Overview of SICKO workflow and data output.

(a) Flow chart of SICKO workflow. [i] C. elegans are age-synchronized as eggs and allowed to grow to adulthood. [ii] Adult worms are challenged with fluorescently labeled bacteria or other microbes. [iii] Following challenge, worms are allowed to crawl on media seeded with non-fluorescent bacteria for 16 hours to wash bacteria from intestine and external surfaces that are not part of an adherent colony. [iv] Individual animals are plated on wells within a single-worm culture environment and [v] imaged daily. [vi] Images are analyzed and compiled. (b) Representative confocal image of an individual C. elegans harboring an intestinal E. coli labeled with GFP following challenge and wash. (c) Image of a single-worm culture environment. Worms are housed on individual wells of a Terasaki tray filled with NGM and seeded with bacteria. Wells are surrounded by an aversive barrier of palmitic acid to prevent fleeing. The Terasaki tray is mounted inside a single-well tray on a custom 3D printed adapter. The space around the Terasaki tray is filled with saturated water absorbing crystals, and the single well tray is wrapped in parafilm to prevent the wells from drying out. (d) Representative images of a single C. elegans harboring a GFP labeled E. coli intestinal colony over time. Images were taken using the GFP channel on a fluorescent stereoscope and processed with the SICKO software. GFP-labeled E. coli is shown in white. (f) Representative heatmap representation of data from a population of C. elegans challenged with fluorescently labeled bacteria. Each row is a single animal and each column is a day. Blue and yellow boxes indicate colony size. Red boxes indicate dead worms and black boxes indicate censored data. Rows above the white line represent animals harboring a bacteria colony, while rows below the white line represent animals without a colony.

C. elegans lacking pmk-1 are more susceptible to E. coli colonization than wild type.

(a) Heatmap representing colony size (based on fluorescent area) for wild type (left) and pmk-1 mutant (right) C. elegans challenged with GFP-labeled E. coli. Death was highly correlated with colonization status in both wild type (p < 0.001, Pearson’s chi-squared test) and pmk-1 (p < 0.001, Pearson’s chi-squared test) animals, and with genotype (p < 0.01, Pearson’s chi-squared test). (b) The proportion of animals harboring a detectable E. coli colony (p = 0.67, two-sided Welch’s t test) and (c) mean colony area (p = 0.36, two-sided Welch’s t test) were not significantly different between wild type and pmk-1 knockout C. elegans 1 day following challenge. (d) Animals lacking pmk-1 developed significantly more late emerging colonies that were not initially detectable that wild type animals (p < 0.01, log rank test). In C. elegans harboring an E. coli colony, the day when the colony was first detected was significantly associated with lifespan in both (e) wild type (p < 0.05, linear regression) and (f) pmk-1 knockout (p < 0.001, linear regression) animals. Points represent values for individual animals. (g) The time between the initial detection of a colony and death was significantly longer for pmk-1 relative to wild type animals (p < 0.05, Welch’s t test). For box-and-whisker and violin plots, center bar or white point represent median, boxes represent upper and lower quartile, whiskers represent the 5th and 95th percentile, and points indicate outliers. Sample sizes: wild type, Ncolonized = 70, Nuncolonized = 171, Ntotal = 241; pmk-1, Ncolonized = 69, Nuncolonized = 101, Ntotal = 170. * p < 0.05, ** p < 0.01, *** p < 0.001, n.s. = not significant for indicated statistical test.

Challenging C. elegans with P. aeruginosa PA14 results in a greater number of more rapidly progressing colonies relative to E. coli OP50.

(a) Heatmap representing colony size (based on fluorescent area) for wild type C. elegans challenges with GFP-labeled E. coli OP50 (left) or mScarlet labeled P. aeruginosa PA14 (right). Death was highly correlated with colonization status for both E. coli (p < 0.001, Pearson’s chi-squared test) and P. aeruginosa (p < 0.001, Pearson’s chi-squared test), and with bacterial species (p < 0.001, Pearson’s chi-squared test). (b) The proportion of animals harboring a detectable colony (p < 0.05, two-sided Welch’s t test) and (c) mean colony area (p = 0.36, two-sided Welch’s t test) were significantly higher for P. aeruginosa than E. coli 1 day following challenge. (d) Animals challenged with P. aeruginosa developed substantially more late emerging colonies that were not initially detectable than animals challenged with E. coli (p < 0.001, log rank test). In C. elegans harboring a colony, the day when the colony was first detected was significantly associated with lifespan for (e) E. coli (p < 0.001, linear regression) and (f) P. aeruginosa (p < 0.001, linear regression) challenged animals. Points represent values for individual animals. (g) The time between the initial detection of a colony and death was significantly longer for P. aeruginosa relative to E. coli challenged animals (p < 0.001, two-sided Welch’s t test). For box-and-whisker and violin plots, center bar or white point represent median, boxes represent upper and lower quartile, whiskers represent the 5th and 95th percentile, and points indicate outliers. Sample sizes: E. coli OP50, Ncolonized = 40, Nuncolonized = 233, Ntotal = 273; P. aeruginosa PA14, Ncolonized = 147, Nuncolonized = 127, Ntotal = 247. * p < 0.05, ** p < 0.01, *** p < 0.001, n.s. = not significant for indicated statistical test.

C. elegans lacking pmk-1 have similar lifespan to wild type animals in both colonized and uncolonized subgroups.

C. elegans harboring a GFP-labeled E. coli OP50 colony are substantially shorter lived than animals without colonies in both (a) wild type (p < 0.001, log rank test) and (b) pmk-1 knockout populations (p < 0.001, log rank test). (c) Lifespan of wild type and pmk-1 knockout animals is not significantly different within the colonized (p = 0.15, log rank test) and uncolonized (p = 0.055, log rank test) subpopulations. The time between initial detection of a colony and death of the animal trended lower but was not significantly different between animals with colonies that were detected late (day 3 or later post challenge) vs. early (day 1 or 2 post-challenge) for both (d) wild type (p = 0.070, two-side Welch’s t test) and (e) pmk-1 (p = 0.13, two-side Welch’s t test) animals. For C. elegans harboring an E. coli colony, the area of the colony on a given day was significantly and negatively associated with remaining lifespan in both (f) wild type (p < 0.01, linear regression) and (g) pmk-1 (p < 0.001, linear regression) animals. Each point represents one animal on one day. For violin plots, center bar or white point represent median, boxes represent upper and lower quartile, whiskers represent the 5th and 95th percentile, and points indicate outliers. Sample sizes: wild type, Ncolonized = 70, Nuncolonized = 171, Ntotal = 241; pmk-1, Ncolonized = 69, Nuncolonized = 101, Ntotal = 170. * p < 0.05, ** p < 0.01, *** p < 0.001, n.s. = not significant for indicated statistical test.

C. elegans challenged with P. aeruginosa PA14 are short-lived relative to animals challenged with E. coli OP50 in both colonized and uncolonized subgroups.

C. elegans harboring colonies are substantially shorter lived than animals without colonies in both populations challenges with (a) GFP-labeled E. coli OP50 (p < 0.001, log rank test) and (b) mScarlet-labeled P. aeruginosa PA14 (p < 0.001, log rank test). (c) Lifespan of P. aeruginosa challenged C. elegans is significantly shorter than that of E. coli challenged animals in both colonized (p < 0.001, log rank test) and uncolonized (p < 0.001, log rank test) subpopulations. The time between initial detection of a colony and death of the animal trended lower but was not significantly different between animals with colonies that were detected late (day 3 or later post challenge) vs. early (day 1 or 2 post-challenge) for both animals challenged with (d) E. coli (p = 0.061, two-side Welch’s t test) and (e) P. aeruginosa (p = 0.11, two-side Welch’s t test). For C. elegans harboring an E. coli colony, the area of the colony on a given day (f) was not significantly associated with remaining lifespan for animals challenged with E. coli (p = 0.97, linear regression), but (g) was significantly and negatively associated with remaining lifespan for animals challenged with P. aeruginosa (p < 0.01, linear regression) animals. Each point represents one animal on one day. For violin plots, center bar or white point represent median, boxes represent upper and lower quartile, whiskers represent the 5th and 95th percentile, and points indicate outliers. Sample sizes: E. coli OP50, Ncolonized = 40, Nuncolonized = 233, Ntotal = 273; P. aeruginosa PA14, Ncolonized = 147, Nuncolonized = 127, Ntotal = 247. * p < 0.05, ** p < 0.01, *** p < 0.001, n.s. = not significant for indicated statistical test.

Faster colony growth is associated with reduced lifespan.

Colony growth rate is significantly and negatively associated with lifespan in both (a) wild type (p < 0.001, linear regression) and (b) pmk-1 (p < 0.001, log rank) animals. (c) Colony growth rate is not significantly different between wild type and pmk-1 animals (p = 0.62, two-sided Welch’s t test). Colony growth rate is significantly and negatively associated with lifespan in C. elegans challenged with both (a) E. coli (p < 0.01, linear regression) and (b) P. aeruginosa (p < 0.01, log rank) animals. (c) Colony growth rate is significantly higher in C. elegans challenges with P. aeruginosa relative to C. elegans challenges with E. coli (p < 0.05, two-sided Welch’s t test). Individual colony growth rate was estimated as the slope of the colony area over time, calculated using linear regression; each point represents an individual animal (panels a-f). Infection severity is significantly higher for (g) pmk-1 vs. wild type animals and for animals challenged with (h) P. aeruginosa vs. E. coli. Infection severity is estimated by adjusting the colony area in each animal for rate of colonization and prior deaths within the same treatment group using the SICKO coefficient (see Methods). For violin plots, center bar or white point represent median, boxes represent upper and lower quartile, whiskers represent the 5th and 95th percentile, and points indicate outliers. Sample sizes: wild type, Ncolonized = 70, Nuncolonized = 171, Ntotal = 241; pmk-1, Ncolonized = 69, Nuncolonized = 101, Ntotal = 170; E. coli OP50, Ncolonized = 40, Nuncolonized = 233, Ntotal = 273; P. aeruginosa PA14, Ncolonized = 147, Nuncolonized = 127, Ntotal = 247. * p < 0.05, ** p < 0.01, *** p < 0.001, n.s. = not significant for indicated statistical test.

Animals lost during wash step.

Counts of total live and dead C. elegans following the wash step (Fig. a[iii]) across each biological replicate experiment.

Sample size per replicate.

Counts of total, colonized, and uncolonized animals in each experimental replicate.