Background

Trichuris muris and T. suis, whipworms of mice and pigs, respectively, invade the mucosa of the large intestine [1–3]. The life cycle is direct with embryonated eggs (each containing a fully developed larva) as infective stage [1, 3]. The embryonated eggs of T. suis (TSO) represent a raw material of the Active Pharmaceutical Ingredient (API) in a medicinal product, which is currently being explored as a treatment for patients with immune-mediated diseases. Such helminthic therapy is founded on the hygiene hypothesis [4] or more broadly, the ‘Old friends’ hypothesis [5, 6]. In more details, a positive immunomodulatory effect of TSO has been demonstrated in patients with inflammatory bowel disease [7–9] and multiple sclerosis [10], although this positive immunomodulatory effect was not demonstrated in a safety and efficacy study in patients with multiple sclerosis [11]. The treatment with TSO is well-tolerated in patients and only transient gastrointestinal symptoms are present following first dosing [10–15].

A first requirement to use TSO as a medicinal drug is that the eggs are biologically active, i.e. that the eggs are able to hatch for the release of larvae. Description of eggs based solely on their morphological development and appearance of larvae inside does not necessarily correlate with the egg hatchability, as larvae may degenerate over an extended period of storage.

In recent years, individual bacterial species were successfully used for hatching T. muris eggs under in vitro conditions [16–18]. However, whether T. suis eggs share the same egg hatching stimulation as T. muris eggs is still unknown. Therefore, in the present study a slight modification of a published method based on Gram-negative Escherichia coli (BL-21) for hatching T. muris eggs [16] was employed to test whether T. muris and T. suis eggs show different ability to hatch when incubated under the same conditions. Further, eggs of the two Trichuris species were incubated with five Gram-positive bacteria.

Here we report how two very closely related nematodes (same genus) respond differently to in vitro hatching when incubated with selected Gram-negative and Gram-positive bacteria.

Methods

Parasite batches

Eggs of T. muris were isolated from faeces of experimentally infected female mice (C57BL/10ScSnOlaHsd) and were embryonated for 3 months at room temperature (23–24 °C) in Milli-Q (ultrapure water purified by filtration and deionization) and in sulphuric acid (H2SO4) pH1, respectively.

Eggs of T. suis were isolated from faeces of experimentally infected female Göttingen minipigs (Ellegaard Göttingen Minipigs A/S, Dalmose, DK). The eggs were embryonated in H2SO4 pH1 for 100 days (±10 days) at 25 °C and then stored at 5 °C (2–8 °C).

Stabilized H2SO4 pH1 was prepared by adding 94 ml 1 M H2SO4, 36 mg potassium sulphate and 1236 mg sodium sulphate in 1000 ml distilled water to prevent any bacterial and fungal growth during storage of T. suis eggs. Storage in sulphuric acid is not known to affect either the development or the infectivity of T. suis eggs [19]. Also, 0.1 N H2SO4 has been used as an embryonation medium for Ascaris suum eggs [20].

To secure comparable preparatory conditions for the batches in the experiment with Gram-positive bacteria, the medium of one T. muris batch was changed to H2SO4 pH1 and stored at 5 °C, and the medium of one T. suis batch was changed to Milli-Q and stored at room temperature for 3 months.

Prior to use, all batches were transferred to Milli-Q.

Bacteria

The E. coli strains (BL-21, M15, SG13009, PMC103, JM109, TUNER, DH5alpha, TOP10) were incubated overnight in Luria Broth (Sigma-Aldrich) at 37 °C on a shaker (250 rounds per minute; rpm). The Gram-positive bacteria Streptococcus hyointestinalis (DSM No. 20770) and Enterococcus caccae (DSM No. 19114) were cultured in tryptic soy broth (Sigma-Aldrich) with added yeast extract (Sigma-Aldrich). Lactobacillus amylovorus (DSM No. 16698, deposited by Dr. S. Konstantinov, Laboratory for Microbiology, Wageningen University and Research Centre Agrotechnology & Food Sciences, Wageningen, The Netherlands), Lactobacillus murinus (DSM No. 20452), and Lactobacillus reuteri (DSM No. 20016) were cultured in MRS broth (Sigma-Aldrich). Cultures were incubated overnight in the presence of CO2 on a shaker at 37 °C (range 220–275 rpm).

The optical density (OD) was measured on SpectraMax M2 (Molecular Devices) at 600 nm (absorbance). The concentration was calculated using E. coli formula: OD × 5 × 108 = bacteria/ml [16].

Incubation with Escherichia coli

A total of 1000 embryonated eggs per well were plated on 24-well plates in RPMI (1640, W/HEPES; CE, GIBCO) with 5 % penicillin (Invitrogen, 100 U/ml) -streptomycin (Invitrogen, 100 μg/ml) and 5 % amphotericin B (Sigma-Aldrich A2942, 250 μg/ml) together with an E. coli suspension.The samples were incubated at 36 °C and examined after 4 and 24 hours. The BL-21 strain (E. coli strain used for hatching T. muris eggs as described in [16]) was tested at three concentrations (5×108, 10×108 and 20×108 bacteria/ml), while the other E. coli strains were tested at one concentration (5×108 bacteria/ml). After the incubation, the wells with T. muris eggs were examined for hatching in subsamples (4–5 × 20 μl), whereas entire wells were examined for T. suis eggs.

Furthermore, the E. coli hatching assay was performed on T. muris eggs embryonated in Milli-Q or H2SO4 pH1 in concentrations of 600–1500 embryonated eggs in 3 ml, according to the original method. The number of released larvae was examined in subsamples for n = 4 per group.

Incubation with Gram-positive bacteria

A suspension of Gram-positive bacterium (5×108 bacteria/ml) was mixed with Trichuris eggs, and five aliquots of approximately 100 embryonated eggs per well were plated on a 96-well plate. Plates were flushed with CO2 and incubated in RPMI-1640 alone or RPMI-1640 containing 5 % penicillin-streptomycin and 5 % amphotericin B at 37 °C. The samples were examined after 4 and 20 h.

In all experiments, released larvae were enumerated using an inverted microscope (Carl Zeiss Primo Vert).

Statistical analysis

Data was analyzed in GraphPad Prism software (version 6) and R version 3.1.2 (R Core Team, 2014). Hatching percentage was calculated based on the number of released larvae and the number of embryonated eggs added per well. Data was log transformed as y = log (y + 1) whenever necessary to fulfill requirements of the normal distribution for the analysis. One-way ANOVA was used to test the effect of different E. coli (BL-21) concentrations on T. muris egg hatching after 4, and 24 h, followed by the Tukey’s post hoc test for multiple comparisons. The effects of different E. coli strains and the time of incubation on T. muris egg hatching were evaluated with the two-way ANOVA. Since the hatching in the control groups was absent or close to 0, they were not included in the analysis of ANOVA. An unpaired Student’s t test was used for pairwise comparisons. Generalized linear mixed model (glmm) fit by maximum likelihood (Laplace approximation) (Bates D, Maechler M, Bolker B and Walker S, 2014; R package version 1.1–7) was used to test the influence of different bacteria and the type of egg storage medium (Milli-Q and H2SO4 pH1) on overnight T. muris egg hatching in RPMI-1640 without antimicrobials. As data contained too many 0 values, the interaction between bacteria and the type of egg storage medium (Milli-Q and H2SO4 pH1) was not included in the analysis. Values of p < 0.05 were considered statistically significant.

Ethical approval for animal use

The experimentation on mice was approved by the local veterinary agency based on Swiss cantonal and national regulations (permission no. 2070), and on minipigs by the Animal Experiments Inspectorate, Ministry of Food, Agriculture and Fisheries of Denmark (permission no: 2012-15-2934-00641).

Results

Egg hatching with Escherichia coli

After 4 h incubation, the two lowest concentrations of E. coli (BL-21) tested (5×108 and 10×108 bacteria/ml) resulted in significantly higher T. muris egg hatching than a concentration of 20×108 bacteria/ml (F2,9 = 54.20, p < 0.0001; Tukey’s post hoc test p < 0.0001). However, no significant difference could be demonstrated after 24 h (F2,9 = 0.93, p = 0.43). No hatching was observed for T. suis at any concentration tested (Table 1).

Table 1 Hatching of Trichuris muris and T. suis eggs after incubation with Escherichia coli (BL-21)
Full size table

The strain of E. coli (ANOVA F7, 32 = 3.52, p = 0.01) and the time of incubation (ANOVA F1, 32 = 12.97, p = 0.001) had a significant effect on the hatching of T. muris eggs (Fig. 1). None of the E. coli strains evaluated induced hatching of T. suis eggs after 4 and 24 h incubation (average hatching was less than 0.2 %).

Fig. 1
figure 1

Trichuris muris eggs incubated with different Escherichia coli strains. Samples were incubated for 4 and 24 hours at 36 °C. The control group represents T. muris eggs incubated without the addition of E. coli. Arithmetic mean (bar) and standard deviation (error bars) for n = 3 per group

Full size image

No significant difference in hatching was observed for T. muris eggs embryonated in H2SO4 pH1 (10.9 ± 6.3 %) (mean ± S.D.) and in Milli-Q (21.1 ± 8.4 %) after 4 h with E. coli at 37 °C (5×108 bacteria/ml) (t6 = 1.9, p = 0.10).

Egg hatching with Gram-positive bacteria

After 4 h, no significant hatching could be demonstrated for T. muris incubated with any of the Gram-positive bacteria in any incubation media (the highest average hatching was 0.6 %). After 20 h, E. caccae (z value = 4.26, p < 0.001) and L. reuteri (z value = 3.42, p < 0.001) induced significantly higher hatching of T. muris eggs as compared to the controls. The effect of L. reuteri on egg hatching did not significantly differ from E. caccae (p = 0.62) (Fig. 2). However, high variation was observed in and in-between experiments with the same bacterium.

Fig. 2
figure 2

Trichuris muris eggs incubated with Gram-positive bacteria in RPMI-1640 without antimicrobials. Samples were incubated for 20 hours at 37 °C. Individual counts of T. muris (Milli-Q or H2SO4 pH1) and median for all experiments with Enterococcus caccae (n = 20), Streptococcus hyointestinalis (n = 10), Lactobacillus reuteri (n = 15), L. amylovorus (n = 15), and L. murinus (n = 10). Controls represent T. muris eggs incubated without the addition of bacteria (n = 25)

Full size image

None of the tested bacteria induced significant hatching of T. suis eggs after 4 (average hatching less than 0.4 %) or 20 h post-incubation (Fig. 3).

Fig. 3
figure 3

Trichuris suis eggs incubated with Gram-positive bacteria in RPMI-1640 without antimicrobials. Samples were incubated for 20 hours at 37 °C. Individual counts of T. suis (Milli-Q or H2SO4 pH1) and median for all experiments with Enterococcus caccae (n = 20), Streptococcus hyointestinalis (n = 10), Lactobacillus reuteri with T. suis in Milli-Q (n = 15), L. reuteri with T. suis in H2SO4 pH1 (n = 20), L. amylovorus (n = 15), and L. murinus (n = 10). Controls represent T. suis eggs without the addition of bacteria (n = 25)

Full size image

Samples incubated with antimicrobials showed consistently low egg hatching of T. muris (Fig. 4) and T. suis eggs (average hatching was less than 0.6 %) after 20 h incubation. In all experiments, the control group with whipworm eggs without the addition of bacteria had low hatching (average hatching less than 0.5 %).

Fig. 4
figure 4

Trichuris muris eggs incubated with Gram-positive bacteria in RPMI-1640 with antimicrobials. Samples were incubated for 20 hours at 37 °C. Individual counts of T. muris (Milli-Q or H2SO4 pH1) and median for all experiments with Enterococcus caccae (n = 20), Streptococcus hyointestinalis (n = 10), Lactobacillus reuteriwith T. muris in Milli-Q (n = 20), Lactobacillus reuteri with T. muris in H2SO4 pH1 (n = 15), L. murinus (n = 10), and L. amylovorus (n = 15). Controls represent T. muris eggs without the addition of bacteria (n = 25)

Full size image

Discussion

The use of individual bacterial species [16–18] or contents from the caecum [18, 21, 22] and the upper colon [22] in hatching T. muris eggs, points to close interactions between commensal bacteria and parasites within their niche. Such interactions may also explain observations of parasite’s ability to modulate the intestinal microbiota [23–27], as shown by an increase of lactobacilli during chronic T. muris infection [26]. In the large intestine of Göttingen minipigs and mice, microbiota is primarily composed of the phyla Firmicutes and Bacteroidetes [28–30], while Proteobacteria comprises 0.6 % in Göttingen minipigs [28].

In the present study, T. muris eggs hatched with E. coli (Proteobacteria), E. caccae (Firmicutes), S. hyointestinalis (Firmicutes), L. reuteri, and L. amylovorus (Firmicutes), while T. suis had overall very low hatching. However, T. muris egg hatching was inconsistent and highly variable with the Gram-positive bacteria.

Due to dietary differences between a mouse and a pig, host-specific bacterial composition may account for the observed discrepancy in the hatching of eggs of the two whipworm species. Although only a few bacteria were presently selected for comparison, selected bacterial species belong to the genera associated with pig and mouse intestinal tract.

Different bacteria may have different mechanisms in which they trigger eggs to hatch. Hayes et al. reported that T. muris eggs hatched through close physical contact with E. coli, and suggested that other mechanisms may exist for Pseudomonas aeruginosa and Staphylococcus aureus [18]. Since the large intestine contains a diverse microbial community (mostly anaerobes) [31], it is possible that a combination of different bacteria with their specific mechanisms, rather than one specific bacterium, is needed to hatch T. suis eggs. Whether a specific bacteria-induced hatching pattern is characteristic of other whipworms eggs (e.g. T. trichiura and T. vulpis) is still undefined. While Areekul et al. reported T. trichiura and T. vulpis co-infections in stools of 10.7 % Trichuris positive children in Thailand [32], other reported human cases of T. vulpis infection are brought into question (reviewed in [33]).

On the other hand, the similarities between pig and human gut microbiota (reviewed in [34]) and positive treatment effect of T. suis eggs in patients with immune-mediated diseases [7–10], indicate that T. suis eggs hatch in the human gut. In addition, pigs have been experimentally cross-infected with T. trichiura, although the development did not reach a patent infection [35]. Therefore, it can be speculated that a common bacterial stimulus exists between T. suis and T. trichiura. Recently, a mix of T. trichiura and T. suis sequence types was identified in three human-originating worms [36]. Even though a few studies either observed the presence of a single male adult worm in the human gut [37] or low grade T. suis egg excretion under experimental infections [35, 38], humans are not considered either natural or suited host for T. suis. Furthermore, during repeated dosing in clinical studies, T. suis eggs are absent in the stool samples [8, 9], which underline that patent T. suis infection is limited to pigs, where the full life cycle is confined.

Although the present study tested a limited number of bacteria, our findings together with previous limited observations on Trichuris cross-species transmission, suggests that Trichuris has a narrower host-specificity window compared to other nematodes, such as Ascaris [39].

To further study the parasite-bacteria interaction and host-parasite specificity, T. suis and T. muris eggs may be cross-incubated with mucosal scrapings from the intestine of a mouse and pig, respectively. Inclusion of other whipworm species in such experiment, e.g. human T. trichiura and canine T. vulpis, may provide further insight into the specific parasite adaptation to the host gut environment.

Conclusions

The present study is the first to compare eggs of two whipworm species in relation to bacterial-dependent hatching. Although T. muris successfully hatched in the presence of the different E. coli strains, the egg hatching showed high variability in the presence of five tested Gram-positive bacteria after overnight incubation without the addition antimicrobials. In contrast, none of the tested bacteria triggered hatching of T. suis eggs. This suggests that the co-evolution of the parasite and host, and the adaptation to host’s microbiota has resulted in unique host-specific hatching stimuli of T. muris and T. suis eggs.