Effect of parasite development upon the host

The first identified stage of infection appeared as red swellings along the length of the intestine (Fig. 1A). The swellings increased in size and number, becoming increasingly opaque, until they started to fill the coelom of the oligochaete (Fig. 1B). At this stage, in live oligochaetes, cylindrical spores could be observed within the swellings. The swellings continued to increase in size causing the body wall of the oligochaete to distend (Fig. 1C). Five days after the initial infection was identified, a portion of the oligochaete ruptured releasing spores, developmental stages and host tissue, into the cell well. The remaining portion of the oligochaete, although alive, was incapable of crawling and survived for several hours before rupturing again for a final time. The released spores were cylindrical 12·2 μm long and 1·6 μm in diameter (Fig. 1D). One oligochaete in the early stage of infection was noted to develop eye-spots midway down its body. The infection persisted in this oligochaete as it underwent binary division. Both resultant individuals were infected with the microsporidian.

Fig. 1. (A) Swellings (caused by infected haemocytes) observed along the line of the intestine during the initial stages of observable infection. Swellings indicated by arrows. (B) Enlargement of infected haemocytes and their increasing opacity within the oligochaete. (C) Distension of body wall of oligochaete by infected cells. (D) Bacilliform spores released from oligochaete. (E) Section through an oligochaete with the early stages of infection. Two infected haemocytes (H) can be observed within the enlarged peri-intestinal sinus (S). Cytoplasmic extensions from an infected haemocyte appear to be interact with the chloragocytes surrounding the sinus (arrow). The intestinal lumen is indicated with an (I). (F) Monokaryotic meronts (arrows) surrounding lobed nucleus (N) of infected cell. (G) Diplokaryotic sporont within hypertrophic haemocyte.

Released spores were difficult to dislodge from the cell well using a pipette, appearing to have an adhesive quality. It was noted that ciliates, introduced into the well with the oligochaete, would rapidly ingest the spores. This appeared to be in preference to the remains of the worm in the well. Even though up to 5 spores were noted within a single ciliate no adverse affects or infection was observed. The ciliates presumably ejected the spores after a period of time as these eventually became evenly distributed over the base of the cell well.

Ultrastructural examination of microsporidian development

The earliest stage of development examined was in an oligochaete with characteristic red swellings down the length of its intestine, but no visible spores. Ultrastructural examination of this oligochaete revealed there to be hypertrophied haemocytes within the peri-intestinal blood sinus (Fig. 1E). These haemocytes contained numerous monokaryotic meronts, which resided in direct contact with the host cell cytoplasm, and consisted of a round nucleus that occupied most of the cell volume (Fig. 1F). Division of the meronts was not noted, but occasionally parasites containing large nuclei with flocculent chromatin, suggestive of nuclear division, were observed. The host cell nucleus was notably enlarged and lobed, while the cell surface possessed many cytoplasmic filaments that appeared to interact with the chloragocytes enclosing the sinus, although direct contact with these cells was not observed. The sinus was markedly enlarged, with the infected cell appearing to be suspended by the filaments in the centre of it.

Sections were examined from oligochaetes when spores were first noted in the infected haemocytes (opaque swellings) and later when the body wall of the oligochaete was distended by spore masses. From these sections, a partial developmental sequence was deduced. The earliest sporogonic stages noted were diplokaryotic sporonts but the transformation of meronts to sporonts was not observed. Initially the sporonts were condensed and contained a well-defined diplokaryon occupying the majority of the cell (Fig. 1G). One sporont was observed apparently in the final processes of dividing by binary division. Sporogenesis occurred directly in the cytoplasm of the host cell and was asynchronous, with sporonts, sporoblasts and mature spores all within the same haemocyte. The initiation of sporogenesis was signalled by the cytoplasm of the sporonts extending, producing an elongate sporoblast, with the diplokaryon positioned at one end (Fig. 2A).

Fig. 2. (A) First stage of sporogenesis with the sporoblast elongating. (B) Transverse section through early sporoblast. Electron-dense material is forming on the surface of the sporoblast with an electron-lucent area immediately beneath it. Arrowheads indicate these two layers. The forming manubrium is composed of 3 layers, an electron-dense outer layer surrounding a granular matrix with an electron-dense core. (C) Transverse section through developing sporoblast. The manubrium is differentiating into layers. The developing exospore and endospore layers now completely encircle the sporoblast with spherical/granular material clearly associated with the exospore. (D) Formation of anchoring apparatus. The exospore and endospore layers of the spore wall are clearly identifiable. The developing anchoring disk (A) caps the manubrium and is surrounded by the polar sac (PS). The manubrium is surrounded by an additional lucent layer (arrow). (E) Transverse section through mature spore. The manubrium is composed of 7 layers with a wide endospore surrounded by a relatively thin exospore. (F) Section through surface of spore wall. The exospore is embedded with granular material. (G) Section through anterior of mature spore. Anchoring disc (A) is at the apex of the manubrium and extends into the polar sac (PS). The electron-dense anterior polaroplast (AP) is followed by a globular posterior polaroplast (PP). (H) Mature spore of Bacillidium vesiculoformis. Structures labelled are diplokaryon (D), posterior dense bodies (B), manubrium (M) and the polar filament (arrow).

During sporogenesis the development of the spore wall and the development of the manubrium occurred in concert. This was clearly observed by transverse sections through sporoblasts by which the development of the spore could be charted. After elongation of the sporoblast, areas of cytoplasm associated with the plasma membrane became increasingly electron-lucent with the surface of the sporoblast becoming thickened and electron-dense (Fig. 2B). The manubrium at this time appeared to consist of an electron-dense outer layer surrounding a granular matrix with an undifferentiated dense core. The changes on the surface of the plasma membrane continued, extending around the sporoblast, with the electron-dense layer becoming recognizable as the developing exospore and the electron-lucent layer beneath it the endospore. Spheroids of material resembling granules measuring approximately 25 nm in diameter were associated with the developing exospore but their origin could not be determined. The manubrium continued to differentiate with the granular layer becoming homogenous and a further electron lucent layer forming around the structure (Fig. 2C, D). The earliest observed sections demonstrating the formation of the anchoring apparatus occurred when the manubrium extended the entire length of the cell, and the developing endospore and exospore were clearly differentiated (Fig. 2D). The developing apparatus, at the anterior of the sporoblast, consisted of a dense anchoring disc, covering the core of the manubrium and capped by a moderately dense polar sac that extended around the anchoring disc and the apex of the manubrium.

As the endospore expanded in diameter, the manubrium continued to differentiate into more layers, and the developing spores became more regular in transverse section. With the development of the endospore, fixation of the spores became increasingly poor. The contents of the vast majority of mature spores were not fixed.

Mature spores were roughly cylindrical with a manubrium composed of 7 layers, an endospore of 100 nm and an exospore of 25 nm (Fig. 2E). The layers of the manubrium alternating electron lucent (layer 1), dense (layer 2), lucent (layer 3), moderate (layer 4), dense (layer 5), light (layer 6) with a moderately dense core (layer 7). The spore surface was covered in dense spherical granules that appeared to be embedded in the exospore (Fig. 2F). In the mature spore the polar sac was a comparatively flat, crescent structure that was composed of dense material (Fig. 2G). The anchoring apparatus appearing as a moderately lucent area beneath the sac, capping the outer layers of the manubrium and appearing continuous with layer 4 of this structure. The apparatus appeared to extend into the polar sac region where this structure became electron lucent. The anterior part of the polaroplast was an electron dense, cup like body that extended back from the polar sac into the spore. It was 1/10 the length of the spore and was lined with two 5 nm thick unit membranes although the internal organization of this area of the polaroplast could not be discerned. The posterior polaroplast did not have a clearly definable structure but appeared to contain electron-dense globular material.

The manubrium extended nearly the entire length of the spore, narrowing to a short polar filament at the posterior. The diplokaryon could be distinguished, with the nuclei lying in tandem along the side of the spore while in the posterior of the spore there were spherical dense bodies of various sizes (Fig. 2H).

The developing spores were initially contained within the cytoplasm of the host cell. As the infection progressed and the spores matured so the cytoplasm became reduced and vacuous, appearing to be replaced by an electron-lucent matrix. The hypertrophied cell at this stage was still contained within the peri-intestinal sinus, being encircled by host cells of uncertain characterization, presumably leucocytes, underlying the chloragocytes of the sinus (Fig. 3A). As the cytoplasm of the infected cell became replaced by the electron-lucent matrix so it increased in size. The body wall of the oligochaete became reduced at points and the infected cell was noted to almost breach the exterior of the worm (Fig. 3B).

Fig. 3. (A) Haemocyte with reduced cytoplasm containing mature spores (S) contained with the peri-intestinal sinus (B) surrounded by a layer of host cells (*). The surrounding chloragocytes of the sinus are labelled with an arrow. Exterior of oligochaete indicated with (E). (B) Enlarged haemocyte containing mature spores (S) breaking through the cuticle of the oligochaete. Chloragocytes surrounding the haematocyte indicated with an arrow. Exterior of oligochaete indicated with (E).

Phylogenetic analysis

1580 bp of sequence (GC content 46·5%) representing the complete SSU, ITS, and partial LSU rDNA of the parasite was successfully amplified and the sequence has been deposited in GenBank, Accession No. AJ581995. The FASTA search revealed that the sequence had the closest homology to the bryozoan parasites, Pseudonosema cristatellae and Trichonosema pectinatellae, and the oligochaete parasite, Bacillidium sp. Parsimony, neighbour-joining and maximum likelihood analyses all gave the same overall tree topography placing the parasite within the clade containing Bacillidium sp., Schroedera plumatellae and Pseudonosema cristatellae (Fig. 4). Therefore, based on ultrastructural characteristics, host type and molecular data, we propose to name this parasite Bacillidium vesiculoformis.

Fig. 4. Phylogenetic tree of microsporidia based on SSU rDNA using maximum likelihood. Bacillidium vesiculoformis is placed within the poorly resolved clade containing both bryozoan (e.g. Pseudonosema cristatellae) and oligochaete (Bacillidium sp.) parasites (highlighted box). The boot strap values (100 replicates) on the tree represent the percentage of bootstrap replicates that gave that topology. Neighbour-joining and parsimony analyses resulted in identical placement of B. vesiculoformis within the tree (data not shown).

Systematic Position

Phylum Microspora, Class Dihaplophasea, Order Dissociodihaplophasida, Family Mrazekiidae

Description of the species

• Name: Bacillidium vesiculoformis n. sp.
• Type host: Nais simplex Piguet, 1906 (Oligochaeta: Naididae)
• Type locality: Southern end of Airthrey Loch, University of Stirling campus. National grid reference NS 804963.
• Infected cell type: Haemocyte
• Type material: Hapantotype material has been deposited in the collection of the Natural History Museum, London. This material comprises embedded blocks of oligochaetes containing early stages of parasite development (registration number 2004[ratio ]6[ratio ]8[ratio ]2) and later developmental stages including sporonts and spores (registration number 2004[ratio ]6[ratio ]8[ratio ]1).
• Etymology: The specific name is derived from the latin for ‘bladder forming’ in reference to the enlargement of the blood sinus around the infected haemocytes.
• Description of spores: Diplokaryotic spores, bacilliform, 12·2 μm±0·4 μm in length, 1·6 μm±0·2 μm (n=20) in width, with a manubrium composed of 7 layers that extends nearly the entire length of the spore, tapering at the posterior to a polar filament. Elongate diplokaryon positioned midway along spore. Electron-dense anterior polaroplast occupying one tenth of the spore length, while the posterior polaroplast appeared globular of uncertain length. Electron-lucent endospore, 100 nm in diameter, surrounded by dense exospore, 25 nm in diameter, embedded with granular material on its surface.

Transmission studies

Neither transmission study induced an observable infection. However, while the oligochaetes were observed to graze on the spores introduced into their well, ingestion of the spores into the bryozoan colonies was more difficult to assess. The majority of spores generally sank quickly out of the water current generated by the lophophore. However, occasional spores were noted to remain in the current to be transported towards the lophophore of the bryozoan's zooid. However, as soon as one spore touched a tentacle the lophophore would immediately retract and the filter-feeding stop. As such, although contact between spores and bryozoa was observed, actual ingestion of the spores was not ascertained. At the end of the experiment, no colonies appeared to be infected.