Expression and Evaluation of Recombinant Plasmodium knowlesi Merozoite Surface Protein-3 (MSP-3) for Detection of Human Malaria

Malaria remains a major health threat in many parts of the globe and causes high mortality and morbidity with 214 million cases of malaria occurring globally in 2015. Recent studies have outlined potential diagnostic markers and vaccine candidates one of which is the merozoite surface protein (MSP)-3. In this study, novel recombinant Plasmodium knowlesi MSP-3 was cloned, expressed and purified in an Escherichia coli system. Subsequently, the recombinant protein was evaluated for its sensitivity and specificity. The recombinant pkMSP-3 protein reacted with sera from patients with P. knowlesi infection in both Western blot (61%) and ELISA (100%). Specificity-wise, pkMSP-3 did not react with healthy donor sera in either assay and only reacted with a few non-malarial parasitic patient sera in the ELISA assay (3 of 49). In conclusion, sensitivity and specificity of pkMSP-3 was found to be high in the ELISA and Western Blot assay and thus utilising both assays in tandem would provide the best sero-diagnostic result for P. knowlesi infection.


Introduction
Malaria remains a major health threat in many parts of the globe and causes high mortality and morbidity especially in sub-Saharan Africa (90%), according to the World Malaria Report (2015). The same report highlights 214 million cases of malaria occurring globally in 2015 which led to 438000 deaths. The four known species of human malaria parasites are Plasmodium falciparum, Plasmodium vivax, Plasmodium malariae, and Plasmodium ovale. Plasmodium knowlesi has been recognized as the fifth malaria causing species in humans and has been reported to cause a large number of human infections in South East Asia since the discovery of a large number of infected patients in Borneo, Malaysia in 2004 [1]. Subsequently, increasing numbers of naturally acquired P. knowlesi infections were found in other Asian countries such as Myanmar, Thailand, Vietnam, the Philippines and Singapore [2][3][4][5][6][7]. On the other hand, in genes [1]. Ethical approval for this study was obtained from the University of Malaya Medical Centre Ethic Committee (MEC Ref. No. 817.18) and informed verbal consent from the donor or the next of kin was obtained for use of these samples in screening. Written consent was found to be unnecessary as verbal consent would be more sufficient and convenient for the purpose of this study. This is especially so in cases of illiterate patients or where patients only understand a specific language or dialect. In this case, it would be easier to get a translator to convey key information about a project as opposed to having multiple translated copies of written consent forms instead. Patient details were noted down for our personal recordkeeping. This consent procedure was approved by the University of Malaya Medical Centre Ethic Committee. Blood samples from 49 non-malarial parasitic infected patients was also collected and was confirmed by commercial ELISA kits. The patient serum in this study was characterized as follows: P. knowlesi human malaria; non-P. knowlesi human malaria (P. falciparum, P. vivax, P. ovale); non-malarial parasitic infection (filariasis, amebiasis, cysticercosis, toxoplasmosis, toxocariasis); and healthy donor serum which was obtained from 56 healthy individuals.

DNA extraction and construction of recombinant expression plasmid
Plasmodium knowlesi DNA was extracted from patient blood sample using a commercial blood extraction kit (Qiagen, Hilden, Germany). A set of primers were designed to allow for PCR product amplification as follows, forward primer MSP3B-F: CCGGATCCATGAAA CGCATTTGG and reverse primer MSP3B-R: CGCGGATCCTTACCAGAATTTCAA. Primers were designed based on the P. knowlesi H strain MSP-3 putative nucleotide sequence, PKH_103010 (Genbank accession no. XM_002259752.1). BamH1 restriction sites were introduced into both the forward and reverse sites in order to facilitate cloning. A single step PCR reaction was carried out according to the following protocol: an initial denaturation step at 95°C for 5 min followed by 30 cycles of 94°C for 1 min, 53°C for 1 min and 72°C for 1 min 30 seconds. Finally, an elongation step at 72°C for 10 min was added to the cycle. The PCR product was then purified and cloned into the pGEM-T1 TA cloning vector (Promega Corp, USA). The PCR product in the cloning vector was then sent for commercial sequencing to confirm the integrity of the gene by aligning the amplified PCR product nucleotide and amino acid sequences with the deposited putative pkmsp3 gene, PKH_103010 (XM_002259752.1) in Genbank. Furthermore, the amplified pkmsp3 sequence was also aligned with MSP-3 orthologues from other Plasmodium species to identify percent identity of the nucleotide and amino acid sequences as well as percent of amino acid matches using the Pairwise Sequence Alignment and Clustal Omega tool provided by the European Bioinformatics Institute (EMBL-EBI) Web Services interface (http://www.ebi.ac.uk/Tools/psa/) [27]. Percent of amino acid matches was calculated taking into account fully conserved residues as well as groups of residues with strongly similar properties with a score of greater than 0.5 in the Gonnet PAM 250 matrix [28]. Sequences from P. knowlesi (Genbank accession no: KT900798-KT900802), P. vivax (Genbank accession no: KC907580.1-KC907583.1, KC907572.1), P. cynomolgi (Genbank accession no: KC907501.1-KC907504.1, KC907501.6) and P. falciparum (Genbank accession no: AM161584.1, AM161586.1, AM161615.1, AM161618.1, AM161642.1) were retrieved from Genbank and used to construct a phylogenic tree using the neighbour-joining method in MEGA6 computational program [29]. Bootstrap proportions were used to assess the robustness of the tree with 1000 bootstrap replicates. The putative size of P. knowlesi MSP-3 was calculated based on the pkmsp3 nucleotide sequence deposited in Genbank utilizing Generunner software version 4.0.9.2 (Hastings Software Inc., Hastings, NY, USA; http://www.generunner. net/). After confirmation of the sequences the pkmsp3 recombinant plasmid was then digested with the BamH1 restriction enzyme and cloned into the T7 promoter-based pRSET A vector; which allows for expression of dually polyhistidine (His)-and Xpress epitope-tagged recombinant proteins in E. coli (Invitrogen, USA). Successfully cloned recombinant pkMSP-3 expression vector was then again sent for sequencing to confirm the presence, integrity, and orientation of the pkmsp3 gene before expression. Transformation into maintenance host E. coli TOP10F' strain was performed for propagation and maintenance of the plasmid. Before expression, the plasmid was transformed into expression host E. coli BL21 (DE3) pLysS strain.

Expression of recombinant pkMSP-3
A single recombinant BL21 (DE3) pLysS clone was inoculated and propagated overnight in Luria Bertani (LB) broth containing ampicillin (100 μg/mL) and chloramphenicol (35 μg/mL) at 37°C with shaking. The overnight culture was then further diluted with LB broth to an optical density (OD) at 600 nm of 0.1 and the culture was then allowed to grow to an OD of 0.4 to 0.6. This was followed by induction of overexpression of the pkMSP-3 protein by the addition of 1 mM isopropyl-1-thio-D-galactopyranoside (IPTG) for 2 hours. Cells were then pelleted at 5000 rpm for 10 min. The supernatant collected was discarded and the cell pellet was then resuspended in 5 mL of Bug Buster reagent (Novagen, USA) per gram of wet cell pellet. One μl of Benzonase (50 mM Tris-HCL, pH 8.0, 20 mM NaCL, and 2 mM MgCl 2 in 50% glycerol) was added per mL of Bug Buster reagent used. The solution containing the cell pellet was allowed to incubate on a rocker till the cell pellet was fully lysed. Insoluble cell debris was separated through centrifugation at 16000 x g for 15 min at 4°C. The insoluble cell pellet was then further processed using 200 μg/mL lyzosyme in the same volume of Bug Buster reagent that had been used during the initial resuspension of the cell pellet. The suspension in lysozyme was then allowed to incubate at room temperature for 10 min and was then further centrifuged at 16000 x g for 15 min at 4°C. The pellet was then washed with 1:10 diluted Bug Buster reagent and the washing was repeated 3 times. The final inclusion body pellet was then re-suspended in Phosphate Buffered Saline (PBS), and a fraction was used for sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) and Western blotting.

Optimization of recombinant pkMSP-3 expression conditions
Optimum expression conditions for maximizing the protein yield was studied by monitoring expression levels of a number of variables including concentration of IPTG used for induction and collection time after induction. The IPTG concentrations studied were 0.5 mM, 0.75 mM, 1.0 mM, and 2.0 mM (S1 and S3 Figs). Cell culture of 1 mL in volume was collected before induction and every hour after induction up to 4 hours. Aside from this, expression levels with induction at different OD was also studied in an effort to identify optimal expression conditions (S2 and S4 Figs). Cells were harvested and analysed by SDS-PAGE and Western blotting.

SDS-PAGE, Coomassie brilliant blue staining and Western blot
Recombinant pkMSP-3 were resolved by 12% SDS-PAGE and stained with Coomassie brilliant blue (Bio-Rad, USA). The separated proteins were also electrophorectically transferred onto polyvinylidene difluoride (PVDF) membranes (Bio-Rad, USA) with protein amounts ranging between 50 to 300 ng and blocked overnight in Tris Buffered Saline (TBS) containing 5% skimmed milk at 4°C. The membranes were washed three times with TBS containing 0.2% Tween-20 (TBS-T) and probed with either anti-Xpress™ antibody (1:5000 dilution) (Invitrogen, USA) or with patient serum (1:200 dilution) with TBS containing 2.5% skimmed milk for two hours with constant shaking at room temperature. After three washes with TBS-T the membrane was treated with biotin-labelled goat anti-mouse IgG (1:2500 dilution) if using anti-Xpress™ antibody; or biotin labelled goat anti-human IgA plus IgM plus IgG (1:2500 dilution) is using human serum (Kierkegaard and Perry Inc., USA) for one hour, followed by streptavidin-alkaline phosphatase (1: 2500 dilution) for one hour. Finally, the membranes were developed by chromogenic nitro-blue tetrazolium/5-bromo-4-chloro-3'-indolyphosphate substrate. The colour was allowed to develop at room temperature in the dark.

Purification of recombinant pkMSP-3
Cell pellets from 50 ml of recombinant pkMSP-3 protein cultures were re-suspended in 8 mL of 6 M urea and added to a ProBond column (Invitrogen, USA) filled with 2 mL of Nickel-NTA resin. The column was then loaded into a MACSmix™ Tube Rotator (Miltenyi Biotec, Germany), and the His-tagged protein was then allowed to bind to the Ni-NTA resin for 4 hours at 4°C. The column is then moved onto a stand and the resin was allowed to settle to the bottom of the column. The supernatant was then aspirated gently and the column was washed with 8 mL of 6 M urea. Washing was done by inverting the tube 10 times to allow the resin to mix with the urea and the column was then returned to the stand to allow the resin to settle. The supernatant is once again aspirated and replaced with 4 M urea. Washing was then repeated and this step was further carried out using 2 M urea. Finally, the protein was eluted with 5 mL of 1 M urea and 250 mM imidazole. The purity of recombinant pkMSP-3 was determined via SDS-PAGE.

Refolding of pkMSP-3
Refolding of the pkMSP-3 protein was carried out via dialysis in PBS as described by Chao et al [30]. The purified polypeptides at a concentration of approximately 0.8 mg/mL were transferred into dialysis cassettes with a molecular weight cut-off point of 10 kDa and dialyzed against PBS for up to 16 hours with 2-4 buffer changes during the duration of dialysis. Usually one dialysis cassette, with 20 mL each in a 1,000-mL beaker was dialyzed against buffer at a ratio of 1:35 of sample volume to dialysis buffer. All dialysis procedures were conducted at 4°C. The dialysis was left overnight in large excess of PBS at 4°C to remove traces of urea. The dialyzed protein was then quantified using a Bradford Assay Kit with Bovine Serum Albumin (BSA) as standard (Bio-Rad, USA) before subsequent downstream experiments. The pkMSP-3 dialyzed protein was evaluated for native refolding and native confirmation via a reducing assay using β-mercaptoethanol as a reducing agent. Recombinant pkMSP-3 protein was boiled in reducing and non-reducing sample buffer with and without β-mercaptoethanol respectively. Samples were then run on an SDS-PAGE gel, stained with Coomassie blue, and native confirmation was evaluated via a difference in mobility in the gel (S5 Fig). concentration of 500 ng of pkMSP-3 protein per well. The wells were then washed with PBS containing 0.1% Tween-20 a total of three times. Blocking buffer (1% BSA in PBS) was added to each well and the plate was left to incubate for 2 hours at 37°C. The wells were further washed 3 times and then incubated with patient serum in a 1:80 dilution of 1% BSA/PBS for an hour at 37°C. Consequently, the wells were then washed 5 times to remove any patient serum left in the well and 1:2500 diluted secondary antibody consisting of peroxidase-labelled goat anti-human IgG (Biorad, USA) was added to each well and the plate was incubated for 1 hour at 37°C. The wells were washed 5 times for a final time after incubation and incubated with 3, 3', 5, 5'-tetramethyl benzidine (TMB) (Amresco, Solon, OH) for 30 min in the dark to allow colour to develop. The reaction was stopped with 2 N H 2 SO 4 . The OD was read at 450 nm and determined by subtracting the OD of the blank containing TMB from the mean OD of the sample. All samples were run in triplicates and the mean OD of each sample was used for further analysis. The cut-off value was calculated as the M N + 2σ of the healthy donor serum group where M N and σ are the mean OD and standard deviation respectively. Samples with a mean OD greater than the cut-off point were considered positive and vice-versa.

Statistical analysis
The sensitivity and specificity of both Western blotting and ELISA for the detection of malarial infection was calculated by using the following formulae: TP/(TP+FN) x 100% for sensitivity and TN/(TN+FP) x 100% for specificity where TP = true positive, TN = true negative, FP = false positive, FN = false negative for malarial infection. The gold standard used for comparison in the analysis of sensitivity and specificity was microscopic analysis of thick and thin smears as well as PCR.

Cloning and expression of recombinant pkMSP-3
The P. knowlesi MSP-3 protein was successfully amplified from genomic DNA by PCR and had an expected size of approximately 1077 base pairs. The gene was cloned into the pRSET A expression vector and transformed into BL21 (DE3) pLysS host cell. Recombinant plasmids were isolated and sequenced to confirm the identity and integrity of the cloned gene before expression. The nucleotide and amino acid sequence percent identity of the pkmsp3 gene was calculated relative to the msp3 of closely related Plasmodium spp. as seen in Table 1. Plasmodium knowlesi msp3 had nucleotide and amino identities of >50% with both P. vivax and P. cynomolgi msp3. A phylogenetic tree seen in Fig 1 was also constructed using sequenced pkmsp3 genes and P. vivax, P. cynomolgi and P. falciparum msp3 genes retrieved from Genbank. It was observed that the sequences clustered into respective Plasmodium groups with pkmsp3 genes branching out closer to the P. vivax and P. cynomolgi msp3 groups as compared to P. falciparum which branches out further away. The high bootstrap value validates the branching that is observed in the phylogenetic tree. The recombinant pkMSP-3 clone was able to successfully exhibit a high level of expression of the pkMSP-3 protein with an approximate size of 34 kDa indicated by the thick band visible in Fig 2 which is not seen in the non-recombinant plasmid control.
Determination of optimal expression conditions was carried out before large scale expression and purification to maximize the yield of protein. A time-course experiment was carried out to determine the optimum length of expression. In tandem with that, different concentrations of IPTG were used (0.5 mM, 0.75 mM, 1.0 mM and 2.0 mM). We found that differing concentrations of IPTG did not yield a significant difference in protein as observed on a Coomassie-stained gel at the same time point. We also observed no difference in protein yield between expression times of 2 hours and 4 hours. Thus, we concluded that the best expression conditions for maximum yield of recombinant pkMSP-3 would be for 2 hours at 37°C with 1.0 mM IPTG used for induction and constant shaking at 250 rpm.

Purification of recombinant pkMSP-3
Recombinant pkMSP-3 was successfully purified under denaturing conditions using a Ni-NTA column. Purified products were observed as a single band at 34 kDa in both a Coomassiestained SDS-PAGE gel (Fig 2) and a Western blot using anti-Xpress as the primary antibody (Fig 3). Subsequent, quantification of pkMSP-3 protein indicated a concentration of protein that ranged from 0.7-0.8 mg/mL for a 400 mL culture.

Western blot analysis of recombinant pkMSP-3 with human serum samples
The purified pkMSP-3 was evaluated by Western Blotting using serum samples from patients with P. knowlesi human malaria, non-P. knowlesi human malaria, non-malarial parasitic infections and healthy donors (Fig 4). The recombinant pkMSP-3 reacted with 61.0% of knowlesi malaria serum samples (25/41) and 14.8% of non-knowlesi malaria serum samples which

ELISA analysis of recombinant pkMSP-3 with human serum samples
The same serum samples used in Western blot were also used in ELISA. For ELISA, 10 μg/mL of recombinant pkMSP-3 was coated in each well and a 1:80 dilution of patient sera was used per patient. The recombinant pkMSP-3 protein reacted with all 41 knowlesi patient serum giving it 100% sensitivity to P. knowlesi patient sera. It also reacted with 70.4% or 19 of 27 nonknowlesi malaria samples; predominantly with P. vivax samples where 11 of the 15 P. vivax samples picked up pkMSP-3. Three non-malarial parasitic patient sera were found to interact with pkMSP-3 indicating a specificity of 97.1%. Results of the patient sera screening against recombinant pkMSP-3 in Western blot and ELISA are shown in Table 2. Sensitivity and specificity of the Western blot assay and ELISA assay are shown in Table 3.

Discussion
The P. knowlesi parasite is able to replicate and complete its blood stage life cycle within 24 hours, the shortest of all human Plasmodium, and thus it is necessary to devise a diagnostic tool that is both sensitive and specific to P. knowlesi. Thus, we sought out to evaluate the serodiagnostic potential of P. knowlesi MSP-3; a previously unstudied protein to assess its efficacy. For the purposes of this study we opted to use the E. coli system for production of the recombinant pkMSP-3. This was due to the ease of use, safety, and known genetic properties of E. coli that make it such an appealing expression host. Furthermore, E. coli is well established and is receptive to foreign DNA when manipulated by genetic manipulation methods. It has the advantage of being able to produce large amounts of protein in an efficient and inexpensive manner which increases its appeal in lieu of other alternatives such as yeast or mammalian cell expression.
The use of recombinant protein can be justified by how advantageous it is in terms of how rapid and simple recombinant protein can be expressed and obtained. This is in light of the fact that the difficulties in long-term culture of P. knowlesi coupled with the complexities of extracting and preparing parasite lysate protein make recombinant P. knowlesi protein an ideal alternative for sustainable production of large amounts of pkMSP-3. The recombinant pkMSP-3 protein was recovered as inclusion bodies expressed in E. coli and there are several reasons that may have contributed to this inclusion body formation. It has been shown that expression of genes that have a large evolutionary distance between the native host and expression host may have a higher propensity for inclusion body formation [31]. This may also extend here where a protein found natively in P. knowlesi is expressed in a non-native E. coli host instead. Furthermore, the protein structure of pkMSP-3 with its alanine-rich central domain increases the overall hydrophobicity of the protein which greatly increases the propensity for inclusion body formation [32]. The inclusion bodies purified in this study were denatured using the Bug ii. P. falciparum  Buster protein extraction kit, and reduction in urea concentration though dialysis against PBS was used to refold the denatured protein [30]. Refolded protein was evaluated using a β-mercaptoethanol reducing assay. In our study, no difference of mobility in the gel was observed between samples boiled in reducing or non-reducing buffer. This was also observed in the native non-solubilized protein. This may be attributed to the fact that pkMSP-3 does not possess any cysteine amino acid residues. Thus, native confirmation of the refolded pkMSP-3 protein through a reducing assay would not be suitable. An alternative to this would be a functional assay to prove that protein structure and function has been restored. Our preliminary observations indicate that mice injected with the pkMSP-3 refolded protein yielded antibodies that were able to recognize antigens on the surface of native P. knowlesi parasites in an immunofluorescence (IFA) assay (Unpublished data). Ideally, crystallization or mass-spectrometry studies for protein refolding are the best option, however, this is one of the limitations of this study due to equipment and facility constraints. During screening of various patient sera using pkMSP-3 recombinant protein we found high sensitivity of the protein (>90%) in the ELISA assay but relatively lower sensitivity in Western blot. This may be due to the loss of protein antigenicity during the denaturing electrophoretic preparation of the antigen for Western blot which does not occur in the ELISA assay allowing preservation of conformational epitopes [33,34]. Furthermore, high-molecularweight proteins transfer poorly to nitrocellulose filters in immunoblot, preventing recognition by the antibodies [35]. In our studies it was also observed that some P. knowlesi patient sera that was used for ELISA screening had borderline activity, where the OD absorbance value of the patient sera hovered slightly above the cut-off value and this may have led to discrepancies in the result [24,36]. Another explanation for the non-reactivity of certain malaria-infected patient sera with pkMSP-3 in both assays is that in certain cases the sera was collected from the patient before adequate production of antibodies against the invading malaria parasite has commenced [37]. Specificity of the pkMSP-3 however, was relatively high (>90%) for both assays indicating that this may be a promising and reliable antigen for use in sero-detection of malarial infection.
It was found that particularly in the ELISA assay, pkMSP-3 cross-reacted with other nonknowlesi malaria sera and this could be due to serological cross-reactivity where previous studies have described cross-reactivity could occur among the Plasmodia family [38][39][40]. In addition to that, it was observed that the pkmsp3 gene possessed high nucleotide and amino acid identities to other members of the Plasmodium species including P. vivax (57.4%) and P. cynomolgi (70.4%) ( Table 1). This is also observed in the phylogenetic tree constructed using pkmsp3 gene sequences where pkmsp3 sequences were seen to branch out closer to P. vivax msp3 and P. cynomolgi msp3 sequences (Fig 1) and thus tallies with the nucleotide and amino acid identity data in Table 1. Antibodies generated against a specific protein may often interact with other closely related proteins with equal amino acid identities and this has been observed in other P. knowlesi protein evaluation studies [22,24]. Cross-reactivity in both assays to nonmalarial and non-knowlesi malaria patient sera may also be due to previous exposure of the patients to P. knowlesi infection which is a possibility we have not ruled out. Indeed, it is not unheard of, as a study done by Wipasa et al has demonstrated that antibodies generated against the malaria parasite during infection are able to last in the human body for a period of 5 years or more [41].
It is noteworthy that, as cross-reactivity was mostly seen in the ELISA assay and not in Western blot it also may be attributed to borderline activity of the sera resulting in a false positive result. There are however, several advantages to utilising the ELISA assay over Western blot assays as an optimized ELISA assay uses marginally smaller amounts of serum for each reaction which is useful when dealing with small volume samples. Throughput is also usually higher and results can be obtained faster by ELISA as compared to Western Blot and this is critical in diagnosis. Lastly, ELISA assays have the ability to quantitatively determine the antibody titer of the serum by measuring the OD value of the sera which correlates to antibody titer.

Conclusion
In conclusion, the recombinant pkMSP-3 protein was successfully expressed and purified in E. coli under optimal conditions. Sensitivity was found to be high when evaluated with an ELISA assay and specificity was found to be high when evaluated using Western Blot indicating that utilising both assays in tandem would provide the best sero-diagnostic results for detecting human malarial infection. Cross-reactivity was observed in evaluation of this recombinant protein and as such further evaluation with non-knowlesi MSP-3 proteins would be ideal. As a future study, it would be interesting to compare and evaluate the results obtained in this study with the P. vivax MSP-3 protein in terms of sensitivity and specificity. Further studies including protein crystallization, mass spectrometry and immunocharacterization of this protein should be carried out to assess its potential as a vaccine candidate.