Restricted expression and photic induction of a novel mouse regulatory factor X4 transcript in the suprachiasmatic nucleus.

The regulatory factor X (RFX) family of transcription factors is characterized by a unique and highly conserved 76-amino acid residue DNA-binding domain. Mammals have five RFX genes, but the physiological functions of their products are unknown, with the exception of RFX5. Here a mouse RFX4 transcript was identified that encodes a peptide of 735 amino acids, including the DNA-binding domain. Its expression was localized in the suprachiasmatic nucleus, the central pacemaker site of the circadian clock. Also, light exposure was found to induce its gene expression in a subjective night-specific manner. Polyclonal antibodies were prepared, and an 80-kDa band was detected in the suprachiasmatic nucleus by Western hybridization. A histochemical study showed a localization of the products in the nucleus. This is the first report on mouse RFX4, which contains the RFX DNA-binding motif. Our investigation may provide clues to the physiological function of RFX4.

Studies using mutants of the RFX orthologues in other organisms have suggested possible functions. Crt1 in Saccharomyces cerevisiae encodes a transcription repressor involved in DNA damage and the replication block checkpoint pathways (6).  in Caenorhabditis elegans plays a critical role in ciliated sensory neuron development (8). dRFX in Drosophila melanogaster is also expressed in the sensory neurons during embryonic development (9,16).
RFX4 was first isolated from human breast cancer cells as a chimeric molecule with the estrogen receptor (17). Two transcripts have been identified in humans (GenBank TM accession numbers AF332192 and AB044245) and one in mice (AK016791), but the mouse transcript fails to encode the RFXspecific DNA-binding domain. This suggests the presence of another RFX4 transcript in mice containing the DBD. Neither target genes nor the physiological role of RFX4 are known, but abundant expression in the testis and some expression in the brain have been shown in humans (11). In this study, a novel mouse RFX4 transcript from the brain was isolated and characterized.
Almost all organisms have an endogenous circadian timekeeper that governs most phenomena in the organism, including its behavior (18). The endogenous oscillator functions on a period of about, but not exactly, 24 h (19 -21). This period is synchronized to the day/night environmental cycle by an entrainment mechanism (22). In mammals, the central oscillator resides in the hypothalamus suprachiasmatic nucleus (SCN) (23,24). Interestingly, the mouse RFX4 transcript was localized to the SCN.

EXPERIMENTAL PROCEDURES
Mice and Tissues-Male mice (C57BL/6 strain) purchased from Japan SLC, Inc. (Hamamatsu, Japan) were trained in light/dark (LD) cycles (12L:12D) for at least 10 days before being transferred to constant darkness (DD). On the 4th day under DD, the cortex and SCN were isolated by micropunching. For gene expression analysis in response to light exposure, the mice were exposed to light at circadian time (CT)14 and sacrificed 20 or 45 min later. For further RFX4 induction studies, mice exposed to light at CT6, -14, or -22 were sacrificed 45 min later. The light exposure intensity was 200 lux. Control mice were sacrificed at the same CT time points but without light exposure. To determine whether the gene expression was rhythmic, the animals were sacrificed at CT2, -6, -10, -14, -18, and -22. Then, the SCN and cortex tissues were perforated using a puncher with a 400-m inner diameter. SCN tissues were obtained from the anterior section, and the cortical tissues were obtained near the most dorsal area from the next caudal section. The pineal gland and retina were prepared as described elsewhere (25).
RNA Preparation, cDNA Synthesis and Sequencing-Total RNA was prepared from the SCN, cortex, and testis with TRIzol (Invitrogen) and then treated with DNase-I (Invitrogen). First strand cDNA was synthesized with an oligo-(dT)18 primer using the SuperScript II First-Strand * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The nucleotide sequence(s) reported in this paper has been submitted to the DDBJ/GenBank TM /EBI Data Bank with accession number(s) AB089184 and AB086957.
Synthesis System according to the manufacturer's instructions (Invitrogen). Reverse transcriptase (RT)-PCR was performed in a 25-l reaction, which included 2.5 units of KOD DNA polymerase (TOYOBO, Osaka, Japan). The primers were F1 (5Ј-TCC ACT AGT TCT TTC TTT TCC CCT TTT ATC-3Ј), R1 (5Ј-TTA TTT GAG TGT GAA ACC ATC CAT ATC TTG-3Ј), F2 (5Ј-ACT TTC TGG TAA CTA CAT CGG CTG AAA AT-3Ј), and R2 (5Ј-CTA GAC CAG GGA CAG AAA TTC ACG TTC TC-3Ј). The PCR conditions were 96°C for 4 min followed by 35 cycles of 98°C for 20 s, 60°C for 10 s, and 74°C for 1.5 min. The RT-PCR products were cloned into the pGEM-T Easy vector (Promega, Madison, WI) and sequenced. To prevent any misreading due to nucleotide misincorporation by the DNA polymerase, RT-PCR was conducted independently three times. For DNA sequencing, three clones were picked from the PCR products obtained from all three independent amplifications. We aligned these 9 sequences, determined the sequence of mNYD-sp10 and bRfx4, and registered the sequences in DDBJ with accession numbers AB089184 and AB086957, respectively.
Messenger RNA Quantification-The amount of mouse RFX4 mRNA was measured by real-time PCR using SYBR-Green PCR Master Mix (Applied Biosystems, Foster City, CA) with the ABI PRISM 7700 (Applied Biosystems) (26). The amount of mRNA was determined through normalization with the glyceraldehyde 3-phosphate dehydrogenase (Gapdh) mRNA signal. The primers were mouse RFX4 (forward, 5Ј-GGC ATA GCG GTG AAG GAG AG-3Ј; reverse, 5Ј-AAA GTC TGG CAG CAA TGT CC-3Ј), mPer2 (forward, 5Ј-GGT CGA CAT TAG ACG GTG CT-3Ј; reverse, 5Ј-CCC CCA AAT TCC GTA TTT CT-3Ј) and GAPDH (forward, 5Ј-GAG GCC GGT GCT GAG TAT GTC GT-3Ј; reverse, 5Ј-GGT GGT GCA GGA TGC ATT GCT G-3Ј). Pre-heating was performed at 95°C for 10 min followed by 50 cycles of 95°C for 15 s, 60°C for 30 s and 78°C for 40 s. Five mice were used for the real-time PCR analysis for each sample point, and the analysis was performed twice.
In Situ Hybridization-Two mice were used for the in situ hybridization analysis at each sample point. Whole brain tissue was fixed in a 4% paraformaldehyde/phosphate-buffered saline solution for 24 h followed by immersion in a 20% sucrose/phosphate-buffered saline solution for 24 h. Then, the brains were sliced into 30-m sections with a Cryostat (Leica, Nussloch, Germany). To prepare the probes for analysis, bRfx4-specific regions were amplified by RT-PCR and cloned into pGEM-T Easy (see Fig. 5C). After DNA sequence confirmation, this fragment was utilized as a probe. The primers employed were the following: forward, 5Ј-ATT ACT GAG TGT CCC CTC GC-3Ј; reverse, 5Ј-GGG TTC CTC CAG TAA CCC AC-3Ј (see Fig. 5C). Radiolabeled probes were synthesized using [␣-33 P]UTP (PerkinElmer Life Sciences) with the manufacturer's protocols for cRNA synthesis. Hybridization was performed as described by Tei et al. (27). The hybridized sections were exposed to BioMax film (Kodak, Rochester, NY). To determine the quantitative number of the transcripts, the optical densities of the SCN from six sections from the rostral to caudal ends were quantified with an Imaging Densitometer (Bio-Rad). The intensities of the six sections were summed after conversion into the relative optical densities produced by the 14 C-acrylic standards (Amersham Biosciences). Data were normalized with respect to the difference between signal intensities in equal areas of the SCN and the corpus callosum.
To verify the antibody specificity, two COS7 transfectants were made with either the mouse bRfx4 or FLAG-tagged mouse bRfx4 constructs. Full-length mouse bRfx4 cDNA was obtained by the PCR with primers forward 278 (5Ј-ATG CAT TGT GGG TTA CTG GAG-3Ј) and reverse 2,485 (5Ј-TCA CTT AGC CCA TCC AGT G-3Ј) using total RNA isolated from the SCN and cloned into the pME18S expression vector. For the FLAG-tagged mouse bRfx4 construct, the pCMV-Tag2 vector was used (Stratagene, La Jolla, CA). Both constructs were confirmed by sequencing.
The SCN and testis were prepared from male mice (C57BL/6), and their extracts were obtained as described previously (28). The extracts of the COS7 cells transfected with bRfx4 or FLAG-tagged bRfx4 were also prepared and used for a control. These extracts were separated by 7.5% SDS-PAGE and then transferred to an Immobilon polyvinylidene difluoride membrane (Millipore, Bedford, MA). The membranes were reacted with an anti-FLAG M2 monoclonal antibody (Sigma) or rabbit anti-RFX4 polyclonal antibodies followed by anti-mouse IgG (Dakocytomation, Glostrup, Denmark) or anti-rabbit IgG (Dakocytomation) treatment. Signal detection was performed with SuperSignal West Dura extended duration substrate reagent (Pierce), and quantitation was performed with a LightCapture (AE-6960, ATTO Corporation, Tokyo, Japan). Immunoprecipitation was performed according to the method described by Morotomi-Yano et al. (11). The lysis buffer consisted of 50 mM Tris-HCl (pH 7.9), 1% Nonidet P-40, 2.5 mM EDTA, 500 mM KCl, and a protease inhibitor mixture (Wako, Osaka, Japan).

RESULTS AND DISCUSSION
A Novel Form of Mouse RFX4 Transcript Is Expressed in the SCN-Recently, the full-length human RFX4 (hereafter referred to as hRFX4) cDNA was isolated, and the interaction of its product with other RFX products was demonstrated. Nonetheless, its physiological function remains unknown. Gene expression analysis of hRFX4 showed that its expression is restricted to the testis (11). In mice, the RFX4 (AK016791) transcript was identified in the testis, but it lacks the RFXspecific DBD (30).
RFX4 expression was examined in several tissues in mice, and its expression was found in the testis, brain, heart, and ovary with expression ratios of 1, 1/36, 1/100, and 1/200, respectively. We failed to detect its expression in other tissues such as kidney, spleen, thymus, liver, muscle, lung, and uterus. These results are consistent with the publicly expressed sequence tag (EST) information. In this study, we focused on the transcript expressed in the brain. Further analyses using the SCN and cortex tissues revealed significant expression in the SCN, which serves as a biorhythm center (Fig. 1).
Two types of alternatively spliced products of RFX4, NYD-sp10 and hRFX4, have been isolated in humans. The hRFX4 product has the DBD motif, which is a characteristic of the RFX transcription factor family (11). NYD-sp10 lacks this domain. In mice, one transcript, AK016791, has been identified, but it lacks the DBD motif (11,30). To date, these findings suggest the existence of a second RFX4 transcript in mice, which encodes an RFX4 homologue containing the DBD motif. A BLAST search (31)  were homologous to hRFX4 or AK016791. PCR using total RNA purified from the mouse brain with primers designed in the 5Ј-ESTs and 3Ј-ESTs that were registered in Mm. 32654 generated a 3.7-kbp fragment. Sequencing this fragment revealed that the transcript was different from the previously reported transcripts. The novel transcript, which we refer to as brainspecific RFX4 (bRfx4), encodes a peptide consisting of 735 amino acid residues containing the B, C, DIM, and DBD domains (Fig. 2). This is the first reported mouse RFX4, which contains the DBD motif. We also obtained a mouse orthologue of NYD-sp10, mNYD-sp10, using primers F2 and R1, which were designed using the NYD-sp10 5Ј-untranslated regions and 3Ј-untranslated regions from the mouse genome (Gen-Bank TM accession number NW_000030) (Fig. 3, A and B, lane  F2 ϩ R1 in the testis). A search for bRfx4 in the public genome sequence data base revealed that the 5Ј-ESTs had been tran-scribed from novel exons upstream of the first exon of the mNYD-sp10 transcript (Fig. 3A).
Taken together, our findings suggest that bRfx4 expression is SCN-specific and that mNYD-sp10 and AK016791 expression is testis-specific. To confirm this, the expression of these mouse RFX4 transcripts was compared in the SCN and the testis using several primers specific for each transcript. We found that bRfx4 is expressed in the SCN, but mNYD-sp10 and AK016791 are only expressed in the testis (Fig. 3). In situ hybridization was also performed with testis sections using a bRfx4-specific probe (Fig. 5C) but failed to detect any signal (data not shown). As mentioned above, we also searched for a mouse hRFX4 orthologue, but we did not find any ESTs in the public data base.
Induction of the bRfx4 Transcript by Light Exposure-Though the disruption of the RFX5 gene causes immunodefi- ciency in mice (12,15), the physiological roles of the other RFX genes in mammals are unknown. The abundant expression of mouse RFX4 in the SCN may indicate that bRfx4 function is related to the circadian clock, because the SCN is the central clock locus for circadian rhythm. After real-time PCR, in situ hybridization was performed using brain sections, and strong signals were found in the SCN with three different probes prepared from the bRfx4 transcript. We also found significant but weak signals in the nucleus accumbens of the cerebellar cortex (data not shown).
The circadian rhythm generated in the SCN is transferred to most of the peripheral tissues in the body. Although peripheral tissues also have their own oscillator, they are thought to be governed by the central clock in the SCN. The circadian generelated phenotypes were examined in bRfx4 expression. One of these phenotypes is an induction of gene expression by light exposure in a subjective night-specific manner. To date, several such genes have been identified in the SCN (32)(33)(34)(35)(36)(37)(38)(39)(40). Another characteristic observed in some of the circadian genes is their rhythmic expression throughout an entire day (41)(42)(43).
The expression change of mouse bRfx4 was analyzed as a result of light exposure in the SCN using real-time PCR. We detected a slight and an ϳ2.5-fold induction at 20 and 45 min after light exposure, respectively (Fig. 4A). No induction was observed in the cortex, pineal gland, and retina (Fig. 4A). Three additional bRfx4 primer sets were utilized, and similar results were obtained (data not shown).
Next, to test whether bRfx4 induction is subjective nightspecific, its expression was measured in the SCN at CT6 (subjective day), CT14 (early subjective night), and CT22 (late subjective night) after light exposure. Whereas significant induction was detected at CT14 and CT22, little induction was observed at CT6 (Fig. 4B). No induction was detected in the cortex at any of the three CT points (Fig. 4B).
To determine whether the expression of the gene is rhythmic, gene expression was measured over a period of 1 day at 4-h intervals, CT2, -6, -10, -16, and -22. We did not detect any significant rhythmic expression in the SCN (Fig. 4C). mPer2 expression was also measured, and clear rhythmic expression was detected, confirming the assay reliability (Fig. 4C).
In situ hybridization was performed, and a clear induction of gene expression was detected in the SCN 45 min after light exposure at CT22 (Fig. 5, A and B). To confirm this, three different probes prepared from bRfx4 cDNA were utilized, each of which obtained similar results. Results using the bRfx4specific probe are shown. Induction of gene expression was detected in all SCN sections: rostral, central, and caudal (Fig.  5A). The sum of the intensities from all six SCN sections covering the entire SCN region is shown in Fig. 5B.
It has been suggested that circadian clock entrainment is regulated at the gene expression level (22). Some light-inducible genes have been isolated in the SCN (32)(33)(34)(35)(36)(37)(38)(39)(40), and tremendous efforts have gone into clarifying the role of the products of such light-inducible genes in the entrainment mechanism, but multiple signal pathways in the mechanism make this difficult using the gene-disrupting approach (44 -46). Here we report on a novel form of mouse RFX4 that is different from the known RFX4 transcript expressed in the testis. The expression of novel RFX4 is clearly localized in the SCN, and its induction occurs only during subjective night.
Several light-induced genes have been identified in the SCN, and a significant number of them have the cyclic AMP-response element sequence, which functions as a cis-controlling element for light induction, in the upstream region of their initiation codon (47,48). Bioinformatic analysis revealed the presence of the cyclic AMP-response element sequence at Ϫ6,807 and Ϫ11,500 of the mouse RFX4 gene. Both sequences are conserved between mice and humans (Fig. 3A). This fact is consistent with the bRfx4 induction in response to light exposure. We also searched for the X-box sequence (GTNRCCNNR-GYAAC) (5, 11) in a 10-kbp region upstream of 19 mouse circadian rhythm genes: per1, per2, per3, bmal1, clock, cry1,2, dbp, dec1,2, e4bp4, rev-erb-␣, tef, tim, vasopressin, prokineticin 2, c-fos, fosB, and nr4a2 to find the target gene of RFX4 and identified the X-box consensus sequence in per1 and per3 genomic regions. However, no X-box exists in human PER1 and PER3 genes, suggesting that these genes are not RFX4 targets.
To obtain evidence for the presence of gene products in the SCN, antibodies for mouse bRfx4 were generated using a synthetic peptide. On the other hand, full-length bRfx4 cDNA was cloned, and a FLAG-tagged bRfx4 gene was prepared and introduced into an expression vector. COS7 transfectants with these constructs were used for antibody verification. The anti-bRfx4 antibody detected a band at ϳ80 kDa, which has the same estimated molecular weight in both COS7 transfectants with bRfx4 or FLAG-bRfx4. The band detected in the FLAGtagged bRfx4 transfectant was also detected by the anti-FLAG monoclonal antibody. Furthermore, the products immunopre- cipitated by the anti-FLAG antibody were detected by the antibody to RFX4 and vice versa (data not shown). These results demonstrated an availability of antibodies and also indicated that the identified bRfx4 encodes an 80-kDa polypeptide in COS7 transfected cells. Interestingly, the band was comprised of four bands with small molecular mass differences, all of which were also detected by the anti-FLAG monoclonal antibody and polyclonal antibodies (Fig. 6A). Different phosphoryl-  5. Induction of the bRfx4 transcript observed by in situ hybridization. The arrows indicate the SCN. The results of in situ hybridization with the whole brain section at CT22 with the bRfx4 transcript probe are shown. A, in situ hybridization with the bRfx4-specific probe with rostral, central, and caudal sections. B, sum of the intensities obtained from all six sections prepared from the brain covering the entire SCN region. The bRfx4 mRNA level in the SCN before light exposure is defined as 100. C, the probe for in situ hybridization, the primers for the real-time PCR, and the locus for the synthetic peptides. ation patterns might explain the multiple bands. Using the antibodies, the SCN and testis were examined. A signal was detected at ϳ80 kDa in the SCN but not in the testis (Fig. 6A). In addition, we examined localization of bRFX4 using the COS7 cells, in which the FLAG-bRfx4 construct was introduced, and detected signals only in the nucleus (Fig. 6B). The results reported in this study suggested that the product encoded by the novel RFX4 transcript plays a role in the entrainment mechanism.