|Home | About | Journals | Submit | Contact Us | Français|
Dendritic cells (DCs) enhance human immunodeficiency virus type 1 (HIV-1) infection of CD4+ T lymphocytes in trans. The C-type lectin DC-SIGN, expressed on DCs, binds to the HIV-1 envelope glycoprotein gp120 and confers upon some cell lines the capacity to enhance trans-infection. Using a short hairpin RNA approach, we demonstrate that DC-SIGN is not required for efficient trans-enhancement by DCs. In addition, the DC-SIGN ligand mannan and an anti-DC-SIGN antibody did not inhibit DC-mediated enhancement. HIV-1 particles were internalized and were protected from protease treatment following binding to DCs, but not from binding to DC-SIGN-expressing Raji cells. Thus, DC-SIGN is not required for DC-mediated trans-enhancement of HIV infectivity.
Dendritic cells (DCs) facilitate infection of cocultured CD4+ T cells with human immunodeficiency virus type 1 (HIV-1). This process, known as trans-enhancement of infectivity (5, 7, 17), requires the internalization of HIV-1 particles into a DC compartment reminiscent of the multivesicular body (1, 12, 21), and transmission has been proposed to occur in the vicinity of a DC-T-cell contact referred to as an infectious synapse (15).
The surface subunit of the HIV-1 envelope glycoprotein gp120 binds to the mannose-binding C-type lectin DC-SIGN (CD209), which is expressed on DCs in mucosal tissues and in vitro on monocyte-derived dendritic cells (MDDCs) (6, 7). Expression of DC-SIGN in the B-lymphoblastoid cell line Raji allows for trans-enhancement of infection of T cells with HIV-1 (7, 22), and it has been proposed that DC-SIGN is similarly involved in the enhancement of HIV-1 infection by DCs (2, 7). In addition, DCs can also be productively infected to some extent with HIV-1 (8, 16, 18, 20). As a consequence, it is thought that transmission of HIV-1 from DCs to T cells occurs by a combination of both recycling of endocytosed particles and release of de novo viral particles from infected DCs. Recently, a number of reports have challenged the relevance of DC-SIGN in transmission of HIV-1 to T cells in various cell-based models (9, 11, 16).
To determine the potential relevance of DC-SIGN in HIV infection, we initially sought to evaluate the effect of down-regulating its expression on the trans-enhancement capacity of Raji cells expressing DC-SIGN (Raji DC-SIGN cells) and in vitro differentiated human DCs. One short hairpin RNA (shRNA) sequence specific for DC-SIGN was designed (GCAGTGGGTGAGCTCTCAGAGAAAT), and for a control, we used an shRNA specific for green fluorescent protein (GFP) (GGCTACGTCCAGGAGCGCACC). Both were introduced in a lentiviral transduction vector (19). DC-SIGN was partially down-modulated from the surfaces of Raji DC-SIGN cells transduced with the DC-SIGN-specific shRNA but was unaltered in cells expressing the control shRNA (Fig. (Fig.1A).1A). To reduce DC-SIGN in human dendritic cells, cord blood-derived CD34+ cells were transduced after expansion in serum-free medium supplemented with stem cell factor, thrombopoietin, Flt3 ligand, and interleukin 6 (IL-6). Following selection in puromycin and differentiation of cord blood-derived dendritic cells (CBDCs) in the presence of granulocyte-macrophage colony-stimulating factor (GM-CSF) and IL-4, DC-SIGN was expressed in more than 43% of the cells transduced with the GFP shRNA vector, but in only 2% of the cells transduced with the DC-SIGN shRNA (Fig. (Fig.1A).1A). Cells were otherwise similar for DC markers (CD14−, CD16+, and HLA-DR+) (data not shown). We further evaluated the ability of these cells to enhance HIV-1 infectivity. Cells were preincubated for 4 h with HIV-1 FF (pHIV-SVluc encoding firefly luciferase) (4) pseudotyped with the JRFL envelope before the addition of preactivated CD4+ target T cells. Luciferase activity was measured 2 days later. Direct infection of T cells alone or in the presence of Raji cells was not detectable (Fig. (Fig.1B).1B). Raji DC-SIGN cells transduced with the GFP shRNA vector enhanced HIV-1 infectivity by more than 18-fold, but this was significantly reduced in cells transduced with the DC-SIGN shRNA vector (Fig. (Fig.1B).1B). Strikingly, this reduction of infectivity was observed despite a moderate down-modulation of DC-SIGN, suggesting that enhancement by Raji DC-SIGN cells requires a critical amount of DC-SIGN at their surface. However, CBDCs transduced with either GFP or DC-SIGN shRNA were equally efficient in enhancing HIV-1 infectivity and showed an increase of luciferase activity of 49- to 80-fold over the luciferase activity of control Raji cells (Fig. (Fig.1B).1B). Activation of the DCs by pretreatment with lipopolysaccharide (as shown by up-regulation of CD86 [data not shown]) gave similar results. We sought to down-regulate DC-SIGN with another shRNA system expressing GFP instead of the puromycin resistance gene used by Arrighi et al. (2, 3). Two shRNA expression vectors were used: LV-si-SIGN11, which effectively down-regulates DC-SIGN and LV-si-SIGN26, which is ineffective. Following culture in GM-CSF and IL-4, GFP-positive (GFP+) cells were sorted and DC-SIGN expression was evaluated. With LV-si-SIGN26, 64.6% of the cells expressed DC-SIGN compared to 4.6% with LV-si-SIGN11 (Fig. (Fig.1C).1C). Again, both cell populations were equally efficient at enhancing HIV-1 infectivity more than 21-fold compared with Raji cells (Fig. (Fig.1D).1D). This is in contrast to the initial reports that described shRNA-mediated down-regulation of DC-SIGN in DCs (2, 3) but is in agreement with a subsequent report (9). However, in those experiments, replication-competent virus was used, and viral production was not evaluated in the absence of target T cells; hence, productive infection of the DCs could not be ruled out (13). In our trans-enhancement assay, direct infection of donor DCs in the absence of T cells was minimal, if detectable, and because we used replication-defective viruses, all the enhancement of HIV-1 infectivity that we measured was generated by the initial viral inoculum.
We next performed enhancement assays in the presence of compounds known to specifically block gp120 interaction with DC-SIGN. Preincubation of Raji DC-SIGN cells with mannan, a strong inhibitor of gp120 binding to DC-SIGN (6), at 0.2 μg/μl or 2 μg/μl completely inhibited enhancement of HIV-1 infectivity. In contrast, there was no effect of mannan at 0.2 μg/μl and only a slight inhibition at 2 μg/μl on the enhancement by CBDCs (Fig. (Fig.2A).2A). In the presence of an anti-DC-SIGN antibody (monoclonal antibody 120612 from R&D), enhancement of HIV-1 infectivity by Raji DC-SIGN cells was similarly abolished, but it was not significantly altered with CBDCs. Similar results were obtained with HIV-1 pseudotyped with the HXB2 envelope and with MDDCs (data not shown). This indicated that enhancement by DCs does not appear to require DC-SIGN or mannan-binding C-type lectins. Early experiments using replication-competent virus indicated that, in DC-T-cell cocultures, productive replication of HIV-1 could be inhibited by treatment with an anti-DC-SIGN antibody (7, 12). However, DC-SIGN is also implicated in cis for efficient productive infection (13), suggesting that the anti-DC-SIGN was blocking cis infection of the DCs and not trans-enhancement.
The interaction of gp120 with DC-SIGN in Raji DC-SIGN cells was previously shown to be associated with an internalization event (12). Since DC-SIGN appeared to be differentially involved in the enhancement of HIV-1 infection in Raji DC-SIGN cells and DCs, we evaluated the sensitivity of Raji DC-SIGN cells and DCs to the aforementioned inhibitors of gp120 capture. As expected, gp120 labeled with fluorescein isothiocyanate (gp120-FITC) (Trinity Biotech) was readily captured by both Raji DC-SIGN cells and CBDCs and only minimally by Raji cells (Fig. (Fig.2B).2B). In Raji DC-SIGN cells and CBDCs, preincubation with mannan or with an anti-DC-SIGN antibody reduced gp120 capture (Fig. (Fig.2B).2B). Similar results were obtained with MDDCs (data not shown). This suggested that most gp120 internalization occurred through interaction with DC-SIGN or other C-type lectins in DCs and Raji DC-SIGN cells. We concluded that the capture of gp120 by DCs does not correlate with the trans-enhancement of infectivity. These findings are consistent with data from Gummuluru et al., describing that HIV-1 capture and transmission by DCs are independent of mannose-binding C-type lectins, while gp120 capture can be inhibited by mannan (10, 11). These results prompted us to investigate the cellular localization of gp120 and bona fide viral particles within DCs and Raji DC-SIGN cells.
Raji DC-SIGN cells and CBDCs were incubated with either gp120-FITC or HIV-1 particles labeled with GFP-Vpr (14). gp120-FITC showed a clear intracellular localization in both Raji DC-SIGN cells and CBDCs (Fig. (Fig.3A).3A). In order to specifically determine whether viral particles were extracellular or intracellular, cells were treated with pronase in order to strip off any extracellular virus. The efficiency of the pronase treatment was confirmed by a decrease in DC-SIGN surface expression (Fig. (Fig.3B).3B). Following pronase treatment, gp120-FITC localization remained identically intracellular in both Raji DC-SIGN cells and CBDCs. The GFP-Vpr signal was protected from pronase treatment in CBDCs, demonstrating that the particles visualized were internalized. Identical observations were made with MDDCs (data not shown). In contrast, after pronase treatment of Raji DC-SIGN cells, the GFP-Vpr signal was no longer visible (Fig. (Fig.3A).3A). Thus, viral particles were located on the outside of the plasma membrane in Raji DC-SIGN cells.
We wished to determine whether these findings could be extended to the trans-enhancement of infectivity. Raji DC-SIGN cells and CBDCs were incubated with HIV-1 RL (pHIV-SVluc encoding Renilla luciferase) JRFL (R5) for 2 h at 37°C. Unbound virus was washed away, and cells were treated with pronase. To normalize for any cell damage due to the pronase treatment, donor cells were infected with vesicular stomatitis virus G (VSV-G)-pseudotyped HIV-1 FF during capture of HIV-1 RL JRFL. In this setting, VSV-G-pseudotyped particles efficiently infect donor cells and can thus be used as an indicator of cell viability. Donor cells were incubated with CD4+ T cells, and luciferase activities were measured 48 h later. Direct infection of Raji DC-SIGN cells and CBDCs by the VSV-G pseudotypes was readily detected but was minimal with HIV-1 RL JRFL. DC-SIGN surface expression was significantly reduced in Raji DC-SIGN cells and CBDCs, demonstrating that the pronase treatment was efficient (Fig. (Fig.4A).4A). trans-Enhancement of HIV-1 infectivity by CBDCs was not altered by the pronase treatment (Fig. (Fig.4B).4B). Similar results were obtained with MDDCs (data not shown). On the other hand, trans-enhancement by Raji DC-SIGN cells was clearly abolished. These results demonstrate that in DCs, but not in Raji DC-SIGN cells, a mechanism protects viral particles. Together, these data show that during trans-enhancement of infectivity, HIV-1 particles are internalized by DCs, but not by Raji DC-SIGN cells.
Our findings indicate that the mechanisms by which Raji DC-SIGN cells and DCs enhance HIV-1 infectivity are fundamentally different. In DCs, DC-SIGN is involved in the internalization of recombinant gp120, but this does not appear to correspond to a physiological function in enhancement of HIV-1 infection. DCs internalize infectious viral particles, leading to their protection from protease treatment and to trans-enhancement of infectivity, but neither is dependent on DC-SIGN.
This work was supported by National Institute of Health grants to D.R.L. C.B. was supported by an NIH Institutional AIDS training grant. N.M. was supported by an EMBO fellowship.
We are grateful to Pamela Hoar, Fred Valentine, and the delivery staff at the NYU Medical Center for providing cord blood samples. We thank Vincent Piguet for providing shRNA constructs, John Hirst for assistance with fluorescence-activated cell sorting, Gretchen Diehl for critical reading of the manuscript, and Lara Vojnov for technical assistance.
Published ahead of print on 20 December 2006.