Gp120 V3-dependent Impairment of R5 HIV-1 Infectivity Due to Virion-incorporated CCR5*

Entry of R5 human immunodeficiency virus type 1 (HIV-1) into target cells requires sequential interactions of the envelope glycoprotein gp120 with the receptor CD4 and the coreceptor CCR5. We investigated replication of 45 R5 viral clones derived from the HIV-1JR-FLan library carrying 0–10 random amino acid substitutions in the gp120 V3 loop. It was found that 6.7% (3/45) of the viruses revealed ≥10-fold replication suppression in PM1/CCR5 cells expressing high levels of CCR5 compared with PM1 cells expressing low levels of CCR5. In HIV-1V3L#08, suppression of replication was not associated with entry events and viral production but with a marked decrease in infectivity of nascent progeny virus. HIV-1V3L#08, generated from infected PM1/CCR5 cells, was 98% immunoprecipitated by anti-CCR5 monoclonal antibody T21/8, whereas the other infectious viruses were only partially precipitated, suggesting that incorporation of larger amounts of CCR5 into the virions caused impairment of viral infectivity in HIV-1V3L#08. The results demonstrate the implications of an alternative influence of CCR5 on HIV-1 replication.

Entry of R5 human immunodeficiency virus type 1 (HIV-1) 2 into a target cell requires cooperative interactions of the viral envelope protein gp120 with the receptor CD4 and the coreceptor CCR5 (or CXCR4 for X4 HIV-1) (1-3). These interactions depend on the concentration and distribution of receptor and coreceptor molecules on the cell surface (4 -7). Cells with a large amount of CD4 only require trace amounts of CCR5 for maximal susceptibility to infection by R5 HIV-1, whereas cells low in CD4 require larger amounts of CCR5 for maximal infection (6,8). Sequential binding of the viral surface glycoprotein gp120 to CD4 and CCR5 initiates R5 HIV-1 infection; CD4 attachment induces a conformational change in gp120 that exposes a CCR5 binding domain (9 -11). The coreceptor-binding site located in the bridging sheet and the V3 loop of gp120 also play a crucial role in interacting with the N-terminal domains of the CCR5 (12)(13)(14)(15)(16). Finally, direct interaction between CCR5 and the V3 loop (35-37 amino acid residues) in gp120 induces structural rearrangements of a fusion peptide of gp41, allowing fusion of viral and cellular membranes (14,15,(17)(18)(19).
Envelope viruses are known to down-modulate the receptor expression on infected cells to prevent reinfection (20,21). Post-entry, HIV infection leads to a rapid and potent downmodulation of CD4 molecules expressed at the cell surface. Three viral gene products, Nef, Env, and Vpu, are involved in trafficking and catabolism of down-modulating CD4. Nef enhances CD4 internalization and directs the receptor to lysosomes for degradation (22)(23)(24)(25)(26)(27), whereas Env and Vpu interfere with the transport of newly synthesized CD4 to the cell surface (28,29). Without strict CD4 down-modulation, CD4 induces trapping and aggregation of nascent progeny virions at the cell surface by the high affinity of gp120 for CD4 (30) and a dramatic reduction in the infectivity of released virions by recruitment and sequestration of gp120 molecules away from budding sites or recruitment of nonfunctional gp120-CD4 complexes at the virion surface (31,32).
On the other hand, strict down-modulation of coreceptor CCR5 is not observed in HIV-infected cells, although CCR5 expression on the cell surface is partially reduced by Nef (33). Partial down-regulation of CCR5 and strict down-regulation of CD4 prevent the superinfection of cells in which viral replication is already progressing. CCR5 binding domains, V3 loop, and the bridging sheet domain of gp120 are not exposed until conformational changes in gp120 are induced by interaction with CD4 on the cell surface; therefore, CCR5 would not trap nascent progeny virions at the cell surface. This may be one reason why CCR5 does not need to be strictly down-regulated after viral entry. However, a presence of an unknown inhibitory effect of CCR5 on R5 HIV-1 replication is possible. This paper addresses whether the level of CCR5 in CD4 ϩ T-cell lines influences R5 HIV-1 replication, including late stages of the viral lifecycle. To evaluate the effect of V3 loop on viral replication with respect to CCR5 expression, 45 replication-competent mutant viruses carrying multiple amino acid substitutions in the gp120 V3 loop were used which were derived from an R5 HIV-1 V3 loop library using HIV-1 JR-FLan as background (34). The library contained a set of random combinations of 0 -10 polymorphic amino acid substitutions observed in 31 R5 clinical isolates. Replication of the viruses in a CD4 ϩ T-cell line PM1 expressing low levels of CCR5 and PM1/CCR5 cells expressing high levels of CCR5 was examined. It was found that the viruses revealed different replication phenotypes with respect to CCR5 expression. The present study focused on a viral clone, HIV-1 V3L#08 , with a replication in PM1 cells comparable with wild type HIV-1 JR-FLan but with dramatically suppressed replication in PM1/CCR5 cells. This is the first report that suppression of replication by high expression of CCR5 is V3 loop-dependent and associated with late stages of viral replication.

EXPERIMENTAL PROCEDURES
Cells and Viruses-The human CD4 ϩ T-cell line PM1 (35) was provided by the National Institutes of Health (NIH) AIDS Research and Preference Reagent Program, Division of AIDS NIAID, NIH, and maintained in RPMI1640 (Invitrogen) supplemented with 10% heat-inactivated fetal calf serum (Vitromex). PM1/CCR5 cells were generated by standard retrovirus-mediated transduction of PM1 cells by coculture with PA317 clone #8 cells transfected with pG1TKneo-CCR5 (36) without cloning. MAGIC5 (37) and 293T cells were maintained in Dulbecco's modified Eagle's medium (ICN Biomedicals) supplemented with 10% heat-inactivated fetal calf serum.
pJR-FL an was created in our laboratory from pJR-FL (kindly provided by Dr. Y. Koyanagi, Kyoto University), incorporating AflII and NheI sites into Env at 6395 and 6562 nucleotides, respectively. R5 viruses, carrying a set of random amino acid substitutions in the gp120 V3 loop, were derived from the HIV-1 V3 loop library (34). For construction of V3 loop mutant viruses, amino acid substitutions were introduced into the gp120 V3 loop of pJR-FL an , as described previously (34). For virus preparation, 293T cells (1 ϫ 10 6 ) were transfected with 10 g of molecular clone DNA using the calcium phosphate Pro-Fection Mammalian Transfection System (Promega). The supernatant was collected at 28 h post-transfection and filtered through a 0.22-m filter unit (Millipore) and stored at Ϫ80°C until use.
Viral Replication Assay-For determination of replication phenotype, 4 ϫ 10 4 of PM1 or PM1/CCR5 cells were infected with 8 ng of p24 Gag for 2 h. After washing twice with PBS, the infected cells were incubated at 37°C in a 5% CO 2 atmosphere. On day 6 post-infection, p24 Gag in the supernatant was measured using a p24 Gag ELISA (Zeptometrix).
Flow Cytometry-Cell surface expression of CD4 and CCR5 was analyzed by flow cytometry. Cells were incubated in the staining solution (3% fetal calf serum plus 0.05% sodium azide in PBS) with the mouse monoclonal antibodies (mAbs) antihuman CD4 (SK3, BD Biosciences Pharmingen) or anti-human CCR5 (2D7, BD Biosciences Pharmingen) at 4°C for 30 min. The cells were washed with PBS, and fluorescein isothiocya-nate-conjugated goat anti-mouse IgG antibody was used for antibody-staining. Flow cytometry was performed with a FACSCalibur flow cytometer (BD Biosciences) and analyzed with BD Cell Quest Version 3.1 software (BD Biosciences Pharmingen).
Quantitative PCR-Virus (8 ng p24 Gag) was pretreated in culture fluids with 690 units of DNase I (Worthington Biochem) and added to PM1 or PM1/CCR5 cells (4 ϫ 10 4 ) for 2 h at 37°C. Cells were then washed and incubated at 37°C for 8 h. Total DNA was purified using the QIAamp DNA blood kit (Qiagen) and eluted in a total volume of 200 l. Two l of DNA was analyzed by real-time quantitative PCR. Late reverse transcription products were detected using primers amplifying the region between nucleotides 685 and 789 of the provirus: forward primer (5Ј-ACATCAAGCAGCCATGCAAAT-3Ј), reverse primer (5Ј-ATCTGGCCTGGTGCAATAGG-3Ј), and probe (5Ј-FAM-CATCAATGAGGAAGCTGCAGAATGG-GATAGA-TAMRA-3Ј). Reactions were performed in triplicate in TaqMan Universal PCR master mix using 0.9 pmol of each primer/l and 0.25 pmol of probe/l. After 10 min at 95°C, reactions were cycled for 15 s at 95°C followed by 1 min at 60°C for 40 repeats on an ABI Prism model 7700 thermal cycler (Applied Biosystems).
Virus Infectivity Assay-For infectivity assay, 5 ϫ 10 3 of MAGIC5 cells (37) were plated 1 day before infection into 48-well tissue culture plates. After absorption of virus for 2 h at 37°C, cells were washed twice with PBS and further incubated at 37°C in 5% CO 2 . At 48 h post-infection, the cells were stained, and the number of blue foci in each well was counted (39).
Western Blot Analysis-Four days post-infection, viruses in the supernatant of HIV-1-infected cells were pelleted by centrifugation at 175,000 ϫ g for 60 min. Viral proteins (10 ng of p24 Gag) were separated by 4 -20% SDS-PAGE, transferred to a polyvinylidene difluoride Immobilon P membrane (Millipore), and blocked with 5% milk in PBST (137 mM NaCl, 8.1 mM Na 2 HPO 4 , 2.68 mM KCl, 1.47 mM KH 2 PO 4 , and 0.05% Tween 20) for 8 h at room temperature. Immunodetection was performed with anti-gp120 antibody (Aalto Bio Reagents) and anti-gp41 antibody (2F5; National Institutes of Health AIDS Research and Preference Regent Program) followed by a secondary antibody conjugated to horseradish peroxidase (Sigma) and Chemi-Lumi One (Nacalai Tesque).
Virus Precipitation Assay-Virus immunoprecipitation assay was performed as described previously (40). Anti-HLA-DR (L243), anti-CCR5 (3A9), and anti-CXCR4 (12G5) mAbs were purchased from BD Biosciences Pharmingen. Anti-CCR5 (T21/8) mAb was purchased from BioLegend. A rat immunoglobulin G1 (IgG1) mAb against hepatitis C virus, Mo-8 (41), was used as a rat isotype-matched negative control. Virus (5 ng of p24 Gag) in PBS containing 3% bovine serum albumin was mixed with the mAb at a concentration of 10 g/ml in a final volume of 100 l and incubated for 12 h at 4°C. Then, 10 l of Pansorbin (Calbiochem), a suspension of heat-killed Staphylococcus aureus cells pretreated for 1 h with 3% bovine serum albumin, was added to the virus/mAb mixture. After incubation for 30 min at room temperature, captured viruses were removed by centrifugation (350 ϫ g for 30 min). p24 Gag in the supernatant was determined by p24 Gag ELISA.
Attention was then focused on the R5 L phenotype, HIV-1 V3L#08 , to ascertain why high expression of CCR5 had a suppressive effect on viral replication. Replication of HIV-1 V3L#08 was markedly suppressed in PM1/CCR5 cells, whereas the virus showed similar replication kinetics to HIV-1 JR-FLan with 8 ng of p24 Gag in PM1 cells (Fig. 3, A and B). On day 6 post-infection, p24 Gag in the supernatant was 33-fold lower in PM1/CCR5 than PM1 cells. However, no replication suppression of HIV-1 V3L#08 was observed in peripheral blood mononuclear cells or macrophages derived from three different donors compared with HIV-1 JR-FLan (data not shown). HIV-1 V3L#08 contained 8 amino acid substitutions: Ile 307 to Val, His 308 to Thr, Ile 309 to Met, Phe 315 to Leu, Thr 317 to Ala, Glu 320 to Asp, Ile 321 to Val, and Asp 324 to Asn in gp120 V3 loop alone (Fig. 3C).
It is possible that suppression of HIV-1 V3L#08 replication in PM1/CCR5 cells was due to high susceptibility of the virus to the chemokine(s) produced by the cells. To exclude this possibility, susceptibility of HIV-1 V3L#08 , HIV-1 V3L#08-A69T , HIV-1 JR-FLan , and HIV-1 JR-FLan-A69T to a ␤-chemokine, RANTES (regulated on activation normal T cell expressed and secreted), was measured, but no differences were detected (data not shown).
Entry Efficiency of HIV-1 V3L#08 into PM1/CCR5 Cells-To investigate whether the entry efficiency of HIV-1 V3L#08 decreased, real-time PCR was utilized to analyze cellular accumulation of the late reverse transcriptase product, gag, synthesized shortly after virus entry into the cells (Fig. 4A). There was no decrease in gag DNA synthesis of HIV-1 V3L#08 in PM1/ CCR5 compared with PM1 cells. Rather, the DNA copies of HIV-1 V3L#08 were 3.3-fold higher in PM1/CCR5 than in PM1 cells. Higher expression of CCR5 could promote more efficient viral entry when CD4 is not significantly expressed, consistent with previous reports for other viruses (1.9 -2.5-fold) (6,8). Moreover, there was no clear difference in DNA synthesis among the four viruses in PM1/CCR5 cells.
In addition, the virus pseudotyped with the envelope proteins of HIV-1 V3L#08 revealed 2.0-fold more efficiency in PM1/ CCR5 than in PM1 cells (Fig. 4B). Luciferase activity of other viruses in PM1/CCR5 cells also increased 1.7-2.2-fold compared with PM1 cells. Measured luciferase activity in cells infected with pseudotyped viruses serves as an indirect estimation of viral entry, integration, and transcriptional activity. The results demonstrate that suppression of HIV-1 V3L#08 replication in PM1/CCR5 cells is not associated with early stages of the viral lifecycle.
Production of Virus from PM1/CCR5 Cells Infected with HIV-1 V3L#08 -A comparison was made of the number of virions generated from PM1/CCR5 and PM1 cells in the presence of a reverse transcriptase inhibitor (AZT) and a CCR5 inhibitor Underlined residues indicate the substitutions that were not detected in the 31 R5 viruses, Phe 315 to Cys or Glu 320 to His or Pro but inevitably incorporated into the library due to combinations of nucleotide substitutions.
(TAK-779) to block secondary infection by nascent progeny virus. The concentrations of AZT and TAK-779 used were 4 and 1 M, respectively, 78-and 32-fold higher than the respec-tive IC 50 levels (the concentration required to inhibit 50% of the blue foci formation in MAGIC5 cells) (data not shown). Note that on day 2 post-infection, the number of viruses generated from   HIV-1 V3L#08 -infected PM1/CCR5 cells was 5.7-fold higher than PM1-infected cells (Fig. 5). HIV-1 JR-FLan , HIV-1 JR-FLan-A69T , HIV-1 V3L#08 , and HIV-1 V3L#08-A69T showed similar results. Production of these viruses generated from PM1/CCR5 cells was 3.8 -8.7-fold higher than from PM1 cells, demonstrating that V3 loop structure of HIV-1 V3L#08 did not affect viral assembly and budding. Increased virus production from PM1/CCR5 cells was    consistent with the higher entry efficiency, as shown in Fig. 4, although the increased levels were not equal. This excluded the possibility that replication suppression of HIV-1 V3L#08 in PM1/ CCR5 cells was due to decreased virus production.
CCR5 Incorporation into HIV-1 V3L#08 Virions-To elucidate the significance of CCR5 levels on decreased infectivity in HIV-1 V3L#08 , we examined whether high expression of CCR5 inhibits incorporation of Env proteins gp120 and gp41 into virus particles. Amounts of gp120 and gp41 on the virus envelopes were compared (Fig. 7). There was no decrease in gp120 and gp41 in HIV-1 V3L#08 generated from PM1/CCR5 cells, indicating that high levels of CCR5 has no effect on the incorporation efficiency of gp120 and gp41 into the virions. A slight increase in gp120 and gp41 incorporation into virions was observed in HIV-1 JR-FLan-A69T and HIV-1 V3L#08-A69T compared to HIV-1 JR-FLan and HIV-1 V3L#08 generated from 293T, PM1, and PM1/CCR5 cells. The increased gp120 and gp41 incorporation into virions by the addition of A69T may be in compensation for the loss of infectivity in HIV-1 V3L#08 from PM1/CCR5 cells.
Another putative mechanism could be that CCR5 is incorporated into virus particles and specifically impaired the intrinsic function of HIV-1 V3L#08 gp120. To examine this possibility, the virions were precipitated using anti-CCR5 mAbs (Fig. 8, A and  B). It has been reported that HLA-DR is incorporated into the viral envelope from the infected cell membrane (42,43). We confirmed that an anti-HLA-DR mAb L243 could precipitate   virions generated from PM1 or PM1/CCR5 cells, whereas anti-E2 protein mAb of hepatitis C virus Mo-8, anti-CD4 mAb SK3, or anti-CXCR4 12G5 could not. mAb 2D7 (44), which could recognize the second extracellular loop of CCR5 on the surface of PM1 or PM1/CCR5 cells via flow cytometry (Fig. 2), failed to precipitate viral particles (Fig. 8, A and B). Similarly, 3A9 (45), an anti-CCR5 mAb recognizing the N terminus via flow cytometry (data not shown), could not precipitate viruses either. Surprisingly, another anti-CCR5 mAb (T21/8), also recognizing the N terminus, partially precipitated viruses generated from PM1 (Fig. 8A) and PM1/CCR5 cells (Fig. 8B) but not from 293T cells (data not shown). Note that HIV-1 V3L#08 generated from PM1/CCR5 cells was almost completely precipitated (98%) by mAb T21/8 (Fig. 8B), whereas HIV-1 V3L#08 generated from PM1 cells and the other three viruses from both cell types were partially precipitated (69 -88%) by mAb T21/8 (Fig.  8A). The results indicate that sufficient levels of CCR5 were incorporated into HIV-1 V3L#08 particles to be precipitated by T21/8. The larger amount of incorporated CCR5 reduced viral infectivity in HIV-1 V3L#08 , although the recruitment of CCR5 into virions and the molecular mechanism of decreased infectivity in HIV-1 V3L#08 are unclear. For comparison of the amount of CCR5 at the surface of the virions, we purified virions by precipitation using anti-gp120 antibody 2G12 from exosomes contaminated in the virion by Western blot analysis. However, we could not detect CCR5 possibly because of low affinity of anti-CCR5 antibodies (data not shown).
Amino Acid Substitutions in the V3 Loop Responsible for R5 L Phenotype-In 45 viral clones, three viruses revealed an R5 L phenotype ( Table 1). The amino acid substitutions Ile 307 to Val and Ile 309 to Met were common to all three viruses, suggesting that these two substitutions could be responsible for the R5 L phenotype. To assess the contribution of the eight substitutions in the HIV-1 V3L#08 V3 loop to the R5 L phenotype, eight V3 loop mutant viruses containing one substitution each were prepared, and their replication phenotypes were determined ( Table 2). Replication of HIV-1 I307V and HIV-1 I309M was sup-pressed in PM1/CCR5 cells, whereas these viruses could replicate in PM1 cells (Ͼ60 ng/ml p24 Gag), consistent with the common substitutions obtained in three R5 L viruses described above (Table 1).
To ascertain the importance of these substitutions, V3 mutant viruses containing random combinations of between two and six of the eight substitutions from HIV-1 V3L#08 were prepared. Most of the recombinant viruses preferentially replicate in PM1 cells (ratio Յ 0.5) ( Table 3). Four of 12 viruses revealed a R5 L phenotype (ratio Յ 0.1) and contained I307V and/or I309M. Viruses containing I307V always showed a low p24 Gag ratio (ratio Յ 0.3), whereas ratios in I309M-containing viruses were not as low (Fig. 9), e.g. 1.4 for HIV-1 V3L#130 (Table  3). In addition, F315L frequently conferred a low p24 Gag ratio combined with other substitutions, although F315L alone could not confer R5 L phenotype. The results show that the R5 L phenotype cannot be conferred by just one or two substitutions; a combination of several substitutions, including key substitutions, is essential. The key amino acid residues at 307 and 309 were directly at the N-terminal of the GPGR sequence in the V3 loop crown, suggesting that the amino acid substitutions adjoining the V3 crown may be important in regulating replication kinetics in cells expressing high levels of CCR5.

DISCUSSION
We investigated the replication ability of 45 viral clones from the R5 HIV-1 V3 loop library and found that the library contained different replication phenotypes with respect to the expression level of CCR5 (Table 1). The report focused on one of the viruses, HIV-1 V3L#08 , that was suppressed 33-fold in PM1/CCR5 cells expressing high levels of CCR5 compared with parental PM1 cells (Fig. 3B). This suppression of viral growth was not associated with early events in the lifecycle but with a decrease in viral infectivity in nascent progeny virus from cells expressing high levels of CCR5. MAGIC5 (37) and NP2/CD4/ hiCCR5 cells expressed CD4 at levels similar to and CCR5 at levels compatible to or higher than that of PM1/CCR5 cells;    DECEMBER 21, 2007 • VOLUME 282 • NUMBER 51 however, HIV-1 V3L#08 showed no CCR5-dependent replication suppression compared with MAGI/CCR5, NP2/CD4/ lowCCR5 expressing a similar level of CD4, and a lower level of CCR5 (data not shown). The results suggest that distribution or localization of CCR5 on the surface in PM1 cell line may be distinct from these adherent cells.

Suppression of HIV-1 Replication by CCR5
Depending on the surface of the host cell, HIV-1 incorporates cell-derived molecules into its envelope (46,47). HLA-DR, ␤ 2 -microgloblin, ICAM-1 (intercellular adhesion molecule 1), and other cellular surface proteins were incorporated into a budding virion (48,49), whereas CD4, CXCR4, and CCR5 were not detectable (50). Using the anti-CCR5 mAb, T21/8, we found CCR5 incorporated into a budding virion (Fig. 8). In contrast, two anti-CCR5 mAbs, 2D7 and 3A9, failed to capture virions from PM1 and PM1/CCR5 cells despite the fact that 2D7 (Fig. 2) and 3A9 (data not shown) could recognize CCR5 at the cell surface, indicating that the epitopes of 2D7 and 3A9 were lost by conformation change or concealed from the antibodies by interaction with other molecule(s) on a virus particle. The epitope of 2D7 is mapped to the second extracellular loop including Lys 171 -Glu 172 by alanine mutagenesis scan (44). The tertiary structure of CCR5 including Ser 7 and 9 IYD 11 in the N terminus and His 88 and Trp 94 in the first extracellular loop are estimated to contribute to the binding of 3A9 by phage displayed peptide libraries (45), whereas T21/8 epitope has not been mapped but should be located in the N terminus of 22 amino acid residues from Met 1 to Lys 22 ; T21/8 was raised against a peptide corresponding to residues 1-22 of CCR5 according to the manufacturer. For viral entry, binding of coreceptors to HIV-1 gp120 is mediated by the V3 loop and the coreceptor-binding site located in the bridging sheet of gp120 (12, 14 -16). The N terminus and the first and second extracellular loops of CCR5 are thought to interact with gp120 (51,52). Loss of 2D7 and 3A9 binding with CCR5 at the viral envelope indicate that conformation of CCR5 at the viral envelope is distinct from that of CCR5 at the host cell surface.
The following scenario could explain the decreased infectivity in HIV-1 V3L#08 from PM1/CCR5 cells. In HIV-1 V3L#08-A69T , HIV-1 JR-FLan , and HIV-1 JR-FLan-A69T generated from PM1 and PM1/CCR5 cells, cellular CCR5 is incorporated into the virion envelope but does not impair viral infectivity due to the low level of CCR5. A similar amount of CCR5 incorporated into HIV-1 V3L#08 from PM1 cells is not sufficient to impair viral  Table  2 and 12 mutant viruses in Table 3 were plotted. Closed circle, mutant virus containing one substitution in the V3 loop (Table 2). Open circle, virus containing a random combination of multiple substitutions in the HIV-1 V3L#08 V3 loop (Table 3).   infectivity either. Larger amounts of CCR5 are recruited by HIV-1 V3L#08 from the PM1/CCR5 cell surface into the virions, and the CCR5 levels are above the threshold needed for suppression of viral replication. HIV-1 V3L#08 from PM1/CCR5 cells was almost completely precipitated by T21/8, indicating that most of the virus particles contained sufficient levels of CCR5 for capture (Fig. 8B), although immunoprecipitation efficiency of virions does not proportionally reflect the amount of CCR5. Otherwise incorporated CCR5 may specifically impair gp120 and gp41 function by an unknown mechanism through the HIV-1 V3L#08 V3 loop or via recruitment of nonfunctional gp120-CCR5 complexes at the virion surface. Loss of infectivity seems to be V3-dependent (Tables 2 and 3), and amino acid substitutions near the V3 tip may play a pivotal role on the complex-formation on virions. However, the role of the point mutation of C1 in the revertant virus (HIV-1 V3L#08-A69T ) on V3 conformation or amount of gp120 and gp41 on virions (Fig. 7) is still murky. Another possibility is that CCR5 interacts with HIV-1 V3L#08 gp120 between virus-virus envelopes in the trans position without CD4; the interaction between viral particles through CCR5 and gp120 leads to decrease of viral infectivity, although this is unlikely. To further elucidate incorporation of CCR5 into a virus particle and decrease of viral infectivity, the molecular behavior of CCR5 and HIV-1 V3L#08 gp120 in PM1/ CCR5 cells needs to be studied.

-
There are several reports for donor-dependent variations in CCR5 expression even among donors who do not possess the CCR5⌬32 allele, with a ϳ20-fold variation in CCR5 expression on peripheral blood mononuclear cells from human donors being reported (53,54). CCR5 expression levels in the target cells might be a source of alternative selective pressure on the development of the HIV-1 envelope in vivo, although the range of expression from PM1 to PM1/ CCR5 cells is supposed to be higher than in CCR5-expressing cells in vivo.
Thus, high CCR5 expression does not always benefit replication in all R5 HIV-1. The same V3 loop sequence for the R5 L phenotype, including HIV-1 V3L#08 , HIV-1 V3L#23 , and HIV-1 V3L#25 , could not be found in the data base of HIV-1 clinical isolates, suggesting that HIV-1 has evolved to overcome problems in viral replication caused by the high expression levels of CCR5 in target cells in vivo. The results demonstrate the significant implications of an alternative influence of CCR5 on R5 HIV-1 replication.