BioSystems, 25 (1991) 75--83
75
Elsevier Scientific Publishers Ireland Ltd.
The troublesome parasites -- molecular and morphological evidence that Apicomplexa belong to the dinoflagellate-ciliate clade JSrn Wolters Institut fi~r Allgemeine M~krobiologie, Christian-Albrechts-Universit(~t Kiel, Olshausenstrasse 40, D-2300 Kiel 1 (Germany) Large insertions and deletions in the variable regions of eukaryotic 16S-like rRNA relative to the archaebacterial structure have been defined as a marker for rapidly evolving taxa. Deletions in the rRNA occur in the diplomonad Giardia and the microsporidian Vairimorpha, whereas insertions occur in Euglenozoa (Euglena and the kinetoplastids), Acanthamoeba, Naegleria, Physarum, Dictyostelium, the apicomplexan Plasmodium, the ciliate Euplotes, and some metazoa. Except Acanthamoeba and Euplotes, all of these protists were previously placed at the base of the eukaryote phylogeny. A re-analysis of the 16S-like rRNA and 5S rRNA data with the neighborliness method revealed a close relationship of Apicomplexa to the dinoflagellate-ciliate clade, most probably closer to the dinofiagellates. Morphological evidence that supports this grouping is the layer of sacs underneath the plasma membrane in all three taxa and the identical structure of trichocysts in the apicomplexan Spiromonas and dinofiagellates. The remaining rapidly evolving organisms might still be misplaced in the 16S-like rRNA trees.
Keywords: Phylogeny; Rioosomal RNA; Apicomplexa; Dinoflagellates; Ciliates.
1. I n t r o d u c t i o n
Within the last decade a large amount of molecular sequence data has been gathered for a great variety of eukaryotes. Since ribosomal RNA is essential for a metabolically independent organism, it is universally distributed from archaebacteria to humans. The largest data sets exist for 5S rRNA, vdth about 125 positions and 319 eukaryotic sequences (Specht et al., 1990), and for 16S-like rRNA, with about 1200 informative positions and 62 eukaryotic sequences (Neefs et al., 1990). Phylogenetic tree.~ derived from 5S rRNA sequence data by unweighted pair group matrix analysis (UPGMA) (Hori and Osawa 1986) always show rhodophytes to be the most primitire eukaryotes followed by the chlorobionts (green algae and land plants sensu Bremer 1985), while Acanthamoeba, Physarum and Dictyostelium as well as the Euglenozoa (euglenids
and kinetoplastids sensu Cavalier-Smith 1981) emerge later. In contrast, phylogenetic trees derived from 16S-like rRNA sequence data by a Fitch-Margoliash method (Sogin et al., 1989; Hendriks et al., 1989) show Acanthamoeba to branch off from the chlorobiont lineage while Dictyostelium, Physarum as well as the Euglenozoa branch off early in the eukaryote phylogeny. How can these different topologies be explained? One is easily persuaded to believe in the 16S-like rRNA data because of the larger number of characters. Of course, a larger number of characters could give a more detailed picture of the phylogeny, but can the 5S rRNA data be completely ignored? A crude cladistic analysis of the 16S-like rRNA data has revealed a quite different position for the Euglenozoa: next to the Eumycota and Metazoa (Wolters and Erdmann, 1986). The present communication focuses on whether these organisms truly show ancestral eukaryotic
0303-2647/91/$03.50 © ]991 Elsevier Scientific Publishers Ireland Ltd. Published and Printed in Ireland
76
traits or whether they might be misplaced by the tree-construction method due to their extremely high evolutionary rates. 2. Materials and methods
The 5S rRNA sequences were obtained in aligned format from the Berlin RNA databank (Specht et al., 1990). Over 430 16S rRNA sequences were aligned using the alignment editor ALMA (Thirup and Larsen, 1990) and manually refined according to common secondary structural elements. Hypervariable regions with length variation in which unambiguous alignment was not possible have been omitted (between positions 0--6, 63--104, 178--221, 405--493, 587--610, 630--652, 837--849, 996--1046, 1129--1143, 1165--1171, 1255--1258, 1435--1465 and 1540-end in the E.coli numbering system), thus reducing the total number of positions to 1380. One hundred eighty positions were universally conserved or showed one single deviation among all three primary kingdoms (excluding mitochondria). Thus, 1200 informative positions were obtained. Phylogenetic trees were constructed from datasets of 84 5S rRNA and 43 16S-like rRNA sequences using a distance-matrix analysis called the neighborliness method (Sattath and Tversky, 1977; Fitch, 1981). 3. Results and discussion
In order to detect high evolutionary rates, attention was focused on the variable regions of the 16S-like rRNA. These regions are usually omitted in the construction of phylogenetic trees because unambiguous alignments are impossible. Beyond a general variability in this region, some species exhibit more extensive insertions or deletions. In five of the nine variable regions defined by Neefs et al. (1990), the archaebacteria show a conserved feature without any considerable length variation and, together with eubacteria, are used as an outgroup representing the primitive state. Large insertions and large deletions then represent a derived state. The number of these insertions and
deletions gives a measure of the evolutionary rate of these organisms. Table 1 lists large insertions and deletions relative to archaebacteria present in various eukaryotic groups. In variable region V7, helix 41 (nomenclature of Neefs et al., 1990) is considerably larger in the Euglenozoa, in the rhizopods Acanthamoeba, Naegleria, Physarum and Dictyostelium, and in the apicomplexan Plasmodium. The only non-protistan group exhibiting this feature is the insects represented by Drosophila and Tenebrio. Variable region V8, comprising helices 43 and 44, shows larger versions in all of the above protistan taxa excluding Dictyostelium. In some species the extra nucleotides can fold into a helix E43-1. Various other insertions in variable regions V2, V4, and V5 are present in some but not all of the seven protistan taxa. In variable region V4, helix E21-9 and an insertion in helix E21-3 are supporting the monophyly of the Euglenozoa. To indicate a higher rate of evolution by using the number of sites with insertions (Table 1), Euglena has the most rapid rate (10 sites), followed by the kinetoplastids (8), Acanthamoeba (7), Naegleria (5), Physarum (4), Plasmodium (3), Euplotes (3), all Metazoa (1), insects (additional 1), mammals (additional 1), and Dictyostelium (1). Because of the well known fact that higher evolutionary rates tend to draw taxa down to the base of a phylogenetic tree (Felsenstein, 1978), all of the above taxa are candidates for potential misplacement. In addition to the evolutionary trend of increase in genetic material apparent in the great majority of eukaryotes, a decrease in genetic material is found in the diplomonad Giardia and the microsporidian Vairimorpha. In variable regions VS, V7 and VS, Giardia and Vairimorpha have helices not only shorter than in all other eukaryotes but also considerably shorter than in all archaebacteria. In variable region V2, Giardia exhibits the eukaryote-specific insertion of 8--9 nucleotides between helices 8 and 9, while Vairimorpha shows an intermediate feature of one nucleotide. Within helix 10, all eukaryotes excluding Giardia and Vairimorpha have two extra helices E10-1 and El0-2. In
?7
~,
~ ~'~
I
I
I
I
I ~
I I I~o
ogN
"~
"~x.
~
~
.~
~~
m
I
°°~'~'~
"~
I
I
I
I
I ~
I=
I
+
~
I ~
,..~
~ ~o ~
I
I I I I I I I I I I I I I I I I I I I
~V
.~
I
+
~ .,~
o
,-.-i
+
-I- +
+
o~ o~
o ~ ~.~
0~ ~
0
I
I
I
I
I
I
~
~
~
~
II
I
~ "~.~ ~ ii
~z
~.~.
O0~ r~
"~ ,
~
0
~
.
V .--k
~ ~
o
"~
~o° o •
~
~
78
Giardia, helix 10 is considerably shorter than in archaebacteria. In Vairimorpha, all of helix 10 and the accompanying helix 11 are deleted. To indicate a higher rate of evolution by using the number of sites with deletions in just these five variable regions (Table 1), Vairimorpha has the most rapid rate (5 sites), followed by Giardia (4). Therefore, these two amitochondrial protists are also candidates for potential misplacement. A mixture of insertions and deletions was not found. This unexpected finding points to the possible existence of two exclusive evolutionary trends, increase and decrease of genetic material. 16S-like rRNA analyses by various laboratories (Sogin et al., 1989; Hendriks et al., 1989) place most of these rapidly evolving organisms at the base of their eukaryotic tree.
In a tree using about 900 positions (Sogin et al., 1989) calculated with a Fitch-Margoliash distance matrix method (Elwood et al., 1985), the taxa branch off in the order Giardia, Vairimorpha, Euglenozoa, Naegleria, Entamoeba (unpublished sequence), Dictyostelium and Plasmodium. In a tree using all positions of the alignment (Hendriks et al., 1989) and calculated with basically the same method (algorithm by DeSoete 1983), the taxa branch off in the order
Giardia, Vairimorpha, Naegleria, Physarum, Euglena, kinetoplastids, Plasmodium and Dictyostelium. This more recent tree contradicts an earlier one (Hendriks et al., 1988) which, like Sogin et al. (1989), placed Euglena and kinetoplastids as sister groups and Plasmodium as the sister group of ciliates. The present analysis of the 16S rRNA data
Giardia
lamblia Vairimorpha n e c a t r i x EUGLENOZOA
Naegleria gruberi Physarum polycephalum D J c t y o s t e l i u m dJscoideum G r a c i l a r i a verrucosa Prorocentrum micans [--~%~ Plasmodium spec. O x y t r i c h a nova DINOZOA Euplotes aediculatus Paramecium t e t r a u r e l i a Tetrahymena thermophila Ochromonas dan|ca Skeletonema costatum CIIROMOPIIYTA Achlya b i s e x u a l i s canthamoeba c a s t e l l a n i i CHLOROBIONTA ASCOMYCOIA
I0
\ MEIAZOA
Fig. 1. Phylogenetie tree of 40 eukaryotes deduced from about 1200 positions of 16S-like rRNA employing the neighborliness method. Thermoproteus tenax, Thermococcus celer and Thermoplasma acidophilum were used as outgroup. Euglenozoa are represented by Euglena gracilis, Crithidia fasciculata, Leishmania donovani and Trypanosoma bruce|, Plasmodium by P. berghei asexual, P. berghei sporozoite, P. falciparum asexual, P. falciparum sporozoite and P. lophurae asexual, Chlorobionta by Chlamydomonas reinhardtii, Chlorella vulgar|s, Nanochlorum eucaryotum, Zamia pumila and Glycine max, Ascomycota by Saccharomyces cerevisiae, Neurospora crassa, and Pneumocystis carinii and Metazoa by Caenorhabditis elegans, Eurypelma californica, A rtemia salina, Drosophila melanogaster, Tenebrio molitor, Xenopus lacy|s, Mus musculus and Homo sapiens. The length bar indicates 10 changes.
79
was restricted to 1380 positions and employed the neighborliness method (Sattath and Tversky 1977, Fitch 1981). In simulation studies (Saitou and Nei 1987; Sourdis and Nei, 1988; Saitou and Imanishi, 1989), this method and the neighborjoining method (Saitou and Nei, 1987) performed slightly better than the Fitch-Margoliash and parsimony methods in recovering a given tree topology under a varying rate of evolution. The phylogenetic tree (Fig. 1) is congruent with that of Sogin et al. (1989) in the placements of Euglena and the kinetoplastids as sister groups and in the placement of dinoflagellates and ciliates as sharing a common ancestor. The apicomplexan Plasmodium, however, is placed as the sister group of the dinoflagellate Prorocentrum. In this respect, the tree is more similar to that of Hendriks et al. (1988). Because the tree of Sogin et al. (1989) includes unpublished data, it is not possible to test whether the different tree topologies are due to the different number of positions used, to the different range of organisms, to the different treeing algorithms, or to a combination of the above. Some rapidly evoMng organisms, as defined by the number of sites with insertions or deletions (Table 1), are always placed late in the eukaryotic tree, e.g. the hypotrichous ciliate Euplotes (Fig. 1). In this case, the sequence of a slower evolving close relative is known, namely Oxytricha. This ensures the correct placement of Euplotes; the high rate of evolution is expressed in its branch length, which is twice as long as its closest relative Oxytricha. A second example is Acanthamoeba, which is always placed late in the eukaryotic tree as the sister group of the Chlorobionta (Fig. 1). The Metazoa are placed as the sister IFoup of Eumycota but can jump to the base of the higher eukaryotes (after the emergence of rhodophytes) depending on the selection of organisms (tree not shown). The position closer to the base is easily explained by the higher evolutionary rate manifested by insertions in the variable regions and therefore considered as an artefact. The remaining taxa of fast-evolving organisms are placed at the base of the eukaryotic tree suggesting that they represent
more primitive eukaryotes. We expect to find, in primitive eukaryotes, features intermediate between those of archaebacteria and those of the majority of higher eukaryotes. None of the taxa (Table 1) have such intermediate features in their variable regions. Interestingly, most of the fast-evolving taxa at the base of the eukaryotic tree are either parasites or rhizopods. Until sequences of non-parasitic relatives are available, this peculiar positioning must be considered uncertain. Their high rate of evolution might, indeed, draw them down to the base of the tree. For the Apicomplexa, this seemed to be the case in the trees of Sogin et al. (1989) and Hendriks et al. (1989). The present analysis is indeed adding support to an earlier analysis (Wolters and Erdmann, 1986) which places the organisms with the most extreme rate of evolution, namely the Euglenozoa, much further up in the tree. Other taxa might be rearranged when more data become available. If, however, the number of reversals (back mutations) in a specific group is higher than the number of derived characters (synapomorphies) it shares with its closest relative, then every treeing method will misplace this organism too far toward the base of the tree. Single nucleotide insertions and deletions are far less frequent than ordinary base substitutions and deserve special attention in the construction and discussion of phylogenetic trees. Table 2 lists all the single-nucleotide insertions and deletions relevant to the discussion of the troublesome group of fast evolving protist rRNAs. The only position congruent with the trees of Sogin et al. (1989), Hendriks et al. (1989) and the present one (Fig. 1) is position 1120.1. Some positions (Table 2) indicate that Naegleria and Physarum might be sister groups: a unique insertion at position 268.1. (E. coli numbering system) and the absence of a nucleotide at positions 287.1 and 860.1 as in archaebacteria. This alternative topology is not surprising since their branch points are unresolved. An unpaired nucleotide in helix 3 at position 29.1 is common for most eukaryotes. According to the primary structure, subsequent deletions often hit the neighboring position 29 as
80
TABLE
2
Single insertions/deletions in recognized monophyletic taxa of eukaryotes in comparison to archaebacteria. Exceptions within groups are indented. Positions are numbered according to the E . c o i l s e q u e n c e . L o w e r c a s e letters denote low frequency o f occurrence. Position
29/29.1
250.1/250.2
268.1
287.1
860.1
870/871
1120.1
1418.1
EUBACTERIA
N
.
.
.
.
.
.
nU
gU
--
ARCHAEBACTERIA
Y
.
.
.
.
.
.
aU
gU
--
-
Giardia
G
.
.
.
.
G
G
-
--
--
C
Vairimorpha
U
.
.
.
.
A
A
--
--
DEL
--
Euglena
U
C
--
--
--
A
C
--
A
--
-
Kinetoplastida Acanthamoeba
U
C
G
U
--
A
C
U
gU
--
--
U
C
U
C
--
A
U
--
U
G
U
Naegleria
U
--
--
--
A
-
--
--
U
--
U
Physarum Dictyostelium Plasmodium
U
--
--
--
A
--
--
--
C
U
U
U
C
U
C
--
A
U
--
U
G
U
U
C
U
U
--
A
U
--
--
G
U
Rhodophyt#
U
--
U
G
--
A
C
--
--
G
U
Prorocentrum Ciliophora Euplotes
U
C
C
C
--
A
U
--
--
G
U
U
C
U
C
--
A
acU
-
-
R
U
Tetrahymena
--
--
G
-
--
G
---
C
Chromophyta
U
C
gU
G
--
A
cU
Ochromonas Pyrenomonas ~' Chlorobionta
-U
C
U
C
--
G
U
U
C
U
uC
--
G
U
--
--
G
U
Pneumocystis
U
C
U
C
-
A
U
--
--
G
U
Ascomycota Metazoa Caenorhabditis
U
C
U
C
--
A
C
--
--
R
u
U
C
U
uC
--
R
aC
-
--
R
cU
--
G
--
C U
--
A rtem ia
Chordata
~Bird et al. (1990). bEschbach et a l . ( 1 9 9 0 ) .
in ciliates, chromophytes and metazoa. It is important to point out that besides in Giardia, Vairimorpha, Naegleria and Physarum, position 29.1 is absent in rhodophytes. A reversal is obvious at position 870 for the kinetoplastids. The primitive condition of positions 250.1 and 250.2 is present in Giardia, Vairimorpha, Euglena (but not in kinetoplastids), Naegleria, Physarum, Euplotes, and Caenorhabditis, while the primitive condition of position 1418.1 occurs in Vairimorpha (but not in Giardia), Euglenozoa and Tetrahymena. All these positions are incompatible with each other so that a
large number of parallel, mostly back mutations have to be assumed in all three phylogenetic trees. Of special importance is position 871: the presence of a nucleotide in Euglenozoa and in the four rhizopods (Acanthamoeba, Naegleria, Physarum and Dictyostelium) represent the primitive character state as in archaebacteria, while Giardia and Vairimorpha like other eukaryotes exhibit a deletion. In the present tree as well as in those of Sogin et al. (1989) and Hendriks et al. (1989), the minimum number of mutations is 4. Interestingly, this position unites
81
Acanthamoeba with all the fast-evolving insertion-bearing organisms found at the base of these trees (except the apicomplexan Plasmodium, which groups with the dinoflagellate-ciliate clade in the neighborliness method). A common origin of the four rhizopod protists is congruent with the 5S rRNA analysis of Hori and Osawa (1986). These troublesome organisms might indeed be misplaced and might all belong to the position occupied by Acanthamoeba, thereby reducing the minimum number of mutations at position 871 to 2. The absence of morphological characters, like mitochondria (as in Diplomonadida and Microsporidia) or flagella (as in Rhodophyta and Eumycota), can be interpreted as either primitive or secondarily derived. Molecular data per se do not provide better information. The absence of a nucleot:ide at a specific position (as in the outgroup) can also be interpreted as either primitive or secondarily derived. In this respect,
ribosomal RNA and morphological data manifest the same problem. Insufficient molecular data can also result in wrong topologies. Therefore, it is not yet justified to use molecular data as proof for alternative morphological hypothesis. For example: if all the above mentioned taxa with higher evolutionary rates are indeed misplaced in the phylogenetic trees and find their proper place much further towards the tip of the tree, then rhodophytes become the most primitive eukaryotes and their lack of flagella becomes a primarily primitive trait. A re-analysis of the 5S rRNA data with the neighborliness method (Fig. 2) obtained results similar to those for 16S rRNA data. Here, the Apicomplexa emerge first before the dinoflagellates and the ciliates. Morphological evidence for a common ancestry of the three groups is currently emerging: the provocative similarity of the inner membrane complex in Apicomplexa (Vivier and Desportes, 1989) with
TIIRAUSTOCIIYTRIACEA
RHODOPItYTA
--[
F Plasmodium b e r g h e i Plasmodium f a l c i p a r u m Cryptl~ecodinium c o h n i i B l e p h a r i s m a japonicum l e t r a h y m e n a thermop|mi|a DINOZOA Paramecium t e t r a u r e l i a Bresslaua vorax Euplotes woodruffi E u p l o t e s eurysLomus
L
Mastigamoeba i n v e r t a n s CItLOROB|ONTA i
OTIIER EUKARYOTES Fig. 2. Excerpt of a phylogenetic tree of 78 eukaryotes deduced from 125 position~ of 5S rRNA employing the neighborliness method. Sulfolobus acidocaldarius, Desulfurococcus mobilis, Pyrodictium occultum, Thermococcus celer, Pyrococcus woesei and Thermoplasma acidophilum were used as outgroup. Thraustochytriacea are represented by Thraustochytrium visurgense and Schizochytrium aggregatum, Rhodophyta by Porphyra yezoensis, Palmaria palmata and Gloiopeltis complanata, and Chlorobionta by Chlamydomonoz sp., Spirogyra sp., Scenedesmus obliquus, Nitella flexilis, Lophocolea heterophylla and Spinacia oleracea. The length bar indicates 1 change.
82
the amphiesma in dinoflagellates, and with the alveoli in ciliates suggests that these layers of sacs underneath the plasma membrane are homologuous structures. A phylogenetic relationship between dinoflagellates and ciliates was first suggested by Taylor (1976). Lynn and Small (1981) suggested that the ciliates derived from a corticoflagellate ancestor with alveoluslike membranous systems beneath the plasma membrane and similar kinetid structures, such as dinoflagellates or the unusual flagellate Colponema (Mignot and Brugerolle, 1975). A common ancestry of dinoflagellates and ciliates was substantiated through 5S rRNA data by Kumazaki et al. (1983) and 16S-like rRNA by Gunderson et al. (1987). Brugerolle and Mignot (1979) discovered a flagellate originally named Spiromonas, which later proved to be indistinguishable from Colpodella Cienkowski 1865 (Patterson and ZSlffel, 1991). Colpodella and another flagellated apicomplexan Perkinsus both bear flagella and might represent missing links between nonflagellated apicomplexans and dinoflagellates. Brugerolle and Mignot (1979) pointed out that the trichocysts of Colpodella are in fact identical in structure to the ones found in dinoflagellates but differ considerably from those of ciliates. This presumed synapomorphy supports the 16Slike rRNA neighborliness tree (Fig. 1) rather than the 5S rRNA tree (Fig.2). Cavalier-Smith (1987) used other features, like ampulliform mitochondrial cristae and a 1-step meiosis, to unite dinoflagellates and apicomplexans in a taxon named Miozoa. Since only one complete 16S-like rRNA sequence from dinoflagellates is available at present, it is unclear whether Apicomplexa are the sister group of, or originated within, dinoflagellates. Apicomplexa might be derived from parasitic genera that exhibit histones as found in the order Syndiniales. Partial 16S-like rRNA sequences from other apicomplexans, Toxoplasma, Sarcocystis and Cryptosporidium (Johnson et al., 1987, 1988, 1990), and from a protist of uncertain affiliation, Blastocystis (Johnson et al., 1989) were obtained by reverse transcription. An analysis of 264 to 369 positions with various tree-construction methods
resulted in different trees, some of which do not support a monophyly of Apicomplexa as the sister group of the dinoflagellate. Taking into account the above problems arising with complete 16S rRNA sequences, any conclusion from a much smaller fraction (Johnson et al., 1987, 1988, 1989, 1990), namely the rejection of an apicomplexan monophyly, seems to be premature. The erection of a supraphyletic taxon named Dinozoa comprising unicellular protists with layers of sacs underneath the plasma membrane (alveoli, amphiesma) is proposed. The term Dinozoa has been used as synonym for dinoflagellates or Dinophyta by Cavalier-Smith (1981) but has not been widely accepted by the scientific community. A redefinition of the term Dinozoa to include related taxa, equivalent to the term Euglenozoa (Cavalier-Smith, 1981), seems justified. The rank of such a taxon (Superphylum, Subkingdom, or Kingdom) should not be specified until an outline of eukaryotic systematics is settled. At present the Dinozoa consists of three recognized phyla, namely Dinoflagellata, Ciliophora and Apicomplexa. If the Apicomplexa are proved to have originated within the phylum DinoflageUata, they must be given a lower rank.
Acknowledgments I wish to thank D.J. Patterson, G. Brugerolle and F.J.R. Taylor for valuable discussions concerning the morphological links between Apicomplexa, dinoflagellates and ciliates, and M. Ragan for providing the Gracilaria verrucosa 16S rRNA sequence prior to publication.
References Bird, C.J., Rice, E.L., Murphy, C.A., Liu, Q.Y. and Ragan, M.A., 1990, Nucleotide sequences of 18S ribosomal RNA genes from the red algae Gracilaria tikvahiae McLachlan, Gracilaria verrucosa (Hudson) Papenfuss and Gracilariopsis sp.. Nucl. Acids Res. 18, 4023. Bremer, K., 1985, Summary of green plant phylogeny and classification. Cladistics 1, 369--385. Brugerolle, G. and Mignot, J.P., 1979, Observations sur le cycle l'ultrastructure et la position syst~matique de
83
Spiromonas perforans (Bodo perforans Hollande 1938), flagell~ parasite de Chilomonas paramaecium: ses relations avec les dincflagelles et sporozoaires. Protistologica 15, 183--196. Cavalier-Smith, T., 1981, Eukaryote kingdoms: seven or nine? BioSystems 14, 461--481. Cavalier-Smith, T., 1987, The origin of eukaryote and archaebacterial cells. Ann. N.Y. Acad. Sci. 503 (Endocytobiology III), 17--54. DeSoete, G., 1983, A least squares algorithm for fitting additive trees to proximity data. Psychometrika 48, 621--626. Elwood, H.J., Olsen, G. and Sogin, M.L., 1985, The smallsubunit ribosomal RNA gene sequences from the hypotrichous ciliates Oxytricha nova and Stylonychia pustulata. Mol. Biol. Evol. 2, 399--410. Eschbach, S., Wolters, J. and Sitte, P., 1991, Primary and secondary structure of the small subunit ribosomal RNA of the cryptomonad Pyrenomonas salina as inferred from the gene sequence: evolutionary implications. J. Mol. Evol., in press. Felsenstein, J., 1978, Cases in which parsimony or compatibility methods will be positively misleading. Syst. Zool. 27, 401--410. Fitch, W.M., 1981, A noa-sequential method for constructing trees and hierarchical classifications. J. Mol. Evol. 18, 30--37. Gunderson, J.H., Elwood, H., Ingold, A., Kindle, K. and Sogin, M.L., 1987, Phylogenetic relationships between chlorophytes, chrysophytes, and oomycetes. Proc. Natl. Acad. Sci. USA 84, 5323--5827. Hendriks, L., van Broeckhoven, C., Vandenberghe, A., van de Peer, Y. and DeWachter, R., 1988, Primary and secondary structure of tile 18S ribosomal RNA of the bird spider Eurypelma californica and evolutionary relationships among eukaryotic phyla. Eur. J. Biochem. 177, 15--20. Hendriks, L., Goris, A., Neefs, J.M., van de Peer, Y., Hennebert, G. and DeWachter, R., 1989, The nucleotide sequence of the small ribosomal subunit RNA of the yeast Candida albicans and the evolutionary position of the fungi among the eukaryotes. System. Appl. Microbiol. 12, 223--229. Hori, H. and Osawa, S., 1986, Evolutionary change in 5S RNA secondary structure and a phylogenetic tree of 352 5S RNA species. BioSystems 19, 163--172. Johnson, A.M., Fielke, R., Lumb, R. and Baverstock, P.R., 1990, Phylogenetic relationships of Cryptosporidium determined by ribosomal RNA sequence comparison. Int. J. Parasitol. 20, 141-147. Johnson, A.M., Illana, S., Hakendorf, P. and Baverstock, P.R., 1988, Phylogenetic relationships of the apicomplexan protist Sarcocysti,~ as determined by small subunit ribosomal RNA comparison. J. Parasitol. 74, 847--860. Johnson, A.M., Murray, P.J., Illana, S. and Baverstock, P.R., 1987, Rapid nucleotide sequence analysis of the small subunit ribosomal RNA of Toxoplasma gondii:
evolutionary implications for the Apicomplexa. Mol. Biochem. Parasit. 25, 239--246. Johnson, A.M., Thanou, A., Boreham, P.F.L. and Baverstock, P.R., 1 9 8 9 , Blastocystis hominis: Phylogenetic affinities determined by rRNA sequence comparison. Exp. Parasitol. 68, 283--288. Kumazaki, T., Hori, H. and Osawa, S., 1983, Phylogeny of Protozoa deduced from 5S rRNA sequences. J. Mol. Evol. 19, 411--419. Lynn, D.H. and Small, E.B., 1981, Protist kinetids: structural conservatism, kinetid structure, and ancestral states. BioSystems 14, 377--385. Mignot, J.P. and Brugerolle, G., 1975, Etude ultrastructurale du flagell~ phagotrophe Colponema loxodes Stein. Protistologica 11, 429--444. Neefs, J.M., van de Peer, Y., Hendriks, L. and DeWachter, R., 1990, Compilation of small ribosomal subunit RNA sequences. Nucl. Acids Res. 18, 2237--2317. Patterson, D.J. and Z51ffel, M., 1991, Heterotrophic flagellates of uncertain taxonomic position, in: The Biology of Free-Living Heterotrophic Flagellates, D.J. Patterson and J. Larsen (eds.) (Oxford University Press, Oxford) in press. Saitou, N. and Imanishi, T., 1989, Relative efficiencies of the Fitch-Margoliash, maximum-parsimony, maximumlikelihood, minimum-evolution, and neighbour-joining methods of phylogenetic tree construction in obtaining the correct tree. Mol. Biol. Evol. 6, 514--525. Saitou, N. and Nei, M., 1987, The neighbour-joining method: a new method for reconstructing phylogenetic trees. Mol. Biol. Evol. 4, 406--425. Sattath, S. and Tversky, A., 1977, Additive similarity trees. Psychometrika 42: 319--345. Sogin, M.L., Edman, U. and Elwood, H., 1989, A single kingdom of eukaryotes, in: The Hierarchy of Life, B. Fernholm, K. Bremer and H. Joernwall (eds.) (Elsevier, Amsterdam) pp. 133--143. Sourdis, J. and Nei, M., 1988, Relative efficiencies of the maximum parsimony and distance-matrix methods in obtaining the correct phylogenetic tree. Mol. Biol. Evol. 5, 298--311. Specht, T., Wolters, J. and Erdmann, V.A., 1990, Compilation of 5S rRNA and 5S rRNA gene sequences. Nucl. Acids Res. 18, 2215--2229. Taylor, F.J.R., 1976, Flagellate phylogeny: a study in conflicts. J. Protozool. 23, 28--40. Thirup, S. and Larsen, N., 1990, ALMA, an editor for large sequence alignments. Proteins 7, 291--295. Vivier, E. and Desportes, I., 1989, The phylum Apicomplexa, in: Handbook of Protoctista, L. Margulis, J.O. Corliss, M. Melkonian and D.J. Chapman (eds.) (Jones and Bartlett, Boston, MA) pp. 549--571. Wolters, J. and Erdmann, V.A., 1986, Cladistic analysis of 5S rRNA and 16S rRNA secondary and primary structure -- the evolution of eukaryotes and their relation to archaebacteria. J. Mol. Evol. 24, 152--166.