What looks like one malaria parasite species may actually be two—and mosquitoes revealed that mixed infections could be far more common than blood tests suggest. P. ovale curtisi and wallikeri circulate sympatrically in time and space, yet mixed infections have rarely been described: fewer than twenty cases were reported across more than thirty-five studies covering 1,515 P.

ovale cases. Because mixed infections of other Plasmodium species are well documented, the technical limitations of detection appeared more plausible than a biological barrier. The study therefore adapted nested and real-time PCR assays to establish criteria for identifying both Poc and Pow, then applied them to P.

ovale-positive isolates from Africa. P. ovale, one of five malaria species known to infect humans, actually represents two distinct species—P. ovale curtisi and wallikeri—that can only be distinguished using molecular diagnostics.

The modified real-time PCR assays targeted 18S rRNA, and an algorithm was developed to detect mixed Poc and Pow infections. The assays were applied to field-collected samples from Tanzania and Cameroon, including infected human blood and P. ovale-positive mosquito midguts from feeding assays.

Both species appeared in roughly ten percent of human blood-stage infections and in a majority of blood-fed mosquitoes, suggesting more frequent co-infection than previously realized. Table one compiles published molecular assays for distinguishing P. ovale, Po curtisi, and Po wallikeri in clinical samples.

It compares nested PCR, real-time qPCR, and amplicon deep sequencing, detailing targets, primers or probes, sample locations, and detected species counts. The table also records assay-specific information such as detection sensitivity, melting temperatures, and cases where mixed infections were confirmed or not detected, showing the range of approaches used for species-level identification.

Sensitivity was tested using dilutions of plasmids containing the species-specific 18S rRNA gene targets for Poc and Pow. The Poc plasmid came from BEI Resources, while the Pow control was created by cloning a first-round nested PCR product from a Sanger-sequenced Pow clinical sample from Tanzania.

Real-time PCR used FastStart Universal Probe Master mix and published primer and probe concentrations on a Bio-Rad CFX Connect system. The modified qPCR assays were tested across plasmid concentrations ranging from ten to the fifth down to ten to the minus first copies per microliter, and a limit of detection was calculated using Probit analysis.

To model mixed infections, Poc and Pow plasmid controls were combined from ten squared down to ten to the minus first copies per microliter in ratios of one to one, one to two, one to five, and one to ten. Each mixture was run ten times, and the mean cycle threshold from those runs was used to develop a classification scheme for mixed infections.

The field comparison used DNA from sixteen Cameroon and twenty-two Tanzania blood samples previously identified as P. ovale positive by an 18S qPCR detecting both species. The Tanzania samples came from symptomatic children in a malaria therapeutic efficacy study and from asymptomatic children and adults in a malaria transmission study.

Samples with lower P. ovale 18S cycle-threshold values were deliberately selected, and the study compared amplification, species concordance, and mixed-species detection between nested PCR and modified qPCR. Real-time PCR was also performed on DNA from seventeen P.

ovale-positive mosquito midguts obtained from feeding assays with colony-reared Anopheles gambiae in Tanzania. The midguts were oocyst-positive by microscopy at day eight after feeding, then underwent DNA extraction and a first-round conventional PCR with Plasmodium genus-specific primers.

A second-round P. ovale qPCR used a one-to-fifty dilution of the first-round product, and positive samples proceeded to species identification. Blood collected during mosquito feeding was tested as well, allowing comparison of species detected in blood and mosquito midguts.

Table 2 reports detection and cross-reactivity for the Poc and Pow real-time PCR assays across 18S plasmid concentrations from ten to the power of five to ten to the power of minus one copies per microliter. For example, at ten to the power of three, the Poc assay produced five of five positives with a mean Ct of thirty-six point seven, while the Pow assay produced two of five at forty-six point three; the reciprocal pattern appears in the wallikeri assay panel.

Distilled-water controls also showed late Poc amplification, supporting the authors’ Ct threshold of below forty-nine for positivity. At higher parasite densities, cross-reactivity could misidentify a secondary species that is not actually present, while lower parasite densities create the opposite concern.

Mock mixtures were therefore used to mimic different Poc and Pow ratios and concentrations. Both species remained detectable in one-to-one mixtures down to their previous detection limits. The Pow assay appeared boosted by Poc at the lowest concentration, likely because of low-level cross-species reactivity that was absent from the corresponding single-species reactions.

In one-to-two and one-to-five mixtures, a cycle-threshold positivity cutoff below forty-nine preserved detection of the minor species, but Pow detection as the minor species was compromised at one-to-ten. An algorithm was developed to prevent false-positive detection of a secondary species in a mixed Poc and Pow infection.

False positives were more likely when the majority species was abundant, with a cycle threshold below thirty-nine, while a cross-reactive signal from the other species could resemble a small minor population above forty-four. The classification system identifies whether both species are present, rather than estimating their relative abundance.

In simulated single-species infections, the approach maintained one-hundred-percent specificity for both species, while mock mixtures produced eighty-four-percent sensitivity for Poc and ninety-four-percent sensitivity for Pow under the stated detection ranges.

Figure two presents a decision tree for interpreting eighteen-S real-time PCR results and identifying Plasmodium ovale species in single or mixed infections. It first asks whether either Poc or Pow cycle-threshold value is below thirty-nine; if not, values up to forty-nine can still support detection of either species.

When one species is clearly abundant, the other Ct value must be below forty-four to classify both species as present, helping limit false-positive calls from cross-reactivity. Among thirty-seven clinical blood samples from Cameroon and Tanzania, results were successfully obtained for seventy percent, or twenty-six samples, using qPCR, compared with thirty-five percent, or thirteen samples, using the published nested PCR conditions.

When nested PCR used more cycles or a larger template volume, successful amplification increased to eighty-five percent, or twenty-three of twenty-seven samples. Species identification was confirmed by Sanger sequencing of nested-PCR products from leukodepleted blood samples.

Among the nineteen samples successfully amplified by both methods, the assays showed good concordance and one-hundred-percent agreement for the major P. ovale species. Real-time PCR additionally detected a second minor species in two of those nineteen samples that nested PCR did not identify.

Across twenty-eight unique infected individuals, fifty-four percent were infected with Poc, thirty-six percent with Pow, and eleven percent had mixed Poc and Pow infections. Table three compares nested and real-time PCR species calls for field samples from Cameroon and Tanzania, using cycle-threshold values and marking concordant or discordant results.

It distinguishes P. ovale curtisi, P. ovale wallikeri, mixed calls, and samples that failed to amplify, including five samples that amplified in neither species assay. The table matters because it makes method agreement and unresolved classifications visible across dried blood spots and leukodepleted blood.

Figure three shows second-round nested PCR products separated on a one percent agarose gel, with the top panel targeting Plasmodium ovale curtisi and the bottom targeting Plasmodium ovale wallikeri. Bands near the expected seven hundred eighty-seven to seven hundred eighty-nine base-pair range for curtisi and seven hundred eighty-two base pairs for wallikeri indicate detectable parasite DNA in selected clinical samples, while controls provide reference points.

The visual illustrates how band presence was used to classify samples in Table three, including results obtained after increasing amplification cycles or template volume. Mosquito-based xenodiagnosis revealed a much higher prevalence of mixed Poc and Pow infections than was identified in human clinical blood samples.

Seventeen oocyst-positive midguts from mosquito feeds on nine infected people were available, and fourteen of seventeen midgut DNA samples amplified successfully in the species assays. Of those fourteen successful samples, eleven, or seventy-nine percent, were positive for both Poc and Pow under the classification scheme.

Only two of the nine carriers had mixed infections detectable in blood at the time of feeding. Mosquito-based xenodiagnosis showed that all but one of the nine carriers—eight of nine—harbored both Poc and Pow and transmitted both species to mosquitoes.

Table four compares species-specific real-time qPCR results in blood samples and the mosquito midguts produced by feeding on them. For each sample, cycle-threshold values are shown for P. ovale curtisi and P. ovale wallikeri, with the final column identifying Poc, Pow, or mixed infections; for example, blood sample MqTZ-3863 was classified as Pow plus Poc, while its midguts included both mixed results and a Poc plus Pow result.

The table matters because mosquito-based xenodiagnosis detected mixed infections in midguts, supporting the authors’ conclusion that this approach can reveal infections not fully represented by clinical blood testing. The modified species-specific assays and classification algorithm showed that mixed Poc and Pow infections occur naturally, at roughly eleven percent in the initial blood survey and eighty-nine percent by mosquito-based xenodiagnosis in this small sample.

Because the two species share the same geographic and temporal distribution, their co-transmission within the same Anopheles mosquitoes is not unexpected. Frequent co-transmission gives the two species ample opportunity to recombine within mosquitoes, even though a species barrier appears firmly established.

The analytical-sensitivity experiments and classification scheme used plasmid controls and were not specifically tested for robustness within blood or mosquito samples. The higher mosquito co-infection finding was surprising and needs validation, while lower assay sensitivity in mosquitoes is expected because of low parasite burdens and PCR inhibition.

The classification scheme is also expected to underestimate mixed infections when the majority species is present at high density. Because the Poc assay was slightly more sensitive than the Pow assay, the surveys cannot establish firm relative prevalence estimates for the two species.

The predominantly mixed midgut samples were not sequenced to verify both species, and some high-cycle-threshold midgut results might reflect contamination, although several observations make that less likely. A modified real-time PCR approach detected mixed P.

ovale curtisi and wallikeri infections in humans and especially mosquitoes, showing that co-infection and co-transmission may be substantially underappreciated.