Article Text

Download PDFPDF

Short report
Rapid detection of a mutation causing X-linked leucoencephalopathy by exome sequencing
  1. Yoshinori Tsurusaki1,
  2. Hitoshi Osaka2,
  3. Haruka Hamanoue1,
  4. Hiroko Shimbo2,
  5. Megumi Tsuji2,
  6. Hiroshi Doi1,
  7. Hirotomo Saitsu1,
  8. Naomichi Matsumoto1,
  9. Noriko Miyake1
  1. 1Department of Human Genetics, Yokohama City University Graduate School of Medicine, Yokohama, Japan
  2. 2Division of Neurology, Clinical Research Institute, Kanagawa Children's Medical Center, Yokohama, Japan
  1. Correspondence to Dr Noriko Miyake, Department of Human Genetics, Yokohama City University Graduate School of Medicine, 3-9 Fukuura, Kanazawa-ku, Yokohama 236-0004, Japan; nmiyake{at}yokohama-cu.ac.jp

Abstract

Background Conventional PCR-based direct sequencing of candidate genes for a family with X-linked leucoencephalopathy with unknown aetiology failed to identify any causative mutations.

Objective To carry out exome sequencing of entire transcripts of the whole X chromosome to investigate a family with X linked leucoencephalopathy.

Methods and results Next-generation sequencing of all the transcripts of the X chromosome, after liquid-based genome partitioning, was performed on one of the two affected male subjects (the proband) and an unaffected male subject (his brother). A nonsense mutation in MCT8 (c.1102A→T (p.R368X)) was identified in the proband. Subsequent PCR-based direct sequencing of other family members confirmed the presence of this mutation, hemizygous in the other affected brother and heterozygous in the proband's mother and maternal grandmother. MCT8 mutations usually cause abnormal thyroid function in addition to neurological abnormalities, but this proband had normal thyroid function.

Conclusion Single-lane exome next-generation sequencing is sufficient to fully analyse all the transcripts of the X chromosome. This method is particularly suitable for mutation screening of X-linked recessive disorders and can avoid biases in candidate gene choice.

  • Next-generation sequencing (NGS)
  • X linked leucoencephalopathy
  • exome sequencing
  • MCT8
  • genetics
  • neurology

Statistics from Altmetric.com

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.

Introduction

High-throughput, next-generation sequencing (NGS) can have a tremendous impact on human genetic research.1 Even personal whole-genome analysis is possible,2 but the cost of obtaining and analysing an entire genome from many people is still unrealistic for many laboratories. Selection and enrichment of regions of interest (genome partitioning) enable us to use NGS efficiently for reasonable numbers of patients with genetic disorders.3–6

Ready-to-use microarray-based and solution-based hybridisation systems are now commercially available. A combination of genome partitioning using these systems and NGS is one of the most promising ways to identify genes causing Mendelian disorders.3 4 6

Here, we performed exome sequencing of entire transcripts of the whole X chromosome to investigate a family with X linked leucoencephalopathy with unknown aetiology after intensive candidate gene analysis by conventional exon-by-exon Sanger sequencing. A single-lane run of NGS on only two family members successfully determined the leucoencephalopathy-causing mutation.

Subjects and methods

A family with X-linked leucoencephalopathy

The proband (III-2) was a 13-year-old boy. He was born to Japanese consanguineous parents (II-3, 4) after an uneventful pregnancy (figure 1A). His birth weight was 3440 g. Congenital horizontal nystagmus was noted as a neonate. Because of his poor weight gain and developmental delay, he was referred to us at age 5 months. He showed progressive spasticity and dystonia with exaggerated deep tendon reflexes as well as myoclonic and tonic seizures, which responded to valproic acid and clonazepam at age 21 months. Brain MRI at 2 years showed diffuse hyperintensity of the frontal lobe on T2-weighted images, suggesting hypomyelination, and normal T1-weighted images (figure 1B,C). The peak latency intervals in auditory brainstem responses (I–V/III–V) were 4.63/2.37 ms, which were elongated compared with those of age-matched controls (4.24±0.08/1.97±0.08 ms (mean±SD)). He was clinically diagnosed with Pelizaeus–Merzbacher disease (MIM#312080), although neither mutation nor duplication was found in PLP1 (RefSeq Gene ID, NM_000533) or GJA12 (NM_020435) (the duplication in GJA12 was not checked). He was never able to follow objects or control his head.

Figure 1

Pedigree and brain MRI of the proband. (A) Family pedigree. (B) T2-weighted image at age 2 years shows diffuse hyperintensity, especially in the frontal lobe. (C) T1-weighted image at 2 years shows nearly complete myelination. (D and E) At age 13 years, both T2 (D) and T1 (E)-weighted images demonstrate complete myelination; the hypomyelination observed at age 2 years can therefore be regarded as delayed myelination. (F) Flow of informatics analysis. A MAQ-based method and NextGENe analysis were performed (III-2). The selection methods included variation relative to the human genome reference sequence; variants mapped to the X chromosome; unknown variants (excluding registered SNPs); variants identified in the proband only (not in his healthy brother); variants in known genes; coding region variants; variants in genes at Xq13.1–q25; and variants common to the two informatics methods. MAQ, Mapping and Assembly with Qualities; SNP, single nucleotide polymorphism.

The dystonia worsened and he is now mechanically ventilated because of tracheomalacia. A thyroid function test at age 13 years indicated all normal levels: free tri-iodothyronine (T3) 1.2 ng/ml (normal range 0.8–1.6 ng/ml), free thyroxine (T4) 6.4 μg/dl (normal range 6.1–12.4 μg/dl) and thyroid-stimulating hormone 1.2 μIU/ml (normal range 0.5–5 μIU/ml). Brain MRI at age 13 years demonstrated improvement of myelination in the white matter, but he still presented with severe mental retardation (figure 1D,E). His younger brother was an 8-year-old boy (III-3) with an almost identical clinical course and MRI findings. His grandparents (I-1, I-2) were both healthy. The elder uncle (II-1) died at age 27 years who, initially, could walk with support but who declined towards the end of his life. Another uncle (II-2) was diagnosed with cerebral palsy and died at 7 months of age of unknown causes.

Informed consent was obtained from the patient's family members in accordance with human study protocols approved by the institutional review board at Kanagawa Children's Medical Centre and Yokohama City University School of Medicine.

Genome-wide single nucleotide polymorphism (SNP) genotyping

Genome-wide SNP genotyping was undertaken for individuals I-1, I-2, II-3, II-4, III-1, III-2 and III-3 using the GeneChip Human Mapping 10K Array Xba 142 2.0 (Affymetrix Inc, Santa Clara, California, USA), according to the manufacturer's protocols. Mendelian errors in the pedigree to exclude conflicted SNPs were checked using GeneChip operating software 1.2 (Affymetrix) and batch analysis in GeneChip genotyping analysis software 4.0 (Affymetrix), with the default settings for a mapping algorithm. Copy Number Analyzer for GeneChip 2.0 was used to validate copy number alterations.7 The linked region with SNPs shared between individuals III-2 and III-3 (not observed in III-1) was checked manually.

Genome partitioning, short-read sequencing and sequence alignment

Genomic DNAs from the proband (III-2) and his unaffected brother (III-1) were used for this study. Three micrograms of DNA were processed using a SureSelect X chromosome test kit (1582 transcripts covering 3053 kb) (Agilent Technologies, Santa Clara, California, USA), according to the manufacturer's instructions. Captured DNAs were analysed using an Illumina GAIIx (Illumina Inc, San Diego, California, USA). We used only one of the eight lanes of the flow cell (Illumina), performing single 76 bp reads for each sample. Image analysis and base calling were performed by sequence control software (SCS) real-time analysis (Illumina) and/or offline Basecaller software v1.6 (Illumina) and CASAVA software v1.6 (Illumina). Reads were aligned to the human reference genome sequence (UCSC hg18, NCBI build 36.1) using the ELAND v2 program (Illumina). Coverage was calculated statistically. Identified variants were annotated based on novelty, impact on the encoded protein, the number and frequency of reads and conservation. NextGENe software v1.99 (SoftGenetics, State College, Pennsylvania, USA) was also used to analyse reads, with the default settings.

Mapping strategy and variant annotation

Approximately 9.9 million reads from III-1 (the unaffected sibling) and 7.8 million reads from III-2 (the proband), which passed the quality control (Path Filter), were mapped to the human reference genome by Mapping and Assembly with Qualities (MAQ)8 and NextGENe software (SoftGenetics) (figure 1F). The bait region of the X chromosome based on the manufacturer's information was carefully evaluated. MAQ was able to align 7 359 688 and 6 614 972 reads to the whole genome for III-1 and III-2, respectively, which were statistically analysed for coverage using a script created by BITS Co Ltd (Tokyo, Japan). SNPs and indels were extracted from the alignment data using another script created by BITS, along with information on registered SNPs (dbSNP build 130). A consensus quality score of ≥40 was used for the SNP analysis in MAQ.

Capillary sequencing

Possible pathological variants were confirmed by Sanger sequencing using an ABI 3500xl or ABI3100 autosequencer (Life Technologies, Carlsbad, California, USA), following the manufacturer's protocol. Sequencing data were analysed by Sequencher software (Gene Codes Corporation, Ann Arbor, Michigan, USA).

Results and discussion

Coverage analysis showed that 78.9% of all the X chromosome transcripts were completely covered by reads, and that 11.6% of transcripts were at least 90% covered. Almost all (99%) of these regions were covered by 20 reads or more (100 reads or more in 97%) by only single-lane sequencing. SNP genotyping was able to delineate the minimal linked region from rs763739 to rs1073455 (UCSC genome browser hg19 assembly, X chromosome coordinates: 76 804 990–126 844 262) (50 Mb). The maximum linked region was from rs1926354 to rs859587 (UCSC genome browser coordinates: 68 404 915–128 933 907) (60.5 Mb). Exome GAIIx sequencing with the two informatics methods identified four potentially interesting changes in the maximum linked region: c.1102AT (p.R368X) in MCT8 (NM_006517; alternatively called SLC16A2); c.1402T→G (p.S468A) and c.1943A→G (p.H648R) in CYLC1 (NM_021118); and c1606G→A (p.D536N) in LRCH2 (NM_020871) (figure 1F). c.1102A→T (p.R368X) in MCT8 was found heterozygously in the proband's healthy mother (II-3) and maternal grandmother (I-2), and hemizygously in the proband and his affected younger brother; each was confirmed by Sanger sequencing (figure 2). This change was not present among 92 normal female controls (0/184 alleles).

Figure 2

Electropherograms of a normal control, a carrier (mother) and the affected proband.

The MCT8 gene encodes a thyroid hormone transporter and is implicated in syndromic X-linked mental retardation, Allan–Herndon–Dudley syndrome and Pelizaeus–Merzbacher-like disease (PMLD).9–12 This nonsense mutation, c.1102A→T (p.R368X), which might lead to nonsense-mediated decay resulting in no protein production, is highly likely to be pathological. Based on the human gene mutation database (http://www.hgmd.cf.ac.uk/ac/index.php), three nonsense mutations in this gene have been previously registered: p.R245X, p.Q335X and p.S448X. The other identified variants, in CYLC1 and LRCH2, are all SNPs because they were identified in normal controls: c.1402T→G (CYLC1): 5/182 alleles, c.1943A→G (CYLC1): 12/184 alleles and c1606G→A (LRCH2): 5/184 alleles. We concluded that the MCT8 mutation was pathogenic in this family.

PMLD caused by MCT8 mutations presents with infantile hypotonia, severe psychomotor development, nystagmus, generalised muscle weakness, dystopia, joint contracture and progressive spastic paraplegia. All affected male subjects develop the disease, while heterozygous female subjects are clinically normal or sometimes show mild thyroid dysfunction.9 12 Brain MRI shows delayed myelination in the first few years of life, which subsequently improves but with residual neurological disability. The unique diagnostic feature of the disease is an abnormal thyroid hormone profile: increased free T3, decreased free T4 and normal thyroid-stimulating hormone.12 The cases we analysed here showed clinical features and brain MRI findings typical of PMLD, but no thyroid hormone abnormalities. Based on regular laboratory testing and conventional PCR-based gene screening, we might have failed, or taken much longer, to identify the causative mutation. Thus, unbiased screening without prior knowledge is one of the advantages of NGS.

Thyroid hormone (T4 and T3) is important in neuronal development and its deficiency in the pre/neonatal stage causes a form of mental retardation called cretinism. T4 is released from the thyroid as a prohormone and is altered to biologically active T3 by iodothyronine deiodinases.13 Active T3 is delivered to the peripheral organs via thyroid hormone transporters. MCT8 is a thyroid hormone-specific transporter and is mainly expressed in the brain and liver.14 15 In MCT8 deficiency, T3 and T4 uptake is impaired and deiodinase 2 is activated.16 This results in increased serum T3 levels because of T3 accumulation in the peripheral blood. In previous reports, the majority of patients showed abnormal levels of thyroid hormones, but some displayed values within the normal range.9 10 12 17 18 The variable range for abnormal thyroid hormone levels might be explained by unidentified modifier effect(s) and/or other transporter(s) that can compensate for MCT8 function.19 Additionally, although MCT8 deficiency has been determined by abnormalities in thyroid function tests, it is unknown what proportion of the patients with MCT8 deficiency show abnormal thyroid function. We suggest that it is important to evaluate thyroid hormone function in PMLD with unknown cause.

Before the exome NGS analysis, we screened PLP1, GJA12, and seven other candidate genes mapped to the linked region: MSN (NM_002444), IGBP1 (NM_001551), SNX12 (NM_013346), OGT (NM_181672), HDAC8 (NM_018486), SH3BGRL (NM_003022.2) and PCDH11X (NM_032967.2). Because we found no pathological changes, we adopted the exome sequencing strategy. We determined that exome sequencing with a single lane for each sample was sufficient to analyse all the transcripts of the X chromosome. In X-linked recessive diseases, male subjects are usually affected, and therefore the single X chromosome is the primary target of exome sequencing. Except for mosaic mutations, the hemizygous (rather than heterozygous) status of disease-related nucleotide changes is relatively easy to detect using all-or-none NGS reads (0% or 100% of reads). There was no difference in the ability of our two informatics methods (MAQ and NextGENe) to detect pathological changes. This approach could equally be applied to the analysis of autosomal recessive diseases that manifest in the offspring of consanguineous relationships.

In conclusion, we rapidly identified a nonsense mutation in MCT8 in a family with X-linked leucoencephalopathy using only a single lane of exome sequencing. This method is powerful for unbiased screening of disease-related mutations in X-linked or recessive conditions.

Acknowledgments

We thank the family for their participation in this study.

References

Footnotes

  • Funding This work was supported by research grants from the Ministry of Health, Labour and Welfare (to HO, HSa, NMa and NMi), a grant-in-aid for scientific research from the Japan Society for the Promotion of Science (NMa), a grant-in-aid for young scientists from the Japan Society for the Promotion of Science (HSa), a grant from the 2010 Strategic Research Promotion of Yokohama City University (NMa), research grants from the Japan Epilepsy Research Foundation (HSa) and a research grant from the Naito Foundation (NMa). The study sponsors had no role in the study design; in the collection, analysis, and interpretation of the data; in the writing of the report; or in the decision to submit the paper for publication.

  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval This study was conducted with the approval of the institutional review board of Kanagawa Children's Medical Center and Yokohama City University School of Medicine.

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