Overexpression of myotonic dystrophy kinase in BC3H1 cells induces the skeletal muscle phenotype.

Myotonic muscular dystrophy is an autosomal dominant defect that produces muscle wasting, myotonia, and cardiac conduction abnormalities. The myotonic dystrophy locus codes for a putative serine-threonine protein kinase of unknown function. We report that overexpression of human myotonic dystrophy protein kinase induces the expression of skeletal muscle-specific genes in undifferentiated BC3H1 muscle cells. BC3H1 clones expressing myotonic dystrophy kinase appear equivalent to differentiated cells with respect to expression of myogenin, retinoblastoma tumor supressor gene, M creatine kinase, beta-tropomyosin, and vimentin. In addition, differential display analysis demonstrates that the pattern of gene expression exhibited by myotonic dystrophy kinase-expressing cells is essentially identical to that of differentiated BC3H1 muscle cells. These observations suggest that myotonic dystrophy kinase may function in the myogenic pathway.

Myotonic muscular dystrophy is an autosomal dominant defect that produces muscle wasting, myotonia, and cardiac conduction abnormalities. The myotonic dystrophy locus codes for a putative serine-threonine protein kinase of unknown function. We report that overexpression of human myotonic dystrophy protein kinase induces the expression of skeletal muscle-specific genes in undifferentiated BC 3 H1 muscle cells. BC 3 H1 clones expressing myotonic dystrophy kinase appear equivalent to differentiated cells with respect to expression of myogenin, retinoblastoma tumor supressor gene, M creatine kinase, ␤-tropomyosin, and vimentin. In addition, differential display analysis demonstrates that the pattern of gene expression exhibited by myotonic dystrophy kinase-expressing cells is essentially identical to that of differentiated BC 3 H1 muscle cells. These observations suggest that myotonic dystrophy kinase may function in the myogenic pathway.
Myotonic muscular dystrophy (DM) 1 is an autosomal dominant myopathy of variable onset, producing pleiotropic effects including skeletal muscle weakness and wasting, cardiac conduction abnormalities, frontal baldness, and cataracts (1,2). Severity of the disease correlates with expansion of an unstable CTG trinucleotide repeat within the 3Ј-untranslated region (UTR) of the gene for DM kinase (DMK) (3)(4)(5)(6)(7). The triplet repeats expand in subsequent generations, producing the genetic phenomenon of anticipation; progeny of affected individuals generally display earlier onset and greater severity of symptoms (8,9).
The product of the DM locus has been identified as a putative serine-threonine protein kinase that shares homology with protein kinases A and C. The full-length cDNA codes for a protein product with an apparent molecular mass of 71 kDa (10,11). 2 The 71-kDa gene product appears to be post-translationally modified to produce a second DMK form with an apparent molecular mass of 84 kDa. 2 Structural predictions from the primary sequence suggest that DMK consists of an amino-terminal kinase domain, an ␣-helical coiled-coil domain, and a carboxyl-terminal transmembrane domain (12). The catalytic activity of DM kinase has been verified (13), but in vivo substrates have yet to be identified. In skeletal muscle, DMK appears to be localized to the neuromuscular junction, while in cardiac muscle the kinase is present in the intercalated disc (10,11). DMK mRNA is present in a variety of tissue types, including muscle, brain, and eye. Quantitation of DMK mRNA concentrations from alleles with expanded triplet repeats in affected individuals has proven difficult, leading to conflicting reports as to expression of mutant alleles (14 -16).
The disease state of DM is likely to be related to the aberrant expression of DMK and the subsequent failure of the kinase to act upon its proper substrates or ectopic interaction with other substrates. In an effort to establish whether DMK operates within a known signal transduction pathway, we constructed stably transfected BC 3 H1 muscle cell lines that constitutively overexpress human DMK. DMK-expressing and untransfected cells were cultured in the presence of 20% fetal calf serum; differentiated BC 3 H1 cells were produced by mitogen withdrawal. Total cellular RNA was isolated and subjected to mRNA differential display analysis, a polymerase chain reaction (PCR)-based technique useful in the identification of differentially expressed genes (17). The pattern of gene expression was substantially changed in DMK-expressing clones, and the new pattern of expression was essentially identical to that of differentiated BC 3 H1 cells. Two differentially expressed gene products, consistent with a skeletal muscle phenotype, were cloned by differential display, sequenced, and used as Northern blot probes. Skeletal muscle-specific ␤-tropomyosin mRNA expression was increased, and vimentin mRNA expression was decreased in DMK-expressing clones. Northern blot analysis of other genes implicated in myogenesis showed that myogenin mRNA was induced in both DMK-expressing and differentiated BC 3 H1 cells. Retinoblastoma tumor supressor (Rb) mRNA was present in untransfected cells but absent from DMK-expressing and differentiated cells. Creatine kinase activity assays demonstrated that DMK-expressing clones up-regulate expression of the M creatine kinase isoform. Thus, despite the presence of mitogens, overexpression of DMK in BC 3 H1 cells induces the skeletal muscle phenotype.

EXPERIMENTAL PROCEDURES
Plasmid Constructs-A coding-only DMK cDNA construct was derived from a full-length (flDMK) cDNA (kindly supplied by R. G. Korneluk) by PCR (Perkin-Elmer) using primers that annealed to the first methionine and termination codons. Primers were designed such that EcoRI restriction sites would be added to the termini of the amplified product: forward primer (5Ј-GGCCGAATTCATGTCAGCCGAGGT-GCGG-3Ј), reverse primer (5Ј-CCGGGAATTCTCAGGGAGCGCGGG-CGGC-3Ј). A 1889-base pair PCR product was amplified and ligated into the EcoRI site of the eukaryotic expression vector pcDNA3 (Invitrogen). This clone, designated pcDNA3/coDMK was sequenced (Sequenase 2.0, United States Biochemical Corp.) to verify integrity of the cDNA sequence. 3Ј-UTR sequences were subcloned from the full-length cDNA in a similar fashion. Forward primer (5Ј-GGCCGAATTCGCCGC-CCGCGCTCCCTGA-3Ј), reverse primer (5Ј-CCGGGAATTCTTTTAT-TCGCGAGGGTCG-3Ј). A 724-base pair PCR product was amplified and ligated into the EcoRI site of pcDNA3; this construct was designated pcDNA3/DMK 3Ј-UTR.
Western Blot Analysis-Cell lysates (20 g) were resolved electrophoretically on a 10% SDS-polyacrylamide gel (Hoefer Minigel). Gels were transferred to a 0.2-m nylon membrane (Rad-Free, Schleicher & Schuell) using a semi-dry transfer cell (Bio-Rad) in a glycine/Tris/ methanol buffer at 10 V for 15 min. The membrane was blocked in phosphate-buffered saline, pH 7.4, 0.1% Tween 20 (PBS-T) containing 5% (w/v) nonfat dry milk for 2-10 h. The primary antibody (anti-DM kinase (10), a polyclonal rabbit antibody kindly supplied by R. G. Korneluk) was diluted 1:5000 in PBS-T and incubated with the membrane for 1 h. After several washes with PBS-T, the membrane was incubated for 1 h with the secondary antibody, horseradish peroxidaselinked goat anti-rabbit IgG, 1:2000 (Jackson Immunoresearch Laboratories). After several washes with PBS-T, the membrane was incubated for 1 min with chemiluminescent detection reagents (ECL, Amersham Life Sciences) and exposed to Kodak X-Omat film for 5-30 s.
RNA Isolation and Northern Analysis-Total RNA was isolated from BC 3 H1 cells using RNA STAT-60 (18) (Tel-Test). RNA (15 g/lane) was resolved by denaturing agarose gel electrophoresis (2% agarose, 2% formaldehyde) and transferred to a nylon membrane (Nytran, Schleicher & Schuell) in a positive pressure transfer cell (Stratagene). Prehybridizations and hybridizations were performed in 10 ml of QuikHyb solution (Stratagene), 10 ml of formamide, and 200 g of yeast tRNA (Sigma R-5636) at 42°C. DNA probes were labeled with [ 32 P]dCTP using the Prime-It II random priming kit (Stratagene); unincorporated counts were removed by passing the probes through Sephadex G-50 Nick spin columns (Pharmacia Biotech Inc.). After overnight hybridization, membranes were washed twice for 15 min at room temperature in 1 ϫ SSC, 0.1% SDS followed by one wash at 58°C for 30 min in 0.25 ϫ SSC, 0.1% SDS. Autoradiography was performed by exposure of the membranes to Kodak X-Omat film for 12-48 h.
mRNA Differential Display-Differential display was done exactly as described by Pardee and Liang (17) (RNAmap, GenHunter). DNA-free total RNA (0.2 g) was reverse transcribed using oligo(dT) primers, followed by PCR with an arbitrary upstream primer in the presence of [ 35 S]dATP. Two arbitrary upstream primers were used: AP-1, 5Ј-AGC-CAGCGAA-3Ј; and AP-4, 5Ј-GGTACTCCAC-3Ј. Four downstream oligo-dT primers were used: T12MG, T12MA, T12MT, and T12MC. A variety of PCR products, representing the 3Ј ends of numerous expressed cDNAs, were visualized by autoradiography on a standard denaturing polyacrylamide sequencing gel. Differentially expressed bands of interest were extracted from the gel and reamplified. The reamplified products were subcloned into the TA cloning vector pCR II (Invitrogen), and fully sequenced in both directions by dideoxy sequencing (Sequenase 2.0). Sequences of interest were compared to the Gen-Bank™ data base at the National Center for Biotechnology Information, using the BLAST alignment algorithm (19) Network Service.

RESULTS AND DISCUSSION
The first attempt to produce a stable DMK-overexpressing cell line was unsuccessful. Stably transfected clones were pro-duced by transfecting BC 3 H1 cells with a plasmid construct containing the full-length DMK cDNA, including the 5Ј-and 3Ј-UTR. DMK mRNA and protein were undetectable in three independently generated full-length clones (Fig. 1, panel A,  lanes 3-6, and panel B, lanes 10 -13). One full-length clone, number 11, expressed very low levels of DMK mRNA but no protein expression could be detected. Given observations regarding the influence of UTR sequences upon gene expression (21)(22)(23)(24)(25), and considering that the genetic lesion associated with DM is located in the 3Ј-UTR of DMK, we removed the UTR to produce a coding-only construct (coDMK). The coDMK construct was derived from the full-length cDNA by PCR using primers that annealed to the start and termination codon sequences. DMK coding sequences were cloned into the pcDNA3 eukaryotic expression vector. Stable transfectants were produced with the coDMK construct and three independently isolated clonal cell lines constitutively expressed high levels of DMK mRNA and protein (Fig. 1, panel A, lane 2, and panel B,  lane 9). Additionally, six independent DMK 3Ј-UTR clones expressing high concentrations of DMK 3Ј-UTR mRNA were produced.
Log phase BC 3 H1 cells growing in high mitogen (20% serum) conditions typically exhibit a fibroblast-like morphology. In low mitogen conditions (1% serum), cells withdraw from the cell cycle, begin expression of muscle-specific genes, and assume an elongated, myoblast-like appearance (26). The coDMK transfectants were phenotypically similar to differentiated BC 3 H1 myoblasts, appearing elongated with numerous cellular outgrowths. coDMK transfectants did not exhibit cell cycle arrest and maintained the "differentiated" morphology in high serum conditions. In contrast, clones expressing DMK 3Ј-UTR alone retained a fibroblast-like morphology identical to untrans- fected BC 3 H1 cells.
In order to identify changes in gene expression associated with DMK or DMK 3Ј-UTR overexpression, we performed differential display analysis of the stable coDMK and 3Ј-UTR cell lines. The patterns of gene expression in the 3Ј-UTR clone and the untransfected control cells were nearly identical (Fig. 2,  panels A-C, lanes 2,3, 5, 6, 8, and 9). In comparison, cells expressing DMK exhibited a markedly different pattern of mRNA expression (Fig. 2, panels A-C, lanes 1, 4, and 7). Expression of numerous mRNAs were affected, apparently representing both up-and down-regulation of several genes. We initially reamplified and subcloned nine differentially ex-pressed cDNA products. The products were both sequenced and used to probe Northern blots to verify that differential display correlated with differential expression. Six of the products were unidentifiable, either showing no significant similarity to sequences in the GenBank™ data base, or matching the 3Ј end of a partially sequenced but uncharacterized cDNA. One product represented mitochondrial sequences and was not examined further.
One product, designated DM114.10, hybridized to an mRNA unique to DMK-expressing cells (Figs. 2C and 3A, lanes 1-3). DM114.10 also hybridized to an mRNA expressed at high concentrations in differentiated BC 3 H1 cells (data not shown). This product was identified as a differentiation-dependent form of ␤-tropomyosin unique to skeletal muscle. Rodents have been shown to produce two different tissue-specific ␤-tropomyosin messages (27). Fibroblasts and smooth muscle produce a 1.1kilobase mRNA transcript; skeletal muscle transcripts are larger (1.2 kilobases) and possess a different 3Ј end produced by alternative splicing and usage of a distal polyadenylation site. Our DM114.10 product is limited to those 3Ј sequences unique to the skeletal muscle mRNA.
The clone designated DM114.07 was identified as vimentin and recognized an mRNA present in untransfected BC 3 H1 cells that became diminished in DMK-expressing cells (Figs. 2B and 3B, lanes 4 -6) as well as differentiated BC 3 H1 cells (data not shown). Studies in rodents and humans have established that vimentin expression is high in satellite cells and regenerating muscle, but undetectable in mature skeletal muscle (28 -31).
Because of the differential expression of ␤-tropomyosin and vimentin, we began to suspect that cells expressing DMK were initiating a portion of a skeletal muscle differentiation program. Although the cognate messages could not be identified, the expression patterns of two of the differential display products, DM114.02 and DM114.03, were consistent with a skeletal muscle phenotype. DM114.02 recognized a transcript present in cells expressing DMK that was absent from untransfected cells (data not shown) and had high similarity to a human cDNA isolated from skeletal muscle (GenBank accession no. Z28822). DM114.03 hybridized to a transcript in untransfected cells that was absent in the DMK clone (data not shown) and shared similarity with a human cDNA from cardiac muscle (GenBank™ accession no. Z33441).
To further assess the differentiation state of DMK-expressing cells, we examined expression of two genes known to be implicated in myogenic development: myogenin (32) and retinoblastoma (Rb) (33,34). A myogenin cDNA probe detected high levels of myogenin mRNA induced in both DMK-expressing and differentiated cells (Fig. 4, A and B). Rb mRNA expression was affected in a reciprocal fashion; it was abundant in untransfected cells but decreased to undetectable levels in DMK-expressing and differentiated BC 3 H1 cells (Fig. 4, C  and D).
Creatine kinase (CK) isoforms serve as markers of skeletal muscle differentiation (35)(36)(37). Undifferentiated myogenic cells predominantly express the B isoform (Fig. 5, lane 1). Following myogenesis, M creatine kinase becomes the major isoform. We perfomed CK activity assays on DMK-expressing clones and observed that with respect to CK activity, DMK-expressing cells appeared to be equivalent to differentiated cells; M creatine kinase activity was strongly increased (Fig. 5, lanes 2  and 3).
Finally, we performed differential display on RNA from untransfected/undifferentiated, differentiated, and DMK-expressing cells. The patterns of mRNA expression shown by DMK-expressing and differentiated cells were nearly identical (Fig. 2D, lanes 11 and 12), and both were substantially different from the pattern of mRNA expression detected in untransfected/undifferentiated cells (Fig. 2D, lane 10).This observation suggests that many of the changes in gene expression associated with DMK overexpression are similar to those induced during differentiation in BC 3 H1 cells.
We have not determined whether the apparent myogenic effect of DMK overexpression is a specific one, or if it is simply a result of the nonspecific activity of an ectopically expressed protein kinase. However, overexpression of protein kinase A or C, the two kinases that share the greatest homology with DMK, actually inhibit muscle-specific transcription (38,39). Although no protein kinase has yet been conclusively implicated in the control of myogenesis, it is likely that kinase cascades are involved in the establishment and maintainance of the differentiated state in muscle (40). It has been speculated that DMK may influence muscle development through an effect upon myogenic gene products (41). Recent observations by Meixell et al. 2 suggest that myogenin can serve as an in vitro substrate for DMK. Ultimately, however, any formal model will rely upon identification of the actual in vivo substrate. It is interesting to note that DMK-expressing cells maintain a differentiated phenotype while actively dividing in high mitogen conditions, suggesting that cell cycle arrest is not a prerequisite for differentiation in these clones.