Abstract
Height is a well-established risk factor for cancer development, with each 10cm increase in height correlating to a ~10% increase in cancer risk. There does not appear to be a simple explanation for this association, reflecting the genetic complexities of both height and cancer. Mendelian (single gene) syndromes have tremendous power to potentially better understand such complex associations. The identified association between Laron dwarfism and decreased cancer risk prompted the hypothesis that there may be other Mendelian growth restriction syndromes where the underlying genetic cause acts in a protective manner against cancer. This thesis aimed to investigate a wide range of Mendelian short stature and microcephaly syndrome genetic variants for potential protective effects. In vitro and bioinformatic investigations were carried out to address this aim.
In vitro experiments aimed to investigate the effects of pathogenic variants for a specific primordial dwarfism (Meier-Gorlin syndrome; MGORS) on the rate of proliferation in cancer cells. Stable RKO cell lines expressing shRNA targeting three MGORS genes were established. Following transfection of variant-encoded MGORS gene vectors, cellular proliferation was tested in an assay measuring DNA content as a proxy for cell number. The results of this assay suggested that two of the three CDT1 pathogenic variants tested may impair cancer cell growth.
Bioinformatic investigations first tested various variant deleteriousness predictor tools, with BayesDel and CADD determined to be the most suitable for the genes under investigation. This project next utilised the UK Biobank (UKB), a large UK-based cohort of ~500,000 individuals. Within this cohort, carriers of pathogenic or deleterious genetic variants in 367 Mendelian growth restriction syndrome genes were identified with the aid of BayesDel and CADD variant effect predictor tools. Logistic and linear regression analyses were used to investigate associations with neoplasm incidence, neoplasm type, and age at first neoplasm diagnosis. While the results of these investigations did not support my ‘protective’ hypothesis, there were multiple tangential findings that are worth further investigation. Carriers of deleterious variants in BLM (RECQL3) and SHOX had an increased risk of neoplasm diagnosis, TSEN15 carriers were associated with a younger age at first neoplasm diagnosis, and OBSL1 carriers had a higher risk of endocrine neoplasms.
The outcomes of this research have the potential to significantly improve approaches to cancer and Mendelian syndromes. This research identified BLM is an ideal candidate for use in polygenic cancer risk prediction and cancer treatment development. Additionally, further investigation of the observed associations between cancer risk and CDT1, TSEN15, SHOX, and OBSL1 variants may improve our understanding, prevention, and treatment of cancers.