Introduction

Plasmodium ovale, most probably first observed in 1900 by Charles F. Craig [1], was formally described and named by Stephens in 1922 because of characteristic oval-shaped infected erythrocytes seen in the blood of soldier on his return to England from East African [2]. Morphologically, P. ovale differs only slightly from P. vivax, making it difficult to distinguish these two species by microscopic examination. This delayed wide acceptance by the malaria community that P. ovale was indeed a distinct species and not a strain of P. vivax. Detailed morphological and biological observations of the course of P. ovale infection in experimentally infected neurosyphilitic patients and in mosquitoes brought conclusive evidence that it was indeed a species in its own right [3]. The biology, pathology and epidemiology of P. ovale were recently reviewed [4], [5]. Briefly, the parasite causes a clinically benign infection where peak parasitaemia does not generally exceed 10,000 parasites per µl, the infection often resolves spontaneously and the relapses due to the activation of hypnozoites are clinically mild and of lesser frequency that those in P. vivax infections. Long considered a parasite confined to Africa it has been recorded in all endemic areas, except America [6]. Nonetheless, its prevalence in sub-Saharan Africa, even in western Africa, rarely exceeds 5%, and it is only occasionally recorded in most of the endemic areas outside Africa.

The advent of molecular assays based on the amplification of genes specific to each parasite, most often the small subunit ribosomal RNA (ssrRNA), has greatly improved and the sensitivity of detection and the identification accuracy of P. ovale [7], [8]. The first epidemiological studies revealed that figures for the prevalence of P. ovale derived from microscopic examination of blood smears were likely to be a gross underestimate of the true values [8], [9], [10]. The observation that some isolates microscopically identified as P. ovale could not be detected by some of the assays based on the amplification of the ssrRNA gene (reviewed in [11]) led to the discovery that this gene in P. ovale occurred as one or other of two types, termed classic and variant [11], [12], [13]. The occurrence of two sequence types was also found for the other P. ovale genes that were analyzed [13], [14], [15]. Recently, two groups working independently (United Kingdom and Thailand) added to this data and concluded that these two variant types defined two distinct non-recombining sympatric species, a proposal made in a joint publication [16], in which they were named the classic type P. ovale curtisi and the variant type P. ovale wallikeri.

The existence of such two species raises a number of important epidemiological and biological questions that are best answered by gathering data from large numbers of cases infected by these species and from mosquitoes that transmit them. This requires a molecular assay for their detection and accurate identification. In a previous study [11], we had designed novel oligonucleotide primer combinations for use in a nested-PCR assay (NPCR) that could indiscriminately detect both P. ovale curtisi and P. ovale wallikeri. Moreover, in the same study it was observed that a real-time PCR assay (Rt-PCR) previously described by us [17] and currently used in our laboratory for the diagnosis of malaria, was able to detect P. ovale curtisi but not P. ovale wallikeri.

An Rt-PCR assay, recently described to distinguish between the two P. ovale species [18], is based on distinguishing the 120 bp fragments from the reticulocyte binding protein 2 (rbp2) gene from the two species by a difference in their melting temperature due to six species-specific single nucleotide polymorphisms. The sensitivity of that assay was not determined, though the results were concordant with a nested PCR based on another parasite gene, the P. ovale sp. tryptophan-rich antigen [18]. We wished to develop an Rt-PCR assay for P. ovale wallikeri based on the same highly conserved non-protein coding ssrRNA gene, that is the target of our previously developed Rt-PCR assays that we routinely use for the detection of the other Plasmodium species [11], [17].

Materials and Methods

Results and Discussion

The 31 samples included in this study (Table 1) formed a subset of samples collected prior to anti-malarial treatment from patients presenting to the University Hospital of Parma. Although initial routine microscopy had failed to detect P. ovale in many, the presence of this parasite was established by NPCR in previous studies [11], [17]. The parasitaemias ranged from <50 to 20,500 parasites/microlitre. All 31 samples were P. ovale positive by the NPCR assay specifically developed to detect either of the two P. ovale species [11]. The samples were negative for Plasmodium species other than P. ovale, except No. 11 and 14 which were also positive for P. falciparum and P. malariae, and No. 22 which was also positive for P. falciparum, as established by nested PCR assay [21].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Characteristics of P. ovale samples and species identification by microscopy, nested PCR and Real time PCR.

https://doi.org/10.1371/journal.pone.0048033.t001

Following optimization of the reaction conditions, the Rt-PCR assay for P. ovale curtisi and the new one for P. ovale wallikeri were evaluated at least in duplicate for their analytical sensitivity using serial dilution of purified plasmids into which the respective ssrRNA gene amplicons were cloned. Using a ten-fold dilution series with plasmid concentrations ranging 5×10⁸ copies per assay to a theoretical value of 0.01 copies, only template aliquots with 50 or more target copies were consistently positive (Ct values ranging from 16.7 to 38 for the highest concentration to the lowest). We noted that for assays with template aliquots containing 5 copies (whose presence was confirmed by NPCR) Ct values between 43 and 54 were obtained in >50% of the plasmid DNA replicates tested. This led us to conduct the assay over 55 cycles. We are aware that Ct values above 40 are generally considered of doubtful significance. However, according to the results we obtained by testing both clinical samples and 5 copies plasmid DNA a cut-off could be established at 54 cycles. Moreover, we feel confident of the specificity of the assay and that false positive results would be rare, since no signal could be detected by either the two P. ovale-specific primer/probe combinations when test assays were initiated with DNA obtained from healthy individuals or from individuals infected with P. falciparum (n = 9), P. vivax (n = 9), P. malariae (n = 5), the non-corresponding P. ovale species, or with a P. knowlesi DNA oligonucleotide corresponding to the target (data not shown). Moreover, no cross-reactivity could be detected with assay using genomic DNA purified from Toxoplasma gondii or Leishmania infantum, neither was a signal detected for any of the no template assays (data not shown).

The different PCR assays were applied to the genomic DNAs purified from the 31 samples collected prior to anti-malarial treatment from 31 patients presenting to the University Hospital of Parma. Twenty of these samples were found infected with P. ovale curtisi and 11 with the P. ovale wallikeri (Table 1). On a single occasion Ct values 53.47 and 50.57 were obtained in duplicate P. ovale wallikeri runs using template isolated from a sample, No. 30, with a very low parasitaemia (Table 1). In this case, the presence of P. ovale wallikeri was confirmed by nested PCR assay. A similar result was observed with sample No. 11 (P. ovale curtisi) (Table 1) in which the nested-PCR assay also confirmed the triple infection with P. falciparum and P. malariae.

The results we present concord with the initial observations on the two P. ovale species [16], [18], namely that the P. ovale wallikeri (previously known as the variant type) is not confined to southeast Asia since nearly all the patients acquired the infection in Africa, and that the two species are generally sympatric in the countries where they occur. Mixed infections with both P. ovale species were not found in any of the samples we have analyzed. Finally, it should be pointed out that nine of the patients presented to the hospital between three months and two years after their arrival to Italy. Thus, it is likely that the samples from these patients contained parasites from relapses, i.e. blood stage infections originating from the activation of a subset of the hypnozoites in the liver. This would indicate that relapses seem to occur in both species, though the low number of patients for which the period of latency was available precluded statistically significant comparisons of the relapse patterns of the two species.

The ability to detect and distinguish the two P. ovale species using the Rt-PCR protocol presented here opens the way to epidemiological investigations of these parasites. As a matter of fact, P. ovale is one of the least studied of the Plasmodium species that infect humans. In the context of the goal of malaria elimination and eventual eradication, it becomes important to investigate P. ovale, given that its prevalence is likely to be substantially higher than previously thought and that it can maintain itself in the human host for long periods as a result of its capacity to produce hypnozoites. Any meaningful investigations of the true epidemiology and biology of the two P. ovale species, whose infections lead to only relatively scanty parasitaemias (peaks of less than 10,000 parasites/µl) even in primary infections [4], will necessitate the application of sensitive and specific molecular methods of detection.