Nature and Phylogenetic Affiliation

Analyses with an electron microscope clearly showed that epixenosomes are always external to the Euplotidium cortex. Since the epixenosome number is quite constant throughout the ciliate's life cycle, their number must increase prior to the binary fission of Euplotidium. These observations led us to interpret epixenosomes not as extrusomes but, rather, as epibionts. Initial evidence for a cellular nature of the epixenosomes was obtained by DAPI staining that showed the presence of DNA within epixenosomes (Verni and Rosati 1990). Subsequent in situ hybridization yielded inconsistent results (Rosati et al. 1998): epixenosomes were specifically labelled with three different eukaryotic probes while no signal was observed with three different prokaryotic probes, including the oligonucleotide EUB338, the sequence of which was considered to be complementary to all bacterial 16S rRNA sequences known at that time (Amann et al. 1991).

A detailed molecular study was then carried out in order to reconstruct the phylogenetic affiliation of E. arenarium epixenosomes (Petroni et al. 2000) based on comparative sequence analysis of genes coding for the small subunit of rRNA (SSU rRNA). Due to the unexpected, ambiguous results mentioned above, different oligonucleotide primers were used for in vitro amplification of the extracted DNA. A distinct amplification product that was suitable for cloning and direct sequencing was obtained employing one certain combination of bacterial with archeal primers. An oligonucleotide probe was designed to specifically target the 16S rRNA sequence from the E. arenarium culture. In situ hybridization with this probe proved that the retrieved 16S rRNA gene sequence in fact originated from epixenosomes (Fig. 3). Comparison of the sequence with the 16SrRNA sequences available in the ARB database (Ludwig and Strunk 1997) revealed that epixenosomes are bacteria phylogenetically related to Verrucomicrobia (Petroni et al. 2000). The latter represents a recently defined bacterial division (Hedlund et al. 1996) that comprises only a few cultivated, diverse species. Among them, members of the genus Xiphinemobacter are endosymbionts of nematodes (Vandekerckhove et al. 2000). However, the majority of the members of this division are so far only recognised by their 16rRNA gene sequences that were recovered as cultivation-independent clones from a variety of habitats. The closest relatives of epixenosomes appear to be two uncultivated marine clones (Fig. 2) and, among the cultivated species, free living bacteria of the anaerobic genus Opitutus (Chin et al. 2001).

- marine clone Artic97B-4, AY028219 sludge clone H2, AF234711

- hot spring clone OPB35, AF027005

- freshwater clone LD19, AF009974

- marine clone Artic95D-9, AY028220

— Verrucomicrobium spinosum, X90515 _|— Prosthecobacter vaneervenii, U 60013

|_r- Prosthecobacter debontii, U60014

[r Prosthecobacter fusiformis, U60015 L Prosthecobacter dejongeii, U60012

— soil clone DA101, Y07576

-rhizosphere clone Riz6E2, AJ244309

soil clone C019, AF013522 I— Xiphinematobacter brevicolli, AF217462 '-Xiphinematobacter americani, AF217460

rt freshwater clone FukuN18, AJ289992

freshwater clone LD29, AF009975 '

-marine clone MB11C04, AY033323 \

-E. arenarium epixenosome, Y19169

-marine clone CHAB-ll-49, AJ240909

freshwater clone SW59, AJ575730 freshwater clone FukuN111, AJ289985 — Opitutus sp. strain VeSm13, X99392 j— Opitutus sp. strain VeGlc2, X99390

Opitutus terrae, AJ229235 -soil clone 0319-6L10, AF234133

contaminated aquifer clone WCHB1 -41, AF050560

-contaminated aquifer clone WCHB1-25, AF050559

- rumen clone RFP12, AF001775


Fig. 2. Maximum likelihood tree showing the phylogenetic position of epixenosomes inferred from a selection of almost full 16SrRNA gene sequences of verrucomicrobia. A filter that retains only 1,349 positions conserved in at least 50% verrucomicrobia was used. The phylogenetic position of epixenosomes is shown. The habitat source of the environmental sequences is indicated before the clone names and the sequence accession numbers. Subdivisions are indicated at the right of the tree. Bar 10 nucleotide substitutions per 100 nucleotides

Fig. 3. Detail of ejected extrusive apparatuses in phase contrast superimposed with the fluorescent signal of epixenosome specific probe

Like other Verrucomicrobiales (Daims et al. 1999), the 16S rRNA gene sequence of epixenosomes has two base exchanges as compared to 16S rRNA gene sequences of other Bacteria that prevent the binding of the bacterial probe EUB338. This could also explain why, in previous experiments of in situ hybridization, oligonucleotide probes, usually used for bacterial detection, did not react with epixenosomes. Based on the data obtained with the specific probes, the positive results obtained with eukaryotic probes appear to represent artefacts, probably caused by unspecific binding.

Morphological Characteristics

Epixenosomes are present in two different forms. Form I, mainly localised in the central region of the epixenosomal band, are spherical, 0.5 |m in diameter and are surrounded by two membranes enclosing a granular matrix in which a clear central zone, containing thin filaments, can be distinguished. They are able to divide by binary fission (Fig. 4A). Form II epixenosomes (Fig. 4B) are mainly localised in the peripheral regions of the band, egg-shaped (2.2 |m in length and 1 |m in width), unable to divide and have a highly complex cytoplasmic organization compared to other prokaryotes. The following well-defined structures are present (Fig. 4B): 1) a dense, apical dome-shaped zone, 2) an inclusion body, 3) an extrusive apparatus, 4) a basket of tubules. This complex structure is gradually acquired following in a well-defined order (Rosati et al. 1993a) during the transformation of form I to form II. The appearance of the EA in the form of a few concentric layers immersed in the cytoplasm is the first sign of the transformation of form I epixenosomes to form II. Already at this early stage the cytoplasm surrounding the lamellar material is to some extent differentiated as preferential digestion with protease can be observed at its level. Successively, the structures become more complex: the concentric layers, which are tightly wound around a central core, increase in number and randomly dispersed tubule-like structures can be seen in the cytoplasm. In a more advanced step, dense material in the apical region of the organism can be observed in addition to the above-mentioned structures. The tubule-like structures become ordered in bundles. Finally, the epixenosomes become oval in shape and acquire the typical structure of form II. Cytochemical studies have revealed a precise localization of some enzymes. For example, active acid phosphatase, an enzyme cytosolic in prokaryotes but contained in membranous organelles in eukaryotes, in epixenosomes appears to be localised only in the space between the two external membranes, while endogenous peroxidase is only present in the inclusion body and at the top of extrusive apparatus. These data indicate that a functional cell compartmentalization corresponds to the structural complexity, in spite of the absence of true internal membranes (Rosati et al. 1996).

Fig. 4A-C. Epixenosomes at different stages. A Type I epixenosomes (from Verni and Rosati 1990); B Type II epixenosome (from Rosati et al. 1996). BT basket tubules; DZ dome-shaped zone; EA extrusive apparatus; IB inclusion body. Arrow points to a para-somal-like vesicle in the Euplotidium cortex. Bars 1 |m. C An ejecting epixenosome at SEM. Bar 5 |m

Fig. 4A-C. Epixenosomes at different stages. A Type I epixenosomes (from Verni and Rosati 1990); B Type II epixenosome (from Rosati et al. 1996). BT basket tubules; DZ dome-shaped zone; EA extrusive apparatus; IB inclusion body. Arrow points to a para-somal-like vesicle in the Euplotidium cortex. Bars 1 |m. C An ejecting epixenosome at SEM. Bar 5 |m

The Dome-Shaped Zone and Inclusion Body

The upper region of epixenosomes form II, just beneath the membrane, is occupied by a granular dense zone, referred to as dome-shaped zone (DZ) for its shape. It is about 0.9 |m thick with an apical hole of 150 nm. It contains DNA and basic proteins (Verni and Rosati 1990; Rosati et al. 1996). Its appearance is reminiscent of that of eukaryotic heterochromatin (Fig. 4B).

A round inclusion body (RB) with a rather constant diameter (200 nm) is localised under the DZ (Fig. 4B). The density is variable, the inclusion body appears empty after in vivo treatments which interfere with polysaccharide metabolism or after inhibition of protein synthesis. Both effects are reversible. These results indicate that RB contains polysaccharides (presumably storage polysaccharides) and enzymes, in particular endogenous peroxidases. After inhibition, the round inclusion body maintains its shape and size and appears surrounded by a thin layer. The latter can be visualised by staining procedures targeting glycosylated substances. These properties distinguish the structure from the storage granules common in prokaryotes and eukaryotes (Rosati et al. 1996).

The Extrusive Apparatus

The extrusive apparatus (EA) occupies most of the basal region of the epixenosome and extends toward the upper region (Fig. 4B). It has been compared to R-bodies described in a few bacterial species of different eubacterial genera. R-bodies are highly insoluble protein ribbons, typically coiled into cylindrical structures within the cells. R-bodies have been studied most intensively in bacteria of the genus Caedibacter, which are obligate symbionts of paramecia to which they confer the killer trait (Preer and Preer 1982). The genus has now been recognised as a polyphyletic group (Beier et al. 2002). Among free-living bacteria, R-bodies have been reported in the hydrogen-oxidising soil species Pseudomonas teniospiralis (Lalucat et al. 1982), the plant pathogen P. avenae (Wells and Horne 1983), the melanogenic marine bacterium Marinomonas mediterranea (Hernández-Romero et al. 2003), and the anoxygenic phototroph Rhodocista centenaria (Favinger et al. 1989). Several classes of R-bodies have been recognised on the basis of physical dimensions, morphology, and the effect of specific physical and chemical treatments. For example, lowered pH, to below 6.5, is apparently required for types 51 (derived from Caedibacter taeniospiralis symbiont of Paramecium tetraurelia) and Pa (synthesised in the free-living bacterium P. avenae) R bodies to unroll. Type 7 and Pt R bodies (synthesised by a symbiont and a free-living bacterium respectively) do not respond to changes in pH but can made to unroll by incubation for about 10 min at 70°C (Pond et al. 1989). They also differ in the way of uncoiling. Marinomonas R-bodies show an outer envelope, which has not been observed in other bacteria (Hernadez-Romero et al. 2003). In most cases, it has been demonstrated that the synthesis of the R-body is determined by extrachromosomal elements that are either phages or plasmids.

Many data point to the conclusion that although different classes of R-bodies exist, there is evolutionary homology among Caedibacter R-bodies. The relationship between the two Pseudomonas R-bodies and between Pseudomonas and Caedibacter R-bodies is less certain (Pond et al. 1989). Quackenbush and Burbach (1983) and Heruth et al. (1994) determined the nucleotide sequence of the genetic determinants for type 51 R-body synthesis and assembly for different strains of C. teniospiralis. Three independently transcribed genes were characterised. The R body-encoding sequences from both strains are identical. This, according to the authors, suggests that it is possible that R-bodies' structure and function requires an extremely high level of sequence conservation. It is now clear that the presence of R-bodies alone is not a phylogenetically meaningful taxonomic marker. R-body production may either be the result of convergent evolution (within different evolutionary lineages of the Proteobacteria) or of a single evolutionary event. In the latter case, this trait was passed to by horizontal gene transfer. Other structures have been described that resemble R-bodies: the cyanobacterial inclusions referred to as scroll-like membrane system (Jensen 1984) and ejectisomes, the extrusive organelles found in two flagellate families (Hausmann 1978; Kugrens et al. 1994), i.e., true cellular organelles in eukaryotic organisms.

The EA of epixenosomes (Fig. 4B) displays a far more complex ultrastructure in comparison with R-bodies. It consists of a ribbon rolled up around a granular central core (150 nm in diameter) containing a bundle of fibrils (20 nm thick and 0.8 |m long) in its upper region. The height of the ribbon decreases, both apically and distally, starting from the innermost to the outermost layer, thereby resulting in an oval shape of the EA. The matrix in which the EA is immersed can be selectively digested with proteases, which causes an unordered appearance of the EA layers. This result indicates that the matrix is proteinaceous and somehow different from the surrounding cytoplasm (Rosati et al. 1993 a). Moreover it is separated from both the cytoplasm and the central core by thin layers in which active adenylate cyclase, a typical membrane bound enzyme in both prokaryotes and eukaryotes, has been detected (Rosati et al. 1996). The chemical nature of the EA ribbon itself is still unclear. It has a specific affinity to phosphotungstic acid, a substance used to stain acidic proteins or mucoproteins in cytochemical tests. Bacteriophages or capsomere-like structures have never been observed.

Further investigations are required to determine whether the synthesis of the whole structure is related to extrachromosomal elements and to evaluate the genetic relationships with the different R-bodies and/or the eukaryotic ejectisomes.

The Basket Tubules

Bundles of regularly arranged tubules, differently oriented with respect to each other, form a basket-like structure around the EA (Fig. 4B). The inner and outer diameter of the tubules is 20-24 nm and 12-14 nm, respectively, and their wall consists of globular structures. These dimensions and structural characteristics, revealed by different preparative techniques, are very similar to that of tubulin-based microtubules. BT share additional characteristics with tubulin microtubules: they are particularly well preserved by fixation with tannic acid, a substance generally used for the maintenance of tubulin microtubules. Furthermore, they are sensible to cold temperature (4°C) and antitubulin drugs such as nocodazole, and a positive immunoreaction was obtained at the BT level with three different antibodies all specific for tubulin. The immunocytochemical analysis was carried out by both fluorescence and electron microscopy (Rosati et al. 1993b).

Tubulin microtubules occur in most eukaryotes. Their acquisition very likely represented an important step in the evolution of the eukaryotic cell by facilitating the engulfment of bacterial endosymbionts, the ancestors of chloroplasts and mitochondria. Cytoplasmic tubules and fibrous structures within the size range of tubulin tubules and tubulin protofilament have been detected in prokaryotes. Whether they are homologous to eukaryotic tubulin microtubules has yet to be established (Bermudes et al. 1994). The leading candidate for an evolutionary precursor of tubulin in the bacterial and archeal domains is the cell division protein FtsZ (Erickson 1995). However, although FtsZ is able to form tubules in vitro (Bramhill and Thompson 1994), the formation of stable tubules has never been described in vivo. Moreover, tubulins and FtsZ share only very few sequence identity (de Boer et al. 1992).

Interestingly enough, two tubulin genes, bacterial tubulin a(btuba) and bacterial tubulin b (btubb) have been detected and sequenced in bacteria of the genus Prosthecobacter (Jenkins et al. 2002). Like epixenosomes, bacteria of this genus belong to the division Verrucomicrobia, but are assigned to a different subdivision than the former (Fig. 2). Despite extensive examination of cell sections, the authors were unable to locate any microtubule-like structures in Prosthecobacter. On the basis of comparative modelling data, the Prosthecobacter tubulins are predicted to be monomeric, unlike eukaryotic a and P tubulins. The latter are known to interact and to form dimers, which then polymerise during microtubule formation. Based on their monomeric nature, the bacterial tubulins are unlikely to form microtubule-like structures. Phylogenetic analyses indicate that Prosthecobacter tubulins are quite divergent and were not recently transferred horizontally from a eukaryote (Jenkins et al. 2002).

In contrast to Prosthecobacter, epixenosomes have well-organised cytoplasmic tubules that resemble eukaryotic tubulin microtubules more than any other bacterial tubular structure described so far. This is supported by the results of an in situ hybridization when the gene encoding for P tubulin in the ciliate Euplotes crassus was used as a probe. Two clusters of gold-particles, revealing that hybridisation took place, were repeatedly found at the DZ where the epixenosome genetic material is localised. The only other structure that was decorated by this method was some chromatin in the Euplotidium macronucleus (Rosati et al. 1998). Future sequencing of the alleged tubulin genes will reveal whether epixenosomes have true tubulins and determine whether these molecules are more closely related to the divergent forms of Prosthecobacter or to eukaryotic forms. The existence of tubulin genes in epixenosomes is presently under investigation but is complicated by the fact that epixenosomes cannot easily be separated from the host and the latter, being a ciliated protozoon, possesses several tubulin genes not yet sequenced.

Was this article helpful?

+1 0

Post a comment