Biochemical Journal

Research article

Involvement of miR-30c and miR-301a in immediate induction of plasminogen activator inhibitor-1 by placental growth factor in human pulmonary endothelial cells

Nitin Patel, Stanley M. Tahara, Punam Malik, Vijay K. Kalra

Abstract

PAI-1 (plasminogen activator inhibitor-1) is a key physiological inhibitor of fibrinolysis. Previously, we have reported PlGF (placental growth factor)-mediated transcriptional up-regulation of PAI-1 (SERPINE1) mRNA expression via activation of HIF-1α (hypoxia-inducible factor-1α) and AP-1 (activator protein-1) in HPMVECs (human pulmonary microvascular endothelial cells), which resulted in elevated PAI-1 in humans with SCA (sickle cell anaemia). In the present study, we have identified the role of post-transcriptional mechanism(s) of PlGF-mediated accumulation of PAI-1 mRNA in HPMVECs by examining the role of microRNAs (miRNAs/miRs) in PlGF-induced PAI-1 mRNA stability. Our results show reduced expression of miR-30c and miR-301a, but not of miR-99a, in response to PlGF, which have evolutionarily conserved binding sites in the 3′-UTR (3′-untranslated region) of PAI-1 mRNA. Transfection of anti-miR-30c or anti-miR-301a oligonucleotides resulted in increased PAI-1 mRNA levels, which were increased further with PlGF stimulation. Conversely, overexpression of pre-miR-30c or pre-miR-301a resulted in an attenuation of PlGF-induced PAI-1 mRNA and protein levels. Luciferase reporter assays using wild-type and mutant 3′-UTR constructs confirmed that the PAI-1 3′-UTR is indeed a direct target of miR-30c and miR-301a. Finally, plasma levels of miR-30c and miR-301a were significantly down-regulated in patients with SCA compared with normal controls. These results provide a post-transcriptional regulatory mechanism of PlGF-induced PAI-1 elevation.

  • adenosine- and uridine-rich element (ARE)
  • microRNA (miRNA)
  • plasminogen activator inhibitor-1 (PAI-1)
  • placental growth factor (PlGF)
  • sickle cell anaemia (SCA)
  • 3′-untranslated region (3′-UTR)

INTRODUCTION

Vascular injury plays a significant role in the pathogenesis of a number of acute and chronic disorders, wherein initial injury triggers a cascade of events, such as inflammation, activation of coagulation and repair processes, so as to maintain vascular integrity. However, a loss in homoeostasis leads to overt pathological disease. Although many factors contribute to vascular dysfunction, PAI-1 (plasminogen activator inhibitor-1) has been shown to play a key role in the pathogenesis of cardiovascular, renal, hepatic and pulmonary disorders [1,2], and hypercoagulability in SCA (sickle cell anaemia) [3,4]. Elevated plasma levels of circulating PAI-1 and tissue factor have been observed in SCA patients under steady state, which increase further during vaso-occlusive crises [5]. The prothrombotic state predisposes patients to both stroke and pulmonary hypertension, two distressing complications of SCA [6].

PAI-1, a secreted 50 kDa protein and a member of the serpin family of serine protease inhibitors, is the main circulating inhibitor of plasminogen activation and hence fibrinolysis [7]. The X-ray crystal structure has revealed that serpins, including PAI-1, are globular proteins, which have an exposed peptide loop and the central reactive loop, the latter being essential for inhibitory mechanisms [7]. PAI-1 is synthesized by a variety of cell types, such as endothelial cells, smooth muscle cells, mononuclear cells, hepatocytes and platelets [8,9]. The expression of PAI-1 is regulated by hypoxia [8,10], insulin [11] and TGF-β (transforming growth factor-β) [12] in a variety of tissue and cell types. The hypoxia-induced expression of murine PAI-1 in monononuclear cells involves cis-elements for Egr-1 (early growth response-1), HIF-1α (hypoxia-inducible factor-1α) and C/EBPα (CCAAT/enhancer-binding protein α) in the promoter region [10]. Oxidative stress-induced PAI-1 expression requires AP-1 (activator protein-1) binding to cis-elements within the promoter [13]. In our recent studies [9], we have shown that PlGF (placental growth factor), released by sickle erythroblasts at a higher level compared with normal erythroblasts, increased PAI-1 expression in primary HPMVECs (human pulmonary microvascular endothelial cells) and monocytes. PlGF-induced PAI-1 (SERPINE1) mRNA expression occurred through activation and binding of the transcription factors HIF-1α and AP-1 to their cognate elements in the promoter region of PAI-1 at 4–6 h post-PlGF stimulation [9]. In addition, we observed that PlGF also increased PAI-1 mRNA at an early time point of 1 h by an unidentified mechanism. Therefore, in the present study, we have studied the role of miRNAs/miRs (microRNAs) in PlGF-mediated post-transcriptional PAI-1 mRNA stabilization.

miRNAs are a class of endogenous small non-coding RNAs of approx. 22–25 nucleotides, which repress gene expression by binding to imperfect complementary sequences in the 3′-UTRs (untranslated regions) of target mRNA, leading to mRNA degradation and translational repression [1417]. In mammals, miRNAs have been shown to be associated with diverse biological processes, including cell proliferation, differentiation, apoptosis, metabolism and morphogenesis [18], innate immunity [1922], kidney dysfunction in diabetes [23,24], tumorigenesis [2527] and leukotriene formation [28]. The 3′-UTR of PAI-1 mRNA contains several putative AREs (adenosine- and uridine-rich elements), and several studies have identified the interaction of different mRNA-binding proteins with these AREs for regulating turnover and stability. However, relatively little is known about the role of miRNAs in PAI-1 mRNA regulation.

In the present study, we show that PlGF treatment of HPMVECs at early time points increased the stabilization of PAI-1 mRNA through a concomitant decrease in the expression of miR-30c and miR-301a. We show further that inhibitor oligonucleotides of miR-30c and miR-301a augmented PAI-1 mRNA expression. Conversely, miR-30c and miR-301a precursor oligonucleotides reduced PAI-1 mRNA expression. We confirmed that these miRNAs directly target the PAI-1 3′-UTR in a sequence-specific manner from reporter studies. We also show that plasma concentrations of miR-30c and miR-301a were significantly reduced in patients with SCA. Our results demonstrate for the first time, to the best of our knowledge, that PlGF-induced PAI-1 expression at early time points is post-transcriptionally regulated by two miRNAs, miR-30c and miR-301a.

MATERIALS AND METHODS

Cells and reagents

HPMVECs were obtained and cultured as described previously [9]. These cells were confirmed to express CD31 and vWF (von Willebrand factor) as described previously [9]. HPMVECs were incubated overnight in EBM-2 (endothelial basal medium-2) containing 2% fetal bovine serum followed by serum deprivation for 3 h prior to treatment with either PlGF (250 ng/ml) or other indicated experimental conditions.

Reagents were obtained as follows: human recombinant PlGF was from R&D Systems; HIF-1α siRNA (small interfering RNA), PlGF siRNA and corresponding control scRNA (scrambled RNA), primary antibodies against PAI-1 and β-actin, and secondary antibodies conjugated to HRP (horseradish peroxidase) were from Santa Cruz Biotechnology; the c-Jun TranSilent siRNA vector was from Panomics; and actinomycin D was from Enzo Life Sciences. The primers used for PCR amplification of PAI-1 3′-UTR and mutagenesis were purchased from Integrated DNA Technologies. Unless otherwise specified, all other reagents were purchased from Sigma.

Human subjects

All blood samples were obtained from children with homozygous SCA at steady state at their elective clinical appointment with routine clinical draws through the Hematology Repository at Cincinnati Children's Hospital Medical Center. All samples were obtained with the informed consent of the patient or parent/legal guardian using Institutional Review Board-approved protocols. The plasma samples were obtained from the same SCA patients and healthy subjects as described previously [29].

Analysis of miRNA expression

Total RNA was extracted from HPMVECs using the mirVana miRNA Isolation kit (Applied Biosystems). For extraction of total RNA (including small RNA molecules) from plasma of human subjects, 400 μl of plasma was mixed completely with TRIzol® reagent at a volume ratio of 1.0:0.8. After chloroform extraction, the aqueous layer was removed and mixed with absolute ethanol (1.25 vol.) and loaded on to the cartridge supplied in the mirVana isolation kit. Total small RNA was eluted according to the manufacturer's instructions. The miRNA expression was determined by using the TaqMan MicroRNA Assay kits for indicated miRNA (Applied Biosystems), according to the manufacturer's protocol. Briefly, 100 ng of total RNA was reverse-transcribed at 42 °C for 15 min, 95 °C for 5 min and 4 °C for 5 min. qRT-PCR (quantitative real-time PCR) (20 μl total reaction) was performed in a 384-well plate at 95 °C for 10 min, followed by 40 cycles of 95 °C for 15 s and 60 °C for 60 s. All reactions were run in triplicate. Gene expression was normalized to a reference gene: 5S rRNA for cell culture samples and RNU6B (U6 small nuclear RNA) for plasma samples. RQ (relative quantitative) levels for miRNA expression were calculated as 2−ΔΔCt by the comparative Ct method [30], where ΔΔCt=(Ct target miRNA of treated sample−Ct reference gene of treated sample)−(Ct target miRNA of control sample−Ct reference gene of control sample).

Transient transfection

HPMVECs (1×106) were resuspended in 100 μl of serum-free RPMI 1640 medium containing appropriate siRNA constructs (50 nM), luciferase reporter plasmids (0.5 μg), 60 pmol of anti-miRNA inhibitor (Life Technologies) or 1 μg of pre-miRNA precursor constructs (Life Technologies) and transfected by nucleofection with the S-05 program in the nucleofector apparatus (Lonza), as described previously [31]. The β-galactosidase plasmid (0.5 μg) was co-transfected with luciferase reporter constructs to monitor transfection efficiency. Following nucleofection, the cells were incubated in EBM-2 complete medium overnight, kept in serum-free EBM-2 for 3 h and treated with PlGF for the indicated time periods. The cell lysates were analysed for luciferase and β-galactosidase activity using appropriate kits (Promega) [32]. Luciferase values were normalized to β-galactosidase values and are expressed relative to the activity of the promoter-less pGL3-Basic vector. For PAI-1 mRNA and miRNA analysis, cells were lysed with TRIzol® and total RNA was extracted. The expression was determined as described above.

Western blot analysis

Cytosolic extracts were isolated from untreated and PlGF-treated cells. Briefly, 2×106 cells were resuspended in 400 μl of cell-lysis buffer for 20 min and cytosolic supernatants were collected after centrifugation at 10000 g for 5 min [32]. The cytosolic extracts were subjected to SDS/PAGE for PAI-1 protein expression. The protein loading was monitored by stripping and reprobing the membrane with a β-actin antibody. The protein bands were visualized using Immunobilon western reagents (Millipore).

PAI-1–3′-UTR plasmid and its mutagenesis

The 455 bp fragment spanning the region between +292 and +746 bp relative to the translation stop codon of PAI-1 mRNA was PCR-amplified using the forward primer containing a HindIII restriction enzyme site and the reverse primer containing an SpeI site, as shown in Supplementary Table S1 (http://www.BiochemJ.org/bj/434/bj4340473add.htm), and a Phusion high-fidelity DNA polymerase (New England Biolabs), according to standard procedures. The human cDNA clone for SERPINE1 (GenBank® accession NM_000602) was used as a template (Origene). The PCR product was cloned downstream of the firefly luciferase reporter gene in pMIR-REPORT plasmid (Ambion). The orientation of the insert relative to the luciferase gene was confirmed by DNA sequencing, and the plasmid was purified using EndoFree Plasmid Maxi Kit (Qiagen). The resulting plasmid is designated as pMIR-PAI-1–3′-UTR. The mutant constructs of PAI-1–3′-UTR in the binding sites for miR-30c, miR-301a or miR-99a were generated using pMIR-PAI-1–3′-UTR as the template with the primers listed in Supplementary Table S1, in accordance with the QuikChange® site-directed mutagenesis protocol (Stratagene). The double-mutant construct for the miR-30c- and miR-301a-binding sites was generated using the miR-30c mutant construct as a template. The mutations were confirmed by DNA sequencing (Microchemical Core Facility, Norris Comprehensive Cancer Center, University of Southern California, Los Angeles, CA, U.S.A.).

Statistical analysis

Results are presented as means±S.E.M. The significant difference in mean values between multiple groups was analysed using a parametric one-way ANOVA, followed by a Tukey–Kramer test using the Instat 2 software suite (Graphpad). A Student's t test was used to evaluate the significant difference between two groups of experiments. *P<0.05, **P<0.01, ***P<0.001, and ns (not significant), P>0.05.

RESULTS

Post-transcriptional regulation of PlGF-mediated PAI-1 mRNA expression at early time periods (1 and 2 h)

We have reported previously [9] that PlGF-induces PAI-1 mRNA expression after 6 h in HPMVECs via transcriptional activation of the PAI-1 promoter by HIF-1α and AP-1. In the same study [9], we observed that PlGF could elevate PAI-1 mRNA levels as early as 1 h post-induction. In the present study, we confirmed those results and observed increased PlGF-mediated PAI-1 mRNA expression at earlier time intervals (1 and 2 h post-induction) using qRT-PCR (Supplementary Figure S1A at http://www.BiochemJ.org/bj/434/bj4340473add.htm). Moreover, PlGF-mediated PAI-1 expression at 1 and 2 h occurred in the absence of transcriptional activation of the PAI-1 promoter, since PlGF activation of the PAI-1 promoter was delayed and observed only at later time points (4 and 6 h post-treatment) (Supplementary Figure S1B).

Next, we assessed the role of the transcription factors HIF-1α and c-Jun in PlGF-induced PAI-1 mRNA expression at early (2 h) and late (4 and 6 h) time points. As shown in Supplementary Figure S1(C), PlGF-mediated PAI-1 mRNA expression at 2 h was not affected by transfection with either HIF-1α siRNA or HIF-1α scRNA. However, PlGF-induced PAI-1 mRNA expression at later time points (4 and 6 h) was significantly attenuated by HIF-1α siRNA, but not with HIF-1α scRNA, as observed previously [9]. Moreover, transfection with c-Jun siRNA, but not c-Jun scRNA, reduced PlGF-induced PAI-1 mRNA expression at 4 and 6 h (Supplementary Figure S1D), as reported previously [9]. In contrast, PlGF-mediated PAI-1 mRNA expression was not affected by transfection of either c-Jun siRNA or c-Jun scRNA at 2 h (Supplementary Figure S1D). Taken together, these results indicate that PlGF-induced PAI-1 mRNA expression at earlier time points (1 and 2 h) most likely occurs by mRNA stabilization and at later time points (4 and 6 h) via transcriptional activation of the PAI-1 promoter by HIF-1α and AP-1.

Next, we performed experiments to determine whether the increase in PAI-1 mRNA levels at the early time period (2 h) in PlGF-treated HPMVECs was a result of increased mRNA stability. The cells were first treated with recombinant PlGF (250 ng/ml) for 2 h, followed by treatment with actinomycin D to inhibit ongoing transcription, and total RNA was isolated at indicated times following induction (10–120 min) for PAI-1 mRNA analysis. A comparison of PAI-1 turnover in the absence and presence of PlGF treatment showed a significant stabilization in mRNA from a half-life of 12±1.5 min to 28±3 min (Supplementary Figure S1E; a description of the method used to determine the half-life can be found at http://www.BiochemJ.org/bj/434/bj4340473add.htm). The kinetics show a biphasic nature of PAI-1 mRNA decline in untreated cells, indicating a second-order degradation process, consistent with an miRNA-dependent mechanism. In contrast, the linear or first-order nature of PAI-1 mRNA decay in PlGF-treated cells implicated a process that was dependent only on mRNA concentration.

Identification of miRNAs involved in PlGF-mediated PAI-1 mRNA stabilization

We examined miRNAs that may be involved in the stabilization and degradation of PAI-1 mRNA. Bioinformatics approaches were used to identify potential miRNA target sites harboured in the 3′-UTR of the PAI-1 mRNA. [17,3335]. We used a web-based miRNA target prediction program (MicroCosm; http://www.ebi.ac.uk/enright-srv/microcosm/cgi-bin/targets/v5/search.pl) that considers complementarity, target site accessibility and the extent of evolutionary conservation. This program predicted ~20 putative miRNA target sites in the PAI-1 3′-UTR. In the present study, we selected five candidates (i.e. miR-30c, miR-301a, miR-99a, miR-299-5p and miR-609) based on good complementarity (ΔG °~−17–27 kcal/mol) and a high degree of site conservation among different mammalian species (Supplementary Figure S2 at http://www.BiochemJ.org/bj/434/bj4340473add.htm). There is only a single predicted target site for each of these miRNAs in the 3′-UTR of PAI-1 mRNA.

Next, we examined whether PlGF affected the expression of these five candidate miRNAs in HPMVECs after 2 h of PlGF treatment. As shown in Figure 1(A), there was an approx. 10- and 4-fold down-regulation in the expression of miR-30c and miR-301a respectively in PlGF-treated cells as compared with untreated cells. However, the expression of miR-99a remained unaltered after PlGF stimulation. The expression of miR-299-5p and miR-609 was not detected in HPMVECs (results not shown) using the TaqMan microRNA assay. In order to identify the role of PlGF in the regulation of miR-30c, miR-301a and miR-99a expression, we utilized an siRNA approach for silencing endogenous PlGF expression. As shown in Figure 1(B), HPMVECs transfected with PlGF siRNA showed elevated expression of both miR-30c and miR-301a by approx. 3.5- and 2.6-fold respectively when compared with PlGF scRNA. However, PlGF siRNA did not change the expression of miR-99a in HPMVECs. It is pertinent to mention that transfection with PlGF siRNA almost completely abolished PlGF mRNA expression by approx. 95%, compared with untransfected cells, whereas PlGF scRNA had no effect on PlGF mRNA expression (results not shown). These results showed that endogenously expressed PlGF regulated the basal expression of miR-30c and miR-301a in endothelial cells.

Figure 1 PlGF down-regulates the levels of miRNA-30c and miR-99a, leading to increased PAI-1 mRNA expression

(A) HPMVECs were treated with PlGF for 2 h, followed by isolation of total RNA. Levels of indicated miRNA candidates were analysed using a specific TaqMan MicroRNA assay for each miRNA, as described in the Materials and methods section. (B) HPMVECs were transfected with either 50 nM PlGF siRNA or PlGF scRNA. Total RNA was isolated after 24 h, followed by qRT-PCR analysis for the indicated miRNAs. In (A and B), the relative expression levels of indicated miRNAs were normalized to 5S rRNA expression. Results are expressed as fold change values compared with control cells. (C) HPMVECs were transfected with indicated the anti-miRNA oligonucleotides, followed by total RNA isolation. PAI-1 mRNA expression was estimated by qRT-PCR. (D) qRT-PCR analysis of PAI-1 mRNA expression in HPMVECs transfected with the indicated anti-miRNA oligonucleotides after 2 h of PlGF treatment. In (C and D), PAI-1 mRNA levels were normalized to GAPDH (glyceraldehyde-3-phosphate dehydrogenase) mRNA expression. (E) Relative expression of the indicated miRNAs in HPMVECs transfected with corresponding anti-miRNA oligonucleotides. The level of miRNA expression observed in cells transfected with the negative control oligonucleotide was considered ‘1’. Values are the means±S.E.M. of triplicate determinations at each time point from three independent experiments. *P<0.05, **P<0.01 and ***P<0.001.

Next, we investigated whether endogenous levels of miR-30c and miR-301a regulated the expression of PAI-1 mRNA in HPMVECs and were thus physiologically relevant. To address this, HPMVECs were transfected with specific antisense RNA oligonucleotides (i.e. anti-miRNAs). As shown in Figure 1(C), transfection of HPMVECs with anti-miR-30c increased the basal expression of PAI-1 mRNA by 3-fold, and transfection with anti-miR-301a produced a 2-fold increase in PAI-1 mRNA expression. However, anti-miR-99a transfection did not affect PAI-1 mRNA expression. These results indicated that miR-30c and miR-301a, but not miR-99a, probably regulated endogenous levels of PAI-1 mRNA in HPMVECs. Subsequently, we examined whether reduced levels of miR-30c and miR-301a observed in HPMVECs, following PlGF treatment, contributed to the increased expression of PAI-1 mRNA, which may be pathophysiologically relevant. As shown in Figure 1(D), transfection of HPMVECs with anti-miR-30c augmented PlGF-mediated PAI-1 mRNA expression by 6-fold compared with a 2.5-fold increase in PAI-1 mRNA expression in response to PlGF alone. Similarly, transfection of anti-miR301a, followed by treatment with PlGF, resulted in a 4-fold increase in PAI-1 mRNA levels compared with a 2.5-fold increase in PAI-1 mRNA expression observed with PlGF alone. However, the transfection of anti-miR99a did not change PlGF-mediated levels of PAI-1 mRNA (Figure 1D). It is pertinent to note that transfection of the indicated anti-miRNAs showed significant reduction (~80%) in the levels of individual miRNAs, as compared with the negative control (Figure 1E). Taken together, these results show that miR-30c and miR-301a, but not miR-99a, are involved in the regulation of basal and PlGF-induced PAI-1 mRNA expression. Furthermore, these findings indicate that the PlGF-mediated reduction in the levels of miR-30c and miR-301a results in the accumulation of PAI-1 mRNA through a decrease in its degradation in HPMVECs.

miR-30c and miR-301a interact with the target-binding sites in the 3′-UTR of PAI-1 mRNA

In order to investigate whether miR-30c and miR-301a directly regulated PAI-1 mRNA expression, an experiment was performed in which the 3′-UTR of PAI-1 mRNA was inserted downstream of the luciferase open reading frame in pMIR-REPORT. As shown in the schematic diagram in Figure 2(A), a region of the PAI-1 3′-UTR (bases +292 to +746), relative to the translation stop codon, containing complementary binding sites for miR-30c, miR-301a and miR-99a and adjacent surrounding sequences was used for reporter construction. Use of this segment also avoided potential interference from other putative miRNA-binding sites. The resulting reporter construct (pMIR-PAI-1–3′-UTR) was co-transfected with the pMIR-REPORT β-galactosidase control plasmid to monitor transfection efficiency in HPMVECs. Our results showed that basal reporter activity was drastically (>90%) reduced in cells transfected with pMIR-PAI-1–3′-UTR when compared with cells transfected with control pMIR-REPORT construct alone, thus demonstrating the destabilizing effect of the PAI-1 3′-UTR on luciferase mRNA expression (Figure 2B). As shown in Figure 2(B), transfection with the pMIR-PAI-1–3′-UTR reporter followed by treatment with PlGF reversed this effect and resulted in a time-dependent (0–2 h) increase in luciferase activity. There was an approx. 4-fold increase in reporter activity by 2 h. It is important to mention that PlGF did not change the reporter activity of the control plasmid pMIR-REPORT (Figure 2B).

Figure 2 PlGF-induced PAI-1–3′-UTR reporter activity is abrogated by miR-30c and miR-99a, but not by miR-99a

(A) Schematic representation of the PAI-1–3′-UTR luciferase reporter plasmid. The region between nucleotides 292 to 746 of the PAI-1 3′-UTR containing the predicted target binding sites for miR-30c, miR-301a and miR-99a was cloned at the 3′-end of luciferase reporter gene in the pMIR-REPORT plasmid. (B) HPMVECs were transfected with either the pMIR-REPORT control plasmid or the pMIR-PAI-1–3′-UTR construct. Following transfection, the cells were treated with PlGF for the indicated time periods. (C) The sequences of predicted miR-30c-, miR-301a- and miR-99a-binding sites within the PAI-1 3′-UTR. The bases in bold highlight ‘seed sequences’, and the point mutations in the seed sequence are indicated by asterisks. (D) HPMVEC were transfected with the pMIR-REPORT control plasmid, the wild-type pMIR-PAI-1–3′-UTR or the indicated PAI-1–3′-UTR mutant constructs for the binding of either miR-30c, miR-301a or miR-99a or both (miR-30c and miR-301a). At 24 h post-transfection, cells were lysed in reporter lysis buffer according to the manufacturer's instructions for the luciferase assay. (E) HPMVECs were transfected with the pMIR-PAI-1–3′-UTR plasmid and the indicated pre-miRNA precursor oligonucleotides. (F) HPMVECs were transfected with the pMIR-PAI-1–3′-UTR plasmid and the indicated anti-miRs, followed by PlGF treatment for 2 h. In (B, DF), each construct was used at 0.5 μg. All luciferase reporters were co-transfected with the β-galactosidase plasmid for normalization of transfection efficiency. The luciferase and β-galactosidase activities were estimated as described in the Materials and methods section. Results are representative of three independent experiments. *P<0.05, **P<0.01 and ***P<0.001. hsa, Homo sapiens; mut, mutant; wt, wild-type.

Next, we determined which target sites in the PAI-1 3′-UTR were responsible for the dramatic reduction in reporter activity. For this we introduced four point mutations in the corresponding ‘seed sequences’ in pMIR-PAI-1–3′-UTR for each miRNA candidate to eliminate any possible binding interactions (Figure 2C). In addition, double mutations were created in the PAI-1 3′-UTR to jointly eliminate miR-30c and miR-301a binding. As shown in Figure 2(D), the suppression of reporter activity was significantly relieved in cells transfected with PAI-1–3′-UTR-miR-30c mutant compared with the wild-type pMIRPAI-1–3′-UTR plasmid. The PAI-1–3′-UTR-miR-301a mutant showed partial recovery of reporter activity in comparison with the wild-type reporter. By contrast, no suppression in activity of the reporter construct pMIR-PAI-1–3′-UTR-miR-99a mutant was observed (Figure 2D), as expected. More interestingly, the double miRNA site mutant completely rescued the repression of reporter activity (Figure 2D). Therefore miR-30c and miR-301a, but not miR-99a, are jointly responsible for post-transcriptional inhibition of PAI-1 expression.

The above results were corroborated further in cells co-expressing specific miRNA precursor molecules and the pMIR-PAI-1–3′-UTR reporter plasmid. As shown in Figure 2(E), overexpression of miR-30c precursor with the pMIR-PAI-1–3′-UTR reporter construct resulted in significant reduction (~90%) in luciferase activity. Similarly, miR-301a precursor transfection attenuated luciferase activity by approx. 60%. In contrast, transfection with miR-99a precursor showed no reduction in luciferase activity. To determine the specificity of these active miRNAs for PAI-1 mRNA, we used a complementary approach utilizing anti-miRNA oligonucleotides. As shown in Figure 2(F), co-transfection of anti-miR-30c with the PAI-1–3′-UTR reporter augmented luciferase activity by 3-fold (lane 3 compared with lane 2), and anti-miR-301a showed a 2-fold increase in reporter activity (lane 4 compared with lane 2), following PlGF treatment for 2 h. As expected, transfection with anti-miR99a had no effect on luciferase reporter activity in the presence of PlGF. Taken together, these results indicate that miR-30c and miR-301a directly target the 3′-UTR of PAI-1 mRNA for turnover, subsequently resulting in the down-regulation of PAI-1 expression.

Overexpression of precursor miRNAs down-regulates PlGF-induced endogenous PAI-1 mRNA and PAI-1 protein expression

Since HPMVECs transfected with either anti-miR-30c or anti-miR-301a showed increased PlGF-mediated PAI-1 mRNA expression, we performed the converse experiment wherein cells were transfected with precursor molecules for miR-30c, miR-301a or miR-99a. Transfection with the pre-miR-30c expression plasmid reduced PlGF-mediated PAI-1 mRNA expression to basal levels when compared with cells treated with PlGF alone (Figure 3A, lane 3 compared with lane 2). Similarly, overexpression of pre-miR-301a resulted in an approx. 80% reduction in PlGF-mediated PAI-1 mRNA expression as compared with PlGF-treated control cells (Figure 3A, lane 4 compared with lane 2). However, transfection with pre-miR99a did not affect PlGF-mediated expression of PAI-1 mRNA (Figure 3A, lane 5 compared with lane 2). Subsequently, we examined whether these miRNAs affected the expression of PAI-1 protein. As shown in Figure 3(B), PlGF treatment of HPMVECs for 3 h led to a significant increase in PAI-1 protein levels, which was completely abrogated when the cells were transfected with the pre-miR-30c expression plasmid (lane 6 compared with lane 2). Similarly, overexpression of miR-301a in HPMVECs reduced PlGF-induced PAI-1 protein expression by approx. 80% (Figure 3B, lane 5 compared with lane 2). However, overexpression of miR-99a did not alter PlGF-induced PAI-1 protein levels (Figure 3B, lane 4 compared with lane 2).

Figure 3 miR-30c and miR-301a affect PlGF-induced PAI-1 mRNA and PAI-1 protein levels

(A and C) HPMVECs were transfected with the indicated pre-miRNA precursor oligonucleotide. At 24 h post-transfection, cells were incubated in the presence (A) or in the absence (C) of human recombinant PlGF for 2 h. Total RNA was isolated using the TRIzol® reagent and were subjected to qRT-PCR analysis of PAI-1 and GAPDH mRNA levels. (B and D) Immunoblot analysis of PAI-1 protein in HPMVECs transfected with the indicated pre-miRNA precursor oligonucleotide and incubated for 3 h in the presence (B) or in the absence (D) of PlGF. The same membrane was stripped and reprobed for β-actin. (E) HPMVECs were transfected with the indicated pre-miRNA precursor oligonucleotide and the relative expression of the indicated miRNAs was estimated using the TaqMan MicroRNA assay. Results are the means±S.E.M. of triplicate determinations at each time point from three independent experiments. Indicated miRNA levels were normalized to 5S rRNA expression. *P<0.05, **P<0.01 and ***P<0.001. The results are representative of three independent experiments. neg., negative.

In addition, we investigated the effect of pre-miR-30c, premiR-301a or pre-miR-99a overexpression on endogenous PAI-1 mRNA and PAI-1 protein expression, in the absence of PlGF stimulation of HPMVECs. As shown in Figure 3(C), transfection of pre-miR-30c and pre-miR-301a singly resulted in a significant reduction in PAI-1 mRNA expression by approx. 120 and 70% respectively compared with untransfected cells. However, overexpression of pre-miR-99a or control precursor molecules had no effect on PAI-1 mRNA level. The effect of precursor molecules for the above miRNA candidates on PAI-1 expression at the protein level was confirmed by Western blotting. PAI-1 levels were significantly reduced in cells transfected with either pre-miR-30c or pre-miR-301a, but not with pre-miR-99a, when compared with cells transfected with negative control precursor molecules (Figure 3D). These results established that miR-30c and miR-301a regulate endogenous levels of PAI-1 expression in HPMVECs. We confirmed the overexpression of each of these pre-miRNAs (i.e. miR-30a, miR-301a and miR-99a) in HPMVECs. As shown in Figure 3(E), transfection of HPMVEC with pre-miR-30c, pre-miR-301a and pre-miR-99a expression plasmids resulted in 10-, 12- and 15-fold increases respectively in the corresponding miRNA levels. Taken together, the results show that miR-30c and miR-301a, but not miR-99a, reduced endogenous basal and PlGF-induced PAI-1 mRNA and PAI-1 protein expression.

Down-regulation of miR-30c and miR-301a levels in plasma from SCA patients

We have shown previously that levels of PlGF are elevated in SCA patients [9], and increased PAI-1 expression in SCA is transcriptionally up-regulated by PlGF via HIF-1α and AP-1 [9]. Our present study shows that PlGF additionally down-regulates miR-30c and miR-301a expression. Hence we wanted to determine whether levels of miR-30c and miR-301a were reduced in plasma obtained from SCA patients. In order to identify any correlation between the elevated levels of PAI-1 and miRNA concentrations in plasma of subjects with SCA, we used the same set of plasma samples for both SCA patients and healthy controls that was used previously for PAI-1 analysis [9]. According to the previous study [9], plasma levels of PAI-1 were significantly higher in SCA patients compared with controls (3.7±1.5 compared with 17±3.2 ng/ml respectively). In the present study, we found that circulating levels of miR-30c, miR-301a and miR-99a were indeed detectable in samples from both SCA patients (n=6) and healthy controls (n=6). The concentrations of both miR-30c and miR-301a were significantly lower, by approx. 13- and 3.8-fold respectively, in plasma from SCA patients when compared with controls (Figure 4). However, there was no significant difference in the concentration of miR-99a between SCA patients and controls (Figure 4). Thus our results show an inverse correlation between the levels of miR-30c and miR-301a and PAI-1 in SCA.

Figure 4 Estimation of miR-30c, miR-301a and miR-99a levels in plasma samples from healthy controls and SCA patients

The plasma concentrations of (A) miR-30c, (B) miR-301a and (C) miR-99a were quantified from healthy controls (n=6) and SCA patients (n=6) using specific the TaqMan MicroRNA assay for each miRNA. The relative expression levels of individual miRNAs (log10 scale on the y-axis) were normalized to RNU6B expression. Each data point represents values from a single subject. The mean value for healthy controls and SCA patients is displayed and represented as a (●).

DISCUSSION

PlGF is a member of the VEGF (vascular endothelial growth factor) family of angiogenic growth factors and was originally shown to be released from placental trophoblast cells and umbilical vein endothelial cells [36]. However, studies have shown that PlGF is also produced by bone marrow erythroid cells, but not by other mature haemopoietic cells [37]. The levels of PlGF are significantly elevated in patients with chronic haemolytic anaemia, such as in SCA and β-thalassaemia, due to increased compensatory erythropoiesis in response to anaemia [38,39]. Our previous studies have identified the role of PlGF in the increased expression of many other genes, such as cytochemokines [38], 5-lipoxygenase-activating protein for leukotriene synthesis [29] and endothelin-1 and its receptor [39]. These genes are involved in different pathologies manifested by SCA, including pulmonary hypertension [40] and lung injury.

Recently, we have demonstrated that PlGF-induced PAI-1 mRNA in HPMVECs via transcriptional mechanisms, which involved activation of HIF-1α and AP-1 complex binding to cis-elements proximal to the promoter of PAI-1 [9]. In addition, we made the observation that PlGF increased PAI-1 mRNA levels as early as 1 h post-stimulation with PlGF, which probably involved post-transcriptional mechanisms such as regulation of mRNA stability. In the present study, we ruled out the possibility of transcriptional regulation of PlGF-mediated PAI-1 mRNA accumulation at early time intervals (1–2 h) in HPMVECs. PlGF failed to drive PAI-1 promoter–luciferase reporter activity at 1 and 2 h post-induction; by contrast, a 5–6-fold increase in luciferase reporter activity was observed at 4 and 6 h of treatment. Moreover, PlGF induced PAI-1 mRNA expression at 2 h was not reduced in cells transfected with HIF-1α siRNA or c-Jun siRNA, whereas a significant reduction in PAI-1 mRNA was observed at 4 and 6 h. The results of both experiments indicate the involvement of post-transcriptional mechanisms in PlGF-induced PAI-1 mRNA expression during early induction. Our results from actinomycin D inhibition during PlGF induction confirmed an increase in PAI-1 mRNA stability by 2-fold at 2 h post-stimulation. The kinetic data of PAI-1 mRNA stability showed that in untreated cells there was a biphasic decline in PAI-1 mRNA, indicative of a second-order degradation process, consistent with an miRNA-dependent mechanism. By contrast, the first-order nature of PAI-1 mRNA decay in PlGF-treated cells suggested that under these conditions the turnover of PAI-1 mRNA was dependent only on mRNA concentration, subject to normal turnover of bulk cytoplasmic mRNA.

The regulation of PAI-1 expression is mainly achieved at the transcriptional level on the basis of the presence of several binding sites for transcription factors within its promoter [810]. By contrast, the mechanisms involved in the post-transcriptional regulation of PAI-1 are poorly understood. Physiological stimuli such as TGF-β [12] and insulin [11] have been shown to regulate PAI-1 levels by the involvement of post-transcriptional processes. PAI-1 mRNA harbours several 3′-AREs, which are involved in rapid mRNA destabilization and degradation [41]. cAMP has been shown to regulate PAI-1 mRNA stability through its AREs; however, the identities of the putative mRNA-binding proteins remain unknown [42]. A recent study identified the role for the ubiquitous RNA-binding protein human-antigen R in AngII (angiotensin II)-mediated PAI-1 destabilization by binding to the ARE in a rat model of kidney fibrosis [43]. Nonetheless, the role of miRNAs in PAI-1 mRNA stability remains to be established.

miRNAs play an important role in the stabilization and degradation of mRNAs [1416,44]. In the present study, we successfully identified candidate miRNAs that showed complementarity to binding sites within the 3′-UTR of PAI-1 mRNA. The miRNA database revealed approx. 20 candidate miRNA-binding sites in the PAI-1 mRNA. We selected five miRNA candidates (miR-30c, miR-301a, miR-99a, miR-299-5p and miR-609), which had the highest degree of species conservation with respect to potential target sites in the PAI-1 mRNA 3′-UTR, for experimental validation. In order to determine which of these candidate miRNAs actually regulated the expression of PAI-1 mRNA, we first determined the level of expression of these miRNAs in HPMVECs following PlGF treatment after 2 h. The levels of miR-30c and miR-301a were significantly reduced in response to PlGF treatment, compared with untreated cells. The presence of miR299-5p and miR-609 was not detected in the assay of HPMVECs under these conditions, whereas miR-99a showed no change. We addressed further the role of endogenous PlGF in the physiological regulation of these miRNAs by studying the consequences of reducing PlGF expression in the cells using PlGF siRNA. Our results showed increased levels of miR-30c and miR-301a in cells transfected with PlGF siRNA, supporting the concept that endogenous PlGF regulates basal expression of miR-30c and miR-301a. Thus three miRNA candidates (miR-30c, miR-301a and miR-99a) were chosen for further study to determine whether they had a biological function in mediating post-transcriptional repression of PAI-1 expression. The effect of miRNA down-regulation on PAI-1 mRNA was examined by transfection of HPMVECs with anti-miR-30c and anti-miR-301a. This resulted in augmented basal PAI-1 mRNA expression, whereas anti-miR-99a had no such effect. Consistent with these results, overexpression of these miRNAs in HPMVECs was performed by transfection with pre-miR-30c, pre-miR-301a or pre-miR-99a singly. In the absence of PlGF, overexpression of miR-30c and miR-301a, but not of miR-99a, resulted in reduced PAI-1 expression. In PlGF-induced cells, PAI-1 mRNA and PAI-1 protein expression induction was antagonized by overexpression of miR-30c and miR-301a.

Luciferase reporter assays were performed to confirm the predicted target sites of miR-30c and miR-301a on PAI-1 mRNA. The effect of individual miRNAs on luciferase reporter activity was antagonized by corresponding mutations within the PAI-1 3′-UTR of the luciferase reporters. A double mutation of the PAI-1 3′-UTR construct for both miR-30c- and miR-301a-binding sites completely abrogated their destabilizing effect on reporter activity, indicating the involvement of both miRNAs in regulating PAI-1 mRNA stability. However, as expected, mutation of the putative miR-99a target site had no effect on reporter activity. Thus our present study shows that miR-30c and miR-301a target the PAI-1 3′-UTR in a direct and sequence-specific manner. Our results were supported further by miRNA overexpression; miR-30c and miR-301a, but not miR-99a, efficiently reduced reporter expression whereas corresponding anti-miRNA oligonucleotides antagonized endogenous miRNA activity and allowed up-regulation of luciferase reporter activity. Our findings are in accordance with those of Shanmugam et al. [44], who reported that S100b-mediated COX-2 (cyclo-oxygenase-2) mRNA stabilization at early time points occurred by the reduction in miR-16 expression in THP-1 monocytic cells. Our results also parallel the miR-409–3p-mediated regulation of FBG (fibrinogen β) mRNA expression, another important molecule in the blood coagulation cascade [45]. Moreover, studies are now emerging regarding the cellular signalling mechanisms and transcription factors involved in expression of miRNAs [46], which would provide another avenue for fine-tuning gene expression in cellular processes at the post-transcriptional level.

To understand the regulation of miR-30c and miR-301a, we examined their gene loci in the human genome (Genome Reference Consortium, version 37). miR-30c is located within a 3′-proximal intron of the NFYC gene (nuclear transcription factor Yγ; a transcription factor) on chromosome 1 in p34.2, whereas miR-301a is located in the first intron of the SKA2 (spindle and kinetochore-associated complex subunit 2) gene located on chromosome 17 at q22. The locations and orientation of miR-30c and miR-301a are depicted in gene schematic diagrams shown in Supplementary Figures S3(A) and S3(B) respectively (http://www.BiochemJ.org/bj/434/bj4340473add.htm). These show that both miRNAs can be co-synthesized from the respective NFYC and SKA2 transcription units. Thus, in the absence of any post-transcriptional regulation, synthesis of these miRNAs would be affected by transcriptional regulators of NFYC and SKA2. Indeed miR-301 has been reported to be expressed from an SKA2 transcript and is observed to be a transcriptional regulator of SKA2 itself [47]. SKA2 protein is involved in mitotic spindle checkpoint silencing [47]. Thus it will be of interest to determine how miR-301a and SKA2 are co-regulated by cis-elements in the SKA2 promoter.

NF-Y, a multimeric transcription factor, is regulated by cellular redox state and plays an important role in the regulation of MHC class II gene expression [48]. NF-Yγ, one of the subunits of NF-Y, recognizes a CCAAT box motif found in promoter and enhancer elements and functions as a co-repressor of agonist-bound mineralocorticoid receptor [49]. An in silico analysis of the NFYC intron, where miR-30c and miR-30e are co-located, indicates the presence of several cryptic promoters and possible transcription factor binding sites (Supplementary Figure S3A). Thus it is possible that both miR-30c and miR-30e originate from a single transcription unit and are regulated independently of NFYC. Further studies are warranted to determine the synthetic origin of miR-30c, i.e. whether it originates from the NFYC transcription unit or from a smaller intronic transcription unit encoding miR-30e and miR-30c.

In conclusion, PAI-1 expression can be regulated by a number of distinct transcription and post-transcriptional mechanisms. Our studies show that PlGF significantly reduces the levels of miR-30c and miR-301a in HPMVECs. The findings obtained with inhibitors or mimics for miR-30c and miR-301a suggest that these miRNAs specifically promote degradation of PAI-1 mRNA in normal cells, whereas PlGF stabilizes PAI-1 mRNA expression by inhibiting the expression of miR-30c and miR-301a. At present, we do not know whether PlGF increases degradation of these miRNAs (miR-30c and miR-301a) or decreases their transcription to stabilize PAI-1 mRNA. Since a single miRNA potentially has many targets, it is possible that these identified miRNAs affect other distinct signalling molecules in signalling pathways, thus potentially regulating the expression of PAI-1. Because increased levels of PAI-1 in SCA can be one of the factors that contribute to a pro-thrombotic state, potentially predisposing patients to stroke and pulmonary hypertension, our studies have uncovered a novel mechanism for the increased expression of PAI-1.

One possible scenario that may lead to abnormally high levels of PAI-1 in SCA would be that PlGF elaborated from erythroid cells keeps levels of miR-30c and miR-301a low. Indeed, our results in plasma of SCA patients demonstrated a significant down-regulation of miR-30c and miR-301a, but not of miR-99a, compared with healthy controls. In the same set of plasma, we observed a reciprocal relationship between PAI-1 and miR-30c and miR-301a levels in SCA. It will be interesting to validate these results in a larger sample population in order to develop these miRNAs as biomarkers for hypercoagulability in SCA patients. A precedent for this has been observed for specific miRNAs that are secreted during tumorigenesis [27]. Future studies in this direction are warranted. As a final thought, RNAi (RNA interference)-based therapeutic approaches may be useful in ameliorating stroke and pulmonary hypertension in patients with SCA. Clinical trials using this approach for the purpose of attenuating viral infections and cancer in humans have ensued [50], suggesting the feasibility of a similar approach in SCA.

AUTHOR CONTRIBUTION

Nitin Patel and Vijay Kalra devised the study and wrote the manuscript. Nitin Patel performed all of the experimental work. Stanley Tahara analysed the genomic locations of miR-30c and miR-301a. Punam Malik collected samples from the SCA patients and normal controls. All authors edited the manuscript prior to submission.

FUNDING

This work was partially supported by the National Institutes of Health [grant numbers CSCC-HL-070595, HL-079916] in addition to institutional support.

Acknowledgments

We thank Dr Trine Fink (University of Aalborg, Aalborg, Denmark) for kindly providing the PAI-1 promoter plasmid, Dr Gangning Liang (Keck School of Medicine, University of Southern California, Los Angeles, CA, U.S.A.) for providing the TaqMan MicroRNA assay for RNU6B, and the Hematology Repository at Cincinnati Children's Hospital Medical Center for help with collection of patient blood samples and appropriate controls. We also thank the Institutional Core of the University of Southern California Research Center for Liver Disease for the use of a sequence detection instrument (supported by the National Institutes of Health [grant number P30-DK 048522]).

Abbreviations: AP-1, activator protein-1; ARE, adenosine- and uridine-rich element; EBM-2, endothelial basal medium-2; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; HIF-1α, hypoxia-inducible factor-1α; HPMVEC, human pulmonary microvascular endothelial cell; miRNA (miR), microRNA; NF-Y, nuclear transcription factor Y; ns, not significant; PAI-1, plasminogen activator inhibitor-1; PlGF, placental growth factor; qRT-PCR, quantitative real-time PCR; RNU6B, U6 small nuclear RNA; RQ, relative quantitative; SCA, sickle cell anaemia; scRNA, scrambled RNA; siRNA, small interfering RNA; SKA2, spindle and kinetochore-associated complex subunit 2; TGF-β, transforming growth factor-β; UTR, untranslated region

References

View Abstract