Inactivation of α2-Macroglobulin by Activated Human Polymorphonuclear Leukocytes

The proteolytic activity of trypsin releases the dye Remazol Brilliant Blue from its high molecular weight substrate, the skin powder (Hide Powder Azure, Sigma), with an increase in absorbance at 595 nm. Active α2- macroglobulin (80 μg/ml) totally inhibits the proteolytic activity of trypsin (14 μg/ml) by trapping this protease. But after a 20 min incubation of α2-macroglobulin at 37°C with 2 × 106 human polymorphonuclear leukocytes activated by N-formyl-L-methionyl-L-leucyl-L-phenylalanine (10−7 M) and cytochalasin B (10−8 M), 100% of trypsin activity was recovered, indicating a total inactivation of α2-macroglobuHn. Incubation with granulocyte myeloperoxidase also inactivates α2-macroglobulin. Hypochlorous acid, a by-product of myeloperoxidase activity, at a concentration of 10−7 M also inactivates α2-macroglobulin, which indicates that an important cause of α2-macroglobulin inactivation by activated polymorphonuclear leukocytes could be the activity of myeloperoxidase.


Introduction
During activation, the polymorphonuclear leukocytes (PMNL) release numerous potent substances, such as active oxygen species, mediators of inflammation and enzymes. I," One of these enzymes, myeloperoxidase (MPO), is included in the primary granules of the polymorphonuclear neutrophils and produces hypochlorous acid from hydrogen peroxide and chloride anion. This oxidant molecule is able to generate chloramines and to destroy bacterial structures resistant to proteinases. 3,4 These potent substances, released by the activated PMNL, normally benefit the host by destroying foreign organisms, and the role of the proteinases would be to allow PMNL to traverse connective tissue barriers by local destruction strictly controlled by antiproteinases.
Matheson et al. have discovered that 0tl-proteinase inhibitor ((Xl-PI) one of the main plasmatic antiproteinases particularly active against elastase, can be inactivated by MPO in the presence of hydrogen peroxide and chloride anion and that this inactivation results from oxidation into sulphoxide of the two methionyl residues of the reactive site of zl-PI.
So, during excessive activation, the neutrophils will release proteinases and MPO together in high concentrations; the oxidative inactivation of zl-PI by MPO-generated factors will result in increasing proteolytic destruction of tissues by proteinases, particularly by free elastase.
( 1994 Rapid Communications of Oxford Ltd 0t2-Macroglobulin (z2-M) is another important plasma antiproteinase presenting a broad reactivity with serine-, cysteine-, aspartic-and metalloproteinases. Its structure and antiproteinase activity have been studied extensively in the past few years. It possesses four identical subunits linked in pairs by disulphide bonds forming two half molecules which are associated by non-covalent bonds. {x2-M contains a bait area of about 35 different amino acids, each being specifically attacked by a particular proteinase. 11 By proteolysis of a specific peptide bond of the bait region and cleavage of a thioester bond, the binding of a proteinase leads to a change of (x2-M conformation, so that the complex proteinase-zi-M will be recognized quickly by macrophage receptors and rapidly eliminated by the reticulo-endothelial system. 12 This conformational change of 02-m leads to the trapping of the enzyme, which then remains active only against low molecular weight substrates, since by steric inhibition the access to the trapped proteinase is prohibited for high molecular weight substrates. 8,13 It has been demonstrated that 0t2-M is vulnerable to ammonium salts, which cause the transition of (x2-M from the S-form to the F-form that differ in electrophoretic mobility ('slow' for S, and 'fast' for F) and that this transition renders z2-M unable to protect large molecular weight substrates from proteolysis. 4a5 To test the antiproteinase activity of z,.-M, Barrett et al. 4 assayed the proteolytic activity of trypsin against a large molecular weight substrate in the presence of (x2-M SO that the trapping of trypsin by (x2-M reduced or totally inhibited the enzymatic activity. But the pre-treatment of (x2-M by ammonium salts (such as methyl ammonium chloride) impaired the inhibitory power of (x2-M on trypsin.
The purpose of this study was to demonstrate that O2-M is highly sensitive to oxidation and that activated PMNL or some of their products of activation can destroy its antiproteinase activity, as they do for 0q-PI. 6,7 The results are in agreement with those reported previously by Reddy et al. 6 Taken together, these findings confirm Weiss' hypothesis of a possible oxidation of {Xa-M by activated neutrophils with possible dramatic consequences on tissue destruction during acute inflammation.
Preparation and activation of polymorphonuclear leukocytes: Polymorphonuclear leukocytes (PMNL) were isolated from human blood of healthy donors following the technique of Borgeat and Samuelsson. TM After a first centrifugation (25 min at 200 x g), the plasma was discarded and the. 'buffy coat' was resuspended in one volume of saline and a half volume of 6% Dextran T 500 in saline, until erythrocytes settled down. The 'foamy coat' was layered over Ficoll-Paque. The PMNL pellets were collected after centrifugation (100 x g, 20 min at 20C) and briefly exposed to Tris (2.06%)-ammo-nium chloride (0.83%) at pH 7.4, to lyse any remaining erythrocytes. After a last centrifugation ( x 106 PMNL in 0.1 ml PBS and 0.9 ml of either a 10 -3 M NBT or ferricytochrome C solution were incubated for 5 min at 37C. After addition of FMLP and cytochalasin B, the absorbance at 522 or 550 nm was followed and compared with the same assay in the presence of 60 /g of superoxide dismutase, the specific enzyme destroying the superoxide anion. Before the test on the activity of trypsin, {Xa-M was preincubated with PMNL for 20 min at 37C in a final volume of i ml ((xi-M final concentration, 800/g/ml). Control assays were made by incubation of 0t,.-M in the same conditions but with non-activated PMNL.
After centrifugation (600 x g, 5 min at 20C), the supernatants were pipetted and used for trypsin activity measurement. Control experiments of the direct action of activated or non-activated PMNL on trypsin activity were carried out.
To rule out a possible release of a trypsin-like enzyme by activated PMNL, which would interfere with the release of the dye from the substrate, we looked at a possible trypsin-like activity in the supernatant of PMNL by using a low molecular weight chromogenic substrate, Chromozym TRY (R) (carbobenzoxy-t-valyl-t-glycyl-t-arginyl-pnitroanilide). ' Trypsin cleaves this substrate releasing p-nitroaniline, the absorbance of which is determined at 405 nm. The working solution of Chromozym TRY (R) was prepared by dilution of 2 ml of a concentrated solution (1.5mg/ml of dimethylsulphoxide) in 25 ml of Tris-HCl buffer (0.1 M, pH 8.0) added to CaCl,. (0.02 M). In the assay tube, 1 ml of the working solution of Chromozym TRY (R) is mixed with 0.8 ml of Tris-HCl buffer and 0.2 ml of the supernatant from activated PMNL. After 10 min of reaction, the absorbance was measured against the same test without supernatant (control test). A standard curve was established with concentrations of bovine trypsin ranging from 2 to 100 ng/ ml.
Effects of various substances on ot2-macroglobulin activity: MPO was purified from PMNL isolated from 30 of blood from healthy donors, according to the method described previously. 21 The purity of the final preparation was checked by the ratio of absorbance at 430 nm vs. 280 nm and by analytical electrophoresis either on 9% polyacrylamide gel at pH 4.6 or on gradient polyacrylamide gel (7-15%) in the presence of 0.1% sodium dodecyl sulphate and 2-mercapto-ethanol, in 0.1 M Tris-glycine buffer at pH 8.3. 22 To demonstrate a possible role played by MPO in 2-M inactivation, the following experiments were carried out. Forty lal of MPO (1.25 10 -M in phosphate buffer 0.1 M, NaCl 0.2 M), 10 I1 of H20 (10 -4 M) and 50/1 of 2-m (1600 /g/ml in Tris-HCl buffer) were incubated for 10 min at 37C. 100/1 of trypsin (14 /g/ml in Tris-HCl buffer) were then added, the volume of the assay tube was adjusted to 0.5 ml with Tris-HCl buffer and a new incubation was performed for 10 min at 37C. Next, the proteolytic activity of trypsin was assayed as described above. To test that the trypsin activity was not directly inhibited by MPO or H202, the same assays were performed in the absence of 2-m (control assay). The enzymatic activity of MPO produces oxidant chlorinated species, particularly hypochlorous acid (HOCl) which could play a role in 0t,.-M inactivation. So, its direct action on {2-M was tested. In each assay, 50 /1 of {2-M (1600 /g/ml) and 50 /1 of HOC1 (10-4M) were incubated for 5 min at 37C. One hundred lal of trypsin (14/g/ml) were added to the assay tube and the volume adjusted to 0.5 ml with Tris-HCl buffer. After 10 min incubation at 37C, the proteolytic activity of trypsin was tested. Control assays were made without O2-M.
The effects of thiourea, an inhibitor of MPO activity, was assessed by incubating PMNL (20 x 106 cells/ ml) for 5 min at 37C with 10/1 of a thiourea solution (10-4 m). 23 The cells were then activated in the presence of O2-m (see above). After incubation for 20 min at 37C, the supernatants were used to test the remaining inhibiting activity of {2-M on trypsin activity. Control tests were made in the same manner with all the reagents except 2-M.
Assay of proteinase inhibitory activity of macroglobulin: The activity of Ot2-M was determined as its inhibition on the proteolytic activity of trypsin against a large molecular weight substrate, hide powder azure. TM Free trypsin breaks the link between the hide powder and a dye, Remazol Brilliant Blue, releasing it in the supernatant where its absorbance is determined at 595 nm (A595). When complexed by O2-M trypsin is inactive against this substrate.
All the assays were prepared in triplicate and the reagents were dissolved in the test buffer, 0.1 M Tris-HC1 at pH 8.1 added to 0.02 M CaC1,. and 0.1% Brij 35. In each of the triplicate assay tubes, 0.1 ml of trypsin (14/g/ml) was incubated for 10 min at 37C with 0.1 ml of O2-m (800/g/ml) in a total volume of 0.5 ml made up with the test buffer. Water (0.5 ml) was then added to the tube, followed by 0.8 ml of the substrate suspension (12.5 mg of hide powder azure per ml in 0.6 M sucrose, 0.1% Brij 35, 0.03% toluene). After an incubation of 20 min at 37C with continuous shaking, 1 ml of water was added and after centrifugation (2 000xg for 5 min), the A595 of supernatant was determined.
To verify the existence of a linear relationship between increasing concentrations of trypsin and the amount of the dye released from a constant concentration of the substrate, a standard curve of 4 to 20/g of trypsin/ml was determined in the absence of {2-M. In the same manner, to verify the existence of a linear relationship between increasing amounts of O2-m and the inhibition of trypsin activity, a constant concentration of trypsin was incubated with increasing amounts of 2-M from 100 to 1 000/g/ml.
Soybean trypsin inhibitor (SBTI) is an efficacious inhibitor of trypsin that blocks the active centre of the enzyme. 24 Consequently, when preincubated with trypsin (5 min at 20C), it has to prevent the enzyme action on the hide powder azure substrate. SBTI was used in these experiments to prove the validity of the test assay of trypsin enzymatic activity. Methylammonium

Results
Using the technique of Barrett, 4 a linear relationship was found between the amount of trypsin incubated in the presence of a fixed concentration of the substrate (10 mg) and the amount of the released dye, for trypsin concentrations ranging from 2 to 20 g/ml (A595 from 0.05 to 0.32). For further assays, a fixed trypsin concentration of 14 lg/ml was chosen giving a reproducible A595 of 0.22 + 0.02 (n 30). The A595 obtained for this trypsin concentration in the absence of O2-m was taken as 100% of trypsin activity ( Fig. 1, Column 1). The enzyme assay was affected by the trypsin inhibitor SBTI, and the trypsin activity (14 /g/ml) was completely inhibited for an SBTI concentration of 15 /g/ml, which represents an SBTI/trypsin molar ratio of 1.1 (Fig. 1, Column 6). A linear relationship was observed between increasing lag/ml) was observed for a methylammonium chloride concentration of 400 mM (Fig. 1, Column 7). At the same concentration, methylammonium chloride was totally inactive on trypsin alone. Fig. 2 shows the increasing production of superoxide anion by PMNL activated by FMLP and cytochalasin B. The absorbance of formazan blue (produced by reaction of tetrazolium nitroblue with  Using the chromozym TRY (R) technique, proteolytic activity sufficient for interfering with the assay conditions could not be detected. Additionally, by themselves, supernatants of activated or non-activated PMNL were unable to release the dye from the substrate. Moreover they did not decrease the release of the dye from the substrate, which indicates that no trypsin inhibitor was produced (Fig. 3, Columns 2 and 3). Activated PMNL destroyed the antiproteinase activity of 0ta-M (Fig. 3, Column 5) since trypsin activity was completely recovered after treatment of O2-M by activated PMNL. This O2-M inactivation was dependent on the activation of PMNL, for non-activated PMNL did not affect the inhibitory power of 02-M on trypsin (Fig. 3, Column 6). When thiourea was added (at the concentration of 10 -4 M) to the incubation medium of O2-M with activated PMNL, it prevented the inactivation of 02-M, resulting in a complete recovery of 02-M inhibition on trypsin activity (Fig. 3, Column 7).
The treatment of O2-M by MPO in the presence of H202 and C1inactivated 0t2-M since all the activity of trypsin was recovered (Fig. 3, Column 10), while the enzyme was not directly affected by the activity of MPO (Fig. 3, Column 8). Hypochlorous acid, HOCl,   (Table 1). But in our assay conditions HOC1 (10-7M), had little effect on trypsin activity (Fig. 3, Column 9). Hydrogen peroxide, at concentrations between 10-7 and 10 -4 M, did not inactivate trypsin (Table 1).
When O2-M is inactivated, its antigenic activity shows little change. Using the immunoprecipitin analysis technique (immunoelectropherogram), a difference in mobility of the inactive human ot2-M, compared with its native active form (Fig. 4)

Discussion
(x2-Macroglobulin (about 1.1 x 10 M) was inactivated by 2 x 10 PMNL. This inactivation was complete, with a total loss of antiproteinase activity, but needed an activation of the cells, since 2-M treated by non-activated PMNL conserved its antiproteinase activity. The control assays allowed us to exclude the capture of O2-M by the cells or the release of an inhibitor directly acting on trypsin. The activation of PMNL can release proteinases (such as elastase) which, by binding to O2-M could have rendered it inactive towards trypsin. But it seems difficult to admit that 2 x 10 cells would have released sufficient amounts of proteinases to completely complex the amount of 2-M used in these assays.
When PMNL are activated, they undergo the 'respiratory burst', triggering the activity of membrane NADPH oxidase leading to the production of excited forms of oxygen, particularly O2 (superoxide anion) from which H20 can be generated. 1,26 PMNL activation also leads to degranulation with the release of proteolytic and hydrolytic enzymes from granules. MPO is an important enzyme present in the azurophilic granules. It uses H202 and halide anions (CI-, I-etc.) for its enzymatic activity, which generates various strong oxidants, such as HOC1 and chloramines. 27,28 The inactivation of O2-M by excited PMNL can be due to the excited forms of oxygen or the action of the granulocytic enzymes. It is unlikely that O2 is responsible for 0t,.-M inactivation. To attack (x-M, O2 needs to reach this molecule, but, in our assay conditions, its access to O2-M is limited by its instability in aqueous media. Moreover, the effects of O on O2-M were tested using the production of O2 by an acetaldehyde-xanthine oxidase system 29 in buffer with added O2-M. In these conditions, the antiproteinase activity of O2-M remained unaffected. Consistent with the conclusion that O2 was not the primary species involved are the results of a previous study by Reddy et al. 16  However, it appears from our observations that this inactivation of O2-M by HOC1 is nonspecific, since trypsin itself can be totally destroyed by HOCl, but only at higher concentrations (10-5M) of this molecule. In the same manner, H202 also has a destroying effect on 0t,.-M, but hydrogen peroxide at concentrations of 10-2M is needed for this destruction. It is probable that inactivation of 2-M occurs by the destruction of the S-S bonds between the four subunits of the protein, leading to their dissociation.
The limited reduction of O2-M produces monomers (180 kDa) which reassociate after proteinase treatment but do not prevent proteinases from cleaving hide powder azure. TM  In our experimental conditions, 2-M (10-9M) was completely inactivated by a 20 min incubation in the presence of excited PMNL, but 2 106 cells per assay were used. Moreover, these cells remained in close contact with {2-M for a long time. While PMNL can exert strong destructive effects on host tissues, 5,2, our experimental conditions are far from being comparable to a normal in vivo situation, where plasma protectors will counteract any PMNL destroying activity on 02-m. Therefore, the extent to which in vivo 2-m inactivation by triggered PMNL could occur and contribute to tissue destruction through imbalance between proteases and antiproteases is still obscure and requires further studies. It is worth remembering, however, that we have measured important plasma concentrations of MPO in inflammation states associated with ARDS, sepsis or acute pancreatitis. 4 On the other hand, decreased plasma and serum concentrations of ,.-M have been reported in ARDS and septic patients. Furthermore, these decreased plasma or serum concentrations of O2-M are measured by routine immunological or immunoelectrophoresis techniques that are unable to easily differentiate the active from the inactive form of O2-M so that the plasma or serum concentrations of the protein measured by these techniques do not.reflect the capacity of 2-M to inhibit proteases. Taken together, these findings indicate that in vivo inactivation of 2-M cannot be ruled out, but as suggested above the occurrence of this phenomenon and its possible involvement in tissue destruction by active protease in inflammatory diseases remain to be firmly demonstrated.