Endothelial cell distributions and migration under conditions of flow shear stress around a stent wire

Abstract

BACKGROUND:

Blood vessels are constantly exposed to flow-induced stresses, and endothelial cells (ECs) respond to these stresses in various ways.

OBJECTIVE:

In order to facilitate endothelialization after endovascular implantation, cell behaviors around a metallic wire using a flow circulation system are observed.

METHODS:

A parallel flow chamber was designed to reproduce constant shear stresses (SSs) on cell surfaces and to examine the effects of a straight bare metal wire on cell monolayers. Cells were then exposed to flow for 24 h under SS conditions of 1, 2, and 3 Pa. Subsequently, cell distributions were observed on the plate of the flow chamber and on the surface of the bare metal wire. Flow fields inside the flow chamber were analyzed using computational fluid dynamics under each SS condition.

RESULTS:

After 24 h, ECs on the bottom plate were concentrated toward the area of flow reattachment. The matching of higher cell density and CFD result suggests that flow-induced stimuli have an influence on EC distributions.

CONCLUSION:

Typical cell concentration occurs on dish plate along the vortexes, which produces large changes in SSs on cell layer.

Keywords

Endothelial cell computational flow dynamics stent cell migration wall shear stress

1. Introduction

Restenosis and thrombosis are occasional complications of artery stenting [1, 2, 3, 4]. Due to the fact that expanding stents exert a pressure on the vascular intima, stent wires injure endothelial cell (EC) layers and expose the underlying smooth muscle cells (SMCs). Previous studies showed that EC layers prevent thrombus adhesion [5] and suppress excessive SMC proliferation in the neointima [6]. Thus, the clinical complications of stents are considered to be a consequence of incomplete endothelialization or endothelial dysfunction around stents.

EC layers are constantly exposed to blood flow in blood vessels, and they respond to the resulting shear stress (SS) in various ways. Several studies suggested that both low and high SS are a factor of arterial dysfunction through inflammatory response, cell apoptosis, macrophage adhesion, and so on [7, 8, 9, 10, 11]. Not only in biological signaling, the influence of SS can also be observed from a morphological view. The shapes and orientations of ECs have been related to local SSs, and EC layers are elongated in the direction of flow under certain levels of SS [12, 13, 14, 15]. Szymanski et al. observed cell migration toward areas of high SS [16].

Due to the fact that SS distributions on vessel surfaces have various effects on EC activities, changes in SSs due to the presence of medical implants have attracted increasing research attention. Moreover, changes in the flow distributions on vessel surfaces have been shown following stent deployment [17], and several flow simulations have been performed in order to characterize the impacts of stents on flow environments [18, 19, 20].

Although numerous studies have reported the computational fluid dynamics (CFD) for stent design, no published studies reported EC responses around stent wires in vivo. Thus, to investigate the influence of SSs on endothelialization around a stent wire, we exposed EC monolayers to a range of flow pressures using a parallel flow chamber with a straight bare metal wire, and then we analyzed the EC distributions around the wire. Due to the fact that direct observations of flow conditions around the wire were untenable, we inferred flow conditions using CFD and correlated flow patterns with EC activities.

2. Materials and methods

2.1 ECs

Human carotid artery ECs were used at passages 7–9 (cell applications) and two-dimensional cultures were grown to confluence in 35 mm dishes coated with a gelatin solution (Wako Pure Chemical Industries, Ltd.). Proliferation media (PM) comprised Medium 199 (Gibco) containing 20% fetal bovine serum (Sigma), 1% penicillin/streptomycin (AUSTRAL Biologicals), and 0.001% human basic fibroblast growth factor (FGF; AUSTRAL Biologicals), which were used for cell culture and flow-exposure experiments.

2.2 Flow exposure

A pulse damper, flow chambers, a reservoir, and a roller pump with silicone tubes were connected in series to construct a flow loop for flow-exposure experiments (Fig. 1). This pulse damper was used to attenuate pulsation and provide a steady flow through the flow chamber. Carbon dioxide gas (5% CO ${}_{2}$ $+$ 95% air) was infused into the reservoir to maintain the pH of the PM as a working fluid for flow exposure. PM were maintained at 37 ${}^{\circ}$ C in a thermostat tank during flow-exposure experiments.

Figure 1.

A schematic of the flow loop.

Figure 2.

A schematic of the flow chamber. Left shows 3D schematics, and right shows numerical meshes for CFD on a cross-sectional plane of chamber.

SSs were applied to ECs in parallel plate flow chambers (Fig. 2) as similar described previously [21]. A NiTi wire with a cross section of 0.406 $\times$ 0.406 mm was cleaned for 20 min using a UV-O ${}_{3}$ cleaning system, and the wire was then fixed to the bottom of the flow chamber by pinching between two silicone gaskets with thin slits.

SSs were applied at 1, 2, or 3 Pa by adjusting the flow rate of the roller pump to 1.08 $\times$ 10 ${}^{-6}$ , 2.16 $\times$ 10 ${}^{-6}$ , and 3.24 $\times$ 10 ${}^{-6}$ m ${}^{3}$ /s, respectively. The wire was fixed at an angle of 90 ${}^{\circ}$ to the flow direction, and comparisons were made with flow conditions of 2.16 $\times$ 10 ${}^{-6}$ m ${}^{3}$ /s and a wire angle of 60 ${}^{\circ}$ to the flow direction.

2.3 Immunofluorescent staining

After 24 h of flow exposure, ECs were fixed in 4% paraformaldehyde for 15 min at room temperature and were then washed five times in phosphate buffered saline (PBS). In order to permeabilize the EC membranes, cells were treated with 1 mL aliquots of 0.2% Triton X-100 (Roche Applied Science) for 5 min and were then washed five times in PBS. In order to detect F-actin filaments, ECs were incubated with Alexa Fluor ${}^{\@setsize{\scriptsize}{9.5pt}{\viiipt}{\@viiipt}{\textregistered}}$ 546 Phalloidin (Gibco) in the dark for 30 min and were then washed twice in PBS prior to nuclear staining for 3 min with 4’, 6-diamidino-2-phenylindole (Gibco).

Fluorescent images of ECs were generated using an inverted laser scanning microscope system (Olympus), and the upper and side surfaces of the wire were observed. Cell densities were calculated from the number of cells adhered on each surface, and the number of nuclei was determined in 0.1 $\times$ 0.1 mm squares to evaluate cell densities in culture dishes. Cell density on wire surface was also calculated on each side (upstream-side, downstream-side, and upper surface). As the evaluation for endothelialization on wire, $t$ -test was adopted between upstream-side and down-stream side.

2.4 Numerical simulations

To calculate wall SS (WSS) distributions in the flow chamber, we used a numerical model (Fig. 2) and performed simulations using a commercial mesher and solver (ICEM-CFD 14.5 and ANSYS Fluent 14.5, ANSYS Inc.). In these analyses, fluids were assumed to have a viscosity of 0.001 Pa $\cdot$ s and 1,000 kg/m ${}^{3}$ and were analyzed as isothermal, incompressible, and Newtonian liquids. Mass flow rates at the inlet boundary were set at 1.08 $\times$ 10 ${}^{-6}$ , 2.16 $\times$ 10 ${}^{-6}$ , and 3.24 $\times$ 10 ${}^{-6}$ m ${}^{3}$ /s to achieve WSSs of 1, 2, and 3 Pa on the dish area, respectively. The outlet boundary pressure was set at 0 Pa for all computations, and a no-slip boundary was adopted on all wall. Tetrahedral mesh of 0.025 mm was adopted around wire surface region, then total mesh number reaches over 6 million. Mesh doubling test shows flow environment consistent in smaller mesh: Average WSS changes $<$ 3%, and reattachment point of flow separation only shifts 0.2 mm.

3. Results

3.1 Flow field

CFD analyses showed that a flow field in the flow chamber was achieved. However, because of the sudden contractions in the fluid domains, a high SS was observed on the top surface of the wire, and vortex flows appeared upstream and downstream of the wire. Consequently, the upstream side of the wire’s surface was exposed to a lower-drag force in a narrow area near the bottom. Moreover, due to the fact that the vortex on the upstream side was small, most of the upstream side of the wire was exposed to an upper-drag force. In contrast, the downstream surface was exposed to upper-drag forces in all areas (see Fig. 3a).

Figure 3.

CFD results as flow fields around the wire. (a) WSS distributions on the dishes. (b) Velocity vector on a cross-sectional plane of the flow chamber.

In order to accommodate flow separation on the downstream side of the wire, we observed separation or reattachment points, which indicated stagnation and contrasting areas of typical low and high SSs (Fig. 3b). Although the tendency of flow patterns was consistent in all 90 ${}^{\circ}$ models, the position of the reattachment flow on the downstream side moved toward the distal side with increases in the inlet flow. An almost constant SS was observed on the dish surfaces upstream and downstream of the vortex, except for areas near the walls.

As in the 90 ${}^{\circ}$ model, the 60 ${}^{\circ}$ model exhibited vortexes according to the flow separation and reattachment upstream and downstream of the wire. Vortex areas along the wire orientation led to typical SS distributions on the bottom of the dish. However, in the 60 ${}^{\circ}$ model, the high-SS area around the reattachment point diminished toward the area near the outlet.

3.2 Cell distributions on culture dishes

Prior to the 24 h flow exposure, we seeded ECs on 35 mm dishes and cultured cells until confluence over three days to confirm the influence of proliferation. Subsequently, we calculated cell numbers after six days of culture and confirmed no significant increases compared with those under static conditions for three days. We also determined 24 h EC proliferation with and without the wire and showed that although some ECs moved onto wire surfaces, the cell numbers in culture plates did not change significantly.

Figure 4.

Cell distributions on the bottom plane of flow chamber. (a) Fluorescent images of EC layers after 24 h flow exposure (top) and the extracted images of EC nuclei (bottom). (b) ECs around the reattachment point, which is estimated from CFD results, under pressure conditions of 2 Pa. Middle and right images are enlarged view of white square A and B on left image.

Figure 5.

Cell densities on the bottom calculated from the number of EC nuclei.

The fluorescent images in Fig. 4a (top) showed the extraction of cell nuclei (bottom) on the dish, and nuclei are represented as black dots in Fig. 4a (bottom), whereas the black area represents cell concentrations or cell enlargement. In Fig. 4b, the enlarged images show reattachment areas under 2 Pa conditions. In order to decipher the influence of EC sizes, the number of ECs was determined for small domains, and EC densities were calculated (Fig. 5).

Compared with SS distributions in CFD analyses, EC distributions showed similar tendencies in high-SS areas, and at the downstream area, typical concentrations of ECs were observed parallel to the wire in both 90 ${}^{\circ}$ and 60 ${}^{\circ}$ models. In contrast, relatively lower EC densities were observed near areas of high cell concentrations, and these differences in EC distributions increased with greater SSs. In the 60 ${}^{\circ}$ model, EC concentrations decreased toward the downstream end, and the results of CFD analyses showed corresponding decreases in SSs toward the downstream end of the 60 ${}^{\circ}$ model.

Figure 6.

Cell densities on wire surfaces.

Figure 7.

Fluorescent images of ECs on wire surfaces at 2 Pa in the 90 ${}^{\circ}$ model.

3.3 Cell distributions on wire surfaces

Comparisons of cell densities on the upstream and downstream surfaces of the wire (Fig. 6) showed no significant differences ( $P>$ 0.2), whereas the downstream side had higher average cell densities under all flow conditions. However, 3 Pa conditions produced higher cell densities on the wire than other conditions ( $P<$ 0.05). The fluorescent images of ECs on the wire surfaces under 2 Pa conditions (Fig. 7) show cell adhesion only in narrow areas near the bottom at the upstream and downstream sides, and no ECs were observed on the upper surface of the wire.

4. Discussion

Previous in vitro studies investigated EC distribution under flow environment using several types of flow chamber. In these results, flow stimuli, especially WSS induced by viscous fluid, has been considered as one of the factors for cell migration, elongation, and orientation [12, 13, 14, 16, 22, 23]. Though our chamber system uses identical flow domain, CFD results (Fig. 3) indicates that even one wire can reconstruct inner flow environment by its deployment. And flow exposure results (Fig. 4) clearly show the typical EC distribution on all cases: the concentration of EC downstream of the wire. As EC concentration downstream of wire matches the area of flow reattachment (which is indicated by high $\to$ low $\to$ high SS change), this EC distribution can be thought as a consequence of SS change.

In addition to the morphological changes of ECs, shear-induced signaling pathways have been defined for each stage of vascular dysfunction in the presence of various mediators [24, 25]. In particular, the expression levels of the matrix metalloproteinases MMP-2 and MMP-9, which degrade all extracellular matrix components, are enhanced by mechanical stimuli [26, 27]. However, our chamber system comprised a single layer of ECs and failed to elucidate EC-SMC interactions, for which several studies show EC-SMC and SMC-EC signaling pathways [28]. Thus, further studies are required to investigate the interactions between ECs and SMCs using coculture models [14].

Sprague et al. previously showed that cell migration onto stent wires decreases from downstream to the upstream of the direction of flow [29]. Hence, SS environments may affect the speed of endothelialization on implanted wires. Although ECs migrated to wire walls in the present study, no significant differences were observed between the upstream and downstream sides. Accordingly, CFD analyses showed small and large recirculation on the upstream and downstream sides, respectively. Moreover, most upstream walls were exposed to the high SS of the main flow, whereas a small area near the bottom received a downward force during recirculation. In contrast, the downstream wall received an upper force during recirculation in all areas. Because of the time limitations of these experiments, the EC coverage on the wire remained limited to small areas near the bottom, but this indicated that surface treatments of wires may facilitate endothelialization under clinical conditions [30]. Moreover, further studies with longer exposure times may show enhanced effects of SS on EC migration.

Separation of flow on wire surfaces causes reattachment of flow downstream of the wire. Due to the fact that flow stagnates at the reattachment point, SS values vary widely around this point, and we accordingly observed higher cell densities around high-SSG areas (Fig. 5). Moreover, due to the fact that the influence of proliferation is likely small over 24 h, changes in cell environments are likely driven by migration. Thus, ECs were concentrated because of cell migration from around the reattachment area, whereas ECs near the wire migrated from the dish to the wire surfaces, suggesting that migration directions vary depending on the chamber geometry and flow conditions.

Local SS environments can be controlled by wire shapes [17, 18, 20], and flow environments have been successfully engineered previously. Moreover, in studies of medical engineering applications, stents were topologically optimized to minimize flow inside aneurysms [31, 32, 33, 34]. Hence, minimizing the recirculation around stenosis wires will likely yield benefits, because extremely low SSs are thought to promote ECs dysfunction [7, 9, 10, 11]. Moreover, due to the fact that stent expansion leads to the deformation of the vessel lumen, Putra et al. applied simulations of fluid-structure interactions to optimize stent topologies and minimize low-SS areas on vessel surfaces [35]. In addition, previous in vivo and in vitro experiments suggested that low-SS environments induce cell degradation [7, 9, 10, 11], further suggesting that optimizing the implant geometry to minimize low-SS areas could prevent flow-induced inflammation after implant treatments.

The current computational approaches allow the simulation of complicated cell interactions and reproduction of tissue morphology [36]. Such approaches may facilitate the development of stents that can achieve rapid endothelialization, warranting the use of EC simulations to model SS-driven EC migration and optimization of devices based on computer simulations of cell activities.

5. Conclusion

In this study, we showed that EC distributions vary under differing SS conditions (1–3 Pa) using a flow chamber system and a wire obstacle. Under these conditions, vortexes downstream of the wire produced large changes in SSs around areas of flow reattachment. Moreover, after 24 h flow exposure, typical concentrations of ECs were observed toward flow reattachment areas. Due to the fact that extremely low SSs can induce inflammatory responses in ECs, areas of low SS may facilitate the topological optimization of endovascular implants.

Footnotes

Acknowledgments

This research was supported by the Fund for the Promotion of Joint International Research (15KK0197) and a Grant-in-Aid for Scientific Research (A) (16H01805). N.K.P would like to acknowledge LPDP of The Ministry of Finance, Republic of Indonesia for its doctorate scholarship program.

Conflict of interest

The authors have no conflict of interest to report.

References

Byrne

Joner

and Kastrati

. Stent thrombosis and restenosis: what have we learned and where are we going? The Andreas Gruntzig Lecture ESC 2014. Eur Heart J. 2015; 36(47): 3320-3331.

Finn

Nakazawa

Joner

Kolodgie

Mont

Gold

et al. Vascular responses to drug eluting stents – importance of delayed healing. Arterioscl Throm Vas. 2007; 27(7): 1500-1510.

Kim

Park

Seo

and Oh

. Delayed severe multivessel spasm and aborted sudden death after Taxus stent implantation. Heart. 2005; 91(2).

Liu

Roubin

and King

. Restenosis after coronary angioplasty – potential biologic determinants and role of intimal hyperplasia. Circulation. 1989; 79(6): 1374-1387.

Asahara

Chen

Tsurumi

Kearney

Rossow

Passeri

et al. Accelerated restitution of endothelial integrity and endothelium-dependent function after phVEGF(165) gene transfer. Circulation. 1996; 94(12): 3291-3302.

Kipshidze

Dangas

Tsapenko

Moses

Leon

Kutryk

et al. Role of the endothelium in modulating neointimal formation – vasculoprotective approaches to attenuate restenosis after percutaneous coronary interventions. J Am Coll Cardiol. 2004; 44(4): 733-739.

Cebral

Ollikainen

Chung

Mut

Sippola

Jahromi

et al. Flow conditions in the intracranial aneurysm lumen are associated with inflammation and degenerative changes of the aneurysm wall. Am J Neuroradiol. 2017; 38(1): 119-126.

Cuhlmann

Van der Heiden

Saliba

Tremoleda

Khalil

Zakkar

et al. Disturbed Blood Flow Induces RelA Expression via c-Jun N-Terminal Kinase 1 A Novel Mode of NF-kappa B Regulation That Promotes Arterial Inflammation. Circ Res. 2011; 108(8): 950-959.

Dong

Yang

Cao

and Qu

. Low shear stress-induced autophagy alleviates cell apoptosis in HUVECs. Mol Med Rep. 2017; 15(5): 3076-3082.

10.

Kwak

Back

Bochaton-Piallat

Caligiuri

Daemens

MJAP

Davies

et al. Biomechanical factors in atherosclerosis: mechanisms and clinical implications. Eur Heart J. 2014; 35(43): 3013-+.

11.

Meng

Tutino

Xiang

and Siddiqui

. High WSS or Low WSS? Complex interactions of hemodynamics with intracranial aneurysm initiation, growth, and rupture: toward a unifying hypothesis. Am J Neuroradiol. 2014; 35(7): 1254-1262.

12.

Galbraith

Skalak

and Chien

. Shear stress induces spatial reorganization of the endothelial cell cytoskeleton. Cell Motil Cytoskel. 1998; 40(4): 317-330.

13.

Levesque

and Nerem

. The elongation and orientation of cultured endothelial-cells in response to shear-stress. J Biomech Eng-T Asme. 1985; 107(4): 341-347.

14.

Sakamoto

Ohashi

and Sato

. Effect of fluid shear stress on migration of vascular smooth muscle cells in cocultured model. Annals of Biomedical Engineering. 2006; 34(3): 408-415.

15.

Sakamoto

Saito

Han

Ohashi

and Sato

. Effect of spatial gradient in fluid shear stress on morphological changes in endothelial cells in response to flow. Biochem Bioph Res Co. 2010; 395(2): 264-269.

16.

Szymanski

Metaxa

Meng

and Kolega

. Endothelial cell layer subjected to impinging flow mimicking the apex of an arterial bifurcation. Annals of Biomedical Engineering. 2008; 36(10): 1681-1689.

17.

Berry

Santamarina

Moore

Roychowdhury

and Routh

. Experimental and computational flow evaluation of coronary stents. Annals of Biomedical Engineering. 2000; 28(4): 386-398.

18.

Benard

Coisne

Donal

and Perrault

. Experimental study of laminar blood flow through an artery treated by a stent implantation: characterisation of intra-stent wall shear stress. Journal of Biomechanics. 2003; 36(7): 991-998.

19.

Duraiswamy

Frank

and Moore

. Blood flow in stented arteries: a parametric comparison of strut design patterns in three dimensions. J Biomech Eng-T Asme. 2005; 127(4): 637-647.

20.

Jimenez

and Davies

. Hemodynamically driven stent strut design. Annals of Biomedical Engineering. 2009; 37(8): 1483-1494.

21.

Ohashi

and Sato

. Remodeling of vascular endothelial cells exposed to fluid shear stress: experimental and numerical approach. Fluid Dyn Res. 2005; 37(1-2): 40-59.

22.

Wang

Baker

Chen

and Schwartz

. Endothelial cell sensing of flow direction. Arterioscl Throm Vas. 2013; 33(9): 2130-2136.

23.

Yoshino

Sakamoto

and Sato

. Fluid shear stress combined with shear stress spatial gradients regulates vascular endothelial morphology. Integr Biol-Uk. 2017; 9(7): 584-594.

24.

Chatzizisis

Coskun

Jonas

Edelman

Feldman

and Stone

. Role of endothelial shear stress in the natural history of coronary atherosclerosis and vascular remodeling – molecular, cellular, and vascular behavior. J Am Coll Cardiol. 2007; 49(25): 2379-2393.

25.

Sena

Pereira

and Seica

. Endothelial dysfunction – a major mediator of diabetic vascular disease. Bba-Mol Basis Dis. 2013; 1832(12): 2216-2231.

26.

Cheng

Tempel

van der Baan

Krams

Daemen

MJAP

van Haperen

et al. Response to letter regarding article, “Atherosclerotic lesion size and vulnerability are determined by patterns of fluid shear stress”. Circulation. 2006; 114(22): E615-E615.

27.

Wang

Chang

Lin

Kuan

and Shyu

. Induction of matrix metalloproteinases-14 and-2 by cyclical mechanical stretch is mediated by tumor necrosis factor-alpha in cultured human umbilical vein endothelial cells. Cardiovasc Res. 2003; 59(2): 460-469.

28.

Lilly

. We have contact: endothelial cell-smooth muscle cell interactions. Physiology. 2014; 29(4): 234-241.

29.

Sprague

Luo

and Palmaz

. Human aortic endothelial cell migration onto stent surfaces under static and flow conditions. J Vasc Interv Radiol. 1997; 8(1): 83-92.

30.

Hayakawa

Nagai

Yoshinari

Makimura

and Nemoto

. Cell-adhesive protein immobilization using tresyl chloride-activation technique for the enhancement of initial cell attachment. Journal of Oral Tissue Engineering. 2004; 2(1): 14-24.

31.

Anzai

Falcone

Chopard

Hayase

and Ohta

. Optimization of strut placement in flow diverter stents for four different aneurysm configurations. J Biomech Eng-T Asme. 2014; 136(6).

32.

Janiga

Daroczy

Berg

Thevenin

Skalej

and Beuing

. An automatic CFD-based flow diverter optimization principle for patient-specific intracranial aneurysms. Journal of Biomechanics. 2015; 48(14): 3846-3852.

33.

Lee

Srinivas

and Qian

. Three-dimensional hemodynamic design optimization of stents for cerebral aneurysms. P I Mech Eng H. 2014; 228(3): 213-224.

34.

Zhang

Anzai

Chopard

and Ohta

. Towards the patient-specific design of flow diverters made from helix-like wires: an optimization study. Biomed Eng Online. 2016; 15.

35.

Putra

Palar

Anzai

Shimoyama

and Ohta

. Comparative study between different strut’s cross section shape on minimizing low wall shear stress along stent vicinity via Surrogate-based optimization. Advances in Structural and Multidisciplinary Optimization. 2017; 2097-2109.

36.

Alt

Ganguly

and Salbreux

. Vertex models: from cell mechanics to tissue morphogenesis. Philos T R Soc B. 2017; 372(1720).