limit AL neuron dendrite outgrowth. A second possibility is that PD173074 blocks a retrograde signaling mechanism in which activation of glial FGFRs and subsequent downstream events normally feed back onto ORN terminals, affecting their ability to signal to AL neurons and thereby control arborization of AL neuron dendrites. There is reason to consider this possibility, as we have shown that blockade of nitric oxide release from ORN terminals leads to a similar AL neuron dendrite overgrowth phenotype in combination with a lack of NP glial migration. We noted in that report that it was necessary to block release of nitric oxide several days before the normal start of glial migration and AL neuron outgrowth, raising the possibility that nitric oxide functioned, in part, to regulate gene expression in glia and/or AL neurons, preparing them to be able to respond to additional signals from ORN axons at the appropriate time. We know, too, from cirradiation and hydroxyurea experiments in which glial cell proliferation was blocked, that we do not see any obvious changes in development of the architecture of the antennal lobe until Glial FGFRs in Glia-Neuron Signaling approximately 75% of the glial cells are eliminated. Given that the complement of glial cells is not greatly reduced in the treated antennal lobes during the period of axon ingrowth, the effects on dendritic outgrowth are not simply a consequence of a severely reduced number of glial cells. Possible mechanisms of FGFR activation Vertebrate FGFRs can be activated by a large number of FGFs, and two FGFs, 17761171” named Pyramus and Thisbe or FGF8-like 1&2, have been described and shown to activate Heartless in Drosophila embryos and in imaginal eye discs of Drosophila larvae. Pyramus, in particular, is an attractive candidate for initiating the events described in the present work, as it induces glial cell proliferation and migration in the Drosophila eye disc. We have searched for Lepidopteran homologs via translated BLAST searches of Lepidopteran ESTs but have found no matches, so the question of whether classical ligands exist for the Manduca FGFR remains CP 868596 cost unanswered. FGFRs and EGFRs in both vertebrates and in Drosophila have been shown to be activated via homophilic and heterophilic interactions in cis and in trans between the IgCAMs L1/ Neuroglian, NCAM/Fasciclin II, and N-cadherin. We know from our previous work that the transmembrane form of Manduca Fasciclin II is expressed by ORN axons and Neuroglian is expressed by ORN axons and central glia. The GPI-linked isoform of MFas, GPI-Fas II, previously was detected in what ” appeared to be a subset of AN glia, but more recently has been found to be expressed by NP, SZ, and AN glia. These results are timely in view of recent 
work describing homophilic interactions between neuronal TM-Fas II and glial GPI-Fas II in Drosophila embryos, in which glial migration along axons is regulated by cell-surface expression of neuronal TM-Fas II. In summary, we have found evidence for activated FGFRs on central and peripheral glia of the primary olfactory pathway of Manduca sexta. Our results indicate that these FGFRs are not required for ORN axon targeting, but are essential for 1) the proliferation and survival of glial cells, 2) the ORN-induced migration of NP glial cells to surround glomeruli, 3) the glia-regulated establishment of normal dendritic territory in the glomeruli, and 4) the glia-induced organization of ORN axon fasciculation through the sort