GATA Factor-dependent Regulation of Cardiac m2 Muscarinic Acetylcholine Gene Transcription*

The m2 subtype is the predominant muscarinic acetylcholine receptor subtype expressed in heart and regulates the rate and force of cardiac contraction. We have previously reported the isolation of the promoter region for the chick m2 receptor gene and defined a region of the chick m2 promoter sufficient for high level expression in cardiac primary cultures. In this manuscript we demonstrate transactivation of cm2 promoter by the GATA family of transcription factors. In addition, we define the GATA-responsive element in the chick m2 promoter and demonstrate that this element is required for expression in cardiac primary cultures. Finally, we demonstrate specific binding of both a chick heart nuclear protein and of cloned chick GATA-4, -5, and -6 to the GATA-responsive element.

The muscarinic acetylcholine receptors (mAChR) 1 belong to the seven-transmembrane domain superfamily of receptors whose biological actions are elicited via activation of GTPbinding regulatory proteins (G-proteins) (1). Five different mammalian mAChR subtypes (m1-m5) and four chicken mAChR subtypes (m2-m5) have been identified and shown to be encoded by separate genes (2,3). The m1, m3, and m5 subtypes preferentially couple to the stimulation of phospholipase C, whereas the m2 and m4 subtypes preferentially couple to the inhibition of adenylyl cyclase. The different mAChR subtypes have unique overlapping expression patterns and are found in the central nervous system, peripheral nervous system, smooth muscle, and heart (4). The m2 subtype is the main mAChR found in mammalian heart (5); however, chick heart expresses predominantly m2 with significant amounts of m4 and m3 (6). Cardiac mAChR activation causes a decrease in both the rate and force of contraction (7). These effects are mediated through the inhibition of adenylyl cyclase activity, activation of an inward-rectifying potassium channel, inhibition of a calcium channel, and inhibition of the hyperpolarization-activated pacemaker current.
Since the initial discovery of multiple muscarinic acetylcholine receptor subtypes, great effort has been made to under-stand the expression patterns and, ultimately, the function of the different subtypes. Although many advances have been made in determining the expression patterns of the different muscarinic receptor subtypes, little is known about the molecular mechanisms that determine these patterns. The isolation of promoter regions for the m1, m2, and m4 receptor genes has been recently reported (8 -11). The mechanisms governing neural-specific expression of muscarinic receptors are beginning to be understood. For example, the m4 receptor contains a repressor element 1/neuron restrictive silencer element that plays a role in determination of neural-specific expression (10). In addition, we have demonstrated activation of the m2 promoter by the cytokines ciliary neurotrophic factor and leukemia inhibitory factor in mammalian neural cell lines. The mechanisms that regulate expression of the m2 receptor in heart are for the most part unknown. Previously, we identified regions of the chick m2 promoter necessary for basal level expression in chick heart primary cultures (9). We sought to determine if the GATA family of transcription factors plays a role in expression of the m2 receptor gene in heart.
The founding member of the GATA family of transcription factors (GATA-1) was first identified as a determinant of lineage-specific gene expression during hematopoiesis that bound to the DNA sequence (A/T)GATA(A/G) found in the promoter region of several erythroid-specific genes (12). In vertebrates, six GATA family members that contain DNA binding domains with two zinc fingers of the CX 2 CX 17 CX 2 C type (13) show distinct but overlapping expression patterns (for review, see Ref. 14). The GATA-1/2/3 subfamily are primarily involved with hematopoiesis, whereas members of the GATA-4/5/6 subfamily from Xenopus, mouse, and chick (15)(16)(17)(18)(19) play a role in expression of lineage-specific genes in a variety of tissues derived from mesoderm, including heart. Two transcripts for chick GATA-5, designated GATA-5 short and long, are produced by the use of alternative exons. The short GATA-5 isoform contains a single zinc finger and binds DNA but has decreased ability to transactivate GATA-responsive promoters (20). Several cardiac-specific genes contain functional GATA sites within their promoter regions and can be transactivated by GATA factors in noncardiac cells (21)(22)(23)(24).
In this manuscript we demonstrate 1) transactivation of m2 reporter constructs by the GATA family of transcription factors, 2) determination of the GATA-responsive element in the m2 promoter, 3) binding of a chick heart nuclear protein to the GATA-responsive element, and 4) regulation of basal level expression of m2 in chick heart primary cultures by the GATAresponsive element.

EXPERIMENTAL PROCEDURES
Cell Culture and Cell Transfection-Primary chicken heart cultures were prepared from 9-day embryos as described and grown in either defined media or medium M199 supplemented with 5% fetal bovine serum and 1% penicillin, streptomycin at 37°C in a humidified 5% CO 2 environment (25). For transfections, cells were plated out into 24-well plates and transfected 60 h later with either Transfectam reagent * This work was supported by National Institutes of Health Grants HL30639 (to N. M. N.) and NS7332 (to M. L. R.) and an American Heart Association of Washington postdoctoral fellowship (to M. L. R.) 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.
(Promega), for cells grown in defined media, or by the calcium phosphate method as described (9,26). For Transfectam transfections, the transfection mixture contained 100 ng/well pSV-␤-galactosidase (Promega) to correct for minor differences in transfection efficiency and 0.037 pmol/well reporter construct (117-134 ng) brought to a final total DNA concentration of 234 ng/well with pGEM4 (Promega) carrier DNA. The calcium phosphate transfection mixture contained 100 ng/well pSV-␤-galactosidase (Promega) and 0.048 pmol (153-183 ng) of reporter construct brought to a final amount of 283 ng with carrier DNA. Cells were lysed 32 h after the addition of transfection mixture. The human choriocarcinoma cell line JEG-3 (American Type Culture Collection, Rockville, MD) was grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum and 1% penicillin, streptomycin at 37°in a humidified 10% CO 2 environment. Cells were plated at a density of 20,000/well into 24-well plates, transfected 48 h after plating, and harvested 48 h later as described (27). The transfection mixture contained 177 ng/well reporter construct and varying amounts of GATA factor expression plasmids. A plasmid containing the lacZ gene driven by the Rous sarcoma virus promoter (RSV-␤-galactosidase) was included to correct for minor differences in transfection efficiencies (28). The total DNA concentration was kept constant by the addition of empty expression vector. COS-7 cells (American Type Culture Collection, Rockville, MD) were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum and 1% penicillin, streptomycin at 37°in a humidified 10% CO 2 environment. Transfections were performed by the calcium phosphate method. Briefly, calcium phosphate precipitates containing 17 g of either GATA factor expression constructs or empty expression vector were added to 10-cm dishes grown to approximately 70 -90% confluency.
Determination of Luciferase and ␤-Galactosidase Activity-Luciferase activities were determined using a Berthold AutoLumat LB 953 as described (27). ␤-Galactosidase activity was determined using the onitrophenyl ␤-D-galactopyranoside colorimetric assay for cells transfected by the calcium phosphate method and the Galactolight chemiluminescent assay (Tropix) for cells transfected with Transfectam (27). The luciferase activity was normalized to ␤-galactosidase activity and in some cases expressed as -fold above the luciferase/␤-galactosidase activity of promoterless luciferase vector (pGL3 Basic).
Preparation of Extracts and Electrophoretic Mobility Shift Assays-Heart cell nuclear extracts were isolated from 9-day embryonic chick heart primary cultures grown in defined media 96 h after plating as described (29). Whole cell extracts were prepared from COS-7 cells 36 h after the addition of transfection mixture as described (30). Electrophoretic mobility shift assays (EMSA) were performed as described (30). Briefly, 10 fmol of double-stranded oligonucleotides labeled with [␥-32 P]ATP using T4 polynucleotide kinase were used as the probe. Experiments were performed with either 1.5-2.5 g of nuclear extracts isolated from 9-day embryonic chick heart primary cultures or 2.2 l of whole cell mini-extracts isolated from COS-7 cells. Binding reaction components were added in the following order: 1) buffer, salts, and nonspecific DNA; 2) protein extracts; 3) cold competitor DNA; and 4) probe. Antibody, either 5 g of anti-HA or 100 ng of anti-myc, was added 10 min after the addition of probe for supershift experiments. The binding reaction was incubated on ice for 45 min and separated on a 5% polyacrylamide 0.25ϫ Tris borate-EDTA gel prerun for 2 h at 4°C. The dried gel was exposed to Kodak X-Omat film at Ϫ70°C with two Dupont lightning intensifying screens for the indicated time. Doublestranded oligonucleotides used were: 3Ј-GATA, 5Ј-GATCATGGAAGA- DNA Constructs-The constructs pNMR26-29 containing 789 bp, 2 kbp, 3.3 kbp, and 8 kbp fragments, respectively, of the m2 promoter cloned into the pGL3 Basic (Promega) luciferase reporter gene vector have been previously described (9). The m2 789-bp reporter construct was used as a template for polymerase chain reaction overlap sitedirected mutagenesis (31) using Pfu polymerase (Stratagene). All polymerase chain reaction constructs were sequenced with the Applied Biosystems Taq Dye Terminator Sequence kit to ensure the GATA to GGTA change and the absence of secondary mutations. The plasmid pAP/GATA was constructed by placing a synthetic GATA, 5Ј-TGCG-GATAAGATAAGGCCGGAATT-3Ј, site upstream of the Ϫ44 to ϩ71 region of the human liver/bone/kidney alkaline phosphatase gene (32). The resulting synthetic promoter was cloned into pGL3 Basic. The PYP-GATA-6 expression plasmid contains the GATA-6 coding region from the sequence PVYVP (amino acid 42) to the stop codon in the expression vector pR18 (33). Expression constructs containing the coding region of the two chicken GATA-5 isoforms (GATA-5S and GATA-5L) inserted into pcDNA3 (Invitrogen) have been reported previously (20). The PYP-GATA-4 expression construct contains the largest cDNA isolated for GATA-4 inserted into pR18 and begins at the conserved PVYVP found in GATA-4-6 (18). The HA-GATA-5 and HA-GATA-6 contain the HA epitope (34) fused in-frame to the initiating methionine of the entire GATA-5(long isoform) and GATA-6 inserted into pcDNA3. The MYC-GATA-5 and MYC-GATA-6 constructs contain five copies of the myc epitope (35) from pCSTϩMT (kindly donated by D. Turner, R. Rupp, and H. Weintraub) fused in-frame to the initiator methionine of GATA-5(long) and GATA-6, respectively, inserted into pcDNA3. Expression constructs for Xenopus GATA-4, -5, and -6 (16) contained the entire coding region inserted into pcDNA3 (Invitrogen).
Western Blot Analysis-Ten microliters of the whole cell extracts used for EMSA were separated on a 10% SDS-polyacrylamide gel and transferred to nitrocellulose using standard conditions (36). Duplicate blots were probed with either anti-HA (5 g/ml final) or anti-myc (100 ng/ml final) monoclonal Ab. After incubation with horseradish peroxidase-conjugated secondary Ab, the blots were developed with the Renaissance chemiluminescent reagent (NEN Life Science Products) and exposed to Kodak X-Omat x-ray film.
Materials-[␥-32 P]ATP was purchased from NEN. All restriction enzymes and T4 polynucleotide kinase were purchased from New England Biolabs. Pfu polymerase was purchased from Stratagene. The anti-HA monoclonal Ab (12CA5) was purchased from Boehringer Mannheim, and the anti-c-myc monoclonal Ab (9E10) was purchased from Santa Cruz Biotechnology Inc. Horseradish peroxidase-conjugated sheep antimouse and donkey anti-rabbit antibodies were purchased from Amersham Pharmacia Biotech.

Cotransfection of Chick GATA-6 Can Transactivate the m2
Promoter via an Element Located within a 789-bp Region-We utilized a heterologous system consisting of the human choriocarcinoma cell line JEG-3, which express very low levels of mAChR, to determine whether chick GATA factors were capable of transactivating m2 reporter gene constructs and what promoter region may be responsible for activation. Preliminary experiments 2 demonstrated that GATA-6 transactivated the m2 promoter at levels higher than GATA-4 or -5. Cotransfection of equimolar amounts of vector alone (pGL3) or either the 789 bp, 2 kbp, 3.3 kbp, or 8 kbp m2 promoter reporter gene constructs along with PYP-GATA-6 and an RSV-␤-galactosidase plasmid, to correct for minor differences in transfection efficiency, resulted in an increase in m2 promoter-driven luciferase expression (45-63-fold above pGL3 alone) compared with basal levels (2-7-fold above pGL3 alone) (Fig. 1). A positive control plasmid pAP/GATA containing a minimal promoter region from the human alkaline phosphatase gene downstream of a synthetic consensus GATA binding site (see "Experimental Procedures") showed an increase from 3-to 26-fold above pGL3 alone when PYP-GATA-6 was cotransfected. Although the various m2 luciferase constructs displayed different basal levels of expression (37), cotransfection of PYP-GATA-6 resulted in similar m2-driven luciferase expression, indicating that the GATAresponsive element is located within the 789-bp promoter fragment.
The m2 Promoter Contains Three Consensus GATA Factor Binding Sites Located within 789 bp from Intron One That Are Sufficient for Transactivation by GATA-4, -5, and -6 in a Heterologous System-Sequence analysis identified three consensus GATA factor binding sites (37,38) located at positions Ϫ150, Ϫ500, and Ϫ580 bp in the m2 promoter ( Fig. 2A). The site present at Ϫ150 bp, designated the 3Ј site, is located 208 bp downstream of the most 3Ј start site of transcription. The two other sites, located at 142 and 222 bp upstream of the most 3Ј start site of transcription, were designated the MID and 5Ј sites, respectively. We wanted to determine if the chick GATA-4, -5, and -6 were capable of transactivating the m2 promoter in our heterologous system. Increasing amounts of the expression constructs PYP-GATA-4, GATA-5L, GATA-5S, and PYP-GATA-6 (see "Experimental Procedures") were cotransfected with the 789-bp m2 reporter gene construct into JEG-3 cells. An RSV-␤-galactosidase plasmid was included to correct for minor differences in transfection efficiencies. The PYP-GATA-4, GATA-5L, and PYP-GATA-6 were capable of transactivating the m2 789-bp promoter, albeit at different levels, whereas the GATA-5S did not transactivate the m2 789-bp promoter (Fig. 2B). The levels of transactivation for PYP-GATA-6 and PYP-GATA-4 peak at 75 ng of expression plasmid/well and decrease with higher amounts of expression plasmid, whereas the GATA-5L shows the highest transactivation when 150 ng/well expression plasmid is cotransfected. In this system, PYP-GATA-6 transactivates the m2 promoter the greatest (43 Ϯ 4-fold increase) with PYP-GATA-4 and GATA-5L transactivating less (18 Ϯ 2 and 6 Ϯ 2). Neither the Xenopus GATA-4, -5, nor -6 were capable of transactivating the m2 promoter in this system. 2 The 3Ј Consensus GATA Factor Binding Site Is Required for Basal Level Expression of m2 in Chick Heart Primary Cultures-Site-directed mutagenesis was used to change the three consensus GATA sites to GGTA in the 789-bp m2 reporter gene construct. The wild type and mutant constructs were transfected into 9-day embryonic primary chick heart cultures grown in either defined media (Fig. 3A) or M199 supplemented with 5% FBS (Fig. 3B). Cultures were transfected using Transfectam reagent (Fig. 3A) or by the calcium phosphate method (Fig. 3B). The two culture and transfection methods were used to ensure that our results were not biased by selective transfection of a subpopulation of the multiple cell types found in chick heart primary cultures. Changing a single adenine to 2 M. Rosoff and N. Nathanson, unpublished observation.

FIG. 2. Schematic of the chick m2 receptor proximal promoter region and determination of the relative level of transactivation for chick GATA-4, -5, and -6 in JEG-3 cells.
A, designation and location of GATA sites with respect to the beginning of intron 1 (0 bp). DNA strand containing the GATA sequence is indicated by arrow above. The most 3Ј start site of transcription located at Ϫ358 bp is indicated by a vertical line with arrow. B, JEG-3 cells were cotransfected with the 789-bp m2 promoter reporter construct and various amounts of expression constructs containing GATA-4, -5, or -6. f, PYP-cGATA4; ࡗ, cGATA5s; q, cGATA5L; OE, PYP-cGATA6. A RSV-␤galactosidase plasmid was included in the transfection mixture to correct for minor differences in transfection efficiencies. The results are expressed as the ratio of luciferase/␤-galactosidase activity.

FIG. 3. Effects of mutating the GATA consensus sites on expression of chick m2 reporter gene constructs in primary heart cell cultures.
Site-directed mutagenesis was used to change the three consensus GATA sites to GGTA. The wild type (wt) and mutant m2 reporter gene constructs were transfected into 9-day embryonic primary chick heart cultures grown in either defined media (A) or M199 supplemented with 5% fetal bovine serum (B). Cultures were transfected using Transfectam reagent (A) or by the calcium phosphate method (B). Results are expressed as -fold over vector (pGL3) alone. The plasmid pSV-lacZ was included in the transfection mixture to correct for minor differences in transfection efficiencies. Results are the average Ϯ S.E. from three to six separate experiments performed with six replicate cultures. guanine nucleotide in the 3Ј-GATA site of the 789-bp promoter region reduced the basal level expression to 25 Ϯ 6 and 22 Ϯ 1% that of the wild type construct for cultures grown in defined media and M199, 5% fetal bovine serum, respectively (Fig. 3, A  and B). Mutating the other consensus GATA sites resulted in little or no change in the basal level expression for cultures grown in defined media. However, mutating the MID-GATA site resulted in decreasing the basal activity to 67 Ϯ 2% that of wild type for cultures supplemented with fetal bovine serum. Any combination of the 3Ј-GGTA with either the 5Ј-GGTA or MID-GGTA resulted in a similar decrease found with constructs containing only the 3Ј-GGTA. Mutating all three GATA sites resulted in slightly greater decreases in basal activity (20 Ϯ 4 and 17 Ϯ 3% that of wild type) for cultures grown in defined media and M199, fetal bovine serum, respectively. Similar results were seen when the mutant constructs were cotransfected with PYP-cGATA6 into JEG-3 cells. 2 A Nuclear Factor Isolated from 9-Day Primary Chick Heart Cultures Grown in Defined Media Shows Specific Binding to the 3Ј-GATA Site in Vitro-Electrophoretic mobility shift assays were performed with nuclear extracts isolated from chick primary heart cultures grown in defined media to determine if the 3Ј-GATA site shown to be functionally important was a binding site for a nuclear protein. A 32 P end-labeled doublestranded oligonucleotide for the 3Ј-GATA site was incubated with nuclear extracts isolated from 9-day embryonic chick heart primary cultures grown in defined media. The products of the binding reaction were separated on a native polyacrylamide gel and subjected to autoradiography. Two different protein-DNA complexes were detected (Fig. 4, A-C). Competition experiments demonstrated that the binding of the slower mobility complex could be competed off, whereas the binding of the faster mobility complex was decreased but not eliminated by 100-fold molar excess of unlabeled 3Ј-GATA probe (Fig. 4A). These results suggest that at least one specific protein binds the 3Ј-GATA site and that the faster mobility band may be composed of either a protein that has higher affinity for binding or two proteins, one of which shows specific binding to this site. Unlabeled 5Ј and MID-GATA probe showed much less competition for binding by both proteins to the 3Ј-GATA site but had some effect when a 100-fold excess was used (Fig. 4, A and B). The mutant 3Ј-GGTA did not compete for binding of either proteins even in a 100-fold molar excess (Fig. 4C). We were unable to reproducibly detect any protein-DNA complexes when the MID and 5Ј-GATA sites were used as probes. 2 Although it is reasonable to postulate that one (or more) of the chicken GATA factors is the nuclear protein that binds to the 3Ј site, the lack of antisera against chick GATA-4, -5, or -6 prevents determination of whether the 3Ј-GATA protein-DNA complex contains either chick GATA-4, -5, or -6.
Chick GATA Factors Expressed in COS-7 Cells Bind Specifically to the 3Ј-GATA Site-Since there are no antibodies available for chick GATA -4, -5, or -6, we epitope-tagged the chick GATA-5 and -6 at the N terminus with either one copy of the HA epitope (39) or five copies of the myc epitope (35). The epitope-tagged GATA-5 and -6 transactivated the 789-bp m2 reporter construct in JEG-3 cells with similar abilities relative to the non-tagged versions (GATA-5L, 6.6 Ϯ 0.9-fold increase; HA-GATA-5, 10.3 Ϯ 1.0; myc-GATA-5, 40.4 Ϯ 4.6; PYP-GATA-6, 41.2 Ϯ 3.6; HA-GATA-6, 75.4 Ϯ 11.9; myc-GATA-6, 135.7 Ϯ 23.3; mean Ϯ S.E., n ϭ 3). We were unable to express functionally active N-terminal myc or HA-tagged PYP-GATA-4. 2 To determine if GATA-4, -5, and -6 were capable of binding to the 3Ј-GATA site in vitro, whole cell extracts isolated from COS-7 cells transiently transfected with expression constructs for chick PYP-GATA-4 and epitope-tagged GATA-5L and 6 were used for EMSA with the 3Ј-GATA site probe (Fig. 5). Experiments performed with extracts from cells transfected with PYP-GATA-4, myc-tagged GATA-5 or GATA-6 (myc-GATA-5, myc-GATA-6), and HA-tagged GATA-5 or -6 (HA-GATA-5, HA-GATA-6) show a single protein-DNA complex that is not present in binding reactions performed with extracts from cells transfected with empty expression vector or untransfected cells. The formation of this complex can be eliminated by including a 100-fold molar excess of unlabeled competitor in the binding reaction (Fig. 5). The addition of monoclonal antibodies to the HA or Myc epitope to the binding reactions produced a "supershift" of the protein-DNA complexes that was proportional to the number of epitopes present in the tagged GATA factors (1 epitope for HA versus 5 for myc). Competition experiments demonstrated that the specificity of chick GATA -4, -5, and -6 for the 3Ј-GATA site (Fig. 6) was similar to results observed with chick heart nuclear extracts (Fig. 4). Western blot analysis demonstrated that the myc-tagged GATA-6 was expressed at higher levels than the myc-tagged GATA-5, and the extent of transactivation of the m2 promoter by the myctagged constructs was proportional to the relative levels of protein expression (Fig. 7). The HA-tagged GATA-5 and GATA-6 are expressed at comparable levels and, in the experiment shown, HA-GATA-5 is expressed at slightly higher levels (Fig. 7). The ability of the HA-tagged GATA constructs to transactivate the m2 promoter do not appear to be dependent FIG. 4. Electrophoretic mobility shift assays using nuclear extracts isolated from 9-day embryonic chick heart cultures grown in defined media. A, competition experiment with 5Ј-GATA site; B, competition with MID GATA site; and C, competition with mutant 3Ј-GGTA site. Nuclear extracts were incubated with 32 P-labeled double-stranded oligonucleotides for the 3Ј-GATA site and electrophoresed on a 5% polyacrylamide 0.25ϫ Tris borate-EDTA gel. The dried gel was exposed to film for 16 h. 0, 0 competitor; P, probe alone; 5Ј GATA, 5Ј m2 GATA site; M GATA, middle m2 GATA site. DNA-protein complexes are shown by arrows. Similar results were seen in two additional experiments performed with extracts isolated from primary cultures prepared separately. on the relative protein levels detected by Western blot analysis (Fig. 7). DISCUSSION We have demonstrated that the chick GATA-4, -5, and -6 transcription factors are capable of transactivating the chick m2 promoter in a heterologous system. In this system, truncated versions of the chick GATA-4 and -6 were capable of transactivating the m2 promoter, in agreement with a recent report demonstrating that mouse GATA-4 containing an Nterminal truncation at the PVYVP motif conserved between mouse, human, chick, and Xenopus shows the same activity as the full-length protein when cotransfected into NIH-3T3 cells (40). In contrast to the GATA-5 long isoform, the GATA-5 short isoform was incapable of transactivating the m2 promoter. Similar results have been reported using a different promoter in COS-7 cells (20). Cotransfection of GATA-5L into chick heart primary cultures with the 789-bp m2 reporter plasmid resulted in an increase in m2-driven luciferase expression, whereas cotransfection of the GATA-5S had no effect. 2 Therefore, it is unlikely that the short GATA-5S isoform acts as a negative regulator of m2 transcription in heart.
Mutation of the 3Ј-GATA site in the m2 promoter drastically decreased the basal level activity in chick heart primary cultures, suggesting that the results from our heterologous system are relevant to cardiac regulation of m2 in vivo. Several reports implicate GATA-4, -5, and -6 involvement in early cardiac development (15-17, 19, 41-44). Targeted gene disruption of GATA-4 in mouse results in an embryonic lethal phenotype with cardiac morphogenic defects (45,46). In chick, the nega-tive chronotropic effect of muscarinic agonists can be detected starting at embryonic day 5 (47). Expression of m2 mRNA did not vary greatly from embryonic day 4 through day 20 but was expressed at levels 10-fold higher than either m3 and m4 (6). Transcripts for chick GATA-4, -5, and -6 were detected in Mini whole cell extracts were prepared from COS-7 cells transfected with the indicated expression constructs. 10-l aliquots were separated on 10% SDS-polyacrylamide gel electrophoresis and transferred to nitrocellulose. The nitrocellulose blot was incubated with either anti-HA (A) or anti-myc Ab (B) and was visualized by chemiluminescent reagent after incubation with horseradish peroxidase-conjugated secondary Ab. The blots were exposed for 30 min (A) or 20 min (B). 0, untransfected; V 1 , pR18 vector alone; V 2 , pCDNA3 alone; 1, PYP-GATA-4; 2, myc-tagged GATA-5; 3, myc-tagged GATA-6; 4, HA-tagged GATA-5; 5, HA-tagged GATA-6. Arrowhead in A indicates specific bands. kd, kilodalton. embryonic day 10 through post-hatched day 6 heart; however, expression at earlier stages has not been investigated (18). In mouse and Xenopus GATA-4, -5, -6, expression was present in mesoderm during gastrulation (15-17, 19, 44). Mutating the GATA sites in the m2 promoter does not completely eliminate basal level expression in chick heart primary cultures. Taken together, these results suggest that although GATA-4, -5, and -6, in combination with other transcription factors, may determine mAChR subtype-specific expression during cardiac differentiation, they are most likely required for but do not regulate chick m2 expression levels during development past embryonic day 4. Chick GATA-4, -5, -6 are incapable of transactivating a mouse m1 subtype mAChR promoter construct in JEG-3 cells, 2 consistent with a specific role for GATA-4, -5, and -6 in the regulation of cardiac mAChR expression.
The identity of the GATA factor(s) that are required for maximal m2 expression in heart remains to be elucidated. Our data suggest that all three members of the GATA-4, -5, and -6 can family bind to and transactivate the m2 promoter. The abilities of the myc-tagged GATA constructs to transactivate m2 transcription in our heterologous system are proportional to the relative levels of protein demonstrated by Western blot analysis and by the intensity of the protein-DNA complexes seen by EMSA. In contrast, the relative abilities of the HAtagged GATA constructs to transactivate m2 did not appear to be dependent on the amount of protein expressed as detected by Western blot analysis. These results suggest that GATA-6 may be better at transactivating transcription of m2 than GATA-5 when the two proteins are expressed at similar levels. This may reflect different binding affinities of GATA-5 or GATA-6 for the 3Ј-GATA element or differential interactions of GATA-5 and GATA-6 with other proteins required for transcription from the m2 promoter. We observed at least two protein complexes that can bind the 3Ј-GATA site in the m2 promoter when chick heart nuclear extracts were used for EMSA but only saw a single complex when GATA-4, -5, and -6 were overexpressed in COS-7 cells. The two complexes may contain different members of the GATA-4, -5, and -6 family or, alternatively, may contain a GATA factor and a cofactor. Recently, a novel factor, dubbed FOG, was found to associate with GATA-1 and synergistically activate transcription of a hematopoietic regulatory region (48). Therefore, the two protein-DNA complexes may contain different combinations of GATA factors and/or cofactors. Preparation of subtype-specific antisera for the chick GATA-4, -5, and -6 will be necessary to clarify which GATA factor(s) are involved with regulation of m2 receptor expression in heart.
The results presented here are the first identification of a transcription factor required for maximum basal level expression of a muscarinic acetylcholine receptor gene. Although we and others have identified promoter regions involved with the determination of neuronal expression of mAChR, the mechanisms determining cardiac-specific expression of mAChRs were previously unknown. Our results provide a framework for further studies to better understand the transcriptional regulation of the m2 gene in both heart and in neural tissue.