INTRODUCTION
Protozoans of the genus Cryptosporidium are intracellular pathogens infecting a broad range of mammalian species including humans. Most Cryptosporidium species complete their development in epithelial cells of the alimentary tract. In amphibians, reptiles, birds and mammals at least 20 valid Cryptosporidium spp. and over 40 Cryptosporidium isolates referred to as genotypes have been summarized currently, and the description of new isolates is still a continuous process (Xiao and Fayer, Reference Xiao and Fayer2008; Fayer, Reference Fayer2009; Fayer and Santin, Reference Fayer and Santin2009).
Wildlife has been considered to be a reservoir of zoonotic diseases and a source of environmental contamination (Appelbee et al. Reference Appelbee, Thompson and Olson2005). However, the application of molecular tools showed that the majority of Cryptosporidium species found in naturally infected wild mammals are host-adapted genotypes and are different from those usually infecting human hosts (Xiao et al. Reference Xiao, Fayer, Ryan and Upton2004; Appelbee et al. Reference Appelbee, Thompson and Olson2005). A few human-pathogenic Cryptosporidium spp. have been isolated from wildlife mammals, but they seem not to represent a major public health concern (Appelbee et al. Reference Appelbee, Thompson and Olson2005). Relatively few molecular studies have been conducted on wildlife mammals and isolates from certain species are therefore still uncharacterized. Matos (Reference Matos, Fayer and Xiao2008) pointed out that less-prevalent species and genotypes, not yet found in humans, could emerge as human pathogens.
The European hedgehog (Erinaceus europaeus L.) is a protected wildlife species, which is probably most frequently presented to veterinarians in Germany (Beck, Reference Beck2007) and the most frequently rehabilitated mammalian wildlife species in the UK (Sainsbury et al. Reference Sainsbury, Cunningham, Morris, Kirkwood and Macgregor1996). Juvenile hedgehogs, born late in the year, are often rescued by humans and brought to rehabilitation centres in autumn or early winter for various reasons, e.g. injuries and/or emaciation. These animals have often insufficient fat resources to survive hibernation (Jensen, Reference Jensen2004). After overwintering they are released back into the wild in the following spring.
Only few case reports on the occurrence of Cryptosporidium infections in hedgehogs are available, presenting Cryptosporidium as a cause of chronic disease, accompanied by anorexia, apathy, progressive weight loss, pale mucous membranes as well as green-slimy and malodorous faeces (Pantchev and Moller, Reference Pantchev and Moller2007) or probably even death of one African hedgehog (Graczyk et al. Reference Graczyk, Cranfield, Dunning and Strandberg1998). In most cross-sectional surveys on large sample numbers from hedgehogs, the detection of Cryptosporidium spp. was not performed. Moreover, reports of molecular characterization of Cryposporidium isolates from these animals are scarce. Genotyping of hedgehog-derived cryptosporidia is documented for only one isolate, identified as ‘bovine genotype’ of Cryptosporidium parvum by sequencing of partial 18S rRNA (Enemark et al. Reference Enemark, Ahrens, Juel, Petersen, Petersen, Andersen, Lind and Thamsborg2002). The present study aimed to obtain the data on the occurrence of Cryptosporidium infections in juvenile hedgehogs overwintering in hedgehog rehabilitation centres and to analyse the genetic diversity among the Cryptosporidium isolates from European hedgehog in Germany with regard to zoonotic or antropozoonotic potential of Cryptosporidium genotypes.
MATERIALS AND METHODS
Sample collection and examination order for Cryptosporidium
In total 188 faecal samples from hedgehogs were provided by 16 geographically distinct rehabilitation centres (Fig. 1). Common reasons for submission were diarrhoea, anorexia and weakness. The sampled animals were assigned into 2 independent groups: the first group (n=74) representing the hedgehogs finishing rehabilitation was sampled between March and April 2007. The second group (n=114) comprised captive animals at the beginning of rehabilitation and were sampled between September 2007 and January 2008. Data on age, feed consumption and faecal consistency of sampled animals were documented and reported by employees of the hedgehog rehabilitation centres. A private diagnostic laboratory provided 2 additional samples, which were tested positive by enzyme immunoassay and were used directly for genotyping.
Spatial distribution of the hedgehog rehabilitation centres collecting the hedgehog samples and isolated subtypes of Cryptosporidium spp. from hedgehogs in Germany.
For Cryptosporidium detection the samples were subjected to a sequence of diagnostic tests. At first immunoassays (IA, an enzyme immunoassay and an immunochromatographic assay) were performed for detection of specific coproantigen. The samples that tested positive for Cryptosporidium coproantigen were verified by microscopic examinations (ME) for the presence of oocysts. In the first group all IA-negative results were verified by ME, in the second group no ME on IA-negative samples was performed.
Examination by immunoassays (IA) and microscopic examination (ME) of samples
For IA a commercial coproantigen ELISA (ProSpectTM, Cryptosporidium Microplate ELISA Assay, Remel, GB) and the immunochromatographic assay (FASTest® CRYPTO Strip, MegaCor Diagnostik GmbH, Austria) were applied. The tests were performed following the manufacturer's directions for use. Briefly, 0·1 g (normal consistence) or 0·3 g (liquid sample) faeces were used for ELISA and the test result was assessed visually. For the FASTest® CRYPTO Strip 0·4 g faeces were homogenized in the supplied buffer solution. The strip was dipped into suspension for 1 min and was left for 5 min on a flat surface. The test was positive if red colouring was visible in the test zone.
The ME was performed by means of modified Ziehl-Neelsen-staining (MZN) and direct immunofluorescence antibody assay (IFA, MERIFLUOR® Cryptosporidium, Meridian Bioscience, Inc. USA). For the modified Ziehl-Neelsen staining, faecal smears were prepared on a microscope slide (24×60 mm) and were subsequently stained as described elsewhere (Henriksen and Pohlenz, Reference Henriksen and Pohlenz1981). For the direct immunofluorescence antibody assay 1 drop of faeces was passed onto the cavity of a microscopic slide and subjected to antibody staining following the manufacturer's directions for use. Stained samples were examined using at least a 400-fold microscopic magnification by light and fluorescent microscopy for MZN and IFA, respectively.
Examination of faecal samples by means of flotation
If enough faecal material was available after IA and ME a flotation technique was conducted, to detect helminth eggs and coccidian oocysts. Faeces (0·5–2 g) were subjected to zinc sulphate flotation (specific gravity=1·3, 704 g/l ZnSO4) by means of centrifugation at 300 g for 5 min. Floated material was transferred to microscopic slides and examined by light microscopy using at least 400-fold microscopic magnification.
DNA extraction
Total DNA was extracted from 20 samples, each of them containing 200–300 mg faeces, using a QIAamp DNA Stool Kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions, with some modifications. The faeces were mixed with buffer ASL and subjected to ultrasonic treatment for 8 min (power 40%, amplitude 0·5) using a Sonopuls GM 70 sonificator (BANDELIN Electronic, Berlin, Germany) to disrupt the oocysts. After sonification, DNA extraction followed the protocol; DNA was eluted in 50 μl of AE buffer.
Polymerase chain reaction (PCR) and restriction fragment length polymorphism (RFLP)
All amplification reactions were prepared in total volumes of 25 μl, consisting of 1× Colorless GoTaq Flexi Buffer (Promega), 0·3 mm of each primer (Table 1), 0·2 mm of each dNTP and 1 Unit GoTaq Flexi DNA Polymerase (Promega), the MgCl2 concentration varied depending on the target being amplified (Table 1). The PCR cycling conditions were 95°C for 2 min, followed by 35 cycles at 94°C for 40 s, at the specific annealing temperature of the respective primer pair (Table 1) for 40 s, at 72°C for 1 min and, finally, 1 cycle at 72°C for 7 min using a thermal cycler (iCycler, BioRad). All DNA samples were separated on 1·5% agarose gel by electrophoresis and photographed under UV light after staining with ethidium bromide.
Primer pairs and conditions used in this study
a Base pair.
b Annealing temperature.
To determine the Cryptosporidium spp. the sequences of 18S ribosomal RNA gene were amplified in a first direct and a subsequent second nested PCR using described primers. The PCR-products of the nested PCR were purified by sodium acetate precipitation (Sambrook and Russel, Reference Sambrook and Russel2001) and subjected to RFLP using the restriction enzymes Ssp I and Vsp I as previously described (Xiao et al. Reference Xiao, Escalante, Yang, Sulaiman, Escalante, Montali, Fayer and Lal1999a). Additionally, twelve 850 bp amplicons from the nested PCR were cloned and sequenced.
The amplicons of the 60 kDa glycoprotein locus (GP60) were generated in direct and nested PCR using previously published primers (Table 1) and conditions (Peng et al. Reference Peng, Matos, Gatei, Das, Stantic-Pavlinic, Bern, Sulaiman, Glaberman, Lal and Xiao2001; Alves et al. Reference Alves, Xiao, Sulaiman, Lal, Matos and Antunes2003).
The partial sequences of the 70 kDa heat shock protein gene (HSP 70) and actin gene were amplified using primer sets (Table 1) developed by Sulaiman et al. (Reference Sulaiman, Morgan, Thompson, Lal and Xiao2000) and Sulaiman et al. (Reference Sulaiman, Lal and Xiao2002) for HSP70 and actin, respectively (Table 1). Twelve amplicons of each partial HSP70 and actin sequences were cloned and sequenced.
Cloning and sequencing procedures
For the phylogenetical studies the PCR products of the 18S ribosomal RNA gene (850 bp), actin gene (1070 bp) and HSP70 gene (1924 bp) were ligated into a pBluescript II SK-Phagemid vector (Stratagene) and used for transformation of chemically competent Escherichia coli XL1-Blue (Stratagene) as previously described (Dyachenko et al. Reference Dyachenko, Beck, Pantchev and Bauer2008). Plasmids from recombinant clones were isolated and sequenced in both directions at the Interdisciplinary Centre for Clinical Research (IZKF) Leipzig using vector-specific (T3- and T7-promoter) primers. Since the sequencing of HSP70 fragments by vector-specific primers does not span the cloned inserts, 2 additional primers in opposite directions were used (Table 1). The products of the nested GP60-PCR were sequenced directly using the primers AL3532 and AL3533 in both directions. The GP60-PCR products of the samples IV-19 and I 45-21 were additionally cloned and sequenced as described above. Cryptosporidium subtypes of GP60 were named using a recently proposed nomenclature (Sulaiman et al. Reference Sulaiman, Hira, Zhou, Al Ali, Al Shelahi, Shweiki, Iqbal, Khalid and Xiao2005), according to the number of trinucleotide repeats TCA and TCG and the ACATCA sequence.
The sets of sequences of samples IV-19, I 45-21, II 60-77, I 64-18, II 77-82 and Garb-51 were deposited in GenBank® (Accession numbers GQ214078-GQ214085 and GQ259136-GQ259151).
Phylogenetic analysis
Phylogenic analysis of partial sequences of GP60, 18S ribosomal RNA, actin gene and HSP70 were conducted using the MEGA 4 software package. The sequences were aligned by ClustalW with the following values: gap opening penalty 15; gap extension penalty 6. The phylogenetic trees were inferred by the Neighbor-Joining method using Kimura-2-parameter model with 10 000 bootstrap replicates.
Statistical analysis
WinEpiscope 2 was used to analyse differences in prevalence as well as to calculate the 95% confidence interval (95% CI). The data from the second group were analysed to determine the possible associations between Cryptosporidium infection and infections with other parasites.
RESULTS
Detection of Cryptosporidium and concurrent infections in faecal samples
In the first group the appraised age of the animals ranged between 6 months and 2·5 years (median 7 months). Cryptosporidium coproantigen was detected in 11 (14·8%) of 74 submitted samples; 1 of the Cryptosporidium coproantigen-containing samples could not be examined further by microscopy as insufficient amounts of faeces were available. Cryptosporidium oocysts were detected in 14 (19·1%) of 73 faecal samples with 3 coproantigen-negative samples that tested positive for oocysts (Table 2). Based on the results of both, IA and ME, 15 (20·2%) sampled animals were considered as positive for Cryptosporidium. Microscopic examination after flotation was performed on 73 of these samples revealing coccidian oocysts (Isospora spp. if sporulated) in 28 (38·3%) cases and eggs of Capillaria spp. in 33 (45·2%) cases (Table 3). Cryptosporidium-positive samples were also positive for shedding of oocysts and Capillaria eggs in 6 and 7 cases, respectively (Table 3).
Results of Cryptosporidium-positive samples showing agreement or disagreement of tests from the first group
(+), positive; (−), negative; i.f., insufficient faeces.
Detection of parasitic pathogens in faecal samples of hedgehogs
a Detection of co-infections was performed on Cryptosporidium-positive samples if sufficient faeces were available.
In the second group the appraised age of the sampled animals ranged between 2 weeks and 24 months (median 2·5 months). In contrast to the first group the oocyst detection was performed only on IA-positive samples. Cryptosporidium-antigen was detected in 45 (39·4%) of 114 submitted samples. Of 45 IA-positive samples 6 samples contained insufficient faeces and the presence of oocysts could not be verified by staining techniques, 4 samples were found to be negative by ME. After Cryptosporidium detection by IA and ME, only 95 faecal samples were available for examination by the flotation technique. Oocysts of Isospora spp. were detected in 7 (7·3%) cases, eggs of Capillaria spp. in 49 (51·5%) cases and finally Brachylaemus erinacei eggs in 8 (8·4%) cases (Table 3).
The shedding of Cryptosporidium coproantigen was accompanied by shedding of Isospora spp. oocysts and helminth eggs in all 34 respective cases. Eleven faecal samples positive for Cryptosporidium coproantigen could not be examined due to lack of enough faecal material. In 34 IA-positive samples examined by flotation, 13 were also positive for Capillaria spp. eggs and 3 for B. erinacei eggs. In each of the 2 cases Cryptosporidium coproantigen detection was accompanied by the presence either of Isospora spp. oocysts and Capillaria spp. eggs or B. erinacei eggs and Capillaria spp. eggs. In 1 case co-infections with Isospora spp., Capillaria spp. and B. erinacei were determined, presenting a 4-fold parasitic infection (Table 3).
Epidemiological data analysis
Since ME was not performed on IA-negative samples in the second group, only IA-positive samples were considered for the data analysis. The comparison between groups showed that juvenile animals brought to the hedgehog station had a 3·7 higher chance (95% CI: 1·7–7·8) of being infected with cryptosporidia (Table 3). On the other hand, Isospora/Eimeria infections were less frequent (crude OR 0·12, 95% CI:0·05–0·31) in younger animals recently brought to rehabilitation centres than in animals that had overwintered in the hedgehog care centres. There was no significant difference in occurrence of Capillaria infections between the two groups. No coherence between Cryptosporidium infections and other parasitic co-infections was observed from the data of the second group.
Subtyping of Cryptosporidium based on 18S ribosomal RNA- and GP60-locus
The PCR-RFLP of partial 18S ribosomal RNA performed on 13 samples identified C. parvum in all samples and as a result, the samples were subtyped at the GP60-locus. The amplification of the partial 60 kDa glycoprotein gene revealed bands that varied in size from 430 bp to 590 bp. In one case (sample II 72-81) a pattern containing 2 bands of 430 and 580 bp was observed.
The sequencing of the GP60-amplicons showed the presence of 3 different subtypes of C. parvum (Table 4). Five samples were identical to the IIc subtype family with subtype IIc A5G3. These sequences showed the highest homology to the IIc A5G3a subtype (AF164491); however, they were not fully identical to them. There were 2 transitions from A to G at positions 307 and 481 (starting from the start codon), which lead to exchange from lysine to glutamic acid in both cases. The GP60 sequence of 1 sample shared homology to the IIa subtype family and was indentified as the IIa A19G1R1 subtype. The other 6 samples presented sequences with low identity to known GP60 sequences (Fig. 2). Sequences of these novel subtypes contain the unusual long TCA and ACATCA (R element) coding 19-23 serines and 10-12 serine-threonine repeats. This 120-126 bp long serine/serine-threonine stretches of coding sequence was followed by a short conserved region coding invariant alanine, proline and lysine residues, which were subsequently followed by highly polymorphic region coding 74 amino acids. The polymorphic amino acid sequence ended with an invariant cysteine, which flanked the second conserved region at position 203 of the reference sequence NINC 1 (AF022929). The novel GP60 sequences are proposed as a new VIIa subtype family, representing probably a novel hedgehog-specific genotype. One of these samples (II 72-81), containing the VIIa subtype, showed 1 band at 430 bp, which was identified as the IIc A5G3 subtype by sequencing (2 bands of different size as noted above).
Genetic variation of various Cryptosporidium isolates of hedgehogs, C. wrairi, C. bovis and C. canis compared to C. pavum bovine genotype (Iowa II)
Neighbor-Joining tree on partial GP60 sequences from Cryptosporidium-positive hedgehogs (in boldface). Bootstrap values greater than 50% are shown.
Multi-locus sequence typing on partial 18S rRNA, actin and HSP70
Twelve samples (sample numbers: I 64-18, IV-19, I 45-21, Gabr-51, II 65-75, II 67-76, II 60-77, II 61-78, II 63-79, II 49-80, II 77-82, Keck-108) of 13 samples subtyped on GP60 locus were selected for multi-locus sequence typing: 5 samples of IIc; 1 sample of IIa as well as 6 samples of the VIIa subtype families (Table 4). The sequencing and alignment of the 834 bp long 18S ribosomal DNA fragment showed minimal differences in sequences among each other and showed the highest identity to C. parvum. However, all 18S rDNA sequences from the samples belonging to the novel subtype family revealed 2 transitions (A to G and G to A) at positions 218 and 430, respectively, and one transversion (A to T) at position 473, and were thus not fully identical to the 18S ribosomal RNA gene of C. parvum IOWAII (Table 5).
Sequence differences of 18S ribosomal RNA gene among isolates of Cryptosporidium parvum, C. hominis, C. wrairi and the Cryptosporidium from hedgehog isolates
Partial sequencing of the actin gene locus showed a high degree of conservation between isolates belonging to C. parvum bovine IIa subtype and the VIIa subtype (Table 4). However, the transition between T and C on positions 514 and 799 lead to exchange of similar amino acids: from leucine to isoleucine and from phenylalanine to leucine for corresponding positions at 172 and 267 in amino acid sequence, respectively. The actin coding sequences from the other subtypes (IIc and IIa) were identical with C. parvum as expected.
Cloning and sequencing of HSP70 from hedgehog samples showed sequences consisting of 1912 bp and 1924 bp for Cryptosporidium parvum (subtype families IIa, IIc) and Cryptosporidium hedgehog genotype (GP60 subtype VIIa), respectively. The A+T content was typical for C. parvum genotypes and was, on average, 58·43% and 58·47% for HSP70 of samples positive for IIc and the VIIa subtypes, respectively. Multiple alignments of the HSP70 gene sequences belonging to IIc, IIa and VIIa subtype families revealed that the HSP70 sequences of the VIIa positive samples are more divergent then the HSP70 of IIa and IIc positive samples. The major difference between HSP70 of bovine C. parvum (Iowa II, cgd2_20) and HSP70 of Cryptosporidium hedgehog genotype (GP60 subtype family: VIIa) was 1 additional repeat of the sequence GGTGGTATGCCA coding amino acids GlyGlyMetPro and leading to the same numbers of repeats actually known from HSP70 of C. hominis. Based on distance matrices of the HSP70 sequences, VIIa subtype-positive samples showed 1·19 nucleotide changes per 100 base pairs compared to 0·27 nucleotide changes of the IIc subtype (Table 4).
In a phylogenetic analysis, a Neighbor-Joining tree was constructed from aligned HSP70 sequences isolated from hedgehog samples (subtypes IIa, IIc and VIIa) and using sequences of closely related C. parvum species, as C. parvum bovine genotype, C. parvum ferret genotype, C. hominis, C. wrairi, C. meleagridis and more divergent Cryptosporidium spp., as C. canis, C. felis, C. baileyi, as well as C. serpentis. The VIIa subtypes of C. parvum from hedgehog samples were segregated to a separate clade between the clusters formed by C. parvum, Iowa II and C. hominis. The branching was supported by high bootstrap values (Fig. 3). However, the Neighbor-Joining trees constructed on alignments of 850 bp fragment of 18S ribosomal RNA and 1070 bp fragment of actin gene could not resolve the VIIa subtype-positive hedgehog samples with significant bootstrap support (not shown).
Phylogenetic relationship of Cryptospordium spp. isolates from hedgehogs (in boldface) on the partial HSP70 sequences as inferred using the Neighbor-Joining method. The Cryptosporidium isolates belonging to the GP60 VIIa subtype family are separated to a different clade as those of C. parvum supported by high bootstrap values. Bootstrap values greater than 50% are shown.
DISCUSSION
The main aim of this investigation was the tentative assessment of occurrence of Cryptosporidium spp. infections and the molecular characterization of Cryptosporidium isolates among hedgehogs. The high rates of parasitic infections in hedgehogs are not surprising and were reported previously (Barutzki et al. Reference Barutzki, Schmid and Heine1984, Reference Barutzki, Laubmeier and Forstner1987; Lowenstein et al. Reference Lowenstein, Prosl and Loupal1991; Pantchev et al. Reference Pantchev, Globokar-Vrhovec and Beck2005). To our knowledge no comparison of the methods for the routine screening for cryptosporidiosis in faecal samples of hedgehogs has previously been reported. The comparison of diagnostic methods on bovine faecal samples showed a lower specificity of the ProspecTTM coproantigen ELISA as compared to the modified Ziehl-Neelsen stain (Brook et al. Reference Brook, Christley, French and Hart2008). The faecal examinations of the first group by several diagnostic tests showed some discrepancies between IA and ME presenting a slight underestimation of the prevalence by IA as 3 antigen-negative samples were positive for oocyst shedding. These discrepancies have been previously documented and are due to different sensitivity and specificity of the applied test, where the Bayesian approach would help to overcome the gold standard problem (Geurden et al. Reference Geurden, Berkvens, Geldhof, Vercruysse and Claerebout2006). Based on examination of data from the first group it would be obvious that the Cryptosporidium prevalence in the second group (juvenile hedgehogs) is underestimated. However, the 4 IA-positive samples, in which no oocysts could be detected, highlight the problem of diagnostic test accuracy as mentioned before (Kehl et al. Reference Kehl, Cicirello and Havens1995; Geurden et al. Reference Geurden, Berkvens, Geldhof, Vercruysse and Claerebout2006). On the other hand it should be kept in mind that the examined hedgehogs originated from a non-randomized population since they were pre-selected due to diarrhoea (37·8% for the first group and 44·4% for the second). Thus, the prevalence of Cryptosporidium infections was probably overestimated when compared to the total population of hedgehogs.
Nevertheless, juvenile hedgehogs rescued and settled to rehabilitation centres (second group) were more frequently infected with Cryptosporidium sp. (estimated prevalence 39·4%, 95% CI: 30·4–48·3%) and had a 3·7 (95% CI: 1·7–7·8) higher chance of being Cryptosporidium coproantigen-positive, as the older animals after several months stay in the rehabilitation centre. However, the difference in the prevalence may be due to age difference or a better health status in the centre or a combination of both.
The prevalence of helminths such as Capillaria spp. was similar in the two groups and even though some animals were treated with anthelmintics in the care centre. The high prevalence of helminth infections has also been reported from feral hedgehogs (Barutzki et al. Reference Barutzki, Schmid and Heine1984). On the other hand, the occurrence of Isospora and/or Eimeria infections was significantly higher in animals that were overwintering in the rehabilitation centres suggesting that these parasites are efficiently transmitted in this environment.
Some Cryptosporidium-positive animals showed a high degree of co-infections and at least 5 juvenile hedgehogs were infected by 3 parasitic species. However, the data analysis presented no coherence of more frequent occurrence of Cryptosporidium infection in animals with double and triple infections. Thus the helminth and coccidian infections do not appear to influence the occurrence of cryptosporidiosis.
The genotyping by nested PCR-RFLP based on the 18S ribosomal RNA gene showed Cryptosporidium parvum in all samples. Previous genotyping of a Cryptosporidium isolate obtained from 1 hedgehog based on partial 18S ribosomal RNA sequencing identified bovine genotype but sequencing of the isolate at a GAG microsatellite locus identified it as a unique subtype (Enemark et al. Reference Enemark, Ahrens, Juel, Petersen, Petersen, Andersen, Lind and Thamsborg2002). The GP60 subtyping was performed to our knowledge for the first time with isolates from hedgehogs and showed 3 different subtypes. Only 1 isolate (sample number II 67-76) was assigned to the zoonotic subtype family IIa (IIa A19G1RI), which was previously described in cattle in England (Brook et al. Reference Brook, Anthony, French and Christley2009) and in a human patient for the first time in Slovenia earlier (Soba and Logar, Reference Soba and Logar2008).
Detection of the IIc subtype family with the subtype IIcA5G3 was surprising as this subtype has never been isolated from non-human origin (Xiao and Fayer, Reference Xiao and Fayer2008) and was considered to be a human specific (anthroponotic) subtype of C. parvum. Although Cryptosporidium IIc subtype from hedgehog was not identical to other IIc subtypes, it had the highest similarity to IIc A5G3a subtype. The occurrence of some parasites in faeces due to ingestion of prey or highly contaminated material is known from some carnivorous or omnivorous mammalians and should be kept in mind. However, hedgehogs are insectivores feeding mostly on arthropods, snails, earthworms and only sometimes on mice (Fons, Reference Fons and Grzimek1988), which to our knowledge can not be a reservoir of Cryptosporidium IIc A5G3 subtype. The strong reaction in ELISA registered in all samples, as well as the high counts of oocysts (in 3 samples over 10 oocysts/per microscopic field), reinforce the idea of natural Cryptosporidium infection by C. parvum IIc subtype. Human infections with the C. parvum IIc subtype have been described in Europe, both parts of America, Australia and Kuwait, but seem to be rare compared to the occurrence of C. pavum IIc infections in developing countries (Wielinga et al. Reference Wielinga, de Vries, van der Goot, Mank, Mars, Kortbeek and van der Giessen2008; Xiao and Fayer, Reference Xiao and Fayer2008). Isolation of IIc A5G3 subtypes from geographically distant centres and from both groups of animals suggests transmission cycles among hedgehogs. However, it is not clear whether hedgehogs significantly contribute to the C. parvum IIc transmission cycle, or they are the only affected species as a result of environmental contamination with oocysts of human origin. Furthermore hedgehogs do not naturally occur in America and Australia, where C. parvum IIc A5G3 have been found.
The other GP60-subtypes could not be assigned to the already known subtype families (I, IIa-IIl, III, IV, V and VI) and were proposed therefore as a new VIIa subtype family, as the GP60 subtypes of C. hominis (subtype family I), C. parvum (subtype family II), C. meleargridis (subtype family III), C. fayeri (subtype family IV), Cryptosporidium rabbit genotype (subtype family V) and Cryptosporidium horse genotype (subtype family VI) have been published so far (Xiao, Reference Xiao2009; Power et al. Reference Power, Cheung-Kwok-Sang, Slade and Williamson2009; Chalmers et al. Reference Chalmers, Robinson, Elwin, Hadfield, Xiao, Ryan, Modha and Mallaghan2009).
Human infections with Cryptosporidium sp. VIIa subtype family have not been described so far and the zoonotical potential of this probably hedgehog-specific genotype is unknown.
The sequencing of partial 18S ribosomal RNA, actin genes of the Cryptosporidium sp. VIIa-positive isolates showed some differences in terms of single nucleotide polymorphisms (SNPs). However, these few SNPs could not resolve VIIa-positive isolates on these loci from C. parvum and closely related genotypes by Neighbor-Joining tree with significant bootstrap support, which is used frequently for phylogenetical analysis of Cryptosporidium spp. sequences (Xiao et al. Reference Xiao, Morgan, Limor, Escalante, Arrowood, Shulaw, Thompson, Fayer and Lal1999 b; Sulaiman et al. Reference Sulaiman, Lal and Xiao2002). On the other hand, using the HSP70 marker, the VIIa subtype family isolates were placed on a separate branch supported by high bootstrap values and could be considered as Cryptosporidium sp. hedgehog genotype, closely related to C. parvum and C. hominis. Based on sequence analysis it is impossible to predict the zoonotical potential of this Cryptosporidium hedgehog genotype. Due to the high degree of conservation of all 3 markers and considering the phylogenetically heterogeneous Cryptosporidium spp. from humans (Leoni et al. Reference Leoni, Amar, Nichols, Pedraza-Díaz and McLauchlin2006), a zoonotical potential of Cryptosporidium sp. VIIa subtype family from hedgehogs may be possible.
From the current point of view, wild animals do not represent a major public health concern (Appelbee et al. Reference Appelbee, Thompson and Olson2005). However, the relatively frequent occurrence of Cryptosporidium infections in juvenile hedgehogs suffering from diarrhoea could mean a possible transmission risk to humans, especially when people have frequent contact to infected animals, e.g. animal keepers at the centres and people, who collect and supply cachectic animals to hedgehog rehabilitation centres. Whether and to what extent the hedgehog could contribute to the transmission routes of zoonotical C. pavum subtypes is so far unknown. Considering the animal welfare, hedgehog rehabilitation centres show a positive effect in reducing of cryptosporidiosis. Improved hygiene in the hedgehog rehabilitation centres would further reduce the infection rates by coccidia.
CONCLUSION
Juvenile hedgehogs presented at rehabilitation centres are frequently infected by Cryptosporidium spp. However, hedgehogs overwintering in human care are significantly less infected by Cryptosporidium then those rescued in the autumn. Subtyping of Cryptosporidium showed the presence of at least 3 GP60 subtype families: IIa; IIc and VIIa. The latter represent the Cryptosporidium sp. hedgehog-specific genotype. Whether Cryptosporidium sp. VIIa subtype is of zoonotic impact is so far unknown. The C. parvum IIc subtype was isolated for the first time from a non-human host and its anthroponotic nature should be revised.
ACKNOWLEDGEMENTS
The authors thank Sandra Gawlowska for her excellent work at the molecular biological laboratory of the Institute of Parasitology. We are grateful to Frank Stöckel, Berit Bangoura, Holger John, Katja Dittmar for the laboratory help during the sample examination periods. We thank a lot Ulli Seewald, the president of the association ‘Pro Igel e.V.’ and the supervisors of the hedgehog care stations from Germany: Karin Oehl, Dielheim; Anette Hübsch, Heidelberg; Gabriele Gaede, Berlin; Siegfried Alber, Neumünster; Buje and Marietta Andersen, Nordsrand; Dr Ilka Preffer, Bad Kösen; Dora Lambert, Berlin; Gabriele Gaede, Berlin Stefanie Meißner, Antenburg; Michaela Kühn, Miltenberg; Rosemarie Adam, Dortmund; Mr Geigenfeind, Schaufling; the supervisors of Igelhaus, Laatzen; the supervisors of Igel-SOS-Donau-Ries E.V., Donauwörth; the supervisors of Igelfreunde Sachsen-Anhalt e.V., Lutherstadt Wittenberg; Beate Kühndelt, München; Mrs Hahn, Haag/Maitenbeth; the supervisors of Arbeitskreis Umweltschutz Bochum e.V., Bochum; Dr Madeiczyk, Coesfeld; Rüdiger Rister, Paderborn; Andrea Hegedüs, Wetzlar; Wilhelm Giesecke, Söhlde; Mr Berger, Frankfurt/Oder. Finally the authors thank Christian Bauer (Institute of Parasitology, Justus-Liebig-University of Giessen, Germany) for his critical reading of the manuscript and helpful suggestions.
