bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

1

2 Circadian regulation of microRNA-target chimeras in Drosophila 3 4 Xiju Xia 1*, Xiaonan Fu 2*, Binbin Wu1, Jinsong Zhu 3#, Zhangwu Zhao 1# 5

6 1 Department of Entomology and MOA Key Lab of Pest Monitoring and Green Management, 7 College of Plant Protection, China Agricultural University, Beijing 100193, China. 8 9 2 The Interdisciplinary Ph.D. Program in Genetics, Bioinformatics, and Computational Biology, 10 Virginia Tech, Blacksburg, VA 24061, USA. 11 12 3 Department of Biochemistry, Virginia Tech, 340 West Campus Drive, Blacksburg, VA 24061, 13 USA 14 15 16 17 18 19 * These authors contributed equally to this manuscript. 20 21 22 # Corresponding authors. [email protected]; [email protected] 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

41 Abstract

42 MicroRNA is critical coordinator to circadian regulation by silencing gene

43 expression. Although many circadian related miRNAs and some of its target are

44 known, the global functional miRNA-mRNA interaction networks remain poorly

45 understand which is hindered by imperfect base-pairing between miRNA and target

46 mRNA. In this study, we used CLEAR (Covalent Ligation of Endogenous

47 Argonaute-bound RNAs) -CLIP (Cross-Linking and Immuno-Precipitation) to explore

48 the regulatory functions of miRNAs in the circadian system by comparing the

49 miRNA-mRNA interactions between the Drosophila wild-type strain w1118 and the Clk

50 mutant Clkjrk. We unambiguously identified thousands of miRNA-mRNA interactions

51 from CLEAR-CLIP data set at unprecedented depth in vivo for the first time. Among

52 them, about 300 miRNA-mRNA interactions were involved in the regulation of

53 circadian, in which miRNAs targeting core clock genes pdp1, tim and vri presented

54 distinct changes in response to Clkjrk. Particularly, the mir-375-timeless interaction

55 from CLER-CLIP shows important effects on circadian, this functional event occurred

56 in the l-LNv neurons. Overexpression of mir-375 in tim neurons caused decreases in

57 TIM content resulting in arrhythmicity of daily locomotion and changes of sleep. This

58 present work provides a global view of miRNA targeting in the circadian rhythm.

59

60 Keywords: CLEAR-CLIP, miRNAs, circadian rhythm, sleep, miR-375

61 62 bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

63 Introduction

64 In animals, the intrinsic circadian clock regulates daily rhythms in physiology and

65 behavior, which are entrained by environmental stimuli such as light and

66 temperature1, 2. This robust timing system maintains rhythmic oscillation even under

67 constant darkness conditions in the fruit fly Drosophila melanogaster.

68 Sleep is one of the established circadian behavior, which is highly associated with

69 health status3. Sleep in Drosophila has been well characterized by certain parameters

70 such as total sleep, sleep bout duration and sleep bout number4-6. Throughout the daily

71 sleep-wake cycle, the flies exhibit two peaks of activity - one before lights on and one

72 before lights off. Genetic dissection in Drosophila indicates that sleep is regulated by

73 the circadian clock. The mutations of the core clock genes result in abnormal sleep7, 8,

74 while the neuropeptide pigment dispersing factor (PDF)-expressing peptidergic clock

75 neurons regulate arousal as well as sleep stability9.

76 The Drosophila rhythmic behavior is maintained by a central clock network in

77 the brain, which is consisted of ~150 circadian neurons10. These clock neurons

78 include ventral lateral neurons (LNvs) (I-LNvs, s-LNvs, and 5th s-LNvs), dorsal

79 lateral neurons (LNds), lateral posterior neurons (LPN) and dorsal neurons (DN1,

80 DN2, and DN3), classified based on anatomical locations and varied expression of

81 core clock genes10-13. Among them, the PDF-positive neurons (I-LNvs and s-LNvs)

82 are essential for normal circadian activity and daily sleep. Their ablation results in

83 arrhythmic flies4. bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

84 The circadian clocks are modulated by a complex transcriptional and translational

85 feedback loop (TTFL) in both mammals and flies5-7. In D. melanogaster, two

86 interlocked clock feedback loops are known to regulate the circadian rhythm. They

87 consist of period (per), timeless (tim), clock (clk), cycle (cyc), vrille (vri) and par

88 domain protein 1e (PDP1e). The transcription factors CLK and CYC form a

89 heterodimer that binds to the E-box of per and tim promoters, thereby promoting their

90 transcription8, 9. The cycling expression of tim mRNA with a peak at Zeitgeber time

91 (ZT) 14 and a trough at ZT 2 is vital for circadian output10, 11. Another feedback loop

92 consists of the clk, cyc, vri and PDP1e8, 9, 12-14. However, the mechanism of how these

93 feedback loops sustain ~24 h rhythm is still unclear, since multi-layered regulation

94 involving post-transcriptional and post-translational regulation are imposed to

95 produce the daily rhythm15-17.

96 The highly conserved and widespread microRNAs (miRNAs) play a critical role

97 in post-transcriptional gene regulation18-22. They are loaded into Argonaute (AGO)

98 proteins to mediate mRNA cleavage or translational inhibition via base pairing with

99 target mRNAs20, 23-25. Disruption of the miRNA biogenesis pathway considerably

100 weakens the rhythm in locomotor activity26. Although the rhythmically expressed

101 miRNAs have been identified27-29, only a limited set of miRNA-target pairs have been

102 verified to affect circadian rhythms26, 30-33. Decoding of the miRNA-mRNA

103 interaction network is difficult because of incomplete complementary pairing between

104 miRNAs and targets as well as context-dependent dynamics of the miRNA-target bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

105 interactions. As a result, some miRNAs such as miR-124 and miR-959-964 have been

106 shown to be involved in circadian rhythms but without known targets29, 34, 35. Thus, a

107 global identification of circadian-related miRNA-target interactions will greatly

108 improve our understanding of the role of miRNAs in circadian regulation.

109 The effort to accurately predict the miRNA-targets is compromised by the

110 complex and dynamic cellular interaction networks36, 37. Immunoprecipitation of the

111 miRNA effector protein Argonaute followed by microarray26 or RNA-seq analysis38

112 narrows down the search scope of potential targets but still relies on bioinformatic

113 analysis to predict miRNA-mRNA pairing. The recently improved experimental

114 method of the CLEAR (Covalent Ligation of Endogenous Argonaute-bound RNAs)

115 –CLIP, which isolates miRNA-mRNA chimeras from endogenous AGO-miRNA

116 -mRNA complexes, permits unambiguous identification of miRNA’s targets, which

117 provide a snapshot of true, physiological miRNA-mRNA interactions in vivo24, 39-41.

118 To systematically investigate the role of miRNAs in circadian rhythms, we used

119 the CLEAR-CLIP assay in Drosophila. Tens of thousands of miRNA-mRNA

120 interactions were detected highlighting the well-established miRNAs targeting

121 features. Our data indicate that miRNAs extensively fine-tune the expression of

122 circadian-relevant genes. Importantly, we show here that the regulation of tim by

123 miR-375 plays an important role in locomotor activity and sleep. This present work

124 provides a global view of the miRNA-mRNA interactions in circadian rhythms and a

125 solid foundation for future functional studies. bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

126 Results

127 Identification of miRNA-mRNA interactions by CLEAR-CLIP

128 The brain clock neurons are the regulatory center of circadian rhythm42, in which

129 the transcription factor Clk is essentially required for rhythmic behavior. Ablation of

130 Clk expression in the mutant Clkjrk makes the oscillation of miRNAs disappear in

131 post-transcriptional regulation29. To uncover the physiological miRNA-mRNA

132 interactions contributing to the circadian rhythm, we applied a modified Ago1

133 CLEAR-CLIP to separately analyze the head and the whole body in Clkjrk and w1118

134 flies (Figure S1). The purified Ago1-miRNA-mRNA complexes could be either

135 self-ligated as chimeric reads or remain unchanged mapping to a single locus (CLIP

136 peaks) (Fig. 1A). In total, we obtained 213,404,300 reads from 8 libraries (Table S1A),

137 of which 88.32% could be mapped to mRNAs, and 1.92% from miRNA-target

138 chimeras (Fig. 1B). The recovery of previously experimentally supported

139 miRNA-mRNA interactions in Drosophila showed a reliable dataset (Table S1B),

140 exemplified by the verified regulation of hid by bantam43 (Fig. 1A).

141 In this study, we discovered 15,825 miRNA-target interactions in total involving

142 239 mature miRNAs and 6,540 mRNA transcripts in Drosophila (Table S2A, B).

143 These miRNA- mRNA reads were further classified to miR-first and miR-last

144 chimeras according to the location of miRNA (Table S1C). The miR-first chimeras

145 (92.55%) were predominant compared with the miR-last chimeras (7.45%), and they

146 showed a higher correlation coefficient between miRNA frequency and miRNA bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

147 abundance (r≥0.79), while the correlation coefficient was low (r =~0.45) for miR-last

148 chimeras (Fig S2A, B). Thus, we focused exclusively on the miR-first chimeras in

149 subsequent data analysis.

150 The target sequences from the miR-first chimeras were enriched with canonical

151 seed which matched within 50 nt of the ligation sites (Fig. 1C). In addition, the mean

152 of predicted free energy between miRNAs and the matched target mRNAs found in

153 chimeras was lower by 2.8 kcal mol-1 than that in randomly matched pairs (Fig. 1D,

154 p<0.001). The strong binding energies of chimeric reads suggest that stable chimera

155 stems come from genuine mRNA-miRNA pairing in miRISC rather than from

156 proximity-induced ligation of non-specific RNAs in the solution. These data indicated

157 that the CLEAR-CLIP chimeras reveal a reliable and high-resolution miRNA-target

158 interaction map.

159 The miRNA-target chimeras preserve functional interaction

160 To further evaluate the physiological functions of miRNA-target interactions, we

161 analyzed the binding motif enrichment of let-7-5p44, miR-34-5p45, miR-305-5p46 and

162 miR-276a47, which have been investigated in circadian regulation. The enriched 7-mer

163 motifs in the targets of these miRNA were reverse-complementary to the miRNA seed

164 regions, indicating the crucial role of the seed sequences in miRNA targeting (see Fig.

165 2A and Fig. S2C). To further examine the evolutionary conservation of

166 chimera-defined target sites, the PhyloP conservation scores of 3’UTR targeting sites

167 were retrieved from multiple alignments of 27 insect genomes. We found that the bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

168 conservation score of chimera-defined targets displayed a marked increase at 6-nt

169 interacted stem relative to flanking regions (Fig. 2B), indicating the close

170 physiological relevance.

171 In addition to assessing the functional outcome of the chimera-identified

172 miRNA-target by CLEAR-CLIP, we also analyzed how the altered RNA levels of

173 miR-305 and miR-34 affected the expression of their target mRNAs based on

174 previously published data related to rhythmicity45, 46. In the CLEAR-CLIP dataset, the

175 miR-305 bound to 598 target sites within 555 mRNA transcripts. Overexpression of

176 miR-305 in Drosophila considerably lowers the abundance of mRNAs bearing

177 chimera-defined miR-305 target sites, compared with the mRNAs that had no

178 miR-305 target site (p=0.005618). And the subset of these mRNAs supported by both

179 chimeras and Ago1 CLIP peaks was also significantly repressed (p=0.0191) (Fig.2C).

180 Functional enrichment analysis of miR-305 targets also demonstrated their important

181 roles in sleep (Fig. 2D), which has been observed in the miR-305 mutant strain46.

182 Furthermore, the mRNA levels of the miR-34 targets were significantly elevated in the

183 miR-34 null mutant (miR-34-5p), compared with w1118 (p<0.001) (Fig. 2E). In this

184 study, GO enrichment analysis of the chimera-defined miR-34 targets linked miR-34

185 to aging and brain disease (Fig.2F), consistent with the previous report that miR-34 is

186 associated with aging and neurodegeneration45. These results verified the

187 physiological significance of the miRNA-mRNA interactions identified by

188 CLEAR-CLIP. bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

189 Clk-oriented miRNA-mRNA interactions are critical for circadian rhythm

190 To explore the whole miRNA-mRNA interactions network in the circadian

191 system, we visualized 9325 unique miRNA-target chimeras supported by

192 corresponding Ago1 CLIP peaks with SOM clustering (Fig. S3A). The

193 miRNA-mRNA interactions were significantly changed in Clkjrk (Fig.3A), implying

194 widespread circadian impact by CLK. Furthermore, by quantitatively identifying the

195 affected miRNA-mRNA interactome (fold change>2), we found that 53.0% of the

196 circadian relevant regulations were altered, respectively (Fig. 3A). For example, the

197 interaction between miR-276 and tim only appeared in the wildtype strain w1118. But

198 the miR-34 -pdp1 pairing only happened within the mutant strain Clkjrk. These

199 phenomena implicate that the miRNAs involved in circadian regulation are

200 dynamically adapting to the systemic changes.

201 To understand the functional impact of changed miRNA-mRNA interactions, we

202 performed a functional enrichment analysis separately using the significantly

203 up/down-regulated genes. Results showed that many circadian-related functional

204 groups such as sleep and visual perception were top-ranked, especially enriched in the

205 fly-head groups, while chemical synaptic transmission is notably affected after Clk

206 mutation (Fig. S3C).

207 Moreover, by grouping the genes involved in circadian phenotype, we found that

208 299 miRNA-mRNA interactions involving 9 core circadian genes were plotted as a

209 miRNA-regulated network (Fig. S3B), in which a majority of interactions appeared to bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

210 target the three Clk-downstream genes pdp1, vri and tim (Fig. 3B). To compare the

211 miRNA-involved regulations in these three genes, we retrieved the interaction pattern

212 under different conditions. Results displayed distinct regulatory features in response

213 to Clk disruption (Fig. 3C). The miRNAs targeting pdp1 acted in a Clk-independent

214 manner. Conversely, the interactions between tim and miRNAs occurred in w1118 but

215 were not detectable in Clkjrk, with exceptions of miR-193-3p and miR-2c-3p. Binding

216 of miRNAs to the vri transcript exhibited a mixed pattern of both Clk-dependence and

217 Clk-independence. The miR-276a/b-3p, miR-14-3p, and miR-210-3p were joint

218 regulators of these three genes (Fig. 3D). Alternatively, miR-375-3p and miR-305-5p

219 only targeted tim in wild type flies (Fig. S4B), in which miR-305-5p had been

220 reported to target tim in several circadian screens48-50. These findings suggest that

221 gene regulation by miRNAs has a profound impact on circadian system stability.

222 Oscillations of Mir-375 expression in pacemaker neurons is affected in Clkjrk

223 Besides the reads mapped to mRNA, the CLEAR-CLIP dataset also included lots

224 of unligated miRNAs, which has been suggested as a reliable indicator of functional

225 miRNA abundance51. To understand the Ago1-bound miRNA changes, we

226 summarized the reads mapped to mature miRNAs. Results showed that only a few

227 miRNAs (miR-375-3p, miR-193-3p, miR-133-3p) related to core circadian genes

228 showed significant changes between w1118 and Clkjrk. Among them, miR-375-3p

229 displayed the largest change, with a 3.28-fold increase detected in the fly head (Fig.

230 4A). Thus, we focused on the miR-375-3p for further study. bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

231 First, the daily expression of miR-375-3p in the fly head was analyzed by

232 qRT-PCR at six time points (ZT0, 4, 8, 12, 16, 20) under LD condition. Results

233 showed that it was rhythmically expressed with the peak at ZT12 in w1118, but this

234 pattern disappeared in Clkjrk (Fig. 4B), implying its dependence on Clk. To investigate

235 the spatial expression of miR-375-3p, we utilized a reporter strain

236 (TI{GAL4}mir-375[KO] ×UAS-Mcd8::GFP) to monitor the expression of mir-375.

237 Results showed that it was expressed in the PDF-expressing I-LNv neurons and

238 co-localized with TIM (Fig.4C).

239 Mir-375 impacts normal circadian rhythm and sleep

240 To further evaluate the role of mir-375 in circadian behavior, we overexpressed

241 mir-375 in tim neurons by using iso:tim-gal4 driver. Results showed that the flies with

242 enhanced expression of mir-375 in tim neurons lost daily locomotor rhythm under

243 both LD and DD conditions. Both morning and evening anticipations disappeared

244 after the overexpression of mir-375 (Table 1, Fig.5A-C), while control flies exhibited

245 normal behavioral rhythms (UAS-LUC-mir-375/+: 95.74% rhythmic, tau=24.5 h,

246 n=47; iso:tim-gal4/+: 100% rhythmic, tau=24.6 h, n=75). In addition, flies with

247 overexpressed mir-375 showed a decrease in sleep bout duration and an increase in

248 sleep bout number at night compared with the control (Fig.5D, E). These results

249 demonstrate that mir-375 is very important in the regulation of the Drosophila

250 circadian system. bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

251 mir-375 influences circadian rhythm and sleep via targeting tim

252 The flies with up-regulation of mir-375 had the same phenotype as the tim null

253 mutants on the regulation of the circadian rhythm (Fig.5C and Fig.S4A), and they

254 both expressed mir-375 in l-LNv neurons. In this study, 11 miRNA-tim

255 chimera-defined target sites were founded across the tim transcript (Fig. S4B). Among

256 them, the miR-375 targeting site was located at its CDS region and the chimeras

257 displayed a stable hybridization stem with lower free energy (Fig. 6A).

258 Furthermore, to address whether miR-375 had an effect on the expression of the

259 core clock gene tim, the mRNA levels of tim were detected daily at six time points

260 (ZT2, 6, 10, 14, 18, 22) in both flies with mir-375 overexpression (OE) in tim neurons

261 and control flies. Results showed that the amounts of tim mRNA were dramatically

262 decreased in mir-375 OE flies at ZT 10, 14 and 18 (Fig. 6B). Similarly, TIM protein

263 levels in I-LNvs neurons, detected at ZT18 and ZT24 by confocal microscopy, also

264 showed significant decreases (31.1% at ZT 18 and 54.37% at ZT24) in mir-375 OE

265 flies compared with control flies (Fig. 6C). This observation was also verified by

266 western blot analysis at ZT18 (Fig. 6D). These results indicate that the TIM in the

267 brain was significantly lower in the mir-375 OE flies than in the control flies.

268 To connect arrhythmic phenotypes caused by overexpressing miR-375 in

269 tim-staining neurons, we simultaneously used the UAS-tim2-5 to rescue this

270 phenotype of overexpressing miR-375, in which the UAS-mCD8-GFP served as a

271 control. The results showed that the UAS-tim2-5 could reduce arrhythmicity to 80% bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

272 compared with the UAS-mCD8-GFP control, confirming that tim is a real target of

273 miR-375 (Fig. 6E). Taken together, all these data demonstrate that the regulation of

274 tim by mir-375 plays an important role in mediating the daily circadian rhythm.

275 Discussion

276 The significance of post-transcriptional regulation in circadian rhythms has been

277 demonstrated by monitoring the core clock gene expression15, 52 and screening

278 miRNAs involved in rhythmic behavior44, 47-49, 53-58. However, elucidation of the

279 underlying molecular mechanism is impeded because only a few of the predicted

280 miRNA-mRNA interactions have been verified. In this study, we captured

281 miRNA-mRNA interaction pairs directly from fly tissue using the recently developed

282 Ago1 CLEAR-CLIP59, 60, which enabled us to detect tens of thousands of

283 miRNA-target interactions. Of these interactions, 75 circadian genes are regulated by

284 61 miRNAs in the network of 299 interactions (see Fig. 1 and Fig. S3B).

285 With the first global miRNA-mRNA interaction profiles in Drosophila, we

286 performed multiple lines of evaluation to assess the quality of the whole dataset. First,

287 CLEAR-CLIP recovers many known miRNA-target interactions such as bantam:hid43

288 and miR-276a:tim47 (Fig. 1). Second, analysis of interactions recovered from chimeras

289 reveals general bioinformatics features for miRNA targeting, including seed

290 enrichment, stable binding energy and evolutionary conservation (Fig.1, Fig.S3).

291 Third, the chimera-defined target mRNAs exhibit significant changes after

292 manipulating the corresponding miRNA (Fig. 2). Fourth, functional enrichment bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

293 analysis of the targets of miR-305 and miR-34 are consistent with the previous

294 phenotypic study of mutant flies45, 46. These together substantiate the reliability of the

295 datasets for miRNA-mRNA interaction, which lay the groundwork for further

296 functional studies of miRNAs.

297 In light of chimera-defined miRNA:target pairs, we were able to globally assess

298 the regulatory roles of miRNAs in the whole clock system consisting of input

299 pathways, endogenous clock pacemaker and output pathways (Fig. 3). It is

300 noteworthy that the disruption of Clk has a broad impact on the miRNA-mRNA

301 interaction profile. The shifted pattern points to considerable changes of Ago1 RISC

302 (RNA-induced silencing complex) assemblies in Clkjrk. One interpretation is that

303 Ago1-miRNA loading is inefficient in Clkjrk flies because the Clk mutation could

304 abolish some of the cyclic miRNA expression53. The miRNA-target interactions may

305 also depend on the availability of miRNA targets, given the central role of Clk in

306 circadian transcription regulation. The drastic changes of circadian mRNAs may alter

307 the Ago loading or retention of the miRNA in RISC61. miRNAs regulating PDP1, VRI

308 and TIM present distinct patterns, in which the miRNAs:tim interaction is

309 Clk-dependent. Despite the different regulatory connection to Clk expression, the

310 patterns may also result from the combinatorial regulation by multiple miRNAs with

311 diverse cyclic expression patterns. The oscillating circadian mRNAs were regulated

312 by multiple miRNAs to ensure correct timing of expression.

313 Our data clearly indicated that the 9 core circadian genes were regulated by 23 bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

314 miRNAs in the brain of fruit flies (Fig. S3). Previous miRNA screens have identified

315 a total of 47 miRNAs as relevant to circadian phenotype 48-50, 12 of those miRNAs

316 were also detected in this study (Fig. S3D). The other miRNAs, which clearly have

317 important roles in circadian rhythm, may indirectly regulate the expression of the

318 circadian genes that have been confirmed to participate in circadian regulation48-50. In

319 addition, we linked 10 miRNAs related to circadian rhythm and mapped 9 interaction

320 sites on the tim transcript. These data suggest a complex regulation of miRNAs on

321 rhythmic output.

322 The current study reports for the first time that miR-375 is involved in circadian

323 rhythm and sleep. At the larval stage of Drosophila, miR-375 is expressed in the

324 salivary glands and hindgut, but detailed expression characterization at adulthood is

325 lacking62. In this study, we confirmed its circadian expression in the I-LNv neurons.

326 Overexpressing miR-375 with the tim driver leads to arrhythmic behavior with loss of

327 morning anticipation and normal sleep pattern. Compared with an 81% arrhythmicity

328 after overexpressing miR-276a in the previous study47, we observed a 100%

329 arrhythmicity. In contrast to the peak expression of miR-276a at ZT22, the highest

330 level of miR-375 appeared at ZT12 with significant repression of TIM in the head at

331 ZT18, as detected by immunoblotting (Fig.6). Multiple miRNAs may participate in

332 the spatiotemporal control of tim expression, coordinating the normal rhythmic

333 output.

334 Furthermore, we also found that overexpression of the mir-305 in tim neurons bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

335 (iso:tim-gal4×UAS-luc-mir-305) caused a unimodal activity and a phase shift in DD

336 condition, while bimodal activity was maintained in an LD condition (Fig. S5). This

337 phenotype was also observed in miR-124 mutant55, 56. Our CLEAR-CLIP data suggest

338 that miR-124 targets shaker (sh) and twenty-four (tyf). The chimera-defined

339 miRNA-target interactions imply a broad impact of miRNAs on many aspects of the

340 circadian system and provide a reliable guide to study the function of individual

341 miRNAs. Extensive investigation is required to comprehensively understand the

342 post-transcriptional regulation of circadian rhythms. This study opens a new avenue

343 for a deep understanding of the post-transcriptional regulation of circadian rhythms.

344

345 Acknowledgments

346 We thank Jeffery Price (University of Missouri at Kansas City) for revision of

347 this manuscript. The work was supported by the Natural Science Foundation of China

348 Grants 31572317 and 31730076 (to ZZ) and National Institutes of Health R01

349 AI122743 (to JZ).

350

351 Author contributions: J.Z. and Z.Z. designed research; X.X., B.W. and X.F.

352 performed research; X.X., X.F. analyzed data; and X.X., X.F., J.Z. and Z.Z. wrote the

353 paper.

354

355 Additional information

356 Competing financial interests: The authors declare no competing financial interests. bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

357 Materials and Methods

358 Fly strains

359 Fly strains were maintained on standard molasses-cornmeal-yeast food in a

360 12L:12D cycle at 25℃ and 60% humidity. The fly lines were used in this study as

361 follows: w1118, ClkJrk, USA-LUC-mir-305/TM3, UAS-LUC-mir-275.mir-305,

362 UAS-LUC-mir-375, UAS-LUC-mir-9c, UAS-Mcd8::GFP, TI{GAL4}mir-375[KO],

363 which were purchased from the Bloomington Drosophila Stock Center. The

364 iso:tim-gal4 was obtained from Yirao’s lab, UAS-tim2-5 from Jeffrey L. Price’s lab

365 and was originally generated by Amita Sehgal’s lab..

366 Ago1 CLEAR-CLIP library construction

367 The protocol was adapted from previous reports 24,39. The collected tissue

368 samples were first ground in liquid nitrogen before ultraviolet irradiation. The purified

369 Ago1-RNA complexes from Drosophila lysates were subjected to stringent wash for

370 removal of nonspecific binding. T4 RNA ligase 1 was added into the complex to

371 promote the formation of miRNA-target chimeras. To track the size of the Ago1-RNA

372 complex, RNA was labeled through ligation with a 32P-labeled 3’ DNA linker. A

373 subsequent size selection on an SDS polyacrylamide gel was used to isolate the

374 Ago1-miRNA-mRNA complexes (>130 kDa) and recover authentic target RNAs for

375 library construction and sequencing.

376 Behavior assay and analysis

377 Adult male flies (2-5 d old) were used to test locomotor activity rhythms. Flies

378 were entrained under LD for 3d and released into constant darkness (DD) for at least

379 6 d at 25°C. Locomotor activity was recorded with Drosophila activity monitors

380 (Trikinetics). Sleep was analyzed by pysolo and GraphPad software. FaasX software

381 was used to analyze behavioral data. Locomotor behavior was analyzed in

382 MATLAB. The details for the experimental protocol and data analysis were described

383 by Chen et al44.

384 Confocal microscopy bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

385 Adult male flies of 5-10 days were collected and their brains were dissected in

386 phosphate buffered saline (PBS). The brains were subjected to 4% paraformaldehyde

387 in PBS for 1 h and then washed three times with the wash buffer (0.05% Triton X-100

388 in PBS) for 10 min at room temperature. Samples were transferred to Blocking buffer

389 (2% Triton-100, 10% normal goat serum in PBS) for an overnight incubation at 4 ℃

390 and incubated with primary antibodies (diluted in blocking buffer) overnight at 4℃.

391 The following primary antibodies were diluted in the blocking buffer: rat-anti-TIM

392 (1:1000, from Jeffrey L. Price), mouse-anti-PDF (1:400, from DSHB). After washing

393 samples three times for 15 min at room temperature, the samples were incubated with

394 secondary antibody (1:200) at 4˚C overnight. The samples were imaged on a Leica

395 SP8 confocal microscope. ImageJ software was used for TIM and PDF quantification.

396 RNA extraction and qRT-PCR analysis

397 Fly heads were collected at the indicated time points and stored at -80℃ before

398 RNA extraction. Total RNA was extracted from the heads by using Trizol reagent

399 (TIANGEN) following the supplier’s instruction. For reverse transcription and

400 real-time PCR of timeless, we used PrimeScript RT reagent kit with gDNA Eraser

401 (TAKARA) and superreal premix plus (SYBR Green) (TIANGEN). The sequences of

402 primers are shown in Table S4. For quantitative analysis of timeless, miR-375 and 2s

403 rRNA, we used a miRcute miRNA first-strand cDNA synthesis kit and a miRcute

404 miRNA qRT-PCR detection kit (SYBR Green) (TIANGEN). The miRNA-specific

405 forward primers used for qRT-PCR are also shown in Table S2.

406 Western blot

407 Flies’ heads were collected at the indicated time points and homogenized in a 1.5

408 mL microtube with RIPA lysis buffer (strong) supplemented with proteinase inhibitors.

409 For immunoblot analysis, proteins were transferred to PVDF membrane. After

410 blocking, the membrane was incubated with actin antibody (1:10000) and TIM

411 antibody (1:4000) for 2 h at room temperature. Image J was used to calculate band

412 intensity. bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

413 Identification of miRNA-mRNA chimeras

414 Ago1 CLEAR-CLIP sequencing reads were preprocessed to remove low-quality

415 reads and 3’ adapter sequences using Flexbar63. Shorter reads (less than 16 nt) were

416 discarded. Reads with identical sequences were collapsed. Three random nucleotides

417 were used in the 5’ adapter to avoid PCR over-amplification. These random barcodes

418 were trimmed prior to mapping.

419 The clean reads containing miRNA sequences were first defined by mapping

420 mature miRNA sequences against sample libraries using BLAST (v2.3.0) with e-value

421 equals to 0.464. Only the best alignment was reported. The remaining sequences

422 excluding miRNAs part were extracted for mapping to the transcripts with BLAST.

423 The reads mapped to rRNA, tRNA and miRNA genes were removed.

424 Peak calling of CLIP cluster

425 The peak calling of CLIP cluster was done as previously described using the

426 pooled reads from three biological samples at each time point65. Briefly, overlapped

427 reads were collapsed to form cluster regions. Cubic spline interpolation (Scipy,

428 http://www.scipy.org/) was performed to determine the location and read number of

429 peaks within a cluster66. Significant peaks were identified by determining reads

430 number cutoffs with the p-value less than 0.01 using Poisson distribution.

431 Motif analysis

432 The analysis of overrepresented motifs of Ago1 CLEAR-CLIP by MEME was

433 performed as previously described59 with the settings: -dna -mod zoops -max w7 -n

434 motifs 1. The motifs identified by meme for targets of each miRNA were then aligned

435 to the reverse-complemented miRNA sequence using FIMO68, with the

436 setting-output-p thresh 0.01. High-confidence motifs were generated with FIMO

437 q-value (FDR) < 0.05, and meme Bonferroni-corrected p-value < 0.05.

438 For de novo motif analysis using Homer, miRNAs with more than 30 unique

439 chimeras-defined sites were kept. The background sequences were randomly selected

440 from other miRNA chimeras. Information score (c) was calculated as previously bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

441 described59 to retain motifs with c≥1. Heatmap with the percentage of the motif

442 presence was created using R heatmap package.

443 RNA duplex structure prediction and binding free energy

444 The predictions of duplex structure for miRNAs and their target sequences were

445 performed by RNA hybrid. Mutations in chimeras sequences for both miRNA and

446 mRNA were omitted by comparing with the original annotation. The target sequences

447 were extended to 100 nt according to the miRNA seed sites distribution in D.

448 melanogaster. The setting of RNA hybrid was modified as previously described to

449 favor the seed pairing59. The minimum free energy of each interaction was calculated

450 by RNA hybrid with default settings67. The shuffled interactions served as controls for

451 the comparison of free energy distribution. The median free energy for each profile

452 was calculated in R for comparison.

453 The duplex heatmap was performed with transformed pairing data (base pairing

454 including G-U match as 1, unpaired site as 0). Cluster numbers (k) 3-10 were tested,

455 with k=5 providing the most meaningful set of distinct categories.

456 Analysis of chimera targets in miRNA perturbation experiments

457 The gene expression data after miRNA perturbation was downloaded45, 46. The

458 log2 fold-change was used to perform cumulative distribution function (CDF)

459 analysis comparing transcripts bearing chimeras-derived miRNA target sites versus

460 those without target sites. The CDFs were also plotted for transcripts in which the

461 chimera-derived miRNA target sites overlap with Ago1 CLIP peaks.

462 Dynamic analysis of miRNA-mRNA interactions

463 To quantitatively analyze the temporal profile of miRNA-mRNA interactions, the

464 chimera-supported Ago1 binding peaks were normalized to the total unique reads in

465 each condition69. In order to visualize the pattern of difference between each time

466 point, the data were loaded into Gene Expression Dynamics Inspector (GEDIv2.1)

467 with default parameter (26x25 grids) to perform the dynamic analysis70. GEDI

468 employs the SOM algorithm to assign the interactions with similar trend into close bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

469 tiles. To select the interaction cluster presenting distinct pattern at a certain time point,

470 we first located the tile with maximum value at local spots displaying unique

471 interactions and expanded the spots following gene Density map in GEDI to the edges.

472 Unique interactions for each category were extracted for phenotype enrichment

473 analysis based on FlyBase annotation.

474 GO and KEGG enrichment analysis

475 The selected gene list was taken as an input for the functional enrichment analysis

476 with DAVID71. The top-ranked GO terms belonging to biological functions were

477 chosen.

478 Statistics analysis

479 Statistics analysis for all indicated data in this study were performed with T-test

480 and p values considered significant at p < 0.05 and extremely significant at p < 0.001.

481 Detailed supplementary information list

482 Fig. S1 Experimental Procedure of Ago1 CLEAR-CLIP.

483 Fig. S2 miRNAs detected in the Ago-RNA complexes and the miRNA-mRNA

484 chimeras.

485 Fig. S3 The miRNA-mRNA interactions involved in Drosophila circadian system.

486 Fig. S4 The miRNAs involved regulation of tim in Drosophila.

487 Fig.S5 Overexpress mir-305/mir-9c in tim-neuron affects phase of circadian

488 locomotor behavior.

489 Table S1A. Read statistics for 8 drosophila CLEAR-CLIP samples; Table S1B.

490 Experimentally Validated miRNA Targets Found in chimeras supported interactions;

491 Table S1C. miRNA-first and miRNA-last distribution found in CLEAR-CLIP.

492 Table S2A Ago1 CLIP peaks.

493 Table S2B Ago1 CLASH chimeras.

494 495 References

496 1. Allada, R. & Chung, B.Y. in Annual Review of Physiology, Vol. 72 605-624 497 (2010). bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

498 2. Johnson, C.H., Elliott, J.A. & Foster, R. Entrainment of circadian programs. 499 Chronobiology International 20, 741-774 (2003). 500 3. Buysse, D.J. Sleep health: can we define it? Does it matter? Sleep 37, 9-17 501 (2014). 502 4. Renn, S.C.P., Park, J.H., Rosbash, M., Hall, J.C. & Taghert, P.H. A pdf 503 neuropeptide gene mutation and ablation of PDF neurons each cause severe 504 abnormalities of behavioral circadian rhythms in Drosophila. Cell 99, 791-802 505 (1999). 506 5. Allada, R. & Chung, B.Y. Circadian organization of behavior and physiology 507 in Drosophila. Annu Rev Physiol 72, 605-624 (2010). 508 6. Zheng, X. & Sehgal, A. Probing the relative importance of molecular 509 oscillations in the circadian clock. Genetics 178, 1147-1155 (2008). 510 7. Rosbash, M., Bradley, S., Kadener, S., Li, Y., Luo, W., Menet, J.S., Nagoshi, 511 E., Palm, K., Schoer, R., Shang, Y. & Tango, C.H.A. Transcriptional feedback 512 and definition of the circadian pacemaker in Drosophila and animals. Cold 513 Spring Harbor Symposia on Quantitative Biology 72, 75-83 (2007). 514 8. Allada, R., White, N.E., So, W.V., Hall, J.C. & Rosbash, M. A mutant 515 Drosophila homolog of mammalian Clock disrupts circadian rhythms and 516 transcription of period and timeless. Cell 93, 791-804 (1998). 517 9. Rutila, J.E., Suri, V., Le, M., So, W.V., Rosbash, M. & Hall, J.C. CYCLE is a 518 second bHLH-PAS clock protein essential for circadian rhythmicity and 519 transcription of Drosophila period and timeless. Cell 93, 805-814 (1998). 520 10. Edery, I., Zwiebel, L.J., Dembinska, M.E. & Rosbash, M. TEMPORAL 521 PHOSPHORYLATION OF THE DROSOPHILA PERIOD PROTEIN. 522 Proceedings of the National Academy of Sciences of the United States of 523 America 91, 2260-2264 (1994). 524 11. HunterEnsor, M., Ousley, A. & Sehgal, A. Regulation of the Drosophila 525 protein timeless suggests a mechanism for resetting the circadian clock by 526 light. Cell 84, 677-685 (1996). 527 12. Kadener, S., Stoleru, D., McDonald, M., Nawathean, P. & Rosbash, M. 528 Clockwork Orange is a transcriptional repressor and a new Drosophila 529 circadian pacemaker component. Genes & Development 21, 1675-1686 530 (2007). 531 13. Cyran, S.A., Buchsbaum, A.M., Reddy, K.L., Lin, M.C., Glossop, N.R.J., 532 Hardin, P.E., Young, M.W., Storti, R.V. & Blau, J. vrille, Pdp1, and dClock 533 form a second feedback loop in the Drosophila circadian clock. Cell 112, 534 329-341 (2003). 535 14. Glossop, N.R.J., Houl, J.H., Zheng, H., Ng, F.S., Dudek, S.M. & Hardin, P.E. 536 VRILLE feeds back to control circadian transcription of Clock in the 537 Drosophila circadian oscillator. Neuron 37, 249-261 (2003). 538 15. Lim, C. & Allada, R. Emerging roles for post-transcriptional regulation in 539 circadian clocks. Nat Neurosci 16, 1544-1550 (2013). bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

540 16. Reischl, S. & Kramer, A. Kinases and phosphatases in the mammalian 541 circadian clock. FEBS Lett 585, 1393-1399 (2011). 542 17. Bae, K. & Edery, I. Regulating a circadian clock's period, phase and amplitude 543 by phosphorylation: insights from Drosophila. J Biochem 140, 609-617 544 (2006). 545 18. Bushati, N. & Cohen, S.M. microRNA functions. Annu. Rev. Cell Dev. Biol. 546 23, 175-205 (2007). 547 19. He, L. & Hannon, G.J. MicroRNAs: small RNAs with a big role in gene 548 regulation. Nature reviews. Genetics 5, 631 (2004). 549 20. Fabian, M.R., Sonenberg, N. & Filipowicz, W. Regulation of mRNA 550 translation and stability by microRNAs. Annual review of biochemistry 79, 551 351-379 (2010). 552 21. Hausser, J. & Zavolan, M. Identification and consequences of miRNA-target 553 interactions--beyond repression of gene expression. Nature Reviews. Genetics 554 15, 599 (2014). 555 22. Chou, C.-H., Chang, N.-W., Shrestha, S., Hsu, S.-D., Lin, Y.-L., Lee, W.-H., 556 Yang, C.-D., Hong, H.-C., Wei, T.-Y. & Tu, S.-J. miRTarBase 2016: updates to 557 the experimentally validated miRNA-target interactions database. Nucleic 558 acids research 44, D239-D247 (2015). 559 23. Bartel, D.P. MicroRNAs: target recognition and regulatory functions. cell 136, 560 215-233 (2009). 561 24. Moore, M.J., Scheel, T.K., Luna, J.M., Park, C.Y., Fak, J.J., Nishiuchi, E., 562 Rice, C.M. & Darnell, R.B. miRNA-target chimeras reveal miRNA 3 563 [prime]-end pairing as a major determinant of Argonaute target specificity. 564 Nature communications 6 (2015). 565 25. Gorski, S.A., Vogel, J. & Doudna, J.A. RNA-based recognition and targeting: 566 sowing the seeds of specificity. Nature Reviews Molecular Cell Biology 18, 567 215-228 (2017). 568 26. Kadener, S., Menet, J.S., Sugino, K., Horwich, M.D., Weissbein, U., 569 Nawathean, P., Vagin, V.V., Zamore, P.D., Nelson, S.B. & Rosbash, M. A role 570 for microRNAs in the Drosophila circadian clock. Genes & development 23, 571 2179-2191 (2009). 572 27. Yang, M., Lee, J.-E., Padgett, R.W. & Edery, I. Circadian regulation of a 573 limited set of conserved microRNAs in Drosophila. BMC genomics 9, 83 574 (2008). 575 28. Chen, X. & Rosbash, M. MicroRNA-92a is a circadian modulator of neuronal 576 excitability in Drosophila. Nature Communications 8 (2017). 577 29. Vodala, S., Pescatore, S., Rodriguez, J., Buescher, M., Chen, Y.-W., Weng, R., 578 Cohen, S.M. & Rosbash, M. The oscillating miRNA 959-964 cluster impacts 579 Drosophila feeding time and other circadian outputs. Cell metabolism 16, 580 601-612 (2012). 581 30. Chen, W., Liu, Z., Li, T., Zhang, R., Xue, Y., Zhong, Y., Bai, W., Zhou, D. & bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

582 Zhao, Z. Regulation of Drosophila circadian rhythms by miRNA let-7 is 583 mediated by a regulatory cycle. Nature communications 5, 5549 (2014). 584 31. Chen, X. & Rosbash, M. mir-276a strengthens Drosophila circadian rhythms 585 by regulating timeless expression. Proceedings of the National Academy of 586 Sciences 113, E2965-E2972 (2016). 587 32. Luo, W. & Sehgal, A. Regulation of circadian behavioral output via a 588 MicroRNA-JAK/STAT circuit. Cell 148, 765-779 (2012). 589 33. Xue, Y. & Zhang, Y. Emerging roles for microRNA in the regulation of 590 Drosophila circadian clock. Bmc Neuroscience 19 (2018). 591 34. Garaulet, D.L., Sun, K., Li, W., Wen, J., Panzarino, A.M., O'Neil, J.L., 592 Hiesinger, P.R., Young, M.W. & Lai, E.C. miR-124 Regulates Diverse Aspects 593 of Rhythmic Behavior in Drosophila. Journal of Neuroscience 36, 3414-3421 594 (2016). 595 35. Zhang, Y., Lamba, P., Guo, P. & Emery, P. miR-124 regulates the phase of 596 Drosophila circadian locomotor behavior. Journal of Neuroscience 36, 597 2007-2013 (2016). 598 36. Schnall-Levin, M., Zhao, Y., Perrimon, N. & Berger, B. Conserved microRNA 599 targeting in Drosophila is as widespread in coding regions as in 3'UTRs. Proc 600 Natl Acad Sci U S A 107, 15751-15756 (2010). 601 37. Fan, X. & Kurgan, L. Comprehensive overview and assessment of 602 computational prediction of microRNA targets in animals. Brief Bioinform 16, 603 780-794 (2015). 604 38. Chi, S.W., Zang, J.B., Mele, A. & Darnell, R.B. Argonaute HITS-CLIP 605 decodes microRNA-mRNA interaction maps. Nature 460, 479-486 (2009). 606 39. Helwak, A. & Tollervey, D. Mapping the miRNA interactome by cross-linking 607 ligation and sequencing of hybrids (CLASH). Nature protocols 9, 711-728 608 (2014). 609 40. Scheel, T.K., Moore, M.J., Luna, J.M., Nishiuchi, E., Fak, J., Darnell, R.B. & 610 Rice, C.M. Global mapping of miRNA-target interactions in cattle (Bos 611 taurus). Scientific Reports 7 (2017). 612 41. Grosswendt, S., Filipchyk, A., Manzano, M., Klironomos, F., Schilling, M., 613 Herzog, M., Gottwein, E. & Rajewsky, N. Unambiguous identification of 614 miRNA: target site interactions by different types of ligation reactions. 615 Molecular cell 54, 1042-1054 (2014). 616 42. Renn, S.C., Park, J.H., Rosbash, M., Hall, J.C. & Taghert, P.H. A pdf 617 neuropeptide gene mutation and ablation of PDF neurons each cause severe 618 abnormalities of behavioral circadian rhythms in Drosophila. Cell 99, 791-802 619 (1999). 620 43. Brennecke, J., Hipfner, D.R., Stark, A., Russell, R.B. & Cohen, S.M. bantam 621 encodes a developmentally regulated microRNA that controls cell proliferation 622 and regulates the proapoptotic gene hid in Drosophila. Cell 113, 25-36 (2003). 623 44. Chen, W., Liu, Z., Li, T., Zhang, R., Xue, Y., Zhong, Y., Bai, W., Zhou, D. & bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

624 Zhao, Z. Regulation of Drosophila circadian rhythms by miRNA let-7 is 625 mediated by a regulatory cycle. Nat Commun 5, 5549 (2014). 626 45. Liu, N., Landreh, M., Cao, K., Abe, M., Hendriks, G.J., Kennerdell, J.R., Zhu, 627 Y., Wang, L.S. & Bonini, N.M. The microRNA miR-34 modulates ageing and 628 neurodegeneration in Drosophila. Nature 482, 519-523 (2012). 629 46. Ueda, M., Sato, T., Ohkawa, Y. & Inoue, Y.H. Identification of miR-305, a 630 microRNA that promotes aging, and its target mRNAs in Drosophila. Genes 631 Cells 23, 80-93 (2018). 632 47. Chen, X. & Rosbash, M. mir-276a strengthens Drosophila circadian rhythms 633 by regulating timeless expression. Proc Natl Acad Sci U S A 113, E2965-2972 634 (2016). 635 48. You, S., Fulga, T.A., Van Vactor, D. & Jackson, F.R. Regulation of Circadian 636 Behavior by Astroglial MicroRNAs in Drosophila. Genetics 208, 1195-1207 637 (2018). 638 49. Goodwin, P.R., Meng, A., Moore, J., Hobin, M., Fulga, T.A., Van Vactor, D. & 639 Griffith, L.C. MicroRNAs Regulate Sleep and Sleep Homeostasis in 640 Drosophila. Cell Rep 23, 3776-3786 (2018). 641 50. Kadener, S., Menet, J.S., Sugino, K., Horwich, M.D., Weissbein, U., 642 Nawathean, P., Vagin, V.V., Zamore, P.D., Nelson, S.B. & Rosbash, M. A role 643 for microRNAs in the Drosophila circadian clock. Genes Dev 23, 2179-2191 644 (2009). 645 51. Flores, O., Kennedy, E.M., Skalsky, R.L. & Cullen, B.R. Differential RISC 646 association of endogenous human microRNAs predicts their inhibitory 647 potential. Nucleic acids research 42, 4629-4639 (2014). 648 52. Lerner, I., Bartok, O., Wolfson, V., Menet, J.S., Weissbein, U., Afik, S., 649 Haimovich, D., Gafni, C., Friedman, N., Rosbash, M. & Kadener, S. Clk 650 post-transcriptional control denoises circadian transcription both temporally 651 and spatially. Nat Commun 6, 7056 (2015). 652 53. Vodala, S., Pescatore, S., Rodriguez, J., Buescher, M., Chen, Y.W., Weng, R., 653 Cohen, S.M. & Rosbash, M. The oscillating miRNA 959-964 cluster impacts 654 Drosophila feeding time and other circadian outputs. Cell Metab 16, 601-612 655 (2012). 656 54. Cusumano, P., Biscontin, A., Sandrelli, F., Mazzotta, G.M., Tregnago, C., De 657 Pitta, C. & Costa, R. Modulation of miR-210 alters phasing of circadian 658 locomotor activity and impairs projections of PDF clock neurons in 659 Drosophila melanogaster. PLoS Genet 14, e1007500 (2018). 660 55. Zhang, Y., Lamba, P., Guo, P. & Emery, P. miR-124 Regulates the Phase of 661 Drosophila Circadian Locomotor Behavior. J Neurosci 36, 2007-2013 (2016). 662 56. Garaulet, D.L., Sun, K., Li, W., Wen, J., Panzarino, A.M., O'Neil, J.L., 663 Hiesinger, P.R., Young, M.W. & Lai, E.C. miR-124 Regulates Diverse Aspects 664 of Rhythmic Behavior in Drosophila. J Neurosci 36, 3414-3421 (2016). 665 57. Chen, X. & Rosbash, M. MicroRNA-92a is a circadian modulator of neuronal bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

666 excitability in Drosophila. Nat Commun 8, 14707 (2017). 667 58. Xue, Y. & Zhang, Y. Emerging roles for microRNA in the regulation of 668 Drosophila circadian clock. BMC Neurosci 19, 1 (2018). 669 59. Moore, M.J., Scheel, T.K., Luna, J.M., Park, C.Y., Fak, J.J., Nishiuchi, E., 670 Rice, C.M. & Darnell, R.B. miRNA-target chimeras reveal miRNA 3'-end 671 pairing as a major determinant of Argonaute target specificity. Nat Commun 6, 672 8864 (2015). 673 60. Helwak, A. & Tollervey, D. Mapping the miRNA interactome by cross-linking 674 ligation and sequencing of hybrids (CLASH). Nat Protoc 9, 711-728 (2014). 675 61. Flores, O., Kennedy, E.M., Skalsky, R.L. & Cullen, B.R. Differential RISC 676 association of endogenous human microRNAs predicts their inhibitory 677 potential. Nucleic Acids Res 42, 4629-4639 (2014). 678 62. Poy, M.N., Hausser, J., Trajkovski, M., Braun, M., Collins, S., Rorsman, P., 679 Zavolan, M. & Stoffel, M. miR-375 maintains normal pancreatic alpha- and 680 beta-cell mass. Proc Natl Acad Sci U S A 106, 5813-5818 (2009). 681 63. Dodt, M., Roehr, J.T., Ahmed, R. & Dieterich, C. FLEXBAR-Flexible 682 Barcode and Adapter Processing for Next-Generation Sequencing Platforms. 683 Biology (Basel) 1, 895-905 (2012). 684 64. Altschul, S.F., Gish, W., Miller, W., Myers, E.W. & Lipman, D.J. Basic local 685 alignment search tool. J Mol Biol 215, 403-410 (1990). 686 65. Zisoulis, D.G., Lovci, M.T., Wilbert, M.L., Hutt, K.R., Liang, T.Y., Pasquinelli, 687 A.E. & Yeo, G.W. Comprehensive discovery of endogenous Argonaute binding 688 sites in Caenorhabditis elegans. Nat Struct Mol Biol 17, 173-179 (2010). 689 66. Haecker, I., Gay, L.A., Yang, Y., Hu, J., Morse, A.M., McIntyre, L.M. & 690 Renne, R. Ago HITS-CLIP expands understanding of Kaposi's 691 sarcoma-associated herpesvirus miRNA function in primary effusion 692 lymphomas. PLoS Pathog 8, e1002884 (2012). 693 67. Kruger, J. & Rehmsmeier, M. RNAhybrid: microRNA target prediction easy, 694 fast and flexible. Nucleic Acids Res 34, W451-454 (2006). 695 68. Grant, C.E., Bailey, T.L. & Noble, W.S. FIMO: scanning for occurrences of a 696 given motif. Bioinformatics 27, 1017-1018 (2011). 697 69. Hafner, A., Stewart-Ornstein, J., Purvis, J.E., Forrester, W.C., Bulyk, M.L. & 698 Lahav, G. p53 pulses lead to distinct patterns of gene expression albeit similar 699 DNA-binding dynamics. Nat Struct Mol Biol 24, 840-847 (2017). 700 70. Eichler, G.S., Huang, S. & Ingber, D.E. Gene Expression Dynamics Inspector 701 (GEDI): for integrative analysis of expression profiles. Bioinformatics 19, 702 2321-2322 (2003). 703 71. Huang da, W., Sherman, B.T. & Lempicki, R.A. Systematic and integrative 704 analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc 705 4, 44-57 (2009). 706 bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

A W1118 clkjrk B miRNA-target chimeras Ago1 CLIP reads Dark Light Dark Dark Light Dark 5’UTR CDS 3’UTR Other noncoding RNAs 1.0100% 100% activity

0.880% 80% tRNA tRNA

snRNA snRNA n 0.6 60%

o 60% snoRNA

ti snoRNA c Locomotor rRNA rRNA 0.440% 40% Fra 3UTR 3UTR

CDS CDS Ago1 CLEAR-CLIP 0.2 20% 20% 5UTR 5UTR 3’OH miRNA 0.0 5’P mRNA 0% 0% W0 WH C0 CH Ago Complex Whole Head Whole Head WW0ho le HWHea d WC0ho le HeCHad body body body body Ligation W1118 clkjrk 1118 jrk 3’OH W clk 3’OH miRNA miRNA 5’P mRNA 5’P mRNA Ago Complex Ago Complex C 00.02.020 miRNA-target chimeras Ago1 CLIP reads 8mer 7mer-m8 ty i s 0.0150.015 7mer-A1

chimera1: TGAGATCATTTTGAAAGCTGATTATTGCTAATTAGTTTTCACAATGATCTCGGTAAAGTTTTGTGGCCT Den 6mer chimera2: TGAGATCATTTTGAAAGCTGATA------GTTTTCACAATGATCTCGGTAAAGTTTTGTGGCC ty

i 6mer-off l 00.01.010 i dme-bantam TTAGTCGAAAGTTTTACTAGAGT-5’ dG = -31.0 kcal/mol b a

|||||||||||| |||||||||| b hid-3’UTR ATTGCTAATTAGTTTTCACAATGATCTCGGTAAAGTTTTGTGGCC-3’ 0.0050.005 Pro

00 50 W1118 -1-10000 --5500 0 50 100 Distance from ligation point (nt)

on Whole ti 0 body bu i 50 1118 tr W D 0.10 s chimeras di Head ty

i 0.08 s ds 0 Shuffled jrk

ea 50 clk interactions r Den

0.06 Whole ty IP i l body i 0 b 0.04 50 jrk a b e CL clk qu Head Pro 0.02

Uni 0 hid 3’UTR 0.00 -35 -30 -25 -20 -15 -10 -5 0 Free Energy (kcal/mol)

Fig. 1 Ago1 CLEAR-CLIP decodes miRNA-mRNA interactions in Drosophila. (A) Schematics of CLEAR-CLIP to detect miRNA-mRNA interactions in Drosophila. Wild-type W1118 and Clk mutant clkjrk fiies with arrhythmic phenotype were collected. The UV-crosslinked sample was homogenized in lysis buffer. Ago1-RNA complex was immunoprecipitated and incubated with T4 RNA ligase to favor the direct ligation of miRNA with its target mRNA fragments. The recovered RNAs were subjected to library construction and sequencing. With the bioinformatics analysis, the chimeric reads could be identified along with the Ago1 binding reads, exemplified by the known bantam targeting hid. (B) The fraction of different RNA classes annotated to Drosophila for miRNA-target chimeras and Ago CLIP reads. (C) Density plot of canonical miRNA seed matches in target regions relative to ligation site for all the chimeras. (D) The median predicted binding energy between miRNA and matching target mRNA found in chimeras was stronger by over 2.8 kcal mol−1 than in randomly matched pairs. bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

A 2 B NNNNNNNNNNNNNNNNNN-5’ miRNA

s t

i 1 b N=25/166 ORF……..NNNNNNNNNNNNNNNNNNNN……Poly(A)

C e-val=2.7e-009 )

0 T

1 2 3 4 5 6 7 TA CCMEME (noT SSC) 01.02.18C 13:10 UGAUAUGUUGGAUGAUGGAGU-5’ dme-let-7-5p 2.0 6nt from the 2 paired duplex

Phylop

s t

i 1 b N=152/633 1.9 Flanking region A A T C e-val=9.9e-038

0 G G

1 2 3 4 5 6 7 G T C AMEME (noA SSC) 01.02.18 13:25 GUCUCGUGGACUACUUCAUGUUA-5’ dme-miR-305-5p 1.8

2

s t i 1 1.7 b N=488/1620 e-val=4.3e-179 CA TG AC

0 T

1 2 3 4 5 6 7 CTGMEME (noC SSC) 01.02.18C 13:39 GUGUUGGUCGAUUGGUGUGACGGU-5’ dme-miR-34-5p 1.6 Evolutionary Conservation Conservation ( Evolutionary N=641/1045 1.5 e-val=3.6e-673 -100 -50 0 50 100 UCUCGUGCCAUACUUCAAGGAU-5’ dme-miR-276a-3p Distance from 5’end of stem in encoded mRNA (nt)

C D GO: biological process Enrichment analysis of miR-305 miR-305 Overexpressed vs Control regulationregulation ofof growthgrowth 1.0 Non-miR-305 hemopoiesishemopoiesis miR-305 chim. sleepsleep 0.8 p = 0.005618 miR-305 chim. + cellcel lmorphogenesis morphogenesis peaks myoblast fusion 0.6 p=0.0191 myoblast fusion regulationregulation of of cellcell shapeshape CDF 0.4 compoundcompound eye eye development development glycolyticglycolytic processprocess 0.2 rhabdomererhabdomere development development imaginalimagin adiscl dis-cderived-derive dwing wing morphogenesis morphogenesis 0.0 0 1 2 3 4 5 6 7 -1.0 -0.5 0 0.5 1.0 0 2 4 6 8 Enrichment fold (-log p.val) E Fold change (log2) F 10 GO: biological process Enrichment analysis of miR-34 miR-34-5p mutant vs Control 1.0 oocyte microtubule cytoskeletonoocyte mpolarizationicrotubule Non-miR-34-5p miR-34-5p chim. pprotonroton t rtransportansport 0.8 p = 9.96e-08 pproteinrotein p phosphorylationhosphorylation miR-34-5p chim. + 0.6 peaks oolfactorylfactory l elearningarning p=1.167e-09 rregulationegulation ooff c cellell s shapehape

CDF 0.4 lolocomotorcomotor r hrhythmythm salivary gland cellsa liautophagicvary gland ccellell d deatheath 0.2 ttricarboxylicricarboxylic aacidcid c cycleycle cell adhesion involved in heart morphogenesis 0.0 heart morphogenesis imaginal disc-derived wwinging mmorphogenesisorphogenesis -1.0 -0.5 0 0.5 1.0 0 2 4 6 8 10 12 Fold change (log2) 0 1 2 3 4 5 6 Enrichment fold (-log10p.val)

Fig.2 miRNA-target chimeras preserve physiological relevant interactions (A)Enriched miRNA binding motifs by chimeras-defined target sites. n, number of motifs found/total number of targets analyzed. E-val, e-value of the motif returned by MEME. Most motifs are complementary to the miRNA seed (bold). (B) Average conservation score of perfect seed matches from miRNA:target interactions located in 3’UTR defined by chimeras. The Phylop conservation score for 27 insect species was downloaded from UCSC.

(C) Cumulative distribution function (CDF) plot shows log2-transformed expression differences for mRNAs identified in miR-305 chimeras comparing the other mRNAs which is not defined as miR-305 targets. The analysis was performed using the transcriptome measurement in miR-305 overexpressed fly versus the wild type control. P-values was calculated from Kolmogorov–Smirnov testing. (E) Cumulative distribution function (CDF) plot shows the change in transcript level after knocking out miR-34-5p. The similar analysis as (A) presents the mRNA expression distribution detected in miR-34-5p chimeras, miR-34-5p chimeras supported by Ago1 peaks and the other mRNAs. (D,F) GO enrichment analysis of chimeras-defined miRNA targets. The top ten biological process terms were plotted. bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

A Clkjrk vs W1118 Whole Body jrk 1118 Fly Head ) Clk vs W ) k k jr 10 Increased interactions jr 10

Clk Decreased interactions Clk 2 Circadian related interactions 2

log 8 8 log (

(

ty i ty i s s

n 6 6 n e e d

d

n n

o 4 4 o ti ti c c

ra miR-34:pdp1 2 ra 2 te te

n miR-34:pdp1 miR-8:vri i n

i

miR-8:vri d d e

0 e 0 z i z l i l a a

rm -2 miR-276:tim -2 miR-375:tim rm

o miR-276:tim

o miR-375:tim n n - - Z Z -2 0 2 4 6 8 10 -2 0 2 4 6 8 10 Z-normalized interaction density (log W1118) 1118 2 Z-normalized interaction density (log2W )

bantam-3p miR-14-3p B P miR-14-3p miR-210-3p miR-276a-3p PDP1 PER miR-317-3p miR-276b-3p miR-277-3p P miR-34-5p miR-2a-3p miR-2b-3p TIM miR-4968-3p miR-316-5p miR-317-3p miR-92b-3p miR-34-5p miR-4968-3p clock V/P box miR-4968-3p miR-9c-5p Degradation

CLK PER TIM miR-133-3p VRI miR-1-3p CYC TIM miR-14-3p miR-14-3p miR-190-5p miR-210-3p miR-307b-5p light miR-2491-3p miR-274-5p miR-276a-3p miR-210-3p miR-1-3p miR-2a-3p miR-276b-3p miR-4968-3p miR-14-3p miR-305-5p miR-34-5p miR-8-3p miR-193-3p miR-34-5p miR-4968-3p miR-210-3p miR-375-3p miR-8-3p miR-276a-3p miR-8-3p CWO miR-276b-3p

C W1118 clkjrk W1118 clkjrk W1118 clkjrk miRNAs D targeting PDP1 PDP1 VRI TIM miR-316-5p miR-277-3p miR-276a-3p miR-276a-3p miR-375-3p miR-317-3p bantam-3p miR-276b-3p miR-276b-3p miR-210-3p miR-316-5p miR-8-3p miR-8-3p miR-2b-3p miR-2a-3p miR-277-3p miR-14-3p miR-276a-3p miR-14-3p miR-4968-3p miR-276b-3p miR-34-5p miR-317-3p miR-1-3p miR-305-5p miR-276a/b-3p bantam-3p miR-14-3p miR-133-3p miR-14-3p miR-34-5p miR-274-5p miR-34-5p miR-375-3p miR-2b-3p miR-210-3p miR-1-3p miR-210-3p miR-4968-3p miR-305-5p miR-2a-3p miR-276b-3p miR-193-3p miR-133-3p miR-193-3p miR-8-3p miR-210-3p miR-190-5p miR-2c-3p miR-274-5p miR-2c-3p miR-1-3p 10 10 miR-190-5p 0 0 10 0 miRNAs miRNAs Normalized peak Normalized peak Normalized peak targeting TIM targeting VRI Height Height Height

Fig. 3 miRNA-mRNA interactions regulates circadian rhythm in Drosophila (A) Scatter plot shows the change of miRNA-mRNA interaction comparing Clkjrk with W1118. The interactions with 2 fold changes are considered as significance. (B) The miRNAs participated regulation in the core network of circadian system. All the miRNA-mRNA interactions were supported by Ago1 peaks. (C) The heatmap shows the change of miRNA-mRNA interactions for PDP1, TIM and VRI at each condition. (D) Venn diagram shows the subsets of miRNAs contribute to the regulation of three core circadian genes. bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

A Ago1-bound miRNAs C TI{Gal4]mir-375KO::mGFP Significant changed miRNAs regulate miRNAs core circadian genes GFP I-LNvs I-LNvs

4 miR-375-3p

PDF 2 I-LNvs I-LNvs (Log2 fold (Log2 change)

miR-193-3p head 0 Fly

118 PDF/GFP W miR-133-3p vs vs I-LNvs

jrk -2 I-LNvs Clk

-2 -1 0 1 2 3 Clkjrk vs W118 Fly body (log2 fold change)

B

5 * W1118 4 clkjrk * 3 level level in heads * 375 - 2 * * miR 1 elative elative

R 0 0 4 8 12 16 20 Zeitgeber Time

Fig. 4, mir-375 oscillations in pacemaker neurons. (A) The changes of Ago1-bound miRNAs abundance between Clkjrk and W1118 comparing whole fly and head. (B) The levels of miR-375 are altered in Clkjrk mutant flies’ heads. (C) w*; TI{GAL4}mir-375KO expression in Drosophila brain. w*; TI{GAL4}mir-375KO flies were crossed with UAS-mGFP files. bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

A AR% R% D 30 Iso:tim-gal4/+ 20 UAS-LUC-mir-375/+ 10 Iso:tim-gal4/UAS-LUC-mir-375 0 sleep for 30 sleep for 30 min 0% 20% 40% 60% 80% 100% Zeitgeber (H) Iso:tim-Gal4/UAS- Iso:tim-Gal4/+ UAS-LUC- mir-375/+ LUC-mir-375 B Iso:tim-gal4/+ UAS-LUC-mir-375/+ Iso:tim-gal4/UAS- Tau=24.6±0.09 Tau=24.5±0.08 LUC-mir-375 E Total sleep sleep bout duration *** 800 800 100 *** 100 *** 80 80 600 *** 600

*** 24) 24) 12) - 12) - - - *** 60 60 Days 400 400 ns

ZT(1 40 40 ZT(1 ZT(12 ZT(12 ns 200 200 20 20 0 0 0 0

C sleep number sleep latency Iso:tim-gal4/+ UAS-LUC-mir-375/+ Iso:tim-gal4/UAS- LUC-mir-375 *** 30 30 100 * 150 150 18.4±1.37 16.6±1.04 16.7±1.12 *** *** n=46 n=41 *** ***

n=60 24) 24)

- ***

20 20 12) -

12) 100 100 - - ** 50 ZT(1 ZT(12 ZT(1 10 10 50 50 ZT(12 Activity

0 0 0 0 0 0630 1830 0630 1830 0630 1830

Fig.5 Overexpression of mir-375 regulate circadian rhythm in Drosophila (A) Overexpression of mir-375 levels causes increased arrhythmicity in flies. AR% represents the percentage of arrhythmic flies, R% represents the percentage of rhythmic flies. (B) Comparison of the circadian locomotor activity in light/dark cycles of controls and mir-375 OE flies. Histograms represent the distribution of activity through 24 h, averaged for flies over three LD days, morning and evening anticipation are lost in mir-375 OE flies. (C) Locomotor behavior under LD cycle and constant darkness. Representative double plotted actograms of iso:tim-gal4/+, UAS-LUC-mir-375/+, iso:tim-gal4/UAS-LUC-mir-375 flies. White indicates the light phase, gray indicates the dark phase. (D) Sleep analysis of male flies with overexpression of mir-375 driven by iso-tim-gal4. The colors of gray dark, light dark and blue represent iso:tim-gal4/+, UAS-LUC-mir-375 and iso:tim-gal4/UAS-LUC-mir-375 sleep traces. (E) Sleep profiles: total sleep, sleep bout duration, sleep number and sleep latency of male flies entrained under three LD cycles are shown, mir-375 levels are upregulated in tim-staining neurons. Data represent mean±SEM (n = 16–32). **P<0.01, ***P< 0.001 determined by Student’s t test. bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

A C ZT18 TIM PDF Merge dme-miR-375 vs timeless ZT18 0.8 AUUCGGUUUGCUUGUUU-5’ |||| ||| |||||| 0.6 UUAUGUGCAGAGCAAGAAGAACAAUGAUA-3’ Control 0.4 **

15 TIM/PDF 0.2 W1118

body OE 375 0.0 - 0 W1118 mir Head ZT24 TIM PDF Merge TIM ZT24 clkjrk 0.8 along body 0.6 Control clkjrk 0.4 *** TIM/PDF Unique CLIP reads distributionCLIP readsUnique Head 0.2 TIM

375 OE 375 0.0 mRNA CDS - 3,500,700 3,500,900 2L: 3,500,500 mir

B D control mir-375 OE 15 1.0 * 0.8 * ZT18 Control mir-375 OE 0.6 10 **

levels in brain levels * TIM 0.4 TIM/ACTIN 5 0.2

timeless ACTIN 0.0

0

Relative 0 4 8 12 16 20 24 Zeitgeber (hours)

E AR% R%

Iso:tim-gal4/UAS-LUC-mir-375 Iso:tim-gal4/UAS-LUC-mir-375;UAS-mGFP Iso:tim-gal4/UAS-LUC-mir-375;UAS-Tim2-5

0% 20% 40% 60% 80% 100%

Fig.6 Tim is a target of mir-375. (A) The Clk-dependent interaction of miR-375 targeting timeless in its coding region. The duplex defined in chimeras was predicted by RNAhybrid. (B) Levels of timeless mRNA are altered in miR-375 OE flies. The relative expression levels were normalized to actin levels and were further normalized to control at ZT2. (error bar indicate SEM, *P < 0.05; two-tailed t test. ) (C) Immunostaining of control (iso-tim-gal4/+; UAS-LUC-mir-375/+) and mir-375 OE (iso-tim-gal4/UAS-LUC-mir-375) flies against TIM (green) and PDF (red) at ZT18 (upper) and ZT24 (lower). Tim signal is normalized to the PDF signal. (n=10-17 hemisphere; error bar indicate SEM, **P < 0.01; ****P < 0.0001; two-tailed t test. ) (D) Western blot against TIM in control and mir-375 OE flies ZT 17, ACTIN is used as loading control, TIM levels are normalize to ACTIN levels. Protein signals are quantitated by using image J. (n=4, error bar indicate SEM, **P < 0.01; two-tailed t test. ) (E) Overexpression of timeless levels recues arrhythmic behavior in mir-375OE flies. AR% represents the percentage of arrhythmic flies, R% represents the percentage of rhythmic flies. bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

A B Input IP

UV 254 nm irradiation AGO1

IgG Grind in the nitrogen Antibody-ago1 - 1. tissue lysis, DNase 2.AGO1 IP Antibody-IgG - 3. High stringency washes 4. RNase treatment RNase ++ + 3’OH 3’OH (KDa) 3’P 5’OH C miRNA 225 3’OH 3’P 3’OH 3’OH 150 AGO1: miRNA: mRNA 5’OH 5’OH mRNA AGO1: miRNA T4 PNK treatment Self ligation 102 1. Dephosphrylation (CIP) 76 2. 3’linker ligation 52

3’OH 38

3’OH 3’OH 31 5’OH

AGO1 D Control 5’P NN AGO1 binding sites CLEAR-CLIP 5’P NN miRNA 5’P NN Chimeric reads 307 RNA extraction and purification 5’linker ligation 240 Chimeric 1. Reverse transcription 190 cDNAs 2. Library construction Enriched 150 miRNAs linker dimer Sequencing 123

Fig. S1 Experimental Procedure of Ago1 CLEAR-CLIP. (A) In Ago1 CLEAR-CLIP, the tissue lysates were prepared from UV cross-linked flies. Endogenous Ago1 was immunopurified and stringently washed to remove non-specific interactions. The RNA ends in the Ago1-complexes were treated with T4 Polynucleotide Kinase and ligated together. RNAs associated with Ago1 were then further purified by size selection using the SDS-PAGE gel and recovered for library construction and sequencing (material and method). (B) Western blot detect Ago1 in input sample, IP with antibody-Ago1 and antibody-IgG. (C) Autoradiogram of the Ago1 associated RNA after digestion with different RNase. (D) PCR products amplified using the recovered RNAs after linker ligation. bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

miR-first miR-last A 0.5 8mer 0.5 8mer 7mer-m8 7mer-m8 7mer-A1 7mer-A1 0.4 6mer 0.4 6mer 6mer-offset 6mer-offset Total Total 0.3 0.3

0.2 0.2

CDF: seed seed presence CDF: 0.1 0.1 CDF: seed seed presence CDF:

0.0 0.0 -200 -100 -50 0 50 100 200 -200 -100 -50 0 50 100 200 Distance from ligation point (nt) Distance from ligation point (nt) B miR-first miR-last 10 W0 r=0.84 10 W0 r= 0.45 WH r=0.88 C0 r=0.86 WH r=0.47 8 CH r=0.79 8 C0 r=0.41 CH r=0.44

6 6

4 4 log2(counts) log2(counts) 2 Chimera abundance Chimera

Chimera abundance Chimera 2

0 0 0 5 10 15 20 0 5 10 15 20 miRNAs abundance miRNAs abundance log2(counts) log2(counts) miRNA position C 1 seed 23

miR-4968-3p miR-14-3p miR-317-3p miR-276a-3p miR-275-3p miR-276b-3p miR-34-5p miR-305-5p miR-1-3p miR-9a-5p miR-7-5p miR-375-3p miR-316-5p miR-2c-3p miR-12-5p miR-13a-5p miR-13b-3p miR-308-3p miR-252-5p miR-989-3p miR-281-3p miR-306-5p miR-210-3p Bantam-3p miR-2b-3p miR-277-3p miR-8-3p miR-11-3p Fraction of miR-9c-5p miR-965-3p motif miR-33-5p miR-314-3p presence miR-2a-3p miR-283-5p 1.0 miR-282-5p miR-278-3p miR-274-5p miR-263a-5p miR-2491-3p 0.5 miR-190-5p miR-184-3p miR-124-3p let-7-5p miR-1000-5p miR-304-5p 0.0 miR-996-3p

Fig. S2 miRNAs detected in the Ago-RNA complexes and the miRNA-mRNA chimeras. (A) The cumulative distribution function of canonical miRNA seed matches in target regions relative to ligation site for miR-first and miR-last chimeras. (B) Correlation plots of miRNA abundance from the whole CLEAR-CLIP libraries and their chimeras. Pearson’s correlation coefficients are shown for each condition. (C) The heatmap shows the distribution of significant enriched 7mer motifs through de novo analyzing all chimeras detected for individual miRNA. The colour intensity is proportional to abundance of 7mer motif in target sequences. The hierarchical clustering was performed in R package gplots2. bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

A W118 clkjrk B Whole Body Whole Fly Fly Head

C GO: biological process Whole Body GO: biological process Fly Head wing morphogenesis chemical synaptic transmission axon guidance long-term memory sarcomere organization neurotransmitter secretion sleep sleep visual perception glucose metabolic process regulation of cell shape dorsal closure cell adhesion negative regulation of translation border follicle cell migration cytoplasmic translation myofibril assembly nervous system development long-term memory wing morphogenesis glucose metabolic process compound eye development septate junction assembly peripheral nervous system development actin filament organization sperm individualization phototransduction locomotor rhythm cell morphogenesis asymmetric neuroblast division cytoplasmic translation determination of adult lifespan locomotor rhythm inter-male aggressive behavior muscle attachment actin filament organization chemical synaptic transmission cell autophagic cell death dorsal closure synaptic vesicle exocytosis

regulation of cell shape wing morphogenesis myofibril assembly axon guidance wing morphogenesis sarcomere organization chaeta development sleep cell morphogenesis visual perception actin cytoskeleton organization regulation of cell shape muscle attachment cell adhesion cell autophagic cell death border follicle cell migration cytoplasmic translation myofibril assembly visual perception long-term memory protein phosphorylation glucose metabolic process actin filament organization septate junction assembly nervous system development actin filament organization phagocytosis phototransduction glucose metabolic process cell morphogenesis ovarian follicle cell development cytoplasmic translation muscle cell cellular homeostasis locomotor rhythm axon guidance muscle attachment oogenesis chemical synaptic transmission epithelial cell apical/basal polarity dorsal closure

0 4 8 12 16 0 4 8 12 16 Enrichment fold (-log10p.val) Enrichment fold (-log10p.val)

D Astroglial miRNAs Sleep relevant related with circadian miRNAs screen behavior (Patricia et. al 2018) (Samantha et. al 2018)

tim cells expressed 16 11 miRNAs screen CLEAR-CLIP identified (Kadener et. Al 2009) 4 miR-210 miRNAs related with 0 miR-274 miR-1 circadian regulation miR-317 miR-133 miR-193 miR-2491 0 miR-276a/b miR-2a 3 miR-307b miR-316 miR-375 miR-305 miR-190 miR-4968 1 miR-2b miR-9c miR-92b

Bantam

miR-8,miR-14 miR-34,miR-277 bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Fig. S3 The miRNA-mRNA interactions involved in Drosophila circadian system. (A) Globally distinct miRNA-mRNA interactome of wild type W1118 and Clkjrk mutant in whole body or head displayed as self-organized maps with the GEDIv2.1. Color bar indicates the normalized peak height. (B) The miRNAs:circadian-genes interacting networks defined chimeras from Drosphila CLEAR-CLIP. The circadian genes were extracted from Flybase with phenotype emphasized in circadian defective behavior. These genes are aligned in circle, color in light red represents core genes in circadian system, color in light green stands for other circadian related genes. miRNAs in red have published results supporting circadian related functions or expression. (C) GO enrichment analysis of chimeras-defined miRNA targets. The top twenty biological process terms were plotted. (D) The Venn diagram of circadian-relevant miRNAs screen.

A tim01 100 15.6±1.24 n=32

50

0 0630 1830 B

Fig. S4 The miRNAs involved regulation of tim in Drosophila. (A) Locomotor profiles of tim01 in LD condition. (B) The distribution of Ago1 peaks along the timeless transcript. miR-276a has been confirmed to target timelss. bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

LD Activity B DD Activity A C Iso:tim-gal4/+ UAS-LUC-mir-9c/+ Iso:tim-gal4/UAS-LUC-mir-9c 100 18.4±1.37 30.6±2.35 11.1±1.15 subjective subjective n=32 n=32 n=32 morning peak evening peak Iso:tim-Gal4/+ Iso:tim-Gal4/+ 50

22.7±1.53 27.2±2.37 bimodal Activity 100 n=16 60 n=14 activity

0 40 0630 1830 0630 1830 0630 1830

Activity 50 20 Tau=24.4h±0.06 Tau=24.5h±0.06 D Iso:tim-gal4/+ UAS-LUC-mir-9c/+ Iso:tim-gal4/UAS-LUC-mir-9c Tau=24.6iso-tim-gal4/+±0.09 TUAS-LUC-miau=23.4±r0.07-9c/+ iso-tim-gal4/UAS-LUC-mir-9c 0 0 0630 1830 0630 1830 UAS-LUC-mir-305/+ UAS-LUC-mir-305/+ 80 30.6±1.92 43.0±4.49 100 n=13 n=16

Days Days 40 Activity 50

0 0630 1830 0 0630 1830

Iso:tim-Gal4/UAS-LUC-mir-305 Iso:tim-Gal4/UAS-LUC-mir-305 25.6 1.25 6028.6±3.73 phase shift n=16 n=15 80 unimodal activity Iso:timiso-tim-gal4/+-gal4/+ 40 E

UAS-LUC-mir-9c/+ Activity UAS-LUC-mir-9c/+ 40 20 iso-tim-gal4/UAS-LUC-mir-9c Iso:tim-gal4/UAS-LUC-mir-9c 0 0 0630 1830 0630 1830

Fig.S5 Overexpress mir-305/mir-9c in tim-neuron affects phase of circadian locomotor behavior. (A) Average activity profiles across three light/dark cycles, mir-305 OE flies (iso:tim-gal4/UAS-LUC-mir- 305) exhibit normal bimodal profiles just like controls (iso:tim-gal4/+; UAS-LUC-mir-305/+). (B) Average activity profiles during the first two DD cycles follow transfer to DD. Controls exhibit a bimodal DD activity profile with peaks around subjective morning and evening. Overexpression of mir- 305 induces a unimodal profile in which evening peak activity is advanced. (C) Comparison of the circadian locomotor activity in light/dark cycles of controls (iso:tim-gal4/+; UAS- LUC-mir-9c/+) and mir-9c OE (iso-tim-gal4/UAS-LUC-mir-9c) flies. Histograms represent the distribution of activity through 24 h, averaged for flies over three LD days, morning and evening anticipation are lost in mir-9c OE flies (dark raw). (D) Locomotor behavior under LD cycle and constant darkness. Representative double plotted actograms of iso:tim-gal4/+, UAS-LUC-mir-9c/+, iso:tim-gal4/UAS-LUC-mir-9c flies. White indicates the light phase, gray indicates the dark phase. (E) Activity analysis of flies with overexpression of mir-9c driven by iso-tim-gal4 in 3 LD and 7 DD.

Table 1 Locomotor activities of flies Genotype Total flies Rhythmic flies (%) Period (h) Power Iso:tim-gal4/+ 75 100 24.6±0.09 56.2±6.2 UAS-LUC-mir-375/+ 47 95.74 24.5±0.08 65.2±7.22 Iso:tim-gal4/UAS-LUC-mir-375 62 0 * * UAS-mir-305/+ 45 93.3 24.3±0.09 97.5±9.62 Iso:tim-gal4/uas-luc-mir-305 47 91.5 23.7±0.13 82.8±9.33 uas-luc-mir-9c/+ 47 100 23.4±0.07 97.9±6.76 Iso:tim-gal4/uas-luc-mir-9c 29 0 * * bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Table S1A. Read statistics for 8 drosophila CLEAR-CLIP samples.

Mapped Chimeric sampleName Raw Reads clean Reads Uniq Reads Reads Reads Percentage W01 22252916 22166118 2607404 1104091 18398 1.67% W02 18229939 18194988 2204986 950071 16954 1.05% W03 20063980 20005202 3417700 1578808 25779 1.02% W04 48304093 47895896 4951811 2946369 68635 1.79% C01 23493584 23438634 2356272 873423 17674 1.02% C02 24206211 24047074 3333769 1296768 25325 1.03% C03 11505949 11487289 1244821 464129 9733 1.05% C04 45347628 45216440 4062840 2094932 51894 1.66%

Table S1B. Experimentally Validated miRNA Targets Found in chimeras supported interactions

miRNA Genes Number of chimeric reads Found in Experiments W1118(whole body), dme-bantam-3p hid 58 W1118(head) dme-let-7-5p ab 12 W1118(head) dme-miR-11-3p E(spl)malpha-BFM 13 W1118(whole body) dme-miR-124-3p gli 20 clkjrk(head) dme-miR-1-3p tutl 14 W1118(head) dme-miR-1-3p CG18542 5 W1118(whole body) dme-miR-1-3p Dl 10 clkjrk(Whole body),clkjrk(head) dme-miR-1-3p Msr-110 8 clkjrk(whole body) dme-miR-1-3p Chd64 13 W1118(head) dme-miR-14-3p EcR 5 W1118(head) dme-miR-14-3p sug 6 W1118(whole body) dme-miR-14-3p IP3K2 4 W1118(whole body) dme-miR-34-5p Su(z)12 2 W1118(whole body) dme-miR-7-5p E(spl)m3-HLH 1 W1118(whole body) dme-miR-7-5p h 21 W1118(head) dme-miR-8-3p wls 2 clkjrk(whole body) dme-miR-8-3p pan 7 clkjrk(whole body)

Table S1C. miRNA-first and miRNA-last distribution found in CLEAR-CLIP

sample W1118 clkjrk

Chimeric Reads head Whole body Head Whole body

miRfirst chimeric 4500 6697 2271 4151

miRlast chimeric 269 590 214 346 bioRxiv preprint doi: https://doi.org/10.1101/622183; this version posted April 29, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.

Table S2. The sequences of primer for real time PCR

tim-yg-F AGTCGCCACTCACCATTCCT tim-yg-R CGTTGTTCTTCTTGCTCTGC actin-F CCAACCGTGAGAAGATGA actin-R GGAGTCCAGAACGATACC 2s TGCTTGGACTACATATGGTTGAGGG miR-375 CGACGTTTGTTCGTTTGGCTTAAGTTA miR-9c GCGATTGTACTTCATCAGGTGCTCTG miR-305 GGCCGTCTTTGGTATTCTAGCTGTAGA