Abstract

Key words

Introduction

Serine proteinase inhibitors are widespread in the plant kingdom, their physiological roles including the regulation of endogenous proteinases during seed dormancy, the reserve protein mobilisation, and the protection against the proteolytic enzymes of parasites and insects. Moreover, they may also act as storage or reserve proteins. Plant serine proteinase inhibitors are grouped into Soybean (Kunitz), Bowman-Birk, potato I and II, and squash families. Several other inhibitor families, such as barley, ragi 1 and 2, and thaumatin, were also suggested (Laskowski & Kato, 1980; Ryan, 1990; Ascenzi et al, 1999).

In the oil-rape (Brassica napus var. oleifera) seed, three distinct classes of proteinase inhibitors (RTI-I, RTI-II and RTI-III) have been identified. In particular, RTI-I, RTI-II and RTI-III represent 4-10%, 13-33% and 60-85% of the total serine proteinase inhibitory activity, respectively. RTI-I and RTI-II belong to the Soybean family, are thermolabile, show a molecular mass of about 19 kDa, and inactivate selectively trypsin. On the other hand, RTI-III is the prototype of a new family of serine proteinase inhibitors from Cruciferae. Indeed, proteinase inhibitors almost identical to RTI-III were also purified from white mustard (Sinapis alba L.) seed (MTI-2) and identified in the Arabidopsis thaliana transcribed genome. Moreover, RTI-III is thermostable, shows a molecular mass of 6.5 kDa, and inactivates both trypsin and chymotrypsin. The amino acid sequence of the 5-oxoPro1-Gly62-RTI-III isoinhibitor contains eight cysteine residues, forming four disulphide bridges Cys5-Cys27, Cys18-Cys31, Cys42-Cys52, and Cys54-Cys57. The reactive site loop of RTI-III isoinhibitors encompasses the Cys18(P₄)-Cys27(P₆’) amino acid sequence. The Arg21 and Ile22 side chains correspond to the P₁ and P₁’ inhibitor reactive site residues, interacting with the S₁ and S₁’ subsites of trypsin and chymotrypsin (e.g. form the potentially scissile P₁-P₁’ inhibitor peptide bond) (Ascenzi et al, 1999).

The present study reports the predicted three-dimensional model of the 5-oxoPro1-Gly62-RTI-III isoinhibitor, built on the basis of the snake venom toxin structure.

Model building of the 5-oxoPro1-Gly62-RTI-III isoinhibitor structure

The amino acid sequence and the disulphide bridge pattern of the 5-oxoPro1-Gly62-RTI-III isoinhibitor have been previously reported (Ascenzi et al, 1999). In the present study, the N-terminal residue of the 5-oxoPro1-Gly62-RTI-III isoinhibitor is considered Pro instead 5-oxoPro. The three-dimensional model of the 5-oxoPro1-Gly62-RTI-III isoinhibitor has been built using the program Modeler (Sali & Blundell, 1993), in the Insight II suite (Insight II User Guide, 1997). This procedure calculates a three-dimensional model by satisfaction of spatial restraints derived from a set of homologous template structures. The three-dimensional model of the 5-oxoPro1-Gly62-RTI-III isoinhibitor was based on erabutoxin b from sea snake (PDB ID: 3EBX; Hatanaka et al, 1994), fasciculin 1 from green mamba (PDB ID: 1FAS; le Du et al, 1992), and cardiotoxin γ from Naja nigricollis (PDB ID: 1TGX; Gilquin et al, 1993). Template amino acid sequences have been aligned according to literature (le Du et al, 1992). The 5-oxoPro1-Gly62-RTI-III isoinhibitor target sequence (Ascenzi et al, 1999) has been manually aligned to the template toxin sequences according to the observed pattern of disulphide bridges (see Fig. 1).

The three-dimensional model of the putative porcine pancreatic trypsin:5-oxoPro1-Gly62-RTI-III isoinhibitor binary complex was also predicted starting from the X-ray crystallographic structure of the serine proteinase:soybean trypsin inhibitor adduct (PDB ID: 1AVW; Song & Suh, 1998) and from the three-dimensional model of the free 5-oxoPro1-Gly62-RTI-III isoinhibitor.

The initial three-dimensional models of the free 5-oxoPro1-Gly62-RTI-III isoinhibitor and of the porcine pancreatic trypsin:inhibitor binary complex have been optimised using molecular dynamics, energy minimization, and manual refinements. Energy minimization and molecular dynamics were performed with the program Discover (Insight II User Guide, 1997), using the consistent valence force field, a dielectric constant dependent from distance to simulate the effect of solvent, and no cross terms. Stereochemical properties of the final three-dimensional models have been assessed with Procheck (Laskowski et al, 1993). All programs have been run under IRIX 6.2 on an O2 Silicon Graphics station.

The three-dimensional model of the 5-oxoPro1-Gly62-RTI-III Isoinhibitor

The 5-oxoPro1-Gly62-RTI-III isoinhibitor shows a high sequence similarity with the Asp1-Gln63-MTI-2 isoinhibitor (69%; see Fig. 1), and the Asn10-Pro68 (70%), Asp28-Cys80 (71%), and Ser30-Cys81 (71%) regions of the putative trypsin inhibitors which have been identified in the Arabidopsis thaliana transcribed genome. The P₁ and P₁’ reactive site residues of the serine proteinase isoinhibitors, forming the scissile peptide bond, are Arg and Ile (Ascenzi et al, 1999). Althought snake venom toxins show a low sequence similarity with serine proteinase inhibitors (<20% identity), some structurally relevant residues (i.e. Cys5, Gly10, Gly11, Cys18, Pro20, Cys27, Cys31, Gly40, Cys42, Gly45, Gly47 Cys52, Cys54 and Cys57 present in the 5-oxoPro1-Gly62-RTI-III isoinhibitor) appear to be conserved (Ascenzi et al, 1999). In particular, the Cys5-Cys27, Cys18-Cys31, Cys42-Cys52 and Cys54-Cys57 disulphide bridge pattern occurring in the 5-oxoPro1-Gly62-RTI-III isoinhibitor is reminiscent of that found in snake venom toxins (le Du et al, 1992; Gilquin et al, 1993; Hatanaka et al, 1994; Ascenzi et al, 1999) (see Fig. 1).

Figure 1: Amino acid sequences of 5-oxoPro1-Gly62-RTI-III isoinhibitor (A; Ascenzi et al, 1999), Asp1-Gln63-MTI-2 isoinhibitor (B; Ascenzi et al, 1999), erabutoxin b (C; Hatanaka et al, 1994), fasciculin 1 (D; le Du et al, 1992), and cardiotoxin γ (E; Gilquin et al, 1993). The arrow indicates the serine proteinase inhibitor Arg(P₁)-Ile(P₁’) scissile reactive site bond (residues are underlined). The disulphide bridge pattern of the 5-oxoPro1-Gly62-RTI-III isoinhibitor (A) is shown. Cys residues are in bold. Amino acid code of every tenth residue of the 5-oxoPro1-Gly62-RTI-III isoinhibitor (A) is represented in italic.

Figure 2: Stereo view of the three-dimensional model of the 5-oxoPro1-Gly62-RTI-III isoinhibitor. White sticks denote disulphide bridges. The Arg21(P₁) and Ile22(P₁’) residues are at the top of the picture. Amino acid residues are labelled with the one-letter code and numbered every ten positions. The picture was created with MOLSCRIPT (Kraulis, 1991).

The three-dimensional model of the 5-oxoPro1-Gly62-RTI-III isoinhibitor (see Fig. 2) has been built on the basis of the similarity of the connectivity of disulfide bridges to those of the snake venom toxins (see Fig. 1). The sequence alignment displays a large deletion in the finger 2 of snake venom toxin structures (see Fig. 1) which is reflected in the absence of a large β-sheet in the 5-oxoPro1-Gly62-RTI-III isoinhibitor model (see Fig. 2). In fact, only two short two-stranded sheets are predicted, including Leu6 and Phe15 in the finger I, and Cys27 and Cys54 in the finger II. The finger III appears not to contain any secondary structure due to the presence of the large insertion relatively to the templates in position Ala33-Lys41 (see Fig. 1) which disrupts the backbone geometry observed in snake venom toxins (le Du et al, 1992; Gilquin et al, 1993; Hatanaka et al, 1994; Ascenzi et al, 1999).

In the 5-oxoPro1-Gly62-RTI-III isoinhibitor, hydrophobic interactions are observed between residues Pro1-Phe17-Phe56, Leu6-Phe15, Pro20-Ala33, Phe26-Pro24, and Val28-Ile43-Val50 (see Fig. 2). Interestingly, Val28 of the 5-oxoPro1-Gly62-RTI-III isoinhibitor seems to play a role similar to the structurally equivalent Tyr25 in erabutoxin b (Hatanaka et al, 1994), Tyr23 in fasciculin 1 (le Du et al, 1992) and in Tyr22 in cardiotoxin γ (Gilquin et al, 1993). The Val28 residue of the 5-oxoPro1-Gly62-RTI-III isoinhibitor is replaced by Tyr in the Asp1-Gln63-MTI-2 isoinhibitor and in the putative trypsin inhibitors which have been identified in the Arabidopsis thaliana transcribed genome (Ascenzi et al, 1999).

The presence of a hydrophobic cluster, organised around a conserved aliphatic or aromatic residue, may play a role in the stability of the β-sheet structure, allowing more flexible regions of the macromolecule, such as the Cys18(P₄)-Cys27(P₆’) reactive site loop of the 5-oxoPro1-Gly62-RTI-III isoinhibitor, to be solvent exposed (see Fig. 2). As already reported for serine proteinase inhibitor reactive sites (Bode & Huber, 1992), this region is strongly connected to the protein core by Cys5-Cys27 and Cys18-Cys31 disulphide bridges. Moreover, Pro20(P₂) and Pro24(P₃’) of the 5-oxoPro1-Gly62-RTI-III isoinhibitor reactive site may play a central role in the achievement of proper conformation for the P₁-P₁’ scissile peptide bond. Thus, the 5-oxoPro1-Gly62-RTI-III isoinhibitor Arg21(P₁)-Ile22(P₁’) scissile peptide bond is in a solvent-exposed region, which corresponds to turn 2 in toxins. However, in toxins, amino acid residues involved in protein-receptor recognition (turns 3 and 4) are located on the opposite site with respect to turn 2 (le Du et al, 1992; Gilquin et al, 1993; Hatanaka et al, 1994). The inhibitor Arg21(P₁)-Ile22(P₁’) scissile peptide bond sequence is consistent with the specificity for trypsin-like serine proteinases, possibly representing the binding site also for chymotrypsin-like enzymes (Ascenzi et al, 1999).

Structural data here reported (see Fig. 2) suggest that the overall architecture of the 5-oxoPro1-Gly62-RTI-III isoinhibitor is reminiscent of that of snake venom toxins, consisting of three-finger shaped loops extending from the cross-over region, and including a two- and a three-stranded anti parallel β-sheets (le Du et al, 1992; Gilquin et al, 1993; Hatanaka et al, 1994). Thus, in the 5-oxoPro1-Gly62-RTI-III isoinhibitor, Cys5-Cys27 and Cys18-Cys31 disulphides link finger I to II, the Cys42-Cys52 disulphide connects finger II to III, and the Cys54-Cys57 disulphide links the C-terminus to finger III. Moreover, the inhibitor cross-over region stabilised by the Cys5-Cys27 and Cys18-Cys31disulphide bridges may represent the stable T-knot scaffold (Ascenzi et al, 1998), around which different three-dimensional structures could be organised.

The three-dimensional model of the porcine pancreatic trypsin:5-oxoPro1-Gly62-RTI-III isoinhibitor binary complex indicates that the inhibitor reactive site fits the enzyme active centre in a way comparable to that observed in the starting serine proteinase:soybean trypsin inhibitor binary adduct (Song & Suh, 1998). In particular, the salt bridge occurring at the enzyme primary specificity pocket between the inhibitor Arg21(P₁) and the serine proteinase Asp189 residues is mantained. Also, the interaction between the carbonyl oxygen of the inhibitor scissile peptide bond Arg21(P₁)-Ile22(P₁’) and the enzyme Ser195 residue is observed. Moreover, hydrophobic contacts occur between the serine proteinase Leu99 and the inhibitor Pro20 residues, as well as between the enzyme Phe41 and the inhibitor Tyr23 chains.

As a whole, the similar three-dimensional architecture of serine proteinase inhibitors from Cruciferae and snake venom toxins, involved in host-parasite and/or prey-predator interaction(s), can shed light on the molecular evolution and ecology of interactions among species.

Acknowledgements

Authors thank Dr. Margherita Ruoppolo and Dr. Lucia Zetta for helpful discussions. This study was partially supported by grants from the Italian Ministry of University, Scientific Research and Technology (target oriented project ‘Biocatalisi e Bioconversioni’), from the Italian Ministry of Agriculture, Food and Forestry Resources (target oriented project ‘Resistenze Genetiche delle Piante Agrarie agli Stress Biotici ed Abiotici’), and from the Italian National Research Council (target oriented project ‘Biotecnologie’).

References

Ascenzi, P., Bolognesi, M., Catalucci, D., Pascarella, S., Ruoppolo, M. & Rizzi, M. (1998) Leech antihemostatic proteins share the T-knot scaffold, a disulphide-reinforced structural motif, Biol. Chem. Hoppe Seyler 379, 1387-1389.

Ascenzi, P., Ruoppolo, M., Amoresano, A., Pucci, P., Consonni, R., Zetta, L., Pascarella, S., Bortolotti, F. & Menegatti, E. (1999) Characterisation of low molecular mass trypsin isoinhibitors from oil-rape (Brassica napus var. oleifera) seed, Eur. J. Biochem., in press.

Bode, W. & Huber, R. (1992) Natural protein proteinase inhibitors and their interaction with proteinases, Eur. J. Biochem. 204, 433-451.

Gilquin, B., Roumestand, C., Zinn-Justin, S., Ménez, A. & Toma F. (1993) Refined three dimensional solution structure of a snake cardiotoxin: analysis of the side-chain organization suggests the existence of a possible phospholipid binding site, Biopolymers 33, 1659-1675.

Hatanaka, H., Oka, M., Kohda, D., Tate, S.-i., Suda, A., Tamiya, N. & Inagaki, F. (1994) Tertiary structure of erabutoxin b in aqueous solution as elucidated by two-dimensional nuclear magnetic resonance, J. Mol. Biol. 240, 155-166.

Kraulis P.J. (1991) MOLSCRIPT: a program to produce both detailed and schematic plots of protein structures, J. Appl. Cryst. 24, 946-950.

Laskowski, M. Jr. & Kato, I. (1980) Protein inhibitors of proteinases, Annu. Rev. Biochem. 49, 593-626.

Laskowski, R.A., MacArthur, M.W., Moss, D.S. & Thornton, J.M. (1993) PROCHECK: a program to check the stereochemical quality of protein structures, J. Appl. Cryst. 26, 283-291.

le Du, M.H., Marchot, P., Bougis, P.E. & Fontecilla-Camps, J.C. (1992) 1.9 Ĺ resolution structure of fasciculin 1, an anti-acetylcholinesterase toxin from green mamba snake venom, J. Biol. Chem. 267, 22122-22130.

Ryan, C.A. (1990) Protease inhibitors in plants: genes for improving defences against insects and pathogens, Annu. Rev. Phytopathol. 28, 425-449.

Sali, A. & Blundell, T.L. (1993) Comparative protein modelling by satisfaction of spatial restraints, J. Mol. Biol. 234, 779-815.

Song, H.K. & Suh, S.W. (1998) Kunitz-type soybean trypsin inhibitor revisited: refined structure of its complex with porcine trypsin reveals an insight into the interaction between a homologous inhibitor from Erythrina caffra and tissue-type plasminogen activator, J. Mol. Biol. 275, 347-363.