An In Vivo Murine Model of Low-Magnitude Oscillatory Wall Shear Stress to Address the Molecular Mechanisms of Mechanotransduction—Brief Report
Objective—Current understanding of shear-sensitive signaling pathways has primarily been studied in vitro largely because of a lack of adequate in vivo models. Our objective was to develop a simple and well-characterized murine aortic coarctation model to acutely alter the hemodynamic environment in vivo and test the hypothesis that endothelial inflammatory protein expression is acutely upregulated in vivo by low-magnitude oscillatory wall shear stress (WSS).
Methods and Results—Our model uses the shape memory response of nitinol clips to reproducibly induce an aortic coarctation and allow subsequent focal control over WSS in the aorta. We modeled the corresponding hemodynamic environment using computational fluid dynamics and showed that the coarctation produces low-magnitude oscillatory WSS distal to the clip. To assess the biological significance of this model, we correlated WSS to inflammatory protein expression and fatty streak formation. Vascular cell adhesion molecule-1 expression and fatty streak formation were both found to increase significantly in regions corresponding to acutely induced low-magnitude oscillatory WSS.
Conclusions—We have developed a novel aortic coarctation model that will be a useful tool for analyzing the in vivo molecular mechanisms of mechanotransduction in various murine models.
Atherosclerosis is an inflammatory disease of the vasculature that is predisposed to localization in regions uniquely characterized by disturbed blood flow and the resultant low-magnitude oscillatory wall shear stress (WSS).1,2 Mechanosensitive pathways have therefore been implicated as important mediators in the pathogenesis of cardiovascular disease. Previous studies have identified likely mechanisms of mechanotransduction and downstream signaling pathways as summarized in numerous reviews.3 However, experimental flow models have largely been limited to in vitro methods, which have highly simplified, nonphysiologic flow environments; lack complex cell-cell and cell-matrix interactions; and have variable conditions (cell line, time course, culture conditions, etc.) between studies. These limitations have led to uncertainties in the applicability of these mechanotransduction pathways to in vivo conditions.
Mouse models of disturbed flow are beginning to be used for the in vivo analysis of molecular mechanisms of mechanotransduction. These models fall into 2 categories: chronic and device-based. Chronic models use regions where the innate morphology produces chronically disturbed flow, including the aortic arch and the brachiocephalic branch.4 The complex morphology in these regions makes for challenging analysis, and the chronic flow environment allows for activation of compensatory mechanisms; thus, studies are limited to the analysis of the atherosusceptible endothelial phenotype or lesion development. Alternatively, device-based models can acutely disturb flow and isolate WSS-induced signaling from chronic compensatory mechanisms.5,6 Although these models have provided insights into the molecular mechanisms involved, the associated hemodynamic environments are very complex. Critical limitations include large alterations of pressure (or lack of characterization of pressure), resulting in an unintended mechanical stimulus, or no oscillations in flow, resulting in a WSS profile that differs from many in vivo regions of pathogenesis. Because of these complexities, there is currently a need for an easily implemented and highly reproducible, acute, in vivo model of low-magnitude oscillatory WSS in which molecular mechanisms of mechanotransduction can be analyzed. To address this need, we hypothesized that using a nitinol clip, we could reliably produce a mouse coarctation model that would induce quantifiable acute changes in WSS that would subsequently result in increased expression of flow-mediated inflammatory proteins.
Detailed methods related to animal care, nitinol clip manufacturing, surgical methodology, computational fluid dynamics (CFD), and histology are described in the Supplemental Data, available online at http://atvb.ahajournals.org. Briefly, we used shape memory nitinol clips with an inner diameter smaller than the aortic diameter of a mouse. For each anesthetized mouse, the aorta was exposed, and a nitinol clip deformed to an open state (Figure 1A) was inserted underneath the aorta. The body temperature of the mouse thermally activated the shape memory recovery of the clip, thereby decreasing the aortic diameter and inducing an aortic coarctation. A CFD model was then created to determine the hemodynamic environment near the coarctation. To assess the biological significance, we stained for either vascular cell adhesion molecule-1 (VCAM-1) expression over an acute time course or fatty streak formation over a chronic time course.
We found that nitinol clips could be used to effectively induce an aortic coarctation. See the Supplemental Results section for the characterization of the coarctation.
Characterization of the Hemodynamic Environment
WSS maps and cross sectional velocity vectors were generated using a CFD model for control, noncoarctation aortas and coarctation aortas (n=3) (Figure 1 and Supplemental Figure III). These models showed unidirectional and relatively high-magnitude WSS in the upstream, thoracic region of control and coarcted mice. The spatial heterogeneity of the WSS was relatively minimal, with some skewing toward the posterior side due to the curvature of the spine. Downstream from the site of the aortic coarctation the model showed low-magnitude oscillatory WSS on the anterior side of the aorta, whereas the control animals showed unidirectional flow with high-magnitude WSS. Low-magnitude oscillatory WSS was observed in aortic coarctation CFD models from 3 different mice, demonstrating the reproducibility of this flow environment.
From these results, we identified 3 regions of interest: the thoracic aorta, the low-magnitude oscillatory WSS region in the abdominal aorta of mice from the coarctation group, and the comparable abdominal region near the celiac branch in control animals. We included the aortic arch as an additional region of interest on the basis of previous studies showing chronic induction of low oscillatory WSS in this region.4 This animal model can therefore be used to compare regions of high-magnitude unidirectional WSS with regions of both acutely and chronically disturbed flow producing low-magnitude oscillatory WSS.
We used quantum dot–based immunohistochemistry to quantify the expression of a representative inflammatory protein, VCAM-1, in our model (n=8 to 10). Control noncoarcted aortas and control aortas with a sham, nonconstricting clip showed low levels of VCAM-1 expression throughout the celiac region of the aorta (Figure 2). Aortas with a coarctation-inducing nitinol clip showed a significant increase in VCAM-1 expression in the region of disturbed flow as identified earlier. This increased VCAM-1 expression was comparable to VCAM-1 expression in regions of chronically disturbed flow and significantly higher than in regions of steady laminar flow (Figure 2).
We used apolipoprotein E−/− mice on a normal chow for 2 months as a model of early lesion formation (n=5). Low levels of fatty streaks formed in the control animals and animals with a sham, nonconstricting clip, whereas fatty streaks formation was significantly increased in the disturbed flow region of the coarctation model (Figure 2).
Our study presents a novel mouse model of acutely disturbed flow. We have thoroughly characterized the hemodynamic environment associated with our model. Using a combination of techniques (in situ micro-CT and ex vivo pressure inflation), we showed that we can effectively constrict the aortic diameter within the nitinol clip with minimal variability. The difference between the 2 measurements can be attributed to the thickness of the vessel wall, which is included in the ex vivo inflation measurements. We further showed that this constriction creates a small degree of stenosis. This stenosis produced minimal pressure drop across the coarctation, as confirmed by blood pressure measurements. The CFD model showed that the coarctation produced low-magnitude oscillatory WSS in the distal region. This coarctation model is therefore a model that can uniquely analyze the response to acute changes in WSS without a significant effect on blood pressure. We further show that the coarctation produces acute inflammatory protein expression as well as chronic fatty streak formation. Mice with a sham, nonconstricting clip showed that the biological response, including both VCAM-1 expression and fatty streak formation, was a product of the altered hemodynamic environment and not the perivascular tissue dissection or an inflammatory response to the nitinol material. The acute shear response downstream from the coarctation can also be compared to the chronic shear response in the aortic arch, thereby allowing for analysis of compensatory mechanisms and the atherosusceptible endothelial phenotype.
We are currently limited in our in vivo imaging methodologies. Using MR, we cannot obtain images at the clip because of magnetic susceptibility artifacts, whereas ultrasound is similarly not possible because of the presence of the clip. To address these limitations, we used an ex vivo perfusion setup to measure the compliance of the aorta. The inflation experiment was used to show that the aortic diameter is constricted within the coarctation and that the clip does not deform under a luminal physiological blood pressure. In addition, we performed a sensitivity analysis on the boundary conditions to show that within a physiological blood pressure range, our parameters (outflow pressure and expansion factor) do not affect the observed flow reversal, although there is some change in the heterogeneous distribution of WSS.
We conclude that this model provides a number of advantages over existing in vivo mouse models of disturbed flow, including simple implementation; high reproducibility; a well-characterized, low-oscillatory-WSS environment; negligible effect on pressure; and both an acute and a chronic disturbed flow environment. This novel model will be an important tool in which to assess the in vivo applicability of shear sensitive signaling pathways, already determined through in vitro experimentation. The availability of transgenic and knockout animals make mice an ideal animal by which to investigate these mechanosensitive molecular mechanisms. Further understanding of how these pathways act in vivo will be important in the development of new therapies to treat cardiovascular disease.
Sources of Funding
This work was supported by National Institutes of Health Grants R01-HL70531, R01-HL090584, and U01-HL080711 and by a predoctoral research fellowship from the Southeast Affiliate of the American heart Association.
Received on: January 4, 2010; final version accepted on: July 7, 2010.
Ku DN, Giddens DP, Zarins CK, Glagov S. Pulsatile flow and atherosclerosis in the human carotid bifurcation. Positive correlation between plaque location and low oscillating shear stress. Arteriosclerosis. 1985; 5: 293–302.
Suo J, Ferrara DE, Sorescu D, Guldberg RE, Taylor WR, Giddens DP. Hemodynamic shear stresses in mouse aortas: implications for atherogenesis. Arterioscler Thromb Vasc Biol. 2007; 27: 346–351.
Cheng C, Tempel D, van Haperen R, van der Baan A, Grosveld F, Daemen MJ, Krams R, de Crom R. Atherosclerotic lesion size and vulnerability are determined by patterns of fluid shear stress. Circulation. 2006; 113: 2744–2753.
Wu JH, Hagaman J, Kim S, Reddick RL, Maeda N. Aortic constriction exacerbates atherosclerosis and induces cardiac dysfunction in mice lacking apolipoprotein E. Arterioscler Thromb Vasc Biol. 2002; 22: 469–475.