| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Immunology and Cell Biology, Forschungszentrum Borstel, Borstel, Germany
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Like all other chemotactic receptors cloned, CXCR-2 is a member of the large superfamily of G protein-coupled receptors (GPCR) that are structurally characterized by seven putative transmembrane segments connected by three intracellular and three extracellular loops and a free extracellular N-terminus. An important feature of all neutrophil GPCR including CXCR-2 is their rapid and reversible disappearance from the cell surface upon ligand binding (9, 10, 11). Using cell lines transfected with CXCR-2, it was demonstrated that this process is due to clathrin-mediated internalization of the receptor into endosomal compartments (12). Recently, heterologous stimuli such as LPS and TNF have also been shown to down-regulate CXCR-2 (13, 14). However, this process is much slower and is irreversible, involving the proteolytic degradation of surface-expressed CXCR-2 rather than its translocation into intracellular compartments (15).
Like many other GPCR the extracellular portion of CXCR-2 contains consensus sites for glycosylation at asparagine residues (N-glycosylation). These sites are located within the N-terminal part of the receptor at the asparagine in position 17 (Asn17) and within the second extracellular loop at Asn186 and Asn197. In general, N-linked oligosaccharides in cell surface receptors may influence their ligand binding, their mobilization to the cell surface, or their resistance to proteolysis (16, 17). However, the significance of glycosylation varies among different receptors, as seen with many GPCR that have been transfected into eukaryotic cell lines; while disruption of potential glycosylation sites did not change the surface expression of the receptors for V2 vasopressin (18) or parathyroid hormone (19), it strongly affected the expression of the calcium receptor (20), the vasointestinal peptide receptor (21), and the ß2-adrenergic receptor (22). Even more complicated, the structure and position of carbohydrate moieties in a given protein may vary with its expression in different tissues, which may change its subcellular distribution as well as its functional properties (16, 17).
In contrast to the above mentioned GPCR, no direct evidence for the glycosylation of chemokine receptors has been provided to date, and nothing is known about its potential implications for receptor physiology. In the present study we focus on neutrophils, the major CXCR-2-expressing cells, to investigate glycosylation of the native receptor. To accomplish this we first characterized the cellular location and glycosylation of CXCR-2 by means of its cross-linking to ligands, immunoprecipitation, and enzymatic removal of N-linked carbohydrates. A distinct, highly glycosylated CXCR-2 variant is expressed on the cell surface, while less glycosylated CXCR-2 variants form an intracellular receptor pool. By glycosidase treatment of surface-expressed receptors we show that full glycosylation is not necessary for signaling and trafficking of the receptor, but is required for its resistance to neutrophil proteases.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Human rIL-8 (72-residue isoform) was purchased from PeproTech
(Rocky Hill, NJ). Human NAP-2 was obtained by 
-chymotrypsin
digestion of homogeneous connective tissue-activating peptide III that
was purified from release supernatants of thrombin-stimulated platelets
using sequential immunoaffinity chromatography, cation exchange
chromatography, and reverse phase HPLC according to a described
protocol (7).
The murine mAbs against CXCR-1 (mAb SE-2) and CXCR-2 (mAb RII115) were described recently (7). They had been generated by immunization with carrier-coupled synthetic peptides (CXCR-1[1–30] and CXCR-2[2–24]) encompassing the amino acids at positions 1–30 in CXCR-1 and 2–24 in CXCR-2, respectively. Both Abs were of the IgG2b isotype.
For conjugation of biotin to mAb RII115 1 mg of purified Ab was incubated with 7.5 µg of biotin-X-NHS (Calbiochem, Bad Soden, Germany) in 1 ml of 0.1 M sodium carbonate buffer (pH 8.3) for 60 min at room temperature. Biotinylated Ab was then separated by exclusion chromatography on a Sephadex G-25 column (Pharmacia, Freiburg, Germany) equilibrated with PBS/0.02% thimerosal.
CXCR-1- and CXCR-2-transfected human embryonic kidney cells HEK 293 (23) as well as correspondingly transfected murine BALB 3T3 cells (24) were provided by A. Ben-Baruch (Laboratory of Molecular Immunoregulation, Frederick, MD) and I. I. Tikhonov (Research Institute of Immunology and Blood Transfusion, Minsk, Belarus), respectively.
Preparation and pretreatment of neutrophils
Human PMN were routinely isolated from citrated blood of single healthy donors by gradient centrifugation on Ficoll-Hypaque as previously described (25). Prepared neutrophils were suspended in Dulbecco’s PBS without Ca2+ and Mg2+ (PBS-D).
To investigate the functional role of glycosylation in surface-expressed CXCR-2, intact neutrophils were suspended in PBS-D (3 x 107 cells/ml) and treated with the indicated concentrations of N-glycosidase F (Roche, Mannheim, Germany) in the presence or the absence of PMSF (200 µM) or aprotinin (10 µg/ml) at 37°C for 60 min. Subsequently, cells were either cooled on ice for precipitation of CXCR-2 or subjected to functional studies as described below. Pretreatment of neutrophils did not alter their viability, which remained >95% in all events as seen by trypan blue exclusion.
Iodination of chemokines and cross-linking to neutrophils
Chemokines were radiolabeled by iodination at tyrosine residues using the chloramine-T method. Before iodination NAP-2 was chemically modified by the introduction of tyrosine residues as previously described (8). For cross-linking experiments PMN (2 x 107/ml) were incubated with 10 nM [125I]IL-8 or [125I]NAP-2 (sp. act., 18 and 11 MBq/nmol, respectively) in PBS-D/1% BSA for 60 min on ice. Cells were washed twice with ice-cold PBS-D to remove free ligands and were incubated with the cross-linker BS3 (Pierce, Rockford, IL) at a final concentration of 2 mM in PBS-D on ice. After 60 min the reaction was stopped by addition of glycine to a final concentration of 100 mM. Cells were washed twice with ice-cold PBS-D and were used for immunoprecipitation of CXCR-2.
Immunoprecipitation of CXCR-2 and deglycosylation of precipitates
For extraction and precipitation of CXCR-2, neutrophils were washed with ice-cold inhibitor buffer (one tablet of Complete (Roche) dissolved in 20 ml of PBS-D) and resuspended in the same buffer to a final concentration of 1 x 108 PMN/ml. Cells were lysed by addition of 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate (CHAPS) to a final concentration of 1%. After 10 min of incubation on ice, cells were vortex mixed for 45 s, and unsolubilized material was spun down at 12,000 x g for 10 min.
For precipitation of total CXCR-2-protein (intracellular and surface-expressed), 1 ml of clear supernatant was incubated with 5 µg of mAb RII115 for 60 min on ice followed by the addition of 35 µl of protein A-agarose solution (Roche). After 60 min of incubation on a roller mixer at 4°C, precipitates were washed four times with inhibitor buffer supplemented with 0.5% CHAPS.
For selective precipitation of intracellular receptors, PMN
(5x107/ml) were pretreated with 60 U/ml of
-chymotrypsin (Serva/Boehringer Ingelheim, Heidelberg, Germany) at
4°C. Subsequently, -chymotrypsin was inactivated by addition of
FCS to a final concentration of 10% and was removed by washing twice
with PBS-D. Successful removal of immunoreactive CXCR-2 from the cell
surface was verified by FACS analysis using mAb RII115. Precipitation
of the remaining CXCR-2 from the cell lysates was performed as
described above.
For selective precipitation of surface-expressed CXCR-2, mAb RII115 (5 µg/ml) was reacted to intact cells (5 x 107/ml in PBS-D/0.1% BSA) for 60 min on ice. After removal of free Ab by washing three times with PBS-D these cells were lysed, and precipitation was performed by addition of protein A-agarose as described above.
For deglycosylation of immunoprecipitated CXCR-2, precipitates (30 µl) were eluted from protein A-agarose by denaturation with 0.5% SDS/4% ME at 95°C. After 10 min of incubation, 90 µl of PBS-D (supplemented with 1.25% CHAPS and 1 mM EDTA, pH 6) were added, and protein A-agarose was spun down by centrifugation. The supernatant was then incubated with the indicated concentrations of N-glycosidase F or endoglycosidase H (both from Roche) at 37°C.
Electrophoresis and Western blotting
The Ab/receptor precipitates obtained with protein A were resuspended in sample buffer (containing SDS and DTT at final concentrations of 2 and 0.6%, respectively), incubated at 95°C for 10 min, and separated electrophoretically in a 10% SDS-PAGE gel. Rainbow Colored Protein Markers (Amersham, Buchler, Braunschweig, Germany) served as a molecular mass standard. Cross-linked receptor/[125I]ligand complexes were identified in the gel using a PhosphorImager (Molecular Dynamics, Krefeld, Germany). In experiments in which CXCR-2 had not been cross-linked, the receptor was transferred onto a polyvinylidene difluoride membrane (Immobilon-P, Millipore, Eschborn, Germany) and detected by Western blotting, using biotinylated mAb RII115 (2 µg/ml) that was finally visualized by means of peroxidase-conjugated streptavidin (Sigma, Munich, Germany) catalyzing a chemiluminescence reaction (BM Chemiluminescence Kit, Roche).
Flow cytometric analysis of CXCR expression on neutrophils
Neutrophils (3x106/ml, suspended in D-PBS/0.1% BSA) were incubated with 2 µg/ml of anti-CXCR-2 mAb RII115 or 5 µg/ml of anti-CXCR-1 mAb SE-2 for 1 h on ice, labeled with fluorescein-conjugated goat anti-mouse IgG (H-L) Ab (Dianova, Hamburg, Germany) at a final concentration of 15 µg/ml, and analyzed in a flow cytometer (Becton Dickinson, Heidelberg, Germany)
Degranulation
Neutrophil degranulation in response to NAP-2 was measured in terms of lysosomal ß-glucuronidase release from cytochalasin B (Cyt B)-pretreated PMN. Cells (1 x 107/ml) were incubated with 5 µg/ml Cyt B (Sigma, Deisenhofen, Germany) for 10 min at 37°C. After supplementation with 1.8 mM CaCl2 and 1 mM MgCl2, 100 µl of cell suspension was added to 100 µl of stimulus solution containing NAP-2 diluted in PBS-D/0.1% BSA. Following incubation for 30 min at 37°C under agitation, the supernatants were recovered by centrifugation. ß-Glucuronidase enzymatic activity in 50 µl of serially diluted supernatants was measured by addition of 50 µl of 0.1 M acetate buffer (pH 4) containing 7.5 mM p-nitrophenyl-ß-glucuronide as a substrate (Sigma, Deisenhofen, Germany). After incubation for 18 h and subsequent acidification by addition of 50 µl of 0.4 M glycine buffer (pH 10.4), hydrolyzed substrate was monitored photometrically at 405 nm. The amount of ß-glucuronidase was expressed as the percentage of total activity in neutrophil lysates obtained with 0.1% Triton X-100.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In a first set of experiments we aimed at characterizing the
molecular properties of CXCR-2 expressed on the neutrophil cell
surface. To selectively label surface-expressed receptors, intact
neutrophils were incubated with and cross-linked to
[125I]IL-8 or
[125I]NAP-2 at 4°C. After washing, detergent
extracts of the cells were prepared and, to exclude interference by
CXCR-1, mAb RII115 specific for CXCR-2 was used for immunoprecipitation
of the receptor. As reported previously, this Ab recognizes the amino
acid sequence FEDFW (positions 6–10) located within the receptor’s
N-terminus and binds to CXCR-2 even in the presence of bound ligand
(7). As seen in our present work, this circumstance also
allowed for immunoprecipitation of CXCR-2 following its cross-linking
to the radiolabeled chemokines. This is shown in Fig. 1
A, where the
immunoprecipitates were subjected to SDS-PAGE and then analyzed by
autoradiography. Regardless of whether
[125I]IL-8 or
[125I]NAP-2 had been used for a ligand, a
radiolabeled protein band migrating with a size of about 64 kDa was
detected. These signals were specific, because they could be completely
abrogated when incubation of cells with the radiolabeled chemokines was
performed in the presence of excess cold ligand (data not shown) or
when immunoprecipitation was performed in the presence of a synthetic
CXCR-2-derived peptide (peptide CXCR-2[2–24]) containing the epitope
required for binding of the Ab RII115 (Fig. 1A).
|
64 kDa) of the receptor-ligand complexes was much
larger than their theoretical size (48 kDa) calculated from the
known amino acid sequence of CXCR-2 (40 kDa) plus a bound chemokine
monomer (8 kDa). We thus examined whether this discrepancy could be
due to glycosylation of the receptor by digesting cross-linked and
immunoprecipitated CXCR-2/[125I]IL-8 complexes
for a constant period of 2 h with increasing concentrations of
N-glycosidase F, which cleaves all types of
asparagine-linked oligosaccharides directly from the protein backbone
(26). After exposure to a low dosage (0.6 U/ml) of the
enzyme, the apparent size of the complex underwent a shift from 64 to
about 55 kDa, while at a more elevated dosage (6 U/ml) a further
reduction to about 46 kDa was observed. The latter band remained
unchanged even when a 10-fold higher enzyme dosage was applied (Fig. 1
B). Because the rIL-8 used for cross-linking is not
glycosylated, these results indicate that neutrophil surface-expressed
CXCR-2 bears asparagine-linked carbohydrates, representing at least
32% of the receptor’s molecular mass. Furthermore, there is
evidence for N-glycosylation of the receptor at least at two
different sites, as indicated by the stepwise removal of two 9-kDa
components with increasing concentrations of
N-glycosidase F. Distinct molecular variants of CXCR-2 are expressed on the neutrophil cell surface and in intracellular compartments
In additional experiments we sought to determine whether resting
neutrophils, apart from expressing CXCR-2 at the cell surface, also
contained intracellularly located receptors. For this purpose cells
were lysed, and detergent extracts were prepared. Because cross-linking
of the receptor(s) to radiolabeled chemokines did not work with this
kind of starting material, we directly immunoprecipitated receptor(s)
from the detergent extract and used biotin-conjugated mAb RII115 in
combination with streptavidin-peroxidase for visualization of CXCR-2
molecules in Western blots. As shown in Fig. 2A, neutrophil extracts
contained at least three major immunoreactive proteins migrating at
about 56, 40, and 38 kDa, and a minor 44-kDa protein. The latter
protein, however, scored with variable intensity or was sometimes
absent in cells from different donors (compare to Figs. 2B
and 3). Recognition of these proteins by the precipitating as well as
the detecting Ab was specific, because no signal was obtained when
either precipitation or detection was conducted in the presence of an
excess of peptide CXCR-2[2–24] (Fig. 2A). To further
exclude that intracellular molecules not related to CXCR-2 but
cross-reactive to the Ab were present in total cell extracts, we also
examined cell lines that reportedly do not express CXCR-2, such as
human embryonic kidney cells (HEK 293) (23) and murine
BALB 3T3 cells (24). Although no immunoreactive proteins
could be precipitated from these cells, we obtained reactive bands with
extracts from the same cell lines that had been transfected with CXCR-2
(data not shown). Thus, our results to date indicated that neutrophil
extracts contain different molecular variants of the receptor, some of
which could be located intracellularly.
|
-chymotrypsin to destroy surface-expressed CXCR-2. Subsequently,
detergent extracts were prepared from the enzyme-treated cells, and the
remaining receptors were subjected to immunoprecipitation and Western
blot analysis. As shown in Fig. 2B, surface-expressed
receptors were represented by one protein corresponding to an apparent
size of about 56 kDa. This result is in good accordance with that
obtained above using the cross-linking method, where a single protein
of about 64 kDa representing highly glycosylated CXCR-2 loaded with a
chemokine monomer was detected (compare to Fig. 1). Subtracting the
size of 8 kDa for a chemokine would yield a molecular size of about 56
kDa for the unloaded receptor. Thus, within the limits of accuracy of
the separation method used, molecules of identical size had become
precipitated from the cell surface by either procedure.
On the other hand, following removal of surface-expressed CXCR-2 by
digestion with -chymotrypsin, precipitation of the remaining
receptors from cell extracts yielded two immunoreactive proteins of 40
and 38 kDa in Western blots, while absolutely no signal was obtained
for the 56-kDa variant (Fig. 2B). The former two proteins
scored with staining intensities practically identical with those of
their counterparts found in untreated control cells. This and their
complete absence from the cell surface (see paragraph above) strongly
indicated an exclusive intracellular location for the 40- and 38-kDa
receptor variants. Likewise, chymotrypsin-insensitive CXCR-2 variants
of 40 and 38 kDa were detected upon digestion of CXCR-2-transfected HEK
293 cells (data not shown); therefore, the occurrence of intracellular
CXCR-2 variants does not appear to be restricted to neutrophils.
Conversely, the total absence of the 56-kDa variant from precipitates
derived from -chymotrypsin-treated neutrophils, regardless of
whether the precipitating Ab was added before or after cell lysis (Fig. 2B), clearly suggests that this receptor variant is
exclusively expressed on the cell surface.
Neutrophil surface-expressed and the intracellular pool of CXCR-2 differ in the degree and type of N-glycosylation
We next examined whether the differences in molecular size
existing between surface-expressed and intracellular variants of
neutrophil CXCR-2 would be due to variable degrees of
N-glycosylation. For this purpose deglycosylation
experiments with immunoprecipitates containing the total receptor
population (surface-expressed and intracellular) as well as with
precipitates containing only surface-expressed receptors were
performed. To distinguish between different types of glycosylation two
species of glycosidases were employed, N-glycosidase F
(cleaving N-linked carbohydrates of the high mannose,
hybrid, or complex type from the protein backbone) and endoglycosidase
H (exclusively removing high mannose or hybrid-type
N-glycans). Upon treatment of immunoprecipitates with
N-glycosidase F and Western blotting, the different bands
representing the total receptor population as well as the 56-kDa band
from the surface precipitate disappeared, and in both samples a rather
broad protein band at about 36–40 kDa became visible (Fig. 3A). After shorter exposure to
the staining reagent this signal appeared as a single sharp band of 37
kDa (not shown). This finding indicates that the size heterogeneity in
CXCR-2 is due to differential glycosylation of an identical protein
core with carbohydrate proportions of about 34, 8, and 3% for the 56-,
40-, and 38-kDa variants, respectively. However, exposure of the
immunoprecipitates to endoglycosidase H only reduced the size of the
intracellular CXCR-2 variants (38 and 40 kDa), while the
surface-expressed 56-kDa protein remained unaffected in both
immunoprecipitates tested (Fig. 3B). These data suggest that
the intracellular receptor variants represent CXCR-2 glycoforms of the
high mannose or hybrid type, while the surface-expressed receptor
carries oligosaccharides further modified to complex structures.
|
N-glycosylation of the 56-kDa receptor is essential for the maintenance of its surface expression
The exclusive expression of only the most highly glycosylated
CXCR-2 variant on the neutrophil surface raised the question of whether
glycosylation would play a role in the receptor’s physiology. To
investigate this we treated intact neutrophils with increasing
concentrations of N-glycosidase F to deglycosylate the
surface-expressed receptor. After removal of the enzyme, Western blot
analysis of immunoprecipitates revealed that treatment with a moderate
dosage of the enzyme (10 U/ml) led to reduction of the receptor’s size
by about 9 kDa (from 56 to 47 kDa), indicating that one of its
two carbohydrate moieties had become removed (Fig. 4). However, the overall signal for the
deglycosylated receptor appeared also markedly decreased, and with a
higher dosage of N-glycosidase F (50 U/ml) the signal
completely disappeared. Since it was reported that neutrophil serine
proteases may down-regulate CXCR-2 expression by proteolytic
degradation (13, 14), we wondered whether the
disappearance of the deglycosylated receptor could be due to its
enhanced susceptibility to proteolysis. Indeed, when
N-glycosidase treatment of neutrophils was performed in the
presence of the serine hydrolase inhibitor PMSF, the signals for
deglycosylated CXCR-2 were much stronger (Fig. 4). Corresponding
results were obtained upon analysis of the receptor’s surface
expression by flow cytometry. Neutrophils exposed to 10 U/ml
N-glycosidase F for 2 h underwent a dramatic reduction
in CXCR-2 expression (from 250 down to 38 median fluorescence intensity
(MFI); Fig. 5), while in the presence of
PMSF (200 µM), aprotinin (10 µg/ml), or FCS (10%) only a moderate
decrease was observed (down to 212, 185, and 127 MFI, respectively;
data not shown). The expression of CXCR-1, which was recorded for a
control, remained unaffected however (Fig. 5). These data demonstrate
that full N-glycosylation of CXCR-2 is essential for the
maintenance of its surface expression by protecting the receptor from
proteolytic attack.
|
|
To investigate the impact of N-glycosylation on
receptor function we first examined whether partial deglycosylation of
surface-expressed CXCR-2 would affect its ligand-induced trafficking,
i.e., its internalization following binding to NAP-2 as well as its
reappearance after removal of the chemokine. To minimize proteolytic
degradation of the receptor during N-glycosidase F treatment
of the cells (which still occurs, although at a reduced level, even in
the presence of PMSF), we chose an incubation time of 1 h with 10
U/ml of the enzyme. These conditions were sufficient for conversion of
the 56-kDa surface molecule into its 47-kDa deglycosylated form, while
there was only a moderate decrease in the receptor’s overall signal
(compare corresponding lanes in Fig. 4). Moreover, as shown by flow
cytometry in Fig. 6, CXCR-2 surface
expression of glycosidase/PMSF-treated neutrophils was only slightly
reduced compared with that of both untreated as well as PMSF-exposed
control cells. Subsequent exposure to increasing concentrations of
NAP-2 for 10 min revealed very similar internalization kinetics for
CXCR-2 in glycosidase/PMSF-treated, untreated, and PMSF-exposed
neutrophils, as characterized by half-maximal down-regulation of
receptor surface expression at NAP-2 dosages of 84, 89, and 120 nM,
respectively (Fig. 6). These results demonstrate that complete
glycosylation of CXCR-2 is not required for NAP-2-induced receptor
down-regulation.
|
, PMSF-pretreated neutrophils, bearing
the 56-kDa receptor at the cell surface, exclusively re-expressed the
same highly glycosylated molecular species. Interestingly,
N-glycosidase F/PMSF-pretreated neutrophils, bearing the
47-kDa deglycosylated CXCR-2, did not re-express any 56-kDa receptors,
but exclusively expressed the 47-kDa variant. While these results
demonstrate that partial deglycosylation of surface CXCR-2 does not
compromise receptor internalization or re-expression, they moreover
indicate that surface receptors are recirculated instead of becoming
mobilized from intracellular stores.
|
Because trafficking is not necessarily a prerequisite for the
ability of surface receptors to transmit ligand-induced signals, we
also examined the impact of deglycosylation on CXCR-2-mediated
neutrophil activation. For this purpose N-glycosidase
F/PMSF-treated and unexposed control cells were tested for their
capacity to respond to increasing concentrations of NAP-2 in terms of
lysosomal degranulation. When neutrophils were treated with
N-glycosidase F at a dosage of 10 U/ml that down-regulated
CXCR-2 expression by 62% (compare to Fig. 6), their degranulation in
response to increasing dosages of NAP-2 was significantly reduced (Fig. 8). As expression of CXCR-1, which also
contributes to NAP-2-induced degranulation (8), was not
affected by treatment with N-glycosidase F (Fig. 5),
complete inhibition of the cellular response was not expectable. On the
other hand, in the presence of PMSF, treatment with
N-glycosidase F did not reduce NAP-2-induced degranulation.
Apparently, neutrophils under conditions causing partial
deglycosylation of CXCR-2 in the absence of proteolytic receptor
degradation show an unchanged response to CXCR-2 ligands. These results
demonstrate that full glycosylation of the surface-expressed CXCR-2 is
not required for receptor signaling. Instead, glycosylation indirectly
contributes to the cellular responsiveness toward CXCR-2 agonists by
preventing proteolytic down-regulation of the receptor.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
By cross-linking studies with radiolabeled IL-8 and NAP-2 and
subsequent immunoprecipitation of the receptor-ligand complexes, we
first obtained evidence that neutrophil surface-expressed CXCR-2 is
represented by a single molecular form, bearing N-linked
carbohydrates. The size of the receptor (56 kDa), as calculated by
subtraction of the Mr for a bound
chemokine monomer (8 kDa) from that of the receptor-ligand complex (64
kDa), was in good accordance with that determined for the unloaded
receptor (56 kDa) directly immunoprecipitated from the cell surface.
With either approach, digestion of the precipitated molecules with
N-glycosidase F indicated the presence of
N-glycosidically bound sugars in surface-expressed CXCR-2 to
a proportion slightly higher than 30% (18 kDa). Considerable
O-glycosylation is likely to be absent, because the removal
of N-glycans reduced the receptor’s apparent size to a
value (37 kDa) already slightly lower than its theoretical
Mr (40 kDa) calculated from its amino
acid sequence.
Three sites of potential N-glycosylation are present within the extracellular portion of CXCR-2 at the asparagine residues located at positions 17, 186, and 197 (5). In fact, our digestion experiments of IL-8-cross-linked and immunoprecipitated CXCR-2 with increasing concentrations of N-glycosidase F revealed that at least two of these sites are used, each carrying an N-linked sugar moiety of 9 kDa. Further support for multiple glycosylation in CXCR-2 was also provided by digestion experiments performed on intact neutrophils in the absence of detergent. The decrease in molecular size of the receptor by only 9 kDa even upon exposure of the cells to high dosages of N-glycosidase F indicated that under these conditions only one of the two N-glycosyl residues had become removed. This phenomenon is probably due to differing accessibilities of the two glycosylation sites for the enzyme, which may be a result of their different locations within the receptor. As found with other GPCR, such as the ß2-adrenergic receptor (22) and the vasointestinal peptide receptor (21), membrane-proximal putative glycosylation sites are often not used, which is thought to be due to sterical reasons (32). Regarding the three consensus sites in CXCR-2, it is therefore most probable that Asn186, being spaced from the adjacent transmembrane domain by only seven amino acid residues, is not occupied by a carbohydrate moiety. In addition, glycosylation at Asn186 appears the more unlikely because of the presence of a proline residue in position 189; this kind of amino acid is generally thought to prevent the attachment of sugars when directly preceding or following the glycosylation sequon (33). Hence, Asn17 and Asn197, both located more distal from the membrane, are the most probable candidates to carry the two 9-kDa oligosaccharides in cell membrane-expressed CXCR-2.
Unexpectedly, our studies also revealed the existence of an intracellular receptor pool in neutrophils. These molecules could be clearly discriminated from the surface-expressed receptor by their molecular characteristics. In fact, following removal of surface-expressed receptors by treatment of intact neutrophils with chymotrypsin, two immunoreactive proteins with apparent sizes of 38 and 40 kDa remained detectable in the precipitates. Our initial reservations that these molecules might represent CXCR-2-unrelated intracellular proteins could be ruled out by two observations. On the one hand, there was no cross-reactivity of the precipitating Ab with proteins from CXCR-2-negative human cell lines, while the same cell lines transfected with the receptor scored positively. On the other hand N-glycosidase F treatment of the different proteins immunoprecipitated from neutrophils abrogated their size heterogeneity by yielding a single 37-kDa molecule, which suggested an identical protein core for the intracellular and the surface-expressed receptors. While these results gave evidence for a lower degree of N-glycosylation in the intracellular CXCR-2 variants compared with the surface-expressed molecule, additional experiments moreover revealed differences in the types of the carbohydrate chains attached. As specified by their sensitivity to cleavage by endoglycosidase H, the intracellular receptor proteins carry oligosaccharides of high mannose or hybrid type, while surface-expressed CXCR-2 carries oligosaccharides of complex structure, which are insensitive to endoglycosidase H. The exclusive intracellular location of mannose/hybrid-type receptor glycoforms, as opposed to the selective surface expression of complex-type glycoforms, suggests that CXCR-2 has to undergo glycosidic modification before it becomes expressed on the cell membrane. Corresponding conditions have been observed with other GPCR, as exemplified by the calcium receptor, which upon transfection into HEK 293 cells was found to occur in various glycoforms differing in susceptibility to endoglycosidase H as well as in subcellular distribution (20, 34). The importance of glycosylation for receptor mobilization was demonstrated by mutagenesis of three of eight glycosylation sites, which led to reduced surface expression and increased intracellular accumulation of the calcium receptor (20). A role for intracellular GPCR other than representing immature glycoforms has been proposed for the fMLP receptor. In resting neutrophils a considerable proportion of this receptor was reported to be located within secretory granules and to become rapidly mobilized to the cell surface upon stimulation with fMLP or platelet-activating factor (35). Although this demonstrates that intracellular pools may also serve as a reservoir readily available for the up-regulation of surface receptors, a corresponding role for intracellular CXCR-2 appears unlikely. In fact, none of various neutrophil-directed stimuli tested (e.g., NAP-2, fMLP, TNF, and G-CSF) induced enhancement of CXCR-2 surface expression, and more importantly, we did not find translocation of intracellular CXCR-2 glycoforms to the surface of NAP-2-stimulated cells.
Instead, we obtained evidence that full glycosylation, i.e., the presence of both carbohydrate moieties in CXCR-2, is relevant for the maintenance of receptor expression on the neutrophil surface. This was seen in deglycosylation experiments performed with intact neutrophils, leading to removal of only one of the carbohydrate chains (see above). FACS analyses as well as immunoprecipitation of CXCR-2 from N-glycosidase F-treated cells revealed that partial deglycosylation was associated with the disappearance of the receptor from the cell surface. Down-regulation of CXCR-2 involved proteolytic activity of serine proteases, as it could be completely prevented when incubation of neutrophils with N-glycosidase F was performed in the presence of serine protease inhibitors such as PMSF or aprotinin. Thus, full glycosylation of CXCR-2 is obviously necessary for protection of the receptor against neutrophil-derived proteases. This appears to be different with CXCR-1, which exhibited unchanged surface expression upon N-glycosidase F treatment of neutrophils. Although it is not known whether CXCR-1 is glycosylated at all or whether potential glycosylation would be susceptible to enzymatic cleavage, this observation suggests that surface expression of this receptor is stabilized by mechanisms different from those existing in CXCR-2.
An impact of carbohydrate components on ligand binding or signaling has been observed with only a few GPCR, while for most GPCR studied glycosylation turned out not to influence these receptor functions (18, 19, 20). Although we found treatment of neutrophils with N-glycosidase F to reduce degranulation in these cells in response to the CXCR-2 agonist NAP-2, this did not indicate that signaling through the receptor depended on the presence of glycosyl residues. In fact, as seen under conditions where proteolytic degradation of CXCR-2 was prevented, partial deglycosylation of the receptor by one of the two 9-kDa moieties did not change the neutrophil degranulation response. Corresponding findings were also obtained from our studies addressing NAP-2-induced trafficking of the receptor. Upon protection from proteolytic attack, partial deglycosylation of CXCR-2 did not change its internalization kinetics in response to NAP-2, nor was there an impact on the degree of receptor redistribution to the cell surface during recovery. Most interestingly, in glycosidase-treated and nontreated cells, reappearing receptors were of the same nature as those found to be surface expressed before ligand-induced down-regulation. Thus, the redistribution of only the fully glycosylated CXCR-2-variant in nontreated cells as opposed to the reappearance of only the partially deglycosylated 47-kDa variant in glycosidase-treated cells provides evidence for the recirculation and reusage of cell surface receptors following NAP-2-induced down-regulation. In conclusion, we have shown by several lines of evidence that full glycosylation in CXCR-2 has no direct impact on the receptor’s functional activity. However, as high glycosylation effectively prevents proteolytic degradation of the receptor, it is required for maintaining neutrophil responsiveness to CXC-chemokines such as NAP-2.
In general, protection of CXCR-2 against proteolytic degradation could be relevant in situations of acute inflammation, where proteases are known to become released from a variety of immune cells as well as from the injured tissue. However, protection through carbohydrate moieties does not appear to be effective upon stimulation of neutrophils themselves by proinflammatory mediators such as TNF or LPS. As reported recently, the latter mediators induce shedding of CXCR-2 by proteases associated with the receptor-bearing cells, and it has been suggested that this event represents a regulatory mechanism for the attenuation of cellular responsiveness to chemokines (15). In context with our observations, receptor glycosylation appears sufficient to shield CXCR-2 from degradation by the presumably low amounts of proteases associated with nonstimulated neutrophils, whereas protection can obviously be overcome by enhancement of proteolytic activity in stimulated cells. Thus, although not directly involved in ligand binding, glycosylation in CXCR-2 may nevertheless represent a regulatory element in the modulation of neutrophil responsiveness to CXC-chemokines.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Andreas Ludwig, Department of Immunology and Cell Biology, Forschungszentrum Borstel, Parkallee 22, D-23845 Borstel, Germany.
3 Abbreviations used in this manuscript: PMN, polymorphonuclear neutrophil granulocytes; NAP-2, neutrophil-activating peptide 2; CXCR-1, CXC-chemokine receptor type 1; CXCR-2, CXC-chemokine receptor type 2; GPCR, G protein-coupled receptor; PBS-D, Dulbecco’s PBS; Cyt B, cytochalasin B; CHAPS, 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate.
Received for publication January 27, 2000. Accepted for publication April 28, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, GROß, GRO
, neutrophil-activating peptide-2, and epithelial cell-derived neutrophil-activating peptide-78 are potent agonists for the type B, but not the type A, human interleukin-8 receptor. J. Biol. Chem. 271:20545.. J. Immunol. 160:4518.-mediated regulation of CXCR1 and CXCR2 chemokine receptor expression. Blood 93:2173. and neutrophil-activating peptide 2 to interleukin 8 receptors on human neutrophils. Proc. Natl. Acad. Sci. USA 89:10542.This article has been cited by other articles:
|
|
 |
|
  J. Bestebroer, K. P. M. van Kessel, H. Azouagh, A. M. Walenkamp, I. G. J. Boer, R. A. Romijn, J. A. G. van Strijp, and C. J. C. de Haas Staphylococcal SSL5 inhibits leukocyte activation by chemokines and anaphylatoxins Blood, January 8, 2009; 113(2): 328 - 337. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Frommhold, A. Ludwig, M. G. Bixel, A. Zarbock, I. Babushkina, M. Weissinger, S. Cauwenberghs, L. G. Ellies, J. D. Marth, A. G. Beck-Sickinger, et al. Sialyltransferase ST3Gal-IV controls CXCR2-mediated firm leukocyte arrest during inflammation J. Exp. Med., June 9, 2008; 205(6): 1435 - 1446. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Z. B. Levashova, N. Sharma, O. A. Timofeeva, J. S. Dome, and A. O. Perantoni ELR+-CXC Chemokines and Their Receptors in Early Metanephric Development J. Am. Soc. Nephrol., August 1, 2007; 18(8): 2359 - 2370. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. W. Nasser, S. K. Raghuwanshi, K. M. Malloy, P. Gangavarapu, J.-Y. Shim, K. Rajarathnam, and R. M. Richardson CXCR1 and CXCR2 Activation and Regulation: ROLE OF ASPARTATE 199 OF THE SECOND EXTRACELLULAR LOOP OF CXCR2 IN CXCL8-MEDIATED RAPID RECEPTOR INTERNALIZATION J. Biol. Chem., March 2, 2007; 282(9): 6906 - 6915. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. N. Moreno, G. Pereira-da-Silva, C. Oliver, M. C. Jamur, A. Panunto-Castelo, and M. C. Roque-Barreira The Macrophage-derived Lectin, MNCF, Activates Neutrophil Migration through a Pertussis Toxin-sensitive Pathway J. Histochem. Cytochem., June 1, 2005; 53(6): 715 - 723. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Murphy, M. McGurk, J. Pettigrew, A. Santinelli, R. Mazzucchelli, P. G. Johnston, R. Montironi, and D. J.J. Waugh Nonapical and Cytoplasmic Expression of Interleukin-8, CXCR1, and CXCR2 Correlates with Cell Proliferation and Microvessel Density in Prostate Cancer Clin. Cancer Res., June 1, 2005; 11(11): 4117 - 4127. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. E. de Salamanca, K. F. Siemasko, Y. Diebold, M. Calonge, J. Gao, M. Juarez-Campo, and M. E. Stern Expression of Muscarinic and Adrenergic Receptors in Normal Human Conjunctival Epithelium Invest. Ophthalmol. Vis. Sci., February 1, 2005; 46(2): 504 - 513. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. J. Paulsen, M. M. Rosenkilde, J. Eugen-Olsen, and T. N. Kledal Epstein-Barr Virus-Encoded BILF1 Is a Constitutively Active G Protein-Coupled Receptor J. Virol., January 1, 2005; 79(1): 536 - 546. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Hiruma, A. Togayachi, K. Okamura, T. Sato, N. Kikuchi, Y.-D. Kwon, A. Nakamura, K. Fujimura, M. Gotoh, K. Tachibana, et al. A Novel Human {beta}1,3-N-Acetylgalactosaminyltransferase That Synthesizes a Unique Carbohydrate Structure, GalNAc{beta}1-3GlcNAc J. Biol. Chem., April 2, 2004; 279(14): 14087 - 14095. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y.-S. Bae, J. C. Park, R. He, R. D. Ye, J.-Y. Kwak, P.-G. Suh, and S. Ho Ryu Differential Signaling of Formyl Peptide Receptor-Like 1 by Trp-Lys-Tyr-Met-Val-Met-CONH2 or Lipoxin A4 in Human Neutrophils Mol. Pharmacol., September 1, 2003; 64(3): 721 - 730. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Bannert, S. Craig, M. Farzan, D. Sogah, N. V. Santo, H. Choe, and J. Sodroski Sialylated O-Glycans and Sulfated Tyrosines in the NH2-Terminal Domain of CC Chemokine Receptor 5 Contribute to High Affinity Binding of Chemokines J. Exp. Med., December 3, 2001; 194(11): 1661 - 1674. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) or the absence (
) of 200 µM PMSF or were incubated without
enzyme in the presence (
) or the absence (•) of PMSF at 37°C for
60 min. Subsequently, cells were stimulated with various concentrations
of NAP-2 or were left unexposed for 10 min. Thereafter, the surface
expression of CXCR-2 was examined using mAb RII115 and was recorded by
flow cytometry as median fluorescence intensity after staining with
fluorescein-conjugated goat anti-mouse IgG. Unspecific
fluorescence, as determined by means of an IgG2b isotype control, was
subtracted (the proportion of unspecific binding is depicted in Fig. 5
