Archives of Biochemistry and Biophysics 557 (2014) 28–35

Contents lists available at ScienceDirect

Archives of Biochemistry and Biophysics

journal homepage: www.elsevier .com/ locate /yabbi

Identification of microRNAs involved in the modulationof pro-angiogenic factors in atherosclerosis by a polyphenol-richextract from propolis

http://dx.doi.org/10.1016/j.abb.2014.04.0090003-9861/� 2014 Elsevier Inc. All rights reserved.

⇑ Corresponding author. Address: Department of Clinical and ToxicologicalAnalysis, Faculty of Pharmaceutical Sciences, Universidade de São Paulo, Av.Professor Lineu Prestes 580, São Paulo, SP, Brazil. Fax: +55 1138132197.

E-mail address: dspa@usp.br (D.S.P. Abdalla).1 These authors contributed equally to this work.

Alejandro Cuevas a,1, Nicolás Saavedra a,1, Marcela F. Cavalcante b, Luis A. Salazar a, Dulcineia S.P. Abdalla b,⇑a Center of Molecular Biology and Pharmacogenetics, Scientific and Technological Bioresource Nucleus (BIOREN), Universidad de La Frontera, Av. Francisco Salazar, 01145Temuco, Chileb Department of Clinical and Toxicological Analysis, Faculty of Pharmaceutical Sciences, Universidade de São Paulo, Av. Professor Lineu Prestes 580, São Paulo, Brazil

a r t i c l e i n f o

Article history:Received 31 January 2014and in revised form 17 April 2014Available online 30 April 2014

Keywords:PolyphenolsPropolismicroRNAsAtherosclerosis

a b s t r a c t

New vessel formation plays a critical role in the progression and vulnerability of atherosclerotic lesions. Ithas been shown that polyphenols from propolis attenuate the progression of atherosclerosis and alsoexert inhibitory effects on angiogenic factors. However, the mechanisms underlying these effects arenot completely understood. Thus, this study aimed to identify microRNAs (miRNAs) involved in the mod-ulation of pro-angiogenic factors in the atherosclerotic plaques of LDL receptor gene knockout mice trea-ted with a polyphenol-rich extract of Chilean propolis. The progression of the atherosclerotic lesions wassignificantly attenuated in treated mice compared with control mice. Using microarray analysis and abioinformatic approach, we identified 29 differentially expressed miRNAs. Many of these miRNAs wereinvolved in biological processes associated with angiogenesis, such as the cell cycle, cell migration, cellgrowth and proliferation. Among them, three miRNAs (miR-181a, miR-106a and miR-20b) were over-expressed and inversely related to the expression of Vegfa (vascular endothelial growth factor A) andHif1a (hypoxia inducible factor 1 alpha). In addition, VEGF-A protein expression was attenuated in histo-logical sections obtained from the aortic sinuses of treated mice. VEGFA is a key pro-angiogenic factor inatherosclerotic plaques, and Hif1a, which is expressed in the necrotic nucleus of the atheroma, is its maininducer. We found a correlation between the over-expression of miR-181a, miR-106a and miR-20b andtheir target genes, Hif1a and Vegfa, which is consistent with attenuation of the atherosclerotic lesion. Inconclusion, our data analysis provides evidence that the anti-angiogenic effects of polyphenols from Chil-ean propolis can be modulated by miRNAs, in particular miR-181a, miR-106a and miR-20b.

� 2014 Elsevier Inc. All rights reserved.

Introduction

Atherosclerosis is a multifactorial chronic disorder character-ized by the progressive thickening and hardening of medium-sizedand large artery walls resulting from fat deposits on their inner lin-ing [1]. It has been observed that advanced atherosclerotic lesionsthat have undergone disruption demonstrate greater intimalneovascularization when compared with stable plaques [2,3].Moreover, it has been shown that increased expression of pro-angiogenic factors in atherosclerotic plaques is associated with

significantly thin fibrous layers [4]. Given this observation, it iswidely hypothesized that angiogenesis contributes to the progres-sion and vulnerability of atherosclerotic lesions with consequentrisks of rupture, thrombus formation and subsequent ischemic dis-ease [3,5].

Angiogenesis is the process of blood vessel formation from pre-existing vessels [6,7]. This process is an essential mechanism inadult life; however, when an imbalance occurs between activatingand inhibitory factors, it may contribute to worsening pathologicalcomplications such as tumor progression [8], Crohn’s disease [9],cartilage destruction in rheumatoid arthritis [10], blindness in dia-betes [11] and atherosclerosis, among others. In atherosclerosis,hypoxia caused by wall thickening and inflammation is an impor-tant factor that stimulates the expression of pro-angiogenic factorssuch as VEGFA, endothelin-1 and matrix metalloproteinase-2.Hypoxia also induces endothelial cell migration to the necroticnucleus [12–14].

A. Cuevas et al. / Archives of Biochemistry and Biophysics 557 (2014) 28–35 29

Current evidence suggests that a polyphenol-rich diet may pro-tect against cardiovascular disease [15–18]. Therefore, rich sourcesof polyphenols have been widely studied to understand their bene-ficial effects. It has been demonstrated that these compounds maymodulate key factors in the pathogenesis of atherosclerosis andangiogenesis; these factors include inflammation [19,20], oxidativestress [19,21] and endothelial dysfunction [22]. Propolis is a poly-phenol-rich resinous substance produced by honeybees (Apis melli-fera) from the exudates of trees and plants [23]. Propolis has beenshown to exhibit several biological activities, including antimicro-bial [24,25], anti-inflammatory [26], antioxidant [27], anticancer[28] and cardioprotective properties. Compelling evidence hasshown that the polyphenol-enriched fraction from propolis canmodulate angiogenesis in both in vitro and in vivo models [29–34].In addition, it is interesting to highlight that polyphenols attenuatethe progression of atherosclerotic lesions in mice [34,35]. Addition-ally, they modulate plasma lipid levels [34,36] and pro-angiogenicfactors [34], alleviating inflammation, improving �NO bioavailabilityand inducing the expression of heme oxygenase-1 [35].

The effects of polyphenols from propolis have been shown in sev-eral studies; however, the polyphenol mechanisms of action under-lying these effects are not fully understood. In this regard, someauthors suggest that the antioxidant activities of polyphenols areresponsible for various biological effects that have been previouslydescribed [37]. Polyphenols from diverse sources have been shownto regulate gene expression, including post-transcriptional regula-tion exerted through microRNAs (miRNAs).2 miRNAs are 20- to24-nucleotide RNAs that bind to the 30-UTR of an mRNA, thus prevent-ing its translation [38]. Polyphenols may modulate the expression ofmiRNAs [39,40]; however, no studies have shown that polyphenolspresent in propolis modulate miRNAs involved in angiogenesis.

Thus, the aim of this study was to assess whether polyphenolsmodulate the expression of pro-angiogenic factors and associatedmiRNAs in the atherosclerotic plaques of C57BL/6 LDLR knockoutmice.

Materials and methods

Preparation of a polyphenolic-rich extract from Chilean propolis

To prepare a polyphenolic-rich extract from Chilean propolis(PCP), a crude propolis sample was obtained from Cunco (La Ara-ucanía, Chile) in a mountainous area (latitude: �38� 580 40.4600,longitude: �72� 10 15.7300). Crude propolis was mixed with 80%ethanol (1:3 w/v proportion) in an amber-colored bottle and thenincubated for 30 min at 60 �C under constant shaking. Then, themixture was filtered using a Whatman No. 1 filter paper to sepa-rate the ethanolic extract from the solid propolis residue. Theextract was left at 4 �C overnight to promote the precipitation ofwaxes and other poorly soluble waste and then centrifuged toobtain a clear supernatant. Subsequently, the PCP was lyophilizedand reconstituted in a 2:1 w/v proportion with 80% ethanol.Finally, the polyphenol content of PCP was quantified by theFolin–Ciocalteu method as previously described [41].

Chemical characterization of the polyphenolic-rich extract fromChilean propolis

The polyphenolic compounds contained in the PCP wereseparated chromatographically on an HPLC system (Shimadzu

2 Abbreviations used: miRNAs, microRNAs; PCP, polyphenolic-rich extract fromChilean propolis; MS, mass spectrometer; Ldlr, low-density lipoprotein receptorHCD, hypercholesterolemic diet; LIMMA, linear modeling for microarray dataMSigDB, Molecular Signatures Databases; IPA, Ingenuity Pathway Analysis; CARM1coactivator-associated arginine methyltransferase 1; HREs, Hif-responsive elementsTSP-1, thrombospondin-1.

;;,;

Prominence, Kyoto, Japan) using a Luna C18(2) column(250 � 3 mm, 5 lm). The solvents were water (A) and methanol(B), both containing formic acid 0.1%. Was used a gradient, increas-ing B from 30–60% between 0 and 80 min, 60–90% (B) from 80 to85 min, maintaining this percentage until the 90th minute andthen returned to 30%. Later, compounds identified by mass spec-trum analysis using a mass spectrometer (MS) equipped with anelectrospray ionization source (Esquire HCT, Bruker Daltonics,Billerica, MA, USA). The electrospray source was operated in nega-tive mode at 3500 V. Helium was used as collision gas. Finally, theidentification of compounds was based on the analysis of UV spec-trum and MS/MS fragmentation pattern.

Animals and treatments

Fifty homozygous low-density lipoprotein receptor-deficientmale mice (Ldlr�/�) were used in the present study. The animalswere fed a 0.5% cholesterol-enriched hypercholesterolemic diet(HCD) for 12 weeks. Starting at week 13, they were divided into2 groups (n = 25) that received the HCD and a daily dose of eitherthe polyphenol-rich extract from Chilean propolis (250 mg/kg ofbody weight, PCP group) or the vehicle (ethanol 0.5 g/kg of bodyweight, control group) by gavage for 4 weeks. The animals receivedfood and water ad libitum and were maintained under a 12-h light/dark cycle and controlled humidity and temperature conditions.After the treatment period, blood samples and cardiac tissues werecollected to analyze the atherosclerotic lesions in the aortic sinus,including the region that contains the aortic valves. All interven-tions were approved by the Commission of Ethics in Animal Exper-imentation of the Faculty of Pharmaceutical Sciences, University ofSão Paulo (Sao Paulo, Brazil).

Serum lipid profile analysis

Serum was obtained from blood samples from each animal tomeasure total cholesterol, triglycerides, HDL cholesterol and LDLcholesterol using colorimetric assays (Labtest Diagnostica SA,Brazil).

Morphometric analysis of atherosclerotic plaques

Aortic sinus samples from cardiac tissues were cryostat-sec-tioned in 10-lm-thick sections and then stained using Oil Red Oto quantify the atherosclerotic lesion areas. Images from stainedslides were obtained using an optical microscope equipped witha digital camera (Nikon DXM 1200C) and corresponding software(Nikon ACT-1C). The quantification of atherosclerotic lesions wasperformed following guidelines previously used [42] to assessimages captured with the ImageJ 1.46r software (NationalInstitutes of Health, Bethesda, Maryland, USA).

Screening for differentially expressed miRNAs in atherosclerotic lesions

miRNA-enriched samples were obtained using the mirVana Iso-lation Kit (Ambion, USA) following the manufacturer’s instructions.The samples were Hy3/Hy5 labeled, hybridized on a miRCURY LNAArray v.16.0 (Exiqon, Denmark) and later scanned using the AgilentG2565BA Microarray Scanner System (Agilent Technologies, Inc.,USA). Fluorescence intensity was obtained from images usingImaGene� 9 Software (miRCURY LNA microRNA Array AnalysisSoftware, Exiqon, Denmark) and later analyzed using the statisticssoftware program R (R Foundation for Statistical Computing,Vienna, Austria) with the Bioconductor/LIMMA (linear modelingfor microarray data) package [43]. The algorithms Normexp [44]and Lowess were used for background subtraction and intra-slidenormalization, respectively.

30 A. Cuevas et al. / Archives of Biochemistry and Biophysics 557 (2014) 28–35

Identification of miRNA targets and signaling pathways potentiallymodulated by differentially expressed miRNAs

To select miRNAs that were potentially associated with the reg-ulation of angiogenesis-related genes, we used algorithms to pre-dict and discriminate as explained below. First, using themirWALK database, we identified the potential targets for all dif-ferentially expressed miRNAs [45]. Then, the gene list was filteredusing gene sets available from Molecular Signatures Databases(MSigDB) and published reports, leaving only those miRNAsrelated to the terms ‘‘Angiogenesis’’ and ‘‘Regulation of Angiogen-esis’’. Moreover, signaling pathways related to these miRNAs wereidentified using the miRSystem database. Additionally, a networkanalysis using the collected data was performed with IngenuityPathway Analysis (IPA, Ingenuity Systems).

Validation of differentially expressed miRNAs and mRNA expression ofpotential targets

Total RNA was isolated using TRI reagent (Ambion, USA.), andan miRNA-enriched fraction was obtained using the mirVana Isola-tion Kit (Ambion, USA). For mRNA quantification, 1 lg of total RNAwas reverse-transcribed using High Capacity RNA-to-cDNA MasterMix (Applied Biosystems Inc., USA), while the cDNA used for miR-NAs analysis was synthesized using miRCURY LNATM Universal RTmicroRNA PCR (Exiqon, Denmark). Both quantification assays wereperformed with real-time PCR in a 7300 Real-Time PCR System(Applied Biosystems Inc., USA) using a reaction containing 50 ngof reverse-transcribed RNA, 200 nM of each specific primer and10 lL of 2� Fast SYBR� Green Master Mix (Applied BiosystemsInc., USA) in a final volume of 20 lL under cycling conditions rec-ommended by the manufacturer. Relative expression was analyzedusing the qPCR database [46], from which the corresponding Cqwas obtained using the Miner algorithm [47]. Rpl13a and U6snRNA were used as references for mRNA and miRNA normaliza-tion, respectively.

VEGF protein expression in atherosclerotic lesions

The expression of VEGF was quantified in 5-lm-thick sectionsobtained from the aortic sinuses of control and treated mice. Thedetection was performed with an appropriate anti-VEGF primaryantibody (Abcam, USA) and visualized using a biotinylated second-ary antibody (Abcam, USA) followed by the addition of streptavi-din-peroxidase. Images were obtained using an opticalmicroscope equipped with a digital camera. Quantitative analysiswas performed using the ImageJ 1.46r software (National Institutesof Health, Bethesda, Maryland, USA).

Statistical analysis

The obtained data were analyzed using GraphPad Prism version5.0a (GraphPad Software, San Diego CA, USA). Data are presentedas the mean ± SD. Differences between groups involving continu-ous variables were evaluated with Student’s t-test. Statistical sig-nificance was set at p < 0.05.

Results

Polyphenolic compounds identified in a polyphenol-rich extract fromChilean propolis

Polyphenols contained in the PCP were identified by LC–MS.This analysis revealed the presence of 35 compounds in the PCPsamples as shown in Fig. 1.

Effect of polyphenols on atherosclerosis progression

The formation of atherosclerotic plaques was visualized usingOil Red O staining of sections of aortic sinus obtained from Ldlrknockout mice, demonstrating the validity of the model. Quantita-tive analysis of atherosclerotic lesion size (Fig. 2) indicated that theprogression of aortic sinus injury was significantly attenuated intreated mice (�30%, p < 0.05) compared with control mice.

Effects of polyphenols from propolis on the lipid profile

There were no significant differences in plasma triglyceride,total cholesterol, HDL cholesterol or LDL cholesterol levels (Table 1)between the PCP-treated and control groups.

Identification of miRNAs differentially expressed between treated andcontrol mice

Normalized values for each miRNA cluster were used for statis-tical analysis. A total of 29 miRNAs were significantly up-regulatedin aortic sinus samples from mice treated with PCP, and a heat mapwas generated to visualize distinguishable miRNA expression pro-files between samples from control and PCP-treated mice (Fig. 3A).

The bioinformatic analysis showed that the differentiallyexpressed miRNAs were associated with multiple biological func-tions including angiogenesis, the cell cycle, cell migration, cellgrowth and cell proliferation, among others (Fig. 3). According tothis analysis, 8 miRNAs (miR-181a, miR-106a, miR-19b, miR-15a,miR-16, miR-17, miR-20a and miR-20b) were related to pro-angiogenic factors. In addition, these miRNAs were closely associ-ated with the VEGF signaling pathway (miRSYSTEM score: 2.090)and the MAPK signaling pathway (miRSYSTEM score: 3.963). ThesemiRNAs were selected for further analysis (Fig. 3C).

Finally, network analyses were generated through the use of IPA(Ingenuity� Systems, www.ingenuity.com). This analysis sug-gested that among the miRNAs selected in the previous analysis,microRNAs with the ‘‘seed’’ sequence AAAGUGC (miR-20b andmiR-106a) and microRNAs with the ‘‘seed’’ sequence AGCAGCAcould be related to the regulation of Vegfa and Hif1a, two importantfactors involved in the initial induction of angiogenesis in athero-sclerotic lesions (Fig. 4).

Validation of differentially expressed miRNAs

The expression of the 8 selected miRNAs was validated by qRT-PCR. Among them, only miR-20b (8.4-fold, p < 0.01), miR-106a(2.4-fold, p < 0.01) and miR-181a (2.7-fold, p < 0.05) were indeedup-regulated (Fig. 5B).

Expression of target genes for differentially expressed miRNAs

Fig. 5C shows the relative expression of the mRNA targets ofvalidated miRNAs. These genes were selected according to theirprobability of being modulated by these miRNAs based on networkanalyses, target prediction analyses and literature reviews. Vegfa,Fgf2 and Hif1a were significantly repressed by 0.52-, 0.64- and0.47-fold, respectively.

VEGFA expression in atherosclerotic lesions

The atherosclerotic lesions of the aortic sinuses from Ldlr�/�

mice treated with PCP were used to evaluate the expression ofVEGF-A by immunohistochemistry (Fig. 6A and B). The results indi-cate decreased expression of the VEGF-A protein (�40%, p < 0.05)in the PCP-treated group compared with the control group(Fig. 6C).

Fig. 1. Polyphenolic compounds found in a polyphenol-rich extract from Chilean propolis. 1: caffeic acid; 2: caffeic acid; 3: ferulic/isoferulic acid; 4: 3,4-dimethyl caffeic acid;5: pinobanksin-5-methyl-ether; 6: quercetin; 7: pinobanksin; 8: quercetin-3-methyl ether; 9: pinocembrin-5-methyl ether; 10: kaempferol; 11: apigenin; 12: quercetin-7-methyl ether; 13: luteolin-5-methyl ether; 14: cinnamilidenacetic acid; 15: pinobanksin-3-O-acetate; 16: isorhamnetin; 17: pinocembrin; 18: caffeic acid benzyl ester; 19:caffeic acid isoprenyl ester; 20: pinobanksin-3-O-acetate; 21: caffeic acid isoprenyl ester; 22: caffeic acid isoprenyl ester; 23: chrysin; 24: caffeic acid phenethyl ester; 25:galangin; 26: methoxy-chrysin; 27: p-coumaric benzyl ester; 28: p-coumaric methyl butenyl ester; 29: pinobanksin-3-O-propionate; 30: caffeic acid cinnamyl ester; 31:pinobanksin-3-O-pentenoate; 32: p-coumaric cinnamyl ester; 33: pinobanksin-3-O-butyrate; 34: pinobanksin-3-O-pentanoate; 35: pinobanksin derivative.

Fig. 2. Atherosclerotic lesion sizes of LDLr�/� mice treated with polyphenols from Chilean propolis and controls. (A) Representative Oil Red O-stained atherosclerotic lesionsfor the control and PCP groups. (B) Lesion areas of atherosclerotic lesions for controls and treated animals presented as the mean ± standard deviation. #Student’s t test,p < 0.05.

Table 1Lipid profile of LDLr knockout atherosclerotic mice.

Control PCP

Total cholesterol (mg/dL) 1858 ± 484 2312 ± 331Triglycerides (mg/dL) 182 ± 46.3 322 ± 160.2LDL cholesterol (mg/mL) 1794 ± 489 2213 ± 297HDL cholesterol (mg/mL) 28.0 ± 13.1 34.6 ± 11.3

Values are shown as the mean ± standard deviation.

A. Cuevas et al. / Archives of Biochemistry and Biophysics 557 (2014) 28–35 31

Discussion

In the last decade, knowledge of microRNA transcriptional con-trol and process modulation has increased considerably. This bodyof knowledge includes studies that suggest a role for these non-coding small RNAs in hypoxia [48] and angiogenesis [49], which

are both important processes associated with atherosclerosis pro-gression. In atherosclerotic ApoE�/� mice, miRNA screeningshowed the over-expression of 75 miRNAs in atherosclerotic aortasin comparison with unaffected controls [50]. Similarly, in humanaortic tissue, a group of miRNAs was differentially expressed inatherosclerotic plaques [51]. Moreover, recent studies suggest thatmiRNAs could be involved in the biological effects exerted by poly-phenols [52].

In the present study, the miRNA profile of atherosclerotic lesionswas analyzed after treating mice with a polyphenol extract fromChilean propolis. In the studied tissues, we observed the expressionof an average of 430 miRNAs registered in miRBase v. 18, whereas27 miRNAs were over-expressed in the aortic sinuses of treatedmice. Then, to guide our analysis and identify those microRNAsrelated to pro-angiogenic factors, we used a bioinformaticsapproach. A selected group of miRNAs was validated by qPCR. That

Fig. 3. Screening for differentially expressed miRNAs. (A) Heat map and hierarchical clustering of significantly over-expressed miRNAs. (B) Heat map of relevant biologicalfunctions obtained by IPA analysis. (C) Selected angiogenesis-related miRNAs for later qPCR validation.

Fig. 4. Relative quantification of miRNAs and their target genes. (A) Angiogenesis-related target genes for selected miRNAs. (B) Differentially expressed miRNAs after qPCRvalidation. (C) Relative expression of predicted target genes. Student’s t-test, #p < 0.05; ##p < 0.01.

32 A. Cuevas et al. / Archives of Biochemistry and Biophysics 557 (2014) 28–35

group included miR-106a, miR-20a, miR-20b, miR-181d, miR-15a,miR-16, miR-19b and miR-17 among the over-expressed miRNAs.Other studies have used a similar approach to evaluate the effectsof polyphenols on the expression of microRNAs [53]; however, tothe best of our knowledge, no other study has replicated our model.In a similar study performed in human aortas obtained by endarter-ectomy, 255 microRNAs were found to be expressed, 23 of whichwere over-expressed in atherosclerotic plaques [51].

The selected miRNAs have been associated with multiple otherfunctions. For example, miR-106a is associated with both the pro-gression of cancer [54,55] and the inhibition of tumor propertiessuch as proliferation and migration [56]. Moreover, some of thesemicroRNAs have a role in the regulation of inflammatory processes.For example, miR-15a was implicated in inflammatory regulationin acute coronary syndrome. miR-15a was shown to target coacti-vator-associated arginine methyltransferase 1 (CARM1) [57] or

Fig. 5. Network analysis showing the relationship between angiogenesis-related genes and miRNAs with AAAGUGC and AGCAGCA seed regions.

Fig. 6. VEGF protein expression in the aortic sinuses of atherosclerotic mice treated with polyphenols from Chilean propolis. (A and B) Expression of VEGF protein in the aortictissues of control and treated atherosclerotic mice, respectively. (C) Relative quantification of VEGF in atherosclerotic lesions of control and treated mice. #Student’s t-test,p < 0.05.

A. Cuevas et al. / Archives of Biochemistry and Biophysics 557 (2014) 28–35 33

cluster microRNA-17/20a/106a, which is implicated in modulatinginflammatory responses in macrophages by targeting signal-regulatory protein a, an essential signaling molecule that modu-lates inflammatory responses in leukocytes [58].

There are few precedents concerning the modulation of thisgroup of microRNAs by polyphenols in similar models. We foundsome commonalities with a study conducted by Milenkovic et al.[59], which evaluated the effects of different polyphenols onmiRNA expression in the livers of ApoE�/� mice. In that work,miR-106a was modulated by hesperidin (30 mg/day) and narangin(30 mg/day), and miR-19a was modulated by curcumin (30 mg/day) [59]. Caffeic acid and quercetin, which are present in our poly-phenolic extract, were also tested, but they failed to modulate anyof the microRNAs over-expressed in our study [59]. The types oftissue analyzed, the treatment doses and the complexity of thetreatment composition might explain these differences. The treat-ment used in the present study consisted of a mixture of polyphe-nolic compounds and not individual polyphenols. In this sense,there may be synergistic effects between the polyphenols present

in the extract [60], thus making it difficult to determine the effectsof an individual compound.

Notably, miR-106a, miR-20a, miR-20b and miR-19b have thesame seed region. Most of the differentially expressed miRNAsdetected by microarray were unrelated to angiogenesis, possiblybecause polyphenols act via several mechanisms of action thataffect multiple targets.

In the pathways analysis performed as previously described[61], we used the miRNAs selected by miRNA-angiogenesis relatedtarget interactions and observed a high association score for theMAPK and VEGF signaling pathways. Finally, IPA was used toestablish a relationship between possible target genes and previ-ously selected miRNAs, revealing the cardiovascular system andangiogenesis as primary functions that were potentially modu-lated. Additionally, this analysis revealed that Vegfa, Runx1 andHif1a would be interesting targets to evaluate. These genes wereassociated with the AAAGUGC (miR-17, miR-20b and miR-106a)and AGCAGCA (miR-15a and miR-16) seed regions. The Fgf2 genewas independently associated with miR-16 and other miRNAs

34 A. Cuevas et al. / Archives of Biochemistry and Biophysics 557 (2014) 28–35

containing this same seed region.qPCR analysis showed differentialexpression for miR-106a, miR-20b and miR-181d, resulting in afalse positive rate of 57% for the total number of miRNAs detectedby the microarray assay, which can be explained by the low num-ber of slides analyzed for each group [62]. The expression levels ofCalca, Hif1a, Cd40, Runx1, Vegfa and Fgf2 were quantified, revealingdecreases in Vegfa, Fgf2 and Hif1a mRNA that correlated with miR-NA over-expression.

Considering the importance of VEGFA in angiogenesis and ath-erosclerotic plaques, we continued to analyze the possible role ofthis gene in the context of the modulatory effects exerted by thepolyphenol extract. In advanced atherosclerotic lesions, the forma-tion of the necrotic core and the subsequent thickening of the arte-rial wall prevents adequate oxygen diffusion from the lumen, whichcauses hypoxia [63,64]. Hypoxia, in turn, promotes the stabilizationof Hif1a–Hif2a, its translocation to the nucleus and the expressionof genes containing Hif-responsive elements (HREs), whichincludes VEGFA [65]. Our data show a decrease in both Hif1a andVegfa gene expression levels in the animals treated with PCP. Theexpression of Vegfa mRNA correlated with the detection of VEGF-A protein in atherosclerotic plaques by immunohistochemistry.The expression of Hif1a and Vegfa might be associated with miR-20b and miR-106a and belong to the same cluster (17–92), giventhat they have the same seed region. Therefore, miR-20b andmiR-106a may be able to recognize the same sequence in the30UTRs of their target genes (AAAGUGC). This cluster has previouslybeen associated with angiogenesis by modulating thrombospon-din-1 (TSP-1), an endogenous inhibitor of that process [66], ordirectly and indirectly regulating the expression of VEGFA [67,68].Moreover, miR-20b can suppress Hif1a mRNA during hypoxia[48]. The mechanism by which polyphenols modulate the expres-sion of these miRNAs remains unclear. However, it might involvea direct effect on the transcriptional activity of cluster 17–92, inhib-iting Hif1a and VEGFA independent of activator factors. In fact, asimilar mechanism has been previously described in the contextof decreased hepatic fat accumulation, which was correlated withthe expression of miR-378 and its host gene, PGC-1b [69].

Independent of oxygen availability, the stabilization of Hif1aand the transcriptional activation of VEGFA can be induced byinflammatory cytokines in an NF-jB-dependent mechanism [70].In prostate cancer cells, stabilization of Hif1a in response to reac-tive oxygen species was observed [71,72], suggesting that potentialantioxidant or anti-inflammatory effects exerted on the atheroscle-rotic plaque may modulate miR-20b or miR-106a and conse-quently the expression of Hif1a and Vegfa. These potentialanti-inflammatory and antioxidant activities have been previouslydescribed for polyphenols [73,74]. Moreover, reactive oxygen spe-cies have demonstrated the ability to modulate the expression ofsome miRNAs; this process has been attenuated with polyphenoltreatment [75]. In fact, there is evidence for the polyphenol-induced over-expression of miR-20b by resveratrol, which modu-lates the expression of Hif1a and Vegfa [76].

In conclusion, our data provide evidence indicating that theover-expression of miR-20b or miR-106a could be a potentialmechanism by which polyphenols from propolis exert modulatoryeffects on Hif1a and Vegfa, two key factors in neovascularization,affecting advanced atherosclerotic plaques.

Acknowledgments

A.C. & N.S. were the recipients of fellowships from CONICYT,Chile. This study was supported by grants from CONICYT (Chile),FAPESP (São Paulo, Brazil) and CAPES (Brazil). N.S. is the recipientof a postdoctoral fellowship from Convenio de Desempeño –Universidad de La Frontera (Chile).

References

[1] P. Libby, Am. J. Cardiol. 98 (2006) 3Q–9Q.[2] J.O. Juan-Babot, J. Martinez-Gonzalez, M. Berrozpe, L. Badimon, Rev. Esp.

Cardiol. 56 (2003) 978–986.[3] R. Virmani, F.D. Kolodgie, A.P. Burke, A.V. Finn, H.K. Gold, T.N. Tulenko, S.P.

Wrenn, J. Narula, Arterioscler. Thromb. Vasc. Biol. 25 (2005) 2054–2061.[4] M.J. McCarthy, I.M. Loftus, M.M. Thompson, L. Jones, N.J. London, P.R. Bell, A.R.

Naylor, N.P. Brindle, Br. J. Surg. 86 (1999) 707–708.[5] P.R. Moreno, K.R. Purushothaman, E. Zias, J. Sanz, V. Fuster, Curr. Mol. Med. 6

(2006) 457–477.[6] G.L. Semenza, J. Cell. Biochem. 102 (2007) 840–847.[7] R. Demir, A. Yaba, B. Huppertz, Acta Histochem. 112 (2010) 203–214.[8] V. Baeriswyl, G. Christofori, Semin. Cancer Biol. 19 (2009) 329–337.[9] I.D. Pousa, J. Mate, J.P. Gisbert, Eur. J. Clin. Invest. 38 (2008) 73–81.

[10] N. Maruotti, F.P. Cantatore, E. Crivellato, A. Vacca, D. Ribatti, Histol.Histopathol. 21 (2006) 557–566.

[11] T.N. Crawford, D.V. Alfaro 3rd, J.B. Kerrison, E.P. Jablon, Curr. Diab. Rev. 5(2009) 8–13.

[12] J. Krupinski, A. Font, A. Luque, M. Turu, M. Slevin, Front. Biosci. 13 (2008)6472–6482.

[13] C.S. Lim, S. Kiriakidis, A. Sandison, E.M. Paleolog, A.H. Davies, J. Vasc. Surg. 58(2013) 219–230.

[14] G.L. Semenza, Nat. Rev. Cancer 3 (2003) 721–732.[15] V. Habauzit, C. Morand, Ther. Adv. Chronic Dis. 3 (2012) 87–106.[16] S.H. Li, H.B. Tian, H.J. Zhao, L.H. Chen, L.Q. Cui, PLoS One 8 (2013) e69818.[17] M. Morrison, R. van der Heijden, P. Heeringa, E. Kaijzel, L. Verschuren, R.

Blomhoff, T. Kooistra, R. Kleemann, Atherosclerosis 233 (2014) 149–156.[18] A. Tresserra-Rimbau, E.B. Rimm, A. Medina-Remon, M.A. Martinez-Gonzalez, R.

de la Torre, D. Corella, J. Salas-Salvado, E. Gomez-Gracia, J. Lapetra, F. Aros, M.Fiol, E. Ros, L. Serra-Majem, X. Pinto, G.T. Saez, J. Basora, J.V. Sorli, J.A. Martinez,E. Vinyoles, V. Ruiz-Gutierrez, R. Estruch, R.M. Lamuela-Raventos, Nutr. Metab.Cardiovasc. Dis. (2014).

[19] Y.W. Kim, X.Z. West, T.V. Byzova, J. Mol. Med. (Berl.) 91 (2013) 323–328.[20] P. Libby, Nature 420 (2002) 868–874.[21] R. Stocker, J.F. Keaney Jr., Physiol. Rev. 84 (2004) 1381–1478.[22] T. Mano, T. Masuyama, K. Yamamoto, J. Naito, H. Kondo, R. Nagano, J. Tanouchi,

M. Hori, M. Inoue, T. Kamada, Am. Heart J. 131 (1996) 231–238.[23] S. Castaldo, F. Capasso, Fitoterapia 73 (Suppl. 1) (2002) S1–6.[24] M. Curifuta, J. Vidal, J. Sánchez-Venegas, A. Contreras, L.A. Salazar, M. Alvear,

Cien. Invest. Agraria 39 (2012) 347–359.[25] N. Saavedra, L. Barrientos, C.L. Herrera, M. Alvear, G. Montenegro, L.A. Salazar,

Cien. Invest. Agraria 38 (2011) 117–125.[26] J.L. Machado, A.K. Assuncao, M.C. da Silva, A.S. Dos Reis, G.C. Costa, S. Arruda, S.

Arruda Dde, B.A. Rocha, M.M. Vaz, A.M. Paes, R.N. Guerra, A.A. Berretta, F.R. doNascimento, Evid. Based Complement. Altern. Med. (2012) 157652.

[27] A.M. Kurek-Gorecka, A. Sobczakb, A. Rzepecka-Stojkod, M.T. Gorecki, M.Wardas, K. Pawlowska-Goral, J. Biosci. 67 (2012) 545–550.

[28] D. Sawicka, H. Car, M.H. Borawska, J. Niklinski, Folia Histochem. Cytobiol. 50(2012) 25–37.

[29] Y. Chikaraishi, H. Izuta, M. Shimazawa, S. Mishima, H. Hara, Mol. Nutr. FoodRes. 54 (2010) 566–575.

[30] S.A. de Moura, M.A. Ferreira, S.P. Andrade, M.L. Reis, L. Noviello Mde, D.C. Cara,Evid. Based Complement. Altern. Med. 54 (2011) 182703.

[31] H. Izuta, M. Shimazawa, K. Tsuruma, Y. Araki, S. Mishima, H. Hara, BMCComplement. Altern. Med. 9 (2009) 45.

[32] M. Keshavarz, A. Mostafaie, K. Mansouri, Y. Shakiba, H.R. Motlagh, Arch. Med.Res. 40 (2009) 59–61.

[33] C. Meneghelli, L.S. Joaquim, G.L. Felix, A. Somensi, M. Tomazzoli, D.A. da Silva,F.V. Berti, M.B. Veleirinho, O. Recouvreux Dde, A.C. de Mattos Zeri, P.F. Dias, M.Maraschin, Microvasc. Res. 88 (2013) 1–11.

[34] J.B. Daleprane, S. Freitas Vda, A. Pacheco, M. Rudnicki, L.A. Faine, F.A. Dorr, M.Ikegaki, L.A. Salazar, T.P. Ong, D.S. Abdalla, J. Nutr. Biochem. 23 (2012) 557–566.

[35] W.M. Loke, J.M. Proudfoot, J.M. Hodgson, A.J. McKinley, N. Hime, M. Magat, R.Stocker, K.D. Croft, Arterioscler. Thromb. Vasc. Biol. 30 (2010) 749–757.

[36] A.R. Yeo, J. Lee, I.H. Tae, S.R. Park, Y.H. Cho, B.H. Lee, H.C. Shin, S.H. Kim, Y.C.Yoo, Prev. Nutr. Food Sci. 17 (2012) 1–7.

[37] J. Coban, B. Evran, F. Ozkan, A. Cevik, S. Dogru-Abbasoglu, M. Uysal, Biosci.Biotechnol. Biochem. 77 (2013) 389–391.

[38] V.N. Kim, Nat. Rev. Mol. Cell Biol. 6 (2005) 376–385.[39] D. Milenkovic, B. Jude, C. Morand, Free Radical Biol. Med. 64 (2013) 40–51.[40] A. Cuevas, N. Saavedra, L.A. Salazar, D.S. Abdalla, Nutrients 5 (2013) 2314–

2332.[41] A.L. Waterhouse, Current Protocols in Food Analytical Chemistry, John Wiley &

Sons Inc., 2001.[42] B. Paigen, A. Morrow, P.A. Holmes, D. Mitchell, R.A. Williams, Atherosclerosis

68 (1987) 231–240.[43] J.M. Wettenhall, G.K. Smyth, Bioinformatics 20 (2004) 3705–3706.[44] J.D. Silver, M.E. Ritchie, G.K. Smyth, Biostatistics 10 (2009) 352–363.[45] H. Dweep, C. Sticht, P. Pandey, N. Gretz, J. Biomed. Inform. 44 (2011) 839–847.[46] S. Pabinger, G.G. Thallinger, R. Snajder, H. Eichhorn, R. Rader, Z. Trajanoski,

BMC Bioinformatics 10 (2009) 268.[47] S. Zhao, R.D. Fernald, J. Comput. Biol. 12 (2005) 1047–1064.[48] P. Madanecki, N. Kapoor, Z. Bebok, R. Ochocka, J.F. Collawn, R. Bartoszewski,

Cell. Mol. Biol. Lett. 18 (2013) 47–57.

A. Cuevas et al. / Archives of Biochemistry and Biophysics 557 (2014) 28–35 35

[49] S. Wang, E.N. Olson, Curr. Opin. Genet. Dev. 19 (2009) 205–211.[50] H. Han, Y.H. Wang, G.J. Qu, T.T. Sun, F.Q. Li, W. Jiang, S.S. Luo, Exp. Mol. Med. 45

(2013) e13.[51] K. Bidzhekov, L. Gan, B. Denecke, A. Rostalsky, M. Hristov, T.A. Koeppel, A.

Zernecke, C. Weber, Thromb. Haemost. 107 (2012) 619–625.[52] C. Boesch-Saadatmandi, A. Loboda, A.E. Wagner, A. Stachurska, A. Jozkowicz, J.

Dulak, F. Doring, S. Wolffram, G. Rimbach, J. Nutr. Biochem. 22 (2011) 293–299.

[53] W.P. Tsang, T.T. Kwok, J. Nutr. Biochem. 21 (2010) 140–146.[54] P. Li, Q. Xu, D. Zhang, X. Li, L. Han, J. Lei, W. Duan, Q. Ma, Z. Wu, Z. Wang, FEBS

Lett. 588 (2014) 705–712.[55] M. Zhu, N. Zhang, S. He, Y. Lui, G. Lu, L. Zhao, FEBS Lett. 588 (2014) 600–607.[56] F. Zhi, G. Zhou, N. Shao, X. Xia, Y. Shi, Q. Wang, Y. Zhang, R. Wang, L. Xue, S.

Wang, S. Wu, Y. Peng, Y. Yang, PLoS One 8 (2013) e72390.[57] X. Liu, L. Wang, H. Li, X. Lu, Y. Hu, X. Yang, C. Huang, D. Gu, Atherosclerosis 233

(2014) 349–356.[58] D. Zhu, C. Pan, L. Li, Z. Bian, Z. Lv, L. Shi, J. Zhang, D. Li, H. Gu, C.Y. Zhang, Y. Liu,

K. Zen, J. Allergy Clin. Immunol. 132 (426–436) (2013) e428.[59] D. Milenkovic, C. Deval, E. Gouranton, J.F. Landrier, A. Scalbert, C. Morand, A.

Mazur, PLoS One 7 (2012) e29837.[60] J.M. Sforcin, R.O. Orsi, V. Bankova, J. Ethnopharmacol. 98 (2005) 301–305.[61] T.P. Lu, C.Y. Lee, M.H. Tsai, Y.C. Chiu, C.K. Hsiao, L.C. Lai, E.Y. Chuang, PLoS One 7

(2012) e42390.[62] W.J. Lin, H.M. Hsueh, J.J. Chen, BMC Bioinformatics 11 (2010) 48.[63] T. Bjornheden, M. Levin, M. Evaldsson, O. Wiklund, Arterioscler. Thromb. Vasc.

Biol. 19 (1999) 870–876.

[64] J.C. Sluimer, J.M. Gasc, J.L. van Wanroij, N. Kisters, M. Groeneweg, M.D.Sollewijn Gelpke, J.P. Cleutjens, L.H. van den Akker, P. Corvol, B.G. Wouters,M.J. Daemen, A.P. Bijnens, J. Am. Coll. Cardiol. 51 (2008) 1258–1265.

[65] S.A. Paul, J.W. Simons, N.J. Mabjeesh, J. Cell. Physiol. 200 (2004) 20–30.[66] M. Dews, A. Homayouni, D. Yu, D. Murphy, C. Sevignani, E. Wentzel, E.E. Furth,

W.M. Lee, G.H. Enders, J.T. Mendell, A. Thomas-Tikhonenko, Nat. Genet. 38(2006) 1060–1065.

[67] S. Cascio, A. D’Andrea, R. Ferla, E. Surmacz, E. Gulotta, V. Amodeo, V. Bazan, N.Gebbia, A. Russo, J. Cell. Physiol. 224 (2010) 242–249.

[68] Z. Hua, Q. Lv, W. Ye, C.K. Wong, G. Cai, D. Gu, Y. Ji, C. Zhao, J. Wang, B.B. Yang, Y.Zhang, PLoS One 1 (2006) e116.

[69] T.I. Jeon, J.W. Park, J. Ahn, C.H. Jung, T.Y. Ha, Mol. Nutr. Food Res. 57 (2013)1931–1937.

[70] H.Z. Imtiyaz, M.C. Simon, Curr. Top. Microbiol. Immunol. 345 (2010) 105–120.[71] N. Gao, M. Ding, J.Z. Zheng, Z. Zhang, S.S. Leonard, K.J. Liu, X. Shi, B.H. Jiang, J.

Biol. Chem. 277 (2002) (1971) 31963–31971.[72] Q. Zhou, L.Z. Liu, B. Fu, X. Hu, X. Shi, J. Fang, B.H. Jiang, Carcinogenesis 28 (2007)

28–37.[73] M.L. Khalil, Asian Pacific J. Cancer Prev. 7 (2006) 22–31.[74] M. Alyane, L.B. Kebsa, H. Boussenane, H. Rouibah, M. Lahouel, Pak. J. Pharm. Sci.

21 (2008) 201–209.[75] J.C. Howell, E. Chun, A.N. Farrell, E.Y. Hur, C.M. Caroti, P.M. Iuvone, R. Haque,

Mol. Vis. 19 (2013) 544–560.[76] P. Mukhopadhyay, S. Das, M.K. Ahsan, H. Otani, D.K. Das, J. Cell Mol. Med. 16

(2012) 2504–2517.