BIOLOGICAL OVERVIEW

takeout (to) is a Drosophila circadian clock-regulated output gene, a transcriptional target of the central clock. The Takeout amino acid sequence shows similarity to two ligand binding proteins, including juvenile hormone binding protein. Takeout mRNA is expressed in the head and the cardia (proventriculus), crop (foregut), and antennae -- structures related to feeding. to expression is induced by starvation, which is blocked in all arrhythmic central clock mutants, suggesting a direct molecular link between the circadian clock and the feeding/starvation response. A takeout mutant has aberrant locomotor activity and dies rapidly in response to starvation, indicating a link between locomotor activity, survival, and food status. It is proposed that takeout participates in a novel circadian pathway target that conveys temporal and food status information to feeding-relevant metabolisms and activities (Sarov-Blat, 2000).

Although the function of most circadian rhythm output genes is unknown, the Drosophila lark gene encodes an RNA binding protein involved in the regulation of eclosion (McNeil, 1999), and the mammalian transcription factor DBP exhibits remarkable circadian oscillations and has been shown to control the circadian expression of known downstream genes (Lavery, 1999). However, the mechanism(s) by which the central clock conveys temporal information to downstream behavioral and physiological processes is largely unknown. Both direct neural connections and humoral control have been proposed. Potent circadian activities are found in serum, suggesting that cycling humoral molecules constitute the connection between the brain and the peripheral pacemakers in mammals (Sarov-Blat, 2000 and references therein).

There are now several hormones known to be involved in circadian rhythms. Arguably, the best known example is melatonin. Others include vasopressin, leptin, and hypocretin. These hormones are involved in controlling different behavioral and physiological output activities. Vasopressin, an antidiuretic neuropeptide involved in salt and water balance, has been suggested to be directly controlled by the central clock transcription factor heterodimer Clock-BMAL1, the mammalian homolog of Clock-Cycle. A similar picture has recently emerged for Albumin D-binding protein (DBP). a PAR (proline and acidic amino acid-rich) leucine zipper transcription factor that is expressed according to a robust circadian rhythm in the suprachiasmatic nuclei, harboring the circadian master clock, and in most peripheral tissues (Sarov-Blat, 2000 and references therein).

In Drosophila, the humoral control of circadian behavior is indicated by classic transplantation experiments. A first candidate peptide factor has been proposed, and very recent results indicate that this putative hormone (PDF) is necessary for free-running locomotor activity rhythms. Recent evidence suggests that Pdf expression is controlled by the central pacemaker. It would not be surprising if other hormones important for other specific circadian output pathways were found (Sarov-Blat, 2000 and references therein).

The sequence of Takeout is suggestive of a second hormone involved in a circadian output pathway; it resembles a group of lipophilic ligand binding proteins. to was found in a PCR-based cDNA subtraction screen, in which poly(A)+ RNA from heads of cycle null mutant (cyc⁰¹) flies was subtracted from wild-type RNA (So, 2000). to gene expression was down-regulated in all of the clock mutants tested. In wild-type flies, TO mRNA exhibits daily cycling expression but with a novel phase, delayed relative to those of the better-characterized clock mRNAs, Period and Timeless. The E-box-containing sequence in the to promoter shows impressive similarities with those of period and timeless. However, it appears that the E box is not involved in the amplitude and phase of the transcriptional cycling of to. The circadian delayed transcriptional phase is therefore most likely the result of indirect regulation through unknown transcription factors (So, 2000).

takeout involvement in feeding was tested by measuring expression levels under starvation conditions. This is because starvation effects are common for genes involved in lipid and glucose metabolism and even in the regulation of appetite. Indeed, TO mRNA levels are increased after 9 to 10 hr of food deprivation, and a 2 hr refeeding reverses the starvation effects. Induction is most striking in the adjacent gastrointestinal tract regions where to expression was not detectable under normal conditions. tim RNA expression in the gut is more widespread than that of to under normal conditions, and to expression appears to expand and coincide with the tim pattern only when the flies are starved. Induction is also observed in the head. The starvation-induced enhanced expression is also detectable at the protein level. The rapid reversal by refeeding is much less obvious at the protein level, most likely because of a delayed effect on protein by mRNA changes. No enhanced expression is observed with heat shock or oxidative stress (Sarov-Blat, 2000).

Interestingly, an increase in to gene expression is most prominent when the flies are deprived of food at the peak of the mRNA cycle, suggesting that functional clock machinery may be required for this starvation-driven increase in expression. Consistent with this notion, induction is not detectable in arrhythmic clock mutants: per⁰¹, tim⁰¹, and cyc⁰¹ flies. The starvation-induced expression of to could therefore be secondary to an effect on a central clock component. To address this possibility, protein levels of Per and Tim were measured under the same starvation protocol. The results show that they are not enhanced, unlike to. Therefore, starvation-induction of to expression is probably not a secondary consequence of general clock gene induction. Rather, it requires and perhaps works through the clock machinery (Sarov-Blat, 2000).

A to deletion mutant was identified during the analysis of to expression levels in various genetic backgrounds. TO mRNA levels are low in the common lab strain ry⁵⁰⁶. In addition, Takeout protein levels in ry⁵⁰⁶ are noncycling and reduced 3-fold, when compared to trough levels of a wild-type (wt) strain. Takeout levels in ry⁵⁰⁶ are also not increased in response to starvation. These effects are not due to the ry⁵⁰⁶ mutation itself, because other lab strains bearing the same ry mutation showed normal to expression. to maps far from ry: to is located at 96B19-96C6, whereas ry is located at 87D11. The ry⁵⁰⁶ mutation was generated by gamma-ray mutagenesis, suggesting that a significant DNA lesion might be present at the to locus. To test this directly, a PCR analysis across the to genomic area was performed in the ry⁵⁰⁶ strain. To facilitate the analysis, the genomic sequence of to was obtained. The PCR results show that the 3' end region of to is missing in ry⁵⁰⁶ flies. Because PCR using another pair of primers reveals the presence of the normal stop codon, the genomic deletion in this ry⁵⁰⁶ strain probably removes the polyA cleavage site or a substantial portion of the 3' UTR, resulting in an unstable transcript (Sarov-Blat, 2000).

Because of the to response to starvation, locomotor activity patterns were examined without food. Flies were transferred to locomotor activity monitors under starvation conditions. Over the subsequent two days, average activity events and the percentage of active flies were monitored as a function of time in LD. The latter parameter approximates survival, since the inactive flies are dead or nearly so. Wild-type flies manifest at least three activity peaks, similar to the morning and evening peaks observed under standard LD conditions. The percentage of active flies slowly decreases during the two days without food. By both criteria, ry506 flies become less active more quickly, i.e., exhibit defects in both circadian activity and survival under these conditions. They show only two peaks of activity, the second significantly smaller than the first, and the average death time (50% inactive flies) was 30 hr for wt flies and 20 hr for the ry⁵⁰⁶ strain. However, the ry⁵⁰⁶ flies were more active than wt flies during the activity decline that followed the first activity peak (Sarov-Blat, 2000).

To confirm that the differences between ry⁵⁰⁶ and the wt strain are the result of the to mutation, two transgenes were introduced into a to mutant background: a tim GAL4 gene and a UAS-to gene. This combination leads to potent to expression in all tim-expressing cells. Two different UAS-to transgenes were examined, and in both cases there was a significant increase in the second activity peak, a small third peaklet, and a dramatically improved survival by about 10 hr of starvation. The transgenic rescue strains also show decreased locomotor activity relative to to after the first peak, even when only active flies are analyzed. No differences in survival rate were observed under heat shock and oxidative stress conditions; in these cases, the to mutant is identical to the wild-type strain. The strain differences indicate that to contributes to wild-type activity patterns and survival under starvation conditions (Sarov-Blat, 2000).

Only under starvation conditions were a survival and locomotor activity phenotype of ry⁵⁰⁶ flies detected, along with convincing to transgenic rescue. The starvation paradigm was inspired by the to expression pattern, which is localized to body tissues that are relevant to food detection and metabolism: the cardia, the crop, and the antennae. mRNA levels are increased by starvation, in the brain as well as these body tissues, and to affects survival and locomotor activity patterns under these conditions. On this basis, it is suggested that the to brain expression is relevant to feeding (Sarov-Blat, 2000).

Before analyzing to body expression and the locomotor activity patterns under starvation conditions, attempts were made to find a to behavioral phenotype under normal conditions. This was also based on a prior report that rosy alleles have a late eclosion phase, although the free-running period of locomotor activity and eclosion is normal. This phenotype was reproducable and a reproducible delayed locomotor activity phase was also found in ry⁵⁰⁶ flies. But the eclosion profiles and the delayed locomotor activity phase of the ry⁵⁰⁶ lines are not rescued with a ry⁺ transgene, and they are also unaffected by the presence or absence of the to deletion (Sarov-Blat, 2000).

to gene expression is not only under starvation control but also under clock gene control. It exhibits daily oscillating expression and is downregulated in all the clock mutants tested. Based on run-on experiments, to cycling is largely transcriptional, as previously described for per and tim. But the TO mRNA cycle peaks several hours after per and tim, and to transcriptional oscillations may not be dependent on a cis-acting E box (Sarov-Blat, 2000).

It is presumed that clock regulation of to expression contributes to metabolic and even behavioral fluctuations that are relevant to food and feeding. This relationship is underscored by the to starvation response and the lack of a response to starvation in the arrhythmic clock mutant backgrounds. The latter observation implies that the upregulation occurs through a clock mechanism or perhaps requires a functional clock. Consistent with this notion, per⁰¹ and tim⁰¹ flies die even more rapidly than to flies under starvation conditions. As all to positive tissues appear to express per and tim, this relationship between the circadian clock and to expression might be intracellular. It is likely that to transcription rates change in response to starvation. A circadian regulation of important output functions is consistent with the proposed circadian regulation of olfactory organ activity in Drosophila and in other insects, as recently shown by physiological assays. It is also consistent with the regulation of PDF in Drosophila.

Many to expression features recall leptin, NPY, VGF, and hypocretin -- mammalian hormones with proposed behavioral as well as metabolic functions. These shared properties include circadian regulation, a role for peripheral as well as brain tissue, a starvation response, secretion, and a relationship to ligands and receptors (ligand binding proteins). The latter connection comes from sequence analysis, which suggests that the shared property of the to superfamily is ligand binding. Since juvenile hormone and the molecules bound by JP29 are lipophilic, one can speculate that Takeout also has an endogenous lipophilic ligand relevant to feeding. The features displayed by the to superfamily also resemble the well-characterized lipocalin protein family, present mostly in vertebrates. Lipocalins are classified as extracellular transport proteins, typified by the retinol binding protein, RBP. RBP serves to regulate retinol release from the liver and to transport the insoluble retinol to peripheral target tissues. In general, these small, secreted transporters also serve to protect the ligands from degradation or chemical modification in the circulation. Lipocalin family members display low levels of overall sequence conservation, i.e., pairwise sequence identity is often below 20%, the threshold for reliable alignment. However, after the initial identification of this family, a growing number of crystallographic structures have been solved, and they reveal a compelling structural similarity (Sarov-Blat, 2000 and references therein).

The levels of this putative Takeout ligand might be clock-regulated. In this case, Takeout would serve to amplify or modulate the signal strength. Alternatively, the ligand concentration might be temporally constant, and the clock would then create the metabolic rhythm by controlling the rhythmic expression of Takeout -- the ligand binding protein. Takeout might also control the timing or duration of signal activity, resembling a proposed function of some secreted odorant binding proteins. Starvation would then further increase the signal (Sarov-Blat, 2000).

Based on circadian as well as starvation regulation, Takeout may contribute to an anticipation of food availability. Alternatively, it may contribute to a proper response to a change in food status. A specialized version of the second possibility is a response to starvation. This might include behavioral as well as metabolic changes caused by the absence of food. A similar speculation has been made in the case of NPY and mammals. Starvation might elicit an increase in locomotor activity, to stimulate a search for food. A heightened anticipatory event fits with the normal circadian regulation of Takeout expression, which occurs despite the constant presence of food. Alternatively, starvation might cause a shutdown of activity, to conserve energy, for example. Although the Takeout behavioral response to starvation is only now being subjected to experimental analysis, the mutant versus rescued activity patterns suggest that the mutant flies decrease their activity levels less rapidly during the first activity decline after about 10 hr of starvation. This could contribute to the less successful survival of the mutant strain. Alternatively, it could be irrelevant to survival rate and an independent manifestation of Takeout function under starvation conditions. It could also be a more subtle consequence of the Takeout-mediated events that also lead to decreased survival. If the latter is the case, however, it is not simply due to an increased percentage of dead (viz. immobile) flies, because the mutant strain is still hyperactive after 10 hr of starvation when immobile flies are excluded from the data. In any case, the to responses appear specific for starvation, because there is no effect of heat shock or oxidative stress on to gene expression. Moreover, there is no difference in survival between the to and wild-type strains in response to these two other stresses (Sarov-Blat, 2000).

Although a more detailed examination of the behavioral phenotype is required, two molecular issues are now of great interest. It is important to identify the putative Takeout ligand, as well as its putative intracellular receptor. Identification of both of these molecules will provide important tools and should help deepen an understanding of the relationship between food and circadian rhythms (Sarov-Blat, 2000).

GENE STRUCTURE

cDNA clone length - 1064

Exons - 3 or 4 (there are two splice variants)

Bases in 3' UTR - 288

PROTEIN STRUCTURE

Amino Acids - 249

Structural Domains

Takeout is a member of a novel family of insect proteins. All members share sequence similarity throughout their entire lengths, and they are all about 250 amino acids in length. However, sequence analysis with an iterated BLAST search suggests that To is also similar to two hydrophobic ligand binding proteins: hemolymph juvenile hormone binding protein JHBP (Touhara, 1993; Du, 1994) and JP29 (Wojtasek, 1995) from moths. These two ligand binding proteins form a superfamily with Takeout and the 0.9 protein. The latter is encoded adjacent to the period gene and is expressed at eclosion (Lorenz, 1989), which is under clock control (Sarov-Blat, 2000).

The sequence similarity between To and the known ligand binding proteins extends throughout the complete sequence. Pairwise alignment of To and the hemolymph JHBP from Manduca sexta shows an overall 24% identity and 54% similarity. Interestingly, both categories of ligand binding proteins are small like the To family members, with about 250 amino acids in all three proteins. In addition, some other To family members show even higher similarity with the ligand binding proteins. For example, a BLAST search against a nonredundant database using family member AI142207 also recognizes JP29, with an expect value of 2 x 10^-13. Such transition homology further strengthens the evolutionary relationship between To and these known ligand binding proteins (Sarov-Blat, 2000 and references therein).

Moreover, the N-terminal ligand binding fragment defined in JHBP (Touhara, 1993) is the best conserved region within the Takeout family. JHBP also has an almost identical secondary structure prediction profile with that of Takeout. Strikingly, the two cysteines implicated in disulfide bond formation and ligand binding are absolutely conserved throughout the family (Wojtasek, 1995), suggesting that Takeout also has a ligand binding property (Sarov-Blat, 2000 and references therein).

Phylogenetic analysis using both maximum likelihood and distance matrix methods showed that Takeout is more closely related to hemolymph juvenile hormone binding proteins than to the nuclear JP29. In addition, the Takeout N-terminal 18 amino acids are predicted to be a signal peptide. Therefore, Takeout may be a secretory protein. To address this possibility, Takeout mRNA and protein levels were measured in both bodies and heads. Consistent with the secretion hypothesis, Takeout body mRNA levels are very low despite significant body protein levels (Sarov-Blat, 2000).

date revised: 4 February 2001

Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.

The Interactive Fly resides on the
Society for Developmental Biology's Web server.