Analysis of MicroRNA Cargo in Circulating Extracellular Vesicles from HIV-Infected Individuals with Pulmonary Hypertension

Abstract

The risk of developing pulmonary hypertension (PH) in people living with HIV is at least 300-fold higher than in the general population, and illicit drug use further potentiates the development of HIV-associated PH. The relevance of extracellular vesicles (EVs) containing both coding as well as non-coding RNAs in PH secondary to HIV infection and drug abuse is yet to be explored. We here compared the miRNA cargo of plasma-derived EVs from HIV-infected stimulant users with (HIV + Stimulants + PH) and without PH (HIV + Stimulants) using small RNA sequencing. The data were compared with 12 PH datasets available in the GEO database to identify potential candidate gene targets for differentially altered miRNAs using the following functional analysis tools: ingenuity pathway analysis (IPA), over-representation analysis (ORA), and gene set enrichment analysis (GSEA). MiRNAs involved in promoting cell proliferation and inhibition of intrinsic apoptotic signaling pathways were among the top upregulated miRNAs identified in EVs from the HIV + Stimulants + PH group compared to the HIV + Stimulants group. Alternatively, the downregulated miRNAs in the HIV + Stimulants + PH group suggested an association with the negative regulation of smooth muscle cell proliferation, IL-2 mediated signaling, and transmembrane receptor protein tyrosine kinase signaling pathways. The validation of significantly differentially expressed miRNAs in an independent set of HIV-infected (cocaine users and nondrug users) with and without PH confirmed the upregulation of miR-32-5p, 92-b-3p, and 301a-3p positively regulating cellular proliferation and downregulation of miR-5571, -4670 negatively regulating smooth muscle proliferation in EVs from HIV-PH patients. This increase in miR-301a-3p and decrease in miR-4670 were negatively correlated with the CD4 count and FEV1/FVC ratio, and positively correlated with viral load. Collectively, this data suggest the association of alterations in the miRNA cargo of circulating EVs with HIV-PH.

Keywords: cocaine; extracellular RNA; pulmonary vascular disease; substance abuse.

Figure 1

(A) Heatmap of the standardized expression of miRNAs that are significantly differentially expressed between HIV + STIM + PAH and HIV + STIM samples. MiRNAs are represented in rows and samples in columns. The normalized expression data were row standardized (zero mean and unit variance) with negative values in green representing relatively low expression and positive red values representing relatively high expression. The data were hierarchically clustered both row and column wise using the Euclidean distance measure and Ward’s linkage function. (B) Volcano plot of the differential expression profile of miRNAs between HIV + STIM + PAH and HIV + STIM samples. The x-axis represents the log of the differential expression ratio, and the y-axis represents the negative log of the false discovery rate. The vertical red perforated lines represent the +/− 1.5-fold-change values. The horizontal red perforated line represents the 0.05 FDR value. MiRNAs that are significantly downregulated in HIV + STIM + PAH relative to HIV + STIM are shown in green and those that are significantly upregulated are shown in red.

Figure 2

Over-representation analysis using miEAA. (A) Significantly upregulated and (B) significantly downregulated miRNAs depicted as a heatmap representing the effected pathways. The diagrams only represent pathways that contain at least four miRNAs and miRNAs that are in at least four pathways. The cells corresponding to an miRNA that is associated with regulated pathways are colored red for upregulated and green for downregulated pathways whereas the not significantly altered pathways are in black. The expression fold difference of the miRNAs is given on top of the heatmap along the column of the corresponding miRNA. The color-bar along the right margin of the heatmap indicates the over (orange)- or under (green)-representation status of the miRNAs in the pathway.

Figure 3

(A) Hierarchically clustered heatmap of differential gene expression between PAH and Control samples in 12 GEO datasets. The log ratio of genes that showed consistent differential expression changes across the 12 samples are shown in the heatmap with red indicating up-regulation and green indicating down-regulation. These genes were selected for their direct association with pulmonary hypertension based on the literature evidence. (B,C) Shown are molecular interaction networks between the significantly differentially expressed miRNAs (HIV + STIM + PAH vs. HIV + STIM) and genes associated with pulmonary hypertension leading to a predicted activation state of the function as deduced by IPA’s causal analysis tools. Solid lines indicate direct molecular interactions and perforated lines indicate indirect interactions.

Figure 4

Quantitative RT-PCR analysis of some of the significantly upregulated (A) and downregulated miRNAs (B) in the plasma-derived EVs from HIV + Stim + PH group compared to EVs from HIV + Stim group. The miRNAs were selected based on criteria of absolute fold-change ≥ 1.5 and p-value ≤ 0.05 as seen in volcano plot, common in the analysis of miRNA gene interaction networks by IPA and gene enrichment analysis; and significant relevance to disease pathology known in the literature and/or previously shown to be associated with HIV infection. * p < 0.05, ** p < 0.01 vs. HIV + STIM.

Figure 5

RT-PCR validation of significantly upregulated miRNAs in plasma-derived EVs from an independent cohort from the University of Pittsburgh. (A) The comparison between plasma-derived EVs from uninfected (UI), uninfected individuals with cocaine use (Coc), HIV-infected individuals (HIV), and HIV-infected individuals with cocaine use (HIV + Coc) (n = 6/group) * p < 0.05, ** p < 0.01 vs. UI and (B) compares EVs from HIV-infected individuals with and without PH individuals (n = 6/group) based on cocaine use * p < 0.05, ** p < 0.01, *** p < 0.001 vs. HIV while (C) shows overall comparison between the two groups (n = 12/group) * p < 0.05, ** p < 0.01 vs. HIV.

Figure 6

RT-PCR validation of selected significantly downregulated miRNAs in EVs from the University of Pittsburgh cohort. (A) Comparison of plasma-derived EVs from uninfected (UI), uninfected individuals with cocaine use (Coc), HIV-infected individuals (HIV), and HIV-infected individuals with cocaine use (HIV + Coc) (n = 6/group) * p < 0.05, ** p < 0.01, *** p < 0.001 vs. UI. (B,C) The levels of downregulated miRNA in EVs from HIV-infected individuals with and without PH individuals showing drug usage (n = 6/group) (C) and overall comparison (n = 12/group). * p < 0.05 vs. HIV.

Figure 7

(A) Correlation analysis of significantly upregulated EV-linked miR-301a-3p in HIV-PH patients with clinical parameters CD4, viral load, DLCO, and FEV1/FVC ratio. (B) Correlation analysis of HIV-PH associated downregulated miR-4670 and miR-5571 with CD4 and viral load in HIV-infected individuals.

Figure 8

Schematic summarizing the miRNA analysis of EVs from people living with HIV with and without pulmonary hypertension. Upregulated molecules are colored red and downregulated molecules are colored green.