Vol. 279, Issue 4, H1460-H1471, October 2000
Influence of erythrocyte aggregation on leukocyte margination
in postcapillary venules of rat mesentery
Mark J.
Pearson and
Herbert H.
Lipowsky
Bioengineering Program, The Pennsylvania State University,
University Park, Pennsylvania 16802
 |
ABSTRACT |
The role of erythrocyte
(red blood cell; RBC) aggregation in affecting leukocyte (white blood
cell; WBC) margination in postcapillary venules of the mesentery (rat)
was explored by direct intravital microscopy. Optical techniques were
refined and applied to relate the light-scattering properties of RBCs
to obtain a quantitative index of aggregate size (G), which,
under idealized conditions, represents the number of RBCs per
aggregate. WBC margination, defined as the radial migration of WBCs to
the venular wall and their subsequent rolling along the endothelium,
was measured as the percentage of the potentially maximal WBC
volumetric flux within the microvessel lumen
(FWBC*). In normal blood,
FWBC* increased exponentially
fourfold, and G increased from 1 to 1.15 as wall shear rates
(
) were reduced from a steady-state value of ~600 to <100
s
1 by proximal occlusion with a blunt microprobe.
Enhancement of aggregation by infusion (iv) of dextran 500 (428 kDa),
to attain a systemic concentration of 3 g/100 ml, resulted in a four-
and sevenfold increase in G and
FWBC*, respectively, as
was reduced below 100 s1. Inhibition of RBC aggregation
by infusion of dextran 40 (37.5 kDa) caused
FWBC* to fall to one-half of its
steady-state level for < 100 s1. Thus it
appears that the well-known increase of WBC margination with reductions
in is strongly dependent on the occurrence of RBC
aggregation. Increasing the extent of RBC aggregation during reductions
in also increased the firm adhesion of WBCs to the endothelium because of an enhanced probability of contact between leukocytes and the postcapillary venular wall.
erythrocyte sequestration; dextran; Wistar-Furth rats; leukocyte-endothelium adhesion
| |
INTRODUCTION |
RECEPTOR-MEDIATED
ADHESION of leukocytes (or white blood cells; WBCs) to
postcapillary endothelium (EC) is recognized as an essential part of
the inflammatory process. It is widely acknowledged that the mechanics
and kinematics of WBC-EC interaction may play a significant role in the
kinetics of the receptor-ligand bond formation that promotes WBC-EC
adhesion. Given that the probability of WBC attachment to the EC is
proportional to the frequency at which WBCs strike EC, many studies
(1, 4, 8, 9, 20, 21, 30, 35) have examined WBC dynamics
during their passage through the capillary network and their subsequent
radial migration toward the venular wall (margination).
It has been hypothesized that WBC margination is augmented by the
dynamic interaction between WBCs and red blood cells (RBCs) that begins
at the exit of the true capillaries. It has been shown by in vitro and
in vivo studies (30) that RBCs tend to push the WBC toward
the walls of postcapillary venules as blood exits the true capillaries.
It has also been hypothesized that the formation of aggregates of RBCs
along the venular centerline (where low shear rates promote
aggregation) may also enhance the radial migration of WBCs toward the
EC as aggregates exclude WBCs from the axial core of RBCs. Vejlens
(35) observed an increase in WBC margination in vivo when
the level of red cell aggregation (RCA) was increased with gelatin or
fibrinogen. Further support for this mechanism of margination was
provided by Nobis et al. (22) and Goldsmith and Spain
(9) by use of small bore glass tubes; they found that RCA
enhanced the margination of WBCs and that in the absence of RCA, there
was no WBC margination. These findings were supported by in vivo
observations of microvascular perfusion with WBC suspensions devoid of
RBCs (4), which revealed little or no WBC margination. Additional in vivo assessments of the role of hydrodynamic factors that
promote margination have focused on the relative roles of WBC-EC
adhesion in venules and hydrodynamic factors. By reversal of the
direction of flow within the microvascular network, a significant amount of margination occurred in arterioles, whereas under normograde flow, virtually no arteriolar WBC margination occurred
(20). The fact that arteriolar margination with flow
reversal occurred to an extent equal to that in venules under normal
flow suggests that RBC-WBC interactions may act as a significant
impetus for margination and subsequent WBC-EC adhesion.
It is therefore apparent that a quantitative assessment of the effect
of RCA on WBC margination may provide valuable insight into the
determinants of WBC margination, particularly in light of the potential
for enhancing RCA with reductions in wall shear rate (). To
date, these processes have been explored only qualitatively because of
a lack of suitable instrumentation for providing a direct quantitative
measure of the extent of RCA. The present study aims to overcome this
deficiency by quantitatively measuring the degree of RCA in
postcapillary venules and seeking correlates with the extent of WBC
margination and reductions in . To this end, in situ methods
for evaluating the light-scattering properties of blood in small
venules have been applied to calculate an aggregation index, based on
the size of RBC aggregates, and to relate this index to the flux of
WBCs rolling along the venular wall. The occurrence of RCA was
increased by reductions in by partial microvessel occlusion
with a blunt microprobe, and the extent of RCA was increased further by
administration of high-molecular-mass dextran (Dx500; 428 kDa). The
results of these manipulations were compared with the effects of
administering a low-molecular-mass dextran (Dx40; 37.5 kDa)
that is known to act as a disaggregating agent.
| |
METHODS |
Animal preparation.
Male Sprague-Dawley, Wistar, or Wistar-Furth rats (240-400 g) were
anesthetized with pentobarbital sodium (35 mg/kg ip). After tracheostomy, a jugular vein and carotid artery were cannulated with
polyethylene tubing for infusions of supplemental anesthetic or dextran
and for monitoring systemic arterial pressure or removing blood
samples, respectively. Systemic arterial pressure was measured with a
strain gage pressure transducer (Century Technology model CP-01). The
intestinal mesentery was exteriorized via a midline abdominal incision
and suffused with warmed (37 ± 1°C) HEPES-buffered (pH = 7.4) Ringer solution (4.20 mM HEPES, 0.126 M NaCl, 22.85 mM
NaHCO3, 3.43 mM KCl, and 2.602 mM CaCl2) with
1% 275 bloom Gelatin (Fisher Scientific). The mesentery was
transilluminated with light from a 150-W xenon lamp, and the image of
small venules was projected by a ×10 eyepiece onto the face of a
silicon diode video camera (Dage MTI, model 67). A long
working-distance Nikon U20 [×13/0.22 numerical aperture (NA)]
objective was used to view vessels in a selected region of the mesentery.
Hemodynamic and blood composition measurements.
RBC velocities (VRBC) were measured along the
vessel centerline by the two-slit photometric technique
(36) by online cross correlation of the photometric
signatures (IPM, model 102), and mean velocities
(Vmean) were obtained from the well-known
empirical relationship Vmean = VRBC/1.6 (2). Luminal diameters
(D) were measured by the video image shearing method (IPM,
model 908) (10). An index of was calculated by
analogy to that of a Newtonian fluid under conditions of Poiseuille
flow, i.e., = 8Vmean/D (15).
Systemic hematocrit and differential WBC counts were measured on blood
samples withdrawn from the carotid catheter. Plasma dextran
concentration was measured from 0.2-ml blood samples obtained via the
carotid artery by use of the anthrone reaction (31). In
brief, plasma from the sample was separated by centrifugation, diluted
1:2000 in phosphate-buffered saline (PBS; Sigma), pH = 7.4, and
added to a 0.2% solution of anthrone in concentrated sulfuric acid.
This solution was incubated in a hot-water bath at 95°C for 16 min,
after which point the reaction was quenched at 4°C by submersion of
the test tube in an ice bath. The absorbance of this solution was
measured at 626 nm (Shimadzu UV-160A spectrophotometer) and compared
with that prepared from standards of known dextran concentration.
WBC margination was quantified as the percentage
(FWBC) of the potentially maximum WBC flux
flowing within a venule that rolled along the EC lining a 120-µm-long
section of a venule with constant (within 5%) diameter
(8). The maximum WBC flux was estimated from the product
of systemic WBC concentration ([WBCsys]; no. per
mm3, determined by a Coulter Counter) and the venular bulk
volumetric flow rate, ignoring the Fahraeus effect for WBCs. The
marginating flux at the wall (fwall; cells/min)
was obtained from video recordings (Panasonic NV 8950) by counting of
the number of WBCs that rolled along the EC, past an arbitrary
reference point, per minute. The percentage of WBCs that marginated
were thus calculated from the expression
![F<SUB>WBC</SUB><IT>=</IT><FR><NU><IT>f</IT><SUB>wall</SUB></NU><DE>[WBC<SUB>sys</SUB>]<IT>V</IT><SUB>mean</SUB>(<IT>&pgr;D<SUP>2</SUP>/4</IT>)</DE></FR><IT>×100</IT>](/content/vol279/issue4/fulltext/H1460/img002.gif)
|
(1)
|
where Vmean is mean blood velocity (in
mm/s) and D is venular diameter (in mm). To enable a
comparison of changes in FWBC accompanying
reductions in 
and minimize its heterogeneity throughout the
microvascular network (8), FWBC was
normalized by dividing by FWBC at high values of
, on the order of 450 s1, i.e.,
FWBC* = FWBC/FWBC,
= 450.
Preparation of dextran solutions.
Dx40 (mol wt 37,500) or Dx500 (mol wt 428,000) solutions were prepared
for infusion into the animal from stock solutions made by dissolving
12 g dextran (Sigma Chemicals) in 25 ml of PBS. The dextran
solutions were then dialyzed overnight against PBS to remove any
impurities and concentrated to ~45% by dialysis against powdered
polyethylene glycol 8000 (Sigma).
Spectrophotometric determination of microvessel hematocrit and
RCA.
Differential spectrophotometry was used to measure the hematocrit
(17) and to provide an index of aggregation
(14) for RBCs within an individual microvessel. To this
end, the attenuation of light at two isobestic wavelengths (520 and 546 nm) was measured in real time with the use of two photomultiplier tubes
(PMTs), each with the appropriate interference filter placed between it and the light transmitted through the microvessel, as described in
detail previously (17). In brief, light transmitted
through a small blood vessel was projected onto the face of a silicon diode video camera. A right-angle microprism (0.7 mm on each leg), interposed between the top eyepiece of the microscope trinocular head
and the video camera, diverted light emanating from a 6 × 6 µm
square region (effective magnification) within the microscope focal
plane to a series of beam splitters and mirrors that steered the beam
of light to the PMTs. Thus each PMT provided a voltage signal
proportional to the light intensity transmitted through the
microvessel, i.e., I520 and I546. A third
photodetector consisted of a phototransistor coupled by a light guide
to the xenon lamp housing to provide a signal proportional to the
incident light intensities, I0-520 and
I0-546, at the two isobestic wavelengths, respectively.
Prior studies (16, 17) have demonstrated that the total
optical density (OD) of blood flowing in a microvessel,
ODtotal, results from the sum of an absorbance term
(ODabs) and a scattering term (ODscat), namely,
|
(2)
|
where OD is defined in terms of incident
(I0) and transmitted (I) light intensities as
log(I0/I). For two closely spaced isobestic wavelengths,
ODscat is equal at each wavelength, and the differential OD
[change in OD (
OD) = ODabs,546 
ODabs,520, where ODabs,546 and
ODabs,520 are the absorbance terms at wavelengths of 546 and 520 nm, respectively] is proportional to the product of
hemoglobin (Hb) concentration, the difference in molecular extinction
coefficients at the two wavelengths, and the pathlength of the
microvessel (diameter), thus yielding the relationship
|
(3)
|
where Hctmicro is the microvessel hematocrit,
D is the vessel diameter (in µm), and K is a
constant dependent on mean cell Hb concentration and the magnitude of
the extinction coefficients at each wavelength (17). In
vitro calibration studies, using small bore glass tubes, yielded a
value of K equal to 51.6 when Hctmicro was
expressed as a percentage (100 × packed cell fraction) (17).
To permit a rapid determination of the magnitude of ODscat,
ODtotal was monitored at 520 nm in real time by inputting
the PMT voltages (proportional to I520 and
I0-520) to an analog log-ratio module (Analog
Devices). Taking advantage of the fact that the molecular extinction
coefficients of Hb at 520 and 546 nm differ by a factor of ~2.0, such
that ODabs,546 
2ODabs,520, and then by
applying Eq. 2 at the two isobestic wavelengths, it is
easily shown (17) that ODscat = ODabs,520 OD.
In practice, the relative magnitudes of ODabs and
ODscat are dependent on the NA of the microscope objective.
As NA is increased, more light is captured by the objective and hence
the magnitude of ODscat is reduced. For the NA used here
(0.22), ODabs and ODscat comprised ~25 and
75% of ODtotal, respectively. This relatively large
contribution of the scattering component permits quantitative estimates
of the size of the scattering particle on the basis of the theory of
light scattering by "large tenuous scatterers" developed by
Twersky (34). As shown therein and validated by in vitro
studies of blood flow in small bore glass tubes (16, 17),
the scattering component of OD may be expressed as
![OD<SUB>scat</SUB><IT>=</IT>−log[<IT>10<SUP>−aX</SUP>+q</IT>(<IT>1−10<SUP>−aX</SUP></IT>)]](/content/vol279/issue4/fulltext/H1460/img006.gif)
|
(4)
|
where X = D(Hctmicro Hctmicro2), and a and q are
defined as
|
(5)
|
The parameter 
b
is the particle
thickness averaged over all spatial orientations; 
is the wavelength
of incident illumination; ni and
n0 are RBC interior and suspending media
refractive indexes, respectively; V is particle volume; 
/2 is the
acceptance one-half angle of the photodetector; S is the
particle surface area, and k is equal to
2
n0/. In physical terms, a
represents the attenuation of the transmitted beam of light due to
scattering at the cell-plasma interface of each particle, and
q represents the ability of the photodetector to accept
light obliquely scattered at an angle to the optical axis. As
schematized in Fig. 1A for a
suspension of dispersed RBCs within the focal plane, only
I0 refracted by individual particles within the acceptance
angle () of the photodetector (objective) will contribute to the
detected I. Hence, a total OD ensues that is greater than that based on
absorption alone for a solution of equal Hb concentration. When RCA
occurs, the effective particle size increases, changing its effective
surface area and volume as well as reducing the effective number of
particles for an equal total Hb concentration in the measuring volume
of the focal plane, as illustrated in Fig. 1B. Thus, in its
simplest terms, with the onset of aggregation, fewer particles are
present to refract light away from the of the objective, and hence
ODscat will tend to decrease relative to ODabs.
It should be noted that ODscat is a parabolic-like function
of the tube hematocrit, Hctmicro (16), and as
Hctmicro approaches 100%, ODscat falls to
zero. The net result of RCA is thus to alter the parameters
a and q as the effective particle diameter,
surface area, and volume change.
View larger version (27K):
[in this window]
[in a new window]
|
Fig. 1.
Schematic representation of the optical basis for the variation in
light scattering due to red blood cell (RBC) aggregation. A:
a disperse suspension of RBCs tends to refract the incident
illumination from the substage microscope condenser at an angle greater
than the acceptance angle () of the microscope objective.
B: with RBC aggregation in a suspension of equivalent total
hemoglobin concentration in the focal plane, particles with increased
surface area and volume refract less light outside the acceptance angle
of the objective, and a greater intensity of light is transmitted
through the suspension (lower scattering optical density). The
effective no. of particles per aggregate (G) was computed on
the basis of the theory of large tenuous scattering particles (see Ref.
34) by use of Eq. 6.
|
|
With a few well-chosen assumptions, one may solve for the volume of the
scattering particle by use of Eq. 4 and the relationship among a, q, S, and V as defined by
Eq. 5, given measurements of Hctmicro,
D, and ODscat. First, if it is assumed that RCA
may result in an average V composed of several RBCs, each of volume V0, then an index of aggregation (G) may be
determined from the ratio G = V/V0, which
reflects the equivalent number of RBCs required to satisfy Eqs.
4 and 5. Second, if it is assumed that b
represents the average diameter of either an individual RBC or a
spherical aggregate composed of G cells, then Eq. 4 may be rearranged to yield (14)
![OD<SUB>scat</SUB><IT>=</IT>−log[<IT>1−</IT>(<IT>1−q<SUB>0</SUB></IT>)<IT>G<SUP>−2/3</SUP></IT>(<IT>1−10</IT><SUP><IT>−a<SUB>0</SUB>G<SUP>1/3</SUP>X</IT></SUP>)]](/content/vol279/issue4/fulltext/H1460/img009.gif)
|
(6)
|
where a0 and q0
are the values of a and q, respectively,
corresponding to the nonaggregated monodisperse state where
G = 1. The values of a0 and
q0 have been determined previously (16, 17) for RBCs (cat blood) by use of glass tubes, with RBCs
suspended in Ringer solution (no aggregation). Similar in vitro
measurements of nonaggregating rat RBCs suspended in Ringer solution
were obtained here for blood flow in small bore glass tubes (40-70
µm in diameter), which gave a0 = 0.116 and q0 = 0.114. Thus, with measurements of
ODscat, Hctmicro, and D, the
G was obtained by the numerical solution (Newton-Raphson
method) of Eq. 6. Implicit to this formulation of
G is the assumption that the precise shape of the scattering particle, whether it be truly spherical or a population of rouleaux of
many orientations, does not affect the ability to empirically characterize the scattering cross section (34) of the
suspended particles as a function of prevailing concentrations and
values.
In vivo application of this procedure for determining G in
11 venules with > 350 s1 yielded values
of G equal to 2.2 ± 0.78 (SD) in the absence of dextran and 2.6 ± 0.55 (SD) in the presence of 3 g/100 ml Dx500, which were significantly greater than in vitro values of
G = 1 at similar values of Hctmicro and
. This discrepancy appeared to arise because of the
sensitivity of q to the mismatch in refractive index between
the walls of the glass tubes (n = 1.51) and the suspending media (n = 1.33). It was found that the
difference between glass and tissue refractive indexes affected
q0 more than a0, which
occurred because of a change in the ability to capture light scattered
at an angle to the optical axis. Thus, to compute in vivo values of
G from Eq. 6, the in vitro value of
a0 was used, and an average value of
q0 was computed for each in vivo experiment as
that value resulting in G = 1 for a disaggregated state
defined by the absence of dextrans with > 350s1.
The results from application of this technique to measure RCA in vitro
(12, 16) are shown in Fig. 2
for the flow of rat RBCs through a 50-µm-diameter glass tube.
Homogeneous (constantly stirred) suspensions of RBCs in Ringer
solution, either with 3% Dx500 or 1% albumin, at a 45% feed
hematocrit were fed into the tube from a pressurized reservoir, and the
mean RBC velocity and were varied by manipulation of the
perfusion pressure. In the case of the nonaggregating Ringer-albumin
solution, G remained invariant with reductions in
from 700 to <50 s1. In the presence of Dx500,
G rose 2.5 times as fell from 200 to 30 s1. To demonstrate the effect of Dx500 concentration on
RCA, presented in Fig. 2B is the variation of G
for concentrations of Dx500 up to 8 g/100 ml at = 50 s1. The occurrence of a maximum value of G at
a Dx500 concentration of ~3 g/100 ml is consistent with in vitro
findings by use of other techniques, such as direct counting of the
number of RBCs per aggregate (6), viscosity indexes
(6), erythrocyte sedimentation rate (6), or
devices based on optical techniques such as the Myrenne aggregometer
(37).
View larger version (22K):
[in this window]
[in a new window]
|
Fig. 2.
Representative variation of the index of aggregation
(G) for Wistar-Furth rat blood, at a feed hematocrit of 40%
flowing in a 50-µm-diameter glass tube, as a function of calculated
wall shear rate (; A) and dextran 500 (Dx500)
concentration in the suspending medium at = 50 s1 (B). In the absence of Dx500, G
remains invariant with but increases 2.5 times as
is reduced below 50 s1. The variation of G
with increasing concentration of Dx500 agrees favorably with trends
established by direct counting of the no. of RBCs per aggregate (see
Ref. 6), reaching a maximum at ~3 g/100 ml.
|
|
To illustrate the subtle variations in the luminal distribution of RBCs
in nonaggregating and aggregating conditions in a small venule (58-µm
diameter), presented in Fig. 3 are scenes from video recordings in the absence (Fig. 3A) and presence
(Fig. 3B) of 3 g/100 ml Dx500. Each scene was recorded by
instantaneously stopping the flow with a proximal microprobe to
eliminate motion artifacts and by capturing the scene before any
further occurrence of aggregation. Before flow cessation, was
~150 s1. In the absence of aggregation, where
G was arbitrarily set to equal 1, the RBCs are seen (Fig.
2A) to be uniformly distributed in the focal plane. In the
presence of Dx500, where G was ~2.0, the RBCs can be seen
(Fig. 2B) to take on a mottled appearance with the presence
of discrete aggregates.
View larger version (122K):
[in this window]
[in a new window]
|
Fig. 3.
Photographs of rat RBCs in a 58-µm-diameter
postcapillary venule that have been brought to a rapid stop after
flowing at a of ~150 s1 before (A)
and after (B) aggregation was enhanced by infusion (iv) of
Dx500 to reach a plasma concentration of 3 g/100 ml. The formation of
RBC aggregates produces a mottled appearance in B, where a
greater no. of larger aggregates can be seen after dextran infusion.
Systemic hematocrits were 47% (A) and 41% (B),
and microvessel hematocrit was estimated to be ~40 and ~33%,
respectively.
|
|
Quantitation of blood volume and red cell entrapment.
To examine shifts in systemic plasma volume and hematocrit accompanying
the infusion of Dx500, blood volumes were determined by injecting 270 µg of Evans blue dye, made to a 12.5% solution with saline, per
kilogram of body weight via the jugular vein. Plasma concentrations of
dye were determined from absorbance measurements at 608 nm from
arterial samples obtained over a 30-min period. The subsequent
exponential decay of plasma dye concentration was extrapolated back to
the time of injection to obtain plasma volume (PVdye). This
process was repeated after injections of Dx500 to determine the effects
of enhanced RCA. To minimize the volume removed from the rat during
plasma volume determinations, the supernatant from the three hematocrit
tubes (~0.2 ml of blood) used to obtain systemic hematocrit
(Hctsys) was collected for absorbance measurements by use
of 40-µl microcuvettes (Starna Cells) to establish PVdye
(plasma volume based on dye concentration). We also used
Hctsys to estimate plasma volume (PVhct), by
ignoring RBC sequestration and by assuming that PVhct = 0.08 × W × (1 Hctsys), where W is the
rat weight in grams. The difference between PVdye and
PVhct was taken as a measure of the volume of RBCs
sequestered in the circulation.
Experimental protocols.
To rule out the systemic effects of dextran toxicity, studies were
initially conducted to elucidate the effects of dextran infusion on
systemic arterial pressure (Part). Dx500 was administered by successive isovolemic exchanges of whole blood with 0.5-1.0 ml
of a 40- to 45-g/100 ml stock solution to raise circulating concentrations in 0.5- to 1.0-g/100 ml increments. Up to 10 exchanges, at 15-min intervals, were carried out before Hctsys was
measured, and supernatants were stored at 20°C for later
measurement of circulating dextran concentrations by the anthrone
method (31).
To examine the relationship between RCA and WBC margination, a single
postcapillary venule (22-65 µm in diameter) was located, and
flow conditions were measured at a position >1 mm downstream of a
bifurcation with no inlets or outlets between. A blunted glass
microprobe was held in a micromanipulator and positioned immediately
downstream of the bifurcation to vary .
Measurements of optical densities, for computation of
Hctmicro and the RCA index (G), were made as
was gradually reduced over a period of up to 70 min before
and after dextran infusion. During the reduction in , video
recordings of WBC margination were made over a 3-min period at each
for subsequent offline measurement of the number of WBCs
rolling or sticking to the vessel wall and the average WBC rolling
velocity. A sufficient amount of dextran was administered to yield a
circulating concentration of ~3 g/100 ml. This concentration of
dextran was chosen because it induces maximal levels of RCA (6,
37), and, because dextran precipitates fibrinogen in human blood
(27), we found for rat blood that 3 g/100 ml Dx500
precipitated <10% of fibrinogen, which we deemed acceptable.
| |
RESULTS |
Systemic effects of dextran infusion.
To establish that the infusion of dextran did not adversely affect
systemic hemodynamic parameters, alterations in systemic Part and Hctsys were monitored in response to
isovolemic dextran infusion, as shown in Fig.
4, A and B,
respectively. Initial studies using Sprague-Dawley rats
(n = 7) revealed that within 1 min of the first Dx500
exchange, a 50% drop in Part occurred that failed to rise
again in the 15-min interval before the next infusion. This initial
Dx500 exchange typically produced a circulating concentration of ~0.8
g/100 ml and a 10% rise in Hctsys, as measured before the
second exchange. With subsequent exchanges of Dx500, Part increased and eventually stabilized to a level similar to its initial
preexchange value, although Hctsys
progressively decreased. This response in Part was likely
due to an anaphylactoid reaction (11). Similar changes in
Part and Hctsys were found with Wistar Rats
(n = 2). In contrast, Wistar-Furth rats revealed no
significant change in Part or rise in Hctsys,
thus suggesting the absence of an anaphylactoid reaction; hence, this
strain of rat was used in the remainder of the studies. However, even
with Wistar-Furth rats, Hctsys decreased as a result of
fluid shifts and RBC entrapment (32).
View larger version (27K):
[in this window]
[in a new window]
|
Fig. 4.
Variations of arterial blood pressure (Part;
A) and systemic hematocrit (Hctsys;
B) in response to a succession of up to 10 bolus infusions
(iv) of Dx500 in Sprague-Dawley (n = 7), Wistar
(n = 2), and Wistar-Furth (n = 9) rats.
Shown are the means ± SE for n animals of each
strain.
|
|
The relative shifts in plasma, RBCs, and total blood volume concomitant
to Dx500 infusion were investigated by use of the Evans blue dilution
protocol. As shown in Fig. 5A,
a progressive decline in Hctsys occurred, as was
seen in Fig. 4B. Before Dx500 infusion, plasma,
RBC, and total blood volumes averaged 9.1 ± 2.3 (SD), 7.4 ± 2.1 (SD), and 15.4 ± 4.4 (SD) ml, respectively. The corresponding
drop in circulating RBCs and rise in plasma volume (Fig. 5B)
suggest that total blood volume remained fairly constant (dotted line
in Fig. 5B). The level of RBC sequestration increased in
proportion to Dx500 concentration (Fig. 5C) and reached 25%
of the total RBC blood volume at a Dx500 concentration of 3.5 g/100 ml.
View larger version (29K):
[in this window]
[in a new window]
|
Fig. 5.
Measurements of blood volume (BV) and Hctsys
at various concentrations of circulating Dx500. A:
Hctsys ( ) decreased with increasing Dx500
concentrations. B: plasma volume (PV) increased
monotonically with increasing Dx500 concentrations ( ), whereas
RBC concentration ([RBC]) diminished ( ) and total BV
remained constant (dotted line). Subscript zero, results normalized
with respect to initial values. C: volume of sequestrated
RBCs increased steadily with increasing Dx500 concentration
( ). The long dashed lines in A and
C are ±95% confidence limits of the linear regression of
the data.
|
|
Venular hemodynamics in response to Dx500.
Presented in Fig. 6 is the time course of
Hctmicro, VRBC, and
observed in 11 postcapillary venules in response to a single infusion
of Dx500 that resulted in a plasma concentration equal to 3.0 ± 0.5 g/100 ml. During both pre- and postinfusion periods of ~70 min,
these parameters remained invariant with time (t-test on
regression slope, P > 0.2). However, after
administration of Dx500, Hctmicro rapidly fell by almost
50%, whereas Hctsys decreased 35%. Although not
significant, RBC centerline velocity and the calculated Newtonian
also decreased by ~30%. A summary of the alterations in
hemodynamic parameters after Dx500 infusion is given in Table
1. Before and after Dx500
infusions, venule diameters were not significantly different (paired
t-test, P > 0.3). The average resting
(nonoccluded venule) centerline RBC velocity decreased significantly
after infusion (P < 0.002, paired t-test),
and a similar fall in occurred in response to Dx500 infusion.
After Dx500 infusion, systemic WBC count decreased significantly,
whereas differential counts of neutrophils and lymphocytes did not vary significantly from 22.0 ± 2.5 (SD) and 76.9 ± 2.3 (SD)%,
respectively, before Dx500 infusion, to 26.1 ± 3.6 (SD) and
73.2 ± 3.4 (SD)%, respectively (paired t-test,
P > 0.17).
View larger version (21K):
[in this window]
[in a new window]
|
Fig. 6.
Microvessel hematocrit (Hctmicro) before
() and after () a single intravenous infusion
of Dx500 to attain a systemic concentration of ~3.0 g/100 ml.
Hctmicro did not change significantly (t-test on
slopes, P > 0.2) with time either before or after the
exchange, but it fell abruptly after the dextran infusion. Values are
means ± SE, and the no. of measurements are given in parentheses
for a total of 11 postcapillary venules.
|
|
As shown in Table 1, the 20% reduction in systemic WBC count in
response to Dx500 infusion resulted in a 39% reduction in the
potential maximum WBC flux within the venule lumen (calculated using
Eq. 1). In contrast, the WBC flux marginating along the venular wall was not significantly altered, suggesting that the fraction of the luminal flux that marginated increased from 0.44 to
0.99% of the luminal flux. Thus the presence of the aggregating agent
Dx500 tended to increase the marginating flux under these resting high
conditions.
To further explore the role of RCA on WBC margination, hemodynamic
measurements were made in a second series of experiments to investigate
the disaggregating effect of Dx40, as shown in Table 1. Under similar
levels of systemic and microvascular hematocrits and with a
similar to that in the Dx500 experiments, a similar 20% reduction in
systemic WBC count was found. However, in contrast to the Dx500
experiments, the marginating flux decreased by ~15% (albeit
statistically insignificant), and the luminal flux that rolled along
the venular wall was unchanged.
To determine whether dextrans adversely affected the WBC population or
promoted WBC-EC adhesion, WBC rolling velocity along the EC, the number
of adherent WBCs per 100 µm of venule length, and the flux of WBCs
rolling along the EC were determined at high ( > 350 s1) and low ( < 250 s1)
values (Fig.
7). The high (>350
s1) values corresponded to those found in the normal
(resting) flow state, whereas the low (<250
s1) values were induced by a gradual reduction in venular
blood flow with a blunted microprobe. After Dx500 infusion, the average WBC rolling velocity decreased significantly by 45% at high and low
(Fig. 7A). The ratio of WBC rolling velocity
(VWBC) to calculated (VWBC/)
was calculated (Fig. 7B) to account for diminished shearing
forces acting on the WBC and revealed a significant decline at all
values in response to Dx500, although a significant reduction
after Dx40 infusion was found only at the high . The significance of this parameter is explored more fully in the
DISCUSSION. After Dx500 infusion, the average number of
WBCs adhering to the EC (per 100 µm of venule length) approximately
doubled (Fig. 7C) at low and high , whereas after
the Dx40 infusion, it increased significantly only at high .
The normalized fractional marginating flux of WBCs (Fig. 7D)
significantly increased and decreased at low for Dx500 and
Dx40, respectively, whereas no significant change was observed at high
.
View larger version (27K):
[in this window]
[in a new window]
|
Fig. 7.
A: white blood cell (WBC) rolling velocity
(VWBC). B: VWBC normalized with
respect to prevailing values. C: no. (#) of WBCs
sticking to the endothelium per 100-µm venule length
(WBCstick). D: WBC marginating flux
(FWBC), normalized with respect to flux at
= 450 s1
(FWBC* = FWBC/FWBC, = 450),
at low (<200 s1) and high (>350 s1)
values, in the absence of dextrans, and in the presence of
dextran 40 (Dx40) or Dx500 at concentrations of 3.0 and 3.2 g/100 ml,
respectively. Significantly different from control (*) or Dx40 ( ) by
use of Mann-Whitney rank sum test (P < 0.001). Values
are means ± SE, and the no. of measurements are given in
parentheses.
|
|
To determine whether dextran resulted in WBC activation, the nitro blue
tetrazolium test (24) was carried out on blood samples taken at the end of an experiment and on WBCs in freshly drawn Wistar-Furth rat blood samples, which were incubated with 3g/100 ml
Dx500 for 1 h. Neither study revealed any sign of WBC activation, nor was there a significant increase in the number of WBCs that showed
pseudopod formation after Dx500 infusion as observed at high
magnification under Nomarski differential interference contrast microscopy.
Effects of reductions on RCA and WBC margination.
In vivo measurements of G are shown in Fig.
8. In the absence of Dx500 (Fig. 8, solid
line), G remained invariant, with reductions in
from 600 to 200 s1 induced with a blunted microprobe
(rank sum test, P > 0.1). As was reduced
further, from 200 to 50 s1, a small yet significant 20%
increase in G was observed compared with values at high
values of between 400 and 500 s1 (rank sum
test, P < 0.001). After infusion of Dx500 to attain a
plasma concentration of 3.0 ± 0.5 (SD) g/100 ml (Fig. 8, dotted line), G increased 40% above control as fell to
250 s1 and then rose significantly by threefold with
further reductions in to 70 s1 (rank sum test,
P < 0.005). Also, Dx500 resulted in a slight elevation
of G for > 350 s1, which may
be the result of the reorientation or increased deformation of RBCs
caused by the increased plasma-dextran viscosity (see DISCUSSION for details).
View larger version (25K):
[in this window]
[in a new window]
|
Fig. 8.
Variation of G in 11 venules with reductions
in for the indicated plasma concentrations of Dx500.
Measurements of G for native plasma () did not
change significantly with reductions in induced by a blunted
microprobe, whereas after infusion of Dx500 (),
aggregation rose sharply as was reduced below 350 s1. * Significantly different from control by use of
Mann-Whitney rank sum test (P < 0.005). Nos. in
parentheses indicate the no. of venules followed by the no. of
measurements taken. Each point represents the mean ± SE.
|
|
The normalized index of WBC margination
(FWBC*) obtained for WBCs in the absence
of dextran was found to increase fourfold as was reduced from
600 to 50 s1 (Fig. 9).
After infusion of Dx500, FWBC* increased
significantly, almost eightfold, as was reduced to 50 s1 (rank sum test, P < 0.03).
FWBC* was also measured before and after a
single exchange of Dx40 (an anti-aggregating agent) to achieve a
circulating concentration of 3.1 ± 0.2 (SD) g/100 ml (Fig.
10). At > 350 s1, the values of FWBC*
obtained with Dx40 were similar to those obtained with Dx500, whereas
at < 350 s1, they were significantly
lower (rank sum test, P < 0.01). Under normal flow
conditions (nonoccluded vessel), RBC centerline velocity (Table 1) did
not significantly change when Dx40 was infused (paired
t-test, P > 0.3). However, after Dx40
infusion, systemic hematocrit dropped from 45.7 ± 1.5 (SD) to
30.9 ± 1.9 (SD)%, which interestingly was about the same as the
drop induced by Dx500. The absolute values of
FWBC for each of the these cases are presented in Table 1.
View larger version (23K):
[in this window]
[in a new window]
|
Fig. 9.
The margination of WBCs, corresponding to the aggregation
measurements of Fig. 8, is presented in terms of the WBC rolling flux
(F; in cells/min), normalized with respect to its value in
each of the 11 venules at high values of = 450 s1, i.e., FWBC* = F/F = 450. The
exponential increase in FWBC* with
decreasing in the absence of dextran () is
accentuated almost 2-fold with plasma levels of Dx500 equal to 3.0 g/100 ml. * Significant increase with Dx500, as assessed by the
Mann-Whitney rank sum test (P < 0.03). Nos. in
parentheses indicate the no. of venules followed by the no. of
measurements taken. Each point represents the mean ± SE.
|
|
View larger version (22K):
[in this window]
[in a new window]
|
Fig. 10.
After a single infusion of Dx40 to achieve a
plasma concentration of 3.2 g/100 ml (), WBC
margination (FWBC*) remained invariant
with reductions in below 350 s1 and was
significantly less (Mann-Whitney rank sum test, P < 0.01) compared with control () with decreasing . Nos.
in parentheses indicate the no. of venules followed by the no. of
measurements taken. Each point represents the mean ± SE.
|
|
| |
DISCUSSION |
Detection of aggregation in vivo.
To explore the relationship between RCA and WBC margination in vivo, a
new technique was employed for measuring the extent of RCA in
microvessels. G was derived from established principles of
the theory of light scattering by a suspension of translucent particles
(34). It should be noted, however, that although the present method draws heavily from a firm theoretical foundation, it
does rely on a certain level of empiricism. Most notable is the lack of
complete applicability of in vitro calibrations to the in vivo
environment, as evidenced by the necessity of defining a disaggregated
state and value for the parameter q (defined by Eq. 5 and appearing in Eq. 6 for obtaining G) at
high values. Because of differences between the refractive
indexes of glass tubes and the vessel wall and its surrounding tissue,
the effective light-gathering ability of the microscope optics differs
for the two applications. Nonetheless, the value G appears
to provide an index of RCA that is sensitive to the extent of
aggregation and its dependence, which under idealized
conditions may be interpreted as the number of RBCs per aggregate.
Physiological effects of dextrans.
Dextrans have been widely used in in vivo and in vitro studies of blood
rheology to induce RCA in an attempt to mimic a variety of blood
disorders, such as some hyperproteinemias and the low flow state.
However, it is well known that dextran infusion into laboratory animals
may induce a variety of adverse effects (11). To
characterize the potential for such effects in the present study, the
effects of dextran on three different strains of rat were investigated
to choose the best strain for this study. Both Sprague-Dawley and
Wistar rats developed an anaphylactoid reaction to systemic
administration of dextran, as evidenced by a precipitous fall in blood
pressure. Even though these rats appeared to adapt to the dextran in
their system, as evidenced by systemic blood pressure returning to
normal, it could not be assumed that the dextran had no other effects
on the circulation. Because no adverse reaction was found in the
Wistar-Furth rat, they were selected for this study of dextran-enhanced RCA.
It should be emphasized that although there have been many studies of
the aggregation of RBCs from different species, few have examined RCA
in rats and none have addressed the variability in RCA among strains of
rats routinely used in physiological studies. Baskurt et al.
(3) found that RBCs from Swiss-Albino rats aggregated only
slightly in native plasma, which agrees with the present findings in
Wistar-Furth rats. However, Ohta et al. (23) found that
RBCs from Wistar-Albino rats aggregate to a greater degree, similar to
the level of RCA for human RBCs. This disparity may also be due, in
part, to methodological differences as well as differences in strains
of rat.
On isovolemic exchange of concentrated dextran solutions with whole
blood, larger decreases in Hctsys were observed (Fig. 5)
than could be attributed to the removal of RBCs from the subject animal
during the exchange with cell-free dextran solutions. A series of blood
volume measurements were conducted that revealed a steady increase in
RBC sequestration with increasing Dx500 concentration, as previously
demonstrated by Simchon et al. (32) in studies of regional
blood flow in dogs. This loss of RBCs from the circulation may be due
to the increased level of RCA that causes rapid RBC settling in regions
of low flow and small vessel plugging, or possibly to cell-dextran-EC
interactions (33). In addition, alterations in colloid
osmotic pressure may have also contributed to a generalized fluid shift
from tissue to vascular compartments that resulted in the increased
plasma volume and diminished Hctsys.
In vivo levels of RCA.
The major goal of the present study has been to elucidate the extent to
which RCA influences WBC margination in postcapillary venules. For
whole rat blood, where RBCs aggregate only slightly, WBC margination
was found to increase exponentially as decreased (8). When aggregation was enhanced with Dx500, the
exponential increase was twofold greater compared with that in the
absence of Dx500 as was reduced below 350 s1
(Fig. 9). When RCA was inhibited with Dx40, WBC margination was greatly
reduced, and fairly constant, below 350 s1 (Fig. 10).
These trends were consistent with the measured exponential increase in
RCA as fell below 350 s1 (Fig. 8).
Prior in vivo studies of the effect of vessel hematocrit
Hctmicro on WBC margination (8) have suggested
that the flux of WBCs at the venular wall is mainly invariant with
alterations of Hctmicro at high and for 15% 
Hctmicro45%. As shown therein, significant variations in
marginating WBC flux with Hctmicro were found only at the
lowest values (<100 s1) and when
Hctmicro exceeded 50% at > 100 s1. In vitro studies by Goldsmith and Spain
(9) showed that at low , with decreasing
hematocrit, there was an increasing enrichment of WBCs relative to
RBCs, but the enrichment of WBCs themselves did not significantly vary
with decreasing hematocrit. Similar results by Nobis et al.
(22) in glass tubes support the general conclusion that
the enhanced margination of WBCs observed here results primarily from
the effect of RCA in contrast to hematocrit variations. Hence the
decrease in Hctmicro caused by the presence of dextran may
play an insignificant role in the margination process, and thus
-dependent RCA appears to be the primary determinant of
increased WBC margination with flow rate reductions.
The present results also suggest that RCA plays a dominant role in
maintaining the marginating WBC flux along venular EC. Although it has
been shown by Schmid-Schönbein et al. (30) that as
many as 94% of all WBCs that exit the true capillaries make rolling
contact with the venular EC, it is apparent that the majority of these
WBCs rapidly mix with the RBC stream unless they are excluded from the
axial stream by RBC aggregates. Enhancement of the WBC flux in the
vicinity of the tube wall near the periphery of the RBC core has been
shown by Nobis et al. (22) to occur in vitro with
reductions in . The twofold increase in rolling WBC flux
observed here (Fig. 9) with the presence of the proaggregating agent
Dx500 (for < 100 s1) and its almost
complete obliteration at all values with the presence of the
disaggregating agent Dx40 (Fig. 10) clearly support a role for RCA in
maintaining the in vivo rolling flux of WBCs. It appears that RCA
either serves to maintain the rolling flux from capillary exit to
venular EC or promotes the radial migration of WBCs that have been
previously entrained by the RBC stream after exit from the capillaries.
In the in vitro studies of Nobis et al. (22), the relative
WBC flux rolling along the wall itself was markedly reduced at all
values, presumably because of the lack of adhesive
interactions between WBCs and tube wall. To explore the role of dextran
on the adhesive interactions between WBCs and EC, the absolute rolling velocity and marginating flux of WBCs were studied. The nearly twofold
rise in WBC flux (compared with control) at low shear in response to
the Dx500 (Fig. 9) and the concomitant threefold increase in the level
of RCA (Fig. 8) suggest that RCA is a major promoter of WBC margination
and subsequent firm adhesion. To explore the effect of dextran on the
adhesive interactions between WBCs and EC, the absolute rolling
velocities and marginating flux of WBCs were studied. As indicated in
Fig. 7A, the significantly lower rolling velocity in the
presence of Dx40, compared with control, and its further attenuation
with Dx500 are suggestive of an enhanced strength of the adhesive bond
between WBCs and EC. Normalization of the VWBC, by division
by , to account for alterations in the forces that tend to
sweep WBCs from the EC (Fig. 7B) yields a consistent
indicator of enhanced adhesive strength, except for the case of low
values in the presence of Dx40. A similar enhancement of the
number of WBCs firmly adhered to the EC also suggests an increased
strength of the WBC-EC adhesive bond (Fig. 7C). However, the
fact that the rolling flux is largely unaffected at high
values for either the disaggregating agent (Dx40) or the proaggregating
agent (Dx500) and is significantly reduced in the presence of Dx40 and
increased with Dx500 at low values suggests that the apparent
increase in adhesion arises because of a greater degree of margination
and a greater probability of contact between WBCs and EC. It is likely
that the reductions in VWBC reflect a shift in the
composition of the marginating population of WBCs to include a greater
proportion of neutrophils, with specific receptors for ligands on the
EC surface, as RCA promotes an increase in the total fraction of all
WBCs that maintain their contact with the EC. The data of Fig.
7B (VWBC/) support this hypothesis. It
has been shown previously (13) that the average rolling
velocity of a population of marginating WBCs is less sensitive to
increases in the or red blood cell velocity (i.e., strength
of the WBC-EC bond) compared with the rolling velocity of an individual
WBC, because of the heterogeneity of adhesive and mechanical properties
of the circulating WBC population.
The increased amounts of firmly adhered WBCs and decreases in
VWBC could have been caused by a number of additional
factors such as tissue changes during the experiment, EC activation, or interference with the normal balance of receptor-ligand interaction at
the WBC-EC interface. However, it seems improbable that increased WBC
adhesion and decreased WBC flux could be the basis for both Dx500 and
Dx40 results. Two additional factors may preclude a role for
dextran-induced EC activation. First, WBC activation would lead to a
large systemic loss of WBCs, and, second, WBC activation would lead to
an increase in WBC margination at higher . Although it may
still be argued that Dx500 might enhance WBC-EC interactions and Dx40
might inhibit them, this possibility appears unlikely because
margination increases at the point at which RCA increases.
The effect of RCA on the resistance to flow in vivo has also been
fraught with controversy. Mchedlishvili et al. (18) found that after infusion of large concentrations of Dx500 into rats, systemic blood pressure increased by ~50%, which was attributed to the increased resistance caused by dextran-induced aggregation. In
the present study, no such increase in systemic pressure was found
after exchange of Dx500 into Wistar-Furth rats, whereas systemic blood
pressure initially decreased for Sprague-Dawley and Wistar rats. Cabel
et al. (5) measured resistance in the circulation of
isolated cat lateral gastrocnemius muscle with 0.5% of dextran 250 present and found venous resistance increased by 30% at high
values and decreased by 80% at low values. In vitro studies
of RCA in vertically positioned small bore glass tubes (7, 28,
29) showed that the apparent viscosity of nonaggregating RBCs
increased threefold as was reduced below 100 s1. When aggregation was increased to normal levels, by
suspending RBCs in plasma, the apparent viscosity changed very little
with decreasing in small (<58 µm) tubes. However, prior in
vivo measurements in the low flow state suggest a dramatic rise in apparent viscosity with reductions in (12).
The current findings of a 28% decrease in RBC centerline velocity
after Dx500 exchange (Table 1), which agrees with Mchedlishvili et al.
(18) and others, suggest that RCA leads to an
increase in microvascular resistance. However, these reductions in
velocity may arise from the increased viscosity of the suspending
medium (
0) after infusion of the Dx500. Because
Poiseulle's law predicts that, in a given vessel and at a given
pressure gradient, volume flow rate is inversely proportional to
viscosity, the effect of the Dx500 may be estimated from the
relationship between bulk viscosity (), 0, and
Hctmicro. Direct in vivo measurements in unbranched
microvessels with blood flow at high values (17) suggest that
|
(7)
|
where a = 0.04 as determined for cat blood
(similar in mechanical properties to rat). With the use of values of
0 = 1.2 mPa · s for rat plasma
(38) and 2.29 mPa · s for 3.0 g/100 ml Dx500
(M. J. Pearson and M. W. Rampling, unpublished
work), an increase in blood viscosity equal to 33% would
follow the infusion of Dx500 and 38% reduction in Hctmicro
noted in Table 1. Thus on the basis of bulk viscometry, RCA would have
very little effect on resistance within individual microvessels at
physiologically relevant values. However, the disruption of
red blood cell aggregates in arteriolar bifurcations and their
formation in venous confluences may introduce an additional source of
increased resistance to flow attendant to RCA, which remains to be
determined by experiments that focus directly on alterations in the
resistance to flow.
In summary, it has been shown that RCA greatly enhances WBC margination
attendant to a reduction in below 350 s1, and
for values above 350 s1, there is little
indication that RCA progresses to the point of affecting WBC
margination. Thus the role of RCA may be of prime importance in
disorders that manifest a combined reduction in values and
increased tendency toward WBC-EC adhesion, such as inflammation, septic
shock, and ischemia. In this regard, the present study may also serve
to establish a framework for evaluating the contribution of a broad
variety of hematological and hemorheological disturbances, such as
elevated levels of fibrinogen, the principal promotor of RCA,
polycythemia, and abnormal RBC deformability and surface charge. With
the availability of the new technique presented here for the
measurement of RCA, it is anticipated that greater insights into these
pathophysiological disturbances may be gained from future studies by
use of the techniques of direct intravital microscopy.
| |
ACKNOWLEDGEMENTS |
We thank Karen Trippett for technical assistance.
| |
FOOTNOTES |
This work was supported by National Heart, Lung, and Blood Institute
Grants HL-28381 and HL-39286.
Address for reprint requests and other correspondence: H. H. Lipowsky, Bioengineering Program, The Pennsylvania State Univ., 205 Hallowell Bldg., University Park, PA 16802 (E-mail:
hhlbio{at}engr.psu.edu).
The costs of publication of this
article were defrayed in part by the
payment of page charges. The article
must therefore be hereby marked
"advertisement"
in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Received 7 January 2000; accepted in final form 6 April 2000.
| |
REFERENCES |
1.
Bagge, U,
and
Karlsson R.
Maintenance of white blood cell margination at the passage through small venular junctions.
Microvasc Res
20:
92-95,
1980[ISI][Medline].
2.
Baker, M,
and
Wayland H.
On-line volume flow rate and velocity profile measurement for blood in microvessels.
Microvasc Res
7:
131-143,
1974[ISI][Medline].
3.
Baskurt, OK,
Farley RA,
and
Meiselman HJ.
Erythrocyte aggregation tendency and cellular properties in horse, human, and rat: a comparative study.
Am J Physiol Heart Circ Physiol
273:
H2604-H2612,
1997[Abstract/Free Full Text].
4.
Blixt, A,
Johnson P,
Braide M,
and
Bagge U.
Microscopic studies on the influence of erythrocyte concentration on the post-junctional radial distribution of leukocytes at small venular bifurcations.
Int J Microcirc Clin Exp
4:
141-156,
1985[ISI][Medline].
5.
Cabel, M,
Meiselman HJ,
Popel AS,
and
Johnson PC.
Contribution of red blood cell aggregation to venous vascular resistance in skeletal muscle.
Am J Physiol Heart Circ Physiol
272:
H1020-H1032,
1997[Abstract/Free Full Text].
6.
Chien, S,
and
Jan KM.
Ultrastructural basis of the mechanism of rouleaux formation.
Microvasc Res
5:
155-166,
1973[ISI][Medline].
7.
Cokelet, GR,
and
Goldsmith HL.
Decreased hydraulic resistance in the two-phase flow of blood through small vertical tubes at low flow rates.
Circ Res
68:
1-17,
1991[Abstract/Free Full Text].
8.
Firrell, JC,
and
Lipowsky HH.
Leukocyte margination and deformation in mesenteric venules of rat.
Am J Physiol Heart Circ Physiol
256:
H1667-H1674,
1989[Abstract/Free Full Text].
9.
Goldsmith, HL,
and
Spain S.
Margination of leukocytes in blood flow through small tubes.
Microvasc Res
27:
204-222,
1984[ISI][Medli
TOP
ABSTRACT
INTRODUCTION
