Tracer-based metabolomics for profiling nitric oxide metabolites in a 3D microvessel-on-a-chip model

Endothelial dysfunction is a common denominator in cardiovascular diseases (CVDs) associated with diabetes, hypertension, obesity, renal failure or hypercholesterolemia. In these disease states, circulating adverse metabolic or hemostatic risk factors drive the progression of inflammation, thrombosis, platelet activation and atherosclerosis. A hallmark of endothelial dysfunction is the reduced bioavailability of nitric oxide (NO), a signaling molecule essential for vascular homeostasis. Numerous studies have focused on NO synthesis by endothelial cells (ECs) using in vitro cultures to understand the pathophysiology of endothelial dysfunction. A limitation of these studies is that the expression of the NO-generating enzyme, endothelial nitric oxide synthase (eNOS), in physiological conditions is modulated by the exposure of the ECs to laminar shear stress, a stimulus that is clearly lacking in most two-dimensional (2D) cultures. Here we developed a tracer-based metabolomics approach to measure NO-specific metabolites with mass spectrometry (MS) and show the impact of unidirectional fluid flow on metabolic parameters associated with NO synthesis using 2D and three-dimensional (3D) platforms. Specifically, we tracked the conversion of stable-isotope labeled NO substrate L-Arginine to L-Citrulline and L-Ornithine to determine eNOS activity. We demonstrated that when human coronary artery endothelial cells (HCAECs) cultured in media containing 13C6,15N4-L-Arginine treated with eNOS stimulator - vascular endothelial growth factor (VEGF), eNOS inhibitor - L-NAME and arginase inhibitor - S-(2- boronoethyl)-L-cysteine (BEC), their downstream metabolites - 13C6,15N3 L-Citrulline and 13C5,15N2 L- Ornithine showed clear responses as measured using Ultra-performance liquid chromatography tandem mass spectrometry (UPLC-MS/MS). In this study, we also assessed the NO metabolic status of a static 2D culture, a 3D microvessel model with bidirectional flow, and our 3D model with unidirectional fluid flow generated by a microfluidic pump. Compared to 2D culture, our 3D model showed significant effects in the control and microvessels exposed to VEGF when Citrulline/Ornithine ratio was analyzed. The obtained result indicates that the 2D static culture mimics more endothelial dysfunction status. Our detection method and 3D model with a unidirectional fluid flow provides a more representative physiological environment that exhibits perfect model to study endothelial dysfunction.