Intermediate hosts of Protostrongylus pulmonalis (Frölich, 1802) and P. oryctolagi Baboš, 1955 under natural conditions in France

Background Protostrongylus oryctolagi and P. pulmonalis are causative agents of pulmonary protostrongyliasis in Lagomorphs in France. These nematodes need usually one intermediate host for its life cycle, a terrestrial snail. However, some studies, mainly in experimental conditions, have identified the species of snails acting as intermediate hosts. Methods In total, 3315 terrestrial snails and 307 slugs were collected in the field in South-Eastern France and analyzed to detect the presence of parasites. Identification of nematode parasites and snails were performed according to morphological and molecular approaches (D2 domain of the 28S rDNA for parasites; 18S and ITS-1 rDNA, COI and 16S mtDNA for snails). Results Eighteen snails were found positive for Protostrongylids larvae. Haplotypes of the larvae corresponding to sequences of P. oryctolagi and P. pulmonalis were detected. Morphological identification of molluscs based on shell characters revealed 4 different morphotypes, and molecular results confirm the membership of these gastropods to the Hygromiidae and revealed 4 different species: Candidula gigaxii, 2 species of Cernuella sp. and Xeropicta derbentina. All infested snails were collected in wine cultures. Conclusion This study displays the first description of intermediate hosts of P. oryctolagi and the first report of X. derbentina as natural intermediate host of P. pulmonalis.


Background
The superfamily of Metastrongyloidea is composed of about 181 species divided into seven families, most of which uses gastropods (generally terrestrial) as intermediate hosts in their life-cycle for the infestation of definitive hosts [1,2].
Since the discovery of the role of snails in the development of Muellerius capillaris by Hobmaier and Hobmaier in 1929 [3], several studies have been implemented in order to identify intermediate hosts and to describe the infestation pathways. Life cycle of Angiostrongylidae has been largely studied, because of its involvement in a human infestation [4][5][6]. The intermediate hosts of Protostrongylidae of domestic ruminants, parasites involved in diseases affecting pulmonary tract of livestock and responsible for economic losses [7], have also been well studied [8][9][10][11][12]. In contrast, developments of Protostrongylidae of Lagomorphs have been studied only in USA [13,14]. Yet, pulmonary protostrongyliasis, disease caused by these nematodes, is frequently encountered and occasionally implicated in the cyclic decline of hare populations in Europe [15,16].
Life-cycle of Protostrongylidae includes a pulmonary adult stage in the definitive host. After emission, the first-stage larvae rejoins the intermediate host, snail or slug, by active or passive penetration [8,9,[17][18][19][20][21][22][23][24]. Unlike the trematodes, first-stage larvae of Protostrongylidae migrate to the muscle of foot where they grow to the third-stage larvae in 25 to 30 days, depending on the external temperature [25,26]. Several land snails and slugs have previously been identified as intermediate hosts of Protostrongylidae [1,27,28]. Using experimental approaches, the authors considered that very few species of snail and slugs are refractory to infection by Protostrongylidae. However there are susceptibility differences depending on species, age, density, activity period of gastropods and mobility of first-stage larvae [29][30][31][32][33]. Actually, young of large size snail species with a life of several years were more susceptible to infection than adults and reverse for little species with a life of one year [34,35]. In addition, a same snails' species can be implicated both in the life cycle of Protostrongylus parasiting hares, and in the cycle of other parasites. For example, the snail species Oxyloma elegans (Risso, 1826) can be infested by P. pulmonalis (parasite of Lagomorphs) and Muellerius capillaris (parasite of small ruminants) [28].
All larvae stages of species of the same genus and even of different genera may be morphologically indistinguishable. Indeed species identification using morphological features is only possible on male adult worms [1]. Hence, to obtain a diagnostic tool to differentiate firststage or third-stage larvae of Protostrongylidae, a molecular approach must be privileged. Among available molecular markers, several studies on nematode species investigating specific identifications used ribosomal DNA, essentially the internal transcribed spacers (ITS1 and 2), known for their inter-specific variability [36][37][38][39][40], and also D2 domain (part of 28S) [41].
Similarly, species identification of snails is not so trivial, when based essentially on morphological criteria. The morphological polymorphism, inter and intra-specific, is extremely important in terrestrial snails depending on many parameters such as age, season, environmental conditions and biochemical factors [42][43][44][45][46][47]. Recently, using of molecular tools, by studying domains such as internal transcribed spacer and 18S rDNA or the COI and 16S mDNA, have facilitated the recognition of snails at a specific level. The combination of the morphological and molecular approaches has helped with the identification of snails difficult to distinguish purely on morphological criteria or only characterizable by specialized malacologists [48][49][50][51][52][53][54][55].
The aim of our study is to investigate and identify the intermediate hosts of Protostrongylids in natural conditions, by a morphological and molecular approach of parasites and snails, to determine the intensity of infestation and the conditions necessary for the development of these parasites in order to better understand their life cycle and assess risk of contamination for hares' populations.

Collection
Living gastropods (snails and slugs) were collected from eight areas of 100 hectares in the South-eastern France ( Figure 1). The sampled sites were selected according to high density of gastropods and hares according to results previously obtained [41], and located in two open environments known to be favorite habitats of hares: grazed grasslands and wine cultures. On each area, samples were realized on 15 circular points of 1 m 2 . For wine areas, points were located in center or at edge of culture. On each point, a raised wooden plank was disposed on one side, providing shelter for snails when it is warm and causes their gathering under it [56]. Collections were implemented monthly from September to November in 2012 and from March to November in 2013 (except in summer, when snails are inactive and buried). The day of the visit was chosen to be a rainy one or after a rainy one, between 7.00 a.m and 9.00 a.m.

Parasite examination
Gastropods collected under planks or on plot, were examined for lungworm larvae. All were stored in separate labeled plastic boxes with ventilation openings and refrigerated (4°C) for a maximum of three weeks. Except for the operculated snails the gastropods were drowned in an airtight jar full of water to kill them and involved the output of the foot out of the snail shell. Foots were recovered and crushed between two heavy glass slides to spread the snail tissue for observation with a stereomicroscope [34,57].
Larvae were observed in some snails in the musculature of the foot. After pre-identification on simple morphological criteria (size, sheath, darkened cuticle) as described by Boev (1984) [1], some larvae were extracted out of muscle, which was easy after dissection or pressing between slide and cover slip. Then they were preserved in 90°ethanol for molecular analysis (Table 1).
After rinsing larvae in distilled water, then drying in an oven at 37°C, DNA extraction was performed with a Qiamp DNA mini kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. Sequencing of the D2 domain of the 28S was used for the identification of nematodes as described previously [41,58] ( Table 2). PCR products were directly sequenced in both directions with the primers used for DNA amplification (Genoscreen, France). Sequence alignment was performed using the ClustalW routine included in the MEGA version 5 software and checked by eye.

Gastropods identification
Shells of positive and negative snails were preserved and primarily identified on morphological and morphometric features according to the collection of continental gastropods, available at the National Museum of Natural History of Paris [59]. In addition, for positive gastropods infested by nematode larvae, a part of foot was conserved in 90°ethanol for molecular analysis (Table 1). DNA extraction of snail was performed using the same method as for nematode larvae. Sequencing of the ITS-1 and 18S of rDNA and mitochondrial COI and 16S were performed using primers and conditions described by   Phylogenetic trees were constructed using the Neighbor Joining (NJ), the Maximum Likelihood (ML), and Minimum Evolution (ME) methods using the MEGA 5 software [61]. For all NJ, ML, and ME analyses, the most appropriate nucleotide substitution model was determined, gaps were treated as missing data and internal node support was assessed by bootstrapping over 500 replicates.

Prevalence
In total, 3622 gastropods were analysed, including 3315 terrestrial snails and 307 slugs. Gastropods were separated depending on their origin, grassland or wine crops. Morphological approach doesn't allow identifying snails at specific level but only at the family or sometimes at the generic level. They belong to two groups and eight families: Cyclophoridae (Prosobranchia) and, from Pulmonata, families of Chondinidae, Clausiliidae, Enidae, Helicidae, Hygromiidae, Vitrinidae and Zonitidae. Out of the 3315 snails, no Prosobranchia were positive for parasites. Eighteen Pulmonata were positive (0.54%), all originated from wine cultures. A total of 123 third-stage larvae were observed and the number of larvae per snail varied from 1 to 52 with an average of 7 larvae.
Identification of slugs was evoked at family level [62,63]. Identification of slugs brings out four families: Agriolimacidae, Arionidae, Limacidae and Milacidae. None slug was positive for protostrongylids larvae.

Identification of larvae
Pre-identification of larvae of Protostrongylids was based on simple morphological criteria. The third stage (infective) larva of this group is generally covered by a double sheath. The upper sheath is thin, transparent, and the lower sheath is compact and rugose with rough traverse folds of brownish color. The larvae along with its sheath usually coil into a ring. (Figure 2) [1,12].
Larvae were then analyzed using molecular markers (D2) in order to identify them at the specific level. For seventeen positive snails, sequences of the larvae are 100% homologous with the haplotype of P. oryctolagi, whereas one snail was positive for larvae whose haplotype corresponds with P. pulmonalis deposited in GenBank.

Identification of positive snails
By pre-morphological identification based solely on shell, the 18 positive snails were all included in the Helicidae and Hygromiidae families, and separated into 4 groups corresponding to their morphotype: A) the first group of white snails including genera such as Xeropicta and Theba (8 snails); B) a group including snails with morphological features of the genus Cernuella (7 snails); C) a group including snails which morphotype corresponds with the genus Candidula (2 snails); D) the last group with one postive snail whose morphological criteria seems to correspond to the Helicella genus. Specimens of each group were found infected with P. oryctolagi larvae, except snails in group A in which the two species of parasites were found.
For each group, two snails were used for molecular analyses (A: sequences EV1 and EV2; B: seq. AT10 and ET3; C: seq. EA1 and ET1), except for the Helicella group where tests failed. We did not take into account this snail for our next results.
An initial study of the combined dataset of 18S, ITS-1, COI and 16S sequences show the membership of all our positive snails to the Hygromiidae family ( Figure 3) and the separation in three different clades corresponding to the different morphotypes. A second study, by analysing the specific 16S and COI allows us to identify some of the species involved in the different groups ( Figure 4). Groupe A is only composed of the species Xeropicta

Discussion
Lungworm infections are frequently found in hares in France [41], but few studies have been conducted to look for intermediate hosts of these parasites in natural conditions. This approach is, however, essential for understanding the development of the parasite life cycle, and thus to evaluate the risk of dispersion and contamination of the parasite in populations of definitive hosts. Our study aims to find these hosts in natura for different species of protostrongylids previously isolated from lungs of hares in France, in order to identify the host-parasite relationship and more precisely the gastropod-nematode relationship, by using techniques previously tested for Trematodes [64][65][66].
Our research of potential intermediate hosts (Gastropods) was therefore carried out on the sites where positive hares had been located. It is important to note that the collection method used here is not suitable for small snail species approx a millimeter in size, living in ground litter and moving very little and very slowly, and already identified  as an intermediate host of Protostrongylidae, e. g: Euconulus fulvus or Vertigo alpestris [28]. Hence we cannot exclude the possibility that we missed some small snail species infested by Protostrongylids, meaning that the number of intermediate host species could be higher than five. Among the 3622 gastropods isolated, no slug was found positive after experimental infection, even if some specimens were previously reported as intermediate hosts.
For terrestrial snails, eight different families were isolated, but only snails belonging to the Hygromiidae were found positive in natural conditions for Protostrongylids larvae. Other families of snails among the Pulmonata have been referenced in the literature, but always experimentally [1,27,28]. The method used to search for larvae in molluscs is different from that usually performed. If using pepsin-digestion is commonly done [67][68][69][70], especially for large molluscs, we prefer the pressing method of snails between two glass plates for small specimens. We do not compare these two methods here. The proportion of infested snails is low (0.54%) but similar to other previous studies [13,71,72]. The number of larvae per snail varied from 1 to 52 with an average of 7 larvae per snail. The mortality rate is higher among the snails which harbour a considerable number of Neostrongylus linearis than the non-infected or moderately infected molluscs [73]. We can suppose that heavily infested individuals have not been collected during this study. Molecular analyzes of the larvae confirm the presence of two species of Protostrongylus, P. oryctolagi and P. pulmonalis, the same species as those previously found on hares in France.
The analysis of the positive snails based on the morphological criteria of the shell highlighted 4 groups, limited at the family or the genus level. Identification at the specific level is extremely difficult because of the important intra and inter-specific polymorphism. To be free from the constraints related to morphological identification, we used molecular tools, as previously described, to identify species incriminated as intermediate hosts [54,55,60].
Larvae of P. pulmonalis was isolated from one individual belonging to the Xeropicta derbentina species. Larvae of P. oryctolagi were observed in four different species of snails: Candidula gigaxii (Pfeiffer, 1850), Xeropicta derbentina Brusina, 1870 and two Cernuella sp. If molecular results confirm the monophyly of the genus Cernuella, the haplotype Helicella obvia sensu Di Napoli et al. [54] most probably belong to the species Cernuella neglecta [74], we will not conclude at the specific level for both haplotypes isolated during our study. In fact, due to the high intra-specific morphological variability for this genus, sometimes including the internal features, identification is often very difficult. Under these conditions and based on the ambiguous molecular results, we consider these two haplotypes as Cernuella sp. pending a more comprehensive study. This is the first report of all these snail species as natural intermediate host of Protostrongylus of Lagomorphs, except Xeropicta derbentina already infested experimentally by Protostrongylus tauricus [1].
Protostrongylidae are not highly specific in their use of intermediate hosts since many species of snail have been shown to be involved in its development [32,75]  in Provence, mainly pasture fields and vineyards, where populations of snails are often very dense [77][78][79].
In our study, all larvae of Protostrongylus were isolated from snails collected in wine areas. All four species of intermediate hosts were also observed in grassland areas, but their abundance seems lower. It is possible that the wine crops are better adapted to the environment allowing snails to release larvae or offer optimal conditions for larvae survival. Third-stage larvae of Muellerius capillaris remain infective for at least six months at −12°C [80], but are supposed to be very sensitive to dry conditions [81] and the wine areas, drier than grassland, do not seem best suited for larval survival. However, third-stage larvae, which are infective for the definitive host by ingestion, do not usually leave the foot when the snail is alive but their output is activated when it dies [75]. We can suggest that chemical treatments used in wine crops could modulate levels of parasitic infestation in snail populations. Indeed, Hock and Poulin [82] observed that pollutants can lead to increase cercariae excretion from aquatic snails.
An experimental study should be initiated to verify whether larvae excretion is facilitated by use of treatments, causing intense drooling and death by desiccation of snails.

Conclusion
Using a combined morphological and molecular approach for parasites and snails, this study displays the first report of natural intermediate hosts of P. oryctolagi and P. pulmonalis, two parasitic nematodes responsible for pulmonary protostrongyliasis in France, disease frequently encountered in hare populations in Europe. Understanding the life cycle of this kind of parasites is essential for risk factor identification. Among the two different habitats sampled, the wine crops seem favourable to the life cycle of hare Protostrongylus. In the future, it would be interesting to implement investigation in other environments in order to identify all habitats representing potential risks of disease for hare populations.
Competing interests I confirm that I have read BioMed Central's guidance on competing interests and that none of the authors have any competing interests in the manuscript.
Authors' contributions AD was responsible for SAGIR network and providing samples. SD realized the morphological identification of snails. CL, CP, HF and DJ analyzed the samples and molecular data and wrote the manuscript. All authors read and approved the final version of the manuscript.