New & Noteworthy
April 18, 2013
One of the ways you can tell a human cell is cancerous is by taking a peek at its genome. Instead of the orderly 23 pairs of chromosomes seen in a normal cell, the cancerous one has a jumbled mess of a genome. There are extra chunks sticking here and there, chunks missing, and lots of other oddities.
Besides looking untidy, this sort of chaos also causes something called copy number variation (CNV). In CNV, there are either more or less than the usual two copies of some genes. Having the wrong number of copies of certain genes can definitely cause problems.
There is some debate out there about whether CNV causes a cell to go cancerous or if it is just an effect of the cancer. In a new study, de Clare and coworkers provide strong evidence that for many genes in the yeast Saccharomyces cerevisiae, having just one copy in a diploid background leads to faster growth, poor cell cycle control, and an aversion to apoptosis (programmed cell death). This argues strongly that CNV can actually cause a cell to go cancerous. This suggestion is strengthened further by the fact that many of the genes they identified are orthologs of human genes that exist as single copies in certain cancers.
Earlier studies from this group looked at the growth rates of over 5,800 heterozygous diploid yeast mutants, each missing one copy of a particular gene, and found around 600 that actually grew faster than wild type. You might not expect such a high number at first blush, since it seems like a single celled organism would have evolved to grow as fast as it can. The authors hypothesized that there must be a strong selective advantage to having these genes, outweighing the fact that they slow down growth.
Looking more closely, they found that the genes in this set were significantly more likely than the average gene to have functions that keep the genome stable, such as DNA damage repair. They were also highly conserved across the Ascomycete fungi, confirming their importance.
The next step was to see whether there might be any connection to human cancer. They took a subset of these genes – 30 genes involved in DNA repair and sister chromatid segregation – and compared them to human genes. Nineteen of the yeast genes had a human ortholog, and 17 of those human genes exist as a single copy in many cancers, suggesting that having only one copy of these genes may contribute to a cell’s cancer phenotype.
If copy number variation of those genes contributes to cancer in human cells, does it confer a cancer-like phenotype on yeast? The researchers found that the heterozygous yeast mutants showed characteristics of cancer cells such as altered cell cycle, a decrease in apoptosis, and lowered sensitivity to anti-cancer drugs. So the increased growth conferred by the mutations comes with a high cost: increased genome instability and cancer-like symptoms.
Because this cancer-like phenotype occurs in yeast, it will be an excellent model to study exactly how particular genes contribute to it. But these findings could also have a more immediate impact on cancer treatment. Certain experimental cancer treatments work by decreasing the activity of the proteins produced by some of these genes. If a treatment only partly knocks down the activity, then it may actually encourage cancer growth. It would mimic the effects of having a single copy of a gene. The authors actually show that this is the case in yeast for some of the drugs they tested.
And this isn’t a worry just for the drug targets themselves. The drugs aren’t completely specific…they can affect other genes too, again mimicking the effects of having a single copy of one of these other genes. Add to this the fact that each genomically jumbled cancer cell may have different proportions of genes, and you have quite a mess. As usual, yeast can swoop in and save the day.
Scientists may be able to use this and other yeast libraries to quickly screen varying amounts of potential new drugs for their effects on growth. Not only that, they’ll be able to identify what pathways these drugs are hitting in addition to the one(s) that are targeted. This should make the process of drug optimization move ahead much more quickly. Thanks yeast!