Sema3A, a prototypical semaphorin, acts as a chemorepellent or a chemoattractant for axons by activating a receptor complex comprising neuropilin-1 as the ligand-binding subunit and plexin-A1 as the signal-transducing subunit. How the signals downstream of plexin-A1 are triggered upon Sema3A stimulation, however, is unknown. Here we show that, in the presence of neuropilin-1, the FERM domain-containing guanine nucleotide exchange factor (GEF) FARP2 associates directly with plexin-A1. Sema3A binding to neuropilin-1 induces the dissociation of FARP2 from plexin-A1, resulting in activation of FARP2's Rac GEF activity, Rnd1 recruitment to plexin-A1, and downregulation of R-Ras. Simultaneously, the FERM domain of FARP2 sequesters phosphatidylinositol phosphate kinase type I isoform PIPKIgamma661 from talin, thereby inhibiting its kinase activity. These activities are required for Sema3A-mediated repulsion of outgrowing axons and suppression of neuronal adhesion. We therefore conclude that FARP2 is a key molecule involved in the response of neuronal growth cones to class-3 semaphorins.

The human gene (PTGS2) encoding an inducible isozyme of prostaglandin-endoperoxide synthase (prostaglandin-endoperoxide synthase 2) that is distinct from the well-characterized and constitutive isozyme (prostaglandin-endoperoxide synthase l), was isolated using a polymerase-chain reaction-generated cDNA fragment probe for human prostaglandin-endoperoxide synthase 2. Nucleotide sequence analysis of the entire human prostaglandin-endoperoxide-synthase-2 gene demonstrated that it is more than 8.3 kb in size and consists of ten exons; this gene is very similar to the murine and chicken prostaglandin-endoperoxide-synthase-2 genes. The structures of exons in the human prostaglandin-endoperoxide-synthase-2 gene were also similar 'to those of the human prostaglandin-endoperoxide-synthase-1 gene (PTGS1). However, the sizes of introns in the human prostaglandin-endoperoxide-synthase-2 gene were generally smaller than those of the human prostaglandin-endoperoxide-synthase-1 gene. Primer-extension analysis indicated that the transcriptional-start site is 134 bases upstream of the translational-initiation site. The sequence of the 1.69-kb region of nucleotides preceding the transcriptional-start site and the first 0.8-kb intron contained a canonical TATA box and various transcriptional-regulatory elements (CArG box, NF-IL6, PEA-1, myb, GATA-1, xenobiotic-response element, CAMP-response element, NF-KB, PEA-3, Sp-1 and 12-0-tetradecanoyl-phorbol-13-acetate-response element). The nucleotide sequence of the 5'-flanking region (275 bp) of the human prostaglandin-endoperoxide-synthase-2 gene showed 63 % similarity to the sequence of murine prostaglandin-endoperoxide-synthase-2/TISIO gene, but essentially no homology to the chicken prostaglandin-endoperoxide-synthase-2 gene, and human and murine prostaglandin-endoperoxide-synthase-1 genes. A fluorescence in situ hybridization study showed that the human genes coding for prostaglandin-endoperoxide synthase 1 (PTGSI) and prostaglandinendoperoxidase synthase 2 (PTGS2) were mapped to distinct chromosomes 9q32-q33.3 and 1q25.2-q25.3, respectively, indicating that these genes are not genetically linked.Prostaglandin-endoperoxide synthase catalyzes the first committed step of the biosynthesis of prostaglandins, thromboxanes and prostacyclin [ 1, 21. Recent studies indicated that at least two distinct isozymes exist for prostaglandin-endoperoxide synthase (prostaglandin-endoperoxide synthase 1 and prostaglandin-endoperoxide synthase 2) [3 -61. The constitutive isozyme prostaglandin-endoperoxide synthase 1 was

The cardiac neural crest, a subpopulation of the neural crest, contributes to the cardiac outflow tract formation during development. However, how it follows the defined long-range migratory pathway remains unclear. We show here that the migrating cardiac neural crest cells (NCCs) express Plexin-A2, Plexin-D1 and Neuropilin. The membrane-bound ligands for Plexin-A2, Semaphorin (Sema)6A and Sema6B, are expressed in the dorsal neural tube and the lateral pharyngeal arch mesenchyme (the NCC "routes"). Sema3C, a ligand for Plexin-D1/neuropilin-1, is expressed in the cardiac outflow tract (the NCC "target"). Sema6A and Sema6B repel neural crest cells, while Sema3C attracts neural crest cells. Sema6A and Sema6B repulsion and Sema3C attraction are diminished either when Plexin-A2 and Neuropilin-1, or when Plexin-D1, respectively, are knocked down in NCCs. When RNAi knockdown diminishes each receptor in NCCs, the NCCs fail to migrate into the cardiac outflow tract in the developing chick embryo. Furthermore, Plexin-A2-deficient mice exhibit defects of cardiac outflow tract formation. We therefore conclude that the coordination of repulsive cues provided by Sema6A/Sema6B through Plexin-A2 paired with the attractive cue by Sema3C through Plexin-D1 is required for the precise navigation of migrating cardiac NCCs.

No abstract