Origin and evolution of the
slime molds (Mycetozoa)
Sandra L.
Baldauf and W. Ford Doolittle
Canadian
Institute for Advanced Research and Department of Biochemistry,
To whom
reprint requests should be addressed. e-mail: sbaldauf@is.dal.ca.
Communicated
by John C. Avise,
review
This
article has been cited by other articles in PMC.
Abstract
The
Mycetozoa include the cellular (dictyostelid), acellular (myxogastrid), and
protostelid slime molds.
However,
available molecular data are in disagreement on both the monophyly and
phylogenetic
position of
the group. Ribosomal RNA trees show the myxogastrid and dictyostelid slime
molds as
unrelated
early branching lineages, but actin and β-tubulin trees place them together
as a single
coherent
(monophyletic) group, closely related to the animal–fungal clade. We have
sequenced the
elongation
factor-1α genes from one member of each division of the Mycetozoa, including
Dictyostelium
discoideum, for which cDNA sequences were previously available. Phylogenetic
analyses of
these sequences strongly support a monophyletic Mycetozoa, with the myxogastrid
and
dictyostelid
slime molds most closely related to each other. All phylogenetic methods used
also place
this
coherent Mycetozoan assemblage as emerging among the multicellular eukaryotes,
tentatively
supported
as more closely related to animals + fungi than are green plants. With our data
there are now
three
proteins that consistently support a monophyletic Mycetozoa and at least four
that place these
taxa within
the “crown” of the eukaryote tree. We suggest that ribosomal RNA data should be
more
closely
examined with regard to these questions, and we emphasize the importance of
developing
multiple
sequence data sets.
Introduction
Olive
defines the Mycetozoa as consisting of three distinct groups (1). The true or
plasmodial slime
molds
(Myxogastria—e.g., Physarum polycephalum) are amoeboflagellates, most of which
develop
into large,
reticulate plasmodia with >104 synchronously dividing nuclei. The cellular
slime molds
(Dictyostelia—e.g.,
Dictyostelium discoideum) are strictly amoeboid, and, under conditions of
nutrient
starvation, aggregate to form large, motile, multicellular slugs (1). The
Protostelia, first
described
in the 1960s (2), are mostly microscopic but morphologically diverse organisms,
with
different
taxa exhibiting various combinations of myxogastrid- and/or dictyostelid-like
traits (1). All
Mycetozoa
share a structurally similar fruiting body consisting of a cellulosic stalk of
one to many sterile
cells
supporting the spore-bearing sori (1). A fourth group of “slime molds,” the
Acrasids, now appear
to be
entirely unrelated, on the basis of both ultrastructural (1) and molecular (3)
data.
Since the
slime molds were first described in the mid-1800s, opinions on the monophyly
and
phylogenetic
affinity of these organisms have varied widely. The striking contrasts in the
trophic stages
of the
myxogastrids and dictyostelids have often led to their being classified as
entirely unrelated.
Furthermore,
the motile slug stage of the dictyostelids, the fungal-like plasmodia of the
myxogastrids,
and the
plant-like fruiting bodies of both have led them, in whole or in part, to be
classified as plants,
animals, or
fungi. In his original five-kingdom scheme of life, Whittaker placed the slime
molds together
at the base
of the fungi (4), while admitting that they stuck out of his mitten scheme
“like a sore thumb”
(5). Olive,
however, argued that the slime molds have little in common with fungi and
should be
classified
as protists (6).
Molecular
phylogenies of rRNA genes show little or no support for a coherent Mycetozoa.
In addition,
these
analyses usually show Physarum as arising early in the tree, among the first
“mitochondriate”
eukaryotes.
These studies include analyses of the small subunit (SSU) or 16S-like rRNA
using whole
sequences
(7) or universally alignable portions only (8), as well as analyses of the
large subunit
(23S-like)
rRNA (9) and 5S rRNA (10). In contrast, actin and β-tubulin trees place Physarum and
Dictyostelium
together with generally high confidence (11–14). Furthermore, these trees,
along with
trees of α-tubulin
(11, 14), RNA polymerase largest subunit (15), and glyceraldehyde-3-phosphate
dehydrogenase
(3), all place the represented mycetozoans among the multicellular eukaryotes,
consistently
closer to animals + fungi than are green plants in all but the polymerase
trees. This position
is also
supported for Dictyostelium alone with a combined maximum likelihood analysis
of 19 proteins
(16).
The protein
synthesis elongation factor-1α (EF-1α) appears to be well suited for
deep-level phylogeny
due to its
slow rate of sequence evolution, its single or low copy number in all taxa
examined to date,
and the
fact that the eukaryote EF-1α tree can be rooted by using closely related
archaebacterial
homologs
(17). To evaluate the origin and possible phylogenetic coherence of the
Mycetozoa, we have
sequenced
the EF-1α-encoding (tef) genes from Physarum polycephalum, Dictyostelium
discoideum,
and an amoeboflagellate protostelid, Planoprotostelium aurantium. Molecular
phylogenetic
analyses of these sequences strongly support the Mycetozoa as a monophyletic
group.
Furthermore,
all methods of analysis place this group among the eukaryote “crown” taxa,
possibly
more
closely related to the animal–fungal clade than are green plants.
Figure 1.
EF-
sequence
alignment
and
intron
positions.
Table 1.
EF-1α
PCR primers
METHODS
Cell
Culture and DNA Extraction.
Planoprotostelium aurantium was grown on the pink
yeast,
Rhodotorula mucilaginosa, on agar plates and in liquid media as described (18).
DNA was
extracted
from 125-ml liquid cultures grown with gentle shaking at 25°C for 7–10 days.
Cells were
harvested by
centrifugation at 500 × g, leaving most yeast cells in suspension. The cell
pellet was lysed
in 0.1 M
EDTA/0.25% SDS containing 50 mg/ml proteinase K for 1 hr at 37°C, extracted
once each
with equal
volumes of chloroform and phenol/chloroform (1:1), and precipitated with
ethanol. After
resuspension
in 10 mM Tris·NaOH, pH 8.0/1 mM EDTA, the DNA was purified once by extraction
with glass
beads. T. Burland (
DNA, D.
Pallotta and A. Laroche (Université Laval,
polycephalum
cDNA, and C. Singleton (
DNA from D.
discoideum.
PCR
Amplification, Cloning, and DNA Sequencing. DNAs were amplified with various
combinations
of the primers described in Table 1. All amplifications used 40 cycles of 1 min
each at
95°C, 50°C,
and 72°C followed by a final, 10-min step at 72°C. Initial amplification
products were
electrophoresed
through low-melting-point agarose gels, from which individual bands were
excised and
melted at
65°C, and 1–5 μl was used in a second round of amplification
in a total volume of 100 μl.
Secondary
amplification products were also separated on low-melting-point agarose gels,
and the
appropriate
bands were excised, extracted from the gel with glass beads, ligated into a
T-tailed vector
(InVitrogen),
and used to transform the competent cells provided (INVaF′).
Positive
clones were initially identified by thermocycle screening of whole colonies
using M13 primers
(19). For
each amplification product, a minimum of five clones were further screened by
partial
sequencing
(20). Final sequencing was done on an Applied Biosystems and/or Licor automatic
sequencer.
Both DNA strands were sequenced in their entirety, and a minimum of one
complete DNA
strand was
sequenced from at least two separate clones to control for Taq DNA polymerase
errors.
An error
rate of approximately 1.2 errors per kilobase of sequence was found, and all
discrepancies
were
resolved by partial sequencing of additional clones.
Phylogenetic
Analyses. Because only four small, well defined areas of length variation are
found
in
eukaryote EF-1α (positions 1–7, 160–164, 217–228, and 450–end,
Fig. 1), sequences were
aligned by
eye. Regions of length variation were omitted from analysis, as were the amino
and carboxyl
termini
(positions 1–20 and 438–end; Fig. 1), which are missing from all PCR-generated
sequences.
The
Toxoplasma gondii tef, encoding 75% of the EF-1α protein, was compiled from the EST
(expressed
sequence tag) database, using only those regions for which at least two ESTs
were
available.
Preliminary distance trees (see below) of all available sequences were used to
trim the initially
large
plant, animal, and fungal clades, with an emphasis on minimizing terminal
branch lengths while
retaining a
broadly representative sampling of taxa. All nonconstitutively expressed,
developmental-stage-specific
EF-1αs were also omitted.
Sequences
were analyzed at the amino acid level and at the nucleotide level, using first
and second
codon
position nucleotides. Amino acid distance analyses utilized the PHYLIP 3.57c
program
PROTDIST
with its Dayhoff weighting matrix (21) and a single outgroup (Desulfurococcus)
as
recommended
(22). Trees were constructed separately by neighbor-joining and by the method
of Fitch
and
Margoliash using 100 replicates of jumbled taxon addition order and global
branch swapping (21).
Distance
analyses of nucleotides used the PHYLIP 3.57c program DNADIST with the Kimura
two-parameter
model and trees constructed by neighbor-joining (21). Distance bootstrap
analyses
consisted
of 100 replicates with trees constructed by neighbor-joining.
Parsimony
analyses of both amino acids and nucleotides utilized the program PAUP 3.1.1
(23).
Shortest
tree searches consisted of 100 rounds of random addition using TBR (tree
bisection-reconnection)
branch-swapping and the steepest descent option (23). Parsimony bootstrap
analyses
consisted of 500 replicates of simple addition holding one tree at each step.
Maximum
likelihood analyses of amino acids utilized the program fastPROTML (24).
Bootstrap values
were
calculated by the RELL method on the 1,000 best trees, using the weighting
matrix of Jones et
al. (25)
normalized to the amino acid composition of the data set (-jf option) on a
semiconstrained
starting
tree. To avoid prohibitively long PROTML analysis times, deeply branching,
single-representative-clade
taxa not directly related to the Mycetozoa, based on the results of
parsimony
and distance analyses (see Results), were omitted. These include Trichomonas,
Entamoeba,
and Glugea for all analyses and also Blastocystis and Stylonychia for analyses
testing
the
monophyly of the Mycetozoa. On the basis of the strong results of the latter
analyses, the
Mycetozoa
were constrained as monophyletic for PROTML analyses testing the phylogenetic
position
of the
group as a whole. Maximum likelihood analysis of nucleotides utilized the
PHYLIP 3.57c
program
DNAML (21) with empirical base frequencies, a transition-to-transversion ratio
of 1.0, and
100
bootstrap replicates.
Figure 2.
Phylogenetic
analyses of
EF-1α amino
acid
sequences
show a
monophyletic,
late-branching
Mycetozoa.
RESULTS
Mycetozoan
tef Gene Sequences and Intron Positions. The 5′ two-thirds of the
Physarum
tef gene was amplified from genomic DNA, while the 3′ half of the gene
was amplified from
cDNA. The
latter was necessary because all primer combinations for the 3′ half of
the gene
preferentially
amplified the retrotransposon Tp1, which constitutes 10–20% of the Physarum
genome
(26). All
12 1F-7R clones screened were identical to each other, as were the 4 2F-10R
clones
screened.
The 3′ and 5′ clones were also identical to each other in their 260
nucleotides of overlap,
suggesting
the presence of a single, active tef locus in this genome. The Physarum tef
gene contains a
single
142-nucleotide intron, which lies at a position identical to that of an intron
found in both
vertebrates
and invertebrates (Fig. 1).
Both
Dictyostelium tef genes, for which cDNA sequences were previously determined
(27), were
amplified
and sequenced in the region covered by primers 1F and 10R (Table 1). A single
147-nucleotide
intron was found in the tef2 gene at amino acid position 53. This intron position
is
clearly
unrelated to that of the Physarum intron, although it is close to another
intron position shared
by
vertebrates and invertebrates (Fig. 1). Otherwise, both Dictyostelium tef
genomic sequences were
identical
to their cDNA sequences, which are also identical to each other at the amino
acid level (27).
Initial
amplification of the protostelid DNA revealed the presence of three tef
sequences (Fig. 1). Two
of these, designated
tef1 and tef2, are very similar to each other and were presumed to be from the
protostelid.
The third sequence appears to be a fungal tef, presumably from the protostelid
food
source (see
below). The two presumed protostelid sequences are intronless and differ at 32
nucleotide
positions,
all of which are silent except for position 377, which gives a glutamate in
tef1 and a glycine in
tef2 (Fig.
1). The five Mycetozoan tef genes show strong codon bias: both the protostelid
and
Physarum
sequences are 74–75% G+C at silent codon positions, versus 32–34% G+C at silent
positions
for the Dictyostelium genes (27).
The third
tef sequence amplified from the protostelid DNA preparation appears to belong
to the
protostelid
food source, Rhodotorula mucilaginosa, as it encodes an insertion found
exclusively in all
animals and
fungi (positions 217–228, Fig. 1; ref. 11) and a two amino acid gap found in
all fungi
(positions
162–163, Fig. 1). The sequence contains four introns ranging in size from 67 to
156
nucleotides.
None of these introns occurs at a previously described intron position (Fig.
1). A fungal
origin of
this sequence was also confirmed by phylogenetic analysis (see below). Some
contamination
of the
protostelid DNA preparation with yeast DNA is expected due to incomplete
separation of cells
prior to
extraction (see Methods) and the presence of undigested yeast cells in
protostelid food
vacuoles.
Molecular
Phylogeny of EF-1α Sequences Strongly Supports a Monophyletic
Mycetozoa.
A data set consisting of all known, constitutively expressed, protistan EF-1α
sequences
and a
representative set of animal, fungal, and angiosperm sequences was analyzed at
the amino acid
and
nucleotide levels, using maximum parsimony and two distance-matrix methods,
with more limited
questions
tested by maximum likelihood analysis (Fig. 2). All phylogenetic methods used
place the
Mycetozoa
together as a monophyletic group, with the myxogastrid and dictyostelid
sequences more
closely
related to each other than either is to the protostelid sequence (Fig. 2). Both
the monophyly of
the
Mycetozoa and of the myxogastrid–dictyostelid clade are strongly supported by
bootstrap analysis
of amino
acid sequences using all methods (81–100% and 89–97%, respectively, Fig. 2).
Nucleotide
analyses
also support both a monophyletic Mycetozoa and a myxogastrid–dictyostelid
clade, although
with
consistently lower bootstrap values (56–84% and 77–92%, respectively).
Parsimony and
distance analyses also place the putative Rhodotorula mucilaginosa tef gene
together
with that
of Puccinia graminis (88–91% bootstrap, Fig. 2). This is consistent with the
current
classification
of Rhodotorula mucilaginosa as a basidiomycete fungus (28) and confirms its
identity.
Our
analyses also show a moderately strong affinity between the basidiomycete and
zygomycete fungi
(76–77%
bootstrap, Fig. 2). This contradicts both SSU rRNA trees and traditional
taxonomy (29) and
warrants
further study.
EF-1α
Phylogeny Tentatively Supports the Mycetozoa as a Sister Group to
Animals +
Fungi. All methods of analysis also place the Mycetozoa within the crown of the
eukaryote
tree, closer to the animal–fungal clade than are the green plants (Fig. 2).
This topology is
reconstructed
in the optimal trees by all methods of analysis used with both amino acids and
nucleotides
(Fig. 2). However, no significant bootstrap support is found for this specific
placement of
the
Mycetozoa or for any other higher order placement of these taxa by any of these
methods.
Inspection
of individual distance bootstrap trees shows that 71% of these actually
reproduce an
animal–fungal–Mycetozoan
clade (Fig. 2), but in 36% of the trees this clade also includes, alone or in
combination,
Porphyra, Stylonychia, Euglenozoa, or Blastocystis (33%, 7%, 6%, and 3% of
total,
respectively).
Only 14% of replicates place the Mycetozoa together with green plants, with or
without
other taxa,
and only 10% place the Mycetozoa deep to animals + fungi + green plants.
Otherwise, the
Porphyra
sequence is also found as the outgroup to an animal–fungal–mycetozoan clade
with green
plants, or
with the Mycetozoa (19%, 27%, and 13% of trees, respectively), whereas the
Stylonychia
sequence is
found most frequently with the Euglenozoa or near the other ciliates (45% and
34% of
trees,
respectively). Because such poorly resolved, unstable branches can obscure
otherwise stable
relationships
among their neighboring branches within a tree (30), analyses were repeated
with the
Stylonychia
and Porphyra sequences deleted.
Distance
analyses with the Porphyra and Stylonychia sequences deleted show 85% bootstrap
support for
the Mycetozoa as members of a larger crown group including green plants,
animals, and
fungi and 70%
bootstrap support for the Mycetozoa as closer to the animal–fungal clade than
are the
green
plants (Fig. 2). Likewise, protein maximum likelihood analysis without these
sequences shows
82%
bootstrap support for a crown placement of the Mycetozoa and 75% support for
their sisterhood
with
animals + fungi (Fig. 2). However, parsimony analysis, albeit the most
refractory to the correction
of
long-branch effects (31), still finds less than 50% bootstrap support for
either relationship.
DISCUSSION
EF-1α
Phylogeny Strongly Supports a Monophyletic Mycetozoa. We have
enzymatically
amplified and sequenced the EF-1α-encoding genes from representatives
of each of the
three
recognized subclasses of Mycetozoa, the cellular (dictyostelid), acellular
(myxogastrid), and
protostelid
slime molds (Fig. 1). Phylogenetic analyses of a broadly representative EF-1α data set
show strong
support for the monophyly of the group by all methods of analysis used (86–100%
bootstrap,
Fig. 2). Strong support for a monophyletic Mycetozoa, represented by
Dictyostelium and
Physarum,
is also found by analyses of actin (72–95% bootstrap, refs. 12 and 13) and of β-tubulin
(74–91%
bootstrap, ref. 14).
The EF-1α data
further subdivide the Mycetozoa into a myxogastrid–dictyostelid clade strongly
excluding
the amoeboflagellate protostelid, Planoprotostelium aurantium (89–97%
bootstrap, Fig.
2). Thus,
the myxogastrid–dictyostelid divergence does not appear to represent the
deepest division
within the
Mycetozoa. This suggests that the differences between these taxa, such as an
amoeboflagellate
versus strictly amoeboid condition and plasmodial versus aggregative
development,
may not be
as profound as many have considered them to be. Both Olive (1) and Spiegel (18)
have
argued that
a strictly amoeboid trophic stage, at least, has probably evolved multiple
times among the
protostelids.
The
Mycetozoa as Members of a Eukaryote Crown Group. Phylogenetic analyses of
EF-1α
sequences also place the Mycetozoa among the multicellular eukaryotes as the immediate
outgroup to
the animal–fungal clade (Fig. 2). This topology is favored by all analytical
methods used
(Fig. 2),
although there is no immediate bootstrap support for this specific topology by
any method.
However,
distance and maximum likelihood analyses of the EF-1α data with the Porphyra and
Stylonychia
sequences deleted show greatly increased bootstrap support for both the
placement of the
Mycetozoa
within the eukaryote crown and for these taxa as more closely related to the
animal–fungal
clade than
are green plants (82–85% and 70–75%, respectively). Since bootstrap values
greater than
70% have
been shown likely to correspond to confidence levels of 95%, except under
conditions of
extreme
substitutional saturation or highly unequal rates (34), both methods seem to
strongly suggest
that the
Mycetozoa are crown eukaryotes, probably more closely related to the
animal–fungal clade
than are
green plants. Nonetheless, unweighted parsimony analysis, a method highly
sensitive to
long-branch
effects (31), still shows no significant support for these or most other major
clades in the
EF-1α tree
(Fig. 2).
Thus the
apparent lack of support for the higher order placement of the Mycetozoa with
the full EF-1α
data set
appears to be due, at least in part, to poor resolution of the branching
positions of several
taxa, most
notably Porphyra and Stylonychia (Fig. 2). Inspection of individual bootstrap
trees shows
that these
sequences are weakly supported at various positions in the tree, Porphyra being
found
mostly among
the “crown” taxa, whereas Stylonychia ranges from among the relatively deeply
branching
ciliates to within fungi. Such unstable branches can decrease bootstrap values,
apparently
even for
relatively distantly related nodes (11, 32). This appears to be due, at least
in part, to a
combination
of the tendency of poorly resolved taxa to obscure underlying tree structure
(30) with the
requirement
of bootstrap analysis, as currently implemented, for strictly monophyletic
groups (33).
Although
increased taxon sampling to break up long branches should help alleviate this
problem with
bootstrap
analysis (22, 35)—with all methods except perhaps parsimony (31)—the gathering
of
protein
sequence data to evaluate ancient divergences is still a relatively slow
process. Nonetheless, it
may still
be possible, with caution, to answer more limited but still highly relevant
questions (11, 33). In
this case,
we are asking only whether the Mycetozoa are early- or late-emerging
eukaryotes, possibly
more closely
related to the animal–fungal clade than are green plants. It is important to
note that we are
in no way
precluding the possibility that other taxa, most notably the red alga Porphyra,
may be more
closely
related to the animal–fungal clade than are the Mycetozoa.
An origin
of the Mycetozoa from within a eukaryote “crown” group—i.e., among animals,
fungi, and
green
plants to the exclusion of most or all protistan lineages represented, is also
supported by
individual
analyses of actin (12–13, 36), RNA polymerase largest subunit (15),
glyceraldehyde-3-phosphate
dehydrogenase (3), and most analyses of α- and β-tubulin (refs. 11 and
14, but see
ref. 36) as well as a combined analysis of all relevant, currently available,
protein data (16).
Furthermore,
both the actin and the combined protein analyses specifically support the
Mycetozoa as
more
closely related to the animal–fungal clade than are green plants (56–60% and
83–86%
bootstrap,
respectively). This relationship is also suggested by analyses of both α- and β-tubulin
(67%
and 73%
bootstrap, respectively, ref. 11), although the rooting of these trees is
clearly problematic
(14, 36).
Although
nucleotide-level analyses of actin place the Mycetozoa closer to animals than
fungi (64%
bootstrap,
ref. 12), this is not supported by amino acid-level analyses of the same data
(11, 13, 36).
Loomis and
Smith (37) also noted strong similarity between animals and Dictyostelium based
on six
small
protein data sets. However, because none of these data sets included an outgroup,
these results
cannot be
meaningfully interpreted. A specific relationship between animals and
Dictyostelium to the
exclusion
of fungi is further ruled out by its lack of an 11- to 13-amino acid insertion
found exclusively
in all
animal and fungal EF-1αs (Fig. 1, ref. 11).
Ribosomal
RNA Phylogeny of Mycetozoa. Although three protein sequence data sets, actin,
α-tubulin, and EF-1α, strongly support a monophyletic
Mycetozoa (refs. 12–14; Fig. 2) and at least
four place
these taxa in the eukaryote crown (refs. 3, 11–15, and 36; Fig. 2), rRNA trees
consistently
show the
Mycetozoa to be polyphyletic as well as early branching (7–10). Physarum,
especially,
appears as
one of the earliest branches of mitochondrial eukaryotes in nearly all rRNA trees
(7–10).
Although
Cavalier-Smith finds very weak evidence for a monophyletic Mycetozoa with SSU
rRNA
(40), this
clade still arises very deeply in the tree. Because a growing body of protein
sequence data
contradicts
these results (3, 11–16, 36), including the data presented here (Fig. 2), it is
necessary to
consider
the possibility that current rRNA trees may be misleading with respect to these
questions.
If the
current rRNA phylogenies are indeed incorrect with regard to the phylogeny of
the Mycetozoa, it
should be
considered that increased taxon sampling has been shown to potentially overcome
many
sources of
both random and systematic error in phylogenetic analyses (22, 35). Thus,
inclusion of
additional
rRNA sequences from all three classes of Mycetozoa, especially protostelids,
might help to
resolve
some of these questions. This is suggested by the results of Spiegel et al.
(41), who analyzed
the first
protostelid molecular sequence, a 310-nucleotide portion of the SSU rRNA gene
of
Protostelium
mycophaga. Analyses of this sequence with a limited set of taxa showed strong
support
for a
monophyletic Mycetozoa, although the method of sequence alignment may have
biased the results
in this
direction (41).
Accuracy of
the Current EF-1α Data Set for Deep-Level Phylogeny. Besides a
relatively
broad representation of the animals, fungi, Mycetozoa, and Apicomplexa (Fig.
2), most of the
EF-1α tree is
still sparsely sampled. Thus most of the deeper branches are only tentatively
resolved,
and the
placement of these taxa in the tree should be considered a general indication
of their true
phylogenetic
position, at best. Perhaps most problematic is the fact that the ciliates do
not form a clade
in the EF-1α tree,
contradicting considerable morphological and molecular data (3, 8–14, 36, 42).
The
instability
of the Stylonychia EF-1α branch was noted above, and the grouping of
Entamoeba with
the ciliate
Euplotes is almost certainly a spurious long-branch attraction as well (Fig. 2,
ref. 31). Better
resolution
of the relationships among the various protistan taxa in the EF-1α tree
will almost certainly
require
both more thorough sampling of taxa and careful analysis of specific questions.
The Glugea
EF-1α is especially noteworthy in that it gives an extremely long branch and
is more
distant
from the rest of the eukaryotes than even the archaebacterium Desulfurococcus
(Fig. 2), more
than twice
as far in distance analyses (21)! The Glugea EF-1α also contains many nonconservative
amino acid substitutions
at otherwise universally conserved positions, including active site residue
changes
unlikely to be compatible with enzymatic function (A. Roger and S.L.B.,
unpublished data).
This
sequence also appears to encode an insertion otherwise found only in animals
and fungi (11). The
latter is
consistent with the placement of the Microsporidia with fungi in α- and β-tubulin
phylogeny
(14, 39).
Thus the Glugea EF-1α may be artefactually drawn toward the base of
the EF-1α tree due
to an
accelerated rate of evolution, as previously observed with the Xenopus EF-1α-derived
protein,
thesaurin
(43).
Implications
of a Mycetozoan Sister Clade to Animals + Fungi. Placement of the
Mycetozoa
among the “crown” eukaryotes is consistent with a large body of data on their
physiology,
biochemistry,
molecular biology, behavior, and development (1, 44–46). Perhaps most notable
among
these is
the Mycetozoan fruiting body, which shows characteristics of true
multicellularity by including
functionally
specialized, nonreproductive cells (1). This is especially striking in the
dictyostelids, where
the
developmental fates of fruiting body cells are predetermined in the slug (1,
44, 45).
Thus, a
growing body of protein sequence data supports a monophyletic Mycetozoa (Fig. 2;
refs. 12–
14), and
all currently available, broadly representative protein data sets support these
taxa as
late-emerging
eukaryotes (Fig. 2; refs. 3, 11–16, and 36). This suggests that the rRNA data
should be
more
closely examined with regard to these questions. In addition, the possibility
that the Mycetozoa
may be more
closely related to the animal–fungal clade than are green plants clearly
warrants further
study. The
results of our work and others (Fig. 2; refs. 3, 11–16, and 36) indicate that
animals, fungi,
and slime
molds may still represent only a small corner of eukaryote diversity, and it
should not be
assumed
that traits shared by these taxa are ancient or universal among eukaryotes. On
the other hand,
these
results support the continued use of mycetozoan taxa as model systems for
studying the origin,
evolution,
and function of a wide range of characteristics of “higher” eukaryotes (44–47).
Acknowledgements
We thank T.
Burland, A. Laroche, D. Pallotta, and C. Singleton for the generous gifts of
DNAs; F.
Spiegel for
Planoprotostelium aurantium cultures and advice on growing them; A. Cohen for
invaluable
help with DNA extractions and screening clones; and D. Edgell, J. Logsdon, J.
Palmer, and
A. Roger
for helpful discussion and for critical reading of the manuscript. This work
was supported by
Medical
Research Council Grant MT4967 to W.F.D.
ABBREVIATIONS
EF-1α, protein
synthesis elongation factor-1α; SSU rRNA, small subunit ribosomal
RNA.
Footnotes
Data
deposition: The sequences reported in this paper have been deposited in the
GenBank database
(accession
nos. AF016239–43).
A
commentary on this article begins on page 11767.
This latter
region is also variable in some protists and archaebacteria. Therefore, it is
only the
combination
of the insertion together with the deletion that defines this as a fungal EF-1α (S.L.B.,
unpublished
data).
Dugesia
japonica (38) EF-1α has a 4-amino acid insertion within this
larger insertion. Although
Microsporidia
may encode a form of the 11- to 13-amino acid insertion, this is consistent
with other
data
suggesting that they may be fungi (refs. 14 and 39; see below).
References
1.Olive, L.
S. & Stoianovitch, C. (1975) in The Mycetozoans (Academic,
2.Olive, L.
& Stoianovitch, C. (1960). Bull. Torrey Bot. Club 87, 120.
3.Roger, A.
J., Smith, M. W., Doolittle, R. F., & Doolittle, W. F. (1996). J.
Eukaryotic
Microbiol.
43, 475485. [PubMed]
4.Whittaker,
R. H. (1969). Science 163, 150160. [PubMed]
5.Whittaker,
R. H. (1969). Science 164, 857. [PubMed]
6.Olive, L. (1969). Science 164, 857. [PubMed]
7.Hendriks, L., De Baere, R., Van de Peer, Y., Neefs, J., Goris, A., & De Wachter, R. (1991).
J. Mol. Evol. 32, 167177. [PubMed]
8.Hinkle, G. & Sogin, M. L. (1993). J. Eukaryotic Microbiol. 40, 599603. [PubMed]
9.De Rijk, P., Van de Peer, Y., Van den
Broeck, I., & De Wachter, R. (1995). J. Mol. Evol.
41, 366375.
[PubMed]
10.Krishnan,
S., Barnabas, S., & Barnabas, J. (1990). BioSystems 24, 135144. [PubMed]
11.Baldauf, S. L. & Palmer, J. D. (1993). Proc. Natl. Acad. Sci. USA 90, 1155811562.
[PubMed]
12.Bhattacharya, D. & Ehlting, J. (1995). Arch. Protistenkde. 145, 155164.
13.Drouin, G., Moniz de Sá , M., & Zuker,
M. (1995). J. Mol. Evol.
41, 841849. [PubMed]
14.Keeling,
P. J. & Doolittle, W. F. (1996). Mol. Biol. Evol. 13, 12971305. [PubMed]
15.Klenk, H.-P., Zillig, W., Lanzendörfer, M., Grampp, B., & Palm, P. (1995). Arch.
Protistenkde. 145, 221230.
16.Kuma, K.-i., Nikoh, N., Iwabe, N., &
Miyata, T. (1995). J. Mol. Evol. 41, 238246.
[PubMed]
17.Baldauf, S. L., Palmer, J. D., & Doolittle, W. F. (1996). Proc. Natl. Acad. Sci. USA 93, 7749
7754.
[PubMed]
18.Spiegel,
F. W. (1982). Protoplasma 113, 165177.
19.Sandhu,
G. S., Precup, J. W., & Kline, B. C. (1989). BioTechniques 7, 689690.
[PubMed]
20.Goode,
B. L. & Feinstein, S. C. (1992). BioTechniques 12, 374375. [PubMed]
21.Felsenstein,
J. (1993) in PHYLIP (Phylogeny Inference Package) (Department of Genetics,
22.Swofford,
D. L., Olsen, G. J., Waddell, P. J., & Hillis, D. M. Hillis, D. M., Moritz,
C., &
Mable, B.
K., eds. (1996) in Molecular Systematics 2nd Ed. (Sinauer,
23.Swofford,
D. L. (1993) in PAUP, Phylogenetic Analysis Using Parsimony (
History
Survey,
24.Adachi,
J. & Hasegawa, M. (1992) in MOLPHY, Programs for Molecular Phylogenetics
I—PROTML,
Maximum Likelihood Inference of Protein Phylogeny, Computer Science
Monographs
no. 27 (Institute of Statistical Mathematics,
25.Jones,
D. T., Taylor, W. R., & Thornton, J. M. (1992). Comput. Appl. Biosci. 8,
275282.
[PubMed]
26.Rothnie,
H. M., McCurrach, K. J., Glover, L. A., & Hardman, N. (1991). Nucleic Acids
Res.
19, 279286.
[PubMed]
27.Yang,
F., Demma, M.,
(
28.Fell, J.
W., Boekhout, T., & Freshwater, D. W. (1995). Stud. Mycol. 38, 129146.
29.Bruns,
T. D., Vilgalys, R., Barns, S. M., Gonzalez, D., Hibbett, D. S., Lane, D. J.,
Simon, L.,
Stickel,
S., Szaro, T. M., Weisburg, W. G., & Sogin, M. L. (1992). Mol. Phylogenet.
Evol. 1,
231241.
[PubMed]
30.Adams,
E. N. (1972). Syst. Zool. 21, 390397.
31.Felsenstein, J. (1978). Syst. Zool. 27, 401410.
32.Lecointre, G., Philippe, H., Van Le, H. L.,
& Le Guyader, H. (1993). Mol. Phylogenet. Evol.
2, 205224.
[PubMed]
33.Sanderson,
M. J. (1989). Cladistics 5, 113129.
34.Hillis,
D. M. & Bull, J. J. (1993). Syst. Biol. 42, 182192.
35.Hillis,
D. (1996). Nature (
36.Philippe,
H. & Adoutte, A. in Evolutionary Relationships Among Protozoa, eds. Coombs,
G. H.,
Vickerman, K., Sleigh, M. A. & Warren, A. (Systematics Assoc.,
37.Loomis,
W. F. & Smith, D. W. (1995). Experientia 51, 11101116. [PubMed]
38.Kobayashi,
M., Wada, H., & Satoh, N. (1996). Molec. Phylogenet. Evol. 5, 414422.
[PubMed]
39.Edlind,
T. D., Li, J., Visvesvara, G. S., Vodkin, M. H., McLaughlin, G. L., &
Katiyar, S. K.
(1996).
Molec. Phylogenet. Evol. 5, 359367. [PubMed]
40.Cavalier-Smith,
T. (1993). Microbiol. Rev. 57, 953994. [PubMed]
41.Spiegel,
F. W., Lee, S. B. & Rusk, S. A. (1995) Can. J. Bot. 73, Suppl. 1,
S738–S746.
42.Lynn, D.
H. & Small, E. B. Margulis, L., Corliss, J. O., Melkonian, M., &
Chapman, D. J.,
eds. (1990)
in Handbook of Protoctista (Jones & Bartlett,
43.Viel,
A., le Maire, M., Philippe, H., Morales, J., Mazabraud, A., & Denis, H.
(1991). J. Biol.
Chem. 266,
1039210399. [PubMed]
44.Gross,
J. D. (1994). Microbiol. Rev. 58, 330351. [PubMed]
45.Kay, R.
R. (1994). Curr. Opin. Genet. Dev. 4, 637641. [PubMed]
46.Burland,
T. G., Solnica-Krezel, L., Bailey, J., Cunningham, D. B., & Dove, W. F.
(1993). Adv.
Microb.
Physiol. 35, 169. [PubMed]
47.Kuspa,
A., Maghakian, D., Bergesch, P., & Loomis, W. F. (1992). Genomics 13, 4961.
[PubMed]