14-3-3 proteins have been postulated to modulate growth cone turning by stabilizing the interaction between the regulatory and catalytic subunits of PKA, thereby reducing PKA activity (Kent et al., 2010). Therefore, an increase in 14-3-3 proteins should lead to a decrease in PKA activity. To assess the levels of active PKA, we used an antibody that

recognizes the activated form PLX4032 mw of the catalytic subunit of PKA: phospho-PKA. Western blotting of lysates from dissociated commissural neurons showed that the levels of phospho-PKA at 3–4 DIV were about one-third lower than the levels at 2 DIV (Figure 4E). PKA phosphorylates the PP-1 inhibitory protein I-1 (phospho-I-1) in growth cones; thus, phospho-I-1 staining is another indicator of PKA activity (Han et al., 2007). Consistent with the decrease in phospho-PKA observed by western blotting, phospho-I-1 staining in commissural neuron growth cones was also significantly lower at 3 DIV compared to

2 DIV (p = 0.0158) (Figure 4F). Hence, the increase in 14-3-3 protein expression at 3 DIV correlated with a decrease in PKA activity. We hypothesized that the increase in 14-3-3 protein levels may mediate the switch in Shh response from attraction to repulsion. To test this hypothesis, we inhibited 14-3-3 activity with R18 (PHCVPRDLSWLDLEANMCLP), a peptide antagonist that inhibits binding of all 14-3-3 isoforms to their Ser/Thr phosphorylated targets. In particular, R18 has been shown to inhibit the binding Ku 0059436 of 14-3-3γ to PKA (Kent et al., 2010). The control WLKL peptide (WLDL mutated to WLKL) does not bind to 14-3-3. Both the R18 peptide and WLKL control peptide were fused to YFP and to Tat to allow entry into cells (Dong et al., 2008). Commissural axons, which are normally repelled by Shh at 3 DIV (Figures 3A–3F), continue to do so in the

presence of the control Tat-WLKL-YFP, with a mean angle turned of −9.5° ± 3.8° (Figures 5A and 5B). Remarkably, in the presence of the inhibitory Tat-R18-YFP, 3 DIV commissural axons were attracted by a Shh gradient, with a mean angle turned of 8.1° ± 3.5° (Figures 5A and 5B). There was a dramatic shift in the distribution of the angles turned from mostly negative in the presence of WLKL, to mostly positive when Linifanib (ABT-869) 14-3-3 proteins were inhibited by R18 (Figure 5A). In contrast, R18 had no effect on net axon growth under the same conditions (Figure S2A). To exclude the possibility of R18 having nonspecific effects, we also used shRNAmir targeted against 14-3-3β and 14-3-3γ, the two isoforms most prominently expressed in postcrossing commissural axons, to knock down 14-3-3 proteins in commissural neurons. Commissural neurons were transfected with plasmids encoding shRNAmir against 14-3-3β or 14-3-3γ. We were able to reduce 14-3-3β and 14-3-3γ protein levels to about 30% of control levels (Figures S2B and S2C).