From Organ to Organelle: towards a Multimodal 3D Cell Atlas of the Malaria Parasite Mosquito stages using Correlative Light, X-ray and volume Electron Microscopy

Volume electron microscopy (vEM) is capable of imaging cellular ultrastructure with nanometer resolution while imaging all cells of a tissue or small organism. This usually happens in a destructive manner as in the case of focused ion beam scanning electron microscopy (FIB-SEM) and serial block-face scanning electron microscopy (SBF-SEM), which makes every sample with a biological event of interest extremely valuable. Another major bottleneck with these methods is the time-consuming image acquisition. For large samples with multiple regions of interest, image acquisition of the entire sample at high resolution can take up to weeks or even months. The midgut of the Mosquito vector has a volume of approximately 0.5 mm3, and to capture the entire volume of the midgut, image acquisition using an SBF-SEM would take up to 7 years and generate 200 TB of data [1]. When only sub volumes of a specimen are of interest for high resolution imaging, targeted workflows are powerful to address these bottlenecks. By utilizing X-ray imaging (µCT and Synchrotron CT) and light microscopy techniques it is possible to produce maps of the sample in a nondestructive way. Using these maps, rare biological events can be precisely targeted, made accessible for imaging using a volume electron microscope and reduces acquisition time substantially [2, 3]. The mosquito stages of the malaria causing parasite Plasmodium are in terms of volumetric electron microscopy largely unexplored, despite their discovery 125 years ago and almost half a million yearly deaths. This is also due to the inaccessibility of the parasite harbouring oocysts inside the mosquito vector. Using µCT imaging of dissected guts of the malaria vector Anopheles, it is possible to identify parasitic regions of interest the midgut, which in turn can be imaged using vEM upon targeting, visualising the organelles of a single parasite cell (Figure 1). By using a µCT, Synchrotron CT (SXCT) and confocal targeting workflow, mapping of different stages of oocysts in the midgut of the mosquito is possible at high resolution and volume electron microscopy of structures of interest is possible in a timely manner. In this case, a malaria infected mosquito midgut was first mildly fixed and then confocal scanned. After confocal scanning, the samples were prepared for classical electron microscopy (heavy metal staining, dehydration in ethanol, resin embedding and polymerization) and then scanned with a lab based µCT scanner (Bruker Skyscan 1127; Figure 2 A) and finally smaller regions of interest were scanned using SXCT. The imaging (µCT/SXCT/Confocal) is first correlated to identify regions of interest (Figure 2 B) (harboring malaria oocysts) before starting electron microscopy acquisition. Then the sample is processed for serial block-face (SBF-SEM) imaging and the and the SBF-SEM overview is correlated with the SXCT imaging to confirm the correct position (Figure 2 C). From this point, high resolution imaging of organelles is performed of the structures of interest (Figure 2 D). By correlating these different imaging modalities, it is possible to target precisely smaller regions of interest in bigger samples and to study rare cellular events. The method presented can be used in a similar way for a variety of samples [4]. A: µCT imaging of a midgut infected with malaria parasite, frame: region of interest. B: volume rendering from SBF-SEM imaging of region of interest in A. C: Segmentation of a single malaria parasite from the SBF-SEM dataset. Correlative, multimodal imaging and targeting of malaria parasites inside the mosquito gut. A: µCT volume rendering of an Anopheles midgut. B: Multimodal image correlation for targeting for volume EM imaging. C: direct single slice SXCT and SBF-SEM comparison to confirm correct position inside the sample. D: High resolution imaging of organelles inside the malaria oocyst and the mosquito tissue.