This is not really an "essay" but if you take a look at the Invertebrate Zoology Take home exam you'll see the instructions. Just to note: (problem of the PDA + the search for the urbilaterian ancestor are for question 1) (question 2 is based on the cteno

BioSystems 73 (2004) 73–83

A “chimera” theory on the origin of dicyemid mesozoans: evolution driven by frequent lateral gene transfer

from host to parasite

Tomoko Noto∗, Hiroshi Endoh Department of Biology, Faculty of Science, Kanazawa University, Kanazawa 920-1192, Japan

Received 13 February 2003; received in revised form 8 August 2003; accepted 2 September 2003

Abstract

The phylogenetic status of the enigmatic dicyemid mesozoans is still uncertain. Are they primitive multicellular organisms or degenerate triploblastic animals? Presently, the latter view is accepted. A phylogenetic analysis of 18S rDNA sequences placed dicyemids within the animal clade, and this was supported by the discovery of a Hox-type gene with a lophotrochozoan signature sequence. This molecular information suggests that dicyemid mesozoans evolved from an ancestral animal degenerately. Consid- ering their extreme simplicity, which is probably due to parasitism, they might have come from an early embryo via a radical trans- formation, i.e. neoteny. Irrespective of this molecular information, dicyemid mesozoans retain many protistan-like or extremely primitive features, such as tubular mitochondrial cristae, endocytic ability from the outer surface, and the absence of collagenous tissue, while they do not share noticeable synapomorphy with animals. In addition, the 5S rRNA phylogeny suggests a somewhat closer kinship with protozoan ciliates than with animals. If we accept this clear contradiction, dicyemids should be regarded as a chimera of animals and protistans. Here, we discuss the traditional theory of extreme degeneration via parasitism, and then propose a new “chimera” theory in which dicyemid mesozoans are exposed to a continual flow of genetic information via eating host tissues from the outer surface by endocytosis. Consequently, many of their intrinsic genes have been replaced by host-derived genes through lateral gene transfer (LGT), implying that LGT is a key driving force in the evolution of dicyemid mesozoans. © 2003 Elsevier Ireland Ltd. All rights reserved.

Keywords:Dicyemid mesozoans; Protistans; Triploblastic animals; Lateral gene transfer; Parasitism; Chimera

1. Introduction

The dicyemid mesozoans, obligate endosymbionts found in the renal system of benthic cephalopods, are one of the simplest multicellular organisms (re- viewed by Furuya and Tsuneki, 2003). They consist of one long axial cell surrounded by a single layer of 20–40 multiciliated somatic cells. The axial cell

∗ Corresponding author. Tel.:+81-76-264-6099; fax: +81-76-264-6099.

E-mail address:[email protected] (T. Noto).

contains a large polyploid nucleus and intracellular stem cells, called axoblasts (Fig. 1). Dicyemids lack distinguishable organs, except for a gonad-like struc- ture that appears during one stage of their life cycle. According to Nouvel (1948), in the late 18th cen- tury, Filippo Calvolini of Italy found small worm-like organisms—dicyemid mesozoans—in octopuses. In 1849, Kölliker named them dicyemids, because they produce two types of embryos in their life cycle. In 1876, Van Beneden called them Mesozoa, to express his belief that the group occupied an evolutionarily intermediate position between the Protozoa and the

0303-2647/$ – see front matter © 2003 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.biosystems.2003.09.002

74 T. Noto, H. Endoh / BioSystems 73 (2004) 73–83

Fig. 1. Dicyemid mesozoans. Light micrograph of nematogen adult (left) and the diagram of the identical adult (right). AB, axoblast (agamete); AC, axial cell; AN, axial cell nucleus; C, calotte; DE, developing embryo; PC, peripheral cell. Bar represents 10�m.

Metazoa. Some investigators have maintained his position (Beneden, 1882; Hartmann, 1925; Hyman, 1940; Dodson, 1956; Lapan and Morowitz, 1974). Conversely, some have proposed that mesozoans have undergone secondary simplification from a worm-like animal as a result of extreme parasitism (Nouvel, 1947; McConnaughey, 1951; Stunkard, 1954). Con- sequently, it has long been controversial whether the dicyemids are truly primitive multicellular organ- isms or secondarily degenerated metazoans. Recent molecular evidence has added to the debate between these two views, and the second view tends to be fa- vored (Katayama et al., 1995; Kobayashi et al., 1999; Pawlowski et al., 1996). In contrast, information on biological traits shows a drastically different aspect of dicyemids. There are no definitive characters support- ing a close kinship of dicyemids with animals, while many show an affiliation with protistans. This situa- tion renders the phylogenetic position of dicyemids enigmatic. This paper highlights the contradiction be- tween molecular information and biological traits in the phylogenetic position of dicyemid mesozoans. We

propose a new theory that resolves this contradiction rationally, leading to the conclusion that dicyemid mesozoans are a chimera organism of animals and protistans.

2. Background

Molecular sequence data are increasingly used to analyze phylogenetic relationships among eukaryotes. A phylogenetic analysis of 5S rRNA data suggested that dicyemids are more closely related to protozoan ciliates than to multicellular animals (Ohama et al., 1984). Halanych (1991)argued that the 5S rRNA molecule is too small to contain phylogenetic informa- tion sufficient for appropriate reconstruction of evolu- tionary relationships, although his tree also indicated a close relationship between mesozoans and protozoan ciliates. By contrast, 18S rDNA analyses placed the mesozoans as triploblastic animals (Katayama et al., 1995; Pawlowski et al., 1996). However, 18S rDNA phylogenies are sometimes misleading (Loomis and Smith, 1990) and may be inadequate to elucidate relationships among groups more than 500 million years old (Rodorigo et al., 1994). In practice, rRNA- and protein-coding gene-based phylogenies can con- tradict each other drastically, as in theTrypanosoma (Alvarez et al., 1996; Germot and Philippe, 1991) and amitochondrial protozoa like microsporidia (Keeling et al., 2000) andEntamoeba(Hasegawa et al., 1993). The validity of molecular sequence data for deduc- ing phylogenetic relationships depends on selecting macromolecules that are ubiquitous, have a highly conserved primary structure, and are functionally conserved during evolution (Müller, 1995).

Since the sequences used to construct phy- logenies of the dicyemid mesozoans so far are RNA-coding genes, a phylogeny based on represen- tative protein-coding genes is needed to provide more robust data. Microtubules are structures that are char- acteristic of eukaryotic cells; they are associated with cell movement via major cytoskeleton components, axonemes, and the 9+ 0 basal body/centriole, sug- gesting that their evolution may have paralleled that of eukaryotes (Edlind et al., 1996). �-Tubulin sequences from a wide variety of eukaryotic species have been reported (Burns, 1991) and used for phylogenetic analyses (Edlind et al., 1996). In order to elucidate

T. Noto, H. Endoh / BioSystems 73 (2004) 73–83 75

the phylogenetic relationship between the dicyemids and other eukaryotes, we cloned and sequenced four �-tubulin genes from two dicyemid species as a representative protein-coding gene. In our�-tubulin phylogeny, dicyemid mesozoans were again placed within higher invertebrates, rather than near lower ones, such as platyhelminths, different from the 18S rDNA analysis (Appendix A). In both trees from 18S rDNA and �-tubulin genes, however, the exact posi- tion of the dicyemid mesozoans within animals was not supported by reliable bootstrap values because of the poor resolutions. These analyses only indicate that the dicyemid mesozoans are involved in the triploblas- tic animals, but not in fungi and protists. Most of the molecular information, including the presence of a Hox-type gene discussed below, strongly suggests that dicyemid mesozoans are triploblastic animals. Is this conclusion fully convincing? We suspect that there is still something wrong with it.

3. Theoretical consideration of the status of dicyemid mesozoans

3.1. Are dicyemids triploblastic animals?

Since Whitman (1883)regarded the simplicity of the mesozoans as not at all primitive, but the result of extreme parasitic degeneration, several investigators, such asStunkard (1954), have strongly maintained this viewpoint. In the last decade, two lines of evidence suggesting that dicyemids evolved degenerately from animals have accumulated. The 18S rDNA phylogeny suggested that dicyemids were triploblastic animals (Katayama et al., 1995; Pawlowski et al., 1996). Re- cently, the presence of a Hox-type gene,DoxC, was reported in the dicyemid mesozoanDicyema orien- tale (Kobayashi et al., 1999). The analysis of the homeodomain sequence indicated that it has the high- est homology with a member of the ‘middle’ group of Hox genes, supporting the 18S rDNA phylogeny. In addition, the so-called ‘spiralian peptide’ motif was confirmed, so the authors advocated the affinity of dicyemids with lophotrochozoans, which consist of brachiopods, annelids, nemertines, platyhelminths, and mollusks including cephalopods, which are the hosts of the dicyemid mesozoans (Aguinaldo et al., 1997; Adoutte et al., 1999). The �-tubulin gene phy-

logeny presented here leads to a similar conclusion, although the resolution of animals was low (Appendix A). Based on this molecular information, there are grounds for classifying dicyemids as triploblastic animals. It is possible that extreme degeneration oc- curred to an unimaginable extent via parasitism. This interpretation requires an explanation of such extreme simplicity; dicyemid adults consist of some 30 cells, which are derived from an axoblast or fertilized egg involving at most 5–9 cell divisions. These cells never divide again during the organism’s life (Furuya et al., 1992, 1994). Considering this, one is compelled to postulate that dicyemids evolved by neoteny from an early embryo at the level of a morula. Recent studies in developmental biology have accumulated much knowledge on the body plan and many genes involved in morphogenesis have been identified (e.g. reviewed by Prince, 2002). The loss of some such genes might have been responsible for the extreme degeneration. This approach might elucidate whether the simplic- ity of dicyemids is really derived from an ancestral animal by parasitic degeneration. Simultaneously, it might be possible to clarify experimentally how the primitive or protistan-like traits were generated or reverted, accompanying the simplification in body construction.

3.2. Why are there so many primitive or protistan-like features?

The extremely simple dicyemid mesozoans lack a nervous system and gut. For this reason,Cavalier- Smith (1993)once placed the phylum Mesozoa in the kingdom Protozoa; this recommendation must show foresight. Now he still gives the Mesozoa the rank of a distinct subkingdom (Cavalier-Smith, 1998). We agree with his proposal, since dicyemids maintain many protistan-like features, and have radical simplification. No articles comprehensively describe the primitive or protistan-like features. Therefore, we summarize these features and discuss them in some detail.

Noting that dicyemid mesozoans had protozoan features,Hartmann (1907)coined the term Moru- loidea for them. Since then, the following evidence of their primitiveness has been noted. They have (1) a double-stranded ciliary necklace, (2) tubular cristae in their mitochondria, (3) endocytic ability from the outer surface, (4) an absence of collagen in the extra-

76 T. Noto, H. Endoh / BioSystems 73 (2004) 73–83

cellular matrix (ECM), (5) cell-to-cell junctions, and (6) distinct phases of asexual (nematogen) and sexual (rhombogen) reproduction.

Freeze-fracture analysis identifies the ciliary neck- lace, a structural array of integral membrane proteins that has been valuable as a genetically fixed mem- brane character for addressing phylogenetic questions (Bardele, 1981). The pattern of protein arrangement in protists is remarkably varied, whereas inverte- brates, including the Porifera and Cnidaria, have a consistent pattern. Animals are characterized by a triple-stranded necklace, while dicyemid mesozoans share a double-stranded necklace structure with pro- tistan ciliates and opalinids (Bardele et al., 1986).

Generally, animals lack the ability to take in food or particulate materials via their outer surface. In con- trast, the dicyemids can take in particulate material, such as ferritin (Ridley, 1968) or host spermatozoa (Nouvel, 1933), from the surface of their peripheral cells by phagocytosis. This characteristic is strikingly different from that of animals. If degeneration in fact occurred, the degenerated ancestor would have had to regain the ability to endocytose material from the outer cell surface, concomitant with the loss of the diges- tive tract. However, no embryos in animals retain the endocytic ability even in the stage of gastrula.

The shape of mitochondrial cristae is a diagnos- tic character for taxa, although it is not necessarily crucial; there are a few instances in which the shape of the cristae alternates within the life cycle, as in Trypanosoma bruceiand certain platyhelminths. An- imals, fungi, and plants generally have mitochondria with plate-like cristae, whereas protistans have either tubular or discoidal cristae (Gray et al., 1998). In di- cyemids, the cristae are tubular, like those of most protistans, throughout their life cycle, unlike most an- imals (Ridley, 1968, 1969).

The synapomorphy that is considered crucial to the affiliation of mesozoans to animals is the presence of collagenous connective tissue, but not multicellu- larity (Willmer, 1990; Cavalier-Smith, 1993, 1998). So far, electron microscopic observation has yet to identify an extracellular matrix (ECM), such as collagen-like structures, in dicyemids (Furuya et al., 1997). Recently, the dicyemid mesozoanKantharella antarcticawas observed by electron microscopy using fibronectin, laminin, and type IV collagen antibod- ies to investigate the ECM (Czaker, 2000). All three

ECM components were located intracellularly, but not intercellularly, unlike the typical ECM. Indeed, fibronectin- and laminin-like molecules have also been confirmed in protistans such as kinetoplastid Leishmania(Del Cacho et al., 1996) and apicom- plexan Eimeria (Lopez-Bernad et al., 1996). These observations strongly suggest that this intracellular distribution of ECM components is primitive. The absence of ECM in dicyemids might be responsi- ble for body organization, which does not reach the tissue level typical of animals (Furuya et al., 1997). The only similar case in animals is the turbellarian group Acoela, which lacks an intercellular matrix (Rieger, 1985). Consequently, a relationship between dicyemids and acoelomates must be considered.

With reference to this problem, cell junctions such as gap junctions (cytoplasmic connections) and adherens-like junctions have been confirmed in di- cyemids, but typical septate junctions are absent (Furuya et al., 1997). The gap junction is thought to function in cell-to-cell communication and the exchange of molecules between neighboring cells. Although lower animals, such as placozoans and sponges, lack gap junctions, a similar channel system is believed to develop. Even in protistans, such junc- tions are observed when cell-to-cell union occurs. For example, in order to synchronize the conjuga- tion process and ciliary movement between pairing partners, a cytoplasmic connection is formed during conjugation in ciliates in which a multicellular state is transiently established. The adherens junction has also been discovered in the multicellular structure of non-metazoan cellular slime molds, coupled with a �-catenin homologue (Grimson et al., 2000). Further- more, it is well known that multicellularity occurred independently many times in the course of evolution, even in protistans (Willmer, 1990; Bonner, 1997). These discoveries outside the animal kingdom show that the potential for cell junction formation had already developed in protistans. Accordingly, inter- cellular junctions are not necessarily crucial to solve phylogenetic relationships.

Finally, dicyemids have distinct phases of asexual and sexual reproduction. In the nematogen phase, larvae develop asexually from a diploid axoblast, whereas in the rhombogen phase, larvae are pro- duced from a fertilized egg. The former appears protistan-like, although regenerative reproduction is

T. Noto, H. Endoh / BioSystems 73 (2004) 73–83 77

observed in some animals. This feature is too dif- ferent from that in triploblastic animals to imagine how dicyemids acquired the alteration of asexual and sexual reproduction.

4. The chimera theory can solve the discrepancy

As mentioned above, dicyemids maintain many protistan-like or extremely primitive features and lack noticeable morphological characters that are shared with animals; however, most of the molecular information obtained so far strongly suggests that dicyemids are true animals. How should this discrep- ancy be interpreted? Here, we present a new theory to resolve this discrepancy. It may be reasonable to regard dicyemids as a chimera of protistans and animals, in which dicyemids acquired many genes from their host via lateral gene transfer (LGT). Sev- eral lines of evidence have recently shown that LGT via phagocytosis occurs with higher than expected frequency (Doolittle, 1998; Schubbert et al., 1997, 1998; Bushman, 2002). For example, theTetrahymena genome project (http://www.tigr.org/tdb/tgi/ttgi/) has determined that approximately 80 genes out of 3500 sequences determined so far came from bacteria, in spite of their free-living mode. Dicyemids are re- stricted to a renal appendage in cephalopods, where they absolutely depend on their host for all nutrients. They have endocytic ability mentioned above and the uptake of host spermatozoa has been observed repeatedly. Furthermore, the calotte (the most ante- rior cells) cilia are stiffer, shorter, thicker and more closely set than those of other peripheral cells, and occasionally penetrate the epithelial cells of the renal appendages, resulting in erosion of the tissue (Ridley, 1968). Therefore, they are exposed to a continual flow of genetic information from the host via their food (fragments of the host tissue and spermatozoa). This situation increased the chance of the dicyemid germline genome taking in host DNA. The obser- vation that two�-tubulin genes fromDicyemodeca contained a short intron at precisely the same site as in the host gene may reflect such gene flow from the host (Appendix A). This assumption reason- ably interprets the inconsistent facts; the presence of lophotrochozoan-like genes, such as Hox-type genes,

and many protistan-like features. A certain laterally transferred gene from the host may have driven multi- cellularization of an ancestral unicellular dicyemid to some extent, and this would have led to multiciliation and polyploidization of somatic nuclei accompanied by DNA rearrangement (Noto et al., 2003), as seen in ciliates. Nevertheless, the intrinsic nature of the putative protistan ancestor might have remained un- changed, resulting in the creation of a ‘chimera.’ In this sense, dicyemids are truly the ‘Mesozoa,’ mak- ing this term even more appropriate. Frequent LGT might be an important driving force in the evolu- tion of dicyemids in particular and in host–parasite relationships in general. This viewpoint will be in- dispensable for clarifying the origin of dicyemid mesozoans.

5. Perspective

To date, only a few genes in dicyemids have been analyzed. If a genome project in dicyemid were car- ried out, or many more genes were analyzed, we ex- pect there to be two major classes of genes identi- fied: animal-like genes and protistan-like genes. At present, the 5S rRNA gene is the only gene that does not show the affiliation of dicyemids to animals. If our theory is true, it may be a vestige derived from its an- cestor. The discovery of additional genes of this type would lend support to our theory. Now we must await the accumulation of such information on the genes of dicyemids.

With reference to this theory, quite recently, fre- quent LGTs were systematically analyzed in protis- tan diplomonads (Andersson et al., 2003). The authors suggest that LGT is a likely source accounting for anomalous phylogeny patterns which are observed in different genes. If LGT events are assumed to be fre- quent in a certain species, an estimation of molecular phylogeny should be cautiously made.

Finally, the collection of completely sequenced mitochondrial genomes has been expanding rapidly (Gray et al., 1998). Generally, mitochondrial DNA (mtDNA) in animals is roughly equal in size, gene content, and genome organization; it ranges from 14 to 20 kb in size and is circular. In contrast, protistan mtDNA is very different from animal mtDNA in that it is extraordinarily diverse in size, form, and gene

78 T. Noto, H. Endoh / BioSystems 73 (2004) 73–83

content. No knowledge of dicyemid mtDNA is yet available, except for the presence of minicircle DNAs encoding cytochrome oxidase I, II, and III (Watanabe et al., 1999). These minicircles all have relatively long non-coding regions. Extrapolating the size of the entire genome based on the ratio of the known coding and non-coding sequences, dicyemids appear to have mtDNA larger than that of animals. This is inconsistent with the general tendency for parasites to downsize their mt genome as an adaptation to parasitism (Gray et al., 1999; Saccone et al., 2000). Indeed, dicyemids seem to maintain a small num- ber of high-molecular weight mitochondrial genes in germ cells, separately from the minicircles (H. Awata, personal communication). The entire mitochondrial genome of dicyemids must be analyzed in detail to determine their phylogenetic relationship.

Acknowledgements

We thank R. Kofuji, T. Hanyuda and K. Ishida, for construction of phylogenetic trees, and H. Awata for a kind supply of her unpublished results, and S. Sakurai, Y. Sasayama and T.G. Doak for encourage- ment to accomplish this work. We are also grateful to A. Sawabe, Y. Sasayama, and K. Yamamoto for sampling and maintenance of the host materials. This work was partially supported by Sasakawa Scientific Research grant.

Fig. 2. Map of the�-tubulin genes from dicyemids and the host octopus. BTov1 (1635 bp) was obtained from the hostO. vulgaris. BTdv1 (1220 bp) and�BTdv2 (1217 bp) or BTda1 (1310 bp) and BTda2 (1327 bp) were obtained fromDicyemasp. orD. antinocephalum, respectively. Horizontal lines represent a putative protein-coding region and introns located in the identical sites are shown as triangles in the same colors. Numbers in the triangles and round brackets denote the intron number and their length in base pair, respectively. Crosses in �BTdv2 represent nonsense mutation.

Appendix A. Phylogenetic analysis of dicyemids from �-tubulin gene sequences

A.1. Characterization ofβ-tubulin genes from two dicyemid species

Four different�-tubulin sequences were character- ized. Two�-tubulin sequences were obtained fromDi- cyemasp. (BTdv1 and�BTdv2) and the other two se- quences fromDicyemodeca antinocephalum(BTda1 and BTda2). Additionally, a�-tubulin gene was cloned from the hostOctopus vulgaris(BTov1). No indels were observed except one pseudogene (�BTdv2) de- scribed below. One intron in the two sequences from Dicyema sp. and four introns in the two sequences from D. antinocephalumwere identified (Fig. 2). They are all short in length, ranging from 20 to 35 bp. The site of the first intron was identical among the four sequences. On the other hand, octopus�-tubulin gene carries one long intron the site of which coincides with the third intron ofD. antinocephalumsequences. One gene fromDicyemasp. (�BTdv2) is thought to be a pseudogene because of three nonsense mutations in the middle regions and two deletions in the 3′ region of the ORF, leading to frameshift mutation. Accord- ingly, the other three sequences were used for the fol- lowing phylogenetic analysis. No signature sequence shared between dicyemids and animals was detected.

It was confirmed that these clones were truly de- rived from dicyemids by Southern blot analysis using

T. Noto, H. Endoh / BioSystems 73 (2004) 73–83 79

Fig. 3. Southern blot analysis. Each lane contains 1�g of DNA from Dicyemasp. (Ds) or 5�g of DNA from the hostO. vulgaris (Ov), which was digested withHindIII. The filters were hybridized with the radiolabeled BTdv1 probe. A distinct signal is detected only on dicyemid DNA, but not on the host DNA.

one of the cloned homolog as a probe (Fig. 3). In fact, contamination of a small amount of the host tissue is usually unavoidable. However, signals were detected only on the dicyemid DNA by Southern blot analysis, but not on the host DNA. Furthermore, the primers used in this study only amplified a 1.6 kb product from the host DNA, whereas a 1.2 kb product from the di- cyemid DNA (data not shown). Both results clearly indicate that the cloned sequence was derived from the dicyemid DNA. On the other hand, a small num- ber of cells of fairly small kinetoplastids,Bodo sp. is usually identified in culture medium after maintenance in vitro for 1–2 weeks. The sequence from theBodo sp. was amplified only after the second PCR and was placed in a protistan clade, near to kinetoplastids as expected (Fig. 4). These observations strongly suggest that there is a fairly small amount of contamination in our dicyemid DNA preparation to undetectable extent by Southern blot analysis, if any.

A.2. Phylogenetic analysis of theβ-tubulin sequences

The amino acid sequences of�-tubulins from a total of 39 eukaryotes were aligned for phylogenetic analy- sis. As mentioned above, neither the dicyemid nor the host sequences had deletions or insertions except for one pseudogene,�BTdv1. Unlike molecules of vary- ing length, such as rDNA, the consistency of these se-

quences facilitates comparison with others. Based on 378 aligned residues, phylogenies were constructed using the Neighbor-Joining (NJ) and Maximum Parsi- mony (MP) methods (Fig. 4). Because both trees are almost identical, here MP tree was demonstrated with bootstrap values by MP (left) and NJ (right). When Trichomonas vaginaliswas used as outgroup, three major branches with high bootstrap values were gener- ated in the MP tree: protistan-plant clade as indicated previously byEdlind et al. (1996), animals, and fungi including microsporidia as shown byKeeling et al. (2000). These large assemblages were monophyletic (bootstrap probability= 73, 97, and 93%, respec- tively). Fungi formed the sister-group to animals, as in the rRNA phylogeny (Hasegawa et al., 1993). The dicyemid mesozoans were positioned within animals, and outside the protistan-plant clade altogether. Their close affinity with triploblastic animals was supported by high bootstrap values, but the bootstrap confidence level for branchings within animal clade, however, was lower. The exact phylogenic relationship of dicyemids and other animals is still not clear. The phylogeny pro- duced using the ML method was almost identical to the tree presented here although bootstrap value has not been calculated (data not shown). According to our �-tubulin phylogenies, dicyemids group with the triploblastic animals, being consistent with the rela- tionships derived from 18S rDNA data.

Appendix B. Materials and methods

B.1. DNA preparation from dicyemids for PCR

Dicyemids were isolated from two different hosts, O. vulgaris and O. dofleini collected in Notojima is- land, Ishikawa prefecture, Japan (Noto et al., 2003). Three different species of dicyemids,Dicyema japon- icum, D. misakienseand D. acuticephalum, usually inhabit O. vulgaris, whereasO. dofleiniharbors only one species,D. antinocephalum(Furuya, 1999). After isolation and 1 week culture of the dicyemids in Ja- marin seawater (JSW; Jamarin Laboratory) containing 10% fetal bovine serum (FBS; Sigma) and 7.625% (w/v) DME/F-12 HAM mixture (Sigma), 10 whole individuals of dicyemids were washed nine times in JSW using a micropipette to remove all octopus host cells, and then digested in 10 mM Tris–HCl, 10 mM

80 T. Noto, H. Endoh / BioSystems 73 (2004) 73–83

Fig. 4. Molecular phylogenetic consensus tree inferred from�-tubulin amino acid sequences by MP. Parsimony analysis was calculated using the program PROTPARS. Bootstrap resampling was accomplished with the use of the program SEQBOOT (1000 replicates) and CONSENCE. The topology of NJ tree showed almost the same of MP tree. Numbers at each branch indicate bootstrap percentage greater than 50%, which was obtained from MP (left) and NJ (right) analysis. The dicyemid mesozoans were placed within the triploblastic animals supported by high bootstrap values. Accession numbers: AB099885, AB099886, AB099887, AB099888 in dicyemids. AB099884 in the hostO. vulgaris. AB099889 in the kinetoplastidBodo sp.

T. Noto, H. Endoh / BioSystems 73 (2004) 73–83 81

EDTA, 150 mM NaCl, and 0.1% SDS at pH 8.0, con- taining 100�g/ml proteinase K, at 55◦C for 2 h. Total DNA was extracted with phenol/chloroform, precipi- tated in ethanol with 2�l of Pellet Paint Co-Precipitant (Novagen), and dissolved in 10�l of sterilized water. The DNA was divided in two aliquots, one of which was used for every PCR as a template.

B.2. Amplification, cloning, and sequencing of β-tubulin genes from dicyemids

�-Tubulin genes were isolated by PCR using B-TU1F (5′-CARTGYGGYAACCARATYGG-3′) and B-TU2R (5′-TCCATYTCGTCCATRCCYTC-3′) primers designed to amplify approximately∼1.2 kbp fragment which account for more than 80% of the dicyemid �-tubulin gene. A mixture of 25�l of 1× Taq DNA polymerase buffer, 0.2 mM dNTPs, 1 �M of each primer, 5�l of template (sample) DNA solution, and 0.5 U ofTaq DNA polymerase (Sawady) was put in a thermal cycler for 35 cy- cles: each cycle consisted of 60 s at 94◦C, 60 s at 50◦C, and 60 s at 72◦C. Cloning and sequenc- ing strategies were described previously (Noto et al., 2003).

B.3. Southern blot analysis

For confirmation of the source of�-tubulin gene, Southern blot analysis using one of clones (BTdv1) as a probe was carried out as described previously (Noto et al., 2003).

B.4. Sequence alignment and phylogenetic analysis

Amino acid sequences were inferred from the PCR product�-tubulin gene sequences, and aligned with homologs from representative species obtained from GenBank. Alignments (378 residues) of the �-tubulin amino acid sequences fromDicyema sp., D. antinocephalum, O. vulgaris, and other taxa were produced using CLUSTAL X (Thompson et al., 1997). These were then modified by eye to optimize them. Phylogenetic trees were constructed using the NJ and MP methods in the programs CLUSTAL X (Thompson et al., 1997) and PHYLIP version 3.6 (Felsenstein, 2002), respectively. All bootstrap values (Felsenstein, 1985) were based on 1000 replicates.

References

Adoutte, A., Balavoine, G., Lartillot, N., de Rosa, R., 1999. Animal evolution. The end of the intermediate taxa? Trends Genet. 15, 104–108.

Aguinaldo, A.M., Turbeville, J.M., Linford, L.S., Rivera, M.C., Garey, J.R., Raff, R.A., Lake, J.A., 1997. Evidence for a clade of nematodes, arthropods and other moulting animals. Nature 387, 489–493.

Alvarez, F., Cortinas, M.N., Musto, H., 1996. The analysis of protein coding genes suggests monophyly of Trypanosoma. Mol. Phylogenet. Evol. 5, 333–343.

Andersson, J.O., Sjogren, A.M., Davis, L.A., Embley, T.M., Roger, A.J., 2003. Phylogenetic analyses of diplomonad genes reveal frequent lateral gene transfers affecting eukaryotes. Curr. Biol. 13, 94–104.

Bardele, C.F., 1981. Functional and phylogenetic aspects of the ciliary membrane: a comparative freeze-fracture study. Biosystems 14, 403–421.

Bardele, C.F., Huttenlauch, I., Schoppman, H., 1986. The ciliate cortex studied by TEM and SEM cryofracture techniques. Symp. Biol. Hung. 33, 115–124.

Beneden, E. van., 1882. Contribution à l’histoire des Dicyémide. Arch. Biol. Paris 3, 195–228.

Bonner, J.T., 1997. The origins of multicellularity. Integr. Biol. 1, 27–36.

Burns, R.G., 1991.�-, �- and �-tubulins: sequence comparisons and structural constraints. Cell Motil. Cytoskel. 20, 181– 189.

Bushman, F., 2002. Constructing the vertebrate genome: evidence from eels that LINEs mobilize SINEs. Mol. Cell. 10, 961– 962.

Cavalier-Smith, T., 1993. Kingdom protozoa and its 18 phyla. Microbiol. Rev. 57, 953–994.

Cavalier-Smith, T., 1998. A revised six-kingdom system of life. Biol. Rev. Camb. Philos. Soc. 73, 203–266.

Czaker, R., 2000. Extracellular matrix (ECM) components in a very primitive multicellular animal, the dicyemid mesozoa Kantharella antarctica. Anat. Rec. 259, 52–59.

Del Cacho, E., Quilez, J., Lopez-Bernad, F., Sanchez-Acedo, C., Gallego, M., 1996. Identification of a fibronectin-like molecule on the surface ofLeishmaniaamastigotes. Vet. Parasitol. 66, 13–18.

Dodson, E.O., 1956. A note on the systematic position of the Mesozoa. Syst. Zool. 5, 37–40.

Doolittle, W.F., 1998. You are what you eat: a gene transfer ratchet could account for bacterial genes in eukaryotic nuclear genes. Trends Genet. 14, 307–311.

Edlind, T.D., Li, J., Visvesvara, M.H., Vodokin, G.L., McLaughlin, G.L., Katiyar, S.K., 1996. Phylogenetic analysis of�-tubulin sequences from amitochondrial protozoa. Mol. Phylogenet. E 5, 359–367.

Felsenstein, J., 1985. Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39, 783–791.

Felsenstein, J., 2002. PHYLIP (phylogenetic inference package) Version 3.6. University of Washington, Seattle.

82 T. Noto, H. Endoh / BioSystems 73 (2004) 73–83

Furuya, H., Tsuneki, K., Koshida, Y., 1992. Development of the infusoriform embryo ofDicyema japonicum(Mesozoa: Dicyemidae). Biol. Bull. 187, 248–252.

Furuya, H., Tsuneki, K., Koshida, Y., 1994. The development of the vermiform embryos of two mesozoans,Dicyema acuticephalum and Dicyema japonicum. Zool. Sci. 11, 235–246.

Furuya, H., Tsuneki, K., Koshida, Y., 1997. Fine structure of dicyemid mesozoans, with special reference to cell junctions. J. Morphol. 231, 297–305.

Furuya, H., 1999. Fourteen new species of dicyemid mesozoans from six Japanese cephlopods, with comments on host specificity. Species Diversity 4, 257–319.

Furuya, H., Tsuneki, K., 2003. Biology of dicyemid mesozoans. Zool. Sci. 20, 519–532.

Germot, A., Philippe, H., 1991. Critical analysis of eukaryotic phylogeny: a case study based on the HSP70 family. J. Eukaryot. Microbiol. 46, 116–124.

Gray, M.W., Lang, B.F., Cedergren, R., Golding, G.B., Lemieux, C., Sankoff, D., Turmel, M., Brossard, N., Delage, E., Littlejohn, T.G., Plante, I., Rioux, P., Saint-Louis, D., Zhu, Y., Burger, G., 1998. Genome structure and gene content in protist mitochondrial DNAs. Nucleic Acids Res. 26, 865– 878.

Gray, M.W., Burger, G., Lang, B.F., 1999. Mitochondrial evolution. Science 283, 1476–1481.

Grimson, M.J., Coates, J.C., Reynolds, J.P., Shipman, M., Blanton, R.L., Harwood, A.J., 2000. Adherens junctions and beta- catenin-mediated cell signalling in a non-metazoan organism. Nature 408, 727–731.

Halanych, K.M., 1991. 5S ribosomal RNA sequences inappropriate for phylogenetic reconstruction. Mol. Biol. E 8, 249–253.

Hartmann, M., 1907. Untersuchungen uber den Generations- wechsel der dicyemiden. Member. Sci. Acad. Belg. 1, 1– 128.

Hartmann, M., 1925. Mesozoa. In: Kukenthal, W., Krumbach, T. (Eds.), Handbuch der Zoologie. DeGruyter & Co., Berlin, pp. 996–1014.

Hasegawa, M., Hashimoto, T., Adachi, J., Iwabe, N., Miyata, T., 1993. Early branchings in the evolution of eukaryotes: ancient divergence of entamoeba that lacks mitochondria revealed by protein sequence data. J. Mol. E 36, 380–388.

Hyman, L.H., 1940. Phylum Mesozoa. In: The Invertebrates 1. Protozoa through Ctenophora 1. McGraw-Hill, New York, pp. 233–247.

Katayama, T., Wada, H., Furuya, H., Satoh, N., Yamamoto, M., 1995. Phylogenetic position of the dicyemid mesozoa inferred from 18S rDNA sequences. Biol. Bull. 189, 81–90.

Keeling, P.J., Luker, M.A., Palmar, J.D., 2000. Evidence from beta-tubulin phylogeny that microsporidia evolved from within the fungi. Mol. Biol. E 17, 23–31.

Kobayashi, M., Furuya, H., Holland, P.W.H., 1999. Dicyemids are higher animals. Nature 401, 762.

Lapan, E.A., Morowitz, H.J., 1974. Characterization of mesozoan DNA. Exp. Cell Res. 83, 143–151.

Loomis, W.F., Smith, D.W., 1990. Molecular phylogeny of Dictyostelium discoideumby protein sequence comparison. Proc. Natl. Acad. Sci. U.S.A. 87, 9093–9097.

Lopez-Bernad, F., Del Cacho, E., Gallego, M., Quilez, J., Sanchez-Acedo, C., 1996. Identification of a fibronectin- like molecule on Eimeria tenella. Parasitology 113, 505– 510.

McConnaughey, B.H., 1951. The life cycle of the dicyemid Mesozoa. Univ. Calif. Publ. Zool. 55, 295–336.

Müller, W.E.G., 1995. Molecular phylogeny of Metazoa (animals): monophyletic origin. Naturwissenschaften 82, 321– 329.

Noto, T., Yazaki, K., Endoh, H., 2003. Developmentally regulated extrachromosomal circular DNA formation in the mesozoan Dicyema japonicum. Chromosoma 11, 359–368.

Nouvel, H., 1933. Recherches sur la cytologie la physiologie et la biologie des Dicyemides. Ann. L’Inst. Océanogr. Mnaco 13, 165–255.

Nouvel, H., 1948. Les Dicyemides. 2e partie: infusoriforme, teratologie du parasitisme, affinites. Arch. Biol. 59, 147– 223.

Ohama, T., Kumazaki, T., Hori, H., Osawa, S., 1984. Evolution of multicellular animals as deduced from 5S rRNA sequence: a possible early emergence of the Mesozoa. Nucleic Acids Res. 12, 5101–5108.

Pawlowski, J., Montoya-Burgos, J.I., Fahrni, J.F., Wuest, J., Zaninetti, L., 1996. Origin of the masozoa inferred from 18S rRNA gene sequences. Mol. Biol. E 13, 1128–1132.

Prince, V., 2002. The Hox Paradox: more complex(es) than imagined. Dev. Biol. 249, 1–15.

Ridley, R.K., 1968. Electron microscopic studies on dicyemid Mesozoa. I. Vermiform stages. J. Parasitol. 54, 975–998.

Ridley, R.K., 1969. Electron microscopic studies on dicyemid Mesozoa. II. Infusorigen and infusoriform stages. J. Parasitol. 55, 779–793.

Rieger, R.M., 1985. The phylogenetic status of the acoelomate organization within the bilateria; a historical perspective. In: Conway Morris, S., George, J.D., Gibson, R., Platt, H.M. (Eds.), The Origins and Relationships of Lower Invertebrates. Clarendon Press, Oxford, pp. 101–122.

Rodorigo, A.G., Bergquist, P.R., Bergquist, P.L., Reeves, P.R., 1994. Are sponges animals? An investigation into the vagaries of phylogenetic inference. In: Van Soest, R.W.M., Van Kempen, M.G., Braekman, J.C. (Eds.), Sponge in Time and Space. Balkema, Rotterdam, pp. 47–54.

Saccone, C., Gissi, C., Lanave, C., Larizza, A., Pesole, G., Reyes, A., 2000. Evolution of the mitochondrial genetic system: an overview. Gene 261, 153–159.

Schubbert, R., Hohlweg, U., Renz, D., Doerfler, W., 1997. Foreign (M13) DNA ingested by mice reaches peripheral leukocytes, spleen, and liver via the intestinal wall mucosa and can be covalently linked to mouse DNA. Proc. Natl. Acad. Sci. U.S.A. 94, 961–966.

Schubbert, R., Renz, D., Schmitz, B., Doerfler, W., 1998. On the fate of orally ingested foreign DNA in mice: chromosomal association and placental transmission to fetus. Mol. Gen. Genet. 259, 569–576.

Stunkard, H.W., 1954. The life history and systematic relations of the mesozoa. Q. Rev. Biol. 29, 230–244.

Thompson, J.D., Gibson, T.J., Plewniak, F., Jeanmougin, F., Higgins, D.G., 1997. The CLUSTAL X windows interface:

T. Noto, H. Endoh / BioSystems 73 (2004) 73–83 83

flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res. 25, 4876–4882.

Watanabe, K., Bessho, Y., Kawasaki, M., Hori, H., 1999. Mitochondrial genes are found on minicircle DNA molecules in the mesozoan animalDicyema. J. Mol. Biol. 286, 645–650.

Whitman, C.O., 1883. A contribution to the embryology, life history, and classification of the dicyemid. Mittheilungen aus der Zoologischen Station zu Neapel 4, 1–89.

Willmer, P., 1990. Invertebrate Relationships: Patterns in animal evolution. Cambridge University Press.