Promoter region of the rat m4 muscarinic acetylcholine receptor gene contains a cell type-specific silencer element.

We describe here the characterization of the rat m4 muscarinic acetylcholine receptor gene and the identification of its regulatory region. Two 5'-noncoding exons are located approximately 5 kilobases upstream from the coding exon, and at least two alternatively spliced variants of m4 mRNA are expressed in the neuronal cell line PC12D. There are two transcription initiation sites. The promoter region is GC-rich, contains no TATA-box, but has two potential CAAT boxes and several putative binding sites for transcription factors Sp1 and AP-2. We assessed the m4 promoter activity functionally in transient expression assays using luciferase as a reporter. The proximal 435-base pair (bp) sequence of the 5'-flanking region produced luciferase activity in both m4-expressing neuronal cell lines (PC12D and NG108-15) and non-neuronal cell lines (L6 and 3Y1B). A longer fragment containing an additional 638-bp sequence produced luciferase activity only in m4-expressing neuronal cell lines. These data suggest that the proximal 435-bp sequence contains a constitutive promoter and that a 638-bp sequence farther upstream contains a cell type-specific silencer element. A consensus sequence for the neural-restrictive silencer element is found within this 638-bp segment.

Muscarinic acetylcholine receptors (mAChR) 1 are the members of the superfamily of G-protein-coupled receptors, which are characterized by the presence of seven putative transmembrane domains. mAChRs are widely expressed in the central and the peripheral nervous system. In the brain, mAChRs are thought to play critical roles in higher functions, including attention, regulation of movement, learning, and memory (1,2). Five subtypes of mAChR (m1-m5) have been identified by molecular cloning (3). Each subtype of mAChRs shows a distinct and complicated distribution in peripheral tissues and brain, suggesting subtype-specific functions (4 -7). Although the coding region of each mAChR subtype has been cloned and sequenced, the precise structures of their 5Ј-noncoding regions and regulatory regions have not been reported. Cloning and analysis of the genetic regulatory elements of these genes should yield important insights into the mechanisms that underlie the tissue-and site-specific expression of each subtype of mAChR.
In this study, we focused on defining the sequences that regulate m4 mAChR gene expression. In mammals, the m4 mAChR gene is expressed predominantly in the central nervous system, although its expression has been detected in some peripheral tissues such as rabbit lung (but not human or rat lung) (8,9). In rat brain, m4 mRNA is present in the cerebral cortex, striatum, olfactory bulb, and pyramidal cell layer of the hippocampus (4,7,10,11). m4 mAChR has been suggested to function not only postsynaptically but also presynaptically in some regions (12,13). As a first step toward elucidating the regulation mechanism of m4 mAChR gene expression, we have characterized the rat m4 mAChR gene and identified its regulatory region.

EXPERIMENTAL PROCEDURES
Cell Culture-PC12D cells (a spontaneously arising derivative of the rat pheochromocytoma-derived cell line PC12 (14), a gift from Dr. M. Sano (Institute for Development Research, Aichi Prefectural Colony, Aichi, Japan)) were maintained in Dulbecco's modified Eagle's medium (Nissui) supplemented with 5% fetal bovine serum and 5% horse serum, L6 (a rat skeletal muscle myoblast-like cell line) and 3Y1B cells (a rat fibroblastoma-derived cell line) in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum, and NG108 -15 cells a hybrid cell line derived from mouse neuroblastoma N18 and rat glioma C6) in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum and HAT medium (10 mM hypoxanthine, 40 M aminopterin, 1.6 mM thymidine, Life Technologies, Inc.). All culture media contained 100 units/ml penicillin G and 100 g/ml streptomycin.
RNA Isolation and Amplification of the 5Ј-Noncoding Region-Total cellular RNA was isolated by the method of Chomczynski and Sacchi (15). Poly (A) ϩ RNA was isolated from total cellular RNA using Oligotex-dT30͗Super͘ (Takara). The 5Ј-noncoding region of m4 mRNA was amplified from 2 g of PC12D poly(A) ϩ RNA by single-strand ligation to single-stranded cDNA-PCR using the 5Ј-Ampli FINDER TM rapid amplification of cDNA ends kit (Clonetech). First strand cDNA was synthesized using an m4 mAChR gene-specific primer m4primer3 (5Ј-gcgaattcGATCATGAGACCTGCCATCTTAGTA-3Ј, complementary to sequence located 432-456 bp downstream from the translation initiation site; lowercase characters denote the linker sequence). An anchor oligonucleotide (supplied with the kit) was ligated to the cDNA, which was then amplified twice by PCR using a primer complementary to the anchor and one of two gene-specific primers: m4primer1 (5Ј-gcgaat-tCACCATCTCCACTGTCTCCAGGT-3Ј, 74 -96 bp downstream from the translation initiation site) or m4primer2 (5Ј-gcgaattcTTGTGGGCT-GCTGTGA-3Ј, 56 -71 bp downstream from the translation initiation site). PCR products were cloned in pBluescript SK(ϩ) and sequenced.
Screening of Genomic Library-A rat genomic library (cloned in bacteriophage vector, a gift from Dr. Richard Mains, Johns Hopkins University, School of Medicine) was screened using a 392-bp XhoI-SmaI fragment of rat m4 mAChR gene coding region (sequence located 787-1179 bp downstream from the translation initiation site) as a probe. Approximately 1 ϫ 10 6 plaques were screened, and three overlapping clones were isolated. One of these clones (JH1411) was subjected to restriction map analysis and was partially sequenced. S1 Nuclease Mapping-A SmaI-HinfI fragment from the m4 genomic clone, which turned out to correspond to residues Ϫ90 to ϩ94 (with the * This work was supported in part by grants-in-aid from the Ministry of Education, Science, and Culture of Japan. 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 GenBank TM /EMBL Data Bank with accession number(s) D78484, D78485.
‡ To whom correspondence should be addressed. Tel.: 81-3-5689-7331; Fax: 81-3-3814-8154; E-mail: mieda@m.u-tokyo.ac.jp. 1 The abbreviations used are: mAChR, muscarinic acetylcholine receptor; bp, base pair(s); kb, kilobase(s); PCR, polymerase chain reaction; NRSE, neural restrictive silencer element; CAT, chloramphenicol acetyltransferase. transcription initiation site designated as ϩ1), was partially filled in at the HinfI site with dA and subcloned in the XbaI and EcoRV sites of pT7Blue vector (Novagen) following partial fill in of the XbaI site with dC and dT. Single-stranded DNA prepared from this construct was used as a template for preparation of single-stranded probe by the "prime cut" method (16). To anneal the primer to the template DNA, 0.2 pmol of T7 primer, 0.2 pmol of the single-stranded template, and 1.5 l of 10 ϫ Klenow buffer (supplied by New England Biolabs) in a total volume of 7.5 l were incubated at 65°C for 5 min. After cooling slowly to room temperature, 0.5 l of 3 mM (each) dGTP, dATP, and dTTP, 5 l of [␣-32 P]dCTP (3000 mCi/mmol) (Amersham), and 2 l of Klenow fragment (1 unit/l, New England Biolabs) were added and incubated at 25°C for 1 h. After addition of 0.75 l of 1 mM (each) dATP, dGTP, dTTP, and dCTP and further incubation for 30 min, the enzyme was heat inactivated at 70°C for 5 min. After digestion with BamHI, which cuts the multicloning site of the vector, the labeled product was extracted with phenol/chloroform and then ethanol precipitated. After denaturation at 95°C for 5 min in 90% formamide, 5 mM EDTA, 0.05% bromphenol blue, and xylene cyanol, the probe was purified by denaturing polyacrylamide gel electrophoresis in 8 M urea. S1 nuclease mapping was performed essentially as described previously (17). The probe was hybridized to 2 g of poly(A) ϩ RNA from PC12D cells or rat kidney at 52°C overnight. 1000 units/ml S1 nuclease (Boehringer Mannheim) was used for digestion at 37°C for 30 min. The S1-resistant products were analyzed using 6% denaturing polyacrylamide gels containing 8 M urea.
Primer Extension-The primer was prepared as follows. A HindIII-ApaI fragment (Ϫ333 to ϩ337) was subcloned in the HindIII and ApaI sites of pBluescript SK(ϩ). Using this construct as a template, the genomic fragment (Ϫ333 to ϩ157) was amplified by PCR using 41-1 primer (5Ј-gcgaattCTGCTTCTGTCTTTCTCTGCT-3Ј, complementary to residues ϩ137 to ϩ157) and the T3 primer. After digestion with SacI and EcoRI, the amplified fragment was cloned in the SacI and EcoRI sites of pBluescript SK(ϩ), and single-stranded DNA was prepared. The antisense strand was synthesized and labeled as described above using a T7 primer. The labeled product was digested with HinfI and EcoRI, and the double-stranded DNA, which corresponds to residues ϩ94 to ϩ157, was purified by nondenaturing polyacrylamide gel electrophoresis. Finally, the labeled antisense strand was isolated by denaturing in 50% formamide, 2.5 mM EDTA and resolving on a nondenaturing polyacrylamide gel. The resultant 64-bp single-stranded DNA was used as a primer. The primer was hybridized to 4 g of poly(A) ϩ RNA from PC12D cells or rat kidney at 34°C overnight as described above. After ethanol precipitation, samples were resuspended and reverse-transcribed with SuperscriptII (Life Technologies, Inc.) as described previously (18). Samples were analyzed using 6% denaturing polyacrylamide gels containing 8 M urea.
DNA Construction and Transient Expression Assays-A 6.3-kb SalI fragment from the 5Ј-end of the m4 genomic clone JH1411 (one of the two SalI sites is derived from the multicloning site of the bacteriophage vector) was subcloned in the SalI site in the pBluescript SK(ϩ). This construct was digested with SacI, and the resultant 3-kb fragment (one of the two SacI sites is derived from the multicloning site of pBluescript SK(ϩ)), which contains the downstream half of the 6.3-kb SalI fragment, was subcloned in the SacI site of the luciferase reporter vector, pGL2-Basic (Promega). This construct was digested with XhoI, which cuts within the multicloning site of the pGL2-Basic, and partially digested with PstI, then blunted and self-ligated to obtain constructs C and D, which contain approximately 1.7-and 1.3-kb SacI-PstI fragments, respectively. To obtain constructs A and B, which contain 435and 1073-bp PstI fragments, respectively, the appropriate KpnI-PstI fragment was removed from construct C using a KpnI site in the vector. pEF-CAT, which contains the bacterial chloramphenicol acetyltransferase (CAT) under the control of the human elongation factor 1␣ gene promoter, was constructed as follows. pBLCAT2 (purchased from ATCC) was digested with BamHI and BglII and then blunted and self-ligated. Digesting this construct with SalI and SmaI, a 1.5-kb fragment containing the CAT gene and polyadenylation signal from SV40 was isolated and cloned in the SalI and SmaI sites of pBluescript SK(ϩ). This fragment was isolated again from pBluescript SK(ϩ) by digesting with XbaI and cloned in the XbaI site of mammalian expression vector pEF-BOS (19).
Transfection was performed using Lipofectamine TM reagent (Life Technologies, Inc.), essentially as recommended in the manual. Cells were seeded in Corning 24-well microtiter plates (0.7 ϫ 10 5 cells per well for NG108 -15, 0.2 ϫ 10 5 cells per well for L6 and 3Y1B) or 12-well microtiter plates (1.6 ϫ 10 5 cells/well for PC12D) and were cultured for 1 day before transfection. 0.45 or 0.3 g of plasmid DNA and 3 or 0.7 l of Lipofectamine TM reagent were used per each well for PC12D cells or other cells, respectively. Cells were incubated with the DNA mixture for 5 h (NG108 -15, L6) or overnight (PC12D, 3Y1B). Cells were harvested 3 days after transfection. Enzyme assays were performed using the Promega Luciferase assay system. The luciferase activity was measured using a scintillation counter (Packard, Tri-Carb 1500) as described in the manual supplied by Promega. We measured the activity of pGL2-Basic vector itself, which contains no insert, to determine background activity. Transfection efficiency was determined by cotransfection of pEF-CAT construct. The CAT activity was measured as described by Nordeen et al. (20).

RESULTS
Structure of the Rat m4 mAChR Gene-To isolate the regulatory region of the m4 gene, we amplified the 5Ј-noncoding region of the m4 mRNA by single-strand ligation to singlestranded cDNA-PCR using poly(A) ϩ RNA isolated from PC12D cells, a subclone of the rat pheochromocytoma-derived cell line PC12. PC12D cells were selected for these studies because they express relatively high levels of m4 mAChR mRNA. 2 Two types of cDNA clones (cDNA 1 and 2, Fig. 1) were obtained with 5Ј sequences that diverged 32 nucleotides upstream from the translation initiation codon. This result suggested the presence of an intron in the 5Ј-noncoding region and that the two types of mRNA were generated by alternative splicing. To identify the upstream exons and the 5Ј-flanking region of the m4 gene, we isolated the rat m4 chromosomal gene. We obtained three overlapping clones from a rat genomic library. One of these clones, JH1411, was chosen for further analysis. The restriction map and partial sequence of the JH1411 clone are shown in Figs. 1 and 2, respectively. A consensus splice acceptor sequence was found immediately 5Ј to the position where the sequences of the cDNA diverged (the sequence "cctccag" with the underline in Fig. 2). Two upstream exons (exons 1 and 2) were mapped to positions approximately 5.3 and 4.4 kb upstream from the translation initiation codon by hybridization analysis (data not shown).
To test the possibility that PC12D cells express m4 mRNA that contains both of the two upstream exons, we carried out reverse transcription-PCR using primers specific to each exon. As shown in Fig. 1, we obtained a cDNA clone that contains sequences from both exons 1 and 2 (cDNA 3). The length of exon 2 was determined to be 158 bp, and a consensus sequence for a splice acceptor site was found next to the 5Ј-end of the exon (Fig. 2). This result suggests that cDNA 2 is an incompletely reverse-transcribed cDNA product derived from a mRNA variant that contains both exons 1 and 2 (mRNA 2). Exon 2 contains an in-frame upstream stop codon, and exon 1 does not contain an in-frame ATG triplet. These data indicate that each m4 mRNA variant encodes the same protein.
Transcription Initiation Sites and Analysis of the 5Ј-Flanking Region-The transcription initiation sites of the rat m4 gene were determined by primer extension analysis and S1 nuclease mapping. For primer extension analysis, we used a 64-bp single-stranded DNA as a primer (Fig. 3C). As shown in Fig. 3A, two bands (161 and 158 bp) were detected using poly(A) ϩ RNA from PC12D cells. This suggests that transcription of the m4 gene initiates primarily around 160 and 157 bp upstream from the 3Ј-end of exon 1. No signal could be detected using poly(A) ϩ RNA from kidney, a tissue that does not express the m4 gene. To confirm that the observed signals in primer extension analysis represent the 5Ј-end of m4 mRNA, we did S1 nuclease mapping. Two clusters of DNA fragments were observed (Fig. 3B). One cluster centered around 96 bp and the other around 93 bp. This result is consistent with the result of the primer extension analysis, indicating that these nucleo- m4 mAChR Gene Promoter tides correspond to the transcription initiation sites. A longer protected product was also observed, which corresponded to the entire region of the S1 probe excluding the sequences derived from the cloning vector. This suggests that there may be at least one more initiation site upstream. Although we tried to identify this upstream initiation site using different probes and primers, we have not yet been successful. We also performed RNase protection assays using several different probes and obtained results similar to those obtained by S1 nuclease mapping (data not shown).
The sequence of the putative promoter region of the rat m4 mAChR gene is rich in GC content (about 70% in the proximal 400 bp) and has no typical TATA-box (Fig. 2). There is a CAAT-box between bases Ϫ593 and Ϫ589 and an inverted CAAT-box ATTGG between bases Ϫ62 and Ϫ57. The putative promoter segment contains several consensus sequences for Sp1 (GGGCGG or CCCGCC) and AP-2 (CC(G/C)C(A/G)GGC) binding sites. In addition, there is a putative neural-restrictive silencer element (NRSE) in an inverted orientation between bases Ϫ857 and Ϫ837 (Fig. 4). Possible AP-1 and Zif268 binding sites are also found in the intron between exons 1 and 2.
Functional Analysis of the m4 Promoter-We assessed the m4 promoter activity functionally by transient expression assay using luciferase as a reporter (Fig. 5). We inserted different portions of the 5Ј-flanking sequence of the m4 gene in the pGL2-Basic vector (Promega) and transfected these into PC12D cells, NG108 -15 cells (a hybrid cell line derived from mouse neuroblastoma N18 and rat glioma C6), L6 cells (a rat skeletal muscle myoblast-like cell line), and 3Y1B cells (a rat fibroblastoma-derived cell line). We corrected for differences in transfection efficiency by cotransfecting with the pEF-CAT and normalizing the luciferase expression data to the measured CAT activity. Construct A containing 435 bp of sequence upstream from the transcription initiation site produced luciferase activity in all cell lines tested. Constructs containing additional upstream sequences (construct B and construct C) produced activities comparable to construct A in m4-expressing neuronal cell lines (PC12D and NG108 -15) but produced much less luciferase activities in non-neuronal cell lines (L6 and 3Y1B) that do not express the m4 gene. Construct D, which lacks the major transcription initiation sites and the proximal 5Ј-flanking sequence, produced little luciferase activity in each cell line tested. These results suggest that a constitutive promoter, which is not cell type-specific, is located in the proximal 435-bp sequence of the 5Ј-flanking region and that a silencer element(s) located upstream suppresses this promoter activity in cell lines where the m4 receptor is not expressed. This hypothesis is consistent with the existence of the NRSE between bases Ϫ857 and Ϫ837, which is contained in constructs B and C but not in construct A. DISCUSSION We have determined the structure of 5Ј-noncoding region of the rat m4 mAChR gene. Two 5Ј-noncoding exons are located approximately 5 kb upstream from the coding exon. PC12D cells express at least two alternatively spliced variants of mRNA (mRNAs 1 and 2 in Fig. 1, the latter is predicted based on the splicing patterns of cDNAs 2 and 3), which originate from these two upstream exons. Also, we have obtained cDNAs that correspond to cDNAs 1 and 2 ( Fig. 1) using RNA from rat brain by reverse transcription-PCR (data not shown). Previous studies have shown that genes encoding rat m1, m3, m5, and porcine m2 mAChR contain at least one intron in their 5Јnoncoding regions and that m4 mAChR gene also has a potential splice acceptor site in its 5Ј-noncoding region (21,22). There are no introns in the coding or the 3Ј-noncoding regions of these genes (3). The porcine m2 mAChR gene has at least two alternatively spliced 5Ј-noncoding exons, although its precise gene structure is not known (23). Another group has reported FIG. 3. Primer extension analysis and S1 nuclease mapping to determine the transcription initiation sites. Panel A, primer extension analysis. A uniformly labeled 64-bp single-stranded DNA primer corresponding to the antisense strand of the rat m4 mAChR gene was hybridized with 4 g of poly(A) ϩ RNA from rat kidney (lane 1) or PC12D cells (lane 2) and then extended with reverse-transcriptase (SuperscriptII, Life Technologies, Inc.). Lanes M, bacteriophage M13 mp18 sequence using the Ϫ40 primer. Panel B, S1 nuclease mapping. A uniformly labeled single-stranded probe corresponding to residues Ϫ90 to ϩ94 was hybridized with 2 g of poly(A) ϩ RNA from PC12D cells (lane 1) or rat kidney (lane 2) and then digested with S1 nuclease. lanes M, bacteriophage M13 mp18 sequence using the Ϫ40 primer. Panel C, strategy to map the transcription initiation sites. The bent arrows indicate the principal transcription initiation sites. Hf, HinfI.
FIG. 4. Alignment of NRSEs between rat m4 mAChR, rat SCG10 (35), rat type II sodium channel (NaII) (36), and human synapsin I (37) genes. For m4 AChR, the sense strand is written at the top and in the orientation 3Ј to 5Ј. The numbering of the bases is as described in Fig. 2. Asterisks show nucleotides that are not seen in other genes.
that two 5Ј-noncoding exons are located about 5.5 kb upstream from the coding exon in the rat m4 gene (24).
The 5Ј-flanking region of the rat m4 mAChR gene lacks a TATA-box, is GC-rich, and contains several potential Sp1 binding sites. These characteristics are typical of promoters for housekeeping genes, which are constitutively expressed. However, recent studies have demonstrated that some highly regulated genes, including immediate early genes, developmentally regulated genes, and tissue-specific genes, also have promoters that lack TATA boxes (25). Genes encoding several other G-protein-coupled receptors, including D 1 (26,27), D 2 (28), and D 5 (29) dopaminergic and ␣ 1B (30) and ␤ 1 (31,32) adrenergic receptors, also have GC-rich promoters that lack TATA boxes. These promoters also contain several putative Sp1 binding sites, although the requirement of Sp1 binding for transcription of these genes is yet to be determined. In some genes that have promoters lacking TATA boxes, binding of Sp1 has been shown to play a critical role in transcription initiation (33,34).
In mammals, the m4 mAChR gene is expressed primarily in neurons. We assessed whether or not the 5Ј-flanking region of the m4 mAChR gene functions as a cell type-specific promoter by transient expression assays using non-neuronal cell lines and m4 mAChR-expressing neuronal cell lines. The fragment containing 1073 bp of the 5Ј-flanking region (Fig. 5, construct  B) was found to be sufficient for cell type-specific expression of the m4 mAChR-luciferase fusion gene. Deletion from the 5Ј side to base Ϫ435 did not result in a loss of promoter activity but rather in a loss of specificity of expression, indicating that the region Ϫ1073 to Ϫ435 represses the expression of m4 mAChR gene in cell lines where the m4 receptor is not expressed. The existence of a putative NRSE between bases Ϫ857 and Ϫ837 is consistent with the repression of the m4 mAChRluciferase fusion gene expression in non-neuronal cell lines, although the involvement of the NRSE in the neuron-specific expression of the m4 mAChR gene remains to be proven. NRSE is a silencer element that is known to regulate neuron-specific expression of several genes including the rat SCG10 (35), rat type II sodium channel (36), and human synapsin I genes (37). However, m4 mAChR is not expressed in all neurons and shows a unique, site-specific expression pattern in the brain. If we assume that NRSE is involved only in determining the neuronspecific expression of the m4 mAChR gene, an additional mechanism may be required to restrict the expression to only a subset of neurons. It remains to be determined whether the 5Ј-flanking region we have isolated contains some elements regulating the specificity of expression among different kinds of neurons. The mouse serotonin 2 receptor (5-HT 2 ) gene is known to possess repressor domains that regulate its neuron-specific expression and an activator domain that allows gene expression in a glial cell line that expresses the 5-HT 2 receptor (38,39). The m4 mAChR is also known to be expressed in some glial cell lines (40), although there is no direct evidence for expression of the m4 mAChR in glia cells in vivo.
In summary, we have determined the structure of the rat m4 mAChR gene and identified its promoter region. It has two 5Ј-noncoding exons, from which at least two alternatively spliced variants of mRNA originate in PC12D cells. The fragment containing 1073 bp of the 5Ј-flanking region is sufficient for cell type-specific (at least neuron-specific) expression.