The wild-type strain N2 (Bristol) was obtained from the Caenorhabditis Genetics Center (CGC, Minnesota, USA), the transgenic strain BJ49 kcls6 [IFB-2::CFP] was obtained from the Leube Lab (RWTH Aachen University, Aachen, Germany). All worms were grown and maintained on nematode growth medium (NGM) seeded with the Escherichia coli strain OP50 at 20°C, according to the routine maintenance protocol (Stiernagle, 2006). Worm populations were synchronized and incubated at 20°C.

Natural C. elegans microbiota isolates Pseudomonas lurida MYb11 and Pseudomonas fluorescens MYb115 were used (Dirksen et al., 2016). The microbiota isolates and the E. coli OP50 control were grown on Tryptic Soy Agar (TSA) plates at 25°C and liquid bacterial cultures were grown in Tryptic Soy Broth (TSB) in a shaking-incubator overnight at 28°C. Peptone free medium (PFM, 3% agar) plates were inoculated with 750 µL OD_600nm of 10, of the overnight cultures and used to maintain worms from L1 to the L4 stage.

Bacillus thuringiensis (MYBt679 and MYBt407) spore aliquots were obtained following a previously established protocol (Borgonie et al., 1995) with minor modifications (Kissoyan et al., 2019). The Bt spore aliquots were prepared in bulk and stored at -20°C, with concentrations of 10⁹-10¹⁰ particles/mL for Bt679 and 10³-10⁴ particles/mL for Bt407. New Bt aliquots were freshly thawed for each assay.

To fluorescently label the C. elegans natural microbiota isolates MYb11 and MYb115, we used a previously published protocol (Wiles et al., 2018) with minor modifications. We performed the bacterial mating overnight and used selective and differential media, MacConkey, to distinguish between E. coli and Pseudomonas spp. colonies to ensure the selection of the desired colonies. We further performed a confirmatory 16S rRNA PCR using primers SHP160(27f) (5’-GAGAGTTTGATCCTGGCTCAG-3’) and SHP161(1495r) (5’CTACGGCTACCTTGTTACGA-3’) and Sanger sequencing, in addition to targeted PCR(TN7) using primers WP11 (5’-CACGCCCCTCTTTAATACGA-3’) and WP256 (5’-CAGCTGCTCTCCTACTACGT-3’).

Bt679 toxins were provided by Elena A. Andreeva and Dr. Guillaume Tetreau from the group of Dr. Jacques-Philippe Colletier (Univ. Grenoble Alpes, CNRS, CEA, Institut de Biologie Structurale, Grenoble, France). Bt679 toxins were purified and the required quality control was performed according to a previously published protocol (Zárate-Potes et al., 2020).

Population growth assays were done to measure the nematode fitness through extrapolating the offspring produced by a single worm after two generations according to a standardized protocol (Dirksen et al., 2016; Kissoyan et al., 2019). Here we used three L4 larvae that grew up from L1 to L4 stage on OP50, MYb11 or MYb115, respectively. Worms were picked onto assay plates containing a lawn of adjusted bacteria (OD_600nm of 10) mixed with Bt679 spores. Worm population size was measured after an incubation period of 5 days at 20°C. The population growth assay was performed at least three times in independent runs.

B. thuringiensis survival assays were done according to the standardized protocol (Nakad et al., 2016; Kissoyan et al., 2019; Zárate-Potes et al., 2021). Worms were grown on 6 cm PFM plates with lawns containing 75 µl of MYb11, MYb115 or OP50 (OD_600nm of 10) mixed individually with different concentrations of the Bt679 spores (1:5, 1:25, 1:50) or with the Bt407 spores (1:5). The pre-prepared Bt spore aliquots were stored in 1 mL tubes, at -20°C, with a concentration range of 10⁹-10¹⁰ particles/mL for Bt679, and 10³-10⁴ particles/mL for Bt407. These were freshly thawed immediately prior to every assay. Bt407 was used as a non-pathogenic Bt control (Sheppard et al., 2013; Zárate-Potes et al., 2020). Worm survival was assessed 24 hours post infection (hpi) as alive or dead using gentle touch with a sterile platinum wire. Similarly, for the survival assay with purified Bt679 toxin, MYb11, MYb115 or OP50 (OD_600nm of 10) were mixed individually with different dilutions (1:10, 1:50, 1:100, 1:1000) of the purified toxin or no toxin (0). Bt survival assays were done in three independent runs, each with four biological replicates per treatment group and around 30 worms per biological replicate.

For the lifespan assay, eggs were harvested by bleaching and allowed to grow to L1 in M9 buffer overnight at 20°C. L1 larvae were transferred to PFM plates with lawns containing 750 µL (OD_600nm of 10) of OP50, MYb11 or MYb115 and raised at 20°C until they reached the L4 larval stage. We did not observe any developmental delay in any of the microbiota treatment conditions and the L4 stage was reached by 46 h in all three bacterial treatment conditions. L4 worms were then picked with a platinum wire pick to fresh PFM plates of the respective bacterium. All PFM assay plates were inoculated with 150 µL of the adjusted bacterial cultures in advance and stored in the fridge at 4°C. 30 min before the picking, plates were transferred to 20°C to ensure that the plates reached room temperature. Every 24 h, alive and dead worms were counted, and all living worms were picked to a new PFM plate until the worms stopped laying eggs. Worms were considered dead when they failed to respond to gentle touch with the platinum wire pick.

To measure nematode fertility in response to the different bacteria, worms were prepared as for the lifespan assay. However, single worms were picked to new plates daily, and the hatched offspring were counted until the worms reached the end of their reproduction period.

To assess differences in bacterial colonization (i.e. number of live bacteria present inside the C. elegans intestine after several washings), synchronized L4 larvae were transferred to new PFM plates every 24 h until they reached: L4, 24 h post L4 (Day 1), 72 h post L4 (Day 3) and 120 h post L4 (Day 5) stage. Worms were washed off the plates with M9T (M9+0.025% Triton-X100) and washed 5 times before being paralyzed with 10 mM tetramisole. This was followed by soft-bleaching, as described in (Dirksen et al., 2020), and the total worm number was counted and homogenized as described in (Pees et al., 2021). Homogenized worms were diluted (1:10 for L4, day 1 and day 3; 1:100 for day 5) and plated onto TSA plates, as well as the undiluted supernatant as a washing control. After 48 h at 25°C, colonies were counted, and the colony-forming units (CFUs) per worm were calculated. The soft bleaching treatment sterilized the worm surface sufficiently as the supernatant was almost always clear of viable bacterial cells.

To visualize the microbiota bacteria in the C. elegans intestine, the transgenic worm strain BJ49 was synchronized, and L1 larvae were exposed for 72 hours to fluorescently labeled bacteria (OP50::dTomato, MYb11::dTomato or MYb115::dTomato). Subsequently, young adult worms were washed with M9T to remove all bacteria that were outside the worms, and then these worms were mounted alive in 10 mM tetramisole on agar-padded object slides using high-performance coverslips (Carl Zeiss Microscopy GmbH, Jena, Germany). All visualizations were performed with a ZEISS LSM 880 confocal laser scanning microscope system (Carl Zeiss Microscopy GmbH). A 40× C-Apochromat water immersion objective with a numerical aperture of 1.2 (Carl Zeiss Microscopy GmbH) was applied, and Immersol™ W (2010) with a refractive index of 1.334 (Carl Zeiss Microscopy GmbH) was used as the immersion medium. The dTomato fluorescence of the bacteria and the CFP fluorescence of the worm intestinal structures were sequentially excited with 561 nm and 458 nm laser light, respectively, and detected with an Airyscan detector operated in the R-S mode (sensitivity mode) without using any emission filters. The system was controlled by the software ZEISS Efficient Navigation 2 (Carl Zeiss Microscopy GmbH). The processing of the Airyscan raw data was performed with the automatic Airyscan processing function of this software.

Statistical analyses were performed with RStudio (Version 1.4.1717). The following statistical tests were used: ANOVA and GLM analysis with Tukey multiple comparison tests (Hothorn et al., 2008) and False Discovery Rate (FDR) or Bonferroni (Benjamini and Hochberg, 1995) correction were used for the survival and bacterial colonization assays respectively. The lifespan data were analyzed with the Kaplan-Meier log rank test with FDR correction for multiple testing (Kaplan and Meier, 1958). Wilcoxon-Rank sum test with FDR correction for multiple testing was used for the bacterial colonization and population growth assays. The fertility data was analyzed using Kruskal-Wallis test, with Wilcoxon pairwise comparison and FDR correction. Statistical analyses are found in the Supplementary Table . Graphs were plotted using ggplot2 (Wickham, 2016) in R (Version 4.1.0) (R Core Team, 2021) and were edited in Inkscape (Version 1.0.1). All treatments were randomized using random codes on plates to prevent experimenter bias.

As a first step to investigate the effects of the protective microbiota isolates P. lurida MYb11 and P. fluorescens MYb115 on C. elegans life-history traits, we grew worms on each of the microbiota isolates MYb11 and MYb115, or the control food bacterium E. coli OP50 and measured nematode fertility (i.e., hatched larvae). We found that both microbiota isolates affected reproductive timing and increased early reproduction up to 22% and 31% on MYb11 and MYb115 respectively, at day two of adulthood compared to worms on OP50 ( Figure 1A and Supplementary Table ). Moreover, worms on MYb11 showed a marginal, but statistically significant, increase in overall fertility.

We next asked if the protective microbiota impact C. elegans lifespan, e.g., due to a potential trade-off between the protective effect and lifespan. When combining the lifespan data from all runs, we found that the overall lifespan of worms was significantly reduced on MYb11 but not on MYb115 compared to worms on OP50 ( Figure 1B ). At the same time, it is important to note that the effect of MYb11 on lifespan was characterized by high variability, with a highly significant negative effect on lifespan in two independent experimental runs, but no significant effect in two other runs ( Supplementary Table ).

We next investigated colonization of the C. elegans intestine by MYb11 and MYb115. Using confocal laser scanning microscopy and fluorescently labeled MYb11, MYb115, and OP50 strains, we confirmed that MYb11 and MYb115 are able to colonize the intestine of young adult worms (Zimmermann et al., 2019; Dirksen et al., 2020; Pees et al., 2021a) and visualized for the first time the colonization of the worm intestine by MYb11 and MYb115 at single cell resolution ( Figures 2A–D and Figure S1 ). While we could not observe single, intact bacteria but only diffuse dTomato fluorescence in the intestines of young adult worms that had been exposed to OP50 ( Figure 2B ), nearly all MYb11 and MYb115 bacteria present in the intestines of young adult worms were intact ( Figures 2C, D , S1 ). Also, visual observations of worms exposed to MYb11 and MYb115 suggest that the intestinal lumina were distended in comparison to the narrow intestinal lumina of worms on OP50. In addition, we quantified the colonization of C. elegans by MYb11 and MYb115 by counting colony-forming units (CFUs) ( Figures 2E , S2 ). Both microbiota isolates, MYb11 and MYb115, showed increased colonization in the intestine at day 1 of adulthood when compared to the control OP50 ( Figures 2E , S2 ). Furthermore, as a first step towards monitoring microbiota colonization in ageing C. elegans, we assessed bacterial load at different days of adult lifespan and observed a gradual increase in the CFU counts as the worm ages ( Figure S2 ).

In parallel to studying C. elegans colonization by MYb11 and MYb115 and the associated effects on worm fertility and lifespan, we were also interested in further exploring the protective capacities of MYb11 and MYb115. In particular, we asked whether the previously observed protective effect of MYb11 and MYb115 against Bt247 extends to another Bt strain, MYBt18679 (Bt679) and its toxins. Bt produces crystal pore-forming toxins (Cry PFTs) that are important virulence factors involved in its pathogenesis (Masri et al., 2015), which damage the nematode via penetrating the target cell membrane introducing pores and eventually leading to cell death (Zárate-Potes et al., 2020). Different strains of Bt produce distinct Cry PFTs and vary in virulence; Bt247 produces the nematicidal toxin Cry6Ba, whereas Bt679 produces the nematicidal toxins Cry21Aa3 and Cry14Aa2 (Schulte et al., 2010; Masri et al., 2015; Hollensteiner et al., 2017; Papkou et al., 2019; Zárate-Potes et al., 2020). We assessed worm population growth (as a proxy for fitness) and worm survival (as a proxy for resistance) of Bt679 spore-infected N2 worms in the presence of MYb11 and MYb115. We found that worms infected with Bt679 showed an increase in population growth in the presence of both MYb11 and MYb115 ( Figure S3 ) and significantly increased survival ( Figure 3A ) compared to infected worms on the E. coli OP50 control. These results align with our previous results on the protective effect of these microbiota isolates against infection with Bt247 (Kissoyan et al., 2019). Thus, MYb11 and MYb115 protect not only against Bt247 but also against Bt679. Since specific microbiota isolates can confer protection against pathogens by directly interfering with the function of bacterial toxins [reviewed in Masson and Lemaitre (2017)], we investigated if MYb11 and MYb115 have a direct effect on Bt679 Cry toxins. Thus, we exposed C. elegans to a mix of purified Bt679 toxins (containing Cry21Aa3 and Cry14Aa2) in the presence or absence of the microbiota isolates MYb11 and MYb115. Intriguingly, we found that worms on MYb11 showed significantly increased susceptibility to the Bt679 toxins ( Figure 3B ), contrary to the protective effect we observed when infecting worms with Bt679 spores ( Figure 3A ). Worms on MYb115 were as resistant to Bt679 Cry toxins as worms on OP50 ( Figure 3B and Supplementary Data File ). However, as worm survival after Bt679 toxin exposure was relatively high for worms on both MYb115 and the OP50 control, it is difficult to draw any conclusion on a potential protective effect of MYb115 against toxin exposure.