Siglec-1 is a novel dendritic cell receptor that mediates HIV-1 trans-infection through recognition of viral membrane gangliosides

Abstract

Dendritic cells (DCs) are essential antigen-presenting cells for the induction of immunity against pathogens. However, HIV-1 spread is strongly enhanced in clusters of DCs and CD4(+) T cells. Uninfected DCs capture HIV-1 and mediate viral transfer to bystander CD4(+) T cells through a process termed trans-infection. Initial studies identified the C-type lectin DC-SIGN as the HIV-1 binding factor on DCs, which interacts with the viral envelope glycoproteins. Upon DC maturation, however, DC-SIGN is down-regulated, while HIV-1 capture and trans-infection is strongly enhanced via a glycoprotein-independent capture pathway that recognizes sialyllactose-containing membrane gangliosides. Here we show that the sialic acid-binding Ig-like lectin 1 (Siglec-1, CD169), which is highly expressed on mature DCs, specifically binds HIV-1 and vesicles carrying sialyllactose. Furthermore, Siglec-1 is essential for trans-infection by mature DCs. These findings identify Siglec-1 as a key factor for HIV-1 spread via infectious DC/T-cell synapses, highlighting a novel mechanism that mediates HIV-1 dissemination in activated tissues.

Figure 1. Siglec-1 is up-regulated in highly trans-infecting LPS mDCs.

(A) (Left) Comparative HIV-1 capture of LPS and ITIP mDCs: cells were cultured with HIV-1, washed, and lysed to measure viral p24^Gag antigen by ELISA. (Right) Comparative transmission of captured HIV-1 from LPS and ITIP mDCs to a reporter CD4⁺ cell line. Graphs show mean values and standard error of the means (SEMs) from two independent experiments including cells from six donors. (B) Plot of SIGLEC genes (in open circles), CD86 and DC-SIGN (in grey circles) computing the fold change in LPS mDCs compared to ITIP mDCs, and the average gene expression across all samples. Circle size is inversely proportional to adjusted p values. Highlighted in red are statistically differentially expressed genes. Analysis was performed with DCs from four donors matured in parallel with the different stimuli. (C) Relative quantification of SIGLEC1 mRNA expression levels in distinct DCs analyzed by qRT-PCR. Measurements were normalized using the endogenous control housekeeping gene Beta Glucuronidase. Data show means and SEMs of samples from six donors. (D) Cell surface expression of Siglec-1 in distinct DCs analyzed by FACS with mAb 7–239-PE. (Left graph) Geometric mean fluorescence intensity (MFI) of Siglec-1. (Right graph) Percentage of Siglec-1 positive cells. Data show mean values and SEM from two experiments, including cells from six donors. (Histograms) Representative profiles of Siglec-1 staining in distinct DCs derived from one donor.

Figure 2. Siglec-1 expressed in LPS mDCs capture distinct ganglioside containing vesicles, such as HIV-1 viral-like particles, liposomes, and exosomes.

(A) Relative capture of VLP_HIV-Gag-eGFP by LPS mDCs that had been pre-incubated with 10 µg/ml of the indicated mAbs or 500 µg/ml of mannan before VLP exposure for 30 min at 37°C. Values are normalized to the level of VLP capture by mock-treated LPS mDCs (set at 100%). Data show mean values and SEMs from three experiments including cells from nine donors. (B) Relative capture of GM1 containing LUV_HIV-tRed by LPS mDCs as described in (A). Data show mean values and SEMs from two experiments including cells from six donors. (C) Relative capture of Exosomes_DiI by LPS mDCs that had been pre-incubated with 10 µg/ml of the indicated mAbs before exosome exposure for 4 h at 37°C. Values are normalized to the level of exosome capture by isotype-treated LPS mDCs (set at 100%). Data show mean values and SEMs from two experiments including cells from five donors. (D) Capture of VLP_HIV-Gag-eGFP by LPS mDCs that had been pre-incubated with decreasing concentrations of α-Siglec-1 mAb 7D2 before VLP exposure for 30 min at 37°C. Titration of α-Siglec-1 mAb 7–239 is shown in Figure S1. Data show mean values and SEMs from three experiments including cells from six donors. (E) Capture of VLP_HIV-Gag-eGFP by LPS mDCs that had been pre-incubated with or without 2 µg/ml of α-Siglec-1 mAb 7D2 previously treated or not with at least a 100-fold molar excess of the indicated human recombinant proteins. Of note, Siglec-14 shares 100% of amino acid homology with Siglec-5 in the V-set domain. Data show mean values and SEMs from three experiments including cells from nine donors. (F) Kinetics of VLP_HIV-Gag-eGFP capture by iDCs (left graph) and LPS mDCs (right graph) compared to the expression of Siglec-1 over time, assessed after LPS addition to mDCs. Cells were pulsed for 1 h at 37°C with VLP_HIV-Gag-eGFP and labeled for Siglec-1 and HLA-DR in parallel at the indicated time points. For comparative purposes, the maximum geometric MFI values obtained by FACS for each donor were set at 100%. Data show mean values and SEMs including cells from three donors. (G) Positive correlation (ρ = 0.9695) between the geometric MFI of captured VLPs and the mean number of Siglec-1 Ab Binding Sites per cell in different DC subtypes (see also Figure S2 to compare VLP capture capacity among LPS mDCs derived from the same donor). Data show values from three experiments including cells from nine donors.

Figure 3. Siglec-1 captures HIV-1 and traffics with the virus to the same sac-like compartment.

(A) Comparative capture of HIV-1 by distinct DCs that had been pre-incubated with 10 µg/ml of the indicated mAbs or 500 µg/ml of mannan for 30 min before viral exposure. Cells were cultured with HIV-1 in the presence of the indicated reagents, washed, and lysed to measure p24^Gag by ELISA. Viral binding at 4°C in LPS mDCs is shown in Figure S3. Data show mean values and SEMs from two experiments including cells from six donors. (B) Comparative capture of HIV-1 by distinct DCs first exposed to the virus and then treated with the indicated reagents for 30 min before washing. Cells were lysed and assessed by p24^Gag ELISA. Data show mean values and SEMs from two experiments including cells from six donors. (C) Comparative capture of HIV-1 by distinct blood myeloid DCs that had been pre-incubated with 10 µg/ml of the indicated mAbs for 30 min before viral exposure as in panel A. Figure S4 depicts Siglec-1 surface expression levels of blood myeloid cells. Data show mean values and SEMs from two experiments including cells from six donors. (D) Confocal microscopy analysis of LPS mDCs pulsed for 4 h with GM1-containing LUV_HIV-tRed, VLP_{HIV-Gag-Cherry} or HIV-1_Cherry, fixed, permeabilized, and then stained for Siglec-1 with mAb 7–239-Alexa 488. (Inset) Merge of the bright field and maximun fluorescence intensity (scale bar: 5 µm). (3D images) Isosurface representation of DAPI stained nucleus and maximum fluorescence intensity of the sac-like compartment where particles and Siglec-1 accumulate are shown in a 3D volumetric x-y-z data field. (Bar graphs) Quantification of the percentage of GM1-containing LUV_HIV-tRed, VLP_{HIV-Gag-Cherry} or HIV-1_Cherry co-localizing with Siglec-1-Alexa 488 7–239 and vice versa, obtained analyzing at least 50 compartments from LPS mDCs of two donors. The mean and standard deviation of the thresholded correlation coefficient of Pearson (obtained considering all the images) were 0.77±0.07, indicating co-localization. See also Movies S1, S2, S3 or Figure S5 to observe the compartment in relation to the plasma membrane or the cytoplasm of the cells. (E) Confocal microscopy analysis showing the sac-like compartment pattern of Siglec-1 in LPS mDCs after internalization of the α-Siglec-1 mAb 7D2. Cells were labeled with the mAb for 30 min at 16°C, revealed with an Alexa 488 secondary Ab, shifted to 37°C for 4 h, and analyzed. (3D image) 3D reconstruction (representative of 69% of the analyzed DCs) was done as in (D). (Inset) Merge of the bright field and maximun fluorescence intensity (scale bar: 5 µm).

Figure 4. Siglec-1 mediates HIV-1 trans-infection to target cells and accumulates at the infectious synapse.

(A) HIV-1 transmission from distinct DCs to a reporter CD4⁺ cell line. DCs were pre-incubated as in Figure 3A, washed, and co-cultured with reporter cells for 48 h. HIV-1 infection of reporter cells was determined by induced luciferase activity in relative light units (RLUs). Data show mean values and SEMs from two experiments including cells from six donors. (B) HIV-1 transmission from distinct DCs first exposed to the virus and then treated with the indicated reagents for 30 min before washing and co-culture with reporter cells. Data show mean values and SEMs from two experiments including cells from six donors. (C) Confocal microscopy analysis of LPS mDCs pulsed with HIV-1_Cherry and then co-cultured with CD4⁺ T cells to reveal Siglec-1 localization. Co-cultures were stained with α-CD4-Alexa 647 mAb to identify the membrane of CD4⁺ T cells, fixed, permeabilized, and labeled with α-Siglec-1-Alexa 488 7–239 mAb. (Left images) Merge of the bright field and the fluorescence of an x-y plane (scale bar: 5 µm). (Right images) Isosurface representation of DAPI-stained nucleus and maximum fluorescence intensity of the compartment where HIV-1_Cherry and Siglec-1 accumulate in the contact zone with a CD4⁺ T cell, shown in a 3D volumetric x-y-z data field. (D) HIV-1 transmission to reporter cells from distinct blood myeloid DCs that had been pre-incubated with 10 µg/ml of the indicated mAbs for 30 min before viral exposure as in (A). Data show mean values and SEMs from three experiments including cells from twelve donors.

Figure 5. SIGLEC1 silencing blocks viral capture and trans-infection, while de novo expression of SIGLEC1 rescues it.

(A) Interference of SIGLEC1. Percentage of LPS mDCs positive for CD14, HLA-DR, Siglec-1, or VLP capture following mock transduction or transduction with nontarget or two different SIGLEC1-specific shRNAs. Data show mean values and SEMs from four experiments including cells from at least four donors. (B) Representative cell surface expression levels of CD14, HLA-DR, or Siglec-1 and VLP_HIV-Gag-eGFP capture profile of LPS mDCs transduced with nontarget shRNA (blue), SIGLEC1 target shRNA (red), or mock transduced (grey). (C) HIV-1 transmission to CD4⁺ reporter cells of LPS mDCs that had been mock-transduced or transduced with nontarget or SIGLEC1-specific shRNA. DCs were pulsed with HIV-1, washed, and co-cultured with reporter cells for 48 h. HIV-1 infection of CD4⁺ reporter cells was determined by induced luciferase activity in RLUs. Data show mean values and SEMs from two experiments including cells from four donors. (D) Transfection of Siglecs in Raji B cells (see also Figure S6 for transfection efficiencies). Capture of VLP_HIV-Gag-eGFP by Raji cells transfected with the indicated expression plasmids for Siglecs or mock transfected. Transfected Raji cells were pre-incubated with 10 µg/ml of the indicated mAbs and exposed to VLPs. See Figure S7 for blocking effect of sialyllactose. Data show mean values and SEMs from two experiments including cells from four transfections. Figure S8 shows results for Siglec transfections in HEK-293T. (E) Representative dot plots from Siglec-1, Siglec-5, and Siglec-7 transfected Raji cells pre-incubated with the indicated mAbs and subjected to VLP capture. (F) HIV-1 transmission from Raji cells transfected with the indicated expression plasmids for Siglecs to reporter CD4⁺ cells. Transfected cells were pre-incubated with the indicated mAbs as in (C) and then exposed to HIV-1. Data show mean values and SEMs from two experiments including cells from four transfections. Figure S9 depicts viral capture and transmission of viruses with or without the envelope glycoproteins.