The gram-negative bacterium Holospora obtusa is an endonuclear symbiont of Paramecium caudatum, which is incorporated into the host cells via the food vacuoles and infects their macronucleus exclusively, but never the micronucleus. Since these two kinds of nuclei originate from a fertilization nucleus, it is assumed that the macronucleus acquires a property necessary for it to be recognized by the bacterium at a certain time during the nuclear differentiation process. We found that this property is acquired by four of the eight postzygotic nuclei as soon as the four nuclei differentiate morphologically into the macronuclear anlagen.
Ciliated protozoa contain two kinds of nuclei, which differ in ploidy, morphology, function and genetic content. The polyploid macronucleus has a high transcriptional activity (Gorovsky & Woodard, 1969; Ron & Urieli, 1977), whereas the diploid micronucleus has condensed, inactive chromatin compared to that of the macronucleus (Rao & Prescott, 1967; Pasternak, 1967) and functions as germ nucleus. Some micronuclear DNA sequences are not present in the macronucleus (Ammermann, Steinbrück, Berger & Henning, 1974; Lauth, Spear, Heumann & Prescott, 1976; McTavish & Sommerville, 1980; Yao, 1982). Despite the apparently different nature of these nuclei, however, both originate from a common fertilization nucleus. How the macro- and the micronucleus, of common genetic origin, differentiate into two morphologically and functionally different nuclei within a cell is unresolved.
In Paramecium caudatum, four species of bacterial endonuclear symbionts are known (Hafkine, 1890; Görtz, 1980) : Holospora elegans, H. undulata, H. obtusa and ms-2. The former two inhabit the micronucleus exclusively, but never the macronucleus or cytoplasm. On the other hand, the latter two inhabit the macronucleus only. The Holospora species can easily infect the specific nucleus via the food vacuoles when a homogenate of symbiont-bearing cells is added to symbiont-free cells (Ossipov & Ivakhnyuk, 1972; Ossipov, Skoblo & Rautian, 1975; Görtz & Dieckmann, 1980; Görtz, 1980). Although the cause of the nuclear specificity of these bacteria has not been clarified, it is at least clear that they possess the ability to infect g specific nucleus, either the macro-or the micronucleus. This ability of the symbionts provides an opportunity to examine the stage at which the postzygotic nuclei differentiate into macro-and micronuclei, with respect to the acquisition of a property necessary for them to be recognized and infected by the symbionts.
We demonstrate here, by using the macronucleus-specific symbiont/f. obtusa, that the macronuclear anlagen acquire the property needed by the macronucleus for an infection with H. obtusa as soon as the postzygotic nuclei differentiate morphologically into the macronuclear anlagen. The result suggests that this property is acquired when macronuclear differentiation begins.
MATERIALS AND METHODS
Strains and culture conditions
The cells used in this study were Paramecium caudatum syngen 3, mating type V, strain 27aG3, and mating type VI, strain 27aG3BC8-l. These strains were kindly supplied by Dr Y. Tsukii, Hosei University. The original Holospora ofitusa-bearing strain C101 (syngen unknown) was collected in Münster, FRG. Later, the symbionts infected strain 27aG3 and this newly infected strain was used for obtaining the symbionts in this study. The culture medium used was 1· 25% (w/v) fresh lettuce juice in Dryl’s solution (Dryl, 1959) inoculated with a non-pathogenic strain of Klebsiella pneumoniae 1 day before use (Hiwatashi, 1968). In ordinary cultures several hundred cells were inoculated into 2 ml culture medium and then 4 ml, 10 ml and 10 ml of fresh medium were added on successive days. Cultures were kept at 25 °C. One or two days after the final feeding, mating reactivity increased to its maximum intensity.
Holospora species show two morphologically distinct forms in their life cycle (Ossipov & Ivakhnyuk, 1972; Ossipov et al. 1975; Görtz & Dieckmann, 1980): a reproductive short form and an infectious long form. The former is observed predominantly in nuclei of vegetatively growing host cells and the latter in that of starved cells. Only the long form is infectious. Infection of macronuclei or macronuclear anlagen of P. caudatum with H. obtusa was achieved as follows. Cultures of symbiont-bearing cells in stationary phase were strained through four layers of fine gauze to remove gross debris. Then the cells were harvested by centrifuging for 3 min at 1000 rev./min, and homogenized by hand in a Teflon homogenizer at 0—4°C. The density of the symbionts in the homogenate was counted with a blood-counting chamber and adjusted to 4 × 105 symbionts/ml by adding Dryl’s solution. Recipient cells and the homogenate containing//, obtusa were mixed in depression slides at 1500cells/ml and 4 × 104 symbionts/ml, respectively, at 25 °C.
Cytological methods for observation of nuclei and symbionts
Cells infected with H, obtusa were harvested by centrifugation in a hand-operated centrifuge, airdried on glass slides, and fixed in a mixture of acetic acid and ethanol (1:3, v/v) for 10 min at room temperature. Preparations were stained by the Feulgen reaction and counterstained with 0 · 25% (w/v) fast green FCF. Observations were made using a differential interference-contrast microscope at a magnification of × 1000. Temporal observation of the symbionts in host macronuclei was made as follows. Cells were pipetted with a few microlitres of culture medium onto glass slides and fixed in 4% (w/v) OsO4 vapour by inverting the slide over a small vial of the fixative for 3 · 4 s. The cells were then observed either unstained, or stained with aceto-orcein.
Observation of food vacuoles
Cell suspensions were mixed with a drop of Indian ink solution for. 15 min at a density of about 1500 cells/ml at 25 °C. Then the cells were fixed and stained by the Feulgen reaction and with fast green FCF as described before. The cells able to form food vacuoles are expected to ingest Indian ink and form many black-coloured food vacuoles.
Time-course of infection with H. obtusa
In order to determine the time needed for infection of the macronucleus of P. caudatum with H. obtusa, symbiont-free cells were mixed with the homogenate of symbiont-bearing cells at 25°C (see Materials and Methods). Then the cells were fixed every 10 min for the first 2 h after mixing, and stained as described. The results are shown in Fig. 1. Ten minutes after mixing the cells with the symbiont-containing homogenate, the macronucleus was infected with H, obtusa in about 20% of the cells. The symbionts were also observed in the cytoplasm, some were in food vacuoles and some apparently outside the food vacuoles. The proportion of cells bearing symbionts in their macronuclei rose to about 60% at 30 min after mixing and to about 100% at 60 min. Mean numbers of the symbionts observed in the individual macronuclei at 10, 30 and 60 min after mixing were about 1, 2 and 3, respectively. The symbionts were never observed in the micronucleus. The data show that the infection of the macronucleus with H. obtusa begins remarkably quickly after mixing. Since the symbionts can be observed not only near the macronucleus but also near the micronucleus, it is evident that the symbionts can infect the macronucleus but not the micronucleus. Similar observations were reported by Ossipov and colleagues (see Ossipov, 1981, for a review).
Infection of macronuclear anlagen of exconjugant cells with H. obtusa
As shown in previous studies (Calkins & Cull, 1907; Klitke, 1916; Saito & Sato, 1961; Mikami & Hiwatashi, 1975; Mikami, 1980), in P. caudatum the synkaryon (fertilization nucleus) normally divides three times successively; four of the resultant nuclei become macronuclear anlagen, one becomes a micronucleus, and the remaining three degenerate. The old macronucleus is transformed into a loosely wound skein (skein formation) at about the time of the third (last) postzygotic division and subsequently breaks into many fragments (macronuclear fragmentation). Recently, it has been found by Mikami (1980) that the determination of the macro-and the micronuclear differentiation occurs immediately after the third nuclear division and is closely related to a brief localization of the nuclei at the opposite ends of the cell, with the prospective macronuclei in the posterior region and the prospective micronuclei in the anterior region. Similar results were also reported in P. tetraurelia (formerly called syngen 4 of P. aurelia, Sonneborn, 1975) (Grandchamp & Beisson, 1981). Hereafter, this stage will be tentatively called the ‘determination stage for nuclear differentiation’. The earliest morphological change of the nuclei after the determination stage is the appearance of heterochromatic aggregates in the four macronuclear anlagen, which is followed by increased stainability with fast green. The heterochromatic aggregates then disintegrate and disappear, accompanied by an increase in the volume of the anlagen (Saito & Sato, 1961).
The high degree of macronuclear specificity of H. obtusa and its rapid infectivity to the macronucleus make it possible to use the symbiont to ask the following question: when do the division products of the synkaryon differentiate into macronuclei with respect to the infection with H, obtusai To examine this problem, 15, 24 and 30 h after the beginning of the mating reaction at 25 °C, cells were mixed with the homogenate of H. obtus a-bearing cells at densities of about 1500cells/ml and 4 × 104symbionts/ml, respectively. Conjugating pairs usually separate about 15 h after the beginning of the mating reaction at 25 °C. The degree of synchrony of the nuclear changes during the conjugation process is relatively high, as the cells enter into the process almost simultaneously during the mating reaction, but it is not absolute. Therefore, in cells harvested at 15, 24 and 30 h after the beginning of the mating reaction, a series of stages from late conjugation to early postconjugation can be observed (Saito & Sato, 1961 ; Mikami & Hiwatashi, 1975), i.e. the third pregamic division, pronuclear exchange between the two mates, synkaryon, three postzygotic divisions, determination stage for nuclear differentiation, eight postzygotic nuclei after the determination stage (post-determination stage) and development of macronuclear anlagen. Two hours after mixing at 25 °C, the cells were fixed and stained as described before.
Uptake of the symbionts into the cells was observed when all eight postzygotic nuclei had migrated from the opposite ends of the cell into a common region, after the determination stage for nuclear differentiation (post-determination stage), which is the stage just before the first visible differentiation of the macronuclear anlagen. Some of the symbionts were observed in food vacuoles and some were apparently outside the food vacuoles, in the host cytoplasm. However, symbionts were not observed in paramecia at earlier stages, i.e. the stages of synkaryon, two or four postzygotic nuclei and the determination stage for nuclear differentiation. After the appearance of the first symbionts within the cells, it was confirmed that none of the postzygotic nuclei immediately before their first visible differentiation were infected with the symbionts. Then, almost simultaneously with the first visible differentiation of the macronuclear anlagen all of them were infected with the symbionts (Fig. 2A, B). Some of the old macronuclear fragments were also infected, but the new micronucleus, which derived from the synkaryon, was not infected. Notwithstanding that the diameter of the first stage of the macronuclear anlagen was shorter than the length of H. obtusa, the symbionts could infect each of the four anlagen. Because of their length, the symbionts appeared to protrude out of the anlagen (Fig. 2A). It should be emphasized that the postzygotic nuclei immediately before the first visible differentiation of the macronuclear anlagen were not infected with the symbionts, despite the presence of symbionts in the cytoplasm.
These results indicate that the macronuclear anlagen acquire the property of the macronucleus that is needed for recognition by and infection with//, obtusa at almost the same time as the first recognizable change in the macronuclear anlagen. Therefore, this strongly suggests that this property is acquired in close association with the initiation of structural or functional differentiation in the macronucleus.
Capacity for food vacuole formation in exconjugant cells
Uptake of the symbionts was not observed in early exconjugant cells, i.e. at the stages of synkaryon, two or four postzygotic nuclei and determination of nuclear differentiation. In order to examine why the uptake of the symbionts did not occur in these cells, their capacity for food vacuole formation was examined. The cells at 15, 24 and 30 h after mixing of complementary mating types at 25 °C were suspended in Indian ink solution for 15 min at a density of about 1500cells/ml at 25 °C, then fixed and stained (see Materials and Methods). Fifteen minutes of incubation with Indian ink solution is sufficient for ingestion of ink and formation of black food vacuoles (see also Takagi et al. 1981). Therefore, the capacity for food vacuole formation at different stages in exconjugant cells can be determined. As mentioned before, by harvesting the cells at three different times (15, 24 and 30 h) after the mixing of complementary mating types, cells at various stages from the third prezygotic division to the development of macronuclear anlagen were obtained. Furthermore, since the incubation time is only 15 min, the nuclear stages of the cells do not change much during the incubation. Therefore, the ability to form food vacuoles at each of the stages from the late conjugation process to the early post-conjugation process can be determined. As a control experiment, cells in stationary phase were incubated in Indian ink solution in the same way. The results are summarized in Table 1. At the synkaryon stage, the cells are still paired at the paroral regions. The pairs usually separate at the stage of two postzygotic nuclei. However, as shown in the table, the exconjugant cells could not form food vacuoles up to the determination stage for nuclear differentiation. Initiation of food vacuole formation was observed after the determination stage. These results agree well with the finding that the first symbiont appeared in exconjugant cells at the post-determination stage, suggesting that the inability of the symbionts to infect early exconjugant stages is due to the inability of the cells to form food vacuoles.
In Paramecium to date, five. Holospora species are known as endonuclear symbionts (Hafkine, 1890; Preer, 1969; Ossipov, Borchsenius & Podlipaev, 1980; Preer & Preer, 1982) : H. obtusa, H. elegans and H. undulata in P. caudatum, H. caryophila (formerly called alpha; Preer, 1969) in P. biaurelia (formerly called syngen 2 of P. aurelia;Sonneborn, 1975), and H. acuminata inP. bursaria. All of these symbionts show a nuclear specificity in their infectivity. In the present study, the uptake of H. obtusa into the cell started at the stage when the cell could form food vacuoles (Table 1). The result strongly suggests that the uptake is mainly by way of the food vacuoles, which agrees with other cytological observations (see Görtz, 1983, for a review).
An electron-microscopic observation of the infection process of P. caudatum with H. obtusa has been reported by Ossipov & Podlipaev (1977). However, many details of this process (i.e. protection against the digestive enzymes in the food vacuoles, migration from the food vacuoles to the macronucleus and nuclear specificity of the infection) are not yet understood. It has been reported that Holospora species infect the host nucleus within 1 – 2 h of the addition of the symbionts to the external medium of symbiont-free cells (Preer, 1969; Ossipov & Podlipaev, 1977; Görtz & Dieckmann, 1980). However, we found that the actual time needed for the appearance of the first symbiont in the macronucleus is less than 10 min (Fig. 1). This rapid infectivity of H. obtusa indicates that it can escape from food vacuoles even earlier. This may protect the symbionts from the digestive enzymes in the food vacuoles, because lysosomal fusion to the newly formed food vacuoles begins 8 min after vacuole formation (Fok, Lee & Allen, 1982). As mentioned before, all Holospora species exhibit nuclear specificity in their habitat. Although the mechanism by which the nuclear specificity is controlled remains unknown, it can be assumed that these symbionts are able to distinguish between the macro-and the micronucleus in some way. Even when the macronucleus is well-infected by H. obtusa, the symbionts are never observed inside the micronucleus, despite the presence of symbionts nearby. Since, as far as we have observed, H. obtusa is unable to pass through the nuclear membrane of the micronucleus, it is suggested that the symbiont may have the ability to distinguish between the nuclear membranes of the two kinds of nuclei, or something associated with them.
Besides the differences in morphology, DNA content and transcriptional activity, several differences have been found between the macro-and micronuclei of ciliates: ribosomal RNA gene amplification (Gall, 1974; Yao, Blackburn & Gall, 1979; Yao, 1981) and lack of some micronuclear DNA sequences in the macronucleus (Ammer-mann et al. 1974; Lauth et al. 1976; McTavish & Sommerville, 1980; Yao, 1982), and the presence of a micronuclear-specific histone (Allis, Glover & Gorovsky, 1979). However, the stage at which these differences develop during the differentiation processes of the two nuclei is not known. In the present study, we determined the stage at which the nuclei acquired the quality needed for infection with H. obtusa ; H. obtusa could infect the macronuclear anlagen as soon as the anlagen become morphologically distinguishable from other postzygotic nuclei. The symbionts could not infect any of the postzygotic nuclei before the occurrence of the first morphological change, or the new micronucleus. The results clearly indicate that the property needed for infection with H. obtusa is acquired at the first stage in the visible differentiation of macronuclear anlagen. As far as we know, the appearance of this property is the earliest of the known characteristics of the macronucleus. This property is assumed to be intimately correlated with a fundamental characteristic of the macronucleus, because it is present from the beginning of macronuclear differentiation to macronuclear degradation (macronuclear fragmentation).
By using micronuclear-specific symbionts, H. elegans or H. undulata, it should be possible to discover when one of the postzygotic nuclei differentiates into the micronucleus, as recognizable by the symbionts. These methods will provide us with a new approach to analysing the timing and the mechanism of macro-and micronuclear differentiation in genetically identical nuclei.
The earlier part of this study was supported by a fellowship from the Alexander von Humboldt-Foundation to M. F., and was carried out while the author was a guest in the laboratory of Dr Klaus Heckmann, Zoologisches Institut der Università!, Münster, FRG.