HDL-transferred microRNA-223 regulates ICAM-1 expression in endothelial cells

Abstract

High-density lipoproteins (HDL) have many biological functions, including reducing endothelial activation and adhesion molecule expression. We recently reported that HDL transport and deliver functional microRNAs (miRNA). Here we show that HDL suppresses expression of intercellular adhesion molecule 1 (ICAM-1) through the transfer of miR-223 to endothelial cells. After incubation of endothelial cells with HDL, mature miR-223 levels are significantly increased in endothelial cells and decreased on HDL. However, miR-223 is not transcribed in endothelial cells and is not increased in cells treated with HDL from miR-223(-/-) mice. HDL inhibit ICAM-1 protein levels, but not in cells pretreated with miR-223 inhibitors. ICAM-1 is a direct target of HDL-transferred miR-223 and this is the first example of an extracellular miRNA regulating gene expression in cells where it is not transcribed. Collectively, we demonstrate that HDL's anti-inflammatory properties are conferred, in part, through HDL-miR-223 delivery and translational repression of ICAM-1 in endothelial cells.

Figure 1. HDL and its components distinctly alter gene expression in human endothelial cells

(a) HCAEC were incubated at 37 °C for 16 h with SUV (final PLPC concentration 3 mmol l⁻¹), (b) lipid-free apoA-I (final concentration 1 mg ml⁻¹), (c) rHDL (final apoA-I concentration 1 mg ml⁻¹) or (d) native human HDL (final total protein concentration 1 mg ml⁻¹). Volcano plots of significant and differential HCAEC gene expression changes are shown for five separate experiments (≥1.5-fold, Benjamini–Hochberg-corrected P < 0.05). Red dots indicate genes with ≥1.5-fold change (Log2 fold change) and P < 0.05 (−Log10 corrected P-value). (e) Venn diagram of overlapping gene changes. Green circle (24 genes) and red circle (1 gene) identify common HCAEC genes that were respectively downregulated or upregulated by SUV, rHDL and native HDL.

Figure 2. HDL inhibits genes associated with inflammation and cholesterol efflux in endothelial cells

ICAM-1 (a) and VCAM-1 (b) mRNA expression in HCAEC incubated with lipid-free apoA-I, rHDL or native HDL as described in the legend to Fig. 1. CSF2 (c), PRMD1 (d), ABCA1 (e) and ABCG1 (f) mRNA levels in HCAEC incubated with native HDL, as quantified by real-time PCR. All data are presented as mean ± s.e.m. (N ≥ 4). Mann–Whitney nonparametric test was used to compare two independent groups with a sample size > 3. When data were paired and samples size was > 3, Wilcoxon-matched-pairs tests were used. *P < 0.05.

Figure 3. HDL transfers miR-223 to human endothelial cells

(a) HCAEC were incubated with native HDL, lipid-free apolipoprotein A-I (apoA-I), discoidal reconstituted HDL (rHDL) or SUV as described in the legend to Fig. 1. Intracellular levels of mature miR-223 were quantified by real-time PCR using individual TaqMan miRNA assays. RQV (normalized to U6) are reported as fold change (N = 5). (b) Primary miR-223 (pri-mir-223, TaqMan Assays), precursor miR-223 (pre-miR-223, SYBR green assays), and mature miR-223 (TaqMan assays) levels in HCAEC. RQV normalized to peptidylprolyl isomerase A (PPIA) (N = 4). (c) Intracellular miR-223 levels in HCAEC after incubation with native HDL (final protein concentration 1 mg ml⁻¹) for 30 min, 1, 4, 8 or 16 h, normalized to U6 and reported as fold change (N = 5–8). (d) Intracellular levels of miR-223 levels in HCAEC incubated with increasing doses of HDL treatments for 24 h. Results are normalized to U6 and reported as RQV (N = 3). (e) Intracellular miR-223 levels in HCAEC pre-incubated with actinomycin D (5 µg ml⁻¹, 8 h) followed by native HDL incubation for 24 h, as quantified by real-time PCR, normalized to PPIA and reported as fold change. (N = 3). (f) Intracellular levels of Dicer1 and miR-223 in HCAEC transfected with 100 nM siRNA against Dicer1, as quantified by real-time PCR. Results are normalized to PPIA and reported as fold change. (N = 4). (g) miRNA profiling of HDL before (pre) and after (post) incubation with HCAEC. Real-time PCR-based TaqMan Low-Density miRNA arrays. Values are presented as a volcano plot of significant and differential HDL-associated miRNA changes (≥1.5-fold, P < 0.05). (h) Validation of miR-223 levels in HDL pre/post incubation with HCAEC, as quantified by individual TaqMan miRNA assays. (N = 4). All data are presented as mean + s.e.m. Unpaired two-tailed Student’s t-test was used when comparing two groups with a sample size ≥3. Mann–Whitney nonparametric test was used to compare two independent groups with a sample size > 3. When data were paired and samples size was > 3, Wilcoxon matched-pairs tests were used. To compare three or more unmatched groups, one-way ANOVA and the Tukey’s multiple comparisons post-test was used. *P < 0.05.

Figure 4. miR-223 directly target ICAM-1 in HCAEC

(a) Normalized (Firefly/Renilla) luciferase activity in HEK 293 cells dual transfected for 48 h with miR-223 and luciferase gene reporters containing the full-length ICAM-1–3′ UTR, a three-nucleotide (814–816) deletion in the predicted miR-223 target site or a control-3′ UTR. N = 3–9 (b) Normalized (Firefly/Renilla) luciferase activity in HCAEC transfected with luciferase gene reporters containing the full-length ICAM-1–3′ UTR, a three-nucleotide (814–816) deletion or control-3′ UTR for 24 h and incubated for a further 24 h with or without native HDL. N = 3–18 (c) ICAM-1 mRNA levels in HCAEC transfected (48 h) with miR-223, control siRNA or targeting pool siRNA ICAM-1. N = 5–7 (d) ICAM-1 protein levels 24, 48 and 72 after HCAEC were transfected with miR-223, ICAM-1 siRNA or control siRNA, as quantified by flow cytometry. N = 4 (e) ICAM-1 protein levels in HCAEC 72 h after miR-223, control siRNA or ICAM-1 siRNA transfection followed by ± TNF-α (2 ng ml⁻¹) for 5 h. N = 3–4 (f) ICAM-1 protein levels in HCAEC transfected with miR-223 hairpin inhibitor, or ICAM-1 siRNA for 24 h followed by native HDL or PBS incubations (16 h) and ± TNF-α for a further 5 h. N = 3–6 (a–f) All data are represented as mean ± s.e.m. (g,h) Neutrophil-binding assays in activated HCAEC as described in the Supplementary Information N = 4. The numbers of adherent neutrophils are presented with values representing the mean + s.e.m. Unpaired two-tailed Student’s t-test was used when comparing two groups with a sample size ≤3. Mann–Whitney nonparametric test was used to compare two independent groups with a sample size > 3. When data were paired and samples size was > 3, Wilcoxon matched-pairs tests were used. To compare three or more unmatched groups, one-way ANOVA and Tukey’s multiple comparisons post-test were used. *P < 0.05, **P < 0.01.

Figure 5. HDL repression of inflammatory gene expression in HCAEC

(a) RQV of miR-223 levels in HDL isolated from human and mouse plasma by fast-protein liquid chromatography (FPLC). Mouse plasma was pooled from six miR-223^−/− mice and 12 WT mice for FPLC separation. (b,c) HCAEC were incubated with pooled native HDL (final total protein concentration 1 mg ml⁻¹) from WT or miR-223^−/− mice for 24 h. (b) miR-223 levels were quantified by real-time PCR and normalized to U6 levels (RQV = 2^{−(miR-223 Ct—U6 Ct)}). (N = 2–3). (c) ICAM-1 and CSF2 mRNA levels were quantified by real-time PCR and normalized to peptidylprolyl isomerase A (PPIA) levels (RQV). (N = 2–3). (d) Predicted miR-223 target site within the 3′ untranslated region (3′ UTR) of the CSF2 (mRNA) gene.