Physiological study on the transgenerational timing mechanism Title in an ( Dissertation_全文 )

Author(s) Matsuda, Naoki

Citation 京都大学

Issue Date 2020-03-23

URL https://doi.org/10.14989/doctor.k22281

学位規則第9条第2項により要約公開; 許諾条件により本文 Right は2020-08-31に公開

Type Thesis or Dissertation

Textversion ETD

Kyoto University

Physiological study on the transgenerational timing

mechanism in an aphid

Naoki Matsuda

Graduate School of Science

Kyoto University

March 2020 Contents

General introduction ...... 2

Chapter 1: Physiological characteristics of the seasonal timer

Introduction ...... 6

Materials and methods ...... 7

Results ...... 11

Discussion ...... 12

Chapter 2: Effects of the seasonal timer on the transcriptomic changes

Introduction ...... 16

Materials and methods ...... 17

Results ...... 21

Discussion ...... 23

Chapter 3: Adaptive significance of the seasonal timer

Introduction ...... 29

Materials and methods ...... 31

Results ...... 33

Discussion ...... 34

General discussion ...... 39

Acknowledgments ...... 43

References ...... 44

Figures ...... 61

Tables ...... 73

1

General introduction

Insects have several types of endogenous timing systems for synchronizing their physiological, developmental and behavioral events with the specific environmental cycles (Saunders 2002; Numata et al. 2015). One of these systems is an oscillatory clock, including circadian (Pittendrigh 1954; Harker 1956), circatidal (Evans

1976; Satoh et al. 2008), circa(semi)lunar (Neumann 1966) and circannual clocks

(Blake 1958; Nisimura & Numata 2001). It is peculiar to these oscillatory clocks that behavioral rhythms persist and free-run under constant conditions without environmental signals. Among them, the circadian clock determining daily timings of behaviors has been studied most intensively (Pittendrigh 1960; Konopka & Benzer

1971), and now its cellular localization and molecular machinery are well-identified

(Tomioka & Matsumoto 2010). Photoperiodism is the other type of endogenous timing system and acts as a seasonal clock in (Tauber et al. 1986; Danks 1987). At higher latitude, photoperiod is a major environmental cue that determines seasonal traits including diapause (Nylin & Gotthard 1998; Nijhout 2003). Although the molecular mechanisms underlying the photoperiodism remains largely unknown, the circadian clock is involved in photoperiodic measurement because a light-dark cycle of 48 hours or longer is regarded as cycles of 24 hours with variable periods of light and darkness by insects (Bünning 1960; Vaz Nunes & Saunders 1999). Moreover, knock-down of clock genes by RNAi disrupts photoperiodism in several species, showing that the circadian clock is a component of the photoperiodic clock (Sakamoto et al. 2009;

Ikeno et al. 2010; Meuti et al. 2015; Mukai & Goto 2016). Therefore, the photoperiodic clock can also be regarded as one of the oscillator clocks. However, some insects have

2 non-oscillatory (hourglass-like) timer mechanisms (Saunders 2002), and they have received less attention than the oscillatory ones.

Most aphid species (: ) show seasonal life cycles consisting of both parthenogenetic and sexual generations (Moran 1992; Simon et al.

2002). From spring to summer, stem mothers, which are sexually produced and hatch from eggs, and their offspring reproduce parthenogenetically and viviparously more than ten generations. In autumn, sexual females and males are produced, and females lay overwintering eggs (Fig. 1). Marcovitch (1923) showed in the strawberry root aphid,

Aphis forbesi Weed, that sexual morph production is induced by short days, and this was the first reported photoperiodism in (Marcovitch 1923). One year later, however, Marcovitch (1924) also reported that sexual morphs do not appear during more than two months after the hatching of stem mothers of A. forbesi even under short days. Lees (1960) showed that the factor responsible for the suppression of sexual morph production in the vetch aphid, Megoura viciae Buckton, measures the total number of days over generations, not that of generations, from the hatching of stem mothers, and therefore he termed it as ‘an interval timer’. In this thesis, however, I call this mechanism a ‘seasonal timer’ to distinguish it from the term ‘interval timer’ used for a mechanism that measures various durations of time in other contexts (MacKay

1977; Nagao & Shimozawa 1987). In contrast to the oscillatory clocks, the seasonal timer is a unidirectional mechanism and is reset by sexual reproduction (Lees 1960;

Saunders 2002). Previous studies in the laboratory have shown the existence of the seasonal timer in several aphid species (Table 1). Except for Drepanosiphum platanoides Schrank in Drepanosiphinae and Eucallipterus tiilae Linnaeus in

Calaphidinae, all these species are in the subfamily Aphidinae, the largest subfamily of

3

Aphididae (approximately 2800/4700 species) (Blackman & Eastop 2007), and they are classified into the two tribes Aphidini and Macrosiphini (Kim et al. 2011; Choi et al.

2018). Possibly, most of species in Aphididae might have the seasonal timer, although it is unclear whether the seasonal timer has evolved once or many times. However, molecular mechanism of the seasonal timer is completely unknown during more than 90 years after the first report of phenomena in which the seasonal timer was involved (Tagu et al. 2005). Additionally, an adaptive significance of the seasonal timer is unclear.

The pea aphid, Acyrthosiphon pisum (Harris), is a suitable species for physiological study on the seasonal timer. Its annual life cycle consists of several parthenogenetic generations and a sexual generation in cold regions (holocyclic life cycle), although in warm regions the sexual generation is not necessary for overwintering and is often eliminated from the life cycle (anholocyclic life cycle). In

Japan, holocyclic populations are distributed mainly in Hokkaido, whereas anholocyclic populations prevail in Honshu to Kyushu (Kanbe & Akimoto 2009). The effects of environmental conditions on the reproductive polyphenism have been extensively studied in this species (Kenten 1955; Lamb & Pointing 1972; MacKay 1987), and its stem mothers can be obtained easily from sexual morphs produced in short-day conditions in the laboratory (Via 1992; Shingleton et al. 2003). Bonnemaison (1972) has shown that sexual morph production is suppressed over several generations from hatching of stem mothers, suggesting that A. pisum has a seasonal timer, although its detailed physiological characteristics are unclear. Moreover, the recent development of the next-generation sequencing and genome resources for this species has enabled transcriptomic approaches to reveal molecular mechanisms of the photoperiodic regulation of the reproductive polyphenism of (International Aphid Genomics

4

Consortium 2010; Le Trionnaire et al. 2013). These technologies can be applied to reveal genomewide transcriptional changes between aphids before and after expiration of the seasonal timer.

In this thesis, I aim to reveal effects of the seasonal timer on mechanisms underlying the reproductive polyphenism and on the seasonal life cycle in A. pisum. In

Chapter 1, I performed successive rearing experiments in the laboratory to examine whether the seasonal timer measures the number of days or generations from hatching.

Additionally, I investigated the effects of rearing temperatures and photoperiods on the duration of the seasonal timer. In Chapter 2, I performed RNA sequencing to identify genes that showed differential expression between before and after expiration of the seasonal timer. In Chapter 3, I reared aphids under natural photoperiods in spring and compared their reproductive patterns between before and after expiration of the seasonal timer to examine whether the seasonal timer suppresses unseasonal productions of sexual morphs.

5

Chapter 1

Physiological characteristics of the seasonal timer

Introduction

In several aphid species, it has been shown that the seasonal timer suppresses sexual morph production under short days. In the vetch aphid, Megoura viciae Buckton, the seasonal timer measures the number of days from hatching, but not the number of generations (Lees 1960). In many previous studies, however, it was unclear whether the number of days or the number of generations from hatching is important (Wilson 1938;

Dixon 1971, 1972; Lushai et al. 1996; Margaritopoulos et al. 2002; Campbell &

Tregidga 2006). Moreover, effects of environmental conditions on the duration of the seasonal timer have been reported only in a few species and fragmentally (Table 1).

Temperature affects the duration for which the seasonal timer can function in M. viciae, whereas rearing photoperiod has no effect on the duration of the seasonal timer in Aphis rubicola Oestlund (Lees 1960; Brodel & Schaefers 1979). Although the physiological and molecular mechanisms suppressing sexual morph production remain unknown

(Tagu et al. 2005), determining the physiological characteristics of the seasonal timer might provide insights into these mechanisms.

Bonnemaison (1972) showed that Acyrthosiphon pisum (Harris) has a seasonal timer by rearing stem mothers and their descendants under several conditions inducing sexual morph production. Sexual morphs did not appear in early generations derived from stem mothers. In the study by Bonnemaison (1972), however, it was not

6 determined whether the duration of the seasonal timer depends on the number of days or the number of successive generations from hatching, or whether the temperature or photoperiod under which aphids were reared has effects on the duration of the seasonal timer.

The present study aimed to examine the physiological characteristics of the seasonal timer. First, I reared lineages of A. pisum that consisted of progenies born at different dates under a constant temperature and photoperiod to determine whether the seasonal timer measures the number of days or the number of generations. Then I reared aphids under different temperatures and photoperiods to examine the effects of the environmental factors on the seasonal timer.

Materials and methods

Insects

The ApL clone of A. pisum (referred to as Sap05Ms2 in Kanbe and Akimoto

2009; named in Ishikawa et al. 2012) was used for experiments. This clone was collected in 2005 from Medicago sativa Linnaeus in Sapporo, Japan and has been maintained in the laboratory. Parthenogenetic females of this clone reared under short days and their progenies produce a small proportion and a large proportion of sexual morphs, respectively (Ishikawa et al. 2012). Parthenogenetic aphid cultures were maintained in plastic tubes (25 mm diameter, 100 mm depth) in which a seedling of the broad bean, Vicia faba L., was maintained by wrapping a bean in wet paper under 16 h light and 8 h darkness (long-day conditions, LD) at 20.0 ± 1.0°C. The light intensity in

7 the photophase produced by a daylight fluorescent lamp was 0.3–2.8 W/m2.

To obtain stem mothers, oviparous females and males were produced under short days. One male and five oviparous females were placed together in a plastic case

(40.0 × 40.0 × 10.0 mm) containing a leaf of broad bean under 10 h light and 14 h darkness (short-day conditions, SD) at 15.0 ± 1.0°C. These females laid eggs on the leaf or on the ceiling of the case after approximately 10 days. These eggs were used for experiments, except for those that did not darken within a few days and were judged as unfertilized (Miura et al. 2003). Fertilized eggs were kept under SD at 15.0 ± 1.0°C for

7-14 days, and then placed under constant darkness at 4.0 ± 1.0°C for approximately three months to ensure embryogenesis (Via 1992; Shingleton et al. 2003). These eggs were transferred to SD at 15°C and hatched within 10 days.

Experimental design

Stem mothers were divided into three groups on the day of hatching and reared under SD at 15°C, SD at 20°C, or LD at 20°C. A maximum of 15 nymphs were reared on a single bean seedling. Each stem mother was transferred to a new seedling on the day of adult emergence and started to reproduce in a few days.

To determine whether the seasonal timer measures the number of days or the number of generations from the hatching, the ability to produce sexual morphs was tested using aphids of two lineages reared in different ways. Progenies of the stem mothers reared under SD at 15°C were kept under the same conditions. After adult emergence, the number of asexual or sexual morphs was counted. A few nymphs that were born within three days of the larviposition onset (older-sisters) were transferred to

8 new seedlings and used to start the ‘older-sister SD lineage’, whereas a few nymphs that were born approximately ten days after the larviposition onset (younger-sisters) were also transferred to new seedlings and used to start the ‘younger-sister SD lineage’ (Fig.

2a). Older- and younger-sisters of each generation were used to produce the subsequent generation in the older- and younger-sister SD lineages, respectively. In this way, aphids of the older- and younger-sister SD lineages were reared over a maximum of ten successive generations.

To examine the effect of temperature on the duration of the seasonal timer, the ability to produce sexual morphs was tested using aphids of lineages reared at different temperatures. Descendants of the stem mothers reared under SD at 20°C were kept under the same conditions, and the older- and younger-sister SD lineages were reared in the same way as those at 15°C (Fig. 2a). In addition to clones derived from overwintered eggs, the stock culture (ApL) that had been maintained parthenogenetically since 2005 was used as a control for testing the ability to produce sexual morphs. Nymphs of the parental clone were transferred from LD to SD on the day of their birth and reared in the same way as the older-sister SD lineage at 15°C and

20°C.

To examine the effect of photoperiod on the duration of the seasonal timer, the ability to produce sexual morphs was tested using aphids reared under different photoperiods. The stem mothers reared under LD at 20°C were used to start ‘the older- sister LD lineage’ in the same way as that under SD (Fig. 2b). Progenies of these stem mothers were divided into two groups on the day of their birth and transferred to SD or kept under LD. The group reared under SD was used to produce the subsequent generation in which the ability to produce sexual morphs was tested. The group reared

9 under LD was used to produce the subsequent generation which were also transferred to

SD or kept under LD. In this way, aphids of the older-sister LD lineage were reared over a maximum of eight generations.

The ability of a lineage to produce sexual morphs was examined by using one to four asexual females from each generation for all lineages. After the larviposition onset, each mother was transferred to a new seedling every 2-3 days and a maximum of

15 nymphs were reared on a seedling. Females with embryos with compound eyes were judged to be parthenogenetic and those with eggs with no compound eye were judged to be sexual. The number of asexual or sexual progenies that each mother produced from the larviposition onset to its death was counted.

Statistical analysis

A generalized linear mixed model (GLMM) with a binomial distribution and logit link was applied to evaluate the effects of the number of days from hatching to the birth of mothers, rearing temperature, sister-lineages and photoperiod on the proportion of sexual progenies, using the package glmmML in R 3.3.2 (R Core Team 2018). The random effects of individual mothers were taken into account to avoid overdispersion.

The effects of days, temperature and sister-lineages were evaluated using samples of the second and subsequent generations of the older- and younger-sister SD lineages at 15°C and 20°C. The effect of photoperiod was evaluated using the samples of the third and subsequent generations of the older-sister lineages under SD and LD at 20°C.

10

Results

The stem mothers produced only parthenogenetic females both at 15°C and

20°C even under SD (Fig. 3). At 15°C, the second generation also produced only parthenogenetic females, and the first sexual progenies were produced by mothers of the third generation born 51 and 57 days after hatching in the older- and in the younger- sister SD lineages, respectively (data not shown). In the older-sister SD lineage at 20°C, the first sexual progenies were produced by a mother of the third generation born 32 days after hatching (data not shown). In the younger-sister SD lineage at 20°C, however, mothers of the second generation born 19 days after hatching produced a small proportion of sexual morphs (data not shown). Although there was no significant difference in the proportion of sexual progenies between stem mothers in the older- sister SD lineage and parthenogenetic females of the stock culture in the first generation, the latter produced a small proportion of sexual morphs under SD (Fig. 3).

In the second generation under SD, however, there were significant differences in the proportions of sexual progenies between mothers in the older-sister SD lineage and parthenogenetic females of the stock culture (Fig. 3).

The proportions of sexual progenies gradually increased over successive generations for approximately 100 days after hatching in both the older- and the younger-sister SD lineages (Fig. 4a-d). On the same day from hatching, the proportions of sexual progenies were larger at 20°C (Fig. 4c, d) than those at 15°C (Fig. 4a, b) both in the older- and the younger-sister SD lineages. The number of days from hatching to birth of mothers and temperature had a significant effect on the proportion of sexual progenies (Table 2a).

11

The number of days from hatching to the birth of mothers that produced 50% sexual progenies was calculated from the results of GLMM analysis (Table 3). The critical number of days for sexual morph production was not different between the older- and the younger-sister SD lineages, although the number of generations at the critical days was larger in the older-sister SD lineage than in the younger-sister SD lineage. Sister-lineage had no significant effect on the proportion of sexual progenies

(Table 2a). The critical number of days for sexual morph production at 15°C was 1.25–

1.39 times that at 20°C, while the number of days per generation at 15°C was approximately 1.7 times that those at 20°C (Table 3).

The proportion of sexual progenies also gradually increased over generations in the older-sister LD lineage (Fig. 4e). Mothers of the third generation first produced sexual morphs. According to the results of GLMM analysis, the photoperiod under which aphids were reared had marginally not significant effect on the proportion of sexual progenies (Table 2b). The critical number of days for sexual morph production under LD was 0.90–0.94 times that under SD at 20°C (Table 3).

Discussion

I showed here that in A. pisum, sexual morph production by stem mothers and by the following generations was inhibited during a few months even under short days.

Such strong suppression of sexual morph production has also been shown in other aphids, e.g., A. forbesi in southern North America, Aphis chloris Koch and M. viciae collected from England, and Myzus persicae (Sulzer) in Greece (Marcovitch 1924;

Wilson 1938; Lees 1960; Margaritopoulos et al. 2002). Lees (1960) suggested that a

12 seasonal timer for sexual morph production is adaptive to avoid producing sexual morphs under short days in early spring. On the other hand, it has also been reported that stem mothers produce sexual morphs under very short days in Drepanosiphum platanoides (Schrank) and Eucallipterus tiliae L. in Scotland (Dixon 1971, 1972). Stem mothers of Acyrthosiphon brevicorne Hille Ris Lambers and Acyrthosiphon svalbardicum Heikinheimo living in high arctic regions, where the season for growth and reproduction is shorter than in temperate regions, produce sexual morphs in natural conditions (Strathdee et al. 1993; Strathdee & Bale 1996). Thus, the strength and duration of the seasonal timer should be adaptive to local climates.

The present results show that in A. pisum, the seasonal timer measured the number of days from hatching, not the number of generations, in accord with the reported findings in M. viciae and A. rubicola (Lees 1960; Brodel & Schaefers 1979).

The number of generations from spring to autumn is variable among sisters in A. pisum because parthenogenetic females continue to reproduce for more than 20 days, which is longer than the time required for nymphal development (MacKay 1987; Ishikawa et al.

2012). Thus, if the seasonal timer measured the number of generations, the response to short days might be different between the younger- and older-lineages in autumn. It is therefore reasonable for aphids to measure not the number of generations but the number of days to produce sexual morphs in an appropriate season.

Lees (1960) showed that the duration of the seasonal timer depends on temperature, and the present results supported this conclusion. Although the effect of temperature on the duration of the seasonal timer was not equal to that on the growth rate (Table 3), high temperature (20°C) advanced the timing of sexual morph production. In M. viciae and A. pisum, however, higher temperature also suppresses

13 sexual morph production: At 25°C, sexual morphs are not produced even under 8 h light and 16 h darkness (Kenten 1955; Lees 1959). Considering these two findings, it is possible that high temperature has contrasting effects on the sexual morph production in parthenogenetic females before and after the seasonal timer has expired (Lees, 1960).

Although sexual morphs were produced only under short days, the seasonal timer also measured the number of days under long days in A. pisum: The ability to produce sexual morphs was recovered at a similar rate irrespective of photoperiod (Fig.

4c, e). In the temperate zone, sexual morph production would be suppressed under long days in summer. A seasonal timer that expires by autumn would be adaptive for responding appropriately to shortened days. The present findings were similar to those of previous studies in M. viciae and A. rubicola: Mothers that had been reared successively under long days and transferred to short days produced many oviparous females (Lees 1960; Brodel & Schaefers 1979). These results altogether showed that the sexual polyphenism in aphids is regulated by two mechanisms, and one of them depends on photoperiod whereas the other does not.

It has been reported that the maternal experience of diapause induced by short days suppresses the response to short days in the progeny generation in Diptera and

Hymenoptera (Henrich & Denlinger 1982; Rockey et al. 1989; Reznik & Samartsev

2015). In Trichogramma spp. parasitoid wasps, the response to short days is suppressed over several generations after termination of a previously experienced diapause (Reznik

& Samartsev 2015). These mechanisms also seem to be adaptive to avoid responding to short days and producing nondiapause progenies in early spring. However, Reznik and

Samartsev (2015) did not distinguish between the number of days and the number of generations.

14

Lees (1960) suggested that the seasonal timer would not result from a cytoplasmic substance which is diluted by cell division until it falls below a critical threshold, because the mechanism of the seasonal timer should resist many dilutions through several generations. Epigenetic mechanisms that alter the expression of DNA, however, can be passed on to one or more successive generations (Ho & Burggren

2010; Sgrò et al. 2016). Although the role of epigenetic modification in reproductive polyphenism has not been studied, methylation levels in juvenile hormone associated genes are higher in winged asexual females than wingless asexual females in A. pisum

(Walsh et al. 2010). In the flesh fly Sarcophaga bullata Parker, the experience of diapause in the maternal generation changes histone acetylation in the progeny generation under the same short-day conditions (Reynolds et al. 2016). Although the physiological and molecular mechanisms of the seasonal timer in aphids are completely unknown, the present study in a model aphid species provides a basis for further investigations on the mechanisms of its seasonal timer, which might also be regulated epigenetically.

15

Chapter 2

Effects of the seasonal timer on the transcriptomic changes

Introduction

Acyrthosiphon pisum (Harris) shows the typical holocyclic life cycle and photoperiodic regulation of the reproductive polyphenism (Kenten 1955; Lamb &

Pointing 1972), and now is an emerging model species for studying polyphenism

(International Aphid Genomics Consortium 2010; Ogawa & Miura 2014). In this species, photoperiodic responses at the transcriptional level have been investigated for opsin genes (Collantes-Alegre et al. 2018), circadian clock genes (Cortés et al. 2010;

Barberà et al. 2017; Barberà et al. 2018), insulin and juvenile-hormone (JH) pathway genes (Ishikawa et al. 2012; Barberà et al. 2019) and a genomewide scale (Cortés et al.

2008; Le Trionnaire et al. 2009). In addition to these transcriptional differences, differential gene splicing might also be involved in the regulation of reproductive phenotypes under different photoperiods (Grantham & Brisson 2018). In stem mothers and their several subsequent generations, the seasonal timer suppresses the sexual morph production under short days in a number of species of aphids, including A. pisum

(Marcovitch 1924; Lees 1960; Bonnemaison 1972). Although Marcovitch (1924) reported a phenomenon in which the seasonal timer is involved more than 90 years ago, molecular mechanisms of its expiration and transfer have not yet been investigated.

Recently, high-throughput RNA sequencing has revealed diverse transcriptional changes underlying induction of diapause and determination of morph, caste and sex in

16 various insects (Huang et al. 2015; Poupardin et al. 2015; Vellichirammal et al. 2016;

Yaguchi et al. 2019). Moreover, this approach may reveal transcriptional differences between insects that show or do not show a maternal effect. The transcript information will be useful for exploring the molecular machinery of photoperiodic and maternal signals. Although many studies have focused on transcriptional differences underlying switches between asexual and sexual morph production after expiration of the seasonal timer in A. pisum (Ishikawa et al. 2012; Le Trionnaire et al. 2013), it remains unclear whether aphids with an operative seasonal timer show different gene expression patterns between long and short days. Therefore, comparisons of transcripts in relation to photoperiod and the seasonal timer are necessary to reveal the relative influence of these two factors on the switch between parthenogenetic and sexual morph productions.

The present study aims to reveal the transcriptomic changes that is regulated by the seasonal timer and photoperiod in the reproductive morph determination of A. pisum. I sequenced transcripts and compared their gene expression levels between groups of aphids that had an operative and an expired seasonal timer under long-day and short-day conditions. Based on functional annotations, I discuss possible molecular mechanisms underlying the reproductive polyphenism regulated by photoperiod and the seasonal timer.

Materials and methods

Insects

The ApL clone of A. pisum was used for experiments. The parthenogenetic

17 aphid cultures were reared in plastic tubes (25 mm diameter, 100 mm depth) in which a seedling of the broad bean Vicia faba L. was maintained by wrapping a bean in wet paper under long-day conditions (16 h light and 8 h dark) at 20.0 ± 1.0°C.

To obtain stem mothers, sexual females and males were produced under short- day conditions (10 h light and 14 h dark) at 15.0 ± 1.0°C, and then fertilized eggs were obtained as described in Chapter 1. Fertilized eggs were incubated for approximately 7 days under the same conditions and then were placed under constant darkness at 4.0 ±

1.0°C for more than 3 months to ensure embryogenesis (Via 1992). These eggs were allowed to hatch under short-day conditions at 15°C. After hatching, these nymphs

(stem mothers) were reared under long-day conditions at 20°C. After the onset of larviposition, their progenies with an operative seasonal timer were divided into two groups as first instar nymphs: one group was reared under long-day, and the other short- day, conditions at 20°C. After their adult emergence, these two aphid groups were termed “OL”, an abbreviation of the operative seasonal timer under long-day conditions, and OS, respectively (Fig. 5).

In the stock culture, in the same way, parthenogenetic females with an expired seasonal timer were divided into two groups as first instar nymphs and reared under long-day or short-day conditions at 20°C. After their adult emergence, these two aphid groups were termed EL and ES, respectively (Fig. 5).

RNA sequencing and differential expression analyses

Heads of adult A. pisum were used for RNA sequencing, because photoperiods are sensed and integrated there (Steel & Lees 1977). These heads were collected from

18 the four sample groups within 7 days after adult emergence at around 7 h after light-on, and pooled in TRIzol Reagent (Thermo Fisher Scientific, Waltham, MA, USA) at -20°C until further use. One sample consisted of 10 aphid heads, and two biological replicates were collected in each sample group (OL, OS, EL and ES). Total RNA was extracted from these pooled samples using TRIzol Reagent according to the manufacturer's instructions. RNA integrity was checked using a NanoDrop spectrophotometer (Thermo

Fisher Scientific) and an Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara,

CA, USA). The extracted RNA was prepared as tagged cDNA libraries for each sample using a TruSeq Stranded Total RNA Library Prep Kit (Illumina, San Diego, CA, USA).

The cDNA libraries were sequenced using a HiSeq 2500 sequencer (Illumina) with the paired-end method and a read length of 100 bp. All reads have been deposited in the

DDBJ Sequence Read Archive (SRA) database under accession number DRA009406.

The sequencing data were uploaded to the Galaxy web platform and were analyzed on the server at usegalaxy.org (Goecks et al. 2010). Adapter sequences were trimmed, and then reads shorter than 50 bp were discarded with the Trim Galore! script

(https://www.bioinformatics.babraham.ac.uk/projects/trim_galore/). These cleaned reads were mapped against the whole genome assembly (v2.0) of A. pisum in the AphidBase genome database with HISAT2 (Legeai et al. 2010; Kim et al. 2015). The mapped reads were then counted with StringTie (Pertea et al. 2015), and genes with zero counts in seven or all of eight samples were excluded from further analyses. Thereafter, analyses were conducted in R 3.5.2 (R Core Team 2018). The transcript abundances were normalized using the trimmed mean of M-values method with the edgeR package

(Robinson et al. 2010; Robinson & Oshlack 2010). A generalized linear model was applied to examine the effects of photoperiods and the seasonal timer on the normalized

19 gene counts with the edgeR package (McCarthy et al. 2012). P-values from the likelihood ratio test were corrected with False Discovery Rate (FDR, Benjamini and

Hochberg, 1995) and genes that had FDR < 0.05 were judged to be differentially expressed (DE). Expression patterns of the DE genes were represented in a heatmap and dendrograms with the gplots package (https://cran.r- project.org/web/packages/gplots/index.html) after the transcript abundances had been transformed to Z-score. The branch lengths of the dendrograms were calculated using

Ward’s method (Ward 1963).

Functional annotations and GO enrichment analyses

Gene Ontology (GO) enrichment analyses were performed with the OmicsBox

1.1 software (BioBam Bioinformatics, Valencia, Spain) with all expressed genes as the reference set and the DE genes as the test set. GO annotations against gene stable IDs were downloaded from BioMart (Smedley et al. 2009), and fine-grained terms were omitted from GO term lists with the GO slim tool. Enriched GO terms were assessed using Fisher’s exact test. Gene descriptions of the enriched GO terms were searched with the Aphidbase genome database (Legeai et al. 2010). For genes for which gene descriptions were uncharacterized, homologs of Drosophila melanogaster genes were searched with BLASTP in the FlyBase 2.0 genome database (Thurmond et al. 2019).

20

Results

RNA sequencing and differential expression analyses

As a result of RNA sequencing, 21-24 million read pairs per sample were obtained, 1.4-3.6% of the obtained reads were discarded, and 96-97% of cleaned reads were mapped on the genome of A. pisum. A total of 17,937 genes showed higher transcript abundances than zero read in two or more samples, and 500 genes (2.8%) showed significantly different expression in relation to photoperiod, the seasonal timer, or both (likelihood ratio test, FDR corrected P < 0.05). Hierarchical clustering of samples based on expression patterns of the DE genes resulted in four groups, coinciding with the sample groups (Fig. 6). The transcriptional differences between long-day and short-day conditions were smaller in samples with an operative seasonal timer (OL and OS) than in those with an expired seasonal timer (EL and ES). The DE genes were classified into four groups, A (311 genes), B (41 genes), C (76 genes) and D

(72 genes), according to their expression patterns. Group A was more highly expressed in ES, which had an expired seasonal timer and was reared under short days, than in the other samples. The expression levels of the DE gene in group B were up-regulated under short-day conditions (OS and ES), whereas those in group C were up-regulated when the seasonal timer had expired (EL and ES). Group D contained two types of genes: One type showed higher expression under long-day conditions (OL and EL), and the other type showed higher expression when the seasonal timer was operating (OL and

OS).

21

Functional annotations and GO enrichment analyses

Enriched GO terms were identified in groups A, C and D (Fisher’s exact test, P

< 0.05), while no enriched GO term was found in group B (Fisher’s exact test, P > 0.05)

(Table 4-6). In terms of the hierarchical structure, the GO terms of biological process in the group A were classified in four branches: chromosome organization, cellular localization, cell population proliferation and catabolic process (Fig. 7). The enrichment of the first three of these GO terms was significant (Fisher’s exact test, FDR corrected P

< 0.05), whereas that of the last one was not significant (Fisher’s exact test, FDR corrected P > 0.05) (Table 4). Moreover, the DE genes with the GO term “cell population proliferation” were associated with chromosome organization

(ACYPI003380, ACYPI007490 and ACYPI009726; see Table 7). Gene descriptions in the databases supported the appropriateness of functional annotations (Table 7).

Notably, more than 10 DE genes were involved in epigenetic modifications: histone methylation, histone acetylation and chromatin remodeling. Groups C and D also contained genes associated with chromosome organization, although this GO term was not significantly enriched there (Fisher’s exact test, FDR corrected P > 0.05). The most highly enriched GO term was circulatory system process and DNA metabolic process in groups C and D, respectively, although their enrichment was not significant (Fisher’s exact test, FDR corrected P > 0.05) (Table 7).

Discussion

Parthenogenetic females of A. pisum produce sexual progenies under short

22 days, when they have an expired seasonal timer (ES), whereas they produce no sexual progeny under long days or when they have an operative seasonal timer (OL, OS and

EL). The present transcriptome analyses based on RNA sequencing show genomewide transcriptional differences in relation to photoperiod and the seasonal timer in A. pisum.

As expected from the reproductive phenotypes, the ES aphids showed transcriptional changes in many genes, compared to the other aphids (OL, OS and EL) (Fig. 6).

Because the RNA samples were obtained from heads of A. pisum, and developing embryos were discarded, our findings support the possibility that the typical transcriptional changes in the ES aphids might be involved in the switch from production of parthenogenetic to sexual progenies in the maternal tissues (group A in

Fig. 6). In contrast, short days induced common transcriptional changes in several genes in the OS and ES aphids, whether or not the seasonal timer was operating (group B in

Fig. 6). These common transcriptional changes might be involved in the initial phase of the reproductive regulation, with which the seasonal timer will not interfere. In fact, opsin genes are differentially expressed in relation to photoperiod in A. pisum with an expired seasonal timer (Collantes-Alegre et al. 2018). In contrast, the OL and OS aphids shared common higher and lower gene expression levels compared to the EL aphids

(groups C and D in Fig. 6), even though these three aphid groups were commonly induced to produce parthenogenetic progenies. Therefore, these genes might be involved in mechanisms underlying maintenance and expiration of the seasonal timer.

Whereas the head is the center of the photoperiodic response and JH transduction in aphids (Steel & Lees 1977; Hardie 1987), the seasonal timer has not been localized yet in an aphid body or in regulatory pathways of the reproductive polyphenism (Le

Trionnaire et al. 2013). If the seasonal timer were located in the ovary and directly

23 regulated embryogenesis, gene expression levels in the head of aphids with an operative seasonal timer would be expected to differ in relation to photoperiod. The results of the present transcriptome analyses suggest that the seasonal timer might be located in the head, and modify the photoperiodic transcriptional responses there. Further investigations will be needed to localize the seasonal timer in the aphid body.

JH is considered to regulate phenotypic determination in the reproductive polyphenism of aphids (Tagu et al. 2005; Ogawa & Miura 2014). Generally, a low level of maternal JH induces production of sexual progeny (Hales & Mittler 1983; Corbitt &

Hardie 1985). In the ApL clone of A. pisum, several JH esterase genes are expressed more highly under short days than under long days (Ishikawa et al. 2012). Contrary to expectation, the DE genes identified by the present results do not include JH esterase genes. One possible explanation for these seemingly contradictory results might be the difference of the methods used for RNA sampling. Total RNA was extracted from the whole body in Ishikawa et al. (2012), whereas it was extracted from the head in the present study in order to focus on transcriptional differences in the maternal tissues.

Although JH is synthesized in the corpus allatum near the brain (Hardie 1987), genes associated with JH degradation are generally highly expressed in the fat body in insects

(Hinton & Hammock 2003; Crone et al. 2007; Liu et al. 2008). Further analyses of the tissues in the thorax and abdomen might show different patterns of expression of JH- related genes in relation to photoperiod and, moreover, in relation to the seasonal timer.

Epigenetic modifications of histones and chromatin have been shown to play important roles in polyphenisms in insects, especially in caste differentiation in ants and termites (Simola et al. 2013; Simola et al. 2016; Glastad et al. 2019; Suzuki et al.

2019). Moreover, it has been reported that post-translational regulation of proteins is

24 differentially activated between different phenotypes in polyphenic insects (Reynolds &

Hand 2009; Poupardin et al. 2015; Zhang et al. 2018). The present results from GO analyses show that upregulation of expression of genes associated with chromosome organization may be correlated with sexual morph production in A. pisum (Table 4).

Histone modification genes are highly diverged in A. pisum, suggesting that these genes might be involved in the wing and reproductive polyphenism (Rider et al. 2010;

Srinivasan & Brisson 2012). However, there is no direct evidence supporting the notion that epigenetic effects are involved in regulating the polyphenisms, and it was reported that pharmacological and RNAi approaches showed the effects on longevity, development rate and fecundity, but not on these polyphenic phenotypes in A. pisum

(Kirfel et al. 2019).

The present study obtained the first results showing transcriptional changes in histone and chromatin modification genes between mothers producing asexual and sexual progenies (e.g., ACYPI001312, ACYPI63818; see Table 7). These transcriptional changes found here are in accord with the changes reported by Vellichirammal et al.

(2016) between mothers producing wingless and winged progenies. Moreover, mothers producing sexual progenies showed higher expression of genes involved in protein processing and degradation in the GO terms “cellular localization” and “catabolic process” (e.g., ACYPI000546, ACYPI001332; see Table 7). All of these findings taken together suggest that both transcriptional and post-translational regulations might be involved in producing the different reproductive phenotypes in aphids. In the fresh fly,

Sarcophaga bullata Parker, the maternal experience of pupal diapause suppresses diapause in the progeny generation under diapause-inducing short days (Henrich &

Denlinger 1982). In that fly, expression levels of histone deacetylase genes in the larval

25 stage differ in relation to photoperiods and, moreover, in relation to the maternal experience of diapause (Reynolds et al. 2016). In the GO term “chromatin organization” in A, pisum, ACYPI49308 showed higher expression when the seasonal timer had expired regardless of photoperiod, although the others depended on photoperiod (Table

7). Based on these results, I suggest that regulations of chromatin organization might be involved both in the seasonal timer and in photoperiodism.

While the seasonal timer is operating, many rounds of DNA replication through parthenogenesis might result in incomplete replications, including shortening of telomeres (Blackburn 1991). Aphids with an operative seasonal timer (OL and OS) had experienced sexual reproduction recently, whereas aphids with an expired one (EL and

ES) had experienced repeated parthenogenesis for many years, and therefore telomeres of the OL and OS are expected to be much shorter than those of the EL and ES (Loxdale

& Lushai 2003). However, it has been reported that telomeres are still active after 10-40 years of parthenogenesis in several aphid species including A. pisum, indicating that telomere lengths are maintained not only in sexual reproduction but also in parthenogenesis (Monti et al. 2011). In the present results, the GO term “DNA metabolic process” was most highly enriched in group D, although the enrichment was not statistically significant (FDR corrected P > 0.05) (Table 4). The effects on categories of genes with these descriptions indicated that genes associated with maintenance of

DNA structure were differentially expressed between aphids with an operative (OL and

OS) and expired (EL and ES) seasonal timer (e.g., ACYPI002321, ACYPI41812; see

Table 7). These transcriptional differences indicate that structural maintenance of DNA, including telomeres, might be involved in expiration of the seasonal timer in A. pisum.

In the Asian tiger mosquito, Aedes albopictus (Skuse), transcriptome analyses of the

26 whole body have shown that genes associated with the GO term “DNA replication”, which is included in the GO term “DNA metabolic process”, are down-regulated in mothers producing diapause eggs (Huang et al. 2015). These transcriptional changes might be involved in developmental differentiations in embryos via regulation of cell divisions. In A. pisum in the present study, however, such transcriptional differences were observed in head samples that did not contain developing embryos. Elucidating how the maternal signals regulating embryonic development are transferred from the head to ovaries will require further investigations (Ogawa & Miura 2014).

In conclusion, the present study revealed that various gene expression alterations under regulations by the photoperiod and the seasonal timer lead to the reproductive morph determination in A. pisum. The GO analyses on these DE genes give novel insights into mechanisms by which the seasonal timer overrides the photoperiodic effect on the morph determination. Previously, it was difficult to explain how the chemical machinery of the seasonal timer resists so many dilutions through multiple generations while the seasonal timer is operating (Lees 1960). However, the fact that epigenetic modifications are transferred to the subsequent generation (Ho &

Burggren 2010; Heard & Martienssen 2014) can offer an explanation for such transfer.

Therefore, it appears to be reasonable that epigenetic modifications are involved in the operation and transgenerational transfer of the seasonal timer in aphids, and the present results support this hypothesis in part. Additional quantifications and functional analyses of the chromatin modification will clarify whether it is involved in the seasonal timer. Moreover, there are emerging technologies for functional analyses of aphid genes, for example, RNAi and CRISPR-Cas9 genome editing (Mutti et al. 2006; Mao & Zeng

2012; Le Trionnaire et al. 2019). The list of DE genes obtained will provide targets for

27 further analyses investigating the involvement of genes in photoperiodism and the seasonal timer in A. pisum.

28

Chapter 3

Adaptive significance of the seasonal timer

Introduction

In many temperate insects, preparation for overwintering, including diapause, is plasticly induced by autumnal short days (Tauber et al. 1986; Danks 1987). However, the photoperiodic responses can sometimes result in maladaptive consequences in the reproductive success under unpredictable climates (Grevstad & Coop 2015; Van Dyck et al. 2015). In many insects, post-diapause development begins when the temperature exceeds the lower threshold in spring regardless of photoperiod (Tauber et al. 1986).

Therefore, it is supposed that the autumnal short-day response might also be induced by spring daylengths if the timing of the diapause termination is advanced by unpredictably high temperatures. This spring diapause might disturb the utilization of seasonal resources. In many insects, however, sensitivity to the photoperiod is actually lost before the end of winter, and therefore the insects do not re-enter diapause in spring

(Tauber et al. 1986). Some insects regain sensitivity to the photoperiod after a certain delay in the same generation, and others do in the subsequent generation (Hodek &

Hodková 1992). For example, the carabid beetle Carabus yaconinus Bates overwinters in the adult stage and reproduces in spring. Some of these overwintered adults survive until the following spring and resume reproduction. Post-diapause adults of this species are refractory to the photoperiod during several tens of days and do not enter diapause even under diapause-inducing short-day conditions. Then they regain the photoperiod

29 sensitivity, which would be required to regulate the timing of the second diapause

(Shintani & Numata 2010). These photorefractory period and recurrent photoperiodic response can produce stable seasonality in the life cycles. In some multivoltine insects, in contrast, sensitivity to photoperiod is lost over several generations following the diapause (Henrich & Denlinger 1982; Reznik & Samartsev 2015). Although this multigenerational effect is also thought to be an adaptation to seasonality, no empirical study has been conducted yet to test this hypothesis.

In many aphid species, sexual females and males are induced by autumnal short days (Marcovitch 1923), whereas spring parthenogenetic generations do not respond to short days (Marcovitch 1924). It is peculiar to the seasonal timer of aphids that the photorefractory period is determined by the total number of days during which parthenogenetic generations are produced from the stem mother (Bonnemaison 1951;

Lees 1960; Brodel & Schaefers 1979). Although more than 90 years have passed since the discovery of the photorefractoriness in aphids (Marcovitch 1924), its adaptive significance has not been demonstrated empirically. If aphids lack the function of the seasonal timer, short days in spring might induce the sexual morph production. Taylor et al. (1998) reported unseasonal appearances of wild males from spring to early- summer in several aphid species, suggesting that daylengths at that time induce sexual morph productions. These males are likely to be produced by clones in which the seasonal timer has not been reset by sexual reproduction in the previous winter (Taylor et al. 1998; Williams & Dixon 2007). Therefore, the seasonal timer is hypothetically proposed to be a mechanism that prevents an unseasonal sexual morph production in spring (Wilson 1938; Lees 1960). This hypothesis is based on the assumption that spring daylengths are short enough to induce sexual morphs. However, daylengths in

30 spring, when stem mothers become active, increase as the season progresses and are likely to be longer than those in autumn, when sexual morphs appear. Therefore, it is necessary to examine empirically whether spring-like daylengths induce autumnal sexual morphs in aphids that lack the function of the seasonal timer. However, it is unclear whether increasing daylengths in spring induce the sexual morph production in this species, although the reproductive response to decreasing daylengths in autumn has been examined (Smith et al. 2011).

The present study aims to determine whether daylengths are short enough to induce the sexual morph production in spring, and the seasonal timer suppresses it in

Acyrthosiphon pisum (Harris). I reared two aphid groups that had an operative or an expired seasonal timer under natural photoperiods and temperatures that would occur in warm and cold regions to examine the effect of variable photoperiods and temperatures on the morph determination in spring. Based on the results obtained, the adaptive significance of the seasonal timer is discussed.

Materials and methods

Insects

The ApL clone of A. pisum was used for experiments. The parthenogenetic aphid cultures were maintained in plastic tubes (25 mm diameter, 100 mm depth) in which a seedling of the broad bean, Vicia faba L., was maintained by wrapping a bean in wet paper under 16 h light and 8 h darkness (LD 16:8 h) at 20.0 ± 1.0°C.

To obtain stem mothers, sexual females and males were produced under LD

31

10:14 h at 15.0 ± 1.0°C, and then fertilized eggs were obtained as described in Chapter

1. Fertilized eggs were incubated for approximately seven days under the same conditions, and then were placed under constant darkness at 4.0 ± 1.0°C for more than three months to ensure embryogenesis (Via 1992).

Rearing experiments

Aphids were reared under natural photoperiods and temperatures in spring in

Kyoto City (35.03°N, 135.79°E) and in Sapporo City (43.07°N, 141.34°E), Japan. After ambient temperatures became warm enough for aphids to develop, fertilized eggs were transferred outdoors on 1 March 2019 in Kyoto and on 17 April 2018 in Sapporo, and they were allowed to hatch and grow as aphids with an operative seasonal timer. In contrast, first instar parthenogenetic females were transferred outdoors from the stock culture on 7 March 2019 in Kyoto and on 21 April 2018 in Sapporo, and they were reared as aphids with an expired seasonal timer. Nymphs of these two groups were reared in plastic tubes and were fed on a seedling of the broad bean. They were kept in a plastic container (530 mm, 370 mm, 180 mm) under a sunshade net, and were protected from rain and direct sunlight. The density of the nymphs was kept at five per seedling to avoid overcrowding and appearance of alatiforms (Sutherland 1969). On the day of emergence, each adult was transferred to a new seedling and was allowed to give birth to nymphs there. After the onset of larviposition, each mother was transferred to a new seedling every two days until her death. The density of the progeny was a maximum of

21 per seedling. After adult emergence of the progenies, they were preserved in 70% ethanol and their reproductive phenotypes were judged using a stereomicroscope (S8

32

APO; Leica Microsystems, Wetzlar, Germany). Males were distinguished by small abdomen and external genitalia. Females with embryos with compound eyes were judged to be parthenogenetic and those with eggs with no compound eyes were judged to be sexual. The number of each morph was counted for each mother. Twenty-nine of

41 stem mothers and 15 of 25 stem mothers gave birth to 10 or fewer intact progenies in

Kyoto and Sapporo, respectively, and were excluded from analyses. Progenies that died before adult emergence were also excluded from analyses.

Statistical analyses were carried out using R 3.5.2 (R Core Team 2018).

Records of daily mean temperatures and daylengths, including civil twilight, were derived from the websites of the Japan Meteorological Agency and the National

Astronomic Observatory of Japan, respectively (Japan Meteorological Agency 2019;

National Astronomic Observatory of Japan 2019).

Results

In Kyoto, fertilized eggs of A. pisum hatched from 2 to 20 March 2019, and adult stem mothers emerged from 6 to 16 April 2019 (Fig. 8a). All of the 12 adult stem mothers with an operative seasonal timer gave birth to parthenogenetic females only, not to any sexual females or males (Fig. 8b). In advance of emergence of these adult stem mothers, adult mothers from the stock culture emerged from 4 to 6 April 2019.

Ten of the 12 adult mothers with an expired seasonal timer from the stock culture gave birth to either sexual females or males, or both of them, and the proportion of sexual progenies ranged from 0% to 89.6% (Fig. 9). The proportion of mothers that produced sexual progenies was significantly higher in the mothers from the stock culture than in

33 the stem mothers (P < 0.001, Fisher’s exact test). In general, sexual females and males were produced in the first and second half of the reproductive sequence, respectively

(Fig. 10). The number of parthenogenetic progenies was significantly smaller for the mothers from the stock culture than for the stem mothers (Fig. 8b). In contrast, the number of males and sexual females were significantly larger for the former than for the latter.

In Sapporo, fertilized eggs of A. pisum hatched from 20 to 28 April 2018, and adult stem mothers emerged from 16 to 26 May 2018 (Fig. 11a). In advance of emergence of these adult stem mothers, adult mothers emerged from the stock culture on 7 May 2019. Parthenogenetic females, but no sexual females or males, were produced by both of these two maternal types (Fig. 11b).

Discussion

The seasonal timer responsible for the photorefractory period has been thought to be adaptive for avoiding an unseasonal response to short days in spring in aphids

(Wilson 1938; Lees 1960). However, there has been no experimental evidence supporting this in any aphid species. In the present study on A. pisum in Kyoto, stem mothers with an operative seasonal timer produced only parthenogenetic females (Fig.

8b). However, parthenogenetic females with an expired seasonal timer produced many sexual females and males, and conversely the number of parthenogenetic progenies they produced decreased (Fig. 8b). The reproductive sequence of the mothers that produced sexual morphs was similar to that of parthenogenetic females reared under short days in the laboratory (MacKay 1987; Ishikawa et al. 2012). These results clearly show that

34 sexual morphs would be induced by short days in spring if the seasonal timer does not function in A. pisum. In Sapporo, however, no sexual females or males were produced regardless of whether the seasonal timer of the mothers had expired (Fig. 11b). It is true that the fecundity of the stem mothers was relatively low. Nevertheless, I consider that the seasonal timer itself would be not deleterious for the fecundity because the differences in the fecundity are caused by a deleterious recessive gene that became homozygous through self-fertilization within the ApL clone (Lees 1960). The different results in the number of sexual progenies between the two sites can be explained by the existence of a critical photoperiod for sexual morph production. In the laboratory, the critical photoperiod under which 50% of mothers produce sexual morphs is between 13 and 14 h at 15°C (Fig. 12). Daylengths that the maternal generation experienced at the nymphal stage were shorter and longer than 14 h in Kyoto and Sapporo, respectively.

(Figs. 8a, 11a). Therefore, the daylengths in the spring were short enough for A. pisum in Kyoto to induce the autumnal response, but not in Sapporo.

The present results first show empirically that the existence of the photorefractory generation is adaptive. In A. pisum, the sexual morph has a reproductive disadvantage over the parthenogenetic female due to the cost of sex and a low developmental rate in an obligate embryonic diapause (Rispe & Pierre 1998; Shingleton et al. 2003). In addition, high ambient temperature induces severe abnormality of the embryogenesis of eggs of A. pisum (Shingleton et al. 2003). Therefore, producing sexual morphs instead of parthenogenetic females would be maladaptive in spring. I thus conclude that the seasonal timer saves the reproductive cost and reduces loss of fitness. Moreover, the seasonal timer is unique as a mechanism that measures the number of days through successive generations, and this type of biological clock has

35 never been reported in other organisms. My experiments showed that such a unique mechanism is not a nonadaptive by-product but an ecologically important trait.

In addition to demonstrating the adaptive significance of the seasonal timer, the present study reveals a climatic difference in the effect of the photorefractory period between cold and warm springs. As mentioned above, sexual morphs are rarely produced under short days by A. pisum clones in warm regions, including Kyoto (Kanbe

& Akimoto 2009). Therefore, the adaptive significance of the seasonal timer might be marginal there. In contrast, clones in cold regions, including in Sapporo, produce sexual morphs under short days. Daylengths are usually long enough for avoiding unseasonal sexual morph productions in Sapporo when stem mothers hatch in spring. However, cold-region clones might face a warm spring with temperatures like the present experimental conditions in Kyoto due to yearly variation in temperatures (Harrington et al. 2007). In this case, stem mothers could hatch earlier under shorter daylengths, and the sexual morph production would be induced in spring, without the seasonal timer.

Furthermore, winged parthenogenetic females might migrate from a cold to warm region before producing sexual morphs (Smith & MacKay 1989). If this clone reproduces sexually and lays overwintering eggs in the warm region, stem mothers could hatch earlier than the cold region, and therefore experience spring short days.

Based on these speculations, I conclude that the seasonal timer contributes to the stability of the seasonal reproductive cycle in A. pisum, and the necessity for the seasonal timer is more dependent on climates than previously expected (Wilson 1938;

Lees 1960). The seasonal timer of aphids will be under selection to adapt their annual life cycles to a local climate and, in some cases, may be lost from their physiological mechanisms regulating the reproductive polyphenism. In Acyrthosiphon species living

36 in high arctic regions, in which the growing season is so short, stem mothers produce sexual progenies both under long days in the laboratory and under natural conditions in spring, suggesting that the function of the seasonal timer is lost or much weakened

(Strathdee et al. 1993; Strathdee & Bale 1996). It is possible that an adaptation to an occasional warm spring is involved in the evolutionary maintenance of the seasonal timer of A. pisum in Sapporo.

Based on the present result, I can explain that the previously reported early appearance of sexual morphs would be caused by spring short days in aphids with the expired seasonal timer. The two aphid species Brevicoryne brassicae L. and

Rhopalosiphum padi L. were shown to have the seasonal timer for suppression of the sexual morph production in the laboratory (Bonnemaison 1951; Lushai et al. 1996). In these species, moreover, weekly captures of males in the wild showed clear bimodal distributions in early-summer and autumn (Taylor et al. 1998). Taken together, it is assumed that the males in early-summer are produced by clones which overwinter as active parthenogenetic females and thus have an expired seasonal timer (Taylor et al.

1998). By contrast, the other aphid species show different seasonal life cycles, in which the sexual morph production might be regulated by other factors, including temperature and host-plant conditions. In the aphids Brachycaudus divaricatae Shaposhnikov and

Kaltenbachiella japonica (Matsumura), sexual females and males generally appear in early-summer and sexually-produced eggs spend a long period from summer to winter

(Akimoto 1985a, b; Rakauskas & Turčinavičienė 2006; Wilkaniec et al. 2016). In the leaf-curling plum aphid, Brachycaudus herichrysi (Kaltenbach), eggs hatch in autumn and stem mothers overwinter (Bell 1983; Williams & Dixon 2007). It needs to be investigated whether these species commonly have the seasonal timer for suppression of

37 unseasonal production of sexual morphs.

38

General discussion

Since the seasonal timer of aphids was firstly reported and some of its physiological features were shown (Marcovitch 1924; Lees 1960), it has been a fascinating mysterious mechanism in chronobiology for many years (Saunders 2002;

Tagu et al. 2005; Ogawa & Miura 2014). During this period, many researchers have revealed physiological, molecular and evolutionary aspects of the reproductive polyphenism in some aphid species including Acyrthosiphon pisum (Harris) (Steel &

Lees 1977; Corbitt & Hardie 1985; Mackay 1989; Kanbe & Akimoto 2009; Ishikawa et al. 2012). In the present study, I performed ecophysiological and trasncriptomic approaches to the molecular mechanism and adaptive significance of the seasonal timer in A. pisum. In Chapter 1, I demonstrated that the seasonal timer of A. pisum measures the number of days from hatching using the Lees’ method (Lees 1960), and thus I used

A. pisum as a model organism in the following studies on the seasonal timer. In

Chapters 2, I showed that the seasonal timer affected expression levels of hundreds of genes, including the several genes associated with the chromatin modification, in A. pisum under short days, taking advantage of the recently developed bioinformatic resources of this species (International Aphid Genomics Consortium 2010; Legeai et al.

2010). These new findings give insights for understanding molecular mechanisms not only of the seasonal timer in aphids, but also of the transgenerational effects in the other insect species (Reznik & Samartsev 2015). In Chapter 3, I showed that the seasonal timer suppressed unseasonal production of males in spring, although it depended on climatic conditions whether the seasonal timer was necessary to produce only parthenogenetic females. The prevalence of anholocyclic populations under mild

39 climates is also known in other aphid species (Dedryver et al. 2001; Delmotte et al.

2001). Therefore, I infer that the seasonal timer adapts to unpredictable temperatures in spring in many aphid species. Considering that the duration of the seasonal timer is under natural selection (Strathdee et al. 1993; Strathdee & Bale 1996; Dedryver et al.

2012), the life cycle and the seasonal timer would have mutually interacted in their evolutionary history in aphids. Furthermore, spring short daylengths might be a potentially diapause-inducing factor also in other short-living insects. Possibly numerous insects rely on the transgenerational mechanism like the seasonal timer, because the similar phenomenon has been reported in Diptera and Hymenoptera

(Henrich & Denlinger 1982; Reznik & Samartsev 2015).

Many insects have two or more types of the endogenous timing system to adapt daily, seasonal and other cyclical environmental changes. It has been investigated whether the molecular machinery of the circadian clock governing the circadian rhythm is identical to that of photoperiodism or the non-circadian rhythms (Bünning 1960;

Palmer 2000; Numata et al. 2015). For example, circadian clock genes are involved in photoperiodic regulation of diapause and development time, and thus act as components of the photoperiodic clock (Sakamoto et al. 2009; Ikeno et al. 2010; Kozak et al. 2019).

In the varied carpet beetle, Anthrenus verbasci Linnaeus, long days play an important role in entrainment of the circannual pupation rhythm (Miyazaki et al. 2005). In this case, the circannual clock of A. verbasci is regulated by photoperiodism in which the circadian clock genes might be involved. In insects that adapt to tidal and lunar cycles, however, circatidal and circalunar clocks do not consist of the circadian clock genes

(Takekata et al. 2012; Kaiser et al. 2016). In aphids, the seasonal timer and the photoperiodic clock cooperate to regulate the seasonal life cycle. In Chapter 1, the

40 duration of the seasonal timer did not vary with photoperiod, showing that the seasonal timer is not regulated by the photoperiodic clock. Therefore, I conclude that the seasonal timer and the circannual clock are differently involved in photoperiodism, although both of them are adaptations for seasonal environmental changes. On the other hand, the photoperiodic time measurement has been suggested to depend on the circadian clock also in aphids (Vaz Nunes & Hardie 1993; Cortés et al. 2010; Barberà et al. 2017). It is worth investigating whether the seasonal timer disrupts the circadian clock and results in production of parthenogenetic females even under short days.

Hourglass-like timers reported in insects and other animals give insights into the molecular mechanism underlying the seasonal timer, although they measure durations shorter than a generation time and are not transferred to the following generations. In the fruit fly, Drosophila melanogaster Meigen, and the tobacco hornworm, Manduca sexta L., larvae have a developmental timer that measures a minimal time for the last larval stage (Mirth and Shingleton 2012; Suzuki et al. 2013).

In both of these cases, insulin/target of rapamycin signaling pathway plays an important role in biosynthesis of ecdysone and determination of metamorphic timing (Mirth and

Shingleton 2012; Hatem et al. 2015). In D. melanogaster, moreover, the duration of the prepupal stage is determined by degradation of Blimp-1 protein, that is related with ecdysone metabolism (Akagi et al. 2016). These results indicate that key proteins work like sand in an hourglass measuring a minimal time for development. In other type of hourglass-like timers regulating seasonal responses, however, epigenetic modifications on DNA or histone play important roles. In the monarch butterfly, Danaus plexippus L., adults enter reproductive diapause under short days, and then terminate diapause endogenously after a few months. The termination timing of the reproductive diapause

41 may be determined by the expression of the Negative Cofactor 2β, which contributes to histone acetyltransferase activity, in D. plexippus (Green & Kronforst 2019). In the

Siberian hamster, Phodopus sungorus Pallas, adults become sexually inactive under short days, and then they become refractory to short days and are sexually reactivated after approximately 20 weeks. The termination timing of the reproductive inhibition may be determined by the level of DNA methylation in the proximal promoter region of deiodinase type III, which are associated with thyroid hormone metabolism, in the hypothalamus in P. sungorus (Stevenson & Prendergast 2013). Taking these instances into consideration, I infer that histone modifications would be important for regulation of the seasonal timer in A. pisum, rather than key molecules regulating the reproductive polyphenism, such as JH.

In conclusion, the present study revealed the diverse effects of the seasonal timer on the reproductive polyphenism of A. pisum, and more generally, pioneered the study on the transgenerational timer using newly developed knowledges and technologies. Such a unique timer might be regulated by epigenetic factors and can function as an adaptive mechanism for periodic environmental changes, similarly to several endogenous oscillatory clocks and hourglass-like timers. These findings contribute to the understanding of unity and diversity in timing mechanisms in insects.

42

Acknowledgments

First of all, I express my cordial gratitude to my supervisor, Dr. Hideharu Numata,

Kyoto University, for his kind guidance and encouragement throughout this study. I am grateful to Dr. Hiroko Udaka, Kyoto University, for her helpful advices, especially in the transcriptome analyses. I express my great appreciation to Dr. Akira Mori and all members of Laboratory of Ethology, Kyoto University, for fruitful discussions during the course of this work, and especially, to Mr. Yoshihito Kuromitsu and Mr. Yusuke

Todoroki for their help in the maintenance of aphid cultures, to Dr. Jun Endo for continuous support for the outdoor experiment, and to Dr. Chihiro Ito for technical advice in the RNA experiment. I am grateful to Dr. Tokitaka Oyama, Kyoto University for critical comments on the draft of this thesis. I thank Dr. Minoru Tamura for allowing me to rear insects outdoors at the Graduate School of Science Botanical Gardens, Kyoto

University.

I am grateful for Dr. Shin-ichi Akimoto and Dr. Takashi Kanbe, Hokkaido

University, for providing the aphid culture, valuable advice on aphid rearing, helpful supports and fruitful discussions in this study. I also express gratitude to members of

Laboratory of Systematic Entomology, Hokkaido University, for their encouragement during my two-months stay in Sapporo for the outdoor experiment.

I thank Dr. Elizabeth Nakajima for linguistic corrections in the draft of this thesis.

Finally, I express gratitude to my wife, Dr. Misaki Ishibashi, and all of my family members for their kind encouragement and continuous support during my Ph.D. course.

43

References

Akagi K, Sarhan M, Sultan A-RS, Nishida H, Koie A, Nakayama T, Ueda H (2016) A

biological timer in the fat body comprising Blimp-1, betaFtz-f1 and Shade

regulates pupation timing in Drosophila melanogaster. Development 143: 2410-

2416.

Akimoto S (1985a) Taxonomic study on gall aphids, Colopha, Paracolopha and

Kaltenbachiella (Aphidoidea: Pemphigidae) in East Asia, with special reference

to their origins and distributional patterns. Insecta Matsumurana 31: 1-79.

Akimoto S (1985b) Occurrence of abnormal phenotypes in a host-alternating aphid and

their implications for genome organization and evolution. Evolutionary Theory 7:

179-193.

Azaryan AG (1966) Some features of intraspecific geographic adaptations in Dysaphis

anthrisci C.B. (Homoptera, Aphidinea). Entomological Review 45: 278-283.

Barberà M, Cañas-Cañas R, Martínez-Torres D (2019) Insulin-like peptides involved in

photoperiodism in the aphid Acyrthosiphon pisum. Insect Biochemistry and

Molecular Biology 112: 103185.

Barberà M, Collantes-Alegre JM, Martínez-Torres D (2017) Characterisation, analysis

of expression and localisation of circadian clock genes from the perspective of

photoperiodism in the aphid Acyrthosiphon pisum. Insect Biochemistry and

Molecular Biology 83: 54-67.

Barberà M, Escrivá L, Collantes-Alegre JM, Meca G, Rosato E, Martínez-Torres D

(2018) Melatonin in the seasonal response of the aphid Acyrthosiphon pisum.

Insect Science in press. https://doi.org/10.1111/1744-7917.12652

44

Bell AC (1983) The life-history of the leaf-curling plum aphid Brachycaudus helichrysi

in Northern Ireland and its ability to transmit potato virus YC(AB). Annals of

Applied Biology 102: 1-6.

Benjamini Y, Hochberg Y (1995) Controlling the false discovery rate: A practical and

powerful approach to multiple testing. Journal of the Royal Statistical Society:

Series B (Methodological) 57: 289-300.

Blackburn EH (1991) Structure and function of telomeres. Nature 350: 569-573.

Blackman RL, Eastop VF (2007) Taxonomic issues. In Aphids as Crop Pests (ed. by HF

van Emden, R Harrington), pp. 1-29. CAB International, Wallingford, UK.

Blake GM (1958) Diapause and the regulation of development in Anthrenus verbasci

(L.) (Col., Dermestidae). Bulletin of Entomological Research 49: 751-775.

Bonnemaison L (1951) Contribution à l'étude des facteurs provoquant l'apparition des

formes ailées et sexuées chez les Aphidinae. Annales des Epiphyties 2: 1-380.

Bonnemaison L (1972) Ètude du “facteur-fondatrice” chez Dysaphis plantaginea Pass.

et Acyrthosiphon pisum Harris (Homoptères, Aphididae). Annales de la Societe

Entomologique de France 8: 607-627.

Brodel CF, Schaefers GA (1979) An “interval timer” for the production of oviparae in

Aphis rubicola. Entomologia Experimentalis et Applicata 25: 1-8.

Bünning E (1960) Circadian rhythms and the time measurement in photoperiodism.

Cold Spring Harbor Symposia on Quantitative Biology 25: 249-256.Campbell

CAM, Tregidga EL (2006) A transgenerational interval timer inhibits unseasonal

sexual morph production in damson-hop aphid, Phorodon humuli. Physiological

Entomology 31: 394-397.

Choi H, Shin S, Jung S, Clarke DJ, Lee S (2018) Molecular phylogeny of Macrosiphini

45

(Hemiptera: Aphididae): An evolutionary hypothesis for the Pterocomma-group

habitat adaptation. Molecular Phylogenetics and Evolution 121: 12-22.

Collantes-Alegre JM, Mattenberger F, Barberà M, Martínez-Torres D (2018)

Characterisation, analysis of expression and localisation of the opsin gene

repertoire from the perspective of photoperiodism in the aphid Acyrthosiphon

pisum. Journal of Insect Physiology 104: 48-59.

Corbitt TS, Hardie J (1985) Juvenile hormone effects on polymorphism in the pea

aphid, Acyrthosiphon pisum. Entomologia Experimentalis et Applicata 38: 131-

135.

Cortés T, Ortiz-Rivas B, Martínez-Torres D (2010) Identification and characterization of

circadian clock genes in the pea aphid Acyrthosiphon pisum. Insect Molecular

Biology 19 Suppl 2: 123-139.

Cortés T, Tagu D, Simon JC, Moya A, Martínez-Torres D (2008) Sex versus

parthenogenesis: a transcriptomic approach of photoperiod response in the model

aphid Acyrthosiphon pisum (Hemiptera: Aphididae). Gene 408: 146-156.

Crone EJ, Zera AJ, Anand A, Oakeshott JG, Sutherland TD, Russell RJ, Harshman LG,

Hoffmann FG, Claudianos C (2007) Jhe in Gryllus assimilis: Cloning, sequence-

activity associations and phylogeny. Insect Biochemistry and Molecular Biology

37: 1359-1365.

Danks HV (1987) Insect Dormancy: An Ecological Perspective. Biological Survey of

Canada, Ottawa, Canada.

Dedryver C-A, Le Gallic J-F, Mahéo F, Parisey N, Tagu D (2012) Delayed setting of the

photoperiodic response in recombinant clones of the aphid species Sitobion

avenae. Ecological Entomology 37: 293-299.

46

Dedryver CA, Hulle M, Le Gallic JF, Caillaud MC, Simon JC (2001) Coexistence in

space and time of sexual and asexual populations of the cereal aphid Sitobion

avenae. Oecologia 128: 379-388.

Delmotte F, Leterme N, Bonhomme J, Rispe C, Simon JC (2001) Multiple routes to

asexuality in an aphid species. Proceedings of the Royal Society B 268: 2291-

2299.

Denlinger DL (2002) Regulation of diapause. Annual Review of Entomology 47: 93-122.

Dixon AFG (1971) The ‘interval timer’ and photoperiod in the determination of

parthenogenetic and sexual morphs in the aphid, Drepanosiphum platanoides.

Journal of Insect Physiology 17: 251-260.

Dixon AFG (1972) The “interval timer”, photoperiod and temperature in the seasonal

development of parthenogenetic and sexual morphs in the lime aphid,

Eucallipterus tiliae L. Oecologia 9: 301-310.

Evans WG (1976) Circadian and circatidal locomotory rhythms in the intertidal beetle

Thalassotrechus barbarae (Horn): Carabidae. Journal of Experimental Marine

Biology and Ecology 22: 79-90.

Glastad KM, Hunt BG, Goodisman MAD (2019) Epigenetics in insects: Genome

regulation and the generation of phenotypic diversity. Annual Review of

Entomology 64: 185-203.

Goecks J, Nekrutenko A, Taylor J, the Galaxy Team (2010) Galaxy: a comprehensive

approach for supporting accessible, reproducible, and transparent computational

research in the life sciences. Genome Biology 11: R86.

Grantham ME, Brisson JA (2018) Extensive differential splicing underlies

phenotypically plastic aphid morphs. Molecular Biology and Evolution 35: 1934-

47

1946.

Green DA, II, Kronforst MR (2019) Monarch butterflies use an environmentally

sensitive, internal timer to control overwintering dynamics. Molecular Ecology

28: 3642-3655.

Grevstad FS, Coop LB (2015) The consequences of photoperiodism for organisms in

new climates. Ecological Applications 25: 1506-1517.

Hales DF, Mittler TE (1983) Precocene causes male determination in the aphid Myzus

persicae. Journal of Insect Physiology 29: 819-823.

Hardie J (1987) The corpus allatum, neurosecretion and photoperiodically controlled

polymorphism in an aphid. Journal of Insect Physiology 33: 201-205.

Harker JE (1956) Factors controlling the diurnal rhythm of activity of Periplaneta

americana L. Journal of Experimental Biology 33: 224-234.

Harrington R, Clark SJ, Welham SJ, Verrier PJ, Denholm CH, Hullé M, Maurice D,

Rounsevell MD, Cocu N (2007) Environmental change and the phenology of

European aphids. Global Change Biology 13: 1550-1564.

Hatem NE, Wang Z, Nave KB, Koyama T, Suzuki Y (2015) The role of juvenile

hormone and insulin/TOR signaling in the growth of Manduca sexta. BMC

Biology 13: 44.

Heard E, Martienssen RA (2014) Transgenerational epigenetic inheritance: Myths and

mechanisms. Cell 157: 95-109.

Henrich VC, Denlinger DL (1982) A maternal effect that eliminates pupal diapause in

progeny of the flesh fly, Sarcophaga bullata. Journal of Insect Physiology 28:

881-884.

Hinton AC, Hammock BD (2003) Juvenile hormone esterase (JHE) from Tenebrio

48

molitor: full-length cDNA sequence, in vitro expression, and characterization of

the recombinant protein. Insect Biochemistry and Molecular Biology 33: 477-487.

Ho DH, Burggren WW (2010) Epigenetics and transgenerational transfer: a

physiological perspective. Journal of Experimental Biology 213: 3-16.

Hodek I, Hodková M (1992). Regulation of postdiapause reproduction by recurrent

photoperiodic response. In Advances in Regulation of Insect Reproduction (ed. by

B Bennettová, I Gelbič, T Soldán), pp. 119-124. Institute of Entomology, Czech

Academy of Science, Ceske Budejovice, Czech Republic.

Huang X, Poelchau MF, Armbruster PA (2015) Global transcriptional dynamics of

diapause induction in non-blood-fed and blood-fed Aedes albopictus. PLoS

Neglected Tropical Diseases 9: e0003724.

Ikeno T, Tanaka SI, Numata H, Goto SG (2010) Photoperiodic diapause under the

control of circadian clock genes in an insect. BMC Biology 8: 116.

International Aphid Genomics Consortium (2010) Genome sequence of the pea aphid

Acyrthosiphon pisum. PLoS Biology 8: e1000313.

Ishikawa A, Ogawa K, Gotoh H, Walsh TK, Tagu D, Brisson JA, Rispe C, Jaubert-

Possamai S, Kanbe T, Tsubota T, Shiotsuki T, Miura T (2012) Juvenile hormone

titre and related gene expression during the change of reproductive modes in the

pea aphid. Insect Molecular Biology 21: 49-60.

Japan Meteorological Agency (2019) Available from URL:

https://www.jma.go.jp/jma/indexe.html.

Kaiser TS, Poehn B, Szkiba D, Preussner M, Sedlazeck FJ, Zrim A, Neumann T,

Nguyen LT, Betancourt AJ, Hummel T, Vogel H, Dorner S, Heyd F, von Haeseler

A, Tessmar-Raible K (2016) The genomic basis of circadian and circalunar timing

49

adaptations in a midge. Nature 540: 69-73.

Kanbe T, Akimoto S (2009) Allelic and genotypic diversity in long-term asexual

populations of the pea aphid, Acyrthosiphon pisum in comparison with sexual

populations. Molecular Ecology 18: 801-816.

Kenten J (1955) The effect of photoperiod and temperature on reproduction in

Acyrthosiphon pisum (Harris) and on the forms produced. Bulletin of

Entomological Research 46: 599-624.

Kim D, Langmead B, Salzberg SL (2015) HISAT: a fast spliced aligner with low

memory requirements. Nature Methods 12: 357-360.

Kim H, Lee S, Jang Y (2011) Macroevolutionary patterns in the Aphidini aphids

(Hemiptera: Aphididae): diversification, host association, and biogeographic

origins. PLoS One 6: e24749.

Kirfel P, Skaljac M, Grotmann J, Kessel T, Seip M, Michaelis K, Vilcinskas A (2019)

Inhibition of histone acetylation and deacetylation enzymes affects longevity,

development, and fecundity in the pea aphid (Acyrthosiphon pisum). Archives of

Insect Biochemistry and Physiology in press. https://doi.org/10.1002/arch.21614

Konopka RJ, Benzer S (1971) Clock mutants of Drosophila melanogaster. Proceedings

of the National Academy of Sciences of the USA 68: 2112-2116.

Kostal V (2011) Insect photoperiodic calendar and circadian clock: independence,

cooperation, or unity? Journal of Insect Physiology 57: 538-556.

Kozak GM, Wadsworth CB, Kahne SC, Bogdanowicz SM, Harrison RG, Coates BS,

Dopman EB (2019) Genomic basis of circannual rhythm in the European corn

borer moth. Current Biology 29: 3501-3509.

Lamb RJ, Pointing PJ (1972) Sexual morph determination in the aphid, Acyrthosiphon

50

pisum. Journal of Insect Physiology 18: 2029-2042.

Le Trionnaire G, Francis F, Jaubert-Possamai S, Bonhomme J, De Pauw E, Gauthier J-P,

Haubruge E, Legeai F, Prunier-Leterme N, Simon J-C, Tanguy S, Tagu D (2009)

Transcriptomic and proteomic analyses of seasonal photoperiodism in the pea

aphid. BMC Genomics 10: 456.

Le Trionnaire G, Tanguy S, Hudaverdian S, Gleonnec F, Richard G, Cayrol B, Monsion

B, Pichon E, Deshoux M, Webster C, Uzest M, Herpin A, Tagu D (2019) An

integrated protocol for targeted mutagenesis with CRISPR-Cas9 system in the pea

aphid. Insect Biochemistry and Molecular Biology 110: 34-44.

Le Trionnaire G, Wucher V, Tagu D (2013) Genome expression control during the

photoperiodic response of aphids. Physiological Entomology 38: 117-125.

Lees AD (1959) The role of photoperiod and temperature in the determination of

parthenogenetic and sexual forms in the aphid Megoura viciae Buckton—I: The

influence of these factors on apterous virginoparae and their progeny. Journal of

Insect Physiology 3: 92-117.

Lees AD (1960) The role of photoperiod and temperature in the determination of

parthogenetic and sexual forms in the aphid Megoura viciae Buckton—II. The

operation of the ‘interval timer’ in young clones. Journal of Insect Physiology 4:

154-175.

Legeai F, Shigenobu S, Gauthier J-P, Colbourne J, Rispe C, Collin O, Richards S,

Wilson ACC, Murphy T, Tagu D (2010) AphidBase: a centralized bioinformatic

resource for annotation of the pea aphid genome. Insect Molecular Biology 19

Suppl 2: 5-12.

Liu S, Yang B, Gu J, Yao X, Zhang Y, Song F, Liu Z (2008) Molecular cloning and

51

characterization of a juvenile hormone esterase gene from brown planthopper,

Nilaparvata lugens. Journal of Insect Physiology 54: 1495-1502.

Loxdale HD, Lushai G (2003) Maintenance of aphid clonal lineages: images of

immortality? Infection, Genetics and Evolution 3: 259-269.

Lushai G, Hardie J, Harrington R (1996) Inhibition of sexual morph production in the

bird cherry aphid, Rhopalosiphum padi. Entomologia Experimentalis et Applicata

81: 117-119.

MacKay PA (1977) Alata-production by an aphid: The “interval timer” concept and

maternal age effects. Journal of Insect Physiology 23: 889-893.

MacKay PA (1987) Production of sexual and asexual morphs and changes in

reproductive sequence associated with photoperiod in the pea aphid,

Acyrthosiphon pisum (Harris). Canadian Journal of Zoology 65: 2602-2606.

Mackay PA (1989) Clonal variation in sexual morph production in Acyrthosiphon pisum

(Homoptera: Aphididae). Environmental Entomology 18: 558-562.

Mao J, Zeng F (2012) Feeding-based RNA interference of a gap gene is lethal to the pea

aphid, Acyrthosiphon pisum. PLoS One 7: e48718.

Marcovitch S (1923) Plant lice and light exposure. Science 58: 537-538.

Marcovitch S (1924) The migration of the Aphididae and the appearance of the sexual

forms as affected by the relative length of daily light exposure. Journal of

Agricultural Research 27: 513-522.

Margaritopoulos JT, Tsitsipis JA, Prophetou-Athanasiadou DA (2002) An interval timer

controls the production of sexual morphs in Myzus persicae (Homoptera:

Aphididae). Physiological Entomology 27: 251-255.

McCarthy DJ, Chen Y, Smyth GK (2012) Differential expression analysis of multifactor

52

RNA-Seq experiments with respect to biological variation. Nucleic Acids

Research 40: 4288-4297.

Meuti ME, Stone M, Ikeno T, Denlinger DL (2015) Functional circadian clock genes are

essential for the overwintering diapause of the Northern house mosquito, Culex

pipiens. Journal of Experimental Biology 218: 412-422.

Mirth CK, Shingleton AW (2012) Integrating body and organ size in Drosophila: recent

advances and outstanding problems. Frontiers in Endocrinology 3: 49.

Miura T, Braendle C, Shingleton A, Sisk G, Kambhampati S, Stern DL (2003) A

comparison of parthenogenetic and sexual embryogenesis of the pea aphid

Acyrthosiphon pisum (Hemiptera: Aphidoidea). Journal of Experimental Zoology

Part B 295B: 59-81.

Miyazaki Y, Nisimura T, Numata H (2005) A phase response curve for circannual

rhythm in the varied carpet beetle Anthrenus verbasci. Journal of Comparative

Physiology A 191: 883-887.

Monti V, Giusti M, Bizzaro D, Manicardi GC, Mandrioli M (2011) Presence of a

functional (TTAGG)n telomere-telomerase system in aphids. Chromosome

Research 19: 625-633.

Moran NA (1992) The evolution of aphid life cycles. Annual Review of Entomology 37:

321-348.

Mousseau TA, Dingle H (1991) Maternal effects in insect life histories. Annual Review

of Entomology 36: 511-534.

Mukai A, Goto SG (2016) The clock gene period is essential for the photoperiodic

response in the jewel wasp Nasonia vitripennis (Hymenoptera: Pteromalidae).

Applied Entomology and Zoology 51: 185-194.

53

Mutti NS, Park Y, Reese JC, Reeck GR (2006) RNAi knockdown of a salivary transcript

leading to lethality in the pea aphid, Acyrthosiphon pisum. Journal of Insect

Science 6: 38.

Nagao T, Shimozawa T (1987) A fixed time-interval between two behavioural elements

in the mating behaviour of male crickets, Gryllus bimaculatus. Behaviour

35: 122-130.

National Astronomic Observatory of Japan (2019) Available from URL:

https://eco.mtk.nao.ac.jp/cgi-bin/koyomi/cande/riseset_en.cgi.

Nelson RJ, Denlinger DL, Somers DE (2010) Photoperiodism: The Biological

Calendar. Oxford University Press, New York, USA.

Neumann D (1966) Die lunare und tägliche Schlüpfperiodik der Mücke Clunio.

Zeitschrift für vergleichende Physiologie 53: 1-61.

Nijhout HF (2003) Development and evolution of adaptive polyphenisms. Evolution

and Development 5: 9-18.

Nisimura T, Numata H (2001) Endogenous timing mechanism controlling the circannual

pupation rhythm of the varied carpet beetle Anthrenus verbasci. Journal of

Comparative Physiology A 187: 433-440.

Numata H, Miyazaki Y, Ikeno T (2015) Common features in diverse insect clocks.

Zoological Letters 1: 10.

Nylin S, Gotthard K (1998) Plasticity in life-history traits. Annual Review of

Entomology 43: 63-83.

Ogawa K, Miura T (2014) Aphid polyphenisms: trans-generational developmental

regulation through viviparity. Frontiers in Physiology 5: 1.

Palmer JD (2000) The clocks controlling the tide-associated rhythms of intertidal

54

animals. BioEssays 22: 32-37.

Pertea M, Pertea GM, Antonescu CM, Chang TC, Mendell JT, Salzberg SL (2015)

StringTie enables improved reconstruction of a transcriptome from RNA-seq

reads. Nature Biotechnology 33: 290-295.

Pittendrigh CS (1954) On temperature independence in the clock system controlling

emergence time in Drosophila. Proceedings of the National Academy of Sciences

of the USA 40: 1018-1029.

Pittendrigh CS (1960) Circadian rhythms and the circadian organization of living

systems. Cold Spring Harbor Symposia on Quantitative Biology 25: 159-184.

Poupardin R, Schöttner K, Korbelová J, Provazník J, Doležel D, Pavlinic D, Beneš V,

Koštál V (2015) Early transcriptional events linked to induction of diapause

revealed by RNAseq in larvae of drosophilid fly, Chymomyza costata. BMC

Genomics 16: 720.

R Core Team (2018) R: A Language and Environment for Statistical Computing. R

Foundation for Statistical Computing, Vienna, Austria.

Raff M (2007) Intracellular developmental timers. Cold Spring Harbor Symposia on

Quantitative Biology 72: 431-435.

Rakauskas R, Turčinavičienė J (2006) Brachycaudus divaricatae Shaposhnikov, 1956 in

Europe: biology, morphology and distribution, with comments on its taxonomic

position (Hemiptera, Sternorrhyncha: Aphididae). Zoosystematics and Evolution

82: 248-260.

Reynolds JA, Bautista-Jimenez R, Denlinger DL (2016) Changes in histone acetylation

as potential mediators of pupal diapause in the flesh fly, Sarcophaga bullata.

Insect Biochemistry and Molecular Biology 76: 29-37.

55

Reynolds JA, Hand SC (2009) Embryonic diapause highlighted by differential

expression of mRNAs for ecdysteroidogenesis, transcription and lipid sparing in

the cricket Allonemobius socius. Journal of Experimental Biology 212: 2075-

2084.

Reznik SY, Samartsev KG (2015) Multigenerational maternal inhibition of prepupal

diapause in two Trichogramma species (Hymenoptera: Trichogrammatidae).

Journal of Insect Physiology 81: 14-20.

Rider SD, Srinivasan DG, Hilgarth RS (2010) Chromatin-remodelling proteins of the

pea aphid, Acyrthosiphon pisum (Harris). Insect Molecular Biology 19 Suppl 2:

201-214.

Rispe C, Pierre J-S (1998) Coexistence between cyclical parthenogens, obligate

parthenogens, and intermediates in a fluctuating environment. Journal of

Theoretical Biology 195: 97-110.

Robinson MD, McCarthy DJ, Smyth GK (2010) edgeR: a Bioconductor package for

differential expression analysis of digital gene expression data. Bioinformatics 26:

139-140.

Robinson MD, Oshlack A (2010) A scaling normalization method for differential

expression analysis of RNA-seq data. Genome Biology 11: R25.

Rockey SJ, Miller BB, Denlinger DL (1989) A diapause maternal effect in the flesh fly,

Sarcophaga bullata: Transfer of information from mother to progeny. Journal of

Insect Physiology 35: 553-558.

Sakamoto T, Uryu O, Tomioka K (2009) The clock gene period plays an essential role in

photoperiodic control of nymphal development in the cricket Modicogryllus

siamensis. Journal of Biological Rhythms 24: 379-390.

56

Satoh A, Yoshioka E, Numata H (2008) Circatidal activity rhythm in the mangrove

cricket Apteronemobius asahinai. Biology Letters 4: 233-236.

Saunders DS (2002) Insect Clocks. 3rd edn. Elsevier, Amsterdam, Netherlands.

Sgrò CM, Terblanche JS, Hoffmann AA (2016) What can plasticity contribute to insect

responses to climate change? Annual Review of Entomology 61: 433-451.

Shingleton AW, Sisk GC, Stern DL (2003) Diapause in the pea aphid (Acyrthosiphon

pisum) is a slowing but not a cessation of development. BMC Developmental

Biology 3: 7.

Shintani Y, Numata H (2010) Adaptive significance of the recurrent photoperiodic

response in a spring-breeding carabid beetle, Carabus yaconinus. Entomological

Science 13: 367-374.

Simola DF, Graham RJ, Brady CM, Enzmann BL, Desplan C, Ray A, Zwiebel LJ,

Bonasio R, Reinberg D, Liebig J, Berger SL (2016) Epigenetic (re)programming

of caste-specific behavior in the ant Camponotus floridanus. Science 351:

aac6633.

Simola DF, Ye C, Mutti NS, Dolezal K, Bonasio R, Liebig J, Reinberg D, Berger SL

(2013) A chromatin link to caste identity in the carpenter ant Camponotus

floridanus. Genome Research 23: 486-496.

Simon J-C, Rispe C, Sunnucks P (2002) Ecology and evolution of sex in aphids. Trends

in Ecology and Evolution 17: 34-39.

Smedley D, Haider S, Ballester B, Holland R, London D, Thorisson G, Kasprzyk A

(2009) BioMart–biological queries made easy. BMC Genomics 10: 22.

Smith MAH, MacKay PA (1989) Seasonal variation in the photoperiodic responses of a

pea aphid population: evidence for long-distance movements between

57

populations. Oecologia 81: 160-165.

Srinivasan DG, Brisson JA (2012) Aphids: a model for polyphenism and epigenetics.

Genetics Research International 12: 431531.

Steel CGH, Lees AD (1977) The role of neurosecretion in the photoperiodic control of

polymorphism in the aphid Megoura viciae. Journal of Experimental Biology 67:

117-135.

Stevenson TJ, Prendergast BJ (2013) Reversible DNA methylation regulates seasonal

photoperiodic time measurement. Proceedings of the National Academy of

Sciences of the USA 110: 16651-16656.

Strathdee AT, Bale JS (1996) Life-cycle and morph production in the Arctic aphid

Acyrthosiphon brevicorne. Polar Biology 16: 293-300.

Strathdee AT, Bale JS, Block WC, Webb NR, Hodkinson ID, Coulson SJ (1993)

Extreme adaptive life-cycle in a high arctic aphid, Acyrthosiphon svalbardicum.

Ecological Entomology 18: 254-258.

Sutherland ORW (1969) The role of crowding in the production of winged forms by two

strains of the pea aphid, Acyrthosiphon pisum. Journal of Insect Physiology 15:

1385-1410.

Suzuki R, Yaguchi H, Maekawa K (2019) Histone modifying genes are involved in the

molting period during soldier differentiation in Zootermopsis nevadensis. Journal

of Insect Physiology 117: 103892.

Suzuki Y, Koyama T, Hiruma K, Riddiford LM, Truman JW (2013) A molt timer is

involved in the metamorphic molt in Manduca sexta larvae. Proceedings of the

National Acadamy of Sciences of the USA 110: 12518-12525.

Tagu D, Sabater-Muñoz B, Simon J-C (2005) Deciphering reproductive polyphenism in

58

aphids. Invertebrate Reproduction & Development 48: 71-80.

Takekata H, Matsuura Y, Goto SG, Satoh A, Numata H (2012) RNAi of the circadian

clock gene period disrupts the circadian rhythm but not the circatidal rhythm in

the mangrove cricket. Biology Letters 8: 488-491.

Tauber MJ, Tauber CA, Masaki S (1986) Seasonal Adaptations of Insects. Oxford

University Press, New York, USA.

Taylor MS, Harrington R, Clark SJ (1998). Unseasonal male aphids. In Aphids in

Natural and Managed Ecosystems (ed. by JM Nieto Nafría, AFG Dixon), pp.

287-293. Universidad de León (Secretariado de Publicaciones), León, Spain.

Thurmond J, Goodman JL, Strelets VB, Attrill H, Gramates LS, Marygold SJ, Matthews

BB, Millburn G, Antonazzo G, Trovisco V, Kaufman TC, Calvi BR, the FlyBase

Consortium (2019) FlyBase 2.0: the next generation. Nucleic Acids Research 47:

D759-D765.

Tomioka K, Matsumoto A (2010) A comparative view of insect circadian clock systems.

Cellular and Molecular Life Sciences 67: 1397-1406.

Van Dyck H, Bonte D, Puls R, Gotthard K, Maes D (2015) The lost generation

hypothesis: could climate change drive ectotherms into a developmental trap?

Oikos 124: 54-61.

Vaz Nunes M, Hardie J (1993) Circadian rhythmicity is involved in photoperiodic time

measurement in the aphid Megoura viciae. Experientia 49: 711-713.

Vaz Nunes M, Saunders D (1999) Photoperiodic time measurement in insects: A review

of clock models. Journal of Biological Rhythms 14: 84-104.

Vellichirammal NN, Madayiputhiya N, Brisson JA (2016) The genomewide

transcriptional response underlying the pea aphid wing polyphenism. Molecular

59

Ecology 25: 4146-4160.

Via S (1992) Inducing the sexual forms and hatching the eggs of pea aphids.

Entomologia Experimentalis et Applicata 65: 119-127.

Walsh TK, Brisson JA, Robertson HM, Gordon K, Jaubert-Possamai S, Tagu D,

Edwards OR (2010) A functional DNA methylation system in the pea aphid,

Acyrthosiphon pisum. Insect Molecular Biology 19 Suppl 2: 215-228.

Ward JH (1963) Hierarchical grouping to optimize an objective function. Journal of the

American Statistical Association 58: 236-244.

Wilkaniec B, Borowiak-Sobkowiak B, Wilkaniec A, Trzciński P, Kozłowska M (2016)

Effect of climate change on seasonal flight activity of aphid males in urban green

area. Acta Scientiarum Polonorum. Hortorum Cultus 15: 157-169.

Williams IS, Dixon AFG (2007) Life cycles and polymorphism. In Aphids as Crop

Pests (ed. by HF van Emden, R Harrington), pp. 69-85. CAB International,

Wallingford, UK.

Wilson F (1938) Some experiments on the influence of environment upon the forms of

Aphis chloris Koch (Aphididae). Transactions of Royal Entomological Society of

London 87: 165-180.

Yaguchi H, Suzuki R, Matsunami M, Shigenobu S, Maekawa K (2019) Transcriptomic

changes during caste development through social interactions in the termite

Zootermopsis nevadensis. Ecology and Evolution 9: 3446-3456.

Zhang G, Storey JM, Storey KB (2018) Elevated chaperone proteins are a feature of

winter freeze avoidance by larvae of the goldenrod gall moth, Epiblema

scudderiana. Journal of Insect Physiology 106: 106-113.

60

Fig. 1. A schematic representation of the typical annual life cycle of holocyclic aphids.

The seasonal timer is considered to expire by autumn and reset by sexual reproduction.

61

Fig. 2. A schematic representation of the experiment in Acyrthosiphon pisum in Chapter

1. (a) In the older- and the younger-sister lineages produced under short-day conditions, aphids born within three days and approximately ten days after larviposition onset were reared successively under short-day conditions at 15°C or 20°C. (b) In the older-sister lineage under long-day conditions, aphids of each generation born within three days after larviposition onset were transferred to short-day conditions, and reared for two successive generations at 20°C.

62

Fig. 3. The proportions of sexual morph progenies under short-day conditions at 15°C

(a) and 20°C (b) in Acyrthosiphon pisum. Stem mothers and parthenogenetic females of the stock culture that had been maintained parthenogenetically since 2005 were used as the first generation, and their parthenogenetic progenies that were born within three days of the larviposition onset were used as the second generation. Each mother produced 27.3 ± 15.5 progenies (mean ± S.D.). Rectangular boxes represent the interquartile ranges of the data, the horizontal lines inside the boxes are the median values, and whiskers extend to the most extreme data point which is no more than 1.5 times of the interquartile range. Open circles are outliers. Significant differences between stem mothers and parthenogenetic females of the stock culture are indicated by asterisks (**P < 0.01, ***P < 0.001, Mann-Whitney U test).

63

Fig. 4. Sexual morph production as a function of the number of days from hatching to the birth of mothers in Acyrthosiphon pisum. Lineages consisted of aphids that were born within three days and approximately ten days after the larviposition onset under short-day conditions at 15°C (a, b, respectively) and 20°C (c, d, respectively), and aphids that were born within three days of the larviposition onset under long-day conditions at 20°C (e). The fitted curves are based on logistic regression between the proportion of sexual progenies and the number of days from hatching. Dotted lines show the mean number of days from hatching to the birth of mothers of each generation. Triangles show the critical number of days from hatching to the production of sexual morphs (see Table 3).

64

Fig. 5. A schematic representation of the experimental design in Acyrthosiphon pisum in

Chapter 2. The upper panel shows the method used for rearing aphids with an operative seasonal timer under long-day and short-day conditions (OL and OS). The lower panel shows a method for rearing aphids with an expired seasonal timer under long-day and short-day conditions (EL and ES). The OL and EL aphids were induced to produce parthenogenetic progenies by long days, whereas the ES aphids were induced to produce sexual progenies by short days. Note that the sexual morph production of the

OS aphids was suppressed by an operative seasonal timer.

65

Fig. 6. Gene expression variation in relation to photoperiod and the seasonal timer in

Acyrthosiphon pisum. The heatmap shows relative expression levels of 500 genes that were differentially expressed in relation to the two variables (likelihood ratio test, FDR corrected P < 0.05). The dendrograms show hierarchical clustering of genes and aphid samples. The branch lengths indicate the degree of difference. OL and OS indicate samples of aphids with an operative seasonal timer under long-day and short-day conditions, respectively. EL and ES indicate samples of aphids with an expired seasonal timer under long-day and short-day conditions, respectively. Color bars beside the heatmap indicate four clusters of genes.

66

Fig. 7. The hierarchical structure of gene ontology (GO) terms of biological process category in Acyrthosiphon pisum. Grey shading indicates enriched GO terms in the differentially expressed gene group A (Fisher’s exact test, P < 0.05).

67

Fig. 8. Photoperiodic morph determination in stem mothers and mothers from the stock culture of Acyrthosiphon pisum in spring in 2019 in Kyoto. (a) Seasonal changes in natural daylength including civil twilight and daily mean air temperature. Grey shading indicates the nymphal developmental period of the mothers. (b) The number of progenies. The number of mothers was 12 for each type of mother. Rectangular boxes represent the interquartile data ranges, horizontal lines inside the boxes are the medians, and whiskers extend to the most extreme data point which is no more than 1.5 times the interquartile range. Open circles are outliers. Asterisks indicate significant differences in the number of progenies between the mother types (*P < 0.05, ***P < 0.001, Student’s t-test in Pf, Mann-Whitney U-test in M and Sf). Abbreviations; Pf: parthenogenetic female, M; male, Sf; sexual female.

68

Fig. 9. Effect of the seasonal timer on morph determination in the progeny of

Acyrthosiphon pisum. Stem mothers (a) and parthenogenetic females originating from the stock culture (b) were used as parents. The parents were reared under natural photoperiods and temperatures in spring in Kyoto. Morph abbreviations; Pf: parthenogenetic female, M; male, Sf; sexual female.

69

Fig. 10. Examples of the reproductive sequence of the parthenogenetic females originating from the stock culture of Acyrthosiphon pisum under natural photoperiods and temperatures in spring in Kyoto. (a) A parent produced parthenogenetic females in the first half and males in the second half in the reproductive sequence. (b) A parent produced sexual females in the first half and males in the second half in the reproductive sequence. Morph abbreviations; Pf: parthenogenetic female, M; male, Sf; sexual female.

70

Fig. 11. Photoperiodic morph determination in stem mothers and mothers from the stock culture of Acyrthosiphon pisum in spring in 2018 in Sapporo. (a) Seasonal changes in natural daylength (including civil twilight) and daily mean air temperature. Grey shading indicates the nymphal developmental period of the mothers. (b) The number of progenies.

The numbers of mothers were 10 for stem mothers and 11 for mothers from the stock culture. Rectangular boxes indicate the interquartile data ranges, horizontal lines inside the boxes are the medians, and whiskers extend to the most extreme data point.

Abbreviations; Pf: parthenogenetic female, M; male, Sf; sexual female.

71

Fig. 12. Effect of photoperiod on morph determination in the progeny of Acyrthosiphon pisum. Parthenogenetic females originating from the stock culture were used as parents.

Percentage of the parents that produced one or more sexual progenies (a) and percentage of the sexual progenies (b) is shown. Rectangular boxes represent the interquartile data ranges, horizontal lines inside the boxes are the medians, and whiskers extend to the most extreme data point which is no more than 1.5 times the interquartile range. Open circles are outliers. Asterisks indicate significant differences in the percentages between photoperiods (***P < 0.001, Fisher’s exact test in (a), Mann-

Whitney U-test in (b)).

72

Table 1. List of aphid species in which the presence of a seasonal timer has been reported.

Species Generationa Temperatureb Photoperiodb Reference

Acyrthosiphon brevicorne - - - Strathdee & Bale 1996

Acyrthosiphon pisum No Yes No Bonnemaison 1972

Chapter 1 of this thesis

Acyrthosiphon svalbardicum - - - Strathdee et al. 1993

Aphis chloris - - - Wilson 1938

Aphis forbesi - - - Marcovitch 1924

Aphis rubicola No - No Brodel & Schaefers 1979

Aphis rumicis - - - Marcovitch 1924

Brevicoryne brassicae No - Unclearc Bonnemaison 1951

Drepanosiphum platanoides - - - Dixon 1971

Dysaphis anthrisci - - - Azaryan 1966

Dysaphis plantaginea - - - Bonnemaison 1972

Dysaphis sorbi - - - Marcovitch 1924

Eucallipterus tillae - - - Dixon 1972

Megoura viciae No Yes Unclearc Lees 1960

Myzus persicae - - - Margaritopoulos et al. 2002

Phorodon humuli - - - Campbell & Tregidga 2006

Rhopalosiphum padi - - - Lushai et al. 1996

Sitobion avenae - - - Dedryver et al. 2012 a Dependence of the seasonal timer on the number of generations. b Dependence of the seasonal timer on temperature or photoperiod. c The seasonal timer works even under long days, but it is unclear whether photoperiod has any effect on its duration.

73

Table 2. Results of GLMM to test the effects of variables on the proportion of sexual morph progenies in Acyrthosiphon pisum.

Variables Coefficient of variables z-value P-value

Estimate S.E.

(a)

Intercept -18.2 2.013 -9.06 < 0.0001

Day 0.126 0.00973 12.92 < 0.0001

Temperature 0.447 0.0928 4.82 < 0.0001

Lineage (younger) -0.484 0.433 1.11 0.264

(b)

Intercept -7.41 0.761 -9.73 < 0.0001

Day 0.116 0.0118 9.75 < 0.0001

Photoperiod (long-day) 0.837 0.444 -1.89 0.0591

(a) The analyzed effects of the number of days from hatching, temperature and the sister-lineage were calculated in the older- and the younger-sister lineages under short- day conditions at 15°C and 20°C. (b) The effects of the number of days from hatching and photoperiod were analyzed in the older-sister lineages at 20°C under short-day and long-day conditions in the third and subsequent generations.

74

Table 3. Duration of the seasonal timer for suppression of sexual morph production in Acyrthosiphon pisum.

Days/generationa

Temperature Photoperiod Lineage (mean ± S.D.) Critical daysb Generation at critical daysc

15°C 10 h light-14 h dark Older-sister 17.9 ± 3.02 90.3 5.0

Younger-sister 28.8 ± 4.27 90.7 3.2

20°C 10 h light-14 h dark Older-sister 11.1 ± 1.92 69.7 6.3

Younger-sister 18.0± 4.87 72.2 4.0

16 h light-8 h dark Older-sister 10.7 ± 2.45 65.3 6.1 a The mean number of days per generation was calculated as the number of days from the birth to the onset of larviposition. b The critical number of days from hatching of the stem mother to the birth of mothers that produced more than 50% sexual morph progenies was calculated from the results of GLMM. c The number of generations on the critical day was calculated from the mean number of days per generation and the critical number of days.

75

Table 4. A full list of enriched gene ontology (GO) terms in the GO “biological process” in Acyrthosiphon pisum.

Gene groupa GO term name GO ID P-valueb FDR

A Protein targeting GO:0006605 3.57E-3 3.87E-2

A Intracellular protein transport GO:0006886 3.57E-3 3.87E-2

A Nucleocytoplasmic transport GO:0006913 4.72E-2 3.33E-1

A Organelle organization GO:0006996 2.35E-3 3.87E-2

A Cell population proliferation GO:0008283 2.11E-3 3.87E-2

A Catabolic process GO:0009056 1.60E-2 1.32E-1

A Cellular component organization GO:0016043 2.18E-2 1.71E-1

A Cellular protein localization GO:0034613 3.57E-3 3.87E-2

A Intracellular transport GO:0046907 9.83E-4 2.46E-2

A Nuclear transport GO:0051169 4.72E-2 3.33E-1

A Chromosome organization GO:0051276 1.27E-4 1.02E-2

A Cellular localization GO:0051641 9.83E-4 2.46E-2

A Establishment of localization in cell GO:0051649 9.83E-4 2.46E-2

A Cellular macromolecule localization GO:0070727 3.57E-3 3.87E-2

A Cellular component organization or biogenesis GO:0071840 2.41E-2 1.83E-1

C Circulatory system process GO:0003013 4.55E-2 1

D DNA metabolic process GO:0006259 2.35E-2 1 a Gene group names follow Fig. 6. b P-values were calculated using Fisher’s exact tests.

76

Table 5. A full list of enriched gene ontology (GO) terms in the GO “cellular component” in Acyrthosiphon pisum.

Gene groupa GO term name GO ID P-valueb FDR

A Nuclear chromosome GO:0000228 4.47E-2 3.32E-1

A Intracellular GO:0005622 1.85E-2 1.49E-1

A Cell GO:0005623 8.80E-3 8.31E-2

A Nucleus GO:0005634 2.32E-6 6.55E-4

A Nucleoplasm GO:0005654 3.39E-3 3.87E-2

A Chromosome GO:0005694 1.22E-2 1.07E-1

A External encapsulating structure GO:0030312 4.95E-3 5.17E-2

A Membrane-enclosed lumen GO:0031974 1.05E-3 2.46E-2

A Nuclear lumen GO:0031981 1.05E-3 2.46E-2

A Organelle lumen GO:0043223 1.05E-3 2.46E-2

A Organelle GO:0043226 2.70E-3 3.87E-2

A Membrane-bounded organelle GO:0043227 1.45E-4 1.02E-3

A Intracellular organelle GO:0043229 2.17E-3 3.87E-2

A Intracellular membrane-bounded organelle GO:0043231 9.85E-5 1.02E-3

A Organelle part GO:0044422 2.42E-3 3.87E-2

A Intracellular part GO:0044424 6.75E-3 6.79E-2

A Nuclear part GO:0044428 1.00E-3 2.46E-2

A Intracellular organelle part GO:0044446 2.42E-3 3.87E-2

A Cell part GO:0044464 1.28E-2 1.09E-1

A Intracellular organelle lumen GO:0070013 1.05E-3 2.46E-2 a Gene group names follow Fig. 6. b P-values were calculated using Fisher’s exact tests.

77

Table 6. A full list of enriched gene ontology (GO) terms in the GO “molecular function” in Acyrthosiphon pisum.

Gene groupa GO term name GO ID P-valueb FDR

A Nucleic acid binding GO:0003673 2.99E-3 3.87E-2

A RNA binding GO:0003723 7.87E-3 7.65E-2

A Histone binding GO:0042393 1.03E-2 9.35E-2

A Heterocyclic compound binding GO:1901363 2.99E-3 3.87E-2

A Organic cyclic compound binding GO:0097159 2.99E-3 3.87E-2

C Peptidase activity GO:0008233 4.55E-4 6.42E-2

C Hydrolase activity GO:0016786 1.51E-2 1

C Catalytic activity, acting on a protein GO:0140096 4.55E-4 6.42E-2

D Oxidoreductase activity GO:0016491 3.49E-2 1 a Gene group names follow Fig. 6. b P-values were calculated using Fisher’s exact tests.

78

Table 7. Genes that were differentially expressed and associated with the enriched gene ontology (GO) terms in Acyrthosiphon pisum.

Gene groupa Gene ID Description E/Ob S/Lc

d d Log2FC FDR Log2FC FDR Chromosome organization A ACYPI000194 Actin-like protein 6B 7.37 1.18E-3 3.24 5.29E-1 A ACYPI001312 Chromatin-remodeling complex ATPase chain Iswi 8.82 7.71E-7 8.79 1.06E-6 A ACYPI003380 Inhibitor of growth protein 5-like 8.31 1.33E-5 8.29 1.85E-5 A ACYPI007490 Inhibitor of growth protein 4-like 6.98 8.61E-3 6.96 1.06E-2 A ACYPI009726 Inhibitor of growth protein 5-like 7.04 1.45E-2 7.02 1.75E-2 A ACYPI062334 Male-specific lethal 3 homolog 8.56 4.23E-6 8.53 5.74E-6 A ACYPI063818 Histone-lysine N-methyltransferase eggless-like 7.55 8.92E-4 7.53 1.14E-3 A ACYPI064701 Histone-lysine N-methyltransferase eggless-like 5.57 3.88E-3 8.31 2.46E-5 A ACYPI21463 Male-specific lethal 3 homolog 9.28 1.78E-8 9.25 2.74E-8 A ACYPI29468 MORF-related gene on chromosome 15e 8.57 2.28E-5 8.54 2.98E-5 A ACYPI41812 X-ray repair cross-complementing protein 5e 4.38 1.19E-3 3.64 1.53E-2 A ACYPI43275 Chromatin modification-related protein eaf3-like 8.19 1.11E-6 3.12 2.09E-1 A ACYPI56670 Histone acetyltransferase KAT8-like 6.49 2.41E-2 6.46 2.94E-2 C ACYPI49308 Chromatin modification-related protein eaf3-like 6.54 1.61E-2 1.65 1 D ACYPI42395 Microtubule-associated protein EB1e -0.747 1 -7.84 1.36E-2 Cellular localization

79

A ACYPI000133 mRNA export factor-like 5.94 1.85E-4 7.26 2.25E-3 A ACYPI000546 Nuclear pore glycoprotein p62-like 8.34 2.99E-5 4.52 5.30E-2 A ACYPI001155 mRNA export factor-like 7.49 1.50E-3 7.46 1.87E-3 A ACYPI004161 Translocating chain-associated membrane protein 1-like 6.56 2.34E-2 6.53 2.89E-2 A ACYPI005876 Mitochondrial import receptor subunit TOM40 homolog 6.43 3.03E-2 6.41 3.69E-2 A ACYPI010046 Mitochondrial import receptor subunit TOM40 homolog 2-like 6.52 2.22E-2 6.50 2.71E-2 A ACYPI24657 Peroxisomal membrane protein PEX13 7.85 2.48E-4 7.82 3.18E-4 Cell population proliferation A ACYPI003380 Inhibitor of growth protein 5-like 8.31 1.33E-5 8.29 1.85E-5 A ACYPI007490 Inhibitor of growth protein 4-like 6.98 8.61E-3 6.96 1.06E-2 A ACYPI009726 Inhibitor of growth protein 5-like 7.04 1.45E-2 7.02 1.75E-2 Catabolic process A ACYPI000140 Hexokinase type 2-like 1.87 5.13E-1 3.12 2.72E-2 A ACYPI001332 Ubiquitin carboxyl-terminal hydrolase 5-like 9.79 3.49E-11 9.76 4.42E-11 A ACYPI001383 Protein 5NUC-like 7.00 1.27E-2 6.97 1.56E-2 A ACYPI001512 Regulator of nonsense transcripts 1-like 4.10 8.44E-3 3.81 2.19E-2 A ACYPI002455 Proteasome subunit alpha type-3-like 2.49 4.74E-4 1.84 3.32E-2 A ACYPI003137 Ubiquitin carboxyl-terminal hydrolase 48-like 5.24 9.75E-6 5.10 2.33E-5 A ACYPI003673 Elongin C-like 7.28 1.41E-3 4.38 1.41E-1 A ACYPI005159 DET1- and DDB1-associated protein 1-like 4.22 1.45E-2 3.42 9.91E-2

80

A ACYPI066898 Lon protease homolog 2, peroxisomal 8.52 6.37E-6 8.49 8.77E-6 A ACYPI072123 Lon protease homolog 2, peroxisomal 9.68 4.70E-10 6.82 7.19E-7 DNA metabolic process A ACYPI002321 Structural maintenance of chromosomes protein 6 8.35 1.15E-5 8.33 1.62E-5 A ACYPI004712 Glucose dehydrogenase [FAD, quinone]-like 7.39 2.05E-3 7.36 2.52E-3 A ACYPI20838 Transposable element Pe 6.35 3.22E-2 6.33 3.92E-2 A ACYPI21463 Male-specific lethal 3 homolog 9.28 1.78E-8 9.25 2.74E-8 A ACYPI41812 X-ray repair cross-complementing protein 5e 4.39 1.19E-3 3.64 1.53E-2 C ACYPI001862 CCHC-type zinc-finger nucleic acid binding proteine 3.63 4.43E-2 -1.65 1 D ACYPI004902 Cuticular protein analogous to peritrophins 3-C precursor -1.45 1.08E-2 0.473 1 D ACYPI006065 Inflatede 0.326 1 -5.80 2.01E-2 D ACYPI062942 Dystrophine -0.727 1 -4.31 4.10E-2 D ACYPI066389 Putative tRNA pseudouridine synthase Pus10e -0.728 5.03E-3 0.157 1 D ACYPI42395 Microtubule-associated protein EB1e -0.747 1 -7.84 1.36E-2 Circulatory system process C ACYPI000734 Glutamyl aminopeptidase isoform 0.650 1.95E-2 0.23 1 a Gene group names follow Fig. 6. b E/O indicates expired/operative seasonal timers. c S/L indicates short-day/long-day conditions.

81 d P-values were calculated using likelihood ratio tests and adjusted using a false discovery rate. e Homologs of Drosophila melanogaster genes were searched with BLASTP.

82