Chapter 9

Static and Dynamic Interactions

between Endothelium and
Circulating Cells in Cancer
Roxana Chotard-Ghodsnia
, Agn`es Drochon , Alain Duperray , Anne
Leyrat
, and Claude Verdier

Biomecanique et Genie Biomedical, CNRS (UMR 6600), UTC, Compi`egne
(France),  Laboratoire de Migration Cellulaire et Infiltration Tumorale (EA2942
INSERM), Institut Albert Bonniot, La Tronche cedex (France),
Laboratoire de
Rheologie, UJF-INPG, CNRS (UMR 5520), BP 53, Grenoble cedex 9 (France)

9.1 Introduction
9.2 Receptors Involved in Interactions between Endothelium and Leukocytes
or Tumour Cells
9.2.1 Selectins
9.2.2 Integrins
9.2.3 Immunoglobulin Superfamily
9.2.4 Cadherins
9.3 Leukocyte or Tumour Cell Adhesion under Flow Conditions
9.3.1 Multistep Process of Leukocyte Adhesion
9.3.2 Adhesive Interactions between Cancer Cells and Endothelium
9.4 In Vitro Devices to Study Circulating Cell-Endothelial Cell Adhesion
9.4.1 Static Assays
9.4.2 In Vitro Flow Assays
9.5 In Vitro Flow Studies of Circulating Cell-Endothelium Adhesion
9.6 Modelling of Circulating Cell-Endothelium Interactions
9.6.1 Experimental Modelling
9.6.2 Mathematical Modelling
9.7 Conclusion
9.8 References

2003 CRC Press LLC

9.1

Introduction

Cellular adhesion to vascular endothelium in the fluid dynamic environment of

the circulation is an important aspect of many physiological and pathological processes. Examples include leukocyte adhesion during recruitment to a site of tissue
injury and cancer cell adhesion during metastasis.
Circulating leukocyte emigration from vasculature into tissues (leukocyte extravasation) is the central event in inflammation. Through extensive studies in recent
years, it has become clear that at least four distinct steps act in sequence to regulate
leukocyte extravasation [1]:

selectin-carbohydrate mediated initial leukocyte tethering and rolling;

activation of integrins on leukocyte surfaces;
transition from rolling to firm adhesion of leukocytes on endothelial cells; and
migration of leukocytes through interendothelial junctions (transendothelial
migration or diapedesis) to extravascular tissue space, following the guidance
of chemo-attractants.
To metastasise, tumour cells must shed into the blood stream (intravasation)
directly by invasion into the tumour-derived vasculature or indirectly by lymphatic
drainage, survive in the circulation, and finally migrate through normal vascular endothelium (extravasation) and proliferate in the target organs. Tumour cell extravasation plays a key role in tumour metastasis. It has been proposed in the literature
that adhesion of circulating tumour cells to the endothelium, mediated by specific
ligand-receptor interactions, is an essential prerequisite step for extravasation to occur. However, the precise mechanisms by which tumour cells penetrate the endothelial cell junction remains one of the least understood aspects of extravasation [2].
Inside the body, the endothelium is continuously subjected to flow induced mechanical stress. These forces induce biochemical signals which can alter the surface
expression of adhesion molecules and, therefore, influence the endothelial monolayers ability to bind circulating cells. These cells contact the endothelium at a rate
dependent upon the fluid dynamics and adhere through receptor-ligand bonds with a
probability that varies with the flow rate. Thus, mechanical stresses and strains play
important roles in interactions between endothelium and circulating cells.
Due to difficulties in characterising both hemodynamic forces acting on circulating cells and expression level of adhesion molecules on vascular endothelium in
vivo, different flow devices have been widely used to study such interactions under
controlled flow conditions. These include in vitro flow devices such as cone-andplate rheometers and parallel-plate flow chambers.
In this chapter, we first describe adhesion molecules involved in interactions between endothelium and leukocytes or tumour cells. We then discuss the in vitro devices to study cell-endothelium interactions with particular emphasis on in vitro flow

2003 CRC Press LLC

devices used to simulate flow conditions in blood vessels. Finally, we describe experimental and mathematical modelling of circulating cell-endothelium interactions
under flow. Understanding the complex interplay among blood flow, cell adhesion,
and vascular biology at the molecular level is crucial for developing new therapeutic
approaches against pathological inflammation and tumour metastasis.

9.2

Receptors Involved in Interactions between

Endothelium and Leukocytes or Tumour Cells

Cell-cell interactions and cell-matrix interactions are mediated by cellular adhesion molecules (CAMs), a diverse group of glycoproteins expressed at the surface
of every cell in the body. These CAMs selectively bind to one another, and play a
critical role in various functions of cellular organisms, such as tissue development
and cohesion, wound healing, cell migration, inflammation, and cancer metastasis.
The recruitment of leukocytes into sites of inflammation involves a cascade of sequential events controlled by the interaction between adhesion molecules expressed
by leukocytes and by the endothelium. Cell migration across endothelial monolayers
involves leukocyte adherence to the endothelium, crawling on the endothelial surface
and penetration between endothelial clefts. This process of leukocyte diapedesis has
been extensively studied and may serve as a paradigm for the mechanisms involved
in tumour cell arrest and extravasation. Several adhesion receptors belonging to different families including integrins, selectins, immunoglobulin-like molecules, and
cadherins (Figure 9.1) have been shown to participate in this mechanism [3]. In the
current model, selectins are implicated in the initial rolling, while adhesion receptors
from the integrin family and the immunoglobulin superfamily are involved in the firm
attachment, flattening, and extravasation of leukocytes; cadherins, and more specifically the endothelial specific VE-cadherin, are involved in the control of endothelial
cell junctions.

9.2.1

Selectins

The selectin family comprises three proteins [4]: E-selectin (CD62E), L-selectin
(CD62L), and P-selectin (CD62P). They all contain a lectin domain, an epidermal
growth factor domain, and a variable number of short consensus repeats of 60 amino
acids present in the complement regulatory proteins. E- and P-selectins are expressed
on endothelial cells, while L-selectin expression is restricted to leukocytes. Selectins
bind to carbohydrate ligands via their lectin domain. It has been shown that tetrasaccharides sialyl-Lewis
and sialyl-Lewis (sLe
, sLe ) are ligands for all the three
selectins [5].

2003 CRC Press LLC

Selectin

Immunoglobulin
Immunoglobulin

Immunoglobulin

Integrins

b
cadherins
cadherins
cytoplasm

Intercellular space
Cell membrane

cytoplasm
Cell membrane

Figure 9.1

Adhesion molecules involved in leukocyte diapedesis. Selectins, expressed by

leukocytes or endothelial cells, interact with glycosylated ligands. Members of
the immunoglobulin superfamily, like PECAM-1, can bind to another molecule
of PECAM-1 at cell-cell junctions. ICAM-1, which is also a member of the
immunoglobulin superfamily expressed by endothelial cells, is a ligand for
leukocyte integrins. VE-cadherin, expressed at interendothelial junction participates in the regulation of vascular permeability.

9.2.1.1

E-selectin

E-selectin (CD62E) is a 115 kD glycoprotein, only expressed on endothelial

cells after activation by interleukin-1 (IL-1), tumour necrosis factor-
(TNF-
), or
bacterial endotoxin such as lipopolysaccharides (LPS). After endothelial cell stimulation, newly synthesised E-selectin is rapidly detected with a maximal surface expression after 3 to 6 hours and a return to basal levels within 24 hours. This rapid
downregulation, has been explained by the release of a soluble form of E-selectin,
and internalisation of the molecule. This regulation of E-selectin expression might
be crucial to control leukocyte accumulation in inflammatory responses. Monoclonal
antibodies specific for E-selectin have been shown to inhibit leukocyte transmigration. It has been suggested that binding of leukocytes to E-selectin on activated en-

2003 CRC Press LLC

dothelium upregulates CD11b (Mac-1) on the leukocytes, and induces an increased

adhesion through an ICAM-1/Mac-1 interaction (see also Sections 9.2.2 and 9.2.3).
9.2.1.2

P-selectin

P-selectin (CD62P) is a single chain glycoprotein of 140 kD, expressed in

platelets and endothelial cells and stored in intracellular organelles. After activation, P-selectin is mobilised to the external plasma membrane within minutes. This
increase in P-selectin expression is transient, and the protein is rapidly internalised
inside the cell, where it is degraded or recycled. P-selectin is also upregulated transcriptionnally by TNF-
. The principal ligand for P-selectin is PSGL-1 (P-selectin
glycoprotein ligand-1), expressed on all leukocytes.
9.2.1.3

L-selectin

L-selectin (CD62L), a 80-100 kD protein, is expressed specifically by a majority

of leukocytes, and interacts with glycosylated endothelial counter-receptors.
Inhibitory experiments with antibodies as well as knock out mice have demonstrated that these selectins are important for the initiation of rolling and adhesion
of leukocytes. Evidence that selectins are involved in metastasis comes from the
fact that sialyl-Lewis
and its isoform sialyl-Lewis have been found on different
types of carcinomas, and that cells expressing these molecules can bind to activated
endothelium [6].

9.2.2

Integrins

Integrins are transmembrane glycoproteins composed of two chains named

and , which are noncovalently linked to each other and form an extracellular binding pocket specific to a given ligand. Both chains are needed for ligand binding. The

chain family is composed of 18 different members, and the family consists of 8
members, but only 24 different types of integrins have been identified. The intracellular domain of integrins binds to the cell cytoskeleton. Integrins are characterised
by low affinity constants for their ligands. This supposes that they must cooperate
with each other in order to produce strong adhesion forces. This cooperation lies on
the formation of clusters at the surface of the cell membrane, where a high concentration of intracellular proteins involved in linking the integrins to the cytoskeleton
is also found. These clusters, known as focal adhesion points, allow the cells to attach tightly to their substrate. A more detailed description of integrins is provided in
Chapter 8.
Integrins expressed by leukocytes and tumours cells are involved in the interacintegrins are exclusively expressed on leukocytes.
tion with endothelial cells.
integrins include four different heterodimers CD11a/CD18 (LFA-1 for lymphocyte
function-associated antigen-1), CD11b/CD18 (Mac-1), CD11c/CD18 (p150,95), and
(CD18) molecule results in
CD11d/CD18. A mutation in the gene encoding the

2003 CRC Press LLC

a genetic disorder: leukocyte adhesion deficiency (LAD). LAD patients show recurrent bacterial infections due to a defect in the effective recruitment of leukocytes in
response to infections, demonstrating the importance of
integrins in the inflammation process.
LFA-1 and Mac-1 are strongly involved in leukocyte adhesion to endothelium
by binding to ICAM (see Section 9.2.3) expressed by endothelial cells. Functional
analysis has demonstrated that CD11a/CD18 (LFA-1) is critical in neutrophil transmigration, and important in transmigration of other leukocyte subtypes. CD11b
(Mac-1) antibodies alone are not as potent an inhibitor of neutrophil transmigration in vitro but they add significantly to the effect of CD11a antibodies in vitro. In
vivo both CD11a and CD11b monoclonal antibodies can reduce inflammation, suggesting both redundancy and a dependence on the inflammatory stimulus and organ
involved. This redundancy is also suggested by the fact that CD11b knock out mice
show no defect in leukocyte transmigration [7].
The most important member of the
integrin subfamily on leukocytes is VLA4 (Very Late Antigen-4, CD49d/CD29). VLA-4 is expressed by most leukocytes and
binds to fibronectin and the immunoglobulin superfamily member VCAM-1 (Vascular Cell Adhesion Molecule-1). VCAM-1 binding occurs with approximately four
times greater affinity than binding to fibronectin.
On tumour cells, the expression panel of integrins is often modified when compared to normal cells, but their specific involvement in cell extravasation is not clear
at the moment. However, over-expression of
v 3 [8] which interacts with PECAM1 (Platelet Endothelial Cell Adhesion Molecule-1, see Section 9.2.3), and
4 1 [9]
which interacts with VCAM-1, on invasive tumour cell participate in adhesion to
endothelium.

9.2.3

Immunoglobulin Superfamily

Many cell adhesion molecules contain one or more immunoglobulin-like domains and have been classified into the immunoglobulin superfamily. These receptors are involved in both homophilic and heterophilic interactions, and play a
major role in the interactions of circulating cells with endothelial cells. Three immunoglobulins which are particularly important in the cascade are intercellular adhesion molecule-1 (ICAM-1) or CD54, vascular cell adhesion molecule-1 (VCAM-1)
and platelet endothelial cell adhesion molecule-1 (PECAM-1).
ICAM-1 (CD54) is a single chain membrane glycoprotein of 80-115 kD, with
five immunoglobulin-like repeats in its extracellular domain [10]. ICAM-1 is moderately expressed on resting endothelial cells, but release of cytokines at sites of
inflammation and immune response such as TNF-
, IL-1 or IFN- results in augmented cellular expression of ICAM-1. ICAM-1 is a ligand for leukocyte integrins
CD11a/CD18 (LFA-1) and CD11b/CD18 (Mac-1). The interaction between Mac1/LFA-1 (CD11a b/CD18) and endothelial ICAM-1 is a well documented adhesion
pathway, which plays an important role in the adhesion and extravasation of leuko-

2003 CRC Press LLC

cytes. ICAM-1 is also a receptor for the major group of rhinoviruses and the malaria
trophozoite plasmodium falciparum. It has been recently shown that fibrinogen is a
ligand for ICAM-1, and that fibrinogen binding to ICAM-1 results in an enhanced
adhesion of leukocytes to HUVEC monolayers, and an increase in transendothelial
migration [11].
VCAM-1 (CD106) is a transmembrane glycoprotein of 110 kD expressed only
on cytokine-activated endothelium. VCAM-1 is a ligand for
4 1 (VLA-4) and

4 7 integrins. VLA-4 binds VCAM-1 through the first and the fourth immunoglobulin domain. Using monoclonal antibodies, several studies have shown that VCAM1 is involved in the transmigration of monocytes and eosinophils, but its involvement
in lymphocyte transendothelial migration remains to be clarified.
PECAM-1 (CD31) is a 130 kD glycoprotein expressed on endothelial cells,
platelets and some leukocytes. CD31 is constitutively expressed on endothelial cells,
and its expression is not modified by cytokines. PECAM-1 can interact both with
itself in a homophilic interaction and with other molecules in a heterophilic interaction. Molecular cloning studies have shown that CD31 is composed of six extracellular immunoglobulin-like domains, a short transmembrane region, and a relatively long cytoplasmic tail of 118 amino-acid containing potential sites for posttranslational modifications. In endothelial cells, CD31 is localised at intercellular
junctions, and thus plays a role in the control of vascular permeability. The high level
of constitutive expression of PECAM-1 in endothelial cells suggests that its function might be regulated, and phosphorylation of the cytoplasmic domain has been
demonstrated. PECAM-1 is directly involved in the process of leukocyte diapedesis
between endothelial cells, as demonstrated by inhibition studies using anti-PECAM1 monoclonal antibodies and soluble recombinant PECAM-1. Leukocytes blocked
in transmigration by anti-PECAM-1 antibodies remain attached to the endothelium,
clearly implicating PECAM-1 in diapedesis rather than in adhesion. PECAM-1 may
participate in tumour cell intravasation by controlling inter-endothelial junctions. In
addition, PECAM can interact with the integrin
v 3 [12] which, as already mentioned, is over-expressed at the surface of invasive cancer cells. This interaction
could contribute to their adhesion to endothelial cells.

9.2.4

Cadherins

The calcium dependent cell adhesion molecules (cadherins) are so called because they have both adhesion and calcium binding sites. This family is described in
detail in Chapter 8. Evidence for a role of this family of adhesion receptors in tumour
cell extravasation is scant. However, the endothelial specific VE-cadherin, which has
been involved in the control of the intercellular endothelial junctions [13] and thus
the control of leukocyte diapedesis, may also play a role in controlling tumour cell
extravasation [14].

2003 CRC Press LLC

9.3

Leukocyte or Tumour Cell Adhesion under

Flow Conditions

9.3.1

Multistep Process of Leukocyte Adhesion

One of the most important aspects of leukocyte extravasation is that it is a multistep process (Figure 9.2): initial contact, primary adhesion (rolling), activation,
secondary adhesion, and transendothelial migration (diapedesis) [1].

I. ROLLING
II. ADHESION
SPREADING

BLOOD FLUX

III. MIGRATION

LEUKOCYTE

ENDOTHELIAL CELL

SELECTINS

INTEGRINS :

IMMUNOGLOBULINS :

aLb2 : LFA-1
aMb2 :Mac-1
a4b1 :VLA-4

IMMUNOGLOBULINS : ICAM-1
PECAM-1

ICAM-1
VCAM-1

CADHERINS :

VE-Cadherin

Figure 9.2

Leukocyte diapedesis: a multistep adhesion cascade.

Initial contact with endothelium is aided by the size of postcapillary venules,

which are the main sites of selective leukocyte extravasation [15]. The diameter

2003 CRC Press LLC

of these venules (approximately 20 to 60 m) is small enough to make frequent

contacts between leukocytes and endothelium, but large enough to make brief contacts. The initial contact is also supported by increased vascular permeability in an
inflammatory situation. This leads to plasma leakage and an increase in local hematocrit which changes the flow characteristics and thus allows more frequent contact
between leukocytes and the vessel wall [16]. Following this initial contact, leukocytes roll slowly along the endothelium for some distance before establishing firm
adhesion. This rolling keeps the cells in close contact with the endothelium, and is
mediated by selectins and their ligands. They can then be activated by substances in
the local environment, by factors bound to the endothelium, or by adhesion receptors
themselves. Activation of leukocytes alters their adhesion characteristics and allows
the establishment of firm adhesion (secondary adhesion). Flattening of leukocytes
upon activation greatly increases the leukocyte-endothelial cell contact area which
allows a large number of bonds to form and decreases fluid forces on the cell. This
permits a highly shear-resistant adhesion. This strong adhesion of leukocytes on
endothelium involves leukocyte integrins (Mac-1, LFA-1,
4 1) and their ligands
on endothelial cells (ICAM-1, VCAM-1). Then, leukocytes can migrate to interendothelial junctions and transmigrate (diapedesis) [3]. This last step is not completely
understood, but it has been shown that ICAM-1, and adhesive proteins expressed at
the interendothelial junctions (VE-cadherin, PECAM-1) are involved.

9.3.2

Adhesive Interactions between Cancer Cells and Endothelium

Despite the striking similarities between the process of leukocyte diapedesis

and tumour cell extravasation, there are differences between leukocytes and circulating tumour cells. Leukocytes are very motile small cells whereas tumour cells
are larger, with a far less ability to migrate. However, these bigger cells can be arrested easily by size constraints in the microcirculation. In addition, these cells can
form multicellular aggregate, by interacting with themselves or with leukocytes and
platelets, and it has been shown that multicellular emboli generate metastases more
efficiently. Morphological evidence indicates that only single cancer cells can enter
and be arrested in the capillaries, whereas multicellular emboli tend to arrest in larger
vessels [17]. In the case of single cancer cells, adhesive macromolecules such as selectins, integrins, cadherins, and immunoglobulins govern the adhesive interactions
between cancer cells and the endothelium. Initial contacts between cancer cells and
the endothelium are weak and transient and likely to be mediated by carbohydratecarbohydrate recognition. This initiates activation of both the endothelium and cancer cells through cytokines, free radicals, bioactive lipids, and growth factors. These
mediators cause the expression of new adhesion molecules which will reinforce the
initial adhesive bonds. Then humoral mediators or integrin-related signalling pathways lead to endothelial cell retraction, cancer cell locomotion, and transendothelial
migration of cancer cells [18].

2003 CRC Press LLC

9.4

In Vitro Devices to Study Circulating CellEndothelial Cell Adhesion

9.4.1

Static Assays

Circulating cell-endothelium adhesion is generally studied using three types of

assays: static assays, in vitro flow assays, and in vivo flow assays. Static assays are
the most straightforward and inexpensive systems available for adhesion studies. In
these systems, the substrate, which can be a cultured endothelial cell monolayer or a
purified ligand, is covered with a cell (leukocyte or cancer cell) suspension for some
period of time. Nonadherent cells are then rinced or centrifuged away, and adherent cells are quantified. These systems allow controlled stimulation of cells. They
also allow quantification of cells which transmigrate through the monolayer onto the
substrate (or through it, in case of a porous substrate), and can give a measure of
the strength of adhesion (with centrifugation) [19]. Variants utilise a multilayered
artificial blood vessel wall into which cells can migrate following diapedesis to examine transmigration and chemotactic stimuli [20]. The common disadvantage of
all static assay systems is the lack of shear forces associated with blood flow. With
static assays, it is impossible to distinguish primary and secondary adhesion events.

9.4.2

In Vitro Flow Assays

Since circulating cell-endothelium interactions take place in the fluid dynamic

environment of blood circulation, one of the obvious questions about these adhesive
interactions is whether they are capable of arresting circulating cells under the conditions of fluid flow. In vitro flow devices were designed to address this problem under controlled flow conditions. Flow assays provide more realistic information than
static assays by allowing discrimination of primary adhesion events (such as rolling)
from secondary adhesion events (like firm adhesion and transmigration). They are
also useful to examine the activation of cells under flow induced mechanical forces.
We will present three major in vitro devices for subjecting cultured cells to laminar
flow, i.e., the cone-and-plate system, the radial flow chamber, and the parallel-plate
flow chamber. Some specific configurations of the parallel-plate flow chamber will
be described in details.
9.4.2.1

Cone-and-Plate System

The cone-and-plate system consists of a stationary plate and a rotating cone

(Figure 9.3). The volume separating the two surfaces is filled with a liquid. The
cells are either cultured adherent to the stationary plate or suspended in the medium.

2003 CRC Press LLC

Figure 9.3

Cone-and-plate system.
The rotation of the upper cone, with an angular velocity  , causes the fluid to move
in a circumferential direction. The gap  between the cone and the plate increases
with the radial position  (   
). As the angle
between the cone and the
plate is small, typically less than one degree, the flow in any local region can be
approximated as the flow between two parallel surfaces. The modified Reynolds
   


) should be less than one to ensure a laminar flow with
number (
negligible secondary flows. The azimutal velocity varies as a function of  (the
distance from the stationary disk):

 



(9.1)

Thus, for a given angular velocity and conic taper, the shear stress exerted by
the moving fluid is constant regardless of the position and described as:



(9.2)

With such a device, Dewey et al. studied the effect of shear stress on endothelial
cell shape and orientation. They showed that cultured endothelial cells aligned in the
direction of flow and that this cell shape change and orientation is quite sensitive to
the magnitude of the applied shear stress and to the time of application [21]. More
recently, Blackman et al. [22] transformed a traditional cone-and-plate device in order to simulate physiological and pathological loading regimes (e.g., arterial wave
forms, repetitive load cycling) experienced by cells in vivo.
9.4.2.2

Radial Flow Chamber

In a radial flow chamber (Figure 9.4), fluid is introduced in the centre, moves out
radially and exits at the edge. So the radial flow apparatus provides an axisymmetric
laminar flow field between two parallel disks. In this geometry, the cross-sectional
area for flow between the two disks increases radially and the radial velocity decreases. Thus the wall shear stress decreases radially and this produces a continuous
range of shear stress values within a given experiment. The wall shear stress can be

2003 CRC Press LLC

Figure 9.4

Radial flow chamber.

estimated by the following formula:

 



 

(9.3)

where  is the flow rate,  the fluid viscosity,  the height of the radial flow channel,
and  the radial position. The typical geometric values are: 140 to 300  gap
height and 2 to 3 cm disk radius. The local Reynolds number must be less than 2000
to ensure laminar flow field. Typical wall shear stresses range from 0 to 4 Pa.
The radial flow chamber has been used in general studies of receptor-mediated
cell adhesion. For instance, Cozen-Roberts et al. [23] used this device to study the
detachment of receptor-coated latex beads from ligand-coated glass surfaces. This
device could be easily adapted to study cell-cell interactions under a range of shear
stresses. To do so, endothelial cells should be cultured on the lower disk and leukocytes or cancer cells suspended in the circulating medium. However, a potential
problem is that many adhesion events are dependent on both shear stress magnitude
and suspended cell activation. Thus, increased cell adhesion near the periphery could
be due either to the lower shear stress there or to longer exposure of suspended cells
to activating substances derived from the endothelial monolayer.
9.4.2.3

Parallel-Plate Flow Chamber

The most commonly used flow device is the parallel-plate flow chamber [24
26]. It consists of a channel with a rectangular cross section of height , width ,
and length . In this device a pressure gradient is created between either end of
the rectangular chamber, causing the fluid to flow inside the channel. The pressure
gradients can be established by either a hydrostatic pressure head or by active pumps.
For the flow to be two-dimensional, the channel height must be much smaller than
its width. For the velocity profile to be fully developed and parabolic over nearly
the entire length of the channel, the entrance length must be small compared to the
channel length. In this type of flow chamber, assuming Newtonian fluid behaviour,

2003 CRC Press LLC

the wall shear stress  is independent of position and is predicted by the following
formula:

  
(9.4)



where  is the flow rate and  the fluid viscosity. The typical geometric values are
100 to 200  gap height, 2 to 3 cm width, and 5 to 7 cm length. The Reynolds
number (  
) is generally less than 100 to ensure a laminar flow. The wall
shear stress can be varied from 0.5 to 5 Pa by changing the flow rate or the gap height
of the flow channel.
9.4.2.4

Linear Shear Stress Parallel-Plate Flow Chamber

In commonly used parallel-plate flow chambers, the shear stress in the whole
field is constant and dependent upon the flow rate and the gap between the two plates.
Whenever one wants to change the shear stress in the field, the flow rate has to be
altered or the gap height has to be changed. These procedures, however, will affect
cell-cell interactions after the cells and the surface are exposed to a different shear
stress in the previous shear flow. The properties of the two-dimensional stagnation
flow permit the design of a flow channel such that the wall shear stress is linearly
distributed along the centre line of the channel.
Usami et al. [27] realised this set up by letting the sides of a flow channel be
coincident with streamlines corresponding to a stagnation flow. The wall shear stress
depends on the axial position  and is described by:

 









(9.5)

where  is the entrance width of the flow channel. Thus, the wall shear stress decreases from the maximum at the entrance (  ) to zero at the exit (  ). With
this design, cell-cell interactions can be studied efficiently over a wide range of shear
stresses using a single flow rate.
9.4.2.5

Side-View Parallel-Plate Flow Chamber

One of the limitations of the parallel-plate flow chamber is that it can only
provide top views of cells that are in contact with the substrate. Some important adhesion and deformation parameters that are related to cell-surface contact under flow
are hard to obtain quantitatively from a top-view chamber. Cao et al. [28] developed
an in vitro side-view flow chamber that permits observations from the side of the
cells in contact with various adhesive surfaces under dynamic flow conditions. This
flow chamber consists of two precision rectangular glass tubes called microslides.
A smaller microslide is inserted into a larger one to create a flow channel with a
flat surface on which either cultured vascular endothelium can be grown or purified
adhesion molecules can be coated. Two optical prisms with a
 chromium-coated
surface are used along the flow channel to generate light illumination and observation pathways. The side-view image of cell-substrate contact can be obtained using
a light microscope. The characteristic data of this flow channel are: 550 m height,

2003 CRC Press LLC

700 m width, and 5 cm long. The wall shear stress can be estimated by the following formula:

   
(9.6)



where  is the correction factor for a rectangular channel with a finite aspect ratio.
Here for a 
 
 ,  varies from 1.4 to 1.5 in the central regions ( 


). For shear stresses ranging from 0 to 3 Pa, the corresponding Reynolds number
is less than 80, and the entrance length less than 2 mm. This design allows to measure
the effects of flow on cell-surface adhesion strength. Moreover, it permits a close
observation of cell deformation and adhesive contact to various surfaces under flow
conditions.
9.4.2.6

Parallel-Plate Flow Chamber with a Porous Bottom Wall

Another configuration of a parallel-plate flow chamber is the one with a porous

bottom wall designed by Chotard-Ghodsnia et al. [29]. It consists of two stainless
steel parts which enclose a pair of parallel glass plates to allow for microscope visualisation. A porous material is held on a nylon screen positioned on the bottom
plate, as shown in the assembled view of this chamber (Figure 9.5). Thus, the circulating medium can flow both along and across the porous bottom wall. The inlet
fluid (     ) is evacuated through a tangential outlet (
  
 ) and a filtrate
outlet (   ). With this flow chamber, cells adhering to the porous material can be
exposed to both a transmural pressure (  ) and a shear stress (  ). This in vitro
system thus allows a more realistic reproduction of flow conditions near the vessels
wall in vivo. Indeed, the transmural pressure might tend to favour cell attachment to
the endothelium whereas the fluid shear stress tends to detach cells from the vessel
wall. This might also affect the signal mechanotransduction.
The hydraulic permeability,  , of the porous bottom wall (

m/(Pa s))
is similar to that of a blood vessel wall (

m/(Pa s) [30]). Thus, the filtrate
flow rate  is very small as compared to input flow rate   , and the tangential

Figure 9.5

Flow chamber with a porous bottom wall.

2003 CRC Press LLC

flow can still be approximated by the plane Poiseuille law. This is confirmed by our
experimental and theoretical analysis and, consequently, the wall shear stress can be
considered as a constant along the flow channel, given by:

 

 



(9.7)

Besides, the transmural flow is governed by Darcys law:

     

(9.8)

where   , the mean transmural pressure, writes:       

 

 , since the pressure profile is linear. We first used this flow chamber to study morphological and biochemical responses of fibroblasts when submitted to tangential
and normal stresses [31]. Our results constitute evidence that transmural pressure
is at least as important as shear stress in regulating cell morphology mediated by
the cAMP pathway (cyclic adenosine monophosphate: an intracellular signal transducer). Further studies should be carried out to study the influence of the transmural
pressure on circulating cell/endothelium interactions.

9.5

In Vitro Flow Studies of Circulating CellEndothelium Adhesion

The rolling of leukocytes on activated endothelium is a critical step in the inflammatory cascade and has received considerable attention in the literature [32].
Leukocytes rolling occurs via the following steps:
1. a receptor-ligand bond forms, exerting an adhesive stress which slows cell
velocity;
2. the slower motion of the cell promotes additional bonding;
3. bonds dissociate at the back edge of contact, causing the cell to tumble forward
in the direction of flow.
These receptor-ligand bonds between the leukocyte and the endothelium exert a friction on the leukocyte, such that its velocity drops well below the hydrodynamic velocity for an unencumbered leukocyte at the same separation distance and wall shear
rate. In the cell biology literature, rolling is often defined as a significant decrease in
velocity to perhaps 50% or less of the hydrodynamic velocity for an unencumbered
cell [33].
Although static assays may detect possible receptor-ligand pairs, they do not
provide information on how ligands interact with receptors under more physiological, dynamic conditions. For cells to adhere under flow, the potential receptor-ligand

2003 CRC Press LLC

pair must quickly react. Once a bond is formed, it must be able to overcome the
drag force exerted by the fluid. Thus, dynamic adhesion studies are better to detect
receptor-ligand pairs that can mediate rolling as compared to static experiments [34].
The most commonly used device for dynamic adhesion experiments is the parallel-plate flow chamber. The endothelial monolayer is cultured on the bottom plate of
the flow channel and leukocytes or tumour cells are perfused to the flow channel at a
rate that produces the desired wall shear stress. A typical observation of leukocytes
rolling over an endothelial monolayer, obtained in our experimental set up, is shown
in Figure 9.6. Some leukocytes are rolling whereas some others are firmly adherent
in a TNF-
-stimulated endothelial monolayer.

Figure 9.6

Leukocyte rolling on endothelial monolayer: one cell (1) is rolling while others
(2 to 5) are adhering.

Such in vitro flow studies showed that distinct sets of adhesion molecules mediate leukocyte rolling and firm adhesion, as widely discussed in a recent review article [35]: (1) L-, E-, or P-selectins are capable of tethering free flowing leukocytes.
Shear stress levels in excess of    are required for optimal selectin-dependent
tethering and rolling of leukocytes; (2)
 -integrins are capable of supporting leukocyte adhesion to activated endothelium under flow; (3)  -integrins cannot initiate

2003 CRC Press LLC

leukocyte adhesion under flow conditions, except possibly at wall shear stresses less
that   , but can support activation-dependent firm adhesion and subsequent migration. Moreover, it has been observed that leukocyte rolling velocity varies with
wall shear stress [24,26,3638] and that the velocity of rolling cells varies considerably with time [37,39]. Kaplanski et al. [39] observed multiple short term arrests (of
2 s duration) of neutrophils on endothelial monolayer under  shear rate. Dong
et al. [40] used a side-view parallel-plate flow chamber to study the influence of cell
deformation on leukocyte rolling adhesion in shear flow. They showed that changes
in shear flow and cell deformability resulted in significant changes in characteristic adhesion binding time, cell-surface contact, and cell rolling velocity. Recently,
the influence of shear stress on leukocyte migration behaviour was also studied in a
parallel-plate flow chamber. Shear flow significantly increases the kinetics of leukocyte transmigration as compared to static conditions [41].
In contrast to leukocytes, the effect of shear stress on cancer cell adhesion to
endothelium has received much less attention in the literature, although cancer cells
and leukocytes have been shown to share some adhesion mechanisms [42]. Several
authors have reported E-selectin-dependent rolling of varied cancer cell types on the
endothelial monolayer under flow conditions [43,44]. A recent work of Moss et al.
[45] showed the importance of studying cancer cell adhesion under flow conditions.
Using a parallel-plate flow chamber, they compared the adhesion of cancer cells
with different metastatic potential and showed that highly metastatic cells were more
adhesive to shear stimulated endothelial cells while nonmetastatic cells were more
adhesive to endothelial monolayer not stimulated. More recent experiments carried
out by Dong et al. [46] with a modified parallel-plate flow chamber, suggested that
shear flow plays a significant role in tumour cell extravasation.

9.6

Modelling of Circulating Cell-Endothelium

Interactions

Although in vitro flow experiments can be more easily controlled than in vivo
studies, there remains a large number of confounding cellular features such as: endothelial cell surface heterogeneity, adhesion molecule expression, cell activation,
and cell deformability. Experimental models were developed to make it easier to
understand the biophysical basis of circulating cell/endothelium interactions. Then
mathematical models could be developed to interpret these experimental results.

2003 CRC Press LLC

9.6.1

Experimental Modelling

Springer and coworkers simplified leukocyte endothelium systems by replacing

the cultured endothelial cells with purified P-selectin reconstituted in lipid bilayers [47,48] and purified E-selectin absorbed to polystyrene slides [49]. Neutrophils
rolled over both substrata with a velocity proportional to the selectin surface density.
Kitayama and coworkers [50] compared the adhesion pattern of colon cancer cells to
that of leukocytes over an E-selectin coated substratum. Adhesive interactions to Eselectin in laminar flow were weaker for colon cancer cells than for leukocytes. Cancer cells showed a stop and roll movement with variable rolling velocities whereas
leukocytes rolled with a relatively stable speed.
Although these experiments removed the variability of the endothelial monolayer, circulating cell activation and deformability was still present and caused complications.
Cell-free systems were then developed to study specific molecular (receptorligand) interactions without the confounding influence of rheology, roughness, signalling, and complex molecular display involved when using cells. In these systems,
coated microspheres are used to model leukocytes and coated substratum to model
endothelial surface.
Brunk and coworkers [51,52] used sLe
-coated polystyrene microspheres over
E-selectin-coated glass substratum in a parallel-plate flow chamber to model leukocyte rolling over endothelium. They showed that these microspheres rolled on the
coated substratum with dynamics similar to those of leukocytes rolling over stimulated endothelial cells. The particle rolling velocity was found to be a function of
wall shear stress and selectin or sLe
surface densities. Moreover, the rolling velocity
varied with time, as also seen in leukocyte/endothelium systems [37,39]. These cellfree studies allowed to understand how the rolling velocity depends quantitatively on
receptor number, ligand density, and shear rate.
Bongrand and coworkers used cell-free systems to measure the lifetime of individual ligand-receptor bonds with a flow chamber [53,54]. To observe the formation
and dissociation of single bonds, they operated with microspheres (2.8 m diameter) much smaller than cells under low shear rates (of the order of a few seconds ).
They determined the velocity, attachment frequency, and duration of binding arrests
of each microsphere and found that microspheres exhibited frequent arrests of fairly
constant duration (approximately 1 s). Their results suggest that the arrests observed
with low receptor densities involved single molecular bonds and the duration of these
bonds was not affected by shear forces [53]. They also showed that the rate of bond
formation is a function of microsphere-substratum distance [54,55].

2003 CRC Press LLC

9.6.2

Mathematical Modelling

Analysis of the forces involved in leukocyte-endothelium adhesion at the equilibrium state was pioneered by Bell [56]. Evans [57] established a theoretical framework of a one-dimensional tape peeling model to compute the adhesion force between a biomembrane and a substrate with constant adhesive strength. To study the
receptor-mediated cell adhesion to a ligand-coated surface, Hammer and collaborators [58] took steps from Bells theory and developed a cell adhesion model taking
receptor-ligand bond kinetics into consideration. This model, called adhesive dynamics, allows to interpret experimental results on rolling cells and to understand
the molecular basis of cell-free rolling systems. We now describe this mathematical
model in more details.
Adhesive dynamics is a computational technique that combines the fluid mechanics analysis of particle motion and the Monte Carlo simulation of bond formation and breakage between receptors and ligands [58]. The adhesive dynamic
method has been extensively described [5860]. Briefly, sLe
-coated microspheres
are modelled as hard spheres with receptors distributed randomly over the surface.
The planar surface coated with E-selectin is assumed to be uniformly reactive (Eselectin density is always larger than sLe
density on the microspheres). Receptors
and ligands are modelled as adhesive springs with spring constant  and equilibrium
length . Each adhesive molecule reacts with the substrate with an overall association rate  (forward reaction) and dissociation rate  (reverse reaction).  is
a function of ligand density, relative velocity between the particle and the surface
(slip velocity  ) and the intrinsic forward rate constant   . The model proposed by
Bell [56] was used to describe  of a bond under applied force  :

    
  

(9.9)

where  is the unstressed dissociation constant rate,  is the Boltzmanns constant,  is the temperature, and  is the bond interaction length (also called reactive compliance, it represents the sensitivity of the dissociation to the applied force).
Flow-induced forces tend to alter the rate of dissociation compared to cases where no
force acts on receptor-ligand bonds. The Bell model takes into account the influence
of these dislodging forces on the dissociation rate.
Each free receptor can become tethered in the time interval ! with a probability:

 
  !

(9.10)

and each tethered receptor can become free in a time interval ! with a probability:

 
  !

(9.11)

During each time step, bond formation and breakage are simulated by a Monte Carlo
sampling of these probability distributions. The Monte Carlo approach simulates the
fate of each individual bond to generate an ensemble of realisations. Each simulation
represents an experimental measurement and can be directly compared to it.

2003 CRC Press LLC

At any instant, the forces, " , exerted by the tethers on the particle can be calculated from the distance between the end points of the attachment, #, and using the
Hookes law: "   # . To obtain the net force and torque acting on the particle, the total force and torque exerted by the bonds on the cell are added to interfacial
forces and to the force and torque exerted by the fluid shear on the cell. The motion
of the particle can be fully described by the following relation:
U  MF

(9.12)

where M is the mobility matrix (known for a sphere near a plane wall in a viscous
fluid [59]), U is a vector with three components of linear velocity and three components of angular velocity, and F is a vector with three components of total force and
three components of the net torque acting on the particle. Thus, once the net force
and torque are known, the velocity of the particle can be calculated.
The simulation begins with a freely moving particle at a constant velocity and
a separation distance (40 nm) greater than the receptor-ligand bond. The position of
the free receptors and tethers, the velocity of the particle, and the net force and torque
exerted on it are known. According to the position of free receptors at time !, tethers
are formed at !  ! following the probability of Equation (9.10). Tethers are broken
at !  ! following Equation (9.11). Then, new positions of free receptors and tethers
are calculated (from velocity and position at !), as well as new net force at !  !.
Finally, the velocity of the particle at !  ! is obtained using Equation (9.12). The
process is repeated until the particle travels across the field of view or a predefined
time is reached. After each simulation, the trajectory of the particle is recorded and
the instantaneous velocity and the average velocity (total displacement divided by
time of interactions) are calculated.
This model can recreate different adhesive behaviours ranging from unencumbered motion of particles to rolling and to firm adhesion [58]. It can simulate the
effect of many parameters on cell rolling and adhesion, such as the density of receptors and ligands, the rates of reaction between receptor and ligand (association
and dissociation rates), the receptor-ligand bond elasticity and the shear rate [61]. To
elucidate the relationship between receptor-ligand functional properties and the dynamics of adhesion, Chang et al. [32] expressed the state diagram for cell adhesion
under flow. This state diagram describes how biophysical properties of adhesion
molecules induce different states of adhesion in flow. These authors showed that
the unstressed dissociation rate,  , and the bond interaction length,  , are the most
important molecular properties that control the dynamics of adhesion. Their results
suggest that adhesive behaviour is primarily determined by the biophysical properties of adhesion molecules [32] however, it can be modulated by cell deformability,
morphology, and signalling.
More recently, Dong and collaborators [40] developed a 2D model that demonstrates the influence of leukocytes rheological properties in regulating rolling on vascular endothelium. This model incorporates both mechanical aspects of cell deformation and biophysical aspects of adhesion bond kinetics. It was based on the assumption that the fluid energy input to a rolling cell would essentially be distributed

2003 CRC Press LLC

into two parts: cytoplasmic viscous dissipation and energy needed to break adhesion
bonds between the rolling cell and its substratum. Both extracellular and intracellular flow fields were solved using finite element methods with a deformable cell
membrane represented by an elastic ring. The adhesion energy loss was calculated
based on the receptor-ligand kinetics equations. They found that the cell-substrate
contact area under high wall shear stress (  ) could be nearly twice that under low
stress (   ) as a result of shear flow-induced cell deformation. An increase in
contact area resulted in more energy dissipation to both adhesion bonds and viscous
cytoplasm, whereas the fluid energy that is transmitted to a cell decreased due to
flattened cell shape. Their results suggest that leukocyte deformability is an essential component that aids in the adhesion process to balance the hemodynamic and
leukocyte-endothelium adhesive forces.

9.7

Conclusion

Metastasis is the growth of secondary tumours at sites distant from a primary

tumour, and is responsible for the majority of failures in cancer treatment. Cells from
a metastatic tumour are able to escape in the surrounding tissues and intravasate into
blood, where they can disseminate through the body. Endothelial cells line every
blood vessel in the organism and regulate the extravasation of blood cells. This
extravasation involves specific junctional proteins and membrane-bound receptors
that control cell-cell interactions. Leukocyte interactions with endothelial cells have
been extensively studied and serve as a model for interactions of circulating cancer
cells with the vasculature. The molecular mechanisms involved in the extravasation
of cancer cells through the endothelial monolayer is not fully understood, and better
experimental approaches are needed to understand this phenomenon. The use of in
vitro flow assays, associated with modern microscopic techniques and mathematical
modelling, will help identifying the different molecules mediating the process of
metastasis. A better understanding of the metastasis mechanisms is crucial for the
development of new therapies. The implication of adhesion molecules in the arrest
of cancer cells in the vasculature strongly suggests that inhibition of these adhesive
interactions could be helpful to inhibit metastasis at the step of extravasation.

9.8

References

[1] Springer, T.A., Traffic signals for lymphocyte recirculation and leukocyte emigration: the multistep paradigm, Cell 76, 301, 1994.

2003 CRC Press LLC

[2] Voura, E.B., Sandig, M., and Siu, C.-H., Cell-cell interactions during
transendothelial migration of tumor cells, Microsc. Res. Tech. 43, 265, 1998.
[3] Johnson-Leger, C., Aurrand-Lions, M., and Imhof, B.A., The parting of endothelium: miracle, or simply a junctional affair?, J. Cell Science 113, 921,
2000.
[4] Ebnet, K. and Vestweber, D., Molecular mechanisms that control leukocyte
extravasation: the selectins and the chemokines, Histochem. Cell Biol. 112, 1,
1999.
[5] Kansas, G.S., Selectins and their ligands: current concepts and controversies,
Blood 88, 3259, 1996.
[6] Takada, A. et al., Adhesion of human cancer cells to vascular endothelium
mediated by a carbohydrate antigen, sialyl Lewis A, Biochem. Biophys. Res.
Commun. 179, 713, 1991.
[7] Lu, H.F. et al., LFA-1 is sufficient in mediating neutrophil emigration in Mac1-deficient mice, J. Clin. Invest. 99, 1340, 1997.
[8] Albelda, S.M. et al., Integrin distribution in malignant melanoma: association
of the 3 subunit with tumor progression, Cancer Res. 50, 6757, 1990.
[9] Tomita, Y. et al., Possible significance of VLA-4 (
4 1) for hematogenous
metastasis of renal-cell cancer, Int. J. Cancer 60, 753, 1995.
[10] Hubbard, A.K. and Rothlein, R., Intercellular adhesion molecule-1 (ICAM-1)
expression and cell signaling cascades, Free Radic. Biol. Med. 28, 1379, 2000.
[11] Sans, E. et al., Analysis of the roles of ICAM-1 in neutrophil transmigration
using a reconstituted mammalian cell expression model: implication of ICAM1 cytoplasmic domain and Rho-dependent signaling pathway, J. Immunol. 166,
544, 2001.
[12] Buckley, C.D. et al., Identification of
v 3 as a heterotypic ligand for
CD31/PECAM-1, J. Cell Sci. 109, 437, 1996.
[13] Dejana, E. et al., Vascular endothelial (VE)-cadherin: only an intercellular
glue?, Exp. Cell Res. 252, 13, 1999.
[14] Liao, F. et al., Monoclonal antibody to vascular endothelial-cadherin is a potent
inhibitor of angiogenesis, tumor growth, and metastasis, Cancer Res. 60, 6805,
2000.
[15] Fiebig, E., Ley, K., and Arfors, K.E., Rapid leukocyte accumulation by spontaneous rolling and adhesion in the exteriorized rabbit mesentery, Int. J. Microcirc. Clin. Exp. 10, 127, 1991.
[16] Tangelder, G.J. et al., Velocity profiles of blood platelets and red blood cells
flowing in arterioles of the rabbit mesentery, Circ. Res. 59, 505, 1986.

2003 CRC Press LLC

[17] Starkey, R. et al., Influence of migratory blood cells on the attachment of tumour cells to vascular endothelium, Int. J. Cancer 34, 535, 1984.
[18] Orr, F.W. et al., Interactions between cancer cells and the endothelium in
metastasis, J. Pathol. 190, 310, 2000.
[19] Charo, I.F., Yuen, C., and Goldstein, I.M., Adherence of human polymorphonuclear leukocytes to endothelial monolayers: effects of temperature, divalent cations, and chemotactic factors on the strength of adherence measured
with a new centrifugation assay, Blood 65, 473, 1985.
[20] Hakkert, B.C. et al., A three-dimensional model system to study the interactions between human leukocytes and endothelial cells, Eur. J. Immunol. 20,
2775, 1990.
[21] Dewey, C.F. et al., The dynamic response of vascular endothelial cells to fluid
shear stress, ASME J. Biomech. Eng. 103, 177, 1981.
[22] Blackman, B.R. et al., In vitro cell shearing device to investigate the dynamic
response of cells in a controlled hydrodynamic environment, Annals Biomed.
Eng. 28, 363, 2000.
[23] Cozen-Roberts, C., Quinn, J.A., and Lauffenburger, D.A., Receptor-mediated
adhesion phenomena: model studies with the radial-flow detachment assay,
Biophys. J. 58, 107, 1990.
[24] Lawrence, M.B., McIntire, L.V., and Eskin, L.V., Effect of flow on polymorphonuclear leukocyte/endothelial cell adhesion, Blood 70, 1284, 1987.
[25] Gallik, S. et al., Shear stress-induced detachment of human polymorphonuclear leukocytes from endothelial cell monolayers, Biorheology 26, 823, 1989.
[26] Jones, D.A. et al., P-selectin supports neutrophil rolling on histaminestimulated endothelial cells, Biophys. J. 65, 1560, 1993.
[27] Usami, S. et al., Design and construction of linear shear stress flow chamber,
Annals Biomed. Eng. 21, 77, 1993.
[28] Cao, J., Usami, S., and Dong, C., Development of a side-view chamber for
studying cell-surface adhesion under flow conditions, Annals Biomed. Eng.
25, 573, 1997.
[29] Chotard-Ghodsnia, R., Drochon, A., and Grebe, R., A new flow chamber for
the study of shear stress and transmural pressure upon cells adhering to a
porous biomaterial, ASME J. Biomech. Eng. 124, 258, 2002.
[30] Smaje, L.H. and Swayne, G.T.G., The effects of compliance on measurement
of hydraulic conductivity in microvessels, Biorheology 21, 171, 1984.
[31] Chotard-Ghodsnia, R. et al., Effect of shear stress and of transmural pressure
on cAMP-dependent responses of cells adhering to a biomaterial, Eur. Phys.
J.-Applied Phys. 17, 155, 2002.

2003 CRC Press LLC

[32] Chang, K.C., Tees, D.F.J., and Hammer, D.A., The state diagram for cell adhesion under flow: leukocyte rolling and firm adhesion, Proc. Natl. Acad. Sci.
USA 87, 11262, 2000.
[33] Chen, S. and Springer, T.A., An automatic braking system that stabilizes leukocyte rolling by an increase in selectin bond number with shear, J. Cell Biol.
144, 185, 1999.
[34] Varki, A.J., Selectin ligands, Proc. Natl. Acad. Sci. USA 91, 7390, 1994.
[35] Konstantopoulos, K., Kukreti, S., and McIntire, L.V., Biomechanics of cell
interactions in shear fields, Advanced Drug Delivery Reviews 33, 141, 1998.
[36] Lawrence, M.B. et al., Effect of venous shear stress on CD18-mediated neutrophil adhesion, Blood 75, 227, 1990.
[37] Goetz, D.J. et al., Dynamics of neutrophil rolling over stimulated endothelium
in vitro, Biophys. J. 66, 2202, 1994.
[38] Rinker, K.D., Prabhakar, V., and Truskey, G.A., Effect of contact time and
force on monocyte adhesion to vascular endothelium, Biophys. J. 80, 1722,
2001.
[39] Kaplanski, G. et al., Granulocyte endothelium initial adhesion, Biophys. J. 64,
1922, 1993.
[40] Dong, C. and Lei, X.X., Biomechanics of cell rolling: shear flow, cell-surface
adhesion, and cell deformability, J. Biomech. 33, 35, 2000.
[41] Kitayama, J. et al., Shear stress affects migration behaviour of polymorphonuclear cells arrested on endothelium, Cellular Immunology 203, 39, 2000.
[42] Takada, A. et al., Contribution of carbohydrate antigens sialyl Lewis A and
sialyl Lewis X to adhesion of cancer cells to endothelium, Cancer Res. 53,
354, 1993.
[43] Giavazzi, R. et al., Rolling and adhesion of human tumor cells on vascular
endothelium under physiological flow conditions, J. Clin. Invest. 92, 3038,
1993.
[44] Tozeren, A. et al., E-selectin-mediated dynamic interactions of breast- and
colon-cancer cells with endothelial-cell monolayers, Int. J. Cancer 60, 426,
1995.
[45] Moss, M.A., Zimmer, S., and Anderson, K.W., Role of metastatic potential
in the adhesion of human breast cancer cells to endothelial monolayers, Anticancer Research 20, 1425, 2000.
[46] Dong, C. et al., In vitro characterization and micromechanics of tumor cell
chemotactic protrusion, locomotion, and extravasation, Annals Biomed. Eng.
30, 344, 2002.

2003 CRC Press LLC

[47] Lawrence, M.B. and Springer, T.A., Leukocytes roll on a selectin at physiologic flow rates: distinction from and prerequisite for adhesion through integrins, Cell 65, 859, 1991.
[48] Alon, R., Hammer, D.A., and Springer, T.A., Lifetime of the P-selectincarbohydrate bond and its response to tensile force in hydrodynamic flow, Nature 374, 539, 1995.
[49] Lawrence, M.B. and Springer, T.A., Neutrophils roll over E-selectin, J. Immunol. 151, 6338, 1993.
[50] Kitayama, J. et al., E-selectin can mediate the arrest type of adhesion of colon
cancer cells under physiological shear flow, Eur. J. Cancer 36, 121, 2000.
[51] Brunk, D.K., Goetz, D.J., and Hammer, D.A., Sialyl Lewis
/ E-selectin mediated rolling in a cell free system, Biophys. J. 71, 2902, 1996.
[52] Brunk, D.K. and Hammer, D.A., Quantifying rolling adhesion with a cell-free
assay: E-selectin and its carbohydrate ligands, Biophys. J. 72, 2820, 1997.
[53] Pierres, A., Benoliel, A.-M., and Bongrand, P., Measuring the lifetime of bonds
made between surface-linked molecules, J. Biol. Chem. 270, 26586, 1995.
[54] Pierres, A., Benoliel, A.-M., and Bongrand, P., Measuring bonds between
surface-associated molecules, J. Immunol. Methods 196, 105, 1996.
[55] Pierres, A. et al., The dependence of the association rate of surface-attached
adhesion molecules CD2 and CD48 on separation distance, FEBS Letters 403,
239, 1997.
[56] Bell, G.I., Models for the specific adhesion of cells to cells, Science 200, 618,
1978.
[57] Evans, E., Detailed mechanics of membrane-membrane adhesion and separation, I. continuum of molecular cross-bridges, Biophys. J. 48, 175, 1985.
[58] Hammer, D.A. and Apte, S.M., Simulation of cell rolling and adhesion on surfaces in shear flow: general results and analysis of selectin-mediated neutrophil
adhesion, Biophys. J. 63, 35, 1992.
[59] Chang, K.-C. and Hammer, D.A., Influence of direction and type of applied
force on the detachment of macromolecularly-bound particles from surfaces,
Langmuir 12, 2271, 1996.
[60] Kuo, S.C., Hammer, D.A., and Lauffenburger, D.A., Simulation of detachment
of specifically bound particles from surfaces under shear flow, Biophys. J. 73,
517, 1997.
[61] Chang, K.C. and Hammer, D.A., Adhesive dynamics simulations of sialylLewis
/ E-selectin-mediated rolling in a cell-free system, Biophys. J. 79,
1891, 2000.

2003 CRC Press LLC