Blastomere Transplantation Allows Clonal Analysis of RPCs

(A) A core network of four key TFs can explain much of the cellular diversity in the retina.

(B) Cells from H2B-GFP, Ptf1a-dsRed double transgenic embryos were transplanted into WT embryos at 3.5 hpf.

(C) Embryos were screened for isolated RPCs at 24 hpf.

(D) At 72 hpf differentiation is completed with radial clones generated by transplanted cells. The asterisk marks a dAC.

(E) Quantification of cell fate distribution in clones generated by WT RPCs into WT hosts. Cell numbers and SDs are indicated as are the percentages of an average clone. Un, unknown. Scale bars represent 20 µm in (C) and 5 µm in (D).

See also Tables S1 and S2.

Intrinsic Impact on Cell Fate

(A) Fate distribution of clones generated by Ptf1a, Atoh7, or Vsx1 morphant cells transplanted into WT hosts. indicates significance for the merged p values of one- and two-cell compared with WT, while # denotes significant difference between one- and two-cell clones within a particular morphant. */# p < 0.05, **/## p < 0.01, ***/### p < 0.001. Error bars depict SEM. The figure legend in (A) is also valid for (B). For the number of clones for the different treatments, see Table S1, and for statistical calculations, see Table S3.

(B) Average clone sizes generated from single RPCs scored at 24 hpf for the various morphants in WT environment.

(C–F) Representative micrographs of clones generated by Atoh7, Ptf1a, or Vsx1 morphant cells in a WT environment, respectively. The red channel is shown individually for the Ptf1a morphant clone in (E) to reveal the fate switch performed by the Ptf1a lineage to PRs, BCs (marked by arrow head) and GCs (marked by hollow arrow head). Arrowhead in (F) indicates a Ptf1a-positive BC. The scale bar in (C) represents 10 µm and is also valid for (D)–(F).

See also Figure S4 and Table S2.

HCs Are Generated as the Intersection between the Lhx1 and Ptf1a Lineages

(A) The retina of a quadruple transgenic zebrafish (Atoh7-gapGFP; Ptf1a-dsRed; Lhx1-GFP; Crx-gapCFP) at 96 hpf. Individual channels with arrowheads illustrating the overlap between GFP and dsRed.

(B) Quadruple transgenic retina at 48 hpf. Arrows indicate Lhx1-GFP, Ptf1a-dsRed double positive cells, i.e., HCs. Individual channels reveal overlap between Crx-CFP and Lhx1-GFP, as indicated by arrowheads. pPR, putative PR.

(C) Combinatorial expression in the quadruple transgenic allows identification of the different cell types in a dissociated sample. Channel designation below is valid for both (C) and (D).

(D) Lhx1-positive PR cell.

(E) Percentages of cells that express Lhx1, Ptf1a, or both Lhx1 and Ptf1a.

The scale bar in (A) represents 20 µm and is also valid for (B), while the scale bar in (C) represents 5 µm and is also valid for (D).

Extent of Extrinsic Feedback during Retina Development

(A) Fate distribution of clones generated by WT cells transplanted into Ptf1a, Atoh7, Vsx1 or Atoh7, Ptf1a double morphant hosts. indicates significance compared with WT, while # denotes significant difference between one- and two-cell clones within a particular morphant. Figure legend is also valid for (B). Error bars depict SEM.

(B) Average clone size generated from single RPCs scored at 24 hpf for WT cells and the various environments previously described.

(C–E) Representative micrographs of WT clones generated in Atoh7, Ptf1a, or Vsx1 morphant environments, respectively.

(F) Checker-plot visualizing the extent of intrinsic regulation and extrinsic feedback in the different Atoh7 morphant scenarios. Color denotes deviation from WT to WT, such that a brighter color corresponds to an increase while a darker corresponds to a decrease of cell numbers of a given cell type.

(G and H) The intrinsic model is modified by the introduction of a delay postponing the onset of neurogenesis in 60% of the RPCs by one division. The resulting clone size (G) and fate (H) distributions match the experimental clones. Inset depicts averages of the experimental values and the values from the modeling with corresponding SDs.

(I–K) As in (F)–(H) but for the Ptf1a morphant environment with the same delay introduced to the intrinsic model.

The scale bar in (E) represents 10 µm and is also valid for (C) and (D). See also Figures S5 and S6 and Table S5.