Towards an in vitro model of Plasmodium hypnozoites suitable for drug discovery

Abstract

Background: Amongst the Plasmodium species in humans, only P. vivax and P. ovale produce latent hepatic stages called hypnozoites, which are responsible for malaria episodes long after a mosquito bite. Relapses contribute to increased morbidity, and complicate malaria elimination programs. A single drug effective against hypnozoites, primaquine, is available, but its deployment is curtailed by its haemolytic potential in glucose-6-phosphate dehydrogenase deficient persons. Novel compounds are thus urgently needed to replace primaquine. Discovery of compounds active against hypnozoites is restricted to the in vivo P. cynomolgi-rhesus monkey model. Slow growing hepatic parasites reminiscent of hypnozoites had been noted in cultured P. vivax-infected hepatoma cells, but similar forms are also observed in vitro by other species including P. falciparum that do not produce hypnozoites.

Methodology: P. falciparum or P. cynomolgi sporozoites were used to infect human or Macaca fascicularis primary hepatocytes, respectively. The susceptibility of the slow and normally growing hepatic forms obtained in vitro to three antimalarial drugs, one active against hepatic forms including hypnozoites and two only against the growing forms, was measured.

Results: The non-dividing slow growing P. cynomolgi hepatic forms, observed in vitro in primary hepatocytes from the natural host Macaca fascicularis, can be distinguished from similar forms seen in P. falciparum-infected human primary hepatocytes by the differential action of selected anti-malarial drugs. Whereas atovaquone and pyrimethamine are active on all the dividing hepatic forms observed, the P. cynomolgi slow growing forms are highly resistant to treatment by these drugs, but remain susceptible to primaquine.

Conclusion: Resistance of the non-dividing P. cynomolgi forms to atovaquone and pyrimethamine, which do not prevent relapses, strongly suggests that these slow growing forms are hypnozoites. This represents a first step towards the development of a practical medium-throughput in vitro screening assay for novel hypnozoiticidal drugs.

Figure 1. Slow growing and normally developing pre-erythrocytic (PE) forms.

Two types of hepatic parasites can be clearly distinguished from days 5 post-infection onwards, in in vitro cultured human or M. fascicularis primary hepatocytes infected with P. falciparum or P. cynomolgi sporozoites, respectively. The representative photomicrographs were made on cultures fixed on Day 5 and Day 11 post-infection. The parasite and host nuclei are stained with DAPI (blue), while the parasites are labelled by an antibody specific to the HSP70 of the two parasite species (green). Mature PE forms (Mat-PE) and uninucleate small PE forms (PE-uni) are clearly distinguishable.

Figure 2. Maturation dynamics of cultured P. falciparum and P. cynomolgi hepatic forms.

The proportion of PE forms at different stages of maturation (as assessed by the extent of nuclear division) is presented for cultures of increasing age. Uninucleate parasite forms correspond to the small forms (PE-uni) shown in Figure 1, while all others correspond to the mat-PE forms. The data was representative of three independent experiments.

Figure 3. Dose-response curves for primaquine, atovaquone and pyrimethamine against the distinct hepatic forms observed for P. falciparum and P. cynomolgi.

The infected primary hepatocyte cultures were exposed to the varying drug concentrations from Day 5 to Day 8 post-inoculation and fixed on D8. The curves obtained for the PE-uni forms are presented in red, while those for the Mat-PE forms are presented in black. The data was representative of two independent assays for pyrimethamine (for both parasite species) and for primaquine (P. falciparum); and of three independent experiments for atovaquone (both parasite species) and primaquine (P. cynomolgi). *, p<0.05.

Figure 4. Selection of P. cynomolgi PE-uni forms prior to screening for hypnozoiticidal activity.

Cultures were treated with atovaquone (ATQ) at 67 ng/ml (182 nM) for three days starting on Day 5 post-inoculation in order to eliminate the mat-PE parasites. Subsequently the cultures were either left untreated, treated for a further two days with ATQ (as above), or primaquine (PQ) at 10 µg/ml (22 nM). The cultures were then stopped by methanol fixation on Day 10 and the hepatic forms present enumerated. The data was representative of three independent experiments. *, p<0.05.