influence of local haemodynamics on leucocyte rolling and chemoattractant-induced firm adhesion in...
TRANSCRIPT
In¯uence of local haemodynamics on leucocyte rolling
and chemoattractant-induced ®rm adhesion
in microvessels of the rat mesentery
X . X I E , P . H E D Q V I S T and L . L I N D B O M
Department of Physiology & Pharmacology, Karolinska Institutet, Stockholm, Sweden
ABSTRACT
Tissue hyperaemia, oedema formation and leucocyte accumulation are characteristic features of the
inflammatory process referable to changes at the microcirculatory level. Here, we used intravital
fluorescence video microscopy to assess relationships among haemodynamical parameters,
leucocyte rolling, and chemoattractant-induced firm adhesion in small venules (13)24 lM) of the rat
mesentery. The rolling leucocyte ¯ux in these vessels was directly proportional to the total leucocyte
¯ux (r � 0.76, P < 0.001), which in turn closely correlated to the venular blood ¯ow (r � 0.77,
P < 0.001). Consequently, the rolling to total leucocyte ¯ux fraction, averaging 39 � 15%, did not
vary with the blood ¯ow and showed no correlation to either blood ¯ow velocity (r � )0.15,
P � 0.42) or wall shear rate (r � )0.06, P � 0.77), indicating that the extent of leucocyte rolling is
not primarily dependent on the ¯uid viscous drag at physiological blood ¯ow rates in vivo. Stimulation
of the mesentery with the chemoattractant fMLP (10)6M) induced ®rm adhesion of rolling
leucocytes. It was found that the number of adherent leucocytes in individual vessels was directly
related to the rolling leucocyte ¯ux (r � 0.78, P < 0.001) and hence to the venular blood ¯ow
(r � 0.47, P < 0.05), while there was no correlation to the wall shear rate (r � 0.27, P � 0.24). The
dependence of the ®rm adhesive response on the blood ¯ow level and the delivery rate of leucocytes
was con®rmed at the whole organ level. Thus, leucocyte accumulation in rat skin lesions was
markedly enhanced when a vasodilator was co-administered with the chemotactic stimulus
compared with chemotactic stimulation alone. The data indicate that, within a physiological blood
¯ow range, the leucocyte response to chemotactic stimulation is largely independent of the prevailing
hydrodynamic shear forces. Instead, manifestation of the ®rm adhesive response, because of its
dependence on the preceding rolling interaction, is clearly related to the blood ¯ow level in the
microvessels, which emphasizes the signi®cance of tissue hyperaemia in in¯ammation.
Keywords blood ¯ow, in¯ammation, intravital microscopy, leucocyte/endothelium interactions, rat
mesentery.
Received 3 February 1998, accepted 5 October 1998
Redness, heat and tissue swelling are clinical signs of
in¯ammation which can be referred to changes at the
microcirculatory level, namely arteriolar dilatation
leading to increased local blood ¯ow, plasma exudation
resulting from increased vascular permeability, and
accumulation of circulating leucocytes. A complex
pattern of mediator actions governs these microvas-
cular events, which in turn mutually may in¯uence
each other. Thus, hyperaemia may potentiate not only
¯uid transfer across the vessel wall but also the
recruitment of leucocytes to the extravascular tissue. It
is clearly indicated by data obtained in various in vivo
models that co-stimulation with vasodilating agents
and chemotactic mediators results in enhanced leuco-
cyte accumulation compared with chemotactic stimu-
lation alone (Issekutz 1981, Issekutz & Movat 1982,
Downey et al. 1988, Raud et al. 1988, Texieira et al.
1993), presumably due to increased delivery rate of
leucocytes with increasing tissue blood ¯ow. On the
other hand, high microvascular ¯ow rates may in¯u-
ence the leucocyte extravasation process unfavourably,
as an increased hydrodynamic force will make it more
dif®cult for the leucocyte to establish ®rm contact with
the vessel wall.
Correspondence: Lennart Lindbom, Department of Physiology and Pharmacology, Karolinska Institutet, S-171 77 Stockholm, Sweden.
Acta Physiol Scand 1999, 165, 251±258
Ó 1999 Scandinavian Physiological Society 251
Leucocyte extravasation is a multistep process
accomplished through sequential adhesion receptor-
dependent interactions with vascular endothelium and
extravascular matrix components (Butcher 1991, Carlos
& Harlan 1994). Slow rolling of the leucocytes along
the venular endothelium, mediated by the selectin
family of adhesion molecules, is the initial event in this
process and a necessary precondition for subsequent
(integrin-mediated) ®rm attachment and transmigration
(Lawrence & Springer 1991, Lindbom et al. 1992, von
Andrian et al. 1992). Through the rolling interaction,
retardation of the fast-¯owing leucocyte is achieved,
enabling the leucocyte to establish ®rm contact with the
vessel wall in the presence of high ¯uid shear. Although
binding via the selectins appears to be relatively shear-
resistant, inverse relationships between leucocyte roll-
ing and wall shear rate has been indicated both in vitro
(Lawrence et al. 1987, Lawrence & Springer 1991) and
in vivo (Atherton & Born 1973, Firrell & Lipowsky 1989,
Ley & Gaehtgens 1991, Damiano et al. 1996). In view
of the close relationship between the extent of leuco-
cyte rolling and the ®rm adhesive response induced by
chemoattractants (Lindbom et al. 1992), these data
suggest that leucocyte adhesion, and hence recruitment
to tissues, will be impeded by increased ¯uid shear.
However, such an implication is dif®cult to reconcile
with the ®ndings of the functional studies referred to
above, i.e. potentiation of leucocyte recruitment to
tissues by increased blood ¯ow.
The present study aimed at investigating relation-
ships between local haemodynamics and leucocyte/
endothelium interactions (leucocyte rolling and ®rm
adhesion) in small venules of the rat mesentery. Direct
intravital microscopic observations in this model
showed that the leucocyte rolling fraction did not vary
in parallel with the blood ¯ow or the wall shear rate in
the microvessels. Instead, there was a proportional
relationship between rolling leucocyte ¯ux and venular
blood ¯ow. In congruity, because of its dependence on
the preceding rolling interaction, the ®rm adhesive
response to chemoattractant stimulation was found to
also correlate with the microvessel blood ¯ow. These
data emphasize the signi®cance of tissue hyperaemia in
in¯ammation for manifestation of the leucocyte
response in host defense and immune reactions.
MATERIALS AND METHODS
Animals
Adult Wistar rats were used in this study. The rats were
anaesthetized with equal parts of ¯uanison/fentanyl
(Hypnorm, 10/0.2 mg mL)1; Janssen Pharmaceutica,
Beers, Belgium) and midazolam (Dormicum,
5 mg mL)1; Hoffman-La Roche, Basel, Switzerland)
diluted 1:1 with sterile water [0.2 mL (100 g body
wt))1 intramuscularly]. The trachea was cannulated to
facilitate spontaneous breathing. A catheter was placed
in the left femoral vein or the left jugular vein for i.v.
administration of supplementary doses of anaesthesia.
Body temperature was maintained at 37 °C by a heating
pad connected to a rectal thermistor.
Intravital microscopy
Laparotomy was performed by a midline incision, and a
segment of the ileum was exteriorized from the peri-
toneal cavity and placed on a heated transparent ped-
estal to allow microscopic observation of the
mesenteric microcirculation. The exposed tissue was
superfused with a thermostated (37 °C) bicarbonate-
buffered saline solution (composition in mM: NaCl 132,
KCl 4.7, CaCl 2.0, MgSO4 1.2, NaHCO3 18) equili-
brated with 5% CO2 in nitrogen to maintain physio-
logical pH. Observations of the mesenteric
microcirculation were made using a Leitz Orthoplan
microscope equipped with water immersion lenses
(´ 25, NA 0.6; ´ 55, NA 0.8). The microscopic image
was televised (Panasonic WV-1550, Panasonic WV-
1900 low light camera) and recorded on video tape
(Panasonic NV-F 100 S-VHS) via a time/date genera-
tor (Panasonic WJ-810) for subsequent off-line analysis.
After positioning under the microscope, a 20±30 min
stabilization period preceded quantitative measure-
ments. In order to detect the free-¯owing leucocytes in
the observed vessel segments, acridine orange
[5 mg (kg body wt))1; Sigma Chemical, St. Louis, MO]
was given intravenously to label the circulating leuco-
cytes, and short periods of ¯uorescent light epi-illumi-
nation (Leitz Ploemopak, ®lter block I2) was applied.
Analyses of blood ¯ow parameters, leucocyte ¯ux and
leucocyte/endothelium interactions (rolling and ®rm
adhesion) were made in small venules (inner diameter
15±25 lm) with stable resting blood ¯ow.
Determination of leucocyte ¯ux and blood ¯ow parameters
Rolling leucocyte ¯ux in the venules was determined
from the recorded video images by counting the num-
ber of rolling leucocytes per minute passing a reference
point in the microvessel under normal light transillu-
mination. The corresponding ¯ux of free-¯owing leu-
cocytes was similarly determined in the ¯uorescent
scenes. The leucocyte rolling fraction was calculated as
the rolling cell ¯ux in percentage of the total leucocyte
¯ux (rolling plus free-¯owing cells). Individual leucocyte
rolling velocities were calculated from the time required
for steady rolling leucocytes to travel a de®ned distance
(100±200 lm) in the microvessel. Distance was mea-
sured by hand with a caliper directly on the TV-screen.
Microvascular blood ¯ow and leucocyte/endothelium interactions � X Xie et al. Acta Physiol Scand 1999, 165, 251±258
252 Ó 1999 Scandinavian Physiological Society
The mean rolling velocity in each vessel was determined
as the average of »10 leucocytes passing in the venule.
Detection of the free-¯owing leucocytes in ¯uorescent
light also permitted, through frame-by-frame analysis,
determination of the velocity of individual free-¯owing
leucocytes by measuring the distance travelled between
two or more successive video frames. The highest cell
velocity among >15 leucocytes analysed in a given
moment was considered to correspond to the maximal
¯ow velocity (vmax) in the vessel, and was used to esti-
mate the mean blood ¯ow velocity (vmean) according to
the relationship vmean � vmax/{2 ) (//2r)2} where /represents leucocyte diameter (»7 lm for the rat) and r
is the vessel radius (Ley & Gaehtgens 1991). Mean
blood ¯ow velocity and vessel radius were used to cal-
culate volume blood ¯ow (Q , nL s)1) and wall shear
rate (c, s)1) according to the equations: Q � pr2vmean
and c � 8vmean/2r, respectively. Wall shear stress
(dyn cm)2) was calculated as the product of viscosity
and shear rate, assuming a constant viscosity of 0.025 P
(Lipowsky et al. 1980).
Leucocyte ®rm adhesion
After recording for analysis of blood ¯ow and leucocyte
¯ux in resting tissue, leucocyte ®rm adhesion was
induced by the chemotactic peptide N-formyl-
methionyl-leucyl-phenylalanine (fMLP; Sigma) admini-
stered topically to the mesentery for 6 min via the
superfusion buffer at a ®nal concentration of 10)6M.
At the end of the stimulation period, the number of
adherent leucocytes (stationary for >1 min) per
100 lm venule length was counted. During the fMLP
stimulation period, rolling velocity of the leucocytes
which became ®rmly attached within the ®eld of
observation was determined through reverse video
play-back analysis of the distance travelled by these cells
every half second prior to their ®rm arrest. Rolling
velocity of those leucocytes which did not adhere
within the same venular segments was similarly deter-
mined.
Skin lesion experiments
In a different group of experiments, skin lesions were
induced in four rats with their backs shaved 16 h prior
to the experiment. In these experiments, zymosan-
activated rat plasma (ZAP) was used as the chemotactic
stimulus. Plasma of heparinized rat blood was incu-
bated with zymosan (10 mg mL)1; Sigma) for 60 min
at 37 °C resulting in complement activation and gen-
eration of C5a. Intradermal injections (100 lL) of PBS,
prostaglandin E2 (PGE2, 10)6M), ZAP (undiluted), or
ZAP plus PGE2 were made in duplicate at alternate
sites in the dorsal region of the animals. The animals
were killed 4 h later with an overdose of anaesthetic,
and the skin lesions were carefully removed and dis-
sected free from fat and muscle tissue. For quanti®ca-
tion of leucocyte in®ltration in the skin lesions, the skin
samples were homogenized in 10 mL 0.5% hex-
adecyltrimethylammonium bromide, freeze-thawed and
centrifuged, whereafter the myeloperoxidase (MPO)
activity of the supernatant was determined. MPO is
abundant in neutrophils and found to be a reliable
marker for neutrophil accumulation in in¯amed tissues
(Lundberg & Arfors 1983). The enzyme activity was
determined spectrophotometrically as the change in
absorbency at 650 nm (20 °C) that occurs in the redox
reaction of H2O2-tetramethylbenzidine catalysed by
MPO, and expressed as units MPO g tissue)1.
Statistics
Values are presented as mean � SD. Calculation of
statistical signi®cance was performed with the
Wilcoxon test for paired samples and the Mann±
Whitney U-test for independent samples. Correlation
between variables was assessed according to Pearson's
product moment correlation.
RESULTS
Leucocyte/endothelium interactions and blood ¯ow
parameters were measured in small venules of the rat
mesentery exposed for intravital microscopy. A total of
30 venules (mean i.d., 18.2 � 3.5 lm; range,
13±24 lm) in 15 rats were used for analysis. Blood ¯ow
velocity in these vessels ranged from 0.1 to 1.4 mm s)1
(mean, 0.7 � 0.3 mm s)1) yielding wall shear rates and
blood ¯ows in the range 46±558 s)1 (mean,
288 � 116 s)1) and 0.02±0.45 nL s)1 (mean, 0.19 �
0.14 nL s)1), respectively. Correspondingly, calculated
wall shear stress varied between 1.2 and 14.0 dyn cm)2
(mean, 7.2 � 2.9 dyn cm)2 (Table 1).
Correlation analysis of ¯ow parameters in individual
vessels revealed that the total leucocyte ¯ux (rolling
plus free-¯owing cells) in the venule was closely related
to the venular blood ¯ow (r � 0.77, P < 0.001). The
rolling leucocyte ¯ux in these vessels ranged from 2 to
53 cells min)1 and was directly proportional to the total
leucocyte ¯ux (r � 0.76, P < 0.001) (Fig. 1a), and,
accordingly, related also to the blood ¯ow (r � 0.51,
P < 0.01) (Fig. 1b). Thus, within the physiological
blood ¯ow range observed, the rolling to total leucocyte
¯ux fraction, averaging 39 � 15% (range, 18±75%), did
not vary in parallel with the venular blood ¯ow
(r � )0.22, P � 0.25). Moreover, leucocyte rolling
fraction showed no correlation to either blood ¯ow
velocity (r � )0.15, P � 0.42) or wall shear rate
(r � )0.06, P � 0.77) (Fig. 2a), indicating that the
Ó 1999 Scandinavian Physiological Society 253
Acta Physiol Scand 1999, 165, 251±258 X Xie et al. � Microvascular blood ¯ow and leucocyte/endothelium interactions
extent of leucocyte rolling is not primarily dependent
on the hydrodynamic force exerted by the blood stream
at physiological ¯ow rates in vivo. Likewise, mean leu-
cocyte rolling velocity, ranging 5.1±37.6 lm s)1
(19.7 � 7.7 lm s)1), showed no correlation to blood
¯ow velocity (r � 0.14, P � 0.48) or wall shear rate
(r � 0.22, P � 0.25) (Fig. 2b).
In resting tissue, only occasional cells were found to
spontaneously adhere to the venular endothelium
during observation periods of 1±2 h. Stimulation with
fMLP (10)6M) induced prompt leucocyte adhesion,
resulting in 16.5 � 7.8 adherent cells per 100 lm ve-
nule length. The number of adherent cells in individual
vessels correlated strongly with the rolling leucocyte
¯ux (r � 0.78, P < 0.001) (Fig. 3) and to a lesser ex-
tent with the venular blood ¯ow (r � 0.47, P < 0.05).
On the other hand, there was no correlation to the wall
shear rate (r � 0.27, P � 0.24). It was found that
leucocytes which adhered to the endothelial lining
(®rmly attached for >1 min) were recruited exclusively
from the rolling population. Moreover, there was a
clear predominance of slowly rolling cells among those
leucocytes which were found to adhere. Thus, the
rolling velocity of those leucocytes which came to arrest
in the observed vessel segment was on average
12.8 � 6.8 lm s)1 whereas leucocytes which did not
adhere, but continued to roll in this segment in its
whole length, were rolling with a signi®cantly higher
velocity, i.e. 32.1 � 16.4 lm s)1 (P < 0.001) (Fig. 4).
Typically, the ®rm adhesive event was not preceded by
Figure 1 Rolling leukocyte ¯ux vs. total leucocyte ¯ux (a) and blood
¯ow (b) in rat mesenteric venules. Linear regression line is superim-
posed.
Figure 2 Leucocyte rolling fraction (a) and leucocyte rolling velocity
(b) as a function of wall shear rate in rat mesenteric venules. Each
data point in panel (b) represents the average velocity of »10 rolling
leucocytes analysed in each venule. Linear regression line is super-
imposed.
Table 1 Microvascular and leucocyte parameters in rat mesenteric
venules
Blood ¯ow velocity (mm s±1) 0.66 � 0.31 (0.09±1.39)
Diameter (lm) 18.2 � 3.5 (13±24)
Wall shear rate (s±1) 288 � 116 (46±558)
Wall shear stress (dyn cm±2) 7.2 � 2.9 (1.2±14.0)
Volume blood ¯ow (nL s±1) 0.19 � 0.14 (0.02±0.45)
Total leucocyte ¯ux (cells min±1) 54 � 29 (10±118)
Rolling leucocyte ¯ux (cells min±1) 20 � 13 (2±53)
Leucocyte rolling fraction (%) 39 � 15 (18±75)
Leucocyte rolling velocity (lm s±1) 19.7 � 7.7 (5.1±37.6)
Values are mean � SD. Range is shown in brackets.
254 Ó 1999 Scandinavian Physiological Society
Microvascular blood ¯ow and leucocyte/endothelium interactions � X Xie et al. Acta Physiol Scand 1999, 165, 251±258
a gradual decline in rolling velocity, but rather, rolling
velocity was quite uniform until 1±2 s prior to the ®rm
arrest. The relationship between rolling velocity and
®rm attachment suggests that a critical rolling velocity
level exists which must not be exceeded in order to
permit transition from a rolling to a ®rm adhesive state.
By statistical means, under the present conditions in the
rat mesentery, an upper limit for rolling velocity at
26 lm s)1 (given by the mean + 2SD) would encom-
pass >95% of the cells capable of adhering in the small
venules studied.
The observed relationship between the (chemoat-
tractant-induced) ®rm adhesive response and the
delivery rate of leucocytes in individual microvessels of
the rat mesentery clearly suggests, if extrapolated to the
whole organ level, that tissue blood ¯ow is an
important factor in the regulation of leucocyte
recruitment to in¯amed tissue sites. In separate
experiments, the existence of such relationship in rat
tissue was veri®ed. In these experiments, zymosan-
activated plasma (ZAP) was used as chemotactic stim-
ulus, because of the documented ef®cacy of C5a
(generated in plasma by zymosan) to induce profound
tissue accumulation of leucocytes (Arfors et al. 1987).
Intradermal injection of ZAP resulted in a close to 3-
fold increase in leucocyte accumulation (as determined
by MPO activity) in the skin lesions [1.32 � 0.77 units
MPO (g tissue))1] compared with PBS injection
[0.51 � 0.21 units MPO (g tissue))1]. Injection of
prostaglandin E2 (PGE2), a vasodilator which does not
induce leucocyte accumulation per se, yielded MPO
activity [0.45 � 0.32 units MPO (g tissue))1] similar to
that for saline injection. However, the leucocyte accu-
mulation in the skin lesions in response to ZAP stim-
ulation was markedly potentiated by co-administration
of PGE2, as re¯ected by an increase in MPO activity to
2.72 � 0.43 units MPO (g tissue))1 (Fig. 5).
DISCUSSION
Our ®ndings demonstrating that chemoattractant-
induced leucocyte accumulation in rat skin is markedly
enhanced when a potent vasodilator is co-administered
with the chemotactic stimulus con®rm previous ob-
servations that tissue blood ¯ow is an important pa-
rameter in determining leucocyte recruitment to
in¯amed tissue sites (Issekutz & Movat 1982, Downey
et al. 1988, Raud et al. 1988, Teixeira et al. 1993). The
present study was undertaken to investigate by which
means the microvascular blood ¯ow level may in¯uence
the leucocyte extravasation process in in¯ammation.
Intravital microscopic observations of microvessels in
the rat mesentery demonstrated a close relationship
Figure 3 Relationship between rolling leucocyte ¯ux in rat mesen-
teric venules prior to chemotactic stimulation and number of adherent
leucocytes in the same venules following challenge with fMLP
(10±6M, 6 min). Linear regression line is superimposed.
Figure 4 Average rolling velocity of leucocytes which came to arrest
(adherent) within de®ned venular segments in the rat mesentery in
response to fMLP (10)6M) stimulation, and of leucocytes which did
not adhere (non-adherent) in the same vessel segments. Data are
means � SD based on analysis of totally 172 leucocytes. P < 0.05.
Figure 5 Leucocyte accumulation, as measured by MPO activity, in
rat skin after local injection of PBS, PGE2, ZAP (activated comple-
ment factors), or ZAP plus PGE2. Mean � SD, n � 4, P < 0.05.
Ó 1999 Scandinavian Physiological Society 255
Acta Physiol Scand 1999, 165, 251±258 X Xie et al. � Microvascular blood ¯ow and leucocyte/endothelium interactions
between the venular blood ¯ow and the rolling ¯ux of
leucocytes. Moreover, there was a strict proportionality
between the chemoattractant-induced ®rm adhesive
response and the rolling cell number, as previously
documented (Lindbom et al. 1992). Consequently, there
is a direct association between the venular blood ¯ow
and the extent of ®rm leucocyte adhesion. In line with
this view are our previous ®ndings that topical ad-
ministration of a vasodilator (acetylcholine) to the rat
mesentery increased venular blood ¯ow, rolling leuco-
cyte ¯ux, and chemoattractant-induced ®rm adhesion
to approximately the same extent (Thorlacius et al.
1995). In an analysis of relationships between small
venular haemodynamics and leucocyte/endothelium
interactions, and their signi®cance for leucocyte re-
cruitment to tissues in in¯ammation, it is important to
stress the fact that variations in tissue blood ¯ow (e.g.
the hyperaemia in in¯ammation) are effectuated
through calibre changes of the feeding arterioles. At the
post-capillary venular level, the corresponding ¯ow
variation will be re¯ected by a change in ¯ow velocity
as these small-sized venules lack the potential of vary-
ing their lumen to any appreciable extent. Conse-
quently, in a given venule, variations in the volume
blood ¯ow will result in proportional changes in blood
¯ow velocity as well as in wall shear stress.
Leucocyte recruitment to extravascular tissue is a
key component in in¯ammatory reactions. In order for
the leucocytes to arrest and establish ®rm contact with
the vascular endothelium in the in¯amed tissue area,
they need to resist the mechanical stress exerted by the
blood stream. To this end, initial rolling along the en-
dothelial lining is a necessary precondition for the ®rm
attachment to occur at physiological blood ¯ow rates
(Lindbom et al. 1992). The rolling interaction, mediated
predominantly by the selectin family of adhesion mol-
ecules (Tedder et al. 1995), serves to retard the motion
of the leucocytes, thereby enabling their ®rm binding to
the vessel wall, which in turn critically depends on the
function of b2 integrins (CD11/CD18) (Arfors et al.
1987). While selectins are capable of supporting a labile
adhesive interaction over a wide range of ¯uid shear,
formation of adhesion bonds via b2 integrins appears
to be markedly shear sensitive. However, once b2
integrin binding (and ®rm cell attachment) is estab-
lished, it becomes shear resistant (Lawrence & Springer
1991). In line with a capacity of the selectins to sustain
leucocyte rolling at high blood ¯ow rates, we found
that neither leucocyte rolling fraction nor the rolling
velocity was quantitatively related to the wall shear rate
in the microvessels. One factor contributing to the
comparably high shear insensitivity of the rolling in-
teraction may be ascribed to the deformability of the
leucocyte in relation to the shear stress imposed. The
rolling leucocytes assume a more ¯attened form with
increasing wall shear rate, and thereby, the contact area
(and adhesion strength) between the leucocyte and the
endothelial surface supposedly will increase (Firrell &
Lipowsky 1989; Damiano et al. 1996).
The apparent insensitivity of the rolling interaction
to varying shear conditions indicated by the present
observations differs from previously reported data on
various degree of inverse proportionality between shear
determinants and rolling parameters in microvessels
in vivo (Firrell & Lipowsky 1989, Ley & Gaehtgens
1991, Perry & Granger 1991, Damiano et al. 1996) and
in ¯ow chamber systems in vitro (Lawrence et al. 1987,
Lawrence & Springer 1991). The reason for these dis-
crepancies may, with respect to the in vivo studies,
pertain to differences in the size of vessels in which
observations have been made. In the present investi-
gation, the category of small-sized venules (15±25 lm)
where the most prominent leucocyte response to che-
moattractant stimulation normally takes place was
chosen for study. Also, in some of the studies quoted,
wall shear rate was lowered well below the normally
prevailing levels by forced reductions in blood ¯ow,
which may have in¯uenced the relationships involved.
The same problem applies to ®ndings obtained in dif-
ferent in vitro models, as the shear stress levels at which
observations typically have been made are more than
one order of magnitude lower than those present at
physiological ¯ow rates in vivo.
Our ®nding that leucocyte rolling velocity varied
within a rather narrow range and showed no correlation
to wall shear rate, suggests that rolling velocity princi-
pally is determined by the rate of formation and
breakage of the adhesive bonds (ToÈzeren & Ley 1992,
Alon et al. 1995) rather than being a function of the
¯uid drag force. However, this conclusion rests on the
fact that we considered only steady rolling leucocytes in
the analysis of rolling velocity and not those which were
intermittently rolling (jumping) along the endothelial
lining. Inasmuch as the ¯ow rate will affect transit time
of intermittently rolling cells similarly as for the free-
¯owing cells, a less restrictive criterion than here used
to delimit the rolling cell population likely accounts for
a seemingly discordant view with regard to the in¯u-
ence of wall shear rate on mean leucocyte rolling
velocity (Ley & Gaehtgens 1991).
We have previously shown that ®rm leucocyte ad-
hesion induced by chemoattractants is directly pro-
portional to the rolling cell number (Lindbom et al.
1992). The ®ndings of the present study extend these
observations to include a dependency of ®rm adhesion
also on the rolling velocity. We found that the average
rolling velocity of those leucocytes which upon
chemotactic stimulation ®rmly adhered in the observed
venular segments was signi®cantly lower than the ve-
locity of those cells which did not arrest but continued
256 Ó 1999 Scandinavian Physiological Society
Microvascular blood ¯ow and leucocyte/endothelium interactions � X Xie et al. Acta Physiol Scand 1999, 165, 251±258
to roll in the same vessel segments. In fact, rolling
velocities in the range of 25 lm s)1 and above were
exceptional among those leucocytes which came to
arrest. The existence of a critical rolling velocity level
above which transition into ®rm adhesion rarely will
take place is consistent with the shear sensitivity of û2
integrins in establishing ®rm binding to the vessel wall
(Worthen et al. 1987, Lawrence & Springer 1991,
Lindbom et al. 1992), and conform with previously
described relationships for spontaneously adhering
leucocytes in the cat mesentery (Perry & Granger 1991).
Even though there exists no clear relationship between
leucocyte rolling and venular shear forces at physio-
logical blood ¯ow rates, profound reductions in blood
¯ow (and shear rate) will ultimately lead to a decrease in
rolling velocity. Consequently, at very low ¯ow rates,
such as in ischaemic conditions, ®rm leucocyte adhe-
sion may be facilitated because of low rolling velocities
present in that situation. Such relationship may account
for the increased spontaneous adhesion found with
decreasing venular wall shear rate associated with
forced reductions in blood ¯ow (Perry & Granger
1991).
Taken together, we have demonstrated that there
exists a quantitative relationship between venular blood
¯ow and the extent of leucocyte rolling on the one
hand, and between leucocyte rolling and the ®rm ad-
hesive response induced by chemoattractants on the
other. Thus, through these relationships, a link between
tissue hyperaemia and leucocyte recruitment, charac-
teristic features of the in¯ammatory response, is pro-
vided. The in¯uence in this process of venular shear
forces on the binding dynamics of the adhesion mole-
cules involved appears to be of minor importance, in-
asmuch as no association between wall shear rate and
the rolling interaction or the ®rm adhesive response
was indicated by the data.
This study was supported by the Swedish Medical Research Council
(14X-4342, 04P-10738), the Swedish Foundation for Health Care
Sciences and Allergy Research (A98110), IngaBritt and Arne Lund-
bergs Foundation, and Karolinska Institutet.
REFERENCES
Alon, R., Kassner, P.D., Carr, M.W., Finger, E.B., Hemler,
M.E. & Springer, T.A. 1995. The integrin VLA-4 supports
tethering and rolling in ¯ow on VCAM-1. J Cell Biol 128,
1243±1253.
Arfors, K.E., Lundberg, C., Lindbom, L., Lundberg, K.,
Beatty, P.G. & Harlan, J.M. 1987. A monoclonal antibody
to the membrane glycoprotein complex CD18 inhibits
polymorphonuclear leucocyte accumulation and plasma
leakage in vivo. Blood 69, 338±340.
Atherton, A. & Born, G.V. 1973. Relationship between the
velocity of rolling granulocytes and that of the blood ¯ow
in venules. J Physiol (Lond) 233, 157±165.
Butcher, E.C. 1991. Leucocyte-endothelial cell recognition:
Three (or more) steps to speci®city and diversity. Cell 67,
1033±1036.
Carlos, T.M. & Harlan, J.M. 1994. Leucocyte-endothelial
adhesion molecules. Blood 84, 2068±2101.
Damiano, E.R., Westheider, J., Tozeren, A. & Ley, K. 1996.
Variation in the velocity, deformation, and adhesion energy
density of leucocytes rolling within venules. Circ Res 79,
1122±1130.
Downey, G.P., Gumbay, R.S., Doherty, D.E. et al. 1988.
Enhancement of pulmonary in¯ammation by PGE2:
Evidence for a vasodilator effect. J Appl Physiol 64,
728±741.
Firrell, J.C. & Lipowsky, H.H. 1989. Leucocyte margination
and deformation in mesenteric venules of rat. Am J Physiol
256, H1667±H1674.
Issekutz, A.C. 1981. Vascular responses during acute
neutrophilic in¯ammation: Their relationship to in vivo
neutrophil emigration. Lab Invest 45, 435±441.
Issekutz, A.C. & Movat, H.Z. 1982. The effect of vasodilator
prostaglandins on polymorphonuclear leucocyte in®ltration
and vascular injury. Am J Pathol 107, 300±309.
Lawrence, M.B., McIntire, L.V. & Eskin, S.G. 1987. Effect of
¯ow on polymorphonuclear leucocyte/endothelial cell
adhesion. Blood 70, 1284±1290.
Lawrence, M.B. & Springer, T.A. 1991. Leucocytes roll on a
selectin at physiologic ¯ow rates: Distinction from and
prerequisite for adhesion through integrins. Cell 65,
859±873.
Ley, K. & Gaehtgens, P. 1991. Endothelial, not
hemodynamic, differences are responsible for preferential
leucocyte rolling in rat mesenteric venules. Circ Res 69,
1034±1041.
Lindbom, L., Xie, X., Raud, J. & Hedqvist, P. 1992.
Chemoattractant-induced ®rm adhesion of leucocytes to
vascular endothelium in vivo is critically dependent on
initial leucocyte rolling. Acta Physiol Scand 146, 415±421.
Lipowsky, H.H., Usami, S. & Chien, S. 1980. In vivo
measurements of `apparent viscosity' and microvessel
hematocrit in the mesentery of the cat. Microvasc Res 19,
297±319.
Lundberg, C. & Arfors, K.E. 1983. Polymorphonuclear
leucocyte accumulation in in¯ammatory dermal sites as
measured by 51Cr-labeled cells and myeloperoxidase.
In¯ammation 7, 247±255.
Perry, M.A. & Granger, D.N. 1991. Role of CD11/CD18 in
shear rate-dependent leucocyte-endothelial cell interactions
in cat mesenteric venules. J Clin Invest 87, 1798±1804.
Raud, J., Dahlen, S.E., Sydbom, A., Lindbom, L. & Hedqvist,
P. 1988. Enhancement of acute allergic in¯ammation by
indomethacin is reversed by prostaglandin E2. Apparent
correlation with in vivo modulation of mediator release.
Proc Natl Acad Sci USA 85, 2315±2319.
Tedder, T.F., Steeber, D.A., Chen, A. & Engel, P. 1995. The
selectins: vascular adhesion molecules. FASEB J 9,
866±873.
Teixeira, M.M., Williams, T.J. & Hellewell, P.G. 1993. Role
of prostaglandins and nitric oxide in acute in¯ammatory
reactions in guinea-pig skin. Br J Pharmacol 110,
1515±1521.
Ó 1999 Scandinavian Physiological Society 257
Acta Physiol Scand 1999, 165, 251±258 X Xie et al. � Microvascular blood ¯ow and leucocyte/endothelium interactions
Thorlacius, H., Raud, J., Xie, X., Hedqvist, P. & Lindbom, L.
1995. Microvascular mechanisms of histamine-induced
potentiation of leucocyte adhesion evoked by
chemoattractants. Br J Pharmacol 116, 3175±3180.
ToÈzeren, A. & Ley, K. 1992. How do selectins mediate
leucocyte rolling in venules? Biophys J 63, 700±709.
Von Andrian, U.H., Hansell, P., Chambers, J.D. et al. 1992.
L-selectin function is required for b2-integrin-mediated
neutrophil adhesion at physiological shear rates in vivo. Am
J Physiol 263, H1034±H1044.
Worthen, G.S., Smedly, L.A., Tonnesen, M.G. et al. 1987.
Effects of shear stress on adhesive interaction between
neutrophils and cultured endothelial cells. J Appl Physiol 63,
2031±2041.
258 Ó 1999 Scandinavian Physiological Society
Microvascular blood ¯ow and leucocyte/endothelium interactions � X Xie et al. Acta Physiol Scand 1999, 165, 251±258