Evaluation of Elastase and Antielastase Balance in Patients with Chronic Bronchitis and Pulmonary Emphysema1 - 3
JIRO FUJITA, NICK L. NELSON, DAVID M. DAUGHTON, CHARLES A. DOBRY, JOHN R. SPURZEM, SHOZO IRINO, and STEPHEN I. RENNARD
Introduction
Pulmonary emphysema is a destructive disease of the peripheral lung that causes progressiveloss of functional alveoli.Severallines of evidence suggest that the activity of proteolytic enzymes, particularly neutrophil elastase, plays an important pathogenetic role in the development of emphysema: (1) instillation of neutrophil elastase can induce emphysema in the lungs of animals (1, 2); (2) the destructive process in emphysema is characterized by the destruction of elastin and the loss of elastic recoil and elastase is capable of hydrolyzing most connective tissue components, including lung elastin (3); (3) persons with genetic deficiency of a major inhibitor of neutrophil elastase, alpha-l-protease inhibitor (alpha-lPI) are particularly susceptible to the development of emphysema (4). Taken together, these observations suggest the "protease-antiprotease imbalance" concept of emphysema in which lung destruction results from an excess of elastinolytic activity (5-7). In this regard, the emphysema that develops in smokers is thought to result in part from a combination of increased elastinolytic activity caused by recruitment of neutrophils into the lower respiratory tract and to loss of elastase inhibitory capacity, perhaps because of partial inactivation of antielastase capacity caused by oxidation or degradation of alpha-l-Pl, This combination of factors is thought to lead to at least a transient excessof elastinolytic activity and hence to lung destruction (8). We evaluated this hypothesis in a group of smokers with chronic obstructive pulmonary disease who uniformly had bronchitis but had variable degrees of emphysema. The relationship between the severity of emphysema as assessed by thin- section computed tomography and diffusion capacity and both the elastase burden and the elastase inhibitory capacity was determined. This study suggests that the severity of emphysema correlates
SUMMARY On the basis of the "protease-antiprotease Imbalance" theory for the pathogenesis of pulmonary emphysema, we hypothesized that measurement of elastase burden and anti elastase capacity In the alveolar space might correlate with emphysema. To evaluate this, the severity of emphysema, the elastase burden, and the elastase Inhibitory capacity were estimated In 28 patients with chronic bronchitis and variable degrees of emphysema, none of whom had congenital deficiency of alpha-1-protease Inhibitor, and all of whom underwent bronchoalveolar lavage. Emphysema was assessed by both computed tomography and diffusing capacity. Toexamine "elastase burden," elastase:alpha-1-protease Inhlbltlor complex and free elastase activity In alveolar lavage fluids were measured. To evaluate "antlelastase" capacity, elastase Inhibiting capacity In alveolar lavage fluid was measured. Elastase burden correlated directly and antlelastase capacity correlated Inversely with emphysema. These data provide direct support for the "protease-antiprotease Imbalance" theory of emphysema In a group of smokers without congenital deficiency of alpha-1-protease Inhibitor. AM REV RESPIR DIS 1990; 142:57-62
directly with the elastase burden and inversely with the elastase inhibitory capacity. Methods Study Population Twenty-nine subjects with chronic bronchitis were studied. All had chronic cough and sputum production for at least 6 months of the year. All of them were without a history of asthma or other atopic disease. None of the subjects was deficient in alpha-l-Pl, Patients with active tuberculosis, a history of pulmonary resection, lung cancer, or a pulmonary infection in the last 3 months were excluded. None of the subjects had clinical, radiologic, or pulmonary function abnormalities suggestive of interstitial lung disease, recurrent pulmonary emboli, or other causes for a reduced diffusion capacity. All had to have an FEV l < 750/0 predicted or an FEVl/FVC ratio < 75%. All subjects were studied under University of Nebraska Institutional Review Board approved guidelines after informed consent had been obtained. History, physical examination, and spirometric data were collected.
Assessment of Emphysema To evaluate the severity of emphysema, two methods were used: (1) computed tomography (CT) of the chest (9-11) and (2) measurement of the diffusion capacity of the lung for carbon monoxide using a single breath technique (12).
Serial CT scans were obtained at a section thickness of 1.5 mm using a 9800 CT unit (General Electric, Milwaukee, WI). Thin 1.5mm sections better display regional anatomy than do standard to-mm-thick sections because of improved spatial resolution (13-16). These images were obtained halfway between the apex of the lung and the carina, at the carina, and halfway between the dome of the right diaphragm and the carina during endinspiration. These were evaluated at several window and level settings and were photographed at a window of 2,000 Hounsfield units (HU) and at the levelof -600 HU. From the computer display module, region-of-interest lines were manually drawn around the lung periphery, and a density measurement was calculated for that section. A similar line was then drawn around the contralateral lung at the same level, and these two numbers were (Received in original form June 14, /989 and in revised form January 5, 1990) 1 From the Pulmonary and Critical Care Section, Department of Internal Medicine, and Department of Radiology, University of Nebraska Medical Center, Omaha, Nebraska; and the First Department of Internal Medicine,Kagawa Medical School, Kagawa, Japan. 2 Supported in part by a grant from the Boehringer Ingelheim Pharmaceutical Company. 3 Correspondence and requests for reprints should be addressed to Stephen I. Rennard, Pulmonary and Critical Care Section, Department of Internal Medicine, University of Nebraska Medical Center, Omaha, NE 68105.
57
58
averaged. Finally, all three levels were averaged for each individual patient. In addition to quantitative analysis of attenuation in Hounsfield units, emphysema was also assessed by two radiologists using a quantitative scoring system that independently rated CT findings based upon a five-point scale (1 = normal, 2 = mild, 3 = moderate, 4 = severe,5 = very severe) based on estimating the percentage of the overall area that demonstrated emphysematous or bullous changes.
Sampling of the Lower Respiratory Tract To sample the lower respiratory tract, flexible fiberoptic bronchoscopy and bronchoalveolar lavage were performed (17).All procedures were performed transorally after topical anesthesia with lidocaine. Premedication consisted of atropine, an inhaled beta-agonist, and sedation with diazepam. There were no complications, although one subject asked that the procedure be terminated because of anxiety. Bronchoalveolar lavage was performed by infusing five 20-ml aliquots of sterile saline at each of three sites (300 ml total). The last four aliquots at each site, representing predominantly alveolar material, were saved for evaluation. Cells were separated from alveolar lavage fluid by centrifugation (300 g for 5 min), and fluids were frozen at -80° C until used for assay. Cell differentials wereevaluated using a modified WrightGiemsa stain after cytocentrifuge preparation. Assessment of Elastase Burden and Antielastase Capacity To evaluate elastase burden in the alveolar lavagefluid, two methods wereused. Freeelastase activity in alveolar lavage fluid was measured using a synthetic substrate, methoxysuccinyl-L-alanyl-L-alanyl-L-prolyl-L-valine p-nitroanilide (Lot 36F-58851; Sigma Chemical Co., St. Louis, MO) (18). One hundred microliters of alveolar lavage fluid wereincubated with 200 IIIof 0.2 mM of substrate in 0.1 M HEPES, 0.5 M NaCI, and 10070 dimethylsulfoxide at pH 7.5. After incubation for 1 h at 37° C, absorbance of the product, p-nitroanilide, was measured at 414 nm. Purified human neutrophil elastase (Lot 87538; Elastin Products Co., Pacific, MO) was used as a standard. Because the majority of elastase might be expected to be inactivated by complexing with alpha-l-Pl, elastase:alpha-l-PI complex in alveolar lavage fluids was measured using a sandwich-type enzyme-linked immunosorbent assay (ELISA) (19, 20). Standard elastase:alpha-l-PI complex was made by incubating 10 ug/ml (final concentration) of human neutrophil elastase and 50 ug/ml (final concentration) of purified plasma alpha-l-Pl (Lot I04F-9400; Sigma) for 1 h at 37° C in 10ml of PBS-Tween. Under these conditions, neutrophil elastase was totally inactivated by complex formation as determined by enzyme assay using a synthetic elastase substrate. This standard solution was defined as 10 ug/ml of complexed elastase. 1\\'0 hundred microliters of goat antihuman elastase (Lot 405708, diluted 1:3,200 with Vollers' buffer; Calbi-
FWITA, NELSON, DAUGHTON, DOBRY, SPURZEM, IRINO, AND RENNARD
was found to give reliable results when assayed in inhibitor excess. However, when assayed in the presence of elastase excess, a decreasing amount of complex was detected (figure 1). For this reason, in the absence of active elastase, the elastase burden was determined as the amount of elastase detected as complex. In contrast, if free enzymatically active elastase was detected, the elastase burden was determined as the amount of active elastase. Only three of 28 patients had free elastase activity detected. In contrast, 12 of 28 patients had immunologically detectable elastase:alpha-l-PI complex. This evidence suggests that measurement of elastase:alpha-l-PI complex may be a more sensitive marker to detect elastase burden than free elastase, because elastase:alpha-l-PI complex can be detected even in antiprotease excess. In two of three patients who had free elastase activity, elastase:alpha-l-PI complex was not detected. This may be due to the ability of excess amounts of neutrophil elastase to degrade elastase.alpha-l-Pl complex or to the inhibition of the measurement of complex by free elastase. Interestingly, these two subjects were brothers who had high levels of alveolar neutrophilia. Thus, it appears to be necessary to measure both free elastase activity and complex to evaluate elastase burden. Elastase burden correlated significantly with the severity of emphysema (figure 2). This was true when emphysema severity was assessed by diffusion capacity (figure 2A, r = -0.559, p < 0.01) or by CT, either using the radiologists score (figure 2B, r = 0.578, p < 0.05) or using Hounsfield units (figure 2C, r = - 0.42, p < 0.001).
ochem Brand Biochemicals, San Diego, CA) was used to coat each well of the micro ELISA plate. After rinsing the plates coated with antihuman elastase, 200 IIIof unknown solutions and standard solutions (10 ug/ml of complexed elastase and its serial dilution) were added to each well and incubated for 30 min at room temperature. The unbound material was rinsed away, and rabbit antihuman alphaI-PI (Lot 6A-5215, 1:500; Behring, La Jolla, CA) was added and incubated for 30 min at room temperature. After rinsing, peroxidaseconjugated goat antirabbit IgG (Lot E 595; ICN ImmunoBiologicals, Lisle, IL) was added and incubated for 120 min at room temperature followedby peroxidase substrate, and the resulting chromophore produced was measured at 492 nm. This sandwich ELISA technique was sensitive to 5 ng/ml elastase: aiphal-PI complex. In order to determine the behavior of this assay over a range of concentrations of elastase and protease inhibitor, a series of experiments was performed with a variety of conditions. In order to correct for variable dilution, albumin concentration was also determined in the bronchoalveolar lavage fluid by ELISA (21). To determine the capacity of the lung to protect itself against neutrophil elastase, the total elastase inhibiting capacity in alveolar lavage was quantified. The total functional inhibitors were titrated by their ability to inhibit the activity of neutrophil elastase. Purified human plasma alpha-l-Pl (Sigma) was used as standard. One hundred microliters of . serial dilutions of alveolar lavage fluid or, for comparison with a known elastase inhibitor, standard alpha-I-PI concentrations were incubated with 100 III of human neutrophil elastase (l ug/ml) for 1 h. After incubation, residual elastase activity was measured as described above. For comparison, the total alphaI-PI concentration was also determined by ELISA.
Statistics All statistical correlations were computed using Spearman's nonparametric correlation coefficient.
Antielastase Protection Direct quantification of antielastase protection revealed marked heterogeneity in this population. Of the three patients with active elastase, two had no detectable inhibitory capacity, and a third, the sole
Results
Elastase Burden The ELISA to quantify elastase complex 2.00 1.75
W
Fig. 1 Effect of varying concentrations of elastase and alpha-1-protease inhibitor on complex detected by ELISA. Elastase in varying concentrations was added to solutions of alpha-1-protease inhibitor either 33 nM (circles) or 160 nM (triangles). Verticle axis: optical density in the ELISA; horizontal axis: molarity of added elastase. The amount of complex detected decreases under conditions of elastase excess.
1.50
o Z
c(
1.25
to
a: 0 en to -e
1.00 0.75 0.50
....
0.25 .025
.25
2.5
25
250
CONCENTRA TION OF ELASTASE (nM)
2500
ELASTASE AND ANTIELASTASE BALANCE IN BRONCHITIS AND EMPHYSEMA
loo1
B .
100lA.
~I
..
.. c
3
"0
•·· =:
10 5
10 5
I c
2
~
!;
D
.
··
:
.
2
~
'!
'ii
'ii
0
:~I
20
10 5
"0
100lC,
~r
0
20
c
59
2
1
o o
o
0
o
o
o +-O--~-CX>-~>-
JIRO FUJITA, NICK L. NELSON, DAVID M. DAUGHTON, CHARLES A. DOBRY, JOHN R. SPURZEM, SHOZO IRINO, and STEPHEN I. RENNARD
Introduction
Pulmonary emphysema is a destructive disease of the peripheral lung that causes progressiveloss of functional alveoli.Severallines of evidence suggest that the activity of proteolytic enzymes, particularly neutrophil elastase, plays an important pathogenetic role in the development of emphysema: (1) instillation of neutrophil elastase can induce emphysema in the lungs of animals (1, 2); (2) the destructive process in emphysema is characterized by the destruction of elastin and the loss of elastic recoil and elastase is capable of hydrolyzing most connective tissue components, including lung elastin (3); (3) persons with genetic deficiency of a major inhibitor of neutrophil elastase, alpha-l-protease inhibitor (alpha-lPI) are particularly susceptible to the development of emphysema (4). Taken together, these observations suggest the "protease-antiprotease imbalance" concept of emphysema in which lung destruction results from an excess of elastinolytic activity (5-7). In this regard, the emphysema that develops in smokers is thought to result in part from a combination of increased elastinolytic activity caused by recruitment of neutrophils into the lower respiratory tract and to loss of elastase inhibitory capacity, perhaps because of partial inactivation of antielastase capacity caused by oxidation or degradation of alpha-l-Pl, This combination of factors is thought to lead to at least a transient excessof elastinolytic activity and hence to lung destruction (8). We evaluated this hypothesis in a group of smokers with chronic obstructive pulmonary disease who uniformly had bronchitis but had variable degrees of emphysema. The relationship between the severity of emphysema as assessed by thin- section computed tomography and diffusion capacity and both the elastase burden and the elastase inhibitory capacity was determined. This study suggests that the severity of emphysema correlates
SUMMARY On the basis of the "protease-antiprotease Imbalance" theory for the pathogenesis of pulmonary emphysema, we hypothesized that measurement of elastase burden and anti elastase capacity In the alveolar space might correlate with emphysema. To evaluate this, the severity of emphysema, the elastase burden, and the elastase Inhibitory capacity were estimated In 28 patients with chronic bronchitis and variable degrees of emphysema, none of whom had congenital deficiency of alpha-1-protease Inhibitor, and all of whom underwent bronchoalveolar lavage. Emphysema was assessed by both computed tomography and diffusing capacity. Toexamine "elastase burden," elastase:alpha-1-protease Inhlbltlor complex and free elastase activity In alveolar lavage fluids were measured. To evaluate "antlelastase" capacity, elastase Inhibiting capacity In alveolar lavage fluid was measured. Elastase burden correlated directly and antlelastase capacity correlated Inversely with emphysema. These data provide direct support for the "protease-antiprotease Imbalance" theory of emphysema In a group of smokers without congenital deficiency of alpha-1-protease Inhibitor. AM REV RESPIR DIS 1990; 142:57-62
directly with the elastase burden and inversely with the elastase inhibitory capacity. Methods Study Population Twenty-nine subjects with chronic bronchitis were studied. All had chronic cough and sputum production for at least 6 months of the year. All of them were without a history of asthma or other atopic disease. None of the subjects was deficient in alpha-l-Pl, Patients with active tuberculosis, a history of pulmonary resection, lung cancer, or a pulmonary infection in the last 3 months were excluded. None of the subjects had clinical, radiologic, or pulmonary function abnormalities suggestive of interstitial lung disease, recurrent pulmonary emboli, or other causes for a reduced diffusion capacity. All had to have an FEV l < 750/0 predicted or an FEVl/FVC ratio < 75%. All subjects were studied under University of Nebraska Institutional Review Board approved guidelines after informed consent had been obtained. History, physical examination, and spirometric data were collected.
Assessment of Emphysema To evaluate the severity of emphysema, two methods were used: (1) computed tomography (CT) of the chest (9-11) and (2) measurement of the diffusion capacity of the lung for carbon monoxide using a single breath technique (12).
Serial CT scans were obtained at a section thickness of 1.5 mm using a 9800 CT unit (General Electric, Milwaukee, WI). Thin 1.5mm sections better display regional anatomy than do standard to-mm-thick sections because of improved spatial resolution (13-16). These images were obtained halfway between the apex of the lung and the carina, at the carina, and halfway between the dome of the right diaphragm and the carina during endinspiration. These were evaluated at several window and level settings and were photographed at a window of 2,000 Hounsfield units (HU) and at the levelof -600 HU. From the computer display module, region-of-interest lines were manually drawn around the lung periphery, and a density measurement was calculated for that section. A similar line was then drawn around the contralateral lung at the same level, and these two numbers were (Received in original form June 14, /989 and in revised form January 5, 1990) 1 From the Pulmonary and Critical Care Section, Department of Internal Medicine, and Department of Radiology, University of Nebraska Medical Center, Omaha, Nebraska; and the First Department of Internal Medicine,Kagawa Medical School, Kagawa, Japan. 2 Supported in part by a grant from the Boehringer Ingelheim Pharmaceutical Company. 3 Correspondence and requests for reprints should be addressed to Stephen I. Rennard, Pulmonary and Critical Care Section, Department of Internal Medicine, University of Nebraska Medical Center, Omaha, NE 68105.
57
58
averaged. Finally, all three levels were averaged for each individual patient. In addition to quantitative analysis of attenuation in Hounsfield units, emphysema was also assessed by two radiologists using a quantitative scoring system that independently rated CT findings based upon a five-point scale (1 = normal, 2 = mild, 3 = moderate, 4 = severe,5 = very severe) based on estimating the percentage of the overall area that demonstrated emphysematous or bullous changes.
Sampling of the Lower Respiratory Tract To sample the lower respiratory tract, flexible fiberoptic bronchoscopy and bronchoalveolar lavage were performed (17).All procedures were performed transorally after topical anesthesia with lidocaine. Premedication consisted of atropine, an inhaled beta-agonist, and sedation with diazepam. There were no complications, although one subject asked that the procedure be terminated because of anxiety. Bronchoalveolar lavage was performed by infusing five 20-ml aliquots of sterile saline at each of three sites (300 ml total). The last four aliquots at each site, representing predominantly alveolar material, were saved for evaluation. Cells were separated from alveolar lavage fluid by centrifugation (300 g for 5 min), and fluids were frozen at -80° C until used for assay. Cell differentials wereevaluated using a modified WrightGiemsa stain after cytocentrifuge preparation. Assessment of Elastase Burden and Antielastase Capacity To evaluate elastase burden in the alveolar lavagefluid, two methods wereused. Freeelastase activity in alveolar lavage fluid was measured using a synthetic substrate, methoxysuccinyl-L-alanyl-L-alanyl-L-prolyl-L-valine p-nitroanilide (Lot 36F-58851; Sigma Chemical Co., St. Louis, MO) (18). One hundred microliters of alveolar lavage fluid wereincubated with 200 IIIof 0.2 mM of substrate in 0.1 M HEPES, 0.5 M NaCI, and 10070 dimethylsulfoxide at pH 7.5. After incubation for 1 h at 37° C, absorbance of the product, p-nitroanilide, was measured at 414 nm. Purified human neutrophil elastase (Lot 87538; Elastin Products Co., Pacific, MO) was used as a standard. Because the majority of elastase might be expected to be inactivated by complexing with alpha-l-Pl, elastase:alpha-l-PI complex in alveolar lavage fluids was measured using a sandwich-type enzyme-linked immunosorbent assay (ELISA) (19, 20). Standard elastase:alpha-l-PI complex was made by incubating 10 ug/ml (final concentration) of human neutrophil elastase and 50 ug/ml (final concentration) of purified plasma alpha-l-Pl (Lot I04F-9400; Sigma) for 1 h at 37° C in 10ml of PBS-Tween. Under these conditions, neutrophil elastase was totally inactivated by complex formation as determined by enzyme assay using a synthetic elastase substrate. This standard solution was defined as 10 ug/ml of complexed elastase. 1\\'0 hundred microliters of goat antihuman elastase (Lot 405708, diluted 1:3,200 with Vollers' buffer; Calbi-
FWITA, NELSON, DAUGHTON, DOBRY, SPURZEM, IRINO, AND RENNARD
was found to give reliable results when assayed in inhibitor excess. However, when assayed in the presence of elastase excess, a decreasing amount of complex was detected (figure 1). For this reason, in the absence of active elastase, the elastase burden was determined as the amount of elastase detected as complex. In contrast, if free enzymatically active elastase was detected, the elastase burden was determined as the amount of active elastase. Only three of 28 patients had free elastase activity detected. In contrast, 12 of 28 patients had immunologically detectable elastase:alpha-l-PI complex. This evidence suggests that measurement of elastase:alpha-l-PI complex may be a more sensitive marker to detect elastase burden than free elastase, because elastase:alpha-l-PI complex can be detected even in antiprotease excess. In two of three patients who had free elastase activity, elastase:alpha-l-PI complex was not detected. This may be due to the ability of excess amounts of neutrophil elastase to degrade elastase.alpha-l-Pl complex or to the inhibition of the measurement of complex by free elastase. Interestingly, these two subjects were brothers who had high levels of alveolar neutrophilia. Thus, it appears to be necessary to measure both free elastase activity and complex to evaluate elastase burden. Elastase burden correlated significantly with the severity of emphysema (figure 2). This was true when emphysema severity was assessed by diffusion capacity (figure 2A, r = -0.559, p < 0.01) or by CT, either using the radiologists score (figure 2B, r = 0.578, p < 0.05) or using Hounsfield units (figure 2C, r = - 0.42, p < 0.001).
ochem Brand Biochemicals, San Diego, CA) was used to coat each well of the micro ELISA plate. After rinsing the plates coated with antihuman elastase, 200 IIIof unknown solutions and standard solutions (10 ug/ml of complexed elastase and its serial dilution) were added to each well and incubated for 30 min at room temperature. The unbound material was rinsed away, and rabbit antihuman alphaI-PI (Lot 6A-5215, 1:500; Behring, La Jolla, CA) was added and incubated for 30 min at room temperature. After rinsing, peroxidaseconjugated goat antirabbit IgG (Lot E 595; ICN ImmunoBiologicals, Lisle, IL) was added and incubated for 120 min at room temperature followedby peroxidase substrate, and the resulting chromophore produced was measured at 492 nm. This sandwich ELISA technique was sensitive to 5 ng/ml elastase: aiphal-PI complex. In order to determine the behavior of this assay over a range of concentrations of elastase and protease inhibitor, a series of experiments was performed with a variety of conditions. In order to correct for variable dilution, albumin concentration was also determined in the bronchoalveolar lavage fluid by ELISA (21). To determine the capacity of the lung to protect itself against neutrophil elastase, the total elastase inhibiting capacity in alveolar lavage was quantified. The total functional inhibitors were titrated by their ability to inhibit the activity of neutrophil elastase. Purified human plasma alpha-l-Pl (Sigma) was used as standard. One hundred microliters of . serial dilutions of alveolar lavage fluid or, for comparison with a known elastase inhibitor, standard alpha-I-PI concentrations were incubated with 100 III of human neutrophil elastase (l ug/ml) for 1 h. After incubation, residual elastase activity was measured as described above. For comparison, the total alphaI-PI concentration was also determined by ELISA.
Statistics All statistical correlations were computed using Spearman's nonparametric correlation coefficient.
Antielastase Protection Direct quantification of antielastase protection revealed marked heterogeneity in this population. Of the three patients with active elastase, two had no detectable inhibitory capacity, and a third, the sole
Results
Elastase Burden The ELISA to quantify elastase complex 2.00 1.75
W
Fig. 1 Effect of varying concentrations of elastase and alpha-1-protease inhibitor on complex detected by ELISA. Elastase in varying concentrations was added to solutions of alpha-1-protease inhibitor either 33 nM (circles) or 160 nM (triangles). Verticle axis: optical density in the ELISA; horizontal axis: molarity of added elastase. The amount of complex detected decreases under conditions of elastase excess.
1.50
o Z
c(
1.25
to
a: 0 en to -e
1.00 0.75 0.50
....
0.25 .025
.25
2.5
25
250
CONCENTRA TION OF ELASTASE (nM)
2500
ELASTASE AND ANTIELASTASE BALANCE IN BRONCHITIS AND EMPHYSEMA
loo1
B .
100lA.
~I
..
.. c
3
"0
•·· =:
10 5
10 5
I c
2
~
!;
D
.
··
:
.
2
~
'!
'ii
'ii
0
:~I
20
10 5
"0
100lC,
~r
0
20
c
59
2
1
o o
o
0
o
o
o +-O--~-CX>-~>-
