Bradykinin Causes Airway Hyperresponsiveness and Enhances Maximal Airway Narrowing Role of Microvascular Leakage and Airway Edema 1 , 2
KEIJI KIMURA, HIROSHI INOUE, MASAKAZU ICHINOSE, MOTOHIKO MIURA, UICHIRO KATSUMATA, TSUNEYUKI TAKAHASHI, and TAMOTSU TAKISHIMA
Introduction
Asthma is characterized by increased airway sensitivity to various stimuli and by a widespread and excessive airway narrowing 0, 2). Asthmatic individuals show a leftward shift of the dose-response curve to histamine or methacholine and progressive airway narrowing with increasing doses of agonists, whereas normal subjects reach a maximal plateau at a mild degree of airway narrowing (3,4). The mechanism of this abnormal airway response in asthma is still unclear. Asthma is recognized to be an inflammatory disorder (5). Increased bronchial vascular permeability and subsequent airway wall thickening (airway edema) are important components of the inflammatory response (6, 7). It was recently reported that the walls of the airways in postmortem specimens obtained from asthmatic subjects are thicker than those of nonasthmatics (8). Theoretically, moderate increases in airway wall thickness that have little effect on baseline resistance can produce a leftward shift of the dose-response curve and a great increase in maximal airway narrowing (9), which is similar to that observed in asthmatic subjects. Furthermore, peribronchial edema could diminish the elastic recoil pressure applied to the external surface of airway walls and may cause airway hypersensitivity. However, there is little direct evidence of the relationship between bronchial edema and abnormal airway response to agonists. Bradykinin (BK) is formed from plasma precursors as part of the inflammatory response and may be an important mediator of inflammatory airway diseases such as asthma (10).BK causes airway microvascular leakage (11) and subsequent airway edema. The aim of this study was to investigate the effect of airway edema on the acetylcholine (ACh)
SUMMARY The relationship between bronchial edema and airway responsiveness was studied in cats in situ. Five cats were exsanguinated, and the bronchial arteries were perfused. We monitored pulmonary resistance (RL), and the provocative dose of acetylcholine (ACh) required to produce a 300% Increase in RL (P0 300 ) was determined. Bronchial vascular permeability was measured by quantifying extravasation of Evans blue (EB) dye. Bradykinin (BK) and ACh were administered via the bronchial arteries to Increase leakage and bronchoconstrlctlon, respectively. BK preperfusion (for 30 min) significantly Increased bronchial vascular permeability to four times the control values (p < 0.05). BK preperfuslon did not alter baseline RL but caused hyperresponslveness to ACh, with log [P0 300 (mole)) of -6.53 ± 0.42 (mean ± SO) and -6.90 ± 0.30, before and after BK, respectively (p < 0.01). Furthermore, the maximal airway narrowing after BK was 58% higher than before BK (p < 0.01). Histologic study showed peribronchial edema after BK. The enhancement of maximal airway narrowing was significantly correlated with the degree of EB dye extravasation. These results suggest that BK causes airway hyperresponslveness to ACh and increases maximal airway narrowing, possibly because of airway edema. AM REV RESPIR DIS 1992; 146:1301-1305
dose-response curve in airways. We employed a bronchial artery perfusate model and used BK to cause airway edema. Methods Preparations Eight cats weighing2.0 to 3.0 kg wereanesthetized intramuscularly with thiopental (30 to 50 mg/kg) and ketamine hydrochloride (30 mg/kg). An endotracheal tube (outer diameter, 6 mm; length 5 em; R = 3.3 cm H 20/L/s) was inserted, and the animals were paralyzed with succinylcholine administered intravenously (25 mg/kg). The lungs were artificially ventilated with a respirator (Model 661; Harvard Apparatus, South Natick, MA) at a tidal volume of 15 to 20 mllkg and a frequency of 15 breaths/min, as previously reported (12-14). The chest was opened, and each cat was exsanguinated from a polyethylene tube inserted into the femoral artery. A glass cannula (outer diameter, 4 mm; inner diameter, 3 mm) was introduced into the aortic arch past the left subclavian artery. The thoracic aorta was tied at two points, the point between the fourth and fifth intercostal arteries, and between the ninth and tenth intercostal arteries, respectively. To isolate the bronchial arteries, the bilateral fifth to ninth intercostal arteries were also tied about 2 em
from their sources. Tyrode solution with 2.5070 dextran 40 (NaC1137mM, KC12.7 mM, MgCI2·6H20 1.0 mM, CaCb·2H20 1.8 mM, NaHC03 11·9 mM, NaH2P04·2H20 0.4 mM, glucose 5.6 mM) bubbled with 95%02 5%C0 2 gas was continuously perfused into the lung through the bronchial arteriesat a rate of 4 mllmin with a syringepump (Model STC521; Terumo, Inc., Terumo, Japan). Dextran 40 was used to keep the colloid osmotic pressure in the Tyrode solution the same as that in the blood. In order to exclude the effect of cyclooxygenase products, the Tyrode solution contained indomethacin 10-6 M. Both ventricles were cut at their midportion to prevent pulmonary congestion. To recover the perfusate, a polyethylene tube was inserted into the azygos vein. The superior and inferior vena cava were also tied. The cervical
(Received in original form September 19, 1991 and in revised form April 24, 1992) 1 From the First Department of Internal Medicine, Tohoku University School of Medicine, Sendai, Japan. 2 Correspondence and requests for reprints should be addressed to Tamotsu Takishima, M.D., First Department of Internal Medicine, Tohoku University School of Medicine, 1-1 Seiryo-machi, Aoba-ku, Sendai 980, Japan.
1301
1302
vagosympathetic trunks were cut bilaterally to prevent vagal reflex bronchoconstriction. A catheter (inner diameter, 1 mm; length, 7 em) was inserted into the distal end of the endotracheal tube to obtain airway pressure~ and connected to a differential pressure transducer (MP 45 ± 50 cm H 2 0 ; Validyne,Northridge, CA). Airflow at the opening of the endotracheal tube was measured with a heated pneumotachograph (Fleisch No. 00) and a differential pressure transducer (MP 45 ± 5 cm H 20 ). Transpulmonary pressure, airflow, and flow-integrated volume were recorded continuously during the experiments with a polygraph (San-ei 8s; San-ei, Tokyo, Japan). As an index of airway caliber, pulmonary resistance (RL) was calculated by a subtraction method (15).
Measurement of Airway Responsiveness ACh (2.8 x 10-8 - 2.8 X 10-4 mol) was administered to the airway via the bronchial arteries. RL was monitored, and the mean consecutive peak value was determined. After RL had returned to the baseline value (about 5 min), this sequence was repeated for each of the incremental concentrations of ACh until no further significant increase in RL was observed. In one group of cats (n = 5), an ACh doseresponse curve was first obtained. Then, in order to increase airway vascular permeability, Tyrode solution with 2.5070 dextran 40 containing 10-7 M BK was perfused to the lung through the bronchial arteries for 30min. This concentration of BK was the maximal one that did not affect the RL in this model. To wash out BK, Tyrode solution without BK was perfused for 10 min. Then, a second ACh doseresponse curve was obtained. In the other group of cats (n = 3), to test the reproducibility in this perfused bronchial arterial model, ACh challenges were repeated two times in the same way described previously except for the BK perfusion. Sham perfusion (Tyrode solution without BK) was performed for 40 min between the first and second ACh challenge. Measurement oj Vascular Permeability The degree of vascular permeability was determined by measuring the content of EB dye extravasated into the airway tissue after removing the intravascular EB dye using a modified method previously reported (16). When RL had returned to the baseline after the second ACh challenge, 10 mg/kg of EB dye (2% solution containing 10 g/dl bovine serum albumin) (17)was administered to the airway via the bronchial arteries. Tenminutes after EB, 150ml of Tyrode solution with 2.5% dextran 40 was perfused to the bronchial arteries at 150 mm Hg pressure to remove intravascular dye. Twosegmental bronchi of the left caudal (diaphragmatic) lobe in each cat were excised and incubated in 4 ml of 100% formamide at 54 0 C for 24 h after measuring wet weight. The concentration of EB dye extracted from the airway tissue was determined by light absorbance at 620 nm wavelength
KIMURA, INOUE, ICHINOSE, ET AL.
Fig. 1. Reproducibility of ACh doseresponse curves in the present model. ACh dose-response curves before and after 40 min sham perfusion in each cat are shown. Open circles indicate the first ACh challenge and closed circles the second one.
S
-8
-7
-6
-5
-4
Acetylcholine
(log (mol»)
Fig. 2. ACh dose-response curves before (open circles) and after (closed circles) BK infusion in each cat.
Acetyldtollne [log (mol»)
(220A spectrophotometer; Hitachi, Tokyo,Japan) and by interpolation on a standard curve of EB dye concentration (0.3-10 ug/ml), EB dye content of each airway was expressed as nanogram per milligram of wet weight tissue.
Histologic Study The right lungs of the cats were frozen with liquid nitrogen at a transpulmonary pressure of 2 em H 2 0 . The frozen lung was transferred to increasing concentrations of celloidin for 4 wk, immersed in xylol for 1 h, then embedded in paraffin, as wepreviously reported (18). They were cut into sections 4 urn thick and stained with Elastica-Masson. We measured airway wall area inside (WAi)and outside the airway smooth muscle (WAo) with a digitalizing tablet coupled to a computer, using the image from a TV monitor connected directly to the microscope (PC-9801;NEC, Tokyo, Japan; LA-5OO; PIAS, Tokyo, Japan) (19). Drugs The following drugs were used: thiopental, ketamine hydrochloride (Tokyo Tanabe, Japan); bradykinin (Peptide Institute Inc., Osaka, Japan); acetylcholine, succinylcholine, indomethacin, Evans blue dye, bovine serum albumin, and formamide (Sigma Chemical Co., St. Louis, MO); dextran 40 (Ohtsuca Pharmaceutical Co., Tokyo, Japan). Statistical Analyses Data are expressed as mean ± SD. To evaluate the sensitivity of the airway response to
ACh, the dose of ACh required to produce a 300% increase in RL (PD 3oo) was determined. In order to estimate the degree of increased maximal airway narrowing after BK, the ratio of RL at maximal plateau of airway narrowing obtained in the second ACh doseresponse curve to that in the first one was calculated. The Student's paired t test was used for statistical evaluation of the data; p < 0.05 was considered significant.
Results
Figure 1 shows the reproducibility of ACh dose-response curves in this model. Good reproducibility of the ACh challenge was obtained in each cat. The ACh dose-response curves from each cat in the BK perfusion group are shown in figure 2. Before BK infusion, ACh administration caused dose-dependent bronchoconstriction and produced a maximal plateau at which RL was 12 times the baseline value. After 10-7 M BK perfusion, the ACh dose-response curve shifted to the left and upward in each case compared with that before BK. Mean log [PD 30 0 (mol)] after BK was significantly lower than that before BK (- 6.90 ± 0.30, - 6.53 ± 0.42, respectively, p < 0.01) (figure 3). Mean RL at maximal plateau of airway narrowing after BK was significantly larger compared with that before
1303
BRONCHIAL EDEMA AND AIRWAY RESPONSIVENESS
was four times higher than that in the control group (147 ± 17.6, 35.5 ± 12.8 ng/mg tissue, respectively) (p < 0.05). The content of EB dye extravasated to the airway tissue and the ratio of RL at maximal plateau of airway narrowing obtained in the second ACh dose-response curve to that in the first one was significantly correlated (r = 0.85, p < 0.05) (figure 5). Histologic study showed peribronchial edema in the bronchi and bronchiole of the BK perfusion group compared with the saline (vehicle for BK) perfusion group. Neither alveolar wall thickening nor extravasated fluid in alveolar space was observed (figure 6). Airway edema was observed only on the wall area outside the airway smooth muscle (table 1).
P
KEIJI KIMURA, HIROSHI INOUE, MASAKAZU ICHINOSE, MOTOHIKO MIURA, UICHIRO KATSUMATA, TSUNEYUKI TAKAHASHI, and TAMOTSU TAKISHIMA
Introduction
Asthma is characterized by increased airway sensitivity to various stimuli and by a widespread and excessive airway narrowing 0, 2). Asthmatic individuals show a leftward shift of the dose-response curve to histamine or methacholine and progressive airway narrowing with increasing doses of agonists, whereas normal subjects reach a maximal plateau at a mild degree of airway narrowing (3,4). The mechanism of this abnormal airway response in asthma is still unclear. Asthma is recognized to be an inflammatory disorder (5). Increased bronchial vascular permeability and subsequent airway wall thickening (airway edema) are important components of the inflammatory response (6, 7). It was recently reported that the walls of the airways in postmortem specimens obtained from asthmatic subjects are thicker than those of nonasthmatics (8). Theoretically, moderate increases in airway wall thickness that have little effect on baseline resistance can produce a leftward shift of the dose-response curve and a great increase in maximal airway narrowing (9), which is similar to that observed in asthmatic subjects. Furthermore, peribronchial edema could diminish the elastic recoil pressure applied to the external surface of airway walls and may cause airway hypersensitivity. However, there is little direct evidence of the relationship between bronchial edema and abnormal airway response to agonists. Bradykinin (BK) is formed from plasma precursors as part of the inflammatory response and may be an important mediator of inflammatory airway diseases such as asthma (10).BK causes airway microvascular leakage (11) and subsequent airway edema. The aim of this study was to investigate the effect of airway edema on the acetylcholine (ACh)
SUMMARY The relationship between bronchial edema and airway responsiveness was studied in cats in situ. Five cats were exsanguinated, and the bronchial arteries were perfused. We monitored pulmonary resistance (RL), and the provocative dose of acetylcholine (ACh) required to produce a 300% Increase in RL (P0 300 ) was determined. Bronchial vascular permeability was measured by quantifying extravasation of Evans blue (EB) dye. Bradykinin (BK) and ACh were administered via the bronchial arteries to Increase leakage and bronchoconstrlctlon, respectively. BK preperfusion (for 30 min) significantly Increased bronchial vascular permeability to four times the control values (p < 0.05). BK preperfuslon did not alter baseline RL but caused hyperresponslveness to ACh, with log [P0 300 (mole)) of -6.53 ± 0.42 (mean ± SO) and -6.90 ± 0.30, before and after BK, respectively (p < 0.01). Furthermore, the maximal airway narrowing after BK was 58% higher than before BK (p < 0.01). Histologic study showed peribronchial edema after BK. The enhancement of maximal airway narrowing was significantly correlated with the degree of EB dye extravasation. These results suggest that BK causes airway hyperresponslveness to ACh and increases maximal airway narrowing, possibly because of airway edema. AM REV RESPIR DIS 1992; 146:1301-1305
dose-response curve in airways. We employed a bronchial artery perfusate model and used BK to cause airway edema. Methods Preparations Eight cats weighing2.0 to 3.0 kg wereanesthetized intramuscularly with thiopental (30 to 50 mg/kg) and ketamine hydrochloride (30 mg/kg). An endotracheal tube (outer diameter, 6 mm; length 5 em; R = 3.3 cm H 20/L/s) was inserted, and the animals were paralyzed with succinylcholine administered intravenously (25 mg/kg). The lungs were artificially ventilated with a respirator (Model 661; Harvard Apparatus, South Natick, MA) at a tidal volume of 15 to 20 mllkg and a frequency of 15 breaths/min, as previously reported (12-14). The chest was opened, and each cat was exsanguinated from a polyethylene tube inserted into the femoral artery. A glass cannula (outer diameter, 4 mm; inner diameter, 3 mm) was introduced into the aortic arch past the left subclavian artery. The thoracic aorta was tied at two points, the point between the fourth and fifth intercostal arteries, and between the ninth and tenth intercostal arteries, respectively. To isolate the bronchial arteries, the bilateral fifth to ninth intercostal arteries were also tied about 2 em
from their sources. Tyrode solution with 2.5070 dextran 40 (NaC1137mM, KC12.7 mM, MgCI2·6H20 1.0 mM, CaCb·2H20 1.8 mM, NaHC03 11·9 mM, NaH2P04·2H20 0.4 mM, glucose 5.6 mM) bubbled with 95%02 5%C0 2 gas was continuously perfused into the lung through the bronchial arteriesat a rate of 4 mllmin with a syringepump (Model STC521; Terumo, Inc., Terumo, Japan). Dextran 40 was used to keep the colloid osmotic pressure in the Tyrode solution the same as that in the blood. In order to exclude the effect of cyclooxygenase products, the Tyrode solution contained indomethacin 10-6 M. Both ventricles were cut at their midportion to prevent pulmonary congestion. To recover the perfusate, a polyethylene tube was inserted into the azygos vein. The superior and inferior vena cava were also tied. The cervical
(Received in original form September 19, 1991 and in revised form April 24, 1992) 1 From the First Department of Internal Medicine, Tohoku University School of Medicine, Sendai, Japan. 2 Correspondence and requests for reprints should be addressed to Tamotsu Takishima, M.D., First Department of Internal Medicine, Tohoku University School of Medicine, 1-1 Seiryo-machi, Aoba-ku, Sendai 980, Japan.
1301
1302
vagosympathetic trunks were cut bilaterally to prevent vagal reflex bronchoconstriction. A catheter (inner diameter, 1 mm; length, 7 em) was inserted into the distal end of the endotracheal tube to obtain airway pressure~ and connected to a differential pressure transducer (MP 45 ± 50 cm H 2 0 ; Validyne,Northridge, CA). Airflow at the opening of the endotracheal tube was measured with a heated pneumotachograph (Fleisch No. 00) and a differential pressure transducer (MP 45 ± 5 cm H 20 ). Transpulmonary pressure, airflow, and flow-integrated volume were recorded continuously during the experiments with a polygraph (San-ei 8s; San-ei, Tokyo, Japan). As an index of airway caliber, pulmonary resistance (RL) was calculated by a subtraction method (15).
Measurement of Airway Responsiveness ACh (2.8 x 10-8 - 2.8 X 10-4 mol) was administered to the airway via the bronchial arteries. RL was monitored, and the mean consecutive peak value was determined. After RL had returned to the baseline value (about 5 min), this sequence was repeated for each of the incremental concentrations of ACh until no further significant increase in RL was observed. In one group of cats (n = 5), an ACh doseresponse curve was first obtained. Then, in order to increase airway vascular permeability, Tyrode solution with 2.5070 dextran 40 containing 10-7 M BK was perfused to the lung through the bronchial arteries for 30min. This concentration of BK was the maximal one that did not affect the RL in this model. To wash out BK, Tyrode solution without BK was perfused for 10 min. Then, a second ACh doseresponse curve was obtained. In the other group of cats (n = 3), to test the reproducibility in this perfused bronchial arterial model, ACh challenges were repeated two times in the same way described previously except for the BK perfusion. Sham perfusion (Tyrode solution without BK) was performed for 40 min between the first and second ACh challenge. Measurement oj Vascular Permeability The degree of vascular permeability was determined by measuring the content of EB dye extravasated into the airway tissue after removing the intravascular EB dye using a modified method previously reported (16). When RL had returned to the baseline after the second ACh challenge, 10 mg/kg of EB dye (2% solution containing 10 g/dl bovine serum albumin) (17)was administered to the airway via the bronchial arteries. Tenminutes after EB, 150ml of Tyrode solution with 2.5% dextran 40 was perfused to the bronchial arteries at 150 mm Hg pressure to remove intravascular dye. Twosegmental bronchi of the left caudal (diaphragmatic) lobe in each cat were excised and incubated in 4 ml of 100% formamide at 54 0 C for 24 h after measuring wet weight. The concentration of EB dye extracted from the airway tissue was determined by light absorbance at 620 nm wavelength
KIMURA, INOUE, ICHINOSE, ET AL.
Fig. 1. Reproducibility of ACh doseresponse curves in the present model. ACh dose-response curves before and after 40 min sham perfusion in each cat are shown. Open circles indicate the first ACh challenge and closed circles the second one.
S
-8
-7
-6
-5
-4
Acetylcholine
(log (mol»)
Fig. 2. ACh dose-response curves before (open circles) and after (closed circles) BK infusion in each cat.
Acetyldtollne [log (mol»)
(220A spectrophotometer; Hitachi, Tokyo,Japan) and by interpolation on a standard curve of EB dye concentration (0.3-10 ug/ml), EB dye content of each airway was expressed as nanogram per milligram of wet weight tissue.
Histologic Study The right lungs of the cats were frozen with liquid nitrogen at a transpulmonary pressure of 2 em H 2 0 . The frozen lung was transferred to increasing concentrations of celloidin for 4 wk, immersed in xylol for 1 h, then embedded in paraffin, as wepreviously reported (18). They were cut into sections 4 urn thick and stained with Elastica-Masson. We measured airway wall area inside (WAi)and outside the airway smooth muscle (WAo) with a digitalizing tablet coupled to a computer, using the image from a TV monitor connected directly to the microscope (PC-9801;NEC, Tokyo, Japan; LA-5OO; PIAS, Tokyo, Japan) (19). Drugs The following drugs were used: thiopental, ketamine hydrochloride (Tokyo Tanabe, Japan); bradykinin (Peptide Institute Inc., Osaka, Japan); acetylcholine, succinylcholine, indomethacin, Evans blue dye, bovine serum albumin, and formamide (Sigma Chemical Co., St. Louis, MO); dextran 40 (Ohtsuca Pharmaceutical Co., Tokyo, Japan). Statistical Analyses Data are expressed as mean ± SD. To evaluate the sensitivity of the airway response to
ACh, the dose of ACh required to produce a 300% increase in RL (PD 3oo) was determined. In order to estimate the degree of increased maximal airway narrowing after BK, the ratio of RL at maximal plateau of airway narrowing obtained in the second ACh doseresponse curve to that in the first one was calculated. The Student's paired t test was used for statistical evaluation of the data; p < 0.05 was considered significant.
Results
Figure 1 shows the reproducibility of ACh dose-response curves in this model. Good reproducibility of the ACh challenge was obtained in each cat. The ACh dose-response curves from each cat in the BK perfusion group are shown in figure 2. Before BK infusion, ACh administration caused dose-dependent bronchoconstriction and produced a maximal plateau at which RL was 12 times the baseline value. After 10-7 M BK perfusion, the ACh dose-response curve shifted to the left and upward in each case compared with that before BK. Mean log [PD 30 0 (mol)] after BK was significantly lower than that before BK (- 6.90 ± 0.30, - 6.53 ± 0.42, respectively, p < 0.01) (figure 3). Mean RL at maximal plateau of airway narrowing after BK was significantly larger compared with that before
1303
BRONCHIAL EDEMA AND AIRWAY RESPONSIVENESS
was four times higher than that in the control group (147 ± 17.6, 35.5 ± 12.8 ng/mg tissue, respectively) (p < 0.05). The content of EB dye extravasated to the airway tissue and the ratio of RL at maximal plateau of airway narrowing obtained in the second ACh dose-response curve to that in the first one was significantly correlated (r = 0.85, p < 0.05) (figure 5). Histologic study showed peribronchial edema in the bronchi and bronchiole of the BK perfusion group compared with the saline (vehicle for BK) perfusion group. Neither alveolar wall thickening nor extravasated fluid in alveolar space was observed (figure 6). Airway edema was observed only on the wall area outside the airway smooth muscle (table 1).
P
