Clearing of hemozoin crystals in malaria parasites enables whole-cell STED microscopy

Malaria is still the deadliest infectious disease caused by a parasite (World Health Organization, World malaria report 2021; https://www.who.int/teams/global-malaria-programme/reports/world-malaria-report-2021). An infection starts with the bite of an infected Anopheles mosquito. While probing for blood, Plasmodium sporozoites are injected into the skin of the host (Sidjanski and Vanderberg, 1997). During the journey of these highly motile cells through the skin, they invade blood vessels to be transported to the liver, where they transform and develop into thousands of merozoites, which are released back into the blood, invading red blood cells (RBCs) (Amino et al., 2006; Sturm et al., 2006). During mass proliferation within RBCs, the clinical symptoms of the disease occur. For continuous development and replication, the parasite relies on the degradation of hemoglobin to acquire nutrients, such as amino acids (Goldberg et al., 1990; Krishnan and Soldati-Favre, 2021; Sherman, 1977), and to make space for its growth (Esposito et al., 2008). Uptake of hemoglobin occurs through cytostome-derived invaginations of the parasite plasma membrane with subsequent formation of vesicles, establishing the digestive vacuole (DV) (Rudzinska et al., 1965). In the most virulent human-infecting Plasmodium species, P. falciparum, the DV appears as a single large organelle, whereas in P. vivax, P. knowlesi, P. ovale and in rodent-infecting P. berghei, hemozoin crystals are resident in smaller vesicles and distributed throughout the entire cytoplasm of the parasite (Fig. 1A). Degradation of hemoglobin within the DV leads to the release of toxic heme, which is rapidly converted into inert hemozoin crystals, which can readily be seen as black structures within the cell (Pandey et al., 2003). Several antimalarial drugs, such as chloroquine, interfere with the formation of hemozoin crystals. The extensive use of chloroquine led to parasites developing point mutations in a protein, termed the chloroquine resistance transporter (CRT) that is localized in the membrane of the DV (Fidock et al., 2000). CRT naturally transports amino acids and iron across the DV membrane (Bakouh et al., 2017; Shafik et al., 2020). Recently, endocytosis has also been defined as a major contributor to artemisinin ‘resistance’. Artemisinin resistance is mediated by mutations in the protein Kelch 13 (K13), leading to reduced endocytosis of hemoglobin (Birnbaum et al., 2020; Mesén-Ramírez et al., 2021). Hemozoin crystals are also found in other stages of the parasite, such as the gametocytes, ookinetes and oocysts, where they are ‘carried over’ from the blood stage.

Owing to the small size of the parasite and its organelles, super-resolution microscopy modalities, such as stimulated emission depletion (STED), structured illumination microscopy (SIM), photo-activated localization microscopy (PALM) and stochastic optical reconstruction microscopy (STORM) have become important techniques for gaining functional insights into parasite biology (de Niz et al., 2017). STED enables fast imaging at a resolution of 30–50 nm (Hell and Wichmann, 1994), whereas SIM permits a resolution of ∼100 nm (Gustafsson, 2000). PALM and STORM support a similar resolution to STED (Betzig et al., 2006), but have not been widely used in parasitology research yet.

STED nanoscopy of Plasmodium parasites has already been used for imaging of hemozoin-free life cycle stages (Kehrer et al., 2022; Prado et al., 2015; Volz et al., 2016). In contrast, imaging of hemozoin-containing blood-stage parasites, relies on the use of adaptive illumination techniques, such as RESCue STED or guided STED, which turns off the laser upon encountering a hemozoin crystal, as contact with the laser would result in immediate cell destruction (Mehnert et al., 2019; Schloetel et al., 2019). Adaptive illumination is particularly challenging when hemozoin is dispersed throughout the entire cytoplasm of the cell, such as occurs in P. berghei, and in addition to that it is so far only available for Abberior instruments. Hence, to date, no whole-cell STED of blood-stage parasites has been achieved.

Clearing of tissues dates back to 1914 and was initially developed for 3D imaging of whole organs to optimize optical properties, and is mostly solvent based (Dodt et al., 2007; Spalteholz, 1914; Tian et al., 2021; Tuchin, 1997). Clearing has been successfully used for imaging of Plasmodium-infected Anopheles mosquitoes (de Niz et al., 2020; Mori et al., 2019). In a recent screen of 1600 hydrophilic chemicals towards their clearing abilities CUBIC-P turned out to be the most effective for decolorization of red blood cells (Tainaka et al., 2018), suggesting that it extracts the heme from hemoglobin.

In this study, we show that hemozoin crystals can be effectively cleared from the malaria parasite using CUBIC-P while preserving the overall structure of the parasite. Decolorization and removal of hemozoin crystals from the DV of fixed P. berghei as well as P. falciparum parasites allowed whole-cell STED imaging of asexual and sexual blood stages as well as ookinetes, the mosquito-infecting motile parasite form. Importantly, the DV membrane could be imaged after CUBIC-P treatment, as demonstrated in a P. berghei parasite line expressing the DV-membrane resident CRT fused to the green fluorescent protein.

The presence of hemozoin crystals in Plasmodium blood-stage parasites limits super resolution imaging techniques, such as STED nanoscopy. CUBIC-P (TCI Chemicals) has been shown to clear red blood cells to optimize their optical properties, suggesting that it solubilizes the color-giving heme from hemoglobin (Tainaka et al., 2018). To test whether CUBIC-P is also able to decolorize the crystallized heme in the form of hemozoin in the Plasmodium parasite, we incubated paraformaldehyde (PFA)-fixed P. berghei-infected RBCs with different CUBIC-P concentrations at 37°C overnight. The following day, cells were washed and observed under a widefield microscope. In the CUBIC-P-treated cells, we observed a slight increase of autofluorescence in RBCs treated with either 10% or 50% CUBIC-P in our three standard imaging settings [Hoechst 33342 excitation (ex) 365 nm, emission (em) 445±50 nm; GFP ex 450±50 nm, em 515±50 nm; mCherry ex 546±12 nm, em 575–640 nm] (Fig. 1B). We furthermore observed swelling of RBCs, which is similar to what has been described before for hydrogel-based clearing techniques (Ono et al., 2019) (Fig. 1C). Both background fluorescence and cell swelling increased with higher CUBIC-P concentration. The size of P. berghei-infected RBCs increased 1.5-fold in cells incubated with 10% CUBIC-P and almost 2-fold for cells incubated with 50% CUBIC-P (Fig. 1C). We observed the same for RBCs infected with P. falciparum and incubated with 10% CUBIC-P (Fig. 1D). In P. falciparum, the hemozoin appears in the light microscope as one distinct area located inside the single large food vacuole. In contrast, the hemozoin in P. berghei is distributed across the entire cell in multiple DVs. Remarkably, none of the parasite stages (trophozoites, schizonts, gametocytes and ookinetes) of P. berghei as well as P. falciparum incubated with 10% or 50% CUBIC-P showed any remaining hemozoin crystals (Fig. 1E,F). We also tested clearing of hemozoin in ookinetes and early oocysts of isolated mosquito midguts 1 day post a blood meal. In midguts treated with CUBIC-P, we could not observe any remaining hemozoin crystals, whereas they were clearly present in non-cleared midguts (Fig. S1). Transmission electron microscopy (TEM) of cleared and non-cleared infected RBCs confirmed the absence of the entire hemozoin crystal, whereas endomembranes, including those likely belonging to the DV could readily be detected (Fig. 1G).

We next tested whether cleared P. berghei parasites are generally able to withstand the strong STED laser without cell damage. To do so, a drop of the respective cell suspension was placed on a microscope slide covered with a cover glass and imaged on an Abberior STED microscope. Parasite-infected RBCs were localized using Hoechst 33342 staining and to test whether the CUBIC-P-cleared parasites are able to withstand the strong STED laser, we made use of the RBC auto-fluorescence. Whereas non-cleared parasites immediately burst upon contact with the STED laser, our cleared parasites resisted even full laser power (Fig. 1H), which enables the imaging of blood-stage parasites with invisible hemozoin for future experiments.

To test whether CUBIC-P treatment would allow staining of selected organelles and structures of infected RBC using standard immunofluorescence protocols, we first aimed to visualize apical secretory organelles (rhoptries) in segmented schizonts and subsets of egress vesicles in gametocytes. Staining of merozoites with NHS-ester has been shown to particularly highlight rhoptries since they are very protein dense organelles (Liffner and Absalon, 2021). We stained cells treated with 10% CUBIC-P and did not observe any significant differences compared to non-treated cells indicating that treatment with CUBIC-P does not have a negative impact on rhoptries. In both situations, we observed a clear staining at the apical part of the cell (Fig. 2A). Gametocytes are particularly challenging to image by STED microscopy, given that they contain hemozoin distributed across the entire cell. For our experiments, we made use of our previously described PAT-GFP|g377-mCherry parasite line (Kehrer et al., 2016). In this line, the endogenous putative panthotenic acid transporter PAT and the gametocyte-specific protein g377 are C-terminally tagged with GFP and mCherry, respectively. Both proteins are essential for P. berghei gamete egress from RBCs in the mosquito midgut and localize to secretory vesicles. As the fluorescent proteins GFP and mCherry used for endogenous tagging in the PAT-GFP|g377-mCherry parasite line are not photostable enough for STED, antibody staining was required to image the vesicles with previously unachieved high resolution. Thus, gametocyte-containing blood was fixed with 4% PFA and cleared with 10% CUBIC-P followed by an immunofluorescence staining against GFP or mCherry in combination with the STED-compatible secondary antibody Atto647N. First observations on a confocal spinning disc microscope showed that both the Pat and the g377 signal were preserved after parasite clearing, demonstrating that antibody staining is still possible in CUBIC-P-treated cells (Fig. 2B). The increased auto-fluorescent background and cell size caused by the clearing agent did not markedly influence imaging. Furthermore, the single antibody stainings against GFP or mCherry were comparable to the initial GFP and mCherry signals; thus epitopes are still accessible for antibodies (Fig. 2B). Imaging of PAT-GFP|g377-mCherry gametocytes and ookinetes stained for anti-GFP in combination with Atto647 using the STED microscope did not result in the cell damage typically observed in non-cleared parasites (Fig. 2C,D), enabling the visualization of individual secretory vesicles at increased resolution, which will improve our understanding of parasite transmission into mosquitoes.

As PAT has been described to (partially) colocalize with g377 and the perforin-like protein pplp2, we additionally used PAT-GFP|pplp2-mCherry parasites (Kehrer et al., 2016). In both lines gametocytes were stained for GFP and mCherry, and colocalization of PAT with g377 or pplp2 was observed with STED microscopy (Fig. 2E). The extent of colocalization was quantified using the Pearson's correlation coefficient, where a value of one stands for perfect correlation and zero for no correlation (Fig. 2F). As described previously, the PAT signal correlated strongly with both, g377 and pplp2.

We found that CUBIC-P-mediated clearing of parasite-infected RBC allowed STED imaging of vesicular structures. To test whether we could image hemozoin-cleared cells at or close to the digestive vacuole we generated a P. berghei parasite line expressing the endogenous chloroquine resistance transporter (PbCRT) fused to GFP via single cross over integration (Fig. S2). Spinning disc microscopy of these PbCRT–GFP-expressing parasites showed that the GFP signal was associated with the black hemozoin crystals localizing to distinct spots within the parasite (Fig. 3A). STED imaging of cleared PbCRT–GFP parasites stained with an anti-GFP antibody (rabbit) followed by an anti-rabbit-IgG Atto647N-conjugated antibody, showed a vesicular pattern reminiscent of the vesicles shown by confocal microscopy. Direct comparison of the recorded STED and confocal image, of the same cell, clearly showed vesicular structures with a defined lumen in the STED image (Fig. 3B).

Here, we established a protocol to elute hemozoin crystals from P. berghei as well as P. falciparum blood-stage parasites, ookinetes and early oocysts using the hydrophilic clearing reagent CUBIC-P. Successful STED imaging of egress vesicles of gametocytes and the chloroquine resistance transporter CRT, as part of the DV membrane, demonstrates the potential of the method for super-resolution whole-cell imaging. Hence, this technique adds a super-resolution modality that should allow imaging of structures with a resolution of ∼50 nm, compared with the 100–150 nm possible with Airyscan or structured illumination techniques.

In the rodent-infecting model parasite P. berghei as well as in most human-infecting parasites (with the exception of P. falciparum), hemozoin crystals are widely distributed in the parasite. This makes it challenging to image larger areas at high resolution by RESCue STED. Our newly established method overcomes this difficulty and allows reliable whole-cell STED imaging of fixed cells. In addition to the use for imaging of infected RBCs, hemozoin clearing will be important for imaging gametocytes, ookinetes and early oocysts, where hemozoin crystals are widely distributed in small vesicles. These mosquito stages have long been neglected by researchers but are currently receiving more attention due to their interesting biology and potential for interference by transmission-blocking vaccines or mosquito targeted anti-parasitic chemicals. Of historical note, the discovery of the oocyst and hence the first experimental indication that malaria is transmitted by mosquitoes was essentially dependent on observing the small hemozoin crystals (Ross, 1897).

CUBIC-P contains 5% Triton X-100, 5% 1-methylimidazol and 10% N-butyldiethanolamine (Tainaka et al., 2018). Treatment of RBCs with CUBIC-P resulted in moderate swelling of the cells (Fig. 1) reminiscent of what occurs in expansion microscopy (Chen et al., 2015). Indeed, 20% 1-methylimidazol (CUBIC-X1) has been used to develop an expansion microscopy protocol based on aqueous chemicals without the need for embedding the tissue in a hydrogel (Murakami et al., 2018). A combination of CUBIC-X with CUBIC-P, as an alternative to the classic expansion microscopy protocol, or just the treatment of parasites with higher concentrations of CUBIC-P might thus be worth further investigation. This might allow to a further increase in the resolution that can be obtained, not only for hemozoin-containing life cycle stages but also for sporozoites and parasites developing in the liver.

Our data suggest that vesicular structures in cleared parasites remain intact and unaltered, as demonstrated for several proteins of egress vesicles. However, we observed that clearing interfered with osmium tetroxide contrasting of membranes for electron microscopy by de-lipidation of the samples. Still, we were able to properly recognize osmiophilic bodies in gametocytes and the quality of our TEM images was sufficient to confirm the absence of hemozoin. Whether the integrity of the cytoskeleton or other organelles, such as mitochondria, apicoplast or other vesicles is affected, needs to be investigated further.

In conclusion, clearing of hemozoin will now enable STED nanoscopy of Plasmodium blood-stage as well as in early mosquito stage parasites. This technique might not only be helpful for the investigation of Plasmodium spp. but also for other blood-feeding parasites.

100 μl of P. berghei of P. falciparum-infected blood was fixed with an equal volume of 4% PFA at 4°C for 48 h. For clearing, cells were washed with PBS, resuspended in 200 µl of 10 or 50% CUBIC-P in PBS and incubated for 24 h at 37°C. Post incubation, cells were washed twice with PBS and stained with respective antibodies and Hoechst 33342 (Sigma). For imaging, a drop of the resuspended cells was added onto a microscope slide and immediately covered with a cover glass.

Cleared samples were permeabilized with 0.5% Triton X-100 in PBS for 15 min at room temperature. After washing cells with PBS, primary antibodies in RPMI were added and incubated for 1 h at room temperature: rabbit anti-GFP (abfinity Roche; 1:40) or rabbit anti-mCherry (Abcam ab183628; 1:500), respectively. Cells were washed three times with PBS and incubated with the secondary antibody goat anti-rabbit-IgG Atto 647N (Sigma 40839; 1:500) in RPMI with 0.1 µg/ml Hoechst 33342 for 1 h at room temperature. Hoechst 33342 was added to the staining solution to visualize parasitic nuclei in infected RBCs. Cells were washed in PBS and stored at 4°C until imaged.

Imaging of cells was performed with a Zeiss CellObserver widefield microscope or with a Nikon spinning disc using either 63× or 100× objectives and exposure times between 50 and 500 ms.

STED microscopy was performed on an Expert Line STED system (Abberior Instruments GmbH, Göttingen, Germany) equipped with an SLM-based easy3D module and an Olympus IX83 microscope body, using a 100× oil immersion objective (NA, 1.4; Olympus UPlanSApo). STED images were acquired using 590 nm and 640 nm excitation laser lines in the line sequential mode with corresponding 615/20 and 685/70 nm emission filters placed in front of avalanche photodiodes for detection. A 775 nm STED laser was used for depletion with pixel dwell time of 10 to 15 μs and 15 nm xy sampling. In addition, confocal images were acquired using 405 nm and 488 nm excitation lasers and 525/50 nm emission filter.

Deconvolution of STED images was performed with Imspector software using the Richardson–Lucy function and 30 iterations or Huygens deconvolution (Scientific Volume Imaging) using classic maximum likelihood estimation (CMLE) algorithm and ‘deconvolution express’ mode with ‘Conservative’ settings. Image processing of widefield and spinning disc pictures including quantification of RBC sizes was undertaken manually with ImageJ/FIJI (Schindelin et al., 2012).

1832 bp of the PbCRT C-terminus was amplified from wtANKA gDNA with primers JK222 (5′-AAAGAATTCCTTATTACTATAGCCGTTGTAGAAAC-3′) and JK223 (5′-AAAGGATCCTGCCCTTGATGTTTCTATAGAAG-3′). The resulting PCR product was digested with EcoRI and BamHI, and ligated in frame with GFP into a vector containing the TgDHFR/TS as selection marker (pL18; modified from Kehrer et al., 2020 preprint). For transfection, the plasmid was linearized with AflII. Transfection of the linearized plasmid was performed as previously described (Janse et al., 2006). After electroporation of schizonts, positive selection was performed by adding pyrimethamine to the drinking water of mice.

Mouse experiments were approved by the Federal State Authority (Regierungspräsidium Karlsruhe) and performed according to Federation of European Laboratory Animal Science Association (FELASA) guidelines. P. berghei experiments were performed using Swiss CD1 mice (25 g, 6–8 weeks old) obtained from Janvier Labs.

20×10⁶ blood-stage parasites were injected intraperitoneally (i.p.) into a naïve (previously untreated) CD1 mouse. At 4 days post infection the blood was harvested by cardiac puncture to set up an overnight ookinete culture. The blood of one mouse was added to 10 ml of ookinete medium (RPMI containing 25 mM HEPES and 300 mg/l, L-glutamine, 10 mg/l hypoxanthine, 50,000 units/l penicillin, 50 mg/l streptomycin, 2 g/l NaHCO₃, 20.48 mg/l xanthurenic acid, 20% FCS; pH 7.8) at 19°C and incubated for 21 h. Fully developed ookinetes were purified via a 63% Nycodenz cushion (Axis-Shield Diagnostics) and fixed with 4% paraformaldehyde.

20×10⁶ blood-stage parasites were injected i.p. into a naive Swiss CD1 mouse. At 4 days post infection female Anopheles mosquitoes were allowed to feed for ∼15 min. Blood-filled midguts were isolated after 24 h fixed in cold 4% PFA for 45 s and washed with cold PBS. The epithelial cell layer was then opened longitudinally using two needles to carefully remove the blood bolus followed by a second fixation with 4% PFA overnight. Nuclei were stained with Hoechst 33342 and imaging was done on a Nikon spinning disc confocal microscope using a 100× objective.

Plasmodium-infected blood for TEM was fixed in 2% glutaraldehyde and 2% PFA in 100 mM cacodylate buffer at 4°C overnight. For final sample processing cells were rinsed three times for 5 min each time in 100 mM cacodylate buffer. Subsequently, samples were fixed in 1% OsO₄ in 100 mM cacodylate buffer for 60 min at room temperature. Then, they were washed twice with cacodylate buffer and twice with ddH₂O (5 min each washing step) and post-contrasted using 1% uranyl acetate in ddH₂O at 4°C overnight. Samples were again washed twice with ddH₂O for 10 min at room temperature. For dehydration, H₂O was replaced with acetone by incubation with increasing amounts of the solvent. To this end, samples were incubated with 30, 50, 70 and 90% acetone solutions for 10 min per step. Finally, two 10-min incubation steps with 100% acetone were performed. Subsequently, samples were infiltrated with 25, 50 and 75% Spurr (Serva Electrophoresis GmbH) for 45 min each followed by an infiltration step with 100% Spurr at 4°C overnight. Samples were embedded at 60°C overnight leading to polymerization of the resin. Sections of 60–70 nm were cut and contrasted with 3% uranyl acetate and lead citrate and were imaged using the transmission electron microscope JEM-1400 (JEOL).

We thank Marta Machado and Severina Klaus for donating P. falciparum 3D7 parasites and Julien Guizetti for comments on the article. We acknowledge support from the Infectious Diseases Imaging Platform (IDIP) at the Center for Integrative Infectious Disease Research, Heidelberg, Germany for microscopy support as well as the Electron Microscopy Core Facility (EMCF), Charlotta Funaya and Marek Cyrklaff for help with electron microscopy.

The peer review history is available online at https://journals.biologists.com/jcs/lookup/doi/10.1242/jcs.260399.reviewer-comments.pdf