Protein C inhibitor directly and potently inhibits ... - Wiley Online Library

33 downloads 0 Views 935KB Size Report
Apr 12, 2007 - *Department of Molecular Pathobiology, Mie University Graduate School of Medicine, Tsu-city, Mie; and Chugai Pharmaceutical Co.,.
Journal of Thrombosis and Haemostasis, 5: 1477–1485

ORIGINAL ARTICLE

Protein C inhibitor directly and potently inhibits activated hepatocyte growth factor activator T. HAYASHI,*1 J. NISHIOKA,*1 N. NAKAGAWA,* H. KAMADA,* E. C. GABAZZA,* T. KOBAYASHI,  A . H A T T O R I   and K . S U Z U K I * *Department of Molecular Pathobiology, Mie University Graduate School of Medicine, Tsu-city, Mie; and  Chugai Pharmaceutical Co., Chemistry Research Department I, Gotenba, Shizuoka, Japan

To cite this article: Hayashi T, Nishioka J, Nakagawa N, Kamada H, Gabazza EC, Kobayashi T, Hattori A, Suzuki K. Protein C inhibitor directly and potently inhibits activated hepatocyte growth factor activator. J Thromb Haemost 2007; 5: 1477–85.

Summary. Background: Hepatocyte growth factor (HGF) plays an important role in tissue repair and regeneration. HGF activator (HGFA), a factor XIIa-like serine protease, activates HGF precursor to HGF. The precursor of HGFA, proHGFA, is activated by thrombin generated at sites of tissue injury. It is known that protein C inhibitor (PCI), an inhibitor of activated protein C (APC), also inhibits thrombin–thrombomodulin (TM) complex. Objectives: In the present study we evaluated the effect of PCI on thrombin-catalyzed proHGFA activation in the presence of TM, and on HGFA activity. Results: PCI did not inhibit thrombin-TM-mediated proHGFA activation, but it directly inhibited activated HGFA by forming an enzyme inhibitor complex. The second-order rate constants (M)1 min)1) of the reaction between HGFA and PCI in the presence or absence of heparin (10 U mL)1) were 4.3 · 106 and 4.0 · 106, respectively. The inhibition of HGFA by PCI resulted in a significant decrease of HGFA-catalyzed activation of HGF precursor. Exogenous HGFA added to normal human plasma formed a complex with plasma PCI, and this complex formation was competitively inhibited by APC in the presence of heparin, but very weakly in the absence of heparin. We also demonstrated using recombinant R362A-PCI that Arg362 residue of PCI is important for HGFA inhibition by PCI as judged from the three-dimensional structures constructed using docking models of PCI and HGFA or APC. Conclusion: These observations indicate that PCI is a potent inhibitor of activated HGFA, suggesting a novel function for PCI in the regulation of tissue repair and regeneration. Correspondence: Koji Suzuki, Department of Molecular Pathobiology, Mie University Graduate School of Medicine, Edobashi 2-174, Tsu-city, Mie 514-8507, Japan. Tel.: +81 59 231 5036; fax: +81 59 231 5209; e-mail: [email protected] 1 T. Hayashi and J. Nishioka equally contributed to the completion of this work.

Received 15 September 2006, accepted 12 April 2007  2007 International Society on Thrombosis and Haemostasis

Keywords: hepatocyte growth factor activator, hepatocyte growth factor, homology modeling, liver regeneration, protein C inhibitor. Introduction Blood clotting is an essential process for proper hemostasis and wound healing. Thrombin is a clotting factor that plays a central role in the promotion of blood coagulation and tissue regeneration in the wound-healing process [1]. Hepatocyte growth factor (HGF) plays a critical role in tissue regeneration by stimulating the proliferation and motility of various cell types, including epithelial and endothelial cells [2]. HGF is synthesized and secreted as an inactive single-chain precursor from liver sinusoidal endothelial cells and megakaryocytes (platelets) [3]. Limited proteolytic activation of this precursor is required for the biological activity of HGF [4]. The most potent activator of HGF precursor is the factor XIIa-like serine protease, HGF activator (HGFA) [5]. HGFA is synthesized in hepatocytes, and its inactive proHGFA circulates in blood as a 98-kDa single-chain precursor protein [6]. Thrombin activates proHGFA by limited proteolysis generating an activated 98kDa HGFA molecule that contains a disulfide-linked 65-kDa heavy chain and a 31-kDa light chain [7]. The HGFA heavy chain is then further cleaved in the systemic circulation by kallikrein releasing the 34-kDa mature form of HGFA [7]. Cell membrane-bound Kunitz-type HGFA inhibitors (HAI) have been recently isolated [8,9], and it has been found that one of them, HAI-1, acts as inhibitor and receptor of HGFA on cell surface [10]. The role of HAI as a physiological inhibitor of HGFA has not been as yet reported. To date, the existence of a potent physiological inhibitor of thrombin-catalyzed proHGFA activation or activated HGFA has not been as yet identified. Protein C inhibitor (PCI) (SERPINA5), a member of the plasma serine protease inhibitor (serpin) family, was isolated from human plasma as an inhibitor of activated protein C (APC), the main protease of the anticoagulant protein C pathway [11]. PCI forms an acyl-bonded complex with APC, and this complex formation is enhanced in the presence of

1478 T. Hayashi et al

heparin [12]. PCI is mainly synthesized in the liver [13], kidneys [13], megakaryocytes [14] and reproductive organs, including testis, epididymis, prostate and seminal vesicles [15]. Because of its broad tissue distribution, human PCI may regulate several physiological and pathological events in which serine proteases are involved, and its molecular targets may include APC of the protein C pathway, plasma kallikrein of the blood coagulation cascade, urinary and tissue plasminogen activators of the fibrinolysis system, acrosin of the fertilization system, and plasminogen activators released during matrix invasion by tumor cells [13]. In addition, it was reported that PCI can potently inhibit the thrombin–thrombomodulin (TM) complex, which is the main activator of protein C [16]. In the present study, we showed that PCI directly and potently inhibits activated HGFA by forming a complex with it independently of heparin; however, PCI did not affect thrombin–TM complex-mediated proHGFA activation. We also demonstrated using recombinant R362A-PCI that Arg residue at position 362 of the PCI molecule is important for HGFA inhibition in the absence of heparin as judged from threedimensional structures constructed using docking models of PCI and HGFA or APC. Materials and methods Materials

Diisopropyl fluorophosphate (DFP), sodium dodecylsulfate (SDS) and unfractionated heparin (160 U mg)1) were obtained from WAKO Pure Chem (Osaka, Japan). Bovine serum albumin (BSA) and 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate (CHAPS) were purchased from Sigma (St Louis, MO, USA), and Protein A-Sepharose was from Amersham Pharmacia Biotech (Uppsala, Sweden). Human plasma kallikrein was purchased from Biogenesis Inc. (Kingstone, NH, USA). The synthetic fluorogenic substrate of APC [Boc-Leu-Ser-Thr-Arg-methylcoumarylamide (MCA)], activated HGFA (Ac-Lys-Thr-Lys-Gln-Leu-Arg-MCA) and plasma kallikrein (Z-Phe-Arg-MCA) were purchased from Peptide Institute (Osaka, Japan). Proteins and antibodies

Recombinant 98-kDa human proHGFA [6] and HGF precursor [5] were prepared as described. The HGF precursor samples consist of a mixture of the major 98-kDa single-chain precursor and the minor two-chain activated species containing a 65-kDa heavy chain and a 31-kDa light chain. The two active forms of HGFA, the 98-kDa HGFA generated from proHGFA by thrombin cleavage and the 34-kDa HGFA generated after sequential processing by thrombin and plasma kallikrein, were prepared (hereinafter Ôactivated HGFAÕ is abbreviated as HGFA) [7]. Purification of PCI and protein C from human plasma, assay for protein C activation and APC isolation were carried out as described previously [12,17]. Thrombin (2500 U mg)1) was prepared from human pro-

thrombin [18]. Recombinant human TM [19] was a kind gift from Asahi Chem., Shizuoka, Japan. HGFA–PCI complexes were prepared by incubating the 34-kDa HGFA (196 nmol L)1) with PCI (200 nmol L)1) at 37 C for 120 min in 50 lL HEPES-buffered saline (50 mmol L)1 HEPES and 100 mmol L)1 NaCl, pH 8.0) containing 2 mmol L)1 CaCl2 and 0.05% CHAPS. Protein concentrations were determined at 280-nm absorbance using extinction coefficients (E1%280 nm) of 14.5 for both protein C and APC [17], 14.1 for PCI, and 10.0 for proHGFA, 98-kDa HGFA, and 34-kDa HGFA. Anti-proHGFA rabbit IgG was prepared from antiserum of rabbits immunized with recombinant human proHGFA using a Protein A-Sepharose column. The resultant anti-proHGFA IgG recognized the 98-kDa and 34-kDa forms of HGFA and proHGFA as demonstrated by Western blotting. The monoclonal anti-PCI murine IgG was prepared as described previously [20]. Measurement of proHGFA, PCI and HGFA–PCI complex

The concentration of proHGFA in plasma was determined by a sandwich-type enzyme immunoassay (EIA) according to a previously described method with some modifications [20]. A calibration curve was prepared using recombinant proHGFA. The concentration of PCI in plasma and in cultured media of baby hamster kidney (BHK) cells was also determined by EIA using peroxidase-coupled anti-PCI rabbit IgG as described previously [21]. The calibration curve was prepared using purified plasma PCI. The concentration of HGFA–PCI complex in plasma was determined by EIA using a monoclonal anti-PCI murine IgG specific for the complex [20] and peroxidase-coupled anti-proHGFA rabbit IgG, according to a modified method used to determine APC–PCI complex [22]. The calibration curve was prepared using purified HGFA–PCI complex. Sodium dodecylsulfate-polyacrylamide gel electrophoresis (SDS–PAGE)

SDS–PAGE was performed by the method of Laemmli [23] using 7.5–15% gradient gels. After electrophoresis, proteins were stained using a silver staining kit (Daiichi Pure Chem, Tokyo, Japan) according to the manufacturerÕs protocol. Activation of proHGFA by thrombin in the presence of TM and in the presence or absence of heparin

Purified proHGFA (141 nmol L)1) was activated by thrombin (0.55 nmol L)1) in the presence of various concentrations of TM and in the presence or absence of heparin (10 U mL)1). The activity of HGFA in the presence of thrombin was determined after treatment with hirudin (5 nmol L)1), a specific inhibitor of thrombin activity, using a peptide substrate, Ac-Lys-Thr-Lys-Gln-Leu-Arg-MCA. The quantity of 7amino-4-methyl-coumarin (AMC) released from the substrate was determined using a RF-510 fluorospectrophotometer  2007 International Society on Thrombosis and Haemostasis

HGFA inhibition by PCI 1479

(Shimadzu, Kyoto, Japan) with excitation at 380 nm and emission at 440 nm. One picomole of 98-kDa or 34-kDa HGFA released 2.18 pmole AMC min)1 in the reaction buffer (50 mmol L)1 Tris–HCl buffer, pH 8.0, containing 100 mmol L)1 NaCl and 2 mmol L)1 CaCl2). Each experiment was repeated at least three times. Effect of PCI on thrombin-catalyzed activation of proHGFA, on HGFA and plasma kallikrein activities

To determine the effect of PCI on thrombin-catalyzed activation of proHGFA, proHGFA (141 nmol L)1) was incubated with PCI (32 nmol L)1) at 37 C for 60 min, and thrombin (0.55 nmol L)1) was added to the reaction mixture in the presence of heparin (10 U mL)1). The HGFA activity was then assayed using the substrate described above. The direct effect of PCIs on HGFA activity was evaluated by incubating the 98-kDa or 34-kDa HGFA (13.6 nmol L)1) with PCI (32 nmol L)1) in the presence or absence of heparin at 37 C for 60 min. The inhibition constant (Ki) of the 98kDa or 34-kDa HGFA by PCI was calculated using the Dixon plot analysis [12]. The second-order rate constant K (M)1 min)1) for inhibition of HGFA by PCI in the presence or absence of heparin was determined as described [9]. The dose-dependent inhibition of HGFA or plasma kallikrein activity by various PCIs was evaluated by incubating the 98kDa or 34-kDa HGFA (13.6 nmol L)1) or plasma kallikrein (6.3 nmol L)1) with PCIs in the presence or absence of heparin as described above. Residual HGFA or plasma kallikrein activity was then assayed using Ac-Lys-Thr-LysGln-Leu-Arg-MCA or Z-Phe-Arg-MCA. Each experiment was repeated at least three times. Plasma samples

Plasma samples were obtained from patients with liver cirrhosis or hepatocellular carcinoma (HCC) with positive diagnosis for hepatitis C virus; the patients were hospitalized for medical or surgical treatment. All patients and healthy control subjects were Japanese with age ranging between 33 and 63 years old (mean 48.0 ± 11.3 years). Informed consent was obtained from all subjects. Homology modeling

The complex formation of APC-PCI and HGFA-PCI was constructed based on the structure similarity with the trypsin– serpin Michaelis complex (PDB accession number 1K9O [24]. All calculations, sequence alignment, homology modeling and structure visualization were carried out using the Molecular Operating Environment (MOE) [25]. The sequence of the serine protease domain of human proHGFA (Swiss-Prot accession number Q04756) and that of the PCI domain of the human PCI precursor (PIR accession number A39339) were aligned to that of trypsin and 1 K90 serpin complex, respectively. The sequence homologies of HGFA with trypsin  2007 International Society on Thrombosis and Haemostasis

and PCI with 1K9O serpin in the complex were 29% and 26%, respectively. The complex of APC-PCI was modeled by replacing trypsin of the 1K9O serpin complex by APC (PDB accession number 1AUT) [26] and the serpin by PCI. In the complex construction described above, there are considerable steric contacts between PCI and proteases. To relax this complex structure, calculation of energy minimization was carried out using AMBER94 force field [27] till the RSM gradient reaches values below 0.05. Before energy minimization, hydrogen atoms were added and force field atomic charges were assigned to proteins. The RCL (reactive center loop) of PCI and amino acids within 5 A˚ from RCL were soaked in 10 A˚ shell of water. During the energy minimization calculation, all atoms except water molecules and amino acid chains within 5 A˚ from the RCL of PCI were fixed. For clarity the water molecules were not displayed from the figures. Preparation of recombinant PCI

Two types of recombinant PCI were prepared using the mammalian expression system as follows. Wild-type PCI cDNA fragment [21] amplified using two synthetic oligonucleotide primers (5¢-CGCGAATTCCTCTGGCAGAGCCTCCGTTTCC-3¢ and 5¢-CGCGAATTCTCAGGGGCGGTTCACTTTGCCAA-3¢; underlined nucleotides indicate Eco RI site) was subcloned into Eco RI site of pBluescript SKII(+) and transformed into E. coli XL10-Gold according to the manufacturerÕs instructions. R362A-PCI cDNA was prepared by QuikChange XL site-directed mutagenesis kit (Stratagene, Cedar Creek, TX, USA) using two synthetic oligonucleotide primers (5¢-CTGAACTCTCAGGCGCTAGTGTTCAAC-3¢ and 5¢-GTTGAACACTAGCGCCTGAGAGTTCAG-3¢; underlined nucleotides indicate mutation sites) and wild-type PCI cDNA as a template. Thereafter, the wild-type PCI cDNA and the mutant R362A-PCI cDNA were respectively inserted into the Cla I/Xba I site of the mammalian expression vector, pRC/CMV (Invitrogen, Carlsbad, CA, USA), and then transfected using Effectene transfection reagent (QIAGEN, Tokyo, Japan) into BHK cells with pSV2/dhfr; these were cultured in high glucose DulbeccoÕs modified EagleÕs medium (DMEM) (Invitrogen) containing sodium pyruvate and glutamate (Invitrogen), supplemented with 10% fetal bovine serum (Equitech-Bio, Kerrville, TX, USA), 0.01% penicillin, and 0.01% streptomycin (Sanko Junyaku, Tokyo, Japan). After transfection, BHK cell lines expressing high amount of wild-type PCI or R362A-PCI were selected and medium containing recombinant wild-type PCI and R362A-PCI was prepared as described previously [28]. The medium was separated from cell debris by centrifugation and frozen at )20C until purification. Purification of recombinant PCI

Recombinant wild-type PCI and R362A-PCI were purified from culture medium as described previously [29].

1480 T. Hayashi et al

Residual HGFA activity (%)

Concentration of each recombinant protein was determined using EIA specific for PCI as described previously [21]. Results Effect of PCI on thrombin-catalyzed activation of proHGFA in the presence or absence of TM

Thrombin-catalyzed activation of proHGFA was enhanced in the presence of dextran sulfate, but TM exerted no effect (data not shown). Next, we examined the effect of PCI on thrombincatalyzed activation of proHGFA in the presence or absence of heparin. As shown in Fig. 1A, proHGFA activation by thrombin was significantly inhibited by PCI. To clarify whether PCI inhibits thrombin-catalyzed activation of proHGFA or directly thrombin-activated HGFA, this latter was incubated with PCI in the presence or absence of heparin. Fig. 1B shows that PCI directly inhibits HGFA and that its inhibitory activity is independent of heparin. Effect of PCI on HGFA amidolytic activity

As shown in Fig. 2, PCI inhibited the 34-kDa HGFA in a dose-dependent manner and independently of heparin. The amidolytic activity of the 98-kDa HGFA was not different from that of 34-kDa HGFA. PCI inhibited the activity of these two HGFA isoforms almost equally, showing apparent inhibition constants (Ki) of 6.1 · 10)9 M and 6.3 · 10)9 M for 34-kDa HGFA and 98-kDa HGFA, respectively. The Ki values calculated by the Dixon plot analysis in the presence or absence of heparin (data not shown) were 3.9 · 10)8 M and 4.0 · 10)8 M, respectively. The second-order rate constants (M )1 min)1) of the reaction between 34-kDa HGFA and PCI in the presence or absence of heparin (10 U mL)1) were 4.3 · 106 and 4.0 · 106, respectively.

75

50

25

0

0

50

100

150

Concentration of PCI (nmol L–1) Fig. 2. Protein C inhibitor (PCI)-mediated dose-dependent inhibition of hepatocyte growth factor activator (HGFA) in the presence or absence of heparin. Activated HGFA (13.6 nmol L)1) was incubated with varying concentrations of PCI in the presence (d) or absence (s) of heparin (10 U mL)1). The amidolytic activity of HGFA was determined using a synthetic substrate specific for HGFA.

Complex formation between HGFA and PCI

For quantitative determination of HGFA inhibition by PCI, EIA specific for the HGFA–PCI complex and SDS–PAGE were performed. As shown in Fig. 3, residual amidolytic activity of HGFA decreased concomitantly with increased HGFA–PCI complex formation, an inverse relationship being observed between HGFA inhibition by PCI and HGFA–PCI complex formation. The band density of a 91-kDa protein, which corresponds to the HGFA–PCI complex, increased in a time-dependent manner and concomitantly with the decrease in 150

100

75 50 25

0 ProHGFA Thrombin Heparin PCI

+ + – –

+ + + –

+ + + +

+ + – +

+ – + +

75 50 25

0 + HGFA – PCI Heparin –

+ + –

+ + +

Fig. 1. (A) Effect of thrombin and protein C inhibitor (PCI) on hepatocyte growth factor precursor (proHGFA) activation in the presence of heparin. ProHGFA (141 nmol L)1) was incubated with PCI (32 nmol L)1) at 37 C for 60 min, and thrombin (0.55 nmol L)1) was added to the reaction mixture in the presence of heparin (10 U mL)1). (B) Effect of PCI and heparin on hepatocyte growth factor activator (HGFA) activity. Activated HGFA (13.6 nmol L)1) was incubated with PCI (32 nmol L)1) in the presence or absence of heparin (10 U mL)1). The amidolytic activity of HGFA was determined using a synthetic peptide substrate specific for HGFA.

100

100

75

50 50 25

0

HGFA-PCI complex (nM)

100

Residual HGFA activity (%)

B Residual HGFA Activity (%)

Activated HGFA (%)

A

100

0 0

50 150 100 Concentration of PCI (nmol L–1)

Fig. 3. Relationship between protein C inhibitor (PCI)-mediated hepatocyte growth factor activator (HGFA) inhibition and HGFA–PCI complex formation. HGFA (13.6 nmol L)1) was incubated with PCI (32 nmol L)1) in the absence of heparin. Residual HGFA amidolytic activity (s) was determined using a synthetic substrate. The amount of HGFA–PCI complex (d) was determined by specific EIA.  2007 International Society on Thrombosis and Haemostasis

HGFA inhibition by PCI 1481 kDa 94

HGFA-PCI complex

67

PCI 43 HGFA

30

PCI HGFA 0

10

20

30

45

60

HGFA-PCI complex (nmol L–1)

60

40

20

0 0

Time (min)

40 20 Incubation time (min)

60

Fig. 4. Complex formation between hepatocyte growth factor activator (HGFA) and protein C inhibitor (PCI). The 34-kDa HGFA (800 nmol L)1) was incubated with PCI (980 nmol L)1) at 37 C for the indicated time intervals. The incubation mixture was treated with 2mercaptoethanol and subjected to sodium dodecylsulfate–polyacrylamide gel electrophoresis followed by silver staining.

Fig. 6. Complex formation of hepatocyte growth factor activator (HGFA) with plasma protein C inhibitor (PCI). The 34-kDa HGFA (50 nmol L)1) (d) or DFP-treated HGFA (50 nmol L)1) (s) was added to 1 mL of human plasma and incubated at 37 C for several hours. At the designated intervals, aliquots of the incubation mixture were taken. The concentration of HGFA–PCI complex was determined by specific EIA.

the bands of the 34-kDa HGFA and 57-kDa intact PCI (Fig. 4).

HGFA–PCI complex formation in plasma in vitro

Effect of PCI on HGFA-catalyzed activation of HGF precursor

In the absence of PCI, almost all of the single-chain HGF precursors were catalyzed by HGFA releasing disulfide-linked active HGF that contained a 65-kDa heavy chain and a 31kDa light chain; however, in the presence of PCI-pretreated HGFA, the single-chain HGF precursor was minimally converted into the two-chain form (Fig. 5). These data indicate that PCI inhibits HGFA-catalyzed activation of HGF precursors.

kDa 94

HGF precursor

67 HGF heavy chain PCI

43

HGF light chain

30

The 34-kDa HGFA was added to plasma in vitro, and the concentration of HGFA–PCI complex was measured by complex-specific EIA. HGFA formed complex with PCI in a time-dependent manner, but DFP-treated HGFA did not (Fig. 6). Effect of APC on HGFA–PCI complex formation

PCI inhibits plasma APC with an apparent Ki of 5.6 · 10)8 M [11] and the 34-kDa HGFA with a Ki of 6.1 · 10)9 M (as described in the present study). To evaluate whether APC and HGFA are competitive substrates of PCI inhibitory activity in plasma, the effect of APC on HGFA–PCI complex formation was examined in the presence or absence of heparin. APC competitively inhibited HGFA–PCI complex formation in the presence of heparin, but exhibited only a weak inhibitory effect in the absence of heparin; in this latter condition PCI effectively inhibits HGFA (Fig. 7). Plasma kallikrein also competitively inhibited HGFA–PCI complex formation in the presence and absence of heparin (data not shown). HGFA–PCI complex formation in human plasma in vivo

Time (min) 0

5 30 + HGFA – PCI

5 30 + HGFA + PCI

Fig. 5. Effect of protein C inhibitor (PCI) on hepatocyte growth factor activator (HGFA)-catalyzed activation of HGF precursor. The HGF precursor, a mixture of the major 98-kDa single-chain precursor and the minor two-chain activated form (61-65-kDa heavy chain and 31-34-kDa light chain), (500 nmol L)1) was incubated for 5 or 30 min with HGFA (10 nmol L)1) alone or HGFA pretreated with PCI (48 nmol L)1). The incubation mixture was treated with 2-mercaptoethanol and subjected to sodium dodecylsulfate–polyacrylamide gel electrophoresis followed by silver staining.  2007 International Society on Thrombosis and Haemostasis

The concentrations of the HGFA–PCI complex, pro-HGFA and PCI in peripheral plasma obtained from normal subjects (n = 15) and from patients with hepatitis (n = 9) or HCC (n = 7) were determined. Figure 8 shows that the plasma concentrations of proHGFA and PCI are significantly (P < 0.05) decreased in patients with HCC compared with normal subjects, and that the plasma concentration of HGFA– PCI complex is significantly (P < 0.01) increased (60 ± 20 pM) compared with normal subjects (27 ± 10 pM). On the other hand, the plasma concentrations of proHGFA

1482 T. Hayashi et al HGFA

APC

HGFA–PCI complex (%)

100

75 37-loop (K37–K38–K39)

50

25 K270 R269 on H-helix

K270 R269 on H-helix

0 0

1

10 2 5 APC/HGFA ratio (mol mol–1)

PCI

50

0

200 150 100 50

HGFA–PCI complex (pmol L–1)

112 ± 50 pM

100

60 ± 20 pM

150

27 ± 10 pM

ProHGFA or PCI (nmol L–1)

Fig. 7. Effect of activated protein C (APC) on hepatocyte growth factor activator– protein C inhibitor (HGFA–PCI) complex formation in the presence or absence of heparin. Varying concentrations of APC were added to a mixture of HGFA (194 nmol L)1) and PCI (179 nmol L)1) in the presence (d) or absence (s) of heparin (10 U mL)1) and incubated at 37 C for 30 min. HGFA–PCI complex was detected by specific EIA. Abscissa indicates molar ratio of APC and HGFA in the incubation mixture.

0 ProHGFA

PCI

HGFA–PCI complex

PCI

Fig. 9. Spatial representation of molecular homology between activated protein C–protein C inhibitor (APC–PCI) (left) and hepatocyte growth factor activator (HGFA)–PCI (right) complexes. APC (green), HGFA (red orange) and PCI (gray) are shown as solid surfaces. In each complex, acidic and basic amino acids are in red and blue colors, respectively. There is no basic residue in the loop structure of HGFA that corresponds with the 37-loop structure of APC. See text for details.

APC

HGFA

APC

HGFA

PCI

PCI

A

B

Fig. 8. Plasma hepatocyte growth factor activator precursor (proHGFA), protein C inhibitor (PCI) and HGFA–PCI complex in patients with hepatocellular carcinoma (HCC) and hepatitis. Concentrations of proHGFA, PCI and HGFA–PCI complex in the plasma of normal subjects (s) and in the plasma of patients with HCC (d) and hepatitis ()). ProHGFA, PCI and HGFA–PCI complex levels were determined by specific enzyme immunoassay.

and PCI were not significantly different between hepatitis patients and normal subjects, but the plasma level of HGFA– PCI complex level was significantly (P < 0.01) increased in patients with hepatitis (112 ± 50 pM) compared with healthy individuals. Three-dimensional model

Figure 9 shows three-dimensional structures with docking models of PCI and APC or HGFA. In these models, the estimated heparin binding sites are the Arg269-Lys270 residues of the PCI H-helix [30] and the 37-loop structure of APC

Fig. 10. Comparative molecular modeling of the RCL region of protein C inhibitor (PCI) with the 37-loop of activated protein C (APC) (left) and hepatocyte growth factor activator (HGFA) (right). APC (green), HGFA (red orange) and PCI (gray) are shown as ribbon (A) and mixed model (B). The reactive center loop structure of PCI is gold-colored. The space occupied by K37, K38 and K39 constructing the 37-loop structure of APC is replaced by I35, G36 and D37 in the homology modeling of HGFA. Hydrogen atoms are not displayed for clarity. The distance from the NH2 residue R362 of PCI to the NH2 residues K37 and K39 of APC of the APC–PCI complex is estimated to be 6.8A˚ and 5.6A˚, respectively. The distance from the NH2 residue R362 of PCI to the carboxyl residue D37 of HGFA of the HGFA–PCI complex is estimated to be 7.6A˚. See text for details.  2007 International Society on Thrombosis and Haemostasis

HGFA inhibition by PCI 1483

B Residual HGFA activity (%)

Residual HGFA activity (%)

A 100 75 50 25 0

100 75 50 25 0

0 50 100 150 Concentration of PCI (nmol L–1)

D Residual kallikrein activity (%)

Residual kallikrein activity (%)

C

0 50 100 150 Concentration of PCI (nmol L–1)

100 75 50 25 0 0 5 10 15 20 25 Concentration of PCI (nmol L–1)

100 75 50 25 0 0 5 10 15 20 25 Concentration of PCI (nmol L–1)

Fig. 11. Inhibition of hepatocyte growth factor activator (HGFA) and plasma kallikrein by recombinant wild-type protein C inhibitor (PCI) and R362APCI in the presence or absence of heparin. HGFA (13.6 nmol L)1) (A, B) or plasma kallikrein (6.5 nmol L)1) (C, D) was incubated with varying concentrations of recombinant wild-type PCI (A, C) or R362A-PCI (B, D) in the presence (d, ) or absence (s, h) of heparin (10 U mL)1). Residual amidolytic activity of HGFA or plasma kallikrein was determined using a synthetic substrate specific for HGFA or plasma kallikrein.

containing Lys37-Lys38-Lys39 [31]. Fig. 10A,B shows the detailed structures of the docking models of PCI and APC or HGFA. In this model, the Arg362 residue, which is the first residue of the s1C strand after RCL of PCI, is estimated to be the nearest residue to the 37-loop structure of APC in the APCPCI complex, and the distance from the NH2 residue Arg362 of PCI to the NH2 residues Lys37 and Lys39 is estimated to be 6.8A˚ and 5.6A˚, respectively. These values indicate the existence of sufficient distance for positive charge repellence between PCI and APC in the absence of heparin. This repellence can be neutralized by heparin when Lys37 and Lys39 residues from APC interact with heparin. On the other hand, the negatively charged amino acid Asp37 of HGFA is located very near (7.6A˚) to the Arg362 of PCI (Fig. 10A, B). The carboxyl residue Asp37 of HGFA is able to interact with the NH2 residue Arg362 of PCI; we speculate that heparin does not affect the inhibition of HGFA by PCI because of the strong interaction between Asp37 of HGFA and Arg362 of PCI. Inhibition of HGFA by wild-type PCI or R362A-PCI in the presence or absence of heparin

As shown in Fig. 11A,B the inhibitory activity of R362APCI on HGFA was markedly decreased compared with  2007 International Society on Thrombosis and Haemostasis

recombinant wild-type PCI, but it was remarkably accelerated in the presence of heparin. We also evaluated the inhibitory activity of R362A-PCI on plasma kallikrein, which is inhibited by PCI independently of heparin; we found that the inhibitory activity of R362A-PCI is weakly suppressed in the presence of heparin compared with that of wild-type PCI (Fig. 11C, D). In addition, we found that R362A-PCI can inhibit APC, and that this inhibition is also accelerated by heparin (data not shown). These data indicate that Arg362 residue of PCI is specifically important for HGFA inhibition. Discussion In the present study, we demonstrated in vitro that PCI is a potent inhibitor of activated HGFA in human plasma by forming a stable acyl-bonded complex with the enzyme, unlike the unstable complex formed between PCI and APC [12]. This HGFA inhibition by PCI decreases the HGFA-mediated proteolytic activation of the single-chain HGF precursor. We have previously demonstrated that platelet-derived phospholipid microvesicles containing phosphatidylethanolamine significantly enhance PCI-mediated inhibition of APC [14]. In this study, however, we found that PCI-mediated inhibition of HGFA is not affected by phospholipids either in solution or in

1484 T. Hayashi et al

solid phase (data not shown). Similar to PCI-mediated inhibition of plasma kallikrein, PCI requires no heparin for HGFA inhibition, thereby suggesting that PCI is able to inhibit HGFA even in the solution phase of plasma, which is devoid of heparin-like glycosaminoglycans commonly found on endothelial surface and surrounding endothelial matrix. Like PCImediated inhibition of plasma kallikrein, the conversion of the 98-kDa HGFA into its 34-kDa form may also occur in the solution phase. Interestingly, we found that PCI bound to heparin is released when PCI forms a complex with HGFA (data not shown); this finding may be due to lower affinity of the HGFA–PCI complex for solid-phase heparin than for HGFA or PCI alone. The release of HGFA–PCI complex presumably occurs following a conformational change in both HGFA and PCI upon binding to glycosaminoglycans. In addition, we found that HGFA inhibition by PCI is significantly affected by APC in the presence of heparin, suggesting that PCI bound to heparin-like glycosaminoglycans has preference for APC inhibition; by contrast, PCI free in the solution phase has preference for HGFA inhibition. To elucidate the mechanism of HGFA inhibition by PCI, we prepared a three-dimensional model of APC, HGFA and PCI, and evaluated whether heparin-independent inhibition of HGFA by PCI is caused by the interaction between Arg362 residue of PCI and Asp37 residue of HGFA. Using recombinant R362A-PCI we found that Arg362 residue of PCI is important for heparin-independent inhibition of HGFA, although heparin weakly enhanced the interaction between HGFA and R362A-PCI. This finding suggests that heparin induces change in the molecular conformation of PCI, making it suitable for HGFA inhibition. R362A-PCI also effectively and heparin-independently inhibits plasma kallikrein, which lacks amino acid residues such as Asp37 of HGFA and 37-loop of APC. These observations suggest that R362A-PCI does not suffer substantial conformational change that may perturb its inhibitory activity on plasma kallikrein, and that Asp37 of HGFA is important for the inhibitory activity of PCI in the absence of heparin. On the other hand, APC contains the 37-loop but lacks amino acid residues such as Asp37, suggesting that heparin can accelerate the interaction between APC and PCI by bridging these two molecules or by inducing conformational changes in PCI. This idea is consistent with the heparin-mediated acceleration of APC inhibition by R362A-PCI. APC has anticoagulant and anti-inflammatory effects, and thus inhibition of APC by PCI may regulate procoagulant and inflammatory responses at sites of tissue injury. In addition, while the thrombin-catalyzed activation of proHGFA mainly occurs in the presence of heparin or negatively charged surfaces [7], HGFA inhibition by PCI may occur in the absence of heparin-like glycosaminoglycans. On the other hand, our present in vitro study showed that formation of HGFA–PCI complex accelerates in culture medium of thrombin-treated liver cells (data not shown), suggesting that PCI-mediated regulation of HGFA may be important during liver regeneration. Further, HGFA–PCI complex was detected in normal

human plasma, and its plasma concentration was significantly increased in patients with hepatitis or HCC, suggesting that PCI may also inhibit HGFA in vivo. In conclusion, the results of this study suggest that PCI plays a crucial role in the control of tissue regeneration and repair during normal wound healing or under pathological conditions by inhibiting the activity of HGFA at sites of tissue injury. Acknowledgements We would like to thank Y. Okagami and S. Hakata for their excellent technical assistance. This study was supported in part by Grant-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology of Japan, from the Japan Society for the Promotion of Science (JPSP) (Nos. 11780443, 12215065, 12217063, 13770877, 14370055, 16591592, 17390276, 18591749), and grants from the Japan Health Science Foundation and the Mie University COE Project Fund. Disclosure of Conflict of Interests The authors state that they have no conflict of interest. References 1 Colvin RB. Wound healing processes in hemostasis and thrombosis. In: Gimbrone MA, eds. Vascular Endothelium in Hemostasis and Thrombosis. Edinburgh: Churchill Livingstone, 1986: 220–41. 2 Kan M, Zhang G, Zarnegar R, Michalopoulos G, Myoken Y, McKeehan WL, Stevens JI. Hepatocyte growth factor/hepatopoietin A stimulates the growth of rat kidney proximal tubule epithelial cells (RPTE), rat nonparenchymal liver cells, human melanoma cells, mouse keratinocytes and stimulates anchorage-independent growth of SV-40 transformed RPTE. Biochem Biophys Res Commun 1991; 17: 331–7. 3 Nakamura Y, Morishita R, Higaki J, Kida I, Aoki M, Moriguchi A, Yamada K, Hayashi S, Yo Y, Matsumoto K, Nakamura T, Ogihara T. Expression of local hepatocyte growth factor system in vascular tissues. Biochem Biophys Res Commun 1995; 215: 483–8. 4 Naka D, Ishii T, Yoshiyama Y, Miyazawa K, Hara H, Hishida T, Kitamura N. Activation of hepatocyte growth factor by proteolytic conversion of a single chain form to a heterodimer. J Biol Chem 1992; 267: 20114–9. 5 Shimomura T, Miyazawa K, Komiyama Y, Hiraoka H, Naka D, Morimoto Y, Kitamura M. Activation of hepatocyte growth factor by two homologous proteases, blood-coagulation factor XIIa and hepatocyte growth factor activator. Eur J Biochem 1995; 229: 257– 61. 6 Miyazawa K, Shimomura T, Kitamura A, Kondo J, Morimoto Y, Kitamura N. Molecular cloning and sequence analysis of the cDNA for a human serine protease responsible for activation of hepatocyte growth factor. Structural similarity of the protease precursor to blood coagulation factor XII. J Biol Chem 1993; 268: 10024–8. 7 Shimomura T, Kondo J, Ochiai M, Naka D, Miyazawa K, Morimoto Y, Kitamura N. Activation of the zymogen of hepatocyte growth factor activator by thrombin. J Biol Chem 1993; 268: 22927–32. 8 Shimomura T, Denda K, Kitamura A, Kawaguchi T, Kito M, Kondo J, Kagaya T, Qin L, Takata H, Miyazawa K, Kitamura N. Hepatocyte growth factor activator inhibitor, a novel Kunitz-type serine protease inhibitor. J Biol Chem 1997; 272: 6370–6.  2007 International Society on Thrombosis and Haemostasis

HGFA inhibition by PCI 1485 9 Kawaguchi T, Qin L, Shimomura T, Kondo J, Matsumoto K, Denda K, Kitamura N. Purification and cloning of hepatocyte growth factor activator inhibitor type 2, a Kunitz-type serine protease inhibitor. J Biol Chem 1997; 272: 27558–64. 10 Kataoka H, Shimomura T, Kawaguchi T, Hamasuna R, Itoh H, Kitamura N, Miyazawa K, Koono M. Hepatocyte growth factor activator inhibitor type 1 is a specific cell surface binding protein of hepatocyte growth factor activator (HGFA) and regulates HGFA activity in the pericellular microenvironment. J Biol Chem 2000; 275: 40453–62. 11 Suzuki K, Nishioka J, Hashimoto S. Protein C inhibitor. Purification from human plasma and characterization. J Biol Chem 1983; 258: 163– 8. 12 Suzuki K, Nishioka J, Kusumoto H, Hashimoto S. Mechanism of inhibition of activated protein C by protein C inhibitor. J Biochem (Tokyo) 1984; 95: 187–95. 13 Suzuki K. Protein C inhibitor (PAI-3): structure and multi-function. Fibrinol Proteol 2000; 14: 133–45. 14 Nishioka J, Ning M, Hayashi T, Suzuki K. Protein C inhibitor secreted from activated platelets efficiently inhibits activated protein C on phosphatidylethanolamine of platelet membrane and microvesicles. J Biol Chem 1998; 273: 11281–7. 15 Laurell M, Christensson A, Abrahamsson PA, Stenflo J, Lilja H. Protein C inhibitor in human body fluids. Seminal plasma is rich in inhibitor antigen deriving from cells throughout the male reproductive system. J Clin Invest 1992; 89: 1094–101. 16 Rezaie AR, Cooper ST, Church FC, Esmon CT. Protein C inhibitor is a potent inhibitor of the thrombin–thrombomodulin complex. J Biol Chem 1995; 270: 25336–9. 17 Suzuki K, Stenflo J, Dahlback B, Teodorsson B. In activation of human coagulation factor V by activated protein C. J Biol Chem 1983; 258: 1914–20. 18 Miletich JP, Broze GJ Jr, Majerus PW. The synthesis of sulfated dextran beads for isolation of human plasma coagulation factors II, IX, and X. Anal Biochem 1980; 105: 304–10. 19 Suzuki K, Hayashi T, Nishioka J, Zushi M, Honda K, Yamamoto S, Hashimoto S. A domain composed of epidermal growth factor-like structures of human thrombomodulin is essential for thrombin binding and for protein C activation. J Biol Chem 1989; 264: 4872–6. 20 Suzuki K, Kusumoto H, Nishioka J, Komiyama Y. Bovine plasma protein C inhibitor with structural and functional homologous

 2007 International Society on Thrombosis and Haemostasis

21

22

23 24

25 26

27

28

29

30

31

properties to human plasma protein C inhibitor. J Biochem (Tokyo) 1990; 107: 381–8. Wakita T, Hayashi T, Tamaru H, Nishioka J, Akita N, Asanuma K, Kamada H, Gabazza EC, Ido M, Kawamura J, Suzuki K. Regulation of carcinoma cell invasion by protein C inhibitor whose expression is decreased in renal cell carcinoma. Int J Cancer 2004; 108: 516–23. Nakase T, Wada H, Minamikawa K, Wakita Y, Shimura M, Hiyoyama K, Tamaki S, Shirakawa S, Deguchi K, Nishioka J, Hayashi T, Suzuki K. Increased activated protein C-protein C inhibitor complex level in patients positive for lupus anticoagulant. Blood Coagul Fibrinol 1994; 5: 173–7. Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970; 227: 680–5. Ye S, Cech AL, Belmares R, Bergstrom RC, Tong Y, Corey DR, Kanost MR, Goldsmith EJ. The structure of a Michaelis serpin–protease complex. Nature Struct Biol 2001; 8: 979–83. Molecular Operating Environment, version 2004.03; Montreal, Canada: Chemical Computing Group Inc., 2004. Mather T, Oganessyan V, Hof P, Huber R, Foundling S, Esmon CT, Bode W. The 2.8 A crystal structure of Gla-domainless activated protein C. EMBO J 1996; 15: 6822–31. Weiner SJ, Kollman PA, Case DA, Singh UC, Ghio C, Alagona G, Profeta S, Weiner PA. A new force field for molecular mechanical simulation of nucleic acids and proteins. J Am Chem Soc 1984; 106: 765–84. Hayashi T, Nishioka J, Shigekiyo T, Saito S, Suzuki K. Protein S Tokushima: abnormal molecule with substitution of Glu for Lys-155 in the second epidermal growth factor-like domain of protein S. Blood 1994; 83: 683–90. Asanuma K, Yoshikawa T, Hayashi T, Akita N, Nakagawa N, Hamada Y, Nishioka J, Kamada H, Gabazza EC, Ido M, Uchida A, Suzuki K. Protein C inhibitor inhibits breast cancer cell growth, metastasis and angiogenesis independently of its protease inhibitory activity. Int J Cancer 2007, in press. Shirk RA, Elisen MG, Meijers JC, Church FC. Role of the H helix in heparin binding to protein C inhibitor. J Biol Chem 1994; 269: 28690– 5. Glasscock LN, Gerlitz B, Cooper ST, Grinnell BW, Church FC. Basic residues in the 37-loop of activated protein C modulate inhibition by protein C inhibitor but not by alpha(1)-antitrypsin. Biochim Biophys Acta 2003; 1649: 106–17.