WO2011025230A2 - Gaba release-regulating agent in cerebellum - Google Patents

Abstract

A GABA (gamma-aminobutyric acid) release-inhibiting agent in the cerebellum and a composition for treating pathological symptoms caused by over-release of GABA in the cerebellum, each comprising a Bestrophin 1(Best1) channel inhibitor as an active ingredient; a GABA release-promoting agent in the cerebellum and a composition for treating pathological symptoms caused by the deficit of GABA in the cerebellum, each comprising a Best1 channel activator as an active ingredient; and a method for screening a GABA release-regulating agent in the cerebellum, which uses Best1 channel as target, are provided.

Description

TITLE OF THE INVENTION

GABA RELEASE-REGULATING AGENT IN CEREBELLUM

FIELD OF THE INVENTION

The present invention relates to a GABA (gamma-aminobutyric acid) release- inhibiting agent in the cerebellum and a composition for treating pathological symptoms caused by over-release of GABA in the cerebellum, each comprising a Bestrophin l(Bestl) channel inhibitor as an active ingredient; a GABA release-promoting agent in the cerebellum and a composition for treating pathological symptoms caused by the deficit of GABA in the cerebellum, each comprising a Bestl channel activator as an active ingredient; and a method for screening a novel GABA release-regulating agent in the cerebellum, which uses Bestl channel as target.

BACKGROUND OF THE INVENTION GABA is a major inhibitory neurotransmitter in the central nervous system of mammals.

GABA is known to act through two modes of action; tonic and phasic modes. Although it is well established that the mechanism of phasic release of GABA involves a Ca ⁺ dependent vesicular release, the source and the mechanism of tonic GABA release still remains a subject of much speculation.

The present invention has been completed by identifying a mechanism for tonic GABA release.

SUMMARY OF THE INVENTION

An embodiment provides a cerebellar GABA release-regulating agent, which comprises a Bestl channel regulator as an active ingredient.

In particular, the embodiment provides a cerebellar GABA release-inhibiting agent, which comprises a Bestl channel inhibitor as an active ingredient.

In addition, the embodiment provides a cerebellar GABA release-promoting agent, which comprises a Bestl channel activator as an active ingredient.

Another embodiment provides a method of regulating a GABA release in cerebellum, by regulating a Bestl channel activity. More specifically, the embodiment provides method of inhibiting a GABA release in cerebellum, by inhibiting a Bestl channel activity.

Alternatively, the embodiment provides a method of promoting a GABA release in cerebellum, by activating a Bestl channel activity.

An embodiment of the present invention provides a use of a Bestl channel regulator as an active ingredient for regulating a GABA release in cerebellum.

In particular, the embodiment provides a use of a Bestl channel inhibitor as an active ingredient for inhibiting a GABA release in cerebellum.

In addition, the embodiment provides a use of a Bestl channel activator as an active ingredient for promoting a GABA release in cerebellum.

Another embodiment of the present invention provides a composition for preventing, improving, alleviating, and/or treating a disease or a symptom caused by over-release or deficit of GABA, the composition comprising a Bestl channel activity regulator as an active ingredient.

In particular, the embodiment provides a composition for preventing, improving, alleviating, and/or treating a disease or a symptom caused by over-release of GABA, the composition comprising a Bestl channel inhibitor as an active ingredient.

In addition, the embodiment provides a composition for preventing, improving, alleviating, and/or treating a disease or a symptom caused by deficit of GABA, the composition comprising a Bestl channel activator as an active ingredient.

Another embodiment of the present invention provides a method of preventing, improving, alleviating, and/or treating a disease or a symptom caused by over-release or deficit of GABA, using a Bestl channel activity regulator as an active ingredient.

In particular, the embodiment provides a method of preventing, improving, alleviating, and/or treating a disease or a symptom caused by over-release of GABA, using a Bestl channel inhibitor as an active ingredient.

In addition, the embodiment provides a method of preventing, improving, alleviating, and/or treating a disease or a symptom caused by deficit of GABA, using a Bestl channel activator as an active ingredient.

Another embodiment of the present invention provides a use of a Bestl channel activity regulator as an active ingredient for preventing, improving, alleviating, and/or treating a disease or a symptom caused by over-release or deficit of GABA. In particular, an embodiment of the present invention provides a use of a Best 1 channel inhibitor as an active ingredient for preventing, improving, alleviating, and/or treating a disease or a symptom caused by over-release of GABA.

In addition, an embodiment of the present invention provides a use of a Bestl channel activator as an active ingredient for preventing, improving, alleviating, and/or treating a disease or a symptom caused by deficit of GABA.

Another embodiment of the present invention provides a method for screening a cerebellar GABA release-regulating agent by contacting a candidate to a cerebellar sample and determining the extent of Bestl channel activation thereafter.

DETAILED DESCRIPTION OF THE INVENTION

The present inventors found that tonic GABAergic inhibition is due to a release of GABA from the cerebellar glial cells via an anion Bestrophin 1 (Bestl) channel. The use of a two-cell sniffer patch technique has confirmed that Bestl channel allows a direct permeation of GABA. Secondly, by employing cell-type specific gene slicing technique and latest optogenetic tool as well as conventional electrophysiolological approach to detect tonic GABA release in adult mice, the present inventors identified that glial cells contain GABA which can be released through Bestl channel and that the release is inhibited by various anion channel inhibitors. The GABA release was found to significantly decrease by a knock-down of targeted gene after lenti viral Bestl-shRNAs were injected into the cerebellar region.

Finally, by combining the cre-lox regulated shRNA system with a hGFAP- CreERT2 transgenic mouse, the present inventors confirmed that attenuation of tonic GABA current due to gene silencing is fully rescued, which indicates that GABA release from glial cells is responsible for ambient GABA. These findings unprecedently conceptualize the role of glial cells and a non- vesicular channel-mediated release mechanism in releasing GABA and highlight the importance of glial integration of neuronal processing.

Since the first observation of tonic GABA current in dentate gyrus granule cells, tonic inhibition has been reported to be distributed differentially throughout the central nervous system including cerebellum, hippocampus, thalamus, cortex, brainstem, and etc. Tonic inhibition dominates over phasic inhibition in controlling the general tone of excitability and carries an important role in information processing of neuronal outputs. The diverse functional roles of tonic inhibition have been implicated in epilepsy, sleep, memory and cognition (Walker and Semyanov, 2008). In the cerebellum, granule cells that provide major excitatory input to Purkiηje cells are highly restrained by continuing tonic GABA inhibition, which is mediated by the extrasynaptic, high-affinity subunit- containing GABA_A receptors (Hamann et al., 2002; Rossi et al., 2003). Tonic GABA inhibition in the cerebellum has been reported to be a critical target for low dose alcohol intoxication that impairs motor behavior (Hanchar et al., 2005). However, its functional significance has been explored only in a limited number of studies, partly due to the lack of understanding of the release mechanism.

Cerebellar granule cells form a unique configuration, called type II glomerulus, and allows accumulation of ambient extracellular GABA along with the glutamatergic mossy fiber, the axons of Golgi cell, the granule cell dendrites, and glial sheaths; the glomerulus is completely enclosed with lamella glial sheaths that retain released GABA. Bergmann glial cells, another unique type of astrocyte in the cerebellum, are located proximal to the Purkinje cells (Fig Ia) and play an important developmental role of providing a scaffold for postnatal granular cell migration and Purkinje cell dendrite maturation. In adults, Bergmann glial cells remain to be a close anatomical and functional partner to Purkinje cells by tightly enwrapping both its somata and synapses in a glial sheath that contains GABAergic synaptic termination from Basket cells and excitatory synaptic terminations from granule cells.

GABA is thought to be synthesized, contained, and released exclusively by neurons in adult brains, but some reports suggest that astrocytes in brainstem and cerebellum contain GABA. In order to determine the presence of GABA in glial cells of adult cerebellum, immunohistochemistry for GABA was performed in GFAP-GFP transgenic mice, in which the somas and the fine processes of GFAP positive astrocytes were labelled with GFP. The immunohistochemistry showed a strong immunoreactivity for antibody against GABA in the somas and the processes of all the GFP-positive Bergmann glial cells as well as in the lamella astrocytes in the granule cell layer (See Fig.5a and b). Notably, the intensity of glial GABA immunoreactivity was in equal or greater magnitude compared to the neighboring neurons. These results provide a potential support for glial release of GABA.

Previous studies have reported that the tonic activation of GABA_A receptors in cerebellar granule cells of adult rats results from an action-potential-independent, nonvesicular release of GABA. These findings are in line with the idea that the source of ambient GABA may possibly be located in glia. Moreover, in a cell line derived from type-2 astrocytes, the activation of purinergic receptor P2X₇ was found to induce the release of [³H]-GABA, which was unexpectedly sensitive to inhibitors of volume- regulated anion channels or HCO₃VCl^" exchangers such as DIDS(4,4'- diisothiocyanatostilbene-2,2'-disulfonic acid) and SITS(4-acetamido-4'- isothiocyanostilbene-2,2'-disulphonic acid). In this context, the present inventors subsequently searched for an anion channel that can serve as a molecular target for GABA release.

Based on the overlapping unique feature of Best 1 among anion channels, Bestl was selected to be the candidate anion channel. Human bestrophin-1 (hBestl), among Bestrophins, was cloned to identify a mutation in autosomal dominant Best vitelliform macular dystrophy. It was proven that hBestl constitutes Cl^" channel that is activated by Ca²⁺ as well as volume change and is readily blocked by Niflumic acid (NFA), NPPB (5-nitro-2(3-phenylpropylamino)-benzoic acid), and DIDS (4,4'- diisothiocyanatostilbene-2,2'-disulfonic acid). In addition to showing much higher permeability to larger anions such as SCN- than Cl-, hBestl displays a significant permeability ratio for HCO3- compared to Cl- (P_HCO₃/P_CI⁼0.44).

On the basis of the finding that GABA release is carried out through Bestl channel in the cerebellar glial cells, a disease or a symptom caused by over-release or deficit of GABA can be prevented, mitigated, and treated by the regulation of Bestrophin 1 channel.

Accordingly, an embodiment according to the present invention provides a GABA release-regulating agent, which comprises a Bestl channel activity-regulator as an active ingredient; a method of regulating a GABA release in cerebellum, by regulating a Bestl channel activity; and a use of a Bestl channel regulator as an active ingredient for regulating a GABA release in cerebellum.

Specifically, an embodiment according to the present invention provides a GABA release-inhibitor, which comprises a Bestl channel-inhibitor as an active ingredient; method of inhibiting a GABA release in cerebellum, by inhibiting a Bestl channel activity; and a use of a Bestl channel inhibitor as an active ingredient for inhibiting a GABA release in cerebellum. Alternatively, an embodiment according to the present invention provides a GABA release-promoter, which comprises a Bestl channel-activator as an active ingredient; a method of promoting a GABA release in cerebellum, by activating a Bestl channel activity; and a use of a Bestl channel activator as an active ingredient for promoting a GABA release in cerebellum.

Another embodiment according to the present invention provides a composition for preventing, mitigating, and/or treating a disease or a symptom caused by over- release or deficit of GABA, which comprises a Bestl channel activity-regulator as an active ingredient; a method of preventing, improving, alleviating, and/or treating a disease or a symptom caused by over-release or deficit of GABA, using a Bestl channel activity regulator as an active ingredient; and a use of a Bestl channel activity regulator as an active ingredient for preventing, improving, alleviating, and/or treating a disease or a symptom caused by over-release or deficit of GABA.

Specifically, an embodiment according to the present invention provides a composition for preventing, mitigating, and/or treating a disease or a symptom caused by over-release of GABA, which comprises a Bestl channel-inhibitor as an active ingredient; a method of preventing, improving, alleviating, and/or treating a disease or a symptom caused by over-release of GABA, using a Bestl channel inhibitor as an active ingredient; and a use of a Bestl channel inhibitor as an active ingredient for preventing, improving, alleviating, and/or treating a disease or a symptom caused by over-release of GABA. Alternatively, an embodiment according to the present invention provides a composition for preventing, mitigating, and/or treating a disease or a symptom caused by deficit of GABA, which comprises a Bestl channel-activator as an active ingredient; a method of preventing, improving, alleviating, and/or treating a disease or a symptom caused by deficit of GABA, using a Bestl channel activator as an active ingredient; and a use of a Bestl channel activator as an active ingredient for preventing, improving, alleviating, and/or treating a disease or a symptom caused by deficit of GABA.

Said Bestrophin 1 , a type of chloride ion channels, is used as a representative example to show that chloride ion channels allow permeation of GABA. Bestrophin 1 genes can be derived from mammals, preferably from rodents or primates, and can be, for instance but not limited thereto, mouse Bestrophin 1 (mBestl) gene (NM Ol 1913, SEQ ID No. 1) or human Bestrophin 1 (hBestl) gene (NM_004183, SEQ ID No. 2).

Said Bestl channel inhibitor may comprise any substance having an inhibiting activity against the expression of Bestl channel or interfering with and/or blocking, directly or indirectly, the activity of Bestl. For instance, the Bestl channel inhibitor can be one or more selected from the group consisting of anion channel blockers and antisense RNAs or shRNAs for Bestl channel-coding nucleotide sequences, without being limited thereto.

Said anion channel blockers can be one or more selected from the group consisting of niflumic acid, flumenamic acid, NPPB (5-nitro-2(3-phenylpropylamino)- benzoic acid), and DIDS (4,4'-diisothiocyanatostilbene-2,2'-disulfonic acid), without being limited thereto. Said antisense RNA can be an antisense RNA for the nucleotide sequence of SEQ ID NO 1 or 2. In addition, said shRNA, indicated by cDNA sequence, can be one or more selected from the group consisting of SEQ ID NOs 3, 4, and 7, without being limited thereto.

5'-

GATCCCCTTGCCAACTTGTCAATGAATTCAAGAGATTCATTGACAAGTTGGC AATTTTTA-3' (SEQ ID NO: 3),

3'- GGGAACGGTTGAACAGTTACTTAAGTTCTCTAAGTAACTGTTCAACCGTTAA AAATTCGA-5' (SEQ ID NO: 4),

5'-

CGCTGCAGTTGCCAACTTGTCAATGAATTCAAGAGATTCATTGACAAGTTGG CAATTTTTGATATCTAGACA-3'(SEQ ID NO: 7).

When GABA release in the cerebellar glial cells is suppressed by inhibiting the activation of Bestl channel, neural GABAergic inhibition, such as tonic inhibition, can be mitigated with the effect of preventing, mitigating, and/or treating a disease or a symptom caused by over-release of GABA. A list of diseases or symptoms caused by over-release of GABA can include epilepsy, sleeping difficulties, memory difficulties, sensory difficulties, cognitive difficulties, motor difficulties, learning difficulties, alcohol addiction, such as low dose alcohol intoxication, and ataxia, without being limited thereto. Therefore, a composition for preventing, mitigating, and/or treating a disease or a symptom caused by deficit of GABA, which comprises a Bestl channel-activator as an active ingredient, can have an effect of prevention, mitigation and/or treatment for one or more diseases or symptoms selected from the group consisting of epilepsy, sleeping difficulties, memory difficulties, sensory difficulties, cognitive difficulties, motor difficulties, learning difficulties, alcohol addiction, such as low dose alcohol intoxication, and ataxia.

In another aspect of the present invention, an Bestl channel activator according to the present invention facilitates the release of GABA in the cerebellum and is able to increase neural GABAergic inhibition, such as tonic GABAergic inhibition, mitigating the excessive neural excitements caused by over-release of excitatory neurotransmitters. Such Bestl channel activator according to the present invention can be any material or substance having the effect of activating Bestl channel, directly or indirectly. For example, the Bestl channel activator can be an agonist of G-protein coupled receptor (GPCR), such as peptide TFLLR and Bradykinin, without being limited thereto. Such Bestl channel activation promotes the release of GABA. And the neural inhibition by the released GABA, e.g., tonic GABAergic inhibition, can have an effect of preventing, mitigation, and/or treating pathological symptoms caused by excessive neural excitement, for example, memory related diseases (Alzheimer, memory loss with aging, and etc.), seizures, excitotoxicity, ischemia, cerebral apoplexy, cerebral hemorrhage, epilepsy, brain injuries, and hypoxia. Accordingly, a composition for preventing, mitigating, and/or treating a disease or symptoms cased by GABA deficits, which comprises Bestrophin 1 charmer activity-inhibitor as an active ingredient, can have an effect of prevention, mitigation, and/or treatment on one or more kinds selected from the group consisting of memory-related diseases (Alzheimer, memory loss with aging, and etc.), seizures, excitotoxicity, ischemia, cerebral apoplexy, cerebral hemorrhage, epilepsy, brain injuries, hypoxia.

According to the present invention, the GABA release-regulating agent and the composition, the method, and/or the use for preventing, mitigating, and/or treating diseases and/or symptoms caused by over-release or deficit of GABA may be one to be administered to mammals, preferably to humans.

Another aspect of the present invention relates to a method for screening a novel cerebellar GABA release-regulating agent, in which the screening is performed using Bestl channel as a target in the cerebellum, more specifically, Bestl channel in the cerebellar glial cells.

The above screening method can comprise the steps of:

preparing a cerebellar sample;

allowing the contact of a candidate to the cerebellar sample; and

observing the activation of Bestl channel in the cerebellar sample.

In the screening method, the candidate is determined to be a GABA release- promoting agent when the Bestl channel is found to be activated, whereas the candidate is determined to be a GABA release-inhibiting agent when the Bestl channel is found to be inactivated.

Determination of whether the Bestl channel is activated can be performed using any method known in the technology field to which the present invention belongs. For example, the determination can be made in a manner in which other channels and receptors are made inactive except the Bestl channel in the cerebellar glial cells, and then inward current changes are measured. Increased inward currents after the treatment of a candidate suggest that the Bestl channel is activated, whereas decreased inward currents after the treatment of a candidate suggest that the Bestl channel is inactivated. Inactivation of other channels and receptors and determination of inward currents are of technology widely known in the technology field to which the present invention belongs, and those skilled in the art can perform easily. For instance, determination of inward currents can be carried out using sniffer patch technique (see 'Lee, C. J. et al. Astrocytic control of synaptic NMDA receptors. J Physiol 581, 1057-81 (2007)', which is incorporated hereto as a reference).

In the screening method according to the present invention, the cerebellar sample can be obtained from mammals, preferably from rodents or primates.

By identifying the mechanism of tonic GABA release via a Bestl channel in the cerebellum, it is expected to more effectively regulate the GABA release as well as GABA-related pathological symptoms.

BRIEF DESCRIPTION OF THE DRAWINGS

Figs. Ia- Id represent GABA containing cerebellar glial cells which express bestrophin 1 channel that can release GABA by direct permeation, wherein

Fig. Ia shows confocal images of immunohistochemistry with antibodies against GFP, Bestl, and GABA in GFAP-GFP transgenic mouse cerebellum;

Fig. Ib is schematic of two-cell sniffer patch; and

Fig. Ic shows the result of the extent of permeation.

Fig. Id shows the degree of GABA release activity under the stated conditions.

Fig. 2a is schematic of Cre-lox regulated pSicoR-shRNA lentivirus construct.

Fig. 2b shows the timeline of experiment with B6 or GFAP-GFP mice.

Fig. 2c shows Bl-shRNA lentivirus containing GFAP-GFP (green) staining, Bestl (magenta) and mCherry (red).

Fig 2d shows the result of glial cell patch recording using the ramp protocol (Vh= -7OmV).

Figs 2e and 2f each shows current-voltage traces in GFAP-GFP mouse with scrambled shRNA injection and Bestl-shRNA injection, respectively.

Fig. 2g illustrates the plotting of averaged current- voltage traces of NPPB sensitive current for each condition.

Fig. 2h represents the current amplitudes at -8OmV obtained from g to compare efflux of Cl^" and GABA

Figs. 3a-3i show that tonic GABA current is inhibited by anion channel blocker and decreased by gene silencing of bestrophin channel, wherein

Fig. 3a shows cerebellar slice of CLMl clomeleon mouse showing bright fluorescent granular cell bodies in granule cell layer and parallel fibers located in molecular layer, separated by translucent Purkinje cell layer (black arrows);

Fig. 3b shows ratiometric imaging of clomeleon revealing the time course of [ClIi change;

Fig. 3c demonstrates a graph that shows the correlation between [Cl^~]j change by NPPB and [Cl ]₁ change by SR;

Fig. 3d is colmeleon imaging from mouse injected with scrambled-shRNA lentivirus;

Fig. 3e is clomeleon imaging from mouse injected with Bl-shRNA;

Fig. 3f is summary of clomeleon imaging in each virus type and each layer;

Fig. 3g shows raw traces of tonic GABA current from granular cell in cerebellar slice (holding potential at -6OmV) of 8wks B6 mice;

Fig. 3h is summary figure of GABAzine-sensitive current from naive, scrambled, and Bl-shRNA injected mice; and

Fig. 3i is summary figure for NPPB sensitive current.

Figs. 4a-4f show that glia specific rescue of Bestl channel restores tonic GABA current, wherein

Fig. 4a is the timeline of experiment for hGFAP-CreERT mice;

Fig. 4b shows a typical glial-specific rescue mechanism;

Fig. 4c shows the whole-cell patch clamp recording from granular cells;

Fig. 4d shows the GABAzine-sensitive currents in case of naive, shRNA and shRNA+Tamoxifen;

Fig. 4e shows the NPPB-sensitive currents in case of naive, shRNA and shRNA+Tamoxifen; and

Fig. 4f is a proposed model of tonic GABA release in cerebellum.

Fig. 5 shows that glial cells strongly express mBestl and GABA in adult mice cerebellum, wherein

Fig. 5a illustrates concocal images of immunocytochemistry for GABA, Bestl and GFAP-GFP in mice cerebellum;

Fig. 5b shows higher magnification of GABA and GFAP-GFP staining in cerebellum; and

Fig. 5c shows higher magnification(X60) of Bestl and GFAP staining.

Figs. 6a-6g shows that tonic GABA release in the cerebellar granular cell layer is Ca2+ dependent, non- vesicular, and inhibited by anion channel blockers, wherein

Fig. 6a shows bright field images of granular and molecular cell layers (upper panel;X40, lower panel;X600) which are used to obtain the results in 6a-6g;

Figs. 6b-6e each shows tonic GABA current recordings with the application of lOOμM Niflumic acid (NFA), incubation with 150μM of BAPTA-AM, incubation with 0.5 μM concanamycin A, and application of 30 μM NPPB, respectively;

Fig. 6f shows summarized results of GABAzine sensitive current with no treatment, concanamycin A treatment and BAPTA-AM treatment; and

Fig. 6g is the block percentage of tonic current by Ca²⁺ sensitive Cl^" channel blockers. Figs. 7a-7d indicate that NPPB does not directly affect GABA receptors, wherein

Figs. 7a and 7b show the results when GABAc expressing HEK293 cells were patch clamped and challenged with lOOμM GABA in the absence and presence of 1 OOμM NFA and 1 OOμM NPPB; and

Figs. 7c and 7d show the application of NPPB which does not affect GABA receptors in cerebellar granule cells, and the magnitudes of GABA induced current with the NPPB application (2 and 5 min) compared to the baseline;

Fig. 8a shows fluorescence images of mCherry signals in shRNA and scrambled RNA injected mice (Scale bars; 200μm (Xl 0) and 50 μm (X40)).

Fig. 8b shows the Bestl knock-down experiment due to gene silencing in hGFAP-Cre mouse. In the left panel, Bl-shRNA lentivirus (red) is exhibited in the hGFAP-Cre mouse injected with tamoxifen. The middle panel shows Bestl stained image in the similar area. The right panel shows the image of GFAP (a marker for Glial cell) stained image in addition to the combined image of left and middle panel. Bestl (red) indicates that lentivirus infected are is similarly stained as the non-infected area. (Scale bar; 100 μM).

Examples

The present invention is further explained in more detail with reference to the following examples. These examples, however, should not be interpreted as limiting the scope of the present invention in any manner.

Example 1: Gene cloning and shRNA virus vector construction

1.1: Cloning of Bestl

For the cloning of full-length mouse bestrophin 1 (mBestl) cDNA, total RNA obtained from cultured astrocytes of P0-P3 postnatal mice (C57BL/6, cerebral cortex; SPF room, Center for Neural Science, KIST, Seoul, Korea) or testis of adult male mice (C57BL/6) was purified, and cDNA was synthesized using Super Script III reverse transcriptase (Invitrogen). Using 21 base primer pair (mBestl-F: 5'- aggacgatgatgattttgag-3¹ (SEQ ID NO: 5), mBestl -R: 5'-ctttctggtttttctggttg-3' (SEQ ID NO: 6)) spanning the open-reading frame based on NCBI database cDNA [GenBank accession numbers NM Ol 1913 XM 129203], PCR was performed to acquire full- length ORF of mBestl. Resulting PCR products were cloned into a pGEM-T easy vector (Promega) and the sequence was verified. 1.2: Plasmid construction of Bestl and expression

In order to express mBestl in mammalian cell, mBestl full-length fragment from pGEM-T easy plasmid (6.65 kb, Promega) was subcloned into pcDNA 3.1 (Invitrogen) by using HindIII(NEB) and Notl (NEB) sites or pIRES2-dsRED (Invitrogen) by using Xbal (NEB) and Xmal (NEB) sites. The pcDN A3.1 -mBestl plasmid was transfected into HEK293T cells (ATCC) with 1/10 quantity of pEGFP-Nl plasmid (Invitrogen) using effectene transfection reagent (Qiagen) to detect mBestl - expressing cells. Cells with green fluorescence were selected when both pcDNA3.1- mBestland pEGFP-Nl were transfected, whereas cells with red fluorescence were selected when pIRES2-dsRED vector plasmid was transfected. One day after transfection, cells were replated onto glass coverslips for electrophysiological recording. The transfected cells were used for patch clamp experiments within 24-36 hrs.

1.3: Bestl shRNA and Ientivirus production

For plasmid-based shRNA expression, the following complementary oligonucleotides were annealed and inserted into the Hindlll/Bglll sites of pSUPER- GFP vector (Oligo Engine):

5'-

GATCCCCTTGCCAACTTGTCAATGAATTCAAGAGATTCATTGACAAGTTGGCA ATTTTTA-3' (SEQ ID NO: 3);

3'- GGGAACGGTTGAACAGTTACTTAAGTTCTCTAAGTAACTGTTCAACCGTTAA AAATTCGA-5' (SEQ ID NO: 4),

(The nucleotide sequence corresponding to of mBestl (563-582) is included. The remaining sequences are included for the purpose of hairpin shape and cloning].

For lenti virus-based shRNA expression, lenti viral vector containing mBestl gene was constructed by inserting synthetic double-strand oligonucleotides 5'- CGCTGCAGTTGCCAACTTGTCAATGAATTCAAGAGATTCATTGACAAGTTGG CAATTTTTGATATCTAGACA-S' (SEQ ID NO: 7) into pstl-Xbal restriction enzyme sites of shLenti2.4 CMV lentiviral vector (Macrogen) and verified by sequencing. Scrambled oligonucleotides-containing-shLenti construct (used as a control for target shRNA that degrades mRNA by recognizing particular sequences and composed of sequences that do not degrade cellar rnRNA, Macrogen) was used as control. The production of lentivirus was performed by Macrogen Inc. (see Dull, T. et al., A third- generation lentivirus vector with a conditional packaging system. J Virol 72, 8463-71 (1998), which is incorporated hereto as a reference).

In the following Example 2, Bestl targeting shRNA lentiviral inserted with SEQ ID NO: 7 was used.

Example 2: Immunochemistry: Verification of co-expression of GABA and Bestl in glial cells

In order to verify the expression of Bestl in the cerebellum, immunochemistry was performed by using antibodies raised against Bestl and GABA in a GFAP-GFP transgenic mouse (FIG 5c).

Adult GFAP-GFP mice (SPF room, Center for Neural Science, KIST, Seoul, Korea) or lentivirus (Bestl targeting shRNA lentivirus in Example 4) injected GFAP- GFP mice were fixed with 4% paraformaldehyde. 30μm sagittal cryostat sections of cerebellum were rinsed in PBS 3 times and incubated for lhr in blocking solution (0.3% Triton-X, 2% normal serum in 0.1 M PBS, sigma). After incubating overnight in blocking solution containing the mixture of rabbit anti-mouse bestrophin antibody (1:100, (Soria et al, 2006)) and chicken anti-GFP antibody (1:1,000, abeam) and guinea pig anti GABA antibody (1 :1,000, Chemicon) at 4^°C on shaker, the cryostat sections were washed 3 times in PBS, and then in Alexa 488, 555 and 647 conjugated with corresponding secondary antibodies. The resulting products were washed 3 times in PBS and mounted with fluorescent mounting medium (Dako, S3023). Confocal series of fluorescence images were obtained using FVlOOO confocal microscope (Olympus). The images were processed using Olympus FLUOVIEW software ver.1.7.

FIGIa represents confocal images of immunohistochemistry with antibodies against GFP, Bestl , and GABA in GFAP-GFP transgenic mouse cerebellum.

FIG. la(upper left) shows a confocal image of GFAP-GFP staining (green) that labels Bergmann glial cells (arrow) and lamella astroctyes (pale blue arrowheads). Purkinje cell (star) is devoid of GFAP-GFP staining. FIG la(upper right) is a confocal image depicting both GFAP-GFP staining and Bestl staining (red). Bestl, expressed in Purkinje cells (star) and other neurons (arrowheads), is also highly expressed in glial cells (pale blue arrowheads) in granular layer and Bergmann glial cells (arrow) in molecular layer. FIG. 1 (lower left) shows merged GFAP-GFP and GABA staining (magenta). In addition to GABAergic neurons, GABA is strongly coexpressed with GFAP-GFP in Bergmann glial cells (arrow) and lamella astrocytes (pale blue arrowheads). FIG. Ia (lower right) represents merged images of GFAP-GFP, mBestl and GABA. According to Fig. Ia, Bestl immunoreactivity intensity is observed in Bergmann glial cells (arrow), lamella astroctyes (pale blue arrowheads), and GABAergic neuron, but not in granular cells.

FIG. 5a-c is a picture showing a strong expression of mBestl and GABA in glial cells of adult mouse cerebellum.

Fig. 5a is an immunohistochemistry confocal images for GABA, Bestl and GFAP-GFP in mouse cerebellum. Immunohistochemical studies show that GABA (First Panel) and Bestl (Second Panel) are intensely expressed in molecular layer than granular cell layer. The third panel indicates that Bergmann glial process occupy the most of region in molecular layer. The last panel shows the merged confocal image of GABA, Bestl and GFAP-GFP.

FIG. 5b shows a higher magnification image of GABA and GFAP-GFP staining in cerebellum. GABA is heavily stained with Perkinje cells and GABAergic interneurons in molecular layer. However, surprisingly, Bergmann glical cells (star) also express strong GABA immunoreactivity in their soma and processes (small black arrowheads). Glial cells in granular layer (arrow) are stained with GABA with lighter intensity, and their processes are also stained with GABA (small white arrows).

FIG. 5c shows a higher magnifϊcation(X60) image of Bestl and GFAP staining. mBestl is highly expressed in glial cells (arrows) of granular layer and Bergmann glial cells (star) of molecular layer. Big arrowhead indicates Perkinje cells that express mBestl but devoid of GFAP staining. Small arrowheads indicate astrocytic process in granular layer. Astrocytic processes also express mBestl. The right most panel shows granular cells heavily stained with DAPI but no mBestl and GFAP immunoreativity were observed. GABA GFAP-GFP GABA+GF AP-GFP Merged with DAPIed

As illustrated in FIG 5a-c, Bergmann glial cell processes, that are located along

Purkiηje cell body and dendritic trees in molecular layer where parallel fibers and climbing fibers are close to and interact with each other, co-expressed Bestl and GABA. These results raise an intriguing possibility that the Bestl channel could serve as a molecular target for glial release of GABA in cerebellum.

Example 3: Two-cell sniffer patch - Verification of Best 1 Channel mediated GABA release

For Bestl channel to mediate release of GABA, it has to permeate GABA upon channel opening. To test for GABA permeability of Bestl channel, a two-cell sniffer patch technique was developed to directly measure GABA release via Bestl channel in sensor HEK293T cells that express GABAc₅ and via other channels that are expressed in source HEK293T cells (Fig. Ib).

pIRES-Bestl-dsRED plasmid (obtained by cloning Bestl using pIRES- dsRED

(Invitrogen)) and GABAc with GFP (obtained by cloning GABAc using pcDNA3.1 (Invitrogen)) were transfected into HEK 293T cells (ATCC) using Effectene transfection reagent (Qiagen). 18~24hrs after transfection, cells were replated together onto glass coverslips for electrophysiological recording and those cells were used for patch clamp experiments within 24~36hrs. For recording, one of adjacent two cells consisting of a dsRED stained cell (Red) transfected with pIRES-Bestl-dsRED and a GFP stained cell (Green) transfected with GABAc with were selectively patched.

The patch pipette internal solution containing 3 or 14OmM GABA which serves as the source of GABA release, and free Ca ⁺ (~4.5 M) at a concentration within physiological range (micromole) which activates Bestl channel, was used. Fig. Ib is a schematic diagram of two-cell sniffer patch illustrating HEK cells which express Bestl (coexpressing dsRed) or GABAc (co-expressing GFP). The picture at the bottom of Fig. Ib depicts intracellular pipetting of 3 or 145mM GABA and 0 or 4.5μM Ca²⁺ into the source. The bright field and fluorescence images of source and sensor cells are also shown at the bottom of Fig. Ib.

For the source of GABA release, the pipette solution containing 3 mM GABA(Tocris) 146 mM CsCl, 5 mM (Ca²⁺)-EGTA[ethylene glycol bis(2-aminoethyl- ether)-N,N,N',N'-tetraacetic acid]-NMDG(N-methyl-D-glucamine), 2 mM MgCl₂, 10 mM HEPES, 10 mM sucrose, 4 mM Mg-ATP and 0.3 mM Na₂-GTP (pH; 7.3) was used. For the sensor, the pipette solution containing 110 mM D-gluconate, 110 mM CsOH, 30 mM CsCl, 2 mM MgCl₂, 4 mM NaCl, 5 mM EGTA(ethylene glycol bis(2-aminoethyl- ether)-N,N,N',N'-tetraacetic acid), 4 mM Mg-ATP, and 0.3 mM Na₂-GTP (pH;7.3) was used. For 3mM GABA Zero Ca²⁺ experiment, 5mM (Ca²⁺)-EGTA-NMDG was replaced by 5mM EGTA-NMDG. For 14OmM GABA experiment, 3mM GABA and 146mM CsCl were replaced by 14OmM GABA. pH was adjusted with CsCl and osmolality was adjusted to 290 mOsmol.

The internal solution containing 150 mM NaCl, 10 mM HEPES, 3 mM KCl^ 2 mM CaCl₂, 2 mM MgCl₂, and 5.5 mM Glucose was used. If the source channel can permeate GABA, GABAc receptor on neighboring cell bind to released GABA and Cl^" inward current would be elicited. Full activation of GABA current was obtained by bath application of 100 μM GABA and normalized for the purpose of comparison.

GABA release was induced by a break-though of membrane patch that goes into a whole-cell configuration in the source cell and such the released GABA was monitored and detected by a neighboring sensor cell. Full activation of GABA current was obtained by bath application of 100 M GABA and the percentage of full activation was calculated (Fig. Ic and Id).

Fig. Ic shows the result of permeation experiment. Permeated GABA is detected in sensor as an inward current (bottom traces). The time period for a membrane break-through to go into whole-cell mode is indicated as a black arrowhead on the source trace (top traces). NPPB was used at lOOμM. Bestl* (Bl*) is a pore mutant Bestl-W93C (Qu et al, 2006). The GABA permeability of said mutant was determined by using cells obtained by transfecting said mutant into HEK293t cell (ATCC). ANO 1 is the recently characterized TMEMl 6 A Ca2+ activated chloride channel (Yang et al., 2008, Caputo et al., 2008; Dr. Park Lab, Gyeongsang National University, Korea). The GABA permeability of said ANOl was determined by using cells obtained by transfecting said mutant into HEK293t cell (ACTT).

Fig. Id is a summary of the extent of GABA release in various conditions.

GABA release detected as a inward current in sensor cell is normalized to a full activation with 100 μM GABA and then the percentage of full activation was calculated. NPPB and NFA were used at 100μM. SKl is small conductance Ca2+ activated K+ channel (Dr. Adelman Lab). The GABA permeability of said SKl was determined by using cells obtained by transfecting said mutant into HEK293t cell (ACTT). Averages are expressed as mean+ SEM (standard error of the mean). Student's t-test was used throughout the experiment (unpaired, 2-tailed).

As illustrated in Fig. Ib, Ic and Id, Bestl channel uniquely displayed a significant permeability for GABA, whereas recently characterized Anol(or TMEM- 16A, Yang et al, 2008, Caputo et al, 2008) or Ca²⁺ activated potassium channel, SKl did not show any permeability for GABA. The GABA release via Bestl was completely abolished by anion channel blockers, NFA and NPPB and was dependent on intracellular Ca²⁺ and GABA concentration (Fig. lc,d). In addition, as shown in Fig. Id, one of the known pore mutants of Bestl, Bestl-W93C (Qu et al, 2006) did not show any GABA permeability, supporting the idea of GABA permeation occurs through the pore of Bestl channel.

In addition, Fig. 7 shows that the inhibition of GABA release by NFA and

NPPB is not obtained by directly affecting the GABAc receptors. In Fig. 7a and 7b, GABAc expressing HEK293 cells were patch clamped and challenged with lOOμM GABA in the absence or presence of lOOμM NFA and lOOμM NPPB. NFA and NPPB application do not have significant impact on GABA release in the GABAc expressing HEK293 cells.

These results raise a possibility that Bestl channels can mediate GABA release through direct permeation in native cells.

Example 4: Transgenic Mouse Experiment - Verification of Bestl channel mediated GABA Release

4.1: Trangenic Mouse Construction

To test whether native Bergmann glial cells express functional Bestl channel that can permeate GABA and further manipulate Bestl channel at the molecular level, a lentivirus carrying a mCherry-tagged small hairpin-forming interference RNA (shRNA), which is under the regulation of Cre-loxP recombination, inducing cell-type specific gene silencing when used in combination with Cre-expressing transgenic mice was constructed (Fig. 2a, Ventura et al., 2004). Fig 2a is Cre-lox regulated pSicoR-shRNA lentivirus construct (Bestl-shRNA cloned with pSicoR vector purchased from ADD Gene). The lentivirus carrying mCherry-tagged shRNA was constructed by attaching mCherry to Cre-lox regulated pSicoR-shRNA lentivirus construct. The two loxP sites are located in the area that includes shRNA under U6 promoter and mCherry under CMV promoter. When Cre recombinase is expressed, it excise out these cassettes, making shRNA inactive.

The Best 1 -targeting shRNA lentivirus (prepared in Example 1.3) was injected streotactically into cerebellar cortex of 6-7 week old GFAP-GFP mouse (SPF room, Center for Neural Science, KIST, Seoul, Korea) (Fig 2b). Fig. 2b shows the time line of experiment with B6 (SPF room, Center for Neural Science, KIST, Seoul, Korea) or GFAP-GFP mice. Mice were injected at 6-7 week age. After 7 days from lentivirus injection into cerebellum, immunohistochemistry or whole-cell recordings were performed. Example 4.2: Immunohistochemical analysis using transgenic mouse

The cells transfected with the lentivirus were extensively distributed in the molecular layer as well as granular cell layer (Fig.2c, right and Fig. 8a). Fig. 2c shows Bl -shRNA lentivirus carrying GFAP-GFP (green) staining, Bestl (magenta) and mCherry (red). Bestl immunoreactivity is significantly reduced in virus injected area compared to uninfected area. Intensities for GFAP-GFP in both area are relatively similar. Right most, the knockdown efficiency is expressed as Bestl intensity normalized by GFAP-GFP intensity after thresholding with GFAP-GFP. The pixel intensity of Bestl immunoreactivity in infected region decreased dramatically compared to the uninfected regions (Fig 2c, far right), confirming both the high efficiency of Bestl-shRNA and specificity of Bestl antibody.

4.3: Recording of whole-cell patch clamp using transgenic mouse

4.3.1: Construction of cerebellum slice

Brain slices were prepared as described in Rossi et al., 2003. For slice recording, either approximately P28 days old or more than 8 weeks old mice were used. Animals were deeply anesthetized with halothane. After decapitation, the brain was quickly excised from the skull and submerged in ice-cold cutting solution (in mM): 250 Sucrose, 26 NaHCO₃, 10 D(+)-Glucose, 4 MgCl₂, 3 myo-inositol, 2.5 KCl, 2 Sodium pyruvate, 1.25 NaH₂PO₄, 0.5 Ascorbic acid 0.1 CaCl₂, 1 Kynurenic acid, pH 7.4. All solutions were gas-treated with 95% O₂-5% CO₂. After trimming both sides of vermis, several parasagittal slices with 250 μM thicknesses containing cerebellar lobes were cut using a microtome (Leica VT 1000) and transferred to extracellular ACSF solution (in mM); 126 NaCl, 24 NaHCO₃, 1 NaH₂PO₄, 2.5 KCl, 2.5 CaCl₂, 2 MgCl₂, 10 D(+)-Glucose, pH 7.4. Slices were incubated for one hour at least at room temperature.

4.3.2: Recording of whole-cell patch

For recording, slices were transferred to an electrophysiological recording chamber (RC-26G, Warner Instruments) which is continuously superfused with ASCF (artificial cerebrospinal fluid, sigma) solution (flow rate; 2ml/min) and controlled by flower controller (Synaptosoft) and a vacuum pump (Charles Austen, model Capex 8C). Slice chamber was mounted on the stage of an upright microscope (Olympus, Japan) and viewed with an X60 water immersion objective with differential interference contrast and infrared optics. Cellular morphologies were visually identified by Imaging Workbench 6.0 (INDEC Systems, Inc), camera controller (Hamamatsu, C4742-95), and light microscope controller (Olympus, TH4-200). Fluorescence images were viewed with mercury lamp (Olympus, U-RFL-T). Whole cell voltage-clamp recording was made from granular cell somata or Bergmann glial cells mostly located in 2-5 cerebellar lobules

For Bergmann glial cell recording patch pipettes (8-10 MΩ) were constructed from thick- walled borosilicate glass capillaries (SC 150F- 10, Warner instrument Corp). For 0 GABA comparison experiment, pipette was filled with an internal solution containing (in mM); 146 CsCl, 5 (Ca²⁺)-EGTA-NMDG, 2 MgCl₂, 8 HEPES and 10 Sucrose, 4 Mg-ATP, and 0.3 Na₂-GTP (pH; 7.3).

For 140 mM GABA experiment, internal solution containing 140 mM GABA, 5 (Ca²⁺)-EGTA-NMDG, 2 mM MgCl₂, 10 mM HEPES ,10 mM Sucrose, 4 mM Mg-ATP, and 0.3 mM Na₂-GTP (pH; 7.3 with CsOH). Osmolality was adjusted to 297 and 290 mOsmol was used. Bergmann glial cells were visually identified by GFP fluoscence image. Holding potential for voltage clamping of Bergmann glial cell was -70 mV.

Pipette resistance for granule cells was typically 10-12 MΩ and pipette was filled with an internal solution containing (in mM) 135 CsCl, 4 NaCl, 0.5CaCl₂, 10 HEPES, 5 EGTA, 2 Mg-ATP, 0.5 Na₂-GTP, 10 QX-314, pH adjusted to 7.2 with CsOH (278-285 mOsmol) was used(Rossi, et al., 2003). With this internal solution, E_ci=0 mV with voltage clamp and holding potential of -60 mV, inward current was elicited.

Electrode junction potentials for Bergmann glial cells recording were corrected but junction potential for granular cell recording was not corrected. Junction potentials were +3.5 mV and -9.7 mV in 0 GABA and 140GABA experiments, respectively.

For puffing experiment, glass electrode (5-6 MΩ) filled with 10OmM was positioned near to the patched granular cell and puffed briefly for 500ms by Picospritzer III (Parker instrumentation) connected with MiniDigi (Molecular Device).

For Purkinje cell recording, patch pipette (2-3 MΩ) was filled with an internal solution containing (in mM) 140 K-gluconate, 10 KCl, 1 MgCl₂, 10 HEPES, 0.02 EGTA, 4 Mg-ATP, 0.4 Na₂-GTP pH adjusted to 7.35 with KOH (Osmol: 278-285) was used. The data recorded from the cell with access resistance over 30MΩ were discarded.

The signals were digitized and sampled at 50 μs intervals with Digidata 1440A

(Molecular Devices) and Multiclamp 700B amplifier (Molecular Devices) using pCLAMP 10.2 sofware (Molecular Devices). Off-line analysis was carried out using Clampfit 10.2 (Molecular Devices), Minianalysis (Synaptosoft, USA), SigmaPlot 10.0 (SPSS) and Excel 2003 (Microsoft).

4.3.3: Drug Application

All the drugs and chemicals used in this study were purchased from Sigma- Aldrich if not mentioned otherwise; Lidocaine N-ethyl bromide (QX-314, Sigma), SR95531 hydrobromide (GABAzine, Tocris), concanamycinA (Tocris), BAPTA-AM (l,2-Bis(2-aminophenoxy)ethane-N,N,N',N'-tetraacetic acid tetrakis(acetoxymethyl ester), Tocris), Fluronic® F-127 (invitrogen), Niflumic acid (Sigma), NPPB (5-Nitro-2- (3-phenylpropylamino)benzoic acid, Tocris), DIDS (4,4'-Diisothiocyanatostilbene-2,2'- disulfonic acid disodium salt hydrate, Sigma). 4.4.4: Data analysis and Statistical analysis

Numerial data was presented as means±S.E.M. The significance of data for comparison was assessed by Student's two-tailed unpaired t test and significance level was displayed as * (p < 0.05), ** (p < 0.01), ***(p < 0.001). Data were filtered at 2kHz, Goldman-Hodgkin-Katz equation was calculated as below and Pχ/Pci- was calculated. E™ = RT/F-ln{PCr[Cr]j + P_x[X]i}/{ PCr[Cr]₀+ Px[X]₀) 4.4.5: Result

The whole-cell patch clamp recordings from Bergmann glial cells in cerebellar slice of naive and lenti virus injected adult mice was performed to search for functional expression of Bestl. The internal solution contained either mostly Cl^" or GABA as an anion, in addition to 4.5 M free Ca²⁺ to activate the endogenous Bestl channel. Anion current was isolated by subtracting the ramp current trace (from 10OmV to -10OmV in 2s) during 50 M NPPB application from that of baseline condition before NPPB application (Fig. 2d). The current- voltage relation of NPPB-sensitive anion current was generated for each cell by transforming the time in ramp trace to voltage (Fig. 2d, 2e, 2f). Fig. 2d shows Glial cell patch recording with ramp protocol (Vh= -7OmV). Anion current is decreased by anion channel blocker, NPPB (50μM) in naϊve GFAPGFP mice. Internal solution of GABA and Cl- are composed as indicated above. Current-voltage traces are generated from each ramp trace. Subtracted current represents NPPB sensitive current. Figs. 2e and 2f show current-voltage traces in GFAP-GFP mouse with scrambled shRNA injection (2e) and Bestl -shRNA injection (2f), respectively.

In Fig. 2g, current-voltage traces of NPPB sensitive current for each condition are averaged and plotted. GABA permeability is calculated using the reversal potential and Goldman-Hodgkin-Huxley equation. The NPPB-sensitive anion current under mostly Cl^" internal solution in naϊve Bergmann glial cells showed averaged reversal potential of -6.9mV (Fig. 2g, black trace, corrected for -3.5mV junction potential), which was not very different from the calculated value from reversal potential of +ImV using the Goldman-Hodgkin-Huxley equation and assuming a contribution of bicarbonate (P_HCO3/P_CI=0.44; QU and Hartzell, 2008). The GABA permeability ratio of NPPB-sensitive anioin current under GABA internal solution was similarly determined to be Po_ABA/Pci⁼0.19(Fig. 2g, green trace). The GABA permeability ratio obtained from naϊve cells was not significantly different from that of scrambled-shRNA expressing Bergmann glial cells (Fig. 2g, blue trace). When the NPPB-sensitive current was isolated from Bestl -shRNA expressing Bergmann glial cells, the conductance as indicated by the slope of the current-voltage trace decreased significantly without shifting the reversal potential (Fig. 2g, red trace)

The outward current measured at 100m V, which represents the influx of Cl^', did not show any significant difference between mostly Cl^" and GABA internal solution (212.23±49.52pA (n=9), 112.52±20.93pA (n=8), p=0.1), indicating that influx of Cl^" was not significantly affected by substitution of Cl^" to GABA internally. Both the inward current measured at -8OmV, which represents the efflux of GABA and outward current measured at 100m V, which represents the influx of Cl^" under mostly GABA internal solution showed significant differences between the scrambled and Bestl - shRNA cells (inward current: -44.62±5.01pA (n=10), -12.99±5.92pA (n=9), p<0.001,

Fig. 2h; outward current: (110.26±23.25pA (n=10), 38.31±12.02pA (n=9), p=0.02).

These results indicate that NPPB-sensitive anion current observed in Bergmann glial cells is mostly mediated by Bestl channel, which displays a significant permeability to GABA at around resting membrane potential.

Example 5: Verification of GABA Release Inhibition by Bestl Silence

5.1: Virus injection

B6 wildtype, GFAP-GFP and hGFAP-CreERT2 trangenic mice (SPF room, Center for Neural Science, KIST, Seoul, Korea) were anaesthetized by intraperitoneal injection of 2% avertin (20 μl/g, sigma) and placed in a stereotaxic frame (David Kopf instrument). pSicoR-blshRNA-mCherry or scrambled virus (Macrogen) was stereo- injected into cerebellar cortex at a rate of 0.2 μl/min (total 2 μl) using syringe pump (Harvard apparatus) and 25 μl syringe (Hamilton company). The coordination of injection site was 1.7mm from the lambda and the depth was 1.5-1.7mm from the skull.

5.2: Clomeleon imaging

To test whether Bestl channel is responsible for tonic GABA release in cerebellum, an optogenetic approach was selected to assess GABA_A receptor mediated [Cl^'Ji movement in granule cells using the Thyl ::CLMl line (Department of Neurobiology, Duke University Medical Center, Durham, North Carolina, USA) of Clomeleon transgenic mice which show exclusive expression in granule cells in the cerebellum (Fig. 3a; Berglund and Augustine 2008). Fig. 3 a is a CFP-YFP FRET image representing cerebellar slice of CLMl clomeleon mouse showing bright fluorescent granular cell bodies in granule cell layer and parallel fibers located in molecular layer, separated by translucent Purkiηje cell layer (black arrows). Green and red squares indicate two regions of interest in molecular layer (green) and granule cell layer (red).

Clomeleon, based on fluorescence resonance energy transfer (FRET), is a genetically-encoded fluorescent indicator for Cl^" in which chloride-sensitive yellow fluroscent protein fused with chloride-insensitive cyan fluorescent protein via a flexible peptide linker (Kuner and Augustine, 2000). This Clomeleon mouse was successfully used to measure the tonic GABA release in cerebellar granule cells with added spatial information (Berglund et al).

Approximately 7-10 days after injecting pSicoR-blshRNA-mCherry virus (Macrogrn) into the Clomelon transgenic mice, the cerebellar slices were prepared under the conventional methods. In brief, the brains were removed from the decapitated mice after anesthetizing with isoflurane and placed in a cold artificial cerebrospinal fluid (ACSF), containing (in mM): 125 NaCl, 2.5 KCl, 1.25 NaH₂PO₄, 26 NaHCO₃, 20 d(+)-glucose, 2 CaCl₂ and 1.3 MgCl₂ (pH 7.4 after bubbling with 95 % O₂/ 5 % CO₂, v/v). A vibratome (LEICA) was used to obtain 200 μm thick saggital section. The slices were then incubated at 36 °C for 30 min prior to use.

For imaging, two ROIs (Regions of Interest) covering the granule cell layer and the molecular layer were drawn. Excitation (440 ± 10 run) and emission (485 ± 15 nm for CFP and 530 ± 15 nm for YFP) filters (Cameleons 2 filter set 71007, Chroma Technologies, Rockingham, VT) were used. Fluorescence excitation was produced by two consecutive 200 to 500 ms long light pulses at 0.5 Hz and fluorescence emission was alternately collected at each wavelength with a back-illuminated, cooled CCD camera with the on-chip multiplication gain control (Cascade 512B, Photometries). Image acquisition was controlled by RatioTool software (ISee Imaging Systems, Raleigh, NC) and a PowerMac G4 (Apple Computer).

As expected, bath application of 10 μM GABAzine (SR95531) markedly decreased [Cl^"]; in granular cell bodies (Fig. 3b, red trace) and parallel fibers in molecular layer (Fig. 3b, green trace) as the block of extrasynaptic GABA_A receptor decreased inward movement of Cl^". Interestingly, the reduction was equally prominent in parallel fibers in molecular layer as in granular cell bodies (Fig. 3f). Application of

10 M NPPB also decreased [Cl^"]i in both layers (Fig. 3b), indicating a decrease of tonic GABA. Fig. 3b shows ratiometric imaging of clomeleon illustrating the time course of [Cl-Ji change. lOμM SR (SR95531, or GABAzine) and anion channel blocker, NPPB ( 1 OμM) decrease [Cl]₄.

The degree of [Cl^"]i change by NPPB was closely correlated with change of [Cl^"

1₁ by GABAzine (Fig. 3c, r=0.96). As shown in Fig. 3c, [C1-]I change by NPPB is highly correlated with [Cl-Ji change by SR.

To test whether this GABA_A receptor activation-induced [Cl^"] j change is Bestl channel dependent, [Cl^"Ji concentration in granule cells of the Clomeleon mice with silenced Bestl gene was measured. [Cl^"] j concentration change by GABAzine was significantly decreased (Fig. 3e,f, p< 0.005) in Bestl -shRNA injected cerebellar slices

(molecular layer: 10.63±1.45mM, granule cell layer: 11.44±1.5mM ,(n=8)) compared to scrambled shRNA injected cerebellar slices (4.29 ±0.91mM, 3.84 ±0.82mM,(n=8)). Fig. 3f summarizes the [Cl^"] i changes due to scrambled shRNA injected cerebellar slices (Fig.

3d) and SR of Bestl -shRNA injected cerebellar slices (Fig. 3e) (p< 0.005). These results indicate that a gene silencing of Bestl channel reduces tonic GABA release detected in soma as well as in parallel fibers of granule cells. 5.3: The Whole-Cell Patch Clamp Recordings

The results from Clomeleon mice were verified by the whole-cell patch clamp recordings in granule cells from adult mice injected with Bestl -shRNA lentivirus. The GABAzine-sensitive current was significantly decreased in Bestl -shRNA lentivirus injected mice (Fig. 3g lower panel, 8.28±0.57pA, n=14) compared to those from naϊve (35.68±4.05pA (n=8), p « 0.001) or scrambled (26.62±2.85pA (n=13), p «0.001), whereas GABAzine-sensitive current between naϊve and scrambled did not show any significant difference (p>0.09). The upper traces of Fig. 3g represents raw traces of tonic GABA current from granular cell in cerebellar slice (holding potential at -6OmV) of 8wks B6 mice. Tonic GABA current is reduced by bath application of 50μM NPPB. Blue arrow indicates GABAzine (SR) sensitive tonic GABA current and orange arrow indicates NPPB-sensitive tonic GABA current. Middle trace is related to a mouse injected with scrambled shRNA lentivirus and bottom trace is related a mouse injected with Bl-shRNA lentivirus.

On the other hand, GAB Azine-sensitive current did not show much difference between naϊve and scrambled mice (p>0.09, Fig. 3h). Fig. 3h is a summary figure of GABAzine-sensitive current from naϊve, scrambled, and Bl-shRNA injected mice. The gene silencing of Bestl channel virtually eliminated the NPPB-sensitive component of tonic GABA current in Bestl -shRNA injected mice (-1.23±3.08pA, n=4), which showed a significant difference when compared to those in naϊve (18.95±2.47pA (n=8), p<0.002) or scrambled mice (12.98±2.57pA (n=4), p<0.02, n=4). Fig. 3i is a summary figure for NPPB sensitive current.

The inhibition of tonic GABA current caused by NPPB was not due to a direct action of this compound on GABA_A receptor expressed in granule cells because NPPB did not have any effect on GABA-induced whole cell current (Fig. 7c and 7d). Figs. 7c and 7d shows that the application of NPPB in the wild B6 mouse did not affect GABA receptors in cerebellar granule cells. The magnitudes of GABA induced current with the NPPB application (2 and 5 min) are also shown.

Other anion channel blockers, NFA and DIDS also blocked the GABAzine- sensitive current significantly (Fig. 6b,g). Fig. 5B shows the tonic GABA current recordings from granule cells when lOOμM Niflumic acid was applied. Fig. 5g indicates the block percentage of tonic current by Ca²⁺ sensitive Cl^" channel blockers. (The ages of mice used: 29.5±0.79, 27±0, 27.5±0.87, 28±0.71, and 74 days (DIDS)).

These results strongly support the idea that NPPB-sensitive tonic GABA release is mediated by Bestl channel in cerebellum and that the tonic GABA current is suppressed by the anion channel blockers as wells as Bestrophin channel gene silencing. Example 6: Tonic GABA Current Recovery by Glia-specific Bestl Channel

Rescue

In order to prevent the glia-specific Bestl from gene silencing, a test was done using hGFAP-CreERT2 mouse injected with tamoxifen and Bestl-shRNA lentivirus to investigate whether tonic GABA release was due to glial Bestl. (Fig. 4a and 4b).

The glia-specific CreERT activation was initiated by intraperitoneal injection of tamoxifen for 5 days prior to lentivirus injection (Fig. 4a,b). Fig. 4a shows the experiment timeline for hGFAP-CreERT mice. Tamoxifen or sunflower oil were injected intraperitoncally once a day for 5 days before lenti virus injection. Under this strategy, the expressed CreERT was transferred to the nucleus to excise out the Bestl - shRNA containing cassette, which then renders Bestl-shRNA inactive (Fig. 4b). As shown in Fig. 4b, Cre-ERT was glia specifically expressed under GFAP promoter and activated by tamoxifen injection but inactivated by shRNA injection.

The effect of glia specific rescue of Bestl was confirmed by immunohistochemistry with the Bestl antibody in either tamoxifen or sunflower oil injected hGFAP-CreERT2 mice (Fig. 8b). Fig. 4c shows the whole-cell patch clamp recording from granular cells. The upper trace shows tonic GABA current from hGFAP- CreERT mice injected with Bl-shRNA lentivirus after sunflower oil treatment whereas the lower left, raw trace indicates tonic GABA current from hGFAP-CreERT mice injected with Bl-shRNA lentivirus after tamoxifen treatment. In the hGFAP-CreERT2 mice treated with sunflower oil and injected with Bestl-shRNA lentivirus, GABAzine- sensitive current was significantly reduced (Fig. 4c, upper trace, 35.68±4.05pA (naive, n=8), 11.27±1.22pA (with sunflower oil, n=9), p < 0.003) to the similar level as that of wild type B6 mice injected with the same lentivirus (Fig 3b). However, in the hGFAP- CreERT2 mice treated with tamoxifen and injected with Bestl-shRNA lentivirus, GABAzine-sensitive currents were fully rescued to the naive animal level (with Tamoxifen: 31.31±2.19pA (n=8), naϊve: 35.68±4.05pA (n=12), p=0.36).

Fig. 4d shows the GABAzine-sensitive currents with and without Tamoxifen treatment. The GABAzine-sensitive currents with and without Tamoxifen treatment showed a significant difference (p«0.001). Fig. 4e shows the NPPB-sensitive currents with and without Tamoxifen treatment. The NPPB-sensitive currents were fully rescued in the tamoxifen treated mice (naϊve: 18.95±2.47pA, n=8; with Tamoxifen: 19.27±2.2pA n=9, p=0.93; without Tamoxifen: 1.93±1.56pA , n=4, p=0.00005). These results indicate that glial Bestl channel is responsible for the majority of tonic GABA release detected in cerebellar granule cells.

The amount of NPPB sensitive and Bestl -mediated tonic GABA release is estimated to be about 70% of the total GABAzine-sensitive current. The source of remaining GABAzine sensitive, NPPB-insensitive or Bestl -independent current is currently unknown and needs further investigation. Cloned Bestrophin channels are known to be activated at low Ca²⁺ concentration range with apparent Kd for activation by Ca²⁺ in the range of ~200nM (Hartzell, et al, 2008). If native Bestl channel has the same Ca²⁺ sensitivity, Bestl channels have to be partially activated at all times, because basal free cytosolic Ca²⁺ is typically around 10OnM, resulting in a constitutive release of GABA through these channels. Consistent with this idea, chelating free cytosolic Ca²⁺ with 25 min BAPTA-AM treatment significantly reduced the GABAzine- sensitive current (Fig. 6c, 6d, 6f, and 6g). Fig. 6c shows the tonic GABA current when incubated with 150 μM of BAPTA-AM in granular cells. Fig 6d shows the tonic GABA current when incubated with 0.5 μM of concanamycin A. As seen in Fig.όc and 6d, the current dramatically decreased when treated with BAPTA-AM but concanamycin A did not have such affect. Fig. 6f shows GABAzine sensitive current with no treatment, concanamycin A-treatment, and BAPTA-AM treatment.

The results from the Clomeleon imaging suggest that tonic GABA can be readily detected from the parallel fibers in the molecular layer. GABA released from the neighboring Bergmann glial processes can serve a role as powerful inhibitor, profoundly affecting the local excitability of parallel fibers and synaptic release of glutamate onto the dendrites of Pukinje cells (Fig. 4f). Fig. 4f is a suggested model for tonic GABA release in cerebellum.

In summary, the present invention demonstrates an unprecedented mechanism of tonic GABA release through a recently characterized bestrophin channel in cerebellar glial cells, a unique role of anion channel in channel-mediated release of transmitter by direct permeation, and a novel glial function in releasing the major inhibitory transmitter GABA to modulate the neuronal excitability. The importance of this channel-mediated release of inhibitory gliotransmitter should provide further understanding of many unexplored physiological roles of glial cells in brain function.

Claims

What is claimed is:

1. A composition for preventing, improving, or treating a disease or a symptom caused by over-release of GABA (gamma-aminobutyric acid), wherein said composition comprises a Bestrophin 1 channel inhibitor as an active ingredient such that said composition inhibits a GABA release in cerebellum.

2. The composition according to Claim 1, wherein said Bestrophin 1 channel inhibitor is one or more selected from the group consisting of anion channel blockers and antisense RNAs and shRNAs (small hairpin RNAs) for Bestrophin 1 channel-coding nucleotide sequences.

3. The composition according to Claim 2, wherein said anion channel blocker is one or more selected from the group consisting of niflumic acid, flumenamic acid, NPPB (5-nitro-2(3-phenylpropylamino)-benzoic acid), and DIDS (4,4'- diisothiocyanatostilbene-2,2'-disulfonic acid).

4. The composition according to Claim 2, wherein said Bestrophin 1 channel-coding nucleotide sequence is the nucleotide of SEQ ID NO: 1 or 2.

5. The composition according to Claim 2, wherein said shRNA is one or more selected from the group consisting of the nucleotides of SEQ ID NOs: 3, 4, and 7.

6. The composition according to any one of Claims 1 to 5, wherein said disease or symptom is selected from the group consisting of epilepsy, sleeping difficulties, memory difficulties, sensory difficulties, cognitive difficulties, motor difficulties, learning difficulties, alcohol addiction, and ataxia.

7. A composition for preventing, improving, or treating a disease or a symptom caused by the deficit of GABA (gamma-aminobutyric acid), wherein said composition comprises a Bestrophin 1 channel activator as an active ingredient such that said composition promotes a GABA release in cerebellum.

8. The composition according to Claim 7, wherein said Bestrophin 1 channel activator is one or more selected from the group consisting of peptide TFLLR and Bradykinin.

9. The composition according to Claim 7 or 8, wherein said disease or symptom is selected from the group consisting of Alzheimer, memory loss with aging, seizures, excitotoxicity, ischemia, cerebral apoplexy, cerebral hemorrhage, epilepsy, brain injuries, and hypoxia.

10. A screening method for a cerebellar GABA release-regulating agent, said method comprising the steps of:

preparing a cerebellar sample;

contacting a candidate material to the cerebellar sample; and

verifying the activation of Bestrophin channel in the cerebellar sample, wherein said candidate material is determined to be a GABA release-promoting agent when the Bestrophin 1 channel is found to be activated, whereas the candidate material is determined to be a GABA release-inhibiting agent when the Bestrophin 1 channel is found to be inactivated.

11. The screening method according to Claim 10, wherein the activation of Bestrophin channel in cerebellar glial cells is verified.

12. The screening method according to Claim 10, wherein the activation of Bestrophin channel is verified by the measurement of an inward current change using sniffer patch technique.

13. A method of preventing, improving, or treating a disease or a symptom caused by over-release of GABA (gamma-aminobutyric acid), wherein said method comprises the steps of:

identifying a patient with the with the disease or a symptom caused by over- release of GABA; and administering an effective amount of a Bestrophin 1 channel inhibitor to the patient, to inhibit a GABA release in cerebellum.

14. The method according to Claim 13, wherein said Bestrophin 1 channel inhibitor is one or more selected from the group consisting of anion channel blockers and antisense RNAs and shRNAs (small hairpin RNAs) for Bestrophin 1 channel- coding nucleotide sequences.

15. The method according to Claim 14, wherein said anion channel blocker is one or more selected from the group consisting of niflumic acid, flumenamic acid,

NPPB (5-nitro-2(3-phenylpropylamino)-benzoic acid), and DIDS (4,4'- diisothiocyanatostilbene-2,2'-disulfonic acid).

16. The method according to Claim 14, wherein said Bestrophin 1 channel- coding nucleotide sequence is the nucleotide of SEQ ID NO: 1 or 2.

17. The method according to Claim 14, wherein said shRNA is one or more selected from the group consisting of the nucleotides of SEQ ID NOs: 3, 4, and 7.

18. The method according to any one of Claims 13, wherein said disease or symptom is selected from the group consisting of epilepsy, sleeping difficulties, memory difficulties, sensory difficulties, cognitive difficulties, motor difficulties, learning difficulties, alcohol addiction, and ataxia.

19. A method of preventing, improving, or treating a disease or a symptom caused by the deficit of GABA (gamma-aminobutyric acid), wherein said method comprises the steps of:

identifying a patient with the disease or a symptom caused by the deficit of

GABA; and

administering an effective amount of a Bestrophin 1 channel activator to the patient, to promote a GABA release in cerebellum.

20. The method according to Claim 19, wherein said Bestrophin 1 channel activator is one or more selected from the group consisting of peptide TFLLR and Bradykinin.

21. The composition according to Claim 19 or 20, wherein said disease or symptom is selected from the group consisting of Alzheimer, memory loss with aging, seizures, excitotoxicity, ischemia, cerebral apoplexy, cerebral hemorrhage, epilepsy, brain injuries, and hypoxia.

22. A use of a Bestrophin 1 channel inhibitor for preventing, improving, or treating a disease or a symptom caused by over-release of GABA (gamma-aminobutyric acid), wherein the Bestrophin 1 channel inhibitor inhibits a GABA release in cerebellum.

23. The use according to Claim 22, wherein said Bestrophin 1 channel inhibitor is one or more selected from the group consisting of anion channel blockers and antisense RNAs and shRNAs (small hairpin RNAs) for Bestrophin 1 channel- coding nucleotide sequences.

24. The use according to Claim 23, wherein said anion channel blocker is one or more selected from the group consisting of niflumic acid, flumenamic acid,

NPPB (5-nitro-2(3-phenylpropylamino)-benzoic acid), and DIDS (4,4'- diisothiocyanatostilbene-2,2'-disulfonic acid).

25. The use according to Claim 23, wherein said Bestrophin 1 channel- coding nucleotide sequence is the nucleotide of SEQ ID NO: 1 or 2.

26. The use according to Claim 23, wherein said shRNA is one or more selected from the group consisting of the nucleotides of SEQ ID NOs: 3, 4, and 7.

27. The use according to any one of Claims 22 to 26, wherein said disease or symptom is selected from the group consisting of epilepsy, sleeping difficulties, memory difficulties, sensory difficulties, cognitive difficulties, motor difficulties, learning difficulties, alcohol addiction, and ataxia.

28. A use of a Bestrophin 1 channel activator for preventing, improving, or treating a disease or a symptom caused by the deficit of GABA (gamma-aminobutyric acid), wherein the Bestrophin 1 channel activator promotes a GABA release in cerebellum.

29. The use according to Claim 28, wherein said Bestrophin 1 channel activator is one or more selected from the group consisting of peptide TFLLR and Bradykinin.

30. The use according to Claim 28 or 29, wherein said disease or symptom is selected from the group consisting of Alzheimer, memory loss with aging, seizures, excitotoxicity, ischemia, cerebral apoplexy, cerebral hemorrhage, epilepsy, brain injuries, and hypoxia.

31. A method of inhibiting release of GABA (gamma-aminobutyric acid) in cerebellum, comprising the step of:

administering an effective amount of a Bestrophin 1 channel inhibitor to a patient in need of inhibiting release of GABA, to inhibit Bestrophin 1 channel activity in a cerebellar glial cell.

32. The method according to Claim 31, wherein said Bestrophin 1 channel inhibitor is one or more selected from the group consisting of anion channel blockers and antisense RNAs and shRNAs (small hairpin RNAs) for Bestrophin 1 channel- coding nucleotide sequences.

33. The method according to Claim 32, wherein said anion channel blocker is one or more selected from the group consisting of niflumic acid, flumenamic acid, NPPB (5-mtro-2(3-phenylproρylarnino)-benzoic acid), and DIDS (4,4'- diisothiocyanatostilbene-2,2'-disulfonic acid) .

34. The method according to Claim 32, wherein said Bestrophin 1 channel- coding nucleotide sequence is the nucleotide of SEQ ID NO: 1 or 2.

35. The method according to Claim 32, wherein said shRNA is one or more selected from the group consisting of the nucleotides of SEQ ID NOs: 3, 4, and 7.

36. A method of promoting release of GABA (gamma-aminobutyric acid) in cerebellum, comprising the step of:

administering an effective amount of a Bestrophin 1 channel activator to a patient in need of promoting release of GABA, to promote Bestrophin 1 channel activity in a cerebellar glial cell.

37. The method according to Claim 36, wherein said Bestrophin 1 channel activator is one or more selected from the group consisting of peptide TFLLR and Bradykinin.

38. A use of a Bestrophin 1 channel inhibitor for inhibiting release of GABA (gamma-aminobutyric acid) in cerebellum.

39. The use according to Claim 38, wherein said Bestrophin 1 channel inhibitor is one or more selected from the group consisting of anion channel blockers and antisense RNAs and shRNAs (small hairpin RNAs) for Bestrophin 1 channel- coding nucleotide sequences.

40. The use according to Claim 39, wherein said anion channel blocker is one or more selected from the group consisting of niflumic acid, fiumenamic acid, NPPB (5-nitro-2(3-phenylpropylamino)-benzoic acid), and DIDS (4,4'- diisothiocyanatostilbene-2,2'-disulfonic acid).

41. The use according to Claim 39, wherein said Bestrophin 1 channel- coding nucleotide sequence is the nucleotide of SEQ ID NO: 1 or 2.

42. The use according to Claim 39, wherein said shRNA is one or more selected from the group consisting of the nucleotides of SEQ ID NOs: 3, 4, and 7.

43. A use of a Bestrophin 1 channel activator for promoting release of GABA (gamma-aminobutyric acid) in cerebellum.

44. The use according to Claim 43, wherein said Bestrophin 1 channel activator is one or more selected from the group consisting of peptide TFLLR and Bradykinin.