The telomere biology disorders (TBDs), or telomeropathies, embody a range of pathological phenotypes ensuing from abnormal telomerase function. The first TBD to be described was dyskeratosis congenita (DKC), a severe inherited multisystem disorder characterized by reticulate skin pigmentation, nail dystrophy, oral leukoplakia and bone marrow failure, presenting with cytopenia of one or more hematopoietic cell lineages. Clinically, telomere shortening due to the premature senescence of stem cells is most prominently displayed in highly proliferating mucocutaneous tissues. The most severe clinical feature of DKC is bone marrow failure, affecting the majority of patients and causing premature mortality. In addition, DKC patients have an increased risk for malignancies, fatal pulmonary complications, and immunodeficiency. At present, DKC is known to be caused by mutations in 11 genes, associated with X-linked recessive inheritance in DKC1, or with autosomal recessive and/or dominant inheritance in TERT, TERC, NHP2, NOP10, ACD, WRAP53, TINF2, RTEL1, CTC1, and PARN [2, 4–6]. Recently, the term DKC has been used to categorize only well-defined childhood symptoms. Based on the phenotype, a severe variant of DKC manifests as Hoyeraal-Hreidarsson syndrome (HHS). This is a rare disorder characterized by bone marrow failure, immunodeficiency, cerebral hypoplasia and intra-uterine growth retardation [3, 7]. DKC has often gone unnoticed due to delayed-onset of mucocutaneous findings. The wide spectrum of clinical presentations and the lack of a conclusive laboratory test for DKC can at times make the clinical diagnosis challenging. Exact genetic diagnosis in DKC is essential due to the limited efficacy of therapeutic options, and the genetic anticipation common in DKC makes timely family counseling a priority.
In the present study, we attested the validity of whole-exome sequencing (WES)-based approaches for routine genetic diagnostics of patients with heterogeneous clinical presentation or suspicion of TBD. In three patients and a brother of one of the patients presenting with varied characteristics of immunological or hematological disorders, we identified monogenic variants in genes associated with TBDs, and supported the genotype-phenotype correlation by demonstrating shortened blood cell telomere length.
This study was performed in Helsinki University Hospital, at the Pediatric and Adult Immunodeficiency Units, at the Department of Adult Hematology and at the Division of Pediatric Hematology-Oncology and Stem Cell Transplantation. Genetic analyses were performed as part of clinical workup in 212 clinically undiagnosed patients with disease phenotypes suggestive of monogenic conditions overlapping Primary immunodeficiency disorders (PIDs). The characterization of the phenotypes of the studied patients is shown in the Additional file 1: Figure S1. After the identification of alterations in TBD-associated genes in Finnish patients, we retrospectively collected their clinical data. This study was conducted in accordance with the principles of the Helsinki Declaration and was approved by the Coordinating Ethics Committee of Helsinki University Hospital. An informed consent was received from patients and in case of children, from their parents.
Genomic DNA of the studied individuals was isolated from peripheral blood through standard salt precipitation protocols. Whole-exome sequencing (WES) was performed in 212 patients. The detailed workflow for reads alignment and variant calling is provided in the Additional file 2.
Briefly, a SureSelect Clinical Research Capture Exome or SureSelect Human All Exon 50 Mb kits (Agilent, Santa Clara, CA, USA) were used. Paired-end sequencing was performed on the HiSeq 1500 or HiSeq 2000 platforms (Illumina, San Diego, CA, USA). The sequencing reads were analyzed using version 2.7 of the in-house developed analysis pipeline (VCP) for quality control and variant identification. Annovar (accessed in May 2017) was used for the annotations and prediction of functional consequences of the identified variants [10, 11]. The sequences were aligned with the GRCh39 reference build of the human genome using the BWA aligner. Downstream processing and variant calling were performed with the Genome Analysis Toolkit, SAMtools, and Picard. Substitution and InDel calls were made with GATK Unified Genotyper.
We considered only variants with a minor allele frequency (MAF) < 0.01, with frequency filtering based on data from Genome Aggregation Database (gnomAD, Cambridge, MA, USA; http://gnomad.broadinstitute.org/; and the SISu project (http://sisu.fimm.fi/) accessed in May 2017, respectively [15, 16]. We filtered according to the predicted consequences at the transcript level, selecting frameshift, in-frame, nonsense, splicing and missense variants. The variants were prioritized according to the predicted effect on the protein, to the conservation of the affected amino acids and in silico prediction tools (included in Annovar), and pathogenicity was predicted according to the American College of Medical Genetics (ACMG) Standards and Guidelines. Where possible, the candidate variants were analyzed in the families for co-segregation with affected relatives. Due to the several described inheritance models, we considered homozygous, compound heterozygous and heterozygous variants. All identified candidate variants were validated by Sanger sequencing (Additional file 2).
Telomere length analysis
Relative telomere length (RTL) was determined by the quantitative PCR method described by Cawthon et al., with minor modifications. Briefly, each sample (17.5 ng DNA from peripheral blood leukocytes) was analyzed in triplicate wells in separate telomere (TEL) and single copy gene (HBG) reactions on an ABI 7900HT instrument (Applied Biosystems), at two separate occasions in 96-well PCR plates. TEL/HBG (T/S) values were calculated by the 2 − ΔCt method, where ΔCt = CtTEL-CtHBG. The RTL value was generated by dividing samples T/S value with the T/S value of a reference cell-line DNA (CCRF-CEM) included in all runs. A standard curve of the reference cell line DNA was included in every run to monitor PCR efficiency. The RTL values were compared to 113 normal controls (age 0–83 years).
Due to the identification of variants in genes associated with defective telomere function, we analyzed the length of the telomeres in the four patients. All patients (P1.1, P1.2, P2, and P3) showed short relative telomere length below the 5th percentile compared to healthy controls of the same age (Fig. 2). The integration of genetic findings, clinical phenotype, and decreased telomere length led to a diagnosis of DKC in P1.1 and P1.2, bone marrow failure in P2, and Hoyeraal-Hreidarsson syndrome in P3.
In the present study, by using hypothesis-free WES we identified disease variants in telomeropathy-associated genes in four patients with clinical phenotypes ranging from mild signs of DKC to SCID.
Telomeropathies include a range of pathological conditions resulting from defective telomere biology, with great variation in terms of clinical severity and age at diagnosis. Not all TBD patients display the classical triad of symptoms (skin pigmentation, nail dystrophy and oral leukoplakia), thus lacking the pathognomonic signs of DKC. In some patients, bone marrow failure or immunodeficiency may be the only presenting symptom. The most severe phenotypical variant of DKC, Hoyeraal-Hreidarsson syndrome (HHS), may clinically present as SCID in children. However, bi-allelic mutations in the HHS-causing gene RTEL1 may also lead to selective NK deficiency. Wide phenotypic variation may contribute to the diagnostic delay, especially in adult patients, as seen in P1.1 and P1.2 who were not diagnosed until the age of 24 and 28 years. In addition to the diverse clinical presentation, mutations in various genes causing telomere-shortening may manifest with increasing severity and shorter telomeres in successive generations, a process called genetic anticipation. Thus, TBD patients with mild phenotype may have offspring potentially manifesting early in life with bone marrow failure or severe immunological aberrations with aggravated complications like rapid deterioration of pulmonary function. Due to the increasing severity in further generations, these disease-causing variants tend to eventually disappear from the pedigree. Likely for this reason, most of the Finnish families with telomeropathies included in this study appeared to have rare, novel mutations. This is fully opposite to the majority of the rare recessive diseases seen in Finland, which are typically caused by one major Finnish founder mutation enriched in the population [22, 23]. Therefore, the possibility of a novel variant needs to be considered for patients with TBD suspicion where no known pathological variants have been found in a targeted-variants approach.
Due to the heterogeneous clinical presentations and the allelic heterogeneity, patients with clinical suspicion of telomeropathies could benefit from the investigation by WES or other applicable NGS-based assays. This is successfully demonstrated in this study, where only one patient within our cohort (P1.1) presented with the classical skin, nail and mucosal findings suggestive of DKC. Furthermore, for one patient (P3) the genetic analysis immediately produced the final molecular diagnosis of HHS, while the initially observed clinical features (SCID) could have led to the use of repeated mistargeted assays not identifying the causal variants.
The identification of variants in TBD-associated genes in patients showing seemingly distinct immune or haematological conditions emphasizes the potential of WES for discerning atypical traits of monogenic syndromes and broadening the phenotypic spectrum of the disease in addition to providing a prompt molecular diagnosis. WES also allows screening for secondary causative variants potentially explaining the untypical phenotypic features. Importantly, WES combined with telomere length measurement further enables identification of novel TBD causing gene defects, given that novel TBD-causing genes are yet to be discovered. Because of the genetic heterogeneity, measurement of telomere length can also be considered for screening of undiagnosed patients with clinical features compatible with TBDs. The quantitative PCR-based assay used in this study is a cost-efficient mean to complement the diagnostics in patients with bone marrow failure or primary immunodeficiency.
From the genetic point of view, early diagnosis of TBD is critical for the counseling provided to the families. Once the disease-causing mutation(s) have been identified, assessment of allelic segregation should be systematically recommended to the entire family, particularly in case of X-linked TBD. As exemplified by DKC1, the disease may be carried silently in females and transmitted to next generations in absence of prenatal diagnosis. Also, some women may present with symptoms due to skewed X-inactivation, therefore further studies of the families are indicated. From the clinical point of view, the treating hematologists should consider the possibility of TBDs when evaluating patients with bone marrow failure, apparently unexplained cytopenias or myelodysplastic syndrome emerging at an early age, even in the absence of the classic mucocutaneous triad. Prompt diagnosis of a telomeropathy in case of atypically presenting TBD, like aplastic anemia, is important to bring to attention the need of monitoring other possibly arising TBD-related organ complications. In addition, immunosuppressive therapy is unlikely to be beneficial, but patients may profit from treatment with anabolic steroids aiming at improving cell counts. Furthermore, patients with severe bone marrow failure can be evaluated for hematopoietic stem cell transplantation. Knowledge of exact genetic diagnosis will help in planning the preparative regimen and follow-up after HSCT, as well as in the choosing of donors among family members.
Using hypothesis-free WES we identified novel disease variants in TBD-associated genes in clinically undiagnosed Finnish patients. The use of WES helped the clinicians to provide a prompt diagnosis of telomeropathy in the absence of the classic pathognomonic signs. We highlight the relevance of achieving a molecular diagnosis and subsequent accurate genetic counseling since effective treatments are lacking for TBDs.