Characterization of a green Stentor with symbiotic algae growing in an extremely oligotrophic environment and storing large amounts of starch granules in its cytoplasm

Distribution of Stentor pyriformis in Japan and its optimal culture conditions

S. pyriformis was described by Johnson in 1893⁶. This algae-bearing Stentor has separated spherical macronuclei without pigmentation, which certainly differentiates it from other Stentor species (see Table S1, Fig. 5B). While the most common algae-bearing Stentor, S. polymorphus assumes a slender trumpet shape (often shortened), S. pyriformis never resembles such a slender trumpet, but assumes a pear or short conical shape, even when it is swimming⁶. Presence or absence of colored pigmentation is also a prominent characteristic for separating Stentor species. Among algae-bearing Stentor spp., S. polymorphus and S. pyriformis only are considered colorless species, whereas colored species are S. amethystinus, S. fuliginosus, S. araucanus, and S. tartari⁸ (Table S1). Therefore, S. pyriformis is a clearly discernible species; however, it remains underexplored. Indeed, we could only find one paper on the new habitats of S. pyriformis⁷, with the exception of the paper of species consolidation of this genus⁸. We confirmed the presence of S. pyriformis at 23 locations (Fig. 1A). This indicates that S. pyriformis is by no means a rare organism. We assume one of the reasons why S. pyriformis has been poorly studied is the difficulty of cultivation. In fact, Johnson⁶ noted that he could not keep them more than a month and never observed any cells in fission. In addition, after five years of failure, it was finally possible to culture S. pyriformis for more than several months. Because of objectively unfounded data that we could not include in the distribution data (Fig. 1A), we noticed the wetlands where we found S. pyriformis were limited to small ponds or bogs locating near the mountain peak or along the ridge (Fig. 1B). That is, the ponds depending on rainfall without inflowing rivers. Because there is no nutrient flowing in, waters in these ponds showed noticeable oligotrophic tendency, i.e., extremely low electric conductivity (Fig. 1A), which gave us some clues on culture.

The most important point of culture for S. pyriformis was keeping the medium lower electric conductivity. We use the KCM medium diluted by 2% with Milli-Q water, and changed medium once a week. A non-photosynthetic cryptophyte, Chilomonas paramecium was selected for food. We selected the food so that it would not itself grow in the culture medium. Growing organisms, like photosynthetic algae, seemed to cause damage to S. pyriformis. Using this culture method, S. pyriformis can be maintained for more than four years (see Table 1). For the organisms not easy to grow in culture, Professor Michael Melkonian mentioned no protist is ‘uncultivable’, there is just human failure³⁰. Here, it just became possible to culture S. pyriformis 120 years after its discovery; however, this method does not always work. S. pyriformis appears to be extremely fragile and disintegrates when any variables are unintentionally altered, that is, the culture is still unstable. When its condition deteriorates, the cells divide unevenly in such a way that a part of the cell is broken. When this happens, the cells become spherical, and the drug drops to the bottom of the dish. It retains this shape for more than a month, but eventually disappears. The doubling time of S. pyriformis remains 3 to 4 weeks, even under favorable conditions (data not shown). We occasionally encountered the blooming of S. pyriformis all over the bottom of the ponds (Fig. 1C). S. pyriformis, therefore, does not seem to be a particularly slow growing species, but our culture method appears to be far from the optimal culture conditions for them. Three S. pyriformis strains used in this study are available from the authors upon request.

Ultrastructure

In this study, we compared the conventional chemical fixation method with the rapid-freezing fixation method for electron microscopic observation. As a result, large vacuoles were observed in the cytoplasm when chemical fixation was used, but not by rapid freezing. Instead, many multi-vesicular bodies were observed in the cytoplasm. The quick-freezing and freeze-substitution method is considered superior in that it can prevent deformation of the intracellular structure compared to chemical fixation³¹. Therefore, it is possible that the originally existing multi-vesicular bodies were artificially disintegrated by chemical fixation, and the constituent biological membranes fused together, eventually forming large vacuoles. To the authors’ knowledge, no intracellular structure similar to the multi-vesicular body in S. pyriformis has been reported in protists. As multi-vesicular bodies of S. pyriformis could only be observed using the freeze-substitution method, similar granules may also be found in other protists if the same technique is used for electron microscopy. In animals, on the other hand, aggregates of secretory vesicles resembling the multi-vesicular bodies of S. pyriformis are present in cardiac telocytes³². The extracellular vesicles form multi-vesicular structures of about 1 μm in diameter and contain materials for intercellular communication that are involved in cardiac physiology and regeneration. Because S. pyriformis cells often form aggregates at the bottom of the pond, some chemicals may be released from the multi-vesicular body, attracting nearby cells and forming aggregates.

Observation by the freeze-substitution method revealed that the symbiosome membrane was in close contact with the symbiotic chlorella. Furthermore, fluffy projections were observed on the cell wall of the symbiotic chlorella. These characteristics were consistent with those of C. variabilis, which is symbiotic in the cells of P. bursaria⁹. The only difference was that in S. pyriformis, the symbiotic chlorella cells were scattered in the cytoplasm, whereas the symbiotic Chlorella in P. bursaria were anchored directly below the cell surface.

Storage granules

The iodine in Lugol’s solution selectively binds to α-1, 4-linked glucose found in polysaccharides, such as starch³³ and glycogen³⁴. The color stained with Lugol’s solution reflects the type of glucose polymer. Starches with high amylose content stain blue-violet (cf. Fig. 4B), high amylopectin stains red–purple, and glycogen stains reddish brown (Table 2). The granules in the cytoplasm of S. pyriformis stained reddish brown with Lugol’s solution (Fig. 4A), suggesting that these granules are composed of α-1,4-linked glucans with high number of α-1,6-linked branch points, either amylopectin-rich starch or glycogen. The pyrenoid of Chlorella spp. is surrounded by a starch sheath of two large plates³⁵. As shown in Fig. 4F,G, the image contrast formed by electron staining of the starch granule in the chloroplast (arrow) was lost by treatment with Lugol’s solution. Although the detailed mechanism is unknown, this observation suggests that electron-stained heavy metals (osmium, lead, and lanthanoid ions) bound to the granules may have been eliminated by iodine in Lugol’s solution. The cytoplasmic granules of S. pyriformis showed the same staining properties as the starch granules in the chloroplasts of symbiotic chlorella, suggesting that both types of granules share chemical characteristics as polysaccharides.

Alveolates make up one of the most diverse and largest groups of protists. They include three major taxa: dinoflagellates, ciliates, and apicomplexan protozoa. All three alveolate lineages store glucose in an α-1,4-linked glucose chain with α-1, 6 branches. Ciliates are known to synthesize glycogen granules. For example, Tetrahymena has glycogen granules between 35 and 40 nm in diameter, each granule being a collection of small γ-granules of 2–3 nm in size³⁶. Dinoflagellates and apicomplexans typically produce more complex and larger spherical starch particles, usually greater than 1 μm in size^37,38. Amylopectin-rich starch and glycogen are very similar polysaccharides, but they differ in granule size and birefringence (Table 2). Starch granules are large, birefringent, and have a high refractive index, but glycogen does not exhibit birefringence, and its granules generally have a size of 300 nm or less. When observed with a polarizing microscope, the starch granules show a Maltese cross pattern. This pattern is derived from the radial arrangement of amylose and amylopectin molecules in granules and is one of the criteria for starch identification. Since the cytoplasmic granules of S. pyriformis are large in size (1–3 μm) and show a typical Maltese cross pattern as shown in Fig. 4E, these granules are likely to be starch granules rich in amylopectin.

Phylogeny of S. pyriformis and its morphology

Relationships of Stentor species were not clearly resolved. BI and ML analyses indicated basal diverging of the S. pyriformis + S. amethystinus clade from others, but NJ analysis did not indicate so (Fig. 5). Recent phylogenetic analyses inclusive of Stentor species also indicated basal diverging of S. amethystinus from the others; however, the monophyly of the others is not highly supported^21,22. Therefore, the one thing that can be said is that S. pyriformis is closely related to S. amethystinus.

For the identification of Stentor species, the shape of macronucleus, presence or absence of cortical pigmentation, and symbiotic algae are very important and iconic characteristics^8,19. S. pyriformis and S. amethystinus share beaded macronuclei and the presence of symbiotic chlorella (Table S1, Fig. 5B). Pigmentation is present in S. amethystinus, but not in S. pyriformis. Pigmentation is a noticeable characteristic, which tinctures the whole body of Stentor cells. The pigment is thought to function as a defense against predators³⁹. However, the kind of pigment compound depends on the species⁴⁰, and the relationship between pigment possession and phylogeny is poor (Fig. 5). Of note, colorless vesicles exist in S. pyriformis (Fig. 2D). The short and fat shape is also a common characteristic for S. pyriformis and S. amethystinus, in this genus with many elongated trumpet shape species^6,8.

Symbiotic algae in S. pyriformis

Algae-targeted PCR products from whole cells of S. pyriformis were sequenced directly, and clear peaks were obtained for each. This shows that all or nearly all of the algal symbionts in each Stentor cell are unified, regardless of samples under long-term culture or nature. In addition, all symbionts were closely related to C. variabilis (Fig. S3), which has been known as a representative symbiont of P. bursaria (Oligohymenophorea), the model organism of multi-algae retaining protists (MARP⁴¹) style symbioses. For the chlorellacean species, the diversity of ITS2 sequence comparisons has often been adopted. For two organisms to compare, ITS2 sequence differences (gaps are counted as a fifth character) usually fall either less than 2% for single species or more than 10% for different species^42,43. This characteristic simply encourages a species concept. The ITS2 sequences of S. pyriformis algae differ only by one nucleotide site out of 248 sites from those of P. bursaria algae (Fig. 6A), which strongly suggests the symbiotic chlorella of S. pyriformis are also C. variabilis. Several Stentor species retain coccoid green algae⁸ (Table S1), but only three algal sequences have been published. Two algal sequences from S. polymorphus belonged to different clades from Chlorellaceae^44,45. As for the other algal sequence of S. amethystinus, the symbiont may belong to Chlorellaceae⁴⁶. This sequence (EF589816) is short (991 bp) and only covers a part of SSU rDNA; therefore, it was not included in our phylogenetic analyses (Fig. S3). The sequence differs from C. variabilis with 10 base changes and 3 indels, indicating that it is not C. variabilis.

In the case of P. bursaria-C. variabilis symbiosis, C. variabilis has been shown to be vastly different from other free-living species. C. variabilis demands organic nitrogen compounds⁴⁷ and leaks nearly half of the photosynthate to outside algal cells^48,49. Furthermore, they are sensitive to the C. variabilis virus (CvV; so-called ‘NC64A virus’), which is abundant in natural freshwater^50,51,52. Therefore, C. variabilis should be considered an already evolved species that is unable to survive without the protection of the host cell⁵³.

Four C. variabilis rDNA sequences obtained from S. pyriformis were identical, with the exception of a nucleotide position in the S1512 intron. Here, the regions without group I introns, i.e., SSU, ITS1, 5.8S, and ITS2 rDNA, are compared among C. variabilis sequences of S. pyriformis and of P. bursaria. Several published sequences cover the above SSU-ITS region, of which varieties are shown as P. bursaria symbiont genotype (PbS-gt) 1 to 3 (Fig. 6A). Due to the small number of sequences, it is still unknown whether these genotypes depend on (or are related to) their living regions. Genotype 1 was from USA and Japan, genotype 2 was from China, and genotype 3 was from Australia. All available sequences for S. pyriformis symbionts were obtained in this study, and they were all from Japan. As a result, all sequences of S. pyriformis symbionts were aggregated into a single genotype SpS, which was distantly related to all P. bursaria symbionts, including those from Japan (Fig. 6B). Five variable sites are found in SSU rDNA among C. variabilis genotypes, of which four are concentrated to that of the symbionts of S. pyriformis (SpS) (Fig. 6A). C/T substitution at alignment position 656 will be a hemi-compensatory base change (hemi-CBC) at the E23_2 helix of SSU rRNA structure (Fig. 6C), whereas the other four sites are at single strand regions (data not shown). Mutations (1821–1828) including comparatively large indels were seen in ITS1 region (Fig. 6A). It was found that all these mutations are assembled in helix 1 (for chlorellacean ITS1 structure, see Bock et al.^54,55). Thermodynamic analysis via Mfold^56,57 predicted that PbS sequences form linear helix 1 similar to the other chlorellacean species, but SpS sequences including the additional nucleotides may form a dichotomous branching of helix 1 (Fig. 6D).

The group I introns inserted in SSU rDNA of S. pyriformis symbionts are identical to those of P. bursaria symbionts^28,58 in terms of numbers (three introns) and insertion sites (S943, S1367 and S1512). The sequences of S943 and S1512 introns are matched more than 99%. However, with respect to the S1367 intron, a large length gap was found (168 nucleotides) at the tip of P8 (Fig. S4). This section has been indicated as a homing endonuclease gene remnant²⁸, and those of S. pyriformis symbionts are presumed to be a more degenerated form than those of P. bursaria symbionts.

At any rate, the symbiotic algae of S. pyriformis were found to be C. variabilis. Because S. pyriformis never lost the symbiotic algae in four years of culture, and all four algae had nearly identical genetic characteristics, the symbiotic relationships between S. pyriformis and C. variabilis can be regarded as stable, or permanent. Although S. pyriformis and P. bursaria share C. variabilis as their endosymbionts, considering the genetic differences depending on their host species, the sharing event has not happened recently. Symbiont sharing among various host species has also been known for some ciliates^41,59 (Carolibrandtia ciliaticola in Fig. S3), and a script to spread a particular algal symbiont has been suggested⁴¹. Given the physiological characters of C. variabilis (mentioned above), this algal species might be an ideal algal symbiont, and it will be no surprise if the other protists also retained C. variabilis as their algal partners. Research on the symbiotic algae that other Stentor spp. have and on host and regional dependencies are awaited.

Adaptation of S. pyriformis to oligotrophic environment in highland marsh

In Japan, S. pyriformis lives only in alpine ponds (Fig. 1), where the winter is cold, and the surface of the pond is always covered with ice. The water in these ponds has low electrical conductivity (~ 10 μS/cm), and there are few living organisms except S. pyriformis, meaning that only little food is available in wintertime. The reason this ciliate is rich in stored carbohydrate granules may be due to its need for nutrients to survive such harsh winter environments.

Preliminary studies suggest that many protists, especially ciliates, may make starch. Large amounts of cytoplasmic granules that show a Maltese cross were observed in chlorella-bearing ciliates such as P. bursaria, while only a small amount of such granules was observed in Euplotes aediculatus, Paramecium caudatum, Blepharisma japonicum, and Tetrahymena pyriformis. Protists with symbiotic algae seem to produce particularly large amounts of stored carbohydrate granules in the cytoplasm, but the mechanism of starch synthesis may be widely shared by ciliates.

P. bursaria has been shown to be more resistant to starvation conditions than the aposymbiotic strain of the same species¹³. Under food-deprived conditions, P. bursaria was interpreted to have survived by digesting symbiotic algae. Resting cyst formation and cannibalism are known as other strategies for protozoans to survive starvation conditions⁶⁰. This study suggests that the use of carbohydrate granules stored in cells may be another possible strategy for ciliates to survive harsh environments such as highland oligotrophic bogs.

Source: Ecology - nature.com