Six3 activation of Pax6 expression is essential for mammalian lens induction and specification

Abstract

The homeobox gene Six3 regulates forebrain development. Here we show that Six3 is also crucial for lens formation. Conditional deletion of mouse Six3 in the presumptive lens ectoderm (PLE) disrupted lens formation. In the most severe cases, lens induction and specification were defective, and the lens placode and lens were absent. In Six3-mutant embryos, Pax6 was downregulated, and Sox2 was absent in the lens preplacodal ectoderm. Using ChIP, electrophoretic mobility shift assay, and luciferase reporter assays, we determined that Six3 activates Pax6 and Sox2 expression. Misexpression of mouse Six3 into chick embryos promoted the ectopic expansion of the ectodermal Pax6 expression domain. Our results position Six3 at the top of the regulatory pathway leading to lens formation. We conclude that Six3 directly activates Pax6 and probably also Sox2 in the PLE and regulates cell autonomously the earliest stages of mammalian lens induction.

Figure 1

Conditional deletion of Six3 in the PLE causes severe lens phenotypes. Compared to age-matched control lenses (A–D), Six3^f/Δ;Le-Cre mutant eyes showed a range of lens defects at representative stages. (A′) At E12.5, the mutant eye appeared smaller, and the shape was abnormal. The shape of the retinal-pigmented epithelium (arrowhead) was defective. (B′) At E14.5, defects in the shape of the mutant eyes were consistently identified, and the lens (arrowhead) was extremely reduced. (C′) Three-month-old isolated mutant eyes exhibited obvious retinal defects such as irregular shape (arrow). The isolated lens was very small and degenerated and appeared to have cataracts (arrowhead). (D′) At 6 months, the eye lids of Six3^f/Δ;Le-Cre mutant mice were closed (arrowhead), suggestive of anophthalmia or microphthalmia. (E–P) H & E–stained sections of control and Six3^f/Δ;Le-Cre-mutant lenses. At E10.5, some mutants exhibited a smaller abnormal lens pit (arrows in F, G). (H) No lens pit (arrow) was present in those with the most severe phenotype, and the retina appeared to have been duplicated (arrowhead). At E12.5, some type I mutants exhibited slightly reduced lens vesicle (arrow in J), and some type II mutants showed greatly reduced ones (K). No lens vesicle was present in type III mutants (arrow in L). Similar phenotypic variations were seen at E14.5. The lens was smaller (N) or very reduced and abnormal (O, higher magnification in P). The lens stalk was frequently observed (arrow in N).

Figure 2

Lens specification and differentiation is not altered in the moderately affected Six3^f/Δ;Le-Cre mutants. Expression of FoxE3 (A′), Prox1 (B′), and β-crystallin (C′) in the moderately affected E12.5 lenses indicated that placode formation and lens differentiation had occurred in some Six3^f/Δ;Le-Cre embryos. Nevertheless, the polarity of the lens vesicle was affected; persistent abnormal FoxE3 expression was detected in the posterior part of the lens vesicle (A′, arrow). (D′) Increased apoptosis was detected in E10.5 Six3-mutant lens pit (arrowhead in rectangle; 3±1% for the control, 25±5% for the mutant, n=3). (E–E′) The percentage of BrdU⁺ cells was normal in E10.5 Six3-mutant lens pit (arrowhead in rectangle; 42±0.6% for the control, 41±1% for the mutant, n=3); however, ectopic BrdU⁺ cells were found in the posterior part of the lens vesicle at E12.5 (arrowheads in F′).

Figure 3

Lens induction and specification are defective in the severely affected Six3^f/Δ;Le-Cre-mutant lens. Cre activity was restricted to the PLE in E9.5 (A) and E10.5 (A′) Six3^f/Δ;Le-Cre-mutant embryos. (B′) Six3 expression was deleted from the mutant PLE (arrow) but was unaffected in the developing optic vesicle (arrowhead). (C′) Pax6 expression in the PLE was drastically reduced (arrow), and that of Sox2 was absent (arrow in D′); however, Pax6 and Sox2 expression in the optic vesicle remained normal (arrowheads in C′, D′). (E′–K′) Lens induction and specification was arrested, as indicated by the lack of thickening or invagination of the PLE and the absence of Sox2 expression. In E10.5 Six3^f/Δ;Le-Cre mutant embryos, Six3 expression in the PLE was deleted (arrows in E′) but not that in the neural retina lineage. In this embryo, the optic vesicle appeared to be duplicated. (F′) Pax6 expression was downregulated in the PLE (area between arrowheads), and that of Sox2 (G′, region between the two arrows) was absent. Expression of lens differentiation markers Prox1 (H′), sFrp2 (I′), and FoxE3 (J′) was not detected in the mutant tissue, where the lens pit should have formed. (K′) Expression of Meis1 remained unchanged.

Figure 4

The timing of Six3 deletion is crucial for the lens phenotype. Ubiquitous, early deletion of Six3 via tamoxifen injection of pregnant Six3^f/Δ;CAGG-Cre-ER dams at E7.8 and E8.5 (A′–F′) and late deletion via injection at E9.5 and E10.0 (G′–I′) were performed. Mutant embryos were analyzed at E10.5. Only early deletion of Six3 caused a failure in lens specification. (A′) Six3 expression was not detected in the PLE (arrow) or optic vesicle (arrowhead). (A′, B′) The PLE did not thicken and the lens placode did not form. (C′, D′) Defective lens induction and specification was confirmed by the absence of Pax6 and Sox2 from the PLE of the mutant embryos (arrows). (E′) Expression of Meis1 remained in the SE. Although Six3 was deleted ubiquitously and the shape of the optic vesicle was abnormal, neural retina expression of Pax6 (arrowhead in C′), and Chx10 (arrow in F′) remained. (G′–I′) Embryos subjected to late Six3 deletion exhibited normal or slightly small lens pits. Pax6 expression was not affected (arrow in H′), and Sox2 was either not affected or was slightly reduced (I′). (J) Graphic representation of the number of analyzed mutant embryos with early or late Six3 deletion per the observed lens phenotype (normal, small, or no lens pit (LP)).

Figure 5

Early deletion of Six3 downregulates Pax6 preplacodal expression. Ubiquitous, early deletion of Six3 by injection of tamoxifen at E7.8 and E8.5 in Six3^f/Δ;CAGG-Cre-ER mutant embryos analyzed at E9.0 (18-somite stage). (A′, D′) Six3 activity was almost completely removed from the PLE (arrowheads) and optic vesicle. Pax6 (B′, E′) and Sox2 (C′, F′) expression was reduced in the PLE, although with varying efficiency.

Figure 6

Six3 expression in the PLE precedes that of Pax6 and is initially unaffected in small eye-mutant embryos. (A) Six3 is expressed in the PLE (arrowhead) of a wild-type 5-somite-stage embryo. (A′) Pax6 expression is not yet detected in that region. At later stages, Six3 (B–D) and Pax6 (B′–D′) are detected in the PLE (arrowheads) and optic vesicles. Six3 expression in the PLE was examined in small eye (Pax6^{Sey-1Neu/Sey-1Neu})-mutant embryos. Pax6 was not required for the initial expression of Six3 in the PLE (arrowheads in E′, F′) but was required for its maintenance in the PLE at later stages (arrowheads in G′, H′).

Figure 7

Six3 is a direct activator of Pax6 and most likely of Sox2 lens lineage expression. (A) Bacterially expressed Six3 protein (full-length) and [³²P]-labeled Pax6 EE, Pax6 SIMO, Sox2 N3, and Sox2 N4 enhancer fragments were subjected to EMSA. Six3-bound DNA fragments are indicated by solid arrowheads and super-shifted bands (following addition of anti-Six3 antibody), by open arrowheads. The binding specificity was determined by the ability of the complex to be competed by 400 × excess of nonradioactive probes. Position of free probe is indicated by the solid arrows. (B) Six3 protein was enriched in selected regions of the Pax6 EE, Pax6 SIMO, and Sox2 N4 enhancers in vivo. Chromatin prepared from 11- to 17-somite stage heads and trunks was used for the ChIP assays and real-time PCR analysis. For Pax6 EE, Pax6 SIMO, and Sox2 N4 enhancers, the calculated ratio when using the heads as a DNA source was significantly higher (>5) than the ratio obtained when using the trunks (1.2–2.97), a result indicating that Six3 protein was bound to those regions. For the Sox2 N3 enhancer, the ratio for the head (1.52) was relatively similar to that for the trunk (1.2–2.97), suggesting that Six3 protein was not bound to that region, at least at that particular stage. (C) Luciferase transcriptional reporter assays showed that Six3 activates Pax6 lens lineage expression in the context of the Pax6 EE. The pCAB-Six3 (blue bars) plasmid activated the Pax6 lens lineage reporter plasmid (Pax6-EE-luc); pCAB-Six3VP16 (green bars) enhanced this activation, and pCAB-Six3EnR (red bars) attenuated it. Each bar represents six replicates (±s.d.). (D) Luciferase reporter assay also showed that Pax6 EE regulates Pax6 lens lineage expression by Six3. Pax6-ΔEE-luc (red bars) attenuated the activation of luciferase activity (blue bars). Each bar represents six replicates (±s.d.). (E) Six3 might also regulate Sox2 lens lineage expression. The pCAB-Six3 slightly increased RL-TK-N4 reporter activity (green bars), and slightly affect that of RL-TK-N3 (red bars), as compared to the activity of the control RL-TK reporter (blue bars). Each bar represents four replicates (±s.d.). (F) Electroporation of pCAB-Six3-IRES-GFP into the SE of chick embryos at HH8 stage (4- to 5-somite stage) specifically expanded the Pax6 expression domain only on the electroporated side (arrow). (G) Proposed gene-regulation network for lens cell fate specification. Solid arrows indicate direct regulation, and dashed arrows indicate potential direct regulation that has not yet been proven.