Register or Login To Download This Patent As A PDF
|United States Patent Application
;   et al.
May 16, 2002
EF-Tu protein encoded on the plastid DNA of the malaria parasite and
protein synthesis inhibitors effective as anti-malarial compounds
The plastid DNA of the malaria parasite Plasmodium falciparum has been
sequenced and found to contain a gene encoding an EF-Tu protein.
Inhibitors of the protein are effective as anti-malarial compounds and
the protein can be used to screen for such inhibitors. Furthermore, the
23S ribosomal RNA encoded on the malaria parasite plastid DNA is a target
for anti-malarial compounds and the antibiotic thiostrepton acts as an
anti-malarial by binding to the RNA.
Clough, Barbara; (London, GB)
; Preiser, Peter; (London, GB)
; Wilson, Robert John Macleod; (London, GB)
Nixon & Vanderhye P.C.
1100 N. Glebe Road, 8th Floor
MEDICAL RESEARCH COUNCIL
May 1, 2001|
|Current U.S. Class:
|Class at Publication:
1. The EF-Tu protein encoded on the plastid DNA of the malaria parasite
2. The protein according to claim 1 which has the sequence labelled
"eftu_pf" in FIG. 2A (SEQ ID NO:2).
3. The protein according to claim 1 or 2 in purified form.
4. A DNA molecule encoding the protein of claim 1 or 2.
5. The DNA molecule according to claim 4 which comprises the sequence
shown in FIG. 2B (SEQ ID NO:1).
6. The DNA molecule according to claim 4 in purified form.
7. The DNA molecule according to claim 4 which is a cloning or expression
8. A host cell transformed with the vector of claim 7.
9. A method of producing the EF-Tu protein encoded on the plastid DNA of
the malaria parasite Plasmodium falciparum, which method comprises (i)
culturing a host cell containing a DNA molecule encoding the protein
under conditions such that the protein is expressed; and (ii) recovering
the protein from the culture.
10. A method of identifying an anti-malarial compound, which method
comprises (i) contacting a compound with the EF-Tu protein encoded on the
plastid DNA of the malaria parasite Plasmodium falciparum; and (ii)
determining whether the compound binds to or inhibits the protein, any
such binding or inhibition being indicative that the compound is an
11. A compound identified by the method of claim 10.
12. A method of preventing or treating infection of a patient with the
malaria parasite Plasmodium falciparum, which method comprises
administering to the patient a compound which inhibits the EF-Tu protein
encoded on the plastid DNA of said malaria parasite.
13. The method according to claim 12 wherein the compound is an
14. The method according to claim 13 wherein the compound is a member of
the kirromycin series of antibiotics.
15. The method according to claim 14 wherein the compound is selected from
the group consisting of kirromycin (mocimycin), aurodox
(1-methylmocimycin), efrotomycin, enacyloxin IIa and GE2270.
16. An antibody specific for the EF-Tu protein encoded on the plastid DNA
of the malaria parasite Plasmodium falciparum.
17. A method of identifying an anti-malarial compound, which method
comprises (i) contacting a test compound with the 23S ribosomal RNA
encoded on the plastid DNA of the malaria parasite Plasmodium falciparum
(pf 23S rRNA.sub.pl) or with a fragment of said RNA containing the GTPase
domain; and (ii) determining whether the compound binds to said RNA or
said fragment, any such binding being indicative that the compound is an
18. A method according to claim 17 which comprises (i) incubating the Pf
23S rRNA.sub.pl or the fragment thereof with the test compound and a
reference compound known to bind to the rRNA or the fragment; (iia)
determining the amount of reference compound that is bound to the rRNA or
the fragment; and (iib) comparing the amount of reference compound bound
to the rRNA or the fragment with the amount that is bound in the absence
of the test compound; wherein any reduction in the binding of the
reference compound in the presence of the test compound compared to the
binding in the absence of the test compound is indicative that the test
compound is competing for binding to the rRNA and that the test compound
could be an anti-malarial.
19. The method according to claim 18 wherein the reference compound is
20. The method according to claim 17 wherein said RNA or said fragment
contains an A residue at the position corresponding to position 1067 in
the 23S rRNA of Escherichia coli.
21. The method according to claim 20 wherein said fragment comprises the
pf 23S rRNA.sub.pl sequence corresponding to the sequence from about
position 1051 to about position 1108 of the 23S rRNA of Escherichia coli.
22. A compound identified by the method of claim 17.
FIELD OF THE INVENTION
 This invention relates to a new protein encoded in the plastid DNA
of the malaria parasite Plasmodium falciparum, to DNA encoding the
protein, to methods of producing the protein, to methods of screening for
anti-malarial compounds, to compounds identified by such screening
methods and to methods of preventing or treating growth of the malaria
BACKGROUND TO THE INVENTION
 The malarial 35 kb circular DNA molecule central to this invention
corresponds to a minor species of DNA distinct from nuclear DNA
discovered in the 1960s (Gutteridge et al. 1971). In the mid-80s the
first study on its purification and molecular analysis was published
(Williamson et al. 1985). Its similarity was noted to a circular DNA in
the related organism Toxoplasma gondii--a well known opportunistic
pathogen in AIDS cases.
 It is important to stress that the malaria parasite and related
apicomplexans are unusual amongst non-photosynthetic organisms in that
they possess two forms of organellar DNA, typically a property of plants.
One form of organellar DNA has been identified as mitochondrial DNA
(mtDNA), whereas the other, the 35 kb circle, we have proposed is the
remnant of a plastid DNA (plDNA), a provenance hitherto unsuspected for
these organisms (Wilson et al. 1991, 94). This plDNA was probably
acquired by an ancient progenitor of the phylum and may be of algal
origin (Williamson et al. 1994). The precise location of these organellar
DNAs in the cell shows they are in separate compartments (Kohler et al
 Thus, there are potentially two organellar protein synthesising
systems of independent prokaryotic origin within the malaria organism
that could be susceptible to inhibition with antibiotics. Although the
malarial mitochondrion is the best characterised of the organelles, its
genetic content is highly idiosyncratic, contributing only incomplete
fragments of two rRNA genes to the machinery required for protein
synthesis. The circular DNA from the putative plastid, on the other hand,
is much more conventional, producing transcripts of four complete rRNA
genes, some twenty tRNA genes, subunits of a typical plastid RNA
polymerase, and a number of ribosomal protein genes organised in modified
SUMMARY OF THE INVENTION
 In sequencing the malarial plastid DNA, we found that it contains a
gene encoding a new EF-Tu protein homologous to the EF-Tu proteins known
in prokaryotes. Thus, the invention provides an EF-Tu protein encoded on
the plastid DNA of the malaria parasite Plasmodium falciparum. The
invention also provides DNA encoding the protein.
 The prokaryotic EF-Tu proteins are known to be important in
controlling the elongation cycle in protein synthesis, and it is known
that inhibition of the proteins by various compounds has an antibiotic
effect. In view of the sequence similarity between the prokaryotic EF-Tu
proteins and our newly-identified malarial plastid EF-Tu protein, we
proposed the theory that the antibiotic compounds which inhibit the
prokaryotic proteins may also inhibit the malarial protein and therefore
be useful as anti-malarials. We tested such antibiotics (e.g. kirromycin
and aurodox) for their anti-malarial effect and found our theory was
correct; the antibiotics were found to be effective anti-malarials both
in vitro and in vivo. Thus, the invention provides a method of preventing
or treating infection of a patient with the malaria parasite Plasmodium
falciparum, which method comprises administering to the patient a
compound which inhibits the EF-Tu protein encoded on the plastid DNA of
said malaria parasite.
 The knowledge provided by the invention of the EF-Tu protein in the
malaria plastid and the fact that its inhibitors are effective
anti-malarials allows the protein to be used in screening for new
anti-malarial compounds. Accordingly, the invention includes a method of
identifying an anti-malarial compound, which method comprises
 (i) contacting a test compound with the EF-Tu protein encoded on
the plastid DNA of the malaria parasite Plasmodium falciparum; and
 (ii) determining whether the compound binds to or inhibits the
protein, any such binding or inhibition being indicative that the
compound is an anti-malarial.
 We also investigated the ability of antibiotics which bind to other
components of the prokaryotic protein synthesis machinery to act as
anti-malarial compounds. As a result of these investigations, it was
found that thiostrepton, which is known to bind to the GTPase domain of
the 23S ribosomal RNA of E. coli, is also able to bind to GTPase domain
of the 23S rRNA encoded on the plastid of the malaria parasite Plasmodium
falciparum (Pf 23S rRNA.sub.pl). Accordingly, the invention provides a
method of identifying an anti-malarial compound, which method comprises
 (i) contacting the compound with the 23S ribosomal RNA encoded on
the plastid DNA of the malaria parasite Plasmodium falciparum (Pf 23S
rRNA.sub.pl) or with a fragment of said RNA containing the GTPase domain;
 (ii) determining whether the compound binds to said RNA or said
fragment, any such binding being indicative that the compound is an
BRIEF DESCRIPTION OF THE DRAWINGS
 FIG. 1 is a schematic illustration of the elongation cycle that
occurs during protein synthesis and shows the points in the cycle at
which various inhibitors operate.
 FIG. 2A shows the amino acid sequence of the EF-Tu protein
according to the invention from the plastid of the malaria parasite
Plasmodium falciparum (pf). The sequence is aligned with sequences of
EF-Tu proteins from other organisms, namely E. coli ("ecoli"), Anacystis
nidulans ("anani"), Cyanophora paradoxa ("cypha") and Cryptomonas phi
 FIG. 2B shows the nucleotide sequence of the tufA gene that encodes
the EF-Tu protein according to the invention.
 FIG. 3A shows a Southern blot of endonuclease-restricted malarial
genomic DNA hybridised with a PftufA-specific PCR product as probe. A
single band for the 35 kb plastid was obtained for each restriction
 FIG. 3B shows cross-hybridisation between endonuclease-restricted
malarial genomic DNA and the yeast tufM gene, indicating the possible
presence of a malarial version of tufM.
 FIG. 4 shows the results of an experiment in which an antisense RNA
probe (about 230 nts) made by in vitro transcription, corresponding to a
portion of the tufA gene encoding domains I and II of the predicted
EF-Tu.sub.pl protein, was used in an RNase protection assay to
demonstrate the presence of tufA transcripts in total RNA extracted from
erythrocytic parasites during the course of a single growth cycle (0-40
 FIG. 5 shows dose-response curves for the effects of fusidic acid,
mocimycin (kirromycin), thiostrepton and GE 2270 on incorporation of
.sup.3H-hypoxanthine and .sup.14C-isoleucine into erythrocytic stages of
P. falciparum grown in cultures over a 36 hour period. The Figure also
shows a dose-response curve for the effect of mocimycin on myeloma cells
 FIG. 6 shows the effects of aurodox and mocimycin on the growth of
P. chabaudi in mice. The solid line is for aurodox, the dotted line is
for mocimycin and the dashed line is for no drug controls. The arrows on
the x-axis indicate days on which three 0.1 ml 100 mM ip injections were
 FIG. 7 shows the sequence of the GTPase region of the plastid 23S
rRNAs of Plasmodium falciparum (Pf) and Toxoplasma gondii (Toxo) (numbers
based on E.coli), showing substitution sites (circled) affecting the
binding of thiostrepton. The alternative nucleotides in Pf cytosolic 28S
rRNA and Pf mitochondrial 23S rRNA are indicated.
 FIG. 8 shows thiostrepton titrations (means of duplicates,
bar=range) in a filter binding assay with transcripts of the GTPase
region of LSU rRNA.
 A) Short 23S transcripts of P.falciparum (Pf) wild type rRNA.sub.pl
(open circle A1067) and mutated forms (open square A1067U and filled
triangle A1067G), as well as Pf 28S rRNA transcripts (filled square), are
compared with an optimized E.coli control transcript (filled circle). For
convenience, nucleotide numbers correspond to E.coli.
 B) T. gondii wild type rRNA.sub.pl transcript (filled triangle) and
mutated transcript (open triangle) compared with control transcripts from
P.falciparum rRNA.sub.pl (open circle) and E.coli (filled circle).
 FIGS. 9 and 10 show the structures of various antibiotics usable in
 FIG. 11 shows slot blots of RNA fractioned on sucrose gradients.
Pretreatment with anisomycin blocked the puromycin-induced shift of the
hybridization signal for P.falciparum cytosolic 23S ribosomes but not the
plastid 16S ribosomes.
 A-C. Blots hybridized with a probe for the cytosolic large subunit
(23S) rRNA. Anisomycin blocked the puromycin-induced shift.
 D-F. The same blots hybridized with a probe for the plastid-encoded
small subunit (16S) rRNA. Anisomycin did not block the puromycin-induced
 FIG. 12 is a slot blot showing the puromycin-induced shift of the
hybridization signal for plastid mRNA specifying EF-Tu.
 FIG. 13 contains immunoblots showing that binding of antibiotics
modifies migration of EF-Tu.GDP in native polyacrylamide gels. Two
segments of the same gel show A) heterologously expressed Pf EF-Tu.sub.pl
protein detected with a malaria peptide-specific antibody and B) E.coli
EF-Tu detected with a specific antibody (Breidenbach et al 1990). Lanes
without antibiotics (1 and 5), lanes with 100 .mu.M antibiotic: GE2270A
(2 and 6), enacyloxin lla (3 and 7), kirromycin (4 and 8). Arrows
indicate uncomplexed EF-Tu.
DETAILED DESCRIPTION OF THE INVENTION
 The EF-Tu Protein
 The function of the EF-Tu protein is in the elongation cycle of
protein synthesis. The cycle is illustrated in FIG. 1. EF-Tu reacts with
GTP and AA-tRNA to form an EF-Tu/AA-tRNA/GTP complex. After binding to
EF-Tu, the AA-tRNA component is transferred to the ribosomal A site with
the release of free EF-Tu-GDP complex and phosphate. The GDP is released
from EF-Tu and the EF-Tu is then ready for another cycle.
 EF-Tu is an exceedingly abundant protein in E. coli, present in
approximately as many copies as there are tRNA molecules. It can bind
every tRNA except for fMet-tRNA.
 The malarial plastid EF-Tu has much sequence identify with known
EF-Tu proteins from other organisms (see FIG. 2A). We have made a
3D-model structure for the malarial plastid EF-Tu protein based on the
crystal structures available for bacterial equivalents (E. coli and T.
thermophilus). This model showed that the bacterial and malarial proteins
are very similar indeed, strongly implying that the malarial plastid
EF-Tu is functional.
 The sequence of the malarial plastid EF-Tu protein of the invention
may be that labelled "eftu_pf" in FIG. 2A, but variations in this
sequence are possible. The protein may, for example, have a sequence
identity with the sequence in FIG. 2A of 80% or more, 90% or more, 95% or
more or 99% or more.
 The sequence of FIG. 2A may be modified by substitution,
deletion,-extension or insertion. A substitution, deletion or insertion
may involve one or more amino acids, typically from 1 to 5, from 1 to 10
or from 1 to 20 amino acids.
 Such modified sequences must retain the functions of the EF-Tu
protein necessary for participation in the elongation cycle of protein
synthesis. In general, the physicochemical nature of the sequence of FIG.
2A should be preserved; the amino acids of a modified sequence should
generally be of a similar charge, size and hydrophobicity/hydrophilicity
as those in the sequence of FIG. 2A. Candidate substitutions are those in
which an amino acid from one of the following groups is replaced by a
different amino acid from the same group:
 H, R and K
 I, L, V and M
 A, G, S and T
 D and E.
 The EF-Tu protein of the invention may be provided in purified
form. The protein may also be provided in pure form and in isolated form.
The protein may, for example, be provided in a preparation in which it
constitutes 10% or more, 40% or more, 80% or more, 90% or more, 95% or
more or 99% or more of the total protein in the preparation by weight.
 The protein will usually be obtained by expression of recombinant
DNA containing the protein, but may also be obtained by biochemical
purification of the protein from the malaria parasite.
 DNA Encoding the Malarial Plastid EF-Tu Protein
 The DNA encoding the EF-Tu protein may have the sequence shown in
FIG. 2B, but variations in this sequence are possible. The DNA molecule
may, for example, have a sequence identity with the sequence shown in
FIG. 2B of 70% or more, 80% or more, 90% or more, 95% or more or 99% or
 A recombinant DNA molecule encoding the EF-Tu protein of the
invention may be obtained using well-known and conventional recombinant
DNA techniques, such as those described in Sambrook et al (1989)
Molecular Cloning: A Laboratory Manual, 2nd Ed., Cold Spring Harbor
Laboratory, Cold Spring Harbor, N.Y.
 Such DNA molecules may be obtained by making a library of
replicable expression vectors. The library may be created by cloning all
the DNA or, more preferably, the plastid DNA of the malaria parasite into
a parent vector. The library may be screened for members containing the
desired nucleic acid sequence, e.g. by means of a DNA probe or antibody.
 The term "replicable expression vector" is used herein to mean a
vector which contains the appropriate origin of replication sequence for
directing replication of the vector. The vector may also contain the
appropriate sequences for expression of the EF-Tu protein. The sequences
for expression of the protein will generally include a transcription
promotor and a translation initiator operably linked to the coding
sequence. The term "operably linked" refers to a linkage in which the
promotor and initiator are connected in such a way to the coding sequence
as to permit expression of the protein. A vector may, for example, be a
plasmid, virus or phage vector. A vector may contain one or more
selectable markers, for example an ampicillin resistance gene in the case
of a bacterial vector or a neomycin resistance gene in the case of a
mammalian vector. A foreign gene sequence encoding the EF-Tu protein
inserted into a vector may be transcribed in vitro or the vector may be
used to transform a host cell.
 According to one embodiment of the invention, there is provided a
host cell transformed with a vector encoding the EF-Tu protein. A vector
and host cell will be chosen so as to be compatible with each other, and
may be prokaryotic or eukaryotic. A prokaryotic host may, for example, be
E.coli in which case the vector may, for example, be a bacterial plasmid
or a phage vector. A eukaryotic host may, for example, be a yeast (e.g.
S.cerevisiae), a chinese hamster ovary (CHO) cell or an insect cell (e.g.
 The invention includes a method of producing the EF-Tu protein
encoded on the plastid DNA of the malaria parasite Plasmodium falciparum,
which method comprises
 (i) culturing a host cell containing a DNA molecule encoding the
protein under conditions such that the protein is expressed; and
 (ii) recovering the protein from the culture.
 Antibodies to the Malarial Plastid EF-Tu Protein
 The invention includes an antibody specific for the EF-Tu protein
of the invention. The antibody is preferably monoclonal, but may also be
polyclonal. The antibody may be labelled. Examples of suitable antibody
labels include radiolabels, biotin (which may be detected by avidin or
streptavidin conjugated to peroxidase), alkaline phosphatase and
fluorescent labels (e.g. fluorescein and rhodamine). The term "antibody"
is used herein to include both complete antibody molecules and fragments
thereof. Preferred fragments contain at least one antigen binding site,
such as Fab and F(ab').sub.2 fragments. Humanised antibodies and
fragments thereof are also included within the term "antibody".
 The antibody may be produced by raising antibody in a host animal
against an EF-Tu protein according to the invention or an antigenic
epitope thereof (hereinafter "the immunogen"). Methods of producing
monoclonal and polyclonal antibodies are well-known. A method for
producing a polyclonal antibody comprises immunising a suitable host
animal, for example an experimental animal, with the immunogen and
isolating immunoglobulins from the serum. The animal may therefore be
inoculated with the immunogen, blood subsequently removed from the animal
and the IgG fraction purified. A method for producing a monoclonal
antibody comprises immortalising cells which produce the desired
antibody. Hybridoma cells may be produced by fusing spleen cells from an
inoculated experimental animal with tumour cells (Kohler and Milstein,
Nature 256, 495-497, 1975).
 An immortalized cell producing the desired antibody may be selected
by a conventional procedure. The hybridomas may be grown in culture or
injected intraperitoneally for formation of ascites fluid or into the
blood stream of an allogenic host or immunocompromised host. Human
antibody may be prepared by in vitro immunisation of human lymphocytes,
followed by transformation of the lymphocytes with Epstein-Barr virus.
 For the production of both monoclonal and polyclonal antibodies,
the experimental animal is suitably a goat, rabbit, rat or mouse. If
desired, the immunogen may be administered as a conjugate in which the
immunogen is coupled, for example via a side chain of one of the amino
acid residues, to a suitable carrier. The carrier molecule is typically a
physiologically acceptable carrier. The antibody obtained may be isolated
and, if desired, purified.
 Assays for Identifying Anti-malarial Compounds that Inhibit the
Malarial Plastid EF-Tu Protein
 As mentioned above, the knowledge provided by the invention of the
EF-Tu protein in the malaria plastid and the fact that its inhibitors are
effective anti-malarials allows the protein to be used in screening for
new anti-malarial compounds.
 Various different assay systems may be used to carry out the
screening, but all the assays have in common that the EF-Tu protein of
the invention is contacted with test compounds and the ability of each
test compound to bind to or inhibit the protein is determined. Any such
binding or inhibition is indicative that the compound could be useful as
an anti-malarial drug.
 The screening assays will generally require one or more controls.
It will generally be desirable to include a positive control in the form
of a compound known to bind to or inhibit the EF-Tu protein, so as to
ensure that the assay system is responding properly. Examples of suitable
positive controls include kirromycin (mocimycin) and aurodox
(1-methylmocimycin), which we have shown through our experiments to be
effective anti-malarials and which are known to inhibit prokaryotic
EF-Tu. It will also generally be desirable to include a negative control
in the form of a sample containing no test compound, so as to obtain a
measurement of the background signal in the assay. If a test compound
gives a signal in the assay above that of the background, this is
indicative that the compound has given a positive result and could be an
 One convenient type of assay system is a "band shift" system. This
involves determining whether a test comopund advances or retards the
EF-Tu protein of the invention on gel electrophoresis relative to the
EF-Tu protein in the absence of test compound. The mobility of GDP
complexed EF-Tu is decreased with GE2270A but increased with enacyloxin
IIa or kirromycin.
 Another convenient type of assay system is a competitive binding
system. Such a system may comprise
 (i) incubating the EF-Tu protein of the invention with a test
compound and a labelled reference compound that is known to bind the
protein (e.g. kirromycin or aurodox);
 (ii) determining the amount of the labelled reference compound that
is bound to the protein; and
 (iii) comparing the amount of bound labelled reference compound
determined in step (ii) with the amount of said compound that binds to
the protein in the absence of the test compound;
 wherein any reduction in the binding of the labelled reference
compound in the presence of the test compound compared to the binding in
the absence of the test compound shows that the test compound is
competing with the reference compound for binding to the protein and
indicates that the test compound could be an anti-malarial.
 The amount of the labelled reference compound bound to the protein
may be measured directly or indirectly. A direct measurement may be
carried out by removing assay mixture containing the unbound labelled
reference compound and measuring the amount of label that is in the
protein fraction. Alternatively, the amount of labelled reference
compound bound to the protein could be determined indirectly by measuring
the amount of label remaining in the assay solution after removal of the
protein fraction, which will be inversely related to the amount that has
bound to the protein.
 In a competitive binding assay system, the EF-Tu protein may be
immobilised on a solid support or may be in solution. The use of
immobilised protein has the advantage that, after the binding reaction is
complete, the protein/labelled reference compound complex may be
separated from the labelled reference compound that remains in solution
by simply removing the solution away from the solid support. If, on the
other hand, the protein is not immobilised during the assay but rather is
in solution, then it will generally be necessary to devise a means for
separating the protein/labelled reference compound complex from the
uncomplexed reference compound before measuring the amount of label. Such
separation could be achieved, for example, by precipitating the protein
using an antibody to the protein or by using a non-specific protein
 Suitable labels for use in the assay systems according the
invention are well-known in the art and include the labels set out above
that may be attached the antibodies of the invention.
 Use of Compounds that Inhibit the Malarial Plastid EF-Tu Protein as
 Compounds that inhibit the malarial plastid EF-Tu protein may be
used as anti-malarial compounds. Such compounds may be identified using
the screening assays described above.
 We have already identified some such compounds. e.g. the
antibiotics kirromycin (mocimycin), aurodox (1-methylmocimcyin),
Efrotomycin (a glycoside of kirromycin), Enacyloxin IIa and GE2270.
Efrotomycin has previously been shown to have the desirable properties of
rapid oral absorption and prolonged plasma half-life.
 The antibiotics in the kirromycin series may be represented by the
general formula: 1
 wherein R.sup.1 is hydrogen or a C.sub.1-C.sub.4 alkyl group (e.g.
methyl); R.sup.2 is hydrogen, C.sub.1-C.sub.4 alkyl or a sugar group
(e.g. a disaccharide); and R.sup.3 is hydrogen, OH or C.sub.1-C.sub.4
 Preferred antibiotics for use in the invention are as follows: 2
 The compounds may be used in either the treatment of an existing
infection by the malaria parasite or in the prevention of such an
infection from occurring in the first place. The dosage regimen will
ultimately be at the discretion of the physician, who will take into
account factors such as the nature of the compound, the severity of any
disease and the weight and age of the patient. However, suitable routes
of administration may include the oral route, the rectal route, the
intramuscular route and the intravenous route. The oral route is
preferred because this is generally the most convenient route for a
patient to take regular doses of the compound without the assistance of a
physician. A typical dose would be from 1 to 1000 mg and such a dose may,
for example, be taken from 1 to 3 times daily.
 In order to be administered to a patient, the compound will be
provided in the form of a pharmaceutical composition containing the
active compound and a pharmaceutically acceptable carrier or diluent.
Typical oral dosage compositions include tablets, capsules, liquid
solutions and liquid suspensions.
 Compounds that Bind to the 23S rRNA Encoded on the Malaria Parasite
 We have found that the 23S rRNA encoded on the plastid of the
malaria parasite Plasmodium falciparum (Pf23S rRNA.sub.pl) is a target
for compounds with anti-malarial activity. More specifically, we have
found that the mechanism of action of the antibiotic thiostrepton, which
was known to have anti-malarial activity, is through binding to the
GTPase domain of the 23S rRNA of the malaria plastid.
 This information allows the design of assays for screening for
further anti-malarial compounds whose mechanism of action operates
through the 23S rRNA. These assays involve contacting each of the test
compounds with the 23S rRNA or a fragment thereof containing the GTPase
binding domain, and measuring any binding of the test compounds to the
rRNA or fragment. Any such binding is of course indicative that the
compound could be an anti-malarial.
 We have already developed one assay for detecting binding to the
23S rRNA. We made a short transcript from DNA encoding the 23S rRNA of
the malaria plastid corresponding to the GTPase domain (about nucleotide
1051 to about nucleotide 1108) and found that the transcript bound
 The binding was specific. It depended to a large extent on the
presence of an A residue at the position corresponding to E. coli
position 1067. A transcript corresponding to the GTPase domain of the E.
coli 23S rRNA (which contained the A at position 1067) was also shown to
bind thiostrepton strongly. Mutation of A1067 to either U or G in the
malaria plastid transcript dramatically reduced binding. A transcript
corresponding to the GTPase domain of the cytosolic malaria 28S rRNA also
bound thiostrepton poorly.
 A screening assay for further anti-malarial compounds can be based
on a competitive binding assay in which the ability of each test compound
to compete with thiostrepton for binding to the Pf23S rRNA.sub.pl is
measured. Such an assay comprises
 (i) incubating the Pf 23S rRNA.sub.pl or a fragment thereof
containing the GTPase domain with the test compound and thiostrepton as a
reference compound (or another reference compound known to bind to the
rRNA or the fragment);
 (ii) determining the amount of thiostrepton (or other reference
compound) that is bound to the rRNA or the fragment; and
 (iii) comparing the amount of thiostrepton (or other reference
compound) bound to the rRNA or the fragment with the amount that is bound
in the absence of the test compound;
 wherein any reduction in the binding of the thiostrepton (or other
reference compound) in the presence of the test compound compared to the
binding in the absence of the test compound is indicative that the test
compound is competing for binding to the rRNA and that the test compound
could be an anti-malarial.
 In a screening assay based on the invention for further
anti-malaria compounds, it would be necessary to use appropriate
controls. A good positive control would be to use a compound known to
compete with thiostrepton (or with the other reference compound) to
ensure that the assay is working properly; a positive result for the
known competitor in the assay would indicate that the assay had worked
correctly. It would also generally be desirable to use a negative control
comprising, for example, a sample in which no thiostrepton or test
compound is present; this would enable the background signal in the assay
to be determined and any signal above the background would indicate
binding to the 23S rRNA.
 The following experiments serve to illustrate the invention.
 Experimental Section
 Materials and Methods
 Polysome preparation and puromycin shift--P.falciparum was grown in
blood cultures (Trager et al 1976) and ribosomes prepared as described
(Sherman et al 1975). Parasitized erythrocytes were lysed for 1 hr on ice
in a buffer containing 0.14% Nonidet P-40 (Trade Name), 25 mM KCl, 10 mM
MgCl.sub.2, 380 mM-sucrose, 6.5 mM .beta.-mercaptoethanol and 50 mM Tris
HCl, pH 7.6. The lysate was centrifuged.times.3 at 10,000 g for 10 min at
4.degree. C. to remove genomic DNA and other cell debris before further
centrifugation at 105,000 g for 1 hr in an SW40 rotor (Trade Name,
Beckman) at 4.degree. C. The resulting pellet was resuspended in 25 mM
KCl, 5 mM MgCl.sub.2 and 50 mM Tris HCl (pH 7.6) and homogenized by hand
(.times.50 strokes) with a glass dounce homogenizer (Wheatstone, USA).
The suspension was centrifuged at 10,000 g for 10 min at 4.degree. C. and
the crude pellet discarded before further centrifugation for 2 hr at
105,000 g in an SW55 rotor (Beckman) at 4.degree. C. The pellet was
resuspended in 10 mM Tris HCl, 10 mM MgCl.sub.2, 100 mM KCl and
homogenized again to give a suspension of ribosomes.
 Polysomes were fractioned on sucrose gradients (20-50% w/v)
prepared in 0.3M KCl, 3 mM MgCl.sub.2 and 1 mM dithiothreitol (DDT) with
0.02M Tris HCl (pH 7.6)--referred to as "high salt" buffer;
centrifugation was at 30,000 g (Beckman Sw40 rotor) for 21 hr at
 In an experiment with RNase (Cox 1969), total polysomes were
incubated with a range of concentrations of RNase (1-13 ng ml.sup.-1
ribosomes, Boehringer) prior to centrifugation for 30 min at 26.degree.
C. In other experiments, polysomes were dissociated to monosomes and
subunits by the incorporation of puromycin; here the total ribosome
preparation was incubated for 20 min at 37.degree. C. with 2 mM puromycin
in the "high salt" buffer to which was added 2 mM GTP, 10 .mu.lml.sup.-1
RNAsin (39 U.mu.l.sup.-1, Promega) and 1 mM DTT. In some experiments,
ribosomes were incubated with both anisomycin (Sigma) and puromycin.
Anisomycin was added at 3 mM for 10 min at 37.degree. C. followed by
incubation with puromycin as above (Cundliffe et al 1974). After ribosome
fractionation on the sucrose gradients, RNA was extracted with
phenol/chloroform/isoamyl alcohol (Chomczynski it al 1987), precipitated
in ethanol and blotted on to nylon membranes (Gene Screen, Trade Name)
using a slot-blot apparatus (Scot-Labs). Hybridization was carried out
with .sup.32P-labelled DNA prepared from either cloned fragments of the
35 kb plDNA of P falciparum, PCR products amplified from it,
oligonucleotides based on its sequence (Wilson et al 1996), or with PCR
products based on the sequence of Pf28S cytosolic rRNA (McCutchen et al
1988). Hybridization signals were quantitated using a Molecular Dynamics
 Antibiotics--Samples of Mocimycin (kirromycin), Aurodox
(N-methylated kirromycin), and Efrotomycin (a glycoside of kirromycin)
were used. Aurodox was dissolved in RPMI-Albumax medium (Grande et al
1977), kirromycin was dissolved in RPMI made alkaline by addition of 1M
NaOH, and efrotomycin was dissolved in ethanol before dilution in culture
medium. Enacyloxin IIA was dissolved in 1% NaHCO.sub.3 prior to dilution
in RPMI-Albumax medium. The antibiotic GE2270A was dissolved in 100% DMSO
before dilution in culture medium. Fusidic acid (Sigma) was dissolved
directly in culture medium, and thiostrepton (Sigma) in 100% DMSO before
dilution in culture medium. A hemisuccinate form of thiostrepton was
prepared as a potassium salt, according to Bodanszky et al 1965.
Incorporation of radiotracers by P.falciparum growing in blood cultures
in the presence and absence of drugs was carried out as described (Strath
et al 1993).
 EF-Tu model--Pf EF Tu.sub.pl was modelled by homology with the
known 3D structures determined by X-ray crystallography of EF-Tu.GTP
(Berchtold et al 1993) and EF-Tu.GDP (Polekhina et al 1996) both from
Thermus aquaticus. Modelling was carried out with the WHAT-IF program
package (Trade Name, Vriend 1990), as described in Tews et al 1996.
Alignments had to be adjusted manually because of small gaps and
insertions. An iterative procedure of the automated model-building
algorithm checked and corrected the alignments until no errors were
detectable. Threee insertions in the Pf EF-Tu.sub.pl sequence had to be
deleted: Leu 190, Pro263 and Leu359-Val363. The final alignment with the
T. aquaticus structure had single residue gaps in the Pf sequence between
Leu41 and Ser42 as well as residues Asn209 and lle210. Co-ordinates for
the C (alpha) backbone were copied from the known structure for
overlapping segments and the atoms for the amino acids Gly, Ala and Pro
were placed directly in their calculated positions. All remaining
residues were assigned to Ala before the order in which side changes had
to be placed was calculated by the algorithm implemented by the program.
Atoms were subsequently placed using a position-dependent amino acid
rotamer library. The model was refined geometrically and re-numbered
according to the P.falciparum sequence.
 Heterologous expression--The malarial plastid tufA gene was
amplified by PCR, cloned into the TA vector (Trade Name, Invitrogen) and
its sequence determined (Wilson et al 1996). Re-cloning into the
expression vector pGEX (Trade Name, Pharmacia) was carried out with a PCR
product generated using 5' and 3' primers providing custom-made
restriction sites. Transfectants in E.coli (strains DH5 alpha, Sure.
JM109) were found mostly to carry deletions within the tufA sequence, but
one clone in JM109 contained the complete insert (sequenced on a single
strand). This was expressed as a fusion protein of the expected size by
induction of mid-log phase cultures with 50 .mu.M isopropyl-.beta.-D-thio-
galactoside (IPTG) at 37.degree. C. or 27.degree. C. The insoluble fusion
protein was solubilized in 5M guanidinium HCl and refolded by dilution
(Lin et al 1991).
 Antibody to an epitope of Pf EF-Tu--A rabbit polyclonal antibody
was prepared against a 13-mer synthetic peptide IQKNKDYELIKSN from domain
I of Pf EF-Tu coupled to polylysine beads (Severn Biotech. Ltd). In
Western blots (ECL protocol, Amersham), this antibody did not cross-react
with EF-Tu from E.coli, nor did an antibody to E.coli EF-Tu react with
the expressed malarial protein.
 Drug binding--Thiostrepton binding to short rRNA transcripts
generated in vitro was assayed according to Ryan et al 1991, as modified
by Clough et al 1997. A band shift method in native 12% polyacrylamide
gel (Cetin et al 1996) was used to demonstrate complex formation between
a resolubilized fraction of the expressed Pf EF-Tu.sub.pl and various
antibiotics. Before electrophoresis and immunoblotting, samples were
incubated on ice for 15 mins in 50 mM imidazole acetate (pH 7.6), 10 mM
NH.sub.4Cl, 10 mM MgCl.sub.2, 1 mM DDT and 100 .mu.M GDP, in a final
volume of 20 .mu.l.
 Evidence for Plastid Protein Synthesis
 Ribosomes from erythrocytic parasites were fractionated by
centrifugation on linear gradients (20-50% sucrose) and RNA was extracted
from fractions collected over the length of the gradients. Slot blots of
the RNA were hybridized with .sup.32P-labelled DNA probes prepared from
either cloned fragments of Pf plDNA, PCR products based on its sequence,
or kinased oligonuceotides. As shown in FIG. 11 C&F, hybridization with
probes for the large.sub.(cytosolic) or small.sub.(plastid) subunit rRNAs
gave signals extending to the bottom of the gradient, indicative of rRNA
incorporated in polysomes. Supportive evidence was obtained by limited
digestion of the total ribosome preparation with RNase (13 ng RNase/mg
ribosomes for 30 min at 26.degree. C.) before fractionation--this causes
dissociation of the polysomes (Cox 1969) and shifted the hybridization
signal up the gradient (data not shown). More specific evidence for a
subset of polysomes belonging to the plastid compartment was obtained by
incubating total ribosomes with 2 mM puromycin in the presence of GTP,
0.3M KCl and 1 mM DDT prior to density gradient fractionation: puromycin
acts as an analogue of the 3' terminal adenosine of aminoacylated tRNAs
and is incorporated into nascent peptide chains, terminating translation
and dissociating polysomes (Gale et al 1981). Incubation with puromycin
caused a shift of both the cytosolic and plastid rRNA hybridization
signals up the gradient (FIG. 11, B&E). The specificity of the
puromycin-shift was confirmed by pre-treating Pf ribosomes with the
antibiotic anisomycin which binds only to eukaryotic ribosomes and
prevents puromycin incorporation (Gale et al 1981). As shown in FIG. 11A,
anisomycin blocked the puromycin-induced shift of the hybridization
signal for Pf 28S cytosolic rRNA, whereas hybridization of the same blot
with a probe for Pf 16S rRNA.sub.pl showed the puromycin-shift of the
plastid subset of polysomes was not blocked (FIG. 11D).
 Similar results were obtained with a probe for an mRNA specified by
the plDNA. FIG. 12 shows the puromycin-induced shift of the hybridization
signal for mRNA specifying EF-Tu.sub.pl.
 To quantitate the relative proportions of the hybridization signals
generated by different species of RNA, slot blots were hybridized with
.sup.32P-labelled oligonucleotides, known amounts of DNA being used as
appropriate standards. The 28S cytosolic rRNA was estimated to be 80-fold
more plentiful than 16S rRNApl and 2000-fold more plentiful than the mRNA
specifying EF-Tu.sub.pl (data not shown). These results and the
puromycin-shifts are consistent with the presence of actively translating
plastid ribosomes in blood cultures of malaria parasites.
 tufA Sequence
 From a combination of cloned DNA fragments and PCR products
amplified from the 35 kb circular DNA of P.falciparum, we derived a 1.23
kb nt sequence whose predicted peptide (calculated M. Wt. 46,633) is
homologous to the elongation factor EF-Tu (FIG. 2A). The malarial gene
lies 45 nts downstream from two ribosomal protein-encoding genes, rps12
and rps7. In this respect, the organization resembles the str operon on
the plDNAs of the flagellate protists Euglena gracilis (Montadon & Stutz,
1984; Hallick et al. 1993) and Astasia longa (Seimelster et al. 1990), as
well as the non-chlorophyte alga Cryptomonas (Douglas, 1991), and the
cyanelle of Cryptomonas paradoxa (Kraus et al. 1990), the intervening fus
gene encoding EF-G in the str operon of bacteria such as E.coli (Zengal
and Lindahl BBA 1050, 317 (1990)) presumably having been transferred to
the nucleus. The short intergenic region upstream of the PfpltufA gene
does not contain an open reading frame or putative leader sequence. At
the nt level, the malarial pltufA gene is extremely rich in adenine and
thymine (A+T) residues (79%) compared to related sequences in the
database, a feature with important consequences for computations intended
to establish the gene's phylogenetic relationships.
 At the predicted peptide level, the malarial sequence is very
divergent from other recorded EF-Tu's (only 45% amino acid identity with
E.coli and 51% identity with Anacystis nidulans). Nonetheless, several
highly conserved regions are evident, including the four segments of
domain I involved in GTP binding. In E.coli, the first three of these
segments carry the consensus elements G18HVDHGK24; D80CPG83; and
N135KCD138. In the malarial sequence there is only one substitution
C136E. Most of the residues defining the GDP binding pocket also are
conserved (in E.coli G23, N135, K136, D138, S173, L175), the only
substitution in the malarial sequence being M139L. In a less well
conserved region (amino acids 180-190, topologically close to the GTP
binding domain). The malarial sequence has an insertion typical of
plastid versions of EF-Tu that is not found in the E.coli gene, and is
only partially present in the mitochondrial equivalent (tufM) of
Saccharomyces cerevisiae (Nagata et al. 1983). Despite the gene's high
A+T content, the predicted malarial EF-Tu peptide is one of the most
highly conserved proteins encoded by the 35 kb circle; however, it is
potentially more basic (calculated pI=8.43) than the versions present in
bacteria or the yeast mitochondrion (Piechulla & Kuntzel, 1983).
 In view of the unknown functional status of the 35 kb circular DNA,
it was of interest to compare the predicted malarial EF-Tu.sub.pl peptide
with the unusual chloroplast form in the Charophycean alga Colochaete
orbicularis, as it has been suggested that the latter may no longer be
functional, there being multiple tufA-like sequences in the nucleus
(Baldauf et al. 1990). Baldauf and colleagues pointed out that the
C.orbicularis EF-Tu.sub.pl amino acid sequence differs in twenty two
sites that otherwise are conserved in all but four of 27 other EF-Tu
sequences. Despite the malarial gene's extreme A+T content, the predicted
EF-Tu peptide has only 6 conservative amino acid substitutions in the
same 22 residues (Table 1). This suggests that the functional domains
encoded by the tufA gene on the 35 kb circle have been maintained under
AMINO ACID SEQUENCE COMPARISON OF
(MODIFIED FROM BALDAUF ET AL. 1990)
C cp eub all Pf
21-22 FS VD VD VD VD
GIT GIT GIT GIT
87 N D D D D
90 N K K K K
V V V V
153 N E E E E
210 L I I I I
227 R D D D D
233 S G G G G
236 L T T T T
241 T R R R K
248 N K K K N
272 K E E E E
286 D N N N N
301 K R
R R R
372 E D D D D
393 V A A A S
401 I V V V I
405 I V V V I
* = Amino acids numbered as in FIG.
C = Coleochaete orbicularis chloroplast tuf A (Baldauf et
cp = cyanobacteria and chloroplast consensus
eub = eubacteria, cyanobacteria and chloroplast consensus
eubacteria, eukaryotes and archaebacteria consensus
Plasmodium falciparum 35 kb circule tufA
 When hybridized with a PftufA-specific PCR product under stringent
conditions, Southern blots of endonuclease-restricted malarial genomic
DNA gave a single band of the size predicted (FIG. 3A). At low stringency
no other bands were revealed that might have corresponded to the
nucleus-encoded mitochondrial gene tufM (Nagata et al, 1983; Wells et al.
1994). The likely presence of a malarial equivalent was indicated,
however, by cross-hydridization at low stringency with a PCR product
based on the yeast tufM gene (FIG. 3B).
 An antisense RNA probe (.about.230 nts) made by in vitro
transcription, corresponding to a portion of the malarial tufA gene
encoding domains I and II of the predicted EF-Tu.sub.pl protein, was used
in an RNase protection assay to demonstrate the presence of tufA
transcripts in total RNA extracted from erythrocytic parasites (FIG. 4).
 Modelling of the Three-dimensional Structure of P.falciparum
 Despite the highly divergent amino acid composition of Pf
EF-Tu.sub.pl a computer model showed conservation of secondary structure
motifs in all three domains of the hypothetical protein. The model is
reliable, with good confidence in the overall folding and also in the
detail of the secondary structure compared with T.aquaticus. Only minor
differences were found in the length of some structural motifs: in domain
I there are small length differences in 5 of the nine alpha helices and
in 3 out of six .beta. strands. The changes are more pronounced in domain
II where the first two .beta.-strands seem to be continuous in Pf, and an
extra .beta.-strand is formed by residues Gly245-Leu249. In domain III,
differences from T. aquaticus are again minor with a slightly different
positioning of two .beta.-strands.
 The model for Pf EF-Tu.sub.pl.GTP showed that the GTP-binding site
is conserved as well as the whole lining towards the GTP-binding pocket.
There is also conservation on the interface between the domains. These
interfaces are exposed when EF-Tu. GTP converts to the effective
EF-Tu.GDP form. In this form of the protein, conserved residues in the
cleft between domains I and III correspond to the site which other
studies have shown kirromycin binds.
 The effects of three classes of compounds on intraerythrocytic
parasites of P.falciparum in vitro, as well as on P.chabaudi in vivo, are
considered below. In assessing the significance of these results it
should be noted that in prokaryotes, kirromycin, whose binding site is at
the interface of domains I and III of EF-Tu.GTP (Mesters et al. 1994),
binds to the ternary complex of tRNA and EF-Tu.GTP preventing the
conformational change required for release from the ribosome upon GTP
hydrolysis, whereas the drug has a different effect on eukaryotic cells.
In the latter, at 100 .mu.M, it blocks RNA synthesis without affecting
DNA or protein synthesis (Schmid et al. 1978).
 Kirromycin: Kirromycin-resistant forms of bacterial EF-Tu are
modified at one of seven amino acids along the opposing interfaces of
domains I and III (Mesters et al 1994 and Abdulkarim et al 1994) and Pf
EF-Tu.sub.pl has a substitution at one of these sites (A375S-E.coli
numbers) that could potentially confer resistance to kirromycin. To test
this possibility, kirromycin (Mocimycin), its methylated derivative
Aurodox or its disaccharide derivative Efrotomycin were incubated with
erythrocytic stages of P.falciparum grown in cultures over a 36 hr
period. The incorporation of both .sup.3H-hypoxanthine and
.sup.14C-isoleucine was inhibited in a dose-dependent fashion, maximum
inhibition being achieved at 100 .mu.M kirromycin (FIG. 5). Similar
levels of inhibition were obtained for all three compounds. In
synchronized cultures, inhibitory effects on ring stage parasites were
observed as early as five hours after exposure to Aurodox and were
maximal after 10 hrs exposure. Once maximal depression of incorporation
had been reached at any particular dose of drug, residual incorporation
continued at a uniform rate thereafter. Vital staining with rhodamine
123, a fluorescent dye that concentrates within the mitochondrion (Divo
et al. 1985) confirmed the parasiticidal effect, loss of specific mt
staining being evident within 2-3 hours (data not shown). Treatment of
parasites with 1 mM Aurodox for 1.5 cell cycles, followed by removal of
the antibiotic by washing and follow-up for 2 weeks in vitro, indicated
the effect was parasite death rather than stasis. Blood cultures of
P.falciparum were about 10 times more sensitive to the antibiotic than a
gram-positive bacterium (Corynebacterium spp) used in parallel bioassays.
 On the basis of these findings, preliminary studies were carried
out on mice infected with P.chabaudi. In the first experiment, mice were
infected and inoculated on the same day intraperitoneally with 0.1 ml of
100 mM Aurodox, a dose calculated to mimic the maximal inhibitory effect
observed in vitro. The Aurodox-treated animals showed a lag in
development of the infection compared with untreated controls indicating
partial inactivation of the infectious inoculum (FIG. 6). In a single
experiment. Mocimycin w as found to be less effective in vivo than
 Enacyloxin IIa: Enacyloxin IIa (ExIIa) is a linear antibiotic
representing a new family of polyenic antibiotics (Watanabe 1992) that
bind to EF-Tu and block transfer to the nascent peptide chain of
aminoacylated-tRNA bound at the A site (Cetin 1996). The profiles for
inhibition of radiotracer incorporation in blood cultures of
P.falciparium incubated with Ex IIa were similar to those with Mocimycin.
 GE2270: This is a thiopeptide antibiotic in the same family as
thiostrepton. It binds to a different site on Ef-Tu than kirromycin and
locks the protein into a different conformation (Landini 1996). This
antibiotic was more inhibitory in blood cultures than either kirromycin
or thiostrepton (FIG. 5).
 Fusidic acid: Fusidic acid, presently in clinical use as a narrow
spectrum antibiotic, was assessed as a potential antimalarial by
titration with P.falciparum in vitro, as described above. Maximum
inhibition of radiotracer incorporation was achieved at a concentration
of 200 .mu.M (FIG. 5). Preliminary experiments with fusidic acid in mice
infected with P.chabaudi found little effect on parasitaemias, even at
toxic dose levels of the drug.
 Thiostrepton: Nucleotide (nt) sequences are available for the
GTPase domain of the 28S rRNA specified by the nucleus (Rogers et al.
1996), as well as the 23S rRNAs specified by the mt and pl large subunit
rRNA genes of the human malaria pathogen Plasmodium falciparum (Pf)
(Feagin, 1992). These data indicate that the high affinity binding site
for the thiazolyl peptide antibiotic thiostrepton, A.sub.1067 in E.coli
(Thompson et al 1991, Ryan et al 1991 and Rosendahl et al 1994), is
conserved in the GTPase domain encoded by the plastid DNA, but modified
to a G in both nuclear and mitochondrial genomes (FIG. 7).
 We have tested thiostrepton to ascertain whether it inhibits Pf and
found reproducible inhibition of uptake of .sup.3H-hypoxanthine and
.sup.14C-isoleucine in blood cultures (50% inhibition at 3-5 .mu.M
thiostrepton). Onset of inhibition of protein synthesis by thiostrepton
was more rapid (5 hrs) than by tetracycline (8 hrs). Specificity was
demonstrated by the lack of effect of viomycin (data not shown), an
unrelated antibiotic that also can inhibit translocation (Kutay et al
 Having established thiostrepton's activity, we asked "does the
antibiotic bind preferentially to the nuclear, mitochondrial or plastid
forms of Pf 28/23 S rRNA?". Evidence that the highest affinity
interaction of thiostrepton is with 23S rRNA.sub.pl was obtained from an
in vitro binding assay (Ryan et al 1991). Short transcripts of wild type
(wt) RNA corresponding to the GTPase domain of Pf 23S rRNA.sub.pl (nts
987-1078) were transcribed in vitro from a PCR product that included a T7
promoter sequence in one of the primers. Mutated malarial rRNA sequences
(E.coli numbers A1067U and A1067G) were obtained by PCR methodology and
transcribed in the same way. Both wild type and modified transcript
sequences were verified prior to thiostrepton binding assays. A positive
control transcript was used based on the 23S rRNA sequence of E.coli with
a mutation (U1061A) that increases stability and binding. FIG. 8A shows
that the mutation Pf.sub.pl (E.coli number A1067U) markedly reduced
thiostrepton binding (.about.14% of wt). An intermediate level of binding
(.about.35% of wt) was obtained with the mutation Pf.sub.pl (E.coli
number A1067G). Thiostrepton binding to a transcript corresponding to the
GTPase domain (nt 1334-1427) of the cytosolic Pf 28S rRNA was .about.10%
of that for Pf 23S rRNA.sub.pl. These data show that the nts crucial for
thiostrepton binding to Pf23 S rRNA are as in E.coli, and that the
plastid form has the highest binding affinity.
 In the same way, we tested a transcript corresponding to the GTPase
domain of the 23S rRNA.sub.pl of Toxoplasma gondii (Tg), a related
apicomplexan that is an important opportunistic pathogen in patients with
AIDS. In this case, the wild type sequence has a substitution at a
different site (E.coli number A1077U)--see FIG. 7, that inhibits binding
by thiostrepton in E. coli (Ryan et al 1991). This was found also to be
the case with a transcript derived from a PCR product covering the GTPase
domain of Tg.sub.pl 23S rRNA (nt 926-1024) (FIG. 8B). Corrective mutation
of the Tg.sub.pl transcript (E. coli number U1077A) conferred a
significant increase (.times.5) in thiostrepton binding (FIG. 8B).
 These thiostrepton binding studies constitute the first direct
evidence that components of the malarial plastid organelle could be
preferentially targeted by drugs. The results complement earlier studies
(Pfefferkon et al 1994 and Beckers et al 1995) which inferred that
toxoplasma's 23S rRNA.sub.pl might be the target of the macrolide
antiobiotic, clindamycin, acting at a different effector site.
 Drugs Bind to Heterologously Expressed EF-Tu
 The material tufA gene in pGEX was expressed as an insoluble fusion
protein in E.coli JM 109. The protein was detected either with antibodies
to the GST tag or with antibodies to a specific peptide sequence in
domain I (IQKNKDYELIKSN) not found on E.coli EF-Tu. Washed inclusion
bodies were dissolved and refolded by dilution (Lin et al 1991). This
yielded a small amount of refolded protein that migrated in native
acrylamide gels as a spontaneously cleaved product and we used this to
show that the expressed protein forms complexes with kirromycin and other
drugs that bind to different sites on EF-Tu. As shown in FIG. 13, the
mobility (Mr) of the expressed malarial protein was advanced or retarded
in these complexes in the same characteristic way described for E.coli
EF-Tu (Cetin et al 1996): the M.sub.r of the GDP form of the complex
decreased with GE2270A, but increased with enacyloxin IIa or kirromycin.
These results show that the heterologously expressed Pf EF-Tu.sub.pl can
adopt a native conformation and bind the classical antibiotic inhibitors.
 Williamson et al., 1985. Mol. Biochem. Parasitol. 14, 199-209.
 Williamson et al., 1994. Mol. Gen. Genet. 243, 249-252.
 Gutteridge et al., 1971. Parasitology 62, 209-219.
 Wilson et al., 1994. Infect. Agents Dis. 3, 29-37.
 Wilson et al., 1991. Parasitol. Today 7, 134-136.
 Wilson et al., 1992. Curr. Genet. 21, 405-408.
 Vaidya et al., 1993. Mol. Cell. Biol. 13, 7349-7357.
 Creasey et al., 1994. Mol. Biochem. Parasitol. 65, 95-98.
 Kohler and Milstein, 1975. Nature 256, 495-497.
 Kohler et al 1997 Science 275 1485-1489.
 Montadon and Stutz, 1984 Nucl. Acids Res. 12, 2851-2859
 Hallick et al., 1993. Nucl. Acids Res. 21, 3537-3544.
 Siemeister et al., 1990. Mol. Gen. Genet. 220, 425-432.
 Douglas, 1991. Nature 350, 148-151
 Kraus et al., 1990. Plant Mol. Biol. 15, 561-573.
 Zenge and Lindahl, 1990. Biochim.Biophys.Acta 1050, 317-322
 Nagata et al., 1983. Proc. Natl. Acad. Sci. USA 80, 6192-6196.
 Piechulla and Kuntzel, 1983 Eur. J. Biochem. 132, 235-240.
 Baldauf et al., 1990. Proc. Natl. Acad. Sci. USA 87, 5317-5321
 Wells et al., 1995. FEBS Letters 358, 119-125
 Mesters et al., 1994. EMBO J. 13, 4877-4885.
 Schmid et al., 1978.FEBS Lett 96, 189-191.
 Divo et al., 1985. Antimicrob. Agents Chemother. 27, 21-27.
 Rogers et al., 1996. RNA 2, 134-145
 Feagin, 1992. Mol. Biochem. Parasitol. 52, 145-148
 Thompson et al., 1991. Biochimie 73, 1131-1135.
 Ryan et al., 1991. J. Mol. Biol. 221, 1257-1268.
 Rosendahl et al., 1994. Nucl. Acids Res. 22, 357-363.
 Kutay et al., 1990. Biochim. Biophys. Acta 1050, 193-196.
 Pfefferkorn et al., 1994. Antimicrob. Agents Chemother. 38, 31-37.
 Beckers et al., 1995. J. Clin. Invest. 95, 367-376.
 Wilson et al (1996) J. Mol. Biol. 261, 155-172
 Trager, et al (1976) Science 193, 673-675.
 Sherman et al (1975) J. Protozool. 22, 568-572.
 Cox, (1969) Biochem. J. 114, 753-767.
 Cundliffe, et al (1974) Proc. Natl. Acad. Sci. USA 71, 30-34.
 Chomczynski et al (1987) Anal. Biochem. 162, 156-159
 McCutchen, et al (1988) Mol. Biochem. Parasitol. 28, 63-68.
 Grande et al (1977) Parasitology 115, 81-89.
 Bodanszky et al (1965) in US Patent Office
 Strath et al (1993) Trans. R. Soc. Trop. Med. Hyg. 87, 211-216
 Berchtold et al (1993) Nature 365, 126-132.
 Polekhina et al (1996) Structure 4, 1141-1151.
 Vriend (1990) J. Mol. Graph. 8, 52-56.
 Tews et al (1996) Nature Struct. Biol. 3, 638-648.
 Lin et al (1991) Biotechniques 11, 748-752.
 Ryan et al (1991) J. Mol. Biol. 221, 1257-1268.
 Clough et al (1997) FEBS Letters 406, 123-125.
 Cetin et al (1996) The EMBO J. 15, 2604-2611.
 Gale et al (1981) The Molecular Basis of Antibiotic Action. John
Wiley & Sons Ltd., London
 Abdulkarim et al (1994) FEBS Letters 352, 118-122
 Watanabe et al (1992) J. Antibiot. 45, 470-475.
 Landini et al (1996) Biochemistry 35, 15288-15294.
 Breidenbach et al (1990) Biochim. Biophys Acta 1048, 209-216.
 (1) General Information
 (i) APPLICANT:
 (A) NAME: CLOUGH et al
 (B) STREET: National Institute for Medical Research, Mill Hill
 (C) CITY: London
 (E) COUNTRY: United Kingdom
 (F) POSTAL CODE (ZIP) NW7 1AA
 (ii) TITLE OF INVENTION: AN EF-TU PROTEIN ENCODED ON THE PLASTID
DNA OF THE MALARIA PARASITE AND PROTEIN SYNTHESIS INHIBITORS EFFECTIVE AS
 (iii) NUMBER OF SEQUENCES: 2
 (iv) COMPUTER READABLE FORM:
 (A) MEDIUM TYPE: Floppy disk
 (B) COMPUTER; IBM PC compatible
 (C) OPERATING SYSTEM: PC-DOS/MS-DOS
 (D) SOFTWARE: PatentIn Release #1.0, Version #1.30 (EPO)
 (v) CURRENT APPLICATION DATA: APPLICATION NUMBER: US NA
 (2) Information for SEQ ID NO: 1
 (i) SEQUENCE CHARACTERISTICS:
* * * * *