Induction of IL-17 and nonclassical T-cell activation by HIV-Tat protein

Histopathological Findings in CNS-IRIS.

CNS-IRIS in the absence of an OI is rare and poorly characterized (3). However, examination of severe inflammation in the absence of an OI can elucidate key pathophysiological mechanisms. Therefore, we evaluated a brain biopsy from such an individual with CNS-IRIS. Magnetic resonance imaging showed diffuse high signal intensity lesions in the white matter (SI Appendix). The brain biopsy specimen showed mononuclear perivascular and scattered parenchymal inflammatory infiltrates in the white matter (Fig. 1A). The invading cells were primarily CD8+ T cells with occasional IL-17+ T cells (Fig. 1 A and B). Previously, elevated IL-17 levels have been detected in the plasma of patients with IRIS (12, 13). We also observed the presence of Tat in a large fraction of infiltrating mononuclear cells, whereas p24 antigen was undetectable (Fig. 1C). This finding suggested that Tat was produced within the CNS compartment despite successful ART.

Detection of Tat in CSF of ART-Treated Individuals.

To determine whether Tat was produced within the CNS during ART, we examined the CSF in a cohort of HIV-infected individuals on ART. Tat was detected in the CSF of three of eight individuals with undetectable (<50 RNA copies per mL) HIV viral load in the blood and CSF and in zero of five CSF samples from HIV negative individuals with pseudotumor cerebri (Fig. 1D). To confirm that Tat was produced during ART, we infected peripheral blood mononuclear cells (PBMCs) with HIV_SF162 and treated the cultures with the protease inhibitor, Darunavir. Cells showed an inhibition of viral replication (Fig. 1E) (P < 0.05) but Tat production was not impacted (Fig. 1F). These data provide evidence that Tat is produced and secreted in patients well controlled on ART and thereby may contribute to chronic inflammation even in the absence of viral replication.

Tat Induces Proinflammatory Gene Expression in T Cells.

To determine changes in gene transcription induced by extracellular Tat in uninfected cells, expression profiling was performed in human primary T cells exposed to Tat by microarray. One hundred gene fragments, representing 94 known genes, were identified as dysregulated in response to Tat (Fig. 2A and SI Appendix). Up-regulated genes included proinflammatory cytokines such as IL1B and IL8. Furthermore, pathway enrichment analysis showed significant dysregulation of IL-17 signaling (SI Appendix). Importantly, this dysregulation occurred without any changes in IL21 or IL23 (SI Appendix). These results were confirmed by immunoaffinity array. Specifically, higher levels of secreted proinflammatory cytokines from PBMCs were observed post Tat exposure including IL-17 (SI Appendix). Because IL-17 can propagate inflammation (14), we determined whether Tat could induce IL-17 secretion in an antigen-independent manner. Tat induced the secretion of both IL-17 (Fig. 2B) and granzyme B (SI Appendix) in a dose-dependent manner. Tat increased IL-17 transcription (Fig. 2C) and also increased the fraction of IL-17 positive T cells (Fig. 2D) predominantly IL-17+/CD4+ cells (SI Appendix). The specificity of the Tat response in primary human T cells was validated following transwell coculture with a Tat secreting, Jurkat–Tat cell line in the absence or presence of Tat neutralizing antibodies (Fig. 2E).

Mechanism of T-Cell Activation by Tat Is Antigen Independent.

It has been previously established that Tat is taken up by cells through endocytosis (9). To define the kinetics of Tat uptake and its subcellular localization, ImageStream analysis was performed on T cells exposed to fluorescently labeled Tat at different time intervals. Tat was rapidly taken up by cells and diffusely spread throughout the cytoplasm with ∼25% of endocytosed Tat entering the nucleus (Fig. 3 A and B). Inhibition of clathrin-mediated endocytosis by chlorpromazine or amiloride resulted in complete inhibition of Tat induced IL-17 induction, whereas the caveolae inhibitor filipin did not significantly inhibit T-cell activation (Fig. 3C and SI Appendix). To determine whether Tat was driving T-cell activation in an antigen-independent manner, T cells were treated with nocodazole (NDZ) to inhibit Lck phosphorylation and downstream signaling from the TCR. Interestingly, NDZ did not inhibit Tat-mediated T-cell activation, suggesting that no additional antigenic stimulus was required to activate T cells in the presence of Tat (Fig. 3D).

Tat-Mediated T-Cell Activation Is Dependent upon VEGFR2.

To elucidate the mechanism of TCR-independent T-cell activation by Tat, we determined the subcellular signaling pathway involved. Gene expression analysis and immunoaffinity array showed that IL-8 and IL-1β (Fig. 2A and SI Appendix) were strongly up-regulated by Tat. Both IL-8 and IL-1β activate VEGFR2 (KDR) gene expression (15). Ingenuity analysis of the VEGFR2 pathway showed an up-regulation of NME2 and PIM1, VEGFR2 responsive genes, suggesting that VEGFR2 was functionally active (SI Appendix). VEGFR2 is expressed on T cells constitutively at low levels (16, 17); however the function of VEGFR2 in lymphocytes is not well understood. To investigate the role of VEGFR2 in Tat induced T-cell activation, T cells were treated with the VEGFR2-specific inhibitor SU1498, which dose-dependently inhibited IL-17 production (Fig. 4A and SI Appendix). This inhibition was confirmed by siRNA-mediated knockdown of VEGFR2 (Fig. 4B and SI Appendix). T-cell activation by Tat was dependent on known VEGFR2 downstream signaling pathways including PI3 kinase, NF-κB, CDKs, and mTOR (Fig. 4C). This activation was independent of PKC and Kv1.3 channel (Fig. 4D), supporting VEGFR2 as the primary signaling pathway. None of the inhibitors alone affected T-cell activation or viability (SI Appendix).

Tat Induces Chromatin Remodeling.

The chromatin of resting T cells is densely packaged and heterochromatic, limiting active transcription (18). To assess the role of Tat in remodeling the epigenetic landscape of T cells, global levels of acetylated histone H3 (H3ac), a marker for euchromatin associated with actively transcribed genes, were measured. A profound increase of global H3ac levels in T cells treated with Tat was observed indicating that Tat changes the global chromatin structure in T cells (Fig. 5 A and B).

Virologic Consequence of Tat Activation of T Cells.

Classically, it has been thought that only activated T cells are capable of being infected (19). We therefore determined the effect of Tat on T-cell proliferation by measuring both thymidine incorporation and carboxyfluorescein succinimidyl ester (Fig. 5C and SI Appendix). Remarkably, no T-cell proliferation was observed. We further investigated whether Tat-induced T-cell activation may allow for enhanced proviral integration in the absence of proliferation.

To investigate this hypothesis, T cells were pretreated with Tat and then infected with HIV. Viral replication was determined by measuring p24 antigen in the supernatant (Fig. 5D). Prior exposure to exogenous Tat strongly increased HIV replication. This effect of Tat was blocked by the addition of neutralizing anti-Tat antibodies, confirming that Tat can increase HIV virulence in an autocrine and paracrine manner.

Cell Isolation, Culture, and Stimulation.

T cells were isolated from healthy donors using a Pan T-cell II isolation kit (Miltenyl Biotec) Primary cells, Jurkat E6.1, and Jurkat–Tat cells and were maintained in Iscove’s modified Dulbecco’s media with 10% (vol/vol) human serum supplemented with 1% (vol/vol) antibiotic/antimycotic solution at 37 °C with 5% (vol/vol) CO₂. A total of 5 × 10⁵ T cells were seeded challenged with no peptide (negative control), 200 nM Tat peptide, or 1 μg/mL phytohemagglutinin (PHA) (Invitrogen). Endotoxin-free recombinant Tat was produced and functionally characterized as described (38). The following pharmacological inhibitors were used for mapping of pathways: Tosyl phenylalanyl chloromethyl ketone (Sigma, 10 µM), RO-32–0432 (Sigma, 100 nM), Margatoxin (Santa Cruz, 100 nM), Genistein (Sigma, 100 µM), LY294002 (Cell Signaling, 100 µM), SNS-032 (Selleck, 0.5 µM), Rapamycin (Selleck, 5 nM), Nocodazole (Sigma, 5 mg/mL), SU 1498 (Invitrogen, 2 µM), Chlorpromazine (Sigma, 30 µM), Amiloride (Sigma catalog, 500 µM), and Filipin (Sigma, 3 µg/mL).

Microarray.

Sample preparation, hybridization, and CEL file creation were performed using Human GeneChip 1.0 ST Array (Affymetrix) following manufacturer’s guidelines. Conversion of CEL files to normalized gene fragment expression data were accomplished using Affymetrix’s Expression Console with the “RMA Sketch” option. Statistical testing of the data by paired t test was accomplished in R. Gene fragments with a test P value < 0.05 and an absolute difference of means ≥ 1.25 fold were deemed differentially expressed. Subsequent gene fragment annotation, network analysis and pathway enrichment analysis was accomplished using IPA.

Flow Cytometry and Cell Imaging.

T cells treated with Tat (200 nM) or unstimulated were fixed and permeabilized before staining with anti–IL-17–PerCPCy5.5 (Biolegend, 1 μL per 10⁶ cells). Analysis was performed on a BD Biosciences LSR II at the Flow Cytometry Core Facility, Johns Hopkins Medical Institute. Flow cytometry coupled with confocal imaging was performed on the Amnis Image Stream Flow Cytometer at the National Institute of Neurological Disorders and Stroke Flow Cytometry Core Facility, National Institutes of Health. For Tat uptake studies, T cells were treated with 200 nM Tat–Venus.

Protein Arrays.

PBMCs were isolated and 1 × 10⁷ cells were untreated or stimulated with Tat for 24 or 72 h. After incubation, supernatants were analyzed by cytokine array (R&D Systems, Human cytokine array, Panel A). Equivalent amounts of protein were applied to the arrays. The intensity of the signal was visualized and interpreted as present (+) or absent (−).

Coculture Stimulation of T Cells.

A total of 1 × 10⁶ T cells were plated in 500 μL in the bottom of a 24-well transwell plate and pretreated for one hour with either anti-Tat polyclonal antibody (Abcam, Ab43014) or an antibody isotype. A total of 1 × 10⁵ Jurkat or Jurkat–Tat cells were plated in 250 μL in the transwell insert. Cells were incubated for 72 h, and supernatants were assayed for granzyme B.

ELISAs.

Concentration of granzyme B (Cell Sciences), IL-17 A/F kit (Cell Sciences), p24 (ZeptoMetrix Retrotek), histone H3 (Abcam), and acetylated histone H3 (Abcam) were measured using ELISA kits according to manufactures instructions. For quantification of Tat by ELISA monoclonal anti-Tat (1D9, 1:500) was used as the capture antibody. This reagent was obtained through the AIDS Research and Reference Reagent Program, Division of AIDS, National Institute of Allergy and Infectious Diseases, National Institutes of Health: HIV-1 Tat monoclonal antibody (1D9) from Dr. Dag E. Helland (39). The detection antibody was a rabbit polyclonal anti-Tat (Abcam Ab43015, 1:500). Signal was amplified with strepavidin–HRP and developed colorimetrically. The Tat ELISA is capable of detecting Tat from clades A and B (range = 100,000–100 pg/mL; fractioned recovery 74.5%).

Immunoblotting.

To visualize histone H3, acetlated histone H3, VEGFR2 and β-actin, protein from nuclear extracts or cell lysates were immobilized on Immobilon-FL membranes and incubated with primary antibodies (histone H3 and acetyl histone H3, Cell Signaling, 1:1,000; VEGFR2, Abcam, 1 µg/ ml; β-actin, Sigma, 1:5,000). Blots were visualized via chemiluminescence following detection with horseradish peroxidase-conjugated secondary antibodies (Cell Signaling, 1:5,000) on Proteinsimple FlourChem M Imager.

Immunohistochemisty.

Standard immunohistochemical methods with Powervision-HRP secondary antibodies were used to determine the subset of T cells present in paraffin embedded, and to determine the presence of HIV p24 (DAKO, 1:10) and Tat (Abcam, 1:500) in serial sections. Heat-mediated antigen retrieval was performed. Sections from an HIV encephalitis patient served as positive controls and sections from HIV (−) autopsy patients served as a negative control for p24. Sections from Jurkat and Jurkat–Tat cell pellets were used as a positive and negative control for Tat.

Knockdown of VEGFR2.

A total of 2 × 10⁷ T cells were transfected with 100 µM RNA of Silencer Select siRNA to VEGFR2 (Life Technologies) or Silencer Select negative control with Lipofectamine 2000 reagent following manufacturers protocol. Control cells were treated with Lipofectamine reagent in the absence of RNA.

Proliferation Assays.

For the carboxyfluorescein succinimidyl ester (CFSE) assay, T cells were labeled with CFSE using a Molecular Probes Cell Trace kit (Invitrogen). Labeled cells were treated with Tat (200 nM), PHA (1 μg/mL) or left unstimulated for 72 h and then analyzed by flow cytometry at 500 nm excitation and 520 nm emission filters. For the thymidine uptake assay, T cells were isolated and treated with Tat (200 nM), PHA (1 μg/mL) or left unstimulated for 5 d. For the last 18 h, cells were pulsed with [3H]thymidine and thymidine uptake was quantified.

Infection of Primary T Cells.

A total of 2.5 × 10⁶ T cells were pretreated for 72 h with either PHA (1 μg/mL) or Tat (200 nM). Cells were then collected and washed, and 2 × 10⁶ cells were infected with 500 TCID₅₀ of HIV-_NL4-3 for 24 h. One week after infection, supernatants were collected and analyzed for Tat and product enhanced reverse transcriptase (40).