Parasitological correlates of Plasmodium ovale curtisi and Plasmodium ovale wallikeri infection

Plasmodium ovale-positive, whole blood specimens stored in the malaria biobank at Public Health Ontario Laboratories (PHOL) from October 2006 to July 2015 were identified and retrieved. For the staining procedure, 2 mL of standard Giemsa stain was diluted in 48 mL of phosphate buffer (pH 7.1) prior to use. The phosphate buffer consists of 65 mL of Na₂HPO₄, 35 mL of NaH₂PO₄, and 900 mL of distilled water, resulting in 4.35 and 2.33 mM concentrations, respectively. For thin films, the slide was immersed in methanol for 10 s, allowed to dry, and then immersed in the working Giemsa solution for 10 min. The slide was then immersed in phosphate buffer solution for 10 s and air dried. Thick films were immersed in the working Giemsa solution for 10 min, immersed in the phosphate buffer for 30 s, and air dried. With each working solution prepared, a control P. falciparum slide was prepared and examined. Parasitaemia, morphological features, pan-aldolase antigen-positivity as determined by the BinaxNOW Malaria test (Alere, Ottawa, ON, Canada), year of import, and country of acquisition that were recorded in the biobank database following initial diagnostic processing were collected and analysed. The study was approved by the Research Ethics Board of Public Health Ontario.

DNA of all specimens was extracted using the DNA Mini Kit Blood or Body Fluid Spin Protocol (Qiagen, Germantown, MD, USA). Prior to use, DNA was stored at −20 °C.

Plasmodium falciparum/P. vivax species-specific duplex [6] and Plasmodium malariae/P. ovale species-specific duplex real-time PCR (qPCR) assays were conducted to confirm microscopy species identification as previously described [12]. All qPCR assays were run under the following conditions using the ABI 7900HT qPCR system: 50 °C for 2 min, 95 °C for 10 min, 95 °C for 15 s, and 60 °C for 1 min (45 cycles); 12.5 µl of TaqMan universal PCR master mix (Life Technologies, Burlington, ON, Canada) and 5 µl of DNA primers and probes with concentrations as previously reported [6, 12, 13] were used, for a final volume of 25 µl per reaction. All qPCR amplification curves were analysed using a manual threshold cycle of 0.02 and an automatic baseline. A cycle threshold (Ct) of <38 was considered to be a positive result.

18S rRNA gene copy numbers were quantified by running the P. malariae/P. ovale species-specific duplex qPCR assay and including serial dilutions of a P. ovale clone (ATCC, Manassas, VA, USA) alongside specimen DNA in triplicates. A linear regression was then constructed based on the logarithm of the gene copy number and Ct values for each concentration of the clone. This equation could then be used to calculate the gene copy number for each banked specimen.

Endpoint PCR of target regions, visualization of the amplicons on agarose gel, and Sanger sequencing were conducted as previously described [12]. Specifically, endpoint PCR of a 396-bp product from the 18S rRNA region was conducted with high-fidelity polymerase AccuPrime Pfx Supermix (Life Technologies, Burlington, ON, Canada) and 200 nM (each) of the primers Plasmo 18S forward (5′-ATTCAGATGTCAGAGGTGAAATTCT-3′) and Plasmo 18S reverse (5′-TCAATCCTACTCTTGTCTTAAACTA-3′). Using an ABI Veriti fast thermal cycler, the cycling conditions were 95 °C for 5 min, followed by 95 °C for 15 s, 58 °C for 30 s, and 68 °C for 30 s for 45 cycles, and then 68 °C for 5 min. Amplicons were visualized on 1% agarose gels with ethidium bromide prior to Sanger sequencing. For sequencing, the same forward and reverse primers were used along with a BigDye Terminator v3.1 cycle sequencing kit (Life Technologies, Burlington, ON, Canada), which was run according to the manufacturer’s recommended conditions. The products were purified by a BigDye XTerminator (Life Technologies, Burlington, ON, Canada) and analysed with an ABI 3130xl genetic analyzer. Forward and reverse sequences of each sample were aligned using Vector NTI software (Life Technologies, Burlington, ON, Canada), and the sequences were verified using a BLAST search to confirm the sub-species identification [14].

Primers for sequencing the dhfr-ts gene were newly designed for this study. All primers were designed using Primer3 [15, 16] and predominantly conserved regions of the complete sequence of the P. ovale dhfr-ts gene (GenBank: EU266602). Degenerate nucleotides were included based on the dhfr-ts gene for P. o. curtisi (GenBank: KP050414) and P. o. wallikeri (GenBank: KP050415). Endpoint PCR of the dihydrofolate reductase-thymidylate synthase region was conducted on all specimens with high-fidelity polymerase Phusion (New England BioLabs, Ipswich, MA, USA) and 200 nM (each) of the primers PocPowDHFR-exF (5′-YTCWACCTTCAGGGGTATCG-3′) and PocPowDHFR-exR (5′-AGTTTWAGCGTGGGRAAAGG-3′), generating an approximately 1700-bp product that covers the 99th amino acid to 1814–1826th amino acid of the sequence, where the end target varies based on the sub-species. The cycling conditions were 98 °C for 30 min, followed by 98 °C for 10 s, 64 °C for 30 s, and 72 °C for 1 min for 40 cycles, and then 72 °C for 10 min using an ABI Veriti fast thermal cycler. Visualization and Sanger sequencing of the PCR products were performed identically compared to the 18S rRNA methodology with the exception of the primers used. For sequencing, PocPowDHFR-exF, PocPowDHFR-exR, and a third internal primer, PocPowDHFR-inR (5′-TTTCCATTKGTTTCCCCTCT-3′) were used. Sequences for each specimen were aligned and verified using a BLAST search to further confirm the sub-species identification. Alignments of P. ovale sequences were performed using Mega6 software [17]. Genetic and peptide sequences between the enrolled specimens and complete P. ovale sequences that were previously reported (GenBank: EU266601, EU266604–EU266618, GQ250090, GQ250091, KP050409, and KP05413–KP050415) [2, 7, 18] were compared.

Reported parasitaemia, morphological features (including Schüffner’s stippling), pan-aldolase antigen-positivity, 18S gene copy number, year of import, and region of acquisition were compared between the two sub-species. Parasitaemia was treated as categorical data by recording if the percentage of parasites observed were <0.1 or ≥0.1%, as the former category is a convention used by PHOL laboratory technicians. Categorical variables were compared using Yates’ corrected Chi squared analysis or Fisher’s exact test, and continuous variables using Mann–Whitney U tests using GraphPad Prism 6 (GraphPad Software, CA, USA). Level of significance was set at p < 0.05.

Records of pan-aldolase antigen result and morphological features were absent for one specimen, resulting in the omission of this isolate from relevant calculations. Median parasitaemia (by microscopy) of specimens positive for P. ovale curtisi was <0.1% (range <0.1–0.8%), while for P. ovale wallikeri it was <0.1% (range <0.1–0.5%), and no difference was found (p = 0.99). There was no difference found in pan-aldolase antigen-positivity, where it was detectable in 11 (50.0%) specimens containing P. ovale curtisi and 11 (42.3%) specimens containing P. ovale wallikeri (p = 0.81). Pan-aldolase positivity for all specimens was 45.8%.

When comparing morphological features (Table 2), it was noted that all eight P. ovale parasites without Schüffner’s stippling were P. ovale wallikeri (p = 0.02) (Fig. 2). Parasitaemia (p = 0.20), 18S rRNA gene copy number (p = 0.50), and pan-aldolase antigen positivity (p = 0.40), were not found to be differentially associated with Schüffner’s stippling. Data on time between specimen collection and receipt were missing for two P. ovale wallikeri specimens without Schüffner’s stippling, and four P. ovale curtisi specimens, where three P. ovale curtisi specimens had Schüffner’s stippling and the other specimen had no morphological features recorded. No significant difference in time for specimen processing between specimens with and without Schüffner’s stippling was noted (p = 0.57).

Partial sequences of dfhr-ts could be obtained in 16 P. ovale wallikeri and seven P. ovale curtisi. In 26/49 isolates, the dhfr-ts region selected could not be amplified by the selected primer set. Several differences in the peptide sequences between P. ovale curtisi and P. ovale wallikeri were observed (Table 3). Non-synonymous mutations were observed within the P. ovale wallikeri isolates as previously reported [2]. Specifically, deletions of two or four amino acids (i.e., TA252–253 or TATA252–255) were identified in six and one P. ovale wallikeri isolates, respectively (Table 3). Furthermore, due to an ambiguity in the nucleic acid sequence, potential Y359F and Y366F mutations were observed. Finally, for one specimen, a H285Q mutation was recorded for one P. ovale wallikeri isolate. While Y359F and Y366F mutations are not expected to change any amino acid properties, a H285Q indicates a change from a basic histidine to a polar, uncharged glutamine. Among the sequences analysed in this study, polymorphisms specific to either sub-species were not observed (Table 4). One P. ovale wallikeri specimen did have a I169L mutation, although this does not reflect any changes in amino acid properties.