HIV-1 Tat C modulates expression of miRNA-101 to suppress VE-cadherin in human brain microvascular endothelial cells

Abstract

HIV-1 infection leads to the development of HIV-associated neurological disorders. The HIV-1 Tat protein has been reported to exert an adverse effect on blood-brain barrier integrity and permeability. Perturbation in permeability is mainly caused by disruptions in adherens junctions and tight junction proteins. We have identified HIV-1 Tat C-induced disruption of VE-cadherin mediated by miRNA-101 in human brain microvascular endothelial cells (BMVECs). HIV-1 Tat C increased the expression of miR-101, which led to downregulation of VE-cadherin. Overexpression of miR-101 resulted into the suppression of VE-cadherin. Inhibition of miR-101 by the miRNA inhibitor enhanced the expression of VE-cadherin. We have demonstrated that VE-cadherin is a direct target of miR-101 using a luciferase reporter assay, which showed that mutated VE-cadherin 3'UTR and miR-101 cotransfection did not change luciferase activity. By overexpression and knockdown of miR-101, we have demonstrated that the expression level of claudin-5 is governed by the expression of VE-cadherin. These findings demonstrate a novel mechanism for the regulation of barrier permeability by miR-101 via posttranscriptional regulation of VE-cadherin in human BMVECs exposed to the HIV-1 Tat C protein.

Figure 1.

Expression of VE-cadherin and other TJPs decreases upon Tat C treatment in a dose-dependent manner. A, Fold change in miR-101 expression level after treatment with increasing doses of HIV-1 Tat C protein and HI-Tat C in BMVECs. BMVECs were harvested for protein and RNA analysis after 12 h of Tat C treatment. Expression of miR-101 was determined by qPCR using the human miR-101-specific TaqMan assay. The expression level of a small RNA, RNU24, was used as a normalizer. Results are shown as the fold change compared with control. Changes in the expression levels of miR-101 are statistically significant. *p ≤ 0.05; **p ≤ 0.005; ***p ≤ 0.0005. B, Western blot analysis demonstrating no change in the VE-cadherin expression level in BMVECs after exposure to the HI-Tat C protein. C, Western blots analysis for VE-cadherin and other TJPs (claudin-5, ZO-1, occludin) of samples treated with increasing doses of Tat C showing a decrease at the protein expression level. D, Graph showing densitometry analysis of the results. Image density of Western blots has been normalized with β-tubulin using ImageJ software. Changes in TJPs and VE-cadherin in BMVECs exposed to the Tat C protein are statistically significant. **p ≤ 0.005; *p ≤ 0.05. For ZO-1, *p ≤ 0.05 at all doses; at 1.5 μg/ml Tat C treatment, **p ≤ 0.005. For occludin, downregulation is statistically significant at all doses of Tat C, *p ≤ 0.05; at 1.5 μg/ml Tat C treatment, **p ≤ 0.005. Downregulation of claudin-5 was statistically significant. *p ≤ 0.05. Results are representative of three biologically repeated experiments and are represented as mean ± SE. E, Fold change in miR-101 expression level in BMVECs after Tat C treatment and empty vector pET-21b purified in a similar manner, showing that upregulation of miR-101 is specific to Tat C. F, Western blot analysis of the pET-21b and Tat C-treated BMVECs showing that downregulation of VE-cadherin is specific to Tat C and is not induced by either buffer or empty vector. All experiments were repeated three times and are represented as mean ± SE.

Figure 2.

Decrease in expression of VE-cadherin and TEER with extended exposure of Tat C to BMVECs and disruption of endothelial barrier function. A, Western blot analysis showing decrease in the expression level of VE-cadherin in BMVECs with the increase in time of exposure of Tat C (500 ng/ml) to BMVECs. B, Graph representative of three independent experiments showing a decrease in VE-cadherin expression with the increase in time of exposure of Tat C (500 ng/ml) to BMVECs. The VE-cadherin expression level started decreasing significantly after 6 h (p ≤ 0.05). C, Graph showing the decrease in TEER value with the increase in time of exposure of Tat C (500 ng/ml) to BMVECs. Increase in the time of exposure of the Tat C protein on BMVECs disrupted the barrier integrity of BMVECs and thereby increased the permeability (p ≤ 0.005). D, Graph showing the dose-dependent decrease in the TEER value after Tat C treatment. E, Bars indicating the fold changes in fluorescence after leakage of fluorescent compound (sodium fluorescein) across the monolayer of BMVECs exposed to Tat C protein (p ≤ 0.05). The migration of sodium fluorescein through control cells was set at 1 and the results are shown accordingly. As the dose of Tat C is increasing, the migration of sodium fluorescein is increasing from upper compartment to the lower compartment of the Transwell insert membrane culture experiment. All experiments were repeated three times and are represented as mean ± SE.

Figure 3.

No significant effect of HIV-1 Tat B or Tat C on the expression of major miRNA biogenesis pathway proteins. A, Western blot analysis for tracking the changes in expression level (if any) of major proteins involved in miRNA biogenesis. BMVECs were exposed for 12 h to Tat B and Tat C proteins (500 ng/ml) and cells were harvested and tested for major proteins (Drosha, Dicer, Ago-2, TRBP). Neither clade of HIV-1 Tat protein (Tat B and Tat C) could exert any significant change in the expression of Drosha, Dicer, Ago-2, or TRBP. B, Densitometry quantitation of Drosha, Dicer, TRBP, and Ago-2 proteins as normalized with image density of β-tubulin by using ImageJ software. Results are representative of three independent biological repeats and are shown as mean ± SE. No significant change can be seen (p ≥ 0.05). C, Bar diagram showing downregulation of miR-101 in BMVECs exposed to the Tat B protein. The expression level of miR-101 was quantified by specific TaqMan probes using RNU24 as a normalizer. D, Western blot analysis of VE-cadherin in Tat B-treated BMVECs demonstrates the differential regulation of VE-cadherin expression. All experiments were repeated three times and are represented as mean ± SE.

Figure 4.

miR-101 binds complementary sequences in CDH5 3′UTR directly and regulates the expression of CDH5 (VE-cadherin). A, Schematic representation of seed sequences in miR-101 and complementary 8-mer binding sites in WT-CDH5 3′UTR (VE-cadherin).The complementary sites are deleted in MUT-CDH53′UTR using a site-directed mutagenesis kit to abrogate the regulatory interaction of miR-101 in the 3′UTR of CDH5. WT and MUT-CDH5 3′UTR reporter constructs were cotransfected with miR-101, miR-29b, and pCMV-β-gal (as a normalizing control) plasmids and luciferase assays were performed in the HeLa cells. B, miR-101 transfection suppressed the luciferase activity significantly (p ≤ 0.005) in the WT-CDH5 3′UTR cotransfection. miR-101 transfection could not represse the luciferase activity when cotransfected with the MUT-CDH5 3′UTR. An irrelevant miR-29b, which does not have the complementary sequences for CDH5 3′UTR, did not perturb the luciferase activity. The luciferase expression level of cells (transfected with only 3′UTR of VE-cadherin) was considered as a control with 100% relative light unit and the results are shown accordingly. All experiments were repeated three times and are represented as mean ± SE. **p ≤ 0.005.

Figure 5.

Overexpression of miR-101 reduces the expression of VE-cadherin in BMVECs. A, Graph showing the fold change in miR-101 and miR 29b in miR-101- and miR-29b-overexpressed BMVECs, respectively, via qPCR using the TaqMan miR-101 and TaqMan miR-29b assays. The miR-101 level was found to be 8.5-fold higher in miR-101-transfected cells compared with vector control (p ≤ 0.05). B, Western blot analysis for VE-cadherin in BMVECs after overexpression of miR-101. Empty vector was used as a negative control and the plasmid pCMV-miR-29b was used as nonspecific miRNA to test the specificity of regulation. In miR-101 overexpression, the VE-cadherin expression level is downregulated significantly (p ≤ 0.05). C, Densitometry image density analysis of VE-cadherin using β-tubulin as a normalizer in ImageJ software. All experiments were repeated three times and are represented as mean ± SE. D, qPCR analysis of VE-cadherin in miR-101- and miR-29b-overexpressed cells to determine changes at the transcript level. Total RNA was reverse transcribed by using random hexamers and transcript levels were quantified with VE-cadherin-specific TaqMan probes using GAPDH as an internal reference. Fold changes were determined with the ΔΔct method. E, Fold change in miR-101 expression level after transfection of scramble miR-101in BMVECs showing no significant change. F, Western blot analysis for VE-cadherin in scramble miR-101-transfected BMVECs showing no change in their protein expression level. All experiments were repeated three times and are represented as mean ± SE.

Figure 6.

Anti-miR-101 transfection rescues the VE-cadherin expression level in BMVECs. A, Anti-miR-101 transfection resulted in the reduced expression of miR-101. An ∼80% reduction in the expression of cellular miR-101 in anti-miR-101-transfected BMVECs was seen compared with scramble-transfected negative controls. *p ≤ 0.05. B, Western blot analysis to determine the regain in level of expression of VE-cadherin in anti-miR-101-transfected BMVECs. After 24 h of anti-miR-101 transfection in BMVECs, cells were again exposed to Tat C protein (500 ng/ml) to test the specificity of the miR-101-mediated control over the expression of VE-cadherin, but the expression of VE-cadherin remained unaffected due to the suppressive effect of anti-miR-101 on miR-101 (p ≤ 0.05). C, Western blot images of VE-cadherin were normalized with β-tubulin and are presented as graph bars. All experiments were repeated three times and are represented as mean ± SE. *p ≤ 0.05.

Figure 7.

Expression level of VE-cadherin influences the expression level of Claudin-5 directly. To distinguish the direct effect of expression level of VE-cadherin on the expression level of claudin-5, three sets of experiments were performed. A, The downregulation of VE-cadherin upon Tat C treatment decreases the expression level of claudin-5 in a dose-dependent manner. B, Transfection of anti-miR-101 in BMVECs showing the same trend of rescued expression of claudin-5 as in the regained expression of VE-cadherin, even in the presence of the Tat C protein. C, Overexpression of miR-101 decreases the expression level of claudin-5 in the same way as that of VE-cadherin. D, Densitometry analysis using ImageJ software of Western blot images for claudin-5 in anti-miR-101 transfection experiments. A significant recovery of claudin-5 expression level can be seen (p ≤ 0.005). All experiments were repeated three times and are represented as mean ± SE. **p ≤ 0.005.

Figure 8.

Proposed model for Tat C-mediated downregulation of VE-cadherin and claudin-5 in BMVECs through miR-101. We propose that expression of VE-cadherin is regulated through miR-101 in BMVECs exposed to the HIV-1 Tat C protein. After HIV-1 Tat C treatment, the expression level of miR-101 increases and targets the expression level of VE-cadherin directly, which in turn influences the expression level of claudin-5 and thereby the permeability in BMVECs.