Cancers display heterogeneity in genetic profiles of the individual cancer cells and in the composition of different malignant and non-malignant cell populations. Such intra-tumor heterogeneity plays a role in treatment response and the emergence of resistance to cancer therapies. Approaches that address this complexity and improve stratification of patients for treatment are therefore highly warranted. Thus, the aims of this thesis were to further develop and apply in situ technologies for expression and mutation analyses of candidate cancer genes to gain a deeper understanding of cancer biology and to study intra-tumor heterogeneity.
In paper I, we established and validated a procedure for scalable in situ hybridization of large gene sets in human formalin-fixed paraffin-embedded tissues for analysis of gene expression. This method was used in paper II for large-scale expression analysis of the tyrosine kinome and phosphatome, two gene families whose members are frequently mutated in many forms of cancers. Systematic, compartment-specific expression mapping at cell type resolution enabled us to identify several novel vascular markers that have gone unnoticed in bulk transcriptomic analyses. In papers III and IV, we used padlock probes for in situ mutation detection in single cells for studies of genetic intra-tumor heterogeneity. In paper III, multiplex detection and genotyping of oncogenic point mutations was demonstrated in routinely processed tissue materials, whereas in paper IV we further the application by demonstrating multiplex detection of fusion gene transcripts.
Collectively, the work presented in this thesis employs in situ-based methods to obtain spatial resolution of gene expression and mutation patterns in normal and cancer tissues, thereby broadening our understanding of the cancer genome.