Introduction

Myxobolus Bütschli, 1882 is a genus with the richest species in myxozoan. It has been reported to have more than 850 species (Eiras et al. 2005; Lom and Dyková 2006), most of which produce morphologically similar or identical spores from genetically related hosts. Therefore, its species identification merely based on their morphological characteristics is difficult and much more confusions will be caused. Fortunately, with the development of molecular biology, more and more new technologies have been widely applied in the classification and identification of Myxobolus species harbored from genetically closely related hosts and infection sites (Bartošová et al. 2009; Fiala 2006; Kent et al. 2001; Khlifa et al. 2012; Morsy et al. 2012; Whipps et al. 2004a, b; Zhao et al. 2008). To avoid mis-identification of myxosporean species, researches have recently suggested that morphology, host or organ specificity, tissue tropism, and molecular data should all be taken into consideration (Bartošová et al. 2009; Cone and Overstreet 1998; Dyková and Lom 2007; Lom and Dyková 2006; Molnár 1994; Morsy et al. 2012; Zhao et al. 2008).

Since Zhao et al. (2008) originally found Myxobolus ampullicapsulatus without causing any disease from the gills of Carassius auratus in Chongqing, China, described it in detail, and obtained its small subunit ribosomal DNA sequence, some researchers have studied this species and the related taxa from goldfish, crucian carp, and allogynogenetic gibel carp (Liu et al. 2012; Xi et al. 2011; Zhang et al. 2009, 2010). Among them, Myxobolus wulii (Wu and Li 1986), M. ampullicapsulatus Zhao et al. (2008) and Myxobolus honghuensis Liu et al. (2012) were found in Carassius auratus gibelio in China and M. wulii from the gills and hepatopancreas of goldfish in Japan (Liu et al. 2012; Xi et al. 2011; Zhang et al. 2010). Phylogenetic analyses of small subunit ribosomal RNA gene sequences showed that M. wulii was closely related to M. ampullicapsulatus (Zhang et al. 2010). Meanwhile, Xi et al. (2011) isolated one Myxobolus species from the gills and pharynx of C. auratus gibelio (Bloch) reared in heavily infected pond in Jiangsu area, China and found that it could cause serious disease for host fish. After sequence comparison among the related Myxobolus species, they considered it as M. ampullicapsulatus based on the high similarity of SSU rDNA sequences. M. honghuensis from C. auratus gibelio with serious disease in Hubei area could be distinguished from M. ampulicapsulatus by its equal polar capsules and more polar filament turns and regarded as an independent species (Liu et al. 2012). In this study, we researched the differences and clarified the confusions among M. wulii, M. ampullicapsulatus, and M. honghuensis based on the genetic and molecular information.

Materials and methods

Sample collection and species identification

The host fish, Carassius auratus auratus were collected during July 2010 and July 2011 from Yongchuan, Chongqing area and Poyang Lake, Jiangxi area, China, necropsied, and examined under the binocular dissecting microscope at ×400 to detect myxosporeans. Plasmodia of three populations of M. ampullicapsulatus were collected from the gills of the host fish, transferred on microscope slides, and ruptured to release spores. After rinsing three times with sterile distilled water, fresh spores were pelleted by centrifugation at 2,000×g, prepared as specimens as previously reported (Zhao et al. 2001), and observed and measured at ×1,000 magnification using a Nikon E600 microscope. The illustrations based on fresh materials were drawn with the aid of the camera lucica and computer software. Photomicrographs were taken using a Nikon DXM1200 microscope at ×1,000 magnification, and measurements based on 25 spores were given in microns (μm) as the arithmetic mean and standard deviation, followed by the range in parentheses. Species identification was conducted as previously reported (Zhao et al. 2001).

DNA extraction, cloning, and sequencing

Genomic DNA of myxosporean isolates from the fish gills were extracted using DNeasy Tissue Kit (QIAGEN). 18S rDNA was amplified using primers ERIB1: 5′-ACC TGG TTG ATC CTG CCA G-3′ and ERIB10: 5′-CTT CCG CAG GTT CAC CTA CGG-3′ at the following conditions: initial denaturation at 94 °C for 5 min followed by 35 cycles of 1 min at 94 °C, 1 min at 56 °C, and 2 min at 72 °C as well as a final elongation at 72 °C for 10 min. Products were purified using a Gel Extraction Kit, measured with the NanoDrop ND-1000 spectrophotometer, inserted into pMD19-T vector, and sequenced in both directions using the ABI PRISM ® 3730 DNA sequencer.

Phylogenetic, sequence, and P-distance analyses

All sequences except those of the newly obtained SSU rRNA genes were obtained using the basic local alignment search tool (BLAST) from GenBank. Sequences selected include those of Myxobolus species which is morphologically similar to M. ampullicapsulatus. Tetracapsuloides bryosalmonae was used as an outgroup taxon. Twenty-two sequences were aligned using Clustal W. Maximum likelihood (ML) analyses were performed using PAUP*4.0b10 (Swofford 2003). Bootstrap confidence values were calculated with a heuristic search using simple sequence addition and 100 replicates. Bayesian analyses were conducted by MrBayes (Ronquist and Huelsenbeck 2003) with parameters setting to 1,000,000 generations and 10,000 trees. Sequences were assembled and manually edited in BioEdit (Hall 1999). In addition, the similarities of seven sequences were calculated by GenBank BLAST, and the P-distance was calculated using MEGA5 (Tamura et al. 2011).

Results

Morphology remarks

Infection sites and plasmodia morphology for three populations of M. ampullicapsulatus collected from Chongqing area and Poyang Lake, Jiangxi area, China show resemblance to the original population of M. ampullicapsulatus (Zhao et al. 2008). Mature spores were pyriform with bluntly pointed apex and rounded posterior in valvular view (Fig. 1 and Table 1). Generally, the population of M. ampullicapsulatus from Poyang Lake possesses narrower and slightly elongated spores compared with that of M. ampullicapsulatus from Chongqing. In addition, the population of M. ampullicapsulatus isolated from goldfish and crucian carp can be distinguished from M. ampullicapsulatus isolated from the pharynx and gills of C. auratus gibelio from Jiangsu by having a lower ratio of length to width and more bluntly pointed apex (Xi et al. 2011), but displays obvious similarity in both spore morphology and host type to M. honghuensis from the pharynx and gills of C. auratus gibelio in Honghu Lake (Liu et al. 2012). Based on spore morphology and host specificity, M. ampullicapsulatus was mainly parasitic on the gills of goldfish and crucian carp. Thus, M. ampullicapsulatus collected from the pharynx and gills of C. auratus gibelio should be classified as M. honghuensis.

Fig. 1
figure 1

Comparison of spores among the populations of M. ampullicapsulatus and M. honghuensis morphologically. a M. ampullicapsulatus CQLg (from Zhao et al. 2008). b M. ampullicapsulatus CQ collected in the present study. c M. ampullicapsulatus JXa.a collected in the present study. d M. ampullicapsulatus JSa.g (from Xi et al. 2011). e M. honghuensis HBa.g (from Liu et al. 2012). Bar = 10 μm

Full size image
Table 1 Morphometrical comparison between populations of M. ampullicapsulatus and M. honghuensis and its synonyms from different sources
Full size table

Sequence analyses

The three new sequences of M. ampullicapsulatus were obtained and submitted to GenBank (Fig. 2), and their lengths are shown in Table 1. Based on GenBank BLAST searches, the three new sequences were most similar to that of M. ampullicapsulatus CQLg (DQ339482) with similarities of 99, 100, and 100 %, respectively, and to that of M. ampullicapsulatus HBa.a (JQ690373) with similarity of 99 % for all (Table 2). Furthermore, Clustal W multiple alignment shows that the coverage length of seven sequences was 1,554 nucleotides, and sequence analyses show that 18S rDNA sequences had little variability among the former five populations of M. ampullicapsulatus (Fig. 3). A total of ten variation sites was scattered in the five sequences. However, 24 variation sites were found in both populations of M. ampullicapsulatus JS and M. ampullicapsulatus CQLg (Zhao et al. 2008; Xi et al. 2011) and only two variation sites in both populations of M. ampullicapsulatus JS and M. honghuensis (Liu et al. 2012). The similarities of the seven sequences are listed in Table 2 and Fig. 3.

Fig. 2
figure 2

Regions of alignment of SSU rRNA gene sequences of populations of M. ampullicapsulatus and M. honghuensis. Regions are separated by vertical lines and the reference sequence. Insertions and deletions were compensated by introducing alignment gaps (dash). Matched sites are represented by dots. Distinct sequence signatures of each clade are shaded

Full size image
Table 2 Estimates of evolutionary divergence between sequences and highly similar sequence. SSU rDNA gene sequence genetic distances (lower triangle) and similarities (upper triangle) among seven sequences
Full size table
Fig. 3
figure 3

Phylogenetic tree generated by maximum likelihood analysis of the aligned small subunit ribosomal RNA gene sequences. Nodal supports are indicated for bootstrap confidence values resulting from ML and posterior probability of Bayesian analysis. “-” reflects minor differences between maximum likelihood and other methods. GenBank accession numbers for each taxon are listed. The distance scale is shown under the tree. Schematic of (a) M. ampullicapsulatus and (b) M. honghuensis. Bar = 10 m

Full size image

P-distance analyses

P-distance inferred from 18S rDNA data for the seven sequences are shown in Table 2. Their P-distances to the former five M. ampullicapsulatus (<0.005) are far smaller than their distances to M. ampullicapsulatus JS and M. honghuensis (0.013–0.016). In addition, the P-distances between the latter two is only 0.001.

Phylogenetic analysis

Since the topology of ML tree is similar to that of Bayesian inference (BI) tree, they were combined into one (Fig. 3). Twenty-one myxosporeans were grouped into two main phylogenetic lineages: freshwater lineage and marine lineage. The result of topologies of the two trees strongly supports that M. ampullicapsulatus and M. honghuensis should be placed into two clades. However, M. ampullicapsulatus JSa.g from the pharynx and gills of C. auratus gibelio in Yancheng, Jiangsu area (Xi et al. 2011) is clustered with M. honghuensis from the same host and infection site. The sister clade of M. honghuensis contains five populations of M. ampullicapsulatus from infected gills of C. auratus auratus. Bayesian analyses strongly support a large clade including a sister group formed by M. wulii with another one consisting of Myxobolus koi and Myxobolus longisporus.

Discussion

Morphological comparison for four populations of M. ampullicapsulatus and M. honghuensis indicates that the former three populations of M. ampullicapsulatus have longer pointed apex and ampullaceous polar capsules with longer neck region (Fig. 1), while M. ampullicapsulatus from Jiangsu area and M. honghuensis possess ampullaceous polar capsules with shorter neck region. Moreover, the spores of M. ampullicapsulatus from Jiangsu area can be distinguished from those of the former M. ampullicapsulatus by having a lower ratio of length to width and more bluntly pointed apex and displaying a similarity to M. honghuensis. Liu et al. (2012) have mentioned that M. honghuensis is differed from M. ampullicapsulatus by subtle morphological differences. As for the host range and tissue tropism, M. ampullicapsulatus from Jiangsu and M. honghuensis were both from the pharynx and gills of C. auratus gibelio (Liu et al. 2012; Xi et al. 2011), whereas M. ampullicapsulatus from Chongqing, Jiangxi, and Hubei were all obtained from the gills of C. auratus auratus (Zhao et al. 2008). All the abovementioned morphological characteristics, host specificity, and tissue tropism play important roles in the classification of the complicated myxosporean species (Li et al. 2012; U-taynapun et al. 2011). Therefore, the different host and parasite locations can be used as good references for taxonomy prior to recognizing the importance of the minor differences in the spore shape and size.

Analysis of variation sites based on the alignment of SSU rRNA gene sequences also revealed that M. ampullicapsulatus obtained previously from Jiangsu area is distinct from the five populations of M. ampullicapsulatus in the study. The sequences of the five M. ampullicapsulatus isolated from C. auratus auratus have high identity (99–100 %) to each other but slightly lower identity (98–99 %) to the sequences of two M. ampullicapsulatus isolated from C. auratus gibelio (Table 2). Moreover, the sequence of M. ampullicapsulatus from C. auratus gibelio is nearly identical to that of M. honghuensis from C. auratus gibelio (99 % over 1,996 nt), with only two variation sites. Genetically, these two sequences are intra-specific variations among those reported species of myxozoans (Ferguson et al. 2008; Molnár et al. 2006; Whipps et al. 2004a, b; Whipps and Diggles 2006; Whipps and Kent 2006; Suo et al. 2010). Therefore, defining the population of M. ampullicapsulatus from C. auratus gibelio as M. ampullicapsulatus is debatable, and M. ampullicapsulatus and M. honghuensis from C. auratus gibelio should be regarded as the same species. Blast analyses indicate that M. ampullicapsulatus JSa.a is distinguishable from the other M. ampullicapsulatus sequenced to date. In conclusion, the result indicates that the intra-species identification for the two Myxobolus species mentioned in the present research should be less than ten variation sites.

Depending on estimates of evolutionary divergence among the seven sequences (Table 2) except M. ampullicapsulatus JSa.a, the genetic distances inferred from 18S rDNA sequences for five populations of M. ampullicapsulatus are very close (<0.005). However, their genetic distances to M. ampullicapsulatus JSa.a and M. honghuensis are greater (>0.013). The population of M. ampullicapsulatus JSa.a is genetically closer to M. honghuensis (0.001), indicating that they are in the same species. The study suggests that M. ampullicapsulatus from C. auratus gibelio was mis-identified and should be M. honghuensis.

ML and BI phylogenetic trees strongly suggest that M. ampullicapsulatus and M. honghuensis should be grouped into two different clades: one contains all the populations for M. ampullicapsulatus except M. ampullicapsulatus JSa.a isolated from the pharynx and gills of C. auratus gibelio in Yancheng, Jiangsu area and the other only includes both M. ampullicapsulatus JSa.a and M. honghuensis Liu et al. (2012) from the pharynx of C. auratus gibelio in Honghu Lake, Hubei area. M. ampullicapsulatus was originally described from the gills of goldfish. The two populations of M. ampullicapsulatus in the present research were also from the gills of goldfish or crucian carp, indicating that tissue specificity is an important factor in the evolution of myxozoans (Blaylock et al. 2004; Burger et al. 2007; Cone et al. 2005; Easy et al. 2005; Eszterbauer 2004; Fiala 2006; Kent et al. 2001; Molnár et al. 2006; Whipps et al. 2004a, b; Zhao et al. 2008).

Based on the discussion above, M. ampullicapsulatus seems to be a gill-specific parasite of C. auratus auratus, whereas M. honghuensis is a pharynx or gill parasite, which has a shift in tissue microhabitat. M. ampullicapsulatus and M. honghuensis can be distinguished with careful scrutiny, as the latter has shorter bottleneck, a lower ratio of length to width, and less polar filament turns. Furthermore, the study infers that M. ampullicapsulatus and M. honghuensis are two different species, and the population of M. ampullicapsulatus from Jiangsu area should be regarded as the same species as M. honghuensis. Taken together, this study highlights the importance of DNA sequence comparisons for distinguishing Myxobolus species and also indicates that morphologically similar or identical spores collected from genetically close hosts are important for species classification. The intra-species identification for the two Myxobolus species mentioned in the present research should be less than ten variation sites and more than ten variation sites for inter-species identification of Myxobolus can be inferred. However, some molecular criterion for the intra-species or inter-species identification based on the comprehensive factors, including sequence length, coverage, similarity, variation sites and genetic distances, should be extracted and proved by much more research data.