The avian schistosome Trichobilharzia franki in mice: Migration, pathogenicity, and the host immune response

Abstract

Cercarial dermatitis (CD; swimmer's itch) is a re-emerging skin disease caused by avian schistosomes, including Trichobilharzia franki. Here, we present morphological, genetic, and experimental evidence confirming the involvement of T. franki in recent CD outbreaks across Czechia. Ocellate furcocercariae were collected from Radix auricularia at four sites and identified as T. franki through ITS1 sequencing. Despite minor morphological differences from previously reported specimens, all isolates belonged to the genetically uniform T. franki “auricularia” clade. Experimental infection of mice with T. franki resulted in a ∼ 60 % penetration rate, accompanied by early-onset scratching and transient weight loss. Gross pathology demonstrated hemorrhages on lung surfaces and splenic atrophy at 2 days post-infection (dpi), along with a prominent enlargement of parotid lymph nodes at both 2 and 7 dpi. Histological examination of the skin revealed viable schistosomula, moderate leukocyte infiltration, epidermal hyperplasia, and the formation of hyperkeratotic crusts at 2 dpi. By 7 dpi, parasites were no longer detectable, but epidermal pathology persisted. In the lungs, eosinophil-rich foci and multifocal hemorrhages were observed at 2 dpi, transitioning to neutrophil-dominated lesions at 7 dpi, despite the absence of detectable schistosomula. Splenocytes from infected mice responded to homologous and heterologous cercarial antigens by producing IFN gamma, IL-4, and IL-10, indicating a mixed Th1/Th2/Treg profile and notable species cross-reactivity. However, parasite-specific IgG remained undetectable at 7 dpi. These findings confirm T. franki as the causative agent of CD outbreaks and underscore its capacity to induce localized and systemic pathology and immune response, cross-reacting with other schistosomes.

Graphical abstract

Highlights

• •Trichobilharzia franki isolated during cercarial dermatitis outbreaks in Czechia.
• •All isolates belong to “auricularia” clade and have low intraspecific variability.
• •First report of T. franki migration and pathology in experimentally infected mice.
• •Infection causes skin hyperkeratotic crust formation and transient lung hemorrhages.
• •Infected mice show Th1/Th2/Treg response and cross-reactivity to other species.

1. Introduction

Avian schistosomes comprise a diverse group of trematodes that infect aquatic snails and waterfowl, which serve as intermediate and definitive hosts, respectively (Loker et al., 2022). Aside from birds, some avian schistosomes can also infect mammals, triggering cercarial dermatitis (CD) (Cort, 1928; Horák et al., 2015). This condition is a complex skin immune reaction against penetrating cercariae, in which immediate dermal inflammation and neuronal irritation combine with IgE-mediated allergic hypersensitivity (Inclan-Rico et al., 2024; Kouřilová et al., 2004). Clinically, CD manifests as maculopapular eruptions associated with intense itching that may last for up to 1–2 weeks (Macháček et al., 2018). Consequently, avian schistosomes and CD outbreaks have considerable socio-economic impacts, particularly in recreational areas (Horák et al., 2015).

Representatives of the genus Trichobilharzia are most often recognized as causative agents of CD outbreaks (Bispo et al., 2024; Horák et al., 2015). Seven Trichobilharzia species have been detected in Europe: T. szidati, T. franki, T. regenti, T. salmanticensis, T. mergi, T. anseri, and T. physellae, recently introduced from North America (Helmer et al., 2021). While some species, such as the viscerotropic T. szidati and the neurotropic T. regenti, have been thoroughly investigated and serve as important laboratory models (Macháček et al., 2022; Vondráček et al., 2022), little is known about the biology and host-parasite interactions of most other species. For instance, there is virtually no information available on the fate of T. franki in mammals, even though this species is often reported in European CD outbreaks (Bispo et al., 2024).

T. franki was first described following CD outbreaks at Lake Tunisee in Baden-Württemberg, Germany (Müller and Kimmig, 1994). Since then, T. franki has been reported in faunistic records throughout Europe, often being associated with CD outbreaks (Fig. 1, Supplementary data). Interestingly, T. franki was also detected in Iran, where it is likely to have been introduced (along with other bird schistosomes) by migratory birds (Ashrafi et al., 2021). The significant role of migratory birds in T. franki distribution was supported by a large phylogenetic analysis conducted by Lawton et al. (2014). While their analysis revealed four distinct clades of T. franki populations in Europe, the clades showed no geographic separation and contained haplotypes from various regions.

Fig. 1.

Reported distribution of Trichobilharzia franki. In the map, the blue colour marks countries where T. franki was reported. The details are provided in the table, in which ID numbers correspond to the marked countries. Stage: A, adults; C, cercariae; E, eggs. Host: Ab, Ampullaceana balthica; Anatids, anatid birds; Ra, Radix auricularia. References are provided as Supplementary data. The map was created by www.mapchart.net. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

The adults of T. franki reside in the venous system of the liver, mesentery, or intestinal submucosa of anatid waterfowl (Ashrafi et al., 2021; Müller and Kimmig, 1994). The eggs are distributed in the intestinal wall and liver, and visceral lesions were observed in experimentally infected ducks (Bayssade-Dufour et al., 2002; Müller and Kimmig, 1994). The mature eggs appear in the feces of experimentally infected ducks 13–15 days post-infection (dpi), and the hatched miracidia infect freshwater snails, Radix auricularia, as intermediate hosts (Kock, 2001; Müller and Kimmig, 1994). However, Jouet et al. (2008) demonstrated that T. franki cercariae can also be emitted from Ampullaceana balthica (formerly known as R. peregra = R. ovata = R. balthica). Indeed, the presence of two distinct clades, T. franki “auricularia” and T. franki “balthica” (the latter originally called “peregra”), was supported by molecular analyses (Jouet et al., 2010). While T. franki “auricularia” has been detected in the entire Europe, T. franki “balthica” was reported (often under the old name “peregra”) across North-Western Europe (Fig. 1). The clades seem to differ in their preferred definitive hosts (Anas platyrhynchos and Cygnus olor for T. franki “auricularia” versus Aythya fuligula and Anas crecca for T. franki “balthica”) and cercarial morphology (T. franki “balthica” is smaller than T. franki “auricularia”) (Jouet et al., 2010; Jouet et al., 2009). Altogether, these findings suggest that T. franki “auricularia” corresponds to the original species described by Müller and Kimmig (1994), but T. franki “balthica” presumably represents a new species.

In this study, we investigated several T. franki “auricularia”-associated CD outbreaks in Czechia and phylogenetically characterized the T. franki “auricularia” isolates. Moreover, one isolate was used to experimentally infect mice; we monitored parasite migration and pathogenicity. Finally, we analyzed the host immune response to the infection and tested for immune cross-reactivity with other schistosome species.

2. Materials and methods

2.1. Parasites

In July–August 2023, Radix auricularia snails, preliminarily identified based on shell morphology, were collected in four recreational water bodies in Czechia, from which outbreaks of human cercarial dermatitis were reported. Locality details are provided in Table 1. The snails were moved to the laboratory and placed individually into beakers with dechlorinated tap water. The next day, they were exposed to lamp light, and cercarial emergence was monitored after one hour. If ocellate furcocercariae were shed, they were fixed in 96 % ethanol for molecular analysis. After confirmation of the species identification (see below), fresh T. franki cercariae were taken for experimental infection of mice or were fixed by hot 4 % formaldehyde for morphological analysis. For mouse infections, the cercariae were isolated from the snails originating from Plzeň (Škodaland pool), as all the snails survived under laboratory conditions and provided a sufficient number of cercariae.

Table 1.

A list of localities from which Radix auricularia snails were collected. The number of positive snails stands for individuals shedding ocellate furcocercariae.

Date of collection	Locality, water body	GPS	Number of Radix auricularia snails
			Total	Positive	Prevalence
13.07.2023	Lukov u Zlína, Dolní bělovodský/lukovský pond	49.2908414 N, 17.7268689 E	9	1	11.1 %
17.07.2023	Proboštov, Proboštovský pond	50.6664067 N, 13.8385653 E	39	2	5.1 %
18.07.2023	Plzeň, Škodaland pool	49.7172092 N, 13.3524572 E	290	5	1.7 %
25.08.2023	Dívčice, Blatec pond	49.1167367 N, 14.3084103 E	75	2	2.6 %

Cercariae of Trichobilharzia regenti and Trichobilharzia szidati and adults of Schistosoma mansoni (Puerto Rican strain), used for comparative purposes in penetration rate and/or cross-reactivity experiments, were obtained from laboratory cycles maintained in the Department of Parasitology, Faculty of Science (Charles University, Prague) as already described (Inclan-Rico et al., 2024; Macháček et al., 2022; Peterková et al., 2024).

2.2. Molecular identification, phylogenetic, and haplotype analysis

Whole cercariae and the part of the snail mantle were used for DNA extraction of the parasite and the intermediate host, respectively, by the DNA isolation kit GeneAll ExgeneTM Tissue SV plus. Polymerase chain reactions (PCR) were performed in 25 μl volume using 50 ng of DNA template, EmeraldAmp Max PCR Master Mix (Takara bio), and 0.5 μl of 10 μM primers for ITS2 region of aquatic snails (ITS2 RIXO TTCTATGCTTAAATTCAGGGG; ITS2 NEWS TGTGTCGATGAAGAACGCAG (Almeyda-Artigas et al., 2000)), and newly designed primers for ITS1 region of avian schistosomes (Schisto_ITS1_tf_F: TGAGACATGCCAGTTGGTGT; Schisto_ITS1_tf_R: TGCAGTCCAGCTTAAAGCCA). The PCR protocols were as follows: aquatic snails: 1× (94 °C, 10 min), 30× (94 °C, 30 s; 50 °C, 30 s; 72 °C, 60 s), 1× (72 °C, 7 min), 4 °C; avian schistosomes: 1× (95 °C, 5 min), 30× (95 °C, 60 s; 58 °C, 30 s; 72 °C, 50 s), 1× (72 °C, 10 min), 4 °C.

Obtained sequences of T. franki were identical; therefore, a single representative sequence (Probostov_Tf_I) was selected and blasted against the NCBI database with the organismal restriction to T. franki (taxid:178317), T. franki haplotype “peregra/balthica” (taxid:1972557), Trichobilharzia sp. haplotype “peregra/balthica” (taxid:860518). Sequences retrieved from GenBank originating from cercariae isolated from snails, which had been identified either morphologically or molecularly, were retained for phylogenetic analysis. If the intermediate snail host was molecularly identified in the original study, its species name is indicated in italics next to the corresponding sequence in the cercarial phylogenetic tree. The phylogenetic analysis was performed in Geneious Prime (v. 2023.0.4) as follows: multiple alignment was performed using Clustal Omega (v. 1.2.3). A phylogenetic tree was constructed using the maximum likelihood analysis using PHYML (v. 3.3.20180621) (Guindon et al., 2010) with the substitution model K80 and 100 bootstrap replicates. The phylogenetic tree was rooted with the T. regenti sequence (EF094537.1).

A haplotype network of T. franki isolates was constructed based on ITS1 sequences. Representative sequences were selected according to a previously inferred phylogenetic tree, specifically from distinct clades corresponding to the “auricularia” and “balthica” types. Sequence alignment was performed using Geneious Prime (v. 2023.0.4) with the Clustal Omega algorithm (v. 1.2.3). All sequences were trimmed and standardized to a uniform length of 551 base pairs. The haplotype network was then generated using the TCS algorithm implemented in PopART (v. 1.7) (Leigh and Bryant, 2015; Templeton et al., 1992), which visualizes genealogical relationships among haplotypes based on statistical parsimony.

2.3. Infection of mice

C57BL/6JOlaHsd female mice were purchased from Envigo/Inotiv and housed in the Centre for Experimental Biomodels, First Faculty of Medicine (Charles University, Prague). Seven-week-old animals were anesthetized by i.p. injection of ketamine/xylazine, and their left pinnae were immersed in 1.5 ml of water with ∼1000 freshly emerged cercariae of T. franki (from Plzeň, Škodaland pool) or other species (T. regenti, T. szidati) used for comparative purposes. After 30 min, the infection process was terminated, and cercariae plus cercarial bodies remaining in the water were fixed and enumerated to calculate the penetration rate.

2.4. Harvesting mice and collecting tissue samples

The mice were harvested at 2 and 7 dpi, along with uninfected control animals age-matched to 7 dpi (n = 6 per time point). The mice were deeply anesthetized with isoflurane, weighed, and bled out after a terminal ketamine/xylazine overdose. Serum was collected by centrifugation (10 min, 1500 g) of clotted blood and stored at −80 °C. The pinnae (control and infected) were cut and fixed in 4 % formaldehyde. After bronchoalveolar lavage (see below), the lungs were extracted and either fixed in 4 % formaldehyde (the left lung) or torn apart in the phosphate-buffered saline (pH 7.4; PBS) to let schistosomula migrate out of the tissue (the rest of the lungs). Finally, pinna-draining superficial parotid lymph nodes (pLN) and spleen were harvested and further processed.

2.5. Histology

Formaldehyde-fixed samples (pinnae, lungs) were dehydrated and mounted in paraffin (Leica HistoCore PEGASUS Plus tissue processor). Five-micrometer sections were cut and routinely stained with hematoxylin & eosin.

2.6. Bronchoalveolar lavage

The trachea was cannulated, and the lungs were lavaged with 0.5 ml of ice-cold PBS. The resultant bronchoalveolar lavage fluid (BALF) was centrifuged (5 min, 500 g, 4 °C), and the pelleted cells were resuspended in PBS, stained with trypan blue, and counted in the Bürker chamber. The residual BALF was stored at −80 °C. Levels of selected cytokines (IL-1β, TNF, IFN gamma, IL-4, IL-5, IL-10) were then measured in undiluted samples by commercially available enzyme-linked immunosorbent assays (ELISA; BioLegend).

2.7. Cell counts in parotid lymph nodes

After removing the surrounding fat tissue, pLNs were harvested into a 1.5 ml tube with ice-cold PBS. The tissue was homogenized by a polypropylene pestle in the tube, filtered via the 70 μm cell strainer, and centrifuged (5 min, 380 g, 4 °C). The pelleted cells were quantified in the Bürker chamber as described above.

2.8. Splenocyte isolation and restimulation, cytokine detection

The harvested spleen was weighed and mechanically homogenized by passing it through a 100 μm cell strainer to prepare single-cell splenocyte suspensions in PBS, as described (Majer et al., 2020). After red blood cell lysis with ammonium-chloride‑potassium buffer, the splenocytes were washed with PBS and counted (Countess II Automated Cell Counter, Invitrogen). The cells were resuspended in the cultivation medium (RPMI 1640 supplemented with 10 % fetal bovine serum, 2 mmol/l l-glutamine, 100 U/ml penicillin, and 100 μg/ml streptomycin) and seeded into a 24-well plate (1.25 × 10⁶ cells/ml).

The cells were treated with a soluble fraction of cercarial homogenate (10 μg/ml) of T. franki (TfH), T. regenti (TrH), and T. szidati (TsH). The homogenates were prepared according to Majer et al. (2020). Briefly, the freshly emerged cercariae were washed and concentrated in sterile, ice-cold PBS supplemented with peptidase inhibitors (Complete Mini EDTA-free, Roche). The cercarial suspension was homogenized by sonication (Vibra Cell) and centrifuged (20 min, 16,000 g, 4 °C) to get the soluble fraction. Additionally, the cells were treated with soluble worm antigens prepared from mechanically homogenized (FastPrep-24 5G, MP Biomedicals) adults of Schistosoma mansoni (SmAWA; (Macháček et al., 2023)). Untreated cells and those stimulated with concanavalin A (1.25 μg/ml) served as negative and positive controls, respectively.

After 72 h of cultivation (37 °C, 5 % CO₂) with the stimuli, the cells were centrifuged (5 min, 300 g, 4 °C), and the supernatants were collected and stored at −80 °C. Levels of selected cytokines (IFN gamma, IL-4, and IL-10) were then measured in undiluted samples by commercially available ELISAs (BioLegend).

2.9. Detection of serum antibodies

The levels of parasite-specific serum antibodies (IgG, IgG1, IgG2a) were analyzed by a home-designed ELISA based on the established protocol (Macháček et al., 2022; Majer et al., 2020). Briefly, the wells of the 96-well plate were coated with 6.25 μg/ml of parasite antigens (TfH, TrH, TsH, or SmAWA; see above) in carbonate buffer overnight at 4 °C. After washing with PBS/0.05 % Tween 20 (PBS-T), the wells were blocked with 5 % nonfat dry milk in PBS-T for 2 h and probed with diluted (1:80) mouse sera for 2 h. Secondary anti-mouse IgG (Merck, A2554; 1:8000), anti-IgG1 (Abcam, ab97240; 1:2500), or anti-IgG2a (Abcam, ab97245; 1:2500) peroxidase-conjugated antibodies were added for 1.5 h. After washing, the reaction was visualized by tetramethylbenzidine liquid substrate (Merck), stopped by 1 M HCl, and read at 450 nm (Tecan Infinite M200 reader).

2.10. Statistical analysis

Quantitative data were visualized and analyzed in GraphPad Prism (v. 10.5). Values obtained from individual mice are shown along with the group average and standard deviation. Specific tests applied are indicated in the figure legends. P-values <0.05 were considered significant and are indicated as follows: *p < 0.05, **p < 0.01, ***p < 0.001. The exact p-value is also shown if it ranges from 0.05 to 0.10.

3. Results

3.1. Morphological and phylogenetic characterization of cercariae

All collected snails releasing ocellate furcocercariae were identified as R. auricularia by ITS2 sequencing. The ocellate furcocercariae from all localities had a pair of pigmented eyes and two pairs of circumacetabular glands (Fig. 2A, B), overall corresponding to Trichobilharzia morphology. Cercarial dimensions, assessed in samples collected in Proboštov and Plzeň, revealed that the cercariae had shorter bodies and tail stems than previously reported for T. franki (Fig. 2C). Nevertheless, the phylogenetic analysis based on ITS1 convincingly demonstrated that cercariae from all four localities belonged to T. franki “auricularia” clade (Fig. 3A), which corresponds to the intermediate hosts from which they were collected.

Fig. 2.

Morphological characterization of cercariae. (A) Ocellate furcocercariae released from R. auricularia, native specimen. (B) Alizarin staining of the cercariae, native specimen. Circumacetabular penetration glands (*) and ducts are visible. (C) Dimensions of the cercariae collected in Proboštov and Plzeň and their comparison to the other published records. Mean ± standard deviation (SD) is shown (except for Jouet et al. (2010) for which SD is not available). Type of fixation and sample size are indicated (FA, formaldehyde).

Fig. 3.

Phylogenetic characterization of cercariae. (A) Phylogenetic tree of Trichobilharzia franki based on the ITS1 region, arrows mark samples collected in the present study. Sequences from cercariae obtained from either morphologically or molecularly identified intermediate hosts were included in the analysis. Names of molecularly identified intermediate hosts are included in branch labels; “auricularia” and “balthica” type refers to Radix auricularia or Ampullaceana balthica (formerly known as Radix balthica = R. peregra = R. ovata), respectively. (B) A haplotype network of T. franki constructed based on ITS1 sequences. The arrow points to the position of isolates collected in this study; I.–IV. mark the main clusters.

Next, haplotype analysis was performed to evaluate the population variability of T. franki “auricularia” in Czechia. Although the samples were collected in four different regions across Czechia (site distances >100 km), all current isolates clustered together in the clade containing haplotypes from Central and Western Europe (Fig. 3B). In total, three major haplogroups were identified. The first group (“auricularia”), in which all Czech isolates were placed, comprised haplotypes from Czechia, Denmark, France, Italy, Poland, Switzerland, and the United Kingdom (Fig. 3B, cluster II.). Phylogenetically adjacent to them was the group containing “auricularia” haplotypes from Iceland, Finland, and Switzerland (Fig. 3B, cluster I.). The third group (“balthica”) included haplotypes shared among Iceland, Norway, Poland, and Switzerland (Fig. 3B, clusters III. + IV.). Collectively, these results show that all collected cercariae correspond to T. franki and that they are all genetically uniform.

3.2. Clinical outcome of T. franki infection in mice

Cercariae of T. franki collected in Plzeň were used to experimentally infect mice. The average penetration rate was approximately 60 %, comparable to that of other avian schistosome species such as T. regenti and T. szidati (Fig. 4A). Mice began scratching their infected pinnae within several hours post-infection, although no additional behavioral changes were observed. Notably, body weight loss was detected at 2 dpi, but the mice recovered by 7 dpi (Fig. 4B). Gross pathological examination revealed no visible lesions or abnormalities in the liver or intestine; however, lung surface hemorrhages with diffuse and indistinct margins were observed in some mice at 2 dpi (Fig. 4C). Despite this, no schistosomula were recovered from the lungs at either time point. The spleen appeared reduced in size at 2 dpi (Fig. 4D), while the pLNs draining the infected pinnae were markedly swollen at both 2 and 7 dpi (Fig. 4E). Indeed, pLNs contained up to 15 times more cells than those draining uninfected pinnae at 7 dpi (Fig. 4F). Together, these findings indicate a pronounced local immune response in the skin and transient, migration-associated lung pathology.

Fig. 4.

Clinical outcome of Trichobilharzia franki infection in mice. (A) Penetration rate of T. franki, T. regenti, and T. szidati, determined as the proportion of cercariae that successfully penetrated the mouse pinna within 30 min. (B) Body weight of infected mice (dpi, days post-infection). Uninfected mice (0 dpi) were age-matched to the 7 dpi group. (C) Occurrence of the lung surface hemorrhages (marked by an arrow in the inset). (D) Relative size of the spleen shown as spleen to body weight ratio. (E) Appearance of the superficial parotid lymph nodes (pLNs) draining uninfected (Ctrl) or infected (Inf) mouse pinnae. (F) Quantification of the pLN cells in uninfected mice (0 dpi) and mice at 7 dpi (pLNs draining control and infected pinnae are shown). Data were analyzed by Kruskal-Wallis and Dunn's multiple comparisons test (A, C), ordinary 1-way ANOVA and Holm-Šídák's multiple comparisons test (B, D), or Brown-Forsythe ANOVA and Dunnett's T3 multiple comparisons test (F). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

3.3. Pathological examination of skin and lungs

At 2 dpi, T. franki schistosomula were detected in the dermis and appeared morphologically intact (Fig. 5B). Moderate intradermal inflammation was observed, characterized by infiltration of leukocytes (Fig. 5C), but these cells did not accumulate around the schistosomula. Prominent pathological features included tissue edema, epidermal hyperplasia (Fig. 5C), and the presence of hyperkeratotic crusts detaching from the skin surface (Fig. 5D). These crusts contained densely packed inflammatory cells; however, no schistosomula were convincingly identifiable. Suprabasal acantholysis (loss of intercellular connections, Fig. 5E) and intraepidermal pustules (Fig. 5F) were rarely observed at this stage. By 7 dpi, schistosomula were no longer detectable in the skin, aside from occasional parasite remnants observed within the hyperkeratotic crusts (Fig. 5G). Epidermal hyperplasia persisted, and intraepidermal pustules were more frequently noted compared to 2 dpi. Intradermal leukocyte infiltration was markedly increased (Fig. 5H), contributing to a noticeable thickening of the dermis.

Fig. 5.

Tissue pathology in mouse skin infected with Trichobilharzia franki. (A) Control skin tissue from uninfected mice. (B) Intact schistosomulum (arrow) in the dermis, 2 days post-infection (dpi). (C) Intradermal infiltration of leukocytes (asterisk) and epidermal hyperplasia (arrow), 2 dpi. (D) Hyperkeratotic crusts (arrow) detaching from the skin surface, 2 dpi. (E) Suprabasal acantholysis (arrow), 2 dpi. (F) Intraepidermal pustula (arrow) and hyperkeratotic crusts (asterisks), 2 dpi. (G) Schistosomula remnants (arrow) in the detached crust, 7 dpi. (H) Intradermal leukocyte infiltration (asterisk), 7 dpi. Hematoxylin & eosin staining, scale bar: 50 μm.

In the lungs, multifocal hemorrhages and perivascular cuffs were observed at 2 dpi (Fig. 6B). No signs of edema, congestion, alveolar wall thickening, or peribronchial inflammatory cuffs were detected at either time point. Also, no increase in bronchoalveolar lavage fluid cellularity or cytokine levels was detected. The most prominent pathological feature was the presence of granulocyte-rich inflammatory lesions (Fig. 6C, D), which were randomly distributed throughout the examined lung lobes. While they were composed mostly of eosinophils at 2 dpi, neutrophil-rich lesions were seen at 7 dpi. These lesions resembled foci typically associated with parasite migration; however, schistosomula or their remnants were never identified within these foci or elsewhere in the pulmonary tissue.

Fig. 6.

Tissue pathology in mouse lungs infected with Trichobilharzia franki. (A) Control lung tissue from uninfected mice. (B) Lung hemorrhage (asterisks) and perivascular cuffs (arrows), 2 days post-infection (dpi). (C) Eosinophil-rich inflammatory lesion, 2 dpi. (D) Neutrophil-rich lesion, 7 dpi. Hematoxylin & eosin staining, scale bar: 50 μm.

3.4. Immune response against homologous and heterologous antigens

Splenocytes harvested at 0 and 7 dpi were treated with cercarial antigens to recall the parasite-specific cytokine production that orchestrates the cellular immune response. While splenocytes from control mice did not react to any stimulation, those obtained from T. franki-infected mice produced IFN gamma, IL-4, and IL-10 when stimulated with homologous cercarial antigen (Fig. 7A–C). A similar cytokine profile was secreted during the heterologous stimulation of splenocytes with cercarial antigens derived from T. regenti and T. szidati. Interestingly, splenocytes from infected mice also partially reacted to unrelated (heterologous) antigens from adult S. mansoni, which increased the levels of IFN gamma and IL-4. Despite the already well-developed cellular immune response, no parasite-specific serum IgG was detected at 7 dpi (Fig. 7D).

Fig. 7.

Immune response of mice infected with Trichobilharzia franki against homologous and heterologous antigens. (A–C) Cytokine production by splenocytes (obtained at 0 and 7 days post-infection; dpi) treated with homogenates from T. franki (TfH), T. regenti (TrH), and T. szidati (TsH) cercariae, as well as from Schistosoma mansoni adults (SmAWA). (D) Levels of IgG recognizing homologous (TfH) or heterologous (TrH, TsH, SmAWA) antigens. Data were analyzed by ordinary 2-way ANOVA and Holm-Šídák's multiple comparisons test.

4. Discussion

T. franki is an epidemiologically significant but experimentally underexplored avian schistosome, commonly found across Europe. Our study provides a comprehensive characterization of T. franki collected during CD outbreaks in Czechia, integrating field surveillance, molecular phylogenetics, and experimental mouse infections. We document its widespread occurrence at four geographically distinct sites and confirm its role as a frequent agent of CD, corroborating reports from other European countries (Fig. 1). Our data demonstrate that T. franki is successfully transmitted in Czech freshwater ecosystems, including recreational lakes, posing a seasonal public health risk and highlighting the need for further research into its biology in mammalian hosts.

All collected cercariae belonged to the T. franki “auricularia” clade, in agreement with the intermediate host reported in the original description by Müller and Kimmig (1994). Although the specimens were smaller than previously reported, cercarial dimensions are known to vary and are unreliable for species identification (Podhorský et al., 2009; Reier et al., 2020). Some Trichobilharzia species even exhibit cercarial polymorphism, producing large and small morphs (Soldánová et al., 2022), making molecular identification indispensable. Recently developed DNA-based methods, such as qPCR, multiplex PCR, and loop-mediated isothermal amplification, now enable rapid, sensitive, and specific detection of avian schistosomes (Procházka et al., 2025; Rudko et al., 2019).

Our haplotype network analysis grouped all current Czech T. franki samples into a cluster shared with continental European isolates of the “auricularia” type. This cluster corresponds to Clade 1/Haplotype 1 in Lawton et al. (2014), who described T. franki diversity in the United Kingdom. Notably, this group is genetically distant from Northern European isolates (e.g., from Finland and Iceland), indicating genetic uniformity of Czech T. franki and a closer relationship to Western European lineages. Of note, the current samples were only two nucleotides different from T. franki isolated in Czechia twenty years ago by Rudolfová et al. (2005), which also belong to the same cluster (II.). These observations support the view that the overall haplotype diversity and population structure of T. franki could be quite stable in time and largely depend on the migratory patterns of their definitive hosts, as suggested by Lawton et al. (2014). Accordingly, ducks and swans in Czechia primarily migrate within Central and Western Europe, with little to no connectivity to northern regions such as Finland or Iceland (Spina et al., 2022a, Spina et al., 2022b), which aligns with the observed population structure.

Despite being a common CD agent, little was known about how T. franki invades mammalian skin. We show that it can successfully penetrate mouse skin, with a success rate of ∼40–80 %, similar to other Trichobilharzia species (Wulff et al., 2007). Therefore, its prominence in CD outbreaks is likely not due to enhanced penetration ability, but rather to the ecological characteristics of its intermediate host, Radix auricularia. This snail is a pioneer species, capable of colonizing newly formed freshwater habitats, such as early-stage ponds with low density of submerged plants (Yang et al., 2025), which are commonly associated with CD outbreaks in Czechia (our unpublished data). Identifying key risk factors will be critical for targeted public health interventions.

Skin penetration by T. franki induces local inflammation with epidermal hyperkeratosis, similar to that caused by other avian schistosomes (Kouřilová et al., 2004; Olivier and Weinstein, 1953). Such epidermal hyperplasia serves to both hinder parasite migration and repair tissue damage (Piipponen et al., 2020). Keratinocyte activation is also seen during early Schistosoma mansoni infection (Bourke et al., 2015), though inflammation is typically milder and resolves faster (Incani and McLaren, 1984). The difference may be due to the ability of human schistosomes, but not avian ones, to suppress neuroimmune signaling that initiates inflammation in naive hosts (Inclan-Rico et al., 2024). While the specific parasite molecules involved remain unidentified, peptidases from penetration glands are likely candidates (Kouřilová et al., 2004). For example, cathepsin B2 from T. regenti, a cysteine peptidase degrading skin components, activates mouse dendritic cells in vitro (Majer et al., 2020). However, further mechanistic studies involving also T. franki are required to address these fundamental questions and interspecific diversity.

The magnitude of T. franki-induced skin inflammation was mirrored in lymphadenopathy of skin-draining pLNs, indicating strong adaptive immune activation. In mice infected with T. regenti, increased cellularity in skin-draining lymph nodes was also observed (Majer et al., 2020), and pLN cells treated with cercarial antigens produced a mixture of Th1/Th2/Treg cytokines (Kouřilová et al., 2004). Such immune polarization was seen even at the systemic level (Majer et al., 2020) and corroborates our observations from mice infected with T. franki. These results suggest that the mixed cytokine profile, combining pro-inflammatory Th1, anti-inflammatory Th2, and T regulatory response, is a common early feature of primary infections with avian schistosomes, irrespective of the particular species.

Interestingly, our cytokine data also reveal substantial immune cell cross-reactivity to diverse avian schistosomes, which could have epidemiological implications. Repeated exposures to different avian schistosome species, on various occasions or at different sites, may lead to continuous sensitization and development of CD in a species-independent manner (Kouřilová et al., 2004; Olivier, 1949). On the other hand, this cross-reactivity could also be utilized in the development of diagnostic kits for a broad spectrum of avian schistosomes. Indeed, antibody cross-reactivity has been shown for several avian schistosome species (Kouřilová and Kolářová, 2002), with anti-cercarial antibodies detectable 7–9 dpi depending on the infection dose (Kouřilová and Kolářová, 2002; Majer et al., 2020). Unexpectedly, splenocytes from T. franki-infected mice also responded to antigens derived from adult human schistosomes, indicating a response to antigens from another parasite genus and life cycle stage. Such cross-reactions, usually seen among cercariae (Kouřilová and Kolářová, 2002), could have important consequences. Notably, Pedersen et al. (1982) observed reduced severity of schistosomiasis in mice pre-exposed to avian schistosomes, raising the possibility of partial cross-protection. This immunoreactivity could have important real-world implications in regions where avian and human schistosomes co-occur (Mudavanhu et al., 2024), suggesting unexplored avenues of parasite interference.

Beyond skin pathology and anti-cercarial immune response, we also followed T. franki somatic migration and pathology. Data from experimentally infected ducks indicated the visceral route via lungs (Bayssade-Dufour et al., 2001; Müller and Kimmig, 1994), similar to the majority of other avian and human schistosomes (Chanová et al., 2007; Haas and Pietsch, 1991; Horák and Kolářová, 2000; Nation et al., 2020; Olivier, 1953). However, we did not detect any lung schistosomula, either naturally released or in histological sections, at 2 or 7 dpi. At these time points, other avian schistosomes, such as T. physellae, T. stagnicolae (Olivier, 1953), or T. szidati (Chanová et al., 2007; Haas and Pietsch, 1991), are commonly detectable. Despite the absence of direct evidence of schistosomula presence in the lungs, we suggest that T. franki migrates to the lungs, as evidenced by the multiple hemorrhages and granulocytic lesions found there. Hemorrhages are commonly associated with schistosome pulmonary migration (Olivier, 1953), and inflammatory lesions occur if schistosomula transit from vessels to alveoli (Crabtree and Wilson, 1986). The seeming absence of schistosomula could be explained by more rapid somatic migration and parasite elimination in the lungs before 2 dpi. For example, Bilharziella polonica has a peak in lung schistosomula number 12–24 h post-infection and is rarely found later (Horák and Kolářová, 2000). Future studies should include earlier time points to clarify migration dynamics and confirm lung involvement.

5. Conclusion

Our results reaffirm T. franki as an emerging agent of CD in Central Europe. Although it does not mature in mammals, its ability to penetrate skin and initiate a strong immune response is sufficient to cause significant pathology. Experimental infections in mice provided new insights into early host responses, including marked skin inflammation, mixed cytokine profiles, and cross-reactivity among (avian) schistosomes. These findings have implications for diagnosis, sensitization, and possible immunological interference with human schistosomes. Future research should focus on early somatic migration, parasite clearance kinetics, and the molecular basis of immune activation.

Availability of data and materials

All data are available within the manuscript.

CRediT authorship contribution statement

Tomáš Macháček: Writing – review & editing, Writing – original draft, Visualization, Supervision, Resources, Methodology, Investigation, Formal analysis, Conceptualization. Roman Leontovyč: Writing – review & editing, Writing – original draft, Visualization, Software, Resources, Methodology, Investigation, Formal analysis. Jan Procházka: Writing – review & editing, Investigation. Alena Revalová: Writing – review & editing, Investigation. Martin Majer: Writing – review & editing, Investigation. Barbora Šmídová: Writing – review & editing, Investigation. Petr Horák: Writing – review & editing, Supervision, Resources, Methodology, Funding acquisition.

Ethics statement

The study was performed in accordance with the European and Czech animal welfare legislation (EU Directive 2010/63/EU and Act No. 246/1992, respectively). The experiments were approved by the institutional (Faculty of Science, Charles University) and ministerial (Ministry of Education, Youth and Sports of the Czech Republic) animal welfare committees (protocol IDs: MSMT-34303/2022‐4, MSMT-33741/2022‐5, MSMT-33979/2022‐4).

Funding

The study was supported by the Czech Science Foundation (GA24-11031S), Charles University institutional funding (Cooperatio Biology, UNCE24/SCI/011, SVV 260678 and 260796), and European Regional Development Fund and Ministry of Education, Youth and Sports of the Czech Republic (CZ.02.1.01/0.0/0.0/16_019/0000759).

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

Supplementary material