Gerald A. Fells

Antielastases of the human alveolar structures. Implications for the protease-antiprotease theory of emphysema.

The current concepts of the pathogenesis of emphysema hold that progressive, chronic destruction of the alveolar structures occurs because there was in imbalance between the proteases and antiproteases in the lower respiratory tract. In this context, proteases, particularly neutrophil elastase, work unimpeded to destroy the alveolar structures. This concept has evolved from consideration of patients with alpha 1-antitrypsin deficiency, who have decreased levels of serum alpha 1-antitrypsin and who have progressive panacinar emphysema. To directly assess the antiprotease side of this equation, the lower respiratory tract of non-smoking individuals with normal serum antiproteases and individuals with PiZ homozygous alpha 1-antitrypsin deficiency underwent bronchoalveolar lavage to evaluate the antiprotease screen of their lower respiratory tract. These studies demonstrated that: (a) alpha 1-antitrypsin is the major antielastase of the normal human lower respiratory tract; (b) alpha 2-macroglobulin, a large serum antielastase, and the bronchial mucous inhibitor, an antielastase of the central airways, do not contribute to the antielastase protection of the human alveolar structures; (c) individuals with PiZ alpha 1-antitrypsin deficiency have little or no alpha 1-antitrypsin in their lower respiratory tract and have no alternative antiprotease protection against neutrophil elastase; and (d) the lack of antiprotease protection of the lower respiratory tract of PiZ individuals is a chronic process, suggesting their vulnerability to neutrophil elastase is always present.

Oxidant-mediated epithelial cell injury in idiopathic pulmonary fibrosis.

Lung inflammatory cells of patients with idiopathic pulmonary fibrosis (IPF) were evaluated for their ability to injure 51Cr-labeled AKD alveolar epithelial cells in the presence and absence of IPF alveolar epithelial lining fluid (ELF). The IPF cells were spontaneously releasing exaggerated amounts of superoxide (O.2) and hydrogen peroxide (H2O2) compared with normal (P less than 0.02). Cytotoxicity of the AKD cells was markedly increased when the IPF inflammatory cells were incubated with autologous ELF (P less than 0.02). The majority of IPF patients had ELF myeloperoxidase levels above normal (P less than 0.002). Incubation of IPF ELF with AKD cells in the presence of H2O2 caused increased cellular injury (P less than 0.01 compared with control), which was suppressed by methionine, a myeloperoxidase system scavenger. IPF patients with high concentrations of ELF myeloperoxidase deteriorated more rapidly than those with low ELF myeloperoxidase (P less than 0.05). Thus, IPF is characterized by an increased spontaneous production of oxidants by lung inflammatory cells, the presence of high concentrations of myeloperoxidase in the ELF of the lower respiratory tract, and a synergistic cytotoxic effect of alveolar inflammatory cells and ELF on lung epithelial cells, suggesting oxidants may play a role in causing the epithelial cell injury of this disorder.

Collagenase in the lower respiratory tract of patients with idiopathic pulmonary fibrosis.

To test the hypothesis that idiopathic pulmonary fibrosis (IPF) is mediated through collagenase present in the lower respiratory tract, we used the fiberoptic bronchoscope to obtain fluid from the lower respiratory tract of 24 patients with IPF, 18 controls and nine patients with sarcoidosis. The fluid was analyzed for a variety of enzymes, including collagenase. Fifteen of 21 patients with IPF showed collagenase activity, whereas normal controls and patients with sarcoidosis showed none (P greater than 0.001, for all comparisons). In two patients with IPF who were re-evaluated after eight to 24 months, the collagenase activity was persistent. Fluid from patients with IPF also contained elevated levels of a non-specific neutral protease (P greater than 0.01 compared with controls), but there was no elastase activity in fluid from patients with IPF or from controls. The collagenase found in lavage fluid in IPF cleaved lung collagen into collagenase-specific TCA and TCB fragments. We conclude that in IPF the collagen of the lung is subjected to sustained lysis, followed by disordered resynthesis, and that the presence of active collagenase in the lower respiratory tract is a specific feature of the alveolitis associated with this disease.

Anti-Neutrophil Elastase Defense of the Normal Human Respiratory Epithelial Surface Provided by the Secretory Leukoprotease Inhibitor

Secretory leukoprotease inhibitor (SLPI), a 12-kD nonglycosylated serine antiprotease with a high capacity for inhibiting neutrophil elastase (NE), is produced by cells of mucosal surfaces including the human lung. The molar concentrations of SLPI in total respiratory tract epithelial lining fluid (ELF) were 56 +/- 10% that of alpha 1-antitrypsin, suggesting SLPI may be more important for the anti-NE protection of the pulmonary epithelial surface than previously thought. However, evaluation demonstrated that SLPI in respiratory ELF was only one-third functional. Studies aerosolizing recombinant SLPI (rSLPI) to sheep demonstrated that in the short term, neither aerosolization and alveolar deposition nor the lavage procedure inactivated the SLPI molecule. In vitro studies with rSLPI demonstrated that exposure to oxidants did not modify the form of the molecule, while exposure to oxidants and NE caused the molecule to be cleaved from 12 to 8 kD. Consistent with this, evaluation of SLPI in lavage fluid of individuals with cystic fibrosis (a condition with oxidants and NE on the respiratory epithelium) showed that the SLPI was degraded. However, evaluation of SLPI in normal ELF by molecular sieve analysis and Western analysis demonstrated an intact 12-kD molecule, suggesting that the partial inactivation of SLPI in normals in vivo is not because it is complexed to NE or exposed to oxidants + NE. Together, these observations demonstrate that SLPI is present in large amounts in respiratory ELF, but since the majority of the SLPI is inactive, it likely does not play a significant role in protecting the normal respiratory epithelium, except perhaps in the upper airways where the levels of SLPI are the highest.

Neutrophil accumulation in the lung in alpha 1-antitrypsin deficiency. Spontaneous release of leukotriene B4 by alveolar macrophages.

The emphysema of alpha 1-antitrypsin (alpha 1AT) deficiency is conceptualized to result from insufficient alpha 1AT allowing neutrophil elastase to destroy lung parenchyma. In addition to the deficiency of alpha 1AT in these individuals resulting from mutations in the alpha 1AT gene, it is recognized that, for unknown reasons, there are also increased numbers of neutrophils in their lungs compared with normal individuals. With the knowledge that alveolar macrophages have surface receptors for neutrophil elastase, we hypothesized that the neutrophil accumulation in the lower respiratory tract in alpha 1AT deficiency may result, in part, from release of neutrophil chemotactic activity by alveolar macrophages as they bind uninhibited neutrophil elastase. Consistent with this hypothesis, alpha 1AT-deficient alveolar macrophages spontaneously released nearly threefold more neutrophil chemotactic activity than normal alveolar macrophages. Analysis of alpha 1AT-deficient macrophage supernates by reverse-phase HPLC, molecular sieve chromatography, radioimmunoassay, and absorption with anti-LTB4 antibody revealed that the majority of the chemotactic activity was leukotriene B4 (LTB4), a mediator absent from normal macrophage supernates. Consistent with this hypothesis, incubation of normal macrophages with human neutrophil elastase resulted in the release of the same neutrophil chemotactic mediator. Furthermore, purified human alpha 1AT was able to prevent the neutrophil elastase from stimulating the macrophages to release the chemotactic factor. Together, these findings suggest that the absence of a normal antineutrophil elastase screen in the lower respiratory tract permits free neutrophil elastase to bind to alveolar macrophages, resulting in the release of LTB4, a process which attracts neutrophils to the alveoli of alpha 1AT deficient individuals, thus accelerating the lung destruction that characterizes this disorder.

Eosinophil-mediated injury to lung parenchymal cells and interstitial matrix. A possible role for eosinophils in chronic inflammatory disorders of the lower respiratory tract.

Eosinophils are a common component of the inflammation of the lower respiratory tract that characterizes the interstitial lung disorders. Bronchoalveolar lavage analyses (n = 680) of 251 patients with interstitial lung disease demonstrated that eosinophils represented greater than 5% of the effector cells comprising the alveolitis in 20% of all lavages. In contrast, lavage of normal individuals (n = 117) showed that eosinophils were never greater than 5% of the total effector cells recovered. To evaluate a possible role for eosinophils in mediating some of the cellular and connective tissue matrix derangements of the lung parenchyma found in interstitial disease, eosinophils were evaluated for the presence of proteases capable of cleaving connective tissue proteins found in the lung and for the ability to mediate cytotoxicity to lung parenchymal cells. Evaluation of guinea pig and human eosinophils demonstrated that eosinophil granules contained a collagenase that specifically cleaved human collagen types I and III, the two major connective tissue components of the human lung parenchyma. In contrast, the eosinophil did not contain an elastase or a nonspecific neutral protease. The eosinophil collagenase appeared to be a metalloprotease, as it was inhibited by ethylenediaminetetraacetate but not by phenylmethanesulfonyl-fluoride or alpha 1-antitrypsin. The eosinophil also has the capacity to injure lung parenchymal cells. Without further stimulation, eosinophils purified from peritoneal exudates of guinea pigs demonstrated spontaneous cytotoxicity for human lung fibroblasts (HFL-1), cat lung epithelial cells (AK-D) and rat lung mesothelial cells (I6B). Under identical conditions, the epithelial cells were more sensitive to eosinophil-mediated cytotoxicity than the fibroblasts or mesothelial cells (P less than 0.01), consistent with the clinical observation that in the interstitial disorders, the alveolar epithelial cells are damaged more commonly than fibroblasts or pleural cells. The eosinophil-mediated cytotoxicity could be partially inhibited by the antioxidants catalase and dimethylsulfoxide suggesting that toxic oxygen radicals play a role in mediating the cellular damage. Importantly, eosinophils purified from bronchoalveolar lavage of human interstitial lung disease also demonstrated spontaneous cytotoxicity for lung epithelial cells. These observations demonstrate that eosinophils are frequent participants of the alveolitis of the interstitial lung disorders and suggest that these cells have the potential to damage the parenchymal cells and collagen matrix of the lower respiratory tract.

Human neutrophil elastase functions as a type III collagen “Collagenase”

Summary Under physiologic conditions, human neutrophil elastase cleaves triple helical human type III collagen into specific fragments in a fashion similar to that of a typical “collagenase”. In contrast, elastase does not attack the triple helical portion of type I collagen. However, unlike a classic collagenase, which is a metalloenzyme, the action of elastase on type III collagen is prevented by serine protease inhibitors. These findings suggest that the concepts of “collagenases” and “elastases” must be broadened and that human neutrophil elastase is capable of mediating a broad range of connective tissue destruction.

Antioxidant macromolecules in the epithelial lining fluid of the normal human lower respiratory tract.

We hypothesized that the alveolar structures may contain extracellular macromolecules with antioxidant properties to defend against oxidants. To evaluate this 51Cr-labeled human lung fibroblasts (HFL-1) and cat lung epithelial cells (AKD) were exposed to a H2O2-generating system and alveolar epithelial lining fluid (ELF) from healthy nonsmokers was tested for its ability to protect the lung cells from H2O2-mediated injury. The ELF provided marked antioxidant protection, with most from a H2O-soluble fraction in the 100-300-kD range. Plasma proteins with anti-H2O2 properties were in insufficient concentrations to provide the antioxidant protection observed. However, catalase, a normal intracellular antioxidant, was present in sufficient concentration to account for most of the observed anti-H2O2 properties of ELF. Depletion of ELF with an anticatalase antibody abolished the anti-H2O2 macromolecular defenses of ELF. Since catalase is not normally released by cells, a likely explanation for its presence in high concentrations in normal ELF is that it is released by lung inflammatory and parenchymal cells onto the epithelial surface of the lower respiratory tract during their normal turnover and collects there due to the slow turnover of ELF. It is likely that catalase in the ELF of normal individuals plays a role in protecting lung parenchymal cells against oxidants present in the extracellular milieu.

Risk factors for emphysema. Cigarette smoking is associated with a reduction in the association rate constant of lung alpha 1-antitrypsin for neutrophil elastase.

The increased risk of developing emphysema among individuals who smoke cigarettes and who have normal levels of alpha 1-antitrypsin (alpha 1AT) is hypothesized to result from a decrease in the antineutrophil elastase capacity of the lower respiratory tract alpha 1AT of smokers compared with nonsmokers. To evaluate this hypothesis we compared the time-dependent kinetics of the inhibition of neutrophil elastase by lung alpha 1AT from healthy, young cigarette smokers (n = 8) and nonsmokers (n = 12). alpha 1-antitrypsin was purified from lavage fluid using affinity and molecular sieve chromatography, and the association rate constant (k assoc) for neutrophil elastase quantified. The k assoc of smoker plasma alpha 1AT (9.5 +/- 0.5 X 10(6) M-1s-1) was similar to that of nonsmoker plasma (9.3 +/- 0.7 X 10(6) M-1s-1, P greater than 0.5). In marked contrast, the k assoc of smoker lower respiratory tract alpha 1AT was significantly lower than that of nonsmoker alpha 1AT (6.5 +/- 0.4 X 10(6) M-1s-1 vs. 8.1 +/- 0.5 X 10(6) M-1s-1, P less than 0.01). Furthermore, the smoker lower respiratory tract alpha 1AT k assoc was significantly less than that of autologous plasma (P less than 0.01). When considered in the context of the concentration of alpha 1AT in the lower respiratory tract epithelial lining fluid, the inhibition time for neutrophil elastase of smoker lung alpha 1AT was twofold greater than that of nonsmoker lung alpha 1AT (smoker: 0.34 +/- 0.05 s vs. nonsmoker: 0.17 +/- 0.05 s, P less than 0.01). Consequently, for concentrations of alpha 1AT in the lower respiratory tract it takes twice as long for an equivalent amount of neutrophil elastase to be inhibited in the smokers lung compared with the nonsmokers lung. These observations support the concept that cigarette smoking is associated with a decrease in the lower respiratory tract neutrophil elastase inhibitory capacity, thus increasing the vulnerability of the lung to elastolytic destruction and thereby increasing the risk for the development of emphysema.

Cells, collagen and idiopathic pulmonary fibrosis

Idiopathic pulmonary fibrosis (IPF) is a usually fatal disorder of lung with clearly defined clinical, roentgenographic, physiologic, morphologic, scintigraphic and bronchoalveolar lavage features. Current concepts of the pathogenesis of this disorder suggest a central role for a chronic alveolitis in causing changes in parenchymal cell populations and derangements in interstitial collagen. Of the many inflammatory and immune effector cells comprising the alveolitis of IPF, it is likely that the neutrophil is the most important mediator of parenchymal damage. To follow the status of lung neutrophils in patients with this disease, two methods have been utilized. Both gallium-67 scanning and bronchoalveolar lavage quantitate the extent of the alveolitis and can be used to stage and follow these patients. The treatment of IPF remains controversial, but it is likely that corticosteroids reduce the alveolitis and prolong the lifespan of these patients.