Improved limit of detection for zoonotic Plasmodium knowlesi and P. cynomolgi surveillance using reverse transcription for total nucleic acid preserved samples or dried blood spots.

Background: Zoonotic P. knowlesi and P. cynomolgi symptomatic and asymptomatic infections occur across endemic areas of Southeast Asia. Most infections are low-parasitemia, with an unknown proportion below routine microscopy detection thresholds. Molecular surveillance tools optimizing the limit of detection (LOD) would allow more accurate estimates of zoonotic malaria prevalence.

Methodology/principal Findings: An established ultra-sensitive Plasmodium genus quantitative-PCR (qPCR) assay targeting the 18S rRNA gene underwent LOD evaluation with and without reverse transcription (RT) for P. knowlesi, P. cynomolgi and P. vivax using total nucleic acid preserved (DNA/RNA Shield) isolates and archived dried blood spots (DBS). LODs for selected P. knowlesi-specific assays, and reference P. vivax- and P. cynomolgi-specific assays were determined with reverse transcription (RT). Assay specificities were assessed using clinical malaria samples and malaria-negative controls. The use of reverse transcription improved Plasmodium species detection by up to 10,000-fold (Plasmodium genus), 2759-fold (P. knowlesi) and 1000-fold (P. vivax and P. cynomolgi). The Kamau et al. Plasmodium genus RT-qPCR assay was highly sensitive for P. knowlesi detection with a median LOD of ≤0.0002 parasites/μL compared to 0.002 parasites/μL for P. cynomolgi and P. vivax. The LODs with RT for P. knowlesi-specific PCRs were enhanced for the Imwong et al. 18S rRNA (0.0007 parasites/μL) and Divis et al. real-time 18S rRNA (0.0002 parasites/μL) assays, but not for the Lubis et al. hemi-nested SICAvar (1.1 parasites/μL) and Lee et al. nested 18S rRNA (11 parasites/μL). The LOD for P. vivax- and P. cynomolgi-specific assays with RT were moderately improved at 0.02 and 0.002 parasites/μL, respectively (1000-fold change). For DBS P. knowlesi samples the use of RT also markedly improved the Plasmodium genus qPCR LOD from 19.89 to 0.08 parasites/μL (249-fold change); no LOD improvement was demonstrated in DBS archived beyond 6 years. The Plasmodium genus and P. knowlesi-assays were 100% specific for Plasmodium species and P. knowlesi detection, respectively, from 190 clinical infections and 48 healthy controls. Reference P. vivax-specific primers demonstrated known cross-reactivity with P. cynomolgi.

Conclusions/significance: Our findings support the use of an 18S rRNA Plasmodium genus qPCR and species-specific nested PCR protocol with RT for highly-sensitive surveillance of zoonotic and human Plasmodium species infections.