Characterization of Elastolytic Protease, Asnilase from Aspergillus nidulans and Its Cytotoxicity on Human Endothelial and Epithelial Cells

Aims: To elucidate the pathogenicity of Aspergillus nidulans , elastolytic protease was isolated from the culture supernatant of the isolate, and its biological activity and cytotoxicity were examined. Methodology: A. nidulans (NBRC 4340) spores were cultured on YCB-elastin medium for three days and elastolytic protease was isolated from the culture supernatant by DEAE-Cellulose anion exchange chromatography. Elastolytic activity was measured by using GAAPLNA (Glutaryl-L-alanylOriginal Research Article


INTRODUCTION
Aspergillosis is one of the opportunistic infectious diseases caused by fungus, and patients with aspergillosis increase with the development of the medical technology. Invasive pulmonary aspergillosis, a common mycosis in immunocompromised hosts undergoing chemotherapy, is caused by inhalation of high concentrations of spores from Aspergillus species, and becomes a significant public health problem. Several species of Aspergillus are responsible for the diseases, aspergillosis. They are A. fumigatus, A. flavus, A. niger and A. terreus [1,2]. A. nidulans is a typical soil fungus with a world-wide distribution and is also a causative agent of aspergillosis in humans and animals [1].
Correlation of elastolytic protease produced by Aspergillus species with ability to cause pulmonary invasive aspergillosis has been discussed [3][4][5][6]. Elastin fibres are found in vertebrate tissues and essential to maintain shape and mechanical rigidity. Since human lung is elastin-rich tissue [7], the destruction of the elastin is thought to be involved in pathogenicity. Various extracellular serine proteases and elastolytic proteases were purified from A. fumigatus that was most common species isolated from invasive aspergillosis, and their properties have been reported [8][9][10][11][12]. Although the presence of protease in different strains of A. flavus, A. terreus, A. parasiticus and A. nidulans were also reported, however, very few studies were made whether the enzyme was related to pathogenesis [13][14][15][16][17][18]. Therefore, the purified elastolytic enzyme was injected into lung tissues directly to see any pathogenic effect of these enzymes [19,20]. It was found from the present study that the enzyme induced the inflammatory cells to accumulate in the lung and hemorrhagic pneumonia was produced. In this paper, we report biochemical properties of novel elastolytic protease from A. nidulans which was named as Asnilase, and investigations of its injury effects on cultural cells.

Materials
A. nidulans NBRC 4340 was obtained from National Institute of Technology and Evaluation Biological Resource Center. Yeast carbon base (YCB) was purchased from Difco Lab. and DEAE-Cellulose was from Whatman Inc. A standard kit for the molecular weight determination was purchased from Sigma-Aldrich Co. LLC., pI marker was from SERVA Electrophoresis GmbH, and Bio-lyte (pH 3.5-10) was obtained from Bio-Rad Laboratories. Inc. Human fibrinogen was supplied by Sigma-Aldrich Co., Ltd. Collagen types I, II, III, and IV were obtained from Wako Pure Chemical Industries, Ltd.

Production of Elastolytic Protease
A sufficient number of A. nidulans spores were cultured in 100 ml YCB-elastin medium (a synthetic medium of 1% yeast carbon base containing 1% elastin) as a nitrogen source. The cultures were incubated for 3 days at 37°C with shaking. The culture broth was filtered through a 0.22 µm pore size membrane filter, and the cellfree culture supernatant was used as the source of crude enzyme.

Isolation and Characterization of Elastolytic Proteinase
One hundred milliliters of the culture filtrate from A. nidulans was applied on a DEAE-Cellulose equilibrated with 10 mM Tris-HCl buffer (pH 7.2) containing 10 mM NaCl. The column was eluted with equilibration buffer and 10 mM Tris-HCl buffer (pH 7.2) containing 0.5 M NaCl.
The molecular mass of the enzyme was determined by SDS-PAGE [21], and the isoelectric point (pI) was estimated by isoelectric focusing/polyacrylamide gel electrophoresis using bio-lyte concentration of 40% (w/v) with a pH range of 3.5-10. Protein concentration was measured by the method of Bradford [22] using bovine serum albumin as the standard.

Assay for Elastolytic Activity
Elastolytic activity of the enzyme was assayed by the diazo coupling method that measuring pnitroanilide (p-NA) released from the synthetic substrate [23]. Briefly, 0.1 ml of the enzyme solution, 20 µl of 50 mM GAAPLNA solution dissolved in DMSO and 0.9 ml 50 mM Tris-HCl buffer (pH 7.5) were incubated for 60 minutes at 37°C. The reaction was stopped by adding 1 ml of 10% trichloroacetic acid. Then 0.2 ml of 0.1% sodium nitrate, 0.2 ml of 0.5% ammonium sulfamate, and 0.2 ml of 0.1% N-1naphthylethylenediamine dihydrochloride were added to the reaction mixture, and the amount of p-nitroaniline released was quantified by measuring the absorbance at 550 nm. One unit of GAAPLNA hydrolase activity was defined as the amount of enzyme that hydrolyzed 1 µmol of substrate per minute.
The effects of inhibitors on elastolytic activity were determined after the incubation of enzyme with inhibitor at 37°C for 10 minutes.

Other Enzyme Activities
Fibrinogenase or collagenase activities of the enzyme were measured by monitoring the time course degradation of these proteins on SDS polyacrylamide gel. Protease activity was assayed using keratin, hide powder azure, azocasein and azoalbumin as a substrate [24][25][26]. Oxidized insulin B chain was used for determination of enzyme specificity. Briefly, the hydrolyzed peptides were fractionated by a reversed-phase HPLC column (Develosil 300 ODS-7), and the fragments were identified using Applied Biosystems 491 protein sequencer and 120A PTH analyzer.

Cytotoxicity Assay
Cultured human cells were trypsinized, resuspended in the medium and seeded into 96multiwell plates at a density of 1.5×10 4 cells/well. Purified enzyme was diluted in sterilized saline at an appropriate concentration and then added to the cells and incubated at 37°C for 24 hours. The viable cell numbers were determined by the colorimetry by means of a cell counting kit based on the tetrazolium salt/formazan system [27,28]. For fluorescence microscopy, cell staining was performed with 0.5 mM of 5-chloromethylfluorescein diacetate, CellTracker TM Green CMFDA in serum-free medium according to the instruction of supplier. Fluorescence microscope system, BZ-X700 (Keyence, Osaka, Japan) was used for the observation.

Isolation and Characterization of Elastolytic Protease
The adsorbed fraction of culture filtrate which was eluted from DEAE-Cellulose column with 0.5M NaCl contained elastolytic protease, and the enzyme activity was found in the fractions from 210 to 240 (Fig. 1). The fraction with elastolytic activity was homogeneous with various tests such as polyacrylamide gel electrophoresis, isoelectric focusing and SDS-PAGE. This protein is named as "Asnilase" (elastase from A. nidulans). Approximately 1.7 mg of Asnilase was obtained from 100 mL of the culture filtrate. The molecular mass of Asnilase was 33,800 Da with the isoelectric point of 4.2. Enzyme specificity of Asnilase was measured by synthetic peptides GAAPLNA, STANA and SAPANA for elastase. Among these substrates, GAAPLNA was hydrolyzed most actively and the specific activity was 334.9 µmole/min/mg protein.
Asnilase is sensitive to high temperature and lost all enzymatic activity at 60°C although 64% activity still remained at 50°C ( Table 1). The enzyme activity of Asnilase was stable over pH range of 6-10 and highest at pH of 7.5. The enzyme activity was completely lost by the inhibitors p-amidinophenyl methylsulfonylfluoride (APMSF) and N-bromosuccinimide (NBS) at the final concentrations of 2.5 mM and pchloromercuribenzoic acid (PCMB) at 1 mM when GAAPLNA was as the substrate. Ethylene glycol tetraacetic acid (EGTA) at 1 mM also showed some inhibitory effect and only 17.9% of enzyme activity retained. However, neither dithiothreitol nor benzamidine had any inhibitory effect.
Several elastolytic proteases [8,10,11,13,14] and alkaline proteases [9,12,15,16] have been isolated from Aspergillus species. Various extracellular proteases were also purified from A. nidulans and characterized [17,18,[29][30][31], and among these enzymes, proteinase I and II [17] and alkaline protease [18] showed similar molecular mass, isoelectric points and stability to Asnilase. The effects of inhibitors on the enzymes derived from A. nidulans, A. fumigatus and A. flavus were compared in Table 1. These elastolytic enzymes were affected by serine protease inhibitors such as APMSF, PMSF and DFP. However substrate specificity was different for these elastolytic enzymes.

Hydrolyzing Effects on Mammalian Tissue Proteins
Two important proteins commonly present in the mammalian tissues were selected for substrate specificity. One is fibrinogen that is important for blood coagulation and the other is collagen that is important for tissues. When human fibrinogen was incubated with Asnilase, Aα chain was shown to be most sensitive to proteolytic digestion and the band was completely disappeared within 15 minutes (Fig. 2, left). The Bβ chain of fibrinogen was also hydrolyzed but it required longer time of incubation. It seems the γ chain of fibrinogen was resistant to hydrolysis with Asnilase. The effect of Asnilase on collagen type I, II, III, and IV were also determined and type IV collagen was significantly hydrolyzed by this enzyme (Fig. 2, right). Type I and III collagen were slightly digested, whereas type II collagen was not degraded. Similar to Asnilase, elastolytic proteases from A. fumigatus and A. flavus digested collagen type I, III and IV, and in addition, the enzyme from A. fumigatus hydrolyzed collagen type II (Table 1).
In order to find which peptide bond can be cleaved by Asnilase, the oxidized insulin B chain was used and its proteolytic specificity was compared with other elastolytic proteases. Asnilase hydrolyzed only His(5)-Leu (6) and Glu(13)-Ala (14) bonds, however enzymes from A. fumigatus and A. flavus cleaved insulin B chain at 7 and 11 sites, respectively [14]. This indicates that the substrate specificity of Asnilase is selective and is different from other elastolytic enzymes.

Incubation was performed for indicated time intervals and the reaction mixture was analyzed by SDS-PAGE
Patterson et al. [1] reported that A. fumigatus is the most common species found in invasive aspergillosis (66% of 261 cases) and responsible for severe illness, while the rates of detection of A. flavus (14%) and A. nidulans (0.4%) are relatively low. The pathogenic difference among these species possibly depends on the substrate specificity. The broad spectrum of A. fumigatus enzyme against various collagen types and other proteins may be responsible for the destruction of various organ and tissues.

Cytotoxic Effect on Human Cells
Cultured human cells (HPAEC: human pulmonary artery endothelial cells, HSAEC: human small airway epithelial cells, HBTEC: human bronchial/tracheal epithelial cells, HPAEpiC: Human pulmonary alveolar epithelial cells) were employed to determine the cytotoxic activity of Asnilase (Fig. 3). After the exposure to the enzyme, the average of living cell count in each well was clearly decreased in a concentration-dependent manner, and the 50% effective concentrations (EC 50 ) were 3.0 µg/ml for HPAEC, 2.0 µg/ml for HSAEC and 6.0 µg/ml for HPAEpiC, respectively. Endothelial cells were most sensitive to Asnilase and viable cell count drastically decreased for the concentrations of 2.0 to 4.0 µg/ml. Pulmonary alveolar epithelial cells also die in a similar manner for the concentrations of 3.0 to 10 µg/ml. The effect of Asnilase to bronchial/tracheal epithelial cells was relatively moderate, and more than 50% of cells were alive at the concentration of 10 µg/ml. The adhered cells to the surface of well were stained with fluorescence dye and observed under a fluorescence microscope. The concentrations of Asnilase which did not cause cell death clearly decreased the adhesive property of the cells to the plate surface. For example, 1.0 µg/mL of Asnilase did not affect the viability of HPAEC with 24 hours exposure (Fig.  3) whereas remaining HPAEC on the plate surface was clearly decreased with 1.0 µg/mL of this enzyme. As shown in Fig. 4B, detachment of most of the cells was obvious and very few adhered cells remained on the plate surface after 24 hours. Time lapse observation of cells indicated that cytomorphologic changes started after 3 hours exposure and detachment of cells gradually occurs subsequently (data not shown). These phenomena are remarkable in HPAEC, but not observed clearly with HBTEC. Reichard et al. [8] reported that Alp (extracellular serine proteinase from A. fumigatus) efficiently detached Vero cells from a plastic surface at concentrations of 1.0 µg/mL after 10 hours exposure. The detached Vero cells were viable and could be propagated in culture, however the effect of Asnilase on HPAEC by was irreversible. The difference between the effect of Alp and Asnilase may be due to difference in cell strain which used. From the mentioned results, Asnilase induce cytomorphologic changes of cells to decrease adhesive property at first, and the exfoliated cells subsequently lose their function and gradually reach to the death.
It may be associated with the pathogenicity of Aspergillus species that HPAEC and HPAEpiC show high susceptibility for elastolytic enzyme. The injury of vascular endothelial cells results in hemorrhage, and bleeding may be enhanced more by the fibrinogenolytic effects of Asnilase. Also, the destruction of alveolar cells by this elastolytic enzyme may facilitate invasion process of Aspergillus. The relatively weak injury characteristic for bronchial cells does not cause the large lesion, but may cause the allergic disease.

CONCLUSION
It is shown in this report that Asnilase, an elastolytic proteinase produced by A. nidulans presumably contribute to pulmonary aspergillus infection by its cytotoxicity and hydrolytic activity on proteins from biotissue.