Susceptibility of Caenorhabditis elegans to Burkholderia infection depends on prior diet and secreted bacterial attractants

Abstract

The nematode Caenorhabditis elegans may be killed by certain pathogenic bacteria and thus is a model organism for studying interactions between bacteria and animal hosts. However, growing nematodes on prey bacteria may influence their susceptibility to potential pathogens. A method of axenic nematode culture was developed to isolate and quantify interactions between C. elegans and potentially pathogenic strains of the Burkholderia cepacia complex. Studying these dynamics in liquid solution rather than on agar surfaces minimized nematode avoidance behavior and resolved more differences among isolates. Most isolates of B. cenocepacia, B. ambifaria and B. cepacia caused 60-80% mortality of nematodes after 7 days, whereas isolates of B. multivorans caused less mortality (<25%) and supported nematode reproduction. However, some B. cenocepacia isolates recovered from chronic infections were much less virulent (5-28% mortality). As predicted, prior diet altered the outcome of interactions between nematodes and bacteria. When given the choice between Burkholderia and E. coli as prey on agar, axenically raised nematodes initially preferred most lethal Burkholderia isolates to E. coli as a food source, but this was not the case for nematodes fed E. coli, which avoided toxic Burkholderia. This food preference was associated with the cell-free supernatant and thus secreted compounds likely mediated bacterial-nematode interactions. This model, which isolates interactions between bacteria and nematodes from the effects of prior feeding, demonstrates that bacteria can influence nematode behavior and their susceptibility to pathogens.

Figure 1. Relative virulence of representative strains of the B. cepacia complex.

Strains vary significantly in their ability to kill C. elegans (mortality at 120 h, F = 264, p<0.0001). A. Percent nematode mortality over time. B. Density of planktonic fraction (OD600) over time. Results are expressed as mean (+/− SEM) of three replicate assays. B. ceno  =  B.cenocepacia; B. cep  =  B.cepacia; B. mult  =  B. multivorans; B. amb  =  B. ambifaria.

Figure 2. Localization of bacteria ingested by C. elegans.

Confocal microscopy of a live nematode that was co-cultured with E. coli DH5α, marked with green fluorescent protein, and B. cenocepacia HI2424, marked with red fluorescent protein, is shown. B. cenocepacia (red, lower left quadrant) grows throughout the nematode gut and forms aggregates on the nematode cuticle, whereas E. coli (green, upper right quadrant) is found only in limited concentration in the mouth.

Figure 3. Effect of prior diet on susceptibility.

C. elegans diet prior to encountering B. cenocepacia strains affects its susceptibility to killing. Strain HI2424 (filled symbols) is more lethal to nematodes that had been starved in buffer (triangles) or fed heat-killed E. coli OP50 (squares) than to nematodes that were grown in axenic CeHR medium (circles); strain AU1054 (open symbols) is less lethal to nematodes that had been fed E. coli or starved in buffer, than to nematodes that were grown in axenic CeHR medium. Results are expressed as mean (+/− SEM) of three replicate assays.

Figure 4. Effect of prior exposure to E. coli on nematode feeding preference.

Nematodes raised in CeHR medium prefer lethal B. cenocepacia strains HI2424 and J2315 over E. coli, but the reverse is true for B. multivorans ATCC17616, which is nontoxic and is preferred by nematodes fed E. coli. †, pairings lethal to nematodes; other interactions supported nematode reproduction. Results are expressed as mean (+/− SEM) of three replicate assays. * p<0.05, 1 sample t-test versus expected 50%.

Figure 5. Nematode preference to bacterial culture supernatants.

C. elegans prefers the spent cell-free culture supernatant of some Bcc strains to that of its prior E. coli diet. Cell-free supernatant from each Bcc strain was spotted opposite that from an E. coli OP50 culture. The fraction of nematodes traveling towards the Bcc supernatant within 1 h (up triangles) and then towards the E. coli supernatant by 3 h (down triangles) is shown. Within the first day, eggs from adult nematodes hatched and neonate migration at 24 h was observed (open circles). The error bars are SEM from three replicate assays.