Caenorhabditis elegans is commonly used as an infection model for pathogenesis studies in 2Pseudomonas aeruginosa. While the standard virulence assays rely on the slow and fast killing or 3 paralysis of nematodes, here we developed a behaviour assay to monitor the preferred bacterial 4 food sources of C. elegans. The type III secretion system is a well-conserved virulence trait that 5is not required for slow or fast killing of C. elegans. However, ΔexsE mutants that are competent 6 for hypersecretion of ExoS, ExoT and ExoY effectors were avoided as food sources in binary 7 assays. Conversely, mutants lacking the secretion machinery or type III effectors were preferred 8 food sources for PAO1. In binary feeding assays, both food sources were ingested and observed 9 in the gastrointestinal tract, but non-preferred food sources were ultimately avoided. Next we 10 developed a high throughput feeding behaviour assay to test a library of 2370 transposon mutants 11 in order to identify preferred food sources. After primary and secondary screens, 37 mutants were 12 identified as preferred food sources, which included mutations in many known virulence genes 13 and that showed reduced virulence in the slow killing assay. We propose that C. elegans feeding 14 behaviour can be used as a sensitive indicator of virulence for bacterial strains that have moderate 15 worm killing activity.
C. elegans is an important model organism for developmental biology and infectious 22 diseases research. The nematode has been used for numerous studies as an infection host for 23Pseudomonas aeruginosa and many other bacteria (Sifri, Begun & Ausubel, 2005). C. elegans is 24 a bacteriovore that forages for bacteria in rotting plants and soil. When C. elegans is fed a lawn 25 of P. aeruginosa PA14, the gut is colonized and numerous virulence factors contribute to worm 26 death over a period of days, also known as slow killing (Tan, Mahajan-Miklos & Ausubel, 1999; 27 Feinbaum et al., 2012). Defects in virulence of transposon mutants fed to synchronized worms 28 are determined by decreased nematode killing kinetics relative to the wild type strain. After 29 ingestion, P. aeruginosa caused gut distension, the production of biofilm-like extracellular matrix 30 in the lumen, penetrated the intestinal barrier and invaded epithelial cells to some extent 31 (Irazoqui et al., 2010). P. aeruginosa PA14 uses a wide range of virulence phenotypes in the slow 32 killing assay (Feinbaum et al., 2012), but when PA14 is grown on rich, high osmolarity medium, 33 the fast killing pathway does not require live bacteria and is due to the production of toxic 34 phenazine compounds (Cezairliyan et al., 2013). P. aeruginosa strains display a range of 35 virulence phenotypes and PAO1 is among the strains with moderate slow killing activity (Lee et 36 al., 2006) . However, PAO1 was shown to induce a rapid, paralytic killing mechanism dependent 37 on hydrogen cyanide production (Gallagher & Manoil, 2001 known that C. elegans does have a feeding preference for cer...