| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* MediCity Research Laboratory, Turku University and National Public Health Institute, Department in Turku, Turku, Finland; and
VTT Biotechnology and Food Research, Espoo, Finland
1Correspondence: MediCity Research Laboratory, Tykistökatu 6A, FIN-20520 Turku, Finland. E-mail: marko.salmi{at}utu.fi
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
70% of granulocyte extravasation into a site of an experimental
inflammation. Using intravital microscopy, VAP-1 blockade is shown to
increase the velocity of the rolling granulocytes and the frequency of
their jerky skippings during the rolling. In addition, the number of
firmly bound leukocytes decreased by 44% when VAP-1 was rendered
nonfunctional. Our results suggest that VAP-1 functions as a molecular
brake early in the adhesion cascade and consequently decreases the firm
adherence; it may also directly influence the transmigration step.
These data elucidate a new interplayer in the granulocyte extravasation
process and provide a novel physiological function for a member of the
monoamine oxidase family.—Tohka, S., Laukkanen, M.-L., Jalkanen, S.,
Salmi, M. Vascular adhesion protein 1 (VAP-1) functions as a molecular
brake during granulocyte rolling and mediates their recruitment
in vivo.
Key Words: inflammation • rolling • extravasation • migration • monoamine oxidases
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 2)
.
According to the prevailing multistep model, the extravasation
cascade is regulated by sequential interplay of different adhesion and
signaling molecules (1
2
3)
. Selectins and their mucin-type
counter receptors mediate initial tethering and rolling interactions
(4
, 5)
. The activation signals are thought to be conveyed
by chemokine binding to their serpentine receptors (6)
.
This triggers a switch of preexisting leukocyte integrins into a higher
avidity stage, which then mediate stable adhesion by binding to
endothelial members of the immunoglobulin superfamily (7
, 8)
. The molecular mechanisms of the transmigration step remain
poorly characterized (9)
.
Vascular adhesion protein 1 (VAP-1) (1)
is a
homodimeric endothelial sialoglycoprotein that belongs to a subgroup of
monoamine oxidases (10
, 11)
. These semicarbazide-sensitive
monoamine oxidases catalyze oxidative deamination of primary amines in
a reaction whose physiological substrates and biological effects remain
mysterious (12
13
14)
. They are strikingly distinct from
monoamine oxidases A and B with respect to their amino acid sequence,
subcellular localization, cofactors, substrate and inhibitor
specificity, and physiological function. We have shown in
vitro that VAP-1 is involved in lymphocyte binding to high
endothelial venules that support lymphocyte, but not granulocyte,
extravasation during physiological recirculation (15
, 16)
.
In those assays with frozen tissue sections, VAP-1 had only a minimal,
if any, role in granulocyte adhesion to endothelial cells. The
potential roles of VAP-1 in the granulocyte extravasation cascade and
its function in inflammatory reactions in vivo have remained
unresolved.
The aim of the present study was to analyze the involvement of VAP-1 in granulocyte accumulation at sites of inflammation in vivo. Moreover, since hydrodynamic shear caused by the blood flow is a critical parameter in all leukocyte–endothelial cell interactions in vivo and cannot fully be modeled in in vitro experiments, we wanted to elucidate the position of VAP-1 during the extravasation cascade using intravital microscopy. Our results show that if VAP-1 is rendered nonfunctional by monoclonal antibodies (mAbs), there is a dramatic reduction in the number of granulocytes infiltrating into an inflamed tissue. We further demonstrate that during granulocyte–endothelial cell interactions, VAP-1 has an important role in the initial rolling step, which leads to clear effects on stable adhesion and extravasation. The current results advance our understanding of development of an inflammatory response and regulation of granulocyte extravasation. Moreover, the biological function of VAP-1 suggests that semicarbazide-sensitive monoamine oxidases can regulate granulocyte trafficking, which is the first clearly defined physiological function for these enzymes.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Hermes-3
against human CD44 (18)
, which does not cross-react with
rabbits, served as an isotype-matched control for TK8–14. An anti-CD31
mAb cross-reacting with rabbits was obtained from Dako (Glostrup,
Denmark).
The construction of chimeric (Ch) anti-VAP-1 mAbs will be
described elsewhere in more detail (M.-L. Laukkanen et al., unpublished
results). In brief, the variable regions of the heavy chain of TK8–14
and 2D10 were amplified by reverse transcription-polymerase chain
reaction using total RNA from the parental hybridoma cells as templates
and degenerate 5' and 3' primers (19)
. Similarly, using
degenerate light-chain 5' primers and the reverse 3' kappa primer, the
light chains of these antibodies were amplified. To construct chimeric
antibodies, the variable regions of heavy and light chains were
subcloned into mammalian expression vectors containing human
heavy-chain constant regions (modified 
2 isotype) and a light-chain
constant region (kappa), respectively, and a mouse heavy-chain promoter
and a signal sequence. The CH2 domain has been modified in this vector
to lack complement activating and Fc receptor binding abilities, and
thus is devoid of all natural effector functions (20)
. A
control chimeric (ChNPLys) was done analogously using the variable
region of the light chain from a mAb against
4-hydroxy-3-nitrophenylacetyl and the variable region of a heavy chain
from a mAb specific to lysozyme.
Mammalian vectors containing the chimeric heavy-chain and light-chain expression cassettes were cotransfected into a nonsecreting rat myeloma cell line YB2/0 by electroporation. The antibody-producing clones were selected with ELISA. The chimeric antibodies were produced in stable transfectants and purified from the conditioned culture medium by protein A Sepharose affinity chromatography and acidic elution. After neutralization and dialysis, the concentration of the chimeric Ab was measured using bicinchoninic acid protein assay. Hence, the resulting chimeric mAbs contain the variable regions of mouse light and heavy chain grafted into an effector function-deficient human IgG2 framework.
Immunohistochemistry and immunoblottings
Acetone-fixed frozen sections from rabbit organs were stained
with an indirect immunoperoxidase technique (15)
using
mouse or chimeric mAbs at 20 µg/ml as primary antibodies. The second
stage reagents were peroxidase-conjugated sheep anti-mouse IgG or
anti-human IgA, M, and G (Dakopatts), as appropriate. The sections were
counterstained with hematoxylin and mounted in Depex.
Endothelial reactivity of anti-VAP-1 mAbs in rabbits was confirmed by ex vivo stainings. In these studies, anti-VAP-1 mAb TK8–14 or isotype-matched negative control mAb Hermes-3 was administered intravenously (i.v.) to rabbits (2 mg/kg). Thereafter, the animals were killed and tissues were collected for preparing frozen sections. In vitro, only the second-stage peroxidase anti-mouse Ig was applied onto the sections before developing the color reaction. Thus, only if the primary antibody has bound to vascular endothelium in vivo can any positive reactivity be detected.
The molecular weight of rabbit VAP-1 was determined by immunoblotting
of 1% Nonidet P-40 lysates made from rabbit liver, which is rich in
VAP-1-positive endothelial cells, using enhanced chemiluminescence and
a previously described protocol (10)
.
Inflammatory models
Two acute granulocyte-dependent peritoneal inflammation models
were used to study the role of VAP-1 in leukocyte extravasation. A
strong inflammation, induced by intraperitoneal (i.p.) injection of
interleukin 1 (IL-1) and proteose peptone, was used to study firm
adhesion and transmigration of granulocytes. A mild local inflammation
induced by superfusion of a part of exposed mesenteric membrane by IL-1
was used to study granulocyte rolling on endothelial cells. Both study
protocols have been approved by the Turku Review Board for the
Laboratory Animal Experimentation.
Chemical peritonitis model
New Zealand white (NZW) rabbits weighing 2 kg were sedated
with an intramuscular (i.m.) injection of 200 µl fentanyl-fluanisone
(Hypnorm) and marginal vein and central artery of the ear were
cannulated. A bolus (2 mg/kg) of anti-VAP-1 mAb TK8–14 or a control
mAb Hermes-3 diluted in sterile phosphate-buffered saline (PBS) was
then administered via the vein. Five minutes later, acute chemical
peritonitis was induced with an i.p. injection of 25 ml sterile PBS
containing 20,000 U human recombinant IL-1 (cross-reacting with
rabbits) and 5% proteose peptone (21)
; 0.1 g of blue
Sepharose 4CL beads as an indicator of a proper injection was included
to the suspension. Hereafter these animals will be called
peritonitis-induced rabbits. In control experiments, only PBS was
injected i.p., and these animals will be referred to as sham-induced
rabbits. Blood was drawn from the central artery before and at 1, 3, 5,
15, 30 min and 2 and 4 h after the mAb injection. At 4 h, the
animals were prepared for surgical anesthesia (see below) and 50 ml
prewarmed RPMI1640 medium was injected into the peritoneal cavity.
After a standardized massage protocol, the lavage fluid was gently
collected into heparin-containing tubes by drawing with a syringe
connected to a plastic tubing, which was inserted into the peritoneal
cavity through a small abdominal incision.
Thereafter, the animals were prepared for intravital microscopy
to enable microscopic analyses of adherent and transmigrated cells in
peritoneum. The mesentery was exteriorized through an abdominal midline
incision and gently spread on a special microscopic stage using a
previously described protocol (16)
. All exposed parts were
covered with saline-soaked gauze and kept moist by continuous dripping
of bicarbonate-buffered superfusion medium thermo-controlled at 37°C.
Leukocyte–endothelial interactions were recorded in real time using
digital video recorder (DHR10000NP, Sony) coupled to a CDD video camera
(Hamamatsu Photonics, Hamal Su City, Japan) attached to an ocular of an
Olympus BX50WI intravital microscope equipped with a water immersion
objective (x 20/0.5 and x 10/0.3 UMPlanFl, 3.3 mm working
distance).
Leukocyte counts from each animal were determined from 0.5 ml blood samples taken into EDTA-anticoagulated tubes (Capiject) using a hemocytometer (Celltac-alpha, Nihon Kohden, Japan) calibrated for rabbit cells. The volume of peritoneal lavage fluid and the presence of the blue indicator beads were recorded, and the number of peritoneal leukocytes was enumerated after trypan blue staining in cell counting chambers. Cytospins were prepared from the peritoneal cells (20,000 cells/slide) and subjected to Diff-Quick staining and differential counting.
In situ superfusion model of inflammation
In a weaker localized inflammatory model, animals were prepared
for intravital microscopy of postcapillary mesenteric venules according
to the published protocols (16
, 22)
with some
modifications. Surgical anesthesia was induced to NZW rabbits with an
i.m. injection of midazolam (Dormicum, 2 mg/kg) and Hypnorm (0.3 ml/kg)
and maintained by repeated injections of 10% Hypnorm (3 ml/kg/h) via a
cannulated marginal ear vein. Physiological saline was continuously
infused i.v. with a computer-controlled syringe pump (6 ml/kg/h). To
assist ventilation, a tracheostomy was performed and a tracheal tube
was inserted. Carotic artery was exposed and a PE 60 catheter was
inserted for blood sampling. Thereafter, a suitable branch of the
mesenteric artery was chosen and dissected free, and a polyethylene
catheter (PE-10) was inserted retrogradely into the vessel.
The animals were transferred to an intravital microscope and an appropriate area of mesenterium was selected for observation (see above). When feasible, two independent venules were monitored simultaneously. Inflammation was induced by superfusing the mesenteric window with recombinant human IL-1ß (3000 U/25 ml PBS injected as multiple small boluses into the superfusion fluid) for 3 h. Blood samples were regularly drawn through the carotic cannula. Centerline velocity of the blood was measured simultaneously using a dual-photodiode method and a commercial velocimeter (CircuSoft Instrumentation, Hockessin, Del.).
Baseline recordings (10 min) were taken after injection of local intra-arterial boluses of saline. Thereafter, 3 x 300 µl boluses of sterile PBS containing control chimeric Ab ChNPLys (100 µg/ml) or anti-VAP-1 chimeric Ab ChTK8–14 (100 µg/ml) alone or together with another anti-VAP-1 chimeric Ab Ch2D10 (100 µg/ml) were given during 45 s. A 10 min period was recorded again and hemodynamic variables were measured simultaneously.
In certain experiments with both inflammation models, animals were
given a 200 µl i.v. bolus of acridine red (1.67 mg/ml in PBS) to
stain the nuclei of leukocytes in vivo. The
leukocyte–endothelial interactions of labeled polymorphonuclear and
mononuclear cells were recorded using stroboscopic epi-fluorescent
illumination and a silicone-intensified target camera as described in
ref 16
.
Analyses of leukocyte–endothelial cell interactions
Hemodynamics were assessed as described in refs 16
, 22
23
24
. In brief, centerline velocities measured were converted
to mean blood flow velocities (Vb) by multiplying
with an empirical factor of 0.625. The Newtonian wall shear rates ()
were approximated as 2.12 x 8 x Vb/d,
where d is the mean diameter of the vessel (measured from 10 different
locations) and 2.12 is a median empirical correction factor. Wall shear
stress (
) was calculated as x 
, where is the
estimated viscosity of blood at 37°C (0.012 Poise).
The videotapes were analyzed off-line for the parameters characterizing leukocyte–endothelial cell interactions, i.e., rolling, stable adhesion, and perivascular extravasation. Leukocyte rolling flux (Rf) is defined as the number of rolling cells that pass a line perpendicular to the vessel wall during 1 min. Leukocyte transit time equals the time measured for a given cell to travel 100 µm in the vessel and is used to count the leukocyte velocity in µm/s (100 µm/transit time). Transit velocity distributions are expressed as number of cells falling into successive 1 µm/s velocity classes (1, 2, 3, 4... 100). Changes in the transit velocity were also analyzed by combining events in 20 successive classes into bigger ones (1–20, 21–40... ) to get enough cells into each class. The median transit velocity was defined as the 1 µm/s class in which cumulative frequency of cells reached 50%.
Rolling patterns were determined for individual successive leukocytes using frame-by-frame analyses. The number of interactions that each leukocyte had with the venular wall within a defined 100 µm vessel segment was scored as follows: 0 = continuous rolling (no change in the rolling pattern; the cell enters and leaves the segment as a rolling cell without any intermittent jumps), 1 = one tethering or detachment event (the cell either enters the segment in a free flow and makes one contact with endothelial layer which results in continuous rolling until the cell leaves the segment or the cell enters the segment as a rolling cell, which then detaches into the free flow during its transit through the segment). Each attachment or detachment event is thus assigned value 1 and the sum is counted for each cell during its transit through the segment. So if a leukocyte enters as a free-flowing cell, tethers to the endothelium, rolls for a while, makes a jump and binds to the endothelium again, and leaves the segment by rolling, it gets a score 3 (two attachments and one detachment within the segment). For different treatments, the mean scores for all cells analyzed before and after Ab injections were counted.
The number of stably adherent leukocytes was determined from the video recordings of the animals subjected to the 4 h chemical peritonitis. The number of leukocytes adherent per 100 µm vessel wall in the areas that supported leukocyte rolling was counted from the playbacks.
The extent of leukocyte transmigration was analyzed from videotapes obtained from the peritonitis-induced and sham-induced animals. The planar surface area of mesenteric membrane where leukocytes covered >50% of the area was measured using a morphometric program (AnalySIS, Soft Imaging System, Germany).
Statistical analyses
Peritoneal leukocyte counts, areas of perivascular
infiltrate, and numbers of firmly adherent cells were compared using
Student’s unpaired t test. The effects of chimeric Ab
injections on hemodynamic parameters and on leukocyte–endothelial cell
contacts were compared to the baseline values obtained from the very
same vessel before the treatment using paired Student’s t
test. Median rolling velocities were compared using analysis of
variance. Statistical significance was set at P<0.05.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
cross-react with rabbit VAP-1. Both mAbs recognize in rabbits an
antigen that displays a tissue distribution similar to that of human
VAP-1. For instance, in the rabbit heart the mAbs stain vessels of
different sizes (Fig. 1A
). An endothelial marker, CD31, displays a similar staining
pattern. Endothelial synthesis of rabbit VAP-1 was confirmed by
administering mAbs i.v. to rabbits and then staining tissue sections
with the second-stage antibody only. Anti-VAP-1 mAb-treated animals
showed intense luminal reactivity, whereas no signal could be detected
from the animals that received the control mAb (Fig. 1B
).
These data show that the primary antibody against VAP-1 has bound to
the luminal surface of vascular endothelial cells in vivo.
Also, the molecular mass of the rabbit VAP-1 antigen [170 kDa under
nonreducing (Fig. 1C
) and 90 kDa under reducing conditions]
is identical to that of human VAP-1 (10)
. Hence, the
rabbit model allowed us for the first time to develop an in
vivo model to test the role of VAP-1 in the extravasation cascade.
|
VAP-1 mediates leukocyte extravasation into sites of acute
inflammation
To study the role of VAP-1 in inflammation in vivo, we
first elucidated whether interference of VAP-1 function with mAbs
affects the net extravasation process in an acute peritonitis model.
The overall extent of peritoneal inflammation was evaluated by counting
the number of extravasated leukocytes in the lavage fluid. The
transmigration process at the perivascular space was also directly
analyzed by intravital microscopy of inflamed mesenteric membrane from
the same animals.
Control animals were injected i.p. with saline only
(sham-induced) and treated with an i.v. bolus (2 mg/kg) of anti-VAP-1
or isotype-matched control mAb. As expected, there were only a few
leukocytes in the lavage fluid at 4 h in either control mAb or
anti-VAP-1 mAb-treated sham-induced animals (Fig. 2A
). The majority of these peritoneal cells (81% on average)
were mononuclear ones (Fig. 2B
), presumably representing
mostly resident macrophages. In videomicroscopic analyses, only very
occasional cells in perivascular space were detectable (Fig. 2C
, F
) in sham-induced animals after either mAb
treatment. Thus, i.p. injection of saline per se does not elicit any
significant inflammatory response.
|
When an acute chemical peritonitis was induced to rabbits by i.p.
injection of IL-1 and proteose peptone (peritonitis-induced animals), a
>20-fold increase in the number of extravasated cells in peritoneal
cavity was seen in control mAb-treated animals (Fig. 2A
). In
peritonitis-induced animals, there was a significant 69% decrease in
the number of extravasated granulocytes in the lavage fluid in
anti-VAP-1-treated rabbits when compared to those that received the
control mAb (Fig. 2A
). Consistent with the granulocytic
nature of the acute inflammation, >97% of the leukocytes were
polymorphonuclears in the lavage fluid of peritonitis-induced animals
(Fig. 2B
). Both anti-VAP-1 and control mAbs caused a
transient drop in systemic leukocyte counts. Notably, however, the drop
was recovered by 15 min in the case of anti-VAP-1 mAb but was more
long-lasting with the control mAb (Fig. 2G
), and hence
cannot explain the beneficial antiinflammatory effects of anti-VAP-1
treatment in the peritonitis-induced rabbits. Inhibition of leukocyte
emigration into the lavage fluid also translated into a clear reduction
in the extent of perivascular infiltration in the VAP-1-treated
peritonitis-induced animals (Fig. 2D
, E
, F
). Thus, blocking of
VAP-1 dramatically attenuates a granulocyte-dependent inflammatory
response in vivo.
Anti-VAP-1 mAb diminishes the number of firmly adherent leukocytes
in inflamed vessels
The net reduction in the number of extravasated cells in
VAP-1-treated animals can be due to the inhibition at the rolling, firm
adhesion, or transmigration step. Therefore, using the intravital
videomicroscopy, we analyzed the number of adherent cells in the
mesenteric vessels of sham-induced and peritonitis-induced animals. In
sham-induced rabbits, very few leukocytes were found firmly adherent to
vascular wall in control mAb-treated animals, and even fewer in
anti-VAP-1-treated rabbits (Fig. 3A
). In peritonitis-induced rabbits, the number of stably
adherent cells was high in the control mAb-treated animals. Most
important, the results showed a statistically significant 44%
reduction in the number of stably bound cells in peritonitis-induced
animals receiving the anti-VAP-1 mAb when compared to the control
mAb-treated animals (Fig. 3A
). Intravital microscopy with
fluorescently labeled leukocytes revealed that the rolling and adherent
cells were almost exclusively (>98%) granulocytes in this model (Fig. 3B
). Thus, VAP-1 contributes critically to the granulocyte
extravasation before the transmigration step.
|
Anti-VAP-1 mAbs and granulocyte rolling
A reduction in the number of firmly adherent leukocytes after
anti-VAP-1 treatment can mean that rolling is normal in these animals,
and the first step affected is stable adhesion, or that rolling is
already impaired and consequently leads to diminished stable adhesion.
The 4 h peritonitis model was not useful to distinguish between
these alternatives for two reasons. First, the severe inflammation with
many firmly adherent cells on the vessel wall led to a situation in
which most rolling took place via leukocyte–leukocyte interactions
(25
, 26)
rather than via leukocyte rolling on endothelial
cells (data not shown). Since VAP-1 is not expressed on any leukocyte,
the possible role of VAP-1 on rolling would be practically impossible
to detect with this model. Second, since systemic administration of
both mAbs caused a transient drop in blood leukocyte counts (see
above), they were not applicable to minute-by-minute analysis of
rolling immediately after their infusion.
To overcome these limitations, we used genetically engineered effector
function-deficient antibodies (complement fixing and Fc receptor
binding capacity; ref 20
). The chimeric antibodies
ChTK8–14 and Ch2D10 were function blocking in endothelial binding
assays with lymphocytes (data not shown) and showed identical tissue
staining patterns in humans and rabbits when compared to the parental
mouse mAbs (Fig. 1A
). The control chimeric antibody did not
recognize anything in rabbits (Fig. 1A
). With these chimeric
mAbs, a local infusion technique into a side branch of mesenteric
artery was applied (22)
to minimize systemic effects of
loading animals with foreign proteins and due to a low yield of the
chimeric mAb. We also developed a model of less severe inflammation in
which almost all rolling occurs between granulocytes and endothelial
cells by locally superfusing mesenteric membrane with IL-1 containing
buffer under an intravital microscope.
In this in situ inflammation model, the number of
rolling cells was 46 ± 12 cells/min immediately after the
surgery. Control or anti-VAP-1 chimeras had no effect on
surgery-induced leukocyte rolling on endothelial cells. The number of
rolling cells significantly increased to 85 ± 21 cells/min
(n=14, P=0.005) after a 3 h IL-1
superfusion. After this stimulation, >94% of the rolling cells were
granulocytes (Fig. 4A
, B
, C
). Administration of either control or anti-VAP-1
chimeras had no significant effects on blood velocity, blood vessel
diameter, shear rate, shear stress, or leukocyte counts nor did they
significantly affect the number of those cells which rolled at velocity
<100 µm/s. (Table 1
, Fig. 4D
, E
, F
).
|
|
VAP-1 blockade regulates the rolling velocity of granulocytes
in vivo
Using the local superfusion model of inflammation, it was apparent
from intravital videomicroscopy that the velocity of the rolling cells
was faster after anti-VAP-1 mAb infusion. The rolling velocity is a
critical parameter in regulation of the adhesion cascade for
determining the time a leukocyte is exposed to local activating stimuli
at an inflammatory focus. Slow rolling velocity is required for an
integrin-mediated stable adhesion to take place (1)
.
Therefore, after a 3 h IL-1 superfusion, chimeric mAbs were
injected via the mesenteric artery and the velocity of individual
rolling cells was determined in detail 3 min and 10 min later.
Treatment with the control chimeric Ab had no effect on the
velocity of rolling cells at the 3 min time point (Fig. 5
). In contrast, the rolling velocity increased significantly in animals
subjected to one or both anti-VAP-1 chimeric Ab. This can be seen from
the cumulative rolling velocity profiles (Fig. 5A
, B
, C
) in
which the velocity of each individual leukocyte is determined. The same
increase in the rolling velocity after neutralization of VAP-1 was
clearly evident from pooled analyses of the distribution of leukocytes
in different velocity classes (Fig. 5D
, E
, F
). When using a
median rolling velocity as a single representative variable, there was
a 20% increase after administration of the anti-VAP chimeric Ab
ChTK8–14 (Fig. 5G
). A more pronounced 35% increase was
seen if both anti-VAP-1 chimeras were applied simultaneously. Ten
minutes after injection of the control chimeric antibody ChTK8–14 and
ChTK8–14+Ch2D10, a 4, 10, and 54% increase in the median rolling
velocities, respectively, were seen. Thus, blocking of VAP-1 causes a
sustained increase in granulocyte rolling velocity.
|
VAP-1 is needed for smooth uninterrupted rolling of granulocytes
The change in the rolling velocity could be due either to an
increase in the velocity of continuously rolling cells or to a change
in the rolling behavior (frequency and duration of rolling contacts and
jumps during transit in the vessel). To distinguish between these two
possibilities, the number of attachments and detachments of each
leukocyte interacting with the vessel wall was enumerated. During the
control period, the majority of cells displayed relatively
uninterrupted smooth rolling or only a few jumps (Fig. 6
). When the function of VAP-1 was blocked with ChTK8–14, alone or
together with Ch2D10, rolling became much jerkier (Fig. 6)
. Thus,
functionally active VAP-1 is needed to secure the granulocyte contacts
with the vessel wall during the rolling phase of the extravasation
cascade.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Peritonitis is a well-established model for an acute granulocyte-dependent inflammation. In the present study, two different inflammatory models were used where the vast majority of rolling and transmigrating cells were granulocytes. The chemical peritonitis causes a brisk inflammation with numerous slowly rolling, adherent, and transmigrating cells. It revealed that ablation of VAP-1 function significantly reduces the number of stably bound and extravasated leukocytes. Nevertheless, it requires loading of animals with milligram quantities of mAbs and shows a prominent component of leukocyte-leukocyte rolling, which even theoretically cannot be inhibited by anti-VAP-1 mAbs. The in situ inflammatory model with a local infusion technique, on the other hand, was designed to dissect leukocyte tethering and rolling of leukocytes directly on endothelial cells with small amounts of locally administered antibodies under mild inflammatory conditions. In this setting, VAP-1 proved to be important already for the very initial granulocyte-endothelial contacts. In contrast to many published reports using intravital microscopy, it should be noted that in this model we were able to analyze the very same vessels with both isotype-matched control and anti-VAP-1 mAbs and to perform careful kinetic analyses of hemodynamic variables throughout the experiments.
The inflammatory response in peritoneum is markedly impaired
after neutralizing the function of several adhesion molecules. On the
endothelial side, the inflammatory response is attenuated by 40–80%
when P-selectin, E-selectin, or ICAM-1 are separately rendered
nonfunctional (27
, 28)
. Nevertheless, an almost complete
inhibition is seen only when ICAM-1 and P-selectin (29)
or
both E- and P-selectin are inactivated simultaneously (27
, 30)
. These data illustrate well the redundancy of the mechanisms
operative at the successive steps of the vital multistep adhesion
cascade. Thus, the 70% inhibition seen after VAP-1 blockade is
highly significant and may be further enhanced by additional
coinhibition of other relevant adhesion molecules. The present results
are the first evidence that inhibition of VAP-1 reduces the
inflammatory response in an in vivo animal model.
Intravital microscopy revealed that VAP-1 is needed to reduce the
rolling velocity of granulocytes during inflammation. In that sense,
its function is reminiscent of that of E-selectin (CD62E). When the
function of inflammation-induced E-selectin is neutralized, the mean
rolling velocity of granulocytes in mice increases (23
, 31)
. In rabbits, neutralization of E-selectin by a 10 min
in situ incubation under no-flow conditions also increased
the absolute rolling velocity in the mesenterium (32)
.
Hence, VAP-1 and E-selectin are probably important as a braking
mechanism, which reduces the velocity of rolling cells enough for the
leukocytes to survey the endothelial cells for signs of inflammation,
that is needed for activation and subsequent stable adhesion.
The stepwise component of rolling is thought to be regulated by
dissociation of clusters of receptor–ligand bonds (33)
.
Thus, the characteristic jerky rolling seen after VAP-1 blockade
suggests that the bond between VAP-1 and its ligand contributes
critically to these clusters. Alternatively, VAP-1 can regulate the
dissociation of other receptor–ligand bonds (presumably those of
selectins) mediating tethering and rolling. Although the regulatory
role of VAP-1 might involve its enzymatic function, which catalyzes the
formation of a signaling molecule
H2O2 and an aldehyde
(11)
, the antibody-dependent inhibition most likely
involves another mode of action since the mAbs used do not inhibit
monoamine oxidase activity of this dual-function molecule.
The rolling flux did not diminish after VAP-1 blockade. Since
this parameter describes how many cells pass a certain line in the
vessel within a given time, it is affected by the number of rolling
cells as well as their velocity. Thus, if the number of rolling cells
remains the same and the rolling velocity increases, it actually
increases the rolling flux. This is the case after E-selectin blockade,
which results in a 1.8-fold increase in the rolling flux
(34)
. Hence, the 35–54% increase in rolling velocity
observed in the absence of increased rolling flux after VAP-1 blockade
means that the number of rolling cells must decrease. We hypothesize
that a more rigorous saturation of VAP-1 by function-blocking
antibodies would also lead to an actual decrease in the number of
rolling cells as a consequence of the increased rolling velocity. In
fact, when the effect of anti-VAP-1 chimeric Abs was analyzed in
vessels with rapidly rolling (100 µm/s) cells, there was a 45%
decrease in the absolute number of rolling cells. These data suggest
that when the endothelial binding avidity of the fast rolling cells is
further weakened by ablating the VAP-1 function, the cells will be
released into the noninteracting pool of leukocytes.
Since interference with VAP-1 affected extravasation more than rolling
or stable adhesion, it suggests that VAP-1 could play a role during the
transmigration step as well. This would be compatible with its
expression in pericytes (35)
. Interaction of leukocytes
with endothelial or pericyte VAP-1 could normally assist their
transmigration during this last phase of the multistep adhesion
cascade. In this case, VAP-1 could share functions with CD31, which is
thought to play a critical role at this step. When the function of this
adhesion and signaling molecule is blocked by mAbs, the number of
rolling or adherent cells does not change, but the cells are trapped at
the basement membrane during the subsequent transmigration process
(36
37
38
39)
.
In conclusion, our results show that a semicarbazide-sensitive monoamine oxidase, VAP-1, plays an important role in granulocyte extravasation in vivo. In fact, its function as an endothelial molecule regulating leukocyte adhesion is the first physiological function ascribed to this family of molecules. Blocking of VAP-1 function with Abs increases the rolling velocity, which translates into reduction of firm adhesion and ultimately into a significant impairment of the emigration response of granulocytes at sites of inflammation. The chimeric anti-VAP-1 antibodies developed here will be useful for further clinical trials as a novel anti-adhesive tool.
|
ACKNOWLEDGMENTS |
|---|
Received for publication April 26, 2000.
Revision received July 28, 2000.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. Jalkanen and M. Salmi VAP-1 and CD73, Endothelial Cell Surface Enzymes in Leukocyte Extravasation Arterioscler. Thromb. Vasc. Biol., January 1, 2008; 28(1): 18 - 26. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Koskinen, S. Nevalainen, M. Karikoski, A. Hanninen, S. Jalkanen, and M. Salmi VAP-1-Deficient Mice Display Defects in Mucosal Immunity and Antimicrobial Responses: Implications for Antiadhesive Applications J. Immunol., November 1, 2007; 179(9): 6160 - 6168. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Sumitran-Holgersson Lock 'n' roll with VAP-1 Blood, September 15, 2007; 110(6): 1706 - 1707. [Full Text] [PDF]
|
|
|
|
|
|
S. Jalkanen, M. Karikoski, N. Mercier, K. Koskinen, T. Henttinen, K. Elima, K. Salmivirta, and M. Salmi The oxidase activity of vascular adhesion protein-1 (VAP-1) induces endothelial E- and P-selectins and leukocyte binding Blood, September 15, 2007; 110(6): 1864 - 1870. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Drenkard, F. M. Becke, J. Langstein, T. Spruss, L. A. Kunz-Schughart, T. E. Tan, Y. C. Lim, and H. Schwarz CD137 is expressed on blood vessel walls at sites of inflammation and enhances monocyte migratory activity FASEB J, February 1, 2007; 21(2): 456 - 463. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Salmi and S. Jalkanen Developmental regulation of the adhesive and enzymatic activity of vascular adhesion protein-1 (VAP-1) in humans Blood, September 1, 2006; 108(5): 1555 - 1561. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Merinen, H. Irjala, M. Salmi, I. Jaakkola, A. Hanninen, and S. Jalkanen Vascular Adhesion Protein-1 Is Involved in Both Acute and Chronic Inflammation in the Mouse Am. J. Pathol., March 1, 2005; 166(3): 793 - 800. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Martelius, V. Salaspuro, M. Salmi, L. Krogerus, K. Hockerstedt, S. Jalkanen, and I. Lautenschlager Blockade of Vascular Adhesion Protein-1 Inhibits Lymphocyte Infiltration in Rat Liver Allograft Rejection Am. J. Pathol., December 1, 2004; 165(6): 1993 - 2001. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Vega, P. Chacon, J. Monteseirin, R. El Bekay, M. Alvarez, G. Alba, J. Conde, J. Martin-Nieto, F. J. Bedoya, E. Pintado, et al. A new role for monoamine oxidases in the modulation of macrophage-inducible nitric oxide synthase gene expression J. Leukoc. Biol., June 1, 2004; 75(6): 1093 - 1101. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Koskinen, P. J. Vainio, D. J. Smith, M. Pihlavisto, S. Yla-Herttuala, S. Jalkanen, and M. Salmi Granulocyte transmigration through the endothelium is regulated by the oxidase activity of vascular adhesion protein-1 (VAP-1) Blood, May 1, 2004; 103(9): 3388 - 3395. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Abella, L. Marti, M. Camps, M. Claret, J. Fernandez-Alvarez, R. Gomis, A. Guma, N. Viguerie, C. C |
