Published online before print October 4, 2005
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Divisions of
* Viral Products, Center for Biologics Evaluation and Research,
Monoclonal Antibodies, and
Therapeutic Proteins, Center for Drug Evaluation and Research, Food and Drug Administration, Bethesda, Maryland
1 Correspondence: Division of Viral Products, Center for Biologics Evaluation and Research, Food and Drug Administration, Building 29B, Room 4NN06, 8800 Rockville Pike, Bethesda, MD 20892. E-mail: zaitseva{at}cber.fda.gov
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: chemokine receptors • cell activation
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The ability of HIV-1 to bind to and enter CD4+ T cells depends on optimal surface expression of CD4 and one of the seven-transmembrane G-protein-coupled chemokine receptors CXC chemokine receptor 4 (CXCR4) and CC chemokine receptor 5 (CCR5), which promote infection with X4 and R5 viruses, respectively [7 , 8 ]. Surface expression of CXCR4 can be down-modulated by cytokines [9 , 10 ], T cell receptor signaling [11 ], and following mitogen stimulation [11 , 12 ]. Yet, activated CD4+ T cells are highly susceptible to infection with X4 viruses, suggesting that factors other than the density of surface CXCR4 may contribute to the efficiency of viral-cell fusion. Despite the important contributions of CXCR4 and CCR5 to HIV-1 entry and pathogenesis, the effects of T cell activation on the function of coreceptors are not well understood.
Although binding of X4 virions to T cells depends on the presence of CD4 and CXCR4, the formation of tri-molecular complexes among the HIV envelope, CD4, and coreceptor molecules at the cell membrane could be affected by coreceptor distribution and conformational changes. Using a panel of monoclonal antibodies (mAb) against well-defined epitopes on CCR5 and CXCR4, both coreceptors were shown to exhibit conformational heterogeneity [13 , 14 ]. Our laboratory reported previously that primary human cells, including monocytes, macrophages, thymocytes, and peripheral blood resting T cells, contain multiple molecular weight (MW) isoforms of CXCR4, which were somewhat cell type-specific [15 ]. Several factors were suggested to play a role in the generation of coreceptor heterogeneity, including receptor oligomerization [16 ] and post-translational modifications of the extracellular domains such as glycosylation and sulfation [17 18 19 20 ]. Ubiquitination of the higher MW CXCR4 species was demonstrated in our previous study [15 ]. Some of the coreceptor post-translational modifications were shown to play a role in the efficiency of viral entry [18 , 20 21 22 ]. However, it is not known whether the increased capability of activated T cells to fuse with X4 viruses correlates with a given post-translational modification and/or a shift in the predominant species of CXCR4 compared with resting cells.
In the current study, primary human T cells were subjected to culture conditions that mimic in vivo cell activation. Using a panel of biochemical and biological methods, T cells were examined for shifts in the predominant CXCR4 coreceptor species and changes in CD4-CXCR4 association, which correlated with the enhanced fusion potential in activated T cells.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
95% pure by flow cytometry. Fresh thymic fragments were obtained at Fairfax Hospital (VA) during cardiac surgery from children (ages 1 month–3 years) with congenital valvular malformations. The tissues were minced, and large aggregates were removed by passage through a nylon mesh. Thymocytes were further enriched by centrifugation over a Ficoll-Paque gradient (Pharmacia Biotech, Uppsala, Sweden). The study received an exempt status by the Research Involving Human Subjects Committee at Center for Biologics Evaluation and Research, Food and Drug Administration (Bethesda, MD). Human-purified T cells and thymocytes were maintained in RPMI medium supplemented with 10% fetal calf serum (FCS; full medium) at 5 x 106/ml at 37°C with 5% CO2 for 24 h or as indicated. T cell activation was induced by adding phytohemagglutinin (PHA; 1 µg/ml, Sigma-Aldrich Co., St. Louis, MO) and interleukin (IL)-2 (20 U/ml, R&D Systems, Minneapolis, MN). In some experiments, cells were treated during the 24-h stimulation phase with Brefedlin A (BFA), an inhibitor of endoplasmic reticulum (ER) to Golgi protein transport (BFA, 10 µg/ml), or with Monensin, an inhibitor of late endosomes (2 µM). All inhibitors were purchased from Sigma-Aldrich Co.
Cell-surface CD4 was removed by incubation of cells with 0.1% Pronase (Calbiochem-Novabiochem Corp., La Jolla, CA) in the presence of DNase 1 (20 U/ml, Ambion Inc., Austin, TX) in phosphate-buffered saline (PBS) for 30 min at 37°C. Following pronase/DNase treatment, cells were washed extensively, first, with cold RPMI supplemented with 20% FCS and then twice, with tissue culture medium. All washes following pronase treatment contained DNase 1.
Analysis of CXCR4 and CD4 surface expression by flow cytometry
The following antibodies were used: rabbit polyclonal antibodies against CXCR4 [23
], rabbit polyclonal antibodies against N-terminal peptide of CXCR4 (Serotec, Oxford, UK), phycoerythrin (PE)-labeled mouse mAb anti-CXCR4 (12G5 mAb), PE-labeled mAb anti-CD4, and fluorescein isothiocyanate-labeled anti-CD25 [anti-IL-2 receptor 
(IL-2R) mAb]. All mAb were purchased from BD Biosciences (San Jose, CA). Indirect staining using rabbit antibodies against CXCR4 was performed using biotin-conjugated goat anti-rabbit F(ab)2 and subsequent incubation with TC-streptavidin; both reagents were from Invitrogen (Carlsbad, CA). Flow cytometry was performed and analyzed with Cell Quest Software on FACSCalibur (BD Biosciences).
Confocal microscopy
Purified anti-CXCR4 antibodies (12G5 mAb, BD Biosciences) were coupled to Alexa® Fluor 546 using a protein-labeling kit (Invitrogen). Mark Marsh (University College London, UK) generously provided mouse anti-CD63 mAb [24
]. Antibodies against human CD4 (Alexa® Fluor-647 conjugate) and goat anti-mouse immunoglobulin G (IgG) antibodies (Alexa® Fluor-488 conjugate) were purchased from Invitrogen.
T cells were fixed in 4% paraformaldehyde for 15 min on ice followed by cell permeabilization in PBS/0.2% gelatin containing 0.05% saponin for 1 h at room temperature. Permeabilized cells were first labeled with mouse anti-CD63 mAb followed by rabbit anti-mouse IgG (Alexa® Fluor-488 conjugate), washed thoroughly, and subsequently, labeled with a mixture of anti-CD4 conjugated to Alexa® Fluor-647 and 12G5 conjugated to Alexa® Fluor 546 mAb. All incubations with antibodies were performed on ice for 1 h. Antibodies were diluted in PBS supplemented with saponin and gelatin. Washings were performed in PBS with saponin. Labeled cells were added to poly-l-lysine-treated coverslips. After a rinse with distilled water, the coverslips were mounted on slides for image acquisition and analysis. To exclude spectral overlap, the settings for image capture were selected using control cells individually labeled with antibodies against CXCR4, CD4, and CD63.
High-resolution (512x512), triply labeled images were acquired using a Zeiss Pascal confocal microscope configured with three lasers (Argon 488 nm/514 nm/543 nm, HeNe 563 nm, and HeNe 633 nm) using a 40x Plan-Neofluar objective lens. Pinhole diameters, detector gain, and lasers were optimized such that images had a full range of pixel intensities (0–255) with little saturation at either end.
HIV-1 envelope-dependent cell fusion assay
Fusion was measured by two techniques: quantification of syncytium formation and flow cytometry (to measure cytoplasmic dye redistribution). To measure numbers of syncytia, resting or PHA/IL-2-activated human T cells were mixed with effector TF228.1.16 cells, which stably express gp160 from HIV-1 IIIB/B10 (x4) [25
] at a 1:1 ratio (105 each cells in 96-well plate) in triplicates. In some experiments, resting and activated T cells were pretreated with 300 nM (Tyr5,12, Lys7)-Polyphemusin II (T22) peptide (Bachem Bioscience Inc., King of Prussia, PA) for 1 h before adding TF228 effector cells to the cell cultures.
Numbers of syncytia were scored following 3–5 h culture
The flow cytometry-based fusion assay was performed following a previously published protocol with slight modifications [26
]. In brief, TF228 effector cells were labeled with the green cytoplasmic dye calcein AM (CaAM; 0.25 µM) in OPTIMEM medium and mixed with target-untreated or PHA/IL-2-activated cells labeled with the orange cytoplamic dye chloromethyl benzoyl amino tetramethyl rhodamine (CMTMR; 17.5 µM) in OPTIMEM medium; both dyes were purchased from Molecular Probes (Junction City, OR). Labeled cells were washed in PBS, mixed at a 1:1 ratio (0.6x106 each cells) in 24-well plates in duplicates, and incubated at 37°C. At multiple time-points, cell cultures were harvested, washed with cold PBS, and incubated with 0.5 ml trypsin/EDTA solution for 1 min at room temperature to disperse aggregates and unfused effector-target complexes. The trypsin cleavage was stopped by adding 10% FCS in RPMI. Cells were pelleted at 5000 rpm in a table-top Eppendorf centrifuge for 10 s, washed with PBS, and subjected to flow cytometry. Dead cells were excluded from the analysis by gating on 7-amino-actinomycin– cells. Cell-cell fusion was quantified as the ratio of the double-positive cell population to the sum of the double-positive and the remaining target (orange) cells [26
].
Solubilization, immunoprecipitation, and Western blotting
The following antibodies were used for immunoprecipitations: anti-CXCR4 mAb 4G10 (kind gift from Edward Berger and Christopher Broder [27
]) and anti-CXCR4 mAb 717 from R&D Systems. The OKT4 hybridoma (anti-CD4 mAb) was obtained from American Type Culture Collection (Manassas, VA). Science Applications International Corp. (Frederick, MD) generated ascites containing 4G10 or OKT4 mAb. Cross-linking of antibodies to protein G-conjugated agarose beads (Pierce, Rockford, IL) for immunoprecipitations was carried out as described previously [15
, 28
] with minor modifications described below.
For direct immunoprecipitations of CXCR4 from total cell extracts, cells were lysed at 20 x 107 cells/ml in Nonidet P-40 (NP-40) lysis buffer containing 1% NP-40, 150 mM NaCl, 10 mM Tris-HCl (pH 7.4), 2 mM EDTA, 5 mM iodacetamide, and protease inhibitors [1 mM 4-(2-aminoethyl) benzenesulfonyl fluoride, 0.5 µg/ml leupeptin, 2 µg/ml aprotinin, and 0.7 µg/ml pepstatin A]. For immunoprecipitation of CD4 and coprecipitation of CXCR4 with CD4, 20 x 107 cells were lysed in 1 ml Brij lysis buffer containing 0.1% Brij 97, 150 mM NaCl, 20 mM Tris (pH 8.2), 5 mM iodacetamide, and 2 mM EDTA, plus protease inhibitors as indicated above. For precipitation of cell-surface CD4 and CXCR4 molecules, activated human T cells (or thymocytes) at 5 x 107 cells/ml were biotinylated by adding 2 mM sulfosuccinimidyl-6-biotinamido-hexanoate-biotin (Pierce) in PBS for 1 h on ice. The reaction was stopped by incubation with glycine at 10 mM for 15 min on ice. After one wash with PBS, cell concentrations were adjusted to 20 x 106 cells/ml, and cells were lysed in NP-40 buffer.
In all experiments, precipitations were performed with cell extracts from 80 x 106 cells in 4 ml NP-40 or Brij lysis buffers using 100 µl packed protein G-conjugated, beaded agarose, conjugated to anti-CD4 or anti-CXCR4 mAb. For precipitation of biotinylated surface proteins, 100 µl packed avidin-agarose beads (Sigma-Aldrich Co.) were used. Precipitated proteins were eluted from beads in 2x Laemmli sample buffer with 8 M urea for 5 min at 95°C and were resolved in 9% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) for 4–5 h at 400 V. Gels were transferred to nitrocellulose membranes, and Western blot analyses were performed. CXCR4 and CD4 proteins were detected using rabbit polyclonal antibodies against CXCR4 at 10 µg/ml [23 ] and rabbit polyclonal antibodies against CD4 at 1 µg/ml (Trinity Biotech, Carlsbad, CA) followed by horseradish peroxidase-conjugated anti-rabbit IgG antibodies (1:10,000, Amersham Biosciences Corp., Piscataway, NJ). Membranes were incubated for 5 min with SuperSignal West Femto maximum sensitivity substrate (Pierce) and exposed to XOMAT-AR Kodak film for varying periods of time. All immunoblots were scanned and subjected to densitometry analysis using AlphaImagerTM imaging system (Alpha Innotech Corp., San Leandro, CA).
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
 View larger version (11K): [in a new window] |
Figure 1. Fusion of T cells with HIV-1 envelope-expressing effector cells is increased after treatment with PHA/IL-2. (A) Purified human T cells were cultured for 24 h in medium alone (open bars) or in the presence of PHA (1 µg/ml) and IL-2 (20 U/ml; hatched bars). At the end of culture, cells were harvested, washed, and cocultured with TF228 effector cells (expressing HIV-1 IIIB envelope). Syncytia were scored at the end of a 4-h culture period. Data are representative of 10 experiments. (B) Purified human T cells, resting or activated for 24 h in the presence of PHA/IL-2, were cultured in medium alone (open bars) on in the presence of 300 nM T22 (solid bars) for 1 h. The T22 peptide was not removed, and TF228 effector cells were added to the cell cultures. Syncytia were scored as described in A. Data are representative of four experiments. (C) Kinetics of cell-cell fusion was monitored by coculturing effector TF228 loaded with CaAM with unstimulated or with PHA/IL-2-activated T cells labeled with CMTMR. At indicated times, cell cultures were harvested and subjected to flow cytometry. Fusion was expressed as the ratio between the percent of double-positive cells and the percent of total CMTMR-positive target T cells. One representative experiment out of three performed with cells derived from three different donors is shown.
|
). The PHA/IL-2 treatment induced an increase in the expression of CD25 on 50% of purified T cells (Fig. 2A
). At the same time, 30% less T cells expressed CD4 following cell activation, in agreement with previous reports demonstrating modest reduction in CD4 levels in response to PHA treatment [32
] (see Fig. 2B
). To test the effects of cell activation on CXCR4 expression, we used mAb 12G5, which recognizes a conformation-dependent epitope in the second extracellular loop of CXCR4, or two rabbit polyclonal antibodies generated against peptides derived from the CXCR4 N terminus. Using mAb 12G5, approximately 30% reduction in surface staining was noted in activated T cells (Fig. 2C)
, in agreement with previously reported data [11
, 12
, 33
]. However, when cells were stained with polyclonal antibodies against the N terminus, no significant differences in the levels of surface CXCR4 between unstimulated and activated T cells were detected (Fig. 2D
and 2E)
. These finding suggested that in activated T cells, some CXCR4 molecules may have undergone conformational changes, making them less accessible to mAb 12G5.
 View larger version (18K): [in a new window] |
Figure 2. Fluorescein-activated cell sorter analysis of surface CXCR4 and CD4 in resting and activated T cells. Resting T cells (thin line) and activated T cells (thick line) were stained with mAb specific for IL-2R (anti-CD25; A), CD4 (B), second extracellular loop of CXCR4 (12G5 mAb; C), and rabbit polyclonal antibodies against CXCR4 N terminus (D, E). Staining with isotype control antibodies is shown in dotted histograms. Data represent five experiments.
|
 View larger version (55K): [in a new window] |
Figure 3. Biochemical analyses of cell membrane CXCR4 and CD4 in activated T cells. Resting and activated T cells from the same donor were subjected to surface biotinylation. Cells were lysed in NP-40 buffer, and the biotinylated proteins were isolated on avidin-conjugated agarose beads. Western blotting (WB) was carried out with rabbit polyclonal antibodies against CXCR4 N-terminal peptide (lanes 1 and 2) and with rabbit anti-CD4 polyclonal antibodies (lanes 3 and 4). In all experiments, the same numbers of T cells, unstimulated and activated, were used. Data represent five experiments with similar results. MWM, .
|
 View larger version (38K): [in a new window] |
Figure 4. T cell activation induced an increase in the amount of CXCR4 coprecipitated with CD4. Resting and activated T cells from the same donor were lysed in NP-40 (A) or Brij (B, C) buffers, and immunoprecipitations (IP) were performed with protein G-conjugated agarose beads cross-linked to (A) mouse anti-CXCR4 mAb or (B, C) mouse anti-CD4 mAb OKT4. Western blotting was carried out with rabbit polyclonal antibodies against CXCR4 N-terminal peptide (A, B) or with rabbit polyclonal antibodies against CD4 (C). Data are representative of five experiments.
|
CXCR4 is found in complex with CD4 in intracellular compartments
The observed increases in the levels of 62–64 kDa CXCR4 species, which coprecipitated with CD4 from activated T cells, could have resulted from the CXCR4/CD4 association that took place on the cell surface or intracellularly. To discriminate between these possibilities, we took advantage of the high sensitivity of CD4 to pronase digestion [34
]. Treatment of activated T cells with pronase/DNase I solution completely removed surface CD4 but only minimally reduced surface CXCR4, as verified by flow cytometry (0–25% in several experiments; Fig. 5
). Prior to biochemical analyses, activated, intact T cells were biotinylated followed by cell lysis. The biotinylated surface proteins were precipitated with avidin-agarose beads and after elution, were subjected to SDS-PAGE. Immunoblotting with anti-CD4 antibodies demonstrated that the pronase treatment indeed removed 97% of surface CD4 in agreement with the flow cytometry results (Fig. 6A
vs. Fig. 5
). It is important that pronase-resistant, intracellular CD4 molecules (16–20% of total cellular CD4) were detected in the extracts of pronase-treated T cells following immunoprecipitation with anti-CD4 antibodies (Fig. 6B) . These protected CD4 molecules most likely represented the intracellular pool of CD4 undergoing internalization and recycling upon activation. It is surprising that 71% of the total cellular 62–64 kDa CXCR4 coprecipitated with CD4 from pronase-treated T cells, compared with untreated cells (Fig. 6C)
. These results demonstrated that a portion of the CXCR4-CD4 complexes was found intact in intracellular compartments.
 View larger version (19K): [in a new window] |
Figure 5. CD4, but not CXCR4, is removed from the surface of PHA/IL-2-activated T cells following treatment with pronase. Activated T cells were left untreated (thin line) or treated with pronase solution (0.1%) in the presence of DNase I (20 U/ml; thick line) for 30 min at 37°C. The CD4 and CXCR4 expression on the surface T cells was determined by flow cytometry. Data are representative of five experiments.
|
 View larger version (19K): [in a new window] |
Figure 6. Removal of cell-surface CD4 molecules only partially diminishes the ability to coprecipitate CXCR4 with CD4 from activated T cells, which were incubated in medium alone (lanes 1, 3, and 5) or in the presence of pronase/DNase solution (lanes 2, 4, 6) as described in Figure 5
. Following treatment, cells were (A) biotinylated (Biot.), lysed in NP-40 buffer, and subjected to precipitation with avidin-conjugated agarose beads; (B, C) cells were lysed in Brij buffer and subjected to immunoprecipitation with the anti-CD4 mAb OKT4. Immunoprecipitated proteins were resolved on 9% SDS-PAGE and reacted in Western blots with anti-CD4 antibodies (A,B) or with rabbit anti-CXCR4 polyclonal antibodies (C). Densitometry analysis of CD4 (MW 55 kDa; A, B) and of CXCR4 (MW 62 kDa; C) bands was performed on scanned images. Numbers under images represent percent of remaining CD4 (A, B) or of CXCR4 (C) in pronase-treated cells compared with untreated cells. Data are representative of three experiments.
|
 View larger version (31K): [in a new window] |
Figure 7. Coprecipitation of CXCR4 with CD4 from human thymocytes after removal of CD4 from the cell surface. Human thymocytes were untreated (lanes 1, 3) or treated with pronase/DNase I solution (lanes 2, 4). (A) Thymocytes were biotinylated and lysed in NP-40 buffer. Surface proteins were subjected to precipitation with avidin-conjugated agarose beads and resolved in SDS-PAGE. Blots were reacted with anti-CD4 (upper panel) or rabbit anti-CXCR4 antibodies (lower panel). (B) Cell extracts were prepared from untreated and pronase-treated thymocytes, subjected to immunoprecipitation with OKT4 mAb. Western blots were reacted with anti-CD4 (upper panel) and anti-CXCR4 rabbit antibodies (lower panel). The numbers under the images in lane 4 represent the percentages of residual CD4 and CXCR4 compared with untreated cells (shown in lane 3). Data are representative of three experiments.
|
 View larger version (58K): [in a new window] |
Figure 8. The 62-kDa species of CXCR4 in thymocytes and activated T cells is ubiquitinated. Human thymocytes (lanes 1 and 3) and activated T cells (lanes 2 and 4) were lysed in Brij buffer, and immunoprecipitation was carried out with OKT4 mAb. Coprecipitated proteins were subjected to SDS-PAGE and reacted in Western blots with rabbit anti-CXCR4 or rabbit antiubiquitin (a-Ub) antibodies. Data are representative of two experiments.
|
 View larger version (26K): [in a new window] |
Figure 9. Effects of BFA on the expression of CXCR4 in activated T cells. (A) T cells were cultured with PHA/IL-2 in the absence or presence of BFA (10 µg/ml) for 24 h, and CXCR4 surface expression was detected by flow cytometry. Staining with isotype control antibody is shown in the shaded histogram. (B) Activated T cells, incubated in the absence or presence of BFA, were lysed in NP-40 buffer. Proteins were immunoprecipitated with anti-CXCR4 mAb, and Western blots were reacted with rabbit anti-CXCR4 antibodies. Data are representative of four experiments.
|
 View larger version (21K): [in a new window] |
Figure 10. Effects of Monensin and BFA treatments on CD4/CXCR4 coprecipitation and on fusion of activated T cells. (A) T cells were incubated with PHA/IL-2 alone or in the presence of BFA (10 µg/ml), Monensin (Mon; 2 µg/ml), or both inhibitors for 24 h. Cell extracts were prepared in Brij lysis buffer. Proteins from total cell extracts were precipitated with OKT4 mAb, and Western blots were reacted with rabbit anti-CD4 (upper) or anti-CXCR4 antibodies (lower). Densitometry was performed on scanned images. The numbers under the blots indicate the percentages of precipitated CD4 and CXCR4 compared with activated T cells cultured without BFA and Monensin. (B) T cells were activated with PHA/IL-2 in the absence (thick line) or in the presence of BFA (dotted line), Monensin (dashed line), or both inhibitors together (thin line). Flow cytometry was conducted after staining with anti-CD4 antibodies. Shaded histogram represents staining with isotype control antibody. (C) T cells were incubated alone or were activated with PHA/IL-2 in the absence or presence of inhibitors and were cocultured with TF228 (IIIB envelope) effector cells. Syncytia were scored after 4 h. Mo, Monensin.
|
To determine the effect of BFA and Monensin treatments on the ability of CXCR4 to associate with CD4, cell extracts were prepared from activated T cells cultured in the absence or in the presence of these inhibitors as described above. Proteins were immunoprecipitated with OKT4 mAb and reacted in the Western blots with anti-CXCR4 antibodies (Fig. 10A , lower panel). BFA and Monensin treatments individually reduced the amount of 62 kDa CXCR4 coprecipitated with CD4 to 71% and 40% of control levels, respectively. An additive reduction in CXCR4 coprecipitated with CD4, down to 24% of control cells, was observed in the extracts from cells treated with PHA/IL-2 and both inhibitors together (Fig. 10A , lower panel). The mechanism of the reduction in the levels of CXCR4 coprecipitated with CD4 detected in Monensin alone-treated cells was not fully understood. As no effect of Monensin on the surface CXCR4 was noted (data not shown), it is possible that only an intracellular pool of CXCR4 residing in Monensin-sensitive compartments is able to form complexes with CD4.
Enhanced fusion of activated CD4+ T cells is diminished in cells treated with BFA and Monensin
Thus far, our combined findings suggested that in activated T cells, a significant portion of CD4 and CXCR4 forms bimolecular complexes and that ubiquitination of newly synthesized CXCR4 molecules may be involved in the process. Such complexes could be formed at the cell surface and enter late endosomes following internalization via clathrin-coated pits. Alternatively, newly synthesized CXCR4 molecules may be targeted directly to the late endosomes, where they associate with CD4 molecules, which were internalized from the cell surface. In both cases, CD4-CXCR4 complexes may recycle from the endocytic compartment to the cell membrane and thus contribute to the enhanced fusion activity of the activated cells. To further elucidate the role of late endosomes in the enhanced X4 HIV-1 fusion, envelope-expressing TF228 effector cells (expressing X4 HIV-1 envelope) were cocultured with activated T cells treated with BFA and/or Monensin (Fig. 10C) . As demonstrated in Figure 1
, activated T cells formed 2.5 times more syncytia than resting T cells. Addition of BFA or Monensin alone reduced the numbers of syncytia back to the levels of fusion observed with unstimulated T cells. When both inhibitors were added, syncytium formation was further reduced to <20% of control values (Fig. 10C)
.
Visualization of CXCR4/CD4 in endosomal compartments in activated T cells
The above results suggested that a Monensin-sensitive intracellular compartment is an important site for association between CD4 and ubiquitinated CXCR4 molecules. Confocal microscopy was used to confirm that CD4 and CXCR4 are colocalized intracellularly in late endosomes. To that end, resting and activated T cells were fixed, permeabilized, and stained with antibodies specific for CD4, CXCR4, and late endosomes (Fig. 11
). CXCR4 and CD4 were detected using 12G5-conjugated Alexa-546 (shown in red) and anti-CD4 conjugated to Alexa-647 (shown in blue). Late endosomes were visualized with anti-CD63 mouse mAb [24
] by indirect staining (green color). In resting T cells, low levels of CD4 and CXCR4 were noted to colocalize within endosomes (shown as white overlay, Fig. 11A
). It is important that high levels of CD4/CXCR4/CD63 colocalization were observed in activated T cells (Fig. 11B)
.
 View larger version (69K): [in a new window] |
Figure 11. Detection of CD4/CXCR4 complexes within endosomal compartments by confocal microscopy. Fluorescence micrographs of untreated (A) and activated (B) cells costained with antibodies against CXCR4 (shown in red), CD4 (shown in blue), and endocytic marker CD63 (shown in green) are provided. Split images of fields containing cells labeled with three antibodies individually are shown in the left three panels in A and B; micrographs representing the superposition of images acquired with individual channels are shown in the right panels in A and B; white indicates areas of triple overlap of green, red, and blue. Provided images represent individual optical sections. One representative experiment out of three is shown.
|
Together, these data suggested that a Monensin-sensitive CD63+ cellular compartment plays an important role in the observed increase in CD4-CXCR4 coprecipitation in PHA/IL-2-activated T cells. Furthermore, these interactions involve primarily newly synthesized CXCR4 molecules, which have undergone ubiquitination. This endocytic compartment may provide the milieu for CD4/CXCR4 associations with subsequent shuttling of the preformed complexes to the cell membrane (Fig. 12 ), where they could contribute to the enhanced HIV-1 fusion activity of activated T cells.
 View larger version (36K): [in a new window] |
Figure 12. A model demonstrating the role of endosomes in intracellular association of CD4 with CXCR4 in activated T cells. T cell activation induces an increase in CXCR4 gene transcription, which is followed by protein synthesis and N-linked glycosylation of the protein within ER. From ER, the newly synthesized N-linked, glycosylated CXCR4 is transported to the Golgi apparatus via a BFA-sensitive mechanism. From the Golgi, the 47-kDa can be transferred directly to the cell surface or could be modified further by addition of ubiquitin (Ub) molecules to its C terminus within the Golgi, thus generating the 62-kDa CXCR4. The ubiquitinated 62-kDa CXCR4 is targeted to late endosome/multivesicular body (MVB). At the same time, following PHA activation, surface CD4 is down-regulated and also enters MVB. CD4 and 62 kDa CXCR4 form complexes within the limiting membrane of MVB, are rescued from degradation, and are transferred to the cell-surface membrane as CD4-CXCR4 complexes.
|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
To identify the mechanisms that contribute to the enhanced HIV fusion, we set up biochemical assays to analyze changes in the predominant species of CXCR4 and in the degree of association between CXCR4 and CD4 molecules in activated T cells. Using Western blot analysis of CXCR4 precipitated from total cell extracts, we identified a correlation between the augmented fusion potential of activated T cells and an increase in the densities of 50 and 62–64 kDa CXCR4 species. Furthermore, following CD4 precipitation, a significant increase in the amount of the 62- to 64-kDa CXCR4, which coprecipitated with CD4 molecules, was observed, suggesting that T cell activation resulted in the accumulation of CXCR4-CD4 complexes. The association between CD4 and CXCR4 proteins occurred only in intact cells and was not simply a nonspecific attachment as a result of the cell lysis conditions. In the previous studies, we have demonstrated that in T cell lines, the CD4-CXCR4 coprecipitation was strictly dependent on the addition of exogenous rgp120 to intact cells at 37°C [23 ]. However, in primary T cells and monocytes, gp120-indpendent CD4-CXCR4 coprecipitation is commonly observed and may reflect constitutive association [15 , 28 ].
The CD4/CXCR4 association could have occurred on the cell surface or in intracellular compartments. A series of experiments, including stripping of surface CD4 with pronase treatment, confocal microscopy, and treatment with BFA and Monensin during cell activation, suggested that in addition to the cell surface, CXCR4-CD4 complexes are found and may be formed in the CD63+ late endosomes. The CD4-coprecipitated CXCR4 molecules from thymocytes or activated T cells reacted with antiubiquitin antibodies in Western blots. Therefore, we identified ubiquitination as an important post-translational modification of CXCR4, which may alter its intracellular trafficking and could target the ubiquitinated isoforms to late endosomes, where they encounter internalized CD4 molecules. Thus, CD4/CXCR4 complexes may be formed in the endosomal compartment with subsequent recycling to the cell membrane (Fig. 12)
Considering that CD4 antigen density on PBMC is <105 molecules per cell [44 ], the levels of surface CXCR4 expression, its membrane distribution, and proximity to CD4 may have an important impact on the efficiency of X4 viral cell entry. Earlier work in our laboratory demonstrated that addition of soluble gp120 to T cell lines enabled coimmunoprecipitation of trimolecular complexes containing gp120, CD4, and CXCR4 [23 ]. Other studies confirmed and extended this finding. Ugolini et al. [45] used confocal and electron microscopy to visualize CXCR4 and CD4 distribution at the cell membrane and demonstrated low-level CXCR4-CD4 colocalization in transfected cell lines, which was enhanced by pretreatment with gp120. However, it was not clear whether in primary cells, CXCR4 can form heterodimers with CD4 prior to gp120 binding and which isoforms of CXCR4 are involved in such constitutive association. This question was addressed in our subsequent study, where we demonstrated that CD4 could be coimmunoprecipitated with CCR5 and CXCR4 in human monocytes in the absence of exogenous gp120 [28 ]. Similar observations with CCR5 were made by another group [46 , 47 ] and were further confirmed by a recent study demonstrating fluorescent resonance energy transfer between CXCR4 and CD4 receptors in transfected human embryonic kidney cells [48 ].
In the current study, we established a correlation between an enhanced fusion activity in activated primary T cells and an increased coprecipitation of CD4 with the 62- to 64-kDa species of CXCR4. Pretreatment of activated T cells with gp120 did not augment the intensity of the 62-kDa band coprecipitated with CD4 (data not shown). These data implied that the cell-activation signals were sufficient to generate saturating numbers of CD4-CXCR4 complexes at the cell surface, which were not sensitive to further addition of soluble gp120.
The finding that the 62- to 64-kDa CXCR4, which spontaneously associated with CD4, is ubiquitinated in primary monocytes [28
], thymocytes, and activated T cells (this manuscript) is intriguing. Ligand-induced ubiquitination of plasma membrane receptors (including T cell receptor and CXCR4) has been shown to play a major role in targeting these proteins for degradation in proteosomes or lysosomes ([35
] and reviewed in refs. [49
, 50
]). However, in other systems, ubiquitination of several transmembrane proteins was shown to take place in the trans-Golgi and was required for their sorting to MVB (reviewed in refs. [51
, 52
]). Specific examples are the yeast permeases Gap-1 and Tat2 and the mouse protein N4WBP5 [53
, 54
]. The mechanism responsible for targeting proteins from trans-Golgi to MVB was suggested to involve Golgi-resident ubiquitin ligases such as Tul1 ligase [55
] and specific recognition of ubiquitin molecules on modified proteins by the members of the Golgi-localized, 
-ear-containing, Arf-binding family [56
].
In our experiments, ubiquitination of CXCR4 was observed in the absence of exposure to the SDF-1 ligand. Therefore, we hypothesized that in activated T cells, CXCR4 molecules are ubiquitinated in the Golgi and are targeted to late endosomes or MVB, found primarily in macrophages but also in T cells [57 ], for association with internalized CD4 molecules. From MVB, the preformed complexes are recycled to the cell membrane. Pronase treatment removed all surface CD4, as described previously [34 ]. Blocking of ER-to-Golgi transport by BFA significantly reduced and in some experiments, completely inhibited accumulation of the 62- to 64-kDa CXCR4 species (Fig. 8 and data not shown), further suggesting that ubiquitination of newly synthesized CXCR4 occurs in the Golgi as part of the post-translational protein modifications and transport. In addition, the proteasome inhibitor, lactacystin, did not increase the levels of 62 kDa CXCR4 (data not shown). These data suggested that ubiquitination of newly synthesized CXCR4 in activated T cells is unlikely to serve as a signal targeting these CXCR4 molecules for 26S proteosomal degradation.
Based on the current knowledge of the role of ubiquitin in targeting biosynthetic molecules to MVB and our findings, we propose a model, whereby cell activation induces CXCR4 gene transcription [11 , 36 ], and following N-linked glycosylation in the ER [18 , 58 ], the CXCR4 molecule is modified further by ubiquitin in trans-Golgi (Fig. 12) . Ubiquitination of the 62-kDa CXCR4 targets it to MVB. In parallel, T cell activation induces dissociation of the cell-surface CD4 from the p56 Lck, which enables entry of CD4 into the endocytic pathway [38 , 59 ]. Therefore, following cell activation and up-regulation of the CXCR4 transcription, biosynthetically derived, ubiquitinated CXCR4 molecules may associate with surface-derived, endocytosed CD4 within late endosomes/MVB. Subsequently, some of the newly formed CD4/CXCR4 complexes may recycle back to the cell membrane and "prime" the cells for HIV fusion. In an alternative model, the newly synthesized CXCR4 molecules are transported directly to the cell membrane, where they undergo ubiquitination by a ligand-independent mechanism and end up in close enough proximity to surface CD4 molecules, which then associate with surface CXCR4 and promote their cointernalization via clathrin-coated pits into late endosomes/MVB with subsequent recycling of some of the internalized complexes back to the cell membrane. As ligand-independent ubiquitination of surface CXCR4 has not been reported, we currently favor the model presented in Figure 12 . The internalization and recycling of CD4 from late endosomes to the cell surface are key elements in both models. The role of endosomes in CXCR4/CD4 association was supported by confocal microscopy, demonstrating colocalization of CXCR4 with CD4 within the CD63+ endocytic compartment. The nature of the CD4-CXCR4 interactions is the subject of further investigations. Indirect protein–protein interactions (i.e., via actin cross-linking or in lipid raft-like domains) can also contribute to the observed, increased association of CD4 with CXCR4 following cell activation and were not excluded at this juncture.
Together, our findings further suggest that following cell activation, CD4-CXCR4 complexes are accumulated within late endosomes/MVB from where they can recycle to the cell surface and augment HIV-1 binding and cell entry. Therefore, pathogens or vaccines that induce significant cell activation may lead to increased risk of X4 virus selection and expansion in viral loads.
Received January 24, 2005; revised July 26, 2005; accepted July 29, 2005.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
association and HIV-1 entry J. Biol. Chem. 277,29484-29489} and the HIV-1 coat protein gp120: a fluorescence resonance energy transfer (FRET) study J. Pharmacol. Exp. Ther. 310,8-17This article has been cited by other articles:
 |
|
 |
|
  F. Reggeti, C. Ackerley, and D. Bienzle CD134 and CXCR4 expression corresponds to feline immunodeficiency virus infection of lymphocytes, macrophages and dendritic cells J. Gen. Virol., January 1, 2008; 89(1): 277 - 287. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Gaibelet, T. Planchenault, S. Mazeres, F. Dumas, F. Arenzana-Seisdedos, A. Lopez, B. Lagane, and F. Bachelerie CD4 and CCR5 Constitutively Interact at the Plasma Membrane of Living Cells: A CONFOCAL FLUORESCENCE RESONANCE ENERGY TRANSFER-BASED APPROACH J. Biol. Chem., December 8, 2006; 281(49): 37921 - 37929. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
