Identification of a murine TEF-1-related gene expressed after mitogenic stimulation of quiescent fibroblasts and during myogenic differentiation.

Fibroblast growth factor (FGF)-1 binding to cell surface receptors stimulates an intracellular signaling pathway that ultimately promotes the transcriptional activation of specific genes. We have used a mRNA differential display method to identify FGF-1-inducible genes in mouse NIH 3T3 fibroblasts. Here, we report that one of these genes, FGF-regulated (FR)-19, is predicted to encode a member of the transcriptional enhancer factor (TEF)-1 family of structurally related DNA-binding proteins. Specifically, the deduced FR-19 amino acid sequence has ~89, 77, and 68% overall identity to chicken TEF-1A, mouse TEF-1, and mouse embryonic TEA domain-containing factor, respectively. Gel mobility shift experiments indicate that FR-19, like TEF-1, can bind the GT-IIC motif found in the SV40 enhancer. The FR-19 gene maps in the distal region of mouse chromosome 6, and analysis of several FR-19 cDNA clones indicates that at least two FR-19 isoforms may be expressed from this locus. FGF-1 induction of FR-19 mRNA expression in mouse fibroblasts is first detectable at 4 h after FGF-1 addition and is dependent on de novo RNA and protein synthesis. FGF-2, calf serum, platelet-derived growth factor-BB, and phorbol 12-myristate 13-acetate can also induce FR-19 mRNA levels. We have also found that FR-19 mRNA expression increases during mouse C2C12 myoblast differentiation in vitro. The FR-19 gene is expressed in vivo in a tissue-specific manner, with a relatively high level detected in lung. These results indicate that increased expression of a TEF-1-related protein may be important for both mitogen-stimulated fibroblast proliferation and skeletal muscle cell differentiation.

In this paper, we report that the murine FR-19 gene is a mitogen-inducible delayed-early response gene predicted to encode a TEF-1-related transcription factor. TEF-1 family members contain a highly conserved DNA-binding motif, the TEA domain (35). The first TEA domain-containing protein identified, human TEF-1, was initially described as a HeLa cell transcriptional activator which could bind the GT-IIC and Sph enhansons of the SV40 enhancer (36,37). Subsequently, it was shown to be involved in transactivation of the SV40 late promoter by large T antigen (38) and to bind to the human papilloma virus type 16 E6/E7 upstream regulatory region (39). Recently, it has been demonstrated that chicken (40) and mouse (41) TEF-1 can bind the M-CAT element, suggesting that these proteins may be involved in cardiac-and skeletal muscle-specific gene regulation. The function of FR-19 is presently unknown; however, our results indicate it may play a role in both mammalian cell growth and differentiation. After the original submission of this manuscript, Yockey et al. (42) reported the cDNA cloning and characterization of this same gene product, which they have named TEFR1.
RT-PCR Assays and cDNA Cloning-An FR-19B cDNA fragment was initially isolated using an RT-PCR-based approach. RNA (1 g) isolated from serum-starved or FGF-1-stimulated cells was converted to cDNA, and PCR was performed using sense SH2 domain and antisense leucine zipper domain primers as described (31). An aliquot of each amplification mixture was subjected to electrophoresis in a 2% agarose gel. X174/HaeIII restriction fragments (Clontech Laboratories) were used as size standards. DNA was visualized by ethidium bromide staining and a ϳ750-bp DNA fragment, termed FR-19, was excised, recovered using the freeze-squeeze method (43), reamplified, and ligated into the cloning vector pCRII (Invitrogen Corp.). The FR-19A cDNA frag-ment was also isolated using RT-PCR. RNA (1 g) isolated from serumstarved or FGF-1-stimulated cells was converted to cDNA and PCR was performed as described, except that only 30 cycles of amplification were used (44). The FR-19B primers were 5Ј-GACAAGCCCATCGACAAT-GATGCA-3Ј (sense) and 5Ј-TCACTGTAGCTTGGGCTTGAC-3Ј (antisense). Primers specific for FGFR-1, described previously (44), were used as a control to demonstrate that equivalent amounts of RNA were used for cDNA synthesis. An aliquot of each amplification mixture was subjected to electrophoresis in a 1.2% agarose gel, and DNA was visualized by ethidium bromide staining. The predicted 553-bp fragment and an additional 682-bp DNA fragment (representing an FR-19A cDNA) were recovered, reamplified, and cloned as described above. The relative abundance of the FR-19A and FR-19B mRNA splice variants in C2C12 cells was determined by RT-PCR using this same pair of primers and identical PCR conditions.
cDNA Library Screening-A mouse Balb/c 3T3 cell cDNA library (gift of T. Lanahan, Johns Hopkins University School of Medicine, Baltimore, MD) was screened with the PCR-derived FR-19B DNA fragment to obtain longer cDNA clones. The DNA fragment was labeled with [␣-32 P]dCTP as described above for the Northern blot hybridization probes. Approximately 2 ϫ 10 5 phage were plated at a density of 2 ϫ 10 4 plaque-forming units/150-mm dish using Escherichia coli C600 Hfl as host. Duplicate plaque lifts (Colony/Plaque screen, DuPont) were prehybridized, hybridized, washed, and exposed to film as described for Northern blots. Five positive phages were purified by one additional round of screening. PCR assays using oligonucleotide primers to the T7 and T3 promoters flanking the Uni-ZAP XR vector cloning site were performed to estimate the sizes of the various cDNA inserts. The phage clone with the largest insert was purified by one additional round of screening and amplified on E. coli C600 Hfl cells. Plasmid was excised from the phage clone using XL-1 Blue cells and R408 helper phage (Stratagene) as described (45).
5Ј-RACE Assays and cDNA Cloning-A cDNA fragment predicted to encode the amino-terminal region of FR-19 was isolated using the 5Ј-AmpliFINDER RACE kit (Clontech Laboratories). Total RNA (10 g) isolated from FGF-1-stimulated NIH 3T3 cells (4 h of treatment) and the antisense primer 5Ј-ACCTGGATGTGGCTGGAGACCTGCTT-3Ј were used in the cDNA synthesis step. PCR amplification was performed according to the manufacturer's instructions. The DNA products were then subjected to electrophoresis in a 1.5% agarose gel and visualized by ethidium bromide staining. Several DNA fragments were recovered, reamplified, and cloned as described above for RT-PCR amplification products.
cDNA Sequence Analysis-Plasmid DNA was purified using a Wizard Miniprep kit (Promega Corp.) and both strands of the entire ϳ1.4-kb FR-19B cDNA clone were sequenced by the dideoxynucleotide chain termination method. Both strands of the RT-PCR-derived FR-19A and FR-19B cDNAs, as well as the FR-19B cDNA fragment isolated by the 5Ј-RACE technique, were also completely sequenced. Since these were PCR-generated DNA fragments that could have misincorporated nucleotides, sequencing was performed on plasmid DNA isolated from several different bacterial colonies. Sequencing was either done automatically using an Applied Biosystems model 373A DNA sequencer or manually using a Sequenase 2.0 kit (U. S. Biochemical Corp.) and [ 35 S]dATP (1000 Ci/mmol, Amersham Corp.). The nucleic acid and deduced protein sequences were compared to sequences in the data base server at the National Center for Biotechnology using the Blast network service (46). Protein sequences were aligned using the University of Wisconsin GCG package.
In Vitro Transcription/Translation and Gel Mobility Shift Assays-A BamHI-KpnI fragment of FR-19B cDNA encoding the entire protein except for the first five amino acids was ligated in-frame into the pRSET-B procaryotic expression vector (Invitrogen). FR-19B mRNA was then transcribed in vitro using T7 RNA polymerase and translated in a rabbit reticulocyte lysate system (Promega) containing [ 35 S]methionine (1000 Ci/mmol; Amersham) according to the manufacturer's instructions. The in vitro translation products were subjected to electrophoresis on a 15% polyacrylamide-SDS gel. The gel was dried and proteins visualized by autoradiography. Gel mobility shift assays were performed essentially as described (47) using in vitro translated protein and the oligonucleotides described by Yasunami et al. (48). Briefly, the double-stranded wild-type GT-IIC oligonucleotide was labeled with [␣-32 P]dCTP (3000 Ci/mmol; Amersham) by filling in with DNA polymerase Klenow fragment (Boehringer Mannheim). Five l of each in vitro translation reaction was used in a binding reaction containing gel shift buffer (12% glycerol, 12 mM HEPES-NaOH (pH 7.9), 4 mM Tris-HCl (pH 7.9), 60 mM KCl, 1 mM EDTA, and 1 mM DTT), 83 g/ml poly(dI-dC), 333 g/ml bovine serum albumin, and ϳ4 ϫ 10 5 cpm (ϳ5 ng) of wild-type GT-IIC oligonucleotide. In some cases, 5 or 50 ng of cold competitor oligonucleotide, either wild-type or mutant (48), were added. DNA binding reactions were preincubated for 5 min at room temperature before the addition of the 32 P-labeled probe. Upon addition of the probe, the reactions were incubated for another 10 min at room temperature. The DNA-protein complexes were resolved by electrophoresis using a 4% polyacrylamide gel and 1 ϫ Tris-glycine buffer (50 mM Tris-HCl, 38 mM glycine, and 2 mM EDTA (pH 8.5)). Gels were dried and subjected to autoradiography at Ϫ80°C with intensifying screens.
Interspecific Mouse Backcross Mapping-Interspecific backcross progeny were generated by mating (C57BL/6J ϫ Mus spretus)F 1 females and C57BL/6J males as described (49). A total of 205 N 2 mice were used to map the Fr19 locus (see text for details). DNA isolation, restriction enzyme digestion, agarose gel electrophoresis, Southern blot transfer, and hybridization were performed essentially as described (50). All blots were prepared with Hybond N ϩ nylon membrane (Amersham). The ϳ1.4-kb FR-19 cDNA probe was labeled with [␣-32 P]dCTP (3000 Ci/mmol; Amersham) using a nick translation labeling kit (Boehringer Mannheim); washing was done to a final stringency of 0.8 ϫ SSCP, 0.1% SDS, 65°C. A major fragment of 11.0-kb was detected in EcoRI-digested C57BL/6J DNA, and major fragments of 13.0-and 8.2-kb were detected in EcoRI-digested M. spretus DNA. The presence or absence of the 13.0-and 8.2-kb EcoRI M. spretus-specific fragments, which cosegregated, was followed in backcross mice. A description of the probes and restriction fragment length polymorphisms for the loci linked to Fr19 including glucose transporter-3 (Glut3), FGF-6 (Fgf6), and Kirsten rat sarcoma oncogene-2 (Kras2) has been reported previously (51). Recombination distances were calculated as described (52) using the computer program SPRETUS MADNESS. Gene order was determined by minimizing the number of recombination events required to explain the allele distribution patterns.

Identification of an FGF-1-inducible mRNA by an RT-PCR-
based Approach-We have been attempting to identify FGF-1inducible genes that encode particular classes of proteins using a targeted differential display technique (30 -34). Briefly, in these experiments, RNA isolated from serum-starved or FGF-1-stimulated NIH 3T3 cells is converted to cDNA using reverse transcriptase and random primers. PCR is then performed using degenerate oligonucleotide primers, and the products are displayed on agarose gels. A pair of primers designed to anneal to DNA sequences encoding proteins with both an SH2 and a leucine zipper structural domain were used in the present study. Amplification products were separated by agarose gel electrophoresis, stained with ethidium bromide, and detected by ultraviolet illumination. The pattern of amplified cDNAs obtained using RNA from serum-starved or FGF-1-stimulated cells (for 2 or 12 h) was similar, except for an ϳ750-bp DNA fragment present in the 12 h post-stimulation lane. This fragment, termed FR-19, was excised from the gel, subcloned, radiolabeled, and used to screen a mouse fibroblast cDNA library in order to isolate longer cDNA clones. Positive phage were isolated, insert sizes were examined, and plasmid was rescued from the phage clone containing the largest insert, which was ϳ1.4 kb in size.
Regulation of FR-19 mRNA Expression in NIH 3T3 Cells-Northern blot hybridization analysis using RNA isolated from either serum-starved cells or cells treated for various lengths of time with FGF-1 was then performed to confirm the RT-PCR results indicating that FR-19 encoded an FGF-1-inducible mRNA. Three major (ϳ6.2, 2.9, and 1.8 kb in size) and one minor (ϳ4.1 kb) FR-19 mRNA species were detected in both untreated and treated cells (Fig. 1A). Increased FR-19 mRNA levels were evident at 4 h after FGF-1 addition, and expression remained high during the remainder of the experiment.
The effect of the RNA synthesis inhibitor actinomycin D on FGF-1 induction of FR-19 mRNA levels was then examined. Serum-starved NIH 3T3 cells were either left untreated or treated with FGF-1 alone, both FGF-1 and actinomycin D, or actinomycin D alone for 4 h. Cells were collected, RNA was isolated, and FR-19 mRNA levels analyzed by Northern blot hybridization. Actinomycin D treatment prevented FGF-1 induction of FR-19 mRNA (Fig. 1B); thus, the increase in FR-19 mRNA expression after FGF-1 addition is likely to be due, at least in part, to transcriptional activation of the FR-19 gene.
We next used the protein synthesis inhibitor cycloheximide to determine whether FGF-1 induction of FR-19 mRNA levels was dependent on de novo protein synthesis. Serum-starved cells were either left untreated or treated with FGF-1 alone, both FGF-1 and cycloheximide, or cycloheximide alone for 4 h. Cells were collected, RNA was isolated, and Northern blot hybridization analysis was performed. Cycloheximide addition alone resulted in a slight increase in FR-19 mRNA levels; nevertheless, FGF-1 induction of FR-19 mRNA levels did not occur in the presence of this inhibitor (Fig. 1C). This indicates that FGF-1-induced FR-19 mRNA expression requires the synthesis of intermediary proteins. Taken together, the results described above indicate that FR-19 is an FGF-1-inducible delayed-early response gene in NIH 3T3 fibroblasts.
We then determined whether calf serum, FGF-2, several of the growth factors present in calf serum (PDGF-BB, TGF-␤1, EGF, IGF-1), or the tumor promoter PMA could also increase FR-19 mRNA levels. In the first experiment, serum-starved cells were either left untreated or treated with 10% calf serum for various lengths of time. RNA was isolated, and Northern blot hybridization analysis was performed. Calf serum increased FR-19 mRNA levels with more rapid kinetics than those observed after FGF-1 treatment. Elevated FR-19 mRNA levels were first detected at 2 h post-serum stimulation and a slightly decreased level of expression was apparent at 24 h ( Fig.  2A). In the second experiment, serum-starved cells were either left untreated or treated with FGF-1, FGF-2, PDGF-BB, TGF-␤1, EGF, IGF-1, calf serum, or PMA for 4 h. RNA was isolated and FR-19 mRNA levels analyzed by Northern blot hybridization. All of these agents except for IGF-1 were able to increase FR-19 mRNA expression (Fig. 2B). PMA was the strongest inducer, while FGF-1, FGF-2, PDGF-BB, and calf serum were slightly less potent. TGF-␤1 or EGF treatment increased FR-19 mRNA expression only slightly above basal levels.
FR-19 cDNA Sequence Analysis-Both strands of the ϳ1.4-kb FR-19 cDNA insert were sequenced by the dideoxynucleotide chain termination method. The nucleotide sequence contained a long open-reading frame and a 202-bp 3Ј-untranslated region with a typical polyadenylation signal followed by a poly(A) tract. However, an initiating AUG methionine codon was not present. Therefore, we used the 5Ј-RACE method (53) to isolate cDNA fragments likely to encode additional amino-terminal protein sequence. DNA sequence analysis of the longest 5Ј-RACE-derived clone (319-bp) indicated that 85-bp of this fragment represented new 5Ј sequence not present in the ϳ1.4-kb cDNA. The 319-bp cDNA clone did not contain an AUG initiation codon; however, it did contain an AUU isoleucine codon flanked by a favorable sequence for translation initiation (54,55). As discussed below, the deduced amino acid sequence of FR-19 has significant identity to human, mouse, and chicken TEF-1. It has been demonstrated that translation of human TEF-1 mRNA is initiated at an AUU codon in vitro and in vivo (36); thus, we consider it likely that the AUU is the initiation codon. If this is indeed the case, the composite nucleotide sequence would encode a protein of 384 amino acids with a predicted molecular mass of 43,650 daltons and an estimated isoelectric point of 6.97. Computer analysis of the predicted FR-19 protein sequence revealed that it contains numerous potential phosphorylation sites as well as a bipartite nuclear targeting motif (amino acid residues 88 -104). This predicted FR-19 protein is referred to as FR-19B to distinguish it from a second predicted isoform, FR-19A (discussed below).
We sequenced both ends of the original RT-PCR-derived ϳ750-bp cDNA clone to identify the FR-19 cDNA sequences that annealed to the oligonucleotide primers. This indicated that the sense SH2 domain oligonucleotide functioned as both a sense and antisense primer. Comparison of the SH2 primer sequence with the appropriate regions of the ϳ1.4-kb FR-19 cDNA sequence revealed that it had ϳ62% and ϳ81% nucleotide sequence identity to sequences within the FR-19B coding and 3Ј-untranslated regions, respectively. Thus, as expected from these findings, the predicted FR-19B protein does not contain a SH2 or leucine zipper structural motif (see below).
FR-19B Amino Acid Sequence Comparisons-A search of the sequence data bases revealed that the deduced amino acid sequence of FR-19B has ϳ89, 77, 77, and 68% overall sequence identity to chicken TEF-1A (40), human TEF-1 (36), mouse TEF-1 (41,56,57), and mouse ETF (48), respectively. These four proteins share a common amino-terminal DNA-binding motif, referred to as a TEA domain (35). There is ϳ93% amino acid sequence identity between the 72-amino acid TEA domains of FR-19B and these other TEF-1-related proteins. An alignment of the deduced FR-19B, chicken TEF-1A, mouse TEF-1, and mouse ETF amino acid sequences is shown in Fig.  3. The predicted FR-19B amino acid sequence is identical to the TEFR1b sequence recently reported by Yockey et al. (42).
Isolation and Sequence Analysis of FR-19A cDNA-The sequence alignment results indicate that the predicted FR-19B protein is missing a stretch of 43 amino acids located just carboxyl-terminal to the TEA domain in other TEF-1-related proteins. Stewart et al. (40) have reported that there are several chicken TEF-1 protein isoforms that are probably generated via alternative splicing of TEF-1 pre-mRNA. Therefore, we investigated whether an FR-19 mRNA splice variant that could encode this missing region was expressed in NIH 3T3 cells. RT-PCR assays were performed using RNA isolated from serum-starved or FGF-1-stimulated cells and a pair of FR-19B oligonucleotide primers that flanked the deleted region. Primers specific for FGFR-1 were also used as a control to demonstrate that equivalent amounts of RNA were used for cDNA synthesis. Two distinct amplification products were detected when the FR-19 primers were used (Fig. 4A). The smaller DNA fragment was of the size predicted for an amplification product of FR-19B cDNA (553 bp), while the larger DNA fragment was ϳ700 bp in size. Each FR-19 cDNA fragment was recovered, cloned, and sequenced. The nucleotide sequence of the FR-19B cDNA was identical to the corresponding region of the ϳ1.4-kb cDNA clone described previously. The nucleotide sequence of the larger FR-19 cDNA fragment was identical to the 553-bp FR-19B cDNA sequence, except that it also contained a 129-bp insertion predicted to encode 43 additional amino acids. We will refer to this cDNA, and the corresponding predicted protein, as FR-19A. The FR-19A insertion has ϳ84, 70, and 44% amino acid sequence identity to the corresponding regions of chicken TEF-1A, mouse TEF-1, and mouse ETF, respectively (Fig. 4B). The deduced FR-19A amino acid sequence within this insertion differs from the TEFR1a insertion sequence reported by Yockey et al. (42) at residue 142, where our cDNA encodes an alanine while the TEFR1a cDNA encodes a leucine. The FR-19A protein isoform would be 427 amino acid residues in length with a predicted molecular mass of 47,968 daltons and an estimated isoelectric point of 8.11.
FR-19 DNA Binding Activity in Vitro-Human TEF-1 (36), mouse ETF (48), and chicken TEF-1 (40) can bind the GT-IIC motif found in the SV40 enhancer sequence. We used gel mobility shift analysis to examine whether FR-19 could also bind this sequence motif. First, either the plasmid vector pRSET or this same vector containing the FR-19B cDNA insert (pRSET/ FR-19B) were used as DNA templates in a coupled in vitro transcription/translation reaction system. Translation products were analyzed by SDS-polyacrylamide gel electrophoresis and autoradiography. A major translation product of ϳ52 kDa in size was synthesized from pRSET/FR-19B, consistent with the predicted size of a protein derived from this construct (Fig.  5A). An aliquot of each lysate was then incubated with a radiolabeled double-stranded GT-IIC oligonucleotide probe and DNA-protein complexes were analyzed by electrophoresis on a non-denaturing polyacrylamide gel and autoradiography. In vitro translated FR-19B formed a specific complex with the GT-IIC-containing oligonucleotide probe (Fig. 5B). This binding was sequence-specific, since it could be competed by unlabeled wild-type GT-IIC oligonucleotides but not by unlabeled mutant GT-IIC oligonucleotides.
FR-19 Gene Localization-The mouse chromosomal location of Fr19 was determined by interspecific backcross analysis using progeny derived from matings of C57BL/6J ϫ M. spretus F 1 X C57BL/6J mice. This interspecific backcross mapping panel has been typed for over 2000 loci that are well distributed among all the autosomes as well as the X chromosome (49). C57BL/6J and M. spretus DNAs were digested with several enzymes and analyzed by Southern blot hybridization for in- formative restriction fragment length polymorphisms using an FR-19 cDNA probe. The 13.0-and 8.2-kb EcoRI M. spretus restriction fragment length polymorphisms (see "Experimental Procedures") were used to follow the segregation of the Fr19 locus in backcross mice. The mapping results indicated that Fr19 is located in the distal region of mouse chromosome 6 linked to Glut3, Fgf6, and Kras2 (Fig. 6). Although 90 mice were analyzed for every marker and are shown in the segregation analysis, up to 183 mice were typed for some pairs of markers. Each locus was analyzed in pairwise combinations for recombination frequencies using the additional data. The ratios of the total number of mice exhibiting recombinant chromosomes to the total number of mice analyzed for each pair of loci and the most likely gene order are: centromere -Glut3 -4/152 -Fgf6 -0/183 -Fr19 -13/128 -Kras2. The recombination frequencies (expressed as genetic distances in centimorgans Ϯ S.E.) are: No recombinants were detected between Fr19 and Fgf6 in 183 animals typed in common, suggesting that the two loci are within 1.6 centimorgans (ϳ2600 kb) of each other (upper 95% confidence limit).
FR-19 mRNA Levels in Mouse Tissues-We next used Northern blot hybridization analysis to examine the tissue distribution of FR-19 mRNA. Six different tissues were obtained from newborn mice and 12 different tissues were obtained from adult mice. In the newborn animals, FR-19 transcripts were expressed at a relatively low level in the heart, intestine, kidney, and liver but at an intermediate level in skin and a high level in lung (Fig. 7A). FR-19 mRNA was also expressed at the highest level in adult lung (Fig. 7B). It was expressed at an intermediate level in skeletal muscle and skin but was undetectable or expressed at a low level in all of the other adult tissues examined. It should be noted that in the newborn tissues, and in some of the adult tissues as well, a relatively small FR-19 transcript (ϳ1.2 kb in size) was expressed. This RNA may not encode an FR-19 isoform, since it does not appear to be polyadenylated (data not shown).
Regulation of FR-19 mRNA Expression in C2C12 Cells-The FR-19 gene is expressed in skeletal muscle tissue (see above); furthermore, previous studies have demonstrated that mouse (41) and chicken (40) TEF-1 can bind to the M-CAT motif, a cis-acting element that is almost identical to the GT-IIC motif and is found in the promoter/enhancer regions of several muscle-specific genes. Therefore, we determined whether FR-19 mRNA levels were regulated during myogenic differentiation in vitro. These experiments were performed using the murine C2C12 skeletal muscle cell line (58). When proliferating C2C12 cells are switched to a growth factor-poor culture medium, they differentiate into multinucleated myotubes and express muscle-specific structural proteins and transcription factors (59,60). To examine FR-19 mRNA expression levels during C2C12 differentiation, cells were first grown to subconfluence in standard growth medium and then either immediately harvested or placed in differentiation medium and then collected at various times thereafter. We also treated NIH 3T3 fibroblasts in an identical manner in order to examine FR-19 mRNA The FGFR-1 primer pair generated a single amplification product of the correct size; therefore, only this region of the gel is presented. The arrowheads denote the two DNA fragments that were recovered, cloned, and sequenced. B, the deduced amino acid sequence for the 43-amino acid insertion predicted to be present in the FR-19A isoform is aligned with FR-19B and the corresponding regions of the same three TEF-1-related proteins described in the Fig. 3 legend. Numbers to the right refer to the last FR-19A or FR-19B amino acids on the lines. Amino acids that are identical in all polypeptides at a given position are boxed. Gaps represented by dashes were inserted to maximize the sequence identity.

FIG. 5. Gel mobility shift analysis using in vitro translated FR-19B.
A, in vitro transcription/translation reactions were performed using either no added DNA template or equivalent amounts of pRSET or pRSET/FR-19B plasmid DNA. Translation products were separated by SDS-polyacrylamide gel electrophoresis and visualized by autoradiography. Molecular masses (in kDa) of protein size standards are shown on the left. B, rabbit reticulocyte lysate was incubated with 32 P-labeled wild-type GT-IIC oligonucleotide probe and DNA-protein complexes analyzed by non-denaturing polyacrylamide gel electrophoresis and autoradiography. In some cases, unlabeled wild-type (wt) or mutant (mt) GT-IIC oligonucleotides were added to the binding reaction. Arrows indicate the positions of the specific complex (C) that was detected and the free probe (F). expression in a non-myogenic cell line that had been subjected to similar growth arrest conditions. RNA was isolated and Northern blot hybridization analysis was performed using several different cDNA probes. The results indicated that FR-19 mRNA levels increased during C2C12 myogenic differentiation, with a maximal level of expression detected at day 3 (Fig.  8A). In contrast, FR-19 mRNA levels decreased when NIH 3T3 fibroblasts were switched to differentiation medium (Fig. 8B). We confirmed that C2C12 myoblast differentiation occurred under our culture conditions by examining cellular morphology (data not shown) and by monitoring both myogenin and muscle creatine kinase mRNA levels, which increased as expected (59,60). We also probed the blots for glyceraldehyde-3-phosphate dehydrogenase mRNA, which is expressed at a similar level in both C2C12 myoblasts and myotubes (60), to demonstrate that equivalent amounts of RNA were present in each gel lane. Glyceraldehyde-3-phosphate dehydrogenase mRNA levels remained constant in C2C12 cells but decreased when NIH 3T3 cells were switched to differentiation medium; however, this did not reflect unequal RNA loading, as estimated by rRNA fluorescence intensity.
We also performed RT-PCR assays to determine the relative abundance of the FR-19A and FR-19B splice variant transcripts in C2C12 cells. Equivalent amounts of RNA isolated from either FGF-1-stimulated NIH 3T3 fibroblasts, C2C12 cells cultured in normal growth medium (myoblasts) or C2C12 cells collected after 5 days in differentiation medium (myotubes) was converted to cDNA and PCR was performed using the same pair of FR-19 primers used for the experiment described in Fig.  4A. In agreement with the Northern blot data, the RT-PCR results indicated that FR-19 mRNA levels were higher in myotubes than in myoblasts (Fig. 9). Furthermore, although NIH 3T3 fibroblasts express predominantly FR-19A mRNA, C2C12 cells (myoblasts or myotubes) express predominantly FR-19B mRNA.

DISCUSSION
The binding of polypeptide growth factors to cell surface receptors triggers a cascade of intracellular biochemical responses, including the transcriptional activation of specific nuclear genes (61,62). It has been demonstrated that at least some of the proteins encoded by mitogen-inducible genes perform functions necessary for cell cycle progression (63)(64)(65). Thus, we have begun to identify and characterize FGF-1-inducible genes in NIH 3T3 fibroblasts in order to gain further insight into mammalian cell growth control. Here, we report that the delayed-early response gene FR-19 encodes a protein with ϳ89, 77, and 68% overall sequence identity to chicken TEF-1A (40), human (36) or mouse (41,56,57) TEF-1, and mouse ETF (48), respectively. These proteins, in combination with Drosophila Scalloped (66), Saccharomyces TEC1 (67) and Aspergillus AbaA (68), define a family of transcription factors that share a highly conserved DNA-binding motif, the TEA domain (35). After the original submission of this manuscript, Yockey et al. (42) reported the isolation and characterization of murine cDNA clones predicted to encode two TEF-1-related protein isoforms, which they called TEFR1a and TEFR1b. These cDNAs were isolated by low stringency screening of an adult mouse cardiac cDNA library with a human TEF-1 cDNA probe. The FR-19 and TEFR1 cDNAs have Ͼ99% nucleotide sequence identity and encode identical proteins, except for a single amino acid difference in the 43-amino acid insertion found in FR-19A (TEFR1a).
The initial FR-19 cDNA fragment was identified using an RT-PCR approach that is conceptually similar to the mRNA differential display and fingerprinting methods used by others to identify differentially expressed genes (69). In our method, (i) cDNA is synthesized using random primers, (ii) PCR assays are performed using pairs of oligonucleotide primers designed to amplify cDNA templates encoding proteins with particular structural motifs, and (iii) amplification products are displayed using agarose gel electrophoresis and ethidium bromide staining. We have found that ϳ50% of the genes identified using this method are in fact FGF-1-regulated; however, cDNA sequence analysis has indicated that most of the genes do not encode proteins with the targeted structural motif(s). In most instances the primers, many of which were degenerate, annealed to regions of low sequence identity. In the case of FR-19, PCR was performed with a sense SH2 domain primer and a degen-erate antisense leucine zipper domain primer; however, only the former primer participated in FR-19 cDNA amplification. In addition, under the PCR conditions used, the SH2 domain primer annealed to two regions of minimal sequence identity, one of which was actually in the 3Ј-untranslated region of the FR-19 cDNA template. It is possible that we can increase the likelihood of successfully identifying certain classes of growthregulated genes by re-designing some of our primers and/or by changing the PCR conditions.
The predicted FR-19 amino acid sequence has the greatest identity (ϳ89%) to chicken TEF-1 (40); thus, the FR-19 gene may be the murine homologue of this chicken locus. However, the two genes do not have similar tissue expression patterns; the chicken TEF-1 gene, but not the mouse FR-19 gene, is expressed at a relatively high level in liver and heart (40). The deduced FR-19 and chicken TEF-1 proteins have a similar degree of sequence identity, ϳ77%, to human (36) or mouse (41,56,57) TEF-1. It appears that at least three chicken TEF-1 isoforms may be expressed from a single gene (40). In comparison to chicken TEF-1A, the TEF-1B and TEF-1C isoforms contain insertions of 13 and 3 amino acid residues, respectively, in two distinct regions that are both carboxyl-terminal to the TEA domain. Both chicken TEF-1A and TEF-1B can bind the GT-IIC motif recognized by human TEF-1 (36), as well as the M-CAT motif, a cis-acting element of similar sequence which is found in several muscle-specific gene promoters (40). We have found that at least two FR-19 mRNA splice variants are expressed in murine fibroblasts and skeletal muscle cells, in agreement with the recent report by Yockey et al. (42). The 43-amino acid insertion predicted to be present in the FR-19A isoform is located at the same position as the 13-amino acid insertion in chicken TEF-1B as well as a 4-amino acid insertion found in the mouse TEF-1 sequence reported by Shimizu et al. (41). Taken together, these results indicate that distinct chicken TEF-1, mouse TEF-1, and mouse FR-19 isoforms may be expressed via alternative splicing of their respective primary transcripts. These isoforms may have unique functional properties; indeed, chicken TEF-1B, but not TEF-1A, can transactivate a heterologous reporter gene (40). Furthermore, Yockey et al. (42) found that mouse TEFR1b (FR-19B) is a stronger transactivator than TEFR1a (FR-19A).
Our experimental results, in combination with previous reports, indicate that there are at least three TEF-1-related genes in the mouse genome: FR-19, TEF-1 (41,56,57), and ETF (48). The FR-19 locus is located on chromosome 6 but the positions of TEF-1 and ETF have not yet been reported. We have compared our interspecific map of chromosome 6 with a composite mouse linkage map that reports the map location of many uncloned mouse mutations (provided from Mouse Genome Database, a computerized data base maintained at The Jackson Laboratory, Bar Harbor, ME). Fr19 mapped in a region of the composite map that lacks mouse mutations with a phenotype that might be expected for an alteration in this locus (data not shown). The distal region of mouse chromosome 6 shares a region of homology with human chromosome 12p; in particular, Fgf6 has been placed on human 12p13. The tight linkage between Fgf6 and Fr19 in mouse suggests that Fr19 will reside on 12p13 as well.
The amino acid sequence identity between FR-19 and either TEF-1 or ETF is ϳ77 and 68%, respectively. TEF-1 and ETF have ϳ68% amino acid sequence identity. The predicted mouse and human TEF-1 proteins have ϳ99% amino acid sequence identity and thus are encoded by homologous genes. Mouse TEF-1 can bind to the A element (which includes an M-CAT motif) found in the rabbit myosin heavy chain-␤ promoter (41,42). This finding, in combination with a report demonstrating FIG. 9. RT-PCR analysis of FR-19A and FR-19B mRNA expression in C2C12 cells. RNA isolated from FGF-1-stimulated (8 h) NIH 3T3 cells, C2C12 myoblasts (Mb), or C2C12 myotubes (Mt) was converted to cDNA and PCR performed using FR-19 sense and antisense oligonucleotide primers. PCR was also performed in the absence of cDNA template (ϪT, minus template). Amplification products were separated by agarose gel electrophoresis and visualized by ethidium bromide staining. DNA size markers (M; in bp) are shown on the left. that disruption of the mouse TEF-1 gene results in heart defects and embryonic lethality (57), suggests that mouse TEF-1 may primarily be involved in cardiac-specific regulation of M-CAT-dependent promoters.
ETF, the most divergent member of the vertebrate TEF-1 family, may be involved in transcriptional control during neural development (48). The ETF gene is regulated during mouse embryogenesis and is primarily expressed in the hindbrain region of 10-day-old embryos (48). Mouse ETF (48), like human TEF-1 (36) as well as chicken TEF-1A and TEF-1B (40), has been shown to bind GT-IIC sequence motifs in vitro. All four of these proteins, as well as mouse TEF-1 (41,56,57), have an identical TEA domain sequence. Previous studies have indicated that the TEA domain found in human TEF-1 is necessary and sufficient for specific DNA binding (36,37). In comparison to the TEA domain of these other TEF-1-related proteins, the FR-19 TEA domain contains five amino acid substitutions. However, three of these substitutions are conservative changes and one other substitution, where a tyrosine residue at position 77 is replaced by a histidine, is located in a region that is not critical for DNA-binding (37). Accordingly, it is likely that FR-19A and FR-19B are also DNA-binding proteins that recognize similar sequence motifs. In support of this possibility, we have found that in vitro translated FR-19B can bind to the GT-IIC motif in gel mobility shift assays. Additionally, Yockey et al. (42) have demonstrated that both TEFR1a (FR-19A) and TEFR1b (FR-19B) can form specific complexes with the M-CAT motif in vitro.
The mouse FR-19, TEF-1, and ETF genes are differentially expressed in adult tissues; thus, they are likely to have unique physiological functions in vivo. We observed that FR-19 mRNA was expressed at the highest levels in lung, in agreement with the results of Yockey et al. (42). TEF-1 mRNA is more widely expressed, with highest levels present in skeletal and cardiac muscle, kidney and lung (41). In contrast, the ETF gene is not expressed to any significant degree in those adult mouse tissues that have been examined to date (brain, testis, liver, kidney, thymus, and heart) (48).
In regard to mouse FR-19, TEF-1, or ETF gene regulation in tissue culture cells, we found that several mitogenic agents and PMA, a tumor-promoting phorbol ester, could induce FR-19 mRNA expression in NIH 3T3 fibroblasts. In FGF-1-stimulated cells, FR-19 mRNA levels are significantly increased at 4 h while in calf serum-stimulated cells, elevated FR-19 mRNA levels are detected earlier, at 2 h post-stimulation. This difference could reflect a more rapid cell cycle transit time following serum treatment; alternatively, serum may contain factors that alter FR-19 gene expression by a mechanism distinct from that used by FGF-1. FGF-1-induced FR-19 mRNA accumulation does not occur in the presence of actinomycin D or cycloheximide. This indicates that FGF-1 treatment promotes the synthesis of a transcription factor required for FR-19 gene activation; however, we have not ruled out the possibility that the actinomycin D is preventing the synthesis of a labile protein that stabilizes pre-existing FR-19 mRNA. Although it has been demonstrated that the mouse TEF-1 gene is expressed in Balb/c 3T3 fibroblasts cultured in vitro (41,42), there have been no reports describing the regulation of TEF-1 expression in fibroblast cell lines.
We have also demonstrated by Northern blot hybridization and RT-PCR analysis that FR-19 mRNA levels increase during C2C12 skeletal muscle cell differentiation in vitro. This finding confirms and extends the data of Yockey et al. (42), which demonstrated that TEFR1 mRNA expression was higher in Sol8 and C2C12 myotubes than in myoblasts. Shimizu et al. (41) have reported that C2C12 myoblasts and myotubes ex-press a similar level of TEF-1 mRNA; thus, the mouse TEF-1 gene does not appear to be regulated during myogenic differentiation. The finding that the FR-19 gene is up-regulated during both growth stimulation (fibroblasts) and growth arrest (myoblasts) is intriguing and suggest that a single TEF-1related protein may have distinct cell type-specific functions. It is possible that the FR-19 protein expressed during muscle differentiation is involved in the up-regulation of M-CAT-dependent genes. However, in the case of at least one such gene, skeletal ␣-actin, increased mRNA expression appears to precede FR-19 mRNA accumulation (70). In fibroblasts, FR-19 could function to activate specific genes encoding proteins required for cell cycle progression. The promoter regions of these cellular genes may contain M-CAT motifs or presently undefined sequence motifs that can also be recognized by  Additional studies that manipulate FR-19 expression levels or inhibit FR-19 function should help to define the role of FR-19 in fibroblast proliferation and myoblast differentiation.