With the discovery that a novel form of the clathrin coat adaptor AP-1 (containing a distinct μ1B subunit) plays VEGFR inhibitor a critical role in basolateral sorting (Fölsch et al., 1999), the elucidation of the machinery for dendritic sorting seemed to be only a matter of time. This expectation turned out

to be far too optimistic. It was soon established that AP-1B is not expressed in neurons, and, as the new decade dragged on, the machinery responsible for recognizing dendritic sorting signals remained as mysterious as ever. In this issue, Farías et al. (2012) finally report progress on this key problem. They identify AP-1 as the missing link and demonstrate its essential role in the sorting of a variety of dendritic proteins, including several neurotransmitter receptors. A recent collaboration between the Rodriguez-Boulan and Bonifacino laboratories showed that AP-1A (the form of AP-1 containing the μ1A subunit)—previously thought to be involved principally in trafficking between the trans-Golgi network, endosomes, and lysosomes—also plays a key role in the sorting of basolateral proteins ( Gravotta et al., 2012). It appears that AP-1A works principally at the trans-Golgi complex while AP-1B acts during endosomal recycling. This result prompted the Bonifacino group to ask whether AP-1A might play a role in dendritic targeting in neurons ( Farías et al., 2012). The authors first performed a rigorous mutational analysis to precisely

identify the dendritic targeting signal in the transferrin receptor (TfR), a protein whose sorting has been well characterized in both MDCK cells and neurons. They identified a tyrosine-based Small molecule library YxxΦ motif in the cytosolic

N-terminal tail of TfR that is essential for its dendritic polarity. Overexpressed wild-type TfR is about ten times more concentrated in the somatodendritic domain than in axons of cultured hippocampal neurons. Mutating the tyrosine residue at position 20 caused TfR to accumulate equally in both the axonal and somatodendritic domains. The structural basis for binding between AP-1A and peptide sorting motifs has not been established, so the authors turned to the homologous AP-2 adaptor, which directs the clathrin-mediated Levetiracetam endocytosis of proteins containing a YxxΦ motif (Kelly and Owen, 2011). Using the known crystal structure of the homologous μ2 subunit, Farías et al. identified residues on the C terminus of μ1A that are likely candidates for interacting with the N-terminal targeting signal of TfR. They then used a yeast two-hybrid screen to characterize the binding between μ1A and the TfR tail. Using this approach, they identified a tryptophan residue (W408) in μ1A that was essential for binding to the TfR tail. Interestingly, the coxsackievirus and adenovirus receptor (CAR), another dendritic protein whose sorting has been well characterized in epithelia, also interacts with μ1A, and this interaction is also disrupted by mutating W408.