Cardiovascular Engineering and Technology

, Volume 1, Issue 2, pp 165–178

Role of Pathologic Shear Stress Alterations in Aortic Valve Endothelial Activation

Authors

  • Daniel Hoehn
    • Department of Aerospace and Mechanical EngineeringUniversity of Notre Dame
  • Ling Sun
    • Department of Aerospace and Mechanical EngineeringUniversity of Notre Dame
    • Department of Aerospace and Mechanical EngineeringUniversity of Notre Dame
Article

DOI: 10.1007/s13239-010-0015-5

Cite this article as:
Hoehn, D., Sun, L. & Sucosky, P. Cardiovasc Eng Tech (2010) 1: 165. doi:10.1007/s13239-010-0015-5

Abstract

Calcific aortic stenosis is the most common aortic valve (AV) disease and is triggered by an active inflammatory process involving endothelial activation and cytokine expression. Interfacing between the leaflet and the surrounding blood flow, shear stress is presumed to play an important role in endothelial injury. This study investigated the hypothesis that pathologic alterations in shear stress magnitude contribute to valvular endothelial activation via BMP-4- and TGF-β1-dependent mechanisms. The fibrosa of porcine AV leaflets was subjected to physiologic, sub-physiologic and supra-physiologic magnitudes of native oscillatory shear stress for 48 h. Endothelial activation was assessed via immunohistochemistry in terms of ICAM-1 and VCAM-1 expressions. Cytokine expression was investigated in terms of BMP-4 and TGF-β1. Pro- and anti-osteogenic media were used to characterize the role of those cytokines in the shear stress-induced pathological response. Supra-physiologic shear stress increased the expression of all biomarkers in a shear stress magnitude-dependent manner. In contrast, neither physiologic nor sub-physiologic shear stress elicited a pro-inflammatory response. While BMP-4 inhibition and supplementation had limited effects on endothelial activation, TGF-β1 supplementation increased the overall leaflet pro-inflammatory state and TGF-β1 inhibition reduced endothelial activation in response to elevated shear stress. Combined TGF-β1 and BMP-4 inhibition completely suppressed shear stress-induced endothelial activation. The results demonstrate that elevated shear stress activates the valvular endothelium on the fibrosa via a BMP-4- and TGF-β1-dependent pathway. The suggested synergy between those cytokines also provides new insights into the transduction of valvular hemodynamic alterations into a pathological response.

Keywords

Aortic valveEndothelial activationShear stressCytokinesAdhesion molecules

Introduction

Calcific aortic stenosis is the most prevalent aortic valve (AV) disease and is present in 8% of the population above 65 years of age.44 The side-specific formation of calcific nodules on the aortic surface of the leaflets14,32,34 contributes to the obstruction of the left ventricular outflow and can lead ultimately to heart failure. The previously accepted theory that linked valvular calcification to a passive wear-and-tear mechanism has lost support due to new developments that have associated the disease with an active process involving inflammation and ossification.17,26,31,38,39 The characterization of calcific lesions suggests that the early stage of the disease is marked by cell proliferation and increased expressions of adhesion molecules, bone morphogenic proteins (BMP) and transforming growth factors-beta (TGF-β).20,30,39 Although the inflammatory stage is thought to be associated with the dysfunction of the leaflet endothelium,42,43 the mechanisms contributing to endothelial activation are not well understood.

Clinical observations, animal and ex vivo studies have suggested that the hemodynamic stress environment experienced by the leaflets may regulate valvular physiology and pathology.15,31,40,41,48,51 Resulting from the relative motion between the leaflet surface and the surrounding pulsatile blood flow, shear stress is an important component of the valve hemodynamic environment.40 The particular valve anatomy and leaflet dynamics give rise to a side-specific shear stress defined by a high unidirectional pulsatile shear stress along the ventricular leaflet surface (ventricularis) and a low bidirectional oscillatory shear stress along the aortic surface (fibrosa).24,49 This complexity has hampered our understanding of the biological processes regulated by the native valvular fluid shear stresses. Although studies have demonstrated the characteristic alignment of valvular endothelial cells perpendicular to steady unidirectional flow5 and the dependence of valvular remodeling activity on steady shear stress magnitude,36 studies on the effects of pathologically relevant shear stress alterations on valvular biology are few. A previous study on the effects of the native oscillatory and pulsatile fluid shear stresses on AV inflammation indicated that simultaneous alterations in shear stress magnitude and pulsatility on the fibrosa could trigger the activation of the endothelium and the upregulation of the cytokines BMP-4 and TGF-β1.45 The significant downregulation of this response following the pharmacological inhibition of BMP-4 or TGF-β1 also suggested a role for those mediators in the shear stress-induced inflammatory pathway.

Although those findings demonstrate the existence of relationships between flow alterations and valvular inflammation, their translation to disease initiation and progression is limited since pathologic valvular hemodynamics is unlikely to result in simultaneous alterations in both shear stress pulsatility and magnitude. For example, the accelerated progression of calcific aortic stenosis19 is accompanied by a reduction in valvular effective orifice area which leads, in turn, to an increased peak aortic velocity.9,33 Similarly, hypertension (i.e., a risk factor for calcific aortic stenosis37) is associated with changes in transvalvular flow rate,21,23,28 which also translates in variations in peak aortic velocity. Those observations may indicate alterations in shear stress magnitude as a key mechanism in the development of valvular disease. Therefore, this study hypothesized that pathological alterations in shear stress magnitude promote valvular endothelial activation via BMP-4- and TGF-β1-dependent mechanisms. This hypothesis was tested via four experiments aimed at: characterizing the effects of normal and pathologic shear stress magnitudes on the expressions of cytokines (BMP-4, TGF-β1) and cell adhesion molecules (ICAM-1, VCAM-1) associated with valvular endothelial activation (experiment 1); investigating the respective role played by BMP-4 and TGF-β1 in valvular endothelial activation in response to pathologic shear stress levels (experiments 2 and 3, respectively); and exploring the synergy between those two cytokines in response to pathologic shear stress magnitudes (experiment 4).

Materials and Methods

Experimental Groups and Conditions

All experiments were conducted on porcine AV leaflets due to the availability of this tissue and its well characterized antibody specificities. The objectives of the study required the use of seven treatment groups: (1) fresh tissue (controls); (2) tissue exposed to shear stress in standard culture medium; (3) tissue exposed to shear stress in standard medium supplemented with the BMP antagonist noggin (100 ng/mL); (4) tissue exposed to shear stress in standard medium supplemented with BMP-4 (10 ng/mL); (5) tissue exposed to shear stress in standard medium supplemented with the TGF-β1 inhibitor SB-431542 (1 μmol/L); (6) tissue exposed to shear stress in standard medium supplemented with TGF-β1 (10 ng/mL); and (7) tissue exposed to shear stress in standard medium supplemented with both noggin (100 ng/mL) and SB-431542 (1 μmol/L). The standard culture medium consisted of Dulbecco’s Modified Eagle’s Medium (DMEM) supplemented with 10% fetal bovine serum (batch #128K0362, Sigma-Aldrich). Noggin is a well-known BMP antagonist that binds to BMPs and prevents its downstream action.12,54 The small molecule inhibitor SB-431542 inhibits specifically the TGF-β type I receptor18,25 and blocks any signaling pathway triggered by the binding of TGF-β1 to its receptor. Given their high binding affinities, noggin and SB-431542 are ideal molecules to investigate the dependence of valvular endothelial activation on BMPs and TGF-β1, respectively. The concentrations considered in this study are consistent with those in previous investigations on the role played by BMPs and TGF-β1 in shear stress- and stretch-induced valvular inflammation.3,45 The assignment of the treatment groups to their respective experiments are summarized in Table 1.
TABLE 1

Experiments and treatment groups considered in this study

 

Group 1

Fresh controls

Group 2

DMEM

Group 3

DMEM + noggin

Group 4

DMEM + BMP-4

Group 5

DMEM + SB-431542

Group 6

DMEM + TGF-β1

Group 7

DMEM + noggin + SB-431542

Experiment 1

X

X

     

Experiment 2

X

 

X

X

   

Experiment 3

X

   

X

X

 

Experiment 4

X

     

X

Since the study focused on the characterization of valvular pathogenesis, mechanical conditioning was only performed along the fibrosa (i.e., surface most prone to calcification14,32,34) using the native shear stress experienced by that leaflet surface. As suggested by simulations of the pulsatile flow through an idealized tri-leaflet valve with a prescribed motion,13,40 the physiologic surface-averaged wall-shear stress experienced by the fibrosa consisted of a bidirectional oscillatory waveform ranging from −8 to +10 dyn/cm2 over a cardiac period of 860 ms, simulating a normal heart rate of 70 beats per minute. As indicated by hot-film anemometry and echocardiographic data, a normal AV is characterized by a peak aortic velocity of 1.35 m/s,35 while mildly and severely stenotic AVs are associated with peak aortic velocities of 2.6 and 5.0 m/s, respectively.33 This increase in velocity caused by the progressive reduction in valvular effective orifice area is expected to translate, in turn, into a proportional increase in leaflet wall-shear stress. Therefore, consistent with this observation, the effects of pathologically relevant shear stress alterations were investigated by scaling up the physiologic shear stress waveform by factors of two and four. Additionally, the dependence of the pathological response on shear stress magnitude was examined by considering two other shear stress waveforms obtained by scaling down the physiologic waveform by similar factors. As a result, the fibrosa was exposed to physiologic (18 dyn/cm2), mild and severe sub-physiologic (4.5 and 9 dyn/cm2, respectively) and mild and severe supra-physiologic (36 and 72 dyn/cm2, respectively) peak-to-peak magnitudes of native bidirectional oscillatory shear stress (Fig. 1).
https://static-content.springer.com/image/art%3A10.1007%2Fs13239-010-0015-5/MediaObjects/13239_2010_15_Fig1_HTML.gif
FIGURE 1

Definitions of the five shear stress conditions considered in this study and imposed on the aortic surface of the leaflets

Tissue Preparation

Porcine hearts were obtained from a local slaughterhouse (Martin’s Custom Butchering, Wakarusa, IN), immediately rinsed in sterile Dubelcco’s phosphate buffered saline (PBS, Sigma-Aldrich) and transported to the laboratory in ice-cold PBS. All subsequent procedures were carried out in a sterile flow hood. Each group consisted of nine circular leaflet samples excised from the basal leaflet region. The nine samples assigned to each treatment group were randomized and selected from different animals. Fresh leaflet samples from the control group (group 1) were processed immediately as outlined in “Immunohistochemical Analysis” section, while samples from the six other groups were conditioned to shear stress in a cone-and-plate bioreactor described and validated previously.46 The whole setup was placed in an incubator at 37 °C and 5% CO2. Tissue from groups 2 to 7 was exposed to shear stress for 48 h, a duration sufficient for biological changes in response to mechanical stimulation to become evident.1,2,27,45,52 Culture medium was continuously perfused during each experiment at a rate of 82 mL/h (i.e., one bioreactor volume/hour) and visually inspected for evidence of contamination.

Immunohistochemical Analysis

The objectives of the biological analyses were 2-fold: (1) to assess whether specific markers associated with endothelial activation were expressed in response to pathologic shear stress levels; and (2) to determine their specific site of expression in the tissue. Immunohistochemistry, which is able to provide such information in one single step, was performed for all analyses. Following shear stress conditioning, tissue samples from groups 2 to 7 were harvested and washed immediately in sterile PBS. For each specimen, the region exposed to flow was separated from the peripheral region that was clamped to maintain the sample in position during the experiments. The specimens were embedded in optimal cutting temperature compound and flash frozen in liquid nitrogen. Tissue sections were then mounted on slides and stored in a −80 °C freezer. Standard immunostaining procedures were used to identify cells positive for VCAM-1 (1:50, Santa Cruz), ICAM-1 (1:50, SouthernBiotech), TGF-β1 (1:25, Santa Cruz), and BMP-4 (1:25, Santa Cruz). Tissue was also probed with vWF (1:200, Sigma) to identify endothelial phenotype and to assess endothelium integrity. Detection of cell apoptosis was performed by a TUNEL assay (Roche Diagnostics).

The intensities of BMP-4-, ICAM-1-, VCAM-1-, and TGF-β1-positive green stains were estimated using Image J (NIH, Bethesda, MD) and normalized by the number of cells visible in each microscope image to yield a quantity consistent to an expression per cell. The number of cells was estimated by counting the number of DAPI-positive nuclei imaged by ultra-violet (UV) epifluorescence. The total expression of a given molecule was assessed by quantifying the intensity of the FITC-positive stain on each image, which was done by computing the integral of the green channel histogram. Cellular expression was calculated as the ratio of the total expression of a given molecule to the number of cells in each image. Apoptosis level was estimated by the ratio of the number of cells with apoptotic fragments as detected under UV epifluorescence using the FITC filter to the total number of cells present in each image.

Statistical Analysis

All quantitative data were expressed as mean ± standard error. The sample size for each experimental group was n = 9. For each experimental condition, the semi-quantitative results were averaged over the nine samples to provide a mean cellular expression. All results were then normalized with respect to the values measured in fresh tissue (group 1) to yield a fold increase with respect to fresh tissue. The data were first analyzed using ANOVA to determine if there was significant contribution by a particular treatment on the measured parameters, followed by the Tukey post hoc test. A p-value of less than 0.05 was used as a measure of statistical significance. All statistical analyses were performed using Minitab 16 (Minitab Inc).

Results

Sub- and Supra-Physiologic Shear Stresses Maintain Endothelium Integrity and Cell Viability

Following the exposure of tissue from group 2 to physiologic, sub-physiologic, and supra-physiologic shear stress, the specimens were analyzed in terms of endothelial integrity and cellular apoptosis. vWF staining (Fig. 2) demonstrated the preservation of the endothelium on the ventricularis and fibrosa at any of the five shear stress levels considered in this study. All samples displayed the same corrugated fibrosa and smooth ventricularis layers as those typically seen in fresh controls. The percentages of apoptotic cells obtained via TUNEL assay in specimens exposed to physiologic, severe sub-physiologic, and severe supra-physiologic shear stress (0.47 ± 0.22%, 0.53 ± 0.26%, and 0.78 ± 0.56%, respectively) were not significantly different from those measured in fresh tissue (0.42 ± 0.21%). Therefore, the acute exposure of the fibrosa to the sub- and supra-physiologic shear stress conditions considered in this study did not affect endothelium integrity and cell viability.
https://static-content.springer.com/image/art%3A10.1007%2Fs13239-010-0015-5/MediaObjects/13239_2010_15_Fig2_HTML.jpg
FIGURE 2

vWF stains in tissue exposed to physiologic, sub-physiologic, and supra-physiologic shear stress (A, aortic surface; V, ventricular surface. Green: endothelial cells; blue: cell nuclei)

Elevated Shear Stress Increases Expression of Pro-Inflammatory Markers on the Fibrosa in a Magnitude-Dependent Manner

The fibrosa of the specimens from group 2 was exposed to different magnitudes of bidirectional oscillatory shear stress in standard culture medium (experiment 1). Under mild and severe supra-physiologic shear stress conditions, positive staining for VCAM-1 and ICAM-1 was found on the endothelial lining of the fibrosa while positive TGF-β1 and BMP-4 staining was detected in both the endothelial and sub-endothelial layers (Fig. 3). In contrast, exposure of the fibrosa to physiologic and sub-physiologic shear stress did not yield any positive staining. Those results are supported by the semi-quantitative analysis that suggests a significant (p < 0.05) 17-fold and 41-fold increase in VCAM-1 and BMP-4 expression, respectively, in tissue exposed to mild supra-physiologic conditions with respect to the fresh controls. Although the analysis demonstrates a threefold and 57-fold increase in ICAM-1 and TGF-β1 expression, respectively, those increases are not statistically significant. Severe supra-physiologic conditions resulted in a 20-fold (p < 0.05), 16-fold, 41-fold (p < 0.05), and 108-fold (p < 0.05) increase in VCAM-1, ICAM-1, BMP-4, and TGF-β1 expression, respectively, relative to fresh controls. Therefore, as the applied shear stress was increased from mild-to-severe supra-physiologic, ICAM-1, and TGF-β1 expressions increased 5-fold and 2-fold, respectively, while VCAM-1 and BMP-4 expressions were essentially maintained constant. This global increase in pro-inflammatory marker expression demonstrates the shear stress magnitude-dependence of the pathological state of the leaflet in response to hemodynamic alterations. Since the results suggest the absence of endothelial activation after exposure to sub-physiologic shear stress, only the mild and severe supra-physiologic shear stress conditions were considered in subsequent experiments (experiments 2–4).
https://static-content.springer.com/image/art%3A10.1007%2Fs13239-010-0015-5/MediaObjects/13239_2010_15_Fig3_HTML.gif
FIGURE 3

Cytokine and adhesion molecule expressions following exposure of the fibrosa to various shear stress magnitudes for 48 h in standard culture medium: immunostaining (A, aortic surface; V, ventricular surface. Blue: cell nuclei; green: positive cells); and semi-quantitative results (A, aortic surface; V, ventricular surface. *p < 0.05 vs. fresh; #p < 0.05 vs. fresh, sub-physiologic or physiologic)

BMP-4 Regulates Adhesion Molecule Expression on the Fibrosa Exposed to Pathologic Shear Stress

In order to determine the role played by BMP-4 in the shear stress-induced valvular endothelial activation (experiment 2), the fibrosa was exposed to elevated (mild and severe supra-physiologic) shear stress in culture medium supplemented with either the BMP antagonist noggin (group 3) or BMP-4 (group 4).

BMP Inhibition Reduces TGF-β1 and Adhesion Molecule Expressions in Response to Elevated Shear Stress

When noggin was added to the standard medium (group 3), positive VCAM-1 staining was detected along the endothelium of the fibrosa while positive staining for TGF-β1 and BMP-4 were detected in both the endothelial and sub-endothelial layers (Fig. 4). The semi-quantitative assessment indicates that noggin treatment on tissue exposed to mild supra-physiologic shear stress magnitude resulted in VCAM-1, ICAM-1, and TGF-β1 expressions statistically similar to those detected in fresh tissue. Under severe supra-physiologic shear stress, the noggin treatment decreased VCAM-1, ICAM-1, and TGF-β1 expressions as compared to the non-noggin treatment (44%, 100%, and 78% (p < 0.05) decrease, respectively). Despite this overall decrease in tissue inflammation, BMP-4 expression remained significantly higher (p < 0.05) than in the fresh controls. For both the mild and severe supra-physiologic shear stress conditions, the noggin treatment had little effect on BMP-4 expression, which remained similar to that detected in tissue from group 2 (standard culture medium).
https://static-content.springer.com/image/art%3A10.1007%2Fs13239-010-0015-5/MediaObjects/13239_2010_15_Fig4_HTML.gif
FIGURE 4

Cytokine and adhesion molecule expressions following exposure of the fibrosa to mild and severe supra-physiologic shear stress for 48 h in medium supplemented with noggin: immunostaining (A, aortic surface; V, ventricular surface. Blue: cell nuclei; green: positive cells); and semi-quantitative results (*p < 0.05 vs. fresh; #p < 0.05 vs. DMEM)

BMP-4 Supplementation Does Not Affect Shear Stress-Induced Endothelial Activation

When the medium was supplemented with BMP-4 (group 4), ICAM-1 and VCAM-1 expression was localized on the endothelium exposed to shear stress, while BMP-4 and TGF-β1 expression was found in both the sub-endothelial and endothelial layers (Fig. 5). Under mild supra-physiologic shear stress, BMP-4 supplementation resulted in increased BMP-4 and ICAM-1 expressions (74% and 245% increase, respectively) and decreased TGF-β1 expression (24% decrease) relative to those obtained in tissue conditioned with standard medium. Similarly, the combination of BMP-4 supplementation and severe supra-physiologic shear stress yielded a 150% increase in both BMP-4 and ICAM-1 expressions and a 42% decrease in TGF-β1 expression relative to the expressions obtained using standard medium. However, for both shear stress conditions, those increases were not statistically significant. VCAM-1 expression following BMP-4 supplementation was statistically similar to that measured in tissue conditioned using standard medium (48% and 9% decrease under mild supra-physiologic and severe supra-physiologic shear stress, respectively). Those results suggest that BMP-4 supplementation had little effect on the pathological state of the tissue in response to elevated shear stress.
https://static-content.springer.com/image/art%3A10.1007%2Fs13239-010-0015-5/MediaObjects/13239_2010_15_Fig5_HTML.gif
FIGURE 5

Cytokine and adhesion molecule expressions following exposure of the fibrosa to mild and severe supra-physiologic shear stress for 48 h in medium supplemented with BMP-4: immunostaining (A, aortic surface; V, ventricular surface. Blue: cell nuclei; green: positive cells); and semi-quantitative results (*p < 0.05 vs. fresh)

TGF-β1 Regulates Adhesion Molecule and BMP-4 Expressions on the Fibrosa Exposed to Pathologic Shear Stress

The role played by TGF-β1 in the shear stress-induced valvular endothelial activation (experiment 3) was investigated by supplementing the standard culture medium with either the specific TGF-β1 inhibitor SB-431542 (group 5) or TGF-β1 (group 6). The fibrosa was exposed to mild and severe supra-physiologic shear stress magnitudes for 48 h.

TGF-β1 Inhibition Prevents Shear Stress-Induced Endothelial Activation

No positive VCAM-1-, ICAM-1-, BMP-4-, or TGF-β1-positive staining could be detected in tissue exposed to either mild or supra-physiologic shear stress following SB-431542 treatment (group 5) (Fig. 6). After exposure to mild supra-physiologic shear stress, TGF-β1 inhibition resulted in a 40%, 90% (p < 0.05), 71% (p < 0.05), and 79% (p < 0.05) decrease in ICAM-1, VCAM-1, BMP-4, and TGF-β1, respectively, relative to the expressions obtained in tissue exposed to similar shear stress with standard culture medium (group 2). After exposure to severe supra-physiologic shear stress, TGF-β1 inhibition resulted in a 79%, 91% (p < 0.05), 63%, and 84% (p < 0.05) decrease in ICAM-1, VCAM-1, BMP-4, and TGF-β1, respectively, relative to the expressions detected in tissue from group 2. Regardless of the degree of supra-physiologic shear stress to which the fibrosa was exposed, the supplementation of the culture medium with SB-431542 returned pro-inflammatory and adhesion molecule expressions to the levels measured in fresh controls.
https://static-content.springer.com/image/art%3A10.1007%2Fs13239-010-0015-5/MediaObjects/13239_2010_15_Fig6_HTML.gif
FIGURE 6

Cytokine and adhesion molecule expressions following exposure of the fibrosa to mild and severe supra-physiologic shear stress for 48 h in medium supplemented with SB-431542: immunostaining (A, aortic surface; V, ventricular surface. Blue: cell nuclei; green: positive cells); and semi-quantitative results (*p < 0.05 vs. fresh; #p < 0.05 vs. DMEM)

TGF-β1 Supplementation Increases Endothelial Activation in Response to Severe Supra-Physiologic Shear Stress

Consistent with the observations made in tissue from other treatment groups, tissue exposed to supra-physiologic shear stress in medium supplemented with TGF-β1 (group 6) expressed ICAM-1 and VCAM-1 in the endothelial layer and BMP-4 and TGF-β1 in both the endothelial and sub-endothelial layers of the fibrosa (Fig. 7). Exposure of the tissue to mild supra-physiologic shear stress resulted essentially in the same levels of expression of cytokines and adhesion molecules as those measured following exposure to a similar condition in standard medium (group 2). VCAM-1, BMP-4, and TGF-β1 expressions decreased by 41%, 8%, and 12%, respectively. Although ICAM-1 expression increased by 246% with respect to the level measured in tissue conditioned with standard medium, this expression was statistically similar to that in fresh tissue. In contrast, TGF-β1 supplementation combined with severe supra-physiologic shear stress significantly increased the pro-inflammatory state of the tissue. As indicated by the semi-quantitative analysis, this combination resulted in a significant (p < 0.05) increase in all biomarker expressions as compared to those in fresh tissue. More importantly, the results reveal a 104% (p < 0.05), 76%, 141% (p < 0.05), and 51% increase in VCAM-1, ICAM-1, BMP-4, and TGF-β1 expressions, respectively, relative to the those measured in tissue exposed to similar stress in standard medium (group 2), suggesting an overall increase in pro-inflammatory state.
https://static-content.springer.com/image/art%3A10.1007%2Fs13239-010-0015-5/MediaObjects/13239_2010_15_Fig7_HTML.gif
FIGURE 7

Cytokine and adhesion molecule expressions following exposure of the fibrosa to mild and severe supra-physiologic shear stress for 48 h in medium supplemented with TGF-β1: immunostaining (A, aortic surface; V, ventricular surface. Blue: cell nuclei; green: positive cells); and semi-quantitative results (*p < 0.05 vs. fresh; #p < 0.05 vs. DMEM)

Combined BMP-4 and TGF-β1 Inhibition Suppresses Cell Adhesion Molecule and Cytokine Expressions on the Fibrosa Exposed to Elevated Shear Stress

The synergistic effects of TGF-β1 and BMP-4 in the shear stress-induced valvular endothelial activation (experiment 4) was investigated by supplementing the standard culture medium with both the specific TGF-β1 inhibitor SB-431542 and BMP antagonist noggin (group 7). The fibrosa was exposed to mild and severe supra-physiologic shear stress magnitudes for 48 h. Under both shear stress conditions, the combined noggin + SB-431542 treatment significantly reduced cytokine and cell adhesion molecule expressions as compared to those measured in tissue exposed to similar conditions in standard medium (Fig. 8). The simultaneous inhibition of BMP and TGF-β1 resulted in a 100%, 76% (p < 0.05), 91% (p < 0.05), and 93% (p < 0.05) decrease in ICAM-1, VCAM-1, BMP-4, and TGF-β1, respectively, relative to the expressions obtained in tissue from group 2 (standard culture medium) after exposure to mild supra-physiologic shear stress. Following exposure to severe supra-physiologic shear stress, the same treatment resulted in a 100%, 98% (p < 0.05), 88% (p < 0.05), and 94% (p < 0.05) decrease in ICAM-1, VCAM-1, BMP-4, and TGF-β1, respectively, relative to the expressions detected in tissue exposed to similar shear stress with standard culture medium. Regardless of the degree of supra-physiologic shear stress, the concurrent supplementation of the culture medium with noggin and SB-431542 brought the pro-inflammatory and cell adhesion molecule expressions back to the levels measured in fresh controls.
https://static-content.springer.com/image/art%3A10.1007%2Fs13239-010-0015-5/MediaObjects/13239_2010_15_Fig8_HTML.gif
FIGURE 8

Cytokine and adhesion molecule expressions following exposure of the fibrosa to mild and severe supra-physiologic shear stress for 48 h in medium supplemented with a combination of SB-431542 and noggin: immunostaining (A, aortic surface; V, ventricular surface. Blue: cell nuclei; green: positive cells); and semi-quantitative results (*p < 0.05 vs. fresh; #p < 0.05 vs. DMEM)

Discussion

The objectives of this study were to explore the acute effects of sub- and supra-physiologic shear stress on valvular endothelium activation, and to characterize the role played by the cytokines BMP-4 and TGF-β1 in that response. The contributions can be summarized under the following three points: (1) exposure of the fibrosa to supra-physiologic levels of shear stress stimulates the expressions of cytokines and cell adhesion molecules within 48 h; (2) the cytokines BMP-4 and TGF-β1 interact to synergistically regulate endothelial activation in response to elevated shear stress; and (3) TGF-β1 has more effect on the shear stress-induced pathological response than BMP-4.

Role of Endothelium in Valvular Pathogenesis

The absence of cytokine and adhesion molecule expression following the exposure of the fibrosa to its physiologic level of bidirectional oscillatory shear stress is consistent with previous studies that demonstrated the key role of physiologic hemodynamic stresses in preserving valvular function.1,50 The onset of a pathological response localized in the endothelial and sub-endothelial layers after exposure of the fibrosa to elevated shear stress magnitudes is also supported by previous results that demonstrated the ability of drastic alterations in shear stress pulsatility to induce a pro-inflammatory response.45 Therefore, this study provides new evidence that the fibrosa is sensitive to its hemodynamic environment and is able to transduce alterations in both shear stress pulsatility and magnitude into a pro-inflammatory response. The detection of cytokine and adhesion molecule expression in the endothelial layer of the fibrosa after exposure to elevated shear stress is consistent with the known athero-prone transcriptional profile of this endothelium43 and indicates the critical role played by the valve endothelial cell population in the transduction of abnormal hemodynamic stimuli into a pathological state.

The results of this study provide further evidence of the distinct differences between the valvular and vascular endothelia.6,11 While regions of the vasculature exposed to low oscillatory shear stress are prone to atherosclerosis and plaque formation,7,8 the present results demonstrate that only supra-physiologic levels of oscillatory shear stress trigger the acute activation of the endothelium lining the aortic surface of the valve leaflets. On the other hand, neither physiologic nor sub-physiologic oscillatory shear stress level elicited a similar pathological response. Therefore, the present findings tend to lend less credibility to the atherosclerotic root of valvular calcification and suggest the involvement of environmental cues such as hemodynamic alterations in valvular pathogenesis.

Shear Stress in Valvular Disease

Although the shear stress alterations considered in this study have not been measured under native disease conditions, the results may yet be pathophysiologically relevant. In fact, the progression of calcific aortic stenosis is accompanied by an increased peak aortic velocity,33,35 which could translate into a 2-fold and 4-fold increase in valvular wall-shear stress magnitude under mild and severe stenotic conditions, respectively. Therefore, the results, which suggest an increased pathological state under elevated shear stress, could also explain the accelerated progression of valvular calcification in patients with stenotic19 and bicuspid4 AVs. Studies on the characterization of the hemodynamic forces experienced by AV leaflets under diseased conditions are currently under way in our laboratory in order to study the impact of pathologically relevant shear stress conditions on valvular inflammation and calcification. Specifically, we are developing computational and experimental methodologies based on fluid structure interaction modeling and particle-image velocimetry, respectively, to capture the native valvular stress state under normal and disease conditions (e.g., bicuspid aortic valve, calcified aortic valve). This hemodynamic characterization will permit to investigate the effects of the native disease-induced shear stresses on valvular pathobiology using the same ex vivo approach as that described in this study.

It is important to note that, although valvular calcific degeneration occurs over a much longer timescale than that considered in this study, the present results demonstrate the existence of acute biological changes in response to mechanical stimulation. Those acute effects are important as they may shape the longer-term mechanisms involved in the progression of valvular disease. Therefore, a complete understanding of valvular degenerative processes can only be obtained by considering both the acute and long-term response to mechanical cues. A new shear stress bioreactor capable of maintaining fresh leaflet tissue under a near-native biomechanical environment for longer periods is being designed in our laboratory and will provide more insights into valvular pathogenesis.

Potential Mechanosensitivity of TGF-β1 and BMP-4

An important contribution of this study is the confirmation of the global decrease in shear stress-induced valvular endothelial activation after silencing TGF-β1 and BMP-4. While the BMP antagonist noggin and the TGF-β1 inhibitor SB-431542 each resulted in a decrease in fibrosa activation following exposure to elevated shear stress, the combined inhibition of those cytokines completely suppressed the shear stress-induced pathological response and returned adhesion molecule expressions to levels similar to those detected in fresh tissue. This observation suggests the key role played by those cytokines and their synergy in the transduction of pathologic hemodynamic alterations into a pro-inflammatory response. This notion is supported by numerous studies that have identified TGF-β1 and BMPs as molecules involved in the early stage of valvular calcification,16,22,29,47 and studies that have demonstrated their side-specific expression on the disease-prone fibrosa.43 Importantly, the detection of TGF-β1 and BMP-4 expression in both the endothelial and subendothelial layers of the fibrosa and the clear dependence of the shear stress-induced endothelial activation on those molecules suggest the mechanosensitivity of those cytokines and their ability to transduce abnormal hemodynamic stresses into a pro-inflammatory response on the fibrosa. Although more studies are needed to confirm this hypothesis, the present results clearly demonstrate strong interactions between BMP-4 and TGF-β1 in valvular pathogenesis.

Model of BMP-4/TGF-β1 Synergy in Shear-Induced Valvular Endothelial Activation

While mechanistic insights into the shear-induced endothelial activation pathway need further specific studies, the present results permit to hypothesize a model of the interplay between TGF-β1 and BMP-4 in response to elevated shear stress. TGF-β1 inhibition and supplementation affected not only adhesion molecule expression but also BMP-4 expression, while BMP-4 inhibition and supplementation mostly affected vascular cell adhesion molecule expression. This suggests that TGF-β1 plays a role upstream of BMP-4 in the shear-induced endothelial activation pathway. Additionally, the reduction in endothelial activation following TGF-β1 inhibition was associated with a significant decrease in TGF-β1 expression. As a specific inhibitor of the TGF-β type I receptor,18,25 SB-431542 should have no effect on TGF-β1 expression, unless TGF-β1 is regulated by downstream mechanisms that depend in turn on the binding of TGF-β1 to its receptors. Therefore, the shear stress magnitude-dependence of valvular inflammation may be caused by, at least, four mechanisms (Fig. 9): (1) the upregulation of TGF-β1 in response to elevated shear stress on the fibrosa; (2) the upregulation of BMP-4 following the binding of TGF-β1 to its receptors; (3) a positive feedback exerted on TGF-β1 expression by the binding of BMP-4 to its receptors; and (4) the upregulation of the adhesion molecules VCAM-1 and ICAM-1 resulting from the upregulation of those cytokines. This four-component pathway controlled by TGF-β1 would explain the drastic effect of the inhibition of this cytokine on adhesion molecule and BMP-4 expressions, and the milder effect of BMP-4 inhibition on the overall pathological response. Although more experiments are necessary to test this pathway, the dominant role of TGF-β1 in valvular calcification has already been suggested in previous studies.31 TGF-β1 induces differentiation of valvular interstitial cells into a bone-like phenotype via apoptosis.10,20,53 TGF-β1 stimulation of canine valvular interstitial cells was also shown to increase calcified nodule formation as compared to BMP stimulation.29 Those results combined with those described in this study clearly demonstrate the importance of BMP-4 and TGF-β1 in the progression of valvular calcification.
https://static-content.springer.com/image/art%3A10.1007%2Fs13239-010-0015-5/MediaObjects/13239_2010_15_Fig9_HTML.gif
FIGURE 9

Hypothetical synergy between TGF-β1 and BMP-4 in the shear stress-induced valvular endothelial activation pathway

Conclusion

This study provides new evidence of the key role played by abnormal hemodynamic forces in the progression of valvular disease. The results suggest the involvement of the leaflet endothelium and the cytokines TGF-β1 and BMP-4 in the transduction of fluid shear stress alterations into a pro-inflammatory response. A more detailed characterization of the molecular mechanisms mediated by TGF-β1 and BMP-4 is needed for the development of targeted pharmacological modalities aimed at preventing the onset or slowing the progression of valvular calcification.

Acknowledgments

The authors thank Steven DeLaurentis and Michael O’Connor (University of Notre Dame) for their assistance with the experiments, Leon Hluchota (University of Notre Dame) for his advice on the bioreactor design and fabrication, and Martin’s Custom Butchering (Wakarusa, IN) for supplying porcine hearts. This work was supported by the Faculty Research Program at the University of Notre Dame.

Copyright information

© Biomedical Engineering Society 2010