Synonymous Mutation in DKC1 Causes Telomerase RNA Insufficiency Manifesting as Familial Pulmonary Fibrosis

Abstract

Background: Idiopathic pulmonary fibrosis (IPF) is the most common of short telomere phenotypes. Familial clustering of IPF is common, but the genetic basis remains unknown in more than one-half of cases. We identified a 65-year-old man with familial IPF, short telomere length, and low telomerase RNA levels. He was diagnosed with a short telomere syndrome after developing hematologic complications post-lung transplantation, but no mutations were identified in a clinical testing pipeline.

Research question: What is the molecular basis underlying the familial IPF and low telomerase RNA levels in this patient?

Study design and methods: We analyzed whole-genome sequence data and performed functional molecular studies on cells derived from the patient and his family.

Results: We identified a previously unreported synonymous variant c.942G>A p.K314K in DKC1, the gene encoding the dyskerin ribonucleoprotein, which is required for telomerase RNA biogenesis. The mutation created a competing de novo exonic splicing enhancer, and the misspliced product was degraded by nonsense-mediated decay causing an overall dyskerin deficiency in mutation carriers. In silico tools identified other rare silent DKC1 variants that warrant functional evaluation if found in patients with short telomere-mediated disease.

Interpretation: Our data point to silent mutation in telomere maintenance genes as a mechanism of familial pulmonary fibrosis. In contrast to DKC1 missense mutations, which primarily manifest in children as dyskeratosis congenita, hypomorphic mutations affecting dyskerin levels likely have a predilection to presenting in adults as pulmonary fibrosis.

Keywords: bone marrow failure; lung transplantation; telomerase.

Figure 1

Rare synonymous germline mutation in DKC1 segregates with idiopathic pulmonary fibrosis (IPF), short telomere length, and low telomerase RNA (TR) levels. A, IPF proband (arrow) and family pedigree, with numbers below indicating age at assessment or diagnosis. “d.” refers to age at death from IPF. The yellow boxes contain the genotype at position 153,999,060 on chromosome X in the GRCh37/hg19 genome assembly. Telogram shows age-adjusted lymphocyte telomere length by flow cytometry and fluorescence in situ hybridization (flowFISH) in the proband and family [arrow and pedigree identifiers as in (A)]. C, Relative TR levels measured by quantitative real-time PCR in lymphoblastoid cell lines [arrow and pedigree identifiers as in (A)]. Each point refers to a control- or patient-derived sample as labeled. Error bars represent SEM and asterisks (∗∗) refer to P < .01 (two-sided, Student t-test). D, Scheme of dyskerin protein with known domains including the nuclear localization signal. The putative position of the DKC1 variant, if it were successfully translated, is near the end of the protein. E, Pherogram shows skewed X-chromosome inactivation pattern in blood-derived genomic DNA of proband’s sister, using HUMARA. It shows the size of CAG repeats in the promoter of the androgen receptor (AR) gene. Allele size with and without digestion with methylation-sensitive restriction endonuclease HpaII is shown. Only one allele is detectable after digestion, consistent with 100% skewing. HUMARA = human androgen receptor gene; NL = nuclear localization; WT = wild type.

Figure 2

Synonymous DKC1 c.942G>A variant causes aberrant mRNA splicing and reduced dyskerin protein levels. A, Relative DKC1 mRNA levels measured by quantitative real-time polymerase chain reaction (qRT-PCR) in lymphoblastoid cell lines (LCLs), using primers in exons 10 and 11 downstream of the mutation. Error bars represent SEM and P values are two-sided, calculated by Student t-test. This experiment was independently replicated twice. B, DKC1 RT-PCR product resolved on a polyacrylamide gel after 0 or 4 h of cycloheximide treatment to inhibit nonsense-mediated decay in LCLs. The location of primers relative to the mutation (lollipop) is indicated in the embedded diagram. C, Sanger sequencing results of excised bands in (B) reveal an aberrant 40-base pair out-of-frame deletion in the proband. D, A diagram showing DKC1 c.942G>A impact on splicing. E, Immunoblot of dyskerin protein in LCLs from four healthy control subjects (C1 through C4) and a patient with a missense T49S DKC1 mutation with normal protein levels compared with the reduced dyskerin levels in the proband and the affected nephew (labeled with pedigree identifier from Fig 1A). Quantification of dyskerin levels relative to actin control is shown. Lane 8 contains protein extract from 293FT cells transfected with DKC1-MycFLAG tag to show specificity. F, Quantitation from three independently harvested protein lysates shows decreased dyskerin protein levels in male mutation carriers compared with control subjects and a T49S DKC1 mutant. Means were compared by Student t-test. Error bars represent SEM and P values are two-sided. cDNA = complementary DNA; CHX = cycloheximide.

Figure 3

Schematic map of disease-causing DKC1 splicing mutations and synonymous variants predicted to affect splicing. Scheme of genomic DKC1 locus showing the positions of potentially splice-altering synonymous variants identified from the Genome Aggregation Database (above the gene axis) at the time of query relative to the numbered exons. Variants shown fulfill the criteria of in silico prediction of > 25% from wild-type based on Human Splicing Finder (HSF) and are summarized from e-Table 1.