Previous Article | Next Article ![]()
Infection and Immunity, September 2005, p. 5936-5944, Vol. 73, No. 9
0019-9567/05/$08.00+0 doi:10.1128/IAI.73.9.5936-5944.2005
Copyright © 2005, American Society for Microbiology. All Rights Reserved.
Department of Immunology, Walter Reed Army Institute of Research, Silver Spring, Maryland 20910,1 Division of Microbiology,2 Division of Parasitology,3 Division of Lab Animals, Central Drug Research Institute, Lucknow 226001, India4
Received 11 March 2005/ Returned for modification 22 April 2005/ Accepted 11 May 2005
|
|
|---|
|
|
|---|
P. cynomolgi, which infects rhesus macaques in southeast Asia, is a closely related species to P. vivax; human transmission of P. cynomolgi has also been reported (25). The two parasites share a similar clinical course of infection (26), a reticulocyte-specific invasion (17), the presence of Schuffner's dots on infected erythrocytes (2), and a dormant liver hypnozoite stage that is responsible for a relapsing blood stage infection (23). P. cynomolgi and P. vivax have similar genomic GC content, and rRNA analysis confirms their close taxonomic relatedness (43). High homology of prime vaccine candidates such as the apical membrane antigen 1, 97% (13); circumsporozoite protein, 90% (16); erythrocyte binding protein, 76% (32); and the 42-kDa fragment of the merozoite surface protein 1 (MSP-1), 75% (31); have been reported. Although it is believed that the P. cynomolgi-rhesus model can serve as a model for testing P. vivax antigens, there is only one previous report where this model was used to test the efficacy of a recombinant P. vivax vaccine (22).
The merozoite surface protein 1 (MSP-1), found on the surface of Plasmodium merozoites, has been a prime vaccine candidate for many years (10). Following its synthesis as a 200-kDa precursor, the MSP-1 molecule undergoes step-wise proteolytic processing resulting in a glycosyl-phosphatidylinositol-anchored 42-kDa protein (MSP-142) on the surface of free merozoites (7). This 42-kDa intermediate undergoes secondary processing at the time of invasion, releasing a 33-kDa soluble polypeptide (MSP-133) and leaving behind on the invading merozoite a glycosyl-phosphatidylinositol-anchored 19-kDa form (MSP-119) (7). Depending on the species, MSP-119 contains 10 or 11 cysteine residues that form five disulfide bonds. Immunization with recombinant MSP-142 and MSP-119 raises antibodies that inhibit invasion of merozoites in vitro (9) and protects the immunized animals against live parasite challenge (29).
We have previously reported a process for the production of correctly folded human vaccine-grade P. vivax MSP-142 protein in the bacterial expression host Escherichia coli Origami(DE3) (14). This bacterial strain has mutations in both the thioredoxin reductase (trxB) and glutathione reductase (gor) genes, which greatly enhances the disulfide bond formation of recombinant proteins expressed in its cytoplasm, hence resulting in a soluble product (6). However, the transposon-mediated genetic modifications of E. coli carried both the tetracycline and the kanamycin resistance genes as selectable markers and therefore made the cells incompatible with the production of human use vaccines that relied on plasmids carrying the Tetr or the Kanr genes. The use of a compatible expression host E. coli BL21(DE3) resulted in the production of an MSP-142 protein that was insoluble and located in the inclusion body fraction of the cells. In this report we have shown that soluble MSP-142 protein can be obtained from the BL21(DE3) cells by in vitro refolding of the inclusion body-derived protein under controlled redox conditions. Both the soluble and refolded P. vivax MSP-142 products appeared to be structurally similar based on biophysical analysis and reactivity with conformational epitope-specific monoclonal antibodies.
It was our goal, therefore, to seek evidence of the immunological equivalence of these proteins in a parasite challenge model. Towards that end we report here the results of a vaccination study comparing the immunogenicity and efficacy of the soluble and refolded P. vivax MSP-142 products in the P. cynomolgi-rhesus model. We have also produced equivalent homologous constructs of P. cynomolgi MSP-142 protein both in their soluble and refolded forms, and these two proteins were also used as positive control immunogens in this study.
|
|
|---|
Expression and purification of soluble PcMSP-142 and PvMSP-142 proteins. Recombinant expression plasmids carrying the PcMSP-142 or PvMSP-142 gene were transformed into E. coli Origami(DE3) strain cells (Novagen, San Diego, CA). Expression was carried out in a 10-liter fermentor (New Brunswick Scientific, New Brunswick, NJ) as described previously (14). Protein expression was induced with a final 0.1 mM concentration of isopropyl-ß-D-thiogalactopyranoside at 25°C for 2 h, and cells were harvested by centrifugation. A two-step purification protocol comprising a nickel-nitrilotriacetic acid (Ni-NTA) column followed by a Q-Sepharose column was used to purify both soluble PvMSP-142 and soluble PcMSP-142 proteins essentially as described previously (14). The purified proteins were dialyzed against phosphate-buffered saline (PBS) prior to use.
Expression and purification of refolded PcMSP-142 and PvMSP-142 proteins. The recombinant plasmids carrying the PcMSP-142 or PvMSP-142 gene were transformed into E. coli BL-21(DE3) strain cells (Novagen, San Diego, CA). The fermentation and induction conditions were identical to those described above for soluble protein expression. To allow upscaling of the procedure and evaluation of different lots of material, the amount of all buffers and resins were indexed to a starting cell weight (scw). Cell paste from both PcMSP-142 and PvMSP-142 fermentations were suspended (10 ml g1 scw) in resuspension buffer (RB; 15 mM sodium phosphate, 450 mM NaCl, pH 7.4). Cells were lysed by microfluidization, and the insoluble inclusion bodies were harvested by centrifugation (12,000 x g, 45 min, 4°C). The pellet was suspended in cold RB (10 ml g1 scw) using a manual glass tissue homogenizer, and the suspension was centrifuged as before. The washed pellet was resuspended in solubilization buffer (SB; 7 M urea in RB) (6.6 ml g1 scw). The pellet was suspended by homogenization, incubated for 20 min on a shaking platform at room temperature (RT), and centrifuged (12,000 x g, 45 min, 10°C). A nickel-nitrilotriacetic acid (Ni-NTA) agarose (QIAGEN, Valencia, CA) column (0.6 ml gm1 scw) was equilibrated with SB, and the cleared supernatant was loaded onto the column. The column was washed with 10 column volumes (CV) of SB containing 20 mM imidazole (pH 7.4), and protein was eluted with 5 CV of SB containing 500 mM imidazole (pH 7.4). The Ni-NTA agarose column-eluted protein was refolded by a rapid 20-fold dilution in refolding buffer (20 mM sodium phosphate, 1 mM reduced glutathione [GSH], 0.05 mM oxidized glutathione [GSSG]; pH 8.0). After refolding (16 h, 22°C) in sealed bottles, the protein was bound to a Q-Sepharose resin column (0.6 ml gm1 scw) pre-equilibrated with 20 mM sodium phosphate (pH 8.0). The Q-Sepharose resin was washed with 10 CV of 20 mM phosphate buffer (pH 8.0), followed by 5 CV of the same buffer containing 50 mM NaCl (pH 8.0). The PvMSP-142 or PcMSP-142 protein was eluted from the column in 20 mM sodium phosphate and 200 mM NaCl (pH 8.0). The fractions containing the protein were pooled and dialyzed against PBS. The endotoxin content of the final product was estimated by a chromogenic Limulus amebocyte lysate (LAL) endpoint assay (Associates of Cape Cod, Falmouth, MA) following manufacturer's instructions.
Formulation, lyophilization, and storage. Purified proteins were quantified by optical density at 280 nm measurement, and the concentration was adjusted to a final 333 µg ml1 in 3.15% sucrose, 5 mM sodium phosphate, 150 mM NaCl, pH 7.0. The protein was dispensed at 1.65 ml per vial, frozen, and lyophilized for 24 h (Flex-Dry MP; FTS System, Stone Ridge, NY). The lyophilized material was stored at 30°C except during transport from WRAIR to Lucknow, India, when they were stored at RT. Analysis by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) of lyophilized preparations of all four MSP-142 proteins indicated that they were stable if stored at 30°C for at least 2 years or at RT for more than 6 months (data not shown).
SDS-PAGE and Western blotting. Purity of the final products was judged by SDS-PAGE using 4 to 12% Bis-Tris NuPage gels (Invitrogen, Carlsbad, CA) run under reducing and nonreducing conditions. For Western blotting, proteins were electrophoretically transferred to a nitrocellulose membrane followed by blocking with Casein blocker (Pierce, Rockford, IL). The blots were incubated in appropriate dilutions of primary and secondary antibodies for 2 h and then washed with PBST (PBS with 0.05% Tween-20) before development using BM-POD substrate (Roche, Indianapolis, IN). Monoclonal antibodies (MAb) 5.14 and 9.14 and the polyclonal rabbit serum against a PvMSP-119 protein produced in the baculovirus infected insect cell system (30) were used to confirm the correct folding of the products.
Animals. Rhesus monkeys (Macaca mulatta) were housed in the primate care facility of the Central Drug Research Institute (CDRI), Lucknow, India, and immunization and challenge protocols were approved by the Institutional Animal Ethics Committee of CDRI and the Committee for the Purpose of Control and Supervision of Experiments on Animals, Government of India. Monkeys were quarantined for 45 days and tested for the absence of common diseases, such as tuberculosis (by skin test and X-ray examination), and blood-borne parasites (by slide exam). Monkeys of either sex were distributed in groups of similar sex ratio and average weight (4.5 to 6.3 kg). Animals in each of the five groups were to receive one of the following preparations: soluble PvMSP-142 protein (PvS), refolded PvMSP-142 (PvR), soluble PcMSP-142 (PcS), refolded PcMSP-142 protein (PcR), or an adjuvant-only control (C).
Immunization schedule and bleeding. Lyophilized protein in each vial was completely resuspended and dissolved in 1.65 ml of sterile distilled water followed by the addition of 3.85 ml of Montanide ISA720 adjuvant. Vials were vortexed for about 2 min to form thick emulsion, and 1.0 ml of the emulsion containing 100 µg of protein was vaccinated per dose. Control monkeys received 1.0 ml of PBS:Montanide ISA720 (30:70, vol/vol). The skin on the back of the monkeys was shaved, and the vaccine was administered subcutaneously at multiple sites. Three doses of the vaccine were administered at 4-week intervals. Animals were prebled 1 week prior to the first day of immunization and then bled every 2 weeks until the date of challenge. Immunization sites were monitored daily for adverse local reaction. The weights of the animals were recorded at 4-week intervals during the study.
Enzyme-linked immunosorbent assay (ELISA). ELISA was performed in 96-well plates essentially as described before (41, 21). Briefly, the wells of the microtiter plates were coated with recombinant PvMSP-142 or PcMSP-142 (40 ng/100 µl/well) diluted in PBS by incubating the plates at 37°C for 1 h and then overnight at 4°C. The plates were blocked by incubation with 300 µl of 3% (wt/vol) nonfat dry milk powder (prepared in PBS) at 37°C for 2 h. After three washes with PBST, the plates were incubated with 100 µl of appropriately diluted antibodies (diluted in 1% milk-PBS) for 2 h at 37°C. The plates were washed thrice with PBST followed by 1.5 h of incubation with peroxidase-conjugated anti-monkey immunoglobulin G (1:2,000 diluted in 1% milk-PBS; Sigma). After washing the plates with PBST, the color was developed by adding the substrate solution (1 mg/ml O-phenylenediamine in citrate-phosphate buffer, pH 5.0, containing 1 µl/ml H2O2). The reaction was stopped after 10 min by adding 5 N H2SO4, and the absorbance was read at 490 nm using a Molecular Devices Spectromax Plus UVmax microplate ELISA reader.
Challenge and parasitemia records. The monkeys in all five groups were challenged with 1 x 104 P. cynomolgi blood stage parasites 4 weeks after the last injection. The Giemsa-stained thick and thin blood smears prepared from individual monkeys were observed under a microscope using a 100x objective. The number of parasites was counted initially from thick smears and subsequently from thin smears and expressed in terms of parasites per cubed millimeter. The blood smears were prepared once daily from day 4 onwards until day 45. After 45 days, the monkeys were cured with chloroquine (34).
Statistical analysis. Group averages were compared by Fischer's least significant difference method (38). Pearson's coefficient of correlation (R2) was calculated by linear regression.
Nucleotide sequence accession number. The sequence determined in the course of this work has been deposited in GenBank under accession no. AY869723.
|
|
|---|
Expression and purification of recombinant proteins.
In the E. coli host strain Origami(DE3), both PvMSP-142 and PcMSP-142 proteins were produced as soluble molecules, and both antigens were purified to homogeneity using an identical two-step purification process. In the E. coli BL21(DE3) host, both PvMSP-142 and PcMSP-142 proteins were localized in the inclusion body fraction. The insoluble proteins were obtained by isolation of the inclusion bodies, resuspension in urea, partial purification on an Ni-NTA column, and refolding using a GSH-GSSH redox couple solution to gain optimal reactivity to a panel of conformation-specific monoclonal antibodies. The ratio of concentration of GSH to GSSG used during the in vitro refolding step was optimized for PvMSP-142 refolding using a checker board of concentration ratios and the reactivity to conformational monoclonal antibodies as an indicator of correct folding (data not shown). The refolded PvMSP-142 and PcMSP-142 proteins were further purified by ion exchange on a Q-Sepharose resin. Figure 1 (top panel) shows the SDS-PAGE analysis of all four MSP-142 products. A major band
45 to 50 kDa represented the MSP-142 monomer. A
33-kDa doublet represented a C-terminal degradation product of MSP-142 as confirmed by N-terminal sequencing and by Western blot analysis using antibodies against the N-terminal 6x His tag. A minor band at
90 kDa on the gels corresponded to the dimer of MSP-142. The combined purity of MSP-1-specific bands in all four products exceeded >99% by laser densitometry. The major band in each of the four lanes migrated as a discreet band under reducing and nonreducing conditions, indicating conformational homogeneity of the disulfide-bonded products. The residual endotoxin content of all four protein preparations was <50 EU/100-µg dose. The soluble and refolded PvMSP-1 and PcMSP-1 proteins reacted on a Western blot with PvMSP-119-specific conformational MAbs 5.14 and E9.14 and with a rabbit polyclonal serum against PvMSP-119 (Fig. 1, bottom panel).
![]() View larger version (100K): [in a new window] |
FIG. 1. Top panel, Coomassie blue-stained SDS-PAGE gel showing 5 µg of the soluble (S) and refolded (R) forms of the P. vivax (Pv) and P. cynomolgi (Pc) MSP-142 products under nonreducing and reducing conditions. The MSP-1-specific bands at 42 and 33 kDa are indicated by arrows. Bottom panel, Western blot reactivity of the four proteins with conformation dependent PvMSP-1-specific mouse MAbs 5.14 and E9.14 and with a rabbit polyclonal serum against PvMSP-119. m, molecular mass marker lanes.
|
![]() View larger version (13K): [in a new window] |
FIG. 2. Group wise mean of the log ELISA titers (y axis) at each of the six bleeds (x axis). The immunization events are denoted by arrows at the bottom. Bleeds were collected at 2-week intervals, and immunizations were administered at 4-week intervals (arrows).
|
|
View this table: [in a new window] |
TABLE 1. Endpoint titers at the time of challenge
|
![]() View larger version (17K): [in a new window] |
FIG. 3. Daily parasitemia records of individual monkeys (parasites mm3, y axis) plotted against days following the day of challenge (x axis). Groups shown from left to right in the top row are adjuvant control, P. cynomolgi MSP-142 soluble, and P. cynomolgi MSP-142 refolded; the bottom row shows P. vivax MSP-142 soluble and P. vivax MSP-142 refolded.
|
![]() View larger version (16K): [in a new window] |
FIG. 4. Group-wise average cumulative parasitemia (parasites mm3) of the 35-day follow-up was plotted for all immunized groups. Only odd-day parasitemias were plotted, and the bars represent the standard errors calculated from the means.
|
![]() View larger version (13K): [in a new window] |
FIG. 5. Parameters chosen to represent parasite burden (open bars). Group-wise mean of primary peak parasitemia (A); secondary peak parasitemia (B); average daily parasitemia (C); number of days when >10,000 parasites mm3 were recorded (D). Standard deviation within group is plotted as a line above each bar.
|
|
|
|---|
A similar level of protection induced by both homologous and heterologous MSP-142 antigens was not surprising given their immunological and structural relatedness. Monoclonal and polyclonal antibodies against several antigens of P. cynomolgi and P. vivax have been shown to be cross-reactive (20, 46), and infection induced by P. cynomolgi sporozoite injection protects humans against a P. vivax challenge (39). Mouse monoclonal and naturally acquired human antibodies against PvMSP-1 also cross-react with a baculovirus recombinant PcMSP-1 protein; in that study deglycosylation of the MSP-1 product improved the cross-reactivity of antibodies (19). Since E. coli-produced PvMSP-142 was naturally nonglycosylated, it was an attractive alternative to employ the P. cynomolgi-rhesus model for the purpose of evaluating the protective efficacy of P. vivax MSP-142 products. The E. coli-produced PvMSP-142- and PcMSP-142-induced cross-reacting antibodies and vaccination with PvMSP-142 induced a similar level of partial protection against P. cynomolgi parasites, as seen in the homologous vaccination groups. While there are few reports of cross-species protection in malaria, the results of this trial suggest the importance of this model for evaluating immune responses to other conserved vaccine candidates of P. vivax.
The protective response to MSP-1 has been shown to be antibody mediated using in vitro invasion inhibition and passive transfer experiments (44). Within the 42-kDa MSP-1, protective epitopes have been mapped to the two epidermal growth factor (EGF)-like domains in P. falciparum MSP-119 (1). Data presented here confirmed a positive correlation between PvMSP-142- and PcMSP-142-specific antibody titers and parameters chosen to indicate parasite burden. Analysis of the amino acid differences between PcMSP-142 and PvMSP-142 might reveal some clues toward the cross-species protection and indicate the importance of equivalent amino acid positions for the binding of invasion inhibitory antibodies. Out of the 86 amino acid differences between PvMSP-142 and PcMSP-142, 11 differences occur in the EGF domains containing the MSP-119 region. Previous studies have used binding of inhibitory and noninhibitory P. falciparum and P. yoelii MSP-119-specific MAbs to PfMSP-1 or PyMSP-119 protein constructs to identify amino acid positions on MSP-119 that are critical for the binding of invasion inhibitory antibodies. In an attempt to understand the cross-species protection observed in this trial, we analyzed each of the 11 amino acid differences between PvMSP-119 and PcMSP-119 in light of the results from previous studies with P. falciparum and P. yoelii MSP-1. The following amino acid differences were seen between PcMSP-119 and PvMSP-119: R6-T, E14-D, Y33-T, D37-E, A38-G, P45-S, M47-V, K52-N, N65-T, K67-S, and E69-K, respectively (amino acids were numbered as reported with the crystal structure of P. cynomolgi MSP-119 [8]). The differences at amino acid positions 14, 37, 38, and 47 are side chain conservative changes. A nonconservative change in P. falciparum MSP-1 at the position corresponding to R6 has been shown to have no effect on the binding of P. falciparum MSP-1-specific inhibitory MAbs 12.8, 12.10, and 5B1 (40). The E14 corresponding position in PyMSP-119 has been shown to be critical for the binding of inhibitory PyMSP-119 MAbs F5 and B6 (5); this position also represents one of the four amino acid polymorphic sites in MSP-119 of P. falciparum. Since the challenge with the P. falciparum FVO allele results in higher protection in the FVO-vaccinated Aotus monkeys compared to the 3D7 MSP-119-vaccinated group (24), we hypothesize that a conserved difference at the E14 position between PcMSP-119 and PvMSP-119 was critical for allowing cross-species protection seen in this study. The A38 corresponding residue in PyMSP-119 is G; an inhibitory MAb, G3, failed to bind PyMSP-119 from isolates that had a nonconservative change, N, at this position (5). We speculate that a conserved amino acid difference, A38-G, at this position may also be critical for allowing cross-protection in this study. The radical charge difference K67-S between PcMSP-119 and PvMSP-119 was similar to the R-G difference seen between the 3D7 and FVO alleles of P. falciparum. The lack of species specificity observed in this trial, despite the presence of a nonconserved change at this position, suggests that allele-specific protective response between the 3D7 and FVO strains of PfMSP-1 (24) may not be a result of a radical substitution at this position. The difference E69-K between PcMSP-119 and PvMSP-119 is also the only known polymorphism reported in the field isolates of PvMSP-119 (35); based on the observations presented here, we suggest that this allelic polymorphism in the field isolates of P. vivax may be immunologically silent. Residues corresponding to all other observed differences between PvMSP-119 and PcMSP-119 were not reported to be targets of inhibitory PfMSP-1 or PyMSP-1 MAbs in the published literature. The three-dimensional crystal structure of PcMSP19 shows that all but one amino acid difference between PcMSP119 and PvMSP119 are concentrated at one end within the EGF domain, and, interestingly, the residues shown to be targets of PfMSP119 inhibitory MAbs 12.8 and 12.10 are not located at the same end of the molecule (8, 40). While the above analysis is an indirect approach towards understanding the lack of species specificity observed in this trial, direct experiments confirming these inferences are required to tie this data to the existing views on the structure-function of MSP-119.
The adjuvant Montanide ISA720 has been used in several human vaccine trials (42, 27). PvMSP-142-Montanide ISA720 adjuvant combination has been shown to induce both B- and T-cell responses in mice (14, 36). We have also reported the induction of high-titer invasion inhibitory antibodies to Montanide ISA720 formulated Apical Membrane Antigen-1 of P. falciparum (15). Encouraged by these findings, we chose to proceed with Montanide ISA720 for the purpose of this trial. As expected, high levels of MSP-1-specific antibodies were induced in all the animals in each of the four MSP-142-vaccinated groups. This immune response, while not resulting in sterile protection, did significantly lower parasitemia in both the homologous and heterologous challenge groups compared to the adjuvant control group. Immunization with MSP-1 along with immunostimulators permissible for human use are known to induce a protective response in the rodent models of malaria (28). MSP-1-induced protection in the new world monkey model has, however, been shown to require the use of Freunds Complete Adjuvant (FCA) (24, 12, 37). Indeed, a high degree of sterile protection has been previously reported using baculovirus-produced P. cynomolgi MSP-142; however, that vaccine was also formulated with FCA (33). The results of this study clearly indicate the ability of MSP-142 with Montanide ISA720 to induce a partially protective immune response in primates using an adjuvant acceptable for human use instead of FCA. Similar partial protection has been reported in two other challenge trials using PvMSP-119 and a nonionic block copolymer in the splenetomized Saimiri monkey model against P. vivax challenge (45, 11).
In summary, we report that E. coli-derived soluble and refolded MSP-142 antigens of P. vivax and P. cynomolgi produced using scalable production processes had comparable immunogenicity in rhesus monkeys. We also show that both P. cynomolgi and P. vivax MSP-142 protein vaccination using Montanide ISA720 adjuvant can induce a partially protective response against a P. cynomolgi challenge. A vaccination induced reduction in parasite burden similar to that reported in this study may be sufficient in reducing mortality in the endemic areas while still not impeding the development of age-related natural immunity against malaria. An Origami strain of E. coli that is sensitive to kanamycin has recently become available. With the knowledge that the soluble and refolded products are immunologically equivalent, we are in the process of comparing the reproducibility of the production process, cost of manufacture, and quality of the final product as criteria to down-select and advance into GMP production of a PvMSP-142-based vaccine.
We thank C. M. Gupta, Director, CDRI, for his support in this international collaborative effort. We thank Shirley Longacre, Pasteur Institute, France, for P. vivax-specific monoclonal and polyclonal antibodies, and Evelina Angov from WRAIR for the expression plasmid used in the study and Jeff Lyon and COL Chris Ockenhouse and for their critical suggestions. We thank Craig Morrissette, WRAIR, for statistical analysis.
The views expressed here are those of the authors and should not be construed to represent those of the U.S. Department of the Army or the U.S. Department of Defense.
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»