Crystal structure of a clip-domain serine protease and functional roles of the clip domains
2005; Springer Nature; Volume: 24; Issue: 24 Linguagem: Inglês
10.1038/sj.emboj.7600891
ISSN1460-2075
AutoresShunfu Piao, Young-Lan Song, Jung Hyun Kim, Sam Yong Park, Ji Won Park, Bok Leul Lee, Byung‐Ha Oh, Nam‐Chul Ha,
Tópico(s)Neurobiology and Insect Physiology Research
ResumoArticle15 December 2005free access Crystal structure of a clip-domain serine protease and functional roles of the clip domains Shunfu Piao Shunfu Piao National Research Laboratory of Defense Proteins, College of Pharmacy and Research Institute for Drug Development, Pusan National University, Jangjeon Dong, Geumjeong Gu, Busan, Korea Search for more papers by this author Young-Lan Song Young-Lan Song Center for Biomolecular Recognition and Division of Molecular and Life Sciences, Department of Life Sciences, Pohang University of Science and Technology, Pohang, Kyungbuk, Korea Search for more papers by this author Jung Hyun Kim Jung Hyun Kim National Research Laboratory of Defense Proteins, College of Pharmacy and Research Institute for Drug Development, Pusan National University, Jangjeon Dong, Geumjeong Gu, Busan, Korea Search for more papers by this author Sam Yong Park Sam Yong Park Protein Design Laboratory, Yokohama City University, Suehiro, Tsurumi-ku, Yokohama, Japan Search for more papers by this author Ji Won Park Ji Won Park National Research Laboratory of Defense Proteins, College of Pharmacy and Research Institute for Drug Development, Pusan National University, Jangjeon Dong, Geumjeong Gu, Busan, Korea Search for more papers by this author Bok Leul Lee Corresponding Author Bok Leul Lee National Research Laboratory of Defense Proteins, College of Pharmacy and Research Institute for Drug Development, Pusan National University, Jangjeon Dong, Geumjeong Gu, Busan, Korea Search for more papers by this author Byung-Ha Oh Corresponding Author Byung-Ha Oh Center for Biomolecular Recognition and Division of Molecular and Life Sciences, Department of Life Sciences, Pohang University of Science and Technology, Pohang, Kyungbuk, Korea Search for more papers by this author Nam-Chul Ha Corresponding Author Nam-Chul Ha National Research Laboratory of Defense Proteins, College of Pharmacy and Research Institute for Drug Development, Pusan National University, Jangjeon Dong, Geumjeong Gu, Busan, Korea Search for more papers by this author Shunfu Piao Shunfu Piao National Research Laboratory of Defense Proteins, College of Pharmacy and Research Institute for Drug Development, Pusan National University, Jangjeon Dong, Geumjeong Gu, Busan, Korea Search for more papers by this author Young-Lan Song Young-Lan Song Center for Biomolecular Recognition and Division of Molecular and Life Sciences, Department of Life Sciences, Pohang University of Science and Technology, Pohang, Kyungbuk, Korea Search for more papers by this author Jung Hyun Kim Jung Hyun Kim National Research Laboratory of Defense Proteins, College of Pharmacy and Research Institute for Drug Development, Pusan National University, Jangjeon Dong, Geumjeong Gu, Busan, Korea Search for more papers by this author Sam Yong Park Sam Yong Park Protein Design Laboratory, Yokohama City University, Suehiro, Tsurumi-ku, Yokohama, Japan Search for more papers by this author Ji Won Park Ji Won Park National Research Laboratory of Defense Proteins, College of Pharmacy and Research Institute for Drug Development, Pusan National University, Jangjeon Dong, Geumjeong Gu, Busan, Korea Search for more papers by this author Bok Leul Lee Corresponding Author Bok Leul Lee National Research Laboratory of Defense Proteins, College of Pharmacy and Research Institute for Drug Development, Pusan National University, Jangjeon Dong, Geumjeong Gu, Busan, Korea Search for more papers by this author Byung-Ha Oh Corresponding Author Byung-Ha Oh Center for Biomolecular Recognition and Division of Molecular and Life Sciences, Department of Life Sciences, Pohang University of Science and Technology, Pohang, Kyungbuk, Korea Search for more papers by this author Nam-Chul Ha Corresponding Author Nam-Chul Ha National Research Laboratory of Defense Proteins, College of Pharmacy and Research Institute for Drug Development, Pusan National University, Jangjeon Dong, Geumjeong Gu, Busan, Korea Search for more papers by this author Author Information Shunfu Piao1, Young-Lan Song2, Jung Hyun Kim1, Sam Yong Park3, Ji Won Park1, Bok Leul Lee 1, Byung-Ha Oh 2 and Nam-Chul Ha 1 1National Research Laboratory of Defense Proteins, College of Pharmacy and Research Institute for Drug Development, Pusan National University, Jangjeon Dong, Geumjeong Gu, Busan, Korea 2Center for Biomolecular Recognition and Division of Molecular and Life Sciences, Department of Life Sciences, Pohang University of Science and Technology, Pohang, Kyungbuk, Korea 3Protein Design Laboratory, Yokohama City University, Suehiro, Tsurumi-ku, Yokohama, Japan *Corresponding authors: National Research Laboratory of Defense Proteins, College of Pharmacy and Research Institute for Drug Development, Pusan National University, Jangjeon Dong, Geumjeong Gu, Busan 607-735, Korea. Tel.: +82 51 510 2528 or 2809; Fax: +82 51 513 6754; E-mail: [email protected] or E-mail: [email protected] for Biomolecular Recognition and Division of Molecular and Life Sciences, Department of Life Sciences, Pohang University of Science and Technology, Pohang, Kyungbuk, Korea. Tel.: +82 54 279 2289; Fax: +82 54 279 2199; E-mail: [email protected] The EMBO Journal (2005)24:4404-4414https://doi.org/10.1038/sj.emboj.7600891 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info Clip-domain serine proteases (SPs) are the essential components of extracellular signaling cascades in various biological processes, especially in embryonic development and the innate immune responses of invertebrates. They consist of a chymotrypsin-like SP domain and one or two clip domains at the N-terminus. Prophenoloxidase-activating factor (PPAF)-II, which belongs to the noncatalytic clip-domain SP family, is indispensable for the generation of the active phenoloxidase leading to melanization, a major defense mechanism of insects. Here, the crystal structure of PPAF-II reveals that the clip domain adopts a novel fold containing a central cleft, which is distinct from the structures of defensins with a similar arrangement of cysteine residues. Ensuing studies demonstrated that PPAF-II forms a homo-oligomer upon cleavage by the upstream protease and that the clip domain of PPAF-II functions as a module for binding phenoloxidase through the central cleft, while the clip domain of a catalytically active easter-type SP plays an essential role in the rapid activation of its protease domain. Introduction Serine protease (SP) cascades amplify signals from physiological or pathological responses in the extracellular region of vertebrates and invertebrates. In mammals, the blood clotting and the complement system employ SP cascades in response to tissue damage and microbial infection, respectively (O'Brien and McVey, 1993). In invertebrates, SP cascades drive diverse biological processes, including embryonic development and immune responses (Krem and Cera, 2002). Many of these signal-amplifying reactions are accomplished by the clip domain-containing SP family, the number of which is expanding rapidly in public genome databases. However, the cellular function and the structure of any clip domain have been unknown. Clip-domain SPs were first identified as enzymes involved in the Toll signaling pathway required for the development of dorsal–ventral structures in Drosophila (Morisato and Anderson, 1995; Anderson, 1998). So far, these proteases have been found only in invertebrates. They consist of a chymotrypsin-like SP domain and one or two clip domain(s) at the N-terminus. The SP domain is responsible for signal amplification by cleaving and activating the downstream SP. The clip domain, usually composed of 37–55 amino-acid residues, is interlinked by three strictly conserved disulfide bonds (Jiang and Kanost, 2000) and connected to the SP domain by a linker of 23–92 residues. Like chymotrypsinogen, the SP domain of the clip-domain proteases begins with the Cys–Gly sequence, and this first cysteine residue forms a disulfide bond with a cysteine residue within the SP domain. The zymogen form of these proteases is converted into the active enzyme by cleavage at the arginine or lysine residue corresponding to Arg15 of chymotrypsinogen, which is located between the two cysteine residues forming the disulfide bond. Owing to this disulfide linkage, the N-terminal fragment including the clip domain remains covalently attached to the SP domain when the zymogens are activated. Prophenoloxidase (ProPO) activation is one of the major immune responses in arthropods (Gillespie et al, 1997; Cerenius and Söderhäll, 2004). Upon injury or infection, proPO in the blood plasma is activated by clip-domain SPs, the so-called proPO-activating factors or enzymes (PPAFs or PPAEs), also known as proPO-activating proteins (PAPs) (Ashida and Brey, 1998; Lee et al, 1998a; Jiang and Kanost, 2000). Recognition of aberrant surfaces or foreign invaders triggers a cascade of PPAFs (Cerenius and Söderhäll, 2004), leading to the activation of proPO(s) (Ashida and Brey, 1998). Phenoloxidase (PO), the active form of proPO, catalyzes the production of quinones, which can crosslink neighboring molecules to form melanin at the injury site or around invading microorganisms. Quinones may also be involved in the production of cytotoxic molecules such as superoxides and hydroxyl radicals that could help to kill the invading microorganisms (Gillespie et al, 1997). The activation of proPO(s), however, needs to be tightly regulated as a local event, because excessive melanization could also cause fatal damages to hosts (Yu et al, 2003; Zhao et al, 2005). Consistently, melanization takes place locally at the site of injury or invading organisms (Söderhäll and Cerenius, 1998). PPAFs have been identified in several insects and a crustacean, including the beetle Holotrichia diomphalia (Kwon et al, 2000), the mealworm Tenebrio molitor (Lee et al, 2002), the silkworm Bombyx mori (Satoh et al, 1999), the tobacco hornworm Manduca sexta (Yu et al, 2003), and the crayfish Pacifastacus leniusculus (Wang et al, 2001). PPAFs can be classified into catalytic and noncatalytic groups, based on the presence or absence of proteolytic activity. The catalytic group (referred to as the easter-type SPs) containing one or two clip domain(s) is structurally related to Drosophila easter, and includes H. diomphalia PPAF-I, -III, B. mori PPAE, and M. sexta PAP-I. The noncatalytic group (referred to as PPAF-II family) contains one clip domain and shares a similar sequence with the catalytic group, but lacks proteolytic activity due to the replacement of the active site serine residue by glycine. It includes PPAF-II of H. diomphalia, serine proteinase homolog (SPH)-I, -II of M. sexta (Jiang et al, 2003), and Drosophila CG5390. Three PPAFs, PPAF-I, -II, and -III of H. diomphalia, were cloned and characterized (Lee et al, 1998a, 1998b; Kwon et al, 2000; Kim et al, 2002). A two-step cleavage mechanism for the activation of proPOs (two highly homologous isozymes proPO-I and -II) was put forth, in which full-length proPOs (79 kDa, ProPO79s) are cleaved into smaller forms (76 kDa, PO76s) by PPAF-I, and one of the smaller forms is converted into the active form of PO (66 kDa, PO66) by PPAF-I or unknown proteases (Kim et al, 2002). In this process, PPAF-III is the upstream protease that cleaves PPAF-II specifically at an arginine residue within the clip domain, and thereby generates the functional form of the protein. PPAF-II is an essential factor for the activation of PO despite the absence of an enzymatic activity (Lee et al, 1998b; Kim et al, 2002). The PPAF-II orthologs are found in a variety of insects, including Drosophila, H. diomphalia, T. molitor, and M. sexta (Lee et al, 1998b, 2002; Yu et al, 2003). However, the detailed mechanism of action of the PPAF-II family in the proPO activation pathway remains poorly understood. To begin understanding the functions of the clip domains, we determined the crystal structure of PPAF-II from H. diomphalia. Structure-based biochemical analyses provided insights into the roles of the clip domains of the noncatalytic and the catalytic SPs in the activation of proPOs and PPAFs. Results Structure determination and overall structure of PPAF-II Initially, we tried to produce easter (Morisato and Anderson, 1995; Anderson, 1998) and persephone (Ligoxygakis et al, 2002) of Drosophila and PPAF-I, -II, and -III of H. diomphalia in insect cells (Piao et al, 2005). Easter and persephone could not be expressed, but all PPAFs were successfully obtained in soluble forms with yields of 1–10 mg per 1 l of High-Five cell culture, which enabled us to attempt the initial screening of crystallization conditions. Only PPAF-II could be crystallized (Piao et al, 2005), and its crystal structure was determined by the molecular replacement method. The crystal-packing interactions between PPAF-II molecules were not extensive, indicating that the protein is monomeric in solution. The PPAF-II gene encodes a protein with 415 amino acids. The N-terminal 24-amino-acid segment of the protein functions as a signal peptide for secretion. The mature protein (392 amino acids) is made up of two parts: the clip domain including the flanking sequence composed of 114 amino acids, and the SP-like (SPL) domain composed of 278 amino acids (Figure 1). The final 2.0 Å resolution model contains residues 22–42 with an N-glycosylated residue, 47–116, 129–173, and 180–415, which include most of the clip and the SPL domains (Figure 2A and B). The missing residues constitute the loop segments flanking the clip domain and a flexible loop in the SPL domain. The small clip domain is bound to the larger SPL domain via hydrogen bonds, hydrophobic interactions, and a salt bridge (Figures 2A and 4B). While the structure of the SPL domain of PPAF-II is similar to chymotrypsin-like proteases (Figure 2C), that of the clip domain is a new fold according to a database search using the program DALI (Holm and Sander, 1993). Figure 1.Sequence alignment and secondary structure assignment. The sequences of the mature forms of four PPAF-II family members, four easter-type SPs, and human chymotrypsin from top to bottom: Hd, Holotrichia diomphalia; Tm-mas, Tenebrio molitor masquerade; Dm, Drosophila; Ms, Manduca sexta; Hs, Homo sapiens are shown. The alignment was performed using CLUSTALW (Thompson et al, 1997), and then adjusted based on the conserved cysteine residues. The secondary structure of PPAF-II is shown above the sequences. The linker between the clip and SPL domains is shown as a blue line on the secondary structure. The red box indicates the clip domain whose disulfide bond linkages are drawn on the secondary structure. The PPAF-II family-specific N-terminal residues are shaded in yellow. Four signature sequences of the PPAF-II family are shaded in magenta and labeled. Conserved amino acids are color coded: blue letters, conserved in the both PPAF-II family and easter-type SPs; magenta, conserved only in the PPAF-II family; green, conserved only in the easter-type SPs. N-glycosylated asparagine is indicated by a hexagon with a tail, the catalytic triad residues (His, Asp, and Ser) by red stars, the cleavage sites by inverted triangles, and the strictly conserved tyrosine residues (Tyr108 and Tyr301) only in the PPAF-II family by blue circles. Download figure Download PowerPoint Figure 2.Overall structure of PPAF-II. (A) A ribbon representation of PPAF-II structure. The clip domain is shown in red, and the PPAF-II family-specific N-terminal segment is in yellow. The SP domain is shown in cyan except for the four signature sequences that are in magenta. Bound calcium ion is in green, and all the three disulfide bonds are drawn with sulfur atoms in yellow. The disordered regions are shown in dashed lines. The nonfunctional catalytic triad residues (Gly–His–Asp) are shown in the ball-and-stick representation. Asn32 is N-glycosylated in the structure. (B) Stereo view of N-glycosylated Asn32 with 2Fo−Fc electron density map contoured at 1.2σ (blue). The N-acetylglucosamine (GlcNAc) molecule covalently bonded to Asn32 is linked to two molecules of fucose (FUC) and one molecule of GlcNAc, which is a paucimannose (a major end product of glycosylation in insect cells) lacking three mannose molecules. (C) Superposition of the Cα traces of the SPL domain of PPAF-II and chymotrypsinogen (gray; PDB code 1CGI). The coloring scheme for PPAF-II and orientation of the proteins are the same as in (A). The signature sequences exhibit quite different Cα positions compared with chymotrypsinogen. Download figure Download PowerPoint Structural feature of the clip domain The clip domain is composed of a high portion of loops and a central, four-stranded, irregular β-sheet (Figure 3A left). The conserved three disulfide bonds knotting the loops and the β-strands together appear critical for the structural integrity of the central β-sheet that serves as the main framework of the clip domain structure. It has been suggested that clip domains may be structurally similar to antimicrobial proteins, β-defensins, based on the identical arrangement of the cysteine residues (Iwanaga et al, 1998). We show here that the clip domain structure is distinctively different from those of β-defensins. First, β-defensins contain three, instead of four, β-strands forming the central antiparallel β-sheet (Schneider et al, 2005) (Figure 3B right). Second, while β2 and β3 are antiparallel with each other in the structure of β-defensins, they are parallel in the structure of the clip domain (Figure 3B). Figure 3.Structure of the clip domain. (A) Central cleft in the clip domain. Left, The secondary structures are shown in blue, and the disulfide bonds are in yellow. The cleavage site is indicated by a blue arrow. The central β-sheet is composed of β0, β1, β2, and β3. Residues substituted in the mutagenesis study and cysteine residues are shown in the ball-and-stick representation. The outer rim of the cleft lined by β2-1-containing loop is in cyan, while the bottom of the cleft composed of a loop between Cys69 and Cys75 is in magenta. Right, Surface representation of the clip domain with the SPL domain as a Cα worm. (B) Comparison of the clip domain with human β-defensin 1. Left, a schematic drawing of the clip domain. The secondary structural elements are numbered in accordance with β-defensin 1. Right, a schematic drawing of β-defensin 1. Note the identical arrangement of the disulfide bonds, but a different topology compared with the clip domain. Download figure Download PowerPoint The clip domain has a noticeable protruding region and a prominent central cleft (Figure 3A). The protruding region consists of nonconserved amino acids between Cys58 (first Cys) and Cys69 (second Cys). The prominent cleft is found in the center of the clip domain (Figures 2C and 3A). The outer rim of the cleft is lined by the β2-1-containing loop (Asp80–Glu104), Cys112 (fifth Cys), and Cys113 (sixth Cys). The bottom of the cleft is composed of the loop flanked by Cys69 (second Cys) and Cys75 (third Cys), whose length in the number of amino acids is invariant in all the clip domains (Figures 1 and 3A). The loops forming the outer rim and the bottom are structurally fixed to the central β-sheet by the disulfide bonds between the four cysteine residues conserved among all the clip domains, suggesting that a similar cleft is present in all the other clip domains. The outer rim of the cleft predominantly consists of hydrophilic residues such as Asp80, Glu93, Glu104, Asp97, Thr91, and Arg99, while the interior surface of the cleft consists of hydrophobic residues including Tyr72, Val78, Phe96, and Val111 (Figure 3A). The shape and the electrostatic nature of the cleft appear to be well suited for binding a hydrophobic moiety. Structural feature of SPL domain The SPL domain of PPAF-II shares a relatively high sequence similarity with chymotrypsin, but it is devoid of a catalytic activity because of Gly353 in place of the invariant serine nucleophile in the catalytic SPs (Figures 1 and 2C). The domain contains two clefts, one of which corresponds to the defective active site cleft containing the Gly–His–Asp triad. It is elongated and shallow, similar to the active site cleft of chymotrypsin (Figure 4C). The cleft may serve as a docking site for binding a peptide segment of a protein. The other cleft is partly composed of a unique signature sequence (designated as Region IV; see below), and accommodates the N-terminal α-helix of a neighboring molecule in the crystal. This crystal packing interaction suggests that the cleft may also serve as a docking site for binding a protein. Figure 4.Structure of the SPL domain. (A) Calcium-binding site. A calcium ion bound to the SPL domain is in green, and the hepta-coordinations are indicated by dotted lines. Region II is represented as a Cα worm. (B) Representative interactions between the clip and SPL domains. The color scheme is the same as in Figure 2A. Hydrogen bonds and a salt bridge are shown in dashed lines. β2-1 of the clip domain and Region I of the SPL domain form a pair of the β-strands, which is stabilized by four hydrogen bonds (not shown). Phe96 and Tyr72 of the clip domain are involved in the hydrophobic interaction with Phe149 of the SPL domain. The conserved Tyr301 and Tyr108 form interdomain hydrogen bonds. (C) Clefts in the SPL domain. Surface representation of the SPL domain is shown with the clip domain as a Cα worm. Download figure Download PowerPoint The structure of PPAF-II reveals two other unique features of the SPL domain compared with the structure of chymotrypsin: the presence of a calcium-binding site and four distinct loops composed of signature sequences. A calcium ion is bound in the loop composed of residues 218–228, and hence referred to as the calcium-binding loop. A similar calcium-binding site was predicted to be present in Drosophila easter based on a three-dimensional modeling study (Rose et al, 2003). The calcium ion is hepta-coordinated with a pseudo-octahedral geometry involving the carboxylates of Glu218 and Glu226, the carbonyl oxygens of Asp220 and Thr223, plus two water molecules (Figure 4A). The residues involved in the calcium coordination are strictly conserved among the PPAF-II family, suggesting the presence of a calcium cage in all the members of this family. The four distinct segments, designated as Regions I–IV, were identified by the superposition of the PPAF-II and chymotrypsinogen structures, showing obvious deviations of these segments from the structure of chymotrypsinogen (Figure 2C). The amino-acid sequences of these segments are signature sequences, since they are highly conserved in the PPAF-II family but distinguished from the corresponding sequences of the easter-type proteases (Figure 1). Clip domain is tightly associated with SPL domain Notably, Regions I–III form the major interface for interaction with the clip domain (Figure 2A). Region I forms paired β-strands with β2-1 of the clip domain. Region II, which is the calcium-binding loop, interacts with the clip domain via Asp220 and Lys225 (Figure 4B). Especially, Asp220 forms a hydrogen bond with Tyr108 that is uniquely conserved in the PPAF-II family (Figure 4B). The calcium-binding loop is fairly rigid, as indicated by a low average temperature factor (24.1 Å2), and is likely to be built to interact with the clip domain through the calcium coordination. Tyr301 in Region III, which is strictly conserved in the PPAF-II family, forms a hydrogen bond with the backbone carbonyl atom of Glu106 in the clip domain. The interaction at the interface is also strengthened by hydrophobic interactions consisting of Phe149 in Region I, the aliphatic part of Lys225 in Region II, and Tyr72 and Phe96 in the clip domain (Figure 4B). Most residues at the interface are conserved in the PPAF-II family, but not in the easter-type SPs. Due to these interactions, the domain boundaries between the clip and SPL domains are barely distinguishable, strongly suggesting that the clip domain should be tightly associated with the SPL domain in solution. In support of this notion, we did not observe the cleavage at Region I by PPAF-III or a preferential cleavage by other proteases, including trypsin and lysyl endopeptidase (data not shown), although it corresponds to the activation loop of the easter-type SPs and contains the cleavage sequence 150Lys-Ile that is conserved in most of the SPs. Presumably, the interdomain interactions between Region I and β-2-1 (Figure 2A) prevent these proteases from accessing Region I. The clip domain also interacts with the SPL domain via the N-terminal segment. Compared with the sequences of the easter-type SPs, the PPAF-II family has a unique insertion of 33 amino acids at the N-terminus (Figure 1). Of these, 20 amino acids interact with the SPL domain at a site remote from the interface formed by Regions I–III (Figure 2A). The presence of this unique N-terminal segment should further restrict movement of the clip domain from the SPL domain. Oligomerization of cleaved, functional PPAF-II PPAF-II is specifically cleaved by PPAF-III at Arg99 within the clip domain and is subsequently converted into the functional form (Kim et al, 2002). The residue is right next to β2-1, which interacts with Region I of the SPL domain (Figure 3A left). Given the availability of recombinant PPAF-II and -III, we inspected whether the cleavage may accompany a quaternary structural change by using a size-exclusion chromatographic column. The analysis revealed that cleaved and functional PPAF-II is a large oligomer with ∼600 kDa molecular mass, whereas uncleaved PPAF-II is an ∼45 kDa monomer (Figure 5A top). Therefore, cleaved PPAF-II becomes functional by forming the quaternary structure. A cleavage-induced conformational change would be small and confined to the region of the paired β2-1/Region I, considering the tight interaction of the clip and SPL domains in uncleaved PPAF-II and no loss of the clip domain following the cleavage. The change into the quaternary structure is likely to be a common property of most PPAF-II family members, as judged based on the high degree of sequence similarity and an observation of a similar oligomerization of a PPAF-II family protein masquerade from T. molitor (S Piao and N-C Ha, unpublished data). In contrast, the striking difference between the two forms was not observed for PPAF-I and -III, whose pro- and active forms were eluted as a monomer from the size-exclusion column (data not shown). Figure 5.Biochemical and mutagenesis analyses of PPAF-II and proPO. (A) Oligomerization of PPAF-II. Top, size-exclusion chromatography of uncleaved PPAF-II and functional PPAF-II. The estimated molecular weight is shown above the peaks. Elution profile is shown and the identity of each peak was confirmed by SDS–PAGE. Bottom, electron microscopic picture of functional PPAF-II. Selected particles are displayed with a cartoon showing the arrangement of the protein molecules at the right side. Scale bars represent 20 nm. (B) Functional PPAF-II forms a complex with PO76s. Three different samples, analyzed using a Superdex S-200 size-exclusion column, are indicated in the rectangular boxes. The uncleaved or functional form of PPAF-II was preincubated with excess amount of PO76s for 30 min at 4°C in the presence of 5 mM CaCl2 before loading the sample on the column. PO76s were eluted as monomeric proteins. The bottom panels show the SDS–PAGE analysis of the fraction containing proteins with >1 MDa molecular weight (left) and an electron microscopic image (right; scale bar: 20 nm). The numbers (1–3) correspond to the subfractions indicated on the chromatogram. Similar results were obtained when 1 mM EDTA was added instead of 5 mM CaCl2 (data not shown). (C) PO activity of PO76s. Fractions of PO76s and the proteins in complex with functional PPAF-II were separated using Superdex S-200 as in (B), and the PO activity of each sample was measured. The amounts of PO76s in the two samples were adjusted to be equal based on the band intensities on an SDS–PAGE. The results represent the mean±s.d. of three separate experiments. Download figure Download PowerPoint Previously, a strong PO activity was observed only in the presence of functional PPAF-II, when proPO79s were incubated with active PPAF-I (Lee et al, 2002). To test whether PPAF-II binds PO76s, which are the major cleaved products generated from proPO79-I and -II by PPAF-I, we purified the PO76 proteins from H. diomphalia plasma and carried out size-exclusion chromatography after incubation of the sample with functional PPAF-II. Indeed, functional PPAF-II formed a tight complex with PO76s (Figure 5B). In a control experiment, PPAF-I or -III did not form a complex with PO76s (data not shown). The intensities of the protein bands on a denaturating gel indicated 1:1 stoichiometry of binding between functional PPAF-II and PO76s (Figure 5B bottom). Therefore, the homo-oligomerization of functional PPAF-II and its interaction with PO76s induce clustering of many molecules of PO76s in close proximity to each other. Consistently, an electron microscopic analysis revealed that 12 molecules of functional PPAF-II form two stacked hexameric rings (Figure 5A bottom), and the quaternary structure of the protein turns into a much bigger ball-like supramolecular assembly in the presence of PO76s (Figure 5B bottom). Clip-domain-dependent activation of PO76s We suspected that the clip domain may play an important role in the intera
Referência(s)