Characterization of the Juvenile Hormone Pathway in the Viviparous Cockroach, Diploptera punctata

Juvenile hormones (JHs) are key regulators of insect development and reproduction. The JH biosynthetic pathway is known to involve 13 discrete enzymatic steps. In the present study, we have characterized the JH biosynthetic pathway in the cockroach Diploptera punctata. The effect of exogenous JH precursors on JH biosynthesis was also determined. Based on sequence similarity, orthologs for the genes directly involved in the pathway were cloned, and their spatial and temporal transcript profiles were determined. The effect of shutting down the JH pathway in adult female cockroaches was studied by knocking down genes encoding HMG-CoA reductase (HMGR) and Juvenile hormone acid methyltransferase (JHAMT). As a result, oocyte development slowed as a consequence of reduction in JH biosynthesis. Oocyte length, fat body transcription of Vg and ovarian vitellin content significantly decreased. In addition, silencing HMGR and JHAMT resulted in a decrease in the transcript levels of other genes in the pathway.


Introduction
Juvenile hormones (JHs) play key roles in regulating growth, development, metamorphosis, aging, caste differentiation and reproduction in insects (as reviewed by [1,2,3]). The multiple processes in which JH is involved and the critical role which JH plays in metamorphosis and reproduction emphasize the importance of elucidating the JH biosynthetic pathway and the factors that regulate its biosynthesis.
JHs are sesquiterpenoid compounds that are synthesized and secreted by specialized, paired endocrine glands, the corpora allata (CA). The complete biosynthetic pathway of JH III (the most widespread and predominant JH homolog in insects) comprises 13 discrete enzymatic steps. This pathway can be divided into two metabolic parts ( Fig. 1): the early portion comprises the mevalonate pathway to the formation of farnesyl diphosphate (FPP) and is conserved in both vertebrates and invertebrates; the later part of the pathway is specific to insects and other arthropods. In this later part, FPP is cleaved to farnesol, which is then oxidized to the  [4] and Nouzova et al. [11]). The Insect-and Arthropod-specific isoprenoid branch of JH biosynthesis is represented in the dashed box. Precursors are in bold and connected by arrows. Enzymes are in italics. Abbreviations for the enzymes are given in brackets.

Materials and Methods Animals
The D. punctata colony was maintained at 27°C in constant darkness and animals were fed lab chow and water ad libitum. To obtain pools of synchronised animals, newly molted female adult cockroaches were picked from the colony, placed in separate containers and provided with water and lab chow. Mated status was confirmed by the presence of a spermatophore.
Tissue collection and RNA extraction D. punctata were dissected in modified cockroach ringer solution (150 mM NaCl, 12 mM KCl, 10 mM CaCl 2 .2H 2 O, 3 mM MgCl 2 .6H 2 O, 10 mM HEPES, 40 mM glucose, pH 7.2) using a dissecting microscope. Samples were flash-frozen in liquid N 2 to prevent RNA degradation and were stored at −80°C until further processing. For each dissected female, basal oocyte lengths were measured to determine the physiological age. CA samples were taken from day 0 to 7 of adult females for the developmental profiles of genes encoding enzymes in the JH biosynthetic pathway. For each time point, three biologically independent pools of 10 animals each were collected. To determine the tissue distribution of the genes of interest, the following tissues were dissected from 3 independent pools of 10 animals each: brain (Br), nerve cord (NC), corpora allata (CA), fat body (FB), midgut (MG), Malpighian tubules (MT), ovary (Ov) from females, and accessory gland (AG) and testes (Te) from males. For the RNAi experiments ( §2.6), 3 biologically independent pools comprising CA from 7 animals were collected. Pooled samples were homogenised with RNase-free pestles and total RNA was extracted using the RNeasy Mini Kit (Qiagen) according to the manufacturer's instructions. An additional DNase treatment (RNase-free DNase set, Qiagen) was performed to eliminate potential genomic DNA contamination. Because of the small size of the CA, RNA from this tissue was extracted using the RNAqueous-Micro Kit (Ambion), followed by the recommended DNase step. Quality and concentration of the resulting RNA samples were measured using a Nanodrop spectrophotometer (Thermo Scientific.).

Sequencing genes involved in the JH biosynthetic pathway
Since no genome or full transcriptome sequence data are currently available for D. punctata, a first set of degenerate primers was developed based on a multiple sequence alignment of known orthologous sequences from different insects. Such an alignment was made for Acetoacetyl-CoA thiolase (Thiol), HMG-CoA reductase (HMGR), Phosphomevalonate kinase (PK), Diphosphomevalonate decarboxylase (PPMD), Isopentenyl diphosphate isomerase (IPPI), Juvenile hormone acid methyltransferase (JHAMT) and the JH target vitellogenin (Vg). Degenerate primers are listed in S1 Table. For Methyl farnesoate epoxidase (CYP15A1) the full sequence was already characterised (GenBank accession number AY509244) [19]. For HMG-CoA synthase (HMGS), Mevalonate kinase (MK), Farnesyl diphosphate synthase (FPPS) and Farnesol dehydrogenase (FOLD), a partial sequence was found in the D. punctata CA EST database [20]. Their sequence was partially confirmed using gene-specific primers (S1 Table). A temperature-gradient polymerase chain reaction (PCR) was run using Taq DNA polymerase (Sigma-Aldrich Co.) with D. punctata day 4 CA cDNA as a template. Bands of the expected size were cut out and further purified using the GenElute Gel extraction Kit (Sigma-Aldrich Co.). The resulting DNA fragments were subcloned into a CloneJET cloning vector using the CloneJET PCR Cloning Kit (Fermentas) and sequenced. Sequences for PMK, PPMD and JHAMT identified using degenerate primers, were too short to submit to NCBI's GenBank; therefore, RACE (Rapid Amplification of cDNA Ends) was performed. Their sequences were confirmed using primers listed in S1 Table. Radiochemical assay (RCA) The in vitro radiochemical assay (RCA) for JH biosynthesis was performed [21,22]. CA were incubated in TC199 medium for 3h, then transferred to new medium containing JH precursors for another 3h incubation. JH biosynthesis in both incubations was measured, and first incubation measurements were used as a control. JH precursors acetyl CoA, mevalonic acid (MA), diphosphomevalonate (DPPM) and farnesol (Sigma-Aldrich, Canada) were dissolved in water before use.
cDNA synthesis and quantitative real-time PCR (q-RT-PCR) cDNA was transcribed from an equal amount of RNA using the SuperScript III First-Strand Synthesis SuperMix for q-RT-PCR in a final volume of 20 μl following the manufacturer's instructions (Invitrogen Life Technologies). All samples were reverse transcribed together in a single run. The resulting cDNA samples were diluted 10-fold with PCR grade water. A calibrator sample was prepared by pooling 5 μl of each cDNA sample. In the same run, negative control reactions were set up without reverse transcriptase enzyme to test for genomic DNA contamination.
q-RT-PCR primers were designed using IDT's (Integrated DNA Technologies) PrimeQuest design tool (http://eu.idtdna.com/PrimerQuest/Home/Index). Primer sets were subsequently validated by determining relative standard curves for each gene transcript using a five-fold serial dilution of the calibrator cDNA sample. Efficiency and correlation coefficients (R 2 ) were determined for each primer pair. Primers used for q-RT-PCR profiling are listed in Table 1.
q-RT-PCR reactions were carried out in triplicate in a total volume of 10 μl containing 5 μl of IQ SYBR Green Supermix (Bio-Rad), 1 μl forward and reverse primer (5 μM), 2 μl of MQwater and 1 μl of cDNA. All reactions were performed using Bio-Rad's CFX384 Touch Real-Time PCR Detection System using a two-step thermal cycling profile: 95°C for 3 min, followed by 40 cycles of 95°C for 10 s and 59°C for 30 s. Upon completion of every run, a dissociation protocol (melt curve analysis) was performed to check for formation of primer dimers. A few Table 1. q-RT-PCR primer sequences and reaction efficiencies and correlation coefficients in the q-RT-PCR assay. For reference gene q-RT-PCR specifications, the reader is referred to Marchal et al. [20]. representative PCR products were run on a 1.2% agarose gel containing GelRedTM (Biotium) and visualised under UV to confirm target specificity. Prior to target gene profiling, previously described housekeeping genes were tested for their stability in the designed tissue distribution, temporal profiling and RNAi experiments. The optimal housekeeping genes were selected using the geNorm and Normfinder software as described previously [23]. For each tested cDNA sample, the normalization factor for the reference genes relative to the calibrator samples was calculated and used to determine the normalized expression levels of the target genes relative to the calibrator [24].

RNA interference (RNAi)
CA cDNA was used as a template to amplify fragments of genes to be used in dsRNA construction. These fragments were subcloned and sequenced as described above. Two separate constructs were designed in different regions of the genes to eliminate off-target effects. Primers used are given in Table 2. dsRNA was synthesized using the MEGAscript RNAi kit (Ambion). The procedure is based on the high-yield transcription reaction of a user-provided linear transcript with a T7 promoter sequence. Transcription was carried out at 37°C overnight. The reaction mixture was treated with DNase I and RNase, and then purified by phase-solid phase adsorption purification, according to the manufacturer's instructions (Ambion). The dsRNA concentration was determined using a Nanodrop spectrophotometer (Thermo Scientific.). Diluted dsRNA was run on a 1.2% agarose gel to examine integrity of the construct and efficiency of duplex formation. The negative control construct (-pJET) was designed in a non-coding region of the pJET 1.2/blunt cloning vector (CloneJet PCR Cloning Kit, Thermo Scientific). Several RNAi trials with different injection timings and dsRNA construct concentrations were investigated to obtain an efficient gene silencing (data not shown). Based on the results of RNAi trials, newly molted mated females were injected with 3 μg HMGR dsRNA on day 0 and day 2, and with 3 μg JHAMT dsRNA on day 1 and day 3. CA, ovary and fat body samples were taken on day 4 as described above. Fat body was immediately stored in liquid N 2 prior to RNA extraction. CA were dissected and cleaned in TC199 medium (GIBCO; 1.3 mM Ca2+, 2% Ficoll, methionine-free) for use in the RCA or flash frozen in liquid N 2 to prevent RNA degradation. Basal oocyte length was measured and the ovaries were collected for vitellogenin measurements ( § 2.7) and histology ( § 2.8).

Vitellogenin ELISA
Single dissected ovaries were homogenized and extracted twice in 50 μl PBS according to Mundall et al. [25]. Total ovary protein content was determined using a Bradford assay [26] performed according to the manufacturer's (Sigma-Aldrich) 96 well plate protocol. Vitellin was quantified using an indirect enzyme-linked immunosorbent assay (ELISA) following the protocol described in Marchal et al. [27]. A rabbit polyclonal antibody made against D. punctata mature egg homogenate was used as the first antibody [28]. Goat anti-rabbit IgG, HRP-linked antibody (1:3000 in 1% BSA) was then added as the second antibody. Wells were treated with 100ul of TMB for 10 min. Absorbance was measured at 650 nm with a Molecular Devices SpectraMax Pus 384 microplate reader.

Histology and microscopy
To view overall structure, a subset of oocytes was fixed for 3 days in 3% glutaraldehyde in 0.1 M phosphate buffer, post-fixed for 1h with 1% OsO 4 in 0.1 M phosphate buffer, dehydrated in ascending ethanol series, and embedded in Spurr's resin. Sections (1 μm) were cut with a Leica EM UC6 ultramicrotome, then mounted and stained with methylene blue and toluidine blue. All light microscopy was conducted using a Leica DMI3000 inverted microscope.

Results
Identification of genes encoding JH biosynthetic enzymes and Vg in D. punctata CYP15A1 was previously identified in D. punctata and the genes encoding HMGS, MK, FPPS and FOLD were present in a CA EST database of D. punctata [19,20]. We confirmed the sequence of HMGS, MK, FPPS and FOLD. For the remaining enzymes in the JH biosynthetic pathway, we aligned (predicted) orthologous sequences from different insect orders and designed degenerate primers based on conserved domains. We succeeded in cloning an additional 5 (partial) sequences for enzymes in the conserved mevalonate pathway: Thiol, HMGR, PMK, PPMD and IPPI and one extra sequence encoding an enzyme in the JH-specific part of the pathway, JHAMT. The complete sequences for HMGR and JHAMT were obtained. The ORFs of HMGR and JHAMT encode proteins of 825 and 274 amino acids in length, respectively. Amino acid alignments of these proteins with known insect orthologs are given in S1 Tissue specificity and developmental transcript profiles during the first gonadotropic cycle are consistent with roles of these genes in JH biosynthesis Using q-RT-PCR, the relative transcript levels of 11 genes in the JH biosynthetic pathway were determined in several tissues of day 4 mated females and males. Relative transcript levels for the Diploptera orthologs of genes in the JH pathway were normalized to transcript levels of the reference genes EF1a and Tubulin [23]. To represent the tissue distribution data for the 11 enzyme-encoding genes in one figure, the transcript levels were normalized to the transcript level of PMK measured in the calibrator sample. The orthologous genes encoding enzymes in the JH biosynthetic pathway were most highly expressed in the CA, which is consistent with the functions of the enzymes (Fig. 2). Most genes appear to be exclusively expressed in the day 4 female CA: Thiol, HMGS, HMGR, MK, PPMD, FPPS, JHAMT and CYP15A1; whereas a few show a broader tissue distribution: PMK, IPPI and FOLD. These latter 3 appear to be expressed not only in nervous tissues but also in peripheral tissues such as the ovary, fat body and Malpighian tubules. Moreover, there are 1000-fold differences in relative mRNA levels of the enzyme-encoding genes, with PMK, PPMD, IPPI and FOLD being present in low abundance, Thiol, HMGS and HMGR being of intermediate abundance and MK, FPPS, JHAMT and CYP15A1 being represented in high abundance (S2 Fig.).
Our next step was to follow the relative transcript levels in the CA of mated females throughout the first gonadotropic cycle. Target gene expression was normalized to transcript levels of EF1a and Armadillo according to a previous study [23]. In general, the expression of the 11 genes is correlated with the in vitro rates of JH biosynthesis in the CA and relative mRNA levels for DippuVg measured in the female fat body (Fig. 3). The relative mRNA levels in the CA for most of the genes of interest were low at the beginning of the adult female stage when JH biosynthesis is low, rose during the vitellogenic portion of the first gonadotropic cycle reaching a peak on day 3-4 and thereafter began to decline on day 5, attaining a low level during oviposition on day 7. However, there appear to be three exceptions to this pattern: PMK, IPPI and FPPS. Relative mRNA quantities for PMK and IPPI were very low compared to other genes and did not display dramatic changes during the first gonadotropic cycle. FPPS transcripts were high on day 0, remained high until day 4 and then declined on day 5 when vitellogenesis slows. Addition of JH precursors stimulates JH biosynthesis in CA with low JH biosynthetic activity in vitro Previous studies suggested that JH synthesis is controlled by the rate of flux of isoprenoids in A. aegypti [11]. To determine the role of other JH precursors in regulating JH biosynthesis, we tested the rate of JH biosynthesis with the addition of JH precursors in the early steps of mevalonate pathway or the addition of farnesol. The addition of acetyl CoA, DPPM or farnesol to the incubation medium had a significant stimulatory effect on JH biosynthesis, whereas MA did not (Fig. 4A). For the first time, we demonstrated that acetyl CoA, as the first precursor in the JH biosynthetic pathway, was able to stimulate JH biosynthesis. The rank order of the stimulatory effects of the different JH precursors on JH biosynthesis is as follows: farnesol > acetyl CoA > DPPM > MA.
We also evaluated the sensitivity of CA to farnesol during the first gonadotrophic cycle. CA were dissected from day 0, 3, 4, 5, and 7, and incubated with medium containing 40 μM farnesol. On days 3 and 4, when the CA show high JH biosynthetic activity, the addition of farnesol had no effect on in vitro JH biosynthesis (Fig. 4B). On the other hand, JH biosynthesis in CA of day 0, 5 and 7 mated females was significantly increased. Injection of HMGR-JHAMT dsRNA resulted in a significant downregulation of the target genes but also of other genes in the JH biosynthetic pathway A systemic RNAi response was observed following injection of HMGR-and JHAMT dsRNA into animals every other day during the first gonadotropic cycle. Relative transcript levels were measured in the CA of day 4 animals using q-RT-PCR. A significant knockdown of 64% and 94% was measured for HMGR and JHAMT, respectively. Off-target effects were investigated by checking the Ct value of housekeeping genes (S2 Table). Moreover, two different dsRNA constructs were used to eliminate off-target effects (primers listed in Table 2). Both constructs yielded a similar phenotype.
q-RT-PCR was used to determine the relative mRNA levels of the other genes involved in the JH biosynthetic pathway using the control and treated CA described above. A significant reduction in relative expression levels was observed for all genes encoding the enzymes directly involved in the pathway, with the exception of the genes encoding FOLD and the epoxidase, CYP15A1 (Fig. 5).  Silencing HMGR and JHAMT resulted in a significant reduction in rates of JH biosynthesis and slows ovarian development The downregulation of HMGR and JHAMT resulted in a 73% reduction in the rates of JH biosynthesis in the CA of day 4 adult females as measured using the in vitro RCA (Fig. 6A). To confirm the role of JH in inducing Vg transcription in the fat body and uptake in the developing basal oocytes, fat body and ovaries were dissected from day 4 control and treated adult cockroaches. The basal oocyte length was significantly decreased following silencing of HMGR and JHAMT. The average oocyte length in the HMGR-JHAMT RNAi animals measured 0.8 mm compared to the control of 1.21 mm (Fig. 6B). The transcript level of the Vg in the fat body of the HMGR-JHAMT dsRNA-treated animals was reduced 80% (Fig. 6C). In addition, the knockdown of HMGR-JHAMT mRNA resulted in a significant reduction in vitellin content compared to the controls (Fig. 6D). We also studied the histology of the developing basal oocytes in control and treated females. Control oocytes were fully patent on day 4, showing large intercellular spaces between the follicle cells and the clear presence of yolk spheres. In the HMGR-JHAMT RNAi animals, on the other hand, patency was not fully induced and as a result, yolk was not deposited in these oocytes (Fig. 7).

Sequence of genes encoding enzymes in the JH biosynthetic pathway
Through in silico data mining and degenerate primer PCR and RACE, we have successfully characterized 11 of the 13 enzyme-encoding genes in the JH biosynthetic pathway of D. punctata. Cheng et al [9] showed that enzymes in the early steps of the mevalonate pathway were generated by single copy genes in many insects, whereas genes encoding FPPS and enzymes in the JH-specific pathway probably underwent gene duplication in Lepidoptera. In D. punctata, only a single copy of these genes was identified. However, at this point, the existence of multiple copies of several of these genes cannot be ruled out and this issue will have to await the assembly of the Diploptera genome.
The full-length HMGR and JHAMT were cloned from D. punctata CA. HMGR encodes a protein of 825 amino acids in length. Following a conserved domain search, HMGR was observed to contain the typical conserved motifs found in members of the HMG-CoA reductase superfamily of proteins. JHAMT encodes a protein of 274 amino acids in length and contains the motifs typically found in AdoMet-dependent methyltransferases. Still missing are sequences for farnesyl diphosphate pyrophosphatase (FPPP) and farnesal dehydrogenase (FALD). Molecular analysis of these genes was long hampered as a consequence of the minute size of the CA but using a newly developed sensitive assay for measuring JH precursor pools employing fluorescent tags, the genes encoding FPPP and FALD were recently characterized in A. aegypti [14,15,29]. These were, however, not included in our study because the availability of only dipteran sequences made their characterization in the phylogenetic basal D. punctata problematic and therefore remain uncharacterized in this species to date.

Transcription of enzymes in the JH biosynthetic pathway partly regulates JH biosynthesis
The 11 orthologs of JH biosynthesis genes cloned in our model D. punctata are predominately expressed in the CA, and their relative transcript levels correlate well with the rates of JH biosynthesis (Figs. 2 and 3). Our results suggest that CA are highly specialized tissues for the biosynthesis of JH, and genes identified in our study are encoding functional enzymes in the JH pathway. Enzymes in the mevalonate pathway are responsible for not only JH production but also for the production of other terpenoids such as defensive secretions and pheromones (as reviewed by [4,30]). It is therefore of interest that all mevalonate enzyme-encoding genes except PMK and IPPI are exclusively expressed in the CA with trace amounts in other tissues. This can be explained by the strikingly high rates of JH biosynthesis in the CA of day 4 adult female D. punctata. For the later enzymes involved in the JH specific steps of the pathway, the transcription of JHAMT and CYP15A1 is CA-specific, whereas FOLD is expressed in many tissues. Similar results were observed in A. aegypti and A. mellifera [11,12,31]. Because of the multiple functions of farnesol, the oxidation of farnesol to farnesal does not appear to be a JH biosynthesis-specific reaction [31].
Changes in rates of JH biosynthesis during the first gonadotropic cycle of D. punctata are dynamic, corresponding to specific reproductive events (see below) and are therefore tightly regulated. The transcript levels of enzyme-encoding genes in JH biosynthesis are highly coordinated with the rates of JH biosynthesis (Fig. 3), suggesting that at least part of the regulation of JH biosynthesis involves coordinated changes in the transcription of the genes in the biosynthetic pathway, in agreement with studies performed in A. aegypti [11] and B. mori [8], although the ability of each gene to regulate JH biosynthesis appears to differ. In addition, the expression patterns of HMGS and HMGR mRNA were similar to the patterns of their enzyme activities [32,33]. Although the transcription of most genes in the biosynthetic pathway showed a correlation with the JH biosynthesis, there are a few exceptions, including PMK, IPPI and FPPS. PMK and IPPI, which are expressed in multiple tissues (Fig. 2), and may not be specifically regulated in the CA. For FPPS, 3 FPPS homologs were found in B. mori [10], and 7 in A. mellifera [12]. Although we currently have identified only 1 FPPS gene in D. punctata, additional FPPS genes may exist in the CA that regulate JH biosynthesis.

Effect of exogenous JH precursors on rates of JH biosynthesis
We have found that JH precursors are able to stimulate JH biosynthesis (Fig. 4A), which suggests that the rate of JH biosynthesis is not only controlled by transcription of the genes in the pathway, but also by the flux of substrates in the pathway. Farnesol stimulates JH biosynthesis by CA from day 0, 5 and 7 mated females. However, application of farnesol to CA showing high JH biosynthetic activity (day 3 and 4) does not result in a significant change in JH biosynthesis (see Fig. 4B). These results suggest that during the first gonadotrophic cycle of D. punctata, the supply of this precursor is rate-limiting in CA with showing low JH biosynthetic activity, whereas there are other factors, including nutrients and neurotransmitters that control JH production at high rates of JH biosynthesis. Nevertheless, 14 JH precursors are involved in the JH biosynthetic pathway, and each JH precursor connects to the 'upstream' and 'downstream' enzymes. Many factors could affect the stimulatory effect of JH precursor in JH production, including cell permeability of the added compounds, the activity of upstream/ downstream enzymes, and the size of other JH precursor pools.
A feedback mechanism involved in the regulation of the transcription of JH biosynthetic genes HMGR and JHAMT have been studied in the JH pathway since they are representative enzymes involved in the well-conserved mevalonate pathway and in the JH-specific portion of the pathway respectively [6,[34][35][36][37][38][39]. We therefore chose to silence HMGR and JHAMT on a posttranscriptional level using RNAi during the first gonadotropic cycle of D. punctata. RNAi trials have shown that fewer injections of HMGR or JHAMT dsRNA and with lower concentrations of construct, did not result in a significant silencing of our target gene mRNA levels (data not shown). We therefore conclude that the RNAi response in D. punctata appears to be systemic, but is not as effective as in another hemimetabolous insect model, the desert locust, Schistocerca gregaria in which smaller quantities of the JHAMT dsRNA construct induced a longer downregulation of the same target gene during the first gonadotropic cycle [6]. A further in-depth study on the RNAi machinery D. punctata should be performed to explain the differences in sensitivity compared to other hemimetabolous insect species.
JH biosynthesis is controlled by many factors including neuropeptides (allatostatins, allatotropins, short neuropeptide F) [40,41], neurotransmitters (octopamine, dopamine and glutamate) [42][43][44] and JH itself [1,16]. Topical application of JH or a JH analog resulted in a suppression of JH biosynthesis, which suggests a negative feedback regulation of CA in D. punctata [45]. In higher animals, cholesterol, the bulk end-product of mevalonate pathway, shows a feedback regulation on the transcription and post-transcription of enzymes in the mevalonate pathway [30,46,47]. Later studies showed that farnesol, an intermediate product in the JH biosynthetic pathway, is able to accelerate the degradation of HMGR, and the increase of mevalonate flux raises the activity of FPPP expressed in CHO cells [48,49]. In our study, regulating transcription of the JH biosynthetic enzymes seems to affect the entire biosynthetic pathway, rather than individual steps. The silencing of HMGR and JHAMT resulted in a significant decrease in the transcript level of several other genes in the pathway (Fig. 5). A possible explanation is that the accumulation of JH precursors as a result of RNAi treatment resulted in a feedback on the expression of other genes in the JH biosynthetic pathway to balance the size of JH precursor pools and the enzyme activity.

The essential role of JH in reproduction
Our results clearly show that select silencing of genes encoding enzymes in the JH biosynthetic pathway effectively reduces JH biosynthesis in vitro (Fig. 6). As a well-known model for the study of the regulation of JH biosynthesis, D. punctata displays a consistent characteristic profile of JH production during the first gonadotropic cyclethat is closely correlated with specific reproductive events such as vitellogenesis and chorionation. Once JH production rises on day 2, vitellogenin synthesis in the fat body commences, and vitellogenin is taken up from the hemolymph and is incorporated into the developing basal oocytes. The elevated JH titer results in patency in the follicular epithelium of the basal oocytes, permitting the uptake of Vg and subsequent yolk formation [27,50]. On day 5, the spaces in the follicular epithelium close as rates of JH biosynthesis decline and Vg transcription in the fat body and circulating Vg levels decrease. At this point, choriogenesis begins and JH biosynthesis remains low until oviposition on day 7 [as reviewed by 16]. Our RNAi study has focused on day 4 of the first gonadotropic cycle, a time when JH titre is high in controls and animals are vitellogenic [51] as characterized by high DippuVg mRNA transcript levels in the fat body and JH-dependent induction of patency in the maturing basal oocytes. Following the reduction in JH production by the silencing of Dip-puHMGR and DippuJHAMT, mRNA levels of DippuVg in the fat body are significantly reduced, incorporation of vitellin into the oocytes is impaired and patency does not occur (Figs. 6 and 7). Previous reports have shown that in S. gregaria, oocyte length is significantly affected following treatment of females with JHAMT dsRNA [6]. RNAi was also used in the cotton bollworm Helicoverpa armigera to effectively silence HMGR, resulting in reduced Vg expression [38]. Each of these studies highlights specific aspects of JH-regulated reproductive events. The results described in the current paper confirm the central role played by JH during the female reproductive cycle of D. punctata.
JH biosynthesis is a fundamental process in regulating insect development, metamorphosis and reproduction [1]. Although the genes directly involved in JH biosynthesis have been characterized using molecular techniques in several insect species, there was no similar study in the hemimetabolous insects. Thus, our study is the first to characterize the majority of the genes directly involved in JH biosynthesis in a hemimetabolous insect and a well-known model for studying the physiology of JH. This paper now provides the molecular tools to study the regulatory mechanisms of JH production in D. punctata (for a first study, see [52]). In addition, the role of JH biosynthetic enzymes and the JH precursor supplies in regulating JH biosynthesis have been demonstrated, as has the critical function of JH as the master regulator of cockroach female reproduction.  [36], the honeybee Apis mellifera (GenBank accession number: XP_623118.1), the red flour beetle Tribolium castaneum (GenBank accession number: XP_973850.1), the mosquito Aedes aegypti (GenBank accession number: XP_001659923.1), the fly Drosophila melanogaster (GenBank accession number: NP_732900.1) and the silkworm Bombyx mori (GenBank accession number: BAF62108.1 [8]) (B) DippuJHAMT with JHAMT from the desert locust, S. gregaria (GenBank accession number: ADV17350.1 [6]), the red flour beetle T. castaneum (GenBank accession number: NP_001120783.1 [53]), the fly D. melanogaster (GenBank accession number: AB113579.1 [54]), the mosquito A. aegypti (GenBank accession number: ABD65474.1 [39]), the silkworm B. mori (GenBank accession number: NP_001036901.1 [7]) and the honeybee A. mellifera (GenBank accession number: AGG79412.1). (PDF)  Table. (Degenerate) primer sequences for cloning of (partial) sequences for orthologous genes encoding JH biosynthesis enzymes in D. punctata. Gene name abbreviations are shown in Fig. 1.