Download Evolution of Plant-Like Crystalline Storage Polysaccharide in the

J Mol Evol (2005) 60:257–267
DOI: 10.1007/s00239-004-0185-6
Evolution of Plant-Like Crystalline Storage Polysaccharide in the Protozoan
Parasite Toxoplasma gondii Argues for a Red Alga Ancestry
Alexandra Coppin,1 Jean-Stéphane Varré,2 Luc Lienard,1 David Dauvillée,1 Yann Guérardel,1
Marie-Odile Soyer-Gobillard,3 Alain Buléon,4 Steven Ball,1 Stanislas Tomavo1
1
Laboratoire de Chimie Biologique, CNRS UMR 8576, Université des Sciences et Technologies de Lille, 59655 Villeneuve dÕAscq cedex,
France
2
Laboratoire dÕInformatique Fondamentale de Lille, CNRS UMR 8022, Université des Sciences et Technologies de Lille,
59655 Villeneuve dÕAscq cedex, France
3
Laboratoire Arago, Observatoire océanologique, CNRS UMR 7628, Université Paris VI, 66651 Banyuls-sur-mer cedex, France
4
Institut National de la Recherche Agronomique, Rue de la Géraudiére, 44316, Nantes cedex 03, France
Received: 17 June 2004 / Accepted: 9 September 2004 [Reviewing Editor: Dr. Patrick Keeling]
Abstract. Single-celled apicomplexan parasites are
known to cause major diseases in humans and animals including malaria, toxoplasmosis, and coccidiosis. The presence of apicoplasts with the remnant of
a plastid-like DNA argues that these parasites
evolved from photosynthetic ancestors possibly related to the dinoflagellates. Toxoplasma gondii displays amylopectin-like polymers within the cytoplasm
of the dormant brain cysts. Here we report a detailed
structural and comparative analysis of the Toxoplasma gondii, green alga Chlamydomonas reinhardtii,
and dinoflagellate Crypthecodinium cohnii storage
polysaccharides. We show Toxoplasma gondii amylopectin to be similar to the semicrystalline floridean
starch accumulated by red algae. Unlike green plants
or algae, the nuclear DNA sequences as well as biochemical and phylogenetic analysis argue that the
Toxoplasma gondii amylopectin pathway has evolved
from a totally different UDP-glucose-based metabolism similar to that of the floridean starch accumulating red alga Cyanidioschyzon merolae and, to a
lesser extent, to those of glycogen storing animals or
fungi. In both red algae and apicomplexan parasites,
isoamylase and glucan–water dikinase sequences are
proposed to explain the appearance of semicrystalline
Correspondence to: Stanislas Tomavo; email: Stan.Tomavo@
univlille1.fr
starch-like polymers. Our results have built a case for
the separate evolution of semicrystalline storage
polysaccharides upon acquisition of photosynthesis
in eukaryotes.
Key words: T. gondii — Plant-like metabolism —
Amylopectin — Floridean starch — Evolutionary
origin — Glucan water dikinase — Isoamylase —
Rhodophyte
Introduction
The vast majority of living cells store glucose in the
form of glycogen. This hydrosoluble polysaccharide
consists of a-1, 4-linked glucans that are branched in
the a-1, 6 position. Bacteria are known to use ADPglucose as a substrate for glycogen synthesis, while
fungi and animal cells use UDP-glucose. The ADPglucose and UDP-glucose pathways differ not only
by the nature of the precursor substrate but also by
other characteristics. These include distinct regulatory mechanisms and different sets of particular
enzymes (Ballicora et al. 2003; Ball and Morell
2003).
Green algae, and land plants accumulate starch as
an insoluble structurally distinct form of a-1, 4-
258
linked, a-1, 6-branched glucose polymer that is
localized within the plastids. It is generally agreed
that starch synthesis has evolved from the pathway
that existed in the cyanobacterial ancestor of plastids. Unlike glycogen, starch is semicrystalline and
organized in a complex granule as a mix of moderately branched amylopectin and mostly linear amylose molecules (Ball and Morell 2003). A wealth of
molecular and genetic data conclusively show that
starch in land plants and green algae is synthesized
through the prokaryote ADP-glucose-based pathway
(Recondo and Leloir 1961; Greenberg and Preiss
1964). Green algae and plants both contain numerous and redundant isoenzymes corresponding to
each step of the pathway (Ball and Morell 2003; Ral
et al. 2005). The significance of this high number of
isoenzymes still escapes us and could be related to
the formation of crystalline polysaccharides. Other
photosynthetic eukaryotes such as the red algae
(Viola et al. 2001), the cyanelle-containing glaucophytes, and organisms which are thought to be derived by secondary endosymbiosis of a red algal
ancestor (dinoflagellates and cryptophytes) synthesize crystalline polysaccharides outside their plastids.
We refer to these as ‘‘floridean starch’’ a name that
was coined to describe the form of storage material
accumulated in the cytoplasm of red algae. The
nature of the pathway used by red algae for storage
polysaccharide synthesis is still under debate. Interestingly, dinoflagellates and cryptophytes both harbor plastids which are thought to be derived from
the rhodoplasts of red algae (Zhang et al. 1999;
Douglas and Penny 1999).
Recent molecular evidence has demonstrated that
the single-cell apicomplexan parasites harbor a plastid-like nonphotosynthetic apicoplast containing
DNA that is somewhat related to the dinoflagellate
rhodoplast DNA (McFadden et al. 1996; Wilson et
al. 1996; Cavalier-Smith 1999; Marechal and Cesbron-Delauw 2001; Fast et al. 2001). However, it is
still controversial whether the apicoplast originated
from a green or a red alga (McFadden et al. 1996;
Wilson et al. 1996; Cavalier-Smith 1999; Marechal
and Cesbron-Delauw 2001; Fast et al. 2001; Köhler
et al. 1997; Funes et al. 2002; Waller et al. 2003; Cai
et al. 2003). Apicomplexan parasites define important
pathogens responsible for widespread diseases such
as toxoplasmosis and malaria in humans or coccidiosis in animal species. Some apicomplexan parasites
such as Eimeria, the agent of coccidiosis in animals
store glucose in the form of amylopectin granules
related to floridean starch (Ryley et al. 1969). Others
such as Plasmodium, the agent of malaria, lack storage polysaccharides. In humans, Toxoplasma gondii
can exist in two distinct forms. At the first stages of
infection, the intracellular parasite divides and invades new cells actively. Upon reaction of the host
and under pressure of the immune system, this rapidly replicating tachyzoite stage switches into a slowly
dividing bradyzoite form. Bradyzoites can be found
within dormant cysts in the brain and muscle tissues.
These encysted bradyzoites can be reactivated and
then differentiated into tachyzoites if the host is immunodepressed. This will lead to fatal toxoplasmic
encephalitis, a common death cause in patients with
AIDS (Luft and Remington 1988; Luft and Remington 1992). Here we present evidence that the
storage polysaccharide accumulated in the cytoplasm
by Toxoplasma gondii and that of the dinoflagellate
Crypthecodinium cohnii can be defined as a genuine
starch. We show that it is synthesized through a
UDP-glucose-based pathway much simpler than that
described for plants. This pathway is very similar to
that we have found to be encoded in the recently
sequenced genome of red alga Cyanidioschyzon merolae (Matsuzaki et al. 2004). We propose that two
particular functions found in all starch synthesizing
organisms (red algae, green algae, plants, apicomplexans) are responsible for the appearance of a
semicrystalline insoluble starch granule instead of
glycogen. Our data shed light not only on the evolution of apicomplexan parasites but also on those of
the green and red lineages of photosynthetic eukaryotes and that of the starch metabolism pathway.
Materials and Methods
Strains and Growth Conditions. The RH strain of Toxoplasma gondii was grown in HepG2 cells and purified as described
(Dzierszinski et al. 1999). However, the culture medium (DulbeccoÕs minimal essential medium containing 10% fetal calf serum)
was left until it became acidic (pH 6.2–6.5), conditions known to
induce bradyzoite and cyst formation. Media and culture conditions for growing the wild type of Chlamydomonas reinhardtii
(137C strain) and the dinoflagellate Crypthecodinium cohnii (ATCC
50050 strain) were previously described (Libessart et al. 1995;
Perret et al. 1993).
Starch Purification. T. gondii, C. reinhardtii, and C. cohnii
cultures were centrifuged and resuspended in 15 ml of phosphatebuffered saline (PBS). Cell suspensions were disrupted three or four
times by a French press (10,000 to 15,000 p.s.i.). Crude starch
pellets were obtained by spinning down the lysate at 10,000 g for
20 min. The pellets were washed twice with water and passed
through a self-formed 90% Percoll gradient. The purified starch
pellets were rinsed in distilled water, centrifuged twice at 10,000 g,
and kept dry at 4C. Starch amounts were measured by the amyloglucosidase assay (Libessart et al. 1995).
Separation of Starch Polysaccharide by Gel Permeation Chromatograghy. Starches (1.5 mg) dissolved in
500 ll of 10 mM NaOH were applied to a Sepharose CL-2B column (0.5 cm · 65 cm) equilibrated in 10 mM NaOH. Fractions of
300 ll were collected at a rate of one fraction each second min.
Glucans in each fraction were detected through the iodine–polysaccharide interaction.
259
Chain-Length Distribution Analysis. Purified starches
(500 lg) were suspended in 55 mM sodium acetate buffer, pH 3.5,
and debranched by 2 units of Pseudomonas amylodermosa isoamylase (Hayashibara Biochemical Laboratory, Japan) at 45C
overnight. The samples were further analyzed by high-performance
anion-exchange chromatography–pulse amperometric detection
(HPAEC-PAD) and capillary electrophoresis.
X-ray Diffraction and Nuclear Magnetic Resonance
(NMR) Analysis. Powder X-ray diffractograms were collected from purified starches as described previously (Buléon et al.
1997). 1H-NMR spectroscopy of purified starches was performed
on a Bruker ASK 400 WB spectrometer. Chemical shifts (expressed
as ppm) were measured by reference to internal dimethylsulfoxided6 (d = 2.52 ppm at 70C). The two-dimensional homonuclear
correlation spectroscopy (COSY) experiments were performed
using the standard Bruker pulse library.
Transmission Electron Microscopy. In vitro-induced
T. gondii bradyzoites were collected by centrifugation, fixed in 2.5%
glutaraldehyde prepared in 0.1 M cacodylate buffer, and post fixed
in 1% OsO4 in the same buffer. After ethanol dehydration, the
pellet was embedded in Epon. Serial thin sections were cut and
collected on longitudinal barred grids. After staining with 2%
uranyl acetate prepared in 50% ethanol and incubation with a lead
citrate solution, sections were observed on a Hitachi H-600 electron
microscope. Transmission electron microscopy (TEM) of Chlamydomonas reinhardtii (137C strain) and Crypthecodinium cohnii
were previously described (Libessart et al. 1995; Perret et al. 1993).
Gene Identification and Reverse Transcriptase–
Polymerase Chain Reaction. The Toxoplasma gondii genes
encoding enzymes involved in amylopectin metabolism were identified at the genome Website, http://www.toxodb.org, using key
word searches or protein ortholog sequences. All genes were
reconstituted in silico using Genewise (Wise 2), Genscan, Genemark. Both exons and introns were determined. The predicted
proteins were aligned using ClustalW and the KozakÕs rule was
used to determine the translational initiation sites according to the
method previously described by Seeber (1997). The following genes
were identified: a-amylase (TGG_994355; gene can be found at
nucleotide positions 55000 to 60000); isoamylase (TGG_994676;
gene at positions 240000 to 255000), D-enzyme (TGG_994353;
gene at positions 170000 to 185000), R1 protein (TGG_994467;
gene at positions 150000 to 185000), a-glucan phosphorylase
(TGG_994281; gene at positions 740000 to 750000), glycogenin
(TGG_994281; gene at positions 540000 to 560000), a-glucosidase
(TGG_994637; gene at positions,15000–40000), indirect debranching enzyme (TGG_994267; gene at positions 150000 to 175000),
UDP-glucose pyrophosphorylase (TGG_994289; gene at positions
535000 to 545000), branching enzyme 1 (TGG_994676; gene at
positions 240000 to 250000), branching enzyme 2 (TGG_994574:
gene at positions 378000 to 389000), and starch (glycogen) synthase
(TGG_994276; gene at positions 60000 to 80000). The transcript
level or expression of these genes was demonstrated by RT-PCR.
To do this, total RNA from T. gondii tachyzoites and bradyzoites
obtained from murine brain cysts was extracted and used for RTPCR as previously described (Dzierszinski et al. 2001). The cDNAs
amplified were cloned and sequenced using an Alfexpress DNA
automatic sequencer (Amersham Pharmacia Biotech).
Enzyme Assays. In vitro-induced bradyzoites of T. gondii,
C. reinhardtii, and C. cohnii extracts were prepared using French
press disruption of cells suspended in 10 ml of cold 50 mM Hepes
buffer, pH 7.0, containing 1 mM MgCl2, 2 mM EDTA, 2 mM bmercaptoethanol, 1 mM PMSF, and a cocktail of protease inhib-
itors. After centrifugation at 2000g for 10 min, the supernatants
were ultracentrifuged at 100,000 g at 4C for 2 h. The supernatants
were recovered and protamine sulfate (20%, w/w) was added. After
10 min on ice, the solutions were centrifuged at 27,000 g at 4C for
30 min. The supernatants were dialyzed against 50 mM Hepes
buffer containing 2 mM EDTA at 4C for 24 h and concentrated
using Centricon-30 (Amicon). Starch or glycogen synthase activity
was determined according to the method described by Karkhanis et
al. (1993) using [14C]ADP-glucose (303 mCi/mmol) or [14C]UDPglucose (304 mCi/mmol) purchased from Amersham.
Phylogenetic Reconstruction. The phylogenetic analysis
was performed using the Phylip package (3.6 version; http://
www.evolution.genetics.washington.edu/phylip.html).
Phylogenetic trees were constructed using the neighbor-joining method.
Distances were calculated using the Jones–Taylor–Thorton method. Positions corresponding to insertions–deletions were excluded
from the analysis. Numbers represent bootstrap replicate values
out of 100 (500 bootstrap replicates were computed). The scale
(down, left) indicates the branch length corresponding to the
number of substitutions per site.
Results and Discussion
Comparative Analysis of the Structure and Morphology of Crystalline Storage Polysaccharide from
T. gondii, C. reinhardtii, and C. cohnii. The presence
of abundant floridean starch granules defines a
cytological marker of the bradyzoite stage (Fig. 1E)
and tachyzoites seem devoid of storage polysaccharides (Tomavo 2001). Like the dinoflagellate Crypthecodinium cohnii (Alveolata) (Fig. 1F), the
bradyzoite stage of Toxoplasma gondii (Apicomplexa)
accumulates a considerable amount of amylopectin in
the cytoplasm .(Fig. 1E), while the unicellular green
alga Chlamydomonas reinhardtii (Chlorophyta) contains amylopectin within the chloroplast (Fig. 1D).
Toxoplasma gondii, together with Cryptosporidium
parvum and the red alga Cyanidionschyzon merolae,
defines one of the first floridean starch containing
organisms whose genome sequence is presently
available (http://www.toxodb.org). We therefore initiated a biochemical characterization of the polysaccharide structure and investigated the pathway of
floridean starch biosynthesis and degradation. The
polysaccharide was purified from tachyzoites that
were induced into bradyzoites in vitro. The purified
starch was dispersed in 10 mM NaOH and subjected
to size exclusion chromatography in CL2B columns.
Results were compared to those obtained with the
unicellular green algae Chlamydomonas reinhardtii,
which synthesizes starch through the use of the ADPglucose pathway (Fig. 1A). We also compared the
results to those gathered from the nonphotosynthetic
floridean starch accumulating dinoflagellate Crypthecodinium cohnii (Fig. 1C), a distant cousin of apicomplexan parasites. It is evident from the
chromatograms displayed in Fig. 1B that only the
high mass amylopectin-like fraction could be recov-
260
Fig. 1. Comparative analysis of the
structure and morphology of
starches from Toxoplasma gondii,
Chlamydomonas reinhardtii, and
Crypthecodinium cohnii. CL2B gel
permeation chromatography of
amylopectin from Toxoplasma
gondii (1 mg) dispersed in 10 mM
NaOH (B) in comparison with
Chlamydomonas reinhardtii (1 mg)
(A) and Crypthecodinium cohnii
(1 mg) (C). The kmax (maximal
absorbance wavelength of the
iodine–polysaccharide complex in
nanometers) is scaled on the left axis
and displayed as a thin solid line in
the polysaccharide containing
fractions. The optical density (d) of
the complex at kmax measured for
each fraction is indicated on the
right axis. The low kmax fraction
excluded from the column defines
amylopectin, while the high kmax
amylose is separated throughout the
column (see arrows). Transmission
electron micrographs of nitrogenstarved Chlamydomonas reinhardtii
(D), Toxoplasma gondii bradyzoite
(E), and a vegetative cell of
Crypthecodinium cohnii (F) showing
storage polysaccharide or starch (s).
ered from Toxoplasma polysaccharide granules, while
the low mass amylose fraction recovered from
Crypthecodinium cohnii (Fig. 1C) and Chlamydomonas reinhardtii (Fig. 1A) displayed a distribution in
size and an iodine interaction identical to those of
plant starches. Crystallinity of the Toxoplasma native
starch granules was ascertained through the use of
wide-angle X-ray diffraction and microscopic observations under nonpolarized and polarized light
(Fig. 2). X-Ray powder diffractograms showed a
characteristic B-type of diffraction pattern (data not
shown), while the granules displayed the classical
maltese cross (Fig. 2B) witnessed in many land plant
starches. The absence of amylose was confirmed by
NMR analysis of the purified polysaccharide (Fig. 3).
Indeed, the spectra are consistent with the presence of
a moderately branched (1% of a-1, 6 linkages) amylopectin-like component.
The purified starches or amylopectin were then
subjected to enzymatic debranching and separation
of the debranched chains by capillary electrophoresis
(Fig. 4). The chain-length histogram distribution
displayed in Fig. 4B shows significant differences
from those of both Chlamydomonas (Fig. 4A) and
Crypthecodinium (Fig. 4C) starches. They remained
much more similar to those of amylopectins than to
those of glycogen. This feature is shared by many
other red algae floridean starches. These patterns of
chain-length histogram distribution obtained by
capillary electrophoresis were also confirmed by
HPAEC-PAD analysis (data not shown).
Crude Extract Analysis of the Starch Pathway. Next, we purified as much crude extract as
possible from Toxoplasma tachyzoites that were induced to differentiate into bradyzoites in vitro. We
monitored both UDP-glucose and ADP-glucose
incorporation into rabbit liver glycogen. Results
displayed in Fig. 5 show that ADP-glucose is preferred by the Chlamydomonas enzymes. However,
significantly more incorporation was found by using
UDP-glucose in Toxoplasma gondii and in Crypthecodinium cohnii. The fact that the activity measured in
the presence of UDP-glucose with the T. gondii ex-
261
Fig. 3. Deciphering the fine structure of storage polysaccharide
by NMR analysis. Superimposed one-dimensional 1H-NMR
spectra of Chlamydomonas reinhardtii, Toxoplasma gondii, and
Crypthecodinium cohnii starches. A Note that the abundance of
peaks seen on right side of the diagram prompted us to represent
only the C. cohnii spectrum. The other two samples display exactly
the same profile in this area. B The two-dimensional 1H-NMR
spectra of C. cohnii starch is shown below. Both T. gondii amylopectin and C. reinhardtii starches display identical NMR spectra.
Fig. 2. Crystallinity of the T. gondii amylopectin ascertained
using nonpolarized light (A) and polarized light (B) microscopy.
The arrows indicate the size (1–3 lm) of one amylopectin granule
and the circles display typical maltese crosses.
tracts was only twice that measured with ADP-glucose is, in our opinion, due to the high background
measured. This in turn is due to the small amount of
parasite crude extract that was available to us.
Comparative Analysis of the T. gondii, C. merolae,
S. cerevisiae, C. reinhardtii and E. coli Storage
Polysaccharide Pathway Genes. We probed the
completed Toxoplasma genome sequence for the
presence of all the enzymes known to be important
for starch or glycogen synthesis. We compared these
to corresponding genes found in the green lineage
(Chlamydomonas reinhardtii), in red algae (Cyani-
dioschyzon merolae), in bacteria (E. coli), and in yeast
(Saccharomyces cerevisiae). It is evident from Table 1
that both ADP-glucose pyrophosphorylase and
ADP-glucose utilizing starch synthase could not be
found in apicomplexans or in red algae. It is also
evident that both C. merolae and T. gondii contain a
UDP-glucose utilizing glycogen (starch) synthase-like
sequence and glycogenins. These enzymes are specific
for the eukaryote UDP-glucose based pathway. In
addition, T. gondii but not C. merolae contains an
indirect debranching enzyme similar to those of fungi
and animals. Indirect debranching enzymes are
bifunctional enzymes that carry both an a-1, 4-glucanotransferase and an amylo-1, 6-glucosidase active
site. Table 2 shows that the four distinct conserved
blocks of sequences that typify the N-terminal sequences of such enzymes were found both in T. gondii
and in Cryptosporidium parvum, another amylopectin
producing apicomplexan (Harris et al. 2004).
Among the genes which distinguish plant starch
metabolism from that of the animal fungal and bacterial glycogen pathways, we found both isoamylase
and R1 (glucan water dikinase activity)-like sequences in C. merolae and T. gondii (Table 1).
262
Fig. 4. Comparative polysaccharide chain-length CL distributions. Histogram of CL distributions of Toxoplasma gondii amylopectin (B) in comparison with Chlamydomonas reinhardtii (A) and
Crypthecodinium cohnii (C) starches. The CL distributions were
obtained after isoamylase-mediated enzymatic debranching
through capillary electrophoresis of 8-aminopyrene-1, 3, 6-trisulfonate (APTS)-labeled fluorescent glucans. The X axis displays the
degree of polymerization (DPS 7 to 36) and the Y axis represents
the relative frequency of chains expressed as percentages.
Differential Expression of Starch Biosynthetic and
Degradation Genes. The expression pattern of the
genes listed in Table 1 was investigated by RT- PCR
in tachyzoites and brayzoites isolated from mouse
brain cysts. Results displayed in Fig. 6 show that
enzymes of amylopectin metabolism are transcribed
in tachyzoites and/or bradyzoites. It is noteworthy
that transcripts coding for enzymes known to be involved in the catabolic functions such as the R1
protein (GWD), a-glucan phosphorylase, a-glucosidase, and a-amylase seem to be preferentially expressed in bradyzoites. The indirect debranching
enzyme transcript, despite being found in tachyzoites,
seems to be preferentially expressed in bradyzoites,
suggesting a catabolic function. Transcripts coding
for enzymes known to be involved in glycogen or
starch synthesis are preferentially expressed in tachyzoites (glycogenin, starch (glycogen) synthase, one
branching enzyme isoform) but can also be detected
in lesser amounts in bradyzoites (Fig. 6). This pattern
is consistent with the production of amylopectin
during differentiation of tachyzoites into bradyzoites
and with the mobilization of the glucose stores during
Fig. 5. Substrate requirement (ADP-glucose and/or UDP-glucose) for transglucosylase reaction catalyzed by Chlamydomonas
reinhardtii, Toxoplasma gondii and Crypthecodinium cohnii starch
(glycogen) synthases. Error bars represent the mean and standard
deviations of three reproducible experiments.
bradyzoite- to- tachyzoite interconversion (Tomavo
2001). Isoamylase, which has been proposed to be
responsible for the crystallization of amylopectin, is
found to be transcribed as a biosynthetic gene, consistent with its hypothetical function (Ball and Morell
2003). We believe that the transcription of D-enzyme
(a-1, 4-glucanotransferase) reflects a requirement of
this function during starch biosynthesis as demonstrated by an analysis of Chlamydomonas mutants
(Colleoni et al. 1999). Surprisingly, the preferential
transcription of one particular branching enzyme
sequence in bradyzoites is also suggestive of some
catabolic function for this particular enzyme. A catabolic function remains to be demonstrated for a
branching enzyme in plants. However, plant mutants
defective for the well-conserved BEI have failed to
display a phenotype consistent with a function in
starch biosynthesis (Ball and Morell 2003).
263
Table 1. Number of genes corresponding to enzymes of glycogen and starch metabolism
UDP-glucose pyrophosphorylase
ADP-glucose pyrophosphorylase
Starch (glycogen) synthase
Branching enzyme
Isoamylase
Indirect debranching enzyme
a-1, 4-Glucanotransferase
Phosphorylase
a-Amylase
a-Glucosidase
R1 protein
Glycogenin
Apicomplexa
(T. gondii)
Red algae
(C. merolae)
Green algae
(C. reinhardtii)
Yeast
(S. cerevisiae)
Bacteria
(E. coli)
1
0
1
2
1
1
1
1
1
1
1
1
1
0
1
1
2
0
2
1
1
1
1
1
1
3
6
3
2
0
2
2
3
0
2
0
3
0
2
1
0
1
0
1
0
3
0
2
1
1
1
1
1
0
1
2
2
1
0
0
Note. The protein-deduced sequences were recovered from Escherichia coli (http://www.genome.wisc.edu/), Saccharomyces cerevisiae (http://
www.yeastgenome.org/) Cyanidiumschyzon merolae (http://merolae.biol.s.u-tokyo.ac.jp/) and Chlamydomonas reinhardtii (http://genome.igipsf.org/chlrel/chlrel.home.html). The Toxoplasma gondii enzymes were identified at the genome Web site (http://www.toxodb.org) by
keyword searches or using protein orthologue sequences. Expression of all T. gondii sequences coding for enzymes involved in amylopectin
metabolism was checked by RT-PCR (Fig. 6) and nucleotide sequencing. The sequences of Cryptosporidium parvuum, another apicomplexan
parasite whose genome has been sequenced, is presently not sufficiently assembled and annotated to allow a systematic identification of full
length genes (http://www.cryptodb.org)
Table 2. Consensus sequences of indirect debranching enzymes in humans, rabbit, yeast, C. parvum, and T. gondii
Consensus sequence
Human
Rabbit
Yeast
C. parvum
T. gondii
Conserved amino
acid in the
catalytic site
N-Terminal
domain I
N-Terminal
domain II
N-terminal
domain III
N-Terminal
domain IV
C-Terminal
peptide
C-Terminal
peptide
209-NVICITDVVYNH
232-VLCITDVVYNH
218-MLSLTDIVFNH
338-GILSACDVVLNH
371-GLLSTIDLVLNH
521-GVRLDNC
544-GVRLDNC
530-DGFRIDNC
643-HAIRLDNC
671-GVRLDNC
*
551-VVAELFT
574-VVAELFT
560-YVVAELFS
673-VYAELF
701-WIFAELFT
*
619-ALFMDITHD
642-ALFMDITHD
662-LFMDCTHD
769-AIFFDCTHD
853-LFYDCTHD
*
1099-WGRDTFJ
1122-WGRDTFI
1083-WGRDVFI
1426-WGRDTFI
1505-WGRDTEI
*
1161-IQDYC
1184-IQDYC
1145-VQDYV
1488-ALDYC
1567-IQDYS
*
Asp-535
Glu-564
Asp-670
Asp-1086
Asp-1147
Note. The GenBank and other database accession numbers of indirect debranching enzymes used are as follows: human (P35573), rabbit
(P35574), yeast (Q06625), C. parvum (Q7YY51) and T. gondii (contig TGG_994267 from http://www.toxodb.org). The amino acids were
numbered from the N-terminal end. Conserved residues are in bold face. Asterisks indicate the amino acid residues that are essential for the
catalytic sites.
Phylogenetic Analyses. Phylogenetic analyses performed with two key enzymes (amylopectin synthase
and the R1 protein) demonstrate that the Toxoplasma
gondii amylopectin synthesis pathway has evolved
from the red algal starch synthesizing machinery
through a secondary endosymbiotic event (Figs. 7
and 8). We believe that these phylogenetic data, together with our enzyme assays (Fig. 5), establish that
apicomplexans (Toxoplasma gondii and Cryptosporidium parvum) and red algae, such as C. merolae, use a
UDP-glucose pathway to build insoluble crystalline
amylopectin. That this is the pathway used by all
floridean starch accumulating organisms is suggested
by the recent characterization of a UDP-glucose
utilizing amylopectin synthase in red algae (Nyvall et
al. 1999). However, it is equally apparent that
Toxoplasma and C. merolae also contain plant-like
genes that are not found in yeast and mammals
(Table 1). These consist of genes that are required in
plants and green algae to breakdown and synthesize
semicrystalline polysaccharides. For instance, one of
the early steps of starch mobilization in plants requires the recently identified R1 protein. This enzyme
has also been named a-glucan water dikinase
(GWD). It has been shown to phosphorylate amylopectin at specific positions and is thought to facilitate subsequent hydrolytic attack through amylases
(Ritte et al. 2002). In addition, both bacteria and
plants polysaccharide catabolism leads to the production of malto-oligosaccharides which are not
generated from glycogen in yeasts or mammalian
cells. Toxoplasma gondii and C. merolae also contain,
264
Fig. 6. Results of semiquantitative RT-PCR of expression of
genes coding for enzymes involved in T. gondii amylopectin
metabolism. Note the bradyzoite stage- specific expression of aamylase, a-glucosidase, R1 protein, and a-glucan phosphorylase,
while the other genes are expressed in both tachyzoite and bradyzoite. Except for glycogenin, the other genes are overexpresed
either in tachyzoites (isoamylase, D-enzyme, branching enzyme 2,
and starch [glycogen] synthase) or in bradyzoites (branching enzyme 1, UDP-glucose pyrophosphorylase, and debranching [indirect] enzyme).
respectively, one or two genes encoding enzymes of
malto-oligosaccharide metabolism such as an a-1, 4glucanotransferase (D-enzyme). This hints that the
evolution of semicrystalline polysaccharides has required the maintenance or acquisition of genes similar to those that are necessary in green algae and
plants to mobilize glucose from such polymers (Table 1). The presence of an isoamylase-related expressed sequence in the genome also hints that
maintenance of this activity is required to ensure the
synthesis of an insoluble semicrystalline form of
storage polysaccharide. Indeed, this enzyme has been
proposed to ensure the processing of the hydrosoluble precursor of amylopectin into mature crystalline
material (Ball and Morell 2003).
Apicomplexans and dinoflagellates are suspected
to have evolved following a secondary endosymbiotic
event. It is likely that the ancestral host cell engulfed a
Fig. 7. Phylogenetic tree of starch, amylopectin, and glycogen
synthases. Unrooted phylogenetic tree from starch (glycogen)
synthases using neighbor joining (NJ). Sequences of starch (glycogen) synthases were aligned with ClustalW (Thompson et al.
1994). Boostrap values are indicated. The scale indicates the branch
length corresponding to the number of substitutions per site. The
GenBank and other database accession numbers of the starch
(glycogen) synthases used for the tree analysis are as follows:
H. sapiens (P13807), M. musculus (Q9Z1E4), O. cuniculus (P13834),
S. cerevisiae 1 (P23337), S. cerevisiae 2 (P27471), N. crassa
(O93869), A. thumefaciens (P39670), S. melilati (P58593), B. halodurans (Q9KDX6), B. subtilis (P39125), C. muridarum (Q9PLC3),
C. pneumoniae (Q9Z6V8), C. tepidum (Q8KAY6), C. trachomatis
(084804), C. perfringens (Q8XPA1), D. radiodurans (Q9RWS1),
E. coli (P08323), F. nucleatum (Q8RF65), H. influenzae (P45176),
L. lactis (Q9CHM9), R. loti (Q985P2), R. tropici (Q9EUT5),
S. penumoniae (Q97QS5), T. caldophilus (P58395), T. maritima
(Q9WZZ7), V. cholerae (Q9KRB6), Y. pestis (Q8ZA78), C. merolae
(locus CMM317C from http://merolae.biol.s.u-tokyo.ac.jp/),
T. gondii (contig TGG_994276 from http://toxodb.org), C. reinhardtii (AAL28128), S. acidocaldarius (Q9hh97), S. solfataricus
(Q97ZD3), P. abyssi (Q9V2J8), P. furisus (Q8TZF1), P. aerophilum
(Q8ZT56), A. majus (O82627), A. thaliana (Q9MAQ0), Z. mays GB
(P04713), O. sativa (P19395), P. sativum (Q43092), S. tuberosum
(Q00775), S. vulgare (Q43134), T. aestivum (P27736), A. thaliana
(Q9FNF2), O. sativa (Q40739), and S. tuberosum (P93568).
red alga that contained the UDP-glucose based
pathway of floridean starch metabolism. This host
cell, like most nonphotosynthetic eukaryotes, probably contained its own UDP-glucose-based pathway
of glycogen synthesis. There is a possibility that
additional genes of the ancestral host were maintained and thus added to the pathway of floridean
starch metabolism after secondary endosymbiosis.
The absence and presence of indirect debranching
enzyme sequences, respectively, in red algae and
apicomplexans could be understood in this light.
However, we cannot presently exclude other mechanisms to explain these differences, such as accidental
lateral transfers during evolution. Evidence for such
transfers can be found in the Arabidopsis genome,
where one of the two branching enzymes is clearly
more related to fungal enzymes than to other plant or
bacterial sequences. A similar fate seems to explain
the presence of one of the two branching enzymes in
the phylogeny displayed in Fig. 9. Indeed, the T.
gondii BE1 sequence is more similar to those of
bacteria than to those of green or red algae. However,
the second BE isoform (T. gondii BE2) seems to have
265
Fig. 8. Phylogenetic tree of dikinases including GWD (a-glucan,
water dikinase [also known as R1 protein], PPDK (pyruvate,
phosphate dikinase), and PPS (pyruvate, water dikinase) family.
Unrooted phylogenetic tree from dikinases using neighbor joining
(NJ). Sequences were aligned with ClustalW (Thompson et al.
1994). Boostrap values are indicated. The scale indicates the branch
length corresponding to the number of substitutions per site. The
GenBank and other database accession numbers of the dikinases
used for the tree analysis are as follows: T. gondii (contig
TGG_994467 from http://toxodb.org), C. merolae (locus
CMT547C from http://merolae.biol.s.u-tokyo.ac.jp/), C. parvum
(EAK88766), A. thaliana 1 (AAG47821), A. thaliana 2
(AAO42141), C. reticulata (AAM18228), S. tuberosum (T07050),
T. aestivum (CAC22583), C. symbosium (P22983), F. bidentis
(S56650), O. sativa (BAA22420), E. histolytica (AAA 18944),
Z. mays (P1 1155), E. coli (S20554), M. maripaludis (AAD28736),
N. meningitidis (NP_273662), P. aeruginosa (AAG05159), and
S. marinus (S51006).
evolved separately from other BE of fungi, animals,
plants, and bacteria. Once again, this could be related
to genes that have been retained from the host cell
that engulfed a red alga.
Despite the presence of these additional plant-like
genes, both Toxoplasma and C. merolae define
remarkably simple systems leading to the biosynthesis of semicrystalline amylopectin polymers. Indeed, a
recent study establishes that multiple forms of enzymes catalysing each synthetic or degradation step is
an ancient characteristic of the green lineage (Ral et
al. 2005). It is indeed remarkable that Ostreoccus
tauri, a unicellular green alga, requires, for example,
six distinct elongation enzymes to build a similar
structure and that these isoenzymes have been conserved throughout the evolution of plants.
The results reported in this paper, together with
our knowledge of storage polysaccharide in green
algae, suggest that there has been a strong drive to
maintain storage polysaccharides into an insoluble
semicrystalline physical state. This has occurred in
several independent lineages using very different
enzymatic toolkits. In agreement with an apicoplast
Fig. 9. Phylogenetic tree of starch (glycogen) branching enzymes
(BE). Unrooted phylogenetic tree from starch (glycogen) branching
enzymes using neighbor joining (NJ). Sequences were aligned with
ClustalW (Thompson et al. 1994). Boostrap values are indicated.
The scale indicates the branch length corresponding to the number
of substitutions per site. The GenBank and other database accession numbers of the starch (glycogen) branching enzymes used for
the tree analysis are as follows: B. subtilis (P39118), E. coli
(P07762), H. sapiens (Q04446), S. tuberosum (P30924), S. cerevisiae
(P32775), T. gondii BE1 (contig TGG_994676 from http://
www.toxodb.org), G. gracilis (AAB97471), S. pneumoniae
(Q97QS8), C. parvum (CAD98370), C. merolae (locus CMH144C
from http://merolae.biol.s.u-tokyo.ac.jp/), E. chrysanthemi
(Q8GQC5), M. tuberculosis (Q10625), O. sativa (Q01401),
M. musculus (NP_083079), F. catus (AAR13899), T. gondii BE2
(contig TGG_994574 from http://www.toxodb.org), A. thaliana
(NP_1 95985), Z. mays (Q08047), X. axonopodis (Q8PR13),
B. japonicum (Q89FD3), R. solanasearum (Q8XT76), and Y. pestis
(Q8ZA75).
of red algal evolutionary origin (Wilson et al. 1996;
Fast et al. 2001), apicomplexan parasites have
inherited the pathway of floridean starch metabolism.
This pathway involves a mosaic of genes of prokaryotic and eukaryotic origin. The green algae and
land plants display a pathway relying entirely on
genes related to the bacterial glycogen metabolism.
Because plastids are presently thought to be monophyletic, it is reasonable to assume that the common
ancestors of the green and red lineages displayed both
eukaryotic and prokaryotic pathways. Subsequently,
when the ancestors of green and red algae diverged
they maintained either the endosymbiont or the host
pathways of glycogen accumulation. This occurred
with a concomitant loss of most of the host or
endosymbiont redundant functions. It is striking that
the cellular localization of storage polysaccharides in
either the endosymbiont or the host cytoplasm correlates with the pathway chosen. Despite the loss of
photosynthesis in apicomplexans, the presence of this
inherited pathway remained suitable and probably
required for those parasites that harbor classical
266
dormant cysts in their life cycles. However, for those
apicomplexans such as Plasmodium species, which
propagate without dormant encysted forms, storage
polysaccharide metabolism was lost. Indeed, screening the full genome database of Plasmodium falciparum (http://www.plasmoDB.org) failed to yield any
sequence involved in glycogen or starch metabolism.
Interestingly in bacteria, the loss of glycogen metabolism has been correlated with the more parasitic and
dependent life cycles (Henrissat et al. 2002).
Concluding Remarks. Our findings indicate that
the storage polysaccharide accumulated in the
cytoplasm of apicomplexans such as Toxoplasma
gondii and Cryptosporidium parvum and that of
the dinoflagellate Crypthecodinium or of the red alga
C. merolae are synthesized using a UDP-glucosebased metabolic pathway. This metabolic pathway is
very similar to the fungal and animal glycogen pathway and distinct from the plant starch pathway. Thus,
the complexity that has been conserved throughout
the evolution of green plants is not per se required to
build semicrystalline polymers in apicomplexan parasites. It is also striking that the only genes that distinguish the parasiteÕs pathway from that of fungi and
animal cells are precisely those that are thought to
have been recruited from the bacterial ancestor of
plastids by green plants to achieve the biosynthesis of
semicrystalline polysaccharides. These data shed light
not only on the evolution of the green and red lineages
of photosynthetic eukaryotes but also on the basic
distinction between hydrosoluble glycogen and semicrystalline starch metabolism. Considerable progress
has been made in understanding the biogenesis of
glycogen in mammals and that of starch in plants. The
different sets of enzymes involved have been identified
and characterized. Developing a better understanding
of amylopectin biogenesis in Toxoplasma gondii is
worthy of further exploration since compounds which
inhibit this process could define potential drugs for
combating apicomplexan-mediated diseases.
Acknowledgments.
For comments, logistical support, and
invaluable technical assistance in the field and lab, we thank Michael Kibe, Marléne Mortuaire, Hervé Moreau, Christian Slomianny, Emmanuel Maes, Frédéric Chirat, Yves Leroy, Florence
Dzierszinski, Brigitte Bouchet, and Bruno Pontoire. We acknowledge the Toxoplasma Genome Sequencing Consortium for making
available the genome database: Preliminary genomic and/or cDNA
sequence data were accessed via http://ToxoDB.org and/or http://
www.tigr.org/tdb/t_gondii/. Genomic data were provided by The
Institute for Genomic Research (supported by NIH Grant
AI05093), and by the Sanger Center (Wellcome Trust). EST sequences were generated by Washington University (NIH Grant
1R01AI045806-01A1). This research was funded by the Centre
National de la Recherche Scientifique (CNRS) through the Action
Thématique Incitative sur Programme et Equipe (ATIPE), the
Programme Inter-organisme de Microbiologie Fondamentale, and
the Agence Nationale de la Recherche sur le Sida (ANRS).
References
Ball SG, Morell MK (2003) From bacterial glycogen to starch:
understanding the biogenesis of the plant starch granule. Annu
Rev Plant Biol 54:207–233
Ballicora MA, Iglesias AA, Preiss J (2003) ADP-glucose pyrophosphorylase, a regulatory enzyme for bacterial glycogen
synthesis. Microbiol Mol Biol Rev 67:213–225
Buléon A, Gallant D-J, Bouchet B, Mouille G, DÕHulst C, Kossman J, Ball SG (1997) Starches from A to C. Chlamydomonas
reinhardtii as a model microbial system to investigate the biosynthesis of the plant amylopectin crystal. Plant Physiol
115:949–957
Cai X, Fuller AL, McDougald LR, Zhu G (2003) Apicoplast
genome of the coccidian Eimeria tenella. Gene 321:39–46
Cavalier-Smith T (1999) Principles of protein targeting in secondary symbiogenesis: euglenoid, dinoflagellate and sporozoan
plastid origins and the eukaryote family tree. J Euk Microbiol
46:347–366
Colleoni C, Dauvillee D, Mouille G, Buleon A, Gallant D, Bouchet
B, Morell M, Samuel M, Delrue B, dÕHulst C, Bliard C, Nuzillard JM, Ball S (1999) Genetic and biochemical evidence for
the involvement of alpha-1, 4-glucanotransferases in amylopectin synthesis. Plant Physiol 120:993–1004
Douglas SE, Penny SL (1999) The plastid genome of the cryptophyte alga, Guillardia theta: complete sequence and conserved
synteny groups confirm its common ancestry with red algae. J
Mol Evol 48:236–244
Dzierszinski F, Popescu O, Toursel C, Slomianny C, Yahiaoui B,
Tomavo S (1999) The protozoan parasite Toxoplasma gondii
expresses two functional plant-like glycolytic enzymes. Implications for evolutionary origin of apicomplexans. J Biol Chem
274:24888–24895
Dzierszinski F, Mortuaire M, Dendouga N, Popescu O, Tomavo S
(2001) Differential expression of two plant-like enolases with
distinct enzymatic and antigenic properties during stage conversion of the protozoan parasite Toxoplasma gondii. J Mol Biol
309:1017–1027
Fast NM, Kissinger JC, Roos DS, Keeling PJ (2001) Nuclear-encoded, plastid-targeted genes suggest a single common origin
for apicomplexan and dinoflagellate plastids. Mol Biol Evol
18:418–426
Funes S, Davidson E, Reyes-Prieto A, Magallon S, Herion P, King
MP, Gonzalez-Halphen D (2002) A green algal apicoplast
ancestor. Science 298:2155
Greenberg E, Preiss J (1964) The occurrence of adenosine
diphosphate glucose: glycogen transglucosylase in bacteria. J
Biol Chem 239:4314–4315
Harris JR, Adrian M, Petry F (2004) Amylopectin: a major component of the residual body in Cryptosporidium parvum oocysts.
Parasitology 128:269–282
Henrissat B, Deleury E, Coutinho PM (2002) Glycogen metabolism loss: a common marker of parasitic behaviour in bacteria?.
Trends Genet 18:437–440
Karkhanis YD, Allocco JJ, Schmatz DM (1993) Amylopectin
synthase of Eimeria tenella: identification and kinetic characterization. J Euk Microbiol 40:594–598
Köhler S, Delwiche CF, Denny PW, Tilney LG, Webster P, Wilson
RJ, Palmer JD, Roos DS (1997) A plastid of probable green
algal origin in apicomplexan parasites. Science 275:1485–1489
Libessart N, Maddelein ML, Van den Koornhuyse N , Decq A,
Delrue B, Mouille G, DÕHulst C, Ball S (1995) Storage, photosynthesis, and growth: the conditional nature of mutations
affecting starch synthesis and structure in Chlamydomonas.
Plant Cell 7:1117–1127
Luft BJ, Remington JS (1988) AIDS commentary. Toxoplasmic
encephalitis. J Infect Dis 157:1–6
267
Luft BJ, Remington JS (1992) Toxoplasmic encephalitis in AIDS.
Clin Infect Dis 15:211–222
Marechal E, Cesbron-Delauw MF (2001) The apicoplast: a new
member of the plastid family. Trends Plant Sci 6:200–205
Matsuzaki M, Misumi O, Shin IT, Maruyama S et al. (2004)
Genome sequence of the ultrasmall unicellular red alga Cyanidioschyzon merolae 10D. Nature 428:653–657
McFadden GI, Reith ME, Munholland J, Lang-Unnasch N (1996)
Plastid in human parasites. Nature 381:482
Nyvall P, Pelloux J, Davies HV, Pedersen M, Viola R (1999)
Purification and characterisation of a novel starch synthase
selective for uridine 5¢-diphosphate glucose from the red alga
Gracilaria tenuistipitata. Planta 209:143–152
Perret E, Davoust J, Albert M, Besseau L, Soyer-Gobillard M-O
(1993) Microtubule organization during the cell cycle of the
primitive eukaryote dinoflagellate Crypthecodinium cohnii. J
Cell Sci 104:639–651
Ral JP, Derelle E, Ferraz C, Wattebled F, Farinas B, Corellou F,
Buléon A, Slomianny MC, Delvalle D, dÕHulst C, Rombauts S,
Moreau H, Ball S (2004) Starch division and partitioning a
mechanism for granule propagation and maintenance in the
picophytoplanktonic green alga Ostreococcus tauri. Plant
Physiol 36:3333–3340
Recondo E, Leloir L (1961) Adenosine diphosphate glucose and
starch biosynthesis. Biochem Biophys Res Commun 6:85–
88
Ritte G, Lloyd JR, Eckermann N, Rottmann A, Kossmann J,
Steup M (2002) The starch-related R1 protein is an alpha-glucan, water dikinase. Proc Natl Acad Sci USA 99:7166–7171
Ryley JF, Bentley M, Manners DJ, Stark JR (1969) Amylopectin,
the storage polysaccharide of the coccidian Eimeria brunetti and
E. tenella. J Parasitol 55:839–845
Seeber F (1997) Consensus sequence of translational initiation sites
from Toxoplasma gondii genes. Parasitol Res 83:309–311
Thompson JD, Higgins DG, Gibson TJ (1994) ClustalW: improving
the sensitivity of progressive multiple sequence alignment
through sequence weighting, position-specific gap penalties and
weight matrix choice. Nucleic Acids Res 22:4673–4680
Tomavo S (2001) The differential expression of multiple isoenzyme
forms during stage conversion of Toxoplasma gondii: an adaptive developmental strategy. Int J Parasitol 31:1023–1031
Viola R, Nyvall P, Pedersen M (2001) The unique features of starch
metabolism in red algae. Proc R Soc Lond B Biol Sci 268:1417–
1422
Waller RF, Keeling PJ, Van Dooren GG , McFadden GI (2003)
Comment on a green algal apicoplast ancestor. Science 301:49
Wilson RJM, Denny PW, Preiser PR, Rangachari K, Roberts K,
Roy A, Whyte A, Strath M, Moore DJ, Moore PW, Williamson
DH (1996) Complete gene map of the plastid-like DNA of the
malaria parasite Plasmodium falciparum. J Mol Biol 261:155–172
Zhang Z, Green BR, Cavalier-Smith T (1999) Single gene circles in
dinoflagellate chloroplast genomes. Nature 400:155–159