Bacterial contact induces polar plug disintegration to mediate whipworm egg hatching

The bacterial microbiota promotes the life cycle of the intestine-dwelling whipworm Trichuris by mediating hatching of parasite eggs ingested by the mammalian host. Despite the enormous disease burden associated with Trichuris colonization, the mechanisms underlying this transkingdom interaction have been obscure. Here, we used a multiscale microscopy approach to define the structural events associated with bacteria-mediated hatching of eggs for the murine model parasite Trichuris muris. Through the combination of scanning electron microscopy (SEM) and serial block face SEM (SBFSEM), we visualized the outer surface morphology of the shell and generated 3D structures of the egg and larva during the hatching process. These images revealed that exposure to hatching-inducing bacteria catalyzed asymmetric degradation of the polar plugs prior to exit by the larva. Unrelated bacteria induced similar loss of electron density and dissolution of the structural integrity of the plugs. Egg hatching was most efficient when high densities of bacteria were bound to the poles. Consistent with the ability of taxonomically distant bacteria to induce hatching, additional results suggest chitinase released from larva within the eggs degrade the plugs from the inside instead of enzymes produced by bacteria in the external environment. These findings define at ultrastructure resolution the evolutionary adaptation of a parasite for the microbe-rich environment of the mammalian gut.


Introduction
Soil transmitted helminths (STH) are parasitic worms that affect nearly 1.5 billion people worldwide [1].The developmental maturation of certain STHs occurs in the gastrointestinal (GI) tract where the parasites encounter trillions of bacteria that are part of the gut microbiota, as exemplified by the whipworm Trichuris.Infection with Trichuris species initiates when embryonated eggs in the environment are ingested, which then hatch in the large intestine where the bacterial community is most diverse and dense [2][3][4].In this environment, the hatched larvae then embed themselves in the intestinal epithelium where they remain anchored as they undergo several molts to become sexually reproductive adult worms [3].Heavy worm burden in individuals infected by the human parasite Trichuris trichiura is associated with colitis, anaemia, and dysentery [5][6][7][8].Single doses of the available anthelmintic drugs for T. trichiura infections display poor efficacy with an average cure rate of less than 50% [9].A better understanding of how Trichuris species have adapted to their host may reveal vulnerabilities that can be targeted for intervention.
Recent studies using the murine parasite Trichuris muris have revealed a remarkable degree of co-adaptations with the mammalian host and the gut microbiota.Bacteria, which trigger egg hatching in vitro, are necessary for T. muris to establish infection [10][11][12].In turn, T. muris colonization alters the microbiota composition of mice [13], which we and others have shown is consequential for the host.In a model of inflammatory bowel disease (IBD), we found the type 2 immune response to T. muris colonization protects against intestinal inflammation by increasing Clostridiales and reducing Bacteroidales species within the gut microbiota [14].Another murine intestinal parasite Heligmosomoides bakeri also participates in three-way interactions with the host and microbiota including inducing an outgrowth of Clostridiales that attenuates allergic airway inflammation [15,16].Colonization of indigenous people in Malaysia with T. trichiura is associated with a similar expansion of Clostridiales and reduction in Bacteroidales in the microbiota, indicating that these relationships are conserved in humans [14].Further, we found that Clostridiales species induce superior egg hatching of T. muris and T. trichiura compared with Bacteroidales species [17].These observations suggest that Trichuris colonization alters the microbiota composition in a manner that may be mutually beneficial for the parasite and mammalian host in certain contexts.
Given this dependence on the microbiota, egg hatching may be a point of vulnerability in the parasite lifecycle.Trichuris eggs are ovoid shaped with a multi-layered shell.A membrane-

PLOS PATHOGENS
Whipworm egg structure during bacteria-mediated hatching like outer vitelline layer covers the entire surface of the egg, and underneath is a middle layer consisting of mainly chitin and a lower lipid layer [18,19].These eggs have two openings, one at each end, that are blocked by chitinous polar plugs [4].Although these plugs are also multilayered, they have a higher proportion of chitin in its middle layer than the rest of the shell [4].During the hatching process (eclosion), the oral spear in the anterior end of the larva pierces the outer vitelline membrane and exits through a polar plug on one side of the egg [20,21].The role of bacteria in this process remains obscure.For the Gram-negative bacterial species Escherichia coli, type 1 fimbriae mediate binding to the polar ends of eggs and are necessary for optimal hatching [10].However, Gram-positive species that do not have fimbriae such as Staphylococcus aureus can also induce efficient hatching in vitro [10,11,17].It is unclear whether the structural events associated with hatching in the presence of Gram-negative and -positive species are similar.
In this study, we show that although many Gram-positive bacterial species are strong hatching inducers, Enterococci fail to trigger hatching.We found that formation of high-density bacterial clusters was associated with faster hatching rates.High resolution 3D volume electron microscopy imaging revealed the ultrastructural organization of the eggshell and larva, and identified striking asymmetrical morphological changes to the polar plug regions that occur upon exposure to E. coli and S. aureus, representative Gram-negative and -positive bacteria that induce hatching.We further demonstrate that the degree of bacterial binding to eggs is directly proportional to the hatching rate yet hatching required bacteria to be metabolically active.Finally, we show that eggs from both T. muris and the human pathogen T. trichiura harbor chitinase activity and propose that this activity plays a role in hatching by degrading the chitinous polar plug of the egg from within.

Bacteria display species-dependent effects on T. muris egg hatching
To establish conditions for investigating structural changes that occur during T. muris egg hatching, we first investigated the degree to which bacterial taxa differ in their ability to induce hatching.Although previous studies have shown that both the Gram-negative species E. coli and the Gram-positive species S. aureus can induce hatching [10,11], only a limited number of bacterial taxa have been examined under the same conditions and it was unclear whether the time course differs between bacterial species.We used previously optimized conditions in which bacteria cultures grown to their maximal density (saturated) are added to embryonated T. muris eggs in a culture well and monitored over time for hatching by light microscopy under aerobic conditions [12] (Fig 1A, S1 Video).Saturated overnight bacterial cultures were previously used for in vitro hatching assays by others in the field and were also shown to induce optimal hatching in our lab [10,12,17].As such, overnight cultures were used for hatching assays unless stated otherwise.We found that E. coli and S. aureus induced comparable levels of T. muris egg hatching by four hours post-incubation, while untreated eggs remained unhatched (Fig 1B).However, a higher proportion of eggs were hatched in the presence of S. aureus at earlier time points compared with E. coli under these conditions (Fig 1B and 1C).Pseudomonas aeruginosa and Salmonella enterica Typhimurium-Gram-negative Proteobacteria related to E. coli-were previously shown to induce T. muris egg hatching under similar conditions [10,17].To increase the number of different Gram-positive bacterial species investigated in this assay, we examined Staphylococcus epidermidis as a member of the same genus as S. aureus, and Bacillus subtilis and Enterococcus faecalis as unrelated taxa.S. epidermidis induced hatching at a similar rate as S. aureus, while B. subtilis also induced efficient hatching, albeit at a modestly slower rate.In contrast, hatching failed to occur in the presence of E.

PLOS PATHOGENS
Whipworm egg structure during bacteria-mediated hatching

PLOS PATHOGENS
Whipworm egg structure during bacteria-mediated hatching faecalis.To determine whether this was specific to E. faecalis, we tested another member of the genus, Enterococcus faecium, and found that it also did not induce hatching (Fig 1D).These differences in hatching rates cannot be explained by the culture media used to grow each bacterial species because S. aureus grown in each of the three media types (BHI, TSB, and LB) induced similar levels of hatching across time points (Fig 1E).
Because we used overnight bacterial cultures based on previously established protocols, conditions were normalized based on volume and growth phase of bacterial cultures rather than bacterial numbers.Consistent with the ability of staphylococci to grow as tightly packed clusters, a higher number of S. aureus (~10 8 colony forming units; CFUs) were added to eggs compared with E. coli (~5x10 7 CFUs) in the above experiments.Therefore, bacterial density may explain differential T. muris egg hatching rates.Indeed, the difference in early hatching was eliminated when we concentrated E. coli to a superphysiological concentration to match the maximal density that is naturally achieved by S. aureus (Fig 1F).This observation indicates that hatching efficiency is sensitive to the number of bacteria present.However, increasing the number of E. faecalis led to negligible hatching (Fig 1F).This lack of hatching was not due to the production of an inhibitory or toxic factor by this bacterium because adding E. faecalis did not negatively impact S. aureus-mediated hatching (Fig 1G).These results show that taxonomically distant Gram-positive and -negative bacteria induce hatching with similar efficiency when similar number of bacteria are added to the culture, although there are Gram-positive taxa such as enterococci that are unable to induce hatching.
Given that hatching occurs predominantly in the anaerobic environment of the cecum [2], we confirmed that E. coli and S. aureus were able to induce hatching in anaerobic conditions (Fig 1H).We previously demonstrated that monocolonization of germ-free mice with E. coli is sufficient to support T. muris development following inoculation with embryonated eggs, although the number of worms recovered is less than conventional mice [12].To determine whether Gram-positive species can support parasite colonization, we monocolonized mice with S. aureus or E. faecalis and inoculated them with two doses of 100 T. muris eggs.Successful colonization was confirmed by plating stool on selective agar (Fig 1I).We recovered adult worms from the cecum from all S. aureus monocolonized mice 42 days after T. muris inoculation, whereas most germ-free control mice did not harbor worms or had reduced numbers of worms (Fig 1J and 1K).There was no significant difference in the number of worms recovered incubation with E. coli (~5 x 10 7 CFU), S. aureus (~10 8 CFU) and E. faecalis (~5 X 10 7 CFU) grown overnight to saturation.(D) Percent of T. muris eggs hatched after incubation in aerobic conditions with overnight cultures of S. epidermidis, B. subtilis, E. faecalis and E. faecium compared with their respective broth controls determined by light microscopy at indicated time points.Colony forming units (CFUs) of each bacterial species added to the eggs are indicated on the graphs.(E) Percent of T. muris eggs hatched after incubation with S. aureus grown overnight in tryptic soy broth (TSB), Luria Bertani (LB) broth and Brain Heart Infusion (BHI) broth.(F) Percent of T. muris eggs hatched at 1 hour after 37˚C incubation with an equal number (~10 8 CFU) of E. coli and E. faecalis.(G) Percent of T. muris eggs hatched after incubation with S. aureus (~3 x 10 8 CFU) and E. faecalis (~3 x 10 7 CFU) alone or together compared with broth controls.For the S. aureus + E. faecalis condition, bacterial cultures were grown separately overnight, and then were mixed the day of the experiment.(H) Percent of T. muris eggs hatched after incubation in anaerobic conditions with overnight cultures of E. coli (~5 x 10 7 CFU) and S. aureus (~2 x 10 8 CFU) compared with their respective broth controls determined by light microscopy at indicated time points.(I) Quantification of CFUs per gram of stool collected from mice monocolonized with S. aureus or E. faecalis.(J) Proportion of germ-free and S. aureus monocolonized mice that had harbored adult T. muris worms after double-dose infection.(K) Number of worms recovered from S. aureus monocolonized mice (n = 9-12 mice per group).(L) Number of worms recovered from E. faecalis monocolonized mice (n = 3 mice per group).Data points and error bars represent mean and SEM from 3 independent repeats for (B), (D), (E), (G), and (H).Dots represent a single well and bars show means and SEM from 3 independent repeats for (C) and (F).~25 eggs per well were assed for (A)-(H).Dots represent a single mouse and bars show means and SEM from 3 independent repeats for (K) and 1 independent experiment for (L).Welch's t test was used to compare area under the curve between each condition and its respective media control for (B), (D), and (H).Kruskal-Wallis test followed by a Dunn's multiple comparisons test was used for (C).Ordinary one-way analysis of variance (ANOVA) test followed by a Tukey's multiple comparisons test was used to compare AUC of hatching induced by different conditions to each other for (E) and (G).Mann Whitney test was used for (F), (K) and (L).Fisher's exact test was used to determine whether there was a significant association between gut microbial composition of mice and the presence of adult worms in the cecum for (J).https://doi.org/10.1371/journal.ppat.1011647.g001

PLOS PATHOGENS
Whipworm egg structure during bacteria-mediated hatching from germ-free mice and E. faecalis monocolonized mice, supporting our observation that E. faecalis is unable to induce T. muris egg hatching (Fig 1L).
Intestinal S. aureus is detected in individuals living in helminth-endemic rural regions [22].Given the propensity of staphylococci to form dense clusters that facilitate hatching, we explored potential associations between S. aureus gut colonization and helminth infection in humans.We previously generated a large metagenomics dataset from stool specimens collected from individuals of a Malaysian rural indigenous population, the Orang Asli, who display a high prevalence of T. trichiura infections [23].Compared with urban controls who were verified as negative for helminth infections, we found that S. aureus was present at higher levels in stool specimens from Orang Asli individuals (S1 Fig) .Strikingly, S. aureus numbers were reduced in Orang Asli individuals following 21 and 42 days of treatment with the anthelmintic albendazole (S1 Fig) .Although causal relationships cannot be determined from this analysis, these observations suggest a bidirectional relationship between S. aureus and T. trichiura, a possibility that should be explored further.Bacterial products may contribute to both egg hatching and sexual development of Trichuris [12].Below we use S. aureus and E. coli as representative Gram-positive and -negative species for hatching assays based on their genetic tractability and ability to grow aerobically, and to facilitate comparisons with prior studies.S. aureus was particularly of interest given its presence in Trichuris infected individuals and that prior work focused on E. coli.

Physical contact with egg is essential for S. aureus-mediated hatching
T. muris egg hatching is reduced in the presence of E. coli when direct contact is inhibited [10,11].Given the structural and growth differences between E. coli and S. aureus, we revisited contact-dependence of hatching using the above conditions.We initially tested whether S. aureus secretes a soluble factor that is sufficient to induce hatching.Similar to our previous finding with E. coli [12], the addition of filtered S. aureus culture supernatant to T. muris eggs did not result in hatching (Fig 2A).It is possible that the presence of T. muris eggs is required for S. aureus to produce a pro-hatching factor, which we would miss in a simple supernatant transfer experiment.To address this possibility, we incubated S. aureus with eggs, collected and filtered the supernatant from the co-culture, and then added this egg-primed supernatant to a new batch of eggs (Fig 2B).The co-culture supernatant was also unable to induce hatching of eggs (Fig 2C).To further determine whether contact is necessary, we separated eggs by placing them inside a 0.4μm transwell insert and added bacteria to the outside well [12].Although moderate amounts of hatching occurred when eggs were segregated from E. coli in this manner, no hatching was observed when eggs were placed in transwells that were incubated with S. aureus (Fig 2D).These findings indicate that direct contact with eggs is required for S. aureus to induce hatching.

Collapse of the polar plug precedes hatching mediated by bacteria
Bacteria-induced T. muris egg hatching has not been examined with ultrastructural resolution.We examined eggs exposed to S. aureus and E. coli by scanning electron microscopy (SEM) and were able to capture eggs at different stages in the hatching process as evidenced by larvae in mid-ejection and morphological changes to the plug not observed in untreated eggs and eggs exposed to E. faecalis, a poor hatching-inducing species (Fig 3A -3E).For both untreated and E. faecalis-treated eggs, the plugs appeared as crater-like structures approximately 5 μm in diameter with an inner surface displaying a slightly rounded wrinkled morphology (Fig 3C and 3D).Diplococci characteristic of Enterococci were present at the plugs of eggs incubated with E. faecalis (Fig 3C).Untreated eggs had little to no visible bacterial cells on their surface

PLOS PATHOGENS
Whipworm egg structure during bacteria-mediated hatching (Fig 3D).However, some untreated eggs were found to harbor bacteria (S2A Fig) .Because eggs were harvested from adult worms isolated from the cecum of mice, these bacteria were likely derived from the mouse gut microbiota [13].We were able to recover a modest number of bacterial colonies when we plated eggs on LB plates, which through colony PCR we identified as Brucella cytisi and E. faecalis (S2B Fig) .The presence of bacteria on eggs may explain why we recovered a small number of worms from a minority of germfree mice.
Eggs incubated with S. aureus and E. coli were enriched for the presence of bacteria at their plugs (Fig 3A and 3B), which displayed a collapsed morphology distinct from untreated eggs (Fig 3D ) and eggs incubated with E. faecalis (Fig 3C).S. aureus and E. coli were densely clustered within or near depressions.Additionally, there was a wooly substance that was closely associated with the bacteria on the plug region (Fig 3A and 3B,right).The vitelline membrane of the plug was also visible and resembled a deflated balloon (Fig 3B , right).Bacteria, while present sporadically across the eggshell, were more densely packed on the plug region than other regions.Examination of eggs incubated with S. aureus and E. coli in which worms were captured mid-ejection showed that, in both instances, larvae exited the egg through the plug, rupturing it along with the outer vitelline layer in the process (Fig 3E).In these eggs in which larvae were partially out of the shell, bacteria were found near the polar plug region (Fig 3E, top and bottom).
We observed more individual bacterial cells per disrupted plug when eggs were incubated with S. aureus than other conditions, likely reflecting their capacity for growing in clusters.

PLOS PATHOGENS
However, the number of visible E. faecalis cells on the polar plug region was comparable to the number of E. coli cells on the structurally distinct polar plug, suggesting that differences in the density of bacterial species at the plug are not sufficient to explain why Enterococci are poor hatching inducers (Fig 3F).Additionally, the diameter of the collar region on eggs in the process of hatching was generally wider than on eggs exposed to bacteria, but that had not yet begun to hatch, consistent with the previously described observation that plugs swell during the hatching process [21] (Fig 3G).These results show that exposure of eggs to strong hatching-inducing bacteria is associated with a collapse and loss of structural integrity of the polar plugs.

Hatching-inducing bacteria trigger disintegration of the polar plug
Advances in the volume electron microscopy technique serial block face-scanning electron microscopy (SBFSEM) present an opportunity to gain additional insight into the above structural changes that we observed in eggs exposed to S. aureus and E. coli.Preparation of samples for ultrastructural electron microscopy imaging normally involves chemical fixation, staining, dehydration and then finally embedding of the specimen in resin [24].However, the impermeability of the eggshell made preservation of the egg contents using routine chemical fixation and high pressure freezing and freeze substitution methods difficult [25].To overcome this challenge, we used a microwave assisted sample preparation method to increase the penetration of fixative and stains into the egg, which enabled identification of structures corresponding to the eggshell, polar plugs, and the larvae (Fig 4A -4C, S2 Video).The eggshell thickness

PLOS PATHOGENS
was comparable between both bacteria-treated eggs, 2.17μm and 2.18μm for E. coli and S. aureus-exposed eggs, respectively (Fig 4A -4C).Cells and other structures within the larvae and the cuticle confining the larvae were visible.Characteristic granules of the larval intestinal tract were observed, and we propose that those granules contain lipids based on the low electron density of the contents (Fig 4A, S2 Video) [20].The plug on one end displayed a more granular and less electron dense morphology compared with the contralateral plug for both S. aureus-and E. coli-exposed eggs.We arbitrarily designated the plug with more evidence of disintegration as the top end of the egg and the contralateral plug as the bottom.The anterior end of the larva was closer to the top plug that was granular and pointed towards the bottom plug in both cases (Fig 4A -4C For Trichuris eggs not treated with bacteria, we found that the fixative did not penetrate through the intact shell even with the microwave assisted fixation method, precluding structural analysis of the larva.It is likely that the addition of bacteria was necessary to increase permeability of the shell allowing fixation of the inside.However, we were able to obtain high resolution images of the eggshell and pole regions of untreated eggs for comparison with bacteria-treated eggs.Eggshells were 2.31μm thick, similar to bacteria-treated eggs (Fig 4A -4D).The polar plugs appeared as uniformly electron dense structures rounded towards the outer surface for egg samples that were not exposed to hatching-inducing bacteria (Fig 4D , right).While being rounded, the height of the plug (Fig 4A, arrow iii) of the untreated egg did not extend far beyond the top of the collar on this untreated egg (3.29μm top, 3.49μm bottom).These intact plugs also contained an electron dense outer vitelline layer that was continuous with the rest of the shell and overlayed a similarly electron dense chitinous layer enclosed by an electron dense, ridged collar (Fig 4D , right).
In contrast, eggs incubated with S. aureus or E. coli displayed plugs that were morphologically distinct from untreated egg plugs.First, they were much less electron dense and non-uniform in electron density.As noted above, these changes were asymmetric.For both bacteria, one plug completely lost structural integrity and appeared more granular and disintegrated than the contralateral plug on the same egg (Fig 4D).Instead of a smooth rounded surface, these plugs displayed indentations on the surface, which was more pronounced for the E. colitreated egg in which the plug had a large depression (Fig 4D and 4E; S2 Video).The contralateral pole was more intact compared with the top pole but was less electron dense compared with plugs on the egg not exposed to bacteria.There were two distinct regions of electron density for these contralateral poles exposed to S. aureus or E. coli, with an outer region of moderate density encompassing an inner less electron dense area (bottom) (Fig 4D , left and middle).In the S. aureus condition, the plug with the disintegrated morphology corresponded with the side associated with the bacterial clusters.Although different in electron density, both plugs on eggs incubated with bacteria were swollen and extended beyond the width of the collar region.Specifically, plugs on E. coli exposed eggs were 9.51μm and 10.43μm in width while their respective collars were 8.87μm and 8.68μm in diameter for the top and bottom, respectively (Fig 4D).Similarly, plugs on S. aureus exposed eggs were 9.46μm and 10.39μm wide while their respective collars were 8.62μm and 9.46μm wide (Fig 4D).Moreover, the heights of the plugs were larger on eggs exposed to E. coli and S. aureus (4.47μm top and 7.40μm bottom, and 5.63μm top and 6.51μm bottom, respectively) than on the untreated egg (Fig 4D).
We repeated the SBFSEM experiment with new bacteria-exposed egg samples to determine whether we can capture additional stages of plug disintegration.In this second run of SBFSEM on S. aureus-exposed eggs, one of the plugs (pole 1, S3A Fig) displayed similar morphology as the contralateral (bottom) plug on the S. aureus-exposed egg in Fig 4D , while the other plug (pole 2, S3A Fig) had a similar electron density to the outer region of the pole 1 plug but was uniformly electron dense.Additionally, the second run of SBFSEM on E. coli-exposed eggs showed plugs with a uniform and high electron density that were not swollen, similar to the plugs seen on eggs that were not treated with bacteria (S3B Fig) .Lastly, in the second run on S. aureus exposed eggs, we observed that the larva appeared to be directly interacting with the inner surface of pole 1 (S3C Fig).This result could be representative of additional and earlier stages of plug degradation.
Together, these results indicate that exposure to hatching inducing bacteria is associated with substantial morphological changes of the polar plugs, including swelling and a decrease in structural integrity.

Degree of bacterial binding is proportional to the rate of T. muris egg hatching
The strict requirement of physical binding for S. aureus-induced hatching and the striking images showing that this bacterial species was densely bound to the plug motivated us to quantitatively define the nature of this interaction.Towards this end, we used spinning-disk confocal microscopy to quantify the location and number of bacteria by incubating eggs with a S. aureus strain that expressed GFP (Fig 5A).At 30 minutes post-incubation, we found that the pole region contained a substantially higher proportion of GFP+ structures compared with the sides of the eggs (Fig 5B).Our cross-species comparisons (Fig 1 ) suggest that hatching efficiency is dependent on the number of bacteria incubated with the eggs.Consistent with this observation, when we added serial dilutions of hatching inducing bacterial species including S. aureus to eggs, we observed that the rate of hatching was proportional to the concentration of bacteria (Fig 5C, S4A-S4C Fig) .This sensitivity to the concentration of the inoculum may explain variability in hatching rates across experiments, especially at early time points.To test whether this relationship between bacterial concentration and hatching rates reflected binding events, we quantified the number and location of bacteria associated with eggs incubated with dilutions of S. aureus-GFP.The percentage of eggs associated with GFP+ puncta and the number of GFP+ puncta per egg after a 30-minute incubation were proportional to the concentration of S. aureus added to the culture (Fig 5D -5F).In contrast to the higher concentration conditions in which almost all GFP+ structures were associated with poles, the proportion of S. aureus bound to the poles versus other parts of the eggshell was reduced at lower concentrations (Fig 5G).These results are consistent with the ultrastructural analyses, suggesting that the degree of S. aureus binding to the polar plug region determines the rate of hatching.
In these experiments, bacteria were bound to both plugs on an egg.As bacterial exposure triggered asymmetric degradation of the polar plugs and hatching occurs through only one plug on an egg, we hypothesized that this could be due to different densities of bacteria bound to each pole.We quantified the number of puncta present on both poles on an egg and observed that bacteria also asymmetrically bind the poles at all concentrations of S. aureus assessed (Fig 5H ).

Bacterial metabolic activity is essential for S. aureus-mediated hatching
Egg hatching was low or completely absent within the first 30 minutes post-incubation with S. aureus (Figs 5C and 6A).Given our results showing a direct relationship between bacterial

PLOS PATHOGENS
Whipworm egg structure during bacteria-mediated hatching

PLOS PATHOGENS
Whipworm egg structure during bacteria-mediated hatching binding and hatching rate, it is possible this delay in hatching reflects the time required for bacteria to bind eggs.However, even at 0 minutes post-incubation with S. aureus-GFP, the majority of eggs were bound by bacteria based on their association with GFP+ puncta (Fig 6B ).By 10 minutes post-incubation, the proportion of eggs bound to bacteria and the number of bacterial cells bound to each egg reached the maximum value (Fig 6B and 6C).S. aureus was consistently enriched at plugs, and the ratio of bacteria bound to plugs versus other regions of the shell was similar across time points (Fig 6D).These findings indicate that bacterial binding is not the rate limiting step for hatching induction.
We previously showed that hatching required metabolically active E. coli, specifically arginine biosynthesis [12].Consistent with a requirement for viability, UV-killed S. aureus cells, even when concentrated up to 4X the equivalent number of live bacterial cells, failed to induce egg hatching (Fig 6E).Also, the bacteriostatic antibiotic chloramphenicol (CAM), which inhibits protein translation in S. aureus [26,27], reduced bacteria-mediated hatching (Fig 6F).This impaired hatching was not due to an effect of CAM on the egg itself because spiking-in S. aureus-GFP that is CAM-resistant at 2 hours after the beginning of the assay immediately restored hatching in the presence of CAM (Fig 6F ).
The above results demonstrate that metabolic activity is important for S. aureus mediated hatching.It is possible that this activity reflects the production of a secreted product that is missing in species such as E. faecalis that are unable to mediate hatching.To test this idea, we added E. faecalis and eggs together inside of a transwell insert and then added S. aureus to the outer compartment, thus allowing E. faecalis to physically associate with eggs while S. aureus products freely pass through the membrane.However, S. aureus was unable to rescue hatching in this setting (Fig 6G).We next determined whether arginine biosynthesis, which is important for E. coli-mediated hatching, is also involved in S. aureus-mediated hatching [12].However, hatching was unaffected when eggs were incubated with a S. aureus mutant that had a nonfunctional argininosuccinate lyase gene (S.aureus argH::Tn) (Fig 6H).In conclusion, egg binding and metabolic activity but not arginine biosynthesis are required for S. aureus to mediate hatching.These results argue against a simple model in which E. faecalis is missing a single metabolite common to E. coli and S. aureus, although we cannot rule out the possible role of an unstable product that is generated through different pathways by Gram-negative and-positive bacteria.

Trichuris eggs harbor chitinase activity
The biochemical process involved in Trichuris egg hatching is unknown.Hatching of eggs from the STH Ascaris lumbricoides is mediated by a parasite-derived chitinase enzyme [28].However, A. lumbricoides hatching occurs independently of bacteria.Thus, we explored the possibility that bacteria provided the chitinase for T. muris eggs to hatch.S. aureus, for example, produces degradative enzymes that have chitinase activity [29].To test this hypothesis, we used a S. aureus mutant with a deletion in a gene encoding a chitinase-related protein, which we refer to as CRP.Hatching occurred at a similar rate in the presence of S. aureus ∆crp and wild-type, indicating that CRP is not required (Fig 7A).In fact, S. aureus mutants deficient in all the major secreted proteases (S. aureus ∆aur∆sspAB∆scpAspl::erm) [30] as well as a S. aureus mutant lacking the three major lipases (S. aureus ∆gehA∆gehB∆gehE) [31] retained the ability to induce hatching with similar efficiency as wild-type bacteria (Fig 7B and 7C).These findings raise the possibility that the enzymatic activity responsible for the disintegration of the poles is derived from the parasite rather than bacteria.
Chitinase is released into the media during A. lumbricoides egg hatching [28].If T. muris also uses chitinase to degrade the polar plug from the inside, we should be able to detect .Data points and error bars represent mean and SEM from 3 independent repeats for (A), (E) and (F).Dots represent mean percentage of 40 eggs that were GFP+ from a single experiment for (B) and number of GFP+ puncta found on a single egg for (C).Dots represent a single well and bars show means and SEM from 3 independent experiments for (G).Bars and error bars show means and SEM and dots represent percent hatching from a single well from 2 independent repeats for (H).~25 eggs per well were assessed for (A), (E), (F) and (H) and ~200 eggs per well were assessed for (G).Kruskal-Wallis test followed by a Dunn's multiple comparisons test was used for (B), (C) and (G).Two-way ANOVA followed by a Sidak's multiple comparisons test was used for (D).Ordinary one-way ANOVA followed by Tukey's multiple comparisons test was used to compare AUC of different conditions tested for (F).

PLOS PATHOGENS
Whipworm egg structure during bacteria-mediated hatching chitinase activity in the post-hatching fluid, similar to A. lumbricoides.Using a previously described assay in which a substrate produces a fluorescent product when cleaved by chitinase [32], we detected chitinase activity in media from eggs incubated with S. aureus, but not in media containing untreated eggs or S. aureus alone (Fig 7D).Release of chitinase was not specific to S. aureus-mediated hatching as chitinase activity was also detected in media from eggs incubated with E. coli (Fig 7E).To determine whether chitinase within T. muris eggs is induced by exposure to bacteria, we measured chitinase activity in crushed eggs with or without preincubation with S. aureus.Chitinase activity was detected in the media containing crushed eggs that were exposed to bacteria as well as in eggs that were not exposed to bacteria (Fig 7F).Thus, eggs harbor chitinase activity independent of bacteria treatment.This experiment also rules out the possibility that bacteria were the source of the chitinase activity.We detected chitinase from crushed T. trichiura eggs (Fig 7G ), indicating that the human parasite also produces this enzyme.We were unable to detect any chitinase activity from crushed unembryonated T. muris eggs (Fig 7H).As unembryonated eggs are unable to hatch, this suggests that chitinase activity is only present in eggs that can hatch.
We tested whether media containing chitinase released from T. muris eggs induce hatching of intact eggs from the external environment ("outside-in").Fluid from hatched eggs was obtained by incubating eggs with bacteria for 4hrs and then filtering the supernatant.Fluid obtained from unhatched eggs was included as a negative control.We observed that fluid from hatched eggs did not induce hatching and was also unable to enhance S. aureus-mediated hatching (Fig 7I ).
Although the S. aureus mutant deficient in proteases mediated egg hatching similarly to wild-type bacteria, the addition of a protease inhibitor cocktail inhibited hatching (Fig 7J).Because the protease inhibitors will not discriminate between bacterial and parasitic target molecules, they may act on either S. aureus or T. muris eggs.

Discussion
In this study we illuminated structural events that occur during bacteria-mediated hatching of T. muris eggs by using multiscale microscopy techniques.Notably, we overcame technical challenges to apply SBFSEM to generate high resolution 3D images of eggs in the process of hatching.This technique along with SEM showed that, despite being taxonomically distant species with significant structural and biological differences, E. coli and S. aureus induce similar morphological changes in the polar plugs of eggs.These structural changes include swelling of the plugs, and disintegration of the plug material, resulting in loss of structural integrity of the plugs that appear as dips and depressions on the plug surface.
A previous study describing the egg hatching cascade showed that prior to hatching, the head of the larva inserts into the plug space and then uses its stylet to tear through the vitelline trichiura eggs.0.02mU of stock chitinase were used as a positive control.(H) Amount of fluorescence detected in wells containing crushed embryonated and unembryonated T. muris eggs.0.02mU of stock chitinase were used as a positive control.(I) Percent of T. muris eggs hatched after incubation with fluid from hatched eggs (purple open circles) and bacteria resuspended in fluid from hatched eggs (purple filled circles).Fluid from unhatched eggs was used as a control (yellow filled and open circles).(J) Percent of T. muris eggs hatched after incubation with a large spectrum protease inhibitor.(K) Graphical depiction of proposed mechanism.Data points and error bars represent means and SEM from 3 independent repeats for (A), (B), (C) and (I)).Dots represent fluorescence detected in a single well and bars and error bars show means and SEM from 2-3 independent repeats for (D), (E), (F), (G) and (H).Bars and error bars show mean and SEM and dots represent percent hatching from a single well from 3 independent repeats for (J).~25 eggs per well were assessed for (A), (B), (C), (I) and (J).Each well contained ~12 eggs for (D) and (E).400 eggs per condition were crushed in (F) and 300 eggs per condition were crushed in (G) and (H).Welch's t test was used to compare area under the curve between mutant S. aureus strains and WT S. aureus for (A), (B) and (C).Kruskal-Wallis test followed by a Dunn's multiple comparisons test was used for (D), (E), (F), (G) and (H).Ordinary one-way ANOVA followed by Tukey's multiple comparisons test was used to compare AUC of different conditions tested for (I).Mann-Whitney test was used for (J).(K) was created using BioRender.com.https://doi.org/10.1371/journal.ppat.1011647.g007membrane of the plug and exit the egg [21].Therefore, it is possible that bacteria-induced plug disintegration plays an important role in hatching by either hollowing out the space in the collar region to accommodate the larval head or by reducing the density of the plug to facilitate movement into the plug space.The difference in electron density we observed between contralateral plugs on the same egg might be indicative of plugs at different stages of degradation, with the less dense plugs being further along in the disintegration process.This model is supported by our comparison of plugs from different eggs, which highlighted the heterogenous degrees of collapse and granularity.
The asymmetrical binding of bacteria to the polar plugs observed could potentially be the cause of different rates of plug degradation and eclosion through one pole.Larval positioning may also be a contributing factor.In SBFSEM images, the head of the larva is located closest to the most degraded plug.The closely related nematode, Caenorhabditis elegans, has been proposed to release enzymes that weaken the eggshell from the head of the larva via pharyngeal pumping during hatching [33,34].Therefore, we speculate that bacterial binding/density triggers larval repositioning and that the plug closest to the head of the larva degrades at a faster rate.This mechanism would also explain why both plugs degrade, despite only one being used for eclosion as degradative enzymes released by the larva would reach both plugs because of diffusion within the perivitelline space.It would be interesting to explore whether similar processes are involved in hatching of other helminths that have polar plugs, such as members of the Capillaria genus [33].Hatching of Capillaria obsignata and Capillaria aerophila eggs also occurs in the bacteria rich intestines of chickens and canines, respectively [35][36][37][38].
The epidemiologic association between S. aureus and Trichuris colonization, and the general absence of information regarding how Gram-positive bacteria mediate hatching, motivated us to investigate hatching of eggs exposed to this bacterium by confocal microscopy.The number of bacterial cells bound to the poles was a critical determinant of hatching efficiency.However, physical association was not sufficient.Although we observed E. faecium bound to egg poles, the two Enterococcus species we tested failed to induce significant amounts of hatching.This result was supported by in vivo data that demonstrated that E. faecalis monocolonized mice, similar to germfree mice, fail to support Trichuris infection above background levels.Given that the microbiome can effect larval development, more work is necessary to confirm that the lack of adult worms was the result of less hatching [12].Also, our finding that bacterial arginine biosynthesis is required for E. coli but not S. aureus calls into question whether a single metabolite is necessary.Our finding that hatching is completely blocked by protease inhibitors implicates proteins, but this role could reflect the impact of protein processing in the physical interaction between bacteria and eggs rather than metabolism.
Given that the polar plug predominantly consists of chitin, we initially considered the possibility that binding led to local production of chitinase derived from bacteria, thereby degrading the pole from the outside.However, hatching remained efficient in the presence of S. aureus mutants lacking candidate enzymes including a homolog of chitinase.Additionally, fluid collected from hatched eggs was unable to induce hatching of unexposed eggs, suggesting that polar plug disintegration is occurring via an "inside-out" mechanism where the larva itself produces the enzymes that facilitate eclosion (Fig 7K).Consistent with this possibility, we detected chitinase within fully mature infective eggs and not in unembryonated eggs.A mechanism by which the larva within the egg is the producer of degradative enzymes would explain why unrelated bacterial taxa can induce hatching-it would be unlikely that all the bacterial species shown to induce hatching possess the same enzyme necessary to dissolve the chitinous plug.
The lack of a genetic system to interrogate T. muris and the impermeable nature of the eggshell pose challenges for elucidating detailed mechanism.Therefore, technical advances may be required to confirm the role for worm-derived chitinase and examine how products released from bacteria relate to this chitinase activity.Although bacteria exposure is not required for the presence of chitinase activity, it may influence the location of the enzymatic activity (Fig 7K).It is possible, for instance, that bacterial exposure triggers the release of chitinase from vesicles within the worm and into the perivitelline space in the egg, as suggested for H. bakeri [39].
Gram-positive species are a prominent component of the gut microbiota and certain taxa are enriched in Trichuris-colonized individuals [17].S. aureus and other staphylococci colonize the gut of humans, especially early in life as the microbiota develops [40,41].It is possible that microbiota composition contributes to the susceptibility of young children to high worm burden and disease.Differences in microbiota composition may also partially explain why helminth infections tend to be aggregated in the host population, with a minority of the population harboring a majority of the worms [42][43][44].Our findings indicate that, although not all bacteria induce hatching of T. muris eggs, those that mediate hatching cause similar structural events to occur at the polar plug.We suggest that targeting enzymatic activity involved in plug disintegration would be difficult if it initiates from within an impermeable eggshell that evolved to be resilient to the external environment.Instead, understanding the chemical and physical interaction between bacteria and eggs may uncover a vulnerability in the life cycle of the parasite that is suitable for intervention.

Experimental model details
Parasite maintenance.Stock eggs of Trichuris muris E strain [14] were propagated and maintained in the NOD.Cg-Prkdc scid /J (Jax) mouse strain as previously described [45].Each egg batch was confirmed to hatch at � 90% in vitro using method below and WT S. aureus before use in subsequent experiments.
Stock eggs of Trichuris trichiura were provided by the Trichuris trichiura egg Production Unit (TTPU) located at the Clinical Immunology Laboratory at the George Washington University.Whipworm eggs were isolated from the feces of a chronically infected human volunteer following a qualified standard procedure that includes a modified Simulated Gastric Fluid (SGF) method.After isolation, the eggs were stored for two months in flasks containing sulfuric acid (H 2 SO 4 ) maintained at 25-30˚C in a monitored incubator.Once embryonation was achieved, the eggs were transferred to a locked and monitored refrigerator at 2-8˚C until further use.Controls for the manufacturing process involved: (i) tests for viability (hatching), which confirmed that more than 80% of the eggs were viable; (ii) species confirmation by polymerase chain reaction (PCR); and (iii) evaluation of the microbiological burden determined by bioburden testing by an outside-certified laboratory.
Bacterial strains.Escherichia coli strain used was a kanamycin resistant transformant of strain BW25113 previously used in our study [12].This strain was chosen because it is the parental strain for a library of single gene knockouts examined in our prior work that served as the foundation for the current study.Staphylococcus aureus strain USA300 LAC clone AH1263 [46], S. epidermidis (ATCC12228) and Enterococcus faecalis OG1RF were provided by V. J. Torres (NYU Grossman School of Medicine) and were originally sourced from Alex Horswill (University of Colorado Anschutz Medical Campus), Eric Skaar (Vanderbilt University) and Lynn Hancock (University of Kansas) respectively.AH1263 was used as wild-type (WT) S. aureus unless otherwise specified.Bacillus subtilis was sourced from the ATCC (ATCC 6633).E. faecium Com15 was kindly provided by Howard Hang (Scripps Research).Frozen glycerol stocks (30% glycerol) of all bacteria were prepared.Glycerol stocks of E. coli was streaked onto Luria Bertani (LB) kanamycin 25 mg/ml plates (NYU Reagent Preparation Core), S. aureus, S. epidermidis and B. subtilis were streaked onto Tryptic Soy Agar (TSA) plates (NYU Reagent Preparation Core) and Enterococcus species were streaked onto bile esculin azide (BEA) plates (Millipore) and incubated overnight at 37˚C.Single colonies of E. coli, Staphylococcus species and Enterococcus species were then spiked into 5mls of LB broth, Tryptic Soy Broth (TSB) and Bacto Brain Heart Infusion (BHI) broth (BD) respectively unless otherwise specified and grown overnight at 37˚C with shaking at 225 rpm.B. subtilis was also spiked into TSB and was grown overnight under the same conditions.To quantify colony forming units, we performed serial dilutions of liquid culture in sterile PBS and plated on LB (Kan) agar for E. coli, TSA for Staphylococcus species and B. subtilis and BEA for Enterococcus species.
S. aureus mutants.The S. aureus ∆aur∆sspAB∆scpAspl::erm mutant, S. aureus ∆gehA∆-gehB∆gehE mutant, S. aureus ∆crp mutant (and its parental strain LAC (ErmS)) and S. aureus GFP strain were all provided by V. J. Torres (NYU Grossman School of Medicine) and were originally sourced from A. Horswill (University of Iowa) [30], Francis Alonzo (University of Illinois at Chicago), Lindsey Shaw (University of South Florida), and the V. Torres lab respectively.S. aureus GFP strain used was a chloramphenicol resistant transformant of strain AH-LAC USA300 transformed with a pOS1 vector containing superfolder gfp with a sarA promoter [47].Frozen glycerol stocks (30% glycerol) of all bacteria were prepared.Glycerol stocks of S. aureus ∆aur∆sspAB∆scpAspl::erm, S. aureus ∆gehA∆gehB∆gehE mutant, S. aureus ∆crp mutant and S. aureus GFP were streaked onto TSA plates with 5 μg/ml of erythromycin, 25 μg/ml Kan/ 25 μg/ml neomycin (Neo), 50 μg/ml Kan/ 50 μg/ml Neo and 10 mg/ml of chloramphenicol respectively (NYU Reagent Preparation Core and V. Torres lab).Single colonies of all species were spiked into 5mls of TSB and grown overnight at 37˚C with shaking at 225 rpm.To quantify colony forming units, we performed serial dilutions of liquid culture in sterile PBS and plated on TSA plates.
Germ-free mice.Germ-free (GF) C57BL/6J mice were bred and maintained in flexiblefilm isolators at the New York University Grossman School of Medicine Gnotobiotics Animal Facility.Absence of fecal bacteria was confirmed monthly by evaluating the presence of 16S DNA in stool samples by qPCR as previously described [48].Mice were transferred into individually ventilated Tecniplast ISOcages for infections to maintain sterility under positive air pressure.Female mice 6-11 weeks of age were used for all experiments in this study.All animal studies were performed according to protocols approved by the NYU Grossman School of Medicine Institutional Animal Care and Use Committee (IACUC) and Institutional Review Board.

Method details
In vitro hatching of T. muris eggs.T. muris eggs were hatched in vitro by mixing 25 μL of embryonated eggs at a concentration of 1 egg/1 μL suspended in sterile water with 10 μL of overnight bacterial culture and 15 μL sterile media in individual wells of a 48 well plate.Media control wells contained an additional 10 μl of sterile media instead of bacterial culture.Plates were incubated at 37˚C and observed every 10 mins, 30 mins and/or hour for 4-5 hrs on the Zeiss Primovert microscope to enumerate the percentage of hatched eggs by counting hatched and embryonated unhatched eggs in each well.Unembryonated eggs, which lack visible larvae and have disordered contents, were excluded due to their inability to hatch.For anaerobic hatching specifically, plates were incubated at 37˚C in an anaerobic chamber and a separate 48-well plate was used per timepoint as the plate needed to be removed from the anaerobic chamber to count hatching each timepoint.Experiments utilizing transwell inserts (Millicell) in Fig 2 were performed as previously described [11].Subcultures were used for transwell

PLOS PATHOGENS
Whipworm egg structure during bacteria-mediated hatching experiments and were made by diluting overnight cultures 1:100 in either TSB for S. aureus, LB for E. coli or BHI for E. faecalis and then incubating at 37˚C while shaking at 225 rpm until 1 × 10 8 CFU/mL.For transwell experiments involving the addition of bacteria to the inner well, 80 μL of either bacterial subculture or media were added to the transwell.For experiments where cell-free supernatant was used, supernatant and cells were isolated by centrifugation and filtration through a 0.22 μm syringe filter or after wash with autoclaved PBS, respectively.Incubation with embryonated eggs was performed by adding 30 μL T. muris eggs to 5 mL of bacterial culture and incubating at 37˚C for four hours.E. coli and E. faecalis were concentrated by centrifuging 4 ml of each and then resuspending the pellets in 400 μl of their respective media.For the co-incubation experiment, 2 ml of S. aureus overnight culture was added to 2 ml of E. faecalis overnight culture and then the mixture was added to eggs.
T. muris in vivo infection in germ-free mice.Female germ-free C57BL/6J mice were monocolonized at 6-11 weeks of age by oral gavage with *1 × 10 9 colony forming units (CFU) of S. aureus or E. faecalis.Subcultures were made by diluting overnight cultures 1:100 in either TSB for S. aureus or BHI for E. faecalis and then incubating at 37˚C while shaking at 225 rpm for ~4 hrs, until 1 × 10 9 CFU/mL was reached.5ml of subcultures were pelleted by centrifugation at 3480 rpm for 10 min and washed once with sterile 1x PBS.Pellets were then resuspended in 500 μL sterile 1x PBS and mice were inoculated by oral gavage with 1x10 9 CFU in a volume of 100 μL.Inoculum was verified using dilution plating of aliquots.
7 and 28 days later, mice were infected by oral gavage with *100 embryonated T. muris eggs.14 days after the second gavage of T. muris eggs, mice were euthanized, and individual worms were collected and enumerated from the cecal contents of mice.Successful monocolonization with bacteria was determined by plating serial dilutions of stool collected from mice and assessing the number of CFUs per gram of stool on TSA plates for S. aureus monocolonized mice and BEA plates for E. faecalis monocolonized mice.
Assessment of prevalence of S. aureus in gut microbiota of helminth colonized people.Sample collection, preparation and genomic sequencing were performed by Tee et al. as described in their publication [23].From the Kraken2 relative abundance tables generated in this study, we determined the percentage of reads that were mapped to S. aureus in these samples.Percentages were then plotted and a statistical analysis was performed on the data.

Electron microscopy
Scanning electron microscopy.T. muris eggs were incubated with S. aureus or E. faecalis for 1hr, E. coli for 1.5 hr or bacteria free media for 1 hr.All eggs were fixed with 2% paraformaldehyde (PFA), 2.5% glutaraldehyde in 0.1 M sodium cacodylate buffer (CB, pH 7.2) at 4˚C for a week.~50 eggs were then loaded into 12 mm, 0.44 mm transwells (#PICM01250, Millipore Sigma) to avoid losing eggs during sample processing.The eggs were washed 3 times with PBS, post fixed with 1% osmium tetroxide (OsO 4 ) in aqueous solution for 1 hour, then dehydrated in a series of ethanol solutions (30%, 50%, 70%, 85%, 95%) for 15 mins each at room temperature.Eggs were then dehydrated with 100% ethanol 3 times for 20 mins each.The eggs were critical point dried using the Tousimis Autosamdri-931 critical point dryer (Tousimis, Rockville, MD), put on the SEM stabs covered with double sided electron conducted tape, coated with gold/palladium by the Safematic CCU-010 SEM coating system (Rave Scientific, Somerset, NJ), and imaged with the Zeiss Gemini300 FESEM (Carl Zeiss Microscopy, Oberkochen, Germany) using secondary electron detector (SE 2 ) at 5 kV with working distance (WD) 18.3 mm.
Serial block face-scanning electron microscopy.T. muris eggs were incubated with S. aureus, E. coli or bacteria-free media and fixed as described above.For further sample processing, we adopted a previously described protocol [49] and made modifications based on personal communications with Rick Webb (Queensland University) using a PELCO Biowave (Ted Pella Inc., Redding, CA) for microwave processing.The detailed sample processing steps are listed in table 1.Samples were embedded with Spurr resin (Eletron Microscopy Sciences, Hatfield, PA) between ACLARE using a sandwich method.
For SBF-SEM imaging, the sample block was mounted on an aluminum specimen pin (Gatan, Pleasanton, CA) using silver conductive epoxy (Ted Pella Inc.) to electrically ground the block.The specimen was trimmed again and sputter coated with 10 nm of gold (Rave Scientific, Somerset, NJ).Serial block face imaging was performed using a Gatan OnPoint BSE detector in a Zeiss Gemini 300 FESEM equipped with a Gatan 3View automatic microtome unit.The system was set to cut sections with 100 nm thickness, imaged with a pixel size of 12 nm and a dwell time of 3.0 μs/pixel, with each frame sized at 50 x 90 μm.SEM beam acceleration was set at 1.5 keV and Focus Charge Compensation gas injection set at 12% (7.9E-04mBar) to reduce charging artifacts.Images of the block face were recorded after each sectioning cycle with a working distance of 6.6 nm.Data acquisition and sectioning were automatically controlled using Gatan Digital Micrograph software to manage imaging parameters.A stack of 250 slices was aligned and assembled using ImageJ.Semi-automated segmentation and video renders were generated with ORS Dragonfly 4.1 (Object Research Systems, Montre ´al, QC).

PLOS PATHOGENS
Whipworm egg structure during bacteria-mediated hatching

Optical microscopy
Live imaging.20 μL of overnight S. aureus-GFP culture were added to 50 μL of embryonated eggs (1 egg/1 μL) and 30 μL sterile media in a 35 mm Petri dishes with No. 1.5 coverglass at the bottom.Mixture was then imaged using a Nikon 40x N.A. 1.3 oil immersion objective lens on a Nikon Eclipse Ti microscope.An environmental chamber set at 37˚C and a lens heater set at 45˚C were both used to observe hatching for approximately 45 minutes.
Binding assay.An in vitro hatching assay was prepared as described above using S. aureus-GFP.After incubating the plate at 37˚C, eggs with bacteria were fixed using 4% PFA, 0.5% glutaraldehyde in PBS at room temperature for 1 hr.Plate contents were then transferred to 1.5 ml Eppendorf tubes.Eggs in this mixture were pelleted by centrifugation at 1000 rpm for 5 mins with a slow brake speed (4).Samples were then washed twice with PBS and then resuspended in PBS for imaging.
Fixed samples were imaged on 35 mm Petri dishes with No. 1.5 coverglass at the bottom using a Nikon 60x N.A. 1.4 oil immersion objective lens on a Nikon Eclipse Ti microscope.Number of puncta on the sides and on the poles were enumerated by eye.
UV killing of S. aureus 2 ml overnight of S. aureus culture were added to 8 ml of sterile PBS in a 150mm Petri dish.The dish was then placed uncovered in a UV Stratalinker 2400 (Stratagene) for 1 hr.Killed bacteria were then transferred to a 15 ml conical tube and then centrifuged at 3480 rpm and 4˚C for 10 mins.Pellets were then resuspended in different amounts of sterile TSB to obtain bacterial suspensions of different concentrations that were then used in an in vitro hatching assay as described above.

Chloramphenicol treatment of S. aureus
Overnight cultures of S. aureus were centrifuged at 3480 rpm and 4˚C for 10 mins, resuspended in fresh, sterile TSB and then incubated with 100 μg/ml of chloramphenicol or 4% ethanol (vehicle control) on ice for 45 mins as previously described [27].Treated bacteria was then used in an in vitro hatching assay described above.After 2 hrs of incubation, 10 μl of untreated chloramphenicol resistant S. aureus (S. aureus-GFP) was added to the wells of the hatching assay and hatching was monitored for 2 more hours.

Chitinase detection assay
In vitro hatching assay was performed in a 96-well black-walled, clear, flat-bottom plate by adding 12.5 μl of eggs to 7.5 μl of sterile media and 5 μl of overnight bacterial culture and incubating the plate at 37˚C for 4 hrs.6.25 μl of 4-Methylumbelliferyl β-D-N,N',N'-triacetylchitotrioside hydrate (4MeUmb, Millipore Sigma) were then added to hatching assay wells as well as an additional well containing 25 μl of chitinase from Streptomyces griseus (positive control, Millipore Sigma).After incubating at 37˚C for 1 hr, reaction was stopped by adding 6.25 μl of 1M glycine/NaOH (NYU Reagent Preparation Core) to all wells.Plate was then read at 355/ 450nm ex/em by using a SpectraMax M3 plate reader (Molecular Devices) [32].
Chitinase activity within eggs exposed to bacteria was measured by first preparing an in vitro hatching assay in a 48-well plate as described above and incubating it at 37˚C for 20 mins.16 wells per condition were prepared.Well contents for each condition were then pooled together into 1.5 ml Safe-lock tubes (Eppendorf) with 1.0 mm diameter glass beads (BioSpec Products).Tubes were then homogenized by performing 4, 20 s homogenization cycles at 4.5

PLOS PATHOGENS
Whipworm egg structure during bacteria-mediated hatching m/s using a bead beater.25 μl of homogenate was then added to a 96-well black-walled, clear, flat-bottom plate and chitinase detection assay was performed as described above.
Chitinase activity within Trichuris eggs that were not exposed to bacteria was measured by first centrifuging equal numbers of eggs (500g, 3 mins, brake = 3 for T. trichiura and 4000rpm, 5 mins, brake = 4 for T. muris) and then resuspending them in sterile TSB.300 eggs were then homogenized and chitinase detection assay performed as described above.In both homogenization experiments, chitinase from Streptomyces griseus (Millipore Sigma) was also homogenized as a positive control.

Determining effect of hatching fluid on eggs
An in vitro hatching assay was prepared in a 48-well plate as described above and incubated at 37˚C for 4 hrs.36 wells per condition were prepared.Well contents for each condition were pooled together, centrifuged at 3480 rpm for 10 mins, and then supernatant was filtered as described above and added to new eggs in another in vitro hatching assay.Hatching was measured over time.
Additional S. aureus overnight cultures were also centrifuged at 3480 rpm for 10 mins and the remaining pellet was resuspended in the same filtered hatching assay fluid obtained above.Resuspended S. aureus was then added to new eggs in another in vitro hatching assay and hatching was measured over time.

Determining effect of protease activity on hatching
Inhibition of protease activity was performed using the cOmplete ULTRA Tablets, Mini, EASYpack Protease Inhibitor Cocktail (Roche) as previously described [52].

Identification of bacteria associated with eggs
10 μl of T. muris eggs were dropped onto LB plates (NYU Reagent Preparation Core) and incubated at 37˚C overnight.Colonies that grew were then picked and added to Eppendorf tubes containing GoTaq G2 Hot Start Green Master Mix (Promega).DNA was extracted by heating the mixture at 95˚C and the 16S rRNA gene was then amplified using 16S-Fwd (CCGATAT CTCTAGAAGAGTTTGATCCTGGCTCAG) and 16S-Rev (CCGATATCGGATCCACGGT TACCTTGTTACGACTT) primers in a PCR reaction.PCR product was sent to Macrogen for sequencing and the sequences were aligned to the EzBioCloud 16S rRNA database for initial taxonomic identification [17,53].

Quantification and statistical analysis
The number of repeats per group is annotated in corresponding figure legends.Significance for all experiments was assessed using GraphPad Prism software (GraphPad).Specific tests are

PLOS PATHOGENS
Whipworm egg structure during bacteria-mediated hatching annotated in corresponding figure legends.p values are also annotated on the figures themselves.A p value of ns or no symbol = not significant.

Fig 1 .
Fig 1. Bacteria display species-dependent effects on T. muris egg hatching.(A) Representative light microscopy image of T. muris eggs induced to hatch after incubation with S. aureus at 37˚C for 45 mins.White arrowhead denotes unhatched egg and black arrowhead denotes hatched egg.(B) Percent of T. muris eggs hatched after incubation in aerobic conditions with overnight cultures of E. coli and S. aureus, compared with their respective broth controls determined by light microscopy at indicated time points.Colony forming units (CFUs) of each bacterial species added to the eggs are indicated on the graphs.(C) Percent of T. muris eggs hatched at 1 hour after 37˚C

Fig 2 .
Fig 2. Physical contact between the bacterial cell and egg is essential for S. aureus mediated hatching.(A) Percent of T. muris eggs hatched after incubation with total overnight S. aureus culture, culture pellet resuspended in fresh media, culture supernatant filtered with a 0.22μm filter, or media control.(B) Experimental approach for determining whether a soluble hatching inducing factor is produced by S. aureus in response to exposure to eggs.Filtered supernatant from S. aureus grown with or without T. muris eggs for 4 h were transferred to a dish containing T. muris eggs.The ability of the supernatant to mediate egg hatching was evaluated over 4 hours.(C) Percent of T. muris eggs hatched after incubation with total overnight S. aureus culture or filtered supernatant obtained from S. aureus incubated with or without eggs as in (B) compared with media controls.(D) Percent of T. muris eggs hatched when placed in a 0.4μm transwell separated from bacteria or control media in the outer well compared with eggs incubated bacteria without transwell separation.Data points and error bars represent the mean and SEM from 3 independent experiments for (A) and (C).~25 eggs per well were assessed for (A) and (C).Dots represent a single well containing ~200 eggs and bars show means and SEM from 3 independent experiments for (D).Ordinary one-way ANOVA test followed by a Tukey's multiple comparisons test was used to compare resuspended pellet and supernatant conditions with total O/N culture for (A) and (C).Two-way ANOVA followed by a Tukey's multiple comparisons test was used for (D).(B) was created using BioRender.com.https://doi.org/10.1371/journal.ppat.1011647.g002

Fig 3 .
Fig 3. Collapse of the polar plug precedes hatching mediated by bacteria.(A, B, C, D) Representative low (left) and high magnification (right) SEM images of T. muris eggs (clear arrowhead) that were exposed to S. aureus for 1 hour (A), E. coli for 1.5 hours (B) and E. faecalis for 1 hour (C) or untreated (D).White arrowheads correspond to bacteria on polar plug regions of the eggs denoted by black arrowheads.Yellow arrow in right panel of (A) and (B) indicates woolly substance present among bacteria.(E) Representative low (left) and high magnification (right) SEM images of hatching T. muris eggs (clear arrowhead) that were exposed to S. aureus for 1 hour (top) and E. coli for 1.5 hours (bottom).White arrowheads correspond to bacteria on polar plug regions of the eggs.Emerging larvae are denoted by white diamonds.(F) Number of bacterial cells visible on polar plugs of T. muris eggs incubated with E. coli, S. aureus, or E. faecalis.Bars showing mean from 2 eggs per condition.(G) Width of collar openings on T. muris eggs that were treated with either E. coli or S. aureus and were either unhatched or in the process of hatching.For low magnification images, scale bar represents 2μm for unhatched egg in (A), (B), (C) and (D) and 10μm for hatched egg in (A) and (B).For high magnification images, scale bar represents 1μm for (A), (B) and (D) and 2μm for (C).Dots represent a single plug and bars show mean and SEM of collar sizes from 4-7 eggs per condition for (G).Two-way ANOVA followed by a Tukey's multiple comparisons test was used for (G).4-10 eggs were imaged per condition.https://doi.org/10.1371/journal.ppat.1011647.g003

Fig 4 .
Fig 4. Bacteria induce asymmetric disintegration of the polar plug.(A) Representative electron micrograph of a longitudinal cross-section of a T. muris egg exposed to S. aureus for 1 hour.Image shows pole associate bacteria (B), granules containing lipids (LG) within the larva, eggshell (ES) and larval cuticle (C).Distances measured are also indicated.The width of the collar opening (i), width of the widest part of the plug (ii), the height of the plug (distance from the top of the collar to the top of the plug) (iii) and the eggshell thickness (iv) were measured for all conditions.Insets show the regions boxed in black dotted line.(B, C) Representative electron micrographs and 3D reconstructions from SBFSEM data of a T. muris egg exposed to S. aureus for 1 hour (B) and E. coli for 1.5 hours (C).For the longitudinal section, the original micrograph (left), the micrograph with color overlays indicating segmented egg components and bacteria (middle) and 3D reconstruction of the egg (right) are shown.For the complete structure, two different angles are shown (left and right).S. aureus (red), E. coli (blue), polar plugs (green), eggshell (yellow) and larvae (purple) are all shown.Scale bars represent 10μm.(D) High magnification images of polar plugs from (B) and (C) on T. muris eggs exposed to S. aureus and E. coli compared with equivalent regions from an egg untreated with bacteria.Outer vitelline layer is denoted by white arrowheads, eggshell is denoted by black arrowheads and associated bacterial cells are denoted by red arrows.(E) 3D reconstruction of polar plugs on T. muris eggs exposed to E. coli.Multiple angles are shown (left to right). 2 eggs each were imaged for all 3 conditions.https://doi.org/10.1371/journal.ppat.1011647.g004 , S2 Video).3D renderings generated from the data collected showed that both S. aureus and E. coli were sporadically present over the entire surface of the egg and enriched at the plugs (Fig 4B and 4C respectively, right; S2 Video).In the case of S. aureus exposed eggs, a large aggregate of a characteristic grape-like cluster of cocci was visible on one of the poles, versus the shell where most bacteria were present as either single cells or in pairs (Fig 4B, far right).

Fig 5 .
Fig 5. Degree of bacterial binding is directly proportional to the rate of T. muris egg hatching.(A) Representative confocal microscopy image of T. muris eggs incubated with S. aureus GFP for 30 minutes.Regions defined as poles and sides are indicated.Scale bar represents 34μm.(B) Percent of GFP+ puncta present on the poles versus the sides of the T. muris eggs.(C) Percent of T. muris eggs hatched after incubation with 10-fold dilutions of overnight S. aureus culture ranging from approximately 10 5 -10 8 CFU.(D) Percent of T. muris eggs associated with GFP+ puncta after incubation with 10-fold dilutions of overnight S. aureus-GFP culture ranging from approximately 10 5 -10 8 CFU.(E) Correlation analysis comparing log 10 CFUs of bacteria used and percent of T. muris eggs with GFP+ puncta present.(F) Number of GFP+ puncta bound per egg from (D). (G) Percent of GFP+ puncta present on the poles of the eggs versus the sides of the eggs from (D). (H) Percent of pole associated GFP+ puncta present on the higher bound pole versus the lower bound pole of eggs from (D).Bars and error bars show means and SEM from 3 independent repeats for (B), (D), (F),(G) and (H).Data points and error bars represent mean and SEM from 3 independent repeats for (C).~25 eggs per well were assessed for (C).Dots represent percentage of 40 eggs that were GFP+ from a single experiment for (D) and (E) and number of GFP+ puncta found on a single egg for (F).Mann-Whitney test was used for (B).Kruskal-Wallis test followed by a Dunn's multiple comparisons test was used for (D) and (F).Simple linear regression was performed for (E).Two-way ANOVA followed by a Sidak's multiple comparisons test was used for (G) and (H).https://doi.org/10.1371/journal.ppat.1011647.g005

Fig 6 .
Fig 6.Bacterial metabolic activity is essential for S. aureus mediated hatching.(A) Percent of T. muris eggs hatched after incubation with 10 8 CFU overnight S. aureus culture at time points indicated.(B) Percent of T. muris eggs that had GFP+ puncta bound after incubation with 10 8 CFU of overnight S. aureus culture for different lengths of time.(C) Number of GFP+ puncta bound per T. muris egg after incubation with 10 8 CFU of overnight S. aureus culture for different lengths of time.(D) Percent of GFP+ puncta present on the poles of the T. muris eggs versus the sides of the eggs after incubation with 10 8 CFU of overnight S. aureus culture for different lengths of time.(E) Percent of T. muris eggs hatched after incubation with UV killed overnight S. aureus culture.(F) Percent of T. muris eggs hatched after incubation with untreated S. aureus, chloramphenicol (CAM)-treated S. aureus (100μg/ml), and CAM-treated S. aureus together with CAM-resistant S. aureus GFP for 4hrs.S. aureus GFP was spiked-in at the 2hr time point.(G) Percent of T. muris eggs hatched when placed in a 0.4μm transwell with either S. aureus (SA), E. faecalis (EF), media (BHI) or nothing, with either SA or just media (TSB) in the outer well.(H) Percent of T. muris eggs hatched after incubation with overnight S. aureus WT and S. aureus argH::Tn.Bars and error bars show means and SEM from 3 independent repeats for (B), (C), and (D).Data points and error bars represent mean and SEM from 3 independent repeats for (A), (E) and (F).Dots represent mean percentage of 40 eggs that were GFP+ from a single experiment for (B) and number of GFP+ puncta found on a single egg for (C).Dots represent a single well and bars show means and SEM from 3 independent experiments for (G).Bars and error bars show means and SEM and dots represent percent hatching from a single well from 2 independent repeats for (H).~25 eggs per well were assessed for (A), (E), (F) and (H) and ~200 eggs per well were assessed for (G).Kruskal-Wallis test followed by a Dunn's multiple comparisons test was used for (B), (C) and (G).Two-way ANOVA followed by a Sidak's multiple comparisons test was used for (D).Ordinary one-way ANOVA followed by Tukey's multiple comparisons test was used to compare AUC of different conditions tested for (F).Mann-Whitney test was used for (H).

Fig 7 .
Fig 7. Trichuris eggs harbor chitinase activity.(A-C) Percent of T. muris eggs hatched after incubation with overnight S. aureus WT, S. aureus ∆crp (A), S. aureus ∆aur∆sspAB∆scpAspl::erm (B) and S. aureus ∆gehA∆gehB∆gehE (C) culture.(D, E) Amount of fluorescence detected in wells containing T. muris eggs hatched in response to S. aureus (D) and E. coli (E).AU = arbitrary unit.5mU of stock chitinase were used as a positive control.(F) Amount of fluorescence detected in wells containing crushed T. muris eggs that were exposed to S. aureus for 20 minutes.0.02mU of stock chitinase were used as a positive control.(G) Amount of fluorescence detected in wells containing crushed T. muris and T.