Using a Cellecta-made shRNA library, groups from the European Institute of Oncology and MD Anderson Cancer Center published the first in vivo pooled RNAi screen in patient-derived tumor xenograft PDX mouse models. The study published in June Cancer Discovery screened 236 epigenetic genes targeted with a pooled shRNA library and identified 17 genes whose inhibition appeared to inhibit growth of either NRAS-mutant or BRAF-mutant grafted melanoma tumors.
Interestingly, although exome profiling showed that the grafted tumors maintained more than 97% of the SNP/indel mutations found in the original patient samples, none of the 17 genes that appeared to be required for tumorigenesis carried relevant cancer mutations or corresponded to known cancer drivers. The results lead the authors to suggest that somatically mutated genes may not necessarily be the most critical ones for tumor maintenance. Also, at least 10 of these implicated genes appeared to be druggable targets which could expand treatment options for melanoma.
Follow-up work led the authors to focus on 4 hits that looked particularly interesting and were confirmed using single shRNA constructs to be essential for tumorigenesis. Both BRAF and NRAS tumors required expression of three of these genes (BAZ1B, SMARCA4, CHD4), but KMT2D was only essential for NRAS tumor proliferation. The authors found that silencing these genes in cultured cells appeared to have little effect on cell proliferation but a strong inhibitory effect on cell migration, suggesting that inhibition of cell migration reduces tumor growth.
Most of the rest of the study focused on the NRAS-specific tumor promoter KMT2D. Although this gene is frequently mutated in cancers, it did not have any cancer-associated mutations in the melanoma samples used in this study. Also, in breast and colorectal cancers, KMT2D overexpression has been associated with poor prognosis.
The authors were able to directly link expression of KMT2D, which methylates histone H3, to the activation of genes involved in cell movement and migration. They also showed that silencing of KMT2D suppressed expression of these genes in NRAS melanoma cells, but not BRAF tumor cells. This suggests that BRAF-mutant tumor formation used an alternative pathway to upregulate cell migration.
It is well known that in vitro cell culture does not adequately model tumorigenesis, and further, that widely propagated cancer cell lines respond very differently than primary tumor cells from the same type of cancers. This in vivo RNAi work on patient-derived xenografts described in this very thorough study by collaborators from two of the world’s premiere cancer centers demonstrates both how important and how challenging it is to go beyond cell culture models and investigate cancer pathology in a more biologically relevant model.