NEW YORK (GenomeWeb) – A team of researchers has explored the roles of thousands of bacterial protein-coding genes with currently unknown function by studying the phenotypes of mutant microbes that lack them.
The team, led by Adam Arkin and Adam Deutschbauer at Lawrence Berkeley National Laboratory and Matthew Blow at the Joint Genome Institute, published its findings today in Nature.
"This is the first really large, systematic experimental effort to try to assign functions to bacterial genes of unknown function," LBNL biologist and senior author Deutschbauer said in a statement. He explained that while it is easy to sequence microbial genomes, "we cannot currently assign confident functions for the majority of genes identified by sequencing."
While thousands of bacterial genomes have been sequenced to date, about a third of the proteins they encode have no predicted function. To investigate their role, the researchers studied the effect of loss-of-function mutations in protein-coding genes under several growth conditions. They then used the mutant phenotypes they observed, in conjunction with comparisons across genomes and species, to come up with gene annotations.
The approach they used, called transposon mutagenesis followed by sequencing (TnSeq), enabled them to assess mutant phenotypes in tens of thousands of mutant bacteria in parallel. And by using a version of the technique that involves random DNA barcoding (RB-TnSeq), they were also able to measure phenotypes across conditions.
For their study, the researchers analyzed 32 diverse bacterial species. For each, they generated a randomly barcoded transposon mutant library and tested the growth of the mutants under a variety of conditions, such as different carbon and nitrogen sources, or in the presence of antibiotics or metals. Sequencing the libraries allowed them to see which protein-coding genes were disrupted in the mutant bacteria, and how the frequency of certain mutants changed under different growth conditions.
In total, they measured mutant phenotypes for almost 12,000 poorly annotated protein-coding genes.
They drew conclusions about the biological function of individual proteins from the mutant fitness data they obtained, either by looking at phenotypes that only occurred under a specific condition or by analyzing similar fitness profiles that are shared by several genes across all conditions.
Doing that, they proposed specific functions for transporter proteins, catabolic enzymes, and uncharacterized protein families and identified several novel potential DNA repair proteins. However, they noted that most of these predictions still need to be experimentally validated.
Overall, the researchers identified functional leads for potential orthologs of just 12 percent of all bacterial proteins lacking good annotations. "Improving this coverage will require a larger effort to generate mutants in more diverse bacteria," they wrote.