Oxygen radical scavengers protect against eosinophilinduced injury in isolated perfused rat lungs KEISAKU STEPHEN
FUJIMOTO, G. KAYES,
Departments of Physiology Alabama, Mobile, Alabama
SAWAKO YOSHIKAWA, AND JAMES C. PARKER and StructuralICellular 36688
SHERRI
Biology,
MARTIN,
College of Medicine,
University
of South
tured parenchymal and bronchoepithelial cells was primarily dependent on oxygen radicals. According to Agosti and co-workers (l), this cytotoxity is mediated by hypohalous acids formed by the eosinophil peroxidase(EPO) H,O,-halide system. Because most previous studies of eosinophil-induced tissue injury have been done in vitro, there are few physiological data on the vascular effects of eosinophils in the intact lung. Recently, Rowen et al. (29) reported that PMA-stimulated human eosinophils caused morphological evidence of acute epithelial and endothelial injury in isolated perfused rat lungs, which was attenuated by catalase pretreatment. However, they did not measure vascular resistance (RT), vascular permeability, or airway resistance (Raw). In physiological studies using isolated rat lungs, we recently demonstrated marked increases in pulmonary vascular permeability and resistance that were proportional to the numbers of PMA-activated eosinophils infused (14). RT increased predominantly in the small veins, and increases in peak airway pressure suggested an increase in Raw. In the present study, we describe further physiological studies using simultaneous measurements of RT, segmental pulmonary vascular resistance, Raw, lung compliance (CL), peak airway pressure (Paw), and vascular endothelial permeability in isolated perfused rat lungs. Vascular permeability was assessed using the capillary filtration coefficient (K,,), and the contribution of Raw phorbol myristate acetate; capillary filtration coefficient; suand CL to Paw was determined by measuring airflows peroxide dismutase; catalase; pulmonary vascular resistance; and volumes. By using these specific measurements of vascular permeability; airway resistance;lung compliance function, quantitative differences in the bronchial and vascular responses can be discriminated between groups. As previously observed, PMA-activated eosinophils EOSINOPHILS have a potent armament of cytotoxic grancaused pulmonary vasoconstriction, increased vascular ular proteins and oxygen free radicals and have been impermeability, and edema formation (14). The increased plicated as a primary effector cell in various types of lung Paw was primarily attributed to bronchoconstriction. injury (1,4, 13, 15, 31). However, the precise physiologiWe tested the hypothesis that lung injury was mediated cal mechanisms by which eosinophils induce tissue injury by eosinophil-generated reactive oxygen metabolites by and lung dysfunction have not been firmly established. pretreatment with the oxygen radical scavengers superIn vitro studies by Ayars et al. (4) demonstrated that oxide dismutase (SOD) and catalase. eosinophils stimulated with phorbol myristate acetate (PMA) produced a nonlytic detachment of cultured alveolar pneumocytes that was not dependent on the generMATERIALS AND METHODS ation of toxic oxygen radicals but rather appeared to be Isolated perfused rat lung preparation. The procedure mediated by granular proteins. Major basic protein also caused lysis of pneumocytes when released in high con- for isolating and perfusing rat lungs has been described in detail (14). Briefly, male Charles River Raleigh rats centrations (4). On the other hand, Davis and co-workers were used in all experiments. The trachea was cannu(9) demonstrated that eosinophil cytotoxity against culFUJIMOTO, KEISAKU, SAWAKO YOSHIKAWA, SHERRI MARTIN,STEPHENG.KAYES,ANDJAMESC.PARKER.OX~~~~XZ~~cal scavengers protect against eosinophil-induced injury in isolated perfused rat lungs. J. Appl. Physiol. 73(2): 687-694, 1992. -The protective effect of oxygen radical scavengerson lung injury induced by activated eosinophils was examined in isolated perfused rat lungs. Eosinophils were obtained by bronchoalveolar lavage from rats infected with Toxocara canis and activated with phorbol myristate acetate (PMA). There were no changesin pulmonary vascular (RT) and airway (Raw) resistances and only minimal changesin vascular permeability assessedusing the capillary filtration coefficient (K& in PMA control lungs and nonactivated eosinophil-treated lungs. In lungs receiving 3 X 10” PMA-activated eosinophils,there were significant increasesfrom baselineof ‘7.3-fold in RT at 30 min, primarily due to the constriction of small arteries and veins; 3.6-fold in KfC at 90 and 130 min; and 2.5fold in Raw. The lungs also becamemarkedly edematous.Both superoxide dismutaseand catalasepretreatment prevented the significant increase in I& and lung wet-to-dry weight ratios and partially attenuated the increasein Raw, but did not significantly inhibit the increasein RT induced by activated eosinophils.Heat-inactivated catalase did not attenuate the eosinophil-induced increasesin &, Raw, or RT. Thus, activated eosinophilsacutely increased microvascular permeability primarily through production of oxygen free radicals. The free radical scavengers superoxide dismutase and catalase partially attenuated the bronchoconstriction but had no significant effect on the vasoconstriction induced by activated eosinophils.
0161-X567/92
$2.00
Copyright0
1992 the
AmericanPhysiologicalSociety
Downloaded from www.physiology.org/journal/jappl at Midwestern Univ Lib (132.174.254.157) on February 12, 2019.
1687.
688
EOSINOPHIL-INDUCED
lated, and the lungs were ventilated with a gas mixture of 20% O,-5% CO,-75% N, by use of a respirator (model 683, Harvard Rodent Ventilator, South Natick, MA), which was set as follows: a tidal volume of 3 ml, a respiratory rate of 40/min, and 2.5 cmH,O positive end-expiratory pressure to keep the lungs in zone 3 conditions. The pulmonary artery and the left atrium were cannulated, and the lungs were immediately perfused using a roller pump (Gilson Minipuls 2, Middleton, WI) and 5% bovine albumin in Krebs bicarbonate buffer maintained at 37OC with a heat exchanger. The lungs and heart were removed, and lung weight was continuously monitored using a spring-counterbalanced force transducer (force displacement transducer FT03D, Grass, Quincy, MA). Once the effluent was visibly clear of formed blood elements, a recirculating perfusion system was used (total volume 30 ml). The flow rate was maintained at ~20 ml/min as long as the lungs remained in an isogravimetric state. When the lungs could not maintain an isogravimetric state, the flow rate was decreased until the lungs were isogravimetric. Arterial, venous, and airway pressures were measured by pressure transducers (Cobe, Lakewood, CO) via thin catheters connected to the cannulas close to the lungs. Lung weight and vascular and airway pressures were recorded continuously on a Grass model 7D polygraph. Estimation of pulmonary capillary pressure, arterial and venous occlusion pressures, and segmental pulmonary vascular resistances. Pulmonary capillary pressure (PC) was
measured by the double occlusion technique (28,35). Inflow and outflow tubing were simultaneously occluded, causing pulmonary arterial pressure (Pa) and pulmonary venous pressure (Pv) to equilibrate at PC. Arterial occlusion pressure (Pao) and venous occlusion pressure (Pvo) were also measured by individually occluding either the inflow or outflow tubing, respectively (28). Segmental resistances were calculated by dividing each of the pressure drops by the perfusate flow (Q) normalized to 100 g initial wet lung weight as follows: RT = (Pa - Pv)/(Q/lOO g); precapillary resistance (Ra) = (Pa - Pc)/(Q/lOO g); postcapillary resistance (Rv) = (PC - Pv)/(Q/lOO g); large and small artery resistances, RLA = (Pa - Pao)/ (Q/100 g) and RSA = (Pao - Pc)/(Q/lOO g), respectively; small and large vein resistances, RSV = (PC Pvo)/(Q/lOO g) and RLV = (Pvo - Pv)/(Q/lOO g), respectively (28). Measurement of Khc. K,, was used as the index of endothelial permeability to fluid, because K,, is determined by hydraulic conductivity when capillary surface area remains constant (10,33). K,, was measured by the time 0 extrapolation method of Drake et al. (10). After a rapid increase in PC produced by raising the venous reservoir increase @P v = 7-8 cmH,O), there was a two-component in lung weight. The initial rapid weight gain was due to an increased lung vascular volume followed by a slow component of weight gain representing filtration of fluid into the pulmonary interstitium. The rate of slow weight gain (AWt/At) was plotted on semilogarithmic graph paper, and the line of best fit was extrapolated to time 0 (t = 0). K,,, was then calculated by dividing the initial rate
LUNG
INJURY
of weight gain at time 0 by the change in capillary pressure and normalizing to 100 g initial wet lung weight K fc , = (AWtlAt),=,/APc
(1)
In cases where Kf c might be underestimated because of low flow rates (~56% of the baseline flow rate) used to prevent excess edema formation (33), PC was raised by increasing the flow rate to 50% of the baseline flow rate and raising the venous reservoir. Raw and CL. Raw was determined using a modification of the single-breath method of Zin et al. (38). A pneumotach was constructed according to Mortola and Nowaraj (22) for use with small animals and situated in line with the tracheal cannula. Airflow was measured using a Validyne differential pressure transducer calibrated for flow and the signal integrated with a Grass 7PlO integrating preamplifier to obtain tidal volumes (VT). A separate cannula was used to monitor tracheal pressure, and the difference between end-expiratory and end-inspiratory pressures @Paw) was determined. Inspiratory and expiratory volume-time curves were recorded at a fast chart speed (50 mm/s) and digitized for analysis by use of a digitizing tablet and IBM PC digital computer. CL was determined using CL = VT/APaw (2) Raw was determined by fitting the expiratory volumetime curve [V(t)] with a least-squares single-exponential curve, where v(t)
= VTemkt
(3)
The rate constant k = l/7 and the time constant 7 = Raw&, so Raw = T/CL. These measurements were not obtained in all experiments of every group, but group sizes ranged from four to six. Measurement of wet and dry lung weights. Initial lung weight was estimated by measuring the weight of the heart, mediastinal tissue, and lungs at the beginning of the experiment and subtracting the weight of all extrapulmonary tissue at the end of the experiment. Lungs excised at the end of the experiment were weighed for final wet lung weight. The wet lungs were then dried in an oven at 60°C for 1 mo and again weighed to calculate the wet-to-dry lung weight ratio (W/D). Preparation of eosinophils. The procedure for isolation of eosinophils has been described in detail (14). Briefly, rats were infected with 2 X 10’ infective eggs of Toxocara canis by gastric intubation under ether anesthesia. On the 14th day after infection, the eosinophils that accumulated in the lungs were collected by bronchoalveolar lavage. After the contaminating red blood cells were lysed with tris(hydroxymethyl)aminomethane HCl-buffered ammonium chloride (pH 7.2), the collected cells were suspended in Dulbecco’s phosphate-buffered saline without Ca2+ and Mg2+ (DBSS; Grand Island Biological, Long Island, NY) with 0.1% gelatin and 5 mM glucose (DBSS-GG). A cytosmear of the collected eosinophils was prepared for differential counting and stained (Hema1 Stain, Danbury, CT), and cell viability was >95% by trypan blue dye exclusion. Lavaged cells were >80% eo-
Downloaded from www.physiology.org/journal/jappl at Midwestern Univ Lib (132.174.254.157) on February 12, 2019.
l
EOSINOPHIL-INDUCED
sinophils [88.3 t 6.9% (SD)], and contaminating cells were almost exclusively mononuclear cells. Actiuation of eosinophils. Before the experiments, 3X lo6 eosinophils suspended in 3 ml of DBSS-GG were incubated with PMA, (10 pglml; Sigma Chemical, St. Louis, MO) at 37°C for 30 min and then centrifuged at 1,000 rpm for 10 min at room temperature (Beckman TJ-6 centrifuge). The supernatant was aspirated, and the pellet of activated eosinophils was resuspended in the last 50 ~1 of supernatant. Experimental p~otocoIs. Isolated perfused lungs were allowed to attain an initial isogravimetric state at a venous pressure of 3.5 cmH,O, and the baseline pulmonary hemodynamics, Paw, Raw, CL, and K,,, were measured. At 30, 90, and 130 min after the addition of eosinophils, the same variables were again measured. The lungs were then grouped as follows: 1) activated eosinophil group (n = 6) 3 X lo6 activated eosinophils were added to the venous’ reservoir 10 min after the baseline measurements; 2) activated eosinophil + SOD group (n = 6): 6,000 U of SOD from bovine liver (Sigma Chemical) were added to the venous reservoir immediately before the 3 X lo6 activated eosinophils; 3) activated eosinophil + CAT group (n = 6): 6,000 U of catalase from bovine liver (Sigma Chemical) were added to the venous reservoir immediately before the activated eosinophil challenge as in the activated eosinophil + SOD group. 4) nonactivated eosinophil group (n = 6): 3 X lo6 eosinophils suspended in 3 ml of DBSS-GG were incubated without PMA, and the eosinophils were added to the reservoir as in the activated eosinophil group; 5) PMA control group (n = 6). 50 ~1 of DBSS-GG containing 10 ,uglml PMA were incubated without eosinophils as previously described and then added to the reservoir as a PMA control; 6) heat-inactivated catalase control group (n = 5): 6,000 U of catalase from bovine liver were boiled for 30 min to abolish enzymatic activity and added to the venous reservoir immediately before 3 X lo6 activated eosinophils. These experiments were performed to determine whether the added protein had any nonenzymatic protective effects. Statistics. The results for each group are expressed as means t SE. Comparisons of variables within groups (time effect) and between groups (treatment effect) were performed using an analysis of variance (ANOVA) followed by a Newman-Keuls multiple range test. Differences were considered significant at P < 0.05.
LUNG
689
INJURY
cantly increased at 30 min after the administration of activated eosinophils, but these variables also returned toward baseline values after 90-130 min. In both SODand catalase-pretreated lungs, the increases in Pa and PC at 30 min were significantly less than that produced by activated eosinophils alone. Figure 1 shows the changes in RT with respect to time in all groups except the heat-inactivated catalase controls (group 6). There were no changes in the PMA control and nonactivated eosinophil groups. In activated eosinophil-treated lungs, RT significantly increased at 30 min to 7.3 times the baseline value and then returned to the baseline value at 90-130 min. The intensity of this pressor response was not significantly affected by pretreatment with either SOD or catalase. The changes in segmental pulmonary vascular resistances with time are summarized in Table 2. There were no changes in any of the segmental vascular resistances in control and nonactivated eosinophil groups. Although both precapillary (Ra) and postcapillary (Rv) resistances were significantly increased at 30 min in all three activated eosinophil groups, there was a greater increase in the Rv. This produced a significant decrease in the ratio of these resistances (Ra/Rv). The peak increases relative to baseline values were greatest in the small arteries (RsA; 2,014-2880%) and veins (Rsv; 1,518-2,217%) compared with that in the large veins (RLv; 223-345%). There was no significant increase in the large artery resistance (RLA). Thus the increase in RT was attributed primarily to the increases in small vessel resistance. Neither SOD nor catalase significantly attenuated the increases in the segmental resistance or RT caused by activated eosinophils. $. Figure 2 shows the changes in K,, with respect to time in groups 1-5. There were no changes in K,, in the control group. In the nonactivated eosinophil-‘treated lungs, Kf,c was significantly increased from a baseline value of 0.23 t 0.02 to 0.44 t 0.03 ml min-’ cmH,O-’ 100 g-’ at 130 min (P < 0.05), but a significant difference from the controls was not observed. In the activated eosinophil-treated lungs, K,, was significantly increased at 30 min and increased further from a baseline value of 0.28 t 0.02 to 1.00 t 0.10 ml min. cmH,O-’ 100 g-l at 90 min (P < 0.01). Pretreatment with either SOD or catalase alone largely prevented the marked increase in K,, induced by activated eosinophils such that the values of K,, at 130 min (0.35 t 0.02 in SOD-pretreated group and 0.41 t 0.04 ml min-l cmH,O-’ 100 g-l in catalase-pretreated group) were not statistically different from that in the control group. Paw, Raw, and CL. Figure 3 shows the changes in Paw with respect to time. Paw increased significantly at 30 min and increased further every time K,,c was measured in all groups. There were no significant differences in Paw between control and nonactivated eosinophil groups, but in the activated eosinophil-treated group, the increased Paw was significantly greater. than in either the control or nonactivated eosinophil groups at every time period. Both SOD and catalase significantly attenuated the increase in Paw at 90 and 130 min compared l
l
l
RESULTS
Pulmonary hemodynamic responses. Table 1 summarized the hemodynamic responses with time of all groups except group 6. There were no changes in any hemodynamic variables with time in the control and nonactivated eosinophils groups. However, in lungs challenged with activated eosinophils, Pa increased by 161-174% at 30 min. Simultaneously with the peak in Pa, the lung weight transiently decreased and then subsequently increased toward baseline levels. Flow rate was decreased at 30 min to -29.4% of the baseline flow rate to maintain an isogravimetric state. Pao and PC were also signifi-
l
l
l
l
l
Downloaded from www.physiology.org/journal/jappl at Midwestern Univ Lib (132.174.254.157) on February 12, 2019.
690
EOSINOPHIL-INDUCED
LUNG
INJURY
1. Pulmonary hemodynamics after challenge with 3 X lo6 activated eosinophils with and without oxygen radical scavengers in isolated perfused rat lungs TABLE
&9
Pa,
ml/min
pv,
cmH,O
cmH,O
Pao,
Pvo,
PC,
cmH,O
cmH,O
cmH,O
Baseline Controls Non-AC Eo AC Eo AC Eo +SOD AC Eo +CAT
Controls Non-AC Eo AC Eo AC Eo +SOD AC Eo +CAT
19.7_tO.l 19.8t0.2 20.2t0.2
lO.Ot0.2 10.3kO.2 10.3t0.3 9.7t0.3
19.5kO.2
9.7t0.2
19.5-to.2
19.7tO.l
lO.lt0.5
19.8t0.2 5.8-tO.6"f
lO.Ot0.2 17.4*0.5*?
6.6+1.2*? 6.3+0.9*-j-
15.3+0.7*t$ 15.8&0.5”?$
Controls Non-AC Eo AC Eo AC Eo +SOD AC Eo +CAT
19.7tO.l 19.8t0.2 17.2t0.9* 16.2t1.7”
9.7kO.3 lO.lt0.3 13.1*0.7*? 11.5t0.4*
14.4+1.9*t
11.6t0.7*
Controls Non-AC Eo AC Eo AC Eo +SOD AC Eo +CAT
19.7to. 1 19.8t0.2 18.7t0.7
9.7kO.3 10.2t0.3 11.4_+0.4*“f
3.3t0.2 3.5to.o 3.5kO.l 3.5tO.l 3.6tO.O 30 min 3.620.1 3.6kO.O 3.4kO.l
6.8t0.2 6.9rt0.2 7.2tO.l 6.8kO.2 6.61~0.3
4.9kO.l 4.8kO.l 4.8kO.l 5.2tO.l 5.0to.o
5.7tO.l 5.8t0.2 6.2t0.2 5.9kO.l 5.9kO.l
6.8t0.4 6.9t0.2 16.1&0.7*'f
5.ltO.l 5.OkO.l 4.8t0.3
5.8kO.l 5.8tO.l 10.5~0.4"~
3.7kO.l
14.5-tO.8”‘r
4.7tO.l
9.3&0.6*“f$
3.6tO.O
14.3+0.5*-t
4.9tO.l
8.7+0.5"?$
90 min 3.4t0.2 3.5kO.l 3.4tO.l 3.5kO.O 3.6tO.l
6.5tO.l 6.9t0.2 10.7-tO.6"t 8.6t0.7* 8.8tl.O*
4.8kO.l 4.9kO.l 4.8t0.2 5.lkO.2 5.0t0.3
5.6tO.l 5.8tO.l 7.5*0.4"t 6.6t0.3 6.6t0.5
6.7kO.l 6.8kO.l 9.5+0.5*-f 7.OkO.41 7.5+0.5$
5.ltO.l 4.920.1 5.OkO.2 5.2kO.l 5.0tO.l
5.7tO.l 5.8tO.l 6.8+0.3*t 6.1+0.1$ 6.0+0.2$
130 rnin
20.220.2
10.6t0.5
3.520.1 3.6tO.l 3.6kO.l 3.5tO.l
19.5t0.2
10.9t0.4
3.6tO.l
are means t SE of 6 rats. Non-AC Eo, nonactivated eosinophils (3 X 106); AC Eo, activated eosinophils (3 X 106); SOD, superoxide CAT, catalase; Q, perfusate flow rate; Pa, pulmonary arterial pressure; Pv, pulmonary venous pressure; Pao, pulmonary arterial pressure; Pvo, pulmonary venous occlusion pressure. Pc, pulmonary capillary pressure. * P < 0.05 vs. baseline; t P < 0.05 vs. controls; 0.05, AC Eo vs. AC Eo +SOD or AC Eo +CAT.
Values dismutase; occlusion
$P
FUJIMOTO, G. KAYES,
Departments of Physiology Alabama, Mobile, Alabama
SAWAKO YOSHIKAWA, AND JAMES C. PARKER and StructuralICellular 36688
SHERRI
Biology,
MARTIN,
College of Medicine,
University
of South
tured parenchymal and bronchoepithelial cells was primarily dependent on oxygen radicals. According to Agosti and co-workers (l), this cytotoxity is mediated by hypohalous acids formed by the eosinophil peroxidase(EPO) H,O,-halide system. Because most previous studies of eosinophil-induced tissue injury have been done in vitro, there are few physiological data on the vascular effects of eosinophils in the intact lung. Recently, Rowen et al. (29) reported that PMA-stimulated human eosinophils caused morphological evidence of acute epithelial and endothelial injury in isolated perfused rat lungs, which was attenuated by catalase pretreatment. However, they did not measure vascular resistance (RT), vascular permeability, or airway resistance (Raw). In physiological studies using isolated rat lungs, we recently demonstrated marked increases in pulmonary vascular permeability and resistance that were proportional to the numbers of PMA-activated eosinophils infused (14). RT increased predominantly in the small veins, and increases in peak airway pressure suggested an increase in Raw. In the present study, we describe further physiological studies using simultaneous measurements of RT, segmental pulmonary vascular resistance, Raw, lung compliance (CL), peak airway pressure (Paw), and vascular endothelial permeability in isolated perfused rat lungs. Vascular permeability was assessed using the capillary filtration coefficient (K,,), and the contribution of Raw phorbol myristate acetate; capillary filtration coefficient; suand CL to Paw was determined by measuring airflows peroxide dismutase; catalase; pulmonary vascular resistance; and volumes. By using these specific measurements of vascular permeability; airway resistance;lung compliance function, quantitative differences in the bronchial and vascular responses can be discriminated between groups. As previously observed, PMA-activated eosinophils EOSINOPHILS have a potent armament of cytotoxic grancaused pulmonary vasoconstriction, increased vascular ular proteins and oxygen free radicals and have been impermeability, and edema formation (14). The increased plicated as a primary effector cell in various types of lung Paw was primarily attributed to bronchoconstriction. injury (1,4, 13, 15, 31). However, the precise physiologiWe tested the hypothesis that lung injury was mediated cal mechanisms by which eosinophils induce tissue injury by eosinophil-generated reactive oxygen metabolites by and lung dysfunction have not been firmly established. pretreatment with the oxygen radical scavengers superIn vitro studies by Ayars et al. (4) demonstrated that oxide dismutase (SOD) and catalase. eosinophils stimulated with phorbol myristate acetate (PMA) produced a nonlytic detachment of cultured alveolar pneumocytes that was not dependent on the generMATERIALS AND METHODS ation of toxic oxygen radicals but rather appeared to be Isolated perfused rat lung preparation. The procedure mediated by granular proteins. Major basic protein also caused lysis of pneumocytes when released in high con- for isolating and perfusing rat lungs has been described in detail (14). Briefly, male Charles River Raleigh rats centrations (4). On the other hand, Davis and co-workers were used in all experiments. The trachea was cannu(9) demonstrated that eosinophil cytotoxity against culFUJIMOTO, KEISAKU, SAWAKO YOSHIKAWA, SHERRI MARTIN,STEPHENG.KAYES,ANDJAMESC.PARKER.OX~~~~XZ~~cal scavengers protect against eosinophil-induced injury in isolated perfused rat lungs. J. Appl. Physiol. 73(2): 687-694, 1992. -The protective effect of oxygen radical scavengerson lung injury induced by activated eosinophils was examined in isolated perfused rat lungs. Eosinophils were obtained by bronchoalveolar lavage from rats infected with Toxocara canis and activated with phorbol myristate acetate (PMA). There were no changesin pulmonary vascular (RT) and airway (Raw) resistances and only minimal changesin vascular permeability assessedusing the capillary filtration coefficient (K& in PMA control lungs and nonactivated eosinophil-treated lungs. In lungs receiving 3 X 10” PMA-activated eosinophils,there were significant increasesfrom baselineof ‘7.3-fold in RT at 30 min, primarily due to the constriction of small arteries and veins; 3.6-fold in KfC at 90 and 130 min; and 2.5fold in Raw. The lungs also becamemarkedly edematous.Both superoxide dismutaseand catalasepretreatment prevented the significant increase in I& and lung wet-to-dry weight ratios and partially attenuated the increasein Raw, but did not significantly inhibit the increasein RT induced by activated eosinophils.Heat-inactivated catalase did not attenuate the eosinophil-induced increasesin &, Raw, or RT. Thus, activated eosinophilsacutely increased microvascular permeability primarily through production of oxygen free radicals. The free radical scavengers superoxide dismutase and catalase partially attenuated the bronchoconstriction but had no significant effect on the vasoconstriction induced by activated eosinophils.
0161-X567/92
$2.00
Copyright0
1992 the
AmericanPhysiologicalSociety
Downloaded from www.physiology.org/journal/jappl at Midwestern Univ Lib (132.174.254.157) on February 12, 2019.
1687.
688
EOSINOPHIL-INDUCED
lated, and the lungs were ventilated with a gas mixture of 20% O,-5% CO,-75% N, by use of a respirator (model 683, Harvard Rodent Ventilator, South Natick, MA), which was set as follows: a tidal volume of 3 ml, a respiratory rate of 40/min, and 2.5 cmH,O positive end-expiratory pressure to keep the lungs in zone 3 conditions. The pulmonary artery and the left atrium were cannulated, and the lungs were immediately perfused using a roller pump (Gilson Minipuls 2, Middleton, WI) and 5% bovine albumin in Krebs bicarbonate buffer maintained at 37OC with a heat exchanger. The lungs and heart were removed, and lung weight was continuously monitored using a spring-counterbalanced force transducer (force displacement transducer FT03D, Grass, Quincy, MA). Once the effluent was visibly clear of formed blood elements, a recirculating perfusion system was used (total volume 30 ml). The flow rate was maintained at ~20 ml/min as long as the lungs remained in an isogravimetric state. When the lungs could not maintain an isogravimetric state, the flow rate was decreased until the lungs were isogravimetric. Arterial, venous, and airway pressures were measured by pressure transducers (Cobe, Lakewood, CO) via thin catheters connected to the cannulas close to the lungs. Lung weight and vascular and airway pressures were recorded continuously on a Grass model 7D polygraph. Estimation of pulmonary capillary pressure, arterial and venous occlusion pressures, and segmental pulmonary vascular resistances. Pulmonary capillary pressure (PC) was
measured by the double occlusion technique (28,35). Inflow and outflow tubing were simultaneously occluded, causing pulmonary arterial pressure (Pa) and pulmonary venous pressure (Pv) to equilibrate at PC. Arterial occlusion pressure (Pao) and venous occlusion pressure (Pvo) were also measured by individually occluding either the inflow or outflow tubing, respectively (28). Segmental resistances were calculated by dividing each of the pressure drops by the perfusate flow (Q) normalized to 100 g initial wet lung weight as follows: RT = (Pa - Pv)/(Q/lOO g); precapillary resistance (Ra) = (Pa - Pc)/(Q/lOO g); postcapillary resistance (Rv) = (PC - Pv)/(Q/lOO g); large and small artery resistances, RLA = (Pa - Pao)/ (Q/100 g) and RSA = (Pao - Pc)/(Q/lOO g), respectively; small and large vein resistances, RSV = (PC Pvo)/(Q/lOO g) and RLV = (Pvo - Pv)/(Q/lOO g), respectively (28). Measurement of Khc. K,, was used as the index of endothelial permeability to fluid, because K,, is determined by hydraulic conductivity when capillary surface area remains constant (10,33). K,, was measured by the time 0 extrapolation method of Drake et al. (10). After a rapid increase in PC produced by raising the venous reservoir increase @P v = 7-8 cmH,O), there was a two-component in lung weight. The initial rapid weight gain was due to an increased lung vascular volume followed by a slow component of weight gain representing filtration of fluid into the pulmonary interstitium. The rate of slow weight gain (AWt/At) was plotted on semilogarithmic graph paper, and the line of best fit was extrapolated to time 0 (t = 0). K,,, was then calculated by dividing the initial rate
LUNG
INJURY
of weight gain at time 0 by the change in capillary pressure and normalizing to 100 g initial wet lung weight K fc , = (AWtlAt),=,/APc
(1)
In cases where Kf c might be underestimated because of low flow rates (~56% of the baseline flow rate) used to prevent excess edema formation (33), PC was raised by increasing the flow rate to 50% of the baseline flow rate and raising the venous reservoir. Raw and CL. Raw was determined using a modification of the single-breath method of Zin et al. (38). A pneumotach was constructed according to Mortola and Nowaraj (22) for use with small animals and situated in line with the tracheal cannula. Airflow was measured using a Validyne differential pressure transducer calibrated for flow and the signal integrated with a Grass 7PlO integrating preamplifier to obtain tidal volumes (VT). A separate cannula was used to monitor tracheal pressure, and the difference between end-expiratory and end-inspiratory pressures @Paw) was determined. Inspiratory and expiratory volume-time curves were recorded at a fast chart speed (50 mm/s) and digitized for analysis by use of a digitizing tablet and IBM PC digital computer. CL was determined using CL = VT/APaw (2) Raw was determined by fitting the expiratory volumetime curve [V(t)] with a least-squares single-exponential curve, where v(t)
= VTemkt
(3)
The rate constant k = l/7 and the time constant 7 = Raw&, so Raw = T/CL. These measurements were not obtained in all experiments of every group, but group sizes ranged from four to six. Measurement of wet and dry lung weights. Initial lung weight was estimated by measuring the weight of the heart, mediastinal tissue, and lungs at the beginning of the experiment and subtracting the weight of all extrapulmonary tissue at the end of the experiment. Lungs excised at the end of the experiment were weighed for final wet lung weight. The wet lungs were then dried in an oven at 60°C for 1 mo and again weighed to calculate the wet-to-dry lung weight ratio (W/D). Preparation of eosinophils. The procedure for isolation of eosinophils has been described in detail (14). Briefly, rats were infected with 2 X 10’ infective eggs of Toxocara canis by gastric intubation under ether anesthesia. On the 14th day after infection, the eosinophils that accumulated in the lungs were collected by bronchoalveolar lavage. After the contaminating red blood cells were lysed with tris(hydroxymethyl)aminomethane HCl-buffered ammonium chloride (pH 7.2), the collected cells were suspended in Dulbecco’s phosphate-buffered saline without Ca2+ and Mg2+ (DBSS; Grand Island Biological, Long Island, NY) with 0.1% gelatin and 5 mM glucose (DBSS-GG). A cytosmear of the collected eosinophils was prepared for differential counting and stained (Hema1 Stain, Danbury, CT), and cell viability was >95% by trypan blue dye exclusion. Lavaged cells were >80% eo-
Downloaded from www.physiology.org/journal/jappl at Midwestern Univ Lib (132.174.254.157) on February 12, 2019.
l
EOSINOPHIL-INDUCED
sinophils [88.3 t 6.9% (SD)], and contaminating cells were almost exclusively mononuclear cells. Actiuation of eosinophils. Before the experiments, 3X lo6 eosinophils suspended in 3 ml of DBSS-GG were incubated with PMA, (10 pglml; Sigma Chemical, St. Louis, MO) at 37°C for 30 min and then centrifuged at 1,000 rpm for 10 min at room temperature (Beckman TJ-6 centrifuge). The supernatant was aspirated, and the pellet of activated eosinophils was resuspended in the last 50 ~1 of supernatant. Experimental p~otocoIs. Isolated perfused lungs were allowed to attain an initial isogravimetric state at a venous pressure of 3.5 cmH,O, and the baseline pulmonary hemodynamics, Paw, Raw, CL, and K,,, were measured. At 30, 90, and 130 min after the addition of eosinophils, the same variables were again measured. The lungs were then grouped as follows: 1) activated eosinophil group (n = 6) 3 X lo6 activated eosinophils were added to the venous’ reservoir 10 min after the baseline measurements; 2) activated eosinophil + SOD group (n = 6): 6,000 U of SOD from bovine liver (Sigma Chemical) were added to the venous reservoir immediately before the 3 X lo6 activated eosinophils; 3) activated eosinophil + CAT group (n = 6): 6,000 U of catalase from bovine liver (Sigma Chemical) were added to the venous reservoir immediately before the activated eosinophil challenge as in the activated eosinophil + SOD group. 4) nonactivated eosinophil group (n = 6): 3 X lo6 eosinophils suspended in 3 ml of DBSS-GG were incubated without PMA, and the eosinophils were added to the reservoir as in the activated eosinophil group; 5) PMA control group (n = 6). 50 ~1 of DBSS-GG containing 10 ,uglml PMA were incubated without eosinophils as previously described and then added to the reservoir as a PMA control; 6) heat-inactivated catalase control group (n = 5): 6,000 U of catalase from bovine liver were boiled for 30 min to abolish enzymatic activity and added to the venous reservoir immediately before 3 X lo6 activated eosinophils. These experiments were performed to determine whether the added protein had any nonenzymatic protective effects. Statistics. The results for each group are expressed as means t SE. Comparisons of variables within groups (time effect) and between groups (treatment effect) were performed using an analysis of variance (ANOVA) followed by a Newman-Keuls multiple range test. Differences were considered significant at P < 0.05.
LUNG
689
INJURY
cantly increased at 30 min after the administration of activated eosinophils, but these variables also returned toward baseline values after 90-130 min. In both SODand catalase-pretreated lungs, the increases in Pa and PC at 30 min were significantly less than that produced by activated eosinophils alone. Figure 1 shows the changes in RT with respect to time in all groups except the heat-inactivated catalase controls (group 6). There were no changes in the PMA control and nonactivated eosinophil groups. In activated eosinophil-treated lungs, RT significantly increased at 30 min to 7.3 times the baseline value and then returned to the baseline value at 90-130 min. The intensity of this pressor response was not significantly affected by pretreatment with either SOD or catalase. The changes in segmental pulmonary vascular resistances with time are summarized in Table 2. There were no changes in any of the segmental vascular resistances in control and nonactivated eosinophil groups. Although both precapillary (Ra) and postcapillary (Rv) resistances were significantly increased at 30 min in all three activated eosinophil groups, there was a greater increase in the Rv. This produced a significant decrease in the ratio of these resistances (Ra/Rv). The peak increases relative to baseline values were greatest in the small arteries (RsA; 2,014-2880%) and veins (Rsv; 1,518-2,217%) compared with that in the large veins (RLv; 223-345%). There was no significant increase in the large artery resistance (RLA). Thus the increase in RT was attributed primarily to the increases in small vessel resistance. Neither SOD nor catalase significantly attenuated the increases in the segmental resistance or RT caused by activated eosinophils. $. Figure 2 shows the changes in K,, with respect to time in groups 1-5. There were no changes in K,, in the control group. In the nonactivated eosinophil-‘treated lungs, Kf,c was significantly increased from a baseline value of 0.23 t 0.02 to 0.44 t 0.03 ml min-’ cmH,O-’ 100 g-’ at 130 min (P < 0.05), but a significant difference from the controls was not observed. In the activated eosinophil-treated lungs, K,, was significantly increased at 30 min and increased further from a baseline value of 0.28 t 0.02 to 1.00 t 0.10 ml min. cmH,O-’ 100 g-l at 90 min (P < 0.01). Pretreatment with either SOD or catalase alone largely prevented the marked increase in K,, induced by activated eosinophils such that the values of K,, at 130 min (0.35 t 0.02 in SOD-pretreated group and 0.41 t 0.04 ml min-l cmH,O-’ 100 g-l in catalase-pretreated group) were not statistically different from that in the control group. Paw, Raw, and CL. Figure 3 shows the changes in Paw with respect to time. Paw increased significantly at 30 min and increased further every time K,,c was measured in all groups. There were no significant differences in Paw between control and nonactivated eosinophil groups, but in the activated eosinophil-treated group, the increased Paw was significantly greater. than in either the control or nonactivated eosinophil groups at every time period. Both SOD and catalase significantly attenuated the increase in Paw at 90 and 130 min compared l
l
l
RESULTS
Pulmonary hemodynamic responses. Table 1 summarized the hemodynamic responses with time of all groups except group 6. There were no changes in any hemodynamic variables with time in the control and nonactivated eosinophils groups. However, in lungs challenged with activated eosinophils, Pa increased by 161-174% at 30 min. Simultaneously with the peak in Pa, the lung weight transiently decreased and then subsequently increased toward baseline levels. Flow rate was decreased at 30 min to -29.4% of the baseline flow rate to maintain an isogravimetric state. Pao and PC were also signifi-
l
l
l
l
l
Downloaded from www.physiology.org/journal/jappl at Midwestern Univ Lib (132.174.254.157) on February 12, 2019.
690
EOSINOPHIL-INDUCED
LUNG
INJURY
1. Pulmonary hemodynamics after challenge with 3 X lo6 activated eosinophils with and without oxygen radical scavengers in isolated perfused rat lungs TABLE
&9
Pa,
ml/min
pv,
cmH,O
cmH,O
Pao,
Pvo,
PC,
cmH,O
cmH,O
cmH,O
Baseline Controls Non-AC Eo AC Eo AC Eo +SOD AC Eo +CAT
Controls Non-AC Eo AC Eo AC Eo +SOD AC Eo +CAT
19.7_tO.l 19.8t0.2 20.2t0.2
lO.Ot0.2 10.3kO.2 10.3t0.3 9.7t0.3
19.5kO.2
9.7t0.2
19.5-to.2
19.7tO.l
lO.lt0.5
19.8t0.2 5.8-tO.6"f
lO.Ot0.2 17.4*0.5*?
6.6+1.2*? 6.3+0.9*-j-
15.3+0.7*t$ 15.8&0.5”?$
Controls Non-AC Eo AC Eo AC Eo +SOD AC Eo +CAT
19.7tO.l 19.8t0.2 17.2t0.9* 16.2t1.7”
9.7kO.3 lO.lt0.3 13.1*0.7*? 11.5t0.4*
14.4+1.9*t
11.6t0.7*
Controls Non-AC Eo AC Eo AC Eo +SOD AC Eo +CAT
19.7to. 1 19.8t0.2 18.7t0.7
9.7kO.3 10.2t0.3 11.4_+0.4*“f
3.3t0.2 3.5to.o 3.5kO.l 3.5tO.l 3.6tO.O 30 min 3.620.1 3.6kO.O 3.4kO.l
6.8t0.2 6.9rt0.2 7.2tO.l 6.8kO.2 6.61~0.3
4.9kO.l 4.8kO.l 4.8kO.l 5.2tO.l 5.0to.o
5.7tO.l 5.8t0.2 6.2t0.2 5.9kO.l 5.9kO.l
6.8t0.4 6.9t0.2 16.1&0.7*'f
5.ltO.l 5.OkO.l 4.8t0.3
5.8kO.l 5.8tO.l 10.5~0.4"~
3.7kO.l
14.5-tO.8”‘r
4.7tO.l
9.3&0.6*“f$
3.6tO.O
14.3+0.5*-t
4.9tO.l
8.7+0.5"?$
90 min 3.4t0.2 3.5kO.l 3.4tO.l 3.5kO.O 3.6tO.l
6.5tO.l 6.9t0.2 10.7-tO.6"t 8.6t0.7* 8.8tl.O*
4.8kO.l 4.9kO.l 4.8t0.2 5.lkO.2 5.0t0.3
5.6tO.l 5.8tO.l 7.5*0.4"t 6.6t0.3 6.6t0.5
6.7kO.l 6.8kO.l 9.5+0.5*-f 7.OkO.41 7.5+0.5$
5.ltO.l 4.920.1 5.OkO.2 5.2kO.l 5.0tO.l
5.7tO.l 5.8tO.l 6.8+0.3*t 6.1+0.1$ 6.0+0.2$
130 rnin
20.220.2
10.6t0.5
3.520.1 3.6tO.l 3.6kO.l 3.5tO.l
19.5t0.2
10.9t0.4
3.6tO.l
are means t SE of 6 rats. Non-AC Eo, nonactivated eosinophils (3 X 106); AC Eo, activated eosinophils (3 X 106); SOD, superoxide CAT, catalase; Q, perfusate flow rate; Pa, pulmonary arterial pressure; Pv, pulmonary venous pressure; Pao, pulmonary arterial pressure; Pvo, pulmonary venous occlusion pressure. Pc, pulmonary capillary pressure. * P < 0.05 vs. baseline; t P < 0.05 vs. controls; 0.05, AC Eo vs. AC Eo +SOD or AC Eo +CAT.
Values dismutase; occlusion
$P
