oxidant generation predominates around calcifying foci...
TRANSCRIPT
Oxidant Generation Predominates Around Calcifying Fociand Enhances Progression of Aortic Valve Calcification
Marcel Liberman, Estevao Bassi, Marina Kamla Martinatti, Fabio Cerqueira Lario,Joao Wosniak Jr, Pablo M.A. Pomerantzeff, Francisco R.M. Laurindo
Objective—We hypothesized that reactive oxygen species (ROS) contribute to progression of aortic valve (AV)calcification/stenosis.
Methods and Results—We investigated ROS production and effects of antioxidants tempol and lipoic acid (LA) incalcification progression in rabbits given 0.5% cholesterol diet �104 IU/d Vit.D2 for 12 weeks. Superoxide and H2O2
microfluorotopography and 3-nitrotyrosine immunoreactivity showed increased signals not only in macrophages butpreferentially around calcifying foci, in cells expressing osteoblast/osteoclast, but not macrophage markers. Such cellsalso showed increased expression of NAD(P)H oxidase subunits Nox2, p22phox, and protein disulfide isomerase. Nox4,but not Nox1 mRNA, was increased. Tempol augmented whereas LA decreased H2O2 signals. Importantly, AVcalcification, assessed by echocardiography and histomorphometry, decreased 43% to 70% with LA, but increased withtempol (P�0.05). Tempol further enhanced apoptosis and Nox4 expression. In human sclerotic or stenotic AV, wefound analogous increases in ROS production and NAD(P)H oxidase expression around calcifying foci. An in vitrovascular smooth muscle cell (VSMC) calcification model also exhibited increased, catalase-inhibitable, calcium depositwith tempol, but not with LA.
Conclusions—Our data provide evidence that ROS, particularly hydrogen peroxide, potentiate AV calcificationprogression. However, tempol exhibited a paradoxical effect, exacerbating AV/vascular calcification, likely because ofits induced increase in peroxide generation. (Arterioscler Thromb Vasc Biol. 2008;28:463-470)
Key Words: calcification � atherosclerosis � antioxidants � valves � free radicals
Degenerative aortic valve (AV) stenosis, the third mostprevalent cardiovascular disease in the elderly,1
shares common risk factors and pathophysiological fea-tures with atherosclerosis.2– 6 Although the role of oxidativestress in atherosclerosis is well explored,7,8 it is unclearwhether redox processes contribute to progression of AVcalcification.2,3,9–11,15,16 Scarce observations provide indirectsupport for this hypothesis.10 In vitro studies showed thatexogenous superoxide, hydrogen peroxide, or other oxidantsincrease the number and activity of calcifying vascular cells(CVCs),11 referred to as a specific subpopulation of cells,derived from (de)differentiation of vascular smooth musclecells,12 pericytes, or mesenchymal cells13 that can producehydroxyapathite in the vascular wall.14 In addition, reactiveoxygen species (ROS) mediate increase in BMP2 expressionand signaling, favoring osteogenesis.2 On the other hand,calcium resorption by osteoclasts is dependent on ROSderived from its own NAD(P)H oxidase,15 whereas nitricoxide induces osteoclast detachment and inhibits calciumresorption.16 Recent data from an experimental mouse modelof aortic stenosis suggested locally increased superoxidegeneration.3 Observational clinical studies with statins indi-
cated possible decrease in calcification progression in hyper-cholesterolemic patients,4 but a prospective clinical trial innormocholesterolemic subjects showed lack of effect.5 In thepresent study, we investigated the occurrence and microto-pography of ROS generation in an experimental rabbit modelof AV calcification and sclerosis and in specimens fromhuman with AV sclerosis or stenosis. In addition, the role ofredox processes in the progression of AV calcification wastested by assessing the effects of 2 antioxidants, tempol andlipoic acid (LA), both in a rabbit model17 and in an in vitromodel of VSMC calcification.18
MethodsAn extended Methods section is available as supplemental material(available online at http://atvb.ahajournals.org).
Human Aortic ValvesAV from patients with stenosis (n�5) or individuals with sclerosis(n�4), collected respectively at surgery or autopsy (�6hs postmor-tem), were analyzed for ROS microtopography, histology, andimmunohistochemistry for NAD(P)H oxidase subunits and com-pared with young autopsied controls (n�6).
Original received September 25, 2007; final version accepted December 12, 2007.From the Vascular Biology Laboratory, Heart Institute(InCor), University of Sao Paulo School of Medicine, Sao Paulo, Brazil.Correspondence to Francisco R.M. Laurindo, MD, PhD, Vascular Biology Laboratory, Heart Institute (INCOR), University of Sao Paulo School of
Medicine, Av. Eneas Carvalho Aguiar, 44, Annex II, 9th floor, CEP 05403-000, Sao Paulo, Brazil. E-mail [email protected]© 2008 American Heart Association, Inc.
Arterioscler Thromb Vasc Biol is available at http://atvb.ahajournals.org DOI: 10.1161/ATVBAHA.107.156745
463
by guest on May 19, 2018
http://atvb.ahajournals.org/D
ownloaded from
by guest on M
ay 19, 2018http://atvb.ahajournals.org/
Dow
nloaded from
by guest on May 19, 2018
http://atvb.ahajournals.org/D
ownloaded from
by guest on M
ay 19, 2018http://atvb.ahajournals.org/
Dow
nloaded from
by guest on May 19, 2018
http://atvb.ahajournals.org/D
ownloaded from
by guest on M
ay 19, 2018http://atvb.ahajournals.org/
Dow
nloaded from
by guest on May 19, 2018
http://atvb.ahajournals.org/D
ownloaded from
by guest on M
ay 19, 2018http://atvb.ahajournals.org/
Dow
nloaded from
by guest on May 19, 2018
http://atvb.ahajournals.org/D
ownloaded from
by guest on M
ay 19, 2018http://atvb.ahajournals.org/
Dow
nloaded from
by guest on May 19, 2018
http://atvb.ahajournals.org/D
ownloaded from
by guest on M
ay 19, 2018http://atvb.ahajournals.org/
Dow
nloaded from
by guest on May 19, 2018
http://atvb.ahajournals.org/D
ownloaded from
by guest on M
ay 19, 2018http://atvb.ahajournals.org/
Dow
nloaded from
by guest on May 19, 2018
http://atvb.ahajournals.org/D
ownloaded from
by guest on M
ay 19, 2018http://atvb.ahajournals.org/
Dow
nloaded from
by guest on May 19, 2018
http://atvb.ahajournals.org/D
ownloaded from
by guest on M
ay 19, 2018http://atvb.ahajournals.org/
Dow
nloaded from
by guest on May 19, 2018
http://atvb.ahajournals.org/D
ownloaded from
by guest on M
ay 19, 2018http://atvb.ahajournals.org/
Dow
nloaded from
by guest on May 19, 2018
http://atvb.ahajournals.org/D
ownloaded from
by guest on M
ay 19, 2018http://atvb.ahajournals.org/
Dow
nloaded from
by guest on May 19, 2018
http://atvb.ahajournals.org/D
ownloaded from
by guest on M
ay 19, 2018http://atvb.ahajournals.org/
Dow
nloaded from
by guest on May 19, 2018
http://atvb.ahajournals.org/D
ownloaded from
by guest on M
ay 19, 2018http://atvb.ahajournals.org/
Dow
nloaded from
by guest on May 19, 2018
http://atvb.ahajournals.org/D
ownloaded from
Rabbit ModelThe rabbit AV calcification model modified from Drolet et al17
involved administration for 3 months of 0.5% cholesterol-enricheddiet�vitamin D2 104 IU/d (HC�vitD rabbits, n�34). SomeHC�vitD rabbits concomitantly received tempol 100 �mol/kg/d(n�15) or lipoic acid (LA) 120 �mol/kg/d (n�11) in drinking water.Controls received 0.5% cholesterol alone (n�9) or normal chow(n�32). Plasma cholesterol, calcium, phosphorus, and creatininelevels were analyzed.
After the 12th week, AV were collected and processed formorphological analysis and collagen histomorphometry, vonKossastaining for calcium detection, immunohistochemistry for NAD(P)Hoxidase subunits, protein disulfide isomerase (PDI), macrophagemarker RAM11, nitrotyrosine, Ki67 and osteoblast markers, immu-nofluorescence for NAD(P)H oxidase subunits, terminal deoxynu-cleotidyl transferase-mediated dUTP nick end-labeling (TUNEL),and TRAP histochemistry for osteoclasts. Quantitative morphomet-ric analysis was performed by Leica Quantimet software.
In Situ ROS GenerationIn situ microfluorotopography of dihydroethidium (DHE) oxidationproducts was performed as described,8 with 3 �mol/L final DHEconcentration. Slides were analyzed by confocal microscopy (ZeissLSM510) with laser excitation at 488 nm and emission at 610 nm,detected by 560 nm longpass filter. We also performed a novelanalogous in situ detection of hydrogen peroxide by 2�,3�-dichlorofluorescein diacetate (DCF-DA), using a similar protocol,with final DCF-DA concentration of 3 �mol/L. Controls, performedby incubating slides for 30 minutes with Peg-SOD (500U/mL) orPeg-catalase (400U/mL), indicated preferential detection of super-oxide with DHE and hydrogen peroxide with DCF-DA. Quantitativeanalysis of fluorescent images was performed with Image J (NIH)software.
Real-Time Polymerase Chain ReactionQuantitative expression of Nox1 or Nox4 mRNA was performed inaortic tissue as detailed in supplement, with primers designedaccording to rabbit sequences.
EchocardiographyValve area, left ventricular mass and dimensions, peak and meanaortic valve flow velocity were assessed by ultrasound/doppler asdetailed elsewhere3 and in supplement, with 12-MHZ phased-arrayprobe and Philips Sonos 5500 system. Exams were performed atbeginning of protocol and after 3 months of intervention.
Calcification AnalysisA semi-quantitative index (range 0 to 3) based on opacity and valvemotility was used for estimating AV echogenicity, an index of valvethickening and calcification, as follows: 0-absent, 1-mild;2-moderate, with preserved motility/opening; 3-severe, with de-creased motility/opening. Data were analyzed on a blinded fashion.Histomorphometry of calcium conglomerates was performed withLeica Quantimet software in von Kossa and hematoxylin-eosinstained sections. Ex vivo 64-row computed tomography of aorticspecimens was performed as detailed in supplement.
In Vitro VSMC Calcification ModelThe model was described by Shioi et al.18 Rabbit aortic vascularsmooth muscle cells cultured in DMEM with FBS10%, at 70%confluence and 4th passage, were supplemented during 14 days with6 mmol/L calcium chloride, alone or in combination with 10 mmol/L�-glycerophosphate, or both in combination with 0.1 �mol/Lvitamin-D3. The latter combination was supplied in the absence orpresence of 100 �mol/L tempol or 100 �mol/L LA for 14 days,changing medium every 3 days. Exogenous hydrogen peroxide(50 �mol/L), catalase (500U/mL), or the NO donor NOC-18(30 �mol/L) were added in specific experiments, changing mediumevery 3 days. We used either primary cells or cells from animmortalized line.19
CVC layer calcium content was assessed by colorimetric assay, asdetailed in supplement. Superoxide production was assessed throughhigh-performance liquid chromatography (HPLC) analysis of DHEoxidation products as described20 and membrane fraction NAD(P)Hoxidase activity through lucigenin (5 �mol/L) chemiluminescence.
StatisticsData are mean�SE. Comparisons were tested through 1-wayANOVA plus Student–Newman-Keuls or Dunnet post-hoc tests, at a0.05 significance level.
ResultsHuman Aortic ValvesHuman stenotic and to a lesser degree sclerotic AV exhibitedthickening, fibrosis, elastolysis, and calcification (notshown). Microtopography of superoxide production showedincreased fluorescence signals in sclerotic, and specially instenotic AV, observed mainly around calcifying foci. Signalsdecreased on specimen incubation with Peg-SOD (Figure 1Athrough 1D). Expression of p22phox and Nox2 NAD(P)H
Figure 1. Microfluorotopography of DHE oxidation products inhuman aortic valves (AV) showing increased fluorescence sig-nals vs Control (A) in sclerotic (B), and specially in stenoticAV (C), observed mainly around calcifying foci (Ca2�), whichdecreased after incubating slides with 500U/mL Peg-SOD (D).NAD(P)H oxidase subunit expression was also increased aroundsuch foci (Ca2�), shown by both p22phox (E) and Nox2 (F)peroxidase-immunostaining. Below, graph representing total flu-orescence intensity (area units [AU], corresponding to panels Athrough D) normalized for control values. Each bar representsmean�SE of at least 3 different valves. *P�0.05 vs Control;#P�0.05 between specified groups.
464 Arterioscler Thromb Vasc Biol March 2008
by guest on May 19, 2018
http://atvb.ahajournals.org/D
ownloaded from
oxidase subunits was increased around such foci (Figure 1Eand 1F; supplemental Figure I).
Rabbit ModelPreliminary experiments with serial echocardiograms showedprogressive increases in AV calcification in HC�vitD rab-bits, with accelerated progression after 8 weeks.
Histological CharacterizationAV calcification accompanied the increased plasma levels ofcholesterol and calcium-phosphorus product, mainly becauseof phosphate increase, and minor increase in serum creatininelevels (supplemental Table I). AV from HC�vitD rabbitsshowed early stages of cusp mineralization (supplementalFigures II and III): thickening caused by collagen infiltrationand increased cellularity, massive macrophage infiltration inthe aortic face, and several conglomerate foci of subendothe-lial calcification. In 3 intact rabbits fed regular chow, tempolwas given at 100 �mol/kg/d for 3 months. There were nochanges in blood biochemical parameters and AV calcifica-tion, showing that tempol alone had no effects in the studyvariables (not shown).
Echocardiographic DataSignificant increase in echogenicity occurred in HC�vitDrabbits (supplemental Figure III). Within the time frame ofstudy, the absolute AV area and aortic jet/left ventricularoutflow tract velocity ratio did not differ among groups(supplemental Figure IV). Final AV area decreased versusbaseline in all HC�vitD rabbits, but such decrease reachedstatistical significance only in HC�vitD rabbits given tem-pol. Concomitantly, there was increase in posterior wallthickness (PWT), revealing concentric ventricular remodelingin all HC�vitD rabbits (supplemental Figure IV). Thus, suchanimal model shows changes consistent with AV sclerosis.
Reactive Oxygen Species GenerationDHE fluorescence microtopography (Figure 2) showed in-creased superoxide signals corresponding to macrophageinfiltrate and an even more pronounced increase, �6-foldabove normal, in cells surrounding calcifying foci, which, asshown below, did not stain for macrophage markers (Figure2A). Signals were importantly reduced after specimen incu-bation with Peg-SOD. Superoxide signals were decreased inLA- and specially tempol-treated rabbits (Figure 2G and 2H).
Figure 2. A, Immunostaining for rabbit macrophages (RAM11 antibody) in AV of HC�VitD rabbit, depicting massive macrophage infil-tration in aortic face, and calcium conglomerate (Ca2�) below. B through H, Microfluorotopography of DHE oxidation products in rabbitAV. B, Normal control; C, HC�VitD; D, Calcium conglomerate from HC�VitD rabbit at 200� magnification; E, Cholesterol-only controlrabbit; F, HC�VitD�in vitro Peg-SOD; G, HC�VitD�tempol; H, HC�VitD�LA. Fluorescence intensity was evident in macrophage infil-tration area (asterisk), but particularly pronounced around calcifying foci (arrows). Signals were markedly reduced by incubating slideswith Peg-SOD (500 U/mL) (F). In both tempol and LA-treated rabbits, fluorescence was markedly reduced (G, H). Below left, Graph rep-resenting fluorescence intensity normalized for Normal control values; Below right, Graph showing number of DAPI-stained nuclei perfield. Letters below each bar correspond to the above panels. Bars represents mean�SE of at least 3 different rabbits. *P�0.05 vs Nor-mal; #P�0.05 between specified groups.
Liberman et al ROS Contribute to Aortic Valve Calcification 465
by guest on May 19, 2018
http://atvb.ahajournals.org/D
ownloaded from
Overall cellularity counts using DAPI nuclei staining showedimportant decrease in cell number around calcifying foci withtempol. Thus, although tempol clearly decreased superoxidesignals in most cells, cell loss appears to have contributed tofluorescence decrease around calcifying foci (Figure 2G).
In situ DCF-DA fluorescence microtopography showedincreased signal in AV cusps in HC�vitD rabbits (Figure 3),which was decreased after specimen incubation with Peg-Catalase, thus indicating that signals reflect mainly hydrogenperoxide. Importantly, hydrogen peroxide signals were in-creased in AV from tempol-treated rabbits (Figure 3E), andmarkedly decreased in LA-treated rabbits (Figure 3F).
In HC�vitD rabbits, nitro/oxidative stress preferentially oc-curred around calcifying foci, and to a lesser extent in macro-phage staining area, as shown from increased 3-nitrotyrosineimmunostaining (Figure 4), which was decreased with LA andparticularly with tempol (supplemental Figure V).
Expression of NAD(P)H Oxidase SubunitsNAD(P)H oxidase subunits p22phox, Nox2, and particularlyits newly-described regulator protein disulfide isomerase19
were strongly expressed around calcification, coincident withthe site of RAM11-negative cells and highest ROS produc-tion (Figure 4). Both Nox2 and protein disulfide isomerasecolocalized with p22phox in such cells (supplemental FigureVI). Analysis of Nox1 and Nox4 mRNA expression byreal-time PCR showed increase in Nox4 expression inHC�vitD rabbits, reflected by increased Nox4/Nox1 ratio,which was especially increased with tempol (Figure 5).
Phenotypic Markers of CVC/OsteoclastsSeveral markers were analyzed for assessing the phenotype ofcells surrounding calcifying foci. Importantly, staining formacrophage marker RAM11 was absent around calcifyingfoci (Figure 2A; supplemental Figure II). There was increasedimmunoreactivity for the osteoblast differentiation markersosteopontin and Cbfa-1 in HC�vitD rabbits (supplementalFigure VIIE and VIIF). Osteopontin mRNA was also in-creased and was unchanged by antioxidants (supplementalFigure VIIC). In addition, osteoclasts were identified bypositive TRAP histochemistry (supplemental Figure VIID).This implicates ROS generation in cellular events related tocalcification process.
Effects of Antioxidants on AV Calcification andCell InfiltrationEffects of tempol or LA in extent of AV calcification wereassessed by 3 methods (Figure 5). Echogenicity, a semiquanti-tative AV thickness/calcification index, was increased inHC�vitD rabbits and unchanged by tempol. Importantly, LA-treated rabbits showed significant 70% decrease in echogenicity,not statistically different vs Normal. Histomorphometric analy-sis, able to detect larger mineral conglomerates, showed in-creased calcification in HC�vitD rabbits, which was furtherenhanced by tempol, and prevented by 43% with LA (Figure5B). Ex vivo 64-row multidetector CT-scan, a method able toidentify only large well-organized calcium deposits, showedincreased opacity signals only in aortas from tempol-treatedHC�vitD rabbits. These findings were independent of calcium-phosphorus product or creatinine levels (supplemental Table I),which were similar among all groups.
Of note, TUNEL-positive cells increased in HC�vitDrabbits and were further enhanced with tempol, especiallyaround calcifying foci, whereas overall cell proliferation,which was increased in HC�vitD rabbits, was decreased withtempol (supplemental FigureVIIIA and VIIIB). Macrophageinfiltration was increased in HC�vitD rabbits and unchangedby tempol or LA, being 2-fold greater in rabbits givencholesterol only (supplemental FigureVIIID). Collagen dep-osition showed minor nonstatistically significant changesamong groups (supplemental Figure VIIIC).
Figure 3. A through F, Microfluorotopography of DCF-DA oxida-tion products in rabbit AV. A, Normal control; B, Cholesterol-onlycontrol rabbit; C, HC�VitD; D, HC�VitD�in vitro Peg-Catalase; E,HC�VitD�tempol; F, HC�VitD�LA. Fluorescence was increasedin HC�vitD (C) and particularly pronounced with tempol (E),decreasing after in vitro incubation with 400 U/mL Peg-catalase(D). Below, Graph representing total fluorescence intensity normal-ized for Normal control values. Each bar represents mean�SEM ofat least 3 different rabbits. Letters below each bar correspond tothe above panels, and H is HC�Vit.D2�tempol�in vitro Peg-catalase (figure not shown). *P�0.05 vs Normal; #P�0.05 betweenspecified groups.
466 Arterioscler Thromb Vasc Biol March 2008
by guest on May 19, 2018
http://atvb.ahajournals.org/D
ownloaded from
In Vitro ModelIn cultured CVCs, calcium deposition increased in proportionwith strength of calcifying stimuli during 14 days. Such depo-sition was further enhanced with tempol, but not with LA(Figure 6). CVC incubation with exogenous 50 �mol/L hydro-gen peroxide amplified mineralization response irrespective ofantioxidant coincubation. Incubation with catalase did notchange basal calcification, but completely prevented increasedcalcification induced by tempol. Incubation with NO donorNOC-18 (30 �mol/L) for 14 days did not change CVC calciumdeposition (not shown). Thus, while exogenous oxidants in-creased cell calcification, the mechanism of calcification appearsto differ in vitro (versus in vivo) enough that baseline CVCcalcification is redox-insensitive. Superoxide production, as-sessed by HPLC analysis of 2-hidroxyethidium (Figure 6), aswell as membrane fraction NAD(P)H oxidase activity (supple-mental Figure IX) was increased with calcifying stimuli, par-tially reduced with tempol and strongly decreased with LA.
DiscussionConvergence between ROS and valve calcification is usuallyregarded in a chronic inflammation context. Thus, in our rabbitmodel, the observed topography of ROS production aroundcalcifying foci from cells displaying phenotypic markers ofosteoblasts or osteoclasts, but not of macrophages, is an impor-tant novel finding, although contribution of other phenotypescannot be excluded. Analogous patterns occurred in human AV
sclerosis or stenosis, despite advanced disease in the latter. Therabbit model—mimicking earlier human disease—displayedmassive AV infiltration of macrophage foam cells exhibitingincreased superoxide signals versus control, but not as marked asin cells around calcification. This probably reflects the role ofROS as signaling intermediates of cellular metabolic or devel-opmental processes occurring during CVC or osteoclast forma-tion and activity.2,6,11,16 Moreover, these data stress that evenwith the likely occurrence of substantial lipoprotein oxidation,cellular signaling still seems of prime or at least comparableimportance as ROS source.7 Thus, our data show that oxidantstress pattern in valve calcification shares partial resemblancewith atherosclerosis, and extends beyond vascular or inflamma-tory cell activation. Of note, tempol and LA decreased superox-ide levels without changes in macrophage density (supplementalFigure VIIID).
Decreased calcification progression with LA in our rabbitmodel indicates that ROS potentiate AV calcification. Thelarge effect of LA verified at echogenicity analysis contrastswith the smaller effect evident at histomorphometry. Echo-genicity reflects AV thickening and microcalcification, sug-gesting LA effects in halting early onset of calcium conglom-erates, because macrophage and collagen densities wereunchanged (supplemental FigureVIIIC and VIIID). Contrar-ily, histology preferentially shows larger mineral conglomer-ates. This could indicate that once a threshold conglomeratesize is reached, further growth occurs rapidly and overrides
Figure 4. A, AV from Normal rabbit; B and C, AV from HC�VitD rabbits at 100 and 1000� magnification, respectively. Panels showperoxidase-immunostaining for 3-nitrotyrosine, p22phox, Nox2, and protein disulfide isomerase (PDI) in each column. Nitro/oxidative stresswas evident around calcifying foci in HC�VitD rabbits, shown by increased 3-nitrotyrosine staining, coincident with expression of NAD(P)Hoxidase subunits and PDI. Dilution of primary antibodies was 1:200 (p22phox, goat), 1:200 (Nox2, goat), and 1:32000 (PDI, mouse).
Liberman et al ROS Contribute to Aortic Valve Calcification 467
by guest on May 19, 2018
http://atvb.ahajournals.org/D
ownloaded from
antioxidant effects, an explanation in line with known expo-nential growth rate of calcification.6 This could also reflect alimitation of our models, related to abnormally high calcifi-cation stimulus attributable to induced increase in calcium-phosphorus product, which can activate CVC differentiationvia phosphate cotransporter Pit-1.21 Although such increase isfound in some diseases, degenerative AV calcification usu-ally occurs under normal calcium-phosphorus product. Infact, in our in vitro model, LA or catalase showed little effectin basal calcification, whereas exogenous peroxide clearlyincreased calcium deposits. Another methodological caveat isthat total DHE fluorescence analysis, used because calciumdeposits and small rabbit AV size limited use of othermethods, lacks specificity for superoxide.20 Thus, whereasPeg-SOD controls indicate that superoxide contributes toDHE fluorescence, other oxidants, heme or peroxidasesmight play additional roles.20 Analogous observations applyto in situ DCF studies regarding hydrogen peroxide detection.On the other hand, 2-hydroxyethidium measurements in CVCin vitro provide more specific evidence for superoxide.20
The NAD(P)H oxidase complex exhibited dynamic expres-sion changes during calcification, with upregulation of Nox4and Nox2, but not Nox1. ROS topography around calcifyingfoci coincided with increased expression of subunits p22phoxand Nox2 and enhanced 3-nitrotyrosine staining. Proteindisulfide isomerase, recently described by us to assistNAD(P)H oxidase activity,19 was strongly upregulated at thislocation. This pattern is in line with known roles of Nox4 incell differentiation/apoptosis and of Nox2 in inflammation.22
The precise correlation between each Nox isoform and celltype remains undefined, considering lack of available infor-mation about Nox subtypes in CVCs/osteoblats. In oste-oclasts, Nox2 is prominent, whereas Nox4 is upregulated onNox2 knockout, both supporting bone resorption.23
Together, our observations implicate hydrogen peroxide asa species contributing to vascular/AV calcification. Tempol,which increased hydrogen peroxide production, promotedenhanced calcification in vivo or in vitro, contrary to LA,which promoted hydrogen peroxide decrease and haltedcalcification progression in vivo. Of note, both antioxidantspromoted comparable superoxide decrease. Moreover, CVCincubation with exogenous hydrogen peroxide promotedenhanced calcification, whereas analogous incubation withnitric oxide donor was without effect. In addition, catalaseprevented tempol-associated increase in CVC calcification.The importance of hydrogen peroxide is in line with observedincrease in Nox4, which may preferentially generate hydro-gen peroxide and, in turn, is induced by this species.22 Indeed,in our CVC, Nox4/Nox1 ratio increased 1.6-fold versuscontrol VSMCs, whereas 14-day exposure to hydrogen per-oxide (see Methods) further enhanced such ratio by 4- to5-fold in both cases (not shown).
Mechanisms whereby ROS enhance vascular/valve calci-fication are yet poorly understood. ROS underlie signalingand expression of BMP family proteins such as BMP22 andBMP4,24 as well as Cbfa-1 and alkaline phosphatase.2,11
Oscillatory shear stress (to which AV is exposed) inducesBMP4 via p47phox/Nox1-dependent NAD(P)H oxidase, fur-
Figure 5. A through C, Graphs depicting effects of antioxidants on AV calcification, analyzed by 3 different methods. A, AV echogenic-ity; B, histomorphometry; C, Ex vivo X-Ray attenuation at computed tomography of aortic specimens (see Methods). D, mRNA expres-sion of NAD(P)H oxidase isoforms Nox4 and Nox1, showing increased Nox4/Nox1 ratio, particularly with tempol. *P�0.05 vs Normal;#P�0.05 vs HC; ¶P�0.05 vs HC�Vit.D2�LA; †P�0.05 vs all other groups.
468 Arterioscler Thromb Vasc Biol March 2008
by guest on May 19, 2018
http://atvb.ahajournals.org/D
ownloaded from
ther inducing intercellular adhesion molecule-1 (ICAM-1)expression and monocyte adhesion.24 ROS may also triggerapoptosis, a potential seed for calcification.2,6,38 Conversely,calcium resorption is enhanced by osteoclast Nox2 or Nox4-derived ROS15,23 and inhibited by nitric oxide.16
Although antioxidant LA effects have been reported inseveral preparations25 and in humans,26 precise mechanismsof LA effects are unclear. LA reacts with peroxynitrite only atlow rates,27 and probably does not scavenge hydrogen per-oxide directly. Although LA might owe its antioxidantactivity to the reduced form dihydrolipoic acid, LA itself canalso scavenge species such as hypochlorous acid thanks to theunique reactivity of its 2,3-dithiolane ring.28 Moreover, LAshares with distinct antioxidants such as resveratrol, but not�-tocopherol or ascorbic acid, the capacity to induce hemeoxygenase-1 via Nrf-2.29
Tempol-mediated increase in calcification and hydrogenperoxide levels provide further evidence that ROS accelerateAV calcification in vivo. Central to this question is themechanism of tempol effects in our system. Similarly to otherreports,30 tempol decreased superoxide levels and nitroty-rosine staining, in line with its proposed mechanisms ofaction, namely SOD-mimetic activity30 and shift of nitroxi-dative stress toward nitric oxide.31 Tempol-induced hydrogenperoxide increase was previously reported in endothelial32 ortumor cells33 and increased at mmol/L tempol concentrationsor high glucose levels. Our EPR assessment of plasma tempol
concentration yielded values �5 �mol/L (supplemental Fig-ure X), well within or even below usual. Explanation forthose effects is yet unclear, but it is unlikely that hydrogenperoxide increase is the usual outcome of possible SOD-liketempol effect, which in itself is debatable.34 Mitochondrialfunction impairment attributable to tempol was reported.32 Inaddition, while reacting poorly with superoxide anionradical, tempol reactivity strongly enhances at lower pHsvia direct reaction with hydroperoxyl radical (OOH�), the pro-tonated uncharged form of superoxide.35 Thus, at least in our invivo model, an unusually strong SOD-like effect might haveoccurred because of decreased pH caused by the carbonicanhydrase activity of osteoclast, which promotes intra- andextracellular acidification.36 This is consistent with lack oftoxicity of high tempol concentrations in intact cells30 and withtempol-induced oxidant stress in the acidic renal medulla.37
Also, we observed higher cell loss around calcifying foci withtempol. Increased progression of valve calcification with tempolmay be clinically relevant, considering that this nitroxide anti-oxidant is already undergoing prospective clinical studies forseveral conditions.30
Collectively, our observations point to a role of redoxprocesses, particularly those resulting in hydrogen peroxideincrease, possibly related to NAD(P)H oxidase activity, in theprogression rate of AV calcification. These results further alink between pathogenesis of AV stenosis and atherosclero-sis, and pave the way to clinically effective interventions
Figure 6. In vitro VSMC calcification model. A, Effects of tempol(T) and LA on calcium deposition with 14-day calcifying stimuli; B,Effects of cell coincubation with exogenous hydrogen peroxide on calcium deposition (†P�0.05 between groups); C, Effects of catalasecoincubation on calcium deposition (†P�0.05 between groups); D, Superoxide production, assessed by 2-hydroxyethidium (EOH) analy-sis. For panels A and D: *P�0.05 vs Control; P�0.05 vs #LA or ¶Tempol-containing media.
Liberman et al ROS Contribute to Aortic Valve Calcification 469
by guest on May 19, 2018
http://atvb.ahajournals.org/D
ownloaded from
capable of slowing such progression, as suggested by protec-tive LA effects.
AcknowledgmentsWe thank Carlos Rochitte for aortic specimen computed tomogra-phy, Leonora Loppnow, Victor Debbas, and Ana Lucia Garippo fortechnical assistance, and Lea Demarchi and Solange Consorti fortissue processing.
Sources of FundingThis work was supported by Fundacao de Amparo a Pesquisa doEstado de Sao Paulo, Conselho Nacional de DesenvolvimentoCientıfico e Tecnologico (Redoxoma Millenium Institute), Financia-dora de Estudos e Projetos.
DisclosuresNone.
References1. Lindroos M, Kupari M, Heikkila J, Tilvis R. Prevalence of aortic valve
abnormalities in the elderly: an echocardiographic study of a randompopulation sample. J Am Coll Cardiol. 1993;21:1220–1225.
2. Johnson RC, Leopold JA, Loscalzo J. Vascular calcification: pathobio-logical mechanisms and clinical implications. Circ Res. 2006;99:1044–1059.
3. Weiss RM, Ohashi M, Miller JD, Young SG, Heistad DD. Calcific aorticvalve stenosis in old hypercholesterolemic mice. Circulation. 2006;114:2065–2069.
4. Rosenhek R, Rader F, Loho N, Gabriel H, Heger M, Klaar U,Schemper M, Binder T, Maurer G, Baumgartner H. Statins but notangiotensin-converting enzyme inhibitors delay progression of aorticstenosis. Circulation. 2004;110:1291–1295.
5. Cowell SJ, Newby DE, Prescott RJ, et al. A randomized trial of intensivelipid-lowering therapy in calcific aortic stenosis. N Engl J Med. 2005;352:2389–2397.
6. Abedin M, Tintut Y, Demer LL. Vascular calcification: mechanisms andclinical ramifications. Arterioscler Thromb Vasc Biol. 2004;24:1161–1170.
7. Chen K, Thomas SR, Keaney JF Jr. Beyond LDL oxidation: ROS invascular signal transduction. Free Radic Biol Med. 2003;35:117–132.
8. Miller FJ Jr, Gutterman DD, Rios CD, Heistad DD, Davidson BL.Superoxide production in vascular smooth muscle contributes to oxi-dative stress and impaired relaxation in atherosclerosis. Circ Res. 1998;82:1298–1305.
9. Massy ZA,Maziere C, Kamel CS, Brazier M, Choukroun G, TribouilloyC, Slama M, Andrejak M Maziere JC. Impact of inflammation andoxidative stress on vascular calcifications in chronic kidney disease.Pediatr Nephrol. 2005;20:380–382.
10. Parhami F, Morrow AD, Balucan J, Leitinger N, Watson AD, Tintut Y,Berliner JA, Demer LL. Lipid oxidation products have opposite effects oncalcifying vascular cell and bone cell differentiation. A possible expla-nation for the paradox of arterial calcification in osteoporotic patients.Arterioscler Thromb Vasc Biol. 1997;17:680–687.
11. Mody N, Parhami F, Sarafian TA, Demer LL. Oxidative stress modulatesosteoblastic differentiation of vascular and bone cells. Free Radic BiolMed. 2001;31:509–519.
12. Steitz SA, Speer MY, Curinga G, Yang HY, Haynes P, Aebersold R, SchinkeT, Karsenty G, Giachelli CM. Smooth muscle cell phenotypic transitionassociated with calcification: upregulation of Cbfa1 and downregulation ofsmooth muscle lineage markers. Circ Res. 2001;89:1147–1154.
13. Tintut Y, Alfonso Z, Saini T, Radcliff K, Watson K, Bostrom K, DemerLL. Multilineage potential of cells from the artery wall. Circulation.2003;108:2505–2510.
14. Watson KE, Bostrom K, Ravindranath R, Lam T, Norton B, Demer LL.TGF-b1 and 25-hydroxycholesterol stimulate osteoblast-like vascularcells to calcify. J Clin Invest. 1994;93:2106–2113.
15. Steinbeck MJ, Appel WH Jr, Verhoeven AJ, Karnovsky MJ. NADPH-oxidase expression and in situ production of superoxide by osteoclastsactively resorbing bone. J Cell Biol. 1994;126:765–772.
16. Mancini, L; Moradi-Bidhendi, N; Brandi, ML; Macintyre I. Nitric oxide,superoxide and peroxynitrite modulate osteoclast activity. BiochemBiophys Res Commun. 1998;243:785–790.
17. Drolet MC, Arsenault M, Couet J. Experimental aortic valve stenosis inrabbits. J Am Coll Cardiol. 2003;41:1211–1217.
18. Shioi A, Nishizawa Y, Jono S, Koyama H, Hosoi M, Morii H. Beta-glycerophosphate accelerates calcification in cultured bovine vascularsmooth muscle cells. Arterioscler Thromb Vasc Biol. 1995;15:2003–2009.
19. Janiszewski M, Lopes LR, Carmo AO, Pedro MA, Brandes RP, SantosCX, Laurindo FR. Regulation of NAD(P)H oxidase by associated proteindisulfide isomerase in vascular smooth muscle cells. J Biol Chem. 2005;280:40813–40819.
20. Fernandes DC, Wosniak J, Pescatore LA, Bertolini MA, Liberman M,Laurindo FR, Santos CX. Analysis of dihydroethidium-derived oxidationproducts by HPLC in the assessment of superoxide production andNADPH oxidase activity in vascular systems. Am J Physiol Cell Physiol.2007;292:C413–C422.
21. Li, X; Ying Yang, H; Giachelli CM. Role of the sodium-dependentphosphate cotransporter, Pit-1, in vascular smooth muscle cell calcifi-cation. Circ Res. 2006;98:905–912.
22. Lassegue B, Clempus RE. Vascular NAD(P)H oxidases: specific features,expression, and regulation. Am J Physiol Regul Integr Comp Physiol.2003;285:R277–R297.
23. Yang S, Madyastha P, Bingel S, Ries W, Key L. A new superoxide-generating oxidase in murine osteoclasts. J Biol Chem. 2001;276:5452–5458.
24. Sorescu GP, Song H, Tressel SL, Hwang J, Dikalov S, Smith DA, BoydNL, Platt MO, Lassegue B, Griendling KK, Jo H. Bone morphogenicprotein 4 produced in endothelial cells by oscillatory shear stress inducesmonocyte adhesion by stimulating reactive oxygen species productionfrom a nox1-based NADPH oxidase. Circ Res. 2004;95:773–779.
25. Packer L, Witf EH, Tritschler HJ. Alpha-lipoic acid as a biologicalantioxidant. Free Radic Biol Med. 1995;19:227–250.
26. Packer L, Kraemer K, Rimbach G. Molecular aspects of lipoic acid in theprevention of diabetes complications. Nutrition. 2001;17:888–895.
27. Trujillo M, Radi R. Peroxynitrite reaction with the reduced and theoxidized forms of lipoic acid: new insights into the reaction of per-oxynitrite with thiols. Arch Biochem Biophys. 2002;397:91–98.
28. Biewenca GP, Bast A. Reaction of lipoic acid with ebselen and hypo-chlorous acid. Methods Enzymol. 1995;251:303–315.
29. Ogborne RM, Rushworth SA, O’Connell MA. Alpha-lipoic acid-inducedheme oxygenase-1 expression is mediated by nuclear factor erythroid2-related factor 2 and p38 mitogen-activated protein kinase in humanmonocytic cells. Arterioscler Thromb Vasc Biol. 2005;25:2100–2105.
30. Soule BP, Hyodo F, Matsumoto K, Simone NL, Cook JA, Krishna MC,Mitchell JB. The chemistry and biology of nitroxide compounds. FreeRadic Biol Med. 2007;42:1632–1650.
31. Schnackenberg CG, Welch WJ, Wilcox CS. Normalization of bloodpressure and renal vascular resistance in SHR with a membrane-permeablesuperoxide dismutase mimetic: role of nitric oxide. Hypertension. 1998;32:59–64.
32. Alpert E, Altman H, Totary H, Gruzman A, Barnea D, Barash V, SassonS. 4-Hydroxy tempol-induced impairment of mitochondrial function andaugmentation of glucose transport in vascular endothelial and smoothmuscle cells. Biochem Pharmacol. 2004;67:1985–1995.
33. Gariboldi MB, Terni F, Ravizza R, Meschini S, Marra M, Condello M,Arancia G, Monti E. The nitroxide tempol modulates anthracyclineresistance in breast cancer cells. Free Radic Biol Med. 2006;40:1409–1418.
34. Goldstein S, Samuni A, Hideg K, Merenyi G. Structure-activity rela-tionship of cyclic nitroxides as SOD mimics and scavengers of nitrogendioxide and carbonate radicals. J Phys Chem A. 2006;110:3679–3685.
35. Krishna MC, Russo A, Mitchell JB, Goldstein S, Dafni H, Samuni A. Donitroxide antioxidants act as scavengers of O2
�. or as SOD mimics? J BiolChem. 1996;271:26026–26031.
36. Laitala T, Vaananen HK. Inhibition of bone resorption in vitro byantisense RNA and DNA molecules targeted against carbonic anhydraseII or two subunits of vacuolar H(�)-ATPase. J Clin Invest. 1994;93:2311–2318.
37. Chen YF, Cowley AW Jr, Zou AP. Increased H2O2 counteracts thevasodilator and natriuretic effects of superoxide dismutation by tempol inrenal medulla. Am J Physiol Regul Integr Comp Physiol. 2003;285:R827–R833.
38. Proudfoot D, Skepper JN, Hegyi L, Bennett MR, Shanahan CM,Weissberg PL. Apoptosis regulates human vascular calcification in vitro:evidence for initiation of vascular calcification by apoptotic bodies. CircRes. 2000;87:1055–1062.
470 Arterioscler Thromb Vasc Biol March 2008
by guest on May 19, 2018
http://atvb.ahajournals.org/D
ownloaded from
Wosniak, Jr, Pablo M.A. Pomerantzeff and Francisco R.M. LaurindoMarcel Liberman, Estêvão Bassi, Marina Kamla Martinatti, Fábio Cerqueira Lario, João
Aortic Valve CalcificationOxidant Generation Predominates Around Calcifying Foci and Enhances Progression of
Print ISSN: 1079-5642. Online ISSN: 1524-4636 Copyright © 2007 American Heart Association, Inc. All rights reserved.
Greenville Avenue, Dallas, TX 75231is published by the American Heart Association, 7272Arteriosclerosis, Thrombosis, and Vascular Biology
doi: 10.1161/ATVBAHA.107.1567452007;
2008;28:463-470; originally published online December 27,Arterioscler Thromb Vasc Biol.
http://atvb.ahajournals.org/content/28/3/463World Wide Web at:
The online version of this article, along with updated information and services, is located on the
http://atvb.ahajournals.org/content/suppl/2008/02/26/ATVBAHA.107.156745.DC1Data Supplement (unedited) at:
http://atvb.ahajournals.org//subscriptions/
at: is onlineArteriosclerosis, Thrombosis, and Vascular Biology Information about subscribing to Subscriptions:
http://www.lww.com/reprints
Information about reprints can be found online at: Reprints:
document. Question and AnswerPermissions and Rightspage under Services. Further information about this process is available in the
which permission is being requested is located, click Request Permissions in the middle column of the WebCopyright Clearance Center, not the Editorial Office. Once the online version of the published article for
can be obtained via RightsLink, a service of theArteriosclerosis, Thrombosis, and Vascular Biologyin Requests for permissions to reproduce figures, tables, or portions of articles originally publishedPermissions:
by guest on May 19, 2018
http://atvb.ahajournals.org/D
ownloaded from
Supplemental Material
Oxidant generation predominates around calcifying foci and
enhances progression of aortic valve calcification
Marcel Liberman, MD
Estêvão Bassi, BSc
Marina Kamla Martinatti, BSc
Fábio Cerqueira Lario, MD
João Wosniak Jr., PhD
Pablo M.A. Pomerantzeff, MD
Francisco R. M. Laurindo, MD, PhD
Supplemental material consists of:
� Expanded methods section
� References
� Table I
� Legends for supplemental figures
� Supplemental figures I to X
Methods (expanded version)
Reagents
All chemicals, including α-lipoic-acid, tempol, cholesterol, PEG-superoxide
dismutase (SOD), PEG-Catalase were from Sigma Chemical, except when specified; DCF-
DA and DHE were from Invitrogen; OCT Tissue Tek embedding compound was from
Fisher Scientific. For immunohistochemistry, the following commercial primary antibodies
and dilutions were used: RAM11 1:200 (Dako Cytomation), Nox2 1:200, p22phox 1:200,
osteopontin 1:200 and Cbfa-1 1:100 (Santa Cruz Biotechnology), nitrotyrosine 1:100
(Upstate Cell Signaling), PDI 1:32000 (Stressgen), Ki-67 1:200 (Novocastra LTD). Non-
commercial antibodies were: osteopontin 1:50 (Hybridoma Bank-University of Iowa) and
Nox2 clone 54.1 1:1600 (kind gift of Dr. J Burritt, University of Montana). Secondary
antibodies were anti-mouse or anti-goat from Vectastain Elite ABC System (Vectastain,
Vector Technologies, Burlingame, CA) for immunohistochemistry. For
immunofluorescense, secondary antibodies were anti-mouse or anti-goat (Calbiochem)
conjugated to rhodamine or FITC. For assessing cell loss in AV tissue samples, we used
“in-situ cell death detection kit” from Roche Applied Science and DNAse from Promega
Corporation. Culture medium DMEM, penicillin (100mg/mL), streptomyicin (100 mg/ml)
and trypsin were from Gibco BRL-Life Technologies (Grand Island); FBS was from
Cultilab, collagenase and elastase were from Worthington Biochemical Corporation. For
cell culture calcium measurement, we used “Calcium Reagent Set” from Teco Diagnostics.
Trizol, PCR, Reverse Transcriptase kits and Sybr Green were from Invitrogen Corporation.
Ethics
This study was approved by an internal scientific institutional Committe and by the
Ethics Committee from Hospital das Clínicas – University of São Paulo Medical School,
and complied with the Guide for the Care and Use of Laboratory Animals (NIH publication
85-23, rev. 1996).
Human aortic valves
Human stenotic AV were collected from patients with isolated tricuspid AV stenosis
(n=5) during surgincal replacement; control (n=6) and sclerotic (n=4) valves were acquired
from corpses undergoing authopsy (not after 6 hours after death). We considered control
valve a tricuspid and thin valve without evidence of subaortic or aortic arch disease from
patients <35 years old. Sclerotic AV were collected from individuals >65 years old,
revealing tricuspid valve thickening, initial calcifying foci, no evidence of rheumatic
disease, any other valve dysfunction and no evidence of aortic dissection or aneurism. All
sclerotic or stenotic AV were analyzed for ROS microtopography, histology and
immunohistochemistry for NADPH oxidase subunits, as described below.
Rabbit AV calcification model
Male NZW rabbits weighing 2500-3000g were housed in individual cages in a
controled room under 12h light-dark cycle at temperature (25oC). After initial quarantine,
rabbits were submitted to a 12-week protocol consisting of supplementation of diet with
cholesterol 0.5% + Vitamin D2 104 IU/day (herein refered to as HC+vitD rabbits, n=34). In
specific experiments, tempol 100 µmol/kg/day (n=15) or lipoic acid (LA) 120µmol/kg/day
(n=11) were administered continuously in drinking water. As controls, some rabbits
received cholesterol 0.5% only (HC, n=9) or regular chow (Normal, n=32). Tempol was
diluted in water and LA in ethanol. Eight normal controls received equivalent ethanol only
and did not differ from the other normal controls.
Biochemical blood analysis
Blood was collected from rabbits for analysis of total cholesterol, calcium,
phosphorus and creatinine, which were measured from plasma by standar methods at
baseline and after 3 months of intervention.
Tissue processing
Rabbit AV were collected after a lethal dose of penthobarbital sodium, rapidly
dissected, fixed in 4% PBS-formaldehyde for 24 hours, and included in paraffin for
morphology and immunohistochemistry, whereas another section was freshly frozen in
OCT compound for in situ ROS detection. Histology was analyzed after hematoxylin and
eosin or Verhoeff-Van Gieson stainings. Histomorphometry was performed using a Leica
Quantimet software using specific stainings for collagen (Masson’s Trichrome), calcium
(Von Kossa) and immunohistochemistry for macrophage infiltration (RAM11 antibody).
Immunohistochemistry and immunofluorescence
Briefly, AV were were cut in 3µm-thick sections in silanized slides for
immunohistochemical staining. Sections were deparaffinized, tripsinized for antigen
recovery and primary antibodies were applied for 12hs. Then, Vectastain Elite ABC System
was used for development, with secondary anti-mouse or anti-goat antibodies conjugated to
streptavidin–biotin complex. Final staining was performed through reaction with 3,3’-
diaminobenzidine (Sigma Chemical Co., St. Louis, MO) and counterstained with Harris
hematoxylin. For immunofluorescence, anti-mouse or anti-goat secondary antibodies
conjugated to rhodamine or FITC were applied. Positive and negative (fetal bovine serum
without primary Ab) controls were used in all cases and stained simultaneously. For total
cell counting employing nuclei staining, DAPI (4’-6-diamidino-2-phenylindole) was used
as a fluorescent dye using excitation-emission wavelenghs of 365nm-395nm.
TUNEL staining
An estimate of apoptotic cells was performed by positive TUNEL staining using
“in-situ cell death detection kit” from Roche, as detailed elsewhere1. Some sections were
pre-incubated with DNAse as positive controls.
TRAP
Osteoclasts were identified by histochemical staining for TRAP2. Sections (3-µm-
thick) were deparaffinized, washed in deionized water for 5 min. and preincubated in 0.2M
acetate buffer (sodium acetate 0.2 M and L(+) tartaric acid 50mM), pH 5.0, for 20 min..
Following 5-min. wash in water, the sections were incubated in the same buffer with added
naphthol AS-MX phosphate (0.5 mg/ml) and Fast Red TR (1.1 mg/ml) for up to 6 hours at
37oC, monitoring until osteoclasts were turn bright red. After rinsing in distilled water, the
sections were counterstained in methyl green and mounted in Keiser’s solution. For
positive controls we used human regenerating bone fracture sections.
In situ aortic valve ROS detection
Dihydroethidium (DHE), a fluorescent dye freely permeable to cells, is trapped to
DNA when oxidized to 2-hydroxyoxyethidium, mainly by superoxide, and ethidium, by
other ROS, heme or peroxidases (ref. 20). Therefore, this probe allows identification of
individual cells producing excess superoxide, although total fluorescence is not specific for
the latter. Microtopography of superoxide is performed at confocal microscopy, with
excitation at 488 nm and emission at 610 nm, detected with a 560nm longpass filter, as
previously described (ref. 8). Longitudinal sections of human and rabbit AV were cut in the
cryostat (30 µm thickness) and placed in glass slides. Samples were incubated with DHE
3µM, coverslipped and mantained in a light-protected humidified chamber at 37°C for 30
minutes. Images were obtained with a Zeiss LSM510 confocal laser scanning microscope
and Axiovision software (Carl Zeiss MicroImaging GmbH, Germany). Simultaneous
paralell acquisition of images was performed always for normal controls and other groups,
using identical laser and software settings. To verify if fluorescense signal was mainly
superoxide-derived, subsequent images was acquired after incubating the slide with 500
U/ml Peg-SOD for 30 min. We adapted this method to perform a novel analogous in situ
detection of hydrogen peroxide using the fluorescent dye 2’,3’-dichlorofluorescein
diacetate (DCF-DA). In this case, since the probe is mainly cytosolic and products do not
bind to DNA, a diffuse staining is verified. The employed protocol was similar to that
described abovesimilar protocol, with a final DCF-DA concentration of 3µM. Controls,
performed by incubating slides for 30 min with Peg-catalase (400U/ml), indicated
preferential detection of hydrogen peroxide.
Morphometric image analysis
For quantification of DHE and DCF fluorescense, TUNEL positive cells, number of
nuclei (DAPI) and 3-nitrotyrosine staining area, we used Image J (version 1.37v; National
Institutes of Health, Bethesda, MD) software (ref. 8).
Echocardiography
Transthoracic echocardiography was performed at baseline and at 12 weeks of
follow-up. Animals were sedated with intramuscular injection of ketamine (30 mg/kg) and
xylazine (3.5 mg/kg). The anterior thorax was shaved for better echo imaging. Images were
acquired with a 12 MHZ phased-array probe connected to a Sonos 5500 imaging system
(Philips Medical Imaging, Andover, Massachusetts). Parasternal long and short axis views
were obtained by applying the transducer to right paraesternal location, and apical five-
chamber view was obtained by applying the transducer to subcostal location. A
comprehensive echocardiographic study was performed, including two-dimensional
imaging, Doppler imaging and M-mode imaging. Images were saved in digital format
(magneto-optical discs) for further blinded fashion one-operator analysis, as still frames
(spectral Doppler and M-mode imaging) or loops (two-dimensional imaging). To minimize
the inter and intra-assay variability, all imaging and analysis were carried out by the same
investigator.
A parasternal short axis view M-mode imaging at mid left ventricular (LV) level
was used to measure the following parameters (average of four cardiac cycles): LV end-
diastolic (LVIDd) and end-systolic (LVIDs) dimensions and ventricular septum (VST) and
posterior wall thickness (PWT). LV mass was obtained by the Devereux formula3. Relative
wall thickness (RWT) was calculated as RWT=2xPWT / LVIDd. Fractional shortening
(FS=[LVIDd-LVIDs]/LVIDd) and LV ejection fraction by the Teichholz4 method were
also calculated.
A parasternal long axis view was used to measure LV outflow tract mid-systolic
dimension (LVOTd). Analysis of the aortic valve anatomy and motility was performed
using paraesternal long axis, short axis and the apical five chambers. The latter view was
used to estimate the peak and mean aortic valve flow velocities, determined by continuous
wave Doppler imaging, averaging the values of four to five beats. The maximal
instantaneous gradient and the mean gradient across the aortic valve were derived from
aortic Doppler velocities by the modified Bernoulli equation. Left ventricular outflow tract
(TVILVOT) was estimated by planimetry, using pulsed wave Doppler, with the volume
sample (adjusted to 0.6 mm) positioned proximal to the aortic valve, as performed in
humans5.
Aortic valve area was calculated by continuity equation5-7
:
π x (LVOTd)2 x TVILVOT
AVA =
4 x TVIAV
Analysis of aortic valve calcification
Echogenicity analysis
A semi-quantitative index (range 0-3) based on opacity and valve motility was used
for estimating AV echogenicity, an index of valve thickening and calcification, as follows:
0-absent, 1-mild; 2-moderate, with preserved motility/opening; 3-severe, with decreased
motility/opening. Data were analyzed on a blinded fashion.
AV calcification area
We measured the percentage of cuspid mineralization (total valve calcification/total
valve area) as the mean of 2 tissue sections stained with Von Kossa staining8 and
hematoxylin-eosin. Data were analyzed with Leica Quantimet software.
Multidetector-row Computed Tomography (MDCT)
Ex-vivo rabbit descending aorta segments were simultaneously scanned in a
Toshiba Aquilion 64 multislice CT (Toshiba Otawara, Japan) and x-Ray attenuation in
Hounsfield units was calculated using mean of 3 measurements for each segment. All
measurements were performed in a Vitrea 2 Workstation (Vital Images Inc, Minnetonka,
MN) by an investigator blinded to the experimental groups.
Real-Time Quantitative Reverse Transcriptase-PCR
Total RNA was purified from descending rabbit aorta fragments, using Trizol. After
checking RNA integrity, it was reverse-transcribed with Superscript II enzyme (Invitrogen)
with random primers. Nox1 and Nox 4 quantification was performed by real-time RT-PCR.
Primer sequences were designed from rabbit sequences for Nox1 and Nox4, kindly
provided by Dr. Bernard Lassegue (Emory University, Atlanta): Nox1- Forward:
CATCATGGAAGGAAGGAGA, Reverse: GCTTCCGGATAAACTCCACA; Nox4-
Forward: CCACAGACTTGGCTTTGGAT, Reverse: TACTGGCCAGGTCTTGCTTT.
cDNA amplification was performed in a Corbett Rotor-Gene RG6000 thermocycler
(Corbett Res. Pty Lim., Sidney, Australia), using SYBR green dye. Copy numbers were
calculated from standard curves generated from Nox1 and Nox4 plasmid templates. Similar
procedures were employed for analysis of osteopontin mRNA, using the following primer
sequences: Forward: CTCCCGGTTAAACACGCTGATTC, Reverse:
AGGATACTGGGCATTAGAGACG. GAPDH amplification was used for normalization
in this case (Forward: TCACCATCTTCCAGGAGCGA, Reverse:
CACAATGCCGAAGTGGTCGT).
EPR analysis of plama tempol concentration
Plasma EPR spectra for tempol detection were recorded at room temperature on a
Bruker ER 200D-SRC upgraded to an EMX instrument, with the same parameters for a
tempol standard solution and plasma samples.
In vitro VSMC calcification model
Aortas from intact rabbits were dissected and cleansed from adventitial tissue.
Primary culture of vascular smooth muscle cells was established after digestion with
collagenase IV (1 mg/mL), elastase (0.125 mg/mL), trypsin inhibitor (0.375 mg/mL) in
DMEM at 37ºC for 4 hours. Cells were seeded and grown in DMEM low glucose, with
10% fetal bovine serum (FBS) and essential aminoacids in an incubator with 5% CO2 at
37ºC. Cells were used at the 4th
passage. Calcification protocol began with cell confluence
of 70%, the culture medium was supplemented with calcium chloride 6mM (CaCl2), CaCl2
6mM+β-glycerophosphate 10mM or both in the presence of Vitamin D3 0.1µM (CaCl2+β-
GP+Vit.D3). The latter combination was supplied in the absence or presence of either
tempol 100µM or LA 100µM. Exogenous hydrogen peroxide (50µM) or catalase 500U/ml
was added in specific experiments. The use of unmodified catalase was based on the fact
that this enzyme is known to have access into smooth muscle cells after a 3-4 hour period9,
which is adequate for the present experimental conditions. All incubations lasted for a total
of 14 days, changing medium every 3 days.
In vitro calcium measurement
After 14 days of calcifying estimuli, cells were washed with PBS and incubated
with HCl 0.6N 1ml/plate for 12hs at 4oC. The supernatant was centrifuged and total
calcium was measured by colorimetric assay through reaction with o-cresolphtalein and
reading in a spectrophotometer absorbing at 570nm after comparing with standard curve
(Teco calcium measurement Kit) and normalized by total protein (Bradford method).
NAD(P)H oxidase activity in membrane fraction
VSMC membrane homogenates were obtained by sequential centrifugation as
described previously (ref. 20). Protein concentration of total homogenate and membrane
fraction was quantified by Bradford method. Membrane fraction (30µl/30µg protein) was
diluted in 815µL PBS/ EDTA 10 µM solution and NADPH 300 µM was added at 37oC.
After baseline counting, lucigenin 5 µM was added into the scintillation tube and
consecutive luminescence countings were registered in a Berthold 9505 luminometer
(EG&Instruments GmbH, Germany). Baseline signal was subtracted from blank (PBS/
EDTA 10 µM) and results normalized by protein concentration and expressed in counts per
minute/mg.
DHE oxidation products by HPLC analysis
Cells with 80% confluence were rinsed twice with PBS/DTPA 0.1mM. All further
procedures were performed in the dark or dim light. Cells were incubated with DHE 100
µM in PBS/DTPA 0.1 mM for 30 min at 37ºC, washed with cold PBS/DTPA 0.1 mM,
extracted with acetonitrile (300 µL/well), scraped, sonicated (3 cycles at 4W for 10s) and
centrifuged (12000g at 4ºC at 10 min). The supernatant was dried in vacuum (Speed
VacPlus SC-110A, Thermo Savant) and kept at -20ºC. Samples were resuspended in 120
µL PBS/DTPA 0.1mM and injected (100 µL) in an HPLC system (Waters Corporation), as
previously published by our laboratory (Fernandes et al, ref. 20). We used NovaPak C18
column (3.9x150 mm, with 5 µm particles), solution A (100% acetonitrile) and solution B
(water; acetonitrile 10%; trichloroacetic acid 0.1%) as a mobile phase with a continous flow
of 0.4 mL/min. The procedure for analysis was initiated, loading solution A 0%, increasing
progressively to 40% during the first 10 min, keeping this proportion for aditional 10 min,
followed by injection of 100% of A for the next 5 min and 0% of A for the last 10 min.
Simultaneous analysis of the remaining DHE and its derived products 2-hydroxyethidium
and ethidium was performed, using respectively UV 245nm (photodiode array W2996
detector) and fluorescence (excitation lengh 510nm and emission 595nm, with fluorescence
W2475 detector). Quantification was performed by comparing the area under the curve of
the peak values obtained for the sample and a standard solution under identical
chromatographic conditions. Results were expressed as the ratio between 2-
hydroxyethidium or ethidium generated per DHE consumed (concentration difference
between initial and remaining DHE).
Statistics
All data are reported as mean±SE. Comparisons were tested through one-way-
ANOVA plus Student-Newman-Keuls (among all groups) or Dunnet (all groups vs.
control) post-hoc tests, at a 0.05 significance level.
References for supplement
1- Wijsman JH, Jonker RR, Keijzer R, van de Velde CJ, Cornelisse CJ, van Dierendonck JH. A new
method to detect apoptosis in paraffin sections: in situ end-labeling of fragmented DNA. J
Histochem Cytochem. 1993;41:7-12.
2- Erlebacher A, Derynck R. Increased expression of TGF-beta 2 in osteoblasts results in an
osteoporosis-like phenotype. J Cell Biol. 1996;132:195-210.
3- Devereux, R. Echocardiographic determination of left ventricular mass in man. Anatomic
validation of the method. Circulation. 1977; 55:613-618.
4- Teichholz, L. Study of left ventricular geometry and function by B-scan ultrasonography in
patients with and without asynergy. New Engl J Med. 1974;291:1220-1226.
5- Quiñones MA, Otto CM, Stoddard M, Waggoner A, Zoghbi W. Recommendations for
Quantification of doppler Echocardiography: A Report From the Doppler Quantification Task Force
of the Nomenclature an Standards Committee of the American Society of Echocardiography. J Am
Soc Echocardiogr. 2002;15:167-184.
6- Otto CM, Pearlman AS: Doppler echocardiography in adults with symptomatic aortic stenosis.
Diagnostic utility and cost-effectiveness. Arch Intern Med .1988;148:2553-2560.
7- Garcia D, Dumensnil JG, Durand LG. Discrepancies between catheter and Dopppler estimates of
valve effective orifice area can be predicted from the pressure recovery phenomenon: practical
implications with regard to quantification of aortic stenosis severity. J Am Coll Cardiol.
2003;41:435-442.
8- Sheehan D, Hrapchak B. 1980. Theory and practice of Histotechnology, 1980, 2nd edition.
Batelle Press, Columbus, OH. pp 226-7.
9- Sundaresan M, Yu ZX, Ferrans VJ, Irani K, Finkel T. Requirement for generation of H2O2 for
platelet-derived growth factor signal transduction. Science. 1995;270:296-299.
Supplemental Table
Table I- Total serum cholesterol, calcium, phosphorus, CaxP product and creatinine at baseline (A)
and after different interventions (B). *p<0.05 vs Normal; †p<0.05 vs HC e
‡p<0.05 vs
HCD+Vit.D2+tempol.
Baseline
Total
cholesterol
(mg/dl)
Total Ca++
(mg/dl) P (mg/dl)
CaxP product
(mg2/dl2)
Creatinine
(mg/dl)
Normal 53±5 15.7±0.1 6.2±0.5 97.1±7.5 1.1±0.04
HC 37±4 16.6±0.5 5.9±0.3 97.5±7.5 1.1±0.1
HC+Vit.D2 57±4 15.9±0.2 6.9±0.5 109.7±7.0 1.1±0.04
HC+Vit.D2+T 36±5 15.6±0.2 6.6±0.4 102±5.9 1.2±0.1
HC+Vit.D2+LA 40±6 16.6±0.3 6.5±0.5 108.1±8.6 1.3±0.1
After 3 months
Total
cholesterol
(mg/dl)
Total Ca++
(mg/dl) P (mg/dl)
CaxP product
(mg2/dl2)
Creatinine
(mg/dl)
Normal 46±6 13.7±0.3 4.6±0.2 63.1±3.7 1.2±0.1
HC 1220±113* 13.8±0.4 4.5±0.3 62.2±4.4 1.4±0.1
HC+Vit.D2 1363±90*‡
15.8±0.4*†
8.0±0.4*†
128.4±7.9*†
1.6±0.1*
HC+Vit.D2+T 971±96* 15±0.4 8.2±0.3*†
122±5.8*†
1.7±0.1*
HC+Vit.D2+LA 1084±114* 14.1±0.4 7.8±0.5*†
110.2±8.4*†
1.6±0.2*
A
B
Supplemental Figure Legends
Figure I
Increased NAD(P)H oxidase expression around calcifying foci (Ca++
) in stenotic
human aortic valves, as shown by Nox2 immunostaining with an antibody distinct from
Fig.4 (clone 59.1 Ab).
Figure II
Histological findings of rabbit AV, showing Normal rabbits (upper panels) and
HC+VitD rabbits (lower panels). The latter exhibits findings analogous to human aortic
stenosis: thickening, macrophage infiltration in the aortic face (asterisk mark), elastolysis
and calcification (arrows). Stainings are: hematoxylin and eosin (left panels) and Verhoeff-
Van Gieson, which also stains strongly for calcium deposits, in addition to elastic fibers
(middle panels). The right panels depict peroxidase-immunostaining for rabbit
macrophages (RAM11 Ab, 1:200 dilution).
Figure III
Echocardiographic images of Normal (left panels) and HC+Vit.D2 aortic valves
(right panels), in diastole (above) and systole (below), depicting increased echogenicity in
AV from HC+VitD rabbits, observed after 12 weeks of intervention.
Figure IV
Echocardiographic variables in the rabbit model. Panel A- absolute AV area after 3
months of intervention, showing little difference among groups; Panel B- increased loss of
in final AV area vs. baseline AV area in all HC+vitD rabbits, which was statistically
significant only with tempol (*p<0.05 vs Normal; #p<0.05 vs HC). HC controls showed no
change vs. normal. Panel C - aortic jet/left ventricular outflow tract velocity ratio, which
was unchanged among groups. Panel D - posterior wall thickness, which was increased in
all HC+vitD rabbits irrespective of antioxidants, whereas HC controls showed minor non-
significant increase.
Figure V
Effects of tempol or LA treatment on 3-nitrotyrosine immunostaining around AV
calcifying foci (Ca++
). (A): Normal control; (B): HC+vitD; (C): HC+vitD +tempol; (D):
HC+vitD +LA. Graph representing staining area (pixel2). Letters below each bar
correspond to the above panels. Each bar represents mean±SE of at least 3 different rabbits.
*p<0.05 vs Normal; #p<0.05 between specified groups.
Figure VI
Confocal microscopy colocalization of p22phox and nox2 (superior panel) and PDI
and p22phox (inferior panel) around calcifying foci in tissue slices from AV cusp of
HC+VitD rabbit. Anti-mouse or anti-goat secondary antibodies were used, conjugated to
rhodamine (red) or FITC (green) (Ca++
= calcifying foci). Dilution of primary antibodies
was 1:200 (p22phox, goat), 1:1600 (Nox2, mouse) and 1:32000 (PDI, mouse).
Figure VII
Phenotypic markers of cells around calcifying foci in HC+VitD rabbits (right
column), compared to normal controls (left column). Panels A,D: TRAP histochemistry
showing osteoclast (arrow); Panels B,E: Cbfa-1 immunostaining (Ab dilution 1:100);
Panels C,F: osteopontin immunostaing (Ab dilution 1:50). Inset: osteopontin RT-PCR,
with lane 1=Normal; 2=HC+VitD rabbit; 3=HC+VitD +tempol; 4=HC+VitD +LA.
Figure VIII
Graphs depicting results of histomorphometric analysis in AV from Normal or
cholesterol-only controls (HC) and HC+vitD rabbits treated or not with tempol or LA.
Panel A: The bar graph represents the number of TUNEL-positive cells per field (AU).
Increased staining in AV cusps from HC+VitD rabbits vs. Normal was observed, which
was enhanced in tempol-treated rabbits. Panel B: The bar graph summarizes increased cell
proliferation around calcifying foci in HC+VitD rabbits and its partial inhibition with
tempol. Cholesterol-only (HC) control rabbits showed minor staining coinciding with
macrophage infiltration area (Figure not shown). Panel C: Collagen accumulation,
expressed as %area at Masson’s trichrome staining. Panel D: Quantification of RAM11
stained area depicting macrophage accumulation in AV. Macrophage infiltration was
unchanged with antioxidants vs. HC+vitD rabbits and was particularly prominent in HC
controls. Each bar represents mean±SEM of at least 3 different rabbits. *p<0.05 vs Normal;
#p<0.05 between specified groups; †p<0.05 vs HC+Vit.D2, HC+Vit.D2+T and
HC+Vit.D2+LA.
Figure IX
Measurement of NADPH oxidase activity in membrane fraction of cultured vascular
smooth muscle cell incubated in vitro with calcifying stimuli for 14 days. Activity was
assessed by lucigenin (5µM) chemiluminescence, as described in supplemental methods.
Figure X
EPR detection of tempol in plasma, yielding average concentration ~5µM after
100µmol/kg/day administration in drinking water. Blood was collected after 3 months of
intervention, before killing the rabbit. Obtained characteristic spectrum was similar to
exogenous standard, and absent in Normal rabbits, which exhibited only the ascorbyl
radical (detail in upper right panel). Identical height of spectrum was obtained after
incubation of plasma samples with ferrycyanide (not shown), indicating that tempol was
not reduced by plasma ascorbate.
Suppl. Figure I
Aortic valve stenosis (negative control) Control valve
Aortic valve stenosis
Aortic valve stenosis (400x)
Ca++
Ca++
Ca++
Suppl. Figure II
Verhoeff-Van Gieson
No
rmal
IH: RAM11 H&E H
C+
Vit
.D2
Rabbit Model Aortic Valves
*
Suppl. Figure III
Normal HC+Vit.D2
Dia
sto
le
Systo
le
Suppl. Figure IV
*p<0.05 vs Normal; #p<0.05 vs HC
A
0.0
0.1
0.2
0.3
Ao
rtic
Va
lve
are
a (
cm
2)
Normal HC HC+Vit.D2 HC+Vit.D2+tempol HC+Vit.D2+LA
0.0
0.2
0.4
0.6
0.8
1.0
Ao
jet/L
VO
T v
elo
city
ra
tio
C
0.0
0.1
0.2
0.3
0.4
* **
Rela
tive P
W th
ickne
ss
D
-0.04
-0.03
-0.02
-0.01
0.00
0.01
0.02
0.03
0.04
* #
Delta
(AV
Afinal -
AV
Ab
ase
lin
e)
cm2
B
Suppl. Figure V
Ca++
D
Ca++
C
A B
Ca++
0
10000
20000
30000
*
*
*
Sta
inin
g a
rea (
pix
el2
)
#
A B C D
50 µm 50 µm
50 µm 50 µm
Suppl. Figure VI
p22phox Nox2 Merge
PDI p22phox Merge
Ca++
Ca
++ Ca
++
Ca++
Ca++
Ca++
Ca++
Ca++
Ca++
Suppl. Figure VII
TR
AP
C
bfa
-1
OP
N
F
20 µm
A
100 µm
D
20 µm
B
50 µm
C
1 4 100 µm
50 µm
E
Suppl. Figure VIII
0
2
4
6
*
*
TU
NE
L +
cells
(A
U)
# A
Normal HC HC+Vit.D2 HC+Vit.D2+tempol HC+Vit.D2+LA
0
5
10
15
20
25
Ki-6
7 +
ce
lls
*
*
# # B
0
5
10
15
20
Colla
ge
n A
rea (
%)
C
0
10
20
30
*
** *
RA
M-1
1 s
tain
ing (%
)
D
†
Suppl. Figure IX
0
10
20
30
40
50
U/m
g p
rote
in * #
#
* ¶ #
Control CaCl2 CaCl2+βGP+Vit.D3 CaCl2+βGP+Vit.D3+T CaCl2+βGP+Vit.D3+LA
Suppl. Figure X
Ascorbil radical
Plasma (Normal) Normal 1 G
10 G Standard (Tempol 36µM)
Plasma (HC+Vit.D2+T)