Suppression of the neutral protease activity of macrophages treated with asbestos in vitro.

Macrophages are often conspicuous in asbestos-induced inflammatory lesions. Chrysotile type B elicits macrophages into the peritoneal cavity of mice which produce high levels of the neutral protease, plasminogen activator; in vitro addition of these same fibers to mouse peritoneal macrophages stimulates enzyme production. It is reported here that, for endotoxin-elicited mouse peritoneal macrophages fed chrysotile type B in vitro, the increased plasminogen activator activity is suppressed by low concentrations of anti-inflammatory steroids. Other active drugs include colchicine and vinblastine. These studies are considered important, as they suggest an approach to controlling the levels of a potentially deleterious enzyme system (PA-plasmin) from macrophages treated with asbestos fibers.


Introduction
Asbestosis is characterized by inflammation and fibrosis surrounding the respiratory bronchioles and alveoli and leads to permanent impairment of lung function (1). Macrophages are found in normal alveoli and are the prominent cell type found in chronic inflammatory lesions, including those induced by asbestos (2). Macrophages are able to release in vitro a number of molecules which are probably important in the initiation and maintenance of an inflammatory reaction, for example, neutral proteases, complement components, endogenous pyrogens, oxygen metabolites, prostaglandins and analogs, proliferation factors for neighboring cells, etc. (3).
For monocyte macrophages in vitro, the activity of the neutral protease, plasminogen activator (PA), shows positive correlations with stimuli [for example, asbestos (4), endotoxin (5), lymphokines (6), and phorbol esters (7,8)] and suppressive drugs [for example, glucocorticoids (9)] associated with an inflammatory response. A role for macrophage PA activity in cell migration from blood to tissue and in the processes of tissue destruction and activation of several inflammatory enzyme cascades has previously been *Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, 10021 U.S.A. Present address: Department of Medicine, University of Melbourne, The Royal Melbourne Hospital, Victoria, 3050, Australia. proposed (9). We have been able to show that IARC chrysotile type B elicits macrophages into the peritoneal cavity of mice which produce high levels of PA (4,10); in vitro addition of these fibers to mouse peritoneal macrophages stimulates enzyme production.
The present studies indicate that the in vitro stimulation of macrophage PA activity can be suppressed by anti-inflammatory glucocorticoids and, under certain culture conditions, by colchicine and vinblastine.

Fibrinolysis Measurement
The peritoneal cells (2 x 106) were plated on 1II-fibrin-coated wells (Linbro Disposo Trays, FB 16-24TC) (9) for 24 h in DB supplemented with 5% heat-inactivated (560C, 30 min) fetal bovine serum (FBS). At this stage, soybean trypsin inhibitor (STI) (100 ,ug/mL) was added to block any fibrinolysis. The cells were washed twice with PBS and placed again in DB + 5% FBS + STI. The particulate preparations were resuspended by passage through a syringe and 26G needle and carefully distributed in the culture medium. The particles were generally added for 2 hr, after which time >95% of the cells contained 20-50 latex particles. The macrophages were then washed three times with PBS and placed in DB + 5% acidtreated FBS (ATFBS) for assay. The acid treatment removes interfering protease inhibitors (9). Fibrinolysis was estimated' by withdrawing aliquots of medium and counting the radioactivity in a Packard Autogamma scintillation spectrometer. The plasminogen dependence of the fibrinolytic activity can be shown by assaying in the presence of ATFBS from which the plasminogen has been removed (9).

Measurement of Secreted Plasminogen
Activator Peritoneal cells (2 x 106) were plated in DB + 5% FBS on regular linbro trays. At 24 hr after plating, the cells were washed twice with PBS and placed again in DB + 5% FBS but also in the presence of the particles for an additional 24 hr. After this period, the cultures were washed twice with PBS and placed in DB containing 0.05% lactalbumin hydrolyzate (LH) (9). After 24 hr, the serum-free condi-tioned media (CM) were collected; also, the cells were washed twice with PBS, lysed by the addition of Triton X-100 (0.2% v/v in water) and scraped from the surface of the dish with a plastic policeman for protein determination.
Serum-free CM were kept frozen at -20°C until assay for PA activity. For this purpose, 1"I-fibrincoated wells of Linbro Disposo Trays were used, and the assay mixture contained 2,ug human plasminogen (purified from outdated human plasma) and 10 M.*L CM in 250 ML 0.1 M Tris chloride, pH 8.1 (9).
All experiments included appropriate controls for all reagents and media. No plasminogen-independent fibrinolysis could be detected when plasminogen was not included in the assay mixture. One unit of PA is defined as the amount that stimulates the release of 10% of the initial radioactivity in 4 hr (9).

Lysozyme
Lysozyme in serum-free CM was assayed by measuring the initial rate of lysis of a suspension of Micrococcus luteus, as described by Gordon et al. (11). Human lysozyme diluted in DB + 0.05 % LH was used as standard.

Lysosomal Enzymes
N-Acetyl-f-D-glucosaminidase and p-galactosidase were assayed as described by Bowers et al. (12). Protein Protein was measured by the method of Lowry et al. (13) with bovine serum albumin as standard.

Results
Enhanced Macrophage PA Activity after Asbestos Exposure in Vitro Phagocytosis of digestible and nondigestible materials enhances the PA activity of mouse peritoneal macrophages (4,5). Figure 1 demonstrates that asbestos fibers can increase the PA activity of peritoneal macrophages obtained from endotoxin-injected mice and also PBS-injected mice as a control population. Comparable results were obtained whether the asbestos addition was allowed for 2 hr or for 24 hr prior to the PA measurement. For asbestos doses less than 100 1.g, there did not appear to be any gross alterations in cell morphology, but with 100,g fiber there was some cell rounding and detachment after 3-4 days. The control, unstimulated peritoneal macrophages released PA activity in the presence of the asbestos, but there was a longer lag period than for the endotoxin-induced cells. The fibrinolytic activity was predominantly (> 95%) plasminogen-dependent, and hence PA activity was being monitored.

Inhibition of in Vitro-Stimulated Fibrinolytic Activity with Drugs
If the assumption is correct that the generation of excessive PA activity around macrophages interacting with asbestos fibers would be deleterious, then it would be desirable to find drugs which might suppress this local proteolysis. In Table 1, a comparison is made of three steroids for their effect on the asbestos-stimulated plasminogen-dependent fibrinolytic activity. The glucocorticoids, dexamethasone and prednisolone, are effective whereas progesterone is not. The data using latex feeding to cells are included for comparison. This inhibition of the increased fibrinolysis persists for at least 72 hr, the maximum assay time. Table 2 lists the relative efficacies of different steroids. Also included for comparison are colchicine and vinblastine, drugs implicated in the control of secretory processes. Certain glucocorticoids are again effective. It is interesting that the activity profile for the steroids is similar to that found for their suppression of the PA activity of thioglycollate-in-  (Table 2) but were quite effective if harvest fluid is assayed for the PA activity from asbestos-treated cells (see below). It can be noted that the PA activity of the endotoxin-primed cells themselves was also inhibited. From Table 1 and 2, it can be observed that Progesterone 39 a Peritoneal cells were cultured for 24 hr on '25I fibrin-coated Linbro wells in the usual manner, before the asbestos or latex feeding. After washing, the solid particles were added in DB + 5% FBS and 100 pg/mL STI for 2 hr at 370C. After washing to remove excess particles, DB + 5% ATFBS was added in presence or absence of drug. Samples assayed as in Fig. 1.   there is a comparable effect of the drugs for cells fed latex particles. The most potent glucocorticoid tested was dexamethasone and, as can be seen from Table 3, it is effective even at doses as low as 10-9 M. This low dose was also found to be effective on thioglycollate-induced macrophages (9).

Enzyme Secretion after Asbestos Exposure in Vitro
The above experiments were all carried out with cells plated directly on the fibrin substrate. Macrophages obtained from endotoxin-injected mice can also secrete elevated PA levels after exposure to asbestos (or latex) in vitro (Table 4); the increase is predominantly extracellular (data not shown), indicating that enzyme synthesis and secretion were tightly coupled (9). There was no effect of asbestos (or latex) on the secretion either of lysozyme or of two lysosomal enzymes, N-acetyl-f-D-glucosamindase and f3- galactosidase (Table 4). Further evidence in support of the concept of differential secretion control for PA and for lysozyme and lysosomal enzymes is the fact that dexamethasone and colchicine, drugs which have a dramatic effect on the extracellular PA levels, fail to alter the levels of lysozyme and lysosomal enzymes from macrophages when they are treated with asbestos or latex particles.

Discussion
Previous studies (4,10). have indicated that intraperitoneal injection of IARC chrysotile type B into mice induces an inflammatory exudate and that macrophages cultured from such exudates have elevated levels of PA. It was also shown that peritoneal macrophages when treated with the same asbestos fibers in vitro are stimulated to produce more PA activity. The possible significance of macrophage PA activity in inflammation has been outlined (9,10). It has been shown above that as little as 0.1 ,ug of the asbestos fibers can raise the PA activity of endotoxin-primed mouse peritoneal macrophages. It is tempting to speculate that, since the in vitro responses resemble those found with latex particles, it is phagocytosis of the asbestos fibers which results in the heightened protease activity. However, until a quantitative analysis of fiber uptake is made, such a conclusion is unwarranted, particularly as other agents, such as phorbol esters (7,8), concanavalin A (7) lymphokines (6) and colonystimulating factors (14), which are presumably not acting via a phagocytic pathway, can all stimulate macrophage PA activity.
The enhanced PA activity is suppressed by antiinflammatory glucocorticoids. It might be profitable if more were known about how glucocorticoids are working in this context; perhaps these drugs are inducing a "macrocortin-like" molecule which has been proposed to account for the manner in which they suppress cellular prostaglandin production (8,15). It is again worth noting that, for macrophages, lysozyme and two lysosomal enzymes activities appear to be under different regulatory control to the PA activity. Studies of carcinogenesis by various forms of asbestos, with and without hydrocarbon carcinogens, have shown that chrysotile can augment considerably the number of malignant tumors produced by benzo[a]pyrene (16,17): these studies and others have suggested that asbestos fibers can function as cocarcinogens (promoters). Interestingly, we have shown that the tumor promoter in skin, namely, 12-O-tetradecanoyl-phorbol-13-acetate (TPA), and asbestos fibers, elicit similar acute inflammatory responses in rabbit skin (18) and, as mentioned above, can both increase the PA activity of macrophages. We have suggested that vascular alterations due to an inflammatory reaction to both asbestos fibers and to TPA ought to be considered in any analysis of the development of neoplasms resulting from their action (8,18).
It is a pleasure to recognize the excellent technical assistance of B. Bunkin. I would also like to acknowledge Dr. E.
Reich, The Rockefeller University, in whose laboratory the initial observations were made. This work was supported by the National Institutes of Health Grant, ES-02014-01, the American Cancer Society grant and by a grant from the National Health and Medical Research Council of Australia. The author is a Senior Research Fellow of the National Health and Medical Research Council of Australia.