Developmental Mouse Brain Common Coordinate Framework

Abstract

3D standard reference brains serve as key resources to understand the spatial organization of the brain and promote interoperability across different studies. However, unlike the adult mouse brain, the lack of standard 3D reference atlases for developing mouse brains has hindered advancement of our understanding of brain development. Here, we present a multimodal 3D developmental common coordinate framework (DevCCF) spanning mouse embryonic day (E) 11.5, E13.5, E15.5, E18.5, and postnatal day (P) 4, P14, and P56 with anatomical segmentations defined by a developmental ontology. At each age, the DevCCF features undistorted morphologically averaged atlas templates created from Magnetic Resonance Imaging and co-registered high-resolution templates from light sheet fluorescence microscopy. Expert-curated 3D anatomical segmentations at each age adhere to an updated prosomeric model and can be explored via an interactive 3D web-visualizer. As a use case, we employed the DevCCF to unveil the emergence of GABAergic neurons in embryonic brains. Moreover, we integrated the Allen CCFv3 into the P56 template with stereotaxic coordinates and mapped spatial transcriptome cell-type data with the developmental ontology. In summary, the DevCCF is an openly accessible resource that can be used for large-scale data integration to gain a comprehensive understanding of brain development.

Keywords: GABA neurons; anatomical segmentation; brain development; common coordinate framework (CCF); multimodal registration; reference atlas.

Extended Data Figure 1.. ADMBA pallium compared to concentric ring topology

Coronal slices comparing the pallium as segmented by ADMBA ontology (left hemisphere) and DevCCF ontology (right hemisphere). ADMBA ontology divides the pallium as dorsal (DPall), medial (MPall), lateral (LPall), and ventral (VPall). In contrast, concentric ring topology represented by the DevCCF divides the pallium into the neocortex (NeoCx), Mesocortex (MesoCx), allocortex (AlloCx), and Pallial Amygdala (APall).

Extended Data Figure 2.. ADMBA ISH gene expression mapping onto DevCCF

(a) E15.5 ISH Pax6 slice (a) from the Allen Institute Developing Mouse Brain Atlas experiment ID 100051660. Left hemisphere overlay is dark contrast generated from inverted image intensity with greyscale color map. Red box represents location of (b). Red arrow indicates identical data registered to DevCCF E15.5 template in (e). (b) high magnification image of Pax6 expression. (c) Inverted intensity image sagittal reconstruction from coronal slices for a single subject containing ISH experiments for Pax6, Lhx2, Phox2b, Pcp2, and Olig3. Red dotted line represents (a,b) slice location. (d) Sagittal reconstruction in (c) after filling missing slices with bSpline interpolation and slice-by-slice alignment to DevCCF E15.5 template. Arrows indicate high gene expression by color. Blue: Pax6, purple: Lhx2, red: Phox2b, yellow: Pcp2, green: Olig3. (e) Localized gene expression of Pax6, Lhx2, Phox2b, Pcp2, and Olig3 (d) overlayed on DevCCF E15.5 MRI FA template. (f) Coronal view of 2D E11.5, E13.5, E15.5, E18.5, P4, P14, P28, and P56 (left to right) Pax6 ISH gene expression signal reconstruction overlayed on 3D DevCCF MRI FA templates.

Figure 1.. DevCCF overview

(a) Morphology and intensity averaged templates from LSFM (blue box) and MRI (yellow block) imaging. The LSFM template was aligned to the undistorted MRI template via multimodal registration. (b) Multimodal DevCCF with anatomical segmentations is established at four embryonic and three postnatal ages. (c) Sagittal slice of various E15.5 multimodal data registered to the E15.5 DevCCF, each highlighting unique anatomical features. From left to right: MRI DWI, MRI FA, LSFM GABAergic neurons from Gad2-Cre;Ai14 mice, LSFM autofluorescence, ISH data Lhx2 and Foxp2 gene expression to collectively guide DevCCF annotations.

Figure 2.. 3D Multimodal Developmental Mouse Brain Templates

(a) 3D DevCCF MRI templates from T2-weighted contrasts. (b) DevCCF LSFM autofluorescence templates (coronal slice) before multimodal registration to the MRI template. (c) Multimodal registration of E15.5 LSFM to MRI DWI templates. Top: horizontal plane. Bottom: coronal plane. (d) DevCCF E15.5 multimodal templates with unique and complementary contrasts. Note distinct anatomical marks differentially highlighted. (e) The CCFv3 template is registered to the P56 DevCCF in stereotaxic coordinates. Left: MRI DWI template midline sagittal slice. Right: co-registered MRI DWI, LSFM autofluorescence, and CCFv3 coronal template coronal slices.

Figure 3.. 3D anatomical labels based on a developmental ontology

(a) DevCCF ontology levels 0 through 5 displayed over an E15.5 sagittal slice. (b) Coronal MRI FA slices through the subpallium with anatomical segmentation overlays on the right hemisphere. Embryonic and postnatal brains are depicted at uniform scales, respectively. (c) Gene expression, antibody and histological staining, different template contrasts registered to the DevCCF P56 to guide segmentations. On the left: midline sagittal MRI ADC slice indicating location of each of the following coronal (labels 1–4) and horizontal (label 5) slices. (1) CCFv3 template; (2) ISH Efnb2 gene expression; (3) LSFM neurofilament staining, (4) MRI FA; (5) LSFM SYTOX nucleus staining with densely packed cerebellar granule cells. (d) 3D renderings of DevCCF annotations, not to scale.

Figure 4.. DevCCF to quantify early emergence of GABAergic neurons in embryonic brains

(a-b) Sagittal images of LSFM imaging from E11.5 Gad2-Cre;Ai14 mice show clusters of GAD2+ neurons in the subpallium (SPall), prosomere 1 and 3 (p1 and p3), rhombomere 1 basal plate (r1B). (c-d) The E13.5 (c) and E15.5 (d) brains show additional GAD2+ neurons in the cerebellar hemisphere (CbH). (e) Average GAD2+ signals at E11.5 (n=7), E13.5 (n=7), and E15.5 (n=5) with 3D rendering overlay on DevCCF templates. Note rapid expansion of GAD2+ neurons at E13.5 and E15.5 from initial clusters at E11.5. (f) In addition to local expansion, GAD2+ neurons migrate to deep (1) and superficial areas (2) of the pallium to establish cortical interneurons. (g) Quantification of GAD2+ signals in developmental neuromeres using DevCCF segmentations at E11.5, E13.5, and E15.5. Additional abbreviation: Pallium (Pall), terminal hypothalamus (THy), peduncular hypothalamus excluding telencephalon (PHy*), prosomere 2 (p2), midbrain (M), prepontine hindbrain (PPH), pontine hindbrain (PH), pontomedullary hindbrain (PMH), and medullary hindbrain (MH).

Figure 5.. DevCCF enables CCFv3 aligned data analysis with developmental framework.

(a) CCFv3 annotations (left hemisphere) and DevCCF P56 annotations (right hemisphere) overlayed on the DevCCF P56 template across five slices from anterior (left) to posterior (right). (b) Sagittal slices with CCFv3 annotations (left), DevCCF (middle), and DevCCF label boundaries overlaid on CCFv3 annotations (right). Neuromeric boundaries of the prosomere (p)1, p2, p3 highlighted with white dashed lines over colored CCFv3 annotations demonstrate shared and non-overlapping boundaries of CCFv3 and DevCCF annotations. (c) Sankey diagram illustrates matched structural relationship between DevCCF and CCFv3 annotations. Size of individual areas represents logarithmic scale of regional volume. (d) Single section of MERFISH spatial transcriptome data with three representative genes (top) and cell type classifications (bottom). (e) Registered spatial transcriptome with CCFv3 segmentation (left) and DevCCF segmentation (right). Colors are the same as in (c). (f) Heatmap of cell-type distribution in CCFv3 (top) and DevCCF segmentations (bottom). Heatmap values show proportion of cells in a region relative to the total number of cells in the subclass. Cell-types are ordered by their parent class, denoted by color bars on the x-axis.

Figure 6.. Web Visualization for DevCCF

(a) Layer panel allows layer selection with a right click and hiding by selecting the eye icon to the left of the layer. (b) Once selected, the layer edit tool enables user modification of layer color, contrast, and brightness. (c) The ontology viewer allows search and selection of individual segmentations and parent regions. When a region is selected in the viewer or ontology tool, the region’s metadata is displayed including region name, abbreviation, and ID. The ontology tool may be dragged to any location in the viewer. (d-f) Neuroglancer allows users to visualize either individual E15.5 LSFM autofluorescence in (d), MRI MTR template in (e), or their overlay (f). (g-i) E15.5 DevCCF templates overlaid (g) and segmentation (h, i). Black dashed line in (i) denotes sagittal slice location of (g,h).