Genetic Ablation of Inositol 1,4,5-Trisphosphate Receptor Type 2 (IP3R2) Fails to Modify Disease Progression in a Mouse Model of Spinocerebellar Ataxia Type 3

Spinocerebellar ataxia type 3 (SCA3) is a rare neurodegenerative disease caused by an abnormal polyglutamine expansion within the ataxin-3 protein (ATXN3). This leads to neurodegeneration of specific brain and spinal cord regions, resulting in a progressive loss of motor function. Despite neuronal death, non-neuronal cells, including astrocytes, are also involved in SCA3 pathogenesis. Astrogliosis is a common pathological feature in SCA3 patients and animal models of the disease. However, the contribution of astrocytes to SCA3 is not clearly defined. Inositol 1,4,5-trisphosphate receptor type 2 (IP3R2) is the predominant IP3R in mediating astrocyte somatic calcium signals, and genetically ablation of IP3R2 has been widely used to study astrocyte function. Here, we aimed to investigate the relevance of IP3R2 in the onset and progression of SCA3. For this, we tested whether IP3R2 depletion and the consecutive suppression of global astrocytic calcium signalling would lead to marked changes in the behavioral phenotype of a SCA3 mouse model, the CMVMJD135 transgenic line. This was achieved by crossing IP3R2 null mice with the CMVMJD135 mouse model and performing a longitudinal behavioral characterization of these mice using well-established motor-related function tests. Our results demonstrate that IP3R2 deletion in astrocytes does not modify SCA3 progression.


Introduction
Spinocerebellar ataxia type 3 (SCA3), also known as Machado-Joseph disease, is a rare autosomal dominantly inherited neurodegenerative disease [1,2] caused by an abnormal polyglutamine expansion within the ataxin-3 (ATXN3) protein [3,4]. This genetic alteration leads to slow neuronal degeneration in specific brain regions and the spinal cord, resulting in a wide variety of clinical manifestations, mainly motor-related [5][6][7]. Despite the welldescribed presence of neuronal death, other non-neuronal cells are involved in SCA3 pathogenesis [5,[8][9][10][11][12]. Astrogliosis is a common pathological feature in SCA3 that has also been found both in humans [5,10,11,13] and in animal models [14] of the disease, suggesting an important role of astrocytes in disease pathogenesis. Astrogliosis has been considered a reaction to neuronal damage; however, studies have shown that astrocytes are activated calcium signalling by genetically ablating IP 3 R2 had a beneficial effect on neuronal protection and motor deficits after stroke [56]. On the contrary, the ablation of IP 3 R2 in the context of amyotrophic lateral sclerosis was detrimental to SOD1 G93A mice, increasing innate immunity that contributed to the significantly shorter lifespan of these animals [57]. These contradictory results have generated controversy about the functional significance of calcium signalling by IP 3 R2 in astrocytes, since it has been assumed that calcium signalling is abolished in IP 3 R2 KO mice [43,58,59]. However, it was later demonstrated that there is calcium signalling generated by other sources in IP 3 R2 KO astrocytes [28], which may compensate to some extent for the absence of IP 3 R2.
To the best of our knowledge, IP 3 R2 has not been linked to SCA3 pathogenesis so far, despite evidence suggesting the involvement of astrocytes and the calcium signalling pathway in SCA3 pathogenesis. In the current study, we aimed to investigate the relevance of IP 3 R2 in the onset and progression of SCA3. Here, we investigated whether IP 3 R2 depletion and the consecutive suppression of global astrocytic calcium signalling would lead to marked changes in the behavioral phenotype of CMVMJD135 mice. Therefore, we crossed the CMVMJD135 mouse model [14] with the IP 3 R2 KO mouse model [42] and performed an extensive longitudinal characterization of the motor phenotype of these mice using a panel of well-established behavioral tests in order to determine the effect of IP 3 R2 ablation in the onset and disease presentation of SCA3.

IP 3 R2 Expression in the CMVMJD135 Mouse Model
To investigate the relevance of IP 3 R2 in SCA3, we first assessed whether Itpr2 gene expression was altered in the Q135 mice as well as to confirm IP 3 R2 deletion in the double mutant animals (described here as IP 3 R2 KO; Q135). We showed that the mRNA levels of Itpr2 are not altered in the brainstem of the Q135 mouse model, a well-known affected brain region in SCA3 (Figure 1a). Importantly, and as expected, IP 3 R2 is not present in the IP 3 R2 KO; Q135 animals as levels of Itpr2 mRNA in these mice were not detected ( Figure 1a). To rule out the possibility that deletion of IP 3 R2 could induce compensatory expression of other IP 3 Rs, we assessed the expression levels of the IP 3 R1 isoform. No differences were found (Figure 1b), suggesting that no adaptive compensations occurred in this constitutional KO strain, as we previously showed for the IP 3 R2 KO model [51].

The Double Mutant Mice Show Progressive Neurological Deficits Similar to CMVMJD135 Animals
Next, we performed a battery of well-established motor behavioral tests to obtain a full characterization of the double mutant to determine the effect of IP 3 R2 ablation in the disease presentation. In this longitudinal behavioral characterization, the animals were tested on several behavioral paradigms at 6,8,12,16,20,24, and 30 weeks of age (Figure 2a), corresponding to early, medium, and advanced disease stages, as previously described for Q135 mice [14]. The analysis of CAG length variation revealed that the Q135 and IP 3 R2 KO; Q135 mice carried a similar number of CAG repeats (Figure 2b), excluding an influence of CAG length variation on the behavior of the different test groups. All mice gained weight at a similar rate until 12 weeks of age ( Figure 2c); however, the Q135 and IP 3 R2 KO; Q135 mice stopped gaining weight throughout the age groups, showing a significantly lower body weight gain compared with their WT-littermates at 20, 24, and 30 weeks of age. In contrast, the control animals continued to gain weight until 30 weeks of age ( Figure 2c). The IP 3 R2 KO; Q135 mice displayed a lower body weight over time when compared to the Q135 mice (Figure 2c). brain region in SCA3 (Figure 1a). Importantly, and as expected, IP3R2 is not present in the IP3R2 KO; Q135 animals as levels of Itpr2 mRNA in these mice were not detected ( Figure  1a). To rule out the possibility that deletion of IP3R2 could induce compensatory expression of other IP3Rs, we assessed the expression levels of the IP3R1 isoform. No differences were found (Figure 1b), suggesting that no adaptive compensations occurred in this constitutional KO strain, as we previously showed for the IP3R2 KO model [51].  In the motor swimming test, no significant differences were found between the Q135 and IP 3 R2 KO; Q135 mice in their swimming performance (Figure 3a), with their latency to traverse the water tank being significantly worse than that of the WT mice during aging. As the disease progresses, the Q135 mice have increasing difficulty in maintaining balance [14]. Both the Q135 and IP 3 R2 KO; Q135 mice showed similar performance on the balance beams (Figure 3b-d).
Next, we evaluated whether the IP 3 R2 KO; Q135 mice showed any sign of movement initiation deficits, seen as a measure of parkinsonism [60,61] a clinical feature of some SCA3 patients, with the adhesive removal test (Figure 4a). Both the Q135 and IP 3 R2 KO; Q135 animals showed an increased latency to remove the nose sticker when compared to their WT-littermates, suggesting movement initiation deficits for both genotypes. Again, no differences were seen in this test between the IP 3 R2 KO; Q135 and Q135 mice (Figure 4a).
Loss of muscular strength is a very early and severe symptom observed in Q135 mice, with it already being observed when the animals were 6 weeks of age [14]. Thus, forelimb and hindlimb strength were also evaluated. In the hanging wire grid (Figure 4b), the Q135 mice showed a significantly lower latency to fall from the grid when compared to their WT-littermates, suggesting that limb muscular strength is diminished in these animals. The IP 3 R2 KO; Q135 mice presented a similar phenotype as the Q135 mice. Accordingly, both the Q135 and IP 3 R2 KO; Q135 mice displayed lower forelimb strength measured by the wire maneuver test (Figure 4c). and IP3R2 KO; Q135 mice carried a similar number of CAG repeats (Figure 2b), excluding an influence of CAG length variation on the behavior of the different test groups. All mice gained weight at a similar rate until 12 weeks of age ( Figure 2c); however, the Q135 and IP3R2 KO; Q135 mice stopped gaining weight throughout the age groups, showing a significantly lower body weight gain compared with their WT-littermates at 20, 24, and 30 weeks of age. In contrast, the control animals continued to gain weight until 30 weeks of age ( Figure 2c). The IP3R2 KO; Q135 mice displayed a lower body weight over time when compared to the Q135 mice (Figure 2c).  Assessment of body weight with age. An independent t-test, repeated measures ANOVA, and oneway ANOVA were carried out. All statistical information is included in Appendix A, Table A1. The data are presented as the mean SEM or as a percentage of animals (%). Asterisks (*) indicate statistical significance between the WT and Q135 mice, while hashtags (##) indicate statistical significance between the Q135 and IP3R2 KO; Q135 mice. The means were considered statistically significant at a p-value * p < 0.05.
In the motor swimming test, no significant differences were found between the Q135 and IP3R2 KO; Q135 mice in their swimming performance (Figure 3a), with their latency to traverse the water tank being significantly worse than that of the WT mice during aging. As the disease progresses, the Q135 mice have increasing difficulty in maintaining balance [14]. Both the Q135 and IP3R2 KO; Q135 mice showed similar performance on the balance beams (Figure 3b-d). (c) Assessment of body weight with age. An independent t-test, repeated measures ANOVA, and oneway ANOVA were carried out. All statistical information is included in Appendix A, Table A1. The data are presented as the mean SEM or as a percentage of animals (%). Asterisks (*) indicate statistical significance between the WT and Q135 mice, while hashtags (##) indicate statistical significance between the Q135 and IP 3 R2 KO; Q135 mice. The means were considered statistically significant at a p-value * p < 0.05. tical significance between the WT and Q135 mice, while hashtags (##) indicate statistical significance between the Q135 and IP3R2 KO; Q135 mice. The means were considered statistically significant at a p-value * p < 0.05.
In the motor swimming test, no significant differences were found between the Q135 and IP3R2 KO; Q135 mice in their swimming performance (Figure 3a), with their latency to traverse the water tank being significantly worse than that of the WT mice during aging. As the disease progresses, the Q135 mice have increasing difficulty in maintaining balance [14]. Both the Q135 and IP3R2 KO; Q135 mice showed similar performance on the balance beams (Figure 3b  In the motor swimming test, the IP3R2 KO; Q135 animals' performance was similar when compared to the Q135 mice. The balance beam test performance of the WT, Q135, and IP3R2 KO; Q135 mice using the 12 mm square beam (b,c) and the 11mm circle beam (d). Significant differences were observed in the beam walk test between the WT and the Q135 animals. However, the IP3R2 KO; Q135 mice performed similarly to the Q135 mice. Repeated measures ANOVA, one-way ANOVA, and the Kruskal-Wallis test were carried out. All statistical information is included in Appendix A, Table A1. The data are presented as the mean SEM or as a percentage of animals (%). Asterisks (*) indicate statistical significance between the WT and Q135 mice. The means were considered statistically significant at a p-value * p < 0.05, ** p < 0.01, and *** p < 0.001. In the motor swimming test, the IP 3 R2 KO; Q135 animals' performance was similar when compared to the Q135 mice. The balance beam test performance of the WT, Q135, and IP3R2 KO; Q135 mice using the 12 mm square beam (b,c) and the 11 mm circle beam (d). Significant differences were observed in the beam walk test between the WT and the Q135 animals. However, the IP 3 R2 KO; Q135 mice performed similarly to the Q135 mice. Repeated measures ANOVA, one-way ANOVA, and the Kruskal-Wallis test were carried out. All statistical information is included in Appendix A, Table A1. The data are presented as the mean SEM or as a percentage of animals (%). Asterisks (*) indicate statistical significance between the WT and Q135 mice. The means were considered statistically significant at a p-value * p < 0.05, ** p < 0.01, and *** p < 0.001. and hindlimb strength were also evaluated. In the hanging wire grid (Figure 4b), the Q135 mice showed a significantly lower latency to fall from the grid when compared to their WT-littermates, suggesting that limb muscular strength is diminished in these animals. The IP3R2 KO; Q135 mice presented a similar phenotype as the Q135 mice. Accordingly, both the Q135 and IP3R2 KO; Q135 mice displayed lower forelimb strength measured by the wire maneuver test (Figure 4c).  hanging wire grid and the (c) wire maneuver test. Assessment of limb strength showed no significant differences between the Q135 and IP3R2 KO; Q135 mice. Repeated measures ANOVA, one-way ANOVA, and the Kruskal-Wallis test were carried out. All statistical information is included in Appendix A, Table A1. The data are presented as the mean SEM. Asterisks (*) indicate statistical significance between the WT and Q135 mice.The means were considered statistically significant at a p-value * p < 0.05 and ** p < 0.01, *** p < 0.001.

Discussion
In this work, we explored the contribution of IP3R2, the main IP3 receptor responsible for somatic calcium signalling in astrocytes, to the progression of motor symptoms in an animal model of SCA3. Our results show that ablation of IP3R2 had no major impact on the motor-related symptoms of a SCA3 animal model, as the IP3R2 KO; Q135 mice displayed a similar motor phenotype when compared to the CMVMJD135 mouse model.
In the brainstem of the Q135 mice, a key affected brain region, the baseline mRNA levels of Itpr2 were not altered when compared to the WT mice. The expression levels of In the adhesive removal test, the IP 3 R2 KO; Q135 and Q135 mice showed worsening of their performance over time. Evaluation of limb muscular strength assessed by the (b) hanging wire grid and the (c) wire maneuver test. Assessment of limb strength showed no significant differences between the Q135 and IP 3 R2 KO; Q135 mice. Repeated measures ANOVA, one-way ANOVA, and the Kruskal-Wallis test were carried out. All statistical information is included in Appendix A, Table A1. The data are presented as the mean SEM. Asterisks (*) indicate statistical significance between the WT and Q135 mice.The means were considered statistically significant at a p-value * p < 0.05 and ** p < 0.01, *** p < 0.001.

Discussion
In this work, we explored the contribution of IP 3 R2, the main IP 3 receptor responsible for somatic calcium signalling in astrocytes, to the progression of motor symptoms in an animal model of SCA3. Our results show that ablation of IP 3 R2 had no major impact on the motor-related symptoms of a SCA3 animal model, as the IP 3 R2 KO; Q135 mice displayed a similar motor phenotype when compared to the CMVMJD135 mouse model.
In the brainstem of the Q135 mice, a key affected brain region, the baseline mRNA levels of Itpr2 were not altered when compared to the WT mice. The expression levels of the homologous isoform (IP 3 R1) were also analyzed to understand whether the deletion of IP 3 R2 could affect other IP 3 Rs expression changes through a direct compensatory mechanism or through interactions between IP 3 R2 and other targets. Our results excluded such compensatory changes.
We confirmed previous observations that Q135 mice display significantly lower body weight gain compared to WT mice and a decline in body weight as the disease progresses. The decline in body weight as the disease progress is most likely associated with the significant atrophy observed in the model [14]. Interestingly, over time, the IP 3 R2 KO; Q135 animals displayed a lower body weight when compared to the Q135 mice. This suggests that IP 3 R2 can be a modulator of this more "peripheral" aspect of the phenotype. IP 3 R2 KO mice have been reported by us [51] and others [47,56] to have a similar body weight when compared to WT mice; however, in the context of SCA3, the absence of IP 3 R2 in the IP 3 R2 KO; Q135 animals appears to have induced the loss of body weight. This suggests that IP 3 R2 could have a detrimental effect on peripheral tissues, and further studies are needed to clarify the functional role of IP 3 R2 in SCA3 mice. Studies with calcium imaging and/or experiments of knock-down of IP 3 Rs in the peripheral tissues [62,63] could help us to elucidate this point and its relevance for SCA3.
Using tests to evaluate core motor deficits, such as those affecting movement, coordination and balance, our data showed that, unlike what was seen for SCA2 [41] the genetic deletion of IP 3 R2 does not alter the disease progression of Q135 animals. Similar results have been described in amyotrophic lateral sclerosis (ALS) mouse models, where the genetic ablation of IP 3 R2 showed no impact on disease onset and motor coordination although worsened muscular strength and decreased survival of the SOD1 G23A mice [57]. At first glance, the involvement of IP 3 R2 in ALS seems controversial, as higher levels of Itpr2 were shown to be detrimental to the mice and therefore, it would be expected that the ablation of this isoform could have beneficial effects [64]. Contrary to our observations and those made in ALS mice, ablation of this receptor in the ischemic brain as well as in aged mouse brains was shown to be neuroprotective, while reducing behavioral deficits [56]. These observations may indicate a distinct role of astrocytes in different contexts and disease aetiologias. The physiological role of IP 3 R2-mediated calcium signalling needs to be studied in different pathological contexts since the absence of IP 3 R2 leads to different outcomes according to the context.
We also performed a battery of tests to evaluate muscular strength and movement initiation, a parkinsonism-related manifestation. In these behavioral dimensions, the performance of the IP 3 R2 KO; Q135 mice was also indistinguishable from that of the Q135 mice. Overall, these data suggest that the deletion of IP 3 R2 had no impact on the motor activity of the IP 3 R2 KO; Q135 mice. Despite the likely suppression of global calcium signalling in astrocytes, the IP 3 R2 KO mice still have IP 3 R2-independent calcium signals through alternative calcium sources and synaptic calcium, such as plasmalemma calcium influx or the mitochondria [27,28,65]. Based on this, it is plausible to question whether the transient calcium signals are robust enough to justify the lack of impact of IP 3 R2 deletion in the SCA3 background. Additionally, the possibility of a compensatory mechanism between IP 3 R2 and other critical players of calcium signalling exists, including G-proteincoupled receptors. Further studies should be performed to unravel the involvement of other calcium signalling pathways in SCA3 onset and progression, including approaches using pharmacological approaches [63,[66][67][68] or the two-photon excitation microscopy technique [69,70].
Globally, our results demonstrate that SCA3-IP 3 R2 double mutants present similar behavior to SCA3 mice, suggesting that the IP 3 R2 receptor is not a major modulator of the onset and progression of SCA3 motor symptomatology.

Animal Generation and Maintenance
In this study, two mouse models (Mus musculus and strain C57BL/6J) were used to understand the potential involvement of IP 3 R2 receptor deletion in SCA3 progression.
The Q135 mice were generated as previously described [14]. The CMVMJD135 mouse model expresses human ATXN3 (cDNA, isoform 3c) under the control of the CMV promoter (ubiquitous expression) at near-endogenous levels [14]. The cDNA variant of the ATXN3 gene carries a repeat tract with the sequence (CAG)2CAAAAGCAGCAA(CAG)129, coding for 135 glutamine [14,71]. The IP 3 R2 knock-out (KO) mouse model was supplied by Prof. Alfonso Araque (Minneapolis, MN, USA) [72] under agreement with Prof. Ju Chen (U.C. San Diego, CA, USA) [42]. Initially, the heterozygous Q135 mice were crossed with homozygous IP 3 R2 KO mice, and the following genotypes were obtained: wild-type (WT), heterozygous Q135 homozygous IP 3 R2 KO (IP 3 R2 −/−), and heterozygous IP 3 R2 KO (IP 3 R2 +/−) mice, as well as a double mutant of heterozygous Q135 and heterozygous IP 3 R2 KO (IP 3 R2 +/−; Q135). To generate the experimental groups, the obtained double mutants (IP 3 R2 +/−; Q135) were crossed with the heterozygous IP 3 R2 KO (IP 3 R2 +/−), obtaining the following genotypes WT, Q135, IP 3 R2 −/−, IP 3 R2 +/−, and IP 3 R2 −/−; Q135. To answer our main question, although we performed the behavior analysis with all the obtained genotypes, we focused on the comparison between the WT, Q135, and IP 3 R2 −/−; Q135 mice. Male mice were used in this study to increase analysis homogeneity (decrease variance) and to allow for a small number of animals. Because the IP 3 R2 heterozygous animals had not been characterized before, we included this group of mice in our analyses. No differences were found in any of the motor-related tests performed when compared to the WT-littermates (see the statistical analyses in Appendix A, Table A1).
The animals were maintained in a conventional animal facility and under standard laboratory conditions, which included an artificial 12 h light/dark cycle, lights on from 8:00 am to 8:00 pm, an ambient temperature of 21 ± 1 • C, and a relative humidity of 50-60%. The mice were given a standard 4RF25 diet during the gestation and postnatal periods and a 4RF21 diet after weaning (3 weeks of age; Mucedola SRL, Settimo Milanese, Italia) as well as water ad libitum. During weaning, the mice were housed in groups of six animals in filter topped polysulfone cages 267 × 207 × 140 mm (370 cm 2 floor area) (Tecniplast, Buguggiate, VA, Italy) using corncob bedding (Scobis Due, Mucedola SRL). Environmental enrichment consisted of soft tissue paper and shredded paper to stimulate the natural behavior of nesting.

Molecular Analysis: Macrodissection, RNA Isolation, cDNA Synthesis, and Real-Time Quantitative PCR Analysis
In order to assess the transcription levels of the Itpr2 gene, relative mRNA levels of the Itpr2 gene were quantified by real-time quantitative reverse-transcriptase polymerase chain reaction analysis (qRT-PCR). For this, the WT (n = 8), Q135 (n = 9), IP 3 R2 KO (n = 3), and IP 3 R2 KO; Q135 (n = 5) animals were euthanized, their brains were harvested, and their brainstems (a key affected region in SCA3) were macrodissected.
Total RNA was isolated from the tissues using TRIZOL (Invitrogen, Waltham, MA, USA) according to the manufacturer's protocol. First-strand complementary DNA (cDNA), synthesized using the iScript™ cDNA Synthesis Kit (Bio-Rad, Hercules, FL, USA), was amplified by qRT-PCR according to the guidelines (Bio-Rad, Hercules, FL, USA).
The primers used in this study (Table 1) were designed using PRIMER-BLAST (NCBI, Bethesda, Rockville, ML, USA; http://www.ncbi.nlm.nih.gov/tools/primer-blast/, accessed on 1 June 2019). Quantification was performed using the Fast Real-Time PCR System (Applied Biosystems, Waltham, MA, USA). The housekeeping beta-2-microglobulin (B2m) gene was used as an internal control. The relative gene expression was determined using the 2 −∆∆Ct relative quantification method and represented as fold change normalized to the mean of the relative expression of the wild-type mice. The PCR cycling conditions used were denaturation at 95 • C for 30 s, 30 cycles at 95 • C for 5 s, annealing at 60 • C for 30 s, plus an extension at 65 • C for 5 s, and a final extension at 95 • C for 5 s. Table 1. Primer sequences used for the analysis of gene expression (qRT-PCR).

Gene
Primer Sequence

Behavior Analysis
Behavioral tests were performed during the diurnal period, and the animals were tested at 6, 8, 12, 16, 20, 24, and 30 weeks of age. Body weight was registered at all time points analyzed. Motor behavior was assessed using the balance beam walk test (12-mm square and 11-mm round beams) and the motor swimming test. Furthermore, movement initiation was evaluated with the adhesive removal test, and muscular strength parameters were assessed using the hanging wire grid and the wire maneuver tests. The behavioral tests are briefly explained below:

Motor Swimming Test
The motor swimming test was used to evaluate voluntary locomotion, by taking advantage of the survival instinct of mice in water environments [73]. Each mouse was gently placed at one end of a 100 cm clear acrylic tank filled with water (15 cm depth), maintained at 23 • C ± 1 • C using a thermostat, and trained for 2 consecutive days (3 trials/animal), to reach a visible black platform at the opposite end. During the following 3 days, the animals were tested, and the time that it took each mouse to cross the water tank was recorded (2 trials/mouse) [73]. The tank was labelled with a blue line to mark the starting position, and the mice swam a distance of 60 cm to complete the trial.

Beam Balance Test
Balance and fine motor coordination were evaluated by the latency of each mouse to traverse different diameters and shape beams to reach a safe platform [73]. The beams consisted of long strips of PVC (1 m) with a 12 mm square cross-section or 11 mm round diameter. The beams were placed horizontally, 50 cm above the bench surface, protected with soft sponges to protect the mice from falls, with one end mounted on a narrow support and the other end attached to an enclosed dark box (20 cm square), into which the mouse could escape. The mice were trained for 3 consecutive days on the 12 mmsquared beam (3 trials/animal), and on the fourth day, the animals were tested on both beams (2 trials/animal). The time each mouse took to cross the beams was recorded and discounted if the animal stopped in the beam. A failed trial was considered when an animal fell or turned around on the beam. Each animal had the opportunity to fail twice on each beam [73]. Because the Q135 mice showed a worsening of the phenotype at 16 weeks of age, which affects the ability to perform the task causing them to fall off the beams very frequently, we analyzed the data by attributing performance scores to the animals as follows: 0-performed 2 trials, 1-performed 1 trial, and 2-performed 0 trials, meaning that it cannot walk on the beam.

Adhesive Removal Test
In the adhesive removal test, the experimenter gently placed a round adhesive (8 mm diameter) on the nose of the mouse and transferred the animal to a cage. The time the animal took to remove the adhesive was registered [61,74].

Hanging Wire Grid Test
In the hanging wire grid test, the mice were placed on top of a metallic horizontal grid and inverted 180 • towards the surface of the bench (protected with soft sponges to protect the mice from falls). With this test, the latency for the mice to fall off the metal grid was evaluated. The maximum time of the test allowed was 120 s [73,75].

Wire Maneuver Test
The mice were picked up by their tails, and their forelimbs were placed to a fixed wire, underneath of which soft sponges were placed to protect them from falls. This test is based on the latency to fall off the metal wire and the maximum allowed time of the test was 120 s [75].

Statistical Analysis
The G*Power 3.1.9.2 software was used to calculate the required sample size, based on a power of 0.8 (obtained from previous data) [76] and a significance level of 0.05 was used for all statistical tests.
All statistical analyses were performed using SPSS 22.0. The statistical analysis of the behavior tests was performed including all the genotypes under study (including IP 3 R2 −/+); however, the results presented in this study were divided into two groups to simplify the visualizations of the key comparisons as well as to answer the study's main question. Behavioral data were analyzed by a repeated-measures ANOVA when the variables were continuous or presented a normal distribution. Values that deviated more than 1.5 interquartile ranges from the mean were considered outliers and excluded from further analyses. The assumption of normality was assessed by qualitative analysis of Q-Q plots and frequency distributions (z-score of skewness and kurtosis) as well as by the Kolmogorov-Smirnov and Shapiro-Wilk tests. The assumption of homogeneity of variances was evaluated by Levene's test. Regarding repeated measurements, sphericity was tested using Mauchly's test and assumed for all tested variables. For the comparison of means between 5 groups, one-way analysis of variance (ANOVA) was used, followed by Tukey HSD or Dunnett T3's test (when data passed on the assumption of homogeneity of variances or when the populations variances were not equal, respectively).
Regarding non-normally distributed data and/or for the comparison of medians of discrete variables across time points, a Friedman's ANOVA was carried out, with pairwise comparisons through the Kruskal-Wallis statistical test.
Effect size measurements are reported for all analyses (Cohen's d for t-tests and eta partial square-n p 2 for ANOVAs). GraphPad Prism 8 was used to create graphs. All figures were created using PowerPoint (version 2305, Microsoft 365 MSO). All statistical information is reported in Appendix A.