Original Paper

Pdf file on Synergy


Acta Biochim Biophys Sin 2006, 38: 507-513


Structural Analysis of Fibroin Heavy Chain signal peptide of Silkworm Bombyx mori



Sheng-Peng WANG1,2, Ting-Qing GUO1, Xiu-Yang GUO1, Jun-Ting HUANG2, and Chang-De LU1*


1 Institute of Biochemistry and Cell Biology, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Shanghai 200031, China;

2 Sericultural Research Institute, Chinese Academy of Agricultural Sciences, Zhenjiang 212018, China


Received: March 14, 2006        Accepted: April 6, 2006

This work was supported by the grants from the National Natural Science Foundation of China (No. 30370326 and No. 30470350)

*Corresponding author: Tel, 86-21-54921234; Fax, 86-21-54921011; E-mail, cdlu@sibs.ac.cn


Abstract        to study the minimal length required for the secretion of recombinant proteins and silk proteins in posterior silk gland, the signal peptide (SP) of the fibroin heavy chain (FibH) of silkworm Bombyx mori was systematically shortened from the C-terminal. Its effect on the secretion of protein was observed using enhanced green fluorescent protein (EGFP) as a reporter. Secretion of EGFP fusion proteins was examined under fluorescence microscope. FibH SPs with lengths of 20, 18, 16 and 12 a.a. can direct the secretion of the reporter, yet those with lengths of 11, 10, 9, 8 and 1 a.a. can not. When the FibH SP was shortened to 12 a.a., the secretion efficiency was decreased slightly and the cleavage happened within EGFP. When 16 a.a. of the FibH SP were used, the secretion of fusion protein was normal and the cleavage site was between the Gly-Ser linker and Met, the starting amino acid of EGFP. These findings are applicable for the expression of foreign proteins in silkworm silk gland. The cleavage site of the SP is discussed and compared with the predictive results of the SignalP 3.0 online prediction program.


Key words        signal peptide; fibroin heavy chain; silkworm, Bombyx mori; rAcMNPV


Most secreted and membrane proteins produced in eukaryotic­ cells are targeted to or translocated across the endoplasmic reticulum (ER) membrane by a short peptide termed the signal peptide (SP). The SPs of nascent preproteins are recognized by signal recognition particles (SRPs) that help them bind and pass through the receptor on ER, then protein is synthesized continually into ER [1]. The SP is cleaved by highly specific signal peptidase during­ or after protein translocation [2]. A model SP has a canonical s­tructure with three regions: an N-terminal domain­ (n-region) which contains a net positive charge, a hydrophobic core domain (h-region) and a polar C-terminal­ domain (c-region) [3,4]. A net positive charge in the n-region is required for efficient translocation across the inner­ membrane. The hydrophobic h-region is the most essential part required for targeting and membrane insertion [4]. The c-region has the least length variability and consists of relatively small and neutral polar residues. This region is very important for recognition and cleavage by signal peptidase [5]. The 3,1 rule states that residues in position 3 and 1 relative to the cleavage site must be small and uncharged, and that large, bulky residues might reside in position 2. The tendency to conserve a distance of four to five residues from the h/c boundary to the cleavage­ site might reflect interaction between the c-region and the active site of signal peptidase [3,4].

The silk gland is a secretory organ of silkworm. Three kinds of silk proteins, fibroin heavy chain (FibH), fibroin light chain and fibrohexamerin/P25 are synthesized and secreted by the posterior silk gland (PSG) [6,7]. The SP of FibH was not clear for a long period. The N-terminal amino acids of FibH has been studied by traditional chromatography methods, but gave indefinite results­ [8]. A potential SP of FibH indicated by EMBL (P05790) is the first 21 amino acid residues according to the nucleotide sequence of the FibH gene (GenBank accession No. AF226688) [9]. This was confirmed experimentally in our laboratory recently using recombinant baculovirus as vector­ [10].

To study the essential region required for secretion of silk proteins and foreign proteins in the silk gland of silkworm, the SP sequence of FibH was systematically shortened from the C-terminal. fusion genes of SP of different length and enhanced green fluorescent protein (SP-EGFPs) were delivered into silkworm silk gland using­ recombinant AcMNPV as a transfer vector. The SP cleavage­ sites were determined by N-terminal sequencing for fusion proteins.



Materials and Methods



Plasmids and gene sequences


Plasmid p5L was cloned previously in our laboratory. It contains the FibH sequence from -874 to +1484 bp including­ the promoter sequence, exon 1, intron sequence and partial­ exon 2 sequence, encoding N-terminal 163 a.a. residues of FibH. Plasmid p5LEGFPhis was constructed from p5L fused with the Egfp gene and a His-tag coding fragment. Plasmid pEGFPhis containing Egfp gene, His-tag coding fragment and SV40 polyA site sequence was constructed in our previous work [10].


Predictive analysis of SP


The N-terminal 70 a.a. of FibH and different fusion proteins, SP-EGFPs, were analyzed using the SignalP 3.0 online prediction program (http://www.cbs.dtu.dk/services/SignalP/) [11,12].


SP shortening by PCR and construction of recombinant­ genes


Promoter of FibH and SP encoding sequence of dif­ferent lengths (1, 8-12, 16, 18 and 20 a.a.) were cloned from plasmid p5L by PCR using primers listed in table 1. BamHI site was introduced downstream from the SP coding­ sequences. All PCR products were cloned into pGEM-T vector (Promega, Madison, USA) and verified­ by nucleotide sequencing (Shanghai BioAisa Biotechnology, Shanghai, China). EGFPhis coding fragment cut from plasmid­ pEGFPhis was fused to the signal sequence by BglII site to form the expression cassette (Fig. 1). The blocked cut site of BamHI/BglII formed a Gly-Ser linker between FibH SP and EGFP. The deduced amino acid sequences­ of different SP-EGFPs are shown in Fig. 1.


Production of recombinant baculovirus


Recombinant baculovirus was generated using the Bac-to-Bac system (Invitrogen, Carlsbad, USA). Plasmid pFFa2 was derived from pFastBacHTa (Invitrogen) with its polyhedrin promoter deleted in our previous work [13]. Expression cassettes of EGFP with FibH SP of different lengths were cloned into donor vector pFFa2, then transferred into Escherichia coli DH10BacDEGT component cells to make recombinant bacmids. Purified bacmids were used to transfect Sf9 cultured cells with Cellfectin (Invitrogen) to produce recombinant virus. All of these procedures referred to our previous works [10,13,14] and Invitrogen instruction manual. Virus titer was determined by the Tissue Culture Infectious Dose 50 method (Adeno Vator Vector System Applications Manual; Qbiogene, Carlsbad, USA), and Sf9 cells were maintained in Grace medium (Invitrogen) supplemented with 10% fetal bovine serum (Invitrogen) at 27 .


Silkworm inoculation and dissection


Silkworm larvae (54A bivotine, Japanese strain) provided­ by the Sericultural Research Institute, Chinese Academy of Agricultural Sciences (Zhenjiang, China) were reared on mulberry leaves at 25 . The recombinant baculovirus was injected into the hemocoele of newly ecdysed fifth instar silkworm larvae with a syringe at the amount of 106 pfu per larva. Approximately 5 d post-injection, the fluorescence of EGFP in silk gland of the silkworm was observed­ and photographed with fluorescence microscope (model MZ FL III; Leica, Heerbrugg, Switzerland) after dissection.


Protein purification by Ni-NTA system


Silk glands dissected from silkworm larvae were rinsed in cold sterile distilled water several times to remove adhesive­ plasma. The PSGs were homogenized with ddH2O, insoluble materials were removed by centrifu­gation at 16,000 g for 10 min; and the supernatant was lyophilized­ to a small volume and stored at 4 for a few hours and then centrifuged again. This cycle was repeated several times until no insoluble materials appeared. The ultimate supernatant­ was purified through the Ni-NTA Purification­ System­ (Invitrogen). All operations were according­ to the instruction manual. Two milliliters of resin was washed with ddH2O several times and balanced with 1ative Purification­ Buffer (50 mM NaH2PO4, pH 8.0, 0.5 M NaCl) before use. Sample for purification was mixed with 1/4 volume of 5ative Purification Buffer, centrifuged and bound in the column for 30-60 min; the column was washed with 8 ml Native­ Wash Buffer (50 mM NaH2PO4, pH 8.0, 0.5 M NaCl, 20 mM imidazole) four times; and the column­ was eluted for the target protein by using 12 ml of Native Elution Buffer (50 mM NaH2PO4, pH 8.0, 0.5 M NaCl, 250 mM imidazole); the fractions were detected­ for fluorescence of EGFP with a fluorescence spectropho­tometer (F-4010; Hitachi, Tokyo, Japan) using­ 490 nm as the excitation wavelength and 510 nm as the emission­ wavelength. The fractions with fluorescent peak were collected and dialyzed against ddH2O at 4 overnight, then concentrated by lyophilizing, and stored at -20 for further analysis.


SDS-PAGE and Western blot analysis


Silk gland homogenates or column purified protein samples were subjected to SDS-PAGE as described by Laemmli [15]. Proteins were transferred onto poly­vinylidene difluoride (PVDF) membrane (Immobilon-P; Millipore, Billerica, USA). The membrane was stained by Coomassie brilliant blue R-250, destained in 50% MeOH, and fully destained in 100% MeOH after photography. Western blot analysis was carried out using horseradish peroxidase-linked mouse monoclonal antibody GFP (B-2) (Santa Cruz Biotechnology, Santa Cruz, USA) and developed­ using 3,3'-diaminobenzidine-tetrachloride (DAB) reagent. After photography the membrane was re-stained using Coomassie brilliant blue R-250.


N-terminal sequencing


For N-terminal sequencing, proteins were transferred onto PVDF membranes from polyacrylamide gel in CAPS buffer [10 mM 3-(cyclohexylamino)-1-propanesulfonic acid (CAPS), 10% MeOH, pH 11.0] as described by Matsudaira [16] with a setting of 90 V and 300 mA for 3 h in a transfer tank (VE-186; Tanon Science & Technology, Shanghai, China). The band of EGFP fusion protein on PVDF membranes was characterized with Western blotting­ of the control lane and excised from the membrane for sequencing on a protein N-terminal sequencer (PE491A; Applied Biosystems, Foster, USA) at the Research­ Center for Proteome Analysis, Shanghai Institutes­ for Biological Sciences, Chinese Academy of Sciences (Shanghai, China).






Construction of recombinant AcMNPVs for expression­ of EGFP with different lengths of FibH SP


The SP of FibH is a typical kind of eukaryotic­ signal peptide and composed of 21 amino acid residues; it contains­ two positive amio acid residues in its n-region (R and K), 11 hydrophobic amino acid residues in its h-region, and five amino acid residues in its c-region. It obeys the 3,1 rule with small and polar amino acid residues, Gly and Thr, at its 1 and 3 positions. The recombinant AcMNPVs for expression of EGFP with FibH SP of different­ lengths were constructed as described in aterials­ and methods. Their structures are shown in Fig. 1.


Expression of EGFP with different lengths of FibH SP in PSG of silkworm


Fusion protein of EGFP with full length FibH SP (5LEGFPhis) could be secreted to the lumen of PSG of silkworm normally. Fluorescence of EGFP was observed in the PSG lumen but not in the cells under fluorescence microscope [Fig. 2(A)]. When the SP of FibH was shortened­ stepwise to 20, 18 and 16 a.a. long, the secretion­ of the fusion proteins was not affected; and the fluo­rescence profiles were the same as the full length FibH SP fusion protein (photograph not shown). when the SP of FibH was shortened to 12 a.a., fluorescence was seen in both the lumen and PSG cells [Fig. 2(B)]. Fusion proteins with 8-11 a.a. long SP of FibH could not be secreted into the lumen, and fluorescence could only be observed in the cells [Fig. 2(C), SP11EGFP]. It was the same as the protein­ SP1EGFP without the FibH SP [Fig. 2(D)].

SDS-PAGE and western blot analysis showed that the secreted­ proteins (SP20EGFP, SP18EGFP and SP16EGFP) had a molecular weight of approximately 28 kDa; and only SP12EGFP had two bands, one of which was approximately 2 kDa larger than the other. The fusion protein with a 163 a.a. long N-terminal (5LEGFPhis) had a molecular­ weight of approximately 53 kDa (Fig. 3) as previously­ reported [10].


Protein purification and N-terminal sequencing


Two fusion proteins (SP12EGFP and SP16EGFP) were purified using the Ni-NTA system. SDS-PAGE and western­ blot analysis showed the results of purification (Fig. 4). SP12EGFP had two bands near the calculated molecular weight and some small bands that might be the products of degradation. The purified SP16EGFP protein had only one main band near the prospective place (Fig. 4). The major bands of SP16EGFP and SP12EGFP were cut for N-terminal sequencing. N-terminal sequences of these two proteins were MVSKGEELFT and EELFTGVVPI, respectively. The cleavage site was between amino acid residues 18 and 19, and behind two linker amino acid residues­ (Gly and Ser) in protein SP16EGFP; and the cleavage­ site was between Gly19 and Glu20 in protein SP12EGFP, which were Gly5 and Glu6 in EGFP. It seemed that the amino acid residues in the linker and EGFP were used as part of the SP in this case, which indicated that the cleavage site moved behind when the SP of FibH was shortened.





SPs have conserved features, and different signal peptide­ sequences through common secretory pathways can be interchanged between different proteins or even proteins of different organisms [17,18]. This work showed that the SP of FibH of silkworm shares the same structural feature with other eukaryotic secretory proteins. The h-region of the SP is essential for recognition by SRPs. When the h-domain was shortened to a limit, the SRPs could not recognize it and the fusion protein failed to secrete as observed­ in this work with SP8EGFP-SP11EGFP. The cleavage of the SP from nascent preprotein is catalyzed by signal peptidase, and the c-region of the SP is important for cleavage by signal peptidase, so a change in this region might alter the cleavage site of the SP. The h-region­ is also very important for the binding of signal peptidase. The distance between the h- and c-region boundaries to the cleavage site is required for catalyzing proper cleavage by signal peptidase. When the h-region is shortened and becomes unsuitable, signal peptidase will find a vicarious cleavage site behind if possible. In this work, with SP12EGFP, for example, the cleavage site moved to between­ Gly5 and Glu6 of EGFP .

Several programs have been developed for the pre­diction of SPs but with varying accuracy [12,19]. A comparison of the experimental results of this work and the predictive results is shown in table 2. We found that the accuracy of cleavage site prediction has been improved notably in the new version of SignalP 3.0 HMM. As predicted by SignalP 3.0 HMM, when the SP is shortened from the C-terminal then linked to EGFP through the two amino acid linker (Gly-Ser), the secretory machinery might find a suitable­ sequence as the h-region of the SP for binding of SRP, then cleave at several a.a. behind by signal peptidase. results in this experiment showed that SP20EGFP, SP18EGFP, SP16EGFP and SP12EGFP could be secreted into PSG lumen like 5LEGFPhis, but SP11EGFP, SP10EGFP, SP9EGFP and SP8EGFP could not. The probabilities of an SP for SP12EGFP and SP11EGFP predicted by SignalP 3.0 HMM were 0.875 and 0.719, respectively. The only difference between SP12EGFP and SP11EGFP is the lack of one Ala residue in the h-region of the SP, and the predicted h-region is changed from 8 (FVILCCAG) to 7 (FVILCCG). According to these results, it seems that FVILCCG is not long enough for the h-region of this SP. This comparison shows that SignalP 3.0 HMM can provide­ accurate predictions with SP20EGFP to SP12EGFP, but not with SP11EGFP. Perhaps the score of probability for SPs should be as high as with SP12EGFP. The cleavage site of signal peptide predicted by SignalP 3.0 HMM is showed by the maximal cleavage site probability (Cmax). The cleavage sites and values of Cmax for SP16EGFP and SP12EGFP predicted by SignalP 3.0 HMM are S18/M19 with a Cmax of 0.833 and G19/E20 with a Cmax of 0.686, respectively. The results of N-terminal sequencing of SP16EGFP and SP12EGFP agree with this prediction.

study of SPs and SP prediction is provoked by the increasing­ biotechnological interest in finding a suitable expression system for large-scale production of commercially interested proteins. Silk gland of silkworm Bombyx mori has long been of interest for the production of large amounts of silk protein in a tissue- and stage-specific manner. Its potential use as a bioreactor has gained more attention [20-22]. FibH makes up 70% of the total silk, so its promoter and SP are preferred for this kind of use. With the purpose of controlling the N-terminal amino acid residues of mature foreign protein expressed in silk gland using a modified FibH SP, and to characterize the functional essential sequence of the FibH SP, the SP sequence was shortened and linked with EGFP reporter. The c-region­ was mutated to two small and neutral linker amino acids, Gly and Ser, the h-region was shortened stepwise and the n-region kept compact. For SP16EGFP, the cleavage site is after Ser, just before the first amino acid residue Met of EGFP. This SP sequence might be used in bioreactors for expressing and secreting mature recombinant protein without­ any extra amino acid residues at the N-terminal.





We thank Dr. Yuan Zhao from the Sericultural Research Institute, Chinese Academy of Agricultural Sciences for kindly providing silkworm eggs and silkworm for this work.





 1   Walter P, Blobel G. Translocation of proteins across the endoplasmic reticulum III. Signal recognition protein (SRP) causes signal sequence-dependent and site-specific arrest of chain elongation that is released by microsomal membranes. J Cell Biol 1981, 91: 557-561

 2    Gierasch LM. Signal sequences. Biochemistry 1989, 28: 923-930

 3    von Heijne G. Patterns of amino acids near signal-sequence cleavage sites. Eur J Biochem 1983, 133: 17-21

 4    von Heijne G. Signal sequences. The limits of variation. J Mol Biol 1985, 184: 99-105

 5    von Heijne G. The signal peptide. J Membr Biol 1990, 115: 195-201

 6    Guo TQ, Wang JY, Wang SP, Guo XY, Huang KW, Huang JT, Lu CD. Loss of posterior silk gland transcription specificity of fibroin light chain promoter due to absence of 41 bp sequence containing possible inhibitor binding sites. Acta Biochim Biophys Sin 2005, 37: 819-825

 7    Inoue S, Tanaka K, Arisaka F, Kimura S, Ohtomo K, Mizuno S. Silk fibroin of Bombyx mori is secreted, assembling a high molecular mass elementary unit consisting of H-chain, L-chain, and P25, with a 6:6:1 molar ratio. J Biol Chem 2000, 275: 40517-40528

 8    Sasaki T, Noda H. Studies on silk fibroin of Bombyx mori directly extracted from the silk gland. III. N-terminal analysis and degradation in slightly alkaline solution. J Biochem (Tokyo) 1974, 76: 493-502

 9    Zhou CZ, Confalonieri F, Medina N, Zivanovic Y, Esnault C, Yang T, Jacquet M et al. Fine organization of Bombyx mori fibroin heavy chain gene. Nucleic Acids Res 2000, 28: 2413-2419

10   Wang SP, Guo TQ, Guo XY, Huang JT, Lu CD. In vivo analysis of fibroin heavy chain signal peptide of silkworm Bombyx mori using recombinant baculovirus as vector. Biochem Biophys Res Commun 2006, 341: 1203-1210

11   Nielsen H, Engelbrecht J, Brunak S, von Heijne G. Identification of prokaryotic and eukaryotic signal peptides and prediction of their cleavage sites. Protein Eng 1997, 10: 1-6

12   Bendtsen JD, Nielsen H, von Heijne G, Brunak S. Improved prediction of signal peptides: SignalP 3.0. J Mol Biol 2004, 340: 783-795

13   Guo TQ, Wang JY, Guo XY, Wang SP, Lu CD. Transient in vivo gene delivery to the silkworm Bombyx mori by EGT-null recombinant AcNPV using EGFP as a reporter. Arch Virol 2005, 150: 93-105

14   Guo XY, Guo TQ, Wang SP, Wang JY, Lu CD. Silk gland specific secretory expression of egfp gene in silkworm Bombyx mori with rAcMNPV system. Arch Virol 2005, 150: 1151-1160

15   Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970, 227: 680-685

16   Matsudaira P. Sequence from picomole quantities of proteins electroblotted onto polyvinylidene difluoride membranes. J Biol Chem 1987, 262: 10035-10038

17   Martoglio B, Dobberstein B. Signal sequences: More than just greasy peptides. Trends Cell Biol 1998, 8: 410-415

18   von Heijne G, Abrahmsen L. Species-specific variation in signal peptide design. Implications for protein secretion in foreign hosts. FEBS Lett 1989, 244: 439-446

19   Menne KM, Hermjakob H, Apweiler R. A comparison of signal sequence prediction methods using a test set of signal peptides. Bioinformatics 2000, 16: 741-742

20   Inoue S, Kanda T, Imamura M, Quan GX, Kojima K, Tanaka H, Tomita M et al. A fibroin secretion-deficient silkworm mutant, Nd-sD, provides an efficient system for producing recombinant proteins. Insect Biochem Mol Biol 2005, 35: 51-59

21   Royer C, Jalabert A, Da Rocha M, Grenier AM, Mauchamp B, Couble P, Chavancy G. Biosynthesis and cocoon-export of a recombinant globular protein in transgenic silkworms. Transgenic Res 2005, 14: 463-472

22 Tomita M, Munetsuna H, Sato T, Adachi T, Hino R, Hayashi M, Shimizu K et al. Transgenic silkworms produce recombinant human type III procollagen in cocoons. Nat Biotechnol 2003, 21: 52-56