Figure 4. Precocious Hoxa5 expression in Dicer mutant leads to a noisy Hox5 –Hox8 boundary in vitro and in vivo.

(a) CAGGs:CreER; Dicer^floxed ESC lines were treated with (+) or without (−) 4-hydroxytamoxifen (4-OHT) and differentiated with caudalized medium (see Experimental Procedures for details) on Day2. Postmitotic neurons were dissociated on Day5 and immunostaining was performed on Days 7 and 8 of differentiation. (b,c) Immunostaining of Day8 MN culture reveals that 4-OHT-treated cells display a significant increase in cells co-expressing Hoxa5 and Hoxc8, whereas non-treated cells manifest robust segregation of Hoxa5^on and Hoxc8^on cells (mean±s.d., n=3 independent experiments, *P<0.01). Scale bar represents 50 μm. (d) Immunostaining of spinal cord sections demonstrating noisy co-expressed Hoxa5/Hoxc8 in E10.5 (sagittal) and E12.5 (horizontal and transverse) Sox1^Cre/+; Dicer^f/f (Dicer^NeuralΔ) embryo sections, whereas mutually exclusive Hoxa5/Hoxc8 expression is present in the control embryos. Scale bar represents 50 μm. (e) Quantification of percentage of co-expressed Hoxa5^on/Hoxc8^on cells against Hoxa5^on cells at brachial spinal cord (number of positive cells per 20 μm ventral spinal cord section) in control and Dicer^NeuralΔ embryos (mean±s.d., n=9 embryos, *P<0.01). (f) Summary of boundary shift of Hoxa5-Hoxc8 in Dicer^NeuralΔ EBs and embryos.