INTRODUCTION
Myxosporeans (Myxozoa) are diverse and widely distributed microscopic, metazoan parasites, known to parasitize mainly aquatic, poikilothermic vertebrates (Lom and Dyková, Reference Lom and Dyková2006). Although myxosporeans parasitize predominantly fish, a number of species of the genera Caudomyxum, Chloromyxum, Hoferellus, Myxidium, Myxobolus and Sphaerospora affect amphibian hosts (Eiras, Reference Eiras2005). Five species of renal myxosporeans are described from amphibians. In 1893, renal myxosporidiasis was reported for the first time from amphibians by Ohlmacher in Bufo lentiginosus. In the same year, Whinery (Reference Whinery1893) described this coelozoic myxosporean from the same host from Illinois and named it Chloromyxum ohlmacheri. In review of Ohlmacher's (Reference Ohlmacher1893) and Whinery's (Reference Whinery1893) articles, Gurley (Reference Gurley1894) referred to this species as C. (Sphaerospora) ohlmacheri. Finally, C. ohlmacheri was redescribed by Desser et al. (Reference Desser, Lom and Dyková1986), who provided a description of its sporogonic stages, and placed the species into the genus Sphaerospora, based on the material originating from tadpoles of Rana catesbeiana from Canada. In European amphibians, myxosporean spores were first recorded from the kidneys of Rana temporaria (in original text referred to as Rana fusca), and Rana kl. esculenta from France by Thélohan (Reference Thélohan1895). Although the parasite was named Leptotheca ranae by Thélohan in Reference Thélohan1895, no adequate description was provided, and L. ranae became a nomen nudum. In 1920, Kudo considered European L. ranae, reported from ranids (but not properly described at that time) to be conspecific with N. American C. ohlmacheri and provisionally placed this myxosporean to the genus Wardia (for details see Desser et al. Reference Desser, Lom and Dyková1986). Ten years later, Morelle (Reference Morelle1929) found a coelozoic myxosporean, similar to that reported by Thélohan (Reference Thélohan1895), in the renal tubules of Rana temporaria in Belgium. These were presumed to be conspecific, and Morelle (Reference Morelle1929) described the parasite using the available name Leptotheca ranae, so the name became valid. Finally, in a recent review of Myxozoa in amphibians and reptiles, Eiras (Reference Eiras2005) assumed L. ranae to be a nomen nudum without further comments.
Another coelozoic, renal myxosporean from European amphibians, Chloromyxum protei was precisely described from Proteus anguinus by Joseph (Reference Joseph1905). Furthermore, Mutschmann (Reference Mutschmann1999) described renal, coelozoic Chloromyxum careni from the Megophrys nasuta from Indonesia, and Duncan et al. (Reference Duncan, Garner, Bartholomew, Reichard and Nordhausen2004) reported mortality, caused probably by the same species in M. nasuta. Recently, renal (histozoic) myxosporean, Hoferellus anurae was described from African hyperoliid frogs, together with a description of pathological changes and introduction of the name for the disease (frog kidney enlargement disease, FKED) (Mutschmann, Reference Mutschmann2004).
Recently, L. ranae was included in the phylogenetic analysis for the first time by Fiala (Reference Fiala2006), based on the same material as this study. In his analysis, in addition to the 2 main lineages recognized by most authors (the marine and freshwater) (for review see Fiala, Reference Fiala2006), a third lineage consisting of 3 freshwater species (Sphaerospora elegans, Sphaerospora truttae, L. ranae) was found. This minor clade represents the basal branch to all (sequenced) myxosporean species.
In our study, we report on the rediscovery of a renal myxosporean in European ranids from the Czech Republic, suggest its conspecificity with L. ranae, provide first the description of its sporogonic stages, a redescription of its spores, emend its taxonomic status and provide phylogenetic analysis inferred from the small-subunit ribosomal DNA (SSU rDNA). In general, the aims of this study were (i) to provide missing information about L. ranae belonging to a unique, recently recognized phylogenetic clade, basal to all other myxosporeans, (ii) to transfer L. ranae to the genus Sphaerospora, and (iii) emend the definition of the genus Sphaerospora.
MATERIALS AND METHODS
Wild-caught animals
During 2000–2005, both adults and larvae of 5 amphibian species from the Czech Republic were examined as part of our long-term research on amphibian parasites: Rana dalmatina, R. temporaria, Bufo bufo, Hyla arborea, Triturus vulgaris. All adult amphibians were collected by hand at night during the breeding season (March, April), placed separately in damp collection plastic boxes and returned to the laboratory within 1 h for processing. Individual amphibians, placed in separate boxes, were kept in 4–6°C for 2–3 days in 2–3 mm of water, with 12/12 h day/night photo-period. Animals were euthanized by pithing within 3 days after collection and processed for routine parasitological examination. The body cavity was opened by longitudinal incision beginning at the lower abdomen and ending at the pectoral girdle region. Blood smears were taken from the exposed heart and stained with Giemsa. The entire gastrointestinal tract, including the liver and gallbladder, was removed, all organs were opened longitudinally and examined for parasites. The lungs, kidneys, spleen, urinary bladder, reproductive organs and coelomic cavity were also examined. From each tissue, both squash preparations and histological preparations were obtained from each dissected specimen. Additionally, methanol fixed kidney smears were prepared and stained with Giemsa. Larval amphibians were collected using a dip-net throughout their development period (April–August). All larvae were placed separately in 100 ml vials in dechlorinated tap water, returned to the laboratory within 1 h, and kept in room temperature for at least 12 h exposed to daylight. Fecal sediment (removed using a Pasteur pipette) from each vial was examined microscopically after concentration by flotation with modified Sheather's sugar solution (s.g. 1·30). Individual larvae were dissected after being examined coprologically as described for adults. Host species were determined according to descriptions reported by Nečas et al. (Reference Nečas, Modrý and Zavadil1997).
Origin of positive samples
Locality 1.: Zaječí potok, Brno, Czech Republic, (16°36′23·07″E, 49°14′15·57″N), 303 m asl.; small forest pond with rich assemblage of macrophytic submerse vegetation, surrounded mainly by broadleaf mixed forest. Amphibians and reptiles recorded (numbers in parentheses indicate number of larvae examined/number of adult amphibians examined): R. dalmatina (200/32), R. temporaria (200/30), H. arborea (30/0), B. bufo (200/20), Triturus alpestris (0/0), T. vulgaris (20/16), Natrix natrix. No fish species occur at the locality. Locality 2.: Růženčin lom, Brno, Czech Republic, 16°40′22·69″E, 49°13′0·78″N, 355 m asl.; several small ponds in large abandoned limestone mine in oak forest. Amphibians and reptiles recorded: R. dalmatina (0/6), B. viridis (0/0), Lacerta agilis. Fish recorded: Rutilus rutilus. Locality 3. Jedovnice, productive lake in the vicinity of the village, Czech Republic, 16°46′30·51″E, 49°20′0·71″N, 484 m asl.; large artificial lake, surrounded by artificial coniferous (Picea abies) forest and wet meadows mosaic. Amphibians and reptiles recorded: R. temporaria (0/18), B. bufo (0/0), N. natrix. Fish recorded: no exact data, at least several cyprinid fish species are reared.
Light microscopy
Small pieces of all examined tissues were compressed between microscope slide and cover-slip and examined as wet mount preparations. Samples for histology were fixed for at least 10 days in 10% buffered formalin, processed for histology as paraffin sections (4–6 μm) and stained with haematoxylin and eosin (H&E). Alternatively, tissues were fixed with 2·5% glutaraldehyde, post-fixed with OsO4, processed in graded acetone series, embedded in resin, stained with toluidine blue and examined as semi-thin (400 nm) sections. Measurements and photographing of all sporogonic stages (except mature spores) were made on stained histological sections, mature spores were examined in squash preparations using Nomarski interference-contrast optics (NIC). Olympus AX 70 light microscope equipped with a calibrated ocular micrometer and DP 70 digital camera was used during the study. All measurements are reported in micrometers (μm). In the morphological section of this paper, we consistently use terms such as valvular, sutural and polar view, and define these terms as follows. Valvular view: perpendicular to sutural plane. Sutural view: perpendicular to longitudinal axis of the spore, parallel with the sutural plane. Polar view: parallel with the longitudinal axis of the spore.
DNA isolation, cloning and sequencing
The total DNA was extracted from infected kidney tissue using the DNeasy TM Tissue Kit (Qiagen, Germany) according to the manufacturer's protocol. Myxozoan specific primers, MyxospecF (5′-TTCTGCCCTATCAACTWGTTG-3′) and MyxospecR (5′-GGTTTCNCDGRGGGMCCAAC-3′) (Fiala, Reference Fiala2006), were used to amplify partial SSU rDNA. PCR was carried out in a 25 μl reaction volume using 10 pmol of each primer, 250 μm of each dNTP, and 2·5 μl 10× PCR Buffer (Top-Bio, Czech Republic) and 1 unit of Taq-Purple polymerase (Top-Bio, Czech Republic). The reactions were run on a Tpersonal cycler (Biometra). Amplification consisted of 30 cycles of 95°C for 1 min, 50°C for 1 min and 72°C for 1 min, then a 10 min incubation at 72°C. The PCR product was isolated from the gel and cloned into pCR® 2.1 TOPO Cloning vector using the TOPO-TA Cloning Kit (Invitrogen, USA). Clones were sequenced from both strands on automatic sequencer CEQTM 2000 (Beckman Coulter, USA) using CEQ DTCS Dye Kit (Beckman Coulter, USA) according to the manufacturer's protocol.
Phylogenetic analysis
The partial SSU rDNA sequence of S. ranae was aligned using the Clustal_X program (Thompson et al. Reference Thompson, Gibson, Plewniak, Jeanmougin and Higgins1997) with 47 myxosporean sequences retrieved from the GenBank. The dataset was aligned with the gap opening/gap extension penalty=8/2 and incongruities were corrected in BioEdit sequence alignment editor (Hall, Reference Hall1999). Buddenbrockia plumatellae and Tetracapsuloides bryosalmonae were set as outgroup.
Phylogenetic analyses were done using the maximum parsimony (MP) and maximum likelihood (ML) methods in the program package PAUP* version 4.0b10 (Swofford, Reference Swofford2001). MP analysis was performed by a heuristic search with tree bisection-reconnection (TBR) branch swapping and 10 random-addition replicates. Gaps were treated as missing data. The matrix was analysed under the transversion/transition (Ts:Tv) ratio 1:2. Bootstrap statistical support was carried out with 1000 replications of heuristic search. For the ML analysis, the likelihood ratio test (LRT) implemented in Modeltest v. 3.6 (Posada and Crandall, Reference Posada and Crandall1998) was used to determine the best model of evolution. Based on the LRT, the ML was performed with the GTR+I+г model of evolution. Clade support was assessed with bootstrapping of 500 replicates.
RESULTS
Myxosporean infections were found in 2 species of frogs of the genus Rana originating from 3 localities in the Czech Republic. This is the first record of myxosporean infection in European amphibians after 78 years. Examination of frog kidneys revealed sporogonic stages of Sphaerospora ranae (Morelle, Reference Morelle1929) n. comb.
Redescription of Sphaerospora ranae (Morelle, Reference Morelle1929) n. comb
Morphology of sporogonic stages:
Only sporogonic stages were observed (Figs 1, 2 and 3). In histological sections, plasmodia of S. ranae often adhere to the microvillous layer of the renal tubular epithelium (Fig. 1D). Pseudopodia-like projections are commonly present on a surface of both early (Fig. 2C) and mature (Fig. 1D) plasmodia. The earliest sporogonic stages (plasmodia) are uninucleate, up to 10 in length, usually occurring in groups (Fig. 2A). Advanced stages were more numerous, represented by immature plasmodia at various stages of development, 12–15 in cross-section, containing a variable number of nuclei and/or secondary and tertiary cells within the enveloping cell (Fig. 2B, C, D, E). There was no evidence of clearly discernible pansporoblasts within plasmodia in paraffin and resin sections. Formations composed of 4 cells contained within another cell(s), representing generative cells enveloped either by valvogenic cells, or possibly by pericytes, were observed occasionally (Fig. 2F). Clearly, these cell-complexes are early stages of spore formation. However, in light microscopy, it was not possible to assess whether these stages represent young pansporoblasts or early spores developing directly by division of generative cells. Formation of these stages, is followed by more advanced spores possessing a distinct ridge along the longitudinal suture on each valve (Fig. 2G, I). In more advanced spores, nuclei of valvogenic cells become indiscernible in histological sections (Fig. 2H, I). After apparent disappearance of the valvogenic cell nuclei, valvogenic cells form a postero-lateral bulge on each spore valve (Fig. 2J, K, L, M). In sections, postero-lateral spore bulges possess lucent vacuoles, seemingly opening outwards, as appears in some spores (Fig. 2J, K). A bilobed lucent structure, about 2–2·5 in breadth, adheres to the posterior part of the young spores still contained within plasmodia (Fig. 2J, K, L, M). This structure is absent in fully mature spores released from plasmodia (Fig. 3B, C, G, 4).
Fig. 1. General aspects of Sphaerospora ranae infection of the kidneys of Rana dalmatina. (B–D) Resin sections stained with Toluidine blue. (A) Infected kidney with several tubules occluded by various sporogonic stages (arrows), paraffin section stained with H&E. (B) Renal tubule occluded by plasmodia possessing mature spores. (C) Glomerular infection (2 infected glomeruli). Both plasmodia (arrow) and free spores are present (arrowheads) within Bowman's capsules. (D) Detail of plasmodia. Two plasmodia are firmly attached (arrowheads) to the microvillous zone of the tubular epithelium. Note numerous pseudopodia-like projections on the surface of plasmodia (arrows).
Fig. 2. Sporogonic stages of Sphaerospora ranae. (A–E) Paraffin sections stained with H&E. (F–K) Resin sections stained with Toluidine blue. (L) Squash preparation, NIC. All (except M.) are the same magnification. (A) A group of early (uninuclear) sporogonic stages. (B) More advanced tetranuclear, undifferentiated plasmodium. (C) Early undifferentiated plasmodium showing pseudopodia-like projections. (D) Early undifferentiated plasmodium attached to the microvillous zone of tubular epithelium. (E) More advanced plasmodium containing numerous secondary and/or tertiary cells. (F) Initial phase of the spore formation. Six-cell complex composed of 2 valvogenic, 2 capsulogenic, 2 sporoplasmic cells. Note that sutural ridge is not developed at this stage. (G) Early spore (more advanced 6-cell complex). All 6 sporogenic cells still contain distinct nuclei. (H) More advanced spore. Nuclei of valvogenic cells are indiscernible at this stage. Note distinct nuclei of both capsulogenic and sporogonic cells. (I) Polar view of spore. Note that valvogenic cells form irregularly folded sheath around the spore surface. (J) Young spore with distinct sutural ridge and reduced valvogenic cells forming postero-lateral bulge (arrowhead) on each spore valve. Note bilobed lucent structure adhering to the posterior part of the spore (asterisk). (K) Mature spore diagonally cross-sectioned outside the sutural plane showing the same structures as in (J). Postero-lateral spore bulge formed by remnants of valvogenic cell (arrowhead) and lucent structure adhering to the posterior part of the spore (asterisk). (L) Mature spore (mechanically liberated from plasmodium) showing the same structures as in (K). Postero-lateral spore bulge formed by remnants of valvogenic cell(s) (arrowhead) and posterior lucent structure (asterisk). Note that outer surface of valvogenic cells is folded. (M) Schematic line-drawing of spore showing section plane (thick line) of the spore from (K). Postero-lateral spore bulge formed by remnants of valvogenic cell(s) (arrowhead) and posterior lucent structure (asterisk). Arrow indicates view (K) direction.
Fig. 3. Morphological features of mature spores of Sphaerospora ranae. (B–K) Images are the same magnification. (A–I) Squash preparations of kidney, NIC. (J–K) Paraffin sections stained with H&E. (A) Mass of spores of S. ranae in infected kidney tissue. (B–C) Details of mature spores. Note orientation of polar filaments, fine granules within sporoplasm and valvogenic cell(s) thickening at the spore apex between the polar capsules. (D–E) Atypically formed spores. (F) Plasmodium containing 4 (only 3 visible) spores. (G) Detail of mature spores (surrounded by spermatozoa) in seminal vesicle contents showing sutural ridge. Note that nuclei of capsulgenic cells are still discernible (arrowhead). (H) Polar view of the spore showing postero-lateral bulges formed by valvogenic cells (arrowhead). (I) Sutural view of the spore showing fine granulation of valvogenic cell(s) cytoplasm (arrowhead). (J) Polar view of the spore sectioned at its posterior part. Note distinct sutural ridge (arrowheads) and folds on the spore surface (arrow). (K) Polar view of the spore sectioned at its apical part. Note that the spore surface is smooth in this apical part of the spore.
Mature spores:
Mature spores (Figs 3 and 4) consist of 2 identical valves joined by a longitudinal suture. From the sutural view, the spores are broadly subspherical or rarely spherical (width/length ratio: 0·90–1·36), roundly pointed anteriorly with 2 distinct postero-lateral irregular bulges (1–2·5 in breadth) formed by valvogenic cells, resulting in a somewhat heart-shaped spore. From the valvular and polar view, the spores are ellipsoidal (Fig. 4C, and Fig. 3H, J, K respectively). Additionally, the surface of the postero-lateral bulges is irregularly folded as observed both in wet mounts (Fig. 2L) and histological sections (Fig. 3J). The surface of the anterior part of the spore is smooth (Fig. 3K). The sutural line is straight, followed by a conspicuous sutural ridge on each valve, clearly visible in both histological preparations (Fig. 2G, I, J) and wet mounts (Fig. 3G). There is a valvular thickening at the apical end of each valve between the 2 polar capsules (Fig. 3B, C). Spore dimensions (Table 1) excluding postero-lateral bulges (R. dalmatina and R. temporaria pooled): 10·1 (9·0–11·0) long, 11·7 (9·0–12·5) wide. The 2 equally sized (4·0–5·0 in diameter) polar capsules are spherical, situated at the pointed apex of the spore, in plane perpendicular to sutural plane. Six to 7 tightly coiled threads of polar filament oriented slightly oblique to the sutural plane, converge towards the spore apex. In Giemsa-stained smears, extruded polar filaments were frequently observed (maximum length 95·0), leaving the spore just beside the sutural ridge at the spore apex. Remnants of capsulogenic cells and especially their nuclei are discernible even within mature spores released from plasmodia (Fig. 3G). Apparently, 2 uninucleate sporoplasm cells occupy most of the remaining space within the spore valves. However, in some spores, there seems to be single, binucleate sporoplasm. A variable number of fine granules is present in the sporoplasm (Fig. 3A, B, C, F, G). Postero-lateral bulges are still present in mature spores (Fig. 3B, C), being somewhat reduced in spores expelled from plasmodia (Fig. 3G). Occasionally, atypical spores with 3 or 4 polar capsules occur (Fig. 3D, E).
Fig. 4. Composite line-drawings of the mature spore of Sphaerospora ranae. (A) Sutural view showing inner composition of the spore. (B) Sutural view showing spore surface. (C) Valvular view showing spore surface.
Squash preparations (Fig. 3F) and paraffin sections (not shown) confirmed that the number of spores within plasmodia varies from 1 to 4, although 1 or 2 spores per plasmodium are visible in most sectioned parasites (Fig. 1D). In most plasmodia, spore development is asynchronous (not shown).
Host spectrum and prevalence
Severe myxosporean infections were detected in 78% (25/32) of R. dalmatina from Locality 1, 50% (3/6) of R. dalmatina from Locality 2, and 17% (3/18) of R. temporaria from Locality 3 (Table 2). All infections were recorded in breeding hosts during their aquatic phenological phase in March and April. No infection was detected in larvae (n=620) of any amphibian species, including R. dalmatina and R. temporaria. No infection was detected in adult R. temporaria (n=30), or any other amphibian species at the Locality 1, where infections are common in R. dalmatina. Comparison of fresh spores and histological examination suggest identity of renal infection in both host species (Table 1).
Table 1. Comparison of spore size of Sphaerospora ranae on the 3 localities with original description
(Note that shorter proportions represent length.)
Table 2. Occurrence of Sphaerospora ranae on the 3 localities
(Prevalence % (number infected/n).)
Localization within host
In both R. dalmatina, and R. temporaria the kidney was the only organ infected. Sporogonic stages of S. ranae (Fig. 1) were localized in the lumen of renal tubules, which were frequently completely occluded by the parasites (Fig. 1A, B). Additionally, glomerular infections were recorded in both species with sporogonic stages (including mature spores) observed in Bowmans' capsules (Fig. 1C). Mature spores released from plasmodia were regularly observed in high quantities in the seminal vesicle (lower part of Wolffian duct expanded during breeding season). Here, the spores sometimes formed large aggregations composed of hundreds of spores. Similar aggregations (as well as single spores) were also detected in sediment in water, in which infected frogs were kept. Histozoic proliferative stages were detected neither in the kidney interstitium nor in blood smears. We did not record significant pathological changes in the tubular epithelium. A focal inflammatory response as evidenced by eosinophilic granule leucocytes was occasionally observed in the kidney interstitium.
Phylogenetic analysis
The length of the partial sequence of S. ranae is 1234 bp with GC content 51·8%. Myxosporean specific primers (MyxospecF and MyxospecR) define the region of SSU rDNA of the length range from 750 to 950 bp in most myxosporean species. The alignment of S. ranae with myxosporean sequences retrieved from the GenBank (1623 included characters, from which 825 were parsimony informative) revealed the presence of 2 areas with extensive nucleotide insertions (expansion segments), in the V4 region within the SSU rDNA of S. ranae. The length of the V4 region is 660 bp (see Table 3. for details and comparisons with other species). The comparison of the V4 region of S. ranae and those of most other myxosporean SSU rDNA sequences (e.g. Ceratomyxa shasta) revealed the V4 region of S. ranae to be significantly longer, 240 bp in the case of C. shasta vs. 660 bp in S. ranae. Similar areas with extensive nucleotide insertions in the V4 region of SSU rDNA are present in the sequences of S. truttae and S. elegans from fish (see Table 3). The lengths of the V4 regions from S. elegans and S. truttae are respectively about 420 and 271 bp longer than that of C. shasta.
Table 3. Comparison of the ranges of V4 regions and the two expansion segments of the SSU rDNA sequences of Sphaerospora elegans, Sphaerospora ranae and Sphaerospora truttae, with their corresponding regions in Ceratomyxa shasta. Numbers in parentheses indicate size (bp) of specific regions.
In both maximum parsimony and maximum likelihood analyses S. ranae clustered with S. elegans, S. truttae and Leptotheca fugu (Fig. 5). In L. fugu, no areas with extensive nucleotide insertions are present within the V4 region. The relationship was supported by high bootstrap support. The 2 areas with inserts in the V4 region (in S. ranae, S. elegans and S. truttae), as a part of highly variable segments of SSU rDNA, were removed from the alignment together with the other ambiguous sites and had no influence on the phylogenetic relationship. In our analysis, 3 main clades, the marine, the freshwater, and a ‘Sphaerosporid clade’ were recognized. The group of S. elegans, S. truttae, S. ranae, and L. fugu representing the ‘Sphaerosporid clade’, formed a sister clade to all other myxosporeans supported with high bootstrap value. Within the ‘Sphaerosporid clade’, S. ranae, S. elegans and S. truttae formed a subclade with maximum bootstrap value, while L. fugu branched as its basal taxon. All available sequences of the SSU rDNA from Sphaerospora spp. were included in the analysis, all of them being distributed in different, non-related lineages.
Fig. 5. The maximum likelihood tree (−ln=16275.3044, α shape parameter=0·8032, proportion of invariable sites=0·1512) of the SSU rDNA sequences of myxosporeans. Buddenbrockia plumatellae and Tetracapsuloides bryosalmonae were set as outgroup. Bootstrap values (ML and MP Ts/Tv=1:2) are indicated for the nodes gaining more than 50% support. GenBank Accession numbers for each taxon are listed. The scale is given under the tree.
DISCUSSION
Since 1929, no myxosporean infection has been recorded in European amphibians. Nevertheless, our research in the Czech Republic revealed a high prevalence of myxosporeans in R. dalmatina, as well as their less frequent incidence in the related species, R. temporaria. Morphological features of this myxosporean closely resemble those in the original description of S. ranae (Morelle, Reference Morelle1929). We found the observed spores to be 10·2 (9·5–11·0) long and 11·8 (10·0–12·5) wide, containing polar capsules 4·0–5·0 in diameter with 6–7 coils of polar filament. Spores reported by Morelle (Reference Morelle1929) were 12·5 long and 16·5 wide, polar capsules 5·5 in diameter, containing 7–8 coils of polar filament. The smaller spore dimensions presented in our study are because of the exclusion of the postero-lateral bulges (typical for S. ranae spores) from the measurements. This exclusion is due to the high morphological and temporal variability of the bulges. Nevertheless, if the breadth of the bulges (1–2·5) is taken into account, our measurements match those from Morelle's data. We consider our isolates to be conspecific with Morelle's S. ranae, as well as with myxosporean observed by Thélohan in R. temporaria and R. kl. esculenta in 1895. Unfortunately, in Morelle's original description, neither type locality nor type material was specified. As a result, topotypic material and/or type material could not be examined. Although the sporogonic stages are presented as line-drawings by Morelle (Reference Morelle1929), no description is provided within the text. This confusing situation can be only addressed by a redescription congruent with current standards, including deposition of (neo)type material and symbiotype sensu Frey et al. (Reference Frey, Yates, Duszynski, Gannon and Gardner1992) in an accredited collection.
We report R. dalmatina to be a new host of S. ranae. According to the current standard, we deposited symbiotype and neotype material of S. ranae from R. dalmatina in the type collection of the Department of Parasitology, University of Veterinary and Pharmaceutical Sciences Brno, Czech Republic. During the 5 years of our study, no myxosporean infection was recorded in syntopically occurring amphibians (B. bufo, H. arborea, T. vulgaris, and T. alpestris). These data suggest that the host specificity of S. ranae is restricted to adults of the European Rana spp. We failed to exclude or prove the infectivity of S. ranae for R. kl. esculenta, as reported by Thélohan (Reference Thélohan1895), because of the absence of this frog species at the studied localities. A relatively narrow host spectrum is also reported in S. elegans and S. truttae, parasitizing only related fish species of the family Gasterosteidae and Salmonidae, respectively.
As in the great majority of myxosporeans, the life-cycle of S. ranae is unknown. However, its life-cycle might involve an actinosporean development. In contrast to fish myxosporeans, contact of anuran hosts with presumed actinosporean stages of S. ranae is rather limited, being restricted to 3 aquatic phenological phases of the host: (i) spring spawning, (ii) larval development, and (iii) hibernation in water (common in R. dalmatina). Infections during spring spawning and larval development are not likely, since the frogs are already parasitized by advanced (sporogonic) stages immediately after the hibernation. In addition, we did not find any signs of infection in 400 tadpoles of both host species. It can be concluded that affinity to aquatic hibernation, or a pre-hibernation stage in water (which is influenced by environmental factors and varies among localities in particular Rana sp.), might be the most important factor influencing the infection process. Possibly, the affinity to water in autumn may result in infections of adult frogs with actinospores, followed by presporogonic proliferative development and early sporogonic development before the hibernation (assuming, that no development occurs during the host's hibernation). We invariably detected mature spores in animals collected immediately after hibernation, when the myxospores are released from frogs into the water via urine and ejaculate during spring spawning (see Results section), potentially infecting a second host. These assumptions are in agreement with Morelle (Reference Morelle1929), who collected R. temporaria during October and November, and found frogs to be infected 2–3 months later in captivity.
It seems that ecology of a potential host (Rana spp.) at a particular locality must coincide temporarily with the S. ranae life-cycle, resulting in its presence at the locality, and the annual timing of S. ranae infections seems to be a result of parasite adaptation to restricted host availability. This may explain rather confusing discrepancies in the occurrence of S. ranae in different host species and/or localities. Such a pattern is better explained by host ecology at the particular locality, rather than by host specifity of a parasite, as suggested by our observations from Locality 1. Here, R. dalmatina usually overwinter in water in the breeding pond, whereas sympatric R. temporaria hibernate terrestrially or in an adjacent forest stream. However convoluted these explanations are, it is evident that the description of the full life-cycle is necessary to understand the epidemiology of infections.
The partial sequence of S. ranae was longer than expected when using particular myxosporean specific primers. This result is due to the presence of 2 areas with extensive nucleotide insertions in the V4 region in the sequence of SSU rDNA. Similar areas with extensive insertions are also found in S. elegans and S. truttae. Only the sequence of S. truttae is almost complete, having a length of 2541 bp. We deduced that the complete sequence of S. ranae would probably have a SSU rDNA of similar length as that of S. truttae. The conservation of the sequences was confirmed by phylogenetic analyses. Sphaerospora ranae with S. elegans and S. truttae formed a well-supported clade and they branched out together with L. fugu, forming a ‘Sphaerosporid clade’, a basal branch to all other myxosporean species. In another analysis Holzer et al. (Reference Holzer, Sommerville and Wootten2004) placed the 2 species of Sphaerospora into close relation with Ceratomyxa sparusaurati, with relatively high bootstrap support. Our phylogeny was more congruent with the results of other analyses (Fiala, Reference Fiala2006; Jirku et al. Reference Jirků, Bolek, Whipps, Janovy, Kent and Modrý2006) and showed a grouping separating S. ranae, S. elegans and S. truttae from L. fugu.
Several myxozoan genera (e.g. Myxidium, Myxobolus, Sphaerospora), characterized by spore morphology, were regarded as paraphyletic or polyphyletic in SSU rDNA-based phylogenetic analyses (Andree et al. Reference Andree, Szekely, Molnar, Gresoviac and Hedrick1999; Kent et al. Reference Kent, Andree, Bartholomew, El-Matbouli, Desser, Devlin, Feist, Hedrick, Hoffmann, Khattra, Hallett, Lester, Longshaw, Palenzuela, Siddall and Xiao2001; Fiala, Reference Fiala2006). The spore morphology alone could not be considered to be sufficient enough for generic classification of myxosporeans, thus the taxonomic status of most myxosporeans remains unclear. According to our analysis, the same applies to the genus Leptotheca, to which S. ranae was originally assigned by Morelle (Reference Morelle1929). The type species of the genus, Leptotheca agilis Thélohan, Reference Thélohan1895, inhabits the gallbladder of the Mediterranean stingray Dasyatis pastinaca (in original description referred to as Trygon pastinaca). As L. agilis possesses features differentiating it clearly from S. ranae, as well as from S. elegans and S. truttae (spore morphology, polysporous plasmodia, localization in gallbladder, marine habitat), and has not yet been included in any phylogenetic analysis, the placement of S. ranae into the genus Leptotheca is unsubstantiated. On the other hand, S. ranae was placed in close relationship with fish Sphaerospora, namely S. elegans (type species of the genus) and S. truttae in our analysis. The close relationship is further supported by the presence of 2 areas with extensive nucleotide insertions in the V4 region in the sequence of SSU rDNA (excluded from our phylogenetic analysis), similar spore morphology, general features of the sporogony, within host localization (coelozoic in excretory system) and freshwater habitat. Additionally, L. fugu was the most basal taxon within the ‘Sphaerosporid clade’. Obviously, L. fugu represents a distantly related lineage as suggested by the topology of the phylogenetic tree, absence of areas with extensive nucleotide insertions in V4 region and basic biological characteristics (marine, histozoic, intestinal) (Tun et al. Reference Tun, Yokoyama, Ogawa and Wakabayashi2000). According to a recent review of myxozoan genera (Lom and Dyková, Reference Lom and Dyková2006), S. ranae possess spore proportions (in fact the only character distinguishing genera Sphaerospora and Leptotheca) characteristic for the genus Sphaerospora, having spores with axial diameter (length) being approximately equal to valvular diameter (width). In Leptotheca, the breadth of the individual spore valve (as measured from midpoint of the suture to the most distant point of the valve) significantly exceeds one half of the axial diameter, and the valvular diameter often significantly exceeds axial diameter. Generally, the character of the sporogonic development observed in S. ranae is congruent with data known from related S. elegans and S. truttae (Feist et al. Reference Feist, Chilmonczyk and Pike1991; McGeorge et al. Reference McGeorge, Sommerville and Wootten1994). The only significant, but striking difference is the presence of 1–4 spores within a plasmodium. This feature distinguishes S. ranae from S. elegans and S. truttae, both possessing disporous pseudoplasmodia. For the reasons mentioned above, we assign S. ranae to be a member of the genus Sphaerospora. It is worth noting, that the genera Sphaerospora and Leptotheca are traditionally assigned to different families, the Sphaerosporidae and Ceratomyxidae, respectively. The shift of S. ranae from one family to another shows the artificial nature of the established myxosporean taxonomy. In conclusion, despite the enormous effort of taxonomists, when all molecular, morphological and life-history data are considered, the status of most myxosporean higher taxa becomes questionable (for review and comparision of phylogenetic relatinships and traditional taxonomy see Fiala, Reference Fiala2006; Lom and Dyková, Reference Lom and Dyková2006).
It is evident that arising taxonomic problems associated with polyphyletic and/or paraphyletic status of most myxosporean genera, as well as higher taxa, can only hardly be solved with the limited data available. Future sequencing of the type species of individual genera and identification of truly synapomorphic characteristics is necessary to join the principles and rules of classical taxonomy with a phylogenetic approach. In this respect, the case of S. elegans (type species of the genus Sphaerospora) is rather exception, and the only other type species sequenced so far is Myxidium lieberkuehni.
Although the spore morphology of all 7 sequenced Sphaerospora (sensu lato) fits well to the current definition of the genus (see Lom and Dyková, Reference Lom and Dyková2006), they belong to unrelated lineages in phylogenetic analyses (see Fig. 5). Sphaerospora elegans, S. ranae and S. truttae form a monophyletic clade, whereas the 4 other sequenced species are scattered in the phylogenetic trees. Interestingly, despite striking life-history differences, the 2 freshwater species from common carp, coelozoic S. renicola (renal tubules) (Dyková and Lom, Reference Dyková and Lom1982) and histozoic S. molnari (gill lamellas) (Lom et al. Reference Lom, Dyková, Pavlásková and Grupcheva1983) are closely related to each other and form a subclade within a clade of histozoic myxosporeans of the genera Myxobolus, Henneguya and Hoferellus. The coelozoic, freshwater S. oncorhynchi (Kent et al. Reference Kent, Whitaker and Margolis1993) appears to be closely related to M. lieberkuehni, hence the type species of the genus Myxidium. In addition, histozoic, marine S. dicentrarchi (Sitja-Bobadilla and Alvarez-Pellitero, Reference Sitja-Bobadilla and Alvarez-Pellitero1992) is obviously a member of Multivalvulida, and its taxonomic status was recently impugned by Diamant et al. (Reference Diamant, Ucko, Paperna, Colorni and Lipshitz2005), who recognized its phylogenetic and life-history differences from other sphaerosporans. Obviously, morphological data alone are of very limited value for the definition of the genus Sphaerospora, providing even less taxonomic and phylogenetic information than in other myxosporean taxa. The spore morphology of Sphaerospora s.l. is very simple, and probably represents a ‘primitive’, basal morphotype retained in most myxosporean lineages.
The availability of sequence of the type species of the genus Sphaerospora (=S. elegans) allows us to identify the clade of ‘true’ Sphaerospora (Sphaerospora sensu stricto). When concluded, the only feature unique for Sphaerospora s.str. was the presence of 2 areas with extensive nucleotide insertions in the V4 region in the SSU rDNA. However, our analysis allows us to point out characteristic combination of features (rather than characteristic features) and enables the emendation of the genus definition. Importantly, this is the first case in the taxonomy of the phylum Myxozoa, when a molecular characteristic is used for definition of a particular taxon.
All named but unsequenced Spaherospora species should be referred to as Sphaerospora incertae saedis to avoid introduction of unnecessary taxonomic discrepancies before more species of Sphaerospora s.l. will be analysed. The same applies to S. dicentrarchi, S. molnari, S. oncorhynchi and S. renicola which await generic reclassification.
Emendation of the genus Sphaerospora Thélohan, 1892
We propose the following (possibly interim) emendation of the genus characterization based on available information: spherical to subspherical spores composed of 2 identical valves. Valvular diameter of spores more or less equal to sutural diameter. Suture straight, longitudinal, joined by prominent sutural ridge. Valves smooth, or with surface ornamentation, sometimes with folded surface and postero-lateral bulges. Two equal polar capsules are spherical to subspherical, opening at the anterior tip of the spore and are situated in a plane perpendicular to the sutural line. Trophozoites (plasmodia) disporic or less often polysporic. Sporogonic stages coelozoic in lumina of renal tubules of freshwater fish and amphibians. Characterized by the presence of 2 areas with extensive nucleotide insertions (expansion segments) in the V4 region of the SSU rDNA sequence. The length of the characteristic V4 region expansion segments of the sequenced Sphaerospora s.str. ranges between 585 and 766 bp (S. elegans 766 bp, S. ranae 585 bp, S. truttae 618 bp, compared to 166 bp in Ceratomyxa shasta).
TAXONOMIC SUMMARY
Sphaerospora ranae (Morelle, Reference Morelle1929) n. comb.
Synonyms: Leptotheca ranae Thélohan, Reference Thélohan1895nomen nudum; Wardia ohlmacheri Kudo, Reference Kudo1920 (in part); Leptotheca ranae Morelle, Reference Morelle1929.
Proposed type host: Rana dalmatina Bonaparte, 1840 (this study).
Other hosts: Rana kl. esculenta (Thélohan, Reference Thélohan1895), Rana temporaria (Thélohan, Reference Thélohan1895; Morelle, Reference Morelle1929; this study).
Proposed type locality: Zaječí potok, Brno, Czech Republic, (16°36′23·07″E, 49°14′15·57″N), 303 m asl. Note: Only R. dalmatina is being infected with S. ranae at this site.
Other localities (for host records at the localities see below): Růženčin lom, Brno, Czech Republic, 16°40′22·69″E, 49°13′0·78″N, 355 m asl.; Jedovnice, productive pond in the vicinity of the village, Czech Republic, 16°46′30·51″E, 49°20′0·71″N, 484 m asl.
Other records: Belgium (Morelle, Reference Morelle1929), France (Thélohan, Reference Thélohan1895). Note: No exact localities are provided by authors.
Site of infection: renal tubules, sporogonic stages coelozoic.
Prevalence: 25/32 (78·1%) of R. dalmatina from the locality Zaječí potok, 3/6 (50%) of R. dalmatina from the locality Růženčin lom, and 3/18 (16·6%) of R. temporaria from the locality Jedovnice.
Neotype material deposited: Paraffin-embedded infected kidney; histological sections of infected kidney; infected kidney preserved in absolute ethanol and digital photomicrographs deposited at the Department of Parasitology, University of Veterinary and Pharmaceutical Sciences Brno, Czech Republic under the collection number R 22/06. The GenBank Accession number for the SSU rDNA sequence of Sphaerospora ranae is EF211975.
Type host: Rana dalmatina symbiotype (sensu Frey et al. Reference Frey, Yates, Duszynski, Gannon and Gardner1992) and liver sample stored in 95% ethanol is deposited in the collection of Department of Parasitology, Faculty of Veterinary Medicine, University of Veterinary and Pharmaceutical Sciences, Brno, Czech Republic (Cat. No. R 22/06).
Remarks: Sphaerospora ranae is clearly distinguishable from other renal myxosporeans parasitizing amphibians. According to spore morphology, both C. protei and C. careni differ (among others) in having 4 polar capsules, whereas H. anurae differs in significantly smaller, striated and mitra-shaped spores with several pointed posterior projections. Lastly S. ohlmacheri differs from S. ranae by the presence of distinct spore surface striations, and absence of postero-lateral bulges. Additionally, in contrast to S. ranae, parasitizing only adult frogs, S. ohlmacheri is known to parasitize both adult anurans, and larval stages of Rana catesbeiana in N America (Desser et al. Reference Desser, Lom and Dyková1986).
This study was supported by grants of the Grant Agency of the Czech Republic nos. 206/03/1544 and 524/03/H133 and by Research Centre “Ichtyoparasitology” (LC522). Permits (nos. MŽP 30446/03-620/6060/03, OOP/5871/01-V853, OOP/6418/99-V556) for work with amphibians were issued by the Ministry of Environment of the Czech Republic. We wish to thank Iva Dyková (IP CAS) for kind help with obtaining the literature, Věra Kučerová, Blanka Cikánová (LEM CAS), Veronika Schacherlová and Marie Flašková (IP CAS) for generous help with processing the samples. Comments and suggestions by anonymous reviewers significantly improved the original manuscript.
