Hepatocyte Growth Factor Activator Inhibitor, a Novel Kunitz-type Serine Protease Inhibitor*

(Received for publication, August 6, 1996, and in revised form, November 7, 1996)

Takeshi Shimomura Dagger , Kimitoshi Denda §, Akiko Kitamura , Toshiya Kawaguchi Dagger , Masahiro Kito Dagger , Jun Kondo Dagger , Shinji Kagaya §, Li Qin §, Hiroyuki Takata §, Keiji Miyazawa and Naomi Kitamura §par

From the Dagger  Research Center, Mitsubishi Chemical Corp., Aoba-ku, Yokohama 227, the § Department of Life Science, Faculty of Bioscience and Biotechnology, Tokyo Institute of Technology, Nagatsuta, Midori-ku, Yokohama 226, and the  Institute for Liver Research, Kansai Medical University, Moriguchi, Osaka 570, Japan

ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
REFERENCES


ABSTRACT

Hepatocyte growth factor (HGF) activator is a serine protease that is produced and secreted by the liver and circulates in the blood as an inactive zymogen. In response to tissue injury, the HGF activator zymogen is converted to the active form by limited proteolysis. The activated HGF activator converts an inactive single chain precursor of HGF to a biologically active heterodimer in injured tissue. The activated HGF may be involved in the regeneration of the injured tissue. In this study, we purified an inhibitor of HGF activator from the conditioned medium of a human MKN45 stomach carcinoma cell line and molecularly cloned its cDNA. The sequence of the cDNA revealed that the inhibitor has two well defined Kunitz domains, suggesting that the inhibitor is a member of the Kunitz family of serine protease inhibitors. The sequence also showed that the primary translation product of the inhibitor has a hydrophobic sequence at the COOH-terminal region. Inhibitory activity toward HGF activator was detected in the membrane fraction as well as in the conditioned medium of MKN45 cells. These results suggest that the inhibitor may be produced as a membrane-associated form and secreted by the producing cells as a proteolytically truncated form.


INTRODUCTION

Hepatocyte growth factor (HGF),1 also known as scatter factor, is a mesenchymally derived heparin-binding glycoprotein that is secreted as an inactive single chain precursor from producing cells, and normally it remains in this form (1), probably associated with the extracellular matrix in the producing tissues (2). In response to tissue damage such as hepatic and renal injury, the inactive single chain form is converted to an active heterodimeric molecule exclusively in the injured tissue by limited proteolysis at a single site (1). The proteolytically activated HGF may be involved in regeneration of the injured tissue, because HGF is a potent mitogen for a variety of cells such as hepatocytes and renal tubular epithelial cells (3-5). Thus, the biological effects of HGF in the injured tissues are regulated through proteolytic processing. This processing is mediated by an enzymic activity that is induced in the injured tissue (1). Four proteases are reported to activate HGF in vitro. These are HGF activator, urokinase, tissue-type plasminogen activator, and blood coagulation factor XIIa (2, 6-9). Recently, we identified HGF activator as a processing enzyme in the injured liver (10).

HGF activator, a novel serine protease, was isolated from fetal bovine serum (6) and from human serum (7). The sequence of the HGF activator cDNA revealed that HGF activator purified from serum is derived from the COOH-terminal half-region of a precursor protein and that the precursor has a similar overall domain organization to blood coagulation factor XII (7). The precursor protein circulates in the blood as an inactive zymogen. The zymogen is activated by limited proteolysis by thrombin, and the NH2-terminal half region of the zymogen is removed by plasma kallikrein in vitro (11). Tissue injury often leads to local blood coagulation. When coagulation is initiated, thrombin and kallikrein are generated from pro-proteins. Activated thrombin and kallikrein may participate in the activation of the HGF activator zymogen and in the removal of the NH2-terminal half of the zymogen, respectively.

The activation potential of HGF activator may be neutralized by the activity of an inhibitor(s). The activity of HGF activator is not inhibited by serum protease inhibitors such as antithrombin III, C1-inhibitor, alpha 2-antiplasmin, and alpha 1-proteinase inhibitor. Furthermore, HGF activator is active in serum (9). Thus, serum protease inhibitors are not responsible for inhibiting the activity of HGF activator. It is likely that cells of tissues produce the inhibitor of HGF activator. We therefore searched human cell lines for the inhibitor and found that conditioned media from various cell lines contained the inhibitory activity for HGF activator. To characterize the molecule for the inhibitory activity, we purified the protein and cloned its cDNA. The nucleotide sequence of the cDNA revealed that the inhibitor is a Kunitz-type of serine protease inhibitor. We designated this newly identified serine protease inhibitor as HGF activator inhibitor (HAI).


EXPERIMENTAL PROCEDURES

Cell Lines and Cell Culture

The following cell lines were obtained from the indicated sources: human melanoma cell lines C32 and A375, human lung carcinoma cell line A549, human lung fibroblast cell line HLF, human hepatoma cell lines Hep G2 and PLC/PRF/5, and human colon carcinoma cell line CCL229 from American Type Culture Collection (Rockville, MD); human stomach carcinoma cell lines HSC-3 and MKN45 and human lung carcinoma cell lines PC-9 and PC-3 from IBL (Gunma, Japan); human lung carcinoma cell line HLC-1 and human stomach carcinoma cell line IG-1 from Department of Physiology, Keio University (Tokyo, Japan); immortalized cell line of human fetal liver NuE from Dr. N. Ishida (Tohoku University, Sendai, Japan). All cells were grown in eRDF medium containing 5% fetal bovine serum. At confluence, the cells were washed twice with serum-free medium and further cultured in fresh serum-free medium for 4 days.

Assay for HGF Activator Inhibitory Activity of Conditioned Medium and Membrane Extract

The conditioned medium was harvested, clarified by centrifugation, and concentrated 20-fold by ultrafiltration using a YM30 membrane (Amicon). The concentrates were assayed for inhibitory activity toward HGF activator. To prepare membrane extract, the cells (1 × 107 cells) were washed three times with Dulbecco's sodium phosphate-buffered saline (PBS) and harvested by a cell scraper. The collected cells were suspended in 10 ml of 20 mM Tris-HCl (pH 8.0), disrupted by sonication at 4 °C, and centrifuged at 1,000 × g at 4 °C for 10 min. The supernatant was again centrifuged at 100,000 × g at 4 °C for 1 h. The pellet was suspended in 1.5 ml of PBS containing 1% Brij 58, sonicated, and centrifuged at 100,000 × g at 4 °C for 1 h. The resultant supernatant (membrane extract) was assayed for inhibitory activity toward HGF activator. Protein concentration in the conditioned medium and membrane extract was measured using a bicinchoninic acid protein assay kit (Pierce) with bovine serum albumin as a standard. The single chain precursor of HGF and HGF activator was prepared as described (6, 7). After 10 µl of 900 ng/ml HGF activator and 40 µl of concentrated conditioned medium or membrane extract were incubated at 37 °C for 30 min, 10 µl of 1.5 mg/ml single chain HGF was added as a substrate, and the mixture was further incubated at 37 °C for 2 h. The mixture was then analyzed by SDS-polyacrylamide gel electrophoresis (PAGE) under reducing conditions. The gel was stained with Coomassie Brilliant Blue, and the bands were scanned using a Flying-Spot Scanner CS-9000 (Shimadzu).

Purification of HAI from the Conditioned Medium of MKN45 Cells

To purify HAI, MKN45 cells were cultured in roller bottles (850, Falcon). At confluence, the cells were washed twice with serum-free medium and further cultured in serum-free medium. After 3-6 days, the conditioned medium was harvested. Ten liters of medium was concentrated to about 300 ml by 30 K OMEGA membrane (FILTRON) and applied to a heparin-Sepharose CL-6B (Pharmacia Biotech Inc.) column (2.5 × 10 cm) pre-equilibrated with PBS. The pass-through fractions were collected and then applied to a ConA-Sepharose (Pharmacia) column (1 × 5 cm) pre-equilibrated with PBS. Proteins were eluted with PBS containing 200 mM alpha -methyl-D-mannoside. The eluates were concentrated, and the buffer was exchanged for 10 mM sodium phosphate (pH 6.8) containing 1 M ammonium sulfate, using a YM30 membrane. The concentrates were applied to a phenyl-5PW (TOSOH) column (0.75 × 7.5 cm) pre-equilibrated with the same buffer. Proteins were eluted with a decreasing linear gradient of ammonium sulfate from 1 to 0 M. The HAI protein was eluted at 500-250 mM ammonium sulfate. The fraction was dialyzed against 20 mM Tris-HCl (pH 8.0) containing 0.05% CHAPS and then applied to an anion exchange PL-SAX (Polymer Laboratories Ltd.) column (0.46 × 5 cm) pre-equilibrated with the dialysis buffer. Elution was performed with a linear gradient of 0-500 mM NaCl. The HAI protein was eluted with 50-200 mM NaCl and then dialyzed against 5 mM sodium phosphate (pH 6.8) containing 0.05% CHAPS. The dialysate was applied to a hydroxyapatite HCA A-4007 (Mitsui-touatu Chemical) column (0.4 × 7.5 cm) pre-equilibrated with the dialysis buffer. The pass-through fraction was collected and then gel-filtrated on an Ashahipak GS520 (Asahi Chemical Industries, Ltd.) column (0.76 × 50 cm) pre-equilibrated with PBS containing 0.05% CHAPS. The HAI protein was collected at 50-30 kDa and finally purified by reverse-phase high performance liquid chromatography (HPLC) on a YMC pack C4 (YMC) column (0.46 × 15 cm). Elution was performed with a linear gradient of 10-50% acetonitrile/isopropyl alcohol (3:7) containing 0.07% trifluoroacetic acid at a flow rate of 1 ml/min for 30 min. The HAI fractions were neutralized with 1 M Tris-HCl (pH 8.0), dried under a vacuum, and then dissolved in PBS containing 0.05% CHAPS. The preparation was analyzed by SDS-PAGE and stained with silver.

The concentration of HAI in the sample was determined by the phenylthiocarbamyl method (12). The final preparation of the protein was hydrolyzed with hydrochloric acid, and the resulting free amino acids were converted to phenylthiocarbamyl derivatives (phenylthiocarbamyl-amino acid) by phenyl isothiocyanate. Phenylthiocarbamylamino acids were separated by a YMC pack ODS-A column (0.46 × 15 cm), and the quantity of HAI in the sample was calculated.

Amino Acid Sequence Analysis

To determine the NH2-terminal amino acid sequence of the purified HAI, the protein that was eluted from the C4 column and dried was reduced with 2-mercaptoethanol in 1 M Tris-HCl (pH 8.6) containing 6 M guanidine hydrochloride and 2 mM EDTA at 40 °C for 2 h. The reduced protein was then carboxymethylated with monoiodoacetic acid at room temperature for 1 h, separated by reverse-phase HPLC on a C4 column, and sequenced using an Applied Biosystems 470A Protein Sequencer. To determine the internal amino acid sequence of HAI, the protein was digested with Achromobacter protease-I. The digested peptides were separated by reverse-phase HPLC on a YMC pack C8 column (0.46 × 15 cm) and sequenced.

Dose Response of the Inhibitory Activity of HAI on HGF-converting Activity of HGF Activator and Factor XIIa

HGF activator (18 ng) or factor XIIa (38 ng) was mixed with various concentrations of HAI in 40 µl of PBS containing 0.05% CHAPS and incubated at 37 °C for 30 min. Five microliters of 1.5 mg/ml single chain HGF in PBS containing 0.05% CHAPS and 5 µl of 100 µg/ml dextran sulfate (Mr cut-off, 500,000, Sigma) were added to the mixture and further incubated (2 h for HGF activator and 24 h for factor XIIa). The mixture was analyzed by SDS-PAGE under reducing conditions. The amounts of single chain HGF and the heterodimeric form were measured by Flying-Spot Scanner. The inhibitory activity of HAI against each protease was estimated by calculating the ratio of the remaining single chain form to total HGF.

cDNA Cloning

Total RNA was prepared from MKN45 cells by acid guanidinium thiocyanate/phenol/chloroform extraction (13), and poly(A) RNA was purified by oligo(dT) affinity chromatography. The primers, 5'-GGNGCNGAYTGYTTRAA-3' and 5'-GGNGCNGAYTGYCTNAA-3' (primer 1), 5'-GTRTCYAANACRAANCC-3' and 5'-GTRTCNAGNACRAANCC-3' (primer 2), 5'-CCNCCRTANACRAANGA-3' and 5'-CCNCCRTANACRAARCT-3' (primer 3), and 5'-CCCCAYAAYTCNACYTG-3' and 5'-CCCCANAGYTCNACYTG-3' (primer 4) (N = A, G, C, or T, Y = C or T, and R = A or G) were chemically synthesized. Using primers 1 and 2, and poly(A) RNA as a template, DNA fragments were amplified by reverse transcription-polymerase chain reaction, and a 56-bp fragment was generated. The DNA fragment was subcloned and sequenced. The primer, 5'-AACAGCTTTACCG-3' (primer 5), which is part of the new sequence, was chemically synthesized. Using primers 3 and 5 and poly(A) RNA as a template, DNA fragments were amplified by reverse transcription-polymerase chain reaction. The products were further amplified using primers 4 and 5, and a 480-bp fragment was generated. Using the fragment as a probe, a human placenta cDNA library (Clontech) was screened to obtain full-length cDNA.

Northern Blotting

Total RNA (10 µg) from MKN45 cells was denatured, electrophoresed (14), and transferred to a nylon membrane (Biodyne). Human adult and fetal multiple tissue Northern blot membranes were purchased from Clontech. The membranes were hybridized at 42 °C for 16 h with the 32P-labeled probe as described (15). The membranes were washed with 1 × SSC containing 1% SDS at 50 °C. The hybridization probe was the 480-bp PCR fragment.


RESULTS

Purification of an Inhibitor of HGF Activator (HAI) from the Conditioned Medium of a Stomach Carcinoma Cell Line

We screened serum-free conditioned media from a variety of human cell lines for HGF activator-inhibitory activity. Six out of 14 conditioned medium samples contained significant inhibitory activity (Table I). Among them, five cell lines (three lung and two stomach carcinoma cell lines) produced high levels of the inhibitory activity. Thus, one (MKN45) of these cell lines was used for purification of a molecule with the inhibitory activity. An inhibitor (HAI) was purified from the serum-free conditioned medium of MKN45 by a seven-step procedure described under "Experimental Procedures". Analysis of the purified protein by SDS-PAGE revealed two bands of 39 and 40 kDa under reducing conditions (Fig. 1). When the NH2-terminal amino acid sequence of the purified protein was analyzed, only one sequence was obtained (Table II). Thus, the apparent heterogeneity of the purified protein on the gel may be caused by a difference in the COOH-terminal sequence or in glycosylation.

Table I.

Inhibition of HGF activator by conditioned medium from human cell lines


Cell line Origin Inhibitory activitya

C32 Human melanoma  -
A375 Human melanoma  -
PC-9 Human lung carcinoma ++
PC-3 Human lung carcinoma +
A549 Human lung carcinoma ++
HLC-1 Human lung carcinoma ++
HLF Human lung fibroblast ±
HSC-3 Human stomach carcinoma ++
MKN45 Human stomach carcinoma ++
IG-1 Human stomach carcinoma ±
PLC/PRF5 Human hepatoma ±
HepG2 Human hepatoma ±
NuE Immortalized cells from human fetal liver ±
CCL229 Human colon carcinoma ±

a ±, less than 10%; +, 10-50%, ++, more than 50%.


Fig. 1. SDS-PAGE of purified human HAI. The final preparation of the C4 reverse phase chromatography was analyzed by SDS-PAGE (12.5% acrylamide) under reducing conditions and stained with silver. Molecular mass markers are shown in kilodaltons on the right.
[View Larger Version of this Image (34K GIF file)]


Table II.

Amino acid sequences of HAI-derived peptides


Peptide no. Amino acid sequencea

N-terminal GPPPAPPGLPAGADCLNSFTAGVPGFVLDTXASVSNGATF
1 VQPQXPLVLK
2 SFVYGGXLGNK
3 DVENTDWRLLRGDTDVRVERK
4 AWAGIDLK
5 DPNQVELWGLK
6 XGTYLFQLTV

a Amino acids not determined are denoted by X.

To determine the internal amino acid sequences, the purified protein was carboxymethylated and digested with Achromobacter protease-I, and the resulting peptide fragments were separated by reverse-phase HPLC. Six partial sequences were determined (Table II). None of the peptide sequences nor the NH2-terminal sequence matched those in the Swiss Prot or NBRF protein sequence data bases, indicating that HAI is a novel protein.

Properties of the Purified HAI

Fig. 2 shows the dose-response curve of the inhibitory activity of HAI. In these reactions, HGF activator (450 ng/ml) was mixed with various concentrations of purified HAI and incubated for 30 min to form an enzyme-inhibitor complex. Then remaining HGF-converting activity in the mixture was measured. The concentration of HAI for 50% inhibition was about 250 ng/ml. Considering the molecular masses of HAI and HGF activator, HAI forms about an equimolar complex with HGF activator within 30 min.


Fig. 2. Dose dependence of the inhibitory activity of HAI toward the HGF-converting activity of HGF activator and factor XIIa. HGF activator or factor XIIa was incubated with various concentrations of HAI. Single chain HGF (sc-HGF) was then added and further incubated. The mixture was then analyzed by SDS-PAGE (A). The inhibitory activity toward HGF activator (open circle ) and factor XIIa (bullet ) was determined as the ratio of the remaining single chain HGF to total HGF (B).
[View Larger Version of this Image (41K GIF file)]


HGF activator is homologous to blood coagulation factor XIIa. Factor XIIa can activate single chain HGF in vitro, although the specific activity of factor XIIa is lower than that of HGF activator (9). We therefore examined whether or not HAI inhibits the HGF-converting activity of factor XIIa and found that it did not, even when a 5-fold molar excess of HAI was incubated with factor XIIa (Fig. 2). Thus, HAI is specific for HGF activator in HGF-converting activity.

Isolation of cDNA Clone and DNA Sequence Analysis

Two hexapeptide sequences, Gly-Ala-Asp-Cys-Leu-Asn and Gly-Phe-Val-Leu-Asp-Thr in the NH2-terminal sequence (Table II), were used to design degenerate oligonucleotide primers for PCR amplification of the sequence for the NH2-terminal region. PCR amplification of MKN45 RNA resulted in a cDNA fragment with the expected size of about 56 bp. The cDNA fragment was subcloned and sequenced. The cDNA clone encoded 19 amino acids, including 7 amino acids in the NH2-terminal amino acid sequence between the two hexapeptide sequences used for the primer design. Thirteen nucleotides from the obtained sequence were used as a 5' primer for further PCR amplification together with a sequence corresponding to the hexapeptide Ser-Phe-Val-Tyr-Gly-Gly in peptide 2 (Table II) as a 3' primer. The PCR amplification products were further amplified using the 5' primer and the sequence corresponding to the hexapeptide Gln-Val-Glu-Leu-Trp-Gly in peptide 5 (Table II) as a 3' primer. The PCR amplification resulted in a cDNA fragment of about 480 bp. The cDNA fragment was subcloned and sequenced. The cDNA clone encoded 160 amino acids, including the sequences of peptides 1, 3, and 4 in Table II. Northern blotting using the PCR clone as a probe revealed that the human placenta produced significant amounts of mRNA of HAI. Thus, a cDNA library from the human placenta (Clontech) was screened using the PCR clone as a probe. Eighty-four hybridization-positive clones were obtained from about 4 × 105 phage. The largest clone was sequenced to determine the primary structure of human HAI. The determined nucleotide sequence of the cDNA is shown in Fig. 3.


Fig. 3. The cDNA and deduced amino acid sequences of human HAI. The cDNA sequence was translated from the first ATG codon in the open reading frame. Nucleotide and amino acid numbers are shown on the left. The cleavage site between the putative signal peptide and the mature protein is indicated by an arrow. Potential N-glycosylation sites are indicated by triangle . Regions corresponding to the amino acid sequences of the peptides derived from purified HAI are underlined at the amino acid sequence. A putative transmembrane region is double-underlined. The Kunitz domains and the LDL receptor-like domain are indicated by boxes 1, 3, and 2, respectively. The cysteine residues in the domains are circled. The possible polyadenylation signal AATAAA is underlined at the nucleotide sequence.
[View Larger Version of this Image (73K GIF file)]


Predicted Amino Acid Sequence of HAI

The amino acid sequence of HAI deduced from the cDNA sequence is also shown in Fig. 3. The translation initiation site was assigned to the first methionine codon because the sequence GCG<UNL>ATG</UNL>G matches a favorable Kozak consensus sequence (16). This methionine is followed by a hydrophobic region (Fig. 4), and the NH2-terminal amino acid of the purified protein is located at 36th residue downstream of the methionine. Thus, the hydrophobic region may represent a signal peptide sequence. Because the reading frame is open upstream of the first methionine codon, it is possible that translation is initiated further upstream beyond the boundary of the cDNA sequence. The open reading frame that starts from the putative ATG codon consists of 513 amino acids, and the protein product has a calculated molecular mass of 56,893. Excluding the putative signal peptide, the mature form of the protein consists of 478 amino acids and has a calculated molecular mass of 53,319. The apparent molecular mass of HAI purified from the conditioned medium of MKN45 cells was about 40 kDa, as determined by SDS-PAGE. Thus, the protein purified from the conditioned medium appears to be a processing product cleaved at the COOH-terminal region. A hydrophobic region of 23 amino acids is present in the COOH-terminal region (Fig. 4), suggesting that the primary translation product is a membrane-associated protein. There are three potential N-glycosylation sites with the canonical Asn-X- (Ser/Thr). A comparison of the predicted protein sequence of HAI with sequences in the Swiss Prot and NBRF protein data base revealed three regions with characteristic structural features. The two regions (residues 250-300 and 375-425) showed extensive similarity to the Kunitz-type sequence of serine protease inhibitors (Fig. 5A). Thus, HAI appears to be a Kunitz-type serine protease inhibitor. The other region (residues 319-353) located between the two Kunitz domains showed similarity to the ligand binding domain of the low density lipoprotein (LDL) receptor and related proteins (Fig. 5B). Other regions did not show clear-cut similarity to any protein sequences within the data base entries.


Fig. 4. Hydropathy profile of HAI. The hydropathy profile of HAI was computed according to Kyte and Doolittle (34), using a window size of 21 residues.
[View Larger Version of this Image (11K GIF file)]



Fig. 5. Alignment of the amino acid sequence of the Kunitz domains (A) and the LDL receptor-like domain (B) of HAI with human homologous protein sequences. APP sequence is from Ponte et al. (19), APP homolog protein from Sprecher et al. (35), Ialpha TI from Kaumeyer et al. (36), TFPI from Wun et al. (28), collagen alpha 3 chain from Chu et al. (37), LDL receptor from Yamamoto et al. (38), very low density lipoprotein (VLDL) receptor from Sakai et al. (39), LDL receptor related protein (LRP) from Herz et al. (40), and perlecan from Murdoch et al. (41).
[View Larger Version of this Image (31K GIF file)]


Detection of Inhibitory Activity toward HGF Activator in Membrane Fraction of MKN45 Cells

The cDNA sequence of HAI suggests that the primary translation product is a membrane-associated protein. We therefore examined whether inhibitory activity toward HGF activator was detected in membrane fraction. Membrane extract and conditioned medium of MKN45 cells were prepared and assayed for the inhibitory activity. Protein concentration in the membrane extract and conditioned medium was 290 and 620 µg/ml, respectively. Significant activity was detected in the membrane extract (Fig. 6). Considering the protein concentration in each fraction, the activity in the membrane fraction was about 80% that in the conditioned medium. These results suggest that HAI may be produced as a membrane-associated form and is secreted as a proteolytically truncated form.


Fig. 6. Inhibitory activity toward HGF activator of conditioned medium and membrane extract of MKN45 cells. HGF activator was incubated with various volumes of conditioned medium or membrane extract of MKN45 cells. Single chain HGF (sc-HGF) was then added and further incubated. The mixture was analyzed by SDS-PAGE (A). The inhibitory activity of conditioned medium (open bars) and membrane extract (closed bars) was determined as the ratio of the remaining single chain HGF to total HGF (B).
[View Larger Version of this Image (50K GIF file)]


Tissue Distribution of HAI mRNA

We determined the size and tissue distribution of HAI mRNA by Northern blotting with poly(A) RNAs from various human tissues (Fig. 7). A major transcript of 2.5 kb was detected in MKN45 cells where we purified the HAI protein. In addition, a minor transcript of 5.6 kb was also detected in MKN45 cells. A transcript of 2.5 kb was detected in a variety of human adult and fetal tissues. Among them, the expression level of HAI mRNA was relatively high in the adult placenta, kidney, pancreas, prostate, and small intestine. It was also high in the fetal kidney. However, although the level was low in the adult lung, it was relatively high in the fetal lung.


Fig. 7. Northern blot analysis of HAI mRNA. Ten micrograms of MKN45 total RNA (A) and 5 µg of poly(A) RNAs from various adult (B) and fetal (C) human tissues were analyzed using HAI cDNA as a probe. Size markers are human rRNA (A) or are indicated in kb (B and C).
[View Larger Version of this Image (36K GIF file)]



DISCUSSION

In this study, we found an inhibitory activity toward HGF activator in serum-free conditioned media of various human cell lines. We purified the inhibitor protein, HAI, from the conditioned medium of MKN45 stomach carcinoma cells. The purified HAI has a molecular mass of about 40 kDa. The primary structure of the protein was predicted from the sequence of the cDNA for human HAI. The structure of human HAI is schematically summarized in Fig. 8. The primary translation product consists of 513 amino acid residues. The NH2-terminal 35 residues may serve as a signal peptide. The mature protein appears to be membrane-bound, because a hydrophobic region of about 20 amino acids is present at the COOH-terminal region. HAI has two well defined Kunitz domains. The Kunitz domain is typically about 60 amino acids in length and contains three disulfide bonds. It was first recognized as the functional domain of bovine pancreatic trypsin inhibitor (17). Thereafter, the domain was found in several mammalian serine protease inhibitors. Thus, one or both of the Kunitz domains found in HAI appear to be responsible for the inhibitory activity of the protein.


Fig. 8. A schematic representation of human HAI. TM indicates the possible transmembrane region.
[View Larger Version of this Image (14K GIF file)]


The first and second Kunitz domains of human HAI show the highest homology (47% identity) to those of human beta -amyloid precursor protein (APP) and human APP homolog protein, respectively. APP is the precursor protein of amyloid beta -protein which is present in neutritic plaque and cerebrovascular deposits in individuals with Alzheimer's disease and Down's syndrome (18). The Kunitz domain is located in the middle of APP (19-21), and it efficiently functions as an inhibitor of several serine proteases (22). The primary translation product of APP has a hydrophobic sequence at the COOH-terminal region, and thus it appears to be a membrane-bound protein. Oltersdorf et al. (23) and Van Nostrand et al. (24) reported that protease nexin II (PNII) is a secreted form of APP. PNII is a protease inhibitor that forms SDS-resistant inhibitory complexes with epidermal growth factor-binding protein, the gamma -subunit of nerve growth factor, and trypsin. Smith et al. (25) reported that an inhibitor of coagulation factor XIa purified from the serum-free conditioned medium of HepG2 liver cells is also a secreted form of APP. Truncated forms of APP are derived from their cognate membrane-associated forms by proteolysis and have apparently lost the cytoplasmic and the transmembrane domains (26). Thus, PNII and factor XIa inhibitor are proteolytically truncated forms of the transmembrane form of APP. HAI purified from the serum-free conditioned medium of MKN45 cells has a molecular mass of about 40 kDa, which is smaller than the predicted molecular mass (53,319) of the primary translation product, and thus it may lack the putative membrane-associated and cytoplasmic domains. Because the HGF activator-inhibitory activity was detected in the membrane fraction of MKN45 cells, it is likely that the primary translation product of HAI is a membrane-associated form. Thus, like PNII and factor XIa inhibitor, the secreted form of HAI appears to be a proteolytically truncated form of the membrane-associated form of HAI.

HAI has two Kunitz domains interrupted by another domain. Two serine protease inhibitors with two or three Kunitz domains have been identified. Inter-alpha -trypsin inhibitor (Ialpha TI) found in mammalian plasma is a high molecular weight glycoprotein with two tandemly repeated Kunitz domains in the light chain, which is one of three polypeptide chains linked by a glycosaminoglycan (27). Tissue factor pathway inhibitor (TFPI), which was also found in mammalian plasma, has three tandemly repeated Kunitz domains (28). TFPI inhibits activated factor X(Xa) directly and, in a Xa-dependent manner, inhibits VII(a)/tissue factor activity by forming a quaternary Xa·TFPI·VII(a)·tissue factor complex. The Kunitz domains in these inhibitors are not interrupted by another domain. Trypsin and chymotrypsin form equimolar complexes with Ialpha TI. Furthermore, a protease inhibitor, which consists of only the second Kunitz domain of the light chain of Ialpha TI (29), has been identified from mast cells (30). This inhibitor, named trypstatin, markedly inhibits factor Xa and tryptase and also inhibits trypsin and chymase. Thus, the second Kunitz domain in the light chain of Ialpha TI is required for the inhibition of serine proteases, whereas the first domain does not seem to be required for this function. Site-directed mutagenesis of TFPI has revealed that the second Kunitz domain is required for the efficient binding and inhibition of Xa, that both Kunitz domains 1 and 2 are required for the inhibition of VIIa/tissue factor activity, but that the third Kunitz domain does not seem to be required for these functions (31). The dose-response curve of the HAI activity showed that HAI purified from the conditioned medium seems to form an equimolar complex with HGF activator. Furthermore, the proteolytic cleavage to produce the extracellular truncated form of HAI appears to occur within the second Kunitz domain, because the size (about 40 kDa) of the protein implies that it consists of about 360 amino acids. Thus, the first Kunitz domain in HAI may be sufficient for it to exert activity.

The other characteristic structural domain in HAI is located between the two Kunitz domains. It consists of about 40 amino acid residues and bears a close resemblance to the "cysteine domain" repeat of the LDL receptor and related proteins. The LDL receptor has seven repeats of the domain (32). Each repeat has six cysteine residues, all of which are involved in disulfide bonds, and in addition it has several negatively charged amino acid residues at the COOH-terminal region. The clustering of cysteine residues in the domain of HAI is similar to that in the repeat of the LDL receptor. Furthermore, the domain of HAI has 8 negatively charged but only 2 positively charged amino acid residues. The negatively charged domain in the LDL receptor is believed to be the binding site for its positively charged apoprotein ligand (32). Although the significance of the negatively charged domain in HAI remains to be established, it may be involved in formation of the inhibitor-enzyme complex, because HGF activator shows high affinity to negatively charged substances.

HGF activator is produced mainly in the liver, and it normally circulates in the blood as an inactive zymogen. In response to tissue injury, the zymogen is activated by proteolytic processing exclusively in the injured tissue (10). The activated HGF activator acquires strong affinity for heparin, which may ensure the localization of the enzyme in injured tissue (10). This localized HGF activator activates single chain HGF that is also associated with a heparin-like molecule on the cell surface. Thus, the activity of HGF activator may be regulated by tissue-derived inhibitors. Because human HAI mRNA is expressed in various tissues, the HAI protein produced by cells of these tissues may be responsible for inhibiting HGF activator. However, HAI mRNA expression is low in some tissues, including the liver and lung. HGF is thought to play a crucial role in repair of liver and lung following injury (1, 33). In these tissues, production of HAI could be induced during tissue repair. Alternatively, another inhibitor(s) may function in these tissues. Characterizations of these inhibitors in injured tissues are needed to understand mechanisms for regulating the activation of HGF.


FOOTNOTES

*   The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The nucleotide sequence(s) reported in this paper has been submitted to the GenBankTM/EMBL Data Bank with accession number(s) AB000095[GenBank].


par    To whom correspondence should be addressed: Dept. of Life Science, Faculty of Bioscience and Biotechnology, Tokyo Institute of Technology, Nagatsuta, Midori-ku, Yokohama 226, Japan. Tel.: 81-45-924-5701; Fax: 81-45-924-5771.
1   The abbreviations used are: HGF, hepatocyte growth factor; HAI, HGF activator inhibitor; PAGE, polyacrylamide gel electrophoresis; PBS, phosphate-buffered saline; CHAPS, 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonic acid; HPLC, high performance liquid chromatography; LDL, low density lipoprotein; APP, beta -amyloid precursor protein; PNII, protease nexin II; Ialpha TI, inter-alpha -trypsin inhibitor; TFPI, tissue factor pathway inhibitor; bp, base pair(s); kb, kilobase pair(s).

REFERENCES

  1. Miyazawa, K., Shimomura, T., Naka, D., and Kitamura, N. (1994) J. Biol. Chem. 269, 8966-8970 [Abstract/Free Full Text]
  2. Naldini, L., Tamagnone, L., Vigna, E., Sachs, M., Hartmann, G., Birchmeier, W., Daikuhara, Y., Tsubouchi, H., Blasi, F., and Comoglio, P. M. (1992) EMBO J. 11, 4825-4833 [Abstract]
  3. Gohda, E., Tsubouchi, H., Nakayama, H., Hirono, S., Sakiyama, O., Takahashi, K., Miyazaki, H., Hashimoto, S., and Daikuhara, Y. (1988) J. Clin. Invest. 81, 414-419 [Medline] [Order article via Infotrieve]
  4. Igawa, T., Kanda, S., Kanetake, H., Saitoh, Y., Ichihara, A., Tomita, Y., and Nakamura, T. (1991) Biochem. Biophys. Res. Commun. 174, 831-838 [Medline] [Order article via Infotrieve]
  5. Rubin, J. S., Chan, A. M.-L., Bottaro, D. P., Burgess, W. H., Taylor, W. G., Cech, A. C., Hirschfield, D. W., Wong, J., Miki, T., Finch, P. W., and Aaronson, S. A. (1991) Proc. Natl. Acad. Sci. U. S. A. 88, 415-419 [Abstract]
  6. Shimomura, T., Ochiai, M., Kondo, J., and Morimoto, Y. (1992) Cytotechnology 8, 219-229 [Medline] [Order article via Infotrieve]
  7. Miyazawa, K., Shimomura, T., Kitamura, A., Kondo, J., Morimoto, Y., and Kitamura, N. (1993) J. Biol. Chem. 268, 10024-10028 [Abstract/Free Full Text]
  8. Mars, W. M., Zarnegar, R., and Michalopoulos, G. K. (1993) Am. J. Pathol. 143, 949-958 [Abstract]
  9. Shimomura, T., Miyazawa, K., Komiyama, Y., Hiraoka, H., Naka, D., Morimoto, Y., and Kitamura, N. (1995) Eur. J. Biochem. 229, 257-261 [Abstract]
  10. Miyazawa, K., Shimomura, T., and Kitamura, N. (1996) J. Biol. Chem. 271, 3615-3618 [Abstract/Free Full Text]
  11. Shimomura, T., Kondo, J., Ochiai, M., Naka, D., Miyazawa, K., Morimoto, Y., and Kitamura, N. (1993) J. Biol. Chem. 268, 22927-22932 [Abstract/Free Full Text]
  12. Stone, K. L., and Williams, K. R. (1986) J. Chromatogr. 359, 203-212 [CrossRef][Medline] [Order article via Infotrieve]
  13. Chomczynski, P., and Sacchi, N. (1987) Anal. Biochem. 162, 156-159 [CrossRef][Medline] [Order article via Infotrieve]
  14. Lehrach, H., Diamond, D., Wozney, J. M., and Boedtker, H. (1977) Biochemistry 16, 4743-4751 [Medline] [Order article via Infotrieve]
  15. Miyazawa, K., Kitamura, A., Naka, D., and Kitamura, N. (1991) Eur. J. Biochem. 197, 15-22 [Abstract]
  16. Kozak, M. (1987) Nucleic Acids Res. 15, 8125-8148 [Abstract]
  17. Laskowski, M., Jr., and Kato, I. (1980) Annu. Rev. Biochem. 49, 593-626 [CrossRef][Medline] [Order article via Infotrieve]
  18. Kang, J., Lemaire, H.-G., Unterbeck, A., Salbaum, J. M., Masters, C. L., Grzeschik, K.-H., Multhaup, G., Beyreuther, K., and Müller-Hill, B. (1987) Nature 325, 733-736 [CrossRef][Medline] [Order article via Infotrieve]
  19. Ponte, P., Gonzalez-DeWhitt, P., Schilling, J., Miller, J., Hsu, D., Greenberg, B., Davis, K., Wallace, W., Lieberburg, I., Fuller, F., and Cordell, B. (1988) Nature 331, 525-527 [CrossRef][Medline] [Order article via Infotrieve]
  20. Tanzi, R. E., McClatchey, A. I., Lamperti, E. D., Villa-Komaroff, L., Gusella, J. F., and Neve, R. L. (1988) Nature 331, 528-530 [CrossRef][Medline] [Order article via Infotrieve]
  21. Kitaguchi, N., Takahashi, Y., Tokushima, Y., Shiojiri, S., and Ito, H. (1988) Nature 331, 530-532 [CrossRef][Medline] [Order article via Infotrieve]
  22. Sinha, S., Dovey, H. F., Seubert, P., Ward, P. J., Blacher, R. W., Blaber, M., Bradshaw, R. A., Arici, M., Mobley, W. C., and Lieberburg, I. (1990) J. Biol. Chem. 265, 8983-8985 [Abstract/Free Full Text]
  23. Oltersdorf, T., Fritz, L. C., Schenk, D. B., Lieberburg, I., Johnson-Wood, K. L., Beattie, E. C., Ward, P. J., Blacher, R. W., Dovey, H. F., and Shinha, S. (1989) Nature 341, 144-147 [CrossRef][Medline] [Order article via Infotrieve]
  24. Van Nostrand, W. E., Wagner, S. L., Suzuki, M., Choi, B. H., Farrow, J. S., and Cunningham, D. D. (1989) Nature 341, 546-549 [CrossRef][Medline] [Order article via Infotrieve]
  25. Smith, R. P., Higuchi, D. A., and Broze, G. J., Jr. (1990) Science 248, 1126-1128 [Medline] [Order article via Infotrieve]
  26. Weidemann, A., König, G., Bunke, D., Fischer, P., Salbaum, J. M., Masters, C. L., and Beyreuther, K. (1989) Cell 57, 115-126 [Medline] [Order article via Infotrieve]
  27. Sailer, J.-P. (1990) Trends Biochem. Sci. 15, 435-439 [CrossRef][Medline] [Order article via Infotrieve]
  28. Wun, T.-C., Kretzmer, K. K., Girard, T. J., Miletich, J. P., and Bronze, G. J., Jr. (1988) J. Biol. Chem. 263, 6001-6004 [Abstract/Free Full Text]
  29. Itoh, H., Ide, H., Ishikawa, N., and Nawa, Y. (1994) J. Biol. Chem. 269, 3818-3822 [Abstract/Free Full Text]
  30. Kido, H., Yokogoshi, Y., and Katunuma, N. (1988) J. Biol. Chem. 263, 18104-18107 [Abstract/Free Full Text]
  31. Girard, T. J., Warren, L. A., Novotny, W. F., Likert, K. M., Brown, S. G., Miletich, J. P., and Broze, G. J., Jr. (1989) Nature 338, 518-520 [CrossRef][Medline] [Order article via Infotrieve]
  32. Südhof, T. C., Goldstein, J. L., Brown, M. S., and Russell, D. W. (1985) Science 228, 815-822 [Medline] [Order article via Infotrieve]
  33. Yanagita, K., Matsumoto, K., Sekiguchi, K., Ishibashi, H., Niho, Y., and Nakamura, T. (1993) J. Biol. Chem. 268, 21212-21217 [Abstract/Free Full Text]
  34. Kyte, J., and Doolittle, R. F. (1982) J. Mol. Biol. 157, 105-132 [Medline] [Order article via Infotrieve]
  35. Sprecher, C. A., Grant, F. J., Grimm, G., O'Hara, P. J., Norris, F., Norris, K., and Foster, D. C. (1993) Biochemistry 32, 4481-4486 [Medline] [Order article via Infotrieve]
  36. Kaumeyer, J. F., Polazzi, J. O., and Kotick, M. P. (1986) Nucleic Acids Res. 14, 7839-7850 [Abstract]
  37. Chu, M.-L., Zhang, R.-Z., Pan, T.-C., Stokes, D., Conway, D., Kuo, H.-J., Glanville, R., Mayer, U., Mann, K., Deutzmann, R., and Timpl, R. (1990) EMBO J. 9, 385-393 [Abstract]
  38. Yamamoto, T., Davis, C. G., Brown, M. S., Schneider, W. J., Casey, M. L., Goldstein, J. L., and Russell, D. W. (1984) Cell 39, 27-38 [Medline] [Order article via Infotrieve]
  39. Sakai, J., Hoshino, A., Takahashi, S., Miura, Y., Ishii, H., Suzuki, H., Kawarabayasi, Y., and Yamamoto, T. (1994) J. Biol. Chem. 269, 2173-2182 [Abstract/Free Full Text]
  40. Herz, J., Hamann, U., Rogne, S., Myklebost, O., Gausephol, H., and Stanley, K. K. (1988) EMBO J. 7, 4119-4127 [Abstract]
  41. Murdoch, A. D., Dodge, G. R., Cohen, I., Tuan, R. S., and Iozzo, R. V. (1992) J. Biol. Chem. 267, 8544-8557 [Abstract/Free Full Text]

©1997 by The American Society for Biochemistry and Molecular Biology, Inc.