The genus Cryptomonas, which belongs to the class Cryptophyceae, was established by Ehrenberg (1831). It is distributed in freshwater habitats worldwide. Cells can be easily recognized by two unequal biflagella, olive-brownish to olive-greenish in color, large ejectisomes lined in the furrow-gullet system, and a peculiar swaying swimming behavior due to the asymmetric shape, dorsally flattened and ventrally concave in lateral view. Cells have two chloroplasts originated from red algae, which contain the accessory pigment phycoerythrin 566 of phycobiliprotein (Hill and Rowan 1989, Clay et al. 1999, Deane et al. 2002, Hoef-Emden and Melkonian 2003). Cryptomonas species display one or two morphotypes within a clonal culture. In the cryptomorph, the inner periplast component (IPC) is made of hexagonal to polygonal plates, whereas in the camphylomorph the IPC is a sheet-like layer (Faust 1974, Brett and Wetherbee 1986, Hill 1991, Hoef-Emden and Melkonian 2003, Hoef-Emden 2007).

Since Ehrenberg (1831, 1832) described six Cryptomonas species, many additional species were added (Pascher 1913, Schiller 1925, 1929, 1957, Skuja 1939, 1948, 1956, Huber-Pestalozzi 1950, Starmach 1974). Traditionally, Cryptomonas species have been characterized by mainly morphological characters, such as cell size, cell

[Fig. 1.] Distributions of the genus Cryptomonas species from Korea.

shape, and internal organization (Bourelly 1970). However, it is difficult to delimit Cryptomonas species due to the paucity of morphological characters and the less-than-adequate visibility of living cells using light microscopy (Pringsheim 1968). For example, an early attempt to organize photosynthetic cryptomonad taxonomy was carried out by Ehrenberg (1838), who included several unrelated genera in the family Cryptomonadina with original short diagnosis. Later, Dujardin (1841) described two families (Xanthodiscaceae and Chilomonaceae) within the order Cryptomonadineae. In his classification system, the family Chilomonadaceae consisted of four genera (Chilomonas, Rhodomonas, Cryptomonas, and Cyanomonas). These genera were characterized by nutritional mode, number and color of chloroplast. Butcher (1967) organized main photosynthetic genera of the Cryptophyceae into three families (Hilleaceae, Hemiselmidaceae, and Cryptomonadaceae) based on complexity of furrow-gullet system with or without ejectisome. He also used number of ejectisome rows in the furrow-gullet system to discriminate between each genus, and thus, described 12 new Cryptomonas species from salt water.

Hill (1991) revised the broad generic definition of the genus Cryptomonas recognized by Butcher (1967) and erected four new genera based on their furrow-gullet system, periplast structure, plastidial complex and rhizostyle; Campylomonas, Geminigera, Storeatula, and Teleaulax. Recently, molecular work (Marin et al. 1998, Deane et al. 2002) has shown that Cryptomonas species grouped together with species of Campylomonas and Chilomonas. Hoef-Emden and Melkonian (2003) synonymized both genera Campylomonas and Chilomonas to the genus Cryptomonas. More recently, Hoef-Emden (2007) revised the genus Cryptomonas again, and provided a secondary structure of the nuclear internal transcribed spacer 2 (ITS2) as a good marker to identify Cryptomonas species. In addition, she emended five species based on molecular signatures as diagnostic characters.

In this study, we report unrecorded Korean Cryptomonas species isolated from freshwaters. We infer phylogenetic relationships among species using a combined nuclear-encoded ITS2, partial large subunit ribosomal DNA (LSU rDNA), and small subunit ribosomal DNA (SSU rDNA), and chloroplast-encoded psbA (photosystem II protein D1) and LSU rDNA sequence data. We also predicted the ITS2 secondary structure of Cryptomonas species.

Specimens were collected from freshwater habitats in Korea (Fig. 1). Live cells were isolated by Pasteur capillary pipette and were brought into a unialgal culture. The cells were cultivated in f/2 medium (Guillard and Ryther 1962, Guillard 1975) with soil extract. The clonal cultures were maintained under a light : dark regime of 14 : 10 at 20-22℃ using cool-white fluorescence lamps with illuminations of 30 μmol protons m^-2 s^-1. The morphology was examined by differential interference contrast with a 100X oil immersion lens (Carl Zeiss Co., Gottingen, Germany). Calibration of magnification was done with grated micrometer. The shape and length of cells, length of the fullow-gullet system, color and number of chloroplasts, presence / absence and number of pyrenoids were examined. Cellular dimensions were determined by measuring 20-25 cells of each taxon from photographic images. Light micrographs were taken with an Axio CamHRc (Carl Zeiss Co.) photomicrographic system attached to the microscope.

Approximately 10 mL of cultures in exponential growth were harvested by centrifugation (4,500 ×g, model 5415; Eppendorf, Hamburg, Germany) for 1 min at room temperature and washed three times with sterilized distilled water. Total genomic DNA was extracted from the pellet using the Dokdo-Prep Blood Genomic DNA Purification Kit (Elpis-Biotech Inc., Daejeon, Korea) following the manufacturer’s blood sample protocol. PCR was performed using specific primers for nuclear ITS2, nuclear SSU rDNA, chloroplast psbA and chloroplast LSU rDNA (Table 1). The PCR amplification was performed on a total volume of 25 μL, containing 0.15 μL of TaKaRa Ex Taq DNA polymerase (TaKaRa Bio Inc., Otsu, Japan), 2 μL of each dNTP, 2.5 μL of 10× Ex Taq buffer, 1 μL of each primer, and 1-10 ng of template DNA. The nuclear ITS2, nuclear SSU rDNA, chloroplast psbA, and chloroplast LSU rDNA were amplified using a PTC-0150 Minicycler (MJ Research, Perkin-Elmer Co., Norwalk, CT, USA) with the following program: 94℃ for 5 min, 30 cycles of 94℃ for 1 min, 37-55℃ for 1 min, and 72℃ for 4 min, 72℃ for 10 min and a 4℃ hold. The PCR products were ~1.5 kb for nr ITS2 partial LSU rDNA, 1.7 kb for nr SSU rDNA, 1.0 kb for cp psbA, and 2.7 kb for cp LSU rDNA and were purified using the Dokdo-Prep PCR Purification Kit (Elpis-Biotech Inc.) according to the manufacturer’s protocol. The purified template was sequenced with internal primers of conserved regions using an ABI 3730xl sequencer (Perkin-Elmer Applied Biosystems, Foster City, CA, USA).

[Table 1.] PCR and sequencing primers

PCR and sequencing primers

The alignment for each gene sequence was aligned by the eye, and was edited using the Genetic Data Environment (GDE 2.4) program (Smith et al. 1994). Unalignable nucleotides were excluded from phylogenetic analyses.

Nuclear ITS2 is likely a suitable marker to identify species according to its degree of conservation (Hoef-Emden 2007), and nuclear ITS2 as well as partial LSU rDNA sequences were used to examine groups of genetically identical strains and to identify species of the strains.

Phylogenetic trees were constructed using Bayesian analysis (BA). Before the BA, we performed a likelihood ratio test using Modeltest, version 3.7 (Posada and Crandall 1998) to determine the best model under the hierarchical likelihood ratio tests (hLRTs) and Akaike Information Criterion (AIC). Evolutionary best-fit model was selected as the GTR + I + Γ model from a combined five gene data (nr ITS2 and nr partial LSU rDNA, nr SSU rDNA, cp psbA, and cp LSU rDNA sequences). The GTR + I + Γ model from the combined data was estimated by the following values; empirical base frequencies (A = 0.2701, C = 0.2014, G = 0.2697, T = 0.2589), substitution rates (A ↔ C = 1.0745, A ↔ G = 3.9918, A ↔ T = 1.9262, C ↔ G = 0.6372, C ↔ T = 8.1268), proportion of invariable sites (0.6217), and gamma distribution shape parameter (0.6929).

BA was performed with MrBayes 3.1.2 (Huelsenbeck and Ronquist 2001). Each analysis was initiated from a random starting tree, and the program was set to run four chains of Markov chain Monte Carlo iterations simultaneously for 2,000,000 generations. Trees and parameters were sampled every 1,000 generations, and the burn-in point was identified graphically by tracking the likelihoods (Tracer V.1.5; http://tree.bio.ed.ac.uk/software/tracer/), and then the first 800 trees were burned to ensure that they had stabilized. A majority rule consensus tree was calculated from the remaining trees to examine the posterior probabilities of each clade. The maximum likelihood (ML) phylogenetic analyses were done using the RAxML 7.0.4 program (Stamatakis 2006) with the general time reversible (GTR) model. We used 1,000 independent tree inferences using the -# option of the program to identify the best tree. Bootstrap values were calculated using 1,000 replicates, using the same substitution model.

Nuclear ITS2 sequences were folded using the mfold server(http://mfold.rna.albany.edu/?q=mfold/RNA-Folding-Form) (Zuker 2003) and RNAfold server (http://rna.tbi.univie.ac.at/cgi-bin/RNAfold.cgi)(Mathews 2004) with default values. A complete secondary structure graph of the nuclear ITS2 of Cryptomonas sp. M1634 was previously published (Hoef-Emden 2007). Prediction of secondary structure was performed according to putative secondary structure graph of Hoef-Emden (2007) as a template. Structures inferred by mfold or RNAfold were examined for common stems, loops, and bulges, and were identified by comparison to a previously published sequence for Cryptomonas sp. M1634.

Nuclear ITS2, partial LSU rDNA, and SSU rDNA, and chloroplast psbA and LSU rDNA sequences of cultured strains of the genus Cryptomonas from Korea were combined for phylogenetic analyses. Combined data were analyzed by using Bayesian and RAxML methods. The resulting phylogenetic tree is shown in Fig. 2. The combined sequence data analyzed in this study were 6,227 nucleotides for 77 strains of Cryptomonas. The average nucleotide frequencies of informative positions were 27.0% adenine, 20.1% cytosine, 27.0% guanine, and 25.9% thymine.

For phylogenetic analyses, 79 strains were used (Table 2). Forty-seven new Cryptomonas strains were included, and two species (Teleaulax acuta and Guillardia theta) were selected as outgroup. Our phylogenetic tree divided into fourteen species and six unidentified strains with high support values (pp = 1.00, ML = 100). Three unidentified strains branched off basally without close relatives, and C. obovoidea, C. commutata, C. erosa, and Cryptomonas sp. CNUCRY75 strain had strong support values (pp = 1.00, ML = 100). Next was C. marssonii with three strains, including a Korean CNUCRY4 strain. C. ovata, C. obovata, C. phaseolus, C. gyropyrenoidosa, C. paramecium, C. borealis, C. lundii, and Cryptomonas sp. Okgeum121810C were recovered as a monophyletic group with only a Bayesian support value (pp = 0.96). C. pyrenoidifera, C. tetrapyrenoidosa, and C. curvata were grouped together with high support values (pp = 1.00, ML = 98). The Cryptomonas sp. CNUCRY284 was closely related to 15 strains of C. curvata, with high supportives (pp = 1.00, ML = 100), while eight strains of C. tetrapyrenoidosa were sistered with 14 strains of C. pyrenoidifera (pp = 1.00, ML = 99).

The lengths of most nuclear ITS2 sequences of the examined strains were between 330 (Cryptomonas sp. CNUCRY284) and 590 nt (Cryptomonas sp. Sinjeongbangjuk080611A). Several features that were reported by Schultz et al. (2005) were also observed in the Cryptomonas nuclear ITS2 (Figs 3-9). For all ITS2 sequences, four-helix structures could be inferred with long third helices (Fig. 3). In all helices II, an unpaired U-U was found (Fig. 5). Three ‘UGGU’ motifs (UGGGU, UGG, GGU or similar) with conserved positions across species were found in helices III, but differed from the results of Schultz et al. (2005). Its position was downstream instead of upstream from the terminal loop (Figs 6-8).

Across species, the helices I and II were quiet variable in their terminal parts and also varied in length (Figs 4 & 5). The numbers of internal loops differed from zero (C. obovoidea, Cryptomonas sp. CNUCRY75, and Okgeum 121810C) to two (C. curvata and C. phaseolus) (Fig. 4).

[Fig. 2.] Bayesian tree of the nuclear internal transcribed spacer 2, partial large subunit ribosomal DNA (LSU rDNA), and small subunit rDNA, and chloroplast psbA and LSU rDNA combined data set. The Bayesian posterior probability and maximum likelihood bootstrap value are shown above or below the branches. Scale bar represents: substitution per site.

[Table 2.] Strains of the genus Cryptomonas used in this study and the GenBank accession numbers for their nr 5.8S-ITS2-LSU, nr SSU, pt psbA and pt LSU rDNA sequences

Strains of the genus Cryptomonas used in this study and the GenBank accession numbers for their nr 5.8S-ITS2-LSU, nr SSU, pt psbA and pt LSU rDNA sequences

[Fig. 3.] Putative secondary structure of the nuclear internal transcribed spacer 2 of the Cryptomonas obovoidea Songgock032611 (from 5’ to 3’ terminus in clockwise direction). In total a sequence of 352 nucleotides were submitted to the mfold server. The presumably correct structure with the four helix domains was found without applying any force options. Helices were numbered in Roman numerals.

The shortest helix I was found in C. obovoidea and Cryptomonas sp. Okgeum121810C, whereas the longest helix I was found in C. pyrenoidifera (Fig. 4). The shortest helix II was found in C. pyrenoidifera CNUCRY27, whereas the longest was found in C. obovata (Fig. 5). In helix II, the proximal parts and an internal loop with a U-U mismatch were highly conserved across species, whereas all were different in length, with various additional internal loops found in the distal parts (Fig. 5). Among the helix III (Figs 6-8), the shortest one was found in C. pyrenoidifera (Fig. 7). The longest helix III was found in Cryptomonas sp. Sinjeong080611A (Fig. 8). It averaged about twice the length of other helices, which also differed markedly between each other (Figs 6-8). Although remarkable differences in length were observed, the structures of the terminal parts were highly conserved. Close to the terminal loop, a highly conserved region across species was found, consisting of an alternating upstream CUCUCU motif paired with a GAGAGGA downstream motif, and included an internal loop (region IIIa in Figs 6-8). A combination of three internal loops with the two other loops protruding to the 5′ side and one symmetrical internal loop in the middle was also found in all helix III (IIIb in Figs 6-8). Two UGGU motifs were always found on the 3′ side opposite to the two outer loops of the IIIb part, whereas a third ‘UGGU’ motif was present upstream from the other two (arrows in Figs 6-8). In the proximal part, the helices were less conserved and showed various numbers and sizes of internal loops apart from one large internal loop which seemed to be present in all helices, but was not conserved in primary sequence across all clades (Figs 6-8). Cryptomonas sp. Yeonra042011B, Dumo2 100311C, and Sinjeong080611A had a branch structure at its middle part but IIIa and IIIb parts were conserved. The nuclear ITS2 sequences were generally not aligned across species in helix IV, which caused difficulty in predicting their structure (Fig. 9). The longest helix IV was found in C. gyropyrenoidosa. C. curvata Sojung032611and Cryptomonas sp. CNUCRY284 had identically short helices IV. Cryptomonas sp. CNUCRY284 strain showed sister relationships with C. curvata clade. Cryptomonas sp. Yeonra043011B and Dumo2 100310C strains included in the same clade was similar helices IV.

Cell shape and length, length of the fullow-gullet system, color and number of chloroplasts, and the presence / absence and number of pyrenoids were examined. The results are summarized in Table 3 and illustrated in Figs 10-12.

[Fig. 4.] Predicted secondary structures of nuclear internal transcribed spacer 2 helix I (clockwise from 5’ to 3’ termini). Parts of the sequences which differ among strains of the same species are labeled by rectangular boxes. In helix I, the proximal parts are marked in brackets and were highly conserved across strains.

[Fig. 5.] Predicted secondary structures of nuclear internal transcribed spacer 2 helix II (clockwise from 5’ to 3’ termini). Parts of the sequences which differ among strains of the same species are labeled by rectangular boxes. In helix II, the proximal parts marked in brackets were highly conserved across strains. In all helices, the conserved un-paird U-U motif was found.

[Fig. 6.] Secondary structure predictions of helix domains III of Cryptomonas curvata Sojung032611, C. gyropyrenoidosa CNUCRY146, C. marssonii CNUCRY4, C. obovoidea Songgock032611, C obovata Hanjeong080611A and C. ovata CNUCRY231 (clockwise from 5’ to 3’ termini). Region IIIa comprised the most conserved part of the nuclear internal transcribed spacer 2. The highly conserved internal three-loop structures of all IIIb regions were accompanied by three UGGU motifs (arrows) which were consistently observed across species.

[Fig. 7.] Secondary structure predictions of helix domains III of Cryptomonas phaseolus Gakgae 043011, C. pyrenoidifera CNUCRY138, C. tetrapyrenoidosa CNUCRY123, Cryptomonas sp. CNUCRY75, Cryptomonas sp. CNUCRY284, and Cryptomonas sp. Okgeum121810C (clockwise from 5’ to 3’ termini). Region IIIa comprised the most conserved part of the nuclear internal transcribed spacer 2. The highly conserved internal three-loop structures of all IIIb regions were accompanied by three UGGU motifs (arrows) which were consistently observed across species.

[Fig. 8.] Secondary structures predictions of helix domains III of Cryptomonas sp. Yeonra043011B, Cryptomonas sp. Dumo2 100311C, and Cryptomonas sp. Sinjeong080611A (clockwise from 5’ to 3’ termini). Region IIIa comprised the most conserved part of the nuclear internal transcribed spacer 2. The highly conserved internal three-loop structures of all IIIb regions were accompanied by three UGGU motifs (arrows) which were consistently observed across species. Helix III of the three strains differed from the helices of the other species, in that it has middle branch.

[Fig. 9.] Presumed secondary structures of helices IV of the nuclear internal transcribed spacer 2 (ITS2). This part of the nuclear ITS2 was the most variable region across species in the alignment. It was also difficult to define 5’ and 3’ termini of this helix.

[Fig. 10.] Light micrographs of genus Cryptomonas. (A & B) C. curvata Begelow S. (C & D) C. gyropyrenoidosa CNUCRY146. (E & F) C. marsonii CNUCRY4. (G & H) C. ovata CNUCRY231. C, contractile vacuole; Cp, chloroplast; E, ejectisome; F, flagellum; G, gullet; MO, maupas oval; Py, pyrenoid; S, starch. Scale bars represent: A-H, 10 μm.

[Table 3.] Morphological points of comparisons among Korean Cryptomonas species

Morphological points of comparisons among Korean Cryptomonas species

[Fig. 11.] Light micrographs of genus Cryptomonas. (A & B) C. obovata Saenae080611D. (C & D) C. obovata Hanjeong080611A. (E & F) C. obovoidea Songgock032611. (G & H) C. phaseolus Gakgae043011. C, contractile vacuole; Cp, chloroplast; E, ejectisome; F, flagellum; G, gullet; MO, maupas oval; Py, pyrenoid; S, starch. Scale bars represent: A-H, 10 μm.

Specimens examined. Sojungsoryuji, Namsan, Korea; Seohojeosuji, Suwon, Korea; Daepyeong swamp, Haman, Korea; Jjokji pond, Changyeong, Korea; 1100goji, Seogwipo, Korea; Chungsan, Taean, Korea; Gungnamji, Buyeo, Korea.

Light microscopy. Cryptomonas curvata was an ovoid-elliptical shape in the ventral view (Fig. 10A & B). The cells were 18-20 μm in length. The length of the gullet with rows of ejectisomes was 2/3 of cell length. Cells had two brown-chloroplasts, each with a pyrenoid. Contractile vacuoles were placed near the gullet. Two flagella were inserted in a vestibulum of the cell invagination.

Specimens examined. Jungmukji, Cheongyang, Korea.

Light microscopy. Cryptomonas gyropyrenoidosa was broad elliptical shape and was flattened in ventral view (Fig. 10C & D). The cells were 20-23 μm in length. The length of the gullet with rows of small ejectisomes was 1/2 of cell length. Cells had brown chloroplasts surrounded by many starch grains. Two red maupas ovals were located in the center of each cell.

Specimens examined. Daesong, Haman, Gyeongnam, Korea.

Light microscopy. Cryptomonas marssonii was a sigmoid shape to asymmetrical in ventral view (Fig. 10E & F). The cells were 20-23 μm in length. Cells had two brown chloroplasts and many starch grains around chloroplasts. Each chloroplast has a pyrenoid. A contractile vacuole was located in the anterior of the cell.

Specimens examined. Songcheonji, Iksan, Korea; Mukawa, Muroran, Hokkaido, Japan.

Light microscopy. Cryptomonas ovata has long-ovoid of elliptical shape in ventral view (Fig. 10G & H). The cells were 32-37 μm in length. The length of gullet with rows of ejectisomes was 3/5 that of cells. Cells had two brown chloroplasts surrounded by many starch grains. Contractile vacuole was located in apical cell and three maupas ovals were located in the center of the cell.

Specimens examined. Saenaebangjuk, Jeongeup, Korea; Hanjeongje, Jeongeup, Korea.

Light microscopy. Cryptomonas obovata had two marked cells which were characterized by presence / absence of pyrenoids (Fig. 11A-D). The cells were 26-36 μm in length. Cells had two brown chloroplasts, each with a terminal pyrenoid. The other cell had biconvex shape in right or left lateral view (Fig. 11C). The cells were 29-38 μm in length. Cells had two red-brown chloroplasts without pyrenoid. Both had ovoid shape in ventral view (Fig. 11B & D). The length of gullet with rows of ejectisomes was 2/3 or 3/4 that of cells. Contractile vacuole located in apical part of the cell and two maupas ovals located in the center of the cell. Two flagella inserted in a vestibulum of the cell invagination.

Specimens examined. Songgock, Nonsan, Korea; Mokmalji, Taean, Korea; Dongbackdongsan, Jeju, Korea.

Light microscopy. Cryptomonas obovoidea had ventral reflex shape in right or left lateral view and ovoid with broad end in ventral view (Fig. 11E & F). The cells were 11-14 μm in length. The length of gullet with rows of ejectisomes was 3/4 that of cells. Cells had two green-brown chloroplasts, each with a pyrenoid. Contractile vacuole located in apical cell. Maupas oval absent. Two flagella inserted in a vestibulum of the cell invagination.

Specimens examined. Angolji, Nonsan, Korea; Gakgaeji, Cheongdo, Korea; Jangdangji, Pohang, Korea; Daepyeong swamp, Haman, Korea; Jeongsan, Cheongyang, Korea.

Light microscopy. Cryptomonas phaseolus had biconvex shape in right or left lateral view and long ovoid with narrow end in ventral view (Fig. 11G & H). The cells were 12-16 μm in length. The length of gullet with rows of ejectisomes was 2/3 that of cells. Cells had two brown chloroplasts, each with a pyrenoid. The cell didn’t have maupas oval when observed in culture. Two flagella were inserted in a vestibulum of the cell invagination.

Specimens examined. Gyorae, Jeju, Korea; Yeongtapji, Daejeon, Korea; Wonseongje, Surok, Baeksan, Korea; Andeokjeosuji, Iksan, Korea; Soiji, Bisan, Eumseong, Korea; Gungnamji, Buyeo, Korea; Sumeunmulbaengdeui, Aewol, Jeju, Korea.

Light microscopy. Cryptomonas pyrenoidifera was biflattened in right or left lateral view and ovoid-elliptical shape in ventral view (Fig. 12A & B). The cells were 18-20 μm in length. The length of gullet with rows of ejectisomes was 2/3 that of cells. Cells had two brown-chloroplasts, each with an obvious pyrenoid. Contractile vacuole was placed at above two maupas ovals. Two flagella were located in a vestibulum of the cell invagination.

Specimens examined. Sinchonje, Naju, Korea; Duungseupji, Taean, Korea; Onsusoryuji, Goseong, Korea; Deokamji, Nonsan, Korea; Yeongildongji, Pohang, Korea; Hudongje, Cheongyang, Korea; Seohojeosuji, Suwon, Korea.

Light microscopy. Cryptomonas tetrapyrenoidosa was

[Fig. 12.] Light micrographs of genus Cryptomonas. (A & B) C. pyrenoidifera Sumeun041911. (C & D) C. tetrapyrenoidosa Onsu032611C. (E & F) Cryptomonas sp. Okgeum121810C. (G & H) Cryptomonas sp. CNUCRY284. (I & J) Cryptomonas sp. Dumo2 100310C. (K & L) Cryptomonas sp. Yeonra043011B. C, contractile vacuole; Cp, chloroplast; E, ejectisome; F, flagellum; G, gullet; MO, maupas oval; Py, pyrenoid; S, starch. Scale bars represent: A-L, 10 μm.

an ovoid shape in ventral view (Fig. 12C & D). The cells were 16-22 μm in length. The length of gullet with rows of ejectisomes was 2/3 that of cells. Cells had two brownchloroplasts, and each chloroplast had two or three pyrenoids facing each other in the left and right chloroplast lobes. Contractile vacuole was located in apical and two flagella inserted in a vestibulum of the cell invagination. Cells had two maupas ovals.

Specimens examined. Okgeumji, Iksan, Korea.

Light microscopy. Cryptomonas sp. Okgeumji121810C has biconvex shape with narrow ends in right or left lateral view and broad ovoid shape in ventral view (Fig. 12E & F). The cells were 13-17 μm in length. The length of gullet with rows of ejectisomes was 1/2 or 2/3 of cell length. Cells had a brown chloroplast with a pyrenoid. Two flagella inserted in a vestibulum of the cell invagination.

Specimens examined. Yeonhwaji, Jeju, Korea.

Light microscopy. Cryptomonas sp. CNUCRY284 had biconvex shape with narrow end in right or left lateral view and ovoid shape in ventral view (Fig. 12G & H). Cells were 18-21 μm in length. Length of gullet with rows of ejectisomes was 3/5 that of cells. Cells had a greenishbrown chloroplast with an obvious pyrenoid. Contractile vacuole above two maupas ovals. Two flagella inserted in a vestibulum of the cell invagination.

Specimens examined. Dumojeosuji, Jeju, Korea.

Light microscopy. Cryptomonas sp. Dumo2 100310C had biconvex with narrow end in right or left lateral view and ovoid shape in ventral view (Fig. 12I & J). The cells were 14-18 μm in length. The length of gullet with rows of ejectisomes was 1/2 or 3/4 of cell length. Cells had a brown chloroplasts with a pyrenoid. Contractile vacuole located in apical cell. Cells didn’t have maupas oval when we established it in culture. Two flagella located in a vestibulum of the cell invagination.

Specimens examined. Yeonra, Hwayangeup, Cheongdo, Gyeongbuk, Korea.

Light microscopy. Cryptomonas sp. Yeonra043011B had flattened shape in right of left lateral view and ovoid shape in ventral view (Fig. 12K & L). The cells were 14-18 μm in length. The length of gullet with rows of ejectisomes was 2/3 of cell length. Cells had a green-brown chloroplasts with a pyrenoid. The cell had two maupas ovals. Two flagella inserted in a vestibulum of the cell invagination.

Although cryptomonad genera were delimited by ultrastructral features, such as periplast and flagellar apparatus, all Cryptomonas species have been described exclusively by light microscopy (Huber-Pestalozzi 1950, Butcher 1967). Pringsheim (1944, 1968) established clonal cultures, and examined morphological characters of the genus Cryptomonas. He recognized that morphological characters cannot be used satisfactorily for delimitation of Cryptomonas species. Most species of the genus Cryptomonas from Korea are ovoid or sigmoid in shape, and cell length ranges from 11 to 38 μm. The ejectisomes were arranged along a gullet located at vestibular region of the cell. The length of gullet with rows of ejectisomes was 1/2-3/4 that of cells. All cells had two chloroplasts with variable color; reddish brown, brown, or green brown. However, the number of pyrenoids was 2 to 6 per cell, but C. obovata Hanjeong080611A has chloroplasts without pyrenoid. The maupas ovals were either present or absent. However, distinction based on morphological characters is unclear for species of the genus Cryptomonas, as suggested by Javornicky (2003) and Hoef-Emden (2007). For example, C. obovoidea had a sigmoid cell shape, but belonged to the species group with ovoid cell shape in our tree. Therefore, species with ovoid cell shape were not monophyletic. Other morphological characters, such as the presence / absence of pyrenoid, were found across clades. Saenae080611A strain of C. obovata clade has pyrenoid, but Hanjeong080611A strain lacked pyrenoid, in spite of being located in the same species clade. Incongruence between morphospecies concept and molecular phylogeny has been reported recently (Hoef-Emden and Melkonian 2003, Hoef-Emden 2007). Therefore, due to lack of clear species-specific morphological characters, new species may have to be defined by molecular signatures, as suggested by Hoef-Emden (2007).

In the molecular phylogenetic tree based on nuclear 18S rDNA, the genus Cryptomonas is known as monophyletic (Deane et al. 2002). Although nuclear 18S rDNA data has been used for resolving phylogenetic relationships among genera of Cryptophyceae (Marin et al. 1998, Deane et al. 2002), the tree did not show obvious resolution within the genus Cryptomonas. The phylogenetic relationships among Cryptomonas species were likely to involve fast-evolving genes which contain more phylogenetic information (Deane et al. 2002). Our phylogenetic tree based on combined nuclear SSU, partial LSU, and ITS2 and plastid psbA and LSU rDNA data have a topology similar to recently published phylogenies (Hoef-Emden 2007). However, our data further clarify the phylogenetic relationships among Cryptomonas species, but there remain areas of uncertainty. For example, the positions of some species (C. ovata, C. obovata, C. phaseolus, C. gyropyrenoidosa, C. paramecium, C. borealis, and C. lundii) were not clearly resolved in the tree. The topologies of our trees obtained by ML and Bayesian analyses are very similar, although the level of support for individual nodes varies considerably. The tree also differs slightly from a tree based on combined nuclear partial LSU and nucleomorph SSU rDNA data (Hoef-Emden et al. 2002, Hoef-Emden 2007). Especially our tree is better supported in internal nodes than that of the previous one (Hoef-Emden 2007).

The nuclear ITS2 in Cryptomonas species showed the typical conserved secondary structure with four helices which are found in most eukaryote lineages (Schultz et al. 2005). Molecular characters were derived from the nuclear ITS2; 1) the highly conserved regions were easily aligned among Cryptomonas species, 2) the less conserved regions provided enough signal to identify synapomorphies or species-specific unique combinations of sequences for clades (species), and 3) the highly variable regions were aligned within but not between species (Hoef-Emden and Melkonian 2003). Most conserved regions in all strains were located in the distal part of helix ？？？. The sequences of the distal parts were not very useful for species identification, but seem to be available to distinguish cryptomonad genera as molecular signatures. Compare to other eukaryotic ITS2 secondary structures, the UGGU motifs have shifted from the 5′ part to 3′ part of the helix ？？？ in previous studies, as well as in this study (Schultz et al. 2005, Hoef-Emden 2007).

The helix ？, ？？, and ？？？b were less conserved regions than helix ？？？a (the distal part of helix ？？？). The proximal part of helix ？ and ？？ had similar sequences among species, but were variable in their terminal parts. Specific structures of helix ？？？b appeared in all strains, despite the differences among sequences. Sequences of helix ？, ？？, and ？？？b in ITS2 secondary structure were very useful for the identification of Cryptomonas species. The helix ？V was varied from species to species because of sequence diversity. Through deduction of the secondary structure, molecular signatures to identify lower rank of genus or species could be determined.