Nanoscopy

The key objective of the M4I Division of Nanoscopy is to gain greater insight into the 3D form of cell proteins, thus paving the way for the development of more effective treatments for diseases such as cancer and tuberculosis. A better understanding of how protein complexes manage healthy, but also diseased, cells will allow drugs and vaccines to target problems more effectively.

For instance, professor Peter Peters and his team are working to improve the vaccination presently used against tuberculosis. This research is based on Peters’s finding in 2007 of how the bacteria that cause tuberculosis behave in within cells. Each day, several thousand people die of tuberculosis, especially in developing countries. The M4I Division of Nanoscopy also aims to develop greater insight into the workings of the immune system, which may in the future lead to an immune response against cancer cells in the human body. 

Project 1 T7SS

In 2007 our team published an article in Cell showing that after prolonged infection in macrophages and dendritic cells, M. tuberculosis translocated from phago-lysosomes to the cytosol and killed the host cell a few days later, while the BCG vaccine strain failed to translocate. We found that this process was dependent on a gene in the extended RD1 region (ext-RD1). The translocation to the cytosol was unexpected, as it contradicted the prevailing dogma. We then focused on BCG with a knock-in of the entire ext-RD1, as this region is now known to encode the ESX-1 secretion system, which was recently identified by one of our former team member as a novel type VII secretion system (T7SS) (Abdallah AM et al., Nat Rev Microbiol. 2007; see figure below). We found that these bacteria translocate to the cytosol of the host cell 7 days after infection. We concluded that the ESX-1 system (in a BCG background) is sufficient for translocation (Abdallah AM et al., J of Immunol. 2011 and Houben et al., Cell Microbiol. 2012). Therefore we believe that the T7SS is the key to virulence in tuberculosis. Using cryo-EM, we therefore investigated the structure of the mycobacterial capsular layer and showed that high levels of the ESX-1-secreted proteins were present in the capsule (Sani et al. 2010).

We are now using cryo-EM single-particle analysis (SPA), and over the next years will investigate recombinant purified proteins of individual gene products from the T7SS. In addition, we are purifying the entire intact T7SS structure (or mutants thereof) using biochemical methods for 3D reconstruction. Our ultimate aim is to image mycobacteria in vitreous sections in human host cells while translocating. The cryo-SPA, X-ray, and NMR data of the T7SS can then be docked into the images from vitreous sections to construct a macromolecular map of the tubercle bacillus within the host cell. The broader objective is to gain insight into the structure and function of the mechanism for type VII-mediated translocation. This should lay the groundwork for the development of novel antibiotics and better vaccines.

Plunge freezing will be used to obtain a thin vitrified film containing isolated mycobacteria whose outer cell wall has been removed. This will then be subjected to cryo-EM imaging. Purified protein complexes composed of RD1-gene products will be obtained from Stewart Cole in Switzerland, Ravi Ravishankar in India and other scientists worldwide, and subjected to SPA as well as tomographic imaging. This process will give rise to 3D images that can be used as models for host-pathogen interaction. The approaches will be developed with Mycobacterium BCG: RD1 in human cells (adapted to biosafety level 2) as a model for host-pathogen interaction, leading to insights that can be used in the development of new tuberculosis vaccines and drugs. An ERC advanced grant draft for this project is currently being prepared.

Project 2 Cryo-electron tomography

Cryo-EM currently provides the ultimate resolution in single-particle analysis (SPA) (<0.5 nm) and in cryo-ET (<2 nm). However, the high information content and low contrast of cryo-EM images often makes it impossible to locate the subcellular structure of interest for cryo-ET. One possible solution may be state-of-the-art correlative light/electron microscopy (CLEM) studies. In these studies, interesting cellular features are identified by imaging GFP-tagged proteins (available for almost all gene products of the genome) using live-cell microscopy, followed by rapid immobilization and cryo-ET. Put simply, cryo-ET provides the extreme resolution needed to construct the detailed 3D image, whereas the proteins are identified by their genetically encoded fluorescent labels. Currently, the limited resolution (~250 nm) of conventional fluorescence microscopy makes it impossible to unequivocally localize the fluorescent labels to the nanomachines. 

To develop this solution, we intend to modify the FEI dedicated cryo-LM holder capable of maintaining the frozen-hydrated EM grid in its vitrified state (recently developed at the Max Planck Institute in Martinsried) for use on the super resolution LM  such that it can work under cryogenic conditions. Using this setup, we will then apply PALM principles to obtain optical  confocal and later super resolution (SR) cryo-images that identify individual proteins in nanomachines at particular subcellular sites. The preparation will subsequently be transferred to the cryo-EM under cryogenic conditions for ET imaging at sub-nanometer resolution. We will screen the samples with a cryo-TEM (200kV) to select those suitable for cryo-ET at NeCEN. Finally, we will deliver combined image modalities with upwards of 20nm correlation accuracy to identify individual macromolecular machines, which should provide a glimpse into the molecular makeup of the cellular complexes at unprecedented resolution levels.

Successful completion of this task depends on development of several new tools:

• An existing cryo-LM holder will be adapted to fit the SR microscope. This is a challenging task because the potential thermal drift on the stage, caused by liquid nitrogen in the cryo-LM holder, must be kept to an absolute minimum. Further, as immersion oil cannot be used at these temperatures, we will employ a 100 x, 0.9 NA dry objective (Fig. 1).
• We will further develop the MAPS navigation tool, which allows for correlation of the SR images with the cryo-EM images with a relocation precision greater than 50 nm. This requires updating of the existing software to high standards and the introduction of fiducial markers.

This project will benefit from a parallel project to develop novel cryo-ultramicrotomy. In particular, we will collaborate with Diatome to adapt their (room-temperature) oscillating diamond knife to obtain cryo-sections with almost no compression in the preparation.

The project will also benefit from another parallel side project with Maastricht Instruments. Briefly, because intact cells are too thick to be imaged by means of EM without sectioning, we will develop an alternative method called the “double grid approach”. In this device, the cells spread voluntarily to ~200 nm thickness, and can be vitrified with our Vitrobot for ET imaging in a close-to-native state. Maastricht Instruments will integrate an objective lens inside our Vitrobot to observe the migrating cells on a golden EM support grid and manufacture “double film” EM grids. The imaging and screening of these thick preparations by way of cryo-EM requires a 200 kV.

The developments described here will lead to a better understanding of the positions and states of nanomachines in normal and diseased cells. We will base validation studies of the new instrument on characterization of the new type VII secretion system (see project 1) involved in this translocation. The advances from these studies will greatly benefit not only cryo-ET, but also the SR imaging community. For example, electron microscopists have long known that chemical fixation methods introduce structural artifacts at tens of nm, whereas SR microscopy in STED and PALM mode is routinely performed on formaldehyde-fixed material. This limits the interpretation of high-resolution SR images from a biological perspective. Cryo-SR LM will image proteins in a much closer-to-native preparation. Importantly, the rate of fluorophore bleaching at cryogenic temperatures will be significantly reduced, which means more photons will emit from each fluorophore, resulting in increased resolution. In theory, this effect should far outweigh the lower light-collecting properties of the dry objective.

To achieve the goals laid out in the STW proposal, industry practitioners must develop new tools in collaboration with academic partners. The interest among companies in realizing this aim is evident from their generous offers to contribute. The project will facilitate the development of commercial products that will be in high demand among cryo CLEM labs (academic and industrial) worldwide; namely, a cryogenic chamber, the integration of this chamber with a state-of-the-art SR microscope, and tools to facilitate rapid relocalization of the preparation from the SR setup to the EM. 

Project 3 Nanofluidic chamber​

Cryo-ET of vitreous sections is an emerging technique that produces high-resolution (2–4 nm) structures of macromolecular complexes in frozen-hydrated state within their native environment. High-pressure frozen cells are sectioned into long ribbons 50nm thick, and viewed using cryo-EM. This approach is technically demanding, and it is difficult to ascertain whether the macromolecular complexes are disrupted by the freezing and sectioning processes. While it has been shown that whole bacterial cells can be visualized without sectioning (Kühner et al. Science 2009), the electron scattering prevents high-resolution imaging for organic samples thicker than 200 nm, which renders mammalian cells (at several microns) too thick. As an alternative to vitreous cryo-sectioning, we recently showed by way of live-cell light microscopy that migrating white blood cells (DCs, T-cells, etc.) can naturally migrate into a 1000 nm thick “nanochamber”. Using nano technology, we now intend to develop a method to vitrify the cells in this state. They are then sufficiently thin to be subjected to a cryo-FIB/SEM procedure making them 200nm, and thus useful for cryo-ET studies of macromolecular complexes within native mammalian cells without the need for cryo-sectioning.

The M4I Division of Nanoscopy will develop this chamber with silicon-based microelectromechanical systems (MEMS) technology in collaboration with colleagues at TU Delft. Two main challenges need to be addressed. First, a chamber must be designed that is even more transparent than the cell, yet still mechanically robust. This chamber should fit seamlessly with the existing high-tech equipment: cryo-EMs, cryo-fluorescent microscopes, and cell preparation tools. Second, high throughput methods must be developed to use the chamber efficiently and obtain resolution within the nanometer range. The envisaged device will allow us to study the behavior of intact mammalian cells and their responses to drug-induced changes at the molecular level. The project is ‘high risk’, but could open up new avenues for nanoscale research on intact mammalian cells. If successful, we intend to investigate human dendritic cells infected with M. marinum as well as the macromolecular interaction of T7SS with the phagolysosomal membrane. 

Project 4 Linezolid effects

At present, drug-induced changes of macromolecular targets cannot be observed within the native cellular environment. Ideally, a technique must be developed that allows for the observation of high-resolution macromolecular structure changes within this environment. One promising antibiotic for the treatment of multi-drug resistant (MDR) tuberculosis is linezolid.

Linezolid acts by inhibiting bacterial protein synthesis. The available X-ray crystal structures of linezolid in complex with 50S ribosome of Haloarcula marismortui (an archaeon) suggest that linezolid may inhibit the formation of the ribosomal initiation complex (Joseph et al., J Med Chem. 2008). As yet, however, the mechanism of action of linezolid on M. tuberculosis has not been definitively identified. This project focuses on the effect of linezolid on ribosomes isolated from mycobacteria, with a view to answering the following questions:

• Does linezolid prevent the formation of functional mycobacterial ribosome initiating complexes in vitro?
• Does linezolid have the same mechanism of action in vivo (within the context of the entire bacterial cell)?

We will evaluate the quality of the ribosome preparation by means of cryo-EM. Using the same technique, we will also study the effect of linezolid on the formation of mycobacterial 70S ribosomes and determine the optimal concentration for inhibiting complex formation. Subsequently, we will investigate the effect of linezolid on in vitro purified mycobacterial ribosomes and on M. tuberculosis cells to determine the minimal inhibitory concentration of antibiotic that has a bactericidal effect on our strains. To this end, we will use fluorescent microscopy to observe bacterial cell death with Live/DEAD BAC/Light assay. Finally, we will prepare M. tuberculosis treated with linezolid for cryo-EM of vitreous sections using high-pressure freezing (HPF) and vitreous cryo-sectioning techniques (Pierson et al. J Structural Biology 2011) or cryo-FIB/SEM lamellae to identify the mechanism of action of drugs on ribosomes inside entire bacterial cells.

This project is intended as a prototype for structural characterization of mycobacterial ribosome. It will provide the research team with scripts and protocols for SP-cryo EM designed to tackle specific issues that may arise during ribosome reconstruction.

Project 5 Technology improvement

If a specimen merely changes the phase of the electron wave (i.e. the specimen is a phase object), provided the electron microscope is at zero focus and there are no aberrations, there will be nothing to observe. That information was obtained from biological specimens in the past was partly due to the poor optical performance of the TEMs, and partly due to the use of a relatively large defocus. However, a significant amount of information is then lost. Contemporary electron microscopes can be equipped with an aberration corrector, and at zero focus on a phase object virtually nothing will be observed. Yet switching to off-focus also results in a loss of information. Thus, a phase plate – similar to the Zernike phase plate in light microscopy – would clearly be desirable. This phase plate should effect a phase shift of pi/2 on the central beam compared to the diffracted beams, and should not block the relevant electrons. In biology, all information beyond 0.01 Å – apart from the phase-shifted central beam – should be passed up to 1 Å through the phase plate. The phase plate must be located in the rear focal plane of the TEM. It will allow all phase information to be translated into amplitude information, which will vastly improve the imaging of biological specimens.

Several research groups have experimented with phase plates. Two types of phase plates can be identified: a) a thin sheet of C, for instance, can be placed such that all diffracted beams have a phase shift of pi/2; or b) the central beam is led through a small tube that effects a phase shift on the central beam. The second approach is the more appropriate. However, although such phase plates have been used successfully for over 5 years, no commercial product is yet available. This can be attributed to the difficulties involved in developing them. Not only must they be extremely small (with ~50 nm precision); the issues of contamination and charging also pose problems. In collaboration with Henny Zandbergen and his team at the Kavli Nanoscience Institute (TU Delft), we have performed tests showing that we can make apertures with the required degree of precision. As we also have solution to the contamination and charging problems, we aim to develop a reliable phase plate of the appropriate dimensions.

Project 6 IN-SENSE

Chronic mental illnesses (CMI) like schizophrenia or the recurrent affective disorders are among the most debilitating (prevalence 1-10%) and costly diseases in the Western world. Research on brain diseases is a focus in FP7. In fact, direct healthcare costs of the two disease categories that IN-SENS is directly addressing, the psychotic and affective disorders cost four times as much as dementia and other neurodegenerative disorders. This is mainly due to the early onset of CMI in adolescence which leads to in most cases lifelong impairments and occupational disability. Yet, even though CMI have dramatic impact on individual patients, their relatives, and society, our biological understanding and, accordingly, our options for novel strategies of efficient treatment or cure of CMI are stalling.

The last decade has brought breakthroughs in the genetics of CMI, like the discovery of the disrupted-in-schizophrenia 1 (DISC1) gene (by a member of this consortium), and others. The identification of these candidate genes now allows analysis of the INter- and intracellular Signalling INSchizophrenia (hence the project acronym IN-SENS). A European perspective is necessary since in most EU countries the infrastructure for the necessary interdisciplinary research is not met on national levels. In Maastricht we will be doing cryo-immunogold EM to localize DISC1 in the healthy and diseased brain.

Project 7 Model tuberculosis

Start date 2014

An urgent task of today’s medical research is to find new ways to treat infections caused by Mycobacterium tuberculosis. There are numerous reports on multidrug-resistant strains of this pathogen, which causes tuberculosis (TB). Most experimental studies of TB rely on animal experiments and there are no existing alternative models such as in vitro tissue models, with the exception of infected single cell cultures. Replacing existing animal models with human in vitro systems is also important because of the questioned validity of these models for TB. The aim of the presented project is to further develop two novel human in vitro models for TB, validate them against tissues from TB patients and to use them instead of animals for testing anti-TB combination regimens based on existing drugs. The models are based on human primary cells and cell lines and the investigation of these models as substitutes for animal models for preclinical drug testing.

The first model is based on macrophages obtained from human blood that are infected with virulent M. tuberculosis. Similar models have been described, but the unique feature of the presented model that the model displays is that under the right conditions, it can mimic clinical latent TB. Understanding latent TB is of importance, since the latent infection remaining in patients on TB treatment is very tolerant towards antibiotics and is causative of multidrug resistance. The so far confirmed results include absence of net growth of the bacteria inside the macrophages over two weeks, alteration of characteristics of the bacteria identical to observations in other studies, suppression of immune activation of the infected cells, tolerance towards antibiotics and most importantly, regrowth of the bacteria during immunosuppression, which is a well-known clinical problem and has so far only been modeled in animals. The second model is based on the establishment of co-cultures of human cell lines and primary immune cells on a filter with a collagen matrix. Human blood monocyte-derived macrophages infected with M. tuberculosis are introduced into the system. The tissues are exposed to air on the apical side, causing the epithelial cells to secrete mucus, further mimicking the microenvironment in the lung. This unique TB model will be characterized within the project. During the present project, the models will be compared with and validated against human TB-infected tissues on an ultrastructural level using advanced electron microscopy (Maastricht University). The proposed project is a translational academic project linking experimental research to clinical testing and including Dutch and Swedish research groups. The models will be used for screening combinations of approved drugs together with TB-antibiotics and for achieving proof-of-principle of adjuvant treatments for a more effective TB treatment. Together, the two models may prove to serve as good substitutes for animal models for TB. If the project is successful, the models can be further developed to mimic different kinds of bacterial and viral infections.