Mechanics, Nonlinearity, and Failure Strength of Lung Tissue Ch 2

DISCUSSION The primary findings of this study are that following PPE-treatment of mice (1) lung elastance decreased and hysteresivity increased, (2) parenchymal tissue fibers failed at a lower stress than in control, (3) dynamic nonlinearity characterized by harmonic distortion (Kd) increased, and (4) total collagen content increased while there was a tendency of loss of elastin of whole lungs in the emphysema group. These physiological observations support the hypothesis that exposure of the lung to PPE leads to a remodeling of the fiber network that significantly alters the mechanical properties and the nonlinear mechanical behavior of the whole organ. Collagen Remodeling in Emphysema In emphysema, the loss of elastin from the alveolar walls appears to be a major event in clinical pathology (1). The breakdown and the degradation of collagen fibers have also reported both in human patients (39) and in rodent models of emphysema (9, 35, 52). Additionally, several studies have demonstrated that after the destruction of alveolar walls, remodeling of collagen fibers as a result of an abnormal repair process contributes to the pathogenesis of emphysema (12, 17, 21, 22, 25, 49). Increases in total amount of collagen of the lungs have been reported in human patients (23, 35). In the present mouse model of emphysema, we also observed a 45% statistically significant increase in total collagen content and a 13% decrease in total elastin content of the whole lung (Table 2), suggesting that collagen remodeling within the lung was indirectly triggered following the elastolytic injury. Previous studies have shown that following the onset and initial progression of emphysema due to the proteolytic injury caused by PPE, synthesis of collagen by lung fibroblasts is considered to be upregulated as part of the repair process of the damaged lung (15, 22). However, the repair process does not restore normal structure and function to the lung leading to pathology and altered physiology (24). Since the extracellular assembly of collagen molecules to fibrils and fibers is sensitive to the composition of the surrounding matrix (22, 24), it is conceivable that the structure and mechanical properties of the newly synthesized collagen differs from those that occur during normal growth. While the details of this abnormal repair process at the molecular level is beyond the scope of the present study, it is important to discuss the physiological consequences. Alveolar Wall and Fiber Failure in Emphysema It has long been proposed that mechanical failure of the alveolar walls plays a pivotal role in the progression of emphysema (50). Recently, Kononov et al. (21) observed the failure of a single alveolar wall in a rat model of PPE-induced emphysema. However, to our knowledge, this is the first study to quantify the failure stress of parenchymal tissue strips from emphysematous lungs. We found evidence that the emphysematous lung tissue breaks at the same strain, but at a stress 40% smaller than the normal tissue (Table 1). In an attempt to interpret these results, we first note that normal collagen fibers are stiffer and stronger than other connective tissue constituents (36). As a consequence, the amount and organization of collagen in the alveolar walls should play a crucial role in determining the stiffness and the failure properties of the lung tissue. For example, when the collagen content of normal lung tissue strips was decreased via in vitro digestion using collagenase, the stiffness of the tissue dropped to 40% of its value before digestion (53). Alternatively, normal developmental changes of the lung increase both the stiffness and the failure strength of the alveolar walls as well as collagen contents during maturation. Indeed, Tanaka and Ludwig (47) reported that the failure stress of normal lung tissue from baby rats was 18 kPa which increased to 28 kPa in adult rats. At the same time, in a follow-up study, Tanaka et al. (46) also found that the collagen content of the lung increased from 18% to about 38% during this normal maturation process. Thus, since in this model of emphysema we found a 45% increase in lung collagen (Table 2), one would expect the stiffness and the failure stress of the tissue to increase. Surprisingly however, our data showed just the opposite behavior: despite the increase in collagen content, both lung elastance and the failure stress decreased by about 30% (Figs. 2 and 4) and 40% (Table 1), respectively, indicating that the total amount of collagen in the lung tissue is not the primary determinant of the mechanical properties in the diseased state. In an attempt to resolve the apparent contradiction between increased collagen content and decreased elastance and failure stress in the emphysematous tissue, we first discuss the relation of lung stiffness and alveolar structure. Since the MLI nearly doubled in the emphysematous group, the number of alveolar walls per unit volume that can resist the deformation of the tissue strip could have decreased. This mechanism alone could account for the lower elastance and lower failure stress even if the mechanical properties of the alveolar walls were similar to those of the normal tissue. However, it is likely that not all of the increase in MLI is due to alveolar wall rupture. The PPE treated lung is softer and at the fixation pressure of 25 cmH2O the alveoli would be more extended than those in the normal lung. Thus, the number of alveolar walls per unit volume in the tissue strip is not necessarily smaller in the PPE treated lung than in the normal. Using microscopic imaging, Brewer et al. (3) recently reported that individual alveolar walls from PPE-treated rats, which also involved collagen remodeling (21), appeared softer and more extensible than those from normal rats. These observations suggest that despite the increased collagen content, the alveolar walls and the collagen fibers are likely to be weaker in the emphysematous lung as a consequence of the process of degrading and remodeling. Indeed, the ultrastructure of collagen from human emphysematous lungs reveals thickened and disorganized fibrils after remodeling (12). Our data then suggest that the stiffness and the failure properties of the remodeled fibers must decrease compared to normal collagen fibers. The reduction in the failure stress of collagen has an important effect on how the structure of the lung evolves during the progression of emphysema. Suki et al. (44) recently developed a fiber network model and argued that since mechanical forces influence the process of tissue breakdown, the alveolar structure must be very heterogeneous and the alveolar walls around the perimeter of severe emphysema lesions or the walls that separate such lesions may be overstretched. In agreement with these predictions, the heterogeneity of the alveolar dimensions was found to be much larger in the emphysematous than in the control lungs, both in the current study and in previous studies (10, 20, 34). Therefore, the alveolar walls in the emphysematous lung may have to oppose larger stresses locally, and as a result, the increased local stresses can promote rupture of the remodeled walls (44), which in turn results in a decreased failure stress observed in Table 1. We thus conclude that mechanical forces are expected to play an important role in the progression of emphysema once the collagen matrix has undergone a critical amount of remodeling. Lung Mechanical Properties The consequences of alterations in the ECM of alveolar walls can be traced to organ level changes in the mechanical properties of the lung. Since the chest wall is very soft in the mouse, at least 90% of H is due to the lung parenchyma (37). Additionally, any change in H must be related to a change in lung mechanics and hence in the discussion that follows we assume that changes in H largely reflect changes in lung mechanics. We have recently developed a new mathematical model which assumes a continuous distribution of H between a minimum and a maximum value (Hmin and Hmax, respectively) (20). We found that all H-related parameters (Fig. 4) as well as the static elastance (Fig. 2) decreased in the emphysematous mice compared to controls. The Hmax represents the stiffest regional elastance in the lung and the collagen should be the most important determinant of its value. Thus, the lower Hmax values in the PPE-treated mice suggest that the ultrastructural changes of remodeled collagen weakens the fibers and the alveolar walls in agreement with the analysis of the failure tests. On the other hand, we speculate that Hmin represents the softest regional elastance, which may be related to the loss of alveolar walls in that region. Thus, it is likely that the lower value of Hmin is a functional consequence of the increased MLI in the emphysematous mice. The hysteresivity is a material property of the lung tissue (14) and it also depends on the microscopic constituents of the alveolar walls. Indeed, changes in ECM composition can cause a change in the hysteresivity in the parenchymal tissue level (33, 53). In the present study, hysteresivity of the PPE-treated mice was higher than that of the control mice, as observed in TGF-α transgenic emphysematous mice (32), in mild emphysematous mice induced by nebulized PPE-treatment (20) as well as in rats (3). In parenchymal tissues of normal guinea pigs, hysteresivity after in vitro digestion with collagenase was significantly higher than that after digestion with elastase (53). This suggests that in the normal lung tissue, the larger the elastin-to-collagen ratio the larger the value of hysteresivity. However, compared to controls, hysteresivity of the emphysematous mice increased (Fig. 5A) while collagen content also increased (Table 2). Taken together, the hysteresivity and elastance results suggest that remodeling in emphysema produces weak and viscous alveolar walls that also fail at lower stresses than those of the normal lung. The values of Raw decreased with increasing PEEP most likely due to the increasing diameters of the airways with lung inflation in both groups (Fig. 5B). Although increased Raw values, which suggest an underlying airway obstruction, were reported in sheep with experimental emphysema following papain treatment (19), Raw values were the same between the groups, perhaps because the effects of PPE treatment on the airways used in this murine model is distinct from that of papain that was used in the ovine model. To our knowledge, increases in Raw have not been reported in mouse models of emphysema yet. In another emphysema model of surfactant protein D deficient mice, Raw became lower than in control (5). Thus, the physiological feature of the present model is similar to the classical physiological alterations in patients with α1-antitrypsin deficiency (2), in which airway conductance was within normal limits, and the primary physiologic defect was a loss of elastic recoil. Since surfactant plays an important role in the mechanical properties of the normal lung, it is conceivable that alterations in surfactant properties in the emphysematous lung could influence our results. Because fewer type II pneumocytes that secrete surfactant were observed in the lungs of human emphysema patients (31), it has been suggested that surfactant plays a protective role against the development of pulmonary emphysema. In PPE-treated mice, administration of surfactant prevents the development of emphysema (30). However, it is not known whether the composition and biophysical properties of lung surfactant change as a consequence of the development of emphysema. While the expression of surfactant protein A messenger RNA is increased in the lungs of the klotho mouse model of emphysema (41), this protein plays a less important role in stabilizing the alveoli than the hydrophobic surfactant proteins (18). One could argue that due to the reduction in lung tissue recoil, surface tension may become even more important in emphysema than in the normal lung. However, emphysema is associated with pronounced heterogeneity at the alveolar level, and it is unclear whether abnormalities in surfactant actually contribute to the development of the disease. The extent to which lung surfactant contributes to recoil in emphysema has also not been well characterized. Further studies would be needed to clarify the role of surfactant in emphysema. Dynamic Nonlinearities Another important physiological finding of this study is that the tissues responsible for generating elastic recoil in the emphysematous lungs also displayed significantly greater nonlinear behavior than control lungs. The mechanical behavior of the normal lung tissue has been characterized as nonlinear (28, 29, 42, 53), and the origin of dynamic nonlinearity has been investigated in various organs (11). In the respiratory system, dynamic nonlinearity is likely related to the ECM components, including the nonlinearly viscoelastic collagen and its interactions with the linear viscoelastic elastin, and the viscous ground substance including mainly proteoglycans (29). Since elastic fibers behave more linearly than collagen fibers (28, 36), tissue nonlinearity could be more related to collagen and in particular, the extent to which collagen fibers are stretched in the alveolar walls. Thus, the dynamic nonlinear behavior of the lung tissue can be considered as a global in vivo assay of collagen function in the intact lung. While the nonlinearity is certainly related to collagen, it is also possible that the physical interaction between collagen and elastin also influences nonlinear behavior. In the present study, we demonstrated for the first time that dynamic nonlinearity of the lung, as characterized by the harmonic distortion index Kd (42), is linearly related to lung elastance parameters (Hmin and Hmax) both in the PPE-treated and the control mice (Fig. 7), as found for normal tissue strips (53). More importantly, despite a decrease in H, the Kd increased in emphysema compared to normal lungs. Specifically, the relationship between Kd and Hmin as well as Hmax shifted to the left in emphysema. This can be accounted for by changes in ECM components. As discussed above, a decrease in elastance and an increase in collagen content suggest that the new collagen in the remodeled alveolar walls must be less stiff than the normal collagen. An increase in Kd on the other hand, suggests that the collagen fibers are either more stretched or inherently different from normal with respect to their nonlinear mechanical behavior in the emphysematous alveolar wall. One may argue that in contrast to tissue strips, in the whole lung airway closure also contributes to harmonic distortion. If a significant portion of the lung is blocked by airway closure and tidal volume remains the same, then a smaller lung will receive the same tidal volume and the lung becomes overstretched. In fact, the Kd was largest at PEEP = 0 in both groups (Fig. 6), which is the condition where recruitment and derecruitment could occur most during oscillations. Furthermore, the decrease in Kd with PEEP implies gradual recruitment. However, neither lung elastance nor Kd decreased when PEEP was increased from 6 to 9 cmH2O. Thus, these data suggest that above 6 cmH2O PEEP, recruitment did not occur and hence it could not influence our data. Therefore, we believe that the relation between Kd and H and consequently the above interpretation of the results, is insensitive to airway closure at least at the higher PEEPs included in this study. In summary, we have characterized the respiratory and lung mechanical properties of a mouse model of emphysema induced by PPE. We observed a decrease in lung elastance and failure strength of the alveolar walls as well as an increase in hysteresivity, dynamic nonlinearity and total lung collagen content. These results suggest that significant collagen remodeling takes place within the alveolar wall, which produces weak but more nonlinear fibers and alveoli that are locally overstretched and hence prone to mechanical failure. These alterations in the micromechanics of the alveolar walls significantly affect organ level lung function in the mouse. FIGURE LEGENDS Fig. 1. Examples of the alveolar structures of lung tissues from the control (A) and the porcine pancreatic elastase (PPE)-treated (B) lungs. Photographs were taken at an original magnification of x 50 from slides stained with hematoxylin and eosin. Fig. 2. Mean ± SD of quasi-static inspiratory pressure-volume relationships from end-expiratory lung volume of the control and the PPE-treated mice. Values are means ± SD. * P < 0.05. Fig. 3. Representative examples of respiratory system resistance (A) and elastance (B) of the control and the PPE-treated mice calculated from the impedance data and corresponding fits of the model to the control and PPE-treated data at a positive end-expiratory pressure (PEEP) of 3 cmH2O. Fig. 4. Dynamic tissue elastance (H) parameters as a function of PEEP. Means ± SD of minimum (Hmin), maximum (Hmax) (A), and mean H (Hmean; B) in the control and PPE-treated mice. * P < 0.05. Fig. 5. Means ± SD of hysteresivity (A), airway resistance (Raw)(B), and tissue damping (G)(C) in the control and PPE-treated mice as a function of PEEP. * P < 0.05. Fig. 6. Harmonic distortion, an index of dynamic nonlinearity of the respiratory system, of the control and the PPE-treated mice as a function of PEEP. * P < 0.05. Fig. 7. Correlations between Hmin (A) or Hmax (B) and harmonic distortion. Data were obtained from of the control (n = 8) and the PPE-treated (n = 9) mice at PEEP = 0, 3, 6, and 9 cmH2O. Fig. 8. Representative examples of parenchymal tissue failure tests of the control and the PPE-treated mice. Notice that the strain axis starts at 2.0. Arrows indicate failure points. Table 1. Tissue fiber rupture test Failure stress (kPa) Strain Control 11.9 ± 1.4 3.33 ± 0.49 PPE-treated 7.2 ± 2.4 3.59 ± 0.93 P = 0.002 P = 0.56 Data are expressed as means ± SD (n = 6 in each group). PPE; porcine pancreatic elastase. Table 2. Lung collagen and elastin contents Dried lung weight Hydroxyproline α-Elastin (mg) (µg/mg) (µg/mg) Control 13.5 ± 3.8 4.38 ± 0.83 47.2 ± 6.0 PPE-treated 13.1 ± 2.4 6.38 ± 1.33 41.2 ± 8.8 P = 0.81 P = 0.002 P = 0.16 Data are expressed as means ± SD. Hydroxyproline and α-elastin contents are expressed as µg/mg of dried lung weight (n = 7 in each group). Fig. 1A and B A B Fig. 2 35 Control PPE-treated Pressure (cmH2O) 30 25 20 * 15 * 10 5 * * * * * 0 0.0 0.2 0.4 0.6 0.8 Volume (ml) 1.0 1.2 Fig. 3A and B Resistance (cmH2O*s/ml) A 1.4 Control data Control model PPE-treated data PPE-treated model 1.2 1.0 0.8 0.6 0.4 0.2 0.0 0 2 4 6 8 10 12 14 16 12 14 16 Frequency (Hz) Elastance (cmH2O/ml) B 40 30 20 10 0 0 2 4 6 8 10 Frequency (Hz) Fig. 4A and B Hmin and Hmax (cmH2O/ml) A 80 Hmax (Control) Hmax (PPE-treated) Hmin (Control) 60 Hmin (PPE-treated) 40 * * * * * * 3 6 9 20 * 0 0 PEEP (cmH2O) Hmean (cmH2O/ml) B 50 Control PPE-treated 40 30 * 20 * * * 6 9 10 0 0 3 PEEP (cmH2O) Fig. 5A and B A 0.25 * Hysteresivity 0.20 * 0.15 0.10 Control PPE-treated 0.05 0.00 0 3 6 9 PEEP (cmH2O) B 0.7 Control PPE-treated Raw (cmH2O*s/ml) 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0 3 6 PEEP (cmH2O) 9 Fig. 5C C 8 G (cmH2O/ml) Control PPE-treated 6 * 4 * * * 6 9 2 0 0 3 PEEP (cmH2O) Fig. 6 10 * Control PPE-treated Kd (%) 8 6 * 4 * 2 0 0 3 6 PEEP (cmH2O) 9 Fig. 7A and B A 12 Control PPE-treated 10 Kd (%) 8 6 4 2 0 0 5 10 15 20 25 30 Hmin (cmH2O/ml) B 12 Control PPE-treated 10 Kd (%) 8 6 4 2 0 0 20 40 Hmax (cmH2O/ml) 60 80 Fig. 8 12 10 s Stress (kPa) 10 8 Control 6 4 PPE-treated 2 0 2.0 2.5 3.0 Strain 3.5 4.0