Conceptual Framework of the Eco-Physiological Phases of Insect Diapause Development Justified by Transcriptomic Profiling

Home , Diapause

Conceptual framework of the eco-physiological phases of insect diapause development justified by transcriptomic profiling

Vladimír Koštála,1, Tomáš St etina a,b, Rodolphe Poupardina,2, Jaroslava Korbelováa, and Alexander William Bruceb

aBiology Centre, Institute of Entomology, Academy of Sciences of the Czech Republic, 37005 Budweis, Czech Republic; and bFaculty of Science, University of South Bohemia, 37005 Budweis, Czech Republic

Edited by David L. Denlinger, Ohio State University, Columbus, OH, and approved June 29, 2017 (received for review May 3, 2017)

Insects often overcome unfavorable seasons in a hormonally ecdysteroids (8, 9). However, any further generalizations about regulated state of diapause during which their activity ceases, insect diapause are strongly complicated by three facts. First, the development is arrested, metabolic rate is suppressed, and toler- insect taxon is enormously rich and diverse. It is believed that ance of environmental stress is bolstered. Diapausing insects pass diapause responses evolve polyphyletically and very rapidly in through a stereotypic succession of eco-physiological phases termed different insect lineages as they encounter diverse environmental “diapause development.” The phasing is varied in the literature, and adversity (10–13). Second, different insect species enter diapause the whole concept is sometimes criticized as being too artificial. in different ontogenetic stages that differ widely in the com- Here we present the results of transcriptional profiling using custom plexity of body architecture and physiology (9). Third, different microarrays representing 1,042 genes in the drosophilid fly, Chymo- insect species enter diapause under various environmental con- myza costata. Fully grown, third-instar larvae programmed for dia- texts. Thus, hibernation is widespread in polar and temperate pause by a photoperiodic (short-day) signal were assayed as they regions, aestivation often occurs in Mediterranean and other dry climate zones, and tropical diapauses may respond to seasonality traversed the diapause developmental program. When analyzing – the gradual dynamics in the transcriptomic profile, we could readily primarily in biotic interactions (14 16). Despite this diversity in distinguish distinct diapause developmental phases associated with insect diapause responses, several phenotypic features occur al- induction/initiation, maintenance, cold acclimation, and termination most ubiquitously: developmental arrest, metabolic suppression, and environmental stress resistance (17), so that some authors by cold or by photoperiodic signal. Accordingly, each phase is char- have proposed the notion of a common genetic “toolkit” for acterized by a specific pattern of gene expression, supporting the diapause (18, 19). According to this scheme, some common physiological relevance of the concept of diapause phasing. Further, diapause-induced gene-expression profiles might be shared we have dissected in greater detail the changes in transcript levels across insect (or even across invertebrate) taxa. However, when of elements of several signaling pathways considered critical for comparing transcriptional profiles linked to dormancy in three diapause regulation. The phase of diapause termination is associ- invertebrate species (flesh fly, fruit fly, and nematode), Ragland ated with enhanced transcript levels in several positive elements et al. (20) found little support for the genetic toolkit hypothesis stimulating direct development (the 20-hydroxyecdysone pathway: and concluded that there may be many transcriptional strate- Ecr, Shd, Broad; the Wnt pathway: basket, c-jun) that are countered gies for producing physiologically similar dormancy responses. by up-regulation in some negative elements (the insulin-signaling One additional generalization on insect diapause posits that pathway: Ilp8, PI3k, Akt; the target of rapamycin pathway: Tsc2 and diapausing insects, although developmentally arrested, pass through 4EBP; the Wnt pathway: shaggy). We speculate such up-regulations a stereotypic succession of eco-physiological phases called may represent the early steps linked to termination of diapause programming. Significance

insects | diapause | development | transcriptomics | microarrays Here we reassess the conceptual framework of insect diapause ’ as a dynamic succession of endogenously and exogenously he obligatory diapause is a fixed component of the insect s driven changes in physiology (“physiogenesis”)byassayingthe Tontogenetic program, whereas the facultative diapause rep- gradual dynamics in the transcriptome as insects traverse the resents an optional alternative pathway to direct ontogeny. diapause developmental program. We show the objectivity and Facultative diapause (herein referred to as “diapause”) is a state eco-physiological relevance of the different phases of diapause of environmentally programmed and centrally regulated developmental arrest, usually accompanied by metabolic suppression, development by describing unique transcriptional profiles in which secures survival over unfavorable seasons and synchro- each phase. Accordingly, the concept should serve future re- nizes the insect life cycle to the seasonality of abiotic environ- searchers as a general platform for the unification of timing “ ” mental factors and biotic interactions (1–5). An improved scales and the interpretation of various -omics data obtained knowledge of diapause is essential for understanding insect life in diverse insect species encountering different ecological situ- cycles and for the development of management strategies for ations. We argue such standardized phasing of diapause devel- economically important insect pests (6) and accurate predictions opment is critical for further molecular dissection of the mechanistic of insect populations’ responses to climate change (7). basis of insect diapause. At a specific sensitive stage, insects perceive environmental token stimuli (any stimulus that signals the upcoming seasonal Author contributions: V.K. designed research; T.S., R.P., and J.K. performed research; change in advance, most often the photoperiod) that reliably A.W.B. contributed new reagents/analytic tools; V.K. and R.P. analyzed data; and V.K. wrote the paper. mark seasonal time (calendar) and switch from direct development to the diapause pathway, typically long before the adverse The authors declare no conflict of interest. period arrives. (Note: in this paper, “direct development” refers This article is a PNAS Direct Submission. to an ontogenetic pathway without intervening diapause.) It is 1To whom correspondence should be addressed. Email: [email protected]. well established that switching between direct development 2Present address: Institut für Populationsgenetik, Vetmeduni Vienna, 1210 Vienna, Austria. and diapause is controlled by the decrease or absence of sig- This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. naling of the basic developmental hormones, juvenoids and/or 1073/pnas.1707281114/-/DCSupplemental.

8532–8537 | PNAS | August 8, 2017 | vol. 114 | no. 32 www.pnas.org/cgi/doi/10.1073/pnas.1707281114 Downloaded by guest on September 25, 2021 “diapause development” or “physiogenesis” (1). This general- (PCA) (Fig. 1A). Early third-instar larvae (nd1 and id1) are most ization also has caused controversies in the past (4, 21, 22); in sensitive to photoperiodic signal. The nd1 larvae will continue in particular, the separation of and naming of the successive phases direct development, whereas the id1 larvae are destined to remains unstandardized and is sometimes criticized as artificial diapause development (24). We found 79 genes differentially (for a review, see ref. 23). expressed (DE) in the id1 and nd1 samples (Dataset S1,Excel We used custom microarrays to compare gene-expression sheet id1 vs. nd1), comprising 7.7% of the total number of profiles for 1,042 mRNA transcripts in the larvae of the droso- 1,042 sequences printed on a custom microarray. In an earlier philid fly Chymomyza costata. The genes were arbitrarily selected study (25) we compared the transcriptional profiles of similar to cover broadly the major structures and processes known or larvae (at the same ontogenetic stages) using RNA sequencing suggested to be involved in insect diapause expression: biological (RNA-seq) and found 1,313 DE sequences, comprising 6.2% of clocks, hormones and signaling cascades, regulators of the cell a total 21,327 sequences. Further, the identity of DE sequences division cycle, energy metabolism and detoxification, response to showed good overlap between two studies: 37 of 45 (82.2%) up- temperature stimulus and cryoprotection, cytoskeleton, biologi- regulated DE sequences in this study were also up-regulated in cal membranes, and transport systems. In addition to comparing the earlier study; similarly, 26 of 34 (76.5%) down-regulated diapause and nondiapause animals, we focused mainly on compar- DE sequences in this study were also down-regulated in the ing the successive eco-physiological phases of diapause develop- earlier study (with a 1.5-fold difference as a threshold for up- or ment, namely, induction, initiation, maintenance, and termination down-regulation). In the RNA-seq study, the short-day–reared, (sensu ref. 23). We also distinguished between the natural (horotelic, diapause-destined larvae (designated as “id1” in this study) sensu ref. 2) and the unnatural (tachytelic, sensu ref. 2) mecha- showed down-regulation of several key genes involved in 20- nisms of diapause termination stimulated by cold and by a long- hydroxyecdysone (20HE) biosynthesis and perception, indicat- day photoperiod, respectively. Our primary aim was to assess ing an inhibition of the developmental hormone signaling. The whether the phases of diapause development are characterized transcriptional pattern further suggested that the hormonal by specific transcriptional profiles that would demonstrate their change was translated into the down-regulation of a series of physiological relevance objectively. Further, we compare the other transcriptional factors, finally resulting in cessation of the phase-specific transcriptional responses observed in C. costata cell-division cycle and deep restructuring of metabolic path- with the analogous responses published for other insects with ways, all robustly reflected in transcriptional patterns (25). the aim of assessing the scope of evolutionary conservation/ Thus, the separation of id1 vs. nd1 clusters on the PC1 vs. convergence in the transcriptional level of diapause regulation PC2 plot (Fig. 1A) is validated by our earlier RNA-seq analysis. (i.e., addressing the applicability of the hypothetical genetic Hence, the separation of id1 vs. nd1 clusters can serve as an toolkit for diapause). accurate gauge for the interpretation of the separation between individual phases of nondiapause or diapause development. Results and Discussion Direct validation analysis, performed with six selected se- Differences Between Diapause and Nondiapause Early Third Instars. quences that exhibited the highest fit to the PCA model (Fig. The clusters of transcriptional profiles of larvae destined for direct 1B), showed a very good match between microarray and RT- development and larvae destined for diapause induction and ini- qPCR methods (Fig. 1C). tiation (nd1 and id1 larvae, respectively; for an explanation of abbreviations, see Fig. S1 and Table S1) are clearly separated along Separation of Successive Phases of Diapause Development. The SCIENCES

the principal component 2 (PC2) axis on the principal component 1 separation of individual phases of diapause development is ob- APPLIED BIOLOGICAL (PC1) vs. PC2 ordination plot of the principal components analysis vious in the PCA ordination plot (Fig. 1A). We performed five

Fig. 1. Gradual change in transcriptional profiles associated with direct and diapause development in third-instar C. costata larvae. (A) Results of PCA analysis exhibiting a clear separation of treatment clusters representing direct nondiapause development (nd1, nd2) and diapause development (id1, id2, id3, id4, md1, md2, ppt, ca, ct1, and ct2). The ellipses were arbitrarily drawn around replicated arrays to help resolution. All treatments had four replicates (three biological replicates plus one technical replicate; solid symbols are used to highlight two technical replicates of one entire microarray). For further explanations, see text, Fig. S1,andTable S1.(B) Eigenvectors are shown for sequences that best fit the PCA model (eigenvector loading >90%). Red ellipses highlight six target sequences selected for validation of microarray analysis by RT-qPCR. The Inset shows the distribution and 90% cutoff circle for all 3,126 eigenvectors (1,042 sequences, each represented in triplicate on our custom microarray). (C) Validation of microarray results using RT-qPCR. (Upper) The heat map shows the log2-nor- malized fluorescence value (NF) of the microarray spot signals of six target sequences. (Lower)Theheatmap

shows log2 relative mRNA abundance (ddCT,meanof five references per target sequence) in RT-qPCR analysis. The blue–red color scale displays a span between the minimum and maximum values in each single row of the heat map.

Koštál et al. PNAS | August 8, 2017 | vol. 114 | no. 32 | 8533 Downloaded by guest on September 25, 2021 additional PCA analyses using five different subsets of micro- decrease with progression of diapause from 1d2 to id4, i.e., array sequences rather than analyzing all 1,042 sequences to- (id2 vs. id1), 6.3% DE sequences of a total 1,042 sequences are gether. The results of such subset PCA analyses (Fig. S2) DE sequences; (id3 vs. id2), 5.2%; and (id4 vs. id3), 3.9%. The confirmed that the clustering and separation of successive phases qualitative aspect of transcriptional changes also tends to evolve of diapause development remain distinct, no matter which subset gradually, as indicated by the changes in gene categories that of sequences is used in the analysis. The specific transcriptional exhibit the highest relative proportion of DE sequences and patterns of individual diapause phases are robust and cover also by the relatively small overlap of DE sequences between the different gene categories/ontologies. We discuss the phases of successive comparisons/phases: i.e., for (id2 vs. id1) vs. (id3 vs. diapause development in their logical order below. id2), 18.7% of 111 DE sequences overlap, whereas for (id4 Diapause induction and initiation. In C. costata, the phases of dia- vs. id3) vs. (id3 vs. id2), 17.9% of 95 DE sequences overlap pause induction and initiation proceed in immediate succession (Dataset S1). and are even partially overlapping. Diapause induction is a Diapause maintenance. The maintenance phase (md1/2) is ecolog- gradual process in which the environmental token stimuli (pho- ically the most explicit phase of diapause development. In this toperiod, thermoperiod) are perceived and, in an interplay with phase, the insect maintains the developmental arrest induced in modifying and directly limiting factors (temperature, population earlier phases. The diapause is maintained despite the environ- density, diet quantity and quality), are transduced into the hor- mental conditions potentially being permissive for the continu- monal changes that cause the expression of a complex diapause ation of development. The insects usually continue responding to phenotype (23, 24). Diapause initiation is characterized by inducing token stimuli that help them maintain diapause/prevent transiently high metabolic activity and continuation of feeding its termination. The maintenance phase is also characterized by and locomotion (although development is already arrested), se- extremely slow metabolism securing only the basic homeostatic curing various needs linked to preparations for the upcoming functions. In our results, the overall static character of the long period of behavioral inactivity and metabolic suppression maintenance phase is reflected in a very small number of DE (23). The process of diapause induction and initiation culminates genes: (md1 vs. id4), 0.7% DE sequences. However, we do ob- at an age of ∼6wkinC. costata (26). Fig. 2 documents how the serve that the number of DE sequences increases markedly transcriptional profile in C. costata larvae changes gradually during the subsequent phases of maintenance: (md2 vs. md1), C costata during the progression through diapause induction and initia- 6.3% DE sequences (Fig. 2). In . , diapause is photo- periodically maintained until practically all larvae die (24). We tion. The quantitative aspect of the gradual transcriptional speculate that the increasing number of DE sequences in the change can be judged by comparing the sizes of the pie charts later phase of maintenance (md2) may reflect a combination of (the bigger the pie, the larger is the proportion of DE se- at least three theoretical factors: (i) the mobilization of alter- quences). Color coding of segments shows the gene functional native energy substrates (when the preferred substrate becomes categories, helping to identify visually the categories with the more limited, the alternative substrate may be recruited, thus highest relative proportions of DE sequences and allowing a requiring the mobilization of different genes); (ii) the need to rough comparison of the qualitative aspect of the gradual tran- replenish and metabolize the energy substrates by occasional scriptional change (for more details, see Dataset S1). We can see feeding [we have shown that half of diapausing C. costata larvae that the magnitude and complexity of the changes gradually die within 2 mo when maintained without food (24), suggesting that occasional feeding is required, and such feeding would also require the recruitment of potentially new gene sets for digestion and metabolism]; and (iii), the need to cope with increasing homeostatic imbalances (with increasing larval age, toxic me- tabolites may accumulate that could also induce a transcriptional response). Diapause termination by long photoperiod. As in many other insects, diapause termination in C. costata requires a specific stimulus: either the lengthening of day (an unnatural stimulus) that leads to tachytelic termination or exposure to cold (a natural stimulus), causing horotelic termination (2, 26). In some other insects, the termination of diapause may be spontaneous, requiring no specific environmental signal (2, 26). In C. costata larvae, the change of the photoperiodic signal from short day to long day resulted in an increase in the relative number of DE sequences: [photoperiod termination (ppt) vs. md1], 8.2% DE sequences (Fig. 2). The rate of ppt-linked change in the transcriptional profile was relatively rapid (i.e., occurring within 1 wk) and was not associated with any overtly observable morphological or behavioral changes. The larvae began their wandering behavior and pupariation only 3–4 wk after transfer to the long photoperiod. Thus, the long photoperiod-induced alteration in 8.2% of the transcriptomic profile represents the result of the perception of a token stimulus (long day) and the transduction of the signal to Fig. 2. Separation of the different phases of diapause development in resuming developmental potential. Additionally, the photoperi- C. costata larvae according to gradual changes in their transcriptional profile. odic termination phase is qualitatively different from the gradual The pie charts indicate quantitative (shown by the relative size of the pie) changes in the transcriptome associated with diapause mainte- and qualitative (colors of segments) aspects of the gradual change (down- nance: only 10 of all DE sequences (6.6% of 151) overlap be- regulation, left pie; up-regulation, right pie) in the transcriptomic profile during the development of third-instar larva under direct development or tween the maintenance (md2 vs. md1) and photoperiodic diapause-promoting conditions. The size of each pie is directly proportional termination (ppt vs. md1) phases. to the percentage of DE sequences detected (of a total 1,042 arrayed se- Cold acclimation followed by diapause cold termination. Cold termi- quences) for the respective treatment comparison (e.g., nd2 vs. nd1). The nation (ct) of diapause is a long process, typically requiring 2- to color of each segment codes for a gene-functional category (see key). Up to 3-mo-long exposure to low temperatures in different insects (23). five (or six) gene-functional categories with the highest percentage of DE se- However, the initial temperature step-down from 18 °C to 4 °C is quences are shown in a specific pie. For more details, see text and Dataset S1. also an important stimulus for another complex phenotypic change,

8534 | www.pnas.org/cgi/doi/10.1073/pnas.1707281114 Koštál et al. Downloaded by guest on September 25, 2021 cold acclimation (ca). The larvae of C. costata become cold accli- that are also characteristic of insect diapause (44–46) (for more mated (i.e., their freeze-tolerance significantly increases) within details, see Fig. S3). relatively short exposures to cold stimuli (27). The cold-acclimation The key molecular components of the 20HE, IS, and TOR phase in C. costata larvae was characterized by relatively complex signaling pathways are well conserved throughout evolution (43, changes in the transcriptomic profile: (ca vs. id4), 19.9% DE se- 46, 47), and a summary of these pathways presented in Fig. 3A quences (Fig. 2), and the gene categories involved in temperature documents that our custom array is well-represented with putative response and cryoprotection were logically those most affected C. costata gene probes of the 20HE, IS, and TOR pathways. (Dataset S1). We placed a number of sequences involved in the However, relatively few transcriptional rearrangements were ob- metabolism and transport of proline and trehalose into an arbitrary served in 20HE, IS, and TOR pathways during the diapause de- gene category that we termed “cryoprotection.” Proline and tre- velopment of C. costata larvae (Fig. 3B; for more details, see halose may play cryoprotective roles when accumulated in high Dataset S1, Excel sheet: 20HE, IS, TOR). Mechanistic interpre- amounts and distributed evenly throughout the larval body during tation of microarray results is difficult for general and specific cold acclimation (27, 28). reasons: The correspondence between mRNA abundance and the Next, we observed a relatively complex rearrangement of the activity of the protein product is generally limited; many elements transcriptome occurring in deeply diapausing larvae that were of the IS and TOR cascades are regulated via phosphorylation kept in constant darkness and at a low temperature, 4 °C (i.e., rather than by transcription; our study lacks resolution at the tissue conditions generally not expected to stimulate any change, i.e. level; and our study relies on the analysis of transcripts in C. costata cold termination). The relative proportion of DE sequences was that were detected as structural homologs of Drosophila mela- 5.4% for the (ct1 vs. ca) comparison and was 5.2% for the (ct2 vs. nogaster genes, but no functional validation is available for most of ct1) comparison (Dataset S1). Only 16 of all DE sequences them. Therefore, we provide only a brief description of observed (14.5% of 110) overlapped between the early (ct1 vs. ca) and late changes that, at this stage, are hypothetically linked to diapause (ct2 vs. ct1) phases of cold termination, possibly indicating that development (see SI Results: Extended Discussion Related to Fig. 3). the late phase of cold termination qualitatively differs from the Wnt signaling pathway. Cold termination of diapause is a process early phase. with an insufficiently understood physiological basis. During diapause termination, developmental potential is covertly re- Genetic Toolkit of Diapause. Since the publication of the pio- stored but need not be overtly realized in the form of resumption neering paper of Denlinger and coworkers (29), a number of of morphogenesis and development (23). In the laboratory, most studies have described differences in gene transcription linked to diapauses can be terminated rapidly after the diapausing animals insect diapause (20, 30–36). Indeed, some authors have attempted are treated with developmental hormones, hormone analogs, or to integrate these studies and to detect commonalities in the dia- some organic solvents (9). During the complex cascades follow- pause transcriptional profiles of different species, searching for the ing hormonal treatment, transcriptional changes were described hypothetical genetic toolkit of diapause (18, 19). Two genes have in the pupae of the flesh fly Sarcophaga crassipalpis (20, 48). been specifically suggested as potential general markers for dis- However, to understand the natural process of horotelic dia- tinguishing between diapause and direct development, phospho- pause termination in the field (or, in the laboratory, but without enolpyruvate carboxykinase (pepck), which is up-regulated in artificial manipulation of the endocrine system), we need to

diapause, and proliferating cell nuclear antigen (pcna), which is know the upstream events preceding the change in hormonal SCIENCES down-regulated in diapause (18). The up-regulation of the pepck signaling. In diapausing pupae of S. crassipalpis, for instance, the gene in the diapause context was explained as reflecting the tran- spontaneous increase in ultraspiracle expression during pro- APPLIED BIOLOGICAL sition between oxygen-rich and oxygen-poor environments, whereas longed exposure to cold might represent one of the upstream the down-regulation of pcna, again in the diapause context, was events that are directly linked to diapause termination (49). linked to the arrest of cell-cycle progression (18). In our study of Some other genes were identified as potential early regulators of C. costata, neither pepck (Seq29 and Seq58063) nor pcna (Seq570) transcripts showed responses consistent with the predictions of the genetic toolkit hypothesis. We have found 45 up-regulated and 34 down-regulated DE genes when comparing the early phase of diapause induction (id1) with the corresponding stage of nondiapause development (nd1), in which neither pecpk nor pcna was represented (Dataset S1, Excel sheet: id1 vs. nd1). However, the pepck sequence was significantly up-regulated during direct development (nd2 vs. nd1), 0.59 log2 fold change (FC) (Seq29), and was down-regulated during the cold-acclimation phase (ca vs. id4), −0.96 log2 FC (Seq29) and −1.42 log2 FC (Seq58063). The pcna sequence was significantly up-regulated during direct development (nd2 vs. nd1), 0.77 log2 FC (Seq570). Accordingly, we agree with Ragland et al. (20) that the evolutionary conservation/convergence of different diapause responses is unlikely to be observable at the level of transcriptional patterns of isolated genes. Rather, a genetic toolkit of diapause may be reflected in the activation/inhibition of Fig. 3. Cooperation of 20HE, IS, and TOR signaling pathways in the regu- whole signaling pathways regulating the hallmark diapause-linked lation of fly larva growth and development. (A) The elements of signaling physiological phenotypes. pathways recognized in D. melanogaster larvae and their putative homologs 20HE, IS, and TOR signaling pathways. Larval growth and direct de- (highlighted in colored rectangles) represented on our C. costata custom velopment are stimulated by the developmental hormone 20HE, microarray. The 20HE signaling cascade is in red, forkhead box O (FOXO) whereas larval diapause is classically explained by the silencing of pathway components are in green, TOR pathway components are in 20HE-mediated signaling (9). In addition, the silencing of the yellow, and the serine-threonine kinase LKB1-activated pathway is shown in brown. White rectangles with gray font indicate elements not represented insulin (IS) and target of rapamycin (TOR) signaling pathways on our custom microarray. (B) Results of microarray transcriptional analysis has often been implicated as an important regulator of insect – in C. costata. Only results in which statistically significant up- or down-reg- diapause (37 41). Active IS and TOR pathways promote cell ulations of signaling pathway components were observed are shown (the proliferation and organismal growth (42, 43), whereas the si- numbers in each box are log2-transformed average FCs in expression). For lencing of these pathways is linked to the suppression of growth/ more details, see SI Results: Extended Discussion Related to Fig. 3 and development and enhanced stress response, phenotypic changes Dataset S1, Excel sheet: 20HE, IS, TOR.

Koštál et al. PNAS | August 8, 2017 | vol. 114 | no. 32 | 8535 Downloaded by guest on September 25, 2021 diapause termination in the embryos of Bombyx mori (50, 51) and Staging of Diapause Development. Larvae of C. costata (Diptera: Droso- in the larvae of Wyeomyia smithii (34). In the pupae of Rhagoletis philidae) were cultured under conditions either promoting direct devel- pomonella, the elements of a calcium-dependent branch of the opment to the pupa and adult stages or inducing diapause in the fully Wnt signaling pathway were found to be moderately up-regulated grown third-instar larval stage (see Fig. S1 and Table S1 for a schematic in late diapause just before termination by cold (52). The Wnt overview of the experimental design). Direct development was promoted pathway is one of the short-range, organ-autonomous (paracrine) under a constant temperature of 18 °C and a long-day photoperiod (16 h signaling systems controlling final cell numbers in larval imaginal light:8 h darkness), whereas diapause was induced by a constant tem- discs and thereby controlling organ size in the adult (53). Because perature of 18 °C and a short-day photoperiod (12 h light:12 h darkness). the suppression of imaginal disc proliferation is a hallmark of Under these conditions, all individuals responded reliably to the photo- diapause in C. costata (54), the Wnt pathway is clearly an im- periodic signal (24, 26, 55). Larvae were sampled at different phases of portant candidate for diapause regulation. However, our custom their nondiapause or diapause development based on our earlier obser- array represents only nine genes putatively belonging to Wnt vations (23, 24, 26). pathway in C. costata, limiting any speculations about their role in diapause. Two of the Wnt pathway-linked transcripts [Seq59352, Analysis of Transcriptomic Patterns Linked to Diapause Development Using basket (JNK) and Seq81736, c-jun] were up-regulated during cold Custom Microarrays. Transcriptomic profiling was based on 1,042 candidate acclimation; this up-regulation seems counterintuitive, because sequences arbitrarily selected from a published Illumina RNA-seq database these genes are positive regulators of gene transcription and the that contains 21,327 putative mRNA transcripts of C. costata (ArrayExpress cell cycle, respectively. In contrast, the concomitant up-regulation accession E-MTAB-3620) (25). The selected genes broadly cover major of shaggy (Seq2621) during the maintenance of diapause and cold structures and processes known or suggested to be involved in insect dia- acclimation may inhibit the Wnt cascade and thus counteract the pause expression. The complete list of the analyzed sequences is given in stimulation of the cell-division cycle potentially caused by up- Dataset S1, Excel sheet: List. All other details on the processing of sampled regulated c-jun. Shaggy is a glycogen synthase kinase 3, GSK-3β, larvae, extraction of total RNA, conversion of RNA to cDNA and labeling it and is a key component of the β-catenin destruction complex. It with cyanine-3-dCTP (Cy3), selection of candidate sequences, designing and functions as a negative regulator in the canonical Wnt cascade synthesis of the probes, printing the custom microarrays, hybridization and (53). It remains currently unknown whether the cold-induced scanning of arrays, quantification of fluorescence signal (Fig. S4), technical validation using qPCR (Table S2), and statistical analysis of results are spec- transcriptional up-regulation of some positive elements (such as ified in SI Materials and Methods. Differential gene-expression analysis be- basket and c-jun) of the Wnt pathway, which is counteracted by the tween consecutive time points was performed to assess the gradual up-regulation of negative elements (such as shaggy), is an early dynamics in transcriptomic patterns as the insects traversed diapause or component of diapause termination. A concomitant up-regulation nondiapause development. In addition, unconstrained PCA was used to re- in positive and negative elements during the cold-acclimation veal clustering effects in global transcriptomic profiles of all treatments phase also occurs in the 20HE, IS, and TOR pathways (for more compared at a single statistical test. details, see SI Results: Extended Discussion Related to Fig. 3). ACKNOWLEDGMENTS. We thank Bettina Fischer (Cambridge Systems Bi- Materials and Methods ology Centre, University of Cambridge) for assistance during the production For a detailed description of materials and methods, see SI Materials and of custom microarrays. This study was supported by Czech Science Founda- Methods. tion Grant 13-01057S.

1. Lees AD (1956) The physiology and biochemistry of diapause. Annu Rev Entomol 19. Yocum GD, et al. (2015) Key molecular processes of the diapause to post-diapause 1:1–16. quiescence transition in the alfalfa leaf cutting bee Megachile rotundata identified by 2. Hodek I (1983) Role of environmental factors and endogenous mechanisms in the comparative transcriptome analysis. Physiol Entomol 40:103–112. seasonality of reproduction in insects diapausing as adults. Diapause and Life Cycle 20. Ragland GJ, Denlinger DL, Hahn DA (2010) Mechanisms of suspended animation are Strategies in Insects, eds Brown VK, Hodek I (Dr W Junk Publishers, The Hague), pp revealed by transcript profiling of diapause in the flesh fly. Proc Natl Acad Sci USA 9–33. 107:14909–14914. 3. Tauber MJ, Tauber CA, Masaki S (1986) Seasonal Adaptations of Insects (Oxford Univ 21. Mansingh A (1971) Physiological classification of dormancies in insects. Can Entomol Press, Oxford, UK), p 411. 103:983–1009. 4. Danks HV (1987) Insect dormancy: An ecological perspective. Terrestrial Arthropods, 22. Tauber MJ, Tauber CA (1976) Insect seasonality: Diapause maintenance, termination, Monograph series No. 1 (National Museum of Natural Sciences, Ottawa) p 439. and postdiapause development. Annu Rev Entomol 21:81–107. 5. Saunders DS (2002) Insect Clocks, eds Steel CGH, Vafopoulou X, Lewis RD (Elsevier, 23. Kostál V (2006) Eco-physiological phases of insect diapause. J Insect Physiol 52: Amsterdam), p 576. 113–127. 6. Denlinger DL (2008) Why study diapause? Entomol Res 38:1–9. 24. Koštál V, Mollaei M, Schöttner K (2016a) Diapause induction as an interplay between 7. Forrest JRK (2016) Complex responses of insect phenology to climate change. Curr seasonal token stimuli, and modifying and directly limiting factors: Hibernation in Opin Insect Sci 17:49–54. Chymomyza costata. Physiol Entomol 41:344–357. 8. Williams CN (1946) Physiology of insect diapause; the role of the brain in the pro- 25. Poupardin R, et al. (2015) Early transcriptional events linked to induction of diapause duction and termination of pupal dormancy in the giant silkworm, Platysamia ce- revealed by RNAseq in larvae of drosophilid fly, Chymomyza costata. BMC Genomics cropia. Biol Bull 90:234–243. 16:720. 9. Denlinger DL, Yocum GD, Rinehart JP (2012) Hormonal control of diapause. Insect 26. Koštál V, Shimada K, Hayakawa Y (2000) Induction and development of winter larval Endocrinology, ed Gilbert LI (Elsevier, Amsterdam), pp 430–463. diapause in a drosophilid fly, Chymomyza costata. J Insect Physiol 46:417–428. 10. Tauber CA, Tauber MJ (1981) Insect seasonal cycles: Genetics and evolution. Annu Rev 27. Kostál V, Zahradnícková H, Simek P (2011) Hyperprolinemic larvae of the drosophilid Ecol Syst 12:281–308. fly, Chymomyza costata, survive cryopreservation in liquid nitrogen. Proc Natl Acad 11. Bradshaw WE, Holzapfel CM (2001) Phenotypic evolution and the genetic architec- Sci USA 108:13041–13046. ture underlying photoperiodic time measurement. J Insect Physiol 47:809–820. 28. Koštál V, Korbelová J, Poupardin R, Moos M, Simek P (2016b) Arginine and proline 12. Emerson KJ, Bradshaw WE, Holzapfel CM (2009) Complications of complexity: In- applied as food additives stimulate high freeze tolerance in larvae of Drosophila tegrating environmental, genetic and hormonal control of insect diapause. Trends melanogaster. J Exp Biol 219:2358–2367. Genet 25:217–225. 29. Denlinger DL, et al. (1995) Diapause-specific gene expression. Molecular Mechanisms of 13. Nylin S (2013) Induction of diapause and seasonal morphs in butterflies and other Insect Metamorphosis and Diapause (Industrial Publishing & Consulting Inc., Tokyo), pp insects: Knowns, unknowns and the challenge of integration. Physiol Entomol 38: 289–297. 96–104. 30. Hayward SAL, Pavlides SC, Tammariello SP, Rinehart JP, Denlinger DL (2005) Temporal 14. Masaki S (1980) Summer diapause. Annu Rev Entomol 25:1–25. expression patterns of diapause-associated genes in flesh fly pupae from the onset of 15. Denlinger DL (1986) Dormancy in tropical insects. Annu Rev Entomol 31:239–264. diapause through post-diapause quiescence. J Insect Physiol 51:631–640. 16. Wolda H (1988) Insect seasonality. Why? Annu Rev Ecol Syst 19:1–18. 31. Robich RM, Denlinger DL (2005) Diapause in the mosquito Culex pipiens evokes a 17. MacRae TH (2010) Gene expression, metabolic regulation and stress tolerance during metabolic switch from blood feeding to sugar gluttony. Proc Natl Acad Sci USA 102: diapause. Cell Mol Life Sci 67:2405–2424. 15912–15917. 18. Poelchau MF, Reynolds JA, Elsik CG, Denlinger DL, Armbruster PA (2013) Deep se- 32. Reynolds JA, Hand SC (2009) Embryonic diapause highlighted by differential ex- quencing reveals complex mechanisms of diapause preparation in the invasive mos- pression of mRNAs for ecdysteroidogenesis, transcription and lipid sparing in the quito, Aedes albopictus. Proc Royal Soc B-Biol Sci 280:20130143. cricket Allonemobius socius. J Exp Biol 212:2075–2084.

8536 | www.pnas.org/cgi/doi/10.1073/pnas.1707281114 Koštál et al. Downloaded by guest on September 25, 2021 33. Urbanski JM, Aruda A, Armbruster P (2010) A transcriptional element of the diapause 61. Pfaffl MW (2001) A new mathematical model for relative quantification in real-time program in the Asian tiger mosquito, Aedes albopictus, identified by suppressive RT-PCR. Nucleic Acids Res 29:e45. subtractive hybridization. J Insect Physiol 56:1147–1154. 62. Petryk A, et al. (2003) Shade is the Drosophila P450 enzyme that mediates the hy- 34. Emerson KJ, Bradshaw WE, Holzapfel CM (2010) Microarrays reveal early transcrip- droxylation of ecdysone to the steroid insect molting hormone 20-hydroxyecdysone. tional events during the termination of larval diapause in natural populations of the Proc Natl Acad Sci USA 100:13773–13778. mosquito, Wyeomyia smithii. PLoS One 5:e9574. 63. Hall BL, Thummel CS (1998) The RXR homolog ultraspiracle is an essential component 35. Kankare M, Salminen T, Laiho A, Vesala L, Hoikkala A (2010) Changes in gene ex- of the Drosophila ecdysone receptor. Development 125:4709–4717. pression linked with adult reproductive diapause in a northern malt fly species: A 64. Hill RJ, Billas IM, Bonneton F, Graham LD, Lawrence MC (2013) Ecdysone receptors: candidate gene microarray study. BMC Ecol 10:3. From the Ashburner model to structural biology. Annu Rev Entomol 58:251–271. 36. Meyers PJ, et al. (2016) Divergence of the diapause transcriptome in apple maggot 65. Thummel CS (2001) Molecular mechanisms of developmental timing in C. elegans and flies: Winter regulation and post-winter transcriptional repression. J Exp Biol 219: Drosophila. Dev Cell 1:453–465. 2613–2622. 66. Edgar BA (2006) How flies get their size: Genetics meets physiology. Nat Rev Genet 7: 37. Tatar M, et al. (2001) A mutant Drosophila insulin receptor homolog that extends life- 907–916. span and impairs neuroendocrine function. Science 292:107–110. 67. Koyama T, Mendes CC, Mirth CK (2013) Mechanisms regulating nutrition-dependent 38. Williams KD, et al. (2006) Natural variation in Drosophila melanogaster diapause due developmental plasticity through organ-specific effects in insects. Front Physiol 4:263. to the insulin-regulated PI3-kinase. Proc Natl Acad Sci USA 103:15911–15915. 68. Gao X, Neufeld TP, Pan D (2000) Drosophila PTEN regulates cell growth and pro- 39. Sim C, Denlinger DL (2008) Insulin signaling and FOXO regulate the overwintering liferation through PI3K-dependent and -independent pathways. Dev Biol 221: – diapause of the mosquito Culex pipiens. Proc Natl Acad Sci USA 105:6777–6781. 404 418. 40. Sim C, Denlinger DL (2013) Insulin signaling and the regulation of insect diapause. 69. Miron M, Sonenberg N (2001) Regulation of translation via TOR signaling: Insights – Front Physiol 4:189. from Drosophila melanogaster. J Nutr 131:2988S 2993S. 41. Sim C, Kang DS, Kim S, Bai X, Denlinger DL (2015) Identification of FOXO targets that 70. Lee JH, et al. (2010) Sestrin as a feedback inhibitor of TOR that prevents age-related – generate diverse features of the diapause phenotype in the mosquito Culex pipiens. pathologies. Science 327:1223 1228. Proc Natl Acad Sci USA 112:3811–3816. 71. Gao X, Pan D (2001) TSC1 and TSC2 tumor suppressors antagonize insulin signaling in – 42. Grönke S, Clarke D-F, Broughton S, Andrews TD, Partridge L (2010) Molecular evo- cell growth. Genes Dev 15:1383 1392. lution and functional characterization of Drosophila insulin-like peptides. PLoS Genet 72. Garami A, et al. (2003) Insulin activation of Rheb, a mediator of mTOR/S6K/4E-BP – 6:e1000857. signaling, is inhibited by TSC1 and 2. Mol Cell 11:1457 1466. 43. Das R, Dobens LL (2015) Conservation of gene and tissue networks regulating insulin 73. Hietakangas V, Cohen SM (2007) Re-evaluating AKT regulation: Role of TOR complex – signalling in flies and vertebrates. Biochem Soc Trans 43:1057–1062. 2 in tissue growth. Genes Dev 21:632 637. 44. Sim C, Denlinger DL (2011) Catalase and superoxide dismutase-2 enhance survival and 74. Grewal SS (2009) Insulin/TOR signaling in growth and homeostasis: a view from the fly world. Int J Biochem Cell Biol 41:1006–1010. protect ovaries during overwintering diapause in the mosquito Culex pipiens. J Insect 75. Teleman AA (2009) Molecular mechanisms of metabolic regulation by insulin in Physiol 57:628–634. Drosophila. Biochem J 425:13–26. 45. Storz P (2011) Forkhead homeobox type O transcription factors in the responses to 76. Riddiford LM (1993) Hormones and Drosophila development. The Development of oxidative stress. Antioxid Redox Signal 14:593–605. Drosophila melanogaster, eds Bate M, Martinez AA (Cold Spring Harbor Lab Press, 46. Martins R, Lithgow GJ, Link W (2016) Long live FOXO: Unraveling the role of FOXO Cold Spring Harbor, NY), pp 899–939. proteins in aging and longevity. Aging Cell 15:196–207. 77. Doctor JS, Fristrom JW (1985) Macromolecular changes in imaginal discs during 47. Hay N (2011) Interplay between FOXO, TOR, and Akt. Biochim Biophys Acta 1813: postembryonic development. Comprehensive Insect Physiology, Biochemistry and 1965–1970. Pharmacology, eds Kerkut GA, Gilbert LI (Pergamon, Oxford, UK), pp 201–238. 48. Fujiwara Y, Denlinger DL (2007) High temperature and hexane break pupal diapause 78. Koyama T, Iwami M, Sakurai S (2004) Ecdysteroid control of cell cycle and cellular in the flesh fly, Sarcophaga crassipalpis, by activating ERK/MAPK. J Insect Physiol 53: commitment in insect wing imaginal discs. Mol Cell Endocrinol 213:155–166. 1276–1282. 79. Zitnan D, Sehnal F, Bryant PJ (1993) Neurons producing specific neuropeptides in the 49. Rinehart JP, Cikra-Ireland RA, Flannagan RD, Denlinger DL (2001) Expression of ec- central nervous system of normal and pupariation-delayed Drosophila. Dev Biol 156: dysone receptor in unaffected by pupal diapause in the flesh fly, Sarcophaga crassi- 117–135.

palpis, while its dimerization partner, USP, is downregulated. J Insect Physiol 47: SCIENCES 80. Harvie PD, Filippova M, Bryant PJ (1998) Genes expressed in the ring gland, the major 915–921. endocrine organ of Drosophila melanogaster. Genetics 149:217–231.

50. Niimi T, Yamashita O, Yaginuma T (1993) A cold-inducible Bombyx gene encoding a APPLIED BIOLOGICAL 81. Richard DS, et al. (1989) Juvenile hormone bisepoxide biosynthesis in vitro by the ring protein similar to mammalian sorbitol dehydrogenase. Yolk nuclei-dependent gene gland of Drosophila melanogaster: A putative juvenile hormone in the higher Dip- – expression in diapause eggs. Eur J Biochem 213:1125 1131. tera. Proc Natl Acad Sci USA 86:1421–1425. 51. Ti X, et al. (2004) Demarcation of diapause development by cold and its relation to 82. Lee G, Park JH (2004) Hemolymph sugar homeostasis and starvation-induced hyper- time-interval activation of TIME-ATPase in eggs of the silkworm, Bombyx mori. activity affected by genetic manipulations of the adipokinetic hormone-encoding – J Insect Physiol 50:1053 1064. gene in Drosophila melanogaster. Genetics 167:311–323. 52. Ragland GJ, Egan SP, Feder JL, Berlocher SH, Hahn DA (2011) Developmental trajec- 83. Ikeya T, Galic M, Belawat P, Nairz K, Hafen E (2002) Nutrient-dependent expression of tories of gene expression reveal candidates for diapause termination: A key life-his- insulin-like peptides from neuroendocrine cells in the CNS contributes to growth tory transition in the apple maggot fly Rhagoletis pomonella. J Exp Biol 214: regulation in Drosophila. Curr Biol 12:1293–1300. – 3948 3959. 84. Colombani J, et al. (2005) Antagonistic actions of ecdysone and insulins determine 53. Jope RS, Johnson GVW (2004) The glamour and gloom of glycogen synthase kinase-3. final size in Drosophila. Science 310:667–670. – Trends Biochem Sci 29:95 102. 85. Layalle S, Arquier N, Léopold P (2008) The TOR pathway couples nutrition and de- 54. Kostál V, Simunková P, Kobelková A, Shimada K (2009) Cell cycle arrest as a hallmark velopmental timing in Drosophila. Dev Cell 15:568–577. of insect diapause: Changes in gene transcription during diapause induction in the 86. Colombani J, Andersen DS, Léopold P (2012) Secreted peptide Dilp8 coordinates drosophilid fly, Chymomyza costata. Insect Biochem Mol Biol 39:875–883. Drosophila tissue growth with developmental timing. Science 336:582–585. 55. Riihimaa AJ, Kimura MT (1988) A mutant strain of Chymomyza costata (Diptera, 87. Garelli A, Gontijo AM, Miguela V, Caparros E, Dominguez M (2012) Imaginal discs Drosophilidae) insensitive to diapause-inducing action of photoperiod. Physiol secrete insulin-like peptide 8 to mediate plasticity of growth and maturation. Science Entomol 13:441–445. 336:579–582. 56. Lakovaara S (1969) Malt as a culture medium for Drosophila species. Drosoph Inf Serv 88. Andersen DS, Colombani J, Léopold P (2013) Coordination of organ growth: Principles 44:128. and outstanding questions from the world of insects. Trends Cell Biol 23:336–344. 57. Rouillard JM, Zuker M, Gulari E (2003) OligoArray 2.0: Design of oligonucleotide 89. Bryant PJ (2001) Growth factors controlling imaginal disc growth in Drosophila. probes for DNA microarrays using a thermodynamic approach. Nucl Acid Res 31: Novartis Found Symp 237:182–194, discussion 194–202. 3057–3062. 90. Hansen IA, Attardo GM, Rodriguez SD, Drake LL (2014) Four-way regulation of 58. Dhami P, et al. (2010) Genomic approaches uncover increasing complexities in the mosquito yolk protein precursor genes by juvenile hormone-, ecdysone-, nutrient-, regulatory landscape at the human SCL (TAL1) locus. PLoS One 5:e9059. and insulin-like peptide signaling pathways. Front Physiol 5:103. 59. Ritchie ME, et al. (2015) limma powers differential expression analyses for RNA-se- 91. Belgacem YH, Martin JR (2007) Hmgcr in the corpus allatum controls sexual di- quencing and microarray studies. Nucleic Acids Res 43:e47. morphism of locomotor activity and body size via the insulin pathway in Drosophila. 60. Ponton F, Chapuis MP, Pernice M, Sword GA, Simpson SJ (2011) Evaluation of po- PLoS One 2:e187. tential reference genes for reverse transcription-qPCR studies of physiological re- 92. Tu MP, Yin CM, Tatar M (2005) Mutations in insulin signaling pathway alter juvenile sponses in Drosophila melanogaster. J Insect Physiol 57:840–850. hormone synthesis in Drosophila melanogaster. Gen Comp Endocrinol 142:347–356.

Koštál et al. PNAS | August 8, 2017 | vol. 114 | no. 32 | 8537 Downloaded by guest on September 25, 2021