Introduction

Various histopathological investigations have been made on digenean larvae in their molluscan hosts (James 1965; Wright 1966; Reader 1971; Huffman and Fried 1985). Previous studies have been concerned with larval trematode infections in various molluscan hosts (Hurst 1927; Pratt and Barton 1941; Cheng 1963a, b; Porter et al. 1967), but none has examined the relationship between Helisoma trivolvis (also known as Planorbella trivolvis) and its larval trematodes. H. trivolvis, a ubiquitous planorbid in North America, is infected with numerous species of larval trematodes (see review in Klockars et al. 2007). Of the larval trematodes reported in H. trivolvis, three occur in NJ and PA (Schmidt and Fried 1997; Klockars et al. 2007). These are Echinostoma trivolvis, Ribeiroia ondatrae, and Zygocotyle lunata. Some incidental information on the histopathology of these species was provided by Huffman and Fried (1990) for E. trivolvis, by Johnson et al. (2004) for R. ondatrae, and by Willey (1941) for Z. lunata; none of the aforementioned papers has detailed the histopathology and cellular reaction in the digestive gland of H. trivolvis in response to daughter rediae and cercariae of these larval trematodes. The purpose of our study is to provide information on the histopathological effects of these trematodes on the digestive gland of H. trivolvis. Incidental to the histopathology study, we made observations on the histochemical presence of keratin in the intramolluscan larval stages, and we report these observations herein. The role of keratin in the intramolluscan larval stages of trematodes is uncertain, although this protein may play a role in cercarial encystment.

Materials and methods

Specimens of H. trivolvis were collected from Delaware Pond, Warren County, NJ, USA (40° 55′ 19.1″ N; 75° 03′ 49.5″ W). The geographic coordinates of the collection site were determined by a global positioning system (Garmin eTrex Legend). Snails were collected by hand from the perimeter of the pond, no more than 1.5 m from the edge. The snails were isolated to determine infection with larval trematodes within 48 h of collection by placing them in individual wells in plastic multiwell trays (Costar Corporation, Cambridge, MA, USA) containing 3 ml of natural spring water in each well. Two 50-W bulbs were placed approximately 30 cm from the trays to maintain the snails at 28–29°C. The trays were examined up to 6 h after snail isolation for cercariae. Live cercariae were examined unstained or stained with 0.01% neutral red. Cercariae of E. trivolvis and Z. lunata were identified by one of us (BF), who has worked with these cercariae for many years. The gymnocephalus cercaria was identified as R. ondatrae by its unique morphological characteristics (Johnson et al. 2004) and by molecular analysis (Wilson et al. 2005). DNA was extracted from the cercariae using the MoBio Ultraclean Tissue DNA Isolation Kit (MoBio Laboratories Inc., Carlsbad, CA, USA). The intertranscribed spacer region 2 (ITS-2) of the ribosomal gene complex was targeted using the primer set 3 S (5′ GGT ACC GGT GGA TCA CGT GGC TAG TG 3′; Bowles et al. 1995) and ITS2.2 (5′ CCT GGT TAG TTT CTT TTC CTC CGC 3′ Hugall et al. 1999) as described in Wilson et al. (2005). Promega Master Mix (Promega Corporation, Madison WI, USA) was used for all polymerase chain reactions (PCR) at 25-µl reaction volume with template and primer concentrations based on manufacturer’s recommendations. PCR product was visualized using a 3% agarose gel for the presence of a 429 bp. The PCR product was sequenced using the Big Dye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems, Forest City, CA, USA) on the AB 3130 Genetic Analyzer (Applied Biosystems, Forest City, CA, USA), following the manufacturer’s protocol. The sequences were compared to GenBank sequences using the basic local alignment tool (BLAST; Altschul et al. 1990) based on the nucleotide–nucleotide BLAST search (blastn) with the default settings. BioEdit Sequence Alignment Editor was used to directly compare the project sequences to R. ondatrae ITS-2 sequences from GenBank (Hall 1999). The sequences were found to be identical to the ITS-2 region of R. ondatrae (Wilson et al. 2005).

Twenty snails were divided into four groups with five snails per group. Three groups contained snails infected with either E. trivolvis, Z. lunata, or R. ondatrae, and the fourth contained snails that were not infected with larval trematodes. All infected snails showed patent larval trematode infections and snails with prepatent infections were not used. During the course of this study, histopathological and histochemical observations were made on tissue sections from most snails in all four groups. Parasitized and nonparasitized snails were removed from their shells and fixed in 10% neutral buffered formalin. For the histopathological studies, tissues were dehydrated in an alcohol series, embedded in paraffin, sectioned at 6 μm, and stained with hematoxylin and eosin. To determine the histochemical presence of keratin, the tissues were dehydrated in an alcohol series, embedded in paraffin, sectioned at 6 μm, and stained with the Ayoub-Shklar method for keratin (Luna 1968).

Tissue damage was evaluated according to criteria established by Malek and Cheng (1974), including mechanical compression and lysis. Cell types were characterized using the criteria of Cheng (1983), granulocytes (variable in size from 12 to 25 μm, with a low nucleus to cytoplasm ratio and the presence of numerous large granules), and hyalinocytes (small, 6–13 um, cells with a high nucleus to cytoplasm ratio, and few granules).

Results

The uninfected digestive gland of H. trivolvis was brown in color and located in the posterior third of the snail. Just beyond the digestive gland was the gonad. However, the gonads of the infected snails were riddled with the intramolluscan stages of the three species of larval trematodes (parasitic castration), and gonadal tissue was seldom apparent. Hence, histopathological observations on the gonads were not made. The digestive gland of the uninfected snails contained numerous tubules surrounded by loose connective tissue containing the visceral hemocoelic space. The tubules were separated from the hemocoel by a thin layer of loose connective tissue and lined with glandular epithelium (Fig. 1). The epithelium was composed of two cell types, serous and mucous cells, and the serous cells were triangular. The cytoplasm at the base of each cell was basophilic, and the nucleus was round and situated close to the base. The cytoplasm toward the apex of the serous cells contained eosinophilic granules. The mucous cells were columnar with flattened nuclei which were crowded against the base of the cells. There was less basophilia at the base of these cells than in serous cells. The periphery of the digestive gland contained muscle. A membrane, comprised of a single squamous epithelium layer overlying a connective tissue layer, enclosed the digestive gland.

Fig. 1
figure 1

The digestive gland of an uninfected snail with numerous tubules surrounded by loose connective tissue containing the visceral hemocoelic space. The tubules (T) were separated from the hemocoel by a thin layer of loose connective tissue and lined with glandular epithelium

Full size image

The digestive glands of H. trivolvis infected with the rediae of E. trivolvis, Z. lunata, and R. ondatrae were mottled with these intramolluscan stages, and the characteristic appearance of each species of redia was readily apparent in the digestive gland. Gross observations of the rediae of E. trivolvis were given in Huffman and Fried (1990) of the rediae of R. ondatrae in Johnson et al. (2004), and of Z. lunata rediae in Willey (1941).

Z. lunata infection of H. trivolvis caused a disruption of the digestive glands; mechanical compression and rupture of the digestive cells were also evident (Fig. 2). A hemocytic infiltration of host tissues surrounding the larval parasites was present. There was no rupture of the tunica propria. The rediae of E. trivolvis were surrounded by clear zones devoid of cells (Fig. 3a). There was a detectable hemocytic response on the part of the molluscan host to the presence of the rediae. Lysis of host cells, mechanical compression, and displacement were seen in the digestive cells (Fig. 3b). The digestive glands were completely disrupted. No rupture of the tunica propria was seen. The rediae of R. ondatrae induced a hemocytic response. Mechanical compression and displacement of host tissues and lysis of cells were also observed (Fig. 4). No rupture of the tunica propria was seen.

Fig. 2
figure 2

Z. lunata infection of H. trivolvis caused a disruption of digestive gland tissue, mechanical compression (arrow), and rupture of the digestive cells

Full size image
Fig. 3
figure 3

a The redia (R) of E. trivolvis was surrounded by clear zones devoid of cells (arrow). A hyalinocyte reaction occurred in response to the redia. b Lysis (L) of host cells, mechanical compression, and displacement were seen in the digestive cells

Full size image
Fig. 4
figure 4

The rediae (R) and cercariae of R. ondatrae induced a hemocytic response. Mechanical compression (arrow) and displacement of host tissues and lysis of cells were observed. No rupture of the tunica propria was seen

Full size image

Digestive gland cells in both infected and uninfected snails contained keratin. A heavy keratin response occurred in the rediae and cercariae of Z. lunata; a moderate response was present in R. ondatrae rediae and cercariae; no keratin was present in E. trivolvis rediae or cercariae.

Discussion

Despite numerous studies of the gastropod digestive gland, there is still uncertainty as to the terminology and classification of cell types. Pan (1958) described three cell types in the digestive gland of Biomphalaria glabrata: goblet, lime, and digestive. Porter et al. (1967) recognized two types in Oxytrema siliqua: liver and calcium cells. Reader (1971) recognized three types in Bithynia tentaculata: absorptive, mucous, and basophilic cells.

Mucous and secretory cells were present in H. trivolvis. The epithelial cells of the digestive gland in this gastropod have morphological characteristics typical of mucous and serous secreting columnar epithelium. These two morphologically distinct types are further differentiated on the basis of their cytoplasmic constituents.

Numerous studies have been made on the destruction of molluscan digestive glands by larval trematodes (Hurst 1927; Pratt and Barton 1941; Porter et al. 1967; Reader 1971; Huffman and Fried 1985). In the present study, we noted certain histopathological effects of the rediae and cercariae of Z. lunata, E. trivolvis, and R. ondatrae on the digestive gland of H. trivolvis.

One difficulty in studying naturally infected snails is that the age of the infection cannot be determined nor can the sequence of the pathology. The primary method of cell destruction in the digestive glands by the three species of trematodes we studied appeared to be through ingestion of host cells. Cheng (1963b) observed similar changes in H. trivolvis infected with Echinoparyphium sp. rediae. All species of rediae in our study exerted mechanical pressure on the digestive gland tubules as evidenced by constricted tubular lumina. It is also possible that the excretory products of the rediae had a lytic effect on host tissue, as evidenced by the clear zones surrounding the parasite. The clear zones were devoid of cells and were probably edematous.

Many workers have noted little or no cellular response to trematode larvae in molluscan hosts (Cheng 1963a; Cheng and Burton 1965; James 1965). Mechanisms by which larval trematodes inhibit the ability of the mollusk to recognize them as foreign are still poorly understood. In the present study, a hemocytic response was noted. Z. lunata elicited a distinct hemocytic response in H. trivolvis. Borges et al. (1998) reported the presence of amebocytes (hemocytes) in the ovotestis and tubular kidney of B. glabrata snails in response to Schistosoma mansoni. In contrast to the work of Borges et al. (1998), most other studies have reported little or no cellular response to larval trematodes in their molluscan hosts (Cheng 1963a; James 1965).

The primary function of keratin, a fibrous protein, is to protect epithelial cells from mechanical and nonmechanical stresses. Other functions include a stress response and apoptosis, as well as unique roles that are keratin specific and tissue specific (Coulombe and Omary 2002). Keratin was found within the rediae and cercariae of Z. lunata and R. ondatrae but not in E. trivolvis. This protein was reported by Dixon (1965) in the metacercarial cyst wall of Fasciola hepatica and in the rediae and cercariae of Philophthalmus megalurus by Huffman and Fried (1985). The latter two parasites encyst on vegetation or on the surfaces of other substrates. Z. lunata also encysts on vegetation and shell surfaces, and R. ondatrae encysts in the lateral line of fish or in the limb buds of ranid tadpoles; E. trivolvis encysts in the kidneys of snails and ranid tadpoles. The presence of keratin in the larval stages of Z. lunata and R. ondatrae may indicate that this protein is involved in cercarial encystment of these two digeneans.