A conserved regulatory logic controls temporal identity in mouse neural progenitors

Abstract

Neural progenitors alter their output over time to generate different types of neurons and glia in specific chronological sequences, but this process remains poorly understood in vertebrates. Here we show that Casz1, the vertebrate ortholog of the Drosophila temporal identity factor castor, controls the production of mid-/late-born neurons in the murine retina. Casz1 is expressed from mid/late stages in retinal progenitor cells (RPCs), and conditional deletion of Casz1 increases production of early-born retinal neurons at the expense of later-born fates, whereas precocious misexpression of Casz1 has the opposite effect. In both cases, cell proliferation is unaffected, indicating that Casz1 does not control the timing of cell birth but instead biases RPC output directly. Just as Drosophila castor lies downstream of the early temporal identity factor hunchback, we find that the hunchback ortholog Ikzf1 represses Casz1. These results uncover a conserved strategy regulating temporal identity transitions from flies to mammals.

Figure 1. Mouse Casz1 Is the Ortholog of Drosophila castor and Is Expressed in RPCs during Mid-Retinogenesis

(A) Meme analysis (Bailey et al., 2009) of murine and Drosophila castor proteins identifies a signature HΩH|C motif, where “Ω” represents a bulky amino acid residue, and “|” divides the 2 zinc finger repeats of the domain (green brackets). Murine Casz1 isoforms contain 11 (Casz1v1) or 5 (Casz1v2) such zinc finger repeats (green boxes). (B–G) Spatiotemporal expression profile of Casz1 during retinogenesis. Arrows in (G) indicate rare amacrine cells that also express Casz1. (H–W) Co-expression of Casz1 (green) and Ki67 (blue) with the RPC marker Vsx2 (red, H–O), or the photoreceptor precursor marker Otx2 (red, P–W), at E16.5 or P0 as indicated. Arrow/circle colors refer to the markers expressed by the indicated cell. Scale bar, 20 μm. RPL, retinal progenitor layer; ONL, outer nuclear layer; INL, inner nuclear layer; GCL, ganglion cell layer.

Figure 2. Conditional Inactivation of Casz1 in RPCs Increases Early-Born Neuron Production at the Expense of Mid-/Late-Born Neuronal Fates

(A) Schematic of Casz1 alleles. Exons are shown as green boxes, zinc finger domains as purple boxes, loxP sites as blue triangles, and FRT sites as purple triangles. (B–D) Immunostaining shows loss of Casz1 protein (purple) in EYFP+ α-cKO cells (arrows). (E) Strategy for clonal inactivation of Casz1 in RPCs. (F–I) Examples of viral heterozygote control (v-Het) or conditional knockout (v-cKO) clones obtained 20 days after infection. Images of EYFP-positive clones (F and H) or camera-lucida drawing (G and I) are shown. R, rod; A, amacrine; B, bipolar; M, Müller. (J) Quantification of clonal composition expressed as fold change over control (dotted line). (K) Clone size distribution in v-Het and v-cKO clones. Bar graphs show mean ± SEM. See also Tables S1 and S2 for statistical details.

Figure 3. Heterochronic Expression of Casz1 in Early RPCs Increases Mid-/Late-Born Neuron Production

(A–F) Examples of clones obtained after infection of retinal explants at E13.5 with retroviral vectors expressing GFP alone (Control: A and B), or GFP and Casz1v1 (C and D) or Casz1v2 (E and F). (B), (D), and (F) show camera lucida drawings of the GFP-expressing cells in (A), (C), and (E). R, rod; A, amacrine; B, bipolar; M, Müller. (G) Quantification of clonal composition expressed as fold change over control. (H) Clone size distribution. Bar graphs show mean ± SEM. See also Tables S1 and S2 for statistical details.

Figure 4. Ikzf1 Represses Casz1 in RPCs

(A–H) Retinal explants transfected 14 days earlier with constructs expressing EGFP alone (Control), Ikzf1-IRES2-EGFP (Ikzf1), or Ikzf1-VP16-IRES2-EGFP (Ikzf1-VP16), stained for the early-born cell-type marker Pax6 (B and D), or the late-born Müller cell marker Ccnd3 (F and H). (I–N) Casz1 staining 24 hr after transfection of E15.5 explants with constructs expressing EGFP alone (Control), Ikzf1, or Ikzf1-VP16. (O) Fluorescence intensity measurements of Casz1 signal in Control, Ikzf1-, or Ikzf1-VP16-expressing cells relative to adjacent untransfected cells. **p < 0.01. Graph shows mean ± SEM. (P) Model of temporal control of RPC competence modified from birthdating data by (Rapaport et al., 2004). RPCs are multipotent, but their probability to give rise to early-, mid-, or late-born cell types changes over time. Ikzf1 and Casz1 manipulations bias RPC production of early (green), mid/late (magenta), and late fates (blue). Ikzf1 is expressed in early RPCs, promotes early-born fates (Elliott et al., 2008), and suppresses Casz1 indirectly (via factor X), while Casz1 is expressed in mid-/late-phase RPCs and increases the probability of generating of mid-/late-stage neurons. Ikzf1-VP16 acts oppositely to Ikzf1 and promotes late fates.