Molecular Investigation of Paramecium bursaria Endosymbiotic Algae: the First Records of Symbiotic Micractinium reisseri from Kamchatka

Paramecium bursaria is a symbiotic ciliate species which cells contain hundreds of algae enclosed in perialgal vacuoles. The aim of the present study was to identify endosymbiotic algal strains of P. bursaria and to define the geographical distribution of the identified species. We analyzed symbiotic strains of P. bursaria originating from distant geographical locations and housed at the Culture Collection of Ciliates and their Symbionts (CCCS) at St. Petersburg University. Based on the obtained results, we identified these strains as Micractinium reisseri, Chlorella vulgaris, and Chlorella variabilis. We did not confirm the occurrence of a division into American and European groups and we guess that this division is only contractual and corresponds to the amount of introns in the 18S rDNA, and that there is no strong correlation with the geographical location. We have demonstrated that the range of M. reisseri is greater than previously supposed. We identified algae strains originating from Southern Europe (Serbia), Western Asia, and from the Far East (Kamchatka) as M. reisseri. Moreover, we identified two strains originating from Europe as C. variabilis, which also contradicts the predetermes about a division into American and European groups.

Paramecium bursaria is a ciliate species that maintains symbiotic relationships with algae. A single cell of P. bursaria possesses about 700 symbiotic algal cells in the cytoplasm (KODAMA & FUJISHIMA 2009). Each endosymbiont's cell is enclosed in a perialgal vacuole membrane derived from the host digestive vacuole membrane, which protects the alga from the host's lysosomal fusion (KARAKASHIAN & RUDZINSKA 1981;GU et al. 2002). The algae cells provide the photosynthetic products to their host and in return they receive carbon dioxide and nitrogen compounds (KODAMA & FUJISHIMA 2009). Furthermore, symbiotic algae are protected from viral infection and are chauffeured to brightly lit areas for optimum photosynthesis (HOSHINA & IMAMURA 2009a). Meanwhile, the presence of algal symbionts minimizes the level of photooxidative stress that P. bursaria is exposed to (HÖRTNAGL & SOMMARUGA 2007).
The symbiotic algae of Paramecium bursaria belong to two classes: Trebouxiophyceae and Chlorophyceae (HOSHINA et al. 2010a;LUO et al. 2010;PRÖSCHOLD et al. 2011), and are characterized by a close morphological similarity. PRÖSCHOLD et al. (2011) have identified four species of symbionts living in a symbiotic relationship with P. bursaria: M. reisseri, C. vulgaris, C. variabilis, and Scenedesmus sp. The Chlorella-clade includes endosymbiotic algae belonging to two genera: Chlorella (HOSHINA et al. 2010a) and Micractinium ). The systematics of Chlorella are constantly being modified since the authors reported from 4 (HUSS 1999;KRIENITZ et al. 2004) to 14 species (BOCK et al. 2011).
The genus Chlorella includes non-motile spherical cells 2-10 µm in diameter, a single nucleus, vacuoles, mitochondria, a few peroxisomes, and a single chloroplast with a pyrenoid surrounded by starch grains (KESSLER & HUSS 1992). Algae from the Micractinium genus, are morphologically similar to those of Chlorella, however cells of Micractinium are equipped with bristles and they are usually organized into colonies (LUO et al. 2005). A surprising fact is that when M. reisseri is isolated from P. bursaria, it does not form bristles and lives as a single cell . Therefore, P. bursaria endosymbionts which belong to two classes are very hard to distinguish through microscopic observations (morphological analysis).
Up-to-date attempts at symbiotic algae identification have been carried out through the microscopic observation and physiological parameters measurement (REISSER 1984;DOUGLAS & HUSS 1986;KESSLER & HUSS 1990), analyzing cell wall structure (TAKEDA 1995), isoenzymes and sensitivity to viruses (LINZ et al. 1999;KVITKO et al. 2001), as well as the content of GC pairs in DNA (KESSLER & HUSS 1990). FOTT and NOVÁKOVÁ (1969) suggest that the morphological and biochemical features which are used as identification tools, as well as the size and the shape of the cell are highly variable parameters and depend on the age of the culture as well as nutrition and environmental conditions. Therefore, the application of molecular markers seems to be a promising tool for algae taxonomy (TAYLOR & HARRIS 2012).
Our objective was to identify symbiont species of P. bursaria strains from the CCCS collection and to define the geographical distribution of the identified species. Taking into account the fact that endosymbionts of P. bursaria are indistinguishable when comparing morphological features, identification based on molecular analyses seemed to be the only way to classify them into a particular taxon.

Molecular methods
Symbiont's DNA was extracted using a GeneJET Plant Genomic DNA Purification Kit (ThermoScientific) according to protocol. Before isolation, culture of P. bursaria was carefully purified using special filters which allowed us to obtain pure culture of P. bursaria cells. 1.5 ml of dense P. bursaria culture was harvested from liquid culture by centrifugation. The pellet was frozen in liquid nitrogen and the mixture was sonicated on ice for 10 s at 40 W. After that, we followed the standard extraction protocol.
For molecular analysis we applied a fragment of the ITS1-5.8S rDNA-ITS2, as the most widely used, marker for Paramecium algal endosymbiont identification (for example BOCK et al. 2011;PRÖSCHOLD et al. 2011 and the other literature cited herein), due to its high degree of nucleotide substitutions, which allows for the comparison of closely related taxa, and which is highly variable among different species, whilst it is conserved within the same species (HOSHINA et al. 2010a). The fragment of ITS1-5.8S rDNA-ITS2 was amplified using primer pairs: ITS1 (WHITE et al. 1990)/ITS2R (primer designed in the present study, Table 2) or ITS1F/ITS2R (primers designed in the present study, Table 2) according to protocol with the following parameters: initial denaturation at 95ºC for 5 min, followed by 30 cycles of denaturation at 95ºC for 1 min, annealing at 54ºC for 2 min, extension at 72ºC for 3 min and a final extension at 72ºC for 5 min. The primers, which would amplify the DNA fragment we were studying, are specific to algae, and were designed according to the following scheme: (I) comparison of several algae sequences available in GenBank and identification of homologous, conservative fragments, (II) application of Reverse Complement software (http://www.bioinformatics.org/sms/rev_comp.html) in order to obtain sequences of reverse primers, (III) determining the Tm for the Forward (sequence 5'-3') and Reverse primer (sequence 3'-5') (the temperature of both primers should be similar) using the Primer Blast program (https://www.ncbi.nlm.nih.gov/tools/primer-blast). PCR amplification for all analyzed DNA fragments was carried out in a final volume of 40 ìl containing 4 ìl of DNA, 1.5 U Taq-Polymerase (EURx, Poland), 0.8 ìl of 20 ìM of each primer, 10 × PCR buffer, and 0.8 ìl of 10 mM dNTPs in a Thermal Cycler PCR (Gstorm). After amplification, the PCR products were electrophoresed in 1% agarose gel for 1 hour at 95V. After that, they were purified from the gel using Nu-cleoSpin Extract II (Macherey-Nagel, Düren, Germany). Cycle sequencing was done in both directions with the application of BigDye Terminator v3.1 chemistry (Applied Biosystems, USA). The primers that were used for amplification were also applied for sequencing. Each sequencing reaction was carried out in a final volume of 10 ìl containing 3 ìl of template, 1 ìl of BigDye Master Mix (1/4 of standard reaction), 1 ìl of sequencing buffer, and 1 ìl of 5 ìM primer. Sequencing products were precipitated using Ex Terminator (A&A Biotechnology, Poland) and separated using the Genomed Company (Poland). Sequences are available in the GenBank database (for accession numbers see Table 1).

Analysis of ITS1-5.8S-ITS2 rDNA fragments variation
We analyzed 20 ITS1-5.8S-ITS2 rDNA fragments (570 bp) of algae including 12 sequences of Micractinium, 3 sequences of C. vulgaris, 5 sequences of C. variabilis, and identified 14 haplotypes. The interspecific haplotype diversity value (Hd) was 0.937 and the nucleotide diversity (ð) was 0.07407 (Table 3).   Taxonomic classification, and reciprocal relationship of the currently studied algal species The phylogram (ML/NJ) constructed on the basis of fragments of ITS1-5.8S rDNA-ITS2 (Fig. 2), isolated from 20 algae strains, revealed strains grouping into three clusters (A, B, and C). The first of them -A is composed of the symbiotic algae of P. bursaria originating from the Danube River in Serbia (MC-SRB9-1) and St. Petersburg in Russia (MC-MS-1). Furthermore, there is also a strain, MC-4 231-1, isolated from P. bursaria collected from Kamchatka (Russia) (Tab. 1). Additionally, the cluster included 6 sequences of Micractinium sp. and 3 sequences of M. reisseri obtained from GenBank. Our strains of this cluster were assigned to M. reisseri after comparing the analyzed sequences with records published in GenBank (97% similarity to the closest match) as well as based on the constructed tree: they form a monophyletic clade together with Micractinium strains with a rather high bootstrap support (ML/NJ: 81/78). The second cluster -B includes symbiotic strains originating from Lake Loch Linnhe, Scotland (CVG-GB15-2) and Kaliningrad, Russia (CVG-KZ-126) (Tab. 1, Fig. 2) and a sequence of C. vulgaris from GenBank.
These strains have been assigned to C. vulgaris based on the grouping with the strain DRL3 (92% similarity and bootstrap values for ML/NJ: 100/100). The third clade -C is composed of strains originating from St. Petersburg, Russia (CVA-BS-7) and the Astrakhan Nature Reserve, Russia (CVA-AZ20-4) and 3 sequences of C. variabilis obtained from GenBank (97% similarity to the closest match, bootstrap values for ML/NJ: 99/93). Our strains of this cluster have been identified as C. variabilis, because of a monophyly with 3 sequences of C. variabilis obtained from GenBank (Fig. 2).
The haplotype network of the fragment of ITS1-5.8S rDNA-ITS2 (Fig. 3) divided the strains into 3 haplogroups. The first one -Micractinium includes 7 haplotypes. Five of them correspond to single strains: 2 of the Micractinium sp. from GenBank and 3 of M. reisseri (newly analyzed strains marked in a darker violet). One of the remaining 2 haplotypes represents 4 strains of Micractinium sp. (GenBank) and the last haplotype represents 3 strains of M. reisseri (conductrix) obtained from GenBank. Molecular variability between particular haplotypes of that haplogroup oscillates from 1 to 10 nucleotide substitutions (Fig. 3). The second haplogroup is composed of 3 unique haplotypes: 2 of newly analyzed strains (marked in darker green) with 4 differences between them and a haplotype of C. vulgaris (strain DRL3) from Gen-Bank which is different from the other two haplotypes mentioned above by over 30 nucleotide substitutions.
And finally, the last haplogroup contains 3 haplotypes. One of them represents 3 strains of C. variabilis from GenBank, and the other two haplotypes (marked in darker red) correspond to the currently studied strains. Molecular variability between particular haplotypes of that haplogroup oscillates from 4 to 15 nucleotide substitutions. In turn, distances between different algal species are much greater: there are about 70-80 nucleotide substitutions between C. vulgaris and C. variabilis, 100-120 nucleotide substitutions between C. vulgaris and Micractinium, and 140-160 nucleotide substitutions between C. variabilis and Micractinium (Fig. 3).

Discussion
The application of molecular analyses is crucial to resolve phylogenetic relationships, especially when the organisms in question are not distinguishable using conventional methods, like microscopic observation or the analysis of physiological parameters. Based on molecular analyses, almost all symbiotic algae of P. bursaria were divided into two groups: American and European (HOSHINA et al. 2004;HOSHINA et al. 2005). According to the results obtained using gene encoding 18S rRNA as a marker, the symbiotic algae of P. bursaria are related to three species: C. vulgaris, C. sorokiniana and C. lobophora (GAPANOVA et al. 2007). HOSHINA and IMAMURA (2009a) described the characteristic geographical distribution of the two groups. The symbiotic algae belonging to the European group originate usually from Great Britain, Germany, Austria and Kaliningrad and The size of the circles is proportional to the haplotype frequency. The median vectors that represent hypothetical intermediates or un-sampled haplotypes are shown as black circles. Haplotypes highlighted by lighter colours represent data from GenBank, whereas those highlighted by darker colours are currently studied strains. Hatch marks on particular branches represent nucleotide substitutions between particular haplotypes (in the case of 10 or more, a corresponding number was given). Analyses were conducted using the Median Joining method in PopART software v. Before our current findings, there were not any reports of M. reisseri occurring in the Far East. Both viruses, CvV (infecting C. variabilis) and MrV (infecting M. reisseri), have been detected from distant regions of the world, but MrV has never been recorded from East Asia ( VAN ETTEN 2003;YAMADA et al. 2006;HOSHINA et al. 2010b). The results obtained in the present study are supported by values of bootstrap reaching 81/78% for the ML/NJ phylogram constructed based on a comparison of the ITS1-5.8S rDNA-ITS2 sequences (Fig. 2). Furthermore, regions of Kamchatka (Russia) are located close to the boundaries of the American group which includes the Far East. PRÖSCHOLD et al. (2011) stated that strains belonging to the American group can be assigned to C. vulgaris or C. variabilis. However, in the present study, strains of C. variabilis were collected from Austria (Wien) and Russia (St. Petersburg and the Astrakhan Nature Reserve).
According to our results, we can conclude that the geographical distribution of M. reisseri is not restricted to only Europe and that the division of symbiotic algae into two groups is only contractual and is related to the number of introns in 18S rDNA (GAPANOVA et al. 2007). Moreover, these differences do not refer to all species of symbiotic algae and what is even more evident is that they don't have a strong connection with the geographical locations of algae. All of the divergences can be due to the fact that, so far, there has been an analysis of strains collected from a few places located very far from each other (for example Western Europe and the Far East). Analyses carried out on symbiotic algae were usually limited to few samples of a particular region. A persistent problem in many of the molecular phylogenetic investigations thus far might be caused by undersampling, which results in systematic errors in phylogenetic reconstruction. For example, some early 18S phylogenies showed a sister relationship between Chlorophyceae and Trebouxiophyceae (KRIENITZ et al. 2001), while more recent studies, that increased taxon sampling, revealed a sister relationship between Chlorophyceae and Ulvophyceae (WATANABE & NAKAYAMA 2007;DE WEVER et al. 2009). Therefore, an analysis which involves dense taxon sampling is important in order to avoid systematic errors in phylogenetic analyses. In order to resolve the phylogenetic relationships between symbiotic algae of P. bursaria originating from all over the world, the next step should be a research extension to new regions from Europe to the Far East and an increase in the taxon sampling of Palearctic and Nearctic ecozones.