The epigenetic modifier DNMT3A is necessary for proper otic placode formation

Abstract

Cranial placodes are thickenings in the ectoderm that give rise to sensory organs and peripheral ganglia of the vertebrate head. At gastrula and neurula stages, placodal precursors are intermingled in the neural plate border with future neural and neural crest cells. Here, we show that the epigenetic modifier, DNA methyl transferase (DNMT) 3A, expressed in the neural plate border region, influences development of the otic placode which will contribute to the ear. DNMT3A is expressed in the presumptive otic region at gastrula through neurula stages and later in the otic placode itself. Whereas neural plate border and non-neural ectoderm markers Erni, Dlx5, Msx1 and Six1 are unaltered, DNMT3A loss of function leads to early reduction in the expression of the key otic placode specifier genes Pax2 and Gbx2 and later otic markers Sox10 and Soho1. Reduction of Gbx2 was first observed at HH7, well before loss of other otic markers. Later, this translates to significant reduction in the size of the otic vesicle. Based on these results, we propose that DNMT3A is important for enabling the activation of Gbx2 expression, necessary for normal development of the inner ear.

Keywords: DNMT3A; Gbx2; Otic; Pax2; Placode.

Fig. 1

DNMT3A is expressed at sites of otic placode formation. DNMT3A transcripts are expressed (A) in the neural plate border region (HH 4), (B, C) the dorsal neural folds and otic placode (HH 6, 8). Dashed circles indicate position of otic precursors/placode. Black lines indicate position of section (A’, B’, C’). Right half of embryo is shown for sections. Arrowheads mark the position of the otic placode/vesicle. HH, staging after Hamburger and Hamilton.

Fig. 2

Morpholino-mediated knock down of DNMT3A reduces otic marker expression and otic vesicle size. DNMT3A MO or control MO were electroporated into the right or left half of the embryo, respectively, at HH4. Morpholinos were FITC-labeled (green, small insets). (A, A’) Gbx2 RNA expression was reduced in the pre-placodal domain at HH7 on the 3A MO electroporated side. Black line indicates level of section (A’). (B) At HH9 Gbx2 RNA expression was reduced in the otic placode on the 3A MO electroporated side. Black line indicates level of section (B’). (C) At HH13 otic vesicle size was reduced on the 3A MO injected side. (D) Graph shows quantification of Gbx2 loss of expression phenotype comparing 3A MO injected sides (n=18) to control MO or uninjected sides (n=26) at HH9-10. (E) DNMT3A morpholino knock down causes reduced Pax2 RNA expression in the otic area. Black line indicates level of section (E’). (F) The graph shows a quantification of the Pax2 loss of expression phenotype. Knock down, n=24; control, n=14. Arrowheads mark the position of the otic placode/vesicle on the 3AMO electroporated side. (G, G’) Expression of epidermal marker Ck19 is unaltered upon DNMT3A MO knockdown (n=9). Bar, 50 μm.

Fig. 3

Loss of Gbx2 is rescued by exogenous DNMT3A. Embryos were electroporated with DNMT3A MO in combination with (A) empty vector or (B) DNMT3A expression construct or (C) embryos were electroporated with control MO and DNMT3A expression vector. (D) Graph represents a quantification of embryos with a strong, mild or no effect on Gbx2 expression phenotype upon DNMT3A MO knock down compared to rescue and overexpression experiments. Knockdown, n=9; rescue, n=13; overexpression, n=7.

Fig. 4

Knock down of DNMT3A does not affect expression of early neural plate border and non-neural ectoderm markers Msx1, Erni, Dlx5 and Six1. HH4 embryos were electroporated with FITC labeled control MO (left side) or DNMT3A MO (right side). RNA levels of neural plate border markers (A) Msx1 (n=8/9), (B) Erni (8/8) and the non-neural ectoderm markers (C) Dlx5 (n=8/8) and (D) Six1 (n=6/7) were not reduced on the DNMT3A MO injected side. Black line indicates level of section (A’–D’). Bar, 50 μm.

Fig. 5

Otx2, Sox2 and Sox3 expression remains unchanged upon DNMT3A knock down. Embryos were injected with 3A MO on the embryonic right side. Morpholinos were FITC-labeled (green, small insets). DNMT3A knock down leaves (A) Otx2 (n=13/13), (B) Sox2 (n=7/7) and (C) Sox3 (n=7/8) expression unaltered. Dashed circles indicate position of otic domain. (D) Model of function of DNMT3A during otic development. In presence of DNMT3A, CpG islands in the putative repressor binding sites are methylated, preventing repressor binding to the target gene. When DNMT3A is decreased, methylation of repressor binding sites is reduced or not present. This allows repressors to bind and prevents target gene expression.