EP4680230A1

EP4680230A1 - Method of treating tuberous sclerosis complex or epilepsy and composition for use therein

Info

Publication number: EP4680230A1
Application number: EP24775441.9A
Authority: EP
Inventors: Yijuang Chern; Christine Chin-jung HSIEH; Yi-Chao Lee; You-Yin Chen; Jim-Min Fang
Original assignee: Academia Sinica
Current assignee: Academia Sinica
Priority date: 2023-03-17
Filing date: 2024-03-15
Publication date: 2026-01-21
Also published as: WO2024196760A1; AU2024241207A1; CN121263180A; JP2026510724A; TW202500138A; KR20250165367A

Abstract

A method of treating tuberous sclerosis complex (TSC) or epilepsy is provided, including administrating to a subject a compound of formula (I), (II) or (III): (I) (II) (III) or a pharmaceutically acceptable salt thereof, wherein X is halogen. A composition for use in a method of treating TSC or epilepsy, including administering to a subject in need thereof the composition comprising a compound of formula (I), (II) or (III) as shown above is also provided.

Description

METHOD OF TREATING TUBEROUS SCLEROSIS COMPLEX OR EPILEPSY

AND COMPOSITION FOR USE THEREIN

FIELD OF INVENTION

[1] The present disclosure relates to a method of treating tuberous sclerosis complex (TSC) or epilepsy and a composition for use in a method of treating TSC or epilepsy.

BACKGROUND OF THE INVENTION

[2] Tuberous sclerosis complex (TSC) is a hereditary disease caused by the gene mutation in either TSC1 or TSC2, which encodes hamartin and tuberin, respectively. These two proteins function together to repress mTOR signaling by inhibiting the small G protein Rlieb that directly activates mTOR. Hyperactivation of mTOR in TSC leads to the formation of hamartomas in multiple organs such as heart, kidney, lung, skin, eyes, and brain as clinical manifestations. The symptoms of TSC are highly variable. The brain pathology is the leading cause of morbidity and mortality. Up to 90% of the patients exhibit neurological symptoms include seizures, such as infantile spasm or status epilepticus, and other cognitive, psychiatric or behavioral deficits.

[3] Among several neurological features, the clinical phenotypes of neurocognitive and neuropsychiatric deficits such as autism spectrum disorders (ASD), attention deficit hyperactivity disorder (ADHD), learning and cognitive impairment, disruptive behaviors and emotional problems are the top problematic but untreatable symptoms of TSC by the current medication. These features are collectively known as TSC-associated neuropsychiatric disorders (TANDs). Most individuals with TSC manifest at least one or more TAND symptoms throughout their lifetime. Individuals with TSC have a relatively high percentage for ASD (up to 40%), as compared to the general population, which is less than 2%. ADHD is also quite common in TSC; and the estimate for ADHD prevalence in TSC is 21% to 50%. Approximately 44% to 64% of TSC patients have intellectual disability, and about one-third of TSC children showed learning difficulties. Other behavioral and emotional problems are also noted in TSC individuals, such as aggression (13%-58%), self-injurious behavior (27%- 41%), anxiety (13%-48%) and depression (19%-43%). The diversity of these clinical manifestations suggested the variability of the gene mutations involved in this disease, which make the treatment for TSC difficult. [4] Epilepsy is commonly found in TSC patients (-90%) during their lifetime. About 70% of these patients are classified as intractable epilepsy, of which patients show no responses to the current antiepileptic drugs (AEDs). Since loss-of-function mutations in TSC 1 or TSC2 lead to constitutive activation of the mTOR, mTOR inhibition has been the main therapeutic strategy in the TSC treatment, including TSC-epilepsy. mTOR inhibitors such as sirolimus (rapamycin) and everolimus have been approved by U.S. FDA in treating TSC-associated epilepsy. However, only 40% of TSC refractory epileptic patients responded to the treatment. In addition, as mTOR inhibitors directly act on cell proliferation and growth, it raises concerns for young children and infants for taking them, especially for long-term use.

[5] Currently, no treatments are available for TSC-related neuropsychiatric disorders and sleeping problems. In addition, conventional anti-epileptic drugs cause great side-effects. One-third of epilepsy patients show no responses to these drugs, and even two-thirds of TSC patients show no therapeutic effect. Effects of conventional mTOR inhibitors on TSC-related refractory epilepsy are limited. Taken together, a safe and effective alternative therapeutics is required for TSC in treating TANDs, and TSC-associated sleep disturbances and epilepsy.

SUMMARY OF THE INVENTION

[6] As described above, one aim of the present invention is to provide a safe and/or effective alternative therapeutics in treating tuberous sclerosis complex (TSC) including TANDs and TSC-associated epilepsy.

[7] The present invention provides a new method of treating TSC by using the compounds of adenosine analogues, an inhibitor of equilibrative nucleoside transporter 1 (ENT1), on the pathogenesis and functional recovery of TSC.

[8] In an aspect of the present invention, a method of treating tuberous sclerosis complex or epilepsy, including administering to a subject in need thereof a compound of formula (I), (II) or (III):

a pharmaceutically acceptable salt thereof, or a composition thereof, wherein X is halogen.

[9] In another aspect of the present invention, a composition for use in a method of treating tuberous sclerosis complex or epilepsy, including administering to a subject in need thereof the composition including a compound of formula (I), (II) or (III) as shown above.

[10] Preferably, the compound is selected from the group consisting of N⁶’-^S-halothien-Z- yl)methyl] adenosine, N⁶-[(4-halothien-2-yl)methyl]adenosine, and N⁶-[(5-halothien-2- yl)methyl] adenosine. More preferably, the compound is selected from the group consisting of N⁶-[(5-iodothien-2-yl)methyl]adenosine, N⁶-[(4-iodothien-2-yl)methyl]adenosine, N⁶-[(3- iodothien-2-yl)methyl]adenosine, N⁶-[(5-bromothien-2-yl)methyl]adenosine, N⁶-[(4- bromothien-2-yl)methyl]adenosine, N⁶-[(3-bromothien-2-yl)methyl]adenosine, N⁶ -[(5- chlorothien-2-yl)methyl]adenosine, N⁶-[(4-chlorothien-2-yl)methyl]adenosine, and N⁶-[(3- chlorothien-2-yl)methyl]adenosine.

[11] Preferably, the compound is selected from the group consisting of N⁶-[(2-halothien-3- yl)methyl]adenosine, N⁶ -[(4-hakithien-3-yl)ineth\4]a€lenosine. and N⁶-[(5-halothien-3- yl)methyl] adenosine. More preferably, the compound is selected from the group consisting of N⁶-[(2-iodothien-3-yl)methyl]adenosine, N⁶-[(4-iodothien-3-yl)methyl]adenosine, N⁶-[(5- iodothien-3-yl)methyl] adenosine, N⁶-[(2-bromothien-3-yl)methyl]adenosine, N⁶-[(4- bromothien-3-yl)methyl]adenosine, N⁶-[(5-bromothien-3-yl)methyl]adenosine, N⁶-[(2- chlorothien-3 -yl)methyl] adenosine, N⁶- [(4-chlorothien-3 -y l)methyl]adenosine, and N⁶- [(5 - chlorothien-3 -yl)methyl] adenosine .

[12] Preferably, the compound, a pharmaceutically acceptable salt thereof, or a composition thereof is administered by an oral, intravenous, intramuscular, subcutaneous, intraperitoneal, or topical route. [13] Preferably, the composition further includes a pharmaceutically acceptable carrier, excipient or vehicle.

[14] Preferably, the epilepsy is TSC-associated epilepsy.

[15] Therefore, the present invention at least provides the following advantages:

1. The claimed method can has a better efficacy in treating cognitive deficits and has no detectable side-effects, when compared to vigabatrin.

2. The claimed method can effectively treat TSC-associated neuropsychiatric disorders (TAND), sleep disturbances caused by TSC, and epilepsy including TSC-associated epilepsy.

BRIEF DESCRIPTION OF THE DRAWINGS

[16] FIGs. 1A-1F illustrates effects of J4 on cognitive deficits and anxiety-like behavior in Tsc2+/- mice according to an embodiment of the present invention. FIG. 1A: The experimental procedures of novel object recognition test. Day 1 was the habituation phase, of which no objects were placed; day 2 was the training phase, of which two same objects were placed; day 3 was the testing phase, of which one familiar object (object A) and one novel object (object B) were placed for the mice to choose from. FIG. IB: The time spent exploring the object A (white bar) and object B (blue bar), in percentage of the total exploration time, was determined for both training phase and the testing phase for the four groups: WT/Veh, WT/J4, Tsc2⁺/- Veh. and Tsc2⁺/-/J4. FIG. 1C: Upper panel, the calculation formula for the discrimination index. Lower panel, the discrimination indices for the four groups. FIG. ID: The schematic diagram of open-field test setup. FIG. IE: Representative tracks for each group were shown. FIG. IF: Time spent in the middle arena as designated in the schematic diagrams shown in (FIG. IE) (red rectangle) was determined for the four groups. Data presented as mean ± SEM.*p < 0.05, n.s. not significant, using unpaired T-test; and One-way ANOVA, with Tukey’s post-hoc test.

[17] FIGs. 2A-2D illustrates effects of J4 on brain microstructural abnormalities in Tsc2+/- mice according to an embodiment of the present invention. FIG. 2A: DKI region-based analysis was performed. Tire indicated brain regions, ACC, EC, CAS, CAI, were shown in coronal view (upper panel), horizontal view (middle panel), and sagittal view (lower panel). FIG. 2B: Mean kurtosis was calculated for each brain region for each group. FIG. 2C: DTI tract-based analysis was performed to examine the integrity of white matter structures, fornix and anterior forceps. FIG. 2D: Fractional anisotropy was determined for these two white matter regions for the four groups. Abbreviations: ACC, anterior cingulate cortex; EC, entorhinal cortex; CA3, cornu ammonis 3; CAI, cornu ammonis 1. Data presented as ± SEM.*/? < 0.05, n.s. not significant, using One-way ANOVA, with Tukey’s post-hoc test.

[18] FIGs. 3A-3K illustrates effects of J4 on the size of dysplastic astrocytes in Tsc2^+,~ mice according to an embodiment of the present invention. FIG. 3A: Fluorescent immunohistochemistry on coronal sections of mice for GFAP and Ibal was performed and hippocampal CAI was visualized. Scale bar: 200 pm; inset, 20 pm. FIG. 3B: GFAP immunoreactivity was compared for WT and Tsc2⁺/- mice with treatments of vehicle or J4 as indicated, in terms of GFAP -positive cells (upper panel), GFAP intensity (middle panel), and area of GFAP immunostaining (lower panel). FIG. 3C: Ibal expression was also compared for the four groups in terms of Ibal -positive cells (upper panel), Ibal intensity (middle panel), and area of Ibal immunostaining (lower panel). FIG. 3D: Hippocampal CA3 was also examined. GFAP intensity (FIG. 3E), area of GFAP immunoreactivity (FIG. 3F) and area of Ibal immunoreactivity (FIG. 3G) was determined for the four groups. FIG. 3H illustrates the schematic diagram of the retrosplenial (RSP) cortex, which was further analyzed. (FIG. 31: Triple labelling of GFAP, Ibal and NeuN was performed to analyze the cell number and morphology of different types of neural cells. Quantitative results of Ibal -positive cells (FIG. J) and NeuN-positive cells (FIG. 3K) were determined for each group. Data presented as mean ± SEM.*p < 0.05, ** p < 0.01, ***p < 0.001, ****p < 0.0001, n.s. not significant, using One-way ANOVA, with Tukey’s post-hoc test.

[19] FIGs. 4A-4C illustrates effects of J4 on the seizure threshold in Tsc2⁺/- mice according to an embodiment of the present invention. FIG. 4A: Seizure susceptibility was determined in WT and Tsc2±'~ mice with treatments of vehicle or J4 as indicated by five intraperitoneal injection of pentylenetetrazol (PTZ) with a dose of 40 mg/kg. The percentage of animals displayed a racine’s score > 4 was shown for every injection. *p < 0.05, using two-tailed Student’s T-test to compare WT/Veh and Tsc2⁺/- /Veh. FIG. 4B: PTZ-induced seizure score was further determined for the four groups during each injection. Data presented as mean ± SEM, *p < 0.05, using two-tailed Student’s T-test to compare WT/Veh and 7s’ Tsc27⁺V/-eh. FIG. 4C: Comparison of the four groups for their PTZ-induced seizure score at the first injection was performed. Data presented as mean ± SEM. *p < 0.05, using One-way ANOVA, with Tukey’s post-hoc test. [20] FIGs. 5A-5C illustrates effects of vigabatrin (VGB) on side-effects in Tsc2⁺'~ mice according to a comparative example of the present invention. FIG. 5A: Novel object recognition test was performed WT and Tsc2^+/~ mice with treatments of vehicle, VGB, or J4 as indicated. The time spent exploring object A or B, presented as percentage of total time, was determined for each group (upper panel). The discrimination index was determined and shown for each group. Data presented as mean ± SEM.*p < 0.05, ** p < 0.01, ***p < 0.001, ****p < 0.0001, n.s. not significant, using unpaired T-test for the comparison of exploring time between objects; and One-way ANOVA, with Tukey’s post-hoc test for multiple comparison of discrimination index. * represents the significance after using post-hoc test to compare with WT/Veh; # represents the significance after post-hoc test to compare with Tsc2⁺/7- Veh. FIG. 5B: The body weight was recorded for the groups as indicated at the beginning of the treatment (at 6-week-old). FIG. 5C: Co-immunostaining of GFAP (red) and Ibal (magenta) on cortical and hippocampal tissues from WT and Tsc2^'~ mice with treatments of vehicle, VGB, or J4 as indicated.

[21] FIGs. 6A-6M illustrates effects of J4 on the survival of Tscl^CKO mice according to an embodiment of the present invention. FIG. 6A: The survival curve showed the survival rate of Tscl^LKO mice with treatments of vehicle or J4 as indicated after gene deletion. FIG. 6B: Western blot data of hippocampal tissue of indicated groups were performed 7 days after gene deletion to visualize the protein expression of S6 and pS6, a-tubulin serves as internal control. FIG. 6C: Quantitative results of the ratio between pS6 and S6 from the Western blot data were performed 7 days after gene deletion. FIG. 6D: Western blot data of hippocampal tissue of indicated groups was performed 21 days after gene deletion to visualize the protein expression of S6 and pS6, a-tubulin was served as an internal control. FIG. 6E: Quantitative results of the ratio between pS6 and S6 from the Western blot data performed 21 days after gene deletion. FIG. 6F: Co-immunostaining of DCX (green) and PV (magenta) of the hippocampal dentate gyrus, and magnified immunostaining results from the indicated rectangles of each group are shown. FIG. 6G. Quantitative bar graphs of PV-positive cell density' (left panel) and DCX- positive area (right panel) are shown. FIG. 6H: Fluorescent immunostaining of GFAP, Sl 00β, and Ibal on cortical layer V for indicated groups is shown. FIG. 61: Quantitative bar graphs of indicated antibodies are shown. FIG. 6J: Nissl staining was performed on the cortex of the indicated groups. FIG. 6K: Quantitative results of the cell size of the cortical layer V are shown. FIG. 6L: Nissl staining was performed on the dentate gyrus hilus of the indicated groups. FIG. 6M: Quantitative results of the cell size of the dentate gyrus hilar cells are shown. Data presented as mean ±SEM. *p < 0.05, ** p < 0.01, ** *p < 0.001, ****p < 0.0001, n.s. not significant, using One-way ANOVA, with Tukey’s post-hoc test for multiple comparison.

[22] FIGs. 7A-7D illustrates effects of J4 on PTZ-induced epileptic seizures and gliosis according to an embodiment of the present invention. FIG. 7A: PTZ at a dose of 35 mg/kg was administered to WT mice with treatments of vehicle (Veh), J4 and vigabatrin (V GB) as indicated. Racine’s score was determined starting from 7^th injection to 21^st injection for each group (left panel). Racine score was compared between the J4 group and the Veh group (right panel). FIG. 7B: The percentage of Racine’s score greater than 4 was determined for the 3 groups (left panel), and compared separately between the J4 group and the Veh group (right panel). FIG. 7C: Co-immunostaining of GFAP and Ibal on the cortical region of WT mice and WT with treatments of Veh, J4, and VGB as indicated. FIG. 7D: Co-immunostaining of GFAP and Ibal on the hippocampal region of WT mice and WT with treatments of Veh, J4, and VGB as indicated. Data presented as mean ± SEM.*p < 0.05, ** p < 0.01, ***p < 0.001, 0.0001, n.s. not significant, using Two-way ANOVA, with Fisher’s LSD post-hoc test for multiple comparison.

[23] FIGs. 8A-8D illustrates effects of J4 on the sleep EEG of Tscl^CKO mice on 22^th day after gene deletion according to an embodiment of the present invention. FIG. 8A: Representation of time-frequency analysis for both electroencephalography, (EEG) and electromyography (EMG) and video determined the stages of awake, non-rapid eye movement (NREM) sleep, and non-rapid eye movement (REM) of WT animal. Awake was characterized by low-amplitude EEG with mixed high-frequency components combined with high-amplitude EMG. NREM sleep was characterized by a relative increase in EEG amplitude consisting of mainly delta and theta frequency components combined with low-EMG tone. REM sleep was characterized by a low-amplitude theta-based EEG combined with muscle atonia and occasional muscle twitches. The amount of awake time (FIG. 8B), the amount of NREM- sleep time (FIG. 8C), and the amount of REM-sleep time (FIG. 8D) were determined in the dark and light phase for WT and Tsci^CKO mice with treatments of Veh or J4 as indicated. The significant differences of amount time of awake, NREM-sleep, and REM-sleep of vehicle- treated WT and J4-treated Tscl^CKO mice vs. vehicle-treated Tscl^CKO mice were indicated by *p<0.05 and ^p<0.05, respectively. These were analyzed by ANOVA with Holm-Bonferroni post-hoc test for multiple comparison. Data was presented as mean ± SEM. DETAILED DESCRIPTION

[24] In one embodiment, a method of treating tuberous sclerosis complex (TSC) or epilepsy is provided, including administrating to a subject a compound of formula (I), (II) or (III): a pharmaceutically acceptable salt thereof, or a composition thereof, wherein X is halogen.

Wherein, the compound of formula (III) is also callas “JMF 1907” herein.

[25] In another embodiment, the compound may be selected from N⁶-[(3-halothien-2- yl)methyl] adenosine, N⁶-[(4-halothien-2-yl)methyl]adenosine, and N⁶-[(5-halothien-2- yl)methyl] adenosine. Preferably, the compound is N⁶ -[(5-iodothien-2-yl)methyl]adenosine, N⁶-[(4-iodothien-2-yl)methyl]adenosine, N⁶-[(3-iodothien-2-yl)methyl]adenosine, ^-[(5- bromothien-2-yl)methyl]adenosine (also called “JMF3464” or “J4”), N⁶-[(4-bromothien-2- yl)methyl] adenosine, N⁶-[(3-bromothien-2-yl)methyl]adenosine, N⁶-[(5-chlorothien-2- yl)methyl] adenosine (also called “JMF3818”), N⁶-[(4-chlorothien-2-yl)methyl]adenosine, N⁶- [(3-chlorothien-2-yl)methyl]adenosine, or a combination thereof.

[26] In another embodiment, the compound may be selected from N⁶-[(2-halothien-3- yl)methyl]adenosine, N⁶-[(4-halothien-3-yl)methyl]adenosine, and N⁶-[(5-halothien-3- yl)methyl] adenosine. Preferably, the compound is N⁶-[(2-iodothien-3-yl)methyl]adenosine, N⁶- [(4-iodothien-3 - yl)methyl] adenosine, N⁶-[(5-iodothien-3-yl)methyl]adenosine, N⁶-[(2- bromo thien-3 -yl)methyl] adenosine, N⁶-[(4-bromothien-S-yl)methyljadenosine, N⁶-[(5- bromothien-3-yl)methyl]adenosine N⁶- [(2-chlorothien-3 -yl)methyl] adenosine. N⁶-[(4- chlorothien-3-yl)methyl]adenosine, or N⁶-[(5-chlorothien-3-yl)methyl]adenosine, or a combination thereof.

[27] In one embodiment, the compound, a pharmaceutically acceptable salt thereof, or a composition thereof is administered by an oral, intravenous, intramuscular, subcutaneous, intraperitoneal, or topical route. [28] Examples

[29] Materials and Methods

[30] Animals

[31] Animals used in this study were treated in accordance with guidelines of the University Committee on the Care and Use of Experimental Animals of Taipei Medical University (Taipei, Taiwan). Mice were housed in an air-conditioned vivarium with free access to food and water and a 12/12-h light/dark cycle. Only male animals (2-3 months old) were used.

[32] The Tsc2⁺/- knockout mouse model (B6;129S4- Tscl^tm1Djk/^fc/J) was purchased from Jackson laboratory (Bar Harbor, ME, USA). The Tscl knockout mouse model ( Tscl^CKO) is generated as follows. The Tscl^{f /}fTg(Camk 2a-CreERT2⁺) mice is first be generated by crossing transgenic mice the carrying tamoxifen inducible CreERT2 under control of the Camk2a promoter (Tg(Camk 2a-cre/ERT2), Jackson Laboratory) with conditional biallelic floxed Tscl mutant mice ( Tscl^tmlD-'^k, Jackson Laboratory).

[33] Drug administration

[34] For Tsc2⁺/- mice, the dosage of J4 administration was 0.06 mg/ml in drinking water containing 1% HPpCD. The administration was started at the age of 6 weeks. The treatment duration was 10 weeks. For Tscl^CKO mice, the dosage of J4 administration was 0.02 mg/ml in drinking water containing 1% HPpCD. The administration was started at the age of 7-9 weeks. The treatment duration was 3 weeks.

[35] Tamoxifen (concentration 20 mg/mL) was dissolved in 90% com oil and 10% ethanol, and then intraperitoneally injected for four consecutive days during light phase of the day (75 mg/kg; day 0-3), when the mice were 7-9 weeks old, to activate CreERT2 to delete Tscl gene in neurons.

[36] Two types of pentylenetetrazol (PTZ) kindling model was used. First, a PTZ kindling model using an every-other-day, low-dose PTZ administration schedule was used.¹ In brief, PTZ was prepared in sterile 0.9% (w/v) NaCl on the day of use at a concentration of 4 mg/ml and injected intraperitoneally at a dose of 40 mg/kg to induce seizures. PTZ was administered every other day when the mice were 9-10 weeks old. A total of 6 injections of PTZ were administered. Second, a similar kindling procedures were followed. To observe the chronic seizures, a dose of 35 mg/kg and at a concentration of 3.5 mg/ml of PTZ was injected every other day for a total of 21 injections when the mice were 7-8 weeks old. [37] Brain slice preparations and immunostaining

[38] Mice were subjected to transcardial perfusion using 10% formalin and then decapitated. After extracted from the skull, the brains were post-fixed with 10% formalin, at 4 °C for overnight. Fixed brains were dehydrated in 30% sucrose in 0.5M PB for 4 days prior to OCT embedding. Sections with 30 pm thickness were obtained using Leica CM 1950 freezing microtome (Leica Biosystems, Wetzlar, Hesse, Germany).

[39] For immunofluorescence staining, the slices were incubated with desired antibodies for overnight at 4°C. After washing with PBS, the slices were incubated with the corresponding Alexa Fluor dye-tagged secondary antibodies at room temperature for 1 hour. After washing by PBS 3 times, tissue slices were mounted onto the slide(s) and mounted with anti-fading mounting medium (Vector Laboratories, Burlingame, CA, USA) and cell nuclei were stained with Hoechst 33258 (Sigma-Aldrich, Missouri, USA). Images were acquired by Leica STP6000 fluorescent microscope (Leica Biosystems), scanned by TissueGnostics (TissueGnostics Gmbh, Vienna, Austria), and visualized with TissueFAXS & HistoFAXS (TissueGnostics Gmbh). For quantitation, for each group, 6 random areas (dimensions of 200 x 200 μm²) of the desired brain region were selected from 3 different animals (N = 3) and analyzed. The immunoreactivity or the intensity of each protein was determined by the “Measure” function of the Fiji/ImageJ software (NIH, Bethesda, MD, USA; https://imagej.net/Fiji). The number of cells was counted by the by the “Analyze Particles” function of the Fiji software and validated by the experimenter.

[40] For Nissl staining, the brain slices were first mounted onto the slides. The slides with the tissues are placed in the cresyl violet acetate solution for 5 minutes. The slides were rinsed briefly with PBS, followed by dehydration in graded alcohols (i.e.,. 50%, 75%, and 95% alcohol). Finally, the slides were cleared in xylene and mounted with mounting medium. Images were acquired by TissueGnostics and visualized with TissueFAXS & HistoFAXS (TissueGnostics Gmbh). For quantitation, for each group, 3 random areas (dimensions of 200 x 200 μm²) of the desired brain region were selected from 3 different animals (N = 3) and analyzed. The number of cells was counted by the by the “Analyze Particles” function of the Fiji software and validated by the experimenter.

[41] PTZ induced seizures

[42] Pentylenetetrazol (PTZ) (Sigma- Aldrich) was prepared by dissolving it in sterile 0.9% (w/v) NaCl on the day of use at a concentration of 3.5 or 4 mg/ml and injected intraperitoneally at a dose of 35 or 40 mg/kg body weight to induce seizures. After PTZ injection, mice were placed in a clear observing cage for 30 min and video recorded for behavioral seizure scoring and seizure frequency quantification. The severity of seizures and scoring were based on published scoring criteria. The Racine scale was modified and described in brief as follows: 0, normal; 1, immobility and lying on belly; 2, head nodding, forelimbs or hindlimbs twitching; 3, myoclonic jerks, tail held up; 4, rearing, clonic seizures, falling on its side; 5, tonic-clonic seizure, wild jumping; 6, death.

[43] Behavioral tests

[44] Before each behavioral test session, mice were habituated in the behavior room for 30 min to 1 h. The novel object recognition (NOR) test consisted of 3 days, and the procedure outline is presented in FIG. 1A. The first day is the habituation, the second day is training, and the third day was the testing. At the first day, during the habituation, mice were allowed to explore freely in an arena with the dimensions of 60 x 60 x 35 cm for 10 min. On the next day, during the training session, mice were placed in the same arena with two identical objects (object A) to familiarize with the object A for 10 min. At the third day, during the testing session, mice were again placed in the same arena with one familiar object (object A) and one novel object (object B). Mice were given 10 min to make a preference choice between two objects. The discrimination index (DI) was calculated by dividing the time difference between the objects A and B by the sum of time exploring the objects A and B (FIG. 1C).

[45] The open-field test (OFT) was carried out using the first day of the NOR test. Mice were allowed to explore the arena freely without any objects for 10 min. A center region with dimensions of 29 x 29 cm was selected (FIG. ID) to observe how much time the animal stayed in the region. Motor tracks of all behavioral tests were video-recorded and analyzed by an open-source Matlab program, OptiMouse^{[3 ]}

[46] MRI acquisition

[47] Magnetic resonance (MR) images were acquired using a 7 Tesla scanner with a 30 cm diameter bore (Bruker Biospec 70/30 USR, Ettlingen, Germany), and the linear volume coil was used to transmit the radio frequency pulses. For the radio frequency signal retrieval, a planar surface coil (T7399V3; Bruker Corp., Billerica, MA, USA) was placed over each mouse’s head. During each MRI session, the subject mouse was anesthetized by inhalation of 3% isoflurane (Attane™ Isoflurane, Minrad Inc., NY, USA), in combination with 20% O2, 75% N2, and 5% CO2. The mouse was fastened to an animal holder and a hot pad set to 37 °C was placed around the abdomen to maintain the body temperature. A life monitoring system and pressure sensor (SA Instruments Inc., New York, NY, USA) were also put under the abdomen of the mouse to monitor the respiratory status. The respiration rate was steady and maintained at between 20 and 40 breaths per minute.

[48] Magnetic field homogeneity was optimized using the fast automated shimming technique by mapping along projections (FASTMAP) with first order shims on an isotropic voxel of 7 x 7 x 7 mm³ encompassing the imaging slices. Turbo spin echo (TSE) T2 images were acquired to check the slice positioning [TR = 2,500 ms, TE = 33 ms, matrix size = 256 x 256 ^x 15, field of view (FOV) = 20 x 20 mm², voxel size = 0.08 x 0.08 x 0.4 mm³, slice thickness = 0.4 mm, 14 horizontal slices]. Diffusion kurtosis images were acquired using DtiEpi SpinEcho sequence. TR = 3750 ms and TE = 31 ms, matrix size = 80 x 80 x 15 pixels, FOV = 20 x 20 mm², slice thickness = 0.4 mm, 15 horizontal slices.

[49] Statistical analysis

[50] For the animal behaviors, NOR discrimination indices, time spent in the middle arena for the OPF were compared among the three groups, WT/Veh, WT/J4, Tsc2⁺/- /Veh, Tsc2⁺/- ZJ4 and using one-way ANOVA and post-hoc analysis of Tukey’s test to determine the statistical significance. For immunostaining image analyses, both the average number of GFAP+ cells, Ibal+ cells, NeuN+ cells, immuno-intensity were assessed using one-way ANOVA with Tukey’s test as post-hoc analysis.

[51] For diffusion kurtosis (DKI) and diffusion tensor imaging (DTI) metrics (MK for ROI- based analysis; FA for tract-based analysis) among the three groups, one-way ANOVA with Fisher’s LSD test as the post-hoc comparison was performed. All the statistical analyses mentioned above were performed using Prism version 8 (GraphPad Software, Inc., San Diego, CA, USA).

[52] Results

[53] J4 ameliorated cognitive deficits and anxiety-like behavior in Tsc2⁺/- mice

[54] The novel object recognition (NOR) test (FIG. 1A) was performed and found that mice treated with vehicle ( Tsc2⁺/- /Veh) were unable to distinguish the familiar object A and the novel object B, while wildtype mice, either treated with vehicle (WT/Veh) or J4 (WT/J4), exhibited significantly higher percentage of exploring time for the novel object B (FIG. IB) When Tsc2⁺/- mice were treated with J4 (Tsc2 ^f /J4; 0.06 mg/ml in drinking water containing 1% HPβCD), the exploring time was increased for the object B. The discrimination index (DI) was calculated for each group. The DI for Tsc2^+/- /Veh group was significantly lower than the other 3 groups (p < 0.05, one- way ANOVA, with Tukey’s post-hoc test), indicating that Tsc2^+/- mice displayed the deficits in learning and memory function. In addition, the treatment of J4 was able to reverse the deficits (FIG. 1C).

[55] Their open-field behaviors are also analyzed by determining the time spent in the center region for each group to determine the anxiety-like behavior for the mice (FIG. ID). The tracks for each group were monitored and analyzed (FIG. IE). From the tracks and the quantification results, Tsc2^+/- /Veh group showed decreased time exploring the center region of the arena, as compared to the WT/Veh and Tsc2^+/-/Veh (p < 0.05, one- way ANOVA, with Tukey’s post-hoc test) (FIG. IF).

[56] J4 ameliorated brain microstructural abnormalities in Tsc2^+/-/ mVieche

[57] Diffusion kurtosis imaging (DKI) analysis was further performed to determine the microstructural integrity of the mice and compare the 3 groups. Several brain regions were analyzed and found that T Tsc2^+/-/Veh mice displayed decreased mean kurtosis (MK) in the following regions: anterior cingulate cortex (ACC), entorhinal cortex (EC), cornu ammonis 1 (CAI), and CA3 (p < 0.05, one-way ANOVA, with Fisher’s LSD post-hoc test) (FIG. 2A). J4 treatment reversed MK in these regions, suggesting that these brain regions were reversible upon the treatment (p < 0.05, one-way ANOVA, with Fisher’s LSD post-hoc test) (FIG. 2B).

[58] Next, diffusion tensor imaging (DTI) tract-based analysis was carried out to delineate the structural connectivity within the brain network. Two white matter (WM) regions, fornix and anterior forceps, represent axon bundles connecting or intertwining gray matter regions such as ACC and the hippocampus (FIG. 2C). It was found that the fractional anisotropy (FA) value was decreased in Tsc2^+/-/Veh mice, and significantly increased after J4 treatment (FIG. 2D) in the fornix, but not anterior forceps. Nevertheless, both DKI and DTI analyses showed that Tsc2^+/-/Veh mice exhibited an abnormal brain microstructure in gray matter (GM) and WM regions, respectively.

[59] J4 reduced the size of dysplastic astrocytes in Tsc2^+/- mice

[60] To find out the possible cellular changes that contribute to the MK changes in MR imaging, immunofluorescence staining was performed with antibodies NeuN, GFAP, Ibal to visualize the number and morphology of neurons, astrocytes, and microglia respectively. The hippocampal CAI region (FIG. 3 A) was first examined. It was found that Tsc2^+/-/Veh mice showed increased GFAP immune-intensity and area of GFAP staining, but not GFAP-positive cells when compared with WT/Veh mice (FIG. 3B). On the other hand, Tsc2^+/-/J4 mice exhibited a decreased area of GFAP staining, but not GFAP intensity, suggesting that the treatment had the ability to reduce the size of GFAP -positive astrocytes, indicating the reduction in astrocytic activation, but not the GFAP expression level (FIGs. 3A and 3B).

[61] Conversely, there was no significant changes in number of Ibal -positive cells, Ibal intensity, or the area of Ibal staining, although an increase trend was observed in Tsc2^+/- 7Veh mice, and a decreased trend was observed in Tsc2^+/-/ J4 mice (FIG. 3C). Further analysis of microglia using other markers is required to confirm the effects of J4 on microglia.

[62] Similarly, in the hippocampal CA3 subfield, the same effect of J4 on GFAP-positive astrocytes was observed (FIGs. 3D-3F). The morphology of Ibal -positive microglia of Tsc2⁺'~ /Veh mice (FIG. 3D, lower panel) was also studied and found that the size and morphology of Ibal -positive cells were indifferent with WT/Veh mice (FIG. 3G). The retrosplenial (RSP) cortical region (FIG. 3H) was further studied to determine the NeuN- and Ibal-positive cells in each group (FIG. 31). It was found that the number of Ibal-positive cells (FIG. 3 J) and NeuN-positive cells (FIG. 3K) did not change significantly in Tsc22⁺7'Veh group when compared to the other groups.

[63] J4 elevated the seizure threshold in Tsc2^+/~ mice

[64] Although the Tsc2^+/- mouse model seems not to have spontaneous seizures, it was speculated that it has a lower seizure threshold when exposed to a chemical assault such as pentylenetetrazole (PTZ). This hypothesis was tested by using a PTZ-induced kindling protocol to observe the seizure score for each group. PTZ was injected every other day for 5 times at a sub-convulsant dose (40 mg/kg, i.p.) and expected WT/Veh to exhibit no convulsive response or signs of myoclonic jerks (stage 1). As expected, the WT/Veh mice showed no response during the first and second injections of PTZ, while Tsc2^+/- /Veh mice showed increased percentage of animals with Racine’s score > 4 (FIG. 4A). At the first injection of PTZ, the seizure behavior of Tsc2^+/- /- Veh mice had an average Racine’s score of 2.8, while that of WT/Veh group was much lower (Racine’s score = 1) (FIG. 4B). When compared to the Tsc2^+/- /- Veh mice, Tsc2^+/-/J4 mice showed more mild seizure behavior (FIG. 4C), indicating the treatment of J4 was able to protect Tsc2+/- mice from PTZ-triggered convulsive behavior.

[65] Vigabatrin caused severe side-effects in Tsc2^+/- mice

[66] The effects of J4 were next compared with an anti-convulsive drug vigabatrin (V GB), which is often used for treating TSC-related seizures, in Tsc2^+/- mice. Tsc2⁺/- mice vigabatrin ( Tsc2⁺/- /VGB) was treated and found that their NOR performance was not reversed like J4 did. (FIG. 5A). Also, vigabatrin caused tremendous weight loss in both WT and Tsc2 " mice (FIG. 5B). In addition, Tsc2⁺/-WGB group showed a substantial augmentation of GFAP-positive astrocytes in the cortical regions, as well as increased Ibal expression in the cortex and the hippocampus, while these side-effects were not seen in Tsc2⁺/-- J4 group (FIG. 5C). These results indicated that J4 has a better efficacy in treating cognitive deficits and has no detectable side-effects, when compared to vigabatrin.

[67] .14 prolonged the survival of Tscl^CKO mice

[68] Next, whether J4 has effects on TSC-related seizures was tested by using a TSC model with spontaneous seizures, Tscl^CKO mouse model. The results showed that J4 was able to prolong the survival of Tscl^CKO mice (FIG. 6A). At day 7 after the removal of Tscl by Tamoxifen, Western blot was performed to determine the levels of pS6 and S6. The vehicle- and J4-treated groups were compared with the WT mice. It was found that J4 treatment was able to reduce the overactivation of pS6 (FIGs. 6B and 6C). Surprisingly, the normalization of the elevated pS6 in Tscl^CKO mice by J4 disappeared when Tsc 1 was deleted by 21 days (FIGs. 6D and 6E) Moreover, J4-treated group showed increased immunoreactivity area of doublecortin (DCX) in the dentate gyrus of hippocampus, where adult neurogenesis takes place, as compared to the Tscl^LKO mice, but no significant changes in parv albumin (PV)-positive cells (FIGs. 6F and 6G). In addition, we also found that J4 also inhibited the proliferative GFAP- positive astrocytes in the cortical layer V of Tscl^CKO mice, which may indicate neuroinflammation after seizures (FIG. 6H and 61). We further found that J4 decreased the size of pyramidal neurons in cortical layer V (FIGs. 6J and 6K) and dentate gyrus hilar cells (FIGs. 6L and 6M). Collectively, the results suggest that J4 transiently inhibited the abnormal activation of pS6 due to the loss of Tscl, increased the adult neurogenesis, reduced reactive astrocytes in cortex, decreased the enlarged cell size, and subsequently extended the shortened lifespan of Tscl^CKO mice.

[69] J4 reduced PTZ-induced epileptic seizures and gliosis

[70] To test whether J4 can be applied to epilepsy in general, a PTZ kindling model was used with an every-other-day, low-dose PTZ administration schedule to examine the effects of J4 and VGB. For each injection, the seizure behavior of the mice was video-recorded and monitored to determine the Racine’s score. At injections of 16^th, 17^th, and 21^st, J4 showed significant differences when compared to the VGB group. Surprisingly, it was noticed that VGB showed a higher average Racine’s score as compared to the Veh group (FIG. 7 A, left panel). Comparing the average score between Veh and J4 separately, it was found that J4 tremendously decreased the average Racine’s score upon each PTZ injection (FIG. 7A, right panel). Similarly, VGB group showed a higher percentage of seizure behavior with a Racine’s score greater than 4, whereas J4 group showed significantly a lower percentage when compared to the Veh and VGB groups (FIG. 7B). Further, double labelling of fluorescent immunostaining of GFAP and Ibal of the cortical region (FIG. 7C) and hippocampal region (FIG. 7D) of each group was performed. The results demonstrated that upon PTZ induction, the Veh and VGB groups showed severe gliosis, while the J4 group exhibited more mild activation of GFAP and Ibal .

[71 ] Effects of J4 on the sleep EEG of Tscl^CKO mice on 22^th day after gene deletion.

[72] The effects of J4 on modulating electroencephalography (EEG) signals during different stages of sleep were further examined. Representation of time-frequency analysis for both electroencephalography, (EEG) and electromyography (EMG) and video determined the stages of awake, non-rapid eye movement (NREM) sleep, and non-rapid eye movement (REM) of WT animal was shown (FIG. 8A). Awake w^fas characterized by low-amplitude EEG with mixed high-frequency components combined with high-amplitude EMG. NREM sleep was characterized by a relative increase in EEG amplitude consisting of mainly delta and theta frequency components combined with low-EMG tone. REM sleep was characterized by a low- amplitude theta- based EEG combined with muscle atonia and occasional muscle twitches. It was found that there was no significant differences in the amount of awake time between vehicle- treated WT (WT/Veh) and Tscl^CKO mice with treatments of vehicle ( Tscl^CKO/Veh) or J4 ( Tscl^CKO/J4) in the light phase. However, the Tscl^CKO/Veh mice presented a significant increase in the amount of awake time as compared with those of WT/Veh and Tscl^CKO/J4 mice in the dark phase (FIG. 8B). It was noticed that there was much shorter amount of NREM- sleep time for Tscl^CKO/Veh as compared with those of WT/Veh and Tscl^CKO/J4 groups in the dark phase. In light phase, there was no difference found in three groups (FIG. 8C). REM sleeping was significantly decreased in Tscl^CKO/Veh group as compared with WT/Veh and Tscl^CKO/J4 groups in both the dark and light phases (FIG. 8D). The significant differences of amount time of awake, NREM-sleep, and REM- sleep of WT/Veh and Tscl^CKO/J4 versus Tscl^CKO/Veh were indicated by *p<0.05 and p<0.05, respectively. These were analyzed by ANOVA with Holm-Bonferroni post hoc test for multiple comparison. Data was presented as mean ± SEM. [73] Conclusion

[74] According to the results above, it was indicated that J4 treatment improved the cognitive function and ameliorated the anxiety-like behavior of Tsc2⁺/- mice. J4 treatment showed efficacy on lowering the seizure threshold in Tsc2⁺/- mice upon pentylenetetrazole (PTZ)-induced seizures and lower adverse effects, when compared to a commonly used AED, Vigabatrin. Moreover, by incorporating non-invasive neuroimaging modalities, diffusion kurtosis and diffusion tensor imaging, it was found that J4 treatment was able to reverse the brain microstructural abnormalities and enhance the connectivity in Tsc2⁺/- mice. Furthermore, J4 on a TSC model with spontaneous seizures, Tscl^LKO mouse model, was tested and the result showed that J4 treatment extended the lifespan. Thus, the present invention provides a new alternative treatment for TSC for treating both TSC-related psychiatric disorders and epilepsy. Consequently, these results suggest that the compounds of the present application are potential candidates for the treatment of TSC.

REFERENCES

1. Dhir, A., Pentylenetetrazol (PTZ) kindling model of epilepsy. Curr Protoc Neurosci, 2012. Chapter 9: p. Unit 9.37. . Shimada, T. and K. Yamagata, Pentylenetetrazole-Induced Kindling Mouse Model. J Vis Exp. 2018(136).

3. Ben-Shaul, Y., OptiMouse: a comprehensive open source program for reliable detection and analysis of mouse body and nose positions. BMC Biol, 2017. 15(1): p. 41. . Koene, L.M.C., et al., Effects of antiepileptic drugs in a new TSC/mTOR-dependent epilepsy mouse model. Ann Clin Transl Neurol, 2019. 6(7): p. 1273-1291.

Claims

WHAT IS CLAIMED IS:

1. A method of treating tuberous sclerosis complex or epilepsy, comprising administering to a subject in need thereof a compound of formula (I), (II) or (III): a pharmaceutically acceptable salt thereof, or a composition thereof, wherein X is halogen.

2. The method of claim 1 , wherein the compound is selected from the group consisting of N⁶- [(3-halothien-2-yl)methyl]adenosine, N⁶-[(4-halothien-2-yl)methyl]adenosine, and N⁶- [(5-halothien-2-yl)methyl] adenosine.

3. The method of claim 2, wherein the compound is selected from the group consisting of N⁶- [(5-iodothien-2-yl)methyl]adenosine, N⁶-[(4-iodothien-2-yl)methyl]adenosine, N⁶-^- iodothien-2-yl)methyl]adenosine, N⁶-[(5-bromothien-2-yl)methyl]adenosine, N⁶-[(4- bromothien-2-yl)methyl]adenosine, N⁶-[(3-bromothien-2-yl)methyl]adenosine, ^-[(5- chlorothien-2-yl)methyl]adenosine, N⁶-[(4-chlorothien-2-yl)methyl]adenosine, and N⁶- [(3-chlorothien-2-yl)methyl]adenosine.

4. The method of claim 1 , wherein the compound is selected from the group consisting of N⁶- [(2-halothien-3 -yl)methyl] adenosine, /V'- [(4-hal oth ien-3 -y Ijmethy l]adenosine, and N⁶- [(5-halothien-3 -yl)methyl] adenosine .

5. The method of claim 4, wherein the compound is selected from the group consisting of N⁶- [(2-iodothien-3-yl)methyl]adenosine, N⁶- [(4-iodothien-3 -yl)methyl] adenosine, N⁶-[(5- iodothien-3-yl)methyl]adenosine, N⁶-[(2-bromothien-3-yl)methyl]adenosine, N⁶-[(4- bromothien-3 -yl)methyl] adenosine, N⁶-[(5-bromothien-3 -yl)methyl] adenosine N⁶- [(2- chlorothien-3-yl)methyl]adenosine, N⁶-[(4-chlorothien-3-yl)methyl]adenosine, and N⁶- [(5-chlorothien-3-yl)methyl]adenosine.

6. The method of claim 1 , wherein the compound, a pharmaceutically acceptable salt thereof, or a composition thereof is administered by an oral, intravenous, intramuscular, subcutaneous, intraperitoneal or topical route.

7. The method of claim 1, wherein the composition further comprises a pharmaceutically acceptable carrier, excipient or vehicle.

8. The method of claim 1, wherein the treating tuberous sclerosis complex comprises treatment for tuberous sclerosis complex (TSC)-associated neuropsychiatric disorders (TAND).

9. The method of claim 1, wherein the treating tuberous sclerosis complex comprises treatment for sleep disturbances caused by tuberous sclerosis complex.

10. The method of claim 1, wherein the epilepsy is TSC-associated epilepsy.

11. A composition for use in a method of treating tuberous sclerosis complex or epilepsy, comprising administering to a subject in need thereof the composition comprising a compound of formula (I), (II) or (III) : or a pharmaceutically acceptable salt thereof, wherein X is halogen.

12. The composition for use of claim 11,, wherein the compound is selected from the group consisting of N⁶-[(3-halothien-2-yl)methyl]adenosine, N⁶ -[(4-halothien-2- yl)methyl]adenosine, and N⁶-[(5-halothien-2-yl)methyl]adenosine.

13. The composition for use of claim 12, wherein the compound is selected from the group consisting of N⁶ [(5-iodothien-2-yl)methyl]adenosine, N⁶-[(4-iodothien-2- yl)methyl]adenosine, N⁶-[(3-iodothien-2-yl)methyl]adenosine, N⁶-[(5-bromothien-2- yl)methyl]adenosine, N⁶-[(4-bromothien-2-yl)methyl]adenosine, N⁶- [(3 -bromothien-2- yl)methyl]adenosine, N⁶-[(5 -chloro thien-2 -yl)methyl] adenosine, N⁶-[(4-chlorothien-2- yl)methyl]adenosine, and N⁶-[(3-chlorothien-2-yl)methyl]adenosine.

14. The composition for use of claim 11, wherein the compound is selected from the group consisting of N⁶-[(2-halothien-3-yl)methyl]adenosine, N⁶-[(4-halothien-3- yl)methyl]adenosine, and N⁶-[(5-halothien-3-yl)methyl]adenosine.

15. The composition for use of claim 14, wherein the compound is selected from the group consisting of N⁶-[(2-iodothien-3-yl)methyl]adenosine, N⁶-[(4-iodothien-3- yl)methyl]adenosine, N⁶-[(5-iodothien-3-yl)methyl]adenosine, N⁶-[(2-bromothien-3- yl)methyl]adenosine, N⁶- [(4-bromothien-3 -yl)methy 1] adenosine, N⁶- [(5-bromothien-3 - yl)methyl]adenosine N⁶-[(2-chlorothien-3-yl)methyl]adenosine, N⁶-[(4-chlorothien-3- yl)methyl]adenosine, and N⁶-[(5-chlorothien-3-yl)methyl]adenosine.

16. The composition for use of claim 11, wherein the composition is administered by an oral, intravenous, intramuscular, subcutaneous, intraperitoneal, or topical route.

17. The composition for use of claim 11, wherein the composition further comprises a pharmaceutically acceptable carrier, excipient or vehicle.

18. The composition for use of claim 11, wherein the treating tuberous sclerosis complex comprises treatment for tuberous sclerosis complex (TSC)-associated neuropsychiatric disorders (TAND).

19. The composition for use of claim 11, wherein the treating tuberous sclerosis complex comprises treatment for sleep disturbances caused by tuberous sclerosis complex.

20. The composition for use of claim 11, wherein the epilepsy is TSC-associated epilepsy.