Yoonsuck Choe - US grants

Detailed network connectivity maps do not exist at present for any mammalian brain circuit. This project will provide the first empirical map of mouse brain network connectivity, focusing on the individual cerebral cortical microcolumn (as determined by Serial Block Face Scanning Electron Microscope; SBF-SEM), and on the entire mouse cortex by integrating light microscopy data (from our Knife-Edge Scanning Microscope; KESM) with that of the SBF-SEM. The project will chart brain networks in the mouse at multiple scales of spatial resolution, and develop interfaces between these levels of description. The combined use of large-scale 3D microscopes, SBF-SEM and KESM, together with automated image analysis and reconstruction methods, will open the internal connectivity of brains of all species to measurement and modeling of brain architecture at a neuronal and subneuronal level of detail. The project will recruit mouse brain data from three sources at three scales: nanoscale (from SBF-SEM, Stanford University); microscale (from KESM, Texas A&M University); and macroscale (from the Mouse Atlas Project, UCLA). A key objective of this project is to develop seamless interfaces across the three levels. Staining, imaging, image processing, and reconstruction methods will be developed to interoperate across these multiple levels. At nanoscale: Stanford is developing high-contrast heavy-element staining methods to be used with a new, automated SBF-SEM for tracing small and tightly packed axons and dendrites over the entire volume of a functional microcircuit. At microscale: We will use the KESM to scan the mouse brain at 300 nm resolution and create an aligned volume data set of select cortical areas. For these two levels, a common heavy element stain will be used. The KESM, in turn, can image conventionally stained tissue to compare to the pre-existing UCLA 3D mouse brain atlas (MAP). The data will be cast into the Mouse Brain Web (MBW), a web-based representation of the mouse brain network which will make possible multi-scale integration of circuitry information across these levels. The availability of such multi-scale information for the mouse will be strongly beneficial to our understanding of human brain function and development. Such information will also contribute to discovering better treatment of disorders, such as epilepsy, and potential regenerative abilities. Data from this project will be made public, along with the software for its integrative storage, retrieval, and analysis.

Detailed network connectivity maps do not exist at present for any mammalian brain circuit. This project will provide the first empirical map of mouse brain network connectivity, focusing on the individual cerebral cortical microcolumn (as determined by Serial Block Face Scanning Electron Microscope; SBF-SEM), and on the entire mouse cortex by integrating light microscopy data (from our Knife-Edge Scanning Microscope; KESM) with that of the SBF-SEM. The project will chart brain networks in the mouse at multiple scales of spatial resolution, and develop interfaces between these levels of description. The combined use of large-scale 3D microscopes, SBF-SEM and KESM, together with automated image analysis and reconstruction methods, will open the internal connectivity of brains of all species to measurement and modeling of brain architecture at a neuronal and subneuronal level of detail. The project will recruit mouse brain data from three sources at three scales: nanoscale (from SBF-SEM, Stanford University); microscale (from KESM, Texas A&M University); and macroscale (from the Mouse Atlas Project, UCLA). A key objective of this project is to develop seamless interfaces across the three levels. Staining, imaging, image processing, and reconstruction methods will be developed to interoperate across these multiple levels. At nanoscale: Stanford is developing high-contrast heavy-element staining methods to be used with a new, automated SBF-SEM for tracing small and tightly packed axons and dendrites over the entire volume of a functional microcircuit. At microscale: We will use the KESM to scan the mouse brain at 300 nm resolution and create an aligned volume data set of select cortical areas. For these two levels, a common heavy element stain will be used. The KESM, in turn, can image conventionally stained tissue to compare to the pre-existing UCLA 3D mouse brain atlas (MAP). The data will be cast into the Mouse Brain Web (MBW), a web-based representation of the mouse brain network which will make possible multi-scale integration of circuitry information across these levels. The availability of such multi-scale information for the mouse will be strongly beneficial to our understanding of human brain function and development. Such information will also contribute to discovering better treatment of disorders, such as epilepsy, and potential regenerative abilities. Data from this project will be made public, along with the software for its integrative storage, retrieval, and analysis.

Detailed network connectivity maps do not exist at present for any mammalian brain circuit. This project will provide the first empirical map of mouse brain network connectivity, focusing on the individual cerebral cortical microcolumn (as determined by Serial Block Face Scanning Electron Microscope; SBF-SEM), and on the entire mouse cortex by integrating light microscopy data (from our Knife-Edge Scanning Microscope; KESM) with that of the SBF-SEM. The project will chart brain networks in the mouse at multiple scales of spatial resolution, and develop interfaces between these levels of description. The combined use of large-scale 3D microscopes, SBF-SEM and KESM, together with automated image analysis and reconstruction methods, will open the internal connectivity of brains of all species to measurement and modeling of brain architecture at a neuronal and subneuronal level of detail. The project will recruit mouse brain data from three sources at three scales: nanoscale (from SBF-SEM, Stanford University); microscale (from KESM, Texas A&M University); and macroscale (from the Mouse Atlas Project, UCLA). A key objective of this project is to develop seamless interfaces across the three levels. Staining, imaging, image processing, and reconstruction methods will be developed to interoperate across these multiple levels. At nanoscale: Stanford is developing high-contrast heavy-element staining methods to be used with a new, automated SBF-SEM for tracing small and tightly packed axons and dendrites over the entire volume of a functional microcircuit. At microscale: We will use the KESM to scan the mouse brain at 300 nm resolution and create an aligned volume data set of select cortical areas. For these two levels, a common heavy element stain will be used. The KESM, in turn, can image conventionally stained tissue to compare to the pre-existing UCLA 3D mouse brain atlas (MAP). The data will be cast into the Mouse Brain Web (MBW), a web-based representation of the mouse brain network which will make possible multi-scale integration of circuitry information across these levels. The availability of such multi-scale information for the mouse will be strongly beneficial to our understanding of human brain function and development. Such information will also contribute to discovering better treatment of disorders, such as epilepsy, and potential regenerative abilities. Data from this project will be made public, along with the software for its integrative storage, retrieval, and analysis.

This award supports the preparation and sharing of high-resolution, high-quality, whole mouse brain neuronal and vascular data obtained from the Knife-Edge Scanning Microscope (KESM). KESM is capable of slicing and imaging whole small animal organs (such as the mouse brain) at less than 1 um resolution within 100 hours, with a resulting data size exceeding 2 TB per specimen. The amount and complexity of the KESM data necessitates innovative approaches in data organization, storage/retrieval, and dissemination. The research team will develop an informatics framework for 3D volume data dissemination and visualization.

The KESM, developed and housed in the PI's laboratory, has been used to dissect and image two whole mouse brains so far, one stained in Golgi to reveal the neuronal morphology, and the other in India ink to show the microvascular network in fine detail. The two data sets are unique in their detail and extent compared to other currently available data sets (orders of magnitude higher resolution along one or more of the x-, y-, and z-axis, and/or higher extent in the imaged volume). Such data enable researchers to conduct a full quantitative analysis of various morphological statistics and their variability across different brain regions and nuclei, estimate morphological parameters for computational simulation, and eventually help link structure to function.

A hybrid approach will be employed that integrates a web-based light-weight data browser and a local data viewer/analyzer, under a multi-scale data scheme. The specific objectives of this project are as follows: (1) Standardized data sets for improved access and interoperability with other data sources, (2) Light-weight web-based volume browser with an open Application Programming Interface (API) for enhanced freedom of access and annotation, and (3) Unit volume viewer for the visualization and analysis of small unit volumes downloaded through the web-based interface.

The data and software tools, including documentation, will be released in the public domain, to build a user/developer community that will help continued use and evolution of the framework.

This Project will support the web-based data sharing of high-resolution, high-quality, whole mouse brain data obtained from an innovative 3D microscopy technology called Knife-Edge Scanning Microscopy (KESM). Furthermore, the data from KESM will be linked to other existing, complementary data sets such as the Allen Brain Atlas containing extensive gene expression data, effectively multiplying the utility of the resource for neuroscientists studying the brain by relating structural data (microstructure) to functional data (gene expression).

The central goal of this project is to devise a web-based data dissemination framework to link and to deliver massive amounts of mouse brain data (up to 2 terabytes per brain) to the neuroscience research community in an easily accessible form. To this end, the research team will design and develop a light-weight, web-based 3D visualization and annotation framework so that the KESM data can be viewed, annotated, and analyzed at multiple resolutions (imagine Google Maps in full 3D). The project is expected to result in a unique resource that provides an unprecedented look into the wiring of the entire mouse brain in the context of large-scale gene expression data. Such a resource can help scientists begin to unravel the intricate functions of the brain, and in the long term enable the construction of intelligent artifacts based on such an understanding.

The project will involve graduate and undergraduate researchers, and a simplified version of the resource will be created to kindle interest and fascination for brain science in K12 students and in the general public. The framework developed in this project is expected to have broader application in geology, meteorology, and even in astronomy where huge amounts of 3D data are gathered routinely yet the access limited due to the massiveness and 3D nature of the data. All data and software resulting from this project will be released in the public domain, available in full at http://kesm.org.

Award Abstract #1256086: Enhanced Knife-Edge Scanning Microscopy
for Sub-micrometer Imaging of Whole Small Animal Organs

This NSF award, made to the Brain Networks Laboratory at Texas A&M University, will help enhance and transform the Knife-Edge Scanning Microscope (KESM), a "one of a kind" microscopy instrument, into a more widely available system for biological research. KESM is capable of sectioning and imaging whole small animal organs at submicrometer resolution (that is, in very fine detail, smaller than the size of a single cell). A prototype KESM instrument was constructed, and its capability was successfully demonstrated by scanning diverse biological organs including whole mouse brains, octopus brains, and the mouse lung at submicrometer resolution. This project will enhance this prototype significantly, and transform it into a robust system that can be replicated and operated with ease by other research groups and industry partners. The enhanced instrument to be built by this project addresses at least two major emerging directions in biological research: (1) "omics" (the study of a total collection of some biological entity, for example, genomics for genes) and (2) multi-scale modeling (modeling systems across a range of scales, for example from the cell up to the whole organ). With the tremendous success of genomics in the 1990's and beyond, biological research is increasingly moving toward various forms of omics. Much of the omics research depends on anatomical information (e.g., connectomics, studying the complete wiring diagram in the brain) or genomic information (e.g., gene expression levels) at the whole-organ-level. Multi-scale modeling has also become a major methodological process in biological research. In many projects in omics and multi-scale modeling, sub-micrometer microscopy data from whole biological organs are essential yet available tools are unable to meet the current demand. KESM is expected to fill this gap. The specific enhancements of the instrument include: (1) enhanced imaging (higher resolution optics and camera, fluorescence imaging [key to genetic and molecular studies]), (2) enhanced mechanical control (more rigid, accurate, and higher resolution motion control), (3) enhanced cutting (vibrating knife, real time monitoring), and (4) enhanced robustness (quality monitoring, calibration). The overall resulting improvement can be summarized as follows: (1) 3X improvement in imaging resolution, (2) 10X improvement in robustness of operation, (3) 10X improvement in imaging speed (compared to competing methods), (4) new fluorescence imaging capability.

Broader impacts: (1) Impact on the research community: The enhanced KESM will allow researchers to obtain high resolution, whole-organ data for multi-scale, omics investigation of various types of biological organs. (2) Impact on education: Microscopic atlases of whole biological organs, such as the web-based KESM mouse brain atlas developed by the project team, will serve as an educational resource for students and educators at all levels (K-12, undergraduate, graduate, and general public). As part of this project, graduate and undergraduate students will be trained in a multidisciplinary environment (neuroscience and computer science). (3) Instrument dissemination plan: The project team will collaborate with a start-up company, to streamline system integration and manufacturing of the KESM instrument for broader dissemination. The design of the instrument and the operational instructions will be made available to the biological research community for those who wish to build their own instrument.