Selection and validation of novel DNA aptamer panel co-specific to Plasmodium falciparum and Plasmodium ovale lactate dehydrogenase Authors: Liga Kunrade, Karlis Pleiko, Laura Sproge, Karina Goluba, Teodors Pantelejevs, Nathan Thompson Wright, Inara Akopjana, Janis Bogans, Kaspars Tars, Una Riekstina Published: 2026 ABSTRACT AbstractMalaria remains a major global health concern. The development of point-of-care diagnostic aptasensors could improve malaria detection, patient management, and eradication efforts. We selected and validated single-stranded DNA aptamers against Plasmodium falciparum lactate dehydrogenase (Pf-LDH), using systematic evolution of ligands by exponential enrichment (SELEX). The five most abundant aptamers were validated with an enzyme-linked oligonucleotide assay (ELONA). The aptamer binding affinities to Pf-LDH varied between 0.35 and 3.08 nM. Truncation of aptamers enhanced binding affinities in most cases, with all truncated aptamers showing comparable or improved dissociation constants (Kd) relative to full-length aptamers, except Pf-LDH5t, which lost its binding affinity. Selectivity analysis against LDH from P. ovale, P. vivax, and P. malariae showed that Pf-LDH1 and Pf-LDH1t specifically recognized Pf-LDH (Kd = 2.74 ± 0.28 and 1.13 ± 0.26 nM, respectively), while Pf-LDH2, Pf-LDH3, Pf-LDH4, and their truncated forms, as well as Pf-LDH5, recognized both Pf-LDH and P. ovale LDH (Po-LDH). Crystallography of Pf-LDH4t revealed a binding site overlapping that of the previously characterized aptamer 2008 but involving distinct molecular interactions. These findings demonstrate the structural versatility of Plasmodium LDH-specific aptamers and provide a novel Pf- and Po-LDH co-specific aptamer panel with strong potential for next-generation aptamer-based malaria diagnostics.Keywords: MT: special issue: innovations in aptamer technology, DNA aptamers, malaria, Plasmodium spp., aptasensors, protein SELEX Introduction Malaria is a deadly infectious disease caused by Plasmodium spp. parasites that are transmitted by Anopheles spp. mosquitos, with Plasmodium falciparum accounting for most malaria-induced deaths.1 According to the World Health Organization (WHO) data, 263 million malaria cases were reported globally in 2023, corresponding to an incidence of 60.4 cases per 1,000 individuals at risk. Africa remains the most affected region, bearing approximately 94% of the global malaria burden; however, the eastern Mediterranean has experienced a 57% rise in malaria incidence since 2021.2 Current malaria diagnostics include microscopic examination of peripheral blood smears, polymerase chain reaction (PCR), and rapid diagnostic tests (RDTs).3 Antibody-based RDTs are aimed to recognize abundant proteins produced by Plasmodium parasites such as lactate dehydrogenase (LDH) or histidine-rich protein 2 (HRP2).4 Conventional RDTs are largely based on lateral flow assays, which utilize antibody-mediated biorecognition to detect biomarkers as fluid samples migrate along a sensor strip.5 Although currently used antibody-based RDTs exert an acceptable limit of sensitivity of >95% and specificity of >90%, their performance is often limited by instability under prolonged tropical temperatures, variability between production batches, relatively high cost, and short shelf-life.4,6 Additionally, existing tests exert a low Plasmodium species differentiation capacity, with up to 20% false-positive rate, low sensitivity in situations of low parasitemia, and a lack of specificity toward mutated P. falciparum.7,8 Failure to detect HRP2 caused by pfhrp2/3 mutation or deletion in P. falciparum parasites has also been reported, limiting its the use as a marker for malaria detection.9 Plasmodium LDH is a more reliable marker for the development of a point-of-care diagnostic device because it is abundantly expressed at the early stage of the malaria and shares 90%–92% genetic identity between Plasmodium spp. and thus could overcome the limitations of HRP2 diagnostic tests.9,10 Accurate malaria diagnosis is a critical challenge in the efforts to eliminate the disease, to avoid unnecessary treatment, and to reduce the risk of drug resistance.11 Because the current gold standard malaria diagnostic tools—microscopy, PCR, and RDT—cannot ensure efficient population-based point-of-care malaria screening, there is an urgent need for improved, cost-effective, and stable point-of-care diagnostic devices. Aptamers are emerging as promising alternatives to antibodies due to their high affinity and specificity to the target, stability, cost-effective synthesis, low batch variability, and ease of modifications with labels or anchor groups.12,13 Systematic evolution of ligands by exponential enrichment (SELEX) is an aptamer selection technique capable of generating high-affinity aptamers against a wide range of targets, from large proteins to small molecules. Several Plasmodium LDH aptamer candidates for malaria detection have been described in the literature, with aptamer 2008 being the most prominent.6,14 Aptamer 2008 is a P. falciparum LDH (Pf-LDH)-specific, single-stranded DNA (ssDNA) aptamer, in which two aptamers bind per Pf-LDH tetramer via a distorted hairpin structure.6 In a similar structure, aptamer pL1, which is co-specific to both Pf-LDH and Plasmodium vivax LDH (Pv-LDH), exhibits aptamer binding to Pv-LDH by folding into a hairpin-bulge, which then promotes binding using both shape complementarity and multiple bridging water molecules.14,15 Identification of pan-aptamers for malaria detection together with Plasmodium spp. specific aptamers is of high importance for the development of a malaria aptasensor that could ensure a wide-range screening and a further proper therapy decision for patients. Here, we report a selection of Pf-LDH-specific aptamer Pf-LDH1t as well as a Pf-LDH and Po-LDH co-specific aptamer panel that exerts a higher species recognition pattern than Pf-LDH aptamers reported before. Results Sequence diversity and dominance analysis across SELEX rounds To quantitatively assess SELEX progression beyond sequence abundance, we evaluated changes in diversity and dominance across rounds R6–R8, using complementary metrics. Effective diversity (exp(H)) showed a pronounced decrease from R6 to R8, indicating a strong reduction in the number of effectively represented sequences within the pool (Figure 1B). Consistently, the fraction of reads contributed by the top 100 sequences increased markedly across rounds, reflecting progressive dominance of the enriched variants (Figure 1D). In parallel, the fraction of singleton reads declined substantially, indicating depletion of low-abundance background sequences (Figure 1C). Together, these trends demonstrate clear convergence of the sequence pool during selection, characterized by loss of diversity and expansion of a limited number of enriched sequence families. Illustration of the protein-SELEX principle, effective diversity, singleton fraction, and dominance of enriched sequences across eight SELEX rounds (A) Schematic of the Pf-LDH SELEX workflow. The image was generated using Biorender software. (B) Effective diversity (exp(H)) decreased sharply from R6 to R8, indicating a strong reduction in sequence diversity and progressive enrichment of a limited subset of sequences. (C) The fraction of singleton reads declined across rounds, demonstrating depletion of low-abundance sequences and reduction of background diversity during selection. (D) The fraction of reads contributed by the top 100 sequences increased from R6 to R8, indicating progressive dominance of enriched sequences and convergence of the library. The fraction of reads accounted for by the top 10 individual sequences (R6, 11.6%; R7, 18.7%; R8, 28.5%) was of similar magnitude to that of the top 10 clusters (R6, 32.0%; R7, 27.7%; R8, 38.4%), indicating that enrichment was driven by a limited number of dominant variants and closely related sequence families; accordingly, we prioritized a subset of the most abundant sequences for downstream validation. All quantitative values underlying these analyses, including per-round diversity metrics (round_diversity_metrics.csv) and cluster-level summaries (cluster_diversity_metrics.csv and cluster_counts_all_rounds.csv), are available upon request. These files enable full reproducibility and independent inspection of diversity, dominance, and clustering behavior across SELEX rounds. Selection of high-affinity aptamers toward Pf-LDH and their validation in vitro Protein-SELEX method was exploited to select ssDNA aptamers toward Pf-LDH (Figure 1A). NGS analysis revealed five enriched aptamer candidates over the course of eight SELEX cycles. These sequences, designated Pf-LDH1, Pf-LDH2, Pf-LDH-3, Pf-LDH4, and Pf-LDH5, share identical 18-nucleotide primer-binding regions at both termini, derived from the initial randomized library, and each features a unique 40-nucleotide central region (5′-ATCCAGAGTGACGCAGCA-N40-TGGACACGGTGGCTTAGT-3′) (Table S1). To evaluate sequence affinity toward Pf-LDH, ELONA was conducted. All five aptamers exhibited low nanomolar dissociation constants (Kd), indicating strong binding to the target—2.74 nM for Pf-LDH1, 2.11 nM for Pf-LDH2, 3.08 nM for Pf-LDH3, 0.35 nM for Pf-LDH4, and 1.94 nM for Pf-LDH5 (Figures 2A and 2B). Representative data of the ssDNA aptamers Pf-LDH1, Pf-LDH2, Pf-LDH3, Pf-LDH4, Pf-LDH5, and their truncated (t) versions’ binding affinity toward Pf-LDH by ELONA (A and B) Assessment of full-length aptamer binding to Pf-LDH (n ≥ 3). (C and D) Assessment of truncated aptamer binding to Pf-LDH (n ≥ 3). Aptamer 2008 was used as a positive control and as a reference. Aptamer GreenB1 specific to integrin-β1 was used as a negative control. Data are represented as the mean ± SEM from ≥3 independent ELONA measurements. Kd, dissociation constant. Next, the Pf-LDH-binding aptamers were truncated by removing the constant primer-binding regions, resulting in unique 40-nucleotide core sequences (Table S1), designated as Pf-LDH1t through Pf-LDH5t. The estimated Kd for the truncated sequences were 1.13 nM for Pf-LDH1t, 0.99 nM for Pf-LDH2t, 0.21 nM for Pf-LDH3t, and 0.33 nM for Pf-LDH4t, while Pf-LDH5t exhibited no detectable binding to Pf-LDH under the assay conditions (Figures 2C and 2D). Thus, truncation led to an improvement in binding affinity for Pf-LDH1t, Pf-LDH2t, Pf-LDH3t, and Pf-LDH4t compared with their full-length counterparts. Notably, the truncated aptamers demonstrated ≥14-fold higher affinity to Pf-LDH than aptamer 2008. Co-specificity of the selected ssDNA aptamers toward other Plasmodium spp. LDHs To assess the species specificity of the selected aptamers, co-specificity was evaluated against LDH from P. vivax (Pv-LDH), P. ovale (Po-LDH), and P. malariae (Pm-LDH), using ELONA. All aptamers except Pf-LDH1 and the truncated Pf-LDH5t recognized Po-LDH with low nanomolar Kd values, as summarized in Figure 3. No detectable binding to Pv-LDH was observed for any of the tested aptamers. Weak binding was detected for both Pf-LDH2, Pf-LDH4, and their truncated forms (Pf-LDH2t and Pf-LDH4t) to Pm-LDH, with estimated Kd values around 1 μM. It should be noted that the binding analysis was performed toward the commercially available SPAN Diagnostics Plasmodium spp. LDH proteins. Co-specificity of selected ssDNA aptamers toward Plasmodium spp. LDH Data are represented as the mean ± SEM from ≥3 independent ELONA measurements. Nb, no binding; Nd, not determined. Kd for aptamer 2008 is representative of the value that has been obtained by in-house ELONA used to validate Pf-LDH aptamer sequences. Unspecific GreenB1 aptamer is used as a negative control. To assess potential off-target effects, cross-reactivity with human LDH isoforms A and B was also tested using ELONA. None of the aptamers exhibited binding to either human LDH subunit, which confirmed their high specificity toward Plasmodium-derived LDH (Figure S1). Selected aptamers exhibit reasonable stability profile in human serum Aptamer stability was evaluated by assessing band integrity via agarose gel electrophoresis following incubation in 10% and 100% human serum for up to 48 h. In 100% human serum, aptamer integrity varied between different aptamers. All aptamers except Pf-LDH4 and Pf-LDH5 were relatively stable in 100% human serum up to 3–6 h, and distinct aptamer bands were present for up to 48 h in 10% human serum, indicating good biological stability (Figure 4). In contrast, aptamer 2008 was stable in both 10% and 100% human serum for only up to 6 h under same conditions. Characterization of Pf-LDH4t binding and stability in 10% and 100% human serum after incubation for up to 24 h at 37oC Stability of aptamer was assessed as the integrity of the aptamer-associated gel electrophoresis band using ImageJ software. The selected aptamers exhibit a varying degree of competition with aptamer 2008 for Pf-LDH-binding sites To indirectly predict whether any of the selected aptamers bind to a site on Pf-LDH other than the one previously reported with aptamer 2008, we performed a competitive ELONA. Self-competition between biotinylated aptamer 2008 and its FAM-labeled version was used to establish a reference IC50 threshold. The self-competition assay yielded an IC50 value of 179 nM, indicating that a substantial excess of the competing aptamer is required to inhibit the test aptamer’s binding (Figure 5F). The experiment to evaluate the ability of aptamer 2008 to compete with the full-length versions of the selected aptamers for Pf-LDH binding was also performed. For most aptamers, IC50 values were comparable to that of aptamer 2008’s self-competition, suggesting a similar or overlapping binding site (Figures 5A–5C, 5E, and 5F). However, a notably higher IC50 value of 566.1 nM was observed for Pf-LDH4, indicating a reduced competition and suggesting that Pf-LDH4 likely binds to a distinct epitope on Pf-LDH (Figures 5D and 5F). Given that the truncated version Pf-LDH4t maintained a similar binding affinity to Pf-LDH and demonstrated enhanced affinity toward Po-LDH compared with its full-length counterpart (Figure 3), Pf-LDH4t was considered as the most promising candidate due to its dual-species recognition capacity and selected for further analysis. Competitive ELONA for selected Pf-LDH aptamers with aptamer 2008 (A) Competition of Pf-LDH1 with 2008. (B) Competition of Pf-LDH2 with 2008. (C) Competition of Pf-LDH3 with 2008. (D) Competition of Pf-LDH4 with 2008. (E) Competition of Pf-LDH5t with 2008. (F) Summary of Pf-LDH aptamer competition with 2008. IC50 values were estimated using GraphPad Prism v.8 non-linear regression fitting to a variable slope. Pf-LDH4t characterization and binding site identification ELONA provides one method for determining aptamer-LDH affinity. To check the reliability of this method and to gain more insight into the thermodynamic parameters that drive aptamer-LDH binding, we next performed isothermal titration calorimetry (ITC) of Pf-LDH4t-Pf-LDH interaction. In-house produced Pf-LDH was used for binding assessment with ITC. The Kd for Pf-LDH4t was approximately 239 ± 50 nM (Figure 6C; Figure S2). Three titrations were performed with different aptamer concentrations in the injection syringe, which suggested that Pf-LDH4t binds with a stoichiometry of 0.24, corresponding to one aptamer molecule interacting with a Pf-LDH tetramer. This is consistent with the X-ray crystallography where we observed that a single DNA aptamer bound to the Pf-LDH tetramer in a groove between two monomers (Figure 7A). There were no traces of electron density at the symmetric groove of the other two monomers nor was there space for another symmetrically placed aptamer due to crystal contacts. Presumably, the binding of aptamer to the one side of tetramer creates asymmetry in the other side of tetramer, preventing the binding of another aptamer copy. However, by aligning aptamer-bound and aptamer-unbound sides of Pf-LDH, few differences in protein structure could be observed. The flexible loop comprising residues 86–93 on chain B was clearly more ordered and in a very different conformation in the monomer where it contacts aptamer (Figure 7B), but it is unclear if or how much these dynamics influence aptamer binding. Our current explanation is that the binding of one Pf-LDH4t aptamer causes a cascading series of small structural changes in the other part of Pf-LDH molecule, leading to a slightly alternative conformation of the aptamer-binding regions that collectively prevent the binding of second Pf-LDH4t aptamer. ITC of aptamer Pf-LDH-4t (100 μM) binding to Pf-LDH (20 μM) support a 1:4 interaction stoichiometry (A) Representative ITC thermogram and isotherm of PfLDH-4t. (B) Fitted thermodynamic parameters of binding (n = 3, technical replicates). (C) Summary data for the experimental concentrations and fitted parameters for the one-site binding model (n = 3, technical replicates). Pf-LDH4t binding site determination on Pf-LDH (A) Binding of aptamer to Pf-LDH tetramer. DNA is shown in red, two aptamer-contacting monomers are shown in two different shades of green, and the other two Pf-LDH monomers are shown in two different shades of blue. (B) Different conformations of loop, comprising residues 86–93 in aptamer (red)-bound monomer (green) vs. aligned protein monomer (blue) from aptamer-unbound end of Pf-LDH. In green monomer, the loop is more ordered and favors binding, while in blue monomer, the same loop is less ordered and, in some places, produces clashes with aptamer (marked with black oval). Disordered parts of loop are shown with dotted lines. In Table S2, the interactions between the Pf-LDH and Pf-LDH4t are summarized. In total, 19 nucleotides from the aptamer were found to be involved in contacts with protein. As judged from buried surface areas, the most extensive interactions were provided by nucleotides G19, A20, G29, and T31, each producing buried surface area greater than 100 Å2 and contacting several residues of protein. Especially important seems adenylate A20, which produced about 240 Å2 of buried surface area. The adenine base of A20 occupies a largely hydrophobic pocket on the protein surface, lined by residues IleA36, TyrA67, AlaA80, GlyA81, PheA82, and IleA105 in A subunit. This protein pocket seems to have high affinity toward nucleotide bases, as the analogous pocket in the B subunit is occupied by a thymine base, resulting in 178 Å2 of buried surface. In total, 24 protein residues are in contact with Pf-LDH4t aptamer. Out of those 24 residues, all are conserved among Pf-LDH, Po-LDH, and Pm-LDH and only 2—His232 and Leu36—are different in Pv-LDH, where they are Leu232 and Val36, respectively. The conserved binding site of Pf-LDH4t aptamer is consistent with very similar binding constants to both Pf-LDH and Po-LDH; however, we cannot easily explain why aptamer largely fails to bind Pm-LDH. Failure to bind to Pv-LDH could be due to different residues in positions 232 and 36; however, even in this case, the binding site seems to change very little, and therefore, complete abolishment of binding is rather surprising. Sequence similarities among the aptamers In Figure S3, sequence alignment of newly discovered aptamers and 2008 is shown. In general, the sequences were quite divergent, but some similarities could be found in the central parts around adenine A20 (Pf-LDH4t numbering) base, bound in the protein pocket in both Pf-LDH4t and 2008. Aptamers Pf-LDH1/1t share an identical stretch of 8 nucleotides before A20 with Pf-LDH5/5t, not found in other aptamers. The corresponding nucleotides in Pf-LDH4t complex were found to make extensive interactions with the protein; therefore, it could be assumed that the exact binding mode of Pf-LDH1/1t/5/5t aptamers is rather different from that of Pf-LDH4/4t. On the other hand, Pf-LDH5/5t and Pf-LDH4t share a stretch of 6 identical nucleotides TCTGTT, just after A20 (C20 in Pf-LDH5t). However, these nucleotides in Pf-LDH4t make relatively few interactions with protein in case of Pf-LDH4t and instead appear to be important in maintaining the stem-loop structure. A similar stem loop cannot be present in Pf-LDH5t, because the sequence lacks corresponding base-pairing nucleotides on the 5' side. Aptamers Pf-LDH2/2t and Pf-LDH3/3t share very little sequence homology with others, and their alignment might be irrelevant. In fact, even for other aptamer pairs, the alignment data should be treated with caution, as it could easily be misleading. For example, upon alignment of 2008 and Pf-LDH4/4t, it might appear that identical tetranucleotide sequence CGGT (nucleotides 11–14 in Pf-LDH4t) must play similar roles in both aptamers, yet its placement in aptamer structure and interactions with protein were entirely different in both cases. Also, aptamers Pf-LDH5 and Pf-LDH5t share exactly the same variable sequence, but only Pf-LDH5 binds to the protein, suggesting that the flanking primer-binding sequences may influence the structure of the central part. Discussion Here, we report the selection and characterization of the Pf-LDH-specific aptamer Pf-LDH1t and a panel of Pf-LDH and Po-LDH co-specific aptamers. The binding of the lead aptamer Pf-LDH4t to Pf-LDH was characterized in detail using ITC, stability assays, and crystallography. To evaluate the performance of our aptamers, we used the published Pf-LDH aptamer 2008 as a reference. Aptamer modifications are often performed to improve the binding ability and stability. Such modifications can include insertion of unnatural nucleotides, chemical modifications, truncation, extension, and spatial structure configuration.16,17,18 We truncated the initially selected sequences by removing the primer-binding sites arising from the initial randomized aptamer library. Truncation improved the binding affinity toward the target protein Pf-LDH for all sequences, while the truncated version of Pf-LDH5t lost its binding ability, indicating either the involvement of primer-binding sites in the interaction with a Pf-LDH or a primer-required binding conformation. There are several aptamers published for Plasmodium LDH. The most well characterized are aptamer 2008, which is Pf-LDH specific, and pL1, which is Pf-LDH and Pv-LDH specific.14,15 The interaction of an aptamer with the target protein can be assessed by multiple methods such as ITC, electrophoretic mobility shift assay (EMSA), and ELONA. Because these methods can result in different Kd values, it is important to keep the methodology the same when comparing affinities. For validation, we used ELONA as the main method, and the Kd values for all tested aptamers were found in the low nanomolar range. We observed that our aptamers exhibited enhanced affinity and species-specific recognition compared with the aptamer 2008, as assessed side-by-side using our in-house ELONA method using Pf-LDH from SPAN Diagnostics. This improved performance suggests potential for the development of aptamer-based RDT technologies that could be particularly valuable in cases of low parasitemia. Furthermore, the increased species differentiation capability could enable more accurate identification of the infectious agent, supporting more appropriate and targeted treatment decisions for patients (Figure 3). To date, there are no reports on P. ovale-specific aptamers. Although P. ovale causes mild-to-moderate malaria and is clinically not considered as concerning as P. falciparum, diagnostic tests capable of differentiating between malarial species are essential for correct diagnosis and treatment. For instance, some RDTs utilizing pan-specific antimalarial antibodies have reduced affinities for P. ovale and P. malariae, leading to lower detection sensitivity for these species.19 Interestingly, all our aptamers, including the long versions and truncated versions, exerted a high co-specificity to Pf-LDH and Po-LDH, while none of them could bind to Pv-LDH (Figure 3). Importantly, our aptamer panel could not bind to human LDH isoforms A and B (Figure S1). LDH from P. vivax, P. malariae, and P. ovale shares approximately 90%–92% sequence similarity with P. falciparum LDH.10 Despite the high similarity, apparently the remaining differences are sufficient to result in distinct aptamer-binding profiles. We evaluated aptamer stability in both 10% and 100% human serum to determine degradation kinetics (Figure 4). Gel electrophoresis analysis revealed that all selected Pf-LDH aptamers maintained band integrity for up to 48 h in 10% human serum. In contrast, aptamer 2008 showed significantly reduced stability, with detectable degradation occurring after 6 h under same conditions. Most aptamers remained stable up to 6 h post-incubation in 100% human serum despite the high nuclease content. However, Pf-LDH4 and Pf-LDH5 displayed markedly reduced stability, with band degradation evident as early as 1 h post-incubation. We also investigated aptamer stability in lysed human blood (data not shown), but the results were inconsistent, possibly due to the presence of anticoagulants in the blood collection tubes, which may have interfered with band integrity assessment. Anticoagulants previously have been found to affect nuclease activity.20,21 For this reason, aptamer stability was mostly assessed in human serum. Aptasensor stability of 3 h or greater in human serum is considered sufficient for aptasensor development.22,23 Although aptamer stability in biological matrices remains a limitation, it should be noted that, in RDTs, stability is typically required only over a relatively short time period sufficient for result acquisition. In addition, biological samples are often diluted in reaction buffers, which may further mitigate nuclease-mediated degradation. Various chemical modifications can be employed to enhance aptamer stability. These include terminal capping, substitution with sulfur atoms, and incorporation of locked nucleic acid nucleotides, all of which have been reported to improve nuclease resistance and, consequently, stability in serum and whole blood.24,25,26 From our panel, Pf-LDH4t was chosen as the lead aptamer due to its equally high affinity for Pf-LDH and Po-LDH and its stability. Considering the competitive ELONA data for Pf-LDH4, we suspected that similarly the truncated version might bind to a different site on Pf-LDH than aptamer 2008, and this warranted further analysis. The ITC-derived affinity (239 ± 50 nM) was weaker than the ELONA-derived affinity (0.33 ± 0.09 nM). This difference may be due to the fact that ITC evaluates binding interactions in solution, while in ELONA, the Pf-LDH protein is immobilized. Immobilization-induced effects, such as the increased local concentration of a binding partner on a surface, may lead to re-binding events and, therefore, higher apparent affinities.27,28 Such discrepancies between various biophysical and biochemical assays are not uncommon. Thus, it is important to assess binding with orthogonal methods. For comparison, the ITC-derived Kd value of aptamer 2008 is 42 nM.6 It should be also noted that LDHs from different sources were used for ELONA and ITC analysis. Aptamer selection and ELONA analysis were performed using commercial Plasmodium spp. LDH proteins, while due to the need for high protein concentrations for ITC, an in-house Pf-LDH protein was produced. Variability in protein source has previously been shown to influence aptamer-target interactions, potentially arising from differences in post-translational modifications or protein conformation.29,30 Such variations may consequently impact the observed binding affinity and specificity of the aptamers. Various detection assays and transducer technologies, including electrical (e.g., potentiometric and impedimetric), magnetic, optical (e.g., surface plasmon resonance and surface-enhanced Raman scattering), and piezoelectric platforms, have been developed to translate aptamer-target interactions into measurable signals.31 Recently, paper-based microfluidic analytical devices (μPADs) have also been reported, utilizing covalent streptavidin immobilization in combination with biotinylated aptamers to enable selective recognition of Pf-LDH. The quantification of clinically relevant Pf-LDH levels in human blood has recently been achieved through aptamer-based electrochemical biosensors, highlighting significant advancements in the development of aptamer-driven diagnostic technologies.32 To further advance such technologies, a detailed understanding of the molecular interface between aptamers and their protein targets is crucial. Structural insights gained from protein crystallography and complementary biophysical approaches provide information on the binding epitopes, conformational changes, and interaction mechanisms. Structure of Pf-LDH in complex with aptamer 2008 has been described by Cheung et al.6 Perhaps, the most striking difference lies in fact that Pf-LDH tetramer binds to two 2008 monomers but only one Pf-LDH4t monomer. In comparing the binding mode of aptamers in both cases, the approximate binding location was the same, comprising a cleft between two Pf-LDH monomers. However, the exact binding mode was very different. Most obviously, aptamer 2008 binds symmetrically to both sides of Pf-LDH tetramer, while Pf-LDH4t aptamer binds only to one (Figure 8A). The total buried surface between the aptamer and protein was also significantly different—1,310 Å in 2008 case vs. 1,732 Å in Pf-LDH4t complex. Comparison of aptamer Pf-LDH4t and 2008 binding site structures (A) Aligned structures of 2008s aptamer (blue) versus our Pf-LDH4t aptamer (red) bound to Pf-LDH (green). (B) Sequence alignment of both aptamers; identical nucleotides are shown in red. (C) Secondary structure of aptamers. Canonical base pairs are indicated by a number of hydrogen bonds, while non-canonical base pairs are marked with dots. (D) Similar and different interactions with protein. In (A–D), Conserved adenine base, marked with rectangle, occupies the same binding pocket in A subunit but is in opposite orientations. Analogous binding pocket in B subunit is occupied by a thymine base, marked with an oval. In the Pf-LDH4t complex, there is unique stacking interaction of a thymine base with tryptophan residue. The corresponding thymidylate is marked with triangle. In sequence alignment, there were some similarities between both aptamers; in central part, there were 4 identical nucleotides followed by one mismatch and 2 more identities (Figure 8B). However, the secondary structures of aptamers were quite different (Figure 8C). In fact, four conserved nucleotides were found to make up a part of stem and did not participate in interactions with protein. Out of aligned nucleotides, only a single adenylate located at the top of stem loops occupied a similar binding position to protein in both cases, and even then the adenine base orientation differed; in the Pf-LDH complex, there was a H-bond to nearby TyrA82 (Figure 8D). The analogous pocket in the other LDH subunit bound to thymine bases in both the 2008 and Pf-LDH4t structures, but the position of the corresponding thymidylate nucleotide was in a different area of the sequence as well as in a different secondary structure motif. However, the protein pocket where adenine/thymine bases were observed seems to be important for the overall binding for a variety of aptamers. Apart from 2008 and Pf-LDH4t complex structures, in a cubane-modified aptamer structure,33 the same binding pocket of Pv-LDH in one subunit was found to be occupied by adenine base in a similar orientation as in 2008Pf-LDH complex case, while the analogous pocket in the neighboring subunit was empty. In the crystal structure of pL1, with Pv-LDH, two thymine bases occupy the same binding pockets in neighboring protein subunits.14 Most of the remaining protein-DNA interactions seem to be unique in each complex; for example, there was a stacking interaction among thymine base and tryptophan residue in Pf-LDH4t complex, which was absent in 2008. We performed in silico modeling of each of the truncated aptamer-Pf-LDH complexes. However, we found significant discrepancies between the Pf-LDH4t in silico model and our crystal structure. In particular, while the binding site on LDH was globally the same, the secondary fold of the aptamer model never matched the fold in the crystal structure. Thus, we suspect that the other aptamer models may also be incorrect. The reasons for these poor results seem to be 2-fold. First, Alphafold and similar programs primarily use the aptamer 2008 structure as a basis for the Pf-LDH-aptamer modeling. Because the structural work presented here is the first to show that alternative aptamer binding orientations can exist, this information was unknown to Alphafold, and the resulting models are correspondingly incorrect. Second, energy profiles from programs such as mFold do not accurately predict the secondary structure of Pf-LDH4t bound to Pf-LDH. Thus, the data that folding algorithms use for constructing high-quality models were incorrect in this instance. Interestingly, Alphafold and MD simulations both indicated, with a high level of confidence, that these in silico models were indeed correct. This serves to underline the now-well-known caveat that modeling is still an imprecise tool despite advances in the past decade. As more diverse aptamer-LDH structures are solved, the predictive modeling accuracy will correspondingly improve. Because Pf-LDH4t binds to Pf-LDH in a 1:4 molar ratio, it may represent an advantageous candidate for use as a capture aptamer in a biosensor platform. This lower binding stoichiometry could help prevent excessive surface coverage on the biosensor interface, which can otherwise lead to reduced target accessibility. By minimizing surface crowding, Pf-LDH4t could promote more efficient target capture and support a more consistent spatial orientation of the analyte, potentially improving signal strength and reproducibility. In the context of a sandwich-format aptamer-based biosensor, this aptamer could be immobilized on the sensor surface, while a second, non-overlapping aptamer or aptamer conjugate served as a detection probe. Such a configuration could enhance sensitivity by promoting a more uniform analyte binding and reducing the background noise. However, at present, it cannot be ruled out that Pf-LDH4t binds to Pf-LDH in a 1:2 molar ratio at some particular conditions, as only minor structural changes appear to be necessary to accommodate Pf-LDH4t at both ends of the Pf-LDH tetramer. Notably, 1:4 molar ratio was observed both by ITC and crystallography, which have vastly different buffer compositions. It is also possible that at low Pf-LDH concentrations, the protein exists, at least partially, in dimeric form, complicating the development of capture-detection aptamer pair. In conclusion, we have identified aptamer Pf-LDH1t as a selective binder of Pf-LDH. Notably, we are the first to report aptamers with co-specificity for both Pf-LDH and Po-LDH, with Pf-LDH4t emerging as a lead candidate due to its unique binding profile to Pf-LDH compared with previously published aptamers. The development of a targeted aptamer panel against P. falciparum and other Plasmodium species could enable next-generation point-of-care multiplex diagnostics, improving malaria detection and supporting global control efforts. Materials and methods Protein-SELEX 20 μg of His6-tagged Pf-LDH protein (SPAN Diagnostics) was bound to magnetic beads (MBs; Dynabeads™ His-Tag isolation and Pulldown, Thermo Scientific) and incubated for 1 h at room temperature (RT). MBs were washed 3× with washing buffer (PBS, 4.5 mM MgCl2 (Thermo Scientific), 0.1% BSA (Sigma-Aldrich), and 5 mM imidazole (Sigma-Aldrich)) on a magnet. 1 μM randomized ssDNA library with 18-nucleotide constant primer-binding regions (5′-ATCCAGAGTGACGCAGCA-N40-TGGACACGGTGGCTTAGT-3′, IDT) was folded by heating at 95°C for 5 min in a folding buffer (PBS, 4.5 mM MgCl2) and cooled to RT over 15 min. The ssDNA library was then mixed with MB-conjugated Pf-LDH and incubated for 1 h at RT. Washing of the complexes was performed one to three times on a magnet. Bound aptamer pool after each washing step was eluted using an elution buffer (PBS, 0.1% BSA, 400 mM imidazole) for 3 min on a magnet. Eluates from each washing step were amplified and biotinylated by PCR, starting with 4 and increasing by 2 cycles until 12 cycles, using Solis Biodyne Ready to load 5× FIREPol master mix. PCR cycle conditions were as follows: 12-min initial denaturation at 95°C, 4–12 cycles of 30-s denaturation at 95°C, 30-s annealing at 56.3°C, 30-s elongation at 72°C using forward primer 5′-ATCCAGAGTGACGCAGCA-3' and reverse primer 5'-Biot-ACTAAGCCACCGTGTCCA-3'. The samples were visualized by agarose gel electrophoresis. The optimal amplification cycle number and eluate, that gave a uniform band, were selected and amplified by preparative PCR. Amplified dsDNA was purified using DNA clean & concentrator kit (Zymo) and bound to MyOne streptavidin C1 MBs (Thermo Scientific) by incubating for 30 min at RT. Washing was performed for 2–3× in folding buffer (PBS, 4.5 mM MgCl2), and the ssDNA aptamer pool was eluted with 20 mM NaOH (Sigma-Aldrich) for 10 min; pH was adjusted with 20 mM HCl (Sigma-Aldrich). ssDNA was purified by oligo purification kit (Zymo) and resuspended in 15 μL of molecular-grade water (Sigma-Aldrich). Selected ssDNA aptamer candidates were used for the next selection cycle. 8 selection cycles were performed. High-throughput sequencing High-throughput sequencing (HTS) was performed as described before.34 Sequencing was done on the Illumina MiSeq platform, using MiSeq Reagent Nano Kit v.2 in the single-read mode for 300 cycles. 30% of PhiX was added to the run. SELEX sequencing analysis and diversity metrics HTS data from SELEX rounds were processed using custom R scripts. FASTQ files were trimmed to retain only the randomized region, and sequences were extracted from each read. Identical sequences were collapsed to generate count tables for each round. To enable a direct comparison across rounds, sequencing depth was normalized by rarefaction. All rounds, except round 5 (which had substantially lower sequencing depth), were randomly subsampled without replacement to a common read depth corresponding to the lowest-depth sample among rounds 6–8. For each round, sequence diversity and convergence were quantified using complementary metrics. Shannon entropy (H) was calculated based on the sequence frequencies and transformed into exp(H), representing the effective number of equally abundant sequences in the pool. Enrichment of dominant sequences was assessed using the top 100 fraction, defined as the proportion of total reads contributed by the 100 most abundant sequences. Loss of low-abundance background sequences was quantified as the singleton fraction (of reads), representing the fraction of reads corresponding to sequences observed only once. To assess convergence at the motif level, sequences were clustered based on Hamming distance ≤ 3, restricted to equal-length sequences. Clustering was performed on the most abundant sequences (top 2,000 by total count across rounds) to ensure computational tractability. All analyses were performed in R (v.4.5.2), using packages including data.table, dplyr, and igraph. Plots were generated using ggplot2. Aptamer truncation Aptamers were truncated by the removal of constant 5′ flanking region ATCCAGAGTGACGCAGCA and 3′ flanking region TGGACACGGTGGCTTAGT, leaving the unique 40-nucleotide-long sequence. Truncated aptamer sequences were generated by Metabion. Direct ELONA 96-well Ni-NTA HisSorb plates (Qiagen) were coated with 100 μL of His6-tagged Pf-LDH, His6-tagged Pv-LDH, His6-tagged Po-LDH, His6-tagged P. malariae LDH (Pm-LDH) (SPAN diagnostics), His6-tagged human LDH A (hLDHA), or His6-tagged human LDH B (hLDHB) (BSP Bioscience) at 2.5 μg/mL in PBS. Blocking was done by 200 μL blocking buffer (PBS, 1% BSA [R&D systems], 100 μg/mL tRNA [Sigma-Aldrich], and 4.5 mM MgCl2 [Invitrogen]) for 1 h at RT. Plates were washed 3× with 230 μL washing buffer per well (PBS, 4.5 mM MgCl2, 0.005% Tween 20). Biotinylated aptamers were folded and added to a plate in 5-fold dilutions ranging from 1 μM to 0.064 nM in a binding buffer (PBS, 4.5 mM MgCl2, 100 μg/mL tRNA), with a final volume of 100 μL per well, and incubated for 1 h at RT. GreenB1 aptamer specific to integrin-β1 that has been described by our group was used as a negative control.35 The samples were run in triplicates. Wells containing 100 μL of binding buffer without aptamer served as a control. After incubation, the plates were washed 3×, as described above. Poly-HRP-streptavidin reagent (Poly-HRP; Thermo Scientific) was diluted 1:2000 in blocking buffer, and 100 μL was added to each well and incubated for 5 min at RT in the dark. The plates were washed 3× in a washing buffer. 100 μL of 1-Step™ TMB ELISA substrate solution (Thermo Scientific) was added to each well and incubated for 5 min at RT in the dark. Reactions were stopped with 100 μL of 1 M H2SO4, and optical density (OD) was measured at 450 nm by using a Tecan Infinite 200Pro spectrophotometer. Kd values were determined using GraphPad Prism 8 software, using a non-linear regression analysis (95% confidence interval). Competitive ELONA with aptamer 2008 Materials, buffers, protein-coating procedure, and aptamer folding were performed as described above. Biotinylated aptamer was diluted in a binding buffer to achieve the final concentration that corresponded to its Kd value—2.75 nM for C8N1, 2.11 nM for C8N2, 3.08 nM for C8N3, 0.35 nM for C8N4, 1.94 for C8N5, and 15 nM for 2008. 100 μL of the binding buffer (PBS, 4.5 mM MgCl2, and 100 μg/mL tRNA) containing the biotinylated aptamer was applied to all wells of the plate except the first row. Binding buffer containing the biotinylated aptamer and 1 μM aptamer 2008 with FAM label ([FAM]-5′-CTGGGCGGTAGAACCATAGTGACCCAGCCGTCTAC-3′) was used as a competitive aptamer, and 125 μL was added to the first row of the plate and then, using 5-fold serial dilutions, was diluted from 1 μM to 0.064 nM in a binding buffer, with a final volume of 100 μL per well, and incubated for 1 h at RT. Wells containing 100 μL of binding buffer without 2008-FAM served as a control. The samples were run in triplicates. Procedure after the incubation of aptamers is identical to the one described for the direct ELONA. IC50 values were obtained by non-linear regression fitting to a variable slope, using the GraphPad Prism 8 software. Aptamer stability in human serum Aptamers were diluted to 1 μM in 100% human serum (Sigma-Aldrich) or 10% human serum. For serum dilutions, a folding buffer (in PBS, 4.5 mM MgCl2) was used. The samples were incubated at 37°C for 0, 1, 2, 3, 6, 24, and 48 h. At each time point, 10 μL of each sample was transferred to a separate tube and mixed with 110 μL of nuclease degradation buffer (1X TBE, 10 M urea).18 Serum without aptamer at a corresponding dilution served as a negative control. The aptamer diluted in folding buffer was used as a positive control. The samples were run in a 3% agarose gel electrophoresis and imaged by UVP imaging system. The results were quantified using gel band densitometry analysis by using ImageJ software. Relative band density was calculated assuming 0 h band as 100%. ITC ITC experiments were performed on a PEAQ ITC instrument (Malvern Panalytical) in PBS as the reaction buffer. LDH and aptamer concentrations were determined using UV-vis absorbance at 280 and 260 nm, respectively. Three ITC experiments were performed with 10 μM LDH in the cell and 50 or 100 μM aptamer in the syringe at 25°C, with 750 RPM stirring and differential power = 4.0 μcal/s. Thermogram integration and isotherm fitting were performed using the PEAQ ITC analysis software (Malvern Panalytical). Isotherms were fitted to a 1:1 binding model with variable stoichiometry and fixed concentrations. Offset subtraction was implemented to account for injection heats. Expression of recombinant protein and crystallography LDH amino acid sequence from Pf-LDH (GenBank: ABY76169.1) was complemented with an N-terminal His6 tag. The corresponding E. coli codon-optimized nucleotide sequence was ordered in company BioCat and cloned in the expression vector pETDuet-1. The protein was produced using standard protocols in E. coli BL21(DE3) cells at 20°C. The cells were lysed by sonication, and protein from the soluble fraction was purified using metal affinity on HisTrap Crude column (Cytiva), followed by gel-filtration on Superdex 200 (Cytiva). Aptamer Pf-LDH4t was dissolved in 20 mM Tris-HCl and 10 mM MgCl2 and mixed with purified Pf-LDH in a molar ratio 1:1 to Pf-LDH monomer. The protein concentration in the mixture was 5 mg/mL. The mixture was incubated for O/N at 4°C and further applied on to a Superdex200 (16/600) column (Cytiva) equilibrated to 20 mM Tris-HCl and 10 mM MgCl2. The chromatogram contained two major peaks—one of higher MW with A260/280 ratio of 1.5, characteristic to that of protein-nucleic acid complexes, and the second peak of lower MW with A260/280 ratio of 2.0, characteristic to that of pure nucleic acids. The higher-MW peak containing the purified Pf-LDH-Pf-LDH4t complex (“4t complex”) was collected and concentrated to 10 mg/mL, using Amicon Ultra-15 10K filtration unit. The 4t complex was crystallized using the sitting drop evaporation technique and several crystallization kits from Molecular Dimensions. The best crystals were obtained with 0.1 M sodium citrate, pH 5.5, 15% PEG6000. The crystals were briefly soaked in mother liquor supplemented with 30% glycerol, flash-frozen in liquid nitrogen, and sent for data collection to beamline i04 of Diamond Light source, UK. The data were collected and autoprocessed in a fully automated mode. The best dataset had a resolution of 1.9Å. The structure was solved by molecular replacement, using protein part of PDB: 3ZH2 as a search model in the CCP4 suite36 program Molrep.37 The DNA chain was manually built-in electron density using COOT38 and structure refined in REFMAC.39 The data collection, refinement, and validation statistics are shown in Table S2. Coordinates along with structure factors were deposited in Protein Data Bank, with accession number PDB: 9RTD. Buried surface areas were determined using the ePISA web server (https://www.ebi.ac.uk/msd-srv/prot_int/cgi-bin/piserver), and protein residues in contact with aptamer were determined using the CCP4 suite program Contact. Data and code availability The data that support the findings of this study are available from the corresponding author (L.K.) upon request. Acknowledgments This project was funded by a EuroNanoMedIII project “Quantitative and storage-stable point-of-care diagnostic device”, Latvian Science Council reg. no. ESRTD/2022/3 and University of Latvia reg. no. ZD2022/21302. N.T.W. was supported by NIH grant 1R15GM148890. K.P. was supported by 1.1.1.9 Research application no, 1.1.1.9/LZP/1/24/049 of the Activity “Post-doctoral Research” “AptaPhage - bringing the power of phage biopanning to aptamer selection”. L.S. was supported by UL doctoral grant “Strengthening Doctoral Research and Development Capacity” project no.: 1.1.1.8/1/24/I/003, University of Latvia reg. no. ESS2025/492d32. Author contributions L.K. and U.R. designed the experiments; K.P. and K.G. performed SELEX of aptamers; L.K. and L.S. validated aptamers using the ELONA method; T.P. performed ITC; I.A, J.B., and K.P. performed crystallography and analyzed aptamer protein interactions; N.T.W. performed in silico modeling. All authors have edited and revised the manuscript. Declaration of interests The aptamer panel presented in this manuscript has been granted a national patent no. LVP2025000016. Footnotes Supplemental information can be found online at https://doi.org/10.1016/j.omtn.2026.102977. Supplemental information Document S1. Figures S1–S3 and Tables S1–S3 Document S2. Article plus supplemental information References Associated Data Supplementary Materials Document S1. Figures S1–S3 and Tables S1–S3 Document S2. Article plus supplemental information Data Availability Statement The data that support the findings of this study are available from the corresponding author (L.K.) upon request.