Reversing GABA polarity corrects synaptic physiology and behavioural decits in young adolescent Syngap1+/- mice

: 11 Haploinsufficiency in SYNGAP1 is implicated in Intellectual Disability (ID) and Autism Spectrum disorder (ASD) 12 and affects the maturation of dendritic spines. The abnormal spine development has been suggested to cause disbalance 13 of excitatory and inhibitory (E/I) neurotransmission at distinct developmental periods. In addition, E/I imbalances in 14 Syngap1 +/- mice might be due to abnormalities in K + -Cl - co-transporter function (NKCC1, KCC2), in a similar manner 15 as in the murine models of Fragile-X and Rett syndromes. To study whether an altered intracellular chloride ion 16 concentration represents an underlying mechanism of altered function of GABAergic synapses in Dentate Gyrus 17 Granule Cells of Syngap1 +/- recordings were performed at different developmental stages of the mice. We observed 18 that neurons at P14-15 of Syngap1 +/- mice had depolarised membrane potential and a decreased Cl - reversal potential. 19 The KCC2 expression was decreased compared to Wild-type (WT) mice at P14-15. concentration that can be counteracted by the small molecule 6BIO. The 6BIO sufficiently restored cognitive, 26 emotional, and social symptoms by pharmacological intervention, particularly, in adulthood.

4 Syngap1 +/mice display improper glutamatergic synapse development and function; however, the trophic role of 75 GABA in shaping synaptic function during development is unknown (Clement et al. 2012;Ozkan et al. 2014;Clement 76 et al. 2013;Jeyabalan and Clement 2016;Kepecs and Fishell 2014;Ben-Ari 2002). Studies have shown that GABA 77 exhibits trophic function -excitatory and inhibitory in early and late stages of development, respectively -regulated 78 by potassium-chloride (Cl -) co-transporters like KCC2 and NKCC1. The NKCC1 expression level is age-dependently 79 regulated and highest in the early stages of the development that allows influx of Cl -. In contrast, the KCC2 expression 80 is high in the later stages of development, thereby, extrudes intracellular Clfrom the neuron. Hence, the intracellular  Rivera et al. 1999). Alteration in the intracellular basal chloride concentration modulates the electrochemical gradient 84 (EGABA) of the chloride ion flow through the chloride conductive GABAA receptor/channels and thus affecting the 85 strength of inhibitory action in a neuronal network. Thus, a switch between excitatory or inhibitory action of GABAA 86 receptor/channels is mediated by the expression level of potassium-chloride-co-transporters. The disruption of the co-87 transporter expression during the critical period impairs synapse formation and function as it has been reported in 88 several ID/ASD-like animal models (He et al. 2014;Hyde et al. 2011;Duarte et al. 2013;Deidda, Allegra, et al. 2015).

89
Besides, reversing the polarity of GABAA mediated chloride flow before the end of the critical period of mouse 90 development by bumetanide corrected EGABA, synaptic function, and phenotypes in Down's, Fragile -X and Rett 91 syndrome-like animal models (He et al. 2019;Deidda, Parrini, et al. 2015;Banerjee et al. 2016). Thus, normalization 92 of the reversal potential for chloride (here called GABA polarity) could represent a putative target for therapeutic 93 intervention in ID/ASD.

94
To restore reversal potentials of chloride GSK-3β (Glycogen synthase kinase-3, multifunctional serine/threonine 95 kinase) has been suggested, which is pivotal in synaptic plasticity (Yao et al. 2002;Salcedo-Tello, Ortiz-Matamoros, 96 and Arias 2011). Increased activity of GSK-3β is a major contributor to the ID/ASD pathophysiology and inhibiting 97 GSK-3β reversed many of their phenotypes (Min et al. 2009;Mines and Jope 2011). Nevertheless, finding potent 98 therapeutics that can target GSK-3β and restore phenotypes, when administered after a critical period of development, 99 has been elusive. Therefore, the opportunity exists to develop a therapeutically viable GSK-3β inhibitor that can 100 restore the phenotypes. We report a GSK-3β inhibitor, 6-bromoindirubin-3`-oxime (6BIO) that has been described to 101 5 cross the blood-brain barrier and to be neuroprotective in an MPTP-based model of Parkinson's disease (Meijer et al. 7 (Merck Millipore, #100983) and centrifuged at 4ºC at 10,000 rpm for 10-minute. The washing step was repeated twice, 154 and the pellet was air-dried. Pellet was resuspended in 20 µl of nuclease-free water (Thermo Fisher Scientific, 155 #AM9932), and RNA concentration was assessed in nanodrop. PCR was performed for cDNA synthesis by using a 156 cDNA synthesis kit (Takara, #TP600) and aliquoted and stored at -80ºC till further use. qPCR reaction was set up in 157 duplicates using SYBR Mix (Roche, #light cycler 480), and probes for the genes mentioned above. Cq values were 158 calculated by the ddCt method, in which dCt was calculated by subtracting test gene value from housekeeping gene 159 value and then taking 2^-dCt as Cq value. The relative mRNA level, 2-ddCt, was plotted.

176
Sections were mounted in Vectashield (Vector, #H-1000) on slides. All the imaging was performed in a confocal laser 177 scanning microscope (LSM 880, Zeiss, India) and analysed using ImageJ software. Dentate gyrus granular cells 178 (DGGC) were chosen for imaging and all the images were acquired with a confocal-fluorescence microscope equipped 179 with a 63X objective. A Z-stacks of 30 images were acquired and images #10 to #24 were considered for analysis.

8
The maximum intensity projection (MIP) image over 15 confocal images of a Z-stack was calculated and analysed.

181
Image thresholds were kept constant to quantify the expression of KCC2 for all genotypes and age groups. The 182 integrated density of the isolated areas was calculated to measure the mean intensity (Madhwal et al. 2020; 183 Mukherjee, and Banerjee 2012). In representative images (MIP images), punctate structures indicate expression of 184 KCC2 (red colour), and SYNGAP1 (green colour) on the YFP background (yellow colour). Thy1-Yfp-Syngap1 +/mice 185 were studied to identify the labelling of excitatory neurons. The whole area of the image represented along with YFP 186 expressing neurons was considered for the analysis. The analysis of genotypes was performed in parallel, and the 187 statistical comparison between groups was evaluated with the Unpaired Student's t-test.

239
In the small molecule study, the mice in the drug group were administered intraperitoneally with 5 mg/kg of 6BIO.

240
All four experimental groups of mice of three different age groups ( Figure

251
The NOR was performed in a custom-made wooden box 35×35×35 cm in JNCASR. On Day-1, mice were familiarised 252 with the arena without the objects for 5-minute. Day-2 and Day-3 sessions included habituation with two familiar 253 objects placed diagonally to each other, and mice were allowed to explore the objects for 10 minutes. The day-4 254 session involved the recognition memory test 24 hours after the last training. The time the mice explored the novel 255 object and a familiar object placed diagonally was taken within 10 minutes (Lueptow 2017;Leger et al. 2013). The 256 time of interaction of the mice (sniffing/touch) to familiar and novel objects was acquired manually using stopwatches.

257
The discrimination index (DI) for novel object recognition test (NOR) [(time spent with novel object -time spent with 258 familiar object)/(time spent with novel object + time spent with familiar object)] X 100, was calculated and taken as 259 a measure of the strength of object recognition (either novel or familiar). DI values above 25 correlated with memory 260 11 and below 25 correlated with a lack of memory. In addition, the mice behaviour was monitored and protocolled by a 261 Handycam (SONY #HDR-CX405).

263
SI and SP were performed in a custom made transparent three-chamber plexiglass box 40.6×21×35 cm in JNCASR.

264
In the first session of the experiment, the mouse was habituated to the three-chambered glass box for 5-minute. In the 265 second session, the chambers were closed with doors, and the mouse was placed in the middle chamber and allowed 266 to explore for 5-minute. In the third session, two steel mesh jars were kept in either of the adjacent chambers -one 267 had a mouse of the same gender and age and the other was empty. The test mouse was introduced into the middle 268 chamber, the doors were removed, and the mouse explored the whole arena for 10-minute (Nakajima et al. 2019).

269
Social interaction was calculated by the time spent between the empty jar and familiar mice. In the last session, the 270 mouse of the same gender and age was introduced into the empty mesh jar, and the test mouse was allowed to explore 271 and interact with both the mice placed beneath the mesh jars. Social preference was calculated by the time spent        Deidda, Parrini, et al. 2015). Thus, we studied whether altered expression of NKCC1 and 313 KCC2 in Syngap1 +/correlates with the observed functional differences regarding GABA-mediated actions. A 314 significant difference in NKCC1 expression between Wild Type (WT) and Syngap1 +/at P8 was detectable using 315 protein biochemical analysis (Figure 2A, and 2B). Notably, a compensatory increase in KCC2 expression was only 316 detectable at P8 and not at P14 in Syngap1 +/- (Figure 2C and 2D). At P14 Syngap1 +/-, the KCC2 expression level was 317 significantly low as compared with WT mice -a similar result as our immunofluorescence analyses ( Figure 2E

326
The results thus far demonstrated that a mutation at Syngap1 caused a reduced neuronal network activity and inhibition 327 due to disabling the switch of GABA polarity during the development of transgenic mice. GABAergic dysfunction is 328 one of the hallmarks of ID/ASD and, thus, a vital candidate target to correct pathophysiological phenotypes. GSK3-β 329 is a well-known regulator of synaptic plasticity (Peineau et al. 2007;Liu et al. 2017). Small molecules such as 6BIO 330 is a potent negative regulator of GSK3-β and by crossing the blood-brain barrier it has been shown to be 331 neuroprotective in an MPTP-based mouse model of Parkinson's disease. Accordingly, we sought whether 332 administration of 6BIO might restore synaptic plasticity, EGABA, and behavioural dysfunction in Syngap1 +/mice. To 333 this end, we studied the effects of 6BIO on mutant mice at three different age groups: Group I: P10-16 (critical period),

334
Group II: P10-80 (critical period till adulthood), and Group III: P30-80 (adulthood; Figure 3A). Clement et al, 2012 335 (Clement et al. 2012) had demonstrated that the SYNGAP1 expression peaks at 14-16-day old mice where synaptic 336 function and dendritic spine morphology was reported at 14-16-day old Syngap1 +/mice. Further, in this study, they 337 have demonstrated that inducing heterozygous knock-out of Syngap1 in adults did not impact synaptic function 338 contrary to knock-out of Syngap1 in P0. Correspondingly, our data shows that KCC2 expression and function (EGABA) 14 is mainly disrupted at P14-16. Thus, we chose this period as a critical period of development as most of our study 340 involved the hippocampus (critical period ends at 3-week) (Chakraborty, Vijay Kumar, and Clement 2021).

352
Remarkably, we found that 6BIO substantially restored memory performance in Syngap1 +/mice.

353
Similar to NOR, a social novelty test (Social Interaction (SI) and Social preference (SP)) was performed for all age 354 groups at P80 (P10-16: Figure 4D and 4G; P10-80: Figure 4E and 4H; P30-80: 4F and 4I). Our results suggested that 355 the time spent with stranger-1 as compared to the empty jar for 6BIO treated Syngap1 +/mice were comparable to WT 356 levels in all age groups. However, we did not observe rescue in social preference for 6BIO treated Syngap1 +/mice in 357 the P10-16 group. Overall, our results suggest that 6BIO is effective in correcting the SI and SP deficits when 358 administered after a critical period of development in Syngap1 +/mice.

362
We observed an increased seizure threshold for the tonic-clonic seizure that matched WT levels in 6BIO administered 363 Syngap1 +/mice at P30-80 ( Figure 5A). We have summarised all the behaviour results in Table 2. The results indicate 364 that rescuing GABAergic-synaptic function is fundamental in correcting behaviour in Syngap1 +/mice as LTP is 365 15 disrupted in these adult mice (Ozkan et al. 2014). Considering the impact of Syngap1 +/mutation as reported in other 366 brain regions (Clement et al. 2012;Clement et al. 2013;Aceti et al. 2015;Ozkan et al. 2014), we performed EGABA 367 from DGGC and LTP from the Schaffer-collateral commissural pathway to validate the effect of 6-BIO in different 368 regions of the hippocampus.

372
suggest small molecules such as 6BIO interventions, particularly after the critical period of development, can correct 373 functional and behavioural deficits in Syngap1 +/mice, indicating an opportunity for novel therapeutics to treat 374 neurodevelopmental disorders.

376
The interplay between GABAergic and Glutamatergic systems is indispensable for the formation, maturation, 377 refinement, and maintenance of synapses in the developing brain. Aberration in any of these processes results in 378 neurodevelopmental and psychiatric disorders due to persistent changes in neuronal activity during development (Ben-379 Ari et al. 1989;Cherubini, Gaiarsa, and Ben-Ari 1991;Owens et al. 1996;Rivera et al. 1999). Based on our results, 380 we consider abnormal activation of neurons to occur upon GABA activation during developmental stages due to 381 increased NKCC1 levels at P8 and a concomitant increase in KCC2 levels at P8 in Syngap1 +/-, contributing to the 382 increased intrinsic excitability (Clement et al. 2012;Kaila et al. 2014), similar to Fragile-X and Rett syndrome (Hinz,

410
Previous studies have demonstrated that the synaptic function and dendritic spine morphology deficits in Syngap1 +/-411 mice were observed either over a single day or in a 2-3 day period within the critical period of development which is 412 comparable with our observation: hippocampus -critical period ends at 3 weeks; somatosensory -7-9 days; mPFC -413 7-8 days; layer 5 -30 days (Clement et al. 2012;Clement et al. 2013;Ozkan et al. 2014;Aceti et al. 2015). These 414 earlier studies strongly imply the stringent regulation of synaptic function and spine morphology during the critical

434
In various studies, targeting the GABAergic circuit has improved synaptic and behavioural deficits observed in the 435 murine models of neurodevelopmental disorders (Banerjee et al. 2016;Deidda, Parrini, et al. 2015;He et al. 2019;436 Braat and Kooy 2015;Cellot and Cherubini 2014;Savardi et al. 2020;D'Hulst and Kooy 2007;Raveendran, Pressey, 437 and Woodin 2020) and, thus, we speculated that restoring GABAergic function to WT levels can correct behavioural 438 deficits in Syngap1 +/mice. Our initial findings described the altered function of Chloride co-transporters (reversal 439 potential of GABA) at P14-15 was restored when 6BIO was administered in Syngap1 +/mice. In support of this, a 440 recent study has shown that the application of BIO compound in a Rett syndrome model restored EGABA, thus, restoring 441 behavioural function (Tang et al. 2019). This also led us to hypothesise that rectifying GABA function can correct 442 behavioural deficits in Syngap1 +/mice.

443
It is essential to understand whether the drug can correct behavioural deficits when administered in young adolescent 444 stages, apart from when given during the critical period of development (P10-16). Based on the paradigm in Figure  18 3A, P10-16 will address whether targeting the critical period alone would be sufficient to correct the phenotype with 446 lasting effects in adult stages. P10-80 will answer whether the drug should be administered from the critical period of 447 development and continue throughout adulthood. P30-80, the most crucial aspect of the drug discovery, would imply 448 whether administration of the drug after a critical period of development corrects synaptic dysfunction and behavioural 449 deficits, which remains a challenge to date. Most compounds can restore these functions when administered during 450 the critical period of development (i.e., P10-16), which makes our study unique and different from the other published 451 results.

452
We have recently shown that 6-bromoindirubin-3`-oxime (6BIO) is neuroprotective in an MPTP-based mouse model 453 of Parkinson's disease, suggesting the potency to cross the blood-brain barrier and inhibit GSK-3β in neurons (Suresh 454 et al. 2017). 6BIO administration in P30-80 (group III) has substantially restored functional (LTP, EGABA (P16)) and 455 behavioural deficits (sociability, memory, and anxiety) in Syngap1 +/mice. However, hyperactivity was not corrected 456 after the neurodevelopmental period suggesting that the target period is during the critical period of development. It 457 appears that social isolation, social preference, and seizure threshold were not impacted by the hard wiring of the 458 neuronal circuit during development as it is corrected in all three strategies, except in group I for SP. Primarily, we 459 demonstrate that administration of 6BIO corrected memory deficits in all therapeutic strategies, particularly after the 460 neurodevelopmental period.

461
Studies have demonstrated that the deficits observed in the Dentate gyrus (DG) in Syngap1 +/is recapitulated in the 462 CA1 region of the hippocampus (Clement et al. 2012). It is shown that LTP was reduced in the CA3-CA1 region while 463 the rest of the study was from the forebrain region of the brain (Ozkan et al. 2014). As demonstrated by previous 464 publications that the effect observed in DG is recapitulated in other parts of the hippocampus and also in other parts 465 of the brain such as the Barrel cortex and mPFC (Clement et al. 2013). Thus, these data demonstrate that the impact 466 of Syngap1 +/global and that prompted us to investigate the effect of 6BIO in the DGGC and CA1 region of the 467 hippocampus. We believe that rectifying the EGABA may correct synaptic function and plasticity as optimal GABA 468 function is needed for suitable excitatory synaptic function leading to plasticity (Ormond and Woodin 2011;Larson   19 Hooper et al. 2007;McCamphill et al. 2020), similar to what we observed in Syngap1 +/model. However, these studies 473 have not shown the level of rescue we observed in Syngap1 +/mice when 6BIO was administered after the critical 474 period of development (i.e., adolescent young adults). Despite this, there is a lack of clarity on the function of GSK3ß 475 in modulating synaptic function. The reason, we believe, is that the pharmacological approaches and shRNA 476 experiments (Liu et al. 2017) could be more acute compared to the conditional deletion, while the latter may induce 477 the risk of measuring secondary effects of GSK-3β deletion. We agree that 6BIO may have other targets that can 478 restore the synaptic function and behaviour deficits to WT levels which need to be explored but not in the scope of 479 this study. However, these experiments are not in the scope of the current study.
Data are presented as single data points and means ± SEM. N: number of mice, n: number of cells. Two way ANOVA, Tukey's multiple comparisons test.