, 2006). We prepared two siRNA cocktails, each containing three siRNAs, one for each of the SMAD1, 5, and 8 isoforms. Transfection of either SMAD1/5/8 knockdown cocktail, but not the nontargeting control siRNA, into the axonal compartment resulted in significant reduction in axonal SMAD levels, measured using isoform-specific antibodies

( Figures S6A–S6C) or using an antibody to SMAD1/5/8 ( Figure 6A). Axonal transfection Palbociclib cost of either siRNA cocktail did not significantly affect transcript levels in cell bodies, as measured by SMAD1, 5, and 8 FISH analysis ( Figures S6D–S6F). These results confirm that compartmentalized siRNA transfection only affected SMAD1, 5, and 8 transcript levels in axons. To determine if axonal SMAD mediates retrograde BMP4 signaling, we applied BMP4 to axons after compartmentalized knockdown of SMAD1/5/8. In these axonal SMAD1/5/8-deficient neurons, retrograde BMP4 signaling was significantly impaired, as measured by reduced induction of nuclear pSMAD1/5/8 and Tbx3 by axonal BMP4 treatment ( Figures 6B, 6C, and S6G). No effect on find more retrograde trafficking of BMP4 endosomes was observed in axons transfected with the SMAD siRNA cocktail ( Figure S6H). These data indicate that axonal SMAD is required for retrograde BMP4 signaling.

Our finding that axonal SMAD1/5/8 mediates retrograde BMP4 signaling suggests that factors which regulate axonal synthesis of SMAD1, 5, and 8 would be required for proper patterning of the trigeminal ganglia. To screen for factors that induce SMAD1/5/8 synthesis, we examined SMAD1/5/8 levels in severed axons after application of different signaling molecules. Standard trigeminal ganglia neuron culturing media contains NT-3 and NGF. However, in this experiment, the axonal compartment of E13.5 trigeminal neurons was Calpain switched to neurotrophin- and BMP4-free media for 24 hr prior to severing. Notably, levels

of axonal SMAD1/5/8 were essentially abolished 24 hr after switching to this media. We first asked if BMP4 induces SMAD1/5/8 in axons. However, treatment of axons with BMP4 did not increase SMAD1/5/8 levels significantly higher than background (Figures 7A and 7B). We next considered the possibility that neurotrophins might regulate axonal SMAD levels, since these molecules have been shown to induce intra-axonal protein synthesis (Cox et al., 2008, Hengst et al., 2009, Yao et al., 2006 and Zhang and Poo, 2002). Application of NGF did not induce a significant increase in axonal SMAD levels, while application of NT-3 caused a small but significant increase in axonal SMAD levels (Figures 7A and 7B). However, despite the effect of NT-3, it is unlikely to be the major physiological regulator of axonal SMAD1/5/8 synthesis since NT-3 is expressed in the ophthalmic, maxillary, and mandibular regions of the face (Arumäe et al.