|
|
||||||||


*
Leukocyte Biology Section, Biomedical Sciences Division, Imperial College School of Medicine, South Kensington, London, United Kingdom; and
Millennium Pharmaceutical Inc., Cambridge, MA 002139
| Abstract |
|---|
|
|
|---|
(CDw123)
and negative for HLA-DR, and their increase in forward scatter (as a
result of cell shape change) in response to chemokines was measured.
Shape change responses of basophils to chemokines were highly
reproducible, with a rank order of potency: monocyte chemoattractant
protein (MCP) 4 (peak at <1 nM)
eotaxin-2 =
eotaxin-3
eotaxin > MCP-1 = MCP-3 >
macrophage-inflammatory protein-1
> RANTES = MCP-2 =
IL-8. The CCR4-selective ligand macrophage-derived chemokine did not
elicit a response at concentrations up to 10 nM. Blocking mAbs to CCR2
and CCR3 demonstrated that responses to higher concentrations (>10 nM)
of MCP-1 were mediated by CCR3 rather than CCR2, whereas MCP-4
exhibited a biphasic response consistent with sequential activation of
CCR3 at lower concentrations and CCR2 at 10 nM MCP-4 and above. In
contrast, responses to MCP-3 were blocked only in the presence of both
mAbs, but not after pretreatment with either anti-CCR2 or
anti-CCR3 mAb alone. These patterns of receptor usage were
different from those seen for eosinophils and monocytes. We suggest
that cooperation between CCRs might be a mechanism for preferential
recruitment of basophils, as occurs in tissue hypersensitivity
responses in vivo. | Introduction |
|---|
|
|
|---|
Basophils respond to many CC chemokines, including eotaxin (16, 17), eotaxin-2 (18), eotaxin-3 (19), RANTES (16, 20, 21), monocyte chemoattractant protein (MCP)5-1 (22), MCP-2 (23), MCP-3 (24), and MCP-4 (16, 25, 26). Besides CCR3, which is believed to mediate principally cell recruitment and chemotaxis (as opposed to degranulation) (16, 17, 27), basophils also express other CCRs, including CCR1 and CCR2. Although the role of CCR1 is still unclear, CCR2 is strongly linked to histamine release and leukotriene secretion (16, 22). Major phenotypic similarities with respect to the expression of chemokine receptors (CCR3 in particular) and adhesion molecules (28) by eosinophils and basophils suggest common pathways of recruitment. However, although there is growing evidence for a crucial role of CCR3 in allergy-related tissue eosinophilia (29, 30, 31, 32, 33), the magnitude and kinetics of basophil influx after allergen challenge parallel neither eosinophil recruitment (6, 8, 9, 10, 11, 13, 14, 15, 32), nor generation of eotaxin or eotaxin-2 at the inflammatory site (6, 32). Thus, the existence of basophil-specific chemoattractants or mechanisms of recruitment need to be given serious consideration.
Having encountered a chemotactic factor in vivo, leukocytes immediately begin to rearrange their cytoskeleton and change their shape to facilitate their attachment to microvascular endothelial cells (34, 35, 36, 37, 38, 39). Such responses can be detected by flow cytometry as changes in either forward or right-angle light scattering of the cells (40, 41). On the basis of this observation, we developed a simple flow cytometric assay, the gated autofluorescence forward scatter (GAFS) assay, to measure chemokine responses of human eosinophils in a mixed population of polymorphonuclear leukocytes (PMNL) (42).
Because basophils make up <1% of leukocytes in the blood of healthy donors and are hence difficult to obtain in quantities relevant for experimental studies, current knowledge on the regulation of basophil migration is still limited. In the present study, we have adapted the GAFS assay to measure responses of human basophils to chemokines in a mixed population of PBMC. We have used this basophil-gated forward scatter (BGFS) assay to dissect the roles of CCRs, particularly CCR2 and CCR3, on basophils, and demonstrate the ability of these chemokine receptors to cooperate, suggesting a possible mechanism for the specific regulation of basophil recruitment in vivo.
| Materials and Methods |
|---|
|
|
|---|
BSA, glucose, mouse IgG1 (clone MOPC 21) and IgG2a (clone UPC
10) control Abs, and FITC-conjugated anti-human HLA-DR mAb (clone
HK14) were obtained from Sigma (Poole, U.K.). PBS (with and without
Ca2+ and Mg2+) and HEPES
were purchased from Life Technologies (Paisley, U.K.). FITC-conjugated
F(ab')2 goat anti-mouse Ig and R-PE-labeled
anti-human CD16 mAb were supplied by Dako (Cambridge, U.K.).
FITC-conjugated anti-human CD14 was obtained from Autogen Bioclear
(Calne, U.K.). Dextran T-500 and Percoll were purchased from Amersham
Pharmacia Biotech (St. Albans, U.K.). Human recombinant eotaxin-2,
RANTES, macrophage-inflammatory protein (MIP)-1
, MCP-1, MCP-2,
MCP-3, MCP-4, and MDC were purchased from PeproTech (London, U.K.).
R-PE-conjugated anti-human CDw123 mAb (clone 7G3), FACS-Flow, and
Cellfix were obtained from Becton Dickinson (Mountain View, CA). Mouse
anti-human CCR3 mAb LS-7B11 (isotype IgG2a) (43), CCR2
mAb LS-132.1D9 (isotype IgG2a), CCR1 mAb LS-2D4 (isotype IgG1), CCR4
mAbs LS-1G1 and LS-2B10 (44), and recombinant human
eotaxin and synthetic eotaxin-3 were produced by LeukoSite/Millennium
Pharmaceuticals (Cambridge, MA). Recombinant human IL-8 was a generous
gift from J. White (SmithKline Beecham, King of Prussia, PA). Stock
solutions of chemokines (10 µM) were prepared in PBS containing 0.1%
BSA. Ten millimolar PBS containing 10 mM HEPES, 10 mM glucose, and
0.1% BSA (pH 7.27.4) was used as assay buffer. Staining buffer
consisted of Ca2+/Mg2+-free
PBS with 0.5% BSA and 0.1% azide.
Cell preparation
Volunteer blood donors were healthy normal subjects or atopics, as defined by a history of asthma, eczema, or hay fever, and symptoms on exposure to common aeroallergens, including pollens and house dust mites. The donors were taking no systemic medication. Blood was sampled according to a local ethics committee-approved protocol and was prepared as previously described (45). Briefly, platelet-rich plasma was removed by centrifugation of citrated whole blood, after which the erythrocytes were removed by dextran sedimentation. PMNL (containing neutrophils and eosinophils) were then separated from PBMC over a discontinuous plasma/Percoll gradient. Any erythrocyte contamination of the PBMC or PMNL pellet was removed by hypotonic shock lysis (46). To identify basophils, PBMC (1 x 108/ml) were stained with R-PE-labeled anti-human CDw123 mAb at 2 µg/ml and FITC-labeled anti-human HLA-DR mAb at 0.6 µg/ml for 10 min at room temperature and washed before use. Similarly, monocytes were distinguished by labeling of PBMC with FITC-conjugated anti-human CD14 mAb at 1/100 dilution (42).
In some experiments, purified basophils were prepared according to a modification of the technique of Tsang et al. (47). Citrated blood was centrifuged as above to remove platelets and plasma. The erythrocyte/buffy coat pellet was diluted with PBS and layered over Histopaque 1077 (Sigma) and centrifuged at 400 x g for 20 min. The resulting PBMC layer was incubated with the StemSep basophil purification Ab cocktail (StemCell Technologies, Vancouver, British Columbia, Canada) and colloidal magnetic particles, and unlabeled basophils were separated from other cell types by passage through a magnetic field according to the manufacturers instructions (StemCell Technologies). Cell purity was assessed by staining of an aliquot with anti-CDw123 and anti-HLA-DR mAbs and was found to be routinely >93% basophils, with a cell yield of 8.5 x 104 to 5 x 105/donor (from 80 ml of citrated blood). Any contaminating cells had forward light scatter (FSC)/side light scatter (SSC) characteristics of lymphocytes.
Measurement of leukocyte shape changes using flow cytometry
Agonist-induced shape change was measured using a modification of the protocol described previously (42). The cells were resuspended in assay buffer containing Ca2+ and Mg2+ and left for a short period (510 min at room temperature) to allow equilibration of intracellular and extracellular calcium. In experiments to investigate receptor usage by chemokines, this incubation period was also used to treat cells with buffer, anti-CCR2 mAb LS-132.1D9, or the anti-CCR3 mAb LS-7B11. Aliquots of cells (5 x 105 PMNL or PBMC) were then mixed with the respective agonist or buffer in 1.2-ml polypropylene cluster tubes (Costar, Cambridge, MA) in a final volume of 100 µl, and the tubes were placed in a 37°C shaking water bath for 4 min (basophils and eosinophils) or 10 min (monocytes). In experiments investigating the kinetics of basophil shape change responses, PBMC were incubated with buffer or eotaxin for the time points indicated before fixation and analysis. After these incubation times, which had been determined to be optimal in preliminary experiments (see Results) or in a previous study (42), the cells were transferred to an ice-water bath, and 250 µl of ice-cold fixative (1x Cellfix diluted 1/4 in FACS-Flow) was added to terminate the reaction and maintain the change in cell shape until analysis. The samples were then analyzed immediately on a FACScalibur flow cytometer (Becton Dickinson).
Basophils were identified as a single population of cells that stained
positive for CDw123 (FL-2 fluorescence channel) and negative for HLA-DR
(FL-1 channel; Fig. 1
)
(48, 49, 50). Anti-CD14-labeled monocytes were detected by
their increased fluorescence on the FL-1 channel. Eosinophils were
distinguished from neutrophils by their higher autofluorescence on the
FL-2 channel (42). FSC, SSC, FL-1, and FL-2 data were
acquired, and acquisition was terminated after 500 (eosinophils and
basophils) or 1000 (monocytes) target events. Shape change of
leukocytes was quantified as percentage of cells showing an increase in
FSC above baseline (for detailed description, see
Results).
|
Immunofluorescence flow cytometry
To investigate the expression of CCR1, CCR2, CCR3, and CCR4 on
leukocytes, PMNL or PBMC were suspended in staining buffer at 5 x
106 cells/ml. Samples were incubated on ice for
30 min with 10 µg/ml rabbit IgG to saturate Fc
receptors and then
with 10 µg/ml anti-CCR1 mAb LS-2D4, 10 µg/ml anti-CCR2 mAb
LS-132.1D9, 3 µg/ml anti-CCR3 mAb LS-7B11, 10 µg/ml
anti-CCR4 mAbs 1G1 or 2B10, or 0.6 µg/ml HLA-DR mAb.
Isotype-matched mouse IgG1 and IgG2a mAbs were used to determine
nonspecific binding. The cells were washed and resuspended in staining
buffer containing FITC-conjugated F(ab')2 goat
anti-mouse Ig on ice for 30 min and then washed again. To saturate
free binding sites of the polyclonal goat anti-mouse Ab, the cells
were incubated with a mixture of mouse IgG1 and IgG2a. Leukocytes were
then counterstained with R-PE-conjugated anti-CD16 (to distinguish
neutrophils from eosinophils in PMNL populations) or CDw123 (2 µg/ml,
to identify basophils in PBMC populations) for 20 min. After a final
wash, the samples were resuspended in fixative and held on ice until
analysis.
To allow for a direct comparison of receptor expression, PBMC and PMNL were analyzed using the same FL-1 setting. Neutrophils were discriminated from eosinophils by their high expression of CD16. Monocytes were identified and gated on a FSC/SSC plot. Basophils were distinguished by their high CDw123 staining. Contaminating dendritic cells did not interfere with the measurement because they either showed the same level of staining (CCR3 and control mAbs) or were totally separated from basophils such that two distinct populations could be observed on a FL-1/FL-2 plot. The characteristics of these populations were also confirmed by anti-HLA-DR staining (0.6 µg/ml). In all instances, receptor density was recorded on the FL-1 channel and represented as mean fluorescence intensity, and nonspecific binding was determined by the relevant control Abs.
Statistical analysis
Data are shown as means ± SEM. Statistical evaluation was performed using the Wilcoxon signed rank test or ANOVA and Bonferronis post test where appropriate.
| Results |
|---|
|
|
|---|
Fig. 1
A shows a fluorescence dot plot of PBMC stained
with anti-HLA-DR-FITC (FL-1) and anti-CDw123-PE (FL-2).
Basophils formed a well-separated population that stained highly
positive for CDw123 but negative for HLA-DR and accounted for 0.53%
of PBMC. Another, smaller population of CDw123 bright cells was also
present among the PBMC, but separated clearly from basophils due to
positive staining for HLA-DR. This population has previously been
identified as monocyte-derived dendritic cells (49). When
the CDw123+/HLA-DRcells were gated (R1) and displayed
selectively on a FSC vs SSC dot plot, they again formed a homogenous
population (Fig. 1
, B and C). Stimulation with
chemokine caused an increase in forward scatter of basophils (Fig. 1
C, 1 nM eotaxin), as compared with buffer (Fig. 1
B). We quantified this effect as percentage of cells in a
higher FSC region, defined as a region that contained only 20% of
cells with high FSC values in a control sample (Fig. 1
B).
When subsequent samples of PBMC from the same donor were exposed to a
chemokine, e.g., eotaxin, a concentration-dependent increase in the
number of basophils in the high forward scatter region could be
observed (Figs. 1
C and 3). At maximal stimulation, up to
80% of basophils were present in this region. Staining with the
anti-CDw123 mAb, directed against the
subunit of the IL-3
receptor, did not interfere with the IL-3 receptor functionally because
basophils labeled in this manner responded to pM concentrations of IL-3
(data not shown).
Comparison of chemokine receptor density on basophils, eosinophils, and monocytes
The expression levels of CCR1, CCR2, and CCR3 on leukocytes are
shown in Fig. 2
. Of these receptors, only
very low levels of CCR1, were found on neutrophils. Basophils and
eosinophils predominantly expressed CCR3, whereas CCR2 was dominant on
monocytes. CCR1 was expressed on all the cell types investigated, its
density being three times higher on monocytes than eosinophils and
basophils. The levels of CCR1 and CCR3 were similar in eosinophils and
basophils, but, importantly, basophils were also positive for CCR2,
which was absent on eosinophils. Despite the original cloning of the
CCR4 receptor from a basophilic cell line, we were unable to detect any
expression of this receptor on peripheral blood basophils using
specific anti-CCR4 mAbs (data not shown).
|
Basophil responsiveness to a range of chemokines was investigated
using the BGFS assay. The order of potency among chemokines was MCP-4
(peak at < 1 nM)
eotaxin-2 = eotaxin-3
eotaxin > MCP-1 > MIP-1
> RANTES = IL-8
(Figs. 3
, A and B).
In contrast, the CCR4 agonist MDC was completely ineffective up to 10
nM. Maximal responses were highest with eotaxin, eotaxin-2, eotaxin-3,
and RANTES, followed by MCP-4 and MCP-1 (Fig. 3
, A and
B). The efficacy of MIP-1
and IL-8 was only
50% that
of eotaxin (Fig. 3
B). Optimal responses to chemokines were
seen after 4 min of stimulation. When basophils were exposed to eotaxin
for longer times, the dose-response curves tended to become more bell
shaped with an apparently reduced maximum efficacy (Fig. 3
C).
|
Parallel experiments determined the responses of monocytes,
basophils and eosinophils. MCP-1 was most potent on monocytes (Fig. 4
A), showing a bell-shaped
dose-response curve with a threshold concentration below 0.1 nM and
responses that peaked at 1 nM. On basophils, MCP-1 was significantly
less potent, but, after its peak at 3 nM, maximal effectiveness was
maintained up to 100 nM (Fig. 4
A). In contrast, MCP-1 did
not elicit a response in eosinophils at 10 nM and below, but induced
marked shape change at 30100 nM, which corresponded to a 30- to
100-fold reduced sensitivity of eosinophils compared with basophils
(Fig. 4
A). MCP-2 was generally less potent than the
other MCPs at inducing shape change in all three cell types, but, of
the three, basophils were most sensitive to this chemokine (Fig. 4
B). MCP-3 induced marked shape change in all three cell
types, with eosinophils being only slightly less sensitive than
monocytes and basophils (Fig. 4
C).
|
Involvement of CCR2 and CCR3 in eosinophil and basophil responses
The dose-response curve to eotaxin was shifted to the right by a
factor of 100 for eosinophils and basophils in the presence of 16
µg/ml anti-CCR3 mAb LS-7B11, whereas the anti-CCR2 mAb
LS-132.1D9 (Fig. 5
, A and
B) and the respective control Abs (data not shown) were
without effect. Similarly, basophil responses to eotaxin-3 were
abolished in the presence of anti-CCR3 (20.7 ± 1.3%,
70.8 ± 1.8%, and 23.7 ± 3.7% basophils responding to
control, 10 nM eotaxin-3, and 10 nM eotaxin-3 plus anti-CCR3 mAb,
respectively; n = 4).
|
In basophils, responses to MCP-2 over the range 010 nM were not
reduced by anti-CCR2 mAb but were almost completely abolished by
anti-CCR3 mAb (Fig. 6
A).
|
Anti-CCR3 mAb inhibited responses to low concentrations of MCP-4 in
basophils, but was ineffective against high concentrations (Fig. 6
C). Conversely, anti-CCR2 mAb had a minimal inhibitory
effect on responses to low MCP-4 concentrations, but significantly
reduced the response to high concentrations of MCP-4 (Fig. 6
C). The combination of anti-CCR2 plus anti-CCR3 was
slightly more effective that any of the mAbs alone, but still could not
abolish responses to high concentrations of MCP-4.
Effect of CCR2 stimulation on CCR3 responsiveness
To investigate whether stimulation of CCR2 could enhance the
response of basophils to CCR3 agonists, basophils were stimulated with
increasing concentrations of eotaxin in the absence or presence of a
low concentration of MCP-1 (0.3 nM), with both chemokines being added
to the cells simultaneously. Stimulation with 0.3 nM MCP-1 caused a
small basophil response. Stimulation of basophils with 0.3 nM MCP-1 in
combination with increasing concentrations of eotaxin resulted in a
left-shift of the dose-response curve to eotaxin, such that the
threshold concentration and EC50 for eotaxin were
decreased by a factor of 3 compared with the response to eotaxin alone
(Fig. 7
). Maximal responses to eotaxin
were not affected by MCP-1, indicating that only sensitivity to, but
not effectiveness of, CCR3 activation was enhanced (Fig. 7
).
|
To investigate whether the patterns of responsiveness observed
above were directly due to agonist stimulation of basophils or
dependent upon indirect stimulation of basophils via other cell types
such as monocytes, experiments were undertaken using highly purified
basophils prepared by negative magnetic selection. Fig. 8
shows that purified basophils exhibited
similar dose-response relationships to the basophils in mixed cell
populations above, after stimulation with eotaxin, MCP-1, or MCP-3. In
two of these experiments, basophil responses to eotaxin, MCP-1, -2, -3,
-4, and MIP-1
were compared in purified and mixed cell populations
from the same donor and found to be identical (data not shown). These
highly purified basophils did not require labeling with anti-CDw123
and anti-HLA-DR, thus also demonstrating that these labeling steps
did not affect basophil responsiveness. Fig. 8
also shows that, in
purified basophil populations, similar patterns of chemokine receptor
usage by eotaxin (CCR3 only), MCP-1 (sequential dependence upon CCR2
and CCR3), and MCP-3 (dependence on more than one receptor) were
observed to those described above.
|
| Discussion |
|---|
|
|
|---|
Previously, we have used a flow cytometric assay of leukocyte shape
change (the GAFS assay) to measure responses of eosinophils,
neutrophils, and monocytes to chemokines in vitro (42).
This sensitive assay is capable of measuring chemokine responsiveness
of individual cell types in mixed cell populations. Basophils have been
historically difficult to study because of their low incidence in the
circulation. We therefore modified the GAFS assay to measure basophil
responses in mixed PBMC preparations, without the need for further
leukocyte purification. We identified basophils using an
anti-IL-3R
(CDw123) mAb (staining predominantly basophils and
monocyte-derived dendritic cells), in conjunction with an
anti-HLA-DR mAb (which labeled the dendritic cells and
distinguished them from basophils).
We found that the kinetics of basophil chemokine responsiveness were
similar to those of eosinophils, being detectable within 1 min and
reaching a maximum response by 4 min. After longer periods of
stimulation, the dose-response curves to eotaxin tended to become more
bell shaped with an apparent reduction in efficacy (Fig. 3
). Eosinophil
responses to MCP-4 are similarly time dependent, although, in contrast
to basophils, eosinophil responses to eotaxin are preserved over time
(42). Basophil shape change was induced by many
chemokines, with rank order of potency of MCP-4
eotaxin-2
= eotaxin-3
eotaxin > MCP-1 = MCP-3 >
MIP-1
> RANTES = MCP-2 = IL-8. Similar patterns of
chemokine responsiveness were observed in highly purified basophil
populations, demonstrating that these rapid responses to chemokines
were not dependent on other cell types (e.g., monocytes), nor were they
modulated by the anti-CDw123/HLA-DR labeling step, which was not
required for the purified cells.
The major human eosinophil chemokine receptor is CCR3, with lower levels of CCR1 expression (42, 51). As we show here, monocytes principally express CCR2, relatively high levels of CCR1, and no detectable CCR3. Basophils show a different pattern of CC chemokine receptor expression, with high levels of CCR3, intermediate levels of CCR2, and low levels of CCR1. The ability of the various chemokines tested to cause basophil shape change reflected these patterns of CCR expression. MCP-4, eotaxin, and eotaxin-2 were the most potent stimulators of basophil shape change. A recently described eotaxin functional homologue, eotaxin-3, also showed similar potency to eotaxin and eotaxin-2. Basophil shape change induced by eotaxin-3 was completely inhibited by a blocking anti-CCR3 Ab.
MCP-1 and MIP-1
were less potent inducers of basophil shape change
than eotaxin, in keeping with the lower levels of expression of their
principle receptors, CCR2 and CCR1, respectively.
IL-8 has been shown previously to cause basophil histamine release
through specific IL-8 receptors (52), and also caused
shape change in basophils. Both IL-8 and MIP-1
were less efficacious
than eotaxin, MCP-1, and MCP-4, causing responses in <50% of the
cells at the concentrations tested. The FSC changes of basophils in
response to chemokine stimulation had some appearances of being "all
or nothing," with discrete movement of increasing numbers of cells
into the higher FSC region upon exposure to increasing concentrations
of chemokine (Fig. 1
). Thus, the lower efficacy of MIP-1
may be
due to the lower level of basophil CCR1 expression (Fig. 2
), with some
cells not expressing sufficient levels of coupled receptors to be able
to respond to ligand. We have previously found eosinophil shape change
induced by chemokines to be pertussis toxin sensitive (I. Sabroe,
unpublished data). Other studies have shown basophil responses to
chemokines including RANTES, MCP-1, and IL-8 to be similarly sensitive
to pertussis toxin (22, 52, 53, 54).
MDC, whose receptor CCR4 was originally described in a basophil-derived cell line (55), failed to elicit any response, and we were unable to detect expression of CCR4 on peripheral blood basophils using specific mAbs. Curiously, at higher concentrations, MDC appeared to cause a reduction in the baseline numbers of basophils that were in the high FSC gate, an effect not seen with other chemokines. This may reflect basophil responses mediated by another MDC receptor, the existence of which has been speculated in other work (56).
The MCP group of chemokines, comprising MCP-1, -2, -3, and -4, all show varying selectivities for CCR2 and CCR3. We postulated that the differing expression patterns of these receptors on eosinophils (CCR3 but no CCR2), monocytes (CCR2 but no CCR3), and basophils (both receptors) may result in leukocyte-specific patterns of chemokine responsiveness. Our data indicated that basophils respond differently to MCPs, when compared with eosinophils and monocytes, and uncovered two patterns of integrated basophil chemokine responses.
Sequential receptor activation
MCP-1 induced a bell-shaped dose-response curve in monocytes (Fig. 4
A), suggesting desensitization of responses at higher
concentrations, and this signaling was completely blocked by an
anti-CCR2 mAb (data not shown). Eosinophils only responded to MCP-1
at high concentrations, in a CCR3-dependent manner (Fig. 5
C). Responses of eosinophils to high concentrations of
MCP-1 using assays of intracellular calcium flux were previously
observed when eotaxin was first described (57). Consistent
with this, one study showed that MCP-1 bound to CCR3, but that the
affinity of MCP-1 for CCR3 was much lower than that of eotaxin
(51). In both that study and work of other groups, CCR3
transfectants were not responsive to MCP-1 (51, 58),
suggesting subtle differences in the coupling of these receptors in
transfectants compared with native eosinophils. Basophils, both
purified and in mixed PBMC populations, showed responses to MCP-1
across a broad concentration range, with signaling being mediated
exclusively by CCR2 at low concentrations and predominantly by CCR3 at
high concentrations (Figs. 5
D and 8). These data suggest
that in the presence of high concentrations of MCP-1 in vivo, CCR2 may
be desensitized, and basophil and eosinophil recruitment may be favored
over monocyte recruitment.
Leukocyte responses to MCP-4 showed an inverse pattern. At low
concentrations, basophil responses to MCP-4, like eosinophil responses
(42), were exclusively mediated via CCR3 (Fig. 6
C), but, at high MCP-4 concentrations, eosinophils showed a
bell-shaped dose-response curve that was not seen in basophils (Fig. 4
D). In eosinophils, this bell-shaped dose-response curve of
MCP-4 signaling is different to the sigmoidal dose-response curve of
eotaxin and suggests different mechanisms of MCP-4 and eotaxin
signaling via CCR3 (42). The maintained responses of
basophils to higher concentrations of MCP-4 correlated with an
increasing monocyte response to MCP-4, likely to be mediated by CCR2.
The combination of anti-CCR3 and anti-CCR2 mAbs was unable to
completely block the responses of basophils to high concentrations of
MCP-4. These data are consistent with either the involvement of a third
receptor (e.g., CCR1) in MCP-4 signaling, or more likely reflect the
inability of the anti-CCR3 Ab to fully block the activity of a
high-affinity ligand at high concentration, consistent with the data
shown in Fig. 5
A. Thus, where high concentrations of MCP-4
are present in vivo, CCR3 may be desensitized and basophil and monocyte
recruitment may be favored over eosinophils. These data are consistent
with recent observations from Ying et al. (32), who
demonstrated that basophil recruitment in the late phase allergic
response in human skin, although not clearly correlated with expression
of a single chemokine, paralleled more closely MCP-4 kinetics. In
contrast, eosinophil recruitment occurred earlier (peaking at 6 h)
compared with basophils (peaking at 24 h), and correlated well
with eotaxin generation.
Both MCP-1 (Figs. 4
A and 5D) and MCP-4 (Figs. 4
D and 6C) appear to exhibit two optimal
concentrations for shape change of basophils, with an early peak
response at 1 nM, a small dip in response at 10 nM corresponding with a
switch in receptor usage, and a subsequent further response at 100 nM.
At these higher concentrations of chemokines, almost all of the
basophils have changed shape and thus further cooperation between CCR2
and CCR3 cannot be demonstrated. It is interesting to note that a
recent paper described two optimal concentrations of NAP-2 in
neutrophil chemotaxis, mediated by the sequential activation of CXCR2
and CXCR1 (59), presumably through regulation of receptor
desensitization and expression (46).
Cooperation of chemokine receptor signaling
MCP-2 proved to be a weak stimulator of leukocyte shape change,
but was a more potent stimulator of basophil than eosinophil responses
(Fig. 4
B). Blocking mAbs demonstrated that basophil
responses to MCP-2 were predominantly mediated by CCR3 (Fig. 6
A). However, an element of cooperative receptor signaling
between basophil CCR3 and another receptor not present on eosinophils
may explain the higher sensitivity of basophils to MCP-2. In keeping
with this, another study has shown that MCP-2 interacts with both the
receptors for RANTES and MCP-1 on basophils (23). To
investigate this further, we measured basophil CCR3-mediated responses
to eotaxin in the presence of low levels of CCR2 stimulation by MCP-1.
Basophil responses to eotaxin were significantly enhanced in the
presence of a low concentration of MCP-1, suggesting that cooperation
between CCR2 and CCR3 may be important in responses of basophils
in vivo.
An alternative pattern of receptor cooperativity was seen in basophil
responses to MCP-3. This chemokine showed potent stimulation both of
eosinophils and monocytes. As therefore anticipated, MCP-3 was a potent
stimulator of basophil shape change. Blockade of basophil CCR2 or CCR3
alone had no significant effect on the response to MCP-3, whereas both
mAbs in combination caused a significant reduction in basophil MCP-3
responses (Fig. 6
). The failure of these Abs to block MCP-3 responses
completely when employed together suggests that a component of basophil
MCP-3 responsiveness is mediated by a third receptor, most likely CCR1.
These data are in keeping with previous observations that basophil
calcium responses to MCP-3 could only be desensitized by RANTES and
MCP-1 in combination, but not by either of these chemokines singly
(24). However, our data suggest that CCR2 and CCR3 may
cooperate to induce basophil activation when stimulated together, so it
was surprising that no reduction in MCP-3 response occurred with these
mAbs used singly. Thus, these data may be explained by maximal
cooperativity being reached when any two of CCR1, 2, and 3 are
stimulated, but we were unable to test this hypothesis further because
the available anti-CCR1 mAb does not cause effective receptor
blockade in the shape change assay (42). Combined blockade
of CCR2 and CCR3 was less effective at inhibiting MCP-3-induced shape
change in purified basophil populations (Fig. 8
), which could be due to
the functional up-regulation of CCR1 during the additional handling
steps required to produce these highly purified cells. An alternative
explanation might rest in receptor heterodimerization. In opioid
receptors, heterodimerization of two different G protein-coupled
receptors results in altered patterns of ligand binding and receptor
signaling (60). The CCR2 G protein-coupled receptor forms
homodimers that regulate responsiveness to MCP-1 (61) and
thus it is possible that a component of the basophil response to MCP-3
is regulated through the formation of CC chemokine receptor
heterodimers.
Taken together, our data demonstrate differential responses to CC chemokines between basophils and eosinophils, regulated by induction of sequential or cooperative receptor signaling. Eosinophils dominate helminth responses (1) and allergen late phase responses (6, 32), whereas basophils dominate cutaneous basophil reactions (8, 11, 12, 13, 14, 15) and tick (3), as well as contact dermatitis responses (9, 10). Differential basophil signaling via sequential and cooperatively acting chemokines could be a mechanism explaining these in vivo differences.
| Acknowledgments |
|---|
| Footnotes |
|---|
2 Current address: Department of Experimental and Clinical Pharmacology, Karl-Franzens-University, Universitätsplatz 4, A-8010 Graz, Austria. ![]()
3 Current address: Section of Allergy and Clinical Immunology, Yale University School of Medicine, New Haven, CT 06510. ![]()
4 Address correspondence and reprint requests to Dr. Ian Sabroe, Leukocyte Biology Section, Biomedical Sciences Division, Imperial College School of Medicine, South Kensington, London SW7 2AZ, U.K. ![]()
5 Abbreviations used in this paper: MCP, monocyte chemotactic protein; MIP-1
, macrophage inflammatory protein-1
; PMNL, polymorphonuclear leukocytes (comprising eosinophils and neutrophils); BGFS, basophil-gated forward scatter assay; FSC, forward scatter; SSC, side scatter. ![]()
Received for publication May 9, 2000. Accepted for publication September 25, 2000.
| References |
|---|
|
|
|---|
pf, J. Norgauer, W. Czech. 1996. Recombinant human eotaxin induces oxygen radical production, Ca2+-mobilization, actin reorganization, and CD11b upregulation in human eosinophils via a pertussis toxin-sensitive heterotrimeric guanine nucleotide-binding protein. Blood 88:3195.
tze, J.-M. Schr
der, H.-D. Flad, E. Brandt. 1997. The CXC-chemokine neutrophil activating peptide-2 induces two distinct optima of neutrophil chemotaxis by differential interaction with interleukin-8 receptors CXCR-1 and CXCR-2. Blood 90:4588.This article has been cited by other articles:
![]() |
S. Main, R. Handy, J. Wilton, S. Smith, L. Williams, L. D. Fou, J. Andrews, L. A. Conroy, R. May, I. Anderson, et al. A Potent Human Anti-Eotaxin1 Antibody, CAT-213: Isolation by Phage Display and in Vitro and in Vivo Efficacy J. Pharmacol. Exp. Ther., December 1, 2006; 319(3): 1395 - 1404. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Mack, M. A. Schneider, C. Moll, J. Cihak, H. Bruhl, J. W. Ellwart, M. P. Hogarth, M. Stangassinger, and D. Schlondorff Identification of Antigen-Capturing Cells as Basophils J. Immunol., January 15, 2005; 174(2): 735 - 741. [Abstract] [Full Text] [PDF] |
||||