European Journal of Pharmacology - Environmental Toxicology and Pharmacology Section, 228 (1992) 213-218
© 1992 Elsevier Science Publishers B.V. All rights reserved 0926-6917/92/$05.00
Activation and damage of cultured airway epithelial cells by human elastase and cathepsin G M a r i e - A n n e N a h o r i , P a t r i c i a R e n e s t o , B. B o r i s V a r g a f t i g a n d M i c h e l C h i g n a r d Unitd de Pharmacologic Cellulaire, Unitd Associge IP / INSERM no. 285, Institut Pasteur, Paris, France
Received 13 July 1992, accepted 4 August 1992
Accumulation of polymorphonuclear neutrophils (PMN) and epithelium damage have often been described during airway inflammation. We studied the effects of two PMN-derived proteinases, namely elastase and cathepsin G, on guinea-pig tracheal epithelial cells in culture. Both proteinases activated tracheal epithelial cells in terms of prostaglandin (PG) E 2 production. A concentration- and time-dependent effect was observed with 10 /zg/ml and 6 h as the optimal conditions for both enzymes. Optical microscopic studies confirmed an effect on tracheal epithelial cells as intercellular gaps were observed upon incubation of the monolayers with proteinases. A small cytotoxic effect was observed after 1 h incubation but remained stable up to 6 h. This cytotoxic effect, more pronounced with elastase than with cathepsin G, was dissociated from PGE 2 formation. Tracheal epithelial cells; Leukocyte elastase (human); Cathepsin G
Injury to airway epithelial cells is a predominant feature in a variety of lung diseases. Another characteristic of some of these disorders is the presence of polymorphonuclear neutrophils (PMN) (Ratliff et al., 1971; Robin, 1979). Indeed, an increase in the P M N count has been observed in the respiratory tract of cigarette smokers, patients with chronic bronchitis and cystic fibrosis. In fact, P M N are thought to play a major role in generating the pathological manifestations of airway inflammation, as shown in animal models of inflammatory e d e m a (Heflin and Brigham, 1981; Shasby et al., 1982). Many investigations have suggested that stimulated P M N may also be an important cause of lung damage in some cases of adult respiratory distress syndrome (Jacob et al., 1980; Tate and Repine, 1983; Brigham and Meyrick, 1984; Boxer et al., 1990). Experimental in vitro models have been developed to investigate the interactions between P M N and epithelial cells. Thus, killing by human P M N of rat alveolar epithelial cells in primary culture has been reported by Simon et al. (1986). In contrast, Ayars et al. (1984) failed to demonstrate such a toxicity but found and characterized a P M N - m e d i a t e d cell detach-
Correspondence to: M.-A. Nahori, Unit~ de Pharmacologie Cellulaire, Unit6 Associ~e Institut Pasteur/INSERM no. 285, Institut Pasteur, 25 rue du Dr Roux, 75724 Paris cedex 15, France. Tel. (33-1) 18.104.22.168; Fax (33-1) 45.68.87.03.
ment from the culture dish. Whatever the intensity of the cell injury, proteinase involvement has been demonstrated in both cases, and recently confirmed by Dunn et al. (1990). Two main proteinases are released from activated PMN, namely elastase and cathepsin G (Weissmann et al., 1972; Ohlsson and Olsson, 1977). Nonetheless, using elastase, Simon et al. (1986) were unable to mimick the effects of P M N on epithelial cells. In contrast, using the same cell type, Rochat et al. (1988) showed that purified cathepsin G increased epithelial permeability and, at higher concentrations, became cytotoxic. Our present aim was to study the potential effect of elastase on guinea-pig tracheal epithelial ceils in culture and to compare it with that of cathepsin G. Since upon activation these cells form mainly prostaglandin (PG) E 2 (Nahori et al., 1991), this p a r a m e t e r was chosen as an index of activation. The optical morphological aspect and the toxicity were also studied. Our results show that both proteinases induce cell activation, a p h e n o m e n o n related to their enzymatic activity.
2. Materials and methods 2.1. M a t e r i a l s
Male Hartley guinea pigs (350-400 g) were obtained from Charles River (Elboeuf, France). Sodium pentobarbitone was from Sanofi Sant6 Animale (Montpellier, France). N-succinyl-Ala-Ala-Pro-Phe-p-nitroani-
214 lide, N-succinyl-Ala-Ala-Ala-p-nitroanilide, phenylmethylsulfonyl fluoride (PMSF), protease (type XIV), hydrocortisone, insulin, epidermal growth factor (EGF), transferrin, cholera toxin, bovine hypothalamic extracts and aprotinin were from Sigma Chemical Company (St Louis, MO, USA). Glutamine, Medium 199 (with Earle salt and red phenol) and foetal calf serum were obtained from Institut Jacques Boy (Reims, France). Antibiotic-antimycotic solution and CNBr-activated Sepharose 4B were from Gibco (Grand Island, NY, USA). Multiwell culture plates (96 wells) of serie Primaria from Falcon were from Becton Dickinson (Oxnard, USA) and CM-Trisacryl M from IBF Biotechnics, (Villeneuve-la-Garenne, France). [125I]tyrosyl methyl ester PGE2 and antibodies against P G E 2 were from the U R I A (Institut P a s t e u r / I N S E R M 207, Prof. F. Dray, Paris, France). Polyethyleneglycol 6000 was from Merck (Darmstadt, Germany). Human plasmatic albumin solution (20%) was from Bio-Transfusion (Lille, France). Chromate sodium 51 (51Cr) was from Amersham (Buckinghamshire, UK).
2.2. Cell culture Cells were cultured according to Wu et al. (1985) as modified by Nahori et al. (1991). Briefly, guinea pigs were anaesthetized with sodium pentobarbitone (30 mg/kg, i.p.), tracheas were removed aseptically, filled with medium 199 containing antibiotics, glutamine and 0.1% protease type XIV and clamped at both extremities. They were then submerged in medium 199 containing glutamine (2 mM), antibiotic-antimycotic solution (1%) (medium A) and kept at 4°C overnight. One end of the tracheas was connected to a needle attached to a syringe filled with 10 ml of medium A and cells were then flushed from the trachea. The resulting suspension was centrifuged for 10 rain at 150 x g and the pellet was resuspended in medium A supplemented with foetal calf serum (10%). This suspension was seeded in culture plates. After 2 days of culture the medium was changed for a fresh one in which foetal calf serum was replaced by hydrocortisone (0.1 lzg/ml), transferrin ( 5 / z g / m l ) , insulin ( 1 / z g / m l ) , E G F (10 ng/ml), bovine hypothalamic extracts (0.2%) and cholera toxin (10 ng/ml). Cultures were maintained in an incubator at 37°C under 95% air, 5% CO 2 and the medium was changed every 2 days and before the beginning of each experiment. Unless otherwise stated, tracheal epithelial cells were used at day 5, a time at which they are at confluence, each well containing around 105 cells (evaluated by their protein content).
2.3. Purification of cathepsin G and elastase Cathepsin G and elastase were purified from human PMN according to the method of Baugh and Travis
(1976) modified by Martodam et al. (1979). Briefly, cells were washed three times and submitted to an hypotonic shock to lyse erythrocytes. PMN were then broken with a Potter-Elvehjem glass homogenizer in order to extract granules containing proteinases. After centrifugation (1000 × g, 15 min, 4°C), the supernatant devoid of nuclei and unbroken cells was collected. The granule pellet obtained after ultracentrifugation (30,000 × g, 30 min, 4°C) was broken with an Ultraturax and centrifuged (1000 × g, 15 min, 4°C). This procedure was repeated twice and pooled supernatants were centrifuged (30,000 × g, 35 min, 4°C). The resulting supernatant was applied onto an affinity chromatography column (aprotinin-Sepharose 4B) in order to separate cathepsin G and elastase from other granule constituents. The mixture of both proteinases was then applied onto an ion-exchange chromatography column (CM-Trisacryl M). The elution was carried out with a sodium gradient and its profile was followed by a spectrophotometric measurement at 280 nm. Fractions of the first peak, corresponding to elastase, and of the second peak, corresponding to cathepsin G, were pooled separately, and after concentration and dialysis, aliquots were stored at -80°C. Enzymatic activities of purified cathepsin G and elastase were determined by following the hydrolysis of their specific synthetic substrates, N-succinyl-Ala-Ala-Pro-Phe-pnitroanilide and N-succinyl-Ala-Ala-Ala-p-nitroanilide, respectively, in the presence of increasing amounts of titrated al-antitrypsin. The linear regression curve obtained allowed us to extrapolate the active site concentration of both proteinases. By spectrophotometric studies, we have also verified that the cathepsin G batch was devoid of elastase activity as well as that there was no cathepsin G in the elastase batch.
2. 4. Preparation of PMSF-treated proteinases Catalytic sites were blocked by incubating cathepsin G and elastase for 60 rain at room temperature in the presence of 1.25 mM PMSF. Mixtures were then dialysed using a microconcentrator in order to remove the inhibitor. PMSF-treated cathepsin G and elastase indeed lost their proteolytic activity, as assessed by measuring the hydrolysis of their specific synthetic substrates. Protein concentration was determined by the method of Lowry et al. (1951).
2.5. Cell stimulation To investigate the proteinase effect, 100 ~1 of medium 199 (with glutamine, antibiotic and 0.1% human serum albumin) containing different concentrations of cathepsin G or elastase were added to the cells in culture. At the end of the incubation period (1, 3
215 and 6 h) the medium was collected and kept at - 2 0 ° C for P G E z measurements.
2.6. Cytotoxicity assay The possible cytotoxic effects of cathepsin G and elastase were tested for different incubation times (1, 3, 6 and 24 h). Tracheal epithelial cells were labelled by overnight incubation with 51Cr (2 /xCi/ml) in the complete growth medium (100/xl/well). Cultures were then washed three times with medium A, and incubated without 51Cr in the presence or absence of proteinases. At the end of the incubation period, supernatants were collected and the radioactivity was measured (cpmex o) using a g a m m a counter (LKB, Wallac, Stockholm, Sweden). Basal releases were determined from cells incubated with the medium alone (Cpmcontrol), and maximal releases were achieved by adding 100 /xl of 1% Triton X-100 on intact cells (Cpmmax). The percentage of 5tCr released in supernatants was calculated as follows: cpm e x p
cpm c o n t r o l
cpm m a x
cpm c o n t r o l
2. 7. Radioimmunoassay of PGE 2 Aliquots (100/xl) of culture medium were collected and the products were assayed directly without extraction. For performing the radioimmunoassay, samples were thawed and incubated overnight at 4°C with iodine-labelled P G E 2 (100/zl) and antiserum (100/zl) in phosphate buffer (10 mM, p H 7.4) containing bovine gamma-globulins (0.3% w / v ) . The next day, bound and free fractions were separated by adding 500 /xl of polyethyleneglycol 6000 at 4°C for 10 min, then the samples were centrifuged (2500 x g for 10 min at 4°C). The radioactivity of the pellets corresponding to the bound fraction was counted for 1 min with a Kontron analytical g a m m a counter. The amounts of P G E 2 were determined by projection on a standard curve and expressed in n g / m l .
analysis of variance ( A N O V A test) and P < 0.05 was considered significant. When the variance analysis showed a significant difference, individual groups were compared using the LSD test (least significant difference).
3.1. Activation of epithelial cells P G E 2, the main arachidonate metabolite produced by stimulated tracheal epithelial cells (Nahori et al., 1991) was measured as an index of activation. As shown in fig. 1, elastase activated tracheal epithelial cells in a concentration-dependent manner. As observed in fig. 2, P G E 2 release induced by elastase (1 /~g/ml) was significant after 3 h of incubation, increased up to 6 h and then plateaued up to 24 h (not shown). A spontaneous formation of P G E 2 from untreated cells was observed at 6 h and became significant at 24 h. At this time point there was no significant difference between control tracheal epithelial cells and elastase-treated tracheal epithelial cells (not shown). Cathepsin G also induced P G E 2 formation at similar optimal conditions as elastase, i.e., 1 0 / z g / m l incubated at 37°C for 6 h (data not shown). Under these conditions the production of P G E 2 triggered by cathepsin G was not significantly different from that induced by elastase (fig. 3). When the catalytic sites of elastase or cathepsin G were blocked by PMSF, both proteinases lost their capacity to activate tracheal ep-
2.8. Morphological studies Tracheal epithelial cell monolayers were observed without any treatment with an optical microscope (Nikon) before and after cathepsin G or elastase addition.
2.9. Statistical analysis For each trachea the different experimental conditions were performed at least in triplicate and the presented data were obtained from at least six different tracheas. Groups were compared using the one-way
I0 HLE (btg/ml)
Fig. I. Effect of elastase on PGE 2 production by tracheal epithelial cells. Tracheal epithelial cells were incubated for 6 h with increasing concentrations of the proteinase and PGE 2 formation was measured by RIA from supernatants. * indicates a significant difference between the proteinase effect and the control, * P < 0.05, ** P < 0.01. Values are means± S.E.M. (n > 6).