In many moths, females produce a sex pheromone to attract conspecific males. Sex pheromones thus play a very important role in pre-mating reproductive isolation among moth species. To date, sex pheromones of more than 1,500 moths have been chemically characterized. Moth pheromones are usually a blend of a few chemical components, and the species specificity is conferred by the combination of components and their blend ratios. Therefore, regulation of sex pheromone blend in females is crucial for the existence of a species. Despite the importance, the regulation of sex pheromone synthesis in female moths, the mechanism of blend ratio control in particular, is not well understood.

　The genus Ostrinia Hubner (Lepidoptera: Crambidae) worldwide includes 21 species, among which eight species inhabit Japan. The adzuki bean borer moth O. scapulalis, one of the eight species, uses a blend of (E)-11- and (Z)-11-tetradecenyl acetates (E11- and Z11-14:OAc) as the sex pheromone, whereas the Asian corn borer moth O. furnacalis, a congener found in Japan, uses a blend of the positional isomers, (E)-12- and (Z)-12-tetradecenyl acetates (E12- and Z12-14:OAc). In this thesis, I studied sex pheromone biosynthesis in O. scapulalis with special reference to the pheromone-gland-specific desaturase, one of the key enzymes that determine the carbon chain configuration of the pheromone components. In parallel, I also investigated the genetic regulation of pheromone biosynthesis by crossing O. scapulalis with O. furnacalis, which uses pheromone components distinctly different from those of O. scapulalis.

1. Fatty acid sex pheromone analogs (FAPA) in Ostrinia scapulalis

　Most of the lepidopteran sex pheromone components are derived from common fatty acids through chain-shortening, desaturation, reduction, and then acetylation (in case of acetates) or oxidation (in case of aldehydes). Unsaturated fatty acids that have the same carbon chain configuration as the sex pheromone components are often called as fatty acid pheromone analogs (FAPAs). Since FAPAs are found specifically in the pheromone gland, these compounds have been assumed to be the immediate precursors of the sex pheromone components. Although control of FAPA production is likely to underlie the control of pheromone production, the dynamics of FAPA biosynthesis remained to be studied.

　(E)-11- and (Z)-11-tetradecenoic acids are FAPAs of O. scapulalis, which are produced from myristic acid (C14) by the function of Δ11-desaturase. I measured the titers of FAPAs in the pheromone gland for three days after the emergence. The titers started to increase at the beginning of the scotophase, reaching a peak at the end of the same phase. FAPAs decreased continuously during the photophase, and started to increase again in the next scotophase. This diel pattern was repeated for three days. These findings suggest that biosynthesis of FAPAs is under photoperiodic control, as in the case of sex pheromone biosynthesis. Strong suppression of the increase of FAPA titers by decapitation suggested that signals from cephalic organs play an important role in the control of FAPA biosynthesis.

　O. scapulalis is known for the intraspecific pheromone blend polymorphism; E-type females produce a pheromone with a 99:1 blend of E11- and Z11-14:OAc, whereas Z-type females produce a pheromone with an opposite blend (E:Z = 3:97). I found that, regardless of the pheromone blend phenotypes, E/Z ratio in the FAPA geometric isomers was stably maintained at about 7:3. This finding suggests that E- and Z-type O. scapulalis use the same pathway for the production of FAPAs, and the blend ratio is regulated at a step in the downstream of FAPA biosynthesis, i.e., either reduction or acetylation.

2. cDNA cloning and in situ hybridization of Δ11-desaturase in O. scapulalis

　Sex pheromones are considered to be produced in a specialized tissue, the "pheromone gland" located in the terminal abdominal segments (8th-10th, TAS) of a moth; however, in many moth species the cells that produce pheromones have not actually been specified. I investigated cells in the TAS that synthesize pheromones in the adzuki bean borer Ostrinia scapulalis, by locating pheromones and their precursors, and mRNA for Δ11-desaturase, a key enzyme in pheromone biosynthesis. I demonstrated that the pheromone components, E11- and Z11-14:OAc, and their fatty acyl precursors were specifically contained in the dorsal part of the TAS. From the TAS of E-type O. scapulalis females, I then cloned a 1217-base cDNA of Δ11-desaturase gene (OscZ/E11, AB232855) that included a complete open reading frame encoding 329 amino acid residues. The deduced amino acid sequence has three histidine boxes which are highly conserved within acyl-CoA desaturases in animals and yeast. RT-PCR and in situ hybridization unequivocally showed that OscaZ/E11 is specifically expressed in the modified epidermal cells located at the dorsal end of the 8th-9th intersegmental membrane.

3. The same set of desaturases are expressed in the pheromone glands of O. scapulalis and O. furnacalis

　While Δ11-desaturase is involved in the pheromone biosynthesis in many moth species including O. scapulalis, involvement of Δ14-desaturase is only known in O. furnacalis. A cDNA of Δ14-desaturase gene (OfuZ/E14) has been cloned from O. furnacalis. Interestingly, although Δ11-desaturase is not utilized in O. furnacalis, transcript of Δ11-desaturase gene has been found in the pheromone gland of this species. Analysis of the cDNA (OfuZ/E11) has shown a very high homology to OscZ/E11. In this chapter, I first examined whether or not Δ14-desaturase gene is expressed in O. scapulalis, which do not use this enzyme for the pheromone production. I found that Δ14-desaturase gene is expressed in O. scapulalis, and I was able to clone a cDNA of this gene from this species (OscZ/E14), which showed a very high homology to OfuZ/E14. Analyses of sex pheromone gland extracts by gas chromatography - mass spectrometry showed that (E)-11- and (Z)-11-tetradecenoic acids, the expected products of OfuZ/E11, were not detected in the pheromone gland of O. furnacalis. Likewise, any traces of E14-16:Acid, Z14-16:Acid, E12-14:Acid and Z12-14:Acid, the expected products of OscZ/E14 and their derivatives, were not detected in O. scapulalis. These findings suggest that functional OfuZ/E11 protein is not present in the pheromone gland of O. furnacalis, and functional OscZ/E14 protein is not present in O. scapulalis. Non existence of OfuZ/E11 in O. furnacalis was finally demonstrated by immunocytochemical analysis of the pheromone gland.

4. Alternative use of two pheromone-gland-specific desaturases in O. scapulalis and O. furnacalis

　To gain an insight into the regulation of sex pheromone biosynthesis in O. scapulalis, which uses E11- and Z11-14:OAc as the pheromone components, and O. furnacalis, which uses E12- and Z12-14:OAc, I crossed these two species and investigated the female sex pheromone composition in F1, F2 and backcross progenies. Surprisingly, all F1 hybrid females predominantly produced saturated tetradecyl acetate, with a small amount of the parental components. Tetradecyl acetate has not been regarded as the sex pheromone component of any Ostrinia species examined to date. Analyses of the pheromone gland components in the backcross and F2 progenies have shown that the pheromone phenotypes were divided into O. scapulalis-like, O. furnacalis-like and F1-like. Based on the frequencies of the phenotypes in the backcross and F2 progenies, the variation in the phenotype was found to be explainable by the existence of a major autosomal locus with two alleles.

In the pheromone gland of F1 females, the expression level of Δ11-desaturase gene was comparable to that in O. scapulalis, a parent that uses Δ11-desaturase for pheromone biosynthesis, and the level of Δ14-desaturase gene transcript was comparable to that of O. furnacalis, the other parent that uses Δ14-desaturas. Nevertheless, the titters of FAPAs, the products of these enzymes, were remarkably low. Taking all findings in the previous and present chapters together, it is inferred that O. scapulalis has a factor that inhibits production of Δ14-desaturase protein, and that O. furnacalis reciprocally has a factor that inhibits production of Δ11-desaturase protein.

In this thesis, I studied sex pheromone biosynthesis in O. scapulalis in comparison with that in a close relative O. furnacalis, with special reference to the pheromone-gland-specific desaturases. I clarified that two species have the same set of desaturase genes (Δ11- and Δ14-), but only one of them is functionally expressed in each species, resulting in a production of completely different set of pheromone components.

　ガ類の雌性フェロモンのほとんどは複数成分のブレンドである．ブレンドの成分比が種特異性を決めるため，生合成における成分比の制御は種にとってきわめて重要である．本論文はアワノメイガ種群の近縁な2種，アズキノメイガOstrinia scapulalis (sca;E/Z11-14:OAcを性フェロモンとする)とアワノメイガO. furnacalis (fur;E/Z12-14:OAcを性フェロモンとする)を用い，アワノメイガ種群における性フェロモン生合成の制御機構の一端を解明しようとするものである．

1.　アズキノメイガのフェロモン腺における性フェロモン類似脂肪酸(FAPA)の変化

　性フェロモンの前駆体で性フェロモンと同じ炭素骨格を持つ脂肪酸(FAPA)の生合成ダイナミクスを調べた．scaのFAPAは14:acylにΔ11位不飽和化酵素が作用して合成されるE/Z11-14:Acylである．scaにはフェロモン成分のE/Z比に多型があり，Eタイプ(E:Z=99:1)とZタイプ(E:Z=3:97)が知られる．しかるにFAPAのE/Z比はE，Z両タイプともにE:Z=7:3であり，性フェロモン成分比とは大きく異なっていた．これはフェロモン成分比がFAPA合成より後の段階で制御を受けることを示す興味深い結果である．

2.　アズキノメイガのフェロモン腺におけるΔ11位不飽和化酵素のクローニングとin situ hybridization

　ガ類雌の性フェロモン合成は雌成虫のフェロモン腺で特異的に行なわれるとの仮説を巧妙な手法によって実証した．すなわち性フェロモン生合成の鍵酵素Δ11位不飽和化酵素に注目し，scaのΔ11位不飽和化酵素遺伝子(OscZ/E11)をクローニングし，RT-PCRおよびin situ hybridizationによってOscZ/E11がフェロモン腺細胞のみで発現していることを確認した．これはΔ11位不飽和化がフェロモン腺細胞のみで生じていることを示す初めての結果である．

3.　アズキノメイガとアワノメイガのフェロモン腺における2種の不飽和化酵素の発現

　furでは性フェロモン(E/Z12-14:OAc)合成にΔ14位不飽和化酵素OfuZ/E14が機能するが，性フェロモン生合成に関与しないΔ11位不飽和化酵素遺伝子OfuZ/E11の発現も確認されている．本研究では，scaからΔ14位不飽和化酵素遺伝子OscZ/E14をクローニングした．しかし，GC-MS分析により，scaではOscZ/E11の反応産物のみが，furではOfuZ/E14の反応産物のみが検出された．さらに，免疫組織化学的実験によってフェロモン腺で発現するΔ11位不飽和化酵素遺伝子はscaでは翻訳されるのに対し，furでは翻訳されないことを示した．

4.　アワノメイガとアズキノメイガの交雑個体における性フェロモン生合成

　実験室内で両種を交雑して雑種第1代(F1)を得た．furとscaの性フェロモン成分は互いに位置異性体であるが，F1では驚くべきことに性フェロモン腺主成分は飽和の14:OAcであり，両親の性フェロモン成分がごくわずかに存在した．また，戻し交雑では親型とF1型が1:1に分離した．これらから，親型，F1型のフェロモン成分の制御は常染色体上に存在する1遺伝子もしくは非常に近接した遺伝子座に存在する遺伝子群によることが示された．F1のフェロモン腺では，Δ11位およびΔ14位不飽和化酵素遺伝子が両親の「活性型」不飽和化酵素遺伝子と同程度の発現量を示したが，これら酵素の反応産物であるFAPAの量は親と比べて少なかった．以上の結果を総合すると，scaではΔ14位不飽和化酵素遺伝子の，furではΔ11位不飽和化酵素遺伝子の翻訳阻害因子が存在すると考えられ，F1は両方の阻害因子を持つために14:OAcが主成分になると考えられた．

　本研究では，性フェロモン生合成に特異的な不飽和化酵素に関する研究を行ない，巧妙な手法を駆使することにより，性フェロモン成分比多型がFAPA合成後に制御されること，フェロモン腺では遺伝的制御により不飽和化酵素が選択的に利用されていること，近縁種間の性フェロモン成分の違いが少数の遺伝子の変異によって生じる可能性を示した．審査委員一同はこれらの成果が学術的にも応用的にも大いに貢献しうるものであり，博士(農学)の学位を授与するに十分な価値を有することを認めた．