Fig. 4. The expression of Nes is reduced by the decreased affinity of Brn2 phosphorylated on Ser362 for the Nes core enhancer during G2-M. (A) Immunostaining showed that only a subset of SOX1-SOX3 and Brn2 immunopositive cells expressed dVenus in the E14.5 cerebral cortex. (B) Flag-tagged Brn2- and Brn2-S362A-expressing 293T cells were synchronized at the G1, S and G2-M phases and labeled with [³²P]orthophosphate. The low level of Brn2 phosphorylation at G1 and S (lane 1, 2) increased during the G2-M phase (lane 3). However, the phosphorylation level of Brn2 S362A at G2-M (lane 5) was the same as in the G1 phase (lane 4). (C) SOX2, Brn2 and Brn2 S362D were used to transfect cells either alone, or as a SOX-POU combination, with a luciferase-encoding reporter plasmid carrying an octamerized nestin core enhancer sequence (8xNes30). Together, SOX2 and Brn2 synergistically activated the 8xNes30 reporter. By contrast, when Brn2 was replaced with Brn2 S362D, the activation was markedly lower. Data represent means ± s.e.m. (D) EMSA showing the lower binding affinity of Brn2 S362D (lane 3) compared with the wild-type Brn2 (lane 2) for the Nes30 probe (black arrow). The binding specificity was confirmed by a super shift assay (black arrowhead; lanes 4, 5). No binding reaction was seen with the lysate of untransfected 293T cells (lane 1). Open arrowhead indicates the free probe. (E) Western blot showing the Brn2 and Brn2 S362D input used for the EMSA. No significant difference was seen between the stability of Brn2 and Brn2 S362D when normalized to -tubulin expression (^*). Scale bar: 25 µm.