by ecancer reporter Clare Samson

The outcome for many cancer patients has substantially improved over the last two decades with the introduction of drugs that target kinases and other proteins in signalling pathways that are disrupted in tumour cells.

However, in almost all cases the tumours eventually develop resistance to these drugs and the cancer returns.

Understanding the processes through which resistance develops will be essential if it is to be overcome and complete cures achieved.

Malignant melanoma, the most aggressive type of skin cancer, is a case in point: about 60% of melanomas carry a mutation in the kinase BRAF and are sensitive to the BRAF inhibitor vemurafenib until resistance arises.

This resistance is known to develop in a small subset of tumour cells, which continue to replicate after the majority of drug-sensitive cells have died.

A group of researchers led by Arjun Raj of the University of Pennsylvania, Philadelphia, USA has investigated the ways through which melanoma develops resistance to vemurafenib at the single-cell level.

They cultured melanoma cells derived from two human patients, exposed both cultures to small doses of vemurafenib and showed that a small fraction of the cells in each culture developed drug resistance.

These resistant cells had proliferated normally before drug addition, proving that they were not in a dormant, ‘persister’ state.

Raj and his co-workers considered two distinct models for the development of resistance in single cells: a heritable model in which resistance results from further mutation and resistant cells cannot revert to a sensitive state, and a non-heritable model in which cells can transition between non-resistant, pre-resistant and resistant states.

They isolated single tumour cells, expanded them into cultures, added vemurafenib and examined the cultures for populations of resistant cells; the results suggested that there was no single, heritable pre-resistance phenotype, and no new mutations were discovered in resistant sub-clones.

The researchers then investigated differences in gene expression between non-resistant and pre-resistant cells.

Most untreated cells expressed several known genetic markers of drug resistance – including WNT5A, AXL, EGFR, PDGFRB, and JUN – at low levels, but a few showed much higher expression levels of these genes.

They stained live melanoma cells with an antibody to EFGR, isolated the cell population with highest EFGR expression levels and exposed the isolated cells to vemurafenib, showing that these cells produced more drug-resistant colonies than the unsorted cell population.

The resistance of these EFGR-high cell cultures dropped after the drug was withdrawn, further indicating that the induced resistance was transient.

The researchers then investigated expression levels of several other resistance-related genes in further cancer cell lines and non-cancerous primary melanocytes, and found that the phenomenon of sporadic gene expression was widespread.

Expression levels of some resistance marker genes were found to correlate, suggesting to Raj and his co-workers that rare cells expressing multiple resistance markers might be even more likely to develop resistance.

To test this, they applied vemurafenib to separate populations of melanoma cells expressing EGFR, NGFR, both receptors and neither, and found as they expected that resistance developed most strongly in the double-positive population.

Cells that expressed high levels of multiple marker proteins developed resistance even more readily.

Taken together, these results suggest that tumour cells can enter a transient state characterised by expression of resistance marker proteins and an ability to readily acquire stable resistance when exposed to a drug.

The expression patterns of drug-exposed, stably resistant cells differ from those with the transient ‘pre-resistant’ phenotype, showing even higher levels of many of the resistance-related proteins.

Finally, Raj and his co-workers explored the mechanisms through which the drug induces a re-programming of tumour cells from the pre-resistant to the stably resistant state by using chromatin to identify which transcription factor binding sites were occupied.

They discovered that a significant change in the occupancy of many such sites occurred after the drug was added, indicating that the cells were re-programmed to express a different range of proteins.

The first change following drug exposure was a loss in binding of the transcription factor SOX-1; later, the transcription factors JUN, AP-1 and TEAD bound to other sites, activating new signalling pathways.

Re-programming is not complete until about 4 weeks after the drug was first added, suggesting that one clinical application of these results might be to inform dosing intervals so pre-resistant cells have enough time to revert to the sensitive state between doses.

Reference
Shaffer, S.M., Dunagin, M.C., Torborg, S.R. and 14 others (2017). Rare cell variability and drug-induced reprogramming as a mode of cancer drug resistance. Nature, published online ahead of print 7 June 2017. doi:10.1038/nature22794