Avoidance behavior independent of innate-immune signaling seen in Caenorhabditis elegans challenged with Bacillus anthracis

Developmental and Comparative Immunology 102 (2020) 103453

Contents lists available at ScienceDirect

Developmental and Comparative Immunology journal homepage: www.elsevier.com/locate/devcompimm

Avoidance behavior independent of innate-immune signaling seen in Caenorhabditis elegans challenged with Bacillus anthracis

T

Michael J. Turnera,b,∗, Justin K. Coxa, Anthony C. Spellmana, Craig Stahlb, Sina Bavarib a

School of Natural Sciences and Mathematics, Department of Science, Mount St. Mary's University, 16300 Old Emmitsburg Rd, Emmitsburg, MD, 21727, USA Molecular and Translational Sciences, United States Army Medical Research Institute of Infectious Diseases (USAMRIID), 1425 Porter Street, Fort Detrick, Frederick, MD, 21702, USA

b

ARTICLE INFO

ABSTRACT

Keywords: Caenorhabditis elegans Bacillus anthracis Avoidance Lifespan Behavioral innate immunity DAF-16

Small organisms, like the nematode C. elegans, are emerging as insightful models in which to study host/pathogen interactions and the evolving interplay between host defenses and microbial offenses. In C. elegans the innate immune response has been shown to be connected to the DAF-2 insulin/insulin-like growth factor 1 (IGF1) signal pathway, a critical transduction pathway that mediates stress response in the worms via the DAF-16 FOXO/forkhead transcription factor. Our studies of the C. elegans’ phenotypes that are associated with behavioral innate immune response (avoidance behavior) and IGF-1 signaling perturbations (lifespan effects) led us to question the cause of the avoidance behavior observed when C. elegans are challenged with B. anthracis. While worms indeed avoid B. anthracis, and this behavior seems to be partly tied to IGF-1 signaling, the bacteria have neither nematocidal nor visible pathogenic effects on the worms. In fact, worms fed B. anthracis alone exhibit extended lifespans. We demonstrate that the extended lifespan phenotype seen in worms fed B. anthracis is likely the result of calorie restriction, and that worms do not eat B. anthracis even when avoidance behaviors have been suppressed. We further demonstrate a large time lag between the onset of avoidance behavior (which occurs upon contact with B. anthracis), and the induction of IGF-1 signaling (which occurs much later) in worms fed B. anthracis. Taken together, our data demonstrate behavioral avoidance that does not appear to be linked to a measurable immune response. We propose that, in some situations, avoidance behaviors categorized as immunological might be more accurately described as broad foraging behaviors induced in worms presented with a non-preferred food choice, or with a food choice that is either difficult or impossible for the worms to ingest.

1. Introduction Lacking an active circulatory system, as well as migratory, antibodyproducing, or phagocytosing cells, the nematode C. elegans is limited to mechanisms of innate immunity to protect itself against would-be pathogens. These mechanisms include physical protection (Taffoni and Pujol, 2015), inducible molecular protection (Dierking et al., 2016; Ermolaeva and Schumacher, 2014), and behavioral protection (Engelmann and Pujol, 2010; Ermolaeva and Schumacher, 2014; Fuchs and Mylonakis, 2006). The nematode's physical protection is provided by a robust outer cuticle that surrounds the worm and creates a barrier between the dangers of the external environment and the inner anatomy of the organism (Taffoni and Pujol, 2015). The worms also possess a grinder, a chewing organ, that functions to grind bacterial cells that are ingested via the mouth before they are passed into the

lumen of the intestine (Avery and Horvitz, 1989). This grinder ensures that the worms can metabolize the nutrients available from the bacteria while minimizing the risk of becoming infected after ingestion. Inducible molecular protection comes when the worms' pathogen-sensing signal pathways upregulate antimicrobial products such as specialized peptides, caenopores, lysozymes, lectins and reactive oxygen species (Dierking et al., 2016; Engelmann and Pujol, 2010). Behavioral protection is manifest in the worms' ability to sense and avoid harmful bacteria, bacterial toxins and other malignant agents by moving away from them (Meisel and Kim, 2014; Pradel et al., 2007; Pujol et al., 2001; Zhang et al., 2005). In C. elegans there are six known signal pathways tied to innateimmune signaling, plus one toll-like receptor whose signal transduction pathway is yet unknown (Gravato-Nobre and Hodgkin, 2005). These six pathways include the TGF-beta-like, IGF-1 insulin-like, programmed

Corresponding author. Mount St. Mary's University, Emmitsburg, MD, USA. Tel.: +1 301 447 5446. E-mail addresses: [email protected] (M.J. Turner), [email protected] (J.K. Cox), [email protected] (A.C. Spellman), [email protected] (C. Stahl), [email protected] (S. Bavari). ∗

https://doi.org/10.1016/j.dci.2019.103453 Received 17 September 2018; Received in revised form 16 July 2019; Accepted 16 July 2019 Available online 18 July 2019 0145-305X/ © 2019 Elsevier Ltd. All rights reserved.

Developmental and Comparative Immunology 102 (2020) 103453

M.J. Turner, et al.

cell death, p38 MAPK, JNK-like and ERK MAP kinase signal pathways. However, of these six pathways, only one has been shown to contribute to all aforementioned modes of innate-immune defense in C. elegans — the IGF-1 pathway; these connections are discussed here briefly. First, molecular innate immunity. The IGF-1 pathway has long been studied as a regulator of worm longevity and stress response (Cypser and Johnson, 2002; Johnson et al., 2001, 2002; Kimura et al., 1997; Ogg et al., 1997), and more recently as part of the worms' innate-immune response (Evans et al., 2008a, 2008b; Kawli and Tan, 2008). The connection between longevity and the immune response has been shown to be correlated with DAF-16 activity; the DAF-16 FOXO/forkhead transcription factor is responsible for the upregulation of key stress response genes and, when activated, confers resistance to stress that results in extended lifespan (Evans et al., 2008b; Kim, 2013; Kimura et al., 1997; Ogg et al., 1997; Wang et al., 2012). The same DAF-16-centric response has been shown to confer resistance to harmful pathogens (Evans et al., 2008b; Gaglia et al., 2012; JebaMercy et al., 2013; Wang et al., 2012). Signaling through the IGF-1 pathway turns off this stress response by phosphorylating DAF-16, which results in DAF-16 retention in the cytoplasm of target cells and thus prevents DAF-16-mediated transcription of stress response genes (Ogg et al., 1997). These studies reveal a strong connection between IGF-1 signaling and the molecular component of the worms’ innate-immune defenses. Second, physical innate immunity. IGF-1 signaling contributes to developmental timing via interactions with the heterochronic genes that regulate larval development (Huang and Zhang, 2011). The connection between this observation and IGF-1 signaling in innate immunity was best established when Ewald et al. (2016) demonstrated that the proper development and maintenance of the hypodermis was dependent on DAF-16. Thus the hypodermis of the worm, which is responsible for secretions that form the worm cuticle, is maintained and preserved, in part, by IGF-1 signaling. These studies demonstrate that the IGF-1 signal pathway is crucial to the formation of the major physical barrier of the worms’ innate immune system – the outer cuticle. Third, behavioral innate immunity. The IGF-1 pathway has also been implicated in the behavioral defenses of C. elegans' innate immunity, namely avoidance behavior (Schulenburg and Ewbank, 2007). Worms with disrupted IGF-1 signaling demonstrate altered ability to avoid both contact with, and ingestion of, Bacillus thuringiensis (Hasshoff et al., 2007). Furthermore, mutants of NPR-1, an effector of IGF-1 signaling in innate immunity, manifest different defensive phenotypes toward Gram-positive and Gram-negative bacteria, and these phenotypes correlate with upregulated IGF-1 signaling (Nakad et al., 2016). Clearly, the IGF-1 signal pathway is a critical component of C. elegans’ innate immune response to would-be pathogens, affecting molecular, physical, and behavioral countermeasures against infection. Given that the primary nutrition source for C. elegans is bacteria encountered in the environment, the worms are very familiar with – and strangely dependent on – potential pathogens. The very food that the worms ingest can have the dual role of providing nutrition but also infecting the host animal. Thus, worms are under strong selective pressure to develop and maintain the ability to discern between nutritious and harmful bacteria. Using C. elegans as a model to study host/pathogen interactions, we observed that worms actively avoid the bacterial species Bacillus anthracis (B. anthracis). Early experiments to ascertain genes that might contribute to this avoidance behavior heavily implicated the IGF-1 signal pathway. These experiments were designed under the assumption that the observed avoidance behavior was tied to an innate-immune response in the worms induced by contact with B. anthracis. However, aside from avoidance behavior, no additional signs of pathogenicity were seen in worms fed B. anthracis; there was no colonization of bacteria in the intestine of the worm, and no negative effect on worm lifespan. In fact, contact with B. anthracis had the surprising effect of extending worm lifespan via a mechanism independent of IGF-1

signaling. The data presented here represent our efforts to understand the cause of the observed avoidance behavior elicited in C. elegans upon contact with B. anthracis. They also suggest causes of worm avoidance independent of innate-immune signaling. Importantly, our findings demonstrate that avoidance behaviors previously attributed to innate immunity, in some circumstances, may be due to other factors. 2. Materials and Methods 2.1. Bacterial and worm strains All worm lines were obtained from the C. elegans Genetics Center (CGC): N2, spe-9 (hc88), DAF-16:GFP (muls61), daf-4 (e1363), sma-3 (e491), dbl-1 (nk3), sma-6 (wk7), sma-2 (e502), age-1 (hx546), daf-2 (e1370), daf-16 (mu86), pdk−1 (sa680), ced-4 (n2273), ced-3 (n1286), ced-9 (n1950), tol-1 (nr2033), sek-1 (ag1), tir-1 (qd4), rab-1 (ok3750), nsy-1 (ag3), pmk-1 (km25), mek-1 (ks54), kgb-1 (um33), mpk-1 (oz140), lin-45 (dx84), mek-2 (n1989), and lin-4 (e912). Worms were maintained on Nematode Growth Medium (NGM) plates using E. coli strain OP50 as a food source. Worms were kept at 15–20 °C, except where otherwise noted. Bacterial lines included Bacillus anthracis (Sterne), Francisella tularensis (LVS), Bacillus thailandensis (E264), Pseudomonas aeruginosa (PA14), Yersinia pestis (PGM-PST), and Escherichia coli (OP50). E. coli, P. aeruginosa, and B. thailandensis were streaked on LB agar plates and overnight cultures were grown in LB medium. B. anthracis was streaked on sheep blood agar plates, and overnight cultures were grown in Tryptic Soy broth. Y. pestis was streaked onto chocolate agar plates and overnight cultures were grown in Meuller-Hinton broth. Fluorescently tagged strains were maintained under necessary antibiotic selection consistent with the plasmids used to transform the bacterial strains: E. coli (OP50)-RFP, E. coli (OP50)-GFP, and B. anthracis (Sterne)-GFP, were cultured in tetracycline (25 μg/ml), kanamycin (50 μg/ml), and ampicillin (100 μg/ml), respectively. The bacterial strains mentioned here are Biosafety Level 2 (BSL-2) pathogens, and all work herein described was performed in BSL-2-rated laboratories. 2.2. Lifespan assays Lifespan assays were performed as previously reported (Kenyon et al., 1993). In summary, synchronized late-L4 spe-9 worms were plated onto NGM plates seeded with the indicated bacterium and incubated at 20 °C. Bacterial species included E. coli, B. anthracis, P. aeruginosa, B. thailandensis, or F. tularensis. After being pipetted onto the NGM plates, bacteria were either allowed to dry without further manipulation (“pooled,” meaning that bacteria did not completely cover the surface area of the plate), or they were spread evenly across the plate and then allowed to dry (“spread”). Each day thereafter, plates were observed for dead worms and the number of dead worms removed from each plate was recorded. Once the plates were empty of worms, the total number of worms taken from the plate was used to calculate the percent survival over time. Treatments were performed in triplicate, and the survival percentage values were averaged and graphed. Plates whereon the bacteria were “pooled” (allowed to dry in a small pool in the middle of the plates) were used to simulate an environment in which worms could actively avoid a harmful substance. Plates whereon bacteria were “spread” (bacteria were spread with sterilized glass rods across the full surface of the plates) were used to simulate an environment in which avoidance behavior was restricted. Statistical analysis was performed for each data set by analyzing the survival curve replicate average from each strain on the described conditions (Fig. 2B-G and Fig. 5A-F) by ANOVA. Time points shown by ANOVA to demonstrate significant differences between conditions were subjected to both Logrank and Chi-square tests to produce a “P” value for the comparison of two conditions (as illustrated in the figure captions). 2

Developmental and Comparative Immunology 102 (2020) 103453

M.J. Turner, et al.

2.3. Binary choice assays

3. Results

Binary choice plates were prepared by placing two 50ul drops (one of each indicated food source) of food on opposing sides of NGM plates. ~75 synchronized worms in M9 medium were pipetted onto binary choice plates and placed in a position equidistant from both food spots. Worms were then incubated at 20 °C for 24 h. After incubation, the number of worms on each food spot was recorded, as was the total number of worms on each plate. Percentages were calculated from these totals. Assays were performed in triplicate, and the degree of avoidance was averaged between replicate samples.

While attempting to establish C. elegans as a model organism in which to study Biological Select Agents and Toxins (BSATs), we noticed that worms on plates of B. anthracis actively avoided segments of the plate covered with bacteria. In order to quantify avoidance behavior we performed binary-choice assays with wild-type (N2) worms, as described by Zhang et al. (2005), and found that N2 worms indeed avoid B. anthracis (Fig. 1a). We hypothesized that the worms were able to sense a substrate secreted by the bacteria that repelled the worms. To test this hypothesis, we repeated the binary choice assay using bacteria that had been killed by incubation at 70 °C for 1 h. Our findings suggest that worms actively avoid the bacteria whether they're alive or dead (Fig. 1a). We further tested for the presence of secreted molecules that might trigger avoidance behavior in worms by resuspending pellets of E. coli in liquid media used to grow overnight cultures of B. anthracis. Worms failed to avoid the E. coli resuspended in B. anthracis medium (Fig. 1a), suggesting that the avoidance behavior manifested by the worms is not induced by a molecular substrate secreted by B. anthracis. Given these observations, we concluded that the avoidance of B. anthracis by the worms must be due to a characteristic of the outer surface of the bacterial cells. We reasoned that if the worms were avoiding B. anthracis in accordance with well-documented innate-immune behavioral activity (Engelmann and Pujol, 2010; Ermolaeva and Schumacher, 2014; Hasshoff et al., 2007; Kim, 2013; Kurz and Tan, 2004; Schulenburg and Ewbank, 2007; Zhang et al., 2005) mutants of the innate-immunity signal pathways would be compromised in their avoidance capabilities. To test this theory, we performed binary choice assays using loss-offunction mutants for 25 genes representating each of the six aforementioned signal pathways known to be associated with innate-immune responses in C. elegans (Table 1 and Fig. 1b). Of the 25 strains assayed, 9 strains demonstrated significantly compromised avoidance behavior (Fig. 1b). These strains represent mutations coming from 4/6 of the innate-immune signal pathways (TGF-beta-like, IGF-1 insulinlike, JNK-like, and ERK MAP kinase) plus the Toll receptor (Fig. 1b). The most heavily implicated of these pathways was the IGF-1 pathway (Fig. 1b and Table 1). In this pathway, loss-of-function mutations to 3/4 protein members tested (DAF-2, DAF-16, and AGE-1) resulted in significantly compromised avoidance behavior in the worms (Fig. 1b). The fourth member, PDK-1, showed the largest average decrease in choice index of any strain tested (Fig. 1b). However, in our hands we saw great variability among plates of this strain, resulting in a less-than-significant “p” value (p = 0.079) (Fig. 1b). Taken together, these data suggest that the avoidance of B. anthracis by the worms may be a behavioral response tied to innate-immune signaling, and that the IGF-1 signal pathway is a major contributing factor in that response. Previous work with C. elegans and B. anthracis has demonstrated that B. anthracis alone is not infectious to the worms; lethal infection by B. anthracis in C. elegans requires the pore-forming protein Cry5B (Kho et al., 2011), a protein common to other Bacillus strains, but not to B. anthracis. In order to evaluate the pathogenesis of B. anthracis in C. elegans, we performed a lifespan assay comparison between worms on four different potential worm pathogens (Fig. 2a). P. aeruginosa has been long established as a worm pathogen that infects and kills the worms (Mahajan-Miklos et al., 1999; Tan et al., 1999a, 1999b). Y. pestis infects the worms by creating a biofilm that covers the worm cuticle, obstructing the mouth of the worms (Darby et al., 2002). Both F. tularensis and B. anthracis have been shown to be conditionally infectious to the worms; the condition of B. anthracis-mediated infection in worms was outlined above. In the case of F. tularensis the worms must be compromised in MAPK signaling in order for the bacteria to become infectious (Jayamani et al., 2017). Upon testing these four pathogens for their nematocidal effects, we found that B. anthracis produced a profound extension of lifespan in the worms; other species had either minimal extensions (F. tularensis), no effect (Y. pestis) or attenuated

2.4. Food choice assays Overnight cultures of E. coli (OP50) and B. anthracis (Sterne) were grown in LB and TSB, respectively. Cell density was normalized between the two species using OD600 values, and then a 1:1 mixture of E. coli/B. anthracis was prepared from which 200uL aliquots were spread onto NGM plates. Once the plates were dry, synchronized, early adult spe-9 (hc88) worms were placed on to the mixture and allowed to feed. Bacteria from the plates were harvested at specific time points (as indicated in Fig. 4 caption), after worms had been removed by a pick, by washing the remaining bacteria off the NGM plates with water, and then using a gDNA isolation kit (Qiagen), according to manufacturer protocol. In each experiment, plates without worms added to them served as a control. 2.5. Semi-quantitative PCR Reactions were set up using Platinum Taq Polymerase (Invitrogen), according to manufacturer's protocol. Species-specific primers targeting the DNA Pol III gene were included in each PCR reaction. In order to provide semi-quantitative measurements, PCR tubes were removed at every 5 cycles of the reaction program, and samples of each reaction at each cycle-point were used to establish and compare amplicon abundance in the linear and exponential phases of the amplification curve. We found that cycling from 15 to 20 rounds produced repeatable data trends that demonstrated quantitative differences between samples. For work reported here, all samples were analyzed after 20 rounds of PCR. The quantification of PCR amplicons was performed by running agarose gels and measuring the band intensity with Image Lab® computer software. Uniform pixel boxes were drawn around each band, and pixel counts were used to quantitate band intensity. The resulting data were used to compare and imply bacterial abundance on food-choice plates. 2.6. Microscopy and DAF-16:GFP timeline Overnight cultures of OP50 and B. anthracis were plated onto 6 cm NGM plates in 200uL aliquots. Bacteria were spread evenly over plates and then allowed to dry at room temperature for 24–36 h. Mixed-stage DAF-16:GFP worms stored at 15 °C were washed off of NGM plates seeded with OP50 and centrifuged for 1 min at 1200×g. The supernatant was removed and worms were washed twice with 10 mL M9 medium. Washed worms were centrifuged and resuspended in 200uL M9, and then aliquoted onto NGM plates covered in either OP50 or B. anthracis and stored at 15 °C. Plates were taken from the 15 °C incubator on an hourly basis, worms of various stages were picked from the plates, placed on a 2% agarose stage in 5 mM levamisole, and then visualized under a high-power microscope. Worms placed on fluorescently labelled bacteria were allowed to feed for 12–16 h and then imaged as described above. Images captured include Bright Field (BF), GFP and RFP; Merge images combine these three. All merge images were comprised of GFP and RFP images that had received identical image-enhancing adjustments (brightness, contrast, etc). 3

Developmental and Comparative Immunology 102 (2020) 103453

M.J. Turner, et al.

Fig. 1. Avoidance of B. anthracis by C. elegans is induced by a cell-surface entity and is compromised in mutants of known innate-immune signal pathways. Synchronized worms were placed equidistant to two food choices: regular food and a potential “pathogen.” After 24 h, the fraction of worms found on each food out of the total number of worms was recorded. The fraction of worms on the regular food was multiplied by (+1), while the fraction of worms on the “pathogen” was multiplied by (−1). The sum of these two numbers was graphed; a score of “1” indicates that 100% of the worms were on the regular food (they absolutely avoided the “pathogen”), a score of “-1” indicates that 100% of the worms were on the “pathogen,” and a score of “0” indicates that the number of worms on the food and the “pathogen” were equal. Each test was done in triplicate, and the error bars represent 99% confidence intervals. (A) N2 worms were presented different “pathogens” to test for avoidance; the “pathogens” are listed across the top of the graph. For each data point, the food choice was between the listed pathogen and E. coli, the worms' normal food source. (B) Mutants of known innate-immune signal transduction pathways were assayed for avoidance behavior upon contact with B. anthracis. Avoidance behavior exhibited for each strain is documented and compared to relative controls: no food (M9—worms avoid this because there is nothing there to eat), and normal food (E. coli—worms choose this food equally with the E. coli on the opposite side of the plate). P-values were obtained by comparing transformed (the arcsine of the square root) proportions of worms on pathogen (the given strain vs N2) using a single-tailed T-test: (***) p 0.0051; (**) p 0.0164; (*) p 0.0388 (*); ( ) p = 0.079. 4

Developmental and Comparative Immunology 102 (2020) 103453

M.J. Turner, et al.

independent of the IGF-1 signal pathway. To further explore a possible role for IGF-1 signaling in response to contact with B. anthracis, DAF-16:GFP worms were placed onto B. anthracis and the localization of DAF-16 was monitored hourly using fluorescence microscopy. We hypothesized that if avoidance behavior was due to IGF-1 signaling and the innate-immune response, DAF-16 localization to the nucleus and subsequent activation of the innate immune and/or stress response via IGF-1 signaling would occur very quickly. This hypothesis was based on the observation that avoidance behavior by the worms, upon contact with B. anthracis, is nearly immediate. As expected, the first cells demonstrating visible translocation of DAF-16 to the nucleus were intestinal cells. However, nuclear localization of DAF-16 was not visible until after 2 h of exposure to B. anthracis (Fig. 3). The time lag between the onset of avoidance behavior (immediate upon exposure) and the activity of the IGF-1 signal pathway via DAF-16 (> 2hrs after exposure) suggests that the avoidance behavior seen is not completely dependent on IGF-1 signaling, or perhaps on any signaling of the worm innate immune system. Our data to this point suggested that the worms actively avoid B. anthracis, and that this avoidance behavior is partially linked to the function of some proteins of key innate-immune signal pathways in the worm. However, they also suggested that B. anthracis is not infectious and has no harmful effect on the worms. Furthermore, the data demonstrated that any traceable innate-immune-related signaling in the worms caused by exposure to B. anthracis occurs long after avoidance behavior ensues. The growing evidence that avoidance, in the case of C. elegans and B. anthracis, may not be the result of an innate-immune response led us to consider an alternative hypothesis: perhaps the worms simply don't recognize B. anthracis as a food source, or at least they prefer other options when available. To test this theory, worms were placed on 1:1 mixtures of E. coli and B. anthracis and allowed to feed. At various time points worms were removed and the abundance of each remaining bacterial species was quantified using PCR and species-specific primers against the DNA Polymerase III gene. We found generally, that the ratio of E. coli to B. anthracis went down over time on plates where worms were allowed to feed, compared with a no-worm control plate (Fig. 4a and b). These data suggest that the worms consume E. coli at a faster rate than they consume B. anthracis. We further tested the worms' ability to choose one bacteria over another by mixing bacterial strains with alternative fluorescent markers (GFP or RFP), and then used fluorescence microscopy to identify the bacteria found in the worms' pharynxes. As expected, worms did not distinguish at all between RFP- or GFP-labelled E. coli strains; every worm that had GFP fluorescence in the pharynx also had RFP fluorescence in the pharynx (Fig. 4c and f). However, worms fed with a mixture of GFP-labelled B. anthracis and RFP-labelled E. coli showed only red E. coli in their pharynx (Fig. 4c and g). These data support the hypothesis that C. elegans don't recognize B. anthracis as a food source and do not ingest B. anthracis. Given these observations, we returned to some of our previous experimental procedures and added some additional controls. First, we reasoned that if worms simply don't recognize B. anthracis as a preferred food source, then mixing B. anthracis together with E. coli should suppress avoidance behavior in the binary choice assay. Consistent with this logic, worms choose the mixed food source as often as they choose pure E. coli (Fig. 1a). Second, worms that experience stress in the form of calorie restriction have been shown to manifest extended lifespans (Kaeberlein et al., 2006). We hypothesized that if worms were failing to recognize B. anthracis as a food source, the extended lifespan of the worms when placed on B. anthracis may simply be accounted for by the fact that the worms weren't eating. Consistent with this hypothesis, the lifespans of worms given no food (“no food” controls) phenocopy the lifespans of worms placed on B. anthracis alone. (Fig. 2b–g). We also reasoned that mixing B. anthracis with E. coli would suppress the lifespan effects seen when worms are placed on B. anthracis alone. Indeed, all lifespan effects seen from any worm strain tested by being placed on

Table 1 Mutants of Signal Transduction Pathways Implicated in Innate Immunity. Strains of each of the mutants listed below (not an exhaustive list of all protein participants in represented signal transduction pathways) were tested for their ability to avoid B. anthracis (Fig. 1b). Strains are organized here, as they are in Fig. 1b, into the signal pathways to which they contribute. Like-colored genes belong to the same signal pathway, the identity of which is indicated by the key.

lifespans (P. aeruginosa) in comparison to worms on the regular food source (Fig. 2a). These data suggest that B. anthracis alone is not harmful to the worms, consistent with previous reports. While B. anthracis did not demonstrate the ability to infect the worms, the extended lifespan manifested by worms exposed to the bacteria provided a phenotypic link to the IGF-1 signal pathway; mutations that lead to constitutive expression of DAF-16 drastically extend lifespan (Dorman et al., 1995; Kenyon et al., 1993). The connection between the IGF-1 genes, avoidance, and lifespan caused us to wonder if other IGF-1-related phenotypes might be altered by exposure to B. anthracis. To address this question, IGF-1 pathway mutations were assayed for lifespan in the presence of regular food or B. anthracis. We found that some of the IGF-1 pathway mutants (daf-2 and daf-16) exhibited a modestly-extended lifespan upon challenge to B. anthracis (Fig. 2b–e). Other mutants of the IGF-1 pathway were not affected by B. anthracis challenge (pkd-1 and age-1) (Fig. 2f and g). These data suggest that the lifespan effect seen in C. elegans upon exposure to B. anthracis is 5

Developmental and Comparative Immunology 102 (2020) 103453

M.J. Turner, et al.

B. anthracis alone – including the mutants of the IGF-1 signal pathway – were suppressed when the worms were placed on the mixed food source (Fig. 2b–g). Taken together, these data strongly support the hypothesis that worms either do not recognize B. anthracis as a food source, or that they will avoid eating it for unknown reasons. We noticed in our experiments with fluorescent bacterial strains that L4 and adult worms placed on B. anthracis alone showed, on rare occasion, ingestion of B. anthracis cells (Fig. 4c and h). We wondered if, under these circumstances, worms may be willing to uptake an otherwise non-ideal food source. Furthermore, we felt this possibility might provide an explanation for the nuclear localization of DAF-16 that is seen only when the worms are on B. anthracis alone (Fig. 3). To test this theory, we placed worms onto plates whereon B. anthracis had been spread completely across the plates, a situation where they could not physically avoid contact with the bacteria. We wondered if worms lacking options would ultimately ingest B. anthracis and, as a result, demonstrate suppressed lifespan effects that would otherwise result from caloric restriction. Indeed, all strains but one, when placed on a ubiquitous pad of B. anthracis alone, showed lifespan ranges more consistent with that of the same worm strain on E. coli (Fig. 5a–f). By restricting space to avoid the bacteria and eliminating alternative food options, we were able to suppress the moderate lifespan extensions seen when plating worms on B. anthracis that were pooled in the middle of the plate (and therefore avoidable) (Fig. 5a–f). Interestingly, DAF-16 mutants maintained their extended lifespan in either situation (pooled or spread B. anthracis), suggesting that the suppression of lifespan effects seen in worms that do ingest B. anthracis is DAF-16-dependent (Fig. 5d).

avoid it? C. elegans' avoidance behavior triggered via an innate-immune response induced by contact with a bacterial pathogen has been shown to correlated with pathogen virulence – and this correlation is independent of cell-wall composition; it has been shown in experiments using Gram-negative E. coli (Anyanful et al., 2009), as well as in experiments using Gram-positive B. thuringiensis (Nakad et al., 2016). Though the worms clearly avoid B. anthracis, the bacteria are not harmful to the worms, suggesting that the avoidance behavior seen our studies may not be tied to innate-immune signaling and pathogen-induced behavioral avoidance. B. anthracis differs significantly from E. coli, the worms' traditional laboratory food source. There are two physical characteristics that might affect the worms' preference for one over the other, or the worms’ ability to detect and/or ingest one or the other. First, the make-up of the cell wall itself. Of all the bacterial strains tested in our hands, B. anthracis was the only Gram-positive bacterial species; all others (including known pathogens used for controls, and E. coli) were Gram-negative (Fig. 2a). Our data suggest that a cell-wall component of the B. anthracis is a key factor affecting avoidance behavior by the worms in our experiments. Heat-killed B. anthracis cells elicited the same avoidance response from worms that live cells did (Fig. 1a), and neither pads of E. coli mixed in media used to grow up B. anthracis, nor pads of mixed B. anthracis and E. coli induced avoidance behavior from the worms (Fig. 1a). It is possible that worms avoid B. anthracis simply because they either do not like, or do not recognize, the bacteria as a food source – and this discernment is linked to an unknown entity on the surface of the B. anthracis cells. The second physical characteristic worth considering here is the size of the bacterial cells. B. anthracis cells measure in the range of 1.0–1.2 μm wide and 3–5 μm long (Spencer, 2003), while E. coli measure an average of 0.5 μm wide and 2 μm long (Koppes et al., 1978). Recent work has demonstrated that worms actively ingest, during foraging, particles from their environment that are 0.5μm–3μm in size, and rarely ingest particles outside of this range (Kiyama et al., 2012). E. coli fall within that ingestible range, and B. anthracis fall outside of it, and our data correlate with these studies. Looking at bacterial abundance over time, on plates where worms were fed mixed E. coli and B. anthracis, E. coli disappeared faster from the plates than B. anthracis (Fig. 4a and b). Furthermore, images of worms fed with fluorescently labelled bacterial strains showed that worms actively ingested E. coli but only rarely ingested B. anthracis (Fig. 4c and d). In the rare occasion that B. anthracis cells were found within worms, those worms were exclusively large; no worms younger than L4 were ever found to have ingested B. anthracis under any circumstance (Fig. 4c and d). Taken together, these data support the idea that the size of the bacterial cell in relation to the size of the worm pharynx may indeed be a determining factor in the worms’ choice (or ability, for smaller worms) to ingest the bacteria. Thus, the avoidance seen in our early studies may simply be the preference of worms for E. coli over B. anthracis based on cell type, or the inability or difficulty of worms to eat B. anthracis based on cell size. Kiyama et al. reported that the uptake of particles outside of the size range commonly ingested by C. elegans was rare in the absence of regular food, and almost completely eliminated in the presence of regular food (Kiyama et al., 2012). Our data exactly correlate with this report. Worms (L4 or adult worms only) on only B. anthracis ingest B. anthracis rarely (~20% of L4s/adults assayed, Fig. 4c). However, when E. coli is mixed with the B. anthracis the fraction of worms (L4 or adult only) that ingest B. anthracis falls to 3% (Fig. 4c). Our data expand the studies of Kiyama et al., which used non-biological particles to demonstrate size preference, to include examples of living cells that appear to be “ignored” by the worms actively looking for food. The correlation of these two studies strongly suggests that C. elegans don't recognize B. anthracis as a food source and that the uptake of B. anthracis in the absence of other food is more a result of the worms' enhanced foraging in the absence of preferred food. To further support this hypothesis, worms given no food phenocopy

4. Discussion Our original assumptions about the observed avoidance behavior of worms placed on B. anthracis were centered on the idea that avoidance of a potentially pathogenic bacterium must be tied to the immune response. In fact, our early experimental data — specifically the compromised avoidance behavior seen in mutants of the known innateimmune signal pathways — seemed to support these assumptions (Fig. 1). The strong implication of the IGF-1 signal pathway was particularly suggestive, given the known role of this pathway in behavioral avoidance and innate-immune signaling (Fig. 1b). Further evidence of IGF-1 signaling in response to B. anthracis exposure was demonstrated using DAF-16:GFP transgenic worms, which showed nuclear localization of DAF-16 after 2–3 h exposure to the bacteria (Fig. 3). Taken alone, these data suggest that IGF-1 signaling is triggered in C. elegans by exposure to B. anthracis, and C. elegans’ avoidance of B. anthracis is partially dependent on IGF-1 signaling. These conclusions strongly support previous work wherein the link between IGF-1 signaling and the avoidance of Gram-positive, Bacillus-genus bacteria by C. elegans has been described (Hasshoff et al., 2007; Nakad et al., 2016). However, the work that established this relationship was performed using Bacillus thuringiensis, a known C. elegans pathogen with nematocidal effects. While our early experimentation with B. anthracis seemed to show similar patterns of pathogen avoidance as work done with B. thuringiensis, B. anthracis did not show any nematocidal effects (Fig. 2). As mentioned previously, B. anthracis is only a conditional pathogen to C. elegans and requires certain virulence factors (the yceGH tellurite resistance genes (Franks et al., 2014) and the pore-forming Cry5b (Kho et al., 2011)) to confer lethality. All reported virulence factors in the lethal infection of C. elegans by B. anthracis are dependent on exogenous Cry5b (Franks et al., 2014; Kho et al., 2011). Our experiments were performed in the absence of exogenous Cry5b and, consistent with the work just cited, B. anthracis did not appear to have any negative effects on the worms. Rather, contact with B. anthracis resulted in extended lifespan, a phenotype that was independent of IGF-1 signaling (Figs. 2 and 5). If B. anthracis is not harmful to the worms, why do they appear to 6

Developmental and Comparative Immunology 102 (2020) 103453

M.J. Turner, et al.

Fig. 2. C. elegans exposed to B. anthracis experience moderate lifespan extension that phenocopies caloric restriction and is independent of the IGF-1 signal pathway. Lifespan assays were performed with young adult worms and carried out at 20 °C or 25 °C. All strains except spe-9 were plated onto FUDRcontaining plates to maintain sterility throughout the assay. Error bars represent 95% confidence intervals. Food on these plates was not spread out and was left “pooled” on the center of the plate and allowed to dry. (A) Worms (spe-9) were plated on to one of 5 possible food sources (E. coli, P. aruginosa, Y. pestis, B. anthracis, or F. tularensis), and assayed for lifespan. The percentage of worms alive of the total number plated at day “0” is graphed over time (days). (B-G) Mutants of the insulin-like signal pathway were plated onto NGM plates containing FUDR and either E. coli or B. anthracis. Lifespan was measured as described above. “P” values for represented curve comparisons are as follows: A: No further statistical analysis performed. B: 0.049, < 0.00001, < 0.00001, 0.389. C: 0.039, < 0.00001, < 0.00001, < 0.00001. D: 0.136, 0.00011, 0.00095, 0.607. E: 0.258, < 0.00001, < 0.00001, 0.31. F: 0.00114, < 0.00001, 0.000021, < 0.00001. G: 0.000038, 0.000011, < 0.00001, 0.00228. 7

Developmental and Comparative Immunology 102 (2020) 103453

M.J. Turner, et al.

Fig. 3. Nuclear translocation of DAF-16 in the intestinal cells of worms exposed to B. anthracis. DAF-16:GFP worms were plated onto one of three food sources: E. coli, B. anthracis, or a 50/50 mixture of E. coli and B. anthracis; all food sources were spread evenly across the plates. Worms on no food serve as a control. At hourly time points, worms were removed from their plates and viewed under a fluorescence microscope to ascertain the cellular localization (cytoplasmic or nuclear) of the DAF-16. The scope of the field is indicated in the top right corner of each image and ranges from 100 to 250μM.

worms on B. anthracis in lifespan assays (Fig. 2), exactly as one would expect if the worms do not ingest (or only very rarely ingest) the B. anthracis. Mixing B. anthracis with E. coli completely suppresses the extended lifespan phenotype because the worms have regular food to eat and therefor cease to express a phenotype consistent with calorie restriction. In situations where worms are given no alternative space or bacteria to forage (plates with B. anthracis spread completely across the agar), active foraging should result in an increased amount of B. anthracis ingested, and subsequent suppression of the life-extension phenotype associated with calorie restriction. This is exactly what our data show; when B. anthracis is spread to completely cove the plate, worms are forced to forage among the bacteria only, the ingestion of B. anthracis increases, and worms show lifespans more consistent with worms on regular food, or worms receiving sufficient nourishment (Fig. 5). Interestingly, under these conditions the suppression of longevity appears to be DAF-16-dependent. We have yet to follow-up on this observation, but one hypothesis we have is that the ability to fully

digest and benefit from the nutritional value of B. anthracis is DAF-16dependent. This hypothesis provides a possible explanation for why worms placed on B. anthracis alone show nuclear translocation of DAF16 in intestinal cells (a phenotype not seen in calorie-restricted worms) after a few hours (Fig. 3). Kumar et al. recently published their findings demonstrating a connection between eating, innate-immune signaling, avoidance and lifespan (Kumar et al., 2019). They report that worms with defective pharynx structures accumulate bacteria in the lumen of the worm intestine, leading to innate-immune signaling and avoidance and, ultimately, the lengthening of life due to calorie restriction (Kumar et al., 2019). However, in our experiments – though we did see IGF-1 signaling in the intestinal cells of worms placed on B. anthracis alone (Fig. 3) – there was no accumulation of bacteria in the worm intestine (Fig. 4d and data not shown), perhaps due to fully functional pharynxes. Furthermore, worms seemed to “avoid” B. anthracis on contact, not after hours of ingestion. Therefore, we do not believe that our data 8

Developmental and Comparative Immunology 102 (2020) 103453

M.J. Turner, et al.

demonstrate the same phenomenon reported by Kumar et al. While we began this work believing that the avoidance we were seeing was part of an innate-immune response, our data are much better explained by the hypothesis that the avoidance behavior seen in our experiments was a manifestation of the worms ignoring B. anthracis as a viable food option. One additional application of this work is worth mentioning. Schelkle et al. recently reported the use of C. elegans as a bioremediation means to decontaminate soil laden with B. anthracis spores (Schelkle et al., 2017). From studies using a GFP-labelled strain of B. anthracis (Sterne), they showed fluorescence images of worms with GFP filling the intestinal lumen of the worms, and bright-field images of spores in the digestive organs of the worms (Schelkle et al., 2017). Their conclusions from these data were that worms actively ingest B. anthracis (Schelkle et al., 2017). We saw no such fluorescence in the intestines of worms fed GFP-labelled B. anthracis. One possible explanation for the discrepancies in these observations is that the studies performed by Schelkle et al. used food mixtures comprised of both vegetative bacteria as well as spores (Schelkle et al., 2017); our studies were performed with vegetative bacteria only. Microscopy performed by Schelkle et al. confirmed that vegetative bacteria ingested by the worms were fully digested (Schelkle et al., 2017), consistent with our observations. However, B. anthracis spores were shown to survive ingestion by the worms (Schelkle et al., 2017). It is possible that the GFP seen in the studies by Schelkle et at. was the result of GFP-positive spores

accumulating in the lumen of the intestine. This explanation is supported by the fact that the size of the B. anthracis spores falls neatly within the easily-digestible size range of food mentioned previously (Carrera et al., 2007; Kiyama et al., 2012). The lack of spores in our experiments would explain our failure to see GFP in the intestine of worms fed B. anthracis-GFP, and may also explain the other phenotypes that Schelkle et al. attributed to ingestion of B. anthracis, but that we did not see (namely lethargy and bloating) (Schelkle et al., 2017). Consistent with our data, their worms showed no negative effects on survival as a result of B. anthracis exposure. One of the major conclusions of the studies conducted by Schelkle et al. was that C. elegans, when combined with a germinant to induce spores to germinate into vegetative bacteria, were an effective bioremediation of soil contaminated with B. anthracis spores (Schelkle et al., 2017). However, these results were only measurably significant with soil samples that had been heavily spiked with B. anthracis spores. Our data challenge the efficacy of such an approach, and the use of C. elegans in the predation of B. anthracis. In our hands, ingestion of B. anthracis only occurred to a significant amount when the worms encountered nothing but B. anthracis; otherwise, B. anthracis was rarely ingested (Fig. 4). It is not likely, in a soil contaminated with B. anthracis spores and in a native environment, that B. anthracis would be the only possible food source for the worms. We suspect that the spiking of soil samples with B. anthracis spores influenced the results shown by Schelkle et al. and the predation effects noted in their data; we believe

Fig. 4. C. elegans preferentially ingest E. coli over B. anthracis when the two are mixed, and rarely ingest B. anthracis even when it is the only available food source. (A-B) N2 worms were plated onto a 1:1 mixture of E. coli and B. anthracis (as determined by OD600 values) and spread evenly across the plates. At given time points, worms were completely removed from plates and gDNA was harvested from total bacteria remaining on the plate. PCR was performed using speciesspecific primers against the DNA Polymerase III locus of the E. coli and B. anthracis genomes. Reactions were set up in tandem, with tubes from each treatment being removed after 10, 15, 20, 25, 30 and 35 rounds of PCR. Reactions were run on agarose gels, and volumes of the expected amplicons were measured. The ratio of the volumes of E. coli/B. anthracis were calculated for each time point and for each PCR sample. (A) The ratios of E. coli/B. anthracis volumes in days 2 and 5 after 20 rounds of PCR (linear phase of cycling). PCR results from bacterial gDNA harvested from plates that had no worms are shown as a control (blue); PCR results from bacterial gDNA harvested from plates upon which worms were allowed to feed are also shown (orange). (B) The differences of the ratios between the “no worm” control and the “worm” treatment described in part “A” are shown. (C–H) Mixed-stage worms were fed bacteria alternatively labeled with fluorescent markers (GFP or RFP). After 12–16 h feeding, worms were counted and imaged using a fluorescence microscope fitted with GFP/RFP filters. (C) Percentage of worms with fluorescence (GFP and/or RFP) in the pharynx is shown for each of five different food mixtures. N = number of worms counted over two experiments. “P” values were determined by Chi-square analysis. (D-H) Representative images of worms fed with different fluorescently-labelled food mixtures. BF = bright field. 9

Developmental and Comparative Immunology 102 (2020) 103453

M.J. Turner, et al.

Fig. 4. (continued)

it likely that such effects would not be seen in a more native and natural environment. Generally speaking, our data suggest an important consideration to make in assaying avoidance phenotypes. Behavior classically

categorized as avoidance may or may not have to do with an innateimmune response. We have demonstrated here what we believe to be an avoidance behavior based more on the preference and/or the ability of worms to ingest bacteria. As such, avoidance behavior by itself is not 10

Developmental and Comparative Immunology 102 (2020) 103453

M.J. Turner, et al.

Fig. 5. The moderate lifespan extension seen in C. elegans upon exposure to B. anthracis is suppressed, in a DAF-16-dependent manner, when worms are forced into contact with the bacteria. Worms were plated on E. coli or B. anthracis that had been spread across the plate evenly (spread food), or left pooled at the center of the plate and allowed to dry (pooled food). Lifespan was measured as described above. Worms were also plated on empty NGM plates as a control (no food). Error bars represent 95% confidence intervals. (Note: Lifespan assays performed with worms on spread E. coli showed no significant difference from worms plated on pooled E. coli). “P” values for represented curve comparisons are as follows: A: < 0.00001, < 0.00001, 0.016, < 0.00001, 0.389. B: < 0.00001, < 0.00001, < 0.00001, < 0.00001, < 0.00001. C: 0.000011, 0.00095, 0.174, 0.065, 0.607. D: < 0.00001, < 0.00001, < 0.00001, 0.115, 0.31. E: < 0.00001, 0.000021, < 0.00001, < 0.00001, < 0.00001. F: 0.000011, < 0.00001, 0.00001, 0.00415, 0.00228.

GFP) strains were provided by the CGC, which is funded by NIH Office of Research Infrastructure Programs (P40 OD010440). E. coli (OP50RFP) was a generous gift from the laboratory of Dr. David Gems at University College London, London, UK. B. anthracis (Sterne-GFP) was graciously provided by Dr. Rekha Panchal's laboratory at USAMRIID, Frederick, MD, USA.

sufficient to assume pathogenicity of candidate bacterial pathogens. Furthermore, molecular evidence – such as active signaling of the innate-immune signal pathways and upregulation of stress-response and/ or immune-response genes – must be coupled with avoidance assays before attributing avoidance behavior to an innate-immune response. This research was funded in part by grants awarded USAMRIID from the U.S. Department of Defense's Defense Threat Reduction Agency (DTRA), and by moneys granted MJT through the Mount St. Mary's University Department of Science. The authors would like to acknowledge and specifically thank Dr. Abigail Kula for assistance with ANOVA studies and for consultation regarding statistical methods, Rhonda Turner for technical assistance with lifespan and feeding assays, and Kirsten Dugan and Brent Dugan for proofreading assistance. Worm strains and E. coli (OP50 and OP50-

References Anyanful, A., Easley, K.A., Benian, G.M., Kalman, D., 2009. Conditioning protects C. elegans from lethal effects of enteropathogenic E. coli by activating genes that regulate lifespan and innate immunity. Cell Host Microbe 5, 450–462. Avery, L., Horvitz, H.R., 1989. Pharyngeal pumping continues after laser killing of the pharyngeal nervous system of C. elegans. Neuron 3, 473–485. Carrera, M., Zandomeni, R.O., Fitzgibbon, J., Sagripanti, J.L., 2007. Difference between the spore sizes of Bacillus anthracis and other Bacillus species. J. Appl. Microbiol. 102,

11

Developmental and Comparative Immunology 102 (2020) 103453

M.J. Turner, et al.

lives twice as long as wild type. Nature 366, 461–464. Kho, M.F., Bellier, A., Balasubramani, V., Hu, Y., Hsu, W., Nielsen-LeRoux, C., McGillivray, S.M., Nizet, V., Aroian, R.V., 2011. The pore-forming protein Cry5B elicits the pathogenicity of Bacillus sp. against Caenorhabditis elegans. PLoS One 6, e29122. Kim, D.H., 2013. Bacteria and the aging and longevity of Caenorhabditis elegans. Annu. Rev. Genet. 47, 233–246. Kimura, K.D., Tissenbaum, H.A., Liu, Y., Ruvkun, G., 1997. daf-2, an insulin receptor-like gene that regulates longevity and diapause in Caenorhabditis elegans. Science 277, 942–946. Kiyama, Y., Miyahara, K., Ohshima, Y., 2012. Active uptake of artificial particles in the nematode Caenorhabditis elegans. J. Exp. Biol. 215, 1178–1183. Koppes, L.H., Woldringh, C.L., Nanninga, N., 1978. Size variations and correlation of different cell cycle events in slow-growing Escherichia coli. J. Bacteriol. 134, 423–433. Kumar, S., Egan, B.M., Kocsisova, Z., Schneider, D.L., Murphy, J.T., Diwan, A., Kornfeld, K., 2019. Lifespan extension in C. elegans caused by bacterial colonization of the intestine and subsequent activation of an innate immune response. Dev. Cell 49, 100–117 e106. Kurz, C.L., Tan, M.W., 2004. Regulation of aging and innate immunity in C. elegans. Aging Cell 3, 185–193. Mahajan-Miklos, S., Tan, M.W., Rahme, L.G., Ausubel, F.M., 1999. Molecular mechanisms of bacterial virulence elucidated using a Pseudomonas aeruginosa-Caenorhabditis elegans pathogenesis model. Cell 96, 47–56. Meisel, J.D., Kim, D.H., 2014. Behavioral avoidance of pathogenic bacteria by Caenorhabditis elegans. Trends Immunol. 35, 465–470. Nakad, R., Snoek, L.B., Yang, W., Ellendt, S., Schneider, F., Mohr, T.G., Rosingh, L., Masche, A.C., Rosenstiel, P.C., Dierking, K., Kammenga, J.E., Schulenburg, H., 2016. Contrasting invertebrate immune defense behaviors caused by a single gene, the Caenorhabditis elegans neuropeptide receptor gene npr-1. BMC Genomics 17, 280. Ogg, S., Paradis, S., Gottlieb, S., Patterson, G.I., Lee, L., Tissenbaum, H.A., Ruvkun, G., 1997. The Fork head transcription factor DAF-16 transduces insulin-like metabolic and longevity signals in C. elegans. Nature 389, 994–999. Pradel, E., Zhang, Y., Pujol, N., Matsuyama, T., Bargmann, C.I., Ewbank, J.J., 2007. Detection and avoidance of a natural product from the pathogenic bacterium Serratia marcescens by Caenorhabditis elegans. Proc. Natl. Acad. Sci. U.S.A. 104, 2295–2300. Pujol, N., Link, E.M., Liu, L.X., Kurz, C.L., Alloing, G., Tan, M.W., Ray, K.P., Solari, R., Johnson, C.D., Ewbank, J.J., 2001. A reverse genetic analysis of components of the Toll signaling pathway in Caenorhabditis elegans. Curr. Biol. 11, 809–821. Schelkle, B., Choi, Y., Baillie, L.W., Richter, W., Buyuk, F., Celik, E., Wendling, M., Sahin, M., Gallagher, T., 2017. Caenorhabditis elegans predation on Bacillus anthracis: decontamination of spore contaminated soil with germinants and nematodes. Front. Microbiol. 8, 2601. Schulenburg, H., Ewbank, J.J., 2007. The genetics of pathogen avoidance in Caenorhabditis elegans. Mol. Microbiol. 66, 563–570. Spencer, R.C., 2003. Bacillus anthracis. J. Clin. Pathol. 56, 182–187. Taffoni, C., Pujol, N., 2015. Mechanisms of innate immunity in C. elegans epidermis. Tissue Barriers 3, e1078432. Tan, M.W., Mahajan-Miklos, S., Ausubel, F.M., 1999a. Killing of Caenorhabditis elegans by Pseudomonas aeruginosa used to model mammalian bacterial pathogenesis. Proc. Natl. Acad. Sci. U.S.A. 96, 715–720. Tan, M.W., Rahme, L.G., Sternberg, J.A., Tompkins, R.G., Ausubel, F.M., 1999b. Pseudomonas aeruginosa killing of Caenorhabditis elegans used to identify P. aeruginosa virulence factors. Proc. Natl. Acad. Sci. U.S.A. 96, 2408–2413. Wang, J., Nakad, R., Schulenburg, H., 2012. Activation of the Caenorhabditis elegans FOXO family transcription factor DAF-16 by pathogenic Bacillus thuringiensis. Dev. Comp. Immunol. 37, 193–201. Zhang, Y., Lu, H., Bargmann, C.I., 2005. Pathogenic bacteria induce aversive olfactory learning in Caenorhabditis elegans. Nature 438, 179–184.

303–312. Cypser, J.R., Johnson, T.E., 2002. Multiple stressors in Caenorhabditis elegans induce stress hormesis and extended longevity. J Gerontol A Biol Sci Med Sci 57, B109–B114. Darby, C., Hsu, J.W., Ghori, N., Falkow, S., 2002. Caenorhabditis elegans: plague bacteria biofilm blocks food intake. Nature 417, 243–244. Dierking, K., Yang, W., Schulenburg, H., 2016. Antimicrobial effectors in the nematode Caenorhabditis elegans: an outgroup to the Arthropoda. Philos. Trans. R. Soc. Lond. B Biol. Sci. 371. Dorman, J.B., Albinder, B., Shroyer, T., Kenyon, C., 1995. The age-1 and daf-2 genes function in a common pathway to control the lifespan of Caenorhabditis elegans. Genetics 141, 1399–1406. Engelmann, I., Pujol, N., 2010. Innate immunity in C. elegans. Adv. Exp. Med. Biol. 708, 105–121. Ermolaeva, M.A., Schumacher, B., 2014. Insights from the worm: the C. elegans model for innate immunity. Semin. Immunol. 26 (4), 303–309. Evans, E.A., Chen, W.C., Tan, M.W., 2008a. The DAF-2 insulin-like signaling pathway independently regulates aging and immunity in C. elegans. Aging Cell 7, 879–893. Evans, E.A., Kawli, T., Tan, M.W., 2008b. Pseudomonas aeruginosa suppresses host immunity by activating the DAF-2 insulin-like signaling pathway in Caenorhabditis elegans. PLoS Pathog. 4, e1000175. Ewald, C.Y., Marfil, V., Li, C., 2016. Alzheimer-related protein APL-1 modulates lifespan through heterochronic gene regulation in Caenorhabditis elegans. Aging Cell 15, 1051–1062. Franks, S.E., Ebrahimi, C., Hollands, A., Okumura, C.Y., Aroian, R.V., Nizet, V., McGillivray, S.M., 2014. Novel role for the yceGH tellurite resistance genes in the pathogenesis of Bacillus anthracis. Infect. Immun. 82, 1132–1140. Fuchs, B.B., Mylonakis, E., 2006. Using non-mammalian hosts to study fungal virulence and host defense. Curr. Opin. Microbiol. 9, 346–351. Gaglia, M.M., Jeong, D.E., Ryu, E.A., Lee, D., Kenyon, C., Lee, S.J., 2012. Genes that act downstream of sensory neurons to influence longevity, dauer formation, and pathogen responses in Caenorhabditis elegans. PLoS Genet. 8, e1003133. Gravato-Nobre, M.J., Hodgkin, J., 2005. Caenorhabditis elegans as a model for innate immunity to pathogens. Cell Microbiol. 7, 741–751. Hasshoff, M., Bohnisch, C., Tonn, D., Hasert, B., Schulenburg, H., 2007. The role of Caenorhabditis elegans insulin-like signaling in the behavioral avoidance of pathogenic Bacillus thuringiensis. FASEB J. 21, 1801–1812. Huang, X., Zhang, H., 2011. The zinc-finger protein SEA-2 regulates larval developmental timing and adult lifespan in C. elegans. Development 138, 2059–2068. Jayamani, E., Tharmalingam, N., Rajamuthiah, R., Coleman, J.J., Kim, W., Okoli, I., Hernandez, A.M., Lee, K., Nau, G.J., Ausubel, F.M., Mylonakis, E., 2017. Characterization of a Francisella tularensis- Caenorhabditis elegans pathosystem for the evaluation of therapeutic compounds. Antimicrob. Agents Chemother. 61. JebaMercy, G., Vigneshwari, L., Balamurugan, K., 2013. A MAP Kinase pathway in Caenorhabditis elegans is required for defense against infection by opportunistic Proteus species. Microb. Infect. 15, 550–568. Johnson, T.E., de Castro, E., Hegi de Castro, S., Cypser, J., Henderson, S., Tedesco, P., 2001. Relationship between increased longevity and stress resistance as assessed through gerontogene mutations in Caenorhabditis elegans. Exp. Gerontol. 36, 1609–1617. Johnson, T.E., Henderson, S., Murakami, S., de Castro, E., de Castro, S.H., Cypser, J., Rikke, B., Tedesco, P., Link, C., 2002. Longevity genes in the nematode Caenorhabditis elegans also mediate increased resistance to stress and prevent disease. J. Inherit. Metab. Dis. 25, 197–206. Kaeberlein, T.L., Smith, E.D., Tsuchiya, M., Welton, K.L., Thomas, J.H., Fields, S., Kennedy, B.K., Kaeberlein, M., 2006. Lifespan extension in Caenorhabditis elegans by complete removal of food. Aging Cell 5, 487–494. Kawli, T., Tan, M.W., 2008. Neuroendocrine signals modulate the innate immunity of Caenorhabditis elegans through insulin signaling. Nat. Immunol. 9, 1415–1424. Kenyon, C., Chang, J., Gensch, E., Rudner, A., Tabtiang, R., 1993. A C. elegans mutant that

12