Mechanism of Transdominant Inhibition of CCR5-mediated HIV-1 Infection by ccr5Δ32*

Human chemokine receptor 5 (CCR5) functions as a co-receptor for Human immunodeficiency virus (HIV-1) infection. CCR5 is a seven-transmembrane cell surface receptor. Recently, a naturally occurring mutation of CCR5, ccr5Δ32, has been described. A small number of Caucasians are homozygouslyccr5Δ32/ccr5Δ32, while a larger number of individuals are heterozygously CCR5/ccr5Δ32. Theccr5Δ32/ccr5Δ32 genotype has been linked to a phenotype that is “highly” protected from HIV-1 infection. On the other hand, several studies have shown that the CCR5/ccr5Δ32 genotype confers “relative” protection from AIDS with onset of disease being delayed by 2–4 years. Although it is known that peripheral blood lymphocytes from heterozygous individuals (CCR5/ccr5Δ32) support ex vivo HIV-1 replication at a reduced level compared with CCR5/CCR5 cells, the molecular basis for this observation is unknown. Here we report on events that post-translationally modify CCR5. We show that CCR5 progresses through the endoplasmic reticulum prior to appearing on the cell surface. Mature CCR5 can be post-translationally modified by phosphorylation and/or co-translationally by multimerization. By contrast, mutant ccr5Δ32, although retaining the capacity for multimerization, was incapable of being phosphorylated. ccr5Δ32 heterocomplexes with CCR5, and this interaction retains CCR5 in the endoplasmic reticulum resulting in reduced cell surface expression. Thus, co-expression in cells of ccr5Δ32 with CCR5 produces a trans-inhibition by the former of ability by the latter to support HIV-1 infection. Taken together, our findings suggest CCR5/ccr5Δ32 heterodimerization as a molecular explanation for the delayed onset of AIDS inCCR5/ccr5Δ32 individuals.

The mechanism through which ccr5⌬32 might affect CCR5function remains to be clarified. To assess this issue we studied the effect of ccr5⌬32 on the processing, stability, and cell surface expression of CCR5. We found that 1) CCR5 is post-translationally phosphorylated upon MIP-1␤ stimulation of cells; 2) intracellularly, CCR5 exists as either CCR5/CCR5 or CCR5/ ccr5⌬32 multimers; 3) CCR5 and ccr5⌬32 are found in different cellular locales (the former is predominantly on the cell surface while the latter is retained in the ER); and 4) co-expression of ccr5⌬32 inhibited surface expression of CCR5 and CCR5-mediated infection by M-tropic HIV-1 isolates.
Cell Culture, Infection, and Transfection-HeLa cells were propagated in Dulbecco's modified Eagle's medium (DMEM) with 10% fetal bovine serum. PBMCs were activated for 3 days in RPMI with 10% fetal bovine serum containing PHA and then washed and resuspended into the same medium without PHA and with 100 units/ml interleukin-2 (Boehringer Mannheim). PBMCs were exposed to M-tropic HIV-1 isolate AD8 (500 units of TCID50) for 1 h at 37°C, washed, and then resuspended into fresh medium. Virus replication was monitored by reverse transcriptase assay as described previously (19). Transfection of HeLa cells was performed using calcium phosphate.
Pulse-Chase Analysis-24 h after transfection, cells were washed twice in DMEM without methionine and cysteine starved for 30 min in the same medium at 37°C. [ 35 S]Methionine ϩ cysteine (translabel ICN) at 1 mCi/ml final concentration was added to the cells. The cells were then incubated at 37°C for the indicated amount of pulse time and then washed and resuspended in DMEM with methionine and cysteine for the indicated chase times.
Confocal Microscopy-HeLa cells were seeded onto coverslips and transfected. 24 h later, cells were fixed with fresh 4% paraformaldehyde, pH 7.0, for 10 min at room temperature. Fixed cells were permeabilized with a 2-min wash in 100% methanol at room temperature followed by several washes in PBS with 4% bovine serum albumin (PBS/BSA). Appropriately diluted primary antibody was incubated with coverslips overnight at 4°C. Excess antibody was removed with four washes in PBS/BSA. Species-specific second antibody conjugated to Texas Red (Cappel) was then reacted with the coverslips for 1 h at room temperature followed with four washes in PBS/BSA. The final samples were mounted onto slides and visualized using a Ziess Axiophot confocal microscope.
Yeast Two-hybrid Assay-Two-hybrid assays were performed according to manufacturer's protocols (CLONTECH).

RESULTS AND DISCUSSION
To ask whether CCR5 forms an oligomer, HeLa cells were transfected with pCMV/CCR5-Flag, which contains the Flag epitope fused to the C terminus of CCR5. 24 h later, cells were pulsed with [ 35 S]methionine ϩ cysteine for 15 min, washed, and then chased in complete medium for 1 h. Extracts prepared from pulsed and pulsed ϩ chased cells were immunoprecipitated using anti-Flag M2. The immunoprecipitates were resuspended into either 1 ϫ loading buffer (Fig. 1A, lanes 1 and 2) or "native" loading buffer without ␤-mercaptoethanol with a reduced amount of SDS (0.25%) (Fig. 1A, lanes 3 and 4). The samples were resolved by SDS-PAGE. The reduced mobility of CCR5 in the pulsed ϩ chased sample (Fig. 1A, lane 2) when compared with the pulsed sample (Fig. 1A, lane 1) is consistent with a post-translational modification. 2 When the same analysis was repeated using native loading buffer (Fig. 1A, lanes 3  and 4), the mobility difference observed for pulsed (lane 3) and pulsed ϩ chased (lane 4) samples was replicated. The native buffer-PAGE revealed additional CCR5-specific bands with sizes consistent with dimeric moieties (d; Fig. 1A, lanes 3 and  4).
To characterize better multimerization potentials, we analyzed CCR5 using the yeast two-hybrid approach (Table I). In this analysis, C-terminal truncation mutants of CCR5 were found to interact with wild type CCR5. However, a mutant deleted in the first 58 amino acids failed to interact with intact CCR5. Thus, in yeasts, the CCR5-CCR5 interactive domain resides in the N-terminal portion of the protein, which encompasses the first transmembrane region.

TABLE I
Yeast two-hybrid analysis of CCR5/CCR5 interaction Yeast were transformed by PIDCCR5wt and the different mutants using LiAc method of Gietz et al. (26). Tax/CCR5 constructs was used as negative control. Tax/Tax interaction (27) represents a positive control. A qualitative colony-lift filter assay was performed using 5-bromo-4chloro-3-indolyl-␤-D-galactopyranoside as substrate. Strongly positive colonies (ϩϩϩϩ) appeared as dark blue within 20 min. Positives (ϩϩϩ) were blue within 1 h.
Ϫ ϩϩϩϩ CCR5 in the same cell reduces susceptibility to infection by M-tropic HIV-1.
The infection results led us to consider how ccr5⌬32 might affect the in vivo presentation of CCR5. To address this question, we used confocal immunofluorescence to visualize potential intracellular influences of ccr5⌬32 on CCR5. Upon staining with specific antiserum, steady-state CCR5 and ccr5⌬32 were found in distinctly different subcellular locales. While CCR5 was found on cell surfaces (Fig. 3A), ccr5⌬32 stained in the ER (Fig. 3B). ER localization was observed for all ccr5 mutants (ccr5/6TM, ccr5/5TM, ccr5/4TM, and ccr5/3TM), except ccr5⌬cyt, which was found on the cell surface (data not shown).
When CCR5 and ccr5⌬32 were co-expressed in the same cell, significantly reduced surface staining of the former was seen (Fig. 3C).
lanes [5][6][7][8] for 2 h in the absence (Fig. 4A, lanes 1, 3, 5, and 7) or presence (Fig. 4A, lanes 2, 4, 6, and 8) of 100 ng/ml of MIP-1␤. Cell extracts normalized for radioactivity were then immunoprecipitated with anti-Flag antibody. In the immunoprecipitates, the amount of 35 S-labeled CCR5 was found to be equivalent between mock-treated and cells treated with MIP-1␤ (Fig. 4A, compare lane 1 with lane 2 and lane 3 with  lane 4). However, in the same experiment, a large difference in 32 P-labeled CCR5 in the presence of MIP-1␤ was observed (Fig.  4A, compare lane 5 with lane 6 and lane 7 with lane 8). These results suggest that MIP-1␤ does not affect the ambient expression of CCR5, but does affect the phosphorylation state of this protein.
The cytoplasmic tail of CCR5 is rich in potential phosphate acceptor sites. To examine the contribution of this portion of the protein to phosphorylation, HeLa cells were transfected with pCMV/CCR5⌬cyt. This construction encodes a truncated form of CCR5 lacking the cytoplasmic tail fused at its C terminus to a HA epitope. In this instance, the experiment was performed as that described in the legend to Fig. 4A except that anti-HA was used for immunoprecipitation. As shown in Fig.  4B, similar amounts of 35 S-labeled CCR5⌬cyt was found in mock-treated and cells treated with MIP-1␤ (compare lane 1 with lane 2). However, we failed to detect any 32 P-labeled CCR5⌬cyt (Fig. 4B, lanes 3 and 4). These results suggest that an intact cytoplasmic tail of CCR5 is essential for phosphorylation.
In summary, our study raises three points: 1) functional CCR5 molecule exists as a multimer; 2) expression of ccr5⌬32 reduces cell surface expression of wild type CCR5; and 3) MIP-1␤ induces the phosphorylation of CCR5 in its cytoplasmic tail. Taken together, these points suggest that there are complex post-translational regulatory events for CCR5. Relevant to the observation of receptor multimerization, it is clear that CCR5 mutants such as ccr5⌬32 are defective in some specific aspects of post-translational processing that render them aberrantly trapped in the ER (see Fig. 3). However, these defective mutants, nevertheless, maintain an ability to multimerize with wild type CCR5 and through this heterocomplexing retains the wild type moiety into the ER. We believe that this mutant-wild type tethering mechanism explains, in part, why PBMCs from CCR5/ccr5⌬32 individuals have lower levels of cell surface CCR5 when compared with CCR5/CCR5 counterparts (18) and why CCR5/ccr5⌬32 individuals have delayed progression to disease (13)(14)(15)(16).
The functional role of phosphorylation on CCR5 is less clear. Likely, this modification governs a step in transducing the MIP-1␤ signal from the cell surface into the cytoplasm/nucleus. Recently, Amara et al. (21) reported that internalization of chemokine receptors (CXCR4 and CCR5), induced by their ligand, contribute to chemokines-mediated inhibition of HIV-1 replication. Interestingly, the cytoplasmic tail of CXCR4 is required for SDF-1-mediated internalization of CXCR4 (21). Whether phosphorylation of chemokine receptors influences ligand-induced internalization is unknown. Another possibility is that chemokine-stimulated CCR5 phosphorylation can exert entry-unrelated effects that ameliorate HIV-1 infection. Indeed, MIP-1␤, MIP-1␣, and RANTES have been shown to have unexpected effects on HIV-1 replication in macrophages (22).
We note that there now exists a plethora of co-receptors for HIV-1 (at least six different ones have been reported; reviewed in Ref. 1), suggesting that ligand-based inhibition of HIV-1 infection would be difficult (23). We note additionally that a protective effect to a CCR2 mutation has been recently described (24). Whether the CCR2 mutation shares functional similarities with the CCR5 ⌬32 mutation (e.g. producing nonproductive transdominant multimerization or affecting phosphorylation) remains to be explored. However, the natural existence of transdominant negative chemokine receptor mutants might be a general principle that explains differential resistance of certain subpopulations for AIDS and that might help guide future intervention strategies for HIV-1.