Identification of Interferon-Stimulated Genes with Antiretroviral Activity

Abstract

Interferons (IFNs) exert their anti-viral effects by inducing the expression of hundreds of IFN-stimulated genes (ISGs). The activity of known ISGs is insufficient to account for the antiretroviral effects of IFN, suggesting that ISGs with antiretroviral activity are yet to be described. We constructed an arrayed library of ISGs from rhesus macaques and tested the ability of hundreds of individual macaque and human ISGs to inhibit early and late replication steps for 11 members of the retroviridae from various host species. These screens uncovered numerous ISGs with antiretroviral activity at both the early and late stages of virus replication. Detailed analyses of two antiretroviral ISGs indicate that indoleamine 2,3-dioxygenase 1 (IDO1) can inhibit retroviral replication by metabolite depletion while tripartite motif-56 (TRIM56) accentuates ISG induction by IFNα and inhibits the expression of late HIV-1 genes. Overall, these studies reveal numerous host proteins that mediate the antiretroviral activity of IFNs.

Graphical abstract

Figure 1

ISG Screening Strategies and Effects of ISGs on HIV-1 and HIV-2 (A) Diagrammatic representations of genes present in the ISG libraries. (B) Schematic representation of the pSCRPSY lentiviral vector. D, splice donor; A, splice acceptor. See Supplemental Experimental Procedures for details. (C and D) Screening strategy to identify candidate genes that protect cells from incoming retroviral infection (B) or reduce infectious virion yield from infected cells (C). (E and F) Effect of ISGs in incoming screens with HIV-1 (D) and HIV-2 (E) in MT4 and THP-1 cells and in production screens with HIV-1 (D) and HIV-2 (E) in 293T cells. Black data points, human ISGs; blue data points, macaque ISGs. Hits that appear twice represent variant transcripts. All values were normalized to the screen average (100 a.u.).

Figure 2

Effects of Human and Macaque ISGs on Diverse Retroviruses (A–F) The effect of ISGs on lentiviruses, FIV (A) and EIAV (B); betaretroviruses, MPMV (C) and HERV-K (D); gammaretroviruses, MLV (E) and CERV2/MLV (F); and a spumaretrovirus, PFV (G). Screen details are as in Figures 1E and 1F except that the indicated cell lines were used.

Figure 3

Putatively Indirect and Direct Effects of ISGs on Incoming HIV-1 and HIV-2 Infection (A) Activation of integrated IFNβ promoter driven luciferase reporter genes in HEK293T and THP-1 cells following transduction with lentiviral vectors encoding macaque and human ISGs. (B) Activation of integrated ISRE-driven GFP reporter gene in HEK293T cells following transduction with lentiviral vectors encoding macaque and human ISGs. (C and D) Infectious titers of HIV-1-GFP and HIV-2-GFP in MT4 cells (C) and THP-1 cells (D) transduced with SCRPSY vectors expressing selected ISGs. Titers are mean + SD.

Figure 4

IDO1 Inhibits HIV-1 and HIV-2 Replication (A and B) FACS plots (A) and quantitation of mean fluorescence intensity (MFI, B) depicting GFP expression following standard or low-dose HIV-2-GFP infection of MT4 cells transduced with vectors expressing no ISG or IDO1. (C) Western blot (WB) analysis of IDO1 expression with or without induction by IFNγ (A549 cells) or doxycycline (Dox) (MT4-IDO1 cells). (D) Infectious titer of HIV-1-GFP (NHG) or HIV-2-GFP in target MT4 cells containing Dox-inducible IDO1, rhT5α, or TagRFP. (E) HIV-1-GFP (NHG) replication in MT4 cells containing Dox inducible IDO1, rhesus TRIM5α (rhT5α), or TagRFP. (F) Production of infectious HIV-1-GFP (NHG) or HIV-2 (VSV-G pseudotyped ROD10) during a single cycle of replication in MT4 cells containing Dox inducible IDO1, rhT5α, or TagRFP. Where indicated (5×), 5-fold higher input was used to negate the “incoming” effect. (G) WB analysis of Gag expression in the same HIV-1 (NHG) infected cells as in (F). (H) Infectious titer of HIV-1 produced and WB analysis of Gag, Env, and Nef expression in MT4 cells containing Dox-inducible IDO1 during a single cycle of HIV-1 (NL4.3) replication. (I) Yield of infectious virions and cell lysate and particulate capsid during a single replication cycle of mixed (in the indicated ratios) unmodified MT4 cells and MT4 cells containing Dox-inducible IDO1. (J and K) Effects of 1-MT (J) or L-Trp (K) on the yield of infectious HIV-1 particles and WB analysis of expression and release of viral Gag proteins during a single cycle of HIV-1 replication in MT4 cells containing Dox-inducible IDO1. Titers are mean + SD.

Figure 5

Endogenous IDO1 Inhibits HIV-1 (A) WB analysis of IDO1 expression in IFNγ-treated A549 cells. (B) Titer of VSV-G pseudotyped replication competent HIV-1-GFP (NHG) on A549 cells following treatment with increasing doses of IFNγ. (C) The yield of infectious progeny virions from a single cycle of HIV-1 replication in A549 cells treated with increasing doses of IFNγ. (D) The yield of infectious progeny virions from a single cycle of HIV-1 replication in A549 cells treated with IFNγ in the presence of 50 μg/ml L-Trp (D and E) or 100 μg/ml 1-MT (D and F). (E) WB analysis of IDO1 expression in IFNγ-treated TZM-bl cells. (F) As in D using TZM-bl cells. Titers are mean + SD.

Figure 6

TRIM56 Overexpression Can Inhibit HIV-1 Replication (A and B) HIV-1 replication in populations of GHOST cells transduced with lentiviral vectors (CCIB) expressing the indicated human (A) or macaque (B) ISGs. (C) WB analysis of GHOST cell clones overexpressing TRIM56. (D) Single round HIV-1 infection of GHOST cells transduced with empty vector (CCIB, filled symbol) or clones overexpressing TRIM56. (E) Spreading replication of HIV-1 in GHOST cells transduced with empty vector (CCIB, red symbol) or the clones overexpressing TRIM56 described in (C). (F) WB analysis of Gag, Nef, Env, and TRIM56 expression and particle release in GHOST vector and GHOST-TRIM56#2 cells during a single cycle of HIV-1 infection. (G) The yield of infectious progeny virions from a single cycle of HIV-1 replication in GHOST-vector and GHOST-TRIM56#2 cells. (H) Adaptation of HIV-1 to GHOST-TRIM56#2 cells via spreading replication. At each discontinuity, virions harvested from the GHOST-TRIM56#2 cells were passaged onto GHOST-vector and GHOST-TRIM56#2 cells. (I) Schematic representation and sequence analysis of an HIV-1 clone (TRIM56R) adapted to replicate in GHOST cells expressing TRIM56. (J) WB analysis of Gag expression and particle release in GHOST vector and GHOST-TRIM56 cells, during a single cycle of HIV-1 and HIV-1/TRIM56R replication.

Figure 7

Multiple Viral Determinants and ISGs Underlie Inhibition of HIV-1 Replication by Endogenous TRIM56 (A and B) Spreading replication of HIV-1 and HIV-1/TRIM56R in GHOSTX4-vector and GHOSTX4-TRIM56#2 cells (A), or chimeric viruses constructed using HIV-1 and HIV-1/TRIM56R (B). (C) WB analysis of TRIM56 expression in clones of MT4-LTR-GFP cells transduced with lentiviral vectors expressing Cas9 only or Cas9 plus one of two TRIM56-targeted guide RNAs (CR1 or CR3). (D and E) HIV-1 replication in the absence (D) or presence (E) of 25U/ml IFNα in clones of control (n = 4) or TRIM56 knockout (n = 7) MT4-LTR-GFP cells described in (C). (F and G) Microarray analysis of the expression levels (in a.u.) for the ∼120 genes most highly induced by 25U/ml IFNα in clones of control (n = 4) or TRIM56 knockout (n = 7) MT4-LTR-GFP cells described in (C) in the presence (G) or absence (F) of IFN.