ISG15 deficiency restricts HIV-1 infection

Abstract

Type I interferons (IFN-Is) are a group of potent inflammatory and antiviral cytokines. They induce IFN stimulated genes (ISGs), which act as proinflammatory mediators, antiviral effectors, and negative regulators of the IFN-I signaling cascade itself. One such regulator is interferon stimulated gene 15 (ISG15). Humans with complete ISG15 deficiency express persistently elevated levels of ISGs, and consequently, exhibit broad spectrum resistance to viral infection. Here, we demonstrate that IFN-I primed fibroblasts derived from ISG15-deficient individuals are more resistant to infection with single-cycle HIV-1 compared to healthy control fibroblasts. Complementation with both wild-type (WT) ISG15 and ISG15ΔGG (incapable of ISGylation while retaining negative regulation activity) was sufficient to reverse this phenotype, restoring susceptibility to infection to levels comparable to WT cells. Furthermore, CRISPR-edited ISG15ko primary CD4+ T cells were less susceptible to HIV-1 infection compared to cells treated with non-targeting controls. Transcriptome analysis of these CRISPR-edited ISG15ko primary CD4+ T cells recapitulated the ISG signatures of ISG15 deficient patients. Taken together, we document that the increased broad-spectrum viral resistance in ISG15-deficiency also extends to HIV-1 and is driven by a combination of T-cell-specific ISGs, with both known and unknown functions, predicted to target HIV-1 replication at multiple steps.

Fig 1. ISG15 deficient, IFN-primed cells are less susceptible to HIV-1 infection.

A: hTERT-immortalized fibroblasts were primed with type I IFN (IFNα2b, 1000 units/mL) for 12 hours, washed and rested for 36 hours followed by infection with HIV-1-VSV-Luc. Cells were lysed 48 hours after infection, upon which HIV-1-luciferase expression was measured. B: Serially diluted HIV-1 (0.1, 0.5, 2.5 and 12.5 μL) was used to infect WT and ISG15-/- cells with and without IFN priming. Infections were done using cell lines from 2 unrelated WT controls and 2 unrelated ISG15-/- patients. Representative experiments are shown. RLU: relative light units. C: The parental ISG15-/- cell lines were complemented with RFP or ISG15 and infected as described above. Complementation with ISG15 abolishes the restriction. All infections were conducted in triplicates. Error bars represent SD. Significance was determined using t-tests (Prism 7 software; **** denotes P<0.0001; ** denotes P<0.01).

Fig 2. ISG15 complementation restores HIV-1 susceptibility.

A, B,: hTERT-immortalized fibroblasts from healthy controls (A) and ISG15-deficient patients (B) were complemented with WT ISG15, ISGylation defective ISG15, or Luciferase RFP as a negative control. Cells stimulated with IFN-I overnight and rested for 36 hours were infected with HIV-1-VSV-GFP reporter virus. Flow cytometry was performed to quantify GFP-positive HIV-1-infected cells. Infectivity normalized to the parental cell-line is shown for mock primed (no IFN) and IFN primed cells. Error bars denote SD. (N = 4) **** denotes P<0.0001; *** denotes P<0.001; ** denotes P<0.01; * denotes P < 0.05 as determined by student’s two-tailed t-test. C, D: hTERTimmortalized fibroblasts from healthy controls (C) or ISG15 knockout healthy controls (D) were complemented with WT ISG15, ISGylation defective ISG15 (ISG15ΔGG), or Luciferase RFP as a negative control. Cells stimulated with IFN-I overnight and rested for 36 hours were infected with HIV-1-VSV-GFP reporter virus. Flow cytometry was performed to quantify GFP-positive HIV-1-infected cells. Infectivity normalized to the parental cell-line is shown for mock primed (no IFN-I) and IFN-I primed cells. Error bars denote SD. (N = 3) **** denotes P<0.0001; *** denotes P<0.001; ** denotes P<0.01; * denotes P<0.05 as determined by a student’s two-tailed T test.

Fig 3. ISG15-deficient primary CD4+ T cells are less susceptible to HIV-1 infection.

A: Primary CD4⁺ T cells were targeted with either non-targeting control (NTC gRNA) or ISG15 (ISG15 gRNA) CRISPR RNPs. On day 3, cells were treated with 0, 10, 100, or 1000 IU/mL IFNα2b for 24 hours. On day 4 (24 hours after IFN treatment) cells were infected with replication-competent HIV-1-GFP. Flow cytometry was performed on Day 10 (6 days post HIV-1 infection) to quantify GFP-positive HIV-1-infected cells. B: ISG15 knockout efficiency in primary human CD4⁺ T cells was assessed by Western blotting. Representative Western blot of CRISPR-targeted CD4⁺ T cells lysed 24 hours post IFNα2b treatment. C: Infection of IFNα2b-primed, CRISPR-targeted CD4⁺ T cells using HIV-1-GFP was performed in duplicate. Duplicates were averaged and percent infectivity was calculated relative to NTC-targeted non-IFN-I treated. Three donors are shown. GFP-positive cells were quantified 6 days post infection. Multiple, unpaired T tests were performed and P value is denoted in the figure. NS denotes T test result was non-significant.

Fig 4. RNA-seq analysis of CRISPR-Cas9 ISG15 knockout in primary CD4⁺ T cells reveals parallels to human ISG15 deficiency.

A: Differential gene expression heatmap between IFNβ stimulated (1000 IU/mL) CRISPR-Cas9 ISG15 knockout primary CD4⁺ T cells and non-targeting/non-edited controls shows elevated levels of interferon-stimulated genes (ISGs) in ISG15^ko primary CD4⁺ T cells. Represented genes are the union of differentially expressed genes identified during IFN_β stimulation of CRISPR-Cas9 ISG15^ko and IFNβ stimulation of the non-targeting controls, respectively. Differentially expressed genes (DEGs) were further filtered against a p-value of p<0.01. Flagged genes represent differentially expressed interferon-stimulated genes. YTHDF3 is not an ISG and is denoted by a star in the heatmap. B: Normalized count bar plots derived from RNA-seq data of select differentially expressed genes during IFN-b stimulation reveal elevated levels of ISGs in ISG15^ko primary CD4⁺ T cells relative to non-targeted guide control cells. Bars represent mean normalized count values. C: Gene set variation analysis of differentially expressed genes identified during IFN-b stimulation of CRISPR-Ca9 ISG15^ko primary CD4⁺ T cells confirms an IFN-I response signature in ISG15^ko primary CD4⁺ T cells. D: UMAP representation of unsupervised clustering results of single-cell RNA of 3,780 ISG15-/- patient (n = 2357 cells) and age-matched healthy control (n = 1423 cells) peripheral blood mononuclear cells. Points are colored by CD4⁺ T cell subset supervised annotation. E: Gene expression UMAPs of classical CD4⁺ T cell marker gene RNA transcripts in ISG15^-/- patient and age matched healthy control PBMCs. Point color intensity is scaled between the 1^st percentile (lo) and the 99^th percentile (hi) of natural log normalized gene expression count values. F: Gene set score distributions for IFN_α signaling in CD4⁺ T cell subsets. The HALLMARKS:IFNa_RESPONSE gene set was extracted from MSigDB Hallmark collection and used as input in Seurat’s AddModuleScore function. G: Average gene expression heatmap of the HALLMARKS: IFNa_RESPONSE gene set in ISG15^-/- patient and healthy control CD4⁺ T cell subsets. Heatmap color intensity represented log normalized average expression. Average expression value for each patient-cluster condition calculated from Seurat’s AverageExpression function.