Skin-penetrating nematodes exhibit life-stage-specific interactions with host-associated and environmental bacteria

Abstract

Background: Skin-penetrating nematodes of the genus Strongyloides infect over 600 million people, posing a major global health burden. Their life cycle includes both a parasitic and free-living generation. During the parasitic generation, infective third-stage larvae (iL3s) actively engage in host seeking. During the free-living generation, the nematodes develop and reproduce on host feces. At different points during their life cycle, Strongyloides species encounter a wide variety of host-associated and environmental bacteria. However, the microbiome associated with Strongyloides species, and the behavioral and physiological interactions between Strongyloides species and bacteria, remain unclear.

Results: We first investigated the microbiome of the human parasite Strongyloides stercoralis using 16S-based amplicon sequencing. We found that S. stercoralis free-living adults have an associated microbiome consisting of specific fecal bacteria. We then investigated the behavioral responses of S. stercoralis and the closely related rat parasite Strongyloides ratti to an ecologically diverse panel of bacteria. We found that S. stercoralis and S. ratti showed similar responses to bacteria. The responses of both nematodes to bacteria varied dramatically across life stages: free-living adults were strongly attracted to most of the bacteria tested, while iL3s were attracted specifically to a narrow range of environmental bacteria. The behavioral responses to bacteria were dynamic, consisting of distinct short- and long-term behaviors. Finally, a comparison of the growth and reproduction of S. stercoralis free-living adults on different bacteria revealed that the bacterium Proteus mirabilis inhibits S. stercoralis egg hatching, and thereby greatly decreases parasite viability.

Conclusions: Skin-penetrating nematodes encounter bacteria from various ecological niches throughout their life cycle. Our results demonstrate that bacteria function as key chemosensory cues for directing parasite movement in a life-stage-specific manner. Some bacterial genera may form essential associations with the nematodes, while others are detrimental and serve as a potential source of novel nematicides.

Keywords: Bacteria; Chemosensation; Parasitic nematode; Sensory behavior; Strongyloides.

Fig. 1

Strongyloides species encounter bacteria at specific points of their life cycle. The life cycles of the skin-penetrating gastrointestinal nematodes S. stercoralis and S. ratti consist of both a parasitic generation and a free-living generation. Developmentally arrested infective third-stage larvae (iL3s) search the environment for a host to infect. Once infection occurs via skin penetration, they develop into 4th stage larvae (L4s) and eventually into parasitic adults within the small intestine of the host. Parasitic adults reproduce asexually and their progeny exit the host in feces. Some of the population develops on feces through the 1st–4th larval stages (L1–L4) and then into free-living adults; the free-living adults reproduce sexually and their progeny develop into iL3s. The rest of the population develops through the 1st–2nd larval stages and then directly into iL3s. S. stercoralis uniquely can undergo autoinfection, whereby the progeny of the parasitic adults develop directly into iL3s within the host. Icons indicate environmental and host-associated bacterial niches the parasites encounter throughout their life cycle. The free-living generation and pre-iL3 life stages encounter fecal/gut bacteria; the iL3s encounter fecal/gut bacteria, host skin bacteria, and other environmental bacteria; and the parasitic life stages that exist inside the host may encounter host gut and fecal bacteria

Fig. 2

S. stercoralis free-living adults are associated with a specific microbiome. a Schematic of the experimental design. Gerbils were infected with S. stercoralis iL3s, and infested feces were harvested from the gerbils on days 14–27 post-infection. Collected feces were made into fecal-charcoal plates and incubated at either 20 °C for 2 days to obtain S. stercoralis free-living adults, or 23 °C for 7 days to obtain S. stercoralis iL3s. Nematodes were then isolated from the fecal-charcoal plates using a Baermann apparatus and washed in buffer before DNA extraction. Sequencing samples included DNA isolated from nematodes; wash buffer control 1, consisting of the buffer used to wash the nematodes prior to DNA extraction; and wash buffer control 2, consisting of the buffer supernatant after washing the nematodes. b–e Results from the amplicon sequencing analysis for Experiments 1 and 2. b, c Principal coordinates analysis (PCoA) of the different sample categories from Experiments 1 (b) and 2 (c). Free-living adult samples clustered separately from the other samples, suggesting a specific microbiome associated with S. stercoralis free-living adults. In contrast, some of the iL3 samples in Experiment 1 clustered with the control samples. In (b), an empty well control was also included. Ellipses representing the 95% confidence region for the sample groups with more than 3 samples were calculated using the ggplot2 and ellipse packages [73]. d, e Relative abundance of ASVs that showed a significant difference across sample categories in Experiments 1 (d) or 2 (e). Box plots are standard Tukey representations. ASVs with an ANCOM W score in the top 40% of all tested features were selected as significant; individual W scores are displayed below the ASV identifiers. ASV identifiers in bold are those that showed a significant enrichment in S. stercoralis free-living adults vs. controls in both sequencing experiments. ASVs in the genera Escherichia-Shigella, Lactobacillus, and Solibacillus were significantly enriched in the free-living adult samples relative to the other samples

Fig. 3

Strongyloides species display life-stage-specific bacterial preferences. a The bacterial panel used to examine the interactions of Strongyloides iL3s and free-living adults with bacteria. The bacteria are categorized according to the major environmental niches where they are likely to interact with Strongyloides species (Additional file 6: Table S1), although we note that some of the bacteria are also found more broadly in the environment. Categories (left to right): skin, fecal/gut, environmental/other. b S. stercoralis free-living adults were robustly attracted to most bacteria tested in a bacterial chemotaxis assay. n = 20–40 trials for each condition, with 75–150 worms per trial. *p< 0.05, ****p< 0.0001, ns = not significant, Kruskal-Wallis test with Dunn’s post-test. c S. stercoralis iL3s did not respond to most bacteria but were attracted to one of the environmental/other bacterial species tested. n = 20–30 trials for each condition, with 300-400 worms per trial. ****p< 0.0001, Kruskal-Wallis test with Dunn’s post-test. d S. ratti free-living adults were robustly attracted to most of the bacterial species tested in a bacterial chemotaxis assay. n = 20–34 trials for each condition, with 75–150 worms per trial. **p< 0.01, ****p< 0.0001, ns = not significant, Kruskal-Wallis test with Dunn’s post-test. e S. ratti iL3s did not respond to most bacteria but were attracted to two environmental/other bacterial species tested. n = 20–26 trials for each condition, with 300–400 worms per trial. *p< 0.05, ***p< 0.001, Kruskal-Wallis test with Dunn’s post-test. Each bacterial species was compared to the LB control; only significant differences are noted. Graphs show the chemotaxis indices for each trial (points), medians (solid lines), and interquartile ranges (dashed lines). Bacteria are color-coded according to the legend shown on the right

Fig. 4

S. stercoralis free-living adults show distinct short-term behavioral responses to bacteria. a Overlaid tracks of individual S. stercoralis free-living adults in a single-worm bacterial chemotaxis assay. Worms were tracked for 20 min or until they left the assay arena. Black crosses represent starting points. For each assay arena, the red circle (left) depicts the experimental zone containing bacteria and the black circle (right) depicts the control zone containing the media control. In the case of the LB control assay, both the experimental zone and the control zone contained LB media. b Worms spent more time in the experimental zone when the experimental zone contained E. coli. **p< 0.01, Kruskal-Wallis test with Dunn’s post-test. The only significant difference is noted. c Worms spent more time in the E. coli lawn than they did in the P. fluorescens lawn. Only animals that reached a bacterial lawn were included in this analysis. d Worms spent more time in the control zone when the experimental zone contained P. fluorescens. *p< 0.05, Kruskal-Wallis test with Dunn’s post-test. The only significant difference is noted. e Worms displayed similar navigational patterns regardless of the bacteria present, as quantified by the distance ratio (total pathlength ÷ linear displacement from the initial point to the final point). No significant differences were detected (Kruskal-Wallis test). n = 24–30 worms for each condition. Graphs show medians (solid lines) and interquartile ranges (dashed lines). All statistical comparisons are relative to the LB control assay

Fig. 5

Fecal/gut bacteria influence S. stercoralis physiology. a Survival of S. stercoralis free-living adult females cultured on either E. coli, E. fergusonii, or P. mirabilis. Worms were placed on plates containing the indicated bacteria as young adults, and percent survival was monitored daily. n = 19 worms for each condition. **p< 0.01, log-rank test with Bonferroni post-test comparing each condition to every other condition. The only significant difference is noted. Error bars show standard error. b Culturing S. stercoralis free-living adults on different fecal bacteria did not affect the number of eggs laid per day. No significant effect of bacteria was detected (two-way repeated measures ANOVA). n = 13–17 worms for each condition. Medians and interquartile ranges are shown. c Culturing S. stercoralis free-living adults on P. mirabilis resulted in fewer hatched eggs per day. Graph shows the percentage of eggs that hatched every 24 h after free-living young adult females were allowed to mate on the indicated bacteria. *p< 0.05, **p< 0.01, ****p< 0.0001, ns = not significant, two-way repeated measures ANOVA with Tukey’s post-test. Statistical significance for each comparison is noted above the graph. n = 12-16 worms for each condition. Medians and interquartile ranges are shown. d P. mirabilis decreased egg hatching. Graph shows the percentage of hatched eggs 48 h after free-living young adult females were allowed to mate on the bacteria. ****p< 0.0001, ns = not significant, Brown-Forsythe and Welch ANOVA with Dunnett’s T3 post-test comparing each condition to every other condition. n = 13–14 worms for each condition. Medians (solid lines) and interquartile ranges (dashed lines) are shown. e There was no significant difference in the percentage of hatched eggs after 48 h when older, gravid free-living adult females were placed on P. mirabilis (unpaired, two-tailed Welch’s t test). n = 12 worms for both conditions. Medians (solid lines) and interquartile ranges (dashed lines) are shown

Fig. 6

P. mirabilis impairs both egg and larval development in S. stercoralis. a Representative images of S. stercoralis embryonic development, a healthy pre-infective larva, and an unhealthy pre-infective larva. Egg development was classified into the categories indicated based on the classifications used for C. elegans [29]. Scale bars = 20 μm. b The percentage of unhatched eggs that fell into each category of embryonic development for worms cultured on either P. mirabilis or E. coli. For each experiment, 1 female and 3 males were cultured on the bacterial lawn for 24 h, after which the worms were removed from the plate and the percentage of eggs in each category was scored. The percentage of eggs in each category was then scored again after 48 h. After 24 h, more undeveloped eggs were observed on P. mirabilis relative to E. coli, suggesting that P. mirabilis interferes with the earliest stages of egg development (left). **p< 0.01, ns = not significant, two-way ANOVA with Sidak’s multiple comparisons post-test. n = 9–10 trials for each condition. After 48 h, nearly all of the eggs on the E. coli plates had hatched; the remaining eggs were almost all undeveloped. In contrast, approximately half of the eggs on P. mirabilis were at the pretzel stage of development while the other half were undeveloped, indicating that culturing on P. mirabilis delays or prevents egg hatching in some eggs that have initiated embryonic development (right). ***p< 0.001, ns = not significant, two-way ANOVA with Sidak’s multiple comparisons post-test. n = 6-10 trials for each condition. c. The percentage of the total progeny consisting of undeveloped eggs, developing eggs, healthy larvae, or unhealthy larvae from the experiment described in (b) after 24 h (left) or 48 h (right). Culturing on P. mirabilis resulted in an increased proportion of eggs that failed to develop and of the worms that hatched, an increased number of unhealthy larvae. Larvae that were classified as unhealthy appeared non-motile and generally remained in a curled position. *p< 0.05, ****p< 0.0001, two-way ANOVA with Tukey’s multiple comparisons test. Comparisons are between the same category and time point; only significant differences are shown. n = 7–10 trials for each condition. Median percentages are represented, and bars represent interquartile ranges. Data in (b) and (c) are from the same experiments