Endotoxemia is associated with changed pulmonary vascular function with respect to vasoreactivity, endothelial permeability, and activation of inducible nitric oxide synthase II (NOSII). However, whether altered passive arterial wall mechanics contribute to this endotoxin-induced pulmonary vascular dysfunction is still unknown. Therefore, we investigated whether endotoxin affects the passive arterial mechanics and compliance of isolated rat pulmonary arteries. Pulmonary arteries of pentobarbital-anesthetized Wistar rats (n = 55) were isolated and exposed to Escherichia coli endotoxin (50 μg/ml) for 20 h. Endotoxin increased pulmonary artery diameter and compliance (transmural pressure = 13 mmHg) in an endothelium-, Ca2+-, or NOSII-induced NO release-independent manner. Interestingly, the endotoxin-induced alterations in the passive arterial mechanics were accompanied by disassembly of the smooth muscle cell (SMC) F-actin cytoskeleton. Disassembly of F-actin by incubation of control arteries with the cytoskeleton-disrupting agent cytochalasin B or the Rho-kinase inhibitor Y-27632 induced a similar increase in passive arterial diameter and compliance. In contrast, RhoA activation by lysophosphatidic acid prevented the endotoxin-induced alterations in the pulmonary SMC F-actin cytoskeleton and passive mechanics. In conclusion, these findings indicate that disassembly of the SMC F-actin cytoskeleton and RhoA/Rho-kinase signaling act as mediators of endotoxin-induced changes in the pulmonary arterial mechanics. They imply the involvement of F-actin rearrangement and RhoA/Rho-kinase signaling in endotoxemia-induced vascular lung injury.
- lung injury
- passive mechanical arterial properties
- lysophosphatidic acid
endotoxemia-associated lung injury is characterized by an impaired pulmonary smooth muscle cell (SMC) contractile response, an increased endothelial permeability, and an increased availability of nitric oxide (NO) due to the induction of the enzyme NO synthase II (NOSII) (26, 28, 31). Besides inducing changes in the active vascular response, endotoxemia might alter the arterial wall mechanical properties, as similarly observed in arteriosclerosis and hypertension (3, 6, 11, 14, 15). Cox (12) reported that hypercholesterolemia was associated with a decreased compliance of iliac arteries in dogs. These changes in the passive arterial wall mechanics are known to be associated with alterations in the composition of the vessel wall, which include changes in the elastin and collagen content and other extracellular matrix components, remodeling of the SMC, and rearrangement of the intracellular F-actin fibers of the cytoskeleton (13, 15, 32).
The cytoskeleton is an integrated, dynamically arranged network of actin fibers, microtubules, and intermediate filaments and is involved in various cellular processes including cell motility, intracellular transport, and SMC contraction. F-actin cytoskeleton rearrangement involves actin repolymerization and provides the cell control of cellular functions such as cell strength and contractility (8, 21, 25, 30, 37). Until now, endotoxemia-induced changes in cellular signaling with respect to cytoskeletal rearrangement have been shown only for human fibroblasts and endothelial cells, murine macrophages, and rat mesangial or hepatic cells but not for pulmonary SMC (5, 9, 17, 22, 34, 38).
The small GTPase RhoA plays a regulatory role in F-actin fiber rearrangement, including the formation of stress fibers and focal adhesion formation, and is involved in the regulation of SMC contraction (2, 20, 29, 30, 35, 36, 43). Rho-kinase is an important mediator of these RhoA-dependent effects on the vasculature. We previously confirmed the involvement of RhoA/Rho-kinase signaling in the reactivity of SMC in pulmonary arteries, as described for other vessels before, by inhibiting the contractile capacity of pulmonary arteries through Rho-kinase inhibition (4, 19, 41, 42).
Because endotoxemia-related changes in the pulmonary passive arterial wall mechanics have never been demonstrated, the first goal of this study was to evaluate the effects of 20 h of in vitro exposure to endotoxin on the passive diameter and compliance of isolated pulmonary arteries. Subsequently, we studied the role of the endothelium, Ca2+ influx, and NOSII activity, which are the common mediators of endotoxin-induced pulmonary vascular injury, in these passive mechanics and determined whether endotoxin altered the mechanical properties in association with changes in the SMC F-actin fiber structure. Finally, we investigated the involvement of the RhoA/Rho-kinase signaling pathway in the endotoxin-induced changes in the pulmonary SMC characteristics.
MATERIALS AND METHODS
Experimental Animals and Tissue Preparation
The Institutional Animal Care and Use Committee of the VU University Medical Center (Amsterdam, the Netherlands) approved the experimental procedures. Wistar rats (males, 308 ± 5 g, n = 55; Harlan, Zeist, the Netherlands) were anesthetized with pentobarbital sodium (60 mg/kg ip, Nembutal; Sanofi Sante). Lungs were removed and pinned in a dissecting dish containing sterile MOPS buffer (4°C, pH 7.4), and a side branch of the main pulmonary artery was isolated (4).
The dissected arteries were cut into two or four segments (axial length 1–2 mm) that were immediately used or incubated in cell culture medium according to the protocol described by Piepot et al. (33). One or two arterial segments were incubated in DMEM containing 50 μg/ml of Escherichia coli lipopolysaccharide (endotoxin, serotype 127:B8; Difco Laboratories) for 20 h at 37°C in 95% humidified air and 5% CO2. The other arterial segments were incubated in control medium and served as time-matched controls essentially as described before (33). One group of control and endotoxin-exposed pulmonary arteries was studied without the application of interventions (n = 7). One group of arterial segments was endothelium denuded by a bolus of 1 ml of air before the start of the incubation period (n = 5). Endothelial damage was confirmed by the absence of the acetylcholine-induced vasodilation. Other pairs of arterial segments were co-incubated with the specific NOSII inhibitor aminoguanidine (10−4 M, Sigma-Aldrich; n = 5), the RhoA activator lysophosphatidic acid (LPA, 10−5 M; Sigma-Aldrich; n = 6), or the Rho-kinase inhibitor Y-27632 (3·10−7 M, Tocris; n = 5). After the incubation period, a paired passive force-length relationship was obtained for both pulmonary arterial segments, as described below. Subsequently, arteries were checked for their contractile (10−6 M thromboxane analog U-46619) and their maximal dilatory (10−4 M papaverin) capacities.
Measurement of the Passive Force-Length Relationship
Isometric force measurements were performed in a double vessel wire myograph (27). Each vessel segment was mounted on two tungsten wires (50 μm diameter) in the organ chamber filled with MOPS buffer. MOPS buffer consisted of (in mM) 145 NaCl, 5 KCl, 2 CaCl2, 1 MgSO4, 1 NaH2PO4, 5 dextrose, 2 pyruvate (Sigma-Aldrich), 0.02 EDTA, and 3 3-(N-morpholino)propanesulfonic acid equilibrated with 95% air and 5% CO2 (37°C, pH 7.4) (all components, except for pyruvate, were obtained from Merck). One wire was connected to a micromanipulator, whereas the other was connected to a force transducer (Kistler Morse no. 46-1003-01). After stabilization (30 min), the pairs of pulmonary arterial segments were stretched in steps to a length that induced a passive force of ∼170 mg/mm (equivalent to a transmural pressure of 13 mmHg).
Absence of Ca2+.
We studied the role of Ca2+ in the passive force-length relationship by investigating the force-length relationship of control and endotoxin-exposed arteries in buffer solution that was free of Ca2+ and was supplemented with 10−2 M of caffeine (Sigma-Aldrich) to induce intracellular Ca2+ depletion (n = 5).
Artificial disassembly of the SMC actin structure.
Cytochalasin B (Sigma-Aldrich) was used to disrupt the F-actin cytoskeleton (n = 4). Freshly isolated pulmonary arteries were endothelium denuded and kept for 40 min in control buffer or buffer containing 5·10−5 M cytochalasin B. Subsequently, a force-length relationship was determined for both arteries.
SMC F-Actin Structure Imaging
To visualize the F-actin structure of the pulmonary SMC layer specifically, we removed the endothelium by mechanical rubbing of the artery lumen. Subsequently, the pulmonary artery segments were perfusion fixed by paraformaldehyde, and the SMC were permeabilized (0.05% Triton X-100, 2 min) and stained (260 nM rhodamine phalloidin, 30 min; Molecular Probes Europe). Arteries were cut open longitudinally and placed on a glass cover slip with the luminal side facing upwards. F-actin fibers were visualized by fluorescence imaging (530 nm; Zeiss, 473028).
After determination of the passive force-length relationship, we calculated the equivalent pressure-diameter (P-D) relationship according to a simplified LaPlace equation (27). All figures in results represent the pulmonary arterial P-D relationship. The diameter corresponding with the applied stretch in the pulmonary arteries was obtained via calculation of the internal arterial circumference (27). The P-D relationship can be described by the following equation (18). where P is the pressure and D is the vessel diameter. From the data points we determined the diameter that corresponded with the first measured pressure that was just >0 and designated that diameter as D0. The curve parameters a and b describe the vessel wall characteristics and are defined as elastic parameters (18). We fixed the size of parameter a at a value of unity based on the following arguments. In general, the estimated values for the parameters a and b showed a negative correlation, implying that a and b are confounders. Also, the setting of parameter a to 1 did not reduce the explained variance, according to the Akaike Information Criterion (data not shown) (1).
For all experiments, a diameter was determined from the curve fit at a calculated pressure of 13 mmHg (D13). This diameter was used as a measure of the diameter increase at a physiological pressure. Furthermore, the compliance at 13 mmHg (C13) was defined as the slope of the P-D relationship (ΔD/ΔP).
The fit parameters b, D0, D13, and C13 are given as means ± SE and were analyzed by paired t-tests or t-tests for independent samples. Differences were considered statistically significant as P < 0.05.
Endotoxin Alters Passive Pulmonary Arterial Mechanics Independently of the Endothelium, Ca2+, and NOSII Activation
Endotoxin induced a rightward shift of the P-D relationship in pulmonary arteries (Fig. 1A), implying an increased diameter at a pressure >0. Figure 1Aa shows the data of an experiment and the applied model fit, whereas Fig. 1Ab shows the model using the average parameter values as described in Table 1 (n = 7). By example, the dotted lines represent the means ± SE of the fit parameter b for control and endotoxin-exposed arteries.
Table 1 lists the mean values of the curve parameters b, D0, D13, and C13 for control and endotoxin-exposed pulmonary arteries under different conditions. Exposure to endotoxin decreased b and increased D13 and C13 compared with control arteries (n = 7). Endothelium removal (n = 5), removal of Ca2+ during the force measurements (n = 5), or NOSII inhibition with aminoguanidine (10−4 M, n = 5) did not prevent endotoxin-induced alterations in b, D13, and C13. Thus endotoxin induces an increase in diameter and compliance at physiological pressure, independently of the endothelium, Ca2+, and NOSII activation.
Endotoxin-Induced Change in the P-D Relationship Is Accompanied by Alterations in the SMC F-Actin Cytoskeleton
To investigate whether the cytoskeleton was involved in the altered passive mechanics of endotoxin-treated arteries, we stained the SMC layer of control and endotoxin-exposed arteries for F-actin with rhodamine-phalloidin. Figure 1B shows that the endotoxin-induced increase in compliance was accompanied by rearrangement of the SMC F-actin fiber structure. Control arteries demonstrated a distribution of F-actin fibers in the axial direction of the SMC, whereas endotoxin-exposed arteries showed a chaotic F-actin distribution pattern in the SMC. These findings were consistent for different pairs of control and endotoxin-exposed pulmonary arteries (n = 5).
To verify whether alterations in the SMC F-actin cytoskeleton induce changes in the P-D relationship as well, we investigated the effects of disassembly of the SMC F-actin structure by cytochalasin B on the passive arterial mechanics. Disassembly of the F-actin structure induced a rightward shift of the P-D relationship (Fig. 2A, n = 4). Figure 2Aa presents a representative experiment and its model fit. The parameter values from Table 1 were used to show the average model fit in Fig. 2Ab. Cytochalasin B and endotoxin induced similar alterations in D13 and C13 (not significant). Figure 2B shows an example of the cytochalasin B-induced changes in the SMC F-actin fiber structure in a control and cytochalasin B-treated artery (n = 4). Cytochalasin B induced a similar chaotic distribution of actin fibers as in the endotoxin-exposed pulmonary arteries (see Fig. 1B). It thus appears that endotoxin and cytochalasin B have similar effects on both the passive P-D relationship and actin cytoskeleton.
Endotoxin-Induced Changes in the P-D Relationship and SMC F-Actin Cytoskeleton Are Mediated by RhoA/Rho-Kinase Signaling
Rho-kinase is involved in the regulation of the cytoskeleton structure and the formation of F-actin stress fibers. We therefore studied the effects of Rho-kinase inhibition by Y-27632 (n = 5) on the P-D relationship in control arteries. Figure 3A shows that Y-27632 increased the artery C13 and D13 and mimicked the endotoxin-induced changes in the passive P-D relationship (Fig. 3Aa, mean parameter values listed in Table 2). Although Y-27632 induced a similar increase in compliance compared with endotoxin, D13 differed between Y-27632- and endotoxin-exposed arteries (P = 0.021). Figure 3B shows the SMC F-actin fiber structure of control pulmonary arteries and after incubation with endotoxin or Y-27632 (n = 4). Incubation of the SMC with Y-27632 resulted in a similar disassembly of the F-actin fiber structure compared with the endotoxin-induced disarray.
We subsequently investigated the effects of co-incubation of control and endotoxin-exposed pulmonary arteries with the RhoA activator lysophosphatidic acid (LPA) on the P-D relationship (Fig. 3Ab). RhoA activation specifically prevented the endotoxin-induced shift of the P-D relationship but did not affect the passive mechanics in control pulmonary arteries (n = 6, mean parameter values listed in Table 2). D13 and C13 of LPA-treated endotoxin-exposed arteries did not differ from control arteries. Figure 3B shows the effects of LPA treatment on the SMC F-actin fiber structure in endotoxin-exposed pulmonary arteries (n = 5). An endotoxin-induced rearrangement of the actin fiber structure was largely prevented by co-incubation with LPA. In contrast, LPA alone did not affect the actin cytoskeleton of control pulmonary arteries (data not shown). Thus RhoA/Rho-kinase signaling is involved in endotoxin-induced alterations of the passive SMC mechanics and F-actin cytoskeleton.
The present study provides evidence that endotoxin increases the passive pulmonary arterial diameter and compliance, independently of the endothelium, Ca2+, or NOSII. In contrast, altered pulmonary mechanics are in close association with disassembly of the pulmonary arterial smooth muscle F-actin cytoskeleton. Moreover, RhoA activation and Rho-kinase inhibition modulate the endotoxin-induced alterations in passive diameter, compliance, and cytoskeleton.
The endotoxin-induced increase of the passive diameter and compliance, as determined by a rightward shift of the P-D relationship, was associated with rearrangement of the smooth muscle cell F-actin cytoskeleton. This was verified by disassembly of the cytoskeleton by cytochalasin B and Rho-kinase inhibition, which as well were associated with a rightward shift of the P-D relationship. Finally, we found that activation of RhoA prevented the endotoxin-induced increases in diameter and compliance and partially inhibited smooth muscle F-actin cytoskeleton disassembly.
The effects of endotoxin on the passive mechanical properties of the pulmonary vascular wall were studied by constructing passive force-length relationships, which were transformed to P-D relationships. From the P-D relationships, we determined the diameter and compliance at the physiological pressure of 13 mmHg (18).
Our data show that we can exclude the involvement of the endothelium, Ca2+, and NO in the endotoxin-induced changes. Ca2+ might be involved in the passive arterial mechanics, since the presence of a constrictor can decrease the vessel diameter at zero pressure (12). Because endotoxin induces the release of vasoconstrictors, this might also affect the P-D relationship (16, 26, 31, 40). However, our data show that Ca2+ is not involved in the endotoxin-associated rightward shift of the P-D relationship.
Another candidate of the endotoxin-induced shift of the P-D relationship may be an activation of NOSII, which is associated with an increased vascular NO availability (7, 23, 28). NO might be involved in the rearrangement of the actin cytoskeleton in hepatic and endothelial cells, thereby changing the passive arterial wall mechanics as well (10, 22). However, we found no effects of NOSII inhibition on the P-D relationships. The finding that removal of the endothelium had no effect also supported the lack of NOS induction in the endothelial cells. Finally, an increased arterial water content due to edema would induce a decreased arterial compliance (14), although we found an endotoxin-induced increased compliance in the pulmonary arteries. Thus our data suggest that the change in the passive P-D relationship is a result of changes in SMC characteristics.
To the best of our knowledge, we are the first to show that endotoxin induces disassembly of the actin fiber cytoskeleton of pulmonary vascular SMC. It has already been reported that endotoxin affects the cytoskeleton structure of other cells, such as endothelial cells, macrophages, mesangial and hepatic cells, and fibroblasts (5, 9, 17, 22, 34, 38). The association between the actin cytoskeleton and mechanical changes were qualitatively mimicked by artificial disassembly of the F-actin fibers by cytochalasin B. Although the relative increase in compliance was higher in cytochalasin B-treated arteries compared with endotoxin-exposed arteries, this might be mainly due to the difference in incubation protocol. Furthermore, it is unknown whether the extent of cytochalasin B-induced cytoskeletal rearrangement is identical to the F-actin disassembly induced by endotoxin. Other investigators have demonstrated that murine vascular SMC lacking the cytoskeleton component desmin exhibit a reduced passive force, suggesting an increased compliance (39). They also showed that the reduction in passive force was paralleled by a disturbed transmission of the signaling process via the cytoskeleton (39). Thus the data from the literature and our findings suggest a causal relationship between the actin cytoskeleton, cellular signaling, and passive mechanical properties.
In general, the relationship between the passive force and applied stretch reflects the characteristics of the arterial wall materials, such as elastin, collagen, and the cytoskeleton. Therefore, the endotoxin-induced shift of the P-D relationship and increased arterial compliance suggests a change in the arterial wall characteristics.
The RhoA/Rho-kinase signaling pathway regulates the rearrangement of the cytoskeleton (2, 20, 29, 35, 36, 43). From our data we conclude that Rho-kinase inhibition induces cytoskeletal rearrangement and associated changes in the passive arterial wall mechanics. This suggests that endotoxin acts similarly, i.e., endotoxin might suppress RhoA/Rho-kinase signaling. Furthermore, RhoA activation prevented passive arterial diameter and compliance changes at physiological pressure and partly inhibited F-actin cytoskeleton disassembly. RhoA activation did not affect the passive smooth muscle properties of control arteries. We hypothesize that the signal transduction of the passive force-length relationship of pulmonary arteries depends on an intact cytoskeleton, which can be modulated by RhoA/Rho-kinase signaling and endotoxin.
We showed that endotoxin-induced alterations in SMC can be altered by RhoA activation or Rho-kinase inhibition. Others showed that endotoxin alters the endothelial cytoskeleton in association with activation of RhoA. Our data suggest an endotoxin-mediated change in RhoA/Rho-kinase signaling, which implies a distinct interaction between endotoxin and RhoA/Rho-kinase signaling for different cell types (17). The endotoxin-induced alterations in the pulmonary passive arterial wall properties and cytoskeleton may play a role in the pathogenesis of sepsis-induced lung injury. It has been reported that lung injury is accompanied with a reduced compliance of the pulmonary vascular tree (24). However, since lung injury in patients is associated with vascular remodeling and pulmonary hypertension, this may account for the paradoxical increased stiffness of the pulmonary circulation in patients when compared with our findings in the in vitro situation. Furthermore, the incubation period of 20 h that we used is probably too short to observe the decreased compliance in patients.
In conclusion, endotoxin rearranges the pulmonary SMC F-actin cytoskeleton, leading to an increased diameter and compliance at physiological pressure. The endotoxin-induced changes in pulmonary SMC behavior are modulated by the RhoA/Rho-kinase signaling pathway. The cytoskeletal damage and altered passive function following endotoxin exposure might play a role in the pathophysiology of vascular lung injury during septic shock. Pharmacological interventions altering the RhoA/Rho-kinase signaling activity may play a future role in the treatment of sepsis-associated lung injury.
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.
- Copyright © 2004 the American Physiological Society