Article Text


ERG phenotype of a dystrophin mutation in heterozygous female carriers of Duchenne muscular dystrophy


PURPOSE Mutations in the dystrophin gene result in Duchenne muscular dystrophy (DMD). DMD is associated with an abnormal electroretinogram (ERG) if the mutation disrupts the translation of retinal dystrophin (Dp260). Our aim was to determine if incomplete ERG abnormalities would be associated with heterozygous carriers of dystrophin gene mutations.

METHODS Ganzfeld ERGs were obtained under scotopic and photopic testing conditions from a family which includes the heterozygous maternal grandmother, the heterozygous mother, and her children, two affected boys and dizygotic twin sibs, an unaffected male and heterozygous female. Southern blot analyses were done to characterise the dystrophin mutation.

RESULTS The dystrophin gene was found to contain a deletion encompassing exon 50. The ERGs in the two affected boys were abnormal, consistent with the DMD ERG phenotype. Serial ERGs of the heterozygous females were abnormal; however, they were less severely affected than the DMD boys. The ERG of the female sib showed a greater abnormality than her mother and maternal grandmother. The unaffected twin had a normal ERG.

CONCLUSIONS The ERG shows abnormalities associated with carrier status in this family with a single exon deletion. A large study of confirmed obligate carriers is planned to clarify further the value of the ERG in detecting female heterozygous carriers of dystrophin gene mutations.

  • muscular dystrophy
  • electroretinography
  • retina
  • dystrophin

Statistics from

Mutations in the gene encoding dystrophin, a 427 kDa membrane cytoskeletal protein, result in Duchenne muscular dystrophy (DMD) and the less severe Becker muscular dystrophy (BMD).1Expression of the dystrophin gene is under elaborate transcriptional and post-transcriptional splicing control. At least seven independent promoters driving the transcription of their alternative first exons in a cell specific and developmentally controlled manner are known.2 Several isoforms of dystrophin are produced in this manner, including full length (427 kDa) cortical, Purkinje, and muscle isoforms, as well as smaller isoforms (140 kDa CNS, 116 kDa Schwann cell, 71 kDa glial cell).3-7 Recently, a novel mouse retinal dystrophin, Dp260, was discovered.8 Mouse retinal dystrophin is derived from a novel exon spliced in frame to exon 30. Two novel human retinal dystrophin isoforms (MW 260 kDa) have been identified, cloned, and characterised in our laboratory (GenBank accession number U27203).9 The two retinal dystrophin proteins differ from each other at the N-termini that are either hydrophobic or hydrophilic and derive from the complex splicing of a unique first exon. The hydrophilic N-terminus in humans shows 100% amino acid identity to Dp260 in the mouse, while the hydrophobic N-terminus sequence in humans is not observed in mice.

Dystrophin is found in significant amounts in the invaginated photoreceptor synaptic complexes in the outer plexiform layer of the retina and appears to be restricted to the photoreceptor plasma membrane bordering the lateral sides of the synaptic invagination.10 Within this invagination are the dendritic tips of depolarising bipolar cells and horizontal cells. Western blot analyses using antisera to different dystrophin domains indicate the presence of 427 kDa, 260 kDa, and 71 kDa immunoreactive proteins in mouse and human retinal extracts.11 12

The electroretinogram (ERG) is a recording of the summed electrical signal produced by the retina in response to a flash of light. A genotype/phenotype correlation is associated with the ERG in DMD.13 An abnormal ERG is seen in DMD/BMD when dystrophin gene mutations disrupt the translation of Dp260.11 14-17In the majority of cases, mutations downstream of exon 30 have an abnormal ERG while mutations upstream of exon 30 have a normal ERG. The ERG shows the following abnormalities: rod isolatedb waves are highly attenuated, the mixed rod-cone response is negative (the a wave is larger than the b wave), and there are abnormalities in the cone mediated ERG. Negative ERGs are associated with a number of unrelated ophthalmic disorders, including X linked and autosomal recessive forms of congenital stationary night blindness, retinal dystrophies, retinoschisis, vascular disorders, retinal toxicity, paraneoplastic melanoma, and degenerative myopia.18 Unlike the above mentioned disorders, the negative ERG in DMD/BMD is not associated with ocular pathology or visual disturbances; indeed, ophthalmic examinations, contrast sensitivity, colour vision, visual fields, and rod dark adaptation thresholds are normal.15 17

Our studies of DMD patients suggest that Dp260 plays a critical role in retinal neurotransmission that is indicated by an abnormal ERG.16 In addition, one might also expect incomplete ERG abnormalities of female heterozygous carriers of dystrophin gene deletions.

In an attempt to understand the genetic mechanisms underlying the abnormal ERG in DMD, we recorded ERGs and performed Southern blot analyses in a family which includes two adult females heterozygous for DMD (maternal grandmother and mother), two affected boys, and their dizygotic twin sibs, an unaffected male and heterozygous female.



The subjects were recruited for participation in this study from The Children’s Mercy Hospital Sections of Ophthalmology and Genetics following the tenets of the Declaration of Helsinki. Informed consent was obtained after the nature and possible risks of the study were explained. The research was approved by the institutional human experimentation committee (IRB).

Subjects included the maternal grandmother (aged 63), the mother (aged 36), and her four children, two boys with DMD (aged 2 and 4 years) and dizygotic twins, an unaffected male and heterozygous female (aged 1 year). These subjects underwent an eye examination, medical examination, ERG, and Southern blot analysis.

An ERG was also recorded in a healthy, homozygous, 35 year old mother of a boy with DMD whose disease was the result of a spontaneous, non-inherited mutation. This mother was previously tested and shown to be negative for carrier status as determined by Southern blot analysis (data not shown). The purpose in recording the ERG in this mother was to compare the results of an adult DMD homozygous female to the adult DMD heterozygous females.


A blood sample was obtained from the family for DNA analysis. Blood was also obtained from a healthy, unrelated female and used as a control. After treating the samples with Bell’s blood lysis buffer, the genomic DNA was extracted using the Applied Biosystems 341 Nucleic Acid Purification System. Approximately 5 μg of the subjects’ genomic DNA was digested with HindIII,BglII, and TaqI. Southern blots and hybridisation were performed as previously described.19 The Southern blots were probed with fragments of the dystrophin cDNA including cDMD1-2a, cDMD2b-3, cDMD4-5a, cDMD5b-7, cDMD8, and cDMD9-10.20 21 The dystrophin cDNA clones were obtained from the American Type Culture Collection. Radioactive probes were generated by random primer labelling (Ambion) of agarose gel electrophoresis purified DNA fragments. Deletions in the dystrophin gene were evaluated by comparison of the autoradiographic patterns with normal controls and the published map of dystrophin genomic DNA.20 22 23 The relative signal of bands on Southern blots were quantitated using Imagequant software and a Molecular Dynamics Phosphoimager. The graph of relative signal was constructed by comparing the ratio of the signal for one band to another band.


Details of the standardised clinical ERG and long duration photopic ERG are described in detail in previous publications.14 16 Briefly, the clinical ERG was recorded under both scotopic and photopic testing conditions. All of the children were sedated with 50 mg/kg oral chloral hydrate syrup for the duration of the test. The ERGs were compared to pooled age and gender matched normal control subjects.

The stimulus flashes and testing conditions were designed to separate rod and cone systems. The rod isolated responses were recorded under scotopic testing conditions using a dim blue stimulus. Mixed rod-cone responses were recorded using higher intensity white stimuli under scotopic testing conditions. After 10 minutes of light adaptation, cone mediated responses were recorded to short (<10 msec) or long (100-200 msec) duration white stimuli under photopic testing conditions. The long duration stimuli separate contributions of the depolarising (ON) and hyperpolarising (OFF) bipolar cells to the photopic ERG.24

The b wave amplitude of the rod isolated response is recorded from the baseline of the wave to the peak of theb wave. The awave amplitude of the mixed rod-cone response was measured from the baseline to the trough of the a wave. The b wave was measured from the trough of the a wave to the peak of theb wave. The b/aamplitude ratio was obtained by dividing bwave amplitude by a wave amplitude. ON response amplitudes were measured from the trough of thea wave to the peak of the ON response (orb wave) and the amplitude ratio was obtained as previously described.



The dystrophin gene was found to contain a deletion in the two male sibs with DMD. After comparison of a series of Southern blots with normal autoradiographic patterns for BglII,HindIII, and TaqI restriction enzyme digests of the dystrophin gene (data not shown), it was determined that a 2.6 kb BglII genomic fragment including exon 50 was deleted (fig 1).

Figure 1

(A) Family pedigree. Open symbols=normal; closed symbols=DMD; half shaded symbols=dystrophin deletion carrier. (B) To the left, Southern blot of genomic DNA from female control and GH and CH (DMD males). Absence of the 2.6 kb genomic band, which includes exon 50, is observed for the two males. To the right, Southern blot of genomic DNA from SH, MH, a female control, BH (normal male twin), and CH (as a control for the 2.6 kb deletion). Family members SH and MH appear to be carriers for the 2.6 kb deletion. Deleted 2.6 kb genomic fragment which corresponds to exon 50 is indicated. Data from the grandmother are not shown.

Deletions of this fragment are consistent with the occurrence of DMD.25 The normal male sib was found to harbour the normal 2.6 kb genomic band in his DNA. The maternal grandmother and mother of the DMD boys were analysed with respect to the 2.6 kb deletion to determine carrier status. A quantitative analysis was performed for the maternal grandmother, mother, and dizygotic twin female in which the 2.6 kb fragment was compared to that of normal female genomic DNA. Results indicate that the signals generated for the 2.6 kb band in the three heterozygous carriers were 50% of the normal signal (fig 2) (data from maternal grandmother not shown).

Figure 2

(A) Southern blot of genomic DNA from SH, a female control, and MH. Genomic fragment corresponding to exon 50 (2.6 kb) and the normal non-deleted 3.5 kb genomic band are indicated. (B) Graph of relative intensity of deleted 2.6 kb fragment in control, SH, and MH indicating 1/2 normal signal. Data from the grandmother are not shown.

This indicates that the maternal grandmother and mother are carriers for the dystrophin gene deletion and that the 2.6 kb deletion is not a de novo mutation. The dizygotic twin female is also a carrier for this deletion which led to a DMD phenotype in her brothers.


The ERGs of the dizygotic twins and one of the affected brothers (CH) are shown in fig 3A.

Figure 3

(A) ERG of normal male twin (BH) at 1 year of age, his twin sister (MH) at 1, 2, 3, and 5 years of age, and their brother (CH) who has DMD. See text for ERG details. Arrows define the O1 and O2 photopic oscillatory potentials. (B) The ERGs of the adult heterozygous female (SH) at two test dates, the ERG of the maternal grandmother (FT), and the ERG of a homozygous female (FF) whose son has DMD as the result of a spontaneous, non-inherited mutation. The most consistent ERG finding in heterozygous females for DMD is a decreased photopic ON response b/a amplitude ratio. The dotted line represents stimulus baseline. The square wave above the msec indicator represents the duration of the stimulus.

The ERG of the unaffected male twin (BH) shows normalb wave amplitudes to both scotopic stimuli. The b/a amplitude ratio of the mixed rod-cone response is 2.3 (normal= mean 2.0, SD 0.4). The response to the short duration photopic stimulus is normal and both O1 and O2 oscillatory potentials are clearly present (arrows). The photopic response to a long duration photopic stimulus of 100 msec shows a typical ON response (or b wave) that rises above the stimulus baseline with a ratio of 2.0. The mean normal ON response b/a amplitude ratio for this age group is 2.0 (SD 0.4). His ERG typifies that of normal subjects in this age group.

The affected 2 year old brother, CH, has ERG findings consistent with dystrophin gene mutations downstream of exon 30. Under scotopic testing conditions, the rod isolated b wave is nearly absent. The b wave of the mixed rod-cone response is reduced resulting in ab/a amplitude ratio of 0.76. The mean normalb/a amplitude ratio for this age group is 1.8 (SD 0.5). The response to the short duration photopic stimulus is missing the O2 oscillatory potential. The response to the long duration photopic stimuli shows a highly attenuated ON response that does not rise above the stimulus baseline with a b/aamplitude ratio of 0.8.

Serial ERGs were recorded in the heterozygous twin sister (MH) at 1, 2, 3, and 5 years of age. At the age of 1 year, the rod isolated response shows a highly attenuated b wave of only 60 μV (mean normal=209 (SD 55) μV). Theb/a amplitude ratio of the mixed rod-cone response is reduced at 1.6 but still within 1 SD. The O2 oscillatory potential to the short duration photopic stimulus is missing. The ON response amplitude of the long duration photopic stimulus is reduced and barely rises above the stimulus baseline with ab/a amplitude ratio of 1.1.

At the age of 2 years, the b wave of the rod isolated response is still attenuated at 100 μV (mean normal=215 (SD 60) μV). The b wave implicit time of the mixed rod-cone response has decreased appropriately with age but the amplitude ratio remains at the lower limits of normal at 1.3. The photopic O2 oscillatory potential is now recognisable but smaller than O1. The photopic ON response ratio is 1.0.

At 3 and 5 years of age, all parameters of the ERG under both photopic and scotopic testing conditions are unchanged. It should be noted that our recording parameters changed for the long duration stimulus to 200 msec ON, 100 msec OFF and the recording sensitivity was increased to visualise the response better. This changes the appearance of the waveform in the illustration but the photopic ON response ratio is still 1.0.

Fig 3B depicts the ERGs of the adult heterozygous female SH, mother of the twins and affected boys described previously. Two ERGs were recorded at different test dates. The ERG of the maternal grandmother (FT) is also shown. For comparison we show the ERG of an adult homozygous female (FF) whose son has DMD as the result of a spontaneous mutation.

The ERG of the heterozygous mother at both test dates does not show the obvious abnormalities that were seen in her daughter under scotopic testing conditions. The rod isolated response is normal. The mixed rod-cone response is normal with a b/aamplitude ratio of 1.8. The amplitude ratio of the pooled normal data in this age group is 1.7 (SD 0.2). However, there is attenuation of the photopic O2 oscillatory potential and the long duration photopic ERG is similar to that of her daughter. The ON response amplitude is decreased. Consequently, the b/a amplitude ratio of the ON response is reduced at 1.2.

The ERG of the maternal grandmother is similar to her daughter’s under scotopic testing conditions. The rod isolated response is normal and the mixed rod-cone response is normal with ab/a amplitude ratio of 1.9. Again, the O2 oscillatory potential and ON response amplitude are attenuated, leading to a b/a amplitude ratio of 1.2.

In contrast, the ERG of FF, the homozygous female, is normal to all stimuli under both scotopic and photopic testing conditions. Her scotopic amplitude ratio is 2.1. The O2 photopic oscillatory potential is present and the photopic ON response is normal with ab/a amplitude ratio of 2.1.


The eye examination included determination of acuity, accommodation, refractive errors, motility, alignment, transillumination of the irides, and fundus examination. All subjects had a normal eye examination and none showed increased macular pigmentation as described by Sigesmund et al.13


The affected male children exhibited the classical signs of DMD, including muscle weakness, calf hypertrophy, and raised serum CK levels of 16 750 IU in the 4 year old and 18 000 IU in the 2 year old. The CK levels of the heterozygous female twin were 2000 IU at 1 year of age and 3443 IU at 5 years of age. CK levels of the mother were 246 IU.

In order to rule out the possibility that the mother and daughter may be manifesting carriers, a physical examination was performed. Physical examination of 5 year old MH showed an alert, active girl with no dysmorphic features. She had normal muscle bulk and tone, no muscle hypertrophy, and normal strength on formal manual muscle testing. Deep tendon reflexes were normal bilaterally and symmetrically in all extremities. Functional examination showed normal gait, stance, and transitions and age appropriate gross motor and fine motor skills. Her developmental history was unremarkable, having achieved independent sitting by 6 months of age and walking by 1 year. Her gross motor milestones were all achieved within a normal sequence and time frame. She does well in preschool and has required no special services. She keeps up with her peers and has normal endurance for prolonged activities.

SH, the mother of MH, similarly showed no clinical signs of muscle weakness with normal strength, normal muscle bulk and tone, intact deep tendon reflexes, and no history of developmental delay, problems with endurance, or any history of deterioration of function over time.


We have shown incomplete ERG abnormalities in three generations of heterozygous female carriers of a dystrophin gene mutation that results in the deletion of exon 50. The most consistent ERG abnormality in the heterozygous females of this family was the reducedb/a amplitude ratio of the ON response to long duration photopic stimuli. All of the heterozygous females showed ON response amplitude ratios of 1.2 or less. The photopic oscillatory potential, O2, is attenuated to the standard short duration clinical stimulus, but this is a subtle finding and difficult to assess. Use of a long duration flash allows the separation of the underlying photopic ON and OFF responses and allows us to evaluate each response independently.

In our study, the scotopic b wave was not a good indicator of carrier status. Scotopic bwaves were normal in the maternal grandmother and mother but were consistently reduced in the daughter at the four test dates. The reduced scotopic b waves in the daughter were not the result of immaturity of the retina or the effect of chloral hydrate sedation, since her unaffected twin brother (also sedated) has adult-like b wave amplitudes at 1 year of age. Our normative data for 1 year of age (n=110) shows scotopic b waves reaching 90% of adult values by this age. What causes the reduced scotopicb wave in the daughter and not in the adult heterozygous females?

One notable difference between the mother and daughter is the significantly higher CK level. However, CK levels should not influence the ERG. CK levels are grossly raised (50-100 times normal) in DMD/BMD boys, yet the ERG phenotype may be normal or abnormal. The difference in ERG phenotype in affected boys is dependent upon the nature of their mutation. If the mutation disrupts the translation of Dp260, the ERG is abnormal.

One possibility for the variable ERG expression is random inactivation of the X chromosome. It is thought that the active X chromosome in manifesting DMD carrier females is the X chromosome bearing the dystrophin gene mutation.26 Manifesting carriers have raised CK levels and muscle weakness; however, none of the female subjects in this study manifests the disease. It is possible that the daughter may have a different X inactivation pattern in the retina that results in the more severe ERG phenotype. Our plan is to continue to follow this family and repeat the ERG in the daughter on a yearly basis to determine at which point (if any) her scotopic bwave amplitude reaches normal values.

In another study, scotopic ERG b wave amplitudes and b/a amplitude ratios were reported as normal in the mothers of 16 DMD boys.17However, the investigators did not do a complete ERG and only recorded the rod-cone mediated response under scotopic testing conditions. Also, the authors indicate that six of the 16 boys had a normal ERG. Therefore, one would not expect to see incomplete ERG abnormalities in the six mothers of the boys with normal ERGs. The results of this previous study are consistent with our adult scotopicb wave findings. We propose that the authors may have found an abnormal ON response b/aratio in the remaining 10 females whose sons had an abnormal ERG had they used a long duration photopic stimulus.

The DMD negative ERG, like the ERG inherited in the family reported in this paper, is the result of a mutation within the dystrophin gene and not the result of a defect in an adjacent gene.27 The ERG in DMD/BMD is a unique entity, easily distinguished from other negative ERG phenotypes and is in no way associated with night blindness.15 17 The question remains, though, as to which dystrophin isoform(s) in the retina play a key role in generating theb wave of the ERG. Mutations upstream of exon 30 are expected to have a normal ERG.13 This would seem to rule out a primary role for the Dp427 isoform as exemplified by the normal ERG in the mdx mouse (Dp427-, Dp260+, Dp71+). An abnormal ERG, described as a severe reduction in bwave amplitude, has been observed in themdx Cv3 mouse model.28 The mdx Cv3mutant mouse has a point mutation at the exon 65 splice junction29 of the dystrophin gene which would affect expression of Dp427, Dp260, and Dp71. Interpretation of the relationship between Dp260 and the abnormal ERG in themdx Cv3 mutant is difficult since the mutation affects all dystrophin isoforms in retina. The family studied in this report eliminates at least one dystrophin isoform from having a role in generation of the human ERGb wave. The affected DMD boys in this family exhibit a single exon deletion of exon 50. The promoter for Dp71 lies between exon 62 and 63. Since the promoter and coding sequence for Dp71 lies downstream of exon 50 in the dystrophin gene, the DMD boys in this family should produce normal Dp71. Therefore, the expression of Dp71 does not allow for a normal b wave in humans. This is in stark contrast to the findings of Kameyaet al,30 who reported a delayedb wave implicit time but normalb wave amplitude in a dystrophin exon 52 gene targeted mouse. This mouse expresses Dp71 in retina, but not full length dystrophin nor Dp260, a scenario we predict is identical in the family described here; however, the ERG phenotypes differ. The ERG of the dystrophin exon 52 gene targeted mouse also differs from themdx Cv3 mouse. An understanding of the physiological mechanisms that influence morphology of theb wave is important to explain these apparent discrepancies.

The b wave is the sum of the electrical activity of more than one cell type but is believed primarily to reflect the activation of depolarising bipolar cells. It is initiated by hyperpolarisation of the photoreceptors (awave). Since the ERG is the sum of events of opposite polarities (negative a wave, positiveb wave), any event that would alter the timing of depolarisation of the bipolar cell will alter the morphology of the b wave. Therefore, the ERG phenotype differences between this family (Dp427-, Dp260-, Dp71+), the dystrophin exon 52 gene targeted mouse (Dp427-, Dp260-, Dp71+), and themdx Cv3 mouse (Dp427-, Dp260-, Dp71-) may represent differences in b wave kinetics that would influence b wave morphology, amplitude, or implicit time. The cellular mechanism underlying the abnormality may be the same but results in a different ERG phenotype. Differences between species are a serious consideration. Besides the likely differences in retinal anatomy, physiology, and circuitry, the role of the unique human Dp260 isoform has yet to be explained. While it appears that, in humans, Dp260 plays the primary role in the production of the b wave, this conclusion is speculative until specific gene targeted animal models for the shorter dystrophin isoforms are produced to test the hypothesis.

Whether the ERG will prove to be a useful test for determining DMD carrier status in women remains unclear. Our data from this family would suggest that the ERG is capable of detecting the carrier status of DMD if the affected male family member has an abnormal ERG. One benefit of using the ERG in determining DMD carrier status is that it is non-invasive. Based on our preliminary results, it may be prudent to test suspected carriers at a young age, and, when testing adults, to do a complete ERG under both scotopic and photopic conditions with no less than 10 minutes of light adaptation. Use of a long duration flash is also highly recommended. A large study should be done to clarify the value of the ERG in detecting female heterozygous carriers of dystrophin gene mutations.


The authors are deeply indebted to the family described in this manuscript for their continued cooperation and interest. The authors acknowledge the expert technical assistance of Lori Moore, R EEG/EP T, of The Children’s Mercy Hospital, Vision Science Laboratory. This work was supported by a grant from The Katharine B Richardson Associates, Kansas City, Missouri (KMF).


View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.