Cytometry in malaria: Moving beyond Giemsa

The article by Bhakdi et al. in this issue of Cytometry Part A on optimizing flow cytometric detection of mouse malaria parasites (1) is, somewhat remarkably, one of only a few dozen publications in the literature (2-54) in which cytometry is applied to the diagnosis, treatment, or biology of this protean disease. The World Health Organization estimates that 350–500 million cases of malaria occur annually, causing at least a million deaths. Of the 270–400 million cases of the severest form of the disease, due to the parasite Plasmodium falciparum, about 70% of these cases are in Africa and about 20% in southeast Asia (55). The diagnosis of malaria is primarily cell-based and involves visual detection of intraerythrocytic parasites by transmitted light microscopy in a peripheral blood smear stained with Giemsa's stain, a mixture of eosin and methylene azure dyes first described over a century ago (56). Identification of the various stages of parasites depends heavily on morphologic information, requiring observation at high power. Although it has been known for many years that methods based on fluorescence microscopy, using acridine orange (57-59) and other dyes (60-62), compare in accuracy with light microscopy (63) and may require less time and a less skilled observer, the required fluorescent microscope has, until recently, been too expensive for most laboratories in areas where malaria is most prevalent. If malaria were more common in affluent countries, we might expect that cytometry would, by now, have supplanted microscopy of Giemsa-stained smears for malaria diagnosis, just as it has for differential leukocyte counting and reticulocyte counting. Demonstrations of the efficacy of flow cytometry for detection, characterization, and counting of malaria parasites date back to the 1970s (20, 21); acridine orange was used for staining in apparatus with 488 nm laser sources, whereas Hoechst dyes, which are more DNA-selective, were favored where UV excitation was available. It was shown in the mid-1980s that different developmental stages of P. falciparum could be identified by flow cytometry based solely on nucleic acid content, without recourse to morphologic information (25, 27); unfortunately, most people in both the malariology and cytometry communities seem to be unaware of this highly significant finding. The genome of P. falciparum was sequenced in 2002 (64); the haploid genome is approximately 23 Mbp in size and contains 80% adenine + thymine (A+T), the highest such percentage found in any organism analyzed to date. P. vivax, P. malariae, and P. ovale, the other species that cause malaria in humans, have not been completely sequenced but appear to have haploid genome sizes of about 30 Mbp and contain about 60% A+T (65, 66). In clinical diagnosis, it is important to distinguish between malaria caused by P. falciparum and malaria due to the other species. In the former, peripheral blood contains a preponderance of haploid and near-haploid forms; in the latter, the later stages of the parasites, with DNA content as much as 16 times as high as that of the haploid forms, are also common in peripheral blood. One would therefore expect to be able to both detect malaria parasites at low levels of parasitemia and distinguish between cases due to P. falciparum and those due to other species by fluorescence cytometry using a DNA-selective stain. Many recent publications on cytometry in malaria (e.g.,16, 19, 39) have used asymmetric cyanine nucleic acid dyes of the SYTO and YOYO series (Molecular Probes/Invitrogen, Eugene, OR). These dyes, structurally related to thiazole orange (4), can be excited with blue or blue-green (488 nm) light and emit in the green or yellow spectral region. Unlike acridine orange, which quenches on binding to nucleic acids, the cyanines enhance fluorescence substantially on binding, typically by a factor of 1,000 or more; this results in lower background fluorescence, which makes it easier to detect smaller (haploid) forms of the malaria parasite. The cyanine dyes mentioned earlier, however, are not DNA-selective; other Molecular Probes products, e.g., Pico Green and Vybrant DyeCycle Green, have similar fluorescence characteristics and are highly DNA-selective. The last named dye also stains DNA stoichiometrically without the need for permeabilization or fixation and may well prove optimal for staining malaria parasites in both diagnostic and experimental settings. Although fluorescence microscopes can now be made substantially less expensive by using light-emitting diodes (LEDs) as light sources (67-69), the accuracy and precision of malaria diagnosis by fluorescence microscopy are still constrained by the limitations of the human observer (70). We believe that low-resolution fluorescence imaging cytometry, shown feasible in 1994 (71) and now possible to implement in a small, rugged, energy-efficient, and extremely inexpensive form using LED illumination and consumer-grade digital camera chips for detection (15,72), may soon be able to bring effective cytometric diagnosis to the resource-poor areas in which it is most sorely needed, as the malaria pandemic is not likely to subside in the foreseeable future. We encourage our colleagues to contemplate any and all cytometric approaches, including digital imaging, to bring malaria under control in the southern hemisphere.