Choose timezone
Your profile timezone:
Powered by Indico
v3.3.4
The purpose of this workshop is to explore areas of common interest in the area of cellular level electron microscopy, between scientists of the Flatiron Institute (CCB and CCM) and the New York Structural Biology Center. Some areas for discussion include methods for automated segmentation, applications of machine learning to the entire pipeline, mining the data (golgi apparatus vs microtubules & spindles vs mitochondrial network vs …), integration of EM with light microscopy and other modes, integration with biophysical modeling of cells and their dynamics, 3D visualization of the complex datasets generated by EM, methods for integrating visualization of complexity across scales and across imaging modalities.
Following the astonishing success of the single-particle cryoEM “resolution revolution”, the next frontier is to explore and understand these molecular machines within the complex environment inside cells. This presents many challenges requiring linking data between image modalities across many scales, a high level of automation, segmenting structures of interest out of noisy and cluttered environments, visualizing the results, and doing all this at a pace that makes it attractive and useful general method. In turn, these exciting challenges present a wealth of opportunities for the development of new approaches and technologies. This presentation will provide a brief introduction to cellular level electron microscopy and will outline some of the challenges and opportunities for advances.
Direct contacts between organelles, such as ER and mitochondria, are emerging as critical platforms for many biological responses in eukaryotic cells. Organelle contacts are emerging as critical signaling platforms in metazoan cells. In non-neuronal cells, many of the important physiological functions played by mitochondria such as Ca2+ uptake and lipid biogenesis require a specialized structural and functional interface with the smooth endoplasmic reticulum (ER). Recent data support a model whereby mitochondrial Ca2+ uptake can only occur upon Ryanodine and/or IP3 receptors-mediated Ca2+ release from the ER at sites of ER-mitochondria contacts, where Ca2+ transiently reaches high enough concentrations to open the mitochondrial calcium uniporter (MCU). The major limiting factors in studying the role of ER-mitochondrial contacts in shaping some of the functional properties of neurons are (1) to map their distribution throughout the dendrites of identified neuronal subtypes over large scales (hundreds to thousands of cubic microns) which requires serial electron microscopy and precise segmentation and reconstructions and (2) Developing a molecular toolkit allowing to manipulate their formation and/or dynamics at single cell resolution. I will report the progress we have made progress on both fronts including some published (Hirabayashi et al. Science 2017) as well as unpublished evidence.
Maps of synaptic connections between neurons, or connectomes, provide crucial information for reverse
engineering the brain. We developed a semi-automated image processing pipeline to reconstruct the connectome from serial electron micrographs. We report first results from the connectome of Megaphragma amalphitanum, a microscopic wasp whose linear size is an order of magnitude smaller than that of Drosophila. One peculiar feature of the Megaphragma is that most neurons lack nuclei. We find that the few neurons with nuclei occupy stereotypical positions in the connectome suggesting that their function requires genetic information.
Joint work with the groups of Alexei Polilov (Moscow),
Harald Hess (Janelia), and Viren Jain (Google).
Modeling and so understanding the dynamics of structures within cells requires input from a variety of observations and preparations -- light microscopy, bleaching, physical and genetic perturbations, and ultrastructure reconstructions using tomography. I'll discuss how the fine-grained tomographic details of microtubule conformation and coupling have driven and assisted our large-scale modeling of nuclear transport and spindle dynamics.
An important step in the cryo-EM pipeline involves taking (centered) particle images and reconstructing a 3d molecular volume.
I'll describe some of the techniques we've developed to improve this step, including:
1. tools for appropriately centering particle images,
2. a simple method for detecting particle heterogeneity and segregating images, and
3. methods for accelerating the image-alignment and molecular reconstruction process.
This is joint work with the flatiron team: Marina Spivak, Joakim Anden, Alex Barnett, and Leslie Greengard.
We have developed a novel approach of visualizing entire spindles in 3D by electron tomography and automatic microtubule segmentation. Using this approach, we can resolve single microtubules, which provides a unique perspective and offers a plethora of completely new information about the microstructure of spindles. By combining electron tomography, light microcopy, and biophysical modeling we aim to develop a detailed and unprecedented understanding of spindle assembly and chromosome segregation during meiosis and mitosis.