Csm, a cardiac-specific isoform of the RNA helicase Mov10l1, is regulated by Nkx2.5 in embryonic heart.

Nkx2.5 (also called Csx) is an evolutionarily conserved cardiac transcription factor of the homeobox gene family. Nkx2.5 is required for early heart development, because Nkx2.5 null mice die before completion of cardiac looping. To identify genes regulated by Nkx2.5 in the developing heart, we performed differential screening in combination with suppression subtractive hybridization using RNA isolated from wild-type and Nkx2.5 null hearts at embryonic day 8.5. One gene that we found to be markedly down-regulated in the hearts from Nkx2.5 null embryos is an isoform of Mov10 like-1 (Mov10l1), a putative RNA helicase expressed in testis. We named this novel isoform as Csm (cardiac-specific isoform of Mov10l1). Csm is identical with the 3' region of the Mov10l1 gene, but its transcript starts from the exon 16 of Mov10l1. The conceptual protein encoded by Csm cDNA contains a helicase motif as well as ATPase and RNA interaction motifs. Csm is expressed specifically in the heart, and its expression in the heart is restricted to cardiac myocytes. Csm potentiated phenylephrine-induced hypertrophic response in cardiac myocytes. Furthermore, transient cotransfection analysis showed that Nkx2.5 transactivates the Csm promoter, suggesting that Nkx2.5 is essential for embryonic Csm expression.

Nkx2.5 (also called Csx) is an evolutionarily conserved cardiac transcription factor of the homeobox gene family. Nkx2.5 is required for early heart development, because Nkx2.5 null mice die before completion of cardiac looping. To identify genes regulated by Nkx2.5 in the developing heart, we performed differential screening in combination with suppression subtractive hybridization using RNA isolated from wild-type and Nkx2.5 null hearts at embryonic day 8.5. One gene that we found to be markedly down-regulated in the hearts from Nkx2.5 null embryos is an isoform of Mov10 like-1 (Mov10l1), a putative RNA helicase expressed in testis. We named this novel isoform as Csm (cardiac-specific isoform of Mov10l1). Csm is identical with the 3 region of the Mov10l1 gene, but its transcript starts from the exon 16 of Mov10l1. The conceptual protein encoded by Csm cDNA contains a helicase motif as well as ATPase and RNA interaction motifs. Csm is expressed specifically in the heart, and its expression in the heart is restricted to cardiac myocytes. Csm potentiated phenylephrineinduced hypertrophic response in cardiac myocytes. Furthermore, transient cotransfection analysis showed that Nkx2.5 transactivates the Csm promoter, suggesting that Nkx2.5 is essential for embryonic Csm expression.
Nkx2.5 (also referred to as Csx) is a member of the NK2 class of homeoproteins (1,2). NK2 class homeoproteins are expressed in a tissue-specific manner, suggesting a role in tissue specification, differentiation, and patterning (3). Murine Nkx2.5 was identified as a homologue of the Drosophila tinman gene (4), which is required for subdivision of the dorsal mesoderm into the cardiac and visceral mesoderm (1,2). In mouse, expression of Nkx2.5 begins as early as embryonic day (ED) 1 7.5 and continues through adulthood (1,2,5). Nkx2.5 null mice die before ED 11 with an arrest of cardiac development before the completion of looping (6,7).
Recently, heterozygous mutations of human NKX2.5 were identified in patients with several types of congenital heart defects (8,9). The most common phenotype was progressive atrioventricular conduction delays and secundum atrial septal defect, but other anatomical abnormalities, such as ventricular septal defect, tetralogy of Fallot, or tricuspid valve abnormalities, including Ebstein's anomaly, and progressive left ventricular failure, were also found (8,9). These findings suggest that two copies of the NKX2.5 allele are necessary for normal heart formation.
We and others have previously reported several genes downregulated in the Nkx2.5-null heart, including atrial natriuretic factor (ANF), B-type natriuretic peptide, CARP, eHAND, MEF2C, myosin light chain 2v, SM22, N-myc, Msx2, Irx4, and Chisel (6, 7, 10 -15). To identify other genes regulated by Nkx2.5 in the developing heart, we performed suppression subtractive hybridization using RNA from hearts of wild-type and Nkx2.5 null mice at ED 8.5. We isolated a member of the RNA helicase family, specifically expressed in the murine heart. The protein encoded by this gene, named Csm, cardiacspecific isoform of Mov10 like-1 (Mov10l1), was markedly downregulated in the Nkx2.5 null heart. In addition, Csm potentiated phenylephrine-induced hypertrophic response in cardiac myocytes.

RNA Preparation, cDNA Synthesis, and Subtractive Hybridization-
Total RNA (150 ng) from hearts of wild-type and Nkx2.5 null embryos at ED 8.5 was treated with RNase-free DNase I and used to generate cDNA using the SMART cDNA synthesis kit (Clontech, Palo Alto, CA). Subtractive hybridization was performed using a PCR-Select cDNA subtraction kit (Clontech). Briefly, two pools of RsaI-digested wild-type heart cDNA were used as tester and ligated to unique adapters. RsaIdigested Nkx2.5 null heart cDNA was used as driver without adapters. Hybridization of the tester population with excess driver and amplification of subtracted species with the two unique adapters in the tester cDNA produced a pool of PCR-amplified fragments theoretically present in wild-type heart cDNA, but not Nkx2.5 null heart cDNA (forward subtraction). Reverse subtraction was also performed using Nkx2.5 null heart cDNA as tester and wild-type heart cDNA as driver to enrich for cDNA-representing transcripts highly expressed in Nkx2.5 null hearts.
Differential Screening-Subtracted PCR fragments were subcloned, and 464 clones were recovered. cDNA inserts of the plasmid were amplified by PCR and were arrayed in duplicate onto nylon membranes. cDNAs from the forward or reverse subtractions were radiola-beled with [ 32 P]dCTP by random priming. Membranes were hybridized to radiolabeled probes at 42°C for 16 h in the presence of 50% formamide. After hybridization, the membranes were washed at 65°C in the presence of 0.1ϫ SSC and 0.1% SDS. Visualization was achieved by exposure to Kodak Biomax MS film (Eastman Kodak, Rochester, NY).
Cloning of Csm cDNA and the Promoter Region-cDNA clones representing Csm were isolated from ED 10 mouse heart cDNA libraries (Stratagene, La Jolla, CA) employing standard methodology. 5Ј-RNA ligase-mediated (RLM)-RACE was performed according to the manufacturer's protocol (Ambion Inc., Austin, TX). Briefly, RNase-free DNase I-treated total RNA (10 g) was treated with calf intestinal phosphatase to remove free 5Ј-phosphates from ribosomal RNA, fragmented RNA, and tRNA. The RNA was then treated with tobacco acid pyrophosphatase to remove the cap structure from full-length mRNA, leaving a 5Ј-monophosphate. The RNA adapter oligonucleotide was ligated to the RNA population using T4 RNA ligase. A random-primed reverse transcription reaction and nested PCR then amplified the 5Јend of a transcript. A 2061-bp fragment of the Csm promoter (positions Ϫ2061 to Ϫ1 in the Csm gene) was isolated from BAC clone (RP23-269G24, Research Genetics, Huntsville, AL).
Cardiac Myocytes Culture-Mouse neonatal cardiac myocytes culture from 1-day-old C57B6 mice was prepared as previously described in rat with a slight modification (16). Briefly, ventricular myocytes were dissociated enzymatically and preplated for 30 min twice to enrich for myocytes. Cells were plated onto 100-mm culture dishes and cultured in cardiac myocyte culture media that contained Dulbecco's modified Eagle's medium/F-12 supplemented with 5% newborn calf serum, 4 g/ml transferrin, 0.7 ng/ml sodium selenite (Invitrogen), 2 g/liter bovine serum albumin (fraction V), 3 mM pyruvic acid, 15 mM HEPES (pH 7.6), 100 M ascorbic acid, 100 g/ml ampicillin, and 5 g/ml linoleic acid. Culture media were changed after 24 h. Myocytes were harvested after 24 h.
Rat neonatal cardiac myocytes, cultured from 1-day-old Wister rats, were prepared as described above with modification. Briefly, ventricles were digested enzymatically, and myocytes were purified over a Percoll gradient. Culture medium was changed to serum-free at 24 h. Myocytes were cultured under serum-free conditions for 24 h before experiments.
Primer Extension Analysis-One microgram of mouse heart mRNA and 1 ϫ 10 5 cpm end-labeled oligonucleotide (antisense for mouse Mov10l1-(2303-2326), 5Ј-AAGGATATATGGGAGAGGCCGGCA-3Ј) were hybridized in hybridization buffer (150 mM KCl, 10 mM Tris, pH 8.3, 1 mM EDTA) at 65°C for 90 min and cooled slowly to room temperature. The sample was mixed with reaction mix (20 mM Tris, pH 8.3, 5.5 mM dithiothreitol, 10 mM MgCl 2 , 150 M dNTP, 150 g/ml actinomycin D, 40 units of SuperScript II reverse transcriptase (RT) (Invitrogen) and then incubated at 42°C for 60 min. The sample was treated with 14 g/ml RNase A at 37°C for 15 min and extracted with phenol/ chloroform. The ethanol-precipitated sample was resuspended with formamide loading buffer. After heating at 65°C for 5 min, the sample was loaded onto an 6% acrylamide/8 M urea gel. After electrophoresis, the gel was dried, and exposures were made to film.
Transfection and Reporter Assay-COS cells cultured in 6-well plates were cotransfected with 3 g of Luciferase reporter construct and 1 g of pTK␤-gal with or without 1 g of Nkx2.5 or Nkx2.5(I183P) by using the calcium phosphate method. Six hours after transfection, cells were washed with PBS, and the medium was changed. Cells were cultured for another 48 h.
Cardiac myocytes were transfected with 3 g of ANF luciferase reporter construct (Ϫ638 ANF Luc, kindly provided by Dr. Kenneth R. Chien, University of California, San Diego, CA) and 1 g of pTK␤-gal with or without 1 g of Csm by using the calcium phosphate method. Two hours after transfection, cells were washed with PBS, and the medium was changed. Cardiac myocytes were cultured for another 24 h and then stimulated with or without phenylephrine (PE, 100 M) for 24 h. Cells were lysed with 200 l of reporter lysis buffer (Promega) and assayed for luciferase activity (by a Promega assay) and ␤-galactosidase activity. Luciferase activity was normalized against ␤-galactosidase activity.
Electrophoretic Mobility Shift Assay-Nkx2.5 protein was prepared by coupled in vitro transcription/translation of a T7-driven Nkx2.5 plasmid in reticulocyte lysate by using a TNT kit (Promega). Labeled DNA probes were incubated with 3 l of programmed lysate in 2 g of bovine serum albumin, 2 g of poly(dI-dC) in 10 mM HEPES, pH 7.9, 50 mM KCl, 1 mM EGTA, 10% glycerol, 2.5 mM dithiothreitol, 7 mM MgCl 2 in a 20-l reaction volume for 30 min at room temperature and separated in 4% polyacrylamide gel with Tris-glycine buffer. All probes were double-stranded and radiolabeled with [␥-32 P]ATP. The following are single-stranded sequences of the probes (putative DNA binding site is underlined, and boldface letters indicate mutated nucleotides): NKE1, GCTGTCTTTCCACTTGAAATGCAGTGG; NKE1M, GCTGTCTTTC-ACAGGTCAATGCAGTGG.
Replication-defective Recombinant Adenovirus and Gene Transfer-HA-tagged Csm cDNA and LacZ cDNA was used to generate recombinant adenovirus expressing Csm (AdCsm) and LacZ (AdLacZ) by using the Adeno-X Expression System (Clontech), respectively. Twenty-four hours after seeding, cardiac myocytes were infected with AdCsm or AdLacZ diluted in the culture media at the multiplicity of infection specified in the text and incubated for 2 h. The viral suspension was removed, and cardiac myocytes were cultured with the serum-depleted culture media.
Amino Acid Incorporation into Proteins-The relative amount of protein synthesis was determined by assessing the incorporation of the radioactivity into a trichloroacetic acid-insoluble fraction as described previously (17). After 24 h in the serum-depleted culture media, cardiac myocytes were stimulated with 100 M PE for 24 h. 0.5 Ci/ml [ 3 H]leucine was added 2 h before harvesting. Cells were quickly rinsed twice with ice-cold phosphate-buffered saline and incubated for 30 min on ice with 5% trichloroacetic acid. After washed twice with ice-cold 5% trichloroacetic acid, cells were solubilized in 0.1 N NaOH. Total trichloroacetic acid-insoluble radioactivity was determined by liquid scintillation counting.
Measurements of Cell Size-Cardiac myocytes were infected with AdCsm or AdLacZ, and 24 h later cells were stimulated with 100 M PE for 48 h. After fixation with 4% paraformaldehyde, cells were stained with TRITC-conjugated phalloidin (Sigma, St. Louis, MO), anti-HAfluorescein isothiocyanate (Roche Applied Science, Indianapolis, IN), and Hoechst 33258 (Sigma). Cells were imaged using a confocal microscope and measured using Image (National Institutes of Health). At least 100 cells were measured per sample.

RESULTS
Identification of Nkx2.5-dependent Gene Csm, Cardiac-specific Isoform of Mov10l1-To identify potential target genes regulated by Nkx2.5 in the developing heart, we performed differential screening in combination with suppression subtractive hybridization using RNA from hearts of wild-type and Nkx2.5 null mice at ED 8.5. Out of 17 clones examined, 9 clones were significantly down-regulated in Nkx2.5 null heart by Northern blot analysis. We chose to focus on a clone containing the 3Ј region of Mov10l1 (18), a putative RNA helicase specifically expressed in testis. We name this gene Csm for cardiacspecific isoform of Mov10l1. Csm was not detectable in Nkx2.5 null heart at ED 9.5 and 10.5 (Fig. 1), indicating that the expression of Csm is strongly dependent on Nkx2.5 in the embryonic heart. Northern blotting of RNA from embryonic hearts as well as from 8 different organs of adult mice revealed the Csm transcript of 1.7 kb in the embryonic and adult hearts ( Fig. 2A). A larger 4.4-kb transcript was detected only in testis, which corresponds to the Mov10l1 transcript as reported previously (18). Csm was abundantly expressed in the mouse heart at a level equivalent to Mov10l1 in the testis. When cardiac myocytes were separated from non-myocytes using primary culture, Csm was found to be expressed in myocytes but not in non-myocytes (Fig. 2B).
Using the partial cDNA fragment as a probe, we isolated a near full-length cDNA clone from a mouse 10-day embryonic heart cDNA library. 5Ј RLM-RACE yielded an additional 30 bp of the 5Ј region. Therefore, the full-length Csm transcript, including 3Ј-polyadenylation consensus sequence, is 1695 bp (Fig. 3). The conceptual protein encoded by the Csm cDNA contains a helicase motif (Fig. 3, boxed) as well as an ATPase (underlined) and RNA interaction motif (dashed line) characteristic of the RNA helicase family (19). Examination of a mouse genomic data base revealed that Csm is identical with the 3Ј region of Mov10l1. Mov10l1 consists of 26 exons, and Csm transcription starts within the exon 16 of Mov10l1 (Fig. 4A).
Interestingly, during the course of this study, Liu et al. (20) reported CHAMP (AF340211), a putative RNA helicase downregulated in MEF2C null mutant hearts. Comparison of the sequences showed that Csm and CHAMP are identical except that CHAMP cDNA is about 300 bp longer than Csm cDNA at the 5Ј-end of the transcript. Transcription of CHAMP starts at exon 14 of Mov10l1 (Fig. 4A). To further determine whether the 5Ј coding sequence of CHAMP is present in Csm transcript, we generated four probes corresponding to exons 14, 15, 16, and 17-20 of Mov10l1 from the testis cDNA and performed Northern blot analysis using mRNAs isolated from the heart and testis. The probes for exon 14 or 15 did not hybridize with RNA from the heart, whereas they hybridized with RNA from testis at 4.4 kb (Fig. 5A). When the entire exon 16 was used as a probe, the 1.7-kb Csm transcript was detected in the heart, whereas the 4.4-kb Mov10l1 transcript was detected in the testis. The hybridization signal of Csm was increased when exons 17-20 were used as a probe, suggesting that the exon 16 (2202-2381) probe did not fully hybridize with Csm mRNA. This result is consistent with 5Ј RLM-RACE showing that Csm transcription starts at the 2252-bp site within the exon 16 of Mov10l1 (Fig. 4B, arrow).
To further prove the 5Ј-end of Csm, two 50-base oligonucleotides, corresponding to antisense for 2202-2251 (which is upstream of the start site that we identified) and 2252-2301 FIG. 1. Csm mRNA expression in Csx/Nkx2.5 null heart. Csm mRNA expression was detected by Northern blot analysis. Blots were made with total RNA (1 g) isolated from hearts of wild-type embryo (ϩ/ϩ) and Nkx2.5 null embryo (Ϫ/Ϫ) at ED 9.5 and 10.5. Csm expression was not detected in Nkx2.5 null heart. A glyceraldehyde-3-phosphate dehydrogenase probe and 28 S RNA were used as a control for assessing RNA loading. The experiments shown represent one of three independent trials, which gave nearly identical results.

FIG. 2. Northern blot analysis of Csm expression in murine tissues and cells.
A, blots were made with mRNA (1 g) isolated from hearts of embryos at ED 10.5 and tissues of an 8-week-old mouse. A band representing 1.7 kb was observed in both embryo and adult heart (arrow). Another band was observed in testis that expressed the 4.4-kb transcript (arrowhead). RNA marker (Mr) was loaded for assessing RNA size. A cyclophilin (Cyph) probe was used as a control for assessing RNA loading. B, blots were made with total RNA (10 g) isolated from mouse cultured neonatal cardiac myocytes (CM) and non-myocytes (NM). A band representing 1.7 kb was observed in CM (arrow) but not in NM. A glyceraldehyde-3-phosphate dehydrogenase probe was used as a control for assessing RNA loading.
(which would be the first 50 bases of the Csm mRNA) of Mov10l1, were used for Northern blot analysis (Fig. 4B). The antisense probe for 2202-2251 did not hybridize with the Csm transcript in the heart, whereas it detected Mov10l1 in the testis (Fig. 5B, left panel). The antisense probe for 2252-2301 detected both Csm in the heart and Mov10l1 in the testis (Fig.  5B, right panel). Further analysis using primer extension revealed that the 5Ј-ends of the cardiac transcript were ϳ48 and 75 nucleotides upstream from the primer located at position 2326 (Fig. 5C), suggesting that Csm has two initiation sites. The band representing 48 nucleotides upstream from the primer was much stronger, suggesting that it is the major site of initiation. The 75-bp band corresponds to the longest transcript identified by 5Ј RLM-RACE. This result indicates that the longest Csm transcript starts at position 2252 of Mov10l1 within exon 16 and does not contain exons 14 or 15 as does the CHAMP transcript.
Nkx2.5 Transactivated the Csm Promoter-Examination of the mouse genomic DNA sequence (NW_000106) reveals that the putative Csm promoter has no TATA box, but two NKE, the consensus binding motif for Nkx2.5 (TYAAGTG) (21), at Ϫ64 (NKE1) and Ϫ1752 (NKE2) (Fig. 6A). Six additional similar sequences to the consensus motif were found within the putative Csm promoter from Ϫ2000 to Ϫ1. To examine whether Csm expression is regulated by Nkx2.5, we isolated a 2061-bp fragment of the Csm promoter (positions Ϫ2061 to Ϫ1 in the Csm gene) and performed reporter assays using the Csm promoter. Wild-type Nkx2.5 transactivated the Csm promoter by 4-fold (Fig. 6B). On the other hand, a point mutant of Nkx2.5 (I183P; i.e. Ile 183 3 Pro), which fails to bind DNA (22), did not transactivate the Csm promoter (Fig. 6B). Furthermore, we made NKE site mutations in the Csm promoter and examined the role of NKE for Nkx2.5 in the Csm promoter. The mutation of the NKE2, Csm(NKE2M)-Luc, did not reduced responsive- ness to Nkx2.5, whereas the mutation of NKE1, Csm(NKE1M)-Luc, significantly reduced responsiveness to Nkx2.5 (Fig. 6C).
The above result suggested that of the two NKEs within Ϫ2061 bp of the Csm promoter, NKE1 is necessary for the full activation of the Csm promoter by Nkx2.5. To confirm that Nkx2.5 binds to NKE1, an electrophoretic mobility shift assay was performed with in vitro translated Nkx2.5 protein and oligonucleotide probes corresponding to NKE1 and NKE1M. As shown in Fig. 6D, Nkx2.5 bound to NKE1, but not NKE1M. These results indicate that Nkx2.5 directly regulates the Csm gene transcription in vitro, at least in part through NKE1 site.
Csm Potentiates Phenylephrine-induced Hypertrophic Response in Cardiac Myocytes-To examine the function of Csm in cardiac myocytes, we first performed Northern blot analysis to determine whether Csm is involved in the endogenous ANF mRNA expression using AdCsm. As shown in Fig. 7A, Csm alone did not induce ANF mRNA expression. However, when cardiac myocytes were stimulated with ␣-adrenergic agonist PE, Csm potentiated PE-induced ANF mRNA expression. We next performed the reporter assay using the ANF promoter and the Csm expression vector. Csm alone did not affect the basal transcriptional activity of the ANF promoter (Fig. 7B, row 2), whereas Csm potentiated PE-induced ANF promoter activation (compare rows 3 and 4 of Fig. 7A). We also examined whether Csm is involved in PE-induced cardiac myocytes hypertrophy. 4 and 6 of Fig. 7C). A similar result was achieved using a higher multiplicity of infection (Fig. 7C, lanes 8 and 10). We then measured cell size. Csm potentiated PE-induced increase of cell size (Fig. 7D). These results indicate that Csm potentiated PE-induced hypertrophic response in cardiac myocytes.

PE stimulated [ 3 H]leucine incorporation by 50% and Csm potentiated PE-induced [ 3 H]leucine incorporation (compare lanes
Csm Expression in Rat and Human Heart-Generally, essential genes are conserved among different species. In the rat heart, RT-PCR revealed the presence of a transcript using primers spanning mouse Mov10l1-(2540 -2831) (exons 18 -20), which correspond to positions within the coding region of Csm (Fig. 8A). Sequencing of the PCR product showed 99% identity with the mouse sequence. However, when this fragment was used for Northern blotting, it did not detect Csm or Mov10l1 in the rat atria or in ventricles, whereas it readily hybridized with the mouse Csm transcript (Fig. 8B).
In the human heart, a transcript was also detected by RT-PCR using specific primers spanning human MOV10L1-(2543-2850) and -(2526 -3078) (Fig. 8C), which also correspond to positions within the coding region of mouse Csm. Northern blot analysis revealed that a probe for human MOV10L1-(2526 -3078) hybridized with RNA samples from human testis but not with the RNA sample from the human heart (Fig. 8D). An oligonucleotide probe mapping within the putative coding region (antisense to human MOV10L1-(2497-2546)) hybridized with RNA samples from human testis, but it did not hybridize . Blots were made with mRNA (1 g) isolated from hearts (H) and testes (T) of 8-week-old mice. Probes were generated from mouse testis cDNA. Probes for exons 16 and 17-20 were hybridized with RNA from heart and testis. However, probes for exon 14 or exon 15 were hybridized with RNA from testis, but not heart. A cyclophilin (Cyph) probe was used as a control for assessing RNA loading. B, Northern blot analysis using oligonucleotide probes corresponding to the antisense sequence for mouse Mov10l1-(2202-2251) and -(2252-2301). Blots were made with mRNA (1 g) isolated from hearts (H) and testes (T) of an 8-week-old mouse. An oligonucleotide probe corresponding to the antisense sequence for mouse Mov10l1-(2252-2301) was hybridized with RNA from hearts and testis. However, an oligonucleotide probe corresponding to the antisense sequence for mouse Mov10l1-(2202-2251) was hybridized with RNA from testis, but not heart. Arrowheads and arrows indicate Mov10l1 and Csm, respectively. A cyclophilin (Cyph) probe was used as a control for assessing RNA loading. C, primer extension analysis was performed using mouse heart mRNA and an antisense oligonucleotide for mouse Mov10l1-(2303-2326). Bands were seen at 48 and 75 bp. Positions in the nucleotide are given by numbers from mouse Mov10l1 (GenBank TM AF285587).
with the RNA sample from the human heart. These findings indicate that the level of expression of Csm and/or Mov10l1 in rat and human heart is below the detection level of Northern blot analysis and is far below the level of Csm expression in the mouse heart.
Examination of human genomic DNA sequence (AL080347 and NT_019197) reveals that human MOV10L1 has 27 exons and that the putative human Csm transcript starts from the exon 17 of MOV10L1. The putative human Csm promoter in intron 16 of MOV10L1 has one similar sequences of consensus binding motif for Nkx2.5 (CWTAATTG) (21) (data not shown).
Csm mRNA was readily detected in wild-type heart at ED 9.5 and 10.5, whereas it was undetectable in Nkx2.5 null heart. This finding indicates that Csm expression depends on Nkx2.5 expression in embryonic heart, suggesting that Csm is either directly or indirectly downstream of Nkx2.5 at this stage of embryonic development. We found two consensus binding motifs for Nkx2.5 (TYAAGTG) (21) within the putative Csm promoter from Ϫ2061 to Ϫ1. We showed transactivation of the Csm promoter by Nkx2.5 in vitro and reduced responsiveness of mutated NKE1 in the Csm promoter to Nkx2.5. These findings suggest that Nkx2.5 directly regulates Csm expression at the stage of embryonic development.
Csm is identical to CHAMP except the 5Ј-end of CHAMP is ϳ300 nucleotides longer. CHAMP mRNA has been reported to be about 2 kb in Northern blot analysis (20). However, we detected a single 1.7-kb transcript in heart. We demonstrate in Fig. 5 that the 1.7-kb mRNA does not contain exons 14 and 15 of mouse Mov10l1 and that the 5Ј-end of Csm is located within exon 16 of Mov10l1. These findings raise the possibility that Mov101l mRNA and CHAMP mRNA are not expressed in heart. Therefore, we further examined whether Mov10l1 mRNA is expressed in mouse heart by using RT-PCR. A PCR product corresponding to Mov10l1 was detected in the mouse heart cDNA (data not shown). Taken together with the results of Northern blot analysis, this finding suggests that, although Mov10l1 RNA is expressed in mouse heart, the amount is below the detection limit of Northern blot analysis. Similarly, al- though CHAMP mRNA may be expressed in the mouse heart, the amount seems below the detection limit of Northern blot analysis. These results indicate that Csm is the predominant isoform of Mov10l1 in the mouse heart.
Sequence analysis reveals that Csm contains a helicase motif as well as an ATPase and RNA interaction motif. Helicases are grouped into major superfamilies of proteins (SFI and SFII) based of the occurrence of seven conserved motifs (19). Csm belongs to SFI, and is most closely related to the Upf1p, which is involved in RNA metabolism (29). Csm is very similar to CHAMP except in the 5Ј region. Overexpression of CHAMP in primary neonatal cardiac myocytes has been reported to block hypertrophic growth (30). Therefore, we examined the function of Csm on hypertrophic response in cardiac myocytes. The hypertrophic response in neonatal cardiac myocytes is characterized by a series of phenotypic changes such as an increase in cell size, increased protein synthesis, an induction of specific genes such as ANF, and increased organization of contractile proteins into sarcomeric units (31). We examined these hypertrophic responses. We showed here that Csm potentiated ANF mRNA expression, amino acid incorporation, and increase of cell size by PE. In contrary to the previous report about CHAMP (30), Csm has promoting effects on hypertrophic response. CHAMP has another 188 amino acids in the N-terminal region compared with Csm. There are five repeated motifs of the sequence (TRNDXQSITN(V/I)) and two ATPase motifs in this region. Therefore, the different function between CHAMP and Csm on hypertrophic response is likely attributed this region.
Although Csm is relatively abundant in the mouse heart, its expression in rat and human hearts is below the detection level of Northern blot analysis. The putative human Csm promoter in intron 16 of MOV10L1 has one similar sequence of consensus binding motif for Nkx2.5 (CWTAATTG) (21). On the other hand, the putative mouse Csm promoter from Ϫ2000 to Ϫ1 has several as described above. Subtle differences in promoter sequences likely explain the difference in Csm expression between human and mouse. This also raises the possibility that Csm is regulated differently by Nkx2.5 between human and mouse. Therefore, our findings may give a caution for future studies of the NKX2.5-regulated molecular pathway in human analyzed by using mouse models.
In summary, Csm is a putative RNA helicase that is markedly down-regulated in Nkx2.5 null hearts. Csm is the predominant isoform of Mov10l1 in the mouse heart. Csm potentiated PE-induced hypertrophic response in cardiac myocytes. However, in humans Csm and/or MOV10L1 expression in the heart is very low. Csm regulation by Nkx2.5 may be quite different between human and mouse.