Glucose oxidase prevents programmed cell death of the silkworm anterior silk gland through hydrogen peroxide production Hiroto Matsui 1 , Motonori Kakei 2 , Masafumi Iwami 1,2 and Sho Sakurai 1,2 1 Division of Biological Sciences, Graduate School of Natural Science and Technology, Kanazawa University, Japan 2 Division of Life Sciences, Graduate School of Natural Science and Technology, Kanazawa University, Japan Introduction Programmed cell death (PCD) plays an important role in the elimination of cells and tissues as part of the progression of animal development. In insects, larval- specific tissues degenerate during metamorphosis in response to 20-hydroxyecdysone (20E), an active form of ecdysteroid. The induction of PCD by 20E has been extensively studied in the salivary glands of Drosoph- ila melanogaster [1], the intersegmental muscles [2] and prothoracic glands of Manduca sexta [3], and the ante- rior silk glands (ASGs) of Bombyx mori [4,5]. The B. mori ASG is a tubular organ consisting of a single layer of cells, and acts as a spinning apparatus for secreting silk thread, which is composed of two dif- ferent proteins produced in the middle and posterior silk glands. The ASG is eliminated at the end of the larval stage through PCD in response to the meta- morphic rise in the 20E concentration [4]. The ASG becomes competent to respond to 20E undergoing PCD late on day 5 of the fifth (last) instar stage, and exhibits full competence after the onset of spinning on day 6, but ASGs before the middle of day 5 do not respond to 20E [6]. Lepidopteran larval epidermis and wing disks lose their sensitivity to juvenile hormone after the change Keywords glucose oxidase; hydrogen peroxide; insect; programmed cell death; silk gland Correspondence S. Sakurai, Division of Life Sciences, Graduate School of Natural Science and Technology, Kanazawa University, Kanazawa 920-1192, Japan Fax: +81 76 264 6215 Tel: +81 76 264 6250 E-mail: ssakurai@kenroku.kanazawa-u.ac.jp (Received 15 November 2010, revised 16 December 2010, accepted 21 December 2010) doi:10.1111/j.1742-4658.2010.07996.x During pupal metamorphosis, the anterior silk glands (ASGs) of the silk- worm Bombyx mori degenerate through programmed cell death (PCD), which is triggered by 20-hydroxyecdysone (20E). 20E triggers the PCD of the ASGs of day 7 fifth instar (V7) larvae but not that of V5 larvae. When V7 ASGs were cocultured with V5 ASGs in the presence of 20E, neither culture of ASGs underwent PCD. The 20E-induced PCD of V7 ASGs was also inhibited when they were incubated in conditioned medium that was prepared by incubating V5 ASGs for 48 h, an indication that V5 ASGs released an inhibitor of 20E-induced PCD during incubation. The inhibitor was purified from conditioned medium and identified as glucose oxidase (GOD). GOD catalyzes the oxidation of glucose to gluconolactone, and generates hydrogen peroxide as a byproduct. We found that hydrogen per- oxide is the molecule that directly inhibits the action of 20E and may act to protect the ASGs from early execution of PCD during the feeding stage. GOD was localized in the inner cavity of the gland, and was discharged to the outside of the ASGs with the silk thread at the onset of spinning. Thus, the spinning behavior, occurring at the beginning of the prepupal period, plays an important role in controlling the time at which ASGs undergo PCD in response to 20E. Abbreviations ASG, anterior silk gland; BHR3, Bombyx hormone receptor 3; BR-C Z1, broad complex Z1; CM, conditioned medium; DHR3, Drosophila hormone receptor 3; GOD, glucose oxidase; NOX, NADPH oxidase; PCD, programmed cell death; RpL3, ribosomal protein L3; 20E, 20-hydroxyecdysone. 776 FEBS Journal 278 (2011) 776–785 ª 2011 The Authors Journal compilation ª 2011 FEBS in commitment that occurs in the fifth instar, and the pupally committed tissues are capable of responding to 20E with pupal differentiation. Whereas the ASGs lose their sensitivity to juvenile hormone on approximately day 4 of the fifth instar [6], they do not exhibit respon- siveness to 20E until late on day 6. This observation led to the hypothesis that an unknown factor(s) may be involved in the control of PCD execution in addi- tion to 20E and juvenile hormone. When we cocultured ASGs from day 5 fifth instar (V5) larvae and V7 larvae with 20E, we found that neither group of ASGs underwent PCD. This result indicates that V5 ASGs inhibit the 20E-induced PCD of V7 ASGs. In addition, when V7 ASGs were cul- tured with 20E in conditioned medium (CM) that was prepared by incubating V5 ASGs for 48 h in the absence of 20E, they did not undergo PCD. This result indicates that an inhibitory factor was released from V5 ASGs into the medium and thereby prevented V5 ASGs from responding to 20E. Purification and identification of the inhibitory fac- tor revealed that it was glucose oxidase (GOD). GOD (EC 1.1.3.4.) is an oxidoreductase that catalyzes the oxidation of glucose to gluconolactone, and generates hydrogen peroxide as a byproduct. Its presence in ani- mals has only been observed in insects. In insects, GOD occurs as a component of salivary secretions in various lepidopterans and hymenopterans [7]. GOD, produced by the adult hypopharyngeal gland of Apis mellifera, is suggested to act as an antibacterial substance in honey, through hydrogen peroxide forma- tion. In addition, the suggested natural functions of GOD include the detoxification of noxious plant components involved in plant defense responses [7]. In contrast, the in vivo function of GOD has not yet been determined. In the present study, we show that, during the incu- bation of V5 ASGs, hydrogen peroxide is generated and glucose is consumed in Grace’s medium, and this hydrogen peroxide is the primary substance involved in inhibiting 20E-induced PCD. As the silkworms entered the prepupal stage, GOD disappeared from the ASG inner cavity, owing to the onset of spinning, which caused the release of GOD from ASGs along with the silk proteins. Thus, ASGs control the timing of their own degeneration by preserving GOD. Results The presence of the PCD inhibitor in V5 ASGs V7 ASGs undergo PCD when cultured with 1 lm 20E (Fig. 1A,G), but V5 ASGs do not undergo PCD under the same conditions [6]. When V7 ASGs were cocul- tured with V5 ASGs, V7 ASGs did not undergo PCD (Fig. 1B,H). Such inhibition of PCD was duplicated with the use of CM in which V5 ASGs had been F f 0 1 2 3 4 5 6 PCD sc o r e Culture period (h) 024487296120144 V7 + V5 V7 + V7 B V7 V5 0 1 2 3 4 5 6 A PCD score Culture period (h) 024487296120144 0 1 2 3 4 5 6 PCD score Culture period (h) 024487296120144 x1 x2 D 0 1 2 3 4 5 6 PCD score Culture period (h) 024487296120144 x1 x2 C 0 0.2 0.4 0.6 0.8 1 Unit/gland day IV1234V01234567 E G g Hh I i Fig. 1. PCD progression of V7 ASGs induced by 20E in vitro and PCD inhibition of V7 ASGs by V5 ASGs. PCD progression is expressed as an increase in PCD score [4]. (A) V7 ASGs responded to 20E by undergoing PCD, but V5 ASGs did not. (B) One V7 ASG was cultured with a V5 ASG (V7 + V5) or two V7 ASGs (V7 + V7) in 0.6 mL of Grace’s medium with 20E. (C) The presence of the inhibitory factor in CM prepared by incubating V5 ASGs in 0.3 mL of Grace’s medium without 20E for 48 h (V5 CM). The activity was assessed for the undiluted CM (·1) or CM doubly diluted with Grace’s medium (·2). (D) The inhibitory activity in V5 ASGs was assessed from the crude extracts of V5 ASGs after adjustment of the extract concentration to 0.5 (·2) or 1 (·1) ASG equivalent with Grace’s medium. (E) Changes in the inhibitory activity in CM through the fourth and fifth instar. The ordinate indicates the activ- ity relative to that in V5 CM (1 U). (F,f) V7 ASGs at 144 h of culture in Grace’s medium alone or (G,g) in Grace’s medium with 1 l M 20E. (H,h) V7 ASGs were cocultured with V5 ASGs in Grace’s med- ium with 20E or (I,i) cultured in V5 CM with 20E. (F–I) Light micros- copy and (f–i) 4¢,6-Diamidino-2-phenylindole staining to show nuclear morphology. See [4] for a detailed description of nuclear morphology (mean ± standard deviation; n = 12–24). Scale bar: 70 lm. H. Matsui et al. Glucose oxidase prevents programmed cell death FEBS Journal 278 (2011) 776–785 ª 2011 The Authors Journal compilation ª 2011 FEBS 777 incubated for 48 h (V5 CM). V7 ASGs did not undergo PCD when cultured in V5 CM (Fig. 1C,I), but twice-diluted V5 CM exhibited no inhibitory activ- ity (i.e. all of the V7 ASGs exhibited PCD scores of 5– 6). These results indicate that V5 ASGs may release an inhibitor of PCD. Furthermore, V5 ASGs also released the inhibitor in NaCl ⁄ P i or Ringer’s solution for insects (Fig. S1). To examine whether the gland itself contains the inhibitor, crude V5 ASG extracts were subjected to the inhibitor assay, either directly or after double dilution. The extracts exhibited inhibitory activity similar to that of V5 CM, indicating that V5 ASGs contained the inhibitor (Fig. 1D). Changes of inhibitor-producing activity during the larval period We examined the inhibitor activity in the CM of ASGs from day 1 of the fourth instar to day 7 of the fifth instar (Fig. 1E). The assays indicated that IV1–IV4 and V7 CM exhibited little or no inhibitory activity, because the CM of five glands from these days did not inhibit PCD. In V6 ASGs, CM prepared with three ASGs had 1 unit of activity, indicating that V6 CM contained one-third of a unit of activity. Purification and identification of the PCD inhibitor After the debris was removed by centrifugation, the V3 CM prepared with NaCl ⁄ P i was concentrated and subjected to a two-step HPLC separation. The chro- matographic peak at 12.9 min on a Mono Q column exhibited inhibitory activity (Fig. 2A). The fractions corresponding to the active peak were then loaded directly onto a Superdex 200 column, and the fractions with retention times of 18–24 min were found to have PCD inhibitory activity (Fig. 2B). Those fractions were combined and subjected to SDS ⁄ PAGE, which revealed a major band at 80 kDa and a minor band at 120 kDa (Fig. 2C). The 120-kDa band was the main component of the peak at 24.3 min, but the peak fraction did not exhibit the inhibitory activity. Therefore, the 80-kDa band was subjected to amino acid sequencing. Ten peptides obtained by trypsin digestion were suc- cessfully sequenced (Fig. S2), and the comparison of these sequences with 14 000 proteins in SilkDB (http:// silkworm.genomics.org.cn/) indicated similarity to a novel protein, with a sequence coverage value of 100% and a score of 1269 (BGIBMGA012872). The amino acid sequence of the novel protein was subsequently submitted to blast analysis with the DNA Data Bank of Japan (DDBJ; http://www.ddbj.nig.ac.jp/). The purified protein exhibited a strong similarity to a Heli- coverpa armigera GOD-like enzyme, with an identity value of 66% and a score of 800, indicating that the inhibitor of PCD may be B. mori GOD. Inhibitory activity of fungal and B. mori GOD V7 ASGs were cultured with 20E in Grace’s medium containing the GOD of Aspergillus niger. This GOD inhibited PCD at a concentration of 0.39 lm, and the inhibition was more prominent at 0.78 lm (Fig. 3A), indicating that A. niger GOD possessed PCD inhibi- tory activity. We examined the GOD activity in the extracts of V3 ASGs, V3 CM, and the active fraction of Superdex 200 200 116 98 66 C 1234 kDa Absorbance (280 nm) 0 0.16 0.32 0 50 100 01020 Retention time (min) A 0.7 M NaCl (%) 020 Retentiontime(min) 40 0 0.08 0.16 B Absorbance (280 nm) Fig. 2. Purification of the PCD inhibitor by two chromatographic steps. The eluent was monitored at 280 nm. (A) Mono Q column chromato- gram. The broken line denotes the NaCl gradient. The inhibitor was eluted at the peak indicated by cross-hatching. (B) Separation on a Superdex 200 column. The cross-hatched fraction (18–24 min) exhibited the inhibitory activity. (C) SDS ⁄ PAGE (6% polyacrylamide) separa- tion of the active fraction in (B). Approximately 5 lg of protein was loaded in each lane. Lane 1: CM prepared by incubation of 2000 ASGs. Lane 2: upper phase of ultrafiltration of the CM. Lane 3: the active peak in (A). Lane 4: the active fraction in (B). Glucose oxidase prevents programmed cell death H. Matsui et al. 778 FEBS Journal 278 (2011) 776–785 ª 2011 The Authors Journal compilation ª 2011 FEBS (Fig. 2B), all of which were adjusted to contain 1 unit of inhibitory activity. GOD activities in the extracts, CM and the active fraction were 67.5 ± 10.6, 67.7 ± 8.3 and 56.9 ± 3.7 nmolÆmin )1 , respectively (Fig. 3B). The activity of 0.78 lm A. niger GOD was 499.7 ± 13 nmolÆmin )1 . Among the V3 larval tissues, GOD activity was found primarily in ASGs. The activ- ity in middle silk glands was very low, and no activity was detected in the other tissues examined (Fig. 3C). Changes in GOD activity in CM indicated that the activity increased from day 0 to day 4, peaked at day 4, and decreased to a low level on day 6 (Fig. 3D). A similar result was obtained for the ASG extracts (Fig. 3E). We next examined GOD gene expression for the three parts of V0 silk glands, and found that the gene was strongly and equally expressed in ASGs and mid- dle silk glands but not in posterior silk glands (Fig. 3C, inset). During the fourth and fifth instar, GOD expression in ASGs was strong in IV2 and V0 larvae, and weak but detectable on the other days (Fig. 3F). The change in the amount of GOD enzyme protein did not directly correspond to the change in gene expression; the enzyme level increased after the decline of gene expression to a low level, and peaked on day 3 (Fig. 3F,G). Mechanism of PCD inhibition by GOD GOD catalyzes the oxidation of glucose to gluconolac- tone, and produces hydrogen peroxide as a byproduct. We examined the substrate specificity of B. mori GOD with the use of crude V3 ASG extracts, and found that only glucose serves as a substrate (Fig. S3). Prior to examining whether the inhibitory activity of GOD was caused by the enzymatic activity of GOD, we needed to reproduce the 20E-induced PCD in medium without glucose (Grace’s medium contains 4.7 mm glucose). Thus, we cultured V7 ASGs with 20E in Ringer’s solu- tion, and found that the ASGs underwent PCD. Next, V3 ASGs were cultured with 20E in Ringer’s solution, in which all ASGs attained PCD scores of 4–6 (Fig. 4A). However, PCD was not induced in a 1 : 1 mixture of Ringer’s solution and Grace’s medium (Fig. 4A), an indication that GOD may inhibit PCD though an enzymatic reaction, using the glucose in Grace’s medium. To examine this issue, we cultured V3 ASGs in Ringer’s solution with 1 lm 20E and 0.39 µ 0.78 0 1 2 3 4 5 6 PCD score 0 24487296120144 A Culture period (h) Stage (day) IV234V01234567 0 100 200 300 400 E 0 100 200 Stage (day) IV2 3 4 V0 1 2 3 4 5 6 7 D Ext CM AnG O D PS B 0 200 400 600 MSG PSG Hemolymph Salivary gland Mid gu t C ASG 0 100 200 H 2 O 2 (nmol·min –1 ·mg –1 ) H 2 O 2 (nmol·min –1 ·mL –1 )H 2 O 2 (nmol·min –1 ) H 2 O 2 (nmol·min –1 ·mg –1 ) F V0 V2 V3 V5 V6 V7 GOD RpL3 IV2 IV4 GOD protein R ela t iv e i nte n sity V0 V2 V3 V5 V6 V7IV2 IV4 G Fig. 3. Characterization of GOD. (A) Fungal GOD inhibits 20E-induced PCD. V7 ASGs were cultured with 20E in Grace’s medium containing A. niger GOD (AnGOD). (B) Pro- duction of hydrogen peroxide by the crude extracts of V3 ASGs (Ext), V3 CM (CM), 0.78 l M of AnGOD, and the active fraction from the Superdex 200 column (PS). Each sample contained 1 unit of inhibitory activity, which is equivalent to that in the V5 CM prepared from one V5 ASG. (C) GOD activity in the extracts of various tissues of V3 lar- vae and (C, inset) GOD expression in the ASG, the middle silk gland (MSG) and the posterior silk gland (PSG) of V0 larvae, as determined by RT-PCR. (D) Developmental changes of B. mori GOD activity in CM, which was prepared using Ringer’s solution, and (E) in ASG extracts [mean ± standard deviation, n = 13 for (A) and n = 3 for (B)–(E)]. (F) Developmental profile of GOD expression, as determined by RT-PCR, and the presence of GOD protein, as deter- mined by western blotting. Gene expression in IV4 and V2–V7 larvae was very faint but detectable. (G) Relative intensity of GOD gene expression (open column) and amount of GOD protein (closed column), depicted on the basis of the data in (F). H. Matsui et al. Glucose oxidase prevents programmed cell death FEBS Journal 278 (2011) 776–785 ª 2011 The Authors Journal compilation ª 2011 FEBS 779 various concentrations of glucose (Fig. 4B). The PCD score decreased with increases in glucose concentra- tion, and PCD was entirely inhibited at 2 mm glucose. These results suggested that hydrogen peroxide, gluconolactone or both may inhibit PCD. Thus, we cultured V7 ASGs with 20E in Grace’s medium supplemented with hydrogen peroxide or gluconolac- tone. Hydrogen peroxide (3 mm) interfered with the progression of PCD, but 10 mm gluconolactone did not inhibit PCD (Fig. S4). Therefore, hydrogen peroxide may be the substance responsible for PCD inhibition. To confirm this finding, we added 0.36 lm catalase to Ringer’s solution containing 2 mm glucose, and cul- tured V3 ASGs with 20E. In this medium, PCD was not inhibited (Fig. 4C), supporting the notion that hydrogen peroxide is the factor that inhibits PCD. During the incubation of V3 ASGs, the concentration of hydrogen peroxide increased, and reached 268.1 ± 88.1 lm at 12 h (Fig. 4D). By contrast, the presence of catalase inhibited the increase in hydrogen peroxide concentration throughout the incubation period. Combining these observations, we concluded that the production of hydrogen peroxide by the enzymatic reaction of GOD results in the inhibition of PCD. Selective suppression of 20E-inducible genes by hydrogen peroxide V7 ASGs were incubated for various lengths of time in V3 CM, which was prepared with Ringer’s solution and adjusted to have 1 unit of inhibitory activity with Grace’s medium. The ASGs were then transferred to fresh Grace’s medium with 20E, and cultured for 144 h. Under these conditions, a 3-h incubation in CM sufficiently inhibited the gland’s responsiveness to 20E (Fig. 5A). Alternatively, V7 ASGs were incubated with 3or6mm hydrogen peroxide for 0–6 h, and then cul- tured in Grace’s medium with 20E. In these conditions, a 3-h incubation with 6 mm hydrogen peroxide inhib- ited PCD (Fig. 5B). In addition, these results indicated that the inhibition was irreversible, because PCD did not proceed after the transfer to Grace’s medium. To examine whether this irreversible inhibition was caused by the toxic effects of hydrogen peroxide, the expres- sion of early and early–late genes was determined in four different media (Fig. 5C). Among the genes exam- ined, expression levels of E75A, broad complex Z1 (BR-C Z1) and Bombyx hormone receptor 3 (BHR3) were greatly reduced, but the expression of other genes was not significantly altered. Because these three genes were suggested to be involved in the early phase of PCD progression in the prepupal period in B. mori [8], the inhibition of PCD may not be attributable to the toxicity of hydrogen peroxide. Explanation for missing GOD activity in V7 ASGs GOD activity in ASGs was rapidly lost from day 5 to day 7 (Fig. 3D). Thus, questions remain regarding the cause of this decreased GOD activity. GOD activity was found in ASG extracts, indicating that GOD may be present in ASG cells, the inner cavity, or both, and that GOD is released into the medium during incubation, 400 300 200 100 0 H 2 O 2 (µM) 013122448 Culture period (h) Glu 2 mM Cat 0.36 µM Glu + Cat 0 1 2 3 4 5 6 PCD score Culture p eriod ( h ) 0 24 48 72 96 120 144 RS RS + 20E RS + GM RS + GM + 20E PCD score 0 1 2 3 4 5 6 Response (%) 100 80 60 40 20 0 A 0 1 2 3 4 5 6 PCD score 00.5 Glucose (mM) 11.52 B C D Fig. 4. Hydrogen peroxide causes PCD inhi- bition. (A) V3 ASGs were cultured with 20E in Ringer’s solution (RS) or a 1 : 1 mixture of Ringer’s solution and Grace’s medium (GM). The ordinate indicates the percentage of the glands that exhibited each score at the end of the culture for 144 h. (B) Inhibi- tion of PCD by glucose in the medium. V3 ASGs were cultured in Ringer’s solution with 20E and various concentration of glu- cose for 144 h. (C) V3 ASGs were cultured with 20E in Ringer’s solution containing glu- cose, catalase (Cat), or both. (D) Generation of hydrogen peroxide by incubation of V3 ASGs in Ringer’s solution (black), Ringer’s solution with 2 m M glucose (white) or Ringer’s solution with 2 m M glucose and 0.36 l M catalase (gray) for the indicated time periods [mean ± standard deviation; n = 12 (A–C) and n = 3 (D)]. Glucose oxidase prevents programmed cell death H. Matsui et al. 780 FEBS Journal 278 (2011) 776–785 ª 2011 The Authors Journal compilation ª 2011 FEBS probably through the cut surface of the cultured ASGs. In addition, ASGs obtained late on day 5 did not complete PCD in Grace’s medium, but ASGs from the middle of day 6 did complete PCD [6]. During this period, larvae begin to spin cocoons. Therefore, we assumed that GOD may be discarded with silk thread through spinning. To test this hypothesis, we interfered with spinning by occluding spinnerets with paraffin in the scotophase of V5 larvae (before the onset of spin- ning) and V6 larvae (after the onset). Then, ASGs were dissected on day 8 and cultured in Grace’s med- ium with 20E. The occlusion on day 5 inhibited the PCD of V8 ASGs, whereas occlusion on day 6 failed to do so (Fig. 6A). In addition, the ASGs of larvae whose spinnerets were occluded on day 5 retained GOD activity, but the ASG extracts from larvae trea- ted on day 6 did not (Fig. 6B). These results indicated the importance of spinning to the loss of GOD activity in ASGs. Finally, we examined the localization of the GOD enzyme by active staining for the silk glands of V3 lar- vae; ASGs were stained reddish–brown (Fig. 6C,c). The stained site in the ASG was restricted to the inner cavity, but ASG cells were not stained. Middle silk glands appeared to be stained faintly at their most anterior region, near the boundary between the ASG and the middle silk gland (Fig. 6D,d). The main bodies of middle silk glands (Fig. 6E,e) and posterior silk glands (Fig. 6F,f) were not stained at all. We also stained fragments of cocoons with o-dianisidine, and observed that the outside layer of the cocoon was partly stained (Fig. 6G,g). Discussion After the onset of spinning, ASGs undergo cell death in response to 20E treatment in vitro, but they do not do so before spinning [6]. Although such differences in responsiveness to 20E have been considered to be caused by a lack of competence to respond to 20E [9], our results clearly indicate that the unresponsiveness of ASGs before spinning is caused by a PCD inhibitory factor produced by the glands themselves. The most striking findings of this study were that the PCD inhib- itory factor is GOD and that the substance responsible for the inhibition of PCD is hydrogen peroxide. –20E +20E RS-CM + 20E H 2 O 2 + 20E EcR-A EcR-B USP-1 USP-2 E75A E75B E74A E74B BR-C Z1 BR-C Z 2 BR-C Z 4 BHR3 RpL3 C 6 R e s p o n se (%) R e s p o n se (%) 0 20 40 60 80 100 1 3 Hours in CM A PCD score 0 1 2 3 4 5 6 H 2 O 2 (mM) Hours with H 2 O 2 3 6 3 0 3 3 6 3 6 6 0 20 40 60 80 100 B Fig. 5. Minimum period of hydrogen peroxide incubation required for PCD inhibition and inhibition of 20E-inducible genes. (A) Two V3 ASGs were incubated in 0.3 mL of Ringer’s solution, and then an equal volume of Grace’s medium was added. Immediately after preparation of the mixed medium, V7 ASGs were incubated in the mixed medium with 20E for the indicated length of time, and then cultured in Grace’s medium with 20E for 144 h. (n = 12) (B) V7 ASGs were incubated in Grace’s medium with 3 or 6 m M hydrogen peroxide for 3 or 6 h with 20E, and then cultured in Grace’s medium with 20E for 144 h. (n = 12) (C) Effects of hydrogen peroxide on 20E-responsive gene expression. V7 ASGs were incubated for 3 h in Grace’s medium without 20E ()20E), Grace’s medium with 20E (+20E), a 1 : 1 mixture of Grace’s med- ium and V3 CM prepared by a 48-h incubation of two V3 ASGs in 0.3 mL of Ringer’s solution (RS-CM+20E), or Grace’s medium with 20E and 6 m M hydrogen peroxide (H 2 O 2 + 20E). EcR, ecdysone receptor; usp, ultraspiracle; BR-C, broad complex; BHR3, Bombyx hormone receptor 3; RpL3, ribosomal protein L3. H. Matsui et al. Glucose oxidase prevents programmed cell death FEBS Journal 278 (2011) 776–785 ª 2011 The Authors Journal compilation ª 2011 FEBS 781 Although hydrogen peroxide is classically considered to be a toxic factor for cells, NADPH oxidase (NOX) in mammalian cells generates reactive oxygen species, including hydrogen peroxide, that play an important role in the regulation of cell proliferation and growth at nontoxic levels (i.e. nanomolar to micromolar level) [10]. Prostate cancer cells produce substantial amounts of reactive oxygen species, in part from stimulation of NOX. Antioxidants or NOX inhibitors inhibit cell proliferation, and instead induce apoptosis [11,12]. Furthermore, exogenous hydrogen peroxide inhibits the activity of caspase-3 [13,14]. These observations suggest that hydrogen peroxide regulates growth and apoptosis in cancer cells. In B. mori fifth instars, the ASGs notably increase in tissue size [15], and both the amount of GOD protein and GOD activity increased, indicating the involvement of GOD-produced hydro- gen peroxide in the growth of ASGs. This phenome- non also suppressed PCD induction by 20E, because 20E-induced PCD depends on the activation of a cas- pase-3-like protein [5,16]. Culturing V7 ASGs in CM irreversibly inhibited 20E-induced PCD, and a 3-h incubation in V3 CM was also sufficient for inhibition. Indeed, the treated ASGs never exhibited any sign of the change in cellu- lar and nuclear morphology specific to PCD or under- went necrosis, even after 144 h in culture. The incubation of V7 ASGs in V3 CM or with 6 mm hydrogen peroxide for 3 h altered the expression of ecdysone-responsive genes; large decreases in the expression of E75A, BR-C Z1 and BHR3 were noted. These genes are upregulated by 20E in V7 ASGs, and are therefore suggested to be involved in the progres- sion of PCD [8]. In the cell death of Drosophila salivary glands, Drosophila hormone receptor 3 (DHR3) is the key gene in the hierarchy of gene expression induced by 20E [17]. During the expression of dopade- carboxylase (the key enzyme for the melanization that occurs shortly after every molting), MHR3,aManduca homolog of DHR3, is the key transcriptional factor in the sequential expression of 20E-induced gene,s result- ing in enzyme expression. These findings indicate that, among the genes inhibited by hydrogen peroxide in V7 ASGs, BHR3 may be the key gene involved in the inhibition of PCD. Expression of the GOD gene, as well as changes in GOD activity and the amount of GOD protein, are not directly connected to one another. In the fourth instar, neither GOD protein nor GOD activity was present, regardless of the high-level expression of the GOD-encoding gene. In the fifth instar, GOD gene expression was high on day 0 and reduced by day 2. However, the amount of GOD protein and its enzy- matic activity peaked on day 3. When performing wes- tern blot analysis, we applied a fixed amount of total protein from crude ASG extracts. If the GOD is pres- ent as a proenzyme, post-translational modifications may contribute to the increase in its enzymatic activity. However, if this was the case with GOD, the signal intensity of GOD in western blot would remain at the original level. Thus, at present, there is no interpreta- tion for the discrepancy between the above-mentioned temporal changes. GOD was localized to the inner cavity of ASGs, suggesting that hydrogen peroxide may be generated in the cavity and act on the cell surface facing the cavity. Thus, the question is how the GOD substrate, glucose, is provided in the cavity. The insect salivary gland con- tains trehalose (a-d-glucopyranosyl-a-d-glucopyrano- side), the major sugar in insect hemolymph [18]. Trehalase is present with the liquid silk in the silk glands, indicating that trehalase is produced by the silk gland cells and is localized to the cavity of ASGs [19,20]. The silk gland and the salivary gland are homologous organs [21], and both produce silk protein [22]. Therefore, it is possible that trehalose is present g C c D E e F f G d g 100 80 60 40 20 0 Response (%) A 0 100 200 B H 2 O 2 ( nmol·min –1 ·mg –1 ) V5 V6 Fig. 6. GOD is discarded with the spinning of silk thread. Effects of spinneret occlusion on the loss of GOD activity from ASGs. Spinnerets of V5 or V6 larvae were occluded with paraffin, and their ASGs were dissected on day 8 to examine their responses to 20E (A) and assess the GOD activity in ASG extracts (B). The ordi- nates in (A) and (B) are the same as in Figs 4A and 3E, respectively [mean ± standard deviation; n = 12 (A) and n = 3 (B)]. (C–G) Locali- zation of GOD in V3 silk glands and a cocoon fragment, as deter- mined by active staining of GOD with o-dianisidine, and (c–g) control tissues incubated without o-dianisidine. (C,c) ASG, (D,d) boundary between the ASG and the middle silk gland, as indicated by an arrowhead, (E,e) the middle silk gland, and (F,f) the posterior silk gland. (G,g) cutting surface of a cocoon. Glucose oxidase prevents programmed cell death H. Matsui et al. 782 FEBS Journal 278 (2011) 776–785 ª 2011 The Authors Journal compilation ª 2011 FEBS in the ASG inner cavity and serves as a substrate for trehalase to produce glucose, from which GOD gener- ates hydrogen peroxide. The available amount of tre- halose, and therefore glucose, may be limited and may therefore limit the hydrogen peroxide present in the cavity. ASGs at noon of day 6 respond to 20E in vitro by undergoing cell death, whereas ASGs from the night of day 5 (i.e. 12–18 h earlier) do not [6]. During this period, larvae begin to spin cocoons, and the first step of cocooning is the construction of a scaffold with silk thread to keep the larvae in a position suitable for completing the cocoon. The silk scaffold remains as the floss around the completed cocoon, and GOD activity was found to remain with the floss. This find- ing indicates that the GOD enzyme in the ASG inner cavity was released with the silk at the early spinning stage. Because the onset of spinning on day 6 is induced by 20E, 20E indirectly induces the responsive- ness of ASGs to 20E. After discarding GOD through spinning, ASGs exhibit overt responsiveness to 20E. The finding that V3 ASGs underwent PCD in Ringer’s solution with 20E shows that they acquire the respon- siveness to 20E 3 days before the onset of spinning, and their responsiveness is covered by GOD in the ASG inner cavity. This may be a typical example of a behavior that is directly involved in the regulation of cellular events at the biochemical and molecular levels. In conclusion, the elimination of GOD from ASGs accompanies metamorphosis, and GOD thus plays a critical role in the control of PCD timing. Experimental procedures Animals B. mori larvae were reared and staged as previously described [23]. V7 larvae that began spinning on day 6 of photophase, followed by gut purge during the scotophase of the same day, were used in the present study. ASGs were obtained during the photophase of individual days. Hormones and in vitro culture 20E (Sigma, St Louis, MO, USA) was dissolved in distilled water (1 mgÆmL )1 ) and stored at )20 °C. ASGs were rinsed with Grace’s insect cell culture medium (Invitrogen, Carls- bad, CA, USA) and cultured individually in 0.3 mL of the medium (pH 6.4, adjusted with NaOH) with 1 lm 20E in 24-well plates (Grainer Bio-One, Frickenhausen, Germany) at 25 °C. The ASGs were observed every 24 h, and the degree of PCD progression was scored according to the changes in cellular morphology. This progression was expressed with a PCD score of 0–6, where a score of 0 indi- cates no change, and a score of 6 signifies completion of PCD with the formation of apoptotic bodies [4]. Nuclear morphology was observed by 4¢,6-diamidino-2-phenylindole staining [5]. Preparation of CM V3 or V5 ASGs were individually incubated for 48 h in 0.3 mL of Grace’s medium, NaCl ⁄ P i (137 mm NaCl, 2.7 mm KCl, 8.1 mm Na 2 HPO 4 , 1.47 mm KH 2 PO 4 , pH 7.4), or Ringer’s solution (128 mm NaCl, 4.7 mm KCl, 1.9 mm CaCl 2 ). After incubation, the medium was recov- ered and stored at 4 ° C until use. After the incubation period, the media were designated as CMs. Crude ASG extraction A fresh ASG was homogenized in 0.15 mL of Ringer’s solution for assays of PCD inhibitory activity or 0.30 mL of 20 mm phosphate buffer (24.4 mm Na 2 HPO 4 , 15.6 mm NaH 2 PO 4 , pH 7.0) for GOD assays. The homogenates were centrifuged at 15 000 g for 10 min at 4 °C. The resulting supernatant was diluted twice with Grace,s medium for the inhibitory activity assay or used directly for the GOD assay. Protein quantities were determined with an RC DC protein assay kit (Bio-Rad, Hercules, CA, USA). PCD inhibitory activity assay PCD inhibitory activity was assayed by culturing V7 ASGs in Grace’s medium with 1 lm 20E for 144 h, and the degree of PCD progression was recorded every day or at the end of the culture period. One unit of inhibitory activity was defined as the activity equivalent to that of the CM pre- pared by incubating one V5 ASG in 0.3 mL of Grace’s medium for 48 h (V5 CM). When the CM of one gland had less than 1 unit of the activity, the number of glands cultured in 0.3 mL of medium was increased to five. Con- versely, the CM was diluted with Grace’s medium when the CM of one gland exhibited 1 unit of activity. Purification and identification of the inhibitor CM (300 mL) was prepared by incubating 2000 individual V3 ASGs in NaCl ⁄ P i . The CM was then centrifuged at 15 000 g for 10 min at 4 °C. The supernatants were pooled and concentrated to 15 mL with a centrifugal filter (Ultra YM-30; Amicon, Beverly, MA, USA). The concentrate was applied to a Mono Q column (0.05 · 5 cm; GE Healthcare, Little Chalfont, UK), equilibrated with 20 mm phosphate buffer (33.6 mm Na 2 HPO 4 , 6.4 mm NaH 2 PO 4 , pH 7.5), and eluted with a linear gradient of 0–0.7 m NaCl in phosphate buffer at a flow rate of 0.8 mLÆmin )1 . The active fraction H. Matsui et al. Glucose oxidase prevents programmed cell death FEBS Journal 278 (2011) 776–785 ª 2011 The Authors Journal compilation ª 2011 FEBS 783 was concentrated to 2.5 mL with an Ultra YM-30, sub- jected to gel filtration (Superdex 200, 0.1 · 30 cm; GE Healthcare) on a column equilibrated with phosphate buf- fer containing 0.15 m NaCl at a flow rate of 0.5 mLÆmin )1 , and eluted under the same conditions. Eluates were col- lected every 3 min, and the active fractions were combined and concentrated to 1 mL. The concentrate was subjected to 6% SDS ⁄ PAGE [24], and the gel was stained with Coomassie Brilliant Blue R-250 without fixation. A protein band was e xcised from t he gel and analyzed by MALDI-TOF MS. MS ⁄ MS w as p erformed with an Applied Biosystems 4800 plus MALDI TOF ⁄ TOF Analyzer (Applied Biosystems, Foster City, CA, USA). A saturated solution of a-cyano- 4-hydroxycinnamic acid in acetonitrile ⁄ water (3 : 7, v ⁄ v) was used as the matrix. GOD activity assay GOD activity was determined by the o-dianisidine method [25]. The reaction mixture (in a total volume of 1 mL) contained 0.17 mm o-dianisidine-HCl (Sigma) in 20 mm phosphate buffer (pH 7.0), 90 mmd-glucose, 60 UÆmL )1 horseradish peroxidase (Sigma), and 34.5 lL of sample solution in phosphate buffer. For the negative control, 34.5 lLof20mm phosphate buffer (pH 7.0) was used instead of the sample solution. The reaction mixture was incubated without sample addition at 35 °C for 5 min, and samples were then added and incubated for 5 min; this was followed by recording of absorbance at 500 nm. GOD activity is expressed as the production rate of hydrogen peroxide (nmolÆmin )1 per mg protein for tissue extracts and nmolÆmin )1 per mL for CM). Active staining of GOD Tissues and cocoon fragments were incubated without sam- ple solution in the above-mentioned reaction mixture at 35 °C for 15 min. After incubation, the tissues and frag- ments were rinsed with Ringer’s solution and observed for the localization of GOD, according to the reddish-brown coloration caused by the oxidation of o-dianisidine. RT-PCR Total RNA was extracted from tissues [26] and treated with RNase-free DNase (Promega, Madison, WI, USA). Com- plementary DNA was prepared from 1 lg of total RNA using anchored oligo-dT [5¢-(T)12(A ⁄ C ⁄ G)(A ⁄ C ⁄ G ⁄ T)-3¢] and Reve Tra Ace reverse transcriptase (Toyobo, Osaka, Japan). For RT-PCR, GOD cDNA was amplified for 35 cycles with the following primers: (forward, 5¢-AACGGC CAGAGGTACACAAC-3¢; reverse, 5¢-GATTCAAACCCA CTGGGAGA-3¢). RT-PCR targeting 20E-induced genes was performed with the same primer sets as described in [8]. RNA encoding ribosomal protein L3 (RpL3) was used as an internal standard and amplified for 25 cycles. PCR products were separated by agarose gel electrophoresis. Western blot analysis Western blot analysis was performed with 8% SDS ⁄ PAGE [27]. The membrane was probed with a primary antibody prepared from rabbits against a peptide (DASVMPS QPTGNPQ) designed from the putative amino acid sequence of B. mori GOD at a 1 : 250 dilution, and this was followed by an incubation with horseradish peroxidase- conjugated goat anti-(rabbit IgG) (Cell Signaling Technol- ogy, Danvers, MA, USA) and detection with an ECL Advance kit (GE Healthcare). Acknowledgements This work was supported by JSPS Grant-in-Aid for Scientific Research 21380035 (to S. Sakurai). References 1 Yin VP & Thummel CS (2005) Mechanisms of steroid- triggered programmed cell death in Drosophila. Semin Cell Dev Biol 16, 237–243. 2 Fahrbach SE, Nambu JR & Schwartz LM (2005) Pro- grammed cell death in insect neuromuscular systems during metamorphosis. In Comprehensive Insect Molecu- lar Science (Gilbert LI, Iatrou K & Gill SS eds), Vol 2, pp. 165–198. Elsevier Pergamon, Oxford. 3 Dai J-D & Gilbert LI (1997) Programmed cell death of the prothoracic glands of Manduca sexta during pupal– adult metamorphosis. Insect Biochem Mol Biol 27, 69– 78. 4 Terashima J, Yasuhara N, Iwami M & Sakurai S (2000) Programmed cell death triggered by insect steroid hor- mone, 20-hydroxyecdysone, in the anterior silk gland of the silkworm, Bombyx mori. Dev Genes Evol 210, 545– 558. 5 Iga M, Iwami M & Sakurai S (2007) Nongenomic action of an insect steroid hormone in steroid-induced programmed cell death. Mol Cell Endocrinol 263, 18–28. 6 Kakei M, Iwami M & Sakurai S (2005) Death commit- ment in the anterior silk gland of the silkworm, Bombyx mori. J Insect Physiol 51, 17–25. 7 Wong CM, Wong KH & Chen XD (2008) Glucose oxidase: natural occurrence, function, properties and industrial applications. Appl Microbiol Biotechnol 78, 927–938. 8 Sekimoto T, Iwami M & Sakurai S (2006) Coordinate responses of transcription factors to ecdysone during programmed cell death in the anterior silk gland of the silkworm, Bombyx mori. Insect Mol Biol 15, 281–292. Glucose oxidase prevents programmed cell death H. Matsui et al. 784 FEBS Journal 278 (2011) 776–785 ª 2011 The Authors Journal compilation ª 2011 FEBS 9 Riddiford LM (1996) Juvenile hormone: the status of its ‘status quo’ action. Arch Insect Biochem Physiol 32, 271–286. 10 Stone JR & Yang S (2006) Hydrogen peroxide: a signal- ing messenger. Antioxid Redox Signal 8, 243–270. 11 Gunawardena K, Murray DK & Meikle W (2000) Vitamin E and other antioxidants inhibit human prostate cancer cells through apoptosis. Prostate 44, 287–295. 12 Brar SS, Corbin Z, Kennedy TP, Hemendinger R, Thornton L, Bommarius B, Arnold RS, Whorton AR, Sturrock AB, Huecksteadt TP et al. (2003) NOX5 NAD(P)H oxidase regulates growth and apoptosis in DU 145 prostate cancer cells. Am J Physiol Cell Physiol 285, 353–369. 13 Hampton MB & Orrenius S (1997) Dual regulation of caspase activity by hydrogen peroxide: implications for apoptosis. FEBS Lett 414, 552–556. 14 Borutaite V & Brown GC (2001) Caspases are revers- ibly inactivated by hydrogen peroxide. FEBS Lett 500, 114–118. 15 Julien E, Coulon-Bublex M, Garel A, Royer C, Cha- vancy G, Prudhomme J-C & Couble P (2005) Silk gland development and regulation of silk protein genes. In Comprehensive Insect Molecular Science (Gilbert LI, Iatrou K & Gill SS eds), Vol 2, pp. 369–386. Elsevier Pergamon, Oxford. 16 Iga M, Manaboon M, Matsui H & Sakurai S (2010) Ca 2+ -PKC-caspase 3-like protease pathway mediates DNA and nuclear fragmentation in ecdysteroid-induced programmed cell death. Mol Cell Endocrinol 321, 146– 151. 17 Lam G, Hall BL, Bender M & Thummel CS (1999) DHR3 is required for the prepupal–pupal transition and differentiation of adult structures during Drosoph- ila metamorphosis. Dev Biol 212, 204–216. 18 Hori K (1975) Digestive carbohydrases in the salivary gland and midgut of several phytophagous bugs. Comp Biochem Physiol 50B, 145–151. 19 Shimada S (1975) Trehalase activity in the silk glands of the silkworm, Bombyx mori (Lepidoptera: bombyci- dae). Can Entomol 107, 1311–1314. 20 Shimada S & Hyashiya K (1975) Hydrolysis of sucrose by the cocoon of the silkworm, Bombyx mori. Insect Biochem 5, 653–657. 21 Tucker AS (2010) Salivary gland adaptations: modifica- tion of the glands for novel uses. Front Oral Biol 14, 21–31. 22 Sehnal F & Sutherland T (2008) Silks produced by insect labial glands. Prion 4, 145–153. 23 Sakurai S, Kaya M & Satake S (1998) Hemolymph ecdysteroid titer and ecdysteroid-dependent develop- mental events in the last-larval stadium of the silkworm, Bombyx mori: role of low ecdysteroid titer in larval pupal metamorphosis and a reappraisal of the head critical period. J Insect Physiol 44, 867–881. 24 Laemmli KU (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage. Nature 227, 680–685. 25 Bergmeyer HU, Gawehn K & Grassl M (1974) Enzymes as biochemical reagents. In Methods of Enzymatic Anal- ysis (Bergmeyer HU eds), Vol 1, 2nd Edn, pp. 457–458. Academic Press, New York, NY. 26 Chomcyznski P & Sacchi N (1987) Single step method of RNA isolation by acid guanidinium thiocyanate–phe- nol–chloroform extraction. Anal Biochem 162, 156–159. 27 Towbin H, Staehelin T & Gordon J (1979) Electropho- retic transfer of proteins from polyacrylamide gels to nitrocellulose sheet: procedure and some applications. Proc Natl Acad Sci USA 76, 4350–4354. Supporting information The following supplementary material is available: Fig. S1. V5 ASGs release the inhibitory factor in Ringer’s solution and NaCl ⁄ P i . Fig. S2. Purification of PCD inhibitory factor. Fig. S3. Substrate specificity of B. mori GOD. Fig. S4. Effects of gluconolactone and hydrogen peroxide on 20E-induced PCD. This supplementary material can be found in the online version of this article. Please note: As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer-reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors. H. Matsui et al. Glucose oxidase prevents programmed cell death FEBS Journal 278 (2011) 776–785 ª 2011 The Authors Journal compilation ª 2011 FEBS 785 . Glucose oxidase prevents programmed cell death of the silkworm anterior silk gland through hydrogen peroxide production Hiroto Matsui 1 ,. arrowhead, (E,e) the middle silk gland, and (F,f) the posterior silk gland. (G,g) cutting surface of a cocoon. Glucose oxidase prevents programmed cell death H.