Apicomplexan rhomboids have a potential role in microneme protein
cleavage during host cell invasion
Timothy J. Dowsea,1, John C. Pascallb,1, Kenneth D. Brownb, Dominique Soldatia,c,*
aDepartment of Biological Sciences, Imperial College London, Alexander Fleming Building, South Kensington Campus, London SW7 2AZ, UK
bSignalling Programme, The Babraham Institute, Babraham Hall, Cambridge CB2 4AT, UK
cDepartment of Microbiology and Genetics, Faculty of Medicine, University of Geneva CMU, 1 rue Michel-Servet, 1211 Geneva 4, Switzerland
Received 25 February 2005; received in revised form 29 March 2005; accepted 3 April 2005
Apicomplexan parasites secrete transmembrane (TM) adhesive proteins as part of the process leading to host cell attachment and invasion.
These microneme proteins are cleaved in their TM domains by an unidentified protease termed microneme protein protease 1 (MPP1). The
cleavage site sequence (IAYGG), mapped in the Toxoplasma gondii microneme proteins TgMIC2 and TgMIC6, is conserved in microneme
proteins of other apicomplexans including Plasmodium species. We report here the characterisation of novel T. gondii proteins belonging to
the rhomboid family of intramembrane-cleaving serine proteases. T. gondii possesses six genes encoding rhomboid-like proteins. Four are
localised along the secretory pathway and therefore constitute possible candidates for MPP1 activity. Toxoplasma rhomboids TgROM1,
TgROM2 and TgROM5 cleave the TM domain of Drosophila Spitz, an established substrate for rhomboids from several species,
demonstrating that they are active proteases. In addition, TgROM2 cleaves chimeric proteins that contain the TM domains of TgMIC2 and
q 2005 Australian Society for Parasitology Inc. Published by Elsevier Ltd. All rights reserved.
Keywords: Apicomplexa; Intramembrane; Cleavage; Rhomboid; Serine; Protease; Toxoplasma gondii
Toxoplasma gondii belongs to the phylum Apicomplexa
which is composed of a large number of protozoan parasites
that cause severe diseases in a wide variety of hosts,
including humans, farm and household animals. The most
notorious member is Plasmodium falciparum, the etiologic
agent of malaria, which is responsible for millions of deaths
every year. T. gondii causes toxoplasmosis, a disease which
can be fatal in immunocompromised patients and may lead
to severe congenital defects if pregnant women develop a
primary infection. T. gondii and other apicomplexans
invade host cells by an active process dependent on the
actomyosin system of the parasite (Dobrowolski and Sibley,
1996; Dobrowolski et al., 1997; Meissner et al., 2002).
Invasion also requires the sequential secretion of the
contents of membrane-bound organelles called micronemes
and rhoptries (Carruthers and Sibley, 1997). Micronemes
discharge diverse adhesive proteins that bind to host cells
(reviewed in Carruthers, 2002; Dowse and Soldati, 2004)
and redistribute towards the posterior pole of the parasite
during invasion (Carruthers et al., 1999). Among these
microneme proteins, the members of the thrombospondin-
related anonymous protein (TRAP) family, including
TgMIC2 in T. gondii, are known to play an essential role
in host cell invasion (Sultan et al., 1997; Yuda et al., 1999;
Templeton et al., 2000; Huynh et al., 2003). In T. gondii,
microneme proteins form complexes, such as TgMIC1/
MIC4/MIC6, TgMIC3/MIC8 and TgMIC2/M2AP (Soldati
et al., 2001; Carruthers, 2002; Dowse and Soldati, 2004),
which are composed of a transmembrane escorter protein
(TgMIC6, TgMIC8 or TgMIC2) essential for the correct
International Journal for Parasitology 35 (2005) 747–756
0020-7519/$30.00 q 2005 Australian Society for Parasitology Inc. Published by Elsevier Ltd. All rights reserved.
*Corresponding author. Address: Department of Microbiology and
Molecular Medicine, Faculty of Medicine, University of Geneva, CMU, 1
rue Michel-Servet, 1211, Geneva 4, Switzerland. Tel.: C44 20 7594 5342;
fax: C44 20 7584 2056.
1These authors contributed equally to this work.
targeting of the complex to the micronemes, and soluble
proteins, some of which exhibit host cell binding properties
(TgMIC1, TgMIC4 and TgMIC3) (Fourmaux et al., 1996;
Garcia-Reguet et al., 2000; Brecht et al., 2001).
After discharge by the micronemes, TgMIC2 is pro-
cessed by two proteolytic activities named microneme
protein protease 1 and 2 (MPP1 and MPP2, respectively)
(Carruthers et al., 2000). Microneme protein protease 1
cleavage results in the release of the ectodomain of TgMIC2
from the parasite surface, an event that is critical for
efficient invasion (Brossier et al., 2003). The release of
TgMIC2 was previously shown to be sensitive to the serine
(Carruthers et al., 2000). 3,4-Dichloroisocoumarin may act
directly or indirectly on MPP1. The C-terminal MPP1
cleavage also occurs on several other transmembrane
proteins such as TgMIC6, TgMIC12, TgMIC8 (Reiss
et al., 2001) and TgAMA-1 (Donahue et al., 2000) as well
as members of the TRAP family in Plasmodium and other
apicomplexan species. The cleavage site of TgMIC6 by
MPP1 was recently mapped by mass spectrometry, and
shown to be within the transmembrane (TM) domain at the
site IAYGG, a sequence conserved in several apicomplexan
transmembrane microneme proteins (Opitz et al., 2002). A
recent study confirmed that TgMIC2 is also cleaved at this
conserved intramembrane site (Zhou et al., 2004). Addition-
ally, a lysine situated 11 residues upstream from the TM
domain of TgMIC2 was shown to be required for MPP1
mediated cleavage and successful invasion (Brossier et al.,
2003). Taken together, these findings suggest that MPP1
activity is most likely essential and ubiquitous throughout
the phylum of Apicomplexa.
Several groups have suggested that a rhomboid-like
protease may be responsible for MPP1 activity (Urban and
Freeman, 2003; Dowse and Soldati, 2004; Kim, 2004;
Sibley, 2004). Rhomboids form a family of polytopic
membrane proteases conserved throughout evolution and
found in bacteria, yeast, plants and animals (Koonin et al.,
2003). The founding member of the family, Drosophila
Rhomboid-1 was characterised as an intramembrane serine
protease, which cleaves the epidermal growth factor
(EGF)-like substrates Spitz, Gurken and Keren, resulting
in the secretion of the soluble factors from the cell (Lee
et al., 2001; Urban et al., 2001, 2002a). Rhomboids are the
only known intramembrane proteases of the serine class,
and are inhibited by the serine protease inhibitor DCI
(Urban et al., 2001). More recent studies have established
that members of this large family of proteases are involved
in controlling processes other than intercellular signalling,
such as mitochondrial membrane fusion (Esser et al., 2002;
Herlan et al., 2003; McQuibban et al., 2003), and new
substrates have been identified for mammalian rhomboids
(Lohi et al., 2004; Pascall and Brown, 2004). Some
rhomboids exhibit broad substrate specificity, recognising
helix-destabilising residues in the luminal region of a TM
domain and cleaving diverse substrates, including
transmembrane adhesins of T. gondii (Urban and Freeman,
Rhomboid-like proteins have recently been identified in
the genome of P. falciparum (Wu et al., 2003), and we
report here the characterisation of TgROM1, TgROM2,
TgROM4, and TgROM5, four members of a family of six
rhomboid-like proteins in T. gondii.
2. Materials and methods
2.1. General reagents
Restriction enzymes were purchased from New England
Biolabs and Invitrogen. The secondary antibodies for
Western blotting were from Biorad or Amersham
Biosciences, and for immunofluorescence were from
Molecular Probes. Reverse-transcriptase-PCR (RT-PCR)
amplifications were performed using the Titan One-Tube
RT-PCR kit (Roche). All PCR reactions were performed
with either the GeneAmp High Fidelity PCR system or
AmpliTaq (Applied Biosystems).
2.2. Parasite strains
T. gondii tachyzoites (RH strain wild-type and
RHhxgprtK) were grown in human foreskin fibroblasts
(HFF) maintained in Dulbecco’s Modified Eagle’s Medium
(DMEM, Gibco) supplemented with 10% foetal calf serum
(FCS), 2 mM glutamine and 25 mg/ml gentamicin. Parasites
were harvested after complete lysis of the host cells and
purified by passage through 3.0 mm filters and centrifugation
2.3. Cloning of the T. gondii rhomboid genes
Plasmids pT8mycTgROM1, pT8mycTgROM2 and
constructed by cloning the corresponding cDNA between
the NsiI and the PacI sites in T8MycGFP-HX (Hettmann et
al., 2000). Total RNA was isolated from freshly released
parasites using TRIzol (Invitrogen). The cDNAs for each
gene were amplified by RT-PCR according to the
manufacturer’s instructions, using the Titan one tube
RT-PCR system (Roche) and the primers listed in Table 1.
The plasmid pT8TgROM1Ty was generated by cloning the
TgROM1 cDNA between the EcoR I and Nsi I sites in the
pT8MLCTy-HX vector (Herm-Gotz et al., 2002).
TgROM4 and pHA-TgROM5 were prepared by cloning
the appropriate cDNA between the BamH I and EcoR I sites
on the pCAN-HA2 vector, a mammalian expression vector
based on pcDNA3 that adds an N-terminal hemagglutinin
(HA) tag. The plasmids pHA-Rho1 and pHA-RHBDL2
carry inserts encoding Drosophila Rhomboid-1 and human
T.J. Dowse et al. / International Journal for Parasitology 35 (2005) 747–756748
2.4. Parasite transfection (RHhxgprtK) and selection
of stable transformants
Transient transfections were undertaken by electropora-
tion as previously described (Soldati and Boothroyd, 1993).
Stable transformants were selected for by hypoxanthine-
expression in the presence of mycophenolic acid and
xanthine as described earlier (Donald et al., 1996). Parasites
were cloned by limiting dilution in 96 well plates and
analysed for the expression of the transgenes by indirect
immunofluorescence assay (IFA).
2.5. Cell fractionation and Western blot analysis
of T. gondii tachyzoites
Crude extracts, or pellet and soluble fractions from
freeze-thaw lysis of T. gondii tachyzoites were subjected to
SDS-PAGE as described previously (Soldati et al., 1998).
Western blot analysis was carried out using on 10%
polyacrylamide gels run under reducing conditions and the
samples that were probed for rhomboid proteins were not
boiled prior to loading. Proteins were transferred to Hybond
ECL nitrocellulose. For detection, the membranes were
incubated with primary antibodies (anti-myc, 1:1,000; anti-
Ty-1, 1:1,000; anti-HA, 1:500) diluted in PBS, 0.05%
Tween20,5% skimmed milk, and then with affinity-purified
horseradish peroxidase-conjugated goat anti-mouse IgG or
goat anti-rabbit IgG (1:3,000) and bound antibodies
visualised using the ECL system (Amersham).
2.6. Indirect immunofluorescence confocal microscopy
All manipulations were carried out at room temperature.
Intracellular parasites grown in HFF on glass slides were
fixed with 4% paraformaldehyde for 20 min. Following
fixation, slides were rinsed in PBS-0.1 M glycine. Cells
were then permeabilised in PBS, 0.2% Triton-X-100 for
20 min and blocked in the same buffer with 2% BSA. Slides
were incubated for 60 min with primary antibodies diluted
in blocking solution, washed and incubated for 60 min with
Alexa 594 goat anti-rabbit or Alexa 488 goat anti-mouse
antibodies (Molecular Probes), diluted 1:3,000 in blocking
solution. Slides were mounted in Vectashield and kept at
4 8C in the dark. Confocal images were collected with
a Leica laser scanning confocal microscope (TCS-NT
DM/IRB) using a 100! Plan-Apo objective with NA
1.30. Single optical sections were recorded with an optimal
pinhole of 1.0 (according to Leica instructions) and 16 times
averaging. All other micrographs were obtained with a Zeiss
Axiophot with a camera (Photometrics Type CH-250).
Adobe Photoshop (Adobe Systems, Mountain View, CA)
was used for image processing.
2.7. Mammalian cell culture and transfections
HEK-293T cells were maintained in DMEM containing
100 i.u. Penicillin, 100 mg/ml streptomycin and 10% (v/v)
fetal calf serum. Cells for transfection were seeded on
35-mm culture dishes (Nunc) that had been pretreated with
poly-L-lysine. After 24 h, the cultures were transfected with
plasmid DNA using Lipofectamine (Invitrogen) as
described (Pascall et al., 2002). Cells were transfected
with DNA constructs encoding a myc-tagged substrate
protein (0.5 mg) and Star (a protein required for Spitz
trafficking; 0.5 mg), with or without a HA-tagged rhomboid
(Drosophila Rhomboid-1 (0.1 mg) or human RHBDL2
(0.1 mg) or a TgROM (0.5 mg). Plasmid pcDNA3 was
included to maintain the total input DNA at 2 mg. Cells
Primers used in this study
TgROM5-11 MfeI/PacI pT8myc
T.J. Dowse et al. / International Journal for Parasitology 35 (2005) 747–756749
and medium were harvested for Western blot analysis of
immunoreactive proteins approximately 24 h after the start
of transfection as described (Pascall and Brown, 2004).
3.1. Toxoplasma gondii possesses six
An agreement for the nomenclature of apicomplexan
rhomboids has been made between our groups and others
(Dowse and Soldati, 2005). Six genes coding for rhomboid-
like proteins (ROMs) have been found in T. gondii, and are
named TgROM1-6. The cDNAs corresponding to
TgROM1, TgROM2, TgROM4 and TgROM5 were ampli-
fied from T. gondii tachyzoite total RNA by RT-PCR,
cloned and sequenced. All attempts to amplify the cDNA for
TgROM3 failed, suggesting that this gene is not transcribed
in the tachyzoite stage. Supporting this view, all ESTs
specific for the TgROM3 gene are found in cDNA prepared
from unsporulated oocysts while none are present in the
abundant tachyzoite EST database (http://www.cbil.upenn.
edu/paradbs-servlet/). TgROM6 exhibits a predicted mito-
chondrial targeting signal at the N-terminus, and clusters
with other mitochondrial rhomboids in a phylogenetic
analysis (Dowse and Soldati, 2005), thus it is very likely
fulfilling a function unrelated to MPP1.
Rhomboids are conserved in apicomplexan parasites,
some being of similar size to Drosophila Rhomboid-1, and
others being significantly larger. The amino acid sequence
alignment of some of the smaller and larger ROMs from
T. gondii, P. falciparum, Plasmodium berghei and Eimeria
tenella are shown in Fig. 1(A) and (B), respectively. All
T. gondii sequences contain six to eight putative TM
domains as predicted by TMpred, (http://www.ch.embnet.
org/software/TMPRED_form.html) and TMHMM (http://
The apicomplexan ROMs contain many of the critical
conserved residues shown to be required for Drosophila
Rhomboid-1 activity. Urban et al. (2001) demonstrated that
mutations of W151, R152, N169, G215, S217 (the putative
catalytic serine) and H281 abolished Rhomboid-1
activity. The putative catalytic serine in most rhomboid-
like proteases is found within the sequence GASG
(Koonin et al., 2003). In the smaller apicomplexan ROMs,
the sequence GAST is found, whereas in the larger ROMs
the sequence GSSG is present and conserved. Residues
corresponding to N169 and H281, the other two residues
proposed to comprise the catalalytic triad of Drosophila
Rhomboid-1 (Urban et al., 2001), are conserved in all
apicomplexan ROMs (Fig. 1(A) and (B)). Another charac-
teristic of rhomboid-like proteins that is also found in the
apicomplexan ROMs is the presence of very short loops
between the predicted TM domains, with the exception of
the loop between the first and second TM domains.
In addition, the larger apicomplexan ROMs contain a
second longer loop between the sixth and seventh TM
domains and also have longer N and C termini than the
smaller rhomboids (Fig. 1(B)).
3.2. Subcellular distribution of the T. gondii ROM proteases
Epitope-tagging each cDNA and expressing the trans-
genes in T. gondii under control of the tubulin-1 promoter
allowed examination by immunofluorescence microscopy
of the subcellular localisations of TgROM1, TgROM2,
TgROM4, and TgROM5. TgROM1 tagged either at the
N-terminus (myc) or at the C-terminus (Ty-1) localised to a
compartment at the apical end of the parasite (Fig. 2(A) and
results not shown), suggesting a microneme localisation
based on the predominant colocalisation with the micro-
neme marker TgMIC4. Additional labelling in the late
secretory pathway is possibly an effect of protein over-
expression as a similar distribution with a significant
accumulation in the secretory pathway was previously
observed with some microneme proteins (Reiss et al., 2001).
Since N- and C-terminal tagging of TgROM1 resulted in the
same localisation, the other ROMs were tagged only with an
N-terminal myc-epitope. mycTgROM2 showed a focused
staining at the apical side of the nucleus, typical of
Golgi staining in T. gondii (Pelletier et al., 2002).
Double immunofluoresce analysis with the Golgi marker
GRASP-YFP (Pelletier et al., 2002), suggests that TgROM2
accumulates in a distinct region of the Golgi stack from
GRASP-YFP, likely to correspond to the trans-Golgi
network (TGN) (Fig. 2(B)). Both mycTgROM4 and
mycTgROM5 localised to the parasite pellicle, potentially
the plasma membrane (Fig. 2(C) and (D)). When
mycTgROM5 is overexpressed, it is localised to the surface,
although the signal is not as homogenous as for
mycTgROM4. Furthermore, unidentified internal structures
are labelled as well as residual bodies, which form in the
parasitophorous vacuole during parasite division. TgMIC4
is present in the interior of these structures, which are
indicative of the parasite poorly tolerating the overexpres-
sion of ROM5. Western blot analysis of the parasites
expressing TgROM1Ty and mycTgROM4 showed that
the tagged proteins are of expected size and separate into the
membrane fraction (Fig. 3).
3.3. TgROM1, TgROM2 and TgROM5 are active proteases
The TM domain of the Drosophila Spitz protein is
cleaved by some, but not all, rhomboids from a variety of
organisms (Urban et al., 2001, 2002b; Urban and Freeman,
2003). However, the efficiency of release of the cleaved
product depends on where processing occurs in the cell.
Drosophila Rhomboid-1 cleaves Spitz in the Golgi and
soluble Spitz is efficiently released from cells (Urban et al.,
2001, 2002a). However, when Rhomboid-1 is artificially
retained in the ER by adding a KDEL retrieval signal to its
T.J. Dowse et al. / International Journal for Parasitology 35 (2005) 747–756750
Fig. 1. Sequence analysis of Toxoplasma gondii rhomboid-like protein (ROM) sequences. (A) Alignment of the shorter ROMs from some apicomplexan
species (TgROM1, AAT29065; TgROM2, AAT29066; TgROM3, AAT39987; PfROM1, PF11_0150; PfROM3, MAL8P1.16; PbRom1, Pb_82b04plc;
PbROM3, Pb_213e08plc; CpROM1, CpIOWA_III_s2; EtROM1, Contig877; EtROM3, Contig7050). Red residues: 40% conserved or identical. Blue
residues: 40% similar. (B) Alignments of the longer ROMs from some apicomplexan species, excluding the divergent N and C termini. (TgROM4,
AAT29067; TgROM5, AAT47708; PfROM4, PFE0340c; PbROM4, Pb_256f11plc; PyROM4, PY04351). In both alignments TMpred predicted
tranmembrane (TM) domains are highlighted in grey, TMHMM predicted TM domains are in bold letters. Residues corresponding to W151, R152,
N169, G215, S217 and H218 of Drosophila Rhomboid-1 are underlined and marked with arrows. Consensus line: * denotes conserved or similar
residues, ! denotes identical residues.
T.J. Dowse et al. / International Journal for Parasitology 35 (2005) 747–756751
C-terminus, the release of ER-cleaved Spitz is inefficient
(Urban et al., 2002a). Furthermore, several bacterial
rhomboids cleave the Spitz TM domain when the proteins
are co-expressed in COS cells but, in some cases, little if any
cleaved product is released, suggesting that these rhomboids
may be retained in the ER of mammalian cells (Urban et al.,
2002b). Consequently, to establish whether a particular
rhomboid can cleave Spitz, it is necessary to look for the
presence of a cleaved product in cell lysates as well as in
medium samples. However, the detection of a cleaved
intracellular product is complicated by the complex protein
band pattern that results from the extensive glycosylation of
Spitz (Schweitzer et al., 1995;Urban et al., 2001, 2002a). To
circumvent this problem, we generated myc-tagged Spitz
from which amino acids 31–77 had been deleted (named
DSpitz), thereby removing one potential N-glycosylation
site and 21 possible O-glycosylation sites.
To determine whether TgROMs are able to cleave the
TM domain of Spitz, HEK293T cells were co-transfected
with constructs encoding myc-tagged DSpitz together with
HA-tagged TgROMs. In Western Blots, HA-tagged
TgROM1 and TgROM5 resolve as single bands of 31 kDa
and O97 kDa apparent molecular weight, respectively,
whereas TgROM2 appears as a series of bands (24–31 kDa
and 41 kDa) (compare Fig. 4(A)). The abundance of this last
band is markedly increased on heating the sample prior to
electrophoresis (results not shown), and may be due
to aggregation of the protein. We have been unable to
demonstrate expression of HA-tagged TgROM4 in
HEK293T cells (not shown). Drosophila Rhomboid-1 and
human RHBDL2, which are known to cleave the Spitz TM
domain (Urban et al., 2001), were used as positive controls
in these experiments and caused the appearance of a 32 kDa
myc-reactive protein, cleaved DSpitz, that was detectable
both intracellularly (arrowhead in Fig. 4(B)) as well as in
Fig. 2. Subcellular localisation of the Toxoplasma gondii rhomboid-like
proteins (ROMs) by epitope tagging (N-terminal myc tag) in tachyzoites.
green, and anti-MIC4 (a-MIC4, micronemal marker) in red. Scale bars 5 mm
(B) Golgi-localisation of TgROM2; a-myc in green or red (upper and lower
plasma membrane; a-myc in green and a-MIC4 in red. Scale bars 2 mm.
(D) Peripheral localisation of TgROM5 at the plasma membrane and in
undefined intracellular structures and residual bodies sometimes produced
after parasite division; a-myc in green, a-MIC4 or a-GAP45 in red. Scale
bars 2 mm.
Fig. 3. Western blot of recombinant epitope-tagged TgROM1 and 4 in
Toxoplasma gondii, probing total cell lysate and soluble and pellet
fractions. (A) Anti-Ty1 (a-Ty1) clearly detects TgROM1Ty1 in the cell
lysate and the pellet but not in the soluble fraction. (B) Anti-myc (a-myc)
clearly detects mycTgROM4 (arrow) in the pellet but not in the soluble
fraction. In both panels, anti catalase (a-catalase) demonstrates that the
cells have been efficiently lysed, as it is present in the soluble fraction, and
only minimal traces are in the pellet.
T.J. Dowse et al. / International Journal for Parasitology 35 (2005) 747–756752
the medium (arrowin Fig.4(C)). This protein was not present
in cells expressing DSpitz alone (Fig. 4(B)) or in cells
expressing an inactive RHBDL2 mutant (results not shown),
suggesting that the 32 kDa protein is cleaved intracellular
DSpitz. Consistent with this, an immunoreactive protein of
this size was present in the medium of cells expressing
Drosophila Rhomboid-1 or RHBDL2 (Fig. 4(C)). The
medium also contains smaller myc-reactive proteins that
proteolysis of the 32 kDa protein after its release from cells.
cleavage of DSpitz as indicated by the appearance of an
intracellular 32 kDa protein, although the effect of TgROM2
was small (Fig. 4(B)). In contrast to the action of Drosophila
Rhomboid-1 or RHBDL2, the cleaved product was barely
detectable in the medium of cells expressing either TgROM1
or TgROM2 suggesting that in HEK293T cells these
Toxoplasma rhomboids are located in the ER and not in the
the presence of TgROM5, the cleaved Spitz product was also
localised in the Golgi apparatus or later in the secretory
pathway of HEK293T cells. These results demonstrate that
TgROM1, TgROM2 and TgROM5 are active proteases that
can cleave the Spitz TM domain.
3.4. TgROM2 cleaves microneme protein transmembrane
In order to test the ability of Toxoplasma rhomboids to
cleave the TM domains of microneme proteins, we replaced
the TM and cytoplasmic domains of DSpitz with the
corresponding regions of MIC2 or MIC12 to generate the
chimeric proteins DSpi/MIC2 and DSpi/MIC12 respect-
ively. Drosophila Rhomboid-1 and human RHBDL2 were
used as positive controls in the assay, as they have
previously been demonstrated to cleave the TM domains
of MIC2 and MIC12 in synthetic constructs (Urban and
Freeman, 2003). In agreement with this, Drosophila
Rhomboid-1 and human RHBDL2 both cleaved the DSpi/
MIC2 and DSpi/MIC12 chimeric substrates efficiently,
resulting in both an intracellular and a secreted cleavage
product (Fig. 5). In contrast, co-expression of either
TgROM1 or TgROM5 with the chimeric substrates failed
to generate any detectable cleaved fragments in either the
cell lysates or in the medium, indicating that, at least in the
context of the chimeric constructs used, these proteases
cannot cleave the TM domains of MIC2 or MIC12. In the
presence of TgROM2, however, DSpi/MIC12 was cleaved
to generate an intracellular fragment, although release of the
cleaved product into the medium was not detected. In
addition, generation of an intracellular fragment from DSpi/
MIC2 was also detectable, although this cleavage was very
weak (Fig. 5). However for neither substrate was
the cleavage catalysed by TgROM2 as efficient as that by
Drosophila Rhomboid-1 or RHBDL2 (Fig. 5).
We report here the characterisation of members of the
rhomboid family of proteases in the parasite T. gondii.
Homologues of these genes are also present in other
apicomplexans including Plasmodium. All these proteins
described here contain multiple predicted membrane-
spanning domains and conserve residues that have been
implicated in the catalytic activity of rhomboid-like serine
proteases. The six genes identified in T. gondii code for
three smaller rhomboids, two larger rhomboids, and one
putative mitochondial rhomboid. Apart from TgROM6,
expressed sequence tags (ESTs) from different life stages
of T. gondii corresponding to each gene are present in the
ToxoDB database with the exception of TgROM3 for
which only ESTs from unsporulated oocysts can be found
(Li et al., 2003) (Table 2). Our failure to amplify the cDNA
of TgROM3 by RT-PCR from tachyzoite total RNA
preparations reinforces the view that this gene is
Fig. 4. TgROMs cleave the Spitz TM domain. HEK293T cells were
transfected with myc-tagged DSpitz and HA-tagged rhomboids as
described in Section 2.7. (A) Western blot of cell lysates using anti-HA
to detect rhomboid protein expression. (B) Western blot of cell lysates
using anti-myc to detect the DSpitz substrate and the cleaved DSpitz
product (arrowhead). (C) Western blot of samples of culture medium using
anti-myc to detect the release of the cleaved DSpitz product (arrow) from
cells. All size markers denote kDa.
T.J. Dowse et al. / International Journal for Parasitology 35 (2005) 747–756753
developmentally regulated. In P. berghei, preliminary
results based on RT-PCR analysis suggest that expression
of PbROM3, the apparent homologue of TgROM3, appears
tobe restricted to
(Dowse, unpublished data). As MPP1 is expressed in
tachyzoites, and possibly in all invasive stages of the
parasite including bradyzoites and sporozoites, the absence
of TgROM3 transcripts in tachyzoites excludes TgROM3
as a candidate for MPP1. As MPP1 activity has previously
been monitored constitutively at the parasite cell surface
(Opitz et al., 2002), the intracellular localisation of the
remaining TgROMs was considered to be key to
determining which if any are likely to correspond to this
protease. By this criterion TgROM4 and TgROM5 are
possible candidates, as they show immunocytochemical
localisation at the parasite cell surface.
In a mammalian cell culture assay system, we have
demonstrated that TgROM1, TgROM2 and TgROM5 are
active proteases that are capable of cleaving the TM domain
of the widely used rhomboid substrate, Drosophila Spitz.
However, of these, only TgROM2 shows any activity
towards chimeric proteins containing the TM and cyto-
plasmic domains of MIC2 and MIC12, although we cannot
dismiss the possibility that TgROM1 and/or TgROM5
would cleave the full-length MIC proteins. Therefore, in this
context, only TgROM2 appears to be a potential candidate
for MPP1. However definitive conclusion on the cleavage of
the gametocyte stage
DSpi/MIC constructs by TgROM2 would require additional
controls using an inactivated mutant form of the protease.
As the cellular localisation of TgROM2 is not that which is
expected for MPP1 the only obvious remaining rhomboid
candidate is TgROM4. Unfortunately, our attempts to
express TgROM4 in mammalian cell cultures have been
unsuccessful and, consequently, we cannot currently
determine whether it is able to cleave the TM domains of
the MIC proteins.
Rhomboid-like proteins have recently been implicated as
candidates for MPP1 based on several lines of evidence
(Urban and Freeman, 2003; Dowse and Soldati, 2004; Kim,
2004; Sibley, 2004): (i) TgMIC2 and TgMIC6 cleavage by
extracellular end, and rhomboids are the only known
intramembrane proteases to cleave near to the luminal face
of their substrates’ TM domains resulting in secretion/
release; (ii) rhomboids are the only known intramembrane
substrates for their activity; (iii) rhomboids are the only
known intramembrane proteases of the serine class and are
sensitive to DCI which is the only inhibitor previously
shown to block MMP1-dependent processing of TgMIC2;
and (iv) rhomboid-1 from Drosophila and human RHBDL2
have previously been shown to cleave MIC2, MIC6 and
MIC12 TM domains (Urban and Freeman, 2003). Our study
now adds three important lines of evidence to support
Fig. 5. TgROM2can useMIC2and MIC12TM domainsasa substrate.HEK293Tcells were transfectedwith myc-tagged DSpi/MIC2orDSpi/MIC12 and HA-
tagged rhomboids as described in the Section 2. (A) Western blot of cell lysates using anti-HA to detect rhomboid protein expression. (B) Western blot of cell
lysates using anti-myc to detect the cleavage products (arrowheads) of the chimeric DSpi/MIC2or DSpi/MIC12substrate proteins. (C) Western blot of samples
of culture medium using anti-myc to detect the release of the cleaved DSpi/MIC2 or DSpi/MIC12 products (arrows) from cells. All size markers denote kDa.
T.J. Dowse et al. / International Journal for Parasitology 35 (2005) 747–756754
this hypothesis: (i) rhomboid-like proteases are present in
T. gondii and conserved throughout the Apicomplexa, as
expected of MPP1; (ii) two rhomboid-like proteins,
as expected of MPP1; and (iii) three T. gondii rhomboids
have been demonstrated to be active, one of which can use
TgMIC2 and TgMIC12 TM domains as substrates. How-
ever, on the basis of these results, we are unable to identify
one of the Toxoplasma rhomboid proteases as MPP1. Of
significant interest is that an examination of the repertoire of
rhomboids proteins in the currently published apicomplexan
genomes also provides evidence for the identification of
MPP1. This survey revealed that the apicomplexan parasites
contain between three and eight genes coding for putative
rhomboid proteases although some of them are very exotic
and might not be active proteases. All Apicomplexa
including Cryptosporidium and Theileria species have at
least one stage of their life cycle which moves by gliding.
Consequently, they all possess at least one or more members
of the TRAP family of microneme proteins, many of which
harbor the conserved cleavage site IAYGG. Based on
phylogenetic analysis, only one rhomboid is commonly
present and conserved in all apicomplexan species and
corresponds to TgROM4, supporting it as a plausible
candidate for MPP1 (Dowse and Soldati, 2005).
The definitive assignment of MPP1 to a rhomboid-like
protease will require an analysis of the biological actions of
these rhomboids, which will be made possible by con-
ditional disruption of the ROM genes. The identification of
apicomplexan proteases that belong to the invasion
machinery will represent a critical step for the development
of novel therapeutic strategies to cure the diseases caused by
this important group of pathogens.
A recent study of the Toxoplasma rhomboids demon-
strated the cleavage of full length MIC2, and synthetic
constructs containing MIC6 and MIC12 TM domains, by
TgROM5. It was also shown that TgROM5 redistributes to
the posterior of the parasite upon microneme secretion. This
led to the conclusion that TgROM5 is responsible for MPP1
activity (Brossier et al., 2005). The discrepancy between
this study and the results reported here might be explained
by the use of different cell lines and constructs for the
This work was funded by a Wellcome Trust Studentship
to TD. JCP and KDB are supported by the Biotechnology
and Biological Sciences Research Council. We are grateful
to Bernardo Foth for his help with the phylogeny analysis
and critical reading of the manuscript.
Brecht, S., Carruthers, V.B., Ferguson, D.J., Giddings, O.K., Wang, G.,
Jakle, U., Harper, J.M., Sibley,L.D., Soldati, D., 2001. The Toxoplasma
micronemal protein MIC4 is an adhesin composed of six conserved
apple domains. J. Biol. Chem. 276, 4119–4127.
Brossier, F., Jewett, T.J., Lovett, J.L., Sibley, L.D., 2003. C-terminal
processing of the Toxoplasma protein MIC2 is essential for invasion
into host cells. J. Biol. Chem. 278, 6229–6234.
Brossier, F., Jewett, T.J.,Sibley,
spatially localized rhomboid protease cleaves cell surface adhesins
essential for invasion by Toxoplasma. Proc. Natl. Acad. Sci. U.S.A.
Carruthers, V.B., 2002. Host cell invasion by the opportunistic pathogen
Toxoplasma gondii. Acta Trop. 81, 111–122.
Carruthers, V.B., Sibley, L.D., 1997. Sequential protein secretion from
three distinct organelles of Toxoplasma gondii accompanies invasion of
human fibroblasts. Eur. J. Cell Biol. 73, 114–123.
Carruthers, V.B., Giddings, O.K., Sibley, L.D., 1999. Secretion of
micronemal proteins is associated with Toxoplasma invasion of host
cells. Cell Microbiol. 1, 225–235.
Carruthers, V.B., Sherman, G.D., Sibley, L.D., 2000. The Toxoplasma
adhesive protein MIC2 is proteolytically processed at multiple sites by
two parasite-derived proteases. J. Biol. Chem. 275, 14346–14353.
Dobrowolski, J.M., Sibley, L.D., 1996. Toxoplasma invasion of mamma-
lian cells is powered by the actin cytoskeleton of the parasite. Cell 84,
Dobrowolski, J.M., Carruthers, V.B., Sibley, L.D., 1997. Participation of
myosin in gliding motility and host cell invasion by Toxoplasma gondii.
Mol. Microbiol. 26, 163–173.
Donahue, C.G., Carruthers, V.B., Gilk, S.D., Ward, G.E., 2000. The
Toxoplasma homolog of Plasmodium apical membrane antigen-1
(AMA-1) is a microneme protein secreted in response to elevated
intracellular calcium levels. Mol. Biochem. Parasitol. 111, 15–30.
Donald, R.G., Carter, D., Ullman, B., Roos, D.S., 1996. Insertional tagging,
cloning, and expression of the Toxoplasma gondii hypoxanthine-
xanthine-guanine phosphoribosyltransferase gene. Use as a selectable
marker for stable transformation. J. Biol. Chem. 271, 14010–14019.
Dowse, T., Soldati, D., 2004. Host cell invasion by the apicomplexans: the
significance of microneme protein proteolysis. Curr. Opin. Microbiol.
Summary of Toxoplasma gondii rhomboids
Accession numberStage of expression (ESTs) Subcellular localisationCleavage of Spitz Cleavage of MICs
T.J. Dowse et al. / International Journal for Parasitology 35 (2005) 747–756 755
Dowse T.J., Soldati D., 2005. Rhomboid-like proteins in Apicomplexa: Download full-text
phylogeny and nomenclature. Trends Parasitol. In press.
Esser, K., Tursun, B., Ingenhoven, M., Michaelis, G., Pratje, E., 2002. A
novel two-step mechanism for removal of a mitochondrial signal
sequence involves the mAAA complex and the putative rhomboid
protease Pcp1. J. Mol. Biol. 323, 835–843.
Fourmaux, M.N., Achbarou, A., Mercereau-Puijalon, O., Biderre, C.,
Briche,I., Loyens, A., Odberg-Ferragut,
Dubremetz, J.F., 1996. The MIC1 microneme protein of Toxoplasma
gondii contains a duplicated receptor-like domain and binds to host cell
surface. Mol. Biochem. Parasitol. 83, 201–210.
Garcia-Reguet, N., Lebrun, M., Fourmaux, M.N., Mercereau-Puijalon, O.,
Mann, T., Beckers, C.J., Samyn, B., Van Beeumen, J., Bout, D.,
Dubremetz, J.F., 2000. The microneme protein MIC3 of Toxoplasma
gondii is a secretory adhesin that binds to both the surface of the host
cells and the surface of the parasite. Cell. Microbiol. 2, 353–364.
Herlan, M., Vogel, F., Bornhovd, C., Neupert, W., Reichert, A.S., 2003.
Processingof Mgm1by the rhomboid-type protease Pcp1is requiredfor
maintenance of mitochondrial morphology and of mitochondrial DNA.
J. Biol. Chem. 278, 27781–27788.
Herm-Gotz, A., Weiss, S., Stratmann, R., Fujita-Becker, S., Ruff, C.,
Meyhofer, E., Soldati, T., Manstein, D.J., Geeves, M.A., Soldati, D.,
2002. Toxoplasma gondii myosin A and its light chain: a fast, single-
headed, plus-end-directed motor. EMBO J. 21, 2149–2158.
Hettmann, C., Herm, A., Geiter, A., Frank, B., Schwarz, E., Soldati, T.,
Soldati, D., 2000. A dibasic motif in the tail of a class XIV
apicomplexan myosin is an essential determinant of plasma membrane
localization. Mol. Biol. Cell 11, 1385–1400.
Huynh, M.H., Rabenau, K.E., Harper, J.M., Beatty, W.L., Sibley, L.D.,
Carruthers, V.B., 2003. Rapid invasion of host cells by Toxoplasma
requires secretion of the MIC2-M2AP adhesive protein complex.
EMBO J. 22, 2082–2090.
Kim, K., 2004. Role of proteases in host cell invasion by Toxoplasma
gondii and other Apicomplexa. Acta Trop. 91, 69–81.
Pellegrini, L., 2003. The rhomboids: a nearly ubiquitous family of
intramembrane serine proteases that probably evolved by multiple
ancient horizontal gene transfers. Genome Biol. 4, R19.
Lee, J.R., Urban, S., Garvey, C.F., Freeman, M., 2001. Regulated
intracellular ligand transport and proteolysis control EGF signal
activation in Drosophila. Cell 107, 161–171.
Li, L., Brunk, B.P., Kissinger, J.C., Pape, D., Tang, K., Cole, R.H.,
Martin, J., Wylie, T., Dante, M., Fogarty, S.J., Howe, D.K.,
Liberator, P., Diaz, C., Anderson, J., White, M., Jerome, M.E.,
Johnson, E.A., Radke, J.A., Stoeckert Jr.., C.J., Waterston, R.H.,
Clifton, S.W., Roos, D.S., Sibley, L.D., 2003. Gene discovery in the
apicomplexa as revealed by EST sequencing and assembly of a
comparative gene database. Genome Res. 13, 443–454.
Lohi, O., Urban, S., Freeman, M., 2004. Diverse substrate recognition
mechanisms for rhomboids; thrombomodulin is cleaved by Mammalian
rhomboids. Curr. Biol. 14, 236–241.
McQuibban, G.A., Saurya, S., Freeman, M., 2003. Mitochondrial
membrane remodelling regulated by a conserved rhomboid protease.
Nature 423, 537–541.
Meissner, M., Schluter, D., Soldati, D., 2002. Role of Toxoplasma gondii
myosin A in powering parasite gliding and host cell invasion. Science
Opitz, C., Di Cristina, M., Reiss, M., Ruppert, T., Crisanti, A., Soldati, D.,
2002. Intramembrane cleavage of microneme proteins at the surface of
the apicomplexan parasite Toxoplasma gondii. EMBO J. 21,
Pascall, J.C., Brown, K.D., 2004. Intramembrane cleavage of ephrinB3 by
the human rhomboid family protease, RHBDL2. Biochem. Biophys.
Res. Commun. 317, 244–252.
Pascall, J.C., Luck, J.E., Brown, K.D., 2002. Expression in mammalian cell
cultures reveals interdependent, but distinct, functions for Star and
Rhomboid proteins in the processing of the Drosophila transforming-
growth-factor-alpha homologue Spitz. Biochem. J. 363, 347–352.
Pelletier, L., Stern, C.A., Pypaert, M., Sheff, D., Ngo, H.M., Roper, N.,
He, C.Y., Hu, K., Toomre, D., Coppens, I., Roos, D.S., Joiner, K.A.,
Warren, G., 2002. Golgi biogenesis in Toxoplasma gondii. Nature 418,
Reiss, M., Viebig, N., Brecht, S., Fourmaux, M.N., Soete, M., Di
Cristina, M., Dubremetz, J.F., Soldati, D., 2001. Identification and
characterization of an escorter for two secretory adhesins in
Toxoplasma gondii. J. Cell Biol. 152, 563–578.
Schweitzer, R., Shaharabany, M., Seger, R., Shilo, B.Z., 1995. Secreted
Spitz triggers the DER signaling pathway and is a limiting component
in embryonic ventral ectoderm determination. Genes Dev. 9, 1518–
Sibley, L.D., 2004. Intracellular parasite invasion strategies. Science 304,
Soldati, D., Boothroyd, J.C., 1993. Transient transfection and expression in
the obligate intracellular parasite Toxoplasma gondii. Science 260,
Soldati, D., Lassen, A., Dubremetz, J.F., Boothroyd, J.C., 1998. Processing
of Toxoplasma ROP1 protein in nascent rhoptries. Mol. Biochem.
Parasitol. 96, 37–48.
Soldati, D., Dubremetz, J.F., Lebrun, M., 2001. Microneme proteins:
structural and functional requirements to promote adhesion and
invasion by the apicomplexan parasite Toxoplasma gondii. Int.
J. Parasitol. 31, 1293–1302.
Sultan, A.A., Thathy, V., Frevert, U., Robson, K.J., Crisanti, A.,
Nussenzweig, V., Nussenzweig, R.S., Menard, R., 1997. TRAP is
necessary for gliding motility and infectivity of plasmodium spor-
ozoites. Cell 90, 511–522.
Templeton, T.J., Kaslow, D.C., Fidock, D.A., 2000. Developmental
arrest of the human malaria parasite Plasmodium falciparum
within the mosquito midgut via CTRP gene disruption. Mol. Microbiol.
Urban, S., Freeman, M., 2003. Substrate specificity of rhomboid
intramembrane proteases is governed by helix-breaking residues in
the substrate transmembrane domain. Mol. Cell 11, 1425–1434.
Urban, S., Lee, J.R., Freeman, M., 2001. Drosophila rhomboid-1 defines a
family of putative intramembrane serine proteases. Cell 107, 173–182.
Urban, S., Lee, J.R., Freeman, M., 2002a. A family of Rhomboid
intramembrane proteases activates all Drosophila membrane-tethered
EGF ligands. EMBO J. 21, 4277–4286.
Urban, S., Schlieper, D., Freeman, M., 2002b. Conservation of intramem-
brane proteolytic activity and substrate specificity in prokaryotic and
eukaryotic rhomboids. Curr. Biol. 12, 1507–1512.
Wu, Y., Wang, X., Liu,X., Wang, Y., 2003. Data-mining approachesreveal
hidden families of proteases in the genome of malaria parasite. Genome
Res. 13, 601–616.
Yuda, M., Sakaida, H., Chinzei, Y., 1999. Targeted disruption of the
Plasmodium berghei CTRP gene reveals its essential role in malaria
infection of the vector mosquito. J. Exp. Med. 190, 1711–1716.
Zhou, X.W., Blackman, M.J., Howell, S.A., Carruthers, V.B., 2004.
Proteomic analysis of cleavage events reveals a dynamic two-step
mechanism for proteolysis of a key parasite adhesive complex. Mol.
Cell Proteomics 3, 565–576.
T.J. Dowse et al. / International Journal for Parasitology 35 (2005) 747–756756