| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* BioXell, Milan, Italy;
Istituto Dermopatico dell’Immacolata, Istituto del Ricovero e Cura a Carattere Scientifico, Rome, Italy; and
Section of General Pathology, Department of Pathology, University of Verona, Verona, Italy
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
4 integrins, namely
4
1 (very late Ag-4)
and the mucosal homing receptor
47 have been shown to
support tethering and rolling (7). Chemokines, which
generate heterotrimeric Gi protein-dependent
signaling pathways, are physiological activators of rapid
integrin-dependent lymphocyte arrest on endothelial cells
(8, 9, 10, 11). Once extravasated, leukocytes navigate through complex chemoattractant gradients by integrating conflicting chemotactic signals within tissues (8, 12). The complexity of the chemokine system appears well suited to convey the large body of information needed to guide the navigation of leukocytes from the blood stream to their final destination within tissues (13, 14). However, the promiscuity and redundancy in ligand-chemokine receptor interactions and the intricate pattern of chemokine receptor expression on leukocytes highlight a scarcely understood level of complexity.
To gain novel insights into chemokine-mediated control of lymphocyte recruitment, we analyzed the microenvironmental presentation of CCR4 ligands C-C chemokine ligand (CCL)3 17 and CCL22 and compared their capacity to trigger rapid integrin-dependent lymphocyte adhesion under conditions of physiological flow as well as their ability to trigger other CCR4-mediated phenomena. Paradoxically, we found that the chemokine CCL17, found on the endothelium, is not the higher potency CCR4 agonist. Interestingly, the dominance of CCL22 over CCL17 correlates with distinct kinetics of CCR4 engagement. By using human Th1 or Th2 cells expressing different levels of the chemokine receptors CCR4 and CXCR3, we find that quantitative differences in chemokine receptor level and agonistic potency of the ligands are two distinct parameters, which are integrated at the cellular level to generate diversity in lymphocyte subset proadhesive responses.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Human Th cell lines were generated as previously described by stimulation of CD8 T cell-depleted cord blood mononuclear cells with 2 µg/ml PHA (Wellcome, Beckenham, U.K.) in the presence of various combinations of cytokines and anti-cytokine Abs (15). Th1 cells were generated by the addition of 5 ng/ml IL-12 (Hoffmann-LaRoche, Nutley, NJ) and 200 ng/ml neutralizing anti-IL-4 Ab (BD PharMingen, San Diego, CA). Th2 cells were generated by the addition of 10 ng/ml IL-4 (BD PharMingen) and 2 µg/ml neutralizing anti-IL-12 Abs 17F7 and 20C2 (Hoffmann-LaRoche). The cells were cultured in complete medium (RPMI 1640 (Sigma-Aldrich, St. Louis, MO) supplemented with 5% FetalClone (HyClone Laboratories, Logan, UT), 2 mM L-glutamine, 1 mM sodium pyruvate, and 100 U/ml penicillin-streptomycin). On day 3, the cultures were washed and expanded in complete medium with addition of 100 U/ml IL-2 (Hoffman-LaRoche). Mouse L1.2 pre-B cells were cultured in complete medium at 1 x 106/ml (5 x 106) and were transfected with 3 µg of full length hCCR4 cDNA (a gift from S. Bhakta, Roche Bioscience, Palo Alto, CA) and 20 µg of Lipofectamine reagent (Life Technologies, Gaithersburg, MD). Forty-eight hours after transfection, cells were cultured in complete medium supplemented with 0.8 mg/ml G418 (Life Technologies). L1.2 stably expressing hCCR4 were sorted by FACS by staining with anti-hCCR4 Ab 1G1 (BD PharMingen).
Cell surface staining
Cells were washed in FACS buffer (50 mM phosphate, 150 mM NaCl, pH 7.4; 1% FetalClone; 0.05% sodium azide) and incubated with anti-hCCR4 (BD PharMingen) or anti-hCXCR3 (R&D Systems, Minneapolis, MN) for 30 min on ice, washed, and analyzed by FACScan flow cytometry (BD Biosciences, San Jose, CA).
Analysis of intracellular calcium mobilization
Indo1-acetoxymethyl ester loading was performed by incubating the cells (5 x 106/ml) in buffer A (HBSS with 10 mM HEPES) with 2.5 µM Indo1-acetoxymethyl ester and 0.05% w/v F-127 Pluronic (Molecular Probes, Eugene, OR) at 37°C for 30 min. The incubation was prolonged for 30 min after the addition of an equal volume of buffer B (HBSS with 10 mM HEPES and 5% FCS). Cells were washed twice in buffer B, resuspended at 1 x 106/ml, and analyzed by FACS before and after stimulation with the indicated chemokines (Dictagene, Epalinges, Switzerland). Data were recorded every 0.5 s as the relative ratio of fluorescence emitted at 395 and 480 nm after excitation at 364 nm. Mean 395/480 fluorescence ratio (y-axis) vs time (x-axis) were calculated using FlowJo analysis software (BD Biosciences). Fold induction of intracellular Ca2+ concentration ([Ca2+]i) was calculated by dividing the peak fluorescence ratio (395 nm:480 nm) of stimulated cells by the peak fluorescence ratio of unstimulated cells. Percentage of cross-desensitization = 100 - [(x - 1/y - 1) x 100], where x represents the fold induction of [Ca2+]i of secondary stimulation and y represents the fold induction of [Ca2+]i of primary stimulation.
Immunohistochemistry
Punch biopsies of normal skin from healthy individuals (n = 2), chronic atopic dermatitis (n = 3), and psoriasis lesions (n = 3) were embedded in OCT, snap-frozen in liquid nitrogen, and stored at -80°C until sectioning. Cryostatic sections of 4 µm were fixed in 5% paraformaldehyde for 10 min, treated with 0.3% hydrogen peroxide to quench endogenous peroxidase activity, and incubated with normal horse serum (Vectastain ABC kit; Vector Laboratories, Burlingame, CA) for 20 min. Double immunostainings were performed with anti-CCL22 mAb (clone 272D; 5 µg/ml) (20) or rabbit anti-hCCL17 Ab (12 µg/ml) (PeproTech, Rocky Hill, NJ) and anti-Factor VIII/von Willebrand factor (vWF) (1/25) (DAKO, Glostrup, Denmark). Single stainings for CCL22, CCL17, and Factor VIII/vWF were also performed; to verify the specificity of the Abs used, sections were subsequently incubated with mouse or rabbit total serum (Vector Laboratories) as controls for the anti-CCL22 or anti-CCL17 Abs, respectively, and with purified mouse IgG1 (BD Biosciences) as controls for the anti-Factor VIII/vWF Abs. Avidin-biotin-peroxidase or avidin-biotin-phosphatase activities were revealed with 3-amino-9-ethylcarbazole and Blue Vector (Vector Laboratories), respectively. For each biopsy, 10 sections were stained with each Ab, and positive cells were evaluated in four adjacent fields at a magnification of x200.
Analysis of chemokine receptor internalization and ligand dissociation rates
Internalization of CCR4 was analyzed by incubating 1 x 106/ml Th2 cells in the absence or presence of various concentrations of CCL17 or CCL22 for the indicated time at 37°C or 4°C. In selected experiments, the cells were incubated for 2 h at 37°C with 10 µg/ml of pertussis toxin (Sigma). Subsequently, Th2 cells were washed twice in ice cold FACS buffer and stained with Ab to CCR4 or an isotype-matched control and analyzed by FACS. Percentage of receptor internalization were (%) = [(x - k)/(y - k)] x 100, where x represents mean fluorescence intensity of CCR4 staining after internalization, y represents mean fluorescence intensity of CCR4 staining before internalization, and k represents mean fluorescence intensity of isotype control. For evaluation of ligand-binding dissociation rates, L1.2-hCCR4 transfectants were washed and resuspended in binding buffer (125 nM NaCl, 25 mM HEPES, 1 mM CaCl2, 5 mM MgCl2, 0.5% BSA, pH 7.0) and incubated for 2 h on ice. For each assay point, 1 x 106 cells in 0.1 ml of binding buffer were incubated on ice in the presence of 250 pM human recombinant 125I-labeled CCL17 or 125I-labeled CCL22 (specific activity, 2000 Ci/mmol; Amersham Pharmacia Biotech, Little Chalfont, U.K.). After 2 h, the cells were washed twice and resuspended in 1 ml of ice cold binding buffer. At various times, the cells were washed twice in ice cold binding buffer and lysed in 2% SDS, and radioactivity was counted with a gamma counter. Nonspecific binding was calculated by addition of a 500-fold molar excess of unlabeled CCL17 or CCL22.
Analysis of lymphocyte adhesion under flow
VCAM-1 was engineered as an IgG fusion protein using human IgG1
CH2-CH3 domains onto which the extracellular domains of human VCAM-1
was fused (kindly provided by Drs. U. Gubler and L. Renzetti,
Hoffman-LaRoche). VCAM-1 sequence was from aa 1–696 ending 2 residues
before the transmembrane domain. The construct was expressed in
Drosophila cells, purified by affinity chromatography from
lysates, and stored at -80°C. Recombinant extracellular domain of
human VCAM-1 (native VCAM-1) purified from Chinese hamster ovary cells
(R&D Systems) was also used with essentially the same results. Before
use, VCAM-1-IgG fusion protein (0.5 mg/ml) was dialyzed against PBS
containing 1% -octyl glucoside. Microcap glass capillary tubes (100
µl capacity; Drummond Scientific, Broomall, PA) were coated for
16 h at 4°C with 20 µl of human VCAM-1 at 2000
sites/µm2. Site densities per square micrometer
of immobilized VCAM-1 were calculated using a
125I-labeled anti-human IgG1 H chain mAb, as
previously described (16). Before use, tubes were washed
and cocoated with 20 µl of 2 µM chemokines for 60 min. After a
washing with PBS, the behavior of interacting Th1 and Th2
lymphocytes was recorded on S-VHS videotape (Panasonic, Verona,
Italy) and analyzed frame by frame, as described
(17). Th1 or Th2 cells were resuspended at a concentration
of 1.5 x 106/ml and injected by applying a
wall shear stress of 2 dyne/cm2 with the help of
a digital syringe pump. After 2–3 min needed to establish laminar
flow, single areas of 0.2 mm2 were recorded for
at least 30 s. Interactions (rolling, arrest, or both) of 1 s
were considered significant and were scored. Lymphocytes that remained
firmly adherent for 10 s were considered fully adherent
(11).
Quantification of chemokine immobilization
Sections (10 mm long) of 100-µl Microcap glass capillary tubes were coated with human VCAM-1 at 2000 sites/µm2, as described above. Human recombinant 125I-labeled CCL17 and 125I-labeled CCL22 (specific activity, 2000 Ci/mmol; Amersham Pharmacia Biotech) were reconstituted at 100 µCi/ml in PBS. A labeled/unlabeled (1/100) mixture of CCL17 and CCL22 was made containing 5 pmol of 125I-labeled chemokines in 100 µl of PBS. Ten microliters of chemokine mixture (corresponding to 50 pmol of chemokine) were added to the capillary tubes to cocoat a 10-mm-long section. For the competition binding assay, radioactive CCL17 was mixed with nonradioactive CCL22 or vice versa (corresponding to 100 pmol of total chemokines). After variable incubation times at room temperature, the tubes were washed with 10 ml of PBS at a flow rate of >10 dyne/cm2. Radioactivity bound to the tubes was quantified with a gamma counter and transformed in number of molecules per square micrometer. Background binding to glass in the absence of VCAM-1 was calculated for both chemokines and was subtracted from the binding in the presence of immobilized VCAM-1. The number of molecules of each chemokine specifically immobilized by one molecule of VCAM-1 was finally calculated.
Statistical analysis
SDs were calculated and statistical significance was assessed by the paired two-tailed Student t test. Values of p < 0.05 were considered to be significant.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Recent data have suggested a critical role for the chemokine
receptor CCR4 in recruitment of cutaneous memory T cells into the
inflamed skin (18, 19, 20, 21). Because CCL17/TARC and CCL22/MDC
have been identified as two distinct ligands of CCR4 (22, 23), we wished to determine their expression pattern in the
inflamed human skin by analysis of lesional skin from atopic dermatitis
and psoriasis patients. In both psoriasis and atopic dermatitis,
expression of CCL17 was detected on endothelial cells of the majority
of vessels (85%) present in both the superficial and reticular
dermis and on tissue-infiltrating cells with the morphology of
dendritic cells (Fig. 1
and data not
shown). In sharp contrast, in both skin disorders, CCL22 was not
detected on endothelial cells, although it was detectable on
tissue-infiltrating dendritic cells (Fig. 1 and data not shown).
Neither CCL17 nor CCL22 could be detected in normal skin (Ref.
20 and data not shown). These results indicate a distinct
pattern of microenvironmental presentation for CCL17 and CCL22 in
inflamed skin, with CCL17 but not CCL22 being specifically presented by
endothelial cells. Notably, this distinct pattern of chemokine
presentation was not markedly different in diseases as diverse as
psoriasis and atopic dermatitis, suggesting that the depicted tissue
distribution of CCL17 and CCL22 could be a general feature of these
chemokines.
|
4
integrin-dependent adhesion of human Th2 cells under conditions of
physiological flow
The results reported above suggested that CCL17 and CCL22 binding
to CCR4 could act at distinct locations, with CCL17 being uniquely
devoted at capturing leukocytes from the blood stream. The role of
CCL17 and CCL22 may not be restricted to regulating T cell recruitment
to the skin given that CCR4 expression is found on activated T cells,
Th2 cells, and regulatory T cells (24, 25, 26), suggesting the
involvement of this receptor in different phases of T cell
localization. This scenario fostered us to compare the ability of CCL17
and CCL22 to promote rapid integrin-dependent adhesion under conditions
of physiological flow. Given the possibility to generate large numbers
of human CD4+ Th cells with a well-characterized
profile of chemokine and adhesion receptors, we initially used in
vitro-generated human Th2 cells, which express high levels of CCR4 as
well as 4 integrins (Fig. 2A) (25, 27).
Because 4 integrins have been shown to support
transient adhesive interactions, namely tethering and rolling due to
their microvillous distribution (28), we were able to
investigate the interactions of Th2 cells on immobilized
41 integrin ligand
VCAM-1, under physiological flow conditions. However, in the absence of
chemokines, VCAM-1 poorly supported transient adhesion of Th2 cells
even at high site density (data not shown). Because CCR4 is highly
expressed on Th2 cells (Fig. 2A), we analyzed the effect of
coimmobilizing CCL17 or CCL22, the two ligands of CCR4, with VCAM-1.
The two chemokines coimmobilized with VCAM-1 induced a powerful
up-regulation of tethering and rolling, rapidly followed by firm
adhesion of Th2 cells (Fig. 2B). Although both chemokines
were similarly efficient inducers of tethering and rolling, CCL22 was
much more efficient than CCL17 in triggering conversion from rolling to
firm adhesion of Th2 cells on VCAM-1 (Fig. 2B). Importantly,
neutralizing anti-CCR4 Abs were able to completely inhibit the
adhesive interactions observed with CCL17 or CCL22 but not CXCL12
(Fig. 2C). These findings confirm the recently reported
rapid up-regulation of
41 integrin-dependent
lymphocyte tethering and rolling by chemokines (29). Most
importantly, they demonstrate that the two CCR4-sharing chemokines
CCL17 and CCL22 elicit qualitatively distinct types of
integrin-mediated interactions under conditions of physiological
flow.
|
To verify that the differences observed were not due to a
different degree of chemokine immobilization on VCAM-1, we next
quantified the number of molecules of CCL17 or CCL22 immobilized in
presence of purified human VCAM-1. Various concentrations of
radiolabeled CCL17 or CCL22 were incubated in the presence of
immobilized VCAM-1 and the number of bound chemokine molecules was
quantified. This experiment revealed a high number of CCL17 and CCL22
molecules that can specifically bind to one molecule of integrin ligand
(Fig. 3). Binding was rather rapid in
that it was detectable within 15 min, reaching a plateau within 30 min.
Saturation was reached at chemokine concentration of 5 µM for both
CCL17 and CCL22 (data not shown). Interestingly, VCAM-1 was capable of
absorbing CCL17 more efficiently than CCL22 (Fig. 3A). The
specificity of these interactions was confirmed by the inability of
immobilized albumin and human IgG to bind the chemokines (Fig. 3A) and by the ability of CCL17 to displace more efficiently
CCL22 binding to VCAM-1 (Fig. 3B). These data show for the
first time that VCAM-1 is able to specifically bind chemokines and
suggest that endothelial integrin ligands can act as highly efficacious
chemokine-presenting molecules. Similar results have been observed with
mucosal addressin cell adhesion molecule-1, and further studies are in
progress to characterize these interactions (D. D’Ambrosio and C.
Laudanna, unpublished results). Importantly, these experiments
show that CCL17 is immobilized 
2- to 3-fold more efficiently than
CCL22, indicating that the lower efficiency of CCL17 in triggering
rapid arrest of Th2 cells under flow is not due to reduced presentation
of this chemokine.
|
The expression level of chemokine receptors on memory/effector T
cells is highly heterogeneous, and it has been suggested to be a
critical parameter regulating rapid lymphocyte adhesion under flow
(17). To evaluate how variations in the expression levels
of chemokine receptors influenced the efficiency of integrin-dependent
adhesion triggered by CCL17 and CCL22, we took advantage of in
vitro-derived human Th1 cells. Th1 cells express levels of
4 integrins similar to Th2 cells but poorly
express CCR4 (Fig. 4A). When
the interactions of Th1 cells on immobilized VCAM-1 in the presence of
coimmobilized CCL17 or CCL22 were analyzed under flow, we found that
coimmobilization of CCL17 with VCAM-1, although able to induce a
moderate up-regulation of Th1 cell rolling, was ineffective in
triggering the complete transition from rolling to firm adhesion (Fig. 4B). By contrast, CCL22 consistently triggered a remarkable
level of firm adhesion of Th1 cells (Fig. 4B). Importantly,
on Th1 cells expressing low levels of CCR4, the difference between
CCL17 and CCL22 in triggering integrin-dependent firm adhesion becomes
an all or none phenomenon (Fig. 4B). Our data indicate that
the agonistic potency of the chemokine cooperates with the expression
level of the receptor to quantitatively and qualitatively regulate
integrin-dependent lymphocyte adhesion.
|
Taken together, our data suggest that CCL22 is more potent than
CCL17 at promoting integrin-dependent arrest of lymphocytes, a
conclusion that is consistent with previous analysis of CCR4 binding
affinity of the two ligands (22, 23) and their ability to
trigger CCR4-mediated signaling. However, because it is conceivable
that different chemokines may be presented on endothelial cells at
different densities, it was of interest to investigate whether
variations in the density of a presented chemokine could compensate for
intrinsic differences among ligands. To this purpose, we assayed the
interactions of Th1 cells on VCAM-1 in the presence of decreasing
amounts of CCL22. The usage of Th1 cells expressing low CCR4 was
instrumental to test the critical role of chemokine density in
suboptimal conditions of chemokine receptor expression. Under these
conditions, coimmobilization of 2 µM CCL17 together with VCAM-1
resulted in a marked induction of Th1 cell tethering and rolling with
only few arrested cells (Fig. 5). In
contrast, 2 µM CCL22 triggered firm adhesion of more cells (Fig. 5).
Analysis of the spectrum of adhesive interactions triggered by
decreasing amounts of immobilized CCL22 revealed that only when it was
immobilized at a concentration 200-fold lower than that of CCL17 did
the two chemokines behave similarly by promoting almost exclusively
rolling (Fig. 5). It should be added that given the 2- to 3-fold
higher efficiency of CCL17 vs CCL22 binding to VCAM-1 (Fig. 3
), the
actual quantitative difference in potency between the two CCR4 ligands
might even be underestimated.
|
Our analysis of the efficiency of induction of rapid adhesion by
CCL17 and CCL22 indicated a remarkable level of dominance of CCL22 over
CCL17. This seemed somewhat surprising given the relatively small
difference of CCR4 binding affinity between the two ligands (22, 23). Thus, we performed a quantitative evaluation of calcium
mobilization, desensitization, and internalization of CCR4, which are
not directly related to integrin activation in response to CCL17 or
CCL22 but provide a more direct measurement of the agonistic potency of
the two chemokines. CCL22 consistently induced more robust calcium
mobilization in Th2 cells (Table I and
data not shown). Moreover, stimulation of Th2 cells with CCL22
followed by stimulation with a 10- or 20-fold molar excess of CCL17 or
vice versa demonstrated an absolute dominance of CCL22 over CCL17 in
the ability to desensitize CCR4 (Table I). Because it is theoretically
possible that CCL22 binds to an unidentified receptor in addition to
CCR4, we have extended our analysis to mouse L1.2 pre-B cells
transfected with hCCR4 receptor (L1.2-hCCR4). L1.2 cells transfected
with hCCR4 receptor but not parental L1.2 cells mobilized calcium in
response to CCL17 or CCL22 (Table I and data not shown). As seen with
Th2 cells, CCL22 was dominant over CCL17 in desensitizing calcium
mobilization in L1.2-hCCR4 even at a 20-fold lower concentration
(Table I).
|
-arrestins and promote receptor sequestration and
internalization (30). Given the hierarchy of CCR4
desensitization induced by CCL17 and CCL22, we investigated how the two
ligands elicited CCR4 internalization. Human Th2 cells were incubated
at 37°C for various times with CCL17 or CCL22, and the cell surface
CCR4 expression was monitored by Ab staining. Strikingly, CCL22
promoted a rapid and potent down-modulation of CCR4 expression, whereas
CCL17 had a minimal effect (Fig. 6A). CCL22 elicited a marked
CCR4 down-modulation after 5 min (Fig. 6A). A dose-dependent
increase in CCR4 internalization was observed, with CCL22 being
effective at a concentration as low as 6.2 nM, whereas CCL17 was poorly
effective even at a 20-fold higher concentration (125 nM) (Fig. 6A). Similar findings were observed using L1.2-hCCR4 cells
(data not shown). To confirm that the reduction of CCR4 staining
intensity was due to sequestration/internalization of CCR4 rather than
binding competition, we performed incubations at 4°C to inhibit
receptor internalization. At 4°C, no significant down-regulation of
CCR4 expression was observed with either CCL17 or CCL22 (Table II). To evaluate the signaling
requirements for ligand-induced internalization of CCR4, we analyzed
the effects of inhibiting inhibitory G subunit
(Gi protein) signaling using pertussis
toxin. Internalization of CCR4 induced by CCL22 was unaffected by
pertussis toxin (Fig. 6B) showing that this is independent
of Gi signaling pathways. In contrast, pertussis
toxin treatment completely inhibited the adhesion of lymphocytes under
flow (data not shown). These data indicate that the hierarchy of
dominance of CCL22 over CCL17 extends beyond the induction of rapid
integrin-dependent adhesion to diverse CCR4-mediated phenomena.
|
|
Previous data have shown that CCL17 and CCL22 bind to CCR4 with
different affinities (22, 23), with CCL22 exhibiting 2-
to 3-fold higher affinity than CCL17. Thus, higher CCR4 affinity could
result in higher receptor occupancy by CCL22 relative to CCL17, and
this may help establish a functional hierarchy between these
chemokines. However, it is evident that the dose range over which CCL22
exhibited clear dominance over CCL17 went well above a 3-fold
difference (Tables I and II and Fig. 5). Thus, the reduced efficiency
of CCL17 in triggering receptor desensitization and internalization as
well as rapid integrin-dependent adhesion under flow is unlikely to be
explained only on the basis of a lower number of receptors
occupied.
This raised the possibility that differences in the modality of
interaction of CCL17 or CCL22 with CCR4 could contribute to establish
the observed phenomenological diversity. Activation of signaling
pathways is critically dependent on triggering the active conformation
of a threshold number of receptors and/or the duration of single
ligand/receptor interactions (31, 32). To evaluate the
kinetics of ligand/receptor interaction, we compared time-dependent
dissociation of CCL17 or CCL22 bound to CCR4. L1.2-hCCR4 cells were
incubated with 250 pM radiolabeled CCL17 or CCL22, and after removal of
nonbound chemokine the amount of radioactivity that remained associated
to the cells was quantified at different times. At time 0, we found
1700 molecules of CCL22 vs 500 molecules of CCL17 specifically
bound per cell, consistent with the reported higher affinity of CCL22
for CCR4 (23). Surprisingly, this experiment revealed a
faster rate of dissociation of CCL22 vs CCL17 (Fig. 6C).
Similar results were obtained using human Th2 cells (data not shown).
These findings indicate that CCL22 and CCL17 bind to CCR4 with
different kinetics, raising the possibility that this difference could,
at least in part, explain the observed phenomenological diversity.
The regulatory role of receptor expression level and agonistic potency as a general paradigm of chemokine-induced integrin-dependent rapid lymphocyte adhesion
We finally wished to test whether the observations made with CCR4
and its ligands represent a general paradigm of regulation of
integrin-dependent rapid lymphocyte adhesion by chemokines. To this
end, we analyzed the adhesive interactions of Th1 or Th2 cells on
immobilized VCAM-1 in response to CXCL9 and CXCL11, which bind
specifically to the CXCR3 receptor that is preferentially expressed on
Th1 cells (Figs. 2A and 4A). Previous work has
shown that CXCL11 displays higher affinity and agonistic potency than
CXCL9 for CXCR3 (33, 34). Analysis of intracellular
calcium mobilization in response to CXCL9 and CXCL11 confirmed that
CXCL11 is dominant over CXCL9 on human Th1 cells as well as on human
CXCR3-transfected L1.2 cells (data not shown). In flow adhesion assays,
both chemokines induced a marked up-regulation in the number of
tethering/rolling and firmly adherent Th1 cells on VCAM-1 (Fig. 7A). CXCL11 was consistently
more efficient than CXCL9 in the conversion of rolling into firmly
adherent Th1 cells (Fig. 7A). When analysis was performed on
Th2 cells, CXCL11 was still able to elicit a significant number of
interactions, a fraction of which consisted of firmly adherent cells
(Fig. 7B). In contrast, CXCL9 was able to induce moderate
rolling but poor arrest of Th2 cells (Fig. 7C). Overall,
these data depict a pattern of interactions induced by CXCL9 and CXCL11
that mirrors that seen with CCL17 and CCL22.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
First, we documented that CCL17 but not CCL22 is expressed on vascular
endothelium in distinct skin inflammatory conditions in vivo (Fig. 1).
Moreover, we showed a more efficient immobilization on VCAM-1 of CCL17
vs CCL22. Notably, CCL22 but not CCL17 is sensitive to CD26 proteolytic
degradation, indicating that CCL17 may be a more stable ligand adept at
presentation by endothelial cells (36, 37, 38).
Together, these results suggest that CCL17 but not CCL22 is devoted to
trigger vascular recognition of CCR4 expressing cells such as cutaneous
memory T cells or Th2 cells.
Starting from these observations, we wished to verify whether CCL17 was
the most efficient proadhesive agonist for CCR4-expressing lymphocytes.
To this end, we used in vitro-derived human Th1 and Th2 cells as a
model system. These cells proved ideal for our analysis because large
numbers can be generated in vitro and exhibit distinctive profiles of
chemokine receptors. To simplify our analysis, we took advantage of the
fact that endothelial vascular ligand VCAM-1 can support, upon
chemokine triggering, lymphocyte tethering, rolling, and arrest under
physiological flow, likely due to subsecond induction of
41 integrin
clustering induced by chemokine receptor signaling (29).
In accordance with this study, we found that immobilization of CCL17 or
CCL22 promotes pertussis toxin-sensitive transient tethering and
rolling of T cells on VCAM-1 (data not shown). The involvement of
integrin 41 in our
setting was confirmed by the inhibitory effect of blocking Abs (data
not shown).
Data obtained with adhesion assays under flow showed striking differences in the ability of chemokines to trigger firm arrest. In Th1 cells (which express low level of CCR4), CCL22 but not CCL17 induced full arrest. The level of chemokine receptor expression was suggested to be a relevant parameter for induction of lymphocyte arrest under flow conditions (39). However, the striking difference between CCL22 and CCL17 cannot be explained simply in terms of lower receptor occupancy, because in Th2 cells, which express high levels of CCR4, CCL22 was still much more efficient than CCL17 in triggering firm adhesion. Furthermore, CCL22 was remarkably more powerful than CCL17 in triggering CCR4 desensitization and internalization, indicating that the dominance of CCL22 over CCL17 extends well beyond the induction of integrin activation to also affect Gi-independent signaling pathways. These differences prompted us to investigate the possible molecular basis by evaluating how the two ligands interact with the receptor. Our findings that the higher affinity ligand CCL22 dissociates from the receptor more rapidly than CCL17 suggest that differential kinetics of chemokine receptor binding may trigger qualitatively different biological responses.
An intriguing possibility is that the frequency of association/dissociation of ligand-receptor complexes is a critical parameter regulating the activation of certain intracellular signaling pathways. This idea is supported by recent data showing that the frequency more than the amplitude of intracellular signaling events controls the specificity and magnitude of gene expression (40, 41) and calmodulin-dependent protein kinase II activation (42). Another possibility is that the conditions of flow impose a tight temporal restrain for a productive signal-generating chemokine receptor engagement on lymphocytes that rapidly flow and contact vessel walls. Coimmobilization of chemokines with VCAM-1 has been reported to be necessary for triggering T cell rolling and adhesion under flow (29). The kinetic parameter inherent to the conditions of flow may help establish a "velocity threshold" for chemokine receptor occupancy that is required for adhesion triggering by the ligand. In this scenario, the efficiency of lymphocyte adhesion under flow would critically depend on the kinetics of chemokine receptor occupancy. Thus, it would be predicted that chemokines with faster rates of receptor association/dissociation, as in the case of CCL22 binding to CCR4, could more easily achieve a kinetic receptor occupancy threshold required for induction of integrin-dependent adhesion under flow. Regardless of the exact mechanism, our data suggest that a different modality of CCR4 engagement by CCL22 induces a quantitatively and/or qualitatively different signal transduction leading to phenomenological diversity.
To investigate whether these findings could be generalized to another
chemokine ligand/receptor system, we also analyzed the activities
triggered by two distinct ligands of CXCR3, a receptor that, opposite
to CCR4, is expressed at higher levels on Th1 cells. CXCL9 and CXCL11
have been reported to exhibit distinct binding characteristics and
agonistic potencies with respect to CXCR3, with CXCL11 being the
dominant ligand (34, 43, 44). Furthermore, a recent study
has reported the selective expression of the low potency CXCL9 on HEV
from inflamed tissue (45). When CXCL9 and CXCL11 were
analyzed for their ability to elicit rapid integrin-dependent adhesion
and calcium mobilization on Th1 and Th2 cells (Fig. 7 and data not
shown), the pattern of induced phenomena mirrored that seen with CCL17
and CCL22, with CXCL11 being the most potent agonist and CXCL9 almost
unable to trigger arrest of Th2 cells. Again, differences in receptor
binding affinity allow a remarkable level of diversification in
signaling and biological responses.
Overall, our data show that lymphocytes presented with different
chemokines binding to same receptor display completely different
biological responses depending on the modality of receptor engagement
and the level of receptor expression. Importantly, our data indicate
that the most potent chemokine not only increases the absolute number
of adherent lymphocytes, but also triggers the arrest of different
lymphocyte subtypes. In this scenario, a quantitative difference
becomes a qualitative change. Notably, the most potent chemokines seem
not to be presented to circulating lymphocytes. This result seems
paradoxical, given that one could expect that the critical conditions
imposed by the flow require the most potent chemokine to be presented
on vascular endothelium. We propose a possible interpretation for these
findings. Although presented less efficiently, the most potent
chemokine may elicit proadhesive responses in different lymphocyte
subtypes expressing high as well as low level of chemokine receptor
(Fig. 8). Thus, by presenting the less
potent chemokine, the immune system adopts a useful strategy to achieve
the selective recruitment of cells expressing only a high level of a
given chemokine receptor, which corresponds to a peculiar functional
phenotype. In this context, it is significant that CCL17, the less
potent CCR4 ligand, is bound more efficiently to VCAM-1 but still does
not trigger significant adhesion of CCR4low Th1
lymphocytes. In contrast, CCL22 seems so powerful that, even if bound
at 20-fold lower concentrations, it is still able to trigger full
arrest of CCR4low Th1 lymphocytes. From the
viewpoint of selectivity, CCL22 could be considered a dangerous
chemokine. Interestingly, expression of all three CXCR3 ligands (CXCL9,
CXCL10, and CXCL11) was detected on the endothelium of atherosclerotic
plaques (46), which are considered the result of a
chronic, rather nonspecific, inflammatory process. Thus, our findings
underscore a potential mechanism of "diversity breaking" in the
immune system, based on expression on the microvasculature of
chemokines able to elicit full proadhesive responses in leukocyte
subtypes expressing either low as well as high levels of a specific
chemokine receptor (Fig. 8).
|
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Daniele D’Ambrosio, BioXell, Via Olgettina 58, Milan, Italy I-20132. E-mail address: daniele.dambrosio{at}bioxell.com
3 Abbreviations used in this paper: CCL, C-C chemokine ligand; hCCR, human CCR; [Ca2+]i, intracellular Ca2+ concentration; vWF, von Willebrand factor; CXCL, C-X-C chemokine ligand.
Received for publication March 13, 2002. Accepted for publication June 26, 2002.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
4 integrins mediate lymphocyte attachment and rolling under physiologic flow. Cell 80:413.[Medline]
2 integrin affinity and mobility changes: differential regulation and roles in lymphocyte arrest under flow. Immunity 13:759.[Medline]
1 integrin-dependent adhesion in human Th2 T lymphocytes. J. Immunol. 164:3292.4 integrin clustering by immobilized chemokines stimulates leukocyte tethering and rolling on endothelial vascular cell adhesion molecule 1 under flow conditions. J. Exp. Med. 192:495. chemoattractant (I-TAC): a novel non-ELR CXC chemokine with potent activity on activated T cells through selective high affinity binding to CXCR3. J. Exp. Med. 187:2009. chemoattractant are allotopic ligands for human CXCR3: differential binding to recepor states. Mol. Pharmacol. 59:707. chemoattractant (CXCL11). J. Immunol. 167:7084.This article has been cited by other articles:
|
|
 |
|
  J. Bayry, E. Z. Tchilian, M. N. Davies, E. K. Forbes, S. J. Draper, S. V. Kaveri, A. V. S. Hill, M. D. Kazatchkine, P. C. L. Beverley, D. R. Flower, et al. In silico identified CCR4 antagonists target regulatory T cells and exert adjuvant activity in vaccination PNAS, July 22, 2008; 105(29): 10221 - 10226. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Miles, E. Liaskou, B. Eksteen, P. F. Lalor, and D. H. Adams CCL25 and CCL28 promote {alpha}4{beta}7-integrin-dependent adhesion of lymphocytes to MAdCAM-1 under shear flow Am J Physiol Gastrointest Liver Physiol, May 1, 2008; 294(5): G1257 - G1267. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Pasvolsky, V. Grabovsky, C. Giagulli, Z. Shulman, R. Shamri, S. W. Feigelson, C. Laudanna, and R. Alon RhoA Is Involved in LFA-1 Extension Triggered by CXCL12 but Not in a Novel Outside-In LFA-1 Activation Facilitated by CXCL9 J. Immunol., March 1, 2008; 180(5): 2815 - 2823. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 X. Yang, Y. Chu, Y. Wang, R. Zhang, and S. Xiong Targeted in vivo expression of IFN-{gamma}-inducible protein 10 induces specific antitumor activity J. Leukoc. Biol., December 1, 2006; 80(6): 1434 - 1444. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Garrood, L. Lee, and C. Pitzalis Molecular mechanisms of cell recruitment to inflammatory sites: general and tissue-specific pathways Rheumatology, March 1, 2006; 45(3): 250 - 260. [Full Text] [PDF]
|
|
|
|
 |
|
 S. M. Curbishley, B. Eksteen, R. P. Gladue, P. Lalor, and D. H. Adams CXCR3 Activation Promotes Lymphocyte Transendothelial Migration across Human Hepatic Endothelium under Fluid Flow Am. J. Pathol., September 1, 2005; 167(3): 887 - 899. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Lecca, C. Priami, P. Quaglia, B. Rossi, C. Laudanna, and G. Constantin A Stochastic Process Algebra Approach to Simulation of Autoreactive Lymphocyte Recruitment SIMULATION, June 1, 2004; 80(6): 273 - 288. [Abstract] [PDF]
|
|
|
|
 |
|
 T. A. Kohout, S. L. Nicholas, S. J. Perry, G. Reinhart, S. Junger, and R. S. Struthers Differential Desensitization, Receptor Phosphorylation, {beta}-Arrestin Recruitment, and ERK1/2 Activation by the Two Endogenous Ligands for the CC Chemokine Receptor 7 J. Biol. Chem., May 28, 2004; 279(22): 23214 - 23222. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Bonecchi, M. Locati, E. Galliera, M. Vulcano, M. Sironi, A. M. Fra, M. Gobbi, A. Vecchi, S. Sozzani, B. Haribabu, et al. Differential Recognition and Scavenging of Native and Truncated Macrophage-Derived Chemokine (Macrophage-Derived Chemokine/CC Chemokine Ligand 22) by the D6 Decoy Receptor J. Immunol., April 15, 2004; 172(8): 4972 - 4976. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) or
firmly adherent (
) cells during 30 s calculated from 12
distinct 0.2-mm2 areas of the capillary tube from 5
experiments. A significant difference between CCL17 and CCL22 induction
of Th2 cell sticking is shown (*, p < 0.001).
C, Chemokine receptor CCR4 is required for
tethering/rolling and firm adhesion of Th2 cells induced by CCL17 and
CCL22 coimmobilized with VCAM-1. Th2 cells were incubated for 15 min at
room temperature with 5 µg/ml anti-hCCR4 Ab (1G1; BD PharMingen)
or isotype-matched control and then assayed in their ability to adhere
under flow to VCAM-1 in the presence of coimmobilized CCL17, CCL22, or
CXCL12. Values are average number ± SD of tethering/rolling (
