570 Biowissenschaften; Biologie
Refine
Has Fulltext
- yes (25)
Is part of the Bibliography
- yes (25)
Year of publication
Document Type
- Journal article (22)
- Doctoral Thesis (2)
- Book (1)
Keywords
- super-resolution microscopy (5)
- Expansion microscopy (2)
- RNS (2)
- biology (2)
- dSTORM (2)
- fluorescence imaging (2)
- imaging the immune system (2)
- infectious diseases (2)
- superresolution (2)
- AI (1)
Institute
- Theodor-Boveri-Institut für Biowissenschaften (22)
- Center for Computational and Theoretical Biology (2)
- Institut für Hygiene und Mikrobiologie (2)
- Medizinische Klinik und Poliklinik II (2)
- Rudolf-Virchow-Zentrum (2)
- Abteilung für Funktionswerkstoffe der Medizin und der Zahnheilkunde (1)
- Graduate School of Life Sciences (1)
- Institut für Organische Chemie (1)
- Institut für Theoretische Physik und Astrophysik (1)
- Institut für Virologie und Immunbiologie (1)
Background
Localization-based super-resolution microscopy resolves macromolecular structures down to a few nanometers by computationally reconstructing fluorescent emitter coordinates from diffraction-limited spots. The most commonly used algorithms are based on fitting parametric models of the point spread function (PSF) to a measured photon distribution. These algorithms make assumptions about the symmetry of the PSF and thus, do not work well with irregular, non-linear PSFs that occur for example in confocal lifetime imaging, where a laser is scanned across the sample. An alternative method for reconstructing sparse emitter sets from noisy, diffraction-limited images is compressed sensing, but due to its high computational cost it has not yet been widely adopted. Deep neural network fitters have recently emerged as a new competitive method for localization microscopy. They can learn to fit arbitrary PSFs, but require extensive simulated training data and do not generalize well. A method to efficiently fit the irregular PSFs from confocal lifetime localization microscopy combining the advantages of deep learning and compressed sensing would greatly improve the acquisition speed and throughput of this method.
Results
Here we introduce ReCSAI, a compressed sensing neural network to reconstruct localizations for confocal dSTORM, together with a simulation tool to generate training data. We implemented and compared different artificial network architectures, aiming to combine the advantages of compressed sensing and deep learning. We found that a U-Net with a recursive structure inspired by iterative compressed sensing showed the best results on realistic simulated datasets with noise, as well as on real experimentally measured confocal lifetime scanning data. Adding a trainable wavelet denoising layer as prior step further improved the reconstruction quality.
Conclusions
Our deep learning approach can reach a similar reconstruction accuracy for confocal dSTORM as frame binning with traditional fitting without requiring the acquisition of multiple frames. In addition, our work offers generic insights on the reconstruction of sparse measurements from noisy experimental data by combining compressed sensing and deep learning. We provide the trained networks, the code for network training and inference as well as the simulation tool as python code and Jupyter notebooks for easy reproducibility.
G-protein-coupled receptors (GPCRs) are hypothesized to possess molecular mobility over a wide temporal range. Until now the temporal range has not been fully accessible due to the crucially limited temporal range of available methods. This in turn, may lead relevant dynamic constants to remain masked. Here, we expand this dynamic range by combining fluorescent techniques using a spot confocal setup. We decipher mobility constants of β\(_{2}\)-adrenergic receptor over a wide time range (nanosecond to second). Particularly, a translational mobility (10 µm\(^{2}\)/s), one order of magnitude faster than membrane associated lateral mobility that explains membrane protein turnover and suggests a wider picture of the GPCR availability on the plasma membrane. And a so far elusive rotational mobility (1-200 µs) which depicts a previously overlooked dynamic component that, despite all complexity, behaves largely as predicted by the Saffman-Delbrück model.
Expansion microscopy (ExM) enables super-resolution fluorescence imaging on standard microscopes by physical expansion of the sample. However, the investigation of interactions between different organisms such as mammalian and fungal cells by ExM remains challenging because different cell types require different expansion protocols to ensure identical, ideally isotropic expansion of both partners. Here, we introduce an ExM method that enables super-resolved visualization of the interaction between NK cells and Aspergillus fumigatus hyphae. 4-fold expansion in combination with confocal fluorescence imaging allows us to resolve details of cytoskeleton rearrangement as well as NK cells' lytic granules triggered by contact with an RFP-expressing A. fumigatus strain. In particular, subdiffraction-resolution images show polarized degranulation upon contact formation and the presence of LAMP1 surrounding perforin at the NK cell-surface post degranulation. Our data demonstrate that optimized ExM protocols enable the investigation of immunological synapse formation between two different species with so far unmatched spatial resolution.
Single-molecule super-resolution microscopy (SMLM) techniques like dSTORM can reveal biological structures down to the nanometer scale. The achievable resolution is not only defined by the localization precision of individual fluorescent molecules, but also by their density, which becomes a limiting factor e.g., in expansion microscopy. Artificial deep neural networks can learn to reconstruct dense super-resolved structures such as microtubules from a sparse, noisy set of data points. This approach requires a robust method to assess the quality of a predicted density image and to quantitatively compare it to a ground truth image. Such a quality measure needs to be differentiable to be applied as loss function in deep learning. We developed a new trainable quality measure based on Fourier Ring Correlation (FRC) and used it to train deep neural networks to map a small number of sampling points to an underlying density. Smooth ground truth images of microtubules were generated from localization coordinates using an anisotropic Gaussian kernel density estimator. We show that the FRC criterion ideally complements the existing state-of-the-art multiscale structural similarity index, since both are interpretable and there is no trade-off between them during optimization. The TensorFlow implementation of our FRC metric can easily be integrated into existing deep learning workflows.
Actin cytoskeleton deregulation confers midostaurin resistance in FLT3-mutant acute myeloid leukemia
(2021)
The presence of FMS-like tyrosine kinase 3-internal tandem duplication (FLT3-ITD) is one of the most frequent mutations in acute myeloid leukemia (AML) and is associated with an unfavorable prognosis. FLT3 inhibitors, such as midostaurin, are used clinically but fail to entirely eradicate FLT3-ITD+AML. This study introduces a new perspective and highlights the impact of RAC1-dependent actin cytoskeleton remodeling on resistance to midostaurin in AML. RAC1 hyperactivation leads resistance via hyperphosphorylation of the positive regulator of actin polymerization N-WASP and antiapoptotic BCL-2. RAC1/N-WASP, through ARP2/3 complex activation, increases the number of actin filaments, cell stiffness and adhesion forces to mesenchymal stromal cells (MSCs) being identified as a biomarker of resistance. Midostaurin resistance can be overcome by a combination of midostaruin, the BCL-2 inhibitor venetoclax and the RAC1 inhibitor Eht1864 in midostaurin-resistant AML cell lines and primary samples, providing the first evidence of a potential new treatment approach to eradicate FLT3-ITD+AML. Garitano-Trojaola et al. used a combination of human acute myeloid leukemia (AML) cell lines and primary samples to show that RAC1-dependent actin cytoskeleton remodeling through BCL2 family plays a key role in resistance to the FLT3 inhibitor, Midostaurin in AML. They showed that by targeting RAC1 and BCL2, Midostaurin resistance was diminished, which potentially paves the way for an innovate treatment approach for FLT3 mutant AML.
Cyanine dyes are exceptionally useful probes for a range of fluorescence-based applications, but their photon output can be limited by trans-to-cis photoisomerization. We recently demonstrated that appending a ring system to the pentamethine cyanine ring system improves the quantum yield and extends the fluorescence lifetime. Here, we report an optimized synthesis of persulfonated variants that enable efficient labeling of nucleic acids and proteins. We demonstrate that a bifunctional sulfonated tertiary amide significantly improves the optical properties of the resulting bioconjugates. These new conformationally restricted cyanines are compared to the parent cyanine derivatives in a range of contexts. These include their use in the plasmonic hotspot of a DNA-nanoantenna, in single-molecule Förster-resonance energy transfer (FRET) applications, far-red fluorescence-lifetime imaging microscopy (FLIM), and single-molecule localization microscopy (SMLM). These efforts define contexts in which eliminating cyanine isomerization provides meaningful benefits to imaging performance.
Translation efficiency can be affected by mRNA stability and secondary structures, including G-quadruplex structures (G4s). The highly conserved DEAH-box helicase DHX36/RHAU resolves G4s on DNA and RNA in vitro, however a systems-wide analysis of DHX36 targets and function is lacking. We map globally DHX36 binding to RNA in human cell lines and find it preferentially interacting with G-rich and G4-forming sequences on more than 4500 mRNAs. While DHX36 knockout (KO) results in a significant increase in target mRNA abundance, ribosome occupancy and protein output from these targets decrease, suggesting that they were rendered translationally incompetent. Considering that DHX36 targets, harboring G4s, preferentially localize in stress granules, and that DHX36 KO results in increased SG formation and protein kinase R (PKR/EIF2AK2) phosphorylation, we speculate that DHX36 is involved in resolution of rG4 induced cellular stress.
Quantifying protein densities on cell membranes using super-resolution optical fluctuation imaging
(2017)
Quantitative approaches for characterizing molecular organization of cell membrane molecules under physiological and pathological conditions profit from recently developed super-resolution imaging techniques. Current tools employ statistical algorithms to determine clusters of molecules based on single-molecule localization microscopy (SMLM) data. These approaches are limited by the ability of SMLM techniques to identify and localize molecules in densely populated areas and experimental conditions of sample preparation and image acquisition. We have developed a robust, model-free, quantitative clustering analysis to determine the distribution of membrane molecules that excels in densely labeled areas and is tolerant to various experimental conditions, i.e. multiple-blinking or high blinking rates. The method is based on a TIRF microscope followed by a super-resolution optical fluctuation imaging (SOFI) analysis. The effectiveness and robustness of the method is validated using simulated and experimental data investigating nanoscale distribution of CD4 glycoprotein mutants in the plasma membrane of T cells.
Biological systems are dynamic and three-dimensional but many techniques allow only static and two-dimensional observation of cells. We used three-dimensional (3D) lattice light-sheet single-molecule localization microscopy (dSTORM) to investigate the complex interactions and distribution of single molecules in the plasma membrane of whole cells. Different receptor densities of the adhesion receptor CD56 at different parts of the cell highlight the importance and need of three-dimensional observation and analysis techniques.
Expansion microscopy (ExM) enables super-resolution imaging of proteins and nucleic acids on conventional microscopes. However, imaging of details of the organization of lipid bilayers by light microscopy remains challenging. We introduce an unnatural short-chain azide- and amino-modified sphingolipid ceramide, which upon incorporation into membranes can be labeled by click chemistry and linked into hydrogels, followed by 4x to 10x expansion. Confocal and structured illumination microscopy (SIM) enable imaging of sphingolipids and their interactions with proteins in the plasma membrane and membrane of intracellular organelles with a spatial resolution of 10-20nm. As our functionalized sphingolipids accumulate efficiently in pathogens, we use sphingolipid ExM to investigate bacterial infections of human HeLa229 cells by Neisseria gonorrhoeae, Chlamydia trachomatis and Simkania negevensis with a resolution so far only provided by electron microscopy. In particular, sphingolipid ExM allows us to visualize the inner and outer membrane of intracellular bacteria and determine their distance to 27.6 +/- 7.7nm. Imaging of lipid bilayers using light microscopy is challenging. Here the authors label cells using a short chain click-compatible ceramide to visualize mammalian and bacterial membranes with expansion microscopy.