Abstract / Introduction
Finding genes that are mutated in human cancers has given us knowledge on how cancers form. This knowledge has allowed us to develop new ways to treat cancer. In this study, researchers analyze the genetic changes that take place during the development of breast and colorectal cancers. They did so by sequencing DNA from breast and colorectal tumors while also browsing RefSeq, a database that details most known human gene sequences.
The researchers used the RefSeq database to find the sequences of known human genes. They used their knowledge of these sequences to develop primers necessary for polymerase chain reaction (PCR). Then, they used PCR to produce many copies of every gene from both tumor samples and cells from healthy people. Afterwards, they looked for differences between the genes from the tumor samples and the same genes from healthy people. They took measures to make sure these apparent differences weren’t due to errors in PCR or natural variations in DNA sequences. From then, they determined the sequences of the tumor genes that differed from normal genes. PCR was run on these specific genes again to make sure that these gene alterations were really somatic.
It was found that 9.4% of the analyzed genes from the tumor samples had mutations that were responsible for some bodily change. Most of these mutations were changes in a single nucleotide pair. There was a strikingly high number of transitions from cytosine to thymine at sites near the promoter for colorectal cancer.
Passenger Mutation Rates
Mutations in cancer are classified as either “drivers” or “passengers”. Drivers have involvement in tumor development and thus have a higher chance of manifesting during cancer progression.
Passengers appear by chance and do not affect tumor development one way or another. However, once they appear, they can remain when the cell they appear in replicates enough times.
The researchers measured the rate in which passenger mutations appear using a technique called high-density oligonucleotide microarrays. They found that roughly 0.55 passengers that result in a change in protein product show up in every one million base pairs of DNA.
Evaluating Mutated Genes
From the data collected up to this point, the researchers pinpointed mutated genes that were the most likely to be drivers. Genes that were considered more likely to be drivers had at least one protein-changing mutation and a sufficient number of mutations per nucleotide. Such genes were termed candidate cancer genes, or CAN-genes. A total of 280 CAN-genes were identified.
Genes that are mutated more frequently than expected during cancer development are likely to be drivers. Using this logic, the researchers assigned scores to the 280 CAN-genes based on their mutation frequency in tumor samples. These scores were called cancer mutation prevalence (CaMP) scores. The genes with the highest CaMP scores were considered most likely to be drivers.
To further determine how often these genes are mutated in cancers, the researchers analyzed a subset of 40 CAN-genes across 96 colorectal cancer patients. These genes’ CaMP scores were among the top half of all CAN-genes. About two-thirds of these genes were found to be mutated in at least one patient. Most were mutated in less than 5% of these patients. A bit over a third of these genes were not mutated in any patients.
Additional Analyses of Mutated Genes
When deciding which genes needed to be studied further, the researchers considered more than just a gene’s mutation rate. They also considered how likely a gene’s mutations were to cause negative effects. To figure out how likely a gene’s mutations were to interfere with protein function, they examined the structures of the mutations. They also looked to see if their mutations happened in the same place as mutations known to cause other diseases.
Analysis of Mutated Pathways
It is biological pathways rather than mutations in single genes that influence tumor development. The researchers set out to find whether certain pathways were more likely to have mutated genes in cancers than other pathways. They found 108 such pathways in breast cancers and 38 in colorectal cancers. The pathways included those that are involved in cell adhesion, the cytoskeleton, and the extracellular matrix. These findings suggest that interactions between cancer cells and extracellular environments are important in cancer development. There were also mutant protein products that interacted with other mutated genes unusually often.
The genomic landscapes of colorectal and breast cancers
In the colorectal cancers studied, the median number of mutations with a bodily effect was 76. This number was 84 for breast cancers. The average number of mutations per colorectal tumor was 49 to 111. In breast cancers, this was 38 to 193, which is less consistent. The average number of mutated CAN genes among the two groups of cancers was 14 to 15. When visualizing these mutations, the researchers found a few of what they called “gene mountains” and many of what they call “gene hills”.
The results detailed in this article add to research done on genetic changes in breast / colorectal cancers. More tumor genes have been sequenced than in past studies. This study also includes data on noncoding mutations while others do not. These contributions give a more complete picture of the genes involved in cancer development. The research team also found better ways to both assess mutation rate and find better ways to identify important mutated genes.
Research in the field of cancer genomics has made practices like analyzing a person’s genome viable approaches to fighting cancer. It can be hard to understand how exactly a mutation influences cancer formation, but new developments are making this more feasible. This study and others like it could open the doors to more personalized cancer treatments. Additionally, mutations in a given cancer can offer insight to guide patient management.