Article Text

PDF

Communications
A new ocular phenotype associated with an unexpected but known systemic disorder and mutation: novel use of genomic diagnostics and exome sequencing
  1. Jacek Majewski1,
  2. Zibo Wang1,
  3. Irma Lopez2,
  4. Sulaiman Al Humaid2,
  5. Huanan Ren1,
  6. Julie Racine2,
  7. Alex Bazinet2,
  8. Grant Mitchel1,
  9. Nancy Braverman1,
  10. Robert K Koenekoop2
  1. 1Department of Human Genetics, McGill University Health Centre, Montreal, Quebec, Canada
  2. 2Department of Ophthalmology, McGill University Health Centre, Montreal, Quebec, Canada
  1. Correspondence to Dr Robert Koenekoop, Department of Ophthalmology, McGill University Health Centre, Montreal, QC H3H 1P3, Canada; robkoenekoop{at}hotmail.com

Statistics from Altmetric.com

Communication

We identified a 28-year-old cognitively normal patient with severe visual loss, absent electrical signals from the photoreceptors by electroretinogram (ERG) and nystagmus due to Leber congenital amaurosis (LCA), associated with hearing loss and Arnold–Chiari malformation. Exome sequencing detected a homozygous PEX1 mutation (p.Gly843Asp). Within a large LCA cohort, we found the mutation again in a 9-month-old baby. Peroxisome biochemical studies on both patients confirmed a peroxisome biogenesis disorder (PBD) in the Zellweger spectrum. We thus demonstrate that these patients, who had isolated LCA on presentation, actually had PBD as the cause of their LCA. Furthermore, the phenotype of the first patient was outside that of a typical Zellweger spectrum, and exome sequencing was instrumental in making this diagnosis possible.

Childhood blindness due to LCA (OMIM 204000) is a severe form of retinal photoreceptor cell degeneration and is defined by the clinical pentad of congenital visual loss, nystagmus (ocular oscillations), amaurotic pupils, retinal degeneration and absent electrical responses from both rods and cones on ERG. Mutations in 15 genes have been associated with autosomal recessive LCA, accounting for ∼70% of the patients. LCA genes are predominantly expressed in the retina, and LCA proteins participate in a variety of retinal pathways. LCA cohort studies have found that the majority of patients have an ocular phenotype only, which correlates with the predominant retinal expression of known LCA genes. However, we recently discovered three ubiquitously expressed ciliary genes that are mutated in LCA and, upon further analyses, are associated with systemic disease. We found that LCA patients with mutations in CEP290 or LCA5 develop olfactory defects. We then identified NPHP5 mutations causing LCA, prompted by protein homology and binding to CEP290, and found that they may result in kidney failure.1–4 These findings indicate the necessity for a paradigm shift in our understanding of the diagnostic boundaries of LCA. In the current study, we investigate this hypothesis.

To identify the causal gene of the first patient, we used a three-step process. First, we performed Arrayed Primer Extension screening (Asper-Ophthalmics, Tartu, Estonia) and eliminated 500 common mutations in 15 LCA genes. We then performed whole-genome single nucleotide polymorphism genotyping (Illumina Infinium HD 660K; Illumina, San Diego, California, USA) and identified homozygous regions, the largest of which spanned 18 Mb on chromosome 7, containing 109 genes, and none of which are known to cause LCA. Using exome capture (Agilent in-solution bead capture, all-exon v1 kits; Agilent, Santa Clara, California, USA), followed by next-generation sequencing using one lane of 76-base Illumina GAIIx reads, we obtained an average of 30× coverage of the targeted 180 000 exons. The data analysis was carried out as described earlier.5 We identified the causal mutation in PEX1, c.2528G→A (p.Gly843Asp), a known disease gene and common mutation.

We confirmed the p.Gly843Asp mutation by Sanger sequencing and found it homozygously in the proband and heterozygously in both parents and the unaffected sib, indicating true homozygosity in the patient. We then asked whether this mutation could be associated with LCA in our cohort of another 100 patients. Using restriction endonuclease enzyme analysis with CspCI, we found the mutation again, this time, heterozygously in another LCA patient, without systemic disease. Sanger sequencing identified the c.2098insT (p.Ile700Tyrfs42X) on the second allele, another well known but severe PEX1 mutation. We saw this baby at age 9 months, blind but healthy, and we diagnosed LCA on the basis of the diagnostic pentad. She was later hospitalised at age 20 months for developmental delay, hypotonia and seizures, and a diagnosis of PBD-Zellweger spectrum was made. In table 1, we show the confirmatory peroxisome metabolic testing performed in both patients.6 7

Table 1

Peroxisome functions in tissues from our patients compared to selected patients from the literature

The p.Gly843Asp PEX1 allele is the most common mutation in the PBD-Zellweger spectrum, a heterogeneous group of autosomal recessive diseases encompassing a continuum of severe (Zellweger syndrome) to less severe diseases (neonatal adrenoleukodystrophy and infantile Refsum disease). Zellweger syndrome usually leads to death before age 1 year, and patients have characteristic dysmorphology, neuronal migration defects, profound hypotonia and poor development. Patients with an infantile Refsum phenotype may survive through adolescence and manifest moderate-to-severe mental retardation, hypotonia, progressive retinopathy and sensorineural deafness and may not always have dysmorphic facies. The PEX1 p.Gly843Asp allele encodes a peroxin with residual protein function.8 9 The presence of at least one p.G843D allele predicts a phenotype milder than Zellweger syndrome.10 Homozygosity for p.Gly843Asp is generally associated with a mild infantile Refsum disease phenotype, which includes systemic disease with mental retardation, dysmorphic facies and other features. PEX1 p.Ile700Tyrfs42X, also a common allele, does not encode a functional protein and is a null allele.

The ocular phenotype of patient 1 consists of 20/400 acuities, nystagmus, amaurotic pupils, constricted visual fields with central scotomas, non-detectable ERGs, a retinal appearance with circular pigment clumps, foveal thinning and loss of foveal autofluorescence (figure 1). To our knowledge, these are the first in vivo detailed retinal architecture and retinal function studies in a blind patient with PBD. Our patient is a 28-year-old successful pharmacist who graduated from a university. In addition to her legal blindness, she has sensorineural hearing loss and an Arnold–Chiari I malformation with multicystic hydrosyringomyelia extending from C7 to the conus. Hearing loss is commonly identified in PBD spectrum of disease, but Arnold–Chiari malformation, certainly, is not. She also developed enamel disease requiring dental surgeries, which has been associated with PBD. The preservation of abnormal intellect remains a very rare occurrence.

Figure 1

(A) Retinal photo of the left eye showing a unique retinal phenotype consisting of marked pigmentary maculopathy, relatively intact perifoveal retina and retinal pigment epithelium (RPE), optic disc pallor with drusen and narrow vessels, RPE atrophy and marked unusual circular pigment clumps of various sizes. (B) Retinal photo of the left eye in the inferior periphery showing the details of the RPE atrophy, the preserved perifoveal area (between the maculopathy and the inferior vascular arcade) and the different-sized pigment clumps. (C) In vivo retinal architecture by optical coherence tomography (Spectralis; Heidelberg Engineering, Heidelberg, Germany), which shows an extremely thin fovea, devoid of cones, and in the periphery, there is retinal remodelling. In the perifoveal area, there remain remnants of outer-segment/inner-segment junctions. (D and E) Fundus autofluorescence (FAF; Heidelberg) of the right and left eyes showing the absence of FAF in the fovea, suggesting an absence of lipofuscin metabolism. Perifoveally, there is remaining FAF in a concentric circle around the fovea. This suggests that lipofuscin metabolism is present and that the RPE–photoreceptor interface is relatively intact.

With a molecular diagnosis of PBD-Zellweger spectrum, we performed an extensive peroxisome function testing in patient 1, considering her unusually mild phenotype (table 1). Her biochemical studies revealed elevations in peroxisomal fatty acid and pipecolic acid levels, normal red blood cell plasmalogens and normal urine bile acid pattern. Studies in fibroblasts were normal, with the exception of reduced phytanic acid oxidation and elevated soluble catalase. As shown in table 1, these values overlap (reduced phytanic acid and pristanic oxidation) with those of few previously reported mild PBD patients; exceptionally, her bile acid pattern was normal.

Our study shows the importance of genomic diagnostics using next-generation sequencing. Using traditional approaches, the first patient had not been suspected of having a peroxisome disorder and might never have been diagnosed. Having uncovered the genetic cause, we have opened the door for medical intervention, treatment and preventive steps in the future. Exome sequencing, combined with in vivo testing of retinal structure and function, is allowing the establishment of new genotype–phenotype correlations (figure 1). Our two patients had markedly different severities of PBD, but both presented initially with LCA. In the first case, the phenotype was dominated by the congenital blindness, and the hearing loss and Arnold–Chiari defects were secondary. She turned out to have one of the mildest forms of PBD reported to date. The second case also illustrates an important lesson, as her phenotype was only ocular (diagnosed as LCA) until age 18 months when she developed the systemic complications of severe PBD. We propose that all babies with LCA should be screened not only for the common PEX1 p.Gly843Asp mutation but also other common PEX1 mutations, to predict systemic disease.

The PEX1 p.Gly843Asp was reported once before in a patient presenting with “isolated LCA”, but this patient had dysmorphic facies, which does not occur in true LCA.10 Also, in re-studying the phenotype, this patient does not meet the strict criteria of LCA because she had normal fixation at birth and gradually lost vision at 6 months and developed acquired nystagmus. Patients with the true diagnosis of LCA have congenital blindness and congenital nystagmus. The dysmorphic patient10 also had papilledema (swelling of the optic nerve heads, signaling brain swelling), motor and speech delay, cerebellar tract signs and myoclonic seizures. These associated features exclude the diagnosis of LCA. Although it is possible that the specific mutation itself may be associated with an “isolated LCA phenotype”, it is also a common mutation, and we propose that mutations in other PEX genes associated with non-Zellweger syndrome phenotypes can also result in LCA before the recognisable onset of typical PBD symptoms. Our currently reported first patient had “true LCA” in that she had congenital blindness and congenital nystagmus, then developed Arnold Chiari malformation (not typical for PBD), hearing loss and enamel disease (typical for PBS). NGS allowed us to discover her mild biochemical abnormalities in keeping with mild PBD.

As a result of the tremendous genetic, allelic and phenotypic variability found to be associated with LCA, in addition to the common occurrence of photoreceptor death in a variety of systemic diseases, which leads to blindness and phenotypic mimicry of LCA, exome sequencing may, in the near future, become the first diagnostic approach for LCA, which will require addressing the ethical problems of the return of unexpected results, such as ours.

Acknowledgments

We thank the patients and their families for their participation. We thank Dr Steven Steinberg at the Kennedy Krieger Institute Peroxisome Disease Laboratory for his extensive biochemical evaluation of patient 1. The Montreal Children's Hospital and McGill University Health Centre scientific and ethic review committees approved this study. Financial support for this work comes from the Canadian Institutes for Health Research, Foundation Fighting Blindness-Canada, Fonds de la Recherche en Sante du Quebec, Reseau Vision, the National Institutes of Health and The Foundation for Retinal Research (to RKK). JM is a recipient of a Canada Research Chair, and this sequencing project was supported by a pilot grant from McGill University Health Centre. We would also like to thank the staff of the Genome Quebec sequencing platforms for expert help and personal involvement in the generation of data.

References

View Abstract

Footnotes

  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval The Research Institute of McGill University Health Centre.

  • Provenance and peer review Not commissioned; externally peer reviewed.

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.