Principal Investigator/Program Director Williams, Robert W.
|Introduction to the Application|
scientific motivation for this program project is to investigate the
complex network of molecules and mechanisms that modulate the
structure of different parts of the brainfrom major subdivisions
such as the hippocampus, thalamus, and cerebellum, down to the level
of discrete subpopulations of neurons and glial cells in single
nuclei. We plan to begin this work by producing the analytic tools,
tissue resources, and genotypes that are essential for the systematic
exploration of the complex genetics of mammalian brain architecture.
The bioinformatic tools and resources that we have been creating have very wide applicability, but our program project is specifically directed at exploiting a powerful new reductionist approach to explore the genetic basis of the very significant structural differences between the brains of different strains of mice. This approach, called either complex trait analysis or QTL analysis, developed rapidly in the late 1990s as a result of the hybridization of quantitative genetics and molecular genetics (Lander and Schork 1994). The suite of techniques associated with this approach has greatly extended the variety of CNS phenotypes that can be subjected to systematic molecular analysis. In neuroscience, complex trait analysis has been embraced by behavioral geneticists and neuropharmacologists (Plomin et al. 1991; Johnson et al. 1992; Crabbe et al. 1994; Takahashi et al. 1994; Kanes et al. 1996; Roubertoux et al. 1998). The time is ripe to apply these techniques to the genetic sources of structural differences in the brain.
Tools and Resources
Taking advantage of rapid advances in both computer technology and molecular genetics, we will assemble an innovative multidisciplinary system that will greatly facilitate experimental analysis of the CNS. We will use this system to investigate the genetic basis of normal variation in mouse CNS. Other neuroscientists and geneticists will also benefit from this project, which will make a huge collection of histological slides available for data collection and stereological analysis over the Internet
Project 1: The Mouse Brain Library.
This integrated, multi-stain, multi-plane digital image resource for neuroscientists is housed at
Project 2: Internet Microscopy Systems for high-resolution imaging.
We have developed an Internet-controlled microscope called the iScope to be used in conjunction with the MBL. This system acquires color images at a resolution of 0.1 µm/pixel. Our goal in this Project is to provide fast access to the entire slide collection using streaming video technology, enabling neuroscientists to acquire high-magnification images of any CNS region for any of the 2000 genetically defined mice in the MBL. Collaborative stereology across the Internet will become feasible.
Project 3: NeuroCartographer for segmentation of the MBL.
Projects 1 and 2 will generate massive collections of images and through-focus series of the brains of several thousand mice. The main objective of Project 3 is to develop extremely efficient ways to extract numbers from a vast collection of digital micrographs and video clips. This ambitious project has begun with the construction of 3D atlases of the mouse brain that will be segmented into hundreds of nuclei and fiber tracts. The project will produce as its output hundreds of quantitative morphometric parameters that will then be analyzed as part of Project 4. Another major function of the NeuroCartographer project is to provide sophisticated navigational tools for use with the MBL and Internet microscopes.
Project 4: The Mouse Neurogenetics Tool Box.
We will develop unique World Wide Web services that make it possible for neuroscientist to rapidly identify and map genes and quantitative trait loci, particularly those related to brain structure and, ultimately, to behavior. The Neurogenetics Tool Box will be tightly integrated with the MBL and NeuroCartographer Projects. Synergy between these components of the program project will allow neuroscientists and geneticists to explore the complex genetics of CNS architecture.
Our project will accomplish three important goals that will encourage further work on the genetics of mouse and human CNS. We expect these projects to illuminate how even seemingly small quantitative variation can catalyze research on the complex network of molecules that control brain size and cell composition.
Universal Access to Neuroanatomical Phenotypes.
By imaging brain sections of several thousand mice at low, intermediate, and high magnification and then putting these images on the web, we will provide universal access to CNS neuroanatomical phenotypes that can be used for systematic quantitative analysis. As part of a pilot project, we have already put images of many brains into the MBL. Even neglecting for a moment our genetic focus, this is an amazing resource with which to explore sex differences, effects of age on the brain (age range is from 25 to 650 days), and differences between strains. We intend to triple the size of this collection over the next five years. Besides adding more genetically defined mice, we will add more depth to the collection. Our image database will span the entire optical range: from a resolution of 25 µm/pixel for entire slides, to &mµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµcro;.5 µm/pixel for individual sections, t&mµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµcro; 1 µm for specific regions of interest such as the hypothalamus and amygdala, down to &mµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµcro;.1 µm/pixel for high-magnification through-focus series (50 x&mµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµµcro;50 µm fields) used for unbiased stereological analysis of cell number and cell size. We will also add fiber-stained sections and anti-ChAT and anti-parvalbumin-labeled sections to begin the inevitable process of incorporating new histological and cytochemical dimensions in the MBL collection.
2. Digital Dissection.
In the NeuroCartographer project, we will develop efficient and accurate methods to digitally dissect the massive MBL database of images and through-focus series. We will initially generate well-aligned images of sections from each brain in the MBL. The aligned set of images will then be semi-automatically segmented into several hundred anatomically defined compartments. Algorithms for segmentation have been developed at a rapid pace over the past decade, and the enormous advances in computer hardware have been a boon to this research. Until we initiated our large-scale bioinformatics project, however, there was no compelling reason to perfect these methods so that hundreds or thousands of brains could be digitally dissected. Now we have the motivation to push segmentation into high gear. Each region of interest (ROI) that we can define using segmentation programs will be associated with a full set of quantitative measures (volume, surface areas, shaped indices, position, and orientation).
3. Comprehensive Genotype Data.
We will supply comprehensive genotype data for an unusually large mapping cross consisting of more than 1400 young and old mice. We will also generate and update databases on genotypes and phenotypes of a large number of recombinant inbred (RI) strains of mice. Our choice of strains and crosses has been thought out carefully with the aim of maximizing the precision and statistical power of QTL analysis. The large BXD set (35 extant strains) will be complemented by a tenth-generation advanced intercross between C57BL/6J and DBA/2J (the G10). Using powerful statistical programs integrated in the Neurogenetics Tool Box (NTB), neuroscientists will be able to exploit these morphometric resources without processing either tissue or DNA. The extensive image resources in the MBL and the matched morphometric databases associated with the NeuroCartographer project will make it comparatively easy to map hundreds of QTLs that specifically modulate the size and cellular architecture of the mouse CNS. Each of these QTLs will be mapped with unprecedented precision because our major cross is unusually large (1400 animals) and because the multigeneration design provides approximately five times the resolution of a corresponding two-generation cross.
|Informatics Center for Mouse Neurogenetics (HBP)|
|Investigating Genetic Variation|
|Generic Resources versus Specific Goals|
|The External Advisory Board|
|Institutional Support for this Program Project|