The FASEB Journal • Research Communication

Dichloroacetate treatment partially regresses establishedpulmonary hypertension in mice with SM22�-targetedoverexpression of the serotonin transporter

Christophe Guignabert,*,†,1 Ly Tu,*,†,2 Mohamed Izikki,*,†,2 Laurence Dewachter,*Patricia Zadigue,† Marc Humbert,*,§ Serge Adnot,† Elie Fadel,*,‡ andSaadia Eddahibi*,†

*INSERM Unity “Pulmonary Hypertension: Pathophysiology and Innovative Therapies,” CentreChirurgical Marie Lannelongue, Universite Paris-Sud 11, Le Plessis-Robinson, France; †INSERMU955, Institut Mondor de Recherche Biomedicale, Creteil, France; ‡UPRES EA2705, CentreChirurgical Marie Lannelongue, Le Plessis-Robinson, France; and §INSERM U764, Hopital Antoine-Beclere, AP-HP, Clamart, France

ABSTRACT Voltage-gated potassium (Kv)1.5 is de-creased in pulmonary arteries (PAs) of patients withidiopathic pulmonary arterial hypertension (IPAH)and in experimental models including mice withSM22�-targeted overexpression of the serotonintransporter (5-HTT). The mechanisms underlyingthese abnormalities, however, remain unknown. Di-chloroacetate (DCA) inhibits chronic hypoxia- ormonocrotaline-induced PAH by inhibiting nuclearfactor of activated T-cells (NFAT)c2 and increasingKv1.5. Therefore, we hypothesized that DCA couldregress established PAH in SM22-5-HTT� mice. Weevaluated pulmonary hemodynamics, vascular re-modeling, NFATc2, and Kv1.5 protein in 20-wk-oldSM22-5-HTT� or wild-type mice treated for 1, 7, and21 d with DCA, cyclosporine-A (NFAT inhibitor), orvehicle. DCA partially reversed PAH in SM22-5-HTT� mice by decreasing proliferation and increas-ing apoptosis in muscularized PAs. Furthermore,serotonin (10�8–10�6 M) dose-dependently increasedPA-smooth muscle cell (PA-SMC) proliferation inculture (EC50�0.97�10�7 M) and DCA (5�10�4 M)vs. PBS markedly reduced the growth of PA-SMCfrom IPAH and control patients treated with thehighest dose of serotonin by 50 and 30%, respec-tively. Finally, although serotonin induces NFATc2activation in PA-SMCs, inhibition of NFATc2 alonewith cyclosporine-A was not sufficient for reversingPAH in this model. Our results support the possibil-ity that DCA may be an interesting agent for investi-gation in patients with PAH.—Guignabert, C., Tu, L.,Izikki, M., Dewachter, L., Zadigue, P., Humbert, M.,Adnot, S., Fadel, E., Eddahibi, S. Dichloroacetatetreatment partially regresses established pulmonaryhypertension in mice with SM22�-targeted overex-pression of the serotonin transporter. FASEB J. 23,000 – 000 (2009). www.fasebj.org

Key Words: pulmonary vascular remodeling � serotonin path-way � NFAT/Kv1.5 axis � PDK inhibitor � Bcl-2/Bax ratio

Pulmonary arterial hypertension (PAH) is a syn-drome of dyspnea, fatigue, chest pain, and syncoperelated to an increase in pulmonary vascular resistance(PVR) of unknown cause that impedes the ejection ofblood by the right ventricle. The most severe form isidiopathic PAH (IPAH), in which a marked PVR in-crease leads to progressive right heart failure and todeath within 2–3 yr of the diagnosis. Although impor-tant insights into the pathobiology of PAH have beengained over the past decade, the mechanisms underly-ing the progressive obliteration of the pulmonary arte-rioles remain incompletely understood (1, 2). Dysregu-lation of the finely tuned balance between proliferationand death of pulmonary artery (PA) smooth musclecells (PA-SMCs), together with pulmonary vasoconstric-tion, thrombosis, and inflammation, play a major rolein the initiation and/or progression of PAH.

Multiple lines of evidence indicate that among thevarious molecules or growth factors involved in the patho-biology of PAH, serotonin [5-hydroxytryptamine (5-HT)]and its transporter (5-HTT) play a central role (3, 4).5-HTT activation induces PA-SMC hyperplasia/hypertro-phy via activation of various intracellular signaling path-ways, including protein tyrosine phosphorylation of GTPase-activating protein (5), rapid superoxide formation,activation of Rho/Rho kinase (ROCK) (6, 7), andextracellular signal-regulated kinase (ERK)1/ERK2 mi-togen-activated protein (MAP) kinase (8, 9). 5-HTTexpression was increased in platelets and lungs frompatients with IPAH, where it predominated in themedia of the thickened pulmonary arteries and in theplexiform lesions (10). Treatment with selective 5-HTT

1 Correspondence: INSERM Unity “Pulmonary Hyperten-sion: Pathophysiology and Innovative Therapies,” CentreChirurgical Marie Lannelongue, 133 avenue de la Resistance,92350 Le Plessis-Robinson, France. E-mail: christophe.guignabert@inserm.fr

2 These authors contributed equally to this work.doi: 10.1096/fj.09-131664

10892-6638/09/0023-0001 © FASEB

The FASEB Journal article fj.09-131664. Published online August 13, 2009.

inhibitors prevented or completely reversed experi-mental hypoxic or monocrotaline PAH (11, 12). Fur-thermore, we recently reported that selective 5-HTT over-expression in SMCs at a level close to that found inhuman IPAH promoted pulmonary vascular remodelingand PAH in mice, causing marked increases in rightventricular systolic pressures (RVSPs), right ventricularhypertrophy (RVH), and muscularization of pulmonaryarterioles (13). PAH in these SM22-5-HTT� mice wors-ened with age. Furthermore, we reported that SM22-5-HTT� mice had no changes vs. wild-type (WT) mice inlung expression levels of endothelin-1, Tie2 receptor,prostacyclin synthase, or members of the bone morpho-genetic protein (BMP) pathway (BMP-RII, BMP-RIA,BMP-RIB, BMP-2, and BMP-4), which suggests thatthese mediators were not involved in establishedPAH in our model. PAH in SM22-5-HTT� micedeveloped despite unaltered 5-HT bioavailability,solely as a consequence of increased 5-HTT proteinexpression in SMCs. However, SM22-5-HTT� miceexhibited a marked decrease in lung expression ofthe voltage-gated potassium (Kv) channels Kv1.5 andKv2.1 (13). Similar abnormalities occur in culturedPA-SMCs from patients with IPAH (14 –16).

In PAH, Kv1.5 channel dysfunction or down-regula-tion may represent a predisposing factor in conjunctionwith other factors and genetic defects (2, 17). InPA-SMCs, Kv channel activity regulates the restingmembrane potential (Em) (15, 18, 19) and cytosolicfree calcium (Ca2�) concentration, because of thevoltage dependence of sarcolemmal Ca2� channels(15) (see Fig. 8). Ca2� elevation triggers vasoconstric-tion and stimulates PA-SMC proliferation. Kv1.5 expres-sion is decreased in pulmonary arteries of patients withIPAH and in various experimental models of PAHincluding the chronic-hypoxia and monocrotalinemodels (13, 14, 16, 17, 20–25). The mechanisms thatunderlie these effects, however, remain unclear. It hasbeen reported that activation of HIF-1� and c-Jun, asignal-transducing transcription factor of the AP-1 fam-ily, are two major drivers of Kv1.5 down-regulation (14,26). However, recent evidence suggests that serotoninmay modulate also Kv1.5 channel expression/activity.By using cultured human PA-SMCs, we reported thatserotonin treatment reduces Kv1.5 mRNA levels, andthis effect is inhibited by 5-HTT inhibitors, stronglysuggesting a relationship between 5-HTT activation andKv1.5 down-regulation (13). Similarly, in rat PA-SMCsand in Ltk� cells stably transfected with human Kv1.5channels, Kv currents were inhibited by serotonin viaactivation of the 5-HT2A receptor (27). However, theeffects of serotonin on expression of PA-SMC ionchannels are not well understood.

Kv channel overexpression achieved by nebulizedadenoviral Kv1.5 gene therapy (28) or oral dichloroac-etate (DCA) therapy (29) improved chronic hypoxia-or monocrotaline-induced PAH, which suggests a ther-apeutic role in PAH (2, 17). DCA inhibits mitochon-drial pyruvate dehydrogenase kinase (PDK), whichinhibits pyruvate dehydrogenase (PDH), a gatekeeping

enzyme for the entry of pyruvate into the mitochon-drial tricarboxylic acid (TCA) cycle (30, 31). DCA is awell-known substance with no major reported sideeffects and constitutes a promising metabolism-tar-geted treatment for PAH. DCA restores not only theexpression and function of Kv1.5 channels via activa-tion of the nuclear factor of activated T-cells (NFAT)c2transcription factor, but also reactive oxygen speciesproduction, mitochondrial Em, and the balance be-tween proliferation and death of PA-SMCs (21, 29,32–35).

The objectives of this experimental study were toevaluate the efficacy of DCA treatment in mice withSM22�-targeted overexpression of 5-HTT (SM22-5-HTT� mice), which spontaneously develop PAH, toexplore the relationship between 5-HTT activation andKv1.5 expression, and to evaluate the role for nucleartranslocation of NFATc2.

MATERIALS AND METHODS

Production of SM22-5-HTT� mice

To generate mice overexpressing 5-HTT specifically in theSMCs, we used the murine SM22 promoter to drive theexpression of the human 5-HTT gene, as described previously(13). These SM22-5-HTT� mice are fertile and have a normallife span and normal growth. Daily gavage was used for 24 h(d 1), 5 d (d 5), and 3 wk (d 21) to administer one of thefollowing drugs: fluoxetine (10 mg/kg/d; Sigma-Aldrich,Saint-Quentin-Fallavier, France), Dichloroacetate (DCA; 80mg/kg/d; Sigma-Aldrich), cyclosporine-A (CsA; 1 mg/kg/d;Sigma-Aldrich), or an equal volume of vehicle (saline). Allanimal experiments were approved by the Institutional Ani-mal Care and Use Committee of the French National Institute ofHealth and Medical Research (INSERM)-Unity 955, Creteil,France.

Assessment of PAH

Mice aged 20 wk were anesthetized with ketamine (60 mg/kgi.p.) and xylazine (10 mg/kg i.p.). The trachea was cannu-lated, and the lungs were ventilated with room air at a volumeof 0.2 ml and a rate of 90 breaths/min. A 26-gauge needleconnected to a pressure transducer was inserted into the rightventricle, and RVSP was recorded immediately. After theassessment of hemodynamic parameters, all mice were killedby cardiac puncture and exsanguination. The heart and lungswere removed en bloc, and the pulmonary circulation wasflushed with 3 ml of saline buffer (37°C). The right lung wasquickly harvested and divided into 2 parts that were immedi-ately snap-frozen in liquid nitrogen and then stored at �80°Cfor Western blot analysis and total RNA extraction. The leftlung was fixed by intratracheal infusion of 4% aqueousbuffered formalin at a pressure of 23 cmH2O. After fixation,a midsagittal slice of the right lung including the apical,azygous, and diaphragmatic lobes was processed for paraffinembedding. Sections (5 �m thick) were cut and stained withhematoxylin-phloxin-saffron and orcein-picroindigo-carmine. Ineach mouse, a total of 50–60 intraacinar vessels accompany-ing either alveolar ducts or alveoli were examined by anobserver masked to the genotype and treatment. Each vesselwas categorized as nonmuscular (no evidence of vessel wallmuscularization) or muscular (SMCs identifiable in all or part

2 Vol. 23 December 2009 GUIGNABERT ET AL.The FASEB Journal � www.fasebj.org

of the vessel circumference). Muscularization was defined asthe presence of typical SMCs stained red with phloxin,exhibiting an elongated shape and square-ended nucleus,and bound by two orcein-stained elastic laminae. The per-centage of muscularized vessels was determined by dividingthe number of muscularized vessels by the total number ofvessels counted in the same experimental group, as describedpreviously. The cell proliferation rate in muscularized vesselswas estimated by proliferating-cell nuclear antigen (PCNA)staining, and cell apoptosis was assessed by terminal deoxy-nucleotidyl transferase biotin-dUTP nick end (TUNEL) (11).The right ventricle (RV) was dissected from the left ventricle �septum (LV�S), and the dissected samples were weighed forcalculation of the Fulton index as RV/(LV�S).

Measurement of systemic arterial pressure (SAP) in mice

To measure SAP, animals were anesthetized with ketamineand xylazine. The right carotid was exposed, and a polyeth-ylene catheter (PE10) was inserted into the carotid artery.Arterial blood pressure was measured by connecting the freeend of the catheter to a pressure transducer, as describedpreviously (36).

Real-time quantitative-PCR (RTQ-PCR) for measurement oflung potassium channel (Kv)1.2, Kv1.5 and Kv2.1 mRNA

Lung mRNAs encoding Kv1.2, Kv1.5, and Kv2.1 were mea-sured by RTQ-PCR as described previously (10). Briefly, lungRNA was extracted using the RNeasy mini-kit (Qiagen,Courtaboeuf, France) including DNase treatment, accordingto the manufacturer’s instructions. Nucleic acid concentra-tions and purity of the total RNA extracted were determinedas the 260/280-nm ratio, and integrity was assessed by visualinspection of ethidium bromide-stained agarose gels. Onemicrogram of each RNA sample was reverse transcribed tocDNA using the SuperScript®II system (Invitrogen, Cergy-Pontoise, France). Primers for PCR were designed usingPrimer Express software (Applied Biosystems, Courtaboeuf,France). To avoid inappropriate amplification of residualgenomic DNA, intron-spanning primers were selected, andinternal control 18S rRNA primers were provided. The am-plification reaction was performed with SYBRGreen mix(Invitrogen) and specific primers. Signal detection and anal-ysis of results were performed using ABI Prism 7000 software(Applied Biosystems). Relative quantification was achievedwith the comparative ��Ct method by normalization for 18SrRNA. All samples were amplified in triplicate from the sameRNA preparation, and the mean value was computed.

Western immunoblot and immunostaining

Protein levels were assessed by Western immunoblot inlung homogenates. Tissues were homogenized and soni-cated in PBS containing protease and phosphatase inhibi-tors (Roche-Diagnostics, Mannheim, Germany). Fifty mi-crograms of protein extracts were resolved on 4 –12%NuPage Bis-Tris gels, and electrotransferred to nitrocellu-laose membranes (GE Healthcare, Orsay, France). Then,membranes were blocked for 1 h (5% milk in 0.1%TBS/Tween) and incubated with rabbit anti-Kv1.2 (1:200)or anti-Kv1.5 (1:100; Sigma-Aldrich, France), with mouseanti-NFATc2 (1:100; Acris-Antibodies, Herford, Germany),or with anti-Bax or anti-Bcl-2 (1:600; Tebu-Bio, Le Perray-en-Yvelines, France). Binding of secondary HRP antibodies(anti-rabbit, 1:5000, Tebu-Bio; anti-mouse, 1:10,000, Cal-biochem, Darmstadt, Germany) was performed. Immuno-reactive bands were detected using the ECLTM chemilumi-

nescent system (GE Healthcare). Equal loading of proteinwas confirmed by blotting for �-actin. The Kv1.5 protein wereimmunocytochemically labeled with primary antibodies forKv1.5 (1:200) as described previously (37).

Isolation and culture of human PA-SMCs

Human PA-SMCs were isolated and cultured as described pre-viously (10). To identify PA-SMCs, we examined cultured cellsfor expression of muscle-specific contractile and cytoskeletalproteins, including �-smooth-muscle-actin, desmin, and vincu-lin. Cells were used between passages 3 and 6. This study wasapproved by the local ethics committee (CPP Ile-de-France,Le Kremlin-Bicetre, France). The patients gave their in-formed consent before the study.

Effect of fluoxetine, DCA, 11R-VIVIT, or CsA treatmentson NFAT nuclear translocation in cultured human PA-SMCsat baseline and after 5-HTT activation

The effects of serotonin on NFAT nuclear translocation werestudied on cultured PA-SMCs from IPAH and control pa-tients. Synchronized cells were exposed to serotonin (Sigma-Aldrich; 10�6 M) with or without fluoxetine (10�6 M) inserum-free medium for 24 h before fixation. In addition, theeffects of DCA (5�10�4 M), 11R-VIVIT (Calbiochem; 4.10�6

M), CsA (10�6 M), and PBS on NFAT nuclear translocation incultured PA-SMCs under basal conditions or in response toserotonin were also evaluated. NFATc2 was detected using amouse monoclonal antibody (1:250) followed by FluoAlexa488-conjugated anti-mouse immunoglobulin antibody (1:200;Invitrogen). Immunofluorescence pictures were taken usinga Zeiss microscope equipped with FluoroArc fluorescencecontrol and Optimas 6.5 software (Carl Zeiss, Oberkochen,Germany).

Effect of fluoxetine, DCA, 11R-VIVIT, or CsA treatmentson proliferation of human PA-SMCs at baseline and after5-HTT activation

Freshly isolated human PA-SMCs were plated on a 24-wellplate and serum starved for 48 h. The effects of DCA (5.10�4

M), 11R-VIVIT (4.10�6 M), CsA (10�6 M), fluoxetine (10�6

M), and PBS on human PA-SMC proliferation under basalconditions or in response to serotonin (10�6 M) were as-sessed using [3H]thymidine incorporation, as described pre-viously (37).

Statistical analyses

Values for each determination are expressed as means � se.Statistical significance was tested using the nonparametricMann-Whitney test or 2-way ANOVA followed by the Student-Newman-Keuls test. Values of P 0.05 were consideredsignificant.

RESULTS

Evaluation of the efficacy of DCA vs. fluoxetinetreatment on spontaneous PAH in SM22-5-HTT�

mice

Six groups of 17-wk-old mice were established, withbalanced numbers of males and females in each geno-type (n6): 1) SM22-5-HTT� and WT mice given 10

3EFFICACY OF DCA TREATMENT IN SM22-5-HTT� MICE

mg/kg/d of the selective 5-HTT inhibitor fluoxetineorally for 3 wk, 2) SM22-5-HTT� and WT mice given 80mg/kg/d of DCA orally for 3 wk, and 3) SM22-5-HTT�

and WT mice given vehicle orally for 3 wk. PAH wasassessed based on RVSP, RVH, and muscularization ofdistal PAs, at the following time points: 24 h (d 1), 5 d(d 5), and 3 wk (d 21) after DCA or fluoxetine orvehicle administration.

In SM22-5-HTT� mice treated with vehicle and stud-ied from d 1 to 21, a marked increase in RVSP, RVH,and numbers of muscularized distal PAs were foundcompared with vehicle-treated WT mice (n5–6 ineach group, P0.001). When SM22-5-HTT� mice weregiven DCA or fluoxetine and compared on d 1 and 5 tovehicle-treated mice, no significant differences inRVSP, RVH, and muscularization of PAs were found(n5–6 in each group, PNS). On d 21, in contrast,these parameters were markedly attenuated in DCA- orfluoxetine-treated SM22-5-HTT� mice compared to

vehicle-treated SM22-5-HTT� mice. Day-21 RVSP values,RVH, and muscularization of PAs in fluoxetine-treatedSM22-5-HTT� mice were similar to those in control WTmice (n5–6 in each group, PNS) (Fig. 1).

There were no changes in RVSP, RVH, and muscu-larization of PAs in WT mice treated with vehicle, DCA,or fluoxetine (n5–6 in each group, PNS). More-over, there were no significant differences betweenSM22-5-HTT� or WT mice given vehicle, fluoxetine, orDCA regarding body weight, SAP, or heart rate (n5–6in each group, PNS) (Table 1).

Effect of DCA and fluoxetine treatments onproliferation and apoptosis in PA walls of SM22-5-HTT� mice

To determine whether PAH attenuation by DCA orfluoxetine in SM22-5-HTT� mice was ascribable to

Figure 1. RVSP, RVH, andpercentage of muscularizeddistal arteries in SM22-5-HTT� and WT mice treatedwith daily oral DCA (80 mg/kg), fluoxetine (Fluox; 10mg/kg), or vehicle (saline)for 1, 5, or 21 d. A) RVSP.B) RVH. C) Percentage ofmuscularized alveolar ductand wall arteries. Bars repre-sent means � se (n5–6).*P 0.05, **P 0.01, ***P 0.001 vs. WT mice; §§P 0.01,§§§P 0.001 vs. vehicle-treated SM22-5-HTT� mice.

TABLE 1. Body weight, heart weight, and mean blood pressure data measured on d 21 in SM22 5-HTT� and WT mice treated withvehicle (saline), DCA, fluoxetine, or CsA

Parameter

Vehicle DCA (80 mg/kg/d) Fluoxetine (10 mg/kg/d) CsA (1 mg/kg/d)

WT SM22-5HTT� WT SM22-5HTT� WT SM22-5HTT� WT SM22-5HTT�

Body weight (g) 28 � 1 28 � 1 27 � 1 26 � 1 28 � 1 27 � 1 26 � 1 26 � 1Heart rate (beats/min) 361 � 13 345 � 18 356 � 12 348 � 15 355 � 19 349 � 17 368 � 23 356 � 18Mean blood pressure

(mmHg) 91.7 � 6.3 89.3 � 8.0 92.2 � 6.4 90.2 � 5.2 89.4 � 7.0 88.5 � 8.9 93.2 � 6.2 91.4 � 9.0n 5 6 6 6 6 6 5 6

4 Vol. 23 December 2009 GUIGNABERT ET AL.The FASEB Journal � www.fasebj.org

decreased proliferation and increased death of PA-SMCs, we assessed the proliferation/apoptosis bal-ance in our 6 groups of mice. Immunohistochemicalstaining of the PCNA showed increased SMC prolif-eration in distal PA walls of SM22-5-HTT� micetreated with vehicle and studied from d 1 to 21compared to WT mice (P0.001) (Fig. 2A, B). Nosignificant difference was found on d 1 betweenSM22-5-HTT� mice treated with DCA or fluoxetinevs. vehicle. On d 5, however, the number of PCNA�nuclei was significantly decreased in fluoxetine-treated SM22-5-HTT� mice but not in DCA-treatedSM22-5-HTT� mice compared to vehicle-treatedSM22-5-HTT� mice. On d 21, the number of PCNA-positive cells was considerably lower in PA walls fromDCA-treated SM22-5-HTT� mice than in walls fromvehicle-treated SM22-5-HTT� mice but was signifi-cantly higher than in walls from control WT mice.SM22-5-HTT� mice treated with fluoxetine for 21 dalso had smaller numbers of PCNA-positive cells in

PA walls, similar to the numbers in WT mice (Fig. 2A,B). We found no differences in numbers of PCNA-positive cells from d 1 to 21 in control WT micetreated with vehicle, DCA, or fluoxetine and studiedfrom d 1 to 21 (n4 –5, PNS).

When we detected in situ cell death by TUNELlabeling on d 5 in distal PA walls from fluoxetine- orDCA-treated SM22-5-HTT� mice, we found a 2-fold anda 4-fold increase, respectively, in the percentage ofapoptotic cells, compared to vehicle-treated SM22-5-HTT� mice (Fig. 2C).

Kv1.5 expression in lungs of SM22-5-HTT� mice andeffect of DCA or fluoxetine treatments

To study interactions between 5-HTT and Kv1.5 duringpulmonary vascular remodeling in SM22-5-HTT� mice,we measured pulmonary expressions of Kv1.2, 1.5, and2.1 on d 21 of treatment with vehicle, fluoxetine orDCA. No significant differences were found in the

Figure 2. Change in num-ber of PCNA-positive cells indistal PAs in SM22-5-HTT�

and WT mice treated withdaily DCA (80 mg/kg), flu-oxetine (Fluox; 10 mg/kg),

or vehicle (saline) for 1, 5, or 21 d. A) Representative photomicrographs of immunohistochemical staining with anti-PCNAantibody. PCNA-positive SMCs (arrows) are visible in medium of distal PAs. B) Semiquantification of histological findings.Percentage of PCNA-positive nuclei (brown) was estimated in the medium of �10 different distal PAs per lung. C) Quantificationof the percentage of apoptotic cells (TUNEL) in the medium of distal PAs on d 5. Bars represent means � se (n5 mice/group).*P 0.05, **P 0.01, ***P 0.001 vs. WT mice; §P 0.05, §§§P 0.001 vs. vehicle-treated SM22-5-HTT� mice. Scale bar 50 �m.

5EFFICACY OF DCA TREATMENT IN SM22-5-HTT� MICE

mRNA level of Kv1.2 or 2.1 channels in SM22-5-HTT�

mice treated or not, compared to WT mice (n5–6 ineach group, PNS). In contrast, as previously reported(13), Kv1.5 mRNA was decreased in lungs from vehicle-treated SM22-5-HTT� mice vs. WT mice. However, theKv1.5 mRNA level was normal in lungs from SM22-5-HTT� mice treated with DCA or fluoxetine, that is,identical to that in WT mice (n5–6 in each group,PNS) (Fig. 3A).

Consistent with the RTQ-PCR results, no differencesin Kv1.2 protein levels were found in any comparisonsof treated or untreated groups. However, Kv1.5 proteinwas significantly decreased in lungs from vehicle-treated SM22-5-HTT� mice, as confirmed by Westernblot (Fig. 3B) and by immunohistochemistry (Fig. 3C).DCA or fluoxetine treatment normalized the Kv1.5protein level to a value similar to that in WT mice(n5–6 in each group, PNS).

Effect of 5-HTT activation on NFAT nucleartranslocation and on protein expression of Kv1.5 incultured PA-SMCs from IPAH or control patientstreated with DCA, 11R-VIVIT, CsA, or fluoxetine

Previous studies showed that DCA decreased NFATtranslocation in DCA-treated cells and that this effectcorrelated with an increase in Kv1.5 expression com-pared to untreated cells (34, 38). To investigatewhether NFAT served as a link between 5-HTT activa-tion and Kv1.5 down-regulation, we conducted parallelevaluations of the NFAT activation states and Kv1.5protein levels in PA-SMCs from patients with IPAH orfrom normal individuals treated with serotonin (10�6

M) or vehicle (PBS) with or without one of the follow-ing: PBS, DCA (5�10�4 M), fluoxetine (10�6 M),11R-VIVIT (4�10�6 M), and CsA (10�6 M). We notedgreater nuclear translocation of NFATc2 and signifi-cantly lower Kv1.5 protein expression in IPAH PA-SMCs

Figure 3. Change in potassium channel (Kv) expression in lung tissues from SM22-5-HTT� and WT mice treated for 21 d withDCA (80 mg/kg/d) or fluoxetine (Fluox; 10 mg/kg/d) or vehicle (saline). A) Quantitative RT-PCR of Kv1.2, Kv1.5, and Kv2.1mRNA normalized for 18S. B) Representative Western immunoblot of Kv1.2 and Kv1.5 protein in lungs from SM22-5-HTT� andWT mice in each group treated with active drug or vehicle and quantification of the Kv1.2/�-actin and Kv1.5/�-actin ratios.C) Representative immunohistochemistry slide showing Kv1.5 labeling in a distal PA in WT vs. SM22-5-HTT� mice. Strongpositive staining was noted for Kv1.5 in SMCs of distal PA from the WT mouse vs. SM22-5-HTT� mouse. Bars represent means �se (n5–6). *P 0.05 vs. WT mice. Scale bar 50 �m.

6 Vol. 23 December 2009 GUIGNABERT ET AL.The FASEB Journal � www.fasebj.org

than in control PA-SMCs. Moreover, serotonin treat-ment of PA-SMCs induced marked nuclear transloca-tion of NFATc2 and a significant decrease in Kv1.5protein expression. In addition, DCA, 11R-VIVIT, CsA,and fluoxetine markedly inhibited nuclear NFATc2translocation and normalized Kv1.5 in both IPAH andcontrol PA-SMCs, compared to vehicle (Fig. 4A, B).

Effect of DCA, the selective inhibitor of NFATtranslocation 11R-VIVIT, CsA, or fluoxetine onproliferation of PA-SMCs

Because NFAT inhibits both Kv1.5 expression and cellgrowth (34, 38), we assessed the effect of 11R-VIVIT,CsA, DCA, or fluoxetine on [3H]thymidine incorpora-tion of cultured PA-SMCs from IPAH or control pa-tients in the presence or absence of serotonin. Seroto-nin (10�8 to 10�6 M) dose-dependently increased PA-SMC proliferation in culture (EC500.97�10�7 M) andDCA (5.10�4 M), 11R-VIVIT (4.10�6 M), CsA (10�6 M)vs. PBS reduced markedly the growth of PA-SMC fromIPAH or control patients treated with the highest doseof serotonin (10�6 M). The growth-stimulating effect of5-HT was totally abolished by the selective 5-HTTinhibitor fluoxetine (Fig. 4C).

NFAT activation in lungs of SM22-5-HTT� mice andeffects of DCA or fluoxetine

To determine whether the diminished Kv1.5 expres-sion in lungs of SM22-5-HTT� mice was ascribable toNFAT activation, we evaluated the pulmonary expres-sion of the active form of NFAT, the Kv1.5 protein, andthe phosphoERK1/2:ERK1/2 ratio at all study timepoints in lungs of SM22-5-HTT� mice given vehicle,fluoxetine or DCA. With fluoxetine or DCA, Kv1.5protein increased gradually to normal levels, and activeNFAT decreased gradually. These changes were accom-panied by a marked reduction in the phosphoERK1/2:ERK1/2 ratio (Fig. 5).

Evaluation of the efficacy of CsA, an indirectinhibitor of NFAT activation, on spontaneous PAH inSM22-5-HTT� mice

Separate experiments were performed with SM22-5-HTT� and WT mice (5 males and 5 females aged 17wk). The animals were given a daily oral dose of 1mg/kg of the pharmacological calcineurin inhibitor,CsA, or an equal volume of vehicle (saline) for 3 wk,based on a previously established protocol (38).

CsA decreased the pulmonary expression of theactive form of NFAT and increased Kv1.5 protein levelsbut did not affect PAH in SM22-5-HTT� mice. Nodifferences were found for RVSP, RVH, or musculariza-tion of peripheral arteries compared to vehicle-treatedSM22-5-HTT� mice (Fig. 6). In addition, there were nosignificant differences between SM22-5-HTT� micegiven CsA and those given vehicle regarding bodyweight, SAP, or heart rate (n5–6, PNS) (Table 1).

Effect of DCA, CsA, or fluoxetine treatments inSM22-5-HTT� mice on the balance betweenantiapoptotic (Bcl-2) and proapoptotic (Bax) factors

Previous studies by our group (11, 39) and others (20,21, 38, 40, 41) have shown that apoptosis of PA-SMCs isnecessary for the regression of pulmonary vascularchanges and PAH. In addition, it is known that NFATactivation inhibits apoptosis via up-regulation of theBcl-2 expression (38). Therefore, we hypothesized thatDCA treatment of SM22-5-HTT� mice modulated thebalance between antiapoptotic (Bcl-2) and proapop-totic (Bax) factors. We therefore measured proteinlevels of Bax and Bcl-2 in lung tissue from SM22-5-HTT� mice treated for 1, 5, or 21 d with one of thefollowing: vehicle (saline), DCA (80 mg/kg/d), fluox-etine (10 mg/kg/d), or CsA (1 mg/kg/d). As shown byFig. 7, DCA treatment led to a rapid and markeddecrease in Bcl-2 expression. On d 5, compared tovehicle-treated SM22-5-HTT� mice, DCA-treated SM22-5-HTT� had a 43% decrease in the Bcl-2/Bax ratio, andfluoxetine-treated SM22-5-HTT� mice a 10% decrease.No change in this ratio was observed in CsA-treatedSM22-5-HTT� mice compared to vehicle.

DISCUSSION

In the first part of the present study, in mice withSM22�-targeted overexpression of the serotonin trans-porter, we showed that DCA treatment partially re-versed established PAH and that fluoxetine normalizedthe RVSP values, RVH, and pulmonary vascular remod-eling. Reversal of established PAH in DCA- or fluox-etine-treated SM22-5-HTT� mice is related to progres-sive restoration of the balance between apoptosis andproliferation of PA-SMCs in the walls of muscularizeddistal PAs. Consistent with this finding, in vitro DCAexposure of human PA-SMCs from IPAH and controlpatients decreased the growth-stimulating effect of5-HT (10�6 M) by �50 and 30%, respectively. Thiseffect was totally abolished by fluoxetine. In addition,we showed that DCA treatment led to rapid andmarked decreases in Bcl-2 expression and in the Bcl-2/Bax ratio compared to vehicle, suggesting that down-regulation of Bcl-2 by DCA might be an importantmechanism in the reversal of pulmonary vascular re-modeling. In the second part of our study, we obtainedin vitro evidence that 5-HTT activation and Kv1.5 down-regulation were connected (Fig. 8). Finally, althoughan abnormal increase in NFATc2 activation was ob-served in lungs from SM22-5-HTT� mice and wascorrected by chronic DCA administration, NFATC2inhibition alone with the indirect inhibitor of NFATactivation CsA was not sufficient to produce PAHreversal in this model.

The present work confirms and extends previousevidence (13) that mice with SM22�-targeted 5-HTToverexpression, in the absence of environmental stim-uli, exhibit aberrant PA-SMC proliferation, pulmonary

7EFFICACY OF DCA TREATMENT IN SM22-5-HTT� MICE

Figure 4. Effect of 5-HTT acti-vation on nuclear NFAT trans-location, Kv1.5 protein expres-sion, and proliferation of IPAHor CTR PA-SMCs treated withDCA (5�10�4 M), fluoxetine(Fluox; 10�6 M), 11R-VIVIT(4�10

�6M), and CsA (10�6 M).

A) Representative microphoto-graphs showing immunoreactiv-ity for the active form of NFATin PA-SMCs treated with seroto-nin (10�6 M) or vehicle (PBS)with or without one of the fol-lowing: PBS, DCA, fluoxetine,11R-VIVIT, and CsA. B) Repre-

sentative Western immunoblot of Kv1.5 and �-actin protein in PA-SMCs treated with active drug or vehicle andquantification of the Kv1.5/�-actin ratios. C) Amount of [3H]-thymidine incorporation was determined in synchronizedhuman PA-SMCs exposed to serotonin (10�6 M) for 24 h following preincubation (24 h) with DCA, 11R-VIVIT, fluoxetine,or CsA. Bars represent means � se (n3). *P 0.05, **P 0.05, ***P 0.01 vs. vehicle; §P 0.05, §§P 0.01, §§§P 0.001 vs. CTR PA-SMCs.

8 Vol. 23 December 2009 GUIGNABERT ET AL.The FASEB Journal � www.fasebj.org

vascular remodeling, and PAH development. PAH inthese SM22-5-HTT� mice worsened with age and devel-oped despite unaltered 5-HT bioavailability, solely as a

consequence of increased 5-HTT protein expression inPA-SMCs. We confirmed that the mechanism causingthe progressive pulmonary vascular remodeling, and

Figure 5. Expression of the active form of NFAT, Kv1.5 protein, phosphoERK1/2, and ERK1/2 in lung tissue fromSM22-5-HTT� and WT mice treated with daily oral DCA (80 mg/kg), fluoxetine (Fluox; 10 mg/kg), or vehicle (saline) for 1,5, or 21 d. A) Representative western immunoblots of NFAT, Kv1.5, phosphoERK1/2, ERK1/2, and �-actin protein in lungs fromboth groups of mice under each condition. B) Quantification of the Kv1.5:�-actin, NFAT:�-actin, and phosphoERK1/2:ERK1/2ratios. Bars represent means � se (n5). *P 0.05 vs. control SM22-5-HTT� mice.

Figure 6. Evaluation of the efficacyof CsA (1mg/kg/d) or vehicle (sa-line) for 3 wk against the estab-lished PAH in SM22-5-HTT� Mice.A) Representative Western immu-noblot of the active form of NFAT,Kv1.5, and �-actin and quantifica-tion of the NFAT:�-actin andKv1.5:�-actin ratios in lungs fromSM22–5HTT� mice either treatedor not treated with CsA. B) RVSP,Fulton index: RV/LV�S, and per-centage of muscularized distal ar-teries. Bars represent means � se(n4). *P 0.05, **P 0.01,***P 0.001 vs. vehicle-treatedSM22-5-HTT� mice.

9EFFICACY OF DCA TREATMENT IN SM22-5-HTT� MICE

RV hypertrophy in SM22-5-HTT� mice was related tochanges in PA-SMC proliferation in peripheral PAs.Our PCNA immunohistochemical studies showed that

20-wk-old SM22-5-HTT� mice exhibited a 5- to 6-foldincrease in the number of PCNA-positive cells in thewalls of muscularized PAs compared to WT mice (58�3and 9�3%, respectively). We reported similar PCNA-positive cell counts in 20-wk-old SM22-5-HTT� mice ina previous study that showed increasing cell prolifera-tion and worsening PAH with advancing age (13). Thefact that this pulmonary vascular remodeling is inducedby SM22�-targeted overexpression of 5-HTT in theabsence of associated or environmental stimuli explainswhy fluoxetine (a selective 5-HTT inhibitor) is moreeffective in reversing established PAH than DCA. Weshowed that regression of established PAH in fluox-etine-treated SM22-5-HTT� mice was associated withmarked inhibition of PA-SMC proliferatrion and, inter-estingly, with a 2-fold increase in cell death comparedto vehicle-treated SM22-5-HTT� mice. On d 5, fluox-etine-treated SM22-5-HTT� mice had a 10% decreasein the Bcl-2/Bax ratio, compared to vehicle-treatedSM22-5-HTT� mice. These results are consistent withrecent work by Zhai et al. (42) showing that theprotective effect of fluoxetine treatment in a rat modelof monocrotaline-induced PAH is ascribable to amarked decrease in proliferation and to an increase inapoptosis related to diminished expression of the anti-apoptotic proteins Bcl-2 and Bcl-XL. Furthermore, sim-ilar to our results, they noted up-regulation of Kv1.5expression in fluoxetine-treated vs. vehicle-treated ratsgiven monocrotaline (42). Taken together, our datasuggest that restoration of the balance between cellapoptosis and proliferation may reverse establishedPAH.

Chronic DCA treatment of 20-wk-old SM22-5-HTT�

mice partially reversed established PAH. DCA partiallyrestored the balance between cell apoptosis and prolif-eration in muscularized PAs of SM22-5-HTT� micecompared to WT control mice. We found that DCA ledto a rapid and marked decrease in Bcl-2 expression thatmight explain the greater percentage of apoptotic cells

Figure 7. Effect of DCA (80 mg/kg/d), fluoxetine (Fluox; 10mg/kg/d), CsA (1 mg/kg/d), or vehicle (saline) treatmentof SM22-5-HTT� mice for 1, 5, or 21 d on the Bcl-2 and Baxexpressions. A) Representative Western immunoblot of Bcl-2,Bax, and �-actin protein in lungs from SM22-5-HTT� micetreated for 1, 5, or 21 d with one of the following: vehicle,DCA, fluoxetine, or CsA. B) Quantification of the Bcl-2/Baxratios. Bars represent means � se (n4–5). *P 0.05, **P 0.01, ***P 0.001 vs. vehicle-treated SM22–5HTT� mice;§§P 0.01 vs. vehicle-treated SM22-5-HTT� mice.

Figure 8. Diagram of the link between 5-HTTactivation and Kv1.5 down-regulation. Intracellu-lar accumulation of serotonin induces SMC pro-liferation via activation of several intracellularsignal transduction pathways, including reactiveoxygen species (ROS), activation of Rho-kinase(ROCK), leading to phosphorylation and nu-clear translocation of extracellular-regulated ki-nase (ERK)1/2 and to dephosphorylation andnuclear translocation of NFATc2. NFATc2 re-mains in the cytoplasm when phosphorylated.Following intracellular accumulation of seroto-nin, or on activation by calcium (Ca2�), NFAT isdephosphorylated by the phosphatase cal-cineurin. Once in the nucleus, NFATc2 canregulate gene expression in coordination withERK signaling, leading to Kv1.5 down-regulationand deregulation of the balance between apopto-sis and proliferation. In addition, it has beenreported that activation of HIF-1� and c-Jun are2 other major drivers of Kv1.5 down-regulation.

10 Vol. 23 December 2009 GUIGNABERT ET AL.The FASEB Journal � www.fasebj.org

on d 5 in distal PA walls from SM22-5-HTT� mice givenDCA, compared to vehicle (20�5 and 6�3%, respec-tively) or fluoxetine (11�2%). Taken together, theseresults constitute strong evidence that DCA decreasesBcl-2 expression and subsequently induces PA-SMCapoptosis in the walls of muscularized PAs. Moreover, ithas been reported also that DCA decreases the amountof Survivin mRNA and increases the amount of PUMAmRNA in endometrial cancer cell lines (33). Thus,further studies are needed to investigate the linksbetween NFAT activation and decreased expression ofthese antiapoptotic factors. Moreover, DCA vs. vehicletreatment of SM22-5-HTT� mice markedly reduced thePA-SMC proliferation of muscularized PAs (percentageof PCNA� cell, 23�1 vs. 52�6%, respectively, by d 21).Although the vascular changes in SM22-5-HTT� micedo not reflect the complex pulmonary vascular lesionsseen in patients with IPAH, our in vivo results togetherwith those obtained from in vitro cultures of humancells from IPAH or control patients support the possi-bility that DCA may be an interesting agent worthy ofclinical trials in patients with PAH. Michelakis et al. (29)previously reported that DCA not only prevented butalso reversed chronic hypoxia-induced PAH, and Mc-Murtry et al. (21) found that DCA reversed PA remod-eling induced by monocrotaline injection in rats via amechanism involving an increase in the mitochondria-dependent apoptosis/proliferation ratio. In addition,evidence is accumulating that cancer cells share manysimilarities with pulmonary vessel cells in human IPAHand that DCA corrects several of these abnormalities,improving both diseases (34, 43, 44).

DCA up-regulates Kv1.5 via an NFAT-dependentmechanism (34), and there is a molecular link incardiac muscle cells connecting 5-HT signaling, Ca2�/calcineurin signaling activation, and hypertrophy (45).We therefore hypothesized that NFAT is a major mo-lecular link between 5-HTT activity and Kv channeldown-regulation. Here we showed that 5-HT, via5-HTT, decreased Kv1.5 expression by inhibiting thenuclear translocation of NFATc2. Immunolocalizationwith an antibody against NFAT demonstrated that invitro 5-HT treatment of human PA-SMCs induced sig-nificant nuclear translocation of NFATc2, and subse-quently a significant decrease in Kv1.5 protein expres-sion was observed. In addition, although Kv1.5 proteinlevels returned to normal with DCA, 11R-VIVIT, CsA, orfluoxetine (but not with vehicle), marked PA-SMCproliferation occurred in response to serotonin (10�6

M), being more pronounced with cells from IPAHpatients than from controls. Interestingly, a similarmarked increase in the active form of NFAT wasobserved in lungs from SM22-5-HTT� mice. Continu-ous DCA or fluoxetine treatment of SM22-5-HTT�

mice reduced this abnormal NFATc2 activation, andsubsequently a restoration of Kv1.5 protein channeland reduction in PA-SMC proliferation in muscularizedPAs were observed, compared to vehicle-treated SM22-5-HTT� mice. Taken together, these results supportthe hypothesis that serotonin, through its transporter,

activates the NFAT/Kv1.5 axis (Fig. 8). Activation ofHIF-1� is another major driver of Kv1.5 down-regula-tion, and abnormal HIF-1� activation has been noted inPA-SMCs from Fawn-hooded rats and from humanswith PAH (14). In PA-SMCs from Fawn-hooded rats,selective HIF-1� inhibition with an HIF-1� dominantconstruct inhibits HIF translocation and subsequentlyincreases Kv1.5 expression (14). Our NFAT immuno-reactivity studies showed that serotonin exposure ofPA-SMCs from normal subjects or IPAH patients wasfollowed by increased NFAT immunoreactivity in thenucleus, which suggests that this phenomenon wasindependent from the HIF-1� activation state. RecentlyWalczak-Drzewiecka et al. (46) described a functionalconsensus NFAT site at position 728 bp that is evolu-tionarily conserved in the HIF-1� promoter, whichsuggests that NFAT and HIF-1�, alone or via coopera-tive effects, are two key players in Kv1.5 down-regula-tion. In addition, Yu et al. (26) convincingly demon-strated that c-Jun also decreased Kv1.5 expression inPA-SMCs.

Oral administration of CsA, an indirect inhibitor ofNFAT activation, returned the Kv1.5 level to normal inSM22-5-HTT� mice, but no beneficial effects on pul-monary hemodynamics or structure were found. In ourin vitro studies, CsA and 11R-VIVIT had similar effectsin diminishing NFAT activation, normalizing Kv1.5protein levels, and inhibiting PA-SMC proliferation.However, we cannot exclude the possibility that 11R-VIVIT treatment may show greater efficacy on estab-lished PAH in SM22-5-HTT� mice. In vivo, a similarCsA dose given for 2 wk partially reversed monocrotal-ine-induced PAH in rats (38). There are several expla-nations for this discrepancy. First, NFAT regulatesmany cytokines known to be central in the pathogenesisof PAH (47), and inflammation is an important com-ponent of PAH in the monocrotaline-induced PAH ratmodel but not in SM22-5-HTT� mice. Second, in our invitro and in vivo studies, we showed that inhibition ofNFAT activation alone was not sufficient to counteractall the effects induced by serotonin via its transporter.Normalization of Kv1.5 expression in vitro neither com-pletely inhibited PA-SMC proliferation induced by se-rotonin nor completely abolished the differences be-tween IPAH and control PA-SMCs. It is possible that thelack of efficacy of CsA in our model is due to the factthat established PAH in SM22-5-HTT� mice is causedby sustained 5-HTT activation, which involves the acti-vation of several other intracellular pathways (6–9). Incontrast to CsA, DCA has a wide spectrum of effects thatinclude dysfunctions related to abnormal NFAT activa-tion, production of reactive oxygen species, fragmenta-tion and/or hyperpolarization of the mitochondrialreticulum, and changes in the apoptosis/proliferationratio, without apparent toxicity (17, 32, 34). Here weshowed that DCA treatment led to rapid and markeddecreases in Bcl-2 expression and in the Bcl-2/Bax ratiocompared to vehicle, which suggests that down-regula-tion of Bcl-2 by DCA might be an important mechanismin the reversal of pulmonary vascular remodeling in

11EFFICACY OF DCA TREATMENT IN SM22-5-HTT� MICE

SM22-5-HTT� mice and chronic hypoxia- or monocro-taline-induced PAH (21, 29).

CONCLUSIONS

We showed that chronic DCA or fluoxetine administra-tion limits the progression of established PAH in SM22-5-HTT� mice. This was directly related to decrease inRVSP, RVH, and normalization of the balance betweenapoptosis and proliferation of PA-SMCs in the walls ofmuscularized distal PAs, thus supporting the possibilitythat DCA may be an interesting agent worthy of clinicaltrials in patients with PAH. In addition, this studyprovides the first evidence, to our knowledge, thatserotonin, via 5-HTT, decreases Kv1.5 expression byinhibiting nuclear NFATc2 translocation in vitro andin vivo.

The authors thank Michel Hamon for helpful discussions.This research was supported by grants from the FrenchNational Institute for Health and Medical Research (IN-SERM), the Ministere de la Recherche.

REFERENCES

1. Humbert, M., Morrell, N. W., Archer, S. L., Stenmark, K. R.,MacLean, M. R., Lang, I. M., Christman, B. W., Weir, E. K.,Eickelberg, O., Voelkel, N. F., and Rabinovitch, M. (2004)Cellular and molecular pathobiology of pulmonary arterialhypertension. J. Am. Coll. Cardiol. 43, 13S–24S

2. Michelakis, E. D., Wilkins, M. R., and Rabinovitch, M. (2008)Emerging concepts and translational priorities in pulmonaryarterial hypertension. Circulation 118, 1486–1495

3. De Caestecker, M. (2006) Serotonin signaling in pulmonaryhypertension. Circ. Res. 98, 1229–1231

4. MacLean, M. R., Herve, P., Eddahibi, S., and Adnot, S. (2000)5-hydroxytryptamine and the pulmonary circulation: receptors,transporters and relevance to pulmonary arterial hypertension.Br. J. Pharmacol. 131, 161–168

5. Lee, S. L., Wang, W. W., and Fanburg, B. L. (1997) Associationof Tyr phosphorylation of GTPase-activating protein with mito-genic action of serotonin. Am. J. Physiol. 272, C223–230

6. Guilluy, C., Eddahibi, S., Agard, C., Guignabert, C., Izikki, M.,Tu, L., Savale, L., Humbert, M., Fadel, E., Adnot, S., Loirand, G.,and Pacaud, P. (2009) RhoA and rho kinase activation in humanpulmonary hypertension: role of 5-HT signaling. Am. J. Respir.Crit. Care Med. 179, 1151–1158

7. Liu, Y., Suzuki, Y. J., Day, R. M., and Fanburg, B. L. (2004) Rhokinase-induced nuclear translocation of ERK1/ERK2 in smoothmuscle cell mitogenesis caused by serotonin. Circ. Res. 95,579–586

8. Lee, S. L., Simon, A. R., Wang, W. W., and Fanburg, B. L. (2001)H(2)O(2) signals 5-HT-induced ERK MAP kinase activation andmitogenesis of smooth muscle cells. Am. J. Physiol. Lung Cell. Mol.Physiol. 281, L646–L652

9. Lee, S. L., Wang, W. W., and Fanburg, B. L. (1998) Superoxideas an intermediate signal for serotonin-induced mitogenesis.Free Radic. Biol. Med. 24, 855–858

10. Eddahibi, S., Humbert, M., Fadel, E., Raffestin, B., Darmon, M.,Capron, F., Simonneau, G., Dartevelle, P., Hamon, M., andAdnot, S. (2001) Serotonin transporter overexpression is re-sponsible for pulmonary artery smooth muscle hyperplasia inprimary pulmonary hypertension. J. Clin. Invest. 108, 1141–1150

11. Guignabert, C., Raffestin, B., Benferhat, R., Raoul, W., Zadigue,P., Rideau, D., Hamon, M., Adnot, S., and Eddahibi, S. (2005)Serotonin transporter inhibition prevents and reverses mono-crotaline-induced pulmonary hypertension in rats. Circulation111, 2812–2819

12. Marcos, E., Adnot, S., Pham, M. H., Nosjean, A., Raffestin, B.,Hamon, M., and Eddahibi, S. (2003) Serotonin transporterinhibitors protect against hypoxic pulmonary hypertension.Am. J. Respir. Crit. Care Med. 168, 487–493

13. Guignabert, C., Izikki, M., Tu, L. I., Li, Z., Zadigue, P., Barlier-Mur, A. M., Hanoun, N., Rodman, D., Hamon, M., Adnot, S.,and Eddahibi, S. (2006) Transgenic mice overexpressing the5-hydroxytryptamine transporter gene in smooth muscle de-velop pulmonary hypertension. Circ. Res. 98, 1323–1330

14. Bonnet, S., Michelakis, E. D., Porter, C. J., Andrade-Navarro,M. A., Thebaud, B., Bonnet, S., Haromy, A., Harry, G., Moudgil,R., McMurtry, M. S., Weir, E. K., and Archer, S. L. (2006) Anabnormal mitochondrial-hypoxia inducible factor-1alpha-Kvchannel pathway disrupts oxygen sensing and triggers pulmo-nary arterial hypertension in fawn hooded rats: similarities tohuman pulmonary arterial hypertension. Circulation 113, 2630–2641

15. Yuan, X. J. (1995) Voltage-gated K� currents regulate restingmembrane potential and [Ca2�]i in pulmonary arterial myo-cytes. Circ. Res. 77, 370–378

16. Yuan, X. J., Wang, J., Juhaszova, M., Gaine, S. P., and Rubin, L. J.(1998) Attenuated K� channel gene transcription in primarypulmonary hypertension. Lancet 351, 726–727

17. Archer, S. L., Gomberg-Maitland, M., Maitland, M. L., Rich, S.,Garcia, J. G., and Weir, E. K. (2008) Mitochondrial metabolism,redox signaling, and fusion: a mitochondria-ROS-HIF-1alpha-Kv1.5 O2-sensing pathway at the intersection of pulmonaryhypertension and cancer Am. J. Physiol. Heart Circ. Physiol. 294,H570–H578

18. Evans, A. M., Osipenko, O. N., and Gurney, A. M. (1996)Properties of a novel K� current that is active at restingpotential in rabbit pulmonary artery smooth muscle cells.J. Physiol. 496(Pt. 2), 407–420

19. Post, J. M., Gelband, C. H., and Hume, J. R. (1995) [Ca2�]iinhibition of K� channels in canine pulmonary artery. Novelmechanism for hypoxia-induced membrane depolarization.Circ. Res. 77, 131–139

20. McMurtry, M. S., Archer, S. L., Altieri, D. C., Bonnet, S.,Haromy, A., Harry, G., Bonnet, S., Puttagunta, L., and Mich-elakis, E. D. (2005) Gene therapy targeting survivin selectivelyinduces pulmonary vascular apoptosis and reverses pulmonaryarterial hypertension. J. Clin. Invest. 115, 1479–1491

21. McMurtry, M. S., Bonnet, S., Wu, X., Dyck, J. R., Haromy, A.,Hashimoto, K., and Michelakis, E. D. (2004) Dichloroacetateprevents and reverses pulmonary hypertension by inducing pulmo-nary artery smooth muscle cell apoptosis. Circ. Res. 95, 830–840

22. Reeve, H. L., Michelakis, E., Nelson, D. P., Weir, E. K., andArcher, S. L. (2001) Alterations in a redox oxygen sensingmechanism in chronic hypoxia. J. Appl. Physiol. 90, 2249–2256

23. Remillard, C. V., Tigno, D. D., Platoshyn, O., Burg, E. D.,Brevnova, E. E., Conger, D., Nicholson, A., Rana, B. K., Chan-nick, R. N., Rubin, L. J., O’Connor, D. T., and Yuan, J. X. (2007)Function of Kv1.5 channels and genetic variations of KCNA5 inpatients with idiopathic pulmonary arterial hypertension. Am. J.Physiol. Cell Physiol. 292, C1837–1853

24. Young, K. A., Ivester, C., West, J., Carr, M., and Rodman, D. M.(2006) BMP signaling controls PASMC KV channel expressionin vitro and in vivo. Am. J. Physiol. Lung Cell. Mol. Physiol. 290,L841–L848

25. Yuan, J. X., Aldinger, A. M., Juhaszova, M., Wang, J., Conte, J. V.,Jr., Gaine, S. P., Orens, J. B., and Rubin, L. J. (1998) Dysfunc-tional voltage-gated K� channels in pulmonary artery smoothmuscle cells of patients with primary pulmonary hypertension.Circulation 98, 1400–1406

26. Yu, Y., Platoshyn, O., Zhang, J., Krick, S., Zhao, Y., Rubin, L. J.,Rothman, A., and Yuan, J. X. (2001) c-Jun decreases voltage-gated K(�) channel activity in pulmonary artery smooth musclecells. Circulation 104, 1557–1563

27. Cogolludo, A., Moreno, L., Lodi, F., Frazziano, G., Cobeno, L.,Tamargo, J., and Perez-Vizcaino, F. (2006) Serotonin inhibitsvoltage-gated K� currents in pulmonary artery smooth musclecells: role of 5-HT2A receptors, caveolin-1, and KV1.5 channelinternalization. Circ. Res. 98, 931–938

28. Pozeg, Z. I., Michelakis, E. D., McMurtry, M. S., Thebaud, B., Wu,X. C., Dyck, J. R., Hashimoto, K., Wang, S., Moudgil, R., Harry, G.,Sultanian, R., Koshal, A., and Archer, S. L. (2003) In vivo genetransfer of the O2-sensitive potassium channel Kv1.5 reduces

12 Vol. 23 December 2009 GUIGNABERT ET AL.The FASEB Journal � www.fasebj.org

pulmonary hypertension and restores hypoxic pulmonary vasocon-striction in chronically hypoxic rats. Circulation 107, 2037–2044

29. Michelakis, E. D., McMurtry, M. S., Wu, X. C., Dyck, J. R.,Moudgil, R., Hopkins, T. A., Lopaschuk, G. D., Puttagunta, L.,Waite, R., and Archer, S. L. (2002) Dichloroacetate, a metabolicmodulator, prevents and reverses chronic hypoxic pulmonaryhypertension in rats: role of increased expression and activity ofvoltage-gated potassium channels. Circulation 105, 244–250

30. Kim, J. W., Tchernyshyov, I., Semenza, G. L., and Dang, C. V.(2006) HIF-1-mediated expression of pyruvate dehydrogenasekinase: a metabolic switch required for cellular adaptation tohypoxia. Cell Metab. 3, 177–185

31. Papandreou, I., Cairns, R. A., Fontana, L., Lim, A. L., andDenko, N. C. (2006) HIF-1 mediates adaptation to hypoxia byactively downregulating mitochondrial oxygen consumption.Cell Metab. 3, 187–197

32. Michelakis, E. D., Webster, L., and Mackey, J. R. (2008) Dichlo-roacetate (DCA) as a potential metabolic-targeting therapy forcancer. Br. J. Cancer 99, 989–994

33. Wong, J. Y., Huggins, G. S., Debidda, M., Munshi, N. C., and DeVivo, I. (2008) Dichloroacetate induces apoptosis in endome-trial cancer cells. Gynecol. Oncol. 109, 394–402

34. Bonnet, S., Archer, S. L., Allalunis-Turner, J., Haromy, A.,Beaulieu, C., Thompson, R., Lee, C. T., Lopaschuk, G. D.,Puttagunta, L., Bonnet, S., Harry, G., Hashimoto, K., Porter,C. J., Andrade, M. A., Thebaud, B., and Michelakis, E. D. (2007)A mitochondria-K� channel axis is suppressed in cancer and itsnormalization promotes apoptosis and inhibits cancer growth.Cancer Cell 11, 37–51

35. Moudgil, R., Michelakis, E. D., and Archer, S. L. (2006) The roleof k� channels in determining pulmonary vascular tone, oxy-gen sensing, cell proliferation, and apoptosis: implications inhypoxic pulmonary vasoconstriction and pulmonary arterialhypertension. Microcirculation 13, 615–632

36. Eddahibi, S., Hanoun, N., Lanfumey, L., Lesch, K. P., Raffestin,B., Hamon, M., and Adnot, S. (2000) Attenuated hypoxicpulmonary hypertension in mice lacking the 5-hydroxytrypta-mine transporter gene. J. Clin. Invest. 105, 1555–1562

37. Eddahibi, S., Guignabert, C., Barlier-Mur, A. M., Dewachter, L.,Fadel, E., Dartevelle, P., Humbert, M., Simonneau, G., Hanoun,N., Saurini, F., Hamon, M., and Adnot, S. (2006) Cross talkbetween endothelial and smooth muscle cells in pulmonaryhypertension: critical role for serotonin-induced smooth musclehyperplasia. Circulation 113, 1857–1864

38. Bonnet, S., Rochefort, G., Sutendra, G., Archer, S. L., Haromy,A., Webster, L., Hashimoto, K., Bonnet, S. N., and Michelakis,E. D. (2007) The nuclear factor of activated T cells in pulmo-nary arterial hypertension can be therapeutically targeted. Proc.Natl. Acad. Sci. U. S. A. 104, 11418–11423

39. Izikki, M., Guignabert, C., Fadel, E., Humbert, M., Tu, L.,Zadigue, P., Dartevelle, P., Simonneau, G., Adnot, S., Maitre, B.,Raffestin, B., and Eddahibi, S. (2009) Endothelial-derived FGF2contributes to the progression of pulmonary hypertension inhumans and rodents. J. Clin. Invest. 119, 512–523

40. Merklinger, S. L., Jones, P. L., Martinez, E. C., and Rabinovitch,M. (2005) Epidermal growth factor receptor blockade mediatessmooth muscle cell apoptosis and improves survival in rats withpulmonary hypertension. Circulation 112, 423–431

41. Schermuly, R. T., Dony, E., Ghofrani, H. A., Pullamsetti, S.,Savai, R., Roth, M., Sydykov, A., Lai, Y. J., Weissmann, N., Seeger,W., and Grimminger, F. (2005) Reversal of experimental pul-monary hypertension by PDGF inhibition. J. Clin. Invest. 115,2811–2821

42. Zhai, F. G., Zhang, X. H., and Wang, H. L. (2009) Fluoxetineprotects against monocrotaline-induced pulmonary arterial hy-pertension is relevant to induction of apoptosis and up-regula-tion of Kv1.5 channels in rats. [E-pub ahead of print] Clin. Exp.Pharmacol. Physiol. PMID: 19298536

43. Michelakis, E. D., Webster, L., and Mackey, J. R. (2008) Dichlo-roacetate (DCA) as a potential metabolic-targeting therapy forcancer. Br. J. Cancer 99, 989–994

44. Voelkel, N. F., Cool, C., Lee, S. D., Wright, L., Geraci, M. W., andTuder, R. M. (1998) Primary pulmonary hypertension betweeninflammation and cancer. Chest 114, 225S–230S

45. Bush, E., Fielitz, J., Melvin, L., Martinez-Arnold, M., McKinsey,T. A., Plichta, R., and Olson, E. N. (2004) A small molecularactivator of cardiac hypertrophy uncovered in a chemical screenfor modifiers of the calcineurin signaling pathway. Proc. Natl.Acad. Sci. U. S. A. 101, 2870–2875

46. Walczak-Drzewiecka, A., Ratajewski, M., Wagner, W., andDastych, J. (2008) HIF-1alpha is up-regulated in activated mastcells by a process that involves calcineurin and NFAT. J. Immu-nol. 181, 1665–1672

47. Macian, F. (2005) NFAT proteins: key regulators of T-celldevelopment and function. Nat. Rev. Immunol. 5, 472–484

Received for publication February 17, 2009.Accepted for publication July 23, 2009.

13EFFICACY OF DCA TREATMENT IN SM22-5-HTT� MICE