Partial Duplication and Poly(A) Insertion in KCNQ1 Not Detected by Next-Generation Sequencing in Jervell and Lange–Nielsen Syndrome
Jervell and Lange–Nielsen syndrome is caused by absence of the voltage-gated potassium current IKs through either homozygous recessive or compound heterozygous mutations in KCNQ1 or KCNE1. We report here a case of Jervell and Lange–Nielsen syndrome with typical clinical features in which clinical genetic testing using next-generation sequencing (NGS) revealed only a known single heterozygous KCNQ1 mutation (R518X), but failed to recognize an unusual and complex 52-bp duplication-insertion. This type of variant has not been previously reported in Long QT syndrome (LQTS) and may therefore account for a portion of genetic variant negative cases.
A 47-year-old woman with congenital deafness, severe QT prolongation (≥800 ms on Holter monitoring; Figure 1A and 1B), and striking exercise-induced QT lability and torsades de pointes was seen in referral. She had her first episode of syncope at 6 years of age, was reported to have marked QT prolongation, started β-blocker therapy, and was then lost to follow-up. At age 40, she had an episode of syncope while working outdoors in the heat, and β-blocker (propranolol at that time) was continued although she admitted to incomplete compliance because of fatigue. At age 46, she had another syncopal episode and was switched to metoprolol 100 mg daily. An implantable cardioverter-defibrillator was recommended but she declined. During an exercise test to assess the adequacy of β-blockade, she developed marked T wave lability followed by a 40-second, self-terminating episode of torsades de pointes, during which she lost consciousness (Figure 1C). Notably, this T wave instability occurred in an unusual (nonalternans) pattern,1 compatible with temporal variability in repolarization with sympathetic activation, as previously suggested.
Causative Mutation Identification
Commercial testing (GeneDx Inc, Gaithersburg, Maryland) using NGS identified a single heterozygous paternally inherited, previously described2 nonsense variant (R518X) in exon 12 of KCNQ1, and array testing revealed no large deletions or insertions in the gene (Methods in the Data Supplement). The clinical picture strongly suggested that a second KCNQ1 variant was present but undetected by NGS. To test this hypothesis, we performed PCR amplification of each exon for Sanger sequencing analysis (primer sets are presented in Table I in the Data Supplement). PCR amplification of exon 15 in the proband produced 2 products, one at the anticipated size of 500 bp and a second larger product of ≈550 bp (Figure 2A). The 500 bp product showed no variation compared with the reference sequence (NM_000218; Figure 2B), whereas the larger product contained a frame-shifting 52-bp insertion comprised of a partial duplication of the 5′ end of exon 15 and an insertion of 36 adenines followed by uninterrupted exon 15 sequence, consistent with a truncated Alu element (Figure 2C).3 The unique 52-bp insertion was a de novo event because targeted sequencing of exon 15 in parental DNA did not detect the insertion. The indel is predicted to change the coding sequence beginning at amino acid 583 with a truncation occurring after amino acid 667 (Figure I in the Data Supplement).
For the 2 identified variants to be the genetic cause of Jervell and Lange–Nielsen syndrome in this patient, they must be located in trans (on separate alleles) to one another. To determine whether this was the case, we used maternally inherited common single nucleotide variants (rs163149, rs163150, and rs1057128; Figures II and III and Table II in the Data Supplement) located upstream of exon 15 to demonstrate that the maternal common single nucleotide variants were on the same allele as the duplication/insertion (Figure IIIB in the Data Supplement; rs163149 shown as an example). Therefore, the proband inherited the variant encoding KCNQ1-R518X from her father, and the poly(A) insertion-partial exon 15 duplication was a de novo event that occurred on the maternally inherited allele (Figure 2D). The 2 variants were thus present in trans, establishing KCNQ1 compound heterozygosity as the pathogenesis for Jervell and Lange-Nielsen syndrome in this case.
Reevaluation of NGS Data
NGS uses massively parallel sequencing to generate short (50–100 bp in this case) reads that are then mapped to a reference sequence. Examination of the original NGS analysis showed a small number of reads that contained the poly-adenine sequence at the 5′ or 3′ edge of the short reads mapped to the reference KCNQ1 sequence. There was a drop-off in read depth at the insertion site, supporting the idea that the algorithm analyzing the NGS discarded reads containing the insertion near their center because they could not be aligned to the reference sequence. When the 52-bp duplication–insertion was included in a modified reference sequence, the insertion-containing variant reads mapped with high confidence, whereas reads from the paternal allele failed to map, leading to a drop-off in read depth (Figure 3).
We identified a novel LQTS variant that is likely composed of a severely 5′-truncated Alu element RNA intermediate insertion (hence the poly(A) tract) and a DNA target site duplication causing the partial exon 15 duplication.3 This unusual type of de novo insertion has not been previously reported in LQTS, either because it is rare or because it is not recognized by conventional testing. NGS has evolved as a low cost, efficient method to generate data on a large number of genes in parallel to screen a targeted panel of arrhythmia susceptibility genes in patients diagnosed with LQTS. Although current NGS algorithms detect single nucleotide variants with high confidence, insertion/deletion variants (indels) are not as sensitively detected. Genetic testing in cases with clear LQTS identifies a likely causative variant in 80% of cases.4 It is possible that a portion of the remaining 20% of cases without known genetic causes could result from intermediate-sized indels like the one described here. Alternatively, undiscovered LQTS-associated genes or deep intronic or regulatory variants that disrupt gene expression or transcript processing may also contribute.
The Vanderbilt VANTAGE Core provided technical assistance for this work.
Sources of Funding
This work was supported by National Institutes of Health grants U19 HL65962 and P50 GM115305; K. Bersell is supported by NIGMS T32 GM07347 through the Vanderbilt Medical-Scientist Training Program and NHLBI F30HL127962. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. VANTAGE is supported in part by CTSA Grant 5UL1 RR024975-03, the Vanderbilt Ingram Cancer Center P30 CA68485, the Vanderbilt Vision Center P30 EY08126, and NIH/NCRR G20 RR030956.
Dr Chung is a consultant for BioReference Laboratories. D. Macaya, Dr Konecki, and E. Venter are employed by GeneDx. The other authors report no conflicts.
The Data Supplement is available at http://circep.ahajournals.org/lookup/suppl/doi:10.1161/CIRCEP.116.004081/-/DC1.
- Received December 14, 2015.
- Accepted May 3, 2016.
- © 2016 American Heart Association, Inc.
- Wei J,
- Fish FA,
- Myerburg RJ,
- Roden DM,
- George AL Jr.