Holospora recta is a micronucleus-specific symbiotic bacterium of the ciliate Paramecium caudatum. This bacterium cannot grow outside the host cell. We isolated the infectious form of this bacterium from the host homogenates by 70% Percoll density gradient centrifugation. DNA in the infectious form of H. recta appeared as a large cluster in the cytoplasmic region near the large periplasm. This was not observed with the infectious form of a macronucleus-specific symbiont of P. caudatum, H. obtusa. The isolated infectious form could infect the micronucleus of P. caudatum and differentiated into the reproductive form within 18 hr after the infection at 25°C. However, not only was this bacterium unable to infect the micronucleus of P. bursaria, but it was also unable to infect the micronucleus of P. multimicronucleatum or P. novaurelia, even though the latter two species are morphologically closely related to P. caudatum. We succeeded in the cryopreservation of this bacterium, as cells stored at −85°C for 127 days maintained their infectivity and reproducibility.
Holospora species are Gram-negative symbiotic bacteria of ciliates of the genus Paramecium. All of the nine Holospora species that have been described (Borchsenius et al., 1983; Fokin, 1989, 1991; Fokin and Sabaneyeva, 1993; Hafkine, 1890; Ossipov et al., 1975, 1980; Preer, 1969) show species-specificity and nucleus-specificity in their habitats. For example, H. obtusa is the macronucleus-specific, but H. recta, H. elegans and H. undulata are the micronucleus-specific bacteria of P. caudatum (Hafkine, 1890). H. caryophila is a macronucleus-specific bacterium of both P. biaurelia (Preer, 1969) and P. novaurelia (Fokin et al., 1996). H. acuminata (Ossipov et al., 1980) is a micronucleus-specific and H. curviuscula (Borchsenius et al., 1983) is a macronucleus-specific bacterium of P. bursaria. H. bacillata (Fokin, 1989; Fokin and Sabaneyeva, 1993) is a macronucleus-specific of P. woodruffi and P. calkinsi, and H. curvata (Fokin and Sabaneyeva, 1993) is a macronucleus-specific bacterium of P. calkinsi. Although they can distinguish Paramecium species and two different nuclei, control mechanisms of species-specific and nucleus-specific infection is not known. H. obtusa infects the macronucleus of H. elegans-bearing cell and H. elegans also infects the micronucleus of H. obtusa-bearing cell, but always a former occupant can remain in the host cell (Fujishima and Görtz, unpublished data), notwithstanding that each Holospora is present in different nuclei in the cell. On the other hand, H. elegans-bearing cell could not produce their offspring even if the bacteria were removed by treatment with antibiotics (Görtz and Fujishima, 1983), and usually Holospora-bearer showed decreased fission ratio, and paramecia do not need holosporas for their growth. Therefore, Holospora species are considered as parasitic bacteria. However, regardless of such harmful effects of the holosporas, they are easily found in paramecia living in relatively cold area such as Northern Europe and the Kamchatka peninsula. Therefore, their existence may give some merits for their host survival in cold temperature.
Holospora species have two distinct forms in their life cycle: a reproductive short form and an infectious long form. When the host cell grows, the reproductive form also grows by binary fission in the host nucleus. When the host ceases binary fission as a result of starvation, the reproductive form ceases binary fission and differentiates into the infectious form. The infectious form is unable to grow by binary fission in the host nucleus, but is able to escape from the host cell (Wiemann, 1989) invades the cytoplasm of a new host through the host digestive vacuole and then enter the target nucleus with its special tip (Görtz and Wiemann, 1989).
Because Holospora species cannot grow outside the host cell, they have to be isolated from mass-cultured host cells when large numbers of cells are needed for biochemical experiments. Furthermore, although infection can be induced by mixing Holospora-free paramecia with Holospora-bearing Paramecium homogenates (Fokin and Sabaneyeva, 1993; Fujishima and Fujita, 1985; Görtz and Dieckmann, 1980; Ossipov et al., 1975, 1993), reproducible results with respect to the timing of the infection and infection ratio are difficult to obtain because of the harmful effect of Paramecium debris in the homogenates. Therefore, isolated infectious forms instead of the host homogenates are needed, especially for studies on early events in the infection process and on recognition mechanism of two kinds of the host nuclei by Holospora species.
Holosporas were first isolated from host homogenates by Preer (1969), who used an Ecteola column to isolate the macronucleus-specific bacterium H. caryophila from P. biaurelia. Later, H. obtusa has been isolated by Fujishima and Nagahara (1985) and H. elegans has been isolated by Schmidt et al. (1987), using Percoll density gradient centrifugation.
In the present study, we attempted to isolate the infectious form of H. recta in a way that maintained its infectivity, and to compare it with H. obtusa with respect to its morphology and timing of differentiation from the infectious form to the reproductive form. We also examined whether the infectious form of H. recta can infect other Paramecium species that are closely related to P. caudatum, as previously shown with H. obtusa (Fujishima and Fujita, 1985). Furthermore, in order to maintain this symbiotic bacterium in the laboratory without culturing, we attempted cryopreservation of the isolated infectious form of H. recta.
MATERIALS AND METHODS
Cell and culture conditions
P20-16, a Holospora recta-bearing strain of P. caudatum (mating type and syngen unknown) and RB-1, a strain of P. caudatum that does not have H. recta (mating type E, syngen unknown) were cultivated in modified lettuce juice medium at 25°C, as described in our previous paper (Fujishima et al., 1990a) except that KH2P04 was used instead of NaH2P04.2H20 in the medium. Other Paramecium strains that were used included the P. multimicronucleatum strain 49B (syngen 2, mating type IV) and strain 53B (syngen 2, mating type III), the H. caryophila-bearmg P. novaurelia strain 91YB1-3 and the H. curviuscula-bearig P. bursaria strain ET50-16. These strains were kindly provided by Dr. D. V. Ossipov, Dr.I.I.Skoblo and Dr. M. Rautian, St. Petersburg State University, Russia.
Isolation of H. recta
Isolation of the infectious form of H. recta was carried out as described previously (Fujishima et al., 1990a, b) with a slight modification. Briefly, 12 liters of medium containing H. recta-bearing paramecia in the stationary phase of growth were harvested by flow-through centrifugation at 700 x g for 30 min. The cells were then washed twice with Dryl's solution (Dryl, 1959) by centrifuging at 800 x g for 3 min at 4°C, washed once with 0.25 M sucrose by the same centrifugation and homogenized in a Teflon homogenizer. The homogenates were centrifuged in 70% (v/v) Percoll (Pharmacia LKB Biotechnology) at 37,500 x g (instead of 38,900 x g) for 30 min at 4°C. A bacterial band denser than 1.13 g/ml was collected and washed twice with 10 mM Na, K-Phosphate buffer (Na, K-PB), pH 6.5, by centrifuging at 2,000 x g for 10 min at 4°C. Densities were visualized with a density marker beads kit (Pharmacia LKB Biotechnology).
Isolated cells of the infectious form were mixed with Paramecium cells at densities of about 1 × 105 bacteria/ml and 1 × 103 paramecia/ ml in a 1-ml depression slide at 25°C. The paramecia were observed under a Nomarski differential-interference-contrast (DIG) microscope 1 hr and 24 hr after mixing.
Cryopreservation of isolated H.recta
Isolated cells of the infectious form of H. recta were suspended in 10 mM Na, K-PB, containing 10% (v/v) glycerol in 1.5-ml serum tubes and immediately frozen at −85°C. Three, 10 and 127 days after freezing, the serum tubes were rapidly warmed with shaking in a water bath at 37°C. As soon as the ice was completely melted, the bacteria were washed once with 10 mM Na, K-PB, by centrifuging at 200 x g for 5 min at 4°C and suspended in the same buffer at 4°C until use.
H. recta in the host micronucleus and those pushed out from the host cells were observed under a Nomarski DIC, phase-contrast and fluorescence optics (Olympus BH2-RFL) at a magnification of × 400 or × 1,000. To observe the infectious form of H. recta, cells isolated from the host homogenates by Percoll density centrifugation were air-dried on a glass slide, stained with 1 μg/ml of the DNA-specific fluorochrome, 4′,6-diamidino-2-phenylindole (DAPI) dissolved in deionized water, and covered with a cover glass.
Morphology of the isolated infectious form of H. recta
The diameter of a micronucleus of P. caudatum is about 5-6 μm. However, when H. recta grew in the micronucleus, the nucleus became swollen (Fig. 1A). In some cases, the nuclear diameter reached more than 30 μm. When a cell was crushed between a slide and a cover glass, many holosporas in the micronucleus were pushed out from the cell (Fig. 1B). As shown in Fig. 1B, H. recta showed different morphologies in the life cycle as seen previously in H. obtusa and H. elegans (Fujishima et al., 1990a; Görtz and Dieckmann, 1980); a reproductive short form (about 2.6 μm in length), an infectious long form (about 10 μm in length) and cells of intermediate lengths in the process of differentiation from the reproductive form to the infectious form, A division constriction was found only in the short form, indicating that reproduction by binary fission occurred only in this form, in agreement with the cases of H. obtusa and H. elegans (Fujishima et al., 1990a; Görtz and Dieckmann, 1980). The fully elongated long form of this bacterium bent a little, and had a rounded end and a tapered end as described previously (Fokin, 1991).
The host homogenates were then centrifuged in 70% Percoll. Cells of the infectious form formed a band near the bottom of the centrifuge tube, between the density marker beads of 1.12 g/ml and 1.13 g/ml. On the other hand, the reproductive and intermediate forms formed another band near the top of the Percoll gradient with the host debris. This showed that the bacterium increases its density during differentiation from the reproductive form to the infectious form. Cells of the infectious form were collected, washed with 10 mM Na, K-PB, and stored at 4°C until use. As shown in Fig. 2A, the isolated infectious form of H. recta consisted of three distinct parts under a phase-contrast microscope: a dark part with a tapered end, a refractive part and a small dark tip at the rounded end as previously seen in H. obtusa and H. elegans (Fujishima and Hoshide, 1988; Görtz and Dieckmann, 1980). In H. obtusa and H. elegans, it was found by transmission electron microscopy that the dark and refractive parts were the cytoplasmic and periplasmic regions, respectively (Wiemann, 1989). On the other hand, the small dark tip was composed of homogeneous materials with low electron densities, and it is known that this special tip is used to infect the host nucleus (Fujishima and Fujita, 1985; Görtz and Dieckmann, 1980). Therefore, the special tip seems to distinguish the two kinds of nuclei in the host and to invade the target nucleus. Furthermore, it was found that this heavy bacterial band did not contain fully elongated intermediate forms that had not yet developed the three distinct intracellular structures. Such the forms were observed in a less dense band together with the reproductive form and the host debris (data not shown). These results suggested that H. recta elongates to the length of the infectious form while it has the lower density of the reproductive form, and then increases its density as it begins to develop the three intracellular structures.
When isolated cells of the infectious form were stained with DAPI, the fluorescence appeared as a large cluster in the bacterium (Fig. 2C). This indicated that DNA is localized at a specific region of the infectious form. As shown in Fig. 2B-D, unlike the infectious forms of H. elegans (Görtz and Dieckmann, 1980) and H. obtusa (Fujishima and Hoshide, 1988; Fujishima et al., 1990a, 1991) the DAPI-positive region of H. recta was apparently seen in the cytoplasmic region (dark region under a phase-contrast microscope) adjacent to the periplasm (refractive region under a phase-contrast microscope).
Infectivity and reproducibility of the isolated infectious form of H. recta
When isolated cells of the infectious form were added to paramecia at 25°C, they soon were engulfed by the host digestive vacuoles, and some of them escaped into the cytoplasm and invaded the host micronucleus. The first bacterium appeared in the micronucleus within 30 min. We could not observe infection of the macronucleus by this bacterium, indicating that the bacterium can distinguish the two kinds of host nuclei and infects only the micronucleus. It was found that the infected bacteria formed constrictions to differentiate into the reproductive short form in the Paramecium micronucleus within 18 hr after the infection.
Infectivity of H. recta to other Paramecium species
We examined whether the isolated infectious form of H. recta also can infect other Paramecium species other than P. caudatum. The bacteria were mixed with cells of P. multimicronucleatum strains 49B and 53B, P. bursaria strain ET50-16 and P. novaurelia strain 91YB1-3, and the Paramecium cells were observed 1 hr and 24 hr after mixing under a Nomarski DIC microscope. H. recta were unable to infect these Paramecium nuclei in two trials for each species.
Cryopreservation of the isolated infectious form of H. recta
If cryopreservation of the isolated infectious form of H. recta is practicable, maintenance of this bacteria would be easy. Previously, we succeeded in the cryopreservation of the infectious form of H. obtusa at −85°C (Fujishima et al., 1991). Using the same method, isolated H. recta cells of the infectious form were frozen in 10 mM Na, K-PB, containing 10% glycerol at −85°C. Three, 10 and 127 days after freezing, the bacteria were thawed, washed and mixed with paramecia to examine their infectivity and reproducibility after infection. In three cases, the bacteria were soon engulfed by the host digestive vacuoles after mixing. Some of them then escaped from the vacuoles to appear in the host cytoplasm, and infected the micronucleus within 1 hr after mixing just as did the non-frozen control bacteria. Within 24 hr after mixing, infection was observed in the majority of the Paramecium cells as it did in the control. We also confirmed that the bacteria that infected the host micronucleus differentiated into the reproductive form within 18 hr after the infection.
In a 70% Percoll density gradient, the infectious form of H. recta formed a band between the densities of 1.12-1.13 g/ ml. In case of the infectious form of H. obtusa, its density was 1.16 g/ml (Fujishima et al., 1990a). Thus, it was found that the density of the infectious form of H. recta was lighter than that of H. obtusa. In the present study, we did not isolate the reproductive form of this bacterium. However, as we succeeded in doing this with H. obtusa (Fujishima et al., 1990a) it seems possible to isolate the reproductive form if micronuclei bearing the reproductive form are isolated first and then the nuclear homogenates are centrifuged in 40% Percoll.
The DAPI-positive region in the infectious form differs between the Holospora species examined. In H. elegans, the DAPI-positive region exactly corresponds to a dark region (cytoplasmic region) under a phase-contrast microscope (Görtz and Dieckmann, 1980). However, in H. obtusa, the DAPI-positive region coincides with a refractive region (periplasmic region) (Fujishima, 1986; Fujishima and Hoshide, 1988; Fujishima et al., 1991). Because the refractive region corresponds to the periplasmic region, and because a narrow cytoplasmic extrusion had been observed in the periplasmic region of H. obtusa under an electron microscope (Görtz et al., 1992), it seems that localization of DNA in this narrow cytoplasmic extrusion makes the refractive region DAPI-positive. In this study, we found that H. recta possesses a DAPI-staining pattern intermediate between those of H. elegans and H. obtusa. Namely, only a limited cytoplasmic region closely adjacent to the large periplasm showed DAPI-fluorescence. Furthermore, although H. obtusa had shown two spots with strong DAPI-fluorescence in the DAPI-positive region (Fujishima and Hoshide, 1988; Fujishima et al., 1990a, 1991) as did H. elegans (Görtz and Dieckmann, 1980), H. recta did not show such spots. These observations indicate that H. recta differs from H. obtusa and H. elegans not only in morphology, but also in the localization of DNA in the infectious forms.
We also found that the timing of the formation of a constriction in the infectious form of H. recta differs from that of H. obtusa. H. recta formed the constrictions to differentiate into the reproductive form within 18 hr at 25°C, while H. obtusa did it between 32-34 hr after infection (unpublished data). Although the timing of the constriction formation in the early infection process has not yet been examined in other Holospora species, our data show that H. recta begins to form it within about half the time needed in H. obtusa.
Paramecium species can be classified morphologically into two groups: an “aurelia group” and a “bursaria group”. The macronucleus-specific H. obtusa infects the macronucleus of not only P. caudatum, but also 14 species of the P. aurelia species complex and P. multimicronucleatum, although maintenance of the infected bacteria is achieved only in specific strains of P. caudatum (Fujishima and Fujita, 1985; Fujishima, 1986). All these Paramecium species belong to the “aurelia group”. The Paramecium species belonging to the bursaria group (P. bursaria, P. trichium, P. calkinsi and P. woodruffi) were not infected by H. obtusa (Fujishima and Fujita, 1985; Fujishima, 1986). Therefore, it seemed that a property of the macronucleus, necessary for it to be recognized and infected by H. obtusa, was present not only in P. caudatum but also in closely related Paramecium species belonging to the “aurelia group”. H. recta, like H. obtusa, could not infect P. bursaria, but unlike H. obtusa, it also could not infect P. multimicronucleatum or P. novaurelia. This indicates that the species-specificity of the infectivity of H. recta is more strictly controlled than that of H. obtusa. This suggests that although macronuclear envelope specific substances distinguished by H. obtusa may be present in various species belonging to the “aurelia group”, but micronuclear envelope specific substances distinguished by H. recta may be present only in P. caudatum or in very limited species if any. On the other hand, infectious forms of Holospora species have a structurally differentiated special tip at the one end (Fujishima and Hoshide, 1988; Fujishima et al., 1990b) and the bacteria always penetrate the host target nuclear envelope with this special tip first. This suggests that substances at this tip have an ability to distinguish some differences between two kinds of the host nuclei and to penetrate the nuclear envelope. Although the substances supposed to be present at the special tip are not yet detected, it may be possible if the isolated infectious forms are used as antigens for obtaining of the monoclonal antibodies against the special tip.
In our previous study (Fujishima et al., 1991), we succeeded in the cryopreservation of the isolated infectious long form of H. obtusa at −85°C without losing their infectivity or reproducibility. Using the same method in the present study, we also succeeded in the cryopreservation of H. recta. Although it is not clear how many days the bacterium frozen at −85°C can maintain its infectivity and reproducibility, this duration may be prolonged with the use of liquid nitrogen. Cryopreservation made it possible preserve this bacterium for a long time without the need for laborious culture procedures.
This study was supported by grants to M. Fujishima from the Japan Ministry of Education, Science, Sports and Culture (Grant-inAid for Scientific Research #05454028 and International Scientific Research Program #07044202).