Kohara for cDNAs, S. Mitani (the Japanese National Bioresource Project) for the dlk-1(tm4024) mutation, W. Xiong and U. Mueller for advice on cell culture, A. Pasquinelli for her generosity in sharing equipment and laboratory space, and Z. Kai, E. Finnegan, and J. Broughton for their time and help in the northern blotting experiment. We thank A.D. Chisholm for critical insights Imatinib datasheet in data interpretation and our laboratory members for discussions and comments on the manuscript. D.Y. was an Associate of the Howard Hughes Medical Institute and is now supported by K99/R00 award K99NS076646. Y.J. is an Investigator of the Howard Hughes Medical Institute. This work was also supported by NIH R01 NS035546

(to Y.J.) and R01 NS057317 (to A.D. Chisholm and Y.J.). D.Y. and Y.J. designed the experiments. D.Y. performed the experiments. D.Y. and Y.J. analyzed and interpreted the data and wrote the manuscript. “
“Analysis of synapse formation in vitro has facilitated great advances in our understanding of synaptic differentiation in CNS. Various molecules that directly regulate

the formation and differentiation of synapses (synaptic organizers) have been identified (Fox and Umemori, 2006). Although differentiation MK-1775 nmr of pre- and postsynaptic sites must be coordinated by reciprocal interaction across synaptic clefts, the mechanisms by which this process is regulated in vivo are not well understood and could differ in different synapse types. For example, first axonal terminals may convert a preexisting shaft synapse into

a spine synapse in neocortical and hippocampal pyramidal neurons (Miller/Peters model) (Harris, 1999; Miller and Peters, 1981; Yuste and Bonhoeffer, 2004). Alternatively, immature dendritic protrusions (filopodia) may capture mobile axonal terminals and induce new synapse formation (filopodial model) (Knott et al., 2006; Okabe et al., 2001; Vaughn, 1989; Ziv and Smith, 1996). Interestingly, a completely different mechanism has been proposed for synapse formation between cerebellar Purkinje cells (PCs) and parallel fibers (PFs), the axons of the granule cells (Sotelo, 1990; Yuste and Bonhoeffer, 2004). In this Sotelo model, dendritic spines are formed autonomously without the influences of presynaptic terminals. Indeed, in the absence of granule cells in weaver or reeler mutant mice, PCs develop spines with almost normal morphology and postsynaptic densities ( Sotelo, 1990). Such spines without presynaptic terminals are called “naked spines” and have been observed transiently during normal development in the cerebellum ( Larramendi, 1969). Nevertheless, little is known about how presynaptic structural changes are induced and how they lead to differentiation of mature synapses. Cbln1 is a C1q family protein, which is produced and secreted from cerebellar granule cells (Hirai et al., 2005).