A novel Rel protein and shortened isoform that differentially regulate antibacterial peptide genes in the silkworm Bombyx mori

Biochimica et Biophysica Acta 1730 (2005) 10 – 21 http://www.elsevier.com/locate/bba

A novel Rel protein and shortened isoform that differentially regulate antibacterial peptide genes in the silkworm Bombyx mori i Hiromitsu Tanaka a, Masafumi Yamamoto b, Yuko Moriyama b, Masafumi Yamao b, Seiichi Furukawa a, Aki Sagisaka a, Hiroshi Nakazawa a, Hajime Mori b, Minoru Yamakawa a,c,* a

Innate Immunity Laboratory, National Institute of Agrobiological Sciences, Owashi 1-2, Tsukuba, Ibaraki 305-8634, Japan b Insect Biomedical Research Center, Kyoto Institute of Technology, Matsugasaki, Sakyo-ku, Kyoto 606-8585, Japan c Institute of Agriculture and Forestry, University of Tsukuba, Tsukuba, Ibaraki 305-8572, Japan Received 13 January 2005; received in revised form 4 May 2005; accepted 19 May 2005 Available online 20 June 2005

Abstract Two cDNAs encoding novel Rel proteins were cloned from the silkworm, Bombyx mori. These cDNA clones (BmRelA and BmRelB) showed identical nucleotide sequences except for the 5V-region. BmRelB cDNA derived probably from an alternatively spliced mRNA lacked 241 bp nucleotides at the 5V-region of the BmRelA cDNA, resulting in a loss of the first 52 amino acids. Expression of antibacterial peptide genes was strongly inhibited upon infection with Micrococcus luteus in transgenic silkworms in which BmRel gene expression was knocked down, suggesting that these two Rel proteins are involved in activation of antibacterial peptide genes. Co-transfection experiments indicated that BmRelB activated the Attacin gene strongly and other genes to a lesser extent, whereas BmRelA activated Lebocin 4 gene strongly and Attacin and Lebocin 3 genes very weakly. The Rel homology domain of BmRelA and BmRelB was shown to bind specifically to nB sites of antibacterial peptide genes. Proline-rich domains of the BmRels were necessary for activation of antibacterial peptide genes. These results illustrate that a minor structural change in Rel proteins can provoke a dramatic differential activation of antibacterial peptide genes, suggesting a novel regulatory mechanism for insect antibacterial peptide gene expression. D 2005 Elsevier B.V. All rights reserved. Keywords: Rel protein; Insect immunity; Antibacterial peptide; Transgenic silkworm; Gene knockdown; Differential transcriptional regulation

1. Introduction The innate immune system in insects and mammals shares structurally and functionally related factors for elimination of invading microbes. Toll homologues and the components involved in signal transduction of immunerelated protein gene expression, such as TAK, FADD and Myd88, are known to have significant conservation between

i The nucleotide sequences reported in this paper have been deposited in the GSDB/DDBJ/EMBL/NCBI nucleotide sequence databases (accession no. AB096087 for BmRelA and AB096088 for BmRelB). * Corresponding author. Innate Immunity Laboratory, National Institute of Agrobiological Sciences, Owashi 1-2, Tsukuba, Ibaraki 305-8634, Japan. Tel.: +81 29 838 6154; fax: +81 298 38 6028. E-mail address: [email protected] (M. Yamakawa).

0167-4781/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.bbaexp.2005.05.007

Drosophila melanogaster and mammals [1]. Rel/NFnB molecules, transcription factors that bind to nB sites and control immune related genes, are also remarkably conserved in Drosophila and mammals [2]. A common feature of Rel/NFnB molecules is that all proteins have a wellconserved domain, so-called Rel homology domain (RHD), involved in DNA binding, dimerization and interaction with Inhibitor nB (InB) [3]. In Drosophila, three mammalian Rel/NF-nB homologues that control antibacterial and antifungal peptide genes have been identified. Two Rel/ NF-nB proteins, Dorsal [4] and Dorsal-related immunity factor (Dif) [5], are activated by the Toll pathway in response to infection with fungi and Gram-positive bacteria. These factors are localized in the cytoplasm and interact with the mammalian InB homologue Cactus in unstimulated cells. In response to an infection, Dif and Dorsal translocate

H. Tanaka et al. / Biochimica et Biophysica Acta 1730 (2005) 10 – 21

into the nucleus by dissociating from Cactus, and activate antifungal peptide genes such as drosomycin [6]. On the other hand, another Rel/NF-nB protein, Relish, is activated by the IMD pathway in response to infection with Gramnegative bacteria [7,8]. Relish comprises the N-terminal Rel homology domain (RHD) and C-terminal ankyrin repeat domain. N-terminal fragment including RHD of Relish is released by endoproteolytic cleavage in response to bacterial infection, and translocates from the cytoplasm to the nucleus activating antibacterial peptide genes such as diptericin [9]. Recently, several Rel/NF-nB homologues from other insects have been cloned and characterized. According to structural features, insect Rel/NF-nB protein can be categorized into two types, Dif-Dorsal type or Relish type. Gambif1 from Anopheles gambiae [10] and A.d.RelA from Allomyrina dichotoma [11] belong to the Dif-Dorsal type because these proteins do not have ankyrin repeat domains and their RHDs have much higher homology with Dorsal than Relish. On the contrary, Relish from Aedes aegypti [12] and SRAM, a 59-kDa nB-binding protein from Sarcophaga peregrina [13], are identified as Relish-type proteins because of the Cterminal ankyrin-repeat domain. In spite of the detailed functional analysis of Rel proteins in the activation of antibacterial and antifungal peptide genes in D. melanogaster [14], their physiological role in other insects still remains obscure. Several antimicrobial peptide genes such as CecropinA [15], CecropinB [16], Attacin [17], Lebocin [18] and Moricin [19] have been identified from Bombyx mori. These genes contain conserved nB motifs in the 5V-upstream region and nuclear factors that bind to nB motif were found by electrophoresis mobility shift assay (EMSA) [20], suggesting the Rel/NFnB homologues also control antimicrobial peptide genes in B. mori. However, the silkworm immune system has not been analyzed well because the lack of transgenic lines with knockout or knockdown-immune genes has been a serious limitation for studying silkworm immunity. We have recently developed a novel method to establish transgenic silkworms, B. mori, using Autographa californica nucleopolyhedrosis virus and piggyBac transposable elements [21]. On the other hand, gene silencing caused by RNA interference has been successfully achieved in B. mori [22]. Taking advantage of the combination of these techniques, we attempted Rel gene silencing in transgenic B. mori to analyze the role of Rel proteins in the activation of antibacterial peptide genes. In this paper, cloning and functional analysis of two novel Rel protein cDNAs from the silkworm, B. mori, are reported. The Rel protein and a shortened isoform likely produced as a result of alternative splicing are structurally identical except for the N-terminal region. The results of functional analysis of these Rel proteins showed that gene expression of some antibacterial peptides were suppressed in BmRel-knockdown transgenic silkworms infected with Micrococcus luteus. The results also indicated that a minor structural difference in the Rel proteins can cause a

11

significant functional difference in regulation of antibacterial peptide gene expression. This suggests a novel regulatory mechanism for insect immune gene expression.

2. Materials and methods 2.1. Insects B. mori (C602 strain) was reared on an artificial diet (Nihonnosanko) in a rearing room at 25 -C under a controlled environment (11 h light and 13 h dark). Fifth instar larvae were used in the experiments. 2.2. cDNA cloning and nucleotide sequencing Total RNA was prepared from the fat body of B. mori larvae and first strand cDNA synthesized as described previously [11]. Forward and reverse degenerate primers to clone B. mori Rel proteins by reverse transcriptase-polymerase chain reactions (RT-PCR) were designed on the basis of conserved RHD sequences described by Barillas-Mury et al. [10]; 5V-GITT(A/C)GITA(T/C)GA(A/G)TG(T/C)GA(A/ G)GG(A/G/C/T)(A/C)G-3V (forward primer) and 5V-AT(A/ G)TCTTC(T/C)TTI(T/G)(G/C)IA(T/C)(T/C)TTITC(A/ G)CA-3V (reverse primer), where I denotes inosine. A 584 base pair (bp) fragment was obtained. 3V-rapid amplification of cDNA ends (3V-RACE) was performed using the RT-PCR fragment as a template, a forward primer, 5V-GACAAGAAGGCCATGAGCGA-3V (F1) and the NotI-(dT18) primer (Amersham Pharmacia Biotech)(reverse primer). An 1177 bp PCR product was subcloned. 5V-RACE was conducted using a 5V-RACE System for Rapid Amplification of cDNA End Reagent Assembly, Ver. 2 (Gibco BRL) with the following primers: 5V-ATCCTTGTTGACGCAGGACA-3V, 5V-GATAGAGCATATCTTGATTG-3V and 5VTGTAGGATAAGTTTTGTTCT-3V. Two kinds of 5V cDNA fragments and a 3V cDNA fragment were isolated. 2.3. Detection of BmRelA and BmRelB mRNA The two 5V cDNA fragments were confirmed to have the same 3V nucleotide sequences by RT-PCR using a combination of a forward primer 5V-CATCGAGGATTTACGCACAC-3V (corresponding to positions 22 to 41) and a reverse primer 5V-TACTGTCGGCCTCAACGTTGAGGT-3V (corresponding to positions 2277 to 2300) or a combination of a forward primer 5V-TCGGACAAGCGAGAGCGAAGAGA3V (corresponding to positions 278 to 300) and a reverse primer 5V-TACTGTCGGCCTCAACGTTGAGGT-3V (corresponding to positions 2277 to 2300) under the following conditions. The reaction mixture including 10 pmol of forward and reverse primers and first strand cDNA corresponding to 0.1 Al of total RNA was kept at 95 -C for 1 min, then 40 cycles of PCR (95 -C for 1 min, 55 -C for 2 min, 72 -C for 3 min) were done. Amplified fragments

12

H. Tanaka et al. / Biochimica et Biophysica Acta 1730 (2005) 10 – 21

were separated by agarose gel electrophoresis and nucleotide sequences of each fragment determined. 2.4. BmRel-knockdown transgenic silkworm Fragments (703 bp) of BmRel gene and 600 bp in the opposite direction of the BmRel gene were amplified from BmRelA cDNA. The primer combination was as follows: 5V-AAAACTGCAGATGGACATCGGCTTAGACGC-3V (forward primer) and 5V-AAAACTGCAGCGTAGATGATGTCGGACACG-3V (reverse primer) for the BmRel gene; 5V-AAAAAGATCTATGGACATCGGCGGAGACGC-3V (forward primer); and 5V-AAAACTGCAGGCTCTGCGTGTGCGAGCGGT-3V (reverse primer) for the BmRel gene encoding opposite directions. These genes produced duplication so as to link the two arms of the inverted repeat. The BmRel gene was designed 103 bp longer than the BmRel gene encoding opposite directions. B. mori cytoplasmic actin A3 promoter fragments were inserted between the EcoT 221 and PstI sites of the pAcpigEGFP vector [21]. The BmRel fragment and SV40 poly A signal fragment were inserted between the BglII and PstI sites, and BglII site of pAcpigEGFP, respectively, resulting in a recombinant transfer vector pAcpigRel. The transfer vector was used to yield a recombinant baculovirus AcpigRel and then the recombinant virus was used as the transformation vector for establishment of transgenic lines as described previously [21]. 2.5. Real-time RT-PCR Primers were designed on the basis of GenBank database accession numbers M24370 for SP2, S78369 for Attacin, D11113 for Cecropin B1, AB003035 for Lebocin 3 and AB019538 for Lebocin 4. To detect BmRelA or BmRelB specifically, a combination of the forward primer 5V-ACGTGTGTTTCCATCCGCGAAGT-3V, and reversed primer 5VCACGAACCCGAAACGCACATT-3V, or a combination of the forward primer 5V-GCAGTCGTTCTTCGTGTGACATCG-3V or reverse primer 5V-CGGCGTCTCCAGCGTGAATG-3V, respectively, were used. Synthesis of cDNA from total RNA was carried out with a First-Strand cDNA Synthesis Kit (Amersham Bioscience). Fifth instar larvae (day 4) from normal and 3rd transgenic generations were inoculated with M. luteus (2104 cells) and total RNA was extracted from the fat body 2 h postinoculation with ISOGEN (Nippon Gene). Total RNA was heated at 65 -C for 10 min and then placed on ice for 2 min. Eleven Al of bulk first-strand cDNA reaction mixture from the FirstStrand cDNA Synthesis Kit (Amersham Bioscience), 1 Al of 200 mM DTT solution, 1 Al of pd(N)6 primer (0.2 Ag/Al) mixture and 2.5 Ag of heat-denatured RNA were added to a sterile 1.5 ml microcentrifuge tube to synthesize the firststrand cDNA, and incubated at 37 -C for 60 min. Real-time PCR was carried out in 96-well plates with a 25 Al reaction volume containing 12.5 Al of 2SYBR Green Master Mix

(PE Applied Biosystems), a total of 21 ng cDNA and 0.2 AM of each forward and reverse primer. Triplicate samples were subjected to denaturation at 94 -C for 10 s, followed by 40 cycles of amplification (94 -C for 10 s and 60 -C for 31 s) using ABI Prism 7000 sequence detection system. Sequence-specific amplification was detected as an increase in the fluorescent signal of SYBER green during the amplification cycles. Expression was normalized against SP2 as an endogenous reference, because SP2 is a major protein in the silkworm fat body. 2.6. Construction of expression vectors The 5V-upstream regulatory regions of B. mori antibacterial peptide genes for Attacin [23], Cecropin B1 [20], Lebocin 3 [18] and Lebocin 4 [18] were amplified by PCR. The regions were as follows: Attacin ( 334 to +38), Cecropin B1 ( 309 to +74), Lebocin 3 ( 585 to 1) and Lebocin 4 ( 585 to 1), where +1 denotes the transcription initiation site in the case of Attacin and Cecropin B1, and the translation initiation site in the case of Lebocin 3 and Lebocin 4. Each PCR fragment was ligated to the pGL3-basic vector containing the luciferase gene (Promega). The coding regions of cDNAs encoding BmRelA (nucleotide 511 to 2369, see data base accession number AB096087) and BmRelB (649 to 2369) were subcloned into KpnI and NotI sites of the pPac-PL vector containing D. melanogaster actin 5C promoter [24] to construct pPacBmRelA and pPacBmRelB. To construct pPacBmRelAA and pPacBmRelBA (see Fig. 8), pPacBmRelA and pPacBmRelB were digested with AseI and then treated with Mung bean nuclease (Nippon Gene) followed by KpnI digestion. Recovered fragments (511 to 1635 for BmRelA and 649 to 1635 for BmRelB) were subcloned into EcoRI and KpnI sites of pBluescript II KS (+) plasmid (Stratagene), in which the EcoRI site was blunted by a Klenow fragment (Nippon Gene). The fragments digested with KpnI and NotI were subcloned into KpnI and NotI sites of pPac-PL vector. pPacBmRelB was digested with PstI, and the ends were blunted with T4 DNA polymerase (Nippon Gene) followed by KpnI digestion to construct pPacBmRelBP (see Fig. 8). The fragment was ligated into BamHI and KpnI sites of pBluescript II KS (+), in which BamHI site was treated with Klenow fragment and Mung bean nuclease. The fragment digested with KpnI and NotI was subcloned into KpnI and NotI sites of the pPac-PL vector. pPacBmRelBP and pPacBmRelAA were digested with KpnI and SmaI and the resultant 7.6 kb and 0.3 kb fragments from pPacBmRelBP and pPacBmRelAA, respectively, were ligated to construct pPacBmRelAP (see Fig. 8). 2.7. Transfection of mbn-2 cells A D. melanogaster cell line, mbn-2 [25], was used for transfection. The mbn-2 cells were maintained at 25 -C in Schneider medium (Sigma) containing 25 mg/l of streptomycin and 25,000 units/l of penicillin G. A mixture of the

H. Tanaka et al. / Biochimica et Biophysica Acta 1730 (2005) 10 – 21

transfection reagent containing 10 Al of DOTAP (Roche Diagnostic), 0.25 Ag of pPac-PL expression vector with BmRelA, BmRelB, Dorsal or Dif, 0.5 Ag of B. mori antibacterial peptide-luciferase fusion gene (pGL3 constructs) and 0.25 Ag of the h-galactosidase expression vector pACH110 [26] was added to 5  105 mbn-2 cells. The medium was replaced with fresh medium after 8 to 10 h. The transfected cells were lysed after 48 h and luciferase activity in the cell lysate was assayed using a Luciferase Assay System (Promega) and a lumicounter (Nition). Similarly, h-galactosidase activity was measured as an internal control using a Galacto-Light (Tropix). 2.8. Recombinant protein production BmRelA (nucleotide no. 558 to 1637) or BmRelB (714 to 1637) cDNA was inserted into the EcoRI and XhoI sites of glutathione S-transferase (GST) expression vector, pGEX6P-1 (Pharmacia). The fusion proteins were expressed in E. coli BL21 and purified by affinity chromatography using glutathione agarose beads according to the method of Ip et al. [27]. 2.9. Electrophoresis mobility shift assay (EMSA) Synthetic oligonucleotides as probes for EMSA were first annealed and purified by polyacrylamide gel electrophoresis. The probes were labeled with (g-32P) ATP (ICN) and purified as described previously [11]. GST –BmRel fusion proteins expressed in E. coli (400 ng) were preincubated at room temperature for 10 min in 20 mM HEPES – NaCl buffer, pH 7.9 containing 100 mM NaCl, 1 mM EDTA, 5% glycerol, 0.5 mM DTT and 50 Ag/ml poly (dIIdC) – poly (dIIdC) (Amersham Pharmacia Biotech). Labeled probes (2 ng, 4  105 dpm) were then mixed with this reaction mixture and incubated for 20 min at room temperature. Electrophoresis was conducted with the reaction mixture and radioactive signals on a dried gel were detected by BAS 2500 (Fuji Film). For competition or antibody supershift experiments, a 50-fold molar excess of nonradioactive double strand DNA or 2.5 Ag of anti-GST antibody (Amersham Pharmacia Biotech) was mixed with preincubation mixture prior to addition of the probes.

3. Results 3.1. Cloning of BmRelA and BmRelB cDNAs and real time PCR for quantification of BmRelA and BmRelB RT-PCR was conducted using RHD primers and a product obtained. Deduced amino acid sequence from the nucleotide sequence of the product revealed a characteristic RHD. Therefore, 3V and 5V-RACE were conducted to obtain a full amino acid sequence of the B. mori Rel protein. The nucleotide and deduced amino acid sequences of the Rel

13

protein (BmRelA) are shown in Fig. 1. This clone contained a single open reading frame composed of 572 amino acid residues (Fig. 1B). BmRelA RHD shows 58.2% homology with Dorsal RHD [4], 55.5% with Gambif1 RHD [10], 60.4% with A.d.RelA RHD [11], 39.8% with Dif RHD [5], 31.4% with Aedes Relish RHD [12], 35.5% with Relish RHD [7], 42.2% with human cRel [28] and 43.8% with human p65 [29] in the amino acid sequence (Fig. 2A). Phylogenetic analysis of amino acid sequences among these Rel RHDs showed that BmRels have strong evolutionary relationship with Dorsal, Gambif1 and A.d.RelA (Fig. 2B). In addition to RHD, a nuclear localization signal (KKRK), a proline rich domain (24.4% proline contents between amino acid number 375 and 419) and leucine zipper motif (amino acid number 551 to 572) were present at the C-terminus. The leucine zipper motif is present in mammalian RelB in the Nterminal region [30] but not in other insect Rel family proteins [4,5,7,10 –12,28,29]. Repetitive amino acid sequences existed in the C-terminal region of BmRelA. Repetitive amino acid sequences (16 amino acid, 6 repeats) were also seen in the mammalian transcription factor, REST [31]. The biological significance of these repeats, however, remains unknown. Another B. mori Rel protein clone designated BmRelB was found using specific 5V and 3V primers. BmRelB cDNA lacked 241 bp nucleotides (nucleotide number 333 to 573) that are present at the 5V-region of BmRel A cDNA (Fig. 1A). Two distinct BmRel mRNAs were detected after RTPCR reaction using fat body total RNA (Fig. 3A). From nucleotide sequence analysis, the upper band was identified to correspond to BmRelA and the lower band to BmRelB. Typical splice and donor sites, ‘‘GT’’ and ‘‘AG’’ [32], were found at the 5V and 3V ends of the 241 bp fragment missing from the BmRelB cDNA, suggesting that alternative splicing is involved in BmRelB mRNA synthesis. Southern blot analysis using a cDNA fragment encoding the RHD region as a probe gave a single band (data not shown), supporting this speculation. The deletion causes the loss of the first methionine codon (ATG) and results in a shift of the translation initiation codon, ATG, from nucleotide number 558 to 714 (Fig. 1A), fitting the Kozak’s translation initiation rule [33]. This ATG codon shift causes a loss of 52 amino acids. No other differences were found between BmRelA and BmRelB. We then quantified BmRelA and BmRelB mRNA by real time RT-PCR using specific primers that can recognize BmRelA and BmRelB differentially (Fig. 3B). The mRNA abundance unit was calculated based on the relative ratio of BmRel mRNA to SP2 mRNA in each mRNA sample. For this, 5th instar larvae were injected with M. luteus, then, mRNA encoding BmRelA and BmRelB analyzed. Both BmRelA and BmRelB were detected in unstimulated conditions. BmRelA RNA was 3.7-fold more abundant than BmRelB mRNA. In Fig. 3A, the band intensity of BmRelB was shown to be much higher than that of BmRelA. Since both cDNAs were amplified to the plateau level in these conditions, BmRelB cDNA was likely amplified more

14

H. Tanaka et al. / Biochimica et Biophysica Acta 1730 (2005) 10 – 21

Fig. 1. Nucleotide and deduced amino acid sequences of BmRelA cDNA. (A) Portion of the nucleotide sequence of the 5V-region of BmRelA cDNA. For the full nucleotide sequence of BmRel A cDNA, see database accession no. AB096087. (B) Deduced amino acid sequence of BmRelA. Open triangles indicate the sites deleted in BmRelB cDNA. Closed triangles denote the putative translation start site of BmRelB. Bold letters show RHD. The box indicates a nuclear translocation signal. The repetitive amino acid sequence is underlined and the proline-rich domain double-underlined. Circled leucine residues indicate the position of the putative leucine zipper.

efficiently than BmRelA DNA under the conditions using the primers described in Fig. 3A. Real-time PCR showed that both mRNAs increased 2 h after M. luteus treatment, suggesting the BmRel coding gene is up-regulated in response to M. luteus infection. 3.2. Effects of BmRel knockdown on the transcription of antibacterial peptide genes in vivo Transgenic silkworms with BmRel knockdown were prepared by a novel method [21] to analyze the effects of the BmRel gene on the transcription of antibacterial peptide genes in vivo. For this, 5th instar larvae from the transgenic

silkworms were injected with M. luteus. Then, transcription of Attacin, Cecropin B1, Lebocin 3 and Lebocin 4 was analyzed by real-time PCR and compared with that of normal silkworms injected with the same bacteria. Results showed the rate of BmRel gene expression in transgenic silkworms was 17.8% that in normal silkworms, indicating that BmRel expression was suppressed significantly, but not completely (Fig. 4). The rates of antibacterial peptide gene expression under these conditions were 21.8% (Attacin), 37.9% (Cecropin B1), 12.9% (Lebocin 3) and 6.9% (Lebocin 4) (Fig. 4), suggesting that the BmRel gene plays an important role in vivo in the activation of antibacterial peptide genes upon infection with M. luteus.

H. Tanaka et al. / Biochimica et Biophysica Acta 1730 (2005) 10 – 21

15

Fig. 2. Comparison and phylogenic analysis between the amino acid sequence of BmRel RHD and other Rel RHDs. (A) Comparison of amino acid sequences among RHDs from BmRel, Dorsal [4], Gambif1 [10], Dif [5], A.d.Rel [11], Relish [7], Aedes Relish R6 [12], human p65 [29] and human cRel [28]. Conserved amino acids are highlighted. Gaps were introduced to maximize sequence alignment. (B) Phylogenetic analysis. Computer-aimed phylogenetic analysis was done by the neighbor-joining method for amino acid sequences among Rel proteins described above. A phylogenetic tree was constructed using CLUSTAL X [45].

3.3. Effects of BmRelA and BmRelB on the activation of antibacterial peptide genes in a cell line D. melanogaster cell line, mbn-2, was chosen to analyze the effects of BmRelA and BmRelB on the activation of B.

mori antibacterial peptide genes, because (1) it was difficult to establish transgenic silkworms in which production of BmRelA and BmRelB mRNAs was separately repressed, (2) the background level of the activation of antibacterial peptide genes by endogenous transcription factors was very

16

H. Tanaka et al. / Biochimica et Biophysica Acta 1730 (2005) 10 – 21

Fig. 4. Comparison of the mRNA levels of BmRel, Lebocin 3, Lebocin 4, Attacin and Cecropin B1 in normal and transgenic B. mori 5th instar larvae injected with M. luteus. Relative quantitation was performed as described in Fig. 3, demonstrating statistically significant differences for BmRel, Lebocin 3, Lebocin 4, Attacin and Cecropin B1.

Fig. 3. Detection of BmRelA and BmRelB and comparison of the relative mRNA levels of BmRelA and BmRelB. (A) Detection of BmRelA and BmRelB cDNA by RT-PCR. A combination of a forward primer 5VCATCGAGGATTTACGCACAC-3V (corresponding to positions 22 to 41) and a reverse primer 5V-TACTGTCGGCCTCAACGTTGAGGT-3V (corresponding to positions 2277 to 2300) (lane 1), or a combination of a forward primer 5V-TCGGACAAGCGAGAGCGAAGAGA-3V(corresponding to positions 278 to 300), and reverse primer 5V-TACTGTCGGCCTCAACGTTGAGGT-3V (corresponding to positions 2277 to 2300) (lane 2) were used for RT-PCR. The RT-PCR products were separated by agarose gel electrophoresis. Upper and lower bands correspond to BmRelA, and BmRelB cDNA, respectively. M denotes DNA size marker. (B) Comparison of the relative mRNA levels of BmRelA and BmRelB in fat body excised from B. mori 5th instar larvae, non-injected (control) or injected with M. luteus. Specific primers described in Materials and methods were used to detect BmRelA and BmRelB. Quantitation was made by real-time detection RT-PCR. The mRNA abundance unit was calculated based on the relative ratio of each BmRel mRNA to SP2 mRNA in each mRNA sample. Relative transcript of BmRelB mRNA (control) was set at 1.0. Results of triplicate experiments are shown.

low (Fig. 5B), (3) Dorsal and Dif from D. melanogaster did not stimulate any B. mori antibacterial peptide gene constructs (Fig. 5B). This provides an ideal system to analyze the interaction between BmRels and B. mori antibacterial peptide genes. Effects of BmRelA and BmRelB on the expression of B. mori antibacterial peptide genes were analyzed 2 days after transfection by measuring luciferase activity. Results showed that BmRelB activated the Attacin gene strongly and Cecropin B1, Lebocin 3 and Lebocin 4 genes to a lesser extent. On the contrary, BmRelA activated the Lebocin 4 gene dramatically and Attacin and Lebocin 3 genes to a lesser extent (Fig. 5B). The results indicate there are two types of antibacterial peptide genes regulated by BmRelA and BmRelB, so representative antibacterial peptide genes, namely Attacin and Lebocin 4, were analyzed for further understanding.

Fig. 5. Effects of BmRelA and BmRelB on the expression of gene constructs consisting of B. mori antibacterial peptide promoter and luciferase reporter gene in mbn-2 cells. Schematic diagram of B. mori antibacterial peptide promoter regions in the gene constructs. The positions of nB sites are indicated. +1 denotes the transcription initiation site for Attacin and Cecropin B1 gene, and 1 indicates 1 bp before the translation initiation site for Lebocin 3 and Lebocin 4. luc denotes the firefly luciferase reporter gene. (B) Relative luciferase activity in the mbn-2 cells. The bars indicate mean T S.D. of enzyme activity (n = 3). pPac and pGL3 indicate pPac-PL vector and pGL3-basic vector, respectively.

H. Tanaka et al. / Biochimica et Biophysica Acta 1730 (2005) 10 – 21

17

Fig. 6. Functional analysis of the nB sites of Attacin, Lebocin 3 and Lebocin 4 genes. (A) Attacin. A nB site of the Attacin promoter region of wild and deletion types are schematically presented. (B) Lebocin 4. Three nB sites in Lebocin 4 genes and deletion mutants are shown. The region of 165 to 158 was deleted to construct the L4mut for the Lebocin 4 construct. The bars indicate mean T S.D. of enzyme activity (n = 3).

3.4. Effects of jB sites on the activation of antibacterial peptide genes by BmRelA or BmRelB The mbn-2 cells were cotransfected with expression vectors and luciferase activity analyzed. When the nB site was deleted, activation of the Attacin gene by BmRelB was dramatically reduced (Fig. 6A). Lebocin 4 gene was activated by BmRelA much more strongly than Lebocin 3 gene (Fig. 6B). There are three nB sites of which two proximal nB sites have identical nucleotide sequences in Lebocin 3 and Lebocin 4 genes. Therefore, the distal nB site of both Lebocin genes, which have different locations and nucleotide sequences (Fig. 6A), is considered to be an important regulatory site. Deletion of the distal nB site ( 166 to 157) of Lebocin 4 gene resulted in a significant

reduction in activation by BmRelA (Fig. 6B), suggesting that the distal nB site plays an important role in the activation of the Lebocin 4 gene by BmRelA. 3.5. Binding of BmRelA and BmRelB to jB sites of antibacterial peptide genes RHD has been shown to be the domain that binds to the nB site [34]. The RHD of BmRelA and BmRelB was examined to determine whether it can bind to the nB site of the Attacin and Lebocin 4 genes. cDNA containing the Nterminal region including the RHD of BmRelA or BmRelB was ligated into a GST expression vector to produce GST – RHD fusion recombinant proteins in E. coli (Fig. 7A). Binding of the recombinant protein to the nB site of Attacin

Fig. 7. EMSA for the binding of BmRelA and BmRelB to the nB site of Attacin and Lebocin 4 genes. EMSA was conducted using recombinant fusion proteins of GST and the N-terminal half of BmRelA or BmRelB containing RHD. (A) Schematic presentation of the fusion protein GST and the N-terminal half of BmRelA or BmRelB. G: GST, A: Fusion protein of GST-BmRelA. B: Fusion protein of GST-BmRelB. (B) Attacin nB site. (C) Lebocin 4 nB site. The nucleotide sequence of the probe is shown and the nB site underlined. + and denote addition and nonaddition, respectively.

18

H. Tanaka et al. / Biochimica et Biophysica Acta 1730 (2005) 10 – 21

or Lebocin 4 genes, shown to be critical for antibacterial protein gene expression, was tested by EMSA. Results showed that the GST –BmRelB RHD recombinant protein can bind to the nB site of Attacin genes (Fig. 7B). No signal was detected with GST protein alone. A supershift of the binding complex was observed when polyclonal antibodies raised against GST were incubated and the mixture analyzed by EMSA (Fig. 7B). In a control experiment, a supershift in EMSA using antibodies raised against IgG was not observed (data not shown). This binding was confirmed to be specific by a competition assay using excess amounts of a cold probe (Fig. 7B). The GST –BmRelA RHD recombinant protein was also demonstrated to bind to the nB site ( 166 to 157) of the Lebocin 4 gene (Fig. 7C). As this binding reaction was inhibited by excess amounts of a cold probe, specificity was confirmed as observed in the Attacin gene (Fig. 5C). The DNA – protein complex signal was not seen with the corresponding region ( 166 to 157) of the Lebocin 3 gene that is not a nB site (data not shown). Reverse combinations of BmRelA RHD and Attacin nB site, and BmRelB RHD and Lebocin 4 distal nB site were also tested to determine their binding ability. Results showed that they can bind well with each other in vitro (data not shown), suggesting that non-selected binding of BmRelA and BmRelB to the nB sites occurs in vivo leading to control of antibacterial peptide gene expression or there are yet unknown mechanisms in vivo controlling the in vivo specific binding of BmRelA and BmRelB to the nB sites of their target antibacterial peptide genes. 3.6. Effects of the C-terminal domain of BmRelA or BmRelB on the activation of Attacin and Lebocin 4 genes Three domains of transcription factors [35], responsible for transcriptional activation have been reported, i.e., acidic domain in GAL4 [36] and RelA (p65) [37], glutamine-rich domain in Sp1 [38] and Dorsal [4], and proline-rich domain in CTF/NF1 [39] and Dif [40]. The proline-rich domains of BmRelA and BmRelB were analyzed to determine whether they play a role in transcription activation. Expression vectors containing different deletion variants of BmRelA and BmRelB cDNA (Fig. 8A) were constructed and mbn-2 cells cotransfected with one of two expression vectors, an expression vector containing Attacin or Lebocin 4 promoter-luciferase reporter gene and an expression vector containing h-galactosidase gene. Luciferase activity of the deletion constructs of the proline-rich domain of BmRelA or BmRelB (BmRelAA or BmRelBA) was dramatically reduced when compared to the activity of a nondeletion control (Fig. 8B). On the contrary, the luciferase activity of the constructs containing the proline-rich domain but a deletion of other C-terminal regions (BmRelAP or BmRelBP) was comparable to that of the nondeletion control for Lebocin 4 and Attacin (Fig. 8B), suggesting the proline-rich domain is indispensable for full expression of transcriptional activation by these Rel proteins.

Fig. 8. Effects of C-terminal truncated BmRelA and BmRelB on the expression of Attacin and Lebocin 4 genes. (A) Schematic presentation of wild and truncated type BmRelA and BmRelB. RHD: Rel homology domain, NLS: nuclear localization signal, PRD: proline rich domain, LZ: leucine zipper motif. Amino acid number from the translation initiation codon of BmRelA is shown (see Fig. 1). (B) Relative luciferase activity. pGL3-basic vector was used as a negative control for cotransfection of mbn-2 cells containing different deletion variants of BmRelA or BmRelB cDNAs.

4. Discussion We have cloned two structurally related cDNAs encoding novel Rel family members (BmRelA and BmRelB) from B. mori. The difference between BmRelA and BmRelB cDNA is that BmRelB cDNA lacked 241 bp containing the BmRelA translation initiation codon, ATG. As a result, BmRelB is supposed to produce a 52 amino acids-shorter product than BmRelA. BmRelA and BmRelB possess typical RHDs at the N-terminal region. Comparison of amino acid sequences among RHDs of insect and human Rel proteins such as BmRel, Dif, Dorsal, Relish, Gambif1, A.d.RelA, Aedes Relish, human cRel and human p65 indicates that the BmRel RHDs have a high similarity to those of other animal Rel proteins. Results of phylogenetic analysis of amino acid sequences among these Rel RHDs showed that BmRels have a strong evolutionary relationship with Dorsal, Gambif1 and A.d.RelA, suggesting they have been derived from a common ancestral gene. However, regions other than the RHD in BmRels showed quite different amino acid sequences from those of other Rel proteins. BmRels are unique in that they have a leucine zipper motif and repetitive amino acid sequences at the C-terminal region. As for the

H. Tanaka et al. / Biochimica et Biophysica Acta 1730 (2005) 10 – 21

leucine zipper motif, only mammalian RelB possesses this motif at the N-terminal region, which is required for full transactivation. However, the leucine zipper motif of BmRels seems not to be necessary for full activation, since deletion of the leucine zipper motif had little effect on the transcriptional activation. BmRel-knockdown transgenic silkworms provided direct evidence that the BmRel gene is involved in the activation of antibacterial peptide genes such as Cecropin B1, Attacin, Lebocin 3 and Lebocin 4. To our knowledge, the transgenic silkworms are the first insect except for D. melanogaster in which activity of a gene is knocked down through subsequent generations. This technique should be useful for analyzing the functions of unknown proteins in B. mori. The most interesting result obtained from the present study is that BmRelA and BmRelB have differential effects on the activation of B. mori antibacterial peptide genes. As far as we know, a differential activation of antibacterial peptide genes by a Rel protein and its shortened isoform has not yet been reported. These results illustrate that a minor structural change at the N-terminal region of the BmRels causes dramatic differential activation of B. mori antibacterial peptide genes. Our results also indicated that the prolinerich domain is essential to activate Attacin and Lebocin 4 genes. Lebocin 4 gene was strongly activated by BmRelA and weakly by BmRelB, suggesting the N-terminal 52 amino acid sequence of BmRelA plays a role as another activator of Lebocin 4 gene expression. On the other hand, the Attacin gene was strongly activated by BmRelB but almost repressed by BmRelA, suggesting that the N-terminal 52 amino acid sequence plays a role as a repressor of the Attacin gene. We speculate from these results that the N-terminal 52 amino acid sequence has dual functions as an activator and a repressor for controlling target antibacterial peptide genes probably by enhancing or suppressing the function of the proline-rich domain. A few studies have reported that the Nterminal region preceding the RHD functions as a transcriptional regulatory domain. In addition to mammalian RelB, A. aegypti Relish has a N-terminal transcriptional activation domain, which corresponds to a glutamine- and histidinerich stretch followed by a long serine-rich region. However, to our knowledge, other Rel proteins have neither transcriptional activation domains nor repressive domains at the Nterminal region. The 52 amino acid sequence of BmRelA is occupied by mainly hydrophobic (61.5%) and acidic (19.2%) amino acid residues (Fig. 9). These characteristic amino acid residues cannot be found at the N-terminal region of insect Rel proteins such as Dif, Dorsal, Relish and A.d.RelA, and even in the mammalian Rel proteins like cRel, which have an extremely short N-terminal sequence preceding the RHD, or p105/p50 [41] containing relatively longer amino acid sequences. On the contrary, Gambif1 (hydrophobic amino acids: 63.0%, acidic amino acids: 19.6%) and human p65 (hydrophobic amino acids: 72.2%, acidic amino acids: 16.7%) possess these characteristic amino acid residues, though they have not been reported as a functional

19

Fig. 9. BmRelA specific amino acid sequence. (A) BmRelA specific 52 amino acid sequence. Bold and underlined letters indicate hydrophobic and acidic amino acids, respectively. (B) Alignment of partial amino acid sequence of BmRelA and Human RelA associate inhibitor (hRAI). Numerals indicate amino acid numbers from the translation initiation codon. Identical amino acids are highlighted.

domain. Interestingly, some transcription factors such as herpes simplex virus type1 virion protein 16 (VP16), RelA/ p65 and p53 have these hydrophobic and acidic amino acid residues involved in transactivation. We demonstrated using yeast GAL4 system that this 52 amino acid sequence activated GAL4-dependent transcription (data not shown), suggesting this sequence has a potential ability as a transcriptional activation domain. The 52 amino acid sequence probably interacts with co-activator or co-repressors to regulate antibacterial peptide genes. In mammalian innate immunity, a similar positive and negative regulation for different genes by a factor has been reported. InB-~ is a member of the InB family which has an ankyrin-repeat and is localized in the nucleus. In general, InB-~ binds to the p50 subunit of NF-nB and inhibits NFnB-dependent transcription [42], whereas InB-~ promotes activation of IL-6 gene [43]. Thus, dual functional nuclear proteins may play an important role in innate immunity of vertebrates and invertebrates. Interestingly, the N-terminal 52 amino acid sequence of BmRelA shows 50% homology with a partial sequence of the ankyrin-repeat of RelAassociated inhibitor (RAI), an inhibitor of human NF-nB [44] (Fig. 9). RAI specifically inhibits the DNA binding activity of p65 (RelA), a NF-nB subunit. It is important to determine whether RAI also has dual functions in the regulation of immune-related protein genes.

Acknowledgements We thank Prof. E. Gateff for the mbn-2 cells, Prof. J-M. Reichhart for the pACH110, pPac-PL and pPac-dorsal plasmids, and Prof. Y. Engstro¨m for the pAct-Dif plasmid. This work was supported in part by a grant from PROBRAIN, Japan and by a Grant-in-Aid (Insect Technology Program) from the Ministry of Agriculture, Forestry, and Fisheries, Japan.

References [1] J.A. Hoffmann, The immune response of Drosophila, Nature 426 (2003) 33 – 38.

20

H. Tanaka et al. / Biochimica et Biophysica Acta 1730 (2005) 10 – 21

[2] N. Silverman, T. Maniais, NF-nB signaling pathway in mammalian and insect innate immunity, Genes Dev. 15 (2001) 2321 – 2342. [3] S. Ghosh, M.J. May, E.B. Kopp, NF-nB and Rel proteins: evolutionarily conserved mediators of immune responses, Annu. Rev. Immunol. 16 (1998) 225 – 260. [4] R. Steward, Dorsal, an embryonic polarity gene in Drosophila, is homologous to the vertebrate proto-oncogene, c-rel, Science 238 (1987) 692 – 694. [5] Y.T. Ip, M. Reach, Y. Engstro¨m, L. Kadalayil, H. Cai, S. GonzalesCrespo, K. Tatei, M. Levine, Dif, a dorsal-related gene that mediates an immune response in Drosophila, Cell 75 (1993) 753 – 763. [6] B. Lemaitre, J.M. Reichhart, J.A. Hoffmann, Drosophila host defense: differential induction of antimicrobial peptide genes after infection by various classes of microorganisms, Proc. Natl. Acad. Sci. U. S. A. 94 (1997) 14614 – 14619. ˚ sling, D. Hultmark, Origins of immunity: Relish, a [7] M.S. Dushay, B. A compound Rel-like gene in the antibacterial defense of Drosophila, Proc. Natl. Acad. Sci. U. S. A. 93 (1996) 10343 – 10347. ˚ sling, M.S. Dushay, I. Ando, S. Ekengren, M. [8] M. Hedengren, B. A Wihlborg, D. Hultmark, Relish, a central factor in the control of humoral but not cellular immunity in Drosophila, Mol. Cell 4 (1999) 827 – 837. [9] S. Stoven, I. Ando, L. Kadalayil, Y. Engstro¨m, D. Hultmark, Activation of the Drosophila NF-nB factor Relish by rapid endoproteolytic cleavage, EMBO Rep. 1 (2000) 347 – 352. [10] C. Barillas-Mury, A. Charlesworth, I. Gross, A. Richman, J.A. Hoffmann, F.C. Kafatos, Immune factor Gambif1, a new rel family member from the human malaria vector, Anopheles gambiae, EMBO J. 15 (1996) 4691 – 4701. [11] A. Sagisaka, H. Tanaka, S. Furukawa, M. Yamakawa, Characterization of a homologue of the Rel/NF-nB transcription factor from a beetle, Allomyrina dichotoma, Biochim. Biophys. Acta 1678 (2004) 85 – 93. [12] S.W. Shin, V. Kokoza, A. Ahmed, A.S. Raikhel, Characterization of three alternatively spliced isoforms of the Rel/NF-nB transcription factor Relish from the mosquito Aedes aegypti, Proc. Natl. Acad. Sci. U. S. A. 99 (2002) 9978 – 9983. [13] H. Shiraishi, A. Kobayashi, Y. Sakamoto, T. Nonaka, Y. Mitsui, N. Aozawa, T. Kubo, S. Natori, Molecular cloning and characterization of SRAM, a novel insect rel/ankyrin-family protein present in nuclei, J. Biochem. 127 (2000) 1127 – 1134. [14] C.A. Brennan, K.V. Anderson, Drosophila: the genetics of innate immune recognition and response, Annu. Rev. Immunol. 22 (2004) 457 – 483. [15] Y. Yamano, M. Matsumoto, K. Sasahara, E. Sakamoto, I. Morishima, Structure of genes for cecropin A and an inducible nuclear protein that binds to the promoter region of the genes from the silkworm, Bombyx mori, Biosci. Biotechnol. Biochem. 62 (1998) 237 – 241. [16] K. Taniai, Y. Kato, H. Hirochika, M. Yamakawa, Isolation and nucleotide sequence of cecropin B cDNA clones from the silkworm, Bombyx mori, Biochim. Biophys. Acta 24 (1992) 203 – 206. [17] M. Sugiyama, H. Kuniyoshi, E. Kotani, K. Taniai, K. Kadono-Okuda, Y. Kato, M. Yamamoto, M. Shimabukuro, S. Chowdhury, J. Xu, S.K. Choi, H. Kataoka, A. Suzuki, M. Yamakawa, Characterization of a Bombyx mori cDNA encoding a novel member of the Attacin family of insect antibacterial proteins, Insect Biochem. Mol. Biol. 25 (1995) 385 – 392. [18] S. Furukawa, K. Taniai, J. Ishibashi, S. Hara, T. Shono, M. Yamakawa, A novel member of lebocin gene family from the silkworm, Bombyx mori, Biochem. Biophys. Res. Commun. 238 (1997) 769 – 774. [19] S. Furukawa, H. Tanaka, H. Nakazawa, J. Ishibashi, T. Shono, M. Yamakawa, Inducible gene expression of moricin, a unique antibacterial peptide from the silkworm (Bombyx mori), Biochem. J. 340 (1999) 265 – 271. [20] K. Taniai, K. Kadono-Okuda, Y. Kato, M. Yamamoto, S. Shimabukuro, S. Chowdhury, J. Xu, E. Kotani, S. Tomino, M. Yamakawa, Structure

[21]

[22]

[23]

[24]

[25] [26]

[27]

[28]

[29]

[30]

[31]

[32] [33] [34]

[35]

[36] [37]

[38]

[39]

[40]

of two cecropin B-encoding genes and bacteria-inducible DNAbinding proteins which bind to the 5V-upstream regulatory region in the silkworm, Bombyx mori, Gene 163 (1995) 215 – 219. M. Yamamoto, M. Yamao, H. Nishiyama, S. Sugihara, S. Nagaoka, M. Tomita, K. Yoshizato, T. Tamura, H. Mori, New and highly efficient method for silkworm transgenesis using Autographa californica nucleopolyhedrovirus and piggybac transposable elements, Biotech. Bioeng. 88 (2004) 849 – 853. G.X. Quan, T. Kanda, T. Tamura, Induction of the white egg 3 mutant phenotype by injection of the double-stranded RNA of the silkworm white gene, Insect Mol. Biol. 11 (2002) 217 – 222. K. Taniai, T. Ishii, M. Sugiyama, A. Miyanoshita, M. Yamakawa, Nucleotide sequence of 5V-upstream region and expression of a silkworm gene encoding a new member of the Attacin family, Biochem. Biophys. Res. Comm. 220 (1996) 594 – 599. M.A. Krasnow, E.E. Saffman, K. Kornfeld, D.S. Hogness, Transcriptional activation and repression by Ultrabithorax proteins in cultured Drosophila cells, Cell 57 (1989) 1031 – 1043. E. Gateff, Malignant neoplasms of genetic origin in Drosophila melanogaster, Science 200 (1978) 1448 – 1459. C. Thisse, F. Perrin-Schmitt, C. Stoetzel, B. Thisse, Sequence-specific transactivation of the Drosophila twist gene by the dorsal gene product, Cell 65 (1991) 1191 – 1201. Y.T. Ip, R.E. Park, D. Kosman, E. Bier, M. Levine, The dorsal gradient morphogen regulates stripes of rhomboid expression in the presumptive neuroectoderm of the Drosophila embryo, Genes Dev. 6 (1992) 1728 – 1739. E. Brownell, N. Mittereder, N.R. Rice, A human rel proto-oncogene cDNA containing an Alu fragment as a potential coding exon, Oncogene 4 (1989) 935 – 942. S.M. Ruben, P.J. Dillon, R. Schreck, T. Henkel, C.H. Chen, M. Maher, P.A. Baeuerle, C.A. Rosen, Isolation of a rel-related human cDNA that potentially encodes the 65-kD subunit of NF-nB, Science 251 (1991) 1490 – 1493. R.-P. Ryseck, P. Bull, M. Takamiya, V. Bours, U. Siebenlist, P. Dobrzanski, R. Bravo, RelB, a new Rel family transcription activator that can interact with p50-NF-nB, Mol. Cell. Biol. 12 (1992) 674 – 684. J.A. Chong, J. Tapia-Ramirez, S. Kim, J. Toledo-Aral, Y. Zheng, M.C. Boultros, Y.M. Altshuller, M.A. Frohman, S.D. Kraner, G. Mandel, REST: a mammalian silencer protein that restricts sodium channel gene expression to neurons, Cell 80 (1995) 849 – 957. S.M. Mount, A catalogue of splice junction sequences, Nucleic Acids Res. 10 (1982) 459 – 472. M. Kozak, An analysis of 5V-noncoding sequences from 699 vertebrate messenger RNAs, Nucleic Acids Res. 15 (1987) 8125 – 8148. C. Kunsch, S. Ruben, C. Rosen, Selection of optimal nB/Rel DNAbinding motifs: interaction of both subunits of NF-nB with DNA is required for transcriptional activation, Mol. Cell. Biol. 12 (1992) 4412 – 4421. P.J. Mitchell, R. Tjian, Transcriptional regulation in mammalian cells by sequence-specific DNA binding proteins, Science 245 (1989) 371 – 378. M. Ptashne, How eukaryotic transcriptional activators work, Nature 335 (1988) 683 – 689. W.S. Blair, H.P. Bogerd, S.J. Madore, B.R. Cullen, Mutational analysis of the transcription activation domain of RelA: identification of a highly synergistic minimal acidic activation module, Mol. Cell. Biol. 14 (1994) 7226 – 7234. A.J. Courey, D.A. Holtzman, S.P. Jackson, R. Tjian, Synergistic activation by the glutamine-rich domains of human transcription factor Sp1, Cell 59 (1989) 827 – 836. N. Mermod, E.A. O’Neill, T.J. Kelly, R. Tjian, The proline-rich transcriptional activator of CTF/NF-I is distinct from the replication and DNA binding domain, Cell 58 (1989) 741 – 753. U.M. Petersen, G. Bjorklund, Y.T. Ip, Y. Engstro¨m, The dorsalrelated immunity factor, Dif, is a sequence-specific trans-activator

H. Tanaka et al. / Biochimica et Biophysica Acta 1730 (2005) 10 – 21 of Drosophila Cecropin gene expression, EMBO J. 14 (1995) 3146 – 3158. [41] M. Kieran, V. Blank, F. Logeat, J. Vandekerckhove, F. Lottspeich, O. le Bail, M.B. Urban, P. Kourilsky, P.A. Baeuerle, A. Israel, The DNA binding subunit of NF-nB is identical to factor KBF1 and homologous to the rel oncogene product, Cell 62 (1990) 1007 – 1018. [42] S. Yamazaki, T. Muta, K. Takeshige, A novel InB protein, InB-~, induced by proinflammatory stimuli, negatively regulates nuclear factor-nB in the nuclei, J. Biol. Chem. 276 (2001) 27657 – 27662. [43] M. Yamamoto, S. Yamazaki, S. Uematsu, S. Sato, H. Hemmi, K. Hoshino, T. Kaisho, H. Kuwata, O. Takeuchi, K. Takeshige, T.

21

Saitoh, S. Yamaoka, N. Yamamoto, S. Yamamoto, T. Muta, K. Takeda, S. Akira, Regulation of Toll/IL-1-receptor-mediated gene expression by the inducible nuclear protein InB-~, Nature 430 (2004) 218 – 222. [44] J.-P. Yang, M. Hori, T. Sanda, T. Okamoto, Identification of a novel inhibitor of nuclear factor-nB, RelA-associated inhibitor, J. Biol. Chem. 274 (1999) 15662 – 15670. [45] J.D. Thompson, D.G. Higgins, T.J. Gibson, CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice, Nucleic Acids Res. 11 (1994) 4673 – 4680.