Peripheral Circadian Clock for the Cuticle Deposition Rhythm in Drosophila Melanogaster

Peripheral circadian clock for the cuticle deposition rhythm in Drosophila melanogaster

Chihiro Ito*, Shin G. Goto*, Sakiko Shiga*, Kenji Tomioka†, and Hideharu Numata*‡

*Graduate School of Science, Osaka City University, Osaka 558-8585, Japan; and †Graduate School of Natural Science and Technology, Okayama University, Okayama 700-8530, Japan

Edited by Jeffrey C. Hall, Brandeis University, Waltham, MA, and approved May 1, 2008 (received for review January 7, 2008) Insect endocuticle thickens after adult emergence by daily alter- 18–21). In the compound eyes, CRY is suggested to act as a nating deposition of two chitin layers with different orientation. repressor of the CLK/CYC transcriptional activity (22). Although the cuticle deposition rhythm is known to be controlled Insect exoskeleton, consisting of epi-, exo-, and endocuticle, by a circadian clock in many insects, the site of the driving clock, the thickens by additional deposition of the endocuticle materials photoreceptor for entrainment, and the oscillatory mechanism after adult emergence. They are secreted by epidermal cells in remain elusive. Here, we show that the cuticle deposition rhythm layers in which the chitin microfibrils exist in different orienta- is regulated by a peripheral oscillator in the epidermis in Drosoph- tions, unidirection and helicoid (23). Such cuticle layers in the ila melanogaster. Free-running and entrainment experiments in endocuticle are well known in a wide range of insects and are vitro reveal that the oscillator for the cuticle deposition rhythm is observed as dark-bright bands in the endocuticle under a No- independent of the central clock in the brain driving the locomotor marski differential interference contrast microscope or polar- rhythms. The cuticle deposition rhythm is absent in null and ization microscope (24). It has been reported that the cuticle dominant-negative mutants of clock genes (i.e., period, timeless, layers increase at the rate of one per day, and the rhythm persists cycle, and Clock), indicating that this oscillator is composed of the even in vitro in cockroaches (25–27), suggesting that the rhythm same clock genes as the central clock. Entrainment experiments is governed by a peripheral oscillator. However, the molecular with monochromatic light–dark cycles and cryb flies reveal that a mechanisms for the cuticle deposition rhythm, including blue light-absorbing photoreceptor, cryptochrome (CRY), acts as a oscillation and entrainment, have not been documented. photoreceptor pigment for the entrainment of the cuticle deposi- Here, we clarified in D. melanogaster that the cuticle deposition rhythm. Unlike other peripheral rhythms in D. melanogaster, tion rhythm at the furca, thoracic apodeme attached to inner the cuticle deposition rhythm persisted in cryb and cryOUT mutant muscles, is controlled by a peripheral circadian oscillator local- flies, indicating that CRY does not play a core role in the rhythm ized in the epidermis and the oscillator involves per, tim, cyc, and generation in the epidermal oscillator. Clk. Furthermore, we suggest the involvement of the per- and tim-independent oscillators because per01 and tim01 flies showed circadian rhythm ͉ clock genes ͉ cryptochrome ͉ entrainment ͉ rhythmic cuticle deposition only under light–dark (LD) cycles of epidermal cell 24 h, but not under LD cycles of 21 or 28 h. Photic entrainment in the cuticle deposition rhythm occurred even when isolated thoraxes were cultured in vitro, indicating that the photorecep- ircadian clocks control daily rhythms at the molecular, tion site resides in peripheral tissues. In contrast to the other physiological, and behavioral levels. Like most organisms, C peripheral oscillators, cryb flies showed a clear cuticle deposition the central circadian oscillator controlling the locomotor activity rhythm without photic entrainment. The entrainability was rhythm in Drosophila is proposed to consist of molecular feed- restored by rescue of cry function. From these results, we back loops. The feedback loops comprise several core clock conclude that CRY is not a core component, but an exclusive genes, such as period (per), timeless (tim), Clock (Clk), and cycle photoreceptor in the cuticle deposition rhythm. (cyc) (1). These genes encode PER, TIM, CLK, and CYC proteins, respectively, and loss-of-function or dominant-negative Results mutations in these genes result in loss of normal circadian Regulation of Cuticle Deposition by a Circadian Oscillator. In Dro- rhythmicity (2–5). CLK and CYC form a heterodimer to activate sophila, the cuticle growth layers are observed only in the transcription of per and tim (2, 4, 6). PER and TIM form another phragma, the second and third furcae in the thorax (28). They are heterodimer, are transferred into the nucleus, and then inacti- the apodemata (i.e., internal cuticular processes of the body wall vate the transcriptional activity of the CLK/CYC heterodimer on which the muscles are inserted). We used the third furca for (6). The feedback loops are entrained and reset by the light- observation because it was easiest to be dissected out, and a layer induced degradation of TIM mediated by a photoreceptor in the other apodemata often was difficult to be discerned in flies pigment, cryptochrome (CRY) (7–10). older than 5 days (28). To characterize the cuticle deposition In addition to the central oscillator, peripheral oscillators exist rhythm, we first examined the cuticle deposition under various in many tissues, including the compound eyes, antennae, wings, light and temperature conditions in WT flies. Alternating bright legs, and Malpighian tubules in adults and the ring gland in and dark layers were observed in the endocuticle (Fig. 1A Left), pupae (11–14). These oscillators are self-sustained and directly with the number of bright layers increasing at the rate of about light-entrainable even in vitro (11–13). In addition, the oscillators one per day, with little variations until day 7 under LD cycles with are clearly independent of the central circadian oscillator(s) in the Zeitgeber period (T) ϭ 24 h at 25°C. After day 7, the increase the brain (15, 16). Previous reports revealed that the peripheral oscillator shares the same genes with the central clock in the brain (12, 17). However, there is a remarkable difference that, Author contributions: K.T. and H.N. designed research; C.I. performed research; C.I., S.G.G., in the peripheral oscillator, CRY functions as a core component S.S., and H.N. analyzed data; and C.I. and H.N. wrote the paper. besides a photoreceptor. In cryb (a cry loss-of-function mutant) The authors declare no conflict of interest. and cry0 (a KO mutant of cry) flies, molecular oscillations of the This article is a PNAS Direct Submission. clock genes and proteins are absent in peripheral tissues (i.e., the ‡To whom correspondence should be addressed. E-mail: [email protected]. compound eyes, antennae, legs, and Malpighian tubules) (10, © 2008 by The National Academy of Sciences of the USA

8446–8451 ͉ PNAS ͉ June 17, 2008 ͉ vol. 105 ͉ no. 24 www.pnas.org͞cgi͞doi͞10.1073͞pnas.0800145105 Downloaded by guest on September 23, 2021 significantly different from the hypothesized values (t test, P Ͻ 0.001). In addition, the variance in the number of bright layers was larger than that in 21, 24, and 28 h (Tukey-type multiple comparison for variations, P Ͻ 0.01) (29). These results indicate that the cuticle deposition rhythm is entrained to LD cycles within the range of T ϭ 21–28 h. Hereafter, we considered that the cuticle deposition rhythm is entrained to LD cycles when the numbers of bright layers are changed with the same pattern as that of WT flies under LD cycles with T ϭ 21–28 h (see Fig. 1C). Thus, the cuticle deposition rhythm showed the three major properties of the circadian rhythm: persistence of an overt rhythm under constant conditions, temperature compensation of its free- running period, and stable entrainment to environmental cues with periods in a limited range Ϸ24 h (30). We conclude, therefore, that the rhythm is controlled by a circadian clock. The variance in the number of bright layers was significantly larger under LD cycles with T ϭ 16, 18, and 48 h than under DD (F test, P Ͻ 0.05) (Fig. 1C), although the mean number was not significantly different among LD cycles with T ϭ 16, 18, and 48 h and DD (Aspin–Welch t test, P Ͼ 0.05), suggesting that the cuticle deposition rhythm did not simply free-run in some individuals outside the entrainment range. In addition, some fraction of flies did not show bright and dark layers under LD cycles with T ϭ 16, 18, and 36 h (8.6, 17.3, and 8.3% of individuals, respectively), but all of the individuals showed Fig. 1. Circadian properties of the cuticle deposition rhythm in adult D. alternating bright and dark layers under LD cycles with T ϭ 48 h. melanogaster.(A) Sagittal view of the endocuticle in the third furca on day 6. ϭ These results indicate that the cuticle deposition rhythm in some Alternating layers are observed in WT flies kept under LD cycles of T 24 h ϭ (Left), but not in the per01 mutant kept under DD (Right). Arrows show the individuals is unstable under extreme LD regimes except T seven bright layers. (Scale bars: 10 ␮m.) (B) The daily increase of growth layers 48 h. We suggest that the cuticle deposition rhythm free-runs in the endocuticle of WT flies under LD cycles of T ϭ 24 h (open circles) and DD with frequency demultiplication and relative coordination out- (filled circles). The dotted line shows the hypothetical values assuming that the side the entrainment range, as reported in various circadian bright layer increases at the rate of one per day (n ϭ 30–43). (C) Entrainment rhythms (30, 31). of the cuticle deposition rhythm of WT flies to LD cycles. The number of bright ϭ ϩ layers on day 6 is shown. Hyperbola: y 1 144/x is the simplified equation Peripheral Clock for the Cuticle Deposition Rhythm. To examine when the rhythm completely entrains to the given LD cycles. Alternating whether the cuticle deposition rhythm is independent of the central layers were not visible in the endocuticle of some individuals when LD cycles PHYSIOLOGY with T ϭ 16, 18, and 36 h (3/35, 9/52, and 3/36, respectively). The data of T ϭ clock(s) in the brain, we tested the cuticle deposition rhythm in 24hinC are the same as those on day 6 in B. Asterisks indicate significant isolated whole thoraxes or furcae cultured in vitro from day 1. Seven differences between the mean number of bright layers and the hypothesized bright layers were formed in the endocuticle of most thoraxes and value (**, P Ͻ 0.01; ***, P Ͻ 0.001; t test) (n ϭ 31–43). furcae cultured under DD on day 6 (Fig. 2A), indicating that the rhythm is generated by a clock within the isolated tissues. It is noteworthy that the cuticle deposition rhythm was entrained to LD rate of cuticle layers declined slightly (Fig. 1B). This cuticle cycles with T ϭ 21–28 h in thoraxes cultured in vitro (Fig. 2A Left), deposition rhythm persisted under constant darkness (DD) (Fig. but not in isolated furcae (Fig. 2A Right). 1B). Hereafter, we compared the numbers of layers on day 6 in We then confirmed the presence of a peripheral clock for the flies kept under various conditions because of the decline of the cuticle deposition rhythm in the furcae by immunohistochemis- increase rate after day 7. try using PER antiserum. PER immunoreactivity was found in The number of layers (mean Ϯ SD) on day 6 under DD at 22.5, the nuclei of epidermal cells in the outermost part of an apodeme 25, and 27.5°C was 7.06 Ϯ 0.83 (n ϭ 33), 7.22 Ϯ 0.83 (n ϭ 36), in 26 of 29 specimens at Zeitgeber time (ZT) 23 (Fig. 2B). No and 6.94 Ϯ 0.72 (n ϭ 32), respectively, with no significant immunoreactivity, however, was observed in all nine specimens difference among the three temperatures (ANOVA, P Ͼ 0.05). at ZT 5. Immunoreactivity was observed in 3 of 23 specimens at The Q10 value of the oscillation rate calculated from the means ZT 8 and 5 of 12 specimens at ZT 11. These results suggest that was 1.09 at 22.5–25°C and 0.85 at 25–27.5°C. Thus, the the epidermal cells include the oscillator controlling the cuticle free-running period of the rhythm is temperature-compensated. rhythm. We next examined the entrainability of the cuticle deposition rhythm in WT flies to LD cycles with various periods (T) for 6 Cuticle Deposition Rhythm in Clock Mutant Flies. To examine the days. If the rhythm is entrained to LD cycles, the number of involvement of clock genes in the cuticle deposition rhythm, we layers would change as shown by the hyperbola in Fig. 1C (y ϭ tested the deposition of the endocuticle in per01, tim01, cyc01, ClkJrk, 1 ϩ 144/x) in accordance with the number of LD cycles. When and cryb mutants under LD cycles with T ϭ 21–28 h and DD. In WT T ϭ 21–28 h, all flies showed bright and dark layers in the and cryb flies, bright and dark layers in the endocuticle were endocuticle, and the mean number of bright layers was changed observed in all individuals under all of the conditions examined along the hyperbola with a little variance (Fig. 1C). The number (Fig. 3). In most of the per01 flies, the endocuticle was thickened of layers in T ϭ 21 h was significantly larger than those in T ϭ without distinct layers under DD (Figs. 1A Right and 3). Most of the 24 h and 28 h (Tukey test, P Ͻ 0.01). Although the mean number tim01, cyc01, and ClkJrk flies also showed similar patterns under DD of layers was significantly higher than the hypothesized value in (Fig. 3). Therefore, the circadian clock genes are involved in the T ϭ 28h(t test, P Ͻ 0.01), its variance was small, and, therefore, timing of change in the chitin arrangement, and a loss of the cuticle the rhythm seemed to entrain to the LD cycles in most individ- deposition rhythm is expressed as an absence of alternating layers uals. On the contrary, when T ϭ 16–18 and 36–48 h, the number in the endocuticle. of bright layers did not change along the hyperbola and was The results under LD cycles were different between per01,

Ito et al. PNAS ͉ June 17, 2008 ͉ vol. 105 ͉ no. 24 ͉ 8447 Downloaded by guest on September 23, 2021 Fig. 4. The effect of monochromatic LD cycles on the entrainment of the cuticle deposition rhythm in WT D. melanogaster. The illumination of blue (wavelength ϭ 470 nm), yellow (583 nm), and red (660 nm) was used instead of the white ﬂuorescent light from the second photophase. The number of bright layers on day 6 is shown (hyperbola: see Fig. 1C). Asterisks indicate signiﬁcant differences between the mean number of bright layers and the hypothesized value (**, P Ͻ 0.01; ***, P Ͻ 0.001; t test) (n ϭ 30–32).

different from 24 h. However, most per01 and tim01 flies showed no bright and dark layers under LD cycles with T ϭ 21 and 28 h. Fig. 2. Localization of the circadian oscillator for the cuticle deposition Therefore, the formation of cuticle layers also can be achieved rhythm in WT D. melanogaster.(A) Persistence of the cuticle deposition by a per-ortim-independent circadian clock. This clock depends rhythm under DD and its entrainment to LD cycles in vitro. Isolated thoraxes on cyc and Clk and oscillates only under LD cycles with T close (Left) and furcae (Right) were cultured in vitro from the second photophase to 24 h. ϭ of the LD cycles with T 21–28 h. DD started from the second scotophase of Because CRY is a photoreceptor for entrainment of the LD cycles with T ϭ 24 h. The number of bright layers on day 6 is shown (hyperbola: see Fig. 1C). Alternating layers were not visible in the endocuticle cuticle deposition rhythm (see next section), we examined the 01 b of some individuals (Left, 7/57 in T ϭ 21 h, 7/56 in T ϭ 24 h, 12/58 in T ϭ 28 h, feature of per-independent rhythm by using per ; cry flies. More 01 b and 8/48 in DD; Right, 9/32 in T ϭ 21 h, 10/34 in T ϭ 24 h, 31/68 in T ϭ 28 h, and than 90% (33 of 35) of per ; cry flies tested had no alternating 21/54 in DD). Asterisks indicate significant differences between the mean layers in the endocuticle even when T was 24 h, suggesting that number of bright layers and the hypothesized value (*, P Ͻ 0.05; **, P Ͻ 0.01; per-independent oscillation needs light input via CRY. ***, P Ͻ 0.001; t test) (n ϭ 23–50). (B) Confocal images of a double-stained furca at ZT 23 by PERIOD-immunocytochemistry (green) and nuclear staining Role of CRY in the Cuticle Deposition Rhythm. To search for the (magenta). (Left) Many immunoreactive spots are found in the surface of the furca. (Center) Propidium iodide (PI) was used for a nuclear staining. The photopigment for entrainment of the cuticle deposition rhythm, original red color for nuclear staining by PI was converted to magenta. (Right) we examined the range of effective wavelengths in WT flies using A view under double exposure through green and red channels. The white monochromatic LD cycles with T ϭ 21–28 h. The rhythm was signals indicate double labeling by anti-PERIOD antiserum and PI. Arrowheads found to be entrained to LD cycles of blue (470 nm), but not to are located in the same positions in each photo. (Scale bar: 20 ␮m.) LD cycles of yellow (583 nm) and red (660 nm) (Fig. 4). These results indicate that entrainment of the cuticle deposition rhythm is achieved by a blue light-absorbing photopigment. We assume tim01 flies and cyc01, ClkJrk flies. No alternating layers were shown that CRY is the pigment because CRY is a blue light-absorbing by 50–80% cyc01 and ClkJrk flies in the endocuticle in all test photopigment to reset and entrain the circadian locomotor conditions, indicating that cyc and Clk are crucial components rhythm in D. melanogaster (1). for the cuticle deposition rhythm (Fig. 3). Most of the per01 and To clarify the role of CRY, we examined the cuticle deposition tim01 flies, in contrast, showed alternating layers when T was 24 h rhythm in cryb. The rhythmic deposition of the endocuticle was (Fig. 3). If the formation of alternating layers with T ϭ 24his clearly formed in most individuals, and the number of bright a direct response to light, it should be observed even when T is layers increased at the rate of about one per day under both LD cycles with T ϭ 24 h and DD (Fig. 5A). Although cryb is a loss-of-function mutant of cry, it leaks out a low level of its encoded protein (32). To exclude the possibility that leaking proteins participate in driving the oscillation for the cuticle deposition rhythm in cryb, we tested the cuticle deposition under DD in another cry mutant, cryOUT, in which no CRY protein is detected by immunohistochemistry (33). Bright and dark layers in the endocuticle were observed in all cryOUT flies, and the number of bright layers on day 6 was 7.13 Ϯ 1.34 (mean Ϯ SD, n ϭ 41), suggesting that circadian cuticle deposition rhythm free-ran in cryOUT. The persistence of the rhythm in cryb and cryOUT under DD shows that, unlike other peripheral circadian rhythms (9, 10, 18–21), cry seems not the core component for the generation of the cuticle deposition rhythm. The cuticle deposition rhythm in cryb flies was not entrained Fig. 3. Effects of clock gene mutations on the cuticle deposition rhythm in to LD cycles with T ϭ 21–28 h (Fig. 5B). We rescued cryb by D. melanogaster. Flies were kept under DD or LD cycles with T ϭ 21–28 h. The act-Gal4-mediated cry expression. Control flies having either a ordinate shows the percentage of the individuals with alternating layers in the b 01 01 UAS-cry responder or an act-Gal4 driver in cry background were endocuticle on day 6. The bars with the same letters in each of per and tim not entrained to LD cycles like cryb flies (Fig. 6). In act-Gal4/ were not significantly different (Tukey-type multiple comparisons for propor- b tions; P Ͼ 0.05) (29). There was no significant difference among all four UAS-cry; cry flies, however, the cuticle deposition rhythm is conditions examined in the WT, cryb, cyc01, and ClkJrk (P Ͼ 0.05, ␹2-test) (n ϭ entrained to LD cycles with T ϭ 21–28 h (Fig. 6), and the range 27–46). of entrainment was the same as that in the WT flies (see Fig. 1C).

8448 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0800145105 Ito et al. Downloaded by guest on September 23, 2021 tubules (15, 16), because the ranges of the Zeitgeber period and effective wavelengths for entrainment were different from the locomotor rhythm; both ranges were distinctively narrower than those in the circadian locomotor rhythm in D. melanogaster (38–41). To investigate whether the rhythm is completely independent of the central clock, transplantation experiments between individuals with different phasing or experiments of specific ablation of central pacemaker cells are necessary. Sim- ilar endocuticle deposition rhythm has been reported in cockroaches and locusts: The cuticle deposition rhythm persists Fig. 5. Circadian properties of the cuticle deposition in cryb adults of D. under DD with a temperature-compensated free-running period melanogaster.(A) The daily increase of growth layers of the endocuticle under of Ϸ24 h (27, 42). The rhythm is neither entrained to LD cycles LD cycles of T ϭ 24 h (open circles) and DD (filled circles) (the dotted line: see nor reset by light in cockroaches (27) but coincides with the 1B). Alternating layers were not visible in the endocuticle of some individuals period of LD cycles such as T ϭ 12 and 48 h in locusts (43). In (4/55 on day 4, 3/49 on day 5, 3/46 on day 6, 7/49 on day 7, and 4/49 on day 8 locusts, it is unclear whether the coincidence of the rhythm with under LD cycles; 1/21 on day 4 under DD). The variance in the number of layers on each day was significantly larger in cryb flies than in the WT ones under LD LD cycles is indeed entrainment of the clock or a direct response cycles (F test, P Ͻ 0.05) (see Fig. 1B). There was no significant difference in the to LD cycles. In contrast, the cuticle deposition rhythm in D. variance of the number of layers in cryb of each day between LD cycles and DD melanogaster is shown to be driven by a system consisting of the (F test, P Ͼ 0.05) (n ϭ 20–51). (B) Entrainment of the cuticle deposition rhythm oscillator and photoreceptors, thus suitable for a study on the to LD cycles. The number of bright layers on day 6 is shown (hyperbola: see Fig. entrainment mechanism of a circadian peripheral oscillator. 1C). Asterisks indicate significant differences between the mean number of Ͻ ϭ bright layers and the hypothesized value (*, P 0.05, t test) (n 31–32). Oscillator Mechanism for the Cuticle Deposition Rhythm. The circadian clock for the cuticle deposition rhythm requires the clock genes per, tim, cyc, and Clk because no alternating bright and These results show the involvement of cry in photic entrainment dark layers were formed in per01, tim01, cyc01, and ClkJrk mutant of the cuticle deposition rhythm. flies in DD. This finding indicates that the epidermal oscillator Discussion shares the same clock genes with the central oscillator of locomotor activity rhythm (2–5) and also would be composed of Peripheral Circadian System for Cuticle Deposition Rhythm. The the molecular feedback loops reported for central circadian cuticle deposition rhythm is crucial for producing the correct clock(s) (1). alternating layers with different chitin orientations, which in- Surprisingly, alternating cuticle layers were observed in the creases the physical strength of the cuticle (23). The present cryb mutant flies in all conditions examined and in cryOUT flies results demonstrated that this rhythm is achieved by the circa- under DD, indicating that CRY is not involved in oscillation for dian clock residing in the epidermal cells. PER signals were cuticle deposition rhythm. This result is in contrast to other detected solidly in the nuclei late at night and weakly during the peripheral tissues, including the antennae, legs, photoreceptor PHYSIOLOGY day as reported for the cerebral pacemaker neurons, photore- cells, and Malpighian tubules, where CRY plays a role as a core ceptor cells of the compound eyes, and the Malpighian tubules component of the circadian oscillator and the circadian molec- of D. melanogaster (34–37). The peripheral oscillator controlling ular oscillations are absent in cryb or cry0 flies in DD (9, 10, the cuticle deposition rhythm seems independent of the central 18–21). In the photoreceptor cells of the compound eyes, CRY clock in the brain, such as that in the antennae and Malpighian functions as a transcriptional repressor of the CLK-CYC- activated transcription (22). This finding explains why circadian clocks in peripheral tissues stop in cryb and cry0 mutants. Because the cuticle deposition rhythm sustains in cryb mutants, the molecular mechanism underlying the endocuticular oscillator seems to be different from other peripheral clocks but resembles the cerebral central clocks. Interestingly, cuticle deposition of per01 and tim01 was rhythmic in LD cycles with T ϭ 24 h. The rhythm is not a direct response to light because the rhythm never persisted in DD and LD cycles with T ϭ 21 and 28 h. The result indicates that there is a per- or tim-independent circadian oscillator. On the contrary, cuticle deposition of ClkJrk and cyc01 was arrhythmic in both DD and all LD cycles examined. Therefore, Clk and cyc must be essential components for the per- or tim-independent circadian oscillator. The oscillator seems to have a narrow range of entrainment, close to 24 h, and seems to be quickly damped under DD because per01 and tim01 flies produced no alternating layers under DD. Fig. 6. Restoration of entrainment ability of the cuticle deposition rhythm The photic input mediated by CRY is required for the per- to LD cycles in cryb flies by act-Gal4-mediated cry expression. The number of independent oscillation because the cuticle deposition rhythm bright layers on day 6 is shown (hyperbola: see Fig. 1C). Entrainment to LD was absent in the double mutant, per01; cryb, even under LD cycles was examined in cryb flies containing the UAS-cry responder (Upper cycles with T ϭ 24 h. The involvement of a per- or tim- Left), cryb flies containing act-Gal4 driver (Upper Right), and act-Gal4/UAS-cry; independent circadian oscillation also has been reported in b cry flies (Lower). Alternating layers were not visible in some individuals locomotor rhythms and in the fluorescent dye incorporation (UAS-cry; cryb; 3/40 in T ϭ 21 h, 10/48 in T ϭ 24 h, and 9/50 in T ϭ 28 h, b ϭ ϭ ϭ rhythm of the salivary gland (38, 39, 44, 45). Under LD cycles act-Gal4/CyO; cry ; 3/50 in T 21 h, 2/40 in T 24 h, and 2/41 in T 28 h, 0 act-Gal4/UAS-cry; cryb; 5/39 in T ϭ 18 h, 4/40 in T ϭ 21 h, 8/42 in T ϭ 24 h, 4/35 with various T, per flies show a limited entrainment range, and in 28 h, and 2/39 in T ϭ 36 h). Asterisks indicate significant differences between entrained flies exhibit a characteristic phase relationship of the 0 the mean number of bright layers and the hypothesized value (*, P Ͻ 0.05; **, locomotor activity rhythm to LD cycles, indicating that per still P Ͻ 0.01; ***, P Ͻ 0.001; t test) (n ϭ 31–41). contains a circadian oscillator system (39). Furthermore, analysis

Ito et al. PNAS ͉ June 17, 2008 ͉ vol. 105 ͉ no. 24 ͉ 8449 Downloaded by guest on September 23, 2021 of this secondary oscillator using temperature cycles with dif- organs and proteins. Cleaned thoraxes were stained with 1% KMnO4. Finally, ferent T reveals the existence of the per-ortim-independent the third furcae were dissected and mounted in glycerol on a slide glass. oscillator that contains Clk and cyc as indispensable components Growth layers were observed at 400ϫ and 600ϫ under a Nomarski differential in the central clock (44). The present results also support this interference contrast microscope (BX50-DIG; Olympus). idea in the peripheral clocks controlling the cuticle deposition rhythm. Although little is known about the per-independent Daily Increase of Cuticle Growth Layers. Flies were exposed to LD 12:12 and DD oscillation in peripheral tissues, the secondary oscillator of the (DD started from 22:00 on the day of adult emergence) and were fixed in the cuticle deposition rhythm driven by CRY-dependent light input photophase and subjective day, respectively, on each day. A distinct eclosion line marked the extent of a preeclosion furca at the end and side of a furca. would serve as a model to reveal its mechanism. This line was bright under a differential interference contrast microscope and was termed an ‘‘eclosion layer’’ (28). In the present study, the eclosion layer Entrainment System for Cuticle Deposition Rhythm. The cuticle also was counted as a bright layer. deposition rhythm in the furca was entrained to LD cycles when the whole isolated thorax was cultured in vitro, but not when the Effect of Temperature on Cuticle Growth Layers. Flies reared at 22.5 or 27.5°C furca separated from the other parts of the thorax was cultured. under LD 12:12 from eggs were kept under DD (DD started from 22:00 on the These results suggest that the primary photoreceptor is present day of adult emergence) at the same temperature. Flies were fixed in the in the thorax and that the known external circadian photore- subjective day on day 6. ceptors, such as the compound eyes, ocelli, and Hofbauer– Buchner’s eyelets (9, 10, 46–48), do not play a primary role in Effect of LD Cycles on Cuticle Deposition Rhythm. Flies were exposed to LD photic entrainment. Although the circadian oscillator driving the cycles with T ϭ 16, 18, 21, 24, 28, 36, or 48 h consisting of equal durations of cuticle deposition rhythm is most likely to be present in the the photophase and scotophase. In these regimes, 9, 8, 7, 6, 5, 4, or 3 LD cycles epidermal cells, the precise site of photoreception for entrain- were given to flies during 6 days (144 h). Flies were fixed in the photophase on ment in the thorax remains unclear. One possible explanation is day 6. that the photoreceptor is in the epidermal cells, but the entrainment system may not be operative under our in vitro conditions. Effect of Monochromatic LD Cycles on Entrainment. White fluorescent light was It seems more probable, however, that the photoreceptor is in used in the first photophase, and monochromatic light was applied from the some tissue other than the epidermal cells in the thorax. second photophase. Flies were exposed to monochromatic LD cycles and fixed in the photophase on day 6. The illumination of the monochromatic light was Identification of the extraretinal receptor in the thorax is an ␭ ϭ interesting next issue. In this case, a humoral factor might be produced with blue LED (SLP-0137A-51, max 470 nm; Sanyo Electric), yellow LED (TLOH180P, ␭ ϭ 583 nm; Toshiba), and red LED (SLP-838A-37S1, ␭ ϭ involved in the entrainment. This possibility would be examined max max 660 nm; Sanyo Electric). The light intensity was adjusted to 2.2–2.9 ϫ 1023 by transplantation of the furca to the abdomen of another fly to photons per s⅐cm2 for 470, 583, and 660 nm with a radiometer/photometer see whether the transplanted furca entrains to LD cycles. The (UDP 161; United Detector Technology). entrainment experiments with monochromatic light and using b cry flies revealed that CRY is the essential photoreceptor for Tissue Culture. Flies were exposed to one LD cycle (T ϭ 21–28 h), and thoraxes the entrainment of the cuticle deposition rhythm. The rescue of or furcae were dissected on the second photophase. They were cultured b photic entrainability in cry flies by act-Gal4-mediated cry ex- individually in 150 ␮l of Schneider’s medium (Invitrogen) including 0.1% pression also supports this view. Therefore, our results provide penicillin-streptomycin (Invitrogen) under LD cycles for 5 days. Cultured tho- evidence that CRY is a predominant photoreceptor for the raxes or furcae were fixed in the photophase. In addition, tissue culture was entrainment in the cuticle deposition rhythm. Thus, the circadian performed under DD. In this case, thoraxes or furcae were isolated on the system for the cuticle deposition rhythm is similar to the central second photophase of LD 12:12, cultured in DD (DD started from the second clock in the brain, which differs from the other peripheral scotophase) for 5 days, and fixed in the subjective day on day 6. oscillators. Immunohistochemistry and Nuclear Staining of Furcae. WT flies reared under LD Materials and Methods 12:12 were fixed at ZT 5, 8, 11, and 23 on day 6 by submersion in 4% Fly Strains’ Rearing and Collection. Flies were reared on standard medium paraformaldehyde in phosphate buffer (pH 7.3) with 0.1% Triton X-100. The containing wheat germ at 25°C under 12-h LD cycles (LD 12:12, 10:00–22:00 third furcae were dissected as whole mounts. The primary and secondary light). Newly emerged flies collected within 6 h from light-on were used in all antibodies used were 1:2,000 anti-PER antiserum (Santa Cruz Biotechnology) experiments. Day 0 is defined as 24 h from light-on of the day of adult and 1:200 biotinylated anti-goat secondary antibody (Jackson ImmunoRe- emergence, and the next 24 h is day 1. Canton-S was used as WT. The following search), respectively. Signals were enhanced by tyramide-biotin system mutants and transgenic strains were used: per01, tim01, cyc01, ClkJrk, cryb, cryOUT (PerkinElmer). Visualization was done with streptavidin-Alexa Fluor 488 (Mo- (2–5, 10, 33), per01; cryb, act-Gal4/UAS-cry; cryb. per01; cryb flies were generated lecular Probes). Some immunostained furcae were counterstained by pro- by per01 and cryb ss1 (aka rec6). act-Gal4/UAS-cry; cryb flies were obtained by pidium iodide. Preparations were imaged by laser-scanning confocal micros- crossing act-Gal4/CyO; cryb and UAS-cry; cryb (49). All experiments were per- copy (FLUOVIEW; Olympus). Digital images were processed by using Adobe formed at 25°C, with exceptional cases described separately. The white fluo- Photoshop 6.0 (Adobe Systems) and Corel Draw 11.0 (Corel). rescent light (FLR20 SW/M or FL15W; Matsushita Electric Works) was used at an 2 intensity of 3 W/m for the photophase in all experiments. ACKNOWLEDGMENTS. We thank J. C. Hall (Brandies University, Waltham, MA) for cryb ss1 (aka rec6), A. Matsumoto (Kyushu University, Fukuoka, Japan) Histological Observation of Cuticle Growth Layers. Flies exposed to various for UAS-cry; cryb flies, R. Stanewsky (Queen Mary University of London, United conditions were fixed in a solution of 3:1 ethanol and acetic acid for 24 h. Kingdom) for cryOUT flies, and S. Tamotsu for permitting us to use the radi- Isolated thoraxes were boiled in 4% sodium hydroxide solution to remove the ometer/photometer and helping in measuring the light intensity.

1. Stanewsky R (2002) Clock mechanisms in Drosophila. Cell Tissue Res 309:11–26. 6. Darlington TK, et al. (1998) Closing the circadian loop: CLOCK-induced transcription of 2. Allada R, White NE, So WV, Hall JC, Rosbash M (1998) A mutant Drosophila homolog its own inhibitors per and tim. Science 280:1599–1603. of mammalian Clock disrupts circadian rhythms and transcription of period and 7. Ceriani MF, et al. (1999) Light-dependent sequestration of TIMELESS by CRYPTO- timeless. Cell 93:791–804. CHROME. Science 285:553–556. 3. Konopka RJ, Benzer S (1971) Clock mutants of Drosophila melanogaster. Proc Natl 8. Emery P, So WV, Kaneko M, Hall JC, Rosbash M (1998) CRY, a Drosophila clock and Acad Sci USA 68:2112–2116. light-regulated cryptochrome, is a major contributor to circadian rhythm resetting and 4. Rutila JE, et al. (1998) CYCLE is a second bHLH-PAS clock protein essential for photosensitivity. Cell 95:669–679. circadian rhythmicity and transcription of Drosophila period and timeless. Cell 9. Emery P, et al. (2000) Drosophila CRY is a deep brain circadian photoreceptor. Neuron 93:805–814. 26:493–504. 5. Sehgal A, Price JL, Man B, Young MW (1994) Loss of circadian behavioral rhythms and 10. Stanewsky R, et al. (1998) The cryb mutation identiﬁes cryptochrome as a circadian per RNA oscillations in the Drosophila mutant timeless. Science 263:1603–1606. photoreceptor in Drosophila. Cell 95:681–692.

8450 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0800145105 Ito et al. Downloaded by guest on September 23, 2021 11. Emery IF, Noveral JM, Jamison CF, Siwicki KK (1997) Rhythms of Drosophila period gene 32. Busza A, Emery-Le M, Rosbash M, Emery P (2004) Roles of the two Drosophila CRYP- expression in culture. Proc Natl Acad Sci USA 94:4092–4096. TOCHROME structural domains in circadian photoreception. Science 304:1503–1506. 12. Giebultowicz JM, Hege DM (1997) Circadian clock in Malpighian tubules. Nature 33. Yoshii T, Todo T, Wu¨lbeck C, Stanewsky R, Helfrich-Fo¨rster C (2008) Cryptochrome is 386:664. present in the compound eyes and a subset of Drosophila’s clock neurons. J Comp 13. Plautz JD, Kaneko M, Hall JC, Kay SA (1997) Independent photoreceptive circadian Neurol 508:952–966. clocks throughout Drosophila. Science 278:1632–1635. 34. Curtin KD, Huang ZJ, Rosbash M (1995) Temporally regulated nuclear entry of the 14. Saez L, Young MW (1988) In situ localization of the per clock protein during develop- Drosophila period protein contributes to the circadian clock. Neuron 14:365–372. ment of Drosophila melanogaster. Mol Cell Biol 8:5378–5385. 35. Zerr DM, Hall JC, Rosbash M, Siwicki KK (1990) Circadian fluctuations of period protein 15. Giebultowicz JM, Stanewsky R, Hall JC, Hege DM (2000) Transplanted Drosophila immunoreactivity in the CNS and the visual system of Drosophila. J Neurosci 10:2749– excretory tubules maintain circadian clock cycling out of phase with the host. Curr Biol 2762. 10:107–110. 36. Hege DM, Stanewsky R, Hall JC, Giebultowicz JM (1997) Rhythmic expression of a 16. Tanoue S, Krishnan P, Krishnan B, Dryer SE, Hardin PE (2004) Circadian clocks in PER-reporter in the Malpighian tubules of decapitated Drosophila: Evidence for a antennal neurons are necessary and sufficient for olfaction rhythms in Drosophila. Curr brain-independent circadian clock. J Biol Rhythms 12:300–308. Biol 14:638–649. 37. Liu X, et al. (1992) The period gene encodes a predominantly nuclear protein in adult 17. Krishnan B, Dryer SE, Hardin PE (1999) Circadian rhythms in olfactory responses of Drosophila melanogaster. Nature 400:375–378. Drosophila. J Neurosci 12:2735–2744. 18. Dolezelova E, Dolezel D, Hall JC (2007) Rhythm defects caused by newly engineered 38. Helfrich-Fo¨rster C (2001) The locomotor activity rhythm of Drosophila melanogaster is null mutations in Drosophila’s cryptochrome gene. Genetics 177:329–345. controlled by a dual oscillator system. J Insect Physiol 47:877–887. 19. Ivanchenko M, Stanewsky R, Giebultowicz JM (2001) Circadian photoreception in 39. Helfrich C, Engelmann W (1987) Evidences for circadian rhythmicity in the per0 mutant Drosophila: Functions of cryptochrome in peripheral and central clocks. J Biol Rhythms of Drosophila melanogaster. Z Naturforsch 42:1335–1338. 16:205–215. 40. Ohata K, Nishiyama H, Tsukahara Y (1998) in Biological Clocks: Mechanisms and 20. Krishnan B, et al. (2001) A new role for cryptochrome in a Drosophila circadian Applications, ed Yvan Touitou (Elsevier, Amsterdam), pp 167–170. oscillator. Nature 411:313–317. 41. Wheeler DA, Hamblen-Coyle MJ, Dushay MS, Hall JC (1993) Behavior in light-dark cycles 21. Levine JD, Funes P, Dowse HB, Hall JC (2002) Advanced analysis of a cryptochrome of Drosophila mutants that are arrhythmic, blind, or both. J Biol Rhythms 8:67–94. mutation’s effects on the robustness and phase of molecular cycles in isolated periph- 42. Neville AC (1965) Circadian organization of chitin in some insect skeletons. Q J Microsc eral tissues of Drosophila. BMC Neurosci 3:5. Sci 106:315–325. 22. Collins B, Mazzoni EO, Stanewsky R, Blau J (2006) Drosophila CRYPTOCHROME is a 43. Neville AC (1963) Daily growth zones in insect skeletons. Acta Physiol Scand 59 Suppl circadian transcriptional repressor. Curr Biol 16:441–449. 213:107–108. 23. Neville AC (1975) Biology of the Arthropod Cuticle (Springer, Berlin). 44. Yoshii T, Sakamoto M, Tomioka K (2002) A temperature-dependent timing mechanism 24. Neville AC (1983) Daily cuticular growth layers and the teneral stage in adult insects: is involved in the circadian system that drives locomotor rhythms in the fruit fly A review. J Insect Physiol 29:211–219. Drosophila melanogaster. Zool Sci 19:841–850. 25. Weber F (1985) Postmolt cuticle growth in a cockroach: In vitro deposition of multi- 45. Weitzel G, Rensing L (1981) Evidence for cellular circadian rhythms in isolated fluores- lamellate and circadian-like layered endocuticle. Experientia 41:398–400. cent dye-labelled salivary glands of wild type and an arrhythmic mutant of Drosophila 26. Weber F (1995) Cyclic layer deposition in the cockroach (Blaberus craniifer) endocu- melanogaster. J Comp Physiol B 143:229–235. ticle: A circadian rhythm in leg pieces cultured in vitro. J Insect Physiol 41:153–161. 46. Helfrich-Fo¨rster C, Winter C, Hofbauer A, Hall JC, Stanewsky R (2001) The circadian 27. Wiedenmann G, Lukat R, Weber F (1986) Cyclic layer deposition in the cockroach clock of fruit flies is blind after elimination of all known photoreceptors. Neuron endocuticle: A circadian rhythm? J Insect Physiol 32:1019–1027. 30:249–261. 28. Johnston JS, Ellison JR (1982) Exact age determination in laboratory and field-caught Drosophila. J Insect Physiol 28:773–779. 47. Helfrich-Fo¨rster C, et al. (2002) The extraretinal eyelet of Drosophila: Development, 29. Zar JH (1999) Biostatistical Analysis (Prentice–Hall, Englewood Cliffs, NJ), 4th Ed, pp ultrastructure, and putative circadian function. J Neurosci 22:9255–9266. 228–229, 563–565. 48. Rieger D, Stanewsky R, Helfrich-Fo¨rster C (2003) Cryptochrome, compound eyes, 30. Johnson CH, Elliott J, Forster R, Honma K-I, Kronauer R (2004) in Chronobiology: Hofbauer-Buchner eyelets, and ocelli play different roles in the entrainment and Biological Timekeeping, eds Dunlap JC, Loros JJ, DeCoursey PJ (Sinauer, Sunderland, masking pathway of the locomotor activity rhythm in the fruit fly Drosophila melanogaster. J Biol Rhythms 18:377–391.

MA), pp 67–105. PHYSIOLOGY 31. Aschoff J, Pohl H (1978) Phase relations between a circadian rhythm and its Zeitgeber 49. Ishikawa T, et al. (1999) DCRY is a Drosophila photoreceptor protein implicated in light within the range of entrainment. Naturwissenschaften 65:80–84. entrainment of circadian rhythm. Genes Cells 4:57–65.

Ito et al. PNAS ͉ June 17, 2008 ͉ vol. 105 ͉ no. 24 ͉ 8451 Downloaded by guest on September 23, 2021