Immunology and Cell Biology (2003) 81, 397–408
Special Feature
Differential roles of VLA-4(CD49d/CD29) and
LFA-1(CD11a/CD18) integrins and E- and P-selectin during
developing and established active or adoptively transferred adjuvant
arthritis in the rat
A N D R E W C I S S E K U T Z , 1 S ATO RU NA K A Z ATO 2 a n d T H O M A S B I S S E K U T Z 1
1
Departments of Pediatrics, Pathology and Microbiology-Immunology, Dalhousie University, Halifax, Nova Scotia,
Canada and 2Otsuka Pharmaceutical Co., Ltd, Tokushima, Japan
Summary The role of the integrins VLA-4 and LFA-1 and of the selectin adhesion molecules in autoimmune
arthritis was investigated. Adjuvant arthritis was induced in Lewis rats by active immunization (s.c.) with Mycobacterium butyricum or by adoptive transfer of immune T cells. With active adjuvant arthritis, Lewis rats develop
maximal polyarticular joint inflammation and migration of radiolabelled (111In and 51Cr) blood neutrophils and
monocytes to the joints 14 days post Mycobacterium butyricum immunization. Using blocking monoclonal antibodies we osbserved that at this stage monocyte recruitment was dependent (85%) on P-selectin plus VLA-4 (α4B1)
and neutrophil recruitment depended (> 80%) on P-selectin plus LFA-1 (CD11a/CD18). E-selectin played a
minimal role in inflammatory cell recruitment to the already inflamed joint. In contrast, during the development
of active adjuvant arthritis, blockade of P-selectin beginning at day 5 post-immunization had no effect on
subsequent arthritis. However, E-selectin blockade at this stage reduced arthritic scores by 70% (P < 0.01) and combined E-selectin plus VLA-4 blockade prevented development of arthritis. Either treatment nearly abolished neutrophil and monocyte recruitment to joints at day 14 and prevented cartilage damage. VLA-4 blockade alone was
less effective. Adoptive T-cell transfer of adjuvant arthritis to naive rats employed spleen/lymph node
lymphocytes from Mycobacterium butyricum immunized rats stimulated with Concanavalin A in vitro (48 h).
E-selectin ± P-selectin blockade had no effect on the development of adoptive arthritis. However, VLA-4 integrin
blockade inhibited adoptive arthritis severity by 55% (P < 0.01). LFA-1 blockade had no effect. In adoptive
adjuvant arthritis, inhibition of arthritis clinically and by histology was essentially complete (> 90%) when E- and
P-selectin blockade was combined with VLA-4 blockade. Thus, in the development of actively induced arthritis
E-selectin plays an important role, likely mediating early antigen reactive T-cell recruitment to joints. In contrast,
VLA-4 and multiple selectin mechanisms are involved in arthritis induction by ex vivo restimulated arthritogenic
T cells. Furthermore, in actively induced adjuvant arthritis, P- and E-selectin and VLA-4 are differently important
in the initiation of arthritis, and at the time of fully developed joint inflammation.
Key words: inflammation, leucocyte, lymphocyte, migration, monocyte, neutrophil, T-cell.
Introduction
Chronic arthritis, most commonly exemplified in humans by
rheumatoid arthritis, is an inflammatory condition characterized by large numbers of leucocytes, including neutrophils
(PMN), monocytes and lymphocytes, infiltrating the joint
space, synovium and periarticular tissues. Leucocyte recruitment to inflamed tissue involves distinct cell adhesion
molecules (CAM). Among these, the selectin family members
P-selectin and E-selectin expressed on inflammatory mediator activated endothelium, primarily mediate the initial capture
of leucocytes from the flowing blood stream and also their
rolling on the postcapillary venule endothelium. These selectins
bind to fucosylated, sialylated, and sulfated carbohydrate
Correspondence: Dr Andrew C Issekutz, Dalhousie University,
IWK Health Care, 5850 University Avenue, Halifax, Nova Scotia,
Canada B3J 3G9. Email: [email protected]
Received 11 June 2003; accepted 19 June 2003.
structures related to sialyl Lewisx expressed primarily on
leucocytes.1,2 This leads to leucocyte activation and firm
adhesion to endothelium, followed rapidly by leucocyte
emigration. The latter process is mediated especially by the
very late activation antigen-4 (VLA-4; CD49d/CD29 or α4β1)
and the β2 (CD11/CD18) integrins, in particular lymphocyte
function-associated antigen-1 (LFA-1; CD11a/CD18) and
Mac-1 (CD11b/CD18). These integrins on the leucocyte
engage the immunoglobulin superfamily adhesion molecules,
intercellular adhesion molecule-1 and -2 (ICAM-1, ICAM-2)
and vascular cell adhesion molecule-1 (VCAM-1) on the
endothelium.3,4 The vascular endothelium in synovium of
patients with rheumatoid and other forms of chronic arthritis,
express P- and E-selectin as well as ICAM-1 and VCAM-1.5–7
A soluble form of E-selectin, ICAM-1 and VCAM-1 are also
elevated in synovial fluid, and in serum of these patients.8–10
Some of these synovial vessel CAM appear to be functional
in mediating leucocyte adhesion ex vivo.
398
AC Issekutz et al.
Inappropriate recruitment of leucocytes is believed to be
key in the development and progression of many autoimmune
diseases, including chronic arthritides.11 Thus the identification of the CAM mediating leucocyte migration into joints,
including the auto-reactive T-cells believed to initiate some
forms of arthritis, is important for developing strategies to
prevent the initiation or progression of chronic arthritis.
Towards this end, we have investigated the CAM mechanisms
for PMN, monocyte and T-cell migration to joints in an
experimental model of human chronic arthritis, including
rheumatoid arthritis and spondyloarthritides, namely adjuvant
induced arthritis. In susceptible strains of rats, polyarticular
arthritis can be induced by a single intradermal (i.d.) or s.c.
injection of killed M. butyricum or M. tuberculosis in
mineral oil.12 This adjuvant arthritis (AA) begins at 10 days
post-immunization in Lewis-strain rats with polyarticular
inflammation and rapidly progressive within 4 days to
maximal clinical severity followed by progression joint
destruction. Adjuvant arthritis is believed to be initiated by
arthritogenic T-cells generated during the immune response to
M. butyricum. It can be transferred to naive rats with spleen
and lymph node T-cells, especially following concanavalin
A stimulation in vitro (48–72 h) or with cloned T-cells
derived from rats responsive to Mycobacterium cell wall
peptidoglycan or Mycobacterium heat shock protein 65.13–15
The polyarticular arthritis is characterized by acute synovitis
as early as 7 days post-immunization, progressing to extensive infiltration of synovium and periarticular tissues by
PMN, monocytes, T-cells (especially CD4 cells) associated
with synovial panus formation, cartilage proteoglycan depletion and cartilage erosion. The talar and metatarsal joints
are most severely affected, with involvement also of carpal
and metacarpal joints, as well as the axial skeleton with
spondyloarthritis.12
Definition of the major CAM mechanisms for the migration of leucocytes to arthritic joints has been greatly facilitated by employing radiolabelled purified blood PMN, blood
monocytes and spleen or lymph node T-cells. Using these
techniques, combined with unique adhesion function blocking mAbs to specific CAM on rat leucocytes, including to the
α chain of VLA-4 (α4, CD49d), LFA-1 (αL, CD11a) and Mac1 (αM, CD11b), to the common β2 (CD18) integrin chain, to
L-selectin on leucocytes, and to E- and P-selectin on vascular
endothelium, we have identified some critical CAM interactions in previous studies which cannot be extensively
reviewed here. However, these showed that monocyte migration to fully developed arthritis 14 days post-immunization
was partially mediated by both VLA-4 and LFA-1; blockade
of both VLA-4 and LFA-1 nearly completely (by 95%) prevented monocyte migration to inflamed joints, demonstrating
that these two integrins had complementary roles in monocyte recruitment.16 Similarly, we demonstrated that PMN
migration to arthritic joints was mediated by a cooperative
function of LFA-1 with Mac-1.17 VLA-4 was also shown to be
present on rat PMN and it also contributed significantly to
the migration of PMN to arthritic joints. However, unlike
the case with monocytes, blocking all of the β2 integrins and
VLA-4 only inhibited PMN migration to the joints by
70–80%.18
Our previous studies of the role of selectins in PMN and
monocyte recruitment to arthritic joints demonstrated that
P-selectin mediated a part of the migration of PMN and
monocytes to fully developed arthritis at day 14, while
E-selectin was not required. Blockade of P-selectin resulted
in 40–60% inhibition of PMN and monocyte migration to
arthritic joints. E-selectin was found to play a cooperative
role with P-selectin in the recruitment of PMN, since combination of E-selectin and P-selectin blockade inhibited PMN
migration to a significantly greater extent (60–80%) than
P-selectin blockade alone.19 However, a significant proportion of PMN and monocyte migration to the arthritic joints
was selectin independent, since treatment with mAb to P-,
E- plus L-selectin in combination, which blocked PMN and
monocyte migration to dermal inflammatory reactions by up
to 90%,20 left a significant proportion of PMN and monocyte
migration to arthritic joints (20–40%) unaffected.
With this background knowledge of the role of integrins
and selectins, we have recently focused our studies on the
relative role of the α4 integrin (VLA-4) and the selectins,
because it is now recognized that the α4 integrin not only
mediates firm adhesion and transendothelial migration of
leucocytes, but can also mediate initial leucocyte capture and
rolling similar to the previously well established function of
P-, E- and L-selectins.21–23 Here we review our recent findings of the role of these CAM in recruitment of PMN and
monocytes to fully developed arthritis, as well as their role in
the development of actively induced arthritis induced by
adoptive transfer of sensitized T-cells.
Materials and Methods
Animals
Adjuvant arthritis was induced in 6–8 week old male Lewis rats
(175–200 g; Harlan Sprague Dawley, Indianopolis, IN, USA) by s.c.
immunization with 0.5 mg of M. butyricum (Gibco, Detroit, MI,
USA) in 0.05 mL mineral oil at two sites on each side of the base of
the tail. Arthritis was scored clinically from 0–4 points per limb and
tail based on severity of swelling, erythema and limitation of movement as described by Taurog et al.12 Rats were also weighed regularly
and joint swelling was quantified by change in diameter measured
with calipers.
Monoclonal antibodies
The following function blocking mAbs to rat adhesion molecules
were used: anti-rat P-selectin (our clone RMP-1, IgG2a and PR849,
IgG1, gift from Dr E Berg, Protein Design Labs); anti-rat E-selectin
(our clone RME-1 and ER890 (from Protein Design Labs), both
IgG1); anti-rat L-selectin (HRL-3 hamster IgG; gift from Dr DC
Anderson, Pharmacia Upjohn, Kalamazoo, MI, USA and Dr M
Miyasaka, Osaka, Japan); anti-α4 integrin (TA-2 mouse IgG1);24
anti-rat LFA-1 (TA-3, mouse IgG1); anti-rat Mac-1 (OX-42, mouse
IgG2a, gift from Dr WD Mason, Oxford, UK); and anti-rat CD18
(WT.3, mouse IgG1, gift from Dr M Miyasaka). All of the mAbs have
been used by us in previous studies in the arthritis model and other
inflammation models, as well as in in vitro leucocyte adhesion
studies.16–19,23 Both in vitro and in vivo studies confirmed the potent
adhesion function blocking activity of these mAbs. Rats were treated
with mAb by i.v. injection of 1 mg for acute short term (2-h)
migration experiments, or 2–3 mg i.v. followed by i.p. injection at
indicated times for longer experiments. Treatment of control animals
consisted of non-blocking, but binding, mAb to the above rat CAM,
Integrins and selectins in arthritis
and isotype control, non-binding mAbs as reported previously.16,18,19
Where indicated, Fab2 fragments of the above mAbs were used.
None of the mAb treatments caused leucopenia or modification of
the circulating half-life of the radiolabelled leucocytes (see below).
Leucocyte isolation radiolabelling and migration studies
Rat blood monocytes and PMN were isolated by hydroxyethyl starch
(Hespan, Dupont Merck, Wilmington, DE, USA) exchange transfusion performed through a needle inserted in a femoral vein of
anaesthetized arthritic rats, as described previously in detail.17,25 The
exchanged blood was collected into heparin and acid-citrate-dextrose
(ACD, formula A, Fenwal-Travenol, Malton, ON, Canada) and the
leucocyte rich plasma was harvested during 1 g sedimentation of red
blood cells. The leucocytes were then separated into PMN and
mononuclear cell fractions using sequential centrifugation on discontinuous Percoll density gradients containing autologous plasma
(10%). Following initial separation of the mononuclear cells above
63% isotonic plasma Percoll, and PMN above 74% plasma-Percoll,
the mononuclear cell layer was made slightly hypertonic to
299 mOsm in calcium and magnesium-free Tyrode’s solution 10%
plasma. Monocytes were then purified on a second discontinuous
plasma-Percoll gradient (40/55/58% Percoll in 10% Tyrode’s
solution) by centrifugation (350 × g, 30 min at 22°C). The purified
PMN (> 95%) and monocytes (> 85% pure) were radiolabelled with
1.5 µCi/107 cells of 111In-oxine (Amersham, Oakville, ON, USA) and
75 µCi/5 × 107 cells of Na251CrO4 (Amersham), respectively as previously described.16,18 After washing, the cells were injected i.v. into
rats, each receiving 4–6 × 106 51Cr-labelled monocytes bearing
approximately 1.5 × 105 cpm together with 5–10 × 106 PMN carrying approximately 5 × 105 cpm of 111In. These cells were allowed to
circulate and migrate for 2 h prior to sacrifice of the rats.
T lymphocytes were isolated from spleens of naive or arthritic
rats by mincing, lysing of red cells with 0.84% NH4C1 and passage
of the cell suspension through nylon wool columns, as previously
described.24,25 The T-cells were labelled with Na251CrO4 for 45 min,
washed and each rat received 1–2 × 107 T-cells bearing approximately 4–7 × 105 cpm, i.v. 20 h prior to sacrifice. In the same animal,
PMN and T-cell or PMN and monocyte migrations were performed
using dual radiolabel experiments.
In some experiments, dermal inflammatory reactions were
induced by intradermal injection (0.05 mL) in duplicate or triplicate
sites of the indicated agents at the time of i.v. injection of the labelled
leucocytes.
Sample collection
At the time the rats were killed, cardiac blood was obtained to
measure the blood leucocyte count, the leucocyte and plasma
associated radioactivity and the mAb concentration in the serum.
Joints were dissected and segments of limbs sectioned to include
carpal, metacarpal, talar and metatarsal joints for analysis as
previously reported.25 The content of 111In and 51Cr in the joints, as
well as in dermal skin biopsies and internal organs, was determined
with a Wallac LKB 1280 gamma counter (Fisher Scientific, Nepean,
ON, Canada) with spectral spillover correction. Accumulated isotope
in the tissue is expressed as cpm/106 cpm injected i.v.
Histological analysis
Tissue samples, including joints, were fixed in 10% phosphate
buffered formalin, decalcified in formic acid and embedded in
paraffin. Sections were stained with hematoxylin and eosin, and
399
cartilage proteoglycan was stained with Safranin O (Difco, Detroit,
MI, USA) for 60 s as shown previously.26
Adoptive transfer of adjuvant arthritis
Lewis rats were immunized s.c. with 0.5 mg M. butyricum in mineral
oil as above and, 12 days later, cervical, axillary, inguinal, popliteal
and paraortic lymph nodes and spleens were removed aseptically.
Spleens were minced and red cells lysed with 0.84% ammonium
chloride in 10 mM potassium bicarbonate, and lymph nodes were
minced and the cells washed with RPMI-1640 (Sigma Chemical Co.,
St. Louis, MO, USA) to prepare single cell suspensions. These lymphocytes were stimulated with 2.5 µg/mL Concanavalin A (Sigma
Chemical Co.) for 48 h in RPMI-5% FCS, 4mM L-glutamine,
0.5 U/mL penicillin and 0.5 U/mL streptomycin (Sigma Chemical
Co.) at a cell concentration of 2.5 × 106 cells/mL in 75 cm2 tissue
culture flasks (NUNC, Naverville, IL, USA) as described by Taurog
et al.13,14 After 48 h of culture (37°C, 5% CO2) the lymphocytes were
harvested and washed with RPMI-1640 and immediately injected
i.v. into naive rats at a dose of 4 × 107 viable cells/100 g body weight.
Rats were scored daily for clinical arthritis, weighed regularly after
lymphocyte transfer, and talar joint swelling was quantified by
caliper measurements.
Statistical analysis
All data are presented as arithmetic means ±1 SEM. Statistical
significance was determined using non-parametric analysis (KruskalWallis test followed by Mann–Whitney U test) or, when the data
was confirmed for normality, it was analysed by ANOVA followed by
post-hoc Bonferroni correction for multiple comparisons.
Results and Discussion
The molecular interactions involved in the adhesion molecule
cascade, resulting in leucocyte emigration from vessels in
response to inflammatory stimuli, have defined an important
role for selectins and VLA-4 in mediating the initial capture
and rolling of leucocytes. These molecular mechanisms have
been defined primarily by the use of intravital microscopy of
exteriorized vasculature (e.g. mesenteric vessels, cremaster
muscle) or in in vitro flow chamber systems.2,3,21–23 Since
such studies showed that disruption of the capture and rolling
phase of leucocyte recruitment essentially prevented leucocyte emigration, we undertook to define the role of specific
selectins and VLA-4 in mediating the intense PMN and
monocyte migration to, and infiltration of, the joints of rats
with AA. Animals were studied at the height of the arthritis,
14–15 days after M. butyricum immunization, and then
again after the rapid accumulation over a 2 h period of i.v.
injected radiolabelled PMN and monocytes in the joint, the
area was quantified. Figure 1 shows the accumulation of 51Cr
labelled blood monocytes and 111In labelled PMN in the talar
joint, which is the most intensely inflamed. Results in this
joint are largely representative of results in the metatarsal and
carpal joints (not shown).27 Monocyte accumulation in the
inflamed joint in arthritic controls was intense, exceeding by
more than 20-fold the migration to normal talar joints in
non-arthritic animals (not shown). As can be seen, treatment
of the animals with mAb to α4 of VLA-4 by the i.v. route at
the time of labelled monocyte injection partially inhibited
monocyte accumulation in the lesions. In contrast, mAb to
E-selectin had no effect while anti-P-selectin treatment
400
AC Issekutz et al.
Figure 1 The effect of α4 (VLA-4) of LFA-1 blockade on the
selectin-independent monocyte and neutrophil migration to the
talar joints of arthritic rats. Fifteen days post-immunization with
M. buturicum, rats with polyarticular arthritis were injected i.v.
with control mAb, or with function-blocking mAbs to the
selectins (open bars) and integrins (solid bars), alone or in
combination as indicated. Immediately afterwards, 51Cr-labelled
blood monocytes and 111In labelled blood neutrophils were
injected i.v. and, after a 2 h migration period, the animals were
euthanized and 51Cr and 111In accumulation in the tissues was
determined as described in Methods. Values for background
labelled monocyte and PMN accumulation in the joints of naive
(non-arthritic) but mAb treated (anti-α4 plus anti-E- and Pselectin or anti-LFA-1 plus anti-E- and P-selectin) animals was
subtracted from the groups (mean talar 51Cr monocytes = 1143
cpm and for 111In PMN = 978 cpm). Values are the means + SEM
of 3–25 animals in each group. *P < 0.05 relative to arthritic
control; **P < 0.01; ***P < 0.001 compared to treatment with
the anti-selectin mAb treated group in each subset of three. (Modified with permission from Birner et al.)27
partially inhibited monocyte accumulation in the joints. There
was no further inhibition by combined anti-E-selectin +
P-selectin ± L-selectin mAb treatment, as shown by the open
bars. This degree of inhibition is comparable to that observed
in previous studies in this model.19 In order to investigate
whether VLA-4 could mediate the remaining monocyte
migration to the joints, separate groups of rats were injected
i.v. with blocking mAb to E- and/or P-selectin plus mAb to
α4 or LFA-1. As shown in Figure 1, blockade of α4 integrin
in addition to E- and P-selectin significantly decreased
monocyte accumulation in the talar joint by a further 58%
relative to animals treated only with mAb to E- + P-selectin.
This resulted in an overall inhibition of monocyte migration
by 83% relative to control mAb treated rats (checkered bar).
Addition of mAb to L-selectin to this combination caused no
further inhibition (left bar set in panel).
We also investigated whether other leucocyte integrins
(aside from α4) involved in monocyte migration, could
mediate the selectin independent monocyte accumulation.
Monoclonal antibody to LFA-1 combined with mAb to Eand P-selectin also significantly inhibited monocyte migration (by up to 67%) to the talar and other joints (not shown)
relative to E- + P-selectin blockade alone. In other combinations with the selectins, LFA-1 blockade had a similar effect
to α4 integrin blockade in the presence of either antiE-selectin, anti-P-selectin, or anti-E- + P-selectin treatment.
Among the selectins, it was clear that P-selectin and one of
these integrins, in particular VLA-4 integrin, were the most
important for monocyte migration to the arthritic joints. This
was even more evident in the carpal joint (not shown but
reported by Birner et al.)27 where P-selectin blockade alone
inhibited monocyte recruitment by 48%, combination of
P-selectin and α4 integrin blockade suppressed migration to
the carpal joints by 80%, while blockade of P-selectin
together with LFA-1 was no more inhibitory than P-selectin
or LFA-1 blockade alone (i.e. 45–55%). Thus, among the
CAM interactions known to mediate leucocyte capture and
rolling, P-selectin and VLA-4 integrin appear to be the two
most important mechanisms for monocyte migration to
inflamed joints in this chronic arthritis model. The role of
LFA-1 appears to be somewhat joint-dependent. The reason
for this is not clear as yet, although variables such as organization of the microvasculature, local blood flow and shear
dynamics during inflammation may play an important role,
since under low shear force conditions LFA-1 may also
mediate leucocyte rolling.28,29 To analyse these events in the
different joints would require intravital microscopy studies of
the inflamed synovium and periarticular tissues.
The role of selectins and α4 (VLA-4) and LFA-1 integrins
in PMN migration to arthritic joints
The lower panel in Figure 1 shows results of simultaneous
measurements of PMN migration to the talar joints in the
same rats reported in the left panel. The PMN migration over
a 2 h period to these joints was intense, being more than
20-fold greater than migration to control, non-arthritic talar
joints in naive rats. This migration was not significantly
affected by α4 integrin blockade, but was partially inhibited
by LFA-1 blockade as previously shown.18 Migration was
slightly inhibited by P-selectin blockade and this inhibition
was significantly enhanced by E-selectin blockade reaching
61% inhibition in the talar joint as also previously shown.19
There was no further inhibition by additionally blocking
L-selectin. These findings suggest that there is an important
role, in particular for P-selectin and E-selectin functioning
in concert, for mediating PMN migration to these joints.
However, there appeared also to be a substantial component
(approximately 40%) of PMN migration which was selectin
independent. As with monocyte migration, E-selectin
blockade alone had no effect. This finding is consistent with
Integrins and selectins in arthritis
other studies including ones of E-selectin knockout mice,
suggesting that E-selectin alone is not an essential selectin
mediating recruitment of inflammatory cells and that Pselectin, and perhaps other mechanisms, may compensate
for loss of the E-selectin function.1,2,30 The data in Figure 1
indicates that α4 integrin (VLA-4) may be an important alternate mechanism to E-selectin, since when α4 was blocked in
combination with E-selectin there was a marked and synergistic inhibition (by 55%) of PMN migration to the joints,
even though individual blockade of E-selectin or α4 had no
effect. VLA-4 appears to serve such an alternate role, also
when with P-selectin, since an additional inhibitory effect
was also observed with α4 plus P-selectin blockade compared
to P-selectin blockade alone. However, this was less evident
in the presence of multiple selectin blockade, suggesting that
VLA-4 (α4) integrin played an important role in substituting
for either E-selectin or P-selectin in mediating a large part
of PMN recruitment, but that it could not substitute for both
E- and P-selectin in serving this function.
It has recently been suggested that LFA-1 on PMN may
also (under some circumstances) mediate PMN rolling, and
that LFA-1 integrin is considerably more abundant on PMN
than the α4 integrin.28,29 Therefore, it was of interest to
evaluate whether the LFA-1 integrin could mediate PMN
recruitment in the absence of the more classically recognized
CAM mechanisms for PMN capture and rolling, the selectins
in particular. These studies indeed show that when P-selectin
mechanisms were blocked, LFA-1 blockade markedly
potentiated the inhibition of PMN migration to the joints,
especially when both E- and P-selectins were blocked. In this
case, inhibition of PMN migration with LFA-1 blockade
caused a further 63% decrease compared to E- and P-selectin
blockade alone. Compared to control antibody treated
animals, overall inhibition was 86%. Such a potent synergistic inhibitory effect was not observed when VLA-4
integrin was blocked in concert with E- and P-selectin,
suggesting that when these endothelial selectin mechanisms
are inoperative, LFA-1 on PMN (perhaps in part because of
its greater abundance on PMN than VLA-4) predominantly
mediates the residual, approximately 30% of PMN migration
of arthritic joints. In comparison to LFA-1 blockade, blockade of the Mac-1 integrin (CD11b/CD18) of the β2 family, in
combination with multiple selectin blockade, had no effect
whatsoever (not shown), demonstrating a relatively unique
role for LFA-1 in mediating selectin independent recruitment
of PMN to inflamed joints.27
Thus, overall from these studies, it appears that for monocyte migration to arthritic joints the two single most important
CAM mechanisms recognized to mediate capture and rolling,
P-selectin and VLA-4 mechanisms, function in concert to
mediate approximately 83% of the overall monocyte recruitment. In contrast, LFA-1 is the predominant selectin independent mechanism of PMN recruitment to the joints with LFA-1
functioning in concert in particular with P-selectin, accounting for up to 85% of the PMN migration to arthritis.27 In the
case of PMN, P- and E-selectin also appear to function in
concert and jointly mediate approximately 60% of the PMN
recruitment as well.19 Although the VLA-4 integrin mechanism also contributes to PMN recruitment under some conditions, this is less important in mediating selectin independent
recruitment than is LFA-1 on PMN.
401
The role of selectins and α4 (VLA-4) integrin in the
development phase of adjuvant arthritis
Having defined the relative contributions of E- and P-selectin
and VLA-4 to PMN and monocyte migration at time of active
arthritis, the importance of these mechanisms during the
development of joint inflammation was examined. Towards
this end, rats were treated with the blocking mAbs, or control
mAb, starting 5 days after immunization with M. butyricum.
By day 5, the immune response to M. butyricum is well underway but this time precedes the development of clinical arthritis
by about 5 days. As shown in Figure 2, treatment of the rats
with mAb to P-selectin from days 5 to 14 had no effect on the
development of clinical arthritis between days 10 to 14, despite
the fact that P-selectin blockade alone inhibited by about
50% of the migration of PMN and monocytes to the joints in
the acute migration experiments shown in Figure 1. In contrast and surprisingly, anti-E-selectin treatment delayed the
onset and markedly decreased the severity of arthritis scores
by over 70%, despite the fact that anti-E-selectin treatment
alone initiated at the height of arthritis (day 14) had no effect
on PMN or monocyte migration to the joints as in Figure 1
and reported previously.19 Combined long-term blockade of
P-selectin and E-selectin with mAb had no greater inhibitory
effect than E-selectin blockade alone. This suggests that in
the development of AA, E-selectin dependent mechanisms
play a major role prior to initial joint inflammation and that
P-selectin is not critical at this phase. This lack of a role for
P-selectin in the development phase of arthritis is in accord
with observations in P-selectin deficient mice in which
collagen induced arthritis is unaffected or even enhanced.31
Since we had shown that the α4 integrin functions in
concert with selectin mechanisms during the active phase of
arthritis, we also examined the role of this integrin during
the developing phase using the same treatment regime. As
shown in Figure 2, treatment with mAb to α4 of VLA-4 integrin delayed the onset and partially inhibited the development
of clinical arthritis by day 14, although significantly less than
anti-E-selectin treatment alone. These findings suggest that
in addition to E-selectin dependent mechanisms, α4 integrin
plays a role in arthritis development. This conclusion was
further supported by the fact that combined blockade of both
the VLA-4 and E-selectin nearly completely prevented
development of clinical arthritis in these animals. The attentuation or prevention of arthritis development by E-selectin
or VLA-4 + E-selectin blockade was manifest not only in the
decreased arthritis scores but also by normal weight gain in
both these groups of animals between day 5 to 14 postimmunization (i.e. increase of 30.2 ± 7.8 g) while control
mAb treated animals lost –11.7 ± 7.9 g (difference P < 0.05).
Furthermore, histology of the joints shown in Figure 3
revealed a marked decrease in the leucocyte infiltration of the
synovium and periarticular tissues ((h) and (e) strain in (e))
compared to control arthritic joints (b). Furthermore, there
was striking preservation of cartilage proteoglycan and cartilage surfaces (Safranin O strain) with either anti-E-selectin
(d) or anti-E-selectin + anti-VLA-4 (f) treatment compared to
the extensive loss of proteoglycan and cartilage erosion in
control mAb treated rats (c).
The severity of joint inflammation was also quantitated by
migration of radiolabelled PMN and monocytes to the joints
402
AC Issekutz et al.
Figure 2 The effect of treatment with mAb to P-selectin, and/or
E-selectin, and/or α4 (VLA-4) integrin on development of adjuvant
arthritis. Treatment with mAb was initiated 5 days after immunization with M. butyricum to induce arthritis. Anti-P-selectin, antiE-selectin, or anti-VLA-4 mAb, or control (anti-pertussis toxin)
mAb were injected i.p. on alternate days as in Methods. Values
are mean ± SEM. *P < 0.05 by Mann–Witney two sample test.
(Modified with permission from 26.) , Controls (n = 22); ––,
Anti-P-sel d5-14 (n = 5); ––, Anti-E-sel d5-14 (n = 10); ––,
Anti-(P & E) sel d5-14 (n = 6); ––, Anti-VLA-4 (n = 5); ––,
Anti-VLA-4 + anti-E-sel d5-14 (n = 4).
at day 14 using methods as in Figure 1. As shown in Figure 4,
rats treated with anti-P-selectin mAb alone from day 5 to 14,
at day 14 the migration over 2 h of radiolabelled PMN and
monocytes to the joints was inhibited by approximately
50–60% shown for the carpal and talar joints, but clearly this
was not of a sufficient degree to impact on the severity of the
arthritis. In contrast, E-selectin blockade alone from day 5 to
14 markedly attenuated the development of PMN and monocyte migration to the joints by 85–95%. This contrasts strikingly with the findings that acute E-selectin blockade at the
height of the arthritis at day 14 had no effect on inflammatory
cell recruitment. This illustrates the differing roles of Eselectin during the developing phase of arthritis compared to
the active phase. Figure 4 also shows that combined blockade
of VLA-4 with E-selectin also resulted in inhibition of PMN
and monocyte migration to the joints, although this was not
greater than E-selectin blockade alone, despite the fact that
the clinical joint inflammation scores and joint swelling
measurements were markedly diminished (Fig. 2). The reason
for this lack of a further decline in monocyte and PMN
migration to joints, despite histological and clinical near
elimination of joint disease, raises interesting questions about
potentially additional functions served by these CAM aside
from mediating leucocyte migration. The VLA-4 integrin for
example may also have a role in leucocyte activation and/or
the balance between leucocyte survival versus apoptosis of
various cell types, including T-cells in inflammation.32–36
Thus, disruption of this integrin’s function, may downregulate joint disease by attenuating lymphocyte or other
leucocyte activation or shortening the survival of the leucocytes that enter the joint. Such additional roles of the CAM
in vivo are being investigated in disease models.
The differing role of E-selectin during the course of
arthritis suggested that E-selectin dependent mechanisms
may play a role in initiation of joint inflammation by
arthritogenic T-cells, believed to be reactive with cartilage
proteoglycan and/or M. butyricum heat shock protein, which
must migrate as recently activated T-cells (T-cell blasts)
to peripheral tissues, including joints, a few days after immunization. These cells are believed to initiate acute synovitis,
observed histologically by day 7 or 8 and clinically by day
10.12,15 To assess whether E-selectin blockade may have had
such an effect, an intradermal delayed type hypersensitivity
(DTH) recall reaction to M. butyricum in immunized animals
was induced and the recruitment of radiolabelled T-cells
to the sites was quantified at 20 h. Migration to these DTH
reactions was decreased by 80% in rats treated with mAb
to E-selectin (day 5–14) while migration of T-cells to
Concanavallin A, polyI:C or IFN-γ + TNF induced dermal
reactions were not significantly depressed (not shown, see
Issekutz et al.).26 Inhibition of the M. butyricum specific
T-cell recruitment to DTH was also observed with animals
treated with anti E-selectin mAb on only days 5 and 7. Such
abbreviated E-selectin blockade attenuated development of
clinical arthritis as well as repeated therapy between days 5
and 14.26 Despite having markedly depressed DTH response
to M. butyricum, the anti-E-selectin and anti-VLA-4 +
E-selectin treated animals did establish an anti-M. butyricum
T-cell response since their lymph node lymphocytes had
normal proliferative responses in vitro to M. butyricum (not
shown).26 Thus, the primary and central immune and memory
response in these animals was not modified by CAM blockade. This supports the notion that E-selectin blockade
prevented the expression of peripheral T-cell dependent
inflammation by interfering with antigen reactive T-cells
reaching non-lymphoid tissues during the initial immune
response. In animals where both VLA-4 and E-selectin were
blocked starting from day 5, the DTH recall reaction was
nearly completely inhibited (95%), although in these animals
migration to IFN-γ + TNF, ConA and polyI:C was also
suppressed due to the important role of VLA-4 in conjunction
with E-selectin in mediating T-cell recruitment to dermal
inflammation as shown recently by us.37
It is now well recognized that resting T-cells express few
ligands for either P- or E-selectin and thus, they do not
recognize these activation induced molecules on vascular
endothelium. However, following activation (e.g. via T-cell
receptor and IL-2), sialylated Lewisx related carbohydrate
ligands are synthesized and displayed on proteins such as
P-selectins glycoprotein ligand-1, E-selectin ligand-1 and
others.38–40 At least under in vitro conditions, expression of
ligands for P-selectin on T-cells appears to be more restricted
than for E-selectin. T-cell migration to skin inflammation is
associated with expression of a highly glycosylated ligand
for E-selectin, the cutaneous lymphocyte antigen.41 In the
context of this knowledge, our findings regarding arthritic
joints and dermal DTH reaction are compatible with the
hypotheses that during the initial immune response to
M. butyricum, activated T-cells primarily utilize E-selectin
to migrate to skin and to joints due to selective expression of
E-selectin ligands without a requirement for P-selectin. Our
findings also suggest that VLA-4 on T-cells, which is also
upregulated in expression and in affinity state for ligands
Integrins and selectins in arthritis
403
Figure 3 Histology of talar joints from normal, control mAb, anti-E-selectin mAb, and anti-VLA-4 plus E-selectin mAb treated rats
immunized with M. butyricum. The hematoxylin/eosin stain in (a) shows a normal non-arthritic joint with thin, paucicellular synovium
(arrows) while (b) is from a control mAb treated arthritic rat displaying synovial expansion and leucocyte infiltration (arrow). (c) is a
Safranin O stained section of control mAb rat showing articular cartilage proteoglycan loss (left arrow) and synovial infiltrate adhering
to and eroding cartilage (right arrow). (d) is from an anti-E-selectin treated rat while (e) and (f) are sections from anti-VLA-4 + E-selectin
mAb treated animals. Note marked decrease in leucocyte infiltrate in synovium in (e) (arrow) with preservation of cartilage proteoglycan and surface integrity in (d) and (f). Original magnification 40×.
such as VCAM-1 following T-cell activation,36,42 may either
cooperate with E-selectin dependent mechanisms or function
as an alternate CAM mechanism on a subset of T-cells for
mediating activated, arthritogenic T-cell migration to the
joints. However, since VLA-4 is expressed on virtually all
T-cells to some degree, its potential role as a modulator of
primary arthritogenic T-cell migration to the joints versus
secondary ‘inflammatory’ T-cell and/or monocyte recruitment to the joints contributing to the amplification phase of
joint inflammation requires further investigation. It appears
from the above results that, during the development phase of
AA, the VLA-4 integrin dependent mechanism appears to be
less critical and less selective than the E-selectin dependent
mechanism.
404
AC Issekutz et al.
Figure 4 The effect of treatment with mAb to P-selectin,
E-selectin and VLA-4 on PMN and monocyte migration to joints
in rats with adjuvant arthritis. Treatment of rats with anti-P- and
anti-E-selectin and VLA-4 mAbs was as in Figure 2 or just 2 h
prior to sacrifice on day 14 with anti-E-selectin mAb as indicated.
PMN and monocyte migration to the joints was determined on
day 14 using 111In or 51Cr labelled rat blood PMN neutrophil and
monocytes, respectively, injected i.v. and allowed to accumulate
in the joints for 2 h. Quantitation was as in Figure 1. Values
shown are mean + SEM of 5 to 10 animals in each group.
*P < 0.01. (Modified with permission from 26.) , Control mAb;
, Anti-P-sel d5-14; , Anti-E-sel d5-14; , Anti-P- + E-sel
d5-14; , Anti-VLA-4 + anti-E-sel d5-14; , Anti-E-sel d14.
The role of selectins and α4 integrin in adoptive T-cell
adjuvant arthritis
In order to further define the molecular mechanisms involved
in T-cell initiation of AA, a model of adoptive transferred AA,
in which lymph node and spleen T-cells collected from rats
with early active AA (day 12) transfer disease to naive rats
following 48 h of ex vivo stimulation of the cells with ConA
was used.14 As shown in Figure 5(a), control rats receiving
transferred lymphocytes begin to develop arthritis by day 5
post-transfer, which progresses rapidly over 5–7 days to
maximum and stable arthritic scores. This, like the active
AA, is a polyarticular arthritis involving primarily the distal
joints of forelimbs and hindlimbs. Using this model, we
tested the above hypothesis that E-selectin is required for
activated arthritogenic T-cell migration to joints to initiate
arthritis. As can be seen from Figure 5(a), administration of
mAb to E-selectin at time of ConA expanded lymphocyte
transfer and subsequently on alternate days up to day 6 had
no effect on the course of arthritis development or severity.
Treatment with mAb to P-selectin alone or in combination
with mAb to E-selectin also had no effect on adoptive AA.
This suggests that, unlike actively induced AA, neither of
these selectins play an essential role in T-cell transfer of AA.
We also examined the role of VLA-4, since α4 blockade in
the active AA had an inhibitory effect on AA development
(Fig. 2). As seen in Figure 5(b), anti-VLA-4 treatment
markedly attenuated the development of arthritis and inhibited the severity by at least 65%, based on arthritis scores.
This was observed even when a single injection of mAb was
administered at the time of lymphocyte transfer. In contrast,
treatment with mAb to LFA-1 administered alone at the time
of lymphocyte transfer did not have a significant effect.
Furthermore, LFA-1 blockade did not potentiate the
inhibitory effect of VLA-4 blockade, suggesting that LFA-1
has little, if any, role in mediating arthritogenic T-cell migration to normal joints and induction of AA following adoptive
transfer. The marked effect of VLA-4 blockade could not be
attributed to lymphocyte depletion or sequestration because
when ex vivo ConA stimulated lymphocytes were radiolabelled with 51Cr immediately prior to i.v. injection, followed
by administration i.v. of control or anti-α4 integrin mAb, the
distribution of radiolabelled lymphocytes in spleen, lymph
nodes (mesenteric, popliteal, cervical, axillary, inguinal,
para-aortic) in lung and liver were comparable. Similarly, the
lymphocyte associated 51Cr in blood was comparable between
these treatment groups and there was no effect on free 51Cr in
the plasma with anti-α4 integrin treatment (not shown). Thus,
antibody, complement or Fc receptor mediated lymphocyte
clearance mechanisms cannot account for the inhibition of
adoptive transfer of the AA body by the anti-α4 mAb treatment. It would be desirable to measure the trafficking of the
radiolabelled arthritogenic lymphocytes to the joints within
the first 24–48 h. However, as has been shown with cloned
T-cells, an extremely small number of antigen reactive T-cells
are sufficient to induce arthritis and the sensitivity of the
radiolabelled lymphocyte method is not sufficient to detect
this process.12,15
As shown in the active AA model, the VLA-4 integrin
plays a complementary role with E-selectin in the development of the arthritis. Therefore, we investigated the potential
contribution of selectin mechanisms to adoptive AA development in conjunction with VLA-4. As shown in Figure 5(c),
combination of P-selectin along with VLA-4 blockade had no
effect or in fact tended to increase the severity of arthritis
when compared to VLA-4 blockade alone. On the other hand,
combination of E-selectin blockade with VLA-4 blockade
slightly delayed the onset and severity of arthritis, but this
was not statistically significant. However, combination of
E- + P-selectin blockade with VLA-4 blockade essentially
prevented development of arthritis. Analysis of these results
is shown in Figure 5(d). This confirmed that blockade of the
VLA-4 integrin significantly delayed the onset of arthritis,
and that the maximum arthritis score and the combined
blockade of the VLA-4 with E- + P-selectin inhibited significantly further both of these parameters relative to VLA-4
blockade alone.
The effects of these adhesion molecule interventions were
striking also at the histological level as shown in Figure 6. At
time of joint examination, 19 days post-lymphocyte transfer,
the joints of control mAb treated rats showed intense periarticular and synovium leucocyte infiltrates with synovial
hyperplasia (Fig. 6(a)), associated with extensive cartilage
surface destruction and even collapse (Fig. 6(b)). As shown
in Figure 6(c), VLA-4 blockade markedly decreased the
Integrins and selectins in arthritis
405
Figure 5 The effect of selectin and integrin blockade on adoptively transferred adjuvant arthritis. Rats were injected i.v. on day 0 with
ConA stimulated (48 h) lymphocytes harvested from arthritic donors, as described in Methods. Immediately thereafter, they received i.v.
injection in (a) of mAb to rat E-selectin or rat E- + P-selectin 3 mg, singly or followed by 2 mg on alternate days (results pooled) as
indicated by the arrows. , Control; , Anti-E-selectin; , Anti-E-Sel+P-sel. In (b), the rats received mAb to VLA-4 (α4) integrin and/or
to LFA-1 (αL, CD11a) 3 mg i.v. as a single dose on day 0. , Control; , Anti-LFA-1; , Anti-VLA-4; , Anti-VLA-4 + LFA-1. In (c),
the indicated combinations of anti-VLA-4 ± selectin antibody were administered by a single injection on day 0. , Control; , Anti-VLA-4
+ P-selectin; , Anti-VLA-4; (), Anti-VLA-4 + E-selectin; , Anti-VLA-4 + E + P-selectin. Rats were scored for clinical arthritis
as in the Methods and followed for up to 19 days. Values are mean ± SEM of 4–10 animals in each group. In (d), statistical analysis of the
time of onset and maximum clinical arthritis in the different treatment groups is summarized. *P < 0.05; **P < 0.021; ***P < 0.01.
, Control; , Anti-VLA-4; , Anti-LFA-1; , Anti-VLA-4 + LFA-1; , Anti-VLA-4 + E-sel; , Anti-VLA-4 + P-sel; , AntiVLA-4 + E-sel + P-sel.
leucocyte infiltration. This was associated with preservation
of cartilage proteoglycan and cartilage surface integrity
as shown in Fig. 6(d). Furthermore, combined blockade of
VLA-4 with E- + P-selectin resulted in joints that were
virtually normal histologically (Fig. 6e,f). In addition to
suppression of the development of arthritis in these animals,
they gained weight normally during the 19 days (antiVLA-4 + E- + P-selectin = 78.5 +/– 12.9 g) and significantly
more rapidly than rats treated with anti-VLA-4 mAb alone
(+ 40.4 +/– 6.0 g; P < 0.05) while control rats failed to gain
weight at all (6.8 +/– 6.7 g, n = 8).
Taken together, these results indicate that the VLA-4
integrin plays a major role in the adoptive T-cell transfer of
AA to naive animals, a role that is analogous to that observed
with active AA, although to a lesser extent in the active AA
model. A major difference between the active versus adoptive
transfer AA is the marked role of an E-selectin dependent
mechanism in the active AA with apparently no requirement
for E-selectin function with the ConA stimulated adoptive
T-cell transfer of AA. However, while E-selectin is not essential for the adoptive AA, one of the two endothelial selectins
E- or P-selectin are important and likely function as an
alternate for the other selectin, most likely mediating T-cell
migration to the naive joints during initiation of AA. While
the results suggest that the adoptive AA model does not
completely reflect the mechanisms mediating active AA, this
406
AC Issekutz et al.
Figure 6 Histology of talar joints from rats with adoptive adjuvant arthritis, and effect of VLA-4 and selectin blockade. (a),
(c) and (e) show hematoxylin and eosin stained sections from joint at day 19 of control mAb, anti-VLA-4 and anti-VLA-4 +
E- + P-selectin mAb treated rats respectively. Correspondingly, (b), (d) and (e) are Safranin O stains for cartilage, in the same
treatment order as (a), (c) and (e). Note intense leucocyte infiltration of synovium (upper arrow) with invasion of periarticular
bone (lower arrow) in (a) and in (b), loss of all articular cartilage (arrow 1), but not of metaphyseal cartilage proteoglycan (arrow
3) and collapse of articular surface (arrow 2). There was preservation of cartilage in (d) and (f) and decrease (in (c)) or absence
(in (e)) of leucocyte infiltration with VLA-4 and VLA-4 plus E- and P-selectin blockade respectively. Original magnification 40×.
may be due to the requirement for ConA stimulation of the
lymphoctyes ex vivo in order to have effective transfer of
AA. To our knowledge, it is not possible to transfer AA with
lymphocytes directly from arthritic to naive animals, at least
not into normal unmanipulated naive recipients (i.e. not
immunosuppressed, irradiated or thymectomized).13,14 It is
possible that during ConA stimulation and culture, the
activated T-cells express sufficient ligands for both E- and
P-selectin and likely increase their expression of VLA-4,
especially in high affinity conformation to an extent greater
than occurs in vivo following active immunization and
antigen stimulation. Such effects on activated T-cells in vitro
have been observed previously, especially with polyclonal
activation. This may skew the balance of CAM mechanisms
Integrins and selectins in arthritis
utilized by the cells for subsequent in vivo interaction with
vascular endothelium.40,42,43 While these results indicate that
caution needs to be exercised in extrapolating observations
made with in vitro cultured cells in regards to the CAM
mechanisms for mediating in vivo trafficking, our results
with active and adoptive transfer AA both emphasize that the
α4 (VLA-4) integrin is an important mechanism for T-cell
initiation of the cascade of events leading to chronic arthritis.
Furthermore, this integrin functions in concert with one or
more of the endothelial expressed selectins, particularly
E-selectin following active immunization resulting in the
initiation of chronic arthritis. The adoptive AA studies
suggest that under certain conditions, perhaps with polyclonal activation in vivo (e.g. by superantigens) a broader
range of CAM mechanisms may mediate activated T-cell dissemination to non-lymphoid tissues, that is not only VLA-4
and E-selectin but also P-selectin. This may be due to expression of a broader range of selectin ligands on T cells under
certain activating conditions.4,40,43 It will now be up to future
studies to investigate whether targeting specific selectins,
such as E-selectin alone and in combination with the VLA-4
integrin, may not only prevent the development of actively
induced or spontaneous arthritis in other models, but also
modify the course of established arthritis. If our hypothesis
that antigen reactive T-cell trafficking to the joints is Eselectin and VLA-4 dependent is correct, then blocking one
or both of these CAM for sufficient time, even in established
arthritis, may eventually downregulate chronic joint inflammation. If this is borne out, then the combination of immunomodulatory therapy, to directly inhibit T-cell activation within
the joint, and E-selectin ± VLA-4 integrin blockade, might
further limit activated T-cell recruitment and chronicity of
arthritis. Investigations to evaluate this strategy are currently
underway.
Acknowledgements
This work was supported by Grant 89036 from the Arthritis
Society of Canada, Grants MOP-7684, MOP-57769 and
MOP-42379 from the Canadian Institutes of Health Research,
and a training grant to S Nakazato from Otsuka Pharmaceuticals Inc. The authors gratefully acknowledge the excellent
technical assistance of Ms C Jordan and Mr D Rowter and the
outstanding secretarial help of Ms M Hopkins.
References
1 Kubes P, Ward PA. Leukocyte recruitment and the acute inflammatory response. Brain Pathol. 2000; 10: 127–35.
2 Ley K, Tedder TF. Leukocyte interactions with vascular endothelium; new insights into selectin-mediated attachment and
rolling. J. Immunol. 1995; 155: 525–8.
3 Springer TA. Traffic signals on endothelium for lymphocyte
recirculation and leukocyte emigration. Annu. Rev. Physiol.
1995; 57: 827–72.
4 Von Andrian UH, Mackay Cr. T-cell function and migration. Two
sides of the same coin. N. Engl. J. Med. 2000; 343: 1020–34.
5 Smith MD, Slavotinek J, Au V, Weedon H et al. Successful
treatment of rheumatoid arthritis is associated with a reduction
in synovial membrane cytokines and cell adhesion molecule
expression. Rheumatology (Oxford) 2001; 40: 965–77.
407
6 Carson CW, Beall LD, Hunder GG, Johnson CM, Newman W.
Soluble E-selectin is increased in inflammatory synovial fluid.
J. Rheumatol. 1994; 21: 605–11.
7 Grober JS, Bowen BL, Ebling H et al. Monocyte-endothelial
adhesion in chronic rheumatoid arthritis. In situ detection of
selectin and integrin-dependent interactions. J. Clin. Invest.
1993; 91: 2609–19.
8 Ilgner S, Stiehl P. Strong LFA-1 and VCAM-1 expression in
histological type II of rheumatoid arthritis. Cell. Mol. Biol.
2002; 48: OL243–9.
9 Bloom BJ, Nelson SM, Alario AJ, Miller LC, Schaller JG.
Synovial fluid levels of E-selectin and intercellular adhesion
molecule-1: relationship to joint inflammation in children with
chronic arthritis. Rheumatol. Int. 2002; 22: 175–7.
10 Klimiuk PA, Sierakowski S, Latosiewicz R et al. Soluble
adhesion molecules (ICAM-1, VCAM-1, and E-selectin) and
vascular endothelial growth factor (VEGF) in patients with
distinct variants of rheumatoid synovitis. Ann. Rheum. Dis.
2002; 61: 804–9.
11 Harris E, D. ed. Joints and connective tissue: The targets of
rheumatoid arthritis. In: Harris E, D. ed. Rheumatoid Arthritis.
Toronto: W.B. Saunders Company, 1997; 3–17.
12 Taurog JD, Argentieri DC, McReynolds RA. Adjuvant arthritis.
Methods Enzymol. 1988; 162: 339–55.
13 Taurog JD, Sandberg GP, Mahowald ML. The cellular basis of
adjuvant arthritis. II. Characterization of the cells mediating
passive transfer. Cell. Immunol. 1983; 80: 198–204.
14 Taurog JD, Sandberg GP, Mahowald ML. The cellular basis of
adjuvant arthritis. I. Enhancement of cell-mediated passive
transfer by concanavalin A and by immunosuppressive pretreatment of the recipient. Cell. Immunol. 1983; 75: 271–82.
15 Van Eden W, Hogervorst EJ, Hensen EJ, van der Zee R,
van Embden JD, Cohen IR. A cartilage-mimicking T-cell epitope
on a 65K mycobacterial heat-shock protein: adjuvant arthritis as
a model for human rheumatoid arthritis. Curr. Top Microbiol.
Immunol. 1989; 145: 27–43.
16 Issekutz AC, Issekutz TB. Monocyte migration to arthritis in the
rat utilizes both CD11/CD18 and very late activation antigen 4
integrin mechanisms. J. Exp. Med. 1995; 181: 1197–203.
17 Issekutz AC, Issekutz TB. A major portion of polymorphonuclear leukocyte and T lymphocyte migration to arthritic joints
in the rat is via LFA-1/MAC-1-independent mechanisms. Clin.
Immunol. Immunopathol. 1993; 67: 257–63.
18 Issekutz TB, Miyasaka M, Issekutz AC. Rat blood neutrophils
express very late antigen 4 and it mediates migration to
arthritic joint and dermal inflammation. J. Exp. Med. 1996; 183:
2175–84.
19 Walter UM, Issekutz AC. The role of E- and P-selectin in
neutrophil and monocyte migration in adjuvant-induced
arthritis in the rat. Eur. J. Immunol. 1997; 27: 1498–505.
20 Walter UM, Issekutz AC. Role of E- and P-selectin in migration
of monocytes and polymorphonuclear leucocytes to cytokine
and chemoattractant-induced cutaneous inflammation in the rat.
Immunology 1997; 92: 290–9.
21 Alon R, Kassner PD, Carr MW, Finger EB, Hemler ME,
Springer TA. The integrin VLA-4 supports tethering and rolling
in flow on VCAM-1. J. Cell. Biol. 1995; 128: 1243–53.
22 Berlin C, Bargatze RF, Campbell JJ et al. alpha 4 integrins
mediate lymphocyte attachment and rolling under physiologic
flow. Cell.1995; 80: 413–22.
23 Johnston B, Issekutz TB, Kubes P. The alpha 4-integrin
supports leukocyte rolling and adhesion in chronically inflamed
postcapillary venules in vivo. J. Exp. Med. 1996; 183:
1995–2006.
408
AC Issekutz et al.
24 Issekutz TB. Inhibition of in vivo lymphocyte migration to
inflammation and homing to lymphoid tissues by the TA-2
monoclonal antibody. A likely role for VLA-4 in vivo.
J. Immunol. 1991; 147: 4178–84.
25 Issekutz AC, Issekutz TB. Quantitation and kinetics of polymorphonuclear leukocyte and lymphocyte accumulation in
joints during adjuvant arthritis in the rat. Lab. Invest. 1991; 64:
656–63.
26 Issekutz AC, Mu JY, Liu G, Melrose J, Berg EL. E-selectin, but
not P-selectin, is required for development of adjuvant-induced
arthritis in the rat. Arthritis Rheum. 2001; 44: 1428–37.
27 Birner U, Issekutz TB, Walter U, Issekutz AC. The role of alpha
(4) and LFA-1 integrins in selectin-independent monocyte and
neutrophil migration to joints of rats with adjuvant arthritis.
Int. Immunol. 2000; 12: 141–50.
28 Henderson RB, Lim LH, Tessier PA et al. The use of lymphocyte
function-associated antigen (LFA)-1-deficient mice to determine the role of LFA-1, Mac-1, and alpha4 integrin in the
inflammatory response of neutrophils. J. Exp. Med. 2001; 194:
219–26.
29 Gaboury JP, Kubes P. Reductions in physiologic shear rates lead
to CD11/CD18-dependent, selectin-independent leukocyte
rolling in vivo. Blood 1994; 83: 345–50.
30 Wolitzky BA. The role of E-selectin and P-selectin in inflammatory disease. In: Peltz G, (ed). Leukocyte recruitment in
inflammatory disease. New York: Springer; 1996; 211–45.
31 Bullard DC, Mobley JM, Justen JM et al. Acceleration and
increased severity of collagen-induced arthritis in P- selectin
mutant mice. J. Immunol. 1999; 163: 2844–9.
32 Leussink I, Zettl K, Jander S, Pepinsky B, Lobb R, Stoll G,
Toyka V, Gold R. Blockade of signaling via the very late antigen
(VLA-4) and its counterligand vascular cell adhesion molecule1 (VCAM-1) causes increased T cell apoptosis in experimental
autoimmune neuritis. Acta. Neuropathol. (Berl.) 2002; 103:
131–6.
33 Westermann J, Engelhardt B, Hoffman JC. Migration of T cells
in vivo: molecular mechanisms and clinical implications. Ann.
Intern. Med. 2001; 135: 279–95.
34 Kanda E, Jin ZH, Mizuchi D, Arai A, Miura O. Activation of
Rac and tyrosine phosphorylation of cytokine receptors induced
by cross-linking of integrin alpha4beta1 and cell adhesion in
hematopoietic cells. Biochem. Biophys. Res. Commun. 2003;
301: 934–40.
35 May AE, Neumann FJ, Schomig A, Preissner KT. VLA-4
(alpha(4)beta(1)) engagement defines a novel activation pathway for beta(2) integrin-dependent leukocyte adhesion
involving the urokinase receptor. Blood 2000; 96: 506–13.
36 Lobb RR, Abraham WM, Burkly LC et al. Pathophysiologic role
of α4 integrins in the lung. Ann. N.Y. Acad. Sci. 1996; 796:
113–23.
37 Issekutz AC, Issekutz TB. The role of E-selectin, P-selectin, and
very late activation antigen-4 in T lymphocyte migration to
dermal inflammation. J. Immunol. 2002; 168: 1934–9.
38 Knibbs RN, Craig RA, Maly P et al. Alpha(1,3)-fucosyltransferase VII-independent synthesis of P- and E- selectin ligands on
cultured T lymphoblasts. J. Immunol. 1998; 161: 6305–15.
39 Lim YC, Henault L, Wagers AJ, Kansas GS, Luscinskas FW,
Lichtman AH. Expression of functional selectin ligands on Th
cells is differentially regulated by IL-12 and IL-4. J. Immunol.
1999; 162: 3193–201.
40 Schottelius AJ, Hamann A, Asadullah K. Role of fucosyltransferases in leukocyte trafficking: major impact for cutaneous
immunity. Trends Immunol. 2003; 24: 101–4.
41 Berg EL, Yoshino T, Rott LS et al. The cutaneous lymphocyte
antigen is a skin lymphocyte homing receptor for the vascular
lectin endothelial cell-leukocyte adhesion molecule 1. J. Exp.
Med. 1991; 174: 1461–6.
42 Jakubowski A, Ehrenfels BN, Pepinsky RB, Burkly LC. Vascular cell adhesion molecule-Ig fusion protein selectively targets
activated alpha 4-integrin receptors in vivo. Inhibition of
autoimmune diabetes in an adoptive transfer model in nonobese
diabetic mice. J. Immunol. 1995; 155: 938–46.
43 Hamann A, Klugewitz K, Austrup F, Jablonski-Westrich D.
Activation induces rapid and profound alterations in the trafficking of T cells. Eur. J. Immunol. 2000; 30: 3207–18.