Ultrastructure and development of Nosema podocotyloidis n. sp. (Microsporidia), a hyperparasite of Podocotyloides magnatestis (Trematoda), a parasite of Parapristipoma octolineatum (Teleostei)

Nosema podocotyloidis n. sp. (Microsporidia, Nosematidae) is described from Podocotyloides magnatestis (Trematoda: Opecoelidae), a parasite of the fish Parapristipoma octolineatum (Teleostei) in the Atlantic Ocean. Electron microscopy reveals that all the stages of the cycle (merogony and sporogony) are diplokaryotic and in direct contact with the cytoplasm of host cells. There is no sporophorous vesicle (pansporoblast). The earliest stages observed are meronts, which have a simple plasmic membrane. Their cytoplasm is granular, rich in ribosomes and contains some sacculi of endoplasmic reticulum. They divide by binary fission into diplokaryotic sporonts. The sporonts have a thick electron-dense wall. Their diplokaryon is slightly less electron-dense than the cytoplasm. The cytoplasm of more advanced sporonts has numerous electron-lucent vesicles. Sporonts with two diplokarya divide by binary fission into diplokaryotic sporoblasts. The older sporoblasts are irregular or elongate and the polar filament is in formation. Their cytoplasm is denser, with ribosomes and lamellae of granular endoplasmic reticulum. The sporoblasts evolve into spores. The mature spores are broadly oval and measure 3.6 (3.1–4.0) × 2.58 (1.8–3.3) μm. Their wall is 100–300 nm thick. The polar tube is isofilar with 11–16 coils, 130–155 nm in diameter and arranged in many layers in the centre of the spore. The polaroplast is divided into two regions: an outer electron-dense cup with granular content and lacking lamellae and an internal region, less electron-dense, composed of irregularly arranged sacs. The posterior vacuole, with an amorphous electron-dense content, is present. The new species is compared with other species of Nosema from trematodes.


Introduction
Podocotyloides magnatestis (Trematoda: Opecoelidae) is a parasite in the gut of the teleostean fish Parapristipoma octolineatum (Valenciennes, 1833) off the coast of Senegal. While studying this parasite, we found that some specimens were hyperparasitised by a microsporidia.
The genus Nosema is identified by the following characters: all the stages of the cycle are diplokaryotic and in direct contact with the cytoplasm of host cells, merogony by binary division and ends with diplokaryotic sporonts, sporogony disposroblastic, spores diplokaryotic and generally ovoidal [9,22].
The genus Unikaryon was established for the type species Unikaryon piriformis, a hyperparasite of Echninoparyphium dunni and Echinostoma audyi, both parasites of the snail Lymnaea rubiginosa [3,4,9]. The essential characters of this species are: nuclei are isolated at all stages of development, the sporogony is disposroblastic and the development occurs in contact with the host cell cytoplasm [5,9,23].
For the genus Pleistophora, essential characters proposed by Canning and Nicholas [6] based on the ultrastructure of the type species P. typicalis from the marine fish Myoxocephalus scorpius are: -nuclei are unpaired in all stages of the development; -merogony stages are bounded by a dense and amorphous wall which detaches from the plasma membrane at the beginning of the sporogony forming the sporophorous vesicle wall; -division of meronts takes place by plasmotomy; -sporogony is polysporous by successive divisions of the sporogonial plasmodium until the formation of uninucleate sporoblasts which develop into spores; -there are numerous spores in the sporophorous vesicle.
The collective group Microsporidium Balbiani, 1984 has been created by Sprague [22]. It regroups the species whose generic positions are uncertain.
In this paper, we describe a new species, assign it to the genus Nosema Naegeli, 1857, and compare it with the species of Nosema from trematodes previously studied using the same techniques. Site of infection: Parenchyma. Development: All stages of merogony and sporogony are diplokaryotic and in close contact with the host cell cytoplasm. Sporogony is diplosporoblastic.
Etymology: After the host name.

Meronts and merogony
The merogonic stages were diplokaryotic cells with a thin plasmic membrane ( Fig. 1B-D). They were in direct contact with the host cell cytoplasm. They were never seen in vacuoles. The cytoplasm was uniform, granular and contained a great number of free ribosomes. There were only traces of an endoplasmic reticulum (Fig. 1B, C). The nuclei were large, paired with their flattened sides in close contact, and sometimes occupied two-thirds of the cytoplasm. The greatest sectioned nuclei measured, on average, 1.7 lm in diameter. The nucleoplasm was slightly less electron-dense than the cytoplasm. Occasionally a central nucleolus was visible (Fig. 1B). Each nucleus was bounded by an envelope consisting of two unit membranes which appeared relatively dense in the contact area (Fig. 1B).
The meronts with one diplokaryon were ovoid (Fig. 1B) or elongate (Fig. 1C). Those with two diplokarya were elongate and divided by binary fission (Fig. 1C) into two daughter cells (meronts). It is unknown if there is more than one bout of merogony. The meronts of the last generation matured into sporonts.

Sporonts and sporogony
At the beginning of the sporogony, the meronts of the last generation transformed with the secretion of a coat of amorphous electron-dense material on the plasmic membrane ( Fig. 2A). The coat is deposited at first in irregular clumps which cover the plasmic membrane. The sporonts thus formed were bounded by a thick electron-dense wall (the thickness attained was about 20 nm), were oval (Fig. 2B) or elongate (Fig. 2C) and had one or two diplokarya. The cytoplasm of early sporonts contained numerous dispersed ribosomes, a small number of electron-lucent vesicles and a few parallel arrays of rough endoplasmic reticulum (Fig. 2B). The diplokaryon was slightly less electron-dense than the cytoplasm. It sometimes occupied more than half the cytoplasm. Nucleoluslike formations were occasionally seen (Fig. 2B).
The sporonts with two diplokarya (Fig. 2C) divided by binary fission into two sporoblasts.

Sporoblasts and sporogenesis
The young sporoblasts were ovoid cells with one central diplokaryon (Fig. 3A). They were bounded by an electron-dense coat with a thickness of approximately 30 nm. The electron-lucent vesicles of the cytoplasm increased in number and the endoplasmic reticulum became more distinct. The older sporoblasts were irregular or elongate (Fig. 3C) and their cytoplasm was full of ribosomes and lamellae of endoplasmic reticulum. The diplokaryon was less electron-dense than the cytoplasm. Distinct nucleoli were occasionally seen (Fig. 3C). The sporogenesis began with the development of a polar tube (Fig. 3B, C). Cross-sections of the immature polar tube appeared as symmetrical rings, each with an electron-dense central axis surrounded by a layer of electronlucent material limited by a membrane (Fig. 3C). The diameter of the immature polar tube was approximately 80-120 nm. In the posterior end of some older sporoblasts, a prominent Golgi apparatus developed near the polar tube (Fig. 3B, C). Its structure was of traditional type. The polaroplast was the last spore structure formed and consisted of highly electron-dense bands (Fig. 3D). At this time, the formation of the endospore wall began. The young spores were diplokaryotic (Fig. 3D).

Mature spores
The mature spores were broadly oval (Fig. 4A). In thin sections, the spore dimensions calculated were 3.6 (3.1-4.0) · 2.58 (1.8-3.3) lm (n = 25). The diplokaryon was found in the posterior part of young spores (Fig. 3D) but it was not clearly visible in mature spores.
The spore wall was about 100-300 nm thick and consisted of three parts: an electron-dense exospore, a median  electron-lucent endospore and an internal unit membrane (Fig. 4A). The endospore was reduced in thickness (100-160 nm) at the anterior end of the spore (Fig. 4B).
The polar tube was isofilar with 11-16 coils, 130-155 nm in diameter, arranged in many layers in the middle region of the spore (Fig. 4A). The fine structure of the coils showed a multilayer structure (Fig. 4C). The polar tube was attached to an anchoring disc, surrounded by the polar sac (Fig. 4B). The widest sectioned disc measured 155 nm in diameter.
The polaroplast, surrounding the anterior part of the polar tube, was divided into two regions: an anterior electron-dense cup, and a posterior region, less electron-dense, composed of irregularly arranged sacs (Fig. 4A, B, D). In the posterior part of the spore was a posterior vacuole, more electron-dense than the cytoplasm (Fig. 4A).

Discussion
The species described here belongs to the genus Nosema Naegeli, 1857 as defined by Larsson [15], Sprague [22] and Canning and Vávra [9]. The following characteristics supported this generic identification: -nuclei were paired as diplokarya at all stages of development; -development at all stages was in direct contact with the host cell cytoplasm; -merogony and sporogony by binary fission of diplokaryotic cells.
Nosema diphterostomi was described as a hyperparasite of adults of Diphterostomum brusinae, an intestinal parasite of Diplodus annularis [16].The most distinctive characters of this species are the low number of coils of the polar tube (6-7 coils), the small diameter of the polar tube (100 nm) and the small size of the spores (2.1 · 1.4 lm).
Nosema dollfusi was described in the sporocysts of Bucephalus cuculus, a parasite of the oyster Crassostrea virginica from Maryland in the USA [21]. The ultrastructure of this species is unknown. This Nosema is differentiated from N. podocotyloidis n. sp. by its host which, in the larval stage, parasitises a mollusc.
Nosema eurytremae is a microsporidian hyperparasite of larvae of the trematodes Eurytrema pancreaticum and Postarmostomum galilinum in the land snail Bradybaena similaris. The most distinctive characters of this species are the coils of the polar tube which are arranged in a single layer close to the spore wall, the polaroplast, which consists of an anterior part composed of laminated membranes and a posterior one composed almost entirely of flattened spindle-shaped sacs, and the hosts which, in the larval stage, parasitise a land mollusc [11].
Nosema gigantica was found in the parenchyma of adult flukes, Allocreadium fasciatasi, living in the gut of the freshwater fish Aplocheilus melastigma from India. Its spores are ellipsoid and measure 7.9 · 4.9 lm [4]. The size of the spores of this species is different from that of N. podocotyloidis n. sp.
Nosema lepocreadii is a microsporidian hyperparasite of adult flukes, Lepocreadium manteri, from the gut of the California grunion, Leuresthes tenuis. It was studied by light and electronic microscopy [7,8]. The most distinctive ultrastructural features of this species are: -the diplokaryotic sporoblasts are not always produced by binary fission of the sporonts but also by multiple fission of elongate sporonts with more than two diplokarya; -the endoplasmic reticulum is abundant in all pre-spore stages; -the polar tube is isofilar with a maximum of 10 coils arranged in a single layer; -the subdivision of the nuclei occurs by ingression of the inner membrane of the nuclear envelope as tongues into the nucleoplasm.
Nosema monorchis is a hyperparasite of Monorchis parvus, an intestinal parasite of Diplodus annularis. It exhibits the following distinctive characters: the great electron opacity and small size of the diplokarya; the polar tube is isofilar with 16-17 coils (90 nm diameter); the polaroplast presents an anterior part with closely packed lamellae and a posterior region with wider or irregularly arranged lamellae and a maximum wall thickness of 220 nm (exospore + endospore) [17].
Nosema strigeoideae is a hyperparasite of larval stages of Diplostomum flexicaudum in the snail Stagnicola emarginata angulata from Michigan in the USA [14]. Its ultrastructure is unknown. It is differentiated from N. podocotyloidis n. sp. by its spores, which measure 4.7 · 3.1 lm, and by its host which, in the larval stage, parasitises a mollusc.
Nosema xiphidiocercariae is a hyperparasite of sporocysts, cercariae and metacercariae in Plagiorchiidae parasites of Lymnaea palustris, a freshwater mollusc from Russia. The live spores measure 4.5 · 2.3 lm and the spores coloured with Giemsa measure 4.0 · 2.3 lm [22]. This Nosema is differentiated from N. podocotyloidis n. sp. by the spore size and by the fact that its host lives in a freshwater mollusc.
Therefore, we consider that N. podocotyloidis n. sp. is different from all other species of Nosema hyperparasites of trematodes.
The most prominent feature of the spores of N. podocotyloidis n. sp. is the anterior part of the polaroplast. It is an electron-dense Moscow Russia [22] cup with granular contents. This type of anterior region of the polaroplast is uncommon. It has been observed only in Nosema lepocreadii, a hyperparasite of Lepocreadium manteri [8]. Usually the polaroplast has two parts, an anterior or outer region with closely-packed lamellae and a posterior or inner region with wider and less regularly arranged lamellae or vesicular units.