Interaction between two domains of the P. yoelii MSP-1 protein detected using the yeast two-hybrid system

Molecular & Biochemical Parasitology 117 (2001) 27 – 35 www.parasitology-online.com.

Interaction between two domains of the P. yoelii MSP-1 protein detected using the yeast two-hybrid system Thomas M. Daly, Carole A. Long 1, Lawrence W. Bergman * Department of Microbiology and Immunology, MCP Hahnemann Uni6ersity, 2900 Queen Lane, Philadelphia, PA 19129, USA Received 2 March 2001; received in revised form 21 May 2001; accepted 24 May 2001

Abstract Several model systems of plasmodia have demonstrated the potential of the merozoite surface protein, MSP-1, to induce protective immunity. However, little is known about the function of this protein or its interaction with other surface molecules that may also serve as immunological targets. To identify potentially significant inter- and intra-molecular interactions involving MSP-1, we have utilized the yeast two-hybrid system. A cDNA activation domain library was constructed from the erythrocytic stages of the murine malarial parasite Plasmodium yoelii yoelii 17XL. A 795 bp region of Py17XL MSP-1 (bait), homologous to the Plasmodium falciparum MSP133 fragment, was inserted into a Gal4p DNA binding domain vector and used to screen the activation domain library (target). Several randomly selected clones that demonstrated bait-target interaction were found to express overlapping regions of Py17XL MSP-1. Deletion constructs further localized the peptide fragments retaining interaction indicating that a region within the MSP-138 fragment interacts with the MSP-1 bait domain. Subsequent studies confirmed this interaction, as both peptides were co-precipitated from cell lysate by a peptide tag-specific antibody. It was observed that the interaction of these two fragments significantly increased the half-life of the MSP-138 within yeast cells. The specific interaction described here demonstrates the potential of this approach to elucidate additional inter- or intra-molecular interactions of Py17XL MSP1 and other malarial proteins. © 2001 Elsevier Science B.V. All rights reserved. Keywords: Malaria; Merozoite surface protein; MSP-1; Yeast two-hybrid

1. Introduction The dramatic resurgence of malaria in recent decades has renewed interest in the possibility of a vaccine or perhaps several vaccines to prevent infection or to limit pathology due to malaria [1,2]. In addition, the emergence and spread of chloroquine resistant P. falciparum has made the search for new anti-malarial drugs more compelling. A significant number of plasmodial proteins derived from various stages of the life cycle have been partially characterized. However, with all the proteins identified, there is still limited information on their biological role in the parasite life cycle. It is clear that new approaches to vaccine development and * Corresponding author. Tel.: + 1-215-991-8376; fax: +1-215-8482271. E-mail address: [email protected] (L.W. Bergman). 1 Present address: Malaria Vaccine Development Unit, NIAID, NIH, 12441 Parklawn Drive Rockville, MD 20852.

chemotherapy will require a much more sophisticated understanding of the biology of these important organisms. Of the numerous blood-stage malaria antigens that have been identified in both infected humans and in animal models of this disease, the clearest rationale for a blood-stage vaccine candidate exists for the merozoite surface protein-1 (MSP-1) [3–7]. MSP-1 appears to be uniformly distributed around the surface of the merozoite and there is evidence to suggest that it is complexed with other parasite proteins [[8 –10], Farley and Long, unpublished observations]. In general, MSP-1 is synthesized as a large (approximately 200 kDa) precursor during schizogony and is processed after merozoite egress from the erythrocyte into a series of proteolytic fragments. Studies by Holder and colleagues reported that the P. falciparum MSP-1 molecule undergoes a two-step proteolytic cleavage at the time of schizont release and erythrocyte invasion [11]. A similar process has been seen with P. knowlesi and P. chabaudi [12,13].

0166-6851/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 6 - 6 8 5 1 ( 0 1 ) 0 0 3 2 9 - 2

28

T.M. Daly et al. / Molecular & Biochemical Parasitology 117 (2001) 27–35

Initially, the intact molecule is cleaved into four fragments of 83, 30, 38, and 42 kDa, although the data indicate that the fragments remain bound together noncovalently on the merozoite surface. Subsequently the C-terminal 42 kDa fragment is further processed into two polypeptides of 33 and 19 kDa [14], the latter including a glycolipid moiety that anchors the molecule to the merozoite surface. The C-terminal 19 kDa proteolytic fragment is retained on the merozoite surface and can be found in newly invaded erythrocytes [15]. The remainder of the complex appears to be released from the merozoite surface during invasion. Since MSP-1 has been one of the better-studied plasmodial antigens, it is frequently used as a target antigen for serological studies as well as a candidate protein for a blood-stage vaccine. Most efforts to use MSP-1 as a vaccine candidate have focused on the conserved C-terminal region of the molecule (designated block 17 in the Tanabe nomenclature) [16]. Earlier results of the Long laboratory first demonstrated that this region of the protein could induce significant, antibody-mediated protective immunity in mice against a normally lethal challenge infection with P. yoelii [17,18]. Recently, as part of attempts to understand the mechanism of action of antibodies to the C-terminus of MSP-1, Holder and Blackman reported that two mabs to P. falciparum MSP-1, which could inhibit invasion in vitro, also inhibited proteolysis of the 42 kDa fragment [19]. They also found that some antibodies to an N-terminal region of MSP-1 were capable of blocking the function of the enzyme inhibitory antibodies, suggesting an interaction between these two regions of the molecule [20]. Despite the studies available on the MSP-1 protein, we have little definitive information on its role or roles in the functioning of Plasmodium species. Its presence on all species of plasmodia examined to date, its regions of conservation, and its complex and specific processing all support its biological importance to the parasite. We have initiated studies utilizing the yeast two-hybrid system to examine the protein– protein interactions involving P. yoelii MSP-1. In this report, we demonstrated that two domains of MSP-1 interact and postulate that this interaction may help explain several observations earlier reported in the literature.

2. Materials and methods

2.1. Experimental animals and parasites About 6 –8-week-old male BALB/cByJ mice were purchased from Jackson Laboratories and housed in our AALAC-approved animal facility. The 17XL lethal variant of Plasmodium yoelii yoelii (Py17XL) was maintained as cryopreserved stabilate. Blood stage infections were initiated by ip injection of parasitized erythro-

cytes. Parasitemias were monitored by microscopic examination of stained blood films.

2.2. Yeast strains and media Yeast strains PJ69-4a (MATa; trp1 -901 ; leu2 -3,112 ; ura3 -52 ; his3 -200 ; gal4 D; gal80 D; GAL2 -ADE2 ; LYS2 ::GAL1 -HIS3 ; met2 ::GAL7 -lacZ) and YJN192 (MATa; leu2 ; ura3 ; trp1 ; prb1 -1122 ; pep4 -3 ; prc1 -407 ), kindly provided by Philip James and Joseph Nickels, respectively, were used in this study. Cells were grown at 30 °C in YPD (1% yeast extract, 2% bactopeptone, 2% glucose), YCA (0.67% yeast nitrogen base, 0.5% casamino acids) or SD (0.67% yeast nitrogen base with either 2% glucose, 2% galactose or 1.5% raffinose) supplemented with appropriate amino acids and bases as required for selection.

2.3. Construction of Py17XL cDNA library BALB/cByJ mice were infected with a uniform inoculum of Py17XL parasites. When parasitemias reached 20–25%, blood was collected from the infected animals and leukocytes were removed by passage over columns of microcrystalline cellulose [21]. Total RNA was then isolated using the guanidinium isothiocyanate method and mRNA isolated using oligo(dT)-cellulose chromatography as described by the manufacturer (Stratagene). cDNA synthesis, unidirectional ligation into the HybriZAP™ vector as a fusion with the Gal4 activation domain and packaging were done according to the manufacturer (Stratagene). The initial library contained approximately 1×106 independent clones. The resulting library was amplified, converted to a phagemid library by in vivo mass excision (vector designated pAD-GAL4) and reamplified as described by the manufacturer (Stratagene). Phagemid library DNA was isolated using the alkaline lysis method (Mega Kit, Qiagen).

2.4. Plasmid constructs To construct the bait domain vector containing the 33 kDa region of P. yoelii MSP-1, a 795 bp region of the Py17XL gene (nucleotides 4369–5163 [22]) was amplified by PCR from Py17XL genomic DNA, isolated as earlier described [23]. Since the actual sites of proteolytic processing within the P. yoelii MSP-1 are not known, this region was chosen by amino acid sequence alignment of P. yoelii and P. falciparum MSP1 molecules. The resulting insert corresponds to amino acids 1394–1659 of P. yoelii MSP-1. The amplified DNA was ligated into the EcoRI site of the pBDGAL4-CAM vector (Stratagene) resulting in an inframe fusion with the Gal4 binding domain and the resulting plasmid was designated Py33pBD. For expres-

T.M. Daly et al. / Molecular & Biochemical Parasitology 117 (2001) 27–35

sion studies, this same region was amplified with an additional 5% ATG codon and inserted into the BamHI and SalI sites of the pESC-TRP expression vector (Stratagene). The resulting fusion added a carboxy-terminal, human c-myc peptide tag (EQKLISEEDL) to the Py33 molecule and was designated Py33pESC. A 462 bp region of the Py17XL MSP1 gene (nucleotides 3907–4368), with an additional 5% ATG codon, was amplified as described above. This fragment was inserted into the NotI and ClaI sites of the pESC-TRP and Py33pESC-TRP vectors and designated Py38TpESC and Py38T+33pESC, respectively. The resulting fusion added a C-terminal FLAG peptide tag (DYKDDDDK) to the Py38T molecule. All insert sequences and fusion junctions were confirmed by DNA sequencing using Big-Dye terminator chemistry (ABI).

2.5. Yeast two-hybrid assays The Py33pBD ‘bait’ construct was introduced into PJ69-4a cells by the lithium acetate-polyethylene glycol method [24] and maintained on SD agar plates without tryptophan (SD-Trp). To confirm the absence of autoactivation by Py33pBD, the pADGal4 ‘target’ vector without an insert was introduced into these cells and the cells were plated on minimal media lacking histidine, leucine and tryptophan (SD-His-Leu-Trp) medium with the addition 0, 2, 4, or 8 mM 3-aminotriazole (3-AT, an inhibitor of the enzyme encoded by the yeast HIS3 gene). Subsequently, a large-scale screen was performed. PJ69-4a cells containing the ‘bait’ vector (Py33pBD) were transformed with the Py17XL cDNA phagemid library as described above and were spread onto one hundred 150 mm plates of SC-HisLeu-Trp agar containing 2 mM 3-AT. Transformed cells were grown at 30 °C for 14 days at which time randomly selected His+ colonies were transferred to SD-Leu-Trp master plates. After 2 days at 30 °C, these colonies were replicated to SD-Ade-Leu-Trp plates and grown for 1 week at 30 °C. Both positive (pBD-p53+ pAD-SV-40 T-Antigen) and negative (pBD-Lamin C+ pAD-SV-40 T-Antigen) controls (Stratagene) were included on each plate. Ade+ colonies were subsequently assayed for b-galactosidase activity as follows. After placement of Whatman 1 filter disks onto SDLeu-Trp plates, His+, Ade+ cells were transferred to the filters and grown at 30 °C for 2 days. Filters were then removed from the plates, frozen in liquid nitrogen and allowed to thaw. After repeating the freeze-thaw cycle twice, these filters were placed onto fresh Whatman 1 disks that had been saturated with substrate buffer (100 mM sodium phosphate pH 7.4, 10 mM KCl, 2 mM MgSO4, 40 mM b-mercaptoethanol and 0.33 mg ml − 1 of X-gal) and incubated at 30 °C. When the positive control had developed sufficient color with respect to the negative, the His+, Ade+ colonies were scored for relative b-galactosidase activity.

29

His+, Ade+, LacZ+ yeast clones were randomly chosen for further analysis. Following overnight growth in SD-Leu broth at 30 °C, cells from each clone were pelleted, vortexed in the presence of lysis buffer, glass beads and phenol:chloroform:iso-amyl alcohol (24:24:1), and the released plasmid DNA was precipitated with ethanol. Following resuspension, each DNA sample was electroporated into competent XL1-blue MRF’ cells, which were plated on LB medium supplemented with 100 mg ml − 1 of ampicillin (LB-amp). A colony from each transformation was subcultured overnight in LB-amp broth and plasmid DNA was isolated by alkaline lysis. A portion of each plasmid DNA preparation was digested with HindIII and analyzed on agarose gels to identify the unique clones. Nucleotide sequence of cDNA inserts from these reactive clones was determined as described above.

2.6. Co-immunoprecipitation and immunoblotting The Py33pESC, Py38TpESC and Py38T+ 33pESC constructs described above, as well as the pESC-TRP vector, were introduced into the protease-deficient yeast strain YJN192 and were maintained on SD-Trp medium. For expression studies, these cells were initially grown overnight at 30 °C in YCA medium supplemented with 1.5% raffinose. The following day, these cultures were diluted into 10–20 vol. of YCA supplemented with 2% galactose (YCAGal) and grown overnight at 30 °C. Subsequently, the cells were pelleted, resuspended in sterile water, repelleted and stored at − 80 °C. For protein half-life studies, cells were grown as described above followed by growth in YCA supplemented with 2% glucose. Following the switch to glucose, aliquots of cells were taken at various times and frozen as described above and cell extracts were prepared from frozen pellets in lysis buffer (25 mM Tris–HCl pH 8.0, 50 mM NaCl, 5mM EDTA, 0.1% Triton X-100) containing a 1:200 dilution of fungal protease inhibitor cocktail (Sigma) as earlier described [25]. The protein concentration was determined by the Bradford method (Protein Reagent, BioRad) and lysates were analyzed directly by SDS-PAGE or used for co-immunoprecipitation as follows. Equivalent amounts of lysate protein in 100 ml of lysis buffer were incubated with 0.2–0.4 mg of rabbit anti-c-myc polyclonal antibody (Santa Cruz Biotechnology) on ice for 30 min. Subsequently, 20 ml of a 50% Protein A-agarose slurry (Sigma) were added and incubated on ice with frequent agitation. After 30 min, the Protein A-agarose was pelleted, washed three times with lysis buffer, resuspended in an equal volume of 2X SDS-PAGE sample buffer [26] and heated to 95 °C for 3 min. Following electrophoresis on 12.5% SDS-PAGE gels, proteins were transferred to nitrocellulose membranes [27]. The expressed tagged proteins were detected with either

30

T.M. Daly et al. / Molecular & Biochemical Parasitology 117 (2001) 27–35

mouse anti-c-myc IgG1 mAb (9E10, Santa Cruz Biotechnology, Inc) or mouse anti-FLAG IgG1 mAb (M2, Eastman Kodak), followed by goat polyclonal antimouse IgG-horseradish peroxidase conjugate (Cappel) using stabilized TMB substrate (Promega) or chemi-luminescence (SuperSignal® West Fempto, Pierce) according to the manufacturer’s instructions.

3. Results

3.1. Partial characterization of the Py17XL cDNA library We have utilized mRNA from an asynchronous infection of mouse red blood cells with P. yoelii 17XL to construct a cDNA activation domain library for use in the two-hybrid system. Due to the asynchronous nature of P. yoelii infection, it was anticipated that mRNA from all blood-stage forms of the parasite would be represented in the library. To determine the quality of the activation domain library, PCR amplification was performed with phage DNA from 15 randomly selected primary library plaques. Using a primer pair complementary to regions just flanking the EcoRI and XhoI insertion sites of the HybriZAP™ vector, ten reactions yielded product ranging in size from approximately 0.7–2.2 kb with an average size of 1.2 kb (Fig. 1A). We believe that the remaining PCR reactions were non-productive rather than failing to contain inserts as initial screening of phage plaques using X-Gal screening yielded greater than 95% of the phage contained inserts. These results suggest that the mRNA was of significant

quality. Due to the high A-T content of the malarial genome, some internal priming of the mRNA by the oligo(dT) primer/linker might be expected during cDNA synthesis. Since the average insert size is approximately 1.2 kb, internal priming would provide representation within the library of more 5% regions from longer mRNA transcripts. Using specific primer pairs for 5%, middle, and 3% regions of Py17XL MSP1, which has a 7.6 kb mRNA primary transcript [28], PCR amplification of library phagemid DNA was performed. As shown in Fig. 1B, a single band of the appropriate size was generated in each of these reactions. The presence of 5% sequence from such a large mRNA transcript suggests that the library was not limited to 3% regions that would have been initiated only at the poly(A) tail. However, this does not rule out the presence of full-length cDNAs from long mRNA transcripts. The unamplified library contained greater than 1× 106 independent clones.

3.2. Py33pBD screening of cDNA library and clone analysis A large-scale transformation of Py33pBD/PJ69-4a cells with the Py17XL cDNA phagemid library yielded 13.2× 106 transformants. This was determined from serial dilutions of the above transformed cells grown on SD-Leu-Trp medium which would require the presence of both binding domain (Trp+) and activation domain (Leu+) vectors. This represents coverage of approximately 10-fold of the cDNA library. For PJ69-4a cells to grow in the absence of histidine, a molecule containing a Gal4 binding domain and a molecule fused to the

Fig. 1. Partial characterization of the Py17XL cDNA expression library. (A) PCR amplification of DNA from 15 randomly selected plaques (primary library) using a primer pair complimentary to regions just flanking the EcoRI and XhoI insertion sites of the HybriZAP™ library vector (lanes 1 – 15). HindIII/EcoRI digested l phage DNA served as a molecular weight marker (lane 16). (B) PCR amplification of excised phagemid (pAD-GAL4) library DNA using primer pairs complimentary to 5%, middle and 3% regions of the PyMSP1 gene (lanes 1 – 3, respectively) assessed by agarose gel electrophoresis. A primer pair complimentary to PyMSP119 with Py17XL genomic DNA served as a positive PCR control (lane 4). HindIII/EcoRI digested l phage DNA served as marker (lane 5).

T.M. Daly et al. / Molecular & Biochemical Parasitology 117 (2001) 27–35

31

Table 1 Aligned amino acid sequences from Ade+His+b-gal+ activation domain library clones

Fifteen randomly selected Ade+His+b-gal+ library clones were grown and plasmid DNA isolated as described above. Following DNA sequence analysis using a primer 5% of the insert junction, the amino acid sequences were deduced and aligned with the homologous region of PyMSP-1. Nine of 15 clones expressed a unique overlapping region of MSP-1 beginning amino-terminal to Py33. The remaining 6 clones were found to be siblings. The amino acid sequence SNYLKIEK was common to all but one clone, 33-30pAD. aEFGTS or R imposed by the EcoRI linker in frame with the Gal4p activation domain. bStart site of clone is approximately 50 amino acids amino-terminal of SNYLKIEK….cThree sibling clones. d Two sibling clones. eOne sibling clones.

Gal activation domain must be brought into sufficient proximity to induce expression of the LYS2::GAL1 HIS3 reporter gene (His+). The growth of transformants on SD-His-Leu-Trp medium containing 2 mM 3-AT indicated that interaction(s) between Py33 (fused to the Gal4 binding domain) and expressed protein(s) from the cDNA library (fused to the Gal4 activation domain) was taking place. One hundred fifty His+ colonies, randomly chosen from 30 of the 100 plates used, were subjected to a second round of selection by plating onto SD-Ade-Leu-Trp medium. Requiring induction of the more tightly controlled GAL2 -ADE2 reporter gene, 86 of 150 colonies grew under this more stringent condition (Ade+). These colonies were additionally screened for b-galactosidase expression under control of the GAL7 promoter. Analysis of 60 Ade+ colonies revealed that 27 of these clones were positive by the b-galactosidase filter assay (b-gal+). Fifteen of these His+, Ade+, b-gal+ clones were subsequently grown and their individual activation domain vectors were isolated as described above. PCR analysis revealed that all of the clones possessed only one plasmid and that the insert sizes ranged from approximately 600 to 1200 bp (data not shown). The DNA sequence of 400–500 bases from the 5% activation domain-fusion protein junction was determined from each of the plasmid preparations using a primer specific for the activation domain (approximately 50 bases 5% of the EcoRI insertion site). Subsequent Blastn analysis (NCBI) of the sequence data revealed that all 15 clones encoded overlapping regions of the Py17XL MSP1 molecule (data not shown). As shown in Table 1, translation of

these sequences into amino acids and subsequent alignment indicated that all were in-frame with the Gal4 activation domain and that they included regions amino-terminal to PyMSP-133. In addition, it was determined that 9 of the 15 clones were unique, with the remaining six being siblings.

3.3. Localization of PyMSP138 and PyMSP133 interaction From the estimated size of the activation domain plasmid, it was possible that the protein–protein interaction represented a dimerization since sequences within the activation domain clones overlapped with the PyMSP-133 region. To further define the regions of interaction, deletion clones were constructed from the original bait vector Py33pBD and library clone 3311pAD, which contained the shortest insert (approx. 600 bp, data not shown). As shown in Fig. 2, removal of 25 amino acids from the amino-terminal end of the 33-11pAD insert (clone 33-11-5T25pAD) was sufficient to abrogate the interaction with Py33pBD as assessed by growth on SD-His-Leu-Trp plates containing 2 mM 3-AT. However, deletion of up to 125 residues from the carboxy-terminal end of clone 33-11pAD did not interfere with interaction with Py33pBD (whereas removal of 150 amino acids eliminated the interaction). These results indicate that a fragment of 77 amino acids or less, located within the PyMSP138 region, is sufficient to produce an interaction with PyMSP133 and this region does not overlap with PyMSP-133 and thus does not represent dimerization of the PyMSP-133 domain. Re-

32

T.M. Daly et al. / Molecular & Biochemical Parasitology 117 (2001) 27–35

Fig. 2. Localization of regions responsible for the PyMSP133 and PyMSP138 interaction. 5% (5T) and 3% (3T) deletion constructs were prepared from the Py33pBD and 33-11pAD plasmids by PCR amplification of DNA fragments with defined starting points. The resulting fragments were cloned into pBD-GAL4-CAM (Py33pBD derived fragments) or pGAD424 (Clontech) (33-11pAD derived fragments) and transformed into strain PJ69-4a as described above. The Py33pBD deletions were tested for their interactions with clone 33-11pAD and an empty pGAD424 vector and the 33-11pAD deletions were tested for their interactions with Py33pBD and an empty pBD-GAL4-CAM vector. The resulting yeast transformants were assessed for growth on SD-His-Leu-Trp plates containing 2.5 mM 3-aminotriazole. The degree of interaction was scored relative to that of the original Py33pBD and 33-11pAD transformed cells, which served as a positive control.

moval of 50 amino acids from the amino-terminal end of the Py33pBD insert (clone Py33-5T50pBD) eliminated the interaction with clone 33-11pAD. The deletion of 100 amino acids from the 3% region of Py33pBD insert (clone Py33-3T100pBD) also disrupted the interaction with clone 33-11pAD, while the removal of 50 residues did not (clone Py33-5T50pBD). These findings suggest that a significantly larger region of PyMSP133, greater than 150 residues, is also required.

3.4. Co-expression and co-immunoprecipitation The yeast two-hybrid system detects protein– protein interactions in the context of DNA-binding domain and activation domain fusion proteins. Therefore, we sought additional direct evidence for the interaction of the domains within the two MSP-1 domains. Thus, the fragments were expressed in yeast cells as distinct molecules to approach the question of whether the molecules could be co-immunoprecipitated. The pESCTRP (Stratagene) vector allows for the simultaneous expression of two inserts, each under control of a tightly controlled, highly inducible GAL promoter and each with a unique peptide tag. For this set of experiments, the Py33 region was inserted into the vector in-frame with a carboxy-terminal human c-myc peptide tag while a fragment derived from activation domain clone 33-11pAD (S1240…Q1393, designated Py38T) insert was cloned in-frame with a carboxy-terminal FLAG® peptide tag. Additional constructs were prepared which contained only one of the inserts in-frame with the

respective peptide tag. A protease deficient yeast strain was transformed, grown and lysates were prepared as described above. As can be seen in Fig. 3A, western blot analysis with anti-FLAG® antibody revealed a unique band with an apparent molecular weight of approximately 25 kDa in lysates from cells expressing the Py38T region (lanes 3 and 4) but not in those from cells transformed with the empty vector or expressing only Py33 (lanes 1 and 2). Similarly, cells expressing the Py33 region and analyzed as above with anti-c-myc antibody demonstrated a unique band of approximately 32 kDa (Fig. 3B, lanes 2 and 4) while those transformed with the empty vector or expressing only Py38T did not (lanes 1 and 3). In addition, a portion of these lysates was used for immunoprecipitation with anti-c-myc antibody, followed by western blot analysis with antiFLAG® antibody as described above. As can be seen in Fig. 3C, immune complexes from lysates of cells expressing both the Py33 and Py38T fragments (lane 4) contained a 25 kDa FLAG®-positive peptide while complexes from lysates of cells expressing only one of the inserts or of cells transformed with the empty vector did not (lanes 1–3). These results suggest that the two molecules interact directly and may be co-immunoprecipitated as a complex from the cytoplasm of yeast cells.

3.5. Stability of co-expressed Py33 and Py38T during induction As seen in Fig. 3A, we consistently observed that the relative amount of Py38T protein (FLAG-tagged) was

T.M. Daly et al. / Molecular & Biochemical Parasitology 117 (2001) 27–35

Fig. 3. Co-immunoprecipitation of the interacting proteins in yeast cell lysates. Western blot analysis was performed on lysates of YJN192 cells transformed with pESC-TRP co-expression constructs following induction, as described above. Equivalent amounts of total lysate protein were electrophoresed (SDS-PAGE) and transferred to nitrocellulose membranes. (A) Western analysis using anti-FLAG® antibody with cell lysates, empty vector (lane 1); Py33pESC-Trp with c-myc tag (lane 2); Py38TpESC-Trp with FLAG® tag (lane 3); Py33 + Py38TpESC-Trp with respective tags (lane 4). (B). Lysate samples were run as in A, but probed with anti-c-myc antibody. (C) Immune-precipitates of the above lysates using the anti-myc antibody were prepared as described above, electrophoresed, and probed with anti-FLAG® antibody as in (A).

significantly less when Py38T was expressed alone as compared with cells expressing both Py38T and Py33 (lane 4), although approximately equivalent amounts of total protein were loaded in each lane. To determine the influence of the co-expression of Py33 on the halflife of Py38T within yeast cells, cells expressing only Py38T and cells expressing both proteins were grown in medium containing galactose to allow expression of the molecules. At T=0, the cells were switched to medium containing glucose, which completely represses expression from the GAL promoter. Aliquots of the cultures were taken at this time point and at 15, 30, 60, 120, and 240 min subsequent to the switch. As described above, cell lysates were prepared, protein concentrations were determined and normalized and equivalent amounts of total protein from all samples were analyzed by western blot with anti-FLAG® antibody. As can be seen in Fig. 4A, the level of Py38T expressed alone (panel I) decreased rapidly following repression of expression with an apparent half-life of approximately 15 min. Additionally we observed (in some cases) a proteolytic fragment of the full length Py38T molecule. However, the Py38T molecule persisted when expressed in the presence of Py33 (panel II) for a significantly longer period with an apparent half-life of approximately 2 h. To determine whether the converse was true, cells expressing Py33 alone and with Py38T were grown and processed as above. Western blot analysis, with anti-c-myc antibody, revealed that the half-life of Py33 appears to be unaffected by the absence (Fig. 4B, panel I) or presence (Fig. 4B, panel II) of Py38T. These results provide additional evidence for an interaction between these two domains of MSP-1 and suggest that the interaction of the two domains protects the Py38T

33

Fig. 4. Stability of co-expressed proteins. (A) Western blot analysis was performed with anti-FLAG® antibody on lysates of Py38TpESCTrp (group I) and Py33 +Py38TpESC-Trp (group II) transformed YJN192 cells taken at 0, 15, 30, 60, 120, and 240 min (lanes 1 –6, respectively) following transfer of the yeast cells from growth in galactose to growth in glucose, allowing complete repression of expression from the GAL promoter. Equivalent amounts of total lysate protein were loaded in each lane. (B) Similar analysis was done with anti-c-myc antibody on lysates of Py33pESC-Trp (group I) and Py33 + Py38TpESC-Trp (group II) transformed cells as in (A).

domain from proteolytic degradation within the yeast cells. We do not know if the observation is also true within the parasite.

4. Discussion The importance of MSP-1 during the erythrocytic phase of malarial infection and its potential as a bloodstage specific vaccine component have been clearly demonstrated. However, we have little knowledge of this molecule’s role(s) in the infectious process. Despite specific proteolytic processing, MSP-1 remains associated with the merozoite surface until the time of erythrocyte invasion. To further our understanding of the interactions between the proteolytic fragments of MSP1 and their association with other merozoite surface molecules, we have employed the yeast-two hybrid system. The yeast two-hybrid system is an in vivo, yeastbased genetic assay that utilizes the reconstitution of specific transcription to reveal interactions between two proteins [30,31]. In this system, the ‘bait’ protein is fused to the DNA-binding domain of the yeast Gal4 protein and is used to screen an expression library or ‘target’ fused to the Gal4p activation domain. It should be noted that full-length cDNA transcripts of mRNA

34

T.M. Daly et al. / Molecular & Biochemical Parasitology 117 (2001) 27–35

are not required, rather only those cDNA regions encoding the interacting domains. Only when an interaction occurs between bait and target proteins is there activation of specific marker genes within the yeast cell. The inclusion of two selectable growth markers (HIS3 and ADE2 ) and one enzymatic marker (E. coli LacZ), each under control of a Gal4p regulated promoter, aids in the elimination of false interactions. This system can often detect weak or transient interactions and purified target proteins or specific antibodies are not required. In addition, as an in vivo assay, the expressed proteins are more likely to be in a native conformation. In this study, we have utilized the 33 kDa region of the P. yoelii MSP-1 (bait), analogous to P. falciparum MSP-133 [11] by sequence alignment, to screen a P. yoelii cDNA library (target) for specific protein – protein interactions. DNA sequence analysis of 15 randomly selected library clones, positive for all three of the two-hybrid selectable markers, revealed nine unique but overlapping inserts while the remaining six were found to be siblings. The region encoded by these inserts was from the MSP-1 molecule and corresponded to an area within the 38 kDa domain analogous to P. falciparum MSP-138 by sequence alignment. However, we could not rule out that the detected interaction represented a dimerization of the 33 kDa domain as the library clones extended into the 33 kDa region. Aminoand carboxy-terminal deletions, within the context of the binding and activation domain vectors, allowed us to map the minimal interacting domains to non-overlapping regions of approximately 75 amino acids from the 38 kDa fragment and approximately 200 amino acids from the 33 kDa fragment. Using sequence alignments between MSP-1 molecules of P. yoelii and P. falciparum West African Wellcome (Lagos) strain, the region of the 38 kDa fragment that we have delineated in these studies, is immediately carboxy-terminal of and does not overlap with the region defined by Davidson and co-workers as interacting with red cells using erythrocyte binding assays [32]. It is interesting to note that the location of these domains, in P. falciparum MSP-1, flanks one of the proteolytic cleavage sites. In addition, Lyons and co-workers described a monoclonal antibody (Mab 710) that required regions of both the 38 and 33 kDa fragment for its epitope, again suggesting that there is an interaction between these two domains of MSP-1 [33]. It is not possible from the present study to determine whether this interaction is intra- or intermolecular. Addressing this question in vivo is difficult due to the haploid nature of the parasite and more likely the essential nature of the MSP-1 molecule [34,35]. In addition, we demonstrated that co-expression of the 33 and 38 kDa regions in yeast cells significantly increased the biological half-life of the 38 kDa fragment. At this time, we do not know whether this is also true within the parasite.

The yeast two-hybrid system has been used with known malarial proteins to demonstrate an interaction [36,37], but to our knowledge this represents the first application of this system in a library-type screen for interacting proteins in plasmodia. In this regard, it should be noted that development of two-hybrid reagents and analytical systems for plasmodia will be generally applicable to a number of interesting cell biology questions concerning these organisms. This system has been used at a genome wide level in yeast as an important tool in the analysis of functional networks within the cell [38], a feature that will become more compelling as the parasite genome projects continue to generate new data. The present study has provided evidence of an interaction between two distinct domains of the MSP-1 molecule. It is interesting to speculate about the potential role of this interaction in the functioning of MSP-1. Several investigators have suggested that the MSP-1 protein participates in the initial binding of the merozoite to the red blood cell, a function that would be facilitated by its apparent location on the entire surface of the merozoite [29,30]. Despite specific proteolytic cleavage, MSP-1 remains on the merozoite surface as a high molecular weight complex. It is possible that the protein–protein interaction, described above, might contribute significantly to maintaining the integrity of the MSP-1 molecule and participate in the formation of a protective structural lattice on the surface of the merozoite. Following the second step of proteolytic processing, the carboxy-terminal 19 kDa domain may function to signal the parasite that it is now within the interior of the erythrocyte. In addition, the usefulness of this yeast system may facilitate the identification of a small molecule that disrupts this potentially critical interaction. The subsequent in vivo consequences of disrupting this interaction may be explored and they may provide a new avenue of therapeutic intervention against the malarial parasite.

Acknowledgements This work was funded by NIH AI48226 to LWB.

References [1] Long CA. Immunity to blood stages of malaria. Curr Opin Immunol 1993;5:548 – 56. [2] Good MF, Kaslow DC, Miller LH. Pathways and strategies for developing a malaria blood-stage vaccine. Ann Rev Immunol 1998;16:57 – 87. [3] Holder AA. The precursor to major merozoite surface antigens: structure and role in immunity. Prog Allergy 1988;41:72 –97. [4] Cooper JA. Merozoite surface antigen-1 of Plasmodium. Parasitol Today 1993;9:50 – 4.

T.M. Daly et al. / Molecular & Biochemical Parasitology 117 (2001) 27–35 [5] Holder AA, Freeman RR. Immunization against blood-stage rodent malaria using purified parasite antigens. Nature 1981;294:361 – 4. [6] Siddiqui WA, Tam LQ, Kramer KJ, Hui GSN, Yamaga KM, Chang SP, Chan EB, Kan SC. Merozoite surface coat precursor completely protects Aotus monkeys against Plasmodium falciparum malaria. Proc Natl Acad Sci USA 1987;84:3014 –8. [7] Majarian WR, Daly TM, Weidanz WP, Long CA. Passive protection against murine malaria with an IgG3 monoclonal antibody. J Immunol 1984;132:3126 –30. [8] McBride JS, Heidrich HG. Fragments of the polymorphic Mr 185 000 glycoprotein from the surface of isolated Plasmodium falciparum merozoites form an antigenic complex. Mol Biochem Parasitol 1987;23:71 –84. [9] Stafford HL, Gunder B, Harris A, Heidrich H, Holder AA, Blackman MJ. A 22 kDa protein associated with the Plasmodium falciparum merozoite surface protein-1 complex. Mol Biochem Parasitol 1996;80:159 –69. [10] Trucco C, Fernandez-Reyes D, Howell S, Stafford WH, ScottFinnigan TJ, Grainger M, Ogun SA, Taylor WR, Holder AA. The merozoite surface protein 6 gene codes for a 36 kDa protein associated with the Plasmodium falciparum merozoite surface protein-1 complex. Mol Biochem Parasitol 2001;112:91 –101. [11] Blackman MJ, Holder AA. Secondary processing of the Plasmodium falciparum merozoite surface protein-1 (MSP-1) by a calcium-dependent membrane bound serine protease: shedding of MSP-133 as a noncovalently associated complex with other fragments of the MSP-1. Mol Biochem Parasitol 1992;50:307 – 15. [12] Blackman MJ, Dennis ED, Hirst EM, Kocken CH, ScottFinnegan TJ, Thomas AW. Plasmodium knowlesi: secondary processing of the malaria merozoite surface protein-1. Exp Parasitol 1996;83:229 – 39. [13] O’Dea KP, McKean PG, Harris A, Brown KN. Processing of the Plasmodium chabaudi chabaudi AS merozoite surface protein 1 in vivo and in vitro. Mol Biochem Parasitol 1995;72:111 – 9. [14] Blackman MJ, Chappel JA, Shai S, Holder AA. A conserved parasite serine protease processes the Plasmodium falciparum merozoite surface protein-1. Mol Biochem Parasitol 1993;62:103 – 14. [15] Blackman MJ, Heidrich HG, Donachie S, McBride JS, Holder AA. A single fragment of a malaria merozoite surface protein remains on the parasite during red cell invasion and is the target of invasion-inhibiting antibodies. J Exp Med 1990;172:379 – 82. [16] Tanabe K, Mackay M, Goman M, Scaife JG. Allelic dimorphism in a surface antigen gene of the malaria parasite Plasmodium falciparum. J Mol Biol 1987;195:273 –87. [17] Daly TM, Long CA. A recombinant 15-kilodalton carboxyl terminal fragment of Plasmodium yoelii yoelii 17XL merozoite surface protein-1 induces a protective response in mice. Infect Immun 1993;61:2462 –7. [18] Daly TM, Long CA. Humoral response to a carboxyl-terminal region of the merozoite surface protein-1 plays a predominant role in controlling blood-stage infection in rodent malaria. J Immunol 1995;155:236 – 43. [19] Blackman M, Scott-Finnigan TJ, Shai S, Holder AA. Antibodies inhibit the protease-mediated processing of a malaria merozoite surface protein. J Exp Med 1994;180:389 –93. [20] Guevara Patino JA, Holder AA, McBride JS, Blackman MJ. Antibodies that inhibit malaria surface protein-1 processing and erythrocyte invasion are blocked by naturally acquired human antibodies. J Exp Med 1997;186:1689 –99.

35

[21] Eling W. Ficoll fractionation for the separation of parasitized erythrocytes from malaria infected blood. Bull WHO 1977;55:105 – 14. [22] Lewis AP. Cloning and analysis of the gene encoding the 230kilodalton merozoite surface antigen of Plasmodium yoelii. Mol Biochem Parasitol 1989;36:271 – 82. [23] Vaidya AB, Arasu P. Tandemly arranged gene clusters of malarial parasites that are highly conserved and transcribed. Mol Biochem Parasitol 1987;22:249 – 57. [24] Gietz D, St Jean A, Woods RA, Schiestl RH. Improved method for high efficiency transformation of intact yeast cells. Nucl Acids Res 1992;20:1425. [25] Santos RC, Waters NC, Creasy CL, Bergman LW. Structurefunction relationships of the yeast cyclin-dependent kinase Pho85. Mol Cell Biol 1995;15:5482 – 91. [26] Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970;227:680 –3. [27] Towbin H, Staehelin T, Gordon J. Electrophoretic transfer of proteins from polyacrylamide to nitrocellulose sheets, procedure and some applications. Proc Natl Acad Sci USA 1979;76:4350 – 4. [28] Burns JM Jr, Daly TM, Vaidya AB, Long CA. The 3% portion of the gene for a Plasmodium yoelii merozoite surface antigen encodes the epitope recognized by protective monoclonal antibody. Proc Natl Acad Sci USA 1988;85:602 – 6. [29] Perkins ME, Rocco LJ. Sialic acid-dependent binding of Plasmodium falciparum merozoite surface antigen, Pf200, to human erythrocytes. J Immunol 1988;141:3190 – 6. [30] Fields S, Song O. A novel genetic system to detect protein – protein interactions. Nature 1989;340:245 – 50. [31] Luban J, Goff SP. The yeast two-hybrid system for studying protein – protein interactions. Curr Opin Biotechnol 1995;6:59 – 67. [32] Nikodem D, Davidson E. Identification of a novel antigenic domain of Plasmodium falciparum merozoite surface protein-1 that specifically binds to human erythrocytes and inhibits parasite invasion, in vitro. Mol Biochem Parasitol 2000;108:79 –91. [33] Chulay JD, Lyons JA, Haynes JD, Meierovics AI, Atkinson CT, Aikawa M. Monoclonal antibody characterization of Plasmodium falciparum antigens in immune complexes formed when schizonts rupture in the presence of immune serum. J Immunol 1987;139:2768 – 74. [34] O’Donnell RA, Saul A, Cowman AF, Crabb BS. Functional conservation of the malaria vaccine antigen MSP-119 across distantly related Plasmodium species. Nat Med 2000;6:91 –5. [35] Cowman AF, Baldi DL, Healer J, Mills KE, O’Donnell RA, Reed MB, Triglia T, Wickham ME, Crabb BS. Functional analysis of proteins involved in Plasmodium falciparum merozoite invasion of red blood cells. FEBS Lett 2000;476:84 –8. [36] Siden-Kiamos I, Vlachou D, Gabrielle M, Beetsma A, Waters AP, Sinden RE, Louis C. Distinct roles for PBS21 and PBS25 in the in vitro ookinete to oocyst transformation of Plasmodium berghei. J Cell Sci 2000;113:3419 – 29. [37] Vlachou D, Lycett G, Siden-Kiamos I, Blass C, Sinden RE, Louis C. Anopheles gambiae lamin interacts with the P25 surface protein of Plasmodium berghei ookinetes. Mol Biochem Parasitol 2001;112:229 – 37. [38] Uetz P, Giot L, Cagney G, Mansfield TA, Judson RS, Knight JR, Lockshon D, Narayan V, Srinivasan M, Pochart P, QureshiEmili A, Li Y, Godwin B, Conover D, Kalbfleisch T, Vijayadamodar G, Yang M, Johnston M, Fields S, Rothberg JM. A comprehensive analysis of protein – protein interactions in Saccharomyces cere6isiae. Nature 2000;403:623 – 7.