Plasmodium falciparum Liver Stage Infection and Transition to Stable Blood Stage Infection in Liver-Humanized and Blood-Humanized FRGN KO Mice Enables Testing of Blood Stage Inhibitory Antibodies (Reticulocyte-Binding Protein Homolog 5) In Vivo

Abstract

The invention of liver-humanized mouse models has made it possible to directly study the preerythrocytic stages of Plasmodium falciparum. In contrast, the current models to directly study blood stage infection in vivo are extremely limited. Humanization of the mouse blood stream is achievable by frequent injections of human red blood cells (hRBCs) and is currently the only system with which to study human malaria blood stage infections in a small animal model. Infections have been primarily achieved by direct injection of P. falciparum-infected RBCs but as such, this modality of infection does not model the natural route of infection by mosquito bite and lacks the transition of parasites from liver stage infection to blood stage infection. Including these life cycle transition points in a small animal model is of relevance for testing therapeutic interventions. To this end, we used FRGN KO mice that were engrafted with human hepatocytes and performed a blood exchange under immune modulation to engraft the animals with more than 50% hRBCs. These mice were infected by mosquito bite with sporozoite stages of a luciferase-expressing P. falciparum parasite, resulting in noninvasively measurable liver stage burden by in vivo bioluminescent imaging (IVIS) at days 5-7 postinfection. Transition to blood stage infection was observed by IVIS from day 8 onward and then blood stage parasitemia increased with a kinetic similar to that observed in controlled human malaria infection. To assess the utility of this model, we tested whether a monoclonal antibody targeting the erythrocyte invasion ligand reticulocyte-binding protein homolog 5 (with known growth inhibitory activity in vitro) was capable of blocking blood stage infection in vivo when parasites emerge from the liver and found it highly effective. Together, these results show that a combined liver-humanized and blood-humanized FRGN mouse model infected with luciferase-expressing P. falciparum will be a useful tool to study P. falciparum preerythrocytic and erythrocytic stages and enables the testing of interventions that target either one or both stages of parasite infection.

Keywords: Plasmodium falciparum; Plasmodium falciparum blood stages; clodronate liposomes; cyclophosphamide; humanized mouse model; reticulocyte-binding protein homolog 5.

Figure 1

Immune modulation with clodronate liposomes and cyclophosphamide leads to a stable Plasmodium falciparum blood stage infection. (A) Timeline showing the protocol for the repopulation of liver-humanized FRGN KO mice with human red blood cells (hRBCs) and subsequent infection with blood stage parasites. This protocol was utilized here to assess the effect of cyclophosphamide on a P. falciparum blood stage infection in FRGN huHep mice. Six mice received an injection of 50 µl CloLip i.v. + 50 µl CloLip i.p. on day −2 and three mice additionally received 150 mg/kg (approximately 150 µl per mouse) cyclophosphamide i.p. All animals were bled 200 µl, and received one i.v. injection of 500 µl hRBCs. On day −1, the animals were bled 100 µl and received an i.p. injection of 700 µl hRBCs. On day 0, the animals were bled 100 µl and received an i.v. injection of 500 µl hRBCs containing 1 × 10⁷ iRBCs. The following days, all mice received an individually determined amount of hRBCs. Parasitemia was followed by daily intravital imaging (B). The IVIS signal of mice treated only with CloLip started decreasing on day 2 postinfection. Only the mice which received both CloLip and cyclophosphamide, showed a stable P. falciparum blood stage infection throughout the 7-day observation period.

Figure 2

Protocol for the repopulation of sporozoite-infected FRGN mice with human red blood cells (hRBCs). (A) Mice are infected by mosquito bite with Plasmodium falciparum NF54 GFP-Luc sporozoites (n = 50 mosquitoes per mouse; 20 min). On the day of infection, mice are injected with 50 µl CloLip i.v. + 50 µl CloLip intraperitoneal (i.p.) and 150 mg/kg cyclophosphamide i.p. These injections are repeated on days 5, 9, 11, and 13 postinfection. On day 5 postinfection, 200 µl blood is drawn and the mice receive one injection of 500 µl hRBCs i.v. On day 6 postinfection, 100 µl blood is drawn and 700 µl hRBCs are injected i.p. The volumes of hRBCs injected the following days are determined individually in order to stabilize the percentages of hRBCs between 50 and 70%. (B) In order to assess the effect of antireticulocyte-binding protein homolog 5 antibody on transition and maintenance of a P. falciparum blood stage infection, blood humanization was achieved as described in (A). On day 5 postmosquito-bite infection, all mice received one i.p. injection of 100 mg/kg of R5.003, R5.016, or control anti-Ebola monoclonal antibody (mAb) (EBL040). All three mAbs were human IgG1. One mouse received PBS as a control. (C) The percentages of hRBCs were assessed by FACS analysis using a CD235ab antibody. All mice show similar percentages of hRBCs throughout the 14-day observation period, independent of the treatment regimen, demonstrating that the differences observed in blood stage infection were not due to varying amounts of hRBCs. The mean ± SD is shown for each treatment group.

Figure 3

Inhibition of Plasmodium falciparum blood stage infection by anti-reticulocyte-binding protein homolog 5 (anti-RH5) antibody. 10 mice were challenged with P. falciparum NF54 GFP-Luc infected mosquitoes (n = 50 per mouse; 20 min). Blood humanization and passive transfer of RH5 and control monoclonal antibodies (mAbs) was achieved using the protocol as depicted in Figure 2B. (A) Mice were imaged daily by IVIS starting 5 days postinfection. They were intraperitoneally injected with 100 μl of Rediject d-luciferin (Perkin Elmer) and imaged after 5 min for a 5-min exposure. Liver and blood-stage burden was assessed by placing an identical region of interest around each mouse and measuring total flux in pixels/second (p/s). The liver-derived IVIS signal peaks on day 6 postinfection and declines thereafter until day 10 due to the transition of the parasite from liver to blood stage infection. After day 10, the blood stage infection increases in all mice except for the ones treated with R5.016. In these mice, the infection decreases further and is undetectable by IVIS by day 11 postinfection. The mean ± SD is shown for each treatment group. (B) Representative IVIS images of R5.003 and R5.016 treated animals from days 6, 9, and 13. The liver-derived IVIS signal on day 6 is comparable for both mice. On days 9 and 13, the mouse treated with R5.003 shows a strong IVIS signal distributed throughout the body due to the increasing parasitemia, whereas the mouse treated with R5.016 shows only a very weak signal on day 9 and no signal is detectable on day 13, indicating that blood stage infection was potently inhibited. (C) P. falciparum parasites per ml of whole blood measured by quantitative reverse transcription (qRT-PCR) targeting P. falciparum 18S RNA as described previously (22) showing a dramatic decrease of infection levels only in the mice treated with R5.016. The mean ± SD is shown for each treatment group. (D) The parasite multiplication rate per 48 h was calculated for days 9–11 from the qRT-PCR data shown in Figure 3C. Controls refer to the mice treated with R5.003, Ebola mAb, or PBS. One control mouse showed low parasitemia on day 9 but reached levels comparable to the other mice on day 11, therefore the PMR is not comparable to the other mice and this data had to be excluded from the analysis. Each dot represents one mouse and the mean ± SD is shown. (E) Giemsa stained thin blood smears were analyzed daily by microscopy. Starting on day 8 postinfection, different developmental stages of asexual parasites could be detected in all animals except for the ones who received R5.016, where no parasites could be detected by microscopy at any time point.