Interferon-Inducible CD169/Siglec1 Attenuates Anti-HIV-1 Effects of Alpha Interferon

Abstract

A hallmark of human immunodeficiency virus type 1 (HIV-1) infection in vivo is chronic immune activation concomitant with type I interferon (IFN) production. Although type I IFN induces an antiviral state in many cell types, HIV-1 can replicate in vivo via mechanisms that have remained unclear. We have recently identified a type I IFN-inducible protein, CD169, as the HIV-1 attachment factor on dendritic cells (DCs) that can mediate robust infection of CD4⁺ T cells in trans Since CD169 expression on macrophages is also induced by type I IFN, we hypothesized that type I IFN-inducible CD169 could facilitate productive HIV-1 infection in myeloid cells in cis and CD4⁺ T cells in trans and thus offset antiviral effects of type I IFN. In support of this hypothesis, infection of HIV-1 or murine leukemia virus Env (MLV-Env)-pseudotyped HIV-1 particles was enhanced in IFN-α-treated THP-1 monocytoid cells, and this enhancement was primarily dependent on CD169-mediated enhancement at the virus entry step, a phenomenon phenocopied in HIV-1 infections of IFN-α-treated primary monocyte-derived macrophages (MDMs). Furthermore, expression of CD169, a marker of type I IFN-induced immune activation in vivo, was enhanced in lymph nodes from pigtailed macaques infected with simian immunodeficiency virus (SIV) carrying HIV-1 reverse transcriptase (RT-SHIV), compared to uninfected macaques, and interestingly, there was extensive colocalization of p27^gag and CD169, suggesting productive infection of CD169⁺ myeloid cells in vivo While cell-free HIV-1 infection of IFN-α-treated CD4⁺ T cells was robustly decreased, initiation of infection in trans via coculture with CD169⁺ IFN-α-treated DCs restored infection, suggesting that HIV-1 exploits CD169 in cis and in trans to attenuate a type I IFN-induced antiviral state.IMPORTANCE HIV-1 infection in humans causes immune activation characterized by elevated levels of proinflammatory cytokines, including type I interferons (IFN). Although type I IFN induces an antiviral state in many cell types in vitro, HIV-1 can replicate in vivo via mechanisms that have remained unclear. In this study, we tested the hypothesis that CD169, a type I IFN-inducible HIV-1 attachment factor, offsets antiviral effects of type I IFN. Infection of HIV-1 was rescued in IFN-α-treated myeloid cells via upregulation of CD169 and a subsequent increase in CD169-dependent virus entry. Furthermore, extensive colocalization of viral Gag and CD169 was observed in lymph nodes of infected pigtailed macaques, suggesting productive infection of CD169⁺ cells in vivo Treatment of dendritic cell (DC)-T cell cocultures with IFN-α upregulated CD169 expression on DCs and rescued HIV-1 infection of CD4⁺ T cells in trans, suggesting that HIV-1 exploits CD169 to attenuate type I IFN-induced restrictions.

Keywords: CD169; HIV; type I IFN; virus entry; virus evasion.

FIG 1

IFN-α treatment of THP-1 cells enhanced HIV-1 infection. (A) Cell surface expression of CD169 on THP-1 and THP-1_CCR5 cells left untreated (NT) or treated with 1,000 U/ml of IFN-α was measured by flow cytometry. (B) Western blot analysis of THP-1 cell lysates left untreated or treated with 1,000 U/ml of IFN-α. Mx2 expression was quantified and normalized to the expression of actin, a loading control. The Mx2/actin ratio of THP-1/IFN cells was further normalized to that of untreated THP-1 cells. (C to E) Infection of HIV-1 pseudotyped with VSV-G (C), replication-competent HIV-1 Lai/YU-2env-luc (D), or Lai-Env-pseudotyped HIV-1 in untreated or IFN-α-treated THP-1 cells (C and E) or THP-1_CCR5 cells (D). Cells were infected with increasing inputs of HIV-1, and replication was quantified 2 days postinfection by measuring luciferase activity in cell lysates. The experiment was performed three times in duplicate, and the data from a representative experiment are shown. (F) Infection of HIV-1 pseudotyped with VSV-G, HIV-1 Bal Env, or replication-competent HIV-1 Lai/YU-2env-luc in untreated THP-1_CCR5 and THP-1_CCR5/IFN cells. The luciferase activity in THP-1_CCR5/IFN cell lysates 2 days postinfection was normalized to that of THP-1_CCR5 cells. The data are the means ± SEMs from five (VSV-G), eight (HIV-1 Bal), or four (Lai/YU-2env) independent experiments performed in duplicate or triplicate. (G) Infection of HIV-1 pseudotyped with VSV-G, HIV-1 Lai Env, or MLV-A Env in untreated THP-1 and THP-1/IFN cells. The luciferase activity in THP-1/IFN cell lysates 2 days postinfection was normalized to that of THP-1 cells. The data are the means ± SEM from seven independent experiments performed in triplicate. (H) Infection of replication-competent HIV-1 Lai and HIV-1 NL4-3 in untreated THP-1 and THP-1/IFN cells at 2 days postinfection was quantified by intracellular p24^gag staining. The data are the means ± SEMs from four independent experiments. Two-tailed P values were calculated using one-sample t test (F and G) or paired t test (H) in GraphPad Prism 5. *, P ≤ 0.05; ***, P ≤ 0.001. ns, not significant.

FIG 2

Exposure of THP-1 cells to IFN-α enhances HIV-1 fusion. (A) Quantitative analysis of late reverse transcription products. Untreated or IFN-α-treated THP-1 (for VSV-G or HIV-1 Lai) or THP-1_CCR5 (for HIV-1 Bal) cells were infected with HIV-1 pseudotyped with VSV-G and replication-competent HIV-1 Lai and HIV-1 Lai/Balenv and lysed at 24 h postinfection for measurement of viral DNA by quantitative PCR. The amount of late reverse transcription products (A) or 2LTR circles (B) in IFN-α-treated THP-1 cells was normalized to that of untreated THP-1 cells. The data are the means ± SEMs from three independent experiments. (C) Binding of HIV-1 particles pseudotyped with VSV-G (n = 7, THP-1), HIV-1 Bal (n = 5, THP-1_CCR5), HIV-1 Lai (n = 7, THP-1), MLV-A (n = 7, THP-1), or replication-competent HIV-1 Lai/YU-2env (n = 3, THP-1_CCR5) to untreated THP-1 or THP-1/IFN cells was quantified by an ELISA, and the percent binding was calculated. The data are the means ± SEMs from independent experiments. (D) Virus fusion of HIV-1 pseudotyped with VSV-G or HIV-1 BalEnv in untreated THP-1_CCR5 or THP-1_CCR5/IFN cells was quantified. Each symbol represents the percentage of BlaM⁺ cells obtained from an independent experiment, and the means ± SEMs are shown. (E) Virus fusion of HIV-1 pseudotyped with VSV-G, MLV-A Env, or HIV-1 Lai Env in untreated THP-1 or THP-1/IFN cells was quantified. Each symbol represents the percentage of BlaM⁺ cells obtained from an independent experiment, and the means ± SEMs are shown. (F) Virus fusion of replication-competent HIV-1 Lai/YU-2env (YU-2) in untreated THP-1_CCR5 or THP-1_CCR5/IFN or HIV-1 Lai and HIV-1 NL4-3 in untreated THP-1 or THP-1/IFN cells was quantified. Each symbol represents the percentage of BlaM⁺ cells obtained from an independent experiment, and the means ± SEMs are shown. Two-tailed P values were calculated using one-sample t test (A and B) or paired t test (C to F) in GraphPad Prism 5. *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001.

FIG 3

IFN-α-induced CD169 expression on THP-1 cells enhances HIV-1 fusion and infection. (A to E) Untreated THP-1 or THP-1_CCR5 or IFN-treated THP-1 or THP-1_CCR5 cells were incubated with isotype-matched control IgG (IgG) or anti-CD169 blocking antibody and infected with VSV-G-pseudotyped HIV-1 (VSV-G) (A), Bal Env-pseudotyped HIV-1 (B), replication-competent HIV-1 Lai/YU-2env (YU-2) (C), Lai Env-pseudotyped HIV-1 (Lai) (D), or MLV-A Env-pseudotyped HIV-1 (E), and virus fusion in these cells was quantified. Each symbol represents the percentage of BlaM⁺ cells obtained from an independent experiment, and the means ± SEMs are shown. (F to J) Untreated THP-1 or THP-1_CCR5 or IFN-treated THP-1 or THP-1_CCR5 cells were infected with VSV-G-pseudotyped HIV-1 (F), Bal Env-pseudotyped HIV-1 (G), replication-competent HIV-1 Lai/YU-2env (H), Lai Env-pseudotyped HIV-1 (I), or MLV-A Env-pseudotyped HIV-1 (J) in the presence of either anti-CD169 blocking antibody or isotype-matched control IgG. Virus infection was quantified by measuring luciferase activity in cell lysates 2 days postinfection and was normalized to the values in untreated cells (no IFN-α). The data are the means ± SEMs from independent experiments performed in duplicate or triplicate (n = 4 for panel F, n = 5 for panel G, n = 4 for panel H, n = 6 for panel I, and n = 6 for panel J). (K) Infection of HIV-1 pseudotyped with HIV-1 Lai Env produced from untreated cells (NT; GSL-enriched virions) or PDMP-treated cells (PDMP; GSL-deficient virions) in THP-1 and THP-1/IFN cells. The luciferase activity in cell lysates 2 days postinfection was normalized to that of untreated THP-1 cells. The data are the means ± SDs from two independent experiments performed in triplicate. Two-tailed P values were calculated using one-way analysis of variance (ANOVA) followed by the Tukey-Kramer posttest (A to J, shown with lines) or the Dunnett's posttest (F to J, shown just above each column, compared to untreated cells [first column]) in GraphPad Prism 5. *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001.

FIG 4

Exogenous CD169 expression in U87 attenuates the effects of ISGs. (A) Cell surface expression of CD169 on parental U87 cells (U87) and CD169-transduced U87 cells (U87/CD169). (B) Western blot analysis of U87 cell or U87/CD169 cell lysates left untreated or treated with 1,000 U/ml of IFN-α. Mx2 expression was quantified and normalized to the expression of actin, a loading control. The Mx2/actin ratio of IFN-α-treated U87 or U87/CD169 cells was further normalized to that of untreated cells. (C) Infection of HIV-1 pseudotyped with HIV-1 Bal Env in untreated U87, IFN-α-treated U87, or IFN-α-treated U87/CD169 cells was quantified 2 days postinfection by measuring luciferase activity in cell lysates. The data are the means ± SDs from two independent experiments performed in triplicate.

FIG 5

Induced CD169 expression on MDMs enhances HIV-1 replication in the presence of IFN-α. (A) Cell surface expression of CD169 on MDMs treated with different doses of IFN-α. (B) Representative FACS profiles of BlaM⁺ MDMs pretreated with 100 U/ml IFN-α for 2 days and infected with BlaM-Vpr containing HIV-1 Lai/YU-2env in the absence or presence of anti-CD169 blocking antibody or isotype-matched IgG. (C and D) The percentage of BlaM⁺ MDMs from multiple infections with HIV-1 Lai/Balenv (C) or HIV-1 Lai/YU-2env (D) and the means ± SEM are shown. Each symbol represents data obtained from cells derived from an independent donor. (E and F) MDMs were pretreated with the indicated doses of IFN-α for 2 days, left uninfected or infected with SIV_mac239 Vpx-containing VLPs, and then incubated with isotype-matched IgG (IgG) or anti-CD169 blocking antibody prior to infection with HIV-1 Lai/Balenv-luc (E) or HIV-1 Lai/YU-2env-luc (F) reporter virus. HIV-1 infection was quantified 3 days postinfection by measuring luciferase activity in cell lysates, and values were normalized to those from untreated MDMs with isotype-matched IgG. The data are the means ± SEMs from six (E) and five (F) independent experiments performed using cells from different donors. Two-tailed P values were calculated using one-way ANOVA followed by the Tukey-Kramer posttest (C and D) or paired t test (E and F) in GraphPad Prism 5. *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001.

FIG 6

CD169⁺ cells in LNs were infected with SHIV in the presence of type I IFN signatures in vivo. (A) Quantification of IP-10 mRNAs in mesenteric LNs. RNA was extracted from frozen tissues which were removed from uninfected (n = 5) or SHIV-infected (n = 5) macaques upon necropsy, and IP-10 mRNA was quantitated by custom digital molecular barcoding using the NanoString nCounter system. Two-tailed P values were calculated using unpaired t test in GraphPad Prism 5. *, P ≤ 0.05. (B) Representative images of immunofluorescence staining of mesenteric LNs of uninfected and SHIV-infected macaques. Frozen tissues were stained for CD169 (green), p27^gag (red), and nuclei (blue). Scale bar = 100 μm. (C and D) Quantitation of CD169⁺ cells and SHIV-infected cells (p27^gag+) in LNs of uninfected (n = 3) and SHIV-infected (n = 4) macaques. CD169⁺ cells and p27^gag+ cells were counted by automated imaging software for 4 to 7 fields per LN. The percentage of CD169⁺ cells in total cells (C) or p27^gag+ cells and doubly positive (CD169⁺ p27^gag+) cells in total cells (D) was calculated, and the means ± SEMs are shown.

FIG 7

CD169 on mature DCs attenuates the effects of ISGs in trans-infection. (A and B) CD4⁺ T cells left untreated (CF) or treated with IFN-α (CF IFN-α) were infected with the indicated doses of X4-tropic (Lai-luc) (A) or R5-tropic (B) HIV-1 (Lai/Balenv-luc) reporter viruses. Alternatively, IFN-DCs were pulsed with the indicated amounts of X4-tropic or R5-tropic viruses, washed, and cocultured with IFN-α-treated CD4⁺ T cells (DC-T IFN-α). Infection in CD4⁺ T cells was quantified at 3 days postinfection by measuring luciferase activity in cell lysates. The data are the means ± SEMs from independent experiments performed in triplicates using cells from three (A) or four (B) different donors. (C) CD4⁺ T cells with or without IFN-α pretreatment were exposed to cell-free HIV-1 Lai-luc as for panel A. Alternatively, IFN-DCs were treated with isotype-matched IgG or anti-CD169 blocking antibody prior to incubation with HIV-1, washed, and added to IFN-α-treated CD4⁺ T cells. Infection in CD4⁺ T cells was quantified at 3 days postinfection by measuring luciferase activity in cell lysates. The data are the means ± SEMs, and each symbol represents data obtained from cells derived from an independent donor. (D) CD4⁺ T cells were infected with cell-free virus or IFN-DC-laden virus as for panel C except that HIV-1 Lai-luc containing BlaM-Vpr was used. HIV-1 fusion in CD4⁺ T cells was quantified at 4 h postinfection. The data are the means ± SEMs, and each symbol represents data obtained from cells derived from an independent donor. Two-tailed P values were calculated using one-sample t test (A and B) or one-way ANOVA followed by the Tukey-Kramer posttest (C and D) in GraphPad Prism 5. *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001.