Journal of Applied Physiology Watch the video to see how APS reaches out to developing nations.
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

J Appl Physiol 103: 803-811, 2007. First published May 31, 2007; doi:10.1152/japplphysiol.00057.2007
8750-7587/07 $8.00
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow All Versions of this Article:
103/3/803    most recent
00057.2007v1
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Jesudason, R.
Right arrow Articles by Suki, B.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Jesudason, R.
Right arrow Articles by Suki, B.

Differential effects of static and cyclic stretching during elastase digestion on the mechanical properties of extracellular matrices

Rajiv Jesudason,1 Lauren Black,2 Arnab Majumdar,1 Phillip Stone,3 and Bela Suki1

1Department of Biomedical Engineering, Boston University, Boston, Massachusetts; 2Department of Biomedical Engineering, University of Minnesota, Minneapolis, Minnesota; 3Department of Biochemistry, Boston University School of Medicine, Boston, Massachusetts

Submitted 12 January 2007 ; accepted in final form 29 May 2007


   ABSTRACT
TOP
 ABSTRACT
METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
Enzyme activity plays an essential role in many physiological processes and diseases such as pulmonary emphysema. While the lung is constantly exposed to cyclic stretching, the effects of stretch on the mechanical properties of the extracellular matrix (ECM) during digestion have not been determined. We measured the mechanical and failure properties of elastin-rich ECM sheets loaded with static or cyclic uniaxial stretch (40% peak strain) during elastase digestion. Quasistatic stress-strain measurements were taken during 30 min of digestion. The incremental stiffness of the sheets decreased exponentially with time during digestion. However, digestion in the presence of static stretch resulted in an accelerated stiffness decrease, with a time constant that was nearly 3x smaller (7.1 min) than during digestion alone (18.4 min). These results were supported by simulations that used a nonlinear spring network model. The reduction in stiffness was larger during static than cyclic stretch, and the latter also depended on the frequency. Stretching at 20 cycles/min decreased stiffness less than stretching at 5 cycles/min, suggesting a rate-dependent coupling between mechanical forces and enzyme activity. Furthermore, pure digestion reduced the failure stress of the sheets from 88 ± 21 kPa in control to 29 ± 15 kPa (P < 0.05), while static and cyclic stretch resulted in a failure stress of 7 ± 5 kPa (P < 0.05). We conclude that not only the presence but the dynamic nature of mechanical forces have a significant impact on enzyme activity, hence the deterioration of the functional properties of the ECM during exposure to enzymes.

stress-strain curve; stiffness; network model; emphysema

THE EXTRACELLULAR MATRIX (ECM) is the structural entity that surrounds and supports the cells of tissues and organs. The ECM is comprised of a number of proteins that are specific for any given tissue. Collagens, proteoglycans, elastin, and glycoproteins all serve specific functions and establish the basic structural composition within the ECM (8). Among these components, elastin plays an important structural and biological role in many tissues, including blood vessels and lungs (6, 28). The elastin fibers consist of two components: an insoluble amorphous component, and a microfibrillar component surrounding the core (20). The insoluble component carries out much of the mechanical function of the ECM, due in part to its ability to store energy with little loss (28). In various diseases, elastin is enzymatically digested by the serine protease, elastase (12, 13). For example, in pulmonary emphysema, the walls of the elastic air sacs of the lungs, the alveoli, are permanently destroyed due to the weakening of the ECM. This process is thought to be triggered by the digestion of elastin (22, 27). Similarly, vessel wall aneurysm, an area of localized blood vessel dilation, is a complication linked to the destruction of elastin in blood vessel walls (16, 30).

Since elastin is an important determinant of the elastic properties of tissues such as the walls of blood vessels, biochemical degradation of elastin is expected to lead to an observable macroscopic compromise of the mechanical function of these tissues. Previous studies have investigated the changes in mechanical properties of elastin-rich ECM sheets during digestion by elastase (2, 11, 17). For example, the elastolytic damage to elastin fibers caused by pulmonary emphysema and aneurysm was modeled by digestion of thin-tissue engineered ECM sheets by elastase, which resulted in a significant decrease in both the stiffness and failure stress of the tissue (2). In another study, the alveolar walls of rat lungs that had been treated in vivo with elastase to mimic emphysema were found to rupture when subjected to mechanical forces similar to those during breathing (11).

The effects of elastolysis on the mechanical and failure properties of tissues have invariably been studied in already digested and/or remodeled tissues. However, in both the lungs and blood vessels, inherent mechanical forces in the form of preexisting tensile stress (prestress) are present even during elastolytic activity. It is conceivable that this prestress also contributes to the breakdown of tissue. Indeed, at the molecular level, mechanical forces can alter enzyme activities in a dramatic fashion (10). If for example a protein undergoes a structural change from a more globular state to an extended state due to stretching, various domains may unfold and become accessible. Such domains may have additional binding sites for enzymes. Alternatively, the prestress itself can exert an extra force on molecular bonds. Both of these effects would facilitate the enzymatic digestion of the tissue. Furthermore, the prestress is not static but cyclically changes both in the lung tissue and the blood vessel wall. The effect of such time-varying forces on enzyme activity is unclear.

In this study, we hypothesized that the presence of static prestress accelerates the enzymatic breakdown of the ECM, which is further amplified when the prestress undergoes cyclic variations. To test this hypothesis, we measured the mechanical properties of cell culture-based ECM sheets at specific time points during elastase digestion while undergoing static or cyclic stretching followed by failure tests. Our results suggest that there is a complex interaction between mechanical forces and enzymatic digestion.


   METHODS
TOP
 ABSTRACT
 METHODS
RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
Cell culture.   ECM sheets were obtained from neonatal rat aortic smooth muscle cells (NNRSMC) isolated from Sprague-Dawley rats (age: 1–3 days) by means of a previously described method (26). The experimental protocol was approved by the Institutional Animal Care and Use Committee at Boston University Medical Center. NNRSMCs are highly elastogenic and produce thin ECM sheets. The NMRSMCs were grown in a culture containing 3.1g/l sodium bicarbonate, 1% sodium pyruvate, 1% penicillin and streptomycin (DV 3.7), and 20% FBS. The medium was changed twice a week, and the culture was monitored routinely by phase contrast microscopy. After a period of 6 wk, the cells were killed with 5% sodium azide in Puck's saline and stored at 4°C. The cultures were then infiltrated with a gelatin solution, which solidified and allowed the thin ECM sheets to be lifted from the flasks without damage. The resulting ECM sheet was thin and comprised of mostly elastin and proteoglycans (2).

Experimental setup.   Failure tests and stress-strain curves were obtained by means of a previously developed uniaxial tissue stretching system (2), consisting of a computer-controlled dual-mode lever arm-force transducer system (model 300B, Aurora Scientific, Ontario, Canada) and a separate, smaller, more sensitive force transducer (model 403A, Aurora Scientific), both attached to an acrylic test stand. The sensitive force transducer was used to measure the stress-strain curves, while the dual-mode system stretched the samples and measured force only during failure tests. The ends of the ECM sheets were attached to the lever arm and sensitive transducer (see Experimental procedure). The sample was then placed inside a tissue bath chamber. A labVIEW (National Instruments, Austin, TX) program was used to operate the stretching system and record displacement and force data. The displacement signals generated by the computer program were led through a digital-to-analog converter and low-pass filtered (901P Filter Bank, Frequency Devices, Haverhill, MA) before being sent to the lever arm. The force response of the sample was recorded by the sensitive force transducer. The recorded signals were first low-pass filtered at 10 Hz (901P Filter Bank, Frequency Devices, Haverhill, MA) and sampled by a data acquisition board (DAQCard-6062E, National Instruments) and connector block (BNC-2110, National Instruments) at a sampling rate of 30 Hz.

Experimental procedure.   The gelatin-infiltrated ECM sheets were cut into strips with dimensions 5 x 15 mm. Small metal plates (5 x 15 mm) were fixed to each end of the sample 5 mm apart with cyanoacrylate glue. This provided a working area of 5 x 5 mm on the ECM strip. The strip was then attached to steel wires connected to the force transducers via the adhered metal plates. After the sample was attached, the bath was filled with 22 ml of PBS, and the entire test stand was placed on a hotplate until the gelatin was completely dissolved (50°C), leaving only the ECM strip in the stretching system. The test stand was then taken off the hotplate, the PBS with the dissolved gelatin was removed from the bath with a transfer pipette, the sample was rinsed 3 times, and the bath was filled with fresh, room-temperature PBS.

After the samples were secured to the system, they were preconditioned by application of three consecutive triangular displacement wave signals, peaking at 25% strain, defined as displacement divided by the initial length of the sample. After preconditioning and a 5-min equilibration time, testing commenced by first measuring the quasistatic stress-strain curve of the samples (see below). Originally, a total of eight experimental groups of ECM sheets were used. In the first four groups no digestion was applied, whereas in the remaining four groups porcine pancreatic elastase (Sigma, St. Louis, MO) was added to the bath at a final concentration of 0.06 IU/ml. In the first (n = 4) and fifth (n = 10) groups no load, in the second (n = 4) and sixth (n = 7) groups a static load, and in the remaining groups a cyclic load was applied to the samples. The stretch during loading was uniaxial. Since in the absence of elastase the load had no effects on the mechanical and failure properties, the first four groups were combined in the final data analysis and presented as the control group (n = 16). In the static load group, the samples were held at 40% strain, whereas in the cyclic load groups cyclic stretching was applied with a wave that consisted of a series of consecutive half-sine waves, peaking at 40% strain at a rate of either 5 (n = 9) or 20 (n = 9) cycles/min. Following the measurement of the first stress-strain curve, the selected load pattern was applied for 5 min, and then the load was reduced to 0% strain for a short period while the second stress-strain curve was measured. After data collection, the same loading pattern was resumed, and the measurement sequence was repeated, collecting stress-strain curves at 10, 20, and 30 min. In a separate set of experiments, additional samples went through identical treatment (n = 5 for each group) as defined above, except that at the end of the stretching protocol, the samples were stretched to ultimate failure to obtain the failure stress and strain values. The protocol is summarized in Fig. 1.


Figure 1
View larger version (32K):
[in this window]
[in a new window]

  		Fig. 1. Schematic of 30-min testing protocol. Shaded areas indicate continuous static or cyclic mechanical stretch to 40% strain. Dark grey circles indicate stress-strain (SS) measurements to 25% strain. Input signals for the mechanical loading and SS measurement are shown in the dashed boxes.

 
During the quasistatic stress-strain measurement, the samples were stretched uniaxially to 25% strain at a rate of 0.75% strain per second, using a triangular wave. Stress (T) was defined as force divided by the cross-sectional area of the sample, and strain ({varepsilon}) was computed as displacement divided by the initial length of the sample. The initial length of the sample was determined during slow stretching as the length of the sample at which the sensitive force transducer first detected noticeable force.

The thickness of the ECM sheets was determined by means of laser scanning confocal microscope (Olympus FV-1000). Since native elastin autofluoresces, no specific labeling was necessary during imaging. First, the emission spectrum was mapped using a 488-nm excitation, and multiple stacked images generated from emission between 500 and 600 nm and differing in z-position with steps of 1.1 microns were collected. The thickness of the samples was then determined from the variation in emission intensity with z-position under 20x magnification as the half-width of the intensity profile. A total of 10 locations were selected, and the final width was the average of 10 local thickness measurements. The thickness determination was done in two samples both before and after digestion and subsequently used in the stress calculation.

Spring network model.   To gain insight into the mechanical behavior of the ECM sheets during digestion in the presence and absence of mechanical forces, we extended a previously developed two-dimensional network of nonlinear elastic springs joined by pin joints (3). The springs represented the mechanical behavior of elastin fibers that make up the ECM. The constitutive relation of the springs was a second-order force-extension relationship obtained as the derivative of the elastic energy (ES) with respect to the extension of the spring ({Delta}l):

Formula 1(1)
where k0 and k1 are the linear and nonlinear spring constants, respectively. The springs were arranged in a hexagonal lattice, with the nodes at the top and bottom boundaries of the network fixed and the internal and lateral nodes free to move. To mimic the heterogeneity characteristic of the fibrous ECM, we first created a homogeneous network with unit length for the springs and 120° resting angle between neighboring springs. The nodes (except those at the top and bottom) were randomly displaced by about 10% of the internode distance, and the new length of each spring was recorded as the initial length. Since hexagonal networks are unstable against uniaxial stretch and shear (23), the network was stabilized by adding elastic resistance against changing the angles at each node. This constraint was implemented by introducing a term in the total elastic energy of the network, which is a function of the change in the angle between two interconnected springs ({Delta}{theta}) as follows:

Formula 2(2)
where the subscript a refers to angular contribution, and r characterizes the magnitude of the contribution. The constant r, also called torsional spring constant, or bond-bending constant (1), has been shown to be related to the mechanical properties of the proteoglycan matrix in compression and shear (3). Parameter values in the model (k0 = 10, k1 = 1, r = 0.11) were selected based on approximate manual adjustment to provide stress-strain curves similar to the experimental ones.

We constructed hexagonal lattices comprised of 698 springs and 225 hexagonal unit cells. Three separate conditions were modeled: control, pure digestion, and digestion during static stretch at a constant 40% uniaxial strain. Stress was calculated by numerically differentiating the total energy of the network. For the two digestion simulations, the breakdown of the ECM was mimicked by a probabilistic procedure, in which the probability P to eliminate a spring from the network was given by:

Formula 3(3)

Here P0 is the load-independent probability that a spring will be cut, and hence it is related to enzyme activity. The quantity {varepsilon}l is the local strain on the spring, and c is a constant characterizing the coupling between the strain on the fiber and the enzyme activity, such that c = 0 if no enzymes are present. In the pure digestion simulation, the network was not stretched, and because {varepsilon}l = 0, the rate of elimination of fibers from the system depended only on P0. In the digestion with static stretch, {varepsilon}l was different from 0, which increased the probability of fiber elimination and hence the rate of network breakdown. Parameter values were chosen as P0 = 0.05 and c = 0.025. This probabilistic breakdown procedure was also implemented with a force-based coupling in which {varepsilon}l in Eq. 3 was replaced with the local force fl on the springs, fl= k0{Delta}l.+ k1{Delta}l2. In this case, the value of c was reduced to c = 0.0025 to emphasize the contribution of the nonlinear term (k1{Delta}l2) of the force-extension relation. To test the effect of nonlinearity on breakdown, simulations were run with k1 = 1 and 100. The breakdown and the calculation of the incremental modulus were computed at regular intervals, representing digestion time in arbitrary units. The network simulations were then repeated 6 times for each condition, and the moduli were averaged.

Stress strain curve analysis.   Both the experimental and the simulated quasistatic stress-strain curves were analyzed by determining the incremental modulus (Y) at 10% strain on the plot of T(t) as a function of {varepsilon}(t). The incremental modulus was defined as the slope of the curve between 9.5 and 10.5% strain. This measure of stiffness was calculated for each time point during experimental testing as well as from each simulation.

Statistical analysis.   Statistical analysis of the pooled incremental modulus Y as a function of time was carried out using two-way repeated-measure ANOVA. Other parameters, such as the failure stress, were analyzed via one-way ANOVA. All pairwise multiple comparisons were analyzed through the Student-Newman-Keuls method. Statistical significance was defined as P < 0.05 for all methods.


   RESULTS
TOP
 ABSTRACT
 METHODS
 RESULTS
DISCUSSION
 GRANTS
 REFERENCES
 
Quasistatic stress-strain properties.   Figure 2A shows typical examples of stress-strain curves before and after 30 min of digestion. Example incremental moduli, Y, normalized by their baseline value at time 0 (Yn) are shown in Fig. 2B as a function of time during digestion.


Figure 2
View larger version (11K):
[in this window]
[in a new window]

  		Fig. 2. A: Example SS curves before (solid line) and after (dashed line) 30 min of elastase digestion. B: Examples of the time course of the normalized incremental modulus (Yn) for pure digestion (bullet) and static stretch ({square}) throughout 30 min of testing with exponential fits (dashed lines).

 
The means and standard deviations of Yn for the different groups were pooled and plotted as a function of time for each testing protocol on a semilog graph (Fig. 3). Note that in all digestion groups, Yn on a logarithmic graph decreases linearly as a function of time, suggesting that Yn follows an exponential decay. As expected, the control group maintained a steady Yn throughout testing without any significant decrease in amplitude. There was a statistically significant decrease in Yn after 20 min from the addition of elastase in the pure digestion and digestion with 20 cycles/min groups compared with the control group (P < 0.05). However, digestion in the presence of static stretch and 5 cycles/min cyclic stretch showed a much greater decrease in Yn with time than the other groups. From 5 min onwards, there was a statistically significant difference between the static stretch group and the pure digestion and digestion with 20 cycles/min stretch groups (P < 0.05). By the end of the 30-min digestion test, the samples under static stretch lost much of their stiffness, exhibiting amplitudes ~2% of their baseline values. Samples digested during cyclic stretch at a rate of 5 cycles/min exhibited a similar decrease in Yn, and the values were also significantly less than those during pure digestion from 5 min onwards. There was no significant difference between the digestion with 20 cycles/min group and the pure digestion group. Among the cyclically stretched groups, the resulting Yn was significantly lower in the 5 cycles/min cyclic stretch group than the 20 cycles/min group from 10 min onwards (P < 0.05). Two-way repeated-measures ANOVA found a significant interaction between time and treatment (P < 0.001), indicating that the rate of stiffness decrease depended on the mechanical loading of the samples.


Figure 3
View larger version (12K):
[in this window]
[in a new window]

  		Fig. 3. Time course of the mean and standard deviation of Yn. Note that the SD bars are not shown for the pure digestion and cyclic stretch groups to enhance clarity of the plot.

 
The Yn parameters as a function of time were fitted with an exponential decay, Yn(t)=e–t/{tau}, as shown in Fig. 2B. The time constant ({tau}) for each test sample was pooled and compared among testing protocols (Fig. 4). The pure digestion group had a time constant of 18.4 min, whereas the static stretch with digestion condition had a significantly lower time constant of 7.1 min (P = 0.012), indicating a nearly 3x faster rate of decay in the presence of static mechanical force. The 5 cycles/min cyclic stretch group had a time constant of 11.2 min. Finally, the 20 cycles/min cyclic stretch group had a mean time constant of 32.4 min, which was significantly higher than that of the digestion with static stretch group (P = 0.021).


Figure 4
View larger version (11K):
[in this window]
[in a new window]

  		Fig. 4. Comparison of the mean and SD of time constants for the pure digestion, static stretch, and cyclic stretch groups in minutes. *P < 0.05 between all groups except pure digestion and digestion with 5 cycles/min cyclic stretch, determined by one-way repeated-measure ANOVA.

 
Failure properties.   An example of the stress-strain curve during failure test is shown in Fig. 5, which also indicates how the failure stress and failure strain values were obtained. The mean failure stress values among the various groups are compared in Fig. 6A. Overall, there was a significant difference among the group means (P < 0.001). The control group exhibited a failure stress of 87.7 ± 20.6 kPa, which was significantly reduced with the introduction of digestion and mechanical loading (P < 0.001). The pure digestion group had a failure stress of 29 ± 14.8 kPa, while the digestion with static stretch resulted in a lower failure stress of 7.2 ± 5.5 kPa. Digestion during cyclic stretch at a rate of 5 and 20 cycles/min resulted in failure stresses of 4.8 ± 2.1 kPa and 10.5 ± 16.3 kPa, respectively. Despite the clear trend in Fig. 6A, there was no significant difference among the groups that were mechanically loaded during digestion.


Figure 5
View larger version (9K):
[in this window]
[in a new window]

  		Fig. 5. An example SS curve collected during the failure test protocol. Indices of failure stress and strain are indicated.

 

Figure 6
View larger version (10K):
[in this window]
[in a new window]

  		Fig. 6. Comparison of the means and standard deviations of the failure stress (A) and failure strain (B). *P < 0.05 between groups.

 
In Fig. 6B, the mean failure strains among the various groups are compared. The control group exhibited failure at 65 ± 30% strain, the pure digestion group at 71 ± 22% strain, the digestion with static stretch group at 67 ± 11% strain, the digestion with 5 cycles/min cyclic stretch group at 62 ± 7% strain, and finally the digestion with 20 cycles/min stretch at 66 ± 9% strain. There was no significant difference between any of the five groups.

Spring network model.   As shown in Fig. 7, the simulated digestion showed that Yn decreased nearly linearly on a semilogarithmic graph. Furthermore, Yn with static stretch exhibited a more rapid and substantial decrease than during simulated pure digestion. These results correspond well with the changes in the experimental incremental modulus for the respective groups (Fig. 3). Simulated digestion in the presence of static stretch with force-based elimination of springs showed no significant difference with the strain-based breakdown at k1 = 1. A statistically significant difference was obtained only when the nonlinear spring constant was k1 = 100. Two-way repeated-measures ANOVA indicated a significant difference in the mean values among the different types of treatment after allowing for the effect of differences in time (P = <0.001).


Figure 7
View larger version (10K):
[in this window]
[in a new window]

  		Fig. 7. Yn throughout the progression of the simulation of digestion. SD bars were obtained from 6 repeated simulations. k1, nonlinear spring constant.

 
Figure 8 displays the configuration of the hexagonal networks after the completion of simulations. Figure 8A displays the network after the simulated pure digestion condition. The deterioration is evenly distributed throughout the network. In the simulated condition of digestion with static stretch (Fig. 8B), areas of cut springs are more localized, creating a heterogeneous network with pockets of defect areas rather than an evenly distributed destruction of springs.


Figure 8
View larger version (59K):
[in this window]
[in a new window]

  		Fig. 8. Final network model landscape for the pure digestion (A) and digestion with static stretch (B) simulations.

 

   DISCUSSION
TOP
 ABSTRACT
 METHODS
 RESULTS
 DISCUSSION
GRANTS
 REFERENCES
 
To our knowledge this is the first study that has investigated the effects of simultaneous enzyme activity and mechanical forces on the mechanical properties of ECM. Our main finding is that there is a strong interaction between enzyme activity and the dynamic nature of mechanical forces within the fibrous elastin network of the ECM. Specifically: 1) the rate of decrease of stiffness of ECM sheets during digestion follows a simple exponential decay independent of the mechanical loading; 2) the time constant associated with the exponential decay systematically varied with the mechanical loading condition, that is, the faster the cyclic stretch, the longer the time constant; 3) at the highest cycling rate, the time constant was 30% higher than during pure digestion, suggesting a possible protective role of cyclic stretching; and 4) the failure stress of the ECM sheets was reduced significantly when digestion occurred in the presence of mechanical forces. Before the physiological implications are presented, we first discuss the limitations of study.

The experimental apparatus used in this study was previously developed by Black et al. (2). This system was designed specifically to test the mechanical and failure properties of thin ECM sheets. The ECM sheets used in this study were very thin (~20 microns) and fragile, which made it necessary to always keep the samples in solution. The sheets were lifted in a gelatin base so that the samples could be loaded onto the stretching apparatus without compromising their integrity. The bath provided a means of melting the gelatin from the ECM sheet. It should be noted that the melting process does not affect the chemistry or crosslinking of elastin. Indeed, studies have shown that aortic elastin isolated with NaOH at 95°C was not different in amino acid composition from elastin isolated with cyano bromide in formic acid at room temperature (19, 25). Furthermore, in a study by Kononov et al. (11), lung tissue strips heated to 55°C to remove agarose showed no effect of heating on the mechanical properties of the strips.

A potential limitation of the study is associated with the ability to accurately measure stress. The force transducer used had a sensitivity of 0.5 mN/volt, which was sufficient to record forces generated by the elastin fibers in the thin sheet. However, stress was calculated as the measured force divided by the cross-sectional area of the ECM sheet. While the width could be measured accurately, accurate assessment of the samples' thickness is always difficult. To determine the thickness, we measured the half-width of fluorescent emission spectrum at several places and found significant intrasample variability of the thickness comparable to intersample variability of the mean thickness, especially after digestion. Thus, we used a single value of the average thickness in the stress calculations. This procedure certainly resulted in some intersample variability of the mechanical parameters. Additionally, the thickness of the sample must have decreased with increasing macroscopic strain. Although some variability may have arisen due to this effect, we do not expect that it could have created a significant difference in the overall evaluation of mechanical data.

During digestion under mechanical forces, we used 40% static or peak strain. To relate this value to the strains on individual fibers in vivo, we note that the lung at functional residual capacity (FRC) is in a prestressed state. Assuming that uniaxial strain varies as the cube root of lung volume, and that 40% of the total lung volume at FRC is tissue (24), we estimate that breathing, including sighs, corresponds approximately to uniaxial strains between 25 and 60%. These numbers are close to those (20 and 67%, respectively) estimated by Sata et al. (21).

The spring network model provides a useful framework for analyzing and interpreting the changes in the mechanical properties of the ECM sheets. The model was a two-dimensional hexagonal network of nonlinear springs combined with linear torsional springs. Because the length-to-thickness ratio of the ECM sheets was over 200 (5 mm over 20 microns), the two-dimensional nature of the model should not be a limitation. The hexagonal structure, however, is certainly not correct. Indeed, the confocal images showed a complex organization of the fiber structure. Nevertheless, the main characteristics of the macroscopic stress-strain properties of the network model does not depend significantly upon microscopic structure, since similar stress-strain curves can be obtained using prestressed rectangular (27) or triangular networks (15). The triangular network is stable and does not need torsional springs. Using a triangular network, Maksym et al. (15) demonstrated that at large strains, stress concentration develops in the network along preferred pathways, which may have implications to the breakdown of the network. While the hexagonal network is unstable, the implementation of the torsional spring in Eq. 2 stabilizes the network with similar force transmission pathways (3) as in the triangular network (15). The constitutive law of single elastin fibers is likely more complicated than the second order force-length relation used in the model. However, the incremental modulus characterizes the slope of the stress-strain curve, and hence the specific form of Eq. 1 is probably not a limitation either. The real ECM is also viscoelastic, and it is possible that the viscoelastic properties of the ECM do influence the rate of decline of Y during digestion or the time constant during cyclic stretching (see below). Viscoelasticty was not included in the model, and hence the model analysis was limited to the case of static stretching. Finally, it should be noted that the time scale of breakdown in the model is arbitrary, and hence only the exponential nature of the decrease of Y and the time constants relative to each other can be compared with the experimental data.

The notion that mechanical forces play a substantial role in the progression of lung disease was first set forth in a study by West in which he argued that the uneven distribution of emphysematous areas in the lung was strongly correlated to the distribution of mechanical forces on the lungs (32). Furthermore, it was observed that patients who undergo lung volume reduction surgery (LVRS) exhibit lung function degradation over time, more so than in the preoperative state (7). In LVRS, a diseased portion of the lung is removed, and the remaining lung expands to fill the void in the thoracic cavity. Degradation is believed to result from the increased mechanical forces required to fill the thoracic cavity. Similarly, increases in systolic blood pressure that lead to increased vessel wall stress play important roles in the progression of diabetes and hypertension (31). More recently, it has been shown that mechanical forces have a significant effect on cellular signaling and hence in remodeling processes of the ECM (9).

In this study, we have shown that mechanical forces directly modulate the effectiveness with which elastase can break down the ECM. In other words, mechanical forces can alter the activity of the enzyme. Specifically, we observed a significantly larger as well as more rapid decrease in the incremental modulus of the ECM sheet when digestion was carried out in the presence of static stretch or prestress and slow cyclic stretch (Fig. 3). In the relaxed state the fibrous network of the ECM is comprised of a number of individual elastin fibers in random orientation. The enzyme will attach at random to binding sites and cleave the bonds, resulting in a gradual decrease of the incremental modulus, Yn. In the presence of prestress, the fibers start to either reorient in the direction of strain or stretch. Whether the reorientation or the stretching occurs first depends on the relative stiffness of the fibers and the proteoglycan matrix (3). Nevertheless, the ECM is heterogeneous, and hence some elastin fibers should undergo stretch. The mechanism of stretching individual fibers is not well known. However, it is conceivable that fibrils and molecules will unfold, making more binding sites available for the enzyme. While this has not been shown for elastin, modules of fibronectin, an important ECM molecule with cell adhesion sites, unfold following the application of tensile force (14). Thus if there are sufficient enzymes in the solution, this should result in an increased rate of decline of the modulus. Another possibility is that the existing prestress on a binding site accelerates the process of cleaving by the elastase. An energy barrier exists between the initial bound state and the final ruptured state (4). This means that the distance between the two atoms has to be increased sufficiently along a reaction coordinate so that the energy barrier is overcome. In the presence of an enzyme, the energy barrier is lowered and the reaction is accelerated (5, 18). If the bond is already stretched due to mechanical forces, the energy barrier is lower than without stretching, and it is likely that the activity of the enzyme is stronger than in the absence of mechanical forces. Hence, this mechanism should also accelerate the decrease of Yn in the ECM. Independent of which mechanism is the dominant, the probability to weaken and break a fiber should increase in the presence of prestress, which we incorporated in our network model using Eq. 3. This simple probabilistic approach was able to fully account for both the increased rate and the exponential nature of the decline of Yn (Fig. 7).

To demonstrate the direct interaction between enzyme activity and mechanical force, the following additional experiment was performed. The digestion protocol described in the Methods was repeated; however, at 4 min from the addition of elastase, the bath was replaced with 20 ml of fresh PBS, and 400 µl of 100x protease inhibitor (Halt Protease Inhibitor Cocktail, Pierce) was also added to stop all enzyme activity. Figure 9 compares digestion with cyclic stretch at 5 cycles/min and inhibited digestion with the same type of stretch. It can be seen that while both groups follow the same trend for the first 10 min, no further decrease in stiffness occurs after 10 min for the inhibition case, even in the continued presence of mechanical forces. The decrease of the stiffness from 4 to 10 min may be the time required for the enzyme to cease activity. Since the modulus decreases only when digestion and mechanical force act simultaneously, the enhanced destruction of the ECM sheet compared with the pure digestion results from the interaction between enzymes and mechanical forces. Since most tissues in the body are always under a prestress, the interaction between mechanical forces and enzymatic digestion should lead to increased local breakdown and subsequent heterogeneity (see below). This in turn should develop further preferential force transmission pathways, which will be at an increased risk of enzymatic cleavage, forming a feedback loop and a subsequent cycle of destruction.


Figure 9
View larger version (9K):
[in this window]
[in a new window]

  		Fig. 9. Comparison Yn for the inhibition protocol.

 
The exponential decrease of Yn with time is a consequence of the random activity of the enzymes. Let us consider for simplicity a network of uniform springs with spring constant k arranged in parallel. The total stiffness of the system is E = Nk, where N is the number of parallel arrays. In the presence of enzymes, each spring is exposed to random elimination with a constant probability P0. At each time step during the simulated digestion, the fraction of springs eliminated is P0. Hence the number of springs is reduced by a factor of (1 – P0). After t time steps, the surviving number of springs is (1 – P0)tN, and the corresponding stiffness is given by Y(t) = (1 – P0)tNk. This expression can be written in the following form:

Formula 4(4)
where Y0 = Nk is the stiffness of the system at t = 0, and the positive time constant is given by {tau} = –1/ln(1 – P0) time steps. In our simulations P0 was 0.05, and hence {tau} = 1/P0 with an error of 2% only. Thus it can be seen that in this simple system, the decay of the stiffness is exponential, and the time constant is very closely the reciprocal of the probability P0 of cutting springs, which in turn is related to the specific properties of the enzyme and the fibers. During static stretch, P0 is replaced by P as defined in Eq. 3, which demonstrates that the exponential nature of Y(t) is retained, but the time constant becomes a hyperbolically decreasing function of the strain in this simple system. The ECM has a complex organization, yet as Fig. 3A demonstrates, the decrease of Yn remains exponential. Our network simulations suggest that even in the presence of heterogeneity with nonlinear and torsional springs, the decrease of stiffness is exponential. Furthermore, it appears that this exponential breakdown is maintained even in the presence of static and dynamic loading of the ECM in agreement with the above arguments. In addition, since Y(t) remained exponential even when Eq. 3 was replaced with a force-based condition (Fig. 7), our experimental data and simulations are unable to distinguish between these two possibilities.

When stretched, the fibers are loaded with a given force. As an elastin fiber breaks, its surrounding fibers will take on an added load that was previously carried by that fiber. As a result, these surrounding fibers become more prone to deteriorate and rupture. The deterioration of fibers is not evenly distributed throughout the sheet since the destruction of a fiber becomes dependent on nearby fibers. This should result in localized areas of degradation and an increased heterogeneity of strain among the fibers of the sheet. To test this, we took light microscopic images of two samples at the conclusion of pure digestion (Fig. 10A) and digestion with static stretch (Fig. 10B). These images are shown in high contrast to obtain a better comparison with the corresponding network simulations in Figs. 10B and 10D, respectively. Note the similarities between experimental and simulated networks. In the case of pure digestion, a homogeneous deterioration of the network of elastin (Fig. 10A) and springs (Fig. 10C) can be seen. In contrast, Figs. 10B and 10D show large heterogeneous destruction through the elastin and spring networks, respectively.


Figure 10
View larger version (60K):
[in this window]
[in a new window]

  		Fig. 10. Comparison of final simulated network model landscapes and images taken during the experimental protocol. A and B depict experimental images of digestion and digestion with static stretch, while C and D are the corresponding simulated zoomed images from the spring network model shown in Fig. 8. Macroscopic strain is in the vertical direction.

 
Unlike the other mechanically loaded groups, the application of cyclic stretch at a rate of 20 cycles/min during digestion did not exhibit any significant difference in the decrease in mechanical properties compared with digestion alone. In fact, the time constant during pure digestion was smaller (18.4 min) than during cyclic stretch at 20 cycles/min (32.4 min), even though this difference did not reach the statistically significant level. The data also suggest that enzymatic activity during cyclic stretch also depends on the frequency. Cyclic stretch at 5 cycles/min during digestion caused a more rapid and greater decrease in Yn than cyclic stretch at 20 cycles/min. It is likely that the amount of time spent at large strains during one given cycle plays a role in the configuration of the fibers. The fibers that comprise the ECM are viscoelastic (6). Consequently, under a macroscopic strain these fibers will take time to unfold, align, and relax. If the rate of strain is high, as in the 20 cycles/min cyclic stretch condition, fibers may not have enough time to reorient and unfold, which may hinder the enzymes from reaching binding sites that are otherwise hidden in the absence of stretch.

Additionally, enzymes can switch between extended and contracted conformations during their function. When an enzyme is adhered to a fiber, and the rate-limiting step of the enzyme is the transition from the extended to the contracted form, a stretching force on the fiber will reduce enzyme activity (29). Alternatively, enzyme activation can also occur if the stretching was sufficient to displace an inhibiting domain from the enzyme's active site. It is thus likely that depending on the specific application, mechanical forces can be a limiting or an enhancing factor. For example, the rate at which restriction enzymes are able to cleave a single DNA molecule depends on the tension along the molecule and the type of enzyme; the induced-fit rate of EcoRV is significantly reduced by mechanical force while that of BamHI is insensitive to force (29). Thus, this may open the possibility of modifying protein and enzyme activities through the application of dynamic force patterns.

Comparison of pure digestion with digestion during stretching indicates that the failure stress was reduced significantly in the presence of mechanical force. Since our data suggest that digestion in the presence of static stretching causes more severe deterioration of the ECM sheet, it is expected that at the end of the digestion, the failure stress is also reduced. However, Fig. 6A shows that there was no difference in failure stress among the cyclic load groups and the pure digestion group. This is somewhat unexpected given that the 20 cycles/min loading resulted in a slower rate of decrease of Yn than the pure digestion group. The reason is likely that Yn was evaluated at small strains, whereas failure occurs at much higher strains. It is interesting that the failure strain was the same across all conditions. The degree to which elastin fibers are degraded have no effect on the failure strain of the ECM sheet. This may imply that individual elastin fibers have an inherent critical strain, independent of other factors, or perhaps proteoglycans that link elastin fibers mechanically also influence the failure strain.

In conclusion, we have characterized the changes in the mechanical properties of ECM sheets as a result of the interaction between mechanical forces and digestion by elastase. We found that while the application of static or slowly varying mechanical forces accelerates the digestion-induced breakdown of ECM, a higher cycling frequency can have a protective role. Thus, the dynamic pattern of stretching is a critical parameter capable of modifying enzyme activity. This raises the possibility of actively controlling enzyme function that is essential in many cellular and extracellular processes by simple mechanical means. Our results may also have clinical implications. During the progression of emphysema, lung volume steadily increases. As a consequence, expanded regions of the lung with significant trapped air that are not participating in breathing-induced cyclic stretching may be at increased risk of progressive deterioration. Furthermore, during mechanical ventilation of COPD patients, a higher rate but smaller amplitude at the expense of a slightly higher end-tidal CO2 concentration might have a positive effect on protecting the lung ECM. This would be in accord with the current notion of permissive hypercapnia in ventilating patients with acute lung injury. Nevertheless, our results must be confirmed in more physiological experiments involving actual lung tissue.


   GRANTS
TOP
 ABSTRACT
 METHODS
 RESULTS
 DISCUSSION
 GRANTS
REFERENCES
 
This study was partially funded by National Heart, Lung, and Blood Institute Grant HL-059215.


   FOOTNOTES
 

Address for reprint requests and other correspondence: Bela Suki, Boston Univ., Dept. of Biomedical Engineering, 44 Cummington St., Boston, MA 02215 (e-mail: bsuki{at}bu.edu)

The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.


   REFERENCES
TOP
 ABSTRACT
 METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 

  1. Arbabi S, Sahimi M. Elastic properties of three-dimensional percolation networks with stretching and bond-bending forces. Phys Rev B Condens Matter 38: 7173–7176, 1988.[Medline]
  2. Black LD, Brewer KK, Morris SM, Schreiber BM, Toselli P, Nugent MA, Suki B, Stone PJ. Effects of elastase on the mechanical and failure properties of engineered elastin-rich matrices. J Appl Physiol 98: 1434–1441, 2005.[Abstract/Free Full Text]
  3. Cavalcante FS, Ito S, Brewer K, Sakai H, Alencar AM, Almeida MP, Andrade JS Jr, Majumdar A, Ingenito EP, Suki B. Mechanical interactions between collagen and proteoglycans: implications for the stability of lung tissue. J Appl Physiol 98: 672–679, 2005.[Abstract/Free Full Text]
  4. Evans E. Probing the relation between force—lifetime—and chemistry in single molecular bonds. Annu Rev Biophys Biomol Struct 30: 105–128, 2001.[CrossRef][Web of Science][Medline]
  5. Florin EL, Moy VT, Gaub HE. Adhesion forces between individual ligand-receptor pairs. Science 264: 415–417, 1994.[Abstract/Free Full Text]
  6. Fung Y. Biomechanics: Mechanical Properties of Living Tissues. New York: Springer-Verlag, 1981.
  7. Gelb AF, McKenna RJ Jr, Brenner M, Schein MJ, Zamel N, Fischel R. Lung function 4 years after lung volume reduction surgery for emphysema. Chest 116: 1608–1615, 1999.[CrossRef][Web of Science][Medline]
  8. Hukins DWL. Connective tissue matrix. London: Macmillan, 1984.
  9. Ingber DE. Cellular mechanotransduction: putting all the pieces together again. FASEB J 20: 811–827, 2006.[Abstract/Free Full Text]
  10. Khan S, Sheetz MP. Force effects on biochemical kinetics. Annu Rev Biochem 66: 785–805, 1997.[CrossRef][Web of Science][Medline]
  11. Kononov S, Brewer K, Sakai H, Cavalcante FS, Sabayanagam CR, Ingenito EP, Suki B. Roles of mechanical forces and collagen failure in the development of elastase-induced emphysema. Am J Respir Crit Care Med 164: 1920–1926, 2001.[Abstract/Free Full Text]
  12. Kuhn C, Yu SY, Chraplyvy M, Linder HE, Senior RM. The induction of emphysema with elastase. II. Changes in connective tissue. Lab Invest 34: 372–380, 1976.[Web of Science][Medline]
  13. Lansing AI, Rosenthal TB, Alex M, Dempsey EW. The structure and chemical characterization of elastic fibers as revealed by elastase and by electron microscopy. Anat Rec 114: 555–575, 1952.[CrossRef][Medline]
  14. Li L, Huang HH, Badilla CL, Fernandez JM. Mechanical unfolding intermediates observed by single-molecule force spectroscopy in a fibronectin type III module. J Mol Biol 345: 817–826, 2005.[CrossRef][Web of Science][Medline]
  15. Maksym GN, Fredberg JJ, Bates JH. Force heterogeneity in a two-dimensional network model of lung tissue elasticity. J Appl Physiol 85: 1223–1229, 1998.[Abstract/Free Full Text]
  16. Marque V, Kieffer P, Gayraud B, Lartaud-Idjouadiene I, Ramirez F, Atkinson J. Aortic wall mechanics and composition in a transgenic mouse model of Marfan syndrome. Arterioscler Thromb Vasc Biol 21: 1184–1189, 2001.[Abstract/Free Full Text]
  17. Moretto A, Dallaire M, Romero P, Ludwig M. Effect of elastase on oscillation mechanics of lung parenchymal strips. J Appl Physiol 77: 1623–1629, 1994.[Abstract/Free Full Text]
  18. Moy VT, Florin EL, Gaub HE. Intermolecular forces and energies between ligands and receptors. Science 266: 257–259, 1994.[Abstract/Free Full Text]
  19. Rasmussen BL, Bruenger E, Sandberg LB. A new method for purification of mature elastin. Anal Biochem 64: 255–259, 1975.[CrossRef][Web of Science][Medline]
  20. Rosenbloom J, Abrams WR, Mecham R. Extracellular matrix 4: the elastic fiber. FASEB J 7: 1208–1218, 1993.[Abstract]
  21. Sata M, Takahashi K, Sato S, Tomoike H. Structural and functional characteristics of peripheral pulmonary parenchyma in golden hamsters. J Appl Physiol 78: 239–246, 1995.[Abstract/Free Full Text]
  22. Snider GL, Lucey EC, Stone PJ. Animal models of emphysema. Am Rev Respir Dis 133: 149–169, 1986.[Web of Science][Medline]
  23. Stamenovic D, Suki B, Fabry B, Wang N, Fredberg JJ. Rheology of airway smooth muscle cells is associated with cytoskeletal contractile stress. J Appl Physiol 96: 1600–1605, 2004.[Abstract/Free Full Text]
  24. Stamenovic D, Yager D. Elastic properties of air- and liquid-filled lung parenchyma. J Appl Physiol 65: 2565–2570, 1988.[Abstract/Free Full Text]
  25. Starcher BC, Galione MJ. Purification and comparison of elastins from different animal species. Anal Biochem 74: 441–447, 1976.[CrossRef][Web of Science][Medline]
  26. Stone PJ, Morris SM, Griffin S, Mithieux S, Weiss AS. Building elastin. Incorporation of recombinant human tropoelastin into extracellular matrices using nonelastogenic rat-1 fibroblasts as a source for lysyl oxidase. Am J Respir Cell Mol Biol 24: 733–739, 2001.[Abstract/Free Full Text]
  27. Suki B, Lutchen KR, Ingenito EP. On the progressive nature of emphysema: roles of proteases, inflammation, and mechanical forces. Am J Respir Crit Care Med 168: 516–521, 2003.[Free Full Text]
  28. Urry DW, Hugel T, Seitz M, Gaub HE, Sheiba L, Dea J, Xu J, Parker T. Elastin: a representative ideal protein elastomer. Philos Trans R Soc Lond B Biol Sci 357: 169–184, 2002.[Abstract/Free Full Text]
  29. van den Broek B, Noom MC, Wuite GJ. DNA-tension dependence of restriction enzyme activity reveals mechanochemical properties of the reaction pathway. Nucleic Acids Res 33: 2676–2684, 2005.[Abstract/Free Full Text]
  30. Vorp DA, Schiro BJ, Ehrlich MP, Juvonen TS, Ergin MA, Griffith BP. Effect of aneurysm on the tensile strength and biomechanical behavior of the ascending thoracic aorta. Ann Thorac Surg 75: 1210–1214, 2003.[Abstract/Free Full Text]
  31. Weir MR. Diabetes and hypertension: blood pressure control and consequences. Am J Hypertens 12: 170S–178S, 1999.[CrossRef][Web of Science][Medline]
  32. West J. Distribution of mechanical stress in the lung, a possible factor in localisation of pulmonary disease. Lancet 1: 839–841, 1971.[CrossRef][Web of Science][Medline]



This article has been cited by other articles:


Home page
 Proc. Natl. Acad. Sci. USAHome page
M. C. Ritter, R. Jesudason, A. Majumdar, D. Stamenovic, J. A. Buczek-Thomas, P. J. Stone, M. A. Nugent, and B. Suki
A zipper network model of the failure mechanics of extracellular matrices
PNAS, January 27, 2009; 106(4): 1081 - 1086.
[Abstract] [Full Text] [PDF]

Home page
 Eur Respir JHome page
X. Trepat and R. Farre
Alveolar permeability and stretch: too far, too fast
Eur. Respir. J., October 1, 2008; 32(4): 826 - 828.
[Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow All Versions of this Article:
103/3/803    most recent
00057.2007v1
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Jesudason, R.
Right arrow Articles by Suki, B.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Jesudason, R.
Right arrow Articles by Suki, B.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Visit Other APS Journals Online
Copyright © 2007 by the American Physiological Society.