Diagnostic Characteristics of Lactate Dehydrogenase on a Multiplex Assay for Malaria Detection Including the Zoonotic Parasite Plasmodium knowlesi

ABSTRACT. Plasmodium lactate dehydrogenase (pLDH) is a common target in malaria rapid diagnostic tests (RDTs). These commercial antibody capture assays target either Plasmodium falciparum–specific pLDH (PfLDH), P. vivax–specific pLDH (PvLDH), or a conserved epitope in all human malaria pLDH (PanLDH). However, there are no assays specifically targeting P. ovale, P. malariae or zoonotic parasites such as P. knowlesi and P. cynomolgi. A malaria multiplex array, carrying the specific antibody spots for PfLDH, PvLDH, and PanLDH has been previously developed. This study aimed to assess potential cross-reactivity between pLDH from various Plasmodium species and this array. We tested recombinant pLDH proteins, clinical samples for P. vivax, P. falciparum, P. ovale curtisi, and P. malariae; and in vitro cultured P. knowlesi and P. cynomolgi. P. ovale-specific pLDH (PoLDH) and P. malariae-specific pLDH (PmLDH) cross-reacted with the PfLDH and PanLDH spots. Plasmodium Knowlesi-specific pLDH (PkLDH) and P. cynomolgi-specific pLDH (PcLDH) cross-reacted with the PvLDH spot, but only PkLDH was recognized by the PanLDH spot. Plasmodium ovale and P. malariae can be differentiated from P. falciparum by the concentration ratios of PanLDH/PfLDH, which had mean (range) values of 4.56 (4.07–5.16) and 4.56 (3.43–6.54), respectively, whereas P. falciparum had a lower ratio of 1.12 (0.56–2.61). Plasmodium knowlesi had a similar PanLDH/PvLDH ratio value, with P. vivax having a mean value of 2.24 (1.37–2.79). The cross-reactivity pattern of pLDH can be a useful predictor to differentiate certain Plasmodium species. Cross-reactivity of the pLDH bands in RDTs requires further investigation.


INTRODUCTION
Global malaria programs have been focused largely on Plasmodium falciparum in sub-Saharan Africa and to a lesser extent P. vivax in Asia and South America. 1 However, the success of malaria elimination requires stopping transmission of all human-infecting Plasmodium species-such as P. falciparum, P. vivax, P. malariae, and P. ovale spp. (P. ovale curtisi of classic type and P. ovale wallikeri of variant type)-that cause natural human malaria via parasite speciation and treatment of infection. [2][3][4] P. malariae and P. ovale spp. are prevalent in Africa and other malariaendemic regions but receive little attention because of clinically mild symptoms. 5,6 In addition, malaria can be caused in humans by simian parasites such as P. knowlesi and P. cynomolgi. P. knowlesi, commonly found in natural hosts including long-tailed macaques (Macaca fascicularis) and pig-tailed macaques (M. nemestrina) in Southeast Asian countries, has been documented in humans as a result of increased human-macaque contacts as a result of habitat disruption and destruction by extensive population growth and deforestation in the areas where natural hosts and the anopheline mosquito vectors carrying P. knowlesi exist. 7,8 P. cynomolgi, often detected with other simian parasites such as P. inui, P. coatneyi, or P. fieldi in the macaque population in Southeast Asia, is a close relative of P. vivax. 9,10 Natural human infection with P. cynomolgi has been reported in Southeast Asia. [11][12][13] Severe malarial symptoms and even death have been described with human P. knowlesi infection. [14][15][16] P. knowlesi has a relatively short asexual life cycle of approximately 24 hours, compared with approximately 48 hours for P. falciparum, P. vivax, and P. ovale, and 72 hours for P. malariae. 17 Consequently, P. knowlesi infection can rapidly replicate to reach a high parasite density and progress to severe malaria similar to that from P. falciparum if there is no prompt medical treatment. 18,19 Although cases of P. knowlesi (misdiagnosed as P. malariae) may have been prevalent at low levels for many decades, a relative surge has arisen as a result of a dramatic reduction in prevalence of P. vivax and P. falciparum over the past decade. 20,21 Identification of P. knowlesi by microscopy is unreliable largely because of its similar morphological features with P. falciparum at the early trophozoite stage and to P. malariae at the other stages, notably trophozoite band forms. 22,23 In addition to the limitations of microscopy to accurately identify P. knowlesi, there are no rapid diagnostic tests (RDTs) specifically designed for P. knowlesi identification. Currently available malaria RDTs use several key biomarkers to differentiate between species: P. falciparum-specific histidine-rich protein 2 (HRP2) or/and lactate dehydrogenase (PfLDH) for P. falciparum, and P. vivax-specific lactate dehydrogenase (PvLDH) for P. vivax. Some malaria RDTs detect pan-malarial biomarkers such as the pan-Plasmodium lactate dehydrogenase (pLDH) or Plasmodium aldolase (pAldo) that can, in principle, detect but not differentiate all Plasmodium species. 24 Several studies demonstrated a risk of cross-reactivity in commercial malaria RDTs (OptiMAL-IT V R [Bio-Rad, Hercules, CA], Entebe Malaria Cassette, Binax-NOW TM Malaria [Abbott, Abbott Park, IL], and Paramax-3 TM [Tulip Diagnostics Ltd, Goa, India]) between parasites of two major Plasmodium species and P. knowlesi parasites, and poor sensitivity in cases of low parasitemia with P. knowlesi infection, revealing the need for caution in data interpretation if confirmatory testing is not conducted. 25,26 For more accurate identification of P. knowlesi, several groups reported the stronger performance of molecular techniques, such as nested polymerase chain reaction (PCR) assay, loop-mediated isothermal amplification (LAMP), and realtime PCR assays. [27][28][29][30] However, these molecular tests are not simple to use or cost-effective. Similarly, there are no P. ovale-and P. malariae-specific RDTs, also leading to the underreporting of these more widely spread forms of human malaria. 31,32 Interestingly, an immunocapture assay has been used to evaluate specificity of monoclonal antibodies against the pLDH isoforms from Plasmodium species. 33,34 A similar reactivity pattern with the pLDH proteins of P. cynomolgi (PcLDH) and P. vivax (PvLDH) indicated the close similarity of pLDH isoforms, whereas the absence of reactivity of the P. falciparum-specific antibodies with other simian parasites except P. knowlesi suggested the possible use of these antibodies to distinguish P. knowlesi from the other simian parasites.
The pLDH is the primary target analyte for P. vivax diagnosis with RDTs. 35 With the emergence of hrp2/hrp3 gene deletions in P. falciparum infections leading to false negative RDT results, pLDH is increasingly a key biomarker for RDTbased diagnosis of malaria. 36,37 It is important to understand the reactivity of different Plasmodium species to the different pLDH assays used in RDTs-namely P. falciparum pLDH, P. vivax pLDH, and PanLDH. In previous studies, we described the development and validation of a laboratorybased multiplex platform, the Human Malaria Array (5-Plex; Quansys Biosciences, Logan, UT), that can quantify HRP2 and pLDH from P. falciparum and P. vivax. 38,39 This multiplex immunoassay, if applied to characterize clinical as well as in vitro cultured samples from these closely related but lesscommon Plasmodium species, may increase our understanding of the performance of current RDTs in these species.
Here, analysis of cross-reactivity of this reference assay was investigated with clinical samples of P. malariae and P. ovale and by using culture-adapted strains of both P. cynomolgi and P. knowlesi. 40,41 We also explored the indicators that can differentiate these less-common parasites from P. falciparum and P. vivax parasites, and the time-course expression of P. knowlesi-specific pLDH (PkLDH) protein in culture by a multiplex reference assay.

MATERIALS AND METHODS
Clinical samples. A total of 116 clinical blood samples (29 P. falciparum, 62 P. vivax, 17 P. malariae, and 8 P. ovale) from febrile patients with malaria mono-infection were obtained from the Foundation for Innovative New Diagnostics (FIND) Specimen Bank. Nested PCR with microscopy was used to identify the species of human malaria parasites in the samples. Parasite density in clinical samples was confirmed by either microscopy or quantitative PCR (qPCR) methods.
In vitro Plasmodium parasite culture. The P. knowlesi strain A1-H.1 was cultured, as previously described. 41 Parasites were grown in RPMI-1640 (Sigma-Aldrich, St. Louis, MO), supplemented with 2 g/L dextrose anhydrous, 0.05 g/L hypoxanthine, 0.5% AlbuMAX II (Invitrogen, Waltham, MA), and 10% (v/v) horse serum (Invitrogen). Human red blood cells (RBCs) were obtained from Duffy antigen-positive individuals from PlasmaLab (Everett, WA), and washed RBCs were added to 2% hematocrit in the culture media. The culture was maintained at 37 C with a hypoxic gas mixture of 90% N 2 , 5% CO 2 , and 5% O 2 . For time-course studies, high parasite density cultures were synchronized in three replicates twice using centrifugation with Nycodenz AG V R (Axisshield Diagnostics Ltd, Dundee, Scotland) gradient solution 4 hours apart, retaining the upper brown portion in the first centrifugation and the lower RBC portion for the second centrifugation, to achieve the highest synchronization rate. After synchronization, synchrony was confirmed, and parasite density in the culture was determined by staining thinfilm smear using a Hemacolor V R staining kit (HARLECO, Sigma Aldrich, St. Louis, MO) and counting parasites in a minimum of 2,000 RBCs using a microscope. Following the second synchronization, the culture was incubated and harvested at different time points. For kinetics of PkLDH, infected RBC (iRBC) cultures that were incubated in the culture tubes in three replicates were harvested at 0, 1, 2, 6, 19, 21, 25, 27, 28, 29, and 30 hours. After centrifugation, the cell culture supernatant was collected, and the remaining iRBC pellets were washed with phosphate-buffered saline (PBS) and resuspended in PBS at the original culture volume. The total number of iRBCs was counted using an Accuri TM C6-UV flow cytometer (BD Biosciences, San Jose, CA) Plasmodium cynomolgi parasites were cultured in complete RPMI media supplemented with GlutaMAX TM (Gibco, Waltham, MA) and 200 mM hypoxanthine (Calbiochem, San Diego, CA) as described previously. 40 Rhesus monkey blood (BioIVT, Westbury, NY) was added to 5% hematocrit, and parasites in a plate were cultured at 37 C in a hypoxic gas mixture of 90% N 2 , 5% CO 2 , and 5% O 2 . Three lots of in vitro culture as both iRBC pellet and culture supernatant were prepared on the second day after synchronization and frozen prior to shipment to PATH (Seattle, WA) for the antigen quantification.
Production of Plasmodium species-specific pLDH proteins. PATH collaborated with the University of Queensland Protein Expression Facility (University of Queensland, St. Lucia, Australia) to develop Plasmodium species-specific pLDH proteins and human LDH protein for use as reference proteins. Based on the DNA sequence information available from GenBank, the full-length ldh genes specific for P. falciparum, P. vivax, P. ovale curtisi, and P. malariae and the human full-length ldh gene were synthesized and cloned into the prokaryotic expression vector pOPIN-NHis3C to express in Rosetta TM (DE3)pLysS competent cells (Novagen, Madison, WI). Bacterial culture was performed in 1 L TB media (Novagen) containing antibiotics at 30 C for around 24 hours. Culture was harvested by centrifugation at 5,000 3 g for 30 minutes at 4 C and resuspended in lysis buffer. After lysing using sonication, the lysate was centrifuged at 20,000 3 g for 30 minutes at 4 C, and the supernatant representing the soluble fraction was isolated. The soluble fraction was subjected to an immobilized metal affinity chromatography using a HisTrap TM Fast Flow column (GE Healthcare, Chicago, IL) followed by size-exclusion chromatography (SEC) 4 using a HiLoad V R 26/600 Superdex V R 200 column (GE Healthcare, Chicago, IL). The SEC elution fractions were analyzed by resolving on 12% SDS-PAGE. The SEC elution fractions of interest were pooled and concentrated using an Amicon V R centrifuge filter (Sigma-Aldrich, St. Louis, MO). The sample was filtered through a 0.22 mm filter and stored at 280 C.
The amount and purity of His-tag-containing recombinant pLDH and human LDH proteins were estimated by absorbance at 280 nm using NanoDrop TM (Thermo Fisher Scientific, Waltham, MA) spectrophotometer and SDS-PAGE, respectively. All recombinant pLDH and human LDH proteins showed more than 95% purity (Supplemental Figure 1) and were kept at 280 C for long-term storage until future use.
Quantification of malaria antigens. The presence and quantification of PfLDH, PvLDH, and PanLDH in iRBC pellets, culture supernatant, clinical samples, and samples spiked with recombinant pLDH proteins were tested using the Q-Plex TM Human Malaria Array (Quansys Biosciences, Logan, UT; hereafter the 5-Plex) to target five different analytes-HRP2, PfLDH, PvLDH, PanLDH, and C-reactive protein (CRP)-according to the manufacturer's instructions. 39 These analytes are identified using pairs of specific antibodies in which capture antibodies are printed as defined spots in a geometric planar array on the surface of each well and a mix of detection antibodies are used to identify the analytes captured on the defined spots. For the pLDH assays, capture antibodies shown to be specific for P. vivax and P. falciparum parasites were individually spotted for the capture of PvLDH and PfLDH, respectively, with a third spot representing a capture antibody to Plasmodium LDH as a PanLDH determinant. A broadly reactive PanLDH antibody was used as the detector antibody for all three assays (PfLDH, PvLDH, and PanLDH). For comparison of reactivity pattern, iRBC pellets of synchronized P. cynomolgi and P. knowlesi cultures and P. vivax clinical samples were diluted at 200 and 40 p/mL of parasite density and tested on the 5-Plex. For the samples at high concentration, which falls outside the standard range, the samples were subsequently diluted 1:20 and 1:400 with assay diluent, and the dilution factors were applied for the final concentration of antigens.
Statistical analysis. All data were analyzed using Graph-Pad Prism (version 5.0; La Jolla, CA). Pearson's correlation coefficient was calculated for the PanLDH level based on the 5-Plex and for parasite density based on microscopy or qPCR. The results relative to ratios of PanLDH/PfLDH and PanLDH/PvLDH are presented as mean 6 SD. Student's t-test was used to determine significant differences between ratio values of PanLDH and species-specific pLDH of recombinant and native proteins, estimated by reference assay. In addition, a two-way analysis of variance (ANOVA) with Sidak multiple comparison test was used to establish statistical significance between groups (P , 0.05).

RESULTS
Performance of the 5-Plex for reactivity to recombinant and native pLDH proteins for P. falciparum, P. vivax, P. malariae, and P. ovale. Dilution series of whole blood samples spiked with recombinant pLDH proteins specific for P. falciparum, P. vivax, P. malariae, and P. ovale curtisi, and clinical whole blood samples infected with each of four Plasmodium species were tested on the 5-Plex to examine the presence and level of reactivity to the assays for PanLDH, PfLDH, and PvLDH. Analysis showed that good specificity on PfLDH and PvLDH spots was found with recombinant PfLDH and PvLDH proteins, respectively (Supplemental Figure 2A). Interestingly, recombinant P. ovale-specific pLDH (PoLDH) and P. malariae-specific pLDH (PmLDH) proteins showed reactivity with both PanLDH and PfLDH spots of the 5-Plex. However, a notable difference was observed with the relative signal from the PanLDH spot, in which both PoLDH and PmLDH recombinant proteins showed relatively less signal, resulting in lower measured concentration as compared with recombinant PfLDH or PvLDH proteins for a given concentration ( Figure 1A).
For the reactivity to native pLDH from four Plasmodium species, the reactivity with PanLDH, PfLDH, and PvLDH spots of the 5-Plex was examined. Native PoLDH and PmLDH in clinical samples behaved similarly to their recombinant counterparts with both the PanLDH and PfLDH spots on the 5-Plex (Supplemental Figure 2B). Furthermore, the relation between level of PanLDH and parasite density was assessed with clinical samples infected by each Plasmodium species ( Figure 1B). The correlation analysis demonstrated that PanLDH level showed significant positive correlations with parasite density determined by either microscopy or qPCR for P. falciparum (N 5 29, r 5 0.670, P , 0.0001), P. vivax (N 5 60, r 5 0.674, P , 0.0001) and P. malariae (N 5 16, r 5 0.834, P , 0.0001) but relatively poor correlation with parasite density for P. ovale (N 5 7, r 5 0.737, P 5 0.059).
The ratio of PanLDH to species-specific pLDH was similar for both native pLDH from clinical samples and the same recombinant proteins across the four Plasmodium species (Table 1, Figure 2). Specifically, the ratio was 1.03 6 0.1 and 1. 6 0.37 for recombinant and native PvLDH proteins (all P values statistically not significant). The PanLDH concentration measurement in the 5-Plex is calibrated to recombinant PfLDH and so the ratio of close to 1 for PanLDH/PfLDH is to be expected.
Performance of reference assay evaluated with in vitro cultures of P. cynomolgi and P. knowlesi. To examine the binding patterns of pLDH from P. knowlesi and P. cynomolgi on each assay for PanLDH, PfLDH, and PvLDH, a series of dilutions (200 and 40 p/mL) of whole blood sample spiked with synchronized P. cynomolgi and P. knowlesi iRBC pellets were tested with the 5-Plex. The results were comparable to that of a clinical P. vivax blood sample. The results showed the similar reactivity between PkLDH and PvLDH proteins for binding to the PanLDH and PvLDH spots, and with comparable PanLDH/PvLDH ratio values between these two Plasmodium species at 2.7 for PkLDH and 2.8 for PvLDH at 200 p/mL of parasite density (Table 2). However, we found that PcLDH of P. cynomolgi only reacted with the PvLDH spot and not to the PanLDH spot.
Time-course study on PkLDH expression over parasite stages of P. knowlesi. For the time-course study, PkLDH protein in the iRBC pellet and supernatant obtained from in vitro culture over 30 hours after synchronization were tested by the 5-Plex. Our results ( Figure 3) demonstrated a life cycle for P. knowlesi of approximately 28 hours based on the time gap between the first ring stage and the second ring stage (Supplemental Table 1), similar to the previously published finding of 27 hours. 39 Although most of the PkLDH was found in iRBC pellets rather than in the supernatant, both pellets and supernatant showed progressively increased concentration of PkLDH (8 times more after 19 hours compared with 1 hour for both pellets and supernatant) during ring and trophozoite stages (Figure 3, Supplemental Table 1). The maximal amount of PkLDH is reached in the pellet during the schizont stage (21 to 27 hours), but a continuous accumulation of PkLDH released from the schizont rupture in the supernatant was observed over this period of time. Interestingly, during the first life cycle, parasite density seems to be decreased during the schizont stage, and the number of parasites in the second life cycle is increased after the released cells infect the fresh cells.

DISCUSSION
The human malaria array (5-Plex) was developed to provide quantitative measurement of the malaria antigens  Any samples showing antigen-negative results by the 5-Plex were excluded from analysis. Depicted are box and whisker plots for ratios of PanLDH/PfLDH for the P. falciparum, P. malariae, and P. ovale columns or ratio of PanLDH/PvLDH for the P. vivax column for each source of pLDH (recombinant and native). The ends of the boxes correspond to the 25th and 75th percentiles, respectively, and the middle line corresponds to the median. Two-way ANOVA plus Sidak multiple comparison test was used to determine the statistical significance against ratio values of P. falciparum. Significant differences (**** P , 0.0001, ** P , 0.001).
HRP2, PfLDH, and PvLDH and a qualitative measure for all malaria species (PanLDH) as well as measure the human biomarker CRP. 39 The pLDH spots used in this reference assay were screened for specificity against PfLDH and PvLDH 38,39 as well as characterized for their utility to screen for hrp2/hrp3 deletions both in whole blood pellets 39 and dried blood spots. 42 We demonstrated that the malaria array is a valuable tool in the detection and diagnosis of P. falciparum and P. vivax, using the described biomarkers. 39,42 However, its capability to diagnose less-common Plasmodium species such as P. malariae and P. ovale, and two simian Plasmodium species, P. knowlesi and P. cynomolgi, which have been recognized to cause human infection, has not been fully explored. As the 5-Plex includes a PanLDH spot detecting all Plasmodium species, in this study, we first evaluated the potential cross-reactivity of pLDH proteins from these parasites against pan-malaria and species-specific spots using both recombinant and native proteins.
The recombinant pLDH proteins of P. falciparum, P. malariae, P. ovale curtisi, and P. vivax offer great diagnostic value for the appropriate evaluation of the assay performance when there are limited clinical blood samples. Our study demonstrated that these recombinant pLDH proteins behave similarly to native pLDH proteins from corresponding Plasmodium species as demonstrated by PanLDH/species-specific pLDH ratio while being distinct from the equivalent host LDH protein (Table 1). However, there seemed to be different binding characteristics resulting in different signal responses from the PanLDH spot on the 5-Plex: the highest signal was observed with recombinant PvLDH followed by recombinant PfLDH, recombinant PoLDH, and then recombinant PmLDH (Supplemental Figure 2A).
The significant variation of PanLDH values in the estimated concentration of recombinant pLDH proteins suggest two possibilities: 1) some minor variance in the amino acid sequence of the conserved region or that of the surrounding area that enhances or reduces the antigen-antibody reaction, and 2) inefficient protein folding during protein expression/purification might have occurred when using a prokaryotic expression system. The second possibility may be addressed by evaluating the native pLDH proteins purified from different Plasmodium species, but obtaining native proteins from live parasites is a major challenge, especially with species of P. vivax, P. ovale, and P. malariae, for which continuous culture systems have not been established unlike P. falciparum strains. [43][44][45] Alternatively, nonhuman primate models could offer the strategies of producing non-P. falciparum parasites and their respective native proteins. 46 The pLDH amount measured as PanLDH correlates to parasite density equally across species, as shown by the normalized binding reactivity of native pLDH proteins with respect to the given parasite number (Supplemental Figure  3). The results suggest that this could be an artifact derived from the protein purification processes.
Additionally, both PmLDH and PoLDH were found to react not only to the PanLDH spot but also to the PfLDH spot, suggesting an overlap in antigenic epitopes of pLDH possessed by P. falciparum, P. malariae, and P. ovale. In our preliminary study, cross-reactivity between P. malariae and the PfLDH band on a few commercial RDTs was also observed (data not shown). It will therefore be of interest to investigate the extent of the level of cross-reactivity, which may have a significant influence on interpreting diagnostic test results. The PanLDH/species-specific pLDH ratio values for P. malariae, P. ovale, and P. vivax were significantly different from that for P. falciparum (Figure 1), suggesting the potential to use this ratio difference in differentiating Plasmodium species on the 5-Plex.
The reactivity patterns of pLDH from P. vivax and P. cynomolgi were compared with P. knowlesi, which is more distantly related to P. vivax. P. cynomolgi causes periodic relapses by dormant hypnozoites, early infectious gametocyte formation, and invasion of Duffy blood group-positive reticulocytes, and thus offers a robust model for P. vivax infection in the rhesus monkey model. 40,47,48 This study  revealed that P. cynomolgi failed to exhibit cross-reactivity to the PanLDH spot of the 5-Plex but did show reactivity to the PvLDH spot (Table 2). Unfortunately, there is no information available on the specific epitopes for PanLDH-specific capture and detector antibodies used in the 5-Plex. We aligned the amino acid sequences of the pLDH proteins from four different Plasmodium species (obtained from the Gene-Bank database) and found a conserved region (the region forming the loop structure) as the potential site recognized by pan-malaria antibodies and several species-specific regions as described previously (Supplemental Figure 4). [49][50][51] We did not find any obvious sequence variance in the potential pan-epitope region of PcLDH protein as compared with three other parasite species including P. falciparum, P. vivax, and P. knowlesi that would explain our observation. It would be interesting to explore binding characteristics of commercially available pan-malaria antibodies to the laboratoryadopted P. cynomolgi strain or recombinant PcLDH reagents and investigate the extent of sequence conservation within the conserved region of PcLDH protein from the cultured P. cynomolgi in a future study.
PkLDH exhibited significant cross-reactivity to both the PanLDH and PvLDH spots, but no or little cross-reactivity with the PfLDH spot on the 5-Plex. A previous study looking at cross-reactivity between pLDH proteins from different Plasmodium species and different antibodies to pLDH highlighted that PkLDH can react to antibodies specific to both PfLDH and PvLDH. 33 In the 5-Plex, only crossreactivity between PkLDH and the PvLDH antibody pair was observed. Additionally, the PanLDH/PvLDH ratio for P. knowlesi was similar to that seen when measuring PvLDH from the P. vivax clinical samples. Although the PanLDH and PvLDH spots on the 5-Plex can detect the presence of the PkLDH antigen, they were unable to distinguish P. knowlesi from P. vivax infection. To differentiate P. knowlesi from other Plasmodium species, development of an expanded array that hosts a P. knowlesi-specific antigen assay is required.
This is the first study to demonstrate the dynamic kinetic expression of PkLDH in a culture system. The results exhibited accumulation of PkLDH over time in culture, except with a slight decrease in PkLDH concentration seen in samples from the late schizont stage at around 25-28 hours. The pellet fraction contains the largest total amount of PkLDH until 29 hours after synchronization. At an inflection point where more PkLDH is contained in the supernatant, the erythrocytic cycle is completed through the rupturing of the schizonts to create new ring-stage parasites. The overall decrease in PkLDH seen starting at hour 25 may be explained by the lability of PkLDH and consequently an overall drop in PkLDH concentration until the parasites mature again. It should also be noted that the reactivity pattern of PkLDH proteins as a function of PanLDH/PvLDH ratio was maintained at 2.11 6 0.24 and 2.35 6 0.24 in both cell pellet and supernatant compartments, respectively, throughout all development stages (Supplemental Table 2).
The differential binding of pLDH from the different Plasmodium species investigated and observed in this study across the 5-Plex PfLDH, PvLDH, and PanLDH assays echoes results from previous studies investigating both antibody reactivity to pLDH and RDT reactivity to pLDH. 26,33,52,53 The results presented here are most likely specific to the particular anti-pLDH antibodies used in this 5-Plex assay. In a similar manner, the pLDH assay components of widely used RDTs should be evaluated for cross-reactivity to the different human malaria species, especially to P. ovale, P. malariae, and P. knowlesi, and also to the less-common species known to infect humans, including P. cynomolgi and P. simium. This will be particularly relevant to countries approaching elimination and/or that have a significant relative burden of P. knowlesi. 54,55 Between the emergence of hrp2/hrp3 deletions in P. falciparum infections and pLDH as the most commonly used target analyte for non-P. falciparum malaria RDTs, understanding the dynamics and crossreactivity of pLDH between the different human Plasmodium species is essential. Additionally, given the known crossreactivities between different Plasmodium species, new diagnostic markers for better differentiating Plasmodium species should be explored.