Insect Molecular Biology (2020) doi: 10.1111/imb.12616 Silencing of the DNA methyltransferase 1 associated protein 1 (DMAP1) gene in the invasive ladybirdHarmonia axyridis implies a role of the DNA methyltransferase 1-DMAP1 complex in female fecundity J. Gegner*, T. Gegner*†, H. Vogel‡ and populations with a biocontrol strain differing in egg- A. Vilcinskas*† laying capacity, suggesting that the DNA methyltrans- *Department of Bioresources, Fraunhofer Institute for ferase 1-DMAP1 complex may influence the invasive Molecular Biology and Applied Ecology, Giessen, performance of this ladybird. Germany; †Institute for Insect Biotechnology, Justus- Keywords: epigenetics, DNMT1, DMAP1, RNA interfer- Liebig-University of Giessen, Giessen, Germany; and ence, fecundity, ovary development, Harmonia ‡Max Plank Institute for Chemical Ecology, Jena, Germany axyridis. Introduction Abstract Individuals of the same species can often be assigned to The invasive harlequin ladybird Harmonia axyridis is a one of several distinct phenotypes, a phenomenon known textbook example of polymorphism and polyphenism as polymorphism when the variation depends on genetic as the temperature during egg development determines differences and polyphenism when an identical genome the frequency of melanic morphs and the number and can give rise to different phenotypes according to the envi- size of black spots in nonmelanic morphs. Recent con- ronment (Simpson et al., 2011). Such phenotypic plasticity cepts in evolutionary biology suggest that epigenetic is thought to be advantageous in dynamic environments mechanisms can translate environmental stimuli into (Levins, 1968; DeWitt et al., 1998), particularly if the adap- heritable phenotypic changes. To investigate whether tations prepare the organism for future environmental epigenetic mechanisms influence the penetrance and changes. The harlequin ladybird Harmonia axyridis (also expressivity of colour morphs in H. axyridis, we used known as the multicoloured Asian ladybird) is a textbook RNA interference to silence key enzymes required for example of polymorphism because it occurs as a number DNA methylation and histone modification. We found of distinct morphs differing, for example, in overall colour that neither of these epigenetic mechanisms affected and the number of the spots on the elytra (Dobzhansky, the frequency of different morphs, but there was a sig- 1924, 1933). These polymorphic colour patterns are regu- nificant impact on life-history traits such as longevity lated by a single locus, which comprises the four major and fecundity. Strikingly, we found that silencing the alleles axyridis, conspicua, spectabilis and succinea, as gene encoding for DNAmethyltransferase 1 associated well as 11 rare alleles (Hosino, 1940; Komai, 1956; Tan, protein 1 (DMAP1) severely reduced female fecundity, 1946; Tan and Li, 1934). The melanic forms (axyridis, con- which correlated with an abundance of degenerated spicua and spectabilis) may be advantageous in cold cli- ovaries in DMAP1-knockdown female beetles. Finally, mates when dark surfaces absorb heat more quickly we observed significant differences in DMAP1 expres- during exposure to sunlight. The frequency of different sion when we compared native and invasive H. axyridis morphs therefore varies amongst populations with different geographical ranges, but it also exhibits seasonal changes First published online 10 September 2019. (Jiang et al., 2008; Jing and Zhang, 2001; Yuan et al., Correspondence: Andreas Vilcinskas, Department of Bioresources, Fraun- 1994). The latter occurs because the temperature during hofer Institute for Molecular Biology and Applied Ecology, Winchester Strasse 2, 35395 Giessen, Germany. Tel: + 49 6419937600; e-mail: the larval development stages influences the frequency of andreas.vilcinskas@agrar.uni-giessen.de melanic morphs (Knapp and Nedvěd, 2013). H. axyridis 148 © 2019 The Authors. Insect Molecular Biology published by John Wiley & Sons Ltd on behalf of Royal Entomological Society. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. 13652583, 2020, 2, Downloaded from https://resjournals.onlinelibrary.wiley.com/doi/10.1111/imb.12616 by Justus-Liebig-Universitat, Wiley Online Library on [01/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Silencing of the DMAP1 gene 149 therefore combines the classic features of both polymor- 2018), we selected HAT, HDAC and DMAP1 genes as tar- phism and polyphenism (Michie et al., 2010). gets for RNAi-mediated silencing. We found that the phe- It is currently unclear how environmental stimuli translate notypic plasticity of H. axyridis was not affected by the into distinct phenotypes encoded by the same genome double-stranded RNA (dsRNA) treatment of any of these (Vilcinskas and Vogel, 2016). Recent concepts predict that target genes, but we observed a striking and unexpected epigenetic mechanisms control transcriptional reprogram- impact on life-history parameters, which we investigated ming resulting in phenotypic plasticity (Flores et al., in the context of differences in gene expression between 2013). To test this hypothesis, we investigated whether invasive, non-invasive and laboratory-bred populations. the frequency of H. axyridis colour morphs in response to temperature is mediated by epigenetic mechanisms. To Results provide an easy readout system, we counted the black spots on the elytra of succinea beetles as previously Temperature dependency of colour polymorphism described (Michie et al., 2010) and assigned them to three In agreement with the results reported by Michie et al. groups: low (0–7), middle (8–15) and high (> 16) spot num- (2010) we found that beetles raised at lower temperatures ber. The F1eggs of the first filial generation (F1) laid by the from the egg stage produced more spots. For the low- F0 generation were kept at three different temperatures temperature group (raised at 15C) all the beetles devel- (15, 21 or 28C) and the emerging beetles were tested to oped more than 16 black spots (high category). In contrast, determine the frequency of each group. We then used in the mid-temperature group (raised at 21C) two thirds of RNA interference (RNAi) to silence genes encoding the beetles belonged to the middle and low spot number enzymes responsible for DNA methylation and histone categories, and in the high temperature group (raised at acetylation/deacetylation, two epigenetic mechanisms that 28C) 92% of the beetles developed fewer than eight spots are known to regulate the initiation of transcription (Glastad and were therefore assigned to the low spot number cate- et al., 2011; Vilcinskas, 2017; Gegner et al., 2019). gory (Table 1). DNA methylation is mediated by enzymes known as These values were used as controls to determine the DNA methyltransferases (DNMTs), which catalyse the impact of silencing the epigenetic regulators HAT, HDAC addition of methyl groups to cytidine residues and favour and DMAP1 in succinea colour morphs. Stage four the formation of compact and inaccessible chromatin, gen- (L4) larvae and F0 beetles were injected with dsRNA corre- erally blocking the interaction between DNA and transcrip- sponding to each gene, but neither the adult beetles devel- tion factors (Jurkowska et al., 2011). In this process the oping from the L4 larvae nor the F1 offspring of the injected DNA methyltransferase 1 associated binding protein F0 parents showed any changes in the distribution of spot 1 (DMAP1) is the key activator of DNMT1, as the loss of this numbers compared to water-injected mock controls. How- protein results in hypomethylation (Lee et al., 2010; Roun- ever, these experiments provided intriguing evidence that tree et al., 2010). RNAi did affect the life-history traits of the F0 and F1 bee- In contrast, histone acetyltransferases (HATs) add ace- tles, and we therefore investigated this aspect in greater tyl groups to the histones, resulting in an open chromatin detail. structure that encourages the binding of transcription fac- tors and promotes gene expression. The activity of HATs is countered by histone deacetylases (HDACs) that Impact of HAT, HDAC and DMAP1 dsRNA treatment on remove the acetyl groups and restore the compact and the development, survival and phenotype of L4 larvae and inaccessible chromatin state (Marks et al., 2003). the F1 generation reared at 15 C Using the comprehensive H. axyridis transcriptome In L4 larvae, the development time from pupae to adults sequence (Vilcinskas et al., 2013) and gene predictions was shortened following the injection of HAT dsRNA on the recently published genome data (Gautier et al., (Fig. 1, Tables S1–S3). Furthermore, significantly fewer Table 1.Colour polymorphism in F1 beetles of the first filial generation (F1) according to the temperature during development. The F0 generation was reared at 21C and separated according to the number of black spots on the elytra (low 0–7, middle 8–15, high > 16). F1 eggs were subsequently reared at three different temperatures (15, 21 and 28C) and the colour pattern of the F1 beetles was recorded F1 beetle 15C 21C 28C F0 beetle High Middle Low High Middle Low High Middle Low High 104 0 0 189 17 15 84 16 94 Middle 111 0 0 133 57 39 31 11 147 Low 132 0 0 68 81 75 12 4 182 © 2019 The Authors. Insect Molecular Biology published by JohnWiley & Sons Ltd on behalf of Royal Entomological Society., 29, 148–159 13652583, 2020, 2, Downloaded from https://resjournals.onlinelibrary.wiley.com/doi/10.1111/imb.12616 by Justus-Liebig-Universitat, Wiley Online Library on [01/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 150 J. Gegner et al. Figure 1. Development and survival of double-stranded RNA-injected stage four (L4) larvae compared to mock-injected controls: n(mock) = 35, n(HAT) = n (HDAC) = n(DMAP1) = 33. (A) Development from L4 larvae to pupae. (B) Development from pupae to beetles. (C) Survival rate. The black lines represent the median and significant differences between groups are depicted by different letters (Tables S1–S4). DMAP1, DNA methyltransferase 1 associated protein 1; HAT, histone acetyltransferase; HDAC, histone deacetylase; d, days; dpi, days postinjection. animals developed from larvae to pupae and from pupae to L4 larvae or the offspring of the F0 beetles. In contrast, adult beetles after RNAi treatment of HAT and HDAC, with the injection of DMAP1 dsRNA resulted in the mortality HDAC dsRNA having the most severe effect on the larva- of 91% of the L4 larvae and, as stated above, there were to-pupa transition. None of the DMAP1 silenced L4 larvae no offspring from the injected F0 parents (Figs 1 and 2, developed into pupae. They either died or remained Tables S1–S7). arrested at the larval stage until the end of the experiment (Tables S1–S4). Following the injection of F0 beetles with HAT or HDAC dsRNA, the F1 generation took longer to Impact of HAT, HDAC and DMAP1 dsRNA treatment on develop from eggs to beetles thanmock controls, whereas life-history parameters of beetles reared at 21 C, and F0 beetles injected with DMAP1 dsRNA failed to produce impact of DMAP1 knockdown on female reproductive any offspring at all (Fig. 2, Tables S5–S7). Compared to organs mock controls, we found that fewer offspring of the F0 bee- Having investigated the effect of RNAi treatment of epige- tles injected with HDAC dsRNA developed from larvae netic regulators on L4 larvae and the F1 generation, we into pupae and from pupae into beetles, and the latter next analysed the impact on the survival, fertility and fecun- was also true for the offspring of beetles injected with dity of the dsRNA injected beetles. The silencing of DMAP1 HAT dsRNA (Tables S4 and S5). For HAT and HDAC, was lethal for 100% of the beetles after 24 days, but injec- the injection of dsRNA did not affect the survival of the tion of HDAC dsRNA had only a slight, nonsignificant effect © 2019 The Authors. Insect Molecular Biology published by JohnWiley & Sons Ltd on behalf of Royal Entomological Society., 29, 148–159 13652583, 2020, 2, Downloaded from https://resjournals.onlinelibrary.wiley.com/doi/10.1111/imb.12616 by Justus-Liebig-Universitat, Wiley Online Library on [01/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Silencing of the DMAP1 gene 151 Figure 2. Development and survival of the first filial (F1) generation after the injection of F0 beetles with double-stranded RNA (compared to mock-injected controls). (A) Development from eggs to larvae. (B) Development from larvae to pupae. (C) Development from pupae to beetles. (D) Survival rate of F1 generation. The black lines represent the median and grey bars show mean values. Error bars show the standard error of the mean. Significant differences between groups are depicted by different letters (Tables S5–S7). HAT, histone acetyltransferase; HDAC, histone deacetylase; d, days. on survival, and HAT-RNAi had no effect at all. The injec- controls (Fig. 4A, B). The hatching rate for the offspring of tion of DMAP1 dsRNA caused the first beetles to die after beetles injected with HDAC or HAT dsRNA was lower than 11 days and the median survival time was 15 days the hatching rate of controls, but there was no difference in (Fig. 3, Table S8). the developmental time of the larvae in each cohort When mock control females were mated with males (Fig. 4C, D). injected with DMAP1 dsRNA, there was no difference in Given that female beetles with silenced DMAP1 the total number of eggs, the hatching rate or offspring showed a significant loss of fertility and fecundity, we dis- development compared to mock control females mated sected their ovaries and recorded the number of ovari- with mock control males, but the number of eggs per batch oles, follicles, mature eggs, and degenerated eggs per was lower (Fig. 4, Tables S9 and S10). In contrast, when ovary. 10 days after injection, the number of ovarioles both parents were injected with DMAP1 dsRNA, or when per ovary was similar in the beetles injected with DMAP1 mock control males were mated with DMAP1 silenced dsRNA and the controls injected with water. However, females, the total number of eggs was significantly lower the ovaries in the DMAP1 dsRNA treatment group con- as well as the number of eggs per batch, and none of the tained significantly fewer follicles and mature eggs, and eggs produced viable larvae (Fig. 4). The injection of a much larger number of degenerate eggs, compared to HDAC dsRNA only marginally reduced the total number the control group (Fig. 5, Tables S11 and S12). There of eggs and number of eggs per batch compared to were no morphological differences between the testes of © 2019 The Authors. Insect Molecular Biology published by JohnWiley & Sons Ltd on behalf of Royal Entomological Society., 29, 148–159 13652583, 2020, 2, Downloaded from https://resjournals.onlinelibrary.wiley.com/doi/10.1111/imb.12616 by Justus-Liebig-Universitat, Wiley Online Library on [01/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 152 J. Gegner et al. expression, and functional analyses to determine that the transcription factor Pannier regulates melanic pattern poly- morphism in H. axyridis and that highly variable discrete colour forms in natural populations result from cis- regulatory allelic variation of a single gene (Gautier et al., 2018). In contrast, polyphenism in insects involves the environmental modulation of phenotypes that may be specified by an invariant genome or may already exist as a number of different allelic variants, thus contributing to the penetrance and expressivity of different alleles. The mechanism that underlies polyphenism is unclear, but epi- genetic mechanisms may explain some of the features of this phenomenon, such as the ability of environmental pressure in one generation to affect the phenotype of later generations (Glastad et al., 2019; Simpson et al., 2011). In the invasive harlequin ladybirdH. axyridis, we confirmed Figure 3. Kaplan–Meyer survival curves of Harmonia axyridis beetles injected with histone acetyltransferase (HAT)/histone deacetylase (HDAC)/ the already reported influence of temperature during the DNA methyltransferase 1 associated protein 1 (DMAP1) double-stranded development of colour morphs (Michie et al., 2010). The RNA (dsRNA) or injected with water as amock control. The survival curve of exposure of eggs to a lower temperature increased the beetles injected with DMAP1 dsRNA differs significantly from those injected with HAT/HDAC dsRNA or with water as a mock control (P < 0.001), but no number and extent of black spots on the elytra of the result- significant differences were observed for other treatments (mock vs. HAT: ing adults. Testing our hypothesis that epigenetic mecha- P = 1.000, mock vs. HDAC=HAT vs. HDAC:P = 0.120) (Table S8). d, days. nisms acting at the level of transcriptional initiation could translate environmental stimuli into the frequencies of dif- males injected with DMAP1 dsRNA and those of mock ferent morphs, we injected dsRNA corresponding to the control males (Fig. S4). genes encoding key regulators of DNA methylation To verify that our findings on female fecundity and ovary (DMAP1) and histonemodification (HATs and HDACs) into development result from DMAP1 knockdown, we com- L4 larvae and F0 adults. However, no effects on the fre- pared relative gene expression levels of DMAP1 dsRNA- quency of different morphs in the adults developing from injected beetles with those of mock-injected controls by injected larvae or the F1 offspring of the injected adults quantitative real-time PCR (qPCR). And indeed our data were observed, indicating that neither DNA methylation showed that DMAP1 gene expression was suppressed nor histone acetylation is solely responsible for the poly- about 3.23-fold ( 1.45 SD) by the dsRNA at 3 days post- phenism observed in H. axyridis. injection (dpi) and that the knockdown remained almost Although our initial hypothesis was rejected, we stable with 2.87-fold ( 1.44 SD) at 10 dpi (Fig. 6A, observed that RNAi treatment of HAT and HDAC genes Tables S13 and S14). influencedH. axyridis life-history parameters including sur- vival and development. Injection of HAT dsRNA did not DMAP1 gene expression level in ve different H. axyridis affect the survival of larvae under cold stress or beetlesfi populations under normal conditions, but it did delay the developmental transition from larvae to pupae under cold stress compared Given the importance of DMAP1 in traits such as fecundity, to controls. In Tribolium castaneum, the inhibition of HAT which are relevant for the invasive performance of H. axyri- by curcumin did not influence larval survival or develop- dis, we investigated whether DMAP1 gene expression dif- ment, but extended the life span of beetles under heat fers between native populations from China and Korea, stress (Bingsohn et al., 2016). Similarly, the longevity of invasive populations from Germany and France, and a adult Drosophila melanogaster was increased after treat- laboratory-reared biocontrol strain with a particularly high ing larvae with the same HAT inhibitor under normal condi- reproductive capacity. The qPCR data showed that tions (Soh et al., 2013). However, the inhibition of HAT with DMAP1 was expressed at a lower level in the native and curcumin did not affect D. melanogaster reproduction (Soh invasive populations compared to the biocontrol strain, et al., 2013) but did reduce the fertility of T. castaneum but there was no difference between the native and inva- (Bingsohn et al., 2016). In our experiments, we found that sive populations (Fig. 6B, Tables S15 and S16). RNAi treatment of HAT reduced the fertility of H. axyridis but had no impact on its fecundity. Discussion Interestingly, injection of HDAC dsRNA caused similar A very recent study combined whole-genome sequencing, effects as shown for HAT, ie reduced fertility but no population-based genome-wide association studies, gene changes in fecundity or survival. However, one key © 2019 The Authors. Insect Molecular Biology published by JohnWiley & Sons Ltd on behalf of Royal Entomological Society., 29, 148–159 13652583, 2020, 2, Downloaded from https://resjournals.onlinelibrary.wiley.com/doi/10.1111/imb.12616 by Justus-Liebig-Universitat, Wiley Online Library on [01/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Silencing of the DMAP1 gene 153 Figure 4. Fecundity and fertility ofHarmonia axyridis beetles injected with histone acetyltransferase (HAT)/histone deacetylase (HDAC)/DNAmethyltransferase 1 associated protein 1 (DMAP1) double-stranded RNA (dsRNA) or with water as a mock control. (A) Total number of batches. (B) Number of eggs per batch. (C) Developmental time from eggs to F1 larvae of the first filial gegneration (F1). (D) Hatching rate of F1 larvae. For mock, HAT, HDAC and DMAP1, both sexes were injected with dsRNA. The black lines represent the median, and significant differences between groups are depicted by different letters (Tables S9 and S10). DMAP1(f), only females injected, mated with mock-injected males; DMAP1(m), only males injected, mated with mock-injected females; d, days. difference was that RNAi treatment of HDAC did not extend component of the machinery is DMAP1 (Lee et al., 2010; the transition times of any stages of development. The Rountree et al., 2010). The RNAi-mediated silencing of HDAC inhibitor valproic acid reduced the survival of T. cas- DMAP1 resulted in severe effects, killing 50% of the trea- taneum larvae and heat-stressed beetles but had no effect ted beetles after 15 days and 100% after 24 days. To the on the longevity ofD. melanogaster even though it reduced best of our knowledge, this is the first report showing the the fertility and fecundity of both species (Bingsohn et al., lethal consequence of silencing the DNA methylation 2016; Gayathri and Harini, 2012). Interestingly, earlier machinery in adult insects, although lethality has been studies concluded that the inhibition of HDAC by sodium observed in insect and mammalian embryos (Flores 4-phenylbutyrate or trichostatin A and sodium butyrate et al., 2013). In the mammalian embryos, demethylated increased the longevity of D. melanogaster but did not DNA induced apoptosis, andwe hypothesize that the same affect reproduction (Kang et al., 2002; Zhao et al., 2005). process is responsible for the lethality observed in These contradicting findings highlight the need for more H. axyridis. studies in insects to investigate the complex regulatory net- Interestingly, we found that the reduced fertility and work of histone acetylation and deacetylation. fecundity caused by DMAP1 silencing was only evident in DNA methylation is also intimately linked to the regula- females, corresponding to the presence of severely degen- tion of genes involved in life-history traits, and a key erated ovaries. To our knowledge this is the first time that © 2019 The Authors. Insect Molecular Biology published by JohnWiley & Sons Ltd on behalf of Royal Entomological Society., 29, 148–159 13652583, 2020, 2, Downloaded from https://resjournals.onlinelibrary.wiley.com/doi/10.1111/imb.12616 by Justus-Liebig-Universitat, Wiley Online Library on [01/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 154 J. Gegner et al. Figure 5. Ovaries of Harmonia axyridis females 10 days after the injection of (A) mock or (B) DNA methyltransferase 1 associated protein 1 (DMAP1) double- stranded RNA, as well as the mean numbers of (C) ovarioles, (D) follicles, (E) mature eggs and (F) degenerated eggs per ovary: n(mock) = 12, n(DMAP1) = 11. Error bars show standard error of the mean and significant differences between groups are depicted by different letters (Tables S11 and S12). [Colour figure can be viewed at wileyonlinelibrary.com]. the essential role of the DNMT1 cofactor DMAP1 in insect RNAi-mediated silencing was also used to demonstrate reproduction has been shown. Until now, similar results that DNMT1 is essential for egg production and embryo were only reported for DNMT1 and DNMT3. For example, viability in the large milkweed bug, Oncopeltus fasciatus for the brown plant hopper, Nilaparvata lugens, it was (Bewick et al., 2019). A recent study using T. castaneum recently reported that the knockdown of DNMT1 and elucidated that DNMT1 is expressed throughout the entire DNMT3 reduced the number of offspring by suppressing life cycle of this model beetle, but the highest levels of mes- ovary development (Zhang et al., 2015). Further, senger RNA (mRNA) transcripts were found in eggs and © 2019 The Authors. Insect Molecular Biology published by JohnWiley & Sons Ltd on behalf of Royal Entomological Society., 29, 148–159 13652583, 2020, 2, Downloaded from https://resjournals.onlinelibrary.wiley.com/doi/10.1111/imb.12616 by Justus-Liebig-Universitat, Wiley Online Library on [01/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Silencing of the DMAP1 gene 155 Figure 6. (A) Changes in DNA methyltransferase 1 associated protein 1 (DMAP1) gene expression in beetles injected with DMAP1 double-stranded RNA relative to mock-injected control beetles at 3 and 10 days postinjection (dpi): DMAP1 gene expression of mock control is represented as baseline. Dots show mean values of three biological replicates (each comprising five males and five females). Error bars indicate standard error of the mean. Significant gene expression knockdown is depicted by asterisks. Significance levels: P < 0.05 (*), P < 0.01 (**) (Tables S13 and S14). (B) Comparison of the DMAP1 gene expression in untreated Harmonia axyridis beetles of two invasive and two native populations relative to untreated individuals of a biocontrol strain. DMAP1 gene expression of the biocontrol strain is represented as baseline. Dots show mean values of 10 biological replicates (each comprising three males and three females). Error bars indicate standard error of the mean. Significant differences from biocontrol strain expressions are depicted by asterisks. Significance level: P < 0.001 (***) (Tables S15 and S16). ovaries. Further, maternal silencing of this gene caused a effects on survival and fecundity. We also showed that developmental arrest in offspring embryos (Schulz et al., DMAP1 was expressed at much higher levels in a biocon- 2018). Taking together these studies and our findings, we trol strain bred for increased fecundity compared to native postulate an evolutionarily conserved role of the and invasive populations. Further research is needed to DNMT1-DMAP1 complex beyond embryonic development determine how epigenetic mechanisms, and in particular including a role in ovary maturation and female fecundity. DNA methylation, promote the success of this highly inva- In our study, DMAP1 silencing did not affect male fecun- sive species. dity or the integrity of the testes. However, we observed a slight decrease in the number of eggs per batchwhenmock Experimental procedures control females (injected with water) were mated with males injected with DMAP1 dsRNA, suggesting that the Rearing and collection of H. axyridis DNMT1-DMAP1 complex plays aminor role in the develop- H. axyridis populations from Germany, France, Korea and China, ment of male reproductive organs. This is supported by the as well as a laboratory-bred biocontrol strain, were collected and fact that DNMT1 is strongly expressed not only in the ova- reared as previously described (Gegner et al., 2018). For mating, ries of N. lugens, but also in the testes of the re ant Sole- adult animals were paired in Petri dishes (94 × 16 mm) and fedfi ad libitumwith pea aphids (Acyrthosiphon pisum). All experiments nopsis invicta (Kay et al., 2018; Lu et al., 2018). were carried out with beetles 2–3 weeks of age representing the In order to gain further insight into the role of the DNA colour morph succinea. methylation machinery in the fecundity of H. axyridis, we compared the expression of DMAP1 in two invasive popu- Analysis of colour polymorphism at different temperatures lations, two native populations and a biocontrol strain, which reportedly differ in their reproductive capacity. The The effect of temperature on colour polymorphism was analysed latter was speci cally bred to achieve a higher daily repro- by separating F0 beetles into three breeding groups according tofi number of black spots on the elytra, namely low (0–7), middle duction rate for pest control applications (Tayeh et al., (8–15) and high (> 16) spot number, as previously described 2012; Tayeh et al., 2015). We found that DMAP1 mRNA (Michie et al., 2010). F1 eggs were maintained at 15, 21 or 28C. was 27% more abundant in the biocontrol strain compared The emerging F1 larvae were reared individually and the colour to native populations from Korea and China and 38%more morphs and spot groups of the resulting F1 beetles were recorded abundant compared to invasive populations from France separately for each breeding group. and Germany. These findings suggest that the DNMT1- DMAP1 complex helps to determine the fecundity and Identification of epigenetic regulators and preparation of fertility of this invasive species and may therefore partly dsRNA contribute to its invasive success. To identify putative H. axyridis HATs, HDACs and DMAP1, the To summarize, we found that the RNAi treatment of epi- corresponding amino acid sequences from Leptinotarsa decemli- genetic regulators can affect the life-history traits ofH. axyr- neata [National Center for Biotechnology Information (NCBI) idis. The knockdown of DMAP1 in particular led to severe accession numbers XP_023029651, XP_023012143 and © 2019 The Authors. Insect Molecular Biology published by JohnWiley & Sons Ltd on behalf of Royal Entomological Society., 29, 148–159 13652583, 2020, 2, Downloaded from https://resjournals.onlinelibrary.wiley.com/doi/10.1111/imb.12616 by Justus-Liebig-Universitat, Wiley Online Library on [01/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 156 J. Gegner et al. XP_023020626] and T. castaneum (NCBI accession numbers of 33 larvae per gene and 35 mock-larvae were monitored for XP_966673, XP_967425 and XP_008195930) were used as 2 months. To investigate the impact of parental RNAi in offspring TBLASTN queries with default parameter settings against our in- beetles, we injected one pair of F0 beetles from the low-spot- house H. axyridis transcriptome assemblies (Vilcinskas et al., number group (development temperature 21C) three times with 2013). Sequences with more than 55% amino acid sequence sim- HAT, HDAC or DMAP1 dsRNA, or with water as a mock-control. ilarity to queries were collected for further analysis. All protein The F1 generation was then reared at 15C and the colour pheno- sequences were aligned in GENEIOUS vR11 (Biomatters Ltd, Auck- type and life-history parameters were recorded. The experiment land, New Zealand) using MUSCLE with default settings. The results was terminated 3 months after injection. The survival of F0 beetles were inspected for regions of high-quality alignment and were was monitored until 35 dpi. Fecundity (number of egg batches and refined manually. During this step, candidates were also scruti- number of eggs per batch) was recorded until the last DMAP1 nized for the presence of conserved amino acid patterns. To find dsRNA-injected female beetle died (24 dpi). We also determined possible gene homologues for the final candidates of HAT, HDAC the hatching rate and the developmental time of F1 larvae. and DMAP1, a BLASTN v. 2.6.0 search (Altschul et al., 1997) was performed against the H. axyridis genome HaxR_v1.0 (Gautier Influence of DMAP1 knockdown on reproductive organs et al., 2018) with default settings and a word size of seven. We could not find any homologues for HAT and DMAP1 in the The influence of DMAP1 knockdown on reproductive organs was H. axyridis genome, as HATmapped only to utg 63 pilon from posi- investigated 10 dpi by dissecting testes and ovaries in tion 2308311 to 2310853 and DMAP1 only to utg_158_pilon from phosphate-buffered saline for analysis by light microscopy. In position 488535 to 490820 (File S1). For HDAC we found two females, we also recorded the numbers of ovarioles, follicles, highly similar homologues, HDAC 4 and 7. One copy mapped mature eggs and degenerate eggs. completely to utg_635_pilon from position 102105 to 141466. The other copy mapped to utg_2690_pilon from position 160556 Gene expression analysis to 169854 and to utg_2046_pilon from position 273426 to 274154 (Files S1 and S2). For each possible homologous RNA isolation, cDNA synthesis and qPCR for relative gene sequence a expression analysis were carried out as previously described,BLASTX (BLASTN v. 2.9.0) search was performed against the non-redundant protein sequences (nr) database using default using the ribosomal protein S3 gene (RPS3) for normalization (for- settings. To exclude off-target effects, the sequence of dsRNA ward primer 5’-GGCTACCAGAACCGACAGAG-3 0 and reverse 0 belonging to H. axyridis was split into all possible 21mers and a primer 5’-GTGCTATGGCGCATAATCCT-3 ; Gegner et al., 2018). search was performed against the H. axyridis genome DMAP1 gene knockdownmediated by dsRNAwas verified withBLASTN HaxR_v1.0 as described above. Hits were only de ned as poten- a total of three biological replicates per treatment and time point,fi tial off-targets if they had mismatches at position 1 and/or 21 and each consisting of 10 pooled individuals with equal sex ratio, in a showed 100% identity in the remaining 19mer. In order to de ne qPCR assay using forward primer 5’-TACCTGCAAATGTGGfi the position of the potential off-targets 10 kb sequences upstream GTC-3 0 and reverse primer 5’-GACCATGTCGCTTCTAAGTTCG- 0 and downstream of these regions were extracted from the 3 . For each replicate 20 beetles (10 males and 10 females) were genome. Ab initio gene prediction was performed using the injected with either DMAP1 dsRNA or water as control. After A pipeline implemented in the O B software 3 days five males and five females per treatment were flash frozenUGUSTUS MICS OX (BioBam Bioinformatics S.L., Valencia, Spain) and the F in liquid nitrogen and pooled for RNA isolation. 10 days after injec-GENESH HMM-based gene structure prediction and gene- nder tool with tion the procedure was repeated for the remaining beetles. Foldfi ΔΔCт standard settings (Salamov and Solovyev, 2000). No off-targets changes in gene expression relative to controls (2 ) (Pfaffl, could be found for our HAT or DMAP1 dsRNA constructs (File 2001; Schmittgen and Livak, 2008) were calculated and the S2, Figs S1 and S2). The possible small-interfering RNAs (siR- 3 and 10 dpi means were tested for statistical significance. NAs) resulting from the HDAC dsRNA (Fig. S3) were determined Population-specific differences in DMAP1 gene expression to address both on-targets described above, as well as two very were determined by qPCR amongst five untreated biological repli- short noncoding sections of the genome (utg_331_pilon from posi- cates representing two invasive populations from Germany and tion 1696692 to 1696710 and utg_893_pilon from position France, two native populations fromChina andKorea, and one bio- 3749566 to 3749584; Files S2 S4). We can thus exclude that the control strain. Six beetles (three males and three females) for each– selected siRNAs could potentially affect other, nontarget genes in treatment (RNAi and water controls) were pooled 10 dpi and the Harmonia axyridis. Utilizing this approach, we selected mRNA experiment was performed twice. For statistical analysis, the sequences for HAT, HDAC and DMAP1 that were then used as results for each invasive population (Germany and France) and templates for the preparation of dsRNAs (Table S17) using an each native population (China and Korea) were calculated and Ambion MEGAscript T7 kit (Applied Biosystems, Foster City, pooled. The populationmeanswere then tested against each other CA, USA). and against the biocontrol strain. Monitoring life parameters after the RNAi treatment of Statistical analysis epigenetic regulators Statistical analysis was carried out using R v. 3.3.3 (R Core We injected 1 μg dsRNA into adult beetles as previously described Team, 2015). The dunnTest function in package FSA (Gegner et al., 2018) or into L4 larvae laterally between the sixth v. 0.8.17 (Ogle, 2016) was used for Kruskal–Wallis multiple and seventh tergites. The life parameters and colour phenotypes comparisons with Bonferroni adjusted P-values to analyse © 2019 The Authors. Insect Molecular Biology published by JohnWiley & Sons Ltd on behalf of Royal Entomological Society., 29, 148–159 13652583, 2020, 2, Downloaded from https://resjournals.onlinelibrary.wiley.com/doi/10.1111/imb.12616 by Justus-Liebig-Universitat, Wiley Online Library on [01/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Silencing of the DMAP1 gene 157 the development and survival rate of the F1 progeny of Dobzhansky, T. (1924) Die geographische und individuelle Varia- dsRNA-injected F0 beetles and of dsRNA-injected L4 lar- bilität von Harmonia axyridis Pall. in ihren Wechselbeziehun- vae, as well as the development, fecundity and fertility of gen. Biologisches Zentralblatt, 44, 401–421. dsRNA-injected beetles, and the ovary parameters of Dobzhansky, T. (1933) Geographical variation in lady–beetles. DMAP1-RNAi females. The survival rate and the numbers The American Naturalist, 67, 97–126. Flores, K.B., Wolschin, F. and Amdam, G.V. (2013) The role of of animals per developmental stage of the F1 progeny of methylation of DNA in environmental adaptation. Integrative dsRNA-injected F0 beetles and dsRNA-injected L4 larvae  and Comparative Biology, 53, 359–372.that developed at 15 Cwere analysed by pairwise compar- Gautier, M., Yamaguchi, J., Foucaud, J., Loiseau, A., Ausset, A., isons using a two-sided Fisher’s exact test with Bonferroni Facon, B. et al. (2018) The genomic basis of color pattern poly- correction provided by the functions fisher.test and fisher. morphism in the harlequin ladybird. Current Biology, 2820, multcomp in package RVAideMemoire v. 0.9–68 (Hervé, 3296–3302. e7. 2017). The survival of F0 beetles reared at 21Cwas inves- Gayathri, D.S. and Harini, B.P. (2012) Adverse effect of valproic tigated using the R package survival v. 2.40–1 (Therneau, acid on mating behaviour and fertility in Drosophila melanoga- 2015) to plot Kaplan–Meier survival curves and to calculate ster. Bioscan, 7, 31–34. differences in survival rates between treatments with the Gegner, J., Baudach, A., Mukherjee, K., Halitschke, R., Vogel, H. log-rank test and Holm-correctedP-values. Statistical anal- and Vilcinskas, A. (2019) Epigenetic mechanisms are involved in sex-specific trans-generational immune priming in the lepi- ysis of qPCR data was carried out according to Gegner dopteran model host Manduca sexta. Frontiers in Physiology, et al. (2018), based on the ΔΔCT method established by 10, 137. Pfaffl (2001) and Schmittgen and Livak (2008). Gegner, T., Schmidtberg, H., Vogel, H. and Vilcinskas, A. (2018) Population–specific expression of antimicrobial peptides con- ferring pathogen resistance in the invasive ladybird Harmonia Acknowledgements axyridis. Scientific Reports, 8, 3600. We thank Dr RichardM. Twyman for editing themanuscript Glastad, K.M., Hunt, B.G., Yi, S.V. and Goodisman, M.A.D. (2011) DNAmethylation in insects: on the brink of the epigenomic era. and Dr Kwang-Zin Lee (Institute for Insect Biotechnology, Insect Molecular Biology, 20, 553–565. Justus-Liebig-University of Giessen) for his helpful advice Glastad, K., Hunt, B. and Goodisman, M. (2019) Epigenetics in and support regarding the experimental setups for the anal- insects: genome regulation and the generation of phenotypic ysis of the beetles’ reproductive organs. The authors diversity. Annual Review of Entomology, 64, 185–203. acknowledge funding provided by the German Research Hervé, M., 2017. Aide–memoire de statistique appliquee a la Foundation (DFG) via the European Research Area Net- biologie –Construire son etude et analyser les resutats a l’aide work (ERA-Net) project BiodivERsA (VI 219/7-1, VO du logiciel R. [online]. Available at: https://cran.r–project.org/ 84171) and by the excellence initiative of the Hessian Min- doc/contrib/Herve–Aide–memoire–statistique.pdf. [Accessed istry of Science, Higher Education and Art (HMWK) sup- 12 April 2017]. porting the LOEWE (Landes-Offensive zur Entwicklung Hosino, Y. (1940) Genetical studies on pattern types on the lady– Wissenschaftlich-ökonomischer Exzellenz) Centre for bird beetle Harmonia axyridis Pallas. Journal of Genetics, 40, 215–228. Insect Biotechnology and Bioresources. Kang, H.L., Benzer, S. and Min, K.T. (2002) Life extension in Drosophila by feeding a drug. Proceedings of the National References Academy of Sciences of the United States of America, 99, 838–843. Altschul, S.F., Madden, T.L., Schaffer, A.A., Zhang, J., Zhang, Z., Komai, T. (1956) Genetics of ladybeetles. Advances in Genetics, Miller, W. et al. (1997) Gapped BLAST and PSI-BLAST: a 8, 155–188. new generation of protein database search programs. Nucleic Jiang, W., Pan, X., Liu, J. and Liu, L. (2008) Studies on the type of Acids Research, 25, 3389–3402. spot on the wing ofHarmonia axyridis Pallas. in Baoding. Hebei Bewick, A., Sanchez, A., Mckinney, E., Moore, A., Moore, P. and Journal of Forestry and Orchard Research, 22, 198–202. Schmitz, R. (2019) Dnmt1 is essential for egg production and Jing, Y. and Zhang, Y. (2001) Studies on the type of spot on the embryo viability in the large milkweed bug, Oncopeltus fascia- wing of Harmonia axyridis Pallas. in Shanxi Province. Journal tus. Epigenetics & Chromatin, 12, 6. of Shanxi Agricultural University, 4, 42–45. Bingsohn, L., Knorr, E. and Vilcinskas, A. (2016) The model beetle Jurkowska, R.Z., Jurkowski, T.P. and Jeltsch, A. (2011) Structure Tribolium castaneum can be used as an early warning system and function of mammalian DNAmethyltransferases. Chem- for transgenerational epigenetic side effects caused by phar- biochem, 12, 206–222. maceuticals. Comparative Biochemistry and Physiology, Part Kay, S., Skowronski, D. and Hunt, B.G. (2018) Developmental C: Toxicology & Pharmacology, 185–186, 57–64. DNA methyltransferase expression in the fire ant Solenopsis DeWitt, T.J., Sih, A. and Wilson, D.S. (1998) Costs and limits of invicta. Insect Science, 25, 57–65. phenotypic plasticity. Trends in Ecology & Evolution, 13, Knapp,M. andNedvěd, O. (2013) Gender and timing during ontog- 77–81. enymatter: effects of a temporary high temperature on survival, © 2019 The Authors. Insect Molecular Biology published by JohnWiley & Sons Ltd on behalf of Royal Entomological Society., 29, 148–159 13652583, 2020, 2, Downloaded from https://resjournals.onlinelibrary.wiley.com/doi/10.1111/imb.12616 by Justus-Liebig-Universitat, Wiley Online Library on [01/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 158 J. Gegner et al. body size and colouration in Harmonia axyridis. PLoS ONE, 8, Tayeh, A., Hufbauer, R.A., Estoup, A., Ravigné, V., Frachon, L. e74984. and Facon, B. (2015) Biological invasion and biological control Lee, G.E., Kim, J.H., Taylor, M. and Muller, M.T. (2010) DNA select for different life histories. Nature Communications, 6, methyltransferase 1-associated protein DMAP1. is a co- 7268. repressor that stimulates DNA methylation globally and locally Therneau, T., 2015. A Package for Survival Analysis in S. version at sites of double strand break repair. The Journal of Biological 2.38. [online] Avialable at: https://CRAN.R–project.org/ Chemistry, 285, 37630–37640. package=survival. [Accessed 13 June 2017]. Levins, R. (1968) In: Evolution in Changing Environments. Vilcinskas, A. (2017) The impact of parasites on host insect epige- Princeton: Princeton University Press. netics. Adv in Insect Physiol, 53, 145–164. Lu, K., Chen, X., Li, W., Li, Y., Zhang, Z. and Zhou, Q. (2018) Insu- Vilcinskas, A. and Vogel, H. (2016) Seasonal phenotype-specific lin-like peptides and DNA/tRNA methyltransferases are transcriptional reprogramming during metamorphosis in the involved in the nutritional regulation of female reproduction in European map butterfly Araschnia levana. Ecology and Evolu- Nilaparvata lugens (Stål). Gene, 639, 96–105. tion, 6, 3476–3485. Marks, P.A., Miller, T. and Richon, V.M. (2003) Histone deacety- Vilcinskas, A., Mukherjee, K. and Vogel, H. (2013) Expansion of lases. Current Opinion in Pharmacology, 3, 344–351. the antimicrobial peptide repertoire in the invasive ladybirdHar- Michie, L.J., Mallard, F., Majerus, M.E. and Jiggins, F.M. (2010) monia axyridis. Proceedings Biological Sciences, 280, Melanic through nature or nurture: genetic polymorphism and 20122113. phenotypic plasticity in Harmonia axyridis. Journal of Evolu- Yuan, R.C., Zhang, F.M., Wen, G.Z., Yu, M., Wang, X.Q., Ma, F.C. tionary Biology, 23, 1699–1707. et al. (1994) Color pattern of Harmonia axyridis in Changbai Ogle, D.H. (2016) In: Introductory Fisheries Analyseswith R. Boca Mountain. Jilin Agricultural Science, 4, 45–54. Raton: Chapman & Hall/CRC. Zhang, J., Xing, Y., Li, Y., Yin, C., Ge, C. and Li, F. (2015) DNA Pfaffl, M.W. (2001) A newmathematical model for relative quantifi- methyltransferases have an essential role in female fecundity cation in real-time RT-PCR. Nucleic Acids Research, 29, in brown planthopper, Nilaparvata lugens. Biochemical and e45–e445. Biophysical Research Communications, 464, 83–88. R Core Team, 2015. R: A Language and Environment for Statisti- Zhao, Y., Sun, H., Lu, J., Li, X., Chen, X., Tao, D. et al. (2005) Life- cal Computing. R Foundation for Statistical Computing. span extension and elevated hsp gene expression inDrosoph- [online] Available at: https://www.R–project.org. [Accessed ila caused by histone deacetylase inhibitors. The Journal of 11 April 2017]. Experimental Biology, 208, 697–705. Rountree, M.R., Bachman, K.E. and Baylin, S.B. (2010) DNMT1 binds HDAC2 and a new co-repressor, DMAP1, to form a com- Supporting Information plex at replication foci. Nature Genetics, 25, 269–277. Additional supporting information may be found online in Salamov, A.A. and Solovyev, V.V. (2000) Ab initio gene the Supporting Information section at the end of the finding in Drosophila genomic DNA. Genome Res, 10, article. 516–4522. Schmittgen, T.D. and Livak, K.J. (2008) Analyzing real-time PCR File S1.BLASTN results for messenger RNA of histone acetyltransferase, his- data by the comparative CT method. Nature Protocols, 3, tone deacetylase and DNA methyltransferase 1-associated protein 1 against the Harmonia axyridis genome HaxR_v1.0. 1101–1108. Simpson, S.J., Sword, G.A. and Lo, N. (2011) Polyphenism in File S2. BLASTX results of putative histone deacetylase homologues against insects. Current Biology, 21, R738 R749. the National Center for Biotechnology Information nonredundant protein– sequences (nr) database. Schulz, N.K.E., Wagner, C.I., Ebeling, J., Raddatz, G., Diddens- de Buhr, M.F., Lyko, F. et al. (2018) Dnmt1 has an essential File S3. BLASTN results for 21mers of double-stranded RNA of histone acet- function despite the absence of CpG DNA methylation in the yltransferase, histone deacetylase and DNA methyltransferase 1-associated protein 1 against the Harmonia axyridis genome HaxR_v1.0. red flour beetle Tribolium castaneum. Scientific Reports, 8, 16462. File S4. Results of the gene prediction for utg_331_pilon from position Soh, J.W., Marowsky, N., Nichols, T.J., Rahman, A.M., Miah, T., 1696 692 to 1696710. Sarao, P. et al. (2013) Curcumin is an early-acting stage- File S5. Results of the gene prediction for utg_893_pilon from position specific inducer of extended functional longevity inDrosophila. 3749566 to 3749584. Experimental Gerontology, 48, 229–239. Table S1. Number of individuals (n), mean, median, standard deviation Tan, C.C. (1946) Mosaic dominance in the inheritance of color pat- (sd) and standard error of the mean (se) for development from double-  terns in the lady-bird beetle, Harmonia axyridis. Genetics, 31, stranded RNA- or mock-injected stage four (L4) larvae at 4 C. 195–210. Table S2. Number of double-stranded RNA- or mock-injected L4 larvae Tan, C.C. and Li, J.C. (1934) Inheritance of the elytral color pat- [n (injected)], survived animals in total [n (alive)], beetles only [n (beetles)], or undeveloped animals [n (L4 larvae)], as well as their survival rate (%). terns of the lady-bird beetle, Harmonia axyridis Pallas. The American Naturalist, 68, 252 265. Table S3. Statistical results for the development of double-stranded RNA-– or mock-injected L4 larvae, reared at 4C, using Kruskal–Wallis multiple Tayeh, A., Estoup, A., Laugier, G., Loiseau, A., Turgeon, J., comparison analysis with Bonferroni corrected P-values. Significance Toepfer, S. et al. (2012) Evolution in biocontrol strains: insight levels: P < 0.05 (*), P < 0.01 (**), P < 0.001 (***). from the harlequin ladybird Harmonia axyridis. Evolutionary Table S4. Statistical results for survival, number (n) of pupae, beetles and Applications, 5, 481–488. undeveloped larvae of double-stranded RNA- or mock-injected L4 larvae, © 2019 The Authors. Insect Molecular Biology published by JohnWiley & Sons Ltd on behalf of Royal Entomological Society., 29, 148–159 13652583, 2020, 2, Downloaded from https://resjournals.onlinelibrary.wiley.com/doi/10.1111/imb.12616 by Justus-Liebig-Universitat, Wiley Online Library on [01/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Silencing of the DMAP1 gene 159 reared at 4C, using Fisher’s exact test for count data with Bonferroni cor- DNA methyltransferase 1-associated protein 1 after knockdown relative to rection for significance levels: P < 0.05 (*), P < 0.01 (**), P < 0.001 (***). mock control at 3 and 10 days postinjection. Table S5. Number of individuals (n), mean, median, standard deviation Table S14.Results of the multiple comparison of means analysis, utilizing a (sd) and standard error of the mean (se) for development and survival from one-way analysis of variance and simultaneous tests for general linear eggs to beetles of offspring of double-stranded RNA- or mock-injected bee- hypotheses, for DNA methyltransferase 1-associated protein 1 (DMAP1) tles at 4C. gene expression of DMAP1-injected beetles and mock controls 3 and Table S6. Statistical results for the development of development of off- 10 days postinjection. Significance levels: P < 0.05 (*), P < 0.01 (**), spring, reared at 4C, of double-stranded RNA- or mock-injected beetles P < 0.001 (***). using Kruskal–Wallis multiple comparison analysis with Bonferroni adjusted Table S15. Number of replicates (n), mean, standard deviation (sd) and P-values. Significance levels: P < 0.05 (*), P < 0.01 (**), P < 0.001 (***). standard error of the mean (se) for fold change in gene expression of Table S7. Statistical results for survival, number (n) of pupae and beetles of DNA methyltransferase 1 associated protein 1 relative to biocontrol the first filial generation (F1), reared at 4C, of double-stranded RNA- or strain. mock-injected parents using Fisher’s exact test for count data with Bonfer- Table S16.Results of the multiple comparison of means analysis, utilizing a roni correction for significance levels: P < 0.05 (*), P < 0.01 (**), one-way analysis of variance and simultaneous tests for general linear P < 0.001 (***). hypotheses, for DNA methyltransferase 1-associated protein 1 gene Table S8. Statistics for survival curves of double-stranded RNA- or mock- expression of native, invasive and biocontrol strains. Significance levels: injected beetles reared at 21C using log-rank test and Holm-P-adjustment. P < 0.05 (*), P < 0.01 (**), P < 0.001 (***). Table S9.Mean, median and standard deviation (sd) for fecundity and fertil- Table S17. Gene-specific primers without the T7 polymerase promoter ity parameters of double-stranded RNA- or mock-injected beetles, reared sequence at the 5 0 end (50-TAA TACGAC TCACTA TAGGGAG-30) gener- at 21C. ating double-stranded RNA of epigenetic regulators and the length of result- ing PCR products: histone acetyltransferase, histone deacetylase and DNA Table S10.Statistical results for fecundity and fertility parameters of double- methyltransferase 1-associated protein 1. stranded RNA- or mock-injected beetles, reared at 21C, using Kruskal– 0 0 Wallis multiple comparison analysis with Bonferroni adjusted P-values. Sig- Figure S1. Sense strand sequence (5 to 3 ) of the histone acetyltransferase ni cance levels: P < 0.05 (*), P < 0.01 (**), P < 0.001 (***). messenger RNA, including forward (dark green) and reverse (light green)fi primer for double-stranded RNA (dark blue) synthesis. Table S11.Number (n), mean, standard deviation (sd) and standard error of 0 0 the mean (se) of ovarioles, follicles, mature and degenerated eggs from Figure S2. Sense strand sequence (5 to 3 ) of the DNA methyltransferase mock- or DNA methyltransferase 1-associated protein 1-injected females 1-associated protein 1 messenger RNA, including forward (dark green) and at 10 days postinjection, reared at 21C. reverse (light green) primer for double-stranded RNA (dark blue) synthesis. Table S12.Statistical results of the one-way analysis of variance comparing Figure S3. Sense strand sequence (5 0 to 30) of the histone deacetylase ovaries of DNA methyltransferase 1-associated protein 1- and mock- messenger RNA, including forward (dark green) and reverse (light green) injected females, reared at 21C. Signi cance levels: P < 0.05 (*), primer for double-stranded RNA (dark blue) synthesis.fi P < 0.01 (**), P < 0.001 (***). Figure S4. Testicles of (A) mock- and (B) DNA methyltransferase Table S13. Number of replicates (n), mean, standard deviation (sd), and 1-associated protein 1-injected Harmonia axyridis males at 10 days standard error of the mean (se) for fold changes in gene expression of postinjection. © 2019 The Authors. Insect Molecular Biology published by JohnWiley & Sons Ltd on behalf of Royal Entomological Society., 29, 148–159 13652583, 2020, 2, Downloaded from https://resjournals.onlinelibrary.wiley.com/doi/10.1111/imb.12616 by Justus-Liebig-Universitat, Wiley Online Library on [01/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License