Myxobolus arariensis, Abrunhosa & Sindeaux-Neto & Santos & Hamoy & Matos, 2018

Abrunhosa, Jacqueline, Sindeaux-Neto, Jose L., Santos, Sidney, Hamoy, Igor & Matos, Edilson, 2018, A new species of myxozoa in the skeletal striated musculature of Rhamdia quelen (Quoy & Gaimard) (Siluriforme: Pimelodidae) Amazonian fish, Marajó island, Brazil, Zootaxa 4482 (1), pp. 164-176 : 166-172

Myxobolus arariensis n. sp.

( Figs 1–3 )

Morphological description. Mature spores are ellipsoidal in shape, with a mean length of 11.4 µm (10.7– 12.6) and mean width of 7.2 µm (6.4–7.9). Each spore contains two polar capsules (PCs) of equal size 4.0 ± 0.7 µm (3.6–4.3) long and 1.9 ± 0.36 µm (1.7–2.2) in width ( Figs. 1B and 3 ) ( Table 1).

Type host. Rhamdia quelen (Quoy & Gaimard 1824)

Site of infection. epaxial and hypaxial layers of the musculature, with plasmodia containing numerous spores.

Type-locality. Arari River , Cachoeira do Arari on Marajó Island, northern Brazil.

Etymology. The species was named for the locality of origin, the Arari River in northern Brazil.

Specimens deposited. Microscope slides containing spores from the muscle layer, prepared using the paraffin technique, stained in Gutierrez and mounted in Entellan were deposited in the International

Protozoan Type Collection of the National Institute for Amazonian Research ( INPA) in Manaus , Amazonas state, Brazil (catalog number: INPA /027). The partial 18S rDNA sequence was deposited in GenBank under accession number MG572219 .

Prevalence. Three of twenty-five R. quelen examined 12% (3/25) had plasmodia of an unknown parasite from the genus Myxobolus .

Remarks. M. arariensis can be differentiated morphologically from all seven Myxobolus species known to infect the muscle tissue of freshwater fish ( Table 1). The new species can be distinguished from M. tasikkenyirensis (Székely et al. 2009a) and M. groenlandicus ( Buchmann et al. 2012) by the different shape of the anterior extremity of the spores, and from M. leptobarbi (Székely et al. 2009b) by the same trait. The length of M. arariensis (11.4 µm) is most similar to that of M. lentisuturalis (Dyková et al. 2002) , which is 11.8 µm long, whereas M. omari (Székely et al. 2009b) is the shortest species, at 7.9 µm. Anomalous spores with a caudal filament and lack of ornamentation on the external wall were also observed ( Fig. 2C ).

Histology. The histological analysis revealed the presence of cysts of M. arariensis lodged in the fibers of the skeletal muscles ( Fig. 4 ). Immature spores were observed in the most external layer of the cyst, with mature spores being found more internally. The cyst wall is thick and fibrous, and the adjacent musculature was compressed, with the sarcoplasm frayed, and evidence of a necrotic reaction caused by this compression.

Molecular data. In the molecular analysis, the specific pair of myxozoan primers (MC5-MC3) amplified 974 bps of the 18S rDNA gene of the spores obtained from the plasmodia found infecting the musculature of R. quelen . The BLAST search of the 18S rDNA sequence data (974 bps) of the Myxobolus species parasitizing R. quelen found no identical myxozoan sequence in GenBank, although a similarity of at least 85% was found with four species: Myxobolus cordeiroi ( KF296353 , 90% similarity), Myxobolus sp. GA2 ( KU 170935 , 86%), Myxobolus lentisuturalis ( AY278563 , 85%), and Myxobolus cultus ( HQ613409 , 85%).

The optimal evolutionary model for maximum likelihood (ML) and Bayesian analysis were determined by jModelTest 3.0 (Posada, 2008) which identified the best evolutionary model as the general time reversible model (GTR + I + G), using Akaike information criteria. Nucleotide frequencies were estimated from the data (A = 0.2574, C = 0.1848, G = 0.2625, T = 0.2326) and six rates of nucleotide substitution calculated as AC = 0.8659, AG = 2.6388, AT = 1.7658, CG = 0.4883, CT = 3.4814, GT = 1.000. The proportion of invariable site was 0.5565 and the alpha value of gamma distribution parameter 0.3612. Two independent runs were conducted with 4 chains for 2 million generations for Bayesian analysis. Ceratomyxa shasta ( AF001579 ) e C. amazonensis ( KX236169 ) was designated as outgroup. Phylogenetic trees were sampled every 100 generation.

Characteristics of some Myxobolus species. Abbreviations: FC = Capsule Formate, SL, spore length, SW = Spore Width, PCL = Polar Capsule Length, PCW= Polar Capsule Width. PC = relative size of the polar capsules (= = equal in size, # = different in size, or equal and different); All measurements are given in micrometers.

In the phylogenetic analysis, trees generated by Bayesian Inference (BI) had similar topologies, but with different support values at some nodes. A strong clustering tendency was found according to phylogenetic affinities. The phylogram indicated the existence of three clades, A, B and C the first paraphyletic, includes species of Henneguya and Myxobolus formed by the freshwater and marine water ( Mugiliformes ).

The clade A subdivide into 2 subclades, A1 and A2. The subclade A1 shows M. arariensis grouping on the same branch with M. cordeiroi and with adjacent subclade with M. marajoensis species, having same host. The other subclade, A2 have the presence of the Myxobolus and Henneguya , corroborating the characteristic of the Myxobolus genus to be paraphyletic. On the other hand, the clades B and C have agrouped species of various Myxobolus parasites of freshwater fish belonging to same and different Order ( Table 3).

The type of host defines a well-supported freshwater and marine water clade of Myxobolus and Henneguya (clade A) that infect fish of the orders Siluriformes , Mugiliformes , Characiformes and Perciformes . However, M. arariensis , which infects silurids, evolved independently from the Myxobolus subclade A1 that infects fish of the families Pimelodidae and Ictaluridae . The tree presented a similar topology for the other clades, clustering according to the taxonomic order of the host. The clade B groups Myxobolus , of the orders Cypriniformes and Perciformes . The clade C is also compounded by Myxobolus groups that parasitize hosts of the different orders Salmoniformes , Siluriformes , Characiformes and Cypriniformes ( Fig. 5 ).

The p distance found between M. arariensis and any other Myxobolus species that had Siluriformes as host ranged from 11.6 to 22.1% ( Table 2), which reinforces the definition of M. arariensis as a new species. The myxosporidian sequences analysed are showing in the Table 3.

……continued on the next page

Figures

FIGURE 1. A. Musculature of Rhamdia quelen (Quoy and Gaimard 1824) infected by Myxobolus arariensis n. sp. showing the whitish, ovoid pseudocyst (arrow) Scale bar: 0.6 mm. B. Fresh spores, frontal view (arrow heads). Scale bar: 10 µm.

FIGURE 2. Fresh spores of Myxobolus arariensis n. sp. from Rhamdia quelen (Quoy and Gaimard 1824). A – The frontal view of the spore. Scale bar: 5 µm. B—Lateral view of the spore. Scale bar: 5 µm. C—Spores of Myxobolus arariensis n. sp. showing external ornamentation (arrow head) and anomalous external morphology (*), with caudal filaments (arrows), but no ornamentation of the external wall. Scale bar: 10 µm

FIGURE 3. Schematic drawing of Myxobolus arariensis n. sp. spore in frontal view found in the muscle. Spore in valvular view, showing its internal organization.

FIGURE 4. A. Histological section of the cyst and musculature of Rhamdia quelen stained with Gutierrez, showing mature spores (MS) and imature spores (IS) of Myxobolus arariensis n. sp. (*), causing bulging (arrow) and disorganization of the muscle fibers (M). B. thick fibrous (F) wall and the adjacent musculature (M).

FIGURE 5. Phylogenetic tree generated by Bayesian Inference (BI) applied to the partial SSU rRNA gene sequences of Myxobolus arariensis sp. n. and related myxosporeans, rooted at Kudoa thyrsites and K. alliaria. The GenBank accession numbers are shown adjacent to the species names. The numbers at the nodes indicate the bootstrap support for the Bayesian Inference. M. arariensis is shown in bold type. The taxonomic order of all the species selected for the phylogenetic analysis is shown. Any values below 50% are indicated by an asterisk. GenBank Accession numbers follow the species name.