Abstract
Lymphangioleiomyomatosis (LAM) is a rare lung tumour that occurs predominantly in women during their childbearing years. The disease is characterised by the invasion of smooth muscle- like cells within the lungs, resulting in the formation of thin-walled cysts and presenting as respiratory insufficiency. While the majority of cases in the clinic are sporadic in nature (Sporadic-LAM; SLAM), the disease may also be seen in combination with the inheritable genetic disorder Tuberous Sclerosis Complex (TSC; TSC-LAM). Although the tissue of origin for LAM remains elusive, the developmental pathology has been delineated. Indeed, loss of function mutations in the TSC1 or more commonly TSC2 genes are causative for both SLAM and TSC. In TSC, the initial mutation occurs in the germline. As such, TSC presents as a multi- tumour syndrome with manifestations including LAM and other phenotypically similar tumours. However, in SLAM, these mutations are acquired somatically. Interestingly, while SLAM occurs exclusively in females, TSC-LAM is observed in some males, although its presentation is much less severe. The research described throughout this thesis uses altered patterns of DNA methylation to investigate novel therapeutic targets and the cell of origin for SLAM, a first for the field.
To begin, a robust patient tissue cohort with age and sex-matched controls was established. Subsequently, histopathology and immunohistochemistry techniques were used to quantify the extent of LAM lesions. Meanwhile, whole genome sequencing was employed to identify patients with germline mutations, indicative of TSC, and subsequently excluded from the SLAM cohort, laying a solid foundation for subsequent analyses. Next, genome wide patterns of DNA methylation and examined in SLAM, using high throughput microarray technologies. First, we reveal the effects of formalin fixation on DNA methylation by performing differential methylation analyses on patient matched fresh frozen and formalin fixed tissue samples. Critically, differential DNA methylation analyses of SLAM tissue with matched control lung uncovered distinct DNA methylation signatures, providing novel insights into mechanisms of the disease’s development. To investigate the cell of origin for SLAM, tissue ontology and advanced computational analyses were employed. Coupled with computational deconvolution methods, hierarchical clustering demonstrated deep clustering of a consensus ‘pure’ LAM cell with uterine and ovarian tissue samples, supporting a female reproductive tract origin for LAM.
ii
Finally, a single patient case study is described, exploring the complex relationship between LAM and other malignancies, with the hopes of paving the way for a personalised medicine approach.