Too much mutation can damage cancer

Cancer is a genetic disease which occurs when growth controlling genes, called oncogenes, are activated inappropriately. This releases cells from the constraints that normally govern their proliferation and survival. So, you would think that the more of these factors there are active in a cell, the better a tumour might grow. In fact, this is definitely not the case: because a new study by William Lockwood from the National Human Genome Research Institute in Bethesda, looking  at human lung cancers, shows that these tumours opt either for a mutation in a gene called EGFR or another one, called KRAS, but never both...

William - These mutations typically happen in specific sets of genes that then are known to cause cancer development. Although these genes are typically seen to be mutated or disrupted in cancer, there's often only one of these mutations in each cancer type. So, there's often not more than one happening together. And our big question was, well why does this occur? Why do cancers only have one mutation and not others? The typical thought has always been that, well the cancers don't need more than one. They only need one of these genes to be mutated in order to cause cancer. So, there is no selection or pressures involved in order to select for both mutations happening at the same time.

Chris - But equally, there's no selection for not having both, is there? So, you ought to see by chance, some cells that do have both if all things are equal.

William - So, exactly. That was our point. We looked at over 600 tumours from patients with lung cancer and that's what we were looking for; is any evidence that these two mutations co-occur, happen in the same cell. And we found no evidence of this as the case.

Chris - Do you think this is just a deleterious thing that one being mutated is good for the cancer? Because it drives the cancer, but get both going down, now it's a case of overkill. Or do you think there's something else going on? In other words, when you go looking, you don't see them because any cells that have both didn't last long enough for you to detect them?

William - That could be the case. We thought there could be some technical issues involved here, but our main theory was that maybe these aren't occurring in the same cell because they actually have some deleterious effect or they actually somehow kill cancer cells that have both these mutations. So, there's a pressure against having both of these genes mutated and that's what we set out to test in this paper. So, we actually have mice that we can engineer to express these mutated genes, specifically in the lung. And these mice, with one of these genes, always get cancer at a specific time points in their life. And what we did is now we engineered the mice to force the expression of both these genes at the same time, and we predicted that if our hypothesis was true, that this would mean that if there was a negative effect of having both genes that we would see an effect on when these mice got tumours. But what we found was that they still got cancer and they still got cancer at the very same time as they do with either gene alone.

Chris - It's not just the system's a bit leaky and it works for one gene, but not for the other sometimes and that's enough to drive a few tumours.

William - No. We figured that it was, at the beginning at least, expressing both these genes. But when we actually took the tumours out of these mice later on, and actually looked to see if both these genes were expressed, we found that only one of the two was expressed. We actually think that there is a selective pressure that these tumours get around having both expression of both these genes and select for just one. So, they're actively actually silencing one of the two mutated genes.

Chris - Why are they doing that?

William - Well. We think that it's obviously the deleterious effects of having both these genes. So, we think that having two of these things on at the same time actually causes too much growth  and proliferation that is actually deleterious to cancer development. So, they just can't really cope with having both these mutations at the same time. So, we think that what they're doing is that they're actually silencing one of the two, which is then sufficient in order to cause cancer.

Chris - Why do you think that, given that cancer is a genetic instability and that  cells that are cancerous are shot full of genetic changes all over the place. Why don't they end up with some sort of compensatory mutation somewhere else in the genome that quenches one of the two damaged oncogenes? So, that the cell can tolerate having both on, it's just that it's dampening down the effect of one of them with a compensatory change elsewhere.

William - It's a really good point. I guess the big thing is we have to remember that this is an evolutionary process. So, in order to, and with selective pressures involved, and this all depends on the order of events, so if you had one mutation in one gene first and then you got the second oncogene mutation. Typically, that cell would die and there'd be no chance for there to be another selective event involved. So, it would all have to occur in random chance. So, you'd have to have that compensatory mutation before you required the second oncogene mutation which would lower the chances of that being the case.

Chris - Do you think you can exploit what you've found? That we might be able to tap into this as way, actually, of making cancers kill themselves?

William - Yes, that's what the most exciting part is, I think, for our future experiments. Is now utilising this knowledge that we have about how we can hyperload cancer cells in order to design therapeutic stratergies. So what we're looking at now is that most therapies out there are typically based on inhibiting a pathway. So these are based on drugs, for example, that shut down mutated oncogenes for example. But now we're actually looking at things that perhaps stimulate signalling in the context of a specific oncogene what we would call more agonistsic therapies. So in this context we'd actually be looking for things that stimulate signalling to a threshold level where then the cells would die.