TOXICOLOGY

AND

APPLIED

PHARMACOLOGY

116,

161-169 ( 1992)

Altered Capillary Filtration Coefficient in Parathion- and ParaoxonInduced Edema in Isolated and Perfused Rabbit Lungs ANNIE Department

DELAUNOIS, PASCAL GUSTIN,

of Pharmacology-Pharmacotherapy-To.xicoiogy. Bd de Colonster.

B 41 Sart-Tilman,

AND MICHEL ANSAY

Faculty qf‘ Veterinary Medicine, 4000 Li?ge, Belgium

University

qf Ligge.

Received February 18. 1992: accepted May 28, 1992

Altered Capillary Filtration Coefficient in Parathion- and Paraoxon-Induced Edema in Isolated and Perfused Rabbit Lungs. DELAUNOIS,A., GUSTIN, P., AND ANSAY, M. (1992). Toxicol. Appl. Pharmacol. 116, 161-169. Changes in pulmonary endothelium permeability and in microvascular hemodynamics induced by parathion (Pth) and paraoxon (Pox), its active metabolite, were investigated in isolated, perfused rabbit lungs. Blood-free perfusate was recirculated through isolated and ventilated lungs in an isogravimetric state and in zone III conditions. The arterial/venous/double occlusion technique was used to divide the total vascular resistance (R,) into four components: arterial, precapillary, postcapillary, and venous. The capillary filtration coefficient (&) was evaluated by measuring the amount of fluid filtering through the endothelium when the arterial and venous pressures were suddenly increased. Pth and Pox induced pulmonary edema by increasing endothelium permeability without changing the hemodynamic parameters at any level of the vascular bed. The KfCvalue increased progressively, reaching a maximum (E,,,.J 60 min after administration of organophosphate (558 f 65% (n = 5) and 707 k 109% (n = 5) of baseline values, for Pth and Pox, respectively). During the next 60 min, it decreased. The time course of Pox-induced changes in KfCwas similar to that obtained with Pth. The concentration-response curve (E,,,) expressed as a percentage of the baseline value versus the logarithm of the molar Pth concentration, ranging from 2 X 1O-5 to 4 X 10m4 M) was linear (y = 1661.1 + 327.3x, r = 0.89,~ < 0.001, n = 14). Piperonyl butoxide (4 X 10e4 M), an inhibitor of cytochrome P450, had a strong protective effect against Pth (4 X 10e4 M)induced alterations of endothelium permeability (n = 5, p < 0.001). The effects of Pox (4 X 10 m4M) on KfCwere completely abolished by pretreatment with 10m5M atropine, as shown by the significantly lower E,, value recorded in atropine-pretreated lungs ( 129 -t 33%, n = 4) than in Pox-treated lungs (707 f 109%, n = 5, p < 0.001). The effects of Pth, on the other hand, were only partially inhibited, since the E,,,,, value recorded in atropine-pretreated lungs (196 k 20%, II = 4) remained significantly higher than that recorded for control lungs (129 + 15%; n = 5; p < 0.05). These results show that isolated and perfused rabbit lungs constitute an appropriate model for studying the direct pulmonary effects of organophosphates. The edema-inducing action of Pth depends on its activation by conversion to Pox in the lung tissue. It can be explained by an increase in

endothelium permeability. This effect is mediated principally by muscarinic receptors. 10 1992 Academic press. I~C. Organophosphates are widely used in agriculture and veterinary medicine as insecticides or as anthelminthics. Accidental or voluntary poisonings have become a serious public health problem due to their number and to the severity of symptoms. Moreover, people and animals chronically exposed to these substances often suffer respiratory troubles or nervous symptoms which are sometimes irreversible (Bartle, 1991) . The anticholinesterasic properties of organophosphates, their impact on the central and autonomic nervous systems, and their acute effects on the respiratory airways are well documented. The muscarinic effects of cholinesterase inhibitors may explain some functional changes, such as bronchoconstriction and increased respiratory secretions. Pulmonary edema, however, also described in cases of acute parathion or diazinon poisoning of men or animals, remains unexplained so far ( Abdelsalam, 1987; Bledsoe and Seymour, 1972; Delaunois et al., 1992). This phenomenon plays a role in the development of acute respiratory failure, which can lead to death of poisoned patients. Pulmonary edema could result from hypoxia, heart failure, effects on the central nervous system, or direct toxicity to the lungs. The aim of the present study was to investigate: ( 1) the effects of parathion (Pth) and its active metabolite, paraoxon (Pox), on endothelium permeability and on partitioning of the vascular resistance in isolated, perfused rabbit lungs; (2) the pharmacological mechanisms by which cholinesterase inhibitors can induce pulmonary edema; and (3) the role of pulmonary biotransformation of Pth in drug toxicity. MATERIALS

AND METHODS

Gene& procedure. Forty albino rabbits (male and female) weighing 2.5-3 kg were anesthetized with a single intramuscular injection of fentanyl (0.2 mg/kg) plus fluanisone ( 10 mg/kg) (Hypnorm, Janssen Pharmaceutics, Beerse, Belgium). The trachea of each animal was isolated and cannulated.

161

0041-008X/92 $5.00 Copyright 0 1992 by Academic Press. Inc. All rights of reproduction in any form reserved.

162

DELAUNOIS.

GUSTIN,

Pancuronium (0.2 mg/ kg) was administered into the marginal ear vein to prevent agonized movements. The animals were ventilated with a small animal respirator (Palmer, Analis, Namur, Belgium). The chest was opened by midsternal incision and 2000 U/kg of Heparin (Liquemine, Roche, Belgium) was injected into the right ventricle before exsanguination. The heartlung block was rapidly removed from the chest and weighed. The two ventricles were opened and glass cannulae (L, 15 mm; ID, 3 mm) secured in the pulmonary artery and left atrium via the corresponding ventricles. The surrounding tissues were included in the ligatures of the pulmonary artery and left atrium, so as to decrease the compliance of these structures (Wangensteen et al., 1977; Kern et al., 1984; Rock et al., 1985). The isolated heart-lungs were suspended on one side of a beam balance. The lung weight was counterbalanced and further weight changes occurring during the experiments were measured by an isometric force transducer (Palmer, Analis, Namur, Belgium) attached to the lever of the balance. This systemwas calibrated before each experiment by suspending weights in place of the lung so that an increase of 1.5 g in lung weight was indicated by a deflection of 1 cm on the recorder paper. The heart-lungs were connected to a recirculating perfusion circuit and perfused with a constant blood flow (20 ml/min/kg). The circuit included an open venous reservoir, an electronic roller pump (NY-7550-62 Masterflex, Bioblock; Illkirch, France) indicating the blood flow, a heat exchanger, and a bubble trap. Care was taken to avoid introducing air into the circuit. The lungs were first flushed until the remainder of blood in the vascular bed was completely removed, and then perfused with 200 ml of artificial blood-free perfusate ( Krebs-Ringer bicarbonate buffer containing 2.5% bovine albumin (Sigma Chemical Co, St Louis. MO). The liquid in the venous reservoir was continually aerated with carbon dioxide and oxygen. The pH and temperature of the perfusate were continually monitored and maintained within physiological ranges (pH 7.47.5; temperature, 37-38°C). The pH was controlled by adding 1 N NaHC03 or by increasing the flow of carbon dioxide. The lungs were ventilated with humidified air at a frequency of 45 cpm, a tidal volume of 5 ml/kg, and an end-expiratory pressure (P,) of 2 cm HzO. They were periodically hyperinflated to avoid atelectasis and covered with a plastic bag to prevent evaporative fluid loss. Arterial (P,) and venous (P,) pressures were measured by means of two thin catheters connected to side ports of the cannulae secured in the pulmonary artery and left atrium and to pressure transducers (P23 ID and P50, Gould, Brussels, Belgium). The pressures were zero referenced at the level of the lung hilus. Equilibration, obtained after IO-15 min. was characterized by an isogravimetric state, i.e., one in which no gain or loss of weight was observed. and by stability of arterial and left atria1 pressures in zone III conditions (P, > P, > P,). In this study, P, and P, ranged from 10 to 17 cm HZ0 and from 3 to 8 cm HzO, respectively. All parameters were recorded on a Gould recorder (TA 2000, Brussels, Belgium). Two lungs which developed high arterial pressure (P, >20 cm HzO) at the onset of perfusion were rejected from the study. Experimental measurements. The total pressure gradient across the vascular bed ( AP,) was partitioned into four components (arterial, precapillary. postcapillary, and venous segments) by application of the double/arterial/ venous occlusion method used by Hakim et al. ( 1982) in dogs and by Gustin et a/. ( 1992) in pigs. Briefly, inflow and outflow occlusions were successively performed using two electromagnetic valves placed around the arterial and venous tubes, near the cannulae, for a brief period (2-3 set) at the end of expiration. When inflow was stopped, the P, first dropped rapidly by an amount equal to the pressure gradient through the large arteries ( AP, = Pa - Par),and then more slowly. due to lung compliance. Clamping the outflow induced an abrupt increase in P, equal to the pressure gradient through the large veins ( AP, = P,, - Pv), followed by a slow rise in P,due to progressive filling of the vascular bed with perfusion liquid. Pa,and P,, are respectively defined as the pressures at the distal end of the arterial segment and the proximal end of the venous segment. Detailed plots of pressure variations obtained in dogs and pigs after arterial and venous occlusions have been published by Hakim ( 1988) and Gustin et a/. ( 1992). The segment including the capillaries, the small arteries, and the small veins, all highly extensible

AND ANSAY

vessels,was divided into pre- and postcapillary components by measuring capillary pressure (PC) by means of the double occlusion method (Dawson et al., 1982). Brief, simultaneous occlusion of both inflow and outflow (for a few seconds) induces an increase in P, and a decrease in Pauntil Paand P, equilibrate at the same pressure, a good estimate of capillary pressure. Resistancesof each segment were calculated by dividing the pressure gradients by the blood flow (Q): Total vascular resistance

R, = -pa- P”

Q

Arterial resistance

R = pa - pa, a Q Precapillary resistance

P,,- PC

R,, = -

Q

Postcapillary resistance

RV’= PC- pvQ Venous resistance R

=P”r-P” ”

Q

However, as the blood flow was constant, the resistances across the various segments of the pulmonary vascular bed are directly proportional to the pressure gradient measurements (AP,; AP,; AP,,; AP,,; AP”). The capillary filtration coefficient ( Krc) was measured by increasing Pa and P,, usually by IO-15 cm Hz0 for 2 to 4 min. PCwas estimated before and at the end of this period in order to determine the hydrostatic pressure difference ( AP,) inducing fluid filtration acrossthe endothehum. The amount of fluid filtering across the endothelium was calculated from the weight-gain plot. At first, due to congestion, the lungs gained weight rapidly for 30 to 60 set following the vascular pressure increases. Thereafter, lung weight increased more slowly due to subsequent fluid filtration. The slopes of the slow phase of the weight increase ( AW/ At) were plotted on a semilogarithmic scale as a function of time and extrapolated to zero time. The value obtained was divided by AP, and by lung weight to yield Kfcexpressed in milliliters (the fluid filtered was assumed to have a density of 1.O) per minute per cm Hz0 per 100 g wet lung weight. Details of this method have been published (Drake and Gabel, 1981; Sumita et al., 1989; Gustin et al., 1992). Plots of lung-weight variation with increasing PCare shown in Fig. 1. Experimental design. Six groups of rabbits were used. In each group, the segmented pressure gradient through the pulmonary circulation and the Kfcvalue were measured at 30-min intervals for 2 hr following measurement of baseline values (to). In the first group (Pth group) (n = 14), we assessedthe effects of Pth on endothelium permeability and on the distribution of the pressure gradient across the various segments of the vascular bed. Just after baseline value measurement, a single dose of drug was added in the venous reservoir. The final concentrations used (in the 200-ml volume of perfusion liquid ) were 2 X 10msM (n = 3), 6 X 10s5 M (n = 3), 10m4M (n = 3), and 4 X 10M4M

(n = 5).

ORGANOPHOSPHATE-INDUCED

PULMONARY

163

EDEMA

TABLE 1 Baseline Values of Total Vascular Resistance (R,), Arterial (R& Precapillary ( R,), Postcapillary (&) and Venous (RJ Resistances Expressed as (cm HzO. liter-’ . min-’ . 100 g-‘) and Capillary Filtration Coefficient (&) (ml. cm H,O-’ . min’ e 100 g-‘) Group of animals Group 1: Pth group (n = 14) Group 2: Pox group (n = 5)

R

&

Rt

&

RV

Kt,

11.94 + 1.12

3.52 F 0.54

2.78 k 0.35

1.59 * 0.22

4.05 f 0.55

3.48 f 0.28

15.55 + 1.25

4.85 k 0.44

3.86 -c 0.17

2.61 k 0.51

4.17 + 0.32

3.22 f 0.93

11.41 +- 1.38

4.34 k 0.52

3.57 + 0.18

1.74 + 0.18

I .76 + 0.24

2.29 f 0.42

14.58 k 1.41

3.67 + 1.06

3.21 k 0.25

2.92 i 0.71

4.78 i 0.22

3.77 f 0.62

15.22 f 0.39

3.92 + 0.77

3.37 2 0.21

3.31 i 0.61

4.56 3~ 0.39

3.71 f 0.41

14.25 zk 2.11

4.90 +_ 0.75

4.87 + 0.71

1.19 +_ 0.32

3.29 2 0.43

3.17 + 0.66

13.82 + 0.7 1

4.24 + 0.28

3.63 f 0.29

2.22 + 0.28

3.73 f 0.39

3.21 + 0.22

Group 3: paranitrophenol (n = 2)

Group 4: PBO group (n = 6) Group 5: atropine group (n = 8) Group 6: control group (n = 5)

Mean for all groups (n = 40)

Note. All values are means f SEM.

In the second group of lungs (Pox group) (n = 5). a single dose of paraoxon (4 X 10m4M) was added in the venous reservoir after baseline value measurement. In the third group (n = 2), paranitrophenol, the principal product of Pth degradation, were used in the same manner in order to verify its absence of effects. In the fourth group (n = 6), to assessthe role of the pulmonary bioactivation of Pth into Pox in drug toxicity, piperonyl butoxide (4 X 10m4M), an inhibitor of cytochrome P450, was added in the venous reservoir 15 min prior to administration of 4 X 10e4 M Pth (n = 4) or Pox (n = 2). Baseline values were measured between the two administrations of drugs. In order to determine the role of acetylcholine in the reaction induced by the organophosphates tested in this work, eight lungs received a bolus of atropine (IO-’ M) before the baseline value measurements. The lungs were then treated with paraoxon (Pox-atropine group; 4 X 10e4 M; n = 4) or parathion (Pth-atropine group: 4 X 10s4 M; n = 4). Fifteen minutes separated the two additions. The propylene glycol (0.2 ml) used in this study as a vehicle was tested for any effect on the various parameters measured and found to have none (group 6; control group, n = 5). Drugs. Parathion-ethyl (Riedel-de Haen, Seelze. Germany), paraoxon ( Riedel-de Hdn, Seelze. Germany), paranitrophenol (Sigma Chemicals) and piperonyl butoxide (Sigma Chemicals) were dissolved and diluted in propylene glycol (0.2 ml). Atropine sulfate was purchased from Merck (Germany). The drug was dissolved in 1 ml distilled water and then diluted 10 times to obtain a concentration of IO-’ M in the circuit. The purity of the organophosphates used was checked by gas chromatography. Sratisrics. Data are expressed as mean values ? standard error of the mean ( SEM ) . Results were subjected to a two-way variance analysis (ANOVA). When the ANOVA was significant, a Student t test was used to compare the means. When it was necessary, the Q values for significance were corrected by Bonferroni’s method to allow multiple comparisons. Differences were considered significant when p < 0.05.

RESULTS

Hemodynamics and the Capillary Filtration Coeficient: Baseline Values The mean baseline value of total vascular resistance (R,) recorded for all groups of rabbit lungs together (n = 40) was

13.82 + 0.7 1 cm Hz0 * liter-’ . min-’ . 100 g-’ lung weight (Table 1). The respective resistances of the arterial (R,), precapillary (R,~ ) , postcapillary (R, 0, and venous (R,) segments, for all groups together, were 30.67 f 2.03%, 26.28 f 2.12%, 16.05 t 2.01%, and 27.00 + 2.82% of R,. Thus, the vessels located upstream from the capillaries (R, and R, !) accounted for a significantly greater percentage of R, (56.95 -t- 4.02% ) than did those located downstream (R,t and R,) (43.05 + 4.02%) (p < 0.05). The mean baseline Kf, value was 3.27 k 0.22 ml. min-’ ~(cm~H20)~‘~100g-‘.Nosignificant difference was observed between the different groups.

Efects of Parathion on Lung Weight, Segmental Vascular Resistances,and IQ, Figure 1 shows typical plots of lung-weight variations throughout the experiment in one control lung and one Pthtreated lung (drug concentration, 4 X 10M4 M). During the & measurements, lung weight increased biphasically due to pulmonary congestion and edema. When the pressures returned to their baseline values, lung weight decreased rapidly to a level depending on the amplitude of edema produced during the Kf, maneuvre. As shown in Table 2, the fluidfiltration-inducing increase in hydrostatic pressure ( AP,) and the duration of the Kf, maneuvre were similar in both groups. However, the amount of fluid filtered through the endothelium, as assessed by the difference between lung weights measured before and after the i& maneuvre ( AW), was significantly greater in Pth-treated than in control lungs (Fig. 1, Table 2) (ANOVA 2, p < 0.00 1). In the control group, no further weight change was observed during the periods of 30 min separating each Kf, measurement. In contrast, the weight of treated lungs continued to increase slightly and

164

DELAUNOIS,

GUSTIN,

hemodynamics. The only parameter to be significantly altered by Pth was &, for which a significantly increased value was recorded 30 mm after Pth administration (p < 0.01). A maximum effect of (E,,,,,) was obtained after 60 min (p < 0.00 1). lower values being recorded at 90 and 120 min. The Pth-induced effects were dose dependent, although the time course was similar (Fig. 3) for all concentrations used: 2 X 10-5, 6 X 10-5, 10-4, and 4 X 10m4 M. The doseresponse curve, calculated by linear regression between the individual E,,, values (expressed as percentages of the baseline Kf, values) and the logarithm of the administered dose was highly significant (r = 0.89; p -C 0.00 1) (Fig. 3 ). The mean Em,, _+ SEM values calculated for the various doses ofPth were 146 k 6% (p < 0.01). 248 I? 25% (a < O.Ol), 347 k 11% (n < O.OOl), and 558 f 65% of baseline (p < 0.001) at 2 X 10p5, 6 X 10-5, 10e4, and 4 X 10m4 M, respectively.

64

56

46. 5 E .P

Kf.c 3 min

40.

t p 1 6

32.

.’ zi

24

16-

6-

1

AND ANSAY

-

-

-

-

time

FIG. 1. Typical plots of lung weight gain recorded throughout the experiment in a control lung (-) and in a parathion (Pth)-treated lung (4 X 10m4M) (- - - -). Capillary filtration coefficient ( Kfc) measurements were performed at 30-min intervals following drug administration. AW represents the difference between lung weights measured before and after the Kfc maneuvre. In the control group, no further weight change was observed during the periods of 30 min separating each Kfc measurement. In contrast, the weight of treated lungs continued to increase slightly and spontaneously until a new isogravimetric state was reached. For details, see text.

spontaneously until a new isogravimetric state was reached (Fig. 1). Figure 2 shows the time courses of variations in the various segmental vascular pressures and of changes in Kf, values for control and Pth-treated lungs (Pth concentration, 4 X 1O-4 M). The drug did not significantly influence pulmonary

Efects of Paraoxon, Paranitrophenol, and Piperonyl Butoxide on K/C and on the Partitioning of Vascular Resistance Time plots of KfCchanges after Pox and Pth administration (drug concentration, 4 X lop4 M) are illustrated in Fig. 4. The maximum effect induced by Pox tended to be higher than that recorded after Pth administration. The difference, however, was not significant. Moreover, we observed no difference in the profile of the time plots obtained with the two drugs. After Pox administration, the vascular pressures in the various pulmonary circulation segments were not significantly altered. Paranitrophenol, administered at a dose of 4 X 10m4 M, had no significant effect on vascular hemodynamics or on endothelium permeability (Fig. 4). The KfC values obtained

TABLE 2 Increaseof Capillary Pressure Inducing Fluid Filtration (hp,), Duration of the Measurementsof the Capillary Filtration Coefficient (K,), and Difference in the Lung Weight Measured Before and After the Kfc Maneuver (AW) at 30-min Intervals in Control Lungs and in Pth-Treated Lungs (4 X 10m4 M)

APc (cm H20) Control Pth Duration (min) Control Pth AW W Control Pth

Baseline

30 min

60 min

13.18 f 1.38 11.25 rt 1.29

8.75 !I 1.75 10.58 + 1.29

14.25 i- 2.0 9.9 2 1.19

2.96 -+ 0.32 3.04 f 0.17

3.4 f 0.4 2.94 f 0.30

3.24 2 0.15 2.94 k 0.24

1.27 t 0.34 1.72 + 0.71

1.27 + 0.36 10.82 + 3.68;

I.84 i- 0.51 9.36 + 2.18**

Note. All values are means + SEM. * Value different from the control value measured at the same time (p < 0.05). ** Value different from the control value measured at the same time (p < 0.01). *** Value different from the control value measured at the same time (p < 0.00 I).

90 min 10.0 t 0.1 9.55 + 0.6 3.35 * 0.35 2.98 k 0.22 0.83 i 0.15 11.06 i 0.88***

120 min 14.69 + 1.34 13.72 k 1.19 3.22 of-0.22 3.1 kO.17 1.71 rt 0.52 6.80 rk 2.39

ORGANOPHOSPHATE-INDUCED

PULMONARY

Pa

165

EDEMA

Pa’

3o a

1

”

b

25-

10

I 0

1 30

I I 60

1 , 120

I 1 90

time

(min)

dI

10

5

5. 0

120

90

time

15

-(+

10

60

30

Pv’

PC C

15

0

(min)

3b

6b

9'0

time

120'

: 0

I

1

I

30

60

90

time

(min)

Pv

1

120

(min)

Kf.c

l2 e

8

4 I 0

I 30

I 60

I 90

time

1 120

(min)

“I

0

3'0

.

6'0

9'0

time

1;o

(min)

FIG. 2. Time courses of variations in arterial (P,), precapillary (I’.,), capillary (P,) , postcapillary (P,,) , and venous (P,) pressures (expressed as cm HzO) and of the capillary fihration coefficient (I&) (expressed as percentage of baseline values) in controls ( n = 5 ) (III) and in parathion (Pth)-treated lungs (n = 5 ) ( + ) . Value different from control value recorded at the same time (*, p i 0.05, **, p < 0.0 I, +*t, p < 0.00 1) . Value different from the baseline value (A, p -c 0.05; AA, p < 0.0 1).

166

DELAUNOIS.

GUSTIN.

AND ANSAY

Pox. The Em,, KfC value recorded in the two lungs which received PBO before administration of Pox ( 7 17 + 65%) was similar to that obtained with Pox alone (707 + 109%).

Kf.c

600

Efects of Parathion Lungs

500 400 300 200 -I

100

0’ 0

30

60

90

120

time

(min)

Emax

000

q

and Paraoxon in Atropine-Pretreated

The effects of Pox on KfC were completely abolished by pretreatment with atropine (Fig. 6): the recorded mean Em,, value was significantly higher in the Pox group (707 + 109% of the baseline value) than in the Pox-atropine group ( 129 + 33% of baseline) (p < 0.01). No difference was observed between the Pox-atropine group and the control group. In contrast with these results, atropine-caused inhibition of Pthinduced effects was not complete. The mean E,,,,, value was reduced from 558 t 65% in the Pth group to 196 5 20% in atropine-pretreated lungs (p < 0.01 ), but the latter value remained significantly higher than the control value ( 129 + 15%) (p < 0.05).

1

DISCUSSION

600

All previous reports of organophosphate-induced pulmonary edema have been based on clinical studies (Abdelsalam, 1987; Tsao et al., 1991; Delaunois et al., 1992) and recently has been described in anesthetized dogs after intravenous injection (Lainee et al., 199 1). Here, we have reproduced the phenomenon in vitro, thus opening the way to investigating the mechanisms involved.

400

200

0: -5,0

I

I -4.5

I

-4,0

-3.5

log

, -3.0

concentration

FIG. 3. The top represents the time course of capillary filtration coefficient (KrJ changes (values expressed as percentages of the baseline value) in control lungs (n = 5) (W) and parathion (Pth)-treated groups. The concentrations used are 4 x lOA M (n = 5) (m), lOA M (n = 3) (*), 6 x 10e5 M (n = 3) (U), and 2 x 10v5M (n = 3) (0). Value different from control value recorded at the same time (k, p < 0.05; It*, p < 0.01; Irk*, p < 0.001). Value different from the baseline value (A, p < 0.05; AA, p < 0.01). The bottom represents the concentration-response curve. Abscissae, logarithms of the molar drug concentrations in the perfusion liquid. Ordinates, maximum effects (E,,) on I&,, (expressed as percentage of baseline values) recorded 60 min after drug administration. The equation of the regression line is y = 1661.1 + 327.3 x (R = 0.89).

Kf,c

800

‘1: 0

after administration of paranitrophenol were not different from control values. Piperonyl butoxide had a strong protective effect against Pth-induced alteration of endothelium permeability (Fig. 5 ) . The mean E,,, value for KfC recorded in lungs pretreated with PBO ( 168 + 11% of baseline) was significantly (p < 0.001) smaller than the corresponding value measured in the Pth group (558 f 65%). It nevertheless remained significantly greater than the value recorded in the control group ( 129 + 15% ) (p < 0.05 ) . PBO did not alter the response to

30

60

120

90

time

(min)

FIG. 4. Time courses of variations of capillary filtration coefficients (K,) expressed as percentage of baseline values) recorded at 30-min intervals in the control group (n = 5) (O), in the parathion (Pth)-treated group (Pth concentration, 4 x lo4 M) (n = 5) (B), in the paraoxon (Pox)treated group (Pox concentration, 4 x 10m4M) (n = 5) (e), and in the paranitrophenol-treated group (4 x 10m4M) (n = 2) (0). No significant difference was observed between the time courses for Pth and Pox. Value different from control value recorded at the same time (*, p < 0.05; **, p < 0.01; *it*, p < 0.001). Value different from the baseline value (A, p < 0.05; AA,p < 0.01).

ORGANOPHOSPHATE-INDUCED

AA ***

T

600

167

EDEMA

A comparison of the baseline values obtained in this study with those recorded in dogs and pigs makes it possible to evaluate physiological factors which may predispose or protect these species to pulmonary edema. The absolute total resistance value calculated in canine lungs from literature averages 16.5 cm Hz0 . liter-’ * min-’ . 100 g-’ (Rippe et al., 1987; Sumita et al., 1989). Our data show that the mean R, value recorded in rabbits ( 13.82 +- 0.7 1 cm H20. liter’ . min-’ - 100 g-‘) is similar to that obtained in dogs, but smaller than that in pigs (55.2 cm H20. liter-’ . min-’ *

Kf,c 700

PULMONARY

EOO-

01 0

30

60

90

time

120

700-

(min) 600-

FIG. 5. Variation of capillary filtration coefficient (K,) values (expressed as percentage of baseline values) as a function of time in control lungs (n = 5) (0) and after administration of 4 x 10m4M parathion (Pth) with (n = 5) (+) or without (n = 5) 03) pretreatment with piperonyl butoxide (PBO) (4 x 10e4 M). PBO alone had no effect (data not shown in this figure). Value different from control value recorded at the same time (k, p < 0.05; Ir+, p < 0.01; *t*, p < 0.001). Value different from the baseline value (A, p < 0.05; A&p < 0.01). Value different from the value recorded at the same time in Pth or Pox groups without pretreatment (WB, p < 0.01, n rnrn, p < 0.001).

5004003002001004 0 I 0

1 60

1 90

time

As shown by the Starling equation

Jv = &c[(pc - pt) -

‘Jd(rp

-

Tt)],

fluid exchange across capillary walls (J,) depends on the capillary filtration coefficient (&), the capillary and interstitial hydrostatic fluid pressures (PC, Pt), the osmotic pressure in tissue fluids ( 7rt) and plasma (r,,), and the osmotic reflection coefficient ( ad). Pulmonary edema could thus result from increased endothelial permeability and/ or capillary pressure. Capillary pressure depends on cardiac blood flow and on the segmental distribution of the pressure gradient across the pulmonary vascular bed. In the isolated, perfused rabbit-lung preparations developed in this study, endothelium permeability was evaluated by directly measuring the capillary filtration coefficient. We further established the distribution of resistance to blood flow in the small and large vessels located upstream and downstream from the capillaries by applying the model of Hakim ( 1988) to assess muscletone variations in vascular smooth muscle at all levels of the vascular bed, as well as variations in capillary pressure. Arterial pressure measurements are not always sufficient for detecting vasoconstriction in a compartment accounting for only a small portion of the total resistance. Moreover, R, does not vary when an increase in the resistance of one segment is counterbalanced by a decrease in another part of the vascular bed (Drake and Gabel, 198 1; Hakim, 1988; Sumita et al., 1989; Gustin et al., 1992).

1 120

(min)

Kf,c 700 600 soo400300zoo1004

04 0

30

60

90

time

120

(min)

FIG. 6. Top: Protective effect of atropine ( 10m5M) against the effects of paraoxon (Pox) on the capillary filtration coefficient (&) expressed as percentage of the baseline values). (IX) control group (n = 5), (+) Pox group (n = 5), (0) Pox-atropine group (n = 4). Bottom: Protective effect of atropine (10m5M) against the effects of parathion (Pth) on the capillary filtration coefficient (&) expressed as percentage of the baseline values. (+) control group (n = S), (l3) Pth group (n = 5), (0) Pth-atropine group (n = 4). Value different from control value recorded at the same time (k, p < 0.05; Ir*, p < 0.01; *+*, p < 0.001). Value different from the baseline value (A,p < 0.05; AA,p < 0.01). Value different from the value recorded at the same time in Pth or Pox groups without pretreatment (m, p < 0.05; n m,p< 0.01; Urn&p< 0.001).

168

DELAUNOIS,

GUSTIN,

100 g-r) (Gustin et al., 1992). In the latter, the vessels located downstream from the capillaries (R,) + R,) accounted for 54.4% of R,. In dogs, these vessels accounted for 53.1% of R, (Hakim, 1988). In rabbits, our results showed that the major site of resistance is located in the vessels upstream from the capillaries, since (R, + R,{) only totalled 43.0% of R,. An obvious conclusion is that the development of capillary hypertension during a vasoconstrictive reaction should be favored in pigs, while rabbits should be relatively protected against this phenomenon. Dogs are in an intermediate situation. Endothelium permeability, however, as assessed by the Kfc value, is significantly greater in rabbits (3.27 + 0.22 ml-min-’ -(cm H@-‘100 g-l) than in dogs (0.11 ml-min-’ -(cm H,O)-’ . 100 g-‘) (Seeger et al., 1986; Sumitaetal., 1989)andpigs(0.19ml~min~‘-(cmH,O)~’~ 100 g-’ ) (Gustin et al., 1992). Our values agree with those mentioned by Kern et al. ( 1984) (2.78 ml * min-’ -(cm H20)-’ - 100 g-l). One might conclude that the high endothelium permeability in rabbits could be factors predisposing these animals to pulmonary edema. Recently, Nemery ( 1987) reviewed ail the toxic effects of organophosphates on the lungs. He concluded that lung injury induced by these compounds could be due to heart failure, hypoxia, central nervous effects, or direct toxicity to the lungs. Without excluding any of these hypotheses, our data clearly demonstrate that parathion directly induces pulmonary edema by increasing the permeability of the endothelium rather than by changing the capillary pressure (Fig. 2 ) . The small spontaneous increase in lung weight recorded in the Pth-treated group between each Kfc maneuvre (Fig. 1) can be explained by this phenomenon. Owing to the Starling equation, it can be seen that when the Kf, is increased, the small fluctuations in PC, often observed after each Kf, measurement, can induce fluid filtration (JV > 0) until a new isogravimetric state is obtained (JV = 0) due to the equilibration of oncotic and hydrostatic pressures. In the control group, these small fluctuations were not sufficient to induce a significant increase in J, because of the small value of Kfc. In the present study, parathion did not develop its effects immediately after administration. The progressive increase in Kfc observed during the first hour following drug administration might be interpreted as the time required by the lung to metabolize Pth to Pox. This latter compound is a more active form since it has an anticholinesterasic effect 15,000 times greater than that for Pth (Eto, 1979). If the delay necessary to reach the maximum effect on the Kfc value is really due to the bioactivation of Pth into Pox, we should have observed an immediate maximal effect after administration of Pox, since this drug is directly active without biotransformation. Our data invalidate this hypothesis, however, since the profile of time courses of Pth- and Pox-induced Kfc changes was similar, as shown by Fig. 4. Another possible explanation is that, acetylcholine being mainly stored in parasympathetic fibers, the inhibition of acetylcholinesterase

AND ANSAY

by Pox had no immediate effect in our preparation. Progressive release of Ach in the pulmonary tissue, resulting in accumulation of the mediator at this level, could be necessary to induce cumulative effects on the target cells (Tanaka and Grunstein, 1984). Involvement of Ach in Pth- and Pox-induced pulmonary edema is demonstrated by the strong inhibition of their Kfcrelated effects by atropine (Fig. 6). However, atropine’s protective effect against Pth was only partial (Fig. 6). The remaining effect on endothelium permeability (35%) in atropine-pretreated lungs may be attributable to sulfur-containing metabolites. Several studies have described toxic effects of the sulfur atom released during Pth metabolism, resulting in inactivation of cytochrome P450 and binding of macromolecules via hydrodisulfide bonds (Kamataki and Neal, 1976; Halpert et al., 1980). Catravas et al. ( 1984) have shown that high concentrations of Ach can induce irreversible edema in isolated and perfused rabbit lungs. This observation is another good argument in favor of the involvement of a cholinergic mechanism in the development of organophosphate-induced edema. Nevertheless, these investigators partly attributed fluid filtration to the hypertensive action of Ach. This was only speculation, however, as endothelium permeability was not measured in their study. The absence of hemodynamic changes in the Pth- and Pox-treated lungs of our study remains unexplained but shows that organophosphate-induced pulmonary edema can be due to changes in the endothelium permeability only. It could be due to differential effects of endogenous and circulating Ach or to some difference in the sensitivity of vascular smooth muscle and endothelial cells to Ach action, since the increased arterial pressure reported by Catravas et al. ( 1984) was only observed at rather high Ach concentrations (greater than 10e6 M). The decrease in the Kfc recorded during the second hour following administration of organophosphates is probably due to overflow of the lung, as suggested by the large increase in lung weight recorded throughout the experiment and by the presence of foam in the trachea. These alterations can be explained by major alveolar injuries which should lead to an infinite conductance of the alveolocapillary barrier. However, accumulation of fluid in the interstitial tissue can induce an increase in P,, an effect which opposes fluid filtration across the endothelium. Under these conditions, the basic principle of Kfc measurement is violated so that the calculated Kfc value is not a good estimation of the actual value. The decrease in the Kfc value recorded after 60 min can be considered as an artefact. We observed a linear relation between the E,,,, values for Kfc and the logarithm of the Pth concentration (Fig. 3 ). The lowest dose eliciting a significant effect on Kfc was 6 X 10e5 M, corresponding to a concentration of 17.5 mg/ml in the perfusion liquid. Toxicokinetic and metabolic studies of Pth in the rabbit (Neal, 1972; Pena-Egido et al., 1988) indicate

ORGANOPHOSPHATE-INDUCED

that this concentration of Pth in the perfusion liquid is approximately that which would be obtained in the plasma after intravenous administration of 10 mg/kg of Pth, the LDzO of Pth in the rabbit. We may thus suspect that the mechanism of organophosphate action identified in vitro could also be involved in vivo. In previous biochemical studies on lung metabolic capacity, it has been demonstrated that Pth is activated, i.e., converted to Pox, by cytochrome P450 located in the microsomal system of the lung and that piperonyl butoxide, administered at a concentration of 4 X lop4 M, prevents 87% of Pox formation in rabbit-liver microsomes (Nakatsugawa and Dahm, 1967; Neal, 1972). However, the functional consequences of this inhibition on the toxicity of Pth have never been explored. The conversion of Pth to Pox is a major determinant in the cascade of events leading to pulmonary edema, as illustrated by the preventive effect (70%) of piperonyl butoxide (4 X 1O-4 M) against the effects of Pth on Kfc (Fig. 5). This could be a new approach to preventing or treating organophosphate poisonings. We conclude that: ( 1) isolated and perfused rabbit lungs constitute an appropriate model for studying the direct pulmonary effects of organophosphates; (2) organophosphate-induced pulmonary edema can be explained by an increased permeability of the endothelium rather than by hemodynamic changes; ( 3) the protective effect of atropine against Pth and Pox demonstrates the involvement of Ach in the reaction. However, sulfur metabolites released during Pth transformation could also contribute to the development of pulmonary edema; and (4) activation of Pth by conversion to Pox is a major determinant of the reaction, as shown by the almost complete inhibition by piperonyl butoxide. ACKNOWLEDGMENTS The authors thank Dr. Nemery for his valuable advice on the protocol. Thanks are also due to S. Bloden and 0. Bonnie for their technical assistance. Financial support was provided by Special Funds for Research from the University of Liege.

REFERENCES Abdelsalam, E. B. (1987). Organophosphorus compounds. 1. Toxicity in domestic animals, Vet. Res. Commun. 11, 211-2 19. Battle, H. ( 199 1) Quiet sufferersof the silent spring. New Sci. 130, ( 1769 ) , 30-35. Bledsoe, F. H., and Seymour, E. Q. ( 1972). Acute pulmonary oedema associated with parathion poisoning. Radiology 103, 53-56. Catravas. J. D.. Buccafusco, J. J., and El Kashef, H. ( 1984). Effects of acetylcholine in the pulmonary circulation of rabbits. J. Pharmacol. Exp. Ther. 231(2),

236-241.

Dawson, C. A., Linehan, J. I-I., and Rickaby, D. A. (1982). Pulmonary microcirculatory hemodynamics. Ann. N. Y. Acad. Sci. 384, 90- 106.

PULMONARY

EDEMA

169

Delaunois, A., Gustin, P., Bloden, S.. and Ansay. M. ( 1992). Fait clinique: Intoxication par le parathion dans une exploitation bovine. &tn. Med. Vet. 136, 137-138.

Drake, R. E.. and Gabel, J. C. ( 198 I). Comparison of technic to measure pulmonary capillary filtration coefficient in dogs. Microvasc. Res. 21, I33141. Eto, M. ( 1979). Organophosphorus pesticides. In Organic and Biological Chemistry (G. Zweig, Ed.), pp. 387. CRC Press. Boca Raton. FL. Gustin, P., Urbain, B., Delaunois. A., Zeimes, K., and Ansay, M. ( 1992). Permeability of the endothelium and partitioning of the pulmonary blood flow resistance in isolated and perfused pig lungs: Effects of breed and age. Vet. Re.r. Commun. 16, 69-82. Hakim. T. S. (1988). Identification of constriction in large versus small vesselsusing the arterial-venous and double-occlusion techniques in isolated canine lungs. Respiration 54,6 1-69. Hakim, T. S., Michel, R. P.. and Chang, H. K. (1982). Partitioning of pulmonary vascular resistance in dogs by arterial and venous occlusion. J. Appl. Physiol. 52, 7 IO-7 15. Halpert, J.. Hammond, D., and Neal, R. A. ( 1980). Inactivation ofpurified rat liver cytochrome P450 during the metabolism of parathion. J. Biol. Chern. 255, 1080-1089. Kamataki, T., and Neal, R. A. ( 1976). Metabolism of parathion by a reconstituted mixed-function oxidase enzyme system:Studies of the covalent binding of the sulfur atom. Mol. Pharmacol. 12, 933-944. Kern. D. F., Kivlen. C.. and Malik, A. G. (1984). Pulmonary capillary filtration coefficient (Kf,c): Blood vs albumin-Ringer’s. Microvasc. Res. 21, 248. Lainee, P., Robineau, P., Guittin, P., Coq, H., and Benchetrit, G. ( 199 1). Mechanisms of pulmonary edema induced by an organophosphorus compound in anesthetized dogs. Fundam. Appl. To.xicol. 1 I, 177- 185. Nakatsugawa, T., and Dahm, P. A. (1967). Microsomal metabolism of parathion. Biochem. Pharmacol. 16, 25-38. Neal, R. A. ( 1972). A comparison of the in vitro metabolism of parathion in the lung and in liver of the rabbit. To.xicol. .4ppl. Pharmacol. 23, 123130. Nemery, B. ( 1987). Lung toxicity of trialkyl phosphorothioates. In Toxicology of Pesticides: E,xperimental, Clinical and Regulatory Aspects (L. G. Costa, and C. L. Galli, Eds.), Springer-Verlag, New York. Pena-Egido, M. J.. Rivas-Gonzalo, J. C., and Marino-Hernandez. E. L. ( 1988 ). Toxicokinetics of parathion in the rabbit. Arch. Toxicol. 61, l96200. Rippe, B., Parker, J. C.. Townsley, M. I., Mortillaro, N. A., and Taylor. A. E. ( 1987). Segmental vascular resistancesand compliances in dog lungs. J. Appl. Physiol. 62, 1206-1215. Rock, P., Patterson, G. A., Permutt, S., and Sylveter, J. T. ( 1985). Nature and distribution ofvascular resistance in hypoxic pig lungs. J. Appl. Physiol. 59, 1891-1901. Seeger, W., Walmrath, D., Menger, M., and Neuhof, H. (1986). Increased lung vascular permeability after arachidonic acid and hydrostatic challenge. J. Appi. Physiol. 61, 1781-1789. Sumita, T., Ishikawa, N., Hashiba, Y., Takagi. K., and Satake, T. ( 1989). S-Hydroxytryptamine receptor subtypes participating in pulmonary edema in dogs. Eur. J. Pharmacol. 164, 69-75. Tanaka, D.. and Grunstein, M. ( 1984). Mechanisms of substance P-induced contraction of rabbit airway smooth muscle. J. Appl. Pttvsiol. 57, 155 l1557. Tsao, T., Juang, Y., Lan, R.. Shieh, W., and Lee, C. ( 1991). Respiratory failure of acute organophosphate and carbamate poisoning. Chest 98, 63 l-636. Wangensteen, 0. D.. Lysaker, E., and Davaryn, P. ( 1977). Pulmonary capillary filtration and flexion coefficients in the adult rabbit. Microvase. Res. 14,81-97.