Changes in Calcium Homeostasis and Gene Expression Implicated in Epilepsy in Hippocampi of Mice Overexpressing ORAI1

Previously, we showed that the overexpression of ORAI1 calcium channel in neurons of murine brain led to spontaneous occurrence of seizure-like events in aged animals of transgenic line FVB/NJ-Tg(ORAI1)Ibd (Nencki Institute of Experimental Biology). We aimed to identify the mechanism that is responsible for this phenomenon. Using a modified Ca2+-addback assay in the CA1 region of acute hippocampal slices and FURA-2 acetomethyl ester (AM) Ca2+ indicator, we found that overexpression of ORAI1 in neurons led to altered Ca2+ response. Next, by RNA sequencing (RNAseq) we identified a set of genes, whose expression was changed in our transgenic animals. These data were validated using customized real-time PCR assays and digital droplet PCR (ddPCR) ddPCR. Using real-time PCR, up-regulation of hairy and enhancer of split-5 (Hes-5) gene and down-regulation of aristaless related homeobox (Arx), doublecortin-like kinase 1 (Dclk1), and cyclin-dependent kinase-like 5 (Cdkl5, also known as serine/threonine kinase 9 (Stk9)) genes were found. Digital droplet PCR (ddPCR) analysis revealed down-regulation of Arx. In humans, ARX, DCLK1, and CDLK5 were shown to be mutated in some rare epilepsy-associated disorders. We conclude that the occurrence of seizure-like events in aged mice overexpressing ORAI1 might be due to the down-regulation of Arx, and possibly of Cdkl5 and Dclk1 genes.


Introduction
Store operated calcium entry (SOCE) is a well-described process in non-excitable cells. It was initially known under the name "capacitative Ca 2+ entry" [1] and has recently been identified also in neurons and called as nSOCE [2][3][4][5]. In the 1980s, several groups observed a phenomenon that decreasing intracellular [Ca 2+ ] activates a transmembrane Ca 2+ current [6][7][8]. During the next two decades of studies, the scientific community established molecular basis of SOCE (reviewed in [9]). Reduction in [Ca 2+ ] ER results in activation of plasma membrane Ca 2+ channels that mediate sustained Ca 2+ influx, which is required for a variety of processes. They include maintaining Ca 2+ oscillations [10], mitochondrial fatty acid oxidation [11], refilling of Ca 2+ stores, and a few other [12]. The essential components of SOCE are stromal interaction molecules (STIM1 and STIM2), which function as sensors of the Ca 2+ concentration inside the endoplasmic reticulum (ER) [13,14] and ORAIs (ORAI1-3), which are highly selective Ca 2+ channels at the plasma membrane (PM) [15,16]. Upon store depletion, STIMs form multimers, which translocate within the endoplasmic reticulum (ER) membrane towards ER-PM junctions, where they activate highly Ca 2+ -selective Ca 2+ -release-activated Ca 2+ channels (CRACs) -ORAIs [17]. CRAC channels function as hexamer complexes that are composed of potentially multiple ORAI homologues [18,19]. Heteromeric channels that are built of endogenous ORAI1 and ORAI2 homologues were found in murine T-cells [20] and in mast cells [21], whereas in the enamel cells [22], dorsal root ganglion (DRG) neurons [23] and in overexpression studies, ORAI1:ORAI3 multimers were observed [24]. In murine immune cells, deletion of Orai1 reduces, whereas deletion of Orai2 increases SOCE, which indicates that ORAI2 is a less conductive subunit of CRAC channels and enables shaping of the magnitude of SOCE and modulation of the immune response [20]. Eckstein et al. postulated that ORAI1 and ORAI2 were the main channel subunits of CRAC channels in enamel cells, noting that ORAI3 can act as an inhibitor of ORAI1 in these cells [22]. In contrast, knockdown of both Orai1 and Orai3 in DRG neurons reduced SOCE compared with Orai1 or Orai3 knockdown alone [23]. These manipulations resulted in altered neuronal excitability [23]. All of these studies have shown that homoand heteromeric CRAC channels demonstrate different SOCE response, which indicates the importance of the stoichiometry of different ORAI homologues that form the CRAC channel. The expression profile of all three homologues of ORAI varies in different tissues, with the highest level of ORAI1 in immune cells [25] and ORAI2 in the brain [26], whereas ORAI3 is abundant in many solid organs and its enhanced gene expression is reported in cancerous cells [27]. Physiologically, the best-characterized homolog is ORAI1 [28,29]. Patients with inherited null or loss of function mutations in ORAI1 suffer from severe combined immunodeficiency-like disease that is accompanied with chronic, often lethal infections and a variety of non-immunological symptoms [30,31].
Our recently published results indicate that overexpression of ORAI1 in the murine brain leads to spontaneous occurrence of seizure-like events in aged animals [32]. In the present study, we extended the phenotyping of the established transgenic line overexpressing ORAI1 in neurons (FVB/NJ-Tg(ORAI1)Ibd; Nencki Institute of Experimental Biology) by the analysis of Ca 2+ response in the hippocampal CA1 pyramidal cells and RNA sequencing (RNAseq) gene expression profiling. Differential expression of selected genes was next confirmed by real-time PCR and validated using the digital droplet PCR (ddPCR) approach. The observed changes in Ca 2+ signals after glutamate treatment or after ER Ca 2+ store depletion suggest a new potential site of action for ORAI1. Our observations were supported by the identification of a set of genes, whose expression was changed in FVB/NJ-Tg(ORAI1)Ibd mice. For some of these genes, such as cyclin-dependent kinase-like 5 Cdkl5, aristaless related homeobox (Arx), and doublecortin-like kinase 1 (Dclk1), mutations in their human homologoues that lead to rare epilepsy-associated disorders were reported. Together, our results provided novel insight into the Ca 2+ dependent actions evoked by CRAC channels that might underlie the occurrence of seizure-like episodes in FVB/NJ-Tg(ORAI1)Ibd mice.

Overexpression of ORAI1 in Neurons Leads to Altered Ca 2+ Response in a Modified Ca 2+ Addback Assay in CA1 Hippocampal Region
Acute brain slices of hippocampi of wild-type and transgenic ORAI1 mice were analyzed by fluorescent imaging. Calcium measurements were performed using Fura-2 acetomethyl ester (AM) indicator that was loaded into the cells by gradual pipetting over the slice with the use of an automated perfusion system. To assess Ca 2+ homeostasis in the ORAI1 overexpressing neurons, we used a Ca 2+ addback assay that was modified to enable the measurement of Ca 2+ release from the ER [33]. According to suggestions that were proposed by Amrhein et al. (2019), the results are presented as raw p-values [34]. Also, the asterisks that are commonly used to signify a particular confidence interval are not shown. As shown in Figure 1A, the basal level of Ca 2+ was slightly higher in neurons of transgenic ORAI1 mice. Stimulation by glutamate led to comparable peaks in all genetic variants tested ( Figure 1B,F). However, the time-course of signal decay following the brief pulse of glutamate tended to be slower in ORAI1 overexpressing neurons ( Figure 1D). The resulting difference in Ca 2+ levels persisted during the subsequent application of ethylene glycol-bis(β-aminoethyl ether)-N,N,N ,N -tetraacetic acid (EGTA) ( Figure 1A,C). These changes were absent in neurons with overexpression of STIM2 ( Figure 1E,G). This may suggest that neurons of ORAI1 transgenic mice exhibit changes in the mechanisms that are responsible for extrusion of Ca 2+ from the cytosol by plasma membrane calcium ATP-ase (PMCA) or Ca 2+ uptake into the ER. Moreover, a tendency towards the increased influx of Ca 2+ ions upon Ca 2+ addback was observed both in ORAI1 and STIM2 overexpressing neurons. Ca 2+ measurements were performed using Fura-2 acetomethyl ester (Fura-2 AM) indicator that was loaded into the CA1 pyramidal neurons of the hippocampal acute brain slices (H). Typically, 20 pyramidal neurons per one slice (n) were analyzed; the slices were isolated from at least 5 animals per genetic variant. The total number of slices analyzed per each genetic variant was 9. (A,E) Averaged time-course of background-subtracted fluorescence signal (expressed as F340/F380) from slices overexpressing ORAI1 and STIM2, respectively. (B,F) Quantification of signal amplitudes observed following glutamate, cyclopiazonic acid (CPA), and Ca 2+ (re)addition in ORAI1 and STIM2 overexpressing neurons, respectively, arbitrary units (a.u.). (C,G) Quantification of F340/F380 values at baseline (average of 0-5th min of the measurement) and following application of ethylene glycol-bis(β-aminoethyl ether)-N,N,N ,N -tetraacetic acid (EGTA) (at 13th and 25th min of the measurement) in ORAI1 and STIM2 overexpressing neurons, respectively, a.u. (D) Time-course of signal decay following glutamate application in ORAI1 and wild-type neurons that was fit by a logarithmic function. Student's t-test was used to check statistical significance of the observed differences; p-values are displayed above the respective charts.

RNAseq Analysis of Transcripts from Hippocampi Indicates Up-Regulation of the Machinery Involved in Ribosome Biogenesis and Ribonucleoprotein Complex Assembly
We next isolated total RNA from hippocampi of wild-type (n = 3, age 7 months) and of FVB/NJ-Tg(ORAI1)Ibd (n = 3, age 7 months) females and performed RNAseq experiments. Based on the collected RNAseq data, we performed functional (pathways, gene groups) analysis of the genes that were differentially expressed between the hippocampi of wild-type and of FVB/NJ-Tg(Orai1)Ibd females. As shown by the Gene Ontology Biological Process groups (GO BP) (Figure 2A, left), the most down-regulated genes were those belonging to hormone transport group and related with import across plasma membrane and ion transport. Most of the up-regulated genes belonged to the ribonucleprotein complex biogenesis, ribosome biogenesis, and cytoplasmic translation (Figure 2A, right). To extract the complex association between the identified GO BP groups we used the category netplot function (cnetplot). The cnetplot depicts the linkages of genes and biological concepts (GO terms) as a network, which allows to visualize the genes that are involved in the enriched pathways and genes that may belong to multiple annotation categories ( Figure S1). Among the up-regulated genes, Orai1 was present, which confirms the identity of transgenic mice ( Figure 2C). The lists of genes that were identified among the particular GO BP groups and that are present in dot plots in Figure 2A are listed in data Supplementary data (Tables S1 and S2). The expression of selected genes, some of which are indicated in the volcano plot ( Figure 2B), was further validated by the real-time PCR and ddPCR methods.  We next performed real-time PCR analysis to verify the RNAseq data using customized arrays designed for the 23 genes that are listed in Table 1. Table 1. A list of selected genes that were chosen for validation using customized arrays. Arx, Dclk1, Cdkl5 and were subsequently taken to ddPCR analysis (framed in green). Table 1. A list of selected genes that were chosen for validation using customized arrays. Arx, Dclk1, Cdkl5 and were subsequently taken to ddPCR analysis (framed in green). As the reference genes, Uba-2 (catalog no. qMmuCID0018845, Bio-Rad), Gapdh (catalog no. qMmuCED0027497, Bio-Rad), and Actin (catalog no. qMmuCED0027505, Bio-Rad) were selected, which meet the criteria of the reference gene selection tool (CFX software, Bio-Rad). RNA was isolated from hippocampi of wild-type (n = 3, age 5 months) and of FVB/NJ-Tg(ORAI1)Ibd females (n = 3, age 6 months). The majority of genes showed a similar level of expression ( Figure 3A and Table 1) that was calculated as 2 −∆∆Ct . The genes which were markedly down-regulated in the transgenic individuals included Arx (aristaless related homeobox) (2 −∆∆Ct = 0.78, p = 0.047), Dclk1 (Serine/threonine-protein kinase) (2 −∆∆Ct = 0.82, p = 0.016), and Cdkl5 (Cyclin-dependent kinase-like 5) (2 −∆∆Ct = 0.85, p = 0.02 when only Actin was used in the calculations as the reference gene). Human mutations in these genes are associated with epilepsy. Also Gabrb2 (Gamma-aminobutyric acid receptor subunit beta-2) and Scn9a (Sodium channel protein type 9 subunit alpha) genes, which are implicated in epilepsy, showed altered expression in the transgenic animals (2 −∆∆Ct = 0.87, p = 0.23; 2 −∆∆Ct = 1.26, p = 0.136, respectively). Because the observed changes in 2 −∆∆Ct values were relatively modest, to validate the changes in expression of the selected genes we used a novel technique, digital droplet PCR (ddPCR, Bio-Rad). This powerful method is based on partitioning of one PCR reaction mixture into up to 20,000 independent reactions, which resulted in high precision, sensitivity, and accuracy of the data [35]. To fulfill the requirements of validation, the RNA from hippocampi was isolated from the next group of animals (3 wild-type vs. 6 transgenic, age-matched). Using customized assays that were dedicated to ddPCR analysis and Uba-2 gene as a reference, we observed a reduction in the copy number of Arx  whereas green corresponds to transgenic ones. The obtained results suggest a down-regulation of expression of Arx, Cdkl5, and Dclk1. The data are presented as a fold change (2 −∆∆Ct ) normalized to the three reference genes (Uba-2, Gapdh, Actin). The statistics was estimated by CFX software (Bio-Rad) using the post hoc method (Tukey's test). (B) The ddPCR analysis of Arx, Cdkl5, Dclk1 using customized assays (Bio-Rad). The number of transcript copies of the selected genes are presented in relation to 1000 copies of the reference gene Uba-2. The results are presented as dot plot, where each sign corresponds to an individual animal (3 wild-type, 6 transgenic). The statistics were calculated by GraphPad Prism software using unpaired t-test. The number of Arx transcripts was about 30 times lower than that of Cdkl5 and Dclk1, therefore, it is presented on a separate chart.

Expression of Hes-5, a Transcriptional Repressor That Is Involved in Neurogenesis, Is Up-Regulated in Tg(ORAI1) Ibd Females
In the next set of the real-time PCR analysis based on RNAseq data, we designed an additional 15 pairs of primers (Table 2).
Using the matrixes that were derived from previous experiments (6 wild-type vs. 6 transgenic mice) and Uba-2 as a reference gene, we identified an up-regulation of Hes-5 (Transcription factor HES-5) expression (2 −∆∆Ct = 2,93, p = 0.0039) ( Figure 4). Among the genes that were analyzed and described in Table 2, we detected a trend towards the up-regulation of Strap (Serine-threonine kinase receptor-associated protein). The protein is a part of the survival of motor neurons (SMN) complex that acts as a catalyst in the assembly of small nuclear ribonucleoproteins (snRNPs), the building blocks of the spliceosome (2 −∆∆Ct = 1.56, p = 0.18). and Strap. The data are presented as a fold change (2 −∆∆Ct ) that was normalized to the reference gene (Uba-2), the statistics was estimated by CFX software (Bio-Rad) using post hoc method (Tukey's test). Table 2. List of selected genes together with primers sequence whose expression profile was analyzed using real-time PCR. Up-regulation of hairy and enhancer of split-5 (Hes-5), a transcriptional repressor, was detected in the hippocampus of Tg(ORA1)Ibd line (highlighted in green). Table 2. List of selected genes together with primers sequence whose expression profile was analyzed using real-time PCR. Up-regulation of hairy and enhancer of split-5 (Hes-5), a transcriptional repressor, was detected in the hippocampus of Tg(ORA1)Ibd line (highlighted in green).

Discussion
The role of SOCE in neurons and its components, such as STIM proteins and Ca 2+ channels, is far from being understood. It becomes clear that not only ORAI and transient receptor potential cation channels (TRPCs) [36], but also voltage-gated dependent channels (VGCC) [37] and α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptors [38,39] are sensitive to STIMs. Our group also demonstrated the involvement of STIM1 in mGlur-LTD phenomenon in CA3-CA1 projection of hippocampus in the novel FVB/NJ-Tg(STIM1)Ibd transgenic line [40]. On the other hand, it is now well established that local Ca 2+ microdomains that appear upon opening of store operated calcium channels activate downstream effectors like calcineurin or adenylyl cyclase in non-excitable cells. The wide scope of processes mediated by Ca 2+ nanodomains that are generated by SOCE is well described in a recent review by Barak and Parekh [9].
In hippocampal neurons nSOCE is believed to be mediated mainly by STIM2 together with Orai2 and Trpc6. It was suggested that Ca 2+ that is delivered by nSOCE is responsible for Ca 2+ /calmodulin-dependent protein kinase II (CaMKII) pathway activation, which contributes to spine maintenance of mushroom spines [41]. Ca 2+ disruption leads to synaptic loss and cognitive decline in aging and Alzheimer's disease [41][42][43]. The role of nSOCE seems to be even more complex as indicated by the recent results by Chen-Engerer et al. [26]. They demonstrated that in CA1 pyramidal cells nSOCE that is mediated by ORAI2 is essential for refilling of IP3 sensitive Ca 2+ stores, whereas ryanodine-sensitive stores are VGCC dependent. The authors postulate that their observations may help to explain the distinct roles of inositol triphosphate receptor (IP3Rs) and RyRs for various forms of synaptic plasticity.
As a consequence of ORAI1 transgene overexpression in mice used in our studies, we modified the level of ORAI1 in neurons, especially in the hippocampus, in which the predominant isoform is ORAI2. The observed trend of a higher amplitude of nSOCE in neurons of Tg(ORAI1)Ibd transgenic mice is in line with the observation by Vaeth et al. (2017), who showed that ORAI2 was a less conductive subunit of CRAC channel in immune cells [20]. This observation was supported by Fura-2 quenching experiments using Mn 2+ , in which we observed faster decay of the signal in ORAI1 overexpressing neurons ( Figure S2). Because Mn 2+ quenches the fluorescence of Fura-2 and store-operated channels are permeable to them, such approach can be used to assess the activity of SOCs. Moreover, a trend for different kinetics of the [Ca 2+ ] decay after glutamate treatment and the subsequent significant differences in the Ca 2+ level in chelating conditions that were described in this work are in agreement with the observations done by Chen-Engerer et al. (2019). They showed the importance of ORAI2-mediated nSOCE in refilling of IP3 sensitive Ca 2+ stores [26]. Due to the fact that glutamate stimulation also evokes IP3 signaling cascade that in turn leads to Ca 2+ release from the ER, subsequent ORAI-dependent refilling of ER Ca 2+ stores might be altered in neurons of our transgenic line [44]. These observations suggest that our transgenic line might be a useful tool for an in-depth analysis of specific molecular targets that are activated by nSOCE in the brain.
RNAseq analysis of the material from Tg(ORAI1)Ibd transgenic line identified up-regulation of the genes that belong to the GO biological processes that are involved in the ribonucleoprotein complex biogenesis, ribosome biogenesis, and assembly, which are all components of translational machinery. It was demonstrated in acinar polarized cells that overexpression of STIM1 increased the density of ribosome-free terminals in rough ER (RER), and facilitated the formation of ER-plasma membrane (ER-PM) junctions and SOCE propagation. If our future validation of these RNAseq data by real-time and ddPCR is positive, it may suggest that Ca 2+ microdomains that are generated by nSOCE serve as a regulatory component of the translation process [45,46].
As previously reported by our group, overexpression of ORAI1 in neurons induced seizure-like symptoms in aged female mice [32]. Application of kainic acid and bicuculline to hippocampal slices from Tg(ORAI1)Ibd females revealed a significantly lower frequency of interictal bursts compared with samples that were isolated from wild-type female or transgenic male mice. The down-regulation of Arx, Cdkl5, and Dclk1 genes that we observed in the present work is in line with the report that loss of function mutations of human aristaless related homeobox (ARX), CDKL5, and DCLK1 are found in epileptic patients [47] (Table 3). It might be speculated that the Ca 2+ nanodomains that appear locally by the activity of CRAC channels affect the expression of these genes via downstream pathways. In conclusion, our data indicate that overexpression of ORAI1 changed the subunit composition of CRAC channels, which caused differences in the observed Ca 2+ responses and induced specific changes in gene expression. By real-time PCR, we detected a down-regulation of Arx, Cdkl5, and Dclk1 genes, human homologues of which are involved in etiology of epilepsy. This down-regulation might explain seizure-like phenomena that were observed in 15-month-old females of the FVB/NJ-Tg(ORAI1)Ibd line.
The PCR product was digested with XhoI and subcloned into the XhoI site of the modified mouse Thy-1.2 expression cassette [66,67].

Brain Slice Preparation
The reagents were purchased from Sigma unless indicated otherwise. Hippocampal sections were prepared according to the protective recovery method [69]. Mice of the age P25 ± 5 were sacrificed by cervical dislocation. Next, brains were isolated and placed in a dish filled with N-methyl-D-glucamine (NMDG)-based artificial cerebrospinal fluid (NMDG-aCSF) solution that was cooled to~2 • C and contained, in mM: NMDG, 92, KCl, 2.5, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), 20, thiourea, 2, glucose, 25, NaH 2 PO 4 , 1.25, NaHCO 3 , 30, MgSO 4 , 10, CaCl 2 , 0.5, Na-ascorbate, 5, Na-pyruvate, 3 N-acetyl-L-cystein (NAC, myprotein.com, Manchester, United Kingdom), 3. Sections that were 350 µm thick were obtained with the use of a vibratome (HM 650V, Thermo Scientific, Waltham, MA, USA). During the slicing procedure, the tissue was submerged in ice-cold NMDG-aCSF solution that was bubbled with carbogen (95%/5% O 2 /CO 2 ). Next, the sections were incubated in carbogenated NMDG aCSF for 10 min (temperature:~33 • C). A 2M NaCl solution was gradually added to the solution according to [69]. The slices were then transferred to HEPES-based aCSF that contained, in mM: NaCl, 82, KCl, 2.5, HEPES, 20, thiourea, 2, glucose, 25, NaH 2 PO 4 , 1.25, NaHCO 3 , 30, MgSO 4 , 2, Na-ascorbate, 5, Na-pyruvate, 3, and 12 mM NAC (temperature:~25 • C). The recordings were started after 1 h of incubation in HEPES-aCSF. The pH of all of the solutions was between 7.3 and 7.4. During the loading procedure, the slices were perfused with aCSF solution at a rate of 4-5 mL/min, at 33 • C, to facilitate incorporation of the dye molecules by the cells. After the loading, the slices were allowed to rest for 20 min, under a constant perfusion of aCSF at a temperature of 25 • C. The consistency of the loading procedure was checked by acquiring fluorescent Fura-2 signal at its isosbestic point (λ = 360 nm). The intensity of the signal excited at this wavelength should not depend on Ca 2+ concentration and thus only reflect the quantity of the incorporated and de-esterified dye alone. The signal was comparable across preparations. Moreover, we did not detect any correlation between the loading efficiency and the magnitude of the recorded Ca 2+ responses.

Ca 2+ Imaging of Neurons in Brain Slices
The measurements were performed using MetaFluor software, a Zeiss Examiner Z.1 microscope (40×/1.0 objective), and the Retiga ELECTRO™ CCD Camera (Teledyne Qimaging, Surrey, BC, Canada). A peristaltic pump was used to perfuse the hippocampal slices with aCSF that contained 126 mM NaCl, 2.6 mM KCl, 20 mM glucose, 1.25 NaH 2 PO 4 , 25 mM NaHCO 3, and 1 mM MgSO 4 . Additionally, aCSF that contained chemicals used for evoking Ca 2+ responses was applied locally through a 100 µm wide tip that was positioned 100 µm above the pyramidal neurons of the CA1 hippocampal area. The pencil was connected to an automated perfusion system ValveLink 8.2 (Automate Scientific, Berkeley, CA, USA). For the first 10 min of the recordings (baseline recording, glutamate application, and signal relaxation), 2.5 mM CaCl 2 was included in the solution that was perfused with the use of the pump. Next, the solution was manually changed to Ca 2+ -free aCSF that contained 0.5 mM (ethylene glycol-bis(β-aminoethyl ether)-N,N,N ,N -tetraacetic acid) EGTA. The pH of aCSF was adjusted to 7.4 at room temperature with carbogen and HCl. The extracellular solution was carbogenated throughout the recordings and heated with an in-line heater (catalog no. 64-0102, controlled by a TC-324B temperature control unit, Warner Instruments) to maintain the temperature in the recording chamber close to 25 • C. The rate of solution flow was 4-5 mL/min.
To assess Ca 2+ homeostasis in the CA1 pyramidal neurons of the hippocampus, a Ca 2+ addback assay that was modified to allow for the quantification of the Ca 2+ efflux from the ER [33] was used. Changes in the fluorescent signal in the neuronal cell bodies were monitored and quantified. For baseline recording, aCSF that contained 2.5 mM Ca 2+ was perfused for 5 min, which was followed by the local application of 100 µM Na-glutamate for 10s. Next, the signal relaxation was monitored for another 5 min in 2.5 mM Ca 2+ aCSF. To deplete the extracellular milieu from the Ca 2+ ions, the solution was changed to aCSF that contained 0.5 mM EGTA for 15 min, which was followed by the local application of 20 µM cyclopiazonic acid (CPA) for 5 min. Next, Ca 2+ was re-added to the solution at a concentration of 5 mM Ca 2+ for 5 min, in the presence of 20 µM CPA. To limit activity-driven Ca 2+ entry, 1 µM Tetrodotoxin (TTX) and 10 µM nifedipine were included in all solutions that were used from the 10th min of the experiment onwards.
For Fura-2 quenching experiments, hippocampal slices were loaded with the dye as described above. For baseline recording, Ringer's solution (composed of, in mM: NaCl, 137, KCl, 5, glucose, 24, HEPES, 10, CaCl 2 , 2, MgCl 2 , 1, pH = 7.3-7.4) was perfused for 5 min, which was followed by application of Ca 2+ -free Ca 2+ depleting solution (Ringer's solution supplemented with 0.5 mM EGTA and 20 µM CPA) for 15 min. Next, Ca 2+ -free Ringer's solution supplemented with 2 mM Mn 2+ (as MnCl 2 ) was locally applied above the cells of interest, in the absence of EGTA and presence of 20 µM CPA. To limit activity-driven Ca 2+ entry, 1 µM TTX and 10 µM nifedipine were included in all solutions. To quantify the rate of Mn 2+ influx, linear part of the curves following Mn 2+ addition (which corresponded to the first 25 s of treatment) were approximated with a linear function and the decay slope was calculated (all calculations performed in Microsoft Excel).

Data Analysis
The acquired signals were quantified and exported with the use of MetaFluor software and further analyzed with Microsoft Excel and GraphPad Prism 5 (USA). Student's t-test was used to check the statistical significance of the observed differences between the groups.

Global Gene Expression Profiling
Total RNA was extracted from three individual hippocampi per variant using RNeasy Lipid Mini Kit (Qiagen: 74804, Hilden, Germany) and treated on-column with DNase using the RNase-Free DNase Set (Qiagen: 79254). Preparation of cDNA libraries and sequencing by Illumina HighSeq 1500 (run type: Paired-end sequencing, read length: 1 × 76 bp) were carried out in cooperation with Core Facility of IIMCB and the Nencki Institute. Sequencing reads (fastq) were filtered by Trimmomatic program, discarding all reads contaminated by sequencing adapter sequence as well as discarding reads whose mean quality in a bin of 20 base pair (bp) was below Q30. Minimal length of read used for mapping was set as 65 bp. Quality filtered reads were aligned to mouse genome (mm10) by STAR program. Statistical analyses, including comparisons to other samples, were done in R environment using NOIseq library [70]. Functional analysis was performed for the genes that were differentially expressed in transgenic versus wild-type samples (raw p-value below 0.05), separately for genes upand down-regulated in transgenic animals. The functional analysis of data was performed using clusterProfiler [71] to find statistically significant Gene Ontology terms. The whole metadata of RNAseq analysis deposit in NCBI's Gene Expression Omnibus [72] and are accessible through GEO Series accession number GSE138370 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE138370).

Real-Time Polymerase Chain Reaction Arrays and Quantitative Real-Time PCR (qPCR)
Hippocampi were dissected from three adult female mice (≥ postnatal day 180 (PD180)) per variant for real-time polymerase chain reaction arrays. The material was homogenized in Qiazol (Qiagen, catalog no. 79306, Germany) and subjected to total RNA isolation according to manufacturer's protocol of RNeasy Lipid Tissue Mini Kit (Qiagen: 74804, Germany). The quality of RNA was checked by absorbance measurements at 260, 280, and 230 nm. For the analysis, only samples whose A260/280 nm and A230/280 nm values exceeded 1.8 were used. There were 1000 ng of RNA templates used to synthesize first-strand cDNA, with the use of iScript reverse transcription supermix (Bio-Rad: 170-8841, USA). The 96-well custom plates that contained primers for target and reference genes (Bio-Rad, USA) and SsoAdvanced Universal SYBR Green Supermix (Bio-Rad:1725274; USA) were used to perform teal-time polymerase chain reaction (RT-PCR, carried out in duplicate). The cDNA (50 ng) was used for each reaction. Small Ubiquitin-like Modifier (SUMO)-activating enzyme subunit 2 (Uba-2), glyceraldehyde-3-phosphate dehydrogenase (Gapdh), and Actin (Actb) were used as reference genes. The reactions were performed using the CFX Connect™ Real-Time PCR Detection System (Bio-Rad). Bio-Rad CFX Maestro 1.1 software was used to analyze the data (Bio-Rad, USA). The reference genes' stability was acceptable (as evaluated by the reference gene selection tool). For calculation of the expression levels, equal efficiencies were assumed for all reactions and the 2 −∆∆Ct method was used. For each gene, the average normalized expression was calculated from the following formula: Normalized expressionGOI = RQGOI/(RQUba2 × RQGapdh × RQActb) 1 3 where: GOI, gene of interest; RQ, relative quantity, calculated as RQ = 2 (−CtGOI) .
The statistical significance of the observed differences in the expression levels between the tested groups was determined with the use of analysis of variance (ANOVA), followed by Tukey's Honestly Significant Difference (HSD) post-hoc test.
The real-time PCR reactions were performed in duplicate for each sample using SsoAdvance Universal SYBR Green Supermix (Bio-Rad) with the following sets of specific gene primers (the primer sequences are listed in Table 2, Sigma). Based on the real-time PCR arrays, SUMO-activating enzyme subunit 2 (Uba-2) was chosen as reference gene. The specificity of the reactions was determined based on dissociation curve analysis. Relative gene expression was calculated using the 2 −∆∆Ct method and Bio-Rad CFX Maestro 1.1 software (Bio-Rad, USA). Six adult female mice (≥ postnatal day 180 (PD180), per variant were taken to perform this experiment.

DdPCR Validation of Selected Genes
Primers for Arx, Cdkl5, Dclk1, and Uba2 were sourced from the EvaGreen ddPCR™ assays (Bio-Rad: dMmuEG5062557; dMmuEG5078166; dMmuEG5084027; dMmuEG5081512, respectively) and diluted according to the kit instructions. For ddPCR technology (QX200 Droplet Digital PCR ddPCR™) System-Bio-Rad, USA) 21 µL reaction mixtures containing 25 ng of cDNA (six transgenics mice vs. 3 wild-type mice variant), primers, and QX200™ddPCR™ EvaGreen Supermix (Bio-Rad: 186-4034) were used. Droplet generation and transfer of emulsified samples to PCR plates (Bio-Rad: 12001925) were performed according to the manufacturer's instructions (Instruction Manual, QX200™ Droplet Generator-Bio-Rad). The ddPCR plate was sealed with a foil heat seal (Bio-Rad, 181-4040) and the PX1™ PCR Plate Sealer (Bio-Rad; 181-4000). The following cycling protocol was established according to the manufacturer's protocol: 95 • C enzyme activation step for 5 min followed by 40 cycles of a two-step cycling protocol (95 • C for 30 s and 58 • C for 1 min). The ramp rate between these steps was slowed to 2 • C/second. The post-cycling protocol was in accordance with the kit instructions (Bio-Rad, 186-4034). The absolute quantity of DNA per sample (copies/µL) was processed using QuantaSoft version 1.7.4.0917 (Bio-Rad, USA). The data were next exported to Microsoft Excel for further analysis.

Conflicts of Interest:
The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.