(C) PLOS One
This story was originally published by PLOS One and is unaltered.
. . . . . . . . . .



Steroid receptor coactivator TAIMAN is a new modulator of insect circadian clock [1]

['Vlastimil Smykal', 'Biology Centre Of The Academy Of Sciences Of The Czech Republic', 'Institute Of Entomology', 'Ceske Budejovice', 'Czech Republic', 'Lenka Chodakova', 'Faculty Of Science', 'University Of South Bohemia', 'Marketa Hejnikova', 'Kristina Briedikova']

Date: 2023-09

TAIMAN (TAI), the only insect ortholog of mammalian Steroid Receptor Coactivators (SRCs), is a critical modulator of ecdysone and juvenile hormone (JH) signaling pathways, which govern insect development and reproduction. The modulatory effect is mediated by JH-dependent TAI’s heterodimerization with JH receptor Methoprene-tolerant and association with the Ecdysone Receptor complex. Insect hormones regulate insect physiology and development in concert with abiotic cues, such as photo- and thermoperiod. Here we tested the effects of JH and ecdysone signaling on the circadian clock by a combination of microsurgical operations, application of hormones and hormone mimics, and gene knockdowns in the linden bug Pyrrhocoris apterus males. Silencing taiman by each of three non-overlapping double-strand RNA fragments dramatically slowed the free-running period (FRP) to 27–29 hours, contrasting to 24 hours in controls. To further corroborate TAIMAN’s clock modulatory function in the insect circadian clock, we performed taiman knockdown in the cockroach Blattella germanica. Although Blattella and Pyrrhocoris lineages separated ~380 mya, B. germanica taiman silencing slowed the FRP by more than 2 hours, suggesting a conserved TAI clock function in (at least) some insect groups. Interestingly, the pace of the linden bug circadian clock was neither changed by blocking JH and ecdysone synthesis, by application of the hormones or their mimics nor by the knockdown of corresponding hormone receptors. Our results promote TAI as a new circadian clock modulator, a role described for the first time in insects. We speculate that TAI participation in the clock is congruent with the mammalian SRC-2 role in orchestrating metabolism and circadian rhythms, and that TAI/SRCs might be conserved components of the circadian clock in animals.

Most living creatures need to synchronize their rhythmic biological functions with the 24-hour day/night cycle of Earth. Instead of pocket watches or cell phones, they use intracellular molecular clocks to anticipate the daily recurring changes. Those circadian clocks are at the genetic level remarkably conserved among mammals and insects. Here, for the first time in insects, we introduce the protein TAIMAN as a new modulator of the circadian clock. The removal of TAIMAN from the adult linden bug Pyrrhocoris apterus or the German cockroach Blattella germanica males slows the clock by 2–4 hours. Circadian clocks often adjust animals’ physiology through various hormones and neuropeptides. Insect endocrinology is dominated by the steroid and steroid-like hormones, ecdysone, and juvenile hormone: and TAIMAN is closely engaged in both juvenile hormone and ecdysone receptor signaling. However, the circadian role of TAIMAN presented here modulates the clock in a hormone-independent way. Importantly, TAIMAN is related to the mammalian SRC-2 protein, which is involved in regulating circadian clock machinery in mice. Thus, the novel role we describe for TAIMAN in insects may have a counterpart in mammalian physiology.

Funding: This work was supported by European Research Council (ERC) under the European Union’s Horizon 2020 Program Grant Agreement 726049 and by the Czech Science Foundation (GACR) grant number 22-10088S. LC‘s salary was supported by FP7-PEOPLE Program Grant Agreement 316790 (INsecTIME). MV was supported by Ministry of Agriculture of the Czech Republic grant number QK1910286. M.H., B.C.H.W, P.C., and D.D. received salaries from ERC Grant (Agreement 726049). V.S., and H.V. received salaries from GACR (grant number 22-10088S). L.C. received a salary from FP7- PEOPLE Program Grant Agreement 316790 (INsecTIME). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Copyright: © 2023 Smykal et al. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Although there are examples of the interaction of circadian proteins with JH and ecdysone signaling pathways, direct involvement of core JH and ecdysone receptors’ proteins in the central circadian clock remains elusive. To clarify this important gap, we used a combination of microsurgical and reverse genetic tools available to the linden bug, and addressed the role of JH and ecdysone signaling in a system where the developmental role can be clearly separated from the function in the fully formed adult organism. We identified steroid receptor coactivator taiman as a new insect clock gene where TAIMAN modulates the pace of the circadian clock in the linden bug P. apterus and German cockroach B. germanica and tai silencing dramatically slowed down clock free-running period in both species by 2–4 hours, while sustaining the clock rhythmic. Interestingly, the effect on the clock seems to be ecdysone- and JH-independent.

Clock crosstalk with JH signaling is more unclear. Shin et al. [ 37 ] described the physical interaction of MET, CYC, and TAI (called FISC) in a female mosquito Aedes aegypti fat body. In their experiments, the presence of CYC in the JH receptor complex facilitated the circadian expression of JH target genes in the fat body peripheral clock. Surgical removal of corpora allata, a JH-producing gland, reciprocally changed the transcription of per and circadian clock target gene Pdp1 in P. apterus fat body [ 38 ]. Further experiments demonstrated that JH acts through MET (but not TAI) together with CLK and CYC in the P. apterus gut to maintain the linden bug reproductive state by controlling reciprocal circadian expression of Pdp1 iso1 and m-cry genes [ 39 ].

The potential crosstalk of the circadian clock and the humoral regulation by JH and ecdysone is reported at different regulatory levels but remains largely elusive. The release of ecdysone in some species display circadian timing [ 26 – 29 ] and Kumar et al. showed EcR expression in Drosophila clock neurons and established ecdysone-induced nuclear receptor (NR) Ecdysone-induced protein 75B (E75/Eip75), an insect homolog of mammalian repressor REV-ERB, as a direct repressor of Clk expression [ 30 ]. In mammals, REV-ERB is opposed by mammalian activator Retinoic Acid Receptor-Related Orphan Receptor α (RORα) [ 31 , 32 ] to keep circadian rhythms. E75B and NR HORMONE RECEPTOR 3 (HR3), an insect homolog of RORα, influence the cycling of Clk, cyc, and d-tim in the ametabolous insect Thermobia domestica [ 33 ]. Interestingly, not HR3 but NR HORMONE RECEPTOR 51 (HR51, also known as UNFULFILLED) was recruited in Drosophila to co-regulate per transcription [ 34 ], and together with E75B constitutes a TTFL. Further, the ecdysone-inducible ABC transmembrane transporter encoded by Early gene at 23 (E23) facilitates lipophilic ecdysone transmembrane transport from the cell, and thus represses ecdysone signaling [ 35 , 36 ]. E23 transcription is also regulated by CLK/CYC complexes, and E23 knockdown increases vri expression [ 36 ].

Insects’ physiological processes at the organismal level are to a great extent synchronized by the steroid hormone ecdysone and sesquiterpenoid juvenile hormone (JH) [ 17 ]. Both hormones have a similar way of action. JH binds to its receptor MET, encoded by the Methoprene-tolerant gene, and thus facilitates its heterodimerization with TAIMAN (TAI), an ortholog of mammalian steroid receptor coactivator, and drives the expression of target genes to regulate development and reproduction [ 18 – 22 ]. Ecdysone binds ECDYSONE RECEPTOR (EcR), heterodimerizes with its partner ULTRASPIRACLE (USP) [ 23 ], and upon replacement of corepressor SMRTER [ 24 ] by transcriptional coactivator TAIMAN, activates expression of target genes [ 25 ].

The complexity of the mammalian circadian clock is high due to duplications of some core circadian genes, such as period (per1-3) and mammalian-type cryptochrome (m-cry1-2, in the mammalian literature often abbreviated as cry1, cry2) and extensive interconnection with metabolism modulators including Steroid Receptor Coactivators [ 7 , 8 ]. The circadian clock architecture in insects is remarkably similar to the mammalian system, depending on a similar set of genes [ 9 ]. However, group-specific differences have accumulated through evolution, and thus comparative studies on various insect taxa with different clock setups can help unveil the plasticity and functionality of the insect clock. For example, cyclorrhaphan Diptera, including the fruit fly Drosophila melanogaster, is the only group that lost mammalian-type CRYPTOCHROME (m-CRY) [ 10 ]. The fruit fly circadian clock relies on d-tim in the repressive feedback loop [ 11 , 12 ], and d-TIM temporal degradation is a primary way of clock synchronization by light [ 13 , 14 ]. Moreover, Drosophila-type CRYPTOCHROME (d-CRY), lost in all Chordata, in Drosophila resets the circadian clock by serving as a neuronal blue light photoreceptor [ 15 , 16 ]. The honeybee Apis mellifera and other Hymenoptera use a more ‘mammalian-like’ circadian clock: they have lost d-cry and d-tim genes. Other insects, such as the linden bug Pyrrhocoris apterus (Heteroptera), combine mammalian and Drosophila clock setup with preserved d-tim but lost d-cry and thus enabled to study evolutionary transition clock state [ 10 ].

CLK in the center of the clock is a target of multiple regulations. Two basic leucine zipper (bZIP) transcription factors, PAR DOMAIN PROTEIN 1 (PDP1) and VRILLE (VRI) activate or repress, respectively, Clk transcription and their transcription is in return activated by CLK/CYC [ 5 ]. Concurrently, CLK-CYC drives the expression of the basic helix-loop-helix (bHLH) transcription factor clockwork orange (cwo). CWO protein enhances the removal of CLK-CYC from the per and d-tim promoters to maintain transcription repression [ 6 ].

Circadian clocks are endogenous oscillators that synchronize multiple physiological and behavioral processes with the day/night cycle. The central insect pacemaker in the brain is a coupled neuronal network creating oscillation and rhythmicity by using several transcriptional-translational interlocked molecular feedback loops (TTFL) [ 1 ]. In Drosophila, transcriptional activators CIRCADIAN LOCOMOTOR OUTPUT CYCLES KAPUT (CLOCK, CLK) [ 2 ] and CYCLE (CYC) [ 3 ], an insect homolog of mammalian BRAIN AND MUSCLE ARNT-LIKE 1 (BMAL1), bind promotors of negative regulators period (per) and Drosophila-type timeless (d-tim) and activate their transcription. Upon transfer to the cytoplasm, per and d-tim mRNA get translated, and proteins, upon phosphorylation, enter the nucleus and inhibit their own transcription by binding to CYC and CLK [ 4 ].

For functional analysis we designed two non-overlapping dsRNA fragments, which target all known B. germanica tai transcripts ( Fig 7A ). B. germanica males were injected with 1 μl of tai fragment #1 or #2 dsRNA, or control lacZ dsRNA, or left without treatment (control intact). After a 4-day entrainment (16 h light/8 h dark), the locomotor activity was measured in constant darkness at 27°C. The portion of rhythmic males was low in the whole experiment, ranging from 46.7% in control intact males to 23.9% and 12.8% in tai dsRNA fr #1 and #2 injected males ( Fig 7C ), respectively. However, tai silenced males with strong rhythmicity had τ 26.8 and 26.03 hours ( Fig 7C and 7D ). The slow pace of the clock in B. germanica and P. apterus males thus seems to be a shared feature of at least two distant insect species.

(A) B. germanica taiman gene locus with mapped transcripts from [ 46 ]. Alternatively spliced exons and the position of RNA interference fragments are depicted. Note the different scales of exons and introns. (B) The domain structure of B. germanica TAIMAN protein, represented by (longest) TAI isoform (1668 amino acids, encoded by HG965205.1 transcript). See Fig 1 for the domain description. (C) B. germanica males were injected with 1 μl of tai fragment #1 or #2 dsRNA, or control lacZ dsRNA, or left without treatment (control intact), and individually placed in Petri dishes partially filled with agar, providing food and water. Males were entrained for 4 days (16 h light/8 h dark) and then their locomotor activity was measured in constant darkness at 27°C. The impact on behavioral rhythmicity is depicted as a percentage of males with strong rhythmicity, complex rhythmicity, and arrhythmicity. Individual free-running period (τ) values of males with strong rhythmicity (from column 3) are shown as individual dots, magenta lines represent the mean ± SEM (Standard Error of Mean). The mean τ, SEM, and statistical difference from the ‘control (intact)’ (p value) (One-way ANOVA with Dunnett’s multiple comparisons post hoc test). ns = statistically nonsignificant (p > .05). See details on FRP determination in Materials and methods. (D) Examples of B. germanica locomotor activity double-plotted actograms of males with strong rhythmicity from (C). The males were either injected with dsRNA (lacZ or tai) or left untreated (control intact males). Males were exposed to light-dark cycles for four days (depicted as white rectangles) and then were released to constant-dark conditions, indicated by a black arrowhead in the control (intact) actogram. Notice predominant nocturnal locomotor activity in all treatments. See Materials and methods for a detailed description of strong and complex rhythmicity phenotype.

B. germanica possesses a single tai gene, which we reconstructed from published genomic contigs and transcripts ( Fig 7A ) [ 46 ; see methods for more details]. Our analysis suggests four published B. germanica tai coding sequences are longer at their 5’end compared to the annotated gene model in (PYGN01000796.1) and encode an extra 30 amino acids at the N terminal end. Predicted B. germanica TAI is very similar to other insect TAI proteins and contains all conserved domains and motifs ( Fig 7B ).

TAI seems to be a robust modulator of the linden bug circadian clock. Whether TAI clock modulatory function is a common feature in other insects is unknown. Therefore, we decided to address taiman role in the circadian clock of the German cockroach Blattella germanica (Blattodea), an insect species that separated from true bugs (Heteroptera) ~ 380 mya [ 45 ].

(A) Two-day-old P. apterus adult females were injected with depicted dsRNA or left untreated, and upon addition of males two days later, were scored for egg-laying and embryonal/larval development. The majority of control females, 92.6% of intact and 88.9% of lacZ dsRNA injected females, laid eggs, whereas the ratio of egg-laying females dropped to 40%, 60%, and 0% in tai-E23-24, tai-E22-24 and tai fr #3 dsRNA injected females, respectively. (B) No effect of tai silencing was observed on male fecundity, as two-day-old P. apterus males, injected with tai fr #3 dsRNA, control lacZ dsRNA, or left untreated, were able to fertilize P. apterus females.

The effect of isoform-specific tai knockdown on P. apterus reproduction was tested by injection of dsRNA in two-day-old virgin females, crossed with males two days later, and scored for an egg laying and presence of embryos/larvae ( Fig 6A ). Knockdown of all tai isoforms, performed by tai fr #3 RNAi, led to complete female sterility. Depletion of most tai isoforms by tai-E23-24 and tai-E22-24 RNAi allowed only 40% and 60% of females, respectively, to lay eggs. Furthermore, embryonic development was blocked in most eggs that were laid by tai-E23-24 and tai-E22-24 dsRNA-treated females. Females injected with tai-iso-C and tai-E21 dsRNA showed only a slight decrease in egg-laying and development of embryos and larvae compared to control intact and lacZ dsRNA-injected females. Interestingly, females crossed with two-day-old tai fr #3 RNAi males (= paternal RNAi) laid eggs at a level comparable to females crossed with control intact and lacZ dsRNA-injected males ( Fig 6B ). Data suggest that paternal tai RNAi has no effect on female egg laying and embryonic/larval development. Contrary to the proposed role of tai isoforms in development, isoform-specific tai RNAi experiments do not support a specific role for any tai isoforms in circadian clock.

We decided to apply isoform-specific tai RNAi in other TAI-govern processes well-established in insects: development and reproduction. Over 93% of intact or egfp- and lacZ-dsRNA-injected larvae developed into adults ( Fig 5A ). In larvae injected with dsRNA fragments targeting the most, tai-E23-24 and tai-E22-24, and all, tai fr #3, isoforms died no later than in the 5 th instar, with 44.1%, 46.9%, and 62.5% of them dying as late 4 th instar larvae, and 8.8%, 9.4%, and 0%, respectively, reaching late L5 instar ( Fig 5A ). Knockdown of isoforms containing exon 21 (tai-E21) permitted reaching late L5 in 85.7% of injected larvae, but only 14.3% of total larvae entered L5/adult ecdysis and died as pharate adults. All larvae injected with tai-iso-C dsRNA developed normally up to the late L5 instar, 52% of them molted to the pharate adult stage and only 16% developed into adults. All tai RNAi fragments tested had a developmental phenotype but differed in the reached developmental stage and penetrance level.

Removal of TAI-C isoform via tai-iso-C RNAi had no effect on τ and was comparable to intact and lacZ dsRNA injected control males ( Fig 2A ). Knockdown of tai isoforms containing alternative exon 21 prolonged τ to 26.10 hrs. Silencing performed by tai-E23-24 and tai-E22-24 RNAi yielded τ of 27.38 and 27.04, respectively, which is close to τ for common tai fr #1 and #2 RNAi ( Fig 2A ). tai fr #3 RNAi, used as a positive control, slowed τ to 28.63 hrs. Based on all these results, the effect of tai isoform depletion on τ seems to be rather additive than dependent on specific isoform knockdown.

We designed and cloned tai fragments targeting (i) isoform-specific exon 20 extension (tai-iso-C RNAi), silencing a single (protein) isoform TAI-C, (ii) alternative exon 21 (tai-E21 RNAi), silencing TAI-A and TAI-D, (iii) exons E23-24 (tai-E23-24 RNAi) or E22-24 (tai-E22-24 RNAi), silencing TAI-A, TAI-B, TAI-D, and TAI-E isoforms ( Fig 1C ). Common tai RNAi fragment #3, silencing all isoforms, was used as a positive control. The experimental design was the same as presented for common tai RNAi fr #1, #2, and #3 (Figs 1 and 2A ). Although the rhythmicity of all males measured in the taiman isoform-specific RNAi experiment was lower, including control intact and lacZ dsRNA-injected males, the ratio of strongly rhythmic males gradually decreased with the number of silenced isoforms and dropped to 43.9% in tai fr #3 RNAi.

The role of TAIMAN in the modulation of the P. apterus circadian clock seems to be JH- and ecdysone-independent. The contrast between striking tai RNAi circadian phenotype and no effect of JH and ecdysone signaling manipulations on males’ free-running period or rhythmicity encouraged us to further support the function of TAI in the linden bug physiology and the circadian clock. We decided to perform isoform-specific tai knockdowns and to test whether TAI clock modulatory function can be associated with specific tai isoform(s). Eight tai isoforms were detected at the level of mRNA, which can be translated into five TAI proteins in P. apterus ( Fig 1D ).

Taken together, out of all our ecdysone signaling manipulations, only taiman silencing had a strong effect on the pace of the linden bug circadian clock, corroborating JH- and ecdysone-independent function of TAI in the linden bug circadian clock.

Similarly, TAI facilitates physical interaction between a WW (tryptophan-tryptophan) domain of YORKIE (YKI), a transcriptional co-activator of the Hippo signaling pathway, and EcR, to regulate Drosophila imaginal disc growth [ 44 ]. We silenced P. apterus yki via RNAi to test its possible effect on the linden bug clock pace, but the τ was comparable to the free-running period in the control (intact) and lacZ dsRNA-injected males ( Fig 4B ).

Ecdysone signaling in the fruit fly is negatively regulated by the Zinc finger C2H2-type transcription factor ABRUPT (AB), which attenuates ecdysone signaling by binding the bHLH domain of TAI by its BTB (Broad-Complex, Tramtrack and Bric a brac) domain [ 42 ]. Although no role for AB in Drosophila circadian rhythm is known and its expression was reported from a subset of dorsal lateral neurons (LNd) rather than from ventral lateral neurons (LNv) expressing EcR, HR51 (= DHR51), and E75B [ 30 , 34 , 43 ], we decided to test ab RNAi effect on the linden bug circadian behavior. Contrary to its function as a negative regulator of TAI in ecdysone signaling, τ of ab-silenced linden bug males was not shorter than the τ in control (intact) males, although the rhythmicity of ab RNAi males was slightly lower ( Fig 4B ).

The ecdysone signaling pathway regulates expression of many target genes, including nuclear receptors (NR) [ 23 ]. We knocked down three NR linked with reported function in Drosophila and Thermobia circadian clock, Hr51, Hr3, and E75. Hr51- and Hr3-silenced linden bug males were rhythmic and their τ was not significantly changed ( Fig 4C ). Interestingly, E75-silenced males were rhythmic between the 1 st and 7 th day in constant darkness but became gradually arrhythmic in the subsequent seven days ( Fig 4A ). The effect was more prominent in E75 RNAi fragment #2, where 85% of males became arrhythmic. Τ determined in the rhythmic males from E75 RNAi fr #2 measurement was significantly longer (24.52 hours) during the first 7 days in darkness when compared to control (intact) males ( Fig 4A ).

In the inverse experiment, injection of the linden bug major ecdysone makisterone A, insect most common 20-hydroxyecdysone (20E), or ecdysone mimic ponasterone A into P. apterus adult males (at Zeitgeber Time = 3–5), affected neither their rhythmicity nor τ, and the phenotypes were comparable to intact, lacZ dsRNA-injected, 0.4% EtOH- and 5% EtOH-treated control males ( Fig 2D ).

Freshly hatched P. apterus 4 th instar larvae were injected with dsRNA or left untreated (intact). (A) Isoform-specific knockdown was more pronounced the more taiman isoforms were targeted, see Fig 1 for details of which isoforms were targeted by given dsRNA fragments. Animals marked with † died within 2–3 days after reaching the depicted developmental stage. Data presented in (B) and (C) were generated in the same experiment but plotted separately for higher clarity. Data of egfp dsRNA-injected control animals are identical in both (B) and (C) plots. Animals marked with * stopped developing, stayed in the instar for up to one month, and ultimately died as L4 (EcR, spo) or L5 (shd) larvae.

Since blocked ecdysone synthesis or ecdysone binding by the receptor had no effect on circadian rhythms, we tested the RNAi efficacy in larval development, where ecdysone signaling plays a crucial role. Firstly, injection of spo dsRNA into freshly ecdysed fifth-instar (= L5) larvae dropped makiA levels to about 4% of the amount detected in control larvae ( S1 Fig ), proving spo RNAi to be an effective way of makiA level reduction. EcR, spo, shd, and to a lesser degree also usp dsRNA, injected into the one-day-old 4 th instar larvae (= L4) blocked ecdysis ( Fig 5B and 5C ), as expected for genes essential for ecdysone signaling. EcR and spo RNAi larvae never molted to the 5 th instar, stayed in the ‘infinite’ 4 th instar for up to one month, and eventually died. The majority (91.7%) of control egfp dsRNA-injected larvae molted to L5 instar within 4–5 days and 90.9% of them developed into adults. Interestingly, 60% of shd and 40% of usp dsRNA injected L4 larvae molted to the 5 th instar, but usp died shortly after the ecdysis, and 88.9% of shd RNAi (L5) larvae became ‘infinite’ L5 larvae. Taken together, larval RNAi of (mainly) EcR and spo proved to be an efficient way to downregulate ecdysone signaling in the linden bug.

Makisterone A (makiA) is a major ecdysteroid of some true bugs (Heteroptera, Pentatomomorpha), including the linden bug P. apterus [ 40 , 41 ] ( S1 Fig ). MakiA relative amount determined by liquid chromatography is high in seven-day-old fifth-instar lacZ dsRNA-injected larvae and the freshly eclosed P. apterus males. The amount drops to about 10% in seven-day-old males ( S1 Fig ). We decided to block ecdysone synthesis by knocking down two crucial cytochrome P 450 enzymes, encoded by spook (spo, CYP307a1) and shade (shd, CYP314a1) genes and ecdysone intracellular receptor by knocking down EcR and usp genes. Linden bug spo-, shd-, EcR- and usp-silenced males were fully rhythmic and their τ was unchanged when compared to the control intact and lacZ dsRNA injected males ( Fig 2C ).

Thus, insensitivity of the linden bug circadian clock to disrupted JH signaling, contrasting to the strong tai knockdown effect, suggested a clock-specific (JH-independent) mechanism for TAI functions. One of the possible mechanisms could be TAI coactivating role in ecdysone signaling, in which TAI binding to EcR/USP complex permits target gene expression [ 25 ].

We decided to test whether TAI, a heterodimerizing partner of MET in a JH receptor complex, could execute its circadian clock modulatory function through MET/JH signaling pathway. However, the knockdown of Met had no effect on the rhythmicity or τ ( Fig 2B ). To further test the requirement of JH for the circadian clock, we surgically removed corpora allata, the endogenous source of JH (allatectomy, CAex) in males 24–48 hours after adult ecdysis that were reared under the long-day (18:6 h light/dark) regime, which is the photoperiod stimulating activity of CA. The rhythmicity of CAex males was fully comparable to intact, sham-operated, and lacZ dsRNA-injected control males ( Fig 2C ). Although allatectomized males’ τ of 24.15 hours is close to the expected 24-h period, it was significantly different from 23.33 hours measured in control (intact) males. The mean τ of males anesthetized by submersion in water was 24.07 hours, being almost identical to CAex-treated males. The period in sham-operated males was 23.69 hours and close to control males in other experiments ( Fig 2C ). Since the allatectomy had no effect on rhythmicity and τ, we decided to perform an inverse experiment, in which we administered JH mimic methoprene and its vehicle, acetone. In both cases, application on adult males induced no change in the rhythmicity or τ ( Fig 2D ), suggesting little or no role of JH in the linden bug circadian clock.

P. apterus males were treated and analyzed as described in Fig 2 . (A) The percentage of males with rhythmic, complex, or arrhythmic behavior after RNA interference against P. apterus E75. The behavior of the males was analyzed separately for the first 7 days in the dark (1–7 in DD, left panel), and for the next 7 days (8–14 in DD, right panel). The mean τ, SEM, and statistical difference from the ‘control (intact)’ (P value) were determined by One-way ANOVA with Dunnett’s multiple comparisons post hoc test (for days 1–7), and unpaired two-tailed t-test (t(24) = 0.9091, p = 0.3723) for intact vs. E75 fr #2. (B-C) P. apterus males were treated as indicated in the first row and their locomotor activity was measured and analyzed as described in Fig 2 (standard 10 days in DD). The mean τ, SEM, and statistical difference from the ‘control (intact)’ (p value) (One-way ANOVA with Dunnett’s multiple comparisons post hoc test). ns = statistically nonsignificant (p > .05), n/a = not applicable.

The males were injected with taiman fr #3 dsRNA, loaded into locomotor activity monitors, and their activity was recorded at 25°C. Males were exposed to light-dark cycles for five days (depicted as white rectangles) and then were released to constant-dark conditions, indicated by a black arrowhead in actogram (A). Examples of (A) control (intact = wild-type), (B) control lacZ dsRNA injected, (C) males after taiman knockdown with strong rhythmicity (= one clear peak of locomotor activity), (D) males after taiman knockdown with complex rhythmicity (= more than one peak of the locomotor activity or free-running period has changed during the measurement). See materials and methods for a detailed description of strong and complex rhythmicity phenotype.

P. apterus males were treated as indicated in the first column and their locomotor activity was measured in constant darkness at 25°C. The impact on behavioral rhythmicity is depicted as a percentage of males with strong rhythmicity, complex rhythmicity, and arrhythmicity. Individual free-running period (τ) values of males with strong rhythmicity (from column 3) are shown as individual dots, magenta lines represent the mean ± SEM (Standard Error of Mean). The mean τ, SEM, and statistical difference from the ‘control (intact)’ (p value) (One-way ANOVA with Dunnett’s multiple comparisons post hoc test for all analyses except for comparison of control (intact) and Methoprene-tolerant (Met) dsRNA, where unpaired two-tailed t-test was used). (A) Silencing of taiman, targeting all tai isoforms (white region) or taiman isoforms (grey region). (B) Silencing of juvenile hormone (JH) receptor Met (t(31) = 0.7152, p = 0.4799) (white region) and components of ecdysone receptor complex (EcR and usp) (grey region). (C) The effect of hormone production removal on the behavioral rhythmicity and τ. Juvenile hormone-producing corpora allata gland removal (CAex) (white region); knockdown of ecdysone-synthetizing enzymes encoded by spook and shade genes (grey region). (D) The effect of hormones or their mimics administration on P. apterus behavioral rhythmicity and τ. Topical JH mimic methoprene application (upper white region); injection: of 20-hydroxyecdysone (20HE) (upper grey region), crustacean ponasterone A (middle white and lower grey regions), and natural ecdysteroid makisterone A (lower white region). Abbreviations and description: control (intact) = non-treated males, control (lacZ dsRNA) = lacZ double-stranded RNA fragment, tai = taiman, tai-iso-C = taiman isoform C, tai-E21 = taiman exon 21, fr #1, #2, and #3 represent non-overlapping dsRNA fragments, Met = Methoprene-tolerant, EcR = Ecdysone receptor, usp = ultraspiracle, anesthetized (in H 2 O) = anesthetized by submersing in water, sham = placebo surgery, CAex = surgical corpora allata removal, spo = spook, shd = shade, acetone = 4 μl of 100% acetone applied, anesthetized (CO 2 ) = anesthetized by CO 2 gas, methoprene = methoprene in acetone applied, 5% EtOH = 5% ethanol in H 2 O, 20E = 20-hydroxyecdysone in 5% EtOH, ponA = ponasterone A in EtOH, makiA = makisterone A. n = number of measured males, rhythmicity: strong = percentage of males with one clear stable period, complex = rhythmicity with more components or unstable period, arrhythmic = no significant period, ns = statistically nonsignificant (p > .05).

The knockdown of tai via RNA interference (RNAi) dramatically slowed the circadian clock FRP (= τ, tau) to 27.42–28.63 hours (Figs 2A and 3 ). τ was evaluated only in males, because contrary to females, their assay performance is not affected by reproduction-related phenomena. All three tested tai non-overlapping dsRNA fragments, targeting all detected isoforms, significantly slowed the pace of the clock (Figs 2A and 3C ), but the ratio of perfectly rhythmic linden bug males was comparable to control (intact) males ( Fig 2A ). The rhythmicity of control (intact) males was in some cases lower than in males with treatment (i.e., strong rhythmicity % of “control (intact)” in Figs 2A and 4A ), most likely due to heterogeneity in the population of cohort used in particular experiment. The ratio of males with more components or unstable period (= complex rhythmicity) was slightly higher in tai RNAi males (Figs 2A and 3D ). The use of three independent tai dsRNA fragments, consonant in the circadian phenotype, strongly supports tai RNAi specificity and excludes the possibility of tai RNAi phenotype being caused by an off-target effect.

(A) taiman transcription starts from alternative exons 1, 2, and 3, which join to exon 4, read in two alternative frames. Exon 21 is alternatively spliced and exon 20 can be extended in an isoform-specific manner. Note the different scales of exons and introns. (B) Domain structure of P. apterus TAIMAN protein, represented by (longest) protein TAI isoform D (1299 amino acids). N-terminal DNA-binding basic-Helix-Loop-Helix (bHLH, blue) domain is followed by two Per-Arnt-Sim (PAS, pink) domains, five centrally located LxxLL motifs (light blue), C-terminal Glutamine-rich regions (green). (C) Positions of taiman isoform-specific RNA interference fragments. Both tai-E21 and tai-iso-C target single exons, tai-E23-24 and tai-E22-24 target two and three exons, respectively. (D) All eight detected P. apterus taiman transcripts with color-highlighted exons encoding protein domains and motifs. Three core taiman RNA interference fragments mapped to tai-A transcript target all tai isoforms. Semi-quantitative expression of tai isoforms/exons. See ONT reads mapping description in Materials and methods for more details.

P. apterus TAIMAN, a steroid receptor coactivator, is encoded in the linden bug by a single tai gene ( Fig 1 ). The tai locus in our in-house assembled genome spans over 767 kbp ( Fig 1A ). Taiman transcripts were identified by mapping full-length Oxford Nanopore Technology transcriptomic reads to the P. apterus taiman genomic contig. Eight identified tai isoforms ( Fig 1D ) are generated by three mechanisms: (i) transcription from three alternative promoters, (ii) alternative read-through of exon 20, and (iii) alternative presence of exon 21. Each of the first three exons is transcribed from an independent promoter and is spliced to the canonical exon 4 ( Fig 1A and 1D ). Translation of six transcripts expressed from exons 1 and 2, tai iso-A-C, and iso-D-F, respectively, starts at the Methionine in exon 4, and those transcripts represent about 80% of tai expression. Transcripts expressed from exon 3 (tai iso-G, H) are prolonged by 40 amino acids (aa) at the N-termini, as a result of an alternative translational start site ( Fig 1A and 1D ). Transcripts tai iso-C, and iso-F represent minor isoforms in the brain. Both contain the read-through exon 20, and are translated into TAI-C protein isoform with extra 60 aa at the C-terminus but lack exons 21–24. Transcripts tai iso-A, iso-D, and iso-G include an alternative exon 21, which prolongs corresponding proteins by 45 aa, together constitute about 1/3 of tai brain expression.

Discussion

Here, we demonstrated taiman as a new modulator of the insect circadian clock. TAIMAN is the only insect ortholog of mammalian Steroid Receptor Coactivators [25] and represents a pleiotropic protein engaged in several signaling pathways, out of which juvenile hormone and ecdysone signaling pathways are the most studied [22,25,37,42,47]. TAI seems to be a robust modulator of insect locomotor circadian rhythms as all tai common dsRNA fragments lengthened the free-running period in at least two insect species, the linden bug P. apterus and German cockroach B. germanica (Figs 2A and 7C). Longer τ was also measured upon P. apterus tai isoform-specific knockdowns, except for tai-iso-C dsRNA targeting rare isoforms (Fig 1D), which left τ unchanged (Fig 2A). Mouse TAI ortholog Steroid Receptor Coactivator-2 has been shown as a transcriptional coactivator of BMAL1/CLOCK [8]. Although the average τ of SRC-2-/- mutant mice was not changed compared to controls, τ varied among individual mice. Moreover, mutant SRC-2-/- mice expressed an abnormal behavioral pattern in locomotor wheel-running behavior [8]. Thus, the TAI/SRC-2 clock modulatory role in two insect species and mice might represent a conservative feature of the circadian clock.

TAI is indispensable for the JH signaling pathway, and tai silencing blocks larval development [22,46,48] and reproduction [22,49]. The function of JH in the insect circadian clock stays elusive. The removal of corpora allata (allatectomy) in the linden bug males had no effect on the pace of the clock (this work), a phenotype found also in the honeybee (Apis mellifera) [50] and bumble bee (Bombus terrestris) [51]. The same circadian clock JH-insensitivity was found upon Met knockdown and when JH mimic methoprene was administered to the linden bug males [this work]. The fact that JH level manipulations are not reflected in the clock pace does not mean that either crosstalk between clock proteins and JH receptor proteins MET and TAI or JH-driven clock genes’ expression is not possible. These interactions are often localized in peripheral tissues, such as in the linden bug P. apterus fat body [38] and gut [39], and the mosquito Aedes aegypti fat body [37], and thus suggest a pleiotropic function of the clock proteins rather than direct involvement of the JH signaling in the control of the circadian rhythms. A notable similarity in tissue-specific clock architecture is found in mammals, where the importance of individual TTFL differs in a tissue-specific manner, which accounts for organ-specific clock gene expression and contributes to the hierarchical organization of the clock [52].

Ecdysone signaling was linked with circadian rhythms in Drosophila [30,36] but the involvement of TAI in the Drosophila circadian clock has not been reported. The ecdysone signaling role in Drosophila contrasts with our findings in the linden bug males, where EcR and usp knockdown had no effect either on rhythmicity or τ.

In D. melanogaster, EcR overexpression, knockdown, or expression of a dominant-negative EcR form in clock neurons caused higher arrhythmicity, weakened rhythm strength, and in some driver-UAS construct combinations lengthened τ [30,36]. On the other hand, Drosophila EcR (and usp) silencing using miRNA overexpression in PDF-positive neurons had a minimal effect on locomotor activity rhythmicity and no effect on τ [34], showing how delicate those experiments are. The linden bug EcR and usp dsRNA were injected into 1-2-day-old adult males, avoiding the detrimental effect of EcR knockdown throughout the development, as described for the Drosophila small ventral lateral clock neurons (s-LNvs) [53]. Moreover, the knockdown of two ecdysone-inducible nuclear receptors, E75B and HR51, in Drosophila s-LNvs, causes weaker circadian behavior or arrhythmicity [30,34,54,55]. Although P. apterus males are rhythmic after Hr51 and Hr3 knockdown and only the Hr51 free-running period is slightly (not significantly) longer, E75 silenced males gradually became arrhythmic and the rhythmic males had slightly longer τ, although only with one RNAi fragment. The onset of arrhythmicity after E75 dsRNA injection was detected at 8–14 days in DD (13–19 days after dsRNA injection) (Fig 4A), contrasting with strong circadian phenotypes published for core circadian genes [10]. In Drosophila, E75B physically interacts with PER and represses Clk expression in insect and mammalian cell cultures, and E75B was postulated as a factor protecting circadian rhythms from stress [30]. Whether P. apterus E75 interacts with the linden bug PER is unknown, but 43.3–50% of P. apterus per RNAi and 15–38% per null mutant males were still rhythmic with τ ~ 19–21 hours [10], suggesting lower dependence of the linden bug clock on PER. Interestingly, Drosophila E75 is not only a 20E-inducible early gene but can be induced by JH III (or JH mimic methoprene) in Drosophila S2 cells in the absence of 20E [56].

Ecdysone titer following daily rhythms was determined in several insect species [26–29] and we thus decided to test the effect of makisterone A (makiA), in insects the most common 20-hydroxyecdysone (20E), and ecdysone-mimic ponasterone A administration on P. apterus males’ rhythmicity and τ (Fig 2D). None of the aforementioned compounds changed the linden bug rhythmicity or τ and agreed with the results obtained after EcR and usp knockdowns. Ecdysone signaling-independent behavioral locomotor activity in linden bug adult males (Fig 2B and 2D) contrasts with the effect of disrupted ecdysone signaling through the bug larval development (Fig 5B and 5C). Although the relative makiA level was not determined in fourth-instar larvae, the amount of makiA in (late) fifth-instar larvae is comparable to levels detected in freshly hatched adult males and the titer gradually decreases throughout the first week of males’ adulthood. The relative amount of makiA in fifth-instar larvae can be effectively decreased by silencing spo (S1 Fig), thus proving the efficacy of spo knockdown. The larval and adult RNAi in P. apterus is equally efficient (compare Met RNAi in [19,22,57]). Larval spo, shd, EcR, and usp knockdown in P. apterus elicited strong developmental phenotype (Fig 5B and 5C), yet silencing of the same genes in adult males affected neither the rhythmicity nor τ, further supporting the linden bug ecdysone-independent clock rhythms (Fig 2C).

The mechanism through which TAI modulates linden bug circadian rhythms is unclear (Fig 8). TAI mammalian ortholog SRC-2 binds BMAL1 through LxxLL protein motifs and pulls down both BMAL1 and CLK [8]. Clk and cyc knockdown in P. apterus resulted in a complete arrhythmicity [10]. LxxLL motifs are known to interface TAI/SRC interaction with nuclear receptors including EcR and USP in insects [25,58]. BMAL1/CLK/SRC-2 drives expression of PER1 by binding to per1 promoter but SRC-2/BMAL1 binding was also enriched in several other known circadian genes, together with genes involved in metabolism [8]. P. apterus and B. germanica TAI-depleted males have a free-running period longer by several hours, suggesting TAI ‘speeds up’ the clock pace in control, non-treated males.

It will be interesting to see whether TAI participates in the Drosophila circadian clock. Although the circadian clock is generally well conserved in insects [59,60], there are several noteworthy unique features of Drosophila clock machinery: For example, the transactivation domain (TAD) is localized only to CLK in the fruit fly, whereas CYC does not have this domain. The same pattern is found only in Cyclorrhapha (a subgroup of Diptera), whereas the TAD is localized to CYC/BMAL1 in the other bilaterians (Bilateria, most animals except sponges, placozoans, cnidarians, and ctenophores). TAD-dependent repression of BMAL1 requires m-CRY [61], a protein that has been lost in Drosophila, and the transition of TAD from BMAL1 to CLK corresponds exactly to the loss of m-CRY [62]. In mice, SRC-2 interacts with BMAL1-CLK and coregulates their function [8]. At this point, it is unclear whether TAI physically interacts with BMAL1 or CLK in P. apterus and B. germanica. Given the above-described modifications of CLK-CYC in Drosophila, it will be interesting to see whether TAI participates in the Drosophila (cyclorrhaphan) clock, and if so, whether the role is similar to that of P. apterus and B. germanica, or whether some Drosophila-specific modifications are identified.

This presented study introduces steroid receptor coactivator TAIMAN as a new insect circadian clock factor. A similar role of insect TAIMAN and SRC-2 in mammals suggests that TAI/SRC clock function has been conserved for more than 500 million years (Fig 9). Although our results suggest that TAIMAN function in the insect circadian clock is ecdysone- and juvenile hormone-independent, the exact mechanism through which TAIMAN modulates the circadian clock function remains elusive.

[END]
---
[1] Url: https://journals.plos.org/plosgenetics/article?id=10.1371/journal.pgen.1010924

Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/

You are viewing proxied material from magical.fish. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.