Pathogens deliver complex arsenals of translocated effector proteins to host cells during infection, but the extent to which these proteins are regulated once inside the eukaryotic cell remains poorly defined. Among all bacterial pathogens, Legionella pneumophila maintains the largest known set of translocated substrates, delivering over 300 proteins to the host cell via its Type IVB, Icm/Dot translocation system. Backed by a few notable examples of effector–effector regulation in L. pneumophila, we sought to define the extent of this phenomenon through a systematic analysis of effector–effector functional interaction. We used Saccharomyces cerevisiae, an established proxy for the eukaryotic host, to query > 108,000 pairwise genetic interactions between two compatible expression libraries of ~330 L. pneumophila‐translocated substrates. While capturing all known examples of effector–effector suppression, we identify fourteen novel translocated substrates that suppress the activity of other bacterial effectors and one pair with synergistic activities. In at least nine instances, this regulation is direct—a hallmark of an emerging class of proteins called metaeffectors, or “effectors of effectors”. Through detailed structural and functional analysis, we show that metaeffector activity derives from a diverse range of mechanisms, shapes evolution, and can be used to reveal important aspects of each cognate effector's function. Metaeffectors, along with other, indirect, forms of effector–effector modulation, may be a common feature of many intracellular pathogens—with unrealized potential to inform our understanding of how pathogens regulate their interactions with the host cell.
Query strain identities and corresponding screening set.
In order to obtain an IDTS-wide genetic interaction map, we modified the automated, high-density replica plating approaches previously developed for analyzing S. cerevisiae double mutants through Synthetic Genetic Array (SGA) analysis (Tong AH, Evangelista M, Parsons AB, Xu H, Bader GD, Page N, Robinson M, Raghibizadeh S, Hogue CW, Bussey H, Andrews B, Tyers M, Boone C (2001) Systematic genetic analysis with ordered arrays of yeast deletion mutants. Science 294: 2364-2368). Instead of looking at double mutants, however, we used yeast genetics to systematically assess the effects of Legionella effector co-expression on yeast growth. Query strains that express one effector were mated to an array of ~330 effectors in groups of ~10 queries at a time ("Analysis Set"). The arrays were then imaged using a high-resolution camera and the spot sizes were quantified using SGAtools (http://sgatools.ccbr.utoronto.ca/) (Wagih O, Usaj M, Baryshnikova A, VanderSluis B, Kuzmin E, Costanzo M, Myers CL, Andrews BJ, Boone CM, Parts L (2013) SGAtools: One-stop analysis and visualization of array-based genetic interaction screens. Nucleic acids research 41: W591-596). Outlier spot sizes flagged by the Jackknife filter (JK) in SGAtools were removed and the average and standard deviation of the remaining values were calculated and normalized to the average empty vector control. This spreadsheet lists all query strains and links them to one or more specific analysis set.
Array_set_annotation.xls
SGA output for analysis sets 1-10.
Query strains that express one effector were mated to an array of ~330 effectors in groups of ~10 queries at a time ("Analysis Set"). The arrays were then imaged using a high-resolution camera and the spot sizes were quantified using SGAtools (http://sgatools.ccbr.utoronto.ca/). Outlier spot sizes flagged by the Jackknife filter (JK) in SGAtools were removed and the average and standard deviation of the remaining values were calculated and normalized to the average empty vector control. This .zip archive includes spreadsheets that encompass the raw SGAtools data output from the paper for analysis sets 1-10.
UrbanusMSB_SGA.zip
Urbanus_SGA_1-10.zip
SGA output for analysis sets 11-20.
Query strains that express one effector were mated to an array of ~330 effectors in groups of ~10 queries at a time ("Analysis Set"). The arrays were then imaged using a high-resolution camera and the spot sizes were quantified using SGAtools (http://sgatools.ccbr.utoronto.ca/). Outlier spot sizes flagged by the Jackknife filter (JK) in SGAtools were removed and the average and standard deviation of the remaining values were calculated and normalized to the average empty vector control. This .zip archive includes spreadsheets that encompass the raw SGAtools data output from the paper for analysis sets 11-20.
Urbanus_SGA_11-20.zip
SGA output for analysis sets 21-30.
Query strains that express one effector were mated to an array of ~330 effectors in groups of ~10 queries at a time ("Analysis Set"). The arrays were then imaged using a high-resolution camera and the spot sizes were quantified using SGAtools (http://sgatools.ccbr.utoronto.ca/). Outlier spot sizes flagged by the Jackknife filter (JK) in SGAtools were removed and the average and standard deviation of the remaining values were calculated and normalized to the average empty vector control. This .zip archive includes spreadsheets that encompass the raw SGAtools data output from the paper for analysis sets 21-30.
Urbanus_SGA_21-30.zip
SGA output for analysis sets 31-37.
Query strains that express one effector were mated to an array of ~330 effectors in groups of ~10 queries at a time ("Analysis Set"). The arrays were then imaged using a high-resolution camera and the spot sizes were quantified using SGAtools (http://sgatools.ccbr.utoronto.ca/). Outlier spot sizes flagged by the Jackknife filter (JK) in SGAtools were removed and the average and standard deviation of the remaining values were calculated and normalized to the average empty vector control. This .zip archive includes spreadsheets that encompass the raw SGAtools data output from the paper for analysis sets 31-37.
Urbanus_SGA_31-37.zip