Proteins constituting the presynaptic machinery of vesicle release undergo substantial conformational changes during the process of exocytosis. While changes in the conformation make proteins vulnerable to aggregation and degradation, little is known about synaptic chaperones which counteract these processes. We show that the cell adhesion molecule CHL1 directly interacts with and regulates the activity of the synaptic chaperones Hsc70, CSP and αSGT. CHL1, Hsc70, CSP and αSGT form predominantly CHL1/Hsc70/αSGT and CHL1/CSP complexes in synapses. Among the various complexes formed by CHL1, Hsc70, CSP and αSGT, SNAP25 and VAMP2 induce chaperone activity only in CHL1/Hsc70/αSGT and CHL1/CSP complexes, respectively, indicating a remarkable selectivity of a presynaptic chaperone activity for proteins of the exocytotic machinery. In mice with genetic ablation of CHL1, chaperone activity in synapses is reduced and the machinery for synaptic vesicle exocytosis and, in particular, the SNARE complex is unable to sustain prolonged synaptic activity. Thus, we reveal a novel role for a cell adhesion molecule in selective activation of the presynaptic chaperone machinery.

The L1 cell adhesion molecule promotes neurite outgrowth and neuronal survival in homophilic and heterophilic interactions and enhances neurite outgrowth and neuronal survival homophilically, i.e. by self binding. We investigated whether exploitation of homophilic and possibly also heterophilic mechanisms of neural stem cells overexpressing the full-length transmembrane L1 and a secreted trimer engineered to express its extracellular domain would be more beneficial for functional recovery of the compression injured spinal cord of adult mice than stem cells overexpressing only full-length L1 or the parental, non-engineered cells. Here we report that stem cells expressing trimeric and full-length L1 are indeed more efficient in promoting locomotor recovery when compared to stem cells overexpressing only full-length L1 or the parental stem cells. The trimer expressing stem cells were also more efficient in reducing glial scar volume and expression of chondroitin sulfates and the chondroitin sulfate proteoglycan NG2. They were also more efficient in enhancing regrowth/sprouting and/or preservation of serotonergic axons, and remyelination and/or myelin sparing. Moreover, degeneration/dying back of corticospinal cord axons was prevented more by the trimer expressing stem cells. These results encourage the view that stem cells engineered to drive the beneficial functions of L1 via homophilic and heterophilic interactions are functionally optimized and may thus be of therapeutic value.