Dysbindin Regulates the Transcriptional Level of Myristoylated Alanine-Rich Protein Kinase C Substrate via the Interaction with NF-YB in Mice Brain

Background An accumulating body of evidence suggests that Dtnbp1 (Dysbindin) is a key susceptibility gene for schizophrenia. Using the yeast-two-hybrid screening system, we examined the candidate proteins interacting with Dysbindin and revealed one of these candidates to be the transcription factor NF-YB. Methods We employed an immunoprecipitation (IP) assay to demonstrate the Dysbindin-NF-YB interaction. DNA chips were used to screen for altered expression of genes in cells in which Dysbindin or NF-YB was down regulated, while Chromatin IP and Reporter assays were used to confirm the involvement of these genes in transcription of Myristoylated alanine-rich protein kinase C substrate (MARCKS). The sdy mutant mice with a deletion in Dysbindin, which exhibit behavioral abnormalities, and wild-type DBA2J mice were used to investigate MARCKS expression. Results We revealed an interaction between Dysbindin and NF-YB. DNA chips showed that MARCKS expression was increased in both Dysbindin knockdown cells and NF-YB knockdown cells, and Chromatin IP revealed interaction of these proteins at the MARCKS promoter region. Reporter assay results suggested functional involvement of the interaction between Dysbindin and NF-YB in MARCKS transcription levels, via the CCAAT motif which is a NF-YB binding sequence. MARCKS expression was increased in sdy mutant mice when compared to wild-type mice. Conclusions These findings suggest that abnormal expression of MARCKS via dysfunction of Dysbindin might cause impairment of neural transmission and abnormal synaptogenesis. Our results should provide new insights into the mechanisms of neuronal development and the pathogenesis of schizophrenia.


Introduction
Schizophrenia is a common and devastating psychiatric disorder. Lack of patient compliance, due to undesirable side effects and efficacy restricted to positive symptoms, highlights the need to develop novel therapeutics. The etiology of the disease remains unknown, but in recent years a convergence of genetic, pharmacological, and neuroanatomical findings suggest that neural transmission [1][2][3][4] and synapse formation [5][6][7][8][9][10][11] are involved in schizophrenia. Recent studies suggest that disturbances of Dysbindin (dystrobrevin-binding protein 1; DTNBP1) are involved in this abnormal neural transmission.
The cause of schizophrenia is thought to involve the combined effects of multiple gene components. Genetic linkage and association studies have identified potential susceptibility genes such as Dysbindin [12,13], Neuregulin [14,15], Catechol-O-methyltransferase [16][17][18] and RG4 [19][20][21][22]. In particular, it has been reported that chromosome 6p is one of the highest susceptibility regions in linkage studies of schizophrenia [23,24]. Among them, genetic variants in a gene 6p22.3 expressing Dysbindin, which is identified as a protein interacting with dystrobevins [25], have been shown to be strongly associated with schizophrenia [12].
In studies on postmortem brain tissue, decreased levels of Dysbindin protein [26] and mRNA [27] have been shown in patients with schizophrenia compared with controls. Chronic treatment of mice with antipsychotics did not affect the expression levels of Dysbindin protein and mRNA in their brains [26,28], suggesting that evidence of lower levels of Dysbindin protein and mRNA in the postmortem brains of schizophrenics is not likely to be a simple artifact of antemortem drug treatment. In addition, previous reports have shown that diverse high-risk single nucleotide polymorphisms (SNPs) and haplotypes could influence Dysbindin mRNA expression [27,29]. These data indicate that the Dysbindin gene may confer susceptibility to schizophrenia through reduced Dysbindin expression.
Several lines of evidence suggest that Dysbindin may be associated with brain function. SNPs in Dysbindin have been associated with intermediate cognitive phenotypes related to schizophrenia such as IQ and working and episodic memory, and a Dysbindin haplotype has been associated with higher educational attainment [30,31]. In addition, several reports suggest the involvement of Dysbindin in cognitive functions [32][33][34]. These findings strongly suggest the importance of Dysbindin in brain function. At the cellular level, Dysbindin is located at both pre-and post-synaptic terminals [26,35], and is thought to be involved in postsynaptic density (PSD) function and the trafficking of receptors (NMDA, GABAergic, and nicotinic). Over-expression of Dysbindin increases glutamate release from pyramidal neurons in cell culture, possibly because of its role in vesicular trafficking [36]. Decreases in Dysbindin mRNA and protein levels have been reported in regions previously implicated in schizophrenia: the prefrontal cortex, midbrain, and hippocampus [26,27]. However, the molecular mechanisms of how decreases in Dysbindin expression may contribute to vulnerability to schizophrenia remain unknown.
Thus, we examined the interacting partners of Dysbindin using yeast two-hybrid analysis in order to help elucidate the function of Dysbindin. These interacting-protein data suggest that Dysbindin is involved in such processes as neurotransmission, cell signaling, the cytoskeleton and transcription. (Matsuzaki S et al. in submission). In addition, our previous reports suggest the following; (1) decreased expression of Dysbindin might increase dopamine release in the brain resulting in the observed abnormal behavior in sdy mice (Dysbindin KO mice) [37,38], (2) Dysbindin is likely involved in dopaminergic or glutamatergic transmission [36,39], (3) Dysbindin is likely involved in neurotransmission by binding with the BLOC1 complex, and with transcription by binding with transcriptionrelated genes (Matsuzaki S et al. in submission), (4) the expression level of Dysbindin may affect the expression of SNAP25 [36,39], (5) Dysbindin may play a key role in coordinating JNK signaling and actin cytoskeleton required for neural development [40]. These findings suggest that Dysbindin may influence neurotransmission and neural development via interaction with other factors or by regulation of transcription.
In a previous paper, we identified several Dysbindin interacting partners including the transcription factor, nuclear transcription factor Y beta (NF-YB) (Matsuzaki S et al. in submission). NF-YB belongs to a family of CCAAT-binding transcription factors, which are important for the basal transcription of a class of regulatory genes and are involved in cellular reactions [41][42][43][44]. Subsequently, in this study, we examined the functional involvement of Dysbindin in transcription via its interaction with NF-YB. As a result, we showed that the NF-YB/Dysbindin complex regulates the transcription of MARCKS via interaction with certain CCAAT sequences, and abnormal NF-YB/Dysbindin interaction could cause alterations such as impaired neural transmission and abnormal development of neurons.

Dysbindin Exists within the Nucleus in Addition to the Cytoplasm
We examined the existence of Dysbindin in the nucleus, because Dysbindin should exist within the nucleus to play a functional role in transcriptional regulation. We used an overexpression vector for Dysbindin tagged with -FLAG or -V5 to check the intracellular localization of Dysbindin. The fractionation study using Dysbindin-FLAG-overexpressing HEK293 cells shows that Dysbindin exists mainly in the cytosol while a small amount exists in the nucleus ( Figure 1A), and Dysbindin-V5 showed the same results (data not shown). These results are in accordance with a previous report [45]. Morphologically, Dysbindin is localized mainly in the cytoplasm with a perinuclear high density region in HEK293 cells and SY5Y cells; however, a faint immunoreaction was also seen within the nucleus ( Figure 1B -a and -b). Furthermore, pretreatment with leptomycine-B (LPB), which inhibits export from the nucleus to the cytoplasm, caused a slight Dysbindin increase in cells, which then showed nuclear localization of Dysbindin ( Figure 1B -c and -d ). These findings suggest that Dysbindin protein is shuttled between the nucleus and the cytoplasm.

Dysbindin Binds to the Transcription Factor NF-YB
Using yeast two-hybrid screening, we identified several transcriptional factors as candidates that may interact with Dysbindin. We selected NF-YB, one of the candidates, and confirmed a Dysbindin-NF-YB interaction by immunoprecipitation assay using HEK293T cells which express NF-YB endogenously ( Figure 1A). HEK293T cells were transfected with expression vectors for Dysbindin-V5, and cell lysates were subjected to immunoprecipitation with anti-V5 or anti-NF-YB antibodies, followed by Western blot analysis with a reciprocal antibody. NF-YB was detected in the immunoprecipitates with an anti-V5 body, comparing to the immunoprecipitates with control IgG (Figure 2A), while Dysbindin-V5 was detected in the immunoprecipitates with an anti-NF-YB antibody, comparing to control IgG (data not shown). Thus, Dysbindin and NF-YB are physiologically associated with each other in transfected mammalian cells.
To further our research, we produced a specific anti-Dysbindin antibody with high titer. The antibody detects endogenous Dysbindin in cell and mouse brain samples, though it did not detect any bands corresponding to Dysbindin from the lysates of Dysbindin knockout mouse brain [40]. The existence of endogenous Dysbindin and endogenous NF-YB in lysates from SH-SY5Y cells was confirmed by Western Blot ( Figure 2B, 3 rd lane of both panels). Immunoprecipitation using the lysates with antibodies for Dysbindin and NF-YB and subsequent Western blot revealed the interaction of endogenous Dysbindin with endogenous NF-YB ( Figure 2B, 2 nd lane of both panels), and this binding was also confirmed using adult mouse brain lysates (data not shown).

Downregulation of Dysbindin Causes Upregulation in Expression Levels of Myristoylated Alanine-Rich Protein Kinase C Substrate (MARCKS)
As shown above, we had revealed an interaction between Dysbindin and NF-YB. This result suggests that Dysbindin may be functionally involved in transcription of some genes regulated by NF-YB. We screened for genes displaying altered expression by means of a DNA chip, using RNA extracts from the Dysbindin or NF-YB knockdown human neural cell line, SH-SY5Y. The expression of either Dysbindin or NF-YB was decreased by the corresponding siRNA for each gene, and the effects of siRNA on Dysbindin or NF-YB were confirmed by Western blot analysis ( Figure S1). The genes showing increased expression in the Dysbindin knockdown cells, as well as in the NF-YB knockdown cells, are listed in Table 1A, while those showing decreased  expression are listed in Table 1B. Next, using the DANASIS 2.0 system or sequencing of the promoter region, we screened for genes having the CCAAT sequence in the promoter region, because NF-YB is known to bind with high specificity to the CCAAT motif in the promoter region of a variety of genes (Table  1, gene names shown in red). We then focused on three genes; Myristoylated alanine-rich protein kinase C substrate (MARCKS) [46][47][48], Phospholipase C beta 4 (PLCB4) [49] and Synaptotagmin 1 (SYT1) [50], because an accumulating number of reports point to the involvement of impaired neural transmission in the schizo-   phrenia pathology. In addition, we considered the involvement of the genes in psychiatric diseases and we narrow down to MARCKS [51] and SYT1 [52]. Interestingly, a previous report suggests the alteration of SYT1 in schizophrenia patients [52]. The paper shows increase of SYT1 mRNA in younger schizophrenia patients group, while it shows decrease of SYT1 mRNA in older schizophrenia patients. These results suggest the complicated and multiple regulation of SYT1 transcriptional regulation. Thus, we examined the functional involvement of the Dysbindin-NF-YB interaction in MARCKS transcription.
To confirm the involvement of the knockdown of Dysbindin or NF-YB in the upregulation of MARCKS, we performed Western blot analysis using Dysbindin or NF-YB knockdown SH-SY5Ycells. Comparing the expression level of the MARCKS protein with that of control cells, Dysbindin knockdown cells showed upregulation of MARCKS protein ( Figure 3A). To confirm the effect of Dysbindin on MARCKS in vivo, we examined the expression of MARCKS protein in the hippocampus with advancing age of the Dysbindin knockout mice, comparing with that found in wild-type mice. In the wild-type mice, a peak in MARCKS protein expression in the hippocampus was identified at postnatal day 15 and 20 ( Figure 3B), and then decreased markedly over time. However, such a decrease was not observed in the Dysbindin knockout mice, where large amounts of Dysbindin protein were still expressed in the hippocampii of older mice ( Figure 3B). These findings suggest that downregulation of Dysbindin may enhance transcription of the MARCKS gene, resulting in the upregulation of MARCKS protein.
We performed chromatin IP analysis using SH-SY5Y cells overexpressing Dysbindin-Flag, to explore the possibility that the Dysbindin-NF-YB complex could affect the transcription of MARCKS via interaction with the promoter region of MARCKS. The cells were stimulated by retinoic acid to induce MARCKS, and were collected as the samples for chromatin IP. PCR products from the chromatin IPs suggest that Dysbindin and NF-YB simultaneously interact with the promoter region of MARCKS, but control IgG experiments did not show this result ( Figure 3C).
These findings indicate that the Dysbindin-NF-YB complex interacts with the promoter region of the MARCKS gene resulting in inhibition of MARCKS transcription.
The Transcriptional Level of the MARCKS Gene Is Regulated by Dysbindin via the NF-YB Binding Motif, CCAAT-2 As shown in Figure 4A, the 592UTR region of the MARCKS gene has two kinds of CCAAT sequences; one CCAAT motif located between UTR 21152 and 2700 and the other located between UTR 2700 and 2614. In this study, we tentatively named the former CCAAT sequence ''CCAAT-1'' and the latter ''CCAAT-2.'' It is well known that NF-YB binds to the CCAAT motif to regulate transcription of target genes. Thus, we examined whether CCAAT motifs are essential to the regulation of MARCKS transcription by means of a luciferase assay, using the following five vectors containing shorter RNA probes; UTR(1152)-Luc, UTR (953)-Luc, UTR(700)-Luc, UTR(614)-Luc, and UTR(462)-Luc ( Figure 4A). These constructs were transiently transfected into SH-SY5Y cells which express Dysbindin and NF-YB endogenously, and luciferase activity in each cell line was measured 24 hours after stimulation with retinoic acid. As baseline, we used luciferase activity detected in the SH-SY5Y cells expressing the UTR(1152)-Luc after retinoic acid stimulation ( Figure 4A). In the cells transfected with UTR (953)-Luc containing both CCAAT sequences and UTR(700)-Luc containing the CCAAT-1 sequence but lacking the CCAAT-2 sequence, luciferase activity remained at baseline level after stimulation with retinoic acid ( Figure 4A). However, luciferase activity was markedly increased in the cells expressing UTR(614)-Luc after retinoic acid stimulation ( Figure 4A). These results suggest that the CCAAT-2 motif plays an important role in inhibition of MARCKS transcription. Furthermore, the SH-SY5Y cells transfected with UTR(462)-Luc lacking CCAAT-1, CCAAT-2 and the Sp1 region showed very low luciferase activity ( Figure 4A), indicating that Sp1 is indispensable for MARCKS transcription.   Figure 4B). Luciferase activity was detected from the SH-SY5Y cells transfected with each probe, and luciferase activity detected in the cells transfected with UTR(1152)-Luc was used as the baseline value ( Figure 4B). Cells transfected with M-UTR(1152)-Luc and those transfected with D4-UTR(1152)-Luc exhibited marked increases in luciferase activity ( Figure 4B), showing that the CCAAT-2 motif plays a key role in inhibition of MARCKS transcription. Furthermore, cells expressing D1-UTR(1152)-Luc, D2-UTR(1152)-Luc or D3-UTR(1152)-Luc exhibited no luciferase activity. These findings suggest that the sequence downstream of the Sp1 region, as well as the Sp1 region itself, is indispensable for MARCKS transcription.
To confirm the involvement of Dysbindin in the altered MARCKS transcription levels via the CCAAT-2 motif, we compared the luciferase activity of UTR(1152)-Luc detected in Dysbindin knockdown cells with that of control cells. As shown in Figure 4C, knockdown of Dysbindin resulted in upregulation of luciferase activity in the UTR(1152)-Luc transfected cells. However, the effect of knockdown of Dysbindin on luciferase activity was not observed in the D1-UTR(1152)-Luc transfected cells ( Figure 4D). These results suggest that Dysbindin regulates MARCKS transcription via the CCAAT2 motif; the NF-YB binding site. On the other hand, since negligible levels of luciferase activity were observed in cells transfected with any of the probes lacking the sequence downstream of the Sp1 region, the sequence downstream of Sp1 appears to be essential for MARCKS transcription ( Figure 4A and 4B).

Discussion
Numerous reports support the role of Dysbindin in the etiology of schizophrenia [13,30,37,[53][54][55][56][57][58][59][60]. Previous studies have reported a decrease in Dysbindin expression in the brains of schizophrenic patients both at the mRNA and protein levels [26,27]. However, the functional involvement of Dysbindin in the neural system is not yet well elucidated. In this study, we examined involvement of Dysbindin in neural transmission and neural formation via transcriptional regulation, because abnormalities in these neural processes are very important in the pathogenesis of schizophrenia.

Regulation of MARCKS Transcription by the Dysbindin/ NF-YB Interaction
As a result of the yeast-two-hybrid assay and immunoprecipitation assay, we revealed an interaction between NF-YB and Dysbindin (Figure 1 and 2). In addition, we showed the binding of NF-YB and Dysbindin to the MARCKS promoter region ( Figure 3C). These findings suggest involvement of this complex in transcriptional regulation of MARCKS. As shown in Figure 4, we found two CCAAT sequence motifs at the 592UTR of the MARCKS gene. Previous reports show that members of the NF-Y family including NF-YB bind to CCAAT sequences and can regulate transcription of a number of genes. Our results suggest that one of the CCAAT sequences, CCAAT-2, is important for MARCKS transcriptional regulation. On the other hand, our luciferase assay results suggest that both the Sp1 region and the sequence downstream of Sp1 are indispensable for MARCKS transcription ( Figure 4A and 4B).

Dysbindin Knockdown Increases MARCKS Protein Levels In Vivo and In Vitro
In accordance with the enhanced MARCKS transcription mediated by the knockdown of Dysbindin, Dysbindin knockdown cells show increased MARCKS levels ( Figure 3A). Next, we examined the expression level of MARCKS in Dysbindin knockout mice. As shown in Figure 3b, in the wild-type mouse brain the peak in MARCKS expression is at postnatal day 15; thereafter decreasing markedly with advancing age until only low levels of MARCKS expression are seen in adults (P45). Comparable alternations in MARCKS expression were also observed in another mouse line, ICR (data not shown). These findings support the hypothesis that MARCKS plays an important role in brain development. However, in the Dysbindin knockout mice, there is no effect on MARCKS expression during the developmental stage, when MARCKS is abundantly expressed in wild-type mice. During this stage, MARCKS transcription may be regulated by multiple molecules, which compensate for the lack of Dysbindin. With increasing age of the mouse, MARCKS expression decreases gradually to a low level of expression in adults (Figure 3b). In contrast, a decrease in MARCKS expression was not observed in Dysbindin knockout mice ( Figure 3b) and even in adult mice brains, a high level of expression of MARCKS was detected. These findings show that Dysbindin likely plays a major role in regulation of MARCKS expression in the adult brain, in contrast to in the developmental stage. Therefore, considering the results in Dysbindin knockout mice, it is likely that MARCKS is expressed at high levels in schizophrenic brains, compared with age-matched control brains.

MARCKS and Neural Transmission
It has been shown that MARCKS impacts on neurotransmission via F-actin and on vesicular transport via synaptic vesicles [46][47][48]. Furthermore, many reports indicate that dopaminergic transmission is increased in the brains of schizophrenics [1][2][3][4]. Dopamine D2 antagonists are an effective treatment in schizophrenia, and dopamine-enhancing drugs mimic psychotic symptoms of schizophrenia. In the schizophrenic brain, the expression of Dysbindin is decreased, resulting in an increase in MARCKS protein expression, which impacts on neurotransmission. Furthermore, we found that decreases in Dysbindin levels upregulate dopamine release [39]. Therefore, the enhanced dopaminergic transmission produced by the lower expression level of Dysbindin may be partially attributable to activation of MARCKS. Thus, the impairment of neural transmission in the schizophrenic brain may be caused by alterations of MARCKS expression levels via changes in Dysbindin.

Dysbindin May Regulate Neural Formation via Alteration of MARCKS Levels
Many studies support the hypothesis that schizophrenia is a neurodevelopmental disease. Disrupted-In Schizophrenia 1 (DISC1) is a gene disrupted by a (1;1) (q42.1; q14.3) translocation that segregates with major psychiatric disorders, including schizophrenia in a Scottish family [61,62]. Previously, we examined the physiological role of the molecular complex composed of DISC1 and its interacting partners, Fasciculation and elongation protein zeta 1 (Fez1) [63] and DISC1-Binding Zinc finger protein (DBZ) [64]. Both the DISC1-Fez1 interaction and the DISC1-DBZ interaction are involved in neurite extension. These reports suggest that abnormalities in the schizophrenia susceptibility genes, such as DISC1, likely cause an impairment of brain development resulting in schizophrenia. In addition, several reports suggest that the PKC signal is involved in psychiatric disorders, as well as other signals such as ERK, which play important roles in neural development. In addition, we previously showed the importance of Dysbindin for growth cone formation [40]. These previous reports suggest that abnormal neural formation could cause psychiatric disorders and that Dysbindin may be one of the important factors in normal neural development. In this study, we demonstrate the transcriptional regulation of MARCKS via Dysbindin and the upregulation of MARCKS by downregulation of Dysbindin. Since MARCKS is involved not only in neural transmission 48 but also in neural developmental processes such as synaptogenesis and maintaining spine morphology [46,47], these results suggest that dysfunction of Dysbindin likely causes the upregulation of MARCKS and may induce abnormal development of the nervous system via alterations of MARCKS levels.
Thus, in this paper, we report the following findings; (1) Dysbindin interacts with NF-YB, (2) NF-YB and Dysbindin bind to the promoter region of MARCKS, (3) one of the CCAAT sequences is likely essential for the transcriptional regulation of MARCKS and (4) the downregulation of Dysbindin upregulates the expression of MARCKS in vitro and in vivo. On the other hand, we previously showed that Dysbindin knockout mice exhibit schizophrenia-like behavior and abnormalities of the dopaminergic system. These phenotypes may be at least partly attributable to over-activation of MARCKS via a decrease in Dysbindin levels.
In conclusion, these results may help shed some light on the causes of schizophrenia, and indicate that the transcriptional regulation of Dysbindin may contribute to schizophrenia. Further studies of Dysbindin and its association with MARCKS and with schizophrenia may reveal novel treatment targets for schizophrenia.

Antibodies
Monoclonal anti-Dysbindin antibody was produced. Briefly, GST-fused human Dysbindin was used as antigen and the Dysbindin protein for ELISA was made by thrombin digestion of GST-Dysbindin. High-titer clones for Dysbindin were selected by ELISA using the Dysbindin protein and the immunoreactivity of
Animals sdy mice (Dysbindin KO mice) and wild-type littermates were provided by the Takeda lab, Department of Psychiatry, Osaka University Graduate School of Medicine. The mice were deeply anesthetized with sodium pentobarbital. Brains (hippocampus) were dissected from each aged mouse. All animal experiments were carried out in accordance with a protocol approved by the Institutional Animal Care and Use Committee of Osaka University.
Immunocytochemistry SY5Y cells were grown on poly-l-lysine-coated four-well chamber dishes at a density of 3610 4 cells/cm 2 . The cells were fixed in 2% paraformaldehyde in 0.1 M PBS, permeabilized, and blocked with 0.02 M PBS containing 0.3% Triton X-100, 3% BSA and 10% goat serum for 30 min at room temperature, and then incubated with antibodies specific for the individual protein.
Confocal microscopy was performed using a Carl Zeiss LSM-510 confocal microscope.

Fractionation Assay
Cells were collected after washing with ice-cold PBS. Cells and brains were homogenized in Tris buffer (20 mM Tris-HCl, pH 7.8, 1 mM EDTA, 150 mM NaCl and protease inhibitor cocktail (Roche)). After homogenization, the homogenized proteins were lysed by the addition of 0.5% NP-40 for 30 min on ice and centrifuged at 500 6g for 10 min to collect the nuclear pellet. The supernatant was collected as the cytosolic fraction.

Immunoprecipitation (IP)
After washing cells with ice-cold PBS, cells were collected and resuspended in 1 mL lysis buffer (20 mM Tris-HCl, pH 7.8, 0.2% NP-40, 1 mM EDTA, 150 mM NaCl and protease inhibitor cocktail (Roche)). Cells were frozen in dry ice/EtOH and stored at 280uC. Cell lysates were incubated on ice for 30 min and then centrifuged for 5 min at 13,6006g. After centrifugation, the supernatants were precleared with protein Sepharose G beads and IP was carried out in lysis buffer with antibody/protein G Sepharose beads for 1 h at 4uC. After washing in lysis buffer, immunoprecipitated proteins were immunoblotted.

Immunoblotting
Aliquots of whole cell lysates or IP lysates separated by SDS-PAGE were blotted onto an Immobilon-P membrane (Millipore), and then incubated with antibodies specific for individual protein.
Proteins were detected by ECL plus Western Blotting Detection System (GE Healthcare), followed by exposure to X-ray films according to the manufacturer's protocol.

RNA Extracts and Microarray
Total RNA was extracted from cells using RNeasy columns (Qiagen) according to the manufacturer's instructions. Five hundred nanograms of total RNA from control and experimental cells was separately amplified and labeled with either Cy3-or Cy5labeled CTP (Perkin Elmer) with an Agilent low input linear amplification kit (Agilent Technologies) according to manufacturer's instructions. After labeling and cleanup, amplified RNA was quantified by UV-vis spectroscopy. One microgram each of Cy3and Cy5-labeled targets were combined and hybridized with a Whole Human Genome Oligo Microarray Kit (G4112F)according to the manufacturer's instructions. Three biological replicates were used at each time point with one of the replicates being a dye reversal of the other two. Microarrays were imaged on a Hitachi image scanner and data analyzed with GeneSpring 6 (Silicon Genetics).

Chromatin Immunoprecipitation (ChIP) Assay
ChIP analysis was performed using a Chromatin Immunoprecipitation Assay Kit (Upstate Biotechnology) according to the manufacturer's instructions. Briefly, protein2DNA complexes were crosslinked with 1% formaldehyde (10 min at room temperature) and cells were harvested. DNA was sonicated to lengths of 50021000 bp. Antibodies specific for individual protein were used for immunoprecipitating protein-DNA complexes overnight at 4uC. PCR was performed with individual specific primer sets for the MARCKS promoter: the proximal CCAAT region, 592GGTTTGCTCTTTGATGCTCTTGAT239 and 592ACTTTCGGGTGGGGTGTAA239

Reporter Assay
Reporter plasmids were transfected into cells using Lipofectamine 2000 (Invitrogen) together with phRG-TK (Renilla reporter for internal control) which monitored transfection efficiency. Luciferase activities were assayed using the Dual Luciferase Assay System (Promega). All assays were performed three times in duplicate and values are shown as means 6 SD. Figure S1 The preparation of mRNAs for microarray analysis. (A-(a) and B-(a)) To prepare RNAs for microarrays analysis, we transfected the siRNA for Dysbindin, NF-YB, or scrambled as a control. The effect of each RNAi was confirmed by Western blot using the antibody for Dysbindin or NF-YB. (A-(b) and B-(b)) The columns and vertical bars denote the means 6 SEM (triplicate independent experiments; P,0.001, Student's t-test). Dysbindin or NF-YB was knocked-down significantly by transfection of the siRNA for Dysbindin or NF-YB, compared with the control. Found at: doi:10.1371/journal.pone.0008773.s001 (1.10 MB EPS)