A long-standing problem of developmental biology is how body size is determined. In Drosophila melanogaster, the insulin/insulin-like growth factor (I/IGF) and target of rapamycin (TOR) signaling pathways play important roles in this process. However, the detailed mechanisms by which insect body growth is regulated are not known. Therefore, we have attempted to utilize systemic nymphal RNA interference (nyRNAi) to knockdown expression of insulin signaling components including Insulin receptor (InR), Insulin receptor substrate (chico), Phosphatase and tensin homologue (Pten), Target of rapamycin (Tor), RPS6-p70-protein kinase (S6k), Forkhead box O (FoxO) and Epidermal growth factor receptor (Egfr) and observed the effects on body size in the Gryllus bimaculatus cricket. We found that crickets treated with double-stranded RNA (dsRNA) against Gryllus InR, chico, Tor, S6k and Egfr displayed smaller body sizes, while Gryllus FoxO nyRNAi-ed crickets exhibited larger than normal body sizes. Furthermore, RNAi against Gryllus chico and Tor displayed slow growth and RNAi against Gryllus chico displayed longer lifespan than control crickets. Since no significant difference in ability of food uptake was observed between the Gryllus chiconyRNAi nymphs and controls, we conclude that the adult cricket body size can be altered by knockdown of expressions of Gryllus InR, chico, Tor, S6k, FoxO and Egfr by systemic RNAi. Our results suggest that the cricket is a promising model to study mechanisms underlying controls of body size and life span with RNAi methods.
One of the essential problems in developmental biology is how organisms attain correct body size within their lifespan. It has been recently postulated that insect body size is controlled via the insulin-like growth factor (IGF) and target of rapamycin (TOR) signaling pathways, which coordinate nutrition and cell growth, as well as steroid and neuropeptide hormones (for review see Oldham & Hafen 2003; Edgar 2006; Mirth & Riddiford 2007). In Drosophila, when the insulin receptor (dInR) is activated by insulin-like peptide binding, the IGF signaling pathway is initiated. Activated dInR activates the insulin receptor substrate (IRS), which is encoded by chico. IRS and the adaptor protein p60 recruit phosphatidylinositide 3-kinase (PI3K) to the cell membrane, which then phosphorylates phosphatidylinositol 4,5-bisphosphate (PIP2) to become PI 3,4,5-trisphosphate (PIP3). The reverse reaction is catalyzed by phosphatase and tensin homologue (PTEN), thereby inhibiting insulin signaling. Increased levels of PIP3 cause phosphoinositol-dependent kinase (PDK) to bind to the cell membrane, via its PH domain, where it activates Akt (or protein kinase B-PKB). Akt promotes growth by inhibiting forkhead box, class O (dFOXO), inhibiting the tuberous sclerosis dimmer (TSC1/TSC2), and promoting the insertion of glucose transporters into the cell membrane. IGF signaling is the primary nutrient sensing mechanism and mediates its growth-regulatory functions largely through signaling pathways (Hietakangas & Cohen 2009). However, the main mediator of cellular nutrient sensing is the protein kinase TOR. TOR receives information regarding levels of cellular amino acids and energy, and regulates the activity of processes involved in cell growth, such as protein synthesis and autophagy, through ribosomal protein S6 kinase (S6k) (Mirth & Riddiford 2007; Hietakangas & Cohen 2009). Recent studies showed that flies deficient for insulin-like peptides are small and severely growth-delayed (Zhang et al. 2009) and that flies carrying mutations in dInR exhibited delayed development and reduced growth (Chen et al. 1996; Shingleton et al. 2005). Furthermore, Drosophila mutants for chico, a Drosophila homologue of vertebrate IRS1-4, were reportedly less than half the size of wild-type flies, and had fewer and smaller cells (Bohni et al. 1999). Loss of the homologue of the tumor suppressor gene, Pten, results in increased PIP3 levels, which are sufficient to compensate for the complete loss of the Insulin/insulin-like growth factor receptor function (Oldham et al. 2002). This reduction of Pten activity is also sufficient to vastly increase organism size. In addition, dFOXO has been shown to be a crucial mediator of insulin signaling in Drosophila, and is needed for the reduction in cell number observed in insulin-signaling mutants (Junger et al. 2003). However, dFOXO null alleles are viable and of normal size. On the other hand, Layalle et al. (2008) found that when the activity of TOR is reduced in the prothoracic gland (PG), the ecdysone peak that marks the end of larval development is abrogated, resulting in extension of the growth duration and an increase in animal size. In addition, it is also known that the epidermal growth factor receptor (EGFR) pathway is used multiple times during organ development and the spatial and temporal features of its signaling can be modified to fit a particular developmental size setting (Shilo 2005). Although the aforementioned signaling pathways are known to be involved in the size control, little is known regarding how body size is actually sensed, or how organ-intrinsic size is controlled.
To elucidate the mechanisms underlying body size control in detail, we have utilized the cricket Gryllus bimaculatus as a novel model organism, because systemic nymphal RNA interference (nyRNAi) can be used effectively in G. bimaculatus embryos (Mito et al. 2007) and nymphs (Nakamura et al. 2008a). G. bimaculatus enlarges via ecdysis and undergoes a gradual, hemimetabolous metamorphosis after the eighth nymphal stage, having an approximate lifespan of 2 months. Its adult organs are formed without pupal formation. Although these characteristics are different from those of holometabolous insects such as Drosophila, the mechanisms involved in body size control and lifespan are likely to be conserved in the cricket. First we examined whether nyRNAi is applicable to study mechanisms of body size control. We cloned Gryllus orthologues of Drosophila genes involved in signaling pathways regulating body weight, such as Gb’InR, Gb’chico, Gb’Pten, Gb’Tor, Gb’S6k, Gb’FoxO and Gb’Egfr. We then performed nyRNAi experiments against those genes to look at their effects on body weight, with the expectation that RNAi against Gb’InR, Gb’chico or Gb’Egfr would result in a decrease in body weight, while RNAi against Gb’Pten or Gb’FoxO would result in an increase in body weight. We found that RNAi against Gb’InR, Gb’chico, Gb’Tor, Gb’S6k or Gb’Egfr reduces body weight, while Gb’FoxO RNAi increases body weight. Unexpectedly, RNAi against Gb’Pten caused no significant effect on body weight. These results indicate that cricket body weight is likely to be altered by systemic RNAi even in nymphal stages, indicating that the cricket is a promising model to study mechanisms underlying controls of body size and lifespan with RNAi methods.
Materials and methods
Rearing of crickets
Crickets (Gryllus bimaculatus) were reared at 28°C in 40% humidity. After injection of double-stranded RNA (dsRNA), nymphs and adults were separated and kept in plastic dishes and cages, respectively, unless otherwise stated (Mito & Noji 2008).
Cloning of Gryllus orthologues of Drosophila InR, chico, Pten, Tor, S6k, FoxO and Egfr cDNAs
Partial cDNA fragments of Gryllus Insulin-like receptor (Gb’InR), chico (Gb’chico), Pten (Gb’Pten), Target of rapamycin (Gb’Tor), ribosomal protein S6 kinase (Gb’S6k), Forkhead box sub-group O (Gb’FoxO) orthologues were cloned as follows. Total RNA was extracted from Gryllus bimaculatus at various embryonic stages using Isogen (Nippon-Gene, Toyama, Japan). mRNA was isolated using an OligotexTM-dT30 Super mRNA Purification Kit (TaKaRa, Kyoto, Japan). cDNA was synthesized using the Superscript First Strand Synthesis Kit (Life Technologies, CA, USA). To isolate a Gb’InR, Gb’chico, Gb’Pten, Gb’Tor, Gb’S6k and Gb’FoxO cDNA fragments by polymerase chain reaction (PCR), we used the degenerate or gene-specific primers. The sequences of primers were as follows:
Double-stranded RNAs (dsRNA) for Gb’InR (386 bp), Gb’chico (438 bp), Gb’Pten (374 bp), Gb’Tor (269 bp), Gb’S6k (497 bp), Gb’FoxO (123 bp), Gb’Egfr (430 bp, spanning the cysteine rich domain) and Discosoma red fluorescent protein DsRed (660 bp), derived from the pDsRed2-N1 (Clontech, CA, USA) were synthesized using the MEGAscript High Yield Transcription Kit (Ambion Japan, Tokyo, Japan), as described previously (Miyawaki et al. 2004). The final concentration of dsRNA was adjusted to 20 μmol/L. RNAi experiments were performed with the cricket nymph as described previously (Nakamura et al. 2008a). For nyRNAi, cricket nymphs were used immediately after molting in the third, fifth and seventh instars. Defined volume of dsRNA solution was injected into the hemocoel of the abdomen of nymphs; 210 nL for third instar, 350 nL for fifth instar and 490 nL for seventh instar nymphs. In all RNAi experiments, DsRed dsRNA was used as a negative control. Nymphs were reared in a mass after hatching. Nymphs treated with dsRNA were kept separately in plastic dishes to measure body weight, food consumption and ecdysis period just after each molt up to the adult stage. After becoming adult, control and RNAi treated adults were kept in plastic cages and lifespans were determined.
Total RNA was extracted from control and nyRNAi-ed crickets at the third, fifth, sixth and seventh instar, 2 days after dsRNA injection, using the Isogen (Nippon-Gene). Total RNA (3–5 μg) was reverse-transcribed to cDNA using the SuperScript First-Strand Synthesis System (Invitrogen) and oligo dT primers (Invitrogen) according to the manufacturer’s instructions. Quantitative-PCR (q-PCR) was performed with the ABI 7900 Real-Time PCR System (Applied Biosystems) as described previously (Nakamura et al. 2008b; Bando et al. 2009). β-Actin was used for normalization of each transcript. The sequences of the q-PCR primers are as follows (forward and reverse, 5′ to 3′):
β-actin, TTGACAATGGATCCGGAATGT and AAAACTGCCCTGGGTGCAT
Gb’InR, TGTATGAAGGTATTGGCCGAGAT and CTCGCTTCGCTCACGATCA
Gb’chico, GGTTGGAGTATTATGACAGTGAGAAAAG and TGAAACAAGTCTTCAAAGCAATGC
Gb’Pten, GGCAACGAAGATGGAGAATTG and GGATGGCGAGAAACAACTTGA
Gb’Tor, TTGCCCTTGGACCCTCAGT and TTGGGTTCTCGGTGGAATTC
Gb’S6k, ACAGGAGCACCACCATTCACT and CTTGCATCAGGCGTCAGGTA
Gb’FoxO, CTCCGCCGGCTGGAA and TCATCCACCACGACGACTTG
Estimation of ability of food uptake
To estimate ability of food uptake (AFU), we measured food weight (F) that each cricket consumed every day and ecdysis period (EP) after each molt. Then, we calculated the amount of food consumed per body weight per ecdysis period (F/W/EP).
Gryllus orthologues of InR, chico, Pten, Tor, S6k, FoxO and Egfr
Partial cDNA fragments of Gryllus orthologues of Gb’InR, Gb’chico, Gb’Pten, Gb’Tor, Gb’S6k, Gb’FoxO and Gb’Egfr were cloned by a PCR method. Since each cloned fragment has the highest homology to the corresponding gene of Drosophila and other organisms, shown by sequence analysis with BLAST program, we presumed that these partial fragments encode Gryllus orthologues of respective Drosophila genes. A 386 bp cDNA fragment of Gb’InR cloned (128 aa) encodes the protein tyrosine kinase catalytic domain (PTKc). This amino acid sequence has 70% and 76% homology to the sequences of Drosophila and human InR, respectively (Fig. 1). A 438 bp cDNA fragment of Gb’chico (146 aa) was obtained, encoding the pleckstrin homology (PH) domain and the phophotyrosine-binding (PTB) domain, both of which are essential for insulin-like growth factor (IGF) signaling via phosphatidylinositol-3-phosphate (PI2P) (Fig. 1). The amino acid sequence of this fragment has 47% and 42% homology to the sequences of Drosophila Chico and human IRS-1, respectively. The 490 bp cDNA fragment of Gb’Pten contains the 77 bp 5′UTR and 413 bp coding region encoding 137 amino acid residues. This fragment, encoding the protein tyrosine phosphatase catalytic (PTPc) domain, exhibits 65% and 68% homology to the sequences of Drosophila and human Pten, respectively (Fig. 1). A 269 bp cDNA fragment of Gb’Tor encodes 89 aa and has 51% and 67% homology to the sequences of Drosophila Tor and human TOR, respectively (Fig. 1). A 497 bp cDNA fragment of Gb’S6k encodes 189 aa containing the serine/threonine kinase catalytic domain and has 67% and 62% homology to the sequences of Drosophila S6k and human S6k, respectively (Fig. 1). A 330 bp cDNA fragment of Gb’FoxO (110 aa) encodes the Forkhead (FH) transcriptional regulation domain. This sequence shows 86% and 82% homology to the sequences of Drosophila and human FoxO, respectively (Fig. 1). The 430 bp cDNA fragment of Gb’Egfr (143 aa) encodes the cysteine-rich domain and exhibits 66% and 43% homology to the corresponding domains of Drosophila and mouse EGFR, respectively (Nakamura et al. 2008b).
Effects of Nymphal RNAi on expression ofGb’InR, Gb’chico, Gb’Pten, Gb’Tor, Gb’S6k and Gb’FoxO
To determine whether knockdown of expression of Gb’InR, Gb’chico, Gb’Pten, Gb’Tor, Gb’S6k or Gb’FoxO would alter cricket body weight, we performed nyRNAi experiments (Nakamura et al. 2008b). As a negative control, we injected dsRNA constructed from cDNA encoding DsRed (Miyawaki et al. 2004; Mito et al. 2007). At first, we measured the amounts of endogenous mRNAs of nyRNAi-ed cricket nymphs by q-PCR at 2 days after injection of dsRNA and were compared with those of negative controls (n = 4). The average ratio of endogenous mRNAs at the third instar was lowered to 0.35 ± 0.09 (n = 4, P < 0.01) in the RNAi-ed nymphs against Gb’InR, to 0.42 ± 0.1 (n = 4, P < 0.01) against Gb’chico, to 0.21 ± 0.09 (n = 4, P < 0.01) against Gb’Pten, to 0.29 ± 0.02 (n = 4, P < 0.01) against Gb’Tor, to 0.10 ± 0.03 (n = 3, P < 0.01) against Gb’S6k and to 0.34 ± 0.07 (n = 3, P < 0.01) against Gb’FoxO, (in triplicate, ± standard deviation. P-values were calculated by Student’s t-test unless otherwise noted), indicating that the nyRNAi treatments were effective (Fig. 2). Next, we measured dependence of the nyRNAi efficiency on the nymphal stage. The average ratio of the amount of Gb’chico mRNA was lowered to 0.42 ± 0.05 (n = 3, P < 0.01) at the fifth instar and to 0.44 ± 0.04 (n = 4, P < 0.01) at the seventh instar, indicating that nyRNAi occurs independent of the nymphal stage. We also estimated duration of the nyRNAi effects. After injection of Gb’chico dsRNA at the third instar, the relative ratio of endogenous Gb’chico mRNA at the sixth instar (2 weeks after injection of dsRNA) was lowered to 0.37 ± 0.03 (n = 3, P < 0.01) (Fig. 2), indicating that duration of the nyRNAi effects is longer than 2 weeks.
Effects of nymphal RNAi against Gryllus InR, chico, Tor, S6k and FoxO on body weight, instar duration and lifespan
The body weight, instar duration (ecdysis periods) and lifespan of RNAi-ed and control DsRednyRNAi nymphs were measured at each molt after injection of dsRNA for each gene at the fifth instar (Fig. 3A–C). RNAi-ed crickets against insulin signaling components were found to be viable. We observed significant reduction of the body weight of Gb’InRnyRNAi, Gb’chiconyRNAi, Gb’TornyRNAi or Gb’S6knyRNAi animal from the sixth instar to adult: The average body weight of Gb’chiconyRNAi adults (415 ± 18 mg, n = 9) was reduced to 50.0% (P < 0.01) of that of DsRednyRNAi adults (831 ± 19 mg, n = 19), that of Gb’InRnyRNAi adults (651 ± 51 mg [n = 5]) to 78.3% (P < 0.01), that of Gb’TornyRNAi adults (651 ± 101 mg [n = 7]) to 78.4% (P < 0.01) and that of Gb’S6knyRNAi adults (677 ± 96 mg [n = 14]) to 81.4% (P < 0.01; Fig. 3A). In contrast, although the change in the average body weight of the Gb’FoxOnyRNAi nymphs was not significant, that of the corresponding adults (942 ± 40 mg, n = 7) increased significantly to 113.4% (P < 0.01) of that of DsRednyRNAi adults (Fig. 3A). Furthermore, we performed the similar experiments with Gb’PtennyRNAi crickets at the third instar. Since Pten is known to inhibit insulin signaling (Mirth & Riddiford 2007), the body weight of Gb’PtennyRNAi crickets was expected to be larger than controls. Although the average body weight of adult Gb’PtennyRNAi crickets (863 ± 31 mg, n = 21) was estimated to be 104% of that of the negative controls (824 ± 22 mg, n = 33), this difference was judged to be insignificant (P > 0.05; Fig. S1A).
We also observed the nyRNAi effects on instar duration and lifespan. The instar duration of Gb’chiconyRNAi or Gb’TornyRNAi nymphs from the fifth to eighth instar was significantly longer than that of the DsRednyRNAi nymphs. For Gb’chiconyRNAi nymphs, it took an average 27.1 ± 0.6 days (P < 0.01) to grow from the fifth instar to adult, and for Gb’TornyRNAi nymphs, it took 31.3 ± 2.9 days (P < 0.01; Fig. 3B), whereas for DsRednyRNAi nymphs, it took an average 21.8 ± 0.3 days. Instar duration of Gb’InRnyRNAi nymphs from the fifth instar to adult was 22.4 ± 0.7 days, that of Gb’S6knyRNAi nymphs is 22 ± 0.4 days and that of Gb’FoxOnyRNAi nymphs is 22 ± 0.3 days, which were not significantly different from that of DsRednyRNAi nymphs (P > 0.05; Fig. 3B). We estimated the mean adult lifespan (days) as the day at survival rate = 50% after becoming adult. The mean lifespan of Gb’chiconyRNAi adults was estimated to be 83.3 days, which was 1.4 times longer (P < 0.01) than the 59.3-day lifespan of the DsRednyRNAi adults (Fig. 3C). The maximum lifespans of DsRednyRNAi and Gb’chiconyRNAi adults were 98 and 123 days, respectively. The mean lifespan of Gb’InRnyRNAi adults was estimated to be 49 days, that of Gb’TornyRNAi adults to be 48 days, that of Gb’S6knyRNAi adults to be 62 days and that of Gb’FoxOnyRNAi adults to be 79 days, all of which were not significant (P > 0.05; Fig. 3C). We also observed effects of nyRNAi against Gb’Pten on instar duration and found that the duration of Gb’PtennyRNAi nymphs from the third instar to adult was 27 ± 0.3 days, which was not significantly different from that of DsRednyRNAi nymphs (P > 0.05; Fig. S1).
Effects of nymphal RNAi for Gryllus chico on body weight and lifespan
Since we found that expression of Gb’chico regulates body weight, instar duration and lifespan, based on our screening results, we focus mainly on the functions of Gb’chico hereafter. To further explore functions of Gb’chico, we performed nyRNAi experiments against Gb’chico at different nymphal stages in comparison with control experiments with DsRednyRNAi and uninjected-normal crickets. We found that there is no significant difference between body weight of DsRednyRNAi and normal crickets. When Gb’chico dsRNA was injected at the third instar, a significant reduction of body weight was observed after becoming the seventh instar (Fig. 4A). The average adult body weight of Gb’chiconyRNAi crickets was 73 ± 3% of that of DsRednyRNAi or 74 ± 3% that of normal crickets (P < 0.05; Fig. 4A). When injected at the seventh instar, significant reduction of body weight was observed after becoming the eighth instar. The average adult body weight of Gb’chiconyRNAi crickets was 74 ± 1% of that of the DsRednyRNAi and 75 ± 1% of that of normal animals (P < 0.01; Fig. 4C). In order to exclude possibilities that nyRNAi affects ability of food uptake directly, we measured the amount of food consumed per ecdysis period. Then, to estimate ability of food uptake (AFU), we calculated the amount of food consumed per body weight per ecdysis period (F/W/EP) (Fig. 4B,D). Although the AFU decreased at the eighth instar and adult in all Gb’chiconyRNAi, DsRednyRNAi and normal crickets, we found that there was no significant difference between Gb’chiconyRNAi crickets and DsRednyRNAi or normal controls (Fig. 4B,D), indicating that the AFU is not affected by RNAi experiments in either category shown in the present study.
To examine whether nyRNAi against Gb’chico affects lifespan in the adult stage, we estimated the lifespan of Gb’chiconyRNAi crickets (Fig. 4E,F). Even when dsRNA was injected at the third instar, the mean lifespan of Gb’chiconyRNAi crickets was estimated to be 90 days, which was longer than the lifespan of the DsRednyRNAi (78 days) or normal (79 days) crickets (Fig. 4E). When injected at the seventh instar, the mean lifespan of Gb’chiconyRNAi crickets was estimated to be a 61-day lifespan, which is longer than the 22-day lifespan observed for DsRednyRNAi (38 days) or normal (22 days) animals (Fig. 4F). The lifespans (22–38 days) of the DsRednyRNAi and normal crickets injected dsRNA at the seventh instar are much shorter than those (78–79 days) of the corresponding animals injected at the third instar. The difference appears to depend on a period of mass-culturing before injection of dsRNA, i.e., the period for the seventh-instar injection is longer than that for the third-instar injection (see Discussion). Since the AFU values were not significantly different among groups (P > 0.05), we concluded that the reduction of body weight and extension of lifespan observed in Gb’chiconyRNAi crickets were due to a defect in the insulin signaling pathway (Fig. 4).
Effects of nymphal RNAi against Gryllus chico on males and females
In Drosophila, chico mutant females exhibit long lifespan but mutant males are slightly short-lived (Clancy et al. 2001). To examine whether nyRNAi against Gb’chico affects body weight, instar duration or lifespan in a sex-specific manner, we compared the corresponding data between males and females of Gb’chiconyRNAi crickets after injection of dsRNA at the third instar. We found that the body sizes of both Gb’chiconyRNAi male and female adults were reduced (Fig. 5A). The body weights of adult Gb’chiconyRNAi males and females were 591 ± 18 mg (n = 19) and 666 ± 21 mg (n = 16), respectively, while those of DsRednyRNAi males and females were 795 ± 21 mg (n = 26) and 892 ± 42 mg (n = 13), respectively, which corresponds to reduction of the control body weight by 26% (P < 0.01) and 25% (P < 0.01) for males and female crickets, respectively (Fig. 6A), indicating that there is no significant difference between male and female (P > 0.05). To examine whether the small body size is due to a decrease in the number of cells or reduction of cell size, we observed morphologies of the eye and leg of Gb’chiconyRNAi crickets in comparison with those of the control crickets (Fig. 5B). We found that in the compound eyes of Gb’chiconyRNAi adults, being smaller than those of DsRednyRNAi adults, the ommatidia were also small (Fig. 5B). In the metathoracic leg, although the leg of Gb’chiconyRNAi adults was shorter than that of DsRednyRNAi adults, five spikes in the tibiae were formed in both Gb’chiconyRNAi and DsRednyRNAi adults (Fig. 5B). These results suggest that the changes observed may be isometric; small body size of Gb’chiconyRNAi adults is due to a proportional reduction of tissues and not loss of parts of tissues.
The instar durations of Gb’chiconyRNAi males and females were longer than those of DsRednyRNAi males and females. For Gb’chiconyRNAi males and females, it took 29.3 and 27.2 days to grow from the third instar to adult stage, respectively, while for DsRednyRNAi males and females, it took 26.9 and 25.8 days, respectively, which are slower by 8% and 5% than the instar durations of DsRednyRNAi males and females (Fig. 6B). In these experiments, although the difference in the instar duration between Gb’chiconyRNAi and DsRednyRNAi crickets at the eighth instar was insignificant, the instar durations of Gb’chiconyRNAi males and females from the third to seventh instar were significantly longer than those of DsRednyRNAi males and females (P < 0.01 and P < 0.01, respectively, Fig. 6B). However, there is no significant difference in the instar duration between male and female (P > 0.05).
Finally, we compared adult mean lifespan of Gb’chiconyRNAi crickets with that of DsRednyRNAi ones when cultured in a mass. Lifespan of both Gb’chiconyRNAi and DsRednyRNAi adults (Fig. 6C) in a mass-cultured condition were shorter than that of separately-cultured condition (Fig. 4E,F, see Discussion); however, the lifespans of Gb’chiconyRNAi adult males and females were estimated to be 12.4 (n = 16) and 26.3 (n = 15) days, which were longer than 11.6 (n = 22) and 21.5 (n = 8) days of DsRednyRNAi males and females, respectively. The maximum lifespans of Gb’chiconyRNAi male and female were 29 and 38 days, whereas those of DsRednyRNAi male and female were 21 and 29 days, respectively, indicating a 7% and 22% extension in the male and female lifespans by nyRNAi against Gb’chico.
In addition to growth phenotypes, Drosophila chico mutant males and females show semi-sterile and sterile phenotype, respectively (Bohni et al. 1999). In order to examine whether the similar phenotype is observed in Gb’chiconyRNAi crickets, we collected their eggs and estimated fertilization rate for Gb’chiconyRNAi or DsRednyRNAi adults. The fertilization rate of eggs laid by DsRednyRNAi females mating with DsRednyRNAi males was 90.2% (Fig. S2). When Gb’chiconyRNAi crickets were used, the fertilization rate was significantly decreased to about 70% of the control rate (P < 0.01, using χ2 test), suggesting that normal expression of Gb’chico is important for fertilization (Fig. S2) in Gryllus, as observed in Drosophila.
Effects of nymphal RNAi for Gryllus Egfr on body size and lifespan
We previously showed that nyRNAi against Gb’Egfr revealed involvement of EGF signaling in leg regeneration (Nakamura et al., 2008b). We confirmed depletion of Gb’Egfr mRNA by nyRNAi previously (Nakamura et al. 2008b). We also found that Gb’EgfrnyRNAi nymphs exhibited remarkable systemic effects of nyRNAi on body size (Fig. 7B) in comparison with experiments with DsRednyRNAi nymphs (Fig. 7A). When dsRNA against Gb’Egfr was injected at the third instar, the average body weight of Gb’EgfrnyRNAi crickets decreased to 59.5% (P < 0.01) of that of the DsRednyRNAi nymphs or 64.3% (P < 0.01) of that of uninjected-normal nymphs at the fifth instar, and 39.5% (P < 0.01) of that of the DsRednyRNAi or 32.4% (P < 0.01) of that of normal nymphs at the sixth instar (Fig. 7C). Unfortunately, all of the examined Gb’EgfrnyRNAi nymphs died by the seventh instar (n = 13), while the negative controls (n = 10) and untreated animals (n = 19) grew up to adulthood. These results indicated a possibility that the EGF signaling pathway is also involved in the determination of body weight in the cricket.
Body size depends on the regulation of both instar duration and growth cessation (Mirth & Riddiford 2007). The highly conserved insulin-signaling pathway is involved in the determination of body size (Mirth & Riddiford 2007; Grewal 2009) and lifespan (Bishop & Guarente 2007; Giannakou & Partridge 2007; Fontana et al. 2010) in nematodes and mammals alike. Thus, it is reasonable to assume that the insulin and TOR pathways play principal roles in controlling nutrition-dependent instar duration in crickets. Since we previously found that systemic nyRNAi can be induced by injection of dsRNA into the cricket hemolymph, we performed nyRNAi experiments to elucidate the mechanisms underlying determination of body size, focusing mainly on the insulin signal pathway. We demonstrate here that Gb’chiconyRNAi adults were significantly smaller than controls. This result is consistent with a small-fly phenotype observed in Drosophila chico mutants (Bohni et al. 1999). In contrast, we found that Gb’FoxOnyRNAi adults were significantly larger than controls. However, our result differs from a phenotype observed in Drosophila FoxO null mutants, which have normal body size (Junger et al. 2003). This Drosophila result was unexpected, because FoxO is a negative regulator of insulin sensitivity (Junger et al. 2003). Thus, our results clearly indicate that the nyRNAi method in the cricket is useful to study mechanisms underlying regulation of body weight/body size by the insulin-signaling pathway. Especially, one of the advantages to use the RNAi method is that timing of knockdown of expression can be chosen freely.
Recent studies in Drosophila showed that reduced InR function affects growth differently in pre-critical weight and post-critical weight larvae (Shingleton et al. 2005). In the pre-critical weight Drosophila larvae, reduced InR activity has no effect on body size, whereas reduced InR activity in the postcritical-weight period results in reduced body size. In Gryllus bimaculatus, critical weight is estimated to be about 0.4 g (unpubl. data), which is attained at the eighth instar. When we performed RNAi experiments against Gb’InR by injecting dsRNA at the fifth instar (before the eighth instar), we observed significant reduction in the adult body weight (Fig. 3A), which is different from the result with Drosophila. Since Drosophila larval development continues until the attainment of critical weight that indicates that a sufficient size has been achieved to support pupariation and maturation (Stern & Emlen, 1999; Nijhout, 2003), the difference may be due to the existence of the pupariation in holometabolous insect species. The effect of Gb’PtennyRNAi treatment was also unexpected, because partial loss of Pten function results in larger Drosophila (Goberdhan et al. 1999; Oldham et al. 2002). This may imply that other insulin signaling components in the cricket compensates loss of Gb’Pten function.
In the case of Gb’TornyRNAi and Gb’S6knyRNAi crickets, reduction of body weight was observed. The link between TOR signaling and body size in invertebrates has been well established (Stanfel et al. 2009). In Drosophila, it has been reported that suppression of the TOR pathway results in prolonged pre-adult development and reduces body sizes (Oldham et al. 2000; Colombani et al. 2003; Kapahi et al. 2004; Katewa & Kapahi 2010). Thus, our result for Gb’TornyRNAi crickets is consistent with previous results with other insects. In the case of S6k, small body size was observed in S6k1−/− mouse mutant (Selman et al. 2009) and dS6K Drosophila mutant (Montagne et al., 1999). Thus, the body weight reduction in Gb’S6knyRNAi crickets is consistent with the results for other animals.
We found that RNAi against Gb’Egfr resulted in smaller body size. A role for EGFR signaling in size determination has not been previously demonstrated in Drosophila (Shilo 2005). This is likely due to the fact that Egfr mutations in Drosophila affect so many tissues that a small body phenotype may be precluded. Interestingly, however, compartment size is reportedly controlled by an EGF ligand from neighboring cells (Parker 2006). Thus, the EGF signaling pathway may be involved in size control. It is interesting to note that Egfr is involved in the regulation of longevity as well as body size in queen honeybee (Kamakura 2011). Since effects of EGFR depletion are rather pleiotropic (Nakamura et al. 2008b) and might be unspecific to some degree, we need further examinations of Gb’Egfr RNAi phenotypes to posit a function of this signaling cascade in body size regulation.
In addition to the body size, we also measured lifespan in Gb’chiconyRNAi animals and found them to be significantly longer than that of controls. This is consistent with data obtained in Drosophila chico mutants, in which median lifespan was extended by up to 48% in homozygotes and 36% in heterozygotes (Clancy et al. 2001). It was reported that in Caenorhabditis elegans, mutation of daf-2 an insulin receptor homologue extended lifespan by reducing the activation of phosphatidylinositol-3-kinase (PI3K) through dephosphorylation and nuclear entry of daf-16 (FoxO homologue) (Bishop & Guarente 2007). Furthermore, mutations in the insulin-signaling pathway have been shown to extend lifespan in C. elegans, Drosophila, mice and humans (for a recent review see Fontana et al. 2010). Mutations that extend lifespan decrease the activity of nutrient-signaling pathways, such as the insulin-like growth factor/insulin and TOR pathways, suggesting that they may induce a physiological state similar to that resulting from periods of food shortage (Fontana et al. 2010). Since the results obtained for Gb’chiconyRNAi nymphs are consistent with those reported so far in organisms ranging from yeast to humans, we conclude that the insulin-signaling pathway may control total body size and lifespan in the cricket.
In our series of experiments on lifespan, we observed that the mean lifespans of control and treated crickets depend on culture conditions; the mean lifespan (78 days) of the controls injected with dsRNA at the third is much longer than that (38 days) of the controls injected at the seventh instar (Fig. 4E,F). Since we reared nymphs in a mass from hatching to the third or seventh instar and cultured them separately after dsRNA injection, the difference appears to be due to difference in a period of the mass-culturing before performing RNAi. Actually, we observed a similar tendency when injected with dsRNA at the fourth and eighth instar (data not shown). Thus, we consider that a high population density in culture negatively affects the cricket lifespan. This may be supported by the fact that adult longevity depends on density population in other insect species such as the grasshopper (Peters & Barbosa 1977). We also found that the mean lifespan of control adults reared in a mass (27 and 26 days for males and females, Fig. 6C) became shorter than that of control adults cultured separately (78 days, Fig. 4E) after performing RNAi at the third instar, suggesting that mating status and laying eggs negatively affect the cricket lifespan. This result is also consistent with the fact that mating status negatively affects lifespan in animals (Hsin & Kenyon 1999; Davies et al. 2005).
In conclusion, we have demonstrated that systemic RNAi against InR, chico, Tor, S6k, FoxO, and Egfr affects body size and that lifespans were changed after systemic RNAi against chico in the cricket. Thus, systemic RNAi in the cricket should be a viable approach to elucidate the mechanisms underlying the determination of the body size and lifespan. The cricket system would facilitate progress of studies on body size and lifespan.
We thank Masato Haji and Kayoko Tada for their technical support. This work was supported by a grant from the Japanese Ministry of Education, Culture, and Sports, Science and Technology to T.M., H.O., and S.N.