Genes Know How to Network

In the loop. A new technique produces complex maps of gene interactions and how they change as cells mature.


From directing the fate of stem cells to determining how tall we grow, the genes in our body act in complex networks. Unraveling these interactions has been a laborious chore. Now, an international collaboration has developed a more comprehensive approach that promises to speed up the process.

The new approach was developed by the Functional Annotation of the Mammalian Genome (FANTOM) international consortium, organized by RIKEN Omics Science Center in Yokohama, Japan. The group, led by genome scientist Yoshihide Hayashizaki, has developed techniques to count how frequently a gene is expressed--the process by which DNA is transcribed into RNA, which in turn codes for a protein. Many of those proteins then affect the expression of other genes, either increasing or decreasing the amount of protein produced. Hundred of genes can influence the activity of other genes at any given time. Complicated, indeed.

The FANTOM team teased out the important players--and their relationships--by trawling through this mass of data with new software. Its first target was the regulatory network involved in controlling the differentiation of THP-1 cells, a line of human leukemia cells used in laboratory experiments. The team identified 29,857 regions on the genome that respond to regulatory proteins and then tracked levels of activity at six instances during the differentiation.

The result was a bird's nest of regulatory pathways. Such pathways have typically been thought of--and drawn--as hierarchical arrangements; one or a few "master" genes at the top set off a cascade of influence that trickles down to the genes below. The FANTOM model, in contrast, has feedback loops through which regulatory proteins boost a gene's expression only to have other regulatory proteins depress it later until the cells reach a new equilibrium.

The new paper is "by far the most complete survey of a cellular differentiation event," says Juha Kere, a molecular geneticist at the Karolinska Institute in Stockholm, who is not affiliated with the FANTOM group. This better understanding of how the network functions could allow scientists to engineer new types of cells, says Hiroshi Tanaka, a bioinformaticist at Tokyo Medical and Dental University.

Two other groups associated with FANTOM put the techniques to work on different topics. A group led by Piero Carninci, a biologist at Omics, looked at how genetic elements known as retrotransposons scattered through genomes affect the synthesis of RNA. The second group, headed by John Mattick, a molecular biologist at University of Queensland, St. Lucia, in Australia, identified a new class of short RNAs that also seem to be involved in regulating gene expression, though their function needs to be confirmed.

All three teams report their results online this week in Nature Genetics.

Follow News from Science