INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Acute lung injury during adult respiratory distress syndrome (ARDS) is characterized by an increase in pulmonary capillary permeability (1) leading to non-cardiogenic pulmonary edema and by an increase in pulmonary artery pressure (PAP) causing an alteration of right ventricular function (4, 5). Although ARDS can occur under neutropenic conditions (6, 7) neutrophils have been implicated in the development of acute lung injury (1, 8) and have been identified as key cells that trigger and perpetuate inflammatory processes (12). The release of chemotactic substances from the endothelium and from resident neutrophils causes sequestration of neutrophils in the pulmonary microvasculature (8, 13). This is considered as an initiating event of ARDS (14) and is accompanied by a remarkable increase in the count of neutrophils and of myeloperoxidase (MPO) activity in the bronchioalveolar lavage fluid (14).

Activated neutrophils may affect the surrounding lung tissue via several potentially pathogenic cellular systems including the production of prostanoids (15), the release of lysosomal proteolytic enzymes (11, 16) and the generation of highly reactive oxygen radicals and intermediates (8, 10, 11, 17). During the oxygen burst of neutrophils several highly reactive radicals and intermediates are generated (17). Hypochlorous acid (HOCl) is synthesized by a MPO-mediated reaction from hydrogen peroxide and chloride. It is the main product of activated neutrophils (16) and is much more reactive than hydrogen peroxide (18). Whereas oxygen radicals cause lipid peroxidation of the liquid phase of the cell membrane, which will lead to a nonspecific increase in the membrane permeability and to cell damage, HOCl as non-radical oxidant causes oxidative modification of free functional groups (19) and subsequent functional alteration of proteins (20). Therefore HOCl that is not expected to cause non-specific effects or cell damage due to lipid peroxidation is a useful model of oxidative stress. Different studies deal with the implication of neutrophils in several models of acute lung injury and provide some evidence that neutrophil-induced changes are triggered by reactive oxygen metabolites (21). The arachidonic acid metabolism was found to be stimulated by reactive oxygen metabolites (24). Highly reactive oxygen metabolites are not only released by PMN but also by endothelial cells. Several models of oxidative stress like the xanthine oxidase system generating superoxide radical (26) or hydrogen peroxide (25) have been investigated. These oxidants are released by other sources than neutrophils as well. However the influence of HOCl that is a neutrophil specific oxidant formed by the MPO on the pulmonary microcirculation has not yet been estimated. Therefore this study characterizes on the one hand the effect of continuous HOCl infusion and on the other hand the effect of the activation of neutrophils having been delayed in the pulmonary microvasculature on the circulation in an isolated rabbit lung model. Furthermore it compares the effects of both stimuli to assess the reproducibility of neutrophil-derived oxidative stress by HOCl.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Isolated Rabbit Lung Model

Preparation of isolated rabbit lungs was performed using the method described in detail by Seeger and coworkers (27). Rabbits of either sex between 2.5 and 3.5 kg were used. The lungs were ventilated with 4% CO₂, 17% O₂, and 79% N₂ (tidal volume 30 ml, frequency 30 min¹). Perfusion flow was gradually increased to the definite flow rate of 100 ml · min¹ with recirculation of the buffer medium (Krebs-Henseleit buffer, total circulating volume 300 ml). Perfusate temperature was 37° C. Pulmonary arterial pressure (PAP), pulmonary venous pressure (PVP), inflation pressure and the weight gain of the isolated organ were continuously registered. After a steady-state period of 30 min only those lungs were selected for the study that showed no signs of leakage at the catheter insertion sights, that had a homogenous white appearance with no signs of edema formation and that had lost weight during the phase of temperature increase and were completely isogravimetric during the steady-state period. After steady-state the PAP ranged between 7 and 10 mm Hg. The inspiratory peak inflation pressure was set 7 to 9 mm Hg. The PVP was adjusted to 2 mm Hg.

Hydrostatic Challenge

Capillary filtration coefficient (K_f,c) was determined gravimetrically from the slope of lung weight gain after a sudden venous pressure elevation of 10 cm H₂O for 8 min (28). K_f,c was related to wet weight lung (WWL) that was calculated from the body weight (BW) using WWL = BW · 0.0024. K_f,c was given as 10⁴ · ml · s¹ · cm H₂O¹ · g¹.

Vascular compliance (C) is defined as the change in vascular volume per change in microvascular pressure. The total rapid change in weight over the first 1 to 2 min after onset of the venous pressure challenge is taken as pure vascular filling and used for the calculation of compliance that was given in g · cm H₂O¹ (28).

Retention (W) was determined as remaining difference of weight before and after a hydrostatic challenge corresponding to the remaining interstitial fluid after normalization of venous pressure and equilibration of a new fluid balance. It is given in grams.

Preparation of PMN

Human PMN were isolated from freshly taken heparinized (10 IU · ml¹) venous blood of normal volunteers. The preparation included dextran enhanced sedimentation, Ficoll-density centrifugation, osmotic lysis of remaining erythrocytes with distilled water. PMN were washed twice and suspended in Hank's balanced salt solution. The preparation of PMN were about 98% pure and viable. They were used within 60 min after preparation.

Experimental Protocol

Two models of oxidative induced lung injury were used: continuous infusion of HOCl and stimulation of neutrophils after having been delayed in the pulmonary microvasculature.

Continuous application of HOCl. The continuous application of HOCl into the arterial line of the system was started at t = 0 min after a 30 min steady-state period and after a baseline hydrostatic challenge at t = 15 min. HOCl that was diluted with perfusate in different concentrations was infused with a constant flow rate of 0.5 ml · min¹, so that HOCl doses of 250, 500, 1,000, and 2,000 nmol · min¹ were administered. Hydrostatic challenges were performed at 30, 60, and 90 min. The results were compared with a control group with continuous saline application.

Stimulated neutrophils. Neutrophils were injected into the arterial line of the preparation at t = 45 min between two steady-state periods of 30 min and two baseline hydrostatic challenges at t = 60 and t = 15 min. The number of neutrophils added to the perfusate was related to the whole circulating volume and given as cells · µl¹ (360 to 3,025 µl¹). Cell counts withdrawn from the venous line indicated that the majority of the PMN (about 95%) were sequestrated in the lung after the first passage. The neutrophils were activated by addition of the chemotactic tripeptide FMLP in a final concentration of 1 µM into the perfusate. Samples for MPO activity measurement were taken 1, 2, 3, 5, 10, 15, 30, 60, and 90 min after PMN stimulation. Hydrostatic challenges were performed at 30, 60, and 90 min as well. The results were compared with control groups either with neutrophil addition without subsequent FMLP injection (group CI) or with FMLP injection without prior neutrophil addition (group CII). Potassium concentration and lactatedehydrogenase (LDH) activity were measured immediately before hydrostatic challenges in both protocols to show that no cell damage occurred.

Substances

All reagents were obtained from Sigma (München, Germany). The concentration of the HOCl stock solution was determined spectrophotometrically ( _290 nm = 350 mol¹ · cm¹) immediately before use (29).

Analytical Methods

Concentrations of potassium and LDH activity were performed by clinical routine methods at the Institute of Clinical Chemistry of the University of Göttingen.

The MPO activity was assayed according to the method described by Klebanoff and coworkers (30). Briefly a sample of 100 µl was added to 2.9 ml assay solution. This solution was made up daily as follows: 26.9 ml H₂O, 3.0 ml 0.1 M sodium phosphate buffer, 0.1 ml 0.1 M H₂O₂ and 0.048 ml guaiacol. The production of tetra-guaiacol was spectrophotometrically measured at 470 nm ( = 26.6 mM¹ · cm¹) for 5 min.

Statistical Methods

Data are given as mean ± SEM. Analysis for statistical significance was performed by the two-tailed Student's t test for unpaired samples.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Continuous HOCl Infusion

The time courses of PAP evoked by continuous HOCl administration and the control group PAP are shown in Figure 1A. Whereas the time course of PAP produced by the application of the lowest HOCl dose (250 nmol · min¹) does not significantly differ from that of the control group, continuous infusion of 500 to 2,000 nmol · min¹ HOCl causes a dose dependent increase in PAP. As shown in Figure 2A, there is no significant difference in the baseline PAP between all groups. The total extent of the pressure rise (given in Figure 2A) as well as the time of maximum PAP (given in Figure 2B) and the start of the pressure rise (see Figure 1A) is dependent on the HOCl dose. The difference between the maximum PAP and the baseline PAP (PAP_max) averages 2.4 ± 0.21 mm Hg, 4.9 ± 0.29 mm Hg, and 4.6 ± 0.25 mm Hg in the 500, 1,000 and 2,000 nmol · min¹ HOCl group. Edema formation that occurred during all experiments except the control group (see Table 1) resulted in a premature termination of the experiments in the 1,000 and 2,000 nmol · min¹ group. The mean recording times are given in Figure 1B. This premature termination due to edema formation may explain that no further pressure development is observed in the 1,000 and 2,000 nmol · min¹ group and that there is no significant difference in the PAP_max between these groups.

View larger version (41K): [in this window] [in a new window]   Figure 1.   Time course of PAP (A) and mean recording times (B) of experiments with HOCl administration. The mean PAP ± SEM of the control group and of the 250, 500, 1,000, and 2,000 nmol · min1 group at certain times and the mean recording times of these groups are shown. *p < 0.01 level of significance versus control group; **p < 0.01 level of significance versus control group and all other groups.

View larger version (52K): [in this window] [in a new window]   Figure 2.   Baseline and maximum PAP (A) and time of maximum PAP (B) of experiments with HOCl administration. Mean ± SEM of the control group and of the 250, 500, 1,000 and 2,000 nmol · min1 group are given. *p < 0.01 level of significance versus baseline PAP and maximum PAP of the control group **p < 0.01 level of significance versus baseline PAP and maximum PAP of every other group.

The results calculated from the data of the hydrostatic challenges (K_f,c, W, and C) are presented in Table 1. The influence of HOCl administration on vascular permeability indicated by these data is time and dose dependent as well. Only slight changes in the vascular permeability are observed in the 250 and 500 nmol · min¹ HOCl group with an up to twofold increase in K_f,c and W. The changes in K_f,c and W due to higher HOCl doses are more severe and occur earlier. Continuous administration of 1,000 and 2,000 nmol · min¹ results in an up to tenfold increase in K_f,c. This increase in K_f,c is accompanied by an increase in fluid retention during hydrostatic challenges and by severe edema formation that is reflected by the total fluid retention, indicating a remarkable vascular leakage. The experiments were terminated when the total fluid retention (W_tot) exceeded 50 g.

As PAP did not exceed 18 mm Hg the formation of hydrostatic edema would not be expected (27). The vascular compliance is found to remain unchanged during HOCl administration indicating that the increase in lung weight is rather due to changes in the microvascular permeability than to increased vascular filling. Perfusate concentrations of potassium and LDH activity (data not shown) show no significant changes as well, implying that the observed effects are not caused by nonspecific cell damage.

PMN Stimulation

Human neutrophils (PMN group: 1,480 ± 323 µl¹, n = 8; CI group; 1,295 ± 426 µl¹, n = 6) were injected into the arterial line of the preparation at t = 45 min between two steady state periods. Cell counts withdrawn from the venous line (data not shown in detail) demonstrated that more than 95% of the cells disappeared from the perfusate within the first passage. The neutrophils were stimulated by injection of FMLP (PMN and CII group: final concentration 1 µM) at t = 0 min. The results of these experiments are given in Figure 3 and in Tables 2 and 3. Neither the data measured during baseline hydrostatic challenges (t = 60/t = 15 min) nor the PAP were significantly altered by addition of neutrophils into the perfusate. Only the perfusate MPO-activity increased slightly after addition of the neutrophils (first arrow) from 2.0 ± 1.5 to 14.9 ± 1.2 mU · L¹ (PMN group) and from 1.7 ± 1.1 to 13.9 ± 1.1 mU · L¹ (CI group) that may be explained by MPO-release due to cell damage during preparation of neutrophils. The time course of PAP as well as the perfusate MPO activity observed during the experiments with stimulated neutrophils (PMN group) shows a rapid increase after neutrophil stimulation (second arrow). Whereas the pressure rise is almost totally reversible, the perfusate MPO-activity remains increased during the whole registration time. The time courses of MPO-activity and PAP of the PMN group differ significantly from those of control (CI and CII) experiments. CII group experiments show a slight increase in PAP and MPO-activity that may be explained by stimulation of resident rabbit PMN. A slight continuous increase in the MPO-activity without any pressure rise is found in CI group experiments indicating a continuous MPO-release from unstimulated PMN due to mechanical stress.

View larger version (28K): [in this window] [in a new window]   Figure 3.   Time course of MPO-activity (A) and PAP (B) of experiments with PMN-stimulation. MPO activity and PAP of the group with stimulated neutrophils (PMN) and control groups (CI: PMN addition without stimulation; CII: no PMN-addition, but FMLP-stimulation) were measured at certain times and are given as mean ± SEM. The first arrow (t = 45 min) shows the addition of neutrophils (PMN, CI) the second arrow indicates the injection of FMLP (PMN, CII). *p < 0.01 level of significance versus CI and CII.

The maximum MPO activity (mean 56.1 ± 7.3 mU · L¹, see Table 3) is measured in the 1-min, 2-min, or 3-min sample. As shown in Figure 4A, there is a strong correlation between number of added cells and maximum MPO activity after neutrophil stimulation. The point of intersection between the axis of ordinates and the regression line (at 23 mU · L¹) indicates that a basal MPO release should be expected in a system without addition of neutrophils (corresponding to CII). Indeed a maximum MPO activity of 14.9 ± 2.3 mU · L¹ that may be released from resident rabbit neutrophils is found in control experiments without neutrophil addition (see Figure 3A control CII).

View larger version (20K): [in this window] [in a new window]   Figure 4.   Correlation of cell count/MPO-release (A) and MPO- release/PAPmax (B). Maximum MPO activity is plotted versus cell count and PAPmax is plotted versus maximum MPO activity for each experiment of the PMN group. Regression lines, correlation coefficients (R) and levels of significance (P) are given.

The PAP_max (mean 8.4 ± 1.1 mm Hg, see Table 3) was recorded 3.7 ± 0.19 min after neutrophil stimulation. The PAP_max of every single experiment correlates closely with the corresponding maximum MPO activity of this experiment (see Figure 4B). The regression line intersects the axis of ordinates near zero. This extrapolation suggests that there is no pressure rise without MPO release. Neither the time of PAP_max nor the time of maximum MPO activity depends on the cell count.

The effects of neutrophil stimulation on K_f,c, W, and C are presented in Table 2. No change in vascular compliance was found. Only the changes in the t = 60 min values of K_f,c and W in the PMN group reached the level of p < 0.05 significance. K_f,c and W measured during the t = 60 min hydrostatic challenges were increased to 81% and 83% of the t = 15 min baseline value. Obviously there is a certain increase in vascular permeability that seems to be reversible after 90 min.

Concentration of potassium and LDH activity in the perfusate samples of all experiments did not significantly alter (data not shown in detail). A relevant cell damage therefore might be excluded.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The aim of the study was to characterize the effect of the neutrophil specific oxidant HOCl on pulmonary microvasculature and to compare its effects with the action of stimulated neutrophils.

Continuous HOCl Infusion

HOCl that is formed from hydrogen peroxide by the MPO acts predominantly via oxidative modification of free functional groups of proteins, especially sulfhydryl groups (19, 20). HOCl is much more reactive than hydrogen peroxide and its targets are distinct from those of hydrogen peroxide (18, 20). HOCl in low concentrations (< 50 µM) has been shown to alter cell membrane function and cellular energy metabolism in a time and concentration dependent manner (20). Our results are consistent with these findings concerning the applicated HOCl concentrations and the dose and time dependence: HOCl doses that were administered in this study are ranging between 250 and 2,000 nmol · min¹ (perfusate flow 100 ml · min¹) and correspond to HOCl perfusate concentrations between 2.5 and 20 µM. Furthermore the observed effects of HOCl application are time and dose dependent as well, i.e., the effects evoked by larger HOCl doses are stronger and occur earlier. A time and dose dependent depletion of the antioxidative cellular defence may be an explanation of these kinetics. This thesis is supported by the observation that the product of mean recording time (Table 1) and HOCl dose of the experiments in which edema formation caused premature termination (1,000 nmol · min¹ group: 83.7 µmol and 2,000 nmol · min¹ group: 78.8 µmol) shows no significant difference. Thus the application of about 80 µmol HOCl may represent the HOCl dose that causes severe edema formation leading to a premature termination of the experiment in this isolated protein-free-buffer-perfused organ model that was used in our study. This dose was not given in the 250 nmol · min¹ and 500 nmol · min¹ group within the recording time of 105 min.

Stimulation of Neutrophils

Neutrophils that were injected into the arterial line rapidly disappeared from the perfusate. Over 95% of the neutrophils were delayed during the first lung passage. This observation is consistent with a variety of reports (22, 23) indicating that even unstimulated PMN adhere to endothelium (1). PMN sequestrated in the lungs may be found as marginated pool of neutrophils in the capillary bed (31) and along the walls of small arteries (2).

In the presented experiments PMN stimulation evoked a rapid MPO release and pressure rise. There was a strong correlation between maximum MPO-activity and cell count on the one hand and between PAP_max and maximum MPO-activity on the other hand. There are two observations that need further discussion: (1) The perfusate MPO activity increased slightly to about 15 mU · L¹ after addition of the neutrophils into the perfusate at 45 min. However this increase did not influence the PAP and K_f,c that remained constant during both steady-state periods. (2) The time course of PAP shows a short maximum and is almost totally reversible whereas the perfusate MPO activity remains increased during the whole registration time. These observations may be explained by the following considerations: The oxygen burst of FMLP-stimulated neutrophils is characterized by a maximum of oxidant generation after about one minute and an overall duration of about 20 to 25 min (32). During the oxygen burst the superoxide radical is formed by the NADPH oxidase and dismutates spontaneously or enzymatically to hydrogen peroxide that is the precursor for the MPO mediated HOCl formation. The mean PAP (given in Figure 3B) declined from 7.1 ± 0.77 mm Hg at t = 2.5 min to about 2 mm Hg after 30 min that is the maximum duration of the neutrophil's oxygen burst after which no hydrogen peroxide generation and subsequently no HOCl formation is expected. This observation suggests that MPO activity without hydrogen peroxide generation does not affect PAP.

Stimulated neutrophils release arachidonic acid metabolites that may affect the pulmonary microvasculature as well. The arachidonic acid metabolism of stimulated neutrophils involves the action of a 5-lipoxygenase on arachidonate to yield a hydroxy-peroxy-intermediate, which is dehydrated to the unstable epoxide leukotriene A₄ (LTA₄) or converted to hydroxy-eicosatetraenoic acid. LTA₄ is either hydrolized to LTB₄ and further metabolized by -oxidation (considered as a major pathway in PMN) or conjugated to form LTC₄ and further cleaved to form LTD₄ and LTE₄ (33). However on the one hand LTB₄ per se that is released by stimulated PMN is reported to have no influence on PAP or microvascular permeability (22, 23) on the other hand LTC₄ and D₄ that have been shown to evoke changes in PAP and K_f,c (34) are not synthesized by PMN alone and need cooperation with other compartments, i.e., endothelial cells (33). This pathway represents a slow reacting process that does not seem to be an appropriate explanation of the observed rapid pressure rise. However it may explain the increase in K_f,c that is observed after 60 min.

Between 360 and 3,025 neutrophils per µl perfusate were administered in this study, i.e., between about 1.1 and 9.1 · 10⁸ neutrophils per single experiment were infused. The isolation of these numbers of neutrophils from whole blood of rabbits would require whole blood from up to six donor animals, that would not be feasible. The preparation of rabbit peritoneal neutrophils that are isolated after the intraperitoneal injection of a stimulus would provide much more neutrophils per donor animal. However peritoneal neutrophils are isolated after migration into the peritoneal cavum. Although this technique is often used, a functional alteration of these cells can not surely be excluded. Therefore resting human neutrophils were prepared from a sufficient volume of whole blood of one donor per single experiment.

The infusion of human neutrophils into an isolated rabbit lung is an established model as well (15, 33, 35). Human neutrophils are slightly larger than rabbit neutrophils (spherical diameter of resting cells 6.4 ± 0.6 µm versus 6.8 ± 0.8 µm, p < 0.05) and human pulmonary capillaries are slightly but not significantly wider than rabbit pulmonary capillaries (31). A percentage of capillary segments (67% in rabbits, 38% in humans) is too narrow to allow neutrophils to pass through them without deforming the neutrophils. Human neutrophils show a higher deformability than rabbit neutrophils. The ratio between longest and perpendicular diameter increased from 1.1 ± 0.1 (both species) to 6.2 ± 3.0 (human) and 1.9 ± 1.0 (rabbit, p < 0.01) during the passage through the pulmonary capillary bed (31). Therefore a trapping of human neutrophils for mechanical reason is not expected. Moreover, neither an increase in vascular permeability nor a pressure rise was observed after infusion of neutrophils, also indicating that a trapping of these cells for mechanical reasons did not occur.

Importance Under In Vivo Conditions

The study investigates the effect of the neutrophil derived oxidant HOCl and of neutrophil stimulation using an ex vivo isolated buffer perfused organ model. In particular the protein-free buffer medium differs from physiological conditions. The absence of plasma proteins may partially modify the effects of neutrophil derived oxidative stress for the following reasons:

(1) Plasma proteinsin particular albuminmay act as potent oxidant scavengers because of their free functional groups (i.e., thiol, thioether and amino groups) (18, 19). Therefore bovine serum albumin that is sometimes used in isolated perfused lung preparations for oncotic reasons (for example 22, 23), may decrease the effects of stimulated neutrophils via oxidant scavenging. However, since the highly reactive oxidant HOCl acts very locally, it is not expected that all produced HOCl is inactivated by the reaction with circulating albumin under physiological circumstances. A certain amount of HOCl will always attack specific targets in close proximity of stimulated neutrophils. Therefore the presence of albumin is expected to modify quantitatively but not qualitatively HOCl- induced effects. The infusion of low concentrations of HOCl (as used in this study) into an albumin containing perfusate would cause an immediate inactivation of the oxidant. Corresponding to others (15, 18, 20, 33) dealing with neutrophil-derived oxidative stress and/or lung injury, we used an albumin free buffer fluid for those reasons.

(2) Other plasma constituents (for example the 1-proteinase-inhibitor) are involved in the regulation of the activity of proteolytic enzymes that are released from neutrophil granules (elastase, serine proteinase, metalloproteinases) (16). The oxidative inactivation of the 1-proteinase-inhibitor due to HOCl-induced oxidative modification of the critical methionyl residue (Met-358) and the oxidative activation of metalloproteinases causes a severe perturbation of the proteinase-antiproteinase balance (for review see [16]). This indirect action of neutrophil-derived oxidative stress is considered a more important in vivo mechanism in the destruction of most tissues than the direct action of HOCl. However special tissues, like the lungs and the vascular bed, have an unexpected greater sensitivity to direct action of oxidants (16). Corresponding to these data, the present study did not find tissue or cell destruction due to administration of low HOCl-doses that is indicated by unchanged perfusate potassium and LDH levels. The increase in PAP and K_f,c occurs after administration of very low HOCl-concentrations (2.5 to 20 µM), whereas tissue destruction and lysis, as described by Weiss (16), were caused by a HOCl concentration of 70 mM. Therefore the results of our study may be explained by HOCl-induced alterations of specific cellular functions, like energy metabolism (20), specific membrane-properties or signal transduction systems (24, 25), that are usually caused by low oxidant concentrations (20). The stimulation of neutrophils as well did not cause cell or tissue destruction (as proved by LDH and potassium measurement) even in the absence of plasmatic antiproteases. On the one hand this lack of proteolytic damage in this isolated lung model may be explained by a too small number of neutrophils that only caused effects comparable to those of low HOCl concentrations. On the other hand lung tissue and endothelial structures may show a special susceptibility to oxidative stress as pointed out by Weiss as well (16). This special susceptibility would explain that the isolated lung model sensitively shows effects of oxidative stress without any tissue damage due to proteolytic activity.

Since this study shows no proteolytic effects in the absence of plasmatic antiproteinases, substitution of antiproteinase would not cause specific effects.

Comparison of Both Models of Neutrophil Derived Oxidative Stress

HOCl infusion as well as stimulation of neutrophils that have been delayed in the pulmonary microvasculature influenced PAP and vascular permeability in an isolated rabbit lung model. However there are some differences between these models: (1) Whereas HOCl infusion caused drastic changes in the vascular permeability and a rather moderate increase in PAP, stimulation of neutrophils evoked a noticeable pressure rise but rather slight changes in K_f,c and W that reached p < 0.05 level of significance only at the t = 60 min value. (2) The pressure rise observed during HOCl infusion started late and increased slowly and continuously whereas the pressure rise induced by stimulated PMN occurred immediately after stimulation, increased rapidly and was reversible. (3) Changes in PAP and vascular permeability due to HOCl administration simultaneously occurred whereas the changes in the vascular permeability after PMN stimulation were observed after PAP_max.

These differences may be explained by the different compartments in which the oxidants act on different targets. Oxidantsin particular HOCl as a highly reactive compound act very locally. Thus, continuous HOCl infusion is expected to affect primarily the endothelial function and to cause an increased vascular permeability first. After the breakdown of the endothelial barrier function representing a condition for the action of HOCl on subendothelial structures, i.e., vascular smooth muscle cells, a pressure rise may occur. By contrast, PMN having been delayed in the pulmonary microvasculature are expected to be found in the capillary bed (31) and in the walls of small arteries (2). Therefore oxidants released during an oxygen burst may markedly influence vascular smooth muscle cells that regulate the vascular tone.

For those reasons the experimental models used in this study are only partially comparable concerning their effect on pulmonary microvasculature.

The effects of continuous HOCl administration as well as of PMN stimulation were not accompanied by a significant increase in perfusate LDH activity and in perfusate potassium concentration. Therefore a significant cell damage can be excluded. This supports the hypothesis that HOCl as nonradical oxidant causes no nonspecific increase in cellular membrane permeability and subsequent cell death that is expected from radical oxidants due to lipid peroxidation. Furthermore our results suggest that low HOCl concentrations cause disturbances of special cellular functions that result in an increase in vascular tone and in a loss of endothelial barrier function. These effects are time and dose dependent and occur after the administration of a certain HOCl dose. This observation suggests that these processes occur after a breakdown of cellular antioxidative defence systems.

HOCl has been proved to be a useful model of oxidative stress that, apart from limitations mentioned above, reproduces the oxidative activity of PMN in the isolated rabbit lung model used in this study. Both models also confirm our hypothesis that PMNs are involved in pulmonary hypertension and edema formation during ARDS via their major oxidant HOCl.