Preclinical techniques to investigate exercise training in vascular pathophysiology

Abstract

Atherosclerosis is a dynamic process starting with endothelial dysfunction and inflammation and eventually leading to life-threatening arterial plaques. Exercise generally improves endothelial function in a dose-dependent manner by altering hemodynamics, specifically by increased arterial pressure, pulsatility, and shear stress. However, athletes who regularly participate in high-intensity training can develop arterial plaques, suggesting alternative mechanisms through which excessive exercise promotes vascular disease. Understanding the mechanisms that drive atherosclerosis in sedentary versus exercise states may lead to novel rehabilitative methods aimed at improving exercise compliance and physical activity. Preclinical tools, including in vitro cell assays, in vivo animal models, and in silico computational methods, broaden our capabilities to study the mechanisms through which exercise impacts atherogenesis, from molecular maladaptation to vascular remodeling. Here, we describe how preclinical research tools have and can be used to study exercise effects on atherosclerosis. We then propose how advanced bioengineering techniques can be used to address gaps in our current understanding of vascular pathophysiology, including integrating in vitro, in vivo, and in silico studies across multiple tissue systems and size scales. Improving our understanding of the antiatherogenic exercise effects will enable engaging, targeted, and individualized exercise recommendations to promote cardiovascular health rather than treating cardiovascular disease that results from a sedentary lifestyle.

Keywords: atherosclerosis; bionengineering; cardiovascular; exercise; preclinical.

Figure 1.

Atherosclerosis progression from endothelial dysfunction to complex plaque formation. Prolonged endothelial exposure to disturbed or physiologically high blood flow initiates endothelial dysfunction and arterial wall injury. Macrophages enter the arterial wall where they phagocytize oxidized low-density lipoprotein (LDL) and transform into foam cells that accumulate to form fatty streaks. Vascular smooth muscle cell (VSMC) migration into the intima promotes an advanced fibrotic plaque phenotype. Plaque calcification may be augmented in athletes due to exercise-driven mechanisms.

Figure 2.

In vitro flow devices are used to study hemodynamic effects on endothelial function. Parallel-plate flow chambers allow live cell imaging but require a large fluid volume that limits metabolite detection. Cone and plate devices require less fluid and enable complex arterial flow waveforms, but cell visualization is challenging. Microfluidic devices are compact and customizable for high throughput studies diverse vascular geometries, but their small scale limits disturbed flow applications. SMC, smooth muscle cell.

Figure 3.

In vitro parallel plate flow chamber used to study disturbed flow. Computational fluid dynamics was used to design gasket geometries that produce disturbed flow (A). Endothelial monolayer alignment in the high shear stress zones validated the in silico simulations (B). Flow effects were then assessed through endothelial morphology (actin alignment) and function [nitric oxide production via diaminofluorescein (DAF) and inflammation via monocyte adhesion] (C). A and C adapted with permission from Sedlak and Clyne (100).

Figure 4.

Atherosclerotic plaques in genetically modified and surgically manipulated mice. ApoE and ldlr knockout mice form replicable lipid lesions at branch points and curvatures along the arterial tree (A). Surgical manipulation of genetically modified mouse carotid arteries (A, dissection image) allows diverse hemodynamic alterations that rapidly induce large plaques with complex tissue composition (B, C, and D). Yellow, orange, and red arrows highlight the unmodified right common carotid artery, surgically manipulated left common carotid artery, and aortic arch. Oil red O histology images confirm lipid deposition 4 wk after surgery. Partial carotid ligation histology includes both oil red O and hematoxylin and eosin staining. ECA, ICA, OA, and STA denote external and internal carotid, and occipital and superior thyroid arteries, respectively. Scale bar denotes 1 mm in dissection image and 100 μm in histology image. Complete ligation and stenosis oil red O histology images are adapted from Chang et al. (180) and Ding et al. (181), respectively.

Figure 5.

Ultrasound characterization of a murine carotid artery before and after partial carotid ligation-induced atherosclerosis (203). B-mode ultrasound visualized vessel morphology at baseline and 21 days after partial carotid ligation (PCL; A), serial B-mode imaging quantified vessel diameter and volume (B), Doppler imaging assessed blood flow velocity and direction (C), and M-mode ultrasound measured vessel pulsatility (D). Novel 4-dimensional ultrasound also assessed spatially resolved vessel pulsatility (E). Red dotted lines highlights the artery, and yellow arrows point to arterial plaques. Scale bar = 0.5 mm.

Figure 6.

Metabolic flux predictions in an endothelial-specific flux balance analysis model. Metabolic enzymes in human umbilical vein endothelial cells (HUVECs) cultured on Matrigel were analyzed by proteomics at 4 and 24 h. The proteomic data were then used to refine a genome scale metabolic model (A). The refined genome-scale metabolic model predicted decreased in glycolytic and increased fatty acid oxidation flux (B). Glycolysis enzymes include hexokinase (HK), phosphofructokinase (PFK), aldolase (ALDO), lactate dehydrogenase (LDH), pyruvate dehydrogenase (PHD). Fatty acid oxidation enzymes include carnitine O-palmitoyltransferase (CPT), medium(M)/short(S)-chain specific acyl-CoA dehydrogenase (ACADM/S), and diffusion octanoyl-CoA (Diff). Isotopic labeling experiments verified the metabolic model, revealing decreased labeled pyruvate (glycolysis) and increased labeled palmitoyl carnitine (fatty acid oxidation) (C). ACADM/S₁ and ACADM/S₂ represent 2 different reactions that are triggered by acyl-CoA dehydrogenase. Figure was modified and reprinted from Patella et al. (385).

Figure 7.

Comparison of time-averaged wall shear stress (WSS) and oscillatory shear index (OSI) in anesthetized, resting, and active apoE KO mice. Rest 1 and 2 represent the estimated lower and upper cardiovascular boundary conditions in resting mice. Active 1 and 2 denote the estimated lower and upper cardiovascular boundary conditions in physically active mice. The qualitative comparison shows increased time-averaged WSS magnitude in active 2 mice (A) but no changes in time-averaged (TA) WSS pattern across groups (B). OSI simulations revealed greater flow reversal in the external carotid artery in active 2 mice than in other groups (C). The quantitative comparison showed greater reversed WSS area with increasing activity in both computational fluid dynamics (CFD) and fluid-structure interaction (FSI) models (D). D highlights the stagnation point (blue) and reversed fluid flow (red) in resting 2 mice using CFD. ECA, external carotid artery; ICA, internal carotid artery; CCA, common carotid artery. Figure was modified and reprinted from De Wilde et al. (305).