Abstract
Juvenile hormone (JH) is critical for multiple aspects of insect development and physiology. Although roles for the hormone have received considerable study, an understanding of the molecules necessary for JH action in insects has been frustratingly slow to evolve. Methoprene-tolerant (Met) in Drosophila melanogaster fulfills many of the requirements for a hormone receptor gene. A paralogous gene, germ-cell expressed (gce), possesses homology and is a candidate as a Met partner in JH action. Expression of gce was found to occur at multiple times and in multiple tissues during development, similar to that previously found for Met. To probe roles of this gene in JH action, we carried out in vivo gce over- and underexpression studies. We show by overexpression studies that gce can substitute in vivo for Met, alleviating preadult but not adult phenotypic characters. We also demonstrate that RNA interference-driven knockdown of gce expression in transgenic flies results in preadult lethality in the absence of MET. These results show that (1) unlike Met, gce is a vital gene and shows functional flexibility and (2) both gene products appear to promote JH action in preadult but not adult development.
THE sesquiterpenoid juvenile hormone (JH) regulates numerous insect functions, including molting, morphology and caste determination, and reproduction (Wheeler and Nijhout 2003). JH has been shown to regulate gene expression in carrying out many of these functions (Dubrovsky et al. 2000; Beckstead et al. 2007; Li et al. 2007; Minakuchi et al. 2008), but a crucial understanding of JH action is lacking, in part because elucidation of the hormone receptor has been difficult (Gilbert et al. 2000; Willis 2007). A likely prospect is the Methoprene-tolerant (Met) gene, originally discovered in a Drosophila melanogaster screen for mutants resistant to the JH insecticidal agonist methoprene (Wilson and Fabian 1986). JH or JH agonists applied to dipteran insects at the onset of metamorphosis result in lethality and morphogenetic defects, such as failure of rotation of the male genitalia that normally occurs during pupal development (Madhavan 1973; Postlethwait 1974), and Met mutants show resistance to these JH effects (Wilson and Fabian 1986). The Met gene product has been shown to possess characteristics of a hormone receptor, such as high-affinity JH binding (Shemshedini and Wilson 1990; Miura et al. 2005), expression in JH target tissues (Pursley et al. 2000; Liu et al. 2009), and JH-dependent transcriptional activity (Miura et al. 2005). Met is a member of the bHLH-PAS transcription factor gene family (Ashok et al. 1998). Recently, a Met-like homolog of the beetle Tribolium castaneum was identified and shown by RNA interference (RNAi) experiments to be necessary for proper larval-larval molts (Konopova and Jindra 2007), a key role for JH in a variety of insects, thus strengthening the likelihood of MET involvement in JH action.
This MET-JH action hypothesis was weakened when a null allele, Met 27, was found to be homozygous viable (Wilson and Ashok 1998). Since JH is involved in molting in many insects, a Met null allele might be expected to result in a lethal phenotype (Riddiford 2008). Perhaps another gene rescues the lethality of Met27 flies. The most likely candidate is a paralogous D. melanogaster gene, germ-cell expressed (gce), identified as a bHLH-PAS gene upon analysis of the sequenced genome (Moore et al. 2000). The level of sequence similarity to Met is ∼60% and highest in the conserved bHLH and PAS domains involved in DNA binding and protein–protein interaction (Dolwick et al. 1993; Huang et al. l993). The function of gce is poorly understood. Recently, however, GCE and MET were shown to induce programmed cell death in D. melanogaster larval fat body tissue during metamorphosis, a result that could be suppressed by methoprene application (Liu et al. 2009). Furthermore, attempts to isolate Met homologous genes from three mosquito species and the beetle T. castaneum (Konopova and Jindra 2007) revealed only a single homolog in each species with higher sequence and intron–exon structure similarity to gce than to Met (Wang et al. 2007). Together, these results suggest that an ancestor gene duplication resulting in gce and Met occurred after the lower–higher dipteran split during evolutionary time. Unlike Met, no gce mutants have been reported, so the consequences of mutated gce are unknown.
To probe the role of gce in JH action, we carried out two sets of experiments: (1) overexpression of gce in both Met + flies to detect any new phenotype and Met mutants to measure possible rescue of Met phenotypic characters and (2) RNAi-driven reduction of gce expression to determine the consequences of insufficient GCE. To overexpress gce and produce gce-RNAi in flies, we employed the GAL4-UAS system (Brand and Perrimon l993), which uses a D. melanogaster gene promoter ligated to yeast GAL4, resulting in a GAL4 protein product to drive expression of gce ligated to a UAS response element in transgenic flies.
MATERIALS AND METHODS
gce overexpression construct:
The GAL4-UAS system (Brand and Perrimon l993) was used to overexpress gce in transgenic flies. A gce-bearing transgenic stock, UAS-gce, was constructed from gce cDNA following RT–PCR amplification. First, total RNA was isolated from adult flies using TRIzol reagent, followed by treatment with RNase-free DNase-I (Invitrogen) and bromochloropropane (Molecular Research Center) reagents. Two-microgram aliquots were reverse transcribed with 200 units/ml of Moloney murine leukemia virus (MMLV) reverse transcriptase (Invitrogen) plus 500 ng of random hexamers (Ambion) in a 25-μl reaction mixture to synthesize first-strand cDNA. The cDNA was amplified by PCR in a reaction mixture consisting of 2.5 μl of 10× PCR buffer, 1 μl Taq DNA polymerase (Invitrogen), 1 μl of 10 mm dNTP mix, 1 μl of 50 mm MgCl2, 1 μl each of 10 mm primer, 1 μl of cDNA, and 16 μl of H2O. The primers used were forward 5′-ATAGGTACCACGATTGCGAAATGTTATGC-3′ and reverse 5′-ATATCTAGAGAAACCCTTCAGTCGAGACC-3′. The product was subcloned into the TOPO vector (Invitrogen), sequenced, and inserted into the pUAST transformation vector (Brand and Perrimon l993). Proper orientation of the fragment in the vector was verified by restriction site analysis and sequencing.
gce RNAi construct:
Expression of gce was also subjected to RNAi-mediated knockdown. A gce RNAi-generating transgene was engineered from a fusion of genomic DNA and reverse complement cDNA, which has been shown to give higher knockdown of the gene of interest (Kalidas and Smith 2002). First, genomic DNA was isolated as follows: 15 flies were homogenized in 200 μl of extraction buffer (1% SDS, 50 mm Tris HCl (pH 8.0), 25 mm NaCl, 25 mm EDTA) and placed in a water bath at 65° for 1 hr. Then 100 μl of 3 m potassium acetate was added, and the sample was placed on ice for 1 hr. The sample was centrifuged, and DNA was precipitated from the supernatant with 2 vol of ethanol (EtOH) and 0.1 vol of 3 m sodium acetate overnight at −20°. The pellet was rinsed in 70% EtOH, air-dried for 5–10 min, and then resuspended in 100 mm Tris HCl (pH 8.0), 1 mm EDTA.
A gce fragment was amplified from the genomic DNA by PCR as described above using forward primer 5′-TATAAGATCTCCGCAGTCGACCATCGATGT-3′ preceded by a BglII adapter site and reverse primer 5′-AGATCTCGAGCTGCCGGCAAATAGTTCAAT-3′ preceded by a XhoI adapter site (for ease of subcloning). The product, which spanned part of exon 6 and intron 6, was subcloned into the TOPO vector (Invitrogen), sequenced, and inserted into the pUAST transformation vector.
To create the cDNA fragment, total RNA was isolated from adult flies, and cDNA was synthesized and amplified as described above using forward primer 5′-AGATTCTAGACACACAGATACCGCAGTCGA-3′ preceded by a XbaI adapter site and reverse primer 5′-TATACTCGAGCAGTGCTGGGTATAAAACGC-3′ preceded by a XhoI adapter site. The product was subcloned into the TOPO vector, sequenced, and inserted into the pUAST transformation vector immediately downstream of the inserted genomic fragment. Proper orientation of the fragments in the vector was verified by restriction site analysis and sequencing.
D. melanogaster transformation:
Germline transformation of either w or w MetW3 embryos was performed by injecting dechorionated embryos with purified UAS-gce or UAS-gce-dsRNA plasmid together with “wings clipped” helper plasmid pπ25.7 (Rubin and Spradling 1982) in a ratio of 2–3:1. Transformant flies were recognized by partial restoration of eye color resulting from the mini-w+ gene included in the pUAST vector. Multiple UAS-gce and UAS-gce-dsRNA transformant lines were isolated and made homozygous for the transgene by selecting for w+ eye color. Each of these lines represents an independent insertion of the transgene into the genome, and lines having the transgene inserted into the third chromosome were selected and used in these studies. Other genes or alleles were introduced into transformant lines by genetic crosses.
gce overexpression and knockdown in GAL4-UAS transgenic strains:
Simple genetic crosses served to bring the UAS-linked and GAL4-linked transgenes together to control expression. Each UAS line was crossed with flies carrying either an actin-GAL4 or a tubulin-GAL4 transgene heterozygous to a TM6 balancer chromosome carrying a Tubby (Tb) dominant mutation, allowing ready identification of larvae/pupae/adults carrying the balancer chromosome in the F1 generation. Similar crosses with tubulin-GAL4/TM6, Lsp2-GAL4, and GawB}dan[AC116]-GAL4 fathers allows tubulin, larval serum protein2, and compound eye promoters, respectively, to drive UAS-gce expression in progeny. The Lsp2 promoter is expressed in the third instar fat body, a JH target tissue (Liu et al. 2009), and GawB}dan[AC116] in the compound eye (FlyBase), thus allowing tissue-specific expression. F1 progeny carrying the TM6 balancer chromosome lack actin-GAL4 and therefore serve as control flies. Similar crosses using w Met27/Y; actin-GAL4 fathers resulted in F1 females that were homozygous for Met27, and using w/w; UAS-gce/UAS-gce mothers, resulted in F1 females that were homozygous/hemizygous for w Met+; both of these crosses were used in some experiments.
Analogous crosses with either w Met27; UAS-gce-dsRNA or w Met+; UAS-gce-dsRNA females allowed either tubulin- or actin-driven expression of the dsRNA transgene in F1 individuals homozygous for either Met27 or Met+.
RT–PCR:
Overexpression or knockdown of gce was measured using RT–PCR. RNA extraction, reverse transcription, and PCR amplification was carried out as described above, except that PCR primers used for gce were forward primer 5′-GGATGCCATCGATCGCAAGT-3′ and reverse primer 5′ GCTTCGTCACTACGCCGAAA 3′. The primers for Rp49 were forward primer 5′-CCGCTTCAAGGGACAGTATC-3′ and reverse primer 5′-ATCTCGCCGCAGTAAACG-3′. Ten microliters of each PCR product was electrophoresed on a 1% agarose gel and stained with ethidium bromide. An image of the gel was captured with the ImageQuant 400 (GE Healthcare) and then analyzed using the ImageQuant TL software (Amersham Biosciences). PCR products were cloned into TOPO vector (Invitrogen) and sequenced with M13 forward primer.
Northern analysis:
Total RNA was isolated from a D. melanogaster Oregon-R strain using TRizol reagent (Invitrogen). The RNA was separated by electrophoresis on a 0.8% formaldehyde–agarose gel and then transferred to positively charged nylon membranes (Roche Diagnostics, Indianapolis). RNA probes were prepared with a DIG RNA labeling kit (Roche) using gce, Met, Rp49, and rRNA sequences as templates. The cDNA used to prepare the probes was as follows: gce cDNA 5′ untranslated region that includes nucleotide +1 to +1683; gce cDNA 5′ coding region +1338 to +2381; gce cDNA 3′ coding region +3340 to +4216; Gce cDNA 3′ untranslated region +3989 to +5530; Met cDNA +151 to +2151; rp49 cDNA +1 to +458; and 18S rRNA gene +500 to +1377. The membranes were hybridized with DIG-labeled RNA probes for 12 hr at 65° with DIG easy hyb (Roche). DIG-labeled RNA was detected with an alkaline phosphatase-conjugated anti-DIG antibody using CDP star (Roche).
Methoprene resistance:
Cultures were assayed for methoprene sensitivity on each of three to five doses of methoprene [isopropyl-(2E,4E)-11-methoxy-3,7,11-tri-methyl-2,4-dodecadienonate; ChemService] and applied as ethanolic solutions (25 μl) to the surface of each culture (food surface area 3.8 cm2) in doses ranging from of 0.27 to 5.4 μg/vial. Mortality occurred predominately during the late pupal (pharate adult) stage; survivors were examined for the methoprene-induced morphogenetic defect of malrotated male genitalia or posterior sternal bristle defects (Madhavan 1973; Wilson and Fabian 1986).
Eye phenotype:
The Met eye-defect phenotype was quantitated by counting defective facets (appearing black) in either pharate adult (nonsurvivors) or adult (survivors) males under ×40 magnification. Preliminary experiments showed that the number of defective facets does not increase between these two stages, thus allowing direct comparison of individuals between the two developmental stages.
Oviposition:
Females were isolated within 8 hr after eclosion and allowed to age for 6 days in the presence of wild-type (Oregon-RC) males and baker's yeast sprinkled on the food surface to maximize oogenesis. They were then placed individually in 28- × 95-mm plastic food vials (Capitol, Fonda, NY) in the continued presence of yeasted food and males at 25°. Eggs were counted during the next 4 days, a time period during development when the ovipositional rate reaches a maximal steady-state value and allows strain comparison (Wilson and Ashok 1998). Egg fertility was noted, and vials having unfertilized eggs were discarded because unmated females are less fecund.
In some experiments, oogenesis was examined by dissecting females in Ringer's solution and censusing vitellogenic oocytes in each ovary using the staging described previously (Wilson and Ashok 1998).
Male courtship:
Males of the appropriate genotypes were taken from cultures at age 3–5 days after eclosion following light etherization. They were allowed to recover for 16–24 hr and then placed individually on food in 25- × 95-mm plastic food vials at 22° with two 4- to 6-day-old post-eclosion Oregon-RC virgin females. Courtship was noted for 1–2 hr, and then males were removed at designated times up to 18 hr, and female fertility was noted in F1 larvae appearing 3 days later, showing that mating had occurred.
RESULTS
Expression of gce:
The gce gene was first detected in embryonic germ cells, giving rise to its name (Moore et al. 2000). We first examined the expression patterns of gce in wild-type (Oregon-RC strain) to detect any differences from those found for Met (Pursley et al. 2000). A Northern analysis showed expression at multiple times during development with periods of low-to-absent expression in early embryonic and pupal stages (Figure 1). The expression of Met was found to be similarly widespread (Pursley et al. 2000), but the pattern was not identical to that of gce.
Northern hybridization analysis of gce transcripts. RNA was isolated from whole bodies at various developmental ages. Five micrograms total RNA was electrophoresed and transferred to nylon membranes. The membranes were hybridized with DIG-labeled gce, Rp49, and rRNA probes. The identity of the 7.5-kb minor band is undetermined.
Expression of gce was also examined by RT–PCR in selected tissues, including two known JH target tissues, ovary and male accessory glands. Expression levels were generally lower than that found for Met, especially in the larval fat body, where it was present at trace levels (Figure 2). Another study has shown expression of gce and Met in a variety of tissues and expression levels (Chintapalli et al. 2007). Clearly, expression of gce is not limited to its namesake tissue in the embryo.
RT–PCR analysis of Met and gce transcripts from selected tissues. The first-strand cDNA was synthesized, and PCR was performed with primer pairs for gce, Met, and Rp49 genes in adult brain, ovary, male accessory gland (MAG), and third instar larval fat body. Values are standard error of the mean (SEM) of three separate determinations expressed relative to those of Rp49. The Met PCR primer sequences are given in Barry et al. (2008).
Overexpression of gce rescues preadult loss of Met function:
The UAS-gce transgene was found to be overexpressed by either tubulin-GAL4 or actin-GAL4, but more strongly by the tubulin promoter (Figure 3 ). Flies having gce overexpressed in a Met+ background showed no obvious novel phenotype—such as lethality or change in development—evident at any stage of development. However, when gce was overexpressed in a Met 27 background, we found rescue of preadult Met27 phenotypic characters. These included both methoprene-conditional and nonconditional characters. The conditional phenotype includes resistance to the well-documented lethal and morphogenetic effects of methoprene. When challenged with methoprene, both Met27; UAS-gce/tubulin-GAL4 and Met27; UAS-gce/actin-GAL4 flies were less tolerant than balancer chromosome sibling controls of the toxic effects at each of four doses (Figure 4 ), showing blockage of the namesake Methoprene-tolerant phenotype of Met27. In contrast, Met27; UAS-gce/TM6 control flies, having a TM6 balancer chromosome substituted for the GAL4 driver chromosome, showed good survival due to Met27 at all four doses tested (Figure 4).
(A) Expression of gce and rp49 in UAS-gce overexpressing strains. Expression was measured by RT–PCR in both larvae and adults. (B) Expression values for gce and rp49 in gce-overexpressed larvae and adults. Values are normalized to those of +; UAS-gce/TM6 sibs. Each value is the standard error of the mean (SEM error bar) of three determinations of larvae or adults of the indicated genotype.
Enhanced methoprene pupal toxicity of gce-overexpressing strains. gce refers to UAS-gce driven by GAL4-actin, GAL4-tubulin, or GAL4-LFB (larval fat body) promoters. Each value is the standard error of the mean (SEM error bar) of 30 pupae, three determinations. Pupal survival was defined as eclosion. Methoprene doses are in micrograms/vial.
Similarly, treatment of D. melanogaster Met+ prepupae with JH analogs results in failure of the complete rotation of the male genital disk that normally occurs during pupal development (Adam et al. 2003), and Met mutants are completely resistant to this effect (Wilson and Fabian 1986). In our experiments, resistance seen in Met27 to this methoprene-induced effect was blocked in Met27; UAS-gce/tubulin-GAL4 flies, resulting in abnormal genital rotation, a response similar to that found for Met+ (Figure 5). The effect was widespread: 91% (N = 100) of males from cultures treated with 0.27 μg/vial of methoprene showed varying degrees of abnormal rotation, compared with 0% (N = 100) of treated Met27/Y; UAS-gce/TM6 control sibs that showed abnormal rotation, as expected due to the Met27mutation. Therefore, the presence of overexpressed gce resulted in striking enhancement of both methoprene toxicity and a morphogenetic response to methoprene, showing that GCE can substitute for MET in the flies to restore methoprene sensitivity approaching Met+ levels.
Rescue of resistance to methoprene-induced failure of male genitalia rotation during pupal development. (A) Normal-appearing genitalia phenotype in w Met27; UAS-gce/TM6 male exposed to 5.4 μg/vial methoprene during larval development. The absence of Met+ protects the fly from methoprene-induced malrotation, and the appearance is normal, having the anal plates posterior and the penis apparatus anterior. (B) Abnormal-appearing genitalia in Met27; UAS-gce/tubulin-GAL4 adult exposed to 0.027 μg/vial methoprene. The genitalia has failed to complete the normal 360° rotation by ∼100°, an effect expected in methoprene-treated flies carrying Met+.
Additionally, gce overexpression rescued a nonconditional Met phenotypic character, that of defective posterior facets in the compound eye of Met mutants (Figure 6). Quantification of the phenotype and its rescue was carried out by counting darkened facets on affected adults under light microscopy (Table 1). The eye phenotype was strongly rescued in MetW3; UAS-gce flies driven by either GAL4 promoter.
Rescue of eye phenotype in MetW3 flies. (A) Nonconditional eye phenotype of disrupted posterior facets (evidenced by darkening) in w MetW3; UAS-gce/TM6 adult. (B) Complete rescue of eye phenotype in w MetW3; UAS-gce/tubulin-GAL4 adult. The light-red eye color in A results from the w+-carrying UAS-gce transgene, and the darker-red color in B from the additional w+-carrying tubulin-GAL4 transgene.
Defective eye facets in gce overexpressed and underexpressed strains
Clearly, overexpression of GCE by a global GAL4 promoter such as actin or tubulin can result in GCE substitution for MET to block the Met phenotype (Figure 4). To examine the blockage effect using a tissue-specific promoter, we drove expression of GCE using a larval fat body promoter in the Lsp2-GAL4 strain. Examination of 50 F1 males showed 100% morphologically normal male genitalia, characteristic of typical Met-induced resistance, showing no substitution by GCE for MET to block the resistance. However, resistance to either the eye defect (Table 1) or methoprene-induced pupal death (Figure 4) in MetW3 or Met27 was either essentially absent (eye phenotype) or partially found (pupal death phenotype), respectively, in the UAS-gce/Lsp2-GAL4 transgenic flies, which showed at least partial substitution by GCE for MET. Therefore, GCE overexpression in a specific tissue can partially substitute for MET to block Met-generated resistance to pupal death and to the Met-generated eye defect, but not to the male genitalia defect. We also examined GCE driven by another GAL4 promoter, GawB}dan[AC116]-GAL4, that is specific for the compound eye. Unlike GCE driven by the Lsp2-GAL4 promoter, the F1 showed complete rescue of the Met eye phenotype, resulting in morphologically normal eyes (Table 1). Therefore, GCE substitution for MET can show tissue specificity.
While rescue of preadult Met phenotypic characters was clear and consistent, the adult phenotype of Met27, consisting of deficient oogenesis/oviposition and male courtship behavior (Wilson and Ashok 1998; Wilson et al. 2003), was not appreciably changed in Met27; UAS-gce/tubulin-GAL4 adults compared to Met27 balancer sibs (Figure 7). Both GAL4 drivers result in gce overexpression in adults, although to a lesser extent than in larvae (Figure 3). Either this level of overexpression is insufficient for rescue or GCE is incapable of substituting for MET in adults.
Fertility of adults overexpressing gce. (A) Eggs laid per female during a 4-day window of optimal oviposition. Each value is the mean of 30 females of the indicated genotype. (B) Males fertilizing newly encountered wild-type females during the indicated time periods; values are cumulative. Met+/Y; UAS-gce/tubulin-GAL4 males were tested only at the 0- to 1-hr period. Each value is the mean of 20 males, five replicates.
Underexpression of gce results in lethality:
Additional insight into the role of gce came to light when we examined the consequences of gce underexpression by using either actin-GAL4 or tubulin-GAL4 to drive expression of gce dsRNA, which results in RNAi directed to gce. Alone, neither of the two GAL4 constructs nor the UAS-gce dsRNA construct alone has a notable effect on survival or fertility in either a Met+ or a Met27 background. However, the level of gce transcript in late larvae expressing gce-RNAi was found to be considerably lowered relative to that in balancer siblings (Figure 8), showing the effectiveness of the dsRNA transgene in reducing gce mRNA. Initally, we examined the phenotype of Met+/Met+; UAS-gce-dsRNA/tubulin-GAL4 animals to understand the consequences of gce reduction in a Met+ background. Embryos were seemingly unaffected, judging from good hatch rates after egg fertilization at 25°, but no role for JH has been reported during D. melanogaster embryogenesis, so this result is perhaps not surprising.
(A) Expression of gce and rp49 in UAS-dsRNA gce larvae having the designated GAL4 promoter or TM6 balancer chromosome as measured by RT–PCR. O-RC, Oregon-RC wild-type strain. (B) Expression values for gce and rp49 in Met27; UAS-gce-dsRNA larvae. Values are normalized to those in Oregon-RC wild-type larvae. Values are standard error of the means (SEM error bars) of three determinations of larvae of the indicated genotype.
However, the consequences for postembryonic viability were severe: Met+; UAS-gce-dsRNA/tubulin-GAL4 homozygotes/hemizygotes, but not TM6 siblings, failed to survive to adulthood, dying as mid-to-late stage larvae or, more commonly, as pupae (Table 2). Visual examination of these pupae revealed that death typically occurred in pharate adults. When driven by the actin-GAL4 construct, gce RNAi expressed in Met+ homozygous progeny resulted in survival in reasonable numbers with many eclosing as adults (Table 2). However, these adults were affected by gce underexpression, since a majority of these females (but not the TM6 sibling adults) died within 2–3 days following eclosion.
Larval and pupal survival values of gce-dsRNA strains
To understand the consequences of gce reduction in a Met background, we generated Met27; UAS-gce-dsRNA/tubulin-GAL4 animals. Again, prepupal death occurred, but lethality was shifted from pharate adults to earlier pupae (0–2 days), showing more severe consequences of gce reduction in a Met27 than in a Met+ background. When driven by the actin-GAL4 construct, gce-dsRNA resulted in lethality, primarily of pharate adults (Table 2).
Therefore, reduction of gce is lethal in Met27 or Met+ preadults when driven by a tubulin promoter but is lethal primarily only in Met27 preadults when driven by the “weaker” actin promoter (Barry et al. 2008). Another possibility is that the slightly larger amount of GCE produced in the actin-driven strain (Figure 8) is sufficient to partially rescue the lethality, resulting in the higher survival seen relative to that with the tubulin-driven promoter. Similar results (not shown) were seen when MetW3 replaced Met27 in these strains, showing the results not to be allele specific.
Methoprene effect during gce underexpression:
It is clear that gce overexpression enhances methoprene toxicity (Figure 4). To determine if methoprene could either rescue or exacerbate the preadult lethality phenotype due to gce underexpression, we examined the consequences of methoprene application when gce is underexpressed. Methoprene clearly acts as a juvenile hormone at both the organismal and molecular level in a variety of insects (Wilson 2004). Since Met27; UAS-gce-dsRNA cultures are preadult lethal (Table 2), we examined Met+; UAS-gce-dsRNA cultures exposed to a range of methoprene doses, including milder, generally sublethal doses (0.27 and 0.54 mg/vial) and stronger doses (2.7 and 5.4 mg/vial). Little difference in the toxicity to gce-dsRNA/actin or TM6 controls was found (Table 3).
Survival of gce RNAi transgenic pupae on methoprene
Methoprene/JH III application to Met27 adults:
JH is required for oogenesis and oocyte production in a variety of insects (Wyatt and Davey 1996), including D. melanogaster (Postlethwait and Handler 1978; Wilson 1982). If Met is involved in JH action, then the poor oogenesis and oviposition previously seen in Met27 females (Wilson and Ashok 1998) might reflect either insufficient JH or some defect in either the JH receptor or an associated component in JH action in these flies. We examined this question by exposing newly eclosed Met27 females to methoprene or JH III applied to the food surface, a valid method of application (Wilson et al. 2003). Daily oviposition was measured during a 5-day window of steady hormone exposure and oviposition. The mean daily oviposition by eight Met27 females (five replicates) exposed to methoprene was 59.6 (2.2 SEM); to JH III, 57.1 (2.0 SEM); and to ethanol, 52.6 (2.2 SEM). Oviposition by methoprene- or JH III-treated adults was higher, but not significantly different from that of ethanol-treated siblings (P = 0.066, one-way ANOVA). Therefore, the poor oviposition in Met27 is not due to significant JH deficiency and is consistent with a phenotype expected to result from an abnormal component of JH reception.
Underexpression of gce magnifies Met eye phenotype:
Since overexpression of gce blocked the preadult Met phenotypic characters (Figures 4–6⇑), we examined the effect of gce underexpression on the MetW3 eye phenotype. The eye-defect phenotype could be examined in moribund MetW3/Y; UAS-gce-dsRNA/actin-GAL4 male pharate adults because eye development in the pharate adult stage appears morphologically normal (except for the facet phenotype). When the Met eye phenotype was compared in actin-driven and TM6-balancer siblings, a significant (but not dramatic) enhancement of the facet defect abundance was evident (Table 1). Met must be lesioned to produce the eye phenotype; if Met+ is substituted for MetW3 in the mother, the eye remains wild type in the F1 Met+/Y; gce-RNAi males, showing that a low GCE level alone does not result in the eye phenotype. Therefore, reduction of gce in MetW3 hemizygous flies results in a mild enhancement of the nonconditional eye phenotype.
DISCUSSION
Our previous work established Met involvement in JH action in D. melanogaster, a conclusion that has been supported by subsequent independent work demonstrating high-affinity JH binding to MET, JH-driven transcriptional activity (Miura et al. 2005), and involvement of a Met homolog in control of metamorphosis in the beetle T. castaneum (Konopova and Jindra 2007). We now show by under- and overexpression studies that the paralogous gene gce in D. melanogaster not only plays a vital role during pupal development but also can substitute for Met, suggesting involvement in the action of this hormone.
Is the rescue due to an abundance of GCE or to expression in tissue not normally expressing endogenous gce? To address the issue of the tissue specificity of Met/gce expression, we turned to recent results demonstrating that larval fat body catabolism, required for completion of metamorphosis, is initiated by ecdysone, MET, and GCE and can be blocked by JH application (Liu et al. 2009). Perhaps high pupal survival of Met mutants following methoprene application results from an absence of methoprene-induced blockage of catabolism and the lowered pupal survival in Met27; gce transgenic flies (Figure 4) results from substitution of GCE for the absent MET. Overexpression of GCE specifically in larvae fat body was carried out using a larval fat body GAL4 driver, and resistance to methoprene-induced mortality and male genitalia malrotation was examined. Methoprene-treated Met27; UAS-gce/lfb-GAL4 was found to be completely (50/50 examined) resistant to the male genitalia defect, indicating no blockage of the Met27 mutation. However, the progeny were only partially resistant to pupal death, showing that GCE can partially substitute for MET in this tissue to block the Met27 mutation and suggesting that tissue-specific, not widespread gce overexpression, may underlie the basis of the GCE substitution effect. Since the larval fat body shows little or no gce (Figure 1), then supplying GCE to this tissue can explain the tissue-specific effect seen in the pupal death phenotype. The eye phenotype was not rescued by larval fat body promoter-driven GCE, but it was completely rescued by compound-eye promoter-driven GCE (Table 1). Therefore, GCE expressed in the larval fat body can partially substitute for MET in the tissue(s) responsible for pharate adult death, but not for eye or male genitalia, demonstrating tissue specificity of expression or utilization of GCE/MET.
We could find little effect of gce overexpression on adult reproductive phenotypes of Met27. JH plays roles in both male and female reproduction in D. melanogaster as well as in many other insects (Wyatt and Davey 1996). Clearly, overexpressed GCE can substitute for MET in Met preadults, but since the adult transgenic fly reproductive phenotypes were similar to those in adults not overexpressing the gce transgene (Figure 7), this substitution appears to be unproductive in adults. This result suggests that MET may be the major player in adults. The presence of only a single Met/gce homolog in three mosquito species (Wang et al. 2007) and in the beetle T. castaneum (Konopova and Jindra 2007) with higher similarity to gce than to Met suggests that this reproductive role for MET evolved following the gene duplication seen in higher Diptera. Both Met and gce are present in the 12 species of D. melanogaster whose genomes have been sequenced (http://flybase.org); therefore, this duplication occurred earlier than the evolutionary divergence of these species that occurred as much as 60 million years ago.
Clearly, gce underexpression can be lethal to either larvae or pupae, especially pupae, and especially in the absence of Met+ (Table 2). The presence of background Met+ allowed greater pupal development, shifting pupal death in RNAi individuals from the early pupal stage to the pharate adult/adult stage, depending on the promoter used and thus presumably on the level of RNAi produced. Underexpression of gce in the absence of Met+, in a Met27background, is more severe and results in a total loss of preadult viability (Table 2). This suggests that both GCE and MET can interact to promote some vital aspect of larval/pupal development. Our previous work using D. melanogaster S-2 cells has shown that MET and GCE can heterodimerize (Godlewski et al. 2006), suggesting a mechanistic basis for MET–GCE interaction. However, underexpression driven by the stronger tubulin promoter results in pupal death in a Met+ background (Table 2), so the absence of Met+ is not a prerequisite for this phenotype. If both MET and GCE are necessary for JH action, then underexpression of both genes could result in death due to a failure of some JH-controlled developmental event, for example. The greater phenotypic severity of gce underexpression in a mutant Met background (Table 2) argues for the JH action scenario.
Tissue-specific expression levels of Met and gce are given for 24 larval and adult tissues in the FlyAtlas database, determined by microarray analysis. The data show robust expression (at least 2/4 “present” calls in the four microarray replicates) for gce in 12 tissues and for Met in 10 tissues. Seven tissues showed robust expression of both genes; interestingly, none are demonstrated JH target tissues. This could mean that (1) the presence of both GCE and MET are not required for JH action, (2) only one is required, or (3) few JH target tissues have been identified (perhaps the more likely explanation). However, there are some surprises in the FlyAtlas data set; for example, neither gene showed good expression in ovary or larval fat body, and only gce showed strong expression in male accessory glands, all demonstrated JH target tissues. Possibly, additional regulatory roles, independent of JH, for either of these transcription factors exist and are reflected in the FlyAtlas data set.
Does this work show MET and GCE involvement in a JH receptor complex? Although the phenotypic characteristics of Met suggest involvement in JH reception, there is no direct evidence for involvement in a bona fide receptor. The disparate levels of transcript found for Met and gce in certain tissues (Figure 2) might suggest separate roles for either or both of the gene products in certain tissues, and not formation of a mandatory heterodimer that might be predicted for a JH receptor. Indeed, one of the roles might involve eye development, which can be disrupted when either Met is mutated (Figure 6; Table 1) or JH titer is low (Riddiford et al. 2010). Likewise, the lack of substantial resistance to methoprene in the Met+; UAS-gce dsRNA/actin-GAL4 flies (Table 3) might seem perplexing, considering the high resistance seen in Met mutants, but this conundrum might simply reflect a lack of strong JH binding by GCE, a possible requirement for resistance to the hormone and insecticide. Ligand binding might be the sole property of MET in a MET–GCE heterodimer, and loss of GCE does affect JH action, but not due to failure of JH binding. Future studies focusing on the role of gce may lead to more rapid progress in defining a JH receptor.
Acknowledgments
We thank J. Herbers for comments on the manuscript and A. Simcox for supplying two of the GAL4 strains. The work was supported by a grant from the National Institutes of Health (AI052290) to T.G.W.
Footnotes
- Received March 25, 2010.
- Accepted May 17, 2010.
- Copyright © 2010 by the Genetics Society of America
