Upregulation of Neuropilin-1 Inhibits HTLV-1 Infection

Abstract

Infection with human T-cell leukemia virus type 1 (HTLV-1) can produce a spectrum of pathological effects ranging from inflammatory disorders to leukemia. In vivo, HTLV-1 predominantly infects CD4⁺ T-cells. Infectious spread within this population involves the transfer of HTLV-1 virus particles from infected cells to target cells only upon cell-to-cell contact. The viral protein, HBZ, was found to enhance HTLV-1 infection through transcriptional activation of ICAM1 and MYOF, two genes that facilitate viral infection. In this study, we found that HBZ upregulates the transcription of COL4A1, GEM, and NRP1. COL4A1 and GEM are genes involved in viral infection, while NRP1, which encodes neuropilin 1 (Nrp1), serves as an HTLV-1 receptor on target cells but has no reported function on HTLV-1-infected cells. With a focus on Nrp1, cumulative results from chromatin immunoprecipitation assays and analyses of HBZ mutants support a model in which HBZ upregulates NRP1 transcription by augmenting recruitment of Jun proteins to an enhancer downstream of the gene. Results from in vitro infection assays demonstrate that Nrp1 expressed on HTLV-1-infected cells inhibits viral infection. Nrp1 was found to be incorporated into HTLV-1 virions, and deletion of its ectodomain removed the inhibitory effect. These results suggest that inhibition of HTLV-1 infection by Nrp1 is caused by the ectodomain of Nrp1 extended from virus particles, which may inhibit the binding of virus particles to target cells. While HBZ has been found to enhance HTLV-1 infection using cell-based models, there may be certain circumstances in which activation of Nrp1 expression negatively impacts viral infection, which is discussed.

Keywords: HBZ; HTLV-1; Nrp1; bZIP; infection; p300/CBP; transcription.

Figure 1

HBZ upregulates genes involved in HTLV-1 infection. (A) Relative COL4A1 mRNA levels in HeLa clonal cell lines expressing wild-type HBZ (HBZ) or carrying the empty expression vector (pcDNA). The graph shows qRT-PCR results averaged from four independent experiments. (B) Relative COL4A2 mRNA levels in HeLa clonal cell lines expressing wild-type HBZ (HBZ) or carrying the empty expression vector (pcDNA). The graph shows qRT-PCR results averaged from six independent experiments. (C) Relative GEM mRNA levels in HeLa clonal cell lines expressing wild-type HBZ (HBZ) or carrying an empty expression vector (pcDNA). The graph shows qRT-PCR results average from eight independent experiments. (D) Relative NRP1 mRNA levels in HeLa clonal cell lines expressing wild-type HBZ (HBZ) or carrying an empty expression vector (pcDNA). The graph shows qRT-PCR results average from five independent experiments. For all graphs, HBZ values are normalized to that of the empty vector (set to 1), and error bars show standard deviations; * p < 0.05; ** p < 0.01. (E) Nrp1 expression in empty vector (pcDNA) and HBZ-HeLa clones. Whole-cell extracts (40 μg for Nrp1 and β-actin, 20 μg for His and β-actin) were analyzed via Western blot using antibodies against Nrp1, HBZ (6xHis tag), and β-actin.

Figure 2

HBZ upregulates NRP1 expression. (A) Deletion of HBZ in ST1 and KK1 ATL-derived cells reduces NRP1 expression. The graph was generated from published microarray data (GEO accession number GSE94409 [52]) and shows the percent reduction in NRP1 transcript levels after inducing CRISPR/Cas9-mediated knockout of HBZ in the ATL-derived cell lines, ST1 and KK1, using two different guide RNAs (sgHBZ_1 and _2). Values are from day 8 post-induction except for sgHBZ_2 in KK1, which is the day 7 value (no day 8 data were provided for this specimen). Data were obtained using GEO2R with calculations based on averaged values from the four array features probing for different regions of the NRP1 transcript. (B) Nrp1 expression in non-infected activated CD4⁺ T-lymphocytes (aCD4) and T-cell lines. Whole-cell extracts (45 μg for Nrp1 and β-actin, 50 μg for Tax and β-actin) were analyzed via Western blot using antibodies against Nrp1, Tax, and β-actin. (C) Nrp1 expression on the cell surface of T-cell lines. Jurkat and ATL-2 cells were labeled with an Nrp1 antibody, fixed, and analyzed using flow cytometry. Histograms are representative of three independent experiments and show relative cell surface labeling as follows: unlabeled cells (CT, light grey) and Nrp1 antibody (dark gray). (D) Relative NRP1 mRNA levels in HTLV-1-immortalized human T-cell lines recently established from peripheral blood lymphocytes (PBL). The graph shows qRT-PCR results averaged from three separate RNA extractions. Values were normalized to those for activated CD4⁺ T-cells (set to 1). Error bars represent standard deviations.

Figure 3

HBZ increases c-Jun and JunB recruitment to an enhancer downstream of the NRP1 gene. (A) HBZ associates with a chromosomal site (enhancer peak) approximately 200kb downstream of the NRP1 transcription start site (indicated by the bent arrow). Peaks of enrichment for HBZ, H3K27ac, and IgG (negative control) at the NRP1 locus in KK1 cells are shown in the IGV Browser. Genomic coordinates are based on the NCBI36/hg18 assembly. Data were obtained from published ChIP-Seq data sets (GEO accession number GSE94732 [52]). (B) HBZ binds to the enhancer region in SLB-1 cells. The graph shows levels of HBZ enrichment at the off-target control site and the enhancer region averaged from four independent ChIP assays using SLB-1 cells transduced to express HBZ with a C-terminal 6xHis tag. (C) c-Jun and JunB are enriched at the enhancer region in ATL-2 cells. The graph shows average levels of factor enrichment at the off-target control site, the enhancer region, and the AP-1 site in the WEE1 promoter (WEE1-AP1). Data are from four (c-Jun) and three (JunB and MafG) independent ChIP assays. (D) Relative NRP1 mRNA levels in HeLa clonal cell lines expressing wild-type HBZ (HBZ-WT), the activation domain mutant (HBZ-MutAD), the leucine zipper domain mutant (HBZ-MutZIP), the translational-defective mutant (HBZ-ΔATG), or carrying the empty expression vector (pcDNA). The graph shows qRT-PCR results average from five independent experiments, with values normalized to that for pcDNA (set to 1). (E) HBZ binds to the enhancer region in HeLa cells. The graph shows levels of HBZ enrichment at the off-target control site and the enhancer region averaged from three independent ChIP assays using HeLa cells expressing HBZ or carrying the empty vector (pcDNA). (F) JunB binds to the enhancer region in HeLa cells. The graph shows levels of HBZ enrichment at the off-target control site and the enhancer region averaged from three independent ChIP assays using HeLa cells expressing HBZ or carrying the empty vector (pcDNA). For all graphs, error bars show standard deviations; * p < 0.05, ** p < 0.01.

Figure 4

p300/CBP is recruited to the NRP1 enhancer. p300 (A) and CBP (B) bind the NRP1 enhancer region. Graphs show average values from three independent ChIP assays using empty vector (pcDNA) and HBZ-expressing HeLa cells. (C) siRNA-mediated depletion of p300 and CBP abrogates activation of NRP1 transcription by HBZ. HeLa clonal cell lines expressing wild-type HBZ (HBZ) or carrying an empty expression vector (pcDNA) were transfected with an siRNA pool targeting p300 and CBP or a non-targeting siRNA pool (Control). The graph shows qRT-PCR results averaged from four independent transfection experiments with values normalized to those for the empty-vector clone (pcDNA) transfected with the non-targeting siRNA pool (set to 1). (D) siRNA-mediated depletion of p300 and CBP. HeLa cells were transfected with an siRNA pool targeting p300 and CBP (p300/CBP) or a non-targeting siRNA pool (Control). Whole-cell extracts (15 μg for p300, 40 μg for CBP and β-actin) were analyzed via Western blot using antibodies against p300, CBP, and β-actin. Inhibition of p300/CBP KAT activity reduces NRP1 transcription in (E) an HTLV-1-infected T-cell line (ATL-2) and (F) an HTLV-1-immortalized primary human T-cell line (CJ4). Cells were treated with A485 (10 μM) or the carrier (DMSO) for 3 h. Graphs show qRT-PCR results averaged from four (ATL-2 cells) and two (CJ4 cells) independent experiments with A485 values normalized to those for DMSO (set to 1). For all graphs, error bars show standard deviations; * p < 0.05; ** p < 0.01.

Figure 5

NRP1 knockdown increases HTLV-1 infection. (A) The flow diagram shows the co-culture/infection assay procedure using HTLV-1-infected cells as donor cells and Jurkat-pminLUC-vCRE cells as target cells. (B) shRNA-mediated depletion of Nrp1 in MT-2 cells increases HTLV-1 infection. Jurkat-pminLUC-vCRE cells were co-cultured with MT-2 cells under puromycin selection following transduction with expression vectors for a negative control shRNA (shGFP) or an shRNA targeting the NRP1 transcript (shNRP1), or co-cultured with non-infectious C8166/45 cells. The graph shows luciferase values averaged from three replicates of a single experiment and is representative of three independent experiments. (C) shRNA-mediated depletion of Nrp1 in MT-2 cells does not affect levels of gp46 (SU) and Gag p19. Whole-cell extracts (50 μg for Nrp1 and β-actin; 15 μg for gp46 and Gag p19) were analyzed via Western blot using antibodies against Nrp1, gp46, Gag p19, and β-actin. (D) shRNA-mediated depletion of Nrp1 in MT-2 cells does not affect levels of cell-free virus. Levels of Gag p19 in clarified culture media were measured using ELISA. The graph shows values averaged from two independent transduction experiments with shNRP1 values normalized to those for shGFP (set to 1). (E) shRNA-mediated depletion of Nrp1 in ATL-2 cells increases HTLV-1 infection. Experiments were performed as described in (B) above. The graph shows luciferase values averaged from three replicates of a single experiment and is representative of two independent experiments. (F) shRNA-mediated depletion of Nrp1 in ATL-2 cells does not affect levels of gp46 (SU) and Gag p19. Western blots were conducted as described in (C) above. (G) shRNA-mediated depletion of Nrp1 in ATL-2 cells does not affect levels of cell-free virus. Experiments were performed as described in (D) above. The graph shows values averaged from three independent transduction experiments with shNRP1 values normalized to those for shGFP (set to 1). (H) The flow diagram shows the co-culture/infection assay procedure using Jurkat cells as donor cells and CHO-LFA-1 cells as target cells. (I) shRNA-mediated depletion of Nrp1 in Jurkat donor cells increases HTLV-1 infection. Jurkat cells were co-transfected with pcDNA3.1, pCRU5HT1-inLuc, and pSG-Tax (no infection, CT) or pCMVHT1, pCRU5HT1-inLuc, pSG-Tax, and the shGFP or shNRP1 vector, and cocultured with CHO-LFA1 cells. The graph shows luciferase values normalized to protein averaged from replicates from three independent electroporation assays. For all graphs, error bars show standard deviations; * p < 0.05, ** p < 0.01.

Figure 6

Overexpression of Nrp1 reduces infection. (A) The flow diagram shows the co-culture/infection assay procedure using HTLV-1-infected SLB-1 cells as donor cells and Jurkat-pminLUC-vCRE cells as target cells. SLB-1 cells were transduced with pLJM1-NRP1 (NRP1) or the pLJM1 empty vector (EV) and placed under puromycin selection. (B) Nrp1 expression in transduced SLB-1 cells. Whole-cell extracts (50 μg for Nrp1, Gag p19, and β-actin; 15 μg for gp46 and β-actin) were analyzed via Western blot using antibodies against Nrp1, gp46, Gag p19, and β-actin. (C) Increased expression of Nrp1 in SLB-1 cells decreases HTLV-1 infection. Jurkat-pminLUC-vCRE cells were co-cultured with SLB-1 cells transduced with pLJM1-NRP1 (NRP1) or the pLJM1 empty vector (EV), or co-cultured with non-infectious C8166/45 cells. The graph shows luciferase values averaged from three replicates of each infection condition from a single experiment and is representative of three independent experiments. Error bars show standard deviations; * p < 0.05, *** p < 0.001.

Figure 7

The ectodomain of Nrp1 is responsible for the inhibition of HTLV-1 infection. (A) Nrp1 is incorporated into HTLV-1 virus particles. Culture media from MT-2, SLB-1, and ATL-2 cells were filtered, ultracentrifuged, and analyzed via Western blot using antibodies against Nrp1 and gp46. (B) The schematic shows full-length Nrp1 and the truncation mutant, Nrp1-Δabc. (C) The flow diagram shows the co-culture/infection assay procedure using HEK293T cells. (D) HTLV-1 infection is not inhibited by an Nrp1 truncation mutant lacking the ectodomain. HEK293T cells were co-transfected with pCMVHT1M, pCRU5HT1-inLuc, and pQCXIP (EV), pQCXIP-NRP1, or pQCXIP-NRP1-Δabc. Luciferase assays were performed 48 h later. The graph shows luciferase values averaged from three independent experiments each performed in triplicate. Error bars show standard deviations; * p < 0.05; ** p < 0.01. (E) Nrp1 expression in transfected HEK293T cells. Whole-cell extracts (50 μg) were analyzed via Western blot using antibodies against Nrp1 (Flag-tagged) and β-actin.