One of these mutants is not like the other

Common sense would suggest that mutations in the same region of the same protein in two different cancers would have at least a similar effect, even if the mutations themselves are slightly different.

But common sense does not necessarily match biology. Such is the case with a gene called SPOP. In recent years, research groups have identified SPOP mutations in subsets of prostate cancer and endometrial cancer patients. The mutations arise in slightly different locations within the same region of the SPOP protein (in its substrate recognition domain, to be precise) in each cancer.

Surprisingly, the mutations have completely opposite effects in the two cancers, despite their close proximity. As a multicenter team including Namrata Udeshi and Steven Carr of the Broad Institute’s Proteomics Platform and led by Broad Cancer Program alum Jean-Philippe Theurillat (now at the Institute of Oncology Research in Southern Switzerland) reported in Nature Medicine, endometrial cancer’s SPOP mutations sensitize tumor cells to treatment with BET inhibitors, a class of compounds being actively pursued in clinical trials. Those in prostate cancer, however, promote resistance to the compounds.

The BroadMinded blog asked Udeshi, Carr, and Theurillat about their findings and the lessons they holds for the approaches scientists take as they translate cancer genetic discoveries into therapeutics. Their answers have been edited for brevity and clarity.

Why did you look into these tumor-specific mutations in SPOP?

Jean-Philippe Theurillat: We discovered the first SPOP mutations in prostate cancer in 2012, and others investigating SPOP in other tumors also found mutations in endometrial cancer. Surprisingly, they were within the same domain but hit other amino acid residues, indicating that somehow the molecular mechanisms affected by the mutations are different.

We had already been working with Namrata and Steve to understand the molecular alterations induced by SPOP mutations in prostate cancers, and decided to apply the same approach to look in an unbiased fashion at the protein in endometrial cancer.

We ended up switching gears a little bit when we found that these mutations greatly alter expression of BET proteins, which are promising therapeutic targets. We thought it would be interesting to investigate the effects of the different mutations on cells’ sensitivity to BET inhibitors. It was one of those projects where you start in one direction and end up going another way.

Where are the SPOP mutations in prostate and endometrial cancers relative to each other?

JPT: The prostate cancer mutations are on the outer surface of a domain that is important for capturing SPOP’s substrate. The endometrial mutations are in the same domain, but in a functionally uncharacterized territory. We don’t know what the mutations there do, but from the data Namrata and Steve produced, it looks like the BET proteins and certain other substrates bind to the endometrial mutant protein better than usual.

Do you know of any other examples of this phenomenon, where cancer-specific mutations in the same protein’s domain have radically different effects?

JPT: There are certain tumor types where you see slightly different mutations between cancers, but not this drastically different outcome.

Do these findings teach us any lessons about how to study the functional consequences of genetic variations?

JPT: It’s frequently assumed in cancer precision medicine that similar mutations will have similar treatment impacts, but one of our main lessons is that you cannot directly infer how a drug will work in two different cancers based on similarity of mutations.

Steve Carr: There are clear clinical implications for these findings, but it’s difficult to find patients in which to test the findings because the mutations themselves are relatively rare.

Namrata Udeshi: From the technology side, we have also shown through this study that we’re there with state-of-the-art proteomics, that we can survey the proteome at deep coverage in an unbiased fashion, that we can see the effects of mutations in putative and known substrates, and that we can achieve the coverage and sensitivity needed to define particular modification states.

Do the findings have implications for drug development?

JPT: I’m not aware of any direct implications for drug development, but it’s interesting to find that specific mutations in a protein like SPOP can cause a gain of function for one substrate and a loss of function for others.

NU: As for this specific case, it would be interesting to look at the proteins that the SPOP mutants interact with, to see whether these other proteins contribute to the mutations’ effects.

Do your results help make an argument for bringing together as many kinds of data as you can about a target before exploiting it?

JPT: Definitely. We’ve been talking about this mutually exclusive mutation pattern in prostate and endometrial cancers, but it goes further. We’re starting to explore additional features of these mutations — for example, the fact that in prostate cancer, one wild type copy of SPOP is always retained, whereas in endometrial cancer this copy is typically lost. We’re starting to see that this context matters for these SPOP mutations, and that the protein interactions affected by the mutation matter as much as the presence of the mutation.

SC: This required analysis of the proteins using mass spectrometry-based proteomics. It ultimately isn’t a problem that could be solved by genomics alone.