Overview of the single-cell RNA-sequencing experiments.

a) Diagram of developmental milestones of SC and TM cells in mice. Red arrows indicate timepoints of sample collection. b) Diagram of single-cell RNA-sequencing experimental design. Dotted lines indicate cutting planes when dissecting the anterior segment and limbal regions. c) Uniform Manifold Approximation and Projection (UMAP) representation of the single cell transcriptomes from all timepoints colored by annotated cell types. d) Barplot showing the proportion of each cell type identified across the sampled developmental stages. Some caution is needed in relating to exact in vivo proportions due to isolation and processing procedures that can alter cell type proportions.

POM/NC-derived cell type analysis.

a) UMAP representation of POM-derived cells colored by high-resolution cell-type annotations. b) UMAP representation of POM-derived cells colored by sample timepoint. c) Dotplot showing the expression of key POM marker genes (columns) in each cell type, where size of dots indicates fraction of genes expressed and color indicates average level of expression. d) Barplot showing the fraction of each POM/NC- derived cell type across timepoints until P21.Data were integrated across ages, P60 omitted as integration with developing clusters obscured mature cell types at this age. Un-identified technical variation resulted in fewer TM3 being profiled at P21. e) UMAP representation of POM/NC derived cells at each timepoint, colored by their cell type annotations (same as in a).

Analysis of TM cell development.

a) 3D view of a segment of the TM across ages represented by the Surface mode in Imaris. The morphology of the SC facing side of the TM is shown. It becomes progressively more ordered along the longitudinal Y axis of the structure as development proceeds. With obvious beam like striations at P21. Scale bar = 50μm b) Visualization of a 1 um thick optically sectioned portion of the TM from a 3D TM reconstruction in the orthogonal XZ plane. Nuclei (DAPI, blue), cell borders (GFP, yellow), and open spaces (magenta, pseudocoloring) of the TM are shown across different developmental stages. At P6.5, the TM is dense with closely packed cells and nuclei (both rounded and elongated). Minimal spaces are evident between the cell borders. As differentiation proceeds (by P10) and inter-trabecular beams and spaces form (P14), the nuclei become elongated and more organized, with increasing volume of space between cells. Clear organization of elongated nuclei and robust inter-trabecular spaces are evident by P21 as major morphogenesis of the TM is complete. Scale bar = 50μm, red dotted line indicates TM border. c) Quantification of TM open space and volume with age. d) Violin plots showing expression distributions of genes enriched in TM cell subtypes across sample timepoints. e) Marker expression across developmental timepoints (immunofluorescence) MYOC – magenta, CRYM – red, a-SMA – green, TFAP2B – green (appears cyan because nuclear co-localization with DAPI). CD31 and yellow dotted lines mark the presumptive SC border. White lines mark TM cells. a: anterior, p: posterior. f) Left UMAP embedding of POM/NC-derived cells between P2.5-P10, colored by pseudo-time (Monocle) where the root node was chosen manually as a random cell in Dev5 cluster and labeled in the plot (see main text).Right: UMAP representation with monocle pseudo-time trajectory colored by cell types. g) Sankey diagram summarizing the development of three TM cell subtypes from clusters Dev1-Dev5.

Analysis of SEC development.

a) UMAP representation of endothelial cells colored by cell subtypes. b) Dotplot summarizing expression of marker genes across endothelial cell types where dot size indicates fraction of cells expressing each gene and color indicates average level of expression. c) UMAP representation of endothelial cells colored by sample timepoints. d) UMAP representation of endothelial cells colored by high-resolution cell type annotations across individual sample timepoints. Asterisk indicates the emergence of Schlemm’s canal endothelial cell cluster (SECs). e) Independent analysis of endothelial cells across developmental timepoints – dotplots of expression of key marker genes across each timepoint for each of the dynamically changing endothelial cell types. f) Dimensional reduction (UMAP visualization) and cell type annotation as defined separately for each timepoint.

Differentiation of IW and OW of SECs.

a) UMAP representation of endothelial cells colored by developmental pseudotime inferred by Monocle. b) Monocle trajectory superimposed on the UMAP representation of endothelial cells colored by cell type annotation, root node being programmatically determined. c) Visualization of marker gene expression plotted on UMAP embedding of endothelial cells. d) Scatter plot showing the average expression level of marker genes in a combined population of VP and SECs across sample collection timepoints. e) Immunofluorescence labeling of key SEC marker genes across developmental and adult timepoints, top panels showing the vascular plane, bottom panels showing the SC plane from the same Z-stack.

Dynamic gene expression analysis during Schlemm’s canal endothelial cell development.

a-b) Gene ontology (GO) enrichment of biological process genes selectively expressed during early development in a: VPs; b: SECs. In all GO plots, x-axis represents the fraction of genes in each pathway which are also identified in the corresponding differentially expressed gene set. Colors indicate adjusted p-value of enrichment, and dot size indicates number of genes in the pathway. c,d) Average expression level of genes in selected GO biological processes enriched in VPs and SECs during early development, across sample collection timepoints. e) GO enrichment of genes expressed during mid-developmental timepoints in SECs. f-i) Average expression level of genes in select processes enriched in VPs or SECs across sample collection timepoints. j) GO enrichment of genes expressed in SECs at late development. k-m) Average expression level of genes in select processes enriched in SECs across sample collection timepoints.

Ligand-target interaction analysis between TM and SC cells.

a-f) Circos plots demonstrating inferred interactions between TM (sender) and SE (receiver) cells across sample collection timepoints stratified by early (P2.5 and P4.5), mid (P6.5 and P10), and late (P14 and P21) timepoints.

Cell type annotation.

a) Heatmap showing the expression of signature genes (rows) across single cells (columns), arranged by seven major cell types. b) UMAP representation of all cells colored by sample collection timepoints, c) UMAP representation of all cells colored by cluster identity at each timepoint.

Identification of POM/NC derived cell types.

a) Visualization of selected marker genes -Tfap2b and Pitx2 - expression plotted on UMAP embedding of all cells. b) Heatmap of signature gene expression across single cells arranged by POM-derived cell types.

3-D analysis of TM development.

This figure describes the anatomic landmarks used to define the TM for Figures 3 and Figures S4 a) En face view of the entire circumference of the aqueous humor drainage structures and surrounding tissues. Fluorescent labeling was achieved by breeding together the Wnt1-Cre2 and mTmG alleles. This gene combination results in the SC and surrounding vasculature expressing tdTomato (red), while the TM, corneal endothelium, ciliary muscle, and lymphatics express GFP (green; see Methods). The ocular tissue was prepared as a whole mount and imaged through the Z-depth using confocal microscopy. Scale bar = 500 µm. b) En face view of a cropped region of the AH drainage structures, as indicated by grey lines in (a). The orientation icon in the upper right corner corresponds to that used in subsequent figures. c-e) YZ views spanning the full depth of the tissue, from the outer to inner surface of the eye. The limbal vascular plexus (LVP), containing both blood and lymphatic vessels, is located closer to the outer surface of the eye compared to the SC and TM. The LVP was optically removed to view the SC and TM. Collector channels (asterisks) are cropped to within 10 µm of their connection to the outer SC surface. f-g) En face views of the region between the white dotted lines in (d) and (e), with other regions cropped out. The SC is visible in (f), while the TM and adjacent structures are shown in (g). The TM exhibits the highest GFP signal intensity. At the posterior boundary of the TM, ciliary muscle fibers (arrows) run orthogonally to the TM. At the anterior boundary of the TM, larger, circular cell borders of the corneal endothelium are visible (hashtags). Dotted lines mark the anterior and posterior boundaries of the TM. h) For analysis, the structures surrounding the TM were further cropped out at the dotted lines in (g). All scale bars (b-h) = 100 µm

TM cell subtype classification.

a) View of the AH-facing side of the TM segments shown in Figure 3 (Imaris Surface mode). As development progresses, the structure becomes increasingly organized along its longitudinal axis. b) Separated panels of TM XZ optical sections from Figure 3. Empty spaces that lacked both nuclei, (DAPI, blue) and cells (GFP, yellow Methods) are marked (magenta, pseudo-color), red dotted line indicates TM border. c) Violin plot showing Crym expression across developmental clusters and ages. Crym is expressed in the Dev1 cluster from P2.5 to P10, which may explain the detection of CRYM protein in the developing TM prior to P6.5 (before TM2 cells are robustly identified). d) Immunofluorescence visualization of TM cell markers shown in Figure 3 at additional developmental ages. MYOC (TM1Hi, magenta) is absent in the TM at early stages but becomes detectable by P10. CRYM (TM2Hi, red) is present at all examined ages, though early expression may reflect Dev1 identity (see above). α-SMA (TM3Hi, green) is consistently expressed across all stages. CD31 and yellow dotted lines demarcate the presumptive SC boundary. White lines outline the TM cell group. a: anterior, p: posterior.

Developmental clusters and TM cell subtypes.

a) Hierarchical clustering of annotated cell types across P2.5-P10 developmental timepoints for all POM/NC derived cell types. b) Barplots demonstrating the proportion of cell cycle phases across sample collection timepoints for the five POM-derived developmental cell types.

GO enrichment analysis of TM developmental marker genes.

a, b) GO biological processes of TM3 subcluster marker genes (compared to other POM/NC derived cell types) at P2.5 and P4.5. c) GO molecular function of TM2 subcluster marker genes at P6.5. d) GO biological processes of TM2 subcluster marker genes at P10. e) GO molecular function of TM1 subcluster marker genes at P14. f) GO biological processes of TM1 subcluster marker genes at P21.

Identification of epithelial cell types.

a) Visualization of selected marker gene expression plotted on UMAP embedding of all cells. b) UMAP representation of epithelial cells colored by subtypes. c) UMAP representation of epithelial cells colored by collection timepoint. d) Heatmap of signature gene expression across single cells arranged by epithelial cell types.

Identification of ciliary body/iris/retinal pigment epithelium (RPE) cell types.

a) Visualization of selected marker gene expression plotted on UMAP embedding of all cells. b) UMAP representation of ciliary body/iris/RPE cells colored by subtypes. c) UMAP representation of ciliary body/iris/RPE cells colored by collection timepoint. d) Heatmap of signature gene expression across single cells arranged by ciliary body/iris/RPE cell types.

Identification of immune cell types.

a) Visualization of selected marker gene expression plotted on UMAP embedding of all cells. b) UMAP representation of immune cells colored by subtypes. c) UMAP representation of immune cells colored by collection timepoint. d) Heatmap of signature gene expression across single cells arranged by immune cell types.

Identification of progenitor cell types.

a) Visualization of selected marker gene expression plotted on UMAP embedding of all cells. b) UMAP representation of progenitor cells colored by subtypes. c) UMAP representation of progenitor cells colored by collection timepoint. d) Heatmap of signature gene expression across single cells arranged by progenitor cell types.

Identification of endothelial cell subtypes.

a) Visualization of selected marker genes plotted on UMAP embedding of all cells. b) Heatmap of signature gene expression across single cells arranged by endothelial cell types. c) Barplots demonstrating the fraction of cell cycle phase across sample collection timepoints for the five endothelial cell types. d) Dotplots representing the expression of selected marker genes (columns) across sample collection timepoints in each of the five endothelial cell types (rows). e) Normalized expression of genes along pseudotime with individual dots colored by chronological time.

GO molecular pathway enrichment analysis of genes across SEC development.

a-d) Enrichment of molecular function pathway genes selectively expressed during SEC development.

GO biological process enrichment analysis of genes across SEC development.

a-d) Enrichment of biological process pathway genes selectively expressed during SEC development.

Expression of pericyte marker genes.

Violin plots representing the expression profile of four selected pericyte marker genes across sample collection timepoints. Dots represent individual cells.

Predicted ligand-target interactions between TM and SC cells during development.

Heatmaps represent the regulatory potential between TM ligands (rows) and predicted receptors/targets in SC cells (columns) across three sample collection timepoints (a: P2.5; b: P10, c: P21). NicheNet prioritized extracellular signaling molecules (y-axis) and their predicted downstream direct and indirect targets in the receiving cell population (x-axis), which may include receptors, signaling mediators, ECM components, and transcriptional targets. Color intensity reflects the NicheNet regulatory potential score.

Expression levels of TM ligands during SC development.

a-f) Dotplots indicating the expression pattern of ligands (ligands) which were predicted to contribute to SC development in each of the three TM cell types (columns) across sample collection timepoints. Dot size is indicative of percent of cells expressing the ligand (not number of cells – due to low numbers of TM1 and TM2 cells at P2.5 and P4.5, the dots should be interpreted with caution at these ages).

Predicted ligand-target interaction between VP-SC and SC-SC interactions.

a) Heatmaps representing the regulatory potential between VP ligands (rows) and predicted receptors/targets in SECs (columns) at P4. b) Heatmaps representing the regulatory potential between SC ligands (rows) and predicted receptors on SC cells (columns) at P4.

Ligand-target interaction analysis between macrophages/pericytes and SECs.

a, b) Circos plot demonstrating inferred interactions between macrophages(sender) and SECs (receiver) at P4.5 and P21 b, c) Circos plot demonstrating inferred interactions between pericytes (sender) and SECs (receiver) at P4.5 and P21.

Ligand-target interaction analysis macrophages-TM and SC-TM.

a, b) Circos plot demonstrating inferred interactions between macrophages(sender) and TM (receiver) cells at P10 and P21 c, d) Circos plot demonstrating inferred interactions between SECs (sender) and TM (receiver) cells at P10 and P21.

Analysis of expression profile of glaucoma risk genes in SC and TM development.

a) Dot plot summarizing expression of glaucoma risk genes from IOP and POAG GWAS across all AS cell types. b, c) GO molecular function and biological process gene enrichment of genes that cause developmental glaucoma. d, e) Average expression level of glaucoma risk genes in SECs and TM cells across sample collection timepoints, developmental glaucoma genes are marked by a rectangular box.

List of antibodies used in the study

List of other resources and reagents

Markers for developing and adult TM and SC cell types