INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The acute phase of acute lung injury (ALI) and the acute respiratory distress syndrome (ARDS) are characterized by the influx of protein-rich edema fluid into the alveolar spaces as a consequence of increased permeability of the alveolar-capillary barrier. Alveolar epithelial damage is a characteristic morphologic feature in patients with ALI/ARDS (1). The loss of epithelial integrity contributes to the alveolar flooding and disrupts normal alveolar epithelial fluid transport which is important for alveolar fluid clearance.

Efficient repair of the alveolar epithelium is therefore crucial for the restoration of the alveolar epithelial barrier and recovery from ALI/ARDS (2, 3). Moreover, disorganized or insufficient epithelial repair may lead to fibrosis (4). The alveolar epithelial type II cell is the progenitor for reepithelialization of the denuded basement membrane (1, 5). Alveolar type II epithelial cells restore the integrity of the alveolar epithelium by proliferation, spreading and migration, and finally by differentiation to alveolar type I cells, restoring the normal morphology and functional properties of the alveolar epithelium. This process is controlled in part by a variety of soluble mediators, including growth factors and cytokines.

Neutrophil granulocytes are the predominant inflammatory cells in the alveolar space of patients with ALI/ARDS (6). Clinical and experimental studies indicate that lung injury is at least in part neutrophil-dependent (7). Pulmonary edema fluid or bronchoalveolar lavage from patients with ALI/ ARDS contain the products secreted by activated neutrophils such as reactive oxygen species (ROS), proteinases, and soluble inflammatory mediators (6). These neutrophil products can be detected in pulmonary edema fluid from patients with ALI/ARDS and are in part responsible for the alveolar epithelial damage present in patients with ALI/ARDS.

The effect of pulmonary edema fluid on alveolar epithelial repair has not been studied. Therefore, our first objective in this study was to determine the effect of pulmonary edema fluid from patients with ALI/ARDS on in vitro alveolar epithelial repair. Because of the extensive damage to the alveolar epithelium in patients with ALI/ARDS and the presence of soluble inflammatory mediators in the pulmonary edema fluid, we hypothesized that pulmonary edema fluid would inhibit alveolar epithelial repair.

To test this hypothesis, we first studied the effect of pulmonary edema fluid from patients with early ALI/ARDS in an in vitro wound healing model using mechanically wounded monolayers of human A549 alveolar epithelial-like cells (8, 9). Pulmonary edema fluid from patients with hydrostatic pulmonary edema was used as a control. Surprisingly, we found an increase in alveolar epithelial repair induced by pulmonary edema fluid from patients with ALI/ARDS when compared with plasma from the same patients or with pulmonary edema fluid from patients with hydrostatic edema. Because we recently found that IL-1 increases in vitro alveolar epithelial repair in rat primary alveolar epithelial cells (9) and biologically active IL-1 has been found in both pulmonary edema fluid (10) and bronchoalveolar lavage fluid (11) from patients with ALI/ARDS, we tested the hypothesis that IL-1 contributes to the increased alveolar epithelial repair by pulmonary edema fluid from patients with ALI/ARDS.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patient Selection and Collection of Pulmonary Edema Fluid

Patients were defined as having ALI or ARDS according to the criteria from the North American-European Consensus Conference (12) and hydrostatic pulmonary edema was diagnosed as described before (13). The study was approved by the Committee for Human Research at the University of California, San Francisco. A random selection of 31 patients with pulmonary edema and plasma samples admitted to the intensive care unit (ICU) at the University of California, San Francisco between 1989 and 1999 was made. Undiluted pulmonary edema samples and plasma samples were collected and prepared by a well-validated method as described before (13). In addition, plasma from six healthy volunteers was individually prepared, pooled, aliquoted, and included in each experiment as control. Total protein concentrations and the rate of alveolar fluid clearance were determined as described before (13, 16, 17).

In vitro Epithelial Wound-healing Assay

The epithelial repair activity was determined using our in vitro epithelial wound-healing assay as described previously (9), except that human A549 alveolar epithelial-like cells (American Type Culture Collection, Rockville, MD) were used. Briefly, A549 epithelial cells were cultured to confluency in 24-well plates for 48 h in minimal essential medium (MEM) containing 10% fetal bovine serum and then mechanically wounded with a pipette tip. Pulmonary edema fluid or plasma from each patient was added to the wounded epithelial monolayers (experiments performed in triplicates) and the area of the wound was measured after 18 h. Preliminary experiments showed that the optimal range of wound-healing activity (50% of maximal effect) was obtained at a 100-fold dilution of the pulmonary edema fluid or plasma in serum-free MEM. The rate of alveolar epithelial repair was expressed as µm²/h and compared with the rate of epithelial repair obtained with the same dilution (100-fold) of pooled plasma from healthy donors (control, shown as 100%). Cell spreading/migration and cell proliferation were determined as described previously (9).

Inhibition experiments were done in presence of IL-1 receptor antagonist (IL-1ra) (R&D Systems, Minneapolis, MN) or recombinant soluble human tumor necrosis factor receptor type 1 (rshTNFR1) (TNF-binding protein 1, Onercept; Serono, Geneva, Switzerland). The concentrations of the inhibitors were chosen based on inhibition curves of a bioassay measuring ICAM-1 upregulation on A549 epithelial cells (10, 11). The effects of human recombinant IL-1 and TNF- (R&D Systems) were studied in serum-free MEM.

Statistical Analysis

Data are presented as mean ± SEM. Statistical analysis was done using Student's unpaired t test, paired t test, or ANOVA, when appropiate. If the ANOVA revealed a significant difference between groups, it was followed by a post hoc Student-Newman-Keuls analysis to determine which groups were statistically different. Chi-square analysis was used for comparisons of the data presented in Table 1. When the expected frequencies were low, Fisher's exact test was used. The results were considered statistically significant if p < 0.05. 

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patient Characteristics

There were 21 patients with ALI/ARDS and 10 patients with hydrostatic pulmonary edema included in this study. The characteristics of the two patient groups are outlined in Table 1. The mean severity of illness (SAPS II) (18) and the lung injury score (LIS) (19) at the time of ICU admission were higher in patients with ALI/ARDS, but the differences did not reach statistical significance. The patients with ALI/ARDS had a longer duration of mechanical ventilation and a higher frequency of impaired alveolar fluid clearance. There was also a tendency for lower survival, more systemic organ dysfunction (liver and renal failure), and a higher frequency of sepsis in the patients with ALI/ARDS (Table 1). The ratio of protein concentrations in edema fluid to plasma, a measure of alveolar-capillary barrier permeability, was 0.92 ± 0.13 in patients with ALI/ARDS and 0.47 ± 0.13 in patients with hydrostatic edema (p < 0.001).

Effect of Pulmonary Edema Fluid from Patients with ALI/ARDS on in vitro Alveolar Epithelial Repair

The epithelial repair activity of pulmonary edema fluid and plasma from patients with ALI/ARDS or hydrostatic pulmonary edema was compared with the epithelial repair activity of pooled plasma from healthy donors. Pulmonary edema fluid from patients with ALI/ARDS enhanced alveolar epithelial repair by 33 ± 4% compared with pooled plasma from healthy donors (p < 0.01) (Figure 1). By contrast, the epithelial repair activity of pulmonary edema fluid and plasma from patients with hydrostatic edema did not differ from the epithelial repair activity of pooled plasma. Plasma from patients with ALI/ ARDS was also not different from pooled plasma from healthy donors (Figure 1).

View larger version (12K): [in this window] [in a new window]   Figure 1.   Effect of pulmonary edema fluid and plasma from patients with ALI/ARDS or hydrostatic pulmonary edema on alveolar epithelial repair. The epithelial repair activity was determined using an in vitro wound-healing assay with A549 alveolar epithelial-like cells. The epithelial repair activity of pulmonary edema fluid or plasma was measured for each sample and expressed as percent increase of alveolar epithelial repair compared with pooled plasma from healthy donors (control). All results are reported as mean ± SEM, *p < 0.01 (compared with control).

Epithelial Repair Activity in Pulmonary Edema Fluid from Patients with ALI/ARDS Over Time

In 11 patients it was possible to obtain sequential samples of pulmonary edema fluid. The activity of pulmonary edema fluid samples from early stages of the disease (< 12 h after intubation) was compared with pulmonary edema fluid samples from the same patients at later stages of the disease (12 to 48 h after intubation). The epithelial repair activity declined over time, but it is still significantly elevated in the pulmonary edema fluid compared with plasma obtained > 12 h after intubation (Figure 2). In contrast, the epithelial repair activity of the sequential plasma samples obtained over time (n = 11) was similar (Figure 2).

View larger version (12K): [in this window] [in a new window]   Figure 2.   Comparison of the epithelial repair activity from sequential pulmonary edema fluid and plasma samples obtained from patients with ALI/ARDS (n = 11). The epithelial repair actvity was determined using an in vitro wound-healing assay with A549 alveolar epithelial-like cells. The epithelial repair activity from pulmonary edema fluid or plasma obtained within the first 12 h after intubation was compared with sequential samples obtained 12 to 48 h after intubation. Data are expressed as percent increase of alveolar epithelial repair compared with pooled plasma from healthy donors (control). All results are reported as mean ± SEM, **p < 0.01, *p < 0.05 (compared with control).

Contribution of Inflammatory Cytokines IL-1 and TNF- to Pulmonary Edema Fluid-induced in vitro Alveolar Epithelial Repair

In order to study the possible contribution of IL-1 and TNF- to the epithelial repair activity of pulmonary edema fluid from patients with ALI/ARDS, we examined the effect of IL-1ra and rshTNFR1 on alveolar epithelial repair induced by pulmonary edema fluid from patients with ALI/ARDS. We first determined concentrations of inhibitors that induced a maximal inhibition of their corresponding cytokines. For that purpose A549 epithelial cells were stimulated to upregulate surface ICAM-1 using rhTNF- (1 ng/ml) and rhIL-1 (20 pg/ml) in the presence of various concentrations of their respective inhibitors, IL-1ra and rshTNFR1 (Figure 3, A and B). The results showed that a 1,000-fold excess of IL-1ra over IL-1 and a 100- to 1,000-fold excess of rshTNFR1 over TNF blocked the induction of ICAM-1 on A549 cells. These ratios of excess of the cytokine inhibitor to the stimulating cytokine were used for further experiments in the wound-healing assay. In all of the pulmonary edema fluid samples from patients with ALI/ ARDS tested (n = 10), the epithelial repair activity was decreased in the presence of 200 ng/ml IL-1ra (Figure 4A). On average, IL-1ra inhibited pulmonary edema fluid-induced alveolar epithelial repair by 46 ± 4% (p < 0.001). In contrast, adding rshTNFR1 (100 ng/ml) to the pulmonary edema fluid from patients with ALI/ARDS did not have a significant effect on the epithelial repair activity (Figure 4B). Pulmonary edema fluid from patients with hydrostatic edema showed no decrease in the epithelial repair activity in the presence of IL-1ra (data not shown). In addition, there was no difference in epithelial repair activity in the presence or absence of IL-1ra in late edema fluid samples (12 to 48 h after intubation), indicating that the repair activity caused by IL-1 is primarily detectable in the most early phase of the disease.

View larger version (11K): [in this window] [in a new window]   Figure 3.   Effect of specific cytokine inhibitors on alveolar epithelial cell activation (ICAM-1 upregulation). (A) Inhibitory effect of IL-1 receptor antagonist (IL-1ra) on IL-1-induced surface ICAM-1 expression in A549 cells. IL-1 concentration was kept constant at 20 pg/ml. (B) Inhibitory effect of recombinant soluble human TNF receptor 1 (rshTNFR1) on TNF--induced surface ICAM-1 expression in A549 cells. TNF- concentration was kept constant at 1 ng/ml.

View larger version (16K): [in this window] [in a new window]   Figure 4.   Effect of inhibitors of IL-1 and TNF- on the epithelial repair activity of pulmonary edema fluid from patients with ALI/ARDS (n = 10). The epithelial repair activity was measured using an in vitro wound-healing assay with A549 alveolar epithelial-like cells. The epithelial repair activity was determined in the presence and absence of IL-1ra (200 ng/ml) (A) or rshTNFR1 (100 ng/ml) (B) and was expressed as µm2/h. Mean epithelial repair activity was decreased by 46 ± 4% in presence of IL-1ra (p < 0.001), whereas there was no significant difference of the alveolar epithelial repair activity in presence of rshTNFR1.

Effect of Recombinant IL-1 and TNF- on in vitro Alveolar Epithelial Repair

In order to confirm that IL-1 enhances in vitro alveolar epithelial repair, the effect of recombinant IL-1 and TNF- was tested in our in vitro wound healing model. There was a concentration-dependent increase in alveolar epithelial wound closure induced by IL-1, but not by TNF- (Figure 5). Interestingly, an effect of IL-1 was detectable at concentrations as low as 1 pg/ml (Figure 5).

View larger version (14K): [in this window] [in a new window]   Figure 5.   Effect of recombinant IL-1 and TNF- on alveolar epithelial repair in vitro. Alveolar epithelial repair was determined using an in vitro wound-healing assay with A549 alveolar epithelial-like cells. IL-1 or TNF- were added to serum-free medium at the concentrations indicated, and the rate of wound closure (µm2/h) was determined after 18 h.

Cell Spreading and Migration, but Not Cell Proliferation Are Mechanisms of in vitro Alveolar Epithelial Repair in A549 Epithelial Cells

Cell spreading, cell migration, and cell proliferation are possible mechanisms of epithelial wound closure (8). Cell spreading/migration was increased at the edge of the wound after incubation with pulmonary edema fluid when compared with plasma from patients with ALI/ARDS (mean internuclear distance, 52.8 ± 11.0 versus 40.2 ± 6.4 µm, p < 0.05). However, cell proliferation was not increased at the wound edge after incubation with recombinant IL-1 or pulmonary edema fluid from patients with ALI/ARDS, as measured by incorporation of BrdU (data not shown). Moreover, there was no statistically significant difference in cell numbers after incubating A549 epithelial-like cells with pulmonary edema fluid from patients with ALI/ARDS compared with patients with hydrostatic edema (n = 10), confirming the hypothesis that the difference in alveolar epithelial wound repair is not due to increased cell proliferation.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The primary objective of this study was to determine the effect of pulmonary edema fluid from patients with ALI/ARDS on alveolar epithelial repair in vitro. On the basis of the pathologic findings in lungs from patients with ALI/ARDS, we hypothesized that pulmonary edema fluid would inhibit alveolar epithelial repair. However, we found that pulmonary edema fluid from patients with ALI/ARDS enhanced in vitro alveolar epithelial repair. Furthermore, the factors that enhanced alveolar epithelial repair must have been produced locally in the distal air spaces of the lung since plasma showed less epithelial repair activity than did pulmonary edema fluid from the same patients. The epithelial repair activity of pulmonary edema fluid was higher early in the course of ALI/ARDS (< 12 h after intubation) and declined over time (12 to 48 h after intubation). Part of the epithelial repair activity was mediated by the early response cytokine IL-1 since pulmonary edema fluid-induced alveolar epithelial repair was significantly inhibited in the presence of IL-1ra.

The early phase of ALI/ARDS is characterized by inflammation in the alveolar space and damage to the alveolar barrier with an increase in endothelial and epithelial permeability (1). Protein-rich edema fluid accumulates in the alveolar space, leading to arterial hypoxemia, decreased lung compliance and respiratory failure. The early phase is followed by a subacute, proliferative phase that is characterized by repair processes resulting in either restoration of the normal architecture of the lung or progressive obliteration of the interstitial and alveolar compartments and lung fibrosis. Early epithelial repair with reepithelialization of the denuded basement membrane is thought to be important for the restoration of the normal architecture of the lung (4, 20).

A variety of growth factors and cytokines might be involved in the alveolar epithelial repair process. Several studies have reported elevated concentrations of growth factors or cytokines in bronchoalveolar lavage or pulmonary edema fluid from patients with ALI/ARDS (6). However, cytokine concentrations have usually been determined using enzyme-linked immunosorbent assays or radioimmunoassays. These assays determine the total or immunoreactive content of the cytokine, without providing any information regarding its biologic activity. Therefore, we modified our in vitro wound-healing system (8, 9, 21) to establish a bioassay with human alveolar epithelial-like A549 cells, which made it possible to determine the net biologic activity of pulmonary edema fluid and plasma from patients on in vitro alveolar epithelial repair.

Because patients with ALI/ARDS have substantial damage to the alveolar barrier with leakage of plasma into the alveolar space, it is possible that the plasma contains growth factors that are responsible for the increase of alveolar epithelial repair by pulmonary edema fluid from patients with ALI/ARDS. Therefore, we compared the epithelial repair activity of the pulmonary edema fluid with that of the plasma obtained simultaneously from the same patients and found significantly higher epithelial repair activity only in the pulmonary edema fluid. These results indicate that biologically active mediators that can enhance alveolar epithelial repair are released into the alveolar space in patients with ALI/ARDS. The epithelial repair process may begin in the early phase of ALI/ARDS since pulmonary edema fluid obtained within the first 12 h after intubation enhanced epithelial repair to a greater extent than pulmonary edema fluid obtained more than 12 h after intubation.

As a control, pulmonary edema fluid and plasma from patients with hydrostatic edema was used. Patients with hydrostatic pulmonary edema represent excellent controls because these patients had sufficiently severe pulmonary edema to require positive-pressure ventilation for respiratory failure. Thus, cytokine production caused by ventilator-induced lung injury is unlikely to be responsible for the enhanced epithelial repair activity observed in pulmonary edema fluid from patients with ALI/ARDS.

Because IL-1 increases alveolar epithelial repair in primary rat alveolar epithelial cells (9), and IL-1 was shown to be biologically active in pulmonary edema fluid from patients with ALI/ARDS (10), we tested the hypothesis that IL-1 contributed to the alveolar epithelial repair induced by pulmonary edema fluid. We found that the epithelial repair activity was inhibited by 46% in the presence of IL-1ra, indicating that IL-1 mediated a major fraction of the enhanced in vitro alveolar epithelial repair. This conclusion was supported by additional experiments with recombinant IL-1 that showed a concentration-dependent increase in alveolar epithelial repair in our in vitro wound-healing model. In contrast, recombinant TNF- did not increase alveolar epithelial repair. As expected, inhibition of TNF- in the pulmonary edema fluid by rshTNFR1 did not affect the epithelial repair activity of the pulmonary edema fluid.

In patients with ALI/ARDS, alveolar macrophages are the main source of IL-1 production (22). The presence of IL-1 in damaged epithelium has been described in nonpulmonary organs, including the skin (23, 24), the cornea (25), and the gastrointestinal tract (26). However, the functional role of IL-1 in the wound-repair process is not clear. Some investigators have speculated that IL-1 is upregulated in epidermal wounds in order to induce an inflammatory response with accumulation of neutrophils to keep the wound sterile (27). Similar mechanisms might be germane to the lung. However, the data in this study demonstrate that IL-1 may have an additional role in ALI/ARDS, specifically to enhance alveolar epithelial repair.

Beside IL-1, other mediators seem to contribute to in vitro alveolar epithelial repair induced by pulmonary edema fluid since the inhibition of IL-1 by IL-1ra was not complete. Previous studies in our laboratory showed elevated concentrations of keratinocyte growth factor, hepatocyte growth factor (15), and transforming growth factor- (28) in pulmonary edema fluid from patients with ALI/ARDS. All these growth factors can increase alveolar epithelial repair in vitro (8, 29) and may contribute to alveolar epithelial repair in vivo.

In a recently published study with primary rat alveolar epithelial cells, we found that the effect of recombinant IL-1 on alveolar epithelial repair is due to cell spreading and cell migration and not to cell proliferation (9). The data presented in this study support the hypothesis that cell spreading and migration are the major mechanisms of edema fluid-induced wound closure in our in vitro model. Cell proliferation was not enhanced in the presence of pulmonary edema fluid compared with plasma and thus did not contribute to the enhanced wound closure in our model.

There are some limitations to this study. First, our in vitro model does not adequately represent the in vivo alveolar environment. For example, there is increasing evidence that reactive oxygen species play an important role in ALI/ARDS. Because reactive oxygen species are extremely unstable, it is possible that the net epithelial repair activity is different in vivo when compared with the ex vivo conditions, although the edema fluid samples were processed immediately after being obtained from the patient. Second, we used the human cell line A549 for our studies and not primary rat alveolar epithelial cells in order not to have any species differences. Although A549 epithelial cells have been used in many studies, they may not respond to all biologic stimuli in the same fashion as primary alveolar epithelial cells. However, we have recently found that the effect of recombinant IL-1 on alveolar epithelial repair is similar in primary rat alveolar epithelial cells and A549 epithelial-like cells (9). Also, the use of A549 epithelial cells allowed us to use human alveolar epithelial cells to study the biologic activity of human pulmonary edema fluid, maintaining consistency in the species that was studied.

In summary, pulmonary edema fluid from patients with ALI/ARDS enhances alveolar epithelial repair in vitro. Part of this activity was produced in the alveolar space since the epithelial repair activity of plasma obtained simultaneously from the same patients was significantly lower and at a level comparable to that of patients with hydrostatic pulmonary edema or that of normal control subjects. The increase in epithelial repair activity was maximal within the first 12 h after intubation and declined over time. Approximately half of the increase in alveolar epithelial repair was mediated by biologically active Il-1. These findings indicate that the epithelial repair process may be induced much earlier in the course of the disease than has previously been appreciated and that the early response cytokine IL-1 may contribute to the repair of the alveolar epithelium. On the basis of these results, coupled with our recent study on primary rat alveolar epithelial cells (9), there may be a novel role for IL-1 in the early repair process after acute lung injury.