L. Blasi F. Denis-Donini S. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 9077-9081Google Scholar, 8Kaltschmidt C. Kaltschmidt B. Baeuerle P.A. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 9618-9622Google Scholar, 9Grilli M. Goffi F. Memo M. Spano P. J. Biol. Chem. 1996; 271: 15002-15007Google Scholar). However, no defined role of NF-κB in neuronal death has been demonstrated in excitotoxic paradigms, and emerging data suggest that NF-κB might play completely different roles under different neurotoxic conditions (10Behl C. J. Neural Transm. 2000; 107: 393-407Google Scholar, 11Qin Z. Wang Y. Chasea T.N. Mol. Brain Res. 2000; 80: 111-122Google Scholar, 12Yabe T. Wilson D. Schwartz J.P. J. Biol. Chem. 2001; 276: 43313-43319Google Scholar, 13Bachis A. Colangelo A.M. Vicini S. Doe P.P. De Bernardi M.A. Brooker G. Mocchetti I. J. Neurosci. 2001; 21: 3104-3112Google Scholar). Glutamate is able to activate NF-κB in mouse and rat cerebellar cultures maintained in serum-containing medium (7Guerrini L. Blasi F. Denis-Donini S. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 9077-9081Google Scholar, 8Kaltschmidt C. Kaltschmidt B. Baeuerle P.A. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 9618-9622Google Scholar, 9Grilli M. Goffi F. Memo M. Spano P. J. Biol. Chem. 1996; 271: 15002-15007Google Scholar). However, by rigorously controlling glial numbers, we found that the prominent κB-binding factor in cultured cortical neurons is composed of proteins distinct from bona fide NF-κB (14Mao X. Moerman A.M. Lucas M.M. Barger S.W. J. Neurochem. 1999; 73: 1851-1858Google Scholar, 15Moerman A.M. Mao X. Lucas M.M. Barger S.W. Mol. Brain Res. 1999; 67: 303-315Google Scholar). This neuronal κB-binding factor (NKBF) is distinct from NF-κB with regard to immunoreactivity, DNA binding preferences, and electrophoretic mobility under both denaturing and nondenaturing conditions. An additional distinction is that glutamate application diminishes NKBF activity; NF-κB is activated by glutamate only in cultures containing significant numbers of glia (14Mao X. Moerman A.M. Lucas M.M. Barger S.W. J. Neurochem. 1999; 73: 1851-1858Google Scholar, 15Moerman A.M. Mao X. Lucas M.M. Barger S.W. Mol. Brain Res. 1999; 67: 303-315Google Scholar). NMDA receptors and calcium were involved in the diminution of NKBF by glutamate, and other data provided correlations between NKBF suppression and excitotoxicity (14Mao X. Moerman A.M. Lucas M.M. Barger S.W. J. Neurochem. 1999; 73: 1851-1858Google Scholar). In light of the fact that Sp1 can compete with Rel factors for κB enhancers (16Hirano F. Tanaka H. Hirano Y. Hiramoto M. Handa H. Makino I. Scheidereit C. Mol. Cell. Biol. 1998; 18: 1266-1274Google Scholar), we explored the contribution of Sp1 family transcription factors to NKBF. Antibody reactivity and probe competition assays indicated that Sp1, Sp3, and Sp4 contribute to NKBF binding activity. Glutamate was able to evoke a substantial reduction in Sp1-related factors in a time- and dose-dependent manner, as detected by DNA binding assays and immunoblot analysis of the Sp1 family polypeptide levels. More importantly, glutamate was capable of influencing expression of NMDA receptor-1 (NR1) expression through the modulation of Sp1-related factors, consistent with previous demonstrations of a role for Sp1 in NR1 transcriptional regulation (17Bai G. Kusiak J.W. J. Biol. Chem. 1995; 270: 7737-7744Google Scholar). Ig/HIV-κB, Sp1, and AP-1 oligonucleotides used in DNA binding assays were supplied by Promega (Madison, WI). Phosphorothioate-modified oligonucleotides used for decoy experiments were synthesized by Oligos Etc. (Wilsonville, OR) with the following sequences: decoy, 5′-TTCAGGGGCCTCCCATACG-3′ (Sp site is underlined); and scrambled control, 5′-GACCATGTCGTCCAGTCGC-3′. Invitrogen supplied all other oligonucleotides, including Ig/HIV-κΒ in assays where it was compared with other probes. The IL6-κB enhancer sequence (AATGTGGGATTTTCCCATGA, κB site is underlined) is derived from the human interleukin-6 gene promoter. Santa Cruz Biotechnology (Santa Cruz, CA) supplied antibodies (listed with their catalog numbers) to Rel family and Sp1 family polypeptides: p50 (sc-114), RelA or p65 (sc-372), Sp1 (sc-59x), Sp3 (sc-644x), and Sp4 (sc-645x). Monoclonal antibody K0043 against RBP-Jκ was generously provided by Dr. T. Honjo (Kyoto University) (18Sakai T. Furukawa T. Iwanari H. Oka C. Nakano T. Kawaichi M. Honjo T. J. Biochem. (Tokyo). 1995; 118: 621-628Google Scholar). Antibody against NR1 (AB1516) was obtained from Chemicon International, Inc. (Temecula, CA). Reverse transcription (RT)-PCR kits were from Clontech (Palo Alto, CA). LipofectAMINE and all enzymes used for cloning were purchased from Invitrogen. Trireagent® is a product of Molecular Research Center, Inc. (Cincinnati, OH). The pGL3-prom reporter plasmid, the pRL-CMV Renillaluciferase expression vector, and the dual-luciferase reporter assay system were obtained from Promega. pPAC-Sp1, pPAC-Sp3, and the control vector pPAC were generously provided by Dr. J. M. Horowitz (North Carolina State University) (19Kennett S.B. Udvadia A.J. Horowitz J.M. Nucleic Acids Res. 1997; 25: 3110-3117Google Scholar); pPAC-Sp4 was obtained from Dr. G. Suske (Philipps-Universitat, Marburg, Germany) (20Hagen G. Dennig J. Preiss A. Beato M. Suske G. J. Biol. Chem. 1995; 270: 24989-24994Google Scholar). Luciferase reporter plasmids pNRL239, pNRL356, and pNRL356mtSp1×2 were generously provided by Dr. G. Bai (17Bai G. Kusiak J.W. J. Biol. Chem. 1995; 270: 7737-7744Google Scholar, 21Bai G. Kusiak J.W. J. Biol. Chem. 1997; 272: 5936-5942Google Scholar). These reporters are derivatives of pGL2-Basic (Promega) with insertions from the NR1 gene promoter. The pNRL239 plasmid contains NR1 promoter region −239 to −1 (translation initiation site as +1). The pNRL356 plasmid contains NR1 promoter region −356 to −1; two Sp1 sites located at −299 to −267 have been specifically mutated in pNRL356mtSp1×2. N9 cells were generously provided by Prof. P. Ricciardi-Castagnoli (22Corradin S.B. Mauel J. Donini S.D. Quattrocchi E. Ricciardi-Castagnoli P. Glia. 1993; 7: 255-262Google Scholar); they were maintained in minimal essential medium with Earle's salts (Invitrogen) supplemented with 10% fetal bovine serum (Invitrogen). Drosophila melanogaster SL2 cells (obtained from Dr. J. C. Azizkhan, Roswell Cancer Institute, Buffalo, NY) were cultured at room temperature in Schneider'sDrosophila medium (Invitrogen) supplemented with 12.5% fetal bovine serum and 10 μg/ml gentamycin sulfate. Primary neuronal cultures were established from the neocortices of 18-day Sprague-Dawley rat embryos as described previously (14Mao X. Moerman A.M. Lucas M.M. Barger S.W. J. Neurochem. 1999; 73: 1851-1858Google Scholar) and were maintained in Neurobasal medium containing B27 supplement (Neurobasal/B27 medium; Invitrogen); 10 μm cytosine arabinoside was present during the first 4 days to suppress glial growth. All neuronal cultures were 8–10 days old at the time of treatment. To minimize background in experiments on expression of Sp1-related factors, the reporter construct pGL3-prom was modified by removal of the six Sp1 sites in its SV40 promoter. PCR was used to amplify the portion of the SV40 promoter downstream of the Sp1 sites: forward primer, 5′-CGAGCTCAGATCTCCCCATCGCTGACTAATT-3′ (the underlinedBglII site was included to facilitate subsequent cloning); and reverse primer, 5′-AAATAACGCGCCCAACACCG-3′. The amplimer was digested by BglII and BamHI, and then the Sp1-less SV40 promoter was used to replace the intact SV40 promoter. The resulting reporter gene with an Sp1-less SV40 promoter was termed pGL3-promΔ. To create a κB reporter construct (pβcE1-GL3), pGL3-promΔ received four copies of the κB element from the human β-amyloid precursor protein promoter, supplied by two insertions of the tandem oligonucleotide cassette: 5′-GGGGTTTCACCG TGT TAG CCA GAC GGGGTTTCACCG A-3′ (κB site is underlined). As a control for efficiency of transfection and survival, primary neurons were cotransfected with pRL-CMV, a plasmid encodingRenilla luciferase under the constitutive control of a cytomegalovirus promoter. When pRL-CMV was used as a transfection control in SL2 cells, the expression of Sp1-related factors consistently elevated Renilla luciferase activity. Therefore, a second control reporter was constructed in which theRenilla luciferase coding sequence replaced the firefly luciferase coding sequence located between NcoI andXbaI in pGL3-promΔ. The resulting reporter was pRL-promΔ, which was used as an efficiency control in experiments with SL2 cells. The specific protocols have been described previously (15Moerman A.M. Mao X. Lucas M.M. Barger S.W. Mol. Brain Res. 1999; 67: 303-315Google Scholar). Antibodies and unlabeled oligonucleotides were added to the binding reaction prior to addition of the radiolabeled NF-κB or Sp1 consensus oligonucleotide. For EMSAs including antibodies, the 15-min incubation period was extended to 60 min at 4 °C in the presence of antibody. For Western blot analysis, highly enriched neocortical neurons were treated with glutamate (50 μm) for the times indicated. After glutamate treatment, cultures were lysed in 80 mm Tris-HCl (pH 6.8) and 2.5% SDS, and total proteins were harvested for Western blot analysis as described previously (15Moerman A.M. Mao X. Lucas M.M. Barger S.W. Mol. Brain Res. 1999; 67: 303-315Google Scholar). The decoy protocol has been described previously (23Mao X. Barger S.W. Neuroreport. 1998; 9: 759-763Google Scholar). Briefly, 500 μl of minimal essential medium containing 3 μm decoy or scrambled oligonucleotide was dispensed into 500 μl of minimal essential medium containing 70 μl of LipofectAMINE suspension dropwise with constant mixing. The DNA/LipofectAMINE mixture was mixed vigorously for 15 s and then incubated for 30 min at room temperature before use. Fifty microliters of the DNA/LipofectAMINE mixture was added dropwise to each well of 24-well plates containing 450 μl of Neurobasal/B27 medium per well. The final concentration for decoy and scrambled oligonucleotides was 150 nm. Only marginal toxicity could be detected after long-term (≥72 h) decoy administration. After 24 h of decoy treatment, total RNA was harvested by Trireagent according to the manufacturer's protocol. The RNA concentration was quantified by absorbance at 260 nm. RNA was immediately used for RT-PCR or stored at −80 °C. The RT-PCR procedure was performed according to the manufacturer's protocols and has been described previously (24Li Y. Liu L. Kang J. Sheng J.G. Barger S.W. Mrak R.E. Griffin W.S.T. J. Neurosci. 2000; 20: 149-155Google Scholar). For NR1, the forward primer was 5′-CCTCCTGAGCGCAAGTACTCT-3′, and the reverse primer was 5′-GCCGTGGATACTTGGAGTGA-3′, creating a 333-bp amplimer. For β-actin, the forward primer was 5′-GTCCTCTGCCATGTGGTTTTC-3′, and the reverse primer was 5′-GCTGCGCTCTCGTAATTGTG-3′, creating a 439-bp amplimer. The PCR was performed at 94 °C for 30 s, 61 °C for 45 s, and 72 °C for 30 s for denaturation, annealing, and extension, respectively (27 and 24 cycles for NR1 and β-actin, respectively). SL2 cells were transfected with plasmids prepared from QIAGEN maxiprep kits by calcium phosphate coprecipitation. Cells were plated in 24-well plates at 106/ml with 0.5 ml of medium/well the day before transfection. Each well received 50 μl of a DNA/calcium phosphate suspension that contained 0.2 μg of reporter plasmid (pNRL series or pβcE1-GL3), 40 ng of pPAC series plasmids, 0.2 μg of pRL-promΔ, and an appropriate amount of inert carrier DNA to equalize to 2.0 μg of DNA. After 48 h, cells were harvested. The dual-luciferase reporter assay system was used to determine luciferase activities following the manufacturer's instructions in a Turner TD-20e luminometer. Transient transfection of primary neocortical neurons followed a calcium phosphate protocol optimized for these cells (25Xia Z. Dudek H. Miranti C.K. Greenberg M.E. J. Neurosci. 1996; 16: 5425-5436Google Scholar). Extensive tests of plasmid/calcium phosphate concentrations and cell density were conduced to further optimize this protocol for our culture conditions. Ultimately, neurons were plated in 24-well plates at 2 × 105/cm2 and treated with cytosine arabinoside in Neurobasal/B27 medium to restrict glial cell numbers. After 8 daysin vitro, neurons were transfected with DNA/calcium phosphate suspension (at 40 μl/well) containing 0.2 μg of pNRL series plasmid, 0.2 μg of pRL-CMV, and 1.6 μg of carrier DNA. After a 3-h transfection, the medium containing DNA/calcium phosphate was removed, and the neurons were washed three times with HEPES-buffered Hanks' balanced salt solution and then maintained in fresh Neurobasal/B27 medium overnight. After this recovery phase, neurons were treated with glutamate (50 μm) for 60 min, followed by two washes with HEPES-buffered Hanks' balanced salt solution; neurons in control conditions experienced the same treatment lacking glutamate. Twenty hours after glutamate exposure, neurons were harvested, and the luciferase activities was determined. Glutamate did not evoke any significant activity change for Renillaluciferase. All experiments were repeated three times. In our investigations of the influence of glutamate on the activity of neuronal κB-binding factors, we found that the prominent factor binding a κB sequence in neurons is not composed ofbona fide NF-κB proteins (15Moerman A.M. Mao X. Lucas M.M. Barger S.W. Mol. Brain Res. 1999; 67: 303-315Google Scholar). This factor, which we termed NKBF, was found to have target sequence specificities that are subtly distinct from those of NF-κB, and its DNA-binding activity is diminished by toxic glutamate treatments in primary neuronal cultures (14Mao X. Moerman A.M. Lucas M.M. Barger S.W. J. Neurochem. 1999; 73: 1851-1858Google Scholar). Sp1 and related factors can compete with NF-κB for binding to a DNA sequence (16Hirano F. Tanaka H. Hirano Y. Hiramoto M. Handa H. Makino I. Scheidereit C. Mol. Cell. Biol. 1998; 18: 1266-1274Google Scholar). Combined with the target DNA specificity of NKBF, this finding suggested that NKBF might include Sp1-related proteins. To test this possibility, EMSA analyses were performed utilizing antibodies directed against Sp1-related factors. Four such factors have been cloned: Sp1 and Sp3 are ubiquitously expressed in most cell and tissue types; Sp2 and Sp4 have more restricted expression patterns (26Suske G. Gene (Amst.). 1999; 238: 291-300Google Scholar), but high levels of Sp4 expression have been reported in the developing central nervous system (27Supp D.M. Witte D.P. Branford W.W. Smith E.P. Potter S.S. Dev. Biol. 1996; 176: 284-299Google Scholar). Therefore, Sp1, Sp3, and Sp4 were investigated in extracts from primary cultures of neocortical neurons probed with an oligonucleotide probe containing an NF-κB-binding site. Antibodies against Sp1 family factors either supershifted or inactivated NKBF (Fig.1 A), and the combination of multiple antibodies was more efficient than an equivalent amount of a single antibody. For example, the combination of 0.9 μg of anti-Sp1 antibody and 0.9 μg of anti-Sp4 antibody (Fig. 1 A,lane 9) altered NKBF more effectively than did 1.8 μg of anti-Sp1 antibody alone (lane 3) or 1.8 μg of anti-Sp4 antibody alone (lane 7). The combination of antibodies against Sp1, Sp3, and Sp4 almost completely abrogated the detection of NKBF (Fig. 1 A, lane 11). These data indicate that NKBF consists of Sp1 family factors. We previously observed an abundant, fast-migrating κB-binding activity (Complex III) in several cell and tissue populations (15Moerman A.M. Mao X. Lucas M.M. Barger S.W. Mol. Brain Res. 1999; 67: 303-315Google Scholar); this complex is insensitive to anti-Sp1 family antibodies. An Sp1-binding activity with the same mobility as NKBF was observed when an Sp1 consensus probe was used in EMSA (Fig. 1 B). With this probe, Complex III was absent, but a separate fast-migrating band appeared (Fig. 1 B,arrowhead). Antibodies against Sp1 family factors affected the NKBF-comigrating complex similarly to their influence on reactions with the NF-κB probe. The fast-migrating band observed with this probe was sensitive to anti-Sp3 antibody. No interactions between antibodies and the probe were observed when nuclear proteins were omitted in the reactions (Fig. 1 B, lanes 12–14). An unlabeled, double-stranded oligodeoxynucleotide containing an Sp1-binding site was able to compete NKBF from a labeled NF-κB probe (data not shown). On the basis of its target sequence specificity and its mobility, we speculated that Complex III might include RBP-Jκ. The core sequence for RBP-Jκ binding is 5′-CCTGGGAAA-3′, which matches the sequence of the 3′-end of the Ig/HIV-κB probe, extending beyond the 3′-end of the κB consensus sequence. Furthermore, mutations in this portion of the probe disrupted Complex III binding (data not shown). Initial assays indicated that a monoclonal antibody (K0043) against RBP-Jκ supershifted Complex III with high efficiency. The supershifted RBP-Jκ complex had a mobility similar to that of native NKBF, potentially confounding interpretation. Therefore, we also performed this test with a sequence that allows Complex III formation, but is deficient in NKBF binding (IL6-κB) (Fig. 1 C). In EMSA of nuclear extracts from N9 microglial cells, anti-RBP-Jκ antibody caused a supershift of Complex III interacting with the IL6-κB probe. Furthermore, this analysis demonstrated that anti-RBP-Jκ antibody did not interfere with the DNA-binding activity of bona fideNF-κB present in the N9 extracts. A similar supershift pattern for Complex III was observed when neuronal nuclear extracts were used (data not shown). Taken together, these data indicate that the neuronal factors binding most prominently to the Ig/HIV-κB sequence are RBP-Jκ and a complex that includes members of the Sp1 family. However, RBP-Jκ binding depends on the bases at the 3′-end of this sequence rather than the core κB consensus sequence. To determine whether Sp1 or related factors can act as transactivating factors at κB cis-elements, we cotransfected an Sp1, Sp3, or Sp4 expression vector with a luciferase reporter plasmid containing four copies of a κB element. These experiments were performed in the Drosophila SL2 cell line, which lacks detectable Sp1-related activity. The reporter construct utilized the κB element from the promoter of the human β-amyloid precursor protein, as this sequence lacks the element responsible for RBP-Jκ binding (15Moerman A.M. Mao X. Lucas M.M. Barger S.W. Mol. Brain Res. 1999; 67: 303-315Google Scholar). Separate cultures were transfected with the above expression vectors alone, and nuclear extracts were prepared to confirm active protein production by EMSA analysis (Fig.2 A). When Sp1 was cotransfected with the κB-dependent reporter plasmid, luciferase activity was elevated over 2-fold (Fig. 2 B). Expression of Sp3 alone did not affect transcription from this particular κB element, and Sp3 did not appreciably affect the induction by Sp1. However, Sp4 expression significantly suppressed basal activity, and the combination of Sp3 and Sp4 significantly suppressed the induction by Sp1. We previously found that glutamate evokes reduction of NKBF DNA-binding activity in a time- and dose-dependent manner (14Mao X. Moerman A.M. Lucas M.M. Barger S.W. J. Neurochem. 1999; 73: 1851-1858Google Scholar). This effect is correlated with glutamate neurotoxicity. To further characterize the relationship of Sp1 to NKBF, we tested the influence of glutamate on the DNA-binding activity of Sp1 and related factors in extracts from neocortical neurons. Glutamate influenced the gel retardation of an Sp1 consensus target probe in a time- and dose-dependent manner (Fig.3, A and B), reminiscent of its effects on NKBF binding to the Ig/HIV-κB probe. Substantial suppression of Sp1-related binding activity was observed only after long-term or high-dose glutamate exposure. This effect appeared to be specific to Sp1-related factors, as glutamate did not have a similar influence on the activities of other transcription factors tested. First, using a similar time course paradigm, we found that the DNA binding of transcription factor AP-1 was inhibited by short-term glutamate exposure and enhanced at longer times of glutamate treatment (Fig. 4 A), behavior that was essentially opposite that of the Sp1-related factors. Second, no significant influence (p = 0.9175) of glutamate on RBP-Jκ was observed (Fig. 4 B). On the other hand, glutamate treatment significantly inhibited the DNA-binding activity of Sp1-related factors (p = 0.0002, glutamate-treatedversus untreated). Surprisingly, the activity of Sp factors was fully restored after prolonged glutamate exposure (Fig.5), whereas glutamate inhibited the activity of Sp1-related factors for relatively short-term treatments (<18 h).Figure 4Glutamate specifically inhibits the activity of Sp factors. Nuclear extracts were prepared from neocortical neurons, and the DNA-binding activity was analyzed by EMSA.A, primary neocortical neurons were treated with 50 μm glutamate for the indicated times, and nuclear proteins (3 μg) were probed with an AP-1 enhancer sequence. For the time points between 0 and 60 min, a chase period followed the glutamate exposure such that the total elapsed time in all conditions was 60 min. *, percentage of control. B, neocortical neurons were treated with 50–300 μm glutamate for 60 min, and the DNA-binding activities of Sp1-related factors and RBP-Jκ were assayed with Sp1 and IL6-κB probes, respectively.View Large Image Figure ViewerDownload (PPT)Figure 5The DNA-binding activity of Sp1-related factors is fully restored after overnight glutamate treatments.Primary neocortical neurons were exposed to 50 μmglutamate for the indicated times. Nuclear proteins were harvested and probed for Sp1-binding activities in EMSA. Arrowheadsindicate fast-migrating complexes. F.P., free probe.View Large Image Figure ViewerDownload (PPT) Interestingly, the diminution of the activity of Sp1-related factors was accompanied by the appearance of a fast-migrating species binding the Sp1 site probe (Fig. 5, lanes 2 and 3,arrowheads). The possibility that this band arose from a fragment of an Sp1-related factor could not be confirmed by supershift, perhaps due to changes in the antigenic determinant. However, degradation of Sp1 family factors was suggested by Western blot analysis. When proteins from neocortical cultures were analyzed with an antibody against Sp1, a 78-kDa immunoreactive band was detected in neuronal cultures. The levels of this species were reduced by short-term glutamate exposure and recovered after long-term treatments (Fig. 6 B). Concomitant with the diminution of the 78-kDa band by glutamate, a smaller species (∼68 kDa) was initially enhanced. Its levels were also diminished by 12 h, perhaps due to loss of the 78-kDa precursor. The levels of Sp3 and Sp4 experienced a similar transient decline, although the precise time course was different for these proteins (Fig.6 B). Specifically, the lowest level of Sp3 was at 1 h, whereas for Sp4, it was at 6 h. Parallel cultures were subjected to EMSA (Fig. 6 A), confirming a correlation between Sp1-related DNA-binding activity and the described changes in Sp1 polypeptide levels. Sp1-binding sites appear to be important for expression of the NR1 glutamate receptor subunit (17Bai G. Kusiak J.W. J. Biol. Chem. 1995; 270: 7737-7744Google Scholar). Together with our demonstration of the attenuation by glutamate of Sp1-related DNA-binding activity, this relationship suggested a potential feedback inhibition of NR1 expression. Therefore, we assayed NR1 polypeptide levels by Western blot analysis in the same samples analyzed for Sp1-related factors. Application of glutamate to neocortical cultures resulted in a reduction of detectable NR1 protein, reaching ∼50% of its control level after a 3-h treatment (Fig. 6 B). Thus, reductions in NR1 expression correlated with the reduction of Sp1-related proteins by glutamate. As a more direct test of the role of Sp1-related proteins in control of NR1 expression, we sought to inhibit Sp1-related DNA-binding activity by loading the cells with
Referência(s)