High Concentrations of Uric Acid Inhibit Angiogenesis via

Circulation Journal Vol.79, November 2015

Circulation JournalOfficial Journal of the Japanese Circulation Societyhttp://www.j-circ.or.jp

including development, cell proliferation, migration, adhesion, apoptosis, and differentiation.8 Several studies have reported that miRNAs play key roles in angiogenesis and cardiovascu-lar disease.9–12 However, whether miRNAs regulate the angio-genesis of endothelial cells under HUA is not known.

In the present study, we first assessed the effect of HUA on angiogenesis. Then, we performed a miRNA expression pro-filing analysis using microarrays in control and uric acid-stimulated endothelial cells. Differentially expressed miRNAs were subjected to bioinformatic analyses, and target miRNAs were identified according to the results. Finally, we predicted and validated the downstream pathway.

s diet components and lifestyles change, hyperurice-mia morbidity is increasing annually.1–3 Hyperurice-mia tends to cause gout and nephrolithiasis, and it is

an independent risk factor for cardiovascular disease and chronic kidney disease, in addition to hypertension, diabetes, and obesity.4–7 Angiogenesis is one of the most important hallmarks of endothelial cells and provides the basis for prop-erly functioning organs. Impaired angiogenesis is a critical cause of cardiovascular disease. Whether high concentrations of uric acid (HUA) affect angiogenesis and the underlying mechanisms of such an affect remain unclear. MicroRNAs (miRNAs) are small, non-coding RNAs that regulate a wide range of physiological and pathophysiological processes,

A

Received March 15, 2015; revised manuscript received July 10, 2015; accepted July 17, 2015; released online August 21, 2015 Time for primary review: 21 days

State Key Laboratory of Kidney Diseases, Department of Nephrology, Chinese PLA Institute of Nephrology, National Clinical Research Center for Kidney Disease, Chinese PLA General Hospital, Beijing (S.Y., Q.H., Y.W., K.H., L.W., Y. Zhang, B.F., Y. Zhuo, W.Z., X.C., D.W.); Department of Cardiology, First Hospital, Peking University, Beijing (S.Y.), People’s Republic of China

The first two authors contributed equally to this work (S.Y., Q.H.).Mailing address: Di Wu, PhD, Department of Nephrology, Chinese PLA General Hospital, State Key Laboratory of Kidney Disease, Beijing

100853, People’s Republic of China. E-mail: wudi@301 hospital.com.cnISSN-1346-9843 doi: 10.1253/circj.CJ-15-0283All rights are reserved to the Japanese Circulation Society. For permissions, please e-mail: [email protected]

High Concentrations of Uric Acid Inhibit Angiogenesis via Regulation of the Krüppel-Like Factor 2-Vascular

Endothelial Growth Factor-A Axis by miR-92aShandong Yu, PhD; Quan Hong, PhD; Yuanda Wang, MD; Kai Hou, MD; Liyuan Wang, MD;

Yang Zhang, MD; Bo Fu, MD; Yunan Zhou, PhD; Wei Zheng, MD; Xiangmei Chen, PhD; Di Wu, PhD

Background: Angiogenesis is a critical component of many pathological conditions, and microRNAs (miRNAs) are indispensable in angiogenesis. It is unclear whether miRNAs regulate angiogenesis in the presence of high concen-trations of uric acid (HUA), and the underlying mechanisms remain unknown.

Methods and Results: It was found that HUA inhibited the angiogenic ability of endothelial cells. miRNA expression profiling was conducted using microarray assays in HUA-stimulated endothelial cells. Eighteen differentially expressed miRNAs were subjected to bioinformatic analyses. The results indicated that miR-92a was negatively regulated and was closely related to angiogenesis. Furthermore, the effects of miR-92a on HUA-stimulated endo-thelial cell angiogenesis and the underlying mechanisms were investigated in dual-luciferase reporter assays, electrophoretic mobility shift assays, immunoblot assays, and tube formation assays. It was determined that Krüppel-like factor 2 (KLF2) is a target gene of miR-92a, and KLF2 binds the vascular endothelial growth factor-A (VEGFA) promoter to inhibit its expression. miR-92a and VEGFA overexpression or KLF2 downregulation alleviates the HUA-mediated inhibition of angiogenesis in endothelial cells in vitro.

Conclusions: This study reported that there is a novel pathway regulating angiogenesis under HUA conditions. In the presence of HUA, miR-92a downregulation increased KLF2 expression, subsequently inhibiting VEGFA, which resulted in decreased angiogenesis. Thus, this study reports a possible mechanism for cardiovascular injury caused by hyperuricemia and suggests that the miR-92a-KLF2-VEGFA axis may be a target for hyperuricemia treat-ment. (Circ J 2015; 79: 2487 – 2498)

Key Words: Angiogenesis; KLF2; miR-92a; Uric acid; VEGFA

ORIGINAL ARTICLEVascular Biology and Vascular Medicine


2488 YU S et al.

is the key functional classification of National Centre for Biol-ogy Information (NCBI).13,14 In general, Fisher’s exact test and the χ2 test were used to classify the GO category, and the false discovery rate (FDR)15 was calculated to correct the P value; the smaller the FDR, the smaller the error in judging the P value. The FDR was defined as

FDR=1– Nk

T,

where Nk refers to the number of Fisher’s test P values that are below χ2 test P values, and T refers to the test statistics. We computed P values for the GOs of all of the differentially expressed genes. Enrichment provides a measure of the sig-nificance of the function; as the enrichment increases, the corresponding function is more specific, which helped us to identify those GOs with more concrete functional descriptions in the experiment. Within the significant category, the enrich-ment, Re, was given by Re=(nf/n)/(Nf/N), where nf is the num-ber of differentially expressed genes within the particular category; n is the total number of genes within the same cat-egory; Nf is the number of differentially expressed genes in the entire microarray; and N is the total number of genes in the microarray.16

miRNA-Gene NetworkThe relationship between the miRNAs and genes was quanti-fied based on their differential expression values and accord-ing to the interactions of the miRNAs and genes in the Sanger miRNA database to build a miRNA-gene network. The adjacency matrix of miRNAs and genes, A=[ai,j], was gener-ated based on the attribute relationships among the genes and miRNAs, where ai,j represents the relationship weight for gene i and miRNA j. In the miRNA-gene network, a circle repre-sents a gene, and a square represents a miRNA, with their relationship being represented by one edge. The center of the network was represented by a degree. The degree indicates the contribution of one miRNA to the genes nearby or the contri-bution of one gene to the miRNAs nearby. The key miRNAs and genes in the network always exhibit the greatest number of degrees.

miRNA-GO NetworkThe miRNA-GO network was built according to the relation-ships of significant GOs and genes, and the relationships among miRNAs and genes. The adjacency matrix of miRNAs and genes, A=[ai,j], was generated based on the attribute rela-tionships among GOs and miRNAs, where ai,j represents the relationship weight of GO i and miRNA j. In the miRNA-GO network, a circle represents a GO, and a square represents an miRNA, with their relationship being represented by one edge. The center of the network was represented by a degree. The degree indicates the contribution of one miRNA to the GOs nearby or the contribution of one GO to the miRNAs nearby. The key miRNAs and GOs in the network always exhibit the largest number of degrees.

Patient Selection, Serum Preparation and RNA ExtractionMale hyperuricemia patients without treatment (n=30) and healthy male age-matched individuals (n=30) were selected to detect miR-92a levels. All of the patients and healthy indi-viduals gave their written informed consent. Demographic and clinical data (eg, biochemical results and duration of disease) were obtained at the time of enrolment. Patients with cardio-vascular disease, diabetes, and kidney disease were excluded (Table 1). The study conformed to the principles outlined in the Declaration of Helsinki and was approved by the local

MethodsUric Acid PreparationUric acid powder was dissolved in a 1 mol/L NaOH solution at a concentration of 40 mmol/L. Then, the uric acid solu-tion was added to the medium at a final concentration of 600 μmol/L, and at a pH 7.2–7.4.

Cell CultureHuman umbilical vein endothelial cells (HUVECs) were pur-chased from the American Type Culture Collection (ATCC) (No. CRL-1730) and cultured in RPMI-1640 medium supple-mented with 10% fetal bovine serum at 37°C in a humidified incubator in a 5% CO2 atmosphere.

Tube Formation AssayHUVECs (1×105) were cultured in a 12-well plate (Greiner) coated with 200 µl of Matrigel Basement Membrane Matrix (BD Biosciences) and divided into control and high uric acid groups (600 μmol/L). Tubes were observed and measured in microscopic fields after 24 h with a computer-assisted micro-scope (Axiovert 100 M equipped with an AxioCam camera; Carl Zeiss, Jena) using Axiovision 4.6 (Carl Zeiss, Imaging Solutions GmbH, Munich).

Microarray Analysis of miRNA Expression and Real-Time Polymerase Chain Reaction (RT-PCR) ValidationCells were seeded in 100-mm dishes and divided into control and HUA groups (600 μmol/L). The cells in the HUA group were cultured in medium containing 600 μmol/L uric acid for 48 h. Total RNA was isolated by using the miRNeasy Mini Kit (QIAGEN) using HUVECs exposed to HUA (600 μmol/L) or normal medium for 48 h. A microarray assay was performed using a service provider (LC Sciences). Briefly, total RNA samples were size fractionated, and small RNAs (<300 nt) were 3’-extended with a poly(A) tail. An oligonucleotide tag was then ligated to the poly(A) tail for fluorescent dye stain-ing; 2 different tags (for Cy3 and Cy5 dyes) were used in dual-sample experiments. Hybridization was performed over-night on an uParaflo microfluidic chip (Chip ID miRHuman 12.0 version; LC Sciences). After the fluorescence images were collected, ratios (Cy3/Cy5, log2 transformed, balanced) and P values were calculated using the Student’s t-test. Sig-nificant signals were considered for those showing P<0.05. We then validated the array data via quantitative PCR (qPCR). Briefly, isolated total RNA was poly-adenylated and reverse transcribed for use in a 2-step qPCR using the NCode miRNA First-Strand cDNA Synthesis and qPCR kits (Invitrogen). The resulting cDNA was subjected to qPCR using the NCode universal reverse primer, in conjunction with a sequence-specific forward primer for selected miRNAs. A master mix was prepared for each amplification reaction, which included SYBR GreenER qRT-PCR SuperMix, the forward primer, the Universal qPCR Primer, ROX reference dye, and template cDNA. RNU6B was used as the internal control for HUVEC miRNAs and cel-mir-39 as a control for serum miRNAs. The reactions were monitored using a preheated real-time instru-ment (ABI StepOne Plus). The PCR conditions were 2 min at 50°C and 10 min at 95°C, followed by 40 cycles of 95°C for 4 s and 57°C for 30 s. The primers used in these assays are shown in Table S1.

Gene Ontology (GO) AnalysisGO analysis was applied to analyze the main function of the genes showing differential expression according to GO, which


2489HUA Inhibits Angiogenesis via miR-92a-KLF2-VEGFA

tubes. Total RNA was isolated from 1 ml of serum using the TRIzol® LS reagent (Life Technologies, Carlsbad, CA, USA) and precipitated with isopropanol containing GlycoBlueTM, according to the manufacturer’s instructions (Ambion,

Ethics Committee. Venous blood was incubated at room tem-perature for 1 h, followed by centrifugation at 500×g for 10 min and then 10,000×g for 30 min at 4°C to separate the serum. The supernatant was transferred to fresh RNase-free

Table 1. Demographic and Clinical Data of Patients and the Control Group

Items Hyperuricemia patients Control

N 30 30

Gender Male Male

Age (years) 46.67±2.90 46.84±3.56

Serum uric acid (μmol/L) 519.68±64.42 351.24±48.26

SBP (mmHg) 123±15 120±17

DBP (mmHg) 78±12 75±11

Duration of disease (months) 6.34±1.87

ALT (U/L) 22.10±4.85 24.32±5.86

AST (U/L) 27.10±5.14 28.54±5.79

Serum creatinine (μmol/L) 72.13±15.61 74.22±14.35

ALT, alanine aminotransferase; AST, aspartate aminotransferase; DBP, diastolic blood pressure; SBP, systolic blood pressure.

Figure 1. High concentrations of uric acid (HUA; 600 μmol/L) inhib-ited tube formation in human umbili-cal vein endothelial cells (HUVECs) in vitro, and differentially expressed microRNAs (miRNAs) between con-trol and HUA-stimulated HUVECs were validated. (A) HUVECs were incubated with 600 μmol/L uric acid or in normal media for 24 h, and tubes length were measured. Rep-resentative micrographs are shown (400×). The results are presented as mean ± SD. *P<0.05 vs. the control group, n=3. (B) The results are pre-sented as mean ± SD. *P<0.05 vs. the control group (one-way analysis of variance), n=3.


2490 YU S et al.

Fulen Gene Co, Guangzhou, China). The assay methodology was similar to that described above.

Western Blot AnalysisFor Western blot analysis, HUVECs were lysed in Radio-Immunoprecipitation Assay (RIPA) lysis buffer (Sigma) for 20 min on ice. After centrifugation for 15 min at 17,000×g (4°C), protein concentrations were determined using bicincho-ninic acid (BCA) protein assay kits (Vigorous Biotechnology). Equal amounts of protein were loaded onto SDS-polyacryl-amide gels and blotted onto polyvinylidene fluoride (PVDF) or nitrocellulose membranes. Western blotting was performed using antibodies directed against KLF2 and VEGFA (rabbit polyclonal; 1:500, incubated 12 h at 4°C; Abcam). Secondary antibodies (1:1,000, incubated 2 h at room temperature) were purchased from Jackson ImmunoResearch. The blots were developed with an enhanced chemiluminescent reagent (Santa Cruz) according to the manufacturer’s instructions and then exposed to X-ray film. Protein bands were quantified using Quantity One software (Bio-Rad). For KLF2 analysis, we extracted nucleoprotein by using Nuclear Extract kits (Active Motif, Carlsbad, CA, USA).

Electrophoretic Mobility Shift AssaysNuclear extracts from HUVECs were prepared using Nuclear Extract kits (Active Motif, Carlsbad, CA, USA). Protein concentrations were determined using BCA protein assay kits (Vigorous Biotechnology). The KLF2-binding site (5’-agCACCagcgctctgtcgggaggcgcagcggttaggtggaccggtcagc-ggactCACCgg-3’) or a mutant site (5’-agTGATagcgctctgtc-gggaggcgcagcggttaggtggaccggtcagcggactTGATgg-3’) was included in the 3’ biotin-labeled and unlabeled single-stranded oligonucleotides. Binding reactions were performed using the LightShift Chemiluminescent Electrophoretic Mobility Shift Assays (EMSA) kit (Pierce, Rockford, IL, USA). The DNA-protein complex samples were analyzed using 6% polyacryl-amide gels.

TransfectionFor overexpression of miR-92a, HUVECs were grown to 50% confluence. MiR-92a mimics, or a control miR (Ambion; concentration of 10 nmol/L) was transfected with, Lipo-fectamine RNAiMAX (Invitrogen) according to the manufac-turer’s protocol. For siRNA-mediated gene knockdown, HUVECs were grown to 60–70% confluence and transfected with GeneTrans II®. Cells were transfected with a validated siRNA targeting KLF2 (Qiagen, Hilden, Germany). A scram-bled siRNA was used as a control (5’- UCAAGAAGC-CAAGGAUAAU-3’). For overexpression of VEGFA, HUVECs were transfected with vectors containing VEGFA cDNA (Origene, Beijing, China) using LipofectamineTM 2000, according to the manufacturer’s protocol (Life Technologies). HUVECs were transfected at 40–60% confluence.

Measurement of VEGFA ConcentrationsSerum VEGFA levels were measured in an enzyme-linked immunosorbent assay for human VEGFA (Quantikine VEGF Immunoassay; R&D Systems, Minneapolis, MN, USA). All samples were analyzed in duplicate and mean values were calculated. Serum VEGFA levels were expressed in pg/ml.

Statistical AnalysisThe results are presented as mean ± SD. The analysis was performed by using IBM SPSS Statistics 17.0.2 software (IBM Corporation, Armonk, NY, USA). Multiple comparisons

Carlsbad, CA, USA). The quantity and quality of RNA were assessed using a NanoDrop ND 1000 (Thermo Fisher, Boston, MA, USA) and an Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA).

Plasmid Construction and Luciferase AssayThe fragment of the KLF2 (NM_016270.2) 3’-UTR contain-ing the miR-92a targeting sequence (GTGCAATA) was cloned into the psiCHECKTM-2 dual-luciferase reporter plas-mid (Promega, USA; cat. no. C8021), to produce psiCHECK-WT-KLF using PCR. The sequence 5’-GTGCAATA-3’ (1,394–1,403 bp) in the 3’-UTR of KLF2, the core binding sequence of miR-92a, was substitutively mutated to ACTACGGC and subcloned into psiCHECKTM-2; the result-ing plasmid was designated psiCHECK-MT-KLF. To carry out the site-directed mutagenesis of the miR92a targeting site, the QuikChange® Site-Directed Mutagenesis Kit (Agilent Technologies, USA; cat. no. 200519) was used.

For the reporter assays, HEK293 cells or HUVECs were cultured to approximately 80% confluence in a 24-well plate and then cotransfected with a dual-luciferase reporter plasmid (psiCHECK-WT-KLF or psiCHECK-MT-KLF) and either miR-92a mimics or miR-92a inhibitors (HEK293 cells) for 48 h. Firefly and Renilla luciferase activities were measured using a Dual-Luciferase Reporter Assay System (Promega cat. no. E1910), and Renilla luciferase activity was then normal-ized to firefly luciferase activity.

Additionally, the KLF2-binding site of vascular endothelial growth factor-A (VEGFA) (5’-agCACCagcgctctgtcgggaggc-gcagcggttaggtggaccggtcagcgga ctCACCgg-3’) or a mutant site (5’-agTGATagcgctctgtcgggaggcgcagcggttaggtggaccggt-cag cggactTGATgg-3’) was cloned into psiCHECKTM-2, and the plasmids psiCHECK-WT-VEGF and psiCHECK-MT-VEGF were obtained, which were subsequently transfected into HEK293 together with the KLF2 vector (purchased from

Table 2. Relative Level of Differential Expression of miRNAs in the HUA Group Compared to the Control Group

No. miRNAs Fold change (HUA/Control)

1 hsa-miR-663 2.23

2 hsa-miR-26b 1.65

3 hsa-miR-1308 −1.35　　 4 hsa-miR-1915 1.23

5 hsa-miR-1977 1.18

6 hsa-miR-638 1.08

7 hsa-miR-191 −0.80　　 8 hsa-miR-1979 0.77

9 hsa-let-7c 0.73

10 hsa-miR-24 −0.67　　11 hsa-miR-1978 0.64

12 hsa-miR-1975 0.62

13 hsa-miR-222 −0.54　　14 hsa-let-7f 0.41

15 hsa-miR-1826 0.40

16 hsa-miR-92a −0.40　　17 hsa-miR-1974 −0.35　　18 hsa-let-7a 0.29

The positive numbers stands for the over-expression of miRNAs in the HUA group, and the negative numbers stands for the down-expression of miRNAs in the HUA group. HUA, high concentrations of uric acid; miRNAs, microRNAs.



significantly upregulated in stimulated HUVECs compared with control HUVECs, whereas 6 miRNAs were expressed at lower levels during HUVEC stimulation (Table 2). The level of each of these 18 filtered miRNAs was verified to be sig-nificantly different between stimulated and control HUVECs by the use of qRT-PCR (Figure 1B, P<0.05).

miR-92a May Be a Positive Regulator Under HUA ConditionsIn the GO analysis, we found that angiogenesis was a highly enriched GO term targeted by both upregulated and down-regulated miRNAs (Table S2; line 1–line 161 upregulated; line 162–line 190, downregulated), and in the miRNA-GO network analysis, we found that the corresponding upregu-lated miRNAs were hsa-let-7a/7c/7f, miR-26b, miR-1977 and miR-663, miR-1915, while the downregulated miRNAs were hsa-miR-92a and miR-222 (Table S3; Figure 2). Among the upregulated miRNAs, although let-7 family miRNAs were predominant in angiogenesis regulation according to the GO analysis, the expression level of the let-7 family did not show

of parametric data were performed through one-way analysis of variance. P<0.05 was considered statistically significant for all analyses.

ResultsHUA Inhibits Angiogenesis In VitroTo investigate whether HUA functionally affects angiogenesis in vitro, we used a HUVEC network formation assay. In this angiogenesis assay, HUA inhibited tubule formation, as shown by reductions in tubule length and number (Figure 1A).

miRNA Expression Profiles and Qualitative Reverse Transcriptase-Polymerase Chain Reaction (qRT-PCR) ValidationmiRNAs play critical roles in angiogenesis. To investigate whether miRNAs regulate angiogenesis under HUA, we per-formed a miRNA expression profiling analysis using microar-rays in control- and uric acid-stimulated endothelial cells. Among the differentially expressed miRNAs, 12 miRNAs were

Figure 2. Micro RNA-gene ontology (miRNA-GO) network analysis of miRNA-microarray chip data. The miRNA-GO network analysis revealed the relationship of targeted GOs with the differential expression of miRNAs. Orange boxes indicate upregulated miRNAs, and blue boxes indicate downregulated miRNAs.


2492 YU S et al.

Figure 3. Micro RNA (miRNA)-gene network analyses of miRNA-microarray chip data and serum miR-92a levels in hyperuricemia patients and healthy individuals. (A) Serum miR-92a levels of healthy subjects and hyperuricemia patients. Eleven males were selected for each group. The ages of the individuals in the healthy and hyperuricemia groups were 50.6±7.3 years and 52.2±8.7 years, respectively, and their serum uric acid concentrations were 372.2±58.3 μmol/L and 521.4±87.4 μmol/L, respectively. Dots indicate the serum miR-92a levels of individuals. Data are presented as mean ± SD. *P<0.05 (analysis of variance), n=30. (B) The miRNA-gene network analysis revealed the relationship between targeted genes and differential miRNA expression. Orange boxes indicate upregulated miRNAs, and blue boxes indicate downregulated miRNAs.



Figure 4. Krüppel-like factor 2 (KLF2) is a direct target of miR-92a. (A) Conserved sites for miR-92a among mammalian KLF2 sequences (Upper panel). Reporter plasmids in which the luciferase coding sequence was fused to the 3’-UTR of KLF2 were transfected into HEK293 cells in conjunction with miR-92a mimics or miR-92a inhibitors. Renilla luciferase activity was normalized to firefly luciferase activity. The results are shown as mean ± SD. *P<0.05 vs. the control group, n=5 (Lower panel). (B) Human umbilical vein endothelial cells were incubated under high uric acid and normal conditions, and KLF2 protein levels were detected by immunoblotting after 48 h. *P<0.05 vs. the control group. #P<0.05 vs. the siRNA control group, n=6. HUA, high concentrations of uric acid.


2494 YU S et al.

Figure 5. Krüppel-like factor 2 (KLF2) inhibited vascular endothelial growth factor-A (VEGFA) by binding to its promoter. Human umbilical vein endothelial cells (HUVECs) were incubated with uric acid (UA: 600 μmol/L) for 48 h before EMSA and immunoblotting analyses. (A) Binding of KLF2 and VEGFA. From left to right, the samples are as follows: labeled probe, labeled probe/nuclear extract protein, labeled probe/nuclear extract protein+unlabeled probe, labeled probe/nuclear extract protein+mutant competitive probe, and labeled probe/nuclear extract+KLF2 antibody. KLF2 binding was eliminated in the presence of an unlabeled probe. A super-shifted band appeared in the presence of the anti-KLF2 antibody. (B) The luciferase activity of the wild-type group was significantly lower than that of control groups in the presence of KLF2. However, the luciferase activity of the mutant type group was similar to that of the controls in the presence of KLF2. *P<0.05 vs. the control groups, n=5. (C) KLF2 knockdown increased VEGFA expression under high UA conditions. HUVECs were transfected with siCon and siKLF2. Both KLF2 and VEGFA expression were detected by immunoblotting, 48 h after the addition of UA. Data are presented as mean ± SD. *P<0.05 vs. the control and siCon groups, #P<0.05 vs. the UA and UA+siCon groups, n=5.



Figure 6. A high uric acid concentration inhibits tube formation in human umbilical vein endothelial cells (HUVECs) via miR-92a regulation of Krüppel-like factor 2 (KLF2)-vascular endothelial growth factor-A (VEGFA). And the serum VEGFA levels of hyperuri-cemia patients were significantly lower than those of healthy individuals. (A) HUVECs transfected with miR-92a mimics and a miR-92a inhibitor. KLF2 and VEGFA expression was assayed by immunoblotting after 48 h of culture. Data are presented as mean ± SD. *P<0.05 vs. the control group, #P<0.05 vs. the miR-92a mimic group, n=5. (B) HUVECs were seeded into 12-well plates for tube formation assays, 12 h after transfection with miR-92a mimics, siKLF2, and VEGFA. Tubes were measured 48 h after incubation with uric acid (600 μmol/L). Representative micrographs are shown (400×). The results are presented as mean ± SD.*P<0.05 vs. the control group, #P<0.05 vs. the high concentrations of uric acid (HUA) group, n=3. (C) Serum concentra-tions were measured via an enzyme-linked immunosorbent assay for human VEGF according to the instructions given by the manufacturer. The VEGFA levels in the serum of healthy individuals were significantly lower than those of hyperuricemia patients. *P<0.05 vs. the control group, n=30.


2496 YU S et al.

expression and inhibited VEGFA expression, resulting in decreased angiogenesis in vitro. Immunoblot analyses showed that miR-92a overexpression reduced KLF2 expression, whereas VEGFA expression was increased (Figure 6A). miR-92a overexpression, KLF2 knockdown, or VEGFA overexpres-sion alleviated the inhibition of angiogenesis in HUVECs mediated by HUA (Figure 6B). Therefore, these results con-firmed our hypothesis. Furthermore, we determined the serum VEGFA levels of hyperuricemia patients and healthy indi-viduals, and we found that the serum VEGFA levels of hyper-uricemia patients were significantly lower than those of healthy individuals (Figure 6C).

Our study showed that miR-92a positively regulated angio-genesis under HUA condition. However, the research of Bonauer et al showed that miR-92a inhibited angiogenesis via integrin-α5 in an ischemic model.9 We speculated that miR-92a regulates angiogenesis via different pathways under dif-ferent conditions. So we determined the integrin-α5 expression level in both the control group and HUA group. There was no difference in integrin-α5 expression level between the control and HUA groups. In addition, we also determined the expres-sion level of another proangiogetic factor, angiopoietin-2, in the control and HUA groups. And the results showed that the expression level of angiopoietin-2 in the HUA group was similar to that of the control group (Figure S2).

DiscussionIt has been demonstrated that serum uric acid levels are posi-tively correlated with the prognosis of metabolic syndrome, diabetes, primary hypertension, atherosclerosis, and chronic kidney disease.20–23 It has also been reported that the patho-physiological changes caused by hyperuricemia do not depend completely on the deposition of uric acid crystals.24 In fact, it is cell damage caused by uric acid that plays a critical role. Angiogenesis is an important endothelial cell function. How-ever, the effect of HUA on angiogenesis in endothelial cells remains unclear. Previous work demonstrated that HUVECs incubated in the presence of 600 μmol/L uric acid show differ-ent reactions compared with normal conditions, and that this concentration of uric acid does not affect the viability of HUVECs.24–26 We therefore selected 600 μmol/L as a high level of uric acid. miRNAs are involved in a wide variety of physiological and pathophysiological processes, and angio-genesis also plays an important role in these processes.9–12 However, there has been no research regarding the effect of miRNAs on angiogenesis under HUA.

The expression of many genes participating in pathophysi-ological processes is altered in the presence of HUA. In this study, microarray analyses were used to identify differentially expressed miRNAs. We focused on miRNAs related to angio-genesis under HUA conditions according to bioinformatic analyses. Among the upregulated miRNAs, the predominant miRNAs (ie, let-7 family members) did not show a significant difference in serum between hyperuricemia patients and healthy individuals; however, one of the downregulated miRNAs, miR-92a, which has been reported to regulate angio-genesis under certain conditions, exhibited lower levels in the serum of hyperuricemia patients than in healthy individuals. We therefore focused on this miRNA and demonstrated the underlying mechanism. We concluded that HUA negatively regulates miR-92a, which increases KLF2 expression to inhibit VEGFA, resulting in reduced angiogenesis. Bonauer et al has reported that miR-92a interacts with integrin-α5 to inhibit angiogenesis under ischemic conditions.9 Given this

a significant difference in the serum of hyperuricemia patients vs. healthy individuals (Figure S1). Among the downregu-lated miRNAs, we noted that under HUA, miR-92a down-regulation was accompanied by repression of angiogenesis, which indicated that miR-92a might be a positive regulator of angiogenesis under these conditions. However, this finding was contradictory to previously reported results. In 2009, Bonauer et al reported that miR-92a inhibited angiogenesis in a limb ischemia model,9 and Hinkel et al also reported that miR-92a inhibition alleviated ischemic damage to the heart.11 Therefore, we focused on miR-92a, and we suspected that HUA might inhibit angiogenesis by negatively regulating miR-92a; specifically, miR-92a might promote angiogenesis under conditions involving HUA. In addition, serum miR-92a levels in hyperuricemia patients were lower than that in healthy individuals (Figure 3A). This result prompted us to investigate the mechanism by which miR-92a regulates angio-genesis under HUA.

miR-92a Affects the Expression and Function of KLF2 In VitroAmong the miR-92a target genes predicted in the miRNA-gene network analysis (Figure 3B), KLF2 has been demonstrated to inhibit angiogenesis.17 We hypothesized that miR-92a regulates KLF2 to influence angiogenesis in HUVECs. To determine whether miR-92a regulates KLF2 expression via binding to sequences in its 3’-UTR, we pre-dicted miR-92a consensus sequences in the KLF2 3’-Untrans-lated Regions (UTR). To validate these putative sequences and determine whether they directly contribute to the regula-tion of KLF2 expression, we utilized dual-luciferase reporter gene assays. We confirmed that miR-92a regulates KLF2 expression by binding to sequences in its 3’-UTR. Cotransfec-tion with miR-92a in HEK293 cells reduced the activity of the reporter gene vector containing wild-type miR-92a targeting sequences (psiCHECK-WT-KLF2). However, cotransfection with miR-92a inhibitors increased the activity of the reporter gene vector. Notably, miR-92a had no effect on the reporter activity of mutated vectors (psiCHECK-MT-KLF2; Figure 4A). These results demonstrated that miR-92a directly regulates KLF2. Additionally, we found that KLF2 protein expression was elevated under HUA (Figure 4B).

KLF2 Regulates VEGFA by Binding to Its Promoter RegionVEGFA is a strong proangiogenic factor,18 and the gene encoding VEGFA is downstream of KLF2.19 Therefore, we hypothesized that KLF2 might inhibit angiogenesis via down-regulating VEGFA under HUA. First, we performed EMSAs and dual-luciferase reporter gene assays to determine whether KLF2 regulates VEGFA directly. As shown in Figure 5A, a super-shifted band appeared after the addition of the KLF2 antibody. In the dual-luciferase reporter gene assays, lucifer-ase activity was significantly lower in the wild-type group than in the control groups in the presence of KLF2. However, the luciferase activity of the mutant type group was similar to that of the controls in the presence of KLF2 (Figure 5B). These results indicated binding of KLF2 to the VEGFA pro-moter. In addition, VEGFA expression increased after KLF2 knockdown under HUA (Figure 5C). These results demon-strated that KLF2 inhibits VEGFA expression under HUA.

HUA Inhibits Angiogenesis by Negatively Regulating miR-92aAccording to the above results, we hypothesized that under HUA, miR-92a was downregulated, which increased KLF



AcknowledgmentsThis work was supported by the Chinese National Natural Sciences Foun-dation (No. 31170810, No. 81470949, and No. 81102673), the Beijing NOVA Program (Z121107002512078) and the Major State Basic Research Development Program of China(2013CB530800). The microar-ray data were analyzed by Genminix Informatics (Shanghai, China).

DisclosuresAll authors declare that there is no conflict of interest associated with their contribution to this article.

References 1. Robinson PC, Taylor WJ, Merriman TR. Systematic review of the

prevalence of gout and hyperuricaemia in Australia. Intern Med J 2012; 42: 997 – 1007.

2. Crittenden DB, Pillinger MH. The year in gout: 2011–2012. Bull NYU Hosp Jt Dis 2012; 70: 145 – 151.

3. Zhu Y, Pandya BJ, Choi HK. Prevalence of gout and hyperuricemia in the US general population: The National Health and Nutrition Examination Survey 2007–2008. Arthritis Rheum 2011; 63: 3136 – 3141.

4. Johnson RJ, Kang DH, Feig D, Kivilghn S, Kanellis J, Watanabe S, et al. Is there a pathogenetic role for uric acid in hypertension and cardiovascular and renal disease. Hypertension 2003; 41: 1183 – 1190.

5. Feig DI, Mazzali M, Kang DH, Nakagawa T, Price K, Kannelis J, et al. Serum uric acid: A risk factor and a target for treatment. J Am Soc Nephrol 2006; 17(4 Suppl 2): S69 – S73.

6. Feig DI, Kang DH, Johnson RJ. Uric acid and cardiovascular risk. N Engl J Med 2008; 359: 1811 – 1121.

7. Numa S, Hirai T, Nakagawa K, Ohara K, Fukuda N, Nozawa T, et al. Hyperuricemia and transesophageal echocardiographic thrombo-embolic risk in patients with atrial fibrillation at clinically low-intermediate risk. Circ J 2014; 78: 1600 – 1605.

8. Bartel DP. MicroRNAs: Genomics, biogenesis, mechanism, and function. Cell 2004; 116: 281 – 297.

9. Bonauer A, Carmona G, Iwasaki M, Mione M, Koyanagi M, Fischer A, et al. MicroRNA-92a controls angiogenesis and functional recov-ery of ischemic tissues in mice. Science 2009; 324: 1710 – 1713.

10. Dang LT, Lawson ND, Fish JE. MicroRNA control of vascular endothelial growth factor signaling output during vascular develop-ment. Arterioscler Thromb Vasc Biol 2013; 33: 193 – 200.

11. Hinkel R, Penzkofer D, Zühlke S, Fischer A, Husada W, Xu QF, et al. Inhibition of MicroRNA-92a protects against ischemia/reperfu-sion injury in a large-animal model. Circulation 2013; 128: 1066 – 1075.

12. Yamada N, Nakagawa Y, Tsujimura N. Role of intracellular and extracellular MicroRNA-92a in colorectal cancer. Transl Oncol 2013; 6: 482 – 492.

13. Gene Ontology Consortium. The Gene Ontology (GO) project in 2006. Nucleic Acids Res 2006; 34: D322 – D326.

14. Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, Cherry JM, et al. Gene ontology: Tool for the unification of biology. The Gene Ontology Consortium. Nat Genet 2000; 25: 25 – 29.

15. Dupuy D, Bertin N, Hidalgo CA, Venkatesan K, Tu D, Lee D, et al. Genome-scale analysis of in vivo spatiotemporal promoter activity in Caenorhabditis elegans. Nat Biotechnol 2007; 25: 663 – 668.

16. Schlitt T, Palin K, Rung J, Dietmann S, Lappe M, Ukkonen E, et al. From gene networks to gene function. Genome Res 2003; 13: 2568 – 2576.

17. Bhattacharya R, Senbanerjee S, Lin Z, Mir S, Hamik A, Wang P, et al. Inhibition of vascular permeability factor/vascular endothelial growth factor-mediated angiogenesis by the Kruppel-like factor KLF2. J Biol Chem 2005; 280: 28848 – 28851.

18. Leung DW, Cachianes G, Kuang WJ, Goeddel DV, Ferrara N. Vas-cular endothelial growth factor is a secreted angiogenic mitogen. Science 1989; 246:1306 – 1309.

19. dela Paz NG, Walshe TE, Leach LL, Saint-Geniez M, D’Amore PA. Role of shear-stress-induced VEGF expression in endothelial cell survival. J Cell Sci 2012; 125: 831 – 843.

20. Wiik BP, Larstorp AC, Høieggen A, Kjeldsen SE, Olsen MH, Ibsen H, et al. Serum uric acid is associated with new-onset diabetes in hypertensive patients with left ventricular hypertrophy: The LIFE study. Am J Hypertens 2010, 23: 845 – 851.

21. Ficociello LH, Rosolowsky ET, Niewczas MA, Maselli NJ, Weinberg JM, Aschengrau A, et al. High-normal serum uric acid increases risk of early progressive renal function loss in type 1 dia-

difference, we speculated that miR-92a regulates angiogenesis via different pathways under different conditions. We deter-mined the expression level of integrin-α5 in HUA and control groups, and the results showed that there was no difference in expression level between the 2 groups. In the present study, miR-92a downregulation increased KLF2 levels, which inhib-ited angiogenesis under HUA. miR-92a overexpression inhib-ited KLF2 expression, thereby promoting angiogenesis. KLF2 is a transcription factor that regulates cell development and differentiation, and Fang and Davies have demonstrated that KLF2 is a target gene of miR-92a.27 We also confirmed that miR-92a could inhibit KLF expression via 3’UTR binding. The inhibition effect of KLF2 to its downstream gene, VEGFA, a strong angiogenesis-promoting factor, has been reported.17 However, whether KLF2 directly binds the promoter region of VEGFA to regulate angiogenesis inhibition is still unknown. Our EMSA results indicated that KLF2 binds directly to the promoter region of VEGFA. In addition, other studies have shown that KLF2 is a protective factor in endothelial cells, and its overexpression increases endothelial nitric oxide synthase (eNOS) expression, which functions to protect endothelial cells.28,29 However, we demonstrated that HUA upregulates KLF2 via the downregulation of miR-92a. In a previous study, we found that HUA result in endothelial cell dysfunction via miRNA-155-mediated inhibition of eNOS.30 We suggest that the regulation of eNOS by KLF2 depends on physiological conditions. Under HUA, miRNA-155 is also involved in the regulation of eNOS expression, and its negative effect might exceed the positive effect of KLF2. These different results indicate the complexity of gene regulation. Genes are regu-lated by complex networks, and different pathways affect the same gene under different conditions, leading to varying results. In addition to our in vitro experiment, we also deter-mined the serum VEGFA levels of hyperuricemia patients and healthy individuals. We found that the serum VEGFA levels of hyperuricemia patients were significantly lower than those of healthy individuals. Although it is difficult for us to inves-tigate the mechanism by which HUA regulate angiogenesis in vivo, the finding that the serum miR-92a and VEGFA levels of hyperuricemia patients are lower than those of healthy individuals partially validates our hypothesis.

Here, we report a possible mechanism of cardiovascular system injury caused by hyperuricemia in vitro. HUA inhibits angiogenesis via miR-92a regulation of KLF2-VEGFA signal-ing. There have been several studies that have shown that miRNAs play an important role in cardiovascular disease,31–33 and there have been many studies that have demonstrated that miRNA modulation is a potential means of treating cardiovas-cular disease.34 Perhaps this pathway could also represent a new target for the treatment of cardiovascular damage caused by hyperuricemia. Nevertheless, some issues remain unclear. We have demonstrated that miR-92a participates in regulating angiogenesis under HUA. Whether the other differentially expressed miRNAs, in addition to the let-7 family, are involved in angiogenesis remains unknown, and determining whether miR-92a plays the same role in angiogenesis in a hyperuricemia model and why the miR-92a levels in serum are negatively correlated with uric acid levels in patients will require further investigation. Additionally, HUVECs are dif-ferent from the endothelial cells of the vasculature of the heart, kidneys and other organs, though we suggest that the underly-ing mechanism of angiogenesis may be the same. Further validation is needed in the future.


2498 YU S et al.

30. Zhang X, Hong Q, Hou K, Wang Y, Wu D, Chen X. High concentra-tion uric acid regulates endothelial function via miR-155. Nan Fang Yi Ke Da Xue Xue Bao 2013; 33: 1141 – 1145 (in Chinese).

31. De Rosa S, Curcio A, Indolfi C. Emerging role of microRNAs in cardiovascular diseases. Circ J 2014; 78: 567 – 575.

32. Izawa H, Amano T. Plasma microRNA-100 as a biomarker of coro-nary plaque vulnerability: A new generation of biomarker for devel-oping acute coronary syndrome. Circ J 2015; 79: 303 – 304.

33. De Paoli F, Staels B, Chinetti-Gbaguidi G. Macrophage phenotypes and their modulation in atherosclerosis. Circ J 2014; 78: 1775 – 1781.

34. Caroli A, Cardillo MT, Galea R, Biasucci LM. Potential therapeutic role of microRNAs in ischemic heart disease. J Cardiol 2013; 61: 315 – 320.

Supplementary FilesSupplementary File 1

Table S1. Forward primers for miRNAs

Table S2. GO analysis of upregulated and downregulated miRNAs

Table S3. miRNA-GO network analysis

Figure S1. Serum let-7 family levels of healthy subjects and hyper-uricemia patients.

Figure S2. There were no differences in the expression level of angiopoietin-2 and integrin-α5 between the high concentrations of uric acid (HUA) group and the control group.

Please find supplementary file(s);http://dx.doi.org/10.1253/circj.CJ-15-0283

betes: Results of a 6-year follow-up. Diabetes Care 2010; 33: 1337 – 1343.

22. Causevic A, Semiz S, Macic Dzankovic A, Cico B, Dujic T, Malenica M, et al. Relevance of uric Acid in progression of type 2 diabetes mellitus. Bosn J Basic Med Sci 2010; 10: 54 – 59.

23. Puig JG, Torres RJ, Ruilope LM, Campo C, Grande C, Sancho T, et al. The pathophysiology of hyperuricemia in essential hypertension: A pilot study. Nucleosides Nucleotides Nucleic Acids 2004; 23: 1197 – 1199.

24. Quan H, Peng X, Liu S, Bo F, Yang L, Huang Z, et al. Differentially expressed protein profile of renal tubule cell stimulated by elevated uric acid using SILAC coupled to LC-MS. Cell Physiol Biochem 2011; 27: 91 – 98.

25. Hong Q, Qi K, Feng Z, Huang Z, Cui S, Wang L, et al. Hyperurice-mia induces endothelial dysfunction via mitochondrial Na+/Ca2+ exchanger-mediated mitochondrial calcium overload. Cell Calcium 2012; 51: 402 – 410.

26. Kang DH PS, Lee IK, Johnson RJ. Uric acid-induced C-reactive protein expression: Implication on cell proliferation and nitric oxide production of human vascular cells. J Am Soc Nephrol 2005; 16: 3553 – 3562.

27. Fang Y, Davies PF. Site-specific microRNA-92a regulation of Kruppel-like factors 4 and 2 in atherosusceptible endothelium. Arte-rioscler Thromb Vasc Biol 2012; 32: 979 – 987.

28. Parmar KM, Larman HB, Dai G, Zhang Y, Wang ET, Moorthy SN, et al. Integration of flow-dependent endothelial phenotypes by Kruppel-like factor 2. J Clin Invest 2006; 116: 49 – 58.

29. Dekker RJ, Boon RA, Rondaij MG, Kragt A, Volger OL, Elderkamp YW, et al. KLF2 provokes a gene expression pattern that establishes functional quiescent differentiation of the endothelium. Blood 2006; 107: 4354 – 4363.

Documents

High Concentrations of Uric Acid Inhibit Angiogenesis via