Enhancer Λ Ig Differentiation, Bind an Essential Site in the Mef2 Proteins

Mef2 Proteins, Required for Muscle Differentiation, Bind an Essential Site in the Ig λ Enhancer

This information is current as Ebenezer Satyaraj and Ursula Storb of September 24, 2021. J Immunol 1998; 161:4795-4802; ; http://www.jimmunol.org/content/161/9/4795 Downloaded from References This article cites 52 articles, 25 of which you can access for free at: http://www.jimmunol.org/content/161/9/4795.full#ref-list-1

Why The JI? Submit online. http://www.jimmunol.org/ • Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average by guest on September 24, 2021 Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 1998 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Mef2 Proteins, Required for Muscle Differentiation, Bind an Essential Site in the Ig ␭ Enhancer1

Ebenezer Satyaraj2* and Ursula Storb3†

The Ig ␭ light chain gene enhancer has two unique essential motifs, ␭A and ␭B. The transcription factors that bind the ␭B motif have been identified as Pu.1 and Pu.1-interacting partner (Pip). We report here that the ␭A site includes a binding site for the myocyte-specific enhancer factor 2 (Mef2) family of transcription factors. Mef2 proteins were first described in muscle cells and, in vertebrates, include four known members designated A to D. Using a ␭A electrophoretic-mobility shift assay (EMSA), in conjunction with a high affinity Mef2 binding site and anti-Mef2 Abs, we show that members of the Mef2 family are present in nuclear extracts of ␭-producing B cells and bind the ␭A site. Functional assays using the chloramphenicol acetyltransferase (CAT) reporter construct containing three copies of the ␭A motif demonstrate that the ␭A sequence can function as an enhancer in

conjunction with the thymidine kinase (TK) promoter and is regulated by Mef2 proteins. Extrapolating from other systems where Downloaded from transcriptional regulation by Mef2 has been studied, other transcription factors may be involved along with Mef2 in transcriptional regulation at the ␭A site. The Journal of Immunology, 1998, 161: 4795–4802.

mmunoglobulins are heteromeric molecules composed of while the other is bound by Pip (Pu.1-interacting partner) recruited light (L)4 and heavy (H) chains. There are two light chains: through specific interaction with Pu.1 (16). Analysis of the human ␭ ␬ ␭ ␭ ␭ I lambda ( ) and kappa ( ). The Ig H and L chains are encoded Ig enhancers has also identified similar A and B domains (17). http://www.jimmunol.org/ by multiple segments that must be somatically recombined to form Analysis of the ␭A site using a transcription factor database a functional gene (1–2). In B cells, expression of Ig genes is (Transcription Factor Search, 1995, Yutaka Akiyama, Kyoto Uni- strictly regulated for cell-type specificity and development stage versity; http://pdap1.trc.rwcp.or.jp/research/db/TFSEARCH.html) specificity. This is accomplished by multiple cis-regulatory ele- identified an AϩT rich consensus binding site for Mef2 (Fig. 1A). ments, promoters, and enhancers (2–4). Additionally, these tran- The Mef2 transcription factor was first identified as a protein scriptional elements have a significant role in V(D)J recombination that binds an AT-rich sequence in the muscle-specific enhancer (1, 5, 6). Enhancers have been identified in the J-C introns of both of the muscle creatine kinase (MCK) gene (18). Mef2 factors ␬ Ј H (7) and genes (8) and also 3 of the C exons of H (9, 10), and bind as homo- and heterodimers to the consensus sequence ␬ ␬ by guest on September 24, 2021 (11) genes. Unlike the H chain and light chain gene locus, the CTA(A/T) TA(G/A), which is found in the control regions of nu- ␭ light chain gene locus is organizationally distinct (12). In the 4 merous muscle-specific genes and growth factor-induced genes murine ␭ locus, two transcriptional enhancers, E␭2-4 and E␭3-1, (19). In the muscle cell, cooperative interaction between Mef2 fac- have been identified and lie 3Ј of the J␭-C␭ gene clusters (13). No tors and myogenic basic-helix-loop-helix (bHLH) factors has been intronic enhancers have been found in the ␭ gene locus. E␭2-4 and shown to regulate muscle-specific transcription (20). Although E␭3-1, which are Ͼ90% homologous, are thought to have evolved Mef2 binding sites are present in many muscle-specific promoters by gene duplication and are believed to function similarly. Two and are important for skeletal and cardiac muscle development (18, distinct domains, ␭A and ␭B, that are essential for function have been defined in the ␭2-4 enhancer (14, 15). Both ␭A and ␭B sites 21), several findings suggest that Mef2 proteins may play a role in bind B cell-specific factors in nuclear extracts (14, 15). The ␭B site nonmuscle gene expression. Four different Mef2 genes have been has two juxtaposed but distinct binding sites that are bound by a identified by molecular cloning experiments and are designated pair of interacting transcription factors (15). One of the composite Mef2A, Mef2B, Mef2C, and Mef2D (22–24). Mef2 proteins belong elements is bound by Pu.1, an Ets family transcription factor (15), to the MADS family of transcription factors (25) (named after the first four proteins in which the MADS domain was first identified: minichromosome maintenance 1 (MCM1), which regulates mat- *Department of Molecular Genetics and Cell Biology and †Committee on Immunol- ogy, University of Chicago, Chicago, IL 60637 ing-specific genes in yeast; AGAMOUS and DEFICIENS, which have homeotic function in flower development; and serum-re- Received for publication April 23, 1998. Accepted for publication June 29, 1998. sponse element (SRE), which regulates serum-inducible and mus- 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 cle gene expression) and share a 56-aa N-terminal MADS box with 18 U.S.C. Section 1734 solely to indicate this fact. domain followed by a 27-aa Mef2 domain, which extends C-ter- 1 This work was supported by National Institutes of Health Grant AI 39535. minal of the MADS domain (26). These protein domains together 2 Current address: Department of Medicine, Northwestern University, Chicago, IL mediate DNA-binding, homo- and heterodimerization, and inter- 60611. action with bHLH proteins (27). 3 Address correspondence and reprint requests to Dr. Ursula Storb, Department of The expression of Mef2 gene products marks early myogenic Molecular Genetics and Cell Biology, 920 East 58th Street, Chicago, IL 60615. E- mail address: [email protected] lineages during embryogenesis (28) and follows specific expres- 4 Abbreviations used in this paper: L, light; H, heavy; EMSA, electrophoretic mo- sion patterns in different tissues (29). After birth, Mef2A, Mef2B, bility shift assay; Mef2, myocyte specific enhancer factor 2; MCK, muscle creatine and Mef2D are expressed ubiquitously (22–24, 26, 30, 31), while kinase; TK, thymidine kinase; CAT, chloramphenicol acetyl transferase; bHLH, basic-helix-loop-helix; MADS, minichromosome maintenance 1 (MCM1) ϩ agamous Mef2C transcripts are restricted to skeletal muscle, brain, and ϩ deficiens ϩ serum-response element (SRE). spleen (26, 31). Mef2C null mice show cardiogenic defects and an

buffer on ice for 30 min before the addition of the remaining components of the binding reaction.

Western blotting For Western blotting, 20 to 40 ␮g of nuclear extract or 5 ␮l of in vitro translated proteins were boiled for 5 min with an equal volume of SDS sample buffer and chilled on ice and resolved on an 8% SDS-PAGE followed by blotting onto nitrocellulose membranes (Hybond ECL, Amer- sham, Buckinghamshire, U.K.). Rabbit anti-Mef2A (cross-reacts with Mef2C), anti-Mef2B, and anti-Mef2D (kind gift of Dr. Ron Prywes, Co- lumbia University, New York) were used as the ﬁrst Ab followed by goat anti-rabbit-horseradish peroxidase (HRP) (Amersham) conjugate. Immu- noblots were developed using a chemiluminescent ECL Western blotting kit (Amersham) and visualized by exposing to x-ray ﬁlm.

In vitro transcription and translation Mef2A and Mef2C cDNAs cloned into the CMV promoter-driven expression vector pcDNAI (Invitrogen, San Diego, CA) (a kind gift of Dr. E. N. FIGURE 1. The ␭A site of the E␭3-1 and E␭2-4 enhancers includes a Olson, Dallas, TX) were used. Mef2A and Mef2C proteins were made by Mef2 binding site. A, Sequences of E␭3-1 and E␭2-4 are shown with the T7-directed in vitro transcription-translation (TNT Kit, Promega, Madison, ␭ ␭ ␭ ␭ WI) in the presence of [35S]methionine and expression confirmed by footprinted sites, A and B motif, and the two E boxes, E1 and E2 (14). Downloaded from B, The Mef2 binding site included in the ␭A site of both E␭3-1 and E␭2-4 SDS-PAGE. are shown in boldface along with the consensus Mef2 binding site. Functional assays using a CAT reporter construct The A3-TKCAT construct was made as follows. A trimer of ␭A cloned in pBluescript II KS (C. Eisenbeis et al., unpublished observations) was amplified embryonic lethal phenotype (27). On the other hand, Mef2B null using the following PCR primers: (forward) 5Ј-CCCGCGGATCCCTAG TG Ј Ј mice are viable and do not show any major defects at birth (32). GATCTTCCACAAGC-3 and (reverse) 5 -GCCCGGGGGATCCCTATCA http://www.jimmunol.org/ In this paper we show that members of the Mef2 protein family CAG-3Ј. PCR primers included BamHI restriction sites to allow cloning into ␭ ␭ a BamHI site in front of the thymidine kinase (TK) promoter that drives a bind the A site of the enhancers of the Ig light chain gene. We bacterial chloramphenicol acetyltransferase (CAT) gene in the TKCAT con- report here the fine analysis and functional properties of this struct (37). J558L cells were transfected by the DEAE-dextran method as interaction. described by Eisenbeis et al. (15). Briefly, cells were grown to a density of 10 ϫ 105 cells per ml. A total of 107 cells were washed twice in TS buffer (137

mM NaCl, 5 mM KCl, 0.4 mM Na2HPO4, 25 mM Tris, 1 mM MgCl2, 0.7 mM Materials and Methods ␮ CaCl2 (pH 7.4)) and resuspended in TS buffer containing 20 g TKCAT Preparation of nuclear extracts reporter construct, 5 ␮gof␤-galactosidase reporter plasmid (pMC1924) (38), J558L myeloma cells and NIH 3T3 fibroblast cells were routinely main- and 0.25 mg of DEAE-dextran (Pharmacia) per ml. For coexpression studies, ␮ by guest on September 24, 2021 tained in DMEM (HyClone, Logan, UT) fortified with 10% FCS (Hy- 20 g of expression plasmids for Mef2 proteins (respective cDNA cloned in Clone) and penicillin G (Life Technologies, Grand Island, NY) and strep- pcDNAI (Invitrogen), a kind gift of Dr. E. N. Olson, Dallas, TX) were in- tomycin (Sigma, St. Louis, MO). A modified protocol based on the cluded, and in these experiments the total amount of DNA used for transfec- protocols of Schreiber et al. (33) and Dignam et al. (34), detailed in Eisen- tion was normalized by using nonrelated plasmid DNA as a “filler.” After 20 beis et al., (16) was used for preparation of nuclear extracts. The nuclear min at room temperature, 15 ml of DMEM containing 0.1 M chloroquine Ϫ diphosphate was added, and the cells were transferred to a 7.5% CO2 incubator extracts were aliquoted and quick-frozen in liquid N2 and stored at 70°C. Protein concentration was estimated by Bradford (35) assay using a kit at 37°C for 1 h. The transfected cells were then pelleted and resuspended in 40 ml of tissue culture medium and incubated in 7.5% CO . After 60 h, cell from Bio-Rad (Hercules, CA). 2 lysates were made from washed cells as described by Gorman et al. (39). A EMSA sample of the lysate was removed to determine the ␤-galactosidase activity (40), and the rest was heated to 60°C for 10 min, chilled on ice for 5 min, and EMSAs were performed as described by Singh et al. (36). Probes were then centrifuged at 15,000 ϫ g for 10 min. CAT assays were performed as made from complementary oligonucleotide pairs with XbaI and BamHI described previously (39). The volume of lysate used in each assay was nor- overhangs, and their sequences are as follows (complementary pairs are malized for ␤-galactosidase activity to control for variation in transfection denoted as “top” and “bottom”). ␭A (top), 5Ј-GATCTTCCA CAAGCTA efficiencies. The TLC plates were analyzed and quantitated using a Phospho- AAATTAGATCTGTGATAGG-3Ј; ␭A (bottom), 5Ј-GATCCCTATCA Imager (Molecular Dynamics, Sunnyvale, CA). CAT activity was calculated CAGA TCTAATTTTAGCTTGTGGAA-3Ј; ␭B (top), 5Ј-GATCTGAAAA in terms of the percent of acetylated chloramphenicol over the total AGAGAAATAAAAGGA AGTGAAACCAAGG-3Ј; ␭B (bottom), 5Ј- chloramphenicol. GATCCCTTGGTTTCACTTCCTTTTATTTCTC TTTTTCA-3Ј; muscle creatine kinase (MCK) (top), 5Ј-CTAGACTCGCTCTAAAAATAACC Ј Ј Results CTGTC-3 ; MCK (bottom), 5 -CTAGAGACAGGGTTATTTTTAGAGC ␭ GAG-3Ј; MEFmt1 (top), 5Ј-CGCTCTAAGGCTAACCCT-3Ј; MEFmt1 The enhancer A site sequence includes a Mef2 binding motif (bottom), 5Ј-AGGGTTAGCCTTAGAG CG-3Ј (mutated nucleotides un- The ␭ enhancers were identified, and essential binding sites and derlined). Annealed oligonucleotides were labeled by filling them with labeled ([32P]dATP and [32P]dCTP) and unlabeled (dGTP and dTTP) nu- regulatory domains were mapped in earlier studies (14). These cleotides and Klenow enzyme and gel-purified on an 8% native polyacryl- results are summarized in Figure 1A, where the sequence of the amide gel. Binding reactions were conducted in 20-␮l volumes containing E␭2-4 and E␭3-1 enhancers of the Ig ␭ light chain gene are shown 10 mM Tris (pH 7.5), 50 mM NaCl, 1 mM DTT, 1 mM EDTA, 5% glyc- with two E-boxes and the sequences of the unique ␭A and ␭B ␮ ␮ erol, 2 g of poly(dI-dC) (Pharmacia, Piscataway, NJ), 0.45 g of soni- DNase I footprinted sites. When the ␭A site sequence was ana- cated salmon sperm DNA, 0.45 ␮g of denatured sonicated salmon sperm DNA, 30,000 to 50,000 cpm of probe and 8 to 10 ␮g of nuclear extract or lyzed using a transcription factor database (Transcription Factor 2 ␮l of in vitro translated protein. They were incubated at room tempera- Search, 1995, Yutaka Akiyama, Kyoto University) an AϩT rich ture for 15 min and electrophoresed on a 4% nondenaturing polyacryl- consensus binding site for Mef2 was identified. Figure 1B shows amide gel at 200 V for 2 h, using 0.5ϫ TBE (1ϫ TBE contains 100 mM the sequence of the ␭A site with the Mef2 binding motif in bold Tris-borate and 2 mM EDTA). For competition assays, 100 ng of the re- ␭ spective annealed cold competitor oligonucleotide was included and, for face. It can be seen that the A-footprinted region includes the Ab supershift experiments, nuclear extract or in vitro translated proteins entire Mef2 binding sequence. were incubated with 2 ␮l of preimmune or immune serum in 1ϫ binding When analyzed by EMSA (Fig. 2), two major complexes are The Journal of Immunology 4797

and forms a major gel-shift complex that comigrates with the lower mobility complex JA1, obtained with J558L nuclear extract (Fig. 3). Mef2C gives a gel-shift band that runs faster than the JA1 complex, which may be an artifact of in vitro synthesis. These Mef2A and Mef2C complexes are successfully competed out by the MCK oligonucleotide (high affinity site for Mef2) but not affected by the mutant site Mef2 mt1. The gel-shift complexes are formed only with reticu- locyte lysates programmed with Mef2A or C mRNA, showing that it is a specific complex (data not shown). Furthermore, the gel-shift complexes formed by Mef2A and Mef2C are clearly supershifted by anti-Mef2 Ab (Fig. 4A, see below). These data suggest that the JA1 complex may include one or more Mef2 proteins. JA2 runs faster than the Mef2 proteins, but its formation is dependent on the presence of an intact Mef2 consensus sequence (Mef2 mt1 does not compete it out, and anti-Mef2 Abs elim- inate this complex (Fig. 4, see below) (Fig. 2). Therefore, we believe that the JA2 complex represents a breakdown product of Mef2 proteins and is unlikely to contain a different protein. Downloaded from Mef2 proteins are present in gel-shift complexes obtained with the ␭A site To determine whether Mef2 protein/s are indeed present in the JA1 and JA2 complexes, we used three rabbit antisera raised against specific Mef2 family members: one raised against a peptide from

Mef2A that is specific for Mef2A but cross-reacts with Mef2C and http://www.jimmunol.org/ two that were reported to be relatively Mef2B and Mef2D specific (41). Anti-Mef2A (Fig. 4A) and anti-Mef2D (Fig. 4C) supershift the bands obtained with J558L nuclear extract and the labeled ␭A oligonucleotide whereas the anti-Mef2B Ab has no effect (Fig. 4B). Anti-Mef2A Ab supershifts bands obtained with both Mef2A FIGURE 2. Two major complexes, JA1 and JA2, form on the labeled and Mef2C proteins as expected, since it is known to be cross- ␭A oligonucleotide. EMSA experiments were performed with labeled ␭A oligonucleotide and J558L nuclear extracts; competitor DNAs (100 ng) reactive. However, anti-Mef2D Ab also supershifts bands obtained ␭ added during the binding reactions are indicated. with Mef2A and Mef2C proteins and the labeled A oligonucleotide, indicating that it also cross-reacts. These Abs do not affect by guest on September 24, 2021 the gel-shift bands containing an unrelated protein (Pu.1 and the formed by the labeled ␭A oligonucleotide and J558L nuclear ex- labeled ␭B oligonucleotide; data not shown). tracts. They are denoted as JA1 and JA2, in decreasing order of These data demonstrate that Mef2 proteins are present and re- their m.w. (there are also minor complexes that we believe to be sponsible for the gel-shift bands JA1 and JA2 that are obtained breakdown products, as will be described later). To determine with the J558L nuclear extract and the labeled ␭A oligonucleotide whether a high affinity site for Mef2 could effectively compete and suggest that Mef2B is not involved but that Mef2A, Mef2C, or with the ␭A site in an EMSA, we chose a sequence from the MCK Mef2D appear to be involved. These antisera exhibit cross-reac- gene, which had been earlier identified as a functional element, tivity. This and other considerations (see next section) suggest that binding muscle-specific factors (18). This MCK oligonucleotide only a subset of Mef2 proteins are present in the complexes with competes for both gel-shift bands (JA1 and JA2), but a mutant J558L nuclear extract. MCK oligonucleotide, MEFmtl (22), with mutations at the conserved site effecting Mef2 binding, fails to compete (Fig. 2). As Mef2A, -C, and -D are present in J558L nuclear extract expected, unlabeled ␭A oligonucleotide also successfully com- EMSA data suggest that one or more members of the Mef2 family petes whereas the ␭B oligonucleotide, which does not contain a interact with the ␭A site. To formally confirm the presence of Mef2 binding site, does not compete. Mef2 family members in J558L cells, we used Western blot as- The only sequence element shared by the ␭A motif and MCK says. Although Mef2 activity has been primarily characterized and oligonucleotide is the Mef2 consensus binding site, and mutations at studied in the muscle cell, there are a number of reports of the the conserved residues affecting Mef2 binding affect the ability of the presence of Mef2 proteins in different tissue and cell types (42), MCK oligonucleotide to compete for gel-shift bands obtained with the but there were no previous reports of Mef2 proteins in J558L cells. labeled ␭A site oligonucleotide. Together, these data suggest that pro- In mammals, Mef2A and Mef2D transcripts are expressed in many tein/s that bind the ␭A site are capable of specifically binding the tissues, while Mef2C transcripts are restricted to muscle, brain, and MCK sequence and are therefore likely to contain one or more of the spleen (26). We analyzed nuclear extracts from J558L myeloma Mef2 proteins or a highly related protein. cells and a non-B cell, NIH 3T3 fibroblasts, using Abs against Mef2 proteins in Western blots (Fig. 5). The immunoblot analysis ␭ In vitro synthesized Mef2 proteins bind specifically to the A shows that both J558L and NIH 3T3 cells have Mef2A and/or site and give gel-shift bands that comigrate with the JA1 Mef2C (since anti-Mef2A Ab cross-reacts with Mef2C) and complex Mef2B. Another laboratory, however, has failed to detect the pres- To directly demonstrate that Mef2 proteins can bind the ␭A site, we ence of Mef2A in J558L cells (43); it is therefore likely that the performed EMSA with in vitro synthesized Mef2A and Mef2C, as signal obtained with the anti-Mef2A Ab could be entirely due to representative members of the Mef2 family. Mef2A binds the ␭A site Mef2C. Both cells also appear to have Mef2D although the anti- 4798 Mef2 PROTEINS BIND THE ␭A SITE IN THE Ig ␭ ENHANCER Downloaded from http://www.jimmunol.org/

FIGURE 3. Gel-shift complex formed by in vitro synthesized Mef2 proteins on labeled ␭A oligonucleotide. EMSA experiments with labeled ␭A oligonucleotide and J558L nuclear extract (lanes 2–4), Mef2A (lanes 5–7), or Mef2C (lanes 8–10) are shown. Competitor DNA (100 ng) added during by guest on September 24, 2021 binding reactions are indicated.

Mef2D signal could also be due to the cross-reactivity it exhibits hancer, we designed a CAT reporter construct, A3-TKCAT. It con- with Mef2A and Mef2C. The appearance of multiple bands in the sists of a bacterial CAT gene driven by a TK promoter, with a trimer immunoblotting analysis done on nuclear extracts has been docu- of ␭A site oligonucleotides cloned upstream of the promoter. A3- mented by others (42)). This heterogeneity is believed to be due to TKCAT reporter constructs were transfected into the J558L B cell posttranslational modifications, such as phosphorylation, and the line, and their CAT activity was assayed (Fig. 6A). Compared with the presence of multiple Mef2 isoforms. promoter-only construct, A3-TKCAT shows a greatly increased CAT The presence of several members of the Mef2 family in the activity, demonstrating the ability of the multimerized ␭A site to act J558L nuclear extract supports the EMSA analysis, indicating that as an enhancer in conjunction with the TK promoter. members of the Mef2 family play a role in the transcriptional regulation of the ␭A site. Of the Mef2 family members, Mef2A is unlikely to be involved for the reason discussed above. Mef2B Overexpression of Mef2 proteins exerts a negative regulatory does not seem to be present in the gel-shift complexes (Fig. 4B). effect on the trans-activating capability of the A3-TKCAT Although the Mef2C gel-shift band runs faster than JA1, it is pos- reporter construct sible that this is a characteristic of the in vitro synthesized protein The functional activity of the A3-TKCAT reporter construct implies and may not be true of the protein made in the cell. That leaves that transcription factors present in the J558L B cells are capable of Mef2C and Mef2D as possible candidates for the gel-shift com- transcriptional activation via the ␭A site. We now wanted to see the ␭ plexes formed with the J558L nuclear extract and the A site. effect of overexpressing Mef2 proteins in J558L cells. To our surprise, cotransfecting with Mef2 cDNAs suppressed the CAT activity of the ␭ ␭ Functional analysis of the A motif in the regulation of the L A3-TKCAT reporter construct by about 50 to 60%, when compared chain enhancer with the CAT activity of the cells transfected with the A3-TKCAT Ig enhancer elements, such as the octamer element (44–46), the ␬B reporter construct alone (Fig. 6A). Mef2 cDNAs cotransfected with site (40, 47), the ␮E3 site (48), and the ␭B site (16), form strong the promoter-only control plasmid TKCAT did not give any signifi- trans-activating elements upon multimerization of single protein bind- cant CAT activity (data not shown). To determine whether the sup- ing domains. For this reason we tested the enhancer function of the A pression is specific, we repeated the cotransfection experiments using site, using it as a trimer. We have not tested the A site as a monomer. a CAT reporter construct with a multimerized ␭B site and cDNAs for In studies done in the muscle system also, the Mef2 site is used as a Mef2A and Mef2C. This reporter construct, B4-TKCAT, has four ␭B dimer (22). To correlate the in vitro ability of Mef2 proteins to bind sites cloned upstream of the TK promoter and serves as an enhancer the ␭A site and in vivo activity of the ␭A domain of the E␭2-4 en- in B cells but is not expected to be bound by Mef2 proteins (15, 16). The Journal of Immunology 4799 Downloaded from http://www.jimmunol.org/ by guest on September 24, 2021

FIGURE 4. A, B, and C, Mef2 proteins are present and responsible for the complexes that bind the labeled ␭A oligonucleotide. A, Anti-Mef2A Ab can supershift JA1 complexes in EMSAs. EMSAs with labeled ␭A oligonucleotide and J558L nuclear extract (lanes 2–4), in vitro synthesized Mef2A (lanes 5–7), or Mef2C (lanes 8–10) and with no Ab (Ϫ) or preimmune serum (PI) or anti-Mef2A Ab (ϩ). Arrow indicates supershifted complex. B, Anti-Mef2B Ab does not supershift JA1 and JA2 complexes. EMSAs with labeled ␭A oligonucleotide and J558L nuclear extract (lanes 2–4), in vitro synthesized Mef2A protein (lanes 5–7), or Mef2C (lanes 8–10) and with no Ab (Ϫ) or preimmune serum (PI) or anti-Mef2B Ab (ϩ). No supershift is seen. C, Anti-Mef2D Ab supershifts JA1 and JA2 complexes. EMSAs with labeled ␭A oligonucleotide and J558L nuclear extract (lanes 2–4), in vitro synthesized Mef2A protein (lanes 5–7), or Mef2C (lanes 8–10) and with no Ab (Ϫ) or preimmune serum (PI) or anti-Mef2D Ab (ϩ). Arrow indicates supershifted complex.

Indeed the B4-TKCAT reporter construct is not affected by overex- therefore clearly affect the enhancer activity of the ␭A site, albeit in a pression of Mef2A and Mef2C (Fig. 6B). Thus, the suppression of the negative manner, since the suppression is clearly speciﬁc and depen- A3-TKCAT construct by overexpressing Mef2 proteins is clearly spe- dent on the amount of the Mef2 protein expressed in these cells. These ciﬁc. data together with the EMSA analysis implicate Mef2 proteins to be Furthermore, the effect of overexpression of Mef2 proteins is dose functionally involved in the transcriptional regulation at the ␭A site. dependent (Fig. 7). Increasing amounts of Mef2C cDNA transfected into J558L cells translates into increase in the suppression of the A3- Discussion TKCAT reporter activity. A similar negative regulatory effect follow- In this report we demonstrate that the ␭A motif of the Ig ␭ light chain ing overexpression of Mef2C cDNA has been described in another enhancer, E␭2-4, includes a binding site for the Mef2 family of tran- system using the c-jun promoter (see Discussion). Mef2 proteins scription factors and is bound in vitro by members of the Mef2 family. 4800 Mef2 PROTEINS BIND THE ␭A SITE IN THE Ig ␭ ENHANCER Downloaded from FIGURE 7. Overexpression of Mef2C negatively regulates the A3- TKCAT reporter activity in a dose-dependent manner. Functional activity of the A3-TKCAT reporter construct in transient transfection assays in J558L cells, either by itself or together with various amounts of Mef2C cDNA expression constructs is shown. As control, promoter-only TKCAT plasmid either by itself or together with 20 ␮g of Mef2C cDNA expression

construct was used. The average CAT activity Ϯ SD of at least two inde- http://www.jimmunol.org/ pendent transfections is given as a percentage of CAT activity of cells transfected with A3-TKCAT alone. FIGURE 5. Mef2 proteins are present in J558L nuclear extract. Western blot analysis of J558L nuclear extract (lane 1), 3T3 nuclear extract (lane 2), Mef2A (lane 3), and Mef2C (lane 4) using anti-Mef2A (A) or anti-Mef2B tion factors essential for muscle cell differentiation, have been shown (B) or anti-Mef2D (C) Abs. to be involved in Ig gene expression. Interestingly, the A3-TKCAT reporter construct, however, exhibits CAT activity when transfected Further, we show that the ␭A motif functions as an enhancer in con- into a non-B cell line, namely NIH 3T3 mouse fibroblast cells (data junction with a TK promoter, and we show functional evidence of the not shown), suggesting that the A site, though essential, may not need role of Mef2 in transcriptional regulation via the ␭A site. To our any B cell-specific factors for its activity in the B cell. The two Ig ␭ by guest on September 24, 2021 knowledge, this is the first time that Mef2 family proteins, transcrip- light chain enhancers, E␭2-4 and E␭3-1, are highly homologous and

FIGURE 6. Overexpression of Mef2 proteins negatively regulates the functional activity of the A3-TKCAT reporter construct. A, Functional activity of A3-TKCAT reporter construct in transient transfection assays in J558L cells, either by itself or with Mef2A, Mef2B, Mef2C, or Mef2D cDNA expression constructs is shown. The average CAT activity Ϯ SD of at least two independent transfections is given as a percentage of CAT activity of cells transfected with A3-TKCAT alone. B, Functional activity of B4-TKCAT reporter construct in transient transfection assays in J558L cells, either by itself or with Mef2A or Mef2C cDNA expression constructs is shown. The average CAT activity Ϯ SD of at least two independent transfections is given as a percentage of CAT activity of cells transfected with A3-TKCAT alone. The Journal of Immunology 4801

are believed to function in a similar manner (14). The ␭A site is In the transient transfection experiments, overexpression of identical between both these enhancers except for a single nucleotide Mef2 proteins had a negative effect on the transcriptional activa- difference that occurs at a variable position in the Mef2 consensus tion of the reporter construct A3-TKCAT. All four Mef2 cDNAs sequence. We therefore believe that the ␭A motif of the E␭3-1 en- that we tested demonstrated this negative effect. The effect was hancer would behave in a similar fashion with regard to the binding specific because it was seen only when the reporter construct had of the Mef2 proteins and enhancer activity in our reporter construct a Mef2 binding sequence (Fig. 6). Suppression of A3-TKCAT by assays. Mef2C was shown to be dose dependent (Fig. 7). Overexpression Mef2 activity, though first identified in the context of muscle spe- of Mef2C has been shown to exert a similar dose-dependent neg- cific enhancers, (18) is not restricted to the muscle cell. A Mef2 con- ative regulatory effect in an LPS-induced c-Jun reporter system sensus site is present in the brain creatine kinase gene promoter (49). (52). The authors argue that the overexpressed Mef2C competes Mef2 sites are also present in the promoters of two immediately early with heterodimers of Mef2 proteins that have the ability to activate gene promoters, c-jun and N10, that are activated in quiescent cells transcription. However, at this time, there is no evidence showing following mitogenic stimulation (50). Mef2 proteins have been shown that heterodimers consisting of more than one Mef2 protein are to be present and capable of binding DNA in a number of cultured cell better trans-activating agents than homodimers formed by a single lines, both lymphoid and nonlymphoid, of human and mouse origin Mef2 protein (E. Olson, unpublished observations). (42). However, until a recent report by Swanson et al. (43), there were The suppression of transcriptional activity by the overexpressed no reports of the presence of Mef2 in the ␭-expressing myeloma cell Mef2 proteins could be explained by proposing a model where the line J558L. Our data clearly show that Mef2 proteins are present in overexpressed Mef2 proteins compete with a “functional transcrip- Downloaded from J558L nuclear extracts. It is therefore formally possible for any of tion complex” that binds and regulates transcription in vivo. This these Mef2 proteins by themselves or as heterodimers to regulate tran- “functional transcription complex” could be a posttranslationally scription at the ␭A site. Using Abs specific for Mef2A (which also modified form of Mef2 protein, such as a phosphorylated form. cross-react with Mef2C, Fig. 4C), Mef2B (Fig. 4B), and Mef2D The overexpressed proteins may not be properly phosphorylated (cross-reacts with MEFA and Mef2C, we were unable to identify a and therefore may be transcriptionally inefficient. On the other specific member in these Ab supershift experiments since all these hand, the overexpressed Mef2 proteins may be identical to the Abs, except anti-Mef2B, independently supershifted the gel-shift endogenous Mef2 proteins, and the excess of Mef2 protein may http://www.jimmunol.org/ sequester other factors needed for transcriptional activation, mak- bands. Since Mef2B is clearly present in the B cell nuclear extract ing them unavailable for the formation of the “functional transcrip- (Fig. 5) and the anti-Mef2B Ab does not supershift the gel-shift bands tional complex,” a phenomenon called squelching (53). Finally, (Fig. 4B), it appears that Mef2B may not be involved in the regulation the Mef2 cDNAs that were used for the cotransfection experiments via the ␭A motif. Based on our data and the data from another lab- were obtained from muscle cells. Different splice forms of Mef2 oratory (43) that failed to detect Mef2A in J558L cells, Mef2C and -D mRNA have been shown to be expressed in different tissues (22– are the likely candidates to bind and regulate transcription via the ␭A 24, 30–32). It is possible that this muscle cell form of Mef2 protein motif. In our EMSA experiments, cell-free-produced Mef2C proteins is not capable of transcriptional activation via the ␭A site but com-

gave a gel-shift band that runs faster than JA1 complex. Furthermore, by guest on September 24, 2021 petes with the “functional transcription complex” for binding to the ␭A motif functions as an enhancer in NIH 3T3 cells that do not the ␭A site. However, when we compared the Mef2C mRNA spe- express Mef2C (Fig. 5) (32). This might suggest that Mef2D may be cies produced in mouse skeletal muscle and J558L cells, we found involved or that there is a degeneracy enabling any Mef2 family mem- that they are the same (data not shown). In summary, it could be ber to bind and regulate transcription. It is also possible that more than postulated that the overexpressed Mef2 protein is capable of bind- ␭ one Mef2 protein binds the A site as a heterodimer or homodimer ing the ␭A site but that it is unable to bring about transcriptional since there is evidence from the muscle system that Mef2 proteins activation or that it is able to squelch the “functional complex.” appear to bind DNA as dimers (51). Loss of function assays have been At this time we have no direct experimental evidence of other done with Mef2 family members. Mef2B knockout mice that were factors being involved along with Mef2 proteins in transcriptional generated do not show any birth defects (32) while Mef2C null mice regulation at the ␭A site. However, the consensus emerging from show cardiogenic defects and embryonic lethal phenotype (27). How- muscle cell gene expression studies strongly suggests that Mef2 ever, the current thinking on Mef2 proteins is that they have overlap- acts as a cofactor for myogenic bHLH proteins in the skeletal ping function in the tissues in which they are expressed (E. Olson, muscle differentiation program but does not activate the program unpublished observations). It is therefore possible that there could be by itself (54). As described earlier, members of the Mef2 family a certain amount of degeneracy with regard to the requirement of a are fairly ubiquitous in their distribution in several cell types. It is specific Mef2 family member for transcriptional regulation in the con- therefore possible that some other protein/s is/are involved in ren- text of a specific enhancer. dering cell-type specificity to this interaction of Mef2 proteins and In the transient transfection assays the ␭A motif functions as en- the ␭A site. In the muscle cell a transcription complex containing hancer in conjunction with the TK promoter in the J558L B cell line, Mef2 and a heterodimer of MyoD and E12, a ubiquitous bHLH indicating that functional transcription factor/s bind and regulate tran- protein (55), regulates transcription (20). The model proposed for scription via the ␭A site. Mutations in the region of the ␭A site that this interaction suggests that a protein-protein interaction between encompass the consensus Mef2 binding site have been shown to sig- Mef2 and a MyoD/E12 heterodimer, with any one of them binding nificantly suppress the enhancer activity using a CAT reporter con- to the DNA, is sufficient to activate transcription (20). It is inter- struct E␭2-4 TKCAT2, with the full-length E␭2-4 enhancer upstream esting to note that the ␭A site is flanked by two E-boxes, which can of the TK promoter (14). These experiments were done with both the bind bHLH proteins such as E12 and E47 (55). One of the E-boxes E␭2-4 and E␭3-1 enhancers, suggesting the importance of the ␭A is located a mere 15 bp upstream. Earlier work using CAT reporter motif in transcriptional regulation. Although these mutations were constructs has shown that both E-boxes are essential for full ac- extensive and it could be argued that the mutations affected the bind- tivity of the E␭2-4 and E␭3-1 enhancers in the mouse (14) and also ing of other factors to the ␭A site, they underscore the importance of in humans (17). Extrapolating from the model proposed for the the Mef2 binding site, which was the main region that was altered in muscle cell, it is possible that a protein-protein interaction may be these mutants. involved between Mef2 protein binding at the ␭A site and bHLH 4802 Mef2 PROTEINS BIND THE ␭A SITE IN THE Ig ␭ ENHANCER

proteins, or with some other factor/s. However it is likely that one 26. Martin, J. F., J. J. Schwarz, and E. N. Olson. 1993. Myocyte-enhancer factor would miss such an interaction in EMSA experiments, which are (MEF) 2C: a tissue-restricted member of the MEF-2 family of transcription factors Proc. Natl. Acad. Sci. USA 90:5282. too harsh to preserve these subtle protein-protein associations; this 27. Qing, L., J. Schwarz, B. Corazon, and E. N. Olson. 1997. Control of mouse cardiac has been seen in the case of muscle-specific enhancers (54). Given morphogenesis and myogenesis by transcription factor MEF2C. Science 276:1401. these observations, it is likely that other factors are involved along 28. Edmondson, D. G., G. E. Lyons, J. F. Martin, and E. N. Olson. 1994. MEF2 gene expression marks the cardiac and skeletal muscle lineages during mouse embry- with the Mef2 proteins in transcriptional regulation at the ␭A site. ogenesis. Development 120:1251. 29. Lyons, G. E., B. K. Micales, J. Schwarz, J. F. Martin, and E. N. Olson. 1994. Acknowledgments Expression of mef2 genes in the mouse central nervous system suggests a role in neruronal maturation. J. Neurosci. 15:5727. We acknowledge Dr. E. N. Olson, University of Texas, Southwestern Med- 30. Leifer, D., D. Krainc, Y.-T. Yu, J. McDermott, R. E. Breitbart, J. Heng, ical Center at Dallas for the pcDNAI expression constructs for Mef2A, R. L. Neve, B. Kosofsky, B. Nadal-Ginard, and S. A. Lipton. 1993. MEF2C, a Mef2B, Mef2C, and Mef2D and Dr. R Prywes, Columbia University, New MADS/MEF2-family transcription factor expressed in a laminar distribution in cerebral cortex. Proc. Natl. Acad. Sci. USA 91:1546. York, for the gift of anti-Mef2 Abs. We thank Drs. T. E. Martin, 31. McDermott, J. C., M. C. Cardoso, Y.-T. Yu, V. Andres, D. Leifer, D. Krainc, G. E. Lyons, J. P. Engler, and N. Michael for critical reading of the paper S. A. Lipson, and B. Nadal-Ginard. 1993. hMEF2C gene encodes skeletal mus- and Drs. H. Singh and A. Brass for useful discussions. cle- and brain-specific transcription factors. Mol. Cell. Biol. 13:2564. 32. Molkentin, J. D., A. B. Firulli, B. L. Black, J. F. Martin, C. M. Hustad, References N. Copeland, N. Jenkins, G. Lyons, and E. N. Olson. 1996. MEF2B is a potent trans-activator expressed in early myogenic lineages. Mol. Cell. Biol. 16:3814. 1. Engler, P., P. Roth, J. Y. Kim, and U. Storb. 1991. Factors affecting the rear- 33. Schreiber, B., P. Matthias, M. M. Muller, and W. Schaffner. 1989. Rapid detec- rangement efficiency of an Ig test gene. J. Immunol. 146:2826. tion of octamer binding proteins with “miniextracts,” prepared from a small num- 2. Stuadt, L. M., and M. J. Lenardo. 1991. Immunoglobulin gene transcription. ber of cells. Nucleic Acids Res. 17:6419. Annu. Rev. Immunol. 9:373. 34. Dignam, J. D., R. M. Lebovitz, and R. G. Roeder. 1983. Accurate transcription Downloaded from 3. Calame, K., and S. Eaton. 1988. Transcriptional controlling elements in the im- initiation by RNA polymerase II in a soluble extract from isolated mammalian munoglobulin and T cell receptor loci. Adv. Immunol. 43:235. nuclei. Nucleic Acids Res. 11:1475. 4. Sen, R., and D. Baltimore. 1989. Factors regulating immunoglobulin-gene tran- 35. Bradford, M. M. 1976. A rapid and sensitive method for quantitation of micro- scription. In Immunoglobulin Genes. T. Honjo, F. W. Alt, and T. H. Rabbitts, eds. gram quantities of protein utilizing the principle of protein-dye binding. Anal. Academic Press, New York, p. 327. Biochem. 72:248. 5. Takeda, S., Y. Zou, H. Bluethman, D. Kitamura, U. Muller, and K. Rajewsky. 1993. 36. Singh, H., R. Sen, D. Baltimore, and P. A. Sharpe. 1986. A nuclear factor that Deletion of the immunoglobulin ␬ chain intron enhancer abolishes ␬ chain gene binds to a conserved sequence motif in transcriptional control elements of im- ␭ rearrangement in cis but not gene rearrangement in trans. EMBO J. 12:2392. munoglobulin gene. Nature 319:154. http://www.jimmunol.org/ ␬ 6. Xu, Y., L. Davidson, F. W. Alt, and D. Baltimore. 1996. Deletion of the Ig light 37. Prost, E., and D. D. Moore. 1986. CAT vectors for analysis of eukaryotic pro- chain gene intronic enhancer/matrix attachment retions impairs but does not abol- moters and enhancers. Gene 45:107. ␬ ␬ ish V J rearrangement. Immunity 4:377. 38. Nielsen, D. A., J. Chou, A. J. MacKrell, M. J. Casadaban, and D. F. Steiner. 1983. 7. Neuberger, M. S. 1983. Expression and regulation of immunoglobulin heavy Expression of preproinsulin-␤-galactosidase gene fusion in mammalian cells. chain gene transfected into lymphoid cells. EMBO J. 2:1373. Proc. Natl. Acad. Sci. USA 80:5128. 8. Queen, C., and D. Baltimore. 1983. Immunoglobulin gene transcription is acti- 39. Gorman, C. M., L. F. Moffat, and B. H. Howard. 1982. Recombinant genomes that vated by downstream sequence elements. Cell 33:741. express chloramphenicol acetyl-transferase in mammalian cells. Mol. Cell. Biol. 9. Matthias, P., and D. Baltimore. 1993. The immunoglobulin heavy chain locus 2:1044. contains another B-cell-specific enhancer close to the ␣ constant region. Mol. Cell. Biol. 13:1547. 40. Murre, C., P. S. McCaw, and D. Baltimore. 1989. A new DNA binding and 10. Pettersson, S., G. P. Cook, M. Bruggemann, G. T. Williams, and dimerization motif in immunoglobulin enhancer binding, daughterless, Myo D and myc proteins. Cell 56:777.

M. S. Neuberger. 1989. A second B cell-specific enhancer 3Ј of the immuno- by guest on September 24, 2021 globulin heavy-chain locus. Nature 307:165. 41. Han, T.-H., and R. Prywes. 1995. Regulatory role of MEF2D in serum induction 11. Meyer, K. B., and M. S. Neuberger. 1989. The immunoglobulin ␬ locus contains of the c-jun promoter. Mol. Cell. Biol. 15:2907. a second, stronger B cell-specific enhancer which is located downstream of the 42. Dodou, E., D. B. Sparrow, T. Mohun, and R. Treisman. 1995. MEF2 proteins, constant region. EMBO J. 8:1959. including MEF2A, are expressed in both muscle and non-muscle cells. Nucleic 12. Storb, U. 1989. Immunoglobulin gene analysis in transgenic mice. In Immuno- Acids Res. 23:4267. globulin Genes. T. Honjo, F. W. Alt, and T. H. Rabbitts, eds. Academic Press, 43. Swanson, B. J., H. M. Jack, and G. E. Lyons. 1998. Characterization of MEF2 New York, p. 303. expression in B and T cells: MEF2C is a B cell restricted transcription factor in 13. Hagman, J., C. Rudin, D. Haasch, D. Chaplin, and U. Storb. 1990. A novel lymphocytes. Mol. Immunol. In Press. enhancer in the immunoglobulin ␭ locus is duplicated and functionally indepen- 44. Gester, T., P. Matthias, M. Thali, J. Jiricny, and W. Schaffner. 1987. Cell type- dent of NF␬B. Genes Dev. 4:978. specificity elements of the immunoglobulin heavy chain enhancer. EMBO J. 6:1323. 14. Rudin, C., and U. Storb. 1992. Two conserved essential motifs of the murine immu- 45. Muller-Immergluck, M. M., W. Schaffner, and P. Matthias. 1990. Transcription noglobulin ␭ enhancers bind B-cell-specific factors. Mol. Cell. Biol. 12:309. factor Oct-2A contains functionally redundant activating domains and works se- 15. Eisenbeis, C., H. Singh, and U. Storb. 1993. PU.1 is a component of a multipro- lectively from a promoter but not from a remote enhancer position in non-lym- tein complex which binds an essential site in the murine immunoglobulin ␭2-4 phoid (He La) cells. EMBO J. 9:1625. enhancer. Mol. Cell. Biol. 13:6452. 46. Tanaka, M., U. Grossniklaus, W. Herr, and N. Hernandez. 1988. Activation of U2 16. Eisenbeis, C., H. Singh, and U. Storb. 1995. Pip, a novel IRF family member, is a snRNA promoter by the octomer motif defines a new class of RNA polymerase lymphoid-specific, PU.1-dependent transcriptional activator. Genes Dev. 9:1377. II enhancer elements. Genes Dev. 2:1764. 17. Glozak, M. A., and B. B. Blomberg. 1996. The human ␭ immunoglobulin en- 47. Pierce, J. W., M. Lenardo, and D. Baltimore. 1988. Oligonucleotide that binds hancer is controlled by both positive elements and developmentally regulated nuclear factor NF-kB acts as a lymphoid specific and inducible enhancer element. negative elements. Mol. Immunol. 33:427. Proc. Natl. Acad. Sci. USA 85:1482. 18. Gossett, L. A., D. J. Kelvin, E. A. Sternberg, and E. N. Olson. 1989. A new 48. Beckmann, H., L.-K. Su, and T. Kadesch. 1990. TFH3: a helix-loop-helix protein myocyte-specific enhancer-binding factor that recognizes a conserved element that activates transcription through the immunoglobulin enhancer ␮E3 motif. associated with multiple muscle-specific genes. Mol. Cell. Biol. 9:5022. Genes Dev. 4:167. 19. Suzuki, E., K. Guo, M. Kolman, Y.-T. Yu, and K. Walsh. 1995. Serum induction 49. Horlick, R. A., G. M. Hobson, J. H. Patterson, M. T. Mitchell, and P. A. Benfield. of MEF2/RSRF expression in vascular myocytes is mediated at the level of trans- 1990. Brain and muscle creatinine kinase genes contain common TA-rich rec- lation. Mol. Cell. Biol 15:3415. ognition protein-binding regulatory elements. Mol. Cell. Biol. 10:4826. 20. Molkentin, J. D., and E. N. Olson. 1996. Defining the regulatory networks for 50. Han, T., W. W. Lamph, and R. Prywers. 1992. Mapping of epidermal growth muscle development. Curr. Opin. Genet. Dev. 6:445. factor-, serum-, and phorbol ester-responsive sequence elements in the c-jun pro- 21. Cserjesi, P., and E. N. Olson. 1991. Myogenin induces the myocyte-specific moter. Mol. Cell. Biol. 12:4472. enhancer binding factor MEF-2 independently of other muscle-specific gene products. Mol. Cell. Biol. 11:4854. 51. Molkentin, J. D., B. L. Black, J. F. Martin, and E. N. Olson. 1996. Mutational 22. Yu, Y.-T., R. E. Breitbart, L. B. Smoot, Y. Lee, V. Madhavi, and analysis of the DNA binding, dimerization, and transcriptional activation do- B. Nadal-Ginard. 1992. Human myocyte-specific enhancer factor 2 comprises a mains of MEF2C. Mol. Cell. Biol. 16:2627. group of tissue-restricted MADS box transcription factors. Genes Dev. 6:1783. 52. Han, J., Y. Jiang, V. V. Kravchenko, and R. J. Ulevitch. 1997. Activation of the tran- 23. Breitbart, R., C. Liang, S. Smoot, D. Laheru, V. Madhavi, and B. Nadal-Ginard. scription factor MEF2C by the MAP kinase p38 in inflammation. Science 386:296. 1993. A fourth human MEF2 transcription factor, hMEF2D, is an early marker of 53. Gill, G., and M. Ptashne. 1988. Negative effect of the transcriptional activator the myogenic lineage. Development 118:1095. GAL4. Nature 334:721. 24. Martin, J. F., J. M. Miano, C. M. Hustad, N. G. Copeland, N. A. Jenkins, and 54. Molkentin, J. D., B. L. Black, J. F. Martin, and E. N. Olson. 1995. Cooperative E. N. Olson. 1994. A Mef2 gene that generates a muscle-specific isoform via activation of muscle gene expression by MEF2 and mygenic bHLH proteins. Cell alternative mRNA splicing. Mol. Cell. Biol. 90:5282. 83:1125. 25. Shore, P., and D. Sharrocks. 1995. The MADS-box family of transcription fac- 55. Atchison, M. L. 1988. Enhancers: mechanisms of action and cell specificity. tors. Eur. J. Biochem. 229:1. Annu. Rev. Cell Biol. 4:127.