micrornas 29b, 133b, and 211 regulate vascular smooth

BASIC RESEARCH www.jasn.org

MicroRNAs 29b, 133b, and 211 Regulate VascularSmooth Muscle Calcification Mediated by HighPhosphorus

Sara Panizo,* Manuel Naves-Díaz,* Natalia Carrillo-López,* Laura Martínez-Arias,*José Luis Fernández-Martín,* María Piedad Ruiz-Torres,† Jorge B. Cannata-Andía,*‡ andIsabel Rodríguez*

*Bone and Mineral Research Unit, Reina Sofia Institute of Renal Research (IRSIN), Renal Research Network (REDinREN) fromCarlos III Health Institute (ISCIII), Hospital Universitario Central de Asturias, Oviedo, Asturias, Spain; †Department of SystemsBiology, Renal Research Network (REDinREN) from Carlos III Health Institute (ISCIII), Faculty of Medicine, University ofAlcalá, Alcalá de Henares, Madrid, Spain; and ‡Department of Medicine, University of Oviedo, Oviedo, Asturias, Spain

ABSTRACTVascular calcification is a frequent cause of morbidity and mortality in patients with CKD and the generalpopulation. The common association between vascular calcification and osteoporosis suggests a link betweenbone and vascular disorders. Because microRNAs (miRs) are involved in the transdifferentiation of vascularsmooth muscle cells into osteoblast-like cells, we investigated whether miRs implicated in osteoblastdifferentiation and bone formation are involved in vascular calcification. Different levels of uremia, hyper-phosphatemia, and aortic calcification were induced by feeding nephrectomized rats a normal or high-phosphorus diet for 12 or 20 weeks, at which times the levels of eight miRs (miR-29b, miR-125, miR-133b,miR-135, miR-141, miR-200a, miR-204, and miR-211) in the aorta were analyzed. Compared with controls anduremic rats fed a normal diet, uremic rats fed a high-phosphorous diet had lower levels ofmiR-133b andmiR-211andhigher levels ofmiR-29b that correlated respectivelywith greater expression of osteogenic RUNX2 andwithlower expression of several inhibitors of osteoblastic differentiation. Uremia per se mildly reduced miR-133blevels only. Similar results were obtained in two in vitromodels of vascular calcification (uremic serum and high–calciumand–phosphorusmedium), andexperimentsusingantagomirs andmimics tomodifymiR-29b,miR-133b,andmiR-211 expression levels in thesemodels confirmed that thesemiRs regulate the calcification process.Weconclude that miR-29b, miR-133b, and miR-211 have direct roles in the vascular smooth muscle calcificationinduced by high phosphorus and may be new therapeutic targets in the management of vascular calcification.

J Am Soc Nephrol 27: ccc–ccc, 2015. doi: 10.1681/ASN.2014050520

A better knowledge of the pathogenesis of vascularcalcification is essential, because it is a frequentcause of morbidity and mortality in the generalpopulation but especially, patients with CKD.Moreover, vascular calcification is frequently asso-ciated to bone mass disorders, which suggests thatfactors involved in the regulation of both entitiescould be linked.1,2

Vascular smooth muscle cells (VSMCs) have thesame mesenchymal origin as osteoblasts, and theyare able, in the presence of several stimuli, includinguremia, to transdifferentiate into osteoblastic cellsin response to signals coming from the internal

milieu, extracellular matrix components, and bio-chemical and physical factors.3,4 This phenotypicmodulation is tightly regulated and involves the

Received May 28, 2014. Accepted June 11, 2015.

Published online ahead of print. Publication date available atwww.jasn.org.

Correspondence: Prof. Jorge B. Cannata-Andía or Dr. IsabelRodríguez, Bone and Mineral Research Unit, Instituto Reina Sofíade Investigación Nefrológica, Hospital Universitario Central deAsturias, C/Julián Clavería s/n, 33006 Oviedo, Asturias, Spain.Email: [email protected] or [email protected]

Copyright © 2015 by the American Society of Nephrology

J Am Soc Nephrol 27: ccc–ccc, 2015 ISSN : 1046-6673/2703-ccc 1

http://www.jasn.org

mailto:[email protected]

mailto:[email protected]

loss of expression of several muscle–related proteins of theVSMCs and the gain of expression of bone-related proteinsaccompanied by macroscopic structural changes, such ascalcifications.5–7

A few recent studies have shown that some microRNAs(miRs) are new potential players in the differentiation of VSMCsinto osteoblast-like cells,8–10 although their role in the develop-ment of calcification is not completely clear. miRs are small(approximately 22 nucleotides), noncoding, single-strandedRNAs that mediate post–transcriptional gene silencing by bind-ing to sites of antisense complementarity in 39 untranslatedregions of target mRNAs. miRs are involved in crucial biologicprocesses, including cell proliferation and differentiation as wellas tissue development.11,12 Recent studies have discovered thatmultiple miRs are important regulators of bone-related genes,including essential transcription factors such as Runt–relatedtranscription factor 2 (RUNX2), the main transcription factorrequired for osteoblast differentiation.13 Several miRs act as neg-ative regulators of osteoblast differentiation and subsequentmineralization through their capacity to silence key transcrip-tion factors of the osteogenesis, such as RUNX2,14,15 Osterix,16

or distal–less homeobox 5.17 In contrast, some miRs promoteosteogenesis by direct downregulation of multiple inhibitors ofosteoblastic differentiation.18

Thus, the aim of this study was to investigate both in vivoand in vitro the role of several miRs already implicated inosteoblast differentiation and bone formation in the processof vascular calcification.

RESULTS

AlteredmiR Levels in Aortas of Chronic Renal Failure RatsIn the in vivo model, all nephrectomized rats showed time-dependent increases in serum levels of urea and creatininecomparedwith the control group (Table 1). In addition, serumP and parathyroid hormone (PTH) increased with time inchronic renal failure (CRF) rats fed a high-phosphorus diet(HPD) compared with both the control group and theirrespective normal–phosphorus diet (NPD) groups. Also, a sig-nificant time–dependent increase in Ca in the aorta was seenin CRF rats fed an HPD (Figure 1) compared with either the

control group or the respective NPD–fed group. Similarresults were obtained for some osteoblastic markers, such asalkaline phosphatase (ALP) and osteocalcin (data not shown).Gene expression analysis of eight miRs potentially involved inosteoblastic differentiation (miR-29b, miR-125, miR-133b,miR-135, miR-141, miR-200a, miR-204, and miR-211)showed an increased expression of miR-29b, whereas theexpressions of miR-133b and miR-211 were decreased inthe aortas of the CRF rats fed an HPD compared with eitherthe control group or the CRF groups fed an NPD (Table 2). Asignificant lesser decrease in miR-133b levels was alsoobserved in the aortas of CRF groups fed an NPD. Therewere no significant differences in the levels of the other miRs.

RUNX2, a target gene for the downregulated miRs (miR-133b andmiR-211),15 showed a significant and time–dependentincreased expression (2.40- and 2.91-fold over control at 12 and

Table 1. Serum biochemical parameters of control and CRF rats after 12 and 20 weeks fed an NPD or an HPD

Biochemical Parameters Control NPD (n=9)CRF 12 wk CRF 20 wk

NPD (n=9) HPD (n=9) NPD (n=8) HPD (n=7)

Urea (mg/dl) 34.369.9 119.4632.5a 142.8650.7a 117.9637.7a 183.7631.7a,b

Creatinine (mg/dl) 0.4260.04 1.2860.42a 1.2960.62a 1.2360.35a 1.8360.44a,b

Calcium (mg/dl) 11.460.6 11.260.6 11.462.1 11.960.5 11.062.1Phosphorus (mg/dl) 4.761.0 5.861.3 9.463.4a,b 4.860.9 12.162.4a,b

PTH (pg/ml) 26618 54637 6376723a,b 74653 14796822a,b

Data represent means6SD.aP,0.05 versus control.bP,0.05 versus the same group on an NPD.

Figure 1. High dietary phosphorus increases aortic calcification inCRF rats. Aortic calcium content of control and CRF rats after 12and 20 weeks (w) of uremia and fed an NPD or an HPD. (A)Means6SD of total calcium content determined by the o-cre-solphtalein complexone method. Control, nine rats; CRF 12wNPD, nine rats; CRF 12w HPD, nine rats; CRF 20w NPD, eightrats; CRF 20w HPD, seven rats. *P,0.05 versus control; #P,0.05versus the same group on an NPD. (B) Representative sections ofaortas stained with von Kossa. Original magnification, 320.

2 Journal of the American Society of Nephrology J Am Soc Nephrol 27: ccc–ccc, 2015


20 weeks, respectively) (Figure 2A), whereas activin A receptortype IIA (ACVR2A), b-catenin interacting protein1 (CTNNBIP1), and histone deacetylase 4 (HDAC4), known

target genes for the upregulated miR(miR-29b),18 showed a significant andtime–dependent decreased expression (Figure2, B–D, respectively) exclusively in the aortasof the CRF rats fed an HPD compared withboth control and CRF rats from the samegroups fed an NPD. Only minor increases inRUNX2 levels were observed in CRF rats fedan NPD (1.38- and 1.63-fold over control at12 and 20 weeks, respectively) (Figure 2A).

miR-29b, miR-133b, and miR-211 AreRegulated by Uremic Serum In VitroAs a first in vitro approximation model,primary VSMCs cultured with mediasupplemented with 15% serum from ure-mic rats showed a significant increase inboth mineralization measured as Cadeposition after 4 and 8 days of culture(Figure 3A) and ALP activity (data notshown). In this in vitro model, the analy-sis of the expression of the three miRsregulated in the in vivo model (miR-29b,miR-133b, and miR-211) as well as that oftheir target genes (RUNX2, ACVR2A,CTNNBIP1, and HDAC4) showed thesame pattern of expression observed inCRF rats fed an HPD (Figure 3, B–D, re-spectively).

miR-29b, miR-133b, and miR-211 AreRegulated by High Phosphorus InVitroTo specifically address the effect of high P inthe vascular calcification process, moreprecise calcifying conditions were used,growing VSMCs with medium supplemen-ted with 2 mM Ca and 3 mM P. In thiscalcifying medium, VSMCs showed a sig-nificant increase in mineralization after4 days of culture, which was measured asCa content andAlizarin red staining, reach-ing the highest calcification level at day 8(Figure 4A). ALP activity was increased inparallel to Ca content (data not shown). Inthismodel, under calcifying conditions, theexpression of miR-29b increased and theexpressions of miR-133b and miR-211 de-creased in calcifying conditions (Figure4B), reproducing the behavior in the invivo and in vitro cultures with uremic se-

rum. In addition, the expression patterns of the studied targetgenes followed those in the previous models: RUNX2 in-creased (Figure 4C), whereas the expressions of the inhibitors

Table 2. Relative levels of eight miRs putatively involved in osteoblasticdifferentiation (miR-29b, miR-125, miR-133b, miR-135, miR-141, miR-200a, miR-204, and miR-211) in the aorta of control and CRF rats after 12 and 20 weeks ofuremia and fed either an NPD or an HPD

miRControl NPD

(n=9)

CRF 12 wk CRF 20 wk

NPD (n=9) HPD (n=9) NPD (n=8) HPD (n=7)

miR-29b 1.0060.32 1.1160.23 2.9260.42a,b 0.9860.24 3.8460.84a,b

miR-125 1.0060.37 0.9160.31 1.0860.26 1.1060.78 1.2460.25miR-133b 1.0060.12 0.7160.07a 0.3260.04a 0. 7560.08a 0.4960.19a,b

miR-135 1.0060.47 1.2960. 84 2.8662.21 1.1260. 30 1.5761.09miR-141 1.0060.10 0.9760.66 2.3061.52 2.4763.45 1.3061.62miR-200a 1.0060.12 0.7760.39 3.0162.37 2.0661.83 1.6861.36miR-204 1.0060.47 0.7860.65 1.2760.58 0.9360.44 1.2761.54miR-211 1.0060.18 0.8960.16 0.5860.10a,b 1.0960.19 0.5460.10a,b

All miRs were corrected by small nuclear ribonucleic acid U6 levels and adjusted for the value in thecontrol group, which was set to one. Data represent means6SD.aP,0.05 versus control.bP,0.05 versus the same group on an NPD.

Figure 2. Aortic gene expression of promoters and inhibitors of osteoblastic differ-entiation in CRF. Expression levels of calcification regulating genes in aorta. RT-qPCRof (A) RUNX2 and (B–D) inhibitors of osteoblastic differentiation (ACVR2A, CTNNBIP1,and HDAC4, respectively) in the aortas of control and CRF rats after 12 and 20 weeksfed an NPD or an HPD. Control, nine rats; CRF 12w NPD, nine rats; CRF 12w HPD,nine rats; CRF 20w NPD, eight rats; CRF 20w HPD, seven rats. Data represent means6SD. GAPDH, glyceraldehide-3-phosphate-dehydrogenase; R.U., relative units.*P,0.05 versus control; #P,0.05 versus the same group on an NPD.

J Am Soc Nephrol 27: ccc–ccc, 2015 microRNAs in Vascular Calcification 3

www.jasn.org BASIC RESEARCH

of bone mineralization ACVR2A, CTNNBIP1, and HDAC4decreased (Figure 4D).

Effects of miR-29b, miR-133b, and miR-211 inCalcification Models In VitroTo show that miR-29b, miR-133b, and miR-211 play a directrole in the regulation of VSMCs calcification, two in vitroapproaches were followed. In the first approach, we recreatedthe changes in miRs levels by overexpressing miR-29b orblocking miR-133b and miR-211 in VSMCs (Table 3). A signif-icant increase in Ca depositionwas observed in three conditions(Figure 5A) accompanied by increased expression of RUNX2when miR-133b or miR-211 was blocked (Figure 5B) and de-creased expressions of CTNNBIP1, ACVR2A, and HDAC4when miR-29b was overexpressed (Figure 5C).

In a second approach, we used the two recognized in vitromodels of calcification (uremic serum and high load of P) after

preventing changes in miRs levels to ana-lyze their functional effect on calcification.VSMCs cultured in either condition weretransfected with an antagomir to miR-29bto inhibit its expression or pre–miR-133bor pre–miR-211 to force the expression ofmiR-133b or miR-211, respectively (Table3). In these transfected VSMCs, the increasein Ca content stimulated by uremic serumwas slightly prevented with the three miRs(Figure 6A), with a clear effect in the targetgenes RUNX2 (Figure 6B), ACVR2A, andCTNNBIP1 (Figure 6C). However, the in-crease in Ca content stimulated by high Caand P was greatly prevented in all cases (Fig-ure 6D); also, the increase inALPactivitywasprevented (data not shown). In this lastmodel, the overexpression of miR-133band miR-211 prevented the induction oftheir target gene RUNX2 compared withcells transfected with a negative control(scrambled) (Figure 6E). When miR-29bwas downregulated, increased expression(prevention of the downregulation) ofsome of their target genes (ACVR2A andCTNNBIP1) was observed in the cells cul-tured with calcifying medium (Figure 6F).We did not find statistically significant dif-ferences in Ca content (Figure 6G) orstimulation or repression of the target genes(Figure 6, H and I) when we transfectedthree miRs simultaneously compared withindividual transfections.

Serum Levels of miR-29b, miR-133b,and miR-211To find out whether these miRs (miR-29b,miR-133b, and miR-211) would serve as

serum biomarkers of vascular calcification in rats, serum levelswere analyzed in all calcified groups. No significant differencesbetween control and calcified rats were found for any of threemiRs analyzed (Table 4).

DISCUSSION

The phenotypic diversity of VSMCs seems to depend on bothinnate genetic programs and signals from the environment.19

However, the mechanisms of the phenotypic modulation andosteogenic differentiation of VSMCs are not completely un-derstood, and they are the subject of intense research. Recentstudies have described multiple miRs as important regulatorsof bone-related genes.10,20 In this study, it has been shownthat three miRs previously implicated in different steps ofbone formation and differentiation (miR-29b, miR-133b,

Figure 3. Uremic serum directly increases VSMC calcification in vitro. Effect of uremicserum in primary VSMCs cultured for 0, 4, and 8 days in vitro. (A) Calcium depositionquantified by the o-cresolphtalein complexone method and a representative image ofAlizarin red staining. (B) miR-29b, miR-133b, and miR-211 relative levels, (C) expres-sion levels of a promoter (RUNX2), and (D) several inhibitors (ACVR2A, CTNNBIP1,and HDAC4) of osteoblastic differentiation determined by RT-qPCR. Data representmeans6SD of three experiments performed in triplicate. GAPDH, glyceraldehide-3-phosphate-dehydrogenase; R.U., relative units. *P,0.05 versus 0 days; #P,0.05 ver-sus their respective controls.



and miR-211) are also endogenous regulators of VSMCscalcification and therefore, could be mediators of vascularcalcification (Figure 7).

In the first approach, eight miRs (miR-29b, miR-125, miR-133b, miR-135, miR-141, miR-200a, miR-204, and miR-211)were selected to be examined using an invivo model of vascular calcification on thebasis of nephrectomized rats fed an HPD.Previous studies have shown in rats thatjust CRF is not enough to induce vascularcalcification. In fact, other additional stim-uli are necessary, such as an HPD21,22 orcalcitriol.23 In addition, rats with normalrenal function under the same stimulusmay exhibit certain alterations in serum,but they do not calcify, despite an HPD.21

The eight selected miRs have a knownrole in the osteogenic differentiation inbone, and therefore, their function in theVSMCs transdifferentiation into osteo-blast-like cells was investigated. It is knownthatmiR-133,miR-135,miR-204, andmiR-211 act as negative regulators of osteoblastdifferentiation and subsequent mineraliza-tion, functionally inhibiting differentiationof osteoprogenitorsby attenuating the essen-tial transcription factor RUNX2.24 Specifi-cally, miR-133 negatively regulates RUNX2by direct targeting of its 39 untranslatedregion.9 Furthermore, miR-133 inhibitsnot only translation of RUNX2 but also, incooperationwithmiR-135, bonemorphoge-netic protein 2–induced osteogenic differen-tiation in mouse myoblasts.14 Moreover,miR-133 was absent in the committed oste-oblast precursor cell line from mouse cal-varia MC3T3. All of these reports suggestthat miR-133 represents a key mechanismfor suppressing the osteoblast phenotype innonosseous mesenchymal cells. Our resultsin VSMCs and CRF in vivo showed lowerlevels ofmiR-133bwith hyperphosphatemia,with an associated increase in RUNX2 levelsin the calcification process. Additionally, aslight decrease in miR-133b levels was attrib-uted to the uremic status. The direct impli-cation of this miR in modulating RUNX2expression was further shown through its di-rected overexpression, which resulted in de-creased RUNX2 expression. In the case ofmiR-211, an endogenous attenuator of thistranscription factor, the effect inCadepositiondid not correlate in proportion with the effectin RUNX2, suggesting thatmiR-211 could ex-

ert its effects on calcification through additional pathways. In fact,miR-211 was described to regulate members of theWnt–signalingpathway, a recognized inducer of osteogenesis.25

Figure 4. Calcifying medium directly increases VSMC calcification. Effects of calcifyingmedium (2 mM calcium and 3 mM phosphorus) in primary VSMCs cultured for 0, 4, and 8days in vitro. (A) Calcium deposition quantified by the o-cresolphtalein complexonemethod and a representative image of Alizarin red staining. (B) miR-29b, miR-133b, andmiR-211 relative levels, (C) expression of a promoter (RUNX2), and (D) several inhibitors(ACVR2A, CTNNBIP1, and HDAC4) of osteoblastic differentiation determined by RT-qPCR. Data represent means6SD of three experiments performed in triplicate. GAPDH,glyceraldehide-3-phosphate-dehydrogenase; R.U., relative units. *P,0.05 versus 0 days.

Table 3. Relative levels of miR-29b, miR-133b, and miR-211 reached in VSMCstransfected with the corresponding pre-miRs and antagomirs, cultured for 4 days inbasal medium (DMEM), DMEM+ 15% uremic serum, and DMEM+ 2mMCa + 3mMP, and determined by RT-qPCR

miR DMEM DMEM + 15% Uremic Serum DMEM + 2 mM Ca + 3 mM P

miR-29b 13536100a 0.3260.13b 0.2260.09b

miR-133b 0.2060.09b 22,55362537a 679461206a

miR-211 0.3060.05b 23826126a 338684a

Levels were normalized in each case with small nuclear ribonucleic acid U6 levels and compared withthe miRs levels in scrambled transfected cells, with levels that were set to one. Data represent means6SD of three experiments performed in triplicate.aCells transfected with the corresponding pre-miR.bCells transfected with the corresponding antagomir.



Not all miRs are functional inhibitors of osteoblastogenesis.In fact, miR-29b was reported significantly upregulated duringosteoblastic differentiation, promoting osteogenesis throughdirect downregulation of multiple inhibitors of osteoblasticdifferentiation, including ACVR2A, CTNNBIP1, andHDAC4.18 The CRF animal model fed an HPD used in ourstudy supports the reported role of miR-29b to modulate thetransdifferentiation of VSMCs into osteoblast-like cells.

In summary, in our in vivo model of CRF and high P, onlythree of eight miRs studied (miR-29b, miR-133b, and miR-211)showed changes in their expression that were associated withRUNX2 expression among other osteogenic-modulatinggenes. However, we cannot discard that the other miRs assayed(miR-125, miR-135, miR-141, miR-200a, and miR-204) mightbe mediators in the calcification process not triggered byhigh P.

In vitromodels reproducing our in vivo results further sup-ported the contribution of these three miRs to calcification.Because uremic serum from patients on dialysis increases theexpression of inducers of calcification, such as RUNX2, andconsequently, the VSMCs calcification,26 in the first in vitromodel, calcification in VSMCs was induced by culturing cellswith uremic serum from CRF rats. VSMCs exposed to uremicserum corroborated the decreased expression of miR-133bobserved in vivo. In the uremic state, several known and un-known uremic toxins and inflammatory cytokines, includingmembers of the TNF-a family, are capable of inducing oste-ogenic genes.27–29 Furthermore, patients with CKD frequentlyhave higher serumP levels, and they aremore prone to developsevere vascular calcification. Clearly, another factor that mayexplain the observed effects of the uremic serum could be theslight increase in P levels, which we could see in our in vitrouremic model showing increased expression of miR-29b and

decreased expression of miR-211 but did not observe in ure-mic and not hyperphosphatemic rats. Therefore, the slightincrease in P may aggravate the in vitro VSMCs calcificationalready triggered by the uremic toxins.30 In fact, the effectsseen in the in vitro uremic model are in the same direction asthose observed in the high–Ca and –P model, although lesspronounced. Both models confirmed a direct effect of themiRs in some of their target genes. However, our experimentaldesign does not allow a greater gene inhibitionwith the pool ofmiRs (even in interfering experiments with small interferingRNAs, gene inhibition is never complete). Similarly, overex-pression of a gene should have a limit, and we were not able toreach a greater expression, even with the blockade of severalexogenous miRs. In any case, our results suggest that theseregulatory molecules contribute, in part, to the mechanismimplicated in the phenotypic changes and mineralization ofVSMCs. However, we cannot rule out the effect on miRs levelsof osteogenic transcription factors, such as Osterix, which areknown to downregulate miR-204/211,31 stabilizing a regula-tory feedback loop.

Additional support of the translational relevance of ourfindings would be obtained by examining whether the levels ofthese candidate miRs in the arteries of patients with CKDcorrelate with the degree of calcification.32 Undoubtedly, thisanalysis is beyond the shape of this study. In addition, it wouldbe important to analyze the effect of these miRs at the proteinlevel, because miRs also work by inhibiting translation.

With the current diagnostic techniques relying mainly onimaging techniques, vascular calcification is observedwhen it isalready well established. The identification of noninvasivebiomarkers of subclinical vascular calcification would allowearly interventions to slow its progression to clinical calcifi-cation. Blood sampling is a minimally invasive method to

Figure 5. Changes in VSMC miR expression that promote calcification. Effects of repression and overexpression of miRs in calciumdeposition and gene expression in primary VSMCs transfected with negative control (scrambled), pre–miR-29b, antagomir to miR-133b,or antagomir to miR-211 and cultured with control media for 4 days. (A) Calcium content quantification by the o-cresolphtaleincomplexone method. Expression of (B) the promoter RUNX2 and (C) several inhibitors (ACVR2A, CTNNBIP1, and HDAC4) of osteo-blastic differentiation determined by RT-qPCR. Data represent means6SD of three experiments performed in triplicate. GAPDH,glyceraldehide-3-phosphate-dehydrogenase; R.U., relative units. *P,0.05 versus scrambled transfection.



identify biomarkers. In various disease conditions, the profilesof circulating miRs vary with the degree of disease progres-sion.33–35 At present, no circulating biomarkers are available

for the diagnosis of subclinical vascular calcification, and thismainly relies on imaging techniques that allow evaluation ofcalcifications. However, great repercussions would result from

Figure 6. Changes in VSMC miR expression that prevent calcification. Effects of repression and overexpression of miRs in the calciumdeposition and gene expression in primary VSMCs cultured for 4 days with two calcifying stimuli: (A–C) DMEM supplemented with 15%uremic serum and (D–I) DMEM supplemented with 2 mM Ca + 3 mM P. (A, D, and G) Calcium content quantification by the o-cresolphtaleincomplexone method in VSMCs transfected with (A and D) negative control (scrambled), antagomir to miR-29b, pre–miR-133b, pre–miR-211,or (G) the pool of three miRs. (B, E, and H) Expression levels of the promoter of osteoblastic differentiation RUNX2 in VSMCs transfectedwith (B and E) pre–miR-133b or pre–miR-211 or (H) the pool of three miRs. (C, F, and I) Expression of several inhibitors of osteoblasticdifferentiation (ACVR2A, CTNNBIP1, and HDAC4) in VSMCs transfected with (C and F) antagomir to miR-29b or (I) the pool of three miRsdetermined by RT-qPCR. Data represent means6SD of three experiments performed in triplicate. GAPDH, glyceraldehide-3-phosphate-dehydrogenase; R.U., relative units. *P,0.05 versus scrambled transfection.



finding a serum biomarker of early vascular calcification. Un-fortunately, none of threemiRs analyzed in our study fulfill thefunction of a biomarker, because no significant differenceswere found in their serum levels as calcification progressed.However, our search was limited to eight miRs. The completeregulatory network of several miRs reported to control osteo-genesis by controlling RUNX236,37 suggests that, in thefuture, a molecular signature of vascular calcification consti-tuted by several miRs could be established.38

In summary, our data show that high P upregulates miR-29band downregulates miR-133b and miR-211, altering, as aconsequence, the expressionof several genes involved inVSMCsphenotypic modulation and osteogenic differentiation (Figure7). These results support a key role of miR regulation in con-trolling osteoblastic differentiation. The possibility of usingthese miRs as therapeutic tools to prevent vascular calcificationawaits animal models using miRs to target specific genes in-volved in vascular calcification.

CONCISE METHODS

In Vivo StudyAnimal Model of Vascular CalcificationThe protocol was approved by the Laboratory Animal Ethics

Committee of Oviedo University and adheres to the National

Institutes of Health Guide for the Care and Use of Laboratory

Animals. The study was performed using 4-month-old (350–400 g)

male Wistar rats (n=42). Rats were anesthetized

using methoxifluorane, and CRF was carried

out by 7/8 nephrectomy as previously de-

tailed.39 The CRF rats were divided in two

groups: a group fed a standard rodent chow

with an NPD (0.6% P, 0.6% Ca, and 20% pro-

tein content) and a group fed an HPD (0.9% P,

0.6% Ca, and 20% protein content; Panlab). An

additional group of rats that was sham operated

(normal renal function) and fed for 20 weeks

with the standard rodent chow was also included in the study as

the control group (n=9). The rats were housed in wire cages and

received diet and water ad libitum. After 12 and 20 weeks, rats were

euthanized using CO2 anesthesia, and serum samples were drawn for

analyses. Aortas were removed, washed two times with saline solu-

tion, and divided into three pieces: the first fragment was used for

RNA extraction, the second fragment was used to determine Ca con-

tent, and the third fragment was fixed, embedded in paraffin, and

sliced for the histologic studies.

Biochemical MarkersSerumurea, creatinine,Ca, andPweremeasuredusing amultichannel

auto analyzer (Hitachi 717). SerumPTHwasmeasured by ELISA (Rat

Intact PTH ELISA Kit; Immutopics) following the manufacturer’s

protocols.

Aortic Calcification AnalysisThe calcification of the abdominal aortas of the rats was analyzed by

two methods: total Ca content and von Kossa staining. To determine

the total Ca content, a piece of the abdominal aorta (the 2-cm segment

proximal to the iliac bifurcation) was homogenized with an Ultra-

turrax (OmniHT) in 0.6 N HCl. After stirring at 4°C for 24 hours,

samples were centrifuged. Ca content was determined in the super-

natant by the o-cresolphtalein complexone method40 (Sigma-

Aldrich), and the pellet was resuspended in lysis buffer (125 mM Tris

and 2% SDS, pH 6.8) for protein extraction and quantification by

Lowry method (Bio-Rad). Ca content was normalized to total cell

protein and expressed as micrograms of Ca per milligram of protein.

To perform von Kossa staining, another fragment of the abdominal

aorta was embedded in methylmethacrylate (Sigma-Aldrich). Five

sections 5-mm thick were obtained using a Polycut S Microtome

(Reicher-Jung) and stained following von Kossa method.

Total RNA Isolation, cDNA Synthesis, and Quantitative PCRThe third fragment of the abdominal aorta from the rats was

homogenized in an Ultraturrax (OmniHT) in TRI Reagent (Sigma-

Aldrich) following the manufacturer’s instructions. Total RNA con-

centration and purity were quantified by UV-Vis Spectrophotometry

(NanoDrop Technologies), which measured absorbance at 260 and

280 nm. Reverse transcription was performed with a High-Capacity

cDNA Reverse Transcription Kit (Applied Biosystems) following the

manufacturer’s instructions. Gene expression was measured by ret-

rotranscription quantitative PCR (qPCR) using a Stratagene

Mx3005P (Agilent Technologies). TaqMan qPCR amplification was

performed with gene-specific primers (Gene Expression Assays;

Table 4. Relative serum levels of miR-29b, miR-133b, and miR-211 in controland CRF rats after 12 and 20 weeks of uremia fed either an NPD or an HPD

miR Control NPD (n=9) CRF 12 wk HPD (n=9) CRF 20 wk HPD (n=7)

miR-29b 1.0060.43 1.4360.27 1.1260.10miR-133b 1.0060.33 0.7560.97 1.6161.12miR-211 1.0060.16 1.0860.36 1.2060.44

All miRs levels were corrected by 4.5S levels and expressed relative to control values set to one. Datarepresent means6SD.

Figure 7. Proposed miRs regulation of vascular calcification. HighP upregulates miR-29b and downregulates miR-133b and miR-211.Although the former downregulates inhibitors of calcification, suchas ACVR2A and CTNNBIP1, the latter increases the promoter ofcalcification RUNX2. Both actions contribute to increase vascularsmooth muscle calcification. 2, Downregulation; +, upregulation.



Applied Biosystems) for selected target genes of the regulated miRs:

RUNX2, ACVR2A, CTNNBIP1, and HDAC4. Rat glyceraldehide-3-

phosphate-dehydrogenase was used as the housekeeping gene. The

relative quantitative evaluation of target genes was performed by

comparing threshold cycles using the DD cycle threshold method.41

Mature miR expression levels were also analyzed from total RNA

using TaqMan miRNA Assays (Applied Biosystems) for rat miR-29b,

miR-125,miR-133b,miR-135,miR-141,miR-200a,miR-204, andmi-

R211 according to the manufacturer’s instructions. qPCR was per-

formed on the resulting cDNAusing the appropriate TaqManmiRNA

Gene Expression Assay. Endogenous rat small nuclear RNA U6 for

tissue sample and exogenously added 4.5S for serum samples were

used as references. The relative quantitative evaluation of target genes

was performed by comparing threshold cycles using the DD cycle

threshold method.41

In Vitro StudyPrimary Cell CulturePrimary VSMCs culture was obtained from aortas of healthy 2-

month-oldWistar rats using explants. Briefly, 12 rats were euthanized

using CO2 anesthesia, and abdominal aortas were removed and

placed immediately in ice-cold PBS with 100 units/ml penicillin

and 100 mg/ml streptomycin (Biochrom AG). After washing with

fresh PBS, aortas were cut lengthwise; the endothelial layer was re-

moved carefully and subsequently cut into 2- to 3-mm2 fragments

(explants). The explants were then placed into fibronectin (10 mg/cm2;

Sigma-Aldrich)–coated six–well plastic culture dishes (CorningCostar)

and immersed in a drop of DMEM (Lonza) supplemented with 20%

FBS (HyClone). The explants were placed in a humidified incubator

(5% CO2) at 37°C until they were attached to the surface of the culture

well. Then, an additional 1ml serum–supplemented (20%)DMEMwas

added. After subconfluence, the tissue fragments were removed, and

cells were enzymatically dispersed (0.25% trypsin and 1 mM EDTA),

pooled, and counted with a Scepter 2.0 (Millipore). Cells were seeded

(105 cells per plate) in 100-mmplates (Nunc)with serum-supplemented

(10%)DMEM.Cells obtained by thismethodwere identified as VSMCs

by the following criteria: (1) the cells grew in the characteristic hill and

valley pattern, and (2) immunostaining was positive fora-smoothmus-

cle actin (mAb from Sigma-Aldrich). We used cells between passages 2

and 8 in three wells per condition, and the experiments were performed

three times each.

Induction of Primary VSMCs CalcificationTo induce VSMCs calcification, two different conditions were used.

For the first condition, VSMCs were culture in DMEM supplemented

with 15% uremic rat serum (a pool of sera from 8-week-old CRF rats

containing 10.8mg/dl Ca, 6.7mg/dl P, and 898 pg/ml PTH). A control

condition with 15% healthy rat serum (a pool of sera containing

10.4 mg/dl Ca, 3.6 mg/dl P, and 25 pg/ml PTH) was used for

comparison. To confirm the effect of P, VSMCs were cultured in a

second condition: DMEM F12 + 0.1% BSA with Ca (final concen-

tration of 2 mM) and P (3 mM; calcifying medium). A control

condition (DMEM F12 + 0.1% BSA; control medium) was used as a

reference. In both cases, Ca deposition was determined 4 and 8 days

after the addition of the stimuli.

Determination of Primary VSMCs CalcificationVSMCscalcificationwasdeterminedby twomethods: totalCa content

and Alizarin red staining. For the first method, VSMCs were washed

three timeswith PBS and decalcifiedwith 0.6NHCl at 4°Cwith gently

shaking for 24 hours. Ca released from cell cultures into the super-

natant was determined colorimetrically using the o-cresolphtaleine

complexone method.40 Cells were collected in lysis buffer (125 mM

Tris and 2% SDS, pH 6.8) for protein extraction and quantification by

the Lowry method (Bio-Rad). Ca content was normalized to total cell

protein and expressed as micrograms Ca per milligram protein. For

the second method, Ca deposits were stained with Alizarin red

following a previously described procedure.42

Transfection with miR Antagomirs and PrecursorsVSMCs were seeded at 5000 cells per cm2 in six-well plates (Corning

Costar). The next day, with 60%–70% confluence, cells were trans-

fected. To force expressions of miR-29b, miR-133b, and miR-211,

cells were transfected overnight with 500 pmol precursors of miR-29b,

miR-133b, andmiR-211 (pre–miR-29b, pre–miR-133b, andpre–miR-211,

respectively; Ambion) using the DharmaFECT transfection reagent fol-

lowing the manufacturer’s instructions. To inhibit miR-29b, miR-133b,

andmiR-211, cells were transfected with an antisense oligonucleotide (an-

tagomir) of miR-29b, miR-133b, and miR-211 (500 pmol anti–miR-29b,

anti–miR-133b, and anti–miR-211, respectively; Ambion). In the ex-

periment to analyze the effect of modifying the expression of three

miRs simultaneously, the amount of each miR mentioned above and

the corresponding transfection reagent were used. Transfection

using a scrambled sequence (miRNA Negative Control; Ambion)

was used as a negative control with the same concentrations and ex-

posure times. VSMCs were cultured for 4 days in control medium or

two different calcifying media (DMEM + 15% uremic serum and

DMEM + 2 mM Ca + 3 mM P), and then, Ca content and gene

expression were assessed.

Statistical AnalysesStatistical comparisons between groups were carried out by one-way

ANOVA and t test. The results were expressed as means6SD. Data

from qPCR were transformed before mean and SD determination by

dividing the values by the mean value of the control group. Differ-

ences were considered significant when P,0.05. All statistical anal-

yses were performed using SPSS 17.0 for Windows.

ACKNOWLEDGMENTS

The authors thankDr. PabloRomán-García,Dr.CristinaAlonsoMontes,

Sara Barrio-Vazquez, Ana Rodríguez-Rebollar, and Socorro Braga for

help. We also thank the reviewers for helpful comments and valuable

suggestions and Dr. Adriana Dusso for critical review of the manuscript.

This study was supported by Plan Nacional de Investigación,

Desarrollo e Innovación (I+D+i) 2008-2011; Plan Estatal de I+D+i

2013-2016; Instituto de Salud Carlos III (ISCIII)–Fondo Europeo de

Desarrollo Regional Grants PI07/0893, PI10/0896, and PI13/00497;

Plan de Ciencia, Tecnología e Innovación 2013-2017 del Principado



de Asturias Grant GRUPIN14-028; Fundación para el Fomento en

Asturias de la InvestigaciónCientíficaAplicada y la Tecnología (FICYT);

Instituto Reina Sofía de Investigación Nefrológica; Fundación Renal

Íñigo Álvarez de Toledo; and Red de Investigación Renal (REDinREN)

from ISCIII Grants RD06/0016 and RD12/0021. S.P. was supported by

ISCIII Grant PI09/0415 and Sara Borrell-FICYT Grant CD11/00258,

and N.C.-L. was supported by ISCIII-RedInRen Grant RD06/0016 and

ISCIII–FICYT Grant CA10/01327. I.R. was supported by FICYT.

DISCLOSURESNone.

REFERENCES

1. London GM, Marchais SJ, Guérin AP, Boutouyrie P, Métivier F, deVernejoul MC: Association of bone activity, calcium load, aortic stiff-ness, and calcifications in ESRD. J Am Soc Nephrol 19: 1827–1835,2008

2. Rodríguez-García M, Gómez-Alonso C, Naves-Díaz M, Diaz-Lopez JB,Diaz-Corte C, Cannata-Andía JB; Asturias Study Group: Vascular cal-cifications, vertebral fractures and mortality in haemodialysis patients.Nephrol Dial Transplant 24: 239–246, 2009

3. Naik V, Leaf EM, Hu JH, Yang HY, NguyenNB, Giachelli CM, SpeerMY:Sources of cells that contribute to atherosclerotic intimal calcification:An in vivo genetic fate mapping study. Cardiovasc Res 94: 545–554,2012

4. Speer MY, Yang HY, Brabb T, Leaf E, Look A, Lin WL, Frutkin A, DichekD, Giachelli CM: Smooth muscle cells give rise to osteochondrogenicprecursors and chondrocytes in calcifying arteries. Circ Res 104: 733–741, 2009

5. Xie C, Ritchie RP, Huang H, Zhang J, Chen YE: Smooth muscle celldifferentiation in vitro: Models and underlying molecular mechanisms.Arterioscler Thromb Vasc Biol 31: 1485–1494, 2011

6. House SJ, PotierM, Bisaillon J, Singer HA, TrebakM: The non-excitablesmooth muscle: Calcium signaling and phenotypic switching duringvascular disease. Pflugers Arch 456: 769–785, 2008

7. Shanahan CM: Mechanisms of vascular calcification in CKD-evidencefor premature ageing? Nat Rev Nephrol 9: 661–670, 2013

8. Balderman JA, Lee HY, Mahoney CE, Handy DE, White K, Annis S,Lebeche D, Hajjar RJ, Loscalzo J, Leopold JA: Bone morphogeneticprotein-2 decreases microRNA-30b and microRNA-30c to promotevascular smoothmuscle cell calcification. J AmHeart Assoc 1: e003905,2012

9. Liao XB, Zhang ZY, Yuan K, Liu Y, Feng X, Cui RR, Hu YR, Yuan ZS, Gu L,Li SJ, Mao DA, Lu Q, Zhou XM, de Jesus Perez VA, Yuan LQ: MiR-133amodulates osteogenic differentiation of vascular smooth muscle cells.Endocrinology 154: 3344–3352, 2013

10. Rangrez AY, M’Baya-Moutoula E, Metzinger-Le Meuth V, Hénaut L,Djelouat MS, Benchitrit J, Massy ZA, Metzinger L: Inorganic phosphateaccelerates themigration of vascular smoothmuscle cells: Evidence forthe involvement of miR-223. PLoS ONE 7: e47807, 2012

11. Bartel DP: MicroRNAs: Genomics, biogenesis, mechanism, and func-tion. Cell 116: 281–297, 2004

12. Chen K, Rajewsky N: The evolution of gene regulation by transcriptionfactors and microRNAs. Nat Rev Genet 8: 93–103, 2007

13. Komori T, Yagi H, Nomura S, Yamaguchi A, Sasaki K, Deguchi K,Shimizu Y, Bronson RT, Gao YH, InadaM, SatoM,Okamoto R, KitamuraY, Yoshiki S, Kishimoto T: Targeted disruption of Cbfa1 results in acomplete lack of bone formation owing to maturational arrest of os-teoblasts. Cell 89: 755–764, 1997

14. Li Z, Hassan MQ, Volinia S, vanWijnen AJ, Stein JL, Croce CM, Lian JB,Stein GS: A microRNA signature for a BMP2-induced osteoblast line-age commitment program. Proc Natl Acad Sci U S A 105: 13906–13911, 2008

15. Huang J, Zhao L, Xing L, Chen D: MicroRNA-204 regulates Runx2protein expression and mesenchymal progenitor cell differentiation.Stem Cells 28: 357–364, 2010

16. Goettsch C, Rauner M, Pacyna N, Hempel U, Bornstein SR, HofbauerLC: miR-125b regulates calcification of vascular smooth muscle cells.Am J Pathol 179: 1594–1600, 2011

17. Itoh T, Nozawa Y, Akao Y: MicroRNA-141 and -200a are involved inbone morphogenetic protein-2-induced mouse pre-osteoblast differ-entiationby targeting distal-less homeobox 5. J Biol Chem284: 19272–19279, 2009

18. Li Z, HassanMQ, Jafferji M, Aqeilan RI, Garzon R, Croce CM, vanWijnenAJ, Stein JL, Stein GS, Lian JB: Biological functions of miR-29b con-tribute to positive regulation of osteoblast differentiation. J Biol Chem284: 15676–15684, 2009

19. Rensen SS, Doevendans PA, van Eys GJ: Regulation and characteristicsof vascular smooth muscle cell phenotypic diversity. Neth Heart J 15:100–108, 2007

20. Rangrez AY, Massy ZA, Metzinger-Le Meuth V, Metzinger L: miR-143and miR-145: Molecular keys to switch the phenotype of vascularsmooth muscle cells. Circ Cardiovasc Genet 4: 197–205, 2011

21. Cozzolino M, Staniforth ME, Liapis H, Finch J, Burke SK, Dusso AS,Slatopolsky E: Sevelamer hydrochloride attenuates kidney and car-diovascular calcifications in long-term experimental uremia. Kidney Int64: 1653–1661, 2003

22. Román-García P, Carrillo-LópezN, Fernández-Martín JL,Naves-DíazM,Ruiz-Torres MP, Cannata-Andía JB: High phosphorus diet inducesvascular calcification, a related decrease in bone mass and changes inthe aortic gene expression. Bone 46: 121–128, 2010

23. Du Y, Gao C, Liu Z, Wang L, Liu B, He F, Zhang T, Wang Y, Wang X, XuM, LuoGZ, Zhu Y, XuQ,Wang X, KongW: Upregulation of a disintegrinand metalloproteinase with thrombospondin motifs-7 by miR-29 re-pression mediates vascular smooth muscle calcification. ArteriosclerThromb Vasc Biol 32: 2580–2588, 2012

24. Sun Y, Byon CH, Yuan K, Chen J, Mao X, Heath JM, Javed A, Zhang K,Anderson PG, Chen Y: Smooth muscle cell-specific runx2 deficiencyinhibits vascular calcification. Circ Res 111: 543–552, 2012

25. Atlasi Y, Noori R, Gaspar C, Franken P, Sacchetti A, Rafati H,Mahmoudi T, Decraene C, Calin GA, Merrill BJ, Fodde R: Wnt sig-naling regulates the lineage differentiation potential of mouse em-bryonic stem cells through Tcf3 down-regulation. PLoS Genet 9:e1003424, 2013

26. Moe SM,DuanD,Doehle BP,O’Neill KD, ChenNX: Uremia induces theosteoblast differentiation factor Cbfa1 in human blood vessels. KidneyInt 63: 1003–1011, 2003

27. Tintut Y, Patel J, Parhami F, Demer LL: Tumor necrosis factor-alphapromotes in vitro calcification of vascular cells via the cAMP pathway.Circulation 102: 2636–2642, 2000

28. Stenvinkel P, Ketteler M, Johnson RJ, Lindholm B, Pecoits-Filho R,Riella M, Heimbürger O, Cederholm T, Girndt M: IL-10, IL-6, and TNF-alpha: Central factors in the altered cytokine network of uremia—thegood, the bad, and the ugly. Kidney Int 67: 1216–1233, 2005

29. Barreto FC, Barreto DV, Liabeuf S, Drüeke TB, Massy ZA: Effects ofuremic toxins on vascular and bone remodeling. Semin Dial 22: 433–437, 2009

30. Moe SM, Chen NX: Pathophysiology of vascular calcification in chronickidney disease. Circ Res 95: 560–567, 2004

31. ChenQ, Liu W, Sinha KM, Yasuda H, de Crombrugghe B: Identificationand characterization of microRNAs controlled by the osteoblast-specific transcription factor Osterix. PLoS ONE 8: e58104, 2013

32. Wen P, Cao H, Fang L, Ye H, Zhou Y, Jiang L, Su W, Xu H, HeW, Dai C,Yang J:miR-125b/Ets1 axis regulates transdifferentiation and calcification



of vascular smooth muscle cells in a high-phosphate environment. ExpCell Res 322: 302–312, 2014

33. Mo MH, Chen L, Fu Y, Wang W, Fu SW: Cell-free Circulating miRNABiomarkers in Cancer. J Cancer 3: 432–448, 2012

34. Zhao A, Li G, Péoc’hM, Genin C, Gigante M: SerummiR-210 as a novelbiomarker formolecular diagnosis of clear cell renal cell carcinoma. ExpMol Pathol 94: 115–120, 2013

35. DieckmannKP, SpiekermannM,BalksT, Flor I, LöningT,Bullerdiek J, BelgeG: MicroRNAs miR-371-3 in serum as diagnostic tools in the managementof testicular germ cell tumours. Br J Cancer 107: 1754–1760, 2012

36. Zhang Y, Xie RL, CroceCM, Stein JL, Lian JB, vanWijnenAJ, SteinGS: Aprogram of microRNAs controls osteogenic lineage progression bytargeting transcription factor Runx2. Proc Natl Acad Sci U S A 108:9863–9868, 2011

37. Hu R, LiuW, Li H, Yang L,ChenC, Xia ZY, Guo LJ, XieH, ZhouHD,WuXP,Luo XH: A Runx2/miR-3960/miR-2861 regulatory feedback loop duringmouse osteoblast differentiation. J Biol Chem 286: 12328–12339, 2011

38. Goettsch C, Hutcheson JD, Aikawa E: MicroRNA in cardiovascularcalcification: Focus on targets and extracellular vesicle delivery mech-anisms. Circ Res 112: 1073–1084, 2013

39. Naves-Díaz M, Carrillo-López N, Rodríguez-Rodríguez A, Braga S,Fernández-Coto T, Lopez-Novoa JM, López-Hernández F, Cannata-AndíaJB: Differential effects of 17beta-estradiol and raloxifene on bone andlipid metabolism in rats with chronic kidney disease and estrogen in-sufficiency. Menopause 17: 766–771, 2010

40. Cardús A, Panizo S, Parisi E, Fernandez E, Valdivielso JM: Differentialeffects of vitamin D analogs on vascular calcification. J Bone Miner Res22: 860–866, 2007

41. Livak KJ, Schmittgen TD: Analysis of relative gene expression datausing real-time quantitative PCR and the 2(-Delta C(T)) Method.Methods 25: 402–408, 2001

42. Gregory CA, Gunn WG, Peister A, Prockop DJ: An alizarin red-basedassay of mineralization by adherent cells in culture: comparison withcetylpyridinium chloride extraction. Anal Biochem 329: 77–84, 2004



micrornas 29b, 133b, and 211 regulate vascular smooth

Documents