|Home | About | Journals | Submit | Contact Us | Français|
Vascular remodeling is essential to proper vessel function. Dramatic changes in mechanical environment, however, may initiate pathophysiological vascular remodeling processes that lead to vascular disease. Previous work by some of our group has demonstrated a dramatic rise in matrix metalloproteinase (MMP) expression shortly following an abrupt increase in carotid blood flow. We hypothesized that there would be a corresponding change in carotid mechanical properties. Unilateral carotid ligation surgery was performed to produce an abrupt, sustained increase in blood flow through the contralateral carotid artery of rats. The flow-augmented artery was harvested after sham surgery or 1, 2, or 6 days after flow augmentation. Vessel mechanical response in the circumferential direction was then evaluated through a series of pressure-diameter tests. Results show that the extent of circumferential stretch (normalized change in diameter) at in vivo pressure levels was significantly different (p<0.05) from normo-flow controls at 1 and 2 days following flow augmentation. Measurements at 1, 2, and 6 days were not significantly different from one another, but a trend in the data suggested that circumferential stretch was largest 1 day following surgery and subsequently decreased toward baseline values. Because previous work with this model indicated a similar temporal pattern for MMP-9 expression, an exploratory set of experiments was conducted where vessels were tested 1 day following surgery in animals treated with broad spectrum MMP inhibitors (either doxycycline or GM6001). Results showed a trend for the inhibitors to minimize changes in mechanical properties. Observations demonstrate that vessel mechanical properties change rapidly following flow augmentation and that alterations may be linked to expression of MMPs.
Vascular remodeling in response to changes in mechanical environment is a fundamental homeostatic process for proper blood vessel function. This remodeling process involves a balance of destructive and constructive processes to break down and rebuild the vessel wall to accommodate new loading conditions. For example, flow-induced outward vascular remodeling is an adaptive process through which blood vessels increase their lumen to normalize hemodynamic shear stresses that have increased with blood flow (Gibbons and Dzau 1994; Guyton and Hartley 1985; Kamiya and Togawa 1980; Kamiya et al 1988; Langille and O'Donnell 1986). The basic mechanisms of this process may play a role in pathophysiological remodeling of arteries observed in atherosclerosis, aortic aneurysms, intracranial aneurysms, brain arteriovenous malformations, and vascular sequelae of head trauma (Gibbons and Dzau 1994; Glagov et al 1987; Hashimoto et al 2001; Hashimoto et al 2006; Hoi et al 2008).
Tissue remodeling appears to be controlled by the orchestrated effects of a number of cytokines and proteinases, including matrix metalloproteinases (MMPs) (Abbruzzese et al 1998; Flamant et al 2007; Galis and Khatri 2002; Tronc et al 2000). In a previous study using rats, we showed that inflammatory cells, such as macrophages, play critical roles in adaptive outward remodeling of the common carotid artery in response to an increase in blood flow (Nuki et al 2009). In this model, the time course of up-regulation of MMPs coincided with that of macrophage infiltration. More importantly, inhibition of MMPs by SB-3CT reduced flow-induced outward remodeling; flow-induced outward remodeling was also reduced in MMP-9 knockout mice (Ota et al 2009). These findings indicate that MMPs, produced by macrophages, play an important role in flow-induced outward vascular remodeling.
Because blood vessels remodel to return cells to homeostatic levels of stress following changes in mechanical environment (Humphrey 2008), it seems logical that dynamic changes in mechanical properties would accompany remodeling. Such changes have not, however, received much attention in the literature, especially in response to increased flow. It is thus not known if such changes occur, how subtle they may be, or what time scale they may occur over in the previously referenced model of outward remodeling. Definition of the magnitude and timing of any alteration in mechanical properties will lend further insight into remodeling phenomena, including issues related to the development of vascular pathophysiology.
Given the role of MMPs in extracellular matrix protein degradation and their observed up-regulation following an abrupt increase in blood flow (Nuki et al 2009), we hypothesized that such a change in flow would also produce a change in carotid artery mechanical properties. To test this hypothesis, we used the described model of carotid ligation in the rat. In this model, ligation of the left common carotid artery augments blood flow in the right common carotid artery (Nuki et al 2009; Ota et al 2009). Results from pressure-diameter tests on the flow-augmented right common carotid show that the mechanical characteristics of the vessel change rapidly following an increase in blood flow and suggest that MMP expression may be at least partly responsible for this alteration.
As has been described previously (Nuki et al 2009), the left common carotid artery of adult male Sprague-Dawley rats (300–350g, 10–12 weeks old) was ligated and the incision closed, according to institutional guidelines. The right common carotid artery was harvested at 1, 2, or 6 days after flow-augmentation. Vessels were also obtained from animals not undergoing flow-augmentation surgery for control measurements. Because changes in the properties of carotid arteries from these animals appeared to be temporally correlated with MMP expression, as measured in prior experiments using the same ligation model (Nuki et al 2009), a follow-up set of experiments was designed to investigate this possible link. In this second set of experiments, animals received one of two broad spectrum MMP inhibitors, GM6001 or doxycycline. GM6001 (10mg/kg) was given intraperitoneally one day before the flow-augmentation and at every 24 hours following. Doxycycline (40 mg/kg/day) was given through drinking water over the same interval as previously described (Lee et al 2006; Pyo et al 2000). Vessels from animals receiving no inhibitor were also tested for comparison. All vessels in the second set of experiments were resected one day after flow-augmentation, the time point at which we observed maximum changes in mechanical properties. In both sets of experiments, four animals were generally used for each group, leading to four successful tests each at 0, 2, and 6 days (3 successful tests at 1 day) in the first set of experiments. For the second set of experiments, 7, 4, and 4 successful tests were run on untreated, doxycyline-treated, and GM6001-treated cases, respectively. Prior to resection, the in vivo length of the carotid was measured. Vessels typically retracted from 11–12 mm to an unloaded length of 8–9 mm upon resection.
To precisely characterize the mechanical response of blood vessels, we have developed an isolated vessel testing system that provides measurement of forces and deformations under a variety of axial and circumferential loading conditions. The apparatus and associated procedures have been described previously (Monson et al 2008). In brief, cross-sections (approximately 0.5 mm thick) were cut from the ends of each sample, and their images were captured to a computer for subsequent diameter and wall-thickness measurement; no repeatable differences were apparent between the two sections, so measurements were averaged for each specimen. The remaining arterial segment was then tied at each end to blunt, 25-gauge needles that had been grooved to minimize slippage during axial stretch. The needles were attached to acrylic blocks whose positions were controlled using a computer-controlled servo-hydraulic load frame [MTS, Model 858 Mini-Bionix] (Fig. 1). Passages in the blocks allowed fluid flow and pressure measurement. To further seal each specimen, a small amount of cyanoacrylate was applied between the vessel ends and the needles. During preparation and testing, specimens were perfused and bathed in physiological saline (Dulbecco’s Modified Eagle’s Medium; DMEM) bath. One of the specimen mounting blocks was connected to a saline delivery system, a linear actuator [Ultramotion, Digit] that drove the position of a syringe. A 200 kPa pressure transducer [Honeywell, MicroSwitch 26PCDFM6G] was mounted on the other block. Applied loads were measured using an 11 N load cell (Transducer Techniques, MDB 2.5), and displacement was monitored through the linear variable displacement transducer (LVDT) built into the load frame. Vessel outer diameter was determined through measurements of video recorded (3 frames/second) using a digital camera (Pixelink, PL-A641), with an attached high-powered zoom lens (FOV: 1.4–17.6 mm). All non-video data were recorded using a standard data acquisition (DAQ) card.
After attachment to the apparatus, each specimen was preconditioned through a series of 3 to 4 tests in which the vessel was stretched axially to target values just over in vivo levels to avoid conditions of buckling, and its length was held constant for approximately two minutes while pressure in the vessel was cycled 5 times between 6.7 and 20.0 kPa (50 and 150 mmHg). Following these initial procedures, another test was conducted in which specimen length was reduced sufficiently to induce buckling and then increased to the maximum value utilized during preconditioning, all while a valve connected to the upper block was open so as to maintain zero pressure; this test provided for identification of the unloaded reference configuration. Finally, the vessel was subjected to a series of three pressurization tests nearly identical to the cyclic preconditioning sequences, each having a different value of axial stretch. Data from the loading phase of the final cycle in each of these three tests were used for determination of mechanical properties.
Considering the vessels to be homogenous circular cylinders, specimen stretch λi was defined as
where i = θ, z correspond to the circumferential and axial directions, respectively, and di, do, and l represent specimen inner diameter, outer diameter, and length in the current configuration; Di, Do, and L are corresponding values in the unloaded state and were determined from digital images of specimen cross-sections using measurement software (National Instruments, Inc.; Vision Assistant). The tissue was assumed to be incompressible to allow calculation of inner diameter in the current configuration. Mean Cauchy stress Ti was defined as
where Pi is internal pressure, F is measured axial load, and A is vessel cross-sectional area, to satisfy the global equilibrium equations.
Outcome parameters (Table 1) were derived from the resulting circumferential stress-stretch curves and analyzed statistically. The chosen parameters were designed to capture vessel mechanical response around in vivo conditions. Mixed-effects linear regression analysis was applied separately for each parameter to test for statistically significant changes at various days after carotid ligation or for different drug treatments, depending on the set of data, while accounting for multiple tests on each specimen and controlling for the contribution of axial stretch (Statacorp; Stata IC 10). Data from the first set of experiments were evaluated using the model specified by Equation 3,
where param refers to the parameter value being predicted, ctrl refers to the mean value of param in control specimens, I_daysi corresponds to 1, 2, or 6 days after surgery and, for example, has values of 1, 0, and 0 for a specimen tested 1 day after surgery, λZC is axial stretch, centered about its mean, and β1i, β2, and β3 are coefficients whose values are adjusted to optimize the fit of the model to the data. Data from the MMP inhibitor experiments were analyzed using the same equation except with I_drugi replacing I_daysi to allow comparison of no drug treatment to the use of one of the inhibitors.
Figure 2 shows the outer diameter-pressure response of a typical specimen for three different levels of axial stretch. As expected, larger axial stretch values resulted in smaller diameters such that the diameter-pressure response curve shifted downward. This anticipated effect of axial stretch was observed for all specimens, confirming the need to account for the influence of axial stretch on measured vessel properties, as shown in the statistical model (Eq. 3).
In comparison to normo-flow controls, carotid arteries exposed to flow augmentation for 1 or 2 days demonstrated a marked increase in circumferential stretch (normalized change in diameter; Fig. 3); this increased stretch was present at pressures of both 80 and 120 mmHg. Figure 3 presents three points for each specimen, each collected at a different level of axial stretch. Accounting for multiple tests with various levels of axial stretch on each specimen, the increase in circumferential stretch was statistically significant (p<0.05) after both 1 and 2 days of increased flow, illustrated by observing that the 95% confidence intervals shown in Figure 4 for Idays_1 and Idays_2 do not cross the mean value (0.0 on the x-axis) of circumferential stretch for control specimens. Changes after 6 days of augmented flow were nearly significant. Figure 4 additionally shows that both axial stretch and its square also had significant, or near significant, influence on circumferential stretch values in the tests, though their effect was to reduce the level of stretch, consistent with Figure 2. Statistically predicted changes in stretch shown in Figure 4 correspond to an increase of 0.1 in axial stretch and its square. Of the outcome parameters identified for study in Table 1, only circumferential stretch was found to significantly change as a result of flow augmentation.
Although circumferential stretch values were significantly different from non-ligated controls at both 1 and 2 days after surgery, results from 1, 2, and 6 days were not statistically different from one another. Figures 3 and and44 both, however, illustrate a trend of circumferential stretch values being highest at day 1 and then returning toward baseline values with time. Notably, the timing of this trend is similar to the expression of MMP-9 in carotid arteries subjected to flow augmentation in the same model, reported elsewhere (Nuki et al 2009). In that report, MMP expression was greatest 1 day following ligation surgery, was somewhat reduced at 2 days, and returned to near baseline values by 7 days. This correspondence motivated additional experiments to explore a possible link between the observed change in mechanical properties and expression of MMP-9.
In order to test whether increased levels of MMPs were responsible for the observed mechanical property changes, we assessed effects of two broad spectrum MMP inhibitors on circumferential stretch. As shown in Figure 5, both doxycycline and GM6001 resulted in vessels with lower mean circumferential stretch values, but differences from untreated controls were not statistically significant. Nevertheless, doxycycline appeared to exert a stronger influence than GM6001. These results suggest a possible role for MMPs in the observed mechanical property changes, but more experiments are clearly needed to establish this connection. While it may be tempting to compare the data of Figure 5 to that of Figure 4, it is important to note that the average axial stretch value in the second dataset was higher than that in the first. This likely explains the lower circumferential stretch values seen in the untreated data of Figure 5 relative to that at 1 day in Figure 4.
Cross-section analysis revealed no changes in unloaded vessel dimensions, including outer diameter, cross-sectional area, and stress-free opening angle, as a function of either time after ligation or drug treatment. Accordingly, there was no evidence of structural remodeling in dimensional changes of cross-sections in the zero-load configuration.
The objective of this work was to explore whether measureable changes in carotid artery mechanical properties occur following an abrupt, sustained increase of blood flow. Results indicate that statistically significant alterations occur rapidly but then subside with time. Initial exploratory studies suggest that MMPs may contribute to the observed changes. These findings lend new insight into how blood vessels respond to changes in their mechanical environment.
To our knowledge, the reported rapid change in vessel properties, followed by a return to baseline behavior, has not been previously reported. While this temporal pattern mirrors previously reported expression of MMP-9 in the same model, it is not clear which components of the vessel wall are responsible for the changes in properties. In discussing this question, it is important to note that smooth muscle activity was not chemically eliminated in experiments. Tests were performed in a standard physiological salt solution (DMEM), considered appropriate by many for measurement of passive properties (Humphrey 2002), and in the absence of flow and, thus, any endothelium-dependent smooth muscle signaling. Data for determining properties were also obtained following preconditioning, eliminating some smooth muscle contraction (Fridez et al 2001), but these conditions don’t remove the possibility of a myogenic response at the basal tone level.
Langille (1996) previously suggested that remodeling in response to a change in flow occurs in two phases – an immediate vascular smooth muscle response followed by later structural alterations in the media. The observed increase in circumferential stretch indicates that carotid arteries previously exposed to flow augmentation expanded more under physiological pressure in pressure-diameter tests than did arteries in which flow was normal before resection. The observed increase in circumferential stretch at 1 day appears to be consistent with smooth muscle-induced dilation in response to increased flow, but it is not clear why this dilation would decrease over the next few days while the increased level of flow persisted.
Consistent with Langille’s timeline, no structural remodeling was observed during the period studied, but this would not rule out all structural alteration. Given the increase in circumferential stretch, it is notable that Eiv, the stiffness of the circumferential stress-stretch curve at 100 mmHg, did not change significantly with flow augmentation. These two results together are consistent with a shift of the circumferential stress-stretch curve to the right, a lengthening of the toe region of the curve without significant alteration of the rest of its shape. Because elastin has been shown to define the low-stiffness response of blood vessels (Roach and Burton 1957), it may be that the mechanism producing the observed alterations affects elastin fibers more than other structural elements of the vascular wall, such as collagen. Given the long time required for elastin synthesis (Humphrey 2009), however, the rapid recovery of control behavior could not be accomplished through replacement of elastin, but contributions from other components may compensate.
Increase in circumferential stretch at 1 day could also be the result of damage to the vessel wall by overload. Cope and Roach (1977) observed changes in toe-region elastic properties of isolated human cerebral vessels subjected to over-pressures in vitro, though the pressures applied in their study were much higher than those present here. Other investigators have reported changes in the internal elastic lamina following increased flow in the carotid or basilar arteries (Hoi et al 2008; Tronc et al 2000; Wong and Langille 1996), though large flow changes were required to produce fragmentation of the lamina. Hoi et al (2008) reported remodeling activity in the basilar artery of rabbits following both unilateral and bilateral ligation of the carotid arteries, but only bilateral ligation resulted in internal elastic lamina fragmentation. In contrast, Nuki et al (2009) showed that luminal diameter increased only slightly during the first 5 minutes after ligation in the current model, so it is unlikely that passive overload occurred. Additionally, it is unlikely that a significantly damaged vessel could return to baseline behavior over a small number of days. Nevertheless, histological study of these vessels would help rule out such damage in future work.
Given these considerations, it is not currently possible to determine the cause for the observed change in properties. It appears, however, that inflammation resulting from the rapid change in mechanical environment produces an alteration in either expected smooth muscle response or the way structural components of the wall interact with one another, or both; related work has suggested that structures coupling extracellular matrix to smooth muscle cells may be disrupted in vascular injury (Jamal et al 1992). Future investigation should include appropriate chemicals to manipulate smooth muscle contributions and allow clear distinction between active and passive response.
While the reported data suggest a trend for maximum increase in circumferential stretch at day 1, followed by a decrease back toward baseline levels, the existence and timing of this trend should be further explored with experiments at earlier and later times. Because results at 1 and 2 days were not statistically different from each other and no measurements were made earlier than 1 day, it may be that maximum change occurs at a time other than one day following flow augmentation. In order to more fully understand the vascular response, it is also important to define whether the trend of recovery suggested by the data actually occurs.
Although neither inhibitor treatment demonstrated a statistically significant effect on mechanical properties, both tended to produce lower circumferential stretch ratios than comparable controls, with doxycycline being more influential than GM6001. In contrast to GM6001, doxycycline is less selective and produces effects other than MMP inhibition, including modulation of various aspects of inflammation (Beekman et al 1997; Kim et al 2005; Lee et al 2003), possibly contributing to its relatively greater influence.
The potential significance of such a rapid change in vessel properties is not known. The observed increase in circumferential stretch for a given level of pressure, however, could indicate a loss of vascular integrity with potential for subsequent dysfunction or initiation of disease. Although none of the carotid arteries tested in this study displayed any tendency to not recover from the increase in compliance, mechanical thresholds for the initiation of such instability should be defined. Additionally, while the initiating mechanical event in this study was an abrupt change in shear stress, the approach of this research should be extended to examine response to other types of changes in mechanical environment, such as those that occur with trauma.
In summary, mechanical properties of carotid arteries change rapidly following exposure to augmented flow. Trends suggest that the affected properties may return to baseline values over time and that the timing of changes may be mediated by MMPs. The reported findings further define how blood vessels respond to changes in their mechanical environment and are believed to lend insight into processes associated with the development of vascular disease and dysfunction.
The authors would like to thank Greg Stoddard, MBA, MPH, for guidance on the statistical analysis.
Support: Funding for this study was provided by NIH K25HD048643 (KLM), NIH R01NS027713 (WLY), NIH R01 NS050173 (GTM), NIH R01NS055876 (TH), NIH P01NS044155 (WLY, TH).
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.