Regulator of Calcineurin 1 (RCAN1) Facilitates Neuronal Apoptosis through Caspase-3 Activation*

Individuals with Down syndrome (DS) will inevitably develop Alzheimer disease (AD) neuropathology sometime after middle age, which may be attributable to genes triplicated in individuals with DS. The characteristics of AD neuropathology include neuritic plaques, neurofibrillary tangles, and neuronal loss in various brain regions. The mechanism underlying neurodegeneration in AD and DS remains elusive. Regulator of calcineurin 1 (RCAN1) has been implicated in the pathogenesis of DS. Our data show that RCAN1 expression is elevated in the cortex of DS and AD patients. RCAN1 expression can be activated by the stress hormone dexamethasone. A functional glucocorticoid response element was identified in the RCAN1 isoform 1 (RCAN1-1) promoter region, which is able to mediate the up-regulation of RCAN1 expression. Here we show that overexpression of RCAN1-1 in primary neurons activates caspase-9 and caspase-3 and subsequently induces neuronal apoptosis. Furthermore, we found that the neurotoxicity of RCAN1-1 is inhibited by knock-out of caspase-3 in caspase-3−/− neurons. Our study provides a novel mechanism by which RCAN1 functions as a mediator of stress- and Aβ-induced neuronal death, and overexpression of RCAN1 due to an extra copy of the RCAN1 gene on chromosome 21 contributes to AD pathogenesis in DS.


Individuals with Down syndrome (DS) will inevitably develop
Alzheimer disease (AD) neuropathology sometime after middle age, which may be attributable to genes triplicated in individuals with DS. The characteristics of AD neuropathology include neuritic plaques, neurofibrillary tangles, and neuronal loss in various brain regions. The mechanism underlying neurodegeneration in AD and DS remains elusive. Regulator of calcineurin 1 (RCAN1) has been implicated in the pathogenesis of DS. Our data show that RCAN1 expression is elevated in the cortex of DS and AD patients. RCAN1 expression can be activated by the stress hormone dexamethasone. A functional glucocorticoid response element was identified in the RCAN1 isoform 1 (RCAN1-1) promoter region, which is able to mediate the upregulation of RCAN1 expression. Here we show that overexpression of RCAN1-1 in primary neurons activates caspase-9 and caspase-3 and subsequently induces neuronal apoptosis. Furthermore, we found that the neurotoxicity of RCAN1-1 is inhibited by knock-out of caspase-3 in caspase-3 ؊/؊ neurons. Our study provides a novel mechanism by which RCAN1 functions as a mediator of stress-and A␤-induced neuronal death, and overexpression of RCAN1 due to an extra copy of the RCAN1 gene on chromosome 21 contributes to AD pathogenesis in DS.
Down syndrome (DS) 4 is the most common genetic cause of mental retardation, affecting approximately one in 800 -1,000 babies (1). Individuals with DS inevitably develop characteristic Alzheimer disease (AD) neuropathology, which includes neuritic plaques, neurofibrillary tangles, and neuronal cell death following middle age (2)(3)(4)(5). DS is a valuable model system for understanding AD pathogenesis. Both DS and AD feature prominent neuronal losses in the hippocampus and cortex, resulting in progressive cognitive deficits in AD patients (5). Apoptosis has been implicated in playing a major role in neuronal cell death in both AD and DS (6 -9). Increased immunoreactivity of activated caspase-3 is observed in neurons of AD and DS (7); however, the underlying mechanism leading to neuronal apoptosis in AD and DS remains elusive.
The development of DS is caused by the presence of an extra copy of human chromosome 21 (10,11). The DSCR1 (Down syndrome critical region 1) gene was identified and located on chromosome 21 (12,13). Dysregulation of a regulatory circuit involving DSCR1-calcineurin-nuclear factor of activated T cells (NFAT) plays an important role in DS development (14). DSCR1 proteins physically interact with calcineurin subunit A and inhibit calcineurin activity in vitro and in vivo (15)(16)(17)(18)(19). DSCR1 was accordingly renamed as RCAN1 (regulator of calcineurin 1) (20). RCAN1 is phosphorylated at Ser 112 by BMK1 (big MAP kinase 1), which is the priming site for the subsequent phosphorylation at Ser 108 by GSK-3 (21)(22)(23)(24). The phosphorylation form of RCAN1 can release the inhibition effect on calcineurin. Furthermore, phosphorylated RCAN1 is a substrate for calcineurin (79). RCAN1 FLISPP motif phosphorylation could increase its ability to inhibit the calcineurin activity (25) and accelerate its degradation rate (26). However, it was reported that GSK-3 kinase can enhance the calcineurin activity through the RCAN1 phosphorylation in the yeast system (23). RCAN1 could also be phosphorylated by NF-B-inducing kinase (NIK), which increases the RCAN1 stability (27). Furthermore, RCAN1 interacts with TAB2 (TAK1 (TGF-␤-activated kinase 1)-binding protein 2) to be phosphorylated by TAK1 (28). The phosphorylated form of the RCAN1 could serve as a calcineurin facilitator. NFAT is a major substrate for calcineurin, and dephosphorylation of NFAT facilitates NFAT nuclear translocation and activation of its target genes' transcription. RCAN1 can repress the NFAT signaling pathway by inhibition of calcineurin (14). Also, as a downstream gene of the NFAT signaling pathway, RCAN1 can be activated by a subset of molecules, including VEGF, angiotensin II, G protein-coupled receptor 54, TNF-␣, thrombin, and other activators of the calcineurin-NFAT pathway, such as calcium ionophore (29,30). RCAN1 can also be activated by dephosphorylation in neural cells via calcium current increase through the L-type calcium channel (31).
RCAN1 has been shown to be involved in cardiac valve development, cardiac hypertrophy, inflammation, angiogenesis, and cancer. RCAN1 has also been implicated in learning and memory. However, the role of RCAN1 in neurodegeneration in AD and DS is unknown. In this study, we show that RCAN1 is overexpressed in cortical tissues from AD and DS patients. To investigate the mechanism of RCAN1 overexpression in AD and DS, we characterized human RCAN1 gene promoters and identified a functional glucocorticoid response element (GRE), through which RCAN1 expression is up-regulated by stress hormone dexamethasone. Here we show that RCAN1 mediates glucocorticoid-induced neuronal apoptosis. We found that overexpression of RCAN1-1 in primary neurons activates the caspase-9 and caspase-3 apoptotic pathway, thereby rendering neurons more vulnerable to apoptosis induced by dexamethasone and A␤. Our data suggest that RCAN1 overexpression may contribute to AD pathogenesis by mediating neuronal death in the brains of DS and AD patients.

EXPERIMENTAL PROCEDURES
Cloning of the RCAN1 Gene Promoter and Construction of Chimeric Luciferase Reporter Plasmids-A forward primer corresponding to Ϫ684 bp (5Ј-ccgctcgaggtcctcttatttttccgctatttc-3Ј) of the transcriptional start site and a reverse primer corresponding to the coding sequence (5Ј-cacaagcttgtcctgcaggtccacctc-3Ј) were used to PCR-amplify the 5Ј-UTR region of the RCAN1 gene exon 1 from the genome of human neuroblastoma cells SH-SY5Y. The DNA fragment was cloned into pGL3-Basic upstream of the luciferase reporter gene to construct pRCANluc-G. The fragment from Ϫ684 to ϩ46 bp was amplified and cloned using primers 5Ј-ccgctcgaggtcctcttatttttccgctatttc and 5Ј-cacaagctttgtcagcagtctcccagc, corresponding to ϩ46 bp of the transcriptional start site on RCAN1 gene exon 1. Primers 5Ј-gctagctagcaatatattgtgaacc and 5Ј-cacctcgagtgagcagacatgtctc were used to amplify the region between Ϫ1650 and Ϫ685 bp in the 5Ј-UTR of RCAN1 gene exon 1. The amplified fragment was inserted in front of the Ϫ684 to 46 bp fragment in pRCANluc-B to obtain the pRCANluc-A. Primers were designed to include restriction enzyme sites so that the resulting PCR-amplified fragment could be easily cloned into the multicloning sites of vector pGL3-Basic (Promega). Deletion plasmids pRCANluc-C, -D, and -E were created from pRCANluc-A by utilization of restriction enzyme sites ApaI, PvuII, and EcoRI, respectively.
Preparation of ␤-Amyloid Fibrils-Synthesized A␤(1-40) was dissolved in sterile double-distilled H 2 O to 1 mM. The dissolved A␤ was incubated at 37°C for 1 h and diluted again with the same volume of PBS (1:1 dilution). A␤ was aged at 37°C for 4 days, allowing the stable growth of A␤ fibrils.
Transfection and Luciferase Assay-Cells were grown to ϳ70% confluence and transfected with 2 g of plasmid DNA/ 35-mm plate using Lipofectamine 2000 (Invitrogen) as described (34). The pCH110 ␤-galactosidase expression plasmid was co-transfected to normalize for transfection efficiency. Cells were harvested 48 h after transfection and lysed in 200 l of 1ϫ Reporter Lysis Buffer (Promega). The luciferase and ␤-galactosidase assays were performed according to the manufacturer's protocols (Promega) as described (35).
Primer Extension Assay-A primer extension assay was performed to determine the transcription initiation site. Total neuronal RNA was extracted from SH-SY5Y cells using TRI-Reagent (Sigma). Yeast tRNA was used as a control. A reverse primer, corresponding to positions ϩ46 to ϩ28 bp, 5Ј-cacaagctttgtcagcagtctcccagc, was synthesized and radioactively end-labeled with [␥-32 P]ATP (Amersham Biosciences). Primer extension and sequencing were performed as described (34).
Gel Shift Assay-Oligonucleotide RCAN1Ϫ272Ϫ237bp (5Јcagctgtcagaaaagcggaactggggacgaggactt) and its antisense strand (5Ј-aagtcctcgtccccagttccgcttttctgacagctg) were synthesized to generate a double-stranded RCAN1-GRE oligonucleotide probe. The GRE consensus oligonucleotide was generated by annealing 5Ј-agaggatcTGTACAggaTGTTCTagat with 5Ј-atctAGAACAtccTGTACAgatcctct (core sequence is italic and uppercase). The RCAN1-GRE probe was end-labeled with [␥-32 P]ATP by T4 polynucleotide kinase. Gel shift assays were performed as described (36). The samples were analyzed by a 4% nondenaturing PAGE gel.
Virus Infection-The RCAN1 expression plasmid pcDNA3.1-RCAN1mycHis was generated as described (37). RCAN1-1 cDNA was cloned into a Semliki Forest virus vector pSFV to generate pSFV-RCAN1, and a green fluorescent protein cDNA was cloned into pSFV to generate pSFV-GFP. To generate the virus, pSFV1-EGFP, pSFV1-RCAN1myc, pSFV-RCAN1GFP, and helper virus pHelper-2 DNA were linearized by SpeI or SapI. Viral particles were generated according to the manufacturer's instructions (Invitrogen). For infection, the viral particles activated with chymotrypsin (Sigma) were added to neuronal cultures at a 1:10 dilution with culture medium. Cells were infected for 1 h at 37°C, followed by replacement of the infection media with conditioned culture medium and incubated for 12-18 h prior to use. Then the infected neurons were monitored for expression of GFP or immunostained for detection of Myc tag by 9E10 antibody to visualize RCAN1 expression in SFV-RCAN1myc-infected neurons.
Antibodies and Immunoblotting-The rabbit anti-RCAN1 polyclonal antibody was raised against the C terminus of RCAN1 protein (RPKPKIIQTRRPEYTPIHLS). The antibody was characterized as being able to specifically detect RCAN1. Brain tissues from AD patients were obtained from the Department of Pathology, Columbia University (New York). Brain tissues of DS abortuses (17-21 gestational weeks) were obtained from University of Maryland Brain and Tissue Bank for Developmental Disorders. The brain tissues or cells were lysed in radioimmune precipitation assay lysis buffer (1% Triton X-100, 1% sodium deoxycholate and 4% SDS, 0.15 M NaCl, 0.05 M Tris-HCl, pH 7.2) supplemented with protease inhibitors (Complete, Roche Applied Science). The lysates were resolved by 12% SDS-PAGE for detecting RCAN1 and 16% Tris-Tricine PAGE for caspase-3 and caspase-9. The immunoblotting was performed as described (33). Rabbit anti-RCAN1 polyclonal antibody (1:1000 dilution) was used to detect RCAN1 expression. Caspase-3 and its cleaved form P20 were detected with a rabbit polyclonal antibody against amino acids 29 -43 of caspase-3 (Sigma, catalog no. C9598, 1:1000). Caspase-3 cleavage was further confirmed with an anti-cleaved caspase-3 antibody (catalog no. 9661) from Cell Signaling. Internal control ␤-actin was analyzed using monoclonal anti-␤-actin antibody AC-15 (Sigma, 1:1000). Caspase-9 was detected with anti-caspase-9 from Stressgen (catalog no. AAP-149).
RT-PCR-RNA was isolated from cells by TRI-Reagent (Sigma). PowerScript TM reverse transcriptase (Invitrogen) was used to synthesize the first strand cDNA from an equal amount of the RNA sample following the manufacturer's instructions. The newly synthesized cDNA templates were further amplified by Platinum TaqDNA polymerase (Invitrogen) in a 25-l reaction. 20 -35 cycles of PCR were used to cover the linear range of the PCR amplification. Two pairs of primers were used to specifically amplify RCAN1 isoform 1 (RCAN1-1) and isoform 4 (RCAN1-4), respectively. Primers for isoform 1 are 5Ј-ggaattcgccaccatggaggaggtggacctg and 5Ј-cgcgtcgactggctgaggtggatcggcgtg. Isoform 4 was amplified by using 5Ј-gggtctgtagcgctttcac and 5Ј-cgcgtcgactggctgaggtggatcggcgtg. A pair of gene-specific primers, 5Ј-tctggatcctcaccaccatggagaaggc and 5Ј-atactcgaggcagggatgatgttctg, was used to amplify a 324-bp fragment of GAPDH as an internal control. The samples were further analyzed on a 1.5% agarose gel. Eastman Kodak Co. Image Station 1000 software (PerkinElmer Life Sciences) was used to analyze the data.
MTT Assay-The MTT (Sigma) was added into culture medium of primary neurons and incubated for 4 h in a 37°C incubator containing 5% CO 2 . The medium was then removed by a syringe, and 2-propanol-HCl was added to dissolve the dye. Lysates were analyzed with a spectrophotometer at a wavelength of 540 nm (Multiskan Ascent, ThermoLab System).
Apoptotic Cell Staining-Rat primary neurons were fixed, permeabilized, and stained with Hoechst 33258 (1 g/ml, Sigma) at room temperature for 15 min. Cells were rinsed twice with PBS and sealed in mount medium (Vector Laboratories). Apoptotic cells were analyzed by a fluorescent microscope (Axiovert 200, Carl Zeiss Inc.). TUNEL staining was performed according to the manufacturer's instructions (Promega and Roche Applied Science). Results were analyzed by fluorescence microscopy.
Mitochondrial Fractionation-90% confluent HEK293 cells were harvested and washed three times with cold 1ϫ PBS. The cells were then suspended in mitochondrial isolation buffer (250 mM sucrose, 10 mM Hepes, 10 mM KCl, 1.5 mM MgCl 2 , 1 mM EDTA) supplemented with Complete Protease Inhibitors (Roche Applied Science)) and incubated on ice for 30 min. Cells were lysed with a glass homogenizer (15-25 strokes) on ice. The homogenate was centrifuged twice at 1000 ϫ g for 5 min in order to get rid of the nuclei and remaining intact cells. The supernatant was removed and centrifuged at 140,000 ϫ g for 15 min. The resultant pellet containing mitochondria was washed with mitochondrial isolation buffer. The supernatant was transferred to a new tube and centrifuged for an additional time in order to remove any remaining mitochondria.

RCAN1 Expression Is Elevated in the Brains of DS Fetus and AD Patients-Previous
studies have shown that RCAN1 mRNA is elevated in the brains of DS and AD patients (17,38), indicating that RCAN1 gene expression is up-regulated at the transcriptional level. To further examine whether RCAN1 is also elevated at the protein level, an antibody specific to RCAN1, DCT3, was used to detect RCAN1 proteins in the brain tissue lysates from AD patients and DS abortuses as well as their agematched controls. A synthetic peptide DCT with sequence PRKPKIIQTRRPEYTPIHLS corresponding to the C terminus of the human RCAN1 protein was used to immunize a rabbit, and a polyclonal antibody DCT3 was raised against the human RCAN1 protein C terminus. To characterize this RCAN1 antibody, HEK293 cells were transfected with empty vector or pRCAN1-mycHis plasmid. Fig. 1A shows that Myc-tagged RCAN1 proteins were detected by both mouse monoclonal 9E10 anti-Myc antibody and rabbit polyclonal antibody DCT3. Overexpressed RCAN1 protein could not be detected with the preimmunization serum pre-DCT. Furthermore, preincubation of DCT3 with excess DCT peptides resulted in clearance of RCAN1-specific antibody, and the precleared antibody could not detect RCAN1 protein (Fig. 1A). These data clearly demonstrate that our DCT3 antibody is able to specifically detect RCAN1 protein.
To examine the localization of RCAN1 protein expression in human brain, human hippocampal and cortical slices were immunostained with DCT3 antibody specifically targeting RCAN1. Immunohistochemistry studies show that DCT3 antibody specifically stains the granule cell layer in the hippocampus and the pyramidal neuronal layers in the cortex (Fig. 1B). RCAN1 expression is detected in both the nucleus and cytosol of neurons. Higher magnification reveals that RCAN1 is preferentially localized in nucleus of pyramidal neurons (Fig. 1B).
These results show that RCAN1 is highly and specifically expressed in neurons in the hippocampus and cortex. Because neurons of the hippocampus and cortex are specifically affected in Alzheimer disease, the high and specific expression of RCAN1 in neurons of the hippocampus and cortex suggests that RCAN1 may be involved in AD pathogenesis.
RCAN1 proteins were examined in the brain tissue lysates from AD patients and DS abortuses and patients as well as their age-matched controls. Western blot analysis using DCT3 antibody showed that levels of RCAN1 protein are significantly elevated in cortical tissue from DS abortuses (156.7 Ϯ 17.04% relative to control, p Ͻ 0.05) (Fig. 1, C and D). RCAN1 protein levels are also significantly increased in the cortex of AD patients (173.4 Ϯ 18.84% relative to control, p Ͻ 0.05) (Fig. 1, E and F). Our data show that RCAN1 is overexpressed in AD and DS brains. The result is consistent with previous reports (39,40). Because nearly all DS patients will develop AD after their 30s, up-regulation of RCAN1 in both DS and AD suggests that RCAN1 may be involved in AD pathogenesis. Increased RCAN1 levels in DS may be attributed to its triplication on chromosome 21. However, the mechanism of RCAN1 up-regulation in AD is unknown.
Cloning and Functional Analysis of the Human RCAN1 Gene Promoter-To investigate the molecular mechanism by which RCAN1 gene expression is up-regulated in AD pathogenesis, we cloned and functionally analyzed a 2000-bp fragment of the 5Ј-flanking region of the first exon of the human RCAN1 gene.
The sequence was deposited to GenBank TM under accession number EF577083 ( Fig. 2A). To determine the transcription start site of the human RCAN1 gene, a primer extension assay was performed. A 32 P-labeled antisense primer (5Ј-tgtcagcagtctcccagc) located 302 bp upstream of the translational start site AUG was used to hybridize with neuronal RNA. The primer extension assay yielded a 45-bp major cDNA product. DNA sequencing gel analysis indicates that the major transcription start site is located at 348 bp upstream from the translation start site. Transcription starts with guanine, and this site was designated as ϩ1 (Fig. 2B). A computer-based transcription factor binding site search revealed that this 2.0-kb 5Ј-flanking region contains several putative regulatory elements, such as AP1, AP2, GATA, OCT1, and USF ( Fig. 2A).
To determine if the 5Ј-flanking fragment obtained from genomic DNA contains the promoter of the RCAN1 gene, we cloned the fragment into a promoterless plasmid vector, pGL3-Basic (Fig. 3A). The pGL3-Basic vector lacks eukaryotic promoter and enhancer sequences upstream of a reporter luciferase gene. Luciferase activity in cells transfected with this plasmid depends on insertion and proper orientation of a functional promoter upstream from the luciferase gene. Luciferase activity is indicative of promoter activity. pRCANluc-A plasmid was constructed to contain a 1.70-kb 5Ј-flanking region from Ϫ1651 to ϩ45 bp of the RCAN1 gene upstream of the luciferase reporter gene. Plasmid DNA was transfected into HEK293 cells and SH-SY5Y cells, and luciferase activity was measured by a FIGURE 1. RCAN1 protein is elevated in brains from AD and DS patients. A, characterization of RCAN1 antibody DCT3. Anti-RCAN1 antibody DCT3 targeting the last 20 amino acids of the human RCAN1 C terminus was raised in a rabbit. Cell lysates from HEK293 cells transfected with plasmid constructs pRCAN1-mycHis or its empty vector pcDNA3.1mycHis(c) were separated on a 12% glycine SDS-polyacrylamide gel. RCAN1 was immunoblotted with DCT3 at a dilution of 1:1000 (lanes 3 and 4). 9E10 antibody targeting the Myc tag fused with RCAN1 was used as positive control (lanes 1 and 2). Lanes 5 and 6 show immunoblots with the preimmune serum drawn from the rabbit before immunization. Lanes 7 and 8 were blotted with DCT3 preincubated with the antigen DCT. B, RCAN1 is enriched in the neuronal layers in human hippocampus and cortex. a, immunohistochemical staining with DCT3 antibody showed that RCAN1 is localized in the granule cell layer in human hippocampus (100ϫ). Higher magnification shows its localization in both cytosol and nucleus in neurons. b, RCAN1 is enriched in pyramidal neuronal layers in cortex (100ϫ). Higher magnification (400ϫ) shows that the characteristic morphology of pyramidal neurons was stained by DCT3 antibody targeting the RCAN1 C terminus. C, representative gels showing Western blots of RCAN1 in tissues from the cortex of DS. Cortical tissues from gestational week 17-21 DS abortuses and their age-matched controls were lysed in 4% SDS radioimmune precipitation assay lysis buffer. 150 g of total protein were separated on a 12% Tris-glycine SDS-polyacrylamide gel. Rabbit polyclonal anti-RCAN1 antibody DCT3 (1:1000 dilution) was used to detect RCAN1. ␤-Actin was detected by a monoclonal anti-␤-actin antibody (AC15 from Sigma) and was used as loading control. D, quantification of Western blots showed that RCAN1 was increased in DS brains. Values represent mean Ϯ S.E. (error bars) (n ϭ 8). *, p ϭ 0.0372 by Student's t test. E, representative gels showing Western blots of RCAN1 in tissues from the cortex of AD patients. F, quantification of Western blots showed that RCAN1 was markedly increased in AD brains. Values represent mean Ϯ S.E. (n ϭ 10). *, p ϭ 0.0121 by Student's t test. Con, control.
luminometer to reflect promoter activity. Compared with an empty vector control, pRCANluc-A-transfected cells had significant luciferase activity at 554.2 Ϯ 44.48 relative luciferase units (RLU) in HEK293 and 461.6 Ϯ 16.82 RLU in SH-SY5Y cells (p Ͻ 0.001 relative to controls) (Fig. 3, C and D, lane A and Basic). However, plasmid pRCANlucR-I containing the RCAN1 promoter sequence from Ϫ1651 to ϩ45 bp in the reverse orientation (Fig. 3C, lane F) and plasmid pRCANluc-F containing the sequence from Ϫ1651 to Ϫ686 lacking the transcription initiation site (Fig. 3C, lane I) have little or no luciferase activity. These results indicate that the 1.70-kb fragment contains the functional promoter of the human RCAN1 gene.
To analyze the transcriptional activity of the RCAN1 promoter, a series of luciferase reporter gene plasmids containing different upstream deletions of the RCAN1 promoter were constructed and transfected into HEK293 and SH-SY5Y cells. The inserts of pRCANluc-B, -C, -D, and -E are the fragments of the RCAN1 promoter from Ϫ684, Ϫ311, Ϫ271, and Ϫ167 bp to ϩ45 bp, respectively (Fig. 3A). Analysis of deletion plasmids indicates that pRCANluc-A, pRCANluc-B, pRCANluc-C, and pRCANluc-D have strong promoter activity in HEK293 and SH-SY5Y cells. Deletion of 969 bp from Ϫ1651 to Ϫ685 resulted in significant reduction of RCAN1 promoter activity from 554.2 Ϯ 44.48 to 450.8 Ϯ 13.48 RLU in HEK293 cells (p Ͻ 0.05) (Fig. 3C, lanes A and B) and from 461.6 Ϯ 16.82 to 352.2 Ϯ 7.10 RLU in SH-SY5Y cells (p Ͻ 0.005) (Fig. 3D, lanes A and B). Deletion of an additional 413 bp from Ϫ685 had no significant effect on the promoter activity in HEK293 cells. pRCANluc-C and pRCANluc-D had similar promoter activity to pRCANluc-B in HEK293 cells, 449.1 Ϯ 33.45 and 442.5 Ϯ 40.73 RLU, respectively (p Ͼ 0.05) (Fig. 3C, lane C and D). However, deletion of 373 bp from Ϫ685 resulted in an increased promoter activity in SH-SY5Y cells from 352.2 Ϯ 7.10 to 457.4 Ϯ 8.75 RLU (p Ͻ 0.005) (Fig. 4D, lanes B and C). The promoter activ-FIGURE 2. Characterization of the human RCAN1 gene promoter. A, the nucleotide sequence of the human RCAN1 gene promoter. A 2000-bp fragment of the 5Ј-flanking region and the first exon of the human RCAN1 gene was isolated from the human genomic library and sequenced. The guanidine at ϩ1 represents the transcription start site. The positions of some of the unique and common restriction enzymes are indicated in italic type. The putative transcription factor binding sites are underlined in boldface type. The codon of the first exon is also indicated. B, a primer extension assay was performed to map the RCAN1 gene transcription start site. Neuronal RNA was extracted by TRI-Reagent, and yeast tRNA was used as a control. A 32 P-labeled primer complementary to the region of ϩ46 to ϩ29 bp was used for both primer extension and the sequencing reaction. The plasmid pRCANluc-D, containing a DNA fragment of Ϫ311 to ϩ46 bp, was used as a sequencing template. The samples were analyzed on 6% denaturing polyacrylamide gel electrophoresis. *, major transcription start site.
ities of pRCANLuc-C and pRCANLuc-A were not significantly different in SH-SY5Y cells (p Ͼ 0.05) (Fig. 3D). Further deletion of 40 bp from Ϫ312 reduced promoter activity to 353.0 Ϯ 12.49 RLU in SH-SY5Y cells (p Ͻ 0.005, Fig. 3D lane D versus lane C). These data suggested that the RCAN1 promoter has cell-specific transcriptional activity.
To identify the minimal promoter region required for RCAN1 gene expression, additional deletion plasmids were constructed. Further deletion of 95 bp from Ϫ272 bp in pRCANluc-D to Ϫ167 bp in pRCANluc-E significantly reduced luciferase activity from 442.5 Ϯ 40.73 to 170.7 Ϯ 2.68 RLU in HEK293 cells (p Ͻ 0.0005) and from 353.0 Ϯ 12.49 to 138.8 Ϯ 3.61 RLU in SH-SY5Y cells (p Ͻ 0.0001) (Fig. 3, C and D). These results indicate that a 317-bp fragment from Ϫ272 to ϩ45 bp contains the minimal human RCAN1 gene promoter.
The RCAN1 Promoter Contains a Glucocorticoid Response Element (GRE)-It was previously reported that RCAN1 mRNA was markedly increased by glucocorticoid induction in pre-B human leukemia cells (41). Neurons in hippocampus and cortex express high levels of glucocorticoid receptor (GR) (42). To investigate if glucocorticoids also activate the RCAN1 gene promoter, a GR expression plasmid was cotransfected with pRCANluc-A into HEK293 cells, and the cells were subjected to 100 nM dexamethasone treatment for 0, 12, 24, 48, and 72 h. A luciferase assay showed that the RCAN1 gene promoter activity was significantly up-regulated in a time-dependent manner by dexamethasone, with 143. 6 (Fig. 4A). Similar results were also observed in SH-SY5Y cells. Up-regulation of RCAN1 promoter activity by dexamethasone suggests that the RCAN1 promoter may contain a GRE.
To identify the GRE site in the RCAN1 promoter, we examined the effect of dexamethasone on the deletion constructs containing various RCAN1 promoter regions, including pRCANluc-A, -B, -C, -D, and -E. pGL3-Basic and pGL3-Promoter, which do not contain any GRE sites, were used as negative controls. Luciferase activity in the cells transfected with pGL3-Basic and pGL3-Promoter plasmids was not affected by dexamethasone treatment (p Ͼ 0.05), whereas luciferase activity in cells transfected with pRCANluc-A, -B, -C, and -D plasmids was markedly increased (Fig. 4B). However, further deletion of 95 bp from Ϫ272 bp in pRCANluc-D to Ϫ167 bp in pRCANluc-E abolished the upregulatory effect of dexamethasone on RCAN1 promoter activity, and dexamethasone had no effect on RCAN1 promoter activity of pRCANluc-E, 83.13 Ϯ 14.81% of control (p Ͼ 0.05) (Fig. 4B, lane E). Thus, the results suggest that a GRE is located in the region of Ϫ272 to Ϫ167 bp in the RCAN1 gene promoter.
To further confirm that the Ϫ272 to Ϫ167 bp region of the RCAN1 promoter contains a GRE, gel shift assays were performed. Double-stranded oligonucleotide probes corresponding to Ϫ272 to Ϫ237 bp, Ϫ251 to Ϫ217 bp, Ϫ231 to Ϫ197 bp, and Ϫ211 to Ϫ166 bp of the RCAN1 promoter were synthesized and end-labeled with 32 P radioactive isotope. A shifted DNA-protein complex band was detected after hybridizing the 32 P-labeled RCAN1Ϫ272Ϫ237bp probe with HeLa nuclear extract (Fig. 4C, lane 2) but not with other labeled oligonucleotide probes. These data suggest that the GRE site may be located in the Ϫ272 to Ϫ237 bp region of the RCAN1 promoter. To determine the binding specificity, GRE consensus oligonucleo-  4). *, p Ͻ 0.01 by Student's t test. C, gel shift assay. The gel shift assay was performed as described under "Experimental Procedures" using a 32 P-labeled double-stranded oligonucleotide probe, RCAN1Ϫ272Ϫ237bp. Lane 1, labeled probe alone without nuclear extract. Incubation of 32 P-labeled RCAN1Ϫ272Ϫ237bp with HeLa nuclear extracts retarded the migration rate of the labeled probe, which formed a new shifted DNA-protein complex band (lane 2). Competition assays were performed by further adding different concentrations of molar excess of unlabeled GRE competition consensus oligonucleotide (lanes 3-5). D, dexamethasone increased the RCAN1-1 mRNA level. RT-PCR was used to amplify RCAN1-1 from SH-SY5Y cells transfected with GR expression plasmid treated with or without 100 nM dexamethasone for 48 h. GAPDH was amplified as the internal control. E, quantification of RT-PCR shows that RCAN1-1 mRNA was elevated ϳ5-fold after dexamethasone treatment. Values represent mean Ϯ S.E. (n ϭ 3). *, p ϭ 0.0003 by Student's t test. F, dexamethasone increased RCAN1 protein expression. Western blot of RCAN1 in SH-SY5Y cells, which were transfected with GR expression plasmid and treated with or without 100 nM dexamethasone for 48 h. 150 g of cell lysates were separated in a 12% Tris-glycine polyacrylamide gel. RCAN1 was detected with anti-RCAN1 antibody DCT3. ␤-Actin was detected by anti-␤-actin and served as the internal control. G, quantification of E shows that RCAN1 protein was increased ϳ2-fold following dexamethasone treatment. The values represent mean Ϯ S.E. (n ϭ 3). *, p ϭ 0.0006 by Student's t test. H, differential activation of RCAN1 alternative promoters in SH-SY5Y cells by dexamethasone. The pRCANluc-A and pDE4Luc plasmids were cotransfected with GR expression plasmid and pCH11O into SH-SY5Y cells. pGL3-Basic was used as a negative control. Cells were treated with or without 100 nM dexamethasone for 48 h before the luciferase assay. The isoform 1 promoter plasmid pRCANluc-A has a higher activity in the neuronal cell line than the isoform 4 promoter plasmid pDE4Luc. *, p ϭ 0.0002 by Student's t test. Dexamethasone increases the promoter activity of isoform 1 promoter pRCANluc-A (*, p ϭ 0.0120 by Student's t test), whereas the drug has no effect on RCAN1 isoform 4 promoter (p Ͼ 0.05 by Student's t test). Values represent mean Ϯ S.E., n ϭ 3. Dex, dexamethasone. tides (5Ј-agaggatctgtacaggatgttctaga-3Ј) were added to compete for GR binding with the labeled RCAN1Ϫ272Ϫ237bp probe. The binding intensity of this shifted band was partially reduced after adding a 10-or 50-fold molar excess of unlabeled GRE consensus competition oligonucleotides, and the shifted band was abolished by the addition of a 100-fold excess of GRE consensus oligonucleotides (Fig. 4C, lanes 3-5). These results confirm that the shifted band represents the complex of RCAN1Ϫ272Ϫ237bp oligonucleotide bound with GR, and the human RCAN1 promoter contains a GRE in the region of Ϫ272 to Ϫ237 bp.
To investigate if this element plays an important role in transcription of the human RCAN1 gene, the effect of dexamethasone on endogenous RCAN1 gene expression was examined. Quantitative RT-PCR was performed to amplify RCAN1-1 mRNA by a pair of gene-specific primers (5Ј-gccaccatggaggaggtggacctg and 5Ј-tggctgaggtggatcggcgtg). Dexamethasone treatment increased the RCAN1-1 mRNA level more than 5-fold in GR-transfected SH-SY5Y cells (p Ͻ 0.0001) (Fig. 4, D and E). To further confirm the regulatory effect of GRs on RCAN1 gene expression, endogenous levels of RCAN1 protein were examined with anti-RCAN1 antibody DCT3. Consistent with our data showing that dexamethasone up-regulated RCAN1-1 expression at the transcription level, endogenous RCAN1-1 protein level was also markedly elevated by dexamethasone treatment (225.49 Ϯ 5.78% relative to control, p Ͻ 0.0001) (Fig. 4, F and G). Taken together, these results demonstrate that the RCAN1-1 gene promoter contains a functional GRE in the region of Ϫ272 to Ϫ237 bp, through which glucocorticoids exert an up-regulatory effect on RCAN1 gene expression.
RCAN1 has a tissue-specific expression pattern. The RCAN1 gene spans ϳ45 kb of genomic DNA and contains seven exons and six introns. Two major isoforms, RCAN1-1 and isoform 4 (RCAN1-4), are generated by alternative splicing of the first four exons. RCAN1-1 has a large form with 252 amino acids (RCAN1-1L) and a short form, RCAN1-1S. RCAN1-1S and RCAN1-4 both consist of 197 amino acids. The isoforms differ only at their N terminus, and the last 168 amino acids of the C terminus, encoded by exons 5-7, are the same in all forms of the isoforms. RCAN1-1 is highly expressed in the central nervous system (CNS), whereas RCAN1-4 is mostly expressed in heart muscle and fetal kidney (13,43). It was reported that the calcineurin-dependent isoform 4 is also highly expressed in areas of the brain in which calcineurin is highly expressed (44,45). We have cloned a 1.7-kb region upstream of the first exon of the human RCAN1 gene containing a functional promoter that controls the transcription of RCAN1-1 (Fig. 2). It has been reported that there is an alternative promoter upstream of exon 4 of the RCAN1 gene, which responds to the calcineurin-NFAT signaling pathway (30,46). Therefore, the RCAN1 gene contains two promoters and two translation initiation codons: one in the 5Ј-UTR of exon 1 and the other in the 5Ј-UTR of exon 4, which are responsible for transcriptional control of RCAN1-1 and RCAN1-4, respectively. To examine whether the effect of glucocorticoids is specific for the RCAN1 exon 1 promoter, a 1200-bp DNA fragment upstream of RCAN1 exon 4 was cloned into promoterless luciferase reporter plasmid pGL3-Basic to generate pDE4Luc. Dexamethasone markedly increased isoform 1 promoter activity (from 175.33 Ϯ 3.74 to 268.86 Ϯ 14.2 RLU, p Ͻ 0.001), whereas it had no effect on isoform 4 promoter activity (from 117.47 Ϯ 2.51 to 112.17 Ϯ 2.39 RLU, p Ͼ 0.05) (Fig. 4H). The data suggest that glucocorticoids specifically up-regulate RCAN1-1 expression but not RCAN1-4 expression.
Overexpression of RCAN1 Induces Neuronal Apoptosis-Our results thus far have shown that RCAN1 expression can be activated by dexamethasone, which has been shown to induce apoptosis in primary neurons. Previous studies have shown that overexpression of RCAN1 deters melanoma tumor growth and suppresses metastasis in mice (29,47,48). Although musclespecific overexpression of RCAN1 does not always result in embryonic lethality (49), there was a report that RCAN1 overexpression could lead to early embryonic fatality in transgenic mice (50), suggesting that RCAN1 may be proapoptotic. RCAN1 is preferentially localized in neurons of the hippocampus and cortex, where there is a marked neuronal loss in the brains of AD patients (38). To investigate whether RCAN1 overexpression could induce apoptosis or exacerbate neuronal death induced by apoptotic inducers, cDNA encoding for the RCAN1 isoform 1 was cloned into a Semliki Forest virus vector, pSFV, to generate pSFV-RCAN1. A green fluorescent protein vector, pSFV-GFP, was used as control. Rat E18 primary neurons were infected with SFV-RCAN1 or SFV-GFP. The infection efficiency is about 30 -40%. A colorimetric MTT assay and Hoechst and TUNEL staining were performed to determine the cell viability of primary neurons. Hoechst staining revealed that there is more nuclear condensation and fragmentation in RCAN1-1-overexpressing neurons exposed to H 2 O 2 (Fig. 5A, compare A2 with A1). The apoptotic features of RCAN1-overexpressing neurons with dexamethasone treatment were confirmed by TUNEL staining (Fig. 5A, A3-A8). Compared with SFV-GFP control, overexpression of RCAN1-1 renders more neuronal death, a 57.90 Ϯ 5.31% apoptosis ratio relative to 33.87 Ϯ 4.86% in control by Hoechst staining (p Ͻ 0.005) (Fig.  5B, lane 2 versus lane 1). RCAN1-1 overexpression also exacerbated the toxicity of H 2 O 2 on neurons, 97.02 Ϯ 0.20% compared with 70.18 Ϯ 0.93% in control (p Ͻ 0.005) (Fig. 5B, lane 4  versus lane 3), and significantly increased the dexamethasoneinduced neuronal apoptosis (80.37 Ϯ 2.80%, compared with 53.81 Ϯ 5.80% in control, p Ͻ 0.001) (Fig. 5B, lane 6 versus lane  5). The proapoptotic effect of RCAN1 was further confirmed by an MTT assay. RCAN1 overexpression resulted in a 54.01 Ϯ 5.17% reduction in A 540 nm relative to control (p Ͻ 0.0001) (Fig.  5C). To examine whether RCAN1 overexpression could further exacerbate neuronal death induced by A␤, a 10 M concentration of aggregated A␤ peptides was added to SFV-infected neurons for 12 h. Consistent with a previous report (51), A␤ induced neuronal apoptosis, as indicated by A 540 nm (69.11 Ϯ 1.45% of control) in the A␤-treated neuronal culture (p Ͻ 0.001) (Fig. 5C, lane 3 versus lane 1). The addition of A␤ to the RCAN1-1-overexpressing neurons further increased neuronal apoptosis, with the MTT assay showing a reduction to 26.39 Ϯ 1.14% of control (p Ͻ 0.0001) (Fig. 5C, lane 4 versus lane 3). These data clearly indicate that RCAN1-1 facilitates neuronal apoptosis.
RCAN1 Overexpression Activates Caspase-9-Our data suggested that overexpression of RCAN1 isoform 1 can activate caspase-3, so we then performed the following experiments to determine the upstream signaling molecules by which caspase-3 is activated. Upstream initiators of caspase-3 include caspase-8, -10, and -9. The cytochrome c-mediated pathway is essential in CNS development and stress-induced cell death (63). Upon exposure to intracellular proapoptotic agents, cytochrome c is relocated from the mitochondria to the cytoplasm, where it binds to Apaf-1 to form the so-called apoptosome, which subsequently recruits procaspase-9 and facilitates autoactivation of caspase-9. The activated caspase-9 then cleaves the downstream effector caspases, such as caspase-3. To investigate whether caspase-9 initiates RCAN1 overexpression-induced caspase-3 activation, caspase-9 activity was measured using the Caspase-Glo 9 assay. Caspase-9 activity was significantly increased by 2.18 Ϯ 0.1-fold in RCAN1-1-overexpressing neurons (p Ͻ 0.005) (Fig. 7A). Furthermore, a Western blot assay revealed that there were higher levels of cleaved forms of caspase-9 and reduced levels of procaspase-9 in RCAN1-1overexpressing HEK293 cells (Fig. 7B). The ratio of the cleaved form to the pro form was increased to 2.80 Ϯ 0.22-fold by RCAN1-1 overexpression (p Ͻ 0.05) (Fig. 7C). The results indicate that RCAN1-1 overexpression in cells activates caspase-9, which then recruits and activates its downstream executor, caspase-3, which subsequently cleaves its substrates and initiates the neuronal apoptotic pathway.
Caspase-11 is another upstream initiator of caspase-3 and mainly associated with inflammation in murine (64,65). Caspase-11 and caspase-3 activation have been observed in the spinal cord of mice with amyotrophic lateral sclerosis (66). To test if the activation of caspase-9 and caspase-3 by RCAN1 has a specific effect or if it has a panactivation effect on all of the caspase families, caspase-11 was examined in rat glial C6 cells transduced with SFV-GFP and SFV-RCAN1. Spleen tissue lysates derived from mice treated with LPS (40 mg/kg) were used as caspase-11 protein markers (67). A monoclonal anticaspase-11 antibody was used to detect caspase-11, and our study showed no significant difference in SFV-GFP-and SFV-RCAN1-transduced C6 cells (data not shown). These results suggest that overexpression of RCAN1 can specifically activate caspase-9 and caspase-3 but not caspase-11.
Activation of caspase-9 requires cytochrome c release from the mitochondria. To investigate whether RCAN1 overexpression affects cytochrome c release, a Western blot assay was used to examine cytochrome c levels in cytosolic and mitochondrial fractions (Fig. 7D). RCAN1-1 overexpression in HEK293 cells resulted in significantly reduced levels of cytochrome c in the mitochondria and markedly increased levels in the cytosol (p Ͻ 0.01) (Fig. 7E). Taken together, our data suggest that the cytochrome c, caspase-9, and caspase-3 signaling pathway is activated by RCAN1-1 overexpression and mediates neuronal apoptosis induced by RCAN1-1 overexpression.

DISCUSSION
Prominent neuronal death, neuritic plaques, and neurofibrillary tangles are neuropathological hallmarks in AD brains (68). The neuronal loss is closely correlated to memory impairment in AD patients. Most studies in AD have been focused on the initial steps leading to neuritic plaque and neurofibrillary tangle formation in AD brains, whereas the mechanism underlying neuronal death still remains elusive. RCAN1 has been shown to play an important role in memory and learning. Although disruption of RCAN1 in mice does not show overt abnormalities in vivo (69), both disruption and overexpression of nebula, a RCAN1 ortholog in Drosophila, lead to severe learning and memory deficits (70). RCAN1 is highly expressed in neurons in the brain and is overexpressed in the brains of AD and DS patients (38,40,43). Recently, we reported that degradation of (RCAN1 is mediated by both the chaperone-mediated autophagy and ubiquitin proteasome pathways, and alteration of the autophagy and ubiquitin proteasome pathways could result in dysfunction of RCAN1 signaling (71). Knock-out of RCAN1 in mice resulted in increased enzymatic calcineurin activity and cleaved calcineurin fragments and caused learning and memory deficits and impaired late phase long term potentiation (45). The effect of RCAN1 on AD and DS pathogenesis is unknown.
Epidemiological studies have revealed that risk factors for AD, including aging, atherosclerosis, stroke, and diabetes, will induce stress and increase oxidative free radicals in the brain, resulting in neuronal death in patients. Furthermore, hippocampal atrophy and severity of cognitive impairment in AD patients are correlated with dysfunctions of the HPA axis (72). Corticosteroids can facilitate neuronal apoptosis in the hip- pocampus as well as memory impairment (73)(74)(75)(76). In this report, we found that RCAN1 transcription is tightly regulated, and the stress hormone dexamethasone can specifically up-regulate RCAN1 gene expression by activating the neuronal tissueenriched isoform 1 promoter of the RCAN1 gene. Neurons in the hippocampus and cortex express a high level of GRs. HPA hyperactivity may result in elevated glucocorticoids, which in turn can up-regulate RCAN1 gene expression by binding to GRE on the RCAN1-1 gene promoter. Indeed, our results showed that RCAN1 protein expression was significantly increased in cortical tissues from DS and AD patients. We showed that the caspase-9 and caspase-3 apoptotic pathways were activated by overexpression of RCAN1-1 in neurons. Previous studies also suggest that activation of caspase-3 increases A␤ production (77,78). Moreover, A␤42 can directly stimulate RCAN1 expression in a U-373 cell line (38). Our studies showed that RCAN1-1 overexpression activated caspase-3. Such results may suggest a vicious cycle of A␤-RCAN1-caspase-A␤ and a link between A␤ accumulation and neuronal death in AD pathogenesis. RCAN1-1 protein was shown to inhibit calcineurin and protects against acute calcium-mediated stress damage, including transient oxidative stress, using hamster fibroblast HA-1 cells (44). It is also reported that disruption of RCAN1 in neurons protects neurons against oxidative stress (79), which is consistent with our findings using primary neuronal cultures. Our study implies that inhibition of RCAN1 may prevent neuronal death and benefit AD patients.
DS patients show AD-associated neuropathological changes in their 30s, which is 20 -30 years earlier than sporadic AD cases (80). We previously reported that A␤ is abnormally elevated in brains of DS patients (81). In addition to overproduction and accumulation of A␤, our current study implies that RCAN1, a gene localized on chromosome 21 and overexpressed in DS, may further contribute to the early onset of AD pathogenesis in DS due to its proapoptotic effect. Abnormal autophagy lysosome and ubiquitin proteasome pathways, resulting in protein accumulation, have been implicated in AD pathogenesis, and our recent report showed that RCAN1 degradation is mediated by both chaperone-mediated autophagy and ubiquitin proteasome pathways (71). Overexpression of RCAN1, due to its triplication in DS, can render neurons more vulnerable to stress and neurotoxic insults leading to hippocampal and cortical neuronal loss. We have demonstrated that A␤ further increases neuronal apoptosis in RCAN1-1overexpressing neurons. AD pathology is characterized by a specific demise of hippocampal and cortical neurons; however, how the specific population of cells is affected remains elusive. Our immunohistochemistry data showed that RCAN1 is specifically expressed in cortical and hippocampal neurons. The specific expression pattern of RCAN1 may explain the particular population of neuronal death in AD because RCAN1 overexpression in neurons leads to neuronal death. Our study provides a novel mechanism by which RCAN1 functions as a mediator of stress and A␤-induced neuronal death, and overexpression of RCAN1 due to an extra copy of the RCAN1 gene on chromosome 21 contributes to AD pathogenesis in DS. Our study suggests that inhibition of RCAN1 signaling may have pharmaceutical potential for reducing neuronal loss and treating cognitive impairments, thereby benefiting AD and DS patients.
It has been reported that the RCAN1 isoform 4 promoter can be regulated by the calcineurin-NFAT signaling pathway, thus forming a negative feedback loop in RCAN1 isoform 4 gene regulation because RCAN1 can inhibit calcineurin activity (14). The negative feedback loop in RCAN1 gene regulation implies that tight regulation of RCAN1 is vital to cell survival. The negative loop may also serve a self-protective function in cells by way of preventing RCAN1 overexpression, leading to cell demise. Previous studies showed that RCAN1 physically binds to calcineurin and inhibits its function (15,17,24,82). Activation of caspase-3 by RCAN1 is probably not due to this known function because previous studies show that calcineurin overexpression induces caspase-3 activation and apoptosis (83,84). Calcineurin promotes apoptosis via BAD dephosphorylation, and there is also evidence indicating that calcineurin inhibitors such as FK506 are neuroprotective (85,86). This suggests that the proapoptotic effect of RCAN1 is independent of its function of inhibiting calcineurin and that RCAN1 is a multifunctional protein. A previous report showed that nebula, the RCAN1 homolog in Drosophila, is crucial for mitochondrial integrity and function (87). Our fractionation and confocal microscopy imaging data indicate that RCAN1 is localized in cytosol, mitochondria, and nucleus, suggesting that RCAN1 is a multifunctional protein. The mitochondrial localization of RCAN1 may contribute to its proapoptotic effect. The underlying mechanism by which RCAN1 overexpression triggers cytochrome c release and initiates cell death remains unknown. Future studies will elucidate the involvement of RCAN1 in the mitochondrial apoptotic pathway.