ACADEMIA

A University of Queensland study mapping the evolution of genes has shed light on the role of gene transfer in bacterial diseases. The study, published recently in Proceedings of the National Academy of Sciences USA, was conducted by three scientists at UQ’s Institute for Molecular Bioscience. The study is believed to be the largest biologically inspired computation carried out in Australia, with over 500,000 hours of supercomputing. The scientists used the national supercomputing facility of the Australian Partnership in Advanced Computing. Dr Robert Beiko, Professor Mark Ragan and Mr Timothy Harlow examined the genomes of 144 species of bacteria, in an effort to map how genes are shared between bacteria. The study highlights lateral genetic transfer – a process where genes are transferred between organisms that aren’t directly related. The map has traced the paths of lateral gene transfer, from the bacteria that donated the genes to the ones that received them. Their results clearly show genetic modification of organisms by lateral transfer is a widespread natural phenomenon, and it can occur even between distantly related organisms, although particularly those which live in a similar environment. The discovery that lateral transfer is so widespread shows how disease-causing bacteria can quickly become resistant to treatment: a bacterium with genes that confer drug resistance can, through lateral transfer, rapidly spread them to other bacteria, instead of just to their own offspring. From the early days of the science of genetics, it was assumed that transfer of genes could only be vertical, i.e. from parents to offspring. But more recently, scientists have become aware of lateral genetic transfer, which occurs in bacteria through methods such as bacterial viruses and direct contact between cells. "The idea of lateral genetic transfer has been around for a few years, but what was missing was a good, hard, rigorous look at it," Professor Ragan said. "Previous studies have either been small scale, or used statistical shortcut methods."