Neutrophil Elastase Activity and Superoxide Production Are Diminished in Neutrophils of Alcohollcs':'

CLIFFORD W. SACHS,4 ROBERT H. CHRISTENSEN, PHILIP C. PRATT, and WILLIAM S. LYNN

Introduction

Centrilobular emphysema (CLE) is a chronic debilitating lung disease of smokers, anatomically characterized by enlargement of the distal airspaces and destruction of alveolar walls (1). Phagocytic inflammatory cells are thought to playa key role in the pathogenesis of CLE by increasing the protease burden and inactivating protease inhibitors (2- 5). Neutrophil elastase (NE) is found in the cytoplasmic azurophil granules of the polymorphonuclear leukocyte (PMN) (6-8). It is of particular interest because of its ability to degrade several structural components of lung tissue, including basement membrane, collagen, and elastin (9). In vivo, NE activity is normally regulated by a plasma serpin, at-antitrypsin (at-AT) (10-12), the major protease inhibitor of the alveolar spaces (13). Hydrogen peroxide, superoxide, and products of myeloperoxidase secreted by stimulated neutrophils and macrophages suppress the elastase inhibitory capacity of at-AT (14, 15) via oxidation of a critical active site methionine residue (16-19). In addition, recent studies indicate that phagocytic inflammatory cells downregulate at-AT activity by proteolytic cleavage of the inhibitor (20-22). Reports that the lung lavage fluid of smokers has decreased elastase inhibitory capacity (23), inactive at-AT with oxidized methionine residues (24), and NE activity in the presence of altered inhibitor (25) provide evidence that in vivo oxidation and proteolysis of this inhibitor are pivotalsteps in the pathogenesis of elastase-mediated damage. Using inflation-fixed human lungs and morphometry, Pratt and Vollmer noted that the consumption of ethanol was associated with a decrease in the prevalence and severity of CLE (26). Correlating smoking and drinking history with quantitative evaluation of histologic and gross pulmonary lesions, these authors found that the probability of the negative association of alcohol consumption with CLE (p < 0.008) was as significant as the

SUMMARY Several lines of evidence indicate that neutrophil elastase and oxidants secreted by phagocytic Inflammatory cells play key roles In the development of centrllobular emphysema. A recent report suggests that ethanol consumption may have a protective role against this disease in smokers. The aim of this study was to examine the effect of ethanol consumption on neutrophil elastase activity and superoxide production of peripheral blood neutrophils. These activities were measured In neutrophils from 52 male intoxicated patients and compared to activities In neutrophlls from 20 male volunteers. Neutrophils from intoxicated patients contained 31%less elastase actiVity than that found in controls, 0.99 ± 0.27 versus 1.44 ± 0.23 I1g110' neutrophlls (p < 0.0001) and produced 25 to 27% less superoxide than controls In response to phorboI12-myrlstate-13-acetate, 0.90 ± 0.17versus 1.2 ± 0.21 nmol/mln/10' PMN (p < 0.0001)or N-formylmethlonylleucylphenyl" alanlne,0.64 ± 0.19 versus 0.88 ± 0.24 (p < 0.001). In follow-up studies of 10 patients admitted for acute alcoholism, elastase activity and superoxide production remained low for 2 to 4 days. After 6 to 10 days, elastase actiVity and superoxide production were significantly greater than they were at Day 0 and approached normal levels. Neutrophlls Isolated from blood samples of healthy abstaining donors, which had been exposed to ethanol or to plasma from inebriated patients for 16 to 20 h, showed no loss of elastase actiVity or superoxlde production. We conclude that the protective role of ethanol against emphysema In smokers stems from the effects of ethanol on marrow precursors of neutrophlls, with the result that neutrophils deficient In elastase and enzymes AM REV RESPIR DIS 1990; 141:1249-1255 . of oxidative metabolism are produced.

positive association of cigarette smoking (0.68 to 0.9 g/dl) induced differentiation with CLE (p < 0.009) (26). Alcohol is of neutrophils, which readily produced reported to inhibit many actions of phago- superoxide in response to soluble stimucytic inflammatory cells, including ad- li (37). If ethanol had similar actions in herence (27,28), migration (29), mobili- vivo, this model would predict that alcozation (30,31), and release of lysosomal hol consumption is associated with the enzymes (32). However, the possible ef- circulation of elastase-deficient neutrofects of ethanol on the activity of neu- phils that secrete superoxide normally. trophil elastase have not been reported. The present study was designed to The azurophilic or primary granules evaluate this hypothesis. To do so, neuthat contain neutrophil elastase and my- trophil elastase activity and superoxide eloperoxidase are formed during the promyelocytic stage of neutrophil differentiation and are distributed to daughter cells(myelocytes) after mitosis (33).There- (Received in original form April 3, 1989 and in fore we initiated a series of experiments revised form September 27, 1989) using the human promyelocytic cell line From the Departments of Pathology and MediHL-60, a model system of myeloid dif- cine, Duke University School of Medicine, and the ferentiation (34-36), to determine the ef- Department of Laboratory Service, Durham Vetfects of ethanol on neutrophil precursors. erans Administration Medical Center, Durham, The major finding of these experiments North Carolina. by an institutional small grant from was that culture of HL-60 promyelocytes the2 Supported Walker P. Inman Fund. with pharmacologic concentrations of 3 Correspondence and requests for reprints should ethanol (0.15 to 0.45g/dl) blocked a three- be addressed to Dr. Philip Pratt, Department of to fivefold increase in NE activity elicited Pathology, Box 3712,Duke University Medical Cenby exposure to recombinant granulocyte- ter, Durham, NC 27710. 4 Supported as a predoctoral fellow in the Duke macrophage colony stimulating factor University Integrated ToxicologyTraining Program, (GM-CSF). Furthermore, exposure of Grant ES 07031 from the National Institute of EnHL-60 cells to higher doses of ethanol vironmental Health Sciences. 1

1249

1250

SACHS, CHRISTENSEN, PRATT, AND LYNN

production were measured in PMN obtained from 52 inebriated patients with plasma ethanol> 0.2 g/dl and were compared to PMN isolated from 20 healthy controls. In 10 patients admitted to an alcoholic rehabilitation program, NE activity and superoxide production in circulating PMN were measured in followup studies during 2 wk of abstinence following intoxication. Methods

Materials Ficoll-Hypaque (LSM) was obtained from Organon Teknika (Durham, NC). Purified neutrophilelastasewas purchasedfrom Elastin Products (St. Louis, MO). All other supplies were obtained from Sigma (St. Louis, MO). Patient and Control Populations The patient population was composed of 52 individuals whose plasma ethanol was greater than 200mg/dl seenin the emergencyroom of the VeteransAdministration Medical Center in Durham, North Carolina. Clinical chemistry tests including ethanol determination were performed on blood plasma obtained at admission after centrifugation of heparinized venous blood samples at 200 x g for 10 min. The plasma-depleted blood samples utilized in this study were kept at room temperature until neutrophil purification was performed. Samples werenot used if they (I) were more than 24 hold; (2) contained considerablehemolysis;or (3)had clumped buffy coats. For 10 of these patients, who had been admitted to the alcohol rehabilitation program, follow-up studies were performed. Each patient was entered in the study after an explanation of its nature and purpose and after each had signed a consent form that had been approved by the hospital ethics committee. Typically, patients admitted to this program for treatment of their alcohol problem are hospitalized for 2 wk. Efforts were made to obtain follow-up blood samples for study 2 to 4 days after admission, 6 to 10 days after admission, and 12to 15days after admission. The control population was composed of apparently healthy laboratory volunteers who stated that they had not consumed alcohol in the preceding week. Preparation of Neutrophils A standard technique employing dextran sedimentation, Ficoll-Hypaquegradients,and hypotonic lysis (38) was used to isolate neutrophils from the lower,plasma-depleted, cellular phase of each blood sample. Since no appreciabledifferenceswereobservedbetween PMN isolated from whole blood or plasmadepleted blood, follow-up studies utilized PMN isolated from whole blood. Counts and viabilityweredetermined with trypan blue and a hemacytometer. Cells isolated in this method were greater than 95% viable. These cells were kept on ice until assay in Dulbecco's

phosphate-buffered saline (DPBS) containing 1 mM dextrose at a final concentration of 101 / ml.

Superoxide Production The maximal rate of superoxide production by stimulated neutrophils was measured in a continuous assay as the superoxide dismutase (SOD), inhibitable reduction of ferricytochrome c (39). Briefly,PMN wereprewarmed to 37° C for 5 min and added to a l-ml cuvette in the thermostated compartment of a dualbeam spectrophotometer (Varian series 634). The final concentrations for this assay were 0.5 to 1 x 106 PMN/ml, 50 JlM horse heart cytochrome c, and 1 mM dextrose dissolved in DPBS. A resting rate of reduction was recorded for 1 min, and the cellswerethen stimulated with 10-1 M formylmethionylleucylphenylalanine (FMLP) or 50 ng/ml of phorbol 12-myristate-13 acetate (PMA). The rates of change in absorbance at 550 nm in the presence of 10 ug/ml of purified bovine erythrocyte SOD were subtracted from the rates obtained in the absence of scavenger. Replicate determinations wereobtained for each agonist. Extraction and Measurement of Neutrophil Elastase A zoo-at aliquot containing 2 x 106 PMN was centrifuged at 400 x g for 5 min. The supernatant was removed and the cell pellet was lysed in 1 ml of 50 mM TRIS-HCl, pH 7.5,0.5070 (wt/vol) cetylammonium bromide, and 0.2% Triton X-l00 and subjected to three cyclesof freezing and thawing in a methanoldry ice slurry (40). The samples werewarmed to room temperature and centrifuged in an Eppendorf microfuge (12,000 x g) for 5 min. The supernatants of these detergent extracts contained greater than 95% of the esterolytic neutrophil elastase activity. Rates of hydrolysis of the sensitive, specific substrate of NE, methoxy-succinyl-alanyl-alanyl-prolylvalyl-p-nitroanaline (MeOSuc-ala-ala-provalpNA) (41), were measured in loo-Jll aliquots of detergent extract supernatant and related to a standard curve constructed with purified neutrophil elastase (between 100and 500 ng/ ml) obtained from a commercial source. Assessment of the In Vitro Effects of Ethanol and Ethanol-containing Plasma on NE and Superoxide Production To determine if measurements made on neutrophils were altered by in vitro exposure to ethanol during the time that elapsed between when the sample was drawn and when the PMN werepurified, the following control experiments were performed. Heparinized venous blood from healthy donors (60 ml) was apportioned equally in six serum collection tubes and allowed to sediment by gravity into plasma and cellular phases. The plasma phase in two of these six samples was spiked with 60 ul of95% ethanol; two sampleswerespiked with 35 ul, and two of these six samples had no added ethanol. Immediately after addition, the four ethanol-containing tubes and controls were mixed by inversion. The sam-

ples were then centrifuged at 200 x g for 10 min, and plasma was removed for determination of ethanol. Neutrophils wereimmediately prepared and studied from one of the samples containing no ethanol and one sample of each set spiked with ethanol as described. The remaining three samples, one from each set, were sealed and kept at room temperature overnight. After 20 to 24 h, the PMN from the remaining three samples were isolated and studied. To determine if superoxide production or NE activity of neutrophils might be altered by in vitro exposure to cytokines or inflammatory mediators released by tissues or inflammatory cellsin response to ethanol intoxication, leukocytes from a healthy abstaining donor were exposed to plasma from intoxicated individuals. All plasma samples used for this study were free of hemolysis and were stored frozen until the time of the experiment, when the samples were thawed and filtered through a 0.45 um filter. Heparinized venous blood of a healthy abstaining donor was centrifuged at 400 x g for 10 min. The resulting buffy coat was carefully removed with a Pasteur pipette, minimizing contamination by erythrocytes.In each experimentaliquots were added to 2 vol of plasma from alcoholics or autologous plasma (controls). After the sampleswereincubated 16to 20h at room temperature, neutrophils were isolated and studied.

Statistical Analysis The number of classes in the histograms shown in figures 1 and 2 was determined by Tukey's rule (the number of classes k to be used in construction of a histogram is the smallest integer k:2k > n, where n is the number of observations) (42). Statistical significance was determined from the results of a one-way analysis of variance (ANOVA) or Wilcoxon-Mann-Whitney U test as indicated in the text (43). Results

During the period of study, neutrophils were purified and studied from 52 male patients with a blood alcohol greater than 200mg/dl, mean 319 ± 97mg/dl(± SD), and from 20 male volunteers. Of the patients in this study, three had a plasma ethanol greater than 500 mg/dl, a level often associated with mortality, but none of these patients died. There was no significant difference between the average age of the intoxicated population and the controls, 46 ± 13 versus 42 ± 13 yr. Significant differences between rates of superoxide production and elastase content of neutrophils from the intoxicated and control groups were readily apparent in the frequency distributions as shown in figures 1 and 2. Figure 1 shows the frequency distribution of the measured elastase content of PMN isolated from intoxicated patients

1251

DECREASED ELASTASE ACTIVITY AND SUPEROXIDE PRODUCTION IN PMN OF ALCOHOLICS

50 Fig. 1. Frequency distribution of elastase content of PMNs isolated from intoxicated patients and controls. Filled bars designate the patient population; open bars designate the controls. The distribution clearly shows the trend for decreased elastase activity in the inebriated patients, although the curves do overlap.

_

40

~

~ 30

c

g.'" 20 ~'"

10 055· 074

075094

095-

115

135-

155-

114

134

154

174

)19 Neutrophil Eiastase/l06PMNs

and controls. The average NE content of the intoxicated patients was 0.99 ± 0.27 versus 1.44 ± 0.23 J.1g/106 PMN in the controls. The difference of these means was statistically significant, p < 0.0001. Significant differences in rates of superoxide production were also observed when either PMA or FMLP was used as stimulus (figure 2). The upper panel of figure 2 displays a frequency distribution of the rates of superoxide production by neutrophils of alcoholic patients and controls in response to the formyl chemotactic peptide FMLP. The average rates of neutrophil superoxide production in response to 10-7 M FMLP were 0.64 ± 0.19and 0.88 ± 0.24nmol/min/Kr PMN in the intoxicated patient and control groups, respectively. The lower panel shows the frequency distribution of rates of superoxide production in response to the phorbol diester, PMA. The average rate of neutrophil superoxide production in response to 50 ng/ml of PMA was 0.90 ± 0.17 in the intoxicated patients compared to 1.20 ± 0.21 nmol/min/Kr PMN in the controls. Levelsof significance for the differences in rates of superoxide production between the two groups were

determined to be p :

1.2 "'Q ~

P:

-'- 0 -

Controls

0.8

....c:

~ Q)

0 .4

o

reviewed patients weremoderate or heavy smokers (> 1.5 packs/day). Five of these smokers werediagnosed as having chronic obstructive pulmonary disease (COPD). No significant differences in the average NE activity or average superoxide production were observed between the patients with COPD and the remaining intoxicated patients. Since none of the patients have died, the nature of the lesions responsible for COPD is not known (1). However, in only one case did the chest radiographs meet validated criteria for emphysema (46);the other four probably had chronic small airway disease (46). Concurrent alcohol-related diseases were as follows: One patient had pneumonia. Two patients had diagnosed hepatic cirrhosis. Abnormal values for the enzymes SGOT, SGPT, LDH, and alkaline phosphatase were observed in 12patients, indicating liver dysfunction. One of these patients had pancreatitis. No significant differences in mean NE activity or mean rates of superoxide production were observed when these patients were compared to the remaining intoxicated patients. Complete blood counts were available for 50 of the inebriated patients, for whom the average white blood cell count (WBC) was comparable to that of the controls, 8.1 ± 2.8 versus 7.0 ± 1.9 X 103 /mrrr'. Of the 50 patients, 38 had normal WBC values; 8 had high WBCs (> 10.8 x 103/mm3 , average 12.9 ± 1.9), and 4 had low WBCs « 4.8 x 103/mm3 , average 4.2 ± 0.4). In 21 patients, including all 4 patients with low WBC and 5 of 8 patients with high WBC, hemoglobin or hematocrit levels were below the normal range of values. Again, no significant differences were observed when NE

2-4

6-10

Doys of Abstin ence

12-15

04 0 E .5 I 0 .N 0

Fig. 4. Comparison of changes in neutrophil elastase activity and stimulated superoxide production of PMN during abstinence following alcoholic inebriation. Values obtained for NE and superoxide production from neutrophils in 10 patients were averaged for the intervals 2 to 4 days, 6 to 10 days, and 12 to 15 days and compared to average values obtained at admission (Day 0). Note that by 12 to 15 days the patients' levels were similar to those of the controls. Open bars indicate the NE activity. Superoxide production in response to FMLP is represented by hatched bars and in response to PMA, filled bars. The error bars show standard deviations. The significance of the difference between the values in the indicated interval and Day 0 average are as follows: (-) p < 0.05; (..J P < 0.005.

activity or rates of superoxide production by PMN of these 21 patients were compared to those of the other intoxicated patients. Discussion

Current concepts of the pathogenesis of smoking-related emphysema suggest that the destruction of lung tissue occurs in localized areas of elastase excess and/or a deficiency of a.-AT (47). Cigarette smoking causes a marked increase in the number of inflammatory cells in the lung (3, 48, 49). The accumulation of neutrophils effectively increases the elastase burden of the lung because the azurophil granules of the PMN are an abundant source of elastase (33). Intratracheal administration of crude granule preparations of PMN or purified NE in laboratory animals results in a lesion that resembles human emphysema (50,51). The accumulation of alveolar macrophages in the lungs of smokers (3, 48, 49) is thought to diminish the functional integrity of the antielastase shield of the lung because smoking stimulates oxidant and protease secretion by alveolar macrophages (15, 52, 53). Two proteases secreted by macrophages, cathepsin B and a metalloprotease elastase, can proteolytically cleave a.-AT, rendering it inactive (21, 22). These processes presumably play important roles in the creation of local areas deficient in inhibitor in which unimpeded activity of NE may destroy lung tissue. Over a period of three to four decades, accumulation of these tissue-destructive processes may ultimately result in sufficient proteolytic damage to account for the clinical manifestations of CLE despite normal levelsof a.-AT in the blood. In this context, the findings of the present study that NE activity and superoxide production by peripheral neutrophils are reduced for several days by alcohol consumption suggest that the observed beneficial and protective role of ethanol in CLE (26) may stem from a similar reduction in the activity of these enzymes in pulmonary neutrophils. PMN isolated from intoxicated patients were found to contain 31% less NE activity than those from the controls and produced 25 to 270/0 less superoxide than did the controls. In lung tissue of smokers, these cells would be less likely to adversely perturb the elastase-antitrypsin balance. It is unlikely that an acute effect of ethanol on mature circulating neutrophils could result in the observed persistent reduction in NE activity and superoxide

DECREASED ELASTASE ACTIVITY AND SUPEROXIDE PRODUCTION IN PMN OF ALCOHOLICS

production. No reduction in superoxide secretion or elastase activity was observed when PMN were exposed to ethanol in vitro for up to 24 h. To address the possibility that an activating inflammatory mediator may be released by tissues as a result of ethanol intoxication, neutrophils in buffy coats wereexposed to plasma from intoxicated patients for 16 to 20 h. There was no significant difference in superoxide production or elastase activity between PMN exposed to autologous plasma or to alcoholics' plasma. This experimental finding does not rule out the possibility that a putative inflammatory mediator not stable to freezing was present in the plasma, nor does it rule out the possibility that PMN of alcoholics are activated by cellsurface stimulation. Acetaldehyde, the primary metabolite of ethanol in the liver, is known to bind nonenzymatically to hepatocyte membranes without affecting hepatocyte function (54). Rat liver membrane vesicles exposed to acetaldehyde have been observed to stimulate superoxide production and degranulation by rat neutrophils (55). It is probable that stimulated PMN degranulation that resulted in a 31010 reduction of elastase activity would decrease the density of the neutrophils because of the concomitant secretion of other granule constituents. In this study, neutrophils wereisolated from other leukocytes on the basis of their greater density; therefore, PMN that had degranulated would be less likely to be isolated by this technique. Furthermore, blood samples that displayed buffy coat aggregation, suggestingthat the leukocytes had been activated, were not used in this study. Because PMN circulate for less than 10h (56), the direct actions of ethanol on mature cells would be of short duration. In contrast, we found that the average NE activity and superoxide production were still low after 2 to 4 days of hospitalization, when plasma tests for ethanol were negative. Significant increases to the low normal ranges of NE activity and superoxide production were observed after 6 to 10days. A further increase in elastase activity was observed after 12 to 15 days. The time course of these changes is consistent with an effect of ethanol on marrow neutrophils or their precursors. Our in vitro studies with the human promyelocytic cell line HL-60 indicate that ethanol inhibits an increase in neutrophil elastase activity specifically elicited with a physiologic regulator of granulopoeisis, GM-CSF (37). Since this is

the stage of neutrophil differentiation at which azurophil granule synthesis occurs (33), our model predicts that regular alcohol consumption results in the persistent production of elastase-deficient neutrophils because of the time required for promyelocytes to differentiate to neutrophils. Kinetic studies have indicated that the mature progeny of myeloblasts take about 10to 11 days to appear in the blood (56). However,the minimal length of time required to increase neutrophil output by increasing the differentiation of myeloblasts is only 5 to 7 days (56). Thus, it would seem that the dramatic increases in NE activity seen in three of the patients within 3 to 4 days should not have occurred. However, a recent study of NE gene expression in myelomonocytic cell lines and inflammatory cells found that the NE gene is expressed only transiently during myeloid differentiation, probably beginning with promyelocyte-to-myelocyte differentiation (57). Therefore, recovery of NE activity in these deficient individuals may have kinetics closer to the generation time required for a myelocyte to differentiate to a neutrophil. Studies of neutrophils in alcoholics have suggested two abnormalities that may be relevant. First, decreased granulocyte reserves have been demonstrated by endotoxin challenge (58). Studies of serum lactoferrin (used to evaluate turnover and activity of neutrophils in vivo) in sober, recently intoxicated, alcoholics suggest that increased turnover and concomitant increased production of peripheral neutrophils occur subsequent to intoxication (59). These findings indicate that ethanol may stimulate marrow egress of neutrophils. Thus, progeny of promyelocytes appear in the bloodstream earlier than in normal healthy donors, similar to what is observed in patients with infection. In the HL-60 model, ethanol was observed to promote neutrophil differentiation. During maturation, these cells acquired the ability to generate superoxide, presumably due to synthesis of respiratory burst enzymes. However,in the present study no marked increase in circulating neutrophils or enhanced ability to generate superoxide was observed in neutrophils of intoxicated humans. In fact the opposite was observed: neutrophils of alcoholics, when maximally stimulated with PMA to secrete superoxide, produced 25% less than controls and, when stimulated with FMLP, 27% less than controls. Thus, extrapolation of this in vitro finding in a leukemic cell line at doses 30% above a generally fatal con-

1253

centration of ethanol and two to three times the average concentrations seen in these patients was clearly inappropriate. The HL-60 cell line can be induced to differentiate to PMN by the organic solvents dimethylsulfoxide (DMSO) and dimethylformamide (DMF) (34-36). The finding that roughly equimolar concentrations ofthese solvents (100mM DMF, 160 mM DMSO, and 140 mM ethanol) suggest that they may be acting by a similar mechanism, possibly by perturbing cell membranes. The approximately twofold difference between the average concentration of plasma ethanol in patients and the differentiating dose of ethanol is a likely source of the disparity in observations. The finding that superoxide production is decreased for 2 to 4 days and recovers by 6 to 10days with no further increase suggests that ethanol may have a more mature target cell than the promyelocyte, possibly the metamyelocyte, which acquires the capacity to generate superoxide (60). No correlations were found between the indirect effects of ethanol, such as the presence of alcohol-related disease, and diminished superoxide production and NE activity. Therefore, the finding that approximately 20% of the intoxicated population had a similar quantity of elastase as the controls might be explained on the basis of variation of individual susceptibility to alcohol. Alternatively, it is possible that neutrophils derived from elastase-deficient promyelocytes in an acutely intoxicated individual would not appear in the blood until several days after intoxication. These studies indicate that alcohol consumption decreases NE activity and superoxide production of circulating PMN via actions on marrow neutrophils or their precursors. The observation that alcohol consumption reduced the prevalence and extent of CLE (26) emphasizes the importance of this finding. We have used severely inebriated patients in this study for two reasons: (1) to have maximum effects for observation, and (2) the ethanol concentrations achieved by alcohol abuse in these patients are equivalent to those that block increases in NE activity elicited by GM-CSF in cultured promyelocytes. These ethanol concentrations are much greater than volunteers could ethically be asked to attain. It is worthy of note that many of the cases described in the morphologic study of emphysema (26) were moderate drinkers who consumed no more than two drinks per day. We postulate that the effects of

1254

SACHS, CHRISTENSEN, PRATT, AND LYNN

ethanol on their leukocytes wereof a lesser degree than observed in this study. Over the many years required to develop significant emphysema (61), lesser reductions of NE activity and superoxide production could result in the observed decreased prevalence and severity of involvement. A recent clinical study (62) showed an interaction between smoking and alcohol consumption that was in a direction opposite to the independent effects of alcohol and smoking. This finding supports the concept that ethanol provides a degree of protection against smoking-related emphysema. Although consumption of alcohol can hardly be recommended as a means of CLE prophylaxis, the fact that no other agent has been noted to have this action indicates that a search for other medications with similar effects on neutrophils or their precursors, but without the inherent risks of ethanol, is clearly appropriate. Acknowledgment The writers thank Evelyn Allen and the other Clinical Chemistry Lab technicians for their assistance throughout the study, and Earl Linthicum for his helpful advice in preparing the illustrations. References 1. NHLBI Workshop Summaries. The definition of emphysema: Report of a National Heart, Lung and Blood Institute, Division of Lung Diseases workshop. Am Rev Respir Dis 1985; 131:764-9. 2. Janoff A. Elastases and emphysema. Current assessment of the protease-antiprotease hypothesis. Am Rev Respir Dis 1985; 132:417-33. 3. Hunninghake GW, Crystal RG. Cigarette smoking and lung destruction: accumulation of neutrophils in the lungs of cigarette smokers. Am Rev Respir Dis 1983; 128:833-8. 4. Snider GL. Pathogenesis of emphysema and chronic bronchitis. Med Clin North Am 1981; 65:647-65. 5. navis J, SalvesenGS. Human plasma proteinase inhibitors. Annu Rev Biochem 1983; 52:655-709. 6. Bretz U, Baggiolini M. Biochemical and morphological characterization of azurophil and specific granules of human neutrophilic polymorphonuclear leukocytes. J Cell Bioi 1974;63:251-69. 7. Spitznagel KK, Dalldorf FG, Leffell MS, et al. Character of azurophil and specific granules purified from human polymorphonuclear leukocytes. Lab Invest 1974; 30:774-85. 8. Janoff A, Scherer J. Mediators of inflammation in leukocyte Iysosomes. IX. Elastinolytic activity in granules of human polymorphonuclear leukocytes. J Exp Med 1968; 128:1137-55. 9. Janoff A, White R, Carp H, Harel S, Dearing R, Lee D. Lung injury induced by leukocytic proteases. Am J Pathol 1979; 97:111-36. 10. Ohlsson K. Neutral leukocyte proteases and elastase are inhibited by plasma aI-antitrypsin. Scand J Clin Lab Invest 1970; 28:251-3. ll. Crystal RG, Brantly ML, Hubbard RC, Curiel DT,States DJ, Holmes MD. The alpha-antitrypsin gene and its mutations. Chest 1989; 95:196-208.

12. Carell RW.aI-Antitrypsin: Molecular pathology, leukocytes, and tissue damage. J Clin Invest 1986; 78:1427-31. 13. Gadek JF, Fells GA, Zimmerman RL, Rennard SI, Crystal RG. Antielastases of the human alveolar structures. Implications for the proteaseantiprotease theory of emphysema. J Clin Invest 1981; 68:889-98. 14. Carp H, Janoff A. Phagocyte-derivedoxidants suppress the elastase-inhibitory capacity of alphaI-proteinase inhibitor. J Clin Invest 1980; 66: 987-95. 15. Hubbard RC, Ogushi F, Gells GA, et al. Oxidants spontaneously released by alveolar macrophages of cigarette smokers can inactivate the active site of a.-antitrypsin, rendering it ineffective as an inhibitor of neutrophil elastase. J Clin Invest 1987; 1980:1289-95. 16. Johnson D, Travis J. The oxidative inactivation of human a.-proteinase inhibitor. J Bioi Chern 1979; 254:4022-6. 17. Ossanna PJ, Test ST, Matheson NR, Regiani S, Weiss SJ. Oxidative regulation of neutrophil elastase-alpha-l-proteinase inhibitor interactions. J Clin Invest 1986; 77:1939-51. 18. Matheson NR, Wong PS, Travis J. Enzymatic inactivation of human alpha-l-proteinase inhibitor by neutrophil myeloperoxidase. Biochem Biophys Res Commun 1979; 88:402-9. 19. Matheson NR, Wong PS, Schuyler M, navis J. Interaction of human alpha-I-proteinase inhibitor with neutrophil myeloperoxidase. Biochemistry 1980; 20:331-6. 20. Beatty K, Beith J, TravisJ. Kinetics of association of serine proteinases with native and oxidized a.-proteinase and a.-antichymotrypsin. J Bioi Chern 1980; 255:3931-4. 21. Banda MJ, Clark EJ, Werb Z. Limited proteolysis by macrophage elastase inactivates human a.-proteinase inhibitor. J Exp Med 1980; 152: 1563-70. 22. Banda MJ, Clark EJ, Werb Z. Regulation of a1pha-l proteinase inhibitor function by rabbit macrophages, J Clin Invest 1985; 75:1758-62. 23. Gadek JE, Fells GA, Crystal RG. Cigarrette smoking induces functional anti protease deficiency in the lower respiratory tract of humans. Science 1979; 206:1315-6. 24. Carp H, Miller F, Hoidal R, Janoff A. Potential mechanism of emphysema: alpha I-proteinase inhibitor recovered from lungs of cigarette smokers contains oxidized methionine and has decreased elastaseinhibitory capacity.Proc Natl Acad SciUSA 1982; 779:2041-5. 25. Martin WJ, Taylor JC. Abnormal interaction of a.-antitrypsin and leukocyte elastolytic activity in patients with chronic obstructive pulmonary disease. Am Rev Respir Dis 1979; 120:411-9. 26. Pratt PC, Vollmer RT. The beneficial effect of alcohol consumption on the prevalence and extent of centrilobular emphysema. A retrospective autopsy analysis. Chest 1984; 85:372-7. 27. MacGregor RR, Spagnuolo PJ, Lentnek AL. Inhibition of granulocyte adherence by ethanol, prednisone, and aspirin measured with an assay system. N Engl J Med 1974; 291:642-6. 28. Brayton RG, Stokes PE, Schwartz MS, Lauria DB. Effect of alcohol and various diseases on leukocyte mobilization, phagocytosis, and intracellular bacterial killing. N Engl J Med 1970; 282: 123-8. 29. Guarneri JJ, Laurenzi GA. Effect of alcohol on the mobilization of alveolar macrophages. J Lab Clin Med 1968; 72:40-51. 30. Gluckman SJ, MacGregor RR. Effect of acute alcohol intoxication on granulocyte mobilization and kinetics. Blood 1978; 52:551-9.

31. Astry C, Warr G, Jakab G. Impairment of polymorphonuclear leukocyte immigration as a mechanism of alcohol induced suppression of pulmonary antibacterial defenses. Am Rev Respir Dis 1983; 128:113-7. 32. Atkinson JP, Sullivan TJ, Parker CWo Stimulation by alcohols of cyclic AMP metabolism in human leukocytes. J Clin Invest 1977;60:284-90. 33. Bainton D, Ulloyot JL, Farquhar MG. The development of neutrophilic polymorphonuclear leukocytesin human bone marrow. Origin and content of azurophil and specific granules. J Exp Med 1971; 134:907-34. 34. Collins SJ, Ruscetti FW, Gallagher RE, Gallo RC. Normal functional characteristics of cultured human promyelocyticleukemia cells(HL-60) after induction of differentiation by dimethylsulfoxide. J Exp Med 1979; 149:969-74. 35. Newburger PE, Chovaniec ME, Greenberger JS, Cohen HJ. Functional changes in human leukemic cell line HL-60. A model for myeloid differentiation. J Cell Bioi 1979; 82:315-22. 36. Harris P, Ralph P. Human leukemic models of myelomonocytic development: a reivew of the HL-60 and U937 cell lines. J Leuk Bioi 1985; 37:407-22. 37. Sachs CW, Christianson RC, Lynn WS, Pratt PC. Ethanol decreases neutrophil elastase activity in differentiating HL-60 promyelocytesand elastase deficient neutrophils are produced in alcoholics. (abstract). Fed Proc 1987; 46:534. 38. Smith CD, Snyderman R. Guanine nucleotide regulatory proteins in receptor-mediated polyphosphoinositide hydrolysisin human leukocytes.Methods Enzymol 1987; 141:261-74. 39. Lynn WS, Sachs CW, Jacobs AR, Lynn DG, Phillips NG. Source and function of keto elaidic acids from lungs of cotton workers. Arch Environ Health 1986; 41:197-207. 40. Yamada M, Kurahashi K. Regulation of myeloperoxidase gene expression during differentiation of human myeloid leukemia HL-60 cells. J Bioi Chern 1984; 259:3021-5. 41. Nakajima K, Powers JC, Ashe BM, Zimmerman M. Mapping the extended substrate binding site of cathepsin G and human leukocyte elastase. J Bioi Chern 1979; 254:4027-32. 42. Iman RL, Conover WJ. In: A modern approach to statistics. New York: John Wiley and Sons, 1983; 35. 43. Noether GE. The Wilcoxon two-sample test. In: Noether GE, ed. Introduction to statistics: a nonparametric approach. 2nd ed. Boston: Houghton Mifflin, 1976; 262-70. 44. Friedenberg WR, Marx JJ. The effect of lithium carbonate on lymphocyte, granulocyte and platelet function. Cancer 1980; 45:91-7. 45. Finch SC. Granulocytopenia. In: WilliamsWJ, Beutler E, Erslev AJ, Rundles RW, eds. Hematology. New York: McGraw-Hili, 1981; 732-72. 46. Pratt PC. Role of conventional chest radiography in diagnosis and exclusion of emphysema. Am J Med 1987; 82:998-1006. 47. Wewers M. Pathogenesis of emphysema. Assessment of basic science concepts through clinical investigation. Chest 1989; 95:190-5. 48. Reynolds HY, Merrill WW. Airway changes in young smokers that may antedate chronic obstructive lung disease. Med Clin North Am 1981; 65:667-89. 49. Ludwig PW, Schwartz BA, Hoidal JR, Niewoehner DE. Cigarette smoking causes accumulation of polymorphonuclear leukocytes in alveolar septum. Am Rev Respir Dis 1985; 131:828-30. 50. Janoff A, Sloan B, Weinbaum G, et al. Experimental emphysema induced with purified human neutrophil elastase: tissue localization of the

DECREASED ELASTASE ACTIVITY AND SUPEROXIDE PRODUCTION IN PMN OF ALCOHOLICS

instilled protease. Am Rev Respir Dis 1977; 127: 540-4. 51. Snider GL, Lucey EC, Christensen TO, et at. Emphysema and bronchial secretory cell metaplasia induced in hamsters by human neutrophil products. Am Rev Respir Dis 1984; 129:155-60. 52. Hoidal JR, Fox RB, LeMarbre PA, Perri R, Repine JE. Altered oxidative metabolic responses in vitro of alveolar macrophages from asymptomatic cigarette smokers. Am Rev Respir Dis 1981; 123:85-9. 53. Rodriguez Rl, WhiteRR, SeniorRM, Levine EA. Elastase release from human alveolar macrophages: comparison between smokers and nonsmokers. Science 1977; 198:313-4. 54. Barry RE, McGivan JD, Haynes M. Acetalde-

hyde binds to liver cell membranes without affecting membrane function. Gut 1984; 25:412-6. 55. Williams AJK, Barry RE. Superoxide anion production and degranulation of rat neutrophils in response to acetaldehyde-altered liver cell membranes. Clin Sci 1986; 71:313-8. 56. Walker RI, Willemze R. Neutrophil kinetics and the regulation of granulopoiesis. Rev Infect Dis 1980; 2:282-92. 57. Takahashi H, Toshihiro N, Basset P, Crystal RG. Myelomonocytic cell lineage expression of the neutrophil elastase gene. J Bioi Chern 1988; 263: 2543-7. 58. Liu, YK. Leukopenia in alcoholics. Am J Med 1978; 54:605-10. 59. Lundin LR, Hallgren R, Venge P. Sequential

1255 studies on serum-levels of lysozyme, lactoferrin, and eosinophil cationic protein in alcoholics after alcohol withdrawal. Scand J Haematol 1980; 25: 431-8. 60. Zakhireh B, Root RK. Development of oxidase activity by human bone marrow granulocytes. Blood 1979; 54:429-39. 61. Pratt PC. Emphysema and chronic airways disease. In: Dail DH, Hamner SH, eds. Pulmonary pathology. New York: Springer-Verlag, 1987;651-70. 62. Garshick E, Segal MR, Worbec TO, Salekin CMS, Miller MJ. Alcohol consumption and chronic obstructive pulmonary disease. Am Rev Respir Dis 1989; 140:373-8.