Alternate receptor usage of neuropilin-1 and glucose transporter protein 1 by the human T cell leukemia virus type 1

Abstract

Recent studies have demonstrated that neuropilin 1 (NP-1) is involved in HTLV-1 entry; however, the role NP-1 plays in this process is not understood. We demonstrated that ectopic expression of human NP-1 but not NP-2 cDNA increased susceptibility to HTLV-1. SiRNA-mediated inhibition of NP-1 expression correlated with significant reduction of HTLV-1 Env-mediated fusion. The vascular endothelial growth factor (VEGF(165)) caused downmodulation of surface NP-1 and inhibited HTLV-1 infection of U87 cells. In contrast, VEGF(165) partially inhibited infection of primary astrocytes and had no significant effect on infection of HeLa cells. VEGF(165) and antibodies to the glucose transporter protein 1 (anti-GLUT-1) were both needed to block infection of primary astrocytes, however, only anti-GLUT-1 antibodies were sufficient to block infection of HeLa cells. HTLV-1 Env forms complexes with both NP-1 and GLUT-1 in primary human astrocytes. The alternate usage of these two cellular receptors may have important implications regarding HTLV-1 neuro-tropism.

Figure 1. Expression of NP-1 and GLUT-1 in primary human astrocytes and cell lines

A&B) RT-PCR analysis of NP-1 and GLUT-1 mRNA expression was performed as described in text. Amplification of β-actin was used as a control for gel loading. The data represent scans of agarose gels stained with ethidium bromide. C) FACS analysis of surface NP-1 and GLUT-1 in cell lines and primary human astrocytes. Polyclonal antibodies were used to analyze NP-1 expression while expression of GLUT-1 was analyzed using GLUT-IgY antibodies. Staining of CHO was performed to provide evidence for the low expression levels of GLUT-1 and NP-1 proteins.

Figure 2. Human Neuropilin 1 (NP-1) acts as a fusion receptor for HTLV-1

CHO cells were transfected with pCNDA3 plasmids encoding either vector, rat NP-1, human NP-1, human NP-2, GLUT-3 or GLUT1 under T7 promoter. Expression of the proteins was accomplished by infection with vaccinia virus encoding T7 RNA polymerase. The CHO cells expressing the different proteins were mixed with HTLV-1 Env-expressing 293 cells that co-express T7-lacZ reporter. Following incubation the extent of Env-mediated cell fusion was quantified by measuring β-galactosidase production. The competence of the transfected cells to undergo cell fusion was verified by VSV fusion (A). The ability of the transfected cells to fuse with the HTLV-1 Env63 is shown in panel B. Expression of rat and human NP-1 was verified by western blotting using a polyclonal antibody against the human NP-1 that cross-reacts with rat NP-1 but not with human NP-2 (C). The results were reproduced at least four times. The control envelope Unc63 is a cleavage mutant of HTLV-1 Env63 that does not promote cell fusion.

Figure 3. SiRNA-mediated reduction of NP-1 expression correlates with significant blocking of HTLV-1 Env-mediated fusion

The siRNA molecules were transfected into U87 cells and the reduction of gene expression was verified by western blot analysis (A). The GLUT-1 gel was exposed 5 times the exposure time of the NP-1 gel (A). The transfected cells were infected with vaccinia recombinant encoding T7 RNA polymerase and mixed with Env-expressing 293 cells expressing T7-lacZ reporter. The potency of the transfected cells to promote HTLV-1 Env-mediated fusion was examined by measuring the amount of β-galactosidase produced (B). VSV Env-mediated fusion was used as a positive control to verify the competence of the transfected cells to undergo cell fusion (C). The numbers 1, 2 and 3 below the lanes correspond to the transfected control, NP-1 or GLUT-1 siRNA respectively. The GLUT-1 gel images shown in panel A was over-exposed to visualize the siRNA effect in U87 cells that express low levels of GLUT-1.

Figure 4. Down-modulation of human NP-1 by VEGF₁₆₅

A) Schematic diagram illustrating the different domains of the human NP-1, the binding site of VEGF₁₆₅, and the anti-NP-1 antibodies used in this analysis. The mouse anti-human BDCA-4 (Neuropilin-1) monoclonal antibody recognizes the a1 domain of the human NP-1. B&C) Flow cytometry analysis of surface expression of the human NP-1 on U87 cells, pre-incubated at 37°C for 30 min in the absence of VEGF₁₆₅ (B) or in the presence of VEGF₁₆₅ (C). Panel D shows another experiment where surface expression of the human NP-1 was analyzed in the presence of VEGF₁₆₅. Following treatment with VEGF₁₆₅ the U87 cells were divided into two aliquots, one was directly stained with anti-NP-1 (Non-permeabilized); the other was treated with 0.3% Triton-X100 for 20 min at 4°C then stained with anti-NP-1 antibodies (Permeabilized).

Figure 5. VEGF₁₆₅ blocks HTLV-1 infection of U87 cells and primary human astrocytes

To examine NP-1-mediated infection the commercially available VEGF₁₆₅ or IL-2 were added to the indicated cell populations at escalating concentrations and incubated for 1 hr at 37°C before infection with the pseudotyped HTLV-1 virus (A&B). To examine the effect of GLUT-1-mediated infection other portions of the same cell populations were incubated with escalating doses of either GLUT-IgY or control IgY antibodies before HTLV-1 infection (C&D). Seventy-two hours post-infection, the cells were lysed and infectivity measured by the amount of luciferase produced (RLU). Unc63 is a control with a mutated envelope protein that is defective in membrane fusion. IL-2 was used as a negative control for the VEGF₁₆₅ activity. The control for GLUT-IgY activity was the normal chicken IgY. This experiment was reproduced at least three times.

Figure 6. HTLV-1 Infection of primary human astrocytes is sensitive to both VEGF₁₆₅ and anti-GLUT-1 antibodies

GLUT-IgY antibodies, VEGF₁₆₅, or both were pre-incubated with primary astrocytes (A), U87 cells (B), or HeLa cells (C) before infection with pseudotyped HTLV-1. The effect on HTLV-1 infection was examined 48 hr post-infection by measuring luciferase production (RLU). Normal chicken IgY was used as a negative control. The results are from one experiment performed in duplicates and are representative of at least three different experiments.

Figure 7. Co-immunoprecipitation of HTLV-1 Env by antibodies to human NP-1

Primary human astrocytes infected with either the WR vaccinia control or with a recombinant vaccinia virus encoding the HTLV-1 Env63 were used to prepare cellular lysates for immunoprecipitation with Anti-NP-1 antibodies. The precipitated proteins were fractionated in 10% SDS-PAGE and blotted. The blot was first probed with anti-NP-1, stripped, reprobed with anti-gp46 (anti-Env), stripped, and then re-probed with anti-GLUT-1 antibodies. Arrows at the right side mark the positions of the different proteins. The positions of the molecular weight markers are indicated at the left side.