R E S E A R C H A R T I C L E

Drug Repositioning Screening Identifies Etravirine as a PotentialTherapeutic for Friedreich’s Ataxia

Giulia Alfedi, MSc,1 Riccardo Luffarelli, MSc,1 Ivano Condò, PhD,1 Giorgia Pedini, MSc,2 Liliana Mannucci, PhD,2

Damiano S. Massaro, PhD,1 Monica Benini, PhD,1,3 Nicola Toschi, PhD,4,5 Giorgia Alaimo, PhD,1,3 Luca Panarello, MSc,1

Laura Pacini, PhD,2 Silvia Fortuni, PhD,1 Dario Serio, BS,1 Florence Malisan, PhD,1 Roberto Testi, MD1,3 andAlessandra Rufini, PhD1,3*

1Laboratory of Signal Transduction, Department of Biomedicine and Prevention, University of Rome “Tor Vergata”, Rome, Italy2Laboratory of Molecular Neurobiology, Department of Biomedicine and Prevention, University of Rome “Tor Vergata”, Rome, Italy

3Fratagene Therapeutics Srl, Rome, Italy4Medical Physics Section, Department of Biomedicine and Prevention, University of Rome “Tor Vergata”, Rome, Italy

5Department of Radiology, Athinoula A. Martinos Center for Biomedical Imaging and Harvard Medical School, Boston, Massachusetts, USA

ABSTRACT : Background: Friedreich’s ataxia is anautosomal-recessive cerebellar ataxia caused by muta-tion of the frataxin gene, resulting in decreased frataxinexpression, mitochondrial dysfunction, and oxidativestress. Currently, no treatment is available for Frie-dreich’s ataxia patients. Given that levels of residual fra-taxin critically affect disease severity, the main goal of aspecific therapy for Friedreich’s ataxia is to increase fra-taxin levels.Objectives: With the aim to accelerate the developmentof a new therapy for Friedreich’s ataxia, we took a drugrepositioning approach to identify market-available drugsable to increase frataxin levels.Methods: Using a cell-based reporter assay to monitorvariation in frataxin amount, we performed a high-throughput screening of a library containing 853 U.S.Food and Drug Administration–approved drugs.Results: Among the potentially interesting candidatesisolated from the screening, we focused our attention onetravirine, an antiviral drug currently in use as an anti–

human immunodeficiency virus therapy. Here, we showthat etravirine can promote a significant increase in fra-taxin levels in cells derived from Friedreich’s ataxiapatients, by enhancing frataxin messenger RNA transla-tion. Importantly, frataxin accumulation in treated patientcell lines is comparable to frataxin levels in unaffectedcarrier cells, suggesting that etravirine could be thera-peutically relevant. Indeed, etravirine treatment restoresthe activity of the iron-sulphur cluster containing enzymeaconitase and confers resistance to oxidative stress incells derived from Friedreich’s ataxia patients.Conclusions: Considering its excellent safety profilealong with its ability to increase frataxin levels and cor-rect some of the disease-related defects, etravirine repre-sents a promising candidate as a therapeutic forFriedreich’s ataxia. © 2019 International Parkinson andMovement Disorder Society

Key Words: drug repositioning; etravirine; frataxin;Friedreich’s ataxia

Friedreich’s ataxia (FRDA) is a progressive degenera-tive disease that affects mainly the nervous system andheart. It is the most common form of inherited ataxia,with an autosomal-recessive inheritance pattern and an

estimated prevalence of 1 in 50.000 individuals in thewhite population.1,2 Symptoms are unremitting andgenerally appear during puberty, although age of onsetvaries from childhood (2–3 years) to adulthood (after

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -*Correspondence to: Dr. Alessandra Rufini, Laboratory of SignalTransduction, University of Rome “Tor Vergata”, Via Montpellier1, 00133, Rome, Italy; E-mail: rufini@med.uniroma2.it

Funding agencies: This study was supported by Fratagene Therapeu-tics, Telethon, National Ataxia Foundation, University of Rome “Tor Ver-gata“, and the European Research Council.

Relevant conflicts of interests/financial disclosures: FratageneTherapeutics is the owner of a patent application US 62/556,794(Methods for treating Friedreich’s ataxia with etravirine).

Full financial disclosures and author roles may be found in theonline version of this article.

Received: 10 July 2018; Revised: 31 October 2018; Accepted: 5November 2018

Published online 00 Month 2018 in Wiley Online Library(wileyonlinelibrary.com). DOI: 10.1002/mds.27604

Movement Disorders, 2019 1

25 years).3 Gait instability and lack of coordination arethe main symptoms of the disease. Additional symp-toms include dysarthria, areflexia, sensory loss, skeletalabnormalities, and left ventricular hypertrophy, whichis the main cause of premature death.4-6

FRDA is caused by a homozygous hyperexpansion ofGAA triplets (from ≈70 to ≈1,000 triplets) within thefirst intron of the gene coding for frataxin (FXN).7,8 Thistype of mutation reduces transcription of the FXN geneattributed to the formation of “sticky” DNA structuresand epigenetic modifications,9 while maintaining a mini-mal residual amount of frataxin (≈ 10–30%), which isessential for survival during embryonic development.10

Frataxin is synthesized in the cytoplasm as a 210 ami-noacid precursor form and subsequently imported intomitochondria. In eukaryotes, the frataxin maturationprocess is mediated by the action of the MPP (mito-chondrial processing peptidase), which operates asequential proteolysis by generating the intermediate(residues 42–210) and mature (residues 81–210) fra-taxin forms.11 Frataxin is involved in several mitochon-drial activities, such as iron metabolism and regulationof iron-sulfur clusters (ISCs) assembly, adenosine tri-phosphate generation, and oxidative stress control.12

Despite that frataxin is expressed ubiquitously in theorganism, its deficiency primarily affects some regionsof the central and peripheral nervous systems,13,14

heart, skeleton, and endocrine pancreas, causing themain clinical and pathological features of the dis-ease.15,16 Currently, there is no approved treatment tocure FRDA or halt the progression of the disease.17,18

Given that reduced amount of frataxin is the principalcause of the disease, the main goal of a specific therapyfor FRDA is to restore physiological frataxin levels.Several therapeutic approaches under development aimat increasing frataxin levels, using both drug reposition-ing19-22 or drug discovery approaches.23-27

Here, we took a systematic drug repositioning approach.To search for drugs capable to increase frataxin levels, wescreened a U.S. Food and Drug Administration (FDA)-approved library comprising 853 different drugs, using afrataxin-reporter fusion construct in a cell-based assay.From the screening, we isolated 19 compounds that pro-mote frataxin accumulation. These drugs belong to differ-ent pharmacological classes and have various knownmolecular targets. In the following validation steps, consid-ering the life-long nature of FRDA disease, we prioritizedcandidate drugs that are known to be tolerated in achronic treatment regimen. In particular, we focused ourattention on etravirine, an antiretroviral drug currently usedfor the life-long treatment of human immunodeficiency(HIV)-infected patients, including children.28,29 Here, weshow that etravirine can promote a significant increase infrataxin levels in several lymphoblastoid cell lines and in pri-mary fibroblasts derived from FRDA patients, withoutshowing any toxic effect. In these cells, etravirine restores

frataxin to levels that are comparable to levels in cells fromthe unaffected carrier mother, suggesting that etravirinecould be therapeutically relevant. Importantly, etravirinetreatment can rescue some phenotypic defects associatedwith frataxin deficiency, such as defects in the activity ofISC-containing enzymes and sensitivity to oxidative stress.Considering its excellent safety profile along with

its ability to induce accumulation of frataxin proteinand improvement of mitochondrial functions, etravir-ine represents a promising candidate as a therapeuticfor FRDA.

Materials and MethodsScreening

The FDA-approved drug library (Selleckchem screen-ing libraries) that was used in the screening was com-posed of 853 drugs. The library was provided in a totalof ten 96-well plates, with the drugs at a stock concen-tration of 10 mM in dimethyl sulfoxide (DMSO). Forthe screening, HEK-293 cells were transiently trans-fected with pCMV-fxn-ProLabel. Twenty-four hoursafter transfection, cells were plated in a 96-well plate ata concentration of 5 × 104 cells per well. Twenty-fourhours after replating, cells were treated with an individ-ual drug per well, at a final concentration of 10 μM.DMSO treatment was used as negative control, whereasMG132-treated cells were included in each plate andconsidered as internal positive control. After 24 hoursthe ProLabel enzyme fragment complementation assaywas performed, using the PathHunter ProLabel Detec-tion Kit (DiscoverX, Fremont, CA), according to themanufacturer’s instructions. Six technical replicateswere performed for each plate.See also the Supporting Methods.

ResultsFDA-Approved Drug LibraryHigh-Throughput Screening

In order to identify drugs that increase frataxinlevels, we performed a high-throughput screening of alibrary containing 853 FDA-approved drugs. To eval-uate the effect of drugs on frataxin abundance, wegenerated a fusion construct between frataxin and Pro-Label that was used as a reporter in an in-cell assay.The system is based on β-galactosidase enzyme frag-ment complementation.32 The 6-kDa ProLabel tagencodes the inactive α fragment of the β-galactosidaseenzyme. When the Ω subunit of the enzyme is added,together with the substrate, the two subunits combineto form an active enzyme that generates a chemilumi-nescent signal, whose intensity correlates to theamount of frataxin-ProLabel fusion present in thecells. This system allows to evaluate subtle variations

2 Movement Disorders, 2019

A L F E D I E T A L

in frataxin levels in a sensitive and accurate manner.33

To validate this system, cells transfected with frataxin-ProLabel were treated with the proteasome inhibitor,MG132. By blocking proteasome-dependent degrada-tion of ubiquitinated target proteins, MG132 allowsaccumulation of frataxin.34 Indeed, treatment of trans-fected cells with MG132 promotes frataxin accumula-tion, which results in a 5-fold increase in the intensityof the luminescence signal generated by the frataxin-ProLabel fusion protein (not shown). To perform thescreening, HEK293 cells transiently transfected withfrataxin-ProLabel were plated in 96-well plates. Theday after, cells were treated with 10 μM of each indi-vidual drug for 24 hours. After addition of the com-plementary β-galactosidase subunit and the substrate,chemiluminescence emission from each individual wellwas measured by luminometer reading (SupportingInformation Fig. S1A,B). Six replicates were performed

for each drug. From the screening, we isolated 19 can-didate drugs, potentially involved in the regulation offrataxin levels, which were selected for furtheranalysis.The identified candidate drugs were individually vali-

dated for their ability to promote frataxin accumula-tion. We initially tested the selected compounds onHEK293 Flp-In cells stably transfected with frataxin(HEK293-fxn) at 10 μM, which is the concentrationused in the screening. This cell line contains a singlecopy of frataxin complementary DNA (cDNA) inte-grated into the genome and therefore, with relativelylow levels of overexpression, enables the detection ofall frataxin forms. As shown in Figure 1A,B, 13 of the19 selected drugs were indeed able to promote at leasta 2-fold increase in frataxin precursor levels. Frataxinprecursor appears as a doublet in western blot analysis;however, the exact nature of these two forms has never

FIG. 1. Candidate drug validation. (A) HEK 293-fxn cells were treated with 10 μM of each candidate drug for 24 hours. Cell extracts were resolved onsodium dodecyl sulfate polyacrylamide gel electrophoresis and analyzed by western blot analysis with antifrataxin (lower panels) or antitubulin antibody(upper panels). Tub, tubulin; pre, frataxin precursor; int, intermediate frataxin; mat, mature frataxin. (B) The table indicates the name of the candidatedrugs and the relative frataxin precursor levels, as measured by the western blot in (A), quantified as the densitometric ratio between frataxin precursorand tubulin for each lane. Data represent the mean from three different independent experiments. S.E.M, standard error of the mean.

Movement Disorders, 2019 3

E T R A V I R I N E R E S C U E S F R A T A X I N D E F I C I E N C Y

been thoroughly investigated. We observed that most ofthe selected drugs are able to upregulate one of the twobands or both. Some drugs also increase the amount ofintermediate form, whereas no effect on the matureform can be appreciated in this experimental setting(see also below). Notably, three of the selected candi-dates (FDA1, FDA11, and FDA5) promote a strikingupregulation of frataxin precursor. Among these threedrugs, we selected etravirine (FDA11), a non-nucleosidereverse transcriptase inhibitor (NNRT), currently usedas a therapeutic for HIV-1 infection. This drug waschosen for further investigation, taking into account itsminimal impact on cell viability and its potential to beused in a chronic therapy without important sideeffects.28

Etravirine Promotes Accumulationof Frataxin Precursor

To further explore the effect of etravirine on fra-taxin levels, we performed a time-course analysis indifferent cell lines and measured frataxin precursorlevels by western blot. Etravirine is able to consis-tently promote a significant increase in frataxin levelsin all the tested cell lines. As shown in Figure 2A, atthe dose of 10 μM, etravirine promotes a rapid androbust increase in the amount of frataxin precursor inHEK293-fxn, already detectable after 4 hours oftreatment and even higher after 12 hours. Moreover,frataxin precursor accumulation can be detected aswell in nontransfected HEK293, expressing onlyendogenous frataxin (Fig. 2B). Most important, etra-virine induces accumulation of the frataxin precursoralso in the lymphoblastoid cell line, GM16203,derived from a FRDA patient (Fig. 2C). We alsoobserved a concomitant reduction of the intermediateform in the three cell lines tested, whereas no changein amounts of the mature form can be observed. Thiscould be attributed to the large amount of rapidlyaccumulated precursor causing a transient stalling inmitochondrial import and processing. As shown inFigure 3, this is overcome when lower doses (3 μM–

300 nM) of etravirine are used.We then investigated whether the observed increase

in frataxin protein levels is attributed to an upregula-tion of the frataxin, messenger RNA (mRNA), eitherpromoted by an increase in the translation rate of fra-taxin mRNA or by its stabilization. To understandwhether etravirine promotes de novo synthesis of fra-taxin mRNA, we metabolically labeled nascent tran-scripts through the incorporation of 5-ethynyl uridine(EU). Through a click chemistry-based reaction, bio-tin was then conjugated to EU-labeled transcripts,thus allowing selective isolation and analysis of newlysynthesized mRNA. We therefore quantitatively ana-lyzed the relative amount of newly synthesized

frataxin mRNA in etravirine-treated HEK293-fxncells, compared to untreated control cells. As shownin Figure 2D, we did not observe any difference in therelative amount of newly synthesized frataxin mRNAin etravirine-treated HEK293-fxn cells, compared tountreated control cells, whereas, as expected, theHDAC inhibitor Vorinostat, which was used as apositive control, significantly upregulated frataxinmRNA synthesis. We then analyzed the effect of etra-virine on the decay of frataxin transcript observedupon treatment with actinomycin. HEK293-fxn cellswere treated with actinomycin for 4, 8, or 12 hoursin the presence or absence of etravirine and frataxinmRNA was quantitated by reverse-transcription poly-merase chain reaction (RT-PCR). As shown inFigure 2E, the half-life of frataxin mRNA is notaffected by etravirine, indicating that etravirine doesnot promote a stabilization of the frataxin transcript.These data suggest that etravirine might be able toenhance the translation efficiency of frataxin mRNA,thus resulting in increased amount of frataxinprotein.

Etravirine Promotes Mature FrataxinAccumulation in FRDA Cells

In order to allow sufficient time for the precursor tobe converted into the mature functional form of fra-taxin, we analyzed the effect of etravirine after a24-hour treatment on FRDA patient-derived cells. Wetested different doses, ranging from 100 nM to 3 μM,on the lymphoblastoid cell line GM16203 and weobserved that concentrations as low as 300 nM are suf-ficient to promote a significant increase of the interme-diate and mature frataxin forms (Fig. 3A,B andSupporting Information Fig. S2).We chose the minimal effective dose (300 nM) to

evaluate the efficacy of etravirine over a broad range ofcells derived from FRDA patients. In particular, weanalyzed seven lymphoblastoid cell lines derived fromFRDA patients and primary fibroblasts derived fromtwo different FRDA patients. Etravirine promoted anincrease in intermediate and mature frataxin levels inall the cell lines tested upon 24-hour treatment(Fig. 3C–E), whereas it did not affect the levels of othermitochondrial proteins (Supporting InformationFig. S3A–D). Average fold increase in mature frataxinlevels promoted by etravirine treatment is reported inFigure 3E for each lymphoblastoid cell line and fibro-blasts tested. Overall, we observed that etravirine wasable to promote a ~50% increase in mature frataxinlevels within 24 hours in cells derived from FRDApatients. Moreover, the graph indicates that there wasno significant correlation between the GAA tripletrepeats number on the longer allele and the ability toupregulate frataxin in response to etravirine. Results

4 Movement Disorders, 2019

A L F E D I E T A L

of these experiments are also summarized in Support-ing Information Table S1. Importantly, frataxin upre-gulation was maintained in cells treated withetravirine for 96 hours, without showing any toxiceffect (Supporting Information Fig. S4A–C), furthersupporting the possibility of using etravirine as achronic therapy for FRDA.

Etravirine Can Functionally Rescue FrataxinDeficiency in FRDA Cells

In order to prove that the increase in frataxin levelscould be relevant for the purpose of restoring the physio-logical conditions, frataxin levels in etravirine-treatedFRDA cells were compared to the levels observed in cells

FIG. 2. Etravirine promotes frataxin precursor accumulation in different cell types. HEK 293-fxn stably expressing frataxin (A), HEK 293 (B), or FRDAlymphoblast cell line GM16203 (C) were treated with 10 μM of etravirine for the indicated time points. Protein extracts were resolved on sodium dodecylsulfate polyacrylamide gel electrophoresis and analyzed by western blot with antifrataxin antibody and antitubulin as a loading control. Etr, etravirine;tub, tubulin; pre, frataxin precursor; int, intermediate frataxin; mat, mature frataxin. One of three experiments performed for each cell line and givingsimilar results is shown. (D) HEK 293-fxn cells stably expressing frataxin were treated with 10 μM of etravirine, 10 μM of vorinostat, or DMSO for 4 or8 hours. Quantitative analysis of newly synthesized FXN mRNA was performed by metabolic labeling of nascent RNA using the Click-iT Nascent RNACapture Kit (Life Technologies, Carlsbad, CA), followed by biotin conjugation and isolation with streptavidin beads, according to provider instructions.Nascent frataxin transcripts levels were quantitated by real-time PCR relative to the expression of the control genes ACTB, GUSB, and ATP5J. Etr,etravirine; Vor, vorinostat. (E) HEK 293-fxn cells stably expressing frataxin were treated with 100 nM of actinomycin D (open triangles) or 100 nM of acti-nomycin D plus 10 μM of etravirine (solid circles) for the indicated times. Relative levels of total frataxin mRNA were measured by real-time PCR andnormalized on expression levels of three control genes, ATP5J, GUSB, and ACTB, using the ΔΔCt (delta delta cycle threshold) method. The graphshows the average of three independent experiments � S.E.M. S.E.M, standard error of the mean.

Movement Disorders, 2019 5

E T R A V I R I N E R E S C U E S F R A T A X I N D E F I C I E N C Y

FIG. 3. Etravirine increases mature frataxin levels in FRDA lymphoblastoid cell lines and primary FRDA fibroblasts. (A) FRDA lymphoblast cells,GM16203, were treated with the indicated doses of etravirine for 24 hours. Cell extracts were resolved on sodium dodecyl sulfate polyacrylamide gelelectrophoresis (SDS-PAGE) and analyzed by western blot analysis with antifrataxin (lower panel) and antitubulin (upper panel), as a loading control.Int, intermediate frataxin; mat, mature frataxin; tub, tubulin. (B) The graph represents the relative mature frataxin levels measured by densitometric anal-ysis of independent blots and normalized with tubulin levels. Data represent the average of three independent experiments � S.E.M. P values were cal-culated with Student’s t test (*P < 0.05; **P < 0.01). (C) Representative blots for each of the seven FRDA lymphoblastoid cell lines. Each cell line wastreated with 300 nM of etravirine for 24 hours. Cell extracts were resolved on SDS-PAGE and analyzed by western blot analysis with antifrataxin andantitubulin, as a loading control. (D) Representative blots for the primary fibroblasts derived from two FRDA patients. Cells were treated and analyzedas described in (A). Etr, etravirine; mat, mature frataxin; int, intermediate frataxin; tub, tubulin. Average fold increase (� S.E.M.) in mature frataxin levels,normalized with tubulin levels, is indicated below each blot. Data represent results from four independent experiments for each cell line. (E) The graphrepresents the average fold increase in mature frataxin levels observed for each FRDA cell line or primary FRDA fibroblasts (specified in the graph leg-end) upon etravirine treatment. FRDA lymphoblastoid cell lines and FRDA primary fibroblasts were distributed along the X-axis by the size of the lon-gest allele. Densitometric analysis was performed on mature frataxin levels and normalized with tubulin levels. Data represent results from fourindependent experiments for each cell line � S.E.M.; solid line and dashed line indicate, respectively, the average mature frataxin levels in untreatedcontrol cells and the average mature frataxin levels in etravirine-treated cells. P values were calculated with Student’s t test (*P < 0.05; **P < 0.01;***P < 0.001; ****P < 0.0001). S.E.M, standard error of the mean.

6 Movement Disorders, 2019

A L F E D I E T A L

derived from the unaffected carrier mother after 24 and48 hours. As shown in Figure 4A,B, frataxin levels inFRDA cells increased over time during etravirine treatmentand compared nicely to levels in unaffected carrier cells.

These data suggest that frataxin accumulation promotedby etravirine treatment could be therapeutically significant.To further explore the therapeutic potential of etra-

virine, we considered its ability to correct some of the

FIG. 4. Etravirine functionally rescue frataxin deficiency in FRDA patient cells. FRDA lymphoblastoid cell line GM16214 and lymphoblastoid cell lineGM16215 derived from the unaffected carrier mother were treated for 24 (A) or 48 hours (B) with 500 nM etravirine or vehicle alone. Cell extracts wereresolved on sodium dodecyl sulfate polyacrylamide gel electrophoresis and analyzed by western blot with antifrataxin antibody and antitubulin as a loadingcontrol. Etr, etravirine; mat, mature frataxin; tub, tubulin. A representative experiment, of four showing similar results, is shown. (C,D) FRDA lymphoblasotidcell line GM16214 was cultured in the presence of 500 nM etravirine (etr) or vehicle only (contr) for 24 hrs (C) or 48 hrs (D). Lymphoblastoid cell lineGM16215 of derived from the unaffected carrier mother was cultured in the presence of vehicle only (contr). Aconitase activity was measured spectrophoto-metrically at 340 nm as described in the Materials and Methods section. Data represent the average of three (C) or five (D) independent experiments �S.E.M. P values were calculated with Student’s t test (*P < 0.05). (E) Lymphoblasts derived from FRDA patient (GM16214) or from the unaffected carriermother (GM16215) were either treated with 500 nM of etravirine for 24 hours or with vehicle only. Cells were then treated with the indicated doses of H2O2

for 16 hours. The graph shows loss of cell viability upon treatment with different doses of H2O2. Cell viability was measured by the XTT assay as describedin the Materials and Methods section. Data represent results from six independent experiments � S.E.M. P values of etravirine-treated GM16214 versusuntreated GM16214 were calculated with Student’s t test (*P < 0.05; **P < 0.01; ***P < 0.001). S.E.M, standard error of the mean.

Movement Disorders, 2019 7

E T R A V I R I N E R E S C U E S F R A T A X I N D E F I C I E N C Y

defects associated with frataxin loss in cells derivedfrom patients. Frataxin participates in the biogenesis ofISCs in cells; thus, frataxin deficiency is associated witha deficit in the activity of many ISC-containingenzymes,35 such as aconitase.36-38 To evaluate whetherthe increase in frataxin levels promoted by etravirinetreatment is sufficient to restore ISC biogenesis inFRDA patient-derived cells, we measured the activity ofaconitase upon etravirine treatment and compared it tothe activity observed in cells derived from the healthycarrier mother. As shown in Figure 4C,D, etravirinewas able to restore aconitase activity in FRDA cellsafter 48 hours.Importantly, cells derived from FRDA patients have

an impaired antioxidant response and show increasedsensitivity to oxidative stress-induced cell death.39-41

We therefore tested whether, by promoting frataxinaccumulation, etravirine could revert this phenotypeand confer resistance to hydrogen peroxide-mediatedstress. We measured the loss of cell viability induced bydifferent doses of H2O2, and we observed that etravir-ine treatment significantly protected lymphoblastoidpatient cells from H2O2-induced cell death (Fig. 4E).Indeed, cell viability in etravirine-treated patient cellswas comparable to what was observed in cells derivedfrom the heterozygous healthy mother. Importantly, nofurther improvement in cell viability was observed inhealthy control cells upon etravirine treatment. Thesedata indicate that etravirine treatment can significantlycorrect functional defects induced by frataxin deficiencyin FRDA patient-derived cells.

Etravirine Promotes Frataxin mRNA Translationin FRDA Cells

We have shown that frataxin accumulation promotedby etravirine is not mediated by enhanced transcriptionor transcript stabilization of frataxin mRNA (Fig. 2D,E). We therefore considered whether a translationalupregulation was taking place upon etravirine treat-ment. In order to evaluate whether etravirine affects theoverall rate of protein synthesis in FRDA lympho-blasts GM16214, we measured the incorporation ofpuromycin into newly synthesized polypeptides, usingthe Sunset method.42 As shown in Figure 5A,B,whereas cycloheximide reduced translation efficiency,we observed no increase in puromycin incorporationin cells treated with etravirine, suggesting that etravir-ine does not promote a general increase in proteintranslation rate. We then tested the hypothesis thattranslation of frataxin mRNA is specifically enhancedby etravirine treatment. To this aim, we fractionatedcell extracts from etravirine-treated and -untreatedFRDA lymphobalstoid cells into translationally activepolysomes and silent messenger ribonucleoproteins(mRNPs), as previously described,43,44 and analyzed

the distribution of frataxin mRNA in the differentfractions. We observed that the percentage of frataxinmRNA associated with the polysomal fraction wassignificantly increased in cells exposed to etravirinefor 24 hours (Fig. 5C,D), whereas no change in distri-bution of actin or glyceraldehyde 3-phosphate dehy-drogenase mRNA could be observed (SupportingInformation Fig. S5A,B). This suggests that etravirineenhances the translation efficiency of frataxin tran-script, resulting in increased amount of frataxin pro-teins in cells derived from FRDA patients.

Discussion

Despite being the most common form of inheritedataxia, FRDA still lacks an adequate approved treat-ment. Finding a therapy for FRDA is therefore a majorunmet medical need. Traditional drug discoveryapproaches are usually very expensive, and many yearsmay be required for a candidate compound to reachmarket approval. Moreover, classical drug discoveryprocesses are associated with a high risk of failure,throughout the development phases. On the otherhand, drug repositioning, which is finding a new indica-tion for an existing and already approved drug, repre-sents an opportunity for accelerating the discovery ofnew therapeutics and may represent a valuable strategy,particularly in the field of orphan diseases, whereresources are often limited.45 This type of approach hasseveral advantages over the classic drug discovery, suchas reduced development costs and shorter time toapproval. These advantages are attributed to the possi-bility of using existing pharmacokinetic, toxicology,and safety data.46,47 With the aim to bypass the costsand risks associated with traditional drug discoveryapproaches and accelerate the identification of an effec-tive treatment for FRDA, we pursued a drug reposition-ing approach. Given that FRDA is mainly caused byinsufficient frataxin synthesis, our main goal was tofind drugs able to increase frataxin levels. Through theuse of a chemiluminescent reporter in a cell-basedassay, we screened an FDA-approved drugs librarysearching for compounds able to promote frataxinaccumulation. Nineteen candidate compounds poten-tially involved in the control of frataxin levels were iso-lated, of the 853 drugs contained in the library. Uponvalidation of the individual drugs, we confirmed thatmost of the identified hits were indeed able to increasefrataxin levels in different cellular systems. However, toprioritize further experimental validation, we consid-ered various aspects, such as the drugs’ effect on cellviability and the potentiality to be used as a chronictherapy, given the chronic nature of FRDA disease.Among other potentially interesting candidates, wefocused our attention on etravirine.

8 Movement Disorders, 2019

A L F E D I E T A L

Etravirine is an NNRT inhibitor of HIV-1 that hasbeen approved (TN Intelence) for the treatment of HIV-1 since 2008.48 Etravirine blocks the RNA- and DNA-dependent DNA polymerase activities of the HIVreverse transcriptase, by binding directly to viralenzyme in two conformationally distinct modes.49 Thisbinding causes a disruption of the enzyme’s catalyticsite, resulting in inhibition of the enzyme activity.50

Etravirine is safe and well tolerated over a long-term

regimen and is currently prescribed as a life-long treat-ment for HIV-positive patients, starting from 6 years ofage.51

Here, we show, for the first time, that the antiviraldrug, etravirine, is able to promote frataxin accumula-tion in cells derived from FRDA patients, restoringphysiological frataxin levels. Importantly, etravirinetreatment does not show any toxic effect on frataxin-deficient cells and does not affect cell viability at any of

FIG. 5. Etravirine promotes frataxin mRNA translation in FRDA cells. (A) FRDA lymphoblastoid cell line GM16214 was treated for 24 hours with 500 nMof etravirine or vehicle alone. Cells were then treated for 30 minutes with 10 μg/mL of puromycin to label nascent polypeptides. Cycloheximide (60 μM)was added where indicated in the last 15 minutes of treatment. Total cell extracts were resolved on sodium dodecyl sulfate polyacrylamide gel electro-phoresis and analyzed by western blot with antifrataxin antibody, antipuromycin antibody, and antitubulin as a loading control. Etr, etravirine; Chx,cycloheximide; tub, tubulin; mat, mature frataxin; puro, puromycin). (B) Densitometric analysis of the blot in (A). Total protein levels as revealed by anti-puromycin antibody were normalized with tubulin levels. (C,D) Translational efficiency of frataxin mRNA was analyzed in FRDA lymphoblastoid cell lineGM16214 trated as in (A). Extracts were fractionated along a 15% to 50% sucrose gradient. Each gradient was collected in 10 fractions while readingthe absorbance at 254 nm. Frataxin mRNA was amplified by RT-qPCR and quantified as PMP (percentage messenger on polysomes).(C) Representative absorbance profile of a sucrose gradient showing heavy polysomes (fractions 1–4), light polysomes (fractions 5–7), and mRNPs(fractions 8–10). Frataxin mRNA was normalized to the amount of an RNA spike-in control. Bars represent the S.E.M (*P < 0.05; **P < 0.01, Student’st test; n = 4) in each fraction. (D) Histogram showing the percentage of frataxin mRNA on heavy polysomes, light polysomes, and mRNPs. Bars repre-sent the S.E.M (*P < 0.05, 24 hours versus T0, Student’s t test; n = 4). S.E.M, standard error of the mean.

Movement Disorders, 2019 9

E T R A V I R I N E R E S C U E S F R A T A X I N D E F I C I E N C Y

the tested doses. Notably, etravirine does not affect therate of neosynthesis of frataxin mRNA, nor does itimpact on the half-life of frataxin transcript. Theseobservations are consistent with the screening methodthat we adopted to select the candidate drugs. Indeed,the frataxin-reporter fusion that we used for the screen-ing is transiently transfected and expressed from acDNA under the constitutively active CMV promoter.Moreover, because the GAA repeats are located withinan intronic region of the frataxin locus, they are notpresent in the cDNA. Drugs that increase the levels ofthe frataxin-reporter fusion are therefore not expectedto regulate the frataxin endogenous promoter and toact through a GAA repeats-independent mechanism.Indeed, we could show that etravirine is able to pro-mote frataxin accumulation in different cell types, bothin transfected cells expressing frataxin from an exoge-nous construct as well as in primary or immortalizedcells derived from FRDA patients, containing GAArepeats expansion. Importantly, we did not observe anycorrelation between the efficacy of etravirine and thepresence or length of the GAA repeats.Etravirine might function by enhancing the rate of

translation of frataxin mRNA or by preventing the deg-radation of frataxin protein. However, when etravirineeffect was compared to the effect of the general protea-some inhibitor, MG132, which is known to prevent thedegradation of frataxin precursor33,34 and increase itslevels, we observed that etravirine-induced frataxinaccumulation is strikingly stronger than what observedupon MG132 treatment, suggesting that de novo pro-tein synthesis is taking place upon etravirine treat-ment and accounts for the massive and rapidaccumulation of frataxin precursor (data not shown).In line with these preliminary observations, we haveindeed shown that etravirine can specifically enhancethe translation efficiency of frataxin transcript, bypromoting a shift of frataxin mRNA from silent iso-lated ribosomes toward the translationally activepolysomal fraction. Our data indicate also that etra-virine does not promote a general increase in proteintranslation rate, but rather selectively enhance fra-taxin translation. Given that the promiscuous natureof drugs is at the base of drug repositioningapproaches,45,52 it is reasonable to assume that etra-virine might interact with a number of proteins withinthe cell. Interaction of etravirine with one of theeukaryotic translation initiation factors (eIFs) or oneof the eIF-binding proteins might promote a more effi-cient formation of the translation initiation complexor might facilitate its attachment to the mRNA, lead-ing to recruitment of ribosomal subunits.53,54 It isknown that upregulation of specific subsets of genes istranslationally controlled55,56 even in conditionswhere a general translational repression is observed.57

It is therefore possible that, together with frataxin, a

number of genes are similarly regulated by etravirine.Ribosome profiling, combined with RNA sequencingapproaches,58,59 will help in the future to achieve amore thorough analysis of genes that are affected byetravirine treatment. Further studies will also berequired to identify the key translation factors andRNA-binding proteins that mediate etravirine actionon frataxin mRNA.Frataxin-deficient cells have an inefficient mitochon-

drial metabolism, which leads to increased generationof reactive oxygen species, coupled with impaired abil-ity to raise an adequate antioxidant response. Thesefeatures make FRDA patient cells particularly suscepti-ble to oxidative stress.40,60 Importantly, we could alsoshow that frataxin accumulated upon etravirine treat-ment could be functionally relevant, in that it can cor-rect some of the pathogenic consequences of frataxindeficiency in cells derived from FRDA patients, such asdeficiency in ISC-containing enzyme activity andincreased sensitivity to oxidative stress. In conclusion,the well-characterized safety profile61 and very low tox-icity induced by etravirine make this drug a promisingcandidate for the development of a new therapy forFRDA. Our data strongly encourage further evaluationof etravirine as a potential therapeutic for FRDA.

AknowledgmentWe thank Prof. Claudia Bagni (Department of Biomedi-

cine and Prevention, University of Rome “Tor Vergata”,Italy; Department of Fundamental Neuroscience, Univer-sity of Lausanne, Switzerland) for her precious supportwith the polysome profiling and interpretation of the data.

References1. Campuzano V, Montermini L, Molto MD, et al. Friedreic’s ataxia:

autosomal recessive disease caused by an intronic GAA triplet repeatexpansion. Science (New York, NY) 1996;271:1423–1427.

2. Harding AE. Friedreich’s ataxia: a clinical and genetic study of90 families with an analysis of early diagnostic criteria and intrafa-milial clustering of clinical features. Brain 1981;104:589–620.

3. Mateo I, Llorca J, Volpini V, Corral J, Berciano J, Combarros O.GAA expansion size and age at onset of Friedreich’s ataxia. Neurol-ogy 2003;61:274–275.

4. Durr A, Cossee M, Agid Y, et al. Clinical and genetic abnormalitiesin patients with Friedreich’s ataxia. N Engl J Med 1996;335:1169–1175.

5. Parkinson MH, Boesch S, Nachbauer W, Mariotti C, Giunti P. Clin-ical features of Friedreich’s ataxia: classical and atypical phenotypes.J Neurochem 2013;126(Suppl 1):103–117.

6. Koeppen AH, Ramirez RL, Becker AB, et al. The pathogenesis ofcardiomyopathy in Friedreich ataxia. PLoS One 2015;10:e0116396.

7. Campuzano V, Montermini L, Lutz Y, et al. Frataxin is reduced inFriedreich ataxia patients and is associated with mitochondrial mem-branes. Hum Mol Genet 1997;6:1771–1780.

8. Marmolino D. Friedreich’s ataxia: past, present and future. BrainRes Rev 2011;67:311–330.

9. De Biase I, Chutake YK, Rindler PM, Bidichandani SI. Epigeneticsilencing in Friedreich ataxia is associated with depletion of CTCF(CCCTC-binding factor) and antisense transcription. PLoS One2009;4:e7914.

10 Movement Disorders, 2019

A L F E D I E T A L

10. Filla A, De Michele G, Cavalcanti F, et al. The relationship betweentrinucleotide (GAA) repeat length and clinical features in Friedreichataxia. Am J Hum Genet 1996;59:554–560.

11. Condò I, Ventura N, Malisan F, Rufini A, Tomassini B, Testi R. Invivo maturation of human frataxin. Hum Mol Genet 2007;16:1534–1540.

12. Gonzalez-Cabo P, Palau F. Mitochondrial pathophysiology in Frie-dreich’s ataxia. J Neurochem 2013;126(Suppl 1):53–64.

13. Koeppen AH, Morral JA, Davis AN, et al. The dorsal root ganglionin Friedreich’s ataxia. Acta Neuropathol 2009;118:763–776.

14. Koeppen AH, Becker AB, Qian J, Feustel PJ. Friedreich ataxia:hypoplasia of spinal cord and dorsal root ganglia. J NeuropatholExp Neurol 2017;76:101–108.

15. Pastore A, Puccio H. Frataxin: a protein in search for a function.J Neurochem 2013;126(Suppl 1):43–52.

16. Koeppen AH, Mazurkiewicz JE. Friedreich ataxia: neuropathologyrevised. J Neuropathol Exp Neurol 2013;72:78–90.

17. Strawser C, Schadt K, Hauser L, et al. Pharmacological therapeuticsin Friedreich ataxia: the present state. Exp Rev Neurother 2017;17:895–907.

18. Tai G, Corben LA, Yiu EM, Milne SC, Delatycki MB. Progress inthe treatment of Friedreich ataxia. Neurol Neurochir Pol 2018;52:129–139.

19. Tomassini B, Arcuri G, Fortuni S, et al. Interferon gamma upregu-lates frataxin and corrects the functional deficits in a Friedreichataxia model. Hum Mol Genet 2012;21:2855–2861.

20. Sahdeo S, Scott BD, McMackin MZ, et al. Dyclonine rescues fra-taxin deficiency in animal models and buccal cells of patients withFriedreich’s ataxia. Hum Mol Genet 2014;23:6848–6862.

21. Santoro A, Anjomani Virmouni S, Paradies E, et al. Effect of diazox-ide on Friedreich ataxia models. Hum Mol Genet 2018;27:992–1001.

22. Libri V, Yandim C, Athanasopoulos S, et al. Epigenetic and neuro-logical effects and safety of high-dose nicotinamide in patients withFriedreich’s ataxia: an exploratory, open-label, dose-escalationstudy. Lancet 2014;384:504–513.

23. Chutake YK, Lam CC, Costello WN, Anderson MP,Bidichandani SI. Reversal of epigenetic promoter silencing in Frie-dreich ataxia by a class I histone deacetylase inhibitor. Nucleic AcidsRes 2016;44:5095–5104.

24. Codazzi F, Hu A, Rai M, et al. Friedreich ataxia-induced pluripotentstem cell-derived neurons show a cellular phenotype that is correctedby a benzamide HDAC inhibitor. Hum Mol Genet 2016;25:4847–4855.

25. Soragni E, Gottesfeld JM. Translating HDAC inhibitors in Frie-dreich’s ataxia. Expert Opin Orphan Drugs 2016;4:961–970.

26. Rufini A, Cavallo F, Condo I, et al. Highly specific ubiquitin-competing molecules effectively promote frataxin accumulation andpartially rescue the aconitase defect in Friedreich ataxia cells. Neuro-biol Dis 2015;75:91–99.

27. Li L, Shen X, Liu Z, et al. Activation of frataxin protein expressionby antisense oligonucleotides targeting the mutant expanded repeat.Nucleic Acid Ther 2018;28:23–33.

28. Katlama C, Clotet B, Mills A, et al. Efficacy and safety of etravirineat week 96 in treatment-experienced HIV type-1-infected patients inthe DUET-1 and DUET-2 trials. Antivir Ther 2010;15:1045–1052.

29. Nelson M, Hill A, van Delft Y, Moecklinghoff C. Etravirine as aswitching option for patients with HIV RNA suppression: a reviewof recent trials. AIDS Res Treat 2014;2014:636584.

30. Chung N, Zhang XD, Kreamer A, et al. Median absolute deviationto improve hit selection for genome-scale RNAi screens. J BiomolScreen 2008;13:149–158.

31. Birmingham A, Selfors LM, Forster T, et al. Statistical methods foranalysis of high-throughput RNA interference screens. Nat Methods2009;6:569–575.

32. Eglen RM. Enzyme fragment complementation: a flexible highthroughput screening assay technology. Assay Drug Dev Technol2002;1(1 Pt 1):97–104.

33. Benini M, Fortuni S, Condo I, et al. E3 ligase RNF126 directly ubi-quitinates frataxin, promoting its degradation: identification of apotential therapeutic target for Friedreich ataxia. Cell Rep 2017;18:2007–2017.

34. Rufini A, Fortuni S, Arcuri G, et al. Preventing the ubiquitin-protea-some-dependent degradation of frataxin, the protein defective inFriedreich’s ataxia. Hum Mol Genet 2011;20:1253–1261.

35. Ye H, Rouault TA. Human iron-sulfur cluster assembly, cellulariron homeostasis, and disease. Biochemistry 2010;49:4945–4956.

36. Rotig A, de Lonlay P, Chretien D, et al. Aconitase and mitochon-drial iron-sulphur protein deficiency in Friedreich ataxia. Nat Genet1997;17:215–217.

37. Bradley JL, Blake JC, Chamberlain S, Thomas PK, Cooper JM,Schapira AH. Clinical, biochemical and molecular genetic correla-tions in Friedreich’s ataxia. Hum Mol Genet 2000;9:275–282.

38. Condo I, Malisan F, Guccini I, Serio D, Rufini A, Testi R. Molecularcontrol of the cytosolic aconitase/IRP1 switch by extramitochondrialfrataxin. Hum Mol Genet 2010;19:1221–1229.

39. Schulz JB, Dehmer T, Schols L, et al. Oxidative stress in patientswith Friedreich ataxia. Neurology 2000;55:1719–1721.

40. Chantrel-Groussard K, Geromel V, Puccio H, et al. Disabled earlyrecruitment of antioxidant defenses in Friedreich’s ataxia. Hum MolGenet 2001;10:2061–2067.

41. Paupe V, Dassa EP, Goncalves S, et al. Impaired nuclear Nrf2 trans-location undermines the oxidative stress response in Friedreichataxia. PLoS One 2009;4:e4253.

42. Schmidt EK, Clavarino G, Ceppi M, Pierre P. SUnSET, a nonra-dioactive method to monitor protein synthesis. Nat Methods 2009;6:275–277.

43. Zalfa F, Giorgi M, Primerano B, et al. The fragile X syndrome pro-tein FMRP associates with BC1 RNA and regulates the translationof specific mRNAs at synapses. Cell 2003;112:317–327.

44. Napoli I, Mercaldo V, Boyl PP, et al. The fragile X syndrome pro-tein represses activity-dependent translation through CYFIP1, a new4E-BP. Cell 2008;134:1042–1054.

45. Sardana D, Zhu C, Zhang M, Gudivada RC, Yang L, Jegga AG.Drug repositioning for orphan diseases. Brief Bioinform 2011;12:346–356.

46. Durães F, Pinto M, Sousa E. Old drugs as new treatments for neuro-degenerative diseases. Pharmaceuticals (Basel) 2018;11. pii: E44.doi: 10.3390/ph11020044.

47. Shameer K, Readhead B, Dudley JT. Computational and experimen-tal advances in drug repositioning for accelerated therapeutic stratifi-cation. Curr Top Med Chem 2015;15:5–20.

48. Usach I, Melis V, Peris JE. Non-nucleoside reverse transcriptaseinhibitors: a review on pharmacokinetics, pharmacodynamics, safetyand tolerability. J Int AIDS Soc 2013;16:1–14.

49. Das K, Clark AD, Jr., Lewi PJ, et al. Roles of conformational andpositional adaptability in structure-based design ofTMC125-R165335 (etravirine) and related non-nucleoside reversetranscriptase inhibitors that are highly potent and effective againstwild-type and drug-resistant HIV-1 variants. J Med Chem 2004;47:2550–2560.

50. Guillemont J, Pasquier E, Palandjian P, et al. Synthesis of novel dia-rylpyrimidine analogues and their antiviral activity against humanimmunodeficiency virus type 1. J Med Chem 2005;48:2072–2079.

51. de Bethune MP. Non-nucleoside reverse transcriptase inhibitors(NNRTIs), their discovery, development, and use in the treatment ofHIV-1 infection: a review of the last 20 years (1989-2009). AntiviralRes 2010;85:75–90.

52. Hernandez JJ, Pryszlak M, Smith L, et al. Giving drugs a secondchance: overcoming regulatory and financial hurdles in repurposingapproved drugs as cancer therapeutics. Front Oncol 2017;7:273.

53. Livingstone M, Atas E, Meller A, Sonenberg N. Mechanisms gov-erning the control of mRNA translation. Phys Biol 2010;7:021001.

54. Grzmil M, Hemmings BA. Translation regulation as a therapeutictarget in cancer. Cancer research 2012;72:3891–3900.

Movement Disorders, 2019 11

E T R A V I R I N E R E S C U E S F R A T A X I N D E F I C I E N C Y

55. Smit E, Caiment F, Piepers J, Kleinjans JCS, van den Beucken T.Translational regulation is a key determinant of the cellular responseto benzo[a]pyrene. Toxicol Lett 2018;295:144–152.

56. Dai A, Cao S, Dhungel P, et al. Ribosome profiling reveals trans-lational upregulation of cellular oxidative phosphorylationmRNAs during vaccinia virus-induced host shutoff. J Virol 2017;91. pii: e01858-16. doi: 10.1128/JVI.01858-16. Print 2017Mar 1.

57. Morita M, Gravel SP, Chenard V, et al. mTORC1 controls mito-chondrial activity and biogenesis through 4E-BP-dependent transla-tional regulation. Cell Metab 2013;18:698–711.

58. Paolini NA, Moore KS, di Summa FM, Fokkema I, t Hoen PAC,von Lindern M. Ribosome profiling uncovers selective mRNA trans-lation associated with eIF2 phosphorylation in erythroid progeni-tors. PLoS One 2018;13:e0193790.

59. Gobet C, Naef F. Ribosome profiling and dynamic regulation oftranslation in mammals. Curr Opin Genet Dev 2017;43:120–127.

60. Bradley JL,Homayoun S,Hart PE, Schapira AH, Cooper JM. Role of oxi-dative damage in Friedreich’s ataxia.NeurochemRes 2004;29:561–567.

61. Allavena C, Katlama C, Cotte L, et al. Long-term efficacy and safety ofetravirine-containing regimens in a real-life cohort of treatment-experiencedHIV-1-infected patients. Infect Dis (Lond) 2016;48:392–398.

Supporting Data

Additional Supporting Information may be found inthe online version of this article at the publisher’sweb-site.

12 Movement Disorders, 2019

A L F E D I E T A L