Tissue-specificity of cancer: linking the rewiring of molecular networks in different tissues and the mutational profiles of different cancer types
Final Report Abstract
With the availability molecular alteration profiles from 10,000s of cancer patients, we begin to realize the large heterogeneity of these profiles over individuals and cancer types. For example, tumor suppressors like BRCA1 or oncogenes like KRAS show a highly tissue-specific tendency to be mutated: in certain cancer types a large fraction of patients carry a mutation in the respective gene (e.g. BRCA1 in breast cancer and KRAS in pancreatic cancer) while in other cancer types these genes are never found mutated (both are rarely mutated in leukemia). Here, I am describing the results of investigating the determinants of the tissue-specificity of molecular alterations in cancer. We considered different types of data to address this question: most importantly expression of the affected genes, functional information and molecular network data. We confirmed that expression in the tissue of origin only poorly correlates with how often a gene is found mutated in a cancer type originating from this tissue. However, we found that genes that are only found mutated in a small number of tissues tend to be involved in molecular functions that deal with environmental perturbations (such as the immune system or DNA damage repair). Based on this observation and by analyzing host-virus and gene-chemical networks, we established that these specific cancer genes tend to interact with environmental chemicals or viral proteins specifically absorbed by the respective tissues. Like point mutations also copy number alterations (CNAs) and DNA methylation (DNAm) show a highly cancer type-specific alteration profile. We found that in patients with a highly elevated or depleted number of CNAs a network of epigenetic modifiers is frequently mutated. Therefore, we investigated how the epigenome of the tissue of origin might influence where CNAs occur. We found that the presence of heterochromatin in the tissue of origin contributes to the recurrence and length of CNAs in the respective cancer type. Similarly, we find that the cancer type-specific amount of differential DNAm is influenced by the specific structural organization of the DNA in the tissue from which the cancer originated. Finally, we investigated proteins and peptides under negative selection in cancer (i.e. that are mutated less frequently then one would expect by chance). Therefore, we analyzed mutation data from thousands of patients and identified a substantial number of genes under negative selection. Many of these genes are related to cancer hallmark functions such as increased proliferation and reprogramming of glucose metabolism. Also, negative selection is tissue-specific: we observed strong signals of negative selection in peptides exposed by the MHC complex to the immune system (epitopes). However, the degree of selection against mutations in epitopes depends on the tissue of origin. We found that cancer types with higher immune activity show stronger negative selection acting against mutations in epitopes. In summary, we identified environmental factors and the chromatin organization of the tissue of origin as important determinants of tissue-specific alteration profiles. While the environment affects selection (e.g. immune system proteins are under positive and immune exposed epitopes under negative selection in a tissue-specific manner), the chromatin organization affects the local rates with which alterations (such as CNAs or DNAm) occur.
Publications
- "A network of epigenetic modifiers and DNA repair genes controls tissue-specific copy number alteration preference." eLife 5 (2016): e16519
Dina Cramer, Luis Serrano, and Martin H. Schaefer
(See online at https://doi.org/10.7554/elife.16519) - "Cell type-specific properties and environment shape tissue specificity of cancer genes." Scientific reports 6 (2016)
Martin H. Schaefer, and Luis Serrano
(See online at https://doi.org/10.1038/srep20707) - "HIPPIE v2. 0: enhancing meaningfulness and reliability of protein–protein interaction networks." Nucleic acids research 45.D1 (2017): D408-D414
Gregorio Alanis-Lobato, Miguel A. Andrade-Navarro, and Martin H. Schaefer
(See online at https://doi.org/10.1093/nar/gkw985) - "Tissue-specific DNA methylation loss during ageing and carcinogenesis is linked to chromosome structure, replication timing and cell division rates." Nucleic acids research (2018)
Marija Dmitrijeva, Stephan Ossowski, Luis Serrano, Martin H. Schaefer
(See online at https://doi.org/10.1093/nar/gky498)
