Introduction

Duchenne muscular dystrophy (DMD) and Becker muscular dystrophy (BMD) are X-linked recessive neuromuscular diseases which are caused by mutations in the gene coding for the 427-kD cytoskeletal protein dystrophin [1, 2]. Approximately 65% of patients have deletions or duplications of one or several exons easily identified by Southern blot analysis or multiplex PCR [3, 4]. In the remaining affected individuals, the disease is presumably due to smaller mutations, the identification of which is hampered by the large size of the coding region (11 kb, 79 exons).

Heteroduplex analysis is a technique based on the formation of heteroduplexes between a wild-type allele and a mutant allele amplified by PCR which allows detection of point mutations [5]. Heteroduplex fragments can be distinguished from the corresponding homoduplexes by electrophoresis on Hydrolink-MDE gel, a modified polyacrylamide-based vinyl polymer. In order to screen for point mutations of the dystrophin gene, we performed heteroduplex analysis on both genomic multiplex PCR products and restricted RT-PCR products of illegitimate transcripts.

Materials and Methods

Patients

A total of 50 (33 DMD and 17 BMD) unrelated Italian patients having no detectable deletions or duplications within the dystrophin gene by both Southern blot analysis and multiplex PCR using 18 primer sets (for promoter and for exons 3, 4, 6, 8, 12, 13, 17, 19, 43, 44, 45, 47, 48, 50, 51, 52, 60) were selected. These patients were characterized at the IRCCS H S. Raffaele, Milano, Ospedali Riuniti, Bergamo and UILDM, Modena. Heteroduplex analysis on genomic PCR products was carried out on 40 exons in all 50 patients. Restricted RT-PCR heteroduplex analysis (RRTH) of illegitimate transcripts was performed on 18 exons in those patients (14 DMD and 4 BMD) in whom blood samples could be obtained for RNA extraction.

RNA Preparation

Peripheral blood lymphocytes were separated from 10 ml of whole blood by centrifugation through a layer of Lymphoprep (Nycomed). Total RNA was extracted from peripheral blood lymphocytes and from skeletal muscle using RNAzol™ B following the manufacturer’s instructions (Cinna Biotecx).

PCR Amplification

Genomic DNA (500 ng) was amplified in a final volume of 100 µl containing 0.75 mM dNTPs, 50 pmol of each primer, 10 µl of 10 × DyNA Zyme buffer, 2 U of DNA Polymerase Dyna™ Zyme (Fynnzymes Oy). Exons 3, 8 and 19 were amplified separately, while other exons were grouped as follows: (4, 44, 48, 51), (16, 46, 49), (6, 43, 50), (13, 47, 52), (12, 17, 45), (2, 29, 32), (7, 41, 42), (25, 54, 55), (20, 60, 65), (64, 76), (34, 68, 74, 78), (70, 72, 75). Primer sequences were derived from Chamberlain et al. [3], Malhotra et al. [6], Beggs et al. [4, 7], Kunkel et al. [8], Roberts et al. [9], Covone et al. [10] and Bebchuk et al. [11]. Thirty cycles of PCR (94°C for 30 s, 54°C for 1 min, 65 °C for 3 min) were carried out, followed by a final extension (65 °C for 3 min) in a Perkin-Elmer 9600 Thermal Cycler.

Nested RT-PCR

Analysis of dystrophin mRNA in PBL was performed according to the method described by Roberts et al. [12] slightly modified. The kit GeneAmp RNA PCR (Perkin-Elmer) was used for reverse transcriptase reactions.

Heteroduplex Analysis

Heteroduplex analysis was carried out as previously described [13].

Restricted RT-PCR Heteroduplex Analysis

RT-PCR sequences, corresponding to reactions 3 and 7[12], were analysed by the program DNA-Strider to obtain fragments ranging from 200 to 500 bp. Region 3 (position: 2344–3594; from exon 17 to exon 25) was digested with the enzyme MspI producing 205-, 294-, 341- and 411-bp fragments. Region 7 (position: 6431–7657; from exon 43 to exon 51) was digested with the enzyme HinfI yielding 341-, 406- and 480-bp fragments. Restriction products were analysed by heteroduplex analysis.

Direct Sequencing

0.2 pmol of PCR products were incubated with 10 U of exonu-clease I (USB) and 2 U of shrimp alkaline phosphatase (USB) at 37 °C for 15 min. After enzyme inactivation, treated PCR products were sequenced using the Sequenase system (USB). Sequencing reactions were carried out on both strands. Primers used were the same as for PCR or internal oligonucleotides (sequences available on request).

Results

Table 1 summarizes the mutations identified by this study: 5 were found in DMD patients and 2 in BMD patients. Mutation No. 4 at the 5′ donor splice site of intron 44 was detected by RT-PCR. The product of reaction 7 was shorter than expected, suggesting that one or more exons had been eliminated from the patient’s transcripts. Digestion of the RT-PCR products from this patient and from a male control with MspI indicated the absence of exon 44 in the patient. DNA sequence analysis showed a G to A transition at position +1 of the 5′ donor splice site of intron 44 (fig. 1). Loss of exon 44 causes a frameshift introducing a stop codon at position 6693 in exon 45.

Fig. 1
figure 1

Direct sequence of the boundary between exon 44 and intron 44 in a normal male control and in patient No. 4.

Full size image
Table 1 Small mutations identified in the dystrophin gene
Full size table

All other mutations were identified as heteroduplex bands by electrophoresis in a Hydrolink-MDE gel. Direct sequencing of the shifted heteroduplex band in patient No. 1 revealed a 4-bp deletion at position 2669–2672 (exon 20), leading to a frameshift and to the formation of a stop codon at position 2742. The 4-nucleotide (GAGA) deletion was localized within a short dinucleotide repeat sequence (GAGAGA). In DMD patient No. 5 direct-sequence analysis showed a deletion of one of the five cytosines present at position 8290–8294 (exon 55), giving rise to a stop codon at position 8382. The origin of both deletions may be due to misalignment during DNA replication or mismatching during meiotic recombination.

In patient No. 2 a heteroduplex band was revealed by RRTH (fig. 2). Direct sequencing of ectopic mRNA showed that the patient’s transcripts contained 8 bp normally present in intron 22, but lacked the first 4 nucleotides of exon 23. As this rearrangement involves the boundary region between intron 22 and exon 23, we also analysed the patient’s DNA. Direct sequencing revealed a 26-bp deletion, extending from nucleotide −21 of intron 22 to the first 5 nucleotides of exon 23 (fig. 3a). The intronic portion of the deleted segment has a palindromic sequence which might have favoured an intramolecular recombination event through the formation of a stem and loop structure (fig. 3b). As the deletion eliminates the 3′ splice site of intron 22, a nearby cryptic 3′ splice site, located 32 nucleotides upstream is activated. The consequence of this rearrangement is a frameshift which produces a stop codon at position 3244.

Fig. 2
figure 2

RRTH performed on region 3 of dystrophin cDNA (see Materials and Methods). Lane 1–14: restriction pattern from digestion of RT-PCR products with the enzyme MspI from a normal control (lane 1) and from DMD or BMD patients (lane 2–14). In lane 6 the DMD patient No. 2 shows a heteroduplex band which migrated slowlier with respect to the 411-bp homoduplex band. On the left, the fragment length is shown.

Full size image

Direct sequencing of the heteroduplex band in DMD patient No. 6 revealed a C to T transition at position 10379 (exon 70). This substitution introduces a termination site in a codon which normally codes for an arginine.

Fig. 3
figure 3

a Sequence of the boundary region between the end of intron 22 and the beginning of exon 23 from a normal control (top) and from the deleted DMD patient (bottom). Nucleotides where the recombination event likely occurred are boxed. Normal and cryptic splice sites are underlined, b Possible secondary structure of the palindrome present at the boundary between intron 22 and exon 23.← = Beginning of exon 23; * = recombination site in patient 2.

Full size image

Both the mutations identified in Becker patients No. 3 and 7 are nonsense. In exon 74, a C to T transition at position 10685 results in the substitution of glutamine 3493 by a stop codon and the subsequent deletion of half the carboxy-terminal domain. In exon 25 a C to T transition also results in the substitution of a glutamine by a stop codon. The truncated protein of only 1102 aa lacks 3/4 of the rod-shaped domain and the entire cysteine-rich and carboxy-terminal domains. The milder phenotype of these patients is inconsistent with the nature of the mutations we found, both of which are expected to lead to severe phenotypes. For the exon 74 nonsense mutation, this paradox may be explained by alternative splicing in exons 71, 72, 73 and/or 74 giving rise to different isoforms of dystrophin [14, 15]. The functional significance of these isoforms has not been clarified yet, but it is likely that they can partially complement the lack of muscular dystrophin. In our patient, the nonsense mutation might favour the event of exon skipping naturally occurring in exon 74, and an overexpression of an isoform internally deleted at exon 74 could allow partial restoration of the severe phenotype.

In the case of patient No. 3 we first postulated that the exon 25 mutation induced skipping of the exon itself. This phenomenon has been shown in a number of genes, such as those for fibrillin and for AMP deaminase [16, 17]. We therefore analysed dystrophin mRNA in ectopic transcripts in the patient, his mother and his sister, who were carriers for the mutation. RT-PCR produced the expected 647-bp fragment containing exon 25 and a 491-bp fragment lacking exon 25. The analysis showed that the 491-bp band was present also in normal individuals, indicating that the exon 25 skipping represents a spontaneous event of alternative splicing in the dystrophin gene (fig. 4). RT-PCR control on skeletal muscle mRNA also showed the presence of transcripts lacking exon 25 indicating the physiological expression of this novel isoform (fig. 4).

Fig. 4
figure 4

Alternative splicing of exon 25. Total RNA from a normal individual was amplified by nested RT-PCR. The first round of PCR, accomplished with primers DMD 3e (CCATCAGAGCCAACAGCAAT) and DMD 4e (CTCTTCAACTGCTTTCTGTA), produced a 1,019-bp fragment. Nested PCR was performed with primers DMD 3g (GCTTTACAAAGTTCTCTGCA) and DMD 4e; the reaction produced a 647-bp fragment containing exons 23, 24, 25 and 26 and a 491-bp fragment lacking exon 25. Lane 1: products of amplification from PBL RNA; lane 2: products of amplification from skeletal muscle RNA. Lane 3: pBR328 dig. Bg/I + pBR328 dig. HinfI.

Full size image

Discussion

We used heteroduplex analysis of 40 exons of the dystrophin gene to search for mutations in 50 unrelated Italian DMD and BMD patients. By this technique we were able to identify 7 novel small mutations. The mutations were randomly spread throughout the gene, confirming the recent observations of Prior et al. [18]. All the mutations detected in DMD patients — a nonsense, a splice site and three frameshift mutations — result in truncated proteins. The severity of the disease is not influenced by the extent of the truncated dystrophin, as the absence of the carboxy-terminal domain is sufficient to determine the DMD phenotype. The cysteine-rich and the carboxy-terminal domains bind to membrane-associated proteins and glycoproteins denominated dystrophin-associated proteins [19]. It has been suggested that dystrophin acts by forming a link between the actin cytoskeleton and the dystrophin-associated proteins, which in turn interact with the extracellular matrix. In DMD patients this link is prevented by the absence of the COOH-terminal regions.

The two nonsense mutations identified in BMD patients are located in exons normally undergoing alternative splicing. In these patients the presence of internally deleted isoforms can partially correct the severe phenotype of the disease. It will be of interest to establish which is the physiological role of these isoforms, if any, to understand the various functions of dystrophin at different stages of development and in separate tissues. Our data agree with the reading frame hypothesis, which associate the DMD phenotype with the presence of a truncated dystrophin and the BMD phenotype with an internally deleted or duplicated dystrophin in which the translational reading frame is maintained [20].

Finally, the novel technique used in this study, RRTH, allows simultaneous screening of several exons, and, since analysis is performed on mRNA and thus only on coding regions, simplifies detection of small mutations within genes.