INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Acute lung injury (ALI) is characterized by an early exudative phase and a late proliferative phase. In the exudative phase alveoli contain proteinaceous fluid, red blood cells, neutrophils, and macrophages. Edema and neutrophils accumulate in the interstitium, and alveolar ducts contain hyaline membranes. Microatelectasis is present, endothelial cells are swollen, and focal destruction of endothelial cells occurs. The exudative phase evolves into the proliferative phase after 1-2 wk approximately. The proliferative phase is characterized by type II epithelial cell hyperplasia and accumulation of fibroblasts and numerous monocytes in the lung interstitium. Normal acinar architecture is destroyed and replaced by thick layers of fibrotic tissue that obliterate alveolar spaces and ducts (1).

Recently some authors have observed an increased number of myofibroblasts and procollagen type I (4) and III (5) producing cells early in the course of ALI, suggesting that the proliferative phase begins much sooner than had been previously appreciated. The elastic system is much less understood, however. Despite advances in the understanding of the structural complexity of the elastic system, the interaction between the elastin and microfibril components of the elastin fiber system remains a matter of speculation, mainly in the face of lung remodeling and repair after ALI. Finally, data relating tissue mechanics to collagen and elastic fibers contents in ALI are not found in the existing medical literature.

This study was undertaken to test the hypothesis that the intrinsic mechanical properties of lung parenchyma may be already altered early in the evolution of ALI, independently of surface film, alveolar flooding, or heterogeneity effects. The mechanical changes would hypothetically reflect modification of the connective fiber complex of the alveolar septa. The model and methods used provide an accurate way of evaluating the relationship between interstitial remodeling (amounts of collagen, oxytalan, and elaunin plus fully developed elastic fibers) and tissue mechanical properties (elastance, resistance, and hysteresivity) in different degrees of paraquat-induced acute lung injury.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Twenty-five Wistar rats (25-300 g) were randomly divided into five groups of five animals each. In the control group (C), saline (5 ml/kg body weight [BW]) was injected intraperitoneally. In P groups, different doses of paraquat (10, 15, 25, and 30 mg/kg BW) were injected intraperitoneally 24 h prior to the measurements. The animals were sedated, anesthetized (pentobarbital sodium [20 mg/kg]), tracheotomized, paralyzed (gallamine triethyliodide [2 mg/kg intravenously]), and mechanically ventilated. Airflow, volume, tracheal, transpulmonary, and esophageal pressures were registered. Respiratory system (rs), lung (L), and chest wall (w) resistive (P1), viscoelastic/inhomogeneous pressures (P2), Ptot (=P1+ P2), static elastance (Est), and the difference between dynamic and static elastances (E) were computed by the end-inflation occlusion method (8). After degassing, the lungs were removed en bloc, and placed in a modified Krebs-Henseleith (K-H) solution (in mM: 118.4 NaCl, 4.7 KCl, 1.2 K₃PO₄, 25 NaHCO₃, 2.5 CaCl₂ · H₂O, 0.6 MgSO₄ · H₂O, and 11.1 glucose) at pH 7.40 and 6° C (11). The pH was controlled with bubbling 95% O₂ and 5% CO₂. Tissue strips of 3 × 3 × 10 mm were prepared, and then suspended vertically in a K-H organ bath maintained at 37° C and continuously bubbled with 95% O₂-5% CO₂. One clip was attached to a force transducer and the other one was fastened to a vertical rod. This fiber-glass stick was connected to the cone of a woofer, which was driven by the amplified sinusoidal signal of a wave-form generator. A side arm of the rod was linked to a second force transducer by means of a silver spring of known Young's modulus, thus allowing the measurement of displacement. The apparatus is described in detail in Faffe and coworkers (14).

Strips were preconditioned by sinusoidally oscillating the tissue during 30 min (frequency = 1 Hz; amplitude large enough to reach a maximal stress of 20 g/cm²). Thereafter the amplitude was adjusted to 5% of the strip's resting length (L₀), the strips were unloaded to a force of 0.8 g, and the oscillation maintained for another 30 min, or until a stable length-force loop was reached. The final basal force was approximately 0.5 g. After stress adaptation, strips were oscillated at a frequency (f) = 1 Hz.

Tissue resistance (R), elastance (E), and hysteresivity () were calculated from the oscillatory recordings according to Fredberg and Stamenovic (14, 15).

Morphometric Analysis

Parenchymal strips were quick-frozen by immersion in liquid nitrogen and fixed with Carnoy's solution (16, 17). Slices 4 µm thick were cut and underwent hematoxylin-eosin and specific staining methods to quantify the collagenous (picrosirius-polarization method [18]) and elastic system fibers (Weigert's resorcin fuchsin method [RF] [19] and Weigert's resorcin fuchsin method modified with oxidation [ORF] [20]) in the alveolar septa. The results were expressed as the amount of elastic and collagen fibers per unit of septum length.

Statistical Analysis

One-way ANOVA and the Spearman correlation test were used when required ( = 5%).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Table 1 shows the mean values of inspiratory flow, tidal volume, respiratory system, lung, and chest wall P's, static elastances, and E obtained in the C and P groups. All mechanical parameters increased with augmenting paraquat doses, except P1,w which remained unaltered (Table 1).

Tissue elastance augmented progressively from the C to P25 group (p < 0.0001). R and tissue hysteresivity increased from C to P10 and from P15 to P25 (p < 0.0001). E, R, and  were similar in the P25 and P30 groups (Table 2).

Figure 1 shows the amounts of elastic and collagenous system fibers in alveolar walls from the lungs of the control (C) and acute lung injury groups (P10, P15, P25, and P30). In the control lung the amount of elastic fibers was 12-fold that of collagen. The amount of collagenous system fibers (col) increased exponentially with the severity of the injury (from C to P30, p < 0.0001, col = 0.06 · e^0.08P, r = 0.95). From C to P30, the fibers of the collagenous system showed a 25-fold increase, whereas the content of the fibers of the elastic system approximately tripled (because of the behavior of the oxytalan component). Elaunin and fully developed elastic fibers contents remained unchanged among the five groups (p = 0.22) (Figure 1). The amount of oxytalan fibers (oxy) remained unaltered among the C, P10, and P15 groups, and increased from P15 to P30 (p < 0.0001) (Figure 1). An exponential regression analysis revealed that the experimental data closely fitted the equation: oxy = 0.002 · e^0.18P (r = 0.95).

View larger version (19K): [in this window] [in a new window]   Figure 1.   Amounts of elastic and collagenous system fibers in alveolar walls from the lungs of the control (C) and acute lung injury groups (10, 15, 25, and 30 mg/kg, P10, P15, P25, and P30, respectively). Symbols represent means of five animals in each group (20 microscopic fields/rat); bars are ± SD. FDEF = fully developed elastic fibers. Considering each elastic or collagenous system fibers, different letters indicate values significantly different (p < 0.05) among the C, P10, P15, P25, and P30 groups, and same letters indicate no significant difference between data points.

Strips histological changes included interstitial edema, intraalveolar hemorrhage, inflammation with polymorphonuclear and mononuclear cells, and hyaline membrane formation in the P25 and P30 groups (severe ALI). P15 and P10 strips had thickened alveolar membranes and increased cellularity compared with C strips, but did not demonstrate hemorrhage or hyaline membrane formation (mild ALI). The difference between P10 and P15 was demonstrated by the increment in cellularity and between P25 and P30 was indicated by the augmentation in intraalveolar hemorrhage.

Figure 2 shows the increased content of collagen and elastic fibers in the P30 group (Figures 2B and 2D, respectively) in relation to control group (Figures 2A and 2C, respectively).

View larger version (121K): [in this window] [in a new window]   Figure 2.   Photomicrographs of parenchymal strips stained with Sirius Red with polarization for collagen in control (A) and para- quat-treated lung (30 mg/kg, P30, B). All brightly birefringent structures, which shine against a dark background, contain collagen molecules (A and B). In C and D, representative fields illustrating elastic fiber system distribution in control (C ) and acute lung injury induced by paraquat (30 mg/kg, P30, D) can be seen. Elastic fibers are stained in black within alveolar walls (arrows). Photographs were taken at an original magnification of ×200 from slides stained by Weigert's resorcin-fuchsin with oxidation.

E was sigmoidally related to the amount of collagen (p = 0.0026) (Figure 3). R increased exponentially with the content of oxytalan fibers (p = 0.0008).

View larger version (18K): [in this window] [in a new window]   Figure 3.   Correlation between physiological and morphometric parameters in parenchymal strips from the lungs of the control (circles), P10 (squares), P15 (triangles), P25 (inverted triangles), and P30 (diamonds) groups. Solid lines represent the best curve fitted to the data. (A) E (dynamic elastance) versus the amount of collagen fibers. (B) R (tissue resistance) versus oxytalan fibers content.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This study provides important evidence concerning the pathophysiology of acute lung injury. The dynamic elastance and tissue resistance of the affected lung parenchyma evaluated in an organ bath were significantly higher than those gathered from control parenchyma. These pronounced changes in mechanical behavior at the tissue level happened even in mildly abnormal lung parenchyma (thickened alveolar membranes and increased cellularity, with no alveolar hemorrhage or hyaline membrane formation). Both collagen and elastic components were increased in the present model of ALI, yielding a process of fibroelastosis that started up early (at 24 h) in the evolution of the lesion.

In this study acute lung injury was produced by paraquat in rats. Paraquat is an herbicide that accumulates predominantly in the lung and induces alveolar epithelial damage due to its action on type II pneumocytes. This is an experimental model of diffuse alveolar damage, which by virtue of its low cost, rapid effect, and simplicity of administration has been used to study acute lung injury (21, 22).

According to current concepts, the early phase of acute lung injury is characterized by an increase in pulmonary endothelial and epithelial permeability. Protein-rich edema fluid accumulates in the interstitium and air spaces of the lung, leading to arterial hypoxemia, decreased lung compliance, and increased work of breathing. This early phase is followed by a subacute fibroproliferative phase that may allow repair of injured lung or result in progressive obliteration of the interstitial and alveolar compartments of the lung (1). The present study provided evidence that the fibroproliferative process may be in fact established at 24 h after the induction of acute lung injury.

Prior studies ascribe changes in lung resistance and elastance resulting from acute lung injury to surfactant dysfunction and/ or loss of functional lung capacity due to alveolar flooding (23, 24). These works suggest that mechanical dysfunction rests on the air-liquid interface and does not, therefore, reside at the tissue level (25). Interestingly, the method used in this study to determine tissue mechanical properties avoids the effects of surface film, alveolar flooding, and airway inhomogeneity. We demonstrated that elastance, resistance, and hysteresivity of lung parenchyma of paraquat-treated rats were significantly increased in comparison with control tissue, suggesting that parenchymal mechanical dysfunction plays an important role in the pathophysiology of ALI. In addition, the modifications in the mechanical behavior at the tissue level happened (1) at an early phase of the lesion, (2) even in mildly abnormal lung parenchyma, and (3) advanced according to the severity of the changes in the connective matrix.

Respiratory mechanics were measured by the end-inflation occlusion method in the same animals to provide data from the whole lung. We do not believe that in vivo and in vitro measurements can be compared in a quantitative way, as they reflect the action of different structural factors (Tables 1 and 2). First the contributions of the tissue properties in the lung are different during nearly isotropic three-dimensional stretching and during uniaxial stretching. Second, the presence of surfactant in the lung may distort the alveolar structure, which may then lead to different manifestations of tissue properties in comparison with the tissue strip without surfactant. In addition, after acute lung injury, parallel heterogeneity, surfactant damage, and alveolar flooding dominate whole lung mechanical changes, with a small contribution of connective matrix remodeling. In an ideal monoalveolar lung, a change in whole lung elastance (EL) can be considered as the sum of the relative changes in tissue plus air-liquid interface components, as they act in series. However, in the presence of marked heterogeneity of lung injury, ventilation of the whole lung tends to be selective to the less compromised territories, that is, those in which the alveolar lining is less damaged, and less edema, less airway narrowing, and less tissue damage can be found. Therefore, EL might underestimate whole lung damage. On the other hand, lung tissue strip elastance (E) reflects unselectively the connective matrix behavior. Under these conditions the change in EL is no more the sum of the partial changes of the strip (connective matrix) and surfactant components. This probably explains why lung dynamic elastance is lower than tissue elastance in severe lesions (P25 and P30, Tables 1 and 2). In fact, depending on the amount and distribution of lung injury, and the ventilator settings, EL can be lower, equal, or higher than E. Another reason that could determine the difference between in vivo and in vitro measurements could be an eventual bias in selecting regions that are strongly affected by paraquat. However, this assertive can be discarded, because we always cut the strip from the same region, that is, the periphery of the left lung. Finally, the relative increases in E and EL cannot be attributed to a poor estimation of dynamic elastance, because the end-inflation technique has been extensively proven reliable (26).

Ingenito and colleagues (29) observed in a model of ALI resulting from intravenous endotoxin administration that rheological properties of lung parenchyma were affected at 48 h after a severe lung injury, and concluded that tissue contribution to lung dysfunction in ALI may vary with both the stage of evolution and the severity of the injury. Indeed, we observed that the parenchymal tissue was already altered at 24 h after the injection of even a small dose of paraquat (10 mg/kg), as indicated by the increased E, R, and  values (Table 2).

In a bleomycin model of lung injury the collagen content of the lungs increased 5-fold in hamsters (30), and rat lungs responded similarly by increasing mRNA for collagen ₁ chains, which reached a maximum in 3 weeks (31). Interestingly, under the present experimental conditions the highest dose of paraquat induced a 25-fold increase in collagen content (Figure 1) from a control value of 0.03 µm²/µm. However, 98% of the exponential increase would be reached only if 62.5 mg/kg BW of paraquat was used. The present results disclosed that collagen content was already elevated 24 h after tissue damage whatever the dose of paraquat used, indicating that the biochemical processes implicated in collagen synthesis are indeed able to react very quickly to the aggression.

The amount of elastic fibers increases in line with lung growth (32), and elastin is in fact responsible for alveolar formation (33). Early in development, the elastic fiber consists of microfibrils that define fiber location and morphology (34, 35). Over time, tropoelastin accumulates within the bed of microfibrils to form the functional, polymeric protein known as elastin. The elastic system has three components defined according to crescent amounts of elastin and fibril orientation: (1) oxytalan fiber composed of a bundle of microfibrils; (2) elaunin fiber made up of microfibrils and a small amount of elastin; and (3) fully developed elastic fibers consisting of microfibrils and abundant elastin (36). In experimental emphysema the decrease in fully developed elastin resulting from proteolytic digestion is eventually followed by its resynthesis (37, 38). It should be pointed out that it is the elastin component of elastic fibers that is greatly affected, without concomitant loss of the microfibrillar component. In line with this result, a 2-fold increase in mRNA for elastin was detected 3 wk after administration of bleomycin to hamsters (30). Nevertheless, the increase in elastin content can be secondary to the production of any of its components. The present study was able to pinpoint the kind of fiber responsible for the elevation of elastin content: the total amount of elaunin and fully developed elastic fibers was not modified by ALI, whereas oxytalan microfibrils content was higher than control after P15. Oxytalan fiber content tripled from a control value of 0.029 µm²/µm after the injection of 30 mg/kg of paraquat. The content of oxytalan fibers could reach 98% of its exponential increase if 27.8 mg/kg BW of paraquat was used. In fact, our findings ruled out the possibility that the destruction of fully developed fibers by paraquat-induced ALI leads to an increase in the microfibrillar component, because the amount of fully developed fibers was not modified. Once more, at least for mild to high degrees of aggression, the tissue response was present at 24 h after the initial lesion.

Previous studies on lung tissue strips challenged with elastase and collagenase showed mechanical changes that agree with the classical model of elastin-dependent elastance and collagen-dependent maximal distension (31). Elastase decreased both E and R in a coupled fashion, so hysteresivity was not modified (12, 39). However, recently, Yuan and coworkers (40) reported in normal animals that both collagen and elastic fibers contribute to tissue elasticity during normal breathing, which contradicts the notion of independent functionality of elastin and collagen fibers. We examined potential correlations between the mechanical and morphometric parameters in ALI to investigate the role of the extracellular matrix components in determining parenchymal mechanics. In the present study E increased sigmoidally with the increment in the amount of collagen until a plateau was reached (Figure 3). In fact E reached its maximum values when the collagen fiber content was 0.41 µm²/µm, which corresponded to a dose of 25 mg/kg BW. R augmented with the increment in oxytalan fiber content reaching 98% of its exponential rise at 0.36 µm²/µm. This value was achieved with a dose of paraquat greater than 25 mg/kg. Hence, 24 h after the induction of ALI with paraquat, collagen and oxytalan fibers were important in determining parenchymal mechanics. However, when the lesion was severe (P25 and P30), the content of elastic and collagen fibers kept increasing whereas tissue mechanical properties reached a steady state.

Our results suggest that the microfibrillar component of elastin is especially relevant in the elastic-dissipative coupling at the level of the connective matrix. Elastase cleaves mainly elastic fibers with a high content of elastin, as occurs in emphysema. Conversely in ALI the increase in the amount of elastic fibers depends on the microfibrillar component (oxytalan fibers).

Thus, both collagen and elastic components were increased in the present model of ALI, yielding a process of fibroelastosis that started up early in the evolution of the lesion. As a consequence, this process should be treated as soon as possible during the exudative phase of the disease.