Conventional cell-biological TEM applications and correlative microscopy

Besides light microscopy support, most of our projects within the institute are on ultra-thin sectioning of tissues, cells and cell components. We hereby collaborate with all departments and research groups having a wet lab. This involves individual projects to answer specific scientific questions as well as long-lasting collaborations. To give examples for individual projects, we recently imaged Leishmania tarentolae cells (collaboration with the Konthur group; Klatt et al., 2013, J. Proteome Res.), studied stress granules in Hela cells upon treatment with Fluorouracil and sodium arsenide (collaboration with the Krobitsch group) and analysed the location of proteins e.g. within mouse sperm cells (Dept. Herrmann). Examples for long-term collaborations are TEM studies related to stem cell differentiation (collaboration with J. Adjaye, MPIMG, now University Düsseldorf), cancer (collaboration with e.g. M. Schweiger, MPIMG, now University of Cologne) and other disease processes such as segmental progeroid disorders (collaboration with Björn Fischer-Zirnsak, research group Mundlos).

Furthermore, the microscopy group collaborated for many years successfully with a large number of national and international research groups e.g. on degenerative brain diseases (AG Wanker, MDC Berlin; AG Multhaup, FU-Berlin). Classical mica-adsorption techniques have been applied to analyse protein-DNA interactions involved in bacterial and yeast replication (AG Speck, Imperial College, London; AG Zawilak-Pawlik, Institute of Immunology and Experimental Therapy, Wrocław; AG Bravo, CIB, Madrid). Another focus lay on the structure, assembly and infection-mechanism of bacterial phages such as SPP1 infecting B. subtilis (collaborations with P. Tavares, CNRS, Gif-sur-Yvette; P. Boulanger, Université de Paris-Sud, Orsay; C. Breyton, CNRS, Grenoble and M. Loessner, ETH Zürich, amongst others). Originally introduced by the former funding director of the institute, Thomas Trautner, SPP1 phages until today serve as a model system for eukaryotic dsDNA viruses. Some of these collaborations ended with the retirement of Rudi Lurz, others are continued e.g. the analysis of DnaA binding to the oriC region in Helicobacter pylori (Donczew et al., 2014, J. Mol. Biol.) and cryo-EM studies on phage assembly intermediates (together with P. Tavares and E. Orlova, Birkbeck College, London).


Figure 1: (A) Montage of 1080 TEM images showing the entire mushroom body calyx region from the Drosophila brain. (B) Zoom into the region marked in (a). Several bouton regions are visible, one of which showing a synaptic active zone is highlighted in (C).

In collaboration with Stephan Sigrist (FU-Berlin), we applied electron tomography to study electron dense specialisations (the so-called “T-bars”) at neuromuscular terminals of Drosophila larvae, comparing the effect of different isoforms of the “Bruchpilot” protein on proper T-bar formation (Matkovic et al., 2013, J Cell Biol). T-bars represent a protein scaffold, which structures the release of synaptic vesicles at the presynaptic active zone. To proof the hypothesis, that age induced memory impairment (AMI) which is also observed in flies and which can be supressed by restoring juvenile polyamine levels (Gupta et al., 2013, Nature Neuroscience) is related to structural rearrangements at the active zone, we implemented Leginon (Suloway et al., 2005, J Struct Biol) for automated data collection on ultrathin-sections. Leginon enables us to image regions up to 500 µm2 at a pixelsize of ~1 nm (Fig. 1). The acquired micrographs (> 1000 per region of interest) are then stitched to one large image using FIJI. Imaging the calyx-region of Drosophila brains from animals at different age and feeding conditions, we could show statistically that the number of active zones and the density of synaptic vesicles per synaptic bouton decreased with age. Spermidine treatment did not restore these changes, instead, it supressed an increase of both, the average size of T-bars and the average number of Bruchpilot molecules within this scaffold, indicating significant restructuring at the active zone with increasing age (Gupta et al., 2017, PLoS Biol).


Figure 2: Correlated light and electron microscopy. Left: Superposition of a confocal fluorescence microscopy image and a TEM montage (480 images recorded at 15.000x magnification) acquired from immune-labelled ultrathin sectioned human iPS cells. The stem cell specific marker protein TRA1-60 was localized using a secondary antibody conjugated with Alexa488 (green) and 5nm immunogold. Blue: DAPI-stained nuclei. Right: Zoom into the marked region of the TEM montage. Arrows indicate immunogold labelled TRA1-60 molecules.

In collaboration with J. Adjaye (University Düsseldorf), we further applied automated TEM to study the differentiation of induced pluripotent stem cells (iPSCs). We found that bile canaliculi, a unique ultrastructural feature of hepatocytes, start to form within the hepatic endoderm (HE) state. Similarly, we could show that iPSCs derived from patient fibroblastes, which potentially can be used as a model to study rare mtDNA-linked genetic diseases, develop a typical triangular shape upon differentiation towards neuron progenitor cells (collaboration with A. Prigione, MDC Berlin, Lorenz et al., 2017, Cell Stem Cell).

These findings clearly demonstrate the potential of automated TEM to visualize the ultrastructure of entire cells and excerpts from tissues. In order to combine ultrastructural information provided by TEM and the power fluorescence microscopy which however is limited to sub-micrometer resolution according to Abbe’s law we now aim at implementing correlative microscopy approaches. In preliminary experiments, we could localize the stem cell specific marker protein TRA1-60 using both, confocal light microscopy and automated TEM-imaging on immuno-labelled ultrathin sections (Fig. 2). In collaboration with M. Schweiger (now University of Cologne) and G. Schönfelder (Bundesinstitut für Risikobewertung, Berlin) we are currently testing correlative approaches including Tokuyasu cryo-sectioning to study cultured cancer cells and Xenocraft tumor models, also aiming at potentially linking ultrastructural information (e.g. the occurrence of autophagosomes) to sequencing data obtained by our in-house sequencing unit.


Go to Editor View