Search

The process of malignant development involves the transformation from the normal cell-of-origin towards pre-invasive cells and eventually invasive cancer cells, which is facilitated by genomic alterations that are acquired during the accumulation of DNA replication errors. Faithful DNA replication is a key requirement for normal cell division which initiates from replication origins distributed throughout the genome, whereas unfaithful DNA replication results in replication errors and accelerates the malignant transformation. In this thesis, I endeavoured to decipher the phylogenies of malignant transformation between pre-invasive and invasive lung lesions, and further explored the role of perturbed DNA replication process from the cell-of-origin to cancer cell lines in remodelling the genomic landscape and contributing to malignant transformation. Pre-invasive lung diseases, including atypical adenomatous hyperplasia (AAH), adenocarcinoma in situ (AIS), minimally invasive carcinoma (MIA) and squamous CIS (carcinoma in situ), are considered to be precursors to lung adenocarcinoma (LUAD) and squamous cell carcinoma (LUSC), respectively. However, the evolutionary relationship between pre-invasive and invasive lung disease remains largely elusive, due to the lack of matched samples within the same patient. Using the TRACERx study, I identified two types of phylogenetic relationships between the pre-invasive and invasive lesions, according to whether the two shared a somatic common ancestor (the SCA group) or not (the non-SCA group). Compared to the non-SCA lesions, SCA pre-invasive lesions harboured elevated mutational burden, increased apolipoprotein B mRNA-editing enzyme catalytic polypeptide-like (APOBEC) mutational signature and exhibited chromosomal instability featured by increased loss of heterozygosity and somatic copy number alterations (SCNAs). Next, I explored how the genomic landscape was reshaped during the malignant transformation by investigating the perturbed DNA replication process. Even during the process of faithful DNA replication, not all replication origins initiate DNA replication at the same time, generating a conserved temporal order of DNA replication (i.e., the replication timing program) which is cell and tissue type specific. The extent to which replication timing alters during malignant transformation and the resultant impacts on the genomic landscape in cancer remain unclear. Thus, I investigated the landscape of altered replication timing (ART) in lung and breast cancer compared to their matched cell-of-origin and demonstrated the interplay between ART and genomic cancer evolution. A systematic shift in replication timing from the cell-of-origin to cancer cell lines was identified in 6%-17% of the genome, half of which were classified as earlier (late-to-early switch), and half later (early-to-late switch) replicated. An increase in mutation load was identified in later replicating regions in comparison to non-altered early replicating regions, whereas the opposite was true for earlier replicating regions showing decreased mutation load. I observed a more prevalent shift in the mutational landscape in breast compared to lung cancer, suggesting that most mutations in breast cancer were accumulated after the alteration in replication timing. Further mutational signature analysis revealed distinct mutational processes related to different DNA mismatch repair (MMR) pathways in lung cancer compared to breast cancer. DNA homologous recombination was prominent in earlier and early replicating regions in breast cancer, whilst in lung cancer, activated DNA MMR pathway were prominent in earlier and later replicating regions. A particular pattern of clustered APOBEC3 mutations (known as APOBEC3 omikli events) was found to be increasingly accumulated in early and earlier replicating regions compared to late/later ones. Finally, earlier (late-to-early switch) replicating genes exhibited an increase in expression in cancer relative to normal tissues, whereas later (early-to-late switch) replicating genes exhibited a decrease in expression. Cancer type-specific driver genes were enriched in earlier replicating regions compared with other non-altered late replicating regions. However, essential genes, which maintained high expression level, tended not to change their replication timing during malignant transformation. Together, the interplay between early/earlier replication timing and the APOBEC3 mutational process sheds lights on the positive selection of cancer-associated genes and thus the malignant transformation of lung and breast cancer. Accumulation of somatic mutations in late/later replicating regions due to their vulnerabilities to other mutagenic insults, contributes to the overall elevated mutation burden in cancer.