Biology

ARTICLE IN PRESS

http://www.elsevier.de/protis Published online date 7 August 2006

1

Correspondin fax +81 29 853 e-mail iinouye 2Current addr Parkville, Victo

& 2006 Elsev doi:10.1016/j

157, 401—419, August 2006

Protist, Vol.

ORIGINAL PAPER

Hatena arenicola gen. et sp. nov., a Katablepharid Undergoing Probable Plastid Acquisition

Noriko Okamoto2, and Isao Inouye1

Graduate School of Life and Environmental Sciences, University of Tsukuba, 1-1-1, Tennodai, Tsukuba, Ibaraki 305-8572, Japan

Submitted February 27, 2006; Accepted May 27, 2006 Monitoring Editor: Robert A. Andersen

Hatena arenicola gen. et sp. nov., an enigmatic flagellate of the katablepharids, is described. It shows ultrastructural affinities to the katablepharids, including large and small ejectisomes, cell covering, and a feeding apparatus. Although molecular phylogenies of the 18S ribosomal DNA support its classification into the katablepharids, the cell is characterized by a dorsiventrally compressed cell shape and a crawling motion, both of which are unusual within this group. The most distinctive feature of Hatena arenicola is that it harbors a Nephroselmis symbiont. This symbiosis is distinct from previously reported cases of ongoing symbiosis in that the symbiont plastid is selectively enlarged, while other structures such as the mitochondria, Golgi body, cytoskeleton, and endomembrane system are degraded; the host and symbiont have developed a morphological association, i.e., the eyespot of the symbiont is always at the cell apex of Hatena arenicola; and only one daughter cell inherits the symbiont during cell division, resulting in a symbiont-bearing green cell and a symbiont- lacking colorless cell. Interestingly, the colorless cells have a feeding apparatus that corresponds to the location of the eyespot in symbiont-bearing cells, and they are able to feed on prey cells. This indicates that the morphology of the host depends on the presence or absence of the symbiont. These observations suggest that Hatena arenicola has a unique ‘‘half-plant, half-predator’’ life cycle; one cell divides into an autotrophic cell possessing a symbiotic Nephroselmis species, and a symbiont-lacking colorless cell, which later develops a feeding apparatus de novo. The evolutionary implications of Hatena arenicola as an intermediate step in plastid acquisition are discussed in the context of other examples of ongoing endosymbioses in dinoflagellates. & 2006 Elsevier GmbH. All rights reserved.

Key words: Hatena arenicola; Katablepharidophyta/Kathablepharida; Nephroselmis symbiont; plant evolution; plastid acquisition via secondary endosymbiosis; ultrastructure.

Abbreviations: EM ¼ electron microscopy; ER ¼ endoplasmic reticulum; ICBN ¼ International Code of Botanical Nomenclature; ICZN ¼ International Code of Zoological Nomenclature; LM ¼ light mi- croscopy; SEM ¼ scanning electron microscopy; SSU rDNA ¼ small subunit ribosomal DNA; TEM ¼ transmission electron microscopy.

g author; 4533

@sakura.cc.tsukuba.ac.jp (I. Inouye). ess: School of Botany, University of Melbourne, ria, Australia.

ier GmbH. All rights reserved. .protis.2006.05.011

ARTICLE IN PRESS

402 N. Okamoto and I. Inouye

Introduction

Eukaryotes are currently classified into five or six supergroups (Baldauf et al. 2000; Baldauf 2003; Bapteste et al. 2002; Nozaki et al. 2003; Simpson and Roger 2002), and eukaryotic autotrophs (e.g., plants and algae) randomly scatter across those supergroups. Eukaryotic autotrophs comprise nine distinct divisions in cell architecture, and this enormous diversity is explained by several en- dosymbiotic events (Bhattacharya et al. 2004; Falkowski et al. 2004; McFadden 2001). It is widely accepted that a primary endosymbiosis between a eukaryote and a cyanobacterial sym- biont gave rise to the three extant primary eukaryotic autotrophs, Glaucophyta, Rhodophyta, and Viridiplantae ( ¼ land plants plus green algae) (see Marin et al. (2005) for an alternative primary endosymbiosis). Subsequently, secondary endo- symbioses occurred between green or red algae and heterotrophic eukaryotic hosts. Two algal divisions (Euglenophyta and Chlorarachniophyta) acquired the plastids of green algae, while four algal divisions (Heterokontophyta, Haptophyta, Cryptophyta, and Dinophyta) and one parasitic phylum (Apicomplexa) acquired those of red algae (although some Dinophyta lost their original plastid and remained colorless or re-acquired different plastids as discussed below). An esti- mated two-thirds of today’s algal diversity resulted from secondary endosymbioses (Falkowski et al. 2004; Graham and Wilcox 2000), and thus this process is important in understanding the evolu- tionary process of plant and algal diversification.

The transition of a symbiont to a plastid involves a series of changes in both the host and the symbiont (Cavalier-Smith 2003; Hashimoto 2005; van der Giezen et al. 2003), which include the establishment of a specific partner alga, lateral gene transfer from the symbiont to the host’s nucleus (Katz 2002), the development of protein- transport machinery to carry proteins from the host cytoplasm to the symbiont (van Dooren et al. 2001), and synchronization of cell cycles so that the symbiont can be passed to host daughter cells during host cell division.

Evidence about plastid integration is accumu- lating (Andersson and Roger 2002; Archibald et al. 2003; Hackett et al. 2004b; Huang et al. 2003; Martin and Herrmann 1998; Martin et al. 2002; Martin 2003a, b; Nozaki et al. 2004; Stegemann et al. 2003), however, the intermediate steps in this process remain largely unknown. Some organisms appear to be in an intermediate stage of plastid acquisition, the best-known examples of which

are the Cryptophyta and Chlorarachniophyta, whose plastids contain a vestige of the symbiont nucleus termed a nucleomorph (e.g. Douglas et al. 2001; Gilson et al., 2006). They are thought to represent a late stage of integration. Early stages of plastid acquisition can be found in the dinoflagellates (for reviews, Hackett et al. 2004a; Morden and Sherwood 2002; Schnepf and Elbrächter 1999), where the most dramatic changes are ongoing. The original plastids of dinoflagellates have been of red algal origin, though some dinoflagellates subsequently lost their original red-algal plastids, which were re- placed by new ones via extra secondary or tertiary endosymbioses. These examples probably reflect stepwise changes in symbiotic conditions during integration (e.g. Hackett et al. 2004a), and are useful to understand the plastid acquisition process.

We discovered an undescribed flagellate, Hate- na arenicola gen. et sp. nov., in October 2000, in an intertidal sandy beach in Japan. The organism appears to be in the process of plastid acquisition. Most cells of H. arenicola in the natural population have a green plastid-like structure with a red eyespot at the cell apex, though it is inherited by only one of the daughter cells during cytokinesis (Okamoto and Inouye 2005a). Molecular phyloge- netic analysis of small subunit ribosomal DNA (SSU rDNA) and ultrastructural observations of the plastid-like structure reveal that it is not a plastid but an autotrophic endosymbiont belonging to the genus Nephroselmis Stein (Prasinophyceae, Vir- idiplantae). We previously reported the symbiotic nature of this association (Okamoto and Inouye 2005a). This paper describes the organism as a new genus and species of katablepharid, a group of flagellates recently designated the phylum Kathablepharida, division Katablepharidophyta (Okamoto and Inouye 2005b). We compare the symbiosis of H. arenicola with other examples of secondary symbioses in dinoflagellates to help elucidate the intermediate steps in the plastid acquisition process.

Results

Description

Hatena arenicola Okamoto et Inouye gen. et sp. nov.

ARTICLE IN PRESS

403Hatena arenicola: Halfway to a Plant?

Latin Diagnosis

Cellulae oblongae secus axem dorsiventrem valde appresae, sine chromatophoro nec vacuola con- tractili; 30-40 mm longae; 15-20 mm latae; ventraliter subapicali cum sulco vadoso longitudinali 3-4 mm longa et ca. 2 mm lata; flagellis crassis binis inaequalibus in sulco insertis; flagello anteriore longiore, altero posteriore breviore; ejectisomis conspicuis distichis prope flagellas longitudinaliter positis; plerumque cum 1-4 endosymbiontis viridis; uni stigma endosymbionti situm ad apicem cellulae.

Holotype: Figure 1A Type locality: Isonoura, Wakayama, Japan (Fig.

1 B-C) Etymology:

Hatena ¼ ‘enigmatic’ in Japanese arenicola ¼ ‘inhabiting sand’ in Latin

Light Microscopy

General Morphology: The cell is flattened along the dorsiventral axis. In the ventral view, it is ovoid, 30-40 mm long and 15-20 mm wide (Fig. 1 A, D-F).

Figure 1. Hatena arenicola gen. et sp. nov. A. Ventral v and an eyespot of the symbiont (arrowhead). B,C. Sam showing two rows of conspicuous Type I ejectisomes ‘‘immature’’ symbiont. G-L. Cell division in Hatena arenic symbiont. Each panel shows a different individual at a di The scale bar is 10 mm in A, D-L.

The cell has a furrow in the subapical region, 3- 4 mm long and ca. 2 mm wide (Fig. 1 D). The long anterior flagellum and shorter posterior flagellum emerge from this furrow, and two rows of ejectisomes are easily visible near the posterior end of the furrow (Fig. 1 D). One large nucleus is located in the middle posterior region of the cell, and the rest of the cytoplasm is mostly occupied by the plastid of the green symbiont. Cells only rarely lacked the symbiont (Fig. 1 E), though some symbionts were not fully developed (Fig. 1 F; see Discussion).

Cell Division: During cell division, one daughter cell inherits the Nephroselmis symbiont while the other does not and becomes colorless. Figure 1 G-L shows cell division in H. arenicola (ventral view). First, the host nucleus moves to the apex between the flagellar insertion and the eyespot (Fig. 1 G). The symbiont contracts to the left side of the host cell (on the viewer’s right in figures) so that the left half of the cell remains green, while the right half becomes colorless (Fig. 1 G). Two new flagella are formed, and one set moves from the right side of the nucleus to the left (flagellar transformation;

iew of a symbiont-bearing cell showing two flagella pling site. D. The same cell in a different focal plane, . E. A cell lacking the symbiont. F. A cell with an ola, where the arrowhead indicates an eyespot of the fferent stage in cell division. N: nucleus. S: Symbiont.

ARTICLE IN PRESS

Figure 2. Hatena arenicola and its symbiont. DIC images are shown in upper column (A, C, E, G) and the fluorescent images of the same cells are shown in lower column (B, D, F, H respectively). Arrow- heads indicate the eyespot. Blue: DAPI-stained nuclei of Hatena arenicola (large fluorescence in the center of the cell) and of the symbiont (smaller dots). Red: Autofluorescence of the symbiont plastid. The scale bar is 10 mm.

0

0.2

0.6

1

a b so

rp tio

n

450 500 550 600 650 700 wavelength

Symbiont of

Hatena arenicola

Chl. b

Chl. b

Chl. a

Chl. a

Figure 3. Microspectrophotometry of the symbiont plastid. Spectrogram shows absorbance similar to chlorophyll a/b-containing plastids.

404 N. Okamoto and I. Inouye

Fig. 1 H). The chromosomes separate (Fig. 1 I). Following nuclear division (Fig. 1 J), cytokinesis results in one green cell with the symbiont and one colorless cell without it (Fig. 1 K-L).

Fluorescence Microscopy: Figure 2 A-H shows DIC and fluorescence images of the same cells. DAPI label (blue) indicates a large host nucleus in the cell center and one to four smaller symbiont nuclei (Fig. 2 B,D,F,H). Each symbiont nucleus is independent of all others and is surrounded by plastid(s), shown in red (due to autofluorescence). Interestingly, the cell with multiple symbiont nuclei has only a single eyespot at the apex of the host cell (Fig. 2 E-G; arrowheads).

Microphotometry: Microphotometry of seven cells shows an absorption pattern characteristic of plastids with chlorophyll a/b. Average absorp- tion is shown in Figure 3. The prominent peaks at 435 and 678 nm represent chlorophyll a absorp- tion, while the smaller peaks at 470 and 650 nm correspond to chlorophyll b absorption.

Uptake of Prey Cells and Symbiont Specifici- ty: Molecular phylogenetic analysis of 16S rDNA

indicates that the symbiont is a member of Nephroselmis (Okamoto and Inouye 2005a). In feeding experiments using a Nephroselmis strain (NIES1417; different from that of the symbiont in 16S rDNA sequence; data not shown), colorless cells of H. arenicola phagocytotically engulfed the alga (Fig. 4 A-I) and tentatively maintained it. However, none developed fully, likely because the strain used in the experiments was not the exact symbiont of H. arenicola. This suggests that symbiont specificity is at the species or strain level. As the H. arenicola cells which engulfed Nephroselmis NIES1417 died, it is unclear whether the Nephroselmis cells were digested.

Crawling Motion: Hatena arenicola displays a conspicuous crawling motion that makes it easy to recognize this organism in a crude sample. Figure 5 A-H illustrates cell motion and flagellar movement. The anterior flagellum produces most of the propulsion, while the posterior flagellum is used to change direction. To move forward, a cell casts the anterior flagellum, which adheres to the substratum by its tip (Fig. 5 A-C), and pulls itself forward (Fig. 5 D, E). It then anchors itself with the posterior flagellum (Fig. 5 F-H), and repeats the process. The tip of the flagellum is sharply bent while it is attached to the substratum, as shown in the supplemental material and in Figure 5 I. The cell bends the posterior flagellum to change direction (not shown).

Electron Microscopy

Electron microscopy (EM) revealed that the green plastid-like structure in the cell is a eukaryotic

ARTICLE IN PRESS

Figure 4. Uptake of Nephroselmis (NIES1417) by Hatena arenicola. A-H were taken at 6-s intervals. I: corresponds to Hatena arenicola and the symbiont of each frame.

Figure 5. Crawling motion of Hatena arenicola. High-speed video images recorded at 0 ms (A), 60 ms (B), 100 ms (C), 130 ms (D), 220 ms (E), 280 ms (F), 300 ms (G), 340 ms (H), respectively. I. SEM image showing the anterior flagellum sharply bent at the distal end (arrowhead). The scale bar is 10 mm in I.

405Hatena arenicola: Halfway to a Plant?

endosymbiont. A single nucleus, usually dorsiven- trally flattened, is located in the middle posterior region of the cell, and its chromatin is always condensed and electron dense (Fig. 6 A,B). Multi- ple mitochondrial profiles are present throughout the cytoplasm, though these could be different sections of a single, large reticulate mitochon- drion. Mitochondrial cristae are tubular (Fig. 6 C), and there is a single large Golgi body near the groove of flagellar insertion between the nucleus and the flagellar apparatus (Fig. 6 D).

Ejectisomes: There are two types of ejecti- somes beneath the membrane: large, Type I ejectisomes sensu Vørs 1992b (Fig. 6 D-E) that are arrayed in two rows near the flagellar insertion (Figs 1 D, 6 F-G); and smaller Type II ejectisomes

sensu Vørs 1992b (Fig. 6 H-I) that are arrayed in numerous rows all over the cell (Fig. 6 F-G). Each ejectisome consists of a coiled ribbon with a gradual depression (Fig. 6 D, H), and bound by a single membrane. Discharged ribbons are slightly curved (Fig. 6 J; type I ejectisome). The ribbon does not have any kink, unlike those of Cryptophyta.

Type II ejectisomes are situated beneath the plasma membrane of the cell (Fig. 7 A,B) except for a small smooth area in the apical region, below which the eyespot is situated (asterisk in Fig. 6 G). The ejectisomes are spaced regularly between longitudinally oriented cytoskeletal microtubular bundles (arrowheads in Fig. 7 A).

Hatena arenicola lacks Type III ejectisomes that are characteristic of the genus Leucocryptos (Vørs 1992b).

Cellular and Flagellar Covering: The plasma membrane of each cell has a characteristic cellular covering composed of a thick inner basal layer (asterisk in Fig. 7 B) and an outer layer of electron-opaque material (arrowhead in Fig. 7 B). This structure extends to the flagellar surface. The outer layer comprises a regularly arrayed and spiraling envelope around the whole flagellum (Fig. 7 C).

Endoplasmic Reticulum: The endoplasmic re- ticulum (ER) does not form a conspicuous stack but is loosely distributed throughout the cell. The rough ER extends beneath the surface of the cell (double arrow in Fig. 7 B).

Flagella and Basal Bodies: The long anterior and shorter posterior flagella emerge from a

ARTICLE IN PRESS

Figure 6. Ultrastructure of Hatena arenicola. A. A longitudinal section of a Hatena arenicola cell showing a nucleus (N). Most of the nuclear content is condensed. The rest of the cytoplasm is occupied by the symbiont (Sym). The symbiont has a large plastid region with multiple pyrenoids. B. A transverse section of Hatena arenicola showing host nucleus (N) and the symbiont (Sym) with the vestigial cytoplasm (asterisk). C. Mitochondrial profiles showing tubular cristae. D. A Golgi body near the flagellar basal bodies (arrows) and Type I ejectisomes sensu Vørs 1992 (nearly longitudinal view). E. A transverse section of a Type I ejectisome (sensu Vørs 1992). F. SEM image of Hatena arenicola showing Type I ejectisomes near the flagellar insertion as well as smaller ejectisomes (Type II ejectisomes sensu Vørs 1992) regularly arrayed over the cell surface. G. Magnified SEM image of the same cell. Note that the apical region of the cell (asterisk) lacks ejectisomes. H. A longitudinal section of Type II ejectisome. I. A transverse section of Type II ejectisome. J. A whole mount TEM image of a discharged type I ejectisome. The scale bar is 10 mm in A, E; 2 mm in B, G; 50 nm in C; 500 nm in D, F; 1 mm in I; 200 nm in H, I.

406 N. Okamoto and I. Inouye

ARTICLE IN PRESS

Figure 7. Surface structure and flagellar transition region of Hatena arenicola. A. A tangential section of the surface of Hatena arenicola. Cytoskeletal micro- tubular bundles (arrows) are arrayed longitudinally below the cell surface; Type II ejectisomes are situated between them. Each pore corresponds to the position of an ejectisome, where the surface sheath is thin. B. A transverse section of the cell periphery showing a bilayered surface sheath. An arrowhead indicates the outer layer. An asterisk indicates the thick basal layer. An arrow indicates the plasma membrane. A double arrowhead indi- cates a profile of rough ER. C. A tangential section of the flagellum shows the outer layer of the surface sheath enveloping the flagellum in a spiraling fashion. D-J. Flagellar transition region of Hatena arenicola. D. Diagram of flagellar transition region reconstructed based on serial ultrathin sections. Each letter in D indicates which part corresponds to one of the transverse sections (E-J). The scale bar is 500 nm in A, C; 200 nm in B, E-J.

407Hatena arenicola: Halfway to a Plant?

shallow subapical furrow. The flagella are coated by a ‘‘surface sheath’’. The diagram of the basal body and the flagellar transition zone (Fig. 7 D) is based on serial sections of the flagellar transition zone (Fig. 7 E-J). At the proximal end of the basal body, there is a cartwheel structure with a central tube. The triplet microtubules of the basal body terminate below the plasma membrane (Fig. 7 E), whereas the doublet microtubules attach to the plasma membrane by a connecting fiber (Fig. 7 F). The flagellar transition region extends above the plasma membrane, and an electron-dense rod

structure is present (Fig. 7 G) in the middle of this region. Surrounding the transition region are outer doublet microtubules lined in a loose, electron- dense material (Fig. 7 H). The flagellar transition region ends in a terminal plate with electron- opaque material at the center (Fig. 7 I). The axonemal central pair begins above the terminal plate (Fig. 7 J). Feeding Apparatus: As reported in Okamoto and Inouye (2005a), colorless cells lacking the symbiont have a complex feeding apparatus at their apex (Fig. 8 A-C), consisting of transverse tubular rings (arrowheads in Fig. 8 B-C) and longitudinal microtubules arrayed in a single layer (arrow in Fig. 8 B-C). These microtubules are different from those that form the cytoskeleton (double arrowheads in Fig. 8 A). Inside the micro- tubular skeletons of the feeding apparatus are several electron-opaque granules, some of which are large and elongate (light gray) and others that are smaller, granulated, and pigmented (Fig. 8 A, C). These granules are restricted to the feeding apparatus, never seen elsewhere in the cell. Symbiont-bearing cells do not have a feeding structure; the corresponding region is occupied by the eyespot of the symbiont (Fig. 8 D-E). Symbiont: The symbionht retains not only a plastid but also its own cytoplasm with a nucleus and mitochondrion(a) (Fig. 9 A-C). Most symbiont cells contain one nucleus, which is often attached to the membrane, and symbiont and host nuclei often face each other (Fig. 9 A-C). The symbiont is bounded by a single membrane (arrowhead in Fig. 9 D), whose origin is unknown. Mitochondria have flat, often degraded cristae, though the extent of degradation varies among individual cells (Fig. 9 E-F). The plastid, which is the largest structure in the symbiont (Fig. 6 A-B), contains multiple pyrenoids bounded by a thin starch sheath (Fig. 9 G). The pyrenoid has shallow invaginations of the thylakoid membrane. The plastid has a single conspicuous eyespot, where the morphological association between host and symbiont is present (see below).

Free ribosomes are densely distributed through- out the symbiont cytoplasm, though no ribosome- bearing membrane (rough ER) is present. Occa- sionally there are flattened, stacked membranes next to the nucleus (Fig. 9 A). This structure would normally be a Golgi body but it is present in a degraded or inactive form, because no Golgi vesicles were seen around the structure. Some randomly shaped vacuoles of unknown origin are also present (Fig. 9 B-C). Because each cell division will result in half the population carrying

ARTICLE IN PRESS

Figure 8. Feeding apparatus of Hatena arenicola. A. A nearly transverse section of the feeding apparatus. Feeding apparatus is distinct from cytoskeletal microtubules (double arrowheads). B. A magnified view showing microtubules (arrow) regularly arrayed in a single layer along the external side of the tubular rings (arrowheads). C. Longitudinal section of the feeding apparatus showing numerous transverse tubular rings (arrowhead) and longitudinal microtubules arrayed in a single layer (arrows). D. An eyespot (e) of symbiont is located at the corresponding place in the symbiont bearing cell. E. A schematic illustration of the feeding apparatus and the corresponding place of the symbiont-bearing Hatena arenicola. The scale bar is 500 nm in A, C-D; 250 nm in B.

408 N. Okamoto and I. Inouye

the symbiont, these morphological varieties likely reflect degradation (see Discussion).

Cytoskeletal structures, including the flagella, basal bodies, the flagellar apparatus, and micro- tubular rootlets, are completely absent. These morphological changes must affect intracellular functions, such as protein synthesis and distribu- tion in the symbiont (see Discussion).

The lysosome of the host cytosol is discontinuous with the symbiont compartment. The lysosome of some cells contains scales of Nephroselmis and Pyramimonas (Fig. 10 A,B), indicating that H. arenicola cells engulf other prey in addition to its Nephroselmis partner. Because Pyramimonas cells are digested, H. arenicola may be partly heterotrophic (see Discussion).

Eyespot: The eyespot is composed of a single- layered sheet of osmiophilic granules (Fig. 11 A) that connects to the inner plastid envelope. Near the eyespot, the plastid, symbiont, and host plasma membranes are tightly layered (Fig. 11 A,B). In some cells, single microtubules are aligned longitudinally between the plasma membrane and the symbiont

membrane overlying the eyespot region (Mts in Fig. 11 C,D). These single microtubules are distinct from the cytoskeletal microtubular bundles (arrow- head in Fig. 11 D), and the space between them lacks Type II ejectisomes; this is consistent with the smooth surface appearance of the eyespot region (Fig. 6 F).

Molecular Phylogeny

Partial SSU rDNA sequences of H. arenicola (AB212285) aligned with the homologues of known eukaryotes were subjected to phylogenetic analyses. The resulting maximum likelihood (ML) tree (Fig. 12) showed the typical topology of the SSU rDNA tree, and H. arenicola was included in the katablepharid clade, which was robustly supported with high bootstrap probability in ML (99%), NJ (100%), and MP (99%) analyses.

Discussion

Hatena arenicola has several unique features that suggest it is in the process of endosymbiosis with

ARTICLE IN PRESS

Figure 10. Lysosome of Hatena arenicola with scales of prasinophytes characteristic to Pyramimo- nas (A) and Nephroselmis (B) respectively. The scale bar is 500 nm.

Figure 9. Ultrastructure of the symbiont. A-C. The symbiont cytoplasm, retaining a nucleus, mitochondria, and sometimes a Golgi body-like vesicle (A) or membranes of random shape (B-C), likely in an intermediate state of integration. D. The single symbiont-enveloping membrane (arrowhead) separates the symbiont compartment (Sym) and host cytoplasm (H). Double membranes of the symbiont plastid are also shown (arrows). E. Mitochondrial profiles that retain flat cristae. F. A relatively degenerated mitochondrial profile. G. A pyrenoid, surrounded by a starch sheath. Random shallow invagination of the thylakoids. The scale bar is 1 mm in A-C; 250 nm in D-G.

409Hatena arenicola: Halfway to a Plant?

a Nephroselmis partner (Okamoto and Inouye 2005a). We will discuss the taxonomy and classification of H. arenicola first, and then focus on its unique endosymbiosis.

Taxonomy

Morphological and molecular analyses of H. arenicola clearly show that it belongs to the recently established division/phylum Katablephar- idophyta (International Code of Botanical Nomen- clature, ICBN)/Kathablepharida (International Code of Zoological Nomenclature, ICZN) (Oka- moto and Inouye 2005b). The katablepharids comprise an ultrastructurally well-defined and small group of heterotrophic flagellates that includes 10 species in two genera: nine species of Katablepharis Skuja (correct spelling in ICBN which will be used in this paper)/Kathablepharis Skuja (original spelling in the ICZN), and one species of Leucocryptos (Braarud) Butcher. Kata- blepharidaceae (ICBN) was originally described by Skuja (1939) based on ovate or cylindrically ovate cell shape, two flagella emerging from a subapical depression, and conspicuous ejectisomes aligned

ARTICLE IN PRESS

Figure 11. Eyespot of the symbiont plastid. A. A longitudinal section of the eyespot (E). B. A magnified view of another cell clearly shows the eyespot granules, the inner/outer envelop of the symbiont plastid (arrows), the single symbiont enveloping membrane (arrowhead), and the host plasma membrane (double arrowhead) associated with each other. C-D. Tangential sections of the eyespot region show single microtubules (Mts) distinctive from the cytoskeletal microtubules (arrowheads) longitudinally situated between the eyespot (E) and the plasma membrane. The scale bar is 250 nm in A; 100 nm in B; 1 mm in C-D.

410 N. Okamoto and I. Inouye

in two rows near the flagellar insertion. Vørs (1992b), and Clay and Kugrens (1999b) emended the family by adding the following ultrastructural features: the entire surface of the cell, including the flagella, is coated with a bilayered surface sheath that appears to form spiraling rows around the cell body; tubular mitochondrial cristae; a complex, truncated conical feeding apparatus and cytoskeleton; a Golgi apparatus situated anteriorly and a centrally located nucleus; and a food vacuole in the posterior part of the cell.

Hatena arenicola shares all these characters, except that the cell is dorsiventrally compressed and the food vacuole is absent. We occasionally observed a vacuole containing scales of prasino- phytes, though its position was anterior to the nucleus. In addition, the flagellar transition zone containing a rod-shaped structure is fundamen- tally the same as that of K. ovalis (Lee et al. 1992). Based on these ultrastructural similarities and the

molecular phylogenetic data, H. arenicola un- doubtedly belongs to the katablepharids.

Currently, all the katablepharids belong to a single family, Katablepharidaceae, whose cells are defined as either ‘‘oblong or cylindrically ovate’’ (Katablepharis; Clay and Kugrens, 1999b; Vørs, 1992b) or ‘‘ovate, pyriform elliptical outline’’ (Leucocryptos; Butcher 1967). The most important distinguishing feature of Leucocryptos is the presence of Type III ejectisomes, which have been found only in Leucocryptos. Because H. arenicola lacks Type III ejectisomes, it does not belong to Leucocryptos.

Hatena arenicola is distinct from all other katablepharids previously described in that it is dorsiventrally compressed with a flat-oval shape, which suggests that it does not belong to the genus Katablepharis. Its crawling motion and the feeding apparatus composed of single-layered microtubules are also distinctive from the other

ARTICLE IN PRESS

0.1

Chlorarachnion CCMP242 Cercomonas longicauda

Heteromita globosa Thaumatomonas seravini

Euglypha rotunda Paulinella chromatophora

Ochromonas danica Phytophthora megasperma

Skeletonema pseudocostatum Pteridomonas danica

Prorocentrum micans Prorocentrum mexicanum

Gymnodinium sp. MUCC284 Pfiesteria sp. B112456

Alexandrium minutum Cryptosporidium parvum

Toxoplasma gondii Prorodon teres

Platyophrya vorax Emiliania huxleyi

Pavlova salina Chlorokybus atmophyticus

Mesostigma viride Arabidopsis thaliana

Fossombronia pusilla Pyramimonas propulsa

Ulothrix zonata 'Chlorella' ellipsoidea Tetraselmis striata

Coleochaete scutata Closterium littorale Cyanophora paradoxa

Cyanoptyche gloeocystis Glaucocystis nostochinearum

Gloeochaete wittrockiana Goniomonas truncata

Rhodomonas mariana Hanusia phi

Geminigera cryophila Cryptomonas ovata

Chroomonas sp. M1318 Hemiselmis brunnescens

Leukocryptos marina Hatena arenicola

Katablepharis japonica Heterophrys marina

Chlamydaster sterni Raphidiophrys ambigua

Rhodella maculata Stylonema alsidii

Bangia sp. Porphyra umbilicalis

Acanthamoeba castellanii Dictyostelium discoideum

Leptomyxa reticulata Hartmannella vermiformis

Scutellospora cerradensis Pneumocystis carinii

Saccharomyces cerevisiae Schizosaccharomyces pombe

Basidiobolus haptosporus Chytriomyces hyalinus

Monosiga brevicollis Clathrina cerebrum

Cirripathes lutkeni

18S rDNA (65species 1252 sites)

100/100/99

99/83/96

64/73/62

54/86/60

58/73/- 52/88/72

53/98/67

100/89/100

100/100/100

99/100/100

99/88/56

56/67/-

100/100/100

90/97/93

86/86/58

54/61/80

100/100/100

-/73/-

100/100/100

58/63/-

99/99/97

67/72/89

72/54/74

99/100/99

95/76/81

76/-/-

100/100/100

82/85/78

75/83/74

91/79/84

100/100/100

75/88/66

99/97/83

75/64/-

65/85/89

74/-/-

76/81/84

93/-/-

-/-/54

53/-/-

50/59/-

92/77/79

ML/NJ/MP

Figure 12. Unrooted eukaryotic tree based on the SSU rDNA. The best tree of the ML method is shown. Bootstrap propotions are shown at the internal branches, in the order of ML/NJ/MP methods. The length of each branch is proportional to the estimated number of substitutions. Bar denotes 10% substitutions per site. For details of the phylogenetic reconstruction methods, see text. The organisms included in the tree are listed in Table 1. Unambiguously aligned 1252 nucleotide positions were used for the analysis.

411Hatena arenicola: Halfway to a Plant?

katablepharids (Clay and Kugrens 1999a, b; Kugrens et al. 1994; Lee and Kugrens 1992; Lee et al. 1991; Okamoto and Inouye 2005b; Vørs 1992a, b). Based on those features, we propose that the organism should be assigned to a new genus, Hatena.

The family-level taxonomy of the katablepharids is still unclear, primarily because little molecular sequence data exist. Until further studies eluci- date katablepharid taxonomy, it is best to include the genus Hatena in the family Katablepharida- ceae.

ARTICLE IN PRESS

AH

Loss of Feeding Appar atus

Plastid: Enlarged Symtiont: Degraded

412 N. Okamoto and I. Inouye

Endosymbiosis

Endosymbiosis is a major driving force in plant evolution, and thus it is important to understand this process. Hatena arenicola may be an im- portant model of early plastid acquisition. Sym- biotic Nephroselmis differs from free-living individuals in having enlarged plastids with a greater number of pyrenoids, degraded subcellu- lar structures, and morphologically distinct eye- spots. Cells of known Nephroselmis species are a maximum of 20 mm in length (Nephroselmis astigmatica; Inouye and Pienaar 1984) and pos- sess a single plastid with a single pyrenoid. The symbiont occupies most of the host cytoplasm, suggesting that the symbiont plastid(s) grows more than ten fold after being engulfed by the host. Pyrenoids also multiply after being engulfed. In contrast, the symbiont cytoplasm loses other major cell components including flagellar appara- tus and microtubular roots, endomembranes such as the ER and transport vesicles, Golgi-like vesicles, and amorphous membranous structures. The dramatic growth of the plastid is in stark contrast to the degradation of the other orga- nelles. Because the cytoplasm is in such a degraded state, it must be difficult to sustain the growth and maintenance of the plastid alone. It is likely that some metabolites from the host cell are used to develop and maintain the plastid.

B

DE

F

G C

Predator phase Plant phase

Formation of Feeding apparatus

Uptake of the partner

Figure 13. Half-plant, half-predator-hypothesis. The solid line indicates a witnessed process, and the broken line indicates a hypothetical process. A-D: A green cell with the symbiont, lacking the feeding apparatus (A) divides (B) into one green (C) and one colorless cell (D). E-G: The colorless cell should form a feeding apparatus de novo and engulfs a Ne- phroselmis cell. G-H: The symbiont plastid selec- tively grows in the host cytoplasm. Because cell division of a colorless cell or a cell with an ‘‘immature’’ symbiont (H) has never been observed, uptake and the subsequent changes in both host and symbiont apparently occur within one generation.

Eyespot Morphology

The morphology of the eyespot is most suggestive of a host-symbiont coordination. The tight layering of the four distinct membranes of different origin implies functional cooperation. The eyespot is an important component of a photo-sensing complex found in various algal groups (Melkonian and Robenek 1984; Gualtieri 2001). It effectively regulates the light received by a photoreceptor located on a nearby membrane, thereby allowing the alga to detect light direction. Preliminary observations have shown that lateral, and not vertical, incidence, is important for the behavior of H. arenicola. Hence, we speculate that the photoreceptor and the regulatory mechanism would be laterally aligned in the cell, so that the eyespot can effectively shade the photorecepter from the lateral incidence. Assuming that the eyespot is functional in H. arenicola, the photo- receptor should be situated in either of the outer or inner plastid membrane, the single endosymbiont envelope, or the plasma membrane. Their mor- phological association can be explained as a

consequence of functional collaboration. Because H. arenicola crawls two-dimensionally, a photo- tactic response to laterally projected light is plausible. To test whether a functional association exists, further investigation of the threshold and the efficiency of phototaxis in both colored and colorless H. arenicola cells are required.

Morphological Changes of the Host and ‘‘Half-plant, half-predator’’ Model

Hatena arenicola cells without a symbiont have a complex feeding apparatus in place of an eye- spot, indicating that symbiont acquisition is accompanied by drastic morphological changes in both the symbiont and the host.

ARTICLE IN PRESS

413Hatena arenicola: Halfway to a Plant?

Such structural changes would necessitate life cycle changes. Based on observations presented in this paper, we propose the ‘‘half-plant, half- predator’’ hypothesis, where H. arenicola switches its lifestyle between that of a plant and a predator (Fig. 13; see also Okamoto and Inouye 2005a for detailed explanation). This proposed life cycle is also supported by the observation that the extent of degradation of symbiont mitochondria and membranous structures varies among individual cells. The presence of prasinophyte scales in a H. arenicola lysosome suggests that H. arenicola lives heterotrophically to some extent, perhaps until it engulfs its symbiotic partner. This repre- sents an intermediate state of trophic alteration. Although there are several assumptions in the model, it promises to lead to new insights and helps elucidate the plastid integration process.

Evolutionary Implications

Extra secondary and tertiary endosymbioses in dinoflagellates may represent evolutionary events that occurred during plastid acquisition (for re- views, see Hackett et al. 2004a; Morden and Sherwood 2002; Schnepf and Elbrächter 1999). The earliest stage is represented by the crypto- phyte symbiont, in which only cytoskeletal com- ponents are lost, as in Amphidinium latum Lebour (Horiguchi and Pienaar 1992), A. poecilochroum Larsen (Larsen 1988), and Gymnodinium acidotum Nygaard (Fields and Rhodes 1991; Wilcox and Wedemayer 1984). The cell cycles of the host and symbiont are not synchronized, and the host cell must repeatedly capture symbionts. The next stage is represented by the symbiont of diatom origin, in which various subcellular structures are lost, except the nucleus, mitochondrion(a) and plastid, as in Durinskia baltica (Levander) Carty et Cox (formerly Peridinium balticum; Chesnick and Cox 1987, 1989; Chesnick et al. 1997; Eschbach et al. 1990; Tippit and Pickett-Heaps 1976; Tomas and Cox 1973) and Kryptoperidinium foliaceum (Stein) Lindemann (Dodge 1971; Eschbach et al. 1990). At this stage, the symbiont divides syn- chronously with the host cell (Chesnick and Cox 1987, 1989; Tippit and Pickett-Heaps 1976), so that the association between the host and symbiont becomes permanent, and repeated uptake of the symbiont is no longer necessary. Finally, the symbiont cytoplasm is reduced, as seen in Lepidodinium viride (Watanabe et al. 1987; Watanabe et al. 1990), Gymnodinium chlorophor- um (Elbrächter and Schnepf 1996), both with symbionts of prasinophyte origin, and Karenia

brevis (Davis) Hansen et Moestrup, Karenia mikimotoi (Miyake et Kominami ex Oda) Hansen et Moestrup, Karlodinium veneficum (Leadbeater et Dodge) Larsen with symbionts of haptophyte origin (Daugbjerg et al., 2000; Inagaki et al., 2000; Tangen and Björnland, 1981). The symbiont of Karenia and Karlodinium species has no remnant of cytoplasm, and can therefore be recognized as an integrated plastid. Based on cytoplasm reduc- tion and the lack of cell cycle synchronization, the symbiosis of H. arenicola can be placed between the cryptophyte and diatom types of symbiosis mentioned above. Nevertheless, their morpholo- gical association suggests an intimate host—sym- biont relationship. Previous studies on symbiosis in dinoflagellates have focused on symbiont degradation only. In this study, we demonstrated that major morphological changes also occur in the host, suggesting that plastid acquisition is not merely ‘‘enslavement’’ where the symbiont is degraded, but also a process during which the host itself changes to establish a new association with the symbiont.

The rigid pattern of asymmetrical inheritance of the symbiont is also suggestive of a partly regulated association. The symbiont always comes to the left side of the host cell (ventral view) before the migration of the host nucleus, implying an interaction between the symbiont compartment and the host cytoskeleton. If the symbiont moved to the division plane, and not to one side of the cell, the symbiont would co- segregate, just as in the division of D. baltica, where the center-positioned symbiont co-segre- gates upon host cytokinesis (Tippit and Pickett- Heaps 1976).

Another question is whether the association of H. arenicola and the symbiont has developed genetic modification. The process of endosym- biosis is hypothesized to include genetic changes such as lateral gene transfer (LGT) from symbiont to host, coupled with evolution of a protein transport machinery from host to symbiont (e.g. Gilson and McFadden 2002). It is unclear when those changes start and how they integrate. Considering host—symbiont intimacy, H. arenico- la would have already experienced or may be experiencing some of these changes. Therefore, the study of LGT or of a protein transport machinery in H. arenicola would be an interesting topic for future studies.

Recently, Marin et al. (2005) reported another example of primary endosymbiosis in Paulinella chromatophora Lauterborn, a freshwater thecate amoeba that bears a cyanobacterium-like

ARTICLE IN PRESS

414 N. Okamoto and I. Inouye

structure. They reported that the symbiosis of P. chromatophora is a more recent event than the origin of all other plastids, based on the molecular phylogeny of ribosomal DNA operon sequences. This is consistent with a morphological feature of the symbiont, namely, a peptidoglycan layer that must have originated from the cell wall of an ancestral cyanobacterium.

The symbiotic relationships of H. arenicola, P. chromatophora, and the dinoflagellates probably represent different intermediate steps in plastid acquisition via primary or secondary endosymbio- sis. Continued study and comparison of these groups should provide further insight into plastid evolution.

Concluding Remarks

Hatena arenicola gen. et sp. nov. is likely in the process of plastid acquisition via secondary endosymbiosis. Although it is in an early inter- mediate stage of acquisition, the two organisms have already established an intimate association in ultrastructure and likely in metabolic function. Based on behavioral and ultrastructural observa- tions, we propose a ‘‘half-plant, half-predator’’ life cycle. Because H. arenicola shows an early intermediate state of plastid acquisition, it should provide further insight into plant evolution. This study provides a foundation for future studies on the topic.

Methods

Sampling and temporary maintenance in the laboratory: Because it is not possible to culture Hatena arenicola in the laboratory, we used crude samples from the natural habitat. Cells were collected at Isonoura Beach, Wakayama Prefec- ture, Japan (Fig. 1 B,C), April—December from 2000 to 2004. Samples were maintained in the laboratory at room temperature in f/2 medium, under ca. 10 mmol photons m�2 s�1, and the light—dark cycle was L:D ¼ 5:19 h.

Morphological observations: Light micro- scopy (LM) and fluorescence microscopy (FM) was conducted using a Leica DMR light micro- scope (Leica Wetzlar GmbH, Wetzlar) and the LM image was taken with a Keyence VB6010 digital chilled CCD camera (Keyence, Osaka). For FM, 40,6-Diamidino-2-phenylindole (DAPI) was used to stain the nucleus. The DAPI fluorescence along with the autofluorescence of the plastid were

observed using a D filter cube (Leica Wetzlar GmbH, Wetzlar).

Microspectrophotometry was performed by majoring three different regions of the symbiont in each of seven cells. Each absorption spectrum was recorded in the range of 300—800 nm with a light microscope (ECLIPSE, Nikon, Tokyo) equipped with a high-resolution multichannel photodetector (MCPD 7000, Otsuka Eelectronics, Osaka) at Okazaki National Institute for Basic Biology, Japan. The average of the three measure- ments was considered the representative absor- bance of each cell. Because the absorptions obtained were almost uniform across the seven cells, their average is shown.

A unialgal culture of Nephroselmis sp. (NIES1417) was established from the same sam- ple site by micropipette isolation and maintained in f/2 medium at 20 1C under ca. 10 mE light intensity, and a light-dark cycle of L:D ¼ 14:10 h. The uptake of Nephroselmis sp. (NIES1417) was photographed under a CKX31 inversed light microscope (Olympus, Tokyo) equipped with a COOLPIX 990 digital camera (Nikon, Tokyo) at 6-s intervals.

High-speed video images were recorded at 200 frames per second using an OPTIPHOT micro- scope (Nikon, Tokyo), equipped with an MHS-200 high-speed video capturing system (Nac Inc., Tokyo). The images were digitized on a Macintosh computer using an NIH imaging program (public domain, developed at the US National Institute of Health; available at http://rsb.info.nih.gov/nih-im- age/) for analysis of the cellular and flagellar motion.

Preparation for transmission electron micro- scopy (TEM) and scanning electron microscopy (SEM) was performed as described elsewhere (Moriya et al. 2000; Okamoto and Inouye 2005b). The observations were made with a JEOL 100CXII electron microscope (JEOL, Tokyo) and a JSM- 6330 scanning electron microscope (JEOL, To- kyo).

Molecular phylogeny: To avoid contamination of the prey genome, single cells of Hatena arenicola were isolated with micropipette into a 0.2-ml PCR tube containing 10 ml of sterilized double distilled water, and immediately frozen at �80 1C for more than 15 min to completely disrupt the cells. The first and the second nested PCR were performed with existing degenerate primer sets (Moriya et al. 2000) using rTaq (TOYOBO, Osaka). Thermal cycling for the first PCR con- sisted of 33 cycles. Annealing temperatures ranged from 50 to 47 1C (six cycles decreasing

ARTICLE IN PRESS

Table 1. Sequences used for the phylogenetic analyses

Organism Accession number

SSU rDNA Katablepharidophyta/Kathablepharida

Hatena arenicola AB212285 Katablepharis japonica AB231617 Leucocryptos marina AB193602

Metazoa Cirripathes lutkeni AF052902 Clathrina cerebrum U42452 Monosiga brevicollis AF084618

Fungi Basidiobolus haptosporus AF113413 Chytriomyces hyalinus M59758 Pneumocystis carinii L27658 Saccharomyces cervisiae V01335 Schizosaccharomyces pombe Z19578 Scutellospora cerradensis AB041344

Amoebae Acanthamoeba castellanii M13435 Hartmannella vermiformis M95168 Leptomyxa reticulata AF293898 Dictyostelium discoideum K02641

Cercozoa Cercomonas longicauda AF101052 Chlorarachnion CCMP242 U03479 Euglypha rotunda X77692 Heteromita globosa U42447 Paulinella chromatophora X81811 Thaumatomonas seravini AF411259

Viridiplantae Arabidopsis thaliana X16077 ‘Chlorella’ ellipsoidea D13324 Chlorokybus atmophyticus M95612 Closterium littorale AF115438 Coleochaete scutata X68825 Fossombronia pusilla X78341 Mesostigma viride AJ250108 Pyramimonas propulsa AB017123 Tetraselmis striata X70802 Ulothrix zonata Z47999

Heterokontophyta Ochromonas danica M32704 Pteridomonas danica L37204 Skeletonema pseudocostatum X85394 Phytophthora megasperma X54265

Alveolata Cryptosporidium parvum X64340 Toxoplasma gondii X75429 Platyophrya vorax AF060454 Prorodon teres X71140 Alexandrium minutum U27499 Gymnodinium sp. MUCC284 AF022196 Pfiesteria sp. B112456 AF218805

Table 1. (continued )

Organism Accession number

Prorocentrum mexicanum Y16232 Prorocentrum micans M14649

Cryptophyta Chroomonas sp. M1318 AJ007279 Cryptomonas ovata AJ421147 Geminigera cryophila U53124 Hanusia phi U53126 Hemiselmis brunnescens AJ007282 Rhodomonas mariana X81373 Goniomonas truncata U03072

Glaucophyta Cyanophora paradoxa X68483 Cyanoptyche gloeocystis AJ007275 Glaucocystis nostochinearum X70803 Gloeochaete wittrockiana X81901

Haptophyta Emiliania huxleyi L04957 Pavlova salina L34669

Rhodophyta Bangia sp. AF043362 Porphyra umbilicalis AB013179 Rhodella maculata U21217 Stylonema alsidii L26204

Centroheliozoa Chlamydaster sterni AF534709 Heterophrys marina AF534710 Raphidiophrys ambigua AF534708

415Hatena arenicola: Halfway to a Plant?

the temperature by 0.5 1C for each cycle, and 27 cycles at a constant temperature). An extension was performed at 72 1C for 1 min, and denaturing was done at 94 1C for 30 s. The final extension period was at 72 �C for 7 min. Thermal cycling for the second PCR consisted of 33 cycles with an annealing step at 53 1C for 30 s, an extension step at 72 1C for 1 min, and denaturing at 94 1C for 30 s, with a final extension period at 72 1C for 7 min. The sequences were determined by direct sequen- cing. Cycle sequencing reaction was performed using a DYEnamic ET terminator cycle sequencing kit (Amersham biosciences, Buckinghamshire), as per the manufacturer’s instructions. Sequencing was conducted with an ABI PRISM 377 DNA Sequencer (Applied Biosystems, California), and sequences were confirmed free of contaminants and not of haptophyte prey origin by comparing at least two cells or performing a BLAST search at the National Center for Biotechnology Informa- tion (NCBI) server (http://www.ncbi.nlm.nih.gov/

ARTICLE IN PRESS

416 N. Okamoto and I. Inouye

BLAST/). The sequence (AB212285) was depos- ited to GenBank.

The SSU sequence was manually aligned to the existing alignments of global eukaryotes (Okamo- to and Inouye 2005b), which include 65 taxa for SSU rDNA (Table 1). To avoid the long-branch attraction (LBA) artifact in SSU rDNA analysis, organisms with an extraordinary evolutionary rate, such as the Euglenozoa, the diplomonads, and the palabasalids, were excluded after confirming that they are not a sister group of the katablephar- ids. A total of 1252 unambiguously aligned nucleotide positions were selected for phyloge- netic analyses.

The maximum likelihood (ML), neighbor joining (NJ), and maximum parsimony (MP) methods for phylogenetic analysis were applied to the data set using PAUP* ver.4.0 b 10 (Swofford 2003). In the ML analysis, the evolutionary model was selected using the Akaike information criterion (AIC) test in Modeltest ver.3 (Posada and Crandall 1998), which selected the general time reversible (GTR) model with rate variation among sites and invariant sites (GTR+I+gamma). The estimated gamma-shape parameter (alpha) of the discrete gamma-distribution was 0.5645, and the propor- tion of invariant sites was 0.3257. A further tree search with GTR+I+gamma model with eight site rate categories that approximates site rate was used to produce the optimal tree. Bootstrap proportion (BP) values for internal branches of the optimal tree of the ML analysis were obtained using PAUP* through 300 bootstrap resamplings for the ML method and 1000 resamplings for the NJ and MP methods.

Acknowledgments

We thank Dr. Masakatsu Watanabe and Dr. Shigeru Matsunaga (Graduate University for Ad- vanced Studies, Japan) for their dedication to the microspectrophotometric work and assistance in sampling, and Dr. Hiroshi Kawai (Research Center for Inland Seas, University of Kobe) for use of his laboratory equipment. We appreciate Dr. Tetsuo Hashimoto and Dr. Miako Sakaguchi for their assistance and advices in molecular phylogeny. We are grateful to Dr. Jeremy Pickett-Heaps for taking movie images of H. arenicola. This research was supported in part by Japan Society for the Promotion of Sciences (JSPS) Grants RFTF00L0162 (to I.I.) and 1612007 (to N.O.). N.O. was also supported by a JSPS Research

Fellowship for Young Scientists as a JSPS Research Fellow.

Appendix A. Supplementary materials

Supplementary data associated with this article can be found in the online version at doi:10.1016/ j.protis.2006.05.011

References

Andersson JO, Roger AJ (2002) A cyanobacterial gene in nonphotosynthetic protists—an early chlor- oplast acquisition in eukaryotes? Curr Biol 12: 115—119

Archibald JM, Rogers MB, Toop M, Ishida K, Keeling PJ (2003) Lateral gene transfer and the evolution of plastid-targeted proteins in the second- ary plastid-containing alga Bigelowiella natans. Proc Natl Acad Sci USA 100: 7678—7683

Baldauf SL (2003) The deep roots of eukaryotes. Science 300: 1703—1706

Baldauf SL, Roger AJ, Wenk-Siefert I, Doolittle WF (2000) A kingdom-level phylogeny of eukaryotes based on combined protein data. Science 290: 972—977

Bapteste E, Brinkmann H, Lee JA, Moore DV, Sensen CW, Gordon P, Durufle L, Gaasterland T, Lopez P, Muller M, Philippe H (2002) The analysis of 100 genes supports the grouping of three highly divergent amoebae: Dictyostelium, Entamoeba, and Mastigamoeba. Proc Natl Acad Sci USA 99: 1414—1419

Bhattacharya D, Yoon HS, Hackett JD (2004) Photosynthetic eukaryotes unite: endosymbiosis connects the dots. Bioessays 26: 50—60

Butcher RW (1967) An Introductory Account of the Smaller Algae of British Coastal Waters. 4. Crypto- phyceae. Fishery Investigations, Series IV. HMSO, London

Cavalier-Smith T (2003) Genomic reduction and evolution of novel genetic membranes and protein- targeting machinery in eukaryote—eukaryote chi- maeras (meta-algae). Philos Trans R Soc Lond B 358: 109—133

Chesnick JM, Cox ER (1987) Synchronized sexu- ality of an algal symbiont and its dinoflagellate host, Peridinium balticum (Levander) Lemmermann. Bio- Systems 21: 69—78

ARTICLE IN PRESS

417Hatena arenicola: Halfway to a Plant?

Chesnick JM, Cox ER (1989) Fertilization and zygote development in the binucleate dinoflagellate Peridinium balticum (Pyrrhophyta). Am J Bot 76: 1060—1072

Chesnick JM, Kooistra WHCF, Wellbrock U, Medlin LK (1997) Ribosomal RNA analysis indicates a benthic pennate diatom ancestry for the endo- symbionts of the dinoflagellates Peridinium folia- ceum and Peridinium balticum (Pyrrhophyta). J Eukaryot Microbiol 44: 314—320

Clay BL, Kugrens P (1999a) Description and ultrastructure of Kathablepharis tenuis sp. nov. and K. obesa sp. nov. — two new freshwater Kathable- pharis (Kathablepharididae) from Colorado Wyom- ing. Europ J Protistol 35: 435—447

Clay BL, Kugrens P (1999b) Systematics of the enigmatic Kathablepharids, including EM character- ization of the type species, Kathablepharis phoeni- koston, and new observations on K. remigera comb. nov. Protist 150: 43—59

Daugbjerg N, Hansen G, Larsen J, Moestrup Ø (2000) Phylogeny of some of the major genera of dinoflagellates based on ultrastructure and partial LSU rDNA sequence data, including the erection of three new genera of unarmoured dinoflagellates. Phycologia 39: 302—317

Dodge JD (1971) A dinoflagellate with both a mesokaryotic and a eukaryotic nucleus. Protoplas- ma 73: 145—157

Douglas S, Zauner S, Fraunholz M, Beaton M, Penny S, Deng LT, Wu XN, Reith M, Cavalier- Smith T, Maier UG (2001) The highly reduced genome of an enslaved algal nucleus. Nature 410: 1091—1096

Elbrächter M, Schnepf E (1996) Gymnodinium chlorophorum, a new, green, bloom-forming dino- flagellate (Gymnodiniales, Dinophyceae) with a ves- tigial prasinophyte endosymbiont. Phycologia 35: 381—393

Eschbach S, Speth V, Hansmann P, Sitte P (1990) Freeze-fracture study of the single membrane between host cell and endocytobiont in the dino- flagellates Glenodinium foliaceum and Peridinium balticum. J Phycol 26: 324—328

Falkowski PG, Katz ME, Knoll AH, Quigg A, Raven JA, Schofield O, Taylor FJR (2004) The evolution of modern eukaryotic phytoplankton. Science 305: 354—360

Fields SD, Rhodes RG (1991) Ingestion and reten- tion of Chroomonas spp (Cryptophyceae) by Gym- nodinium acidotum (Dinophyceae). J Phycol 27: 525—529

Gilson PR, Su V, Slamovits CH, Reith ME, Keeling PJ, McFadden GI (2006) Complete nucleotide sequence of the chlorarachniophyte nucleomorph: Nature’s smallest nucleus. Proc Natl Acad Sci USA 103: 9566—9571

Graham LE, Wilcox LW (2000) Algae. Prentice-Hall, NJ

Gualtieri P (2001) Morphology of photoreceptor systems in microalgae. Micron 32: 411—426

Hackett JD, Anderson DM, Erdner DL, Bhatta- charya D (2004a) Dinoflagellates: a remarkable evolutionary experiment. Am J Bot 91: 1523—1534

Hackett JD, Yoon HS, Soares MB, Bonaldo MF, Casavant TL, Scheetz TE, Nosenko T, Bhatta- charya D (2004b) Migration of the plastid genome to the nucleus in a peridinin dinoflagellate. Curr Biol 14: 213—218

Hashimoto H (2005) The ultrastructural features and division of secondary plastids. J Plant Res 118: 163—172

Horiguchi T, Pienaar RN (1992) Amphidinium latum (Dinophyceae), a sand-dwelling dinoflagellate feed- ing on cryptomonads. Jpn J Phycol (Sorui) 40: 353—363

Huang CY, Ayliffe MA, Timmis JN (2003) Direct measurement of the transfer rate of chloroplast DNA into the nucleus. Nature 422: 72—76

Inagaki Y, Dacks JB, Doolittle WF, Watanabe KI, Ohama T (2000) Evolutionary relationship between dinoflagellates bearing obligate diatom endosym- bionts: insight into tertiary endosymbiosis. Int J Syst Evol Microbiol 50: 2075—2081

Inouye I, Pienaar RN (1984) Light and electron microscope observation on Nephroselmis astigmati- ca sp. nov. (Prasinophyceae). Nord J Bot 4: 409—423

Katz LA (2002) Lateral gene transfers and the evolution of eukaryotes: theories and data. Int J Syst Evol Microbiol 52: 1893—1900

Kugrens P, Lee RE, Corliss JO (1994) Ultrastruc- ture, biogenesis, and functions of extrusive orga- nelles in selected non-ciliate protist. Protoplasma 181: 164—190

Larsen J (1988) An ultrastructural study of Amphi- dinium poecilochroum (Dinophyceae), a phago- trophic dinoflagellate feeding on a small species of cryptophytes. Phycologia 27: 366—377

Lee RE, Kugrens P (1992) Relationship between the flagellates and the ciliates. Microbiol Rev 56: 529—542

ARTICLE IN PRESS

418 N. Okamoto and I. Inouye

Lee RE, Kugrens P, Mylnikov AP (1991) Feeding apparatus of the colorless flagellate Katablepharis (Cryptophyceae). J Phycol 27: 725—733

Lee RE, Kugrens P, Mylnikov AP (1992) The structure of the flagellar apparatus of two strains of Katablepharis (Cryptophyceae). Br Phycol J 27: 369—380

Marin B, Nowack ECM, Melkonian M (2005) A plastid in the making: evidence for a second primary endosymbiosis. Protist 156: 425—432

Martin W (2003a) Gene transfer from organelles to the nucleus: frequent and in big chunks. Proc Natl Acad Sci USA 100: 8612—8614

Martin W (2003b) The smoking gun of gene transfer. Nature Genet 33: 442

Martin W, Herrmann RG (1998) Gene transfer from organelles to the nucleus: How much, what hap- pens, and why? Plant Physiol 118: 9—17

Martin W, Rujan T, Richly E, Hansen A, Cornelsen S, Lins T, Leister D, Stoebe B, Hasegawa M, Penny D (2002) Evolutionary analysis of Arabidopsis, cyanobacterial, and chloroplast genomes reveals plastid phylogeny and thousands of cyanobacterial genes in the nucleus. Proc Natl Acad Sci USA 99: 12246—12251

McFadden GI (2001) Primary and secondary en- dosymbiosis and the origin of plastids. J Phycol 37: 951—959

Melkonian M, Robenek H (1984) The Eyespot Apparatus of Flagellated Green Algae: a Critical Review. In Round FE, Chapman DJ (eds) Progress in Phycological Research, Vol. 3. Biopress Ltd., Bristol, pp 93—268

Morden CW, Sherwood AR (2002) Continued evolutionary surprises among dinoflagellates. Proc Natl Acad Sci USA 99: 11558—11560

Moriya M, Nakayama T, Inouye I (2000) Ultrastruc- ture and 18SrDNA sequence analysis of Wobblia lunata gen. et sp. nov., a new heterotrophic flagellate (stramenopiles, incertae sedis). Protist 150: 835—846

Nozaki H, Matsuzaki M, Misumi O, Kuroiwa H, Hasegawa M, Higashiyama T, Shin T, Kohara Y, Ogasawara N, Kuroiwa T (2004) Cyanobacterial genes transmitted to the nucleus before divergence of red algae in the Chromista. J Mol Evol 59: 103—113

Nozaki H, Matsuzaki M, Takahara M, Misumi O, Kuroiwa H, Hasegawa M, Shin-i T, Kohara Y, Ogasawara N, Kuroiwa T (2003) The phylogenetic position of red algae revealed by multiple nuclear

genes from mitochondria-containing eukaryotes and an alternative hypothesis on the origin of plastids. J Mol Evol 56: 485—497

Okamoto N, Inouye I (2005a) A secondary symbio- sis in progress? Science 310: 287

Okamoto N, Inouye I (2005b) The katablepharids are a distant sister of the Cryptophyta: a proposal of Katablepharidophyta divisio nova/Kathablepharida phylum novum based on SSU rDNA and beta tubulin phylogeny. Protist 156: 163—179

Posada D, Crandall KA (1998) Modeltest: testing the model of DNA substitution. Bioinformatics 14: 817—818

Schnepf E, Elbrächter M (1999) Dinophyte chloro- plasts and phylogeny — a review. Grana 38: 81—97

Simpson AGB, Roger AJ (2002) Eukaryotic evolu- tion: Getting to the root of the problem. Curr Biol 12: R691—R693

Skuja H (1939) Beitrag zur Algenflora Lettlands. II. Acta Hortic Bot Univ Latviensis 11/12:41—168

Stegemann S, Hartmann S, Ruf S, Bock R (2003) High-frequency gene transfer from the chloroplast genome to the nucleus. Proc Natl Acad Sci USA 100: 8828—8833

Swofford DL (2003) PAUP*: Phylogenetic Analysis Using Parsimony (* and other methods), Version 4.0b10. Sinauer Associates, Inc., Sunderland, MA

Tangen K, Björnland T (1981) Observations on pigments and morphology of Gyrodinium aureolum Hulburt, a marine dinoflagellate containing 190- hexanoyloxyfucoxanthin as the main carotenoid. J Plankton Res 3: 389—401

Tippit DH, Pickett-Heaps JD (1976) Apparent amitosis in the binucleate dinoflagellate Peridinium balticum. J Cell Sci 21: 273—289

Tomas RN, Cox ER (1973) Peridinium balticum (Levander) Lemmermann, an unusual dinoflagellate with a mesocaryotic and an eukaryotic nucleus. J Phycol 9: 91—98

van der Giezen M, Cavalier-Smith T, Martin W, Wilkins A, Doolittle WF, Allen JF, Leaver CJ, Whatley FR, Howe CJ (2003) Genomic reduction and evolution of novel genetic membranes and protein-targeting machinery in eukaryote—eukar- yote chimaeras (meta-algae) — Discussion. Philos Trans R Soc Lond B 358: 133—134

van Dooren GG, Schwartzbach SD, Osafune T, McFadden GI (2001) Translocation of proteins across the multiple membranes of complex plastids. Biochim Biophys Acta Mol Cell Res 1541: 34—53

ARTICLE IN PRESS

419Hatena arenicola: Halfway to a Plant?

Vørs N (1992a) Heterotrophic amoebae, flagellates and heliozoa from the Tvärminne area, Gulf of Finland, in 1988—1990. Ophelia 36: 1—109

Vørs N (1992b) Ultrastructure and autecology of the marine, heterotrophic flagellate Leucocryptos marina (Braarud) Butcher 1967 (Katablepharidaceae/Katha- blepharidae), with a discussion of the genera Leucocryptos and Katablepharis/Kathablepharis. Europ J Protistol 28: 369—389

Watanabe MM, Suda S, Inouye I, Sawaguchi T, Chihara M (1990) Lepidodinium viride gen. et sp. nov. (Gymnodiniales, Dinophyta), a green dinoflagel-

late with a chlorophyll a- and b-containing endo- symbiont. J Phycol 26: 741—751

Watanabe MM, Takeda Y, Sasa T, Inouye I, Suda S, Sawaguchi T, Chihara M (1987) A green dinoflagellate with chlorophylls a and b: morphology, fine structure of the chloroplast and chlorophyll composition. J Phycol 23: 1987

Wilcox LW, Wedemayer GJ (1984) Gymnodinium acidotum Nygaard (Pyrrophyta), a dinoflagellate with an endosymbiotic cryptomonad. J Phycol 20: 236—242