This delay in the appearance of Tbx3 is consistent with a requirement for Tbx3 transcription and translation following retrograde BMP4 signaling. Similar increases in pSMAD and Tbx3 levels in the soma following axonal BMP4 treatment were also observed using western blotting ( Figures S1H and S1I). Notably,

the retrograde signal generated by axonal application of BMP4 is as robust as BMP4 signals generated by activation of BMP4 receptors in the cell body, based on their similar ability to elevate pSMAD and Tbx3 levels ( Figures S1J and S1K). We next asked if the retrograde signal is conveyed to the cell body by molecular motors. Application of the dynein inhibitor erythro-9-[3-(2-hydroxynonyl)] adenine (EHNA) (Penningroth, 1986) Small molecule library to the axonal compartment blocked the appearance of pSMAD1/5/8 and Tbx3 induced by axonal application of BMP4 (Figures 1B and 1C). Furthermore, expression of dynamitin, which disassembles the dynactin complex required for dynein motor transport (Burkhardt et al., 1997), similarly blocked the increase in nuclear pSMAD by BMP4 treatments in axons (Figure 1D). Together, these data suggest that signaling molecules conveyed from distal axons in a dynein-dependent manner mediate axonal BMP4 signaling. In the case of the neurotrophins,

retrograde signaling has been shown to be mediated by “signaling endosomes” containing endocytosed receptors that are translocated in an active ligand-bound form to the cell body where they activate their effectors (Cosker et al., learn more 2008 and Ibáñez, 2007). To test whether a similar mechanism may be involved in retrograde BMP4 signaling, we first asked whether the activity of BMP receptors is required in the cell body after axonal application of BMP4. Dorsomorphin, DMH2, and LDN193189, selective and reversible inhibitors of BMP type I receptors ALK2, ALK3, and ALK6, block BMP signaling, but not signaling by related TGF-β family for members (Yu et al., 2008). Application of these inhibitors to the cell body blocked

the increase in pSMAD1/5/8 and Tbx3 that was induced by application of BMP4 to axons (Figures 1B, 1C, and S1L–S1N). These data suggest that retrograde BMP4 signaling results in the appearance of active BMP4 receptors in the cell body, which is required for retrograde BMP4 signaling. In these experiments, we confirmed that cell body levels of SMAD1/5/8 were unchanged (Figure S1F) and no signs of cytotoxicity were observed during the time course of the experimental treatments (Figure S1O). Local action of the inhibitor in the microfluidic chambers was also confirmed (Figure S1P). To further test the idea that retrograde BMP4 signaling involves axonally derived activated BMP4 receptors, we monitored the translocation of BMP4 from the axon to the soma.