Myocyte enhancer factor 2 (MEF2) is a key modulator of the expression of the prothoracicotropic hormone gene in the silkworm, Bombyx mori Kunihiro Shiomi 1 , Yoshihiro Fujiwara 1 , Tsutomu Atsumi 1 , Zenta Kajiura 1 , Masao Nakagaki 1 , Yoshiaki Tanaka 2 , Akira Mizoguchi 3 , Toshinobu Yaginuma 4 and Okitsugu Yamashita 4,5 1 Faculty of Textile Science and Technology, Shinshu University, Nagano, Japan 2 National Institute of Agrobiological Sciences (NIAS), Ibaraki, Japan 3 Graduate School of Science, Nagoya University, Aichi, Japan 4 Graduate School of Bioagricultural Sciences, Nagoya University, Aichi, Japan 5 Chubu University, Aichi, Japan Organisms have adapted to seasonal fluctuations by evolving internal clocks and neuroendocrine systems to anticipate variations in living conditions [1]. In insects, prothoracicotropic hormone (PTTH) secretion appears to be triggered by a particular set of environ- mental signals, including the photoperiod and tem- perature [2–4]. PTTH stimulates the prothoracic glands to synthesize and release ecdysone, the steroid necessary for molting, metamorphosis, and the termin- ation of pupal diapause [2–4]. PTTH was first purified and sequenced from the silkworm, Bombyx mori [5,6]. The PTTH gene is constantly expressed during larval– pupal development [7], and the peptide is produced exclusively in two pairs of lateral PTTH-producing neurosecretory cells (PTPCs) in the brain [8]. From there it is transported via axons to the corpora allata and then released into the hemolymph. The PTTH titer in the hemolymph has been shown to correlate Keywords baculovirus; Bombyx mori; MEF2; metamorphosis and diapause; PTTH Correspondence K. Shiomi, Faculty of Textile Science and Technology, Shinshu University, Ueda, Nagano, 386-8567, Japan Fax: +81 268 21 5331 Tel: +81 268 21 5338 E-mail: shiomi@giptc.shinshu-u.ac.jp Database The sequences reported in this paper have been deposited in the DDBJ database under Accession no. AB121093. (Received 14 April 2005, revised 24 May 2005, accepted 31 May 2005) doi:10.1111/j.1742-4658.2005.04799.x Prothoracicotropic hormone (PTTH) plays a central role in controlling molting, metamorphosis, and diapause termination in insects by stimulating the prothoracic glands to synthesize and release the molting hormone, ecdysone. Using Autographa californica nucleopolyhedrovirus (AcNPV)- mediated transient gene transfer into the central nervous sytem (CNS) of the silkworm, Bombyx mori, we identified two cis-regulatory elements that participate in the decision and the enhancement of PTTH gene expression in PTTH-producing neurosecretory cells (PTPCs). The cis-element media- ting the enhancement of PTTH gene expression binds the transcription fac- tor Bombyx myocyte enhancer factor 2 (BmMEF2). The BmMEF2 gene was expressed in various tissues including the CNS. In brain, the BmMEF2 gene was expressed at elevated levels in two types of lateral neurosecretory cells, namely PTPCs and corazonin-like immunoreactive lateral neurosecre- tory cells. Overexpression of BmMEF2 cDNA caused an increase in the transcription of PTTH. Therefore, BmMEF2 appears to be particularly important in the brain where it is responsible for the differentiation of lat- eral neurosecretory cells, including the enhancement of PTTH gene expres- sion. This is the first report to identify a target gene of MEF2 in the invertebrate nervous system. Abbreviations AcNPV, Autographa californica nucleopolyhedrovirus; BmMEF2, Bombyx mori myocyte enhancer factor 2; CLI-LNCs, corazonin-like immunoreactive lateral neurosecretory cells; CNS, central nervous system; DIG, digoxigenin; EGFP, enhanced green fluorescence protein; MADS box, MCM1-Agamous-Deficiens-Serum response factor box; PTPCs, PTTH-producing neurosecretory cells; PTTH, prothoracicotropic hormone; SG, subesophageal ganglion. FEBS Journal 272 (2005) 3853–3862 ª 2005 FEBS 3853 closely with the ecdysteroid titer [3,9]. Fluctuations of PTTH titer in hemolymph consequentially act as a pacemaker in the neuroendocrine regulation of develo- pment by varying the secretion of ecdysone. The tim- ing of the increase in hemolymph PTTH titer on the day of wandering is photoperiodically controlled in B. mori [3]. In addition, in larvae of Heliothis vires- cens, expression of the PTTH gene declines sharply at the onset of larval wandering behavior and remains low during pupal diapause [10]. Thus, analysis of the molecular mechanisms controlling PTTH secretion in PTPCs is important for understanding the termination of pupal diapause as well as the induction of molting and metamorphosis. In the current study, we developed a convenient sys- tem for transiently transferring genes into the central nervous system (CNS) of B. mori using the recombin- ant baculovirus, AcNPV [11]. Using this system, we have been able to preferentially express the enhanced green fluorescence protein (EGFP) reporter gene under control of the PTTH promoter in PTPCs [11]. We used this system to investigate the molecular mecha- nisms controlling PTTH secretion by PTPCs. In the present report, we focused on the regulation of PTTH gene expression and found that the Bombyx myocyte enhancer factor 2 (BmMEF2) binds to the PTTH pro- moter and enhances its gene expression. Thus, the PTTH gene was identified as the first known target gene of MEF2 in the invertebrate nervous system. BmMEF2 appears to be particularly important in the brain where it causes the differentiation of lateral neuro- secretory cells by enhancing PTTH gene expression. Results Expression of the PTTH reporter gene is regulated by two cis-elements To determine the cis-elements participating in the regu- lation of PTTH gene expression, we performed repor- ter gene analysis using an AcNPV-mediated gene transfer system. We first examined whether the repor- ter gene construct containing EGFP under control of nucleotides )879 to +52 of the PTTH promoter (v[PT ⁄ EGFP]) [11] is expressed in the somata and neurites of PTPCs (Fig. 1). The fluorescence was localized within two pairs of lateral cells in the proto- cerebrum, and a faint signal was found in many cells throughout the brain lobes and at the midline in the subesophageal ganglion (SG) (Fig. 1A). The axons emanating from the somata of the two pairs of lateral cells extend towards the pars intercerebral with some arborization (Fig. 1A, box and Fig. 1B), run contralat- eral after crossing the pars intercerebral (Fig. 1A,B, arrow), and then project into the corpora allata with varicosites (Fig. 1F). Immunohistochemical staining with an anti-PTTH IgG to visualize endogenous PTTH produced by PTPCs identified two pairs of lateral cells in the protocerebrum (Fig. 1C) [8]. Merging Cy3 (anti- PTTH) with EGFP (v[PT ⁄ EGFP]) signals in the somata and axons of the cells (Fig. 1D,E) revealed many gran- ules on the cell surface, although most of the EGFP sig- nal was localized preferentially in the nucleus (Fig. 1E). Furthermore, the Cy3 signals in the PTPCs projected into the corpus allatum–corpus cardiacum complex where most of the signal overlapped with the EGFP signal (Fig. 1F). Thus, the neurosecretory cells in the brain of Bombyx expressing v[PT ⁄ EGFP]-derived EGFP corresponded to PTPCs. The results also sug- gest that the sequence of the PTTH promoter from nucleotides )879 to +52 contains cis-regulatory ele- ments that drive PTTH gene expression in PTPCs. We constructed six recombinant AcNPVs carrying different upstream regions of the PTTH gene fused with the EGFP reporter gene. EGFP fluorescence was observed in PTPCs (Fig. 1G–M). We also measured the fluorescence intensity in somata and compared it with the intensity of the recombinant AcNPV (v[PT ⁄ EGFP]) carrying nucleotides )879 to +52 of the PTTH promoter (n ¼ 22) (Fig. 1R). Progressive deletion of the 5¢-upstream region, either from nucleo- tides )208 to +52 or from )180 to +52, had no signi- ficant effect on EGFP expression (97.6 ± 4.0%, n ¼ 21 and 95.5 ± 7.7%, n ¼ 25, respectively; Fig. 1G–I). However, recombinant AcNPVs carrying nucleotides )167 to +52 and or )119 to +52 of the PTTH pro- moter caused an abrupt decrease in the expression of EGFP in the PTPCs (49.7 ± 19.0%, n ¼ 46 and 44.3 ± 16.9%, n ¼ 42, respectively; Fig. 1J,K). Using a recombinant AcNPV carrying nucleotides )105 to +52 of the PTTH promoter, EGFP expression was faint, and no fluorescence signal was observed in some pupa (9.9 ± 9.8%, n ¼ 37; Fig. 1L,L¢). No expression was observed when nucleotides )60 to +52 of the PTTH promoter were used (2.1 ± 4.1%, n ¼ 9; Fig. 1M), although faint signals on small cells were still detected in the lateral brain. Injection with recom- binant AcNPVs carrying nucleotides )879, )208, )180, )167, )119, )105, or )60 to +52 of the PTTH promoter resulted in hemolymph virus titers of 9.01, 7.42, 9.92, 8.64, 9.04, 9.15, and 10.15 · 10 6 pfuÆmL )1 , indicating that there was no significant difference in the ability of the various viral constructs to infect the pupae. In addition to injection of pupae with AcNPVs, we also examined the effect of injections into day 0 of Enhancement of PTTH gene expression by MEF2 K. Shiomi et al. 3854 FEBS Journal 272 (2005) 3853–3862 ª 2005 FEBS fifth instar larvae (Fig. 1N–Q). As in pupal brain, fluorescence due to injection of v[PT ⁄ EGFP] was observed in two pairs of lateral neurosecretory cells (Fig. 1N) that corresponded to the PTPCs (data not shown). In the larval brain infected with recombinant AcNPVs carrying nucleotides )180 to +52, )167 to +52, or )105 to +52 of the PTTH promoter, the relative fluorescence intensities of EGFP were 82.3 ± 47.2% (n ¼ 35) (Fig. 1O), 40.1 ± 15.7% (n ¼ 20) (Fig. 1P), and 5.7 ± 5.1% (n ¼ 20) (Fig. 1Q), respect- ively. Injection with recombinant AcNPVs carrying nu- cleotides )879, )180, )167, or )105 to +52 of the PTTH promoter resulted in hemolymph virus titers of 3.52, 5.01, 2.92 and 4.27 · 10 8 pfuÆmL )1 , indicating that there was no significant difference in the ability of the different viral constructs to infect the larvae. Thus, using EGFP reporter gene analysis and serial deletion of the PTTH promoter, we identified two potential cis- regulatory elements: (a) a 61 bp sequence from nucleo- tide )180 to )119 that participates in the enhancement of PTTH gene expression, and (b) a 15 bp sequence from nucleotide )119 to )105 that helps direct the expression of the PTTH gene expression in PTPCs. It appears that these two cis-elements are functionally conserved during larval–pupal development. The MEF2-binding sequence is important for enhancing PTTH gene expression To identify the trans-activating factors that enhance PTTH gene expression, we searched the 61 bp sequence from nucleotide )180 to )119 of the PTTH gene for transcription factor-binding sites using mat- inspector (http://www.genomatix.de/). As shown in Fig. 2A, we found that the DNA sequence bound by MEF2, C ⁄ TTA(A ⁄ T) 4 TAG ⁄ A [12], is conserved at the 5¢-upstream region from nucleotides )180 to )151 of the PTTH gene (CACAATGGTT CTATTTTAAG GATTTATCAC; MEF2 binding consensus underlined; Fig. 2A, wt). A gel-mobility shift assay using a 30-bp double- stranded oligonucleotide encoding nucleotides )180 to )151 of the PTTH promoter (Fig. 2A, wt) as a probe showed a shifted band (Fig. 2B, lane 1) that was pro- gressively lost upon incubation with increasing concen- trations of unlabeled wt oligonucleotide (Fig. 2B, lanes 2–4). We further synthesized three double-stranded oligonucleotides as competitors to analyze the sequence specificity of the protein bound to the wt oligonucleotide. In oligonucleotide M1, the MEF2 consensus binding sequence was disrupted by mutation A OL B r S G CA CA O L J I K FGH D C B E ML L' R -879 -208 -180 -167 -119 -105 -60 0 25 50 75 100 Relative fluorescence intensity (%) C o n s t r u c t s G H I J K L M P Q N O Fig. 1. Identification of cis-regulatory elements controlling PTTH gene expression in brain PTPCs of B. mori using AcNPV- mediated reporter gene analysis. Fluores- cence microscopy was used to visualize EGFP expression in the brain–SG complexes of larvae (N–Q) and pupae (A–M) injected with recombinant AcNPVs expressing v[PT ⁄ EGFP] carrying nucleotides )879 to +52 (A–G and N), )208 to +52 (H), )180 to +52 (I, O), )167 to +52 (J, P), )119 to +52 (K), )105 to +52 (L, L¢,Q),or)60 to +52 (M) of the PTTH gene. The axon emanating from the somata (light blue arrowheads and enlarged image shown in E) runs towards the midline of the brain with some arborization (boxed area in A), contralateral after crossing the midline (arrow in A–D), and then projects to the corpus allatum (F). Magnified images (B–E and G–Q) show the somata and axon indicated by the box in (A). In B-F, the PTPCs in a v[PT ⁄ EGFP]-injected pupae were exam- ined by immunohistochemistry with an anti-PTTH mAb (magenta). The EGFP fluores- cence was visualized by the green color. The relative fluorescence intensity of the PTPCs are shown as the percentage compared with v[PT ⁄ EGFP]-injected pupa (R). Br, brain; SG, subesophageal ganglion; CA, corpus allatum; OL, optic lobe. Scale bar ¼ 50 lm. K. Shiomi et al. Enhancement of PTTH gene expression by MEF2 FEBS Journal 272 (2005) 3853–3862 ª 2005 FEBS 3855 of 2 bp (Fig. 2A, M1). M2 contained the sequences of optimal targets for MEF2 expressed in mouse brain [12] (Fig. 2A, M2). In M3, 3 bp were mutated, but they are not within the MEF2 consensus binding sequence (Fig. 2A, M3). Even at a 100-fold excess, M1 was unable to compete the binding of protein to wt (Fig. 2B, lane 5). However, M2 eliminated protein binding to wt (Fig. 2B, lane 6); in fact, competition by M2 was stronger than with unlabeled wt (Fig. 2B; lanes 2–4). The shifted band also decreased in the pres- ence of M3 (Fig. 2B, lane 7) or anti-BmMEF2 (MADS) IgG (Fig. 2B, lane 10), but nonimmune serum had no effect (Fig. 2B, lane 9). Thus, we found that a protein in Bombyx brain bound to the MEF2 consensus binding sequence, and its binding was pre- vented by an antiserum that recognizes the MADS box of BmMEF2. We examined this further using recombinant AcNPVs carrying nucleotides )180 to +52 of the PTTH promoter and the M1, M2,orM3 sequence (Fig. 2A,C). Based on fluorescence intensity, the M3 virus was as effective (92.7 ± 5.4%, n ¼ 18; Fig. 2C, panel M3) at mediating EGFP expression as the wt virus (Fig. 2C, panel wt). The M2 virus resulted in an enhanced level of fluorescence (113.3 ± 10.1%, n ¼ 18; Fig. 2C, panel M2), and faint EGFP expression was observed with the M1 virus, which contains a disruption of the MEF2 binding consensus (36.4 ± 8.4%, n ¼ 18), although EGFP fluorescence was never completely eliminated by this construct (Fig. 2C, panel M1). Injection with recombinant AcNPVs wt, M1, M2, and M3 resulted in hemolymph virus titers of 8.96, 7.35, 9.68, and 6.66 · 10 6 pfuÆmL )1 , respectively, indicating that there was not a significant difference in the ability of the different virus constructs to infect the pupae. Thus, the expression of EGFP was altered by mutation of the MEF2 consensus binding sequence in the PTTH promoter, a region important for enhancing reporter gene expression. These findings suggest that the Bombyx MEF2 homolog binds to the MEF2 consensus binding sequence in the PTTH promoter, enhancing PTTH gene expression. Cloning of the Bombyx MEF2 (BmMEF2) cDNA We next cloned the MEF2 cDNA from the brain–SG complex in Bombyx using a PCR-based strategy with degenerate primers corresponding to the MADS-box and the MEF2 domain [13], regions that are highly conserved across a variety of organisms. A 2716-bp sequence containing the 5¢- and 3¢-untranslated regions of MEF2 (Accession no. AB121093) was obtained by RT-PCR and rapid amplification of cDNA ends. The open reading frame was from nucleotides +748 to A C B Fig. 2. Mutational analysis of the MEF2 consensus sequence (A) by gel mobility shift assay (B) and reporter gene analysis (C). The MEF2 consensus binding sequence in mouse brain [12] is boxed, and the 10 bp MEF2 core binding sequence is shown in capital letters. In the gel mobility shift assay (B), the double-stranded oligonucleotide encoding from )180 to )151 of the PTTH gene (wt) was use as a probe. Oligo- nucleotides M1, M2, and M3 were used as competitor DNAs. The mutated nucleotides are shown in bold. NS, normal rabbit serum; Ab, anti-BmMEF2 (MADS). The shifted band is indicated by an arrow. Reporter gene expression was performed using a recombinant AcNPV car- rying nucleotides )180 to +52 of the PTTH promoter and the nonmutated sequence (wt) or the M1, M2, or M3 mutant sequences. The somata of PTPCs are indicated by light blue arrowheads. Enhancement of PTTH gene expression by MEF2 K. Shiomi et al. 3856 FEBS Journal 272 (2005) 3853–3862 ª 2005 FEBS +1965 and encoded a predicted 404-amino acid pro- tein (Fig. 3A). A MADS box and an adjacent MEF2 domain are encoded within an 86-amino acid N-ter- minal sequence (Fig. 3A). These two regions are highly conserved in MEF2s from various organisms (Fig. 3B). The Bombyx sequence is most similar to that of Dro- sophila melanogaster (D-MEF2), with 96% amino acid sequence identity in the MADS box and MEF2 domain (Fig. 3B). Developmental expression of BmMEF2 We examined the developmental expression of BmMEF2 in various tissues during embryonic and postembryonic development by RT-PCR. BmMEF2 mRNA was first detected on day 3 after oviposition (Fig. 4A, lane 2) and was detected thereafter through- out embryogenesis, although the signal intensity of the hybridized band decreased on day 9 after oviposition (Fig. 4A, lane 4). During postembryonic development, BmMEF2 mRNA was detected in various tissues con- taining the brain–SG complex (Fig. 4A, lanes 5–15). Intense signals were detected in the mixture of integu- ment and muscle at both larval and pupal stages (Fig. 4A, lanes 9 and 15) as well as in the fat body at the pupal stage (Fig. 4A, lane 12). The PTTH mRNA was exclusively expressed in the brain–SG complex during postembryonic development (Fig. 4A, lanes 20– 30). During embryonic development, hybridized signals were detected from day 3 (Fig. 4A, lane 17), which corresponded to BmMEF2 expression (Fig. 4A, lane 2), although the signals were faint compared with those in larval and pupal brain–SG complexes. Next, we specifically examined the distribution of BmMEF2 mRNA in the CNS by RT-PCR (Fig. 4B). Although PTTH mRNA was detected exclusively in brain (Fig. 4B, lane 1), the BmMEF2 mRNA was detected in the SG and the first thoracic ganglion (T1) as well as in the brain (Fig. 4B, lanes 1–3). Furthermore, we determined the localization of BmMEF2 mRNA in brain by whole-mount in situ hybridization (Fig. 4C–I). Using an antisense BmMEF2 RNA as a probe, we observed hybridization throughout the brain, but it was particularly concen- trated in cells within the lateral region of the protocer- ebrum (Fig. 4C, blue box) and at the periphery of the tritocerebrum (Fig. 4C, red box). Signals were not detected when the sense strand RNA was used as a probe (Fig. 4D). In the tritocerebrum, there were intense hybridization signals that were reproducibly detected in  20 cells in each hemisphere (Fig. 4E). In contrast, in the lateral protocerebrum, the hybridiza- tion signals were relatively weak, and different num- bers of positive cells were observed among the 240 specimens (190 specimens with no positive cells, 26 with one positive cell, 12 with two positive cells, 9 with three positive cells, and 4 with four positive cells). Thus, in many specimens, hybridized signals in positive large cells of the lateral brain were similar to levels of neighboring cells. To identify the lateral cells, we performed immuno- histochemistry with the anti-PTTH IgG after in situ hybridization. In some specimens, there were two A B Fig. 3. Deduced amino acid sequence of the Bombyx MEF2 (A). The MADS box and the MEF2 domain are shown in red and blue, respectively. Alignment of the MADS box and the MEF2 domain of BmMEF2 with that of (abbreviations and Accessions nos shown in parentheses) Mus musculus MEF2A (mMef2a; U30823), Xenopus laevis Mef2a (xMef2a; BC046368), Homo sapiens MEF2A (hMEF2A; BC013437), Gallus gallus MEF2A (cMef2a; AJ010072), Danio rerio mef2a (zMef2a; BC044337), Cyprinus carpio MEF2A (CcMEF2A; AB012884), Caenorhabditis elegans mef-2 (Cemef-2; U36199), Podocoryne carnea Mef2 (PcMef2; AJ428495), Coturnix coturnix japonica qMEF2D (qMEF2D; AJ002238), Rattus norvegicus MEF2D (rMEF2D; AJ005425), Halocynthia roretzi MEF2 (As-MEF2; D49970), and Drosophila melanogaster D-MEF2 (D-MEF2; U07422) (B). Identical amino acids are indicated with *. K. Shiomi et al. Enhancement of PTTH gene expression by MEF2 FEBS Journal 272 (2005) 3853–3862 ª 2005 FEBS 3857 lateral cells showing BmMEF2 mRNA expression (Fig. 4F) that also were stained with anti-PTTH IgG (Fig. 4G). Moreover, in a few specimens, we found that a hybridization signal in a lateral cell corresponded to anticorazonin-immunoreactive cells (Fig. 4H,I), although four corazonin-like immuno- reactive lateral neurosecretory cells (CLI-LNCs) were found in the lateral region of the brain in each hemisphere [14]. Regulation of PTTH gene expression by controlling BmMEF2 expression To investigate whether the expression of BmMEF2 affects PTTH gene expression, we constructed two recombinant AcNPVs, v[PT ⁄ MEFs] and v[PT ⁄ MEFi], which were designed to overexpress and silence BmMEF2 mRNA under control of the PTTH promo- ter, respectively. When injected at 10 2 pfu per pupa, the virus titers in hemolymph for v[PT ⁄ EGFP], v[PT ⁄ MEFs], and v[PT ⁄ MEFi] were 5.12, 5.37, and 4.97 · 10 6 pfuÆmL )1 , respectively, and when injected at 10 6 pfu per pupa, the virus titers were 1.98, 1.18, and 1.57 · 10 7 pfuÆmL )1 . Therefore, we concluded that there was no significant difference in the ability of the different virus constructs to infect the pupae. Using RT-PCR, we first investigated the effect of infection with AcNPV on the amounts of mRNAs transcribed from the BmMEF2, PTTH, and actin A3 genes. When v[PT ⁄ EGFP] was injected at 10 2 pfu per pupa, we could clearly detect the BmMEF2 mRNA, and we could also detect it in noninjected pupae (Fig. 5A, lanes 1, 2). However, when v[PT ⁄ EGFP] was injected at 10 6 pfu per pupa, there was a slight decrease in the amount of the BmMEF2 and PTTH mRNA (Fig. 5A, lanes 3 and 10). Also, there were no changes in the actin A3 mRNA (Fig. 5A, lanes 15, 16, and 17). These results suggest that the AcNPV infection causes a decrease in the amount of both BmMEF2 and PTTH mRNA. When v[PT ⁄ MEFs] was injected, there was a higher level of BmMEF2 mRNA than in pupae that were not injected or that were injected with v[PT ⁄ EGFP] (Fig. 5A, lanes 4 and 5). Injection of v[PT ⁄ MEFi] at both 10 2 and 10 6 pfu per pupa caused a large reduc- tion of the BmMEF2 mRNA compared with non- injected and v[PT ⁄ EGFP]-injected pupa (Fig. 5A, lanes 6 and 7). Thus, the two recombinant AcNPVs, v[PT ⁄ MEFs] and v[PT ⁄ MEFi], were able to induce overexpression and suppression of the BmMEF2 gene, respectively. In addition, the amount of PTTH mRNA was also increased by injection with v[PT ⁄ MEFs] (Fig. 5A, lanes 11 and 12). However, v[PT ⁄ MEFi] did not cause elimination of the PTTH mRNA (Fig. 5A, lanes 13 and 14). Thus, BmMEF2 activated but was not essential for PTTH gene expression. F G I H BC D E ActA3 MEF2 Br SG T1 lane: 1 2 3 PTTH A 2h 3d 5d 9d BS MG FB SL IM BS MG FB OV TS IM MEF2 lane: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 ActA3 PTTH lane:1617181920 21222324252627 282930 lane: 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 Fig. 4. Developmental profiles of expression of BmMEF2 and PTTH genes. RT-PCR and Southern blot analysis (A, B) were performed during embryogenesis 2 h (2 h) and 3 (3d), 5 (5d), and 9 (9d) days after oviposition (Em) and on day 4 in fifth instar larvae (LV4) as well as on day 3 in pupae (P3). BS, brain–subesophageal ganglion complex; MG, midgut; FB, fat body; SL, silk gland; IM, integument and muscle; OV, ovary; TS, testis; Br, brain; SG, subesophageal ganglion; and T1, first thoracic ganglion. Whole-mount in situ hybridization was performed in pupal brain by using antisense (C, E–I) and sense (D) RNA of the BmMEF2 gene as probes. Magnified images of the periphery of the tritocerebrum (E) and lateral brain (F–I) are shown by the boxes in red and blue, respectively (C). The hybridized signals in lateral brain (F, H) were examined by immu- nohistochemistry with a monoclonal anti-PTTH IgG (magenta) (G) or an anti-corazonin IgG (green) (I). Scale bar ¼ 100 lm. PT/EGFP PT/MEFiPT/MEFs C 10 2 10 6 10 2 10 6 10 2 10 6 (PFU/pupa) lane: 1 2 3 4 5 6 7 MEF2 lane: 15 16 17 18 19 20 21 PTTH ActA3 lane: 8 9 10 11 12 13 14 Fig. 5. Effect of PTTH gene expression on the overexpression and silencing of the BmMEF2 gene. RT-PCR was performed on pupal brain injected each of three recombinant AcNPVs (v[PT ⁄ EGFP], v[PT ⁄ MEFs] and v[PT ⁄ MEFi]) at 10 2 and 10 6 pfu per pupa as well as noninjected pupal brain (c). Levels of BmMEF2 (lanes 1–7), PTTH (lanes 8–14), and ActinA3 (lanes 15–21) mRNAs were exam- ined. Enhancement of PTTH gene expression by MEF2 K. Shiomi et al. 3858 FEBS Journal 272 (2005) 3853–3862 ª 2005 FEBS Discussion We previously developed a system using AcNPV for transient gene transfer into the CNS of the silkworm, B. mori [11]. This system allows reporter gene analysis of many constructs, enabling the identification of the cis-elements in vivo. Furthermore, the system is highly reproducible and can be set up within 2 weeks of con- struction of the recombinant plasmids. In this study, we used this system to identify the cis-elements and a transcription factor responsible for expression of the PTTH gene in vivo. Within the PTTH promoter, we identified two cis-regulatory elements participating in (a) the decision to express the PTTH gene and (b) the enhancement of PTTH gene expression. Our results indicate that the 5¢-upstream region from nucleotides )119 to )105 of the PTTH gene participates in the decision to express the PTTH gene (Fig. 1). We analyzed the cis-regula- tory elements and a trans-activating factor participa- ting in the decision to express the PTTH gene. The 5¢-upstream region of the PTTH gene from )180 to )151 is similar to the MEF2 consensus bind- ing sequence of a variety of organisms. MEF2 belongs to the family of MADS box transcription factors, which bind to DNA as homo- and heterodimers through the consensus MEF2 binding sequence, C ⁄ TTA(A ⁄ T) 4 TAG ⁄ A [12]. This sequence is found in the upstream regions of numerous genes including muscle-specific genes, and plays a critical role in the differentiation of cells during the development of multicellular organisms [15]. There are four isoforms (A–D) of mammalian MEF2, and they have high homology within the 56-amino-acid MADS box at their N-termini and within an adjacent 29-amino-acid region referred to as the MEF2 domain. The MADS box is essential for DNA binding and dimerization, and the MEF2 domain plays an important role in DNA binding affinity as well as an indirect role in dimerization. The C-terminal portion of MEF2C is required for its transcriptional activation [13]. The N-terminal 86 amino acids of BmMEF2 are highly conserved and include a MADS box and a MEF2 domain. We found that BmMEF2 binds to the consensus sequence via its MADS box and can acti- vate transcription of the target gene in B. mori as well as MADS box-containing genes in various other organisms. Furthermore, the BmMEF2 gene is expressed in various tissues containing muscle and neural tissues in Bombyx as well as in D. melanogaster and various vertebrates. Consequently, correlation between the structure and gene expression profiles sug- gests that the BmMEF2 is a structural and functional analog of MEF2 proteins in various organisms. Fur- thermore, it has been speculated that BmMEF2 is responsible for the regulation of fundamental cellular processes in various tissues. In this study, we demonstrated that the MEF2 bind- ing sequences in the PTTH promoter enhance expres- sion of the EGFP reporter gene in PTPCs and are important for binding of the BmMEF2 protein. Fur- thermore, overexpression of the BmMEF2 gene can induce PTTH gene expression. Thus, it appears that BmMEF2 plays a role in the enhancement of PTTH gene expression in PTPCs. A single MEF2 gene, d-mef2, has been identified in D. melanogaster, and the isoforms of the D-MEF2 protein act as functional ana- logs of the vertebrate forms that participate in muscle differentiation [16,17]. Furthermore, D-MEF2 protein is expressed in Kenyon cells in the mushroom bodies of larval and adult brains, suggesting that these pro- teins are responsible for the differentiation of the Ken- yon cells and for morphogenesis of the mushroom body learning center [18]. However, the target genes for D-MEF2 have not been identified, and MEF2 functions have not been determined in the insect ner- vous system. Thus, our findings are the first identifica- tion of a gene that is a target of MEF2 in the invertebrate nervous system. Expression of the PTTH gene was first detected on day 3 of embryogenesis. In addition, Adachi-Yamada et al. [7] showed that it is constantly expressed during larval–pupal development. The correlation between PTTH and BmMEF2 gene expression suggests that BmMEF2 activates PTTH expression throughout embryonic and postembryonic development. We found that the BmMEF2 gene is preferentially expressed not only in PTPCs, but also in CLI-LNCs. In Manduca sexta [19] as well as in Bombyx [14], CLI- LNCs are identified as type Ia 1 neurosecretory cells. These cells coexpress PERIOD and various peptides, such as FMRFamide, and leu-enkephalin [20,21]. Furthermore, the genes of Antheraea pernyi, timeless, and period are also expressed exclusively in four pairs of cells in protocerebral lateral neurosecretory cells, which are likely type Ia 1 neurosecretory cells. The close anatomical localization between PTPCs and CLI- LNCs suggests that there are routes of communication between these two cell populations that may be important for the circadian control of PTTH release [22]. Although it is not known whether the two types of neurosecretory cells communicate, the BmMEF2 gene may be activated via a common specialized mech- anism in both PTPCs and CLI-LNCs and may thereby participate in the terminal differentiation processes of these lateral neurosecretory cells. K. Shiomi et al. Enhancement of PTTH gene expression by MEF2 FEBS Journal 272 (2005) 3853–3862 ª 2005 FEBS 3859 Aizono and Shirai suggested that muscarinic acetyl- choline receptor-induced signal transduction was involved in the control of PTTH release in B. mori [23]. Activation of phospholipase C and the subsequent activation of both protein kinase C and calmodulin- dependent kinase were essential in this signaling path- way. Furthermore, MEF2 is known to act as an endpoint for growth factor signaling pathways [24]. Although we identified BmMEF2 as a factor that enhances PTTH gene expression, BmMEF2 may parti- cipate in several other cellular processes that regulate PTTH secretion through signaling pathways. Thus, it will be important to further investigate the signal transduction pathway by which extracellular signals regulate insect functions including molting, metamor- phosis, and diapause. Experimental procedures Animals The polyvoltine strain, N4, of B. mori was used throughout these experiments. Eggs were incubated at 25 °C under con- tinuous darkness. Larvae were reared on an artificial diet (Silkmate-2M, Nosan Co., Yokohama, Japan) at 25–27 °C under a 12 h light ⁄ 12 h dark cycle. Larvae and pupae used in the experiments were collected within 1 h after each ecdysis (referred to as day 0) to synchronize their subse- quent development. Pupae were kept at 25 °C to allow adult development. Injection of recombinant AcNPV was performed according to Shiomi et al. [11]. Preparation of recombinant AcNPV Recombinant AcNPVs were prepared according to the manufacturer’s instructions (Invitrogen, Carlsbad, CA, USA) and Shiomi et al. [11]. For reporter gene analysis, six DNA fragments encoding the PTTH gene (Accession no. AB186492) promoter from nucleotides )208 to +52, )180 to +52, )167 to +52, ) 119 to +52, )105 to +52, and )60 to +52 were PCR-amplified from pPT ⁄ EGFP [11], which contains the PTTH gene promoter from nucleotides )879 to +52. The forward primers included a Sal I site, and the reverse primer included a NcoI site. PCR products were digested with SalI and NcoI and then inserted into pPT ⁄ EGFP lacking the promoter region of the PTTH gene. Recombinant plasmids were sequenced, and recom- binant AcNPVs were prepared according to the manu- facturer’s instructions. To create three mutants in the promoter region between nucleotides )180 and )151 of the PTTH gene (Fig. 2), we performed PCR amplification using the same reverse primer described above and one of three forward primers, each of which encoded a SalI site. The titers of budded virions were determined using the BD BacPAK Baculovirus Rapid Titer Kit (BD Biosciences, Palo Alto, CA, USA). Reporter gene analysis Four days after injection with recombinant AcNPV, the brain–SG complex of larvae and pupae was dissected out in NaCl ⁄ P i and mounted onto a hole-slide glass with 1 : 4 Fluoroguard Antifade reagent (Bio-Rad, Hercules, CA, USA) in NaCl ⁄ P i . EGFP fluorescence was detected using an ECLIPSE E600 microscope (Nikon Co., Tokyo, Japan) equipped with a DP50CU digital camera (Olympus Co., Tokyo, Japan). Digital images of the brain–SG complex were scanned using view finder lite, version 1.0 (Pixera Co., Los Gatos, CA, USA) at a sensitivity of 400 and an exposure of 1 ⁄ 15 s. Using NIH image 1.62 (http://www.rsb.info.nih.gov/ nih-image/), the relative fluorescence intensity of the PTPCs was determined as the intensity of the individual cells relative to the mean pixel fluorescence for the entire somata (S) of the brain. Fluorescence images were converted to grayscale and inverted into black and white images. An area adjacent to the area of interest (A) and an area from an image lacking a spe- cimen (N) were scanned as the background signals. When PTPCs were not visible, the focal plane was adjusted to faintly signals on small cells in the same field. The relative fluorescence intensity was calculated as follows: Relative fluorescence intensity (%) ¼ 100 · ([(S) – (A) – (N)] ⁄ [(A) – (N)]) for the virus of interest ⁄ ([(S) – (A) – (N)] ⁄ [(A) – (N)]) for the virus carrying nucleotides )879 to +52 of the PTTH promoter (v[PT ⁄ EGFP]) [11]. In situ hybridization and immunohistochemistry In situ hybridization was performed as described by Sato et al. [25] with some modifications. The procedures prior to proteinase K treatment were adapted from Shiomi et al. [11]. Brain–SG complex was treated for 5 min with 10 lgÆmL )1 proteinase K (Roche, Indianapolis, IN, USA) and then hybridized with digoxigenin (DIG)-labeled sense and antisense RNA probes, respectively. The DIG-labeled RNA probes were prepared with a DIG RNA labeling kit (Roche) using BmMEF2 cDNA as a template. BmMEF2 cDNA encoding from nucleotides +654 to +854 (Acces- sion no. AB121093) was amplified by PCR and inserted into the pCR-XL-TOPO vector (Invitrogen) in sense and antisense directions from the T7 promoter. DIG-labeled RNA was detected with an alkaline phosphatase-conjugated anti-DIG IgG using a DIG nucleic acid detection kit (Roche). For immunohistochemistry, we used an anti- PTTH monoclonal IgG (3E5mAb) [8] and an anti-corazo- nin rabbit polyclonal IgG [14]. The immunoreaction procedures were adapted from Shiomi et al. [11]. EGFP fluorescence and anti-PTTH immunofluorescent staining were detected using a Radiance 2000 confocal microscope Enhancement of PTTH gene expression by MEF2 K. Shiomi et al. 3860 FEBS Journal 272 (2005) 3853–3862 ª 2005 FEBS (Bio-Rad). Images were adjusted and assembled in Adobe photoshop cs (Adobe systems Inc., San Jose, CA, USA). Gel-mobility shift assay Cell extract was prepared from a mixture of the brain–SG complex from day 1, 3, and 5 pupae according to Ueda and Hirose [26] with some modifications. A double-stran- ded synthetic oligonucleotide corresponding to the PTTH promoter encoding nucleotides )180 to )151 (Fig. 2) was end-labeled with T4 polynucleotide kinase and [ 32 P]ATP[c P] and then used as a probe. Incubation and electrophoresis were performed according to Ueda and Hirose [27]. The BmMEF2 (MADS) antibody (Qiagen, Valencia, CA, USA) was generated by immunizing rabbits with a peptide enco- ding the 15 N-terminal amino acids of BmMEF2. Cloning of the B. mori MEF2 (BmMEF2) cDNA Poly(A) + RNA was directly purified from brain–SG complex of day 3 pupae using Dynabeads oligo(dT) 25 (Dynal Biotech LLC., Brown Deer, WI, USA). RT-PCR was performed using degenerate primers based on the sequences of the MADS box and the MEF2 domain [13] common to several organisms (Accession nos AB01288, U66569, AJ005425, BC011070, BC040949, AJ002238, U66570, Z19124, X83527, D49970, and U36198): 5¢-CAGGTGACCTTYAMCA- ARMG-3¢ (forward) and 5¢-TCRTGDGGYTCRTTR- TAYTC-3¢ (reverse). The full-length cDNA sequence was determined using a SMART RACE cDNA amplification kit (Clontech, Mountain View, CA, USA). Finally, the full- length BmMEF2 cDNA (Accession no. AB121093) was amplified by RT-PCR. RT-PCR and Southern hybridization Eggs were collected 2 h and 3, 5, and 9 days after oviposi- tion. Various tissues were dissected from day 4 fifth instar larvae and day 3 pupae. Total RNAs were extracted from eggs and various tissues using TRIzol reagent (Invitrogen) and then subjected to poly(A) + RNA purification using Dynabeads oligo(dT) 25 (Dynal). Poly(A) + RNA from the brain–SG complex was directly purified using Dynabeads Oligo (dT) 25 (Dynal). First-strand DNA was synthesized using a SMART RACE amplification kit (Clontech). PCR amplification was carried out on mRNAs for BmMEF2, PTTH, and actin A3. The BmMEF2 cDNA was amplified from nucleotides +654 to +2685 (Accession no. AB121093), the PTTH cDNA from +34 to +708 (Acces- sion no. D90082), and the actin A3 cDNA from +70 to +498 (Accession no. U49854). PCR products were subjec- ted to electrophoresis, transferred to Hybond-N + nylon membranes (Amersham, Little Chalfont, Bucks, UK), and then hybridized with the 32 P-labeled internal oligonucleo- tides encoding nucleotides +908 to +937 of BmMEF2, +252 to +275 of PTTH, or +308 to +331 of actin A3. Overexpression and RNA interference for BmMEF2 mRNA Two recombinant AcNPVs were constructed for over- expression (v[PT ⁄ MEFs]) or silencing (v[PT ⁄ MEFi]) of the BmMEF2 gene. To obtain v[PT ⁄ MEFs], the BmMEF cDNA corresponding to the open reading frame was inser- ted downstream of the PTTH promoter. The PCR product of the BmMEF2 cDNA was ligated to the recombinant plasmid pPT ⁄ EGFP [11] after excision of the EGFP cDNA by digestion with NcoI and XhoI. To obtain the v[PT ⁄ MEFi], we constructed two inverted repeat DNAs corres- ponding to the 1.2-kbp BmMEF2 cDNA fragment from nucleotides +748 to +1934. These were inserted down- stream of the PTTH promoter with 200-bp spacer sequences consisting of the intron sequence (nucleotides +369 to +568) of the DH-PBAN gene [28] as described by Giordano et al. [29]. Acknowledgements This research was funded by grants from the Research for the Future Program from the Japan Society for the Promotion of Science (JSPS-RFTF99L01203). Addi- tional support was provided by Grants-in-Aid (17688003 and 17658027) from the Ministry of Educa- tion, Science, Sports and Culture of Japan. We are also indebted to the Division of Gene Research, Research Center for Human and Environmental Sci- ences, Shinshu University, for providing the facilities for these studies. References 1 Malpaux B, Migaud M, Tricoire H & Chemineau P (2001) Biology of mammalian photoperiodism and the critical role of the pineal gland and melatonin. J Biol Rhythms 16, 336–347. 2 Truman JW & Riddiford LM (1974) Physiology of insect rhythms. 3. The temporal organization of the endocrine events underlying pupation of the tobacco hornworm. J Exp Biol 60, 371–382. 3 Mizoguchi A, Dedos SG, Fugo H & Kataoka H (2002) Basic pattern of fluctuation in hemolymph PTTH titers during larval–pupal and pupal–adult development of the silkworm, Bombyx mori. Gen Comp Endocrinol 127, 181–189. 4 Bollenbacher WE & Granger NA (1985) Endocrinology of the prothoracicotropic hormone. In Comprehensive Insect Physiology, Biochemistry and Pharmacology K. Shiomi et al. Enhancement of PTTH gene expression by MEF2 FEBS Journal 272 (2005) 3853–3862 ª 2005 FEBS 3861 (Kerkut GA & Gilgert LI, eds), Vol. 7, pp. 109–151. Pergamon Press, Oxford. 5 Kataoka H, Nagasawa H, Isogai A, Ishizaki H & Suzuki A (1991) Prothoracicotropic hormone of the silkworm, Bombyx mori: amino acid sequence and dimeric structure. Agric Biol Chem 55, 73–86. 6 Kawakami A, Kataoka H, Oka T, Mizoguchi A, Kimura-Kawakami M, Adachi T, Iwami M, Nagasawa H, Suzuki A & Ishizaki H (1990) Molecular cloning of the Bombyx mori prothoracicotropic hormone. Science 247, 1333–1335. 7 Adachi-Yamada T, Iwami M, Kataoka H, Suzuki A & Ishizaki H (1994) Structure and expression of the gene for the prothoracicotropic hormone of the silkmoth Bombyx mori. Eur J Biochem 220, 633–643. 8 Mizoguchi A, Oka T, Kataoka H, Nagasawa H, Suzuki A & Ishizaki H (1990) Immunohistochemical localiza- tion of prothoracicotropic hormone-producing neuro- secretory cells in the brain of Bombyx mori. Dev Growth Differ 32, 591–598. 9 Mizoguchi A, Ohashi Y, Hosoda K, Ishibashi J & Kataoka H (2001) Developmental profile of the changes in the prothoracicotropic hormone titer in hemolymph of the silkworm Bombyx mori: correlation with ecdyster- oid secretion. Insect Biochem Mol Biol 31, 349–358. 10 Xu WH & Denlinger DL (2003) Molecular characteriza- tion of prothoracicotropic hormone and diapause hor- mone in Heliothis virescens during diapause, and a new role for diapause hormone. Insect Mol Biol 12, 509–516. 11 Shiomi K, Kajiura Z, Nakagaki M & Yamashita O (2003) Baculovirus-mediated efficient gene transfer into the central nervous system of the silkworm, Bombyx mori. J Insect Biotechnol Sericol 72, 149–155. 12 Andres V, Cervera M & Mahdavi V (1995) Determina- tion of the consensus binding site for MEF2 expressed in muscle and brain reveals tissue-specific sequence con- straints. J Biol Chem 270, 23246–23249. 13 Molkentin JD, Black BL, Martin JF & Olson EN (1996) Mutational analysis of the DNA binding, dimeri- zation, and transcriptional activation domains of MEF2C. Mol Cell Biol 16, 2627–2636. 14 Roller L, Tanaka Y & Tanaka S (2003) Corazonin and corazonin-like substances in the central nervous system of the Pterygote and Apterygote insects. Cell Tissue Res 312, 393–406. 15 McKinsey TA, Zhang CL & Olson EN (2002) MEF2: a calcium-dependent regulator of cell division, differentia- tion and death. Trends Biochem Sci 27, 40–47. 16 Lilly B, Galewsky S, Firulli AB, Schulz RA & Olson EN (1994) D-MEF2: a MADS box transcription factor expressed in differentiating mesoderm and muscle cell lineages during Drosophila embryogenesis. Proc Natl Acad Sci USA 91, 5662–5666. 17 Nguyen HT, Bodmer R, Abmayr SM, McDermott JC & Spoerel NA (1994) D-mef2:aDrosophila mesoderm- specific MADS box-containing gene with a biphasic expression profile during embryogenesis. Proc Natl Acad Sci USA 91, 7520–7524. 18 Schulz RA, Chromey C, Lu MF, Zhao B & Olson EN (1996) Expression of the D-MEF2 transcription in the Drosophila brain suggests a role in neuronal cell differ- entiation. Oncogene 12, 1827–1831. 19 Copenhaver PF & Truman JW (1986) Metamorphosis of the cerebral neuroendocrine system in the moth Manduca sexta. J Comp Neurol 249, 186–204. 20 Homberg U, Davis NT & Hildebrand JG (1991) Pep- tide-immunocytochemistry of neurosecretory cells in the brain and retrocerebral complex of the sphinx moth Manduca sexta. J Comp Neurol 303, 35–52. 21 Wise S, Davis NT, Tyndale E, Noveral J, Folwell MG, Bedian V, Emery IF & Siwicki KK (2002) Neuroanato- mical studies of period gene expression in the hawk- moth, Manduca sexta. J Comp Neurol 447, 366–380. 22 Sauman I & Reppert SM (1996) Molecular characteriza- tion of prothoracicotropic hormone (PTTH) from the giant silkmoth Antheraea pernyi: developmental appear- ance of PTTH-expressing cells and relationship to circa- dian clock cells in central brain. Dev Biol 178, 418–429. 23 Aizono Y & Shirai Y (1995) Control of prothoracicotro- pic hormone release with neurotransmitters in silkworm. In Molecular Mechanisms of Insect Metamorphosis and Diapause (Suzuki A, Kataoka H & Matsumoto S, eds), pp. 119–128. Industrial Publishing & Consulting, Tokyo. 24 Cavanaugh JE (2004) Role of extracellular signal regu- lated kinase 5 in neuronal survival. Eur J Biochem 271, 2056–2059. 25 Sato Y, Ikeda M & Yamashita O (1994) Neurosecretory cells expressing the gene for common precursor for dia- pause hormone and pheromone biosynthesis-activating neuropeptide in the suboesophageal ganglion of the silk- worm, Bombyx mori. Gen Comp Endocrinol 96, 27–36. 26 Ueda H & Hirose S (1990) Identification and purifica- tion of a Bombyx mori homologue of FTZ-F1. Nucleic Acids Res 18, 7229–7234. 27 Ueda H & Hirose S (1991) Defining the sequence recog- nized with BmFTZ-F1, a sequence specific DNA bind- ing factor in the silkworm, Bombyx mori, as revealed by direct sequencing of bound oligonucleotides and gel mobility shift competition analysis. Nucleic Acids Res 19, 3689–3693. 28 Xu WH, Sato Y, Ikeda M & Yamashita O (1995) Mole- cular characterization of the gene encoding the precur- sor protein of diapause hormone and pheromone biosynthesis activating neuropeptide (DH-PBAN) of the silkworm, Bombyx mori and its distribution in some insects. Biochim Biophys Acta 1261, 83–89. 29 Giordano E, Rendina R, Peluso I & Furia M (2002) RNAi triggered by symmetrically transcribed transgenes in Drosophila melanogaster. Genetics 160, 637–648. Enhancement of PTTH gene expression by MEF2 K. Shiomi et al. 3862 FEBS Journal 272 (2005) 3853–3862 ª 2005 FEBS . MADS box at their N-termini and within an adjacent 29 -amino-acid region referred to as the MEF2 domain. The MADS box is essential for DNA binding and dimerization, and the MEF2 domain plays an. Xenopus laevis Mef 2a (xMef 2a; BC046368), Homo sapiens MEF 2A (hMEF 2A; BC013437), Gallus gallus MEF 2A (cMef 2a; AJ0100 72) , Danio rerio mef 2a (zMef 2a; BC044337), Cyprinus carpio MEF 2A (CcMEF 2A; AB0 128 84),. cells throughout the brain lobes and at the midline in the subesophageal ganglion (SG) (Fig. 1A) . The axons emanating from the somata of the two pairs of lateral cells extend towards the pars intercerebral