Allen B. Cohen

Studies on the Pathogenesis of the Adult Respiratory Distress Syndrome

Bronchoalveolar lavage (BAL) fluid was obtained from 24 sequentially studied patients with adult respiratory distress syndrome (ARDS) for assessment of potential activating and mediating factors. Proteolytic activity of the fluids was observed by measuring cleavage of radiolabeled proteins of the contact (Hageman factor) and complement systems. Proteolytic activity was observed in 17 of 24 (71%) patients with ARDS, and BAL fluid of the 7 ARDS patients without demonstrable, active, enzyme exhibited inhibitory activity for the proteolytic activity. The enzymes cleaved Hageman factor, prekallikrein, plasminogen, high molecular weight kininogen, C4, C3, C5, and Factor B of the complement system. Cleavage of the contact system proteins producted fragments similar or identical in size to the fragments observed during activation of these molecules, although continued incubation invariably reduced the protein to small peptide fragments. None of 7 normal individuals, and 29 of 99 patients (29%) with other forms of pulmonary disease contained measurable enzymes. The proteolytic activity in BAL fluid of ARDS patients was blocked by diisopropylphosphofluoridate (0.1 mM), Trasylol, soybean trypsin inhibitor, and normal plasma, or plasma deficient in inhibition of the first component of complement. Alpha(1)-proteinase inhibitor (alpha1-PI)-deficient plasma failed to inhibit the proteolytic activity and addition of alpha1-PI to the deficient plasma reconstituted the inhibition. MUCH OF THE PROTEOLYTIC ACTIVITY OF THE BAL FLUID FROM ARDS PATIENTS WAS IDENTIFIED AS NEUTROPHIL ELASTASE: the fluids cleaved elastin and synthetic peptide substrate of neutrophil elastase, neutrophil elastase antigen was present in the BAL fluids as determined immunologically using antineutrophil elastase, alpha1-PI was the major inhibitor in plasma, and the enzyme was inhibited by diisopropylphosphofluoridate but not chelation. In addition, purified neutrophil elastase produced cleavage fragments of proteins of the contact system similar to those of the BAL fluids. In each of the seven BAL fluids of ARDS patients that did not reveal active elastase, alpha1-PI was present in active form (as determined by (125)I-trypsin binding). In 9 of the 17 patients with active elastase in the BAL fluid, alpha1-PI antigen was present in the fluid, but was inactive (no binding of (125)I-trypsin). Immunoelectrophoretic analysis of elastase and alpha1-PI throughout proteins in these BAL fluids revealed the presence of both elastase and alpha1-PI that migrated with the same R(f), suggesting the presence of an enzyme-inhibitor complex. Free, inactive alpha1-PI was also observed in these fluids. The data reveal that in BAL fluids from all 24 patients with ARDS, leukocytic elastase and/or alpha1-PI exist. A complex of elastase and alpha1-PI was observed in BAL fluids, and in some cases where active enzyme and alpha1-PI coexisted, free, but inactive alpha1-PI was present.

Human plasma kallikrein releases neutrophil elastase during blood coagulation.

Elastase is released from human neutrophils during the early events of blood coagulation. Human plasma kallikrein has been shown to stimulate neutrophil chemotaxis, aggregation, and oxygen consumption. Therefore, the ability of kallikrein to release neutrophil elastase was investigated. Neutrophils were isolated by dextran sedimentation, and elastase release was measured by both an enzyme-linked immunosorbent assay, and an enzymatic assay using t-butoxy-carbonyl-Ala-Ala-Pro-Val-amino methyl coumarin as the substrate. Kallikrein, 0.1-1.0 U/ml, (0.045-0.45 microM), was incubated with neutrophils that were preincubated with cytochalasin B (5 micrograms/ml). The release of elastase was found to be proportional to the kallikrein concentration. Kallikrein released a maximum of 34% of the total elastase content, as measured by solubilizing the neutrophils in the nonionic detergent Triton X-100. A series of experiments was carried out to determine if kallikrein was a major enzyme involved in neutrophil elastase release during blood coagulation. When 10 million neutrophils were incubated in 1 ml of normal plasma in the presence of 30 mM CaCl2 for 90 min, 2.75 micrograms of elastase was released. In contrast, neutrophils incubated in prekallikrein-deficient or Factor XII-deficient plasma released less than half of the elastase, as compared with normal plasma. The addition of purified prekallikrein to prekallikrein-deficient plasma restored neutrophil elastase release to normal levels. Moreover, release of elastase was enhanced in plasma deficient in C1-inhibitor, the major plasma inhibitor of kallikrein. This release was not dependent upon further steps in the coagulation pathway, or on C5a, since levels of elastase, released in Factor XI- or C5-deficient plasma, were similar to that in normal plasma, and an antibody to C5 failed to inhibit elastase release. These data suggest that kallikrein may be a major enzyme responsible for the release of elastase during blood coagulation.

Increased interleukin-8 concentrations in the pulmonary edema fluid of patients with acute respiratory distress syndrome from sepsis

OBJECTIVE To test the hypothesis that significantly higher concentrations of interleukin-8 (IL-8) are found in the pulmonary edema fluid and plasma of patients with a septic vs. a nonseptic etiology of acute respiratory distress syndrome (ARDS). DESIGN Prospective measurement of IL-8 concentrations in previously collected edema fluid and plasma. SETTING Adult intensive care units at a university medical center. PATIENTS There were 27 patients with ARDS (16 patients with a septic etiology and nine patients with a nonseptic etiology) plus eight control patients with hydrostatic pulmonary edema. MEASUREMENTS AND MAIN RESULTS IL-8 was present in the pulmonary edema fluid of all patients with ARDS, but the median IL-8 concentration was higher in the edema fluid of patients with ARDS associated with sepsis (84.2 ng/mL, n = 16) compared with the ARDS patients without sepsis (14.8 ng/mL, n = 11) (p < .05). In patients with cardiogenic edema, IL-8 concentration (5.0 ng/mL,n = 8, p < .05) was significantly lower than those values in patients with ARDS. Median plasma concentration of IL-8 was increased in septic individuals (1.3 ng/mL), but these concentrations were not significantly higher than in patients with a nonseptic etiology of ARDS (0.35 ng/mL) (p = .14) or those patients with cardiac failure (0.21 ng/mL). CONCLUSIONS The high concentrations of IL-8 in pulmonary edema fluid, coupled with the relatively low concentrations of IL-8 in the plasma, suggest that the lung was the primary source of IL-8 in the patients with ARDS. The markedly increased concentrations of IL-8 in the pulmonary edema fluid of patients with ARDS from sepsis suggests that this group of patients may be particularly suitable for potential trials directed at inhibiting the activity of this important chemokine.

Inactivation of Factor XIa by Plasma Protease Inhibitors: PREDOMINANT ROLE OF α1-PROTEASE INHIBITOR AND PROTECTIVE EFFECT OF HIGH MOLECULAR WEIGHT KININOGEN

Factor XIa is a plasma protease that, by activating Factor IX, plays an important role in the early phase of the intrinsic pathway of blood coagulation. Four plasma protease inhibitors, alpha(1)-protease inhibitor, antithrombin III, C1-inhibitor, and alpha(2)-plasmin inhibitor, have been reported to inactivate human Factor XIa, but their quantitative contribution to the inactivation of Factor XIa in plasma has not been fully assessed. Using purified systems, we observed that the second-order rate constants for the reaction of Factor XIa with alpha(1)-protease inhibitor, antithrombin III, and CI-inhibitor were 4.08, 10, and 14.6 M(-1) min(-1) x 10(3), respectively. The pseudo-first-order rate constants, at plasma concentration of the inhibitors, were 1.86 x 10(-1), 4.68 x 10(-2), and 2.4 x 10(-2) min(-1), respectively. These kinetic data predict that alpha(1)-protease inhibitor should account for 68%, antithrombin III for 16%, and C1-inhibitor and the equipotent alpha(2)-plasmin inhibitor each for 8% of the total inhibitory activity of plasma against Factor XIa. The rate of inactivation of Factor XIa in various plasma samples specifically deficient in inhibitors was consistent with these predictions. Factor XI, the zymogen form of Factor XIa, circulates in plasma associated with the contact system cofactor, high molecular weight kininogen (HMW kininogen). Kinetic analysis indicated the existence of a reversible bimolecular Factor XIa-HMW kininogen complex with a dissociation constant (K(d)) = 0.17 muM. The light chain derived from HMW kininogen decreased the inactivation rate of Factor XIa by C1-inhibitor with a K(d) of 0.08 muM for a complex of Factor XIa and the light chain derived from HMW kininogen. The protective effect of HMW kininogen was confirmed by the finding that the inactivation rate of Factor XIa in kininogen-deficient plasma was increased over normal plasma. The present study confirms that alpha(1)-protease inhibitor is the major inhibitor of Factor XIa in plasma, and that the formation of a reversible complex between Factor XIa and HMW kininogen decreases the rate of inactivation of the enzyme by its inhibitors.

Biological and kinetic characterization of recombinant human macrophage inflammatory peptides 2 alpha and beta and comparison with the neutrophil activating peptide 2 and interleukin 8

We examined the biological and kinetic characteristics of two new members of the intercrine family of cytokines. Human macrophage inflammatory peptides 2 alpha and beta (huMIP-2 alpha and beta) were compared to human interleukin 8 (huIL-8), neutrophil activating peptide 2 (huNAP-2), and N-formyl-L-methionyl-L-phenylalanine (fMLP). The huMIP-2 peptides were the least potent cytokine tested in triggering neutrophil degranulation. They were also less potent neutrophil chemotaxins than fMLP or huIL-8. However, they were more effective than NAP-2 in stimulating chemotaxis of neutrophils. The binding studies showed that huMIP-2 peptides could interact with specific receptors on human blood neutrophils. Moreover, huMIP-2 peptides competed for up to 60% of the huIL-8 binding sites on neutrophils whereas huIL-8 competed for almost 100% of either of the huMIP-2 peptide binding sites. These data suggest the huMIP-2 peptides have little or no affinity for 40% of the huIL-8 receptors. In addition, detectable amounts of mRNA for huMIP-2 alpha were found in samples from human alveolar macrophages stimulated with Staphylococcus aureus, toxic shock syndrome toxin-1 (TSST), but not in samples stimulated with S. aureus enterotoxin A (SEA) or Escherichia coli endotoxin (lipopolysaccharide = LPS). In conclusion, huMIP-2 alpha and beta are weak neutrophil stimulating agents, which may increase inflammation in diseases such as toxic shock syndrome.

A Unique Elastase in Human Blood Platelets

Previous investigations suggested that elastolytic activity found in platelets could be due to contamination by neutrophil elastase. In the present study, the lysate of blood platelets free of detectable neutrophils was examined for elastase-like activity using tertiary-butyloxycarbonyl (tBOC)-ala-ala-pro-ala-aminomethyl coumarin (I), tBOC-ala-ala-pro-val-aminomethyl coumarin (II), and succinyl-tri-ala-rho-nitroanilide (SAPNA), and for elastolytic activity using 3H-labeled dog and human lung elastins. The platelet lysate degraded I at a higher rate than II, while the reverse was true of neutrophil elastase. The rate of degradation of I, II, and SAPNA by the lysate increased with reaction time up to 20 min. The rate of I, II, and SAPNA degradation by the lysate was decreased by the presence of 0.5 M NaCl, whereas NaCl greatly potentiated their degradation by neutrophil elastase. Plasma alpha 2-macroglobulin inhibited elastolysis by the platelet lysate, whereas plasma alpha 1-antitrypsin did not. The lysate activity was inhibited by diisopropyl fluorophosphate, phenylmethylsulfonyl fluoride, elastatinal, Trasylol, and furoyl-saccharin. The optimum pH for platelet lysate activity was 8.5-9.0, as in other studies using elastin as substrate. The pH 4.5 eluate obtained after incubation of the lysate with dog lung elastin at neutral pH exhibited the same catalytic properties as the activity in the lysate. The different substrate and inhibitor specificities and the failure of IgG specific for neutrophil elastase to remove elastase-like and elastolytic activities from the lysate indicate that a unique elastase occurs in platelets.

Neutrophil turnover in normal rabbit lungs.

Neutrophil turnover was studied in the blood and alveoli of normal rabbits. Blood neutrophil turnover was examined by two different methods. In the first method, donor rabbit neutrophils were labeled in vivo by injecting tritium-labeled thymidine intravenously. After 72 h recipient rabbits received blood from the donors. The decline of the specific radioactivity of blood neutrophils was used to determine that their half-life was 4.03 h. In the second method, rabbit peritoneal exudative neutrophils were elicited with oyster glycogen. These cells were labeled with 111Indium oxine and infused into the blood of recipient animals. By their decline in specific radioactivity, the half-life of the blood neutrophil was 4.08 h. These half-lives are not significantly different. Lung lavage was performed on the animals that received the 111Indium-labeled neutrophils and the turnover time of the lung neutrophil was found to be 2.63 h. The turnover of the alveolar neutrophil pool accounted for only 0.19% of the total turnover of the blood neutrophils. Therefore, the lung appears to contribute only minimally to the total capacity of the body to dispose of neutrophils.

Search for drugs that may reduce the load of neutrophil azurophilic granule enzymes in the lungs of patients with emphysema.

Neutrophil elastase and myeloperoxidase probably play an important role in the development of pulmonary emphysema. We have analyzed drugs from the major classes of agents that alter neutrophil function to determine if there are drugs in use today that can reduce the load of neutrophil elastase or myeloperoxidase in the lungs of smokers. Eleven representative drugs were tested for their ability to inhibit chemotaxis and degranulation. None of the drugs inhibited chemotaxis in a dose-response fashion at concentrations achievable in human plasma. Sulfinpyrazone, phenylbutazone, and auranofin completely inhibited the release of azurophilic granules (myeloperoxidase) and tertiary granules (beta-D-glucuronidase) when formyl-L-methionyl-L-leucyl-L-phenylalanine (FMLP) was used as the stimulant, and inhibited azurophilic granule release by 69%, 19%, and 64% respectively, but not tertiary granule release when macrophage-conditioned media was used as the stimulus. In conclusion, none of the drugs tested are inhibitors of chemotaxis; however, three are excellent inhibitors of azurophilic granule enzyme release. Of these three, sulfinpyrazone, a drug that is not currently used clinically for its antiinflammatory effects, is the least toxic and should be considered as a potential drug to reduce the elastase and myeloperoxidase load in the lungs of smokers who are developing emphysema.

A synthetic peptide which specifically inhibits heat-treated interleukin-8 binding and chemotaxis for neutrophils

Interleukin-8 (IL-8) is a peptide which is secreted by stimulated human monocytes and which is chemotactic for human neutrophils. We synthesized three overlapping peptides spanning the amino-terminal region of the IL-8 sequence. None of the peptides retained the chemotactic activity of the native molecule. One of the peptides, IL-8(3–25), inhibited the neutrophil chemotactic activity of recombinant IL-8 (rIL-8) which had been preheated to 40°C but did not reduce neutrophil chemokinesis, or the chemotactic activity of unheated rIL-8, FMLP, C5a or LTB4. Interleukin-8 exhibited similar binding kinetics and chemotaxis for neutrophils regardless of whether it had been pretreated at 40°C.In addition, IL-8(3–25) was also able to decrease the binding of prehead IL-8 to neutrophils. IL-8(3–25), which can self-associate, binds directly to receptors on the neutrophil. The data suggest that heat-treated, but not untreated, IL-8 causes the IL-8(3–25) multimers to disaggregate, allowing the monomeric peptide to directly bind to the IL-8 receptor and thus inhibiting IL-8/receptor binding.

Reduction of neutrophil elastase load in the lungs of patients with emphysema by reducing neutrophil enzyme secretion or chemotaxis.

In recent years, a large body of scientific data has led investigators to formulate the hypothesis that emphysema in patients with a deficiency of alpha,proteinase inhibitor or in cigarette smokers may be caused by the unrestrained action of neutrophil elastase. In studies carried out over the last several years, we determined that neutrophils migrate through normal lungs.* The pool size of airspace neutrophils was 6 x 104 neutrophils per gram of lung,3 and 38-42% of the pool turns over per hour.4 These observations indicate that lungs are always exposed to a small pool of neutrophils with a rapid turnover. Two chemotactic factors for neutrophils are secreted by human alveolar macr ~ p h a g e s . ~ One factor has a molecular weight of 350, whereas the other factor is a peptide with an apparent molecular weight of about 8,000. Human alveolar macrophages secrete a peptide with a molecular weight of about 8,000, which is capable of releasing the azurophilic granule enzymes elastase and myeloperoxidase from human neutrophils, which were bound to membranes.6 There is now no question that neutrophil elastase, myeloperoxidase. and P-glucuronidase (azurophilic granule enzyme constituents) can be found in bronchoalveolar lavage fluids from normal subjects. The fate of these enzymes is