They found that endogenous coexpression
of ephrin-As in LMCM Z-VAD-FMK cells leads to a reduction in free (ligand-unbound) EphAs, thereby sharpening the differences between LMCM and LMCL neurons in their ability to bind ephrin-As in trans. The authors then performed a series of manipulations of ephrin expression in vivo by chick in ovo electroporation and assessed their effects on LMC axon pathway choices. Knockdown of ephrin-A5 in the spinal cord resulted in the change of trajectories of many LMCM axons, which chose to grow dorsally instead of ventrally, suggesting that they became more sensitive to ephrin-As in the ventral limb mesenchyme. Accordingly, gain-of-function experiments showed that ephrin-A5 overexpression in spinal motor Ulixertinib order neurons was sufficient to reroute some LMCL axons into the ventral trajectory, consistent with their loss of responsiveness to ephrin-As in the target tissue. Similar results were obtained with a mutated
version of ephrin-A5, which is defective in trans- but not cis-interactions ( Carvalho et al., 2006), confirming that the guidance errors are a consequence of altered cis-interaction between ephrins and Ephs on the same membrane and not of reverse signaling by the overexpressed ephrin-A5 activated in trans by limb-expressed EphAs. All findings were reproduced in the other respective LMC subpopulation by manipulations of ephrin-B expression. While these experiments in chick provide strong evidence for the existence of cis-attenuation in vivo, future studies with conditional mouse mutants in which the expression of ephrins can be controlled first in specific LMC subpopulations would be valuable. The authors then proceeded to directly test the responsiveness of
LMC axons to ephrins in the stripe assay. In agreement with the in vivo observations, LMCM neurons that were not repelled by stripes of exogenous ephrin-As under control conditions started avoiding these stripes upon knockdown of endogenous ephrin-As. Moreover, to exclude any influence of interaxonal interactions on LMC axon behavior, Kao and Kania reproduced their findings in a stripe assay with low-density dissociated LMC cultures. These functional data were complemented by visualization of Eph proteins on the LMC axons, revealing an inverse correlation between the level of ephrin expression and the abundance of free Ephs on the cell surface. Since ephrin-As in LMC neurons were previously shown to engage in trans-interactions with exogenous EphAs and mediate attractive responses ( Marquardt et al., 2005), the authors then extended their stripe assay to test the involvement of ephrin-As in cis-attenuation versus reverse signaling by challenging the axons with ephrinA- and EphA-coated stripes. Interestingly, it turned out that these two modes of ephrin-A interaction exist in different neurons, with cis-binding prevailing in LMCM and trans-binding in LMCL cells ( Figure 1).