"MicroRNAs are affecting the majority of protein-coding genes, either at a functional level or an evolutionary level," says Andrew Grimson, a post-doctoral fellow in Bartel's lab.
In order to make a protein, a gene codes for a specific molecule called messenger RNA, or mRNA. Each mRNA molecule contains a blueprint for making a protein. A microRNA can bind to a short sequence on a targeted mRNA and suppress protein production.
In a paper published last January in the journal Cell, Bartel's lab, in collaboration with Chris Burge's lab at MIT, presented evidence that one third of human genes are regulated by microRNAs. In this new study, published online Nov. 24 in Science, the researchers demonstrate that microRNAs affect the expression or evolution of the majority of human genes.
Nearly all genes, the authors explain, contain short sequences that match portions of microRNAs. Some of these potential microRNA target sites are evolutionarily "conserved," meaning that they show up in the same spot on the same gene across species as disparate as the mouse and the chicken. The authors of last January's Cell paper showed that thousands of human genes contain microRNA sites that are conserved in this way. To the extent that evolution has preserved these sites more than would be expected by chance, scientists have regarded them as sites that microRNAs target.
But is a matching sequence all that's required for microRNA targeting and gene regulation, and do nonconserved sites also have the potential to disrupt protein production?
In the new study, scientists in the Bartel lab designed an experiment that zeroed in on these nonconserved targets. Grimson took mRNAs whose target sequences were not conserved and exposed them to microRNAs, which latched on without a problem. The experiment proved that a matching sequence is generally sufficient to disrupt mRNA's ability to make protein.
But while Grimson showed that, at least in the lab, microRNAs could regulate mRNAs with nonconserved sites, the researchers still didn't know the extent to which nonconserved mRNAs coexisted with their matching microRNAs in the natural cell environment. To answer this question, the researchers turned to gene expression patterns of different types of mouse cells.
Kyle Kai-How Farh, a graduate student in Bartel's lab, found that mRNAs with nonconserved sites were generally absent in cells with corresponding microRNAs--more absent than statistical models suggested. The researchers concluded that over the course of evolution many mRNAs, in order to maintain their functions and ensure fitness of the organism, have quickly lost sites that pair up with microRNAs.
In addition to the thousands of cases where genes have avoided microRNA targeting, Farh also investigated the opposite extreme, cases where genes have maintained microRNA target sites over the course of evolution. He found that as immature muscle cells stop dividing and become mature muscle cells, microRNAs are activated and suppress genes that are no longer needed at such high levels in the mature muscle. "Many of these evolutionarily conserved microRNA targets are known to be active in the processes of cell proliferation, development, and cancer," says Farh. "Our genomes have good reason to maintain the microRNA targeting sites necessary for turning down these genes at the appropriate place and time."
An emerging idea is that microRNAs often act to reduce the quantity of protein a gene produces without shutting it off all together. "We think the microRNAs are sometimes having what you can call a dampening effect," says Bartel, who is also a Howard Hug hes Medical Institute investigator and MIT professor of biology. "They appear to be helping cells achieve optimal levels of proteins."
"MicroRNAs are leaving an evolutionary footprint on the majority of the mammalian genome," says Grimson. "Some genes are trying to preserve beneficial microRNA sites and others are evolving in order to avoid developing harmful ones."