The Promiscuous CC Chemokine Receptor D6 Is a Functional Coreceptor for Primary Isolates of Human Immunodeficiency Virus Type 1 (HIV-1) and HIV-2 on Astrocytes (original) (raw)

Abstract

The role of coreceptors other than CCR5 and CXCR4 in the pathogenesis of human immunodeficiency virus (HIV) disease is controversial. Here we show that a promiscuous CC chemokine receptor, D6, can function as a coreceptor for various primary dual-tropic isolates of HIV type 1 (HIV-1) and HIV-2. Furthermore, D6 usage is common among chimeric HIV-1 constructs bearing the gp120 proteins of isolates from early seroconverting patients. D6 mRNA and immunoreactivity were demonstrated to be expressed in HIV-1 target cells such as macrophages, peripheral blood mononuclear cells, and primary astrocytes. In primary astrocytes, an RNA interference-mediated knockdown of D6 expression inhibited D6-tropic isolate infection. D6 usage may account for some previous observations of alternative receptor tropism for primary human cells. Thus, D6 may be an important receptor for HIV pathogenesis in the brain and for the early dissemination of virus in the host.

The cellular tropism of human immunodeficiency virus types 1 and 2 (HIV-1 and -2), the etiological agents of AIDS, is largely determined by the usage of chemokine coreceptors for cell entry (7). The most important chemokine receptor for the in vivo pathogenesis of HIV infections is CCR5, and it appears to be essential for efficient transmission (18). It is expressed on subsets of target CD4+ T cells as well as on other important cell targets such as macrophages and immature dendritic cells (7). The majority of viral isolates from asymptomatic HIV-positive people are CCR5 tropic (R5 viruses) (3, 9), and individuals homozygous for a 32-bp deletion in the CCR5 coding sequence (∼1% of Caucasians) are largely resistant to infection (24). During the course of infection in the individual, viruses may arise that can use several other coreceptors, most notably CXCR4 (X4 viruses) (29). While the use of CXCR4 is associated with disease progression in about 50% of HIV-1 subtype B infections, it is not a prerequisite for the development of clinical AIDS (18).

In contrast, many strains of HIV-1 and HIV-2 have been shown to use a variety of alternative chemokine receptors in vitro. HIV-2 primary isolates have a broader range of coreceptors than HIV-1 isolates (16, 27) and can use alternative receptors, including CCR2b, CCR3, CCR8, and CXCR6, all of which can be expressed on T cells and/or macrophages (7). The inability of viruses to use these receptors on their primary targets has argued against their importance in vivo (18). Indeed, most dual-tropic (R5X4) isolates of HIV-1 cannot use CXCR4 to productively infect macrophages, despite its expression on these cells (28).

Lymphocytes and macrophages are not the sole targets of HIV in vivo. Once the virus infects its host, usually through the genital or rectal mucosa, it can colonize distinct anatomical sites (4). In addition to the lymphoid tissue, viral infection can be found in astrocytes and microglia in the brain, where it can induce neuropathy and dementia. Other sites may include the kidneys, the liver, and cells of the male reproductive tract. Many of these cell types express little or no CD4, and their expression of CCR5 and CXCR4 varies. How alternative coreceptors might participate in the establishment of nonlymphoid reservoirs is thus an important issue, especially since many of these sites are immunoprivileged and poorly accessible to antiretroviral drugs. A recent study showed that a subset of HIV-1 and HIV-2 isolates, including some isolated from the brain, were able to replicate in primary adult astrocytes, brain microvascular endothelial cells (BMVEC), and macrophages by using an alternative coreceptor(s) (36). The identity of this receptor is unclear, but it was sensitive to a blockade of a variety of CC chemokines, including RANTES, eotaxin, and viral macrophage inflammatory protein 1 (vMIP-1), encoded by Kaposi's sarcoma-associated herpesvirus (KSHV). The expression of known receptors for these ligands did not correlate with the cell types susceptible to infection.

Since any potential HIV reservoir would require the transit of virus across endothelial barriers, the virus may either transcytotically penetrate the barrier, cross within an infected cell, or directly infect the endothelial cell itself (6). One recently identified endothelial chemokine receptor, known as D6 or CCBP2, is expressed on the surfaces of lymphatic vessels (20). As with all chemokine receptors, D6 is a seven-transmembrane-spanning protein. However, unlike “classical” chemokine receptors, D6 lacks a DRYLAIVHA G-protein interacting domain and fails to mobilize intracellular calcium upon ligand binding in 293 cells (21). It binds a broad range of inflammatory CC chemokines, including RANTES, eotaxin, and macrophage chemoattractant protein 1 (MCP-1). D6 is constitutively endocytosed (and perhaps transcytosed) and targets bound chemokine ligands for endosomal destruction (10, 12, 35). These processes have led to speculation that D6 (and its close relative the Duffy antigen) has a role in regulating the concentrations of inflammatory chemokines (10) and their endothelial transport (17, 19). D6 expression is also found in the placenta, on myeloid and lymphoid precursor cell lines, and on macrophages in inflamed tissue (8, 22). It is likely that D6 expression or that of related receptors will be found on many other cell types, especially those with roles in tissue homeostasis and repair.

We found that cloned human D6 could function as an efficient coreceptor for a variety of primary HIV-1 and HIV-2 isolates, including those that can use an alternative receptor on primary astrocytes and macrophages. D6 mRNA and protein expression could be detected on these cell types, and RNA interference (RNAi) experiments demonstrated that it could act as the functional coreceptor in this context. Interestingly, a proportion of gp120s amplified from the plasmas of early HIV-1 seroconvertors could also use D6 to enter cells.

MATERIALS AND METHODS

Cells and viruses.

293T cells and the human glioma cell line NP2, expressing CD4 and a variety of human chemokine receptors (30), were maintained in Dulbecco's modified Eagle medium (Invitrogen, United Kingdom) supplemented with 10% fetal calf serum and 500 U/ml penicillin and streptomycin. The primary HIV-1 and HIV-2 isolates used for this study are listed in Table 1. All primary isolates were propagated in fresh human peripheral blood mononuclear cells (PBMC) stimulated with phytohemagglutinin (Sigma, United Kingdom) and recombinant human interleukin 2 (Roche, Germany).

TABLE 1.

HIV-1 and HIV-2 strains used for this study, with their R5 and X4 coreceptor tropisms and known alternative coreceptors

Virus (derivation)a Coreceptor tropismb Known alternative coreceptorsc Reference
HIV-1
HAN-2 (PI) R5/X4 CCR2b, CCR3, CCR8, GPR1 35
2028 (PI) R5/X4 CCR3, CCR8 27
Gun-1v (TCLA) X4/R5 CCR2b, CCR3, CCR8, GPR1 35
2076 (PI) R5/X4 CCR3, CCR8 27
SF-2 (TCLA) X4/R5 CCR3, CCR2b Unpublished
RF (TCLA) X4
IIIB (TCLA) X4
2044 (PI) X4 (M-tropic) CCR8 27
SF162 (PI) R5 CCR3 35
YU2 (PI molecular clone) R5 CCR3 Unpublished
BaL (PI molecular clone) R5 CCR3 Unpublished
HIV-2
ETP (PI) X4/R5 CCR1, CCR2b, CCR3 26
prCBL20 (PI) X4/R5 CCR1, CCR2b, CCR3 15
MIR (PI) X4 CCR1, CCR2b, CCR3 15
TER (PI) R5 CCR1, CCR3, CXCR6 26
JAU (PI) R5 CCR1, CCR2b, CCR3, CCR8 26
ALI (PI) R5 CCR1, GPR15 26
MIL (PI) X4/R5 CCR3 26
prCBL23 (PI) X4/R5 CCR1, CCR2b, CCR3 15
Rod ACR23 (TCLA) X4 CCR3 26

Recombinant viruses with early seroconvertor envelopes.

A cohort of patients presenting with primary HIV-1 infection were recruited as described previously (1). A cDNA encoding the envelope glycoprotein subunit gp120 of the viral RNA in serum was synthesized by reverse transcriptase PCR (RT-PCR) and cloned into the env gene of the molecular clone HXB2, and the viruses thus constructed have been described in detail elsewhere (1). To produce infectious virus, 4 μg of plasmid encoding the recombinant molecular clone was transfected into subconfluent 293T monolayers using Fugene6 (Roche) per the manufacturer's instructions. The virus was harvested at 48 h posttransfection, snap-frozen, and stored in liquid nitrogen.

Cloning and expression of human D6 in NP2/CD4 cells.

The open reading frame encoding human D6 is contained within a single exon on chromosome 3 (15). Human genomic DNA isolated from PBMC was used as a template for a PCR using the following primers with flanking EcoRI and SalI restriction sites: forward, GCGCGAATTCCACCATGGCCGCCACTGCCTCTCC; and reverse, GCGCGTCGACTCAGGCTGATTTATTCCCC. The resulting 1.2-kb fragment was cloned into the retroviral expression vector pBabe-Puro. D6 expression vectors were sequenced and transfected into subconfluent 293T cells cotransfected with the murine leukemia virus packaging construct pHIT60 and pMD-G, encoding the envelope glycoprotein of vesicular stomatitis virus (VSV) (32). The supernatant from these cells was used to transduce NP2/CD4 cells. Puromycin-resistant cells were screened for D6 expression by flow cytometry using an anti-human D6 goat polyclonal antibody (Alexis, United Kingdom) and a secondary donkey anti-goat phycoerythrin-conjugated antibody (Abcam, Cambridge, United Kingdom).

Viral titrations and inhibition assays.

NP2/CD4 cells or variants expressing human chemokine receptors were seeded at 104 cells per well into a 48-well plate 24 h prior to infection. Dilutions of viral supernatants were added to the cells for 2 h at 37°C. The medium was then replaced, and the cells were grown for another 72 h. The cells were then fixed in ice cold acetone-methanol (1:1 [vol/vol]), stained in situ for p24 expression using a 1:1 mixture of the anti-p24 monoclonal antibodies EVA365 and EVA366 (NIBSC, Potters Bar, United Kingdom) and a secondary goat anti-mouse beta-galactosidase-conjugated antibody (Aalto), and visualized by X-Gal (5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside) staining. Titers were determined by enumerating virus-infected foci by light microscopy. For chemokine inhibition assays, an appropriate concentration of chemokine was added to the NP2 cells 20 min prior to infection. The cells were then infected with 200 focus-forming units (FFU) of virus (as determined for that particular cell line) in the presence of chemokine. The cells were then further processed as described above.

Primary cell cultures.

The isolation of the primary human astrocytes used for this study was described previously (36). The cells were maintained in Dulbecco's modified Eagle medium supplemented with 10% fetal calf serum. Primary macrophages were isolated from fresh PBMC by adherence and cultured for 7 days prior to use in RPMI supplemented with 10% pooled human AB serum (Harlan, United Kingdom).

Ad-CD4 transduction.

An adenoviral (Ad) vector encoding human CD4 was propagated in 293 cells and used to transduce astrocytes 24 h prior to infection as described previously (36).

RNAi.

Small interfering RNA (siRNA) oligonucleotides directed against the D6 coding sequence (si1, AAGGCTGCCTCTCTGCAAAGT [from nt 108]; si2, AAGATGGTGAGCACTCTTTAT [from nt 518]) and the 3′ untranslated region of the mature mRNA (si3, AAACCCTTGGCTCAAGCAATT [from nt 2758]) were purchased from QIAGEN (United Kingdom). The oligonucleotides were transfected into subconfluent NP2/CD4/D6 or astrocyte cultures using Oligofectamine (Invitrogen) per the manufacturer's instructions. The cells were then grown for 72 to 120 h before replating and infection, with D6 expression assessed by fluorescence-activated cell sorting.

RESULTS

Efficient usage of D6 as a coreceptor by primary isolates of HIV-1 and HIV-2.

The D6 chemokine receptor was cloned by PCR from human genomic DNA and ligated into the retroviral expression vector pBabe-Puro. Murine leukemia virus-based vectors packaging this construct were produced and used to transduce the human glioma cell line NP2 expressing CD4, and puromycin-resistant cells were selected. D6 expression was confirmed by flow cytometry (see Fig. 4B).

FIG. 4.

FIG. 4.

D6 expression on primary HIV target cells. (A) RT-PCR for D6 mRNA in primary human astrocytes (lanes 1), brain microvascular endothelial cells (lanes 2), and primary macrophages (lanes 3). Cellular glyceraldehyde-3-phosphate dehydrogenase mRNA served as a control for reverse transcription reactions with or without the RT enzyme. (B) Flow cytometry for D6 surface expression. Control goat immunoglobulin G (solid peaks) or polyclonal goat anti-human D6 was used to stain NP2/CD4/D6 cells, primary human macrophages, primary adult astrocytes, and stimulated human PBMC cultures costained with a CD14 or CD4 monoclonal antibody conjugated to fluorescein isothiocyanate. D6 expression was revealed with a secondary anti-goat phycoerythrin-conjugated antibody.

We first sought to investigate whether laboratory and primary isolates of HIV-1 and HIV-2 could use D6 as a functional coreceptor. These isolates and their known coreceptor usages are listed in Table 1. Neither highly passaged T-cell-line-adapted X4 strains (IIIB, RF, and SF2), nor R5 monotropic strains (YU2, BaL, and SF162) grew significantly on NP2/CD4/D6 cells compared to NP2/CD4 cells (Fig. 1A). However, a variety of primary and dual-tropic isolates could grow on these cells relatively efficiently; in particular, HIV-1 Han-2, 2028, Gun-1v, and 2076 had titers up to 3 to 4 orders of magnitude above the background. Not all dual-tropic isolates were capable of using D6. Both the R5X4 strain 89.6 and the macrophage-tropic primary X4 isolate 2044 failed to plate on D6-expressing cells. Primary HIV-2 isolates, which often have a broader receptor tropism than HIV-1 (16), were also tested for D6 tropism. Several HIV-2 strains (MIR, prCBL-20, and ETP) could be plated on D6-expressing cells (Fig. 1B). Interestingly, D6 tropism did not correlate closely with R5X4 tropism (MIR does not use CCR5), but the D6 titers were broadly equivalent to the R5 titers for ETP and prCBL-20. Therefore, D6 can act as an HIV coreceptor, and tropism for D6-expressing cells is relatively common among primary dual-tropic isolates of HIV-1 and HIV-2.

FIG. 1.

FIG. 1.

D6 as a functional chemokine receptor for primary isolates of HIV-1 (A) and HIV-2 (B). Viruses were titrated on NP2/CD4 cells overexpressing no chemokine receptor, D6, CXCR4, or CCR5 and stained in situ for viral antigen expression at 72 h postinfection. Titers are expressed in focus-forming units (FFU)/ml.

Blockade of HIV infection of D6-expressing cells by human and viral CC chemokines.

D6 is a promiscuous CC chemokine binding receptor whose probable function is to regulate inflammatory chemokine concentrations by mediating their endocytosis and endosomal destruction. Interestingly, while many CC chemokines such as RANTES, MCP-1, and eotaxin bind to D6, commercially available recombinant MIP-1α (LD78α), one of two nonallelic forms in humans, fails to bind (23). It does, however, represent a useful control when assessing the chemokine blockade patterns of HIV infection of D6-expressing cells. Figure 2 shows that HIV-1 Han-2 plating on NP2/CD4/D6 cells could be blocked by increasing concentrations of RANTES and eotaxin, but not by recombinant human MIP-1α (LD78α). This blockade could also be achieved for HIV-1 2028 and the seroconvertor MM3 (data not shown). The dependency on D6 was formally demonstrated by the ability of anti-D6 antibodies and siRNAs directed against D6 cDNA (si1 and -2), but not those directed against the mature cellular mRNA 3′ untranslated region (si3), to block the infection of HIV-1 Han-2 (Fig. 2B and C). Interestingly, a chemokine encoded by KSHV, vMIP-1, was also able to block the infection of D6-expressing cells (Fig. 2A). These results suggest firstly that viral chemokines such as vMIP-1 may be D6 ligands and thus subject to D6-mediated transcytosis/endosomal degradation. Secondly, among the primary HIV-1 isolates that were able to use D6 in the above assays were viruses such as HIV-1 HAN-2 and Gun-1v, which were shown to productively infect astrocytes by a receptor sensitive to RANTES and vMIP-1 (36).

FIG. 2.

FIG. 2.

Blockade of D6-mediated infection of NP2/CD4/D6 cells. (A) HIV-1 Han2 was used to infect NP2/CD4/D6 cells in the presence of the indicated concentrations of human recombinant chemokines. Infection is expressed as a percentage of the titer of the virus on untreated NP2/CD4/D6 cells. (B) Blockade of HIV-1 Han2 using 5 μg polyclonal anti-human D6 antibody, processed as described for panel A. (C) Titers of HIV-1 Han2 on NP2/CD4/D6 cells transfected or mock transfected with siRNA oligonucleotides directed against the coding sequence of the D6 cDNA (si1 and si2) or the 3′ untranslated region of the mature D6 mRNA that is not contained within the expression cassette. Titers are expressed in FFU/ml. All error bars represent standard deviations of the means of three experiments.

Envelope glycoproteins derived from early HIV-1 seroconvertors are able to plate on D6-expressing cells.

The currently accepted view of HIV-1 transmission suggests that only R5-using strains are passed through sexual contact (18). However, since traditional virus isolation methods are done by coculture with stimulated PBMC, it is possible that there is a bias against the isolation of strains using alternative coreceptors. Recently, we cloned envelope glycoproteins from seroconvertors presenting with symptoms of primary HIV infection (1). We assessed these recombinant viruses for D6 use (Fig. 3). Interestingly, five of seven individual viral envelopes conferred the ability to infect NP2/CD4/D6 cells with efficiencies about 1% that of CCR5. Viruses derived by cloning envelopes from the laboratory strains IIIB and YU2 failed to show any infection of D6-expressing cells (not shown), implying that the observed results for patients MM1, MM3, MM8, MM23, and MM27 were not a result of PCR errors during cloning. Thus, the functional use of D6 as an alternative coreceptor by these viruses may have implications for viral transmission and dissemination.

FIG. 3.

FIG. 3.

D6 is a coreceptor for viral envelopes derived from early seroconvertors. Titration was performed with HIV-1 HXB2 molecular clones encoding the gp120 proteins of viruses derived from early seroconvertors. Recombinant viruses were titrated on NP2/CD4 cells expressing CD4 alone or in combination with D6, CCR5, or CXCR4.

Expression of D6 by primary human macrophages, PBMC, brain microvascular endothelial cells, and astrocytes.

D6 is primarily expressed on lymphatic endothelial cells, where it is proposed to perform a regulatory/scavenging role in chemokine homeostasis (19). Expression has also been noted in the placenta, the liver, cord blood, and monocytic precursor cell lines such as U937 and THP-1 (22). Since there were similarities in the natures of the viruses that could use D6 as a coreceptor, the patterns of blockade of infection by human and viral chemokines, and the as yet unidentified coreceptors on human astrocytes, BMVEC, and macrophages (36), we sought to assess whether these cell types did indeed express D6. Total RNAs from these cell types were extracted, and RT-PCRs were performed for D6 mRNA. Figure 4A shows that adult human astrocytes, BMVEC, and monocyte-derived macrophages were RT-PCR positive for D6 mRNA. We further determined whether the D6 protein could be detected on the surfaces of monocytes, macrophages, and astrocytes by flow cytometry (Fig. 4B). In all cases, the surface expression of D6 could be detected on these potential HIV target cells. Also, interleukin-2-stimulated human PBMC cultures expressed D6 on a subset of CD14- and CD4-positive cells. Cultured astrocytes did not express CCR5, CCR3, or CXCR4, consistent with previous reports (36), although low levels of CD4 could be detected by RT-PCR (not shown).

D6 is a functional HIV-1 coreceptor on astrocytes.

Given the expression of D6 on primary cultured HIV target cells, we determined whether this molecule could serve as a functional HIV-1 coreceptor. Since macrophages express a variety of chemokine receptors, the relative contribution of D6 would be difficult to determine. We therefore sought to assess whether D6 usage conferred astrocyte tropism to primary HIV-1 and HIV-2 isolates. Cultured primary astrocytes express very low levels of CD4, with CCR5, CXCR4, and major alternative coreceptor expression being undetectable by RT-PCR (36). The expression of CD4 was increased by transduction with an adenoviral vector expressing CD4 (Ad-CD4). The HIV-1 primary isolates Han-2 and 2028, the chimeric viruses bearing seroconvertor envelopes from patients MM1, MM2, MM3, and MM23, and the HIV-2 isolates prCBL-20 and MIR could each infect the primary astrocyte culture 001A after transduction with AdV-CD4, with all save 2028 and MM1 able to infect in the absence of exogenous CD4 expression (Fig. 5A). Prototypic X4 (NL4.3) and R5 (BaL) strains of HIV-1 could not infect these cells under either circumstance.

FIG. 5.

FIG. 5.

D6 is a functional coreceptor on primary human astrocytes. (A) HIV-1 and HIV-2 primary isolates and recombinant viruses bearing seroconvertor envelopes were titrated on primary adult astrocyte 001A cultures that were preexposed or not to an adenoviral vector encoding human CD4. Cells were fixed and stained at 72 h postinfection. (B) siRNA-mediated knockdown of D6 in human astrocytes. Astrocyte cultures 001A and 003 were transfected with D6 siRNAs (1:1:1 [si1:si2:si3]) or an irrelevant siRNA. Five days later, the cells were replated and transduced with Ad-CD4 overnight, and the indicated viruses were titrated on them and processed as described above. The results are expressed as percentages of infection compared to mock-transfected astrocytes. Error bars represent standard deviations for three experiments. (C) RNAi-treated astrocytes were infected with a VSV-G pseudotyped HIV-1-based vector encoding GFP. At 48 h posttransduction, GFP-positive colonies were enumerated by florescence microscopy. (D) Flow cytometric analysis of surface expression of D6 on 003 astrocytes 5 days after transfection with an irrelevant siRNA or D6 siRNA compared to D6 expression on untransfected cells (solid peaks).

To assess the role of D6 in the infection of these astrocytes, siRNA oligonucleotides directed against D6 or an irrelevant target were transfected into 001A astrocytes, and the cells were cultured for 5 days, replated, and transduced with Ad-CD4. The knockdown of D6 surface expression by D6 siRNAs, but not the irrelevant siRNA, was determined by fluorescence-activated cell sorting (Fig. 5D). Figure 5B shows that Han-2, 2028, MM3, and MIR infection was inhibited in astrocytes transfected with D6 siRNAs but not the irrelevant siRNA. Inhibition ranged from 60% for Han-2 to 98% for MIR and demonstrated that the D6 on cultured astrocytes is a functional HIV coreceptor. Interestingly, the MM2 virus, which did not use D6 above the background level on NP2/CD4/D6 cells and plated efficiently on astrocytes, was unaffected by D6 RNAi, suggesting that there are other potential coreceptors to be defined on these cells. These results were confirmed with a second astrocyte culture, 003, for HIV-1 Han-2 and 2028. In this case, however, the chimeric virus MM3 was not efficiently inhibited by D6 RNAi, perhaps reflecting differences in other putative receptor expression levels between these cells. The RNAi treatment did not substantially affect the transduction of the astrocytes by a VSV-G-pseudotyped HIV-1-based vector encoding a green fluorescent protein (GFP) marker (Fig. 5C), indicating that secondary nonspecific effects of the siRNAs did not interfere with virus infection. Thus, D6 is a functional receptor on primary astrocytes for some primary HIV-1 and HIV-2 strains. In addition, other coreceptors exist on primary astrocytes for some of these strains and for envelopes derived from early seroconvertors.

DISCUSSION

In this report, we show that the promiscuous CC chemokine receptor D6 can be used as a coreceptor for primary dual-tropic HIV-1 and HIV-2 strains on primary human astrocyte cultures. D6 binds a range of CC chemokines, and the infection of D6-expressing cells can be blocked by RANTES and vMIP-1. This is a hallmark of previously unidentified HIV coreceptors on astrocytes, brain microvascular endothelium, macrophages, and PBMC (36). All these cells were shown to express D6 in culture at both the RNA and protein levels. Interestingly, gp120 envelope glycoproteins of HIV-1 derived from early seroconverting patients (<30 days after the onset of primary symptoms) (1) were tropic for D6-expressing cells and astrocytes, suggesting that D6 might be an important coreceptor for the colonization of the brain and subsequent neuropathies.

D6 and its close relative, the Duffy antigen, form a distinct group of broad-specificity chemokine receptors, recently termed “interceptors,” expressed primarily on endothelia (19). They lack interaction domains to couple them to cellular G proteins but appear to undergo constitutive endocytosis, delivering their cargo for endosomal destruction (10, 12, 35). D6 is expressed on lymphatic vessels, where it may regulate inflammatory chemokine concentrations (20). The full tissue expression of D6 remains to be defined, but the data herein on D6 expression on macrophages and astrocytes suggest that many cell types, especially those with tissue homeostasis and support functions, are likely to express receptors like D6.

HIV infection of the brain leads to dementia in up to 30% of AIDS patients (26). HIV-1 and -2 infect microglia, astrocytes, and occasionally, neurons (11), although the contribution of cells other than macrophages to brain pathology remains controversial (5). Brain-tropic isolates have been shown to require less cell surface CD4 for efficient infection (25), while other studies have suggested a role for the mannose receptor in astrocyte infection (14). While R5 viruses predominate in the brain (13, 14) and CCR5 can be found on some astrocytic cells in situ (11), the cultured astrocytes used in this study were negative for CCR5 and CXCR4 (36). Yet they were still susceptible to a range of HIV-1 and HIV-2 primary isolates, including chimeric viruses bearing the envelopes from early seroconvertors. Previous studies showed that this infection was sensitive to a blockade by RANTES, eotaxin, and KSHV vMIP-1, indicating that a subclass of CC chemokine receptors was involved (36). However, none of the known vMIP-1 receptors could be detected in these cells. Here we show that not only is D6 a functional receptor for many of these viruses, but also that infection with this receptor is sensitive to inhibition by RANTES, eotaxin, and vMIP-1. The high concentrations needed to block infection compared to those published for CCR5 probably reflect the cell biology of D6; its lack of G protein signaling and its constitutive endosomal recycling mean that ligand-induced surface down-regulation is much harder to achieve (35). We could demonstrate D6 surface expression on primary astrocytes, and siRNAs against D6 mRNA inhibited the infection of these cells by many of the same viruses. Thus, D6 was at least one of the receptors identified in the previous study. The tropism for astrocytes of MM2, despite its inability to use D6 on NP2 cells, and the various amounts of interference seen for some viruses (especially MM3) suggest that other alternative coreceptors can mediate infections of these cells. Interestingly, U87/CD4 astroglioma cells also express D6 and are able to support infection by Han-2, Gun-1v, prCBL-20, and MIR (data not shown). D6 usage by HIV-1 and -2 may be important for brain pathogenesis. We could detect D6 on other primary cell targets of HIV, i.e., PBMC subsets and macrophages, which also bear vMIP-1-sensitive coreceptors (36), suggesting that D6 may play a role as an alternative coreceptor in lymphoid compartments.

The common use of D6 by gp120s derived from early seroconverting patients may also be important for pathogenesis. All these viruses use CCR5 efficiently, and while CCR5 is the most important receptor for the transmission, establishment, and progression of HIV infection (18), the ability to use alternative coreceptors may allow the colonization of distinct cellular and anatomical niches where CCR5 is limited or absent. Several seroconvertor envelopes tested here were tropic for astrocytes lacking CCR5 and CXCR4, although it appears that at least one (MM2) can use an alternative to D6. We postulate that alternative coreceptors such as D6 may be important for virus dissemination in the host to nonlymphoid tissues early in infection.

The tissue known to express by far the highest levels of D6 mRNA is the syncytiotrophoblastic layer of the placenta (22). Trophoblastic cells can be productively infected with HIV as well as mediating the transplacental transport of virus (2, 4, 31, 33). Therefore, any role of D6 in prenatal mother-to-child transmission should be investigated.

The widespread expression of D6 on lymphatic endothelia has been proposed to regulate inflammatory chemokine concentrations and leukocyte extravasation. The constitutive nature of D6 endocytosis and recycling means that receptor desensitization and down-regulation are difficult to achieve (35). Thus, it will be interesting to assess whether chemokine analogues and small-molecule inhibitors proposed as HIV antivirals can block D6-mediated infection. The blockade by vMIP-1 indirectly suggests that D6 might also regulate the function of KSHV-derived chemokines. Since Kaposi's sarcoma is a neoplasm of the lymphatic endothelium, driven in part by a virally encoded chemokine receptor (34), the role of “interceptors” such as D6 in KSHV pathogenesis merits further investigation.

We concluded that D6 is a functional coreceptor for HIV-1 and -2 on primary astrocytes and perhaps other relevant viral target cells and may represent an important alternative receptor for virus dissemination during infection.

Acknowledgments

We thank Keith Aubin, David Marchant, and Sophie Holuigue for reagents and helpful discussions.

This work was supported by the Medical Research Council, UK.

The authors have no competing financial interests.

REFERENCES