The 14 kb muscle isoform of the Duchenne muscular dystrophy (DMD) gene is expressed primarily in skeletal and cardiac muscle. Transcription of the muscle isoform is induced as myoblasts differentiate into multinucleated myotubes and transcript levels are increased a further 10-fold in mature skeletal muscle. In previous studies we have demonstrated that the core muscle promoter of the human DMD gene contains sequences that regulate the induction of DMD gene expression with myoblast differentiation. However, direct injection studies have indicated that the activity of the core muscle promoter in mature skeletal muscle is 30-fold lower than in immature myotubes. This discrepancy between endogenous transcript levels and core promoter activity suggested that additional transcriptional elements are involved in the regulation of DMD gene expression in muscle. In this report we present evidence for the existence of a muscle-specific enhancer within intron 1 of the human DMD gene. Functional analysis of HindIII fragments from within a 36 kb region surrounding muscle exon 1 of the human DMD gene resulted in the identification of a 5.0 kb fragment within muscle intron 1 that consistently provided high levels of reporter gene expression in both immature and mature skeletal muscle. Sequences within this 5 kb fragment were shown to be functionally independent of position and orientation and to be inactive in fibroblasts, properties that are consistent with the definition of a muscle-specific enhancer. Although this enhancer provided a 30-fold increase in transcription from a SV40 viral promoter in mature skeletal muscle, only a 3-fold increase was observed from the DMD core muscle promoter. Intron 1 enhancer activity alone is therefore insufficient to account for the discrepancy between endogenous transcript levels and core muscle promoter activity in immature and mature skeletal muscle and points to the existence of additional enhancer elements in other regions of the DMD gene. This report provides the first evidence for the involvement of a transcriptional enhancer in DMD gene regulation in muscle and impacts on our understanding of the functional consequences of mutations at the 5'-end of gene. In this regard, deletions in this region in X-linked dilated cardiomyopathy patients provides indirect evidence for a role for this enhancer in regulating DMD gene expression in cardiac muscle.
The product of the Duchenne muscular dystrophy (DMD) gene in muscle is a 427 kDa cytoskeletal protein (dystrophin) that is one component of a large glycoprotein complex at the surface membrane of muscle fibers (1 -4 ). The 14 kb transcript encoding the muscle isoform of the DMD gene is spliced from more than 79 exons spread over 2500 kb or ~1.5% of the entire length of the human X chromosome (5 ,6 ). The predicted transcription time of 14-24 h based on published elongation rates for RNA polymerase II has been verified experimentally (7 ). DMD gene transcripts are most abundant in skeletal and cardiac muscle (8 ), consistent with the primary involvement of these tissues in the pathogenesis of the disease. However, DMD gene transcripts represent only 0.02-0.12% of total mRNA in mature skeletal muscle (8 ) and have been reported to be 10-fold lower in immature myotubes (8 ,9 ). This would imply that the muscle promoter of the DMD gene is relatively weak and provides low rates of transcript initiation in immature and mature skeletal muscle. Alternatively, DMD muscle promoter activity may be high but transcript accumulation may be limited by factors such as the length of time needed to transcribe the gene, the 30-40% differential between transcript initiation and completion (7 ), or the possibility that completed transcripts have a relatively short half-life in the cell.
In previous studies we have demonstrated that genomic fragments containing at least 150 bp of sequence upstream of muscle exon 1 in the human DMD gene can mediate the induction of chloramphenicol acetyltransferase (CAT) reporter gene expression as primary human, primary mouse and rat H9C2(2-1) myoblasts differentiate into multinucleated myotubes in vitro (10 ). This result was consistent with observations of an up-regulation of endogenous DMD gene expression with myoblast differentiation (7 ,9 ,11 ). Sequence analysis of the 150 bp `minimal promoter' region identified domains that were highly homologous to E-box (12 ,13 ) and CArG box (14 -16 ) motifs which are binding sites for the myogenic determination gene (MyoD, myf5, myogenin, MRF4) and serum response factor [SRF; CArG box binding factor (CBF)] families of transcription factors which have been shown to play important roles in creatine kinase, acetylcholine receptor, and skeletal and cardiac actin gene regulation in muscle. Functional studies by Gilgenkrantz et al. (17 ) demonstrated that the minimal muscle promoter region of the DMD gene could not be trans-activated by MyoD, but that mutation of the CArG box motif resulted in a significant decrease in promoter activity. This regulatory domain was also shown to compete for the same transcription factor that binds to CArG box motifs within the cardiac actin gene promoter. Taken together, these results suggested that muscle-specific DMD gene expression was regulated at least in part by mechanisms common to other muscle-specific genes (18 ).
Transcriptional enhancers are defined by their ability to elevate gene expression in a tissue- or development-specific manner independent of their position and orientation relative to a core promoter. Enhancer elements have been identified in a number of muscle-specific genes to date and in most cases have been shown to play a critical role in the regulation of these genes in skeletal and/or cardiac muscle. For example, expression of the muscle creatine kinase (MCK) gene in skeletal and cardiac muscle (19 ,20 ) is regulated by a muscle-specific basal promoter and two enhancer elements: one located 1.1 kb upstream of the transcription start site and a second located within intron 1 approximately 1.0 kb downstream of exon 1 (19 -21 ). By itself the MCK basal promoter provides relatively low levels of reporter gene expression in skeletal and cardiac muscle. The upstream enhancer is required for high levels of MCK gene expression in skeletal and cardiac muscle, while the downstream enhancer has been reported to function in skeletal muscle only (19 ).
The critical role that transcriptional enhancers play in the regulation of muscle-specific genes raised the question of whether enhancers were also involved in regulating DMD gene expression in skeletal and cardiac muscle. Indirect evidence for enhancer involvement in DMD gene transcription was provided by the identification of a X-linked dilated cardiomyopathy patient with a mutation within the DMD gene that removes the muscle promoter, muscle exon 1, and part of muscle intron 1 (22 -24 ). Skeletal muscle in this patient was phenotypically normal and dystrophin could be detected immunocytochemically. No dystrophin was detected in cardiac muscle. Neuronal- and cerebellar purkinje cell-specific promoters lying upstream and downstream of muscle exon 1 were found to be selectively activated in the skeletal muscle of this patient, raising the possibility that this deletion removes a cardiac-specific enhancer and repositions a skeletal muscle-specific enhancer to allow it to activate these non-muscle promoters in patient skeletal muscle (25 ,26 ). However, the recent description of XLDC arising from a point mutation in the 5' splice site of the first intron of the DMD gene (27 ) is more difficult to interpret in terms of differential transcriptional consequences in skeletal and cardiac muscle.
In this report we present direct evidence for the presence of a muscle-specific enhancer element within intron 1 of the human DMD gene. A comparison of the relative transcriptional activities of DMD muscle promoter-reporter gene constructs in immature myotubes and mature skeletal muscle revealed inconsistencies between these activities and endogenous DMD gene transcription patterns. A functional survey of fragments from within a 36 kb region surrounding muscle exon 1 identified a 5 kb HindIII fragment within muscle intron 1 that displayed properties consistent with other muscle-specific enhancers. Although this enhancer was seen to have positive effects on DMD muscle promoter activity in both immature and mature skeletal muscle, inconsistencies between these activities and endogenous transcript levels remain unresolved.
A 700 kb long-range SalI (S), EagI (E) and BssHII (B) restriction map of the 5'-end of the human DMD gene is shown in Figure 1 . The relative positions of restriction sites and the first exons of the brain (BE1) and cerebellar purkinje cell (CPE1) transcripts are based on the long-range map proposed by Gorecki et al. (28 ). Our observation of unique SalI and EagI restriction sites within a 36 kb cosmid clone (XJcos8; 10 ) that contains muscle exon 1 (ME1) established that muscle exon 1 and the cerebellar purkinje cell promoter are separated by 15-65 kb of muscle intron 1 sequence. Repositioning of ME1 increased the distance between the muscle and brain promoters to between 170 and 220 kb. The size of muscle intron 1 was estimated to be 120-155 kb and exon 3 was estimated to lie 315-390 kb downstream of exon 2 (28 ). The relative positions of exons 3 and 7 in Figure 1 are based on our mapping of these exons to three other overlapping cosmid clones (unpublished data).
Several indirect lines of evidence have suggested that additional muscle-specific transcriptional regulatory elements may exist within the large introns surrounding muscle exon 1 of the DMD gene. In this study we present evidence demonstrating that a muscle-specific enhancer is located within the first intron of the human DMD gene. These sequences are localized to within a 5 kb HindIII fragment that is positioned downstream and adjacent to a 3.4 kb HindIII fragment containing muscle exon 1 and the core muscle promoter. Transient transfection and direct injection of reporter gene constructs demonstrated that this enhancer element functions in both immature myotubes and mature skeletal muscle. In vitro studies demonstrated that these sequences function in a position- and orientation-independent manner in immature muscle but are inactive in fibroblast cultures, characteristics consistent with a true muscle-specific enhancer. The intron 1 enhancer was also shown to have a positive influence on the transcriptional activity of the core muscle promoter in both immature and mature skeletal muscle, although the degree of transcriptional activation was significantly less than that observed from heterologous viral promoters. This result was somewhat unexpected in view of a similar study that had shown that the upstream enhancer of the muscle creatine kinase (MCK) gene was more effective in activating transcription from its endogenous promoter (20-fold) than from a heterologous viral promoter (15-fold) in mature skeletal muscle (20 ). However, Gilgenkrantz et al. (17 ) have shown that the SV40 enhancer, which has been reported to have activity in skeletal muscle similar to that of the myosin light chain 1/3 enhancer (35 ), generated only a 4-fold increase in reporter gene expression in immature myotubes when positioned upstream of the core muscle promoter of the DMD gene. We have also observed that the SV40 enhancer has little or no influence on the transcriptional activity of the core muscle promoter in mature skeletal muscle (data not shown). Taken together, these results suggest that the muscle promoter of the DMD gene is less responsive to the influence of enhancer sequences than are heterologous viral promoters, perhaps reflecting the fact that the core promoter contains sequences that tightly regulate it's own transcriptional activity at different stages of muscle development (10 ,17 ).
The relatively low transcriptional activity of the intron 1 enhancer-DMD core promoter combination also raises the possibility that this enhancer functions primarily in cardiac muscle tissue. Deletion of the muscle promoter and a portion of muscle intron 1 in a X-linked dilated cardiomyopathy (XLDC) patient has been reported to result in aberrant expression of DMD gene transcripts from the brain and cerebellar purkinje cell-specific promoters in skeletal but not in cardiac muscle (25 ,26 ). It was suggested that this deletion might remove a cardiac-specific transcriptional regulatory element, raising questions regarding the potential involvement of the intron 1 enhancer described in this report in regulating DMD gene transcription in cardiac muscle. Although the molecular events governing cardiac gene expression are not well understood, cardiac-specific transcriptional control elements have been identified within the promoter regions of the muscle creatine kinase gene (20 ), the cardiac troponin T gene (36 ), the cardiac myosin light chain 2 gene (37 ) and the slow/cardiac troponin C gene (38 -40 ). Specific sequence elements within the upstream enhancer of the muscle creatine kinase gene have also been implicated in cardiac muscle-specific gene regulation (19 ). It will be interesting to determine whether the DMD intron 1 enhancer contains cardiac-specific transcriptional control elements and the degree to which it is involved in regulating DMD gene expression in cardiac muscle. In this regard it may be noteworthy that the intron 1 enhancer was initially identified by its activity in the H9C2(2-1) myogenic cell line which was originally derived from embryonic rat heart and which reportedly retains residual cardiac-specific transcriptional properties (41 ,42 ). However, a recent report describing XLDC arising from a splice mutation at the 5'-end of intron 1 (27 ) is difficult to explain in terms of potential mechanistic effects on the activity of this enhancer.
The role of the muscle intron 1 enhancer element in the regulation of endogenous DMD gene transcription in skeletal and cardiac muscle remains unclear. Our data suggests that this enhancer element does not by itself account for the discrepancies between endogenous DMD gene transcript levels and the activity of the core muscle promoter in early myotubes and mature skeletal muscle and raises the possibility that additional as yet unidentified enhancer elements may be involved in the regulation of DMD gene transcription in muscle. An alternative interpretation is that these discrepancies reflect species or development-related differences in transcriptional, translational or post- translational processing of reporter gene expression plasmids used in these studies. However, to our knowledge significant tissue- or species-specific differences in reporter gene processing have not been reported. Furthermore, the unusually high levels of sequence conservation observed within the DMD minimal promoter region (17 ) and the consistencies in DMD promoter function observed in primary human, primary mouse and rat H9C2(2-1) myotubes in vitro (10 ) make it unlikely that these discrepancies reflect species-specific variations in transcriptional activity. In any event, the identification of a muscle-specific transcriptional enhancer within the DMD gene has important implications on our understanding of the transcriptional regulation of this gene in muscle, on the organization of functional domains at the 5'-end of gene and on the significance of mutations within this region. Indirect evidence suggests that this enhancer may be involved in regulating DMD gene expression in cardiac muscle, and that one or more skeletal muscle-specific enhancers are also involved in regulating DMD gene expression. As no other enhancer domains were identified in our survey of 36 kb of genomic sequence, any additional enhancers would be expected to be positioned at least 16 kb upstream or 20 kb downstream of muscle exon 1.
The isolation of a cosmid clone (XJcos8) containing the 275 bp muscle-specific exon 1 of the human Duchenne muscular dystrophy gene has been described elsewhere (10 ). Individual HindIII fragments isolated from this cosmid clone were initially subcloned into the ptkGH vector (43 ) for preliminary screening for enhancer activity. Construction of the pHB918CAT construct containing the core human muscle-specific DMD gene promoter has been described (10 ). All additional chloramphenicol acetyltransferase (CAT) expression constructs were prepared in the pBLCAT2 vector containing the enhancerless thymidine kinase (tk) promoter upstream of the CAT gene (44 ). For example, the 5 kb HindIII fragment from within intron 1 (DME.H5) was excised from the ptkGH vector and subcloned upstream and in both orientations relative to tkCAT.
For direct injection experiments DMD promoter and intron 1 enhancer fragments were subcloned into the pGL2 luciferase reporter gene vector series (Promega). Constructs prepared include the 850 bp human DMD muscle promoter fragment cloned into either the pGL2-Basic (promoterless, enhancerless) vector or the pGL2-Enhancer vector (containing the SV40 enhancer downstream of the luciferase gene) at the HindIII site in the positive orientation; and the intron 1 enhancer fragment cloned into the SalI-BamHI sites downstream and in the negative orientation relative to the DMD promoter in the pGL2-Basic vector and the SV40 promoter in the pGL2-Promoter vector (containing an enhancerless SV40 promoter). Genomic HindIII fragments from within the region surrounding muscle exon 1 were isolated from the XJcos8 cosmid and subcloned upstream of the SV40 promoter in the pGL2-Promoter vector. Constructs were tested by transient transfection of H9C2 myoblasts prior to use in direct injection experiments.
H9C2(2-1) myoblasts were seeded at a density of 100 cells/mm2 on plastic culture dishes in Alpha MEM containing 16 mM glucose, 10% fetal bovine serum and 40 [mu]g/ml gentamycin (Growth medium). Cells were allowed to recover for 24 h at 37oC in a 5% CO2 atmosphere and fresh Growth medium was added 3-5 h prior to transfection. Cells were transfected using the modified calcium phosphate precipitation protocol (45 ) using 10 [mu]g test plasmid and 10 [mu]g or the reference plasmid pXGH5 (43 ) per 100 mm dish. Plasmids were prepared by double-banding on ethidium bromide-cesium chloride gradients. The pXGH5 reference plasmid contains the human growth hormone gene driven by the mMT-1 promoter and was included in all experiments as a control for transfection efficiency. Transfected cells were washed three times in Alpha MEM and allowed to recover in Growth medium for 24 h. Cells were exposed to Fusion medium (Alpha MEM containing 16 mM glucose, 5% horse serum and 100 U/ml penicillin: 0.1 mg/ml streptomycin) for 72 h to allow for differentiation prior to harvesting for CAT and growth hormone assays. NIH 3T3 fibroblasts were maintained in Growth medium at all times.
Gene transfer was carried out in tibialis anterior (TA) muscles of 6-8 wk old male C57BL mice (Iffa Credo, France) under anesthesia (sodium pentobarbital, 75 mg/kg IP) as described previously (30 ). Plasmid DNA was injected into normal (mature) or regenerating (immature) TA muscles, which were removed 3 days later and assayed for luciferase reporter gene activity. Regeneration had been induced 3 days previous to gene transfer by injection of a snake-venom derived cardiotoxin (34 ) such that the muscle would contain predominantly young myotubes at the time of gene transfer and late myotubes or young myofibers at the time of assay. For each construct to be tested individual TA muscles were each injected with 10 [mu]g DNA (50 [mu]l at 0.2 mg/ml in PBS).
Chloramphenicol acetyltransferase (CAT) assays were performed on transfected cell extracts according to the dual phase-diffusion procedure (46 ). Briefly, cell extracts (25 [mu]g protein) were added to reaction mixtures containing 100 mM Tris-HCl (pH 7.8), 1.0 mM chloramphenicol and 0.1 mM acetyl CoA [2-5 [mu]Ci [3H]acetyl-CoA (198 mCi/mmol; Dupont Canada Inc., Mississauga, Ont.)] in a final volume of 250 [mu]l. Reactions were overlaid with 5 ml of Econofluor (Dupont Canada Inc.) and assayed at 30-60 min intervals. CAT activities were calculated using values which fell within the linear range of the reaction and were converted from c.p.m. to units CAT using standard CAT activity values determined in each experiment. Human growth hormone levels in 20 [mu]l aliquots of Fusion medium sampled just prior to harvesting cells for CAT assays were determined using a dual-antibody radioimmunoassay (Nichols Institute Diagnostics, Los Angeles, CA) according to the manufacturer's protocol. Values were converted to ng growth hormone/ml using a standard curve prepared with each experiment. Cell extract protein concentrations were determined using the procedure of Lowry et al. (47 ). CAT activities are expressed as units CAT/mg protein/ng growth hormone. Luciferase assays were performed on TA muscles removed from mice 3 days after injection of DNA. Luciferase activity was measured using the Promega `Luciferase Assay System' as described previously (30 ). The protein concentrations in mouse muscle extracts were determined using the BioRad microassay procedure. Results were expressed as group means +- SEMs of RLU/mg protein/s. As no significant differences were observed in reporter gene activities in mouse TA muscles injected and harvested under these different conditions the results were pooled and reported as promoter activities in mature muscle.
We would like to thank Robert Whalen for his assistance with the direct injection experiments. Portions of this work were supported by grants from the Medical Research Council of Canada (H.J.K and H.L.D.), the University of Ottawa (H.L.D.), the Muscular Dystrophy Association (R.G.W. and P.N.R.), the Muscular Dystrophy Association of Canada (R.G.W. and P.N.R.), and the Canadian Networks of Centres of Excellence (R.G.W.).
Human Molecular Genetics
Pages
Introduction
Results
Discussion
Materials And Methods
Plasmid construction
Cell lines and transfections
Direct injection of mouse skeletal muscle
Reporter gene assays
Acknowledgements
References
REFERENCES
This page is maintained by OUP admin. Last updated Thu Oct 31 15:27:49 GMT 1996. Part of the OUP Journals World Wide Web service.Copyright Oxford University Press, 1996