Crystal structure of a fungal protease inhibitor from Antheraea mylitta Sobhan Roy a,1 , Penmatsa Aravind b,1 , Chaithanya Madhurantakam a,2 , Ananta Kumar Ghosh a , Rajan Sankaranarayanan b, * , Amit Kumar Das a, * a Department of Biotechnology, Indian Institute of Technology, Kharagpur 721302, India b Centre for Cellular and Molecular Biology, Council of Scientific and Industrial Research, Hyderabad, India article info Article history: Received 3 October 2008 Received in revised form 19 December 2008 Accepted 20 December 2008 Available online 19 January 2009 Keywords: Antheraea mylitta Canonical protease inhibitor Reactive site Disulfide linkage Fungal protease inhibitor abstract Indian tasar silk is produced by a wild insect called Antheraea mylitta. Insects do not have any antigen– antibody mediated immune system like vertebrates but they produce a wide variety of effector proteins and peptides possessing potent antifungal and antibacterial activity to combat microbial attack. Anthe- raea mylitta expresses a fungal protease inhibitor AmFPI-1, in the hemolymph that inhibits alkaline pro- tease of Aspergillus oryzae for protection against fungal infection. AmFPI-1 is purified from the hemolymph, crystallized and the structure is solved using the single isomorphous replacement with anomalous scattering (SIRAS) method to a resolution of 2.1 Å. AmFPI-1 is a single domain protein pos- sessing a unique fold that consists of three helices and five b strands stabilized by a network of six disul- fide bonds. The reactive site of AmFPI-1 is located in the loop formed by residues 46–66, wherein Lys54 is the P 1 residue. Superimposition of the loop with reactive sites of other canonical protease inhibitors shows that reactive site conformation of AmFPI-1 is similar to them. The structure of AmFPI-1 provides a framework for the docking of a 1:1 complex between AmFPI-1 and alkaline protease. This study addresses the structural basis of AmFPI-1’s specificity towards a fungal serine protease but not to mam- malian trypsin and may help in designing specific inhibitors against fungal proteases. Ó 2009 Elsevier Inc. All rights reserved. 1. Introduction Proteases are ubiquitous in living systems and show an intimate relationship with proteins from synthesis to their breakdown. Pro- teases take part in various physiological processes such as food digestion, blood clotting, clot dissolution, embryogenesis, tissue regeneration, defense mechanisms and immune responses (Las- kowski et al., 2000). In general, protease activity is modulated by means of regulated expression/secretion, proteolytic activation of their inactive forms, auto-inactivation or autolysis, transportation and inhibition by specific protease inhibitors (Bode and Huber, 2000; Fodor et al., 2005). Hence protease inhibitors play an impor- tant role in all living organisms through regulation of proteolysis. During the last two decades, a number of protease inhibitors have been studied to understand their physiological role and their po- tential use as therapeutic agents. Multiple families (48) of structurally distinct protease inhibi- tors have been identified with the capability to inhibit serine, cys- teine, metallo and aspartyl proteases (Rawlings et al., 2004). These inhibitors interact with proteases through various structural ele- ments, such as N- or C-terminal exposed loops, either separately or in association with other similar structural elements (Otlewski et al., 2005). Serine protease inhibitors are classified into three cat- egories: canonical, non-canonical and serpins. Canonical inhibitors form the largest family and act through a standard mechanism of inhibition (Laskowski and Kato 1980; Krowarsch et al., 2003). These inhibitors are rigid, stable and mostly composed of purely b-sheet or mixed a/b topologies. However, some canonical prote- ase inhibitors are only a-helical or form irregular structures rich in disulfide bridges (Otlewski et al., 2005). About 18 families of canonical inhibitors have been identified which show a common three-dimensional architecture forming the exposed loop sur- rounding the reactive site. These residues, in the reactive loop have a b-strand architecture and the P 1 residue (Schechter and Berger nomenclature, Schechter and Berger, 1967) has a 3 10 -helical con- formation. Inhibition is achieved through several intermolecular interactions between the protease active site and the inhibitor reactive loop: formation of two hydrogen bonds between the car- bonyl oxygen of P 1 and the amides of the oxyanion hole, and a short contact between the P 1 carbonyl carbon and the catalytic ser- ine (Otlewski et al., 2005). Non-canonical protease inhibitors, like hirudin and haemedin interact with the active site of serine prote- ases like thrombin, through their N-terminal tails in such a manner that the N-terminal tail is inserted into the enzyme active site, 1047-8477/$ - see front matter Ó 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.jsb.2008.12.010 * Corresponding authors. Fax: +91 40 27160591 (R. Sankaranarayanan), +91 3222 278707 (A.K. Das). E-mail addresses: sankar@ccmb.res.in (R. Sankaranarayanan), amitk@hijli. iitkgp.ernet.in (A.K. Das). 1 These authors contributed equally. 2 Present address: Center for Infectious Medicine, Department of Medicine, F59, Karolinska University Hospital, Hudding, Stockholm, 141 86, Sweden. Journal of Structural Biology 166 (2009) 79–87 Contents lists available at ScienceDirect Journal of Structural Biology journal homepage: www.elsevier.com/locate/yjsbi