Neurons contain abundant subsets of highly stable microtubules that resist depolymerizing conditions such as exposure to the cold. Stable microtubules are thought to be essential for neuronal development, maintenance, and function. Previous work has indicated an important role of the microtubule-associated protein STOP in the induction of microtubule cold stability. Here, we developed STOP null mice. These mice were devoid of cold-stable microtubules. In contrast to our expectations, STOP−/− mice had no detectable defects in brain anatomy but showed synaptic defects, with depleted synaptic vesicle pools and impaired synaptic plasticity, associated with severe behavioral disorders. A survey of the effects of psychotropic drugs on STOP−/− mice behavior showed a remarkable and specific effect of long-term administration of neuroleptics in alleviating these disorders. This study demonstrates that STOP is a major factor responsible for the intriguing stability properties of neuronal microtubules and is important for synaptic plasticity. Additionally, STOP−/− mice may yield a pertinent model for study of neuroleptics in illnesses such as schizophrenia, currently thought to result from synaptic defects.

STOP proteins are microtubule-associated, calmodulinregulated proteins responsible for the high degree of stabilization displayed by neuronal microtubules. STOP suppression in mice induces synaptic defects affecting both short and long term synaptic plasticity in hippocampal neurons. Interestingly, STOP has been identified as a component of synaptic structures in neurons, despite the absence of microtubules in nerve terminals, indicating the existence of mechanisms able to induce a translocation of STOP from microtubules to synaptic compartments. Here we have tested STOP phosphorylation as a candidate mechanism for STOP relocalization. We show that, both in vitro and in vivo, STOP is phosphorylated by the multifunctional enzyme calcium/calmodulin-dependent protein kinase II (CaMKII), which is a key enzyme for synaptic plasticity. This phosphorylation occurs on at least two independent sites. Phosphorylated forms of STOP do not bind microtubules in vitro and do not co-localize with microtubules in cultured differentiating neurons. Instead, phosphorylated STOP co-localizes with actin assemblies along neurites or at branching points. Correlatively, we find that STOP binds to actin in vitro. Finally, in differentiated neurons, phosphorylated STOP co-localizes with clusters of synaptic proteins, whereas unphosphorylated STOP does not. Thus, STOP phosphorylation by CaMKII may promote STOP translocation from microtubules to synaptic compartments where it may interact with actin, which could be important for STOP function in synaptic plasticity.

cADPR (cADP-ribose), a metabolite of NAD+, is known to modulate intracellular calcium levels and to be involved in calcium-dependent processes, including synaptic transmission, plasticity and neuronal excitability. However, the enzyme that is responsible for producing cADPR in the cytoplasm of neural cells, and particularly at the synaptic terminals of neurons, remains unknown. In the present study, we show that endogenous concentrations of cADPR are much higher in embryonic and neonate mouse brain compared with the adult tissue. We also demonstrate, by comparing wild-type and Cd38-/- tissues, that brain cADPR content is independent of the presence of CD38 (the best characterized mammalian ADP-ribosyl cyclase) not only in adult but also in developing tissues. We show that Cd38-/- synaptosome preparations contain high ADP-ribosyl cyclase activities, which are more important in neonates than in adults, in line with the levels of endogenous cyclic nucleotide. By using an HPLC method and adapting the cycling assay developed initially to study endogenous cADPR, we accurately examined the properties of the synaptosomal ADP-ribosyl cyclase. This intracellular enzyme has an estimated K(m) for NAD+ of 21 microM, a broad optimal pH at 6.0-7.0, and the concentration of free calcium has no major effect on its cADPR production. It binds NGD+ (nicotinamide-guanine dinucleotide), which inhibits its NAD+-metabolizing activities (K(i)=24 microM), despite its incapacity to cyclize this analogue. Interestingly, it is fully inhibited by low (micromolar) concentrations of zinc. We propose that this novel mammalian ADP-ribosyl cyclase regulates the production of cADPR and therefore calcium levels within brain synaptic terminals. In addition, this enzyme might be a potential target of neurotoxic Zn2+.