retinoic acid attenuates inducible nitric oxide synthase (nos2) activation in cultured rat cardiac...

J Mol Cell Cardiol 33, 933–945 (2001)

doi:10.1006/jmcc.2001.1356, available online at http://www.idealibrary.com on

Retinoic Acid Attenuates Inducible NitricOxide Synthase (NOS2) Activation inCultured Rat Cardiac Myocytes andMicrovascular Endothelial CellsSandrine Grosjean1,2, Yvan Devaux3, Carole Seguin3, Claude Meistelman1,Faiez Zannad3, Paul-Michel Mertes4, Ralph A. Kelly2 andDan Ungureanu-Longrois1

1Department of Anesthesia and Intensive Care, C.H.U. Brabois, Vandœuvre-les-Nancy, France,2Brigham and Women’s Hospital, Cardiovascular Division, Harvard Medical School, Boston MA, USA,3UPRES-EA 971068, Universite Henri Poincare Nancy I, City, France, 4Laboratory of Physiology,University of Reims-Champagne-Ardennes, Reims, France

(Received 24 October 2000, accepted in revised form 1 February 2001, published electronically 19 March 2001)

S. G, Y. D, C. S, C. M, F. Z, P.-M. M, R. A. K D. U--L. Retinoic Acid Attenuates Inducible Nitric Oxide Synthase (NOS2) Activation in Cultured RatCardiac Myocytes and Microvascular Endothelial Cells. Journal of Molecular and Cellular Cardiology (2001) 33,933–945. The inducible NO synthase (NOS2) in cardiac tissue contributes to myocardial and coronary in-flammation and dysfunction. Several natural (endogenous) hormones such as retinoic acid, the active metaboliteof vitamin A, have the ability to attenuate NOS2 activation in inflammatory cells. The aim of this study was toinvestigate the effect of RA on NOS2 activation in cultured cardiac microvascular endothelial cells (CMEC)and adult rat ventricular myocytes (ARVM). CMEC were stimulated either with a combination of 10 �g/mllipopolysaccharide (LPS) and 50 IU/ml interferon-� (IFN-�) or with a combination of 1 ng/ml interleukin-1� (IL-1� )+IFN-� whereas ARVM were stimulated with 1 ng/ml IL-1� and 50 IU/ml IFN-� in the absence or presenceof all-trans retinoic acid (atRA). Activation of the NOS2 pathway was estimated by measurement of mRNA(Northern blot) and protein (Western blot) expression, enzyme activity by conversion of [3H] -arginine to [3H]-citrulline, and nitrite accumulation. NOS2 mRNA half-life was studied in CMEC and ARVM in the presence ofactinomycin D. In CMEC and ARVM stimulated with a combination of LPS and/or cytokines, atRA (10−6, 10−5 )significantly (P<0.05) attenuated NOS2 mRNA and protein expression, enzymatic activity and reduced supernatantnitrite concentration. Upon stimulation with LPS/IFN-�, atRA significantly decreased NOS2 mRNA half-life. Thiswas not seen after stimulation with IL-1�/IFN-�. These results document for the first time an effect of RA onNOS2 activation in cardiac cells. They may contribute to the characterization of the immunomodulatory effectsof retinoids in myocardial and coronary inflammatory disorders. 2001 Academic Press

K W: Nitric oxide synthase; Lipopolysaccharide; Interferon type II; Retinoic acid; Macrophages.

product nitric oxide (NO), a major component ofIntroductionthe innate immune system, have been implicatedin the myocardial dysfunction of septic shock,1 inThe inducible nitric oxide synthase (NOS2) and its

Please address all correspondence to: D. Ungureanu-Longrois, M.D., Ph.D., Department of Anesthesia and Intensive Care, C.H.U.Brabois, Rue du Morvan, 54511 Vandoeuvre Cedex, France. Fax: 33 3 83 15 36 88. E-mail: [email protected] by the Department of Anesthesia and Intensive Care and UPRES-EA 971068, C.H.U. Brabois, Rue du Morvan, 54511Vandoeuvre Cedex, France.

0022–2828/01/050933+13 $35.00/0 2001 Academic Press

S. Grosjean et al.934

the myocardial inflammatory disorders of tryp- NOS2 expression in CMEC and ARVM. Interestingly,our results suggest that the effect of atRA dependsanosomiasis,2 in cardiac failure,3 and accelerated

graft arteriosclerosis.4,5 Both beneficial6–8 and detri- on the pro-inflammatory stimuli used to activatethe NOS2 pathway.mental2,9,10 effects of NOS2 pathway activation in

the myocardium have been reported. The intensityof NOS2 activation, the type of stimulus used toactivate NOS211 as well as the redox-dependent Materials and Methodsbiochemical NOx species12 modulate the biologicaleffects of NOS2 activation. Animals

The intensity of NOS2 activation is the resultof a complex interaction between activators and The present study was performed according to in-inhibitors. Activators of the NOS2 pathway have stitutional guidelines for animal use and care andbeen thoroughly characterized.3,13 Several studies to the recommendations of the Ministere Francaishave demonstrated that natural inhibitors of the des Affaires Sociales et de la Solidarite NationaleNOS2 pathway activation exist and include, but concerning animal studies.are not limited to, glucocorticoids,14 TGF-�1,15 inter-leukin-416 and interleukin-10.17

The hormonal and/or nutritional status of thePreparation of the reagentshost before a pro-inflammatory stimulus or chal-

lenge with an infectious agent could also modulateSalmonella typhimurium or Escherichia coli serotypeNOS2 activation. Carotenoids, vitamin A and itsB12 lipopolysaccharide (LPS) (Sigma, Saint Louis,metabolites such as retinoic acid are considered toMO, USA) was dissolved in phosphate buffered sa-be both hormones and micronutriments,18 and haveline (PBS) nominally calcium and magnesium freereceived much attention because they are known(Eurobio, Les Ulis, France) at a concentration ofto modulate the host immune response and in-10 �g/�l and was stored at −20°C. Murine re-flammatory reaction through yet poorly char-combinant IFN-� (Gibco BRL, Bethesda, MD, USA)acterized mechanisms. Interestingly, vitamin Awas dissolved in 0.1% (w/v) bovine serum albuminstatus has been shown to modulate morbidity andin PBS solution at a concentration of 100 IU/�lmortality in experimental models19,20 and humanand stored at −70°C. All-trans retinoic acid wasinfectious diseases.21 Recent trials using carotenoidsdissolved in DMSO, stored at −70°C and handledcould not demonstrate a beneficial effect of long-sheltered from light. All other reagents were fromterm carotenoids supplementation in patients withSigma unless otherwise stated and were of thecoronary artery disease.22 For all these studies,highest quality grade. All reagents were nominallythe mechanisms that could underlie the biologicalLPS-free.effects of vitamin A or its metabolites are poorly

documented.It has been shown recently that retinoic acid,

the active metabolite of vitamin A, attenuates NOS2 Isolation and culture of cardiac microvascularendothelial cells and myocytesactivation in several cell types in vitro. Most studies

have concerned macrophages23 but some have in-Cardiac microvascular endothelial cells (CMEC) andcluded keratinocytes.24 Only two studies have con-

cerned vascular smooth muscle cells25,26 and there adult rat cardiac myocytes (ARVM) were isolated aspreviously described.27–29 Briefly, adult male Wistarare no published studies on the effect of retinoic

acid on NOS2 activation in cardiac tissue. rats (Charles River, Wilmington, MA, USA) wereasphyxiated in carbon dioxide, the heart wasThe aims of the present study were: (i) to in-

vestigate the effects of all-trans retinoic acid (atRA) quickly removed and perfused in a retrograde man-ner on a Langendorff apparatus with DMEM (Gibcoon cultured cardiac microvascular endothelial cells

and myocytes and compare them with the effects BRL) to remove blood cells. Left ventricles wereimmersed in 70% ethanol for 30 s to devitalizein macrophages, where the effects of atRA on NOS2

pathway activation have already been dem- epicardial mesothelial cells and endocardial endo-thelial cells. For CMEC, after removing one quarteronstrated; (ii) attempt to document whether re-

tinoids modulate NOS2 activation in a cell- and/or of the left ventricular free wall and septum, thehearts were minced finely and incubated in 0.2%stimulus-specific manner.

We report here that co-administration of atRA (w/v) collagenase (Worthington Biomedical, Lake-wood, NJ, USA) for 20 min at 37°C. Trypsin (Gibcoand different pro-inflammatory stimuli attenuates

Retinoic Acid Attenuates Inducible NO Synthase Activation in Cardiac Tissue In Vitro 935

BRL) was subsequently added (final concentration ml IFN-� (Gibco BRL), in the presence or absenceof atRA at a concentration of 10−6 or 10−5 .0.3 mg/ml) and incubated for another 20 min. After

centrifugation of the dissociated cells at 100×g Similarly, adult rat ventricular myocytes weretreated for 24 h with 1 ng/ml interleukin-1� (IL-for 5 min, the cells were resuspended in DMEM

supplemented with 20% (v/v) fetal calf serum (FCS) 1�; Calbiochem, La Jolla, CA, USA), and 50 IU/mlIFN-�, in the absence or presence of atRA, 10−6 orand penicillin/streptomycin (Gibco BRL). After 2 h,

attached cells were washed with DMEM and cul- 10−5 .Rat alveolar macrophages were treated for 16 htured in DMEM with 20% FCS (v/v) in 5% CO2/

95% O2 at 37°C. with a combination of 10 �g/ml LPS and 50 UI/mlIFN-� in the absence or presence of atRA 10−5 .Cardiac microvascular endothelial cells were

grown to confluence in DMEM and FCS 20% (GibcoBRL) and serum-starved for 12 h before ex-perimentation. All experiments for CMEC were con- Analysis of NOS2 mRNA expression by Northern blotducted in DMEM without phenol red enriched withinsulin, transferrin and sodium selenite as pre- Total RNA was extracted using a commercial so-

lution according to the manufacturer’s instructionsviously described.11

For ARVM, hearts were washed with Krebs– (Trizol, Gibco-BRL). Five micrograms of total RNAwere electrophoresed through a 1% formaldehyde-Henseleit buffer and perfused with nominally cal-

cium-free buffer containing 0.05% (w/v) agarose gel, blotted onto a nylon membrane (Genes-creen Hybridization Transfer, NEN Life Sciencecollagenase and 0.03% (w/v) hyaluronidase for

30 min. The atria and great vessels were removed, Products, Boston, MA, USA) by capillary transfer,and then fixed by UV cross-linking. The molecularthe remaining ventricular tissue was minced and

trypsin (0.02 mg/ml) and deoxyribonuclease probes used were a 217 bp fragment representinga portion of the rat NOS2 cDNA sequence31 and a(0.02 mg/ml) (Worthington Biochemical) were

added. After a 20-min enzymatic digestion, the cells 536 bp fragment representing a portion of the ratGAPDH cDNA sequence. The following primerswere centrifuged at 50×g for 3 min. The myocytes

were resuspended in DMEM and plated on laminin- were used: for NOS2, sense 5′-(GAG ATC AAT GCAGCT GTG)-3′ corresponding to base pair 1342–coated dishes (Costar, Cambridge, MA, USA), and

used the following day for experiments. 1359; antisense 5′-(AGA ATG GAG ATA GGA CGT)-3′ was complimentary to base pair 1541–1558 ofMyocytes were maintained in DMEM (Gibco BRL)

enriched with albumin, creatine, carnitine, insulin, the cDNA sequence of the rat vascular smoothmuscle cell NOS2 isoform.27 For GAPDH, thetaurine and triiodothyronine as previously de-

scribed.11 primers were selected from a published sequence ofrat GAPDH; sense 5′-(CAA GAT GGT GAA GGTCGG TGT GAA CG)-3′ corresponded to base pair846–871 and the antisense 5′-(CAC AGT CTT CTGIsolation and culture of rat alveolar macrophagesAGT GGC AGT GAT GG)-3′ was complimentary tobase pair 1382–1407. The identity of the amplifiedAlveolar macrophages were obtained by broncho-

alveolar lavage of male Wistar rats, 350–400 g cDNA used as probes was confirmed by sequencingat our facility.weight (Iffa Credo, Saint Germain sur l’Arbresle,

France) as previously described.30 The first cellular The cDNA probes were radiolabelled with [32P]by the random primer labeling technique (Gibcopellet was resuspended (5×105 cells/ml) in phenol

red-free DMEM (Gibco BRL) supplemented with 1% BRL). Hybridization was performed with QuickHyb

(Stratagene, La Jolla, CA, USA), according to the(v/v) heat-inactivated FCS, 1% glutamine (w/v),penicillin and streptomycin 100 IU/ml (designated manufacturer’s instructions. The membranes were

washed with 1× SSC/0.1% SDS solution (SSC con-here as culture medium), plated on tissue cultureplates (Costar) and incubated at 37°C in 10% CO2/ tains 0.15 NaCl and 0.015 sodium citrate) for

20 min at room temperature, followed by 0.2×air. One hour after plating, non-adherent cells wereremoved by washing the cultures twice with DMEM. SSC/0.1% SDS for 30 min and 0.1× SSC for 10 min

at 65°C. The membranes were exposed on a Kodak-AR film at−70°C for 8 h. NOS2 mRNA abundancewas normalized to GAPDH mRNA. The density ofExperimental protocolthe digitized blots was measured with a densityanalysis software (Scion Image, Scion Corp., Fre-Cardiac microvascular endothelial cells were treated

for 24 h with 10 �g/ml LPS from E. coli and 50 IU/ derick, MD, USA).


Measurement of atRA effect on NOS2 mRNA stability (brain isoform) expressed in rat cerebellum. After a60-min wash, a secondary goat anti-rabbit antibody

The effect of atRA on NOS2 mRNA stability was conjugated to alkaline phosphatase was added for1 h at a dilution of 1:10 000. The blots were washedinvestigated as follows. Cultured CMEC or ARVM

were treated with LPS/IFN-� or IL-1�/IFN-�, re- five times with TBST for 10 min, followed by de-tection of immunoreactive proteins by the enhancedspectively, as described above. After 18 h, 10 �g/

ml actinomycin D were added, and cells were har- chemiluminescence method using Renaissance

(NEN Life Science Products) as a substrate. Mem-vested at 0, 3, 6, 9, and 12 h thereafter, a time-frame used by previous reports.32–34 Extracted brane-bound chemiluminescence was detected with

Kodak-AR film (Eastman Kodak Co., Rochester, NY,mRNA was separated on a 1.2% formaldehyde-agarose gel and the density of the blots was cor- USA).

NOS2 protein abundance was estimated by meas-rected for differences in loading by calculating theratio to GAPDH mRNA. The resulting curve was uring the density of the digitized blots with a density

analysis software (Scion Image) and expressed aslinearized by using the natural logarithm for theoptical density values measured at each time point. arbitrary units.Any difference in the slope of decrease of mRNAwas assessed by linear regression.

Assessment of NO synthase activity: conversion of[3H]L-arginine to [3H]L-citrulline

Analysis of NOS2 protein expression by Western blotThe cells were detached with a cell scraper andresuspended in 0.25 ml 20 m Tris-HCl pH 7.4,

Cultured CMEC and ARVM were harvested on ice0.5 m EDTA, 0.5 m EGTA, 1 m DTT, 1 �

in a lysis buffer of the following composition: 20 mTHB4, 1 � leupeptin and 0.2 m PMSF. The cells

Tris-HCl, 0.5 m EDTA, 0.5 m EGTA, 200 �were disrupted by three cycles of freezing/thawing

phenylmethylsulfonyl fluoride (PMSF), 1 � pep-and vortexed. To assess NOS2 enzymatic activity,

statin, 1 � leupeptin, 50 IU/ml aprotinin, 1 �25 �l of cell homogenate corresponding to ap-

tetrahydrobiopterin (THB4, Research Biochemicalsproximatively 50 �g of protein were incubated at

International, Natick, MA, USA), 1 m dithiotreitol37°C for 1 h in 150 �l 50 m HEPES (pH 7.4 at

(DTT). The samples were sonicated three times for37°C), 1 m EDTA, 0.5 m NADPH, 5 � FAD,

8 s on ice and then stored at−80°C until analysis.5 � FMN, 10 �g/ml calmodulin and 50 n [3H]-

Protein concentration was assessed by the Bradfordarginine (Amersham Pharmacia Biotech Inc., Pi-

method (Bio-Rad Protein Assay Kit, Bio-Rad, Her-scataway, NJ, USA). Calcium chelators EDTA and

cules, CA, USA). Macrophages (5×105 cells) wereEGTA were used in order to measure the NOS

lysed in 200 �l sample buffer containing 0.5 Tris-enzymatic activity of calcium-independent NOS,

HCl pH 6.8, 10% (w/v) SDS, 10% (v/v) glycerol,i.e. NOS2. The enzymatic reaction was stopped by

5% (v/v) �-mercaptoethanol and 0.05% (w/v) bro-adding 2 ml of 20 m HEPES (pH 5.5) and 5 m

mophenol blue.EDTA solution. The final solution was placed into

The samples were mixed in loading buffer con-Dowex -50W×8 columns (sodium form, 200 Mesh,

taining SDS, DTT and bromophenol blue. ProteinsBioRad) equilibrated with a solution of cold HEPES

(20 �g) were separated through a 7.5% poly-20 m (pH 5.5 at 4°C). The columns retained [3H]-

acrylamide electrophoresis gel (Bio-Rad) and trans-arginine whereas [3H]-citrulline was eluted with

ferred electrophoretically in transfer buffer [25 m2 ml distilled water. Enzymatic activity was ex-

Tris-HCl pH 7.2, 193 m glycine, 0.05% (w/v) SDS,pressed in pmol/mg protein/hour. Under these con-

20% (v/v) methanol] onto a PVDF membrane (Bio-ditions, the enzymatic reaction was linear with time

Rad) overnight at 4°C. After blocking for one hourand with the concentration of total cell homogenate

at room temperature with 5% (w/v) non-fat dryprotein added. Protein concentration was measured

milk in TBST [0.1 Tris-HCl pH 7.5, 9% (w/v)by the Bradford method (Bio-Rad).

sodium chloride and 0.1% (v/v) Tween 20], themembrane was probed for one hour at room tem-perature with a polyclonal NOS2 antibody (Trans- Measurement of nitrite concentration in the culture cell

supernatantduction Laboratories, Lexington, KY, USA) at adilution of 1:2500. The specificity of this antibody

Nitrite concentration measurement in the cellhas been described elsewhere.35 In Western blotexperiments, it did not cross-react with rat NOS3 supernatant was performed as previously de-

scribed.29 Briefly, 150 �l of the supernatant were(endothelial isoform) expressed in rat aorta or NOS1


added to 900 �l of Griess reagent (0.75% sulf- abundance from 1.63±0.18 to 1.10± 0.21 (thatis 69% of LPS/IFN-�-stimulated CMEC) andanilamide, final concentration in 0.5N HCl/0.75%

naphthylethylene diamine). Nitrite concentration 0.99±0.18 (that is 62% of LPS/IFN-�-stimulatedCMEC), P<0.05, for stimulated CMEC treated withwas measured as absorbance at 540 nm. Standard

curves were obtained using known concentrations atRA 10−6 and 10−5 , respectively, and from1.81±0.40 to 1.08±0.27 (that is 77% of IL-1�/of sodium nitrite (linear relationship from 0.1 to

50 �). A standard curve was constructed for each IFN-�-stimulated ARVM) and 0.57±0.09 (that is63% of IL-1�/IFN-�-stimulated ARVM), P<0.05,experiment. Samples were assayed in duplicate.

Cultured CMEC and ARVM were stimulated as for stimulated ARVM treated with atRA 10−6 and10−5 respectively (Fig. 1).described under Experimental protocol in the pres-

ence or absence of atRA at 10−6 or 10−5 . In macrophages stimulated with LPS/IFN-� andcoincubated with 10−5 atRA, NOS2 mRNA ex-Alveolar macrophages were grown in the culture

medium and stimulated for 24 h with 10 �g/ml LPS pression was reproducibly and significantly de-creased from 0.73±0.03 to 0.55±0.04 (that isand 50 IU/ml IFN-� and in the absence or presence

of atRA at concentrations ranging from 10−8 to 61% of LPS/IFN-�-stimulated macrophages),P<0.05 (Fig. 1).10−4 . In all experiments the effect of the vehicle

(DMSO) used to dissolve atRA was used as a neg-ative control. Twenty-four hours after incubation inthe presence of the study drugs, the cell supernatant Effect of atRA on NOS2 mRNA half-life in CMEC and

ARVMwas harvested, centrifuged at 1500×g and storedat −20°C until analysis.

In order to investigate the potential mechanismsthrough which atRA decreases NOS2 mRNAabundance in CMEC and ARVM we performed ex-Statistical analysisperiments in which mRNA transcription wasstopped by the addition of actinomycin D.Statistical analysis was performed with the Statview

In cultured CMEC stimulated with LPS/IFN-�,IV software (Abacus Concept, Inc., Berkeley, CA,NOS2 mRNA half-life was unchanged as reflected byUSA). Results are expressed as mean±... Stat-the horizontal slope in Figure 2(a). This is consistentistical difference among several groups was testedwith the very long half-life of NOS2 mRNA reportedby non-parametric analysis of variance (Kruskall–by Vodovotz et al.36 In the presence of 10−5 atRA,Wallis test). Statistical difference between twoNOS2 mRNA half-life was significantly shortenedgroups was determined by the Mann–Whitney non-(P=0.048) to approximately 360 min. When CMECparametric test. Comparison of NOS2 mRNA half-were stimulated with IL-1�/IFN-�, NOS2 mRNAlife was performed by linear regression analysis.half-life was not modified by atRA (180 min vA P value less than 0.05 was the criterion of165 min) [Fig. 2(b)]. Thus atRA had different effectssignificance.on NOS2 mRNA stability according to the type ofpro-inflammatory stimulus used.

In cytokine-stimulated ARVM, NOS2 mRNA half-life was estimated to be approximately 260 minResults[Fig. 2(c)]. Addition of atRA 10−5 significantly(P=0.023) decreased NOS2 mRNA abundance, butEffect of atRA on NOS2 mRNA expression in CMEC,

ARVM, and macrophages did not change its half-life (210 min, P=0.98 forslope difference) [Fig. 2(c)]. These results are con-

As previously reported,27,31 under control conditions sistent with an effect of atRA on NOS2 gene tran-scription with no change in NOS2 mRNA half-lifeno NOS2 mRNA, protein or activity could be de-

tected in cultured CMEC and ARVM. In un- in ARVM.stimulated (control) macrophages, NOS2 mRNAand protein expression was not detectable.

In CMEC stimulated with LPS/IFN-� and in ARVM Effect of atRA on NOS2 protein expressionstimulated with IL-1�/IFN-� there was a re-producible de novo induction of NOS2 mRNA (Figure In the absence of stimulation by LPS and/or cyto-

kines, NOS2 protein expression (as studied by West-1). Addition of atRA (10−6 and 10−5 ) to stim-ulated CMEC and ARVM significantly reduced NOS2 ern blot experiments) in CMEC, ARVM, and

macrophages homogenates was not detectable.mRNA expression normalized for GAPDH mRNA


(a)

(b)

(c)


(a) 1.62±0.26 to 0.71±0.21, P<0.05 for stimulatedARVM treated with atRA 10−5 (Fig. 3).

In macrophages stimulated with LPS and IFN-�,atRA 10−5 reproducibly and significantly(P<0.02) decreased the expression of the NOS2protein from 2.81±0.48 to 1.52±0.27 (Fig. 3).

(b)Effect of atRA on nitrite accumulation in the cell culturesupernatant

Nitrite concentration was below the limit of detectionin the supernatant of control CMEC or ARVM. InCMEC stimulated with LPS/IFN-� there was markednitrite accumulation that was sig-nificantly (P<0.05) reduced in the presence of atRA

(c) 10−5 (4.39±0.39 v 2.42±0.32 �). Nitrite con-centration was below the limit of detection in thesupernatant of ARVM stimulated with IL-1�/IFN-�.

Control macrophages exhibited a low con-centration of nitrite in the supernatant. Macrophagestimulation with LPS/IFN-� increased nitrite ac-cumulation in the supernatant from 2.25±0.16 to11.51±0.77 � (P<0.05). Treatment with atRA10−4 and 10−5 , but not atRA 10−6 , significantlyFigure 2 Effect of atRA on NOS2 mRNA half-life. Cul-reduced the nitrite accumulation in the super-tured CMEC were stimulated with LPS/IFN-� (a) or IL-natant from 11.51±0.77 to 3.25±0.73 and1�/IFN-� (b) for 16 h, or ARVM were stimulated with

IL-1�/IFN-� (c). Actinomycin D was added and mRNA 6.85±0.48 �, respectively (P<0.05).extracted 3, 6, 9 and 12 h thereafter. When the stimulusused was LPS/IFN-�, atRA 10−5 significantly reducedNOS2 mRNA half-life (∗P=0.048). When the stimulusused was IL-1�/IFN-�, RA 10−5 significantly decreased Effect of atRA on NOS2 enzyme activity as measured byNOS2 mRNA abundance without changing its half-life. conversion of [3H]L-arginine to [3H]L-citrulline

The baseline NOS2 activity in unstimulated controlCMEC and ARVM was low (0.2±0.07 and0.13±0.02 pmol/mg/h, respectively). In CMECstimulated with LPS/IFN-� and in ARVM stim-In CMEC stimulated with LPS/IFN-� and in ARVM

stimulated with IL-1�/IFN-� there was reproducible ulated with IL-1�/IFN-� there was a reproducibleincrease in NOS2 enzymatic activity. Addition ofde novo expression of a 130 kDa protein recognized by

a NOS2-specific antibody (Fig. 3). Addition of atRA atRA 10−5 resulted in a decrease of NOS2enzymatic activity from 0.8±0.13 pmol/mg/h to(10−6 and 10−5 ) to stimulated CMEC and ARVM

significantly reduced NOS2 protein expression (op- 0.55±0.12 pmol/mg/h (P=0.088) in stimulatedCMEC and from 0.58±0.14 pmol/mg/h totical density expressed in arbitrary units) from

3.01±0.53 to 1.10±0.21 and 0.95±0.33, 0.3±0.07 pmol/mg/h in stimulated ARVM (P=0.048).P<0.05 for stimulated CMEC treated with atRA 10−6

and 10−5 respectively (Fig. 3); and from In contrast to what was observed in CMEC

Figure 1 Effect of all-trans retinoic acid on NOS2 mRNA expression. Representative Northern blot analyses of totalRNA extracted from cultured rat CMEC (a), ARVM (b) or macrophages (c). Cells were incubated in control medium(Control) or stimulated with a combination of LPS and IFN-� (for CMEC and macrophages) or IL-1� and IFN-� (forARVM) as described under Materials and Methods in the absence (LPS/IFN-� or IL-1�/IFN-�) or presence of 10−5 atRA (LPS/IFN-�+atRA or IL-1�/IFN-�+atRA). Hybridization was performed with a 32P-labelled NOS2 cDNA fragmentand GAPDH cDNA sequence to correct for loading differences. Relative abundance of NOS2 mRNA was significantlyreduced in the presence of atRA. The experiments were performed three or more times with similar results. Results ofthe densitometric analysis are presented in the Results section.


(a)

(d)

(b)

(e)

(c)

(f)

Figure 3 Effect of atRA on NOS2 protein expression. Representative Western blot analysis of total cell homogenatefrom rat CMEC (a), ARVM (b) and macrophages (c) incubated in control medium (Control) or stimulated for 24 h asdescribed under Material and Methods with a combination of LPS and IFN-� (for CMEC and macrophages) or IL-1�and IFN-� (for ARVM) in the absence (LPS/IFN-� or IL-1�/IFN-�) or presence of atRA 10−5 (LPS/IFN-�+atRA or IL-1�/IFN-�+atRA). Staining the membranes with Ponceau Red checked equal loading of the wells in the electrophoresisgel and equal transfer to the membranes. A NOS2-specific polyclonal antibody recognized a protein whose molecularweight was approximately 130 kDa. (d–f) Bar graphs representing the densitometric analysis of NOS2 protein opticaldensity (arbitrary units) are shown for CMEC (d), ARVM (e), and macrophages (f), treated with LPS/IFN-� or IL-1�/IFN-� in the absence or presence of atRA 10−6 or 10−5 . Results are expressed as mean±... of densitometryanalysis arbitrary units; the experiments were performed three or more times with similar results; ∗P<0.01 as comparedto LPS+IFN-� group (CMEC and macrophages) or IL-1�+IFN-� group (ARVM).

and ARVM, unstimulated macrophages showed Nevertheless, although the basal NOS activity ofunstimulated macrophages varied from experimentbackground conversion of [3H] -arginine to [3H]

-citrulline (2.25±2 pmol/mg/h) that could be to experiment, treatment with LPS and IFN-�always resulted in a several-fold increase in [3H]inhibited by N6-monomethyl -arginine (-NMMA)

(data not shown) suggesting the presence of -arginine to [3H] -citrulline conversion. In thepresence of atRA 10−5 , NOS2 activity of LPS andconstitutive NOS activity as previously reported.37


IFN-�-stimulated macrophages was significantly segments able to trans-activate heterologous pro-moters.44–46 RXR can heterodimerize with other(P<0.01) and reproducibly reduced from 10.7±10

to 5.4±4 pmol/mg/h. nuclear receptors such as triiodothyronine re-ceptor, the vitamin D receptor, the peroxisomeproliferating activated receptors (PPARs) and mod-ulate gene transcription.47 Retinoic acid modulatesDiscussiongene expression through complex mechanismsthat are mostly transcriptional but also non-The main findings of the present study are

that: (i) all-trans retinoic acid at therapeutic transcriptional.48 In addition, RARs and RXRshave been shown to be important modulators ofconcentrations (1–10 �)38 attenuated the ac-

tivation of the NOS2 biosynthetic pathway in the transcriptional machinery through complexprotein-protein interactions with other tran-cultured rat cardiac microvascular endothelial

cells and myocytes in a manner similar to that scription factors such as AP-149,50 andNF�B.51,52 Retinoids have been successfully used asobserved in inflammatory cells such as alveolar

macrophages; (ii) the effects of atRA on the NOS2 therapy of several human diseases with clinicallyacceptable side effects.38pathway in cardiac cells are dependent on the

pro-inflammatory stimulus used to induce NOS2.When LPS and IFN-� were used, atRA decreasedNOS2 mRNA half-life whereas when IL-1� and Effects of atRA on LPS/IFN-�-stimulated NOS2

pathwayIFN-� were used, the effect of atRA was consistentwith decreased NOS2 gene transcription.

Retinoic acid may have had toxic effects inNitric oxide derived from NOS2 activation is anon-specific immune effector of activated macro- the experiments performed in this study. Several

arguments are against this hypothesis the mostphages and other immune and non-immunecells,39 part of the innate immune response to important being: (i) the relatively modest effect

of retinoids on NOS2 pathway activation inpathogens. Several approaches aimed at mod-ulating NOS2 activation have been investigated. macrophages stimulated with LPS alone; (ii) the

absence of signs of cellular toxicity upon inspectionMost of these have used relatively non-specificNO synthase inhibitors such as -NMMA.40 At- of cells cultured in the presence of atRA and

(iii) the yield of protein (mg/ml) in cellulartempts to modulate NOS2 activation specificallywithout interfering with the constitutive NO syn- homogenates from cells incubated in the presence

or absence of atRA was similar. These observationsthases (NOS1 and NOS3) activities could benefitfrom using modulators of NOS2 induction rather support the idea that retinoic acid-mediated at-

tenuation of the LPS and/or cytokines-stimulatedthan inhibitors of NOS2 enzymatic activity. Glu-cocorticoids have been reported to inhibit NOS2 NOS2 pathway in our experiments was not due

to toxic effects of atRA.transcription41 but their anti-inflammatory effectsare probably too broad thus explaining their lack The effects of atRA on NOS2 activation are prob-

ably complex. A computer analysis of the publishedof beneficial effect in human septic shock trialswhen administered in high doses.42,43 In order to 1845 bp rat NOS2 gene promoter demonstrated a

putative RARE.53 Recently, Sirsjo et al. demonstratedcharacterize other potential inhibitors/modulatorsof NOS2 transcription that have already been that RAR� decreased the expression of a gene con-

struct consisting of the murine NOS2 gene promoterused in human clinical practice, we have in-vestigated the effects of retinoids on NOS2 ac- fused to a luciferase reporter gene transfected into

cultured rat vascular smooth muscle cells, consistenttivation.Retinoids are members of the superfamily of with a direct effect of atRA via RAR� on NOS2 gene

transcription.26 Our results demonstrate that re-steroid/thyroid hormone nuclear receptors andact as ligand-induced transcription factors.44.45 tinoic acid decreased NOS2 mRNA half-life after LPS/

IFN-� stimulation but not after IL-1�/IFN-� stimu-Their pleiotropic effects are mediated by twodistinct classes of receptors, the retinoic acid lation in CMEC and ARVM. Retinoic acid could mod-

ulate NOS2 gene expression and activity either byreceptor (RAR) and the retinoid×receptors (RXR)that bind to DNA as homo- or heterodimers44–46 direct mechanisms mediated by RAR� binding to the

putative RARE in the promoter of NOS2 gene26,54 orRAR and RXR recognize specific sequences on theDNA designated retinoic acid response elements by protein–protein interaction with other tran-

scription factors such as AP-155 and/or NF�B51,52(RARE). These are found in the regulatory regionsof several natural target genes and in DNA both reported to be involved in the modulation of


NOS2 gene expression.56,57 In addition RA has been the effects of retinoids are cell type-dependent,62 anobservation which is consistent with the previouslyshown to modulate cytokine gene expression in cul-

tured cells58,59 and could therefore indirectly decrease described context-dependent effects of retinoicacid.63NOS2 gene expression by attenuating the gene ex-

pression of pro-inflammatory cytokines. Clinically relevant vitamin A/retinoic acid deficitshave been described with mild alcohol con-Inducible nitric oxide synthase enzymatic activity

could also be influenced by atRA as previously sumption64 and with sepsis.65 Thus, vitamin A de-ficiency and decreased RA availability couldreported for glucocorticoids shown to modulate

mRNA expression of the -arginine transporters14 influence the hormonal status of the host andmodulate the host’s immune and inflammatoryand the enzymes involved in the metabolism of

tetrahydrobiopterin, an essential cofactor for NO response. The mechanisms responsible for the effectsof vitamin A on the complex host/pathogen inter-synthase enzymatic activity.14 Differential effects

on NOS2 mRNA/protein expression and cofactors actions, which include, but are not limited to, theNOS2 pathway,66,67 need further studies. Sup-could explain the discordance between the effects of

atRA on NOS2 mRNA/protein and NOS2 enzymatic plementation with retinoic acid could be used astherapy in situations in which the NOS2 pathwayactivity found in CMEC in the present study. Indeed,

NOS2 enzyme activity assessed by the conversion is activated, such as cytokine administration forcancer therapy68 or septic shock. Administration ofof [3H] -arginine to [3H] -citrulline is performed

with cofactors in excess, whereas nitrite ac- retinoids in clinical practice for cancer therapy hasbeen reported to be well tolerated.38cumulation in the cellular supernatant is the result

of the global enzyme activity with cofactors at their In conclusion, we have confirmed and extendedprevious work and shown that retinoids attenuateconcentrations in intact cells.

Taken together, the results presented here are NOS2 pathway activation upon stimulation withLPS and cytokines in cardiac cells through a pro-consistent with an effect of atRA on NOS2 gene

transcription when the stimulus is IL-1�/IFN-�, and inflammatory stimulus-specific mechanism. Furtherstudies are necessary to document the moleculardecreased mRNA half-life when LPS is added to

IFN-� (in CMEC). Signaling through LPS and IFN- mechanisms by which retinoids attenuate NOS2pathway activation and to confirm the effects of� for NOS2 induction is different in that IFN-�

increases NOS2 gene transcription whereas LPS retinoids on the NOS2 pathway in vivo.stabilizes NOS2 mRNA without any effects on genetranscription.60 The interaction between atRA andLPS could not be investigated in ARVM because AcknowledgementsLPS per se does not modulate NOS2 gene expressionunder the experimental conditions of this study. Sandrine Grosjean was financed by grants from(S. Grosjean, D. Ungureanu-Longrois, unpublished Societe Francaise d’Anesthesie-Reanimation andobservations). from a scholarship of Ministere Francais des Affaires

Etrangeres (Programme Lavoisier). Yvan Devaux,Paul-Michel Mertes and Dan Ungureanu-Longrois

Potential implications of the present study were financed by a grant awarded to UPRES-EA971068. All authors except RAK are members of

The biological and potential clinical implications of ARISC and AIMAR. The technical assistance ofthese results are at the moment speculative because Minying Pu and the secretarial assistance of Claudethe vast majority of studies on atRA and NOS2 Baillot and Rebecca Clement for this manuscriptactivation have been performed in vitro. Never- are gratefully acknowledged.theless, recent reports have demonstrated a bene-ficial effect of vitamin A supplementation in severalanimal sepsis models.20

ReferencesOur results confirm and extend to three new cell

types (i.e. cardiac microvascular endothelial cells, 1. U-L D, B JL, K RA,S TW. Myocardial contractile dysfonction incardiac myocytes and alveolar macrophages) pre-the systemic inflammatory response syndrome: rolevious studies23,25,26,61 that demonstrated an effect ofof a cytokine-inducible nitric oxide synthase inthe vitamin A metabolite, atRA, on NOS2 ac-cardiac myocytes. J Mol Cell Cardiol 1995; 27:

tivation. The fact that retinoic acid attenuates the 155–167.activation of the NOS2 pathway in several cellular 2. H H, C J, W M, J LA, M

SA, F SM, W LM, B VL, Btypes is important because it has been shown that


CJ, Y N, C M, S J, T expression by interleukin-4 in primary mousemacrophages. J Leukoc Biol 1994; 55: 227–233.HB. Expression of cardiac cytokines and inducible

form of nitric oxide synthase (NOS2) in Try- 17. B C, V Y, N C. Macrophagedeactivation by interleukin 10. J Exp Med 1991;panosoma Cruzi-infected mice. J Mol Cell Cardiol

1999; 31: 75–88. 174: 1549–1555.18. B R, G MH, G JB. Vitamin A3. K RA, B JL, S TW. Nitric oxide

and cardiac function. Circ Res 1996; 79: 363–380. metabolism: new perspectives on absorption, trans-port and storage. Physiol Rev 1991; 71: 951–990.4. K J, G-J T, M JS, R

ME. Exacerbated transplant arteriosclerosis in in- 19. H EA, S JI, B SA, V JJ.Vitamin A supplementation improves macrophageducible nitric oxide-deficient mice. Circ 1998; 97:

2059–2065. function and bacterial clearance during ex-perimental salmonella infection. Proc Soc Exp Biol5. L-W A, C CL, A S, W TC,

H RH, L CJ. Inducible nitric oxide Med 1989; 191: 47–54.20. E M, L K, D P, S T,synthase expression in coronary arteries of trans-

planted human hearts with accelerated graft ar- E O. Beneficial effects of pre-treatment withvitamin A on cardiac and pulmonary functions interiosclerosis. Am J Pathol 1997; 151: 919–925.

6. G Y, J WK, X YT, T XL, B W, W endotoxaemic pigs. Acta Anaesthesiol Scand 1996;40: 538–548.WJ, H H, L VE, P P, Y Z, Q Y,

B R. The late phase of ischemic preconditioning 21. W U, T A, B T, HLA, K H, D UI. Vitamin A deficiencyis abrogated by targeted disruption of the inducible

NO synthase gene. Proc Natl Acad Sci USA 1999; predisposes to Staphylococcus aureus infection. InfectImmun 1996; 64: 209–214.96: 11507–11512.

7. H C, K G, M U, M H, 22. G ER, B JA, K MR, STA, N DW, S MM, M JS,S GA, B F. Defective nitric oxide

effector functions lead to extreme susceptibility of H RW. Mortality associated with low plasmaconcentrations of beta carotene and the effect ofTrypanosoma cruzi-infected mice deficient in gamma

interferon receptor or inducible nitric oxide syn- oral supplementation. JAMA 1996; 275: 699–703.23. M K, MQ T, T S, P R,thase. Inf Immun 1998; 66: 1208–1215.

8. MM JD, N C, H C. Altered re- A BB. Inhibition by all-trans retinoic acidof tumor necrosis factor and nitric oxide productionsponses to bacterial infection and endotoxic shock

in mice lacking inducible nitric oxide synthase. by peritoneal macrophages. J Leukoc Biol 1994;55: 336–342.Cell 1995; 81: 641–650.

9. G SA, A MA, M RN, K- 24. B PA, L G L, K S, C O,F C, I F, M-P JM, D P, GW, K RA, P MA, P JM.

Inducible nitric oxide synthase and tumor necrosis M MD, A M. CD23-mediated nitricoxide synthase pathway induction in human ker-factor in animal models of myocardial necrosis

induced by coronary artery ligation or isoproterenol atinocytes is inhibited by retinoic acid derivatives.J Invest Dermatol 1996; 106: 1182–1186.injection. J Card Fail 1999; 5: 236–245.

10. K IA, S JD, M T, K 25. H K, O’S KM, R P,W MR. Inhibition of nitric oxide synthesisLH. A dichotomous role for nitric oxide during

acute Toxoplasma gondii infection in mice. Proc Natl in vascular smooth muscle by retinoids. Br JPharmacol 1994; 113: 1448–1454.Acad Sci USA 1997; 94: 13955–13960.

11. U-L D, B JL, S 26. S A, G AC, O A, T H, AM, K H, F U, H GK.WW, O I, L S, L CJ, K

SL, M T, K RA, S TW. Induction of Retinoic acid inhibits nitric oxide synthase-2 ex-pression through the retinoic acid receptor-�.nitric oxide synthase activity by cytokines in

ventricular myocytes is necessary but not sufficient Biochem Biophys Res Commun 2000; 270: 846–851.27. B JL, U-L D, Sto decrease contractile responsiveness to �-ad-

renergic agonists. Circ Res 1995a; 77: 494–502. WW, K L, L CJ, L S, KRA, S TW, M T. Induction of NO synthase12. F P, D H, A I, R RA,

S R. Peroxynitrite is a major contributor to in rat cardiac microvascular endothelial cells byIL1-� and IFN-�. Am J Physiol 1995; 268: H1293–cytokine-induced myocardial contractile failure. Circ

Res 2000; 87: 241–247. H1303.28. N M, C WW, G ME, E13. N C, X QW. Regulation of biosynthesis of

nitric oxide. J Biol Chem 1994; 269: 13725–13728. Ø, K RA, S TW. Isolation and charac-terization of human and rat cardiac microvascular14. S WW, U-L D, S GK,

S TW, K RA. Glucocorticoids regulate endothelial cells. Am J Physiol 1993; 264: H639–H652.inducible nitric oxide synthase by inhibiting te-

trahydrobiopterin synthesis and -arginine trans- 29. U-L D, B JL, O I,S WW, K L, L CJ, Kport. J Biol Chem 1996; 271: 23928–23937.

15. V Y, B C, P J, N C. Mech- SL, M T, K RA, S TW. Contractileresponsiveness of ventricular myocytes to iso-anisms of suppression of macrophage nitric oxide

release by transforming growth factor �. J Exp proterenol is regulated by induction of nitricoxide synthase activity in cardiac microvascularMed 1993; 178: 605–613.

16. B C, V Y, P J, X QW, N C. endothelial cells in heterotypic primary culture.Circ Res 1995; 77: 486–493.Mechanism of suppression of nitric oxide synthase


30. B JL, U D, K RA, K L, severe sepsis and septic shock. N Engl J Med 1987;317: 653–658.P D, M T, S TW. Abnormal

43. H L, P P, Y E. Effect of high-contractile function due to induction of nitric oxidedose glucocorticoid therapy on mortality in patientssynthesis in rat cardiac myocytes follows exposurewith clinical signs of systemic sepsis. N Engl J Medto activated macrophage-conditioned medium. J1987; 317: 659–665.Clin Invest 1993; 91: 2314–2319.

44. C P. The molecular and genetic dissection31. B JL, U-L D, Sof the retinoid signaling pathway. Gene 1993; 135:WW, P D, M TA, K M,223–228.T Z, L CJ, D AJ, K RA,

45. E RM. The steroid and thyroid hormoneS TW, M T. Cytokine-inducible nitric oxidereceptor superfamily. Science 1988; 240: 889–895.synthase (iNOS) expression in cardiac myocytes. J

46. C P. The retinoid signaling pathway: mo-Biol Chem 1994; 269: 27580–27588.lecular and genetic analyses. Sem Cell Biol 1994;32. DN MR, C CL, L DM. Pho-5: 115–125.sphatidyl serine is involved in the reduced rate of

47. U K, G V, G CK, R MG,transcription of the inducible nitric oxide synthaseE RM. Retinoic acid and thyroid hormonegene in macrophages from tumor-bearing mice. Jinduce gene expression through a common re-Immunol 1997; 158: 1810–1817.sponsive element. Nature 1988; 336: 262–265.33. L P, S A, S G. Modulation

48. A J, L F, P J. Adipsin geneof inducible nitric oxide synthase mRNA stabilityexpression in 3T3-F442A adipocytes is post-by tetrahydrobiopterin in vascular smooth muscletranscriptionally down-regulated by retinoic acid.cells. Biochem Biophys Res Commun 1998; 243:J Biol Chem 1991; 266: 1157–1161.137–141.

49. L R, K SJ, A P, R AB, S34. R YS, L JH, S JH, H JH, L YS, LMB, K M, W RL. Interleukin-1 stimulatesYM, H GM. Acetaminophen inhibits iNOS geneand all-trans retinoic acid inhibits collagenaseexpression in RAW 264.7 macrophages: differentialgene expression through its 5′ activator protein-1regulation of NF-�� by acetaminophen and sa-binding site. Mol Endocrinol 1990; 4: 973–980.licylates. Biochem Biophys Res Commun 2000; 272:

50. P L, E C, B C. Suppression758–764.of collagenase gene expression by all-trans and 9-35. B J, L S, U-L D,cis retinoic acid is ligand dependent and requiresD G, F T, A M. Induction ofboth RARs and RXRs. J Cell Biochem 1995; 57:diaphragmatic nitric oxide synthase by E. coli575–589.endotoxemia in rats. Role on diaphragmatic con-

51. G J, P LLL, L TJ, C SW,tractile function. J Clin Invest 1996; 98: 1550–S RA. Retinoic acid inhibits the regulated1559. expression of vascular cell adhesion molecule-1 by36. V Y, K NS, P M, M J, cultured dermal microvascular endothelial cells. J

P J, N C. Inactivation of nitric oxide Clin Invest 1997; 99: 492–500.synthase after prolonged incubation of mouse 52. N SY, K BY, C SW, H SJ, M X,macrophages with IFN-� and bacterial lipo- T G, I SY, L JW, K TS. Retinoidspolysaccharide. J Immunol 1994; 152: 4110–4118. inhibit interleukin-12 production in macrophages

37. M PR, B L, R A, H L. through physical associations of retinoid X receptorConstitutive nitric oxide production by rat alveolar and NF-��. J Biol Chem 1999; 274: 7674–7680.macrophages. Am J Physiol 1998; 274: L360–L368. 53. E W, K D, H R, P J.

38. M JR, F SR, H C, DG F, Molecular cloning of the rat inducible nitric oxideG WA, Y CW, W RP J. Clinical synthase gene promoter. Biochem Biophys Respharmacology of oral all-trans retinoic acid in Commun 1996; 223: 752–756.patients with acute promyelocytic leukemia. Cancer 54. G AC, Z W, G A, S A. SyntheticRes 1992; 52: 2138–2142. retinoids improve survival in rodent model of

39. MM JD, X QW, N C. Nitric oxide endotoxic shock. Eur J Surg 2000; 166: 165–169.and macrophage function. Annu Rev Immunol 55. S R, R P, Y N, K SA,1997; 15: 323–350. R LJ, B J, V IM, E RM.

40. H BG, B TR, S J, D Retinoic acid is a negative regulator of AP-1AJ, O J, C RD, S RL. Inhibition of responsive genes. Proc Natl Acad Sci USA 1991;nitric oxide synthesis during endotoxemia promotes 88: 6092–6096.intrahepatic thrombosis and an oxygen radical- 56. X QW, W R, N C. Promoter of themediated hepatic injury. J Leukoc Biol 1992; 52: mouse gene encoding calcium-independent nitric390–394. oxide synthase confers inducibility by interferon-�

41. K H, E C, I-B I, F- and bacterial lipopolysaccharide. J Exp Med 1993; U. Glucocorticoids inhibit the induction 177: 1779–1784.of nitric oxide synthase II by down-regulating 57. E W, K D, H R, P J.cytokine-induced activity of transcription factor Molecular cloning of the rat inducible nitric oxidenuclear factor-kappa B. Mol Pharmacol 1996; 49: synthase gene promoter. Biochem Biophys Res15–21. Commun 1996; 223: 752–756.

42. B RC, F CJJ, C TP, S GJ, 58. M K, MQ T, T S, P R,M CA, B RA. A controlled clinical trial of A BB. Inhibition by all-trans retinoic acid

of tumor necrosis factor and nitric oxide productionhigh-dose methylprednisolone in the treatment of


by peritoneal macrophages. J Leukoc Biol 1994; of retinoic acid receptors. Cell 1992; 70: 1007–1019.55: 336–342.

59. H K, O’S KM, R P, 64. W XD. Chronic alcohol intake interferes withretinoid metabolism and signaling. Nutr Rev 1999;W MR. Inhibition of nitric oxide synthesis

in vascular smooth muscle by retinoids. Br J 57: 51–59.65. S RD. Vitamin A, immunity, and infection. ClinPharmacol 1994; 113: 1448–1454.

60. W A, O S, C L, E H. Dual mech- Infect Dis 1994; 19: 489–499.66. C MT, N FE, H CE. In vitaminanism for the control of inducible-type NO synthase

gene expression in macrophages during activation by A deficiency multiple mechanisms establish a regu-latory T helper cell imbalance with excess Th1 andinterferon-� and bacterial lipopolysaccharide. J Biol

Chem 1994; 269: 8324–8333. insufficient Th2 function. J Immunol 1994; 152:1515–1522.61. M K, S H, I H, N H. Effects

of retinoids on the production of tumor necrosis 67. C MT, N FE, C TY, H CE.Vitamin A down-regulation of IFN-gamma synthesisfactor-alpha and nitric oxide by lipopolysaccharide-

stimulated rat Kupffer cells in vitro: evidence for in cloned mouse Th1 lymphocytes depends on theCD28 costimulatory pathway. J Immunol 1996; 156:participation of retinoid X receptor signaling path-

way. Cell Biochem Funct 1997; 15: 95–101. 2674–2679.68. F G, T M, B R, V M, M-62. G V, V PM, Z B, L M. Retinoic acid

interleukin-1-induced cytokine synthesis in human M, B A, B C, G PP, RC, C C. Impaired left ventricular filling ratesmonocytes. J Leukoc Biol 1993; 54: 125–132.

63. N S, S M, K P, D B, N- induced by treatment with recombinant interleukin2 for advanced cancer. Br Heart J 1994; 71: 166– H, C P. Promoter context- and re-

sponse element-dependent specificity of the 169.transcriptional activation and modulating functions

retinoic acid attenuates inducible nitric oxide synthase (nos2) activation in cultured rat cardiac...

Documents