The blood samples used in this study were collected from residents in 3 regions on April 2019. A total of 777 blood samples from Gimpo-si (n=319), Paju-si (n=308), and Yeoncheon-gun (n=150) from local health centers, where malaria cases are diagnosed in a relatively higher number compared to other endemic areas from nation-wide surveillance of 2005 through 2018, were obtained (Fig. 1A). In addition, we used serum samples from Haman-gun (n=80), Gyeongsangnam-do as vivax nonendemic areas for comparison of antibody response with those of endemic areas. Blood samples confirmed to be either vivax positive or negative by testing with a rapid diagnosis test kit and microscopy were collected in endemic areas of Korea. The blood samples collected in BD Vacutainer EDTA tubes (Becton Dickinson, Franklin Lakes, NewJersey, USA) were delivered to Kangwon National University; confirmation by blood films for microscopic examination and RDT was performed. The plasma and packed red blood cells were separated and stored at −80°C for protein array and genomic DNA isolation, respectively. The Institutional Review Board approved this study at Kangwon National University (no. 2019-03-002-001).

The pooling method for genomic DNA extraction was used, as described in our previous study [9]. Briefly, 20 μl each of 10 different samples was pooled for extraction of genomic DNA using an AccuPrep Genomic DNA Extraction Kit (Bioneer Corp., Daejeon, Korea) in accordance with the manufacturer’s instructions. Nested PCR was performed as described in a previous report [10]. Briefly, for the primary reaction, 4 μl of the eluted gDNA was mixed in a 20 μl reaction mixture using Accupower premix (Bioneer) with rPLU1_F (TCAAAGATTAAGCCATGCAAGTGA) and rPLU5_R (CCTGTTGTTGCCTTAAACTTC) primers. In the second nested PCR, 2 μl of the first amplicon was used as a template under the same 20 μl reaction mixture conditions as described above. rPLU3_F (TTTTTATAAGGATAACTACGGAAAAGCTGT) and rPLU4_R (TACCCGTCATAGCCATGTTAGGCCAATACC) were used for the nested PCR. The PCR conditions were as follows: step 1, 95°C for 5 min; step 2, denaturation at 95°C for 30 sec; step 3, annealing at 55°C for 1 min; step 4, extension at 72°C for 2 min in the first PCR and 30 sec in the second PCR. Steps 2 to 4 were repeated 30 times with the annealing temperature changed to 60°C, and step 4 was performed for 10 min. The amplified products were visualized on 1.5% agarose gels.

CareStart^TM Malaria Pf/PAN (HRP2/pLDH) Ag Combo RDT (AccessBio Inc., Somerset, NewJersey, USA) was used with whole-blood samples from endemic areas according to the manufacturer’s manual. Blood samples from patients with vivax malaria were used as a positive control.

Genomic DNA extracted from the whole-blood sample of vivax malaria patients was used as a template for gene amplification of pvmsp1-19 (GenBank accession no. XM_001614792.1; Fig. 2A). The N-terminal region of pvlsa (XM_001615278.1; Fig. 3A) and full-length pvcsp-vk210 (XM_001613018.1; Fig. 2B) were synthesized by Macrogen (Seoul, Korea). The central repeat domain of pvcsp was specially designed from previous sequence analysis reports (vk210 genotype mostly) of Korean vivax isolates. Segments of the genes were amplified using a high-fidelity KOD-plus Kit (TOYOBO Co., Osaka, Japan) with the following pairs of gene-specific primers: pvmsp1-19 forward (gggcggatatctcgagAACGAGTCCAAGGAAATATTATCCCAG) and reverse (gcggtacccgggatccGTCCGTCATTTTACATTCAGCTTCAGG); pvcsp-vk21 forward (gggcggatatctcgagATGAAGAACTTCATTCTCTTGGCTGTTTCTT) and reverse (gcggtacccgggatccCTAATTGAATAATGCTAGGACTAACAATATGACTAGC); and pvlsa-n forward (gggcggatatctcgagATGAGTTCCTTCTGGAGCTTCATCAAAT) and reverse (gcggtacccgggatccCTAACTCCTCGAGATGTCATCGTTGCT). Small letters indicate the plasmid-derived sequences for in-fusion cloning, and BamHI and XhoI restriction enzyme sites are indicated as italicized and underlined letters. The amplicons were purified and cloned into the pEU-E01-His-Tev-N2 plasmid vector (CellFree Sciences, Matsuyama, Japan) designed for the wheat germ cell-free protein expression system by the In-fusion cloning method (Takara Bio USA Inc., Mountain View, California, USA). The sequences of cloned plasmids were confirmed. The recombinant proteins were synthesized by the wheat germ cell-free protein expression system using the bilayer translation reaction method, as described previously [11]. The expressed recombinant proteins were confirmed by western blot analysis under reducing conditions using an anti-His-tagged antibody (Qaiagen, Hilden, Germany).

Amine-coated glass slides were prepared as described in our previous study [12]. The crude recombinant protein antigens (1 μl) were spotted onto amine-coated glass slides in duplicate and incubated at 37°C for 2 hr. After blocking with 5% BSA in PBS containing 0.1% Tween-20, the recombinant protein-coated slides were reacted with each of the individual serum samples in duplicate. Fluorescence was detected with an Alexa Fluor 546-conjugated goat anti-human IgG secondary antibody (Invitrogen, Carlsbad, California, USA) and an Innopsys IS300 scanner (InnoScan, Carbonne, France). Mean fluorescence intensity (MFI) plus 2 standard deviations (2SD) was used as the cut-off value. The MFI was normalized by dividing the MFI by the cut-off value, and the normalized MFI from patients was subtracted from the normalized MFI from healthy serum.

Statistical analysis was performed using the Statistical Package for the Social Sciences (SPSS) version 16.0 (SPSS Inc., Chicago, Illinois, USA). The unpaired t-test was used for nonparametric independent 2-group comparisons between the healthy individual group and the vivax malaria patient group. Differences with P<0.05 were regarded as statistically significant. MFI values were plotted using GraphPad Software version 5.0 (GraphPad Software, La Jolla, California, USA).

In this study, a passive case detection method was carried out by local health authorities, and a total of 777 residents from 3 areas, Gimpo-si (n=319; age 70.8 on average and 93-31 on range), Paju-si (n=308; 70.5 and 95-23), and Yeoncheon-gun (n=153; 59.2 and 90-20), were recruited for the detection of clinical malaria (Fig. 1). All collected blood samples were confirmed by blood smear and RDT (data not shown) as well as the nested PCR method (data not shown) for malaria infection. Consequently, there was no positive case based on the 3 methods; however, the amplified gene was observed in the positive control sample.

To analyze serological positivity for P. vivax proteins, 3 antigens, P. vivax merozoite surface protein 1 (PvMSP1), P. vivax circumsporozoite protein (PvCSP), and P. vivax liver-stage antigen (PvLSA), were investigated (Fig. 2A, B). The full-length C-terminal region of PvMSP1 (aa 1622-1706), including a combination of the central repeat region sequence of PvCSP (aa 1-377) and the N-terminal region of PvLSA (aa 1-436), was used for recombinant protein expression. The recombinant proteins were successfully expressed by the wheat germ cell-free expression system, and the proteins were detected by western blot analysis (Fig. 2C).

The protein array method was carried out to analyze the seropositivity of the plasma samples collected from 777 participants as well as 16 vivax patients and healthy individuals as controls. The serum samples were diluted (1:25) and reacted with 3 P. vivax antigens bound on array chips. For the control, PvLSA-N showed higher positivity to patient sera (87.5%) than PvCSP-VK210 (62.5%) and PvMSP1-19 (68.7%). However, one healthy serum sample was positive for PvLSA-N but not for the others (Fig. 3; Table 1). Among plasma collected from residents of 3 local surveyed areas, the seropositivities of PvLSA-N (20.2±6.8%) and PvCSP-VK210 (15.1±6.3%) were higher than 18.8% and, 16.3%, respectively, in Haman residents but not in cases of antibody titers against PvMSP1-19 (Fig. 3; Table 2).

We also issued questionnaires about previous exposure to malaria infection to the residents of the surveyed areas. Gimpo-si and Yeoncheon-gun had 21 and 4 cases of malaria history, respectively, with exposure to vivax infection from a few years to 50–60 years previously (Fig. 1B). Seropositivity among the cases with a history was compared with those without a history and samples from nonendemic areas. Interestingly, there was a significant difference between history and nonhistory from endemic areas as well as the nonendemic area (Haman-gun) for PvLSA-N, but this was not observed for the other antigens (Fig. 4A–C). In addition, seropositivity to PvLSA-N among history samples was much higher than that of nonhistory and nonendemic area samples (Fig. 4D), suggesting that PvLSA-N is a protein marker to detect the specific antibody remaining in people who have been infected with the malaria parasite.