The dystrophin gene defective in Duchenne muscular dystrophy (DMD) is extreme in size and complexity with several promoters which direct expression of different isoforms in different tissues. In contrast with adult skeletal muscle which expresses 427 kDa dystrophin, fetal muscle tissue expresses the 71 kDa ubiquitous isoform Dp71 as well as 427 kDa muscle dystrophin. To examine Dp71 expression in fetal muscle further, we have monitored its expression pattern in differentiating myogenic cultures of human fetal muscle origin. The presence of transcripts initiated from the Dp71 promoter was demonstrated by quantitative RT-PCR. The level of transcript expressed from the Dp71 promoter did not change significantly during myogenic differentiation, consistent with the housekeeping nature of the promoter. Measurements to determine the stability of the Dp71 mRNA indicated that it has a half-life of ~20 h and, therefore, is somewhat more stable than the larger 14 kb muscle dystrophin mRNA (t1/2 = 16 h). In contrast with the constant level of Dp71 transcript during myogenic differentiation, the level of Dp71 protein increased significantly, perhaps due to changes in translation efficiency or protein stability. These results demonstrate expression and posttranscriptional upregulation of Dp71 in human fetal myogenic cultures.
Duchenne muscular dystrophy is an X-linked lethal disorder characterized primarily by progressive degeneration of skeletal muscle and caused by mutations in the dystrophin gene. This complex gene encodes a set of at least seven protein isoforms of various sizes and tissue distributions, each translated from a unique message initiated from one of the seven distinct promoters in the gene. Three high molecular weight isoforms (427 kDa), expressed in skeletal muscle (1 ), neurons of the cerebral cortex (2 -4 ) and purkinje cells of the cerebellar cortex (2 ), are the result of transcripts initiated from three distinct promoters at the 5' end of the gene, upstream of a common exon 2. Shorter isoforms of size 260 kDa (5 ), 140 kDa (6 ), 116 kDa (7 ) and 71 kDa derive from transcripts that contain unique first exons located in introns 29, 44, 55 and 62. The first three of these have tissue specificity for the retina, central nervous system and peripheral nerve respectively.
The shortest isoform, Dp71, is present in the liver cell line HepG2 and in many adult tissues including brain, liver, testis, lung and kidney but is not detectable in adult skeletal muscle (8 ,9 ). The transcripts are initiated from a promoter 8 kb upstream of exon 63 (9 -11 ). Dp71 contains the cysteine-rich and C-terminal domains of dystrophin and is missing the N-terminal actin binding and spectrin-like repeat domains present in the largest isoforms (10 ). Seven amino acids are encoded by Dp71 exon 1 generating a novel N-terminus. The remainder of the amino acids are encoded by exons 63-79 with alternative splicing of exons 71 and 78 (12 ). Dp71 is undetectable in fetal brain but the levels increase during development and are detectable in adult brain. Conversely, Dp71 is expressed at detectable levels in fetal muscle but not in adult muscle tissue (13 ). The function of Dp71 is unknown but it localizes to the membrane of HepG2 cells (14 ) and when expression is forced in myogenic cells not expressing the 427 kDa dystrophin isoform it localizes to the membrane and restores the dystrophin-associated glycoprotein (DAG) complex to the membrane (15 ,16 ). Despite this, Dp71 is unable to correct the muscle defect in mouse models of DMD indicating that its function differs from dystrophin in muscle (15 -16 ).
Since Dp71 is expressed in fetal, but not adult skeletal, muscle, its level of expression in cultured fetal myoblasts and differentiated myotubes is not readily predictable, and measurement of both mRNA and protein expression during myogenic differentiation in culture has not been reported. Such measurements are important, as expression in fetal myogenic cultures provides a system for study of the regulation and function of Dp71 in muscle; the primary tissue affected in individuals with DMD. Furthermore, previous studies have focused on muscle tissue which contains multiple cell types, the use of clonal human fetal myogenic cultures for these studies will clarify if Dp71 is expressed in this cell type. We monitored expression of Dp71 mRNA by quantitative RT-PCR and observed a significant level of transcript in fetal myoblasts but no increase in this level as cells fused to form multinucleated myotubes, a result consistent with the housekeeping nature of the Dp71 promoter. This was in contrast with dystrophin transcript levels which, as expected, increased during differentiation. The Dp71 mRNA half-life was measured to be 20 h, not grossly different from the dystrophin mRNA half-life of 16 h. In contrast with the relatively constant Dp71 mRNA levels, we observed an increase in the level of Dp71 protein, indicating posttranscriptional upregulation of Dp71 expression during fetal myotube formation.
The Dp71 isoform previously characterized in other tissues is encoded by a transcript containing a novel exon 1 spliced to exons 63-79 of the dystrophin gene with alternative splicing of exons 71 and 78 (10 -12 ). To determine whether Dp71 transcripts are expressed in human fetal myogenic cultures, myoblasts were plated at high density in growth medium, allowed to recover for 24 h, and then exposed to fusion medium to induce differentiation into multinucleated myotubes. Total RNA was isolated from cultures after 24 h in fusion medium and Dp71 mRNA was detected by RT-PCR. Template cDNA was made by reverse transcription using primer 79R in exon 79. Dp71 transcripts were subsequently amplified using the same reverse primer and the forward primer 741 in the novel Dp71 exon 1. The Dp71 amplification product expected from this reaction is 1879 bp if the RNA is not alternatively spliced. A single product of approximately this size was amplified from RNA of human fetal myogenic cultures (Fig. 1 A).
To monitor the expression profile of Dp71 and dystrophin transcripts during myogenic differentiation, clonal human fetal myoblasts were plated at high density, allowed to recover in growth medium for 24 h, then exposed to fusion medium to induce differentiation and dystrophin gene expression. Cell samples were collected 0, 10.5, 23.5, 29.8 and 37.4 h after the addition of fusion medium to the cultures and total RNA was isolated from each sample. Quantitative RT-PCR was used to monitor transcript accumulation. To quantitate RNA levels by RT-PCR accurately, we introduced an in vitro transcribed control RNA (cRNA) as a standard with which to compare the endogenous RNA. A control template (Dp71-1-79) for in vitro transcription (Fig. 2 ) was constructed by cloning part of the Dp71 cDNA, which included the novel exon 1 and exons 63-78 and a portion of exon 79 of the dystrophin gene, into a vector with a T7 promoter. A 118 bp insertion was introduced into exon 65 so that the products obtained for the control and endogenous RNAs following RT-PCR could be separated by gel electrophoresis.
In other studies we measured the stability of the dystrophin mRNA in fetal myogenic cultures and obtained a half-life of 16 h (17 ). Since full length muscle dystrophin mRNA and Dp71 mRNA are both expressed in fetal myogenic cultures, and therefore in the same cellular environment, it was interesting to compare the stability of these transcripts which share many sequences at the 3' end of the message. For these experiments fetal myogenic cultures were grown in fusion medium for 40 h and then exposed to 5 [mu]g/ml actinomycin D, an inhibitor of transcription. Total RNA was prepared from cultures exposed to actinomycin D for 0, 1, 2, 4, 7, 11 and 23 h. The decay of the Dp71 mRNA was monitored by quantitative RT-PCR. Control RNA was prepared using the template Dp71-1-79 (Fig. 2 ). A measured amount of cRNA was combined with a measured amount of total RNA from the cultured cells and cDNA was made by reverse transcription using primer 926 in exon 72. The Dp71 mRNA and cRNA were amplified using the forward primer Dp71-1 in the first exon of Dp71 and the reverse primer 69R in exon 69. The decay of the endogenous Dp71 mRNA was monitored and the results are shown in Figure 4 . The amount of endogenous RNA decreases modestly relative to the control RNA. Using a phosphorimager the amount of control and endogenous RNA was determined and the results used to calculate the absolute amount of Dp71 mRNA in each sample. This analysis was done for three different myogenic clones. The graph in Figure 5 plots the average percent maximum against the time the cells were exposed to actinomycin D. These results show that ~20 h are required for 50% of the Dp71 mRNA to decay, indicating this mRNA is somewhat more stable than the dystrophin mRNA.
To examine the expression profile of the Dp71 protein during differentiation, total protein extracts were prepared from myogenic cultures. Protein accumulation was examined using Western blots stained with the monoclonal antibody Mandra 1 which detects both dystrophin and Dp71. The protein samples were prepared from the same clone and simultaneous with the RNA samples used for the quantitative RT-PCR experiments shown in Figure 3 . The signal was revealed by ECL (enhanced chemiluminescence) and exposure to X-ray film (Fig. 6 A).
The initial characterization of Dp71 by others established that the protein contains only the cysteine-rich and C-terminal domains of dystrophin and that it is expressed in most tissues except adult skeletal muscle (8 -9 ). More recent studies have detected Dp71 in fetal muscle by Western blot (13 ) and at very low levels in adult skeletal muscle tissue by Western blot analysis of immunoprecipitated proteins (19 ). Since muscle tissue contains non-myogenic cell types it is difficult to assess the cellular origin of the Dp71 protein in these studies. We have shown expression of both Dp71 protein and transcript in clonal cultures containing only myogenic cells, indicating that Dp71 is expressed in human fetal myogenic cells. This is consistent with a previous report that the Dp71 transcript is expressed at low levels in the rat myogenic cell line L185 (20 ).
Our analysis of Dp71 expression during myogenic differentiation indicated that the transcript level was relatively constant over the first 38 h of this process. This was not surprising as studies of the Dp71 promoter have shown it to be a housekeeping type promoter, high in GC content with several potential Sp1 binding sites and no TATA box (11 ). This type of promoter is expected for genes expressed in many different tissues and is not likely to be responsive to differentiation signals. In contrast with the Dp71 transcript, the level of full-length (14 kb) dystrophin transcript increased sharply following differentiation. This result is similar to that obtained in differentiating rat glial cultures where the Dp71 transcript remains constant while the level of dystrophin transcript increases (20 ). This increase is consistent with transfection studies showing that expression from the muscle promoter, which contains several myogenic regulatory sequences, is upregulated during myogenic differentiation (21 ). The RT-PCR strategy we designed to assay Dp71 RNA used primers in exon 72 or 79 for reverse transcription and therefore would not have detected the apo-dystrophin-3 transcript which contains dystrophin exons 63-70 (22 ).
Using actinomycin D we have shown that the Dp71 mRNA has a half-life of ~20 h, a value somewhat greater than that obtained for muscle dystrophin mRNA (t1/2 = 16 h) (17 ) in the same cellular environment. The fact that these values are not grossly different supports the concept that Dp71 and dystrophin transcript accumulation profiles differ primarily because of differential regulation at the promoters for these transcripts.
The use of independent promoters within the same gene to activate transcription in different tissues is a common regulatory mechanism. Independent promoters are used less frequently to drive expression of vastly different isoforms in the same tissue, but one other example is the gene encoding myosin light chain kinase and its smaller isoform telokin (23 ). This mechanism may be used less commonly, than for instance, multigene families, because nested transcription units active in the same gene may tend to cause promoter occlusion. This has been observed for the ADH gene in Drosophila (24 ), which contains two closely linked promoters. Our results indicate that in the case of the dystrophin gene an increase in expression from the upstream dystrophin muscle promoter does not significantly alter expression from the downstream Dp71 promoter. Perhaps the great distance between the promoters due to the large size of the dystrophin gene prevents such effects.
In these studies we also examined expression of the Dp71 and dystrophin proteins during fetal myogenic differentiation in culture. Protein levels were monitored by Western blot analysis using the monoclonal antibody Mandra 1 which detects both proteins. Dystrophin protein levels are very low initially and begin to increase at ~23 h after exposure of the cultures to fusion medium (Fig. 7 ). This is consistent with our previous studies which showed that 16 h are required for transcription of this gene (>2300 kb) and that under our culture conditions mature transcript begins to accumulate ~21 h after exposure to fusion medium (18 ).
In contrast with the relatively constant levels of Dp71 transcript during myogenic differentiation, the level of Dp71 protein increased significantly over this same time period. Since transcript levels stay fairly constant the mechanism by which protein levels increase may involve changes in translational efficiency. Alternatively, the effective stability of the protein may change during differentiation. The effective stability is determined by two components, the intrinsic stability and a dilution factor dependent on the rate of cell division (25 ). During the process of differentiation cells stop dividing and if the protein is stable it will appear to accumulate as it is no longer diluted by cell division.
Studies of Dp71 expression in mouse indicate that although the protein is expressed in fetal muscle the levels decrease and become undetectable in adult muscle tissue. Therefore, although the level of Dp71 protein increases during the early formation of multinucleated myotubes, it is downregulated as the animal matures indicating another level of regulation in vivo.
As observed for other tissues (10 -12 ), Dp71 is alternatively spliced for exons 71 and 78 in cultured fetal muscle. Our analysis of the Dp71 transcript in these cultures indicated that a portion of Dp71 mRNAs are missing exon 71. This deletion does not change the reading frame and therefore results in removal of 13 amino acids from the protein (10 ). Similarly, a portion of transcripts are missing exon 78, the penultimate exon. In this case the reading frame is changed and the last 13 amino acids are replaced by 31 amino acids (10 ). Our Western blot analysis confirmed the presence of multiple Dp71 isoforms at the protein level in these cultures. Western analysis of Dp71 expressed from a transgene in an mdx myoblast cell line identified a doublet where the larger protein was shown to be phosphorylated (15 ) and treatment with alkaline phosphatase gave a single small band. Therefore some of the Dp71 isoforms we observed by Western blot may be due to differences in phosphorylation.
The 427 kDa dystrophin protein expressed in muscle is localized to the muscle sarcolemma (26 ). Muscle dystrophin binds cytoskeletal actin at its N-terminus (27 ) and associates with a membrane-associated glycoprotein complex through the cysteine-rich C-terminal domains (28 -30 ). One member of the dystrophin-associated glycoprotein (DAG) complex interacts with a component of the extracellular matrix (31 ) and therefore dystrophin links the intra- and extracellular matrices across the muscle membrane. It has been proposed that through these interactions dystrophin may stabilize the muscle membrane and prevent rupture induced by repeated rounds of contraction. Clearly, Dp71 would not play such a role as it does not contain the actin-binding domain of dystrophin. Alternatively, Dp71 may play a role in cell to cell contact or may help organize the glycoprotein complex (13 ). Dp71 will go to the membrane and restore the glycoprotein complex in muscle of mdx mice indicating that it can interact with components of the complex (15 ). Studies have shown that some components of the DAG complex bind to domains of dystrophin that are present in Dp71. For example, the syntrophins bind to amino acids in exons 73 and 74 (32 ,33 ) and [beta]-dystroglycan binds directly to residues of the cysteine-rich and C-terminal domains (34 ). Furthermore, alternative splicing may modulate the nature of the proteins that associate with Dp71. Since Dp71 is expressed in myoblasts and newly formed myotubes its expression overlaps with dystrophin and may compete for proteins like these that interact with the cysteine-rich C-terminal domains. The precise role of Dp71 in fetal muscle is not known but its presence in the primary tissue affected in DMD patients makes the observation interesting and potentially significant.
Clonal populations of myogenic cells were prepared from muscle of a 10 week human fetus. The procedures to prepare clonal human myoblast cultures have been described (21 ). A myogenic clone was expanded at low density in cloning medium [F12 medium (Gibco-BRL) containing 20% fetal bovine serum and one Bullet Kit (Clonetics) which includes 0.5 ml human epidermal growth factor (10 [mu]g/ml), 5.0 ml insulin (100 mg/ml), 0.5 ml GA-1000 (gentamycin at 50 mg/ml and amphotericin at 50 [mu]g/ml), 5.0 ml Fetuin (50 mg/ml), 0.5 ml dexamethasone (0.39 mg/ml) and 5.0 ml bovine serum albumin (50 mg/ml)]. The cells were then plated at high density in cloning medium, allowed to recover for 24 h, and then transferred to fusion medium (alpha minimal essential medium; Gibco-BRL) containing 16 mM glucose, 2% fetal bovine serum and 100 U/ml penicillin and 0.1 mg/ml streptomycin, to induce differentiation into multinucleated myotubes.
Total RNA was isolated from myogenic cultures as described (35 ). RNA concentration was determined from OD260 nm.
The plasmid Dp71-1-79 shown in Figure 2 contains the last 27 bps of Dp71 exon 1 followed by dystrophin cDNA sequences 9433-11806 which includes dystrophin exons 63-78 and a portion of exon 79. There is a 118 bp fragment from the HaeIII phi X ladder (Gibco-BRL) inserted in exon 65. It was constructed from the plasmid Dp71-118 which is a Bluescript vector (Stratagene) containing dystrophin exons 63-70 following the Dp71 first exon and a 118 bp insertion in exon 65. To make Dp71-1-79, the plasmid Dp71-118 was digested with AccIII and NdeI to remove dystrophin cDNA sequences downstream of 9996. A 1.8 kb AccIII-NdeI insert containing dystrophin cDNA sequences 9996-11806 was isolated from the plasmid p9-14 (ATCC 57677) and ligated into the above plasmid.
Dp71-1-79 plasmids were isolated from bacteria by alkaline lysis and purified using equilibrium centrifugation in CsCl (36 ). Purified plasmid was linearized using StuI and control RNA (cRNA) was generated by in vitro transcription from a T7 promoter using the T7 MEGAscript kit (Ambion) as described (18 ). The concentration of cRNA was determined from OD260 nm.
Quantitative RT-PCR required a first step to produce cDNA by reverse transcription of samples containing a measured amount of cRNA and total muscle cell RNA. This was followed by PCR of a dilution series of these samples using a radiolabeled primer. The detailed conditions used for reverse transcription and PCR have been described (37 ). The RT-PCR products were separated by non-denaturing polyacrylamide gel electrophoresis. The gels were dried and autoradiograms obtained. A phosphorimager and imagequant software (Molecular Dynamics) were used to quantitate the amount of control and endogenous RT-PCR products.
Cultured myogenic cells were washed twice with PBS-CIT and scraped into 1 ml of fresh PBS-CIT. The cells were collected by centrifugation and resuspended in 200 [mu]l of solubilization buffer (0.1 M Tris, pH 8.0, 1% SDS, 10 mM EDTA, 10 mM DTT, 0.01 mM PMSF, 5 [mu]g/ml aprotinin, 0.1 mM benzamidine, 0.5 [mu]g/ml leupeptin, 0.5 [mu]g/ml trypsin-chymotrypsin inhibitor and 0.025 [mu]g/ml pepstatin A). Cells were then passed through a 25 gauge needle three times, boiled for 2 min and centrifuged for 4 min at 10 000 r.p.m. The supernatant was aliquoted and stored at -80oC.
Powder of frozen muscle tissue and brain tissue were made using a mortar and pestle pre-cooled on dry ice. Crushed tissue (~50-100 mg) was added to 500 [mu]l of solubilization buffer and vortexed. The solution was passed through a 25 gauge needle three times, boiled for 2 min and centrifuged for 4 min at 10 000 r.p.m. The supernatant was aliquoted and stored at -80oC.
Protein concentration was determined by the Lowry assay (38 ). Protein extract from cultured myogenic cells (60 [mu]g) and protein extract from muscle tissue and brain tissue (10 [mu]g) were size-fractionated by gel electrophoresis using a 7% SDS-polyacrylamide gel. Proteins were transferred to nitrocellulose using the Mini-Transblot Electrophoretic Transfer Cell (BioRad). Non-specific binding sites were blocked by shaking the blot in TBS-T [10 mM Tris-HCl (pH 8.0), 150 mM NaCl, 0.1% (v/v) Tween-20] containing 5% dry skim milk overnight at 4oC. The blot was then washed once in TBS-T for 15 min at room temperature and twice for 5 min with fresh buffer at room temperature. The primary antibody was diluted (1:1000 for NME-1 and 1:300 for Mandra1) in TBS-T containing 1% dry skim milk and the blot incubated for 2 h at room temperature in this mixture. The monoclonal antibody Mandra1 was generously provided by G. E. Morris and detects an epitope within amino acids 3558-3684 in the C-terminal domain of dystrophin (39 ). It does not detect utrophin and has been shown to detect the originally described isoform of Dp71 which is missing exons 71 and 78 of the dystrophin gene (10 ). The polyclonal antibody NME-1 is specific to Dp71 exon 1 (G. Y. Dally and P. N. Ray, manuscript in preparation). The membrane was then washed as described above and incubated in TBS-T containing 1% dry skim milk and a secondary antibody linked to horseradish peroxidase (Amersham) diluted 1:2000. The signal was revealed by enhanced chemiluminescence (Amersham).
We thank C. L. Duff for preparation of the human myogenic clones. We also thank Glenn E. Morris for his generous gift of the monoclonal antibody Mandra1. This research was supported by grants from the Medical Research Council of Canada and the Canadian Genetic Diseases Network. C.N.T. was the recipient of a postgraduate scholarship from the Natural Sciences and Engineering Research Council of Canada.
Human Molecular Genetics
Pages
Introduction
Results
Detection and splicing of Dp71 transcripts in fetal myogenic cultures
Expression of Dp71 RNA during differentiation of fetal myogenic cultures
Dp71 mRNA half-life in myogenic cultures
Dystrophin and Dp71 protein accumulation during differentiation of fetal myogenic cultures
Discussion
Materials And Methods
Human myogenic cell culture
RNA preparation
Construction of the control template Dp71-1-79
Preparation of cRNA by in vitro transcription
Quantitative RT-PCR
Preparation of protein extracts
Western blot analysis
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