Previously studied for its role in processing olfactory information in the antennal lobe, GABA also may shape development of the olfactory pathway, acting either through or on glial cells. Early in development, the dendrites of GABAergic neurons extend to the glial border that surrounds the nascent olfactory lobe neuropil. These neuropil glia express both GABAA and GABAB receptors, about half of the glia in acute cultures responded to GABA with small outward currents, and about a third responded with small transient increases in intracellular calcium. The neuronal classes that express GABA in vivo, the local interneurons and a subset of projection neurons, also do so in culture. Exposure to GABA in culture increased the size and complexity of local interneurons, but had no effect on glial morphology. The presence of glia alone did not affect neuronal morphology, but in the presence of both glia and GABA, the growth-enhancing effects of GABA on cultured antennal lobe neurons were eliminated. Contact between the glial cells and the neurons was not necessary. Operating in vivo, these antagonistic effects, one direct and one glia mediated, could help to sculpt the densely branched, tufted arbors that are characteristic of neurons innervating olfactory glomeruli.

In adult olfactory nerves of mammals and moths, a network of glial cells ensheathes small bundles of olfactory receptor axons. In the developing antennal nerve (AN) of the moth Manduca sexta, the axons of olfactory receptor neurons (ORNs) migrate from the olfactory sensory epithelium toward the antennal lobe. Here we explore developmental interactions between ORN axons and AN glial cells. During early stages in AN glial-cell migration, glial cells are highly dye coupled, dividing glia are readily found in the nerve and AN glial cells label strongly for glutamine synthetase. By the end of this period, dye-coupling is rare, glial proliferation has ceased, glutamine synthetase labeling is absent, and glial processes have begun to extend to enwrap bundles of axons, a process that continues throughout the remainder of metamorphic development. Whole-cell and perforated-patch recordings in vivo from AN glia at different stages of network formation revealed two potassium currents and an R-like calcium current. Chronic in vivo exposure to the R-type channel blocker SNX-482 halted or greatly reduced AN glial migration. Chronically blocking spontaneous Na-dependent activity by injection of tetrodotoxin reduced the glial calcium current implicating an activity-dependent interaction between ORNs and glial cells in the development of glial calcium currents.

The compound eye of the tobacco hornworm moth Manduca sexta contains green-, blue-, and ultraviolet-sensitive photoreceptors. Electroretinogram spectral-sensitivity measurements were recorded from different regions of the retina in order to broadly map the distribution of the three receptor types. The relative contribution of the three receptors to spectral-sensitivity curves was estimated by fitting theoretical curves based on the absorption spectra of the three rhodopsins. This analysis indicated that the dorsal retina is green and ultraviolet dichromatic, with green-sensitive cells greatly predominating. The ventral retina is trichromatic with a substantial population of blue- and ultraviolet-sensitive receptors. We previously showed that flower visitation for nectar feeding is mediated mainly by blue-sensitive cells. Their localization in the ventral retina seems an appropriate adaptation of the receptor mosaic, since the moths hover above flowers as they feed.

Larvae of the tobacco hornworm moth Manduca sexta were reared on either a carotenoid-supplemented or a carotenoid-deficient diet. The former yields fortified adults with normal visual function, whereas visual sensitivity and rhodopsin content are reduced by 2−4 log units in the compound eyes of the deprived moths reared on the latter. We characterized the retinoids of fortified retinas and investigated the recovery of visual function in deprived moths that were provided with retinaldehyde as a source of photopigment chromophore. Retinoids were identified and measured by high-performance liquid chromatography (HPLC). Fortified retinas contained mainly 3-hydroxyretinaldehyde (R3); 11-cis R3 predominated in dark-adaptation, all-trans in light-adaptation, indicating that R3 is the photopigment chromophore. No retinoids could be measured in deprived eyes. Retinaldehyde (R1) was delivered to the retinas of deprived moths by “painting” solutions of 11-cis or all-trans R1 in dimethylsulfoxide (DMSO) on the corneal surfaces of the compound eyes or on the head capsule between the eyes. 11-cis R1 induced rapid recovery: during 3 days, sensitivity rose to within a log unit of that measured from fortified animals. By 7 days, sensitivity was close to normal. Although rhodopsin and P-face particle densities of photoreceptor membranes increased, neither rose to the levels found in fortified animals. All-trans R1 induced only a slight increase in sensitivity that could have resulted from some nonspecific isomerization of the all-trans to the 11-cis isomer; we found no evidence for a retinal isomerase that functions in darkness. Small amounts of R3 were measured in recovering retinas, indicating some conversion of R1 to R3. However, the chromophore of most of the rhodopsin that was synthesized must have been R1. It is possible that rhodopsin did not reach normal levels in the retina even after a week of recovery because the normal chromophore R3 was not provided. Although the rhodopsin that initially formed in recovering moths may have resulted from the association of the chromophore with pre-existing opsin, the extent of eventual recovery indicates that opsin synthesis was stimulated by 11-cis R1.

Two distinct opsin-encoding cDNAs, designated MANOP1 and MANOP2, were isolated as 3′ fragments from the sphingid moth Manduca sexta. They were obtained by reverse transcription of retinal RNA and amplification with the polymerase chain reaction (PCR) using a degenerate primer designed to an amino-acid sequence conserved in arthropod opsins. The cDNA fragments labelled bands at approximately 1.8 kb on Northern blots of retinal RNA extracts. Levels of opsin message were compared in retinas from normal moths, whose diets were fortified with carotenoid precursors of the Manduca rhodopsin chromophore, 3-hydroxyretinal, and those reared on carotenoid/retinoid (vitamin A) deficient diets. The chromophore-depleted retinas contained more opsin mRNA; this was particularly true for MANOP2. Thus, the chromophore is not required for opsin gene transcription in Manduca.