(Received for publication, July 16, 1996, and in revised form, October 9, 1996)
From the Gladstone Institute of Cardiovascular Disease,
¶ Department of Medicine, and the Daiichi Research
Center, University of California,
San Francisco, California 94141-9100
Two forms of the monocyte chemoattractant protein-1 receptors (the type A monocyte chemoattractant protein 1 (MCP-1) receptor CCR-2A and the type B MCP-1 receptor (CCR-2B) have been recently cloned and found to differ only in their terminal carboxyl tails. Here, we report that the two isoforms are alternatively spliced variants of a single MCP-1 receptor gene. Sequencing of the gene revealed that the 47-amino acid carboxyl tail of CCR2B was located in the same exon as the seven transmembrane domains of the receptor, and the 61-amino acid tail of CCR2A was in a downstream exon. Examination of freshly isolated human monocytes by reverse transcriptase-polymerase chain reaction revealed that CCR2B was the predominant isoform and that message levels of both CCR2A and CCR2B decreased as the monocytes differentiated into macrophages. In stably transfected cell lines, CCR2B trafficked well to the cell surface, but CCR2A was found predominately in the cytoplasm. Equilibrium binding studies revealed that those CCR2A receptors that successfully trafficked to the cell surface bound MCP-1 with high affinity (Kd = 310 pM), similar to CCR2B. In signaling studies, both CCR2A and CCR2B mediated agonist-dependent calcium mobilization, as well as inhibition of adenylyl cyclase. Creation of chimeras between CCR2A and the human thrombin receptor revealed that the cytoplasmic retention of CCR2A was due to its terminal carboxyl tail. Progressive truncation of the carboxyl tail indicated that a cytoplasmic retention signal(s) was located between residues 316 and 349. These data indicate that the alternatively spliced form of the human MCP-1 receptor (CCR2A) binds MCP-1 with high affinity and is a functional receptor and that expression at the cell surface is controlled by amino acid sequences located in the terminal carboxyl tail.
Chemokines are small (8-10 kDa), secreted basic peptides that are
involved in the directed migration and activation of leukocytes (see
Refs. 1, 2, 3, 4 for recent reviews). They can be divided into two groups
based on the arrangement of the first two of four conserved cysteines.
The - or "C-X-C" branch is characterized by the
presence of a single amino acid between the first two cysteines and
includes interleukin-8 (IL-8),1 GRO (
,
, and
), NAP-2, and platelet factor 4. In the
- or "C-C"
branch, the first two cysteines are adjacent. Members of the
-branch
include monocyte chemoattractant protein 1, 2, and 3 (MCP-1, MCP-2, and
MCP-3), RANTES (
egulated on
ctivation,
ormal
cell
xpressed and
ecreted), macrophage inflammatory proteins 1
and 1
(MIP-1
and MIP-1
), and eotaxin. In general, C-X-C
chemokines are chemotactic for neutrophils, whereas the C-C chemokines
are chemoattractants for monocytes and lymphocytes. The recently
identified peptide eotaxin is specific for eosinophils (5). A novel
peptide, lymphotactin, which lacks the first and third cysteines, has
been isolated from progenitor T cells (6) and may represent the first
member of a third branch of the chemokine family.
We recently reported the cloning of two MCP-1 receptors, type A (CCR2A)
and type B (CCR2B), that differed only in their terminal carboxyl
tails, suggesting that they were derived from a single gene via
alternative splicing (7). Inspection of the amino acid sequence of the
type B carboxyl tail revealed a 32% identity with both the
MIP-1/RANTES receptor (CCR1 (8)) and the eotaxin receptor (CCR3 (9,
10)), as well as a 21% identity with CCR4 (11). In contrast, this
region of CCR2A bore little resemblance to other chemokine receptors.
In functional studies, both forms of the MCP-1 receptor conferred
robust, agonist-dependent calcium mobilization in
Xenopus oocytes (7). In human embryonic kidney (HEK)-293
cells, the type B receptor was shown to inhibit adenylyl cyclase and
rapidly mobilize intracellular calcium via coupling to
G
i (12). Signaling studies of the type A receptor,
however, were hampered by its relatively poor surface expression in
transfected cells.
To further elucidate the biology of CCR2A, we have cloned and sequenced
the human MCP-1 receptor gene. We have examined the relative
expression of CCR2A and CCR2B in monocytic cell lines and in human
monocyte/macrophages and have established that the signal for
cytoplasmic retention of CCR2A is found in its carboxyl tail.
Finally, we report that type A receptors that successfully traffic to
the cell surface bind MCP-1 with high affinity and couple to
Gi to mediate signal transduction.
MCP-1 was obtained from R&D Systems, Inc. (Minneapolis, MN). Indo-1 AM was purchased from Molecular Probes, Inc. (Eugene, OR). Lipofectamine, G418, and minimal essential medium with Earle's balanced salt solution were obtained from Life Technologies, Inc. M1 and M2 monoclonal antibodies to the Flag epitope were purchased from Kodak. Fluorescein isothiocyanate (FITC)-conjugated goat anti-mouse antibody was purchased from Calbiochem. Flow cytometry grade FITC-conjugated goat anti-mouse antibody was from Zymed (South San Francisco, CA). Ficoll/Hypaque was from Pharmacia Biotech Inc. Horseradish peroxidase-conjugated goat anti-mouse antibody was obtained from Bio-Rad. Epitope-tagged thrombin receptor (13) was generously provided by Dr. Shaun Coughlin (University of California, San Francisco).
Isolation of Genomic ClonesA Lambda FIX II human genomic
library (Stratagene, San Diego, CA) was screened with a full-length
CCR2B cDNA probe under conditions of high stringency. A total of
1.3 million plaques were screened from which four independent clones
were isolated. Hybridizations were carried out in 50 ml of 50%
formamide, 5 × SSC, 0.1% Denhardt's, 0.1 mg/ml salmon sperm
DNA, and approximately 5 × 106 cpm/ml radiolabeled
probe at 42 °C for 14-20 h. The cDNA probe was labeled with
[-32P]dCTP by the random priming method (Prime-It kit,
Stratagene) to a specific activity of 108-109
cpm/µg. Filters were prehybridized in the same solution at 42 °C for 1-2 h before addition of the probe. Membranes were washed by
rinsing with 100 ml of an 0.2 × SSC, 0.1% SDS solution at room temperature for 30 min, followed by washing at 65 °C for 2 h. The positively hybridizing plaques were purified. From one clone (
FixII-A2), overlapping EcoRI and PstI
fragments were subcloned into pBSSK and sequenced to generate a map of
the gene. Restriction enzyme digests and sequence analysis of a second
clone (
FixII-17A), as well as Southern hybridization analysis of all
four genomic clone isolates, were performed to confirm our map of the
MCP-1 receptor gene. The MCP-1 receptor genomic sequence has been
deposited in GenBank (accession numbers U80923[GenBank] and U80924[GenBank]).
The sequence encoding the "Flag"
epitope (DYKDDDD) preceded by the prolactin signal sequence
(MDSKGSSQKGSRLLLLLVVSNLLLCQGVVS) (13) was ligated onto the 5 end
of the CCR2A and CCR2B cDNAs. These modified cDNAs were
subsequently subcloned into the mammalian expression vector pcDNA3
(InVitrogen, San Diego, CA). MCP-1RA receptor mutants with the
cytoplasmic tail terminating at either position 316 or 349 were made
using standard PCR methods. Leucine to alanine point mutants were made
as described (14). Thrombin/MCP-1 receptor chimeras were constructed
using a two-step overlapping PCR mutagenesis procedure (15). The
resulting PCR product encoded a chimeric receptor in which the first
371 amino acids were derived from the thrombin receptor (14), and the
carboxyl terminus was derived from the end of the seventh transmembrane
domain and carboxyl tail of either CCR2A or MCP-1R to give a
thrombin/MCP-1 receptor junction of ... IDPLIY/AFVGEK... .
All constructs made by PCR were completely sequenced.
Buffy coats were purchased from Irwin Memorial Blood Bank (San Francisco, CA). Human monocytes and macrophages were isolated and prepared as described (16). Cells were washed several times to remove the Percoll, and monocytes were quick frozen. The cells were greater than 80% monocytes, as determined by neutral red uptake. To obtain macrophages, monocytes were cultured in RPMI 1640 with 10% autologous serum for the indicated number of days.
mRNA Isolation and RT-PCRmRNAs from monocytes,
macrophages, and cultured cell lines were isolated using Trizol reagent
(Life Technologies, Inc.) following the manufacturer's instructions.
Total RNA was treated with RNase-free DNase (Boehringer Mannheim) at
0.1 units/µg total RNA at 37 °C for 15 min. RT-PCR was performed
using a GeneAmp RNA PCR kit (Perkin-Elmer), according to the
manufacturer's instructions, and 1 µg of DNase-treated RNA. PCR
conditions for first strand synthesis were 42 °C for 15 min,
99 °C for 5 min, and 5 °C for 5 min using the Perkin-Elmer GeneAmp PCR system 9600. Second step PCR conditions were incubation at
95 °C for 105 s, 95 °C for 15 s, and 60 °C for
30 s. After 34 cycles, the reaction was terminated by heating to
72 °C for 7 min. The resulting products were analyzed on a 2%
agarose gel. For detection of CCR2A and CCR2B, a common 5 primer was
used in conjunction with two carboxyl tail-specific 3
-antisense
primers. In control experiments using CCR2A and CCR2B plasmid DNA, we
identified PCR primer pairs that amplified both receptors equally.
HEK-293 cells were grown in minimal essential medium with Earle's balanced salt solution as described (12). COS-7 cells were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum (Hyclone Labs, Logan, UT) and 1% penicillin/streptomycin, at 37 °C in a humidified 5% CO2 incubator. Both transiently and stably transfected cell lines were created using Lipofectamine according to the manufacturer's instructions. Transiently transfected cells were assayed 2 days after the introduction of plasmid DNA. Stable cell lines were obtained by growing transfected cells at low density for 2 days, followed by the addition of G418 (800 µg/ml). Surviving colonies were assayed by Northern blot, and clones expressing high levels of receptor RNA were expanded for further studies.
Flow CytometryCells were detached using 1 mM EDTA, pelleted by centrifugation, and incubated with the M1 antibody to the Flag epitope in minimal essential medium/Earle's balanced salt solution at 1:1000 dilution at room temperature for 1 h, with shaking. Following incubation, cells were washed three times with PBS and resuspended in medium containing FITC-conjugated goat anti-mouse antibody (Zymed) in the dark for 30 min at room temperature with shaking. Cells were washed three times, resuspended in PBS with propidium iodide (20 µg/ml), and analyzed for receptor expression by flow cytometry.
Immunofluorescence MicroscopyCells were grown in six-well dishes (35 mm) and were used when 30-50% confluent. After removing the medium, the cells were washed once with PBS, fixed in 4% paraformaldehyde (in PBS) for 15 min at room temperature, and washed twice for 5 min each in Quench solution (150 mM sodium acetate and 0.1% nonfat dry milk in PBS). To permeabilize cells, lysis buffer (Quench solution plus 0.5% Triton X-100) was added for 15 min. Cells were washed three times, 5 min each, in wash solution (15 mM sodium acetate, 0.1% nonfat milk) followed by incubation in primary antibody. Monoclonal antibody M1 was added at 1:1000 dilution (in wash solution) incubated for 1 h at room temperature, and the cells were washed three times, 5 min each, and incubated in the dark with FITC-conjugated goat anti-mouse antibody at 1:1250 dilution for 40 min. Following the incubation, the cells were washed three times for 5 min each, rinsed once in PBS, and mounted in Vectorshield (Vector, Burlingame, CA) with coverslips. Cells were viewed using a Bio-Rad MRC-600 Laser Scanning Confocal Imaging System.
Inositol Phosphate Formation AssayTotal inositol phosphate was measured as described previously (17).
Binding, Calcium Fluorimetry and Adenylyl Cyclase AssaysThese assays were performed as described previously (12).
Quantitation of Cell Surface Receptors by ELISATo compare
surface expression of receptors in transiently transfected cells,
ELISAs were done 2 days after transfection. Transfections were carried
out in 24-well dishes, and each construct was examined in triplicate.
For stable lines, 2 × 105 cells/well were plated in a
24-well dish 1 day before the assay. Cells were fixed in 4%
paraformaldehyde (in PBS) for 10 min on ice, washed twice in PBS, and
incubated with M1 monoclonal antibody for 1 h at room temperature,
as above. Cells were then rinsed once with PBS and incubated 30 min
with a horseradish peroxidase-conjugated goat anti-mouse antibody
(second antibody). The cells were washed three times with PBS.
Developing solution (1 mg/ml
2,2-azinobis(3-ethylbenzthiazolinesulfonic acid), 0.1 M
sodium citrate, 0.1 M sodium phosphate, pH 4.0, 1 µl/ml
hydrogen peroxide) was added, and color development was read at 450 nm
using a Vmax (Molecular Devices, Palo Alto, CA) microtiter plate reader.
We have previously
identified two MCP-1 receptor cDNAs (CCR2A and CCR2B) that differed
only in the predicted amino acid sequence of their terminal carboxyl
tails (7). To test the hypothesis that these cDNAs represented
alternatively spliced variants of a single gene and to elucidate
potential splicing mechanisms, we cloned the gene for the human
MCP-1 receptor. Four independent genomic clones were identified by
hybridization with the CCR2B cDNA probe. Overlapping
EcoRI and PstI fragments from one clone, FixII-A2, were subcloned into pBSSK and sequenced. Both
isoforms of the receptor were encoded in three exons spanning
approximately 7 kb of genomic sequence (Fig. 1).
The 5-untranslated region (UTR) of both isoforms is interrupted by a
long intron (~3 kb) followed by a long exon (exon 2) of 1.9 kb that
codes for the remaining 1.9-kb sequence of CCR2B mRNA. Exon 3 lies
260 bp downstream of the end of exon 2 and codes for the
carboxyl-terminal tail and the entire 3
-UTR of the CCR2A isoform.
Therefore, both forms of the receptor have identical 5
ends composed
of exon 1 and the 5
-half (981 bp) of exon 2. The remaining 3
-half of
exon 2 is absent from the CCR2A cDNA, giving rise to unique
carboxyl tails (and 3
-UTR) in CCR2A versus CCR2B (Fig. 1,
bottom). Strong consensus sequences for canonical 5
donor
and 3
acceptor splice sites (18) were found at all proposed
intron/exon splice junctions, including the two required to synthesize
CCR2A (Fig. 1, bottom). Consensus AATAAA polyadenylation signal sequences were found ~20-45 bp upstream from the 3
end of
both exon 2 and exon 3 (Fig. 1, arrows), although at the end of exon 2, these sequences are mismatched at one base. Analyses of
positively hybridizing bands on Southern blots of the genomic clones,
as well as human genomic DNA, were consistent with the physical map
shown in Fig. 1 (data not shown).
To determine the relative expression of type A and
type B MCP-1 receptor mRNAs in human monocytes and macrophages, we
performed RT-PCR using an upstream (sense) primer corresponding to the
seventh transmembrane domain and a downstream (antisense) primer
derived from either the type A or type B receptor carboxyl tails.
Preliminary reactions were performed using plasmid cDNAs and P1
genomic clones as templates, and primer pairs were selected that gave
robust, reproducible, and comparable amplification of CCR2A and CCR2B. As shown in Fig. 2, amplification of CCR2A or CCR2B
cDNAs with these primers resulted in PCR products of the expected
size (281 and 485 bp, respectively), without detectable
cross-reactivity. CCR2B was amplified in the monocytic THP-1 and
MonoMac6 cell lines but not in the K562 erythroleukemia cell line (Fig.
2B). Freshly isolated human monocytes were also strongly
positive by PCR. Macrophages were obtained by maintaining human
monocytes in culture for up to 7 days, during which time their
scavenger receptor activity (i.e. internalization of
acetyl-low density lipoprotein, which is associated with monocyte to
macrophage differentiation) increased (data not shown). Less PCR
product was seen in the 7-day monocyte/macrophage cultures, suggesting
that the CCR2B receptor mRNA levels had fallen. A similar
expression pattern was seen for the type A MCP-1 receptor (Fig.
2A), although the PCR products were less abundant. Freshly isolated human monocytes and 3-day monocyte/macrophages were positive for CCR2A by PCR, and less product was seen in the 7-day macrophages. An additional band of approximately 1.4 kb was also detected using the
primer specific for the type A receptor (Fig. 2A, top
band). This size corresponds to mRNA from which the second
intron has not been spliced, as evidenced by the identically sized band
obtained in the P1 lambda phage clone (Fig. 2A, lane
9). This transcript would likely encode the type B MCP-1 receptor.
The longer transcript was also seen in THP-1 and MonoMac6 cells but not
in K562 cells. These data suggest that CCR2B is the predominant form of
the MCP-1 receptor in both monocytic cell lines and human
monocyte/macrophages and also indicate that receptor mRNA levels
fall as monocytes mature into macrophages.
Expression of the Two Forms of the MCP-1 Receptor in Transfected Cells
In an attempt to identify functional differences between
the two forms of the receptor, we transfected HEK-293 cells with cDNAs for CCR2A and CCR2B. To facilitate quantitation of receptor expression, we incorporated Flag epitopes (13) at the amino terminus of
the receptor. Previous work from our laboratory has shown that
incorporation of this epitope does not alter ligand binding or receptor
activation (19). Analysis of multiple independent cell lines by flow
cytometry revealed that CCR2B, but not CCR2A, was highly expressed at
the cell surface (Fig. 3A). Similar results were obtained using immunofluorescence, which further revealed intense
staining for CCR2A in permeabilized cells but not on the surface of
intact cells (Fig. 3B). Western blot analysis indicated that
CCR2A was at least as abundant as CCR2B in these cells (data not
shown). These data indicated that whereas CCR2B was expressed at the
cell surface and in the cytoplasm, CCR2A was found almost exclusively
in the cytoplasm.
Thrombin/MCP-1 Receptor Chimeras
Since the two MCP-1
receptors differ only in their cytoplasmic tails, it was likely that
sequences within the 61-amino acid carboxyl tail of the type A receptor
were responsible for its low expression at the cell surface. To test
this hypothesis, chimeras were created in which the cytoplasmic tails
of either the type A or type B form of the MCP-1 receptor were ligated
to the end of the seventh transmembrane domain of the human thrombin
receptor. Trafficking studies of the thrombin receptor have previously
shown that it is expressed at the cell surface, as well as in an
intracellular pool (13, 20). cDNAs encoding the wild-type and
chimeric receptors were transiently transfected into HEK-293 cells.
Replacement of the carboxyl tail of the thrombin receptor with that of
CCR2B resulted in a chimera that was expressed at the cell surface, as
well as intracellularly (Fig. 4A). In
contrast, when the cytoplasmic tail of CCR2A (amino acids 314-374 (7))
was used in place of the thrombin receptor tail, the chimera was not
detected at the cell surface. Permeabilization of these cells revealed
that this receptor was expressed at high levels in the cytoplasm (Fig.
4B) in a manner reminiscent of the distribution of wild-type
CCR2A (see Fig. 3B). Cell surface expression of each of
these constructs, quantitated by ELISA (Fig. 5),
revealed that the CCR2A cytoplasmic tail reduced surface expression of
the thrombin receptor to a level comparable with that of CCR2A. In
contrast, replacement of the wild-type tail with that of CCR2B did not
significantly reduce thrombin receptor expression at the cell
surface. Similar results were obtained in transfected COS-7 cells
(data not shown). These data indicate that the poor surface expression
of CCR2A is attributable to its cytoplasmic tail and further suggest
that sequences in this region contain trafficking information
applicable to other seven transmembrane domain receptors.
CCR2A Cytoplasmic Tail Deletion and Mutation Analysis
To
determine the amino acid sequences responsible for intracellular
retention of CCR2A, two additional constructs were made with the
cytoplasmic tail truncated at either position 349 or 316. We also
prepared a construct in which the two leucine residues at positions 350 and 351 were changed to alanine. Surface expression of these altered
receptors in transiently transfected cells was quantitated by ELISA
(Fig. 6). In both HEK-293 and COS-7 cells, the surface
expression of CCR2A was significantly lower than CCR2B. Removal of the
carboxyl 25 amino acids of the cytoplasmic tail (CCR2A-349) reduced
expression of the receptor to below that of the wild-type CCR2A.
However, deletion of virtually all of the cytoplasmic tail (CCR2A-316)
resulted in a receptor whose surface expression approached that of
CCR2B. Mutation of the Leu-Leu motif to Ala-Ala (CCR2A-L/A) did not
affect surface expression. These results suggest that the cytoplasmic
tail of the CCR2A contains both positive (between 349 and 374) and
negative (316-349) signals for cell surface expression.
Immunofluorescence studies confirmed these results and further revealed
intense cytoplasmic staining of cells transfected with CCR2A,
CCR2A-349, and CCR2A-L/A (data not shown).
The truncation mutants were examined further in signaling experiments.
In transiently transfected COS-7 cells, little phosphatidylinositol turnover was detected with wild-type CCR2A or either of the two carboxyl tail truncation mutants (Fig. 7). As recently
shown by our group (17), co-transfection with the cDNA for the
chimeric G-protein Gqi5 significantly enhanced
agonist-dependent signaling of CCR2A. Signaling by
CCR2A-316 was only comparable with that of CCR2A, despite the
significantly higher surface expression of the truncated receptor.
CCR2A-349, which was poorly expressed at the cell surface, did not
signal as well as CCR2A. Given the robust surface expression of
CCR2A-316 relative to wild-type CCR2A, these data suggest an important
role for the receptor carboxyl tail in receptor coupling to
G-proteins.
Ligand Binding and Signaling Characteristics of CCR2A and CCR2B
To further study signaling and ligand binding, we screened
a number of stably transfected HEK-293 cell lines and identified those
which had the highest levels of CCR2A (25-50% of CCR2B) at the cell
surface. In equilibrium binding assays, 125I-labeled MCP-1
bound specifically to CCR2A-transfected cells (Fig. 8).
Scatchard analysis of these data revealed a dissociation constant
(Kd) of 310 pM, very similar to the
Kd of 260 pM determined previously for
binding to CCR2B (12).
We have previously shown that the activation of type B MCP-1
receptor signals mobilization of intracellular Ca2+ in an
agonist-dependent manner, with an EC50 value of
approximately 3 nM (12). Agonist-dependent
mobilization of intracellular calcium was also seen in
CCR2A-transfected cells, although the magnitude of the calcium response
was considerably reduced and the apparent EC50 was higher
(30 nM) (Fig. 9). As was the case for CCR2B
(12), the calcium flux was blocked by pretreatment of the cells with pertussis toxin (data not shown). CCR2A also inhibited adenylyl cyclase
in response to MCP-1 binding (Fig. 10). As compared
with CCR2B, the dose-response curve for inhibition of cyclase by CCR2A was shifted approximately 5-fold to the right, with IC50
values of 0.8 and 1.4 nM, respectively.
We have previously reported the cloning of cDNAs encoding two functional human MCP-1 receptors, designated CCR2A and CCR2B (7). Because these two receptors differed only in their terminal carboxyl tails, we speculated that they arose via alternative splicing of a single gene. Here, we report the organization of the MCP-1 gene and the identification of exons encoding the alternative carboxyl tails of CCR2A and CCR2B. We found that both forms of the receptor were present in human monocytes and macrophages and that the poor cell-surface expression of CCR2A was due to its terminal carboxyl tail. Finally, we have provided evidence that CCR2A receptors that do reach the cell surface are functional in ligand binding and signaling assays.
Sequence analysis of human genomic clones indicated that the MCP-1
receptor gene spans at least 7 kb, comprises 3 exons bounded by
canonical splice site consensus sequences, and contains multiple polyadenylation signals at the 3 ends of exon 2 and exon 3. Whereas exon 1 encodes part of the 5
-UTR of both receptor isoforms, exon 2 encodes the entire open reading frame and 3
-UTR of CCR2B. Exon 3, on
the other hand, codes for only the carboxyl tail and 3
-UTR of CCR2A.
Near the center of exon 2 and at the beginning of exon 3 lie strong 5
donor and 3
acceptor splice site consensus sequences, respectively,
that define the region of the gene removed to form CCR2A. Therefore,
both isoforms share a common 5
end composed of exon 1 and the 5
-half
of exon 2 but differ in their carboxyl tails and 3
-UTRs.
Regulation of expression of the alternative MCP-1 receptor transcripts
may be controlled by recognition of alternative polyadenylation signals
located at the end of exon 2 in a mechanism similar to that shown for
the regulation of membrane-bound versus secreted forms of
immunoglobulin µ chains in B lymphocytes (21, 22, 23). In the MCP-1
receptor gene, if exon 2 polyadenylation signals are recognized,
cleavage and polyadenylation would be expected to produce CCR2B
transcripts. However, if these signals are not recognized, mRNA
synthesis would continue through exon 3, thus providing the 3 acceptor
site necessary for the splicing and synthesis of CCR2A mRNA. In
this case, full-length pre-mRNA encoding both carboxyl tail and
3
-UTR isoforms would be synthesized, and any regulation of MCP-1
receptor isoform expression would be predominantly controlled by the
degree of splicing of these pre-mRNAs. Interestingly, in our
earlier Northern blot analysis of THP-1 and MonoMac6 cell RNAs, the
major hybridizing band detected by carboxyl tail-specific probes to
both CCR2A and CCR2B was 3.5 kb (7). This is the transcript size
predicted if both exons 2 and 3, in their entirety, are included in the
mRNA and may imply that the majority of the mRNA retains the
potential to be spliced to yield CCR2A. Examination of these blots also
revealed the presence of lower molecular weight forms in the CCR2A but
not the CCR2B lanes, consistent with a small amount of splicing to
produce fully processed CCR2A (7).
We have used RT-PCR to examine the relative expression of CCR2A and CCR2B in monocytic cell lines and freshly isolated human monocyte/macrophages at the mRNA level. Quantitative determination of the expression of CCR2A and CCR2B at the protein level is not feasible at the present time because high affinity, carboxyl tail-specific antibodies are not yet available. CCR2B was readily detected in monocytic cell lines, as well as in freshly isolated human monocytes, whereas CCR2A was less abundant. To examine MCP-1 receptor expression in macrophages, human monocytes were maintained in culture for up to 7 days, during which time they differentiate into cells that morphologically resembled macrophages (16). At the same time, scavenger receptor activity increased in the macrophages, compared with freshly isolated monocytes, as expected (24). In the case of both the type A and type B MCP-1 receptors, mRNA levels fell as the monocytes differentiated into macrophages. Taken together, these data are consistent with earlier studies in which radiolabeled MCP-1 failed to bind to macrophages (25, 26, 27) and suggest a possible mechanism for limiting the mobility of macrophages at sites of inflammation or injury.
Significant differences were observed in the cell-surface expression of
CCR2A and CCR2B. In a number of different transfected cell types,
including HEK-293 cells, RAT-1 fibroblasts, COS-7 cells, Jurkat T
cells, and RBL-2H3 basophilic cells, CCR2A was expressed at relatively
low levels at the cell surface, as compared with CCR2B. CCR2A was
synthesized in these cells, as revealed by the very abundant staining
of the cytoplasm after permeabilization. A similar intracellular
localization of a G-protein-coupled receptor has been observed for the
M2-4H receptor (28) and has been postulated to represent
a reserve pool of receptors that could be rapidly transferred to the
plasma membrane. A significant intracellular pool of thrombin receptors
has also been noted in both transfected cells and endothelial cells,
and evidence for translocation of these receptors to the cell surface
has been reported (20). The physiological significance of the
intracellular pool of CCR2A receptors remains to be determined.
Through the creation of receptor chimeras, we found that the poor trafficking of CCR2A to the cell surface was due to its terminal carboxyl tail and that this region of CCR2A could cause retention of other seven transmembrane domain receptors in the cytoplasm. Further support for this hypothesis came from truncation studies, which revealed that removal of the carboxyl tail significantly increased surface expression, as compared with wild-type CCR2A, and implicated amino acids 317-349 as containing the signal(s) for cytoplasmic retention. Examination of the amino acid sequence of the CCR2A tail did not reveal the presence of known retention signals (29) but did reveal the presence of two adjacent leucine residues at positions 350 and 351. Di-leucine motifs function as internalization signals for some receptors, as well as sorting signals in the Golgi membrane (30, 31). For example, a di-leucine motif in the cytoplasmic tail contributes to lysosomal targeting and endocytosis of CD3 (32). Mutation of the leucines at positions 350 and 351 to alanines, however, had no effect on the cell-surface expression of CCR2A.
A limited number of transfected cell lines were identified in which
CCR2A was expressed at the cell surface at 25-50% of the level of
CCR2B. Permeabilization of these cells revealed extremely intense
staining for CCR2A, and only in such lines was CCR2A detected on the
cell surface at greater than 10% of the level of CCR2B. The type A
receptors that were expressed on the cell surface bound MCP-1 with the
same high affinity as CCR2B (12), mobilized intracellular calcium, and
mediated inhibition of adenylyl cyclase in a dose-dependent manner. The reduced magnitude of the intracellular calcium response, as
compared with CCR2B, and the approximately 5-fold increase in the
IC50 for cyclase inhibition are likely due to the relative paucity of CCR2A at the cell surface. An alternative interpretation is
that CCR2A couples less efficiently than CCR2B to G-proteins. In this
regard, Hasegawa et al. (33) have reported that carboxyl tail splice variants of the G-protein-coupled EP3 receptor differed in
their efficiency of coupling to Gi. In our case, the
disparity in the number of cell-surface receptors precludes a more
definitive interpretation of the data. Despite extensive efforts, we
have not been able to isolate a transfected CCR2A cell line with
surface expression equal to that of CCR2B. Therefore, the question
remains unanswered as to whether differences in the carboxyl tails of the two forms of the MCP-1 receptor cause qualitatively different G
protein coupling.
The biological significance of the existence of two forms of the MCP-1 receptor is not clear at this time, in part because we have not identified a cell that expresses CCR2A at high levels. There is precedent, however, for multiple isoforms of chemokine receptors. The two forms of the IL-8 receptor differ in their tissue distributions. IL-8RB is the predominant form in monocytes and lymphocytes, whereas IL-8RA and IL-8RB are present in equal amounts in polymorphonuclear leukocytes (34). Recently reported differences in the affinity of IL-8 for the two forms of the receptor have led to the suggestion that IL-8RB may be relatively more important for the initiation of chemotaxis (34). In contrast, the lower affinity of IL-8 for IL-8RA indicates that this receptor would only become activated in the presence of high concentrations of IL-8 (e.g. at sites of acute inflammation). Similarly, two isoforms of the thromboxane A2 receptor with alternatively spliced carboxyl tails have been found to differ in their tissue distribution (35). Our data suggest that CCR2B is the predominant form of the MCP-1 receptor in human monocytes and macrophages. It is possible that CCR2A is the major form of the receptor in other leukocytes or that its expression is increased upon activation of monocyte/macrophages. It is also possible that CCR2A will be found in non-hematopoietic cells. The elucidation of the structure of the MCP-1 receptor gene represents an important first step in understanding the role of this receptor in normal homeostasis and in response to infection or injury.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBankTM/EMBL Data Bank with accession number(s) U80923[GenBank] and U80924[GenBank].
We thank Drs. Robert Pitas and Karl Weisgraber for a critical reading of the manuscript, Dr. David Sanan for assistance with microscopy, John Carroll and Amy Corder for graphics and photography assistance, Angela Chen for manuscript preparation, and Gary Howard and Stephen Ordway for editorial assistance.