**4. Discussion**

#### *4.1. Green Algae in Endosymbiosis Belonging to the Chlorellaceae*

Zoochlorellae or *Chlorella*-like algae living as endosymbionts in ciliates and other protozoa are known for a long time ([7] and references therein). Interestingly, most of these green algae belonged to the *Chlorella* clade of the Trebouxiophyceae. Within this clade, six out of the seven species (highlighted with green circles in Figure 2) exclusively occurred in endosymbiotic associations. Only *Chlorella vulgaris* could be found free-living in various habitats (see details in reference [18]). *C. vulgaris*, *C. variabilis*, and *Micractinium conductrix* formed an obligate endosymbiont in *Paramecium bursaria* [7,29]. *Meyerella planctonica* is the endosymbiont of another green *Paramecium* (*P. chlorelligerum* [10]). The genus *Carolibrandtia* was discovered to be the endosymbionts of the ciliates *Pelagodileptus trachelioides*, *Cyclotrichium viride*, and *Stokesia vernalis* [11,12].

*Micractinium tetrahymenae* sp. nov. represented a second species within this genus that lives in symbiosis with a ciliate. In contrast to *M. conductrix*, a species which is the obligate endosymbiont of *Paramecium bursaria* [7,29] and *Coleps primhirtus* [30], *M. tetrahymenae* formed only under anoxic or microaerobic conditions a symbiotic association with *Tetrahymena utriculariae* [14]. This demonstrated that *M. tetrahymenae* is a facultative endosymbiont. However, whether this species can also occur free-living needs further investigations. No entry in GenBank could be found in the BLASTn search (100% coverage, 97% identity) using our SSU and ITS sequence (2452 bp). It is also unknown if *Tetrahymena utriculariae* would be able to live in symbiosis with other green algae belonging to the Chlorellaceae.

#### *4.2. Taxonomy and Systematics of the Genus Micractinium*

Morphologically, both endosymbiotic *Micractinium* species were difficult to distinguish from each other. *M. tetrahymenae* sp. nov. was slightly smaller than *M. conductrix* (3–8 vs. 4–10 μm). Both species showed no bristle formation under the chosen culture conditions. Three other species of *Micractinium*, all occurring free-living, were known to be bristle-less (*M. inermum*, *M. simplicissimum*, and *M. singulare* [31,32]). Colony formation among *Micractinium* species was not always observed. The morphological features of all currently accepted species are summarized in Table 1.The taxonomy and systematics of spiny coccoid green algae is very confusing and unclear for two major reasons: (i) Most species were described based on field samples and no type material of these species is available for comparative studies; often only pictures were presented as holotypes [2]; (ii) cultured material such as strains of *Micractinium* were unicellular and without any bristles, which made it almost impossible to distinguish them from members of the genus *Chlorella*. Luo et al. [3,4] demonstrated that bristle and colony formation is an inducible defense mechanism against grazing of the rotifer *Brachionus calcyflorum.* Phylogenetic analyses such as those presented in Figure 2 revealed the close relationship between *Chlorella* and *Micractinium*. In contrast to *Micractinium*, the monophyly of *Chlorella* was not supported in our analyses. However, the molecular signature described by Pröschold et al. [7], the CBC at the end of helix III in ITS-2 (G-C in *Chlorella* vs. C-G in *Micractinium*), remained.

Traditionally both genera belonged to two different families. The family Chlorellaceae comprised algae reproducing exceptionally by autospores without sexual reproduction or zoosporogenesis. Other important criteria for separation of Chlorellaceae was composition of cell wall, which consisted of 2–3 layers containing obligatory cellulose and an outside layer of sporopollenin [2]. Unfortunately, this feature was based on the investigation of *Chlorella fusca* (now *Scenedesmus abundans*, Chlorophyceae) and the cell wall of "true" *Chlorella* species did not contain sporopollenin [33]. In contrast, the family Micractiniaceae contains algae, in which sexual reproduction, but no production of zoospores, is known. The cells are arranged in colonies consisting out of 2 up to 256, and were covered with bristles. The cell walls contain cellulose, without sporopollenin [34,35]. In summary, the differences between both families were the presence of sexual reproduction and bristles in Micractiniaceae. However, phylogenetic analyses have revealed that both families were polyphyletic (see [36] and references therein).

Hegewald and Schnepf [34,37] revised the representatives of the family Micractiniaceae based on morphological, ultrastructural investigations using SEM and TEM. They studied living cultures and some formaldehyde-fixed type material to explore the nature of spines and bristles used for the di fferentiation at generic level within this family. By definition, bristles contained, in contrast to spines, no cellulose and only proteins in their appendices. In addition, the formation of both is di fferent. Whereas spines were formed before the cell walls were produced, bristles were exhibited after the cells are covered by the rigid cell wall. Considering these features, they revised the genus *Micractinium* by transferring several species to this genus, which were originally as species of other genera, such as *Golenkinia* and *Golenkiniopsis*. The genus *Micractinium* comprised four species, *M. pusillum*, *M. appendiculatum*, *M. elongatum*, and *M. parvulum*, according to Hegewald and Schnepf [34,37], and the complicated synonymy were provided therein. However, the validation of these taxonomical combinations needs to be proven.



The latter species was transferred to another genus, *Hegewaldia*, based on phylogenetic analyses of SSU and ITS rDNA sequences [6]. In addition, they also transferred *Diacanthos belenophorus* to *Micractinium*, which was assigned to the Micractiniaceae by Hegewald and Schnepf [38].

Interestingly, it is the occasional occurrence of the sexual reproduction in the family Micractiniaceae. The oogamy was observed in *Micractinium pusillum* by Nygaard [39], Lund [40], Korschikov [41], and Hegewald [42], and in *Hegewaldia parvula* by Iyengar and Balakrishnan [43], Starr [44], and Ellis and Machlis [45], originally assigned as *Golenkinia minutissima*. The ultrastructure of the spermatozoid was investigated by Moestrup [46], who showed that the spermatozoid had an untypical structure of the flagella (9 + 1). The presence of sexual reproduction in *Micractinium* and *Hegewaldia* and its absence in *Chlorella* could be potential criteria for distinguishing the genera. However, Fucikova et al. [47] found in *Chlorella* meiotic genes and genes that were transcribed during sexual reproduction, in only asexually reproducing trebouxiophytes. This questioned the traditional concept of genera.

As already pointed out Hegewald and Schnepf [34], even the formation of bristles considered as a good morphological feature, is not a stable feature. The morphology and length of bristles is polymorphic and dependent on temperature and media. For example, they observed that *Micractinium strigonense* Hortobagyi sometimes have different bristles (thick and delicate) and occurred sometimes without bristles. Considering these observations, they proposed to synonymize several species, which is unfortunately illegitimate.

The high phenotypic plasticity and the lack of stable morphological and ultrastructural characters requested a new generic and species concept within the Chlorellaceae and Micractiniaceae. These traditional families should be rejected according the phylogenetic analyses of molecular marker genes. Considering the SSU and ITS sequences, new species were described from Japan [31] and Antarctica [32]. The integrative approach used in this study clearly demonstrated that *Micractinium* contained eight species (Figure 2). The morphological features of those species as well as the remaining species of Hegewald and Schnepf [34,37] were compared in Table 1. The comparison and judgement of traditional features and molecular data is quite difficult. For example, both endosymbiotic species showed only small morphological differences and were not considered as members of *Micractinium* without phylogenetic analyses. However, our study showed that both are separate species based on the CBC approach, as demonstrated in Figure 7. On the other hand, molecular data provided an inflation of new species descriptions, when the traditional literature was not considered and no strains are available in public culture collections. As an example, Chae et al. [32] described *Micractinium variabile* based on SSU and ITS rDNA sequences. Morphologically, this species is very similar to *M. quadrisetum*, which is unfortunately not available in culture. Therefore, it is possible that both species represent only one species. According to the ICN, *M. quadrisetum* would have priority against *M. variabile*. As described in the results, only little genetic differences among *Micractinium* species could be discovered. In particular, *M. inermum* and the three species described by Chae et al. [32] had identical V9 regions and little differences in ITS-1 and ITS-2, but they differed by two CBCs and three HCBs (Figures 5–7). Considering the ITS-2/CBC approach, we do not propose any taxonomic changes without further investigations.

#### *4.3. Ecology and Distribution of Micractinium*

The genera *Chlorella* and *Micractinium* have different ecological patterns and are distributed in various habitats. Whereas *Chlorella* has a worldwide distribution in almost all kinds of habitats, it seems that *Micractinium* is restricted to freshwater habitats. Species of *Chlorella* were found aquatic in freshwater and marine habitats [18,36], symbiotic in ciliates and heliozoa [7], and terrestrial [48,49]. *Micractinium* species were only observed in freshwater habitats [2,31,32,36], in wet soils [5], and symbiotic in ciliates ([7,14] and this study). The occurrence of *M. tetrahymenae* in the traps of *Utricularia* is exceptional. Whereas *Tetrahymena* species are widely distributed in the bladder traps of different *Utricularia* species, only one record is known of the green *Tetrahymena utriculariae* [14]. No other record of the occurrence of a *Micractinium* species in such traps have been reported in microbiome studies [50]. Simek et al. [15] studied the ecology and dynamic of trap communities and found that the endosymbiosis of *Micractinium* in *Tetrahymena* contributed significantly for the survival of the ciliate in such harsh environment. No other *Utricularia* species have had mixotrophic ciliates in their traps until now [50].

#### *4.4. Interactions between Tetrahymena Utriculariae and Micractinium Tetrahymenae*

The role of the *Micractinium* symbiont in *Tetrahymena utriculariae* has been studied in experiments by Simek et al. [15]. The ciliate has a flexible life strategy. It can live in different aquatic environments under oxygenic conditions, or if captured in bladder traps of *Utricularia* under anoxic conditions. For this flexibility, the endosymbiotic *Micractinium* is absolutely necessary. It has been demonstrated that aposymbiotic *Tetrahymena* had the highest growth rate, if exclusively bacterial food is present. However, if cultivated with both bacterial food and symbiotic *Micractinium*, *Tetrahymena* had a reduced growth rate, but after 44 days of cultivation, 80% of the Tetrahymena cells reestablished the symbiosis with the algae [15]. These experiments clearly demonstrated that both organisms formed a symbiotic association depending on the environment. The main profit for the host is that the algae produced the oxygen through photosynthesis. If the algae also provided nutrients to the host, this has not been investigated so far. *Micractinium tetrahymenae* benefited from CO2 production of the host and stable conditions inside the host, whereas outside in ciliates the environment was very harsh (low pH 4.3 and anoxic) [51]. The endosymbiosis with this *Micractinium* species is probably essential for *Tetrahymena*, because the green algae were included in cys<sup>t</sup> formation [14].
