Patient cohort
The PTHS-like cohort included patients who were genetically counselled for PHTS and who tested negative for gPVs in the coding regions of PTEN. Patients were included if they were ≥18 years and were diagnosed with: i) ≥2 major PHTS features (breast-, endometrial-, thyroid cancer, macrocephaly, hamartomas, ganglioneuromas, trichilemmomas); ii) 1 major feature and ≥2 minor PHTS features (colorectal-, renal cancer, benign breast disease, benign thyroid disease, lipomas, vascular malformations); iii) 1 non-cancer major feature; or iv) any PHTS-associated cancer at age ≤35 years. All pathology-related features were extracted from the Dutch Nationwide Pathology Databank (PALGA).
Of 281 individuals of that were tested negative for PTEN, A total of 62 PHTS-like patients met the inclusion criteria and were successfully sequenced. From this cohort 37 individuals were diagnosed with breast cancer; six with endometrial cancer; sixteen with thyroid cancer; three with renal cancer and six with colorectal cancer. Twenty-three individuals have macrocephaly and eight individuals were diagnosed with Lhermitte Duclos (n=2), hamartomas (n = 2), ganglioneuromas (n = 2) or trichilemmomas (n = 2).
Targeted long-read sequencing of PTEN
In order to apply a T-LRS approach for PTEN, Long-range PCR amplicons sizing between ~8-16kb with a 1 kb overlap were designed to cover the full length of PTEN (108.31 kb). This targeted amplicon-based approach was chosen as it can also be applied on older DNA samples from PHTS-like patients. The design of the T-LRS approach for PTEN was successful, using a total of nine amplicons to capture the region of interest. The T-LRS was performed on a Pacific Biosciences Sequel II sequencer system. On average, each sample was sequenced at a median read depth of 254 reads [median range: 156-373 reads per sample]. After variant prioritization and in silico predictions for pathogenicity, two variants of interest were identified in two individuals in the 5’UTR and 3’UTR of PTEN.
T-LRS in PHTS-like patients allows for identification of variants in non-coding regions within PTEN. Functional analysis of these variants of interest will be carried out in order to determine their effect on expression of PTEN. Upon confirmation of the pathogenicity of these variants, this approach may lead to a genetic diagnosis for 3% of the tested PHTS-like patients.
Whole genome long-read sequencing
In addition to the T-LRS, eight individuals were selected for whole genome LRS. These individuals were also tested with T-LRS but remained negative for any putative causative variants. Whole genome LRS was performed on a Pacific Biosciences Revio sequencer system. So far, no variants have been identified that explain the phenotype of the patients.
Functional validation of variants in PTEN
In addition to the follow-up of variants that were identified through T-LRS, variants of uncertain significance (VUSs) in PTEN were also assessed for functional follow-up. These VUSs were previously identified in an ERN-GENTURIS PHTS cohort (n=510). All variants were classified as a VUS using the PTEN-specific ACMG guidelines. In order for these variants to be eligible for functional follow-up they needed to have a spliceAI score of >0.1 with no published functional data to date. Subsequently, these variants were then considered for a splice assay. Amplicons spanning one or two PTEN exons that harboured the identified VUSs were designed and cloned into a minigene vector using gateway cloning. Site-directed mutagenesis was used to generate the variants of interest and splice assays with mutant and wild type constructs were performed. Eight PTEN VUSs were identified for further functional testing. We identified five missense variants (c.77C>A, p.(Thr26Asn); c.89C>A, p.(Pro30Gln); c.94A>T, p.(Ile32Phe); c.801G>T, p.(Lys267Asn); c.929A>G, p.(Asp310Gly)) and three intronic variants (c.80-1G>C; c.165-18T>A; c.253+5G>C). These eight variants were distributed over six exon-intron regions, captured in five amplicons ranging between ~2.5-5.2 kb in size. The mini-gene splice assay has currently shown that variant c.253+5G>C results in vitro in altered splicing of PTEN.
