PubMed@Masakazu SAGA:28257485 JSONTXT

{"target":"https://pubannotation.org/docs/sourcedb/PubMed@Masakazu%20SAGA/sourceid/28257485","sourcedb":"PubMed@Masakazu SAGA","sourceid":"28257485","text":"Deep sequencing of the prothoracic gland transcriptome reveals new players in insect ecdysteroidogenesis\r\n\r\nEcdysteroids are steroid hormones that induce molting and determine developmental timing in arthropods. In insect larva, the prothoracic gland (PG) is a major organ for ecdysone synthesis and release. Released ecdysone is converted into the active form, 20-hydroxyecdysone (20E) in the peripheral tissues. All processes from ecdysone synthesis and release from the PG to its conversion to 20E are called ecdysteroidogenesis and are under the regulation of numerous factors expressed in the PG and peripheral tissues. Classical genetic approaches and recent transcriptomic screening in the PG identified several genes responsible for ecdysone synthesis and release, whereas the regulatory mechanism remains largely unknown. We analyzed RNA-seq data of the silkworm Bombyx mori PG and employed the fruit fly Drosophila melanogaster GAL4/UAS binary RNAi system to comprehensively screen for genes involved in ecdysone synthesis and/or release. We found that the genes encoding δ-aminolevulinic acid synthase (CG3017/alas) and putative NAD kinase (CG33156) were highly expressed in the PG of both B. mori and D. melanogaster. Neither alas nor CG33156 RNAi-induced larvae could enter into the pupal stage, and they had a lower abundance of the active form ecdysteroids in their prolonged larval stage. These results demonstrated that alas and CG33156 are indispensable for ecdysteroidogenesis.\r\n\r\nIntroduction\r\n\r\nSteroid hormones play crucial roles in the regulation of many biological processes in vertebrates and invertebrates. In insects, growth is controlled by molting and metamorphosis, and characteristic developmental events are regulated by steroid hormones called ecdysteroids. During the larval stage, ecdysone is synthesized from cholesterol through serial oxidation/hydroxylation steps in the prothoracic gland (PG) and secreted into the hemolymph. The released ecdysone is converted to the biologically active form of ecdysone, 20-hydroxyecdysone (20E), in peripheral tissues. The sequence of biological processes from ecdysone synthesis in the PG to 20E production in the peripheral tissues is called ecdysteroidogenesis. As the timing of molting and metamorphosis are determined by ecdysone production and secretion from the PG, it is important to understand the regulatory mechanisms of ecdysteroidogenesis.\r\n\r\nGenes expressed predominantly in the PG are expected to be involved in ecdysone synthesis and release. In the last decade, analyses of the PG gene expression using a plethora of molecular techniques has successfully identified genes that are indispensable for ecdysone biosynthesis. For example, Rieske-domain oxygenase Neverland, which is an enzyme for 7, 8-dehydrogenation of cholesterol, Cyp306a1/phantom, and Cyp307a1/spook, which are collectively called Halloween genes and encode cytochrome P450 monooxygenases, were identified via cDNA microarray analyses or fluorescent differential display. In addition to ecdysteroidogenic enzymes, recent studies showed additional factors made PG as an ecdysteroidogenic organ, such as an ecdysone-specific ABC transporter, atet, and a transcription factor, ooija board.\r\n\r\nHigh-throughput RNA sequencing (RNA-seq) is an effective and quantitative method for transcriptomic analysis. On an RNA-seq platform, a huge number of sequences in a cDNA library from a tissue of interest are determined. RNA-seq has been applied to characterize distinct tissues in various insect species. To identify novel factor(s) responsible for ecdysone synthesis and/or release, we have conducted RNA-seq analysis of the PG of the silkworm Bombyx mori. Focused screening of G protein-coupled receptors (GPCRs) in the PG transcriptome revealed a GPCR, BNGR-B2, as a receptor for the pigment-dispersing factor that enhances ecdysone synthesis and release in the PG. However, further players are considered essential for PG ecdysteroidogenesis and they would be discovered in the RNA-seq data.\r\n\r\nHere we report detailed analysis of the RNA-seq data and the use of an additional screening method, the GAL4/UAS-based RNA interference (RNAi), in Drosophila to reveal genes responsible for ecdysteroidogenesis in the PG. We analyzed the expression profile of 21,302 genes predicted in the Bombyx mori genome and transcriptomically screened for identified genes expressed more in the PG than in the brain. Screening via Drosophila RNAi highlighted two genes, alas and CG33156, as candidate genes responsible for ecdysteroidogenesis in the PG. Knockdown of alas or CG33156 in the PG caused developmental arrest at the third instar larva with no morphogenetic abnormality, along with a low abundance of 20E during an extended larval phase. These results suggested that alas and CG33156 are required for proper ecdysone synthesis and/or release in insect larval development.\r\n\r\nMaterials and methods\r\n\r\nAnimals\r\n\r\nDrosophila melanogaster strains were maintained on standard cornmeal medium at 25°C under 12-h light:12-h dark photoperiod. yw was used as the wild type fly strain. UAS-Dcr2; phm-GAL4, and UAS-Dcr2; 2-286-GAL4 were used for in vivo RNAi. Virgin females with UAS-Dcr2; phm-GAL4 or UAS-Dcr2; 2-286-GAL4 were crossed to UAS-Inverted Repeat (IR) males. Females were allowed to lay eggs overnight, and the numbers of adult or pupae were counted until day 14 or day 8, respectively. Flies with UAS-IR constructs were obtained from the Vienna Drosophila RNAi Center (VDRC) and the National Institute of Genetics (NIG-FLY). The UAS-mCD8::GFP line was obtained from the Bloomington Drosophila Stock Center.\r\n\r\nTranscriptomic analysis of silkworm PG\r\n\r\nRNA-seq of Bombyx mori PG and brain in wandering larvae were previously performed as outlined in (Accession number: DRA002282). Sequence reads were mapped to the reference model transcripts according to the B. mori genome annotations (http://silkworm.big.ac.cn/jsp/download.jsp). The number of reads uniquely aligned to the genome without any mismatches were digitally counted. Parts per million (ppm) to the total mapped reads were calculated for each gene, and the ppm ratio of PG to the brain was calculated to compare gene expression levels.\r\n\r\nQuantitative RT-PCR\r\n\r\nThe ring gland (RG), central nervous system, salivary gland, fat body, Malpighian tubules, trachea, midgut, and epidermis were dissected from wandering yw third instar larvae. For each sample, tissues were collected from 40 larvae, and three independent, biological replicates were performed. Total RNA was isolated using the TRIzol reagent, and the isolated RNA was treated with RQ DNaseI (Promega) to digest genomic DNA. Reverse-transcription was performed using the PrimeScript RT reagent kit (Perfect Real Time) (TaKaRa) with both random hexamers and oligo (dT) primer, following manufacturers’ instructions. Real-time PCR was performed with the Thermal Cycler Dice and SYBR Premix ExTaq II (TaKaRa). The expression levels were normalized to rp49, and the ΔΔCt method was used to calculate the expression ratio for each gene. Averages of triplicate samples were calculated for each tissue, and the highest expression level among tissues was set to 1. Samples with Ct values over 38 were excluded. The primers used in this study are listed in S1 Table.\r\n\r\nIn situ hybridization\r\n\r\nNucleotide fragments corresponding to candidate genes were amplified by RT-PCR and subcloned into the pGEM-T vector (Promega). Primers used are listed in S1 Table. Sense or antisense RNA probes were synthesized using the DIG labeling mix (Roche) with T7 and SP6 RNA polymerase. RG-brain complexes were dissected from yw wandering larvae and fixed in 4% cold formaldehyde/PBS. Subsequent procedures were carried out accordingly as previously reported but with some modifications. To detect RNA probes, the specimens were incubated in PBS with 0.3% Triton X-100 containing anti-DIG antibody conjugated with alkaline phosphatase, and 4-nitroblue tetrazolium chloride-5-bromo-4-chloro-3-indolyl-phosphate (NBT-BCIP) was used as a chromogenic substrate.\r\n\r\nMeasurement of ecdysteroids\r\n\r\nFemales with phm-GAL4/ TM3 Ser Kr-GFP were crossed to males homozygous for UAS-IR and allowed to lay eggs on standard cornmeal food for 6 h. GFP larvae (negative control), with genotype UAS-IR and TM3 Ser Kr-GFP, were collected at 96 and 120 (±3) h after egg laying (h AEL). Both wandering larvae and white pupae were observed at AEL 120 in the controls, and therefore, both were collected at 120 h AEL. On the other hand, GFP negative larvae, with genotype phm-GAL4 and UAS-IR, were collected at 96, 120, and 144 (±3) h AEL. Five pupae or ten larvae were collected for each sample in a 1.5-ml tube and stored at -20°C until further use. Collected larvae were homogenized with a close-fitting plastic pestle and sonicated for 10 min in 200 μl of buffer containing 50 mM Tris-HCl, pH 7.5, 150 mM NaCl, and 2 mM EGTA. Steroids were extracted from the samples with 1-butanol (Wako). Two-fold volumes of 1-butanol were added to sonicated samples and shaken vigorously for 5 min. After centrifuging at 1,000 ×g, supernatants were transferred to a new tube. Steroid extraction using 1-butanol was repeated two more times, and the solution was pooled in the same tube. The solvent was evaporated in a centrifugal evaporator, and the residue was dissolved in 50 μl of methanol (Wako). Dissolved samples were shaken vigorously for 5 min and sonicated for 10 min. After centrifuging at 17,800 ×g, the supernatant was transferred to a glass vial. Ten microliters were injected into a LC-MS/MS system to measure steroid amount. Calibration curves were generated with a 0.5–500 ng/ml dilution series of ecdysone, 20E and Makisterone A. Conditions for the LC-MS/MS analysis, Prominence gradient HPLC system (Shimadzu), and the triple quadrupole QTRAP 5500 mass spectrometer (AB SCIEX) are described in. Conditions to detect Makisterone A are shown in S2 Table. Makisterone A was purchased from Santa Cruz Biotechnologies. Ecdysone and 20-hydroxyecdysone were purchased from Sigma-Aldrich.\r\n\r\nMorphology of RG\r\n\r\nThe offspring of UAS-Dcr2; UAS-mCD8::GFP; phm-GAL4/TM6B Tb, and UAS-IR/UAS-IR flies were reared on standard media, and RG-brain complexes were dissected from day 5 and 8. RG-brain complexes were then observed using a U-MNIBA2 fluorescent microscope (BX-60, Olympus).\r\n\r\nResults\r\n\r\nScreening of genes expressed preferentially in the B. mori PG\r\n\r\nThe PG is an organ that specifically synthesizes and secretes ecdysone; therefore, higher expression of responsible genes for ecdysteroidogenesis in the PG was expected as compared to other tissues. To identify such genes, we performed RNA-seq using the PG and brain of B. mori wandering larva. The brain was used as a non-steroidogenic control organ. Fifty million single-end reads of 36 base pairs were determined per organ, and 20,084,675 reads from the PG (total of 42,418,141 reads) and 18,628,667 reads from the brain (total of 47,649,669 reads) were uniquely aligned to the B. mori genome (http://silkworm.big.ac.cn/jsp/download.jsp). The number of reads aligned to a gene was counted, and ppm of the mapped reads to all of the uniquely aligned reads was calculated to estimate the expression level of each gene. The ratios of ppm in the PG to those in the brain were calculated to compare gene expression levels. To evaluate the quantification performance of our RNA-seq, we analyzed the expression ratios of some housekeeping genes as well. In the top 30 genes expressed in the PG, we found 15 genes encoding ribosomal proteins (S3 Table). The mean and standard deviation of expression ratios of these 15 ribosomal protein genes were 1.22 and 0.144, respectively. In addition, the expression ratios of three elongation factor genes, EF1α, EF1γ, and EF2 were 1.87, 1.22, and 1.63, respectively (S3 Table). These results suggested that comparing ppm of any gene of interest was suitable to evaluate differential gene expression in the PG and brain.\r\n\r\nNext, we searched for genes potentially involved in ecdysteroidogenesis in the PG. Of the 21,302 genes predicted in the genome database, the top 1,000 highly expressed genes covered a significant portion of the PG transcriptome (13,957,412 reads; 69.5% of all mapped sequence reads). We searched gene fold-ratios \u003e10 in the top 1,000 genes. Following this criterion, we obtained 62 candidate genes that showed preferential expression in the PG (Fig 1, Table 1). Previously reported ecdysteroidogenic genes, such as neverland (Bmb000191, Bmb028931), spook (Bmb008079), short-chain dehydrogenase/reductase (known as nm-g/shroud Bmb002129), disembodied (Bmb019913), torso (Bmb009410), silk gland factor3/ventral vein lacking (Bmb024277), and Npc1a (Bmb016744, Bmb016745, Bmb020860) were in this candidate list (Table 1, Fig 1). Due to an incomplete silkworm genome database, RNA-seq reads mapped to different contigs derived from the same genes. For example, sequence reads for neverland mapped to Bmb028931 and Bmb000191 that encoded amino acid residues 114–301 and 302–453 of the Neverland protein (NP_001037626.1). Although the mapping was not integral, both split sequences were detected as genes specifically expressed in the PG. Therefore, other responsible genes for ecdysone synthesis and release possibly existed in the list of candidate genes.\r\n\r\nExpression profiles of the top 1,000 genes expressed in the PG.\r\n\r\nEach data point represents ppm in the PG (x-axis) and relative expression ratio compared with the brain (y-axis) of each gene. Genes that are known to be involved in ecdysone biosynthesis are represented by triangles.\r\n\r\nList of genes expressed \u003e10 times in the PG than in the brain.\r\n\r\nGene Name\tppm\tRatioPG/Br\tDescription\tAccession #\tFly homolog\t \tBmb012212\t552.0\t12.4\tactin [Aedes aegypti]\tXP_001659963.1\tCG10067/act57B\t \tBmb025648\t200.9\t47.4\taminopeptidase N-12 [Bombyx mori]\tAFK85028.1\tCG11956/SP1029\t \tBmb007487\t2108.3\t24.1\tchitinase-related protein 1 [Danaus plexippus]\tEHJ65741.1\tCG18140/cht3\t \tBmb040752\t248.2\t77.1\tchitinase-related protein 1 precursor [Bombyx mori]\tNP_001036861.1\tCG18140/cht3\t \tBmb024372\t263.7\t50.6\tcytochrome P450 18a1 [Bombyx mori]\tNP_001077078.1\tCG6816/Cyp18a1\t \tBmb019913\t966.7\t114.0\tcytochrome P450 302A1 [Bombyx mori]\tNP_001036953.1\tCG12028/dib\t \tBmb008079\t79847.5\t9721.9\tcytochrome P450, family 307, subfamily a, polypeptide 1 precursor [Bombyx mori]\tNP_001104833.1\tCG10594/spo\t \tBmb007257\t314.9\t130.4\tendonuclease-reverse transcriptase [Bombyx mori]\tADI61816.1\t--\t \tBmb002562\t385.8\t12.1\tinositol 1 [Papilio xuthus]\tBAM18672.1\tCG4026/IP3K1\t \tBmb012355\t14470.0\t60.8\tinsulin-related peptide binding protein precursor [Bombyx mori]\tNP_001040473.1\tCG15009/ImpL2\t \tBmb035169\t796.4\t29.6\tinsulin-related peptide binding protein precursor [Bombyx mori]\tNP_001239349.1\tCG15009/ImpL2\t \tBmb022329\t307.5\t18.1\tjuvenile hormone esterase [Bombyx mori]\tAAR37335.1\tCG10175\t \tBmb021298\t308.8\t10.3\tmalate dehydrogenase [Bombyx mori]\tNP_001093280.1\tCG5889/Men-b\t \tBmb031565\t221.1\t823.6\tneuropeptide receptor A34 [Bombyx mori]\tNP_001127750.1\tCG30340\t \tBmb000191\t11416.7\t7089.2\tneverland [Bombyx mori]\tNP_001037626.1\tCG40050/nvd\t \tBmb028931\t4236.1\t7173.9\tneverland [Bombyx mori]\tNP_001037626.1\tCG40050/nvd\t \tBmb024993\t646.6\t12.6\tno hit\t--\t--\t \tBmb038048\t178.8\t64.1\tno hit\t--\t--\t \tBmb021395\t1788.7\t13.6\tPREDICTED: 5-aminolevulinate synthase, nonspecific, mitochondrial-like [Bombyx mori]\tXP_004922915.1\tCG3017/Alas\t \tBmb005099\t457.7\t236.8\tPREDICTED: adenylate cyclase type 5-like [Bombyx mori]\tXP_004932850.1\tCG42513\t \tBmb005846\t1095.1\t214.7\tPREDICTED: adenylate cyclase type 5-like [Bombyx mori]\tXP_004932850.1\tCG42513\t \tBmb037415\t157.6\t366.9\tPREDICTED: adenylate cyclase type 5-like [Bombyx mori]\tXP_004932850.1\tCG42513\t \tBmb030773\t142.7\t241.7\tPREDICTED: bile salt-activated lipase-like isoform X1 [Bombyx mori]\tXP_004924253.1\tCG10175\t \tBmb004665\t272.4\t23.2\tPREDICTED: chaoptin-like [Bombyx mori]\tXP_004923576.1\tCG1744/chp\t \tBmb034135\t191.1\t18.8\tPREDICTED: chaoptin-like [Bombyx mori]\tXP_004923576.1\tCG1744/chp\t \tBmb040810\t503.4\t13.4\tPREDICTED: chaoptin-like [Bombyx mori]\tXP_004923576.1\tCG1744/chp\t \tBmb007256\t198.3\t230.9\tPREDICTED: ecdysone-inducible protein E75-like [Bombyx mori]\tXP_004924954.1\tCG18023/Eip78\t \tBmb021498\t256.3\t18.8\tPREDICTED: facilitated trehalose transporter Tret1-like [Bombyx mori]\tXP_004923499.1\tCG8234/Tret1-2\t \tBmb025408\t230.0\t34.8\tPREDICTED: FH1/FH2 domain-containing protein 3-like [Bombyx mori]\tXP_004929944.1\tCG42610/Fhos\t \tBmb038593\t169.3\t15.0\tPREDICTED: GRAM domain-containing protein 1B-like [Bombyx mori]\tXP_004929648.1\t--\t \tBmb001251\t590.6\t24.2\tPREDICTED: GRAM domain-containing protein 1B-like [Bombyx mori]\tXP_004930811.1\tCG34394\t \tBmb005705\t446.1\t554.0\tPREDICTED: lipase member H-like [Bombyx mori]\tXP_004931604.1\t--\t \tBmb002715\t307.8\t11.4\tPREDICTED: LOW QUALITY PROTEIN: papilin-like [Bombyx mori]\tXP_004921811.1\t--\t \tBmb015275\t313.1\t42.6\tPREDICTED: LOW QUALITY PROTEIN: papilin-like [Bombyx mori]\tXP_004921811.1\tCG33103/Ppn\t \tBmb021601\t1003.6\t21.7\tPREDICTED: LOW QUALITY PROTEIN: papilin-like [Bombyx mori]\tXP_004921811.1\tCG33103/Ppn\t \tBmb000943\t238.4\t10.9\tPREDICTED: LOW QUALITY PROTEIN: titin-like [Apis florea]\tXP_003693102.1\tCG1915/sls\t \tBmb036727\t300.0\t17.7\tPREDICTED: luciferin 4-monooxygenase-like, partial [Bombyx mori]\tXP_004927714.1\tCG8834\t \tBmb000190\t652.9\t3040.8\tPREDICTED: multiple coagulation factor deficiency protein 2 homolog isoform X2 [Bombyx mori]\tXP_004922199.1\t--\t \tBmb018637\t312.0\t10.0\tPREDICTED: NAD kinase-like isoform X2 [Bombyx mori]\tXP_004926312.1\tCG33156\t \tBmb016744\t1289.5\t19.0\tPREDICTED: niemann-Pick C1 protein-like isoform X1 [Bombyx mori]\tXP_004924425.1\tCG5722/Npc1a\t \tBmb016745\t753.7\t16.0\tPREDICTED: niemann-Pick C1 protein-like isoform X1 [Bombyx mori]\tXP_004924425.1\tCG5722/Npc1a\t \tBmb020860\t799.9\t13.3\tPREDICTED: niemann-Pick C1 protein-like isoform X1 [Bombyx mori]\tXP_004924425.1\tCG5722/Npc1a\t \tBmb000196\t199.7\t310.0\tPREDICTED: nose resistant to fluoxetine protein 6-like [Bombyx mori]\tXP_004933619.1\tCG3106\t \tBmb030828\t363.3\t23.2\tPREDICTED: phosphatidylinositol 3-kinase catalytic subunit type 3-like [Bombyx mori]\tXP_004925089.1\tCG5373/Pi3K59F\t \tBmb036448\t646.9\t10.6\tPREDICTED: pogo transposable element with ZNF domain-like [Bombyx mori]\tXP_004922578.1\t--\t \tBmb020216\t178.1\t11.6\tPREDICTED: protein FAM135A-like [Bombyx mori]\tXP_004931668.1\tCG32333\t \tBmb002130\t151.1\t17.6\tPREDICTED: serine protease easter-like [Bombyx mori]\tXP_004932592.1\tCG16705/SPE\t \tBmb008527\t145.1\t50.1\tPREDICTED: serine/threonine-protein kinase prpf4B-like [Bombyx mori]\tXP_004921796.1\tCG2054/cht2\t \tBmb033191\t143.7\t13.2\tPREDICTED: shootin-1-like [Bombyx mori]\tXP_004933010.1\t--\t \tBmb025563\t366.1\t17.3\tPREDICTED: solute carrier family 2, facilitated glucose transporter member 1-like [Bombyx mori]\tXP_004923552.1\tCG1086/Glut1\t \tBmb003383\t464.6\t84.0\tPREDICTED: solute carrier organic anion transporter family member 3A1-like [Bombyx mori]\tXP_004932455.1\tCG7571/Oatp74D\t \tBmb030561\t1017.9\t67.2\tPREDICTED: solute carrier organic anion transporter family member 3A1-like [Bombyx mori]\tXP_004932455.1\tCG7571/Oatp74D\t \tBmb010938\t249.7\t35.5\tPREDICTED: transmembrane protein nessy-like [Bombyx mori]\tXP_004933932.1\tCG9655/nes\t \tBmb004658\t149.7\t11.2\tPREDICTED: ubiquitin-conjugating enzyme E2 R2-like [Bombyx mori]\tXP_004922372.1\tCG7656\t \tBmb002531\t161.2\t31.3\tPREDICTED: uncharacterized protein LOC101738974 [Bombyx mori]\tXP_004926099.1\tCG8420\t \tBmb039058\t221.7\t30.8\tputative trypsin-like serine protease [Danaus plexippus]\tEHJ67268.1\tCG7996/snk\t \tBmb002258\t613.9\t19.4\tras-related GTP-binding protein Rab3 [Bombyx mori]\tNP_001037620.1\tCG7576/rab3\t \tBmb038774\t153.3\t42.6\tserine protease inhibitor 11 precursor [Bombyx mori]\tNP_001139704.1\tCG10913/Spn6\t \tBmb032080\t152.8\t17.0\tserine protease inhibitor 3 precursor [Bombyx mori]\tNP_001040318.1\tCG11331/Spn27\t \tBmb002129\t462.1\t2152.3\tshort-chain dehydrogenase/reductase [Bombyx mori]\tNP_001171333.1\tCG12068/sro\t \tBmb024277\t2449.4\t20.6\tsilk gland factor 3 [Bombyx mori]\tNP_001037456.2\tCG10037/dfr\t \tBmb009410\t354.3\t1320.2\ttyrosine-protein kinase receptor torso [Bombyx mori]\tNP_001164049.1\tCG1389/torso\t \t\r\n\r\nPhenotypic screening using the Drosophila GAL4/UAS RNAi system\r\n\r\nMany of the ecdysteroidogenic enzymes and regulatory mechanisms for PG function are conserved between Bombyx and Drosophila. The GAL4/UAS-based RNAi system in Drosophila is a suitable approach for comprehensive screening. Therefore, we used RNAi screening in Drosophila to assess the contribution of candidate genes identified in our B. mori RNA-seq analysis to ecdysone synthesis and/or release. Drosophila genes homologous to the Bombyx PG-preferential genes were searched via BLASTX against Fly Base (http://flybase.org/) with a cut-off value of 1e-5. Out of 62 genes obtained from the Bombyx RNA-seq expression profile, 11 genes were known to be involved in ecdysone biosynthesis or ecdysone metabolism as described above, and nine genes (Bmb007257, Bmb024993, Bmb038043, Bmb038593, Bmb005705, Bmb002715, Bmb000190, Bmb036448, and Bmb033191) had no homologous genes in the Drosophila genome (Table 1). After excluding the redundant B. mori genes (Bmb022329/Bmb030773, Bmb012355/Bmb035169, Bmb007487/Bmb040752, Bmb015275/Bmb021601, Bmb003383/Bmb030561, Bmb005099/Bmb005846/Bmb037415, and Bmb004665/Bmb034135/Bmb040810), 33 homologous genes in Drosophila were determined (Table 1). We obtained 53 transgenic lines carrying IR sequences corresponding to 27 out of the 33 candidate genes (S4 Table). To knockdown each target gene in the PG specifically, male flies with the IR construct were crossed to UAS-Dcr2; phm-GAL4/TM3 Ser females, and the number of adults emerged was counted. Out of 53 lines corresponding to the 27 genes tested, 11 lines corresponding to eight genes showed a significant decrease of adult flies as compared to controls (chi-square test, p \u003c 0.01) (Table 2). The other six genes were not examined because RNAi lines were not available. To determine their lethal phase, the males were crossed to Dcr2; phm-GAL4/TM6B Tb females to observe if the RNAi strain with the homozygous UAS-IR genotype was viable. The RNAi larvae were fed for 7 days after oviposition, and their developmental stages were classified according to spiracle and mouth hook morphology. In all nine lines examined, we observed either no or significantly less (chi-square test, p \u003c 0.01) pupae in which the target gene was disrupted (Table 2). CG10913/Serpin6 (Spn6) RNAi animals died at the second instar larval stage. Other RNAi larvae could grow up to the third instar, but they could not form pupae except for some pupariated individuals. To confirm that the developmental abnormality did not result from unexpected expression of phm-GAL4 other than in the PG, the UAS-IR males were crossed to another PG-expressing GAL4 driver line, Dcr2; 2-286-GAL4/TM6B Tb females, and the number of pupae was counted. For all nine lines that showed developmental arrest by phm-GAL4, this same phenotype was observed when 2-286-GAL4 was used (Table 2). These results suggested that these genes play crucial roles in Drosophila molting or metamorphosis.\r\n\r\nResults of RNAi driven by phm-GAL4 or 2-286-GAL4.\r\n\r\n\t\t\t\t\tDcr2; phm-GAL4/TM3 Ser\tDcr2; phm-GAL4/TM6B Tb\tDcr2; 2-286-GAL4/TM6B Tb\t\t \t\t\t\t\t\tnumber of adult\tnumber of pupa\tnumber of pupa\t\t \tGene Name\tRatio PG/BR\tDescription\tFly homolog\tRNAi line\tcontrol\tRNAi\tcontrol\tRNAi\tcontrol\tRNAi\tlethal stage\t \tBmb021395\t13.6\t5-aminolevulinate synthase [Bombyx mori]\tCG3017/alas\tN 3017R1\t148\t0\t129\t0\t23\t0\tL3\t \t\t\t\t\tN 3017R2\t91\t12\t89\t34\t269\t0\tL3\t \tBmb018637\t10.0\tPredicted: NAD kinase-like isoform X2 [Bombyx mori]\tCG33156\tN 6152R3\t60\t0\t104\t0\t48\t0\tL3\t \t\t\t\t\tN 6152R4\t237\t10\t249\t53\t81\t0\tL3\t \tBmb000943\t10.9\tPredicted: titin-like [Apis florea]\tCG1915/sls\tV 47298\t136\t0\t40\t0\t144\t0\tL2/L3\t \t\t\t\t\tV 47301\t111\t0\t167\t0\t23\t0\tL2/L3\t \tBmb001251\t24.2\tPredicted: GRAM domain-containing protein 1B-like [Bombyx mori]\tCG34394\tV 23105\t121\t1\t131\t23\t100\t32\tL3\t \tBmb002258\t19.4\tras-related GTP-binding protein Rab3 [Bombyx mori]\tCG7576/rab3\tV 100787\t91\t0\t156\t0\t98\t1\tL3\t \tBmb038774\t42.6\tserine protease inhibitor 11 precursor [Bombyx mori]\tCG10913/Spn6\tN 10913R1\t203\t1\t152\t0\t42\t0\tL2\t \tBmb000196\t310.0\tPredicted: nose resistant to fluoxetine protein 6-like [Bombyx mori]\tCG3106\tV 42701\t226\t21\t--\t--\t--\t--\t--\t \tBmb002531\t31.3\tPredicted: uncharacterized protein LOC101738974 [Bombyx mori]\tCG8420\tN 8420R2\t305\t45\t--\t--\t--\t--\t--\t \t\r\n\r\nIn the “RNAi line” column, V and N indicate that the RNAi lines were obtained from VDRC and NIG-FLY, respectively.\r\n\r\nExpression analysis for candidate genes\r\n\r\nGenes that specifically function in ecdysone synthesis or release are expected to be expressed in the RG, which is a complex organ mostly comprised of PG cells. To examine tissue distribution of candidate genes, the expression levels in several tissues (RG, central nervous system, salivary gland, trachea, fat body, gut, Malpighian tubule, and epidermis) of wandering yw third instar larvae were analyzed by real-time RT-PCR. alas and CG33156 were expressed most predominantly in the RG (Fig 2A). However, other candidate genes did not show this RG-preferential expression pattern, even though knockdown of these genes caused developmental defects. In addition, whole mount in situ hybridization was conducted to confirm gene expression in the RG of wandering yw third instar larvae. Regarding alas and CG33156, we showed that antisense probe specific signals were detected in the RG (Fig 2B–2E), whereas no clear signals were detected with probes for the other six candidate genes (data not shown). alas and CG33156 showed typical expression profiles for genes responsible for ecdysone synthesis or release, and while the other six genes might be expressed in the PG, they may be less expressed or temporally restricted during development.\r\n\r\nExpression analysis of candidate genes.\r\n\r\n(A) Quantitative real-time PCR for candidate genes in several tissues from wandering third instar larvae of D. melanogaster. Rg, ring gland; Br, brain with ventral nerve cord; Sg, salivary gland; Tr, trachea; Fb, fat body; Gu, Gut; Mt, Malpighian tubule; Ep, epidermis. The expression level of each gene was normalized using rp49 as an endogenous control. Relative expression levels were determined using the ΔΔCt method. Among tissues examined, the highest average expression level was set to 1. (B-E) In situ hybridization analysis for candidate genes in the ring gland. alas and CG33156 transcripts were detected with antisense probes but not with sense probes. B: alas antisense, C: CG33156 antisense, D: alas sense, and E: CG33156 sense.\r\n\r\nEcdysteroid titers and RG morphology in RNAi larvae\r\n\r\nSince alas and CG33156 were predominantly expressed in the RG and disrupting their expression in the PG resulted in the developmental arrest at the larval stage, loss of gene function was expected to cause defects in ecdysteroid synthesis and/or release. To test this hypothesis, we measured ecdysteroid titers in the larvae using an LC-MS/MS quantification method established previously. In addition to ecdysone and 20E, the abundance of Makisterone A (MakA), which is another major ecdysteroid in Drosophila, was determined. In our experiments, all animals were third instar larvae at 96 h after egg laying (96 h AEL), while both wandering larvae and prepupae coexisted at 120 h AEL in the control group. Therefore, we examined ecdysone, 20E, and MakA titers at 96 h AEL feeding larvae, 120 h AEL in wandering larvae, and 120 h AEL in white prepupae from control groups. On the other hand, all larvae were still in the third instar at 96 h, 120 h, and 144 h AEL in the RNAi groups. Levels of 20E and MakA in the control groups were low during the feeding stage and increased with regular development. In contrast, these steroid titers in alas or CG33156 RNAi larvae remained at basal levels up to 144 h AEL (Fig 3). On the other hand, ecdysone was not detected in any samples examined; therefore, it is not clear that Alas and CG33156 are responsible for ecdysone synthesis or release. These results suggested that PG function is critically disrupted by interfering with alas or CG33156, and particularly, their larval arrest phenotype resulted from 20E and MakA deficiency during the third instar larval stage.\r\n\r\nEcdysteroids in D. melanogaster RNAi larvae.\r\n\r\nWhole body 20E and MakA titers of RNAi and control larvae were measured by LC-MS/MS. (A) 20E and MakA titers of alas RNAi animals. (B) 20E and MakA titers of CG33156 RNAi animals. Data are from three or four biological replicates, which consisted of 10 larvae or 5 pupae for each data point. W: wandering larvae and P: prepupae. Values are mean ± S.E.M. and asterisks indicate under quantitation limit (2.5 pg/larva or 5 pg/pupa).\r\n\r\nTo examine if the ecdysteroids depletion was due to the failure of RG organogenesis, the RGs of RNAi larvae were visualized by expressing mCD8::GFP under the control of phm-GAL4. RG morphology of alas or CG33156 RNAi larvae was normal as compared to controls at day 5 in the third instar larval stage (Fig 4A–4C). These RNAi larvae continued feeding after control animals became pupae, and they grew larger in the extended larval stage up to 10 days. In the prolonged larval stage, the RGs of RNAi larvae hypertrophied at day 8 with growth of their bodies (Fig 4D and 4E). In addition, the GFP signal in the RG appeared to be stronger probably due to additional expression of GFP during the extended larval period. These results suggested that these genes affect neither differentiation nor growth of the RG cells.\r\n\r\nRG morphology of D. melanogaster RNAi larvae.\r\n\r\nRG from day 5 or day 8 larvae were visualized by mCD8::GFP under the phm-GAL4 driver. (A) Representative image of the RG from day 5 control larva (phm \u003e mCD8::GFP). (B) and (D) Representative images of the RG from day 5 and day 8 alas RNAi larvae (phm \u003e mCD8::GFP, alas inverted repeat), respectively. (C) and (E) Representative images of the RG from day 5 and day 8 CG33156 RNAi larvae, respectively (phm \u003e mCD8::GFP, CG33156 inverted repeat). The morphology of the RG of RNAi larvae did not differ from that of control larvae at day 5, but the RG of RNAi larvae became larger during their extended larval period.\r\n\r\nDiscussion\r\n\r\nIn the present study, RNA-seq of target tissues unraveled many genes implicated in regulation of ecdysone synthesis and/or release as being expressed preferentially in the PG of B. mori. Most genes required for ecdysteroidogenesis are evolutionally conserved between Bombyx and Drosophila. To assess the function of the candidate genes obtained from our silkworm RNA-seq screening, we conducted Drosophila GAL4/UAS-based RNAi screening for fly orthologs of these candidate genes. Via this analysis, we observed that eight genes showed developmental defects when RNAi was driven by phm-GAL4 or 2-286-GAL4. Particularly, all RNAi lines of alas and CG33156 showed developmental defects. Real-time PCR and in situ hybridization confirmed that alas and CG33156 were predominantly expressed in the RG. Although the other six candidate genes, Spn6, sls, CG3106, CG34394, rab3, and CG8420, showed developmental defects upon RNAi, their expression was either not specific to the PG or very low, thus indicative that they may not be relevant to ecdysone synthesis and/or release.\r\n\r\ns\r\n\r\nRNAi animals of alas or CG33156 were still larvae and showed low ecdysteroid titers at 144 h AEL when normal animals have completed metamorphosis. These results indicated that both genes are responsible for proper ecdysteroid synthesis and/or release in the larvae. alas encodes δaminolevulinic acid synthase, which catalyzes the formation of aminolevulinic acid from succinyl-CoA and glycine and it is the first step of heme biosynthesis. CG33156 is a predicted gene with NAD kinase domain and expected to associate with NADP(H) production. Heme and NADP(H) are required for normal cell physiology, and disrupting the production of these molecules caused morphogenetic defects during embryogenesis in vertebrates. However, morphology of the RG in the RNAi larvae seemed to be normal at day 5 and then the RG enlarged and showed enhanced GFP expression at day 8. These results strongly suggested that these genes affected neither morphogenesis nor cell survival of the RG but rather the function of the PG as a steroidogenic organ. Some steroidogenic enzymes such as, Spo, Phm, Dib, and Sad, are members of the cytochrome P450 monooxygenase (CYP) family, which requires heme as a cofactor for electron transfer and NADPH as an electron donor for their enzymatic function. Therefore, alas and CG33156 might be essential factors for CYP function in ecdysteroid synthesis, whereas details remain to be seen. There are two different genes encoding ALAS isozyme in vertebrates. One gene (alas1) encodes an isozyme ubiquitously expressed and another (alas2) encodes an erythrocyte-specific isozyme. In contrast, both the fruitfly and silkworm have a single alas gene in respective genomes. In Drosophila, expression of alas has been reported in seven pairs of oenocytes and two unidentified cells in the anterior during embryogenesis. Furthermore, this study revealed that alas was highly expressed in the RG, with low expression was detected in other tissues in wandering larva. Such high expression of alas in the PG may help to meet the demand of CYPs during steroidogenesis. In addition, CG33156 was predominantly expressed in the RG. In humans, there are two types of NADK, cytosolic NADK1 and mitochondrial NADK2, and they are expressed in most human tissues. CG33156 showed high similarity to human NADK1 by BLASTP search (identity = 63.4%, E-value = 5.6e-133). In the Fly Base, there were two more homologs to human NADK (CG6145 to human NADK1, identity = 54.1%, E-value = 1.8e-120; CG8080 to human NADK2, identity = 36.4%, E-value = 2.6e-64). All of them have a conserved NAD kinase domain and may function as a NAD kinase in vivo. When CG33156 were solely knocked down in the RG, the RNAi animals could not make puparium, and neither CG6145 nor CG8080 could compensate for the lack of CG33156 function. Thus, CG33156 might have a pivotal role in ecdysone synthesis in the RG as a provider of NADP(H). Further precise functional analysis of Alas and CG33156 will provide much insight into the temporal dynamics of ecdysone synthesis and/or release.\r\n\r\nIn conclusion, our RNA-seq analysis of the Bombyx PG and phenotypic analysis of RNAi in Drosophila revealed two new players in the regulation of ecdysone synthesis and/or release. The genes analyzed in this study are just the tip of the iceberg, and the foundation we have formed may help to accelerate further elucidation of insect steroid synthesis and release. Particularly, future work purposed to construct the ecdysone transcriptional network in the PG and searching for target genes encoding effector molecules will shed light on the molecular basis of insect hormone production, release, and regulatory mechanisms.\r\n\r\nSupporting information\r\n\r\n\r\n","tracks":[]}