Background
Over the last decade, Myanmar, the country in the Greater Mekong Subregion that historically has had the highest burden of malaria, has made considerable progress toward malaria elimination. Since 2010, the number of malaria cases in Myanmar has decreased by more than 94% [
1], corresponding to increased malaria control and treatment efforts [
2]. This progress has presented new challenges for malaria control and elimination programmes to target interventions effectively and allocate limited resources to optimize impact on malaria transmission. These challenges include heterogeneous and focal transmission in geographic “hotspots” and population “hotpops” [
3,
4] as well as low-density asymptomatic infections, which are thought to be a silent reservoir of malaria transmission and represent the majority of infections in low-transmission settings [
5‐
7]. Ultrasensitive PCR (usPCR) methods have revealed that in areas of Myanmar, an estimated 85–99.5% of infections are asymptomatic [
2,
8‐
10], making them potentially important targets for malaria elimination. While usPCR has proven valuable for characterizing malaria distribution in Myanmar, it is not optimal for long-term surveillance, as it cannot be performed in the field and has high equipment and reagent costs [
11]. These challenges are exacerbated in elimination zones, where low transmission levels mean that large sample sizes are required for precise prevalence estimates [
12].
Antibody biomarkers represent promising tools for malaria surveillance in elimination settings that may mitigate some of the challenges of measuring parasite prevalence with usPCR. Antibody responses can reveal past exposure in addition to current infection status, adding substantially more information to population-level measures of transmission [
13,
14]. Antibody biomarkers are also readily detectable in field-friendly diagnostic formats similar to commercially available lateral flow-based rapid diagnostic tests (RDTs), and have been targeted in point-of-contact tests for other infectious pathogens, such as HIV [
15] and the hepatitis C virus [
16]. Additionally, it is likely that in individuals with low-density infections, malaria-specific antibodies are more readily detectable than parasite-specific antigens currently detected by currently available RDTs.
Protein microarrays are useful for screening antibody responses against large panels of antigens and protein fragments to identify potential biomarkers of exposure. In this format, antigens are printed onto nitrocellulose pads, antibodies in samples bind to antigens, and fluorescently-labelled secondary antibodies enable quantitative readout of signal. Protein microarrays have been used extensively to screen malaria-specific antibody responses, including those that elicit protection from uncomplicated and severe disease [
17‐
20] and transmission [
21,
22], as well as for markers of exposure [
14,
23‐
25]. In most studies, the sample matrix probed on protein microarrays has been human serum, the gold standard for serological studies. However, the complexity of sample processing and cold chain requirement for collecting and transporting serum is a major limitation in hard-to-reach places where malaria is prevalent. Dried blood spots (DBS) are potentially useful alternatives to serum; they are simple to collect, durable, easy to transport, and present minimal biohazard risk. In this study, the performance of DBS was directly compared to that of serum in a matched, head-to-head study using protein microarrays populated with antigen fragments from the
Plasmodium falciparum 3D7 reference genome (Pf250) [
26]. It was also hypothesized that long-term exposure to low-density infections, which have been shown to persist for months [
27], may elicit antibody signatures that could serve as biomarkers of infection. This hypothesis was tested by probing DBS on protein microarrays to identify antibody responses as markers for usPCR-positive (usPCR+)
P. falciparum malaria infections.
Discussion
This study demonstrated that DBS are viable samples for probing IgG responses on malaria protein microarrays. Despite a loss of signal relative to serum, IgG activity from DBS was successfully quantified, and similar trends were seen for antibodies derived from DBS and serum. Antibodies eluted from DBS were also found to be useful for identifying informative IgG responses in individuals who have asymptomatic, subpatent P. falciparum malaria.
Matched DBS and serum samples were probed on Pf250 protein microarrays, and approximately 30% of the IgG activity in serum was recovered from DBS. This recovery was lower than expected, as ELISA-based methods have reported up to 70% recovery of malaria IgG [
30], and the DBS elution procedure was designed to approximate the 1:100 serum dilution in the standard Pf250 protocol. It is unlikely that sample degradation played a role in this loss of IgG activity from DBS, as the DBS and serum were placed in the freezer at the same time, a strength of this study. As such, the loss of IgG activity can likely be attributed to the comparatively gentle, overnight DBS elution protocol. Many elution methods exist, and while harsh conditions may decrease antibody activity by denaturing or shearing antibodies, gentler methods such as the one used in this study may not completely elute all of the available antibodies. Additional optimization may identify ideal conditions for maximizing elution while minimizing antibody denaturation.
Although recovery of IgG activity from DBS was lower than expected, similar antigen and population trends were observed for antibodies obtained from both sample matrices. There was a large overlap of serorecognized antigens, with over 65% of the fragments that were recognized in serum also being serorecognized in DBS eluates. Average antibody response signal against serorecognized fragments was strongly correlated between the two sample matrices, indicating that DBS eluates provided reliable quantitative results compared to serum. Similarly, despite the overall lower signal from DBS, population trends in seroreactivity by gender and by age were comparable between the two sample matrices. Thus, IgG activity measured on protein microarrays from DBS can be relied on for observing antigen and population trends in malaria-endemic areas. These results likely extend to infections beyond malaria for which serosurveillance is routinely used to track exposure or vaccination coverage, and DBS collection would be more logistically feasible and cost effective than serum collection for serosurveillance of a number of infectious pathogens.
The case-control study aimed to identify serological signatures of asymptomatic
P. falciparum infections. Previously collected and banked DBS were eluted using the same method as the serum/DBS comparison study and probed on Pf250 microarrays. While the study was designed to identify markers of asymptomatic infection, it is important to note that serological profiles also reflect past malaria exposure, which is likely associated with infection status. Among the 18 antibody responses that affected the odds of usPCR positivity, 15 were positively correlated and three were negatively correlated with
P. falciparum infection. The antibody response with the greatest magnitude of negative or positive association with asymptomatic infection was to CSP, a leading malaria vaccine candidate [
33]. In this study, IgG activity against CSP was negatively correlated with
P. falciparum-positivity, consistent with CSP antibodies having a protective effect against infection as previous studies have reported [
34,
35]. Many of the other antibody responses identified here have also been identified in other studies as markers of
P. falciparum exposure or asymptomatic infection using similar protein microarray methods with serum samples [
14,
17,
18,
36,
37]. For example, five of the 18 antibody responses identified here were also implicated by Helb et al. as predictors of the number of days since previous
P. falciparum infection or of infection incidence in the last year [
14]. These consistent findings between the present study and previous work provide further evidence that DBS are useful samples for profiling
P. falciparum-specific immune responses on protein microarrays.
The antibody responses against the 18 antigens in Table
2 could be combined to classify individuals as usPCR-positive and usPCR-negative with modest performance (AUC = 0.772). Model performance was inversely related to age, working better in younger age groups (0–10, AUC = 0.854) compared to older age groups (41+, AUC = 0.558). This age effect is likely due to repeated malaria exposure contributing to increased breadth and intensity of malaria-specific circulating antibodies in older people that obfuscated the signatures attributable to usPCR-positivity.
The test characteristics of the model relative to usPCR were modest across all samples, with 58% sensitivity, 85% specificity, 88% NPV, and 52% PPV, suggesting that these serological signatures alone would not be sufficient for accurately identifying usPCR-positive individuals. However, it is important to note that all of the individuals included in this case-control study were RDT-negative. While the model does not accurately capture all usPCR-positive infections, it does capture more of the asymptomatic reservoir than standard RDTs currently used for malaria diagnosis, though at the cost of reduced specificity, particularly for adults or individuals with a history of previous exposure. Given this improved sensitivity over RDTs, serological markers could be useful intermediate tools for identifying exposure to ultra-low density malaria infections, particularly if serosurveillance strategies target individuals with less previous exposure, such as children, as sentinels for recent population exposure.
Conclusions
Screening malaria-specific antibody responses on protein microarrays can help identify immune factors that elicit protection from malaria infection, disease, and transmission, as well as for markers of past exposure. Most malaria protein microarray work has relied on serum as a sample matrix, which requires immediate laboratory processing and continuous cold chain. This study demonstrated that malaria-specific immune responses can be detected, quantified, and analysed from DBS on protein microarrays. As a proof-of-concept, DBS from asymptomatic P. falciparum cases and controls were probed to identify a panel of antibody responses associated with increased odds of asymptomatic malaria infection. The use of DBS on protein microarrays enables high-dimensional serological studies in populations where serum collection, transport, and storage would be otherwise impossible.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.