|Home | About | Journals | Submit | Contact Us | Français|
Spinocerebellar ataxia type 2 (SCA2) is a neurodegenerative disorder caused by a polyglutamine expansion within the Ataxin-2 (Atxn2) protein. Purkinje cells (PC) of the cerebellum fire irregularly and eventually die in SCA2. We show here that the type 2 small conductance calcium-activated potassium channel (SK2) play a key role in control of normal PC activity. Using cerebellar slices from transgenic SCA2 mice we demonstrate that SK channel modulators restore regular pacemaker activity of SCA2 PCs. Furthermore, we also show that oral delivery of a novel selective positive modulator of SK2/3 channels (NS13001) alleviates behavioural and neuropathological phenotypes of aging SCA2 transgenic mice. We conclude that SK2 channels constitute a novel target for SCA2 treatment and that the developed selective SK2/3 modulator NS13001 holds promise as a potential therapeutic agent for treatment of SCA2 and possibly other cerebellar ataxias.
Spinocerebellar ataxia type 2 (SCA2) belongs to the family of polyglutamine expansion (polyQ) disorders. This group of degenerative and hereditary diseases also comprises Huntington's disease (HD), dentatorubropallidoluysian atrophy (DRPLA), spinobulbar muscular atrophy (SBMA) and other SCAs, including SCA1, SCA3 (Machado-Joseph disease), SCA6, SCA7, and SCA17 (Carlson et al., 2009; Matilla-Duenas et al., 2009; Orr and Zoghbi, 2007). In these polyQ disorders, an unstable CAG expansion within the disease-causing gene encodes an elongated polyQ tract, which is associated with a progressive neuronal degeneration accompanied by different clinical manifestations that depend on the function and expression pattern of the affected protein. A common feature of SCAs is a progressive cerebellar ataxia (Klockgether et al., 1998). In SCA2, disease pathogenesis is caused by polyQ expansion of more than 57 repeats in the ataxin-2 (Atxn2) protein (Pulst et al., 1996), the function of which is not well understood. Interestingly, polyQ repeat expansions of intermediate length in Atxn2 have been associated with amyotrophic lateral sclerosis (ALS) and parkinsonian symptoms (Ross et al., 2011; Simon-Sanchez et al., 2005). The cerebellar ataxia in SCA2 is associated with a loss of Purkinje cells (PCs) and generation of cytoplasmic inclusions (Huynh et al., 2000; Liu et al., 2009). The nuclear inclusion bodies characteristic of other polyQ disorders are not prominent in SCA2 (Huynh et al., 2000; Lastres-Becker et al., 2008). The reason for PC degeneration in SCA2 and other SCAs is not fully understood (Bezprozvanny and Klockgether, 2010; Kasumu and Bezprozvanny, 2010; Matilla-Duenas et al., 2009).
PCs exhibit a tonic pacemaking activity that is crucial for the correct encoding of cortical cerebellar information to deep cerebellar nuclei and further to other motor coordination areas (Ito, 2002). In a recent study, we demonstrated that pacemaking activity of PCs is abnormal in aging SCA2 mice (Kasumu et al., 2012). Similar disruptions of PC pacemaking have been reported in slices from mouse models of SCA3 (Shakkottai et al., 2011) and episodic ataxia type-2 (EA2) (Walter et al., 2006). Small conductance Ca2+-activated K+ channels (SK channels) play a key role in control of regular tonic firing in PCs (Womack and Khodakhah, 2003) and the two broad-specificity SK/IK channel activators chlorzoxazone (CHZ) and 1-ethyl-2-benzimidazolinone (1-EBIO) normalize PC firing and exert beneficial effects in a mouse model of EA2 (Alvina and Khodakhah, 2010a, b; Walter et al., 2006). Three subtypes of SK channels are expressed in the brain (Adelman et al., 2012; Kuiper et al., 2012) with the SK2 isoform predominant in PCs (Cingolani et al., 2002; Hosy et al., 2011; Sailer et al., 2004). Thus, the SK2 channel subtype is the most attractive pharmacological target for treatment of cerebellar ataxia. Indeed, we here directly demonstrate that SK2 controls normal PC pacemaking. We also demonstrate that application of the SK/IK modulator NS309 and the SK3/SK2 modulator CyPPA restore regular pacemaker activity of cerebellar PCs from SCA2 transgenic mice. We further report development of a novel potent and selective modulator of SK2/3 channels (NS13001) and show that oral delivery of this compound resulted in significant beneficial effects in the transgenic mouse model of SCA2. We conclude that NS13001 holds promise as a potential therapeutic agent for treatment of SCA2 and possibly other cerebellar ataxias.
SK channels play a key role in the control of pacemaking in PCs (Womack and Khodakhah, 2003). To confirm that positive SK modulators can exert a modulatory effect on PC firing, we performed a series of whole-cell recordings of PC activity in rat cerebellar slices using NS309 (3-oxime-6,7-dichloro-1H-indole-2,3-dione), a high potency positive modulator of SK/IK channels (Table 1) (Strobaek et al., 2004). In the majority of experiments, regular action potential (AP) firing of PCs occurring in the frequency range 20-30 Hz (average frequency 23 ± 5 Hz (n = 6 PCs)) was recorded under control conditions resulting in aninterspike interval close to 50 ms (Fig 1A, 1B). Bath application of 3 μM NS309 caused a marked reduction in the firing frequency to 9 ± 3 Hz (n=6 PCs) (Fig 1A) and a concomitant increase in the interspike interval to above 150 ms (Fig 1B). The effect of NS309 was reversible and the spontaneous activity of PCs was restored towards the initial frequency following washout (Fig 1A). Prolonged exposure to NS309 frequently lead to a complete silencing of the cell (not shown) but upon extended wash-out, pacemaker activity could be restored. The bee venom toxin apamin specifically blocks SK channels (Adelman et al., 2012). Within minutes of application, 200 nM apamin changed the firing pattern of PCs from tonic to regular frequency bursts separated by silent periods (Fig 1A). The interspike interval within each burst was reduced to less than 15 ms in the presence of apamin (Fig 1B). Three subtypes of SK channels are expressed in the brain (Adelman et al., 2012; Patko et al., 2003; Sailer et al., 2002; Sailer et al., 2004). The SK2 isoform is highly expressed in PCs (Cingolani et al., 2002; Hosy et al., 2011; Sailer et al., 2004). In order to evaluate the importance of this subtype for regulating the intrinsic firing properties of PCs, we challenged regularly firing cells with the highly potent and selective synthetic SK2 toxin inhibitor Lei-Dab7 (Shakkottai et al., 2001). We confirmed the subtype specificity of the Lei-Dab7 batch available to us (Fig S1). In experiments with recombinant channels we demonstrated that Lei-Dab7 inhibits hSK2 channels with IC50 equal to 7 ± 1 nM (n = 3). In contrast, Lei-Dab7 inhibited recombinant hSK3 channels with IC50 equal to 1.8 ± 0.6 μM (n = 3) and recombinant hSK1 channels with IC50 equal to 27 ± 11 μM (n = 3). Thus, our batch of Lei-Dab7 is at least 200-fold more specific for SK2 channels when compared to SK1 and SK3 channels, in agreement with the published observations (Shakkottai et al., 2001). We discovered that at 100 nM, a concentration which strongly inhibits SK2 but does not affect SK1 or SK3 channels (Fig S1), Lei-Dab7, like apamin, changed the rat PC firing pattern from tonic into pronounced bursting (Figure 1C). The results with Lei-Dab7 support the key role of SK2 channels in control of PC pacemaking activity, in agreement with recently reached conclusions (Hosy et al., 2011).
Previously, we discovered that cerebellar slices from aging SCA2-58Q (58Q) transgenic mice have a significantly higher fraction of bursting PCs when compared to slices from age-matched wild type mice (Kasumu et al., 2012). Consistent with these findings, we found that most PCs (91 ± 10%) in slices from 24 week old wild type mice exhibited tonic activity (Fig 2A) characterized by stable firing rates (Fig 2C). In contrast, fewer PCs (64 ± 9%) in slices from 24-week old 58Q mice exhibited tonic activity. PCs in slices from 24-week old SCA2-58Q mice instead exhibited bursting activity (Fig 2B). We observed various firing patterns in bursting 58Q PCs. Some PCs fired regular high-frequency bursts separated by brief silent periods, a pattern that we named “persistent bursting” (Figure 2B, 2D). Other PCs displayed periods of relatively constant frequency firing separated by short periods of silence or increased frequency of firing. These cells were classified as “transiently bursting”. In the previous study, we reasoned that the burst firing pattern of SCA2 PCs reflects the deteriorating health and loss of metabolic control of these cells and that this might be casually connected with the impaired motor performance of aging SCA2 mice (Kasumu et al., 2012). Thus, agents that can switch bursting SCA2 PCs to tonic firing may have a potential therapeutic value for SCA2.
To investigate if modulators of SK channels can rescue the abnormal firing of SCA2 PCs, we performed slice recording experiments with slices from 24 week old 58Q mice in the presence of the pan-SK channel modulator NS309 and the selective SK2/3 modulator CyPPA (cyclohexyl-[2-(3,5-dimethyl-pyrazol-1-yl)-6-methyl-pyrimidin-4-yl]-amine) (Hougaard et al., 2007) (Table 1). Application of 5 μM NS309 switched “persistently bursting” 58Q PCs to tonic firing pattern (Fig 2E, n = 11 of 11 PCs). Application of 5 μM NS309 also switched “transiently bursting” 58Q PCs to tonic firing pattern (Fig 2G, n = 7 of 7 PCs). CyPPA was less effective than NS309 in experiments with “persistently bursting” PCs since application of 5 μM CyPPA was only able to switch “persistently bursting” 58Q PCs to tonic firing pattern in 5 out of 11 experiments (Fig 2F). However, similar to NS309, application of 5 μM CyPPA switched “transiently bursting” SCA2-58Q PCs to tonic firing pattern (Fig 2H, n = 6 of 7 PCs). These experiments suggested that activation of SK2 channels provides a potential strategy for restoring tonic firing of PC cells in aging SCA2 mice. When compared to NS309, the lower efficacy of CyPPA in these experiments is likely to be due to the relatively lower potency of this compound as an SK2 channel modulator (Hougaard et al., 2007), (see also Table 1).
The SK channel activators CHZ, 1-EBIO and SKA-31 used in previous studies with EA2 and SCA3 ataxic mice (Alvina and Khodakhah, 2010a, b; Shakkottai et al., 2011; Walter et al., 2006) have low potency and lack subtype selectivity (Table 1). CyPPA (Fig 3A) is a well-characterized and selective positive modulator of SK2/3 channels (SK3 > SK2 >>> SK1 = IK) (Hougaard et al., 2007) (Table 1, Fig S2), whereas NS309 is the potent pan-selective IK/SK channel modulator (IK > SK1 = SK2 = SK3) (Strobaek et al., 2004) (Fig 3A, Table 1, Fig S2). In our studies we set out to develop an SK channel modulator that combines potency of NS309 and selectivity of CyPPA. To achieve this goal, a chemical optimization program based on the CyPPA scaffold was conducted at NeuroSearch (Palle Christophersen, personal communication), leading to the novel compound NS13001 (4-Chloro-phenyl)-[2-(3,5-dimethyl-pyrazol-1-yl)-9-methyl-9H-purin-6-yl]-amine) (Fig 3A) (Eriksen et al., 2008). The procedures for chemical synthesis of NS13001 (Fig S3) are described in supplementary online materials. In experiments using inside-out patches, 1 μM NS13001 potently activated hSK3, less potently hSK2 and had no activating effect on hSK1 (Fig S2) or hIK channels (data not shown). Thus, NS13001 recapitulates the basic subtype selectivity properties of the lead molecule CyPPA, but with the potency comparable to NS309. When compared to NS309 and CyPPA, NS13001 also exerted less off-target effects. In the micromolar concentration range NS309 blocks hERG channels (IC50 = 1.3 μM) (Strobaek et al., 2004) and CyPPA blocks voltage-gated sodium channels (IC50 = 11 μM) (Hougaard et al., 2007), whereas NS13001 had no effect on these channels at concentrations as high as 10 μM (data not shown).
The mechanism of NS13001 action was characterized in more detail, using hSK3 channels. Fig 3B shows current-voltage (I-V) relationships measured at symmetrical K+ and with Ca2+ buffered at 0.2 μM Ca2+ or 10 μM Ca2+ (solid lines). Application of NS13001 in the range from 0.001 μM to 10 μM to the inside of the patch at a [Ca2+]i of 0.2 μM resulted in a concentration-dependent increase in the hSK3 current (Fig 3B, broken lines). The maximal activation of hSK3 channels by NS13001 was 90% of the level observed at 10 μM cytosolic Ca2+ which activates SK channels maximally (Fig 3B). The characteristic inward rectification of the current was maintained when the channels were activated by NS13001 at low Ca2+ (Fig 3B). Fig 3C depicts the hSK3 current recorded at -75 mV as a function of time showing the effects of increasing concentrations of NS13001. To quantify the concentration-dependence of NS13001, a number of similar experiments were performed with patches from hSK1, hSK2 and hSK3 expressing cells and the current response in each experiment was normalized to the size of the current recorded at 10 μM Ca2+ for the same patch. The normalized data were averaged and plotted as a function of NS13001 concentration for each hSK subtype (Fig 3D). Data fitted by the Hill equation (solid line) yielded an EC50 value for hSK3 activation of 0.14 μM, a Hill coefficient of 1 and an efficacy of 91 % (Fig 3D, Table 1). For hSK2 the EC50 was 1.6 μM, the Hill coefficient was 1.4 and the efficacy was 90% (Fig 3D, Table 1), whereas concentrations of NS13001 higher than 10 μM were needed to induce only a marginal increase in the hSK1 current (Fig 3D, Table 1). These results confirmed that the subtype selectivity properties of NS13001 is SK3 > SK2 >>> SK1.
To further understand the mechanism of NS13001 action, we evaluated the Ca2+-dependence of its effects on SK channels. At very low cytosolic Ca2+ concentrations (≤ 0.01 μM Ca2+), the application of 1 μM NS13001 did not result in increased activity of hSK3 channels (Fig 3E, 3F). In contrast, in the middle range Ca2+ concentrations (0.2 μM Ca2+), application of 1 μM NS13001 induced a large, reversible increase in the current level at all membrane potentials tested (Fig 3E, 3F). At high cytosolic Ca2+ concentration (10 μM Ca2+), the application of NS13001 induced a small, but significant reduction in the hSK3 current level (Fig 3E, 3F). To test the effects on the SK Ca2+-dependence more generally, we determined the Ca2+-activation curves of hSK1, hSK2 and hSK3 in the absence or presence of 1 μM NS13001. In the absence of compound, the Ca2+-dependence for all 3 SK channel subtypes is virtually identical, with EC50 of 0.42 μM Ca2+ (Hougaard et al., 2007) (Fig 3G, dashed line). The presence of 1 μM NS13001 induced a pronounced left-ward shift in the Ca2+-activation curve of hSK3, resulting in an EC50 for Ca2+ of 0.11 μM (Fig 3G). Presence of NS13001 also resulted in a small reduction in the maximal activity of hSK3 channels at the highest Ca2+ concentrations (Fig 3G). Similarly, 1 μM NS13001 also increased the apparent Ca2+-sensitivity of hSK2, with a new EC50 = 0.18 μM (Fig 3G) and induced a small reduction of the hSK2-mediated current at high Ca2+ concentrations (Fig 3G). In contrast to hSK2 and hSK3 channels, 1 μM NS13001 did not have any stimulating effect on the Ca2+ dependence of hSK1 channels, but also resulted in a reduction of the current at high concentrations of Ca2+ (Fig 3G). Based on the obtained results, we conclude that NS13001 primarily acts as a potent and selective positive allosteric modulator of SK2 and SK3 channels.
In order to test the effects of SK channel positive modulators on the motor performance of symptomatic SCA2 mice, we evaluated the effects of oral delivery of NS13001 and CyPPA to a group of 9 months old 58Q mice. NS309 was not utilized in these studies as our pilot studies demonstrated that this compound is less stable in vivo and has significantly inferior pharmacokinetic and brain penetration properties compared to NS13001 and CyPPA (data not shown). Thus, although NS309 provides a powerful tool for in vitro experiments with brain slices (Fig 1 and Fig 2), it is less appropriate for in vivo long term studies with repeated dosing. The in vivo studies were designed following the same general principles as in the previous evaluation of dantrolene feeding and 5PP overexpression in 58Q mice (Kasumu et al., 2012; Liu et al., 2009). Briefly, 9-month old WT and 58Q mice were subdivided into 6 treatment groups with 10-15 mice in each group (Table 2). We confirmed that the average weight of the mice was similar for each group (Fig S4). The baseline motor performance was evaluated by beamwalk (11 mm round and 5 mm square beams) and accelerating rotarod assays. Consistent with the previous results (Kasumu et al., 2012; Liu et al., 2009), 9 months old 58Q mice were impaired in both assays when compared to age-matched wild type mice (Fig 4; Fig S5). Following the baseline test, the mice were orally fed daily with 30 mg/kg NS13001 or10 mg/kg CyPPA. The dose of CyPPA was chosen based on the previous in vivo studies with this compound (Herrik et al., 2012; Vick et al., 2010). The higher dose of NS13001 was chosen based on its more favorable target selectivity profile. In pilot experiments, we established that 1 hour after oral delivery of 30 mg/kg of NS13001 to adult mice the concentration in blood plasma was 16 μM (8 μM after 6 hours). After 1 hour, the concentration in the brain was 17 μM (data not shown).
The 58Q and WT mice were fed with the compound orally starting at 9 months of age for 3 consecutive weeks with the control mice fed with the vehicle alone. After three weeks on this dosing regimen, feeding of the compounds was halted for 3 days. Following this brief washout period, motor coordination of each mouse was re-tested using identical beamwalk and rotarod assays. The average body weight for all 6 groups remained constant after feeding with the compounds (Fig S4). All 3 groups of WT mice displayed similar levels of motor performance after drug treatment. The latency of crossing the 11 mm beam was reduced for all 3 groups of WT mice (Fig 4A), presumably due to learning the task. There was no difference in the number of footslips that mice made crossing the 11 mm beam (Fig 4B). Testing on the 5 mm square beam revealed that the latency to cross and the number of footslips remained the same for all 3 groups of mice (Fig S5).
The control (vehicle-fed) group of 58Q mice traversed the 11 mm and 5 mm beams post-treatment similarly as prior to treatment (Fig 4A, Fig S5) and there was no significant difference in the number of footslips (Fig 4B, Fig S5). The rotarod performance of the vehicle-fed group of 58Q mice remained the same as prior to treatment (Fig 4C). In contrast to WT mice, NS13001-fed 58Q mice demonstrated significantly improved performance following treatment with the compound. In the 11 mm beam task a decrease in the latency to traverse the beam (Fig 4A; p < 0.01) and a decreased number of footslips (Figure 4B; p < 0.05) were observed. Both effects were replicated on the 5 mm beam (Fig S5). There was also a significant increase in the latency to fall off the accelerating rod following feeding of 58Q mice with NS13001 (Figure 4C, p < 0.05). Treatment with CyPPA also resulted in improved motor performance of 58Q mice, although beneficial effects were less pronounced than for NS13001. Treatment of 58Q mice with CyPPA resulted in a reduced latency (p < 0.05) to cross the 11 mm and 5 mm beams (Fig 4A, Fig S5) but had no significant effect on the number of footslips on either beam (Fig 4B, Fig S5) or on the rotarod performance of these mice (Fig 4C). The lower efficacy of CyPPA in these assays is probably explained by its lower potency (Table 1) and lower dosing level in the in vivo experiments. Following initial evaluation (Fig 4, Fig S5), we attempted to determine if effects of the drugs were reversible. All mice were returned to the home cages and re-tested again 2 months later. However, at this time-point the mice were 13 months old and many 58Q and WT had difficulty completing the motor tasks due to inability to remain on the balance beam or the rotating rod. For this reason we could not clearly discern treatment reversibility from other effects in this study.
At 13 months of age (2 months after drug treatment was finished), all mice were sacrificed and processed for neuropathological analysis. In previous studies, we demonstrated that quantification of dark cell degeneration (DCD) provides the most reliable and most sensitive way to score excitotoxic PC death in 58Q mice (Kasumu and Bezprozvanny, 2010; Kasumu et al., 2012). DCD has also been used to quantify excitotoxic PC death in SCA7 and SCA28 (Custer et al., 2006; Maltecca et al., 2009). DCD is a form of PC death characterized by morphological changes in PCs identifiable by transmission electron microscopy (TEM) of slices from 58Q mice (Fig 5A, right panel). In contrast, most PCs in age-matched WT mice look normal (Fig 5A, left panel). To analyse DCD, cerebellar sections from each of the 6 experimental groups of mice (Table 2) were processed for TEM and the number of normal, moderately and severely degenerated PCs was quantified. According to (Kasumu and Bezprozvanny, 2010; Kasumu et al., 2012), PCs spherical in shape and with regular alignment in the PC layer were classified as “normal” (Fig 5B, left panel). PCs with slight shrinkage compared to surrounding PCs and with moderately electron-dense cytosol that is not as dark as the nucleus, were classified as “moderate” (Fig. 5B, middle panel). PCs with markedly shrunken and electron-dense cytosol with similarly darkened nucleus were classified as “severe” (Fig. 5C, right panel). Consistent with our previous data (Kasumu and Bezprozvanny, 2010; Kasumu et al., 2012), we found that in samples from 58Q control mice, 11% of PCs were normal, 51% were moderately degenerated and 38% were severely degenerated (n = 296 PCs; Fig. 5C; Table 2). Also consistent with our previous data (Kasumu and Bezprozvanny, 2010; Kasumu et al., 2012), most cells were healthy in samples from age-matched WT control mice, in which 71% of PCs were scored as normal, 22% were moderately degenerated and 7% were severely degenerated (n = 258 PCs, Fig 5C, Table 2). The WT mice treated with NS13001 or CyPPA did not exhibit a significant change in the number of normal cells and moderately affected cells (Fig 5C, Table 2). In contrast, the samples from the 58Q mice exposed to NS13001, the fraction of normal PCs was increased to 43%, moderately degenerated cells reduced to 37% and degenerated cells reduced to 20% (n = 288 PCs; Fig. 5C, Table 2). Similarly, in samples from CyPPA-treated 58Q mice the fraction of normal cells was increased to 32%, the fractions of moderately degenerated cells and severely degenerated cells were reduced to 35% and 33%, respectively (n = 222 PCs; Fig. 5C, Table 2). When compared to vehicle-treated 58Q mice, the increase in the fraction of “normal” cells in 58Q mice treated with NS13001 or CyPPA was statistically significant (p < 0.001; Fig. 5C, Table 2). Similar to behavioural studies (Fig 4, Fig S5), the beneficial effects of CyPPA in DCD assay with 58Q mice were less pronounced than the beneficial effects of NS13001 in the same assay (Fig 5C, Table 2).
The cerebellum plays an essential role in learning and control of coordinated movements. The precision and speed of these movements requires exact timing of cerebellar output. The inhibitory projections from the PC to the deep cerebellar nuclei (DCN) constitute the sole output of the cerebellar cortex (Ito, 2002). Recent electrophysiological analysis confirmed that PC electrical activity is tightly coordinated at millisecond resolution (de Solages et al., 2008; Heck et al., 2007; Person and Raman, 2012). In slices, PCs spontaneously fire action potentials at a constant frequency in the range 17 – 150 Hz (Llinas and Sugimori, 1980a, b; Nam and Hockberger, 1997; Raman and Bean, 1997, 1999; Smith and Otis, 2003; Womack and Khodakhah, 2002). It is generally believed that this endogenous pacemaking activity of PCs represents the crucial background activity for correct encoding of the integrated cerebellar cortex information to DCN and other motor coordination areas. Cerebellar PCs are affected in many ataxias (Carlson et al., 2009; Matilla-Duenas et al., 2009; Orr and Zoghbi, 2007), and massive PC death is observed at the end stage of disease for many ataxic patients. However, it is becoming evident that early symptoms of ataxia may result not from PCs death but from PCs dysfunction and loss of firing precision. Consistent with this hypothesis, disruptions of regular PCs pacemaking activity have been uncovered in studies with mouse models of EA2 (Walter et al., 2006), SCA3 (Shakkottai et al., 2011), and SCA2 (Kasumu et al., 2012). Based on these findings, it has been argued that drugs that can normalize the regular firing of PCs may provide therapeutic benefit for ataxic patients (Rinaldo and Hansel, 2010; Shakkottai et al., 2004; Shakkottai et al., 2011; Walter et al., 2006).
There are multiple ion conductances that control the spontaneous electrical activity of PCs (Llinas and Sugimori, 1980a, b; Raman and Bean, 1997, 1999). Small conductance Ca2+-activated K+ channels (SK channels) emerged as one of the principle channel types involved in precise control of PC pacemaking (Womack and Khodakhah, 2003). A number of small molecule modulators of SK channels have previously been identified (Table 1), enabling pharmacological manipulation of SK channel activity in ataxic mouse models. The two broad-specificity SK/IK channel activators chlorzoxazone (CHZ) and 1-ethyl-2-benzimidazolinone (1-EBIO) normalized PC firing and exerted beneficial effects in a mouse model of EA2 (Alvina and Khodakhah, 2010a, b; Walter et al., 2006). Short-term exposure of SCA3 mice to SKA-31, a riluzole analogue optimized for positive modulation of SK channels (Table 1), provided benefit in a mouse model of SCA3 (Shakkottai et al., 2011). These results supported the hypothesis that positive modulators of SK channels may offer therapeutic benefit for treatment of ataxia. Indeed, riluzole yielded promising results in a recent phase II study in a mixed population of ataxia patients (Ristori et al., 2010), an effect that was interpreted as the ability of riluzole to facilitate the activity of SK channels (Table 1). Despite these promising results, most agents used in previous studies of ataxia had low potency, poor specificity and suboptimal blood brain permeability properties (Table 1).
Three subtypes of SK channels are expressed in the brain (Adelman et al., 2012; Kuiper et al., 2012; Patko et al., 2003; Sailer et al., 2002; Stocker, 2004; Stocker and Pedarzani, 2000). The SK2 isoform is predominant in PCs (Cingolani et al., 2002; Hosy et al., 2011; Sailer et al., 2004), while high levels of SK3 channels are expressed in cerebellar granule cells (Stocker and Pedarzani, 2000). SK3 single knockout mice lack a clear motor phenotype (Bond et al., 2000), but showed increased dopamine release in the striatum and certain changes in models of depression and anxiety (Jacobsen et al., 2009; Jacobsen et al., 2008). A naturally occurring SK2 loss-of-function mutation in mice (frissonant mice) causes prominent motor deficits (Callizot et al., 2001). In experiments with the SK2-specific synthetic toxin inhibitor Lei-Dab7 (Shakkottai et al., 2001) we now demonstrate the essential role of SK2 channels in the control of PCs spontaneous activity (Fig 1C). These results are in agreement with a recent report (Hosy et al., 2011). Furthermore, we demonstrated that the potent pan-SK channel modulator NS309 converted the “burst” firing pattern of aging PC cells from SCA2 transgenic mouse model to a tonic firing pattern (Fig 2E, 2G). The SK2/3-selective modulator CyPPA was also able to restore the “tonic” firing pattern of some SCA2-PCs but appeared less effective (Fig 2F, 2H). The difference between the effect of NS309 and CyPPA is most likely due to the substantially lower potency of CyPPA in activating SK2 channels (Table 1, Fig S2). Based on all these results we concluded that the SK2 channel subtype is the most attractive pharmacological target for treatment of cerebellar ataxia. We therefore set out to develop a new selective and potent SK2/3 positive modulator with CyPPA as lead molecule with improved pharmacokinetic and brain penetration properties.
NS13001 is a novel molecule (Fig 3A) (Eriksen et al., 2008) identified in an optimization program based on CyPPA (Palle Christophersen, personal communication). Our electrophysiological experiments revealed that NS13001 recapitulates the basic subtype selectivity properties of the lead molecule CyPPA (hSK3>hSK2>>>hSK1), with a potency comparable to NS309 (Fig 3, Fig 1S, Table 1). When compared to CyPPA or NS309, NS13001 is also considerably more stable towards metabolic degradation by liver microsomes in vitro (data not shown) and achieves significantly higher plasma and brain concentrations following oral administration to rats (data not shown). Similar to CyPPA (Hougaard et al., 2007), NS13001 acts as an allosteric modulator of SK2/3 channels, which increased their sensitivity to activation by cytosolic Ca2+ (Fig 3G). In our experiments, we evaluated a potential efficacy of N13001 and CyPPA in a transgenic mouse model of SCA2. In this mouse model human Atxn-58Q transgene is expressed under control of PC-specific promoter (Huynh et al., 2000), resulting in progressive development of motor symptoms and loss of PCs (Huynh et al., 2000; Kasumu and Bezprozvanny, 2010; Kasumu et al., 2012; Liu et al., 2009).
Both NS13001 (30 mg/kg) and CyPPA (10 mg/kg) were fed to SCA2 mice for 3 consecutive weeks starting at 9 months of age. The dose of CyPPA utilized in these studies was chosen based on the previous in vivo studies with this compound (Herrik et al., 2012; Vick et al., 2010) and its potential off-target effects (Hougaard et al., 2007). The higher dose of NS13001 was chosen based on its more favorable target selectivity profile. Following drug treatment, these mice were evaluated in motor coordination assays (balance beam walk and accelerating rotatod). We discovered that both NS13001 and CyPPA significantly improved performance of SCA2 mice in beamwalk assays, although effects of NS13001 were more pronounced (Fig 4A, 4B, Fig S5). The effects were specific, as performance of age-matched wild type mice was not significantly affected by either compound (Fig 4, Fig S5). NS13001, but not CyPPA, demonstrated efficacy in the rotarod assay (Fig 4C). Overall, these data strongly indicated that oral exposure to positive modulators of SK2/3 channels have the potential of improving motor performance of aging SCA2 mice. Much to our surprise, the benefit of administering NS13001 and CyPPA to SCA2 mice extended beyond improved motor performance after 3 weeks. At the conclusion of the study, we evaluated excitotoxic PC death in 13 months old SCA2 mice by quantifying their dark cell degeneration (DCD) status (Kasumu and Bezprozvanny, 2010; Kasumu et al., 2012). We found that PCs in SCA2 mice were partially protected from DCD (Fig 5C, Table 2) by both NS13001 and CyPPA. Similar to the behavioural assays, the degree of protection appeared to be greater in NS13001-fed mice than in CyPPA-fed mice (Fig 5C, Table 2).
What is an explanation of these findings? And what is the physiological target of NS13001 and CyPPA in these experiments? Both compounds have significantly higher selectivity for SK3 than for SK2 (Table 1, Fig 3D, Fig S1). However, in contrast to SK2, SK3 channels are not prominently expressed in PCs and the most likely molecular target of these compounds are thus SK2 channels (Cingolani et al., 2002; Hosy et al., 2011; Sailer et al., 2002; Sailer et al., 2004). High levels of SK3 channels are expressed in cerebellar granule cells (Sailer et al., 2002; Sailer et al., 2004; Stocker and Pedarzani, 2000) and in dopaminergic neurons of the substantia nigra (Sailer et al., 2002; Sailer et al., 2004). In addition, both SK3 and SK2 channels are present in DCN (Sailer et al., 2004; Shakkottai et al., 2004; Stocker and Pedarzani, 2000). It cannot be excluded that some of the behavioural effects of NS13001 and CyPPA are due to activation of SK3 channels in non-PC neurons. However, as SK3 single knockout mice lack a clear motor phenotype (Bond et al., 2000), this is not very likely. Thus, we propose that the beneficial effects of NS13001 and CyPPA in the SCA2 mouse model are primarily due to ability of these compounds to potentiate activity of SK2 channels in PCs of aging SCA2 mice.
There are also several potential explanations for the observed beneficial effects. The first explanation is that NS13001 and CyPPA converted “bursting” to “tonic” pattern of PCs in aging SCA2 mice (Fig 2) and helped information processing in cerebellum of these mice by restoring regular firing of PCs. The second explanation is that NS13001 or CyPPA induced low frequency tonic firing pattern of all PCs in aging SCA2 mice. PC action potentials are coupled to increases in the intracellular Ca2+ concentration in PC soma and dendrites due to opening of P-type voltage-gated Ca2+ channels (Sabatini et al., 2001). Handling of this Ca2+-influx puts PCs in conditions of latent metabolic stress, being proportional to the frequency of spontaneous firing, and likely to be strongly amplified during periods of uncontrolled bursting. Decreased frequency of PCs tonic firing and in particular reversion of bursting in the presence of NS13001 or CyPPA most likely reduces Ca2+ influx and leads to much lower metabolic demand of these cells. The reduction of Ca2+ influx is particularly critical for PCs in SCA2 mice, which already have supranomal cytosolic Ca2+ signals due to pathogenic interactions between mutant ataxin-2 and InsP3R1 (Kasumu and Bezprozvanny, 2010; Kasumu et al., 2012; Liu et al., 2009). It is possible that during the drug administration period PCs in SCA2 mice were exposed to a 3-week “metabolic holiday”, giving them a chance to recover from Ca2+ overload and rejuvenate. This latter explanation is consistent with DCD data collected 2 months after drug feeding was discontinued (Fig 5C, Table 2), which demonstrated long lasting neuroprotection in drug-exposed SCA2 mice. These results lead us to suggest that NS13001 and related compounds may exert not only symptomatic but also neuroprotective effects for cerebellar ataxia patients.
The proposed explanation of NS13001 or CyPPA ability to protect PCs in SCA2 mice is consistent with studies in Parkinson's disease (PD) field, where it was demonstrated that reducing voltage-dependent Ca2+ influx during pacemaker firing of substantia nigra (SNc) neurons leads to neuroprotection in models of PD (Chan et al., 2009, 2010; Surmeier, 2007; Surmeier et al., 2010) and possibly in PD patients (Becker et al., 2008; Ritz et al., 2010) (but see (Louis et al., 2009; Simon et al., 2010)). In case of studies in PD models, the Ca2+ influx in SNc neurons was reduced not by slowing down pacemaking activity of these cells but by pharmacological block of CaV1.3 voltage-gated Ca2+ channels which mediate most of Ca2+ influx in these cells during spontaneous activity (Chan et al., 2010). Similar to PCs, SK channels are involved in the control of firing rates of dopaminergic neurons in the SN neurons (Johnson and Wu, 2004; Kuznetsov et al., 2006; Shepard and Bunney, 1988), which express high levels of SK3 channels (Sailer et al., 2002; Sailer et al., 2004). Downregulation of SK channels and increased bursting frequency of SN neurons have been related to PD-linked genetic mutations (Bishop et al., 2010). If NS13001 indeed acted in our experiments by reducing “metabolic burden” on SCA2 PC cells, it is likely that NS13001 and related compounds may offer potential benefit not only for cerebellar ataxias but also for PD and for other neurodegenerative disorders that affect SK2 or SK3-expressing pacemaking neurons. This hypothesis is consistent with recently reported neuroprotective effects of CyPPA in experiments with dopaminergic neuronal cultures (Benitez et al., 2011; Herrik et al., 2012). The data in the current manuscript suggest that NS13001 should exert even more potent protective effect than CyPPA on SNc neurons both in vitro and in vivo. Evaluation of NS13001 in animal models of PD and other neurodegenerative disorders will be required to test these predictions.
The most serious potential side effects related to using modulators of SK2/3 channels are likely to be related to potential memory and learning impairments (Kuiper et al., 2012). Hippocampal-dependent memory tasks were potentiated by blocking SK channels with apamin (Stackman et al., 2002; Vick et al., 2010) and impaired by transgenic overexpression of SK2 channels (Hammond et al., 2006; Stackman et al., 2008). Transient down-regulation of SK2 and SK3 channels was reported during spatial learning paradigm in rats (Mpari et al., 2010). Moreover, systemic administration of 15 mg/kg of CyPPA resulted in object memory encoding deficits in mice (Vick et al., 2010). Based on these results it is likely that the “therapeutic window” for usage of NS13001 and other SK2/3 modulators for treatment of neurodegeneration will be eventually determined by the balance between neuroprotective effects on bursting cells (such as PC cells in ataxias, SNc cells in PD) and memory impairing effects in hippocampus. Future studies with animal models of disease and human clinical trials will be needed to find an appropriate dosage and delivery regiment for these compounds to achieve maximal benefit with minimal side-effects.
Cerebellar ataxias are a group of genetic disorders that are caused by progressive dysfunction and death of cerebellar PCs. SK channels play a key role in control of PC firing rates (Womack and Khodakhah, 2003) and a number of previous studies suggested that pharmacological modulators of SK channels may exert beneficial effects in cerebellar ataxia mouse models (Alvina and Khodakhah, 2010a, b; Shakkottai et al., 2011; Walter et al., 2006). Riluzole yielded promising results in a recent phase II study in a mixed population of ataxia patients (Ristori et al., 2010), an effect that was suggested to be related to the ability of riluzole to facilitate the activity of SK channels. Despite these promising results, most agents used in previous studies of ataxia had low potency and poor specificity (Table 1). We here report a novel compound NS13001 that acts as a potent and selective positive modulator of SK2/3 channels. We demonstrate that SK2 channels play a key role in control of pacemaking activity of cerebellar PCs and established that application of SK modulators restores tonic firing pattern of bursting PCs from aging mouse model of SCA2. We demonstrated that 3 weeks oral feeding of NS13001 resulted in improved performance of aging SCA2 mice in motor coordination assays and reduced PC degeneration in these mice. Similar, but less pronounced, positive effects were observed in SCA2 mice fed with CyPPA. The most likely mechanism responsible for beneficial effects of NS13001 and CyPPA is a reduction in Ca2+ influx and related metabolic stress due to normalized spontaneous activity of SCA2 PCs. From these results we conclude that NS13001 holds promise as a potential therapeutic agent for treatment of SCA2 and possibly other cerebellar ataxias. We reasoned that NS13001 may also be useful for treatment of other neurodegenerative disorders that affect pacemaking cells expressing SK2/3 channels, such as for example dopaminergic neurons in SNc (Benitez et al., 2011; Chan et al., 2009, 2010; Surmeier, 2007; Surmeier et al., 2010). Evaluation of NS13001 in animal models of ataxia, PD and other neurodegenerative disorders will be required to test these predictions. The most serious potential side effects related to using modulators of SK2/3 channels are likely to be related to potential memory and learning impairments (Hammond et al., 2006; Stackman et al., 2008; Vick et al., 2010), which may eventually determine a limit on clinically useful doses of these compounds for treatment of neurodegeneration.
Cyclohexyl-[2-(3,5-dimethyl-pyrazol-1-yl)-6-methyl-pyrimidin-4-yl]-amine (CyPPA) and 3-Oxime-6,7-dichloro-1H-indole-2,3-dione (NS309) were previously described (Hougaard et al., 2007; Strobaek et al., 2004). (4-Chloro-phenyl)-[2-(3,5-dimethyl-pyrazol-1-yl)-9-methyl-9H-purin-6-yl]-amine (NS13001) is a novel compound (Eriksen et al., 2008) and its synthesis is described in SOM. Apamin was purchased from Sigma Aldrich. Lei-Dab7 was synthesized in the Sabatier laboratory by following published procedures (Shakkottai et al., 2001).
Procedures involving wildtype rats were conducted in strict accordance with the guidelines described in the Guide for Care and Use of Laboratory Animals, the policies adopted by the Society for Neuroscience, and the Danish Committee for Experiments on Animals. SCA2-58Q mice on C57/B6 background (Huynh et al., 2000) were kindly provided to our laboratory by Dr Stefan Pulst (Univ of Utah) and have been used in our previous studies (Liu et al., 2009). In these mice the expression of human Atx2-58Q transgene is driven by the PC-specific L7/pcp2 promoter (Huynh et al., 2000). The mice were back-crossed to FVB/N background for at least 6 generations in our laboratory as previously described (Kasumu and Bezprozvanny, 2010; Kasumu et al., 2012). The SCA2-58Q (FVB) male hemizygotous mice were bred to wildtype (WT) FVB/N females to generate mixed litters. The pups were genotyped by PCR for the presence of human Atxn2 transgene and parallel experiments were performed with transgenic and wild type littermates. All mice were housed in a temperature-controlled room at 22-24°C with a 12hr light/dark cycle. Mice had access to standard chow and water ad libitum. All procedures were approved by the Institutional Animal Care and Use Committee (IACUC) of the UT Southwestern Medical Center at Dallas in accordance with the National Institutes of Health guidelines for the Care and Use of Experimental Animals.
Human embryonic kidney (HEK) 293 cell lines stably expressing human SK1, SK2 and SK3 proteins have been previously described (Hougaard et al., 2009). Cells were cultured in Dulbecco's Modified Eagle's Medium (DMEM, Gibco, Life Technologies, Nærum, Denmark) enriched with 10% fetal calf serum (FCS, Gibco) at 37°C and 5% CO 2. One day prior to electrophysiological experiments, the cells (approximately 75% confluency) were washed once with phosphate buffered saline (PBS), harvested by TrypLE™ Express (Gibco) treatment and transferred to petri dishes containing cover 3.5 mm diamter coverslips (VWR international, Herlev, Denmark).
SK-mediated membrane currents were recorded at room temperature using the inside-out configuration of the patch-clamp technique as previously described (Hougaard et al., 2007). Glass patch pipettes of 2 MΩ resistance were used in recordings using EPC-9 amplifier and Pulse software (HEKA, Lambrecht, Germany). In all experiments a solution with a high K+ concentration was applied to the extracellular side of the membrane (in mM): 154 KCl, 2 CaCl2, 1 MgCl2 and 10 HEPES, pH adjusted to 7.4 with 1 M KOH. The intracellular solutions contained (in mM): 154 KCl, 10 HEPES, 10 EGTA, or a combination of EGTA and NTA (10 mM in total). Concentrations of MgCl2 and CaCl2 required to obtain the desired free concentrations (Mg2+ always 1 mM, Ca2+ 0.01 – 10 μM) were calculated (EqCal, Cambridge, UK) and added. The intracellular solutions were adjusted to pH 7.2 with 1 M KOH. The currents were elicited by applying a 200 ms linear voltage ramp from -80 to +80 mV every 5 s from a holding potential of 0 mV.
The recordings of PC activity from rat cerebellar slices were performed essentially as described (Kaffashian et al., 2011). Briefly, Sprague Dawley rats (14-18 days old; Taconic, Ry, Denmark) were decapitated and brains rapidly dissected out into ice-cold artificial CSF (aCSF) of the following composition (in mM): 124 NaCl, 4 KCl, 8 MgSO4, 2.5 CaCl2, 1.25 NaH2PO4, 26 NaHCO3, 11 glucose, saturated with 95% O2/5% CO2. 300-μm parasagittal cerebellar slices were cut using a vibrating tissue slicer (VT1200 Leica, Ballerup, Denmark) and placed in a home-made holding chamber at room temperature in aCSF (composition like above except that MgSO4 was reduced to 1.2 mM), bubbled with 95% O2/ 5% CO2. Slices were left to recover for a minimum of 1 h prior to experiments. Individual slices were transferred to a submersion-style recording chamber (Luigs & Neumann Ratingen, Germany) and perfused at 2 ml/min with aCSF maintained at 30 °C using a feedback-controlled heater (Warner Instruments, Hamden, CT, USA). PCs for whole-cell current-clamp recordings were visualised at 40x using an up-right Olympus microscope (BX51WI) equipped with oblique illumination. Patch pipettes of resistance 3-7 MΩ were filled with the pipette solutions containing (in mM): 135 CH3KSO4, 10 KCl, 10 HEPES, 1 MgCl2, 2 Na2-ATP, 0.4 Na-GTP, pH 7.2 with 1 M KOH. Following high resistance seal formation the membrane was ruptured by suction and recordings were performed using an EPC-9 amplifier (HEKA, Lambrecht, Germany). Experimental control, data acquisition, and basic analyses were done with the Patchmaster (HEKA) software package.
Recordings of spontaneous PC activity from WT and 58Q mice at 24 weeks of age were perfomed as previously described (Kasumu et al., 2012). Briefly, the mice were anesthetized with a ketamine/xylazine cocktail and transcardially perfused with ice-cold aCSF containing (mM) 85 NaCl, 24 NaHCO3, 25 glucose, 2.5 KCl, 0.5 CaCl2, 4 MgCl2, 1 NaH2PO4, 75 sucrose. Solutions were equilibrated with 95% O2/5% CO2. Subsequently, the cerebellum was dissected and 300 μm thick sagittal slices were made with a VT1200S vibratome (Leica). Slices were allowed to recover in aCSF containing (in mM) 119 NaCl, 26 NaHCO3, 11 glucose, 2.5 KCl, 2.5 CaCl2, 1.3 MgCl2, 1 NaH2PO4 at 35°C for 30 minutes and then transferred to roo m temperature before recordings were made. The external bath used for recording was the same as the recovery aCSF in addition to containing 100 μM picrotoxin (PTX) and 10 μM 6,7-dinitroquinoxaline-2,3-dione (DNQX), equilibrated with 95% O2/5% CO2. All recordings were made within 5 hours after dissection. The recording chamber was heated to 34-35°C using P H1 heated holder (Warner Instruments, Hamden, CT). Loose-patch recordings were made according to (Hausser and Clark, 1997; Kasumu et al., 2012; Smith and Otis, 2003) to evaluate spontaneous activity of PCs. Briefly, 1-3 M glass pipettes were filled with the internal solution containing 140 mM NaCl buffered with 10 mM HEPES pH 7.3 and held at 0 mV. A loose patch (less than 100M ) configuration was established at the PC soma as close to the axon hillock as possible. Spontaneous action potential currents were recorded for 5-60 minutes from each cell using Axon Multiclamp 700B amplifier (Molecular Devices, Sunnyvale, CA). The five-minute recordings were analysed for tonic or burst firing as we previously described (Kasumu et al., 2012). Cells were characterized as firing tonically if they fired repetitive non-halting spike trains for 5 minutes. A cell was characterized as bursting if it had more than 5% of the interspike intervals that fell outside of 3 standard deviations from the mean of all interspike intervals in that cell. The analysis of instantaneous firing rates was performed using Clampfit 10.2 (Molecular Devices, Sunnyvale, CA). Data was plotted as the instantaneous firing rate every 2 seconds for the entire recording duration. From the plot of firing rates, bursting PCs were further categorized into 2 groups. Persistently bursting PCs were identified by a continuous presentation of bursts, each separated from the next by a period of silence (< 1 minute), throughout the duration of the recording. Transiently bursting PCs were identified by the presence of long periods of relatively constant tonic firing separated by short intermittent bursts. Once a burst firing pattern was observed during the first 5 minutes of recordings, the bath solution was switched to aCSF containing 5 μM NS309 or 5 μM CyPPA for at least 15 minutes to determine the effect of the compound on the firing pattern of that PC.
The drug feeding protocol was adapted from our previous study (Liu et al., 2009). NS13001 (30mg/kg) or CyPPA (10mg/kg) was suspended in the vehicle (0.5% HPMC-Corn flour suspension). The mice were fed orally 5 consecutive days (Monday to Friday) with 2 rest days (Saturday and Sunday) for 3 consecutive weeks starting at 9 months of age. Control groups of mice were fed with the vehicle (0.5% HPMC-Corn flour suspension) alone. Rotarod and beamwalk tasks were used to assess motor coordination as previously described (Kasumu et al., 2012; Liu et al., 2009). At baseline (prior to drug feeding), mice were trained on the beamwalk task to traverse 3 separate beams of differing diameters. A round plastic 17mm beam, a round plastic 11mm beam and a wooden square 5mm beam were used for training. Mice were given 3 consecutive training trials on 3 consecutive days on each beam. On the 3rd day, the mean latencies to traverse the entire length of the 11mm and 5mm beams were recorded and analysed for every animal in all 6 groups. After testing on the beamwalk task, mice were given a 3-day wait period and subsesquently trained on the accelerating rotarod task. Mice were trained to walk on a rotating rod accelerating at 0.2 rpm. Mice were trained for 4 consecutive days with 3 consecutive trials per day. The mean latency to fall off the accelerating rod was recorded and analysed for every animal in all 6 groups. After baseline testing, mice were fed for 3 consecutive weeks with the allotted compounds. Following drug feeding, mice were left alone for 3 days and then re-tested in motor tasks. Specifically, mice were trained on the beamwalk with 3 trials per beam on day 1 and tested on day 2. After a 3-day waiting period, mice were trained on the accelerating rotarod (Columbus instruments, Columbus, OH) with 3 consecutive trials on day 6 and tested on day 7.
Quantification of dark cell degeneration (DCD) status was performed as previously described (Kasumu and Bezprozvanny, 2010; Kasumu et al., 2012). Briefly, 5-6 mice in each group were sacrificed at 13 months of age. Mice were euthanized with pentobarbital and transcardially perfused with PBS followed by 2% paraformaldehyde/2% glutaraldehyde in 0.1M cacodylate buffer. The cerebellum was dissected out and cut into 1mm3 saggital sections and post-fixed in 1% osmium tetroxide. The specimen were subsequently stained en bloc with aqueous 1% uranyl acetate and lead citrate, dehydrated through a graded ethanol series, and embedded in EMbed 812 resin. Each cerebellum was cut into thinner 70 nanometer-thick sections and placed on copper grids, which were stained with aqueous 2% uranyl acetate and lead citrate. Sections from each animal were examined on a FEI Tecnai G2 Spirit Biotwin transmission electron microscope operated at 120 kV. Digital images were captured with a SIS Morada 11 megapixel side mount CCD camera. At least 5 mice were analysed per group with 2 grids made from different areas of the sections. PCs were judged to be in 1 of 3 stages- normal, moderate or severe. Normal PCs are spherical in shape and have regular alignment in the PC layer. The nucleus is also distinctly darker than the cytosol. Moderately degenerated PCs have slight shrinkage and moderately electron-dense cytosol that is almost as dark as nucleus. Severely degenerated PCs have markedly shrunken and electron-dense cytosol with similarly darkened nucleus. These PCs are usually not regularly aligned in the PC layer. The processing of samples for DCD analyses were performed by an investigator that was blind to genotype and treatment group. Quantification of DCD status of PCs was performed by an investigator that was also blind to mouse genotype and treatment group. The average percentage of normal, moderate and severely degenerated cells was calculated for each treatment group and plotted.
Differences between groups were judged by a two-tailed Student's unpaired t test using a significance level of 0.05.
We thank Leah Taylor for administrative assistance. Vibeke Meyland-Smith and Susanne Kalf Hansen are greatly acknowledged for their help with patch clamp experiments and preparation of cerebellar slices. Dr. Nicolas Andreotti is thanked for his help with the synthesis of LeiDab-7. A.W.K. is a Howard Hughes Medical Institute Med into Grad scholar. IB is a holder of the Carl J. and Hortense M. Thomsen Chair in Alzheimer's Disease Research. Funding: This study is supported by Neurosearch A/S (CP, RLCB, IB), by the NIH grants R01NS056224, R01NS38082 and R01NS074376 (IB) and by the contract with the Russian Ministry of Science 14.740.11.0924 (IB).
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.