(Received for publication, January 15, 1997, and in revised form, February 11, 1997)
From the Cell Biology Unit, Glaxo Wellcome Medicines
Research Centre, Gunnels Wood Road, Stevenage, Hertfordshire SG1 2NY,
United Kingdom, the ¶ Departement de Biochimie Medicale, Centre
Medicale Universitaire, 1224 Champel, Geneva 1224, Switzerland,
Geneva Biomedical Research Institute, Glaxo Wellcome Research
and Development, 14 Chemin des Aulx, 1228 Plan les Ouates, Geneva 1228, Switzerland, and ** Affymax Research Institute, Palo Alto,
California 94304
Chemokines are chemotactic proteins which play a
central role in immune and inflammatory responses. Chemokine receptors
are members of the seven transmembrane G-protein coupled family and have recently been shown to be involved in the entry of human immunodeficiency virus (HIV) into target cells. To study chemokine endocytosis in detail we have used novel site-specific chemistry to
make a fluorescently labeled CC-chemokine agonist (rhodamine-MIP-1) and antagonist (NBD-RANTES). We have also generated a CHO cell line
stably expressing a hemagglutinin-tagged version of the CC-chemokine receptor 1 (CCR1), and using these reagents we have examined the receptor-mediated endocytosis of CC-chemokines by confocal microscopy. Our studies reveal that the agonist was internalized and accumulated in
transferrin receptor-positive endosomes whereas the antagonist failed
to internalize. However, receptor-bound antagonist could be induced to
internalize by co-administration of agonist. Analysis of receptor
redistribution following chemokine addition confirmed that
sequestration was induced by agonists but not by antagonists.
Chemokines are a large family of chemotactic proteins which
regulate leukocyte activation and recruitment to sites of inflammation. They can be divided into two main classes, the CXC- and the
CC-chemokines based on the spacing of the first two cysteine residues.
Both CXC- and CC-chemokines bind to seven transmembrane
G-protein coupled receptors (GPCRs)1 which
in most cases are promiscuous in that they will bind more than one
ligand with high affinity (1). Although receptor-mediated endocytosis
of certain GPCRs such as the 2-adrenergic receptor has
been well documented, the endocytic pathways utilized by most GPCRs are
still uncharacterized. Following stimulation by agonist, the
2-adrenergic receptor is rapidly phosphorylated by a
specific GPCR kinase. This uncouples the interaction between the
receptor and the G-protein and allows binding of
-arrestin which
acts as an adapter to recruit the receptor into clathrin-coated
vesicles (2). Several studies have shown that GPCRs are internalized via clathrin-coated vesicles, although this may not be a universal mechanism as alternative endocytic pathways have been described for a
number of different ligands (3-10).
Chemokines such as RANTES (regulated on activation normal T cell
expressed and secreted), MIP-1, and MIP-1
have been shown to be
the major HIV-suppressive factors produced by CD8+ T cells
(11), and a number of recent studies have gone on to show that
chemokine receptors are co-receptors along with CD4 for the entry of
HIV into cells (12-21). This has raised the possibility of using
chemokines therapeutically to block HIV entry. Receptor-mediated endocytosis of chemokines has not been studied in detail, but an
understanding of this process is clearly important if we are to develop
selective therapeutic agents which block HIV entry into cells. To this
end we have studied the receptor-mediated endocytosis of CC-chemokines
by CCR1 as a prototype for the CC-chemokine receptor family.
Recombinant chemokines were expressed and purified from a bacterial expression system as described (22). Fluorescent chromophores were conjugated by chemically coupling to the amino terminus according to the procedures described (23). The formation of rhodamine-MIP-1 necessitated certain modifications of this technology.2
Receptor ExpressionThe full length cDNA encoding CCR1
was cloned by reverse transcriptase-PCR from the human eosinophilic
cell line EOL-3 using specific primers based on the published sequence
(24). The cDNA was subcloned into a pcDNAneo1-based mammalian
cell expression vector (p12ca5) after the addition of the HA epitope
tag (YPYDVPYASLRS) to the 5 end of the receptor cDNA by PCR. The
sequence of the receptor cDNA was verified prior to transfection
into CHO-K1 cells. The CHO-K1 cells were maintained in Dulbecco's
modified Eagle's medium-F12 medium containing 10% heat-inactivated
fetal calf serum, 2 mM glutamine, and 100 units/ml
penicillin/streptomycin (complete medium) and were harvested by
trypsinization and resuspended at 2 × 107 cells/ml in
20 mM HEPES buffer pH 7.3 containing 150 mM
NaCl. Five hundred-microliter aliquots of cells were electroporated with 30 µg of p12ca5 plasmids containing the receptor cDNA at 260 V, 960 µF, using a Bio-Rad Gene Pulser. After electroporation, cells
were transferred to fresh medium and allowed to recover for 48 h
before addition of 600 µg/ml geneticin (G418). Fourteen days after
electroporation, individual G418-resistant colonies were isolated by
ring cloning and maintained in complete medium containing G418.
Individual clones were then tested for binding to the anti-HA
monoclonal antibody 12CA5 (Boehringer), and cells expressing high
levels of receptor were selected by a fluorescence-activated cell
sorter using an anti-mouse fluorescein isothiocyanate conjugate. The
resultant cell population was designated CHO-CCR1 and was subsequently
maintained in complete medium.
Biological activity of the chemokines or modified chemokines was assessed by their ability to induce monocyte chemotaxis in micro-Boyden chambers as described previously (22, 25). Assays were performed either on the THP-1 cell line or on human peripheral blood monocytes.
Immunofluorescence Confocal MicroscopyAll experiments were performed on CHO-CCR1 cells grown overnight on chamber slides (Nunc). For chemokine endocytosis experiments the cells were washed twice with ice-cold buffer (phosphate-buffered saline with 1% bovine serum albumin). Fluorescent chemokines (500 nM) were incubated with the cells for 4 h on ice followed by three washes in ice cold buffer to remove unbound ligand. Fresh medium was added, and the cells were either kept at 4 °C or incubated at 37 °C for 15 or 30 min. At the end of this incubation cells were fixed, permeabilized, and processed for confocal microscopy by standard procedures. For immunofluorescence studies following incubation with chemokines the fixed cells were permeabilized and incubated with an anti-transferrin receptor monoclonal antibody (H68.4, a kind gift from Professor Colin Hopkins, Laboratory for Molecular Cell Biology, University College London) or with the anti-HA antibody 12CA5 (Boehringer). All slides were examined with a Leica TCS4 confocal scanning laser microscope.
To study CC-chemokine endocytosis we generated fluorescently
labeled RANTES and MIP-1 using site-specific chemistry to attach the
fluorophore specifically to the terminal amino group of the polypeptide
chain (23, 26). Using this technique we conjugated rhodamine to
MIP-1
and 7-nitrobenz-2-oxa-1,3-diazole-4-yl (NBD) to RANTES. These
fluorescently labeled CC-chemokines retained their ability to bind
specifically to CCR1 although the binding affinity was slightly reduced
compared with the unmodified agonist (23, 26). Prior to using these
ligands to study endocytosis we assessed their biological activity in
chemotaxis assays. Rhodamine-MIP-1
acted as a full agonist and
induced maximal chemotaxis at a concentration of 100 nM
(Fig. 1A). However, NBD-RANTES had no
biological activity in a chemotaxis assay. We have previously shown
that modification of the amino terminus of RANTES has a profound
influence on its biological properties (25). Failure to cleave the
initiating methionine from bacterially expressed RANTES (Met-RANTES)
produced a protein which was devoid of agonist activity but was a
potent antagonist. Similarly, conjugation of the fluorophore NBD to the amino terminus of RANTES also converted it from an agonist to an
antagonist. As shown in Fig. 1B both Met-RANTES and
NBD-RANTES can fully antagonize the chemotactic activity of RANTES on
THP-1 cells.
For endocytosis studies the human CCR1 cDNA was cloned by reverse
transcriptase-PCR, and an HA epitope tag was placed at the extreme
amino terminus. The tagged receptor was transfected into CHO cells, and
lines stably expressing the receptor were cloned. Radioligand binding
assays confirmed that the tagged CCR1 expressed in CHO cells (CHO-CCR1)
retained high affinity ligand binding and ligand specificity (data not
shown). In the initial experiments the fluorescent rhodamine-MIP-1
and NBD-RANTES were bound to CHO-CCR1 cells for 4 h at 4 °C at
a final concentration of 500 nM. Unbound ligand was washed
off with ice-cold buffer prior to warming the cells to 37 °C for
timed periods up to 30 min (Fig. 2). The agonist,
rhodamine-MIP-1
, was effectively internalized and accumulated in
perinuclear vesicles (panels A-C) whereas the antagonist,
NBD-RANTES, remained almost entirely on the cell surface (panels
D-F). This is consistent with other GPCRs for which it has been
shown that receptor internalization is dependent upon agonist
stimulation and is inhibited by antagonists (4, 6, 9, 10). However,
since we have made NBD-RANTES we were able to perform the experiment to
study the effect of agonist stimulation upon the receptor-mediated
endocytosis of a fluorescent antagonist. Binding of NBD-RANTES in the
presence of an equimolar concentration of RANTES showed that the
antagonist could be induced to cluster and internalize in the presence
of agonist (panels G-I). Although the internalized
antagonist appeared in more peripheral endosomes rather than in the
perinuclear structures in which the agonist accumulated, this
observation nonetheless demonstrated that agonist-bound receptor could
induce the sequestration of antagonist-occupied receptor. These
findings suggest that activation of the GPCR kinase by agonist will
induce phosphorylation, and consequently sequestration, of receptors
occupied by agonists or antagonists. The significance of this relates
to how a virus uses a GPCR to gain entry to a cell. These findings
suggest that HIV would not have to act as an agonist to be internalized
via CC-chemokine receptors provided there is sufficient chemokine
agonist present to induce receptor internalization. This is consistent
with recent findings, using chimeric receptors, that the regions of
CCR5 required for viral entry and for chemokine signal transduction are
distinct (27). It may also explain the observation that in certain
macrophage cultures addition of CC-chemokines actually enhances rather
than inhibits HIV replication (28).
To define the endocytic compartment into which rhodamine-MIP-1 was
being delivered we performed dual-labeling studies with the transferrin
receptor. Rhodamine-MIP-1
was bound to CHO-CCR1 cells at 4 °C,
and receptor-bound ligand was subsequently allowed to internalize for
30 min at 37 °C. Cells were fixed, permeabilized, and stained using
an antibody to the transferrin receptor (Fig. 3).
Analysis of the cells by confocal microscopy revealed that at 4 °C,
the rhodamine-MIP-1
decorated the plasma membrane (panel B) whereas the transferrin receptor staining was both on the
plasma membrane and in numerous endocytic vesicles (panel
A). Following a 30-min warm up the rhodamine-MIP-1
was
internalized into perinuclear endosomal vesicles (panel D)
and showed an exact co-localization with the transferrin receptor
(panel C). Since it is well documented that the transferrin
receptor is a marker for the clathrin-coated pit endocytic pathway it
is reasonable to conclude that the CCR1 is also internalized via this
mechanism.
As a final study we decided to examine the receptor redistribution
following addition of agonist or antagonist. For these experiments
various chemokines were added to CHO-CCR1 cells at 37 °C for 30 min
after which time the cells were washed, fixed, and permeabilized, and
the CCR1 receptor was visualized by staining with the anti-HA
monoclonal antibody (Fig. 4). The various chemokines were added at a concentration of either 50 nM or at 1 µM. Without chemokine addition the receptor was
predominantly seen on the plasma membrane (panel A) and
addition of the antagonists NBD-RANTES (panels B and
C) or Met-RANTES (panels D and E) did
not induce any significant receptor redistribution. However the
agonists MIP-1 (panels F and G) and RANTES
(panels H and I) both induced receptor
sequestration, although MIP-1
appeared more effective in this
respect than RANTES. Even at 50 nM, MIP-1
induced most of the CCR1 to redistribute from the plasma membrane to perinuclear vesicles, and at 1 µM receptor down-regulation was almost
complete.
These observations are the first detailed description of chemokine and chemokine receptor endocytosis, and they show that CCR1 is likely to be internalized via clathrin-coated pits. We demonstrate that both ligand and receptor accumulate in perinuclear endosomes and that receptor internalization is dependent upon agonist stimulation. Provided that all CC chemokine receptors behave like CCR1, does this give us any insight into how we might use chemokines as therapeutic agents to prevent HIV entry into cells? The protective effects of chemokines may be due to two factors. The chemokine may act as a competitive inhibitor preventing binding of the virus to the receptor or it may down-regulate surface receptors so that there are no receptors available for the virus. If the second explanation were true then antagonists, which do not induce receptor sequestration, would not be expected to block HIV entry. Recently published data (29) and our own studies3 confirm that chemokine antagonists are effective at inhibiting HIV entry into target cells. Consequently our study strongly suggests that the observed HIV suppressive effects of chemokines are due to competitive inhibition for receptor binding and not due to receptor down-regulation. These findings have significant implications for the design and discovery of novel therapeutic agents targeted at blocking HIV entry into cells.
We thank Emily Tate, Frederic Borlat, Raphaelle Buser, and Marc-Olivier Montjovent for technical assistance.