Isolation and characterization of a humoral factor that stimulates transcription of the acyl-CoA-binding protein in the pheromone gland of the silkmoth, Bombyx mori.

Acyl-CoA binding protein (ACBP) is a highly conserved 10-kDa intracellular lipid-binding protein that binds straight-chain (C14-C22) acyl-CoA esters with high affinity and is expressed in a wide variety of species ranging from yeast to mammals. Functionally, ACBP can act as an acyl-CoA carrier or as an acyl-CoA pool maker within the cell. Much work on the biochemical properties regarding the ACBP has been performed using various vertebrate and plant tissues, as well as different types of cells in culture, the regulatory mechanisms underlying ACBP gene expression have remained poorly understood. By exploiting the unique sex pheromone production system in the moth pheromone gland (PG), we report that transcription of a specific ACBP termed pheromone gland ACBP is triggered by a hemolymph-based humoral factor. Following purification and structure elucidation by means of high resolution electrospray ionization mass spectrometry and NMR analyses, in conjunction with stereochemical analyses using acid hydrolysates, the humoral factor was identified to be beta-D-glucosyl-O-L-tyrosine. Examination of the hemolymph titers during development revealed that the amount of beta-D-glucosyl-O-L-tyrosine dramatically rose prior to eclosion and reached a maximum of 5 mg/ml (about 1 mg/pupa) on the day preceding eclosion, which was consistent with the effective dose of beta-D-glucosyl-O-L-tyrosine in stimulating pheromone gland ACBP transcription in vivo. Furthermore, in vitro assays using trimmed PG indicated that beta-D-glucosyl-O-L-tyrosine acts directly on the PG. These results provide the first evidence that transcription of some ACBPs can be triggered by specific humoral factors.

Acyl-CoA-binding protein (ACBP) 1 is a 10-kDa intracellular lipid-binding protein that specifically binds medium-and longchain fatty acyl-CoA esters with high affinity and is structur-ally highly conserved from yeast to mammals as well as insects (1). Although the in vivo function of this protein has yet to be clarified in detail, in vitro investigations have revealed that ACBP protects acyl-CoA esters from hydrolysis (2)(3)(4); consequently, it can function as both an acceptor and a donor of acyl-CoA esters (3) by regulating their availability for various metabolic purposes such as mitochondrial ␤-oxidation (4 -6), microsomal glycerolipid synthesis (4), and phospholipid synthesis (7,8). ACBP can also generate an intracellular acyl-CoA pool that can serve a number of biochemical purposes (9,10), because overexpression of ACBP in yeast cells results in a substantial increase in cellular acyl-CoA content (11).
In mammals, whereas the ACBP gene is a typical housekeeping gene (12), its expression differs markedly among different cell types; a high level of ACBP expression occurs in hepatocytes, steroidogenic cells, and adipocytes (reviewed in Ref. 13). The expression of ACBP is partly correlated with increased lipogenesis and occurs, for instance, in 3T3-L1 preadipocytes during in vitro differentiation, a process that is accompanied by a marked accumulation of triacylglycerol (TG) and de novo fatty acid synthesis (14,15). Although the molecular mechanisms regulating ACBP expression have remained largely unknown, two functional regulatory elements, a sterol regulatory element-binding protein (SREBP)-binding site and a nuclear factor Y (NFY)-binding site, have been identified in the proximal promoter of the human ACBP gene (16). Recently, it has also been demonstrated that ACBP is a novel peroxisome proliferator-activated receptor (PPAR) target gene and that the PPAR-response element in intron 1 of the rat ACBP gene is a bona fide PPAR-response element (17).
In insects, many species of female moths produce speciesspecific sex pheromones to attract conspecific male moths. Their pheromone components are generally synthesized de novo through long-chain fatty acyl intermediates in the abdominal pheromone gland (PG), a functionally differentiated organ responsible for sex pheromone production. In the course of examining the molecular mechanisms underlying sex pheromone production in the silkmoth, Bombyx mori, we found that two distinct ACBPs termed pheromone gland ACBP (pgACBP) and midgut ACBP (mgACBP) are specifically expressed in the PG cells during pheromonogenesis (18). Northern blot analyses using various tissues also revealed that mgACBP, but not pgACBP, is highly expressed in the midgut during larval feeding stages, suggesting that there are at least two distinct ACBPs with different physiological functions in the moth and that both ACBPs simultaneously participate in sex pheromone production (18). In B. mori, the PG cells accumulate a large number of lipid droplets within the cytoplasm just before eclosion (19). The lipid droplets contain TGs composed of unsaturated C 16 and C 18 fatty acids in addition to the pheromone (bombykol, (E,Z)-10,12-hexadecadien-1-ol) precursor, ⌬10,12-hexadecadienoic acid, which is present as a major component. These droplets play a significant role in storing the bombykol precursor and eventually releasing it for pheromone production at eclosion in response to the neurohormone designated as pheromone biosynthesis-activating neuropeptide (PBAN) (19 -21). Because the PG cells must retain large amounts of pheromone precursors during pheromonogenesis, one of the functions of the ACBPs expressed in the PG seems to be to donate acyl-CoAs for the synthesis of the TGs in the lipid droplets.
Apart from these findings, we found that both of the ACBPs are expressed simultaneously in the PG and are up-regulated on the day preceding eclosion (18). Based on these observations, we hypothesized that there might be some kinds of physiological cues that trigger transcription and thus regulate the expression of the ACBPs in the PG. In this report, we describe the presence, isolation, and characterization of a humoral factor that appears in the hemolymph just before eclosion and that stimulates pgACBP transcription. Present results provide the first evidence that transcription of some ACBPs can be triggered by specific humoral factors.

EXPERIMENTAL PROCEDURES
Insects-B. mori eggs (Shuko ϫ Ryuhaku) were purchased from Katakura Kogyo (Matsumoto, Japan), and larvae were raised on an artificial diet as described (22). Bioassays were performed using 5-10 female pupae 3 days before eclosion (day Ϫ3). Pupal age was determined based on the morphological characteristics as described (21).
Plasma Preparation-B. mori hemolymph was collected into a chilled microcentrifuge tube containing 0.05% phenylthiourea and 0.5 mM p-APMSF in 0.2 M phosphate buffer (pH 6.0) using the "flushing out" method by injecting 0.1 M phosphate buffer containing 150 mM KCl and 10 mM EDTA (pH 6.0) into the hemocoel (23). For purification, the hemolymph collected from female pupae (day Ϫ1) was immediately centrifuged at 4°C for 10 min at 500 ϫ g to remove hemocytes, and the resulting supernatant (plasma) was mixed with an equal volume of Ϫ20°C acetone. Following centrifugation at 4°C for 10 min at 20,000 ϫ g, the supernatant was passed through an Amicon Ultra PL-5 membrane (Millipore). The pass-through fraction was concentrated with the SpeedVac System (Thermo Savant) and designated as the plasma fraction.
Isolation Procedures-The plasma fraction prepared from 500 female pupae (day Ϫ1) was loaded onto a TSK-GEL Amide-80 column (4.6 ϫ 250 mm; Tosoh Co.) and eluted isocratically with 70% CH 3 CN at a flow rate of 1.0 ml/min. The bioactive fraction that eluted at 12-13 min was then loaded onto a CAPCELL PAK C 18 column (4.6 ϫ 150 mm; pore size, 5 m; Shiseido Co.) and eluted with 5% CH 3 CN in 0.1% CF 3 COOH/ H 2 O at a flow rate of 0.5 ml/min.
Mass Spectrometry-High resolution electrospray-ionization mass spectrometry (HR-ESIMS) was carried out in the positive ion mode using a JEOL JMS-T100LC spectrometer with methanol as the mobile phase at a flow rate of 0.2 ml/min (internal standard: sodium trifluoroacetate). The purified humoral factor and synthetic ␤-D-glucosyl-O-Ltyrosine were dissolved in water at a concentration of 1 g/ml, and 10-l aliquots were subjected to analysis.
NMR Analysis-The one-dimensional 1 H at 600 MHz and 13 C NMR spectra at 150 MHz, and two-dimensional DQF-COSY, HMQC, and HMBC NMR spectra at 25°C were recorded with a JEOL JNM-ECA600 spectrometer equipped with a Nalorac 3-mm gradient inverse double resonance probe head. The purified humoral factor (100 g) (as a trifluoroacetate salt) was dissolved in 350 l of D 2 O with 0.75% 3-(trimethylsilyl) propionic-2,2,3,3-d 4 acid, sodium salt as the internal reference. 1  Hydrolysis and Stereochemistry Analysis-The purified humoral factor (100 g) was hydrolyzed in a vacuum-sealed tube with 6 N HCl at 110°C for 20 h. DL-Amino acid analysis of the resulting hydrolysates was carried out according to Hayashi and Sasagawa (26). The purified humoral factor (1 mg) was also hydrolyzed with 1 N HCl at 100°C for 12 h. The resulting hydrolysates were loaded onto a TSK-GEL Amide-80 column and eluted with 70% CH 3 CN at a flow rate of 1.0 ml/min. The glucose fraction, which eluted at 7-8 min, was evaporated, redissolved in 1.5 ml of water, and then subjected to measurement of its optical rotation with a polarimeter DIP-370 (Jasco, Japan) at a wavelength of 589 nm in a cell with a path length of 10 cm at a precision of 0.001°.
In Vitro Assay-Trimmed PGs were placed in individual wells of a 48-well plate and incubated in the presence or absence of ␤-D-glucosyl-O-L-tyrosine at 25°C in 200 l of insect Ringer's solution. After incubation, five trimmed PGs were combined and homogenized on ice in 100 l of 0.2 M phosphate buffer (pH 6.0) containing 0.05% phenylthiourea and 0.5 mM p-APMSF with a glass homogenizer. The homogenate was centrifuged at 1,000 ϫ g at 4°C for 10 min. The resulting supernatant was further centrifuged at 10,000 ϫ g at 4°C for 60 min to prepare the cytosolic fraction, which was then used to quantify the amount of ␤-D-glucosyl-O-L-tyrosine present by means of HPLC using a TSK-GEL Amide-80 column as described above. The pgACBP transcript levels following incubation were also measured by RT-PCR as described above.
Fluctuation of the ␤-D-Glucosyl-O-L-Tyrosine Content in the Hemolymph-To measure the fluctuation of the ␤-D-glucosyl-O-L-tyrosine titers in the hemolymph, plasma fractions were prepared daily from fifth instar larvae, and the amount of ␤-D-glucosyl-O-L-tyrosine was quantified by means of HPLC using a TSK-GEL Amide-80 column as described above. ␤-D-Glucosyl-O-L-tyrosine measurements were performed every 12 h from sample sets that spanned day Ϫ4 pupae to day 1 adults.
Measurement of the Total Volume of the Hemolymph in the Pupae-Aliquots (10 l) of Evans Blue (WAKO chemicals) solution (5 mg/ml) were injected into 10 female pupae. After maintaining for 15 min at 25°C, the hemolymph (20 l each) was collected into a chilled microcentrifuge tube and diluted 50-fold with H 2 O, and the absorbance at 610 nm was measured.

RESULTS
Presence of a Humoral Factor That Triggers pgACBP Transcription in the PG-In our previous paper, Northern blot analyses revealed that the pgACBP transcript is specific to the PG, where it undergoes significant up-regulation 1 day prior to adult emergence (18). This up-regulation within the PG was again confirmed in the present RT-PCR analyses using PGs prepared 3 days and 1 day before eclosion (day Ϫ3 and day Ϫ1) (Fig. 1A, lanes 1 and 2). To examine whether humoral factor(s) present in the hemolymph could stimulate pgACBP transcription, we injected various B. mori plasma samples into day Ϫ3 female pupae (Fig. 1A, lanes 3-8). The results indicated that injection of the female plasma taken 1 day prior to eclosion (lane 6) or immediately after eclosion (lane 7) caused a marked increase of the pgACBP transcript at 18 h postinjection. Although this activity in the plasma seemed to gradually increase during pupal development (lanes 4 -6), no activity was detected in the plasma taken from either fifth instar larvae or female moths 1 day after eclosion (lanes 3 and 8). Since pgACBP transcripts were detectable as early as 12 h postinjection (Fig.  1B), we elected to monitor the progress of our purification of the plasma humoral factor by assaying pgACBP transcript levels 12 h after injection of the test samples.
Isolation of the Humoral Factor in the Female Pupal Hemolymph-Preliminary experiments revealed that the humoral factor in the pupal plasma (day Ϫ1) could be recovered in the fraction that not only failed to be precipitated with 50% acetone but that also passed through an Amicon Ultra PL-5 membrane (Millipore), suggesting that the humoral factor was a low molecular weight substance. For purification, we prepared this plasma fraction, which was equivalent to 500 female pupae (day Ϫ1), and performed isocratic HPLC separation using a TSK-GEL Amide-80 column ( Fig. 2A). The bioactive fraction that eluted from the column at 12-13 min was further purified using a C 18 column with the activity recoverable in a single peak that eluted at 5.5 min (Fig. 2B).  13 C NMR spectra of the purified humoral factor showed the presence of a ␤-glucosyl unit and a tyrosine-like amino acid unit (Fig. 3), results that were supported by two-dimensional DQF-COSY, HMQC, and HMBC NMR experiments (data not shown). Low field chemical shifts of an anomeric proton at 5.12 (d, 7.6 Hz) ppm suggested that the phenolic hydroxyl group of the tyrosine unit was glucosylated. This glucosylation was confirmed by a long range correlation between the anomeric proton and a quaternary carbon at 159.0 ppm. These results indicate that the purified factor is ␤-glucosyl-O-tyrosine. Further analyses, both amino acid analysis and acid hydrolysis, of the purified factor indicated that the factor is composed of L-tyrosine and D-glucose ([␣] 589 ϭ ϩ46.7 (c ϭ 0.35, H 2 O)). Taken together, these results demonstrate that the humoral factor in the plasma is ␤-D-glucosyl-O-L-tyrosine.

Structure Elucidation of the Humoral Factor-HR-ESIMS
Biological Activity of Synthetic ␤-D-Glucosyl-O-L-Tyrosine-To confirm the biological activity of ␤-D-glucosyl-O-L-ty-rosine, we synthesized ␤-D-glucosyl-O-L-tyrosine and examined its ability to stimulate transcription of the ACBPs in the PG. When we injected various concentrations of synthetic ␤-D-glucosyl-O-L-tyrosine into day Ϫ3 female pupae, the levels of pgACBP mRNA increased in a dose-dependent manner; the transcript was detected at doses of 0.5, 1.0, and 2.0 mg/pupa (Fig. 4A). These biological activities are consistent with those of the purified humoral factor (data not shown). At these concentrations, more than 1.0 mg/pupa seemed to be required to obtain the comparable level of pgACBP transcript of the untreated day Ϫ1 pupa (Fig. 1A, lane 2). Because injection of L-tyrosine and/or D-glucose alone failed to elicit a transcriptional response, even at concentrations of 2.0 mg/pupa, the covalent bond between Ltyrosine and D-glucose is indispensable for biological activity (data not shown). In contrast to the elevation of the pgACBP transcript levels, ␤-D-glucosyl-O-L-tyrosine essentially had no effect on the transcription of mgACBP (Fig. 4B).
To examine whether ␤-D-glucosyl-O-L-tyrosine acts directly on the PG cells to stimulate pgACBP transcription, we prepared trimmed PGs from female pupae (day Ϫ3) and incubated each PG for 12 h in the presence of ␤-D-glucosyl-O-L-tyrosine at concentrations ranging from 0 -5 mg/ml. Similar to the in vivo assays described above, RT-PCR revealed that the levels of pgACBP mRNA in the trimmed PG increased in a dose-dependent manner (Fig. 5A). These results demonstrate that the humoral factor ␤-D-glucosyl-O-L-tyrosine acts directly on the PG to stimulate pgACBP transcription. In addition, the titer of ␤-D-glucosyl-O-L-tyrosine within the cytosolic fraction of PGs increased proportionately with the concentration of ␤-D-glucosyl-O-L-tyrosine used in the incubations (Fig. 5B). At a concentration of 5.0 mg/ml, the titer in the cytosolic fraction reached a maximum (0.6 g/PG) at 6 h (Fig. 5B), whereas the pgACBP transcript appeared at 12 h (Fig. 5C). lymph during development (i.e. from the larval fifth instar to the adult stage) (Fig. 6). Although there was no significant difference in the patterns between female and male, two prominent peaks in the ␤-D-glucosyl-O-L-tyrosine titer appeared during the molting stages; the smaller of the two surges occurred at the larval-pupal molt and reached a maximum level (1 mg/ml) on the day of pupation, whereas the larger pulse occurred during the pupal-adult molt. This second surge began 3-4 days prior to eclosion, reached a maximum of as much as 5 mg/ml at 1 day before eclosion, and then rapidly declined to the lowest level (0.2 mg/ml) within 2 days of emergence. When we measured the total volume of the hemolymph in female pupae (Table I), it was estimated that a single female pupa (day Ϫ2) contains, on average, ϳ190 l of hemolymph. This estimation implies that the effective dose of ␤-D-glucosyl-O-L-tyrosine (more than 0.5 mg/pupa) is comparable with the ␤-D-glucosyl-O-L-tyrosine titers around 1 day before eclosion (0.95 mg/pupa). These results therefore indicate that the upsurge of ␤-D-glucosyl-O-L-tyrosine in the hemolymph just before eclosion is indeed responsible for the up-regulation in pgACBP transcription during pheromonogenesis. DISCUSSION ACBP is a highly conserved 10-kDa N-acetylated polypeptide that is expressed in a wide variety of species ranging from yeast to mammals (29). Because ACBP binds straight-chain (C 14 -C 22 ) acyl-CoA esters with high affinity and thereby protects them from hydrolysis (2-4), functionally, ACBP can act as an acyl-CoA carrier or an acyl-CoA pool maker within the cell to increase total cellular acyl-CoA content (9 -11). In previous studies, we have demonstrated that two distinct ACBPs are specifically expressed in the PG during pheromonogenesis and undergo up-regulation on the day prior to eclosion in the female B. mori moth (18). In the PGs of many moth species, various long chain fatty acyl-CoAs participate as precursors or intermediates in the biosynthesis of species-specific sex pheromones (30), suggesting that ACBPs expressed in the PG could possibly function as carriers or cellular deposits for the acyl-CoAs utilized in pheromone biosynthesis.
Whereas numerous studies regarding the biochemical properties of the ACBP have been performed using various vertebrate and plant tissues as well as different types of cells in culture, the regulatory mechanisms underlying ACBP gene expression have remained poorly understood; only the involvement of transcription factors such as SREBP and PPAR␥ has been reported (16,17). In the present study, by exploiting the unique sex pheromone production system in the moth PG, we have provided the first evidence that transcription of a specific ACBP is triggered by a hemolymph-based humoral factor. To initiate the present experiments, we hypothesized that there would be a specific physiological cue that triggered ACBP transcription within the PG cells 1 day prior to eclosion. Because insects contain an open circulatory system, the PG is continuously exposed to the hemolymph; consequently, we examined whether the hemolymph 1 day prior to eclosion could stimulate pgACBP transcription and found that a transcriptional activator was indeed present (Fig. 1A). Following purification and structure elucidation by means of HR-ESIMS and NMR analyses, in conjunction with stereochemical analyses using hydrolysates, the active factor in the hemolymph was identified as ␤-D-glucosyl-O-L-tyrosine (Figs. 2 and 3). The validity of the structural determination was confirmed by comparing the purified factor with synthetic ␤-D-glucosyl-O-L-tyrosine.
The in vivo assays using day Ϫ3 pupae revealed that ␤-Dglucosyl-O-L-tyrosine does stimulate pgACBP transcription but that it requires as much as 1.0 mg/pupa to obtain transcript   levels comparable with that observed in the PGs of untreated day Ϫ1 pupa (Fig. 4). Quite interestingly, we found that the ␤-D-glucosyl-O-L-tyrosine titer in the hemolymph dramatically rises prior to eclosion and reaches a maximum of 5 mg/ml (about 1 mg/pupa) on the day preceding eclosion (Fig. 6). This relationship between the effective dose of ␤-D-glucosyl-O-L-tyrosine and its upsurge in the hemolymph just before eclosion is consistent with its proposed role in the up-regulation of pgACBP during pheromonogenesis. Furthermore, the in vitro assays using trimmed PG (Fig. 5) indicate that ␤-D-glucosyl-O-L-tyrosine acts directly on the PG to stimulate pgACBP transcription, although the molecular mechanism behind how the ␤-D-glucosyl-O-L-tyrosine signal stimulates the transcription of pgACBP within the PG cells remains to be clarified. In mammals, two functional regulatory elements, a SREBP-binding site and an NFY-binding site, have been identified in the proximal promoter of the human ACBP gene (16). In addition, it has been reported that the ACBP gene is a novel PPAR␥:retinoid X receptor target gene and that PPAR␥:retinoid X receptor activates transcription through an intronic PPAR-response element in both human and rodent ACBP (17). In the present experiments, we found that there is a time lag of several hours between the incorporation of ␤-D-glucosyl-O-L-tyrosine into the PG cells and the appearance of the pgACBP transcript (Fig. 5, B and C). This finding suggests that transcription of pgACBP does not result from a direct interaction with ␤-D-glucosyl-O-Ltyrosine itself. We also found that the titer of ␤-D-glucosyl-O-L-tyrosine in the hemolymph rapidly declines following adult emergence (Fig. 6), whereas we had previously reported that the pgACBP transcript level remained high for several days after eclosion (18). These results suggest that ␤-D-glucosyl-O-L-tyrosine stimulates pgACBP transcription through a signal transduction cascade that probably includes some as yet unidentified transcription factor similar to SREBP, NFY, and/or PPAR␥. The presence of ␤-D-glucosyl-O-L-tyrosine in the hemolymph has been reported in several insects (31)(32)(33)(34). In addition, it has been demonstrated in the tobacco hornworm, Manduca sexta, that ␤-D-glucosyl-O-L-tyrosine is synthesized in the fat body by the action of a specific ␤-glucosyltransferase present there (35). In M. sexta, the regulation of ␤-D-glucosyl-O-L-tyrosine synthesis and hydrolysis is under the control of hormones that regulate molting and metamorphosis; the decline in juvenile hormone titer after the last larval ecdysis initiates ␤-D-glucosyl-O-L-tyrosine synthesis (36), whereas the major pulse of 20hydroxyecdysone triggers hydrolysis of ␤-D-glucosyl-O-Ltyrosine by a ␤-glucosidase in the fat body (33). In insects, L-tyrosine is an important precursor for the diphenols and quinones that cross-link cuticular proteins during tanning or sclerotization, the process of which results in the hardening and pigmentation of the new cuticle following the molt (33). Consequently, the function of ␤-D-glucosyl-O-L-tyrosine has long been suggested to serve as a reservoir of L-tyrosine for the eventual incorporation into the newly formed cuticle following the molt and/or metamorphosis. This idea seems to be consistent with the present finding that an increase in the ␤-D-glucosyl-O-L-tyrosine titer occurs, regardless of sex, at the time of pupation and eclosion (Fig. 6). Alternatively, since we have unequivocally demonstrated that ␤-D-glucosyl-O-L-tyrosine triggers pgACBP transcription during pupal-adult metamorphosis in the female moth, our present results also suggest a multifunctional role for ␤-D-glucosyl-O-L-tyrosine.
We found that ␤-D-glucosyl-O-L-tyrosine essentially failed to stimulate mgACBP transcription (Fig. 4B). Since we have detected mgACBP transcriptional activation capabilities in a hemolymph fraction other than the ␤-D-glucosyl-O-L-tyrosine fraction, it is likely that up-regulation of this protein in the PG is under the control of another humoral factor.