ebony encodes an β-alanyl-dopamine synthase regulating β-alanyl conjugation of dopamine and histamine, thus 'trapping' these biogenic amines preventing their further function (see Borycz, 2002 and Richardt, 2003). This enzymatic activity regulates fly pigmentation, photoreceptor activity and behavioral rhythmicity. It has been suggested that glia may be required for normal circadian behavior, but glial factors required for rhythmicity have not been identified in any system. This study shows that a circadian rhythm in Drosophila Ebony (N-β-alanyl-biogenic amine synthetase) abundance can be visualized in adult glia and that glial expression of Ebony rescues the altered circadian behavior of ebony mutants. Molecular oscillator function and clock neuron output are normal in ebony mutants, verifying a role for Ebony downstream of the clock. Surprisingly, the ebony oscillation persists in flies lacking PDF neuropeptide, indicating it is regulated by an autonomous glial oscillator or another neuronal factor. The proximity of Ebony-containing glia to aminergic neurons and genetic interaction results suggest a function in dopaminergic signaling. A model for ebony function is presented wherein Ebony glia participate in the clock control of dopaminergic function and the orchestration of circadian activity rhythms (Suh, 2007).

Because life has evolved in the presence of daily geophysical cycles, most organisms have acquired the ability to adapt the timing of physiological processes to external cycles using an intrinsic time-keeping device called a circadian clock. Both forward genetic and molecular screens in Drosophila and other organisms have identified genes encoding integral components of the circadian oscillator. In the fruit fly, the core oscillator mechanism governing behavioral rhythmicity is comprised of two interconnected molecular loops that result in circadian changes in PER and TIM clock protein abundance and the cyclical feedback repression of clock gene transcription. In addition to the core transcriptional loops, posttranscriptional factors have been identified that are required for the modulation of clock protein stability, activity, or nuclear entry. Although there has been significant progress in delineating clock mechanisms, less is known about the molecular and cellular output pathways that control organismal physiology and behavior (Suh, 2007).

Two behaviors are widely employed to assay circadian rhythmicity in Drosophila: eclosion (the emergence of the adult from the pupal case) and adult locomotor activity. Mutation of a clock element affects rhythms in both eclosion and activity, since the same or a molecularly similar clock regulates both behaviors. In contrast, several mutations have been reported to affect only one of these two behaviors; i.e., to have rhythm-specific effects on circadian behavior. These findings indicate that genetically separable output pathways mediate the circadian control of the two different processes. Mutations in ebony, for example, selectively perturb the locomotor activity rhythm, causing arrhythmicity, but have no effect on the adult eclosion rhythm (Newby, 1991). Such a rhythm-specific effect suggests that Ebony acts downstream of the clock mechanism to orchestrate the circadian control of locomotor activity (Suh, 2007).

Multiple microarray-based studies have identified Drosophila transcripts exhibiting rhythmic daily changes in abundance. These studies verified cycling for all of the known clock genes and, importantly, identified hundreds of other genes that show robust circadian changes in abundance within head tissues. Of note, ebony RNA was shown to exhibit robust circadian cycling in two independent studies (Claridge-Chang, 2001; Ueda, 2002). These results are consistent with the behavioral studies discussed above, which suggest that Ebony protein functions in a clock output pathway (Suh, 2007).

The most obvious phenotype of ebony mutants is defective sclerotization and cuticle pigmentation, although they also exhibit altered rhythms, vision (Hotta, 1969), and courtship behavior (Kyriacou, 1978). Consistent with these phenotypes, Ebony protein can be detected in the hypodermis (which produces the cuticle), the visual system, and other brain regions (Richardt, 2002). In the fly visual system, Ebony is localized exclusively to glia including neuropile and epithelial glia (Richardt, 2002), and it is thought that Ebony functions in a metabolic pathway (Borycz, 2002; Richardt, 2003) that may terminate the action of histamine, the photoreceptor cell neurotransmitter. Based on studies of the pigmentation phenotype of ebony mutants, it was shown that Ebony protein has β-alanyl-dopamine (DA) synthase (BAS) enzymatic activity (Hovemann, 1998), and consequently mutants are lacking N-β-alanyl-dopamine (NBAD) in peripheral and neural tissues (Perez, 2004) and have elevated levels of DA and β-alanine in both types of tissues (Hodgetts, 1973; Ramadan, 1993). Recently, it was reported that the Ebony enzyme has a broader substrate specificity than anticipated from previous studies: purified Ebony can conjugate β-alanine to several different biogenic amines, including DA, serotonin (5-HT), histamine, tyramine, and octopamine (Richardt, 2003); hence, it is now considered a β-alanyl-biogenic amine synthase (Suh, 2007).

It is known that DA, 5-HT, and other biogenic amines have neuromodulatory activity in Drosophila and other insects. Together with the behavioral defects of ebony mutants, these findings suggest a model for the circadian function of Ebony, in which clock output regulates Ebony (BAS) activity, and consequent changes in biogenic amine-related signaling within a specific group of neural cells of the fly brain. This study shows that Ebony-containing glia are localized close to clock cell projections, that there is a PER/TIM-dependent control of rhythmic ebony expression within a discrete population of glial cells, and that Ebony enzymatic activity is required within glia for the clock control of locomotor activity. Cellular and molecular analyses indicate that Ebony acts downstream of the clock to control locomotor activity and that Ebony-containing glia are positioned near DA and 5-HT neurons of the larval and adult brains, consistent with the idea that these glia are required for the modulation of aminergic functions. A genetic interaction between ebony¹ and an allele of the fly dopamine transporter gene (dDAT) suggests that dopaminergic transmission has a role in rhythmicity in vivo. That glia may function in rhythmicity is consistent with a genetic mosaic study that implies a role for PER/TIM-containing glia in the regulation of activity rhythms (Suh, 2007).

Studies of Ebony indicate that glia have an essential role in the orchestration of circadian locomotor activity. It is of interest that previous studies in both mammals and insects have suggested that glia might be important for the control of rhythmic physiological events. Cultured cortical astroglia that express per-luciferase transgenes, for example, show circadian rhythms of bioluminescence that may depend on diffusible signals from neurons of the suprachiasmatic nuclei; these studies suggest that such glia contain autonomous oscillators that can be reset by environmental stimuli or by interactions with clock neurons. In Drosophila, previous investigations have shown that the clock proteins PER and TIM can be detected in neurons and glia of the optic lobes and protocerebrum, and PER protein abundance fluctuates according to a circadian rhythm in both cell types. Consistent with roles for PER in neurons and glia, genetic mosaic analysis has suggested that per expression in either cell type might be sufficient for rhythmicity (albeit weak rhythmicity with glial expression). The current results indicate that Ebony is localized to glia of at least two types: those containing PER and TIM and a second class in which clock protein expression is not detectable. In the first class of cells, it seems likely that rhythmic ebony expression is controlled by an intracellular PER/TIM-based oscillator. In the latter class, ebony expression is most likely regulated by direct or indirect interactions with clock cells (Suh, 2007).

It is now an accepted axiom that neuron-glia interactions are critical for neuronal development and function. In addition to serving support roles in the mature nervous system, glial cells influence the developmental specification of neurons, migration, myelination, synapse number, and synaptic transmission. In Drosophila, studies have provided a detailed understanding of glial cell development and revealed the transcriptional mechanisms underlying the differentiation of this class of neural cells. Previous studies have shown that glial cells function in the phagocytosis of neuronal debris during development and documented roles for glia in injury-induced neuronal degeneration. Of note, a glial-specific receptor known as Draper has been described that is part of a neuron-glia signaling mechanism mediating such injury-induced responses. Insect glia have also been implicated in neurotransmitter uptake/recycling, based on studies of GABA, acetylcholine, or glutamate uptake. Finally, studies of Drosophila repo mutants have demonstrated that glial support is important for neuronal survival in insects, similar to results obtained in mammals (Suh, 2007).

Perhaps more relevant for behavior, recent studies have shown that certain types of mammalian glia (astrocytes) can regulate the excitability of neurons through the regulated release of “gliotransmitters” (glutamate, ATP, adenosine, cytokines, and growth factors), and it has become apparent that there are reciprocal neuron-glia signaling systems that regulate neuronal excitability. Although certain aspects of this dynamic communication system are beginning to be understood, clearly much remains to be learned about the specific factors that regulate neuron-glia communication. These studies have identified a glial-specific factor (Ebony) and a subpopulation of glia within the fly nervous system that function with clock neurons to regulate circadian activity rhythms. It seems likely that intercellular communication between the neuronal and glial elements of the fly circadian system is important for the temporal coordination of activity (Suh, 2007).

Does PDF release contribute to the control of rhythmic ebony expression? Colocalization studies, using antibodies for Ebony and PDF, show that greater than 80% of the Ebony-containing glia reside in close proximity to PDF neuronal somata or projections—this is the case for glia that reside in the lateral and dorsal protocerebrum and the optic medulla of the adult brain. Previous immunoelectron microscopy results show that certain PDF-containing varicosities are adjacent to glial cells of the optic medulla. Ebony-containing glia have been observed near varicosities of the PDF neuronal projections, which probably contain dense core vesicles (DCVs), and adjacent to other regions of the projections. Surprisingly, however, it appears that PDF is not essential for the regulation of the ebony rhythm. It is therefore postulated that communication between Ebony glia and clock neurons, if it occurs, is mediated by factors other than the PDF neuropeptide (Suh, 2007).

A transgene expressing an enzymatically dead form of Ebony does not provide behavioral rescue for ebony mutants; thus, BAS activity is essential for Ebony's circadian function. As indicated previously, BAS can conjugate β-alanine to many different aminergic neurotransmitters, including DA, 5-HT, histamine, octopamine, and tyramine (Richardt, 2003; Perez, 2004). Interestingly, it has been demonstrated that Ebony glia are situated near histamine release sites of photoreceptor cells in the lamina, and it has been suggested that BAS activity conjugates histamine to β-alanine to terminate action of the transmitter (Richardt, 2002). This study has shown that Ebony-containing cells are in close proximity to dopaminergic and serotonergic neurons of the larval and adult brains, suggesting a role for BAS in terminating DA and 5-HT action. A genetic interaction between e¹ and DAT^fmn, a DAT mutant, strongly suggests that Ebony has a role in dopaminergic signaling. The rhythmic production of Ebony (BAS) may result in a circadian modulation of DA action and in turn rhythmic regulation of locomotor activity. Alternatively, circadian changes in BAS activity may result in the rhythmic production and release of N-β-alanyl-dopamine (NBAD) with high levels of NBAD driving locomotor activity. Two lines of evidence support this idea: (1) NBAD is presumably highest during the day, the time of maximal activity; (2) the e¹ mutation, a protein null, eliminates NBAD, and this mutation suppresses the hyperactivity of DAT^fmn flies even though the double mutant is predicted to have high synaptic levels of DA (Suh, 2007).

This study has shown that Ebony glial expression is regulated in a circadian manner and that the protein is required within glia for normal behavioral rhythmicity. The localization of Ebony-containing glia near clock cells and aminergic neurons suggests an explicit model for Ebony regulation and function in the circadian system. According to this model, ebony transcription is regulated either directly by a PER/TIM-dependent oscillator within glia (for those glia containing PER and TIM) or by the release of an unidentified output factor from clock neurons. Consequently, diurnal changes in ebony-encoded and glial-localized BAS activity lead to rhythms in the conjugation of biogenic amines to β-alanine and generation of NBAA product (NBAD in glia near DA neurons). Such a diurnal modulation of amine action may help shape the temporal organization of the daily bouts of locomotor activity. This model, of course, implies the existence of a glial amine transporter that mediates the uptake of synaptic amines into glia, although such a system has not yet been identified in Drosophila (Suh, 2007).

Furthermore, it is postulated that the production of NBAD, which is high during the subjective day, serves as a bioactive compound to drive locomotor activity during the daytime. The observation that e¹ mutants exhibit selective daytime deficits in locomotor activity is consistent with this idea. According to this model, NBAD is released from glia and acts on dopaminergic or other neurons to regulate excitability and/or transmitter release. There is no evidence in the literature that β-alanyl-amine conjugates have bioactivity, but this is certainly a possibility given that many other glial compounds have such activity. Obviously, NBAD may not regulate locomotor activity, by itself, as it is presumably high throughout the day, given the profile of Ebony production, whereas locomotor activity is bimodal, with bouts occurring at dawn and dusk. An alternative model for the role of Ebony in the regulation of activity is that the unconjugated amine (i.e., DA) provides excitatory drive for behavior and that its modification by BAS activity decreases such excitation. However, such a model is not consistent with the presumption that NBAD levels are highest during the daytime, the time of maximal activity, nor with the observation that the DAT^fmn;e¹ mutant, which probably has high DA levels, is not hyperactive (Suh, 2007).

Finally, it is known that Drosophila tyrosine hydroxylase (TH) RNA is transcribed according to a circadian rhythm, with high abundance occurring during the subjective day, and it is thus a good assumption that TH enzymatic activity and DA production is maximal during the day. Such a profile of DA production may explain why a high constitutive expression of Ebony (BAS) in glia can restore rhythmic behavior. Because TH production is presumably still rhythmic in ebony mutants, high NBAD levels would be expected to occur in such flies only during the daytime, thus permitting behavioral rhythmicity (Suh, 2007).