Abstract
Shwachman-Diamond syndrome (SDS) is an inherited bone marrow failure syndrome (IBMFS) with gastrointestinal manifestations (pancreatic insufficiency) and cytopenias, primarily neutropenia. Skeletal dysplasias and short stature are frequent. Patients with SDS are at significantly increased risk of myelodysplastic syndrome and acute myeloid leukemia. More than 90% of patients have autosomal recessive inheritance of germline pathogenic variants in SBDS, a ribosome biogenesis gene. The most common variants are c.258+2T>C (disrupts the donor splice site of intron 2, leading to an 8bp deletion and a premature protein truncation due to a frameshift); and c.183_184TA>CT (introduces an in-frame stop codon). To date, the vast majority of pathogenic alleles reported have been single nucleotide variants (SNVs) or small insertions/deletions. Rare cases have been reported with SBDS exon 3 deletions.
The National Cancer Institute's IBMFS study is a longitudinal cohort study with 521 families enrolled, including 54 SDS or SDS-like families. Through clinical testing or whole exome sequencing all but nine families have had their disease-causing alleles identified. Three of the nine families had a known single pathogenic variant in SBDS. Array comparative genomic hybridization (aCGH) was uninformative in all but one.
We initially focused on the family with a known SBDS c.258+2T>C and potential deletion on aCGH. The 27 year-old (yo) male proband was diagnosed with SDS at 5yo due to a history of malabsorption, requiring pancreatic enzyme supplementation, failure to thrive (FTT), short stature and neutropenia. He also has metaphyseal dysplasia, scoliosis, psychomotor retardation and learning disability. A bone marrow (BM) biopsy at 21yo showed hypoplasia with mild dysplastic changes and a del20q clone. aCGH showed a possible large deletion, but the resolution was too low to determine exact breakpoints. Therefore, long-range Single Molecule, Real-Time (SMRT) Sequencing (PacBio Systems) was undertaken. This identified a read spanning a 19kb deletion with exact coordinates (hg19: chr7:66,436,397-66,455,294), and was confirmed with polymerase chain reaction and Sanger sequencing. This deletion removes part of intron 4, all of exon 5 and the 3'UTR of SBDS. Western blotting for SBDS showed significant decrease in protein production from cultured fibroblasts. This deletion in combination with the known SBDS variant and western blot is consistent with the SDS phenotype in this patient.
We then designed a targeted SMRT sequencing panel consisting of the four published SDS genes: SBDS, EFL1, DNAJC21 and SRP54 to evaluate the remaining families. The second proband evaluated was a 19yo female heterozygous for the SBDS c.258+2T>C variant. She had a history of malabsorption requiring enzyme supplementation and FTT in childhood. She has short stature, deformity of the distal metaphysis of the ulnae, and ongoing neutropenia with a hypocellular BM but no cytogenetic clones. Targeted SMRT sequencing identified a deletion in SBDS which deletes all of exon 3 and part of the surrounding introns (872bp, hg19: chr7: 66,457,992-66,458,863). Western blotting for SBDS showed a decrease in protein production in cultured fibroblasts consistent with a diagnosis of SDS. Sequencing of the family members indicated the deletion was paternal in origin, while the SNV was maternal.
A deletion of SRP54 was identified in two affected siblings from a third family; both siblings have a history of neutropenia and mild BM dysplasia. The deletion encompasses exon 8 of SRP54, which encodes part of the G-domain, critical for SRP54 cotranslational function, and is near other reported autosomal dominant pathogenic variants. One affected individual in this family also has a single SBDS c.183_184TA>CT variant, but had normal SBDS protein levels by western blot. In this family, the SBDS variant is likely not disease causing as individuals with heterozygous pathogenic variants of SBDS are generally asymptomatic, but rather the dominant SRP54 deletion is the disease causing genetic alteration in this SDS-like family.
No large deletions or insertions were identified in DNAJC21. Analysis of EFL1 is on-going. These cases illustrate the advantages of using long-read sequencing methodologies to identify large deletions which are not readily found using short-read sequencing, and identify large deletions as a novel mechanism of SDS inheritance.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.
This icon denotes a clinically relevant abstract