Biomolecular Sensing Processing and Analysis - Rashid Bashir and Steve Wereley

[13]P.D Maker, D.W.Wilson, and R.E. Muller. Fabrication and Performance of Optical Interconnect Analog Phase Holograms made by E-beam Lithography. In R.T. Chen and P.S. Guilfoyle (eds.), Optoelectronic Interconnects and Packaging, Proc. SPIE CR62, pp. 415–430, 1996.

[14]D. Modarress, D. Fourguette, F. Taugwalder, M. Gharib, S. Forouhar, D.Wilson, and J. Scalf. Design and Development of Miniature and Micro Doppler Sensors. 10th International Symposium on Application of Laser Techniques to Fluid Mechanics, Lisbon, Portugal, 2000.

[15]D. Modarress and D.A. Johnson. Investigation of Turbulent Boundary Layer Separation Using Laser Velocimetry. AIAA J., 17(7):1979.

[16]D. Modarress and D. Tan. Application of LDA to Two-phase Flows. Exp. Fluids, 1:1983.

[17]A.A. Naqwi and W.C. Reynolds. Dual Cylindrical Wave Laser Doppler Method for Measurement of Skin Frictionin Fluid Flow, Report No. TF-28, Stanford University, 1987.

[18]J. Turunen and F.Wyrowski (Eds.). Diffractive Optics for Industrial and Commercial Applications, John Wiley & Sons, 1998.

[19]D.W. Wilson, P.D. Maker, and R.E. Muller. Binary Optic Reflection Grating for an Imaging Spectrometer.

Diffractive and Holographic Optics Technology III, SPIE Proceedings, Vol. 2689, Jan. 1996.

[20]D.W.Wilson, J.A. Scalf, S. Forouhar, R.E. Muller, F. Taugwalder, M. Gharib, D. Fourguette, and D. Modarress. Diffractive optic fluid shear stress sensor. Diffractive Optics and Micro Optics, OSA Technical Digest Optical Society of America, Washington DC, pp. 306–308, 2000.

[21]D.W. Wilson, P.K. Gogna, R.J. Chacon, R.E. Muller, D. Fourguette, D. Modarress, F. Taugwalder, P. Svitek, and M. Gharib. Diffractive Optics for Particle Velocimetry and Sizing. Diffractive Optics and Micro Optics, OSA Technical Digest, Optical Society of America, Washington DC, pp. 11–13, 2002.

[22]D.W. Wilson, P.D. Maker, R.E. Muller, P. Mouroulis, and J. Backlund. Recent Advances in Blazed Grating Fabrication by Electron-beam Lithography. Paper 5173–16, Proc. SPIE, 2003.

[23]Y. Yeh and H.Z. Cummins. Localized Fluid Flow Measurements with a He-Ne Laser Spectrometer. Appl. Phys. Lett., 4:176–178, 1964.

18

Vascular Cell Responses to Fluid Shear Stress

Jennifer A. McCann, Thomas J. Webster, and Karen M. Haberstroh

Weldon School of Biomedical Engineering, Purdue University, West Lafayette, IN, 47907

ABSTRACT

The development of several vascular diseases is linked to both blood flow properties and cellular behavior in the arterial and venous systems. For instance, atherosclerosis development is dependent on the blood flow profile, shear stress rate, and resulting cellular responses in the arteries. Specifically, in regions of disturbed flow behavior, cells demonstrate both altered morphology and phenotype. Based on this clinical knowledge, in vitro fluid flow studies have been performed on vascular endothelial and smooth muscle cells to understand the process of disease initiation and development. Ultimately, results of such studies will provide knowledge regarding key pathways involved in disease progression. Moreover, this information will be critical when designing effective drug therapies in the clinical setting.

18.1. INTRODUCTION

Cardiovascular diseases (coronary heart disease, stroke, hypertension, valvular heart disease, and diseases of the vessels) affect approximately 62,000,000 Americans annually (American Heart Association). One of the most predominant forms of vascular disease and the leading cause of death in the developing world [47] is atherosclerosis, or hardening of the arteries, which results in a narrowed vessel lumen due to an accumulation of lipids, fibrous elements, and smooth muscle cells [45, 68, 91]. In this pathological condition, the cells and mechanical properties of the affected vessel, as well as the blood flow properties through the affected region, are altered as a direct result of plaque formation.

Two hypotheses regarding the relationship between blood flow and atherosclerosis initiation have received much attention. The first states that higher than normal shear stresses

FIGURE 18.1. Oil-Red-O staining shows lipid-rich atherosclerotic lesions in the thoracic aorta arch of a Watanabe heritable hyperlipidemic rabbit. (Reprinted from Molecular Medicine Today, Vol 5, Topper and Gimbrone, Blood flow and vascular gene expression: fluid shear stress as modulator of endothelial phenotype, 40–46, Copyright 1999, with permission from Elsevier)

cause an injury to the endothelial cell layer and induce smooth muscle cell proliferation [69], while the second states that regions with lower than normal shear stresses and disturbed flows (e.g., regions with flow separation, non-laminar flow patterns, etc.) are more prone to plaque initiation [7, 15, 47].

Most research supports the second theory; it has been observed that flow streamlines separate at branches, curvatures, and bifurcations of large arteries, resulting in complicated flow patterns. For example, Figure 18.1 demonstrates the formation of lesions at a bifurcation in the thoracic aorta of a Watanabe heritable hyperlipidemic rabbit. In addition, the abdominal aorta develops atherosclerotic lesions that are located mainly along the posterior wall of the aortic bifurcation and extend proximally towards the renal arteries [7]. A flow study in an anatomically correct model revealed that flow patterns within the abdominal aorta distal to the renal arteries had regions of low shear stress and flow disturbances, with maximal velocities at the anterior wall and flow reversal occurring at the posterior walls [51]. Moreover, an in vivo study (performed with human autopsy subjects) showed a direct relationship between areas of low wall shear stress and atherosclerotic plaques in the abdominal aorta [61].

Corresponding with these defined flow patterns is the fact that endothelial cells within these specific vessel regions experience very different shear exposure than cells just microns away; these cells become proatherogenic, inducing a cascade of pathological events. The effects of atherosclerosis on endothelial and smooth muscle cell functions are discussed in further detail below.

18.1.1. Vessel Physiology

Arteries are thick, muscular, tubular structures, which serve primarily as nonthrombogenic conduits for oxygenated blood [68, 91]. The arterial wall can be divided into three

circumferential layers: the intima, the media, and the adventitia. The intima comprises the innermost layer of arterial walls; it lies inside the internal elastic lamina and is directly exposed to blood flow [91]. It contains a lining of longitudinally oriented endothelial cells (which lay on the internal elastic lamina), connective tissue, and a few subendothelial smooth muscle cells [68, 91]. The media is the middle layer of arterial walls and is bounded by the internal and external elastic lamina. It is composed primarily of circumferentially oriented smooth muscle cells, which control vessel diameter and are embedded in an extracellular matrix of elastin and collagen [68, 91]. In larger arteries, elastic lamellae are also present. The amount of each component as well as the arrangement of this layer varies along the arterial tree, thus providing the necessary elasticity and mechanical properties required by the specific vessel. The adventitia is the outermost layer of arterial walls and lies outside the external elastic lamina [91]. It is made of fibrous connective tissue, with elastic fibers and fibroblasts as well. Most of its support and strength is provided by collagen. Furthermore, this layer is penetrated by small blood vessels (the vasa vasorum), which provide nutrients to the inner layers of the vessel wall [68, 91]. Finally, blood flow occurs in the lumen, or the hollow center of the blood vessel.

As was previously mentioned, endothelial cells, smooth muscle cells, fibroblasts, and macrophages are found within the various layers of the arterial wall. While each participates in homeostatic mechanisms, endothelial cells and smooth muscle cells are most actively involved in these processes.

Endothelial cells form the principal layer between blood elements and the arterial wall and actively participate in vessel homeostasis. They are thin, flat, and elongated cells, with prominent nuclei, many mitochondria, extensive endoplasmic reticulum, and Golgi apparatus [68, 91]. These cells rest on a basement membrane made primarily of collagen type IV [91]. Endothelial cells form a selective permeability barrier between the blood and underlying areas [68]. These cells produce nitric oxide (NO), prostacyclin (PGI2), and other compounds which promote vasodilation and inhibit platelet aggregation, smooth muscle cell proliferation, and monocyte adhesion [32, 56]. Furthermore, these substances control the growth, differentiation, and function of smooth muscle cells and macrophages [56, 68].

Smooth muscle cells are primarily located in the media; they modulate vessel function by controlling vessel wall tone in response to mechanical and neural stimuli [64, 91]. These cells are spindle shaped with elongated nuclei and are characterized by a basal lamina, which surrounds individual cells. Smooth muscle cells can take on two forms: contractile or synthetic. The contractile phenotype contains many cytoplasmic filaments with less mitochondria and endoplasmic reticulum. The synthetic phenotype is characteristic of proliferating smooth muscle cells; proliferating cells have fewer contractile filaments, an increased amount of synthetic organelles, and an increased production of secretory proteins [64, 91]. Furthermore, in this phenotype, smooth muscle cells are much more responsive to mitogens. Current research supports the idea that multiple subpopulations of smooth muscle cells exist within the media, each with distinct developmental lineages and functional characteristics [27]. This is not surprising given the fact that smooth muscle cells must perform a wide range of functions, often at the same time. For instance, one group of cells may focus on maintaining vessel wall tone while another group is involved in damage repair. In this capacity, smooth muscle cells within different vessel areas may respond appropriately to local mechanical or chemical factors.

18.1.2. Vessel Pathology

In healthy vessels, cells function to provide a smooth, nonthrombogenic environment for blood flow. In contrast, endothelial dysfunction is characteristic of diseased vessels; this is an important factor in the development of atherosclerosis and its resulting disturbed flows. Specifically, injury to the endothelial cell layer (ranging from cell denudation to alterations in normal cell functions like expression of new surface factors, release of growth factors, etc. [91]) contributes to the initiation of lesion formation. For instance, when injured, endothelial cells express leukocyte adhesion molecules such as vascular cell adhesion molecule (VCAM)-1 and intercellular adhesion molecule (ICAM)-1 [24, 32, 43]; their involvement in recruiting monocytes and leukocytes suggest a potential role for these molecules in the development of atherosclerosis. While endothelial cells normally participate in lipid uptake, this process is enhanced in the diseased state and results in fatty streak (and ultimately plaque) formation. Finally, one of the most important endothelial cell responses to injury is decreased production of nitric oxide, which leads to impaired vasodilation and increased systemic blood pressure.

In the atherosclerotic plaque, smooth muscle cells become activated by growth factors and cytokines (e.g., platelet derived growth factor (PDGF)-β, interleukin (IL)-1) released by macrophages, platelets, and other cells. Subsequently, smooth muscle cells produce and secrete extracellular matrix components, proliferate at increased rates, migrate into the vessel lumen, and have altered gene expression [32]. The extracellular matrix components these cells release (e.g., collagen, elastin) make up the fibrous cap of the plaque. While this was first thought to be part of the natural progression of atherosclerosis, some now believe this to be a defense mechanism against the progression of the disease since the cap serves to cover the more dangerous lipid core [24]. It seems that plaques with more smooth muscle cells compared to the number of inflammatory cells rupture less often than those with a larger number of inflammatory cells within the lipid core.

While endothelial and smooth muscle cells are key components of disease development and progression, these cells do not act alone in the vessel, rather, atherosclerosis is often thought of as an inflammatory disease critically affected by the role of monocytes and other inflammatory cell types. For instance, injured endothelial cells express E and P selectins on their surface in the atherosclerotic state; expression of these adhesion molecules leads to a ‘sticky’ vessel wall and mediates rolling of inflammatory cells along the endothelial cell layer [13, 32, 68]. Once attached, these inflammatory cells are able to migrate between the endothelial cells and into the intima, where they differentiate into macrophages.

Monocyte-derived macrophages are present in large numbers in atherosclerotic lesions [32]; these cells are the major source of foam cells and the most predominant cell in the fatty streak [32, 68]. Furthermore, they secrete biologically active products which affect endothelial and smooth muscle cell function [32]. For instance, activated macrophages secrete IL-1, which acts on smooth muscle cells to increase their rate of proliferation.

As a result of this cellular cascade, atherosclerotic vessels become hardened and possess narrowed vessel lumens; each of these factors ultimately affects blood flow. For example, the vessel surface becomes rough as plaques protrude into the blood, which leads to thrombus formation and further blocks blood flow. In addition, a reduced luminal area results in increased blood velocity through the diseased vessel. Furthermore, atherosclerotic vessels

lose their distensibility and are easily ruptured due to the hemodynamic forces associated with blood flow.

18.2. HEMODYNAMICS OF BLOOD FLOW

Before discussing studies aimed at understanding cellular responses to flow, a brief discussion of fluid flow through the circulatory system is required. The circulatory system is a closed loop flow system, with two circulations (pulmonary and systemic) arranged in series. The pulmonary circulation delivers deoxygenated blood from the right side of the heart to the lungs and then returns it to the left side of the heart, and the systemic arteries and arterioles deliver oxygenated blood from the left side of the heart to the body tissues under high pressures. Nutrient and gas exchange occurs at the tissue level in the capillary beds. Finally, the systemic veins transport deoxygenated blood back to the heart under low pressures.

The circulatory system itself is quite complex, thus precise mathematical representation of the system is difficult. The most obvious considerations are: the heart which is a non-simple pump, the vessels which are different sizes and possess diverse mechanical properties, and the blood which is a non-homogeneous fluid. Furthermore, the vessels of the circulatory system are not inert tubes, rather they are constantly remodeling to accommodate altered pressures and flow rates. Despite these problems, there are some simple relationships that are used to describe blood flow throughout the body. The primary considerations in the rate of blood flow through a vessel are the pressure between the two points of flow, the resistance which the vessel provides to flow, and the vessel area.

Based on these important parameters, three equations relating blood flow rate to vessel area, pressure, and resistance are:

where Q is the volumetric flow rate, ν is the blood flow velocity, A is the cross-sectional area of the vessel, P is the pressure gradient between the two points of interest, r is the vessel radius, µ is the viscosity of the blood, l is the distance between the two points, and R is the resistance between the two points [10]. Note that Equation (18.3) is Poiseuille’s Law.

In addition to these relationships, the unitless Reynolds number (NR ) is a useful quantity for predicting flow conditions:

where ρ is the fluid density and D is the vessel diameter. Laminar flow is typical of healthy vessels and occurs when NR is below 2000; such flow is characterized by streamlines parallel to the vessel axis. A value of NR above 3000 is indicative of turbulent flow, with irregular fluid motion and mixing. This type of flow is much less efficient than laminar flow and is

FIGURE 18.2. Schematic diagram of the forces resulting from blood flow which act on an artery wall. (Reprinted and adapted from New Surgery, Vol 1, Araim, Chen, and Sumpio, Hemodynamic forces: effects on atherosclerosis, 92–100, Copyright 2001, with permission from Landes Bioscience)

characteristic of diseased vessels. NR values between 2000 and 3000 describe transitional flow.

As a result of blood flow, vessel walls are continuously subjected to a range of mechanical stimuli. The most predominant forces experienced by the arteries are shear stress, pressure, and cyclic strain or stretch due to the nature of blood flow [7, 40, 83]. These forces (shown schematically in Figure 18.2) affect the vessel physically, while also altering the function and signaling pathways of arterial cells. Therefore, the involvement of each of these mechanical forces has been investigated with regard to the maintenance of vessel homeostasis and the development of vascular diseases.

Perhaps the most widely studied force in the vascular system is shear stress, which results from the friction created as blood flows parallel to the vessel wall. Shear stress has been shown to modulate vessel homeostasis and cellular functions in many ways, including altered endothelial cell morphology and function; Figure 18.3 is a representation of the various endothelial cell responses to both laminar and atherogenic flows. The expression for wall shear stress, τ (measured in dynes/cm2), can be obtained based on Poiseuille’s Law:

As stated above, µ is the viscosity of the blood, Q is the volumetric flow rate, and r is the vessel radius [10]. Normal shear stress values in the venous system range from 1 to 6 dynes/cm2, and in the arterial system from 10 to 70 dynes/cm2 [47]. More importantly, abnormal levels of shear stress are directly correlated with the development of cardiovascular diseases. Though not as widely studied, vessel pressure and strain also contribute to cellular responses and overall vessel health. Hydrostatic pressure within a

18.3.2. Parallel Plate Flow Chamber

The original parallel plate flow chamber consisted of a carefully milled polycarbonate plate, a rectangular Silastic gasket (which creates the channel depth), and a glass slide to which cells are attached [25]. A vacuum at the periphery of the slide held the system together, while also ensuring a uniform channel depth. The chamber was located between two fluid reservoirs, and the flow rate was controlled by a fluid pressure differential. Namely, a constant fluid flow was maintained by supplying fluid to the top reservoir at a higher rate than fluid flowed over the cells. Any excess fluid drained down the overflow manifold into the bottom reservoir. The polycarbonate plate had two ports: an entrance and exit for fluid flow, thereby allowing for continuous flow. The dimensions of the original chamber were 0.022 cm channel height and 2.5 cm channel width. In this chamber geometry, the shear stress over the cells can be easily calculated for a given flow rate, Q (cm3/s), using the following equation for plane Pouiselle flow:

width, and τ (dynes/cm ) is the wall shear stress [25]. This corresponded to a shear stress of roughly 10 dynes/cm2 for a flow rate of 0.2 cm3/s.

The parallel plate flow chamber is the most widely used technique for studying shear stress effects on cells; therefore this system has undergone several modifications. For instance, one version consists of a machine milled polycarbonate plate with two ports for fluid flow, a glass slide to which the cells were attached, and a machined top (Figure 18.5; [48]). The polycarbonate plate possessed two milled recesses: one created the flow channel and the other served as a shelf where the glass cover slide rested. The height difference between the two recesses determined the channel height. Screws located on the chamber periphery held the entire chamber together. The specific dimensions of the flow chamber used in these studies were as follows: 3.429 cm width (b), 0.03302 cm height (h), and 7.4 cm length (l). As in the original parallel plate flow chamber, fluid was delivered to the top reservoir at a faster rate than the fluid flow over the cells, thereby providing continuous flow. Additionally, a valve located prior to the plate inlet was used to control the flow rate. In another parallel plate flow chamber design used by Nauman et al. [55], an adjustable flowmeter was introduced at the chamber outlet to more precisely control flow (as opposed to changing the height difference between the upper and lower reservoirs).

In all parallel plate studies, flow is assumed to be steady, laminar, and uniform over the length of the plate. Therefore, all cells in the chamber are thought to be exposed to the same flow conditions regardless of their location. However, when dealing with modified chambers and dimensions on the order of microns to hundreds of microns, standard machining tolerances (which occur in even the most carefully machined and assembled parallel plate flow chambers) can cause non-negligible variations in the chamber height and local shear stresses. Sensitive cell responses, like those involving gene expression, may not be consistent over the chamber area in the environment of these non-uniform flows. Therefore, flow characterization studies are often used to determine whether flow is uniform in such modified parallel plate flow chambers and, in turn, whether these changes in flow uniformity influence cell responses.