P

Gp

AMGa

b

a

ARRA

KDSMEISIFHBE

gsl�shf1gb

r

SB

1h

Pharmacological Research 72 (2013) 9– 24

Contents lists available at SciVerse ScienceDirect

Pharmacological Research

jo ur n al hom epage: www.elsev ier .com/ locate /yphrs

erspective

LPG0492, a novel selective androgen receptor modulator, improves muscleerformance in the exercised-mdx mouse model of muscular dystrophy�

nna Cozzoli a,1, Roberta Francesca Capogrossoa,1, Valeriana Teresa Sblendorioa,aria Maddalena Dinardoa, Catherine Jagerschmidtb, Florence Namourb,iulia Maria Camerinoa, Annamaria De Lucaa,∗

Sezione di Farmacologia, Dipartimento di Farmacia – Scienze del Farmaco, Università degli Studi di Bari “Aldo Moro”, Bari, ItalyGalapagos SASU, Romainville, France

r t i c l e i n f o

rticle history:eceived 23 November 2012eceived in revised form 14 March 2013ccepted 15 March 2013

eywords:uchenne muscular dystrophyelective androgen receptor modulatorsdx mouse model

xercise performancen vivo forelimb forcekeletal musclesometric and eccentric muscle contractionibrosis

a b s t r a c t

Anabolic drugs may counteract muscle wasting and dysfunction in Duchenne muscular dystrophy (DMD);however, steroids have unwanted side effects. We focused on GLPG0492, a new non-steroidal selectiveandrogen receptor modulator that is currently under development for musculo-skeletal diseases such assarcopenia and cachexia. GLPG0492 was tested in the exercised mdx mouse model of DMD in a 4-weektrial at a single high dose (30 mg/kg, 6 day/week s.c.), and the results were compared with those from theadministration of �-methylprednisolone (PDN; 1 mg/kg, i.p.) and nandrolone (NAND, 5 mg/kg, s.c.). Thisassessment was followed by a 12-week dose-dependence study (0.3–30 mg/kg s.c.). The outcomes wereevaluated in vivo and ex vivo on functional, histological and biochemical parameters. Similar to PDN andNAND, GLPG0492 significantly increased mouse strength. In acute exhaustion tests, a surrogate of the 6-min walking test used in DMD patients, GLPG0492 preserved running performance, whereas vehicle- orcomparator-treated animals showed a significant increase in fatigue (30–50%). Ex vivo, all drugs resultedin a modest but significant increase of diaphragm force. In parallel, a decrease in the non-muscle area and

istopathologyiochemical biomarkerslectrophysiology

markers of fibrosis was observed in GLPG0492- and NAND-treated mice. The drugs exerted minor effectson limb muscles; however, electrophysiological biomarkers were ameliorated in extensor digitorumlongus muscle. The longer dose-dependence study confirmed the effect on mdx mouse strength andresistance to fatigue and demonstrated the efficacy of lower drug doses on in vivo and ex vivo functionalparameters. These results support the interest of further studies of GLPG0492 as a potential treatmentfor DMD.

© 2013 Elsevier Ltd. All rights reserved.

Abbreviations: DMD, Duchenne muscular dystrophy; SARM, selective andro-en receptor modulators; PDN, �-methylprednisolone; NAND, nandrolone; s.c.,ub-cutaneous; i.p., intra-peritoneal; DIA, diaphragm; EDL, extensor digitorumongus; GC, gastrocnemius; CK, creatine kinase; LDH, lactate dehydrogenase; TGF-1, transforming growth factor-�1; S.E.M., standard error of the mean; S.D.,tandard deviation; gm, total membrane ionic conductance; Hz50, frequency foralf-maximal tetanic contraction; MT, mechanical threshold; R, rheobase voltage

or contraction; PGC-1�, peroxisome proliferative activated receptor� coactivator- �; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; HPRT1, hypoxanthineuanine phosphoribosyl transferase; IGF-1, insulin-like growth factor-1; BDNF,rain-derived nerve factor.� Perspective articles contain the personal views of the authors who, as experts,eflect on the direction of future research in their field.∗ Corresponding author at: Unità di Farmacologia, Dipartimento di Farmacia –

cienze del Farmaco, Università degli Studi di Bari “Aldo Moro”, Via Orabona 4, 70125ari, Italy. Tel.: +39 080 5442245.

E-mail address: annamaria.deluca@uniba.it (A. De Luca).1 These authors equally contributed to the work.

043-6618/$ – see front matter © 2013 Elsevier Ltd. All rights reserved.ttp://dx.doi.org/10.1016/j.phrs.2013.03.003

1. Introduction

Duchenne muscular dystrophy (DMD) and the homologuedystrophic condition in the mdx mouse model, are caused by X-chromosome gene mutations that lead to the absence of the proteindystrophin [1,2]. Dystrophin is a subsarcolemmal component of amulti-molecular network (the dystrophin–glycoprotein complex,DGC) that ensures a physical linkage between the intracellularcytoskeleton and the extracellular matrix. The widely accepted the-ory is that dystrophin provides mechanical stability to myofibreduring contraction. The absence of dystrophin triggers a complexand still unclear sequence of events that eventually leads to pro-gressive myofibre degeneration, failing regeneration and fibrosis,thus resulting in muscle weakness and wasting [1]. Various key

events in the process have been identified, and many of them arerelated to muscle activity, such as recurrent states of ischemiadue to impaired nitric-oxide induced vasodilatation, alterationsin calcium homeostasis via mechano-sensitive pathways, an early

1 cologic

asdsttbtmqeosdmsibatdf[ritertbgscwcpmwnpgdwmtiahmap

2

oDliTtAeih

0 A. Cozzoli et al. / Pharma

nd self-sustained inflammatory response paralleled by oxidativetress, and an impaired balance between protein synthesis andegradation [3–10]. The complexity of the causal and temporalequence of the pathological events justify the effort expendedowards therapeutic strategies aimed at replacing or correctinghe gene defect, although many obstacles have to be overcomeefore the clinical use of such innovative strategies [11]. Untilhen, pharmacological approaches are compelling for the enhance-

ent of muscle function and reduction of wasting to improve theuality of life of DMD patients [12]. Among the various strategiesxplored, there is a particular interest in anabolic agents, basedn their ability to exert a stimulatory action on muscle mass andtrength while promoting protein synthesis. In this context, variousrugs have been tested with a variable level of success, includingyostatin-blocking agents, �2-adrenoceptor agonists and anabolic

teroids [13–17]. The latter have produced controversial resultsn both pre-clinical and clinical studies. In fact, nandrolone haseen reported to worsen the pathological progression in mdx mice,lthough the androgen-sensitive muscle levator ani was showno present a milder pathological progression compared to theiaphragm [18,19]. Oxandrolone has been observed to exert dif-erential protection in relation to the duration of the treatment20,21]. In addition, the reduction of the expression of several geneselated to the muscle regeneration program has been described,mplying that oxandrolone has the ability to decrease regenera-ion and metabolic muscle demands [22]. Some of the controversialffects observed with anabolic steroids may be related to theirelevant effects on androgen-sensitive tissues other than skele-al muscle, with the possibility that beneficial effects are maskedy the action of the steroids on off-target sites. Selective andro-en receptor modulators (SARM) are non-steroidal compoundsynthesised for more specific actions on bone and skeletal mus-le [23,24]. One such compound is GLPG0492, a small moleculeith remarkably low activity on the prostate compared to mus-

le and bones. Because of its profile, we decided to assess theotential benefits of GLPG0492 in the treadmill-exercised mdxice, as a model of DMD. The first phase of the study was a 4eek-treatment in which 30 mg/kg GLPG0492 was compared withandrolone (5 mg/kg), the classic anabolic steroid, and �-methyl-rednisolone (PDN; 1 mg/kg), the clinically used gold-standardlucocorticoid [12]. In the second phase, a treatment with multipleoses of GLPG0492 (0.3, 3 and 30 mg/kg) was performed up to 12eeks, in order to assess the dose- and time-dependent effects. Aultidisciplinary in vivo and ex vivo approach was used to evaluate

he outcome of the treatment on disease-sensitive indices and clin-cally relevant parameters and to elucidate the drug mechanism ofction on dystrophic muscle. The results suggest that GLPG0492 isighly promising as an adjuvant compound for improving perfor-ance in dystrophic patients, although its complex mechanism of

ction deserves further investigation to better assess its therapeuticotential.

. Methods

The animals housing, all the in vivo experiments and mostf the ex vivo studies were done at the Unit of Pharmacology,epartment of Pharmacy-Drug Sciences, University of Bari (Italy),

egally authorised to these experimental procedures by the Ital-an Minister of Health (Authorisation n.◦ 219/95 – A 19/5/1995).he animal research protocol was designed in conformity with

he Italian Law for Guidelines for Care and Use of Laboratorynimals (D.L. 116/92), and European Directive (2010/63/UE). Thexperimental procedures used respected the standard operat-ng procedures for pre-clinical tests in mdx mice available onttp://www.treat-nmd.eu/research/preclinical/SOPs/.

al Research 72 (2013) 9– 24

2.1. In vivo experiments

2.1.1. Animal groups, treadmill running and drug treatmentA total of 56 mdx male mice of 4–5 weeks of age (C57BL/10ScSn-

Dmdmdx/J; Jackson Laboratories, USA), and homogeneous for bodyweight, fore limb force and exercise resistance were used. Themice, either vehicle or drug treated, underwent a 30 min run-ning on an horizontal treadmill (Columbus Instruments, USA)at 12 m/min, twice a week (keeping a constant interval of2–3 days between each trial), for 4–6 weeks or 12 weeks. Theexercise protocol facilitates the estimation of drug efficacy onpathology-related parameters [25,26] (http://www.treat-nmd.eu/resources/research-resources/dmd-sops). For the 1st study thefollowing groups were used: 13 vehicle-treated mice (7 on cornoil + ethanol and 6 on sterile water), 6 mice treated with GLPG0492at 30 mg/kg s.c.; 5 mice treated with nandrolone at 5 mg/kg s.c. and6 mice treated with PDN i.p. (1 mg/kg). A dose of GLGP0492 in thehigh-effective range on other animal models was chosen to avoidfalse negative data in this first test in dystrophic animals. The treat-ment started one day before the beginning of the exercise protocol,and continued until the day of sacrifice. The dose was formulatedso to inject 0.1 mL/10 g body weight. Vehicle for both GLPG0492and nandrolone was 10% corn oil in 90% ethanol. Particular care inanimal handling and environment was used to avoid any animaldiscomfort and stress during daily injection. However, due to theoily nature of the vehicle, skin bruises of various entities occurred.These did not appear to influence animal well being and muscleforce vs. those without skin reactions; however for the second study(12-week treatment) the drug doses were formulated so to injectsub-cutaneously 0.05 mL/10 g body weight, in order to minimisethe risk. No skin bruises were observed in the 12-week treatment.This second study included the following groups: 7 vehicle-treatedmice and three groups treated with GLPG0492: 5 mice at 0.3 mg/kg;8 mice at 3 mg/kg and 6 mice at 30 mg/kg. The dose scaling waschosen based on the result of the first study and to evaluate dose-and time-dependency of drug effects. In both studies, drugs wereadministered once a day, 6 days a week (Monday to Saturday).The one day-off protocol, suitable for chronic treatment regimen,has been decided for internal organisation. Animals for both the1st and the 2nd study were randomly assigned to each group.Non-exercised (sedentary) mdx and wild-type (wt; C57BL/10ScSn)mice, treated or not with vehicle were left free to move in thecage, without additional exercise and monitored at the same timepoints of exercised counterparts, if needed. Every week all micewere monitored for body weight and forelimb force by means of agrip strength meter (Columbus Instruments, USA); the end of the4th–5th week was considered for statistical analysis [25,26]. Forthe second study the additional time points were the 8th and the12th weeks. At these time-points, as well as at the beginning of theprotocol (time 0), an exercise resistance test on treadmill was alsoperformed. All mice were made running on a horizontal treadmillfor 5 min at 5 m/min, then increasing the speed of 1 m/min eachminute. For each mouse, the total distance run until exhaus-tion was measured. At the end of the 4th and of the 12th week ofexercise/treatment the ex vivo experiments were started [27]. Eval-uation of in vivo parameters was performed in a blind fashion. Dueto the time-consuming nature of some of the ex vivo experimentsno more than one-two animals could be sacrificed per day. Theanimals continued to be exercised/treated until the day of sacrifice.

2.2. In vitro studies

2.2.1. Muscle preparations and organ biopsiesAnimals of 8–10 and 16–18 weeks of age belonging to the differ-

ent groups were anaesthetised with 1.2 g/kg urethane i.p. Extensordigitorum longus (EDL) muscle of one hind limb was removed and

ologic

rTrafcfss

rtdla

2

tn2tdtdScttttailtcepac0d1sdcsl1tledfoetmL(

2m

ow

A. Cozzoli et al. / Pharmac

apidly placed in the recording chamber for isometric recordings.he contralateral EDL muscle was used for the electrophysiologicalecordings. Gastrocnemius (GC) muscle from one side was removednd processed for histology, while the contralateral one was snaprozen in liquid nitrogen and stored at −80 ◦C until use for bio-hemical and real-time PCR analysis. The same procedure was usedor various parts of the left half-side of Diaphragms (DIA), whiletrips of the right half-side were used for isometric tension mea-urements.

Soon after mouse death, different organs were removed andapidly weighed in order to evaluate potential effect of drugreatment on other tissues. Hind limb bones were removed foretermination of morphometry by microCT, while plasma was col-

ected for determination of creatine kinase, lactate dehydrogenases well as testosterone and GLPG0492 level.

.2.2. Isometric contractionEDL muscle and strips of diaphragm were securely tight at

endon insertion and placed in a muscle chamber containing theormal physiological solution (in mM): NaCl 148; KCl 4.5; CaCl2.0; MgCl2 1.0; NaHCO3 12.0; NaH2PO4 0.44 and glucose 5.55, con-inuously gassed with 95% O2 and 5% CO2 (pH = 7.2–7.4). For theiaphragm, the ribbon side was fixed to a 25 g isometric forceransducer (FORT25, WPI, Inc, FL, USA) connected to a TCI 102 trans-ucer interface and an MP 100 acquisition unit (Biopac Systems,anta Barbara, CA, USA), while the opposite end was fixed to ahamber hook. The EDL muscle was connected by one tendon athe dual-mode muscle lever-system transducer (300C dual moderansducer, Aurora Scientific, ON; Canada). An electrical stimula-ion field was obtained with two axial platinum wires connectedo a stimulator (LE 12406, 2Biological Instruments, VA, Italy). Aftern equilibration period (30 min), the preparation was stretched tots optimal length (L0; measured with an external calliper), i.e. theength producing the maximal twitch to a 0.2 ms 40 V pulse. Thenhe stimulation protocol was initiated. Maximal twitch tension andontraction kinetic (time to peak and half relaxation time) werevaluated as the mean values from 5 single twitches elicited byulses of 40 V and 0.2 ms (every 30 s); Maximal tetanic tensionnd frequency for half-maximal activation (Hz50) have been cal-ulated from the force–frequency curve constructed with trains of.2 ms 40 V pulses from 10 to 140 Hz; train duration was 450 ms foriaphragm and 350 ms for EDL muscle. A protocol of 10 tetani at00 Hz (450 ms and 350 ms for diaphragm and EDL, respectively)paced by 5 s intervals allowed to calculate the muscle fatigue as %rop of tension at the 10th pulse vs. the first tetanus. The eccentricontraction has been measured in EDL muscle with a 120 Hz tetanictimulation of 500 ms having a gradual stretch up to 10% during theast 200 ms. The protocol consisted of 10 of such tetani spaced by5 s; the stretch-induced damage was estimated as the % drop ofension between 1st and 10th tetani. Data was collected and ana-ysed using AcqKnowledge software vs. 3.8 (Biopac System). At thend of experiments, samples were removed, cleaned from tendons,ried and weighted. Diaphragm strips had variable mass (rangingrom 7 to 10 mg) due to the preparation procedure. Absolute valuesf tension were normalised by cross sectional area according to thequation sP = P/(Mass/Lf × D) where P is absolute tension; Mass ishe muscle mass from tendon to tendon; D is the density of skeletal

uscle assumed to be 1.06 g/cm3; Lf is determined by multiplying0 by previously determined muscle length to fibre length ratios0.44 for the EDL and 1 for the diaphragm) [27,28].

.2.3. Electrophysiological recordings by intracellular

icroelectrodes

EDL muscles were bathed at 30 ± 1 ◦C in the normal physi-logical solution (see composition above) continuously gassedith 95% O2 and 5% CO2 (pH = 7.2–7.4; 27 ± 1 ◦C). Standard two

al Research 72 (2013) 9– 24 11

intracellular microelectrode current clamp method was used tomeasure the membrane electrical properties of muscle fibres,among which membrane resistance (Rm), according to the cableequation (fibre input resistance of 140 � cm2) [25,26]. The totalmembrane conductance (gm) was calculated as 1/Rm in normalphysiological solution.

The mechanical threshold (MT) was determined in the presenceof tetrodotoxin (3 �M) using a two microelectrode “point” voltageclamp method [25,26]. In brief, the two microelectrodes (spacedabout 50 �m) were inserted into the central region of a superficialfibre, continuously viewed using a stereomicroscope (100× mag-nification). Depolarising command pulses of duration ranging from500 to 5 ms (0.3 Hz) were progressively increased in amplitudefrom the holding potential (H) of −90 mV until visible contraction.The threshold membrane potential (V, in mV) was read on a digi-tal sample-and-hold millivoltmeter at the various pulse durationst (in ms) for each fibre; mean values at each t allowed the construc-tion of a “strength-duration” curve. Rheobase voltage (R, in mV)and the time constant (�) to reach the rheobase was obtained bya non-linear least square algorithm using the following equation:V = (H − R exp (t/�))/(1 − exp (t/�)) [25,26].

2.2.4. Muscle histology and bone micro CTGastrocnemius (GC) muscles and hemidiaphragm were dissec-

ted by surrounding tissue, fixed in paraffin, routinely processedin descending alcohol and then paraffin wax embedded. Six �mtransversally cut sections were stained with haematoxylin–eosinto calculate the percentage of both healthy myofibres withperipheral nuclei (peripherally nucleated fibres) and regener-ating/regenerated myofibres, showing central nuclei (centrallynucleated fibres). Morphometric analysis was performed by usingan Image J analysis software on the entire muscle section, selectingabout 5 fields/section. For each muscle two sequential cross-sections were used; at least three animals per experimental groupwere assessed [26,29]. MicroCT scans of the metaphyseal regionof tibias were performed at an isotropic resolution of 8 �m, usinga threshold of 275 to obtain trabecular bone structural parame-ter (BV/TV) using the Scanco Medical �CT scanner CT (�CT 20;Scanco Medical AG, Bassersdorf, Switzerland). The mean tissue ofthe scanned area was between 0.5 and 1.25 mm from the growthplate in the trabecular bone of the tibial proximal metaphysis.The morphometric measurements were performed blinded to thetreatment for each mouse and validated by a second independentexperimenter, always in a blind fashion.

2.2.5. Determination of transforming growth factor-ˇ1 levelTotal and active TGF-�1 protein was measured by enzyme-

linked immunosorbent assay (ELISA), according to the manufac-turer’s instructions (RandD System, Minneapolis, MN, USA). Briefly,10–20 mg of frozen muscle tissue was homogenised in 500 �L of asolution containing 1% Triton X-100, 20 mM Tris pH 8.0, 137 mMsodium chloride, 10% glycerol, 5 mM ethylendiaminetetraaceticacid and 1 mM phenylmethylsulphonyl fluoride [26,27,29]. TGF-�1levels were expressed as pg of TGF-�1/�g of total protein.

2.2.6. Plasma levels of creatine kinase, lactate dehydrogenase,testosterone and GLPG0492

Blood was collected from ventricular camera soon after ani-mal death in heparin rinsed tubes. The blood was centrifuged at3000 × g for 10 min and plasma was separated. Creatine kinase (CK)and lactate dehydrogenase (LDH) determination was performed bystandard spectrophotometric analysis by using diagnostic kits (Sen-

tinel, Farmalab – Italy). CK and LDH were determined on the sameday of plasma preparation.

Serum testosterone levels were evaluated at day 7 in each groupusing a RIA kit (Orion Diagnostica, Finland). In addition, circulating

1 cologic

ld

2

RsRwag

BpB1((serto(o23sgIt

2

(r[mswaadO(t

R

3

3t

gltwc

Awe

2 A. Cozzoli et al. / Pharma

evels of GLPG0492 were quantified at steady-state at 3 or 5 h post-ose by LC–MS/MS in Galapagos bio-analytical department.

.2.7. Real-time PCRFor each DIA and GC muscle sample, total RNA was isolated with

Neasy Fibrous Tissue Mini Kit (Quiagen) and quantified by using apectrophotometer (ND-1000 NanoDrop, Thermo Scientific). TotalNA (400 ng) was used for reverse transcription. Synthesis of cDNAas performed by using random hexamers (annealed 10 min, 25 ◦C)

nd Superscript II reverse transcriptase (Invitrogen–Life Technolo-ies) incubated at 42 ◦C for 50 min.

Real-time PCR was performed in triplicate using the Appliediosystems Real-time PCR 7300 Fast system. TaqMan hydrolysisrimer and probe gene expression assays were ordered by Appliediosystems with the following assay IDs: Insulin-like growth factor

(encoded by igf1 gene) assay ID: Mm 00439560 m1; Follistatinencoded by Fst gene) assay ID: Mm 00514982 m1, myogeninencoded by myog gene) assay ID: Mm 00446194 m1 and Peroxi-ome proliferator activated receptor � coactivator-1 � (PGC1�;ncoded by ppargc1a gene), assay ID: Mm 01208835 m1. Eacheaction was carried on as singleplex reaction. The setup of reac-ions consisted 8 ng of cDNA, 0.5 �L of primer and probe set, 5 �Lf TaqMan Fast Universal PCR master mix No AmpErase UNG (2×)Applied Biosystems) and H2O nucleotide free for a final volumef 10 �L, under the following PCR conditions: step 1, 95 ◦C for0 s; step 2, 95 ◦C for 3 s; and step 3, 60 ◦C for 30 s; steps 2 and

were repeated 45 times. The results were compared with atandard curve and normalised to expression of the housekeepingene glyceraldehyde-3-phosphate dehydrogenase (gapdh) (assayD Mm 99999915 g1) and hypoxanthine guanine phosphoribosylransferase (hprt1) (assay ID Mm 00446968 m1).

.3. Statistics

All data is expressed as mean ± standard error of the meanS.E.M.). The S.E. estimate for the fitted rheobase (R) values (andelative statistical analysis) were obtained as previously described25,26]. Statistical analysis for direct comparison between two

eans was performed by unpaired Student’s t test. Multipletatistical comparisons between groups were performed by one-ay ANOVA, with Bonferroni t test post hoc correction for

llowing a better evaluation of intra- and inter-group vari-bility and avoiding false positive. The recovery score by therug treatment has been evaluated according to the Standardperating Procedures described in the Treat-NMD web site

http://www.treat-nmd.eu/research/preclinical/DMD SOPs) usinghe following equation:

ecovery score = [mdx treated] − [mdx untreated][wildytpe] − [mdx untreated]

× 100.

. Results

.1. Effect of GLPG0492 on forelimb strength and resistance toreadmill exercise

A time-dependent increase in body weight was observed in allroups in both studies (Figs. 1 and 2). The increase was significantlyower in the PDN-treated mice, agreeing with previous observa-ions [25,30]. In the second study, the group of mdx mice treatedith GLPG0492 at 3 mg/kg showed lower body weight increases

ompared with the other treatment groups (Fig. 2A).

The mouse strength increased age-dependently in all groups.

t time 0 (i.e., 4 weeks of age), all of the mdx mice groups wereeaker than wild-type (wt) mice group. The weakness was exac-

rbated by the exercise protocol in the groups of vehicle-treated

al Research 72 (2013) 9– 24

mdx animals in both studies (Figs. 1 and 2). In agreement withprevious studies, this weakness was less evident in non-exercisedmdx mice (data not shown; [26,30]). Significant protection againstthe 4-week exercise-induced weakness was observed in the groupof mice treated with GLPG0492, nandrolone and PDN (Fig. 1B). Toavoid any possible bias due to individual differences in body weight,the normalised strength was calculated. This analysis showedsignificant protection from GLPG0492 and PDN, whereas slight pro-tection was observed with nandrolone (Fig. 1C). The dose- andtime-dependent study confirmed the ability of the SARM to sig-nificantly protect mouse strength, even at lower doses, and tomaintain efficacy over time (Fig. 2B). In agreement with the firststudy phase, significant protection against exercise-induced weak-ness was observed in the mice treated with GLPG0492 at 3 and30 mg/kg after 4 weeks of treatment. After 8 weeks of treatment,all three doses tested led to a significant increase in the absoluteforelimb strength, and this protection was maintained for up to12 weeks. At this stage, the highest value of forelimb strengthwas found in the 30 mg/kg-treated mice. A similar time-dependenttrend was observed when the strength was normalised to bodyweight (Fig. 2C).

Another index of neuromuscular function is the resistance totreadmill exercise, which is representative of the 6-minute walktested clinically used for DMD patients, and is measured as themaximal distance that each mouse is able to run in an acuteexhaustion test. The test was performed at the beginning of exper-imental section (T0) and at T4, T8 and T12 according to theduration of the treatment. A significant decrease in resistance wasobserved in the untreated exercised group, whereas a differentdegree of protection was observed in the GLPG0492-treated mice(Figs. 1D and 2D). In fact, after 4 weeks of treatment, 30 mg/kgGLPG0492 was the most effective drug to protect against fatiguewhen comparing treated mice running a longer distance to vehicle-treated mice (+62.5%). The recovery score was 26%. In contrast,PND- and nandrolone-treated mice did not show any improvement(Fig. 1D).

In the dose- and time-dependent experiments, we confirmedthat after 4 weeks of treatment, GLPG0492-treated mice couldrun for a greater distance than the untreated group. Lower doses(0.3 and 3 mg/kg) were more effective than 30 mg/kg. This latterdose, however, significantly increased the resistance to exercisewith respect to vehicle-treated counterparts, with a comparable,although greater, efficacy with respect to the first study. The slightlydifferent improvement from 30 mg/kg dose between the two stud-ies was attributed to the expected intergroup variability. After 12weeks of treatment, GLPG0492, at all three doses, significantlyincreased the total distance run. No clear dose relationship couldbe observed on this parameter; the maximal activity was con-stantly maintained over 12 weeks with the 3 mg/kg GLPG0492 dose(Fig. 2D).

3.2. Effect of GLPG0492 on mass of muscles andandrogen-sensitive tissues

Muscle performance can be a direct consequence of the hyper-trophic effect exerted by anabolic drugs. Thus, we first evaluatedthe outcome of drug treatment on the masses of the EDL, soleusand levator ani muscles, taken as normalised values over bodyweight, to evaluate the hypertrophic action of the drug on fast- andslow-twitch and androgen-sensitive muscle phenotypes, respec-tively. The mass of other androgen-sensitive tissues, such as heartand prostate, was also evaluated. The results of both the 1st and

2nd studies are shown in Figs. 3 and 4, respectively. As shown, nosignificant changes by GLPG0492 were observed in any of the mus-cles sampled, which exclude a phenotype-dependent hypertrophicdrug effect in the mdx mice. In contrast, a significant increase in

A. Cozzoli et al. / Pharmacological Research 72 (2013) 9– 24 13

Fig. 1. Effect of 4-week treatment with GLPG0492 and comparators on the in vivo parameters of mdx mice. The figure shows the in vivo parameters at the beginning (T0) andafter 4 (T4) weeks of the protocol for wild-type (Wt) and mdx mice treated either with corn oil (Mdx + V1) or with 30 mg/kg GLPG0492 (Mdx + GLPG0492), 5 mg/kg nandrolone(Mdx + NAND), water (Mdx + V2) or 1 mg/kg �-methylprednisolone (Mdx + PDN). The drugs were given 6 days per week. In each graph, the bars are the means ± S.E.M. from5 to 7 animals. Significant differences between groups were evaluated using the ANOVA test for multiple comparisons and the Bonferroni t-test post hoc correction. In (A),the bars show the body weight values (body weight) in g. No significant differences were observed between the values of mdx mice (either treated or not) using an ANOVAtest. In (B), the bars show the maximal forelimb strength (Forelimb force) in kg. The ANOVA test did not indicate any significant differences at time 0 (T0). A significantdifference was found at time 4 (T4) (F > 5.79; p < 0.005). The post hoc Bonferroni t-test results are indicated as follows: *significantly different with respect to Wt mice withp < 0.003; ◦significantly different with respect to respective vehicle-treated mdx exercised mice with 0.007 < p < 0.01. In (C), the bars show the normalised fore limb forcevalues (Normalised forelimb force) calculated by normalising for each mouse the fore limb strength to the respective body weight. The ANOVA test did not show significantdifference for time 0 (T0). A significant difference was found for time 4 (T4) (F > 5.8; p < 0.006). The post hoc Bonferroni t-test results are as follows: *significantly differentw ◦

to mdr pect ta .

tmhctGowaadHnt3bfd

ith respect to Wt mice with 0.0006 < p < 0.03; significantly different with respectunning in a treadmill exhaustion test. All values are significantly different with ress follows: *significantly different with respect to wt mice with 5.5 × 10−7 < p < 0.01

he levator ani muscle weight was observed in nandrolone-treatedice. No changes were observed in the mass of the prostate and

eart in both the 1st and the 2nd studies. However, a signifi-ant decrease in the normalised prostate weight after 12 weeks ofreatment suggested potential anti-androgen activity at the highestLPG0492 dose (30 mg/kg). As potential target organs, the massesf the liver, spleen, and kidney were measured, and few changesere observed in both studies (Figs. 3B and 4B). In the 1st study,

significant increase in liver weight was observed with all drugs,nd significant increases in the kidney and spleen weights wereetected in the nandrolone and GLPG0492 groups, respectively.owever, after 12 weeks of treatment, a significant increase in kid-ey weight was observed for all three doses of GLPG0492, whereas aendency towards a lower spleen weight was observed at the 3 and

0 mg/kg doses. The differences observed in the 2nd study coulde related to the longer treatment duration and adaptation (i.e.,or the liver and spleen) or exacerbation (kidney) of tissue-specificrug activity.

x exercised mice with 0.003 < p < 0.02. In (D), the total distance (in m) is shown foro wt animals at both T0 and T4. The post hoc Bonferroni t-test results are indicated

3.3. Effect of GLPG0492 on the functional parameters of isolatedmuscles

3.3.1. DiaphragmResistance to exercise may be dependent on the contractile per-

formance of respiratory muscles. Therefore, the decreased exerciseperformance of mdx mice may be due to the serious impairmentin the diaphragms of mdx mice. We tested whether the enhancedresistance to running observed with GLPG0492 correlated withan improvement of diaphragm contractile efficiency. The maximaltwitch and tetanic forces of diaphragm (DIA) strips were measuredby isometric contraction. A significant increase in twitch tensionwas observed in the nandrolone-treated group, while a similartrend was found with GLPG0492; however, all the three drugs

tested in the 1st study significantly increased the DIA tetanic ten-sion after 4 weeks of treatment (Fig. 5A and B). The recovery scoresfor tetanic tension were 35% and 37% for GLPG0492 and nandrolone,respectively.

14 A. Cozzoli et al. / Pharmacological Research 72 (2013) 9– 24

Fig. 2. Dose- and time-dependent effect of GLPG0492 on in vivo parameters of mdx mice. The figure indicates at various time points, from the beginning (T0) up to 12 weeksof protocol (T12), the in vivo parameters of wild-type (Wt) and mdx mice treated with corn oil (Mdx + V1) or with GLPG0492 (Mdx + GLPG0492) at 0.3, 3 and 30 mg/kg.The drugs were given 6 days per week. In each graph, the values, as the means ± S.E.M., from 5 to 8 animals are indicated. The significant differences between groups wereevaluated using an ANOVA test for multiple comparison and the Bonferroni t-test post hoc correction. In (A), the bars indicate the body weight values (Body weight) in g.The ANOVA test did not indicate a significant difference for BW at time 0, time 4 and time 6. A significant difference was found for BW at time 8 (F > 3.9; p < 0.02) and time12 (F > 3.8; p < 0.03). The post hoc Bonferroni t-test results are indicated as follows: *significantly different with respect to Wt mice with 0.006 < p < 0.01 and ◦significantlydifferent with respect to mdx exercised mice with p < 0.05. In (B), the bars indicate the maximal forelimb strengths (forelimb force), in kg at either the beginning (FmaxT0), the 4th (Fmax T4), 8th (Fmax T8) and the 12th (Fmax T12) week of the protocol. The ANOVA test indicated a significant difference at T0 (F > 9; p < 0.0006), T4 (F > 11;p < 0.0002), T8 (F > 3.76; p < 0.02) and T12 (F > 5.4; p < 0.006). The post hoc Bonferroni t-test results are indicated as follows: *significantly different with respect to wt micewith 9.3 × 10−8 < p < 0.02 and ◦significantly different with respect to mdx exercised mice with 3.6 × 10−6 < p < 0.01. In (C), the normalised forelimb force values (normalisedforelimb force) were calculated by normalising, for each mouse, the forelimb strength to the respective body weight. The ANOVA test indicated significant differences at alltime points from T4 onward (F > 4; p < 0.02). The post hoc Bonferroni t-test results are indicated as follows: *significantly different with respect to wt mice with p < 0.0005and ◦significantly different with respect to mdx exercised mice with 0.002 < p < 0.02. In (D), the total distance (in m) run in an exhaustion test on the treadmill is shown.Significant differences between groups were evaluated by ANOVA test and Student’s t test. All values are significantly different with respect to wt animals at correspondingt < 0.02f signifi

so4hitafmc

3

tcf

ime point. A significant difference was found at T4 (F > 5.4; p < 0.007), T8 (F > 4; p

ollows: *significantly different with respect to wt mice with 0.0009 < p < 0.02 and ◦

After 12 weeks of treatment with scaling doses of GLPG0492, aignificant increase in both twitch and tetanic tension was observednly in 0.3 mg/kg GLPG0492-treated mice, with recovery scores of1% and 50%, respectively; a slight increase was observed at theigher doses (Fig. 5C and D). No significant changes were observed

n the contraction kinetics or in other calcium-dependent parame-ers (twitch/tetanus ratio and Hz50) after 4 or 12 weeks of exercisend/or treatment (data not shown). A trend of reducing muscleatigue was observed after 4 weeks in the drug-treated groups pri-

arily with GLPG0492 and PDN treatment, but this trend was notonfirmed after 12 weeks of any of the GLPG0492 doses.

.3.2. Extensor digitorum longus muscle

The contractile profile of the EDL muscle was also monitored

o evaluate the muscle-specific action of GLPG0492 as well as theontribution of limb muscles to the general increase in mouseorce observed in vivo. No significant amelioration was observed

), and T12 (F > 5; p < 0.009). The post hoc Bonferroni t-test results are indicated ascantly different with respect to mdx exercised mice with 0.009 < p < 0.03.

on twitch or tetanic forces in the drug-treated animal groups in thefirst study. In addition, no significant differences were observedbetween experimental groups in the contraction and relaxationtimes, the calcium-dependent parameters and fatigue, although apartial recovery from fatigue was observed with nandrolone andPDN (data not shown). We then evaluated whether lower dosesand/or prolongation of treatment may produce a greater effecton the limb muscles. The prolonged treatment with GLPG0492exerted a significant increase in EDL muscle twitch force at 3 and30 mg/kg, whereas the increase was not significant at the lowestdose (Fig. 6A). An increase in tetanic tension ranging between 10%and 20% for the different doses was also observed; however, theincrease was not significant (Fig. 6B). We then evaluated the abil-

ity of GLPG0492 to protect dystrophic EDL muscle against fatigueand eccentric contraction force drop. As shown in Fig. 6C, a wt-like resistance to fatigue was observed in EDL muscles treated withGLPG0492 at 3 and 30 mg/kg, although the values did not reach the

A. Cozzoli et al. / Pharmacological Research 72 (2013) 9– 24 15

Fig. 3. Effect of 4-week treatment with GLPG0492 and comparators on the weight of androgen-sensitive tissues and other potential target tissues. Each bar represents themean ± S.E.M. from 5 to 10 animals and shows the tissue mass normalised with respect to the individual body weight of mdx mice treated with either vehicle (corn oil andwater; Mdx + VTOT) or with 30 mg/kg GLPG0492 (Mdx + GLPG0492), 5 mg/kg nandrolone (Mdx + NAND) or 1 mg/kg �-methylprednisolone (Mdx + PDN). The drugs were given6 days per week. In (A), the figure shows the weight of androgen-sensitive tissues, i.e., the heart, prostate, levator ani, EDL and soleus muscles. The normalised values for thelevator ani have been scaled by a factor of ten for graphical reasons. The ANOVA analysis and Bonferroni t test indicated significant differences only for the levator ani weight( he figut sis anp

stwwden

Ffaaiwi

F > 4; p < 0.015). ◦Significantly different vs. mdx vehicle-treated (p < 0.05). In (B), the liver have been scaled by a factor of ten for graphical reasons. An ANOVA analy

< 0.04); ◦significantly different vs. mdx vehicle-treated (p < 0.02).

tatistical significance vs. untreated mdx mice, most likely due tohe high inter-individual variability. In addition, partial protectionas observed for the force drop induced by eccentric contraction

hen using GLPG0492 at 0.3 and 3 mg/kg, but not at the highestose tested (Fig. 6D). The analysis of the calcium-dependent param-ters indicated no significant alteration at any dosage, suggestingo effect of the drug on calcium homeostasis.

ig. 4. Dose- and time-dependent effect of GLPG0492 on the weight of androgen-sensitivrom 5 to 8 animals and show the tissue mass normalised with respect to the individual bt 0.3, 3 or 30 mg/kg (Mdx + GLPG0492). The drugs were given 6 days per week. In (A), theni, EDL and soleus muscles. The normalised values for the levator ani have been scaled bndicated significant differences only for prostate weight (F > 12; p < 5.4 × 10−5); ◦signific

eight of the spleen, liver and kidneys. The normalised values for the liver have been scalndicated significant differences only for kidney weight (F > 19; p < 1.9 × 10−6); ◦significan

re shows the weights of the spleen, liver and kidneys. The normalised values ford the Bonferroni t test indicated significant differences only for liver weight (F > 3;

This issue was addressed at the myofibre level by measuring thevoltage threshold for contraction (mechanical threshold; MT) usingelectrophysiological experiments. A shift of MT and rheobase volt-

age towards more negative potentials is a typical hallmark of mdxEDL muscle fibres and is indicative of calcium homeostasis distress,resulting from greater release, slower reuptake or higher basalcytosolic levels [4,25]. The 4-week treatment with GLPG0492 at

e tissues and other potential target tissues. Each bar represents the mean ± S.E.M.ody weight of mdx mice treated with either corn oil (Mdx + V1) or with GLPG0492

figure shows the weights of androgen-sensitive tissues, i.e., heart, prostate, levatory a factor of ten for graphical reasons. An ANOVA analysis and the Bonferroni t testantly different vs. mdx vehicle-treated (p < 1.4 × 10−5). In (B), the figure shows theed by a factor of ten for graphical reasons. An ANOVA analysis and Bonferroni t testtly different vs. mdx vehicle-treated (p < 0.03).

16 A. Cozzoli et al. / Pharmacological Research 72 (2013) 9– 24

Fig. 5. Effect of GLPG0492 on isometric contraction on diaphragm muscle. (A) and (B) list the normalised values of the maximal isometric twitch (sPtw measured in kN/m2)and tetanic tension (sP0 measured in kN/m2) of the diaphragm strips from wt and mdx mice, treated or not, from the first study. The figures list the following groups: wild-type mice (Wt) and mdx mice treated with vehicle (water or corn oil: Mdx + VTOT), 30 mg/kg GLPG0492 (Mdx + GLPG0492), 5 mg/kg nandrolone (Mdx + NAND) or 1 mg/kg�-methylprednisolone (Mdx + PDN). The drugs were given 6 days per week. Each bar is the mean ± S.E.M. for 4–7 animals per group. The significant differences betweengroups were evaluated by ANOVA test for multiple comparison (F values) as follows: F = 3; p < 0.05. The Bonferroni t-test post hoc correction was used to estimate significantdifferences between individual mean values and are indicated as follows: *significant difference vs wt (0.001 < p < 0.05); ◦significant difference vs. Mdx + VTOT (p < 0.01). In (C)and (D), the normalised values of the maximal isometric twitch (sPtw measured in kN/m2) and tetanic tension (sP0 measured in kN/m2) of diaphragm strips from WT and mdxmice, treated or not, belonging to the second study are shown. The figures show the wild-type mice (Wt) and mdx mice treated with vehicle (only corn oil: mdx + V1) or withG er weed ons (Fa 5) and

3tnTGmTvktt

Mtttrrdnat

o

LPG0492 at 0.3, 3 or 30 mg/kg (mdx + GLPG0492). The drugs were given 6 days pifference between groups was evaluated by the ANOVA test for multiple comparisnd the results are indicated as follows: *significant difference vs wt (0.001 < p < 0.0

0 mg/kg led to a significant shift in the potential for fibre contrac-ion towards wt values at all durations of depolarising pulses. Bothandrolone and PDN were less effective than GLPG0492 (Fig. 7A).he fit of the data points indicated that the rheobase voltage ofLPG0492-treated EDL myofibres was similar to that of the wt andore positive than the values obtained with nandrolone and PDN.

his latter compound exerted the same effect observed in pre-ious studies [25,30]. An improvement was also observed in theinetics for reaching the rheobase. In fact, the time constant forhe GLPG0492-treated EDL myofibres was remarkably shorter thanhat of the untreated myofibres (Fig. 7B and C).

The results from increasing the duration of the treatment onT are shown in Fig. 8A–C. The strength-duration curves for the

hree doses of GLPG0492 were all shifted to more positive poten-ials when compared with the untreated groups, suggesting thathe effect was maintained over time. However, a dose-dependentelationship was not observed, suggesting a saturating process orather the occurrence, at higher drug doses and longer treatmenturation, of other effects on calcium-dependent handling mecha-isms that mask the beneficial drug effect on this parameter. Such

mechanism could partially account for the absence of effects onhe calcium-dependent contractile parameters described above.

A similar trend was observed for the effect of the drug treatmentn the decrease of gm (mostly due to a decrease in chloride channel

k. Each bar represents the mean S.E.M. for 4–7 animals per group. The significant values) as follows: F = 3; p < 0.05. A Bonferroni t-test post hoc correction was used◦significant difference vs Mdx + V1 (p < 0.01).

conductance, gCl), a typical cellular hallmark of myofibre sufferancein diaphragm and exercised EDL muscle of mdx mice. In the firststudy, all of the drug treatments resulted in a significant increase ingm, and the effect was particularly evident with GLPG0492, whichproduced effects comparable to those of PDN [25] (Fig. 7D).

In the dose- and time-dependent study, all of the doses ofGLPG0492, after 12 weeks, significantly counteracted the exercise-induced decrease of gm in EDL muscle; however, the effect wasnot greater than the effect found after 4 weeks, and again, a dose-dependent relationship was not observed (Fig. 8D).

3.4. Plasma level of creatine kinase (CK) and lactatedehydrogenase (LDH)

A marked elevation of plasma creatine kinase is a typicaldiagnostic marker of muscular dystrophy, along with the func-tional impairment and the altered histological profile. Similarly,an increase in lactate dehydrogenase is observed and is a sign ofmetabolic distress. As shown in Table 1, no significant ameliora-tion was observed with any of the drugs on CK or LDH. A slight,

but not significant reduction of LDH was observed in GLPG0492-treated mdx mice in the 4-week study. The absence of an effect ofPDN on CK and LDH has been already observed in previous studies[25,30].

A. Cozzoli et al. / Pharmacological Research 72 (2013) 9– 24 17

Fig. 6. Effect of GLPG0492 on isometric and eccentric contraction of extensor digitorum (EDL) muscle. The figure shows the contractile parameters of the isolated EDL musclesfrom wt and mdx mice treated with either corn oil (Mdx + V1) or with GLPG0492 at 0.3, 3 or 30 mg/kg (Mdx + GLPG0492). The drugs were given 6 days per week. In (A),the normalised values for maximal isometric twitch (sPtw measured in kN/m2) are shown. ANOVA test indicated significant differences with F = 4 and p < 0.05. The post hocBonferroni t-test results are indicated as follows: *significant difference vs. wt (p < 0.05) and◦ vs Mdx + V1 (0.005 < p < 0.05). In (B), the normalised values of maximal isometrictetanic tension (sP0 measured in kN/m2) are shown. ANOVA test indicated significant differences with F = 4 and p < 0.03. The post hoc Bonferroni t-test results are indicated asfollows: *significant difference vs wt (0.01 < p < 0.05). In (C), the muscle fatigue, defined as the percentage drop of force at the 10th pulse with respect to the first contraction,is shown. No significant difference was observed as evaluated with ANOVA. A Bonferroni t-test indicated significant differences, and the results are indicated as follows:*significant difference vs wt (p < 0.005). In (D), the percentage of tension reduction during eccentric contraction (calculated as the drop at the 10th pulse vs the tension atthe first eccentric stimulus) is shown. An ANOVA test indicated significant differences with F = 4 and p < 0.02. The post hoc Bonferroni t-test results are indicated as follows:*significant difference vs wt (p < 0.05). Each bar represents the mean S.E.M. for 4–7 anima

Table 1Effect of GLPG0492 on the plasma biomarkers of mdx mice.

CK (U/L) LDH (U/L)

1st studyWt 1448 ± 144 (5) 580 ± 67 (5)Mdx + VTOT 13,805 ± 4129 (10) 5866 ± 1947 (10)Mdx + GLPG0492 30 mg/kg 11,556 ± 1514 (5) 3857 ± 777 (5)Mdx + NAND 5 mg/kg 16,434 ± 2751 (5) 4817 ± 1075 (5)Mdx + PDN 1 mg/kg 17,739 ± 1840 (6) 7124 ± 2275 (6)2nd studyWt 1634 ± 291 (5) 716 ± 376 (5)Mdx + V1 11,438 ± 2059 (7) 3695 ± 559 (7)MDX + GLPG 0492 0.3 mg/kg 15,435 ± 2845 (5) 4977 ± 957 (5)Mdx + GLPG0492 3 mg/kg 16,199 ± 1902 (8) 3666 ± 354 (8)Mdx + GLPG 0492 30 mg/kg 18,847 ± 1766◦ (6) 6962 ± 1495◦ (6)

Columns are as follows: groups of mice used: wild-type C57BL10 mice (Wt) andMdx mice treated with vehicle (water or corn oil; Mdx + VTOT, or only corn oil:Mdx + V1) (1st and 2nd studies); Mdx mice treated with GLPG0492 30 mg/kg(Mdx + GLPG0492), nandrolone 5 mg/kg (Mdx + NAND) or �-methylprednisolone1 mg/kg (Mdx + PDN) (1st study); Mdx mice treated with GLPG0492 0.3 mg/kg(Mdx + GLPG0492 0.3 mg/kg), 3 mg/kg (Mdx + GLPG0492 3 mg/kg) and 30 mg/kg(Mdx + GLPG0492 30 mg/kg) (2nd study). The table shows the plasma levels ofcreatine kinase (CK) and lactate dehydrogenase (LDH) measured by standard spec-trophotometric methods. The values are the means ± SEM. The number of animalsis indicated in brackets. For each parameter, the significant differences betweengroups were evaluated using the ANOVA test for multiple comparisons (F values) andthe Bonferroni t-test post hoc correction. No significant difference was observed byANOVA. The post hoc Bonferroni t-test results are indicated as follows: ◦significantlydifferent with respect to mdx exercised mice with 4.9 × 10−9 < p < 0.01.

ls per group.

In line with these observations, no effect on both CK and LDHwas observed at the end of 12 weeks of treatment with any of theGLPG0492 doses (Table 1).

3.5. Effect of GLPG0492 on muscle histology, markers of fibrosisand bone morphology

Representative pictures of the histology profile of the mdxdiaphragm and GC muscles for the various experimental conditionsare shown in Fig. 9. Both muscles clearly presented the typical dys-trophic features, such as the alteration of the muscle architecture,with the presence of areas of necrosis, infiltrates and a large non-muscle area, likely due to the deposition of fibrotic and adiposetissue. A large variability in fibre size and the presence of centro-nucleated fibres (CNF), isolated or in clusters near necrotic fibres,were also present along with other infiltrates, such as mononuclearinflammatory cells. The alterations were still present in the groupsof treated muscles, showing a very high inter-individual variability(see as example Fig. 9), which barely allowed for the estimation ofthe qualitative signs of amelioration.

However, with respect to the first study, the morphometricanalysis indicated a significant decrease in the total area of mus-cle damage (necrosis + non muscle tissue) in the diaphragms ofGLPG0492- and nandrolone-treated mice but not in PDN-treated

ones. Interestingly, this result was entirely due to a reduction inthe non-muscle area (Fig. 10A). In contrast, minor effects of thetreatments were observed on GC, including a slight reduction ofnecrosis by nandrolone and PDN (from 9.0 ± 5% of vehicle treated

18 A. Cozzoli et al. / Pharmacological Research 72 (2013) 9– 24

Fig. 7. Effect of 4-week treatment with GLPG0492 and comparators on functional cellular parameters in extensor digitorum longus (EDL) muscle fibres of mdx mice. In (A),the data, expressed as the means ± S.E.M. from 14 to 30 values from 2 to 5 preparations, show the voltages for the contraction of EDL myofibres (mechanical threshold) atincreasing pulse duration in wild-type mice (WT, black circles) and in mdx mice treated with either vehicle (corn oil and water; Mdx + VTOT, white circles), 30 mg/kg GLPG0492(white triangles), 5 mg/kg nandrolone (upside-down black triangles) or 1 mg/kg PDN (white rhombus). The drugs were given 6 days per week. The voltage threshold valuesof myofibres of mdx mice treated with 30 mg/kg GLPG0492, 5 mg/kg nandrolone or 1 mg/kg PDN were significantly more positive with respect to those of mdx mice treatedwith vehicle (p < 0.03 or less by Student’s t test) at each pulse duration. For some data points, the standard error bar is not visible because it is smaller than the symbol size.In (B) and (C), the rheobase voltage, in mV and the time constant, in ms, with relative standard errors, have been calculated from the fit of data points of the voltage-durationcurves in A, using the equation described in Section 2. In (D), the total resting membrane ionic conductances (gm) in �S/cm2 of EDL muscle fibres of the same experimentalgroups described in A are shown. The bars represent the means ± SEM from the number of 3–5 prep/25–37 fibres. For each parameter, the significant differences betweengroups were evaluated using ANOVA for multiple comparisons (F values) and the Bonferroni t-test post hoc correction. Significant differences were found for rheobase voltage(F > 4; p < 0.003) and gm (F > 7; p < 0.0002). The post hoc Bonferroni t-test results are indicated as follows: *significantly different with respect to wt mice with p < 0.05 and◦

tw

obTmtcbiogNd

significantly different with respect to mdx exercised mice with p < 0.02.

o 3.2 ± 0.8% and 3.4 ± 1%, respectively), which was not observedith GLPG0942.

After 12 weeks of treatment, a blinded qualitative analysis basedn arbitrary units did not highlight any marked improvements foroth muscles following GLGP0492-treatment at any dose (Fig. 9).hese results were also confirmed with a quantitative morpho-etric analysis (data not shown). Therefore, the decrease in the

otal area of damage observed with the short-treatment durationould not be confirmed in DIA and/or GC muscle. However, a slightut significant decrease of centro-nucleated fibres was observed

n the diaphragms of mice treated with 0.3 mg/kg (from 48 ± 2.4%

f vehicle treated to 41 ± 1.8%; p < 0.025 by Student’s t test), sug-esting a potential reduction of degeneration–regeneration cycles.o remarkable changes were observed in this parameter at otheroses in DIA and GC muscles.

Due to the reduction in non-muscle area observed in thediaphragm after the 1st treatment, the level of the pro-fibrotic TGF-�1 protein were measured using an enzyme-linkedimmunosorbent assay (ELISA) in parallel with the total protein con-tent in diaphragm muscle.

The classic Bradford assays indicated an increase of total pro-tein content in the treated muscles from both the 1st and the 2ndstudy. For the latter, the 0.3 mg/kg treatment showed the great-est increase (total protein content of 9487 ± 2787 �g/mL; n = 5,vs. 3940 ± 251 �g/mL; n = 5, corn-oil treated mice) followed by3 mg/kg with total protein content of 6048 ± 785 �g/mL (n = 5);

30 mg/kg showed only a modest increase (5085 ± 342 �g/mL;n = 4). These results corroborate the anabolic action of GLPG0492on dystrophic muscle, with lower doses already exerting a ceilingeffect. Fig. 10B shows the effect of drug treatment on the TGF-�1

A. Cozzoli et al. / Pharmacological Research 72 (2013) 9– 24 19

Fig. 8. Dose- and time-dependent effect of GLPG0492 on functional cellular parameters in extensor digitorum longus (EDL) muscle fibres of mdx mice. In (A), the data,expressed as the means ± S.E.M. from 27 to 41 values from 3 preparations, show the voltages for the contraction of EDL myofibres (mechanical threshold) at increasing pulseduration in wild type mice (WT, black circles) and in mdx mice treated with either corn oil (Mdx + V1, white circles) or GLPG0492 at 0.3 (white square), 3 (black square)or 30 mg/kg (white triangles). The drugs were given 6 days per week. The voltage threshold values of the myofibres of mdx mice treated with GLPG0492 at any dose weresignificantly more positive with respect to those of mdx mice treated with vehicle (p < 0.01 or less by Student’s t test). For some data points, the standard error bar is notvisible because it is smaller than the symbol size. In (B) and (C), the rheobase voltages, in mV and time constant, in ms, with relative standard errors, respectively, have beencalculated from the fit of data points of the voltage-duration curves in A, using the equation described in Section 2. In (D), the total resting membrane ionic conductances(gm) in �S/cm2 of EDL muscle fibres of the same experimental groups described in A are shown. The bars represent the means ± SEM from the values of 2–3 prep/21–41fibres. For each parameter, the significant differences between groups were evaluated using an ANOVA test for multiple comparisons (F values) and the Bonferroni t-test posth 03) af nd ◦si

lswrt(p

sooo

3

pioI

oc correction. Significant differences were found for rheobase voltage (F > 4; p < 0.0ollows: *significantly different with respect to wt mice with 1.1 × 10−13 < p < 0.02 a

evels. After 4 weeks of treatment, both GLPG0492 and nandroloneignificantly reduced the total and active TGF-�1 levels. This effectas maintained over 12 weeks, and an inverse dose-dependent

elationship was observed, with 0.3 mg/kg being the most effec-ive dose. A similar trend was observed for the active TGF-�1 levelFig. 10C), again corroborating the ceiling effect observed for somearameters with low doses of GLPG0492.

The potential outcome of drug treatment on the trabecular bonetructural parameter (BV/TV) values was assessed after 12 weeksf treatment. The exercised mdx mice presented a 30% decreasef BV/TV vs. wt; however, no significant effect of GLPG0492 wasbserved (data not shown).

.6. Plasma level of drug and testosterone

Due to the limited dose-dependent effect observed on various

arameters, the plasma levels of GLPG0492 were determined dur-

ng the long-term treatment. As shown in Fig. 11, the plasma levelsf GLPG0492 were in the range expected from a single acute dose.nterestingly, a dose-dependent relationship was observed.

nd gm (F > 22; p < 1.3 × 10−6). The post hoc Bonferroni t-test results are indicated asgnificantly different with respect to mdx exercised mice with p < 1 × 10−6.

We also evaluated the possible impact of the treatment ontestosterone levels. First, we estimated the basal differencesbetween wt and mdx mice at 8 weeks of age and the impact of4 weeks of exercise on those differences. As shown in Fig. 12A, aslightly higher value of testosterone was observed in mdx vs. wtmice and in the exercised groups vs. the non-exercised groups.These differences were not significant. The 12-week treatment withGLPG0492 did not lead to significant changes in the testosteroneplasma levels at doses of 0.3 or 3 mg/kg. A slight reduction wasobserved at 30 mg/kg, but the reduction did not reach significance,likely because of the great inter-individual variability (Fig. 12B).

3.7. Effect of GLPG0492 on target-gene expression

To elucidate the mechanism of action of GLPG0492 ondystrophic muscle, gene expression analysis was performed.

A few genes were selected as potential targets of SARMaction based on the available evidence. In particular, insulin-like growth factor-1 (IGF-1) and follistatin, genes involvedin control of muscle mass; myogenin, a marker of muscle

20 A. Cozzoli et al. / Pharmacological Research 72 (2013) 9– 24

Fig. 9. Effect of GLPG0492 on the histological profile of mdx mouse muscles. Samples of haematoxylin–eosin staining showing the morphological profiles of diaphragm (DIA)and gastrocnemius (GC) muscles from mdx mice either untreated (Vehicle) or treated with GPL0492 at different dosages (0.3, 3, and 30 mg/kg). The drugs were given 6 daysp at thd s accon

r(m

frtbgNdat

4

euTct

er week. For qualitative comparison, a typical profile of a wt GC muscle is shownystrophic muscle, with great variability in fibre dimension, large areas of necrosion-muscle tissue are also visible. The images are at 20× magnification.

egeneration; and peroxisome-proliferator-receptor �-coactivatorPGC)-1� (ppargc1˛), a modulator of muscle metabolism and of

echano-transduction signalling were chosen.The analysis focused on the two doses, 0.3 mg/kg and 3 mg/kg,

or which the effects on other parameters eliminated the occur-ence of non-specific/toxic effects. The values were normalised tohe levels of the GADPH and HPRT1 housekeeping genes; however,ecause no differences were observed with the two housekeepingenes, only the effects normalised for GADPH are shown in Fig. 13.o significant changes were observed in both gastrocnemius andiaphragm muscles, although slight increases in IGF-1, myogeninnd PGC-1� were observed in the 0.3 mg/kg-treated group in gas-rocnemius muscle.

. Discussion

The present study is the first in vivo and ex vivo pre-clinicalvaluation of GLPG0492, a novel selective androgen receptor mod-

lator (SARM), on a mouse model of Duchenne muscular dystrophy.he rationale for the study was the expected benefit of an anabolicompound to counteract the muscle wasting that occurs in dys-rophic conditions without the unwanted side effects typical of

e top of the figure. The sections clearly show the poorly homogenous structure ofmpanied by mononuclear infiltrates and/or small regenerating fibres. The areas of

anabolic steroids [18–22]. Our studies show a clear improvement inthe in vivo performance of GLPG0492-treated mdx mice in terms ofboth increases in animal strength and resistance to exercise fatigue,which predict a potential benefit for dystrophic patients. In fact,the test used to evaluate fatigability to exercise performance isconsidered representative of the 6-min walk test used for DMDpatients.

The in vivo effects observed after short-term treatment withthe high dose of 30 mg/kg were greater than those observed withthe natural comparator, the steroid nandrolone, as well as thoseof prednisolone, the “gold standard” for pharmacological man-agement of Duchenne patients [12]. Importantly, the beneficialeffects observed for the in vivo parameters was maintained overtime and was evident at 10- and 100-fold lower doses, empha-sising the remarkable sensitivity of dystrophic animals towardsthis compound. Accordingly, the lack of a dose-dependent rela-tionship observed with many parameters might be because thelow doses are near the maximal effective ones. In fact, the

plasma levels of GLPG0492 in chronically treated mice wereclearly dose-dependent and comparable to those after acute dos-ing, ruling out potential saturation in the pharmacokinetic ofGLPG0492.

A. Cozzoli et al. / Pharmacological Research 72 (2013) 9– 24 21

Fig. 10. Effect of GLPG0492 treatment on fibrosis markers of mdx mouse muscle. In(A), the bars show the percentage of area of muscle damage (left) and the percentageof non-muscle area (right) of diaphragm muscle of mdx mice treated either with cornoil (Mdx + V1) or with 30 mg/kg GLPG0492 (Mdx + GLPG0492), 5 mg/kg nandrolone(Mdx + NAND), water (Mdx + V2) or 1 mg/kg �-methylprednisolone (Mdx + PDN),as measured by haematoxylin–eosin staining. The drugs were given 6 days perweek. Each bar is the mean of at least 3 muscles/approximately 10 fields per mus-cle. Significant differences between groups were evaluated using an ANOVA test,and the Bonferroni t-test post hoc correction. The results are indicated as follows:◦significantly different with respect to mdx mice treated with corn oil p < 0.03. In (B),the bars show the levels of total (left) and active TGF-�1 (right), in diaphragm mus-cle, in mdx mice treated with either vehicle (corn oil, Mdx + V1) 30 mg/kg GLPG0492(Mdx + GLPG0492), or 5 mg/kg nandrolone (Mdx + NAND), as measured by ELISA.Each value is the mean ± S.E.M. from 4 to 5 preparations. An ANOVA test for mul-tiple comparisons between the groups did not indicate any significant differencein TGF-�1 levels. The post hoc Bonferroni t-test results are indicated as follows:◦significantly different with respect to vehicle-treated mdx mice, p < 0.03. In (C), thebars show the levels of total (left) and active TGF-�1 (right) in diaphragm musclefor mdx mice treated with either corn oil (Mdx + V1), or with GLPG0492 at 0.3, 3 or30 mg/kg (Mdx + GLPG0492), as measured by ELISA. The drugs were given 6 days perwew

msdmaedat

Fig. 11. Dose-related plasma levels of GLPG0492 in the mouse. GLPG0492 plasmalevels were assessed over an 8-h period after s.c. delivery of 0.3 mg/kg (A), 3 mg/kg

eek. Each value is the mean ± S.E.M. from 4 to 5 preparations. Significant differ-nces between groups were evaluated using Student’s t test. ◦Significantly differentith respect to mdx exercised mice 0.05 < p < 0.025.

Surprisingly, no significant drug effect was observed on grossuscle mass and histologically on fibre diameter (data not shown),

uggesting a lack of hypertrophic action. These findings are quiteifferent from the results observed with other approaches, such asyostatin-inhibitors or �2 agonists, for which hypertrophic action

ppears to be the mechanism responsible for the functional ben-

fit in dystrophic mice [13,14,16]. Whether hypertrophy is a realesired effect in enhancing performance is still controversial. Inddition, drug-mediated hypertrophy may be more evident duringhe growing phase of dystrophic animals [15,17,31]. We actually

(B), or 30 mg/kg (C) of the compound into wild-type mice receiving a single acutedose (black circle with a slash) or into mdx mice receiving chronic dosing (blackcircle).

tested the SARM in growing young adult mdx mice, showing thatthe absence of hypertrophy is not an age-related issue.

Importantly, the effect of androgens and SARM on muscle

mass has often been observed under conditions of experimentallyinduced atrophy, including those from corticosteroids, hypogo-nadism, or ageing sarcopenia [32–34]. In mdx mice, the basallevels of testosterone were slightly higher than in wt, whereas the

22 A. Cozzoli et al. / Pharmacological Research 72 (2013) 9– 24

Fig. 12. Plasma testosterone levels in wt and mdx mice. In (A), the bars show theserum testosterone levels of 8-week-old wild type and mdx mice either exercisedfor 4 weeks (WT EXER; MDX EXER) or not (WT SED; MDX SED). Each bar is themean ± S.E.M. from 5 to 6 animals. Significant differences between groups wereevaluated by Student’s t test. *Significantly different with respect to wt mice withpi

eb[thnrspaHGlep

w

Fig. 13. Effect of GLPG0492 on expression of gene targets of SARM action in DIAand GC muscle of mdx mice. Real-time PCR analysis was performed for insulin-likegrowth factor-1 (IGF-1) and follistatin, genes involved in the control of muscle mass;myogenin, a marker of muscle regeneration; and peroxisome proliferator receptor�-coactivator (PGC)-1�, a modulator of muscle metabolism and of mechano-transduction signalling. The figure shows the normalised values of the target geneswith respect to a housekeeping gene (GADPH) for vehicle (Mdx + V1); 0.3 mg/kgGLPG0492 (Mdx + 0.3 mg/kg GLPG0492) and 3 mg/kg GLPG0492 (Mdx + 3 mg/kgGLPG0492) in diaphragm (left side; DIA) and gastrocnemius (right side; GC). The

< 0.05. In (B), the bars show the effect of GLPG0492 on plasma testosterone levelsn mdx mice. Each bar is the mean ± S.E.M. from 5 to 7 animals.

xercise-induced increase was slightly lower, suggesting a possi-le distress but not failure of the axis muscle activity-testosterone35–38]. GLPG0492, as with many developed SARMs, acts as a par-ial agonist of androgen receptors [23,24]. Thus, the absence ofypertrophy in our experimental settings may be related to theormal hormonal conditions of the animals used and the occur-ence of antagonism of testosterone actions. This hypothesis isupported by the slight decrease in the testosterone level and thearallel decrease in prostate weight observed in the mdx micefter prolonged administration with the highest dose of GLPG0492.owever, the treatment of the non-exercised mdx mice withLPG0492 at 3 mg/kg for 4 weeks produced minor, if any, ame-

ioration of the in vivo performance compared to that observed inxercised mice (data not shown), which is likely related to the lower

henotype severity.

Importantly, DMD patients experience delayed pubertyith reduced testosterone levels, often related to the use of

drugs were given 6 days per week. Each value is the mean ± S.E.M. from 4 to 5preparations.

corticosteroids [39], which predicts a greater sensitivity to SARMswith respect to the mdx animals.

From a functional perspective, an improvement in contractionstrength was observed with the isolated muscles of mdx mice, anddifferent sensitivities for the diaphragm and hind limb EDL musclewere observed. These results imply that there are muscle-specificactions of the drug possibly due to differences in either drug distri-

bution or androgen-receptor expression. The slight but significantincrease in strength recorded in the diaphragm is important from atherapeutic perspective, as this muscle presents impairment more

ologic

sadtttaipec

ooscwsaaritsliawafifimtotfoBb

bctrapbgtrsdtos

5

iecaww

[

[

[

[

[

[

[

[

[

[

[

A. Cozzoli et al. / Pharmac

imilar to that in patients [2,40]. Importantly, the EDL muscle waslso partially protected by GLPG0492, with respect to fatigue andamaging eccentric contraction, during the longer treatment dura-ion. In parallel, the cellular indices of myofibre distress such ashe mechanical threshold and resting sarcolemmal ionic conduc-ance were ameliorated in treated EDL myofibres, although thismelioration was neither time- nor dose-dependent. The functionalmprovement in both respiratory and limb muscle may, at leastartially, account for the observed in vivo effects; however, thexact correlation between the in vivo and ex vivo performances isurrently under further evaluation.

No marked signs of amelioration, such as reduced degenerationr enhanced regeneration, were evident in the histological analysisf the GLPG0492-treated muscles, despite the increase in proteinynthesis, which suggests anabolic activity. In addition, the bio-hemical plasma markers of muscle sufferance, such as CK and LDH,ere not improved. Also, no changes were found in the expres-

ion of genes that can be the target of androgen-like drugs andre involved in regeneration and growth, such as IGF-1, follistatinnd myogenin or in muscle metabolism, such as PGC-1�. Theseesults raise questions about the mechanism of action of the SARMn the skeletal muscle of dystrophic animals. However, it is impor-ant to emphasise that GLPG0492-treated diaphragms showed aignificant decrease in TGF-�1, a marker of fibrosis, which paral-eled, in the first study, a decrease in non-muscle area. Although notnvestigated herein, the reduction of TGF-�1 could be related to thebility of androgens to interfere with the WNT-� catenin pathway,hich eventually results in the inhibition of TGF-�1 production

nd signalling [23]. This action is important, due to the key role ofbrosis in dystrophinopathies and the potential usefulness of a lessbrotic respiratory muscle to contract and sustain training perfor-ance. Additionally, part of the amelioration of the alterations in

he in vivo performance may result from the action of GLPG0492n other systems, such as the peripheral and central nervous sys-ems. In fact, androgens have been claimed to enhance motor nerveunction and increase the number of synaptic contacts, the densityf acetylcholine receptors and the level of neurotrophins such asDNF [38,41]. These potential effects in dystrophic animals need toe addressed in future studies.

Despite the tissue-selective action, the mechanism of action ofoth steroid hormones and SARMs on skeletal muscle is not yetlear, and our data indicate that this is particularly true for dys-rophic animals. Androgen receptors (AR) are classical intracellulareceptors that modulate the transcription of specific genes oncectivated. Several tissue specific co-activators and co-repressorslay a key role in the action of testosterone [42]. SARMs have alsoeen claimed to interfere with AR and modulate a different set ofenes [23]. In skeletal muscle, some of these co-activators belongo the supervillin family, a little known family of proteins withelevant actions as cytoskeletal components [43,44]. Some of theupervillin-related proteins, such as archivillin co-segregate withystrophin at the costamere level [44]. Then, the absence of dys-rophin may affect the localisation and function of co-activatorsf AR, leading to a difference in androgen-mediated physiologicalignalling. These aspects deserve additional specific studies.

. Conclusions

The results of this pre-clinical study suggest that GLPG0492 maymprove the quality of life of DMD patients due to its ability tonhance functional performance. Further studies are necessary to

larify its mechanism of action and its potential role as a primarygent to counteract muscle wasting or as an important adjuvantith other therapies, such as corticosteroids. Pre-clinical studiesith low doses of GLPG0492 in combination with PDN will be

[

al Research 72 (2013) 9– 24 23

essential to demonstrate the potential advantages and clear thera-peutic applications in DMD.

Conflicts of interest

None.

Acknowledgments

We gratefully acknowledge the support of Charley’s Fund andthe Nash Avery Foundation.

References

[1] Hoffman EP, Dressman D. Molecular pathophysiology and targeted ther-apeutics for muscular dystrophy. Trends in Pharmacological Sciences2001;22:465–70.

[2] Grounds MD, Radley HG, Lynch GS, Nagaraju K, De Luca A. Towards developingstandard operating procedures for pre-clinical testing in the mdx mouse modelof Duchenne muscular dystrophy. Neurobiology of Disease 2008;31:1–19.

[3] Rando TA, Disatnik MH, Yu Y, Franco A. Muscle cells from mdx mice havean increased susceptibility to oxidative stress. Neuromuscular Disorders1998;8(1):14–21.

[4] Fraysse B, Liantonio A, Cetrone M, Burdi R, Pierno S, Frigeri A, et al. The alterationof calcium homeostasis in adult dystrophic mdx muscle fibers is worsened bya chronic exercise in vivo. Neurobiology of Disease 2004;17:144–54.

[5] Chen YW, Nagaraju K, Bakay M, McIntyre O, Rawat R, Shi R, et al. Early onsetof inflammation and later involvement of TGFbeta in Duchenne muscular dys-trophy. Neurology 2005;65:826–34.

[6] Rolland JF, De Luca A, Burdi R, Andreetta F, Confalonieri P, Conte Camerino D.Overactivity of exercise-sensitive cation channels and their impaired modula-tion by IGF-1 in mdx native muscle fibers: beneficial effect of pentoxifylline.Neurobiology of Disease 2006;24:466–74.

[7] Lai Y, Thomas GD, Yue Y, Yang HT, Li D, Long C, et al. Dystrophins carryingspectrin-like repeats 16 and 17 anchor nNOS to the sarcolemma and enhanceexercise performance in a mouse model of muscular dystrophy. Journal ofClinical Investigation 2009;119:624–35.

[8] Shkryl VM, Martins AS, Ullrich ND, Nowycky MC, Niggli E, Shirokova N. Recipro-cal amplification of ROS and Ca(2+) signals in stressed mdx dystrophic skeletalmuscle fibers. Pflügers Archiv 2009;458(5):915–28.

[9] Whitehead NP, Yeung EW, Froehner SC, Allen DG. Skeletal muscle NADPH oxi-dase is increased and triggers stretch-induced damage in the mdx mouse. PLoSONE 2010;5(12):e15354.

10] Villalta SA, Rinaldi C, Deng B, Liu G, Fedor B, Tidball JG. Interleukin-10reduces the pathology of mdx muscular dystrophy by deactivating M1macrophages and modulating macrophage phenotype. Human MolecularGenetics 2011;20(4):790–805.

11] Pichavant C, Aartsma-Rus A, Clemens PR, Davies KE, Dickson G, Takeda S, et al.Current status of pharmaceutical and genetic therapeutic approaches to treatDMD. Molecular Therapy 2011;19:830–40.

12] Bushby K, Finkel R, Birnkrant DJ, Case LE, Clemens PR, Cripe L, et al. Diagnosisand management of Duchenne muscular dystrophy, part 1: diagnosis, and phar-macological and psychosocial management. Lancet Neurology 2010;9:77–93.

13] Bogdanovich S, Krag TO, Barton ER, Morris LD, Whittemore LA, Ahima RS, et al.Functional improvement of dystrophic muscle by myostatin blockade. Nature2002;420(6914):418–21.

14] Patel K, Amthor H. The function of Myostatin and strategies of Myostatin block-ade new hope for therapies aimed at promoting growth of skeletal muscle.Neuromuscular Disorders 2005;2:117–26.

15] Amthor H, Macharia R, Navarrete R, Schuelke M, Brown SC, Otto A, et al. Lack ofmyostatin results in excessive muscle growth but impaired force generation.Proceedings of the National Academy of Sciences of the United States of America2007;104(6):1835–40.

16] Gehrig SM, Koopman R, Naim T, Tjoakarfa C, Lynch GS. Making fast-twitch dys-trophic muscles bigger protects them from contraction injury and attenuatesthe dystrophic pathology. American Journal of Pathology 2010;176(1):29–33.

17] Murphy KT, Ryall JG, Snell SM, Nair L, Koopman R, Krasney PA, et al.Antibody-directed myostatin inhibition improves diaphragm pathology inyoung but not adult dystrophic mdx mice. American Journal of Pathology2010;176(5):2425–34.

18] Krahn MJ, Anderson JE. Anabolic steroid treatment increases myofiber dam-age in mdx mouse muscular dystrophy. Journal of the Neurological Sciences1994;125(2):138–46.

19] Souccar C, Gonc alo Mdo C, Buck Hde S, Lima-Landman MT, Lapa AJ. Mild dys-trophic damage in the androgen-sensitive levator ani muscle of the mdx mouse.Neuromuscular Disorders 2005;15(1):48–55.

20] Fenichel G, Pestronk A, Florence J, Robison V, Hemelt V. A beneficial effect of

oxandrolone in the treatment of Duchenne muscular dystrophy: a pilot study.Neurology 1997;48(5):1225–6.

21] Fenichel GM, Griggs RC, Kissel J, Kramer TI, Mendell JR, Moxley RT, et al. A ran-domized efficacy and safety trial of oxandrolone in the treatment of Duchennedystrophy. Neurology 2001;56(8):1075–9.

2 cologic

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

4 A. Cozzoli et al. / Pharma

22] Balagopal P, Olney R, Darmaun D, Mougey E, Dokler M, Sieck G, et al. Oxan-drolone enhances skeletal muscle myosin synthesis and alters global geneexpression profile in Duchenne muscular dystrophy. American Journal of Phys-iology - Endocrinology and Metabolism 2006;290(3):E530–9.

23] Bhasin S, Jasuja R. Selective androgen receptor modulators as function pro-moting therapies. Current Opinion in Clinical Nutrition and Metabolic Care2009;12(3):232–40.

24] Narayanan R, Mohler ML, Bohl CE, Miller DD, Dalton JT. Selective androgenreceptor modulators in preclinical and clinical development. Nuclear ReceptorSignaling 2008;6:e010.

25] De Luca A, Pierno S, Liantonio A, Cetrone M, Camerino C, Fraysse B, et al.Enhanced dystrophic progression in mdx mice by exercise and beneficial effectsof taurine and insulin-like growth factor-1. Journal of Pharmacology and Exper-imental Therapeutics 2003;304(1):453–63.

26] De Luca A, Nico B, Liantonio A, Didonna MP, Fraysse B, Pierno S, et al. A multi-disciplinary evaluation of the effectiveness of cyclosporine a in dystrophic mdxmice. American Journal of Pathology 2005;166(2):477–89.

27] Burdi R, Rolland JF, Fraysse B, Litvinova K, Cozzoli A, Giannuzzi V, et al. Multiplepathological events in exercised dystrophic mdx mice are targeted by pentoxi-fylline: outcome of a large array of in vivo and ex vivo tests. Journal of AppliedPhysiology 2009;106(4):1311–24.

28] Brooks SV, Faulkner JA. Contractile properties of skeletal muscles from young,adult and aged mice. Journal of Physiology 1988;404:71–82.

29] Cozzoli A, Nico B, Sblendorio VT, Capogrosso RF, Dinardo MM, Longo V, et al.Enalapril treatment discloses an early role of angiotensin II in inflammation-and oxidative stress-related muscle damage in dystrophic mdx mice. Pharma-cological Research 2011;64(5), 482–92.a.

30] Cozzoli A, Rolland JF, Capogrosso RF, Sblendorio VT, Longo V, Simonetti S,et al. Evaluation of potential synergistic action of a combined treatment withalpha-methyl-prednisolone and taurine on the mdx mouse model of Duchennemuscular dystrophy. Neuropathology and Applied Neurobiology 2011;37(3),243–56.b.

31] Shavlakadze T, Chai J, Maley K, Cozens G, Grounds G, Winn N, et al. A growth

stimulus is needed for IGF-1 to induce skeletal muscle hypertrophy in vivo.Journal of Cell Science 2010;123(Pt 6):960–71.

32] Jones A, Hwang DJ, Narayanan R, Miller DD, Dalton JT. Effects of a novel selectiveandrogen receptor modulator on dexamethasone-induced and hypogonadism-induced muscle atrophy. Endocrinology 2010;151(8):3706–19.

[

al Research 72 (2013) 9– 24

33] Gao W, Reiser PJ, Coss CC, Phelps MA, Kearbey JD, Miller DD, et al. Selectiveandrogen receptor modulator treatment improves muscle strength and bodycomposition and prevents bone loss in orchidectomized rats. Endocrinology2005;146(11):4887–97.

34] Hoffman JR, Kraemer WJ, Bhasin S, Storer T, Ratamess NA, Haff GG, et al. Positionstand on androgen and human growth hormone use. Journal of Strength andConditioning Research 2009;23(5 Suppl.):S1–59.

35] De Luca A. Pre-clinical drug tests in the mdx mouse as a model of dys-trophinopathies: an overview. Acta Myologica 2012;31(1):40–7.

36] Vingren JL, Kraemer WJ, Ratamess NA, Anderson JM, Volek JS, Maresh CM.Testosterone physiology in resistance exercise and training: the up-streamregulatory elements. Sports Medicine 2010;40(12):1037–53.

37] West DW, Phillips SM. Anabolic processes in human skeletal muscle: restor-ing the identities of growth hormone and testosterone. Physician andSportsmedicine 2010;38(3):97–104.

38] Verhovshek T, Cai Y, Osborne MC, Sengelaub DR. Androgen regulatesbrain-derived neurotrophic factor in spinal motoneurons and their target mus-culature. Endocrinology 2010;151(1):253–61.

39] Bianchi ML, Biggar D, Bushby K, Rogol AD, Rutter MM, Tseng B.Endocrine aspects of Duchenne muscular dystrophy. Neuromuscular Disorders2011;21(4):298–303.

40] Willmann R, Possekel S, Dubach-Powell J, Meier T, Ruegg MA. Mammaliananimal models for Duchenne muscular dystrophy. Neuromuscular Disorders2009;19:241–9.

41] Bleisch WV, Harrelson A. Androgens modulate endplate size and ACh recep-tor density at synapses in rat levator ani muscle. Journal of Neurobiology1989;20:189–202.

42] Heemers HV, Tindall DJ. Androgen receptor (AR) coregulators: a diversityof functions converging on and regulating the AR transcriptional complex.Endocrine Reviews 2007;28(7):778–808.

43] Ting HJ, Yeh S, Nishimura K, Chang C. Supervillin associates with andro-gen receptor and modulates its transcriptional activity. Proceedings of theNational Academy of Sciences of the United States of America 2002;99(2):

661–6.

44] Oh SW, Pope RK, Smith KP, Crowley JL, Nebl T, Lawrence JB, et al. Archvillin,a muscle-specific isoform of supervillin, is an early expressed componentof the costameric membrane skeleton. Journal of Cell Science 2003;116(Pt11):2261–75.