(19) TZZ Z _T

(11) EP 2 540 295 A1

(12) EUROPEAN PATENT APPLICATION

(43) Date of publication: (51) Int Cl.: 02.01.2013 Bulletin 2013/01 A61K 31/404 (2006.01) A61P 25/00 (2006.01) A61P 25/28 (2006.01) (21) Application number: 11171532.2

(22) Date of filing: 27.06.2011

(84) Designated Contracting States: (72) Inventors: AL AT BE BG CH CY CZ DE DK EE ES FI FR GB • Briault, Sylvain GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO 45000 ORLEANS (FR) PL PT RO RS SE SI SK SM TR • Perche, Olivier Designated Extension States: 45380 LA CHAPELLE SAINT MESMIN (FR) BA ME (74) Representative: Thomas, Dean et al (71) Applicants: Cabinet Ores • Centre national de la recherche scientifique 36 rue de St Pétersbourg 75016 Paris (FR) 75008 Paris (FR) • Centre Hospitalier Régional d’Orléans 45032 Orléans Cédex 1 (FR)

(54) Compositions for the treatment of Fragile X syndrome

(57) The present invention relates to a composition a fluoro-oxindole or a chloro- oxindole for use in the treat- comprising a maxi-K potassium channel opener the use ment of fragile X syndrome. in the treatment of fragile X syndrome. More specifically the present invention relates to a composition comprising EP 2 540 295 A1

Printed by Jouve, 75001 PARIS (FR) 1 EP 2 540 295 A1 2

Description ularly shyness, limited eye contact, memory problems and difficulty with facial encoding and recognition. Many [0001] The present invention relates to compositions individuals with FXS also meet the diagnostic criteria for for the alleviation of neuropsychiatric symptoms and in autism. Most females who have the syndrome experi- particular those of Fragile X syndrome. 5 ence symptoms to a lesser degree because of their sec- [0002] Neuropsychiatry is the branch of medicine deal- ond X-chromosome; however, females can develop ing with mental disorders attributable to diseases of the symptoms just as severe as their male counterparts. nervous system, this discipline is closely related to neu- While full mutation males tend to present with severe rology but neurology and neuropsychiatry are typically intellectual disability, the symptoms of full mutation fe- practiced separately. Neurology is a medical specialty 10 males run the gamut of minimally affected to severe in- dealing with disorders of the nervous system such as tellectual disability, which may explain why females are pathologies of the central, peripheral and autonomic under diagnosed relative to males, leading to the persist- nervous systems, as well as the connective tissue asso- ence of the anomaly in the population. ciated with the nervous system. [0009] At the present time no medicament has been [0003] The basis of neuropsychiatric conditions can be 15 approved for the treatment of FXS. due to a single gene aberration or a combination of sev- [0010] FXS is considered an important therapeutic tar- eral mutations and/or environmental cues. get firstly as it is the most common form of mental retar- [0004] An example of a neuropsychiatric disorder dation attributable to a single gene mutation and also caused by a single gene aberration is Fragile X syndrome because as the development of drugs to treat neuropsy- (FXS). FXS or Martin-Bell syndrome, is a genetic syn- 20 chiatric diseases such as mental retardation or autism is drome which results in a spectrum of characteristic phys- becoming an increasingly important commercial objec- ical and intellectual limitations as well as a number of tive, an approach based upon looking at single- gene dis- emotional and behavioral aberrations which range from orders and the use of these as a window into the field severe to mild in manifestation. In particular sufferers of treating of neuropsychiatric diseases such as mental re- FXS can present symptoms of mental disorder upon the 25 tardation or autism is one of the most attractive develop- autism spectrum as well as epilepsy. ment strategies available. [0005] FXS is the most common form of mental disa- [0011] Workers seeking the means to treat FXS have bility (MD) due to a genetic defect. In 5 to 25% MD is shown that excessive glutamate accumulates at the syn- associated with behavioral disorders in the autism spec- apse of a mouse model of FXS and is responsible for the trum. With a prevalence of 1/1000 males and 1/4000 fe- 30 cognitive deficits seen in this mouse model of FXS (Huber males, in France a country of approximately 60 million K.M. et al. Proc. Natl. Acad. Sci. U. S. A. 99, 7746-7750 people, more than 40,000 FXS patients are currently un- (2002)). Workers also found that reducing expression of dergoing treatment. a specific glutamate receptor, mGluR5, reverses these [0006] The molecular basis of FXS is an abnormal ex- symptoms in the animal model (Dölen G. et al. Neuron pansion of a CGG triplet located at Xq27 in the 5’ un- 35 56, 955-962 (2007)). translated region of the FMR1 gene, which leads to a [0012] As a consequence of this work mGluR5 inhibi- decrease in transcription of the gene. There are three tors were developed, notably by Merck & Co Inc and sub- generally accepted states of the chromosome region in- sequently have been taken into clinical trials. volved in FXS which relate to the length of the repeated [0013] A second class of drug exemplified by arba- CGG sequence: Normal (29-31 CGG repeats) (not af- 40 clofen, has also been proposed which stimulates recep- fected by the syndrome), Premutation (55-200 CGG re- tors of gamma amino-butyric acid(B) or GABA(B) and peats) (not affected by the syndrome), Full Mutation thereby dampens glutamate signaling. Clinical trials are (more than 200 CGG repeats) (affected). also underway to test arbaclofen on children and adults [0007] The FMR protein (FMRP) acts as a suppressor with fragile X syndrome, as well as an open- label trial of of translational activity particularly in neuronal dendrites 45 children with autism. that control the localization and expression of a set of [0014] The Applicants have now discovered a further synaptic specific proteins. It has been proposed that the class of drugs which can be used to specifically treat the loss of negative feedback as a consequence of the drop symptoms of FXS and also may be generally applicable in FMRP production, results in abnormalities in the struc- to the treatment and alleviation of the symptoms of autism ture and functioning of the synapses which in turn cause 50 spectrum disorder and/or other classes of neuropsychi- the FXS phenotype. atric disorders. [0008] Aside from intellectual disability, prominent [0015] In particular the Applicants have investigated characteristics of FXS include an elongated face, large the effects of agents which act upon potassium channels or protruding ears, flat feet, larger testes (mac-as the means to treat the symptoms of specifically FXS roorchidism) and low muscle tone. Speech may include 55 and more generally those of autism. cluttered or nervous speech patterns. Behavioral char- [0016] Surprisingly the inventors have established that acteristics may include stereotypic movements (e.g., this class of compound can reverse characteristic symp- hand-flapping) and atypical social development, partic- toms associated with FXS and therefore in accordance

2 3 EP 2 540 295 A1 4 with a first aspect of the present invention therefore there 130015), indole carboxylates such as Indole-3-carboxy- is provided a composition comprising a maxi-potassium lic acid (CGS 7181, CGS 7184), Arylpyrrole (NS 8), Di- channel opener for use in the treatment of a neuropsy- hydrosoyasaponin-1 (DHS-1), Terpene, Cerebroside chiatric disorder. (Baifuzi), Mallotoxine (Rottlerin), Arylquinoline, Aryloxin- [0017] In accordance with this aspect of the present 5 dole, Pimaric acid (PiMa), Dichlorodehydroabietic (diCl- invention the neuropsychiatric disorder may in particular DHAA), Flavonoid (Naringenin, Phloretin), Benzofuroin- show symptoms/behaviors characteristic of autism spec- dole (LDD 175), Benzimidazolinone (1-EBIO), Fluoro trum disorders. propionamide [(S)-N-(4-benzoylphenyl)-3,3,3-trifluoro- [0018] The autism spectrum, also called autism spec- 2-hydroxy-2-methyl-propionamide], Unoprostone (and trum disorders (ASD) or autism spectrum conditions10 unoprostone isopropyl), Benzothiadiazolamine [6-(trif- (ASC), is a spectrum of neuropsychiatric conditions char- luoromethoxy) benzothiazol-2-amine], ethylbromide acterized by widespread abnormalities of social interac- tamoxifen, epoxyeicosatrienoic acid, estradiol-17 beta, tions and communication, as well as restricted interests Diphenylurea (NS 1608). and repetitive behavior. [0024] One class of potassium channel openers which [0019] Specifically in accordance with the present in- 15 selectively function on cells having high intracellular cal- vention there is provided a composition comprising a cium concentrations are 3- phenyl substituted oxindol de- maxi-K potassium channel opener for use in the treat- rivatives, as described in U. S. Patent Nos. 5,565,483 ment of fragile X syndrome. and 5,602,169. Halo-oxindoles such as fluoro-oxindole [0020] In accordance with the present invention a and chloro-oxindole compounds are within the above- Maxi-K channel also known as a BK channel (Big Potas- 20 described class and are capable of acting selectively as sium) or slol or BKCa (Big Potassium Calcium) channel, maxi-K channel openers on cells having high intracellular are ion channels characterized by their large conduct- calcium concentration. ance of potassium ions (K+) through cell membranes. [0025] This class of potassium channel opener has These channels are activated (opened) by changes in been characterized as being sensitive to the intracellular membrane electrical potential and/or by increases in con- 25 calcium concentration of cells and demonstrated to be centration of intracellular calcium ion (Ca 2+). Opening of most effective under conditions of increased intracellular maxi-K channels allows K + to passively flow through the calcium concentrations, e.g. micro molar range, while be- channel, down the electrochemical gradient. Under typ- ing minimally effective or not at all effective under normal ical physiological conditions, this results in an efflux of physiological intracellular calcium concentrations. K+ from the cell, which leads to cell membrane hyper- 30 [0026] These compounds do not act to an appreciable polarization (a decrease in the electrical potential across extent to open maxi- K potassium channels in cells having the cell membrane) and a decrease in cell excitability (a normal, moderate or low intracellular calcium concentra- decrease in the probability that the cell will transmit an tion action potential). [0027] In WO02/30868, this new class of Potassium [0021] As with other potassium channels, Maxi-K 35 channel opener were proposed for use as neuroprotec- channels have a tetrameric structure. Each monomer of tive agents, wherein dysfunction of Potassium channels/ the channel-forming alpha subunit is the product of the over accumulation of calcium is associated with neurons KCNMA1 gene. Modulatory beta subunits (encoded by being at risk during neurodegenerative conditions, such KCNMB1, KCNMB2, KCNMB3, or KCNMB4) can asso- as those which occur during acute ischemic stroke. ciate with the tetrameric channel. 40 [0028] Maxi-K ion channels are proteins that react to [0022] In accordance with the present invention a substantialincreases inintracellular Ca 2+ and membrane maxi-K channel opener is one which activates maxi-K depolarization by markedly increasing potassium (K +) ef- potassium channels in cells under conditions of high in- flux, rapidly hyperpolarizing the membrane and reducing tracellular calcium concentration and does not signifi- further voltage-dependent Ca2+ influx (V. K. Gribkoff et cantly activate maxi- K potassium channels in cells under 45 al., Adv. Pharmacol., 37:319-348 (1997)). low or normal concentrations of intracellular calcium. [0029] Although clinical trials were undertaken to as- [0023] As guidance, non- limiting examples of high lev- certain the effects of 3-phenyl substituted oxindol deriv- els of intracellular calcium are typically considered to be atives on ischemic stroke victims, no statistically signifi- in the high nanomolar (e.g. greater than about 250 or 300 cant effect was observed and therefore studies ceased nM) to micromolar range (e.g. about 1 to 10 mM); normal 50 (B. S. Jensen, CNS Drug Reviews, Vol. 8, 4 No. pp. or physiological levels of intracellular calcium are typical- 353-360 (2002)). ly considered to be in the range of about 50 nM to 250 [0030] Morespecifically inaccordance with the present nM and low levels of intracellular calcium are typically invention there is provided a composition comprising a considered to be in the range of about 5 to 50 nM.A large halo-oxindole selected from the group fluoro- oxindole or number of molecules are known to have the ability to55 a chloro-oxindole for use in the treatment of fragile X open maxi-K channel such as 3- phenyl substituted ox- syndrome. indol derivatives, halo-oxindoles, benzimidazolone (NS [0031] As indicated above 3- phenyl substituted oxin- 004, NS 1619), benzimidazole, amiodarone (KBdol derivatives, such as fluoro- oxindole and chloro-oxin-

3 5 EP 2 540 295 A1 6 dole compounds have previously been characterized as [0040] For a better understanding of the invention and maxi-K channel openers. to show how the same may be carried into effect, there [0032] Suitable fluoro-oxindoles for use according to will now be shown by way of example only, specific em- the present invention include 6( )-(5-chloro-2-methoxy- bodiments, methods and processes according to the phenyl)-1, 3-dihydro-3-fluoro-6-(trifluoromethyl)-2H-in- 5 present invention with reference to the accompanying dol-2-one; (3S)-(+)-(5-chloro-2-methoxyphenyl)-1, 3-di- drawings in which: hydro-3-fluoro-6-(trifluoromethyl)-2H-indole-2-one and (3S)-(-)-(5-chloro-2- methoxyphenyl)-1, 3-dihydro-3-flu- Figure 1: Analysis of transcriptional and protein lev- oro-6- (trifluoromethyl)-2H-indole-2-one. els in human patients with FXS. The expression of [0033] A suitable chloro-oxindole for use according to 10 the FMR1 gene and KCNMA1 gene is calculated us- the present invention include ((6)-(5-chloro-2-methoxy- ing the expression chip EXON1.0 Affymetrix (Panel phenyl)-1, 3-dihydro-3-chloro-6- (trifluoromethyl)-2H-in- A) and verified by quantitative PCR. FXS patients dol-2-one); (3S)-(+)-(5-chloro-2-methoxyphenyl)-1, 3-di- have almost no expression of the FMR1 gene but hydro-3-chloro-6-(trifluoromethyl)-2H-indol-2-one; and show no abnormalities in the expression of (3R)-(-)-(5-chloro-2-methoxyphenyl)-1,3-dihydro-3- 15 KCNMA1. Quantification of proteins from each of chloro-6-(trifluoromethyl)-2H-indol-2-one. these genes was obtained by Western blot (Panel [0034] In particular the Applicants have performed a B). FXS patients, the inventors observed the ab- number of experiments using the fluoro-oxindole com- sence of FMRP and the inventors also saw a signif- pound (3S)-(+)-(5-Chloro-2-methoxyphenyl)-1, 3-dihy- icant decrease of 46% in the amount of the KCNMA1 dro-3-fluoro-6- (trifluoromethyl)-2H-indol-2-one which is 20 protein. also called BMS-204352 and this compound represents Figure 2: Analysis of transcriptional and protein lev- a preferred embodiment of the present invention. els in mice FMR1 KO and wild type mice. The ex- [0035] Although the applicants have performed exper- pression of the FMR1 gene (Panel A) and gene iments using 3-phenyl substituted oxindol derivatives KCNMA1 (Panel B) was studied by quantitative PCR such as (3S)-(+)-(5-Chloro-2-methoxyphenyl)-1, 3-dihy- 25 in different organs (brain, testis, liver). FMR1 KO dro-3-fluoro-6- (trifluoromethyl)-2H-indol-2-one, given mice exhibit, as expected, virtually no expression of the therapeutic effect is due to the opening of the maxi- the FMR1 gene but show no abnormalities in the K channel, it is clear that all the various substances expression of the KCNMA1 gene. Quantification of known to open Potassium channels will have a similarly the FMR1 and KCNMA1 proteins from these genes therapeutic effect. 30 was obtained by Western blot (Panels C and D re- [0036] According to a second aspect of the present spectively). In the brain, FMR1 KO mice, no FMRP invention a method of treating fragile X syndrome in an was detected and a decreased amount of the KCN- individual in need thereof, comprising: administering to MA protein (a 42% reduction compared to wild- type the individual an effective amount of a maxi-K channel mice (Panel D)) was observed. opener, said opener having opener activity on maxi-K 35 Figure 3: Immunolocalization of the FMRP protein potassium channel proteins in neuronal cells having a and the KCNMA1 protein was observed in the cer- high intracellular calcium concentration, while having no ebellum. The distribution of FMRP [left column A significant opener activity on maxi- K potassium channel (x20) and B (x40 dotted area)] and the distribution proteins in neuronal cells having normal or low intracel- of KCNMA1 [right column C (x20) and D (x40 dotted lular calcium concentration. 40 area)] was carried out in wild type mice. FMRP is [0037] In particular where the maxi-K channel opener present in all the cerebral neurons. BKCa protein is is selected from the group consisting of fluoro-oxindole found in the cerebellum in Purkinje cells (D, arrows). compounds and chloro-oxindole compounds. Nuclei were labeled with DAPI. [0038] Most particularly wherein the fluoro-oxindole Figure 4: Analysis of social interaction as per the compound is selected from the group consisting of ( 6)-(5- 45 Crawley test. The social activity of wild type mice chloro-2-methoxyphenyl)-1, 3-dihydro-3-fluoro-6-(trifluor- and FMR1 KO mice, treated with vehicle or BMS- omethyl)-2H-indol-2-one; (3S)-(+)-(5-chloro-2-methoxy- 204352, was studied (n = 16 per group). FMR1 KO phenyl)-1, 3-dihydro-3-fluoro-6-(trifluoromethyl)-2H-indo- mice exhibit a lack of social interest significantly (p le-2-one and (3S)-(-)-(5- chloro-2-methoxyphenyl)-1, 3-di- = 0.036) lower than wild type controls (Panel A). An hydro-3-fluoro-6- (trifluoromethyl)-2H-indole-2-one. 50 injection of BMS-204352 (2 mg/kg) significantly in- [0039] Alternatively wherein the chloro-oxindole com- creased (p = 0.037) the social interest of FMR1 KO pound is selected from the group consisting of ((6)-(5- mice (Panel B). An injection of BMS-204352 (2 mg chloro-2-methoxyphenyl)-1, 3-dihydro-3-chloro-6- (trif- / kg) did not significantly alter the social activity of luoromethyl)-2H-indol-2-one); (3S)-(+)-(5-chloro-2-me- wild-type mice (Panel C). The observation of an in- thoxyphenyl)-1, 3-dihydro-3-chloro-6- (trifluoromethyl)- 55 creased social interest following administration of 2H-indol-2-one; and (3R)-(-)-(5-chloro-2-methoxyphe- BMS-204352 to FMR1 KO mice was not due to an nyl)-1,3-dihydro-3-chloro-6-(trifluoromethyl)-2H-indol-2- increase in locomotor activity (Panel D). one. Figure 5: Analysis of direct social interaction be-

4 7 EP 2 540 295 A1 8

tween test mouse (male) with a female. The social - Expression Assays activity of wild type mice and FMR1 KO mice, treated with vehicle or BMS-204352, was studied (n = 9 per RNA isolation group). FMR1 KO mice have a deficit of social in, terest in the female compared to control animals, 5 [0043] Frozen brain tissue samples were homoge- however, an injection of BMS-204352 (2 mg / kg) nized in TRIZOL reagent (Invitrogen Life Techonolgies, significantly increased social interaction of FMR1 KO Carlsbad). Total RNA was extracted using a standard mice and did not have a significant effect upon the chloroform protocol. RNA integrity was evaluated by us- wild type control. ing RNA 6000 Nano LabChips on an Agilent 2100 Bio- Figure 6: Analysis of mouse behavior using Y Maze 10 analyzer (Agilent Technologies, Foster City). RNA purity to test for non- social novelty preference. Preference was assessed by the ratio of spectrophotometric absorb- for novelty of wild mice and FMR1 KO, treated with ance at 260 and 280 nm (A260/280nm) using NanoDrop vehicle or BMS-204352, was studied (n = 9 per ND-1000 (NanoDropInc, Wilmington). All chips were pre- group). FMR1 KO mice show a deficit/ lack of interest pared according to the manufacturer’s instructions. Total in the novel environment significantly lower than the 15 RNA degradation was evaluated by reviewing the elec- wild type mice controls. An injection of BMS- 204352 tropherograms and the quantification of preserved 18S (2 mg / kg) significantly increased the time spent by and 28S peaks. the FMR1 KO mice in the novel enviroment. EXON1.0 Affymetrix [0041] There will now be described by way of example 20 a specific mode contemplated by the Inventors. In the [0044] cDNA was synthesized from RNA total using following description numerous specific details are set the GeneChip® WT (Whole Transcript) Sense Target La- forth in order to provide a thorough understanding. It will bellingand Control Reagentskit as described by the man- be apparent however, to one skilled in the art, that the ufacturer (Affymetrix). The sense cDNA was then frag- present invention may be practiced without limitation to 25 mented by UDG (uracil DNA glycosylase) and APE 1 these specific details. In other instances, well known (apurinic/apyrimidic endonuclease 1) and biotin- labelled methods and structures have not been described so as with TdT (terminal deoxynucleotidyl transferase) using not to unnecessarily obscure the description. the GeneChip® WT Terminal labelling kit (Affymetrix, Santa Clara, USA). Hybridization was performed using Materials and Methods 30 5 mg of biotinylated target, which was incubated with the GeneChip® Human Exon 1.0 ST array (Affymetrix) at - Western Blot 45°C for 16-20 hours. Following hybridization, non-spe- cifically bound material was removed by washing and [0042] Proteins of interest were detected by Western detection of specifically bound target was performed us- blot. Mice brain was, homogenized in RIPA buffer [NaCl 35 ing the GeneChip® Hybridization, Wash and Stain kit, 0,15M, EDA 1mM, Tris10 mM, Nonidet P40 1%, SDS and the GeneChip® Fluidics Station 450 (Affymetrix). 0,2%] and incubated on ice for 10 before being centri- The arrays were scanned using the GeneChip® Scanner fuged at 11 000 g for 20 min at 4°C. The protein content 3000 7G (Affymetrix) and raw data was extracted from of the supernatant was quantified using BCA protein as- the scanned images and analyzed with the Affymetrix say (Pierce). 60 mg of protein was loaded onto a 12% 40 Power Tools software package (Affymetrix). The micro- SDS-polyacrylamide gel. Proteins were then blotted onto array data has been deposited in the Gene Expression a nitrocellulose membrane. After blocking (Tris- buffered Omnibus Database. saline containing 0.1% Tween 20 and 10% Non fat dry milk, TTBS) the membrane was incubated overnight at Real time quantitative PCR 4°C with anti- FMRP or anti- KCNMA1 (scbt-28739, Santa 45 Cruz biotechnology and APC- 021, Alomone, repectively) [0045] For each tested gene, commercially available diluted in TTBS1X at 1: 2000. After TTBS1X washes, sec- primers and TaqMan probes were used from Agilent Bi- ondary antibody was incubated 1 hour at 1: 4000 (Amer- osystem technologies. PCR was carried out in a 50m l shamBiosciences). Immunopositive signalswere detect- volume containing 1X PCR buffer Master Mix (Applied ed using an ECL plus chemiluminescence detection kit 50 Biosystems) and cDNA. For the TaqMan assay, amplifi- (Amersham Biosciences). Then blots were stripped and cation reactions were initially heated to 95°C for 10 min- reprobed with anti-GAPDH antibody (Abcam). Quantifi- utes and then subjected to 45 cycles of 94°C for 15 sec- cation of immunoreactive bands was done using Image onds, 60°C for 30 seconds, and 72°C for 30 seconds in J (Image J Software). On each blot, there were 2 samples Mx3005p (Agilent Biotechnology, Agilent). Fluorescent per time and three blots were done per antibody. 55 data were collected during the 72°C step. The PCR threshold cycle (Ct), which is defined as the fractional cycle number at which the fluorescence reaches 10 times the standard deviation (SD) of the baseline, was deter-

5 9 EP 2 540 295 A1 10 mined by the software. Average Ct for duplicate stand- one side to the other. ards and clinical samples was calculated by the software. [0051] Nonsocial activities: rearing (standing on the Standard curve equations were calculated by regression hind limbs sometimes with the forelimbs against the walls analysis of average Ct versus the log10 of the standard of the cage) and digging. Time spent in self-grooming copy number. The viral copy numbers in the clinical sam- 5 (the animal licks and mouths its own fur) was analyzed ples were calculated automatically by the data analysis separately, since this is sometimes considered to repre- software. sent repetitive behavior. [0046] In these assays primer sets KCNMA1: Mm- 00516078_ml and FMR1: Mm00484415_m1 were used Sociability and preference for social novelty in the in mouse studies and KCNMA1: Hs00266938_m1 and 10 three compartment test FMR1: Hs00924547_m1were used in human studies, both from Agilent Biosystem technologies. Apparatus.

Behavioral procedures [0052] The testing apparatus employed here is similar 15 to the one previously described by others (Moy et al., [0047] Three behavioural tests were made on mice 2004). It consisted of 3 compartments : a central cham- which were 12 weeks of age, the tests were a direct social ber (45x18x25 cm) connected on each side to another interaction test with a female, then a three- compartment compartment (45x20x25 cm) through a small rectangular test for sociability and a spontaneous alternation test in opening (15x5 cm). The floors and walls (1 cm thick) of a Y-maze. Two experimental groups (n=10 for both) were 20 all compartments were made of transparent Plexiglas. studied, the vehicle group (DMSO 1/80th, Tween 80 Each side compartment contained a round stimulus cage 1/80th, Saline Solution) and the BMS-204352 treated (10 cm in diameter, 7 cm high) made of wire mesh (hole groups(DMSO containing BMS- 204352 (2mg/kg) 1/80 th, size: 0.7x0.7 cm) covered by a plastic roof (5 cm high). Tween 80 1/80 th, Saline Solution). Administration of each A metal weight was attached to the roof in order to keep solution was made 30 minutes before the three tests. 25 the stimulus cage stable. Each stimulus cage was placed at a distance of 6 cm from the back wall and 4 cm from Direct social interaction with a female the sides. Tracking images from a camera above the center of the apparatus were analyzed with Ethovision Apparatus. (Version 3.1, Noldus Technology, Wageningen, The 30 Netherlands). Procedure. Experimental and stimulus [0048] Direct social interaction was assessed in a mice (4-week old NMRI male mice) were individually 30x15x22 cm plastic cage with 3 cm of sawdust and a housed in standard polycarbonate cages provided with plastic roof with a 17x8 cm central opening. sawdust, food, and water bottles and left undisturbed in the experimental room for about 10 min before testing Procedure. 35 began. Each experimental subject was then introduced in the middle of the central compartment and allowed to [0049] Experimental and stimulus mice were habituat- explore the apparatus for 3 trials of 5 min each: ed to the experimental room as before. Each experimen- tal mouse was then introduced into the testing cage and Trial 1 (habituation): the stimuli cages were empty; left to habituate for 5 min. An unfamiliar stimulus mouse 40 basal levels of exploration were assessed. (NMRI adult female) was then introduced into the testing Trial 2 (sociability): a stimulus mouse was intro- cage through the roof opening. The testing session lasted duced in one of the stimulus cages, while a novel 3 min, but was stopped immediately if aggressive epi- object (a plastic grey cylinder, 6 cm in diameter, 2 sodes occurred. The testing cage was cleansed with wa- cm high) was introduced in the opposite cage (sides ter and the sawdust was renewed between sessions. 45 were counterbalanced within experimental groups); Testing sessions were recorded and videos were ana- preferential exploration of the social versus non-so- lyzed with Observer XT (version 7, Noldus, The Nether- cial novel stimuli was measured. lands), taking only the experimental animal into account. Trial 3 (social novelty preference): as trial 2, but the One observer who was unaware of the genotype of the object was exchanged for a novel stimulus mouse; animals scored both frequency and duration for each of 50 preferential exploration of the novel versus familiar the following behavioral categories and elements: social stimulus was evaluated.

Variables measured [0053] At the end of each trial the experimental animal was confined in the central compartment by means of [0050] Affiliative behaviors: sniffing the head and the 55 two Plexiglas magnetic doors for 30 sec. At the end of snout of the partner, its anogenital region, or any other the third trial the apparatus as well as the object and the part of the body; allogrooming (grooming the partner); stimuli cages were cleansed with water and dried. Vari- traversing the partner’s body by crawling over/ under from ables measured. Exploration of each stimulus was as-

6 11 EP 2 540 295 A1 12 sessed by measuring the time spent in each contact area, most significant. However, this observation has not been a 20622 cm area containing the stimulus cage (see Fig. verified in humans and therefore the inventors first step 1). A percentage score was also computed for the last was to ascertain the existence of an anomaly in human two trials as follows: patients suffering from FXS. 5 [0059] In a first set of experiments, a transcripteomic For trial 2: Sociability score= 1006Tsocial stimulus/ analysis was performed using DNA microarrays (Tsocial stimulus + Tnon-social stimulus), EXON1.0 Affymetrix (GENOTRANS), for the genes For trial 3: Social novelty preference score=FMR1 and Kcnma1, the results were confirmed by RT- 1006Tnovel social stimulus/(Tnovel social stimulus PCR (7900 HT, Applied Biotechnology). + Tfamiliar social stimulus). 10 [0060] Briefly RNA from patients (FXS) and healthy controls was extracted from lymphoblastoid cell lines. [0054] Finally, the total distance moved in the entire [0061] As shown in (Figure 1A), the expression of apparatus was measured in meters in each trial. FMR1 drops over 95% in patients with FXS, which con- firms the phenotype of fragile X syndrome. Moreover, no • Spontaneous alternation 15 difference in expression of the KCNMA1 gene was ob- served between patients and healthy controls. Apparatus. [0062] Proteomic analysis was also performed by Western blot for FMRP and KCNMA1, using the protein [0055] Spontaneous alternation was assessed in a extracts from the same lymphoblastoid cell lines. This grey, plastic Y-maze, placed on a table 80 cm high and 20 analysis, showed a reduction of over 45% of the protein located in the middle of a room containing a variety of KCNMA1in FXS patients(Figure 1B), thisreduction com- extra maze cues. The three arms of the Y-maze were parable to that observed in FMR1 KO mice by Liao L et similar in appearance and spaced at 120 cm from each al, 105:15281-15286 PNAS (2008). other. Each arm was 42 cm long and 8 cm wide. The [0063] In conclusion, therefore the inventors have con- entire maze was enclosed by a wall 15 cm high and 0.5 25 firmed that an anomaly in KCNMA1 (BKCa) is found in cmthick. Tracking images from a camera above themaze FXS patients. were analyzed with Ethovision. [0064] The inventors therefore have tested whether that a molecules which can open the BKCa channels Procedure. such as BMS-204352, may be used as a therapeutic 30 agent for treating human FXS. [0056] Mice were habituated to the experimental room as before and then introduced at the end of one of the 2. Study of the effects of fluoro-oxindoles on FMR1 arms and allowed to explore the maze for 5 min. Alloca- KO Mice tion of the start arm was counterbalanced within experi- mental groups. Variables measured. An entry into one 35 2.1 Phenotype Studies of the arms was scored by an observer unaware of the genotype of the animals when all four paws of the animal [0065] The similarities between the behavior observed were placed inside an arm. Thus, if an animal made the in FMR1 KO mice and FXS patients, make this mouse following sequence of arm choices: A, B, C, B, A, B, C, strain a good model to study the behavioral features of A, the total number of alternation opportunities would be 40 FXS syndrome (The Dutch-Belgian Fragile X Consor- six (total entries minus two) and the percentage alterna- thium 1994, and Crusio Bernardet 2006). tion would be 67% (four out of six). [0066] FMR1 KO mice are available from the Centre de Neurosciences Intégratives et Cognitives, Bordeaux Results (CNRS UMR 5228). 45 [0067] This is strain was obtained by the standard pro- 1. Levels of KCNMA1 in human FXS and non-FXS tocolof "The Dutch- Belgian Fragile X Consortium (1994)" subjects by inserting a neomycin cassette in exon 5 of the FMR1 gene by homologous recombination, inactivating the [0057] The central role of FMRP as regulator of neu- gene. ronal translation means that any change in its function 50 [0068] Froma phenotypic point ofview, FMR1 KO mice contributes to disruption of the translation of other pro- exhibit characteristics of FXS subjects. Thus, male mice teins and thus via this cascade of disruption alters the have macroorchidism similar to human FXS subjects and protein profile of the cell. also display similar behaviors such as increased levels [0058] This effect has been clearly shown (Liao L et of anxiety, stereotypical/repeated behaviors and im- al, 105:15281-15286 PNAS (2008)), in a primary culture 55 paired social interaction. of neurons from the FMR1 KO mouse (a mouse model [0069] The inventors have also validated this mouse of FXS). Among the variations observed, a 50% decrease model from a molecular point of view by analyzing the in the level of the protein KCNMA1 appears to be the expression of the FMR1 and KCNMA1 genes/ gene prod-

7 13 EP 2 540 295 A1 14 ucts. [0080] The data presented in Figure 4 shows that: [0070] As shown in Figure 2, the inventors observed that wild type mice show greatly increased levels of ex- a. There is a difference of "social interaction" (p = pression of the FMR1 gene in comparison to the FMR1 0.036) between wild type mice and untreated mice KO mice, whatever the tissue studied (brain, testis, liver). 5 kb FMR1 (Figure 4A), The residual expression of the FMR1 gene in FMR1 KO b. That the injection of a single dose of BMS- 204352 mice probably corresponds to a form of "splicing" of the (2mg/kg) in adult FMR1 KO mice (3.5 months) sig- untranslated FMR1 mRNA. nificantly improved (p = 0.037) its interest in the new [0071] In comparison and similarly to that seen in hu- mouse (Figure 4B), mans, the KCNMA1 gene, is highly expressed in the brain 10 c. Statistical analysis shows no significant difference and testis and less in other tissues (Figure 2B) and shows between wild-type mice treated and untreated (p = no difference in expression between wild type mice and 0.22) (Figure 4C), FMR1 KO mice. The amount of protein KCNMA1 pro- d. The improvement observed in FMR1 KO mice is duced is however decreased by 42% (Figure 2C and D) not a consequence of an increase in overall locomo- in FMR1 KO mice brain, validating the theory that FMRP 15 tive activity (Figure 4D). reductionleads to decrease of the translation ofKCNMA1 mRNA. 2.2.2 Test of female interaction [0072] In a parallel series of experiments to better un- derstand the operation of the BKCa channel in connec- [0081] In this test, a female mouse of another strain, tion with FMRP, the inventors studied by immunohistol- 20 but of the same age is introduced in the environment of ogy the distribution in the mouse brain of FMRP and the mouse tested. KCNMA1 in wild mice. [0082] Affiliative behavior of the mice tested (Sniffing, [0073] From Figure 3, the BKCa channel is expressed grooming ...) is recorded over the course of a 5 minute in the cerebellum, in Purkinje cells (right panels C and period. D, arrows), but also in the cortex, brainstem, and the25 [0083] The data presented in Figure 5. show: hippocampus. Furthermore, FMRP is in the brain, but exclusivelylocated inside neurons(Figure 3, leftcolumn). 1. That there is a difference "direct social interaction" (p = 0.04) between untreated wild type mice and un- 2.2 Behavioral Studies treated FMR1 KO mice, and 30 2. That the injection of a single dose of BMS 204 352 [0074] To study the therapeutic effect of the molecule (2mg/kg) in adult FMR1 KO mice significantly im- BMS-204352, the inventors focused on behavioral pa- proves social interaction (p = 0.04). rameters that have the greatest effect upon impairment in patients with FXS and mouse FMR1 KO, namely social 2.2.3 Y Maze Test interactions. 35 [0075] To do this, the inventors used three behavioral [0084] This test involved placing the subject animal in tests: a maze which is in the form of a Y, but intitially has one closed arm. 1. Social preference test (the Crawley test), [0085] After 5 minutes of habituation in the initial maze, 2. Test for direct social interaction (interaction with 40 the third arm is opened and the time the animal spends a female), and in the new arm is recorded. 3. Preference for novelty using a Y maze. [0086] The data presented in Figure 6. show:

[0076] Each of these experiments was performed us- 1. There is a difference of "preference for novelty" ing adult male mice (3.5 months) with or without intra- 45 (p = 0.04) between wild mice and mouse FMR1 KO peritoneal injection of BMS-204352 at 2 mg/kg. untreated, and 2. That the injection of a single dose of BMS -204 2.2.1 Crawley Test 352 (2mg/kg) in adult FMR1 KO mice significantly enhances their preference for novelty (p = 0.04). [0077] In this test, the mouse is introduced into an en- 50 vironment in which two other mice are present, these two [0087] Thus, the injection of a single dose of BMS- other mice being called the "stimuli". 204352 to adult male FMR1 KO mice significantly im- [0078] One of the stimuli mice is already known to the proves their performance in the three different behavioral subject mouse and the second is completely unknown. tests. This result confirms the inventors hypothesis that [0079] The parameter measured is the time spent with 55 cognitive and behavioral problems are at least partly re- each of the two stimuli mice. A wild type mouse would lated to an abnormality of BKCa and that therefore a ther- be expected to spend a longer time with the unknown apy which alleviates the effects of the BKCa dysfunction mouse. also can alleviate the cognitive and behavioral problems.

8 15 EP 2 540 295 A1 16

[0088] Furthermore, the observation of improvement (3S)-(+)-(5-chloro-2-methoxyphenyl)-1, 3-dihydro- after single-dose treatment of adult mice suggests that 3-chloro-6- (trifluoromethyl)-2H-indol-2-one; and behavioral problems are wholly or partly a consequence (3R)-(-)-(5-chloro-2-methoxyphenyl)-1,3-dihydro-3- of abnormal function and not of a structural change in chloro-6-(trifluoromethyl)-2H-indol-2-one. the BKCa channel. 5 [0089] The present invention therefore provides the basis of a potential way to treat FXS.

References 10 [0090]

1. Liao L, et al. PNAS, 105:15281-15286 (2008). 2. Laumonnier F, et al. Am J Psychiatry, 163: 1622-1629 (2006). 15 3. Jensen BS, CNS drug reviews, 8: 353-360 (2002). 4. Gribkoff VK, et al. Nature medicine 2001, 7: 471-477. 5. Zoute L, et al. Organic & biomolecular chemistry, 1:1833-1834 (2003). 20 6. Hamashima Y, et al, Journal of the American Chemical Society, 127:10164-10165 (2005). 7. Sausbier M, et al. PNAS, 101:9474-9478 (2004). 8. Bernardet M, et al. TheScientificWorldJournal, 6: 1164-1176 (2006). 25 9. Consorthium TD-BFX: Fmr1 knockout mice: a model to study fragile X mental retardation. The Dutch-Belgian Fragile X Consortium. Cell, 78: 23-33 (1994). 10. Huber K.M. et al. Proc. Natl. Acad. Sci. U. S. A. 30 99, 7746-7750 (2002). 11. Dölen G. et al. Neuron 56, 955-962 (2007). 12. V. K. Gribkoff et al., Adv. Pharmacol., 37: 319-348 (1997). 35

Claims

1. An opener of a maxi- K potassium channel for use in the treatment of fragile X syndrome. 40

2. The use according to claim 1 wherein said maxi-K potassiumchannel opener is selected from the group comprising a fluoro-oxindole or a chloro-oxindole. 45 3. The use according to claim 2 wherein said fluoro- ox- indole is selected from the group comprising: ( 6)-(5- chloro-2-methoxyphenyl)-1, 3-dihydro-3-fluoro-6-(tri- fluoromethyl)-2H-indol-2-one; (3S)-(+)-(5-chloro-2- methoxyphenyl)-1, 3-dihydro-3-fluoro-6- (trifluoro- 50 methyl)-2H-indole-2-one and (3S)-(-)-(5-chloro-2- methoxyphenyl)-1, 3-dihydro-3-fluoro-6- (trifluoro- methyl)-2H-indole-2-one.

4. The use according to claim 2 wherein said chloro- 55 oxindole is selected from the group comprising: ((6)-(5-chloro-2-methoxyphenyl)-1, 3-dihydro-3- chloro-6- (trifluoromethyl)-2H-indol-2-one);

9 EP 2 540 295 A1

10 EP 2 540 295 A1

11 EP 2 540 295 A1

12 EP 2 540 295 A1

13 EP 2 540 295 A1

14 EP 2 540 295 A1

15 EP 2 540 295 A1

16 EP 2 540 295 A1

17 EP 2 540 295 A1

18 EP 2 540 295 A1

19 EP 2 540 295 A1

20 EP 2 540 295 A1

REFERENCES CITED IN THE DESCRIPTION

This list of references cited by the applicant is for the reader’s convenience only. It does not form part of the European patent document. Even though great care has been taken in compiling the references, errors or omissions cannot be excluded and the EPO disclaims all liability in this regard.

Patent documents cited in the description

• US 5565483 A [0024] • WO 0230868 A [0027] • US 5602169 A [0024]

Non-patent literature cited in the description

• HUBER K.M. et al. Proc. Natl. Acad. Sci. U. S. A., • GRIBKOFF VK et al. Nature medicine, 2001, vol. 7, 2002, vol. 99, 7746-7750 [0011] [0090] 471-477 [0090] • DÖLEN G. et al. Neuron, 2007, vol. 56, 955-962 • ZOUTE L et al. Organic & biomolecular chemistry, [0011] [0090] 2003, vol. 1, 1833-1834 [0090] • V. K. GRIBKOFF et al. Adv. Pharmacol., 1997, vol. • HAMASHIMAY etal. Journal of the American Chem- 37, 319-348 [0028] [0090] ical Society, 2005, vol. 127, 10164-10165 [0090] • B. S. JENSEN. CNS Drug Reviews, 2002, vol. 8 (4), • SAUSBIER M et al.PNAS, 2004, vol. 101, 353-360 [0029] 9474-9478 [0090] • LIAO L et al. PNAS, 2008, vol. 105, 15281-15286 • BERNARDET M et al. TheScientificWorldJournal, [0062] [0090] 2006, vol. 6, 1164-1176 [0090] • LAUMONNIER F et al. Am J Psychiatry, 2006, vol. • The Dutch-Belgian Fragile X Consortium. Cell, 1994, 163, 1622-1629 [0090] vol. 78, 23-33 [0090] • JENSEN BS. CNS drug reviews, 2002, vol. 8, 353-360 [0090]

21