A different spectrum of DMD gene mutations in local Chinese patients with Duchenne/Becker muscular dystrophy

Duchenne muscular dystrophy (DMD) and Becker muscular dystrophy (BMD) are X-linked recessive, allelic disorders.^1,2 They are the most common form of hereditary muscular dystrophy, with an incidence of about 1/3300 male births.³ DMD is referring to the more severe phenotype with delayed walking in early childhood, pseudohypertrophy of muscles, a sky high serum creatine kinase level and progressive deterioration in muscle power; and most patients become confined to a wheelchair by the age of 12.³ BMD is a milder phenotype with later onset of muscle weakness and slower progression; and most patients remain ambulatory beyond the age of 12.³ Immunohistochemical staining on muscle biopsy specimen typically shows complete absence of dystrophin in DMD and a fainter and patchy dystrophin staining in BMD.⁴ It is also well known that about 20% of patients have mental retardation.³

The gene, DMD, was mapped to chromosome Xp21. It is by far the largest human gene ever known; it consists of 79 exons and spans more than 2.3 Mb.⁵ Since the cloning of the gene in 1987,⁶ exon deletions and duplications were found to be the most common molecular defects underlying the disease, accounting for about 60% and 6% of cases respectively.^6-10 They were found to cluster at two hotspots, a 5' minor hotspot and a 3' major hotspot.⁹ Traditionally Southern blotting was employed to detect these deletions, and 6－8 hybridizations with cDNA probes were required for sufficient coverage of the gene. With the advent of the multiplex polymerase chain reaction (mPCR) technique, it became the more favored choice, because it was a much faster and less labour demanding technique. It was designed to cover 18－20 exons at the two hotspots and was shown to detect 98% of the deletions identified by Southern blotting.¹¹ The rest of the cases were caused by heterogeneous small mutations scattered along the gene.

The Clinical Genetic Service (CGS) is a tertiary referral centre and is the only centre in Hong Kong that provides DMD gene analysis as a service. Since the availability of the test in 1996, it had been noted that the percentage of patients with exon deletions in this locality was lower than the 60% commonly quoted in the literature. However, whether this observation is statistically significant could not be confirmed until a sizable number of cases were collected. The present study was a retrospective review of the clinical and molecular findings in Hong Kong Chinese patients with the diagnosis of DMD or BMD, with the objectives of studying the spectrum of DMD mutations and looking for any genotype-phenotype correlation. This was the first series of Chinese patients who had gone through such comprehensive DMD gene analysis and gave a more complete picture of DMD/BMD in Chinese.

METHODS

Patients
Patients selected in this study had been referred from paediatricians and neurologists across the territory since 1996, when DMD molecular testing became available in our laboratory, till the end of 2004.

These patients were reviewed by a clinical geneticist and had to satisfy the following criteria: (1) clinical diagnosis of DMD or BMD was reliable; (2) DMD gene testing had been performed; and (3) being Chinese. The patient files were reviewed and whether they had a DMD or BMD phenotype was determined as follows.

DMD: wheelchair bound by age of 12; not yet age 12, but total absence of dystrophin by immunohistochemical staining on muscle biopsy. BMD: ambulatory beyond age 12; not yet age 12, but both clinical severity and immunohistochemical staining result support a BMD phenotype. Indeterminate: not yet age 12, and no documentation of immunohistochemical staining.

The following data were also extracted from the patient files: sex, mental retardation, and family history.

We had tested 67 local Chinese index patients, 65 were male and 2 were female. The distribution of muscle phenotype was 30 DMD to 15 BMD to 22 indeterminate. Mental retardation was present in 14 (22.2%) of 63 patients who had documentation of mentality. A positive family history was present in 19 (28.4%) of all 67 patients.

Methods
Detection of exon deletions and duplications
Multiplex PCR (mPCR) was the first line investigation that all of the 67 patients had received. We modified the test to make it semi-quantitative, so that we could also detect exon duplications and identify female carriers. Multiplex ligation-dependent probe amplification (MLPA), a new technique that detects exon deletions and duplications and has the advantage of whole gene coverage, was introduced only in 2004.

Semi-quantitative mPCR: Twenty pairs of fluorescence-labelled primers, divided into 3 sets, were used. Set I covered exons 4, 8, 12, 17, 19, 44, 45, 48 and 51. Set II covered exons 1 (and the muscle-type promoter), 3, 6, 13, 43, 47, 50, 52 and 60. And set III covered exons 42 and 53. Primer pairs for set I and II were as described by others previously.^11-13 PCR was carried out in a final volume of 15 µl using 100 ng genomic DNA extracted from peripheral blood, 1×FX buffer, 0.2 mmol/L of each dNTPs, 0.3 µmol/L of each primer pair and 0.2 unit AmpliTaq Gold polymerase. For set I, the PCR condition consisted of an initial denaturation at 94˚C for 12 minutes, followed by 19 cycles of 94˚C for 45 seconds, 63˚C for 4 minutes and 63˚C for 10 minutes. For set II, the PCR condition consisted of an initial denaturation at 94˚C for 12 minutes, followed by 18 cycles of 94˚C for 1 minute, 59˚C for 1 minute and 70˚C for 3 minutes, and a final elongation at 70˚C for 10 minutes. For set III, the PCR condition consisted of 25 cycles of denaturation at 94˚C for 48 seconds, annealing at 59˚C for 48 seconds, and extension at 72˚C for 2.5 minutes, followed by a final extension at 72˚C for 10 minutes. The PCR products of set I and II were subject to Genescan analysis in an ABI PRISM 310 Genetic Analyzer. The quantity of the PCR products was determined by the fluorescence signal intensity. The signal intensity from each exon in a multiplex was divided by that from each of the other exons, and this ratio was compared to those obtained from normal control samples. In male patients, a deletion would simply lead to absence of signal from the respective exons. Duplication would lead to a roughly two-fold increase in signal. In female carriers, a deletion would lead to a roughly 1/2 decrease in signal, while a duplication would lead to a roughly 1.5 fold increase in signal. The PCR products of set III were simply electrophoresed in 1.8% agarose gel.

MLPA analysis: The MLPA DMD test kit (SALSA P034/P035) was obtained from MRC-Holland, the Netherlands. The 80 probes were divided in two probe mixes P034 and P035. Total 50-500 ng of genomic DNA extracted from peripheral blood was used. MLPA reactions were performed according to the manufacturer's instructions. PCR products were analyzed on an ABI PRISM 3100 Genetic Analyzer. The peak area of each fragment was compared to that of a control sample. The manufacturer recommended that one determine the relative peak area of each fragment by dividing each measured peak area by the sum of all 45 peaks of that sample. This relative peak area was then divided by that of the corresponding probe obtained from a control DNA sample. In male patients, a deletion would result in total absence of signal from the corresponding exons, while duplication would result in a two-fold increase. In female patients or carriers, a deletion would result in a 35%－55% decrease in signal of the corresponding exons, while duplication would result in a 30%－55% increase.

Detection of small mutations
Denaturing high performance liquid chromatography (DHPLC) screening followed by sequencing was introduced in 2003.

DHPLC: Primers were designed to cover each of the 79 exons as well as the flanking intronic regions. Totally, 86 segments of size ranging from 170 to 600 base pairs were analyzed. The sequence of the primers and the respective annealing temperatures for PCR were obtained from the Leiden Muscular Dystrophy website (http://www.dmd.nl). For male patients, PCR amplicons were mixed in a 1:1 ratio with those from an unaffected male (wild type). For female patients or carriers, PCR amplicons were analyzed directly. The amplicon/mix was heated to 95˚C in a thermocycler for 5 minutes, followed by a slow reduction in temperature (at a rate of －0.1˚C/4 seconds.) until a temperature of 25˚C was reached. The hetero- and homoduplexes thus formed were analyzed at pre-determined temperatures in the WAVE^® system (Transgenomic Inc., USA). The resultant chromatograms were compared with the wild type for any variation in shape or retention time. All PCR amplicons with a chromatogram different from the wild type were subjected to direct DNA sequencing.

Direct sequencing: Purified PCR products were subjected to cycle sequencing in a 20 µl reaction with the ABI BigDye Terminator V1.1 cycle sequencing kit. The purified extension products were then sequenced on an ABI PRISM 310 or 3100 Genetic Analyzer.

Statistical methods
Binomial probability was calculated to determine if the percentage of exon deletions was significantly less than 60%.

Chi-square (χ²) and Fisher exact tests were applied, where appropriate, to study genotype-phenotype correlation.

RESULTS

Exon deletions and duplications
Of the 67 patients, 28 (41.8%) were found to have exon deletions (n=23, 34.3%) and duplications (n=5, 7.5%) (Table 1 and Fig. 1). The percentage of deletions was significantly less than the 60% commonly quoted in the literature (binomial probability P < 0.00002), while that of duplications was close to that reported by other studies. Only one deletion (del exons 65－76) lied outside the hotspots and was not detected by mPCR but was detected incidentally during DHPLC screening. Only one exon duplication (dup exon 2) was not detected by mPCR but was detected by MLPA.

view in a new window

Table 1. Exon deletions and duplications detected in 28 patients

view in a new window

Fig. 1. Schematic diagram illustrating the distribution of 28 exon deletions and duplications across the DMD gene. Vertical rectangles represent different exons; exons in black are those actually tested in mPCR. Thick horizontal lines represent different deletions and duplications. Lines in red = DMD; lines in green = BMD; lines in black = indeterminate. Twenty-six deletions/ duplications were identified by mPCR. An exon 2 duplication was identified by MLPA. A deletion encompassing exons 65-76 was identified by PCR during DHPLC screening and was subsequently confirmed by MLPA.

Small mutations
Small mutations were found in 23 (34.3%) patients, including 13 (56.5%) nonsense mutations, 4 (17.4%) splice site mutations, 4 (17.4%) small deletions and 2 (8.7%) small insertions (Table 2 and Fig. 2). Twenty-one mutations were predicted to result in a truncated protein. Two splice site mutations (patients 30 and 34) were predicted to result in exon skipping without frameshift. Twenty-two mutations were identified by DHPLC screening followed by sequencing in 31 patients who had negative findings after mPCR testing. One small deletion (c.2086_2098del13) was detected incidentally by mPCR.

view in a new window

Table 2. Small mutations detected in 23 patients

view in a new window

Fig. 2. Schematic diagram illustrating the distribution of the small mutations. Arrows in red = DMD; arrows in green = BMD; arrows in black = indeterminate; broken arrows = mentally retarded. The 13 bp deletion in exon 17 was identified by mPCR. The rest were identified by DHPLC screening following by sequencing.

No mutation
Sixteen (23.9%) patients had no DMD gene mutation found, 7 of which had gone through testing with mPCR, MLPA and DHPLC, while 9 had only received mPCR but no further analysis because of the lack of DNA specimen.

Genotype-phenotype correlation
Correlation between the type of mutation and clinical severity (Table 3)
The patients were divided into 3 groups according to different types of mutations, namely exon deletions, exon duplications and small mutations. The relative proportion of DMD, BMD and indeterminate phenotypes in each group was compared to that of the overall distribution among all 67 patients. No significant correlation was found using χ² test (χ² = 4.6; df = 6; P = 0.596).

view in a new window

Table 3. Correlation between the type of mutation and clinical severity (n)

Correlation between the type of mutation and mental retardation (Table 4)
Only 63 patients were included in the analysis because 4 patients did not have documentation of mental status. No significant correlation was found using Fisher exact test.

view in a new window

Table 4. Correlation between the type of mutation and mental retardation (MR), positive family history and maternal carriers (n)

Correlation between the type of mutation and maternal carrier status (Table 4)
Only 44 patients whose mothers had been tested were included; the 16 patients with no mutation identified were not included because the 4 carrier mothers in this group were known only from the positive family and thus there would be an ascertainment bias. There were significantly fewer maternal carriers in the group with exon deletions (two tailed Fisher exact probability, P = 0.03).

Correlation between the type of mutation and positive family history (Table 4)
Small mutations were more likely to be associated with a positive family history (two tailed Fisher exact probability, P = 0.02).

Correlation between the location of exon deletion/ duplication and clinical severity (Table 5)
Ten mutations took place at the 5' hotspot and 15 at the 3' hotspot. Two mutations were so large that both hotspots were involved. One mutation took place outside the hotspots. It appeared that BMD was more likely to be caused by a mutation at the 5' hotspot. However, statistical significance was not attained by Fisher exact test (P = 0.09). On the contrary, DMD phenotype was significantly less common in the group with mutations at the 5' hotspot (Fisher exact probability, P = 0.046).

view in a new window

Table 5. Correlation between the location of exon deletion/duplication and clinical severity (n)

Correlation between the location of small mutation and clinical severity
The small mutations showed a widely scattered distribution in the gene and there was no apparent correlation between the location of mutation with severity (Fig. 2).

Correlation between the location of small mutation and mental retardation
All 4 mentally retarded patients among 19 with documentation of mental status had mutations located in the 3' half of the gene, in exons 58, 61, 64 and 70, respectively (Fig. 2).

DISCUSSION

We found that local Chinese patients with DMD or BMD had significantly fewer exon deletions of the DMD gene, while the percentage of duplications was close to those reported by others. A selection bias in the patient sample was unlikely because these patients were referred to our Service for DMD gene analysis from different clinicians from different institutes across the territory. These patients had not been tested for DMD gene deletions or duplications elsewhere before they were referred to us. Testing was offered as long as the patients had a compatible clinical phenotype. Furthermore, the lack of significant correlation between the type of mutation and the severity argued against the possibility of selection bias.

In fact, our finding was not new. A study of another 24 Hong Kong DMD families using Southern blotting with cDNA probes found deletions in only 37.5%,¹⁴ which was remarkably close to the 34.3% found in the present study. A literature search for other Chinese studies of DMD gene mutations also revealed a generally low percentage of deletions, ranging from 37% to 62.3%,^15-21 with only one study finding over 60% deletions.²¹ It implied that there existed an ethnic difference in predisposition to such deletions. It was also possible that even among ethnic Chinese a regional variation existed, as suggested by the discordant results of all studies including the present one. We also looked at the results from another Mongoloid population, the Japanese. A similar pattern was found, with the percentage of deletions ranging from 33% to 64%.^22-30 Whether this ethnic and regional difference in predisposition to deletions was a result of environmental differences or genomic variations or both is a subject of future research.

We showed that DMD phenotype was significantly less common in those patients with deletions/ duplications at the 5' hotspot. On the contrary, BMD phenotype was more common, similar to what had been found by Lindlof.³¹ However, statistical significance was not attained, probably because of the small number of BMD patients. Other than this, correlation between the genotype and clinical severity among patients with exon deletios/ duplications could not be observed. This was not unexpected, because most of these mutations were identified by mPCR. The design of mPCR had its intrinsic shortcoming of incomplete coverage of the gene; less than 1/3 of the coding exons were covered. Therefore, it only allowed a crude estimation of the extent of deletions/duplications. In a separate evaluation of the efficacy of MLPA in detecting exon deletions/duplications of the DMD gene, 45.8% of the deletions and duplications were found to have a size larger than previously determined by mPCR.³² This false estimation of the extent of the deletions/duplications in turn precluded determination of the occurrence of frameshift, which is a major factor that determines the muscle phenotype. Those with frameshift are more likely to have DMD, while those without frameshift are more likely to have BMD. The correlation was as high as 92%.⁷ MLPA, with the advantage of complete gene coverage, can overcome this problem and enable more accurate estimation of the extent of deletion/duplication and hence genotype-phenotype correlation.³²

Among the 23 small mutations, including the one incidentally picked up by mPCR, 15 were novel mutations. No one mutation occurred twice within this patient sample, and they had a widely scattered distribution in the DMD gene. All 4 mutations associated with mental retardation were located at the 3' end of the gene, which was in accord with previous finding that a disruption of the translational reading frame in the C-terminal region was particularly associated with mental retardation.³³ Four (23.5%; patients 36, 42, 43 and 46) of the 17 point mutations were the result of C to T or G to A changes at CpG dinucleotides, implying that deamination of methylcytosine was a significant but not major player in the genesis of point mutations.

The overall mutation detection rate in this series was 76.1%. However, 9 patients received only mPCR. The detection would have been more than 80% if more DNA of these 9 patients was available for further analysis. Seven patients had gone through testing by mPCR, DHPLC and MLPA, and the mutations were still elusive. It was unlikely that we had missed a major deletion or duplication. Unidentified intragenic small mutations were still possible. These mutations might be located well within the intronic regions so that they were not covered by the PCR; or their locations were covered by the PCR but further adjustment in the melting temperature in DHPLC screening was required to detect them. Other possibilities were defects of control elements, genetic or epigenetic, involved in the regulation of transcription.

REFERENCES

1. Kingston HM, Thomas NS, Pearson PL, Sarfarazi M, Harper PS. Genetic linkage between Becker muscular dystrophy and a polymorphic DNA sequence on the short arm of the X chromosome. J Med Genet 1983; 20:255-258.

2. Kingston HM, Sarfarazi M, Thomas NS, Harper PS. Localisation of the Becker muscular dystrophy gene on the short arm of the X chromosome by linkage to cloned DNA sequences. Hum Genet 1984; 67:6-17.

3. Emery AEH. Duchenne and other X-linked muscular dystrophies. In: Rimoin DL, Connor JM, Pyeritz RE, Korf BR, eds. Emery and Rimoin's principles and practice of medical genetics, 4th ed. London: Harcourt Publishers Limited; 2002: 3266-3284.

4. Arahata K, Hoffman EP, Kunkel LM, Ishiura S, Tsukahara T, Ishihara T, et al. Dystrophin diagnosis: comparison of dystrophin abnormalities by immunofluorescence and immunoblot analyses. Proc Natl Acad Sci U S A 1989; 86: 7154-7158.

5. Tennyson CN, Klamut HJ, Worton RG. The human dystrophin gene requires 16 hours to be transcribed and is cotranscriptionally spliced. Nat Genet 1995; 9:184-190.

6. Koenig M, Hoffman EP, Bertelson CJ, Monaco AP, Feener C, Kunkel LM. Complete cloning of the Duchenne muscular dystrophy (DMD) cDNA and preliminary genomic organization of the DMD gene in normal and affected individuals. Cell 1987; 50:509-517.

7. Koenig M, Beggs AH, Moyer M, Scherpf S, Heindrich K, Bettecken T, et al. The molecular basis for Duchenne versus Becker muscular dystrophy: correlation of severity with type of deletion. Am J Hum Genet 1989; 45:498-506.

8. Read AP, Mountford RC, Forrest SM, Kenwrick SJ, Davies KE, Harris R. Patterns of exon deletions in Duchenne and Becker muscular dystrophy. Hum Genet 1988; 80:152-156.

9. Den Dunnen JT, Grootscholten PM, Bakker E, Blonden LA, Ginjaar HB, Wapenaar MC, et al. Topography of the Duchenne muscular dystrophy (DMD) gene: FIGE and cDNA analysis of 194 cases reveals 115 deletions and 13 duplications. Am J Hum Genet 1989;45:835-847.

10. Hu X, Ray PN, Murphy EG, Thompson MW, Worton RG. Duplicational mutation at the Duchenne muscular dystrophy locus: its frequency, distribution, origin, and phenotype-genotype correlation. Am J Hum Genet 1990;46:682-695.

11. Beggs AH, Koenig M, Boyce FM, Kunkel LM. Detection of 98% of DMD/BMD gene deletions by polymerase chain reaction. Hum Genet 1990;86:45-48.

12. Chamberlain JS, Gibbs RA, Ranier JE, Caskey CT. Multiplex PCR for the diagnosis of Duchenne muscular dystrophy. In: Innis MA, Gelfand DH, Sninsky JJ, et al, eds. PCR protocols: a guide to methods and applications. San Diego: Academic Press; 1990: 272-281.

13. Abbs S, Yau SC, Clark S, Mathew CG, Bobrow M. A convenient multiplex PCR system for the detection of dystrophin gene deletions: a comparative analysis with cDNA hybridisation shows mistypings by both methods. J Med Genet 1991;28:304-311.

14. Lau YL, Srivastava G, Wong V, Liu YT, Ho FC, Yeung CY. Deletions, duplications and novel restriction fragment length polymorphism in Duchenne and Becker muscular dystrophies. Clin Genet 1992;41:252-258.

15. Ko TM, Tseng LH, Chiu HC, Hsieh FJ, Lee TY. Dystrophin gene deletion in Chinese Duchenne/Becker muscular dystrophy patients via multiplex DNA amplification. J Formos Med Assoc 1992;91:951-954.

16. Yang J. Diagnosis of pseudohypertrophic muscular dystrophy with multiplex polymerase chain reaction (PCR). Zhonghua Shen Jing Jing Shen Ke Za Zhi (Chin) 1991; 24:322-325, 382.

17. Zeng YT, Chen MJ, Ren ZR, Qui XK, Huang SZ. Analysis of RFLPs and DNA deletions in the Chinese Duchenne muscular dystrophy gene. J Med Genet 1991;28:167-170.

18. Soong BW, Tsai TF, Su CH, Kao KP, Hsiao KJ, Su TS. DNA polymorphisms and deletion analysis of the Duchenne-Becker muscular dystrophy gene in the Chinese. Am J Med Genet 1991;38:593-600.

19. Zhang J. Genetic analysis of 60 Duchenne muscular dystrophy (DMD) or Becker muscular dystrophy (BMD) patients using dystrophin cDNA. Acta Acad Med Sini (Chin) 1993;15:399-404.

20. Yang RC, Yang SL, Jin SH, Chen HW, Jong YJ, Chen SS, et al. Detection of dystrophin gene deletion in Chinese Duchenne/Becker muscular dystrophy patients utilizing multiplex polymerase chain reaction. Gaoxiong Med J 1994;10:1-8.

21. Yuge L, Hui L, Bingdi X. Detection of gene deletions in Chinese patients with Duchenne/Becker muscular dystrophy using CDNA probes and the polymerase chain reaction method. Life Sci 1999; 65: 863-869.

22. Sugino S, Fujishita S, Kamimura N, Matsumoto T, Wapenaar MC, Deng HX, et al. Molecular-genetic study of Duchenne and Becker muscular dystrophies: deletion analyses of 45 Japanese patients and segregation analyses in their families with RFLPs based on the data from normal Japanese females. Am J Med Genet 1989;34: 555-561.

23. Nakajima T, Matsuo M, Nakamura H, Fujiwara Y. Southern blot and PCR analyses of dystrophin gene deletions in Japanese patients with Duchenne muscular dystrophy. Kobe J Med Sci 1991; 37:21-33.

24. Asano J, Tomatsu S, Sukegawa K, Ikedo Y, Minami R, Iida M, et al. Gene deletions in Japanese patients with Duchenne and Becker muscular dystrophy. Jinrui Idengaku Zasshi 1990;35:159-168.

25. Sakuraba H, Ishii K, Shimmoto M, Yamada H, Suzuki Y. A screening for dystrophin gene deletions in Japanese patients with Duchenne/Becker muscular dystrophy by the multiplex polymerase chain reaction. Brain Dev 1991;13:339-342.

26. Kitoh Y, Matsuo M, Nishio H, Takumi T, Nakajima T, Masumura T, et al. Amplification of ten deletion-rich exons of the dystrophin gene by polymerase chain reaction shows deletions in 36 of 90 Japanese families with Duchenne muscular dystrophy. Am J Med Genet 1992;42:453-457.

27. Hiraishi Y, Kato S, Ishihara T, Takano T. Quantitative Southern blot analysis in the dystrophin gene of Japanese patients with Duchenne or Becker muscular dystrophy: a high frequency of duplications. J Med Genet 1992;29:897-901.

28. Hiyama K, Kodaira M, Satoh C, Karakawa T, Kameo H, Yamakido M. Detection of partial deletion and partial duplication of dystrophin gene in Japanese patients with Duchenne or Becker muscular dystrophy. Jpn J Hum Genet 1993;38:169-176.

29. Katayama S, Takeshita N, Yano T, Ubagai T, Qiu XJ, Katagiri Y, et al. Deletion detection for diagnosis of Duchenne muscular dystrophy in the Japanese population--comparison between the polymerase chain reaction and the Southern blot analysis. Jpn J Hum Genet 1993;38:177-184.

30. Imoto N, Arinami T, Hamano K, Matsumura K, Yamada H, Hamaguchi H, et al. Topographic pattern of the rearrangement of the dystrophin gene in Japanese Duchenne muscular dystrophy. Hum Genet 1993;92:533-536.

31. Lindlof M, Kiuru A, Kaariainen H, Kalimo H, Lang H, Pihko H, et al. Gene deletions in X-linked muscular dystrophy. Am J Hum Genet 1989;44:496-503.

32. Lai KK, Lo IF, Tong TM, Lam ST. Detecting exon deletions and duplications of the dystrophin gene with a novel technique - Multiplex Ligation-dependent Probe Amplification (MLPA). Clin Biochem 2006; 39: 367-372

33. Lenk U, Hanke R, Thiele H, Speer A. Point mutations at the carboxy terminus of the human dystrophin gene: implications for an association with mental retardation in DMD patients. Hum Mol Genet 1993;2:1877-1881.