Toward a workbench for rodent brain image data systems architecture and design

We present a novel system for storing and manipulating microscopic images from sections through the brain and higher-level data extracted from such images. The system is designed and built on a three-tier paradigm and provides the research community with a web-based interface for facile use in neuroscience research. The Oracle relational database management system provides the ability to store a variety of objects relevant to the images and provides the framework for complex querying of data stored in the system. Further, the suite of applications intimately tied into the infrastructure in the application layer provide the user the ability not only to query and visualize the data, but also to perform analysis operations based on the tools embedded into the system. The presentation layer uses extant protocols of the modern web browser and this provides ease of use of the system. The present release, named Functional Anatomy of the Cerebro-Cerebellar System (FACCS), available through The Rodent Brain Workbench (http:// rbwb.org/), is targeted at the functional anatomy of the cerebro-cerebellar system in rats, and holds axonal tracing data from these projections. The system is extensible to other circuits and projections and to other categories of image data and provides a unique environment for analysis of rodent brain maps in the context of anatomical data. The FACCS application assumes standard animal brain atlas models and can be extended to future models. The system is available both for interactive use from a remote web-browser client as well as for download to a local server machine.

Topographical organization of pathways from somatosensory cortex through the pontine nuclei to tactile regions of the rat cerebellar hemispheres

The granule cell layer of the cerebellar hemispheres contains a patchy and noncontinuous map of the body surface, consisting of a complex mosaic of multiple perioral tactile representations. Previous physiological studies have shown that cerebrocerebellar mossy fibre projections, conveyed through the pontine nuclei, are mapped in registration with peripheral tactile projections to the cerebellum. In contrast to the fractured cerebellar map, the primary somatosensory cortex (SI) is somatotopically organized. To understand better the map transformation occurring in cerebrocerebellar pathways, we injected axonal tracers in electrophysiologically defined locations in Sprague-Dawley rat folium crus IIa, and mapped the distribution of retrogradely labelled neurons within the pontine nuclei using three-dimensional (3-D) reconstructions. Tracer injections within the large central upper lip patch in crus IIa-labelled neurons located centrally in the pontine nuclei, primarily contralateral to the injected side. Larger injections (covering multiple crus IIa perioral representations) resulted in labelling extending only slightly beyond this region, with a higher density and more ipsilaterally labelled neurons. Combined axonal tracer injections in upper lip representations in SI and crus IIa, revealed a close spatial correspondence between the cerebropontine terminal fields and the crus IIa projecting neurons. Finally, comparisons with previously published three-dimensional distributions of pontine neurons labelled following tracer injections in face receiving regions in the paramedian lobule (downloaded from http://www.rbwb.org) revealed similar correspondence. The present data support the coherent topographical organization of cerebro-ponto-cerebellar networks previously suggested from physiological studies. We discuss the present findings in the context of transformations from cerebral somatotopic to cerebellar fractured tactile representations.

Expression mapping of tetracycline-responsive prion protein promoter: digital atlasing for generating cell-specific disease models

We present a digital atlas system that allows mapping of molecular expression patterns at cellular resolution through large series of histological sections. Using this system, we have mapped the distribution of a distinct marker, encoded by the LacZ reporter gene driven by the tetracycline-responsive prion protein promoter in double transgenic mice. The purpose is to evaluate the suitability of this promoter mouse line for targeting genes of interest to specific brain regions, essential for construction of inducible transgenic disease models. Following processing to visualize the promoter expression, sections were counterstained to simultaneously display cytoarchitectonics. High-resolution mosaic images covering entire coronal sections were collected through the mouse brain at intervals of 200 microm. A web-based application provides access to a customized virtual microscopy tool for viewing and navigation within and across the section images. For each section image, the nearest section in a standard atlas is defined, and annotations of key structures and regions inserted. Putative categorization of labeled cells was performed with use of distribution patterns, followed by cell size and shape, as parameters that were compared to legacy data. Among the ensuing results were expression of this promoter in putative glial cells in the cerebellum (and not in Purkinje cells), in putative glial cells in the substantia nigra, in pallidal glial cells or interneurons, and in distinct cell layers and regions of the hippocampus. The study serves as a precursor for a database resource allowing evaluation of the suitability of different promoter mouse lines for generating disease models.

Micro3D: computer program for three-dimensional reconstruction visualization, and analysis of neuronal populations and barin regions

This article presents a computer program, Micro3D, designed for 3-D reconstruction, visualization, and analysis of coordinate-data (points and lines) recorded from serial sections. The software has primarily been used for studying shapes and dimension of brain regions (contour line data) and distributions of cellular elements such as neuronal cell bodies or axonal terminal fields labeled with tract-tracing techniques (point data). The tissue elements recorded could equally well be labeled with use of other techniques, the only requirement being that the data collected are saved as x,y,z coordinates. Data are typically imported from image-combining computerized microscopy systems or image analysis systems, such as Neurolucida (MicroBrightField, Colchester, VT) or analySIS (Soft Imaging System, Gmbh, Münster, Germany). System requirements are a PC running LINUX. Reconstructions in Micro3D may be rotated and zoomed in real-time, and submitted to perspective viewing and stereo-imaging. Surfaces are re-synthesized on the basis of stacks of contour lines. Clipping is used for defining section-independent subdivisions of the reconstruction. Flattening of curved sheets of points layers (e.g., neurons in a layer) facilitates inspection of complicated distribution patterns. Micro3D computes color-coded density maps. Opportunities for translation of data from different reconstructions into common coordinate systems are also provided. This article demonstrates the use of Micro3D for visualization of complex neuronal distribution patterns in somatosensory and auditory systems. The software is available for download on conditions posted at the NeSys home pages (http://www.nesys.uio.no/) and at The Rodent Brain Workbench (http://www.rbwb.org/).

Database and tools for analysis of topographic organization and map transformations in major projection systems of the brain

Integration of dispersed and complicated information collected from the brain is needed to build new knowledge. But integration may be hampered by rigid presentation formats, diversity of data formats among laboratories, and lack of access to lower level data. We have addressed some of the fundamental issues related to this challenge at the level of anatomical data, by producing a coordinate based digital atlas and database application for a major projection system in the rat brain: the cerebro-ponto-cerebellar system. This application, Functional Anatomy of the Cerebro-Cerebellar System in rat (FACCS), is available via the Rodent Brain WorkBench (http://www.rbwb.org). The data included are x,y,z-coordinate lists describing exact distributions of tissue elements (axonal terminal fields of axons, or cell bodies) that are labeled with axonal tracing techniques. All data are translated to a common local coordinate system to facilitate across animal comparison. A search capability allows queries based on, e.g. location of tracer injection sites, tracer category, size of the injection sites, and contributing author. A graphic search tool allows the user to move a volume cursor inside a coordinate system to detect particular injection sites having connections to a specific tissue volume at chosen density levels. Tools for visualization and analysis of selected data are included, as well as an option to download individual data sets for further analysis. With this application, data and metadata from different experiments are mapped into the same information structure and made available for re-use and re-analysis in novel combinations. The application is prepared for future handling of data from other projection systems as well as other data categories.

Three-dimensional topography of corticopontine projections from rat sensorimotor cortex: comparisons with corticostriatal projections reveal diverse integrative organization

The major cortical-subcortical re-entrant pathways through the basal ganglia and cerebellum are considered to represent anatomically segregated channels for information originating in different cortical areas. A capacity for integrating unique combinations of cortical inputs has been well documented in the basal ganglia circuits but is largely undefined in the precerebellar circuits. To compare and quantify the amount of overlap that occurs in the first link of the cortico-ponto-cerebellar pathway, a dual tracing approach was used to map the spatial relationship between projections originating from the primary somatosensory cortex (SI), the secondary somatosensory cortex (SII), and the primary motor cortex (MI). The anterograde tracers biotinylated dextran amine and Fluoro-Ruby were injected into homologous whisker representations of either SI and SII, or SI and MI. The ensuing pontine labeling patterns were analyzed using a computerized three-dimensional reconstruction approach. The results demonstrate that whisker-related projections from SI and MI are largely segregated. At some locations, the two projections are adjoining and partly overlapping. Furthermore, SI contributes significantly more corticopontine projections than MI. By comparison, projections from corresponding representations in SI and SII terminate in similar parts of the pontine nuclei and display considerable amounts of spatial overlap. Finally, comparison of corticopontine and corticostriatal projections in the same experimental animals reveals that SI-SII overlap is significantly larger in the pontine nuclei than in the neostriatum. These structural differences indicate a larger capacity for integration of information within the same sensory modality in the pontocerebellar system compared to the basal ganglia.

In vivo tracing of major rat brain pathways using manganese-enhanced magnetic resonance imaging and three-dimensional digital atlasing

The magnetic resonance imaging (MRI)-detectable T1 contrast agent manganese (Mn2+) has recently been introduced as a neural tracer in rodents, birds, and monkeys. We have tested to what extent this in vivo method is useful for three-dimensional (3-D) survey of connectivity patterns in the rat somatosensory system. A commonly available 3 T human clinical MRI scanner was used to trace neural pathways following focal injection of manganese chloride (MnCl2) in the somatosensory cortex. Six to 10 h after MnCl2 injection, we found significant signal enhancement in major projection systems, including corticocortical, corticostriatal, corticothalamic, corticotectal, corticopontine, and corticospinal pathways. To facilitate the assignment of anatomic localization to the observed Mn2+ signal enhancement, we registered the MRI data with a 3-D digital reconstruction of a stereotaxic rat brain atlas. Across-animal comparison using the digital model allowed demonstration of a corticothalamic 3-D topographic organization in agreement with previously published two-dimensional topographic schemes based on classical neural tracing data. We conclude that anterograde MnCl2/MRI tracing allows rapid analysis of topographic organization across multiple brain regions. The method allows a higher data throughput for 3-D studies of large-scale brain connectivity than conventional methods based on tissue sectioning.

Clustered and laminar topographic patterns in rat cerebro-pontine pathways

Novel neuroanatomic approaches for investigating topographic maps at a systems level include combined use of sensitive neural tracing techniques and computerized methods for three-dimensional reconstruction from serial sections. Application of these methods have allowed discovery of new principles of topographic organization in the rat pontine nuclei. The pontine nuclei are intercalated in the large pathways connecting the cerebral cortex to the cerebellum. In rat, cerebropontine projections are characterized by multiple delineated terminal fields. It is generally accepted that these projections are topographically organized. The presence of widespread axonal clusters in the pontine nuclei is typically interpreted to represent a complex scheme of organization. In recent anatomic investigations of somatosensory corticopontine projections in young and adult rats, a somatotopic distribution of axonal clusters, concentrically organized in an inside-out fashion, has been reported. This review summarizes the topographic principles of organization proposed for somatosensory corticopontine projections, and discusses the possibility that widely segregated corticopontine terminal fields are located inside lamellar volumes. This organizational pattern may be explained by mechanisms operative during development, and resembles the patterns of organization previously described in cat and monkey. Possible implications of this architecture are discussed in relation to map transformations from the cerebral cortex to the cerebellum.

Comparison of semi-automatic and automatic data acquisition methods for studying three-dimensional distributions of large neuronal populations and axonal plexuses

Neuroanatomy is in need of high throughput methods for reliably recording the distribution of tissue elements across large brain regions. We compared two methods for recording the spatial distribution of identified neuronal elements such as tracer labelled cell bodies or axonal plexuses. The methods compared were computerized image-combining microscopy (semi-automatic method), which is a user controlled method providing feedback during digitization, and digital camera technology with image analysis software (automatic method). Both methods were applied to biotinylated dextran amine labelled axonal plexuses and FluoroRuby labelled neuronal cell bodies, in the pontine nuclei of the rat. Coordinates were assigned to the labelled elements using both methods. The ensuing distribution patterns were compared, section by section, and in three-dimensional reconstruction. The experienced investigator, using the semi-automatic method, could detect individual axons, fragments of axons, weakly labelled elements, and overlapping cell bodies, better than the automatic system. Nevertheless, both methods detected the overall distribution of the labelled axons and cells investigated. Automatic methods provide opportunities for efficient large-scale data acquisition of labelled neuronal elements.

Three-dimensional computerised atlas of the rat brain stem precerebellar system: approaches for mapping, visualization, and comparison of spatial distribution data

Comparisons of microscopical neuroanatomic data from different experiments and investigators are typically hampered by the use of different section planes and dissimilar techniques for data documentation. We have developed a framework for visualization and comparison of section-based, spatial distribution data, in brain stem nuclei. This framework provides opportunities for harmonized data presentation in neuroinformatics databases. Three-dimensional computerized reconstructions of the rat brain stem and precerebellar nuclei served as a basis for establishing internal coordinate systems for the pontine nuclei and the precerebellar divisions of the sensory trigeminal nuclei. Coordinate based diagrams were used for presentation of experimental data (spatial distribution of labelled neurons and axonal plexuses) from standard angles of view. Each nuclear coordinate system was based on a cuboid bounding box with a defined orientation. The bounding box was size-adjusted to touch cyto- and myeloarchitectonically defined boundaries of the individual nuclei, or easily identifiable nearby landmarks. We exemplify the use of these internal coordinate systems with dual retrograde neural tracing data from pontocerebellar and trigeminocerebellar systems. The new experimental data were combined, in the same coordinate based diagrams, with previously published data made available via a neuroinformatics data repository (www.nesys.uio.no/Database, see also www.cerebellum.org). Three-dimensional atlasing, internal nuclear coordinate systems, and consistent formats for presentation of neuroanatomic data in web-based data repositories, offer new opportunities for efficient analysis and re-analysis of neuroanatomic data.

Three-dimensional topography of corticopontine projections from rat barrel cortex: correlations with corticostriatal organization

Subcortical re-entrant projection systems connecting cerebral cortical areas with the basal ganglia and cerebellum are topographically specific and therefore considered to be parallel circuits or "closed loops." The precision of projections within these circuits, however, has not been characterized sufficiently to indicate whether cortical signals are integrated within or among presumed compartments. To address this issue, we studied the first link of the rat cortico-ponto-cerebellar pathway with anterograde axonal tracing from physiologically defined, individual whisker "barrels" of the primary somatosensory cortex (SI). The labeled axons in the pontine nuclei formed several, sharply delineated clusters. Dual tracer injections into different SI whisker barrels gave rise to partly overlapping, paired clusters, indicating somatotopic specificity. Three-dimensional reconstructions revealed that the clusters were components of concentrically organized lamellar subspaces. Whisker barrels in the same row projected to different pontine lamellae (side by side), the somatotopic representation of which followed an inside-out sequence. By contrast, whisker barrels from separate rows projected to clusters located within the same lamellar subspace (end to end). In the neostriatum, this lamellar topography was the opposite, with barrels in the same row contacting different parts of the same lamellar subspace (end to end). The degree of overlap among pontine clusters varied as a function of the proximity of the cortical injections. Furthermore, corticopontine overlap was higher among projections from barrels in the same row than among projections from different whisker barrel rows. This anisotropy was the same in the corticostriatal projection. These findings have important implications for understanding convergence and local integration in somatosensory-related subcortical circuits.

Rat somatosensory cerebropontocerebellar pathways: spatial relationships of the somatotopic map of the primary somatosensory cortex are preserved in a three-dimensional clustered pontine map

In the primary somatosensory cortex (SI), the body surface is mapped in a relatively continuous fashion, with adjacent body regions represented in adjacent cortical domains. In contrast, somatosensory maps found in regions of the cerebellar hemispheres, which are influenced by the SI through a monosynaptic link in the pontine nuclei, are discontinuous ("fractured") in organization. To elucidate this map transformation, the authors studied the organization of the first link in the SI-cerebellar pathway, the SI-pontine projection. After injecting anterograde axonal tracers into electrophysiologically defined parts of the SI, three-dimensional reconstruction and computer-graphic visualization techniques were used to analyze the spatial distribution of labeled fibers. Several target regions in the pontine nuclei were identified for each major body representation. The labeled axons formed sharply delineated clusters that were distributed in an inside-out, shell-like fashion. Upper lip and other perioral representations were located in a central core, whereas extremity and trunk representations were found more externally. The multiple clusters suggest that the pontine nuclei contain several representations of the SI map. Within each representation, the spatial relationships of the SI map are largely preserved. This corticopontine projection pattern is compatible with recently proposed principles for the establishment of subcortical topographic patterns during development. The largely preserved spatial relationships in the pontine somatotopic map also suggest that the transformation from an organized topography in SI to a fractured map in the cerebellum takes place primarily in the mossy fiber pontocerebellar projection.

Anatomic evidence of a three-dimensional mosaic pattern of tonotopic organization in the ventral complex of the lateral lemniscus in cat

The ventral complex of the lateral lemniscus (VCLL, i.e., the ventral and intermediate nuclei) is composed of cells embedded in the fibers of the lateral lemniscus. These cells are involved in the processing of monaural information and receive input from the collaterals of the fibers ascending to the inferior colliculus. Whereas tonotopic organization is a feature of all other nuclei of the auditory system, this functional principle is debated in the VCLL. We have made focal injections of the tracer biotinylated dextran amine into different frequency band representations of the inferior colliculus in cat. Retrogradely labeled cells and terminal fibers (collaterals of efferent local axons and other ascending lemniscal fibers) were found in the ipsilateral VCLL. The spatial distribution of the labeling was analyzed using three-dimensional (3-D) reconstruction and computer graphical visualization techniques. A complex topographic organization was found. In all cases, labeled fibers and cells were distributed in multiple clusters throughout the dorsoventral extent of the VCLL. The shape, size, and location of the labeled clusters suggest an interdigitation of clusters assigned to different frequency-band representations. But an overall mediolateral distribution gradient was observed, with high frequencies represented medially and lower frequencies progressively more laterally. We conclude that the clusters may represent discontinuous frequency-band compartments as a counterpart to the continuous laminar compartments in the remaining auditory nuclei. The 3-D orderly mosaic pattern indicates that the VCLL preserves the spectral decomposition originated in the cochlea in a way that facilitates across-frequency integration.

Topographical organization in the early postnatal corticopontine projection: a carbocyanine dye and 3-D computer reconstruction study in the rat

We have explored basic rules guiding the early development of topographically organized projections, employing the rat corticopontine projection as a model system. Using anterograde in vivo tracing with 1,1',dioctadecyl-3,3,3',3'-tetramethylindocarbocyanine perchlorate (DiI), we studied the distribution of labelled fibers in the pontine nuclei in relation to cortical site of origin during the first postnatal week. Labelled corticopontine fibers enter the pontine nuclei in distinct, sharply defined zones. The putative terminal fibers typically occupy lamella-like subspaces. Related to changes in cortical site of origin, we describe mediolateral, internal to external, and caudorostral distribution gradients in the pontine nuclei. Fibers originating in the anterolateral cortex occupy an internal central core, while implantations at increasing distance from the anterolateral cortex produce 1) more externally located lamellae, and 2) a caudal to rostral shift in fiber location. Previous investigations have shown that pontocerebellar neurons migrate into the ventral pons in a temporal sequence (Altman and Bayer [1987] J. Comp. Neurol. 257:529). The earliest arriving neurons occupy the central core and later arriving neurons settle in more externally and rostrally located subspaces. We hypothesize that the earliest arriving corticopontine fibers grow into the then only available zone of pontocerebellar neurons (central core), attracted by a diffusible chemotropic cue. Later arriving fibers grow into correspondingly later and more externally and rostrally located contingents of pontocerebellar neurons. Thus, we propose that the topographical organization in the early postnatal corticopontine projection is determined by simple temporal and spatial gradients operative within source (cerebral cortex) and target region (pontine nuclei).

Semi-automatic data acquisition for quantitative neuroanatomy. MicroTrace--computer programme for recording of the spatial distribution of neuronal populations

We present a computer programme, MicroTrace, designed for user-guided digitisation of objects in biological sections. The programme is optimised for recording the spatial distribution of neuronal structures, such as large populations of tracer-labelled cell bodies or axonal plexuses, regional borders, and surfaces. System requirements are a PC running Microsoft Windows, a microscope equipped with stepping motors, and a drawing tube. A computer generated drawing area, surrounded by menus and icons, is projected into the microscope field of view via the drawing tube. Different 'object' icons are assigned to individual object categories (cell types, surfaces, etc.). Digitisation is performed by pointing the cursor at objects in the section. Computer graphical symbols are superimposed on the digitised objects. All object categories are digitised, before moving the stage to other fields of view by manipulating the joystick or scroll bars. Movement of the microscope stage is accompanied by a translation of the graphical image, so that continuous feedback on the progress of the digitisation is provided. MicroTrace can readily be adapted to the specific needs of the user. We show its use in different experimental neuroanatomical techniques. Two-dimensional images and three-dimensional reconstructions of neuronal distribution and surfaces are demonstrated.