Skip to main content
Premium Trial:

Request an Annual Quote

Comparative Genomics Study Characterizes Legionella Effectors

NEW YORK (GenomeWeb) – In Nature Genetics today, researchers from Israel and the US reported on efforts to find virulence contributors known as effectors in more than three dozen Legionella species.

"De novo sequencing of 38 Legionella species allowed us to explore the diversity of effectors used by the Legionella genus, to study the evolution of these special proteins, and to infer new potential functions mediated by them," wrote the researchers led by co-senior authors Gil Segal of Tel Aviv University and Howard Shuman of the University of Chicago.

The team noted that hundreds of such effectors have already been described in the Legionnaires' disease culprit L. pneumophila. Using new genome sequences for 38 species, the group set out to find and characterize effector sequences in other Legionella species.

The search led to nearly 5,900 candidate effectors. But beyond a set of seven core effector groups, the researchers saw very little overlap between effectors used by various Legionella species. And given the sequence composition of the species-specific effectors, they speculated that at least some of the effector-coding sequences may have been nabbed by horizontal gene transfer.

Past studies have started uncovering the suite of effector proteins that help the Legionnaires' disease-causing pathogen L. pneumophila shuttle around hundreds of proteins involved in infiltrating and controlling host cells.

For the current study, Segal and colleagues set out to find effector candidates in other Legionella species, including pathogenic bugs and other more innocuous members of the genus.

Using Illumina HiSeq instruments, the team sequenced DNA libraries for 38 Legionella isolates, which were then compared with genome sequences for three species sequenced previously: L. pneumophila, L. longbeachae, and L. drancourtii.

The resulting genomes spanned between 2.37 million bases and more than 4.8 million bases apiece and contained more than 16,400 orthologous protein-coding gene groups. Of those, the researchers found that just 1,054 orthologous genes were shared across all 41 Legionella genomes considered.

Using several dozens of the most highly shared proteins, the team put together a phylogenetic tree for the Legionella species that let them define three main clades.

The deepest branching clade contained L. oakridgensis, L. londiniensis, and L. adelaidensis, while the two remaining clades housed 15 species, and 22 species, respectively. A species called L. geestiana formed an outgroup in relation to the other species, the researchers noted, in keeping with relationships reported in past studies.

The researchers first searched the genomes for sequences resembling a well characterized secretion system called Icm/Dot type IVB, which carts around some 300 effector proteins in L. pneumophila.

From there, they focused on potential effector proteins themselves, uncovering between 52 and 247 candidate genes in the sequenced bugs. Interestingly, the size of these apparent effector repertoires appeared to vary depending on the clade in question.

Although the team was able to place the 5,885 potential effector genes into 608 orthologous groups, some 42 percent of these were found in only one species and the orthologous effector groups that were shared occurred in just a handful of species. Of the seven core effector genes, meanwhile, only one was shared across all 41 species.

The architecture and apparent sources for these effectors varied as well. For example, the team detected signs of effectors produced through rearrangements of conserved domains, pseudogenization, or potential horizontal gene transfer, with potentially pathogenic species typically undergoing effector gains.

Based on these and other findings, the study's authors suspect that "certain Legionella species, including the most pathogenic ones, acquire genetic information from outside the Legionella genus more frequently than others and adapt these sequences to encode functional effector proteins."